A R T I C L E
MOZ-TIF2, but not BCR-ABL, confers properties of leukemic
stem cells to committed murine hematopoietic progenitors
Brian J.P. Huntly,1,* Hirokazu Shigematsu,2Kenji Deguchi,1Benjamin H. Lee,1,3Shinichi Mizuno,2
Nicky Duclos,1Rebecca Rowan,1Sonia Amaral,1David Curley,1Ifor R. Williams,4Koichi Akashi,2
and D. Gary Gilliland1,5,*
1Division of Hematology, Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts 02115
2Department of Cancer Immunology and AIDS, Dana Farber Cancer Center, Boston, Massachusetts 02115
3Department of Pathology, Brigham and Women’s Hospital, Boston, Massachusetts 02115
4Department of Pathology, Emory University, Atlanta, Georgia 30322
5Howard Hughes Medical Institute, Harvard Medical School, Boston, Massachusetts 02115
*Correspondence: firstname.lastname@example.org, email@example.com
To better understand the origin of leukemic stem cells, we tested the hypothesis that all leukemia oncogenes could
transform committed myeloid progenitor cells lacking the capacity for self-renewal, as has recently been reported for MLL-
the oncogenes MOZ-TIF2 and BCR-ABL, respectively. MOZ-TIF2-transduced progenitors could be serially replated in
methylcellulose cultures and continuously propagated in liquid culture, and resulted in an acute myeloid leukemia in vivo
that could be serially transplanted. In contrast, BCR-ABL transduction conferred none of these properties to hematopoietic
progenitors. These data demonstrate that some, but not all, leukemia oncogenes can confer properties of leukemic stem
cells to hematopoietic progenitors destined to undergo apoptotic cell death.
adult life, undergo an increased number of cell divisions, and
therefore may have a greater opportunity than more short-lived
cells to acquire the minimum number of mutations required for
malignant transformation (Bradford et al., 1997; Chesier et al.,
1999; Lansdorp, 1997). Phenotypic evidence also implicates
somatic stem cells as the target cells for transformation. Both
stem cells and tumor cells express similar levels of telomerase
(Morrison et al., 1996) required to maintain telomere length and
prevent replicative senescence. In human AML, the distinctive
leukemia “stem cell” that is sufficient to generate AML in the
NOD-SCID mouse model (the SCID leukemia initiating cell, SL-
IC) has a similar phenotype (CD 34?/CD 38?) to that of normal
SCID-repopulating cells (Lapidot et al., 1994) and to the cell
compartment enriched for true human hematopoietic stem cells
(HSC) (Manz et al., 2002). Furthermore, recurrent cytogenetic
in this cell compartment in patients with AML, chronic myeloid
leukemia (CML), and ALL (Deininger et al., 2000; Haase et al.,
1995; Mehrotra et al., 1995; Quijano et al., 1997). Indeed, the
Philadelphia chromosome, the cytogenetic hallmark of CML,
can be detected in cells of the myeloid, erythroid, megakaryo-
cytic, and B-lymphoid lineages, suggesting transformation of a
Limitless self-renewal potential is one of the hallmarks of all
cancers (Hanahan and Weinberg, 2000). It had been thought
that all cells of a tumor contained such potential. However,
recent evidence from human malignancies, including acute my-
eloid leukemia (AML) (Bonnet and Dick, 1997), acute lympho-
blastic leukemia (ALL) (Cobaleda et al., 2000), and breast (Al-
Hajj et al., 2003) and CNS (Hemmati et al., 2003; Ignatova et
al., 2002; Singh et al., 2003) cancer has demonstrated the exis-
tence of subpopulations of cells, delineated by differential ex-
pression of surface markers, that exclusively contain the ability
to recapitulate the human disease upon transplantation into
recipient NOD-SCID (non-obese diabetic/severe combined im-
munodeficient) mice. The tumors that develop are phenotypi-
cally very similar to the original human tumor, and serial trans-
plantation of the tumor to secondary recipients confirms the
true stem cell nature of these subpopulations of malignant cells.
It has been suggested that the target cells for malignant
transformation are adult somatic stem cells (Reya et al., 2001),
and that oncogenic mutations appropriate the capacity for self-
renewal inherent in these cells. Stem cells persist throughout
S I G N I F I C A N C E
The existence of cancer stem cells has been demonstrated in leukemias and solid-organ cancers. It has been suggested that they
result from transformation of adult somatic stem cells with inherent self-renewal capacity. We report that some leukemia-associated
oncogenes, but not others, can confer properties of leukemic stem cells to committed murine myeloid progenitors, and exclude
retroviral insertional events as the sole explanation for these findings. These data provide tools for the identification of cellular
programs that confer self-renewal properties to somatic cells previously destined to undergo terminal differentiation and apoptotic
cell death. Characterization of these programs may identify new targets for therapeutic intervention in leukemia, and genes that
confer properties of self-renewal to adult somatic cells for regenerative therapeutic purposes.
CANCER CELL : DECEMBER 2004 · VOL. 6 · COPYRIGHT © 2004 CELL PRESS 587
A R T I C L E
cell with multilineage differentiation potential such as an HSC
(Deininger et al., 2000; Martin et al., 1980).
poietic progenitors without self-renewal capacity may also be
potential targets for transformation, and that this could explain
phenotypic differences between leukemias associated with the
same characteristic molecular abnormality (Fialkow et al., 1981;
Griffin and Lowenberg, 1986). However, committed progenitors
have engaged programs for terminal differentiation and apo-
ptotic cell death, and do not have the properties of leukemic
stemcells. Ithasbeen recentlyreportedthat retroviraltransduc-
ted myeloid progenitors and demonstrates induction of AML
following their transplantation into irradiated mice (Cozzio et al.,
2003). However, although progenitorstransduced with transfor-
mation-disabled mutants of MLL-ENL did not engender this
phenotype, these findings did not address whether this is a
capability conferred by all leukemia-associated oncogenes, ei-
ther alone or in concert with possible retroviral integration ef-
In this study, we tested the hypothesis that all leukemia-
associated oncogenes have the capacity to confer self-renewal
properties of leukemic stem cells to hematopoietic progenitors
that inherently lack the ability to self-renew. We used multipa-
rameter flow cytometry to isolate purified populations of com-
mitted myeloid progenitors that lack self-renewal capacity
(Akashi et al., 1999; Na Nakorn et al., 2002). These populations
were then transduced with either the MOZ-TIF2 (Carapeti et al.,
1988; Deguchi et al., 2003) or BCR-ABL oncogenes (Deininger
et al., 2000; Sattler and Griffin, 2003; Wong and Witte, 2001),
as representative members of the transcription factor/cofactor
or tyrosine kinase fusion oncogene families, respectively, and
their self-renewal and transformation properties were analyzed
with in vitro and in vivo systems and compared with similarly
transduced HSC and whole bone marrow. The contributions of
the initial cell of transduction to the eventual phenotype of the
disease were also assessed utilizing the same experimental
Figure 1. Schema of in vitro and in vivo experiments used to test the ability
of MOZ-TIF2 and BCR-ABL to confer properties of self-renewal to committed
In brief, flow-sorted CMP and GMP from Ly5.2 mice were transduced with
retroviral vectors expressing the MOZ-TIF2 or BCR-ABL oncogenes along with
either the neomycin resistance gene or GFP. For the in vitro studies, serial
replating assays were performed following the initial selection of neomycin-
resistant transduced cells in the first methylcellulose plate. Colonies which
formed in the 3rdor 4thplates weresubsequently tested for the ability to grow
in IL-3-supplemented liquid culture, and these cell lines were subsequently
transplanted into irradiated mice. In the in vivo studies, CMP and GMP
transduced with vectors expressing MOZ-TIF2 or BCR-ABL and GFP were
directly injected into lethally irradiated congenic Ly5.1 mice and assessed
for the development of leukemia. Where leukemia developed in primary
recipients, the long-term self-renewal of the initially transduced progenitors
was further tested in serial transplant of leukemic cells to sublethally irradi-
ated secondary recipients.
MOZ-TIF2, but not BCR-ABL, confers properties
of self-renewal to committed myeloid
progenitors in vitro
(Lavau et al., 1997). Transduction of whole bone marrow with
certainleukemia oncogenesconfers theability toserially replate
in vitro, although the target population(s) of cells that acquire
the ability to serially replate is unknown (Lavau et al., 1997).
The ability of CMP and GMP transduced with MOZ-TIF2 or
BCR-ABL to serially replate was tested as shown in the schema
in Figure 1. CMP or GMP were transduced with MSCV-pgk-Neo
cells in the first plating. In control experiments, untransduced
CMP and GMP did not serially replate (Figure 2A). We also
observed that CMP or GMP transduced with BCR-ABL did not
serially replate, indicating that BCR-ABL lacks the ability to
alter the programs for terminal differentiation and apoptosis in
committed myeloid progenitors. However, in striking contrast,
CMP or GMP transduced with MOZ-TIF2 did serially replate
(Figure 2A). After 4 replatings, the MOZ-TIF2-transduced colo-
nies were small and tightly packed, with morphology consistent
with small CFU-blast (Figure 2C). Cytospins of single colonies
monocytic morphology and hemophagocytosis, features evi-
(Borrow et al., 1996; Carapeti et al., 1988) (Figure 2D), as well
asinmurinemodelsofleukemiainducedby MOZ-TIF2 (Deguchi
et al., 2003).
Mutational analysis of MOZ-TIF2 demonstrates
the same domain requirement for self-renewal
as for leukemic transformation
Leukemia induction by MOZ-TIF2 has been shown to minimally
require the nucleosome binding motif of MOZ (abrogated in
the C543G mutant) and the CBP interaction domain of TIF2
(abrogated by the LXXLL mutant), but histone acetyltransferase
mechanism of transformation by MOZ-TIF2, therefore, involves
CANCER CELL : DECEMBER 2004
A R T I C L E
Figure 2. Invitro analysisdemonstrates thatMOZ-
TIF2, but not BCR-ABL, can confer properties of
self-renewal to committed myeloid progenitors
A: Bar chart of serial replating experiments. In
contrast to untransduced progenitors (normal
HSC, NHSC, NCMP, and NGMP) and progenitors
transduced with BCR-ABL, CMP and GMP trans-
duced with MOZ-TIF2 demonstrated replating
potential to the 4thplating.
B: Serial replating experiments with MOZ-TIF2 mu-
tants, demonstrating that the nucleosomal bind-
ing site (mutated in C543G mutant) and CBP
interaction domain (mutated in LXXLL mutant)
necessary for transformation are also required to
confer serial replating ability. Results presented
are the mean (? SD) of 2 (untransduced HSC,
CMP, GMP, and MOZ-TIF2 mutants) or 3 experi-
ments (BCR-ABL- and MOZ-TIF2-transduced CMP
and GMP) performed in duplicate. The MCID
mutant contains both the nucleosomal binding
site and CBP interaction domain.
round platings for BCR-ABL- and MOZ-TIF-trans-
duced GMP. While no colonies were seen on the
BCR-ABL plate, numerous small tightly clustered
CFU-blast could be seen in the MOZ-TIF2 plate.
D: Morphology of cells derived from colonies of
the 4thround of plating for MOZ-TIF2-transduced
CMP and GMP. Primitive myelomonocytic cells
were demonstrated with prominent hemopha-
gocytosis (right panels), features of human MOZ-
E: Line graph demonstrating the sustained
duced CMP and GMP in IL-3-supplemented liq-
aberrant recruitment of CBP to MOZ nucleosome binding sites
(Deguchi et al., 2003). To test the requirements for self-renewal
in committed progenitors, CMP and GMP were transduced with
either the C543G and LXXLL mutants, or the MOZ-CID (MOZ-
CBP interacting domain) mutant containing the minimal domain
requirement for transformation, and were serially replated in
methylcellulose as above. The results (Figure 2B) show serial
replating only for the MOZ-CID, demonstrating that the domain
requirements for self-renewal are the same as for leukemic
transformation, and link self-renewal to the aberrant transcrip-
tional program initiated by MOZ-TIF2 that contributes to leuke-
transduction of whole bone marrow (Figure 5A, and data not
shown). Cell lines derived from MOZ-TIF2-transduced progeni-
tors were also capable of generating AML, with 1 out of 2 GMP
cell lines and 2 of 3 CMP cell lines resulting in AML when
transplanted into lethally irradiated recipient mice (latency 103–
162 days) (data not shown).
Taken together, these data indicate that MOZ-TIF2, but not
BCR-ABL, confers the ability to serially replate in vitro, one
surrogate for self-renewal that is not present in normal commit-
ted hematopoietic progenitors, nor indeed in normal HSC (Fig-
ure 2A). Importantly, the capacity of MOZ-TIF2 to confer the
ability to serially replate cannot be attributed to retroviral inser-
tional mutagenesis. Retroviral transduction with point mutants
of MOZ-TIF2 that are not leukemogenic, or with an unrelated
but fully active leukemia oncogene, BCR-ABL, does not confer
MOZ-TIF2-transduced committed progenitors
can be propagated in liquid culture,
and these cell lines can induce AML in mice
MOZ-TIF2-transduced CMP and GMP derived from the third
and fourth rounds of replating were amalgamated and could be
subsequently propagated in liquid culture containing IL-3 (Fig-
ure 2E), demonstrating sustained growth for at least 8 weeks.
Morphological analysis of these cells at subsequent time points
not shown). Flow cytometric analysis of the cell lines demon-
strated a similar surface immunophenotype whether they were
derived from either CMP or GMP, which was similar to the
phenotype observed in MOZ-TIF2 leukemic cells derived from
MOZ-TIF2-, but not BCR-ABL-, transduced committed
myeloid progenitors can generate AML in mice
tors in vivo. CMP, GMP, or whole bone marrow mononuclear
cells (BM MNC) derived from Ly5.2 mice were transduced with
MOZ-TIF2 or BCR-ABL and directly injected into lethally irradi-
ated Ly5.1 congenic recipients according to the schema in Fig-
ure 1. The control experiments with unfractionated BM MNC
showed that mice transplanted with BM MNC transduced with
CANCER CELL : DECEMBER 2004589
A R T I C L E
Figure 3. MOZ-TIF2, but not BCR-ABL, generates
leukemia in transduced committed myeloid pro-
A: Mice transplanted with BCR-ABL-transduced
BM MNC succumb to a fatal myeloproliferative
disease (MPD) within 30 days of transplantation,
while mice transplanted with BCR-ABL-trans-
duced CMP and GMP do not develop an MPD
over a prolonged time period (p ? 0.0001). In
contrast, mice transplanted with MOZ-TIF2-trans-
duced BM MNC, CMP, and GMP develop AML,
with no significant difference in latency or pene-
trance (p ? 0.75)
B: Histology of representative mice transplanted
with BCR-ABL- (left panel) and MOZ-TIF2- (right
panel) transduced GMP. No histological abnor-
mice. In contrast, MOZ-TIF2 transplanted mice
demonstrated the presence of peripheral blood
(PB) and bone marrow (BM) blasts and extensive
tissue infiltration of organs including the liver,
trasting histology was demonstrated in mice
transplanted with transduced CMP.
C: Flow cytometric analysis of spleen and bone
marrow single cell suspensions from mice trans-
planted with BCR-ABL and MOZ-TIF2-transduced
GMP. Relative fluorescence for the myeloid
marker Mac-1 and the congenic marker Ly 5.2
are shown on a logarithmic scale. Quadrant
markers segregate positive and negative cells
Mac-1-positive Ly5.2 (CD 45.2) leukemic cells
(upper right quadrant) are present in the bone
marrow and spleen of mice transplanted with
MOZ-TIF2- but not BCR-ABL-transduced cells.
D: Southern analysis of proviral integration in
mice transplanted with BCR-ABL- and MOZ-TIF2-
transduced progenitors. Oligoclonal integration
is demonstrated for MOZ-TIF2-transplanted mice,
but no integration is seen with BCR-ABL mice.
either MOZ-TIF2 or BCR-ABL developed AML or a myeloprolif-
previously described (Deguchi et al., 2003; Wong and Witte,
2001) (Figure 3A). In consonance with observations in the serial
replating assays, mice transplanted with MOZ-TIF2-transduced
CMP and GMP developed AML, with a similar long latency to
mice transplanted with BM MNC (Figure 3A). In contrast, none
of the mice transplanted with BCR-ABL-transduced CMP or
GMP showed evidence of disease upon sacrifice at between
113 and 240 days, a time point at which all mice transplanted
with BCR-ABL-transduced BM MNC had been sacrificed for
development of myeloproliferative disease (Figure 3A).
AML (Figures 3B [right panel], 4A, 4B, and 5A). These features
included leukemic infiltration of the spleen, submandibular
gland, and submental lymph nodes, liver, lung, and bone mar-
row. The presence of leukemic blasts was seen in the peripheral
blood, and occasional mice demonstrated granulocytic sarco-
mata. Analysis of the leukemic cells from these mice showed
expression of GFP and Ly 5.2 by FACS analysis and oligoclonal
retroviral integration by Southern blotting (Figures 3C and 3D).
FACS analysis indicates that the hematopoietic
compartment with properties of leukemic stem
cells in MOZ-TIF2-associated leukemias
lies downstream of GMP
FACS analysis, gated on the leukemic cell populations by GFP
orwhole bonemarrow hadvery similarpatterns ofexpression of
surface markers, although the proportion of cells expressing
some markers varied. The surface phenotype of the leukemic
negative, and expressed Mac-1 with varying degrees of Gr-1
and c-kit positivity (Figure 5A). There was uniform reduction of
myeloid progenitor groups within this leukemia population, with
MOZ-TIF2-associated leukemias are virtually
indistinguishable regardless of the initial
cell population transduced
MOZ-TIF2-associated leukemias arising from transduced com-
mitted progenitors populations (CMP and GMP) had a similar
tissue distribution, histology, and immunophenotype to those
arising from unfractionated bone marrow, and recapitulated
many of the phenotypic characteristics of human MOZ-TIF2
CANCER CELL : DECEMBER 2004
A R T I C L E
Figure 4. MOZ-TIF2 leukemias are almost identical regardless of the initial cell population transduced
A: Tissue distribution of disease is similar for mice transplanted with MOZ-TIF2-transduced BM MNC, CMP, and GMP. Leukemic mice invariably demonstrated
splenomegaly and neck masses resulting from infiltration of the submandibular glands and submental lymph nodes.
B: Representative morphology of peripheral blood (PB) and histology of the submandibular masses (SM mass), spleen, and lung are shown for leukemic
mice transplanted with transduced BM MNC, CMP, and GMP. Primitive myelomonocytic cells were demonstrated in the peripheral blood, infiltrating the
lung parenchyma and effacing the normal splenic and submental lymph node architecture.
cursors) and only residual numbers of GMP. However, there
was a population of c-kit ?/Fc?RII?/CD34? cells that were not
present in wild-type mice (Figure 5B). Taken together, these
data indicate that there is impaired differentiation associated
with the development of leukemia, and that the population of
cells with properties of leukemic stem cells is downstream of
the transduced CMP and GMP.
Purified HSC were transduced with MOZ-TIF2 using the same
protocol, and serial replating and direct transplantation assays
were performed as described above for MOZ-TIF2-transduced
CMP or GMP. In consonance with results in CMP and GMP,
HSC transduced with MOZ-TIF2 demonstrated serial replating
ability and were able to grow in liquid culture in an IL-3-depen-
dent manner (data not shown). Direct transplantation of MOZ-
TIF2-transduced HSC also led to the development of AML in
mice with a similar penetrance but a slightly longer latency than
in transduced CMP, GMP, or whole BM MNC, perhaps relating
to their lower cell dose and transduction efficiency at trans-
plantation (Figure 6A).
The leukemias that developed were indistinguishable mac-
roscopically from other MOZ-TIF2 associated leukemias, again
MOZ-TIF2 transduction of HSC also results in AML
that is indistinguishable from other MOZ-TIF2-
We next tested whether MOZ-TIF2 could also engender acute
myeloid leukemia when transduced into the HSC compartment.
CANCER CELL : DECEMBER 2004 591
A R T I C L E
Figure 5. FACS analysis of MOZ-TIF2-associated
leukemias demonstrates that the compartment
with leukemia stem cell properties lies down-
stream of GMP
A: Similar expression patterns of Gr-1, Mac-1,
c-kit, CD45.2, CD19, and B220 are shown for un-
gated leukemic bone marrow cells from mice
transplanted with transduced BM MNC, CMP,
and GMP. Relative fluorescence for the markers
indicated on the x and y axes are shown on a
logarithmic scale. Quadrant markers segregate
positive and negative cells, and quadrant per-
centages are given.
B: Myeloid progenitor analysis of bone marrow
from representative animals transplanted with
MOZ-TIF2-transduced BM MNC, CMP, and GMP.
A wild-type animal (WT) is shown for comparison.
Relative fluorescence for the markers indicated
on the x and y axes are shown on a logarithmic
scale. The values shown represent the percent-
age proportion of total bone marrow cells ana-
lyzed. Stem cell and progenitor populations are
demonstrating mild splenomegaly and characteristic submental
masses. Microscopically, the pattern of organ infiltration and
the histological appearance of the leukemic cells were charac-
teristic of other MOZ-TIF2-associated leukemias, as was ex-
pression of immunophenotypic markers (Figures 6B and 6C).
Analysis of myeloid progenitor populations also showed these
populations to be reduced (Figure 6D), again suggesting that
the compartment with self-renewal capacity lies downstream
CMP and GMP are populations of cells that are irrevocably
committed to terminal differentiation and apoptotic cell death.
Remarkably, transduction of these committed progenitors with
MOZ-TIF2 confers properties of self-renewal in serial replating
assays, and induces acute monocytic leukemia that can be
serially transferred into secondary recipients.
The histological and immunological phenotypes of the
monocytic MOZ-TIF2-associated leukemias are virtually indis-
tinguishable regardless of whether the initially transduced cell
population was whole bone marrow mononuclear cells, CMP,
GMP,or HSC.The explanationthatwe favorfor thisobservation
is that MOZ-TIF2 confers a specified phenotype regardless of
the transduced cell type among whole bone marrow, CMP,
GMP, or HSC populations. The phenotype includes conferring
ing assays and development of leukemia in vivo that can be
serially transplanted, and impaired differentiation downstream
of GMP in the monocytic lineage. A similar pattern of differentia-
tion has also recently been described for transduction of hema-
topoietic progenitors with the oncogene MLL-ENL (Cozzio et
An important question is whether these results can be ex-
plained as a consequence of retroviral insertional mutagenesis.
To address this question, we have undertaken two types of
control experiments. First, we have retrovirally transduced CMP
AML induced by transduction of myeloid progenitors
produces leukemia in secondary recipient mice
The long-term self-renewal properties of leukemic cells derived
from MOZ-TIF2-transduced CMP or GMP were further tested
by transplantation into sublethally irradiated secondary recipi-
ents at a range of cell doses from 1 ? 106bone marrow cells
with ? 1 ? 105cells developed leukemia with latencies of be-
tween 32 and 53 days, with an occasional mouse transplanted
with 1 ? 104cells also developing disease. Mice below the
threshold dosage of 1 ? 104did not develop leukemia with a
follow up of 84 days. The leukemias were histologically and
immunophenotypically indistinguishable from the primary dis-
demonstrated oligoclonal retroviral integration, as did the pri-
mary disease (data not shown).
CANCER CELL : DECEMBER 2004
A R T I C L E
Figure 6. Transduction of HSC with MOZ-TIF2 also
engenders AML in recipient mice
A: Kaplan-Meier graph comparing the latency
and incidence of AML in recipients of MOZ-TIF2-
transduced HSC with transduced CMP, GMP,
and BM MNC.
B and C: Representative histology and immu-
nophenotype from a MOZ-TIF2-transduced HSC-
associated AML demonstrating the similarity to
other MOZ-TIF2-associated leukemias. Relative
fluorescence for the markers indicated on the x
and y axes are shown on a logarithmic scale.
tive cells and quadrant percentages are given.
D: Myeloid progenitor analysis, again demon-
strating a decrease in myeloid progenitor sub-
sets, suggesting that the compartment with
self-renewal potential lies downstream of GMP.
Relative fluorescence for the markers indicated
on the x and y axes are shown on a logarithmic
mia in the murine BMT system using whole bone marrow. These
lack the ability to confer properties of leukemic stem cells to
CMP or GMP, indicating that retroviral insertional mutagenesis
with nonleukemogenic mutants of MOZ-TIF2 alone is not suffi-
cient to confer such properties. Second, retroviral transduction
of BCR-ABL did not confer properties of leukemic stem cells to
CMP or GMP. These data demonstrate that retroviral insertional
mutagenesis, even in collaboration with a fully active leukemo-
genic tyrosine kinase, is not sufficient to cause leukemia. To-
exclude the possibility that an active MOZ-TIF2, but not BCR-
ABL, can collaborate with mutations induced by retroviral muta-
progenitors. Limiting dilution analysis indicated that the fre-
quency of the leukemia repopulating cell was low in this assay
mia (Lessard and Sauvageau, 2003) and in human leukemic
cells transplanted into NOC-SCID recipient mice (Bonnet and
Dick, 1997; Hope et al., 2004). Taken together, these data sup-
port the existence of a subset of leukemic cells that retain long
term self-renewal potential.
What is the nature of the qualitative difference between
BCR-ABL and MOZ-TIF2 in conferring properties of self-renewal?
differences between these two oncogenes. BCR-ABL expres-
sion in humans, and in mouse models, confers proliferative
and/or survival advantage to hematopoietic progenitors without
affecting differentiation, and is associated with a chronic myelo-
proliferative phenotype characterized by leukocytosis and nor-
mal differentiation. In contrast, MOZ-TIF2 is associated with
acute monocytic leukemia in humans and in murine models of
disease characterized by impaired hematopoietic differentiation
model in Figure 8, where BCR-ABL confers proliferative and
survival signals to the committed myeloid progenitors, which
may allow transient expansion of downstream progeny, but
does not confer properties of self-renewal. In contrast, MOZ-
TIF2 confers properties of self-renewal while also impairing my-
eloid differentiation. This allows preleukemic expansion of this
compartment and may facilitate the subsequent development
of AML. The basis for the difference between BCR-ABL and
MOZ-TIF2 in conferring properties of leukemic stem cells to
CMP and GMP remains to be discerned, but it is clear that
leukemia-associated oncogenes vary in their ability to do so.
In addition, these data may have important clinical implica-
tions for treatment of leukemias associated with the BCR-ABL
oncogene. If, as our data suggests, BCR-ABL signals are dis-
pensable for maintenance of self-renewal, then small molecule
inhibitors of BCR-ABL may not target the relevant “leukemia
stem cell” in this disorder (Graham et al., 2002)—indeed, it
appears that imatinib therapy as a single agent, though quite
effective in remission induction, is rarely if ever curative in BCR-
ABL-positive CML (Melo et al., 2003).
It is not certain how many, or what type of, leukemia onco-
progenitors. However, MLL-ENL has recently been shown to
CANCER CELL : DECEMBER 2004 593
A R T I C L E
Figure 7. Transplantation of leukemic cells from
animals transplanted with MOZ-TIF2-transduced
CMP and GMP causes identical leukemia in sec-
ondary recipients in a cell dose-dependent
A: Kaplan-Meier graph of survival for secondary
recipients transplanted with leukemic cells from
primary mice transplanted with either MOZ-TIF2-
transduced CMP or GMP. All animals trans-
planted with greater than 1 ? 104cells devel-
oped leukemia, while no animals transplanted
with less than 1 ? 104developed disease.
B: Representative histology demonstrating the
tribution of MOZ-TIF2 disease in secondary recipi-
confer similar properties of self-renewal and to generate AML
in committed myeloid progenitors (Cozzio et al., 2003). We
speculate that this ability may be a generic property of certain
scription factors. For example, there is evidence for increased
self-renewal of AML1-ETO-expressing progenitors and HSC,
both in murine models and transduced human CD34? cells (de
Guzman et al., 2002; Mulloy et al., 2002, 2003; Higuchi et al.,
2002; Rhoades et al., 2000; Tonks et al., 2003), although as yet,
no evidence exists for the reestablishment of such properties
in progenitors which inherently lacked self-renewal potential.
In addition, analysis of gene expression profiles in the U937
hematopoietic precursor cell line stably expressing AML1-ETO,
PML-RAR?, or PLZF-RAR? demonstrated the induction of
genes involved in the maintenance of the stem cell phenotype
(Alcalay et al., 2003; Muller-Tidow et al., 2004). This suggests
fusions may havea wider relevance in thepathogenesis of AML.
that leukemogenic mutations often occur in HSC, as expression
of MOZ-TIF2 in this population also resulted in leukemia. The
data only indicate that if such mutations arise in committed
progenitors, they are capable of transforming committed pro-
genitors to cells with the properties of leukemia stem cells.
BCR-ABL or inactive MOZ-TIF2 mutants with MOZ-TIF2 itself
should allow for definition of the transcriptional programs that
confer properties of self-renewal to hematopoietic stem cells.
These programs may be targets for therapeutic intervention in
ties of self-renewal to adult somatic cells in other therapeutic
contexts, such as tissue regeneration.
Figure 8. Development of AML is facilitatedby the properties of self-renewal
conferred to myeloid progenitors by MOZ-TIF2
The normal differentiation of an HSC to a granulocyte through specific
myeloid progenitor intermediates is shown in the left column. In the center
column, transduction of myeloid progenitors with BCR-ABL may provide
a proliferation and survival signal to these progenitors without impairing
differentiation. This in turn may lead to an expansion of the granulocyte
compartment, but without self-renewal is insufficient to generate leukemia
in the recipient mice. In contrast, transduction of myeloid progenitors with
MOZ-TIF2 (right column) both restores self-renewal to this compartment and
impairs its further differentiation. In combination, these properties allow pre-
leukemic expansion of the compartment and may facilitate the develop-
ment of AML upon the acquisition of further cooperating mutations.
Cell staining and sorting
Bone marrow mononuclear cells were flushed from the leg bones of C57B6
mice (Taconic, Germantown, NY). For transplants using transduced bone
marrow mononuclear cells, the mice were treated with 5FU 150 mg/kg
(Sigma, St. Louis, MO) intraperitoneally 6 days prior to sacrifice. The cells
were then washed and the red cells lysed on ice with RBC lysis buffer
CANCER CELL : DECEMBER 2004
A R T I C L E
(Gentra, Minneapolis, MN). HSC, CMP, and GMP populations were sorted
and analyzed as previously reported (Na Nakorn et al., 2002).
conjugated with fluorescein isothiocyanate (FITC), phycoerthyrin (PE), or
biotin. Binding of biotinylated primary antibodies was detected using PE-
conjugated streptavidin (Immunotech, Westbrook, ME) or FITC-conjugated
avidin (Southern Biotechnology, Birmingham, AL). Cells were washed once
in staining buffer followed by two-colored flow cytometric analysis with a
FACScan (Becton Dickinson, San Jose, CA). A minimum of 10,000 events
was acquired showing FITC and PE fluorescence signals of viable cells
gated on the basis of forward and side scatter signals.
Retroviral constructs, retroviral transduction,
and bone marrow transplant assay
The MOZ-TIF2, MOZ-TIF2 mutants, and BCR-ABL MSCV-pgk-Neo and
MSCV-IRES-GFP vectors and the production of retroviral supernatants are
as previously described (Dash et al., 2002; Deguchi et al., 2003). Bone
marrow MNC, HSC, and CMP were all incubated overnight in IMDM (Life
Technologies, Rockville, MD) supplemented with murine IL-11 10 ?g/ml and
SCF 10 ?g/ml (both R&D Systems, Minneapolis, MN), and transduced the
next day. GMP cells were transduced immediately following sorting. For
transduction, cells were plated in IMDM supplemented with IL-11 and SCF
in 6-well plates (Becton Dickinson, Franklin Lakes, NJ) coated with 40 ?g/ml
Fibronectin (Takara, Japan). Retroviral supernatant was added along with
Polybrene 4 ?g/ml and the cells were incubated overnight at 37?C in 5%
CO2. The cells were then harvested, counted, and resuspended in 0.5 ml
Hank’s balanced salt solution (Life Technologies, Rockville, MD). For BM
MNC, each of 4 mice received 1 ? 106cells. For HSC, CMP, and GMP, 3
experiments each with transplantation of 2 mice for each group were per-
formed. The transplanted cell dose range was 4–12 ? 103for HSC, 3–9 ?
104cells for CMP, 3–8 ? 104for GMP, and 1 ? 106cells for each of
the MOZ-TIF2-transduced myeloid cell lines. Transduction efficiencies were
measured for each batch of MOZ-TIF2 and BCR-ABL viral supernatants
used in these experiments by either GFP expression on FACS analysis or
the ratio of colony number in the presence or absence of G418 selection,
and were comparable for each progenitor population between constructs.
Secondary transplants using bone marrow from primary leukemic animals
initially transplanted with either MOZ-TIF2-transduced CMP or GMP were
performed at limiting dilution using doses of 1 ? 106, 1 ? 105, 1 ? 104, 1 ?
103, or 1 ? 102, with 5 mice transplanted for CMP and GMP at each dosage.
The bone marrow from CMP- and GMP-induced MOZ-TIF2 leukemias was
matched for leukemic infiltration (CMP 45% and GMP 39% of cells GFP-
positive). Progenitor cells and cell lines were transplanted along with 0.5 ?
106BM MNC from C57B6.SJL mice (Stock # 002014, Jackson Laboratories,
Bar Harbor, ME). Cells were injected into lethally irradiated (650 rads ? 2)
C57B6.SJL recipients (Jackson Laboratories, Bar Harbor, ME) via the lateral
tail vein. Secondary transplant recipient mice received sublethal irradiation
Southern blotting for proviral insertion and clonality
Genomic DNA was prepared from single-cell suspensions of tumor cells
using a PUREGENE DNA isolation kit according to the manufacturer’s proto-
col (Gentra Systems, Minneapolis, MN). Twenty micrograms of genomic
DNA was digested with EcoRI, which cuts once within the proviral sequence
and once within the integrated locus, and was subjected to electrophoresis
and hybridization according to standard protocols.
We acknowledge the members of the Gilliland and Akashi labs for helpful
discussion. This work was supported by NIH grants CA66996 and DK50654,
and a Leukemia and Lymphoma Society Center grant. (D.G.G.). D.G.G. is
an Investigator of the Howard Hughes Medical Institute and is a Doris Duke
Distinguished Clinical Scientist. B.J.P.H is a Senior Clinical Fellow of the
Leukemia Research Fund (UK).
Received: March 3, 2004
Revised: August 25, 2004
Accepted: October 21, 2004
Published: December 20, 2004
Akashi, K., Traver, D., Miyamoto, T., and Weissman, I. (1999). A clonogenic
common myeloid progenitor that gives rise to all myeloid lineages. Nature
Alcalay, M., Meani, N., Gelmetti, V., Fantozzi, A., Fagioli, M., Orleth, A.,
Riganelli, D., Sebastiani, C., Cappelli, E., Casciari, C., et al. (2003). Acute
myeloid leukemia fusion proteins deregulate genes involved in stem cell
maintenance and DNA repair. J. Clin. Invest. 112, 1751–1761.
Serial replating assay and generation of IL-3-dependent
myeloid cell lines
Serial replating assays were performed as previously described (Lavau et
al., 1997) for 4 platings. Excess cells from the 3rdand 4thplatings were
amalgamated and each total cell population was expanded in RPMI supple-
mented with 20% FCS and IL-3 10 ?g/ml (R&D systems, Minneapolis, MN).
These cells were then counted weekly on an inverted microscope by trypan
Al-Hajj, M., Wicha, M., Benito-Hernandez, A., Morrison, S., and Clarke, M.
(2003). Prospective identification of tumorigenic breast cancer cells. Proc.
Natl. Acad. Sci. USA 100, 3983–3988.
Bonnet, D., and Dick, J.E. (1997). Human acute myeloid leukemia is orga-
nized as a hierarchy that originates from a primitive hematopoietic cell. Nat.
Med. 3, 730–737.
Borrow, J., Stanton, V.P., Andresen, J.M., Becher, R., Behm, F.G., Chaganti,
R., Civin, C., Disteche, C., Dube, I., Frischauf, A., et al. (1996). The transloca-
ferase to the CREB-binding protein. Nat. Genet. 14, 33–41.
Murine tissues were fixed for at least 24 hr in 10% neutral buffered formalin
(Sigma, St. Louis, MO), dehydrated in alcohol, cleared in xylene, and infil-
trated with paraffin on an automated processor (Leica, Bannockburn, IL).
Tissue sections (4 ?m thick) were placed on charged slides, deparaffinized
in xylene, rehydrated through graded alcohol solutions, and stained with
hematoxylin and eosin (H&E).
Bradford, G., Williams, B., Rossi, R., and Bertoncello, I. (1997). Quiescence,
cycling, and turnover in the primitive hematopoietic stem cell compartment.
Exp. Hematol. 25, 445–453.
Carapeti, M., Aguiar, R., Goldman, J., and Cross, N. (1988). A novel fusion
between MOZ and the nuclear receptor coactivator TIF2 in acute myeloid
leukemia. Blood 91, 3127–3133.
Immunophenotypic analysis of leukemic cells
Single cell suspensions of bone marrow, spleen, submandibular glands, and
lymph nodes, and granulocytic sarcomata if present, were prepared. Red
blood cells were lysed in ammonium chloride solution (150 mmol/l NH4Cl,
10 mmol/l KHCO3, 0.1 mmol/l EDTA [pH 7.4]) for 5 min at room temperature.
The cells were washed in PBS with 0.1% NaN3and 0.1% bovine serum
albumin (BSA; staining buffer). To block nonspecific Fc receptor-mediated
oma line (anti-CD16/CD32; cell line American Type Culture Collection, Rock-
ville, MD) for 20 min on ice. Aliquots of 0.5 to 1.0 ? 106cells were then
stained for 20 min on ice with monoclonal antibodies specific for B220
(CD45R), CD19, CD34, CD45.2 (Ly5.2), c-kit, Gr-1, Mac-1, CD4, or CD8
Chesier, S., Morrison, S., Liao, X., and Weissman, I. (1999). In vivo prolifera-
tion and cell cycle kinetics of long-term self-renewing hematopoietic stem
cells. Proc. Natl. Acad. Sci. USA 96, 3120–3125.
Cobaleda, C., Guitierrez-Cianca, N., Perez-Losada, J., Flores, T., Garcia-
etic cell is the target for the leukemic transformation in human Philadelphia-
positive acute lymphoblastic leukemia. Blood 95, 1007–1013.
CANCER CELL : DECEMBER 2004595
A R T I C L E Download full-text
man, I.L. (2003). Similar MLL-associated leukemias arising from self-
renewing stem cells and short-lived myeloid progenitors. Genes Dev. 17,
and leukemic transformation of a myelomonocytic precursor by retrovirally
transduced HRX-ENL. EMBO J. 16, 4226–4237.
Lessard, J., and Sauvageau, G. (2003). Bmi-1 determines the proliferative
capacity of normal and leukaemic stem cells. Nature 423, 255–260.Dash, A., Williams, I., Kutok, J., Tomasson, M., Anastasiadou, E., Lindahl,
K., Li, S., Van Etten, R., Borrow, J., Housman, D., et al. (2002). A murine
model of CML blast crisis induced by cooperation between BCR/ABL and
NUP98/HOXA9. Proc. Natl. Acad. Sci. USA 99, 7622–7627.
isolation of human clonogenic common myeloid progenitors. Proc. Natl.
Acad. Sci. USA 99, 11872–11877.
H., Hiebert, S.W., and Klug, C.A. (2002). Hematopoietic stem cell expansion
and distinct myeloid developmental abnormalities in a murine model of the
AML1-ETO translocation. Mol. Cell. Biol. 22, 5506–5517.
Martin, P.J., Najfeld, V., Hansen, J.A., Penfold, G.K., Jacobson, R.J., and
Fialkow, P.J. (1980). Involvement of the B-lymphoid system in chronic my-
elogenous leukaemia. Nature 287, 49–50.
Mehrotra, B., George, T., Kavanau, K., Avet-Loiseau, H., Moore, D.N., Will-
netically aberrant cells in the stem cell compartment (CD34?lin-) in acute
myeloid leukemia. Blood 86, 1139–1147.
Deguchi,K.,Ayton,P., Carapeti,M.,Kutok,J.,Snyder,C., Williams,I.,Cross,
N., Glass, C., Cleary, M., and Gilliland, D. (2003). MOZ-TIF2-induced acute
myeloid leukemia requires the MOZ nucleosome binding motif and TIF2-
mediated recruitment of CBP. Cancer Cell 3, 259–271.
Melo, J.V., Hughes, T.P., and Apperley, J.F. (2003). Chronic myeloid leuke-
mia. Hematology (Am. Soc. Hematol. Educ. Program) 2003, 132–152.
Deininger, M., Goldman, J., and Melo, J. (2000). The molecular biology of
chronic myeloid leukemia. Blood 96, 3343–3356.
in hematopoietic cells is associated with self-renewal potential. Immunity 5,
Fialkow, P., Singer, J., Adamson, J., Vaidya, K., Dow, L., Ochs, J., and
Moohr, J. (1981). Acute nonlymphocytic leukemia: heterogeneity of stem
cell origin. Blood 57, 1068–1073.
Muller-Tidow, C., Steffen, B., Cauvet, T., Tickenbrock, L., Ji, P., Diedrichs,
S., Sargin, B., Kohler, G., Stelljes, M., Puccetti, E., et al. (2004). Translocation
products in acute myeloid leukemia activate the Wnt signalling pathway in
hematopoietic cells. Mol. Cell. Biol. 24, 2890–2903.
Graham, S.M., Jorgensen, H.G., Allan, E., Pearson, C., Alcorn, M.J., Rich-
mond, L., and Holyoake, T.L. (2002). Primitive, quiescent, Philadelphia-posi-
tive stem cells from patients with chronic myeloid leukemia are insensitive
to STI571 in vitro. Blood 99, 319–325.
Mulloy, J.C., Cammenga, J., MacKenzie, K.L., Berguido, F.J., Moore, M.A.,
and Nimer, S.D. (2002). The AML1-ETO fusion protein promotes the expan-
sion of human hematopoietic stem cells. Blood 99, 15–23.
Griffin, J., and Lowenberg, B. (1986). Clonogenic cells in acute myeloblastic
leukemia. Blood 68, 1185–1195.
Haase, D., Feuring-Buske, M., Konemann, S., Fonatsch, C., Troff, C., Ver-
beek, W., Pekrun, A., Hiddemann, W., and Wormann, B. (1995). Evidence
for malignant transformation in acute myeloid leukemia at the level of early
hematopoietic stem cells by cytogenetic analysis of CD34? subpopulations.
Blood 86, 2906–2912.
Mulloy, J.C., Cammenga, J., Berguido, F.J., Wu, K., Zhou, P., Comenzo,
R.L., Jhanwar, S., Moore, M.A., and Nimer, S.D. (2003). Maintaining the self-
renewal and differentiation potential of human CD34? hematopoietic cells
using a single genetic element. Blood 102, 4369–4376.
Na Nakorn, T., Traver, D., Weissman, I., and Akashi, K. (2002). Myeloeryth-
roid-restricted progenitors are sufficient to confer radioprotection and pro-
vide the majority of day 8 CFU-S. J. Clin. Invest. 109, 1579–1585.
Hanahan, D., and Weinberg, R. (2000). The hallmarks of cancer. Cell 100,
D.H., Bronner-Fraser, M., and Kornblum, H.I. (2003). Cancerous stem cells
can arise from pediatric brain tumors. Proc. Natl. Acad. Sci. USA 100,
Quijano, C., Moore, D.N., Arthur, D., Feusner, J., Winter, S., and Pallavicini,
19- marrow compartment in children with acute lymphoblastic leukemia.
Leukemia 11, 1508–1515.
J.R. (2002). Expression of a conditional AML1-ETO oncogene bypasses
embryonic lethality and establishes a murine model of human t(8;21) acute
myeloid leukemia. Cancer Cell 1, 63–74.
Reya, T., Morrison, S., Clarke, M., and Weissman, I. (2001). Stem cells,
cancer, and cancer stem cells. Nature 414, 105–111.
Rhoades, K.L., Hetherington, C.J., Harakawa, N., Yergeau, D.A., Zhou, L.,
Liu, L.Q., Little, M.T., Tenen, D.G., and Zhang, D.E. (2000). Analysis of the
role of AML1-ETO in leukemogenesis, using an inducible transgenic mouse
model. Blood 96, 2108–2115.
Hope, K., Jin, L., and Dick, J. (2004). Acute myeloid leukemia originates
from a hierarchy of leukemic stem cells classes that differ in self-renewal
capacity. Nat. Immunol. 5, 738–743.
Sattler, M., and Griffin, J.D. (2003). Molecular mechanisms of transformation
by the BCR-ABL oncogene. Semin. Hematol. 40, 4–10.
Ignatova, T.N., Kukekov, V.G., Laywell, E.D., Suslov, O.N., Vrionis, F.D., and
Steindler, D.A. (2002). Human cortical glial tumors contain neural stem-like
cells expressing astroglial and neuronal markers in vitro. Glia 39, 193–206.
Singh, S.K., Clarke, I.D., Terasaki, M., Bonn, V.E., Hawkins, C., Squire, J.,
and Dirks, P.B. (2003). Identification of a cancer stem cell in human brain
tumors. Cancer Res. 63, 5821–5828.
Lansdorp, P. (1997). Self-renewal of stem cells. Biol. Blood Marrow Trans-
plant. 3, 171–178.
Tonks, A., Pearn, L., Tonks, A.J., Pearce, L., Hoy, T., Phillips, S., Fisher, J.,
Downing, J.R., Burnett, A.K., and Darley, R.L. (2003). The AML1-ETO fusion
gene promotes extensive self-renewal of human primary erythroid cells.
Blood 101, 624–632.
Lapidot, T., Sirard,C., Vormoor, J., Murdoch, B.,Hoang, T., Caceres-Cortes,
J., Minden, M., Paterson, B., Caligiuri, M.A., and Dick, J.E. (1994). A cell
initiating human acute myeloid leukaemia after transplantation into SCID
mice. Nature 367, 645–648.
Wong, S., and Witte, O. (2001). Modeling Philadelphia chromosome positive
leukemias. Oncogene 20, 5644–5659.Lavau, C., Szilvassy, S.J., Slany, R., and Cleary, M.L. (1997). Immortalization
CANCER CELL : DECEMBER 2004