Monoclonal antibody-mediated targeting of CD123, IL-3 receptor alpha chain, eliminates human acute myeloid leukemic stem cells.
ABSTRACT Leukemia stem cells (LSCs) initiate and sustain the acute myeloid leukemia (AML) clonal hierarchy and possess biological properties rendering them resistant to conventional chemotherapy. The poor survival of AML patients raises expectations that LSC-targeted therapies might achieve durable remissions. We report that an anti-interleukin-3 (IL-3) receptor alpha chain (CD123)-neutralizing antibody (7G3) targeted AML-LSCs, impairing homing to bone marrow (BM) and activating innate immunity of nonobese diabetic/severe-combined immunodeficient (NOD/SCID) mice. 7G3 treatment profoundly reduced AML-LSC engraftment and improved mouse survival. Mice with pre-established disease showed reduced AML burden in the BM and periphery and impaired secondary transplantation upon treatment, establishing that AML-LSCs were directly targeted. 7G3 inhibited IL-3-mediated intracellular signaling of isolated AML CD34(+)CD38(-) cells in vitro and reduced their survival. These results provide clear validation for therapeutic monoclonal antibody (mAb) targeting of AML-LSCs and for translation of in vivo preclinical research findings toward a clinical application.
- [show abstract] [hide abstract]
ABSTRACT: Most human acute myeloid leukaemia (AML) cells have limited proliferative capacity, suggesting that the leukaemic clone may be maintained by a rare population of stem cells. This putative leukaemic stem cell has not been characterized because the available in vitro assays can only detect progenitors with limited proliferative and replating potential. We have now identified an AML-initiating cell by transplantation into severe combined immune-deficient (SCID) mice. These cells homed to the bone marrow and proliferated extensively in response to in vivo cytokine treatment, resulting in a pattern of dissemination and leukaemic cell morphology similar to that seen in the original patients. Limiting dilution analysis showed that the frequency of these leukaemia-initiating cells in the peripheral blood of AML patients was one engraftment unit in 250,000 cells. We fractionated AML cells on the basis of cell-surface-marker expression and found that the leukaemia-initiating cells that could engraft SCID mice to produce large numbers of colony-forming progenitors were CD34+ CD38-; however, the CD34+ CD38+ and CD34- fractions contained no cells with these properties. This in vivo model replicates many aspects of human AML and defines a new leukaemia-initiating cell which is less mature than colony-forming cells.Nature 03/1994; 367(6464):645-8. · 38.60 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The hematopoietic system is controlled by growth factors (cytokines) which can influence cell survival, proliferation, differentiation and functional activation. The study of cytokine receptor expression by flow cytometry could allow us to differentiate between normal and tumoral cells. We analyzed the expression of the interleukin-3 (IL-3) receptor a chain (CD123) in 22 normal samples and in a wide panel of hematologic malignancies using flow cytometry. We found that CD123 was expressed in the myeloid progenitor subpopulation but in contrast, normal lymphoid progenitors lacked CD123. We analyzed the CD123 expression pattern in 64 patients with acute leukemia, 45 with acute myeloid leukemia (AML) and 19 with acute lymphocytic leukemia (ALL) (13 B-cell lineage ALL and 6 T-cell lineage ALL). All the AML cases except two patients with M7 and all the B-cell lineage ALL patients were CD123 positive. In contrast, all the T-cell lineage ALL cases tested were CD123 negative. We also studied the CD123 expression pattern in 122 patients with a B-cell chronic lymphoproliferative disease (B-CLPD). CD123 was positive in three situations: 1) typical cases of hairy cell leukemia showed a specific, strong CD123 expression, 2) a subgroup of atypical chronic lymphocytic leukemia with a marked CD11c expression was also CD123 positive, and finally 3) transformed B-CLPD showed the phenomenon of transition from CD123 negativity to CD123 positivity simultaneuosly with morphologic changes. In summary, our data show high expression of IL-3 receptor a chain in hematologic malignancies. Given the high frequency of CD123 reactivity in blast cells in contrast to in normal precursors, this antigen could be applied to the study of minimal residual disease in acute leukemia. CD123 is expressed with a characteristic pattern in cases of hairy cell leukemia.Haematologica 01/2002; 86(12):1261-9. · 5.94 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Recombinant human (rh) interleukin-3 (IL-3) stimulated the proliferation and differentiation of erythroid, granulocyte, macrophage, eosinophil (Eo), and mixed colonies as well as megakaryocytes from human bone marrow cells. rh IL-3 was a weaker stimulus than rh granulocyte-macrophage colony-stimulating factor (GM-CSF) for day 14 myeloid cell colonies. At day 7 of incubation, rh IL-3 stimulated a few G, M, and Eo clusters but no colonies. This loss of responsiveness of myeloid cells to rh IL-3 was accentuated with further differentiation of the cells. rh IL-3 stimulated very few or no clones after five-day incubation with enriched promyelocytes and myelocytes, whereas rh GM-CSF was an efficient stimulus. Responsiveness to rh IL-3 was completely lost in postmitotic mature neutrophils. Incubation of these cells with rh IL-3 did not result in enhanced antibody-dependent cell-mediated cytotoxicity (ADCC) of tumor cells or superoxide anion production after stimulation with formyl-methyl-leucyl-phenylalanine (FMLP), although they could be stimulated by rh GM-CSF. In addition, preincubation of neutrophils with different concentrations of rh IL-3 failed to increase or decrease their response to rh GM-CSF. In contrast to neutrophils, mature Eos could be stimulated by rh IL-3 to kill antibody-coated tumor cells. These results show that cells of the neutrophilic myeloid series lose their responsiveness to h IL-3 as they differentiate and suggest that although h IL-3 may be an important therapeutic agent to use for hematopoietic regeneration in vivo, the lack of stimulation of mature neutrophil function makes it an unlikely sole candidate as adjunct therapy for treatment of infectious diseases.Blood 12/1988; 72(5):1797-804. · 9.06 Impact Factor
Cell Stem Cell
Monoclonal Antibody-Mediated Targeting
of CD123, IL-3 Receptor a Chain, Eliminates
Human Acute Myeloid Leukemic Stem Cells
Liqing Jin,1,6Erwin M. Lee,2,6Hayley S. Ramshaw,3Samantha J. Busfield,4Armando G. Peoppl,1Lucy Wilkinson,5
Mark A. Guthridge,3Daniel Thomas,3Emma F. Barry,3Andrew Boyd,5David P. Gearing,4Gino Vairo,4Angel F. Lopez,3
John E. Dick,1and Richard B. Lock2,*
1Division of Cell and Molecular Biology, University Health Network, Toronto, ON M5G 1L7, Canada
2Children’s Cancer Institute Australia for Medical Research, University of New South Wales, Sydney 2031, Australia
3The Division of Human Immunology, Centre for Cancer Biology, Adelaide 5000, Australia
4CSL Limited, Melbourne 3052, Australia
5Queensland Institute of Medical Research, Brisbane 4029, Australia
6These authors contributed equally to this work
Leukemia stem cells (LSCs) initiate and sustain the
acute myeloid leukemia (AML) clonal hierarchy and
possess biological properties rendering them resis-
of AML patients raises expectations that LSC-tar-
geted therapies might achieve durable remissions.
We report that an anti-interleukin-3 (IL-3) receptor
a chain (CD123)-neutralizing antibody (7G3) targeted
AML-LSCs, impairing homing to bone marrow (BM)
and activating innate immunity of nonobese diabetic/
severe-combined immunodeficient (NOD/SCID) mice.
7G3 treatment profoundly reduced AML-LSC engraft-
ment and improved mouse survival. Mice with pre-
established disease showed reduced AML burden in
the BM and periphery and impaired secondary trans-
plantation upon treatment, establishing that AML-
LSCs were directly targeted. 7G3 inhibited IL-3-medi-
ated intracellular signaling of isolated AML CD34+
CD38?cells in vitro and reduced their survival. These
results provide clear validation for therapeutic mono-
clonal antibody (mAb) targeting of AML-LSCs and for
translation of in vivo preclinical research findings
toward a clinical application.
is less than 30%, with progressively worse prognosis for more
of work is emerging in experimental systems that predicts LSCs
may lie at the heart of posttreatment relapse and chemoresist-
ance. AML is organized as a cellular hierarchy sustained by
LSCs at their apex (Bonnet and Dick, 1997; Guan and Hogge,
2000; Guzman et al., 2001; Hope et al., 2004; Lapidot et al.,
of self-renewal while still generating rapidly proliferating progen-
to conventional chemotherapeutics that target proliferating cells
(Bonnet and Dick, 1997; Guan and Hogge, 2000; Guzman et al.,
2001; Hope et al., 2004; Ishikawa et al., 2007; Lapidot et al.,
1994; Wang and Dick, 2005). In addition, minimal residual
disease occurrence and poor survival have been attributed to
the ability of LSCs to engraft NOD/SCID mice (Pearce et al.,
2006) and high CD34+CD38?frequency at time of diagnosis in
ative that new treatments are developed to complement estab-
lished chemotherapy by specifically eliminating AML-LSCs for
2006; Aribi et al., 2006; Morgan and Reuter, 2006; Stone, 2007).
As with normal hematopoietic stem cells (HSCs), very little is
known of the molecular regulation that governs the self-renewal,
differentiation, andsurvivalofAML-LSCs, althoughbothof these
ability, and expression of some surface markers including the
CD34+CD38?immunophenotype (Bhatia et al., 1997; Bonnet
and Dick, 1997; Lapidot et al., 1994). CD123, which is widely re-
ported to be overexpressed on AML blasts, CD34+leukemic
progenitors, and AML-LSCs in comparison with normal HSCs
(Florian et al., 2006; Graf et al., 2004; Hauswirth et al., 2007;
Jordan et al., 2000; Munoz et al., 2001; Riccioni et al., 2004;
Sperr et al., 2004; Testa et al., 2002; Yalcintepe et al., 2006),
represents a promising cell-surface target for the development
of therapeutics that specifically target AML-LSCs but not
HSCs. CD123 is the a subunit of the IL-3 receptor (IL-3R), the
major binding protein for IL-3, which together with CD131 (bc)
forms the functional heterodimeric high-affinity IL-3R. The
binding of IL-3 to CD123 is species specific and leads to activa-
tion of the receptor that promotes cell survival and proliferation
(Bagley et al., 1997; Miyajima et al., 1993).
Overexpression of CD123 on AML cells confers a range of
growth advantages over normal HSCs; AML cells proliferate
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 31
extensively with IL-3 treatment in vitro (Budel et al., 1989; Miyau-
chi et al., 1987; Pebusque et al., 1989; Vellenga et al., 1987), and
some AML samples secrete cytokines including IL-3 (Elbaz and
Shaltout, 2001; Guan et al., 2003; Nowak et al., 1999). Moreover,
high-level CD123 expression on AML cells correlates with the
level of IL-3-stimulated and spontaneous signal transducer and
activator of transcription 5 (STAT5) activation, the proportion of
cycling cells, a more primitive cell-surface phenotype, and resis-
tance to apoptosis (Graf et al., 2004; Testa et al., 2002, 2004).
Clinically, high CD123 expression in AML is associated with
higher blast counts at diagnosis and a lower complete remission
rate that results in reduced survival (Graf et al., 2004; Testa et al.,
2002, 2004). Collectively, these data point to the significance of
CD123 expression in leukemia cell stimulation and AML patient
The increased expression of CD123 on LSCs compared with
HSCs presents an opportunity for selectively targeting AML-
LSCs with a therapeutic antibody. Besides the possibility that
IL-3 is required for LSC functions, an antibody to CD123 could
stimulate host immune-mediated mechanisms for cell killing.
An antibody with both IL-3R-neutralizing and innate immunity-
activating properties could represent an ideal therapeutic
candidate for clinical testing. The mAb 7G3, raised against
CD123, has previously been shown to inhibit IL-3-mediated
proliferation of leukemic cell lines (Sun et al., 1996). While
AML-LSCs are often cited to be enriched in the CD34+CD38?
fraction, recent reports have demonstrated that other fractions,
such as the CD34+CD38+subpopulation, also have NOD/SCID
repopulating capacity (McKenzie et al., 2006; Taussig et al.,
2008). In this report, we show that CD123 is highly expressed
on the bulk of AML cells as well as the CD34+CD38?fraction
compared to normal hematopoietic cells. Importantly, we
demonstrate that 7G3 targeting of CD123 in the absence of
exogenous human cytokines impairs AML-LSCs in vivo. This
occurs through at least two mechanisms involving inhibition of
homing of CD34+CD38?cells and engraftment of AML-LSCs in
the NOD/SCID xenograft model, as well as activation of innate
immunity in NOD/SCID mice. As a prerequisite for the potential
role of 7G3 in inhibiting IL-3-mediated growth advantages on
AML-LSCs, we demonstrate that both the unsorted and the
CD34+CD38?subpopulations of AML cells proliferate and
survive via IL-3-mediated intracellular signaling pathways and
that these are inhibited by 7G3 in vitro. The recent characteriza-
of human malignancies (Wang, 2007), as well as their relative
resistance to conventional chemotherapy and radiotherapy
(Rich and Bao, 2007), supports the broad applicability of our
approach and provides rationale for the progression of AML-
LSC-targeted therapeutics from preclinical evaluation to clinical
Ex Vivo 7G3 Treatment Selectively Inhibits AML
Engraftment in NOD/SCID Mice
Since AML-LSCs are central to long-term AML growth and they
are difficult to assay in vitro, we used the SCID-leukemia initi-
ating cell (SL-IC) assay to determine whether 7G3 can directly
target AML-LSCs and inhibit their repopulating ability. Ex vivo
7G3 incubation markedly reduced the engraftment of 10 of 11
primary AML samples in sublethally irradiated NOD/SCID mice
to a mean of 11.4% ± 1.9% of isotype-matched (IgG2a-treated)
controls (p = 0.00021, Figure 1A, Table 1). This reduction in
engraftment was sustained in five of seven samples when as-
sessed between 8 and 10 weeks following inoculation (5.7% ±
1.7% of controls, p = 0.004). Ex vivo 7G3 treatment inhibited
the engraftment of AML-8 harvested at both diagnosis and
relapse to a similar extent. AML-5 was the only AML sample in
which engraftment was not reduced by ex vivo 7G3 treatment.
Although the reason for this is unknown, it is noteworthy that
AML-5 is a monosomy 7 sample (noted for poor prognosis)
with relatively low CD123 expression (Table 1).
We next investigated the sensitivity of normal cord blood (CB)
and BM (NBM) to 7G3 using the same strategy as for AML
samples to determine if there was differential targeting of normal
HSCs. When measured at 4–11 weeks postinoculation, 7G3
significantly reduced the engraftment of only two of five normal
samples (Figure 1B and Table 1). The inhibitory effect of 7G3
on the engraftment of normal cells (76.5% ± 8.9% engraftment
relative to IgG2a controls) was significantly less (p < 0.0001)
than against AML cells. Additionally, 7G3 treatment did not alter
the differentiation profiles of the engrafted normal human hema-
topoietic populations (data not shown). Furthermore, to demon-
strate the clinical relevance and specificity of 7G3 treatment
against LSCs and not normal HSCs, we showed that a mouse
anti-human HLA-A,B,C antibody indiscriminately inhibited the
engraftment of two AML and three normal samples (see Fig-
ure S1 available online). Independent analysis at two different
institutions (Sydney and Toronto) revealed that CD123 expres-
sion on AML CD34+CD38?cells (relative fluorescence index
[RFI] 38.2 ± 6.6) was significantly higher than on their normal
counterparts (RFI 9.6 ± 1.6) (Figure 1C and Table 1). The engraft-
ment levels of ex vivo 7G3-treated samples were inversely
correlated with the intensity of CD123 expression on the
CD34+CD38?population (Figure S2; Spearman R = ?0.69).
Taken together, we can conclude that normal HSCs are consid-
erably less sensitive to 7G3 thanAML-LSCs, due, at least in part,
to their relatively low levels of cell-surface CD123 expression.
The reduction in AML engraftment caused by ex vivo 7G3
treatment was also associated with improved survival. Mice
transplanted with IgG2a- or 7G3-treated AML-9 cells exhibited
median survival of 11.5 and 24 weeks, respectively (Figure 1D),
with 40% of the 7G3 group surviving beyond the end of the
experiment (25 weeks), in contrast with the control group, in
which no mice survived beyond 20 weeks.
7G3 Inhibits AML Homing Capacity in NOD/SCID Mice
To gain insight into the mechanism whereby 7G3 inhibited AML-
LSC engraftment, we investigated the influence that antibody
binding had on AML cell trafficking, since the SL-IC assay
requires AML-LSCs to traffic to the BM in order to survive and
were performed on two AML samples (AML-8-rel and -9)
following ex vivo 7G3 treatments. 7G3 reduced the homing effi-
ciency of AML-9 in the BM to 12.2% ± 2.7% and in the spleen to
9.4% ± 2.4% of controls (Figure 2A), and inhibited the homing of
AML-8-rel in the BM to 34.7% ± 5.6% (Figure 3A) and in the
spleen to 46.9% ± 3.5% of controls. To better distinguish the
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
32 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.
effects of 7G3 on AML homing, lodgment, and proliferation, ex
vivo-treated AML-8-rel cells were transplanted intravenously
(i.v.) via the tail vein or directly into the right femur (RF). The intra-
femoral (IF) approach circumvents the AML-LSC trafficking/
homing processes associated with the circulation (Mazurier
et al., 2003). While 7G3 remained effective in significantly
reducing the engraftment in both the injected femur and the non-
injected bones, IF inoculation did attenuate the inhibitory effects
of 7G3 on engraftment in comparison with i.v. inoculation
In order to more directly demonstrate 7G3 inhibition of
AML-LSCs, we investigated the impact of 7G3 treatment on
CD34+CD38?cells since AML-LSCs are significantly enriched
in this fraction (Bonnet and Dick, 1997). The number of
CD34+CD38?cells from AML-8-rel and AML-9 homing to the
BM was reduced by ex vivo 7G3 treatment to 8.4% ± 0.018%
and 12.0% ± 4.3% of control, respectively (Figure 2C). Similarly,
the number of AML-9 CD34+CD38?cells homing to the spleen
was reduced to 3.8% ± 1.5% of control. To further confirm this
CD34+CD38?cells from AML-9 following ex vivo antibody treat-
ment. The homing efficiency of human cells in the 7G3-treated
group was reduced to 7.8% ± 1.7% of IgG2a controls in the
BM and 11.2% ± 0.84% in the spleen (Figure 2D). Consistent
with the observation that 7G3 diminished AML-LSC homing
capacity, ex vivo 7G3 treatment reduced the number of
CD34+CD38?cells in the BM xenografts of three AML samples
(Figure 2E). By contrast, the number of CD34+CD38?cells
present in xenografts established from four independent normal
hematopoietic samples following ex vivo 7G3 treatment was
81.9% ± 11.6% of IgG2a controls (p = 0.19, data not shown).
Collectively, we can conclude that 7G3 inhibits not only homing
but also lodgment and proliferation of AML-LSCs in the BM
7G3-Mediated Inhibition of AML-LSC Homing
and Engraftment Is Fc Dependent
In order to determine whether the inhibitory effects of 7G3 are Fc
mediated, the homing efficiency of AML cells was examined
following treatment with F(ab0)2 fragments of various CD123-
targeting antibodies. Incubation of AML-8-rel with two MAbs
clones, 6H6 and 9F5, that bind CD123 but are weakly neutral-
izing reduced the homing efficiency in the BM to a similar extent
as 7G3 (Figure 3A). In contrast, when AML-8-rel cells were
treated ex vivo with 7G3 or 6H6 F(ab0)2 fragments, the inhibitory
effects of each antibody on AML homing were attenuated.
In addition, the Fc requirement for inhibition of NOD/SCID
repopulation was also examined. While ex vivo incubation of
AML-9 and AML-10 cells with 7G3 or 9F5 significantly reduced
their ability to repopulate mouse BM, the corresponding F(ab0)2
Figure 1. Ex Vivo 7G3 Treatment Selectively Inhibits the Repopulating Ability of AML Primary Cells in NOD/SCID Mice
(A) Percentage of human AML cells in the BM of mice transplanted with 7G3 or IgG2a control-treated AML cells at indicated time points. n = 3–10 per treated
(B) Levels of human engraftment in the BM of mice transplanted with 7G3 or IgG2a control-treated CB and NBM cells. Bars of CB represent the results from three
separate experiments. n = 4–6 mice per group in each experiment.
(C) CD123 expression on total and CD34+CD38?fractions of AML and normal cells. Each point represents an individual sample. Bars represent the mean.
(D) Kaplan-Meier survival curves of mice transplanted with IgG2a or 7G3 ex vivo-treated AML-9 cells. Survival curves were compared by log rank test. n = 10 per
group. Error bars represent mean ± SEM; *p < 0.05, **p < 0.01, and ***p % 0.0001 between selected groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 33
antibody fragments were ineffective (Figure 3B), despite 7G3
F(ab0)2 retaining its IL-3Ra-neutralizing activity (data not shown).
The requirement for Fc regions to inhibit homing and re-
population, combined with the reduced efficacy of 7G3 when
trafficking in the circulation was circumvented by IF transplanta-
tion, strongly supports a role for the innate immune system in
mediating at least a portion of the inhibitory effects of 7G3.
CD122+Cells Contribute to 7G3-Mediated Inhibition
of AML Homing and Repopulation in NOD/SCID Mice
While NOD/SCID micearedevoid of functional TandBcells, and
are defective in complement fixation, they retain residual levels
of innate effector activity (principally due to NK cells and macro-
phages) that can affect stem cell engraftment. To determine
whether residual NOD/SCID innate immunity contributed to the
inhibitory effects of 7G3 on LSCs, mice were injected with anti-
CD122 mAb prior to transplantation with ex vivo 7G3-treated
AML-8-rel cells. In the IgG2a control-treated groups, leukemic
engraftment in the CD122+cell-depleted mice was increased
to 113.3% ± 2.8% of nondepleted mice (Figure 3C), reflective
of our earlier data showing increased detection of HSCs in
such recipients (McKenzie et al., 2005). We found that depletion
of CD122+cells significantly attenuated the ability of 7G3 to
reduce leukemia engraftment from 82.7% ± 9.4% to 39.8% ±
a significant difference still remained between 7G3-treated and
IgG2a-treated groups. Similarly, when we used NOD/SCID inter-
leukin-2 receptor g chain null mice (NOD/SCID/IL-2Rgnull), which
have lower residual NK cell activity than NOD/SCID mice (Ito
et al., 2002), we observed similar attenuation, but not complete
ablation, of the inhibitory effects of ex vivo 7G3 treatment on
AML-1 engraftment in the BM (33.3% ± 12.6% of control
compared with 1.1% ± 0.9% of control for NOD/SCID mice,
Additionally, anti-CD122 antibody treatment also partially
attenuated the ability of 7G3 to block AML cell homing to the
BM observed in both AML-8-rel and AML-9. As shown in
Figure 3D, the homing efficiency of AML-9 cells treated with
7G3 was 8.4% ± 1.4% of control, and this was attenuated to
18.2% ± 3.1% with depletion of CD122+cells. The number of
Table 1. Patient Characteristics, CD123 Expression, and Effects of mAb 7G3 on Engraftment of AML and Normal Hematopoietic Cells
Effect of 7G3
as % Control)
270/FM1 NormalMutant270Deceased Apheresis1962.226.1
475/MM5aTrisomy 8Mutant 300DeceasedApheresis 273.536.54.7
653/M M4Eo Inv16Mutant 300Deceased BM 5044.9 24.21.5
780/M M5NA NA122 NAApheresis 451.9 802
847/FM4 NA NA33DeceasedApheresis436 8.28 4611.1
47/F M4 NANA33DeceasedApheresis 436 6.6 52.417.3
11 78/MM2Normal Wild-type166 DeceasedBM 260.2
NBM-3was aCD34+sortednormal BMsample.BM,bone marrow. CB,cordblood. NA,notavailable.NE, noengraftmentincontrols. WCC,peripheral
blood white cell count (3109/L).
aAt diagnosis (this column).
bFAB criteria (this column).
cFrom date of initial diagnosis (this column).
dPercent of total population (this column).
eRFI of CD34+CD38?population (this column).
fMean engraftment in the ex vivo 7G3-treated group as a percentage of the IgG2a-treated group, based on Figures 1A and 1B (this column).
gSample had very low proportion of CD34+cells.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
34 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.
CD34+CD38?cells that homed to the BM of mice was also
reduced to 5.3% ± 1.1% of control (Figure 3E), and this number
was only marginally increased by the addition of anti-CD122
antibody (8.2% ± 1.9% of control, Figure 3E).
engraftment and homing of AML cells in NOD/SCID mice is
activity caused by NK and/or other CD122+-dependent cells,
and specific inhibitory effects of 7G3 on AML-LSC homing and
7G3 Reduces AML Burden in NOD/SCID Mice
Several in vivo treatment strategies were adopted to determine
whether direct injection of 7G3 into NOD/SCID mice affected
AML engraftment: (1) administering 7G3 to the mice 6 hr before
cell transplantation almost completely ablated AML-1 engraft-
ment in mouse BM to 1.3% ± 0.9% of IgG2a control at 5 weeks
posttransplantation (Figure 4A); (2) initiating 7G3 treatment at
24 hr posttransplantation, to allow for LSC homing, also reduced
the engraftment of two of three AML samples at 5 weeks post-
transplantation (Figure 4B), indicating that early administration
of 7G3, when the leukemic burden is low, can efficiently impair
the engraftment of AML cells in NOD/SCID mice; (3) commen-
cing 7G3 or IgG2a administration 28 days posttransplantation,
in an established disease model, and continuing treatment until
time of sacrifice, a significant reduction in the BM burden of
AML was seen in two of five samples, likely reflective of the
heterogeneity of AML seen clinically. AML-2 responded to 7G3
with reductions in BM engraftment at 9 and 14 weeks posttrans-
(Figure 4D). Moreover, while some AML samples did not have
of 7G3 treatment at either 4 or 28 days posttransplantation,
a significant reduction in AML burden in the liver and spleen, but
not the peripheral blood, was observed (Figures 4E–4G).
To further assess the clinical potential of a CD123-targeting
mAb, 7G3 treatments were commenced in mice at 35 days post-
transplantation with NBM cells and caused no significant reduc-
tion in BM infiltration when administered for 8 days or continu-
ously for 5 weeks (Figure 4D and data not shown). Moreover,
7G3 caused no significant impairment of multilineage engraft-
ment of normal cells (data not shown). Together, these data
suggest that 7G3 is biologically active in vivo and can repress
the growth of AML with lesser effects on normal human hemato-
Since murine NOD/SCID cells do not bind 7G3, we carried out
preclinical toxicity studies in a more relevant large animal model,
the cynomolgus monkey. A chimeric variant of mAb 7G3 was
engineered that maintains CD123 binding specificity and
neutralization activity reformatted with a human IgG1Fc region.
This model permits evaluation of any effect on resting hemato-
poiesis where there is a source of endogenous IL-3 and a normal
immunesystem. ThemAbwasadministered byi.v.infusion once
Figure 2. Inhibition of AML-LSC Homing Contributes to the Inhibitory Efficacy of 7G3
(A) Homing efficiency of AML-9 cells to the BM and spleen of mice following ex vivo 7G3 treatment from two separate experiments. n = 3–6 per group.
(B) Engraftment of ex vivo antibody-treated AML-8-rel cells in the injected femur (RF) and whole BM (WBM) after i.v. (IV) or intrafemoral (IF) transplantation. n =
4–5 mice per group.
(C) Absolute number of CD34+CD38?AML cells homed in the BM and spleen of NOD/SCID mice injected with ex vivo 7G3-treated leukemic cells. n = 2–3 or
5 mice per group for AML-8 and AML-9, respectively.
(D) Homing efficiency of sorted CD34+CD38?AML-9 cells after ex vivo treatment into both BM and spleen of mice. n = 3 mice per group.
(E) The number of CD34+CD38?cells in the AML graft of mouse BM transplanted with AML-1, -5, and -9 after ex vivo IgG2a or 7G3 treatment. Each symbol
represents a single mouse; horizontal bars indicate the mean. Error bars represent mean ± SEM; *p < 0.05, **p < 0.01, and ***p % 0.0001 between IgG2a
and 7G3 groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 35
weekly for 4 consecutive weeks at 0, 10, 30, and 100 mg/kg to
a total of 32 cynomolgus monkeys (16 males and 16 females).
CD123 binding by the chimeric variant was confirmed to be
equivalent to the original parent 7G3 mouse mAb, and binding
of both MAbs to cynomolgus CD123 was also demonstrated.
There were no antibody-related effects on clinical observations
nor on a comprehensive list of hematological parameters
measured over 70 days after the first antibody treatment (data
not shown). Overall, these data indicate that a CD123-targeting
antibody does not exert adverse effects on normal hematopoi-
esis and are consistent with our NOD/SCID mouse experiments
demonstrating that 7G3 treatment can specifically inhibit AML
In Vivo Treatment with 7G3 Targets AML-LSCs
are targeted, serial transplantation was performed following
in vivo 7G3 treatment. While 10 weeks of 7G3 treatment did not
overtly decrease the engraftment of AML-10 in the BM or spleen
of primary engrafted mice (Figure 5A), the AML cells harvested
from 7G3-treated mice had significantly impaired homing ability
to the BM and spleens of secondary recipient mice compared
with IgG2a-treated controls (Figure 5B). The repopulation ability
was also significantly impaired: while eight of nine secondary
recipient mice transplanted with untreated control cells were
engrafted, only three of eight mice inoculated with cells from
7G3-treated mice showed evidence of engraftment in the BM
the proportion of CD34+CD38?primitive cells in the BM
(Figure 5D). Similar results were obtained in an independent
experiment with AML-9 cells (Figure S4). In addition, when anti-
body treatment was combined with a suboptimal dose of cytara-
another independent AML sample (AML-10), 7G3 again caused
a marked reduction in the proportion of secondary mice
engrafted (Figure S5). Collectively over three experiments, 26 of
27 (96%) secondary mice showed evidence of engraftment by
cells harvested from IgG2a-treated mice, while only 12 of 23
(52%) were engrafted by cells from 7G3-treated mice. These
results demonstrate that in vivo 7G3 administration specifically
targets AML-LSCs in NOD/SCID mice, resulting in decreased
homing and engraftment in secondary recipients.
7G3 Inhibits Spontaneous and IL-3-Induced
Proliferation of Primitive AML Cells In Vitro
Due to the lack of cross-reactivity between the human and
mouse IL-3 and CD123 systems, the ability of 7G3 to eliminate
LSCs through targeting IL-3 signaling pathway is unable to be
directly tested. To determine whether blocking IL-3 signaling
can be one of the 7G3 inhibitory functions on AML-LSCs, we
incubated different subtypes of primary AML cells with 7G3 or
IgG2a in the medium containing IL-3. 7G3 inhibited exogenously
added IL-3-induced proliferation in 32 of 35 primary AML
samples (Figure 6A). Interestingly, 7G3 inhibited the growth of
cells in nine of the samples to 50%–75% of control in the
absence of exogenous IL-3, suggesting that these samples
may possess an autocrine/paracrine IL-3 pathway or alternate
growth mechanisms that can be blocked by 7G3. This profound
Figure 3. Fc Region of the Antibody and Innate Immunity Mediate 7G3 Antileukemic Effects
(A) Homing efficiency of AML-8-rel cells to the BM following ex vivo treatment with IgG2a, 7G3, 7G3 F(ab0)2, 6H6, 6H6 F(ab0)2, or 9F5. n = 3 per group.
(B) Engraftment of AML-9 and -10 in the BM of mice following ex vivo IgG2a, 7G3, 7G3 F(ab0)2, 9F5, or 9F5 F(ab0)2 treatment. n = 5 per group.
(C and D) 7G3-mediated inhibition of AML-9 engraftment (C) and homing efficiency (D) was attenuated in mice depleted of CD122+cells (+). n = 3 per group.
(E) Numbers of CD34+CD38?AML-9 cells homed to the BM of irradiated NOD/SCID mice with (+) or without (?) CD122+cell depletion. n = 3 per group. Data are
representative of results obtained with two AML samples. Error bars represent mean ± SEM; *p < 0.05, **p < 0.01 between selected groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
36 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.
inhibition by 7G3 was IL-3 specific since 7G3 had no effect on
GM-CSF-induced cell proliferation (Figure 6B). In order to more
directly link the 7G3-mediated reduction in proliferation to prim-
itive AML cells, we demonstrated that 7G3 was able to signifi-
cantly reduce IL-3-mediated survival of CD34+CD38?CD123+
tion, 7G3 significantly reduced the survival of CD34+CD38?cells
fromtwo samples (AML-14 and AML-15) in the absence of exog-
enously added IL-3. These data verify that 7G3 inhibits IL-3-
and CD34+CD38?CD123+cell survival through binding to
7G3 Blocks IL-3-Mediated Signaling in AML Cells
We next tested whether 7G3 inhibited leukemic cell growth by
The IL-3R bcchain (CD131) was found to be coexpressed with
CD123 on CD34+primary AML cells measured by both flow
cytometry and PCR analyses (data not shown). Furthermore,
IL-3-induced CD131 activation in primary AML cells and TF-1
assessed by tyrosine phosphorylation was inhibited by 7G3 in
a concentration-dependent manner (Figure 6D and Figure S6,
respectively). Inhibition of downstream STAT5 phosphorylation
was also observed in TF1, bulk, and CD34+CD38?AML cells
(Figure 6E), as well as inhibition of both STAT5 and Akt phos-
clones, 9F5 and 6H6 (Sun et al., 1996), were ineffective at inhib-
iting IL-3-mediated proliferation (data not shown) or signaling
(Figure S6), demonstrating that different CD123 epitopes are
functionally distinct. Collectively, these in vitro studies establish
that 7G3 has the potential to also target LSCs by blocking IL-3-
mediated signaling. Thus, in a clinical context, CD123 targeting
has the potential to deliver antileukemic effects via activation
of host immunity and inhibition of the IL-3 pathway.
In this report, we show that AML-LSCs can be targeted with the
CD123-specific 7G3 mAb, resulting in impaired human AML cell
engraftment and proliferation in NOD/SCID mice and improved
long-term survival. The mechanism of LSC impairment by 7G3
treatment in the NOD/SCID model appeared complex and multi-
factorial, involving inhibition of LSC homing to the BM niche, and
stimulation of residual innate immunity in NOD/SCID recipients.
Although the consequences of blocking huIL-3 signaling with
7G3 in LSCs cannot be fully assessed in NOD/SCID mice, the
in vitro data we generated showed marked impairment of the
signaling, survival, and proliferation of primitive CD34+CD38?
AML cells. Since this subpopulation is highly enriched for
LSCs, this result strongly suggests that impairment of IL-3
signaling will also be part of the multifactorial mechanism of
action of 7G3 in a human context. Collectively, our results
Figure 4. The Schedule of 7G3 Administration Influences Its Antileukemic Efficacy in NOD/SCID Mice
(A) Engraftment levels of AML-1 cells in the BM of mice treated with a single dose of IgG2a or 7G3 (300 mg) 6 hr prior to transplantation.
(B) Percentage of AML cells in the BM of mice (n = 5–6 per group) when the treatment was commenced 24 hr posttransplantation for four doses.
(C) Engraftment levels of AML-2 in the BM of mice when treatment was initiated at day 28 posttransplantation for 9 weeks’ duration. n = 3–5 per group for each
(D) Percentage of human AML-1 or NBM cells in the BM of mice after four doses of IgG2a or 7G3 starting on day 28 (AML-1) or 35 (NBM) posttransplantation.
(E–G) Assessment of leukemia infiltration in the liver (E), spleen (F), and peripheral blood (G) for the experiments in which initiating 7G3 treatments at 4 or 28 days
did not cause a significant reduction in leukemic burden in the BM. Data are collected from six experiments for the different organs with n = 6–53 mice per group.
Each symbol represents data from an individual mouse as a percent of average control for each experiment. Horizontal bars indicate the median in (E)–(G) and
mean in (A) and (D). Otherwise results were expressed as mean ± SEM; *p < 0.05, **p < 0.01 between IgG2a and 7G3 groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 37
demonstrate that CD123 is an important marker for the targeting
of LSCs and downstream progenitors that are capable of rapid
proliferation. Our studies also show that, while the NOD/SCID
mouse strain is immune deficient due to depleted T, B, and NK
ical testing of antibody-mediated immunotherapy.
Targeting LSCs by means of the 7G3 antibody against CD123
is an attractive approach, since (1) this receptor has been widely
shown to be selectively overexpressed in LSCs; (2) the IL-3R
classically stimulates multiple biological functions; and (3) 7G3
ating ADCC by effector cells providing additional and specific
efficacy against leukemic cells, which a small molecule inhibitor
of downstream signaling (e.g., JAK/STAT) may not be able to
provide. Initial in vitro characterization showed that 7G3 robustly
impaired IL-3 binding to its receptor in a broad panel of AML
samples, thereby preventing IL-3-dependent CD131 tyrosine
phosphorylation and downstream signaling, which are required
to promote both cell survival and proliferation (Guthridge et al.,
2000, 2006). Furthermore, inhibition of AML proliferation by
7G3 in the absence of exogenous IL-3 in 9 of 32 samples
suggests that there is autocrine or paracrine secretion of IL-3
in some AML samples at physiologically significant levels. These
experimental data are consistent with other reports demon-
strating the expression of IL-3 mRNA and protein in primary
AML samples (Guan et al., 2003; Nowak et al., 1999), as well
as elevated serum IL-3 levels associated with leukemic burden
in AML patients (Elbaz and Shaltout, 2001). By contrast, normal
CD34+CD38?cell proliferative potential is not affected by IL-3
(De Bruyn et al., 2000), and lineage-negative NBM cells did not
have detectable IL-3 mRNA expression (Guan et al., 2003), sug-
poiesis in IL-3-deficient mice (Nishinakamura et al., 1996). In
AML,the level of CD123expression and responsiveness to cyto-
kines including IL-3 have been associated with poor prognosis
(Graf et al., 2004; Testa et al., 2002, 2004; Tsuzuki et al., 1997).
Thus, 7G3 inhibition of the CD123 signaling pathway in the
context of AML patients, many of whom are likely to express
high levels of circulating IL-3 (Elbaz and Shaltout, 2001), may
provide significant additional benefit beyond the mechanisms
we have already uncovered with the NOD/SCID model.
cells clearly contribute to the action of 7G3. Depletion of innate
immunity in NOD/SCID mice with anti-CD122 mAb significantly,
but not completely, attenuated the inhibitory effects of 7G3 on
the homing and repopulating abilities of AML-LSCs. Additional
evidence supporting this mechanism of 7G3-mediated inhibition
of LSC function includes evidence that the Fc portion of 7G3 is
critical for its activity, as well as the reduced potency of 7G3 in
a NOD/SCID strain without NK cell activity. These findings
support the further modification of 7G3 to enhance ADCC
Our experiments provide two key findings that support the
development of MAbs targeting CD123 as a novel therapy for
AML. First, combined with those from other groups, our data
showed that CD123 was highly expressed on the surface of
CD34+CD38?populations enriched for AML-LSCs compared
Figure 5. 7G3 Inhibits Self-Renewal Ability of AML-LSCs
(A) Engraftment levels of AML-10 cells in BM and spleen after 10 weeks of 7G3 or IgG2a treatment. The schedule of antibody treatment is shown in the schematic
(B–D) (B) Homing efficiency, (C) levels of engraftment in the BM and spleen, and (D) the percentage of CD34+CD38?cells in the BM of secondary recipient mice.
Mice in (C) and (D) were analyzed at 12 weeks posttransplantation. Each symbol represents a single mouse, and horizontal bars indicate the mean value. *p <
0.05, **p < 0.01 between the two groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
38 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.
to their normal hematopoietic counterparts from both newborn
CB and adult BM. Reduction of AML engraftment by ex vivo
7G3 treatment with less effect on normal HSCs, in comparison
with the nonspecific ablation of both normal and AML sample
engraftment by the antibody against HLA-A,B,C epitope, is
consistent with the CD123 expression data. Similarly, in vivo
7G3 treatment appears to preferentially reduce AML engraft-
of engraftment in secondary recipients demonstrates that 7G3
treatment targets the AML-LSCs in vivo, impairing LSC homing
and reducing the repopulation of secondary recipients similar
to the data from ex vivo treatment. Overall, this establishes
7G3 as a compelling LSC therapeutic in this preclinical model.
Interestingly, the IF injection method established that at least
a part of the action of 7G3 on LSC homing occurred during lodg-
ment in microenvironmental niches and not during circulation
through the blood, or during extravasation across endothelial
membranes, since this method directly bypasses the latter
processes by delivering cells to the femoral cavity.
The clinical potential for a CD123-targeting mAb is supported
bythree lines of evidence. First, our study has shownthat ex vivo
or in vivo 7G3 treatments selectively target AML cells compared
with their normal counterparts. Second, toxicity testing in pri-
mates has shown that a chimeric IgG1 variant of 7G3 had no
significant effects on any measured hematological parameters
over 70 days. If normal hematopoiesis or HSCs had been
within this time frame. Third, the same chimeric variant mAb is
being investigated in a phase I clinical trial as weekly treatment
of patients with relapsed or refractory or high-risk AML. To
date, with a total of >180 infusions administered to 26 patients
comprising five dose-level cohorts up to 10 mg/kg, no signal of
Figure 6. mAb 7G3 Inhibits Proliferation of Primary AML Cells
(A) Inhibition of primary AML cell proliferation by 7G3. Each line represents an individual AML sample exposed to the three different conditions. n = 35.
(B) Concentration-dependent effects of 7G3 on the proliferation of a primary AML sample induced by GM-CSF (0.1 ng/ml) or IL-3 (1 ng/ml). Data are represen-
tative of results obtained with 21 different AML samples.
(C) 7G3 inhibits IL-3-mediated survival of isolated CD34+CD38?CD123+primary AML cells. The percentage of surviving cells is shown.
(D) Western blot showing that 7G3 inhibits IL-3-induced CD131 tyrosine phosphorylation in a dose-dependent manner. n = 2 AML samples.
(E) 7G3 inhibits IL-3-induced phosphorylation of STAT5 in TF-1, primary AML, and sorted CD34+CD38?AML cells shown by representative histograms of intra-
cellular FACS (red, no IL-3; blue, IL-3 with IgG2a; green, IL-3 with 7G3); bar graphs represent the cumulative data in sorted CD34+CD38?cells. Mean ± SEM in
triplicates (B) or duplicates (C and E); *p < 0.05, **p < 0.01, ***p % 0.0001 between indicated groups.
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 39
treatment-related toxicity has been detected from hematology,
biochemistry, or vital signs. Other than two mild infusion reac-
tions, only one serious adverse event, an infection, was consid-
ered possibly related to treatment with the mAb. The incidence
unrelated adverse events have been consistent with complica-
tions and risks of AML (A.W. Roberts et al., 2008, ASH Annual
Although 7G3 effectively targeted AML-LSCs, it was most
successful under conditions where the leukemic burden was
low. When 7G3 treatment began at 4 weeks posttransplantation,
BM engraftment was significantly impaired in only two of five
primary AMLs. However, in this model of established AML,
7G3 distinctly reduced the AML burden in peripheral hematopoi-
etic tissues (spleen, liver) in the majority of the samples we
tested, perhaps due to a greater access of innate immune cells
eliminating 7G3-coated AML cells (Fujii et al., 2007). While
sample-to-sample variability was encountered at high leukemic
burden, the increased effectiveness of 7G3 at low leukemic
burden suggests a potential application of anti-CD123 treatment
during remission following treatment with conventional chemo-
The concept of antibody targeting of malignancy is well estab-
lished. For example, several MAbs directed at hematological
malignancies have been evaluated in clinical trials, including
rituximab (which targets CD20) and epratuzumab (CD22) in B
cell malignancies, alemtuzumab (CD52) in chronic lymphocytic
leukemia, daclizumab (CD25) in T cell malignancies, and gemtu-
zumab ozogamicin (CD33) in AML. However, these mAb thera-
pies are unlikely to target CSCs, and, while impressive cytore-
duction and clinical responses have been observed, none are
curative. Therefore, the multifaceted properties of 7G3 shown
in this preclinical model of AML support a broader proposal for
CSC-targeted cancer drug development (Wang, 2007) in which
potential therapies that target key traits of CSCs are identified
and tested using primary patient samples in relevant in vivo xen-
otransplantation models. The ongoing clinical evaluation of
a chimeric CD123 mAb in advanced AML (http://clinicaltrials.
gov/ct2/show/NCT00401739?term=CSL360&rank=1) will be the
first of its kind to test whether the significant activity of an LSC-
targeted mAb therapy in the xenograft models shown in this
study translates into a clinical benefit for patients. Ultimately,
this clinical testing will also provide more definitive proof of
a role for IL-3 in the pathology of AML.
AML Patient Samples, Normal Hematopoietic Cells, and Cell Lines
Patient samples were collected after informed consent according to institu-
tional guidelines, and studies were approved by the Royal Adelaide Hospital
Human Ethics Committee, Melbourne Health Human Research Ethics
Committee, Research Ethics Board of the University Health Network, and
the South Eastern Sydney and Illawarra Area Health Service Human Research
Ethics Committee. Diagnosis was made using cytomorphology, cytogenetics,
and leukocyte antigen expression and evaluated according to the French-
American-British (FAB) classification. Mononuclear cells were enriched by
Lymphoprep (Axis-Shield PLC, Dundee, Scotland) or Ficoll (GE Healthcare,
Uppsala, Sweden) density gradient separation and frozen in liquid nitrogen.
Human CB and NBM cells were obtained from full-term deliveries or consent-
ing patients receiving hipreplacement surgery, orcommercially fromCambrex
Corporation (East Rutherford, NJ) and Lonza (Basel, Switzerland), respec-
tively, and processed as previously described (Mazurier et al., 2003).
Ex Vivo Antibody Treatment
Thawed AML or normal hematopoietic cells were incubated with IgG2a, 7G3,
7G3 F(ab0)2, 9F5, 9F5 F(ab0)2, 6H6, or 6H6 F(ab0)2 (10 mg/ml) for 2 hr in X-VIVO
10 medium (Cambrex Corporation or Lonza) supplemented with 15%–20%
bovine serum albumin-insulin-transferrin (StemCell TechnologiesInc, Vancou-
ver, Canada) at 37?C before i.v. transplantation into sublethally irradiated
NOD/SCID mice for repopulating assays. Engraftment was measured at
4–10 weeks at two different time points.
Xenotransplantation of Human Cells into NOD/SCID Mice
and In Vivo Antibody Treatment
Animal studies were performed under the institutional guidelines approved by
the University Health Network/Princess Margaret Hospital Animal Care
Committee or the Animal Care and Ethics Committee of the University of
New South Wales. Transplantation of human cells into NOD/SCID mice was
performed as previously described (Mazurier et al., 2003). Briefly, all mice
received sublethal irradiation 24 hr before i.v. or IF transplantation with either
5–10 3 106human AML cells, 3 3 105lineage-depleted CD34+CB cells, 8 3
106BM cells, or 1 3 106sorted CD34+BM cells per mouse. Anti-CD122
antibody purified from the hybridoma cell line TM-b1 (generously provided
by Professor T. Tanaka, Hyogo University of Health Sciences) (Tanaka
et al., 1993) was injected intraperitoneally (i.p.) immediately after irradiation
(200 mg/mouse). Engraftment levels of human AML and normal hematopoietic
cells were evaluated by the percentage of huCD45+cells by flow cytometry
(Lock et al., 2002). The number of CD34+CD38?AML cells in the BM and
spleen was also calculated based on the average number of cells harvested,
and the engraftment levels and percent of CD34+CD38?AML cells in each
mouse. To measure effects on LSC activity, secondary transplantations
were also performed by i.v. transplantation of 7–10 3 106AML cells isolated
from the BM (two femurs and two tibias) of IgG2a- or 7G3-treated primary
mice into secondary recipient mice.
For in vivo testing, IgG2a or 7G3 (300 mg per injection) was injected i.p. into
mice three times a week with schedules described in the legends to each
therapeutic reagent Ara-C as described in Figure S5.
In Vivo Homing Assay
Homing assays were performed on ex vivo 7G3-treated cells, sorted
CD34+CD38?cellsfrom primarypatient samples,or cellsharvested from previ-
vested from BM and spleen of mice transplanted 16 hr previously were stained
etry for human cells using 5 3 104to 5 3 106collected events. Homing effi-
ciency of human cells into the mouse tissues was calculated based on the
number of total huCD45+cells in the tissue and the number of cells injected.
Cell Staining, Sorting, and Flow Cytometry
For flow cytometry, cells were stained as previously described (Bonnet and
Dick, 1997; Lock et al., 2002) with conjugated anti-human antibodies against
CD15, CD14, CD19, CD33, CD34, CD38, and CD45 (BD Biosciences, or Bio-
Legend, CA). CD123 expression was measured with anti-CD123 clone 9F5,
and RFI was determined by the ratio of the geometric mean of the 9F5-stained
signal to matched isotype control. Stained cells wereanalyzed using FACScan
or FACSCalibur flow cytometers (BD Biosciences). For sorting, cells were
stainedwithanti-humanantibodies against CD34,CD38,and CD123,and pro-
using Moflo and BD Aria cell sorters (BD Biosciences).
Survival Analysis of CD34+CD38?CD123+AML Cells
Sorted cells plated at 1.5 3 105cells/ml in IMDM/0.5% FCS were treated with
150 mg/ml 7G3 or IgG2a (clone BM4) for 30 min prior to addition of 1 nM IL-3.
Cells were analyzed for survival at 48 and 72 hr by staining with 1:100 Annexin
V-FLUOS (Roche, Basel, Switzerland) as described previously (Guthridge
et al., 2006). Absolute cell number was also assessed by addition of 50 ml
Flow-Count fluorospheres (Beckman Coulter).
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
40 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.
AML cell-growth responses to IL-3 or GM-CSF were measured by3H-thymi-
dine assay as previously described (Lopez et al., 1988). Briefly, 2 3 104mono-
nuclear cells per well in 96-well plates were stimulated with IL-3 (1 ng/ml) or
GM-CSF (0.1 ng/ml) in the presence of 0–10 nM 7G3 or IgG2a in 200 ml
IMDM + 10% HI-FCS (Hyclone, UT) for 48 hr with 0.5 mCi of3H-thymidine
(MP Biomedicals Australasia, Sydney, Australia) added for the last 6 hr of
cell harvester (PerkinElmer Life and Analytical Sciences, Melbourne, Australia)
and counted using a Top Count (PerkinElmer). All cytokines were supplied by
R&D Systems (MN).
Cytokine Signaling Assays
Phosphorylation of signaling proteins was detected by immunoprecipitation
and immunoblots. TF-1 and primary AML cells were washed and rendered
quiescent overnight before incubation with IgG2a, 9F5, 6H6, or 7G3
(0–100 nM) for 20 min on ice. Cells were then stimulated with 1 nM IL-3 for
10 min at 37?C. Cells were lysed in NP-40 lysis buffer, and CD131 was immu-
noprecipitated using 1C1 and 8E4 antibodies conjugated to Sepharose beads
(Guthridge et al., 2004). Immunoprecipitates were subjected to SDS-PAGE
and immunoblotting as previously described (Guthridge et al., 2004). Anti-
bodies used were the following: antiphosphotyrosine mAb 4G10 (Upstate
Biotechnology Inc, NY), anti-phospho-Akt Ser473 (Cell Signaling Technology
Inc, MA), and, anti-phosphorylated STAT5 mAb (Zymed Laboratories Inc,
CA). All antibodies were used according to manufacturers’ instructions. Blots
were stripped and reprobed with antibody to bc (1C1) as a loading control.
For intracellular FACS, quiescent TF-1, bulk, and sorted primary AML cells
were stimulated with 20 ng/ml IL-3 plus 20 mg/ml IgG2a or 7G3 for 1 hr. Sorted
subpopulations were incubated with 150 mg/ml 7G3 or IgG2a for 30 min on ice
before stimulation with 1 nM IL-3 for 15 min. Cells were fixed with BD Cytofix
Buffer (BD Biosciences), methanol permeabilized, and stained with anti-phos-
phoSTAT5 (BDBiosciences) or isotype control. Cells werethenanalyzed using
a FACSCalibur flow cytometer (BD Biosciences).
groups was determined using the unpaired, two-sided Student’s t test, or the
nonparametric Mann-Whitney U test. Survival curves were compared using
the log rank test.
Supplemental Data include six figures and can be found with this article online
This work was supported by Children’s Cancer Institute Australia for Medical
Research and by research grants from CSL Limited (A.B., A.F.L., J.E.D., and
R.B.L.). Children’s Cancer Institute Australia for Medical Research is affiliated
with the University of New South Wales and Sydney Children’s Hospital.
H.S.R. is the Peter Nelson Leukaemia Research Fund Senior Research Fellow.
The authors wish to thank Dr. Mark Minden (University Health Network, Tor-
onto, Canada) and Naomi Sprigg, Rosemary Hoyt, and Dr. Andrew Roberts
(Royal Melbourne Hospital, Melbourne, Australia) for providing patient biopsy
specimens; Thu Tran, Soo Min Heng, and Dean Inwood (Department of Radi-
ation Oncology, Prince of Wales Hospital, Sydney, Australia) for mouse irradi-
ation; CSL Protein Group for antibody reagents; and Dr. Andrew Roberts for
helpful discussions. The authors wish to declare the following potential
conflicts of interest: employees of CSL Limited (S.J.B., D.P.G., and G.V.)
and consultants of CSL Limited (A.F.L. and R.B.L.).
Received: July 15, 2008
Revised: April 1, 2009
Accepted: April 30, 2009
Published: July 1, 2009
Abutalib, S.A., and Tallman, M.S. (2006). Monoclonal antibodies for the treat-
ment of acute myeloid leukemia. Curr. Pharm. Biotechnol. 7, 343–369.
Aribi, A., Ravandi, F., and Giles, F. (2006). Novel agents in acute myeloid
leukemia. Cancer J. 12, 77–91.
Bagley, C.J., Woodcock, J.M., Stomski, F.C., and Lopez, A.F. (1997). The
structural and functional basis of cytokine receptor activation: lessons from
the common beta subunit of the granulocyte-macrophage colony-stimulating
factor, interleukin-3 (IL-3), and IL-5 receptors. Blood 89, 1471–1482.
Bhatia, M.,Wang,J.C., Kapp,U.,Bonnet,D.,and Dick,J.E.(1997).Purification
of primitive human hematopoietic cells capable of repopulating immune-defi-
cient mice. Proc. Natl. Acad. Sci. USA 94, 5320–5325.
Bonnet, D., and Dick, J.E. (1997). Human acute myeloid leukemia is organized
as a hierarchy that originates from a primitive hematopoietic cell. Nat. Med.
Budel, L.M., Touw, I.P., Delwel, R., Clark, S.C., and Lowenberg, B. (1989).
Interleukin-3 and granulocyte-monocyte colony-stimulating factor receptors
on human acute myelocytic leukemia cells and relationship to the proliferative
response. Blood 74, 565–571.
De Bruyn, C., Delforge, A., Lagneaux, L., and Bron, D. (2000). Characterization
of CD34+ subsets derived from bone marrow, umbilical cord blood and mobi-
lized peripheral blood after stem cell factor and interleukin 3 stimulation. Bone
Marrow Transplant. 25, 377–383.
Elbaz, O., and Shaltout, A. (2001). Implication of granulocyte-macrophage
colony stimulating factor (GM-CSF) and interleukin-3 (IL-3) in children with
acute myeloid leukaemia (AML); malignancy. Hematology 5, 383–388.
Estey, E., and Dohner, H. (2006). Acute myeloid leukaemia. Lancet 368, 1894–
Florian, S., Sonneck, K., Hauswirth, A.W., Krauth, M.T., Schernthaner, G.H.,
Sperr, W.R., and Valent, P. (2006). Detection of molecular targets on the
surface of CD34+/CD38– stem cells in various myeloid malignancies. Leuk.
Lymphoma 47, 207–222.
Fujii, H., Trudeau, J.D., Teachey, D.T., Fish, J.D., Grupp, S.A., Schultz, K.R.,
and Reid, G.S. (2007). In vivo control of acute lymphoblastic leukemia by
immunostimulatory CpG oligonucleotides. Blood 109, 2008–2013.
Graf, M., Hecht, K., Reif, S., Pelka-Fleischer, R., Pfister, K., and Schmetzer, H.
(2004). Expression and prognostic value of hemopoietic cytokine receptors in
acute myeloid leukemia (AML): implications for future therapeutical strategies.
Eur. J. Haematol. 72, 89–106.
Guan, Y., and Hogge, D.E. (2000). Proliferative status of primitive hematopoi-
etic progenitors from patients with acute myelogenous leukemia (AML).
Leukemia 14, 2135–2141.
Guan, Y., Gerhard, B., and Hogge, D.E. (2003). Detection, isolation, and stim-
ulation of quiescent primitive leukemic progenitor cells from patients with
acute myeloid leukemia (AML). Blood 101, 3142–3149.
Guthridge, M.A., Stomski, F.C., Barry, E.F., Winnall, W., Woodcock, J.M.,
McClure, B.J., Dottore, M., Berndt, M.C., and Lopez, A.F. (2000). Site-specific
serine phosphorylation of the IL-3 receptor is required for hemopoietic cell
survival. Mol. Cell 6, 99–108.
Guthridge, M.A., Barry, E.F., Felquer, F.A., McClure, B.J., Stomski, F.C., Ram-
of the GM-CSF/IL-3/IL-5 receptors mediates hematopoietic cell survival
through activation of NF-kappaB and induction of bcl-2. Blood 103, 820–827.
Guthridge, M.A., Powell, J.A., Barry, E.F., Stomski, F.C., McClure, B.J., Ram-
shaw, H., Felquer, F.A., Dottore, M., Thomas, D.T., To, B., et al. (2006). Growth
factor pleiotropy is controlled by a receptor Tyr/Ser motif that acts as a binary
switch. EMBO J. 25, 479–489.
Guzman, M.L.,Neering, S.J., Upchurch, D., Grimes, B., Howard, D.S., Rizzieri,
D.A., Luger, S.M., and Jordan, C.T. (2001). Nuclear factor-kappaB is constitu-
tively activated in primitive human acute myelogenous leukemia cells. Blood
Hauswirth, A.W., Florian, S., Printz, D., Sotlar, K., Krauth, M.T., Fritsch, G.,
Schernthaner, G.H., Wacheck, V., Selzer, E., Sperr, W.R., et al. (2007).
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc. 41
Expression of the target receptor CD33 in CD34+/CD38?/CD123+ AML stem
cells. Eur. J. Clin. Invest. 37, 73–82.
Hope, K.J., Jin, L., and Dick, J.E. (2004). Acute myeloid leukemia originates
from a hierarchy of leukemic stem cell classes that differ in self-renewal
capacity. Nat. Immunol. 5, 738–743.
Ishikawa, F., Yoshida, S., Saito, Y., Hijikata, A., Kitamura, H., Tanaka, S.,
Nakamura, R., Tanaka, T., Tomiyama, H., Saito, N., et al. (2007). Chemo-
therapy-resistant human AML stem cells home to and engraft within the
bone-marrow endosteal region. Nat. Biotechnol. 25, 1315–1321.
Ito, M., Hiramatsu, H., Kobayashi, K., Suzue, K., Kawahata, M., Hioki, K.,
Ueyama, Y., Koyanagi, Y., Sugamura, K., Tsuji, K., et al. (2002). NOD/SCID/
gamma(c)(null) mouse: an excellent recipient mouse model for engraftment
of human cells. Blood 100, 3175–3182.
Jin, L., Hope, K.J., Zhai, Q., Smadja-Joffe, F., and Dick, J.E. (2006). Targeting
of CD44 eradicates human acute myeloid leukemic stem cells. Nat. Med. 12,
Jordan, C.T., Upchurch, D., Szilvassy, S.J., Guzman, M.L., Howard, D.S., Pet-
tigrew, A.L., Meyerrose, T., Rossi, R., Grimes, B., Rizzieri, D.A., et al. (2000).
The interleukin-3 receptor alpha chain is a unique marker for human acute
myelogenous leukemia stem cells. Leukemia 14, 1777–1784.
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 initi-
ating human acute myeloid leukaemia after transplantation into SCID mice.
Nature 367, 645–648.
Lock, R.B., Liem, N., Farnsworth, M.L., Milross, C.G., Xue, C., Tajbakhsh, M.,
Haber, M., Norris, M.D., Marshall, G.M., and Rice, A.M. (2002). The nonobese
diabetic/severe combined immunodeficient (NOD/SCID) mouse model of
childhood acute lymphoblastic leukemia reveals intrinsic differences in bio-
logic characteristics at diagnosis and relapse. Blood 99, 4100–4108.
Lopez, A.F., Dyson, P.G., To, L.B., Elliott, M.J., Milton, S.E., Russell, J.A., Jutt-
ner, C.A., Yang, Y.C., Clark, S.C., and Vadas, M.A. (1988). Recombinant
human interleukin-3 stimulation of hematopoiesis in humans: loss of respon-
siveness with differentiation in the neutrophilic myeloid series. Blood 72,
Mazurier, F., Doedens, M., Gan, O.I., and Dick, J.E. (2003). Rapid myeloeryth-
roid repopulation after intrafemoral transplantation of NOD-SCID mice reveals
a new class of human stem cells. Nat. Med. 9, 959–963.
McKenzie, J.L., Gan, O.I., Doedens, M., and Dick, J.E. (2005). Human short-
term repopulating stem cells are efficiently detected following intrafemoral
transplantation into NOD/SCID recipients depleted of CD122+ cells. Blood
vidual stem cells with highly variable proliferation and self-renewal properties
comprise the human hematopoietic stem cell compartment. Nat. Immunol. 7,
Miyajima, A., Mui, A.L., Ogorochi, T., and Sakamaki, K. (1993). Receptors for
granulocyte-macrophage colony-stimulating factor, interleukin-3, and inter-
leukin-5. Blood 82, 1960–1974.
Miyauchi, J., Kelleher, C.A., Yang, Y.C., Wong, G.G., Clark, S.C., Minden,
M.D., Minkin, S., and McCulloch, E.A. (1987). The effects of three recombinant
growth factors, IL-3, GM-CSF, and G-CSF, on the blast cells of acute myelo-
blastic leukemia maintained in short-term suspension culture. Blood 70,
Morgan, M.A.,andReuter, C.W.(2006).Molecularlytargeted therapies inmye-
lodysplastic syndromes and acute myeloid leukemias. Ann. Hematol. 85,
Munoz, L., Nomdedeu, J.F., Lopez, O., Carnicer, M.J., Bellido, M., Aventin, A.,
Brunet, S., and Sierra, J. (2001). Interleukin-3 receptor alpha chain (CD123) is
widely expressed in hematologic malignancies. Haematologica 86, 1261–
Nishinakamura, R., Miyajima, A., Mee, P.J., Tybulewicz, V.L., and Murray, R.
(1996). Hematopoiesis in mice lacking the entire granulocyte-macrophage
colony-stimulating factor/interleukin-3/interleukin-5 functions. Blood 88,
Nowak, R., Oelschlagel, U., Gurth, H., Range, U., Albrecht, S., Krebs, U.,
Hietschold,V.,andEhninger,G.(1999).Relations between IL-3-inducedprolif-
eration and invitro cytokinesecretion of bone marrow cells from AMLpatients.
Cytokine 11, 435–442.
Pearce, D.J., Taussig, D., Zibara, K., Smith, L.L., Ridler, C.M., Preudhomme,
C., Young, B.D., Rohatiner, A.Z., Lister, T.A., and Bonnet, D. (2006). AML
engraftment inthe NOD/SCID assay reflects the outcome of AML:implications
for our understanding of the heterogeneity of AML. Blood 107, 1166–1173.
Pebusque, M.J., Fay, C., Lafage, M., Sempere, C., Saeland, S., Caux, C., and
Mannoni, P. (1989). Recombinant human IL-3 and G-CSF act synergistically in
stimulating the growth of acute myeloid leukemia cells. Leukemia 3, 200–205.
Riccioni, R., Rossini, A., Calabro, L., Diverio, D., Pasquini, L., Lococo, F.,
Peschle, C., and Testa, U. (2004). Immunophenotypic features of acute
myeloid leukemias overexpressing the interleukin 3 receptor alpha chain.
Leuk. Lymphoma 45, 1511–1517.
Cell 1, 353–355.
Sperr, W.R., Hauswirth, A.W., Florian, S., Ohler, L., Geissler, K., and Valent, P.
(2004). Human leukaemic stem cells: a novel target of therapy. Eur. J. Clin.
Invest. 34 (Suppl 2), 31–40.
Stone, R.M. (2007). Novel therapeutic agents in acute myeloid leukemia. Exp.
Hematol. 35, 163–166.
Sun, Q., Woodcock, J.M., Rapoport, A., Stomski, F.C., Korpelainen, E.I., Bag-
ley, C.J., Goodall, G.J., Smith, W.B., Gamble, J.R., Vadas, M.A., et al. (1996).
Monoclonal antibody 7G3 recognizes the N-terminal domain of the human
interleukin-3 (IL-3) receptor alpha-chain and functions as a specific IL-3
receptor antagonist. Blood 87, 83–92.
Tanaka, T., Kitamura, F., Nagasaka, Y., Kuida, K., Suwa, H., and Miyasaka, M.
(1993). Selective long-term elimination of natural killer cells in vivo by an anti-
interleukin 2 receptor beta chain monoclonal antibody in mice. J. Exp. Med.
Taussig, D.C., Miraki-Moud, F., Anjos-Afonso, F., Pearce, D.J., Allen, K.,
Ridler, C., Lillington, D., Oakervee, H., Cavenagh, J., Agrawal, S.G., et al.
(2008). Anti-CD38 antibody-mediated clearance of human repopulating cells
masks the heterogeneity of leukemia-initiating cells. Blood 112, 568–575.
Testa, U., Riccioni, R., Militi, S., Coccia, E., Stellacci, E., Samoggia, P., Lata-
gliata, R., Mariani, G., Rossini, A., Battistini, A., et al. (2002). Elevated expres-
sion of IL-3Ralpha in acute myelogenous leukemia is associated with
enhanced blast proliferation, increased cellularity, and poor prognosis. Blood
Testa, U., Riccioni, R., Diverio, D., Rossini, A., Lo Coco, F., and Peschle, C.
(2004). Interleukin-3 receptor in acute leukemia. Leukemia 18, 219–226.
Tsuzuki, M., Ezaki, K., Maruyama, F., Ino, T., Kojima, H., Okamoto, M., Yama-
guchi, T., Nomura, T., Miyazaki, H., Wakita, M., et al. (1997). Proliferative
effects of several hematopoietic growth factors on acute myelogenous
leukemia cells and correlation with treatment outcome. Leukemia 11, 2125–
van Rhenen, A., Feller, N., Kelder, A., Westra, A.H., Rombouts, E., Zweegman,
S., van der Pol, M.A., Waisfisz, Q., Ossenkoppele, G.J., and Schuurhuis, G.J.
(2005). High stem cell frequency in acute myeloid leukemia at diagnosis
predicts high minimal residual disease and poor survival. Clin. Cancer Res.
Vellenga, E., Ostapovicz, D., O’Rourke, B., and Griffin, J.D. (1987). Effects of
recombinant IL-3, GM-CSF, and G-CSF on proliferation of leukemic clono-
genic cells in short-term and long-term cultures. Leukemia 1, 584–589.
Wang, J.C. (2007). Evaluating therapeutic efficacy against cancer stem cells:
new challenges posed by a new paradigm. Cell Stem Cell 1, 497–501.
Wang, J.C., and Dick, J.E. (2005). Cancer stem cells: lessons from leukemia.
Trends Cell Biol. 15, 494–501.
3 receptor subunits on defined subpopulations of acute myeloid leukemia
blasts predicts the cytotoxicity of diphtheria toxin interleukin-3 fusion protein
against malignant progenitors that engraft in immunodeficient mice. Blood
Cell Stem Cell
Antibody Targeting of Leukemia Stem Cells
42 Cell Stem Cell 5, 31–42, July 2, 2009 ª2009 Elsevier Inc.