Leukemia (2011) 25, 1249–1258
& 2011 Macmillan Publishers Limited All rights reserved 0887-6924/11
www.nature.com/leu
ORIGINAL ARTICLE
Expression of CD34 and CD7 on human T-cell acute lymphoblastic leukemia
discriminates functionally heterogeneous cell populations
B Gerby1, E Clappier1,2, F Armstrong1, C Deswarte1,3, J Calvo1, S Poglio1, J Soulier4, N Boissel5, T Leblanc5, A Baruchel5,
J Landman-Parker3, PH Roméo6, P Ballerini1,7 and F Pflumio1
1
Laboratoire des Cellules Souches Hématopoı̈étiques et Leucémiques, Institut de Radiobiologie Cellulaire et Moléculaire (IRCM),
UMR967 INSERM, Université Paris 7, Université Paris 11, Commissariat à l’Energie Atomique (CEA), Fontenay-aux-Roses, France;
2
Département de Génétique, Hôpital Robert Debré, AP-HP, Paris, France; 3Service d’Hématologie et d’Oncologie Pédiatrique,
Hôpital A Trousseau AP-HP/Université Pierre et Marie CurieFParis 6, Paris, France; 4Equipe Remaniements du Génome et
Cancers, INSERM U944 et Laboratoire d’Hématologie AP-HP, Institut Universitaire d’Hématologie Université Paris Diderot,
Hôpital Saint-Louis, Paris, France; 5Services d’Hématologie Pédiatrique et d’Hématologie Adulte, Hôpital Saint-Louis et Robert
Debré, AP-HP, Paris, France; 6Laboratoire de la Régulation de la Transcription des cellules Souches, Institut de Radiobiologie
Cellulaire et Moléculaire, Inserm U967, Université Paris 7, Commissariat à l’Energie Atomique (CEA), Fontenay-aux-Roses, France
and 7Service d’Hématologie Biologique, Hôpital Trousseau AP-HP, Université Pierre et Marie Curie, Paris, France
Leukemia-initiating/repopulating cells (LICs), also named leukemic stem cells, are responsible for propagating human acute
leukemia. Although they have been characterized in various
leukemias, their role in T-cell acute lymphoblastic leukemia
(T-ALL) is unclear. To identify and characterize LICs in T-ALL
(T-LIC), we fractionated peripheral blood cell populations from
patient samples by flow cytometry into three cell fractions by
using two markers: CD34 (a marker of immature cells and LICs)
and CD7 (a marker of early T-cell differentiation). We tested
these populations in both in vitro culture assays and in vivo for
growth and leukemia development in immune-deficient mice.
We found LIC activity in CD7 þ cells only as CD34 þ CD7 cells
contained normal human progenitors and hematopoietic stem
cells that differentiated into T, B lymphoid and myeloid cells.
In contrast, CD34 þ CD7 þ cells were enriched in LICs,
when compared with CD34CD7 þ cells. These CD34 þ CD7 þ
cells also proliferated more upon NOTCH activation than
CD34CD7 þ cells and were sensitive to dexamethasone and
NOTCH inhibitors. These data show that CD34 and CD7
expression in human T-ALL samples help in discriminating
heterogeneous cell populations endowed with different LIC
activity, proliferation capacity and responses to drugs.
Leukemia (2011) 25, 1249–1258; doi:10.1038/leu.2011.93;
published online 13 May 2011
Keywords: T-ALL; leukemic-initiating cells; xenograft
Introduction
T-cell acute lymphoblastic leukemia (T-ALL) accounts for 15%
of newly diagnosed ALL in children and 25% in adults. T-ALL is
generally characterized by a high tumor burden with high
peripheral blood blasts counts and extra-medullar localizations
that represent high-risk clinical features. Although intensive
chemotherapy has markedly improved the prognosis for patients
with T-ALL, relapses still occur, often with a dismal outcome.
Different analyses have revealed that T-ALL is a highly
heterogeneous disease;1 despite considerable efforts to predict
Correspondence: Dr F Pflumio, Laboratoire des Cellules Souches
Hématopoı̈étiques et Leucémiques, UMR967 INSERM, IRCM, CEA, 18
route du panorama, Fontenay-aux-Roses 92260, France.
E-mail: [email protected]
Received 7 September 2010; revised 28 February 2011; accepted 28
March 2011; published online 13 May 2011
outcome based on biological markers, genetic or immunophenotypic criteria are still limited that correlate clearly with
prognosis.
T-ALL is thought to result from the transformation of T-cell
progenitors blocked either early or late in differentiation.2
Normal human T-cell differentiation proceeds notably through
acquisition or loss of surface markers, such as CD34 and CD7,
the former marker being progressively lost as maturation
progresses while CD7 expression persists and other markers
such as CD1a, CD4, CD8 and CD3 appear.3 Immature T-ALL
often bears CD34 and CD7 surface markers but can lack others
such as CD4, CD8 or CD3. The reverse is not always true for
cortical or mature T-ALL, as aberrant expression of surface
markers such as CD34 can be detected with CD4, CD8 and
CD3.4
The cancer stem cell hypothesis posits that the cells that make
up a cancer are functionally heterogeneous and organized in a
hierarchy that mimics normal differentiation, with a fraction of
cells endowed with self-renewal properties (that is, cancer stem
cells) at the apex of the hierarchy.5 By identifying and
characterizing these stem cells in leukemia one hope to learn
more about the heterogeneity of cancer cells, how these stem
cells initiate and propagate the disease and, in the long-term, to
develop specific therapeutic strategies aimed at eradicating the
cancer and preventing relapse.6
Whether T-ALL is organized as a hierarchy of cells, as
previously described for acute myeloid leukemia,7,8 chronic
myeloid leukemia9 and pediatric B-cell acute lymphoblastic
leukemia,10,11 is still unknown. There are few published studies
of cell populations enriched in leukemic stem cell activity in
human T-ALL, although we and others have previously shown
that T-cell leukemia-initiating cell (T-LIC) activity can be
explored by using in vivo xenografts and in vitro assays.12,13
Indeed, both these studies demonstrated that T-ALL contained a
T-LIC activity using serial transplantation experiments into
immune-deficient mice that are considered hallmarks for
studying cancer stem cells. Moreover, T-LICs were proposed
to reside into the population expressing CD34, but not CD7 or
CD4, in reference to immature T-cell progenitors.12
Here, we have investigated the T-LIC activity of distinct
cell populations in peripheral blood samples from cortical/
mature T-ALL patients by using a combination of in vitro
studies and transplantation into immune-deficient non obese
Functional heterogeneity in human T-ALL
B Gerby et al
1250
diabetic-severe combined immune deficient (also called NODSCID or NS) or NOD-SCIDgc/ (NSG) mice. We find that these
cell fractions are heterogeneous in their functional properties. In
contrast to previous reports,12 CD34 þ CD7 cells did not contain
leukemia-initiating activity but underwent normal hematopoietic
differentiation in vitro and in vivo. In fact, leukemia-initiating
activity was enriched in CD34 þ CD7 þ cells, compared with
CD34CD7 þ cells. Moreover, we demonstrate that combination
of dexamethasone, a commonly used drug in T-ALL therapy, and
inhibitors of NOTCH signaling inhibit proliferation and activity of
the stem cell-containing CD34 þ CD7 þ cell population.
Materials and methods
Mice
NOD/Ltsz-scid/scid (NS) and NOD.Cg-Prkdc(scid) Il2rg(tm1Wjll)/
SzJ (NSG) (Jackson Laboratory, Bar Harbor, ME, USA) immunedeficient mice were housed in pathogen-free animal facilities at
the CEA, Fontenay-aux-Roses, France. Adult NS and NSG mice
were irradiated at 3 Gy and anesthetized with isoflurane before
intravenous retro-orbital injection. All experimental procedures
were done in compliance with French Ministry of Agriculture
regulations (animal facility registration number: A920322) for
animal experimentation.
Human samples
All primary samples were obtained with the informed consent of
the patients in accordance with national ethics rules. White
blood cells were isolated by Ficoll centrifugation, immunophenotyped and used directly or frozen in fetal calf serum
containing 10% dimethyl sulfoxide. Normal B, T and myeloid
cells were excluded by flow cytometry based on their expression
of lymphoid CD19, CD3(hi) TCRab(hi) CD4(hi) or CD8(hi) and
myeloid CD14/CD15 markers. In most cases, T-ALL blasts
constituted 490% of cells contained in patient peripheral blood
samples. Human umbilical cord blood CD34 þ cells were
isolated by using immuno-magnetic purification (Myltenyi
Biotech, Paris, France).
(BD Biosciences, Le Pont de Claix, France). Cell sorting was
performed on a Moflow cell sorting cytometer (Triple-laser MLS
System, Dako Cytomation, Fort Collins, CO, USA).
In vitro drug sensitivity
Sorted normal and leukemic cells were cultured for a week with
MS5-DL1 stromal cells,13 in the presence of a range of
concentrations of dexamethasone (0.1–500 nM; Sigma-Aldrich,
D-4902), DAPT (0.5–5 mM) or DAPT (5 mM) þ dexamethsaone
(0.1–50 nM), or with the dimethyl sulfoxide/vehicle alone.
Cultured cells were harvested and cell cycle commitment was
measured by immunolabeling for Ki67 (BD Pharmingen,
Grenoble, France) and analyzed by flow cytometry.
Gene expression
Quantitative PCR analysis was performed as previously described with the same primers as in Armstrong et al.13 Three
measurements were performed for each sample and data were
normalized over b2m values.
T-cell receptor (TCR) gene rearrangements
Specific clonal rearrangements of TCRg and TCRd genes were
detected in leukemic cells at diagnosis according to standard
procedures.14 Briefly, genomic DNA was subjected to several
multiplex PCR assays for amplification of recombined V(D)J
gene segments. Fluorescently labeled PCR products were
analyzed by GeneScanning using an ABI PRISM 7900HT
sequence detection system (Applied Biosystems, Foster City,
CA, USA). Clonal rearrangements present in leukemic cells at
diagnosis were then searched for in sorted populations and
immune-deficient mouse-derived cells by PCR and fragment
size analysis.
Notch1 mutations
Sequencing was performed according to Weng et al.15
TLX3 and SIL-TAL1 genomic rearrangements
Culture conditions
Peripheral blood cells from T-ALL patients were cultured as
previously described.13 Briefly, the cells were grown in
complete medium containing reconstituted a-MEM supplemented with 10% fetal calf serum (6450; Stem Cell Technologies,
Vancouver, Canada) and 10% human serum (J Boy, Reims,
France), in the presence of 50 ng/ml recombinant human
stem cell factor (Amgen, Neuilly-sur-Seine, France), 20 ng/ml
rhFlt3-Ligand (Diaclone, Besançon, France), 20 nM insulin
(Sigma-Aldrich, St Quentin Fallavier, France) and 10 ng/ml
rhIL-7 (R&D System, Lille, France). g-Secretase inhibitor (GSI)
(N-[N-(3,5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl
ester; DAPT, Calbiochem, Strasbourg, France) was diluted into
dimethyl sulfoxide and added in the culture medium three times
a week at the indicated concentration.
The TLX3 chromosomal translocation was detected by fluorescence in situ hybridization on interphase nuclei using the TLX3
fluorescence in situ hybridization DNA-specific dual color
translocation probe (Dako, Trappes, France). The SIL-TAL1
genomic micro-deletion was detected by genomic PCR of the
flanking regions using a fluorescently labeled primer, followed
by fragment size analysis.
Statistical analysis
Values are presented as the mean or median±s.d. Statistical
comparisons between groups were done using the Student’s
t-test, *Po0.05; **Po0.01; ***Po0.001 were considered
statistically significant.
Results
Flow cytometry
Cells were stained with fluorescein (FITC-), phycoerythrin (PE-),
PE-cyanin 7 (PC7-) and allophycocyanin (APC-) conjugated
mouse anti-human monoclonal antibodies specific for CD45,
CD34, CD38, CD7, CD4, CD8, CD19, CD14 and CD15 (1:25
dilution; Beckman Coulter, Villepinte, France). Cell-surface
marker expression was analyzed on a FacsCalibur cytometer
Leukemia
Heterogeneous expression of CD34, CD7 and CD4
in T-ALL cells
We first aimed at measuring the proportion of cells positive for
surface markers useful for the search of T-LIC. We determined
the proportion of CD34-, CD7- and CD4-positive cells in
samples of peripheral blood mononuclear cells (MNC) from 18
Functional heterogeneity in human T-ALL
B Gerby et al
1251
Figure 1 Phenotypic and functional cell heterogeneity in human T-ALL. Analysis of CD34, CD7 and CD4 expression in mature/cortical T-ALL
patient samples in three representative samples M77, M66 and M78. Expression of CD7 and CD4 was analyzed in CD34 þ CD7 (upper panels) or
CD34 þ CD4 (lower panels) gated populations. Percent of cells is indicated under every plot.
newly diagnosed cortical/mature T-ALL patients. Normal peripheral mature CD4 þ and CD8 þ , single positive T cells, which
represented up to 5% of the mononucleated cells, were excluded
by gating out CD3(hi) TCRab(hi) CD4(hi) and CD8(hi) cells, allowing
us to characterize only immature ‘blast’ cells. The data showed
that T-ALL patient samples exhibit very variable proportions of
CD34 þ blast cells (typical samples are shown in Figure 1a). Some
of them contained very few (o1%) CD34 þ cells, whereas others
contained up to 85% CD34 þ blasts that were mostly CD7 þ
(Figure 1a; Supplementary Table 1). Interestingly, all samples
contained a small proportion of CD34 þ CD7 cells that were
CD4 (Figure 1a). When CD34 was plotted against CD4, we
found that most CD34 þ CD4 cells were CD7 þ , except in some
samples where they remained CD7 (Figure 1a). These results
indicated that T-ALL is heterogeneous in terms of cell surface
CD34, CD7 and CD4 expression.
CD34 þ CD7 cells undergo normal hematopoiesis
A previous study reported that the leukemic progenitor cell
population of T-ALL is contained in CD34 þ CD7 and
CD34 þ CD4 cells.12 As we found that T-ALL-derived
CD34 þ CD7 cells are mostly CD4 (see above), we reasoned
that the T-LIC activity would more likely to be found in the
CD34 þ CD7 cell population. Interestingly, the absolute numbers of CD34 þ CD7 cells were high (3–9 108/l) compared
with healthy subjects (2–5 107/l)16 related to the fact that TALL patients have a hyper-leukocytosis. We first co-cultured the
sorted CD34 þ CD7 cells from seven samples in contact with a
stromal cell line expressing the human NOTCH ligand DL1
(MS5-DL1), which allows human progenitor cells to differentiate
into T cells (La Motte-Mohs et al.17 and Supplementary Figure 1)
and human T-ALL to proliferate.13 In these conditions, T-ALLderived CD34 þ CD7 cells progressively expressed CD7 while
they downregulated CD34 (Figure 2a) and CD4 þ CD8 þ T cells
were detected 7 weeks later. Interestingly, CD34 þ CD7 cells
generated also myeloid cells in these conditions (Supplementary
Figures 2A and B). When the NOTCH pathway was blocked
using a GSI called DAPT, the production of CD7 þ T cells, but
not CD14 þ /CD15 þ myeloid cells, was severely impaired,
indicating that CD7 þ T-cell growth was specifically inhibited
and that DAPT was not toxic for myeloid cell develop-
ment (Supplementary Figures 2A and B). Upon culture with
control MS5 (MS5-Ctl) cells, CD34 þ CD7 cells generated a
wider range of hematopoietic cell types: CD14 þ /CD15 þ
myeloid, CD19 þ B lymphocyte and CD34 þ progenitor-like
phenotypes with only low levels of CD7 þ cells detected
compared with co-culture with MS5-DL1 cells (Figure 2a;
Supplementary Table 2).
These data suggested that T-ALL-derived CD34 þ CD7 cells
contained normal progenitors. Likewise, we observed that
human umbilical cord blood CD34 þ CD38/low cells cocultured with MS5-DL1 or MS5-Ctl cells produced similar
proportions of differentiated T, B and myeloid cells (Supplementary Figure 1). Moreover, the genetic abnormalities characteristic of the tested T-ALL samples they were derived from
were not detected in human cells generated during culture. For
instance, CD34 þ CD7-derived myeloid and T cells from the
M60 sample did not express TLX1 transcripts and those from the
M77 sample were negative for the TLX3 rearrangement
(Supplementary Figure 2C; Figure 2b). Moreover, only in 1/4
tested samples we could detect at very low level (p1/1000) the
clonal TCR rearrangement of leukemic cells in CD7 þ T cells
recovered after culture (M79, Supplementary Table 2), probably
explained by a weak contamination of the starting population
with leukemic cells. Altogether, these findings indicated that
T-ALL-derived CD34 þ CD7 cells contained normal hematopoietic progenitors capable of multi-lineage differentiation but
no leukemic cell development in vitro.
To assess the potential T-LIC activity of these cells, we
injected 2.5–5 104 sorted CD34 þ CD7 cells into NS and
NSG mice. Mice remained healthy during the entire period of
observation. Engraftment analysis showed that 12/16 NS/NSG
mice contained detectable (40.1%) levels of human CD45 þ
hematopoietic cells, but engraftment did not exceed 5% in
bone marrow (BM) and 7% in spleen (Figure 2c; Supplementary
Table 3). Phenotypic analysis showed these cells were mainly
CD19 þ B cells and CD14 þ /CD15 þ myeloid cells with only
a minority of CD7 þ T cells (Figures 2c and 3a). In all, 4/14
mice also contained human CD7 þ CD4 þ CD8 þ T cells in their
thymus (Figure 2c; Supplementary Table 3). In comparison,
CD34 þ CD7 cells from T-ALL patients generated myeloid cells
as well as B and T cells in NSG mice in similar proportions to
those generated by CD34 þ umbilical cord blood cells, although
Leukemia
Functional heterogeneity in human T-ALL
B Gerby et al
1252
Figure 2 Functional analysis of T-ALL CD34 þ CD7 cells. (a) 5 104 CD34 þ CD7 cells sorted from M77 T-ALL were co-cultured with either
MS5-DL1 or MS5-Ctl stromal cells for, respectively, 7 and 3 weeks. Expression of various cell surface markers was monitored every week for
co-cultures with MS5-DL1 cells and on week 3 for co-cultures with MS5-Ctl cells. Percent of cells is indicated under every plot. Data are
representative of seven (M53, M60, M77, M79, M70, M71 and M105) T-ALL. (b) TLX3 rearrangement was tested by fluorescence in situ hybridization
in M77 patient diagnosis sample (left), in the progeny of M77-derived CD34 þ CD7 cells after 21 days in co-culture with MS5-DL1 cells (middle) and
16 weeks after transplantation into NSG mice (right). Yellow arrows indicate the presence of a TLX3 rearrangement. (c) 5 104 CD34 þ CD7 cells
sorted from M77 T-ALL were transplanted into NSG mice. The bone marrow (BM), spleen and thymus of each mouse were analyzed 16 weeks after
transplantation for the presence of human CD45 þ cells. Engrafted CD45 þ human cells were further characterized with myeloid (CD14 and CD15),
lymphoid (CD7, CD4, CD8 and CD19) and CD34 subset-specific antibodies. Plots are representative of 10 mice and 2 independent samples. (d) TCRg
and TCRd gene rearrangements of human cells recovered from BM and thymus of NSG mice. Representative of three mice. Left histograms are results
of the corresponding patient sample. The color reproduction of this figure is available at the Leukemia journal online.
at lower levels (Supplementary Table 3). The cells recovered
from NSG mice transplanted with CD34 þ CD7 cells from
sample M77 did not have the TLX3 rearrangement originally
present in the parent sample (Figure 2b). Analysis of the clonal
rearrangements of the TCR genes (TCRg and TCRd) indicated
that the CD34 þ CD7-derived cells did not exhibit the specific
clonal rearrangements present in the sample at diagnosis
(Figure 2d); human CD7 þ cells present in the thymus of
transplanted mice contained polyclonal TCRg and TCRd gene
rearrangements reminiscent of normal T-cell maturation.
Furthermore, when these T cells from the thymus were
transplanted into secondary recipient mice, they failed to
initiate human T-ALL (data not shown). Together, these findings
demonstrate that CD34 þ CD7 cells, isolated from T-ALL
patients contain normal hematopoietic stem cells (HSCs) and
progenitor cells but no T-ALL leukemic cells.
T-LIC activity is enriched in CD34 þ CD7 þ cells
We know that transplantation of mononucleated cells
from T-ALL patients into immune-deficient mice results in
Leukemia
leukemia.13 Furthermore, we found a correlation between the
proportion of CD34 þ cells and the LIC activity in several
samples we tested (Supplementary Figure 3), yet we have
demonstrated above that CD34 þ CD7 cells from most T-ALL
patients do not have LIC activity (Figure 3a). To investigate
further the relationship between CD34, CD7 and LIC activity,
we sorted CD34 þ CD7 þ and CD34CD7 þ cells and transplanted them into NS and NSG mice. We analyzed human
CD45 þ CD7 þ leukemic cells in the BM and spleens of the mice
as a measure of leukemia development.13 When doses of 1–
5 104 CD34 þ CD7 þ cells/mouse were transplanted, 26/29
(89.6%) mice harbored human leukemic cell infiltration (that is,
41% CD45 þ CD7 þ cells) in their BM, whereas when similar
numbers (5 104 cells/mouse) of CD34CD7 þ cells were
transplanted, only 11/24 (46%) mice did so (Figure 3a). Moreover, engraftment levels were higher in mice transplanted
with CD34 þ CD7 þ compared with mice transplanted with
CD34CD7 þ cells (63 versus 1.5% CD45 þ CD7 þ cells,
Po0.001; Figure 3a). We also observed the leukemic cells
spread to the spleen: 21/24 (88%) mice transplanted with
CD34 þ CD7 þ cells contained human leukemic CD45 þ CD7 þ
Functional heterogeneity in human T-ALL
B Gerby et al
1253
Figure 3 T-LIC activity of T-ALL CD34 þ CD7 þ and CD34CD7 þ cells. (a) Left panels: CD34 þ CD7, CD34 þ CD7 þ and CD34CD7 þ cells
sorted from eight T-ALL patient samples were injected into NS mice (M61}, M66& and M78J) or NSG mice (M71K, M77~, M87’, M89m
and M22 ); (2.5–5 103 or 1–5 104 cells/mouse; 3–5 mice/group). Unsorted cells were used as positive controls (0.1–5 106 cells/mouse).
Percent of human leukemic CD45 þ CD7 þ cells in BM and spleen was monitored between 6 (NSG mice) and 12 (NS mice) weeks after
transplantation. Right panels: Secondary transplants performed with leukemic cells recovered from primary mice transplanted with CD34 þ CD7 þ
and CD34CD7 þ cells from M61 (&; 2.5 104 blast cells/NS mouse), M71 (J; 5 104 blast cells/NS mouse) and M22 (’; 5 104 and 5 103
blast cells/NSG mouse; 3–5 mice/group) T-ALL. Leukemic CD45 þ CD7 þ cell development was measured 10 weeks after transplant. Median
engraftment is indicated above each plot; *Po0.05, **Po0.01, ***Po0.001. (b) TCRd gene rearrangements and SIL-TAL1 micro-deletion analyses
in unfractionated cells and in sorted CD34 þ CD7 þ and CD34CD7 þ cells from M87 T-ALL patient. Results from human CD45 þ CD34 þ CD7 þ and CD34CD7 þ -derived cells engrafted into NSG mice are also shown. Representative of two T-ALL.
cells in their spleens, compared with 10/24 (42%) mice
transplanted with CD34CD7 þ cells; leukemic engraftment
level in the spleen was significantly higher for CD34 þ CD7 þ
compared with CD34CD7 þ cells (20 versus 0.05%, Po0.001;
Figure 3a). Moreover, mice transplanted with CD34 þ CD7 þ
cells reproducibly developed splenomegaly and pale bones
characteristics of high leukemia invasion (Supplementary Figure
4). Thus, mice transplanted with CD34 þ CD7 þ cells reproducibly developed high levels of invading human leukemia.
Injecting fewer cells of either CD34 þ CD7 þ or CD34CD7 þ
cells into mice confirmed that the CD34 þ CD7 þ cells were
enriched in T-LIC activity compared with CD34CD7 þ cells.
Transplantation of 2.5–5 103 CD34 þ CD7 þ or CD34CD7 þ
cells engrafted, respectively, 66% (8/12) and 28% (2/7) of mice
Leukemia
Functional heterogeneity in human T-ALL
B Gerby et al
1254
Figure 4 NOTCH activation in CD34 þ CD7 þ and CD34CD7 þ T-ALL cells. (a) 5 104 CD34 þ CD7 þ (white diamonds) and CD34CD7 þ
(white squares) cells sorted from M61, M78 and M66 T-ALL were co-cultured with MS5-Ctl or MS5-DL1 cells for 30 days in the presence (GSI;
filled symbols, dashed lines) or absence (vehicle; open symbols, plain lines) of 10 mM DAPT. Human CD45 þ CD7 þ leukemic cells were harvested
and counted by flow cytometry at each time point. Representative of eight T-ALL. (b) Phenotype of unfractionated, CD34 þ CD7 þ and
CD34CD7 þ cells (105 cells/well) from M66 T-ALL after 1 week of culture with MS5-DL1 cells. (c) TAL1 gene expression was measured in
cultured M66 T-ALL cells generated at day 7 of culture from unfractionated or fractionated cells. Mean±s.d. of two measurements normalized over
B2microglobulin mRNA levels are shown. Representative of two T-ALL. (d) Deltex-1, pTa and HES-1 gene expression was monitored in M66 and
M78 CD34 þ CD7 þ (white bars) and CD34CD7 þ (black bars)-derived cells co-cultured 7 days with MS5-Ctl or MS5-DL1 cells. Data are mean
values±s.d. of three measurements normalized over B2microglobulin; *Po0.05, **Po0.01, ***Po0.001. Representative of three T-ALL.
with 6 and 0.6% median engraftment levels. As shown above,
5 104 CD34CD7 þ cells were significantly less efficient than
2.5–5 103 CD34 þ CD7 þ cells (Figure 3a). In one case (M105,
Supplementary Table 1), limiting cell numbers were transplanted and leukemia development was followed up during up
to 9 weeks for the lowest cell dose. Limiting dilution and kinetic
analysis confirmed the more aggressive behavior and the
LIC enrichment of CD34 þ CD7 þ cells as compared with
CD34CD7 þ cells, although as seen in other samples
(Figure 3a), CD34 þ CD7 cells also contained LIC activity
(Supplementary Figure 5).
The engrafted human cells exhibited the same genetic
abnormality identified at diagnosis (for example, the SIL-TAL1
genomic micro-deletion for M61 and M87 T-ALL) and they
had the specific clonal TCR gene rearrangements identified
at diagnosis in their respective leukemic cells (Figure 3b and
data not shown). Thus, the CD34 þ CD7 þ human cells that
engrafted immune-deficient mice derived from the patients’
leukemic cells.
To test whether the human leukemic cells recovered from
primary mice contained cells with self-renewal properties, we
transplanted human leukemic CD45 þ CD7 þ cells recovered
from some primary transplanted mice into secondary recipients.
Leukemia
The engraftment efficiency observed in primary mice, namely
that cells originating from CD34 þ CD7 þ cells engrafted better
compared with CD34CD7 þ cells, was reproduced in the
secondary mice (Figure 3a), indicating that CD34 þ CD7 þ cells
were endowed with self-renewing capacity.
Together, these results indicate that selection of CD34 þ
CD7 þ cells in T-ALL patient samples enriches for T-LIC activity.
Differential responses of CD34 þ CD7 þ and
CD34CD7 þ cells to NOTCH activation
Signaling through the NOTCH pathway is important in the
development/maintenance of T-ALL and is implicated in T-LIC
activity.13 In this section, we assessed the growth capacity of
CD34 þ CD7 þ and CD34CD7 þ populations purified from
eight T-ALL patients (M61, M66, M71, M69, M73, M78, M22
and M105) during co-culture with MS5-Ctl and MS5-DL1. In all
tested samples, except one, CD34 þ CD7 þ cells grew faster than
did CD34CD7 þ cells when co-cultured with MS5-DL1 cells
for 4 weeks, although variability in growth curves was observed
between individual samples (Figure 4a; Supplementary Figure
6A). In accordance, a greater proportion of CD34 þ CD7 þ cells
expressed the cell proliferation marker Ki-67 than did
Functional heterogeneity in human T-ALL
B Gerby et al
1255
Figure 5 Proliferation and T-LIC activity of CD34 þ CD7 þ and CD34CD7 þ T-ALL cells after treatment with dexamethasone and GSI.
(a) CD34 þ CD7 þ and CD34CD7 þ cells from the M88 T-ALL sample were sorted and co-cultured for 1 week with MS5-DL1 stromal cells, in the
presence of 0.1–500 nM dexamethasone, 0.5–5 mM DAPT, or 5 mM DAPT þ 0.1–50 nM dexamethasone, or with the drug diluent (vehicle). Cultured
cells were harvested, labeled for the proliferation marker Ki67 and measured by flow cytometry. Results indicate mean±s.d. of duplicated wells.
(b) CD34 þ CD7 þ cell population from M105 T-ALL sample was co-cultured for 1 week with MS5-DL1 cells, in the absence (Vehicle) or in the
presence of 5 nM dexamethasone þ 5 mM DAPT. Cultured cells were harvested and absolute number of Ki67 þ cells was measured by flow
cytometry. Data are mean±s.d. of three measurements. (c) Limited cell numbers of vehicle or dexamethasone þ GSI (5 nM þ 5 mM)-treated
CD34 þ CD7 þ cells from M105 T-ALL were transplanted into NSG mice after a 1-week in vitro treatment period. Percent of human leukemic
CD45 þ CD7 þ cells in BM was monitored 5 weeks after transplantation. Statistical comparisons between groups were done using the Student’s
t-test; *Po0.05, **Po0.01, ***Po0.001. ND: not done. (d) Kinetic follow-up of T-ALL development after 5 101 vehicle (black triangles) or
dexamethasone þ GSI (5 nM þ 5 mM, white triangles) treated CD34 þ CD7 þ cells transplantation. Shown are % CD45 þ CD7 þ cell in BM sampled at
5, 7 and 9 weeks after transplantation. Solid bars represent the medians of leukemia engraftment.
CD34CD7 þ cells after 7 days co-culture with MS5-DL1
stromal cells (Supplementary Figure 7). Surprisingly, CD34 þ
CD7 þ cells from 7/8 samples also generated leukemic cells
upon co-culture with MS5-Ctl cells, although at lower levels
than with MS5-DL1 cells, whereas the CD34CD7 þ fraction
did not (Figure 4a; Supplementary Figure 6A). Cultured CD34 þ
CD7 þ - and CD34CD7 þ -derived human cells had similar
patterns of expression of CD4, CD8 and CD34, although CD34
expression was lost during culture with MS5-DL1 stromal
support (Figure 4b). Cultured cells continued to express their
oncogenic marker (for instance, TAL1 for M66 T-ALL; Figure 4c),
indicating they were of leukemic origin. High level (10 mM) of
GSI and NOTCH signaling inhibitor DAPT completely inhibited
the growth of both CD34 þ CD7 þ and CD34CD7 þ cells in coculture with either MS5-DL1 or MS5-Ctl stromal cells (Figure 4a;
Supplementary Figure 6A). These data suggested that NOTCH
signaling is implicated in T-LIC cell growth with MS5-DL1 and
also with MS5-Ctl cells, which are known to express baseline
levels of the NOTCH ligands DL1 and jagged1-2.18,19 We
further evaluated NOTCH activation in leukemic cells derived
from sorted cell fractions after 7–30 days in co-culture by
measuring expression of the NOTCH target genes Deltex-1,
HES-1 and pTa (Figure 4d; Supplementary Figure 6B). Expression levels were higher in leukemic cells recovered from
MS5-DL1 co-cultures than in those from MS5-Ctl co-cultures,
consistent with the growth curves described above. Most
importantly, NOTCH target genes were expressed significantly
more in CD34 þ CD7 þ cells than in CD34CD7 þ cells,
whatever MS5 stromal cells used (Figure 4d; Supplementary
Figure 6B). Therefore, our data demonstrate that CD34 þ CD7 þ
and CD34CD7 þ cell populations have differential sensitivity
to NOTCH activation in vitro, which correlates with their
different growth rates.
Combination of dexamethasone and GSI reduces
proliferation and T-LIC activity of CD34 þ /CD7 þ cells
The corticoid-like molecule dexamethasone is commonly used
during the induction period of chemotherapy of T-ALL
patients.20 We next investigated whether CD34 þ CD7 þ cells
were sensitive to dexamethasone and whether combining
dexamethasone with DAPT would interfere with their proliferation compared with dexamethasone alone. Fractionated cells
from four T-ALL patient samples (M22, M18, M88 and M105)
were co-cultured with MS5-DL1 cells for 7 days in the presence
of increasing concentrations of dexamethasone (0.1–500 nM) or
DAPT (0.5–5 mM) or a combination of the two, dexamethasone
(0.1–50 nM) þ DAPT (5 mM). For two samples (M22 and M88),
the cell fraction CD34CD7 þ was also tested. We observed a
consistent dose-dependent sensitivity of CD34 þ CD7 þ cell
fraction to treatment with either dexamethasone or DAPT,
whereas CD34CD7 þ cells behaved more variably as a result
of their low activation status when cultured with MS5-DL1
stromal cells (Figure 5a; Supplementary Figure 8A). Interestingly,
treatment with a combination of DAPT and 5–50 nM dexamethasone additively inhibited cell growth of both cell fractions.
Leukemia
Functional heterogeneity in human T-ALL
B Gerby et al
1256
Finally, in order to test the efficiency of DAPT þ dexamethasone treatment on T-LIC activity, we transplanted residual
cells recovered from cultures of CD34 þ CD7 þ cells from two
T-ALL (M18 and M105), treated or not with dexamethasone and
DAPT, into NSG mice (Figure 5b and c; Supplementary Figure
8B). M105 T-ALL cells were injected at limiting dilution. We
found that although T-LIC activity was still detected in drugtreated cells, the levels of leukemic cells in the BM and spleens
of mice transplanted with the drug-treated CD34 þ CD7 þ cells
were decreased compared with control, vehicle-treated cells
(Figure 5c). Furthermore, the limiting dilution experiment
combined with the kinetic analysis of drug-treated cells
compared with controls indicates that treatment reduced the
T-LIC frequency in the CD34 þ CD7 þ cell compartment (Figures
5c and d; Supplementary Figure 8). As only few cells (410-fold
less) were recovered in culture after 7 days treatment with
dexamethasone þ DAPT compared with vehicle-treated control
cells (Figure 5b; Supplementary Figure 8B), these data show
that combined dexamethasone þ DAPT treatment dramatically
reduces the overall T-LIC activity of CD34 þ CD7 þ cells.
Discussion
Current chemotherapeutic agents were originally developed to
act against the bulk of malignant cells rather than any putative
cancer-initiating/repopulating cells. Consequently, LICs may
resist current chemotherapies and may contribute to relapse of
the disease.6 Several studies indicate a central role of initiating
cells in the pathogenesis of leukemias, implying that targeting
these cells specifically could help to eradicate the disease and
prevent relapses. Toward this goal, our work is aimed at
investigating heterogeneity of cell populations in human T-ALL
and at providing new insights to help the cure of such blood
disorders.
We first demonstrate here that CD34 þ CD7 cells in the
blood of T-ALL patients neither generate blast cells in vitro nor
promote leukemia in immune-deficient mice. In fact, in all, but
one, samples tested, these cells give rise to human B, T and
myeloid hematopoietic cells that are not of leukemic origin,
indicating the presence of normal HSC and progenitor cells in
this population. This is similar to what has been described for
CD34 þ CD19CD10CD38 cells in B-ALL patients10,21 and is
consistent with the clinical observation that patients treated with
high-dose chemotherapy eventually restore normal hematopoiesis, implying that residual normal progenitors/HSC are present.
In agreement with a previous study,22 hematopoietic reconstitution from CD34 þ CD7 cells was quiet inefficient compared
with CD34 þ cells isolated from umbilical cord blood, reflecting
the presence of high progenitor/HSC ratio in patient blood or
an ontogeny difference between both cell sources.23 Interestingly, the absolute numbers of CD34 þ CD7 normal cells in
peripheral blood of T-ALL patients are high compared with
healthy subjects, suggesting either an abnormal proliferation or
a mobilization of the normal HSC/progenitors in patient blood
possibly related to high leukemic cell infiltration of the BM.
Our observation that CD34 þ CD7 T-ALL cells do not have
LIC activity appears to contradict an earlier study by Cox et al.12
that reported that CD34 þ CD4 and CD34 þ CD7 cells were
capable of NOD/SCID engraftment. This might be explained by
the fact that the cell populations in the two studies were
different. In our study here, the CD34 þ cells were both CD7
and CD4, whereas in the study by Cox et al. the cell
populations were either CD7 or CD4, not both. As we
demonstrated here, the CD34 þ CD4 cells from certain T-ALL
Leukemia
samples contained CD7 þ cells; this raises the possibility
that the CD34 þ CD4 cells in the study by Cox et al. were
CD34 þ CD7 þ CD4, like the cells that we found had LIC
activity. Another possibility would be that either study did not
work with similar T-ALL cohorts: we mostly studied mature/
cortical T-ALL and probably immature T-ALL would have
greater LIC activity in more primitive-like populations.
In fact, in our experiments T-LIC activity is present only in
CD7 þ cells, implying that leukemia development or maintenance relies on a cell population already engaged in T-cell
differentiation.3 This conclusion is consistent with the recently
published work on Lmo2-induced T-cell leukemia in mouse,
that takes its first steps in the thymus,24 and with the fact that
most genomic rearrangements in T-ALL that are thought to be
primary oncogenic events occurred during TCR V(D)J region
recombination and thus probably occur in differentiating T-cell
precursors.1 Moreover, we found that T-LICs were particularly
enriched in CD34 þ CD7 þ cells compared with CD34CD7 þ
cells. The enrichment of LICs in the CD34 þ CD7 þ cell fraction
might correlate with a high aggressiveness in patients, as it has
been recently reported a correlation between the expression of
CD34 and the prognosis of T-ALL.4
Our results indicate that LIC activity is also detected, although
to lower levels, in the CD34CD7 þ cell population. This
implies that in T-ALL samples in which this population is major,
the LICs are at least as numerous in CD34CD7 þ populations as
in CD34 þ CD7 þ populations. These findings are consistent with
a very recent paper on acute myeloid leukemia that reported
that, in certain samples, LIC activity was found in CD34
cell populations.25 This suggests that additional markers,
such as the CD90 (Thy1) marker26 or Hoechst dye exclusion,
will be necessary to further purify LICs from T-ALL samples.
Transplantation into immune-deficient mice may select for
cells that have an advantage to grow in a xenogenic environment and/or that escape from residual mouse immune activity,
as previously reported for normal and acute myeloid leukemia
cells where CD47 expression help escaping from the mouse
macrophages.27,28 CD34 is traditionally seen as an immature
cell marker.16 On the other hand, CD34 belongs to a sialomucin
protein family with anti-adhesive properties, implicated in cell
trafficking and migration of hematopoietic progenitor cells to the
BM.29,30 CD34 expression on CD7 þ T-ALL cells is mostly not
related to an immature stage of differentiation but rather to an
aberrant expression as T-ALL samples we studied belong to
the mature/cortical T-ALL immunologic subtypes, and thus
differ from the early T-cell precursor leukemia subtype recently
characterized.31 Thus, expression of CD34 might confer
migration or adhesiveness properties to leukemic cells that
may contribute to the LIC activity of CD34 þ CD7 þ cells.
We previously demonstrated that signaling through NOTCH
is crucial for the maintenance/amplification of LICs from T-ALL
patients.13 We extend this finding here by showing that the
CD34 þ CD7 þ population is more sensitive to NOTCH activation than CD34CD7 þ T-ALL cells. Consistent with this,
CD34 þ CD7 þ cell proliferation was easily inhibited by the
NOTCH signaling inhibitor DAPT, but also by the corticoid-like
molecule dexamethasone, these drugs having additive effects
on the proliferation of this particular T-ALL cell fraction,
as previously reported for T-ALL cell lines and a few patient
samples.32 This finding is surprising because one might expect
the LICs to be quiescent and thus more resistant to drugs than the
non-T-LIC cell fraction.33 This has been recently addressed in
another study on T-ALL.34 However, as T-LIC phenotype, T-ALL
samples and the experimental design differ between our study
and the Chiu’s study, comparison of results may be difficult. One
Functional heterogeneity in human T-ALL
B Gerby et al
1257
explanation may reside in the assay we used, that is, NOTCH
activation during co-culture with MS5-DL1 cells. However, as coculture of leukemic fractions with stromal cells may reflect the in
vivo niche context, combining dexamethasone and GSI treatment
of CD34 þ CD7 þ enriched T-LIC very efficiently decreased T-LIC
activity, indicating that such treatment could be interesting for
patients. It is, however, necessary to acknowledge the fact that
inhibiting NOTCH pathway as a therapeutic strategy might be
limited to samples where PTEN and/or FBW7 are not mutated or
lost as it would interfere with GSI sensitivity.35,36 In fact, it would
be of great interest to test such primary samples in vitro and in vivo
with combination of drugs as most previous studies have been
done with cell lines that probably do not faithfully reflect the
behavior of patient samples.
Conclusions
Altogether, the findings reported in this paper show that cells in
human T-ALL samples are heterogeneous in terms of their
CD34, CD7 and CD4 marker expression and in terms of normal
hematopoiesis and initiation/repopulation of leukemia
in immune-deficient mice. We find the LIC activity enriched
in CD34 þ CD7 þ cells, but not in CD34 þ CD7 cells that give
rise to normal hematopoiesis. We further show that NOTCH
activation is crucial for LIC activity from CD34 þ CD7 þ cells
and inhibition of such pathway can efficiently interfere, in
combination with corticoid-like molecule, with leukemia
development in xenograft models. These findings provide a
base for studying and understanding LICs in T-ALL, and new
avenues for the development of promising therapeutics.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgements
We acknowledge all the patients and their parents for providing
samples for our studies. Cell sorting was performed by the flow
cytometry platform of IRCM with the help of N Dechamps
and J Baijer (Fontenay-aux-Roses, France). C Joubert, J Tilliet and
V Neuville in IRCM provided excellent animal colony care. BG
received a fellowship from the Ligue Nationale Contre le Cancer
and from the Société Française d’Hématologie. EC is a fellow of
Institut du Cancer (INCA). FA was paid by the CreMEC
consortium. This work was supported by grants from INSERM,
CEA, Université Paris 7, Association Laurette Fugain, INCA and
the Cancéropole, Ile de France.
Author contributions
BG, EC, FA, CD, SP and JC performed experiments; BG, EC, PB
and FP designed the experiments, analyzed the results and wrote
the manuscript; JS and PHR helped with critical experiments
and with writing the manuscript; NB, TL, AB, JLP and PB
provided patient samples and shared crucial informations.
References
1 De Keersmaecker K, Marynen P, Cools J. Genetic insights in the
pathogenesis of T-cell acute lymphoblastic leukemia. Haematologica 2005; 90: 1116–1127.
2 Ferrando AA, Look AT. Gene expression profiling in T-cell acute
lymphoblastic leukemia. Semin Hematol 2003; 40: 274–280.
3 Blom B, Spits H. Development of human lymphoid cells. Annu Rev
Immunol 2006; 24: 287–320.
4 van Grotel M, van den Heuvel-Eibrink MM, van Wering ER, van
Noesel MM, Kamps WA, Veerman AJ et al. CD34 expression is
associated with poor survival in pediatric T-cell acute lymphoblastic leukemia. Pediatr Blood Cancer 2008; 51: 737–740.
5 Dick JE. Stem cell concepts renew cancer research. Blood 2008;
112: 4793–4807.
6 Reya T, Duncan AW, Ailles L, Domen J, Scherer DC, Willert K
et al. A role for Wnt signalling in self-renewal of haematopoietic
stem cells. Nature 2003; 423: 409–414.
7 Bonnet D, Dick JE. Human acute myeloid leukemia is organized
as a hierarchy that originates from a primitive hematopoietic cell.
Nat Med 1997; 3: 730–737.
8 Hope KJ, Jin L, Dick JE. Acute myeloid leukemia originates from a
hierarchy of leukemic stem cell classes that differ in self-renewal
capacity. Nat Immunol 2004; 5: 738–743.
9 Wang JC, Lapidot T, Cashman JD, Doedens M, Addy L, Sutherland
DR et al. High level engraftment of NOD/SCID mice by primitive
normal and leukemic hematopoietic cells from patients with
chronic myeloid leukemia in chronic phase. Blood 1998; 91:
2406–2414.
10 Castor A, Nilsson L, Astrand-Grundstrom I, Buitenhuis M,
Ramirez C, Anderson K et al. Distinct patterns of hematopoietic
stem cell involvement in acute lymphoblastic leukemia. Nat Med
2005; 11: 630–637.
11 Cox CV, Evely RS, Oakhill A, Pamphilon DH, Goulden NJ, Blair A.
Characterization of acute lymphoblastic leukemia progenitor cells.
Blood 2004; 104: 2919–2925.
12 Cox CV, Martin HM, Kearns PR, Virgo P, Evely RS, Blair A.
Characterization of a progenitor cell population in childhood
T-cell acute lymphoblastic leukemia. Blood 2007; 109: 674–682.
13 Armstrong F, de la Grange PB, Gerby B, Rouyez MC, Calvo J,
Fontenay M et al. NOTCH is a key regulator of human T-cell acute
leukemia initiating cell activity. Blood 2009; 113: 1730–1740.
14 van Dongen JJ, Langerak AW, Bruggemann M, Evans PA,
Hummel M, Lavender FL et al. Design and standardization of
PCR primers and protocols for detection of clonal immunoglobulin
and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT983936. Leukemia 2003; 17: 2257–2317.
15 Weng AP, Ferrando AA, Lee W, Morris JPT, Silverman LB,
Sanchez-Irizarry C et al. Activating mutations of NOTCH1 in
human T cell acute lymphoblastic leukemia. Science 2004; 306:
269–271.
16 Krause DS, Fackler MJ, Civin CI, May WS. CD34: structure,
biology, and clinical utility. Blood 1996; 87: 1–13.
17 La Motte-Mohs RN, Herer E, Zuniga-Pflucker JC. Induction of
T-cell development from human cord blood hematopoietic stem
cells by Delta-like 1 in vitro. Blood 2005; 105: 1431–1439.
18 Han W, Ye Q, Moore MA. A soluble form of human Delta-like-1
inhibits differentiation of hematopoietic progenitor cells. Blood
2000; 95: 1616–1625.
19 Bennaceur-Griscelli A, Pondarre C, Schiavon V, Vainchenker W,
Coulombel L. Stromal cells retard the differentiation of CD34
(+CD38(low/neg) human primitive progenitors exposed to cytokines
independent of their mitotic history. Blood 2001; 97: 435–441.
20 Veerman AJ, Kamps WA, van den Berg H, van den Berg E,
Bokkerink JP, Bruin MC et al. Dexamethasone-based therapy for
childhood acute lymphoblastic leukaemia: results of the prospective Dutch Childhood Oncology Group (DCOG) protocol ALL-9
(1997-2004). Lancet Oncol 2009; 10: 957–966.
21 Kong Y, Yoshida S, Saito Y, Doi T, Nagatoshi Y, Fukata M et al.
CD34+CD38+CD19+ as well as CD34+CD38-CD19+ cells
are leukemia-initiating cells with self-renewal capacity in human
B-precursor ALL. Leukemia 2008; 22: 1207–1213.
22 Meek B, Cloosen S, Borsotti C, Van Elssen CH, Vanderlocht J,
Schnijderberg MC et al. In vitro-differentiated T/natural killer-cell
progenitors derived from human CD34+ cells mature in the
thymus. Blood 2010; 115: 261–264.
23 Lansdorp PM, Dragowska W, Mayani H. Ontogeny-related
changes in proliferative potential of human hematopoietic cells.
J Exp Med 1993; 178: 787–791.
24 McCormack MP, Young LF, Vasudevan S, de Graaf CA,
Codrington R, Rabbitts TH et al. The Lmo2 oncogene initiates
Leukemia
Functional heterogeneity in human T-ALL
B Gerby et al
1258
25
26
27
28
29
leukemia in mice by inducing thymocyte self-renewal. Science
2010; 327: 879–883.
Taussig DC, Vargaftig J, Miraki-Moud F, Griessinger E,
Sharrock K, Luke T et al. Leukemia-initiating cells from some
acute myeloid leukemia patients with mutated nucleophosmin reside in the CD34() fraction. Blood 2010; 115:
1976–1984.
Yamazaki H, Nishida H, Iwata S, Dang NH, Morimoto C. CD90
and CD110 correlate with cancer stem cell potentials in human
T-acute lymphoblastic leukemia cells. Biochem Biophys Res
Commun 2009; 383: 172–177.
Takenaka K, Prasolava TK, Wang JC, Mortin-Toth SM,
Khalouei S, Gan OI et al. Polymorphism in Sirpa modulates
engraftment of human hematopoietic stem cells. Nat Immunol
2007; 8: 1313–1323.
Majeti R, Chao MP, Alizadeh AA, Pang WW, Jaiswal S, Gibbs Jr KD
et al. CD47 is an adverse prognostic factor and therapeutic antibody target on human acute myeloid leukemia stem cells.
Cell 2009; 138: 286–299.
Nielsen JS, McNagny KM. Influence of host irradiation on longterm engraftment by CD34-deficient hematopoietic stem cells.
Blood 2007; 110: 1076–1077.
30 Nielsen JS, McNagny KM. Novel functions of the CD34 family.
J Cell Sci 2008; 121 (part 22): 3683–3692.
31 Campana D, van Dongen JJ, Mehta A, Coustan-Smith E, WolversTettero IL, Ganeshaguru K et al. Stages of T-cell receptor protein
expression in T-cell acute lymphoblastic leukemia. Blood 1991;
77: 1546–1554.
32 Real PJ, Tosello V, Palomero T, Castillo M, Hernando E, de
Stanchina E et al. Gamma-secretase inhibitors reverse glucocorticoid resistance in T cell acute lymphoblastic leukemia. Nat Med
2009; 15: 50–58.
33 Trumpp A, Essers M, Wilson A. Awakening dormant haematopoietic stem cells. Nat Rev Immunol 2010; 10: 201–209.
34 Chiu PP, Jiang H, Dick JE. Leukemia-initiating cells in human
T-lymphoblastic leukemia exhibit glucocorticoid resistance. Blood
2010; 116: 5268–5279.
35 Palomero T, Sulis ML, Cortina M, Real PJ, Barnes K, Ciofani M
et al. Mutational loss of PTEN induces resistance to NOTCH1
inhibition in T-cell leukemia. Nat Med 2007; 13: 1203–1210.
36 O’Neil J, Grim J, Strack P, Rao S, Tibbitts D, Winter C et al. FBW7
mutations in leukemic cells mediate NOTCH pathway activation
and resistance to gamma-secretase inhibitors. J Exp Med 2007;
204: 1813–1824.
Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)
Leukemia