| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Department of Genetics, University Erlangen, 91058 Erlangen [S. S., M. B., M-P. G-C., O. Z., R. S.], and Children’s Hospital, University Tübingen, 72076 Tübingen [J. G.], Germany
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
On the molecular level, 11q23 aberrations affect the mixed lineage leukemia gene (MLL, also called HRX or ALL-1). MLL codes for a large 3968-amino-acid protein with limited but significant homology to the Drosophila epigenetic regulator trithorax (trx; 5, 6, 7, 8 ). In the fruit fly, TRX is required to maintain, but not to initiate, the expression pattern of the homeobox group genes that control the body plan (9) . TRX is believed to act by establishing an open state of chromatin that enables transcription of its target genes. Knockout studies in mice suggest a similar function also for MLL (10, 11, 12) . In vitro assays studying the hematopoietic capability of either MLL-/- embryonic stem cells or MLL-/- fetal liver cells demonstrated an essential role for MLL in hematopoiesis (13 , 14) . All cloned 11q23 translocations consistently combine the 5' portion of the MLL gene in frame with the corresponding fusion partners. The translation of the respective chimeric RNAs yields fusion proteins with novel combinations of functional domains. MLL contributes two DNA binding motifs that recognize DNA without strict sequence specificity: the AT-hooks and the CXXC or methyltransferase domain. The role of the respective fusion partners is less clear. There is no apparent unifying feature among the 21 fusion partners cloned to date, except that several of them are either directly or indirectly involved in transcriptional activation. The three most frequent fusion partners AF4, AF9, and ENL [t(4;11), t(9;11), t(11;19)] contain transactivator domains (15 , 16) and, consequently, it has been shown that the fusion protein MLL-ENL is a pleiotropic transcriptional transactivator when tested in transient reporter gene assays (17) .
Despite substantial sequence information about MLL and its fusion
partners, the molecular mechanisms of hematopoietic cell transformation
by MLL fusion proteins remain elusive. Two studies demonstrate the
fundamental capability of the translocation products MLL-ENL
[t(11;19)] and MLL-AF9 [t(9;11)] to act as leukemogenic
proto-oncogenes. Retroviral transduction of primary mouse hematopoietic
precursor cells with a virus expressing MLL-ENL immortalized
myelomonocytic precursors. These cells were able to elicit a myeloid
leukemia on reintroduction into mice. However, MLL-ENL-induced
leukemias were only observed after a prolonged latency period of 
3
months (18)
. In a second study, a translocation t(9;11)
was reconstructed at the chromosomal level by a "knock-in"
strategy. Modified ES cells were used to generate chimeric mice and the
animals were followed in their hematopoietic development. Although an
expanded myeloid cell population derived from cells carrying the
knock-in allele was detectable shortly after birth, these mice finally
succumbed to mostly myeloid leukemias after a latency period of more
than 4 months (19
, 20)
. Surprisingly, a fusion of
MLL with the bacterial lacZ gene was able to
cause similar myeloid leukemias in this system. Notably, in the case of
MLL-lacZ expression, the elapsed latency period before the appearance
of overt leukemias was approximately twice as long as that observed
after expression of the authentic fusion protein MLL-AF9
(21)
. The use of a model with additional uncontrolled
genetic alterations accumulating during the life span of the animals
makes an investigation of the progression toward full leukemia
difficult. Therefore, the immediate consequence of MLL fusion protein
expression on cell physiology as well as the nature of potentially
cooperating oncogenes is unclear.
In this study, we used a retroviral transduction approach to investigate the direct effects of MLL-ENL on the development of primary hematopoietic cells. We show here that a block in myelomonocytic development caused by MLL-ENL is dependent on the proto-oncogene c-myc, and that it is readily reversible. Hematopoietic progenitors transformed by MLL-ENL are growth factor dependent and apoptosis competent, and a normal genetic program of cell differentiation can be induced by G-CSF.3 A suppression of Myc activity precludes transformation by MLL-ENL, whereas an ectopic expression of Myc cooperates with the fusion protein. The consequences of these findings for the etiology of leukemias with MLL rearrangements are discussed.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
DNA Manipulation and Plasmid Generation.
Genomic DNA was purified according to standard methods
(22)
. For retroviral transductions, all cDNAs were
inserted into either the retroviral vector pMSCV-neo or
pMSCV-pac (23)
. The cDNAs of the human
c-myc gene and the dominant-negative c-myc mutant
lacking the amino acids 40–178 of the Myc transactivation domain
(
Myc) were a gift from D. Eick (24
, 25)
.
RT-PCR, and cDNA Arrays.
Polyadenylated messenger RNA for RT-PCR was purified by oligo
d(T) chromatography on latex beads (Oligotex; Quiagen Inc) according to
the manufacturer’s instructions. Primers used for RT-PCR of
MLL-ENL, Myc, and Myc were: MLL forward (gcaaacagaaaaaagtggctccccg)
and ENL reverse (accatccagtcgtgagtgaacccct); myc forward
(cctagtgctgcatgaggggacac) and myc reverse (gcaccagagtttcgaagctgttc);
and Myc forward (ctcgatcctccctttatccagcc) and Myc reverse
(cctcatcttcttgttcctcctcag). The cDNA array analysis was done with the
ATLAS mouse cDNA expression array (version 1) from Clontech. The
evaluation of the hybridization results was performed with the Clontech
Atlas Image analyzer software.
Virus Production and Retroviral Transduction of Mouse Primary
Hematopoietic Cells.
High titer retrovirus supernatants were produced by transient
transfection of the packaging cell line Phoenix-E by a standard
calcium-phosphate precipitation method (22)
. The viral
titers were in the range of 1 x 106 to >1 x 107 colony-forming units/ml depending on the size
of the insert in the retroviral vector. The retroviral transduction of
primary hematopoietic cells was done according to the method of Lavau
et al. (18)
. An overview of the procedure is
given in Fig. 5A
. BMCs, enriched in noncycling hematopoietic
precursors, were recovered from 5-FU-primed Balb/C mice. Previous to
infection, the BMCs were activated by a cytokine mixture (IL-3, IL-6,
and SCF). After infection by spinoculation, the infected cells were
returned to activation medium as above. After overnight activation,
duplicates of 1 x 104
infected
cells were plated under the appropriate drug selection in 0.5 ml each
MethoCult murine methylcellulose medium with the addition of SCF, IL-3,
IL-6, and GM-CSF. For double transductions, equal volumes of different
virus supernatants were mixed before infection, and the infected BMCs
were selected simultaneously with G418 and puromycin. Colonies in one
of the duplicate wells were stained by the addition of 50 µl of 1
mg/ml p-iodonitrotetrazolium violet in PBS (INT; Sigma
Chemical Co.). INT is converted by living cells to a brown-violet
insoluble tetrazolium salt. The colonies in the remaining well were
resuspended in medium, counted, and replated as above.
|
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). As expected, this
lead to the outgrowth of a myelomonocytic cell population with a high
content of early myeloid precursor cells (Fig. 1A)
. The presence of unrearranged retroviruses was verified by
Southern blot analysis, and multiple integration events could be
detected (Fig. 1B)
. The presence of MLL-ENL RNA was
confirmed by RT-PCR (Fig. 1C)
. Three different,
independently transduced cell populations were generated for the
subsequent experiments (labeled md11 for murine cells transduced with
derivative 11). Because there was no significant difference in the
experimental results obtained with each of them, only one
representative example is shown. A similar MLL-ENL-transduced cell
population has been shown previously to elicit acute leukemias after
injection into syngeneic mice (18)
.
|
|
. IL-3 deprivation lead to a rapid and efficient cell death
within 24 h. Cells kept without IL-3 showed the characteristic
signs of programmed cell death, including DNA fragmentation and
positive TUNEL staining (Fig. 2, B and C)
. The
cells could be kept continuously in culture for more than 6 months in
medium supplemented with saturating amounts of IL-3 without showing any
signs of senescence. Normal BMCs, transduced with a control retroviral
vector, did not proliferate with IL-3 as growth factor (not shown).
|
. G-CSF also initiated a profound proliferation arrest
(Fig. 3B)
. The visible morphological features of
differentiation were accompanied by a strong up-regulation of
myeloid-specific cell surface markers as detected by FACS staining.
Confirming their primitive myeloid nature, md11 cells that were grown
in IL-3 displayed an intermediate surface level of GR-1 and Mac-1
(CD11b). Addition of G-CSF for 6 days elicited a marked shift of both
markers toward higher expression levels, which indicated the presence
of more mature cells of the myeloid lineage (Fig. 3C)
. A
statistical evaluation of the respective cultures according to
morphological criteria supported the view that the MLL-ENL-dependent
maturation block was not absolute but could be overcome by the addition
of the differentiation factor G-CSF (Fig. 3D)
. However,
occasionally after prolonged culture (>4 months), md11 cells
spontaneously generated subclones that lost the differentiation
capability and that were able to outgrow the remaining population (not
shown).
|
. Two macrophage/monocyte-specific genes coding for a chemokine C-C
receptor and for M-CSF (macrophage colony-stimulating factor)
were up-regulated in response to G-CSF. The RNA levels of the remaining
16 genes decreased in response to this cytokine. The majority of the
genes that were down-regulated after G-CSF treatment coded for
growth-indicators. Several of these are involved in the cell-cycle
regulatory circuit controlled by c-myc. Next to the RNA for
the proto-oncogene c-myc itself, the RNAs of four
established c-myc target genes (coding for the HSP60
mitchondrial heat-shock protein, the blast marker CD71, ornithine
decarboxylase, and the E2F dimerization partner DP-1; Refs. 26
, 27
) were less abundant in G-CSF-treated cells. In addition, the
expression of the proto-oncogene pim-1, which is known to
cooperate with c-myc (28)
, and of the
myc-regulatory transcription factor Nm23-M2
(29)
also decreased under the influence of G-CSF. The
remaining RNAs that were down-regulated after G-CSF treatment
corresponded to the following genes: proliferating cell nuclear
antigen (PCNA), cyclin A, the transcription
factors A10 and ATF4, the heat-shock proteins
HSP86 and HSP84, Rac1,
integrin-ß, and the DNA-binding protein YB1.
|
Myc) or authentic Myc itself. Cells
harboring both viruses were selected by simultaneous addition of G418
and puromycin. The Myc protein is capable of forming heterodimers
with its normal cellular counterpart Max. The Myc/Max heterodimers
are still able to bind to E-box-containing promoters but lack the
transcriptional transactivator function (24)
. As controls,
BMCs were coinfected with MLL-ENL/pMSCV-neo, Myc/pMSCV-pac,
Myc/pMSCV-pac, and with pMSCV-neo/pMSCV-pac (Neo/Puro). The
double-transduced cells were tested for their self-renewing capacity,
the phenotypic and growth properties, and their differentiation
capabilities. In the standard methylcellulose replating assay (Fig. 4A)
, colonies will form in tertiary rounds of plating only if
a cell with self-renewing capabilities has been immortalized.
Nontransformed cells will exhaust their proliferative potential and
terminally differentiate. In this assay, the coexpression of
dominant-negative Myc completely inhibited MLL-ENL-induced colony
formation. Apparently, this was not attributable to a simple
cell cycle and growth suppression because an equal number of primary
MLL-ENL/Myc and MLL-ENL/neo colonies were observed (Fig. 4B)
. In addition, the total cell count of MLL-ENL/Myc and
MLL-ENL/Neo cells was comparable in the primary round of plating (Fig. 4C)
. In contrast, cells cotransduced with MLL-ENL and
wild-type Myc consistently yielded a higher number of colonies in
third-round platings when compared with the MLL-ENL/Neo control cells.
In the absence of MLL-ENL, neither cells expressing Myc nor cells
expressing Myc formed colonies in third-round platings (Fig. 4B
and not shown). Similar results were obtained when the
proliferation of the cells in response to IL-3 was measured 3 days
after explantation from third-round methylcellulose cultures.
MLL-ENL/Myc cells proliferated more vigorously and reached higher cell
densities than MLL-ENL cells (Fig. 4D)
. Both cell
populations could be grown continuously (>4 weeks) in liquid medium
without any signs of decrease in vitality. Cells coexpressing the
dominant-negative Myc, however, proliferated only weekly and stopped
growth completely after 2 weeks in liquid culture. The
Puro/Neo negative control cells did not grow in liquid culture
(not shown). Myc also had an effect on the differentiation potential of
md11 cells. MLL-ENL/Myc cells remained arrested in an earlier state of
myeloid differentiation (Mac-1 negative) and showed less spontaneous
differentiation than MLL-ENL cells (Fig. 5A)
. Additionally,
G-CSF did not induce growth arrest and differentiation of MLL-ENL/Myc
cells but, instead, stimulated a proliferative response similar to that
from IL-3 (Fig. 5B)
. G-CSF treated MLL-ENL/Myc cells
up-regulated surface Mac-1, but only a small subpopulation showed an
increase in GR-1. The generation of a GR-1+
subpopulation after G-CSF treatment was accompanied by the appearance
of a few mature granulocytes in cytospin preparations (Fig. 5, A and C)
. The majority of the cell population,
however, remained arrested in a premature state and continued to
proliferate in the presence of G-CSF. Despite the growth promoting
effects of Myc, it did not render the cells growth-factor independent.
MLL-ENL/Myc cells responded up to an optimum G-CSF concentration of
0.5 ng/ml with an increase in the proliferation rate and did not
grow without this cytokine. Higher than optimal G-CSF concentrations
lead to a slight reduction in the proliferation rate (Fig. 5D)
. The integrity of the Myc and Myc retroviral inserts
was verified by immunoblot analysis of protein extracts prepared from
the viral packaging cells (Fig. 5E)
. The MLL-ENL RNA as well
as the Myc and the Myc transcript was detected by RT-PCR in
MLL-ENL/Myc and MLL-ENL/Myc cells (Fig. 5F)
. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The md11 cells described here are in several aspects significantly different from cells that were established from 11q23 leukemias: (a) a characteristic clinical feature of this type of leukemia is the frequent treatment failure attributable to the resistance of the leukemic cells to high doses of chemotherapeutic agents (30) . Recent reports underscore these observations and describe an unusual resistance of four different patient-derived cell lines against apoptosis (31 , 32) . md11 cells, in contrast, are highly apoptosis sensitive; (b) unlike md11 cells many of the cell lines derived from patients carrying a 11q23 translocation were not dependent on growth factors other than those contained in the serum of the culture medium; and (c) most of these lines did not show any differentiation capability (33 , 34) . These differences are supportive of the hypothesis that the expression of an MLL-ENL fusion protein might only be the first genetic event and has to be followed by additional mutations before acute leukemia arises. These secondary mutations would render the cells finally apoptosis-resistant, growth-factor independent, and refractory toward differentiation stimuli. Interestingly, a series of studies that recreated a t(9;11) translocation in a natural setting by a genomic knock-in strategy observed a distinct two-phase etiology of leukemia (19 , 20) . Dobson et al. (20) detected an elevated level of myeloid cells in the blood of the AF9 knock-in mice shortly after birth but no clinical signs of leukemia were present yet. After up to 6 months of latency, the animals finally succumbed to acute leukemias. This would be consistent with the initial presence of a population of myeloid cells with a partial block in differentiation similar to md11 cells. A subsequent genetic event or events might convert this initial myeloproliferative disease into an acute leukemia. The multistep nature of leukemias with an 11q23 translocation is also corroborated by the discovery of three genes that have been found mutated in leukemia cells that are isolated from patients. Dominant-negative mutants of the transcription factor Ikaros and an abrogation of INK4A expression by promoter methylation were frequent in these leukemias (35 , 36) . In addition, in rare cases, activated ras mutants have been detected (37) .
The experiments presented here suggest that the proto-oncogene c-myc is not only a likely candidate for a secondary mutation but it seems to be also intricately involved in the biological function of MLL-ENL. The nuclear phosphoprotein c-Myc controls cell cycle progression and has a central role in growth regulation. Consequently, it is frequently found as an activated oncogene in human neoplastic disease (38) . The expression of c-myc is normally tightly controlled. Forced myc expression can drive quiescent cells into cycle and is able to block differentiation. When coexpressed with an oncogenic partner such as ras, c-myc can induce transformation of primary cells. A plausible model to explain the effects of Myc on the MLL-ENL-induced transformation would postulate that MLL-ENL cooperates with Myc in a common pathway to activate the transcription of an overlapping set of target genes. Alternatively, MLL-ENL and Myc might act in parallel pathways that converge on essential targets that cannot be sufficiently activated by either MLL-ENL or Myc alone. The intricate connection of MLL fusion proteins and Myc was corroborated in a recent study of Galoian et al. (39) . In these experiments, not only MLL-ENL but also three other MLL fusion proteins functioned as specific transactivators for promoters that contained myc-type E-boxes. Most importantly, this transactivator property of MLL-ENL was essentially dependent on the presence of the Myc protein itself. The MLL fusion proteins did not transactivate E-box promoters in Myc-/- cells. Because the levels of Myc protein were unaltered in cells expressing MLL fusion proteins, the cooperation apparently augments the intrinsic ability of Myc to activate certain target promoters. It is easy to imagine how a moderate increase in the concentration of Myc targets might lead to a reversible differentiation arrest that can be overcome by a strong maturation stimulus, e.g., by G-CSF. An additional overexpression of the Myc protein might amplify the cooperative effect with MLL-ENL and finally drive the cells beyond a threshold level that makes the maturation block irreversible. Although the activation of Myc targets seems to be necessary for the MLL-ENL transformation mechanism, it is not sufficient. An overexpression of Myc alone did not lead to transformation of the hematopoietic cells and, therefore, MLL-ENL must have additional effects that are required to elicit a true transformation. The availability of the bone marrow transduction system will allow us to dissect additional molecular details of the leukemogenesis caused by MLL-ENL and to identify other potentially cooperating oncogenes.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by grants from the Deutsche
Forschungsgemeinschaft (SFB466-C7, SFB473-B10, SL27/4-1) and a Ria
Freifrau-von-Fritsch Stiftung career development award. 
2 To whom requests for reprints should be
addressed, at Department of Genetics, University of Erlangen,
Staudtstrasse 5, 91058 Erlangen, Germany. Phone: 49-9131-8528527; Fax:
49-9131-8528526; E-mail: rslany{at}biologie.uni-erlangen.de
3 The abbreviations used are: G-CSF, granulocyte
colony-stimulating factor; GM-CSF, granulocyte macrophage
colony-stimulating factor; IL, interleukin; SCF, stem cell factor;
RT-PCR, reverse transcription-PCR; BMC, bone marrow cell; MTT,
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; TUNEL,
terminal dideoxy nucleotidyltransferase nick-end labeling; FACS,
fluorescence-activated cell sorter; 5-FU, 5-fluorouracil.
4 Internet address:
http://www.stanford.edu/group/nolan/.
Received 2/13/01. Accepted 6/28/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in acute lymphoblastic leukemia with chromosomal translocation t(4;11).. Leukemia, 13: 1539-1547, 1999.[Medline]
This article has been cited by other articles:
|
|
 |
|
  J. Chen, D. A. Santillan, M. Koonce, W. Wei, R. Luo, M. J. Thirman, N. J. Zeleznik-Le, and M. O. Diaz Loss of MLL PHD Finger 3 Is Necessary for MLL-ENL-Induced Hematopoietic Stem Cell Immortalization Cancer Res., August 1, 2008; 68(15): 6199 - 6207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Behbod and J. M. Rosen Will cancer stem cells provide new therapeutic targets? Carcinogenesis, April 1, 2005; 26(4): 703 - 711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Cartwright, C. McLean, A. Sheppard, D. Rivett, K. Jones, and S. Dalton LIF/STAT3 controls ES cell self-renewal and pluripotency by a Myc-dependent mechanism Development, March 1, 2005; 132(5): 885 - 896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. B. Zeisig, T. Milne, M.-P. Garcia-Cuellar, S. Schreiner, M.-E. Martin, U. Fuchs, A. Borkhardt, S. K. Chanda, J. Walker, R. Soden, et al. Hoxa9 and Meis1 Are Key Targets for MLL-ENL-Mediated Cellular Immortalization Mol. Cell. Biol., January 15, 2004; 24(2): 617 - 628. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Johnson, W. Chen, W. Hudson, Q. Yao, M. Taylor, T. H. Rabbitts, and J. H. Kersey Prenatal and postnatal myeloid cells demonstrate stepwise progression in the pathogenesis of MLL fusion gene leukemia Blood, April 15, 2003; 101(8): 3229 - 3235. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
-Myc antibody; kDa,
Mr in thousands. F, RT-PCR
analysis of the expression of MLL-ENL (left panel), Myc
(middle panel), and 
