Constitutive activation of the Wnt/b-catenin signalling pathway ... | Nature

Oncogene (2005) 24, 2410–2420

& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc

Constitutive activation of the Wnt/b-catenin signalling pathway in acute myeloid leukaemia Maria Simon1, Victoria L Grandage1, David C Linch1 and Asim Khwaja*,1 1

Department of Haematology, Royal Free & University College Medical School, 98 Chenies Mews, London WC1E 6HX, UK

The b-catenin protein is at the core of the canonical Wnt signalling pathway. Wnt stimulation leads to b-catenin accumulation, nuclear translocation and interaction with T-cell factor/lymphoid enhancer factor (TCF/LEF) transcription factors to regulate genes important for embryonic development and proliferation. Wnt/b-catenin can promote stem cell self-renewal and is dysregulated in colon carcinoma. We have examined the role of the Wnt pathway in the development of acute myeloid leukaemia (AML) and find that the b-catenin protein is readily detected in primary AML samples. Using transfection of a TCF/LEF reporter construct into primary AML cells and normal human progenitors, we find increased reporter activity in 16/25 leukaemia samples. Retrovirally mediated expression of a mutant active b-catenin in normal progenitors preserves CD34 expression and impairs myelomonocytic differentiation. Activation of TCF/LEF signalling decreases factor withdrawal-induced apoptosis of normal progenitors. A significant proportion of AML cases show aberrant expression of components of the Wnt pathway including Wnt-1, Wnt-2b and LEF-1. These results provide evidence for the involvement of the Wnt/b-catenin pathway in the pathogenesis of AML. Oncogene (2005) 24, 2410–2420. doi:10.1038/sj.onc.1208431 Published online 14 February 2005 Keywords: leukaemia; b-catenin; CD34; differentiation

Introduction The Wnt family consists of at least 19 secreted proteins that regulate developmental processes (Smalley and Dale, 1999). The modules that transduce Wnt signals are highly conserved throughout evolution and in the canonical Wnt pathway a central role is played by the b-catenin protein. b-Catenin was first identified as a protein that links cell adhesion molecules of the cadherin family to the cortical actin cytoskeleton, but later genetic and biochemical analyses also placed it within the Wnt pathway (Conacci-Sorrell et al., 2002). In the steady state, levels of free (non-cadherin-bound) *Correspondence: A Khwaja; E-mail: [email protected] Received 26 January 2004; revised 21 October 2004; accepted 7 December 2004; published online 14 February 2005

cytosolic b-catenin are suppressed by proteasomal degradation. This process is regulated by a complex of proteins that include axin, adenomatous polyposis coli (APC), casein kinase 1 (CK1a) and glycogen synthase kinase-3 (GSK-3b) (Giles et al., 2003). Phosphorylation of b-catenin by CK1a and GSK-3b on key N-terminal residues targets it for ubiquitination and breakdown in the proteasome. Wnt binding to cognate Frizzled receptors results in the activation of the Dishevelled protein (Dsh) by an unknown mechanism. Dsh suppresses the phosphorylation of b-catenin, resulting in its accumulation and translocation to the nucleus. Here it binds to transcription factors of the T-cell factor/ lymphoid enhancer factor (TCF/LEF) family and stimulates the expression of target genes (Kikuchi, 2003). The Wnt/b-catenin pathway is known to be involved in tumour formation in several cell types. Truncation or loss of the APC protein as seen in colorectal cancer results in the accumulation of b-catenin and increased expression of the cell cycle regulators c-Myc and cyclin D1 (Bienz and Clevers, 2000; Giles et al., 2003; Kikuchi, 2003). Inactivating mutations of axin, with resulting activation of b-catenin signalling, have been described in hepatic carcinoma. Point mutations in the N-terminal regulatory region of b-catenin are found in a number of epithelial tumours. These prevent phosphorylation or ubiquitination of b-catenin and are often found in colorectal tumours that do not have mutations in APC (Giles et al., 2003; Kikuchi, 2003). Recently, overexpression of Dsh has been implicated in the development of mesothelioma and lung cancer (Uematsu et al., 2003a, b). In haemopoiesis, murine yolk sac and fetal liver progenitors have been shown to express Wnt-5a and -10b as well as Frizzled receptors (Austin et al., 1997). Incubation of fetal liver progenitors with conditioned media from cell lines expressing Wnt-1, -5a or -10b led to amplification of cell number. Similar results were reported by Van Den Berg et al. (1998), who found expansion of colony forming cells in response to coculture of progenitors with Wnt expressing feeder cells. Wnt proteins are also mitogenic for B-cell precursors, and the LEF-1 transcription factor plays a key role in B-cell development (Reya et al., 2000). Recently, Weissman’s group has shown that expression of a mutant activated b-catenin in murine haemopoietic progenitors results in an expansion of the haemopoietic

b-Catenin in AML M Simon et al

2411

stem cell (HSC) pool and that inhibition of Wnt signalling results in reduced HSC function (Reya et al., 2003). In this study, we have looked for activation of the Wnt/b-catenin pathway in acute myeloid leukaemia (AML). We have chosen to investigate primary material for its biological relevance; cell lines may not give an accurate representation of normal progenitor biology or of the molecular pathogenesis of AML. This latter point was recently well illustrated by Quentmeier et al. (2003), who showed that mutations of the Flt3 tyrosine kinase are rare in AML cell lines (four of 69), whereas large studies using primary material have indicated that such mutations are present in over 30% of cases (Kottaridis et al., 2001).

Results Expression of b-catenin in normal and leukaemic myeloid cells First we examined expression of b-catenin protein in normal human CD34 þ progenitors induced to undergo myeloid differentiation in liquid culture in response to cytokines SCF, IL-3, G-CSF and GM-CSF. Under these conditions, there is extensive cell division and differentiation with loss of expression of CD34 and acquisition of maturation markers such as CD11b (Watts et al., 2002) (Figure 1). We found that freshly isolated CD34 þ cells expressed readily detectable

Figure 1 Expression of b-catenin in normal haemopoietic progenitors and AML blasts. (a) b-Catenin expression (by Western blot) and expression of cell surface markers (by flow cytometry) were measured in normal human CD34 þ progenitors induced to undergo myeloid differentiation in vitro. (b) Expression of b-catenin in 26 separate cases of AML was compared with that of normal myeloid (NM) progenitors (6 days in myeloid culture). The fraction of cells expressing CD34 for each sample is given below the appropriate lane (ND ¼ not done). Patient identifying number for this study is shown below. (c) Expression of b-catenin in total extracts (T) and nuclear fraction (N) from three cases of AML. Samples have also been probed for Rac1 to show that the nuclear fraction has not been contaminated by membrane/cytoplasmic components Oncogene


2412

b-catenin by Western blot and that this was downregulated by days 5–7 of differentiation (Figure 1). At this stage, the cells have the morphological appearance of myeloblasts – the majority of cells have downregulated CD34, express CD33 but there has not yet been a significant increase in CD11b expression. We compared the expression of b-catenin in 26 cases of AML with that in normal myeloid precursors, which had been in CD34 þ cell-initiated myeloid expansion culture for 6 days. Unlike the case in normal progenitors, we found a variable level of expression of b-catenin in the AML cases (Figure 1), which did not appear to correlate with the level of expression of CD34 as assessed by flow cytometry, that is, there were cases with high levels of b-catenin and low CD34 expression as well as cases with low levels of b-catenin and high CD34 expression. Normally, in cells not exposed to Wnt signals, b-catenin is complexed with cadherin molecules at the cell membrane and links their intracytoplasmic domain to the actin cytoskeleton. Fractionation of AML blasts showed the presence of b-catenin in the nuclear fraction indicating potential transcriptional activation of this pathway (n ¼ 3) (Figure 1). Development of a technique for high-level transient transfection of primary normal progenitors and AML cells Establishing whether a signalling pathway is constitutively active in primary tumour cells, compared with their normal counterparts, is often difficult. The mere presence of components of a pathway cannot be used as

a

Activity of the Wnt/b-catenin/TCF–LEF pathway in normal progenitors and AML blasts Having successfully established high-level transient transfection of primary normal and leukaemic cells, we GFP

GFP vs CD34-PE

GFP

Control

a surrogate for activity. In order to establish if the Wnt/ b-catenin pathway was active in primary normal CD34 þ cells or in AML blasts, we wished to utilize a technique to assay transcriptional activity using a reporter construct originally used to identify activation of this signalling module in cell lines derived from patients with colon carcinoma (Korinek et al., 1997). In order to do this, we needed to develop a method of achieving high levels of transfection in these primary cells. First, we investigated if we could transfect normal CD34 þ progenitors using the technique of modified electroporation known as nucleofection (Trompeter et al., 2003) using a GFP expressing plasmid. We found that normal CD34 þ cells rapidly express GFP following nucleofection with transfected cells detected within 3 h (Figure 2). Larger plasmids including an Akt-GFP fusion vector were also efficiently transfected with overall transfection levels in CD34 þ cells of 5479% (Figure 2). Having established successful transfection in normal progenitors, we used the same technique in primary AML blasts. Mean transfection levels were 3976% (Figure 2) and the CD34 þ fraction of AML cells (which includes putative leukaemia-initiating cells) was transfected to a similar degree as the CD34-negative fraction (data not shown).

CD34

b

Akt-GFP fusion protein

c 70

Phase

GFP

%GFP positive

60 50 40 30 20 10 0 CD34

AML

Figure 2 Efficient transfection of normal CD34 þ progenitors and AML blasts by nucleofection. (a) Normal human CD34 þ progenitors were nucleofected with a control (left panel) or a GFP plasmid (centre panel) and GFP expression assessed by flow cytometry after 3 h. The right panel shows CD34 þ cells nucleofected with a GFP plasmid and examined the following day in combination with surface labelling for CD34. (b) CD34 þ cells were nucleofected with an Akt-GFP fusion plasmid and expression assessed by fluorescence microscopy. (c) CD34 þ normal progenitors or AML blasts were nucleofected with a GFP plasmid and GFP expression assessed by flow cytometry after 24 h. Mean7s.e. results for GFP expression for normal CD34 þ cells (n ¼ 7) and AML cells (n ¼ 14) are shown Oncogene


2413

utilized the luciferase reporter plasmid TOP-FLASH, which is activated in response to binding by TCF/LEF transcription factors, which lie downstream from Wnt/ b-catenin (Korinek et al., 1997). We confirmed that this reporter was functional in the 293T cell line when cotransfected with an active b-catenin mutant (Figure 3a). We transfected TOP-FLASH, or the control plasmid FOP-FLASH, into normal CD34 þ progenitors and myeloid leukaemia blasts. In early normal CD34 þ cells (days 1–3 of culture), where bcatenin was detected at the protein level, TOP-FLASH reporter activity averaged 1.570.4 (fold activity relative to that obtained with FOP-FLASH, n ¼ 3 separate donors). b-Catenin is regulated by phosphorylation of N-terminal residues by GSK-3b in a multiprotein complex including APC and axin resulting in its degradation via the proteasome. We found that TOPFLASH activity in normal CD34 þ cells could be significantly increased by incubation with the GSK-3b

a Ratio TOP/FOP-FLASH

15 10

6 5 4 3 2 1 0

4 2 0

5 control beta-catenin

CD34+

CD34 plus LiCl

1 2 3 4 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

e

Ratio TOP/FOP-FLASH

Ratio TOP/FOP-FLASH Ratio TOP/FOP-FLASH

d

12 10 8 6

20

0 20 18 16 14 12 10 8 6 4 2 0

b

293T

25

c

inhibitor lithium chloride (Davies et al., 2000) (Figure 3b). Increased TOP-FLASH reporter activity compared with normal progenitors was detected in a high proportion of cases with AML in the absence of any exogenous growth factors (Figure 3c). Using an arbitrary cutoff of twice the average seen with normal CD34 þ cells, increased reporter activity was detected in 16 out of 25 cases. Increased activity was seen in 6/7 cases with an activating internal tandem duplication of the Flt3 tyrosine kinase compared with 10/18 without an Flt3 mutation, a difference in incidence that did not reach statistical significance (P ¼ 0.06). To confirm that the TCF/LEF pathway was present and could be regulated in primary AML cells, we incubated blast cells transfected with the TOP-FLASH reporter with purified recombinant Wnt-3a and found an increase in reporter activity (Figure 3d). This increase in reporter activity could also be achieved by transfection of a constitutively active mutant b-catenin. Further, we were

control Wnt-3a S37A beta-catenin

Case 1

Case 2

6 5 4 3 2 1 0

- + CD34

control Axin

Case 3

Case 4

Figure 3 Activation of a TCF/LEF reporter in normal CD34 þ progenitors and in AML cells. (a) A control vector or a mutant active b-catenin (S37A) was coexpressed with the TCF/LEF reporter TOP-FLASH or a control reporter FOP-FLASH in 293T cells and luciferase activity measured after 24 h. The result is shown as a ratio of TOP-FLASH/FOP-FLASH activity after correction for transfection efficiency. No result is shown for case 5 due to very high levels of cytotoxicity following transfection. (b) Normal CD34 þ progenitors were nucleofected with the TOP-FLASH reporter and incubated overnight without or with the GSK3 inhibitor lithium chloride (10 mM). (c) Blasts from 25 AML cases were nucleofected with the TOP-FLASH reporter and luciferase activity measured in the absence of exogenous growth factors. Hatched bars highlight patients with Flt3 mutations. Filled bars show values for normal CD34 samples incubated without () or with ( þ ) added cytokines. An arbitrary cutoff of twice the average seen with normal CD34 þ cells (n ¼ 3) is shown as a horizontal line. Patient identifying numbers for this study are shown on the x-axis and match those in Figures 1 and 5. (d) Effect of the addition of Wnt-3a (100 ng/ml) or the coexpression of constitutively active b-catenin on the activation of TOPFLASH in two cases of AML. (e) Expression of axin inhibits the constitutive basal TOP-FLASH activity seen in two cases of AML Oncogene


2414

able to show that transfection of axin, which inhibits the b-catenin/TCF pathway, reduced activity of the TOPFLASH reporter (Figure 3e). Effect of expressing a mutant b-catenin on the proliferation and differentiation of normal human CD34 þ progenitors Having shown that the TCF/LEF pathway is constitutively active in a significant proportion of AML cases, we wished to investigate the impact this could have on normal myeloid development in order to assess its

a

potential role in leukaemogenesis. We expressed a constitutively active b-catenin (S37A mutant) in normal primary human CD34 þ cells by retroviral transduction (Orford et al., 1997). The S37A b-catenin mutant cannot be fully phosphorylated by GSK-3b and is therefore less susceptible to proteasomal degradation. Using a bicistronic vector expressing GFP from an IRES, CD34 þ cells expressing b-catenin or a control vector were sorted by flow cytometry and incubated in cytokines including GM-CSF and G-CSF in liquid culture (Figure 4). Cells were counted and immunophenotyped with CD34, CD11b (to examine granulo-monocytic differ-

b β-catenin

Day 3

IκBα %positive

5

20

2.5 control beta-catenin

70 *

40 *

30

0

4

6

8

β-catenin

25

GFP

GFP

control

β-catenin

GFP

GFP

12

30 * CD14

%positive

control beta-catenin

20 10

d

0

CD11b

%positive

control

control beta-catenin

50

*

7.5

25

15

60

10

* 30

Con Con α-cat α-cat

c

Day 6

35

20 15

*

*

10 5 0

4

6

8

12

e

control

β-catenin

Figure 4 Activation of the b-catenin pathway impairs the differentiation of normal human haemopoietic progenitors. GFP expressing cells were flow sorted after transfection of normal haemopoietic progenitors with control or active S37A b-catenin vectors and placed in myelomonocytic differentiation cultures. (a) Expression of b-catenin by Western blot in control and S37A b-catenin transfected primary human haemopoietic progenitors after 6 days in myeloid culture (duplicate cultures). (b) Expression of CD34 in control and S37A b-catenin expressing cells after 3 or 6 days in culture (left two panels, mean7s.e., n ¼ 6, *Po0.05 by Wilcoxon’s test). (c, d) Expression of myelomonocytic markers CD11b (c) and CD14 (d) in control and S37A b-catenin expressing cells incubated in myeloid differentiation culture (representative flow cytometric histograms and mean7s.e. of eight experiments). *Po0.05 by Wilcoxon’s test. (e) Morphology of cytospins of cells expressing control vector or activated b-catenin after 14 days in culture (May–Grunwald–Giemsa stain) Oncogene


2415

a Wnt1

Wnt2B LEF1

Jurkat

CD34+

GAPDH 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 patient number --------------------------------------------------AML cases-------------------------------------------------

b Jurkat

CD34+

Wnt10B 1 2 3 4 5 6 7 8 9 10 patient number --------------AML cases--------------

Figure 5 Expression by RT–PCR of Wnt pathway components in AML blasts and normal progenitors. (a) Expression by RT–PCR of LEF-1, Wnt-2b, Wnt-1 and GAPDH in AML blasts, normal CD34 þ cells and a positive control (Jurkat cell line). Patient identifying number for this study is shown below. (b) Expression by RT–PCR of Wnt-10b in a subset of AML blasts, normal CD34 þ cells and a positive control (Jurkat cell line)

entiation) and CD14 (monocytic differentiation) antibodies at various time points. Proliferation under these conditions was equivalent in control and active bcatenin expressing cells (Figure 4). Cells expressing active b-catenin maintained higher levels of expression of CD34 compared with control cells (Figure 4b). Cells expressing active b-catenin had impairment of myelomonocytic differentiation with attenuated expression of CD11b and CD14 (Figure 4c and d). Morphological examination confirmed these findings with a reduction in mature monocyte/macrophages in particular in cells expressing b-catenin (Figure 4e).

mutations in the APC gene did not reveal abnormalities in the regions that are frequently mutated in colon cancer. We found that a significant proportion of AML cases were positive by RT–PCR for the LEF-1 transcription factor, whereas expression in normal progenitors was not detectable (Figure 5). In addition, there was constitutive expression of Wnt-1 and Wnt-2b in a number of AML samples (Figure 5). Both these Wnt family members can activate signalling via the bcatenin pathway. Wnt-10b was expressed in both normal and leukaemic progenitors.

Effect of activation of the TCF/LEF pathway on progenitor cell survival

Discussion

Activation of b-catenin/TCF–LEF signalling by lithium chloride inhibition of GSK-3b resulted in partial protection of CD34 þ cells from factor withdrawalinduced apoptosis. Cell survival measured by flow cytometric evaluation of annexin V binding/propidium iodide exclusion was 3777% in cells incubated without growth factors for 48 h compared with 5878% in the presence of 10 mM lithium chloride (n ¼ 4, P ¼ 0.04). Normal CD34 þ progenitors expressing retrovirally transduced active b-catenin were also protected from apoptosis with 48% survival at 72 h compared with 17% for control vector expressing cells. Investigation of potential mechanisms by which b-catenin/ TCF signalling is upregulated in AML Sequencing of the N-terminal region of b-catenin, which is mutated in colon cancer, in the AML cases did not reveal any abnormality. Further, investigation for

The Wnt signalling pathway is a key regulator of a number of embryonic developmental events and contributes to tumour formation when it is constitutively activated (Conacci-Sorrell et al., 2002; Moon et al., 2002; Giles et al., 2003). In addition to their role in the embryo, Wnts are also involved in stem cell self-renewal and differentiation throughout life. This is best characterized in the intestine where the proliferation and differentiation of stem cells at the base of colonic crypts is controlled by the b-catenin/TCF pathway (Batlle et al., 2002; van de Wetering et al., 2002). Mutations resulting in constitutive activation of b-catenin/TCF lead to impaired differentiation and the maintenance of a crypt progenitor-like phenotype (van de Wetering et al., 2002). b-Catenin/TCF also appears to regulate the epidermal stem cell compartment with activation associated with increased self-renewal and impaired differentiation (Huelsken et al., 2001). In lymphopoiesis, Wnt/LEF-1 signalling appears to be involved in proliferation of the B-cell precursor compartment (Reya et al., 2000). Oncogene


2416

There is some recent controversy as to the role of Wnt/b-catenin in haemopoiesis: work from Weissman and colleagues showed that Wnt/b-catenin may play a significant part in regulation of HSC function (Reya et al., 2003; Willert et al., 2003). They demonstrated that expression of a mutant b-catenin in murine HSC leads to their expansion in vitro and an increased ability to reconstitute haemopoiesis in transplantation experiments (Reya et al., 2003). Overexpression of axin, an inhibitory molecule in this pathway, resulted in inhibition of HSC function (Reya et al., 2003). In addition, exposure of HSC to purified Wnt-3a in vitro resulted in their expansion (Willert et al., 2003). In contrast, using inducible Cre-loxP-mediated inactivation of b-catenin, Cobas et al. (2004) have shown that bone marrow progenitor self-renewal and multilineage repopulation capabilities are not dependent on this pathway. Wnt-5a has been shown to increase human HSC repopulating ability (Murdoch et al., 2003), but may also act as a tumour suppressor in haemopoietic tissue (Liang et al., 2003). This effect may be independent of b-catenin however (Liang et al., 2003); indeed Wnt-5a has been shown in some circumstances to oppose the canonical Wnt/b-catenin signalling pathway (Topol et al., 2003). We have found that b-catenin is expressed in primary human CD34 þ progenitor cells and is downregulated with myeloid differentiation such that CD33 þ CD34 cells that are at the myeloblast stage have very low/ undetectable protein levels. Primary AML blasts have a range of expression of b-catenin, and this does not appear to correlate with CD34 expression, indicating that the link of b-catenin downregulation with myeloid differentiation may be uncoupled in leukaemia. In addition, we have found that b-catenin is detectable in the nuclear fraction of AML blasts, which is suggestive of involvement in transcriptional activation (Giles et al., 2003). These findings of altered expression patterns of b-catenin have also recently been described in progenitor cells from patients with blast-crisis transformation of chronic myeloid leukaemia (CML) (Jamieson et al., 2004). Haemopoietic progenitor cell self-renewal and differentiation are mediated by the interaction of a number of transcription factors and cell fate may depend on relatively subtle changes in these pathways. In order to assess the activity of the Wnt pathway in normal CD34 þ progenitors and AML blasts, we measured changes in transcriptional activity of this pathway using a TCF/LEF reporter (Korinek et al., 1997). This is a well-established technique in other cell types, in particular in colon carcinoma cell lines, but requires efficient transfection of a reporter construct, which is notoriously difficult in primary haemopoietic cells. We utilized the nucleofection technique, which has been reported to be effective in other primary cell types (Lai et al., 2003; Trompeter et al., 2003), and were able to achieve high levels of transient transfection in both CD34 þ cells and in AML blasts. This showed a low but detectable level of TCF/LEF reporter activity in normal CD34 þ cells. Elevated reporter activity (Xtwo-fold the Oncogene

average of the normal CD34 þ cell level) was found in 16 of 25 cases of AML. In six of seven cases with an Flt3-ITD, there was increased activity but this was not statistically significantly different from those without a mutation. It has been reported that the Wnt pathway can be activated by signalling via hepatocyte growth factor receptor tyrosine kinase (c-met) as well as by the Ron tyrosine kinase (Danilkovitch-Miagkova et al., 2001). It remains to be seen if Flt3 can directly activate this pathway. Wnt/b-catenin has previously been implicated in the development of human leukaemia: McWhirter et al. (1999) showed that the E2a-Pbx fusion protein found in precursor B-acute lymphoblastic leukaemia (ALL) activates the expression of Wnt-16 resulting in a potential autocrine signalling loop. Chung et al. (2002) showed that b-catenin protein was expressed in leukaemia cell lines and primary cells to varying levels. Inhibition of the Wnt/b-catenin pathway in Jurkat T-lymphoblastic cells impaired proliferation and increased susceptibility to apoptosis in response to Fas ligation. Recent papers have shown the involvement of g-catenin in AML associated with translocations involving AML1-ETO and PML-RARa (Muller-Tidow et al., 2004; Zheng et al., 2004). Jamieson et al. (2004) have shown that granulocyte–macrophage progenitors in blast-crisis CML have enhanced self-renewal capability as a result of the aberrant activation of the Wnt/ b-catenin pathway. One of the defining features of AML is impaired differentiation of the malignant cells to mature granulocytes and monocytes. In some cases, this is associated with specific chromosomal abnormalities affecting transcription factors that play a role in normal myeloid differentiation. These include translocations involving core binding factors such as the AML1-ETO fusion in t(8;21) and the CBFb/MYH11 fusion in inv16 and these fusion proteins have a dominant-inhibitory effect on the product of the remaining normal CBF allele (Kelly and Gilliland, 2002). We expressed a mutant form of bcatenin (S37A mutant), which cannot be fully phosphorylated by GSK-3b and is therefore less susceptible to proteasomal degradation (Orford et al., 1997), in normal primary human CD34 þ cells by retroviral transduction. We found that under conditions of maximal cytokine stimulation, progenitor cells expressing mutant b-catenin had impaired myelomonocytic differentiation. There was no significant effect of mutant b-catenin on proliferation under these culture conditions: this could be due to the maximal rate of cell division in response to high cytokine concentrations in vitro. The characteristics of altered self-renewal properties demonstrated by Reya et al. (2003) and abnormal differentiation shown here are typical of cancer ‘stem’ cells (Reya et al., 2001). It has been shown that the Wnt/ b-catenin pathway may be abnormally activated in malignancy by a number of different molecular mechanisms. These include the loss of APC or axin (or their functional inactivation), mutations in the b-catenin gene resulting in protein stabilization, and aberrant expression of Wnt family members Dishevelled or LEF-1 (Giles et al., 2003; Kikuchi, 2003; Uematsu et al.,


2417

2003a). We could not detect mutations in b-catenin or APC in our AML samples but could detect aberrant expression of Wnt-1, Wnt-2b and LEF-1. Further work is in progress to characterize other elements of the Wnt/ b-catenin signalling module in these samples. It may be that a number of different upstream genetic lesions can lead to the activation of the Wnt/b-catenin/TCF axis. A large number of b-catenin/TCF target genes have been identified and in colorectal cells include c-myc and cyclin D1 (Giles et al., 2003; Kikuchi, 2003). Overall, it has been suggested that b-catenin/TCF targets often work to repress differentiation (Giles et al., 2003; Kikuchi, 2003). Reya et al. (2003) found moderate upregulation of HOXB4 and Notch1 in murine progenitors expressing activated b-catenin. Both these factors have been implicated in stem cell selfrenewal (Varnum-Finney et al., 2000; Antonchuk et al., 2002; Stier et al., 2002). Preliminary microarray gene expression results from human CD34 þ cells transfected with a mutant b-catenin as detailed above show a reduction in expression of myelomonocytic and erythroid specific genes (M Simon and A Khwaja, unpublished data). In conclusion, by using experimental techniques designed to examine primary human haemopoietic tissues, we have found that the Wnt/b-catenin/TCF pathway is constitutively active in many cases of AML. We show that activating this pathway in normal progenitors results in impaired myelomonocytic differentiation. In conjunction with recent data showing that the Wnt/b-catenin pathway can promote the selfrenewal of normal stem cells and leukaemic progenitors (Reya et al., 2003; Jamieson et al., 2004), these results suggest that aberrant activation of this pathway may contribute significantly to the malignant phenotype of AML, which is typified by impaired differentiation and dysregulated self-renewal and survival signalling. The identification of Wnt/b-catenin/TCF abnormalities in AML provides a novel target for the development of rationally designed therapeutics. Materials and methods CD34 and AML cell culture Human CD34 þ cells were isolated from the peripheral blood of haematologically normal individuals undergoing stem cell mobilization by positive immunomagnetic selection (Watts et al., 2002). Samples were obtained according to local ethics committee guidelines. CD34 þ cells were induced to undergo myeloid differentiation as previously described in the presence of SCF, IL-3, GM-CSF and G-CSF (Watts et al., 2002). AML blasts were isolated from the peripheral blood or bone marrow of patients at first presentation, frozen in liquid nitrogen in 10% DMSO and recovered for experiments. Samples with poor recovery or very low cell numbers were discarded and constituted five of 31 samples leaving 26 cases for evaluation (Table 1). Samples constituted more than 90% blasts. There were no patients with FAB types M3 or M6 and seven cases had Flt3-ITD detectable by PCR (Kottaridis et al., 2001). Only one patient had an Flt3-D835 mutation – this was in addition to an Flt3-ITD. Cells were routinely cultured in RPMI þ 10% FCS.

Cell lysates and Western blots Total cell lysates and nuclear extracts were made as previously described (Ardeshna et al., 2000). Antibodies used were b-catenin (BD-Transduction Labs), tubulin (Roche), IkBa (Santa Cruz) and Rac1 (BD-Transduction Labs). Nucleofection The nucleofector device and human CD34 kits were obtained from Amaxa (Koeln, Germany) and used in accordance with the manufacturer’s protocol (U-08). For GFP expression, 1–5 106 CD34 þ cells or AML blasts were used with 1–2 mg of the pEGFPN1 plasmid (Clontech) or a plasmid encoding an Akt-GFP fusion protein (provided by J Downward, CRUK, London). Cells were analysed by flow cytometry (Epics, Beckman-Coulter) or fluorescence microscopy. Reporter assays The TOP-FLASH and FOP-FLASH luciferase reporter constructs were provided by H Clevers (Utrecht, Holland) and the Renilla luciferase plasmid (pRL-CMV) was obtained from Promega. A total of 1–5 106 CD34 þ cells or AML blasts were nucleofected with 2 mg of either TOP-FLASH or FOP-FLASH plus 0.4 mg of pRL-CMV and incubated overnight in RPMI þ 10% FCS for AML cells or IMDM þ 20% FCS supplemented with or without cytokines for normal CD34 þ cells. TOP-FLASH or FOP-FLASH activity was corrected for Renilla activity using the Dual Luciferase Kit (Promega) and results are expressed as a ratio of corrected TOP-FLASH/FOP-FLASH. The axin construct was kindly provided by Dr F Costantini (Columbia University, New York) (Fagotto et al., 1999). Wnt-3a was purchased from R&D Systems. Retroviral transduction and flow cytometry The S37A b-catenin mutant was provided by S Byers (Georgetown University, Washington, DC, USA) and was cloned into the PMX-IRES-GFP vector. FLYRD18 producer cell lines expressing empty vector or b-catenin were generated as described (Kelly et al., 2000) and flow-sorted twice to obtain high GFP expressing pools using an EPICS Elite flow cytometer (Beckman-Coulter). Peripheral blood CD34 þ cells were prestimulated for 48 h in IMDM þ 20% FCS with 50 ng/ ml SCF, 50 ng/ml Flt3-L, 10 ng/ml IL-3 and 10 ng/ml IL-6. Six-well plates were coated with Retronectin (Takara, Shiga, Japan) at 20 mg/ml for 2 h, blocked with 2% BSA in PBS for 1 h, washed and loaded twice with retroviral supernatant from FLYRD18 producer cells at 30 min intervals. CD34 þ cells were then added and the infection repeated 24 h later. Transduction rates ranged between 15 and 65%. GFP-positive cells were then isolated by flow-sorting 48 h after the first infection. Cells were immunophenotyped at intervals using PEconjugated CD34 (BD Pharmingen), CD33, CD14 and CD11b (all Dako). PCR and RT–PCR Identification of Flt3-ITD, D835 mutation (Kottaridis et al., 2002) and N-terminal b-catenin mutations (Abraham et al., 2002) by PCR was as described. APC mutations were screened for by using heteroduplex analysis. DNA was amplified using Optimase Polymeraset (Transgenomic Limited, Crewe, UK). Primers for APC covering nt 762–1181 (1F 50 -GCAGAG AGGTCATCTCAGAAC-30 , 1R 50 -GGCCCGAGCCTCTTT ACTGCC-30 ), nt 3874–4315 (2F 50 -CTTTGTCATCAGCT Oncogene

2418

Oncogene

Table 1 FAB type

01 02 03 04 05 06 07 08 09 10 11 12

M5a 21 MDS M2 M1 M5 M1 M2 M5b M2 N/A M4 M5 (secondary AML) M5a M2 M2

89 38 47 80 257 33 10 52 117 NK 165 44

46,XX N/A Trisomy 21 46,XX 46,XY 46,XY Trisomy 11 46,XX del(9) q12q34 NK Constitutional 45,X 46,XX

162 47 8

16 17 18 19 20 21

M5 M1

69 140

M2 M1 M2 (relapsed)

42 34 36

22 23

M1 M2

14 35

24 25 26

M1 M4 Transformed ET

13 14 15

Presenting WBC

401 233 146

N/A: not applicable; NK: not known

Cytogenetics

Age

Flt3/Ras status

Length of first CR

Current status

Survival

Cause of death

64 64 43 51 45 38 43 66 37 66 67 60

ITD/EX1 WT/WT ITD+D835/WT WT/EX1 ITD/WT ITD/WT ITD/WT WT/WT WT/EX2 WT/WT ITD/WT WT/WT

11 refractory N/A 18 months 16 months N/A >46 months 7 months 11 Refractory Transplant in first CR NK N/A 11 refractory

Dead Dead Alive Dead N/A Alive Alive Dead Dead NK Dead Dead

2 weeks 8 months 4 years 19 months

Leukaemia Aplasia CR2 Leukaemia

46 months 46 months 6 weeks 11 months NK 2 days 6 months

CR1 CR2 Leukaemia TRM NK Leukaemia Leukaemia

46,XY,del(10) No dividing cells 45,XY,add(8)(q22)add(15)(p1)add(21)(p1) Complex 46,XY

40 49 69

WT/EX1+2 WT/WT WT/EX2

11 refractory >5 years 23 months

Dead Alive Dead

1 month 5 years 40 months

Haemorrhage CR1 Leukaemia

41 62

WT/WT ITD/WT

6 months 12 months

Dead Alive

9 months 4 years

Leukaemia (CNS) CR4

t(8;22) N/A 46,XX,r(7)(?p?q)t(9;22)(q34;q11) 46,XX 46,XY

27 72 61

WT/WT WT/WT WT/WT

Transplant in CR1 Treated palliatively 11 refractory

Alive Dead Dead

37 months 1 month 6 weeks

CR1 Leukaemia Leukaemia

43 69

NA/NA WT/EX1

Alive

30 months

CR1

46,XX Inv16(p13q22) Complex

28 56 81

ITD/WT WT/EX2 WT/WT

Transplant in CR1 >6 months. Lost to follow-up Mini allo in CR1 >28 months Treated nonintensively

Dead Alive Dead

7 months 28 months 1 month

TRM CR1 Leukaemia


Patient number

Patient characteristics


2419 GAAGATG-30 , 2R 50 -CTATCTGGAAGATCACTGGGG30 ) and nt 4250–4640 (3F 50 -CGTTCAGAGTGAACCATG CAG-30 , 3R 50 -CCCATTGTCATTTTCCTGAAC-30 ) were used. PCR products were gel purified, denatured and analysed using denaturing HPLC on a Transgenomic WAVEt. Patterns from patient samples were compared with those from DNA from normals and from the SW480 colon carcinoma cell line as a positive control. PCR products from samples with altered patterns were directly sequenced using the CEQt DTCS Quick Start kit and a CEQt 8000 Genetic Analysis System (Beckman Coulter Inc., Fullerton, USA). cDNA was generated from RNA isolated from normal or leukaemic cells using an Ambion kit (Ambion, Huntingdon, UK) and PCR carried out for LEF-1 (50 -TTCTCCACCCA

TCCCGAGAAC-30 forward primer and 50 -CTGAGGCTTC ACGTGCATTTAG-30 reverse primer), Wnt-1 (50 -ATGGGG CTCTGGGCGCTGTTG-30 forward and 50 -CACACGTGCA GGATTCGATGG-30 reverse), Wnt-2b (50 -CACCTGCTGG CGTGCACTCTCAGA-30 forward and 50 -GGGCTTTGCAA GTATGGACGTCCACAGTA-30 reverse) andWnt-10b (50 GGAGGGCGGCCCCAGAGTTCC-30 forward and 50 -AAG CTGCCACAGCCATCCAACAGG-30 reverse). Acknowledgements MS was supported by Leukaemia Research, UK, VLG by the Kay Kendall Leukaemia Fund and AK by the Medical Research Council, UK.

References Abraham SC, Klimstra DS, Wilentz RE, Yeo CJ, Conlon K, Brennan M, Cameron JL, Wu TT and Hruban RH. (2002). Am. J. Pathol., 160, 1361–1369. Antonchuk J, Sauvageau G and Humphries RK. (2002). Cell, 109, 39–45. Ardeshna KM, Pizzey AR, Devereux S and Khwaja A. (2000). Blood, 96, 1039–1046. Austin TW, Solar GP, Ziegler FC, Liem L and Matthews W. (1997). Blood, 89, 3624–3635. Batlle E, Henderson JT, Beghtel H, van den Born MM, Sancho E, Huls G, Meeldijk J, Robertson J, van de Wetering M, Pawson T and Clevers H. (2002). Cell, 111, 251–263. Bienz M and Clevers H. (2000). Cell, 103, 311–320. Chung EJ, Hwang SG, Nguyen P, Lee S, Kim JS, Kim JW, Henkart PA, Bottaro DP, Soon L, Bonvini P, Lee SJ, Karp JE, Oh HJ, Rubin JS and Trepel JB. (2002). Blood, 100, 982–990. Cobas M, Wilson A, Ernst B, Mancini SJ, MacDonald HR, Kemler R and Radtke F. (2004). J. Exp. Med, 199, 221–229. Conacci-Sorrell M, Zhurinsky J and Ben-Ze’ev A. (2002). J. Clin. Invest., 109, 987–991. Danilkovitch-Miagkova A, Miagkov A, Skeel A, Nakaigawa N, Zbar B and Leonard EJ. (2001). Mol. Cell. Biol., 21, 5857–5868. Davies SP, Reddy H, Caivano M and Cohen P. (2000). Biochem. J., 351, 95–105. Fagotto F, Jho E, Zeng L, Kurth T, Joos T, Kaufmann C and Costantini F. (1999). J. Cell Biol., 145, 741–756. Giles RH, van Es JH and Clevers H. (2003). Biochim. Biophys. Acta, 1653, 1–24. Huelsken J, Vogel R, Erdmann B, Cotsarelis G and Birchmeier W. (2001). Cell, 105, 533–545. Jamieson CH, Ailles LE, Dylla SJ, Muijtjens M, Jones C, Zehnder JL, Gotlib J, Li K, Manz MG, Keating A, Sawyers CL and Weissman IL. (2004). N. Engl. J. Med., 351, 657–667. Kelly LM and Gilliland DG. (2002). Annu. Rev. Genomics Hum. Genet., 3, 179–198. Kelly PF, Vandergriff J, Nathwani A, Nienhuis AW and Vanin EF. (2000). Blood, 96, 1206–1214. Kikuchi A. (2003). Cancer Sci., 94, 225–229. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW, Vogelstein B and Clevers H. (1997). Science, 275, 1784–1787. Kottaridis PD, Gale RE, Frew ME, Harrison G, Langabeer SE, Belton AA, Walker H, Wheatley K, Bowen DT, Burnett AK, Goldstone AH and Linch DC. (2001). Blood, 98, 1752–1759.

Kottaridis PD, Gale RE, Langabeer SE, Frew ME, Bowen DT and Linch DC. (2002). Blood, 100, 2393–2398. Lai W, Chang CH and Farber DL. (2003). J. Immunol. Methods, 282, 93–102. Liang H, Chen Q, Coles AH, Anderson SJ, Pihan G, Bradley A, Gerstein R, Jurecic R and Jones SN. (2003). Cancer Cell, 4, 349–360. McWhirter JR, Neuteboom ST, Wancewicz EV, Monia BP, Downing JR and Murre C. (1999). Proc. Natl. Acad. Sci. USA, 96, 11464–11469. Moon RT, Bowerman B, Boutros M and Perrimon N. (2002). Science, 296, 1644–1646. Muller-Tidow C, Steffen B, Cauvet T, Tickenbrock L, Ji P, Diederichs S, Sargin B, Kohler G, Stelljes M, Puccetti E, Ruthardt M, deVos S, Hiebert SW, Koeffler HP, Berdel WE and Serve H. (2004). Mol. Cell. Biol., 24, 2890–2904. Murdoch B, Chadwick K, Martin M, Shojaei F, Shah KV, Gallacher L, Moon RT and Bhatia M. (2003). Proc. Natl. Acad. Sci. USA, 100, 3422–3427. Orford K, Crockett C, Jensen JP, Weissman AM and Byers SW. (1997). J. Biol. Chem., 272, 24735–24738. Quentmeier H, Reinhardt J, Zaborski M and Drexler HG. (2003). Leukemia, 17, 120–124. Reya T, Duncan AW, Ailles L, Domen J, Scherer DC, Willert K, Hintz L, Nusse R and Weissman IL. (2003). Nature, 423, 409–414. Reya T, Morrison SJ, Clarke MF and Weissman IL. (2001). Nature, 414, 105–111. Reya T, O’Riordan M, Okamura R, Devaney E, Willert K, Nusse R and Grosschedl R. (2000). Immunity, 13, 15–24. Smalley MJ and Dale TC. (1999). Cancer Metast. Rev., 18, 215–230. Stier S, Cheng T, Dombkowski D, Carlesso N and Scadden DT. (2002). Blood, 99, 2369–2378. Topol L, Jiang X, Choi H, Garrett-Beal L, Carolan PJ and Yang Y. (2003). J. Cell Biol., 162, 899–908. Trompeter HI, Weinhold S, Thiel C, Wernet P and Uhrberg M. (2003). J. Immunol. Methods, 274, 245–256. Uematsu K, He B, You L, Xu Z, McCormick F and Jablons DM. (2003a). Oncogene, 22, 7218–7221. Uematsu K, Kanazawa S, You L, He B, Xu Z, Li K, Peterlin BM, McCormick F and Jablons DM. (2003b). Cancer Res., 63, 4547–4551. van de Wetering M, Sancho E, Verweij C, de Lau W, Oving I, Hurlstone A, van der Horn K, Batlle E, Coudreuse D, Haramis AP, Tjon-Pon-Fong M, Moerer P, van den Born M, Soete G, Pals S, Eilers M, Medema R and Clevers H. (2002). Cell, 111, 241–250. Oncogene


2420 Van Den Berg DJ, Sharma AK, Bruno E and Hoffman R. (1998). Blood, 92, 3189–3202. Varnum-Finney B, Xu L, Brashem-Stein C, Nourigat C, Flowers D, Bakkour S, Pear WS and Bernstein ID. (2000). Nat. Med., 6, 1278–1281. Watts MJ, Somervaille TC, Ings SJ, Ahmed F, Khwaja A, Yong K and Linch DC. (2002). Br. J. Haematol., 118, 117–123.

Oncogene

Willert K, Brown JD, Danenberg E, Duncan AW, Weissman IL, Reya T, Yates III JR and Nusse R. (2003). Nature, 423, 448–452. Zheng X, Beissert T, Kukoc-Zivojnov N, Puccetti E, Altschmied J, Strolz C, Boehrer S, Gul H, Schneider O, Ottmann OG, Hoelzer D, Henschler R and Ruthardt M. (2004). Blood, 103, 3535–3543.