Original Article

Leukemia (2006) 20, 1829–1839. doi:10.1038/sj.leu.2404342; published online 3 August 2006

The Mll-Een knockin fusion gene enhances proliferation of myeloid progenitors derived from mouse embryonic stem cells and causes myeloid leukaemia in chimeric mice

C T Kong1,2,5, M H Sham2,5, C W E So3, K S E Cheah2, S J Chen4 and L C Chan1

  1. 1SH Ho Foundation Research Laboratories in the Department of Pathology, Hong Kong Jockey Club Clinical Research Centre, Hong Kong SAR, China
  2. 2Department of Biochemistry, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong SAR, China
  3. 3Section of Haemato-Oncology, Institute of Cancer Research, Sutton, UK
  4. 4State Key Lab of Medical Genomics, Shanghai Institute of Hematology, Ruijin Hospital Affiliated to School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Correspondence: Professor LC Chan, Department of Pathology, Division of Haemotology, The University of Hong Kong, University Pathology Building, Queen Mary Hospital, 102 Pokfulam Road, Pokfulam, Hong Kong SAR, China. E-mail: chanlc@pathology.hku.hk

5These authors contribute equally to this work.

Received 24 March 2006; Revised 14 June 2006; Accepted 21 June 2006; Published online 3 August 2006.

Top

Abstract

Rearrangement of the mixed lineage leukaemia (MLL) gene with extra eleven nineteen (EEN) was previously identified in an infant with acute myeloid leukaemia. Using homologous recombination, we have created a mouse equivalent of the human MLL-EEN allele and showed that when MllEen/+ embryonic stem (ES) cells were induced to differentiate in vitro into haemopoietic cells, there was increased proliferation of myeloid progenitors with self-renewal property. We also generated MllEen/+ chimeric mice, which developed leukaemia displaying enlarged livers, spleens, thymuses and lymph nodes owing to infiltration of MllEen/+-expressing leukemic cells. Immunophenotyping of cells from enlarged organs and bone marrow (BM) of the MllEen/+ chimeras revealed an accumulation of Mac-1+/Gr-1- immature myeloid cells and a reduction in normal B- and T-cell populations. We observed differential regulation of Hox genes between myeloid cells derived from MllEen/+ ES cells and mouse BM leukemic cells which suggested different waves of Hox expression may be activated by MLL fusion proteins for initiation (in ES cells) and maintenance (in leukemic cells) of the disease. We believe studies of MLL fusion proteins in ES cells combined with in vivo animal models offer new approaches to the dissection of molecular events in multistep pathogenesis of leukaemia.

Keywords:

MLL-EEN, mouse model, myeloid leukaemia, multistep pathogenesis

Top

Introduction

The mixed lineage leukaemia (MLL) gene (also referred to as HRX, ALL-1 and Htrx) located on human chromosome 11q23 is a major regulator of Hox homeobox genes, which are required for normal embryonic development and establishment of haemopoiesis.1, 2, 3, 4 The MLL gene is frequently rearranged as a result of chromosomal translocations in leukaemias to create chimeric genes consisting of 5' MLL sequences fused in-frame to 3' partner gene sequences. MLL fusion genes are present in 5–10% of childhood and adult leukaemias as well as in the majority of infant leukaemias and secondary leukaemias arising in patients treated with topoisomerase II inhibitors.5, 6, 7, 8 On the whole, leukemic patients with MLL fusion genes have a poor prognosis on conventional chemotherapeutic regimens with the worst outcome seen in the infant group.9, 10

The pathogenic properties of MLL fusion genes have been confirmed both from in vitro cellular transformation studies as well as in vivo mouse models of leukaemias developed from either the homologous recombination approach in embryonic stem (ES) cells or via retrovirus transduction/transplantation (RTT) of enriched haemopoietic progenitor cells (reviewed by Ayton and Cleary,11 Daser and Rabbitts12). To date, nine mouse models of MLL fusion genes (i.e. ENL, CBP, ELL, AF10, AFX, FKHRL1, AF1p, GAS7 and SEPT6)13, 14, 15, 16, 17, 18, 19, 20, 21, 22 have been generated through the RTT approach compared to four (i.e. AF9, ENL, CBP and AF4)23, 24, 25, 26, 27, 28 achieved through the gene targeting approach. Consistent with gene profiling studies from patients with MLL translocations,29, 30, 31 mouse models have confirmed deregulation of Hox genes mediated either through transactivation of the MLL fusion protein or dimerization of MLL contributed by the fusion partner to be a critical step for leukemic transformation. However, the requirement for specific subsets of Hox genes in leukemic transformation in mouse models mediated by MLL fusion genes appears to vary depending on the type of fusion partner and on the approach used for generating the mutant mice.32, 33, 34, 35 Besides MLL-AF936 and MLL-CBP26 mouse models, there have been no other studies performed to study the early effects of MLL fusion protein on haemopoietic development and its impact on Hox expression in pre-leukemic phase. These studies are pertinent given the multistep pathogenesis of human leukaemias and the in utero origin of some MLL fusion proteins.

The MLL-EEN gene was first identified in an infant with acute myeloid leukaemia.37 We have shown that EEN, a member of the endophilin family involved in clathrin-mediated endocytosis,38, 39, 40, 41 is localized in different subcellular compartments,42 and as an adaptor protein can act as tumour suppressor in rat fibroblast Ras transformation assays.43 In vitro studies on the oncogenic role of MLL-EEN have shown that when HL-60 cells are transfected with the fusion gene, the cells show increased proliferation and are more resistant to cell death induced by serum deprivation.44 In order to study the biological role of MLL-EEN fusion gene in relation to the multistep pathogenesis of leukaemia in a physiological relevant setting, we created MllEen/+ mouse ES cells and observed its effect on in vitro haemopoiesis and the pathogenesis of leukaemia in MllEen/+ chimeric mice. Our results showed that MllEen/+ ES cells led to enhanced proliferation of myeloid progenitors derived from embryoid bodies. The chimeric mice we generated developed leukaemia characterized by organomegaly and infiltration with immature myeloid cells after a latency of at least 3.5 months. Our results also revealed different patterns of Hox gene expression between myeloid cells derived from MllEen/+ ES cells and that observed in leukemic mouse bone marrow (BM) cells – this suggests that MLL fusion gene differentially regulates HOX gene expression at different stages of leukemic development. Taken together, we demonstrated a pathogenic role of MLL-EEN in the development of leukaemia and showed that studies of ES cells combined with in vivo animal models can open new approaches to the dissection of molecular events in the multistep pathogenesis of leukaemia.

Top

Materials and methods

Construction of Mll-Een targeting vector

The Mll genomic fragment containing exons 8 and 9 of Mll (Probe A, Figure 1a) was used as a probe to screen a 129Sv/Ev mouse lambda phage genomic library (kind gift of J Rossant). Positive Mll genomic clones containing the translocation breakpoint region from the BamHI site of exons 3–10 were purified and subcloned into the pBluescript KSII+ vector (Stratagene, La Jolla, CA, USA). The Een complementary DNA (cDNA) plasmid (SH3p8) (kind gift of BK Kay)45 was amplified from exons 2 to 10 of Een and subcloned into the IRES-EGFP-loxP-Neo vector. The IRES-EGFP-loxP-Neo vector was derived from the IRES-EGFP plasmid (BD Biosciences, Pharmingen, San Diego, CA, USA) and the pGK loxP Neo plasmid (kind gift of JD Huang). The resulting Een-IRES-EGFP-loxP-Neo DNA fragment was inserted after exon 8 of Mll by bacterial recombination46 to generate the Mll-Een knockin targeting vector (Figure 1b). The inclusion of IRES-EGFP facilitates detection of cells that expressed the Mll-Een fusion protein. Integrity of the DNA sequences of the various vectors was confirmed after each cloning step.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Mll-Een knockin targeting construct. (a) Construction of the Mll-Een targeting plasmid. The 'Een-IRES-EGFP-loxP-Neo' PCR fragment is inserted after exon 8 of Mll by bacterial recombination method. The size of the restriction fragments corresponding to the wild-type Mll and targeted Mll-Een are 5.5 and 3.5 kb EcoRI fragments, respectively, when detected using the 3' probe. The 28 and 13 kb AgeI fragments correspond to the wild-type Mll and targeted Mll-Een, respectively, when detected using the 5' probe. IRES indicates internal ribosomal entry site; EGFP, enhanced GFP; Neo, kanamycin-resistance gene. (b) Comparison of the breakpoint sequences found in human MLL-EEN patient and the constructed mouse Mll-Een-targeted allele. The sequence of the Mll-Een breakpoint region of the targeted allele was confirmed by sequencing genomic DNA derived from targeted ES cells. (c) Southern blot hybridization for genotyping ES clones with wild-type and Mll-Een-targeted alleles. DNA samples were digested with AgeI and EcoRI and hybridized with 5' and 3' probe, respectively.

Full figure and legend (234K)

Generation of MllEen/+ ES cells and chimeric mice

The Mll-Een knockin targeting vector was introduced into L4 ES cells established from 129 Sv/Ev blastocysts in our Transgenic Core Facility (R Lung and KSE Cheah, unpublished) and ES cell clones carrying the desired targeted MllEen/+ allele were identified by Southern blot analysis. Chimeric mice were generated by injecting MllEen/+ ES cells into C57BL/6 blastocysts. The MlllacZ/+-targeted ES cells are E14 (129/Ola) cells with fusion of lacZ into exon 8 of the Mll locus (kind gift of TH Rabbitts).47 Both L4 cells and E14 ES cells were from the 129 strain of mouse.

Haemopoietic differentiation of MllEen/+ ES cells

The schedule for in vitro differentiation of ES cells to the haemopoietic lineage is summarized in Figure 2a. ES cells were maintained as stem cells on mitomycin C inactivated feeder layer. At 48 h before differentiation, the ES cells were adapted to Iscove's modified Dulbecco's medium (IMDM), and on day 0, ES cells (5 times 103) were added to 1 ml of IMDM medium containing stem cell factor (SCF) and 1% methylcellulose as described in the supplier's instructions (Stem Cell Technologies, Vancouver, BC, Canada). Embryoid bodies (EBs) were formed in culture by plating on bacteriological dishes. At day 5, 7 and 9, the total number of EBs and cells dissociated from EBs were counted. Colony-forming assays were performed by culturing 5 times 104 cells in 1 ml of IMDM containing 1% methylcellulose, 15% fetal bovine serum (FBS), 2 mM L-glutamine, 150 muM monothioglycerol (MTG), 1% bovine serum albumin, 10 mg/ml insulin, 200 mug/ml transferin, 100 ng/ml mSCF, 30 ng/ml mouse interleukin-3 (mIL-3), 30 ng/ml IL-6 and 40 ng/ml granulocyte–monocyte-colony-stimulating factor (GM-CSF). At day 7, the numbers of colony-forming units-granulocyte, erythroid, macrophage, megakaryocyte (CFU-GEMM), colony-forming units-granulocyte and macrophages (CFU-GM) and colony-forming units-erythroid (CFU-E) colonies were scored for cell clusters of more than 200 cells. For fluorescence-activated cell sorter (FACS) analysis, the culture was scaled up 10 times by plating 5 times 105 ES cells in 10 ml of culture medium.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

In vitro haemopoietic differentiation of ES cells and self-renewal assay. (a) Overview of the culture protocol for myeloid differentiation of ES cells and self-renewal assay. ES cells were cultured in suspension with SCF, which enhances development of haemopoietic progenitors. EBs with various lineages of progenitor cell developed were harvested at days 5, 7 and 9, and cultured in cytokines IL-3, IL-6, SCF and GM-CSF. Pooled population of haemopoietic colonies were dissociated and replated at a 7-day interval for self-renewal assay. (b) Representative morphology of EBs derived from Mll+/+, Mlllac/+ and MllEen/+ ES cells. (c) Typical CFU-GEMM, CFU-GM and CFU-E colonies derived from day 9 harvested Mll+/+, Mlllac/+ and MllEen/+ EBs. (d) Representative morphology of Mll+/+, Mlllac/+ and MllEen/+ colonies from the self-renewal assay. (e) Total number of EBs per 5 times 103 ES cells plated at day 5, 7 and 9. (f) Total number of CFU-GM, CFU-GEMM and CFU-E colonies derived from 5 times 104 EBs cells plated. (g) Total number of CFU-GEMM and CFU-GM colonies derived from 1 times 105 cells plated in each round of the self-renewal assay. ES, embryonic stem cell, EBs, embryoid bodies; SCF, stem cell factor; IL, interleukin; GM-CSF, granulocyte–monocyte-colony-stimulating factor; CFU, colony-forming unit; CFU-GEMM, CFU-granulocyte, erythroid, macrophage, megakaryocyte; CFU-GM, CFU-granulocyte and macrophages; CFU-E, CFU-erythroid. Error bar of the bar chart indicated the standard derivations between three independent experiments each were duplicated. Error bar of the bar chart for E–H indicated the standard deviations between three independent experiments, each were duplicated. Photos were taken by Nikon Eclipse TS100 microscope with times 100 magnification, unless specified.

Full figure and legend (243K)

Replating assays for self-renewal property

Cells (1 times 105) from haemopoietic colonies derived from day 7 and 9 EBs were cultured in semisolid methylcellulose medium containing haemopoietic cytokines as described in the previous section. At day 7, the numbers of CFU-GEMM, CFU-GM and CFU-E colonies were counted, and the colonies harvested and plated at a concentration of 1 times 105 cells/ml. Replating was performed at a 7-day-interval to determine self-renewal. A similar replating assay was performed under similar conditions on BM cells from wild-type and MllEen/+ chimeric mice obtained by flushing the mouse femur with IMDM medium with 10% FBS.

Mouse pathology and immunophenotypic analysis

MllEen/+ chimeric mice were killed when they showed signs of distress. Following post-mortem examination, tissues were fixed in 10% formalin, embedded in paraffin wax, sectioned and stained with haematoxylin and eosin (H&E) by routine procedures. Cytospins of BM and peripheral blood (PB) films were fixed in methanol:acetone (80:20) and stained with Wright–Giemsa stain (Sigma Aldrich, St Louis, MO, USA) to assess cell morphology under light microscopy.

Immunophenotypic analysis was performed by labelling the cells harvested from colony assays or from BMs and organs of MllEen/+ chimeric mice with a combination of fluorochrome-conjugated antibodies against c-kit (biotin), Mac-1 (CD11b, PE), Gr-1 (biotin), B220 (Cy5), IgM (PE), CD4 (Cy5) and CD8 (PE) (BD Biosciences, USA). A secondary streptavidin Per-CP substrate was applied to the biotin-labelled antibody. The percentage of Mll-Een-expressing cells was reflected by the percentage of green fluorescence protein (GFP)-expressing cells. The immunophenotypic analysis was performed with FACScan instrument ('Epics-Altra', Beckman Coulter) with analysis of 10 000 cells in each sample.

Semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR) analysis

Semiquantitative RT-PCR was performed to study the mRNA expression level of Hoxa7, Hoxa9, Hoxa10, Hoxb4 and Meis1. Total RNA was isolated from cells using TRIzol (Life Technologies, Carlsbad, CA, USA) as recommended by the manufacturer. Two micrograms of total RNA were denatured and reverse transcribed using oligo (dT)15 primers in a 20 mul reaction. Thereafter, 1 mul of cDNA was amplified in a 20 mul PCR reaction mixture in the presence of specific primers (sequences are available upon request) and within the linear range of amplification cycles. Total RNA extracted from day 14.5 embryo was used as positive control and PCR reaction without cDNA template was used as negative control.

Top

Results

Generation of MllEen/+ ES cells

In the patient with MLL-EEN translocation, exon 6 of MLL is fused in-frame with exon 2 of EEN. To recapitulate the fusion breakpoint found in the human patient, the cDNA of Een was fused in-frame with exon 8 of Mll, which is equivalent to exon 6 of MLL (Figure 1a). The alignment of the breakpoint region of MLL-EEN fusion and the Mll-Een knockin mutant is illustrated in Figure 1b.

The targeting vector was introduced into ES cells and five out of six hundred clones screened were identified as targeted clones by Southern blot analysis using 5' and 3' external probes (Figure 1c). The targeted ES cells contain the designed gain-of-function Mll-Een fusion gene that mimics the corresponding gene rearrangement found in human leukaemia and the IRES-EGFP that facilitates detection of cells, which express the Mll-Een fusion protein. As the targeted Mll locus is disrupted by the Een cDNA, the Mll-Een-targeted allele (designated MllEen) also carries a concurrent heterozygous loss-of-function mutation of Mll.

MllEen/+ ES cells leads to enhanced proliferation of myeloid progenitors

To investigate the effect of Mll-Een fusion protein on early haemopoietic development, targeted MllEen/+ ES cells were subjected to in vitro haemopoietic differentiation and their proliferation and differentiation capacities were examined according to the time schedule shown in Figure 2a. We used two independent MllEen/+-targeted ES cell clones (clone 1F and clone 4H), and compared the results with wild-type ES cells (Mll+/+) and ES cells with an artificial Mll-lacZ fusion allele (MlllacZ/+)47 representing heterozygous loss-of-function mutation of Mll. In the absence of LIF, all three ES cell lines gave rise to EBs as spherical masses of cells surrounded by cellular envelopes with no significant difference in their morphology and growth during the first three days of culture (Figure 2b). From day 5 onwards, however, there was an approximately two-fold increase in the number of EBs generated from MllEen/+ ES cells when compared with Mll+/+ and MlllacZ/+ ES cells (Figure 2e). As all MllEen/+ ES cells and EBs expressed GFP throughout the culture (data not shown), this suggests that one of the early effects of the Mll-Een fusion protein on ES cell development is to facilitate the commitment of a higher number of ES cells to form EBs.

We next investigated the haemopoietic colony forming potential of EBs generated from MllEen/+, Mll+/+ and MlllacZ/+ ES cells. Cell suspensions from day 5, day 7 to day 9 EBs were assayed for their potential to differentiate into CFU-GEMM, CFU-GM and CFU-E. In contrast to EBs from Mll+/+ and MlllacZ/+ ES, which gave rise to very small numbers of CFU-GEMM, CFU-GM and CFU-E colonies, day 7 and day 9 EBs, from MllEen/+ ES cells produced greater numbers of and larger size colonies (Figure 2c and f). Immunophenotypic analysis of pooled colonies (day 16) generated from day 9 from MllEen/+ EBs showed that more than 70% of the cells were Mac-1+ immature myeloid cells (Figure 3a). These results suggest that cells from MllEen/+ EBs can differentiate into myeloid lineage with enhanced proliferative capacity.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Immunophenotyping of haemopoietic cells in the self-renewal assay. (a) Pooled population of haemopoietic colonies (day 16) were dissociated and dual-labelled with anti-c-kit/Mac-1. (b) CFUs from the replating assay were harvested, dissociated and dual-labelled with anti-c-kit/Mac-1 antibodies. Analysis was not carried out on Mll+/+ and MlllacZ/+ cells on the third and fourth repalting because there were no colonies in the culture. (c) FACS analysis (Beckman Coulter, Altra Epics sorter) of EGFP, which marks the expression of Mll-Een was analysed in cells from haemopoietic colonies (day 16) and self-renewal colonies.

Full figure and legend (78K)

Myeloid progenitors generated from MllEen/+ EBs have self-renewal property and specific Hox gene profile

We next assessed the proliferative capacity of MllEen/+, Mll+/+ and MlllacZ/+ myeloid colonies for self-renewal by performing replating assays at 7-day intervals and, at simultaneous time points, RT-PCR analysis of Hoxa7, a9, a10, b4 and Meis1. From the first replating (day 23), MllEen/+ myeloid colonies continued to increase in size and number, maintained a high proportion of CFU-GEMM up to the fourth replating (Figure 2d and g) and expressed GFP (Figure 3c). Immunophenotypic analysis showed that cells from MllEen/+ myeloid colonies before replating (day 16) had a more immature myeloid phenotype with two- threefold increase of c-kit+/Mac-1+ (Figure 3a) compared with cells from Mll+/+ and MlllacZ/+ colonies. These differences in phenotype were maintained at the fourth replating of MllEen/+ (day 44) cells (Figures 3b). Taken together, these results indicate that MllEen/+ myeloid progenitors not only proliferate faster but have enhanced self-renewal property compared to Mll+/+ and MlllacZ/+ derived progenitors.

MllEen/+ chimeric mice develop myeloid leukaemia

We investigated the ability of Mll-Een fusion protein in contribution to leukaemogenesis from 15 chimeric mice generated from two independent MllEen/+ ES cell clones (clone 1F and clone 4H). We observed no germline transmission of the MllEen/+ mutant allele upon examination of 250 offspring obtained from breeding the chimeric mice with wild-type C57BL/6. This suggests that either the knockin MllEen/+ mutant allele is embryonic lethal or that the targeted ES cell clones did not contribute to the germline in any of the chimeric mice.

Of these 15 chimeric mice, nine mice (3.5–11 months) were found dead with two in poor conditions; and six were killed when they showed signs of distress. Of 13 mice suitable for tissue examination, there were signs of leukaemia in 11 mice (seven male and four female, the youngest being 3.5 months) as evident by pathologically enlarged thymuses, livers, spleens and lymph nodes, presence of blasts in peripheral blood and infiltration with immature myeloid cells in BM and other organs (Figure 4a–c, Table 1). Immunophenotypic analysis of cells from BMs and pathologically enlarged organs by flow cytometry revealed high levels of EGFP expression, which marked Mll-Een-expressing cells and confirmed the leukemic origin of these cells contributing to organomegaly (Table 1). We also observed infiltration of leukemic blasts in varying degrees in other non-haemopoietic organs, that is, interstitium of the kidney, epicardium and interstitium of the myocardium, interstitium between seminiferous tubules and muscle coat of vas deferens were also observed (data not shown). Of the three chimeric mice with no pathological evidence of leukaemia, two appeared normal (e.g. cases 4 and 9; Table 1), with case 4 having low levels of Mll-Een-expressing cells (2% EGFP-positive cells) in the BM; case 9 was not amenable to analysis.

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Development of leukaemia in MllEen/+ chimeras. (a) Gross tissue morphology of the liver, spleen and mesenteric lymph node (LN) of the wild-type (WT) and MllEen/+ chimera (bar=1 cm). (b) Bone marrow and peripheral blood film of typical MllEen/+ chimera showing presence of blast cells. (c) Representative histological sections from femur, liver, spleen, thymus and mesenteric lymph node of MllEen/+ chimera showing infiltration of immature myeloid (IM) cells. (d) FACS analysis on single-cell suspension of BM, spleen, thymus and mesenteric lymph node dual-labelled with anti-Mac-1/Gr-1 antibodies. Data on the third row showed results of analysis gating of EGFP-positive cell population. (e) FACS analysis on cell suspension of spleen and mesenteric lymph node dual-labelled with anti-B220/IgM, and thymus and mesenteric lymph node dual-labelled with anti-CD4/CD8, respectively. N, neutrophils; W, white pulp; R, red pulp; F, lymphoid follicle; C, cortex; Med, medullary and In, infiltration of leukemic cells. Original magnification as indicated.

Full figure and legend (579K)


We next performed cell marker studies of myeloid, B- and T-cell lineages on cells from pathologically enlarged organs that were infiltrated by leukemic cells expressing Mll-Een in two mice (cases 6 and 5 with 84% and 50% Mll-Een-expressing cells, respectively). As shown in Figure 4d, the predominant phenotype seen in the total cells as well as the gated population, that is, cells expressing Mll-Een in case 6 was an increase in Mac-1+/Gr-1- (immature myeloid) populations in spleen, thymus, mesenteric lymph node and BM compared to wild type. We also observed increase in blasts in peripheral blood and BM (Figure 4b) as well as absence of normal B220+/IgM+ mature B-cell population in the spleen and mesenteric lymph node, absence of normal CD4+/CD8+ double-positive T cells in the thymus, absence of CD4 and decrease of CD8 single positive T cells in mesenteric lymph node (Figure 4e, second row). In the other less severely affected mouse (i.e. case 5), there were also decrease in normal B- and T-cell populations in spleen and thymus but to a lesser extent (Figure 4e, third row).

Taken together, the pathological and immunophenotypic findings showed evidence of leukaemia in the chimeric mice and demonstrated a pathogenic role of Mll-Een in the development of myeloid leukaemia.

Mll-Een increases the self-renewal of BM myeloid progenitors

We next investigated whether mouse BM cells expressing Mll-Een have self-renewal property by performing replating assays in the presence of haemopoietic cytokines as described previously. Mll-Een enhanced proliferation and self-renewal of myeloid progenitors up to the fourth replating compared to the wild-type BM cells, which exhausted their proliferation capacity on the second replating (Figure 5a).

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

In vitro clonogenic assay and expression analysis of BM cells from MllEen/+ chimeras. (a) Total number of CFU derived from 1x105 BM cells of wild-type and MllEen/+ chimeras. (b) CFU derived from BM cells of wild-type and MllEen/+ chimeras showing predominant cell population with c-kit+, Mac-1+ and Gr-1. (c) The left and right panel shows the typical very compact and less-dense colonies derived from BM of MllEen/+ chimeras, respectively. Both colonies expressed EGFP under fluorescence microscopy (Nikon ECLIPSE TS100, original magnification, times 40). (d) RT-PCR analysis of Hox and Meis expression in MllEen/+ pooled cell population isolated from day 7 hemopoietic colonies and self-renewal assay. (e) RT-PCR analysis of Hox and Meis1 expression in BM cells of MllEen/+ chimeras with high (80%) and low (2%) percentage of EGFP expression, respectively. RNA integrity was examined and the amount was normalized to GAPDH. Reaction with RNA from E14.5 whole-mouse embryo or without cDNA template was used as positive and negative control, respectively.

Full figure and legend (196K)

Immunophenotyping analysis of day 7 colonies expressing Mll-Een showed higher levels of c-kit positivity with fewer Mac-1+/Gr-1+ cells compared to the wild–type, which suggests that these cells were at a more immature state (Figure 5b). EGFP-positive Mll-Een colonies appeared as very compact colonies and contained cells resembling immature myeloid cells by Wright–Giemsa staining (Figure 5c). These results show that Mll-Een myeloid cells can undergo self-renewal with impairment of terminal differentiation resembling features typically seen in acute leukaemia.

Hox expression profiles are different between Mll-Een myeloid population derived from ES cells and in the leukemic mice

As Hox genes are believed to be major mediators for MLL-mediated leukemogenesis, we assessed the effect of Mll fusion on Hox gene expression in the early developmental stage (e.g. myeloid progenitors derived from Mll-Een ES cells) and in full-blown leukaemia (e.g. MllEen/+ leukemic cells) by semiquantitative RT-PCR. Interestingly, we observed that although myeloid progenitors derived from MllEen/+ ES cells expressed Hoxa7, a9, b4 and Meis1 before replating, Hoxa7 and a9 were downregulated and only the expression of Meis1 and Hoxb4 were maintained in the subsequent replating (Figure 5d). Conversely when we examined the expression of Hox genes in the myeloid leukemic cells from chimeric animals, Hoxb4 which was up-regulated in MllEen/+ ES cell-derived myeloid progenitors, was not expressed in the leukemic cells. Instead, we observed upregulation of Hoxa7, a9, a10 and Meis1 in cells harvested from the leukemic mouse (with 80% MllEen/+-expressing cells) but not in the non-leukaemia chimeric mouse (with 2% MllEen/+-expressing cells) (Figure 5e). These results indicate that Mll-Een may differentially regulate the expression of Hox genes during leukemic development.

Top

Discussion

The multistep model of leukaemia development in which genetic mutations accumulate over time leading to the development of clinical disease is supported by the findings in clinical samples as well as from the long latency required to develop overt malignancy in mouse models of human leukaemias. Although there are several mouse models of MLL fusion leukaemias to date, the early effects of the MLL fusion protein on haemopoieitic development have only been examined in the MLL-AF9 and MLL-CBP mice.36, 26 In both studies, the fusion protein was shown to induce expansion of myeloid progenitors and enhanced self-renewal with the requirement of secondary events for the development of fully established leukaemia. In order to define the role of MLL fusion protein in the pathogenesis of leukaemia in a multistep model, we have used mouse ES cells as the biological platform not only to examine the effects of Mll-Een on the in vitro haemopoietic development of ES cells but also on the pathogenesis of leukaemia development. This was achieved by creating a mouse equivalent of the human MLL-EEN allele by targeted insertion of Een cDNA into the Mll locus of ES cells via homologous recombination and subsequently generated MllEen/+ chimeric mice.

We showed that primary MllEen/+ ES cells were enhanced to form Ebs, which can generate myeloid progenitors with enhance proliferation in a replating assay in the presence of haemopoietic cytokines. This is in contrast to Mll+/+ and MlllacZ/+ myeloid progenitors, which proliferate but mature and loose their proliferative capacity after the second replating. The finding that MllEen/+ ES myeloid progenitors could be replated up to four times is indicative of self-renewal property, which is necessary but not sufficient on its own for the development of acute leukaemia without further cellular perturbations such as maturation arrest and apoptotic resistance. These self-renewing myeloid progenitors as expected were also shown to have a more immature myeloid phenotype compared to Mll+/+ and MlllacZ/+ myeloid progenitors.

When the MllEen/+ ES cells were put back to blastocyst, we have been successful in generating 15 MllEen/+ chimeric mice. No germline transmitted heterozygous mice were created, suggesting that no targeted ES cells could contribute to the germline, or early or high level of expression of Mll-Een is incompatible with life. The 15 MllEen/+ chimeric mice have relatively low levels of chimerism based on coat colour. Among 13 chimeric mice analysed, there was evidence of leukaemia in 11 mice based on histopathological examination and immunophenotypic studies, which showed infiltration with immature myeloid cells in BM and in enlarged organs such as the liver, spleen, thymus and lymph node. Interestingly, although signs of leukaemia were observed in both male and female chimeras, most of the chimeras that died suddenly were female (six out of seven). It is coincidental that the only human leukaemia case of MLL-EEN fusion we reported37 was an affected female. Apart from its functional role in abnormal haemopoiesis, it is not clear whether Mll-Een has other sexually biased functions, which might aggravate the disease conditions.

We also demonstrated that BM progenitor cells expressing Mll-Een could undergo self-renewal in a replating assay. These results provide strong evidence for a pathogenic role of MLL-EEN for the development of myeloid leukaemia, although the long mean latency of 7 months suggests additional secondary events may be necessary for the induction of leukaemia. While part of the oncogenicity of MLL-EEN can be attributed by homodimerization via the coiled-coil domain of EEN, the substantially shorter disease latency of Mll-Een mice compared with those of Mll-lacZ,47 suggested that Mll-Een is a more potent oncoprotein than just an MLL-dimer. This is also in keeping with our findings that only MllEen/+ but not MlllacZ/+ ES cells were able to maintain in vitro proliferation of myeloid progenitors.

Previous studies showed that transduction of murine progenitor cells with various MLL fusion genes led to the expression of Hoxa1, a3, a5, a7, a9, a10, a11 and Meis1 genes.20, 32 However there is evidence from other studies which shows that a single Hox gene such as Hoxa9 under particular experimental settings is absolutely required for MLL-ENL-mediated transformation32 but not so in MLL-GAS7 or MLL-AF9 mediated transformation.33, 34, 35 Thus the pathogenic significance of these genes either as whole profiles, subsets or combinations in the development of leukaemia remains to be further clarified. We have previously shown that MLL-EEN can transactivate HoxA7 promoter in HL-60 cells transfected with the fusion gene44 and our results on Hox gene expression in the MllEen/+ chimeric mice is in agreement with previous RTT mouse models, which show upregulation of Hoxa7, a9, a10 and Meis1.32, 34, 35 However, neither of these critical Hoxa genes were expressed in the active proliferating Mac-1+ cells derived from MllEen/+ ES cells. Instead, we detected overexpression of Hoxb4, another Hox gene that functions to promote self-renewal. This result is reminiscent of recent findings that in hES cell-derived HSCs,48 the expression patterns of HoxA and HoxB genes are differentially regulated when compared with in vivo HSCs. Moreover, it has been demonstrated that both Hoxa9 and Hoxb4 have the ability to expand HSCs, but overexpression of Hoxb4 will result in enhanced self-renewal without induction of leukaemia.49, 50 Therefore, Hoxb4 might have distinctive roles in the control of self-renewal of the MllEen/+ ES cells. Although we cannot exclude the possibility that the observed Hox expression profiles may simply reflect the developmental status of the cells, these findings also suggest that different waves of Hox expression may be activated by MLL fusion proteins for initiation (in ES cells) and maintenance (in leukaemic cells) of the disease. The consecutive replating ability of Mll-Een-transformed myeloid progenitors derived directly from ES cells in the absence of both Hoxa7 and Hoxa9 strongly indicate the functional redundancy of Hox genes in MLL-mediated leukaemia. Consistent with the previous observation on MLL-GAS7 and MLL-AF9, our results provide further evidence that MLL fusion proteins are unlikely to be absolutely dependent on a single Hox gene for transformation. It will thus be crucial to determine the functional interplay among different Hox genes in MLL-mediated leukaemia.

In conclusion, we have shown that the Mll-Een fusion gene leads to enhanced proliferation of myeloid progenitors derived from mouse ES cells and the development of myeloid leukaemia in chimeric mice. The advantage of the ES cell system is that it can be used to compare the biologic effects of diverse MLL fusion proteins on in vitro haemopoietic development. To provide further insight into the molecular pathogenesis of leukaemia mediated by MLL-EEN, it will be important to generate conditional mouse mutants so that different haemopoietic developmental stages or lineages can be studied. It will also be pertinent to examine other biological mechanisms through which EEN contributes to leukemogenesis as a fusion partner to MLL in future mouse models of the disease. EEN is unique among the fusion partners of MLL, which can dimerize but potentially can function as a tumour suppressor43 and also in endocytosis which is of particular interest in view of mounting evidence linking deregulation of endocytosis leading to aberrant intracellular signalling and cancer.51, 52

Top

References

  1. Ziemin-van der Poel S, McCabe NR, Gill HJ, Espinosa III R, Patel Y, Harden A et al. Identification of a gene, MLL, that spans the breakpoint in 11q23 translocations associated with human leukemias. Proc Natl Acad Sci USA 1991; 88: 10735–10739. | PubMed | ChemPort |
  2. Djabali M, Selleri L, Parry P, Bower M, Young BD, Evans GA. A trithorax-like gene is interrupted by chromosome 11q23 translocations in acute leukaemias. Nat Genet 1992; 2: 113–118. | Article | PubMed | ISI | ChemPort |
  3. Yu BD, Hess JL, Horning SE, Brown GA, Korsmeyer SJ. Altered Hox expression and segmental identity in Mll-mutant mice. Nature 1995; 378: 505–508. | Article | PubMed | ISI | ChemPort |
  4. Hess JL, Yu BD, Li B, Hanson R, Korsmeyer SJ. Defects in yolk sac hematopoiesis in Mll-null embryos. Blood 1997; 90: 1799–1806. | PubMed | ISI | ChemPort |
  5. Look AT. Oncogenic transcription factors in the human acute leukemias. Science 1997; 278: 1059–1064. | Article | PubMed | ISI | ChemPort |
  6. Super HJ, McCabe NR, Thirman MJ, Larson RA, Le Beau MM, Pedersen-Bjergaard J et al. Rearrangements of the MLL gene in therapy-related acute myeloid leukemia in patients previously treated with agents targeting DNA-topoisomerase II. Blood 1993; 82: 3705–3711. | PubMed | ISI | ChemPort |
  7. Domer PH, Head DR, Renganathan N, Raimondi SC, Yang E, Atlas M. Molecular analysis of 13 cases of MLL/11q23 secondary acute leukemia and identification of topoisomerase II consensus-binding sequences near the chromosomal breakpoint of a secondary leukemia with the t(4;11). Leukemia 1995; 9: 1305–1312. | PubMed | ISI | ChemPort |
  8. Broeker PL, Super HG, Thirman MJ, Pomykala H, Yonebayashi Y, Tanabe S et al. Distribution of 11q23 breakpoints within the MLL breakpoint cluster region in de novo acute leukemia and in treatment-related acute myeloid leukemia: correlation with scaffold attachment regions and topoisomerase II consensus binding sites. Blood 1996; 87: 1912–1922. | PubMed | ISI | ChemPort |
  9. Chen CS, Sorensen PH, Domer PH, Reaman GH, Korsmeyer SJ, Heerema NA et al. Molecular rearrangements on chromosome 11q23 predominate in infant acute lymphoblastic leukemia and are associated with specific biologic variables and poor outcome. Blood 1993; 81: 2386–2393. | PubMed | ISI | ChemPort |
  10. Pui CH, Behm FG, Downing JR, Hancock ML, Shurtleff SA, Ribeiro RC et al. 11q23/MLL rearrangement confers a poor prognosis in infants with acute lymphoblastic leukemia. J Clin Oncol 1994; 12: 909–915. | PubMed | ChemPort |
  11. Ayton PM, Cleary ML. Molecular mechanisms of leukemogenesis mediated by MLL fusion proteins. Oncogene 2001; 20: 5695–5707. | Article | PubMed | ISI | ChemPort |
  12. Daser A, Rabbitts TH. The versatile mixed lineage leukaemia gene MLL and its many associations in leukaemogenesis. Semin Cancer Biol 2005; 15: 175–188. | Article | PubMed | ISI | ChemPort |
  13. Lavau C, Szilvassy SJ, Slany R, Cleary ML. Immortalization and leukemic transformation of a myelomonocytic precursor by retrovirally transduced HRX-ENL. EMBO J 1997; 16: 4226–4237. | Article | PubMed | ISI | ChemPort |
  14. Lavau C, Du C, Thirman M, Zeleznik-Le N. Chromatin-related properties of CBP fused to MLL generate a myelodysplastic-like syndrome that evolves into myeloid leukemia. EMBO J 2000; 19: 4655–4664. | Article | PubMed | ISI | ChemPort |
  15. Lavau C, Luo RT, Du C, Thirman MJ. Retrovirus-mediated gene transfer of MLL-ELL transforms primary myeloid progenitors and causes acute myeloid leukemias in mice. Proc Natl Acad Sci USA 2000; 97: 10984–10989. | Article | PubMed | ChemPort |
  16. DiMartino JF, Miller T, Ayton PM, Landewe T, Hess JL, Cleary ML et al. A carboxy-terminal domain of ELL is required and sufficient for immortalization of myeloid progenitors by MLL-ELL. Blood 2000; 96: 3887–3893. | PubMed | ISI | ChemPort |
  17. DiMartino JF, Ayton PM, Chen EH, Naftzger CC, Young BD, Cleary ML. The AF10 leucine zipper is required for leukemic transformation of myeloid progenitors by MLL-AF10. Blood 2002; 99: 3780–3785. | Article | PubMed | ISI | ChemPort |
  18. So CW, Cleary ML. MLL-AFX requires the transcriptional effector domains of AFX to transform myeloid progenitors and transdominantly interfere with forkhead protein function. Mol Cell Biol 2002; 22: 6542–6552. | Article | PubMed | ChemPort |
  19. So CW, Cleary ML. Common mechanism for oncogenic activation of MLL by forkhead family proteins. Blood 2003; 101: 633–639. | Article | PubMed | ChemPort |
  20. So CW, Lin M, Ayton PM, Chen EH, Cleary ML. Dimerization contributes to oncogenic activation of MLL chimeras in acute leukemias. Cancer Cell 2003; 4: 99–110. | Article | PubMed | ISI | ChemPort |
  21. So CW, Karsunky H, Passegue E, Cozzio A, Weissman IL, Cleary ML. MLL-GAS7 transforms multipotent hematopoietic progenitors and induces mixed lineage leukemias in mice. Cancer Cell 2003; 3: 161–171. | Article | PubMed | ISI | ChemPort |
  22. Ono R, Nakajima H, Ozaki K, Kumagai H, Kawashima T, Taki T et al. Dimerization of MLL fusion proteins and FLT3 activation syngergize to induce multiple-lineage leukemogenesis. J Clin Invest 2005; 115: 919–929. | Article | PubMed | ISI | ChemPort |
  23. Corral J, Lavenir I, Impey H, Warren AJ, Forster A, Larson TA et al. An Mll-AF9 fusion gene made by homologous recombination causes acute leukemia in chimeric mice: a method to create fusion oncogenes. Cell 1996; 85: 853–861. | Article | PubMed | ISI | ChemPort |
  24. Collins EC, Pannell R, Simpson EM, Forster A, Rabbitts TH. Inter- chromosomal recombination of Mll and Af9 genes mediated by cre-loxP in mouse development. EMBO Rep 2000; 1: 127–132. | Article | PubMed | ISI | ChemPort |
  25. Forster A, Pannell R, Drynan LF, McCormack M, Collins EC, Daser A et al. Engineering de novo reciprocal chromosomal translocations associated with Mll to replicate primary events of human cancer. Cancer Cell 2003; 3: 449–458. | Article | PubMed | ISI | ChemPort |
  26. Wang J, Iwasaki H, Krivtsov A, Febbo PG, Thorner AR, Ernst P et al. Conditional MLL-CBP targets GMP and models therapy-related myeloproliferative disease. EMBO J 2005; 24: 368–381. | Article | PubMed | ISI | ChemPort |
  27. Chen W, Li Q, Hudson WA, Kumur A, Kirchhof N, Kersey JH. A murine Mll-AF4 knockin model results in lymphoid and myeloid deregulation and hematological malignancy. Blood 2006; 108: 669–677. | Article | PubMed | ChemPort |
  28. Metzler M, Forster A, Pannell R, Arends MJ, Daser A, Lobato MN et al. A conditional model of MLL-AF4 B-cell tumourigenesis using invertor technology. Oncogene 2006; 25: 3093–3103. | Article | PubMed | ChemPort |
  29. Armstrong SA, Staunton JE, Silverman LB, Pieters R, den Boer ML, Minden MD et al. MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet 2002; 30: 41–47. | Article | PubMed | ISI | ChemPort |
  30. Ross ME, Mahfouz R, Onciu M, Liu HC, Zhou X, Song G et al. Gene expression profiling of pediatric acute myelogenous leukemia. Blood 2004; 104: 3679–3687. | Article | PubMed | ISI | ChemPort |
  31. Kohlmann A, Schoch C, Dugas M, Schnittger S, Hiddemann W, Kern W et al. New insights into MLL gene rearranged acute leukemias using gene expression profiling: shared pathways, lineage commitment, and partner genes. Leukemia 2005; 19: 953–964. | Article | PubMed | ChemPort |
  32. Ayton PM, Cleary ML. Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9. Genes Dev 2003; 17: 2298–2307. | Article | PubMed | ISI | ChemPort |
  33. Kumar AR, Hudson WA, Chen W, Nishiuchi R, Yao Q, Kersey JH. Hoxa9 influences the phenotype but not the incidence of Mll-AF9 fusion gene leukemia. Blood 2004; 103: 1823–1828. | Article | PubMed | ChemPort |
  34. Zeisig BB, Milne T, Garcia-Cuellar MP, Schreiner S, Martin ME, Fuchs U et al. Hoxa9 and Meis1 are key targets for MLL-ENL-mediated cellular immortalization. Mol Cell Biol 2004; 24: 617–628. | Article | PubMed | ISI | ChemPort |
  35. So CW, Karsunky H, Wong P, Weissman IL, Cleary ML. Leukemic transformation of hematopoietic progenitors by MLL-GAS7 in the absence of Hoxa7 or Hoxa9. Blood 2004; 103: 3192–3199. | Article | PubMed | ISI | ChemPort |
  36. Johnson JJ, Chen W, Hudson W, Yao Q, Taylor M, Rabbitts TH et al. Prenatal and postnatal myeloid cells demonstrate stepwise progression in the pathogenesis of MLL fusion gene leukemia. Blood 2003; 101: 3229–3235. | Article | PubMed | ChemPort |
  37. So CW, Caldas C, Liu MM, Chen SJ, Huang QH, Gu LJ et al. EEN encodes for a member of a new family of proteins containing an Src homology 3 domain and is the third gene located on chromosome 19p13 that fuses to MLL in human leukemia. Proc Natl Acad Sci USA 1997; 94: 2563–2568. | Article | PubMed | ChemPort |
  38. Schmidt A, Wolde M, Thiele C, Fest W, Kratzin H, Podtelejnikov AV et al. Endophilin I mediates synaptic vesicle formation by transfer of arachidonate to lysophosphatidic acid. Nature 1999; 401: 133–141. | Article | PubMed | ISI | ChemPort |
  39. Simpson F, Hussain NK, Qualmann B, Kelly RB, Kay BK, McPherson PS et al. SH3-domain-containing proteins function at distinct steps in clathrin-coated vesicle formation. Nat Cell Biol 1999; 1: 119–124. | Article | PubMed | ISI | ChemPort |
  40. Ringstad N, Gad H, Low P, Di Paolo G, Brodin L, Shupliakov O et al. Endophilin/SH3p4 is required for the transition from early to late stages in clathrin-mediated synaptic vesicle endocytosis. Neuron 1999; 24: 143–154. | Article | PubMed | ISI | ChemPort |
  41. So CW, So CK, Cheung N, Chew SL, Sham MH, Chan LC. The interaction between EEN and Abi-1, two MLL fusion partners, and synaptojanin and dynamin: implications for leukaemogenesis. Leukemia 2000; 14: 594–601. | Article | PubMed | ChemPort |
  42. Cheung N, So CW, Yam JW, So CK, Poon RY, Jin DY et al. Subcellular localization of EEN/endophilin A2, a fusion partner gene in leukaemia. Biochem J 2004; 383: 27–35. | Article | PubMed | ChemPort |
  43. Yam JW, Jin DY, So CW, Chan LC. Identification and characterization of EBP, a novel EEN binding protein that inhibits Ras signaling and is recruited into the nucleus by the MLL-EEN fusion protein. Blood 2004; 103: 1445–1453. | Article | PubMed | ChemPort |
  44. Liu H, Chen B, Xiong H, Huang QH, Zhang QH, Wnag ZG et al. Functional contribution of EEN to leukemogenic transformation by MLL-EEN fusion protein. Oncogene 2004; 23: 3385–3394. | Article | PubMed | ChemPort |
  45. Sparks AB, Hoffman NG, McConnell SJ, Fowlkes DM, Kay BK. Cloning of ligand targets: systematic isolation of SH3 domain-containing proteins. Nat Biotechnol 1996; 14: 741–744. | Article | PubMed | ISI | ChemPort |
  46. Zhang Y, Buchholz F, Muyrers JP, Stewart AF. A new logic for DNA engineering using recombination in Escherichia coli. Nat Genet 1998; 20: 123–128. | Article | PubMed | ISI | ChemPort |
  47. Dobson CL, Warren AJ, Pannell R, Forster A, Rabbitts TH. Tumorigenesis in mice with a fusion of the leukaemia oncogene Mll and the bacterial lacZ gene. EMBO J 2000; 19: 843–851. | Article | PubMed | ISI | ChemPort |
  48. Wang L, Menendez P, Shojaei F, Li L, Mazurier F, Dick JE et al. Generation of hematopoietic repopulating cells from human embryonic stem cells independent of ectopic HOXB4 expression. J Exp Med 2005; 201: 1603–1614. | Article | PubMed | ChemPort |
  49. Sauvageau G, Thorsteinsdottir U, Eaves CJ, Lawrence HJ, Largman C, Lansdorp PM et al. Overexpression of HOXB4 in hematopoietic cells causes the selective expansion of more primitive populations in vitro and in vivo. Genes Dev 1995; 9: 1753–1765. | PubMed | ISI | ChemPort |
  50. Antonchuk J, Sauvageau G, Humphries RK. HOXB4-induced expansion of adult hematopoietic stem cells ex vivo. Cell 2002; 109: 39–45. | Article | PubMed | ISI | ChemPort |
  51. Polo S, Pece S, Di Fiore PP. Endocytosis and cancer. Curr Opin Cell Biol 2004; 16: 156–161. | Article | PubMed | ChemPort |
  52. Bache KG, Slagsvold T, Stenmark H. Defective downregulation of receptor tyrosine kinases in cancer. EMBO J 2004; 23: 2707–2712. | Article | PubMed | ISI | ChemPort |
Top

Acknowledgements

This study is supported by research grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. HKU7269/02M, HKU1/98C and HKU2/02M). We thank Cary So for technical support, Jian Dong Huang for help with construction of targeting vector, Florence Loong for help with histology and Chi Leong So, Keith Leung, Christine Ng and Carly Lam for help with gene targeting in ES cells and generation of mutant mice.