Mutant N-RAS induces erythroid lineage dysplasia in human CD34+ cells.

RAS mutations arise at high frequency (20-40%) in both acute myeloid leukemia and myelodysplastic syndrome (which is considered to be a manifestation of preleukemic disease). In each case, mutations arise predominantly at the N-RAS locus. These observations suggest a fundamental role for this oncogene in leukemogenesis. However, despite its obvious significance, little is known of how this key oncogene may subvert the process of hematopoiesis in human cells. Using CD34+ progenitor cells, we have modeled the preleukemic state by infecting these cells with amphotropic retrovirus expressing mutant N-RAS together with the selectable marker gene lacZ. Expression of the lacZ gene product, beta-galactosidase, allows direct identification and study of N-RAS-expressing cells by incubating infected cultures with a fluorogenic substrate for beta-galactosidase, which gives rise to a fluorescent signal within the infected cells. By using multiparameter flow cytometry, we have studied the ability of CD34+ cells expressing mutant N-RAS to undergo erythroid differentiation induced by erythropoietin. By this means, we have found that erythroid progenitor cells expressing mutant N-RAS exhibit a proliferative defect resulting in an increased cell doubling time and a decrease in the proportion of cells in S + G2M phase of the cell cycle. This is linked to a slowing in the rate of differentiation as determined by comparative cell-surface marker analysis and ultimate failure of the differentiation program at the late-erythroblast stage of development. The dyserythropoiesis was also linked to an increased tendency of the RAS-expressing cells to undergo programmed cell death during their differentiation program. This erythroid lineage dysplasia recapitulates one of the most common features of myelodysplastic syndrome, and for the first time provides a causative link between mutational activation of N-RAS and the pathogenesis of preleukemia.

Mutant RAS selectively promotes sensitivity of myeloid leukemia cells to apoptosis by a protein kinase C-dependent process.

RAS mutations arise at high frequency in human malignancy and have been shown to play a role in the disruption of both normal differentiation and proliferation. In addition, RAS influences a number of intracellular signaling pathways, which impinge on proteins that regulate programmed cell death. In this study, we have examined whether this oncogene can influence the activation of the apoptotic process induced by a range of therapeutic agents used to treat leukemia, and we have identified the downstream targets of RAS mediating the observed changes in sensitivity. Using myeloid leukemia cells (P39) retrovirally transduced with mutant H-RAS, we found that the influence of this oncogene was highly dependent on the inducer used: whereas RAS had no significant effect on spontaneous apoptosis or on the response to the cytotoxic drugs (doxorubicin or 1-beta-arabinofuranosylcytosine), P39-RAS cells showed a strongly augmented response to all-trans-retinoic acid (ATRA) in both the induction of apoptosis and differentiation. Because, under some circumstances, RAF has been associated with promoting apoptosis, we examined whether the activation of this kinase by mutant RAS could be responsible for the augmented response to ATRA. However, constitutive activation of RAF did not alter the apoptotic sensitivity of these cells, making it unlikely that RAS promotes apoptosis by stimulating this kinase. Nor did we find that BCL-2 was differentially down-regulated in P39-RAS cells. Rather, we found that the activation of protein kinase C (PKC) by low-dose phorbol ester could almost entirely recapitulate transformation by RAS, in terms of promoting both apoptosis and differentiation after treatment with ATRA. Moreover, the RAS-induced phenotype could be completely abolished by a specific inhibition of PKC under conditions that had no effect on the response of control cells. In conclusion, we have shown that mutant RAS promotes differentiation-associated cell death in P39 cells by stimulating the activity of PKC, which is itself an important regulator of myeloid differentiation. PKC activation, in turn, powerfully synergizes with the PKC-independent action of ATRA. This work identifies a possible explanation for the ability of this oncogene to promote myeloid differentiation of hematopoietic cells. Clinically, it raises the possibility that although leukemias expressing mutant RAS may not show an altered response to cytotoxic agents, they may show enhanced sensitivity to differentiation therapy with ATRA.

Aberrant expression of p21RAS but not p120GAP is a common feature of myelodysplasia.

The RAS gene family encodes signal transducing proteins which are involved in the regulation of cell growth and differentiation. Constitutively 'activating' point mutations of RAS have been detected at high frequency in preleukaemia and acute leukaemia, however, the distribution of expression of p21RAS in normal or preleukaemic primary haematopoietic cells has not been studied. We have examined the expression of p21RAS and its negative regulator/downstream effector, p120GAP, in combination with lineage-specific cluster of differentiation markers in normal and preleukaemic myeloid bone marrow cells using flow cytometry. Normal bone marrow was characterized by low and uniform levels of p21RAS expression throughout all lineages analysed. In contrast, 28% (9/32) of patients with myelodysplastic syndrome (MDS) over-expressed p21RAS. In three of these patients a single over-expressing peak of p21RAS expression was observed, with no evidence of a population exhibiting expression within the normal range. In 6/32 MDS patients over-expression of p21RAS was observed only in a subpopulation of the myeloid cells. Follow-up samples were analysed in three of these six patients; over-expression was confirmed in each patient and in two patients a relative expansion of the over-expressing cell population was observed. Eight out of nine of the patients with aberrant p21RAS expression were diagnosed with low-risk MDS. From 21 MDS patients screened for p120GAP expression, no reduction or loss of expression was observed, however, one AML patient demonstrated a heterogeneous pattern of expression. p120GAP expression was lower (P < 0.05) in the AML group than in normals. The results of the study suggest that over-expression of the RAS gene product, p21RAS, may represent an alternative or additional mechanism of activation of the RAS signalling pathway and that this may play a role in leukaemogenesis, however, there is no evidence from this study that loss of p120GAP expression is a feature of myelodysplasia.

Expression of AML1-ETO in human myelomonocytic cells selectively inhibits granulocytic differentiation and promotes their self-renewal.

The t(8;21) translocation is one of the most frequent translocations in acute myeloid leukaemia (AML), giving rise to the AML1-ETO fusion protein (or RUNX1-CBF2T1). This abnormality is associated with myelocytic leukaemia with dysplastic granulopoiesis. Here, we demonstrate that when expressed in a normal human (CD34(+)) progenitor population, AML1-ETO selectively inhibits granulocyte colony formation but not monocyte colony formation. In bulk liquid culture, we found that though AML1-ETO transiently inhibited the proliferation of CD34(+) cells, it promoted long-term growth of myeloid cells for more than 80 days, suggesting that differentiation was inhibited. In support of this, cultures expressing AML1-ETO demonstrated enhanced retention of colony-forming capacity. Phenotypic examination of AML1-ETO cultures revealed a defect in granulocytic differentiation in terms of retention of CD34(+) cells within the culture and delayed CD11b upregulation. Morphologically, granulocyte terminal differentiation in AML1-ETO-expressing cells was inhibited by 83+/-5%, giving rise to a build-up of early to intermediate granulocytes that exhibited a number of morphological features associated with t(8;21) leukaemias. In contrast, AML1-ETO had little or no effect on monocytic differentiation. Taken together, these results suggest that expression of AML1-ETO selectively inhibits the differentiation of granulocytic cells and promoted extensive self-renewal, supporting a causal role for t(8;21) translocations in leukaemogenesis.

Mutant RAS inhibits neutrophil but not macrophage differentiation and allows continued growth of neutrophil precursors.

Mutational activation of RAS is the most common molecular abnormality in myeloid leukemias. In order to better understand its role in leukemogenesis, we have devised a model based on the multipotent cell line, FDCP-mix. We show that expression of mutant RAS in FDCP-mix strongly inhibits terminal neutrophil differentiation under the influence of G-CSF plus GM-CSF at the metamyelocyte stage, whereas macrophage differentiation was unaffected. In addition, whereas control cultures differentiated and became postmitotic under these conditions, FDCP-mix cells expressing mutant RAS continued to proliferate indefinitely while maintaining a metamyelocytic phenotype. Labeling of these cultures with the fluorescent tracking dye, PKH26, showed that this extended proliferative capacity resulted from continued division of metamyelocytes in the culture. Dissection of the growth factor response of these cells demonstrated that GM-CSF was critical in maintaining proliferation and inhibiting the differentiation of these cells. We further show the block in neutrophil differentiation could be partially overcome by treatment with low-dose Ara C, suggesting that maintenance of cell cycle progression may be partly responsible for the anti-differentiation effect of this oncogene. These findings suggest that activation of RAS is able to specifically inhibit terminal neutrophil differentiation and in so doing promotes continued division of metamyelocyte cells.

Oncogenic RAS genes impair erythroid differentiation of erythroleukaemia cells.

RAS mutations occur frequently in acute myeloid leukaemia and myelodysplasia, suggesting a functional role for this oncogene in leukaemogenesis. We show here, for the first time, that both N-RAS and H-RAS can impair erythroid differentiation of erythroleukaemia cells induced with hexamethylene bisacetamide. Transformation by RAS allowed extended proliferation in the presence of inducer and also inhibited maturation as measured by impaired haemoglobinization and reduction in cell size. These data provide an interesting counterpoint to the effect of mutant RAS on monocytic cells, where it has a potentiating effect on differentiation and may indicate a causal link between the activation of RAS and erythroid lineage dysplasia in preleukaemia.

Upregulation of p21 RAS levels in HL-60 cells during differentiation induction with DMSO, all-trans-retinoic acid and TPA.

The role of p21 RAS in the proliferation and differentiation of myeloid cells has been studied by analysing the changes in the level of expression of p21 RAS proteins by flow cytometry upon differentiation down the granulocytic and monocytic pathways. Differentiation resulted in upregulated p21 RAS expression despite a marked decline in the number of dividing cells. On the other hand, growth inhibition, without differentiation, resulted in a decline in expression. Cell cycle analysis showed that the increase in p21 RAS occurred throughout the cell cycle. These results suggest that p21 RAS has a role in the process of myeloid differentiation.

Alternative effects of RAS and RAF oncogenes on the proliferation and apoptosis of factor-dependent FDC-P1 cells.

Despite the fact that RAF-1 lies immediately downstream of p21RAS in the MAP kinase-signalling cascade, recent evidence in non-haematopoietic environments suggest that RAS and RAF can transduce signals through alternative pathways specific to a particular cell type. Since mutational activation of RAS occurs at high frequency in human leukaemia, we have investigated the contribution of signalling from mutant RAF in mediating the transforming effects of the N-RAS oncogene in the growth factor-dependent cell line, FDC-P1. Independent activation of N-RAS extended the period of exponential growth leading to an increased saturating density under optimal growth conditions. Under conditions of growth factor withdrawal, cells expressing mutant RAS, but not control cells, demonstrated protection against apoptotic death. Although RAF promoted cell proliferation in a similar manner to that observed in FDCP-RAS cells, expression of mutant RAF was not as effective at protecting these cells against apoptotic death following growth factor withdrawal. The results suggest that RAS utilises RAF-dependent signals in promoting the proliferation of FDC-P1 cells but the anti-apoptotic effects of this oncogene are mediated through a RAF- and BCL-2-independent pathway.

γ-Catenin is overexpressed in acute myeloid leukemia and promotes the stabilization and nuclear localization of ß-catenin.

Canonical Wnt signaling regulates the transcription of T-cell factor (TCF)-responsive genes through the stabilization and nuclear translocation of the transcriptional co-activator, ß-catenin. Overexpression of ß-catenin features prominently in acute myeloid leukemia (AML) and has previously been associated with poor clinical outcome. Overexpression of γ-catenin mRNA (a close homologue of ß-catenin) has also been reported in AML and has been linked to the pathogenesis of this disease, however, the relative roles of these catenins in leukemia remains unclear. Here we report that overexpression and aberrant nuclear localization of γ-catenin is frequent in AML. Significantly, γ-catenin expression was associated with ß-catenin stabilization and nuclear localization. Consistent with this, we found that ectopic γ-catenin expression promoted the stabilization and nuclear translocation of ß-catenin in leukemia cells. ß-Catenin knockdown demonstrated that both γ- and ß-catenin contribute to TCF-dependent transcription in leukemia cells. These data indicate that γ-catenin expression is a significant factor in the stabilization of ß-catenin in AML. We also show that although normal cells exclude nuclear translocation of both γ- and ß-catenin, this level of regulation is lost in the majority of AML patients and cell lines, which allow nuclear accumulation of these catenins and inappropriate TCF-dependent transcription.

CD200 expression suppresses natural killer cell function and directly inhibits patient anti-tumor response in acute myeloid leukemia.

Upregulation of the immunosuppressive cell surface glycoprotein, CD200, is a common feature of acute myeloid leukemia (AML) and is associated with poor patient outcome. We investigated whether CD200 overexpression on AML cells could specifically compromise patient natural killer (NK) cell anti-tumor responses. We found that CD200(hi) patients showed a 50% reduction in the frequency of activated NK cells (CD56(dim)CD16(+)) compared with CD200(lo) patients. Additionally, NK receptor expression (NKp44 and NKp46) on these cells was also significantly downregulated in CD200(hi) patients. To assess whether NK cell activity was directly influenced by CD200 expression, we examined the effect of ectopic expression of CD200. These assays revealed that both NK cell cytolytic activity and interferon-γ response were significantly reduced toward CD200(+) leukemic targets and that these targets showed increased survival compared with CD200(-) cells. Similarly, NK cells isolated from AML patients were less functionally active toward CD200(hi) autologous blasts from both cytolytic and immunoregulatory perspectives. Finally, blocking CD200 alone was sufficient to recover a significant proportion of NK cell cytolytic activity. Together, these findings provide the first evidence that CD200 has a direct and significant suppressive influence on NK cell activity in AML patients and may contribute to the increased relapse rate in CD200(+) patients.

OGG1 is a novel prognostic indicator in acute myeloid leukaemia.

OGG1 (8-oxoguanine DNA glycosylase) constitutes a key component of the DNA base excision repair pathway, catalysing the removal of 8-oxoguanine nucleotides from DNA, thereby suppressing mutagenesis and cell death. We found that OGG1 expression was significantly downregulated by the RUNX1-ETO fusion protein product of the t(8;21) chromosome translocation in normal haematopoietic progenitor cells and in patients with acute myeloid leukaemia (AML). Further examination of OGG1 expression in 174 AML trial patients using Affymetrix microarrays showed that the prevalence rate of OGG1 expression was 33% and correlated strongly with adverse cytogenetics. OGG1-expressing patients had a worse relapse-free survival and overall survival and an increased risk of relapse at 5-years of follow-up. There remained a trend towards increased relapse rate among OGG1-expressing patients, even after adjusting for other known risk factors in comprehensive stratified analyses. We also determined a trend for OGG1 expression to have a more adverse impact on disease outcome in the context of the FLT3-ITD mutation. This study highlights OGG1 as a valuable prognostic marker that could be used to sub-stratify AML patients to predict those likely to fail conventional chemotherapies but those likely to benefit from novel therapeutic approaches that modulate DNA repair activity.

Transcriptional dysregulation mediated by RUNX1-RUNX1T1 in normal human progenitor cells and in acute myeloid leukaemia.

The t(8;21)(q22;q22) occurs frequently in acute myelogenous leukaemia and gives rise to the transcription factor fusion protein, RUNX1-RUNX1T1 (also known as AML1-ETO). To identify the genes dysregulated by the aberrant transcriptional activity of RUNX1-RUNX1T1, we used microarrays to determine the effect of this mutation on gene expression in human progenitor cells and during subsequent development. Gene signatures of these developmental subsets were very dissimilar indicating that effects of RUNX1-RUNX1T1 are highly context dependent. We focused on gene changes associated with the granulocytic lineage and identified a clinically relevant subset of these by comparison with 235 leukaemia patient transcriptional signatures. We confirmed the overexpression of a number of significant genes (Sox4, IL-17BR, CD200 and gamma-catenin). Further, we show that overexpression of CD200 and gamma-catenin is also associated with the inv(16) abnormality which like RUNX1-RUNX1T1 disrupts core binding factor activity. We investigated the functional significance of CD200 and gamma-catenin overexpression in normal human progenitor cells. The effect of IL17 on growth was also assessed. Individually, none of these changes were sufficient to recapitulate the effects of RUNX1-RUNX1T1 on normal development. These data provide the most comprehensive and pertinent assessment of the effect of RUNX1-RUNX1T1 on gene expression and demonstrate the highly context-dependent effects of this fusion gene.

Sequential changes in MHC antigen expression induced by the v-Ki-ras oncogene.

A series of early-passage cell lines were transformed with the v-Ki-ras oncogene with the aim of examining the effect of an activated ras gene on the ability of these cells to express major histocompatibility complex (MHC) antigens. These cell lines were found to undergo multiple phenotypic changes upon transformation and subsequent proliferation. At early passage, the predominant effect of ras was an increased ability to express class II antigens when induced with interferon gamma (IFN gamma). For class I antigens, maximum levels of expression induced with IFN gamma were largely unaffected, however, decreased sensitivity to induction with this lymphokine was noted. With subsequent in vitro or in vivo passage, both class I and class II antigen inducibility was attenuated. The latter phenotypic change was found to be transferable by coculture, implicating a soluble IFN gamma antagonist. Conditioned media from ras-transformed cells treated to activate their latent transforming growth factor beta (TGF beta) content mediated similar changes in MHC antigen inducibility, suggesting that TGF beta may be involved in modulating MHC antigen expression in ras-transformed cells.