RAS proteins are small GTPases that play a central role in transducing signals that regulate cell proliferation, survival, and differentiation. The RAS proteins interact with a common set of activators and effectors; however, they associate with different microdomains of the plasma membrane as well as other endomembranes and are capable of generating distinct signal outputs. Mutations that result in constitutive activation of RAS proteins are associated with ∼30% of all human cancers; however, different RAS oncogenes are preferentially associated with different types of human cancer. In myeloid malignancies, NRAS mutations are more frequent than KRAS mutations, whereas HRAS mutations are rare. The mechanism underlying the different frequencies of RAS isoforms mutated in myeloid leukemia is not known. In this study, we compared the leukemogenic potential of activated NRAS, KRAS, and HRAS in the same bone marrow transduction/transplantation model system. We found that all three RAS oncogenes have the ability to induce myeloid leukemias, yet have distinct leukemogenic strengths and phenotypes. The models established here provide a system for further studying the molecular mechanisms in the pathogenesis of myeloid malignancies and for testing targeted therapies. [Cancer Res 2007;67(15):7139–46]

RAS proteins are small GTPases that act as molecular switches to transduce signals from activated receptors. They do so by cycling between a GDP-bound inactive state and a GTP-bound active state. When in its GTP-bound state, RAS can bind to and activate a range of downstream effector proteins, which may then result in diverse cellular outcomes like cell proliferation, survival, differentiation, and neoplastic transformation (reviewed in refs. 1, 2).

Three RAS genes code for four highly homologous RAS proteins, NRAS, HRAS, and KRAS4B/KRAS4A (splice variants). These proteins have identical effector binding domains and hence can interact with the same set of downstream effectors. However, due to differences in their posttranslational modifications, they have different trafficking routes and localize to distinct microdomains of the plasma membrane and other endomembranes (3). As a result of this, they may have access to different effector pools and may be capable of generating distinct signal outputs (4). Indeed, RAS isoforms have been shown to differ in their abilities to activate various downstream proteins (57). Oncogenic versions of HRAS are better than NRAS or KRAS at transforming fibroblast cells, whereas NRAS is better at transforming hematopoietic cells (8). Gene knockout studies further highlight these differences. Knocking out NRAS or HRAS or both in mice results in essentially normal animals, whereas KRAS-deficient mice are embryonic lethal (9, 10).

Nearly 30% of human cancers, including solid tumors and hematologic malignancies, are associated with mutations in RAS genes. Interestingly, mutations in different RAS isoforms are preferentially associated with cancers of different organs (11). For example, KRAS mutations are found in nearly 90% of pancreatic cancers. In myeloid malignancies, NRAS mutations are more frequent than KRAS mutations, whereas HRAS mutations are rare. The mechanism underlying the different frequencies of RAS isoforms mutated in myeloid malignancies is not known.

The leukemogenic potential of oncogenic RAS has been studied in animals by transgenic as well as bone marrow transduction/transplantation (BMT) models. Transgenic mice expressing HRAS under the mouse mammary tumor virus promoter/enhancer developed B-lymphoblastic leukemia, whereas expression of HRAS in a BMT model induced B and T lymphoid leukemia/lymphoma (12, 13). Transgenic mice expressing NRAS under the IgH Eμ enhancer or the hMRP8 promoter developed T lymphoid leukemias or epithelial tumors (14, 15). Expression of NRAS under the Moloney murine leukemia virus long terminal repeat (Mo-MuLV LTR) in a BMT model induced myeloid malignancies with a long latency and incomplete penetrance (16). These studies suggested that activation of RAS by itself might not be sufficient to induce myeloid leukemias. However, recently, others and we have shown that expression of activated mutants of NRAS and KRAS can efficiently induce myeloid leukemias in mice (1719). Expression of oncogenic NRAS using a BMT model induces an acute myeloid leukemia (AML)– or chronic myelomonocytic leukemia (CMML)–like disease in mice, whereas expression of oncogenic KRAS under its endogenous promoter in a conditional knock-in strain gives rise to a CMML-like disease in all the mice. Because oncogenic RAS proteins were studied in different model systems, it is not clear whether the difference in phenotypes of RAS oncoproteins is due to the different methods used to express the oncogenes or due to differences in their intrinsic leukemogenic potentials. Given that RAS proteins have both shared and distinct biochemical and biological functions, direct comparison of their leukemogenic potentials could provide insights into the mechanism of RAS leukemogenesis and help to identify critical targets of RAS for developing therapies.

In this study, we sought to compare NRAS, KRAS, and HRAS leukemogenesis by expressing them in the same model system. We find that all NRAS, KRAS, and HRAS have the potential to induce myeloid leukemia in mice, but differ in terms of their potency and disease phenotype.

DNA constructs. The NRASD12, KRASD12, and HRASV12 genes were amplified by PCR from expression sequence tags for human NRAS (Genbank accession no. N44803) and HRAS (Genbank accession no. AF493916) and cDNA for KRAS (Genbank accession no. AF493917; UMR cDNA Resource Center). The activated mutants of NRAS, KRAS, and HRAS were generated by PCR, by introducing mutations in codon 12 of these genes, which changed glycine to aspartic acid (G12D) in NRAS and KRAS and into valine (G12V) in HRAS. The changes in codon 12 were GGT to GAT for NRASD12 and KRASD12 and GGC to GTC for HRASV12. The genes were cloned into 5′-NotI and 3′-ClaI restriction enzyme sites in the murine stem cell virus (MSCV) vector, downstream of an internal ribosomal entry site (IRES) from the encephalomyocarditis virus (EMCV). All NRASD12, KRASD12, and HRASV12 were expressed as myc-tagged proteins downstream of the IRES element, whereas green fluorescent protein (GFP) was expressed under the control of the MSCV LTR. The control vector, MSCV-IRES-GFP (MIG) consists of only the GFP gene downstream of the EMCV-IRES. The fidelity of PCR cloning was confirmed by sequencing.

Retrovirus production and determination of viral titer. Bosc23 cells, a retroviral packaging cell line, were used for producing retroviruses used for infecting cell lines as well as primary bone marrow as previously described (20). Briefly, 2 × 106 Bosc23 cells were plated ∼12 h before transfection. Five micrograms of the DNA of interest were transfected into the Bosc23 cells by the calcium phosphate precipitation method. The transfection medium was replaced with fresh medium ∼10 h posttransfection. The viral supernatant from the transfected cells was collected 48 h after transfection. The viral titer was calculated in transducing units (TU) by multiplying the percentage of NIH3T3 cells expressing GFP and the number of cells seeded. The titers in this experiment were ∼4 × 105 TU/mL for NRAS, KRAS, and HRAS (matched) and ∼5 × 106 TU/mL for KRAS-high titer condition and vector control.

Bone marrow transduction/transplantation. Bone marrow cells were isolated from 6- to 8-week-old, 5-fluorouracil (5-FU; 250 mg/kg) treated, donor BALB/c mice (Taconic Farms). They were infected with retroviruses, and cultured for 2 days in vitro, in the presence of cytokines. They were then transplanted, by injecting 4 × 105 cells, into the tail vein of each of the lethally irradiated (2 × 450 rad, 3 h between each dose) female recipient BALB/c mice as described (21). The recipient mice were monitored for signs of disease from day 14 posttransplantation.

Flow cytometry analysis. Peripheral blood cells or bone marrow cells (0.5–1 million) were used for staining with each of the following antibodies, either alone or in combination, after blocking with purified anti-mouse CD16-CD32 (2.4G2, BD PharMingen). The following antibodies used in the analysis were also purchased from the same source: phycoerythrin-conjugated Gr-1 (RB6-8C5), B220 (RA3-6B2), CD19 (1D3), Thy-1.2 (30-H12), CD86, CD31, CD115 (M-CSFR), and Ter-119; biotinylated Mac-1 (M1/70), CD34 (RAM34), CD38 (90), CD16/32, and c-Kit (2B8); allophycocyanin-conjugated CD45 (30-F11) and streptavidin-allophycocyanin. Phycoerythrin-conjugated anti-mouse F4/80 (CI:A3-1) and anti-mouse macrophages-monocytes (MOMA-2) were purchased from Serotec. Flow cytometry was done on the FACSCalibur machine, and data were analyzed using FlowJo (Tree Star) software.

Hematopathologic analysis. Blood was collected from mice by tail bleed, and 3 μL were diluted in 3 mL of Isoton II (Fisher Scientific). WBC counts were measured using the Coulter Counter model Z1 (Coulter), after lysing the RBC with ZAP-O-Globin (Beckman Coulter). Hematocrit was measured by capillary centrifugation on a microhematocrit centrifuge (StatSpin). Smears, cytospin, and touch preparations of blood and other murine tissues were stained with Hema 3 stain set (Fisher) for routine identification of cell morphology. Histology was done on 4-μm-thick paraformaldehyde-fixed, paraffin-embedded tissue sections. After deparaffinization, the sections were stained by H&E.

Southern blot analysis. Genomic DNA isolated from liver or spleen tumor cells using the Promega Wizard Genomic DNA Purification Kit (Promega) was digested with either XbaI or BglII. Fifteen micrograms of digested DNA were electrophoresed on 1% agarose gel and transferred onto Zeta-probe GT membrane (Bio-Rad). The membrane was hybridized with a 32P-labeled 0.7 kb NcoI-SalI fragment corresponding to GFP as a probe. The blot was stripped with 0.1% boiling SDS solution and reprobed with 32P-labeled 1.4 kb irf-4 DNA fragment as a loading control. To generate a control for single-copy proviral integration, we did single-cell sorting of 32D cells transduced with MSCV-BCR/ABL-IRES-GFP. We then tested the single-cell clones by Southern blotting, and isolated a cell line with a single-copy provirus (Fig. 5A , lane 1).

Western blot analysis. Cell lysates were prepared from infected NIH3T3 cells and from ACK-treated single-cell suspensions of mouse tissues by adding equal volume of 2× Laemmli sample buffer to the cell suspensions in PBS. For assays using 32Dcl3 cells, lysates were prepared after withdrawing interleukin 3 for 12 h. The lysates were heated at 100°C for 5 min and centrifuged to remove debris. Lysates were then resolved on 6% to 18% gradient polyacrylamide gels, transferred to nitrocellulose membranes, and blotted with the following primary antibodies: anti-Ras (RAS10, Upstate Biotechnology); anti-actin (AC40, Sigma); anti-myc tag 9E10 monoclonal antibody (from conditional medium of 9E10 hybridoma cell line); anti-dynamin (BD Biosciences); and pAkt, Akt, pMek1/2, Mek1/2, pErk42/44, Erk42/44, pS6rp, and S6rp (All 1:1,000, Cell Signaling Technologies). Horseradish peroxidase–labeled goat anti-mouse IgG or goat anti-rabbit IgG (Southern Biotechnology) was used as secondary antibodies. Calculation of ratios of RAS to actin or dynamin was done on a Macintosh computer using the public domain NIH Image program (developed at the NIH and available online).3

Oncogenic NRAS, KRAS, and HRAS induce fatal diseases in mice with different latencies. To compare the leukemogenic potentials of activated mutants of NRAS, KRAS, and HRAS, we constructed retroviral vectors expressing myc-tagged NRASD12, KRASD12, and HRASV12 (D12 and V12 mutations were chosen based on their being most frequent mutations found in the corresponding RAS genes; Fig. 1A). The expression of the activated RAS proteins in NIH3T3 cells was confirmed by Western blotting (Fig. 1B). We then infected bone marrow cells isolated from 5-FU–treated mice with retroviruses containing NRAS, KRAS, or HRAS or vector alone and transplanted these cells into lethally irradiated syngeneic recipient mice. 5-FU is a pyrimidine analogue that induces cell cycle arrest and apoptosis by inhibiting the ability of the cell to synthesize DNA (22). Its role here is to eliminate the proliferating hematopoietic precursor cells and to enrich and stimulate hematopoietic stem cells.

Figure 1.

Retroviral vectors and RAS expression. A, diagrammatic representation of retroviral vectors used in this study. Bicistronic vectors were used that could coexpress RAS and GFP. Vector control expresses GFP alone. Positions of restriction enzyme sites as well as probe used for Southern blot are indicated. B, Western blot analysis of lysates prepared from NIH3T3 cells infected with NRAS, KRAS, and HRAS retroviruses, as well as the vector control (lanes 1, 2, 3, and 4, respectively) using anti-panRAS and Myc-tag antibodies. Detection of dynamin using an anti-dynamin antibody was used as a loading control.

Figure 1.

Retroviral vectors and RAS expression. A, diagrammatic representation of retroviral vectors used in this study. Bicistronic vectors were used that could coexpress RAS and GFP. Vector control expresses GFP alone. Positions of restriction enzyme sites as well as probe used for Southern blot are indicated. B, Western blot analysis of lysates prepared from NIH3T3 cells infected with NRAS, KRAS, and HRAS retroviruses, as well as the vector control (lanes 1, 2, 3, and 4, respectively) using anti-panRAS and Myc-tag antibodies. Detection of dynamin using an anti-dynamin antibody was used as a loading control.

Close modal

Using this system, we have shown that activated NRAS rapidly and efficiently induces a CMML- or AML-like disease in mice (19). In this study, NRAS is used as a control and the data presented here are from the new experiment where we compare NRAS, KRAS, and HRAS leukemogenesis. We found that oncogenic mutants of all three RAS isoforms are capable of inducing fatal diseases in mice. Under the same experimental conditions (same retroviral titers and same pool of donor bone marrow cells), the disease latency of HRASV12 mice was significantly shorter than NRASD12 and KRASD12 mice (P < 0.001 and P < 0.0001, respectively). Although majority of the NRASD12 mice died faster than KRASD12 mice, the difference in their disease latencies was not statistically significant (P = 0.2; Fig. 2; Table 1). All vector control mice remained healthy during the course of the experiment.

Figure 2.

Cumulative survival of oncogenic NRAS, KRAS, and HRAS mice. Cumulative survival curves of mice transplanted with oncogenic NRAS, KRAS, and HRAS or vector control (MIG) transduced bone marrow cells were generated by Kaplan-Meier survival analysis. Donor bone marrow cells were transduced under closely matched titers of NRAS, KRAS, and HRAS12 retrovirus: ∼4 × 105 TU/mL for all RAS and 5 × 106 TU/mL for KRASD12 high titer (HT) and vector control.

Figure 2.

Cumulative survival of oncogenic NRAS, KRAS, and HRAS mice. Cumulative survival curves of mice transplanted with oncogenic NRAS, KRAS, and HRAS or vector control (MIG) transduced bone marrow cells were generated by Kaplan-Meier survival analysis. Donor bone marrow cells were transduced under closely matched titers of NRAS, KRAS, and HRAS12 retrovirus: ∼4 × 105 TU/mL for all RAS and 5 × 106 TU/mL for KRASD12 high titer (HT) and vector control.

Close modal
Table 1.

Summary of hematopathologic characteristics of oncogenic NRAS, KRAS, and HRAS mice

ConstructNo. animalsDiseaseLatency D*WBCs, ×103/μLHematocrit (%)Liver weight (g)Spleen weight (g)Pulmonary hemorrhage
MIG§ None N/A 6–12 53.6 ± 3 1.0 ± 0.1 0.1 ± 0.04 — 
NRASD12 AML 49.5 6–26 21.5 ± 2 1.7 ± 0.3 0.8 ± 0.5 — 
 CMML 62.5 42–103 26.5 ± 4.5 1.9 ± 0.2 1.0 ± 0.45 — 
KRASD12 CMML 64 14–73 27 ± 6 1.7 ± 0.5 0.9 ± 0.1 — 
KRASD12-HT CMML 40 10–57 20 ± 6.6 1.45 ± 0.3 0.9 ± 0.3 — 
HRASV12 AML 43 8–24 23 ± 10 1.4 ± 0.1 0.7 ± 0.2 ++ 
ConstructNo. animalsDiseaseLatency D*WBCs, ×103/μLHematocrit (%)Liver weight (g)Spleen weight (g)Pulmonary hemorrhage
MIG§ None N/A 6–12 53.6 ± 3 1.0 ± 0.1 0.1 ± 0.04 — 
NRASD12 AML 49.5 6–26 21.5 ± 2 1.7 ± 0.3 0.8 ± 0.5 — 
 CMML 62.5 42–103 26.5 ± 4.5 1.9 ± 0.2 1.0 ± 0.45 — 
KRASD12 CMML 64 14–73 27 ± 6 1.7 ± 0.5 0.9 ± 0.1 — 
KRASD12-HT CMML 40 10–57 20 ± 6.6 1.45 ± 0.3 0.9 ± 0.3 — 
HRASV12 AML 43 8–24 23 ± 10 1.4 ± 0.1 0.7 ± 0.2 ++ 
*

Median latency within the particular group.

WBC count is given as the range of WBCs for the diseased mice within a particular disease group.

Hematocrit, liver, and spleen weights are average ± SD.

§

Mice transduced with vector control, MSCV-IRES-GFP (MIG).

Not applicable.

High titer.

Activated NRAS, KRAS, and HRAS induce myeloid leukemias with distinct phenotypes. We sacrificed moribund mice to determine the disease phenotype in each case. As shown previously, the mice receiving NRASD12 transduced bone marrow cells succumbed to either of two diseases. Some mice developed leukocytosis, anemia, enlarged spleens, and livers. The leukemic cells were predominantly mature granulocytes and monocytes. This disease resembled human CMML. The other disease was characterized by the presence of immature myeloid cells, severe anemia and hepato-splenomegaly, and resembled human AML (19). The frequencies of these different diseases seemed to correlate with the retroviral titers used, with higher titers favoring AML (Fig. 3 and data not shown).

Figure 3.

Immunophenotyping of leukemic cells from diseased oncogenic NRAS, KRAS, and HRAS mice. Bone marrow or peripheral blood cells were isolated from diseased mice, stained with antibodies to various cell surface markers as indicated, and subjected to flow cytometry analysis. GFP expression is along the X axis, and Y axis shows expression of the cell surface marker specified over each column.

Figure 3.

Immunophenotyping of leukemic cells from diseased oncogenic NRAS, KRAS, and HRAS mice. Bone marrow or peripheral blood cells were isolated from diseased mice, stained with antibodies to various cell surface markers as indicated, and subjected to flow cytometry analysis. GFP expression is along the X axis, and Y axis shows expression of the cell surface marker specified over each column.

Close modal

Analysis of the peripheral blood, bone marrow, livers, and spleens of all diseased KRASD12 mice by smears and touch preparations revealed the accumulation of mainly mature myeloid cells. Flow cytometry analysis of peripheral WBC showed two major populations. One expressed the granulocytic surface markers (Mac-1+, Gr-1hi, CD16/32+) and the other expressed monocytic cell surface markers (Mac-1+, Gr-1lo/−, F4/80+, M/CSFR+, Moma-2+, CD16/32+, CD31+, CD86+, and CD38+; Fig. 3). The level of leukocytosis varied among the mice, with the majority developing WBC counts higher than 50,000/μL, but some with counts between 10,000/μL and 50,000/μL (Table 1). However, even in these mice, most cells were mature myeloid cells. These mice also develop anemia and have enlarged livers and spleens due to leukemic infiltration and extramedullary hematopoiesis (Table 1). This disease is similar to the CMML-like disease seen in some NRASD12 mice.

Because mice expressing NRASD12 develop either an AML-like or a CMML-like disease that correlate with retroviral titers, we wanted to check if this was true for KRASD12. We therefore conducted the experiment under different titer conditions, ranging from extremely high titers to much lower titers and found that the disease latency does correlate with the retroviral titer used—mice receiving higher titers succumb to disease faster than those receiving lower titers of the same virus (Fig. 2 and data not shown). However, unlike oncogenic NRAS, all KRAS mice succumbed to a similar CMML-like disease regardless of the titers used (Table 1 and data not shown).

Analysis of peripheral blood, bone marrow, livers, and spleens from HRASV12 mice revealed the abnormal presence of immature myeloid cells. Flow cytometry analyses show that the majority of the GFP-positive cells are Mac-1+ CD38+ and CD16/32+ but Gr-1lo/-, M/CSFR, and MOMA-2. A fraction of the leukemic cells also express CD117 (c-Kit), CD34, Thy1.2, F4/80, CD86, and CD31, indicating a mixed blast and differentiated monocytic population (Fig. 3). These mice also develop severe anemia, as well as massively enlarged livers and spleens (Table 1). This disease resembles the AML-like disease seen in some NRASD12 mice. Importantly, when we did experiments using lower titers of HRAS, we found that all the mice still succumbed to an AML-like disease (data not shown).

Although both NRAS and HRAS induce an AML-like disease in mice, the invasiveness of the two diseases is drastically different. NRASD12 mice showed some leukemic infiltration in the lung tissue, but mice receiving HRASV12 showed massive infiltration of lungs with leukemic cells, along with widespread pulmonary hemorrhages (Fig. 4). This high invasiveness of the tumor correlates with the shorter disease latency of HRAS mice.

Figure 4.

Invasiveness of NRAS- and HRAS-induced AML-like disease in mice. Paraffin sections of lung tissue from vector control and diseased NRAS and HRAS mice were stained with H&E. Images are magnified as indicated.

Figure 4.

Invasiveness of NRAS- and HRAS-induced AML-like disease in mice. Paraffin sections of lung tissue from vector control and diseased NRAS and HRAS mice were stained with H&E. Images are magnified as indicated.

Close modal

In addition to these predominant phenotypes, majority of the KRAS and the HRAS mice (similar to the NRAS mice) have increased mast cells in the peripheral blood, bone marrow, livers, and spleens. This indicates that all oncogenic RAS mutants can also induce mastocytosis in mice.

NRAS-, KRAS-, and HRAS-induced myeloid leukemias are oligoclonal. To further compare the leukemogenic potentials of activated mutants of RAS isoforms, we examined the clonality of the leukemias by Southern blot analysis (Fig. 5). Restriction enzyme BglII, which recognizes a unique site in the RAS proviral DNA, was used to check proviral integration, and XbaI, which cleaves the RAS proviral DNA at the LTRs, was used to check the integrity and the total amount of proviruses (Fig. 1A). Genomic DNA isolated from a cloned MSCV-BCR/ABL-IRES-GFP–infected 32D cell line was used as a single-copy provirus control (Materials and Methods and Fig. 5, lane 1). After probing with 32P-labeled GFP, the blot was stripped and reprobed with a 32P-labeled irf-4 DNA fragment to check for loading. We found that the majority of the leukemias induced by NRAS, KRAS, and HRAS are oligoclonal. This suggests that all oncogenic RAS genes may require secondary oncogenic events to induce leukemias.

Figure 5.

Proviral integration in leukemic cells from oncogenic NRAS, KRAS, and HRAS mice. A, genomic DNA isolated from leukemic cells of diseased RAS mice (NRAS CMML: lanes 2–3, NRAS AML: lanes 4–6, HRAS: lanes 7–9, KRAS: lanes 10–13) was digested with BglII to analyze proviral integration. A 0.7 kb 32P-labeled GFP fragment was used as the probe. Lane 1 is a control to establish levels of a single copy of the provirus. The blot was stripped and reprobed with a 1.4-kb 32P-labeled irf4 probe as loading control. B, genomic DNA was digested with XbaI (which cuts within the LTRs) to show the intact provirus. Top, MSCV-BCR/ABL-IRES-GFP single-copy proviral control (larger in size than RAS provirus). Bottom, loading control.

Figure 5.

Proviral integration in leukemic cells from oncogenic NRAS, KRAS, and HRAS mice. A, genomic DNA isolated from leukemic cells of diseased RAS mice (NRAS CMML: lanes 2–3, NRAS AML: lanes 4–6, HRAS: lanes 7–9, KRAS: lanes 10–13) was digested with BglII to analyze proviral integration. A 0.7 kb 32P-labeled GFP fragment was used as the probe. Lane 1 is a control to establish levels of a single copy of the provirus. The blot was stripped and reprobed with a 1.4-kb 32P-labeled irf4 probe as loading control. B, genomic DNA was digested with XbaI (which cuts within the LTRs) to show the intact provirus. Top, MSCV-BCR/ABL-IRES-GFP single-copy proviral control (larger in size than RAS provirus). Bottom, loading control.

Close modal

Expression levels of RAS and activation of RAS downstream signaling pathways by oncogenic NRAS, KRAS, and HRAS. To gain insights into the mechanism underlying different phenotypes of the RAS isoforms, we compared the signaling pathway activation by the oncogenic RAS proteins. We infected 32Dcl3 cells, a murine myeloid progenitor cell line, with retroviral constructs expressing activated NRAS, KRAS, and HRAS and sorted for GFP-positive cells to generate nearly pure cell lines. As controls, we generated cell lines expressing GFP alone (empty vector) or the BCR/ABL oncogene. We then checked for the activation of signaling proteins Mek, Erk, Akt, PDK1, mTOR, and phospho-S6 ribosomal protein (pS6rp; Fig. 6A). We found that none of the RAS isoforms activated the mitogen-activated protein kinase pathway above basal levels (as evidenced by phosphorylation of Mek and Erk) in 32D cells, whereas BCR/ABL did. In contrast, oncogenic NRAS and HRAS showed increased phosphorylation of components of the phosphatidylinositol 3-kinase pathway (pPDK1, pAkt, pMTOR, pS6rp), whereas KRAS is a much weaker activator, if at all, of this pathway. It is possible that this differential activation of the phosphatidylinositol 3-kinase pathway underlies the phenotypic differences among the RAS isoforms.

Figure 6.

Expression levels of oncogenic RAS proteins and activation of downstream pathways. A, myeloid precursor cell line, 32Dcl3, cells transduced with vector control (MIG), NRASD12 (N), KRASD12 (K), HRASV12 (H), and BCR/ABL (p210) were starved for 12 h, lysed and separated on a 6% to 18% polyacrylamide gel. After transfer, the membrane was probed for proteins downstream of RAS, as indicated. B, membrane was also probed for levels of RAS expression, with dynamin as loading control. Ratios of oncogenic RAS to dynamin are indicated. C, liver lysates from leukemic NRAS, KRAS, and HRAS mice were separated on a 6% to 18% gradient gel and probed with anti-RAS antibody.

Figure 6.

Expression levels of oncogenic RAS proteins and activation of downstream pathways. A, myeloid precursor cell line, 32Dcl3, cells transduced with vector control (MIG), NRASD12 (N), KRASD12 (K), HRASV12 (H), and BCR/ABL (p210) were starved for 12 h, lysed and separated on a 6% to 18% polyacrylamide gel. After transfer, the membrane was probed for proteins downstream of RAS, as indicated. B, membrane was also probed for levels of RAS expression, with dynamin as loading control. Ratios of oncogenic RAS to dynamin are indicated. C, liver lysates from leukemic NRAS, KRAS, and HRAS mice were separated on a 6% to 18% gradient gel and probed with anti-RAS antibody.

Close modal

Next, we checked for levels of expression of exogenous RAS in these cell lines. We observed that in myeloid progenitor cells, the expression levels of RAS proteins are drastically different, with HRAS expression the highest and KRAS the lowest (Fig. 6B). The lower expression level of KRAS may account for, at least partially, the reduced activation of the phosphatidylinositol 3-kinase pathway in KRAS-expressing 32D cells.

We have previously shown that overexpression of oncogenic NRAS in leukemic cells may contribute to the development of AML versus CMML. We went on to test whether differential expression of oncogenic RAS also correlates with the phenotypic differences among the NRAS, KRAS, and HRAS mice. We checked levels of RAS expression in tumors from NRAS, KRAS, and HRAS mice and found that, in general, RAS proteins are expressed at higher levels in AML (HRAS and NRAS) than CMML (NRAS and KRAS; Fig. 6C).

In this study, we have compared the leukemogenic potential of oncogenic NRAS, KRAS, and HRAS in the same BMT model system. We found that although all three RAS isoforms have the potential to induce myeloid leukemias in mice, they do so at different rates and produce distinct phenotypes. Previously, we have modeled NRAS leukemogenesis using the same system and shown that oncogenic NRAS can induce an AML- or CMML-like disease in mice. Here, we show that oncogenic HRAS always induces an AML-like disease, whereas oncogenic KRAS invariably induces a CMML-like disease in mice.

Expression of oncogenic KRAS in a conditional knock-in mouse strain has been shown to exclusively induce a CMML-like disease (17, 18). The result that oncogenic KRAS also invariably induces CMML in the mouse BMT model indicates that the BMT model can be comparable with the conditional knock-in model and suggests that different oncogenic RAS proteins have distinct leukemogenic potentials. Consistent with the different biological activities of oncogenic N versus KRAS in mice, recent studies showed that NRAS and KRAS mutations are preferentially associated with distinct cytogenetic subgroups in human AML (23). The mouse BMT model system is advantageous in performing the in vivo structure-function analyses of cancer-related genes by screening a large number of mutants. Such studies help to identify critical targets in RAS leukemogenesis.

Interestingly, the KRAS protein level is lower than NRAS and HRAS in myeloid cells, although the same expression vector was used for all three RAS oncogenes and they expressed equally in NIH3T3 fibroblast cells (Fig. 1B). One possibility is that KRAS is regulated posttranslationally in hematopoietic cells. Alternatively, high expression of KRAS might activate signaling pathways that negatively regulate cell proliferation and/or survival. Further study of this unique property of KRAS may help to develop therapies for KRAS-positive malignancies.

HRAS mutations are rare in human hematologic malignancies. Recent findings show that patients with germ line mutations in HRAS are not particularly predisposed to leukemias (24). However, we found in this study that activated HRAS can be a potent oncogene in the induction of myeloid leukemia. This seeming discrepancy could be due to differences in regulation of gene expression. Retrovirus-mediated transgene expression varies in target cells depending on its site of integration in the host genome. The combination of certain integration sites and the strong retroviral LTR promoter/enhancer can result in very high expression of the transgene in target cells. Indeed, we found that HRAS leukemic cells express high levels of GFP and HRAS (Figs. 3 and 6). In contrast, HRAS expression is low in normal hematopoietic cells—the lowest among the three RAS genes (25). It is, therefore, possible that HRAS has the potential to induce myeloid malignancies when it is overexpressed, but normally its expression is tightly regulated in hematopoietic cells, making it difficult for cells with HRAS mutations to develop into tumors. Consistent with this idea, overexpression of human HRAS was found to be responsible for tumors induced by chemical carcinogens in mice (26). These observations suggest that overexpression of HRAS, in addition to its oncogenic mutation, plays an important role in tumorigenesis. However, we cannot rule out the possibility that the different oncogenic potential of HRAS in human versus mouse could be due to species-specific differences.

We have previously shown that oncogenic NRAS induced AML-like or CMML-like disease seems to correlate with the level of NRAS expression, with high expression favoring AML and lower expression correlating with CMML. We found here that this correlation extends to all RAS oncogenes (Fig. 6). This result suggests that overexpression of RAS oncoproteins also plays an important role in the pathogenesis of AML. Consistent with this idea, overexpression of wild-type and oncogenic RAS is indeed common in human AML (27, 28).

It has been shown that overexpression of oncogenic RAS in fibroblast and epithelial cells cause senescence or apoptosis (29). We show here that all oncogenic RAS proteins can efficiently induce myeloid malignancies. This highlights the fact that the outcome of RAS signaling is very dependent on cellular context. It is possible that myeloid cells contain factors that can overcome the negative signaling activated by RAS oncogenes. Consistently, c-Myc has been shown to require a second oncogenic hit to suppress apoptosis in inducing lymphoid but not myeloid leukemia (30). Identifying factors that permit RAS transformation in myeloid cells may help to develop therapies for human myeloid malignancies.

The Raf/Mek/Erk pathway was the first downstream signaling pathway of RAS to be identified and it is required for RAS transformation of fibroblast cells (31). However, we found here that activation of Mek and Erk was not detected in 32D myeloid progenitor cells expressing any one of the three RAS oncogenes (Fig. 6). This result is consistent with the finding that there is no activation of Mek and Erk in leukemic cells from the oncogenic KRAS conditional knock-in mice (17). These observations suggest that myeloid leukemogenesis by oncogenic RAS may involve unique RAS signaling networks. Identification of critical molecular events in RAS leukemogenesis is important for developing targeted therapies for myeloid malignancies involving RAS. The models established here provide a system to achieve this goal.

Note: Current address for R. Subrahmanyam: Laboratory of Cellular and Molecular Biology, National Institute on Aging, NIH, Baltimore, MD 21224.

Grant support: National Heart, Lung, and Blood Institute (HL083515, R. Ren).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Campbell SL, Khosravi-Far R, Rossman KL, Clark GJ, Der CJ. Increasing complexity of Ras signaling.
Oncogene
1998
;
17
:
1395
–413.
2
Khosravi-Far R, Campbell S, Rossman KL, Der CJ. Increasing complexity of Ras signal transduction: involvement of Rho family proteins.
Adv Cancer Res
1998
;
72
:
57
–107.
3
Hancock JF. Ras proteins: different signals from different locations.
Nat Rev Mol Cell Biol
2003
;
4
:
373
–84.
4
Rocks O, Peyker A, Bastiaens PI. Spatio-temporal segregation of Ras signals: one ship, three anchors, many harbors.
Curr Opin Cell Biol
2006
;
18
:
351
–7.
5
Voice JK, Klemke RL, Le A, Jackson JH. Four human ras homologs differ in their abilities to activate Raf-1, induce transformation, and stimulate cell motility.
J Biol Chem
1999
;
274
:
17164
–70.
6
Walsh AB, Bar-Sagi D. Differential activation of the Rac pathway by Ha-Ras and K-Ras.
J Biol Chem
2001
;
276
:
15609
–15.
7
Yan J, Roy S, Apolloni A, Lane A, Hancock JF. Ras isoforms vary in their ability to activate Raf-1 and phosphoinositide 3-kinase.
J Biol Chem
1998
;
273
:
24052
–6.
8
Maher J, Baker DA, Manning M, Dibb NJ, Roberts IA. Evidence for cell-specific differences in transformation by N-, H- and K-ras.
Oncogene
1995
;
11
:
1639
–47.
9
Esteban LM, Vicario-Abejon C, Fernandez-Salguero P, et al. Targeted genomic disruption of H-ras and N-ras, individually or in combination, reveals the dispensability of both loci for mouse growth and development.
Mol Cell Biol
2001
;
21
:
1444
–52.
10
Koera K, Nakamura K, Nakao K, et al. K-ras is essential for the development of the mouse embryo.
Oncogene
1997
;
15
:
1151
–9.
11
Bos JL. ras oncogenes in human cancer: a review.
Cancer Res
1989
;
49
:
4682
–9.
12
Hawley RG, Fong AZ, Ngan BY, Hawley TS. Hematopoietic transforming potential of activated ras in chimeric mice.
Oncogene
1995
;
11
:
1113
–23.
13
Sinn E, Muller W, Pattengale P, Tepler I, Wallace R, Leder P. Coexpression of MMTV/v-Ha-ras and MMTV/c-myc genes in transgenic mice: synergistic action of oncogenes in vivo.
Cell
1987
;
49
:
465
–75.
14
Haupt Y, Harris AW, Adams JM. Retroviral infection accelerates T lymphomagenesis in E mu-N-ras transgenic mice by activating c-myc or N-myc.
Oncogene
1992
;
7
:
981
–6.
15
Kogan SC, Lagasse E, Atwater S, et al. The PEBP2βMYH11 fusion created by Inv(16)(p13;q22) in myeloid leukemia impairs neutrophil maturation and contributes to granulocytic dysplasia.
Proc Natl Acad Sci U S A
1998
;
95
:
11863
–8.
16
MacKenzie KL, Dolnikov A, Millington M, Shounan Y, Symonds G. Mutant N-ras induces myeloproliferative disorders and apoptosis in bone marrow repopulated mice.
Blood
1999
;
93
:
2043
–56.
17
Braun BS, Tuveson DA, Kong N, et al. Somatic activation of oncogenic Kras in hematopoietic cells initiates a rapidly fatal myeloproliferative disorder.
Proc Natl Acad Sci U S A
2004
;
101
:
597
–602.
18
Chan IT, Kutok JL, Williams IR, et al. Conditional expression of oncogenic K-ras from its endogenous promoter induces a myeloproliferative disease.
J Clin Invest
2004
;
113
:
528
–38.
19
Parikh C, Subrahmanyam R, Ren R. Oncogenic NRAS rapidly and efficiently induces CMML- and AML-like diseases in mice.
Blood
2006
;
108
:
2349
–57.
20
Gross AW, Zhang X, Ren R. Bcr-Abl with an SH3 deletion retains the ability To induce a myeloproliferative disease in mice, yet c-Abl activated by an SH3 deletion induces only lymphoid malignancy.
Mol Cell Biol
1999
;
19
:
6918
–28.
21
Zhang X, Ren R. Bcr-Abl efficiently induces a myeloproliferative disease and production of excess interleukin-3 and granulocyte-macrophage colony-stimulating factor in mice: a novel model for chronic myelogenous leukemia.
Blood
1998
;
92
:
3829
–40.
22
Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action and clinical strategies.
Nat Rev
2003
;
3
:
330
–8.
23
Bowen DT, Frew ME, Hills R, et al. RAS mutation in acute myeloid leukemia is associated with distinct cytogenetic subgroups but does not influence outcome in patients younger than 60 years.
Blood
2005
;
106
:
2113
–9.
24
Aoki Y, Niihori T, Kawame H, et al. Germline mutations in HRAS proto-oncogene cause Costello syndrome.
Nat Genet
2005
;
37
:
1038
–40.
25
Shen WP, Aldrich TH, Venta-Perez G, Franza BR, Jr., Furth ME. Expression of normal and mutant ras proteins in human acute leukemia.
Oncogene
1987
;
1
:
157
–65.
26
Maruyama C, Tomisawa M, Wakana S, et al. Overexpression of human H-ras transgene is responsible for tumors induced by chemical carcinogens in mice.
Oncol Rep
2001
;
8
:
233
–7.
27
Gougopoulou DM, Kiaris H, Ergazaki M, Anagnostopoulos NI, Grigoraki V, Spandidos DA. Mutations and expression of the ras family genes in leukemias.
Stem Cells
1996
;
14
:
725
–9.
28
Stirewalt DL, Appelbaum FR, Willman CL, Zager RA, Banker DE. Mevastatin can increase toxicity in primary AMLs exposed to standard therapeutic agents, but statin efficacy is not simply associated with ras hotspot mutations or overexpression.
Leuk Res
2003
;
27
:
133
–45.
29
Cox AD, Der CJ. The dark side of Ras: regulation of apoptosis.
Oncogene
2003
;
22
:
8999
–9006.
30
Luo H, Li Q, O'Neal J, Kreisel F, Le Beau MM, Tomasson MH. c-Myc rapidly induces acute myeloid leukemia in mice without evidence of lymphoma-associated antiapoptotic mutations.
Blood
2005
;
106
:
2452
–61.
31
Shields JM, Pruitt K, McFall A, Shaub A, Der CJ. Understanding Ras: “it ain't over 'til it's over”.
Trends Cell Biol
2000
;
10
:
147
–54.