Nras^G12D/+ promotes leukemogenesis by aberrantly regulating hematopoietic stem cell functions

Hematopoietic stem cells (HSCs) undergo self-renewing divisions to sustain life-long hematopoiesis. HSC self-renewal is regulated by networks of proto-oncogenes and tumor suppressor genes.¹  During leukemogenesis, genetically altered HSCs play an important role in initiating and/or maintaining leukemia phenotypes.^2,3  These mutant HSCs often gain a competitive advantage over wild-type HSCs through increased cell division. In most cases, increased HSC proliferation tends to be associated with reduced HSC self-renewal and subsequently HSC depletion.⁴  In contrast, increased HSC cycling is associated with normal or increased self-renewal in HSCs deficient for p53^5,6  or p18INK4c.^7,8  Thus, it remains unclear how the proliferation and self-renewal of HSCs are differentially regulated during leukemogenesis.

Ras proteins are master signaling switches that couple extracellular stimuli to the intracellular response machinery. Oncogenic RAS mutations are prevalent in essentially all human cancers.⁹  In human leukemias, although both NRAS and KRAS are mutated at significant frequencies, oncogenic NRAS mutations are significantly more prevalent than oncogenic KRAS mutations.¹⁰  In mice, expression of one copy of oncogenic Nras (Nras^G12D/+) from its endogenous locus in hematopoietic cells leads to expansion of myeloid progenitors, increased long-term reconstitution of bone marrow (BM) cells, and a chronic myeloproliferative neoplasm (MPN) after a prolonged latency.^11-13  However, acute expression of Nras^G12D/+ in a pure C57BL/6 background does not induce hyperactivated granulocyte macrophage colony-stimulating factor signaling or increased proliferation in myeloid progenitors.¹³  In contrast to Nras^G12D/+, stronger activation of oncogenic Ras signaling, such as Kras^G12D/+ and two copies of oncogenic Nras (Nras^G12D/G12D), leads to hyperactivated cytokine signaling, hyperproliferation, and expansion of myeloid progenitors as well as an acute MPN.^2,3,13,14  It is thus unclear how Nras^G12D/+ induces expansion of myeloid progenitors and promotes MPNs. We hypothesized that Nras^G12D/+ promotes leukemogenesis by aberrantly regulating HSC function.

Here, we show that endogenous Nras^G12D/+ induced proliferation and increased self-renewal and myeloid differentiation bias in HSCs. ERK1/2 is constitutively hyperactivated in Nras^G12D/+ HSCs, and downregulation of MEK/ERK signaling attenuates Nras^G12D/+ HSC phenotypes.

Cre expression was induced through intraperitoneal injection of 250 μg (Sigma-Aldrich) or 7.5 μg/g body weight (GE Healthcare) polyinosinic-polycytidylic acid (pI-pC) every other day for two times. All experiments were conducted with the ethical approval of the International Association for Assessment and Accreditation of Laboratory Animal Care at the University of Wisconsin-Madison. Additional materials and methods are provided in the supplemental Data.

We previously reported that more than 90% of recipient mice transplanted with Nras^G12D/+ BM cells develop a chronic MPN, closely resembling human chronic myelomonocytic leukemia (CMML).¹¹  To determine whether Nras^G12D/+ HSCs are required for the initiation and/or maintenance of chronic MPNs, we transplanted different populations of cells from primary Nras^G12D/+ mice or recipients that had developed chronic MPNs. Nras^G12D/+ HSCs (Lin⁻ CD41⁻ CD48⁻ c-Kit⁺ Sca-1⁺ CD150⁺)¹⁵ , but not more committed myeloid progenitors, efficiently initiated CMML-like phenotypes in primary recipients, and only CMML cells containing HSCs could re-establish the disease in secondary recipients (supplemental Table 1). These data indicate that Nras^G12D/+ HSCs are required to initiate and maintain CMML-like phenotypes in recipient mice and thus serve as MPN-initiating cells. Similar observations are reported in human chronic lymphocytic leukemia¹⁶  and in a mouse model of BCR-ABL–induced chronic myelogenous leukemia,¹⁷  suggesting a common role of genetically altered HSCs in the chronic phase of leukemias. Noticeably, CMML-like phenotypes that developed in the secondary recipients required a long latency similar to that observed in the primary recipients, suggesting that just as in Kras^G12D/+-initiated MPNs,²  additional mutations promoting CMML development did not occur in Nras^G12D/+ HSCs. With the second hits, we believe that myeloid progenitors gain short-term self-renewal capability but are not fully transformed to CMML-initiating cells, which explains why these cells fail to maintain disease phenotypes in secondary recipient mice.

We further investigated how endogenous oncogenic Nras signaling regulates HSC function to promote leukemogenesis. We first examined the HSC compartment in Mx1-Cre; Nras^G12D/+ mice (described in Wang et al¹¹ ) at different time points after pI-pC injections. The conditional Nras^G12D allele was recombined in almost 100% of HSCs in these mice, although the recombination frequency was 0% to 35% in Mx1-Cre; Nras^G12D/+ mice without pI-pC treatment (supplemental Figure 1). The day of the first pI-pC injection is referred as day 1. On days 5 and 12 (Figure 1A),¹⁸  compared with control HSCs, Nras^G12D/+ HSCs displayed a moderate but consistent expansion (∼25% increase). We did not observe pI-pC–induced Sca1 expression in total BM cells or defined HSCs (supplemental Figure 2). The numbers of multipotent progenitors (MPPs) and LSK cells in G12D/+ mice were also significantly elevated compared with controls (Figure 1B-C).

Expression of endogenous Nras^G12D/+ increased the division of HSCs, but not whole BM (WBM) cells. On day 12, consistent with a previous report,¹⁹  only ∼0.5% of control HSCs were in S/G₂/M. In contrast, a significant increase of Nras^G12D/+ HSCs was detected in S/G₂/M, suggesting an increased proliferation in these HSCs (Figure 1D and supplemental Figure 3A). This result was further confirmed by using 5-ethynyl-2′-deoxyuridine (EdU) in vivo incorporation (Figure 1E and supplemental Figure 3B). We noticed that the increased proliferation in Nras^G12D/+ HSCs is moderate compared with the strong hyperproliferation of Kras^G12D/+ HSCs.²⁰ 

To determine whether the proliferation of Nras^G12D/+ HSCs was correlated with increased HSC self-renewal, we transplanted highly purified control or Nras^G12D/+ HSCs along with wild-type WBM cells into lethally irradiated mice (Figure 1F). Compared with control cells, Nras^G12D/+-derived blood cells were greatly overrepresented in peripheral blood of recipients in all lineages (supplemental Figure 4A). We euthanized some of the recipients 16 weeks posttransplant. Control donor HSCs represented ∼25% of the recipients’ BM HSC compartment, and Nras^G12D/+ HSCs comprised >90% (Figure 1F). On the basis of our quantitative analysis, a single Nras^G12D/+ HSC generated ∼10-fold as many HSCs as a single control HSC. Secondary transplantation further supported the increased self-renewal potential in Nras^G12D/+ HSCs (Figure 1G and supplemental Figure 4B). These results indicate that Nras^G12D/+ signaling enhances HSC self-renewal capability. The increased self-renewal potential in Nras^G12D/+ HSCs is in sharp contrast to reduced self-renewal and depletion in HSCs expressing stronger oncogenic Ras, such as Kras^G12D/+ or Nras^G12D/+G12D21  (G.K., Y.-I.C., and Jing Z., unpublished data), suggesting a distinct regulation of HSC function in an oncogenic Ras activity-dependent manner.

To investigate the signaling mechanism(s) underlying aberrant Nras^G12D/+ HSC function, we studied the MEK/ERK and AKT signaling in HSCs and their downstream MPPs²²  (Figure 2A) by using a recently developed HSC phospho-flow method.²³  In the absence of serum and cytokines, ERK1/2 was constitutively hyperactivated in Nras^G12D/+ HSCs (twofold over control HSCs) but not in MPPs (Figure 2A). The constitutive hyperactivation of ERK1/2 was inhibited by AZD6244, a potent MEK inhibitor (supplemental Figure 5). In contrast, AKT activation in Nras^G12D/+ HSCs and MPPs was comparable to that in controls (Figure 2A). Upon stimulation with interleukin-3, ERK1/2 was further hyperactivated in Nras^G12D/+ HSCs but not in MPPs (Figure 2B). The hyperactivation of ERK1/2 in Nras^G12D/+ HSCs is in sharp contrast to the normal pERK1/2 level in Nras^G12D/+ MPPs and Lin⁻ cKit⁺ progenitor cells,¹³  indicating that HSC is the most sensitive cell type to oncogenic Nras mutation.

To better understand the molecular mechanisms underlying oncogenic Nras signaling-induced aberrant HSC function, we performed a microarray analysis (accession number: GSE46948) using highly purified HSCs. Compared with control HSCs, Nras^G12D/+ HSCs did not show significant differential expression in genes or pathways that are associated with HSC self-renewal⁴  (supplemental Figure 6A-B).

Because Nras^G12D/+ HSCs support leukemia development predominantly in myeloid cells (supplemental Table 1 and Wang et al¹¹ ), we sought to determine whether these HSCs are associated with a myeloid differentiation bias. In support of this idea, GSEA of our HSC microarray results identified a gene signature associated with myeloid differentiation in Nras^G12D/+ HSCs (supplemental Figure 6C). Consistent with the GSEA results, the common myeloid progenitor (CMP) compartment was greatly expanded in Nras^G12D/+ mice (Figure 2E, Li et al,¹²  and Wang et al¹³ ), although common lymphoid progenitor (CLP) compartment was comparable to that in control mice (supplemental Figure 6D). Our current result of myeloid differentiation bias in Nras^G12D/+ HSCs explains the selective expansion of myeloid progenitor compartment in Nras^G12D/+ mice, in which granulocyte macrophage colony-stimulating factor signaling and cell proliferation are indistinguishable from those in control cells.^11,13 

To test whether hyperactivation of MEK/ERK signaling contributes to the effects of Nras^G12D/+ on HSCs, we treated control and Nras^G12D/+ mice with vehicle or AZD6244,²⁴  an efficacious and specific MEK inhibitor (Figure 2C). AZD6244 prevented the expansion of Nras^G12D/+ HSC and CMP compartment (Figure 2D-E) and restored slightly enlarged spleen to a normal size in Nras^G12D/+ mice (supplemental Figure 7). In addition, we generated Nras^G12D/+; Erk2^+/− mice to downregulate the MEK/ERK signaling using a genetic approach. Loss of one copy of Erk2 partially downregulated ERK signing (supplemental Figure 8) and significantly decreased the cycling of Nras^G12D/+ HSCs without affecting the cycling of WBM cells (Figure 2F). The slightly enlarged spleen remained a similar size (supplemental Figure 9).

In summary, our data suggest that moderate activation of oncogenic Ras/MEK/ERK signaling contributes to the increased long-term competitiveness in mutant HSCs during leukemogenesis.

The authors thank Dr Kevin Haigis and Dr Tyler E. Jacks for generously providing conditional oncogenic Nras mice. The authors are grateful to Drs Emery Bresnick and Qiang Chang for helpful discussion and critical comments on the manuscript and are appreciative of Dr Ya-Huei Kuo for sharing her EdU labeling protocol. The authors thank the University of Wisconsin Carbone Comprehensive Cancer Center for use of its Shared Services to complete this research.

This work was supported by the National Cancer Institute (NCI) (1R01CA152108), a Shaw Scientist Award from the Greater Milwaukee Foundation, an ASH Scholar Award from the American Society of Hematology (ASH), a V Scholar Award from the V Foundation for Cancer Research, and an Investigator Initiated Grant from the University of Wisconsin Carbone Comprehensive Cancer Center (J.Z.), a grant from the National Natural Science Foundation of China (31271457; J.W.), and a grant from the Canadian Institutes of Health Research (MOP-119327; S.M.). S.M. holds the Canada Research Chair in Cellular Signaling. This work was also supported in part by a grant from the NCI/National Institutes of Health Comprehensive Cancer Center (P30 CA014520-UW).

Contribution: J.W. and G.K. provided experimental design and execution; Y.L., J.D., Y.-I.C., S.R.T., A.D., and Jingfang Z. provided experimental execution; X.Z. performed microarray analysis; E.A.R. provided histopathologic analysis and edited the manuscript; M.K.S.-E.-L. and S.M. generated the Erk2 conditional knockout allele; and Jing Z. provided experimental design and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Jinyong Wang, Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, 190 Kaiyuan Rd, Guangzhou Science Park, Guangzhou, China 510530; e-mail: wang_jinyong@gibh.ac.cn; and Jing Zhang, McArdle Laboratory for Cancer Research, 1400 University Ave, University of Wisconsin, Madison, WI; e-mail: zhang@oncology.wisc.edu.