Inactivating Icmt ameliorates K-RAS-induced myeloproliferative disease

Abstract

Hyperactive signaling through the RAS proteins is involved in the pathogenesis of many forms of cancer. The RAS proteins and many other intracellular signaling proteins are either farnesylated or geranylgeranylated at a carboxyl-terminal cysteine. That isoprenylcysteine is then carboxyl methylated by isoprenylcysteine carboxyl methyltransferase (ICMT). We previously showed that inactivation of Icmt mislocalizes the RAS proteins away from the plasma membrane and blocks RAS transformation of mouse fibroblasts, suggesting that ICMT could be a therapeutic target. However, nothing is known about the impact of inhibiting ICMT on the development of malignancies in vivo. In the current study, we tested the hypothesis that inactivation of Icmt would inhibit the development or progression of a K-RAS-induced myeloproliferative disease in mice. We found that inactivating Icmt reduced splenomegaly, the number of immature myeloid cells in peripheral blood, and tissue infiltration by myeloid cells. Moreover, in the absence of Icmt, the ability of K-RAS-expressing hematopoietic cells to form colonies in methylcellulose without exogenous growth factors was reduced dramatically. Finally, inactivating Icmt reduced lung tumor development and myeloproliferation phenotypes in a mouse model of K-RAS-induced cancer. We conclude that inactivation of Icmt ameliorates phenotypes of K-RAS-induced malignancies in vivo.

Figure 1

Inactivation of Icmt inhibits cell proliferation induced by expression of an endogenous K-RAS^G12D allele in mouse embryonic fibroblasts. (A) Proliferation of immortalized Icmt^fl/+K^LSL and Icmt^fl/flK^LSL cells infected with Cre- and β-gal-adenoviruses. Cre-adenovirus-treatment of Icmt^fl/+K^LSL and Icmt^fl/flK^LSL fibroblasts produced Icmt^Δ/+K^G12D and Icmt^Δ/ΔK^G12D cells that expressed endogenous K-RAS^G12D and that had one or both Icmt alleles inactivated, respectively. Data are mean plus or minus SEM of one cell line assayed in triplicate. Similar results were obtained in 5 independent experiments. (B) Cells from a focus formation assay taken 12 days after plating (original magnification ×10; 0.22 NA objective). (C) Proliferation of primary Icmt^fl/+K^LSL and Icmt^fl/flK^LSL cells infected with β-gal- (2 left bars) and Cre-adenoviruses (2 right bars). A total of 10⁴ cells were plated and counted after 23 days. Data are one cell line for Icmt^fl/+K^LSL and the mean of 2 cell lines for Icmt^fl/flK^LSL. (D) PCR amplification of genomic DNA from cells of a typical experiment in panel A showing the deletion of the Icmt^fl allele and the stop cassette in the promoter of the Kras2^LSL allele and the appearance of the Icmt^Δ band and the activated Kras2^G12D band in Cre-adenovirus-treated cells (lanes 2 and 4). (E) Western blots showing levels of K-RAS, total RAS, and ACTIN in total cellular extracts from cells in panel A. Similar results were obtained with a different pair of cell lines. For panels B, D, and E, the genotypes of cell lines i through iv are shown in panel A.

Figure 2

Icmt deficiency reduces the accumulation of immature myeloid cells and tissue infiltration in mice with K-RAS–induced MPD. (A) White blood cell counts of Icmt^fl/+K^LSLM and Icmt^fl/flK^LSLM mice. Blood was analyzed before and weekly after pI-pC injections. (Icmt^fl/+K^LSLM: week 0-8, n = 11-19; week 9-14, n = 4-6; Icmt^fl/flK^LSLM: week 0-7, n = 7-20; week 8-14, n = 1-5). (B) White blood cells were evaluated in blood smears from Icmt^fl/+K^LSLM (▭) and Icmt^fl/flK^LSLM mice (▬) before pI-pC injections (n = 6-7), at 5 to 7 weeks (n = 5), and at 13 to 14 weeks after pI-pC injections (n = 2-3). Shown is the percentage of immature cells (ie, myeloblasts, myelocytes, metamyelocytes, band cells, and pelgeroid cells). (C) Photographs of typical blood smears from panel B (original magnification ×100; 1.30 NA oil objective). (Top right panel) → represents band cell; ⇉, pelgeroid cell; ▲, erythroblast. The large cell to the right of the erythroblast is a myeloblast. (D) Hemoglobin concentrations measured in the blood samples shown in panel A. (E) Hematoxylin and eosin–stained sections of liver and spleen of mice killed 13 weeks after pI-pC injections (original magnification ×20; 0.50 NA objective). (F) Kaplan-Meier curve showing survival of Icmt^fl/+K^LSLM (n = 21) and Icmt^fl/flK^LSLM (n = 22) mice.

Figure 3

Inactivation of Icmt reduces splenomegaly and colony growth of splenocytes in mice with K-RAS–induced MPD. (A) Spleen weight of Icmt^fl/+K^LSLM (n = 3, 5, and 9 at the 3 indicated times, respectively) and Icmt^fl/flK^LSLM mice (n = 4, 6, and 9). Average spleen weight of wild-type mice (dashed line; n = 20) was 3 mg/g body weight. (B) Flow cytometry showing an increased percentage of CD11b/Gr-1 double-positive cells in spleens of Icmt^fl/+K^LSLM compared with Icmt^fl/flK^LSLM and wild-type mice. Shown are representative scatter plots from one mouse of each genotype and the mean percentage of double-positive cells (n = 4 for all genotypes). The increase in double-positive cells in the spleens of Icmt^fl/+K^LSLM mice was statistically significant (P = .043 vs Icmt^fl/flK^LSLM; P = .009 vs control). (C,D) Methylcellulose colony assays of splenocytes (n = 3-4) cultured in the absence (C) and presence (D) of exogenous growth factors (SCF, IL-3, IL-6, and EPO). CFU indicates colony-forming unit; GEMM, granulocyte-erythroid-macrophage-megakaryocyte; BFU-E, burst-forming unit erythroid; GM, granulocyte-macrophage; G, granulocyte; M, macrophage. (E) Colony growth of splenocytes (n = 3-4 for each genotype) seeded in methylcellulose at the indicated concentrations of GM-CSF. (F) May-Grünwald-Giemsa–stained cytospins of individual colonies from experiment in panel D (original magnification ×100; 1.30 NA oil objective).

Figure 4

Mislocalization of RAS proteins in Icmt-deficient splenocytes and analyses of downstream RAS effectors in CD11b-positive splenocytes. (A) Intracellular localization of RAS proteins in splenocytes. Total cell extracts (T) from splenocytes were fractionated into soluble (S) and membrane (P) fractions and analyzed on Western blots with a pan-RAS antibody. Note the accumulation of RAS proteins in the soluble (S) fraction in splenocytes of Icmt^fl/flK^LSLM mice. (B) Western blots showing levels of GTP-bound RAS and downstream effectors in serum-starved and GM-CSF-stimulated CD11b-positive splenocytes pooled from mice of the same genotype (control, n = 5; Icmt^fl/+K^LSLM, n = 3; Icmt^fl/flK^LSLM, n = 3).

Figure 5

Knockout of Icmt reduces growth factor-independent colony growth of bone marrow cells from mice with K-RAS–induced MPD. (A,B,D) Methylcellulose colony assays of bone marrow cells (n = 3-5 for each genotype) cultured in the absence (A) and presence (B) of growth factors (SCF, IL-3, IL-6, and EPO), and the indicated concentrations of GM-CSF (D). (C,E) Colony growth of bone marrow cells from Icmt^fl/+M (n = 3) and Icmt^fl/flM mice (n = 3) (without the Kras2^LSL allele) cultured in the presence of growth factors (C) or the indicated concentrations of GM-CSF (E). (F) BFU-E formed by bone marrow cells (n = 2 for each genotype) cultured in the indicated concentrations of EPO. (G,H) PCR amplification of genomic DNA from individual colonies. The genomic DNA used in panel G is from colonies from experiments in Figures 3D and 5B. DNA used in panel H is from experiments in panel C.

Figure 6

Inactivation of Icmt improves survival and reduces autonomous colony growth of hematopoietic cells in a second model of K-RAS–induced malignancies. (A) Survival of Icmt^fl/+K^LSLLC (n = 24), Icmt^fl/flK^LSLLC (n = 14), and control (n = 10) mice. Methylcellulose colony assay of bone marrow cells (B) and splenocytes (C) (n = 3 for each genotype) cultured in the absence of exogenous growth factors. (D) Photographs from typical experiment in panel C.