Disruption of the ARF transcriptional activator DMP1 facilitates cell immortalization, Ras transformation, and tumorigenesis

Abstract

The DMP1 transcription factor induces the ARF tumor suppressor gene in mouse fibroblasts, leading to cell cycle arrest in a p53-dependent manner. We disrupted sequences encoding the DNA-binding domain of DMP1 in mouse embryonic stem cells and derived animals lacking the functional protein. DMP1-null animals are small at birth, and males develop more slowly than their wild-type littermates. Some adult animals exhibit seizures and/or obstuctive uropathy, each of unknown cause. The growth of explanted DMP1-null mouse embryo fibroblasts (MEFs) is progressively retarded as cells are passaged in culture on defined transfer protocols; but, unlike the behavior of normal cells, p19(ARF), Mdm2, and p53 levels remain relatively low and DMP1-null MEFs do not senesce. Whereas the establishment of cell lines from MEFs is usually always accompanied by either p53 or ARF loss of function, continuously passaged DMP1-null cells readily give rise to established 3T3 and 3T9 cell lines that retain wild-type ARF and functional p53 genes. Early-passage DMP1-null cells, like MEFs from either ARF-null or p53-null mice, can be morphologically transformed by oncogenic Ha-Ras (Val-12) alone. Splenic lymphocytes harvested from both DMP1-null and ARF-null mice exhibit enhanced proliferative responses in long-term cultures when stimulated to divide with antibody to CD3 and interleukin-2. Although only 1 of 40 DMP1-null animals spontaneously developed a tumor in the first year of life, neonatal treatment with dimethylbenzanthracene or ionizing radiation induced tumors of various histologic types that were not observed in similarly treated DMP1(+/+) animals. Karyotypic analyses of MEFs and lymphomas from DMP1-null animals revealed pseudodiploid chromosome numbers, consistent with the retention of wild-type p53. Together, these data suggest that ARF function is compromised, but not eliminated, in animals lacking functional DMP1.

Figure 1

Targeted disruption of DMP1 in mice and expression of DMP1 protein in tissues. (A) Schematic representations of a portion of the murine DMP1 locus (top) and the DMP1 targeting vector (middle). Solid boxes denote exons with 5′ ends to the left. AvrII sites important for analysis of deletions are indicated. When genomic DNA is digested with AvrII and hybridized with the HindIII–AvrII probe, the wild-type locus yields a 5-kb band and the targeted locus (bottom) produces a 12-kb fragment. (B) Southern blot analysis of tail DNAs from F₂ animals. The sizes of AvrII fragments are indicated. (C) Tissue distribution of DMP1 protein in an 8-week-old mouse. A lysate from wild-type MEFs infected with full-length DMP1 virus was used as a positive control. The sources of tissues are indicated at the top; arrowheads at right indicate the positions of differentially phosphorylated DMP1 isoforms (Hirai et al. 1996).

Figure 2

Growth of neonatal DMP1^−/−, DMP1^+/−, and DMP1^+/+ mice. The body weights of littermates were measured every 2–3 days. DMP1^−/− mice were 20%–30% smaller than their control littermates at birth. Male DMP1^−/− mice remained smaller even as adults; female knockout mice eventually became indistinguishable from their DMP1^+/+ or DMP1^+/− littermates.

Figure 3

Expression of DMP1, p19^ARF, and p16^INK4a proteins in MEFs of different genotypes. Genotyped MEFs, as indicated at the top of each panel, were passaged following a 3T9 protocol, and cell pellets were collected at each passage. Proteins were extracted and sonicated in buffer containing 0.5% NP-40 and protease inhibitors, and DMP1, p19^ARF, p16^INK4a, and actin levels were determined by direct immunoblotting with antibodies to the indicated proteins. Representative results from two different clones with each genotype are shown. Early-passage levels from 2 to 7 are indicated at the bottom.

Figure 4

Growth kinetics and morphology of MEFs. (A) Cell proliferation on a 3T3 protocol. At 3-day intervals, the total number of cells per 60-mm-diam. culture dish were counted prior to redilution to 3 × 10⁵ per dish for the next passage. Data were plotted from four to six embryos of each genotype (see Table 1). Error bars indicate s.d. from the mean. (♦) DMP1^−/− cells; (█) DMP1^+/+ cells. (B) Cells from DMP1^−/− (♦) and DMP1^+/+ (█) MEF strains at passages 5, 10, and 15 were seeded at 1 × 10⁵ cells per culture in 60-mm-diam. dishes. Duplicate cultures were harvested at daily intervals, and the total number of cells per culture were determined. Data from four to six different strains of the same genotype were pooled. Error bars indicate s.d. from the mean. (C) Photomicrographs of MEFs at passage 18. Wild-type (clone 49) MEFs were senescent; DMP1^−/− MEFs (clone 14) were not.

Figure 5

MEF cell lines established from DMP1^−/− embryos retain wild-type p53. (A) Genotyped MEFs (as indicated at the bottom) were passaged on a 3T3 protocol. p19^ARF, p53, and Mdm2 expression were determined from cell lysates prepared at passages 5, 20, and 30 (indicated at the top) using actin as a control for protein loading. (B) Immunoprecipitation of metabolically labeled p53 proteins with isoform-specific antibodies. Lysates from eight MEF clones (top) harvested 5–10 passages postestablishment after metabolic labeling with [³⁵S]methionine were immunoprecipitated with specific monoclonal antibodies to wild-type (Ab 246) or mutant (Ab 240) p53 protein. Established clones from DMP1^+/+ or DMP1^+/− embryos (genotypes indicated at bottom) generally produced both mutant and wild-type p53 proteins; five of six randomly chosen clones from DMP1^−/− embryos synthesized only wild-type p53. DMP1^−/− clone 28 contained a minor population of cells synthesizing mutant p53. (C) Induction of p53 and Mdm2 in irradiated MEF cell lines. Established MEF cell lines (genotypes indicated at bottom) were irradiated (10Gy) and cultured for the times (hr) indicated at the top. Cell lines from DMP1^−/− MEFs (clones 1, 14, and 22) showed rapid accumulation of both p53 and Mdm2 following exposure to ionizing radiation; a DMP1^+/− line (clone 43) containing high levels of mutant p53 showed a minimal Mdm2 response.

Figure 6

Early-passage DMP1^−/− MEFs are transformed by oncogenic Ha-Ras. (A) Focus-forming assay. Genotyped MEFs (upper left) were infected with retrovirus expressing Ha-Ras (Val-12). Infected cells were selected with puromycin for 48 hr. A total of 1 × 10⁴ infected cells were mixed with 3 × 10⁵ uninfected cells and seeded onto 100-mm-diam. plates. Fourteen days later, transformed foci were stained with Giemsa and photographed. (B) Morphology of transformed foci. Genotypes are indicated in inserts at top. Cells in the upper left panel were infected with vector; all other cells were infected with retrovirus expressing Ha-Ras (Val-12). (C) Expression of p19^ARF, p53, p16^INK4a, and Ha-Ras proteins in virus-infected cells. Genotyped passage-3 MEFs were infected with empty vector or virus encoding Ha-Ras (Val-12), and infected cells were harvested 2, 3, and 4 days postinfection (numbers at top). Lysates were analyzed by direct immunoblotting with specific antibodies to the indicated proteins.

Figure 7

Growth of stimulated splenic T-lymphocytes in culture. (A) Freshly isolated DMP1^−/− (█) and DMP1^+/+ (○) splenic lymphocytes were stimulated with 0.25 μg/ml soluble anti-CD3 and 100 U/ml human recombinant IL-2. (B) After 3 weeks of primary stimulation, the cells were restimulated with 0.25 μg/ml soluble anti-CD3 and 100 U/ml human recombinant IL-2. Cell numbers were determined at the indicated intervals.