Spontaneous transformation of murine epithelial cells requires the early acquisition of specific chromosomal aneuploidies and genomic imbalances

Abstract

Human carcinomas are defined by recurrent chromosomal aneuploidies, which result in a tissue-specific distribution of genomic imbalances. In order to develop models for these genome mutations and to determine their role in tumorigenesis, we generated 45 spontaneously transformed murine cell lines from normal epithelial cells derived from bladder, cervix, colon, kidney, lung, and mammary gland. Phenotypic changes, chromosomal aberrations, centrosome number, and telomerase activity were assayed in control uncultured cells and in three subsequent stages of transformation. Supernumerary centrosomes, binucleate cells, and tetraploidy were observed as early as 48 hr after explantation. In addition, telomerase activity increased throughout progression. Live-cell imaging revealed that failure of cytokinesis, not cell fusion, promoted genome duplication. Spectral karyotyping demonstrated that aneuploidy preceded immortalization, consisting predominantly of whole chromosome losses (4, 9, 12, 13, 16, and Y) and gains (1, 10, 15, and 19). After transformation, focal amplifications of the oncogenes Myc and Mdm2 were frequently detected. Fifty percent of the transformed lines resulted in tumors on injection into immunocompromised mice. The phenotypic and genomic alterations observed in spontaneously transformed murine epithelial cells recapitulated the aberration pattern observed during human carcinogenesis. The dominant aberration of these cell lines was the presence of specific chromosomal aneuploidies. We propose that our newly derived cancer models will be useful tools to dissect the sequential steps of genome mutations during malignant transformation, and also to identify cancer-specific genes, signaling pathways, and the role of chromosomal instability in this process.

Figure 1. Experimental scheme used for spontaneous transformation of murine epithelial cells

This figure describes the experimental design used for isolation and culture of murine epithelial cells derived from bladder, cervix, colon, kidney, lung, and mammary glands. Multiple techniques were used to monitor morphological and genomic changes that occurred throughout culture during spontaneous transformation. We identified three sequential stages: pre-immortal, immortal, and transformed based on distinct morphologies.

Figure 2. Pre-immortal, immortal and transformed stages exhibit distinct morphologies

Sequential stages designated as (A) pre-immortal, (B) immortal, and (C) transformed for bladder epithelial cells. Sequential stages designated as (D) pre-immortal, (E) immortal, and (F) transformed for cervical epithelial cells. Note the pronounced formation of foci in (C) and (F). Actin staining (green) of mammary epithelial cells designated as pre-immortal (G), immortal (H) and transformed (I). DNA was counterstained with DAPI (orange pseudo-color). Note that at the immortal stage; the actin fibers are more disorganized compared with pre-immortal. At the transformed stage, actin becomes concentrated into dense figures. In (H) we show a binucleated cell (center of image). (I) Note the enlarged nucleus (center) resulting from increased DNA content.

Figure 3. Chromosomal aberrations during spontaneous transformation

Representative karyotypes after SKY analysis of the pre-immortal stage from cells derived from the cervix (A) and colon (B). These karyotypes are all near-tetraploid (±4n) with recurrent chromosomal aneuploidies (loss of Chrs 4, 7, and 18). White arrows denote numerical aberrations. (C) Example of an anaphase bridge. DNA is stained with DAPI (blue) and cytoplasm with cytokeratin (red). (D) Additional examples of specific chromosomal aberrations, specifically dicentric chromosomes (upper and bottom panel). (E) Examples of profound chromosomal instability in immortal bladder cells in the form of acentric fragments, chromosome breakage, and tri-radial exchanges. (F) Example of metaphase spread from transformed kidney cell line analyzed by SKY containing multiple complex chromosome aberrations. (G) FISH analysis of transformed kidney cells. DAPI stain shows chromosomes in blue, chromosome 17 (red signal), and chromosome 2 (green signal). Note the presence of enormous numbers of double minute chromosomes (dmin) derived from chromosome 2. (H) FISH analysis using DNA probe for Myc (yellow) in a transformed bladder cell, with DNA counterstained with DAPI (pseudo-color red). Massive amplification of the Myc oncogene is seen in an interphase nucleus and micronuclei.

Figure 4. Chromosomal gains and losses and patterns of aneuploidy

(A) Chromosomal gains and losses based on SKY analysis of cell lines from all six organs at the transformed stage (green, gains; red, losses). X-axis: chromosomes in numerical order starting with 1; Y-axis: total percentage of cases with a given aberration. Note that some chromosomes are subject to frequent gains (e.g., chromosome 15) or losses (chromosome 4). Dashed areas indicate that the chromosome is only partially gained or lost. In addition to chromosomes that are subject to imbalances in all tissues, we observed also tissue specific imbalances. Note that some aberrations occurred in all cases of a given tissue. (B) Distribution of ploidy levels determined by SKY analysis. These circular plots indicate ploidy ranges of the different tissues for the three different stages in different colors as determined by the number of chromosomes. For instance, regarding lung, the majority of the pre-immortal cells were diploid (2n=40), or near-diploid, ±2n). This was followed by a frequent tetraploidization (or near-tetraploidization) when the cells immortalized, and became near-triploid by the sequential loss of chromosomes when transformed. In contrast, regarding the bladder, most cells became immediately near-tetraploid at the pre-immortal stage, whereas in the cervix we observed a substantial fraction of triploid (or near-triploid) cells. Each radial line indicates an individual cell for which cytogenetic analysis was conducted in all three stages.

Figure 5. Cross-species comparison of genomic imbalances in mouse (MMU) and human (HSA) chromosomes

Visualization of genomic imbalances observed in human carcinomas (our own data derived from CGH analysis colon, breast, cervical, and lung cancers) and those detected in the mouse models (mammary gland, lung, colon, bladder, kidney, and cervix). (A) Ideogram of mouse chromosome 4 (MMU4) mapped with human orthologous regions. HSA 1p and 9p are frequently lost in human carcinomas, as are the orthologous regions on MMU 4. Note that the tumor suppressor gene CDKN2A/INK4A/p16, which maps to chromosome arm 9p, is included in the region frequently lost within MMU4, also containing the orthologous gene Cdkn2a/Ink4a/p16. (B) Display of chromosome arm lost from HSA3 in human cancers homologous to MMU 9. (C) MMU 15 contains two homologous segments from human chromosomes frequently subject to gains (HSA 8q and 5p). These regions contain the Myc oncogene, and the RNA component of the telomerase complex, Tert. All respective homologous segments are highlighted in color on MMU and HSA chromosomes. Data for mouse orthologous regions was obtained from: http://useast.ensembl.org/Mus_musculus/Location/Synteny/

Figure 6. Binucleate cell analysis

(A) Bi-nucleate mammary cells stained for actin (green) to visualize cytoskeletal structure; DNA stained with DAPI (pseudo-color red); (B) Binucleated cells (arrows) in bladder culture, passage 0; (C) Images taken from live-cell imaging of a primary C57BL/6 normal lung cell undergoing the process of binucleation. Arrows identify the number of nuclei present within the cell.

Figure 7. Centrosome analysis

Examples of centrosome amplifications. Centrosome abnormalities were observed early in the transformation process (48 hr). (A), Bladder cell exhibiting centrosome duplication; (B) Binucleated mammary cell with multiple centrosomes; and (C) mitotic colon-derived cell. (Yellow = FITC-labeled anti γ-tubulin; red = DNA). (D) Graph displaying numbers of cells containing abnormal centrosome counts for 17 cell cultures and their derived cell lines at three stages of spontaneous transformation.

Figure 8. Average telomerase activity

Graphical display of the average telomerase activity levels using the TRAP-assay for normal uncultured cells from six different organs, and for spontaneously transformed cells at two stages of transformation (immortal and transformation). Tert-deficient cells served as the negative control, and murine E4 cells that over-express telomerase enzyme activity, were used for normalization.