Immortality of cancers: a consequence of inherent karyotypic variations and selections for autonomy

Abstract

Immortality is a common characteristic of cancers, but its origin and purpose are still unclear. Here we advance a karyotypic theory of immortality based on the theory that carcinogenesis is a form of speciation. Accordingly, cancers are generated from normal cells by random karyotypic rearrangements and selection for cancer-specific reproductive autonomy. Since such rearrangements unbalance long-established mitosis genes, cancer karyotypes vary spontaneously but are stabilized perpetually by clonal selections for autonomy. To test this theory we have analyzed neoplastic clones, presumably immortalized by transfection with overexpressed telomerase or with SV40 tumor virus, for the predicted clonal yet flexible karyotypes. The following results were obtained: (1) All immortal tumorigenic lines from cells transfected with overexpressed telomerase had clonal and flexible karyotypes; (2) Searching for the origin of such karyotypes, we found spontaneously increasing, random aneuploidy in human fibroblasts early after transfection with overexpressed telomerase; (3) Late after transfection, new immortal tumorigenic clones with new clonal and flexible karyotypes were found; (4) Testing immortality of one clone during 848 unselected generations showed the chromosome number was stable, but the copy numbers of 36% of chromosomes drifted ± 1; (5) Independent immortal tumorigenic clones with individual, flexible karyotypes arose after individual latencies; (6) Immortal tumorigenic clones with new flexible karyotypes also arose late from cells of a telomerase-deficient mouse rendered aneuploid by SV40 virus. Because immortality and tumorigenicity: (1) correlated exactly with individual clonal but flexible karyotypes; (2) originated simultaneously with such karyotypes; and (3) arose in the absence of telomerase, we conclude that clonal and flexible karyotypes generate the immortality of cancers.

Keywords: Muller’s ratchet; aneuploidy proximate carcinogen; clonal and flexible cancer karyotypes; growth advantages of aneuploidy; karyotypes of immortal clones of telomerase-deficient mice; karyotypic linkage of immortality and tumorigenicity; long preneoplastic latency; low probability of speciation; selection for cancer-specific autonomy; sub-speciation via karyotypic drift.

Figure 1. Karyotype of the immortal tumorigenic HA1 cell line, which originated late after transfection of human kidney cells with overexpressed telomerase. The chromosomes were prepared from metaphases and labeled by hybridization with chromosome-specific colors as described in “Materials and Methods.”

Figure 2. Three-dimensional arrays of 20 karyotypes of the immortal tumorigenic human cell lines HA1 and HA1-2 and BJ based on the cytogenetic data listed in Table 1. HA1 and HA1-2 lines originated late from human kidney cells (Fig. 1), and BJ from human skin cells about 60 generations after transfection of mass cultures with artificially overexpressed telomerase and two overexpressed oncogenes (text). The karyotype arrays are three-dimensional tables, which show the chromosome numbers of 20 karyotypes on the x-axis, the copy numbers of each chromosome on the y-axis and the number of karyotypes arrayed on the z-axis. By forming parallel lines, the chromosome copy numbers of 20 karyotypes indicate that the three lines are highly clonal. A minority of non-clonal copy numbers differing from the clonal values in narrow margins of ± 1, stands out above or below these lines (Table 1). This indicates that all three lines have clonal but flexible karyotypes.

Figure 3. Cellular phenotypes of human cen3tel skin cells from a centenarian female 37 (A), 100 (B), 166 (C) and 1014 (D) generations after transfection with overexpressed telomerase. The culture had a fibroblastic cellular phenotype at PD37 (A) and a less defined fibroblastic phenotype at generation PD100 (B). At generation PD166, the culture had acquired a polymorphic cellular phenotype typical of cancer cells and simultaneously became tumorigenic and immortal (C). This new polymorphic phenotype was unchanged at PD1014 (D).

Figure 4. Aneuploidy among karyotypes of 20 human cen3tel skin cells 37 (A) and 100 (B) generations after transfection with overexpressed telomerase. The 20 karyotypes were arrayed as described for Figure 2. The data show that 60% of the cen3tel cells were randomly aneuploid 37 generations after transfection (Table 2), and that 100% were randomly aneuploid 100 generations after transfection (Table 3). Moreover, the data show that the degree of aneuploidy increased sharply from generation 37 to generation 100 after transfection. The partial clonality of trisomy 7 was native of the cen3-skin biopsy studied here (see text).

Figure 5. Karyotypic stability and drift of the tumorigenic clone cen3tel PD166 at the time of isolation (A) and after 451 (B) and 848 (C) unselected generations in vitro. As a test for immortalization, karyotype arrays prepared from the original clone cen3tel PD166 and descendents isolated 451 (PD617) and 848 (PD1014) unselected generations later were compared. These karyotype arrays show that the original cen3tel PD166 karyotype was mostly stable, but was also variable within narrow margins over 848 cell generations (Table 4). The copy numbers of 21 of 33 (64%) PD166-specific chromosomes including marker chromosomes were clone-specific during 848 generations, while 12 of 33 (36%) were generation-specific or cell-specific drifting within narrow margins of ± 1 over 848 generations (Table 4 and Fig. 6). Based on the high clonal stability over 848 unselected generations, we conclude that the cen3tel PD166 clone is immortal but flexible, exceeding the operational Hayflick threshold of 50 generations by 798 clonal generations.

Figure 11. Karyotypic stages during carcinogenesis of a virtual organ of 10⁶ cells: (1) The induction of random aneuploidy in an arbitrary 1% of normal cells by a carcinogen, such as overexpressed telomerase (Fig. 4) or SV40 virus (Figs. 9 and 10); (2) Self-catalyzed progression of aneuploidy over many cell generations up to 100% of the initiated cell population, as for example in the human skin cells 100 generations after aneuploidization by overexpressed telomerase shown in Figure 4; (3) The stochastic evolution of an autonomous cancer stem cell with a new, clonal flexible karyotype and phenotype. The new autonomous cancer cell supplants the 10⁶ non-cancerous precursors in our Petri dish within 20 generations after its origin (Figs. 5 and 8); (4) Self-catalyzed karyotypic variations of cancer cells generate new unselected subspecies, spontaneously (Figs. 5and 8) and drug-resistant or metastatic subspecies, selectively (see text).

Figure 6. Two individual karyotypes of the cen3tel clone PD166 at its origin (A) and after 848 generations of unselected growth in culture at PD1014 (B). About 64% of the chromosomes of the two karyotypes share PD166-specific copy numbers or are common markers (Table 4 and Fig. 5). The two karyotypes differ in generation-specific and cell-specific chromosomal copy numbers and markers [labeled in (A and B)] that could only be distinguished as such by the statistical analyses of the karyotype arrays shown in Figure 5.

Figure 7. (A) Karyotype of the immortal tumorigenic clone cen3telS2 PD84, which arose from an independent preneoplastic aliquot of the cen3tel mass culture 84 generations after it was transfected with overexpressed telomerase. (See above Results, “How does overexpressed telomerase induce…”) (B) Karyotype of cen3telS2 PD143, which supplanted the PD84 clone 59 generations after it was identified at PD84.

Figure 8. Karyotype arrays of the immortal, tumorigenic clones cen3telS2 PD84 (A) and a sub-clone cen3telS2 PD143 (B) based on data of Table 5. The cen3telS2 PD84 clone arose spontaneously in an independently growing preneoplastic aliquot of the cen3tel mass culture (from which also cen3tel PD166 originated) 84 generations after transfection with overexpressed telomerase. Upon further cultivation for 59 generations, the cen3telS2 PD84 clone was supplanted by a more transformed and faster growing variant, cen3telS2 PD143. Comparative karyotype arrays revealed that PD143 is a near-tetraploid, clonal variant of cene3teS2 PD84 (Table 5).

Figure 9. Karyotypes of immortal clones of SV40 virus-transformed lung cells, FL1 (A) and tail cells, FT1 (B) from a telomerase-deficient mouse (see also Table 6). The transformed clones arose together with their new individual karyotypes 2–3 mo after infection of the cells with SV40 virus.

Figure 10. Karyotype arrays of immortal, neoplastic clones of SV40 virus-transformed lung cells, FL1 (A and B) and tail cells, FT1 (C) from a telomerase-deficient mouse. The clones appeared 3 mo after infection by SV40 tumor virus together with their new clonal and flexible karyotypes. Karyotype arrays of the lung-derived focal culture, FL1 were determined 35 and 65 generations after isolation of the focal culture. The arrays show that the FL1 karyotype was near-triploid with near clonal FL1-specific chromosome numbers and chromosome copy numbers (see also Table 6). The karyotype array of the tail-derived focal culture, FT1 was determined 25 generations after isolation of the focal FT1 clone. The array shows that the FT1 karyotype was hypo-diploid with 94–100% clonal chromosome numbers and chromosome copy numbers.