|Cancer||Cell death||Cell cycle||Cytoskeleton||Exo/endocytosis||Differentiation||Division||Organelles||Signalling||Stem cells||Trafficking|
Cell Biology International (2007) 31, 14471455 (Printed in Great Britain)
Origin of diplochromosomal polyploidy in near-senescent fibroblast cultures: Heterochromatin, telomeres and chromosomal instability (CIN)
Kirsten H. Walenab*
aViral and Rickettsial Disease Laboratory, California Department of Health Services, 850 Marina Bay Parkway, Richmond, CA 94804, USA
bCROMOS, 763 Ocean Avenue, Richmond, CA 94801, USA
The near-senescence associated phenomena of increases in cells with chromosomal damage (CIN) and in endopolyploid mitotic cells were analyzed for possible inter-relationships by cytogenetics. Gross chromosomal abnormalities in all phases of mitosis were analyzed in situ. Hetrochromatization of telomeres, centromeres and interstitial chromatin regions (i.e., chromocenters/SAHF) were shown to be specific occurrences in the near-senescent phase. Stickiness between such chromatin regions caused breakage/fragmentation by anaphase-pulls on clumped chromosomes. Gluey heterochromatin is therefore, seen as a cause of CIN in near-senescence. Detrimental effects on chromosomes from heterochromatin have been observed for decades, and can be explained from chromatin remodeling in epigenetics. A consequence of genomic damage was re-replication to polyploidy of arrest-escaped cells with G2/M-DNA content. This second synthetic period produced diplochromosomal cells that cycled by bi-polar division into genome reduced cells. This sequence from h-chromatization to CIN and further to cycling endopolyploidy is believed to be a basic mechanism for the production of genetic variability in neoplasia.
Keywords: Epigenetics, Checkpoint controls, Re-replication, G2/M cells, Mitosis, Endopolyploidy.
*Correspondence address: VRDL, California Department of Health Services, 850 Marina Bay Parkway, Richmond, CA 94804, USA. Tel.: +1 510 234 3375; fax: +1 510 234 3127.
In recent model systems of carcinogenesis the suggested cellular events involve changes of diploid to polyploid cells which can escape from cell cycle arrest, induced by polyploid check point controls (Storchova and Pellman, 2004; Margolis, 2005). Mitosis of such escaped cells is assumed to produce aneuploidy (the hallmark of tumor cells) resulting from multipolar misdivisions. This latter outcome, especially for aged cell-related cancers, is further thought to be augmented in genetic variability from unstable chromosomes with eroded telomeres (i.e., ends of chromosomes). This pathological chromosomal condition is a favorite hypothesis for the induction of chromosomal instability (CIN) in neoplasia (Hahn, 2004; Shay and Roninson, 2004). These ideas are partially supported by recent findings in early, pre-cancerous lesions, which include: presence of senescent cells, eroded telomeres, CIN and polyploidy (Barrett et al., 2003; Collado et al., 2005). This latter change to polyploidy can occur in many ways, as for instance, through cell fusion, acytokinesis or endoreplication with multiple DNA synthetic (S) periods and no mitosis (Edgar and Orr-Weaver, 2001). Cells from this latter route to polyploidy, when entering mitosis, present with so-called diplochromosomes (i.e., pairs of sister chromosomes) (D'Amato, 1989), which was observed in presenescence of in vitro diploid cells (Walen, 2006, 2007). Since these cells propagated with a new type of cycling from, for example, 4C-2N to 8C-4N and back to 4C-2N, the cellular pathway to the origin of these unusual cells is explored in the present study.
From previous studies, it is well known that immediately before cells become arrested in G1 (senescence) there are increasing frequencies of both polyploid cells and cells with spontaneous chromosomal aberrations including: dicentrics, exchange figures, breaks, fragments and micronuclei (Hayflick and Moorhead, 1961; Saksela and Moorhead, 1963; Miles, 1964; Thompson and Holliday, 1975; Sherwood et al., 1988; Walen, 2005, 2006). The unexplained issue in these studies is the sudden and simultaneous frequency-peaks of these two different classes of cells. The propagating polyploid cells with diplochromosomes became a major subpopulation just before senescence (Walen, 2006). For this report, the important feature of this type of mitotic endocycling is that these cells evaded polyploid-check point controls (Andreassen et al., 1996; Nigg, 2001; Margolis, 2005). This was not observed for polyploid cells that were present in progressive cell growth to senescence. Such cells were present at a constant low incidence level (Saksela and Moorhead, 1963). It was, therefore, assumed that diplochromosomal polyploidy originated from a process that was special for the near-senescent phase.
One route to such chromosomal complexes is through arrest of genome-damaged cells by abortive cell cycling, which for mitotic cells is a premature reversion to interphase cells (also known as mitotic catastrophe [Roninson et al., 2001; Erenpreisa et al., 2005]) (Nagl, 1978; Hartwell and Kastan, 1994; Elledge, 1996; Storchova and Pellman, 2004). Arrested cells with G2/M-DNA content (bichromatid chromosomes) can escape from arrest and re-replicate, and become endopolyploid cells with diplochromosomes. Since cells with spontaneous chromosomal abnormalities also peaked in frequency in the near-senescent stage, it was hypothesized that these latter cells with genomic-damage gave rise to endopolyploidy through re-replication. This sequence to polyploidy is the most common process for higher animals (Nagl, 1978).
However, at the base of this latter suggestion is the question of what causes the sudden high rate of spontaneous chromosomal damage in pre-senescence? As mentioned, telomere attrition (Harley, 1991) occurs in progressive cell expansions to senescence and is considered to be one cause of CIN (Ducray et al., 1999; Gisselsson et al., 2001; Desmaze et al., 2003; Feldser et al., 2003; Shay and Roninson, 2004). However, the precise mechanism(s) of this process to CIN from eroded telomeres is not known. In fact, the origin/cause of CIN is a major question in tumor biology (Rajagopalan et al., 2003; Sharpless and DePinho, 2004). Moreover, in tumor progression of colon cancer it is suggested that the cause of CIN acts as “a heritable alteration that affects genetic instability” (Nowak et al., 2002). One such mechanism suggested here is that CIN is linked to chemical restructuring of normal euchromatin (to h-chromatin) or to changes in constitutive h-chromatin which can be inherited as epigenetic changes (Goldberg et al., 2007). Such chromatin changes have long been known to be expressed as condensed nuclear chromatin bodies that are permanent alterations, and seemingly do not affect cell survival or ability to enter mitosis (e.g., condensed X chromosomes) (Brown, 1966; Bernstein et al., 2007). This suggestion of h-chromatin as an underlying source of CIN is based on the following information. (1) Recent studies on h-chromatization of telomeres (Chan and Blackburn, 2002; Owen-Hughes and Bruno, 2004) and of interstitial chromosomal regions, which were referred to as “senescence associated heterochromatic foci, SAHF” (Narita et al., 2003; Collado et al., 2005). In addition, inherited epigenetic changes (h-chromatic regions) have recently been proposed as “progenitors” of neoplasia (Feinberg et al., 2006; Jones and Baylin, 2007). (2) There is a ubiquitous presence in cancer cells of h-chromatic, condensed nuclear chromatin bodies (i.e., chromocenters) and of abnormal mitoses with misalignment on metaphase plates of clumped chromosomes (De May, 1996). (3) “Historical” observations have been made of chromosomal detrimental effects from changes of eu- to h-chromatic chromosomal regions (Brown, 1966; Swanson, 1968; Nagl, 1978; Therman and Susman, 1993).
These facts and suggestions lead to the question of how h-chromatin can affect chromosomal stability. The historical works showed that h-chromatic chromosomal regions could be sticky and fuse together: in interphase cells to condensed, chromocenters and in mitosis to clumps of chromosomes. Chromosomal clumping will have damaging effects on chromosomal segregations in mitosis (Swanson, 1968; Steinbeck, 1998). Furthermore, a recent article uses the phrase “Slow and sticky chromosomes” in regard to asymmetric, anaphase chromosomal segregations (Hardy and Zacharias, 2005). Consequently, the present study on CIN in normal, diploid cells (intact p53, Rb, p16, p21 etc. functions) is a demonstration of effects on mitosis from sticky h-chromatin, and it addresses the following specific issues of: (1) endopolyploidization of cells with genomic damage from the process of CIN, and (2) earlier works with demonstrations of differential genomic responses to h-chromatization of chromosomal regions.
2 Materials and methods
2.1 Cell culture, fixation and staining
The data presented here were derived from the same human, diploid, fibroblast cell strain from female, fetal lung tissue (strain L645) that was used for the preceding papers on senescence and on endocycling of diplochromosomes (Walen, 2005, 2006, 2007). In an on-going quality control program this cell strain is routinely monitored at different passage levels by karyology and for Mycoplasma infection. The cells were cultured in Eagle's medium (MEM), supplemented with 10% fetal bovine serum, 100
Briefly, cells were detached by trypsin/versene treatment and split 1 to 3 at weekly, intervals until cell growth became noticeably reduced at passage numbers 31 to 33 (p31–p33). At this point of slow growth, one quarter volume of old spent medium from the flask to be transferred was added to fresh culture medium. (Experience had shown better cell-attachment for old cells from such a mixture than from fresh medium only.) The split ratio was also changed to 1 to 2 for a few additional passages (2–4) before an almost abrupt disappearance of mitotic cells (p37–p40). Passage numbers from 30 to 40 with present split ratio translate into 45–60 population doublings, which is the growth period generally considered to be the near-senescence phase, and was in earlier studies on fibroblast cells called phase III (Hayflick and Moorhead, 1961). The attainment of this pre-senescent phase is generally achieved by a forced, almost synchronized, cell population-expansion which is not expected to occur in vivo. However, it has been shown to be a useful model system (see Section
During this growth period just before total cell senescence, the cells were grown on chamber slides and coverslips and were variously harvested: 3 and 5 days for the slides and 7, 10 and 12 days for coverslips after seeding. The number of cells/clumps (2–10 cells) that was seeded varied with number of days for growth. For the slide cultures it was as previously described (Walen, 2007). The coverslips, with longer growth-periods, received only about 2000 cells/clumps each. A mild in situ hypotonic treatment was done before fixation in 3 parts absolute methyl alcohol and 1 part acetic acid (3:1). All cell preparations were Giemsa (R66)-stained: (2% in pH 6.8, sodium and potassium phosphate buffer). The illustrations are from a Zeiss compound microscope with an attached Zeiss Icon camera.
2.2 Hypotonic treatment and consequences
The present cell preparations were either untreated with hypotonic solution (1 to 1 of 0.4% KCl and 0.4% Na-citrate) or were given a short exposure of only 4 instead of the usual 8
2.3 Assessment of the level of polyploidy
The genome content of interphase nuclei of approximately 2N, 4N and 8N was judged by two criteria (Walen, 2006): (1) nuclear size in relationship to the number of Barr bodies (h-chromatic X chromosomes), and (2) size differences in nuclear volumes. The first method was used in the original work on cell senescence by Hayflick and Moorhead (1961). This was based on the fact that Barr bodies are associated with the nuclear membrane, always being situated in the same general area of the upper quadrant at the apex of the cell. In the presence of multiple chromocenters, this method becomes invalid, but it proved to be a useful ploidy- quantification system in their absence (Hayflick and Moorhead, 1961). The second method used nuclear size of early diploid prophases with intact nuclear membranes (2N, 4C; C
There is a temporal relationship in the appearance of cells with chromosomal damage and an increase in endomitotic polyploid cells in the near senescent phase. In different studies of diploid cell strains, the average incidences of these abnormalities were &007E;20% and 15%, respectively, but under the classification of hyperploidy, the percentages were much higher at 62% and 92% (Harley, 1991). In a more recent study of the same cell strain used in the present study, the incidences of polyploidy and of telophase bridges were 37% and 10% respectively, whereas diploid mitosis was reduced from an average of 3% to 0.7% in passage (p) 34 (&007E;53 population doublings, PDs; see; Walen, 2006, for detailed quantifications). This heterogeneous cell population also contained cells with one giant nucleus and large senescent flat cells with more than one diploid-sized nucleus. Another noticeable change in this near senescent phase was presence of two types of cells with spotted condensed nuclear chromatin, viz. small dots (pepper-corns) which often were paired and larger bodies known as chromocenters (Fig. 1J,M; SAHF, see Section
Gross chromosomal abnormalities in mitoses and heterochromatization of interphase nuclei. (A, B) Chromosomal clumping in metaphase. (C, D, E, G, K) Anaphase segregations of clumped chromosomes. (H) Small bridges with h-chromatic centromeres and telomeres. (F) Genome fragmentation. (L) Stickiness between telomeres and centromeres. (I,J, M) Interphase nuclei with pairs of h-chromatic telomeres and chromocenters (note; both daughter nuclei). Microscopic magnification: A,B,D,F,G–M: ×500; C,E: ×50.
(A–D) Cells with interphase nuclei surrounded by micronuclei from reverted mitotic division figures. (E, F) Polyploid nuclei with 2 and 4 Barr bodies. (G) A metaphase with diplochromosomes—pairs of sister chromosomes. (H) Early passage metaphase (p30) with two dicentrics and one not fully condensed chromosome. (I, J) A normal fibroblastic cell area (p32) as compared to polyploid cell growth (p34) with larger, rounder, and more condensed nuclei. Arrows point to clumped chromosomes in polyploid (left) and diploid (right) cells, and at lower edge (J) there is an abnormal anaphase segregation-figure of clumped chromosomes. Microscopic magnification: A,B,D: ×200; C,E–H: ×500; I,J: ×125.
3.1 Chromosomally damaged mitoses
In the present replicative progression to phase III with a 1:3 split ratio of the cultures, the peaking frequencies of gross chromosomal damage and polyploidy occurred in p33–35 (i.e. &007E;50 to 54 PDs). From previous studies in the evaluation of polyploidy and telophase bridges (Walen, 2006), it became apparent that there were increasing incidences of abnormal clustering and misalignments of the chromosomes in metaphase plates and abnormal divisions in anaphases (Fig. 1A–E). Another abnormality was presence of irregularly shaped nuclei either as single nuclei or pairs. These interphase cells most often contained micronuclei that showed chromosomal structures (Fig. 2A–D). Morphologically abnormal cells were from aborted mitotic divisions (i.e. premature reversion of mitotic to interphase cells). The occurrences of these abnormalities together with presence of normal mitotic cells are presented for 7-day harvests of coverslip cultures for p33 to 36 (Table 1). By p36, the number of mitosis at day 7 was too low for a statistically valid total cell number which was therefore obtained by combining the numbers in growth at 7, 10, and 12 days after seeding. The overall information associated with increasing passages is reduction in normal and increases in abnormal mitoses (i.e. chromosomally and aborted mitotic cells). The data only show a trend for the various cellular alterations, since such incidences vary in different experiments (Harley, 1991). The rise in reversion to interphase cells of meta-, ana- and telo-phase cells clearly demonstrates cell cycle arrest of mitotic cells with genomic damage.
Percent chromosomal abnormalities and aborted mitoses in seven-day growths on coverslips at different passage levels
In earlier passages (<p33), there were occasional cells with 2–4 clustered chromosomes, whereas at p35 and 36 most of the chromosomes were either in one or several clustered groups. Ana- to telophase figures (Fig. 1C–F,G,K) which showed clumps of chromatin, bridges and breakage/fragmentation of the chromosomal material, were especially prominent in p34. These particular abnormalities for p34, with the exception of telophase bridges, were present in &007E;half the total of 38% di- and polyploid mitotic cells. An arrow points to breakage in centromeric region that has produced an arm in Fig. 1K, and several small fragments either as pairs or single dots are also present. In non-chromosomally clustered cells (from p31), the more conventional types of aberrations were present, including dicentric and not completely condensed chromosomes (Fig. 2H, arrows).
Simultaneous with the increase in chromosomally abnormal and aborted mitoses (Table 1), polyploid cells accumulated (Walen, 2006). In endomitosis, they showed diplochromosomes, and in endoreplicated interphase nuclei there was an increase in Barr bodies (Fig. 2E–G). The two interphase nuclei shown here have 2 and at least 4 such condensed chromatin bodies (arrows), corresponding to tetra- and octo-ploid cells (see Section
Condensed dot-like nuclear chromatin was often closely associated in pairs which indicate h-chromatic telomeres (Figs. 1M and 2E). Similar dot-pairs of condensed chromatin can also be seen in a cell with premature chromatid separations that have formed small bridges (Fig. 1H, arrows). These bridges show a repeated pattern of darkly staining dots (knobs), one each on the 2 pole-oriented centromere regions and 2 dots (pairs) on telomere associations in the middle of the bridge (i.e. a pattern of 1-2-1 dots). Similar dots on chromosomal ends were also present in other abnormal anaphase figures (Fig. 1G). Alignments of chromosomal ends with one another led to the formation of small chains of chromosomes, which were only observed in gently, hypotonically treated metaphases of early passages (Fig. 1L, arrows, p31). In the insert of this figure, alignments between ends and centromeric regions are evident; a presumed #20 chromosome is bent in the h-chromatic centric area and is in close contact with ends of two G-group chromosomes. The bending that stretches the h-chromatic centric area was observed earlier for a B-group chromosome (Walen, 2006, figure 4r).
In addition to cells with condensed dot-like pairs of chromatin, cells with larger condensed chromocenter bodies were present, especially in passages that neared the total cell senescent phase. The quantification of these cells, however, proved elusive because their numbers varied in different and the same passages, and no pattern was evident. Their presence in dying cultures (debris and rounded cells) was most pronounced, as previously reported (Hayflick and Moorhead, 1961). However, an association with cell death of this type of h-chromatization may not be absolute (Narita et al., 2003) since both daughter cells contained such chromocenters (Fig. 1J,M, see Section
The relevance of data from cell cultures to in vivo happenings is often questioned. Therefore, it is noteworthy that in studies on early pre-neoplastic changes, there has been a steadily increasing acknowledgement of polyploidy as an intermediate stage in neoplasia (Bibbo et al., 1989; Shackney et al., 1995; Margolis, 2005). In addition, 2 specific disease conditions—ulcerative colitis and Barrett's esophagus (Galipeau et al., 1996; Rabinowitch et al., 1999; O'Sullivan et al., 2002; Barrett et al., 2003)—have contributed data suggestive of a multi-step sequence of events very similar to the present conclusions of cellular changes, involved in potential neoplastic, mutational changes. Briefly, the beginning cellular alterations in these diseases were found to be associated with CIN and, from observations of shortened telomeres, their induction of CIN was suggested. CIN preceded the appearance of 4N (G2-tetraploid) cells. These polyploid cells, however, did not show mitotic activity in spite of the presence in them of gene-profiles necessary for such entry (Barrett et al., 2003). The cytogenetic data discussed below add new information on the origin and process of CIN, and on polyploidization that can lead to mitotic endocyclings of genomes, with the result of bichromatid chromosomes in G1.
4.1 Heterochromatin, chromosomal damage and polyploidization
Chromatin division into eu- and h-chromatin was originally based on cytological appearances (Heitz, 1933)—even staining of nuclear euchromatin in contrast to darkly, condensed chromatin bodies for h-chromatin (e.g. Barr body; Fig. 2E,F). Telomeres have been found to contain condensed chromatin, and a pattern of paired dots versus random dots in interphase nuclei was discussed (Brown, 1966; Suja and Rufas, 1994; Dernburgh et al., 1995). The cell age-related h-chromatization of interstitial chromosomal regions, SAHF (Narita et al., 2003), are from morphology and time of appearance most likely the same nuclear bodies as the well known chromocenters (Heitz, 1933), which also are a phase III phenomenon (Fig. 1I,J,M) (Hayflick and Moorhead, 1961; Miles, 1964). However, from the presence of chromocenters in both daughter nuclei (Fig. 1J,M) an absolute association of SAHF with a cell death program does not seem likely (Narita et al., 2003). A more likely explanation would be a quantitative effect on cell survival from such chromatin remodeling (Gregory and Shiekhattar, 2004). Moreover, chromocenters can be decondensed (the dispersion phase) in mitotically active cells (Heitz, 1933; Bernstein et al., 2007). Another phase III happening was appearance of cells with h-chromatic telomeres and centromeres in both interphase and mitotic cells. Stickiness between these h-chromatic regions was expressed in various chromosomal configurations such as chains, bridges (Fig. 1H,L) and dicentric chromosomes (Fig. 2H, arrows) with end-to-end telomere associations. The dicentrics were earlier linked to the occurrence of ana-telophase bridges, which were found to be significantly increased in the late passages before senescence (Benn, 1976; Harley, 1991; Saltman et al., 1993; Ducray et al., 1999; Walen, 2006). Another type of adhesion of h-chromatic regions was expressed by chromosomal clumping in both meta- and ana-phase figures (Fig. 1A–E,G,K). The essence of all of these observations in regard to CIN is that at least 3 occurrences of h-chromatization (telomeres, centromeres and SAHF/chromocenters) that causes chromosomal aberrations from stickiness are specific happenings for the near senescent phase.
The clumping of metaphase chromosomes together with the presence of interphase-reverted mitotic cells with micronuclei (i.e. aborted mitoses; Figs. 1A–D, 2A–D; in Table 1, “aborted meta-, ana- and telo-phases”) constitutes strong evidences for operative cell cycle arrest mechanisms in this near-senescent cell population (Storchova and Pellman, 2004). (The decrease in anaphase chromosomal figures and the increases in aborted ana- and telo-phase nuclei for p35 and p36 were unexpected (Table 1). Chromosomal sorting in metaphase is a possible explanation given by Ogle et al. (2005). Furthermore, it is apparent from descriptions of nuclei in earlier studies on phase III cells that abnormal interphase nuclei from abortive mitotic cycling were also observed in these studies (Hayflick and Moorhead, 1961; Miles, 1964). As mentioned earlier, re-replication is a common mechanism among cell cycle arrested cells with G2/M-DNA (two chromatids)for the creation of endopolyploidy. Therefore, it is concluded that the origin of cells with diplochromosomes (i.e. requiring 2 consecutive periods of DNA synthesis and no mitosis) in near-senescence is a result of re-replication of arrest-escaped, genome damaged cells. This means that for normal, diploid cells, CIN precedes the possible development of aneuploidy (Rajagopalan et al., 2003).
Re-replication of the escaped G2/M cells into endopolyploidy is usually an end to continued cell replication because of arrests from checkpoint controls for polyploidy (Margolis, 2005). However, the present cells with diplochromosomes (Fig. 2G) evaded arrest and divided by bipolar whole genome segregations into cycling cells with bichromatid chromosomes in telophase (4C for 2N cells) that condensed into interphase nuclei (Walen, 2007). Technically, this division of a tetraploid cell with 8C-DNA content (92 chromosomes in pairs) reduces the genome content to a diploid state with 4C-DNA content (46 chromosomes). This latter level of 4C-DNA has been previously documented in a fraction of similar senescent cells by both microphotometry and flow cytometry methods (Grove and Mitchell, 1974; Yanishevsky et al., 1974; Sherwood et al., 1988). But, the cells were interpreted to be single chromatid, tetraploid (4C) cells in G1 in both instances. A similar situation of 4C cells with regenerative potentials constituted a small fraction of the total of 2C cells in normal epithelial skin tissue (Gelfant, 1966). These cells, however, were from cell growth experiments described as diploid cells, “resting” with a G2 (i.e. bichromatid chromosomes) chromosomal content, as discussed here for diplochromosomes. This observation on the epidermis is intriguing from the viewpoint of stem cell biology in neoplasia (Feinberg et al., 2006). The question, however, is what methodology would reveal potential genomic endocycling (e.g., 2N–4C to 4N–8C and back to 2N–4C) in normal “cell turnover” tissues? So far, the reality of such cycling rests on cytogenetic demonstrations (Grell, 1946; Walen, 2007).
4.2 Heterochromatin and chromosomal instability (CIN) in near-senescence
The second attribute of sticky chromosomes is their behavior in anaphase, in which the cells showed bridges, and severe breakage and fragmentations of the chromosomal materials (Fig. 1C–E,G,K). Presence of dot-fragments and chromosomal arms in anaphase (Fig. 1K) showed breakage in h-chromatic telomere and centromere regions. The implication of chromosomal breakage from anaphase-pulls on clumped/associated h-chromatic regions was first demonstrated for radiation induced chromosomal stickiness (Swanson, 1968). More recently for genome unstable tumor cell lines it was found that prevention of anaphase movements resulted in a significant reduction of cells with genomic damage (Gisselsson et al., 2001).
Currently the chromatin changes to heterochromatin are seen in numerous studies on the chemistry of chromatin remodeling (Gregory and Shiekhattar, 2004; Owen-Hughes and Bruno, 2004; Bernstein et al., 2007). Both histone/protein and folding changes are implicated in different types of h-chromatic changes. These can vary in chromatin lengths and in number of regions affected. In this remodeling, it is concluded that the chemically changed chromatin can act as “glue” in promoting “clustering of heterochromatic regions” (Grewal and Moazed, 2003). There is now, notably, a chemical reason for the several decade-long observations of stickiness between h-chromatic regions.
4.3 General effects of heterochromatin on chromosomal instability—CIN
These effects can be summarized as follows: (1) segregational abnormalities in mitosis and arrest in the cell cycle; (2) high breakability compared to euchromatin (e.g. chromosome 9 centric inversions; Mondello et al., 2000); (3) chromosomal replication that is out-of-step (i.e. allocycly, Fig. 2H, arrowhead); (4) such regions can be under-replicated (Ionov et al., 1993; Edgar and Orr-Weaver, 2001; Desmaze et al., 2003); and (5) a resulting gene silencing as in the variegated position effect. This latter phenomenon was recently demonstrated for a newly formed telomere in human cells (Baur et al., 2001). But perhaps the uniqueness of hetero- as compared to eu-chromatin is best expressed in the classical study on the “Dissociation-Activator” system in corn (Zea mays) (McClintock, 1984). In this system, small h-chromatic elements could leave their genomic places and insert themselves in new positions (i.e. transposons), which resulted in activation or silencing of nearby genes. The part relevant to the occurrence of CIN in pre-senescence is the genome-destabilizing system becomes active after trauma (stress) due to repeated healings of broken chromosomal ends. In analogy to this Ds-Ac system, the erosion of the telomeres in growth towards senescence is also a result from repeated breakage and healing processes (Gire et al., 2004; Zou et al., 2004). Transposons, as heterochromatic elements and with involvements in gene-silencing, have recently been described for mammalian cells (Bernstein et al., 2007). Therefore, dysfunctional telomere-activated instability of transposons could be an explanation for total genome fragmentations (Fig. 1F).
On the other hand, there is another less well known feature of h-chromatin that may involve another type of CIN (Marx, 2002; Feldser et al., 2003). This type is best known from advanced tumors in which high frequencies of inter- and intra-chromosomal exchanges occur. The remarkable fact is that most of the break-points for the exchanges occurred in h-chromatic regions (Gisselsson et al., 2001). In Drosophila, added h-chromatin increased the frequencies of somatic crossing over more than several hundred-fold and the breakpoints were in h-chromatic regions (Brown et al., 1962; Walen, 1964). In human cells without somatic pairing (in contrast to Drosophila), h-chromatic regions can get together (i.e. synapsis) by so-called ectopic pairing (Nagl, 1978; John, 1988), which may facilitate exchange between different chromosomes. The presence of h-chromatin in malignant tumor cells has often been found to display characteristic patterns from the nuclear locations of the chromocenters (Habers and Sandritter, 1968; Therman and Susman, 1993). Clearly, the listed multitude and variety of genomic responses (CIN) which can be caused by h-chromatization are impressive, and are seen as quite pertinent to recent discussions on the origin of CIN and its effects/behavior in tumor biology (Rajagopalan et al., 2003; Sharpless and DePinho, 2004).
This report presents new and previous findings of the near-senescent phase of normal, diploid cells in context of operative cell cycle checkpoint controls of genome damaged cells. It shows that gain of polyploidy is composed of many steps that are inter-dependent. It also opens the door for future correlations between types of chromatin remodeling (i.e. heterochromatization) and their effects on CIN.
The well defined phase III immediately before total cell senescence was in the present study and earlier found to be a dynamic stage for several cellular changes that can be arranged in an inter-dependent sequence of events as follows: (1) senescence associated h-chromatization which induces CIN, (2) cell cycle arrest of cells with genomic damage from CIN, (3) escape from arrest of cells with G2/M-DNA content, (4) skip of mitosis and re-replication into diplochromosomal polyploidy, and (5) a bipolar endocycling of these special polyploid cells. Perfect endoploidy-divisions from, for example, 4N–8C to 2N–4C are not expected to occur either in vitro or in vivo. CIN caused by permanent h-chromatization, would continue to operate and give rise to misdivisions with the possible result of aneuploidy with chromosomal abnormalities.
Applying this sequence of events to the non-replicating 4N cells in vivo (e.g. in Barrett's esophagus mentioned above), there is a possibility that the 4N (G2-tertraploid) cells were really G2/M/4C cycling cells “resting” in G1. Flow-cytometry and micro-photometry of interphase cells cannot distinguish between G2/M (4C) diploid and G1 tetraploid (4C) cells. Considering these in vivo possibilities together with earlier and present in vitro observations, the outlined sequence of cellular events (1 to 5 above) becomes a very likely process in the creation of genetic variability for neoplasia.
I would like to thank Drs. David Schnurr and Chief Carol Glaser of the California Department of Health Services, Viral and Rickettsial Disease Laboratory, for laboratory space and use of equipment during this investigation. The diploid human lung cells were kindly made available for this study by Dr. Schnurr. To Dr. F.L. Schuster of this laboratory, I am very grateful for critique with helpful suggestions and editing of the manuscript. I also wish to thank Mr. Chao Pan of this laboratory, for computer assistance.
Andreassen PR, Martineau, SN, Margolis, RL. Chemical induction of mitotic checkpoint override in mammalian cells results in aneuploidy following a transient tetraploid state. Mutat Res 1996:372:181-94
Barrett MT, Pritchard, D, Palanca-Wessel, C, Anderson, J, Reid, BJ. Molecular phenotype of spontaneously arising 4N (G2-tetraploid) intermediates of neoplastic progression in Barrett's esophagus. Cancer Res 2003:63:4211-7
D'Amato F. Polyploidy in cell differentiation. Caryologia 1989:42:183-211
De May RM. . The art and science of cytopathology 1996:vol. I:
Dernburgh AF, Sedat, JW, Cande, WZ, Bass, HW. Cytology of telomeres. Telomeres 1995:
Ducray C, Pommier, J-P, Martins, L, Boussin, FD, Sabatier, L. Telomere dynamics, end-to-end fusions and telomerase activation during the human fibroblast immortalization process. Oncogene 1999:18:4211-23
Galipeau PC, Cowan, DS, Sanchez, CA, Barrett, MT, Emond, MJ, Levine, DS. 17p (p53) allelic losses, 4N (G2/tetraploid) populations, and progression to aneuploidy in Barrett's esophagus. Proc Natl Acad Sci U S A 1996:93:7081-4
Gelfant S. Patterns of cell division: the demonstrations of discrete cell populations. Methods in cell physiology. II 1966:359-95
Gisselsson D, Jonson, T, Petersen, A, Strombeck, B, Cin, PD, Hoglund, M. Telomere dysfunction triggers extensive DNA fragmentation and evolution of complex chromosome abnormalities in human malignant tumors. Proc Natl Acad Sci U S A 2001:98:12683-8
Habers E, Sandritter, W. Gesteigerte Heterochromatizierung als pathologisches Prinzip. Deutsch Med Wochenschr 1968:93:269-71
Ionov Y, Peinado, MA, Malkhosyan, S, Shibata, D, Perucho, M. Ubiquitous somatic mutations in simple repeated sequences revealed a new mechanism for colonic carcinogenesis. Nature 1993:363:558-61
John B. The biology of heterochromatin. Heterochromatin: molecular and structural aspects 1988:1-147
Narita M, Nunez, S, Heard, E, Narita, M, Lin, AW, Hearn, SA. Rb-mediated heterochromatin formation and silencing of E2F target genes during cellular senescence. Cell 2003:113:703-16
Nowak MA, Komarova, NL, Sengupta, A, Jallepalli, PV, Shih, le-M, Vogelstein, B. The role of chromosomal instability in tumor initiation. Proc Natl Acad Sci U S A 2002:99:16226-31
O'Sullivan JN, Bronner, MP, Brentnall, TA, Finley, JC, Shen, W-T, Emerson, S. Chromosomal instability in ulcerative colitis is related to telomere shortening. Nat Genet 2002:32:280-4
Rabinowitch PS, Dziadon, S, Bronner, MP, Emond, MJ, Crispin, DA, Haggitt, RC. Pancolonic chromosomal instability precedes dysplasia and cancer in ulcerative colitis. Cancer Res 1999:59:5148-53
Walen KH. Human diploid fibroblast cells in senescence; cycling through polyploidy to mitotic cells. In Vitro Cell Dev Biol Anim 2006:42:216-24
Yanishevsky R, Mendelsohn, ML, Mayall, BH, Cristofalo, VJ. Proliferative capacity and DNA content of aging human diploid cells in culture: a cytophotometric and autoradiographic analysis. J Cell Physiol 1974:84:165-70
Received 13 February 2007/24 April 2007; accepted 27 June 2007doi:10.1016/j.cellbi.2007.06.015