Block in Anaphase Chromosome Separation Caused by a Telomerase Template Mutation

See allHide authors and affiliations

Science  07 Mar 1997:
Vol. 275, Issue 5305, pp. 1478-1481
DOI: 10.1126/science.275.5305.1478


Telomeres are essential for chromosome stability, but their functions at specific cell-cycle stages are unknown. Telomeres are now shown to have a role in chromosome separation during mitosis. In telomeric DNA mutants of Tetrahymena thermophila, created by expression of a telomerase RNA with an altered template sequence, division of the germline nucleus was severely delayed or blocked in anaphase. The mutant chromatids failed to separate completely at the midzone, becoming stretched to up to twice their normal length. These results suggest a physical block in mutant telomere separation.

Telomeres “cap” the termini of eukaryotic chromosomes. Chromosomes lacking telomeres undergo fusion, degradation, and extremely high loss rates (13). However, there is little information on the mechanism by which telomeres ensure chromosome stability, or at what cell-cycle stage their job is performed.

In the ciliated protozoan Tetrahymena thermophila the transcriptionally active, polygenomic macronucleus divides amitotically, whereas the diploid germline micronucleus, with its chromosomal complement of five pairs of metacentric chromosomes, divides mitotically (4). The telomeric DNA tracts of the two nuclei have the same terminal GGGGTT repeat sequence, although the tracts are markedly different in overall length (5). Whereas macronuclear telomeres play a crucial role in amitotic macronuclear divisions (6), the function of micronuclear telomeres has not been examined. Micronuclear chromosomes are trancriptionally quiescent and mostly dispensable for vegetative cell divisions (7), facilitating studies of the role of telomeres in mitotic chromosome stability.

In T. thermophila, the wild-type (WT) telomerase RNA (TER1 RNA) contains the template sequence for the synthesis of GGGGTT telomeric repeats. Mutating this sequence generates corresponding mutations in the telomeric DNA (6). In this study, the template was changed to a sequence predicted to synthesize GGGGTTTT telomeric repeats, and this mutated gene, ter1-43AA, was introduced on a high-copy vector into WT T. thermophila cells (8). Four days after transformation with ter1-43AA, many cells were enlarged, and the cell population doubling rate was much slower than that of control cells transformed with WT TER1 (9). By 5 days, most ter1-43AA transformants were grossly enlarged and misshapen (Fig. 1), and their doubling rate had markedly decreased (9). The cells ceased to divide within 18 to 30 population doublings after transformation, except for rare revertants arising from loss of the introduced ter gene (9). This gradual onset of cell death, reminiscent of previously characterized telomerase RNA mutants in Tetrahymena (6), is interpreted as a cumulative effect of the cells using the altered telomerase during growth.

Fig. 1.

Cellular and nuclear phenotypes of T. thermophila ter1-43AA mutant. (A and B) Nomarski optics (same magnification; bar, 20 μm). (C to G) 4′,6′-Diamidino-2-phenylindole (DAPI) fluorescence (same magnification; bar, 10 μm). (A, C, and D) Tetrahymena thermophila cells 5 days after transformation with the WT ter1 gene. (B, E to G) Tetrahymena thermophila cells 5 days after transformation with the ter1-43AA gene. Cells were fixed in 3.7% formaldehyde in buffer A [15 mM Pipes (pH 7.0), 80 mM KCl, 20 mM NaCl, 0.5 mM EGTA, 2 mM EDTA, 0.5 mM spermidine, 0.2 mM spermine, 1 mM dithiothreitol] and stained with DAPI (0.25 μg/ml). In (C) and (E), interphase micronuclei are positioned in a pocket immediately adjacent to the macronucleus, at roughly 7 and 11 o'clock, respectively. In (D), (F), and (G), elongated micronuclei are indicative of anaphase.

Five days after transformation, the fraction of ter1-43AA transformant cells with micronuclei in mitotic anaphase (Fig. 1, F and G) was over fivefold higher than in WT transformants, indicating a severe delay or block at this stage of the cell cycle. Furthermore, the anaphase micronuclei in many highly enlarged mutant cells appeared stretched out and were up to twice as long as wild-type anaphase micronuclei (compare Fig. 1G to Fig. 1D), but with no visible separation of chromosomes at the midzone. These results suggested that the mutant telomeres prevented complete chromosome segregation. The proportions of elongated (Fig. 1F) and “hyperelongated” (Fig. 1G) micronuclei, taken as an average from quantitation of seven independent cell lines 7 days after transformation, were 26 and 25%, respectively, resulting in an average total mitotic index of 51% (10). In contrast, quantitation of WT micronuclei demonstrated an average of 7% elongated micronuclei, and none was hyperelongated. These data, coupled with an ∼16-hour population doubling time (WT transformant control doubling time was ∼4 hours) (9), indicated that the average time in anaphase for ter1-43AA micronuclei was ∼8 to 9 hours—nearly 30 times longer than the 15- to 20-min anaphase of WT micronuclei. The hyperelongation of mutant micronuclei indicated that anaphase in the mutant cells continued to progress despite the failure of chromatids to separate.

To study the structure of the ter1-43AA anaphase micronuclei, we used three-dimensional, high-resolution wide-field fluorescence microscopy (11). The ter1-43AA micronuclei were hyperelongated, and the distances between the leading edges (the centromeres) were ∼15 and ∼17 μm, respectively (Fig. 2, A and B). However, the daughter chromosome sets were still unseparated. In contrast, WT anaphase micronuclei with a comparable distance (15 μm) between leading centromeres showed a pronounced (2.7 μm) gap between the trailing free chromosome ends (Fig. 2E). Chromosome strands in the ter1-43AA hyperelongated micronuclei nearly always appeared continuous, and sometimes tautly stretched from one pole to the other, with no obvious separation between chromatids (Fig. 2A). Rarely, one free chromosome end was clearly detected, in an anaphase figure that otherwise showed an overall lack of chromosome segregation between the daughter poles (Fig. 2B, asterisk). This observation, and inspection of large numbers of mutant cell anaphase figures, indicated that separation of sister chromatids was detectable by this imaging method and was rare. Several thousand mutant anaphase micronuclei were scanned, but only one showed chromosome separation at the midzone (Fig. 2C). This midzone gap measured 2.8 μm, yet the distance between the leading daughter centromeres was 27 μm—nearly twice the distance between the centromeres in WT nuclei with comparably separated daughter chromatids. In WT anaphase micronuclei, the chromatid arms were generally not individually resolved. In contrast, in the hyperelongated mutant micronuclei, individual strands, possibly representing multiple arms bundled together, were readily visualized. Their number varied between anaphase micronuclei, perhaps reflecting chromosome missegregation in the ter1-43AA mutant.

Fig. 2.

High-resolution fluorescence microscopy of micronuclei in anaphase. (A to C) ter1-43AA micronuclei. (D to E) Wild-type micronuclei in control cells. Cells were fixed, stained with a 1:1000 dilution of OliGreen (Molecular Probes, Eugene, Oregon), and washed in buffer A. Cells were pipetted onto polylysine (1 mg/ml)-coated #1.5 cover slips, allowed to dry briefly, and layered with 20 μl of fluorescence mounting medium (Vectashield; Vector Laboratories, Burlingame, California). Subsequent control studies indicated that the drying step did not significantly affect the data qualitatively or quantitatively (27). Three-dimensional (3D) data sets were collected with an Olympus 60× 1.4 numerical aperture oil immersion lens on a computer-controlled wide-field microscopy system and cooled charge-coupled device (CCD) camera (11). Cells were imaged in 3D by moving the sample through the focal plane of the objective lens at 0.25-μm increments and recording an image with the CCD camera at each position. Out-of-focus light was removed by a constrained iterative deconvolution algorithm with an empirical point-spread function (28). Processed data were examined and manipulated by means of the IVE software package developed for 3D images (29). (A) and (B) are serial projections, where 12 sequential single-plane images 0.25 μm thick are converted into four sequential projections 0.75 μm thick. In (B), a free end is marked (*). (C) to (E) are single-plane images. Bars, 2 μm.

To investigate whether the block in micronuclear mitosis in the ter1-43AA mutant caused the gradual cell death observed, we quantitated the micronuclear phenotype and cell division capability of transformant lines. Random single cells were isolated and scored for these properties 24 hours after isolation. In all seven transformant cell lines analyzed in one experiment, although different lines varied quantitatively, the frequency of hyperelongated micronuclei correlated strongly with cell death. For example, on day 5, most C7 transformant line cells were grossly enlarged, with 50% of the visible micronuclei in mitotic anaphase, and 50% of these being hyperelongated (Fig. 3). Of the single cells isolated at this time point, 47% remained undivided 24 hours later. On day 7, 75% failed to divide, and all cells had ceased dividing by day 9. In contrast, on day 5, the C4 cellular phenotype was less severe, and only 6% failed to divide, but on day 7, large numbers of hyperelongated micronuclei had appeared, and fewer cells divided. On day 9, many C4 cell nuclei appeared degraded, and nearly all cells failed to divide (Fig. 3). Together, these data strongly suggest that the extreme lag or block in micronuclear anaphase of ter1-43AA mutants caused the block in cell division.

Fig. 3.

Micronuclear hyperelongation correlates with lack of cell division. After the establishment of transformed cell lines, the cell concentration was monitored and cells were diluted as needed into fresh medium containing paromomycin (100 μg/ml). At 5, 7, and 9 days after electroporation, (i) the phenotype of 60 micronuclei on average per cell line was assessed for elongation and hyperelongation as described (10), and (ii) individual cells were isolated to determine division capability. For the latter analysis, 35 random single cells per cell line were isolated into fresh medium and division was scored 24 hours later.

The nuclear and cellular phenotypes of ter1-43AA mutant cells allowed inference of their cellular progression (Fig. 4). In WT Tetrahymena cells, the micronucleus completes mitosis before the macronucleus undergoes its amitotic division (12), after which cytokinesis occurs (Fig. 4, A to F). Despite the inability of ter1-43AA mutant cells to complete micronuclear division, some aspects of cell cycle progression took place, as evidenced by the extreme stretching of the mutant anaphase micronucleus, indicating continuation of anaphase (Fig. 4I), visible cleavage furrow initation (Fig. 4, H and I), and grossly enlarged cytoplasm (Fig. 4J). Such cell cycle progression is predicted, because in ciliates, commitment to cell division occurs during micronuclear anaphase (13). The ter1-43AA cells that had reached a terminal phenotype (unable to divide, but still motile) were usually grossly enlarged and misshapen, with both nuclei visibly degraded (Fig. 4K). In many such cells, the micronucleus was no longer visible and the macronucleus was splayed and partially fragmented. An alternative terminal phenotype was an unusually small, misshapen cell containing an interphase micronucleus or no visible micronucleus (9). Both terminal phenotypes accounted for the lack of further increase in mitotic micronuclei 7 to 9 days after transformation (Fig. 3).

Fig. 4.

Fluorescence microscopy indicates ter1-43AA micronuclei remain blocked in anaphase as cell cycle progresses. Tetrahymena thermophila cells transformed with (A to F) the WT ter1 gene and (G to K) the ter1-43AA gene (same magnification, bar, 10 μm). Cells were stained with DAPI as described in the legend to Fig. 1 and images of different cells were compiled into an inferred progression.

These findings provide strong evidence that a mutant telomeric DNA sequence, generated by the expression of an altered telomerase RNA, causes a severe delay or block in completing mitotic anaphase. Although the sister chromatids begin to separate and are pulled apart as anaphase progresses, they are unable to segregate to the daughter poles. What is the nature of the anaphase arrest? A DNA damage cell-cycle arrest triggered by mutant telomeres (2, 3, 14) would be expected to occur in G2, before mitosis (15). In contrast, cell division of the Tetrahymena ter1-43AA mutant is blocked only late in the progression of anaphase, and many aspects of the cell cycle continue. These results, and the extremely high failure rate of chromatid separation, are consistent with a physical block in telomere separation during or before anaphase. Therefore, we propose that telomeres of sister chromatids normally are associated until metaphase, and that resolution of this association before chromosome segregation in anaphase is abrogated by the mutated telomeric DNA sequence.

How sister chromatids cohere to one another until anaphase is unknown (1618), but telomeres may be involved. Normal telomere-telomere associations, seen cytologically in a variety of organisms (19) and occurring specifically in G2 in Schizosaccharomyces pombe (20), could be mediated through single-stranded DNA tails (21) or telomere-binding factors. The latter include the TBP proteins of hypotrichous ciliates, the Rap1p or Sir proteins of yeasts, and hTRF of humans (22, 23), or the telomerase ribonucleoprotein itself. A failure of telomere separation in the ter1-43AA mutant could result from covalent fusion of the mutant telomeric DNA, or it could be mediated through aberrant association of telomere-binding factors. Altered spatial organization in the nucleus, or an altered telomere structure, could also make mutant telomeres inaccessible to factors that normally act to separate sister chromatids. For example, the yeast DNA untangler topoisomerase II is required for anaphase chromatid separation (24), as is proteolysis of noncyclin proteins such as Cut2 (18, 25). The cut2 and cut1 mutant phenotype (26) is strikingly similar to that of Tetrahymena ter-43AA: The main bodies of the sister chromosomes are pulled apart, but the telomeres remain localized late in anaphase, while the cell cycle continues (20). Further investigations into cells with altered telomere structure will help define the mechanism by which sister chromatids cohere and separate in a timely manner.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.

Stay Connected to Science

Navigate This Article