Jan n&v final.indd

Miguel Godinho Ferreira
Each round of DNA replication results in the erosion of telomeres, the ends of linear chromosomes. Telomerase counteracts
telomere shortening by synthesizing new DNA sequences in a time period restricted to late S and G2 cell-cycle phases, even
though the enzyme is active throughout the cell cycle. A recent study directly implicates cyclin dependent kinase 1 (Cdk1) in the
regulation of telomere elongation during the cell cycle.
In recent years, the telomere field has been fac-ing an interesting riddle: although telomerase purified from cell extracts is active throughout the cell cycle, telomeres are only extended in late S and G2 phases. Carol Greider and col-leagues now provide evidence for the direct involvement of the cell-cycle master regulator Cdk1 (ref. 1).
Telomeres have unique properties that dis- tinguish them from damage-induced DNA double-strand breaks (DSBs). Alterations in these properties have been shown to promote genomic instability associated with cancer, as well as to ultimately limit the proliferation of both normal and cancer cells. Telomeres are composed of small tandem arrays of duplex repeats with a G-rich 3ʹ strand protruding beyond the double-stranded region to form an overhang. Both double- and single-stranded telomeric regions must be maintained, and
protective proteins bind specifically to each Figure 1 Telomere elongation follows the cycles of activation of Cdk1. Cdk1 regulation allows for two
structure. These proteins fulfill two main func-
stages in the cell cycle: in early G1 (low Cdk1 activity), telomeres, together with DSBs, are refractive tions: first, they protect chromosome-ends to 5ʹ-end degradation. Consequently, unresected ends at DSBs favour NHEJ and a refractory status for telomeres. During S, G2 and M phases (high Cdk1 activity), an MRN-dependent nuclease activity from illegitimate DNA repair, which can cause degrades 5ʹ-ends thus generating extended ssDNA both at telomeres and DSBs. These extended fusions or extensive degradation; second, they 3ʹ-overhangs allow for the engagement of telomerase at telomeres and homologous recombination- promote the complete replication of chromo- some-ends by engaging telomerase — a reverse transcriptase that uses an integral RNA compo- Similarly to other organisms, yeast telomer- The MRX complex is involved in the formation nent as a template to lengthen the G-rich strand ase activity can be reconstituted from extracts of the telomeric 3ʹ-overhang in S-phase cells, derived from different phases of the cell cycle4. thus allowing the binding of Cdc13 and the During telomere DNA replication, leading- However, in synchronized cells, telomere elon- subsequent recruitment of telomerase5,6.
strand synthesis generates a blunt-ended mol- gation is observed only in late S phase3. Frank et ecule, whereas lagging-strand synthesis leaves al. used a telomere-healing assay that relies on 3ʹ-overhangs at DSBs was recently shown to be behind a small G-rich overhang. To elongate the formation of a new telomere ‘seed’ (gener- under the control of Cdk1 (ref. 2). This finding these ends, telomerase requires an extended 3ʹ- ated in vivo by HO endonuclease) that is subse- begged the question of whether Cdk1 regula- overhang that is provided by a nuclease activ- quently elongated by telomerase1. This method tion of MRX was also involved in generating ity under the control of Mre11–Rad50–Nbs1 was previously used to show that a telomere the telomeric 3ʹ-overhang and, consequently, (MRN), a protein complex recently shown to seed resulting from a DSB could be extended telomere elongation. Using a drug-regulatable be involved in relaying the cell-cycle status to by telomerase in G2/M phases, but not in G1- Cdk1 mutant7, Frank et al. showed that tel- arrested cells. It also provided a means to test omere elongation requires Cdk1 activity, pos- Telomere elongation has been best studied the genetic requirements for telomerase activ- sibly through the MRX-dependent generation in budding yeast using two different assays3,4. ity at telomeres. Apart from requiring telom- of the 3ʹ-overhang both at telomere seeds and erase, its regulators (Est1-3, Tlc1 and Cdc13) at native telomeres1. This result is reminiscent Miguel Godinho Ferreira is in the Telomere and and DNA polymerases α and δ4, elongation of the current model for cell-cycle regulation of Genome Stability Laboratory, Instituto Gulbenkian de of this broken-end also needs the MRX com- DSB repair, in which low Cdk1 activity favours Ciência, Apartado 14, 2781-901 Oeiras, Portugal.
e-mail: [email protected] plex5 (the budding yeast version of MRN). non-homologous end joining (NHEJ) DNA NATURE CELL BIOLOGY VOLUME 9 | NUMBER 1 | JANUARY 2007
repair in G1 because of inefficient resection of response, similar to DSB repair, so that ATM origins of replication), homologous recombina- DSBs2. Conversely, high Cdk1 activity favours and Mre11 are specifically recruited to telom- tion repair takes over as the major DSB repair homologous recombination repair in the eres in G2 phase12. Thus, passage of a repli- pathway and telomeres become available for remainder of the cell cycle, due to the activation cation fork could be a prerequisite for MRN telomerase. If the extrapolation from DNA rep- of MRN-dependent resection of DSBs, making binding and processing of telomere ends. This lication holds true, we would expect that multi- them better substrates for homologous recom- hypothesis would also explain the seemingly ple mechanisms linked to Cdk1 are involved in bination repair. Thus, an attractive unifying conflicting result from the two telomere-elon- controlling the elongation of telomeres. These theme seems to be emerging (Fig. 1): telomeres gation assays3,4: in one assay, telomeres could may include cell cycle-dependent protein deg- and damage-induced DSBs are viewed by the be elongated from broken telomere seeds in radation (for example, an essential component cell in a similar manner8. When Cdk1 activity G2/M blocked cells, but this was not the case of the telomerase complex, Est1, is degraded by is low in G1, MRN is unable to promote resec- in an independent assay, where telomeres the proteosome in G1; ref. 14), activating and tion at terminal DNA ends, leading to NHEJ were shortened by removing internal telomere inhibitory phosphorylation events dependent at DSBs and exclusion of telomerase from tel- repeats by the flp recombinase. In the latter on Cdk1 and physical exclusion mechanisms omeres. Later in the cell cycle, Cdk1 becomes assay, although telomere length increases in (such as nuclear exclusion of regulatory com- active, permitting MRN activity both at DSBs S phase, they fail to elongate further in G2/ ponents). What is missing in this puzzle is the and at telomeres; this engages the homologous M-blocked cells3, suggesting that telomeres discovery of critical Cdk1 substrate(s) in tel- recombination (HR)-repair machinery and tel- would be refractory to telomerase later in the omerase at their respective substrates.
cell cycle. However, normal telomere length is However, like in every good riddle, several recovered in subsequent cell cycles, as replica- 1. Frank, C. J., Hyde, M. & Greider, C. W. Mol. Cell 24,
new questions arise. Perhaps the most pressing tion forks traverse the telomere and allow MRN et al. Nature 431, 1011–1017 (2004).
one is the role of DNA replication in the resec- 3. Marcand, S., Brevet, V., Mann, C. & Gilson, E. Cell cycle restriction of telomere elongation. Curr. Biol. 10,
tion of telomere-ends. Resection of telomere Studies from cell-cycle regulation of DNA 4. Diede, S. J. & Gottschling, D. E. Cell 99, 723–733
ends and consequent telomere lengthening replication have provided the groundwork for was found to require the passage of a replica- most of our current understanding of ‘once-per- 5. Diede, S. J. & Gottschling, D. E. Curr. Biol. 11, 1336–
tion fork3,9. One possible explanation could be cell cycle’ events. Similar to the establishment of 6. Larrivee, M., LeBel, C. & Wellinger, R. J. Genes Dev. 18,
that, although Cdk1 activation of MRN can competent origins of replication, cycles of Cdk1 et al. Nature 407, 395–401 (2000).
occur at newly formed DSBs, MRN can only activation work as a binary switch between two 8. Ferreira, M. G. & Cooper, J. P. Genes Dev. 18, 2249–
access telomeres after they have been opened periods in the cell cycle13. In one stage Cdk1 9. Dionne, I. & Wellinger, R. J. Nucleic Acids Res. 26,
by the passage of a replication fork. Consistent activity is lacking (early G1), cells establish with this hypothesis are the observations that competent origins of replication, possess higher 10. Aylon, Y., Liefshitz, B. & Kupiec, M. EMBO J. 23,
MRX-dependent resection of DSBs can occur levels of NHEJ and telomeres are refractive to 11. Vodenicharov, M. D. & Wellinger, R. J. Mol. Cell 24,
in late G1 (in a cdc7ts block) but telomeres telomerase. In the second stage, which com- 12. Verdun, R. E., Crabbe, L., Haggblom, C. & Karlseder, J. are left unresected in an early S-phase block prises late G1 until M phase, Cdk1 is active, Mol. Cell 20, 551–561 (2005).
(hydroxyurea arrest)10,11. In human cells, rep- DNA replication can be initiated (in addition 13. Diffley, J. F. Curr. Biol. 14, R778–R786 (2004).
14. Osterhage, J. L., Talley, J. M. & Friedman, K. L. Nature
lication of telomeres triggers a DNA-damage to the consequent inhibition of new competent Struct. Mol .Biol. 13, 720–728 (2006).
NATURE CELL BIOLOGY VOLUME 9 | NUMBER 1 | JANUARY 2007

Source: http://bioem.sysu.edu.cn/lecture/structure/ncb0107-22.pdf