**2. Roles of proteins involved in DNA end-processing in telomere maintenance**

DNA DSBs are repaired by HR or non-homologous end-joining (NHEJ) [1]. *S. cerevisiae* MRX is involved in both HR and NHEJ [2]. MRX is suggested to be involved in the processng of DSB ends in HR repair. Recently, several other proteins involved the processing have been discovered. Some of the these proteins are also involved in the processing of telomere ends. In this section, the roles of these proteins in telomere maintenance will be discussed.

Roles of DNA Repair Proteins in Telomere Maintenance 599

telomere assay mimic wild-type telomere end. In wild-type *S. cerevisiae* cells, 3' singlestranded overhangs increase in S phase at telomeres (Wellinger, Wolf, and Zakian 1993) (Dionne and Wellinger 1996). In contrast, 3' single-stranded overhangs can be detected at telomeres throughout the cell cycle in the absence of *S. cerevisiae* Ku, suggesting that Ku inhibits resection at telomere (Gravel et al. 1998) (Polotnianka, Li, and Lustig 1998). This function of Ku is conserved in *S. pombe* Ku (Kibe et al. 2003). However, proteins involved in the resection of telomere ends are not well studied in *S. pombe*. In *S. pombe*, Dna2 is involved

Fig. 1. Model for DNA end-processing at *S. cerevisiae* telomere. DNA replication will create blunt-end at leading-strand telomere and 3' single-stranded overhangs at lagging-strand telomere after removal of the last RNA primer. Similar to the case at DSB ends, MRX and Sae2 play a major role to produce 3' single-stranded overhangs at telomeres. Sgs1-Dna2 and

*S. pombe* Taz1 binds telomeric double-stranded DNA (Cooper, Watanabe, and Nurse 1998). Deletion of *taz1* causes massive telomere elongation. Asynchronous wild-type *S. pombe* cells have small amount of 3' single-stranded overhangs (Kibe et al. 2003). In contrast, *taz1* disruptant has very long 3' single-stranded overhangs (Tomita et al. 2003). In this mutant background, roles of MRN, Ku, Dna2, and Exo1 are studied (Fig. 2). MRN and Dna2 are responsible for the production of 3' single-stranded overhangs (Tomita et al. 2004). But, 3' single-stranded overhangs are produced by concomitant deletion of Ku and MRN, suggesting that unknown nuclease can produce the overhangs in the absence of both MRN and Ku in *taz1* disruptant. Exo1 is not involved in this activity. Telomere ends in *taz1* disruptant is partially unprotected. Indeed, RPA foci and Rad22Rad52 foci are produced at telomere in *taz1* disruptant (Carneiro et al. 2010). Therefore, proteins involved in the resection in *taz1* disruptant may not be same as that in wild-type cells. However, Dna2 is

the proteins involved in the resection in *taz1* disruptant are also involved in the resection in

Δ

Δ **cells** 

background, suggesting that some of

Exo1 can provide compensatory activities to produce 3' single-stranded overhangs.

**2.3 Proteins involved in DNA end-processing in** *S. pombe taz1*

involved in the resection in both wild-type and *taz1*

wild-type cells.

in the resection of telomere ends (Tomita et al. 2004).

#### **2.1 Roles of proteins involved in DNA end-processing at DSB ends**

Role of *S. cerevisiae* MRX in HR is well studied both in vivo and in vitro (Mimitou and Symington 2009) (Mimitou and Symington 2008) (Zhu et al. 2008) (Gravel et al. 2008) (Cejka et al. 2010) (Niu et al. 2010). MRX cooperates with Sae2 to initiate 5' resection at DNA DSB end. Although both MRX and Sae2 have nuclease activities, it remains unclear the contribution of these nucleases to DSB resection. The resultant 3' single-stranded ovehangs are further resected by two redundant pathways. One is dependent on Sgs1 helicase, a conserved RecQ family member, and the Dna2. Dna2 has both helicase and nuclease domains, but nuclease activity is enough for DSB resection (Zhu et al. 2008). The other is dependent on Exo1 5'-3' exonuclease. *S. cerevisiae* Yku70-Yku80 heterodimer (Ku) binds to DSB ends and recruits downstream NHEJ factors (Critchlow and Jackson 1998). Ku inhibits 5' resection by MRX (Mimitou and Symington 2010) (Shim et al. 2010). Similar model is proposed in *S. pombe*(Tomita et al. 2003)*. S. pombe* Mre11-Rad50-Nbs1 (S. cerevisiae Xrs2 homologue) complex (MRN) is also suggested to be involved in 5' resection at DNA DSB end. *S. pombe* Ku also inhibits 5' resection by MRN. In the absence of MRN, Exo1 can resect DSB ends. Contribution of *S. pombe* RecQ helicase Rqh1 and Dna2 in the resection of DSB ends remains unclear. It has been shown that human BLM, a RecQ helicase family, and DNA2 interact to resect DNA end and helicase activity of BLM and nuclease activity of DNA2 are required for this reaction (Nimonkar et al. 2011). The functional conservation of these proteins from yeast to human suggests that the functions of these proteins in *S. pombe* are also conserved.

#### **2.2 Roles of proteins involved in DNA end-processing in telomere maintenance in**  *S. pombe* **and in** *S. cerevisiae*

Telomere ends should not be recognized as DSB ends, because telomere ends should no be repaired by HR or NHEJ. However, proteins involved in HR or NHEJ are also involved in telomere maintenance (Longhese et al. 2010). The chromosome end replicated by laggingstrand synthesis has 3' single-stranded overhangs. In contrast, the chromosome end replicated by leading-strand synthesis is blunt-end. However, most eukaryotes have 3' single-stranded overhangs at both ends, suggesting that the chromosome end replicated by leading-strand synthesis is resected (Wellinger et al. 1996; Makarov, Hirose, and Langmore 1997). *S. cerevisiae* MRX is suggested to be involved in this resection (Diede and Gottschling 2001). However, MRX independent resection has been suggested, which may be produced at lagging-strand telomere after DNA replication without any nuclease activity (Larrivee, LeBel, and Wellinger 2004). MRX mainly binds to the leading-strand telomere, further suggesting that MRX is involved in this resection at leading-strand telomere (Faure et al. 2010). An inducible short telomere assay revealed that artificial telomere ends is resected by the same DNA repair factors (Bonetti et al. 2009) (Longhese et al. 2010) (Iglesias and Lingner 2009) (Fig. 1). MRX and Sae2 act in the same resection pathway. Concomitant inactivation of Sae2 and Sgs1 abolishes end resection, suggesting that they have redundant function for the resection. Dna2 acts redundantly with Exo1, but not with Sgs1, suggesting that Dna2 supports Sgs1 activity. The lack of Sgs1, Dna2 or Exo1 by itself does not affect the resection, suggesting that Exo1 and Sgs1-Dna2 may less important for the resection than MRX and Sae2. These results were obtained by using artificial telomere, which initially produces blunt-end telomere by nuclease. However, leading-strand synthesis in wild-type cells also produces blunt-end telomere. Consistently, Sae2 and Sgs1 also play redundant functions in natural telomere end-processing (Bonetti et al. 2009), suggesting that an inducible short

Role of *S. cerevisiae* MRX in HR is well studied both in vivo and in vitro (Mimitou and Symington 2009) (Mimitou and Symington 2008) (Zhu et al. 2008) (Gravel et al. 2008) (Cejka et al. 2010) (Niu et al. 2010). MRX cooperates with Sae2 to initiate 5' resection at DNA DSB end. Although both MRX and Sae2 have nuclease activities, it remains unclear the contribution of these nucleases to DSB resection. The resultant 3' single-stranded ovehangs are further resected by two redundant pathways. One is dependent on Sgs1 helicase, a conserved RecQ family member, and the Dna2. Dna2 has both helicase and nuclease domains, but nuclease activity is enough for DSB resection (Zhu et al. 2008). The other is dependent on Exo1 5'-3' exonuclease. *S. cerevisiae* Yku70-Yku80 heterodimer (Ku) binds to DSB ends and recruits downstream NHEJ factors (Critchlow and Jackson 1998). Ku inhibits 5' resection by MRX (Mimitou and Symington 2010) (Shim et al. 2010). Similar model is proposed in *S. pombe*(Tomita et al. 2003)*. S. pombe* Mre11-Rad50-Nbs1 (S. cerevisiae Xrs2 homologue) complex (MRN) is also suggested to be involved in 5' resection at DNA DSB end. *S. pombe* Ku also inhibits 5' resection by MRN. In the absence of MRN, Exo1 can resect DSB ends. Contribution of *S. pombe* RecQ helicase Rqh1 and Dna2 in the resection of DSB ends remains unclear. It has been shown that human BLM, a RecQ helicase family, and DNA2 interact to resect DNA end and helicase activity of BLM and nuclease activity of DNA2 are required for this reaction (Nimonkar et al. 2011). The functional conservation of these proteins from yeast to human suggests that the functions of these proteins in *S. pombe*

**2.2 Roles of proteins involved in DNA end-processing in telomere maintenance in** 

Telomere ends should not be recognized as DSB ends, because telomere ends should no be repaired by HR or NHEJ. However, proteins involved in HR or NHEJ are also involved in telomere maintenance (Longhese et al. 2010). The chromosome end replicated by laggingstrand synthesis has 3' single-stranded overhangs. In contrast, the chromosome end replicated by leading-strand synthesis is blunt-end. However, most eukaryotes have 3' single-stranded overhangs at both ends, suggesting that the chromosome end replicated by leading-strand synthesis is resected (Wellinger et al. 1996; Makarov, Hirose, and Langmore 1997). *S. cerevisiae* MRX is suggested to be involved in this resection (Diede and Gottschling 2001). However, MRX independent resection has been suggested, which may be produced at lagging-strand telomere after DNA replication without any nuclease activity (Larrivee, LeBel, and Wellinger 2004). MRX mainly binds to the leading-strand telomere, further suggesting that MRX is involved in this resection at leading-strand telomere (Faure et al. 2010). An inducible short telomere assay revealed that artificial telomere ends is resected by the same DNA repair factors (Bonetti et al. 2009) (Longhese et al. 2010) (Iglesias and Lingner 2009) (Fig. 1). MRX and Sae2 act in the same resection pathway. Concomitant inactivation of Sae2 and Sgs1 abolishes end resection, suggesting that they have redundant function for the resection. Dna2 acts redundantly with Exo1, but not with Sgs1, suggesting that Dna2 supports Sgs1 activity. The lack of Sgs1, Dna2 or Exo1 by itself does not affect the resection, suggesting that Exo1 and Sgs1-Dna2 may less important for the resection than MRX and Sae2. These results were obtained by using artificial telomere, which initially produces blunt-end telomere by nuclease. However, leading-strand synthesis in wild-type cells also produces blunt-end telomere. Consistently, Sae2 and Sgs1 also play redundant functions in natural telomere end-processing (Bonetti et al. 2009), suggesting that an inducible short

**2.1 Roles of proteins involved in DNA end-processing at DSB ends** 

are also conserved.

*S. pombe* **and in** *S. cerevisiae*

telomere assay mimic wild-type telomere end. In wild-type *S. cerevisiae* cells, 3' singlestranded overhangs increase in S phase at telomeres (Wellinger, Wolf, and Zakian 1993) (Dionne and Wellinger 1996). In contrast, 3' single-stranded overhangs can be detected at telomeres throughout the cell cycle in the absence of *S. cerevisiae* Ku, suggesting that Ku inhibits resection at telomere (Gravel et al. 1998) (Polotnianka, Li, and Lustig 1998). This function of Ku is conserved in *S. pombe* Ku (Kibe et al. 2003). However, proteins involved in the resection of telomere ends are not well studied in *S. pombe*. In *S. pombe*, Dna2 is involved in the resection of telomere ends (Tomita et al. 2004).

Fig. 1. Model for DNA end-processing at *S. cerevisiae* telomere. DNA replication will create blunt-end at leading-strand telomere and 3' single-stranded overhangs at lagging-strand telomere after removal of the last RNA primer. Similar to the case at DSB ends, MRX and Sae2 play a major role to produce 3' single-stranded overhangs at telomeres. Sgs1-Dna2 and Exo1 can provide compensatory activities to produce 3' single-stranded overhangs.

#### **2.3 Proteins involved in DNA end-processing in** *S. pombe taz1*Δ **cells**

*S. pombe* Taz1 binds telomeric double-stranded DNA (Cooper, Watanabe, and Nurse 1998). Deletion of *taz1* causes massive telomere elongation. Asynchronous wild-type *S. pombe* cells have small amount of 3' single-stranded overhangs (Kibe et al. 2003). In contrast, *taz1* disruptant has very long 3' single-stranded overhangs (Tomita et al. 2003). In this mutant background, roles of MRN, Ku, Dna2, and Exo1 are studied (Fig. 2). MRN and Dna2 are responsible for the production of 3' single-stranded overhangs (Tomita et al. 2004). But, 3' single-stranded overhangs are produced by concomitant deletion of Ku and MRN, suggesting that unknown nuclease can produce the overhangs in the absence of both MRN and Ku in *taz1* disruptant. Exo1 is not involved in this activity. Telomere ends in *taz1* disruptant is partially unprotected. Indeed, RPA foci and Rad22Rad52 foci are produced at telomere in *taz1* disruptant (Carneiro et al. 2010). Therefore, proteins involved in the resection in *taz1* disruptant may not be same as that in wild-type cells. However, Dna2 is involved in the resection in both wild-type and *taz1*Δ background, suggesting that some of the proteins involved in the resection in *taz1* disruptant are also involved in the resection in wild-type cells.

Roles of DNA Repair Proteins in Telomere Maintenance 601

recombination (Watt et al. 1996) (Onoda et al. 2001). *S. cerevisiae* Sgs1 and Top3 migrate and disentangle a double Holliday junction (dHJ) to produce non-crossover recombination products in vitro (Cejka et al. 2010). This activity is also detected in human RecQ helicase BLM and human topoisomerase IIIa (Wu and Hickson 2003). Mutant of *S. pombe* RecQ helicase *rqh1* is sensitive to DNA damage and has high frequency of recombination under normal growth conditions and following DNA damage, suggesting that Rqh1 is also involved in HR repair both positively and negatively (Murray et al. 1997) (Stewart et al.

As mutation of *S. cerevisiae SGS1* does not affect telomere length, Sgs1 has no apparent role in telomere maintenance in the presence of telomerase activity (Watt et al. 1996). However, the double mutant between telomerase RNA component *TLC1* and *SGS1* shorten telomeres at an increased rate per population doubling and Sgs1 affects telomere-telomere recombination in the absence of telomerase, demonstrating that Sgs1 plays roles at telomere in the absence of telomerase activity (Johnson et al. 2001) (Cohen and Sinclair 2001) (Huang et al. 2001). Xshaped structures are accumulated at telomeres in senescing *tlc1 sgs1* double mutants and these structures are suggested to be the recombination intermediates related to hemicatenanes. This result suggests that Sgs1 is required for the efficient resolution of telomere recombination intermediates in the absence of telomerase (Lee et al. 2007; Chavez, Tsou, and Johnson 2009).

Human RecQ helicase WRN binds to telomere in S phase in primary human IMR90 fibroblasts and is required for efficient replication of the G-rich telomeric DNA strand, suggesting that WRN is required for replication of telomeric DNA in telomerase-negative primary human fibroblasts (Crabbe et al. 2004). In Werner syndrome (WS) cells, replicationassociated telomere loss results in the chromosome fusions, causing genomic instability (Crabbe et al. 2007). The life span of normal human skin fibroblasts derived from WS patients can be extended by expression of the catalytic subunit human telomerase reverse transcriptase (hTERT) (Wyllie et al. 2000; Ouellette et al. 2000). These facts demonstrate that dysfunctional telomere is a major determinant of the premature aging syndrome and WRN plays important role at dysfunctional telomere and telomerase activity can suppress the defect in WRN deficient cells. Consistently, Wrn-deficient mouse, which has telomerase activity, has no disease phenotype, but telomerase-Wrn double null mouse elicits a Wernerlike premature aging syndrome (Chang et al. 2004). Telomere sister chromatid exchange (T-SEC) increases in cells from telomerase-Wrn double null mouse, suggesting that WRN are required to repress inappropriate telomere recombination (Laud et al. 2005) (Multani and Chang 2007). Human WRN and other RecQ helicase BLM co-localizes with telomere in human cancer cells that lack telomerase, ALT cells (Johnson et al. 2001; Opresko et al. 2004; Lillard-Wetherell et al. 2004). As telomeres in ALT cells are maintained by HR, human WRN and BLM are suggested to be involved in the recombination at telomere in ALT cells.

Possible roles of WRN in telomere maintenance will be discussed in the next section.

Pot1 is conserved from *S. pombe* to human and binds to single-stranded telomeric DNA sequence specifically (Baumann and Cech 2001). Deletion of *S. pombe pot1* causes rapid

**3.4 Functional interaction between RecQ helicase and POT1 in** *S. pombe*

**and in human** 

1997) (Doe et al. 2000) (Caspari, Murray, and Carr 2002).

**3.2 Roles of RecQ helicase in telomere maintenance in** *S. cerevisiae*

**3.3 Roles of RecQ helicase in telomere maintenance in mammals** 

Fig. 2. Model for DNA end-processing at dysfunctional telomere. 3' single-stranded overhangs are produced by MRN and Dna2 in *S. pombe taz1*Δ cells (Left). Ku inhibits unknown nuclease, but not nuclease activity depending on MRN-Dna2. 3' single-stranded overhangs are produced by Pif1 or Exo1 in *S. cerevisiae cdc13-1* cells (Right). Unknown nuclease is suggested to function together with Pif1 helicase.

#### **2.4 Proteins involved in DNA end-processing in** *S. cerevisiae cdc13-1* **cells**

*S. cerevisiae* Cdc13 binds telomeric single-stranded DNA (Garvik, Carson, and Hartwell 1995). *cdc13-1* temperature sensitive mutant is used to study proteins that are involved in the resection at uncapped telomeres (Lydall 2009). These studies revealed that the singlestranded DNA at telomeres in *cdc13-1* mutants resembles a DSB end. However, there are some differences between these ends (Fig. 2). In *cdc13-1* mutants at high temperature, Pif1 helicase and Exo1 are redundantly involved in the resection of uncapped telomere (Dewar and Lydall 2010). It remains unclear how Pif1 contribute to the resection. As Pif1 has no nuclease activity, involvement of the unknown nuclease is suggested to cleave singlestranded DNA unwound by Pif1 helicase. Sgs1 also contributes to resection of telomeres in *cdc13-1* mutants (Ngo and Lydall 2010). However, unlike *pif1 exo1* double mutant, resection of telomeres in *cdc13-1* mutant background occurs in *sgs1 exo1* double mutant, demonstrating that Pif1 and Exo1 play major roles in the resection of uncapped telomere at high temperature.
