**3.4 DSB repair: Nonhomologous end joining and homologous recombination**

ROS, especially •OH generated by ionizing radiation or via the Fenton reaction (Karanjawala et al., 2003; Clark, 2008), also have a high potential to cause double-strand breaks (DSB) (Karanjawala et al., 2002; Karanjawala et al., 2003). DSB require repair mechanisms distinct from photolyases, NER and BER. Therefore cells primarily depend on either the nonhomologous DNA end joining pathway (NHEJ) or homologous recombination (HR). NHEJ is an error-prone repair pathway, which directly ligates the free DNA ends together. In animals, the pathway is discussed to start with the binding of the heterodimeric Ku70/Ku80 complex to a DNA end. This step is required for employment of DNAdependent protein kinase (DNA-PK) and Artemis endonuclease that process the DNA ends (Ma et al., 2002), while rejoining and ligation is performed by the XRCC4/LIG4/XLF complex (Grawunder et al., 1997; Barnes et al., 1998) (Fig. 4). The processing of the DNA ends can result in deletions or insertions and is the reason why NHEJ based repair often results in mutations in the repaired DNA. Current research in plants indicates that the NHEJ pathway is conserved among plants and animals. Ku70 and Ku80 related proteins as well as the Artemis-like protein SNM1/PSO1 are expressed in *Arabidopsis* and rice, and *Arabidopsis* mutants affected in these proteins become hypersensitive to -irradiation and the chemotherapeutic agent bleomycin, a double-strand break inducing chemical, which is in agreement with their roles in NHEJ (Tamura et al., 2002; Friesner and Britt, 2003; Gallego et al., 2003; Molinier et al., 2004a; Kimura et al., 2005; Kimura and Sakaguchi, 2006; Charbonnel et al., 2010). Likewise, XRCC4 and Lig4 homologues have been described in plants, and functionally connected to NHEJ (West et al., 2000; Friesner and Britt, 2003; Kimura and Sakaguchi, 2006; Waterworth et al., 2010).

In contrast to the error prone NHEJ pathway, HR is a more accurate repair mechanism that uses homologous DNA strands as templates for repair activities (Boyko et al., 2006a; Boyko et al., 2006b; Li and Ma, 2006; Osman et al., 2011). Several alternative pathways may exist that allow HR based repair of DSBs, however, good evidence is provided for at least two alternative pathways in plants. One is the synthesis-dependent strand annealing (SDSA) mechanism which involves the meiotic recombination11/Rad50/X-ray sensitive 2 (MRN) complex (Waterworth et al., 2007; Ronceret et al., 2009; Amiard et al., 2010). The MRN complex is discussed to function as a first sensor of double strand breaks. It generates single strand DNA at the DSB sites that can be used as templates to mediate HR by RecA and Rad51 homologues (Lin et al., 2006; Li et al., 2007; Markmann-Mulisch et al., 2007; Odahara et al., 2007; Vignard et al., 2007; Waterworth et al., 2007; Odahara et al., 2009; Ronceret et al., 2009; Amiard et al., 2010; Chittela and Sainis, 2010; Devisetty et al., 2010; Ko et al., 2010; Schaefer et al., 2010; Wang et al., 2010; Ko et al., 2011) (Fig. 4). However, the precise subsequent steps of Holliday structure formation, cleavage by endonucleases and dissociation into two DNA chains is only poorly understood in plants. Alternatively to SDSA, plants also use the single strand annealing (SSA) mechanism (Tissier et al., 1995; Ayora et al., 2002; Blanck et al., 2009; Mannuss et al., 2010). SSA requires a double strand break between two repeated sequences that are oriented in the same direction. Adjacent to the break, single-stranded DNA is created so that the repeated sequences can be used as

Doucet-Chabeaud et al., 2001; Chen et al., 2003). However, although a similar role of plant PARP1 and PARP2 in damaged DNA recognition and initiation of DNA repair is likely, a

ROS, especially •OH generated by ionizing radiation or via the Fenton reaction (Karanjawala et al., 2003; Clark, 2008), also have a high potential to cause double-strand breaks (DSB) (Karanjawala et al., 2002; Karanjawala et al., 2003). DSB require repair mechanisms distinct from photolyases, NER and BER. Therefore cells primarily depend on either the nonhomologous DNA end joining pathway (NHEJ) or homologous recombination (HR). NHEJ is an error-prone repair pathway, which directly ligates the free DNA ends together. In animals, the pathway is discussed to start with the binding of the heterodimeric Ku70/Ku80 complex to a DNA end. This step is required for employment of DNAdependent protein kinase (DNA-PK) and Artemis endonuclease that process the DNA ends (Ma et al., 2002), while rejoining and ligation is performed by the XRCC4/LIG4/XLF complex (Grawunder et al., 1997; Barnes et al., 1998) (Fig. 4). The processing of the DNA ends can result in deletions or insertions and is the reason why NHEJ based repair often results in mutations in the repaired DNA. Current research in plants indicates that the NHEJ pathway is conserved among plants and animals. Ku70 and Ku80 related proteins as well as the Artemis-like protein SNM1/PSO1 are expressed in *Arabidopsis* and rice, and *Arabidopsis* mutants affected in these proteins become hypersensitive to -irradiation and the chemotherapeutic agent bleomycin, a double-strand break inducing chemical, which is in agreement with their roles in NHEJ (Tamura et al., 2002; Friesner and Britt, 2003; Gallego et al., 2003; Molinier et al., 2004a; Kimura et al., 2005; Kimura and Sakaguchi, 2006; Charbonnel et al., 2010). Likewise, XRCC4 and Lig4 homologues have been described in plants, and functionally connected to NHEJ (West et al., 2000; Friesner and Britt, 2003; Kimura and

In contrast to the error prone NHEJ pathway, HR is a more accurate repair mechanism that uses homologous DNA strands as templates for repair activities (Boyko et al., 2006a; Boyko et al., 2006b; Li and Ma, 2006; Osman et al., 2011). Several alternative pathways may exist that allow HR based repair of DSBs, however, good evidence is provided for at least two alternative pathways in plants. One is the synthesis-dependent strand annealing (SDSA) mechanism which involves the meiotic recombination11/Rad50/X-ray sensitive 2 (MRN) complex (Waterworth et al., 2007; Ronceret et al., 2009; Amiard et al., 2010). The MRN complex is discussed to function as a first sensor of double strand breaks. It generates single strand DNA at the DSB sites that can be used as templates to mediate HR by RecA and Rad51 homologues (Lin et al., 2006; Li et al., 2007; Markmann-Mulisch et al., 2007; Odahara et al., 2007; Vignard et al., 2007; Waterworth et al., 2007; Odahara et al., 2009; Ronceret et al., 2009; Amiard et al., 2010; Chittela and Sainis, 2010; Devisetty et al., 2010; Ko et al., 2010; Schaefer et al., 2010; Wang et al., 2010; Ko et al., 2011) (Fig. 4). However, the precise subsequent steps of Holliday structure formation, cleavage by endonucleases and dissociation into two DNA chains is only poorly understood in plants. Alternatively to SDSA, plants also use the single strand annealing (SSA) mechanism (Tissier et al., 1995; Ayora et al., 2002; Blanck et al., 2009; Mannuss et al., 2010). SSA requires a double strand break between two repeated sequences that are oriented in the same direction. Adjacent to the break, single-stranded DNA is created so that the repeated sequences can be used as

detailed *in planta* functional description is still missing for these proteins.

Sakaguchi, 2006; Waterworth et al., 2010).

**3.4 DSB repair: Nonhomologous end joining and homologous recombination** 

complementary strands to anneal the ends of the break, after which non-homologous tails are detached and nicks can be ligated (Tissier et al., 1995; Puchta, 2005; Blanck et al., 2009; Mannuss et al., 2010). Because HR is less likely to cause changes in the genetic information than NHEJ, it is likely that the extent to which either NHEJ or HR repair pathways are employed in DSB repair may impact genome evolution in living organisms.

Fig. 4. ROS induced repair of double-strand breaks.

Two alternative pathways for which strong evidence is present to exist in plants are shown. Double-strand breaks can either be repaired by nonhomologous DNA end joining pathway (NHEJ; left hand side), or by homologous recombination (HR; right hand side).

#### **3.5 DNA repair in chloroplasts**

Based on their high metabolic activities in respiration and photosynthesis, organelles are centers of ROS production. Both mitochondria and chloroplasts possess their own repair

Recognition and Repair Pathways of Damaged DNA in Higher Plants 213

phytoestrogen with unknown function. Both compounds are associated with effecting cell proliferation, cell cycle, and apoptosis of mammalian cells (Ndebele et al., 2010; Delmas et al., 2011) and may be signaling molecules in plants to trigger specific responses upon UV

A critical factor that is discussed as a regulator of DNA repair pathways in response to increased ROS accumulation in the cell appears to be ADP-Ribose/NADH Pyrophosphohydrolase AtNUDX7 (Ishikawa et al., 2009). AtNUDX7 belongs to the family of Nudix hydrolyases, which catalyze the hydrolysis of dinucleoside polyphosphates, nucleoside di- and triphosphates, nucleotide sugars, and coenzymes in plants and animals (McLennan, 2006; Kraszewska, 2008). AtNUDX7 substrates are ADP-Ribose (ADP-Rib) and NADH which are converted to AMP plus ribose 5-phosphate and nicotinamide mononucleotide (NMNH) plus AMP, respectively (Ogawa et al., 2005). The protein may have a central function for the homeostasis of NAD+ pools by supplying ATP via nucleotide recycling from free ADP-Ribose molecules. This may be critical for the cell since substantial PARP activity can significantly lower NAD+ and ATP levels; such a depletion of cellular energy can result in necrotic cell death (Ha and Snyder, 1999; Virag and Szabo, 2002; De Block et al., 2005). As discussed for BER, DNA damaged-induced, PARP-dependent poly(ADP)-ribosylation of proteins is considered a critical step for recognition of damage to be converted into intracellular signals that can trigger DNA repair programs or cell death. Consequently, AtNUDX7 is up-regulated upon abiotic stress (salinity, drought, high light, paraquat), and plants constitutively overexpressing AtNUDX7 become less susceptible to

A microarray study has shown that UV-C, bleomycin, or biotic stress factors elicit a hypersensitive response and increased H202 levels in the cell (Molinier et al., 2005). Although each stress elicited specific responses, the authors could also find 209 genes that were commonly up-regulated, while 54 were similarly down-regulated by all three stress treatments. Among the commonly regulated genes were components of signaling pathways, transcription factors, and genes connected with an oxidative stress or defense response. Cellcycle genes were also down-regulated after genotoxic stress exposure, as was earlier noted for animals (Dasika et al., 1999). However, the authors also noted that in Arabidopsis expression of only a comparably few number of repair genes was induced, and concluded that the plant must be mainly relying on existing synthesized proteins. It will be interesting to see whether results are broadly applicable to other plant species or whether the tested

A central regulator of the fate of damaged cells between apoptosis or cell cycle arrest and DNA repair in animals is the tumor-suppressing p53, sometimes dubbed the "guardian of the genome". This transcription factor controls not only cell cycle genes like p21 and apoptosis factors like PUMA, NOXA, and BAX but also various components of major DNA repair pathways such as CSB, DDB2 and XPC (NER); FANCC (DNA crosslink repair) and MSH2, MLH1, and PMS2 (mismatch repair) (Gatz and Wiesmuller, 2006; Brady and Attardi, 2010). There is also evidence that it has a more direct role in BER, interacting with APE1 and OGG1 and thereby enhancing the excision of oxidized DNA bases (Gatz and Wiesmuller, 2006; Vigneron and Vousden, 2010). Additionally p53 seems to recognize and bind directly to certain DNA structures e.g. Holliday junctions and mismatches where it represses the

exposure.

these stress conditions (Ishikawa et al., 2009).

conditions and responses are specific for Arabidopsis.

**3.7 ATM/ATR dependent regulation of DNA repair** 

pathways, and it appears to be that they have most of the repair pathways that are also found in the nucleus (for an excellent review about mitochondrial repair see (Boesch et al., 2011). We will focus here on chloroplast repair and briefly summarize recent findings.

The chloroplast genome is in general relative small, but gene numbers can vary significantly between species ranging from, for example, 54 in *Helicosporidium sp. ex Simulium jonesii* and up to 301 in *Pinus koraiensis* (http://chloroplast.cbio.psu.edu/); (Cui et al., 2006). These genes and their corresponding proteins are crucial for proper functioning of the organelle and hence survival of the plant and it is not surprising that chloroplasts have several repair pathways.

A recent report described a rice CPD photolyase to mediate repair of direct DNA damage caused by UV light (Takahashi et al., 2011), and also an earlier report describes PHR2, a class II photolyase predicted to be in chloroplasts of *C. reinhardtii* (Petersen et al., 1999). There is also strong evidence in *Arabidopsis* for two bifunctional DNA glycosylase/lyase of the *E. coli* Endonuclease III/Nth type and an APE to be involved in repair of ROS based DNA damage (Gutman and Niyogi, 2009). The authors can show specific localization of the three proteins to the chloroplast and specific activities *in vitro*. However, single or even triple mutants affected in the three proteins do not display any apparent developmental defects or increased sensitivities to photo-oxidative stress (e.g. UV- and high light or methyl viologen), from which the authors concluded that additional, yet unknown BER repair pathways exist in chloroplast (Gutman and Niyogi, 2009).

Currently no clear data are available for NER activities in the chloroplasts and only poor evidence is currently present on whether or how chloroplasts repair DSBs. Work on the green algae *Chlamydomonas reinhardtii* showed presence of a chloroplast-located RecA homolog, which is inducible in expression by DNA-damaging reagents (Nakazato et al., 2003). In addition, *Arabidopsis* T-DNA insertion mutants affected in a chloroplast localized RecA (*cpRecA*) homolog have increased amounts of single-strand DNA, altered structures of chloroplast DNA, and chloroplasts showed signs of reduced function after four generations post T-DNA insertion (Rowan et al., 2010). Yet, further data for additional repair proteins is still missing, as well as strong evidence for HR or NHEJ activities in chloroplasts of higher plants. Recent findings, however, indicate that chloroplasts repair DSBs using microhomology-mediated end joining (MMEJ) (Kwon et al., 2010). This repair mechanism requires only very short (2–14 bp) regions of homology, and is discussed as a potential backup to NHEJ in eukaryotes (Heacock et al., 2004; Bennardo et al., 2008). Although Kwon and co-workers provide strong evidence for MMEJ in chloroplasts, it is currently open which proteins mediate this repair.

#### **3.6 Physiological responses after UV and ROS exposure**

Besides immediate repair processes, it is also critical for plant cells to generate a physiological environment in which further DNA damage is prevented or at least reduced. A common physiological response to UV exposure in plants appears to be the accumulation of anthocyanin and flavonoids, potentially as a photoprotective or ROS quenching mechanism (Ng et al., 1964 ; Yatsuhashi et al., 1982; Takeda and Abe, 1992; Ye et al., 2010). It is interesting to note that some plants like grape vine (*Vitis vinifera*) or common bean (*Phaseolus vulgaris*) accumulate resveratrol or coumestrol, respectively, in response to UV exposure (Langcake and Pryce, 1977; Beggs et al., 1985). Resveratrol is a protective phytoalexin that is produced primarily under biotic stress conditions, while coumestrol is a

pathways, and it appears to be that they have most of the repair pathways that are also found in the nucleus (for an excellent review about mitochondrial repair see (Boesch et al., 2011). We will focus here on chloroplast repair and briefly summarize recent findings. The chloroplast genome is in general relative small, but gene numbers can vary significantly between species ranging from, for example, 54 in *Helicosporidium sp. ex Simulium jonesii* and up to 301 in *Pinus koraiensis* (http://chloroplast.cbio.psu.edu/); (Cui et al., 2006). These genes and their corresponding proteins are crucial for proper functioning of the organelle and hence survival of the plant and it is not surprising that chloroplasts have several repair

A recent report described a rice CPD photolyase to mediate repair of direct DNA damage caused by UV light (Takahashi et al., 2011), and also an earlier report describes PHR2, a class II photolyase predicted to be in chloroplasts of *C. reinhardtii* (Petersen et al., 1999). There is also strong evidence in *Arabidopsis* for two bifunctional DNA glycosylase/lyase of the *E. coli* Endonuclease III/Nth type and an APE to be involved in repair of ROS based DNA damage (Gutman and Niyogi, 2009). The authors can show specific localization of the three proteins to the chloroplast and specific activities *in vitro*. However, single or even triple mutants affected in the three proteins do not display any apparent developmental defects or increased sensitivities to photo-oxidative stress (e.g. UV- and high light or methyl viologen), from which the authors concluded that additional, yet unknown BER repair pathways exist

Currently no clear data are available for NER activities in the chloroplasts and only poor evidence is currently present on whether or how chloroplasts repair DSBs. Work on the green algae *Chlamydomonas reinhardtii* showed presence of a chloroplast-located RecA homolog, which is inducible in expression by DNA-damaging reagents (Nakazato et al., 2003). In addition, *Arabidopsis* T-DNA insertion mutants affected in a chloroplast localized RecA (*cpRecA*) homolog have increased amounts of single-strand DNA, altered structures of chloroplast DNA, and chloroplasts showed signs of reduced function after four generations post T-DNA insertion (Rowan et al., 2010). Yet, further data for additional repair proteins is still missing, as well as strong evidence for HR or NHEJ activities in chloroplasts of higher plants. Recent findings, however, indicate that chloroplasts repair DSBs using microhomology-mediated end joining (MMEJ) (Kwon et al., 2010). This repair mechanism requires only very short (2–14 bp) regions of homology, and is discussed as a potential backup to NHEJ in eukaryotes (Heacock et al., 2004; Bennardo et al., 2008). Although Kwon and co-workers provide strong evidence for MMEJ in chloroplasts, it is currently open

Besides immediate repair processes, it is also critical for plant cells to generate a physiological environment in which further DNA damage is prevented or at least reduced. A common physiological response to UV exposure in plants appears to be the accumulation of anthocyanin and flavonoids, potentially as a photoprotective or ROS quenching mechanism (Ng et al., 1964 ; Yatsuhashi et al., 1982; Takeda and Abe, 1992; Ye et al., 2010). It is interesting to note that some plants like grape vine (*Vitis vinifera*) or common bean (*Phaseolus vulgaris*) accumulate resveratrol or coumestrol, respectively, in response to UV exposure (Langcake and Pryce, 1977; Beggs et al., 1985). Resveratrol is a protective phytoalexin that is produced primarily under biotic stress conditions, while coumestrol is a

pathways.

in chloroplast (Gutman and Niyogi, 2009).

which proteins mediate this repair.

**3.6 Physiological responses after UV and ROS exposure** 

phytoestrogen with unknown function. Both compounds are associated with effecting cell proliferation, cell cycle, and apoptosis of mammalian cells (Ndebele et al., 2010; Delmas et al., 2011) and may be signaling molecules in plants to trigger specific responses upon UV exposure.

A critical factor that is discussed as a regulator of DNA repair pathways in response to increased ROS accumulation in the cell appears to be ADP-Ribose/NADH Pyrophosphohydrolase AtNUDX7 (Ishikawa et al., 2009). AtNUDX7 belongs to the family of Nudix hydrolyases, which catalyze the hydrolysis of dinucleoside polyphosphates, nucleoside di- and triphosphates, nucleotide sugars, and coenzymes in plants and animals (McLennan, 2006; Kraszewska, 2008). AtNUDX7 substrates are ADP-Ribose (ADP-Rib) and NADH which are converted to AMP plus ribose 5-phosphate and nicotinamide mononucleotide (NMNH) plus AMP, respectively (Ogawa et al., 2005). The protein may have a central function for the homeostasis of NAD+ pools by supplying ATP via nucleotide recycling from free ADP-Ribose molecules. This may be critical for the cell since substantial PARP activity can significantly lower NAD+ and ATP levels; such a depletion of cellular energy can result in necrotic cell death (Ha and Snyder, 1999; Virag and Szabo, 2002; De Block et al., 2005). As discussed for BER, DNA damaged-induced, PARP-dependent poly(ADP)-ribosylation of proteins is considered a critical step for recognition of damage to be converted into intracellular signals that can trigger DNA repair programs or cell death. Consequently, AtNUDX7 is up-regulated upon abiotic stress (salinity, drought, high light, paraquat), and plants constitutively overexpressing AtNUDX7 become less susceptible to these stress conditions (Ishikawa et al., 2009).

A microarray study has shown that UV-C, bleomycin, or biotic stress factors elicit a hypersensitive response and increased H202 levels in the cell (Molinier et al., 2005). Although each stress elicited specific responses, the authors could also find 209 genes that were commonly up-regulated, while 54 were similarly down-regulated by all three stress treatments. Among the commonly regulated genes were components of signaling pathways, transcription factors, and genes connected with an oxidative stress or defense response. Cellcycle genes were also down-regulated after genotoxic stress exposure, as was earlier noted for animals (Dasika et al., 1999). However, the authors also noted that in Arabidopsis expression of only a comparably few number of repair genes was induced, and concluded that the plant must be mainly relying on existing synthesized proteins. It will be interesting to see whether results are broadly applicable to other plant species or whether the tested conditions and responses are specific for Arabidopsis.
