**2.** *S. cerevisiae* **as a model system to study DDB2-mediated GG-NER in chromatin**

It has been demonstrated that DDB2 is the initial lesion detection factor in GG-NER (Tang et al., 2000; Wakasugi et al., 2002; Fitch et al., 2003b; Pines et al., 2009). Although it has been implicated in the recruitment of XPC to CPD sites (Fitch et al., 2003b); how DDB2 transfers these identified lesions to XPC remains controversial. It is believed that ubiquitination of DDB2 leads to its degradation at damage sites and this degradation is required for CPD repair. However, there are several lines of evidence disputing this model, including: 1) inhibition of ubiquitination-mediated DDB2 degradation in mouse via Cul4a ablation enhances CPD repair (Liu et al., 2009), 2) DDB2 degradation is not stimulated by either DNA binding or XPC association (Luijsterburg et al., 2007), and 3) crystal structures suggest that DDB2 and XPC can co-localize on the lesion (Min & Pavletich, 2007; Scrima et al., 2008). Therefore, we try to explore the budding yeast as a simplified, alternative model system to begin to dissect the role(s) of ubiquitination in DDB2-mediated GG-NER.

### **2.1 Galactose induced expression of DDB2 in** *S. cerevisiae*

As discussed in the introduction, DDB2 has no homolog in budding yeast. However, conservation of the GG-NER pathway and interacting partners such as DDB1 are known (Zaidi et al., 2008). Therefore, we hypothesized that DDB2 would act in a physiological relevant manner in budding yeast GG-NER. We first cloned the DDB2 gene into a low copy number, galactose inducible yeast expression vector. The cloning results in a fusion protein;

*Saccharomyces cerevisiae* as a Model System to

at 30 °C for 48 hours. *∆rad26* (A). *∆rad16* (B).

to its function in DNA damage detection.

Study the Role of Human DDB2 in Chromatin Repair 235

Fig. 3. DDB2 expression suppresses UV sensitivity of *∆rad26* mutant, but not *∆rad16* mutant. BY4741 (WT) cells expressing DDB2 or empty vector were diluted 1/10 and plated on galactose media. Cells were exposed to UV irradiation at dose indicated and grown in dark

As discussed in the introduction, both DDB2 and Rad16 are necessary for lesion identification *in vivo* and are part of E3 ubiquitin ligase complexes (Verhage et al., 1994; Mueller & Smerdon, 1995; Shiyanov et al., 1999; Tang et al., 2000; Wakasugi et al., 2002; Fitch et al., 2003b; Groisman et al., 2003; Ramsey et al., 2004; Pines et al., 2009). Therefore, it was surprising that DDB2 was unable to suppress the *∆rad16* UV sensitive (Fig. 3B). Our data suggest that despite similarities in their biochemical properties, on a gross functional level DDB2 and Rad16 are not analogs. It should be noted that Rad16 has also been implicated in post-incision processes (Reed et al., 1998) while DDB2 has not. It is therefore plausible that DDB2 and Rad16 have analogus functions in the lesion identification step of GG-NER, but

In addition, we found that DDB2 was not able to significantly suppress UV sensitivity of any other knockout strains, including *∆rad7* cells (data not shown). These data are consistent with no known DDB2 homolog in budding yeast. The observed DDB2-dependent suppression of TC-NER deficient UV sensitivity is consistent with reported DDB2

To assess if DDB2 is functioning in a physiologically relevant manner, we first examined the phenotypic effects of mutant DDB2 on DDB2-dependnet suppression of *∆rad26* UV sensitive phenotype. Several DDB2 mutations identified in XPE patients are known to interfere with its ability to function properly in GG-NER. It has been reported that a point mutation changing lysine 244 to glutamic acid (DDB2 K244E) results in inability of DDB2 to make contact with DNA lesions (Scrima et al., 2008) (Fig. 4A). However, this mutation does not alter the ability of DDB2 to interact with DDB1 in the Cul4a E3 ubiquitin ligase complex, therefore its role in ubiquitination is not altered. When this damage recognition deficient mutant DDB2 was introduced into *∆rad26* cells, it was unable to suppress *∆rad26* UV sensitivity (Fig. 5). This suggests that the observed DDB2-conferred UV resistance is linked

this post-incision function of Rad16 is unable to be rescued by DDB2 expression.

**2.3 DDB2 mutations abrogate its ability to suppress** *∆rad26* **UV sensitivity** 

stimulation of GG-NER (Wakasugi et al., 2001; Wakasugi et al., 2002).

DDB2 fused with V5His6 tag (Fig. 2A). Both the empty plasmid vector and the DDB2 containing plasmid were transformed into *S. cerevisiae*. As expected, when cells were grown in the presence of galactose, DDB2 protein was produced as identified by Western blot using both V5 and DDB2 antibodies (Fig. 2B and data not shown). No protein was detectable at the calculated molecular weight of DDB2 in the empty vector control using the same Western blot technique (Fig. 2B).

Fig. 2. Expression of DDB2-HIS in S. cerevisiae. (A) Schematic of DDB2 fusion cloned into pYCT/C2 expression vector. (B). Western blot (WB) using V5 antibody to detect expression of DDB2 containing or empty vector. (C) Glucose addition (4%) stops production of DDB2 detected by Western blot using V5 antibody, equal amount of total protein was verified using coomassie blue staining. BY4741 is the wild type (WT) strain used in these experiments.

To access the efficacy of the galactose induction 4% glucose was added to the media. Rapid shut down of the galactose inducible promoter is presumed due to the significant decrease in DDB2 protein levels 30 min post addition of glucose (Fig. 2C). This observed decrease in DDB2 protein levels is likely due to normal protein turnover in the absence of nascent DDB2 transcription and subsequent translation. These data confirm that DDB2 is expressed in *S. cerevisiae* cells under the control of the galactose promoter.

#### **2.2 DDB2 suppresses UV sensitivity of** *∆rad26* **cells**

Next we identified genetic background in which a DDB2-dependent phenotype could be observed. We screened several yeast strains in which various NER proteins were deleted. The strains tested were *∆rad7* and *∆rad16* in which only TC-NER is active, *∆rad26* in which only GG-NER is active, and *∆rad1* in which the core pathway is defective and therefore there is no active NER. The spotting assay was used to determine DDB2 dependent suppression of UV sensitivity. Clearly, DDB2 expression suppresses the UV sensitive phenotype of *∆rad26* cells (Fig. 3A). Survival curve experiments verified these findings (Data not shown).

DDB2 fused with V5His6 tag (Fig. 2A). Both the empty plasmid vector and the DDB2 containing plasmid were transformed into *S. cerevisiae*. As expected, when cells were grown in the presence of galactose, DDB2 protein was produced as identified by Western blot using both V5 and DDB2 antibodies (Fig. 2B and data not shown). No protein was detectable at the calculated molecular weight of DDB2 in the empty vector control using the same

Fig. 2. Expression of DDB2-HIS in S. cerevisiae. (A) Schematic of DDB2 fusion cloned into pYCT/C2 expression vector. (B). Western blot (WB) using V5 antibody to detect expression of DDB2 containing or empty vector. (C) Glucose addition (4%) stops production of DDB2 detected by Western blot using V5 antibody, equal amount of total protein was verified using coomassie blue staining. BY4741 is the wild type (WT) strain used in these

To access the efficacy of the galactose induction 4% glucose was added to the media. Rapid shut down of the galactose inducible promoter is presumed due to the significant decrease in DDB2 protein levels 30 min post addition of glucose (Fig. 2C). This observed decrease in DDB2 protein levels is likely due to normal protein turnover in the absence of nascent DDB2 transcription and subsequent translation. These data confirm that DDB2 is expressed in *S.* 

Next we identified genetic background in which a DDB2-dependent phenotype could be observed. We screened several yeast strains in which various NER proteins were deleted. The strains tested were *∆rad7* and *∆rad16* in which only TC-NER is active, *∆rad26* in which only GG-NER is active, and *∆rad1* in which the core pathway is defective and therefore there is no active NER. The spotting assay was used to determine DDB2 dependent suppression of UV sensitivity. Clearly, DDB2 expression suppresses the UV sensitive phenotype of *∆rad26* cells (Fig. 3A). Survival curve experiments verified these findings (Data not shown).

*cerevisiae* cells under the control of the galactose promoter.

**2.2 DDB2 suppresses UV sensitivity of** *∆rad26* **cells** 

Western blot technique (Fig. 2B).

experiments.

Fig. 3. DDB2 expression suppresses UV sensitivity of *∆rad26* mutant, but not *∆rad16* mutant. BY4741 (WT) cells expressing DDB2 or empty vector were diluted 1/10 and plated on galactose media. Cells were exposed to UV irradiation at dose indicated and grown in dark at 30 °C for 48 hours. *∆rad26* (A). *∆rad16* (B).

As discussed in the introduction, both DDB2 and Rad16 are necessary for lesion identification *in vivo* and are part of E3 ubiquitin ligase complexes (Verhage et al., 1994; Mueller & Smerdon, 1995; Shiyanov et al., 1999; Tang et al., 2000; Wakasugi et al., 2002; Fitch et al., 2003b; Groisman et al., 2003; Ramsey et al., 2004; Pines et al., 2009). Therefore, it was surprising that DDB2 was unable to suppress the *∆rad16* UV sensitive (Fig. 3B). Our data suggest that despite similarities in their biochemical properties, on a gross functional level DDB2 and Rad16 are not analogs. It should be noted that Rad16 has also been implicated in post-incision processes (Reed et al., 1998) while DDB2 has not. It is therefore plausible that DDB2 and Rad16 have analogus functions in the lesion identification step of GG-NER, but this post-incision function of Rad16 is unable to be rescued by DDB2 expression.

In addition, we found that DDB2 was not able to significantly suppress UV sensitivity of any other knockout strains, including *∆rad7* cells (data not shown). These data are consistent with no known DDB2 homolog in budding yeast. The observed DDB2-dependent suppression of TC-NER deficient UV sensitivity is consistent with reported DDB2 stimulation of GG-NER (Wakasugi et al., 2001; Wakasugi et al., 2002).

#### **2.3 DDB2 mutations abrogate its ability to suppress** *∆rad26* **UV sensitivity**

To assess if DDB2 is functioning in a physiologically relevant manner, we first examined the phenotypic effects of mutant DDB2 on DDB2-dependnet suppression of *∆rad26* UV sensitive phenotype. Several DDB2 mutations identified in XPE patients are known to interfere with its ability to function properly in GG-NER. It has been reported that a point mutation changing lysine 244 to glutamic acid (DDB2 K244E) results in inability of DDB2 to make contact with DNA lesions (Scrima et al., 2008) (Fig. 4A). However, this mutation does not alter the ability of DDB2 to interact with DDB1 in the Cul4a E3 ubiquitin ligase complex, therefore its role in ubiquitination is not altered. When this damage recognition deficient mutant DDB2 was introduced into *∆rad26* cells, it was unable to suppress *∆rad26* UV sensitivity (Fig. 5). This suggests that the observed DDB2-conferred UV resistance is linked to its function in DNA damage detection.

*Saccharomyces cerevisiae* as a Model System to

defects in various steps of the ubiquitin pathway.

model system will facilitate deciphering such questions.

photolesions. Biochemistry 40(8): 2495-2501.

in human cells.

**3. Conclusion** 

**4. Acknowledgment**

**5. References** 

from the Concern Foundation.

276(51): 48175-48182.

Study the Role of Human DDB2 in Chromatin Repair 237

Taken together, these data suggest that exogenously expressed DDB2 is acting in a physiologically relevant manner. Additionally, our findings indicate that the DNA damage recognition function of DDB2 is essential for the observed suppression of UV sensitivity. We also found that DDB2 function is dependent on interaction with Mms1, a subunit of an E3 ubiquitin ligase. These observations are consistent with what is reported for DDB2 function

Studies in *Saccharomyces cerevisiae* have made major contributions to our understanding of NER. Here, we present evidence suggesting that *S. cerevisiae* can be used to dissect the roles of human DDB2 in initiating NER in chromatin. Since DDB2 functions are regulated by the ubiquitin pathway and DDB2 itself is a component of an E3 ligase, it will be interesting to explore the regulation of DDB2 functions by ubiquitination, using yeast mutants with

Ubiquitination is a well studied post-translational modification and recent data suggest multiple fates of ubiquitin modified proteins (Sadowski & Sarcevic, 2010). It will be important to determine if ubiquitination of DDB2 promotes its degradation or controls DDB2 association with chromatin. The budding yeast system described here will also provide an alternative system to screen the effect(s) of various DDB2 lysine mutations to determine which amino acid residue(s) is modified. Additionally, as reviewed by Kirkin et al., ubiquitin signaling is altered in many cancers (Kirkin and Dikic 2010), suggesting a potential role of ubiquitination in regulating DNA binding proteins such as transcription factors and repair proteins. Therefore, it will be interesting to determine what, if any, role ubiquitination plays in the chromatin association of other DNA binding proteins, specifically transcription factors and repair proteins. The utilization of the budding yeast

The project described in this article was supported by Award R01ES017784 from the National Institute of Environmental Health Sciences and by a young investigator award

Chen, X., Zhang, Y., Douglas, L., and Zhou, P. 2001. UV-damaged DNA-binding proteins

Conconi, A., Bespalov, V.A., and Smerdon, M.J. 2002. Transcription-coupled repair in RNA polymerase I-transcribed genes of yeast. Proc Natl Acad Sci U S A 99(2): 649-654. Douki, T. and Cadet, J. 2001. Individual determination of the yield of the main UV-induced

Fitch, M.E., Cross, I.V., Turner, S.J., Adimoolam, S., Lin, C.X., Williams, K.G., and Ford, J.M.

are targets of CUL-4A-mediated ubiquitination and degradation. J Biol Chem

dimeric pyrimidine photoproducts in DNA suggests a high mutagenicity of CC

2003a. The DDB2 nucleotide excision repair gene product p48 enhances global genomic repair in p53 deficient human fibroblasts. DNA Repair (Amst) 2(7): 819-826.

Fig. 4. Crystal structure of DDB2 mutations modified from crystal structure solved by Scrima et al. (A) Lysine to glutamic acid substitution at aa 244 predicted to effect DDB2 DNA interaction. Red residue indicates site of mutation. Yellow indicates damaged DNA strand. (B) Deletion of aa 349 and substitution of proline for leucine at aa 350. This mutation is predicted to effect the DDB2 DDB1 interaction. Red indicates site of mutation. Mutant DDB2 was constructed by site directed mutagenesis.

Another mutation that affects DDB2's function prevents the interaction with its in vivo partner DDB1 (Nichols et al. 2000). This mutation was also constructed and is a complex mutation, consisting of both a deletion of amino acid 349 and a proline substitution for leucine at amino acid 350 (DDB2 L350P) (Fig. 4B). Like DDB2 K244E, this mutation also abrogated DDB2's ability to suppress UV sensitivity in *∆rad26* cells (Fig. 5). These data suggest DDB2-conferred UV resistance is dependent on a conserved interacting partner.

Fig. 5. DDB2 mutations and deletion of Mms1 (DDB1 homolog) abrogate suppression of UV sensitivity in *∆rad26* cells.

Although Mms1 has been identified as the budding yeast DDB1 homolog (Zaidi et al., 2008), there are no reports of it being involved in NER. However, our previous observation suggesting DDB2 function requires a conserved interacting partner prompted us to test DDB2 function in the absence of Mms1. To test this, wild type DDB2 was expressed in the *∆rad26∆mms1* double mutant and UV sensitivity was accessed by spotting assays. Indeed, this reciprocal experiment verified that Mms1 is necessary for DDB2-dependent suppression of UV sensitivity (Fig. 5).

Taken together, these data suggest that exogenously expressed DDB2 is acting in a physiologically relevant manner. Additionally, our findings indicate that the DNA damage recognition function of DDB2 is essential for the observed suppression of UV sensitivity. We also found that DDB2 function is dependent on interaction with Mms1, a subunit of an E3 ubiquitin ligase. These observations are consistent with what is reported for DDB2 function in human cells.
