Contributions of Histone H3 Nucleosome Core Surface Mutations to Chromatin Structures, Silencing and DNA Repair

Histone H3 mutations in residues that cluster in a discrete region on the nucleosome surface around lysine 79 of H3 affect H3-K79 methylation, impair transcriptional silencing in subtelomeric chromatin, and reveal distinct contributions of histone H3 to various DNA-damage response and repair pathways. These residues might act by recruitment of silencing and DNA-damage response factors. Alternatively, their location on the nucleosome surface suggests a possible involvement in nucleosome positioning, stability and nucleosome interactions. Here, we show that the yeast H3 mutants hht2-T80A, hht2-K79E, hht2-L70S, and hht2-E73D show normal nucleosome positioning and stability in minichromosomes. However, loss of silencing in a subtelomeric URA3 gene correlates with a shift of the promoter nucleosome, while nucleosome positions and stability in the coding region are maintained. Moreover, the H3 mutants show normal repair of UV lesions by photolyase and nucleotide excision repair in minichromosomes and slightly enhanced repair in the subtelomeric region. Thus, these results support a role of those residues in the recruitment of silencing proteins and argue against a general role in nucleosome organization.


Introduction
Chromatin structure serves as a central regulator for DNAassociated cellular processes in eukaryotic cells, including transcription, replication, and repair. The primary components of chromatin are the highly conserved histone proteins, which provide the structural foundation for the organization of DNA. Molecular genetic, biochemical and genome wide studies have elucidated many aspects of the structural and functional roles of histones in genome organization. Here we investigate a set of nucleosome surface mutants in yeast with respect to their potential impact on chromatin structure, silencing and DNA repair.
Eukaryotic genomes are folded in arrays of nucleosome cores connected by linker DNA and further condensed into compact fibers and additional levels of higher order structures. While nucleosome cores are formed by intranucleosomal histone-histone and histone-DNA contacts, the close proximity of nucleosomes in arrays and higher order structures may require internucleosomal interactions of histones and DNA as well as a contribution of nonhistone chromosomal proteins [1,2]. Nucleosome cores are cylindrical particles containing of about 147 bp DNA wrapped around an octamer of core histones, two each of H2A, H2B, H3, and H4. The histones are folded in the centre of the core and have flexible N-terminal tails that protrude from the particle [3]. The tails may interact with the DNA or histones of adjacent nucleosomes, thereby contributing to higher order structures [4,5]. The histone fold domains show an irregular surface with a distinct charge distribution that has also been implicated in nucleosome-nucleosome contacts to promote chromatin higher order structure formation [3,4,6].
In chromatin of the budding yeast S. cerevisiae, nucleosome cores are connected by short linker DNA [7,8] and reduced cleavage by micrococcal nuclease (MNase) between nucleosomes suggested that some nucleosomes can be in close face-to-face contact [9,10,11,12]. The nucleosome arrays are frequently interrupted by nucleosome free regions (NFRs) that occur at promoters and 39 ends of genes and at origins of replication. A major fraction of nucleosomes is positioned [8,13,14]. Positioning is mediated by boundaries such as NFRs that restrict statistical distribution of nucleosomes, the DNA-sequence, and chomatin folding [8,9,11,12,13,14,15,16,17,18,19,20]. Despite the detailed information on the genome wide arrangement of nucleosomes in yeast, the folding into higher order structures remains unclear and evidence supporting compact fibers in yeast is controversial [21,22,23].
Heterochromatin is a specialized higher order chromatin structure that restricts access of proteins to DNA and silences gene expression. In S. cerevisiae silent chromatin regions are found at telomeres, at the silent mating-type loci, and at the ribosomal DNA. Subtelomeric silencing depends on spreading of the silencing complex, which consists of Sir2, Sir3, and Sir4 proteins, , and hht2-E73D (MFY19) were digested with MNase (0, 25, 50, 100 and 200 U/ml). DNA was purified and separated on a 1% agarose gel containing ethidium bromide. M is a 2 log DNA ladder (New England BioLabs). (B) Schematic representation of the minichromosomes YRpFT35 and YRpFT38. The minichromosomes contain one (Umid) and two copies (UmidA and UmidB) of sequences of the URA3 coding region (Umid) inserted in the TRP1ARS1 circle. Indicated are: the TRP1 gene, the autonomously replicating sequence ARS1, the EcoRI-XbaI fragment used to generate probes for indirect end-labeling; nucleosome positions (circles) determined by MNase digestion. Four nucleosomes are tightly packed in YRpFT35 forming a tetranucleosome (dark circles) [10,42]. (C) Nucleosome footprints in YRpFT35. Chromatin (C lanes) and DNA (D lanes) was digested with MNase, cut with XbaI, run on a 1% agarose gel, blotted and hybridized with the XbaI-EcoRI probe.
from the telomeres and requires the histone deacetylase activity of Sir2 [24]. Genetic screens have identified H3 and H4 residues on a specific surface located at the H3-H4 histone-fold motif, which are important for transcriptional silencing of RNA polymerase II dependent reporter genes. Mutated amino acid residues that impaired silencing cluster around and include lysine 79 of H3 (H3-K79) [25,26,27]. H3-K79 is a site for methylation, by the conserved histone methyltransferase Dot1. Dot1 can add 1-3 methyl groups per residue, influenced in part by another histone modification, ubiquitylation of H2B [28]. H3-K79 is hypermethylated in transcriptionally active and hypomethylated in silenced chromatin, suggesting that hypermethylation of histone H3-K79 limits silencing to discrete loci by preventing the binding of Sir proteins elsewhere along the genome [29,30]. Recent work established that Sir3 binds two locations on the nucleosome core, the LRS surface (Loss of Ribosomal DNA Silencing) and the Nterminal histone tails [31,32,33,34].
In addition to silencing, some of the histone mutants compromise the DNA damage response. UV irradiation causes the formation of cyclobutane pyrimidine dimers (CPDs) and pyrimidine-(6-4)pyrimidone photoproducts (6-4 PPs), which both can be repaired by nucleotide excision repair (NER). In yeast, CPDs can also be repaired by a CPD specific photolyase that reverses the damage in a light-dependent reaction (photoreactivation, PR). Unrepaired lesions can be tolerated and bypassed by post replication repair (PRR) pathways [35,36]. Deletion of DOT1 results modest sensitivity towards UV and ionizing irradiation, and increased resistance to methylmethane sulfonate (MMS), suggesting that methylation of H3-K79 plays distinct roles in the repair of specific forms of DNA damage [37,38,39,40]. Histone H3 point mutations identified in the vicinity of H3-K79 that have distinct effects on H3-K79 methylation states showed varying degrees of UV-sensitivity and genetic interactions with UV-damage response pathways, suggesting that H3-K79 methylation states may be modulated in response to UV damage via a trans-histone regulatory pathway. In particular, hht2-L70S and hht2-T80A were found to cause no additional UVsensitivity when combined with an NER mutation (rad1) and therefore act within the NER pathway, while hht2-E73D revealed additional UV-sensitivity indicating a role outside of NER [41].
While such genetic analysis has contributed to the identification of functionally important histone domains, many questions remain as to the mechanism by which these domains operate. They may serve as binding sites for proteins involved in silencing and the DNA damage response. On the other hand, residues that map on the histone octamer surface might play a role in establishing contacts between nucleosomes in arrays and higher order structures, thereby affecting nucleosome positioning and stability. We therefore investigated whether mutations that affect silencing and repair (hht2-T80A, hht2-K79E, hht2-L70S, hht2-E73D) do so as a consequence of altered nucleosome arrangements. We show that all mutants maintained nucleosome positions and stability as well as the capacity to repair UV-lesions. However, loss of silencing in the subtelomeric URA3 gene correlated with an altered position of the promoter nucleosome and slightly enhanced repair of subtelomeric chromatin. These results support a role of those residues in the recruitment of silencing proteins and argue against a general role in nucleosome organization.

Results
Histone H3 mutants maintain nucleosome positioning and stability in minichromosomes To investigate whether histone H3 mutations in close proximity to methylatable H3-K79 (hht2-T80A, hht2-K79E, hht2-L70S, and hht2-E73D; hereafter abbreviated hht2 mutants) affect chromatin structure, we used yeast strains in which both genomic loci coding for histones H3 and H4 (HHT1-HHF1 and HHT2-HHF2) were disrupted and replaced with either an HHT2 wild-type or an hht2 mutant allele of H3 on a centromeric plasmid [26]. The relative expression levels of wildtype H3 and the H3 mutants were very similar [41]. To test whether these H3 mutations affect the stability of nucleosomes, nucleosome positioning, and nucleosome-nucleosome contacts, the strains were transformed with circular minichromosomes (YRpFT35 and YRpFT38; Figure 1 B) that were shown to have distinct chromatin structures in strains containing the wild type set of histone genes (S288c) [10]. Both minichromosomes were generated by insertion of one (YRpFT38) or two tandem copies (YRpFT35) of Umid sequences (corresponding to 3.5 nucleosomes of the URA3 coding region) in the TRP1ARS1 circle. Both minichromosomes showed (i) nuclease sensitive regions (NSRs) at the TRP1 promoter (EcoRI site), at the origin of replication (ARS1), and a non-functional NSR at one end of Umid; (ii) four imprecisely positioned nucleosomes on TRP1; and (iii) positioned nucleosomes (I and III, R1, R2, R3, R4) in an untranscribed region. In addition, YRpFT35 revealed a long nuclease-resistant footprint in UmidA and the flanking region consistent with four close-packed nucleosomes forming a ''tetranucleosome''. Different structures on the two Umid sequences in YRpFT35 demonstrate that the DNA sequence does not determine nucleosome positioning in these regions. The presence of positioned and tightly packed nucleosomes and nucleasesensitive regions makes these minichromosomes suitable substrates to test nucleosome positioning, stability, and nucleosome-nucleosome contacts. Moreover, the extrachromosomal nature of those circular minichromosomes allows one to investigate structures independent of chromosomal position effects and to use supercoiling assays in addition to conventional nuclease digestions [42].
Chromatin was digested with increasing amounts of MNase; DNA was purified, separated on a 1% agarose gel, and stained with ethidium bromide. Bulk chromatin from all tested stains displayed clear nucleosomal ladders and a similar accessibility to MNase (Figure 1 A) indicating that the H3 mutations do not affect global chromatin organization.
To investigate the arrangement of the nucleosomes in the minichromosomes, MNase cutting sites were displayed by indirect end labeling from the XbaI site of YRpFT35 and YRpFT38 ( Figure 1 C and D). Cleavage sites in chromatin were compared with those in naked DNA. Regions that were protected from cleavage in chromatin and encompass 140-160 bp were interpreted as positioned nucleosomes [43]. The cutting patterns in YRpFT35 and YRpFT38 were similar in all tested strains and therefore independent of the histone mutation (  Thus, the hht2 mutants did not dramatically affect positions of spaced nucleosomes, nor contacts of tightly packed nucleosomes, nor the nuclease sensitive regions (ARS1 and TRP1 promoter). Moreover, the strong footprints indicate that nucleosomes were not remarkably destabilized.
Since MNase hydrolyses DNA and RNA [44] and since different yeast strains and chromatin preparations might vary in RNA contents, we performed codigestion experiments with MNase [42] to quantitatively compare the stability of chromatin containing mutant H3 histones with wild-type chromatin. Nuclei from HHT2 wild-type cells containing YRpFT35 (MFY15) were mixed and codigested with nuclei from cells expressing a wild type or one of the hht2 alleles and containing YRpFT38. Digestion kinetics was assessed by indirect end labeling with a TRP1 fragment that detects both minichromosomes (Figure 2 A). Vice versa, the experiment was repeated with wild type cells containing the short minichromosome YRpFT38 (MFY57) and mutants containing the long minichromosome YRpFT38 (Figure 2 B). In both co-digestion series, the top bands representing the undigested YRpFT35 and YRpFT38 minichromosomes decreased with similar kinetics. Scans of the top bands of the samples digested with 0, 25, and 50 U/ml MNase did not manifest pronounced differences in chromatin susceptibility to degradation by MNase As an alternative assay to investigate chromatin and nucleosomes stability in vivo, we tested the superhelical density of the circular minichromosome YRpFT35. To this end, DNA was purified from wild type and hht2 mutant cells. Plasmid topoisomers were separated in chloroquine-agarose gels and analyzed by Southern blotting. As shown in Figure 3, the topoisomer distributions of YRpFT35 were similar in all mutants and indistinguishable from the wild type strain. The results confirm that none of the hht2 mutations substantially destabilized nucleosomes nor affected nucleosome density in YRpFT35.
The H3 mutations maintain nucleosome positioning in the coding region of a subtelomeric URA3 gene but alter the position of the promoter nucleosome Genes placed near telomeres in S. cerevisiae are transcriptionally repressed by binding of a silencing complex containing Sir2, Sir3 and Sir4 proteins [45,46,47]. The hht2 mutants used in this study were identified in a screen for loss of silencing of a URA3 reporter gene that was inserted at the telomere on the left arm of chromosome VII (TEL-VII-L) [26]. The results are consistent with the observation that the surface on the nucleosome around K79 serves as a binding site for Sir3 [48,49] and that loss of subtelomeric URA3 silencing in the hht2 mutants is due to an inability of the silencing complex to bind and spread to this region [50]. Alternatively, it is also conceivable that the hht2 mutations affect chromatin structure in subtelomeric regions and/or that loss of silencing proteins affects chromatin structure. In support of this idea, a recent chromatin excision approach revealed differential contributions of H3 and H4 residues on conformational properties of heterochromatin [51].
We therefore investigated chromatin footprints of URA3 at the telomere TEL-VII-L ( Figure 4). The results revealed positioned nucleosomes in the adh4a region and characteristic chromatin structures of the URA3 gene as observed previously in the normal locus in chromosome V, the subtelomeric regions on chromosomes V, and in various minichromosomes, namely a nuclease sensitive promoter (arrow heads) and six positioned nucleosomes (U1 to U6) and a nuclease sensitive 39end (arrow head). Nucleosome U1 is flanked by the TATA box and contains the major transcription start sites [9,12,19,20]. The arrangement of the nucleosomes was similar in all tested strains. Therefore the hht2 mutants do not disrupt chromatin structure near the telomere, which complements the results obtained with the minichromosome.
There was however a significant difference in positioning of the promoter nucleosome U1 between wild type and mutant cells. Nucleosome U1 at the 59 end can occupy two extreme positions (circles a and b), defined by double bands at the 59 end of the gene [12,19]. In HHT2 wild-type cells (MFY15) the nucleosome (U1a) partially protects the cutting site at the TATA box and the lower band of the double bands at the 59 end is more intense (Figure 4 C). On the other hand, all hht2 mutants (MFY16, MFY17, MFY18, and MFY19) showed a stronger cut at the TATA box relative to other sites and a more pronounced upper band of the double bands at the 59 end. This indicates that in cells where URA3 was silenced (HHT2) the nucleosome preferentially adopts position U1a covering the TATA box, whereas in the hht2 mutants, when URA3 is transcribed, this nucleosome was shifted towards the coding region to position U1b exposing the TATAbox. A similar nucleosome shift at the 59 end of a subtelomeric URA3 gene was previously found in sir3D cells [19] suggesting that the rearrangement of the nucleosome at the URA3 promoter depends on binding of silencing proteins to this region. Whether this change of U1 is a cause or consequence of transcription is unknown. However, the fact that only one nucleosome was affected argues against a general role of the H3 mutants in nucleosome stability and nucleosome contacts. Telomere length is not affected by the H3 mutations Telomeres are nucleoprotein structures at the end of chromosomes that protect the chromosome ends from degradation, endto-end-fusion, and play additional roles in genome stability and subtelomeric transcriptional silencing. In S. cerevisiae, telomeric DNA is composed of arrays of heterogeneous TG 1-3 sequences, normally about 300 bp in length. The length of the telomeric DNA repeats, which are generated by the activity of telomerase, is regulated by the action of various factors including histonemodifying enzymes. Telomeres serve as nucleation sites for subtelomeric heterochromatin formation, and elongation of telomeric DNA increases subtelomeric silencing, whereas several mutations that disrupt telomeric silencing also decrease the length of telomeres [52,53,54,55,56].
To investigate whether loss of subtelomeric silencing in hht2 mutants correlates with changes in length of telomeric DNA, we grew yeast cultures in YPD to an A 600 of about 2 and isolated DNA. The DNA was then digested with XhoI, separated on a 1% agarose gel and the length of telomeric DNA was analyzed by Southern blotting with a Y9-specific probe ( Figure 5). Strains with sir3D, sir4D and yku70D mutations revealed telomere shortening as described [55,56], but the hht2 mutants as well as sir2D maintained the normal telomere length. Thus, loss of subtelomeric silencing in the hht2 mutants is not accompanied by significant changes in length of telomeric DNA.

The hht2 mutants show normal accessibility of UV lesions to photolyase and nucleotide excision repair (NER)
Among the mutants used here, hht2-L70S, hht2-E73D, hht2-T80A showed varying degrees of UV-sensitivity and genetic interactions with UV-damage response pathways. In particular, hht2-L70S and hht2-T80A were found to act within the NER pathway, while hht2-E73D revealed increased UV-sensitivity, indicating an additional role outside of NER [41]. Since packaging of DNA in chromatin affects accessibility and repair of UV lesions [57], we analyzed repair of CPDs by NER and photolyase to assess the impact the H3 mutants.
To measure DNA repair of UV lesions, yeast cultures were UVirradiated with 150 J/m 2 to generate about 0.3 CPDs/kb, and either exposed to light for photoreactivation or incubated in the dark for NER. DNA was purified, cut at CPDs with T4endonuclease V and the cutting sites were displayed by indirect endlabelling using alkaline gels and strand specific probes for YRpFT35 and the subtelomeric URA3 region (Figures 6 and 7).
DNA of non-irradiated cells showed the intact restriction fragment (top band) irrespective of T4-endonuclease V treatment (Figures 6 A and 7 A, lanes 1 and 2). DNA of irradiated cells showed the fraction of intact restriction fragments in the mock-treated lanes (2T4 EndoV) and a specific band pattern and a weaker top band when cut at CPDs with T4-endonuclease V (+T4 EndoV). The bands represent the yields and distribution of CPDs along the DNA fragment. Top and bottom strands revealed different patterns demonstrating strand specificity. CPD bands disappeared rapidly when cells were exposed to photoreactivating light (Figures 6 A and  7 A, lanes 3 to 9) due to the combined activity of photolyase and NER (PR+NER). The bands disapeared more slowly by NER alone when cells were incubated in the dark (Figures 6 A and 7 A, lanes 13 to 20, NER). Repair of CPDs was quantified as time-dependent increase of the fraction of intact restriction fragments (Figures 6 B  and 7 B). The results revealed efficient repair of YRpFT35 by both pathways in wild type and all H3 mutants (Figure 6 B). Thus, the hht2 mutations did not generally affect DNA accessibility and repair in the minichromosomes. None of the H3 mutants were defective in NER. Repair was also efficient in the subtelomeric region of chromosome VII containing URA3 (Figure 7 B). In this locus, however, repair by photolyase in wild type cells was slightly reduced compared with repair of the H3 mutants ( Figure 7B, 15 minutes NER+PR) and NER might be mildly enhanced in the hht2-T80A mutant. This result correlates with loss of URA3 silencing and reflects an altered accessibility in compact silenced and open nonsilenced chromatin as shown previously [19]. Taken together, there were only subtle differences in CPD repair between wild type and hht2 mutants, which provides additional in vivo evidence that the mutants do not generally affect DNA accessibility in chromatin.

Discussion
We have studied the histone H3 mutations hht2-T80A, hht2-K79E, hht2-L70S, and hht2-E73D that alter amino acid residues in an area on the nucleosome surface with distinct roles in silencing and the DNA-damage response pathways [25,26,39,41]. We show that those mutants have little effect on chromatin structures and conclude that they rather act by recruitment of proteins involved in silencing and DNA-damage response than by direct or factor mediated nucleosome interactions and positioning.
Structural studies pointed out that histone residues (and their posttranslational modifications) on the solvent exposed nucleosome surface might impact higher order structures by promoting interactions between nucleosomes in chromatin arrays and within condensed chromatin structures [3,4,6,58]. Here, we used a minichromosome system to investigate nucleosome positioning in vivo, since previous studies established that the nucleosome positions in those constructs are not determined by the DNA sequence, but mainly by NFRs acting as boundaries and chromatin folding [9,10]. Moreover, those minichromosomes contained nucleosomes in close proximity, may be in face-to-face contact (tetranucleosomes), and, hence, were considered to be sensitive to mutations affecting the nucleosome surface. However, none of the H3 mutants showed effects on nucleosome positioning in the minichromosomes, nor on Figure 5. Analysis of telomere length in hht2 mutants. The DNA from the indicated strains was isolated, digested with XhoI, separated on a 1% agarose gel and transferred to a Zeta-Probe GT membrane and hybridized with a Y9-specific probe (lanes 2-10). Lane 1 shows a molecular size marker. doi:10.1371/journal.pone.0026210.g005 the supercoil density, the general sensitivity to MNase, or the accessibility to DNA repair enzymes. In addition, none of the H3 mutants tested here revealed an effect on nucleosome repeat length of genomic chromatin. Similarly, a H3-K79R mutant studied previously revealed very little differences in MNase accessibility compared to wildtype H3 [59]. Thus, at this level of analysis, there is no indication that the H3 residues studied so far significantly contribute to the general organization of chromatin. Furthermore, these results are consistent with the structural analysis of methylated H3-K79, which indicates that the effects of this modification are limited to localized structural changes to the nucleosome surface, and do not impact higher-order structures [60].
To test whether the impact of the mutants on silencing was related to changes in chromatin or recruitment of factors, we investigated the subtelomeric region containing URA3 as a reporter gene. With the exception of the promoter nucleosome, the footprints of nucleosomes in the URA3 gene were unchanged in the hht-mutants and sir3-deletion strains and slightly enhanced repair of the whole subtelomeric domain was observed in the mutants compared with the wild type. Those observations are entirely consistent with our previous studies using silencing deficient sir3-deletion strains that fail to recruit the silencing complexes [19]. We take this as an indication that the histone H3 residues act by binding factors rather than by changing chromatin structures at the level of nucleosomes. Our interpretation is consistent with genetic and biochemical evidence supporting direct interactions of the nucleosome core surface surrounding H3-K79 (the LRS surface) with the bromo-adjacent homology (BAH) domain of Sir3 to establish silencing [27,31,33,34].
Several of the H3 mutants (L70S, T80A, E73D, Q76R) showed enhanced UV-sensitivity indicating a role in the DNA-damage response. Epistasis analysis of UV-survival with NER mutants (rad1) indicated that L70S and T80A act within the NER pathway, while E73D and Q76R revealed partially additive effects as shown for dot1 [39,41]. Our measurements of CPD repair by NER in presence and absence of photoreactivation, however, revealed no obvious difference between T80A, E73D and K79E, neither in minichromosomes nor in the subtelomeric region. Hence, neither the CPD accessibility (assayed by photoreactivation) nor the NER capacity appeared to be severely compromised. This observation does not support the hypothesis that histone H3 residues might define a binding site for repair factors [41] facilitating damage recognition and NER. However, deficiencies in NER have been reported for an H3-K79R mutation at silent mating locus HML, suggesting that the role of this domain in NER may be locusspecific and dependent on local chromatin structure [59]. Epistasis analysis of these mutations indicates that this domain additionally operates within other UV-damage response pathways [41], thus it is reasonable to anticipate that repair factors that bind to this domain in H3 may exist in these pathways. Consistent with this, DNA damage checkpoint factor Rad9 has been suggested to bind to methylated H3-K79 in response to double-stranded breaks [61] but additional studies are needed to determine if the interaction is direct or mediated by an additional nucleosome-binding recruit-ment factor, or if such an interaction arises in response to UV damage.

Chromatin analysis by micrococcal nuclease
Yeast cells were grown at 30uC in 5 liters synthetic complete (SC) dropout media lacking tryptophan to an A 600 of about 0.8-1.2 and either a crude nuclei extract was prepared essentially as described [63] and suspended in buffer A (20 mM Tris, pH 8.0; 150 mM NaCl; 5 mM KCl; 1 mM EDTA; 1 mM PMSF) (Figures 1 C and 3) or spheroplasts were prepared as described [64] and suspended in buffer B (20 mM Tris, pH 8.0; 150 mM NaCl; 5 mM KCl; 1 mM EDTA; 0.15% NP40; 1 mM PMSF; 1 mg/ml pepstatin; 1 mg/ml leupeptin) (Figures 1 A, 1 D, and 2). 2 ml aliquots of the crude nuclei extract or 400 ml aliquots of the spheroplasts were supplemented with CaCl 2 to a final concentration of 1 mM and digested with micrococcal nuclease (1-200 U/ ml, Roche Diagnostics) at 37uC for 5 minutes. In case of the crude nuclei extracts, the reactions were terminated by addition of 3 ml 2.56 buffer G2 (2 M guanidine HCl; 75 mM Tris, pH 8.0; 75 mM EDTA; 12.5% Tween-20; 1.25% Triton X-100; 200 mg/ ml RNase A (Sigma); 300 mg/ml Proteinase K (Roche Diagnostics)) and incubation for 2 h at 50uC. Genomic DNA was isolated using Genomic-tips 100/G (QIAGEN), and dissolved in 10 mM Tris, pH 8.0; 1 mM EDTA. In the case of the digested spheroplasts, the reactions were terminated by addition of 40 ml stop solution (10% SDS; 100 mM EDTA) and the DNA was isolated and dissolved in 10 mM Tris, pH 8.0; 1 mM EDTA. Figure 6. Efficient repair of CPDs by photolyase and NER in YRpFT35. Yeast cells containing YRpFT35 were irradiated with 150 J/m 2 of UV light, either put on ice to measure the initial damage (lanes 11 and 12), exposed to photoreactivating light (PR+NER) for 15 to 120 minutes (lanes 3 to 10), or kept in the dark for NER (30 to 240 minutes, lanes 13 to 20). The DNA was purified, cut at CPDs with T4-endonuclease V (T4+, odd lanes) or mock treated without T4-endonuclease V (T42, even lanes). To display the CPD cutting sites in YRpFT35, the DNA was digested with XbaI (X), separated on 1.5% alkaline agarose gels, transferred to Zeta-Probe GT membranes and hybridized with strand specific EcoRI-XbaI probes ( Figure 1 Mapping micrococcal nuclease cutting sites by indirect end labeling The DNA was cut with XbaI or HhaI to map the minichromosomes and adh4D::URA3-TEL-VII-L, respectively, electrophoresed on 1% agarose-TBE (89 mM Tris-borate; 2 mM EDTA; pH 8.3) gels, transferred to Zeta-Probe GT membranes, and hybridized with 32 P-labeled DNA probes [65]. Probes were generated by random hexanucleotide-primed DNA synthesis using a HexaLabel DNA labeling Kit (Fermentas), [a-32 P]CTP (Amersham Biosciences), and short DNA templates as indicated in the Figure legends. MNase digestion patterns were analyzed using a PhosphorImager screen (Amersham Biosciences) and ImageQuant software (Molecular Dynamics).

Plasmid DNA topology analysis
Superhelical density of the minichromosomes YRpFT35 was determined as described [66]. Briefly, yeast cells were grown at 30uC in SC dropout media lacking tryptophan to an A 600 of about 1-1.5, mixed with an equal volume of buffered ethanol/toluene (20 mM Tris, pH 8.0; 95% ethanol; 3% toluene) prechilled to 220uC followed immediately by addition of 0.5 M EDTA to a final concentration of 10 mM. Spheroplasts were prepared by incubation with Zymolyase 100T (Seikagaku Kogyo Co., Ltd.) for 30 minutes at 30uC. The DNA was isolated using Genomic-tip 100/G (QIAGEN), electrophoresed on a 0.75% agarose-TBE gel containing 1 mg/ml chloroquine (Sigma) and transferred to Zeta-Probe GT membranes. YRpFT35 topoisomers were probed with a 32 P-labeled EcoRI-XbaI fragment of the TRP1 gene. Topoisomers were analyzed using a PhosphorImager.

Telomere length analysis
Genomic DNA was prepared from yeast strains grown in YPD to an A 600 of about 2. DNA was digested with XhoI, separated on a 1% agarose-TBE gel, transferred to a Zeta-Probe GT membrane, and hybridized with a 32 P-labeled DNA probe. The probe was generated by random hexanucleotide-primed DNA synthesis as described above using a short Y9 specific DNA template, which was generated by PCR from genomic yeast DNA using the primers Y9-1 (59-TGCCGTGCAACAAACACTAAAT-CAA-39) and Y9-3 (59-CGCTCGAGAAAGTTGGAGTTTT-TCA-39).

UV irradiation and repair by photolyase and nucleotide excision repair
UV irradiation of yeast cultures and repair of UV lesions was done as described previously [67]. Briefly, yeast cultures were grown in 6 liters SC dropout media lacking tryptophan at 30uC to a density of about 1610 7 cells/ml, resuspended in SD (0.67% yeast nitrogen base without amino acids, 2% dextrose) to about 3610 7 cells/ml. Suspensions were irradiated with UV light by use of Sylvania G15T8 germicidal lamps (predominantly 254 nm) at a dose of 150 J/m 2 (measured by an UVX radiometer; UVP Inc., Upland, Calif.) and supplemented with adenine, uracil, and the appropriate amino acids. For photoreactivation and NER, the cell suspension was exposed to photoreactivating light (Sylvania type F15 T8/BLB bulbs, peak emission at 375 nm) at ,1.3 mW/cm 2 (measured by an UVX radiometer with a 365 nm photocell) and 27-30uC. For NER alone an aliquot was incubated in the dark at room temperature. Samples were chilled on ice, genomic DNA was isolated using Genomic-tips 100/G (QIAGEN), and dissolved in 10 mM Tris, pH 8.0; 1 mM EDTA. All steps until lysis of cells were done in yellow light (Sylviania GE Gold fluorescent light) to prevent undesired photoreactivation.

CPD analysis by indirect end labeling
DNA was digested with XbaI or HhaI and nicked at CPDs with T4-endonuclease V (Epicentre) in 50 mM Tris, pH 7.5; 5 mM EDTA or mock treated with the same buffer. The DNA was electrophoresed on 1.5% alkaline agarose gels, transferred to Zeta-Probe GT membranes, and hybridized with 32 P-labelled strand specific DNA probes. Strand-specific probes were generated by primer extension using the DNA fragments described in the Figure legends, [a-32 P]CTP (Amersham Biosciences), and Taq DNA polymerase (Fermentas) for 20 cycles. The membranes were analyzed using a PhosphorImager and ImageQuant software. The CPD content (CPDs/top strand and CPDs/bottom strand) was calculated using the Poisson expression 2ln(RFa/RFb), where RFa and RFb represent the signal intensity of the intact restriction fragment of the T4-endonuclease V and mock-treated DNA, respectively [68]. CPD repair was expressed as percentage with respect to the initial damage (0 min = 100% damage).