DNA binding by the Rev1 BRCT region : implications for biological and structural function

(1)

DNA binding by the Rev1 BRCT region : implications for biological and structural function

Groote, F.H. de

Citation

Groote, F. H. de. (2011, February 9). DNA binding by the Rev1 BRCT region : implications for biological and structural function. Retrieved from https://hdl.handle.net/1887/16451

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/16451

Note: To cite this publication please use the final published version (if applicable).

(2)

53

Chapter II

DNA Binding Properties of the Rev1 BRCT Region

(3)

DNA Binding Properties of the Rev1 BRCT Region

54 Abstract

Rev1 is a eukaryotic DNA polymerase of the Y family involved in translesion synthesis (TLS) at damaged templates, a major damage tolerance pathway that allows DNA replication. Uniquely among the Y family polymerases, the N-terminal part of Rev1, dubbed the BRCA1 C-Terminal homology (BRCT) region, includes a BRCT domain.

While most BRCT domains mediate protein-protein interactions, Rev1 contains a predicted α-helix N-terminal to the BRCT domain, which was shown to endow the BRCT region of human Replication Factor C (RFC) with DNA binding capability. Here, the DNA binding properties of the BRCT region of yeast and mouse Rev1 were studied. The experiments presented here show that, similar to RFC, the BRCT region of Rev1 is specifically capable of binding to a synthetic 5’ phosphorylated primer terminus. This DNA binding strongly depends on the extra α-helix N-terminal to the BRCT domain. Surprisingly, an additional 20 amino acids N-terminal to the predicted α-helix are also required to confer specificity for the 5’ phosphate. Rev1 binding to DNA also appears more complex than that of RFC, exhibiting very slow exchange kinetics. These DNA binding characteristics are discussed in view of the proposed recruitment of Rev1 by 5’ primer termini, downstream of stalled replication forks.

(4)

Chapter II

55 Introduction

Advancing replication forks may stall at DNA lesions due to the inability of the high fidelity polymerases δ and ε to accommodate modified nucleotides in their active sites. If not resolved, these stalled forks can collapse, resulting in chromosomal breakage or inappropriate cell death. Stalled replication forks can be rescued by a DNA damage tolerance pathway called DNA translesion synthesis (TLS) [Waters et al.

2009; Sutton et al. 2001]. In TLS, the high fidelity replicative polymerase (pol) is replaced by a specialized, low fidelity polymerase that synthesizes a short stretch of DNA past the site of damage. The active site of the low fidelity polymerase is generally much more open than that of the replicative polymerases thereby accommodating bulky adducts or abasic sites in the damaged DNA strand. It is thought that in eukaryotes TLS occurs via the action of two polymerases, typically one of the Y family polymerases Rev1, η, ι or κ [Ohmori et al. 2001] and Polζ, a B family polymerase that efficiently extends mismatched primers. Translesion synthesis is a double-edged sword in that it allows the completion of genomic replication and yet is responsible for introducing the majority of mutations [Lawrence, 2004]. It is therefore clear that a high level of regulation must occur to ensure that only the optimal polymerase can access the 3’ terminus and carry out TLS.

Rev1 is a unique member of the Y family polymerases since its catalytic activity is limited to DNA-dependent deoxycytidyl transferase activity [Nelson et al. 1996], in particular opposite abasic sites and a variety of different types of adducted guanines [Zhang et al. 2002]. Rev1 is responsible for the majority of DNA damage-induced mutagenesis and has been shown to be involved in the generation of resistance to the anticancer drug cisplatin [Lin et al. 2006]. However, the polymerase activity is not essential for TLS mediated by Rev1. For example, introduction of a polymerase deficient, but otherwise wild type Rev1 protein, did not lead to a change in

(5)

56

mutagenesis rates but did cause changes in the mutation spectrum [Otsuka et al.

2005; Ross et al. 2005]. Therefore, it is currently thought that the polymerase activity of Rev1 is only required for certain types of DNA damage [Waters et al.

2009]. Conversely, mutations outside of the polymerase domain have dramatic effects on bypass of a broad range of DNA damage [Otsuka et al. 2005; Nelson et al.

2000]. Rev1 can function as a binding platform for other Y family polymerases and Polζ [Guo et al. 2003] and presumably it is this function that is critical for Rev1- mediated TLS. This suggests an additional, regulatory, role of Rev1 in TLS in the selection of other Y family polymerases and recruitment of Polζ to template lesions.

In addition to the deoxycytidyl transferase domain, mammalian Rev1 consists of three other functional domains: an N-terminal BRCA1 C-terminal homology (BRCT) domain, two ubiquitin binding motifs (UBM) in the C-terminal region of the protein and a C-terminal domain which binds to other Y family polymerases and/or polymerase ζ [Waters et al. 2009]. The UBMs mediate binding of Rev1 to monoubiquitinated PCNA which is produced in response to DNA damaging reagents such as UV light. This association with ubiquitinated PCNA is responsible for the localization of Rev1 into repair foci upon DNA damage induction. The BRCT domain is required for DNA damage-induced mutagenesis in yeast and vertebrate cells [Otsuka et al. 2005; Lemontt, 1971]. Recently, it has been shown that the Rev1 BRCT domain is required for efficient binding of Rev1 to the processivity clamp PCNA [Guo et al.

2006] and the Rev7 subunit of polζ [D’Souza et al. 2006]. In addition, the BRCT domain is responsible for localization of Rev1 to sites of (presumably stalled) DNA replication in the absence of induced damage [Tissier et al. 2004]. Interestingly, the BRCT domain is also required for recruitment of Rev1 to sites of double strand DNA breaks [Hirano and sugimoto, 2006]. Finally, mutations in the BRCT domain strongly reduce Rev1-dependent mutagenesis in yeast and mammals, although the domain is not essential for survival [Otsuka et al. 2005; Lemontt, 1971, Jansen et al. 2005].

(6)

Chapter II

57

Despite the fact that the BRCT domain of Rev1 clearly plays an important role in the regulation of TLS, a molecular mechanism of the function of the BRCT domain remains elusive. BRCT domains form a wide superfamily and are mainly found in proteins involved in DNA repair or cell cycle checkpoints [Callebaut and Mornon, 1997; Bork et al. 1997]. BRCT domains consist of 90-100 amino acids and are defined by their conserved fold, which consists of a four stranded, parallel β-sheet surrounded by three α helices. The majority of BRCT domains mediate a range of protein-protein interactions including homo-dimerization, hetero-dimerization with and without other BRCT domains and phosphopeptide binding (reviewed in [Glover et al. 2004]). In addition, some BRCT domains have clearly been shown to bind DNA [Allen et al. 1998].

Recently the DNA binding properties of the BRCT region of the large subunit of human Replication Factor C (RFC) have been characterized. It was found that DNA binding does not only depend on the BRCT domain itself, but also on extra amino acids flanking it [Kobayashi et al. 2006]. In the same study, it was noted that Rev1 exhibited sequence similarity with the BRCT region of RFC. Here it is shown that Rev1 from multiple species displays comparable, but more complex, DNA binding properties with respect to a 5’ phosphorylated, double stranded DNA substrate, and the requirement for amino acids outside of the conserved BRCT domain.

(7)

58 Methods

Expression and purification of Rev1 fragments

BL21 pLySs cells (Novagen), were transformed with the pET20bmRev1(1-131) or pET20bscRev1(1-251) plasmids and cultured in LB medium containing ampicillin (50μg/ml) and chloramphenicol (34μg/ml) at 25 °C. Protein expression was induced at an OD600 of 0.6 with 1 mM IPTG and growth was continued overnight at 18 °C.

Cells were collected by centrifugation at 5000 rpm for 20 minutes. The cell pellets were resuspended in binding buffer containing 20 mM Tris-HCl (pH 7.5), 5 mM Imidazole and 500 mM NaCl, and stored at -80 °C. Cells were lysed by thawing on ice, and passed through a French press at 1500 psig twice. The recovered lysate was centrifuged at 35000 rpm and 4 °C for 45 minutes using a Beckman Ti35 rotor. The supernatant was applied to a 5ml HisTrap HP Ni²⁺ chelating column (GE healthcare) equilibrated in the manufacturer’s suggested binding buffer supplemented with PMSF. The column was washed with 20 mM Tris-HCl (pH 7.9); 60 mM Imidazole and 500 mM NaCl and protein was eluted with 20 mM Tris-HCl (pH 7.9); 500 mM Imidazole and 500 mM NaCl. The eluted fraction was diluted to a final concentration of 250 mM NaCl by adding 25 mM Tris-Hcl (pH 7.9), and applied to a 5 ml SP ion exchange column (GE healthcare) equilibrated in 25 mM HEPES (pH 7.5); 50 mM NaCl and 1 mM DTT, followed by elution using a linear gradient of 50 mM – 1 M NaCl in 20 column volumes. A DNA peak was eluted at 455 mM and a protein peak was eluted at 675 mM NaCl. Fractions containing the protein peak were pooled, and loaded on a 116 ml (1.6 x 60 cm) Superose 12 (GE Healthcare) column equilibrated in 25 mM HEPES (pH 7.5); 50 mM NaCl and 1 mM DTT at 1 ml/min. Functional mRev1(1-131) eluted at 101 ml. An identical procedure was used to express and purify the N-terminally truncated mRev1 proteins and scRev1(1-251) which eluted at a volume of 77 ml.

(8)

Chapter II

59 Construction of Rev1 expression plasmids

DNA fragments encoding amino acids 1 to 131 (mRev1(1-131)), 21 to 131 (mRev1(21-131)) and 41 to 131 (mRev1(41-131)) of mouse Rev1 were amplified by PCR and cloned into a pET20b plasmid for bacterial expression. A DNA fragment encoding amino acids 1 to 251 of yeast Rev1 was PCR amplified and inserted into the bacterial expression vector pET20b in an identical manner. All PCR products were digested with NdeI and XhoI and ligated with pET20b, treated with the same restriction enzymes, to give respectively pET20bmRev1(1-131), pET20bmRev1(21- 131), pET20bmRev1(41-131) and pET20bscRev1(1-251). Each construct contained a non-encoded Met at the N-terminus and a 6 his affinity tag at the C-terminus. All constructs were verified by sequence analysis.

Preparation of oligonucleotides

All oligonucleotides were commercially synthesized (Microsynth), and are presented in table 1. For radiolabeling, 2 pmol of the selected DNA oligonucleotide was treated with T4 polynucleotide kinase (New England Biolabs) at 37 ^oC with 10 μCi (6000 Ci/mmol) γ³²P-ATP as a substrate. To ensure complete phosphorylation of the oligonucleotide, unlabeled ATP was added to a final concentration of 50 μM after 90 minutes of incubation and the reaction was continued for 30 minutes. For 3’

radiolabeling 2 pmol of the selected DNA oligonucleotide was treated with Klenow fragment (New England Biolabs) at 37 ^oC with 10 µCi (6,000 Ci/mmol) a ³²P-dCTP as a substrate. To ensure a high percentage of base pairing, the hairpin oligonucleotides, were denatured at 100 °C, and slowly cooled to room temperature. Non reacted radioactive nucleotides were removed using Sephadex G25 spin columns (GE Healthcare) following the standard protocol.

(9)

60

(Table 2.1). DNA sequences and secondary structures used in this study. Shaded residues indicate the hairpin structure, and underlined residues indicate dsDNA. The DNA sequence is presented 5’ to 3’. All 5’ phosphate modifications were introduced enzymatically. The 3’phosphate modification of 3’SHP was introduced when synthesized commercially.

Name Sequence Structure

GAP0 ACTCCACCCTCATGGGTGGAGTCGTGTGAGC TCATGCTCACACG

GAP3 ACTCCACCCTCATGGGTGGAGTCTTGTGCTG GTGTCATCACCAGCAC

GAP10 ACTCCACCCTCATGGGTGGAGTCTTGTAATGA GTGTGGTCGTCATCGACCACAC

5’SHP₁ CTCGAGGTCGTCATCGACCTCGAGATCA

5’SHP₂ CTCGAGGTCGTCATCGACCTCGAG

3’SHP ACTGCTCGAGGTCGTCATCGACCTCGAG

ssDNA TGGGGTGGGGT

dsDNA

TGCAGATTGCGCAATCTGCA (self annealing)

Electrophoretic Mobility Shift Assay (EMSA)

DNA binding was detectedusing a gel retardation assay. The indicated amount of protein was diluted in a buffer consisting of 10 mM HEPES, pH7.8, 2 mM MgCl2, 0.1 mM EDTA, 100 µg/ml bovine serum albumin,15% glycerol, (0-0.8) µg/ml poly(dI-dC) (Boehringer Mannheim) and 2 mM dithiothreitol in a total volume of 15 µl. 20 fmol of the ³²P-labelled GAP3 oligonucleotide (see Table 1) in 5 µl was added, incubated on ice for 30 minutes, and applied to a non-denaturing 8% Tris-glycine acrylamide

(10)

Chapter II

61

gel containing 2% glycerol. Electrophoresis was performed at 80 V for 20 minutes and subsequently 120 V for 40 minutes at 4 °C in 25 mM Tris-HCl pH 8.5, 200 mM glycine. The gel was dried and radioactivity was detected using X-Ray film (BioMax, Kodak).

Quantification of DNA binding

Radioactivity was detected using a phosphorimager (Biorad). For each reaction, the amount of protein-DNA complex was calculated as a percentage: percent bound = ([counts in shifted DNA]/[total counts per lane])*100. For competition binding experiments The competition efficiency was normalized and calculated from the percent bound by the following formula: (1 – [percent bound with specific competitor]/[percent bound without competitor]). A competition efficiency of 1 indicates that all of the ³²P labeled oligonucleotide is replaced by the competitor in the protein-DNA complex.

Results.

Construct formation and purification of proteins.

The BRCT domain of mouse Rev1 encompasses amino acid residues 52 to 119, whereas the analogous residues in S. cerevisiae Rev1 are 169 to 247. In both Rev1 proteins there is a region with unknown function N-terminal to the BRCT domain.

Interestingly, this region may contain an α-helix as predicted by the PSIPRED secondary structure prediction server (Figure 2.1A) [McGuffin et al. 2000]. A similar configuration for the BRCT region of RFC p140 was recently experimentally determined, where it was demonstrated that the α-helix N-terminal to the BRCT domain of RFC p140 was required for DNA binding [Kobayashi et al. 2006]. In those studies it was speculated that the Rev1 BRCT region can also bind DNA. To investigate this possibility, constructs expressing the complete BRCT region of mouse

(11)

62

(Figure 2.1) (A) Amino acid sequence alignment of Rev1 and RFC p140 sequences generated by ClustalW. Residues with greater than 90% identity are shaded in black, 75% identity in dark grey and 50% identity in light grey. The secondary structure is indicated below the alignment. For the first 44 amino acids of mmRev1 and hsRev1 and the full scRev1 sequence the secondary structure was predicted using PSIPRED, for the other sequences the secondary structure is based on experimental data. (B) Domain composition of mmRev1, showing the BRCT domain (BRCT), the deoxycytidyl transferase domain (DTD), ubiquitin binding motifs (UBM) and the C-terminal domain (C-term).The details of the BRCT region used in this study, including the position of the N-terminal α-helix, are shown below. (C) Domain composition of the analogous region of scRev1 with details of the BRCT region using in this study.

and yeast Rev1 were generated, including the N-terminal α-helix and the BRCT domain, fused to a C-terminal his6-tag, were generated resulting in mRev1(1-131) and scRev1(1-251), respectively (Figures 2.1B, 2.1C). In addition, two N-terminal

(12)

Chapter II

63

truncations in the mouse Rev1 BRCT region were made based on the predicted secondary structure: (I) deletion of N-terminal residues up to the α-helix, resulting in mRev1(21-131), and (II) deletion of all amino acids N-terminal to the BRCT domain, resulting in mRev1(41-131) (Figure2. 1B). All four proteins were expressed in E. coli and subsequently purified using nickel affinity chromatography, ion exchange chromatography and gel filtration chromatography.

To study DNA binding of the complete BRCT region of mouse Rev1, increasing amounts of protein were titrated into a ³²P-labelled, double hairpin oligonucleotide containing a 3 nucleotide (nt) single stranded gap (5’P GAP3, Table 2.1).

Subsequently, DNA binding was analyzed using a gel retardation assay (EMSA). The oligonucleotide was based on the optimal substrate for structure-specific binding of the RFC p140 BRCT region but included an additional 3’ terminus to account for all possible common DNA structures. Migration of labeled 5’P GAP3 was retarded in the presence of mRev1(1-131) indicating that this protein binds the 5’P GAP3 oligonucleotide (Figure 2.2A). Quantitation of the DNA binding yielded a KD of 160 nM (Figure 2.2B), in comparison to approximately 10 nM for RFC p140(375-480) [Kobayashi et al. 2006]. However, since the EMSA is a non-equilibrium assay this value should be viewed as approximate. Similar results were found for the BRCT region of yeast Rev1, however, with a lower affinity (Figure 2.2C). To exclude the possibility that Rev1 BRCT binding to 5’P GAP3 reflects a general DNA binding property of BRCT regions, an unrelated BRCT domain from human BRCA1 (aa1646- 1863) was titrated into a similarly labeled DNA. Even at protein to DNA ratios up to 2000:1 no DNA binding was observed (not shown). Furthermore, other biologically relevant in vitro DNA binding has been shown at similarly high ratios [Ma et al.

2004]. For subsequent experiments aimed at delineating the specificity of the Rev1- DNA interaction the mouse Rev1 protein was used.

(13)

64

(Figure 2.2). DNA binding by the mouse and yeast Rev1 BRCT region. (A) mRev1(1-131) was titrated into a constant amount of ³²P 5’ labeled GAP3 oligonucleotide (20 fmol) at the indicated ratios. The position of the DNA-protein complex is denoted by ◄, and free DNA by

◄. (B) The binding data in A were quantitated and fit to a single site binding curve to yield a K_D of approximately 160 nM. (C) scRev1(1-251) was titrated into a constant amount of ³²P labeled GAP3 oligonucleotide (20 fmol) at the indicated ratios. DNA binding by the scRev1 BRCT region. Lane marked “-” no protein, lane marked “mR” 6 pmol mRev1(1-131).

Specific DNA binding requires amino acid residues N-terminal to the conserved BRCT domain

In a previous study of the BRCT region of RFC p140, DNA binding was dependent on a stretch of 28 residues N-terminal to the conserved BRCT domain [Kobayashi et al.

2006]. To investigate whether the predicted α-helix of mRev1, N-terminal to the BRCT domain, plays a role in DNA binding, mRev1(1-131), mRev1(21-131) and mRev1(41-131) were assayed using the EMSA and the 5’P GAP3 substrate (Figure 2.3). At a protein:DNA ratio where mRev1(1-131) shifts nearly 100% of the input

(14)

Chapter II

65

(Figure 2.3). The effect of N-terminal truncation on the DNA binding capability of the mouse Rev1 BRCT region. 16 pmol of the indicated protein was added to 20 fmol of ³²P 5’ labeled GAP3 oligonucleotide, shown in the first lane of each gel. The contribution of non- specific dsDNA binding to complex formation was visualized by adding increasing amounts of poly(dI-dC) (respectively 0.23, 0.47, 0.80 µg/ml) during incubation, where 1 µg/ml corresponds to a nucleotide ratio of 3750:1 with respect to labeled DNA substrate. DNA- protein complex is denoted by ◄ and free DNA by ◄. Proteins are (A) mRev(1-131), (B) mRev(21-131) and (C) mRev(41-131) shown schematically in each panel.

DNA, mRev1(21-131), which still contains the predicted N-terminal α-helix, bears considerably less DNA binding activity (compare Figure 2.3A lane 1 with Figure 2.3B lane 1). Additional removal of the N-terminal α-helix resulted in very weak DNA binding (compare Figure 2.3A lane 1 with Figure 3.3C lane 1). To investigate DNA structure-specific versus nonspecific dsDNA binding, a competition experiment was performed with poly(dI-dC) using different nucleotide ratios of up to 3000:1 of poly(dI-dC):5’P GAP3. Addition of poly(dI-dC) has a moderate effect on DNA binding by mRev1(1-131) (Figure 2.3A) suggesting that DNA binding is in part due to non- specific dsDNA binding. DNA binding by both the mRev1(21-131) and mRev1(41-131) proteins was severely disrupted by the addition of poly(dI-dC) suggesting that the majority of DNA binding by these two proteins is non-specific to the DNA backbone (Figure 2.3B-C). Taken together this data suggests that the complete N-terminal

(15)

66

(Figure 2.4) Determination of structural features of the DNA substrate required for binding.

The indicated unlabeled 5’ phosphorylated oligonucleotide was premixed with 20 fmol of ³²P 5’ labeled GAP3 oligonucleotide at the indicated ratios and incubated with 6 pmol mRev1(1- 131). The competition efficiency was determined by comparing binding of GAP3 in the presence of competitor to that in the absence of competitor. A competition efficiency of 1 indicates that all of the GAP3 DNA is displaced by the competitor DNA in the complex.

Unlabeled substrates are respectively: 5’P GAP3 and 5’P GAP10 (A), 5’P GAP3 and 5’P GAP10 (B), 5’P GAP3 and 5’P SHP1 (C) and 5’P SHP₁ and 5’PSHP₂ (D).

BRCT region of mouse Rev1 plays a role in DNA binding which involves unique structural features of the 5’P GAP3 substrate.

Structural Features of DNA Required for High Affinity Binding by mRev1(1-131) To investigate the substrate specificity of mRev1(1-131), competition between ³²P labeled GAP3 and various unlabeled DNA oligonucleotides for mRev1(1-131) binding was determined using the EMSA. In this experiment, the labeled GAP3 and

(16)

Chapter II

67

unlabeled competing oligonucleotides were mixed prior to addition of mRev1(1- 131). DNA binding was quantitated using a phosphorimager and presented in graphic form as competition efficiency where a score of 1.0 means complete abolition of binding to the labeled substrate (Figure 2.4, Figure 2.5). The effect of the length of ssDNA in a gapped substrate was studied, by competing ³²P labeled 5’P GAP3 with 5’P GAP10, a 5’ phosphorylated oligonucleotide containing a 10 nt ssDNA gap. 5’P GAP10 exhibited similar competition efficiency to 5’P GAP3 itself, indicating that a ssDNA region longer than 3 nt has no stimulating effect on binding by mRev1(1-131) (Figure 2.4A). In contrast, 5’P GAP0, a 5’ phosphorylated oligonucleotide containing a single strand nick, competed significantly less efficiently for binding mRev1(1-131) than 5’P Gap3, suggesting that a stretch of ssDNA is important for mRev1(1-131) binding (Figure 2.4B). Competition with a 5’

phosphorylated, single hairpin with a 3’ ssDNA overhang (5’P 5’SHP1) is not quite as efficient as GAP3 itself suggesting a minor effect of the 3’ dsDNA terminus of GAP3 (Figure 2.4C). The competition efficiency is approximately 10 fold reduced using a substrate containing a hairpin and a blunt end (5’SHP2) suggesting a strong requirement for at least 3 nt of the template strand (Figure 2.4D).

In order to probe more subtle requirements for DNA binding of mRev1(1-131), a series of substrates were investigated, that were substantially less efficient competitors. Note that to achieve reasonable levels of competition it was necessary to titrate these substrates to a roughly 10 fold greater ratio to 5’P GAP3 (Figure 2.5) than those depicted in Figure 2.4. Non-phosphorylated (5’OH) GAP3 competes very poorly with the 5’phosphorylated version (Figure 2.5A), clearly demonstrating that binding of mRev1(1-131) to gapped substrates strongly depends on a 5’ phosphate.

Consistent with the results of the poly(dI-dC) titration, addition of blunt ended, non- phosphorylated 20 base pair (bp) dsDNA competed, albeit weakly, with GAP3 confirming a moderate affinity of mRev1(1-131) for dsDNA (Figure 2.5B). The

(17)

68

(Figure 2.5) Determination of secondary structural features of the DNA substrate required for binding. The indicated unlabeled 5’ phosphorylated oligonucleotide was premixed with 20 fmol of ³²P 5’ labeled GAP3 oligonucleotide at the indicated ratios and incubated with 6 pmol mRev1(1-131). The competition efficiency was determined by comparing binding of GAP3 in the presence of competitor to that in the absence of competitor. A competition efficiency of 1 indicates that all of the GAP3 DNA is displaced by the competitor DNA in the complex.

Unlabeled substrates are respectively: 5’OH GAP3 and 5’OH GAP3 (A), 5’P GAP3, ssDNA and 5’P ssDNA B), and 5’P GAP3, 3’OH 3’SHP and 3’P SH1 (C).

competition efficiency of the non-phosphorylated 20 bp oligonucleotide was greater than that of non-phosphorylated GAP3 while a non-phosphorylated 3’SHP oligonucleotide with 9 bp failed to elicit any competition (Figure 2.5B,C). Together this data suggests a requirement for more than 9 bp for non-specific binding to dsDNA and that the dsDNA should be contiguous for maximal binding. A 3’

phosphorylated single hairpin (3’SHP) also competed with GAP3 for mRev1(1-131) binding, but somewhat less efficiently than the blunt ended dsDNA, indicating that

(18)

Chapter II

69

mRev1(1-131) can bind the phosphorylated 3’terminus of dsDNA, albeit extremely weakly (Figure 2.5C). Neither non-phosphorylated ssDNA, nor 5’ phosphorylated ssDNA resulted in detectable competition with the labeled GAP3 substrate, even at a 50:1 ratio (Figure 2.5D). Together these results strongly suggest that the BRCT region of mouse Rev1 preferentially binds to the recessed, phosphorylated 5’ end of a primer template junction (5’ primer terminus).

The Rev1 BRCT region – DNA complex has a long lifetime

Preliminary competition binding experiments suggested that order of addition of the DNA substrate was important. Specifically, if the unlabeled competitor was added after mRev1(1-131) was mixed with labeled GAP3, the extent of competition was sensitive to the length of the incubation period. These data suggested the possibility of slow exchange kinetics. This question was addressed in two different ways. First, a time course analysis of competition binding was performed using the EMSA (Figure 2.6A). mRev1(1-131) was pre-incubated with unlabeled GAP3 for 30 minutes, conditions under which equilibrium binding is established. Subsequently, the complex was challenged with ³²P labeled GAP3 for various periods of time. Figure 2.6A clearly shows that at short time periods only a small amount of the unlabeled DNA was displaced by the labeled DNA. However, after 30 minutes the system had reached equilibrium since the amount of labeled molecules in protein-DNA complexes was equal to that when ³²P labeled GAP3 was pre-mixed with unlabeled DNA (Figure 2.6A, lanes 5 and 6). Second, the sensitivity of the 5’ ³²P to phosphatase treatment was investigated at various times after mixing the protein and DNA.

Naked 5’ ³²P GAP3 is nearly 100% dephosphorylated within time required for the DNA to enter the gel in this assay (Figure 2.6B). In contrast, after preincubation with mRev1(1-131), phosphorylated GAP3 DNA is still observable even after 15 minutes of phosphatase treatment. A similar experiment was performed using mRev1(21- 131) under conditions where approximately 20% of the 5’ ³²P GAP3 was bound. No

(19)

70

(Figure 2.6) mRev1(1-131) releases 5’ phosphorylated DNA with slow kinetics. (A) Time- dependent binding of ³²P GAP3 by mRev1(1-131). 1 - no protein added. 2 mRev1(1-131):³²P GAP3-complex at a ratio of 400:1. Lanes 3 to 5 - ³²P GAP3 was added for 5’, 15’ or 30’

respectively at an equimolar ratio to unlabeled GAP3 that was preincubated for 30’ with mRev(1-131). 6 - equimolar ratio of labeled and unlabeled GAP3 were both preincubated with mRev1(1-131) for 30’. (B) DNA binding by mRev1(1-131) protects the 5’ phosphate from phosphatase treatment. Rev1(1-131), or mock assay without protein, was incubated with ³²P GAP3 at a ratio of 400:1 for 30 minutes. Subsequently, calf intestinal phosphatase was added for the indicated time before running on gel. The control reactions show that phosphatase treatment was complete before the DNA enters the gel. DNA-protein complex is denoted by

◄ and free DNA by ◄.

protection from phosphatase could be observed, further suggesting that the first 20 are required for structure-specific DNA binding (not shown).

Other possible binding partners of the Rev1 BRCT region

To study a possible direct physical interaction between PCNA and the Rev1 BRCT region, purified PCNA was immobilized on MagnaBind beads, and incubated with mRev1(1-131) (Figure 2.7A). The immobilized protein was pulled down magnetically and beads were washed with buffer. The immobilized protein was pulled down again, and resuspended in denaturing buffer. Samples were subjected to SDS-PAGE

(20)

Chapter II

71

(Figure 2.7) Interaction between mRev1(1-131) and PCNA. (A) Immobilized (I.M.) PCNA was used to pull down mRev1(1-131) and vice versa. (B). Lane 1: purified PCNA; Lane 2: purified mRev1(1-131). Lane 3 mix of both proteins before pulldown; Lane 4: buffer wash; Lane 5 final pulldown; lanes 6-8: Control experiment without immobilized protein.

and Western Blotting using mouse anti-His6 as primary antibody (Figure 2.7A).

MagnaBind beads without protein were used as a control. mRev1(1-131) is still visible in the wash fractions but not in the eluted fractions To eliminate the possibility that immobilization blocks the possible interaction between PCNA and Rev1, the experiment was repeated with immobilized mRev1(1- 131) (Figure 2.7B). Here, PCNA is visible in the wash fraction, but not in the elution fraction which demonstrates that there is no physical interaction between mRev1(1- 131) and PCNA.

It has been reported that the yeast BRCT domain is able to bind a phosphorylated peptide. To see if the mouse Rev1 BRCT region has similar binding properties a peptide library was designed to compete with DNA binding. The library consisted of 8000 peptides of the following sequence ISRSTSSZXXNKQTKB, divided into 20 pools for each possibile amino acid Z, each with all 400 possible amino acid combinations XX. In a competition binding experiment, a series of respectively 2.24, 1.12, 0.56,

(21)

72

(Figure 2.8) Competition binding study of a phospho-Serine peptide library versus DNA. Lane 1: Free ³²P5’P GAP3; Lane 2: 2.0 pmol mRev1(1-131) incubated with 40 fmol ³²P GAP3 in a ratio of 50:1. Lane 3-6: Similar to lane 2 but in the presence of a peptide library. Respectively 2.24, 1.12, 0.56 and 0.28 pmol of each peptide was present in lanes 3-6. A schematic representation of the library is shown at the left.

and 0.28 pmol of each possible peptide was competed with 60 fmol of ³²P labeled GAP3 to bind to 2.0 pmol mRev1(1-131) (Figure 2.8). Even at a ratio of 1.12 times the amount of peptide over protein, no difference in the amount of shifted DNA was observed. This suggests that mRev1(1-131) has no, or very low affinity for the phosphorylated peptide in comparison to DNA binding.

Discussion

DNA binding by the Rev1 BRCT region

Clearly, the present study shows that the BRCT domain containing region of mammalian and yeast Rev1 proteins binds DNA. The conservation of DNA binding across two such distantly related species argues for its biological relevance.

Apparently there are two modes of binding, a non-specific interaction with dsDNA (Figure 2.9A) and a highly specific interaction with a recessed, 5’ phosphorylated primer terminus (Figure 2.9B). In mouse Rev1, the conserved BRCT domain is sufficient for weak, non-specific dsDNA binding that is increased by the presence of aa’s 12-31 residues N-terminal to the BRCT domain, which contain a predicted α-

(22)

Chapter II

73

(Figure 2.9) Model of DNA binding by the Rev1 BRCT region. (A) Non-specific interactions with dsDNA contribute to a fast sliding along the DNA, likely via interaction with the phosphate backbone, while the N-terminal region of the protein remains unfolded. (B) Once the 5’

Phosphate is bound, the N-terminal region undergoes a structural change to form a protein- DNA complex with a substantially longer lifetime.

helix. However, in contrast to the BRCT region of the large (p140) subunit of RFC [Kobayashi et al. 2006], the protein consisting of the predicted N-terminal α-helix plus the BRCT domain is not sufficient for high affinity, 5’ phosphate specific binding.

An additional 20 aa’s N-terminal to the predicted α-helix are required for structure specific, high affinity binding to DNA, which we refer to as the 20NT region (Figure 2.9). Preliminary NMR analysis of the different mouse Rev1 proteins used in this study suggests that the 20NT region is completely unfolded (not shown), an observation that is consistent with the PSIPRED prediction of the total absence of any secondary structure.

(23)

74

DNA binding is not uncommon for BRCT domains. In addition to the previous work on the BRCT region of RFC p140, the BRCT domain has been implicated in direct DNA binding of X family polymerases [Allen et al. 1998] and of bacterial NAD(+) dependent ligases [Feng et al. 2004; Wilkinson et al. 2005]. The BRCT domain of the bacterial NAD(+) dependent ligases binds both dsDNA and ssDNA while the specificity of both the Rev1 and RFC p140 BRCT region is clearly for recessed, phosphorylated 5’ primer termini of dsDNA. However, amongst these various DNA binding proteins, Rev1 is the only one with the slow binding kinetics that are observed here. The slow off rate, combined with the relatively high affinity, results in a protein-DNA complex with a half life of about 5 minutes that sterically blocks access to the 5’ phosphate. The measured approximate affinity and koff suggest a kon

on the order of 10⁴ per second, five orders of magnitude slower than diffusion limited binding. A possible explanation for the slow on rate is that high affinity binding to the 5’ phosphate requires folding of the 20NT which could be rate limiting.

Other possible binding partners of the Rev1 BRCT region.

A non-physiological 3’ phosphorylated oligonucleotide has 20-100 fold lower affinity than a 5’ phosphorylated oligonucleotide, but nonetheless is able to bind mRev(1- 131). This weak residual affinity is likely due to direct interaction between the phosphate and the protein since the non-phosphorylated oligonucleotide could not bind at all. In a previous report, a number of BRCT domains from a variety of proteins, including Rev1, were shown to bind phosphopeptides [Yu et al. 2003]. Of the BRCT phosphopeptide interactions that have been confirmed by structural biology, all involve tandem repeats of the BRCT domains, which is readily understood since the phosphopeptide binding site is formed by the BRCT interface.

Thus it is possible that the phosphopeptide binding by yeast Rev1 BRCT simply reflected the residual binding of the conserved BRCT domain to phosphate, similar

(24)

Chapter II

75

to 3’ phosphorylated oligonucleotide binding. However, since yeast Rev1 contains a much larger N-terminal region than the mammalian homolog, it remains possible that a nonconserved domain cryptically resides in this region and provides phosphopeptide specific binding. Competition DNA binding assays were carried out with mRev1(1-131) using a phosphopeptide library, similar to that reported earlier [Yu et al. 2003], but no effect on DNA binding was observed. NMR analysis of inorganic phosphate binding to mouse Rev1(1-131) revealed only subtle changes in the spectrum of the protein suggesting at best an extremely weak interaction (not shown). From this the conclusion is drawn that, at least for the mammalian protein, there is no phosphopeptide binding. Although it seems likely that mouse Rev1(1- 131) cannot bind phosphopeptides, that does not per se rule out protein binding. It is also important to note that DNA binding and protein binding within the mRev1(1- 131) protein are not necessarily mutually exclusive. A number of laboratories have reported protein-protein interactions involving the BRCT region of Rev1. Among the potentially most relevant interactions is the binding to Rev7 [D’Souza et al. 2006], a subunit of polζ, and PCNA [Guo et al. 2006]. However, a direct physical interaction between the BRCT region of Rev1 and Rev7 or PCNA has not yet been shown. It is important to consider the possibility that co-immunoprecipitation occurs via common binding to DNA present in these assays when interpreting these results.

During the purification of the various Rev1 proteins used in this study, it was observed that significant amounts of DNA copurify during metal affinity and size exclusion chromatography, which is separated from the protein only after cation exchange chromatography. Rev1 appears to bind PCNA in both a ubiquitin independent and a ubiquitin dependent manner [Guo et al. 2006]. For mammalian Rev1, binding to non-ubiquitinated PCNA is dependent on the first 240 aa, which contain the BRCT region [Guo et al. 2006]. Presently it is not possible to say whether this interaction is mediated by common DNA binding, by a BRCT-PCNA interaction or by a non-canonical PCNA interacting peptide as found in other Y family polymerases

(25)

76

[Hishiki et al. 2008]. Attempts to pull down purified PCNA with purified, immobilized mRev1(1-131) and vice versa were not successful. Interestingly, published data suggests that Rev1 interaction with non-ubiquitinated PCNA may enable recruitment into “replication factories” during S phase of cells that have not been treated with DNA damaging reagents [Guo et al. 2006].

In conclusion, a bona fide interaction of the Rev1 BRCT region with a phosphorylated 5’ primer terminus has been presented here. Comparable to DNA binding by the RFC BRCT region, amino acids N-terminal to the BRCT domain are essential for DNA binding. In addition to RFC, 20 amino acids at the N-terminus (defined as the 20NT) are predicted to be unstructured. Most likely the 20NT undergoes a change in confirmation upon DNA binding. 5’ termini are only present in Okazaki fragment and in postreplicative gaps [Lopes et al. 2006], and it has been proposed to perform a critical role in organizing TLS events [Jansen et al. 2007]. Rev1, including the BRCT domain, could be an important mediator of these events [Jansen et al. 2005].

Furthermore, single stranded DNA could stimulate the binding of either the 3’ or the 5’ primer terminus by Rev1 [Masuda et al. 2006]. Further analysis of the binding properties of the Rev1 BRCT region should give more insight into how the BRCT region contributes to the function of Rev1 at the 5’ terminus of a postreplicative gap.

(26)

Chapter II

77 References

Allen BL, Uhlmann F, Gaur LK, Mulder BA, Posey KL, Jones LB, Hardin SH. DNA recognition properties of the N- terminal DNA binding domain within the large subunit of replication factor C. Nucleic Acids Res. 1998 Sep 1;26(17):3877-82.

Bork P, Hofmann K, Bucher P, Neuwald AF, Altschul SF, Koonin EV. A superfamily of conserved domains in DNA damage-responsive cell cycle checkpoint proteins. FASEB J. 1997 Jan;11(1):68-76.

Callebaut I, Mornon JP. From BRCA1 to RAP1: a widespread BRCT module closely associated with DNA repair. FEBS Lett. 1997 Jan 2;400(1):25-30.

D'Souza S, Walker GC. Novel role for the C terminus of Saccharomyces cerevisiae Rev1 in mediating protein-protein interactions. Mol Cell Biol. 2006 Nov;26(21):8173-82.

Edmunds CE, Simpson LJ, Sale JE. PCNA ubiquitination and REV1 define temporally distinct mechanisms for controlling translesion synthesis in the avian cell line DT40. Mol Cell. 2008 May 23;30(4):519-29.

Feng H, Parker JM, Lu J, Cao W. Effects of deletion and site-directed mutations on ligation steps of NAD+-dependent DNA ligase: a biochemical analysis of BRCA1 C-terminal domain. Biochemistry. 2004 Oct 5;43(39):12648-59.

Glover JN, Williams RS, Lee MS. Interactions between BRCT repeats and phosphoproteins: tangled up in two. Trends Biochem Sci. 2004 Nov;29(11):579-85.

Guo C, Fischhaber PL, Luk-Paszyc MJ, Masuda Y, Zhou J, Kamiya K, Kisker C, Friedberg EC. Mouse Rev1 protein interacts with multiple DNA polymerases involved in translesion DNA synthesis. EMBO J. 2003 Dec 15;22(24):6621- 30.

Guo C, Tang TS, Bienko M, Parker JL, Bielen AB, Sonoda E, Takeda S, Ulrich HD, Dikic I, Friedberg EC. Ubiquitin- binding motifs in REV1 protein are required for its role in the tolerance of DNA damage. Mol Cell Biol. 2006 Dec;26(23):8892-900.

Hirano Y, Sugimoto K. ATR homolog Mec1 controls association of DNA polymerase zeta-Rev1 complex with regions near a double-strand break. Curr Biol. 2006 Mar 21;16(6):586-90.

Hishiki A, Hashimoto H, Hanafusa T, Kamei K, Ohashi E, Shimizu T, Ohmori H, Sato M. Structural basis for novel interactions between human translesion synthesis polymerases and proliferating cell nuclear antigen. J Biol Chem.

2009 Apr 17;284(16):10552-60.

Jansen JG, Fousteri MI, de Wind N. Send in the clamps: control of DNA translesion synthesis in eukaryotes. Mol Cell.

2007 Nov 30;28(4):522-9.

Jansen JG, Tsaalbi-Shtylik A, Langerak P, Calléja F, Meijers CM, Jacobs H, de Wind N. The BRCT domain of mammalian Rev1 is involved in regulating DNA translesion synthesis. Nucleic Acids Res. 2005 Jan 13;33(1):356-65.

Kobayashi M, Figaroa F, Meeuwenoord N, Jansen LE, Siegal G. Characterization of the DNA binding and structural properties of the BRCT region of human replication factor C p140 subunit.J Biol Chem. 2006 Feb 17;281(7):4308-17.

Lawrence CW. Cellular functions of DNA polymerase zeta and Rev1 protein. Adv Protein Chem. 2004;69:167-203.

Lemontt JF. Mutants of yeast defective in mutation induced by ultraviolet light. Genetics. 1971 May;68(1):21-33.

Lin X, Okuda T, Trang J, Howell SB. Human REV1 modulates the cytotoxicity and mutagenicity of cisplatin in human ovarian carcinoma cells.Mol Pharmacol. 2006 May;69(5):1748-54.

(27)

78

Lopes M, Foiani M, Sogo JM. Multiple mechanisms control chromosome integrity after replication fork uncoupling and restart at irreparable UV lesions. Mol Cell. 2006 Jan 6;21(1):15-27.

Ma Y, Lu H, Tippin B, Goodman MF, Shimazaki N, Koiwai O, Hsieh CL, Schwarz K, Lieber MR. A biochemically defined system for mammalian nonhomologous DNA end joining. Mol Cell. 2004 Dec 3;16(5):701-13.

Masuda Y, Kamiya K. Role of single-stranded DNA in targeting REV1 to primer termini. J Biol Chem. 2006 Aug 25;281(34):24314-21.

McGuffin LJ, Bryson K, Jones DT. The PSIPRED protein structure prediction server. Bioinformatics. 2000 Apr;16(4):404-5.

Nelson JR, Gibbs PE, Nowicka AM, Hinkle DC, Lawrence CW. Evidence for a second function for Saccharomyces cerevisiae Rev1p. Mol Microbiol. 2000 Aug;37(3):549-54.

Nelson JR, Lawrence CW, Hinkle DC. Deoxycytidyl transferase activity of yeast REV1 protein. Nature. 1996 Aug 22;382(6593):729-31.

Ohmori H, Friedberg EC, Fuchs RP, Goodman MF, Hanaoka F, Hinkle D, Kunkel TA, Lawrence CW, Livneh Z, Nohmi T, Prakash L, Prakash S, Todo T, Walker GC, Wang Z, Woodgate R. The Y-family of DNA polymerases. Mol Cell. 2001 Jul;8(1):7-8.

Otsuka C, Kunitomi N, Iwai S, Loakes D, Negishi K. Roles of the polymerase and BRCT domains of Rev1 protein in translesion DNA synthesis in yeast in vivo.Mutat Res. 2005 Oct 15;578(1-2):79-87.

Ross AL, Simpson LJ, Sale JE. Vertebrate DNA damage tolerance requires the C-terminus but not BRCT or transferase domains of REV1. Nucleic Acids Res. 2005 Mar 1;33(4):1280-9.

Sutton MD, Walker GC. Managing DNA polymerases: coordinating DNA replication, DNA repair, and DNA recombination. Proc Natl Acad Sci U S A. 2001 Jul 17;98(15):8342-9.

Tissier A, Kannouche P, Reck MP, Lehmann AR, Fuchs RP, Cordonnier A. Co-localization in replication foci and interaction of human Y-family members, DNA polymerase pol eta and REVl protein. DNA Repair (Amst). 2004 Nov 2;3(11):1503-14.

Waters LS, Minesinger BK, Wiltrout ME, D'Souza S, Woodruff RV, Walker GC. Eukaryotic translesion polymerases and their roles and regulation in DNA damage tolerance.Microbiol Mol Biol Rev. 2009 Mar;73(1):134-54.

Wilkinson A, Smith A, Bullard D, Lavesa-Curto M, Sayer H, Bonner A, Hemmings A, Bowater R. Analysis of ligation and DNA binding by Escherichia coli DNA ligase (LigA). Biochim Biophys Acta. 2005 May 20;1749(1):113-22.

Yang L, Mullen JR, Brill SJ. Purification of the yeast Slx5-Slx8 protein complex and characterization of its DNA-binding activity. Nucleic Acids Res. 2006;34(19):5541-51.

Yu X, Chini CC, He M, Mer G, Chen J. The BRCT domain is a phospho-protein binding domain. Science. 2003 Oct 24;302(5645):639-42.

Zhang Y, Wu X, Rechkoblit O, Geacintov NE, Taylor JS, Wang Z. Response of human REV1 to different DNA damage:

preferential dCMP insertion opposite the lesion. Nucleic Acids Res. 2002 Apr 1;30(7):1630-8.