Molecular Determinants of Polyubiquitin Recognition by Continuous

Subscriber access provided by West Virginia University | Libraries

Article

Molecular determinants of polyubiquitin recognition by continuous ubiquitin binding domains of Rad18 Trung Thanh Thach, Namsoo Lee, Donghyuk Shin, Seungsu Han, Gyuhee Kim, Hong Tae Kim, and Sangho Lee Biochemistry, Just Accepted Manuscript • DOI: 10.1021/bi5012546 • Publication Date (Web): 10 Mar 2015 Downloaded from http://pubs.acs.org on March 12, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

1

Molecular determinants of polyubiquitin recognition by continuous ubiquitin binding

2

domains of Rad18

3

Trung Thanh Thach, Namsoo Lee, Donghyuk Shin, Seungsu Han, Gyuhee Kim, Hongtae Kim

4

and Sangho Lee*

5

Department of Biological Sciences, Sungkyunkwan University, Suwon 440-746, Korea

6 7

*Corresponding author: Sangho Lee

8

Phone:

+82-31-290-5913

9

Fax:

+82-31-290-7015

E-mail:

[email protected]

10 11 12

Funding Information:

13

This work was supported by the Rural Development Agency through the Woo Jang Chun

14

Program (PJ009106), and the Basic Science Research Program, through the National Research

15

Foundation of Korea (NRF), funded by the Ministry of Education (NRF-2013R1A1A2059981),

16

the Pioneer Research Center Program (2012-0009597) through NRF, funded by the Ministry of

17

Education, Science, and Technology.

18

1 Environment ACS Paragon Plus

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19

ABBREVATIONS

20

DDR: DNA damage response

21

DSB: DNA double-strand break

22

UBZ: Ubiquitin-binding zinc finger

23

UIM: Ubiquitin-interacting motif

24

ELRM: Extended ligand recognition motif

25

IRIF: Ionizing radiation-induced foci

26

SAXS: Small-angle X-ray scattering

27

NMR: Nuclear magnetic resonance

28

BLI: Bio-layer interferometry

29


Page 2 of 49

Page 3 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

30

ABSTRACT

31

Rad18 is a key factor in double-strand break DNA damage response pathways by associating

32

with K63-linked polyubiquitylated chromatin proteins through its bipartite ubiquitin binding

33

domains UBZ and LRM with extra residues in between. Rad18 binds K63-linked polyubiquitin

34

chains as well as K48-linked ones and mono-ubiquitin. However, the detailed molecular basis of

35

polyubiquitin recognition by UBZ and LRM remains unclear. Here, we examined the interaction

36

of Rad18(201-240), including UBZ and LRM, with linear polyubiquitin chains that are

37

structurally similar to the K63-linked ones. Rad18(201-240) binds linear polyubiquitin chains

38

(Ub2, Ub3, Ub4) with similar affinity to a K63-linked one for diubiquitin. Ab initio modeling

39

suggests that LRM and the extra residues at the C-terminus of UBZ (residues 227-237) likely

40

form a continuous helix, termed ‘extended LR motif’ (ELRM). We obtained a molecular

41

envelope for Rad18 UBZ-ELRM:linear Ub2 by small-angle X-ray scattering and derived a

42

structural model for the complex. The Rad18:linear Ub2 model indicates that ELRM enhances

43

the binding of Rad18 with linear polyubiquitin by contacting the proximal ubiquitin moiety.

44

Consistent with the structural analysis, mutational studies showed that residues in ELRM affect

45

binding with linear Ub2, not monoubiquitin. In cell data support that ELRM is crucial in Rad18

46

localization to DNA damage sites. Specifically E227 seems to be the most critical in

47

polyubiquitin binding and localization to nuclear foci. Finally, we reveal that the ubiquitin-

48

binding domains of Rad18 bind linear Ub2 more tightly than those of RAP80, providing a

49

quantitative basis for blockage of RAP80 at DSB sites. Taken together, our data demonstrate that

50

Rad18(201-240) forms continuous ubiquitin binding domains, comprising UBZ and ELRM, and

51

provides a structural framework for polyubiquitin recognition by Rad18 in the DDR pathway at a

52

molecular level.

53 3 Environment ACS Paragon Plus

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 49

54

INTRODUCTION

55

Protein ubiquitylation is an emerging post-translational modification regulating a range of

56

intracellular processes beyond the canonical function of proteasomal protein degradation

57

Ubiquitylation involves covalent attachment of one or more 76-residue-long ubiquitin moieties

58

to Lys residues of the target protein by a series of enzymes referred to as E1 ubiquitin-activating

59

enzymes, E2 ubiquitin-conjugating enzymes and E3 ubiquitin ligases. In the case of

60

polyubiquitylation, multiple ubiquitin moieties form a chain, linked by a specific Lys residue

61

among seven such residues on ubiquitin. Linkage-specific polyubiquitin chains are known to be

62

involved in different biological processes. For example, K63-linked polyubiquitins are associated

63

with DNA damage repair and other signaling processes, while K48-linked ones are associated

64

with canonical cytoplasmic protein degradation (5). Linear or M1-linked ones have been reported

65

to function in a similar fashion to K63-linked ones. Such covalently attached ubiquitin chains are

66

recognized non-covalently by a variety of proteins harboring ubiquitin-binding domains (5-7).

67

The DNA damage response (DDR) pathway at double-strand break (DSB) sites is one of the

68

intracellular processes regulated by ubiquitylation

69

genomic instability, resulting to tumorigenesis and cell death. Rad18, an E3 ubiquitin ligase and

70

a ubiquitin-binding protein, functions in DNA damage bypass and post-replication in yeast and

71

vertebrates. As an E3 ubiquitin ligase, Rad18 promotes monoubiquitylation of proliferating cell

72

nuclear antigen at stalled replication forks

73

sensitivity to various DNA-damaging agents and enhanced genomic instability

74

ubiquitin-binding protein, Rad18 binds directly to K63-linked polyubiquitin chains and recruits

75

Rad51C to DNA damage sites, facilitating homologous recombination repair

76

binding domain of Rad18 prevents the recruitment of other DDR mediators, such as 53BP1,

(1-4)

.

(2)

. DNA damage can lead to mutations and

(8-10)

. Mice and chicken cells lacking Rad18 show


(11, 12)

. As a

(13)

. A ubiquitin-

Page 5 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

77

RAP80, and BRCA1 to DSB sites by blocking access of these proteins to ubiquitylated

78

chromatin (14).

79

Rad18 harbors multiple domains. At the N-terminus, there is a RING domain (residues 25-64;

80

human numbering) typical of E3 ubiquitin ligases. The C-terminal region contains two domains

81

responsible for binding Rad6 (residues 340-395) and polymerase η (residues 401-445)

82

UBZ domain (ubiquitin binding zinc finger; residues 201-224) and SAP domain (SAF-A/B,

83

Acinus, and Pias; residues 243-282) are located in the middle. The SAP domain serves as a

84

DNA-binding domain (15).

85

The region including the UBZ domain (residues 186-240) is responsible for ubiquitin recognition

86

in both intra- and intermolecular interactions

87

monoubiquitylated Rad18 is the major form in the cytoplasm

88

between the monoubiquitin moiety of Rad18 and the UBZ domain is required for Rad18 auto-

89

monoubiquitylation (16). The UBZ domain is also involved in targeting Rad18 to DSB sites via a

90

direct interaction in polyubiquitin

91

ubiquitylated chromatin proteins, (13, 17) such as H2A/X (18, 19).

92

Biochemical and structural characterizations of Rad18 UBZ have advanced recently. The UBZ

93

domain of Rad18 is a type-4 C2HC zinc finger. A recent NMR structure of the Rad18 UBZ

94

domain (residues 198-227) revealed a β1-β2-α fold (20). Binding of Rad18 UBZ to monoubiquitin

95

is mediated by hydrophobic interactions by residues in the β1-strand and α-helix as well as

96

electrostatic ones, including D221 of Rad18, a conserved residue among type-3 and -4 UBZ

97

domains (UBZ3 and UBZ4, respectively). UBZ4 from Rad18 binds monoubiquitin with an

98

apparently moderate affinity (Kd) of 42 µM in vitro, while Kd values of UBZ4 from Wrnip1 and

(13)

(15)

. The

(13)

. Rad18 undergoes monoubiquitylation and the (16)

. Intramolecular interaction

and localizing Rad18 in nuclei by recognizing


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

99

UBZ3 from polymerase η are 37 and 81 µM, respectively

Page 6 of 49

(15, 17, 21)

. Rad18 UBZ binds linkage-

100

specific polyubiquitin chains, K63-linked and K48-linked ones, with different affinities: Kd is

101

30 µM for K63-Ub2 and 106 µM for K48-Ub2

102

polyubiquitin chains increase regardless of linkage specificity: 17 nM for K63-Ub5 and 36 nM

103

for K48-Ub5

104

specific polyubiquitin chains has been reported.

105

A recent study reveals that the minimal ubiquitin-binding region consists of residues 201-232, a

106

few extra residues beyond the C-terminus of the UBZ domain (residues 201-224)

107

sequence motif adjacent to the C-terminus of UBZ domain, called the LR motif (LRM; residues

108

232-240) enhances ubiquitin recognition and provides specificity among other ubiquitin binding

109

proteins involved in the DSB response

110

nuclear foci, presumably by participating in recognizing polyubiquitylated proteins localized in

111

the nuclear foci. These reports suggest that the two ubiquitin-binding domains of Rad18, UBZ

112

and LRM, mediate ubiquitin recognition together. Overexpression of UBZ-LRM reportedly

113

suppresses the association of 53BP1, RAP80, and BRCA1 to DSB sites via blockage of

114

K63-linked polyubiquitin moieties of chromatin proteins

115

competition in binding polyubiquitin moieties of the chromatin proteins among proteins involved

116

in the DSB pathway. However, molecular details on the interaction of UBZ-LRM with ubiquitin

117

and blockage of other ubiquitin binding proteins at DSB sites by UBZ-LRM remain poorly

118

understood.

119

To obtain mechanistic insights and functions on ubiquitin recognition by the UBZ-LRM of

120

Rad18 at the molecular level, we investigated which residue(s) contributed to binding ubiquitin

121

quantitatively. Furthermore, we derived a molecular model for Rad18:linear Ub2 in solution by

(17)

. Binding affinities of Rad18 UBZ with longer

(13)

. However, no information about the binding of Rad18 with other linkage-

(22)

. A

(22)

. LRM is required for the recruitment of Rad18 to

(14)

. This suggests that there may be


Page 7 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

122

small-angle X-ray scattering combined with molecular modeling. The recruitment of Rad18 to

123

DNA damage sites through polyubquitin recognition by new region, termed ELRM, was also

124

examined. Finally, the competitive binding to linear polyubiquitin between Rad18 and RAP80

125

was monitored quantitatively. Collectively, our results provide insight on how Rad18 recognizes

126

polyubiquitin and functions in the DDR pathway at a molecular level.

127


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

128

MATERIALS AND METHODS

129

Plasmids, cloning, and mutagenesis

130

Genes encoding various regions of Rad18 were amplified by PCR from full-length human Rad18

131

cDNA (accession number, NM_020165.3), then inserted between BamHI and EcoRI restriction

132

enzyme sites in a parallel GST2 vector (23). Genes encoding hexa-histidine-tagged monoubiquitin

133

(Ub) or linear Ub2, Ub3, or Ub4 were inserted between the BamHI and EcoRI sites in a parallel

134

His2 vector. To prepare plasmids encoding monoubiquitin mutants for linkage-specific

135

polyubiquitin synthesis (77D, K63C, K48C), the gene encoding monoubiquitin was cloned into

136

pET-3a (Novagen) using restriction enzymes NdeI and EcoRI first and the resulting plasmid was

137

subject to site-directed mutagenesis. To express Rad18 UBZ-ELRM constructs in mammalian

138

cells, Rad18(171-240) fused with SV40 nuclear localization signal was cloned in pEGFP-C2

139

vector (Clonetech) between EcoRI and BamHI restriction enzyme sites. The resulting plasmid,

140

pEGFP-Rad18(171-240)-SV40NLS, encodes EGFP-Rad18(171-240)-linker (GG)-SV40NLS

141

(PKKKRKV) fusion protein. Site-directed mutagenesis was performed with the QuikChange kit

142

(Stratagene) based on the protocol supplied. The identities of all constructs were confirmed by

143

DNA sequencing.

144 145

Protein expression and purification

146

Linear polyubiquitin chains in which the C-terminal G76 of the preceding ubiquitin moiety is

147

directly linked to the N-terminal M1 of the next ubiquitin moiety were expressed as a

148

hexahisitidine tagged proteins: His-Ub2, His-Ub3, and His-Ub4. Rad18 was expressed as a GST


Page 8 of 49

Page 9 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

149

fusion protein. All proteins including Rad18 mutants were overexpressed in Escherichia coli

150

BL21(DE3) strain. Cells were inoculated in LB medium, incubated with gentle shaking at 37°C

151

up to an OD600 of 0.8-1.0, induced with 500 µM isopropyl-D-thiogalactoside, and grown further,

152

overnight at 20°C. The cells were harvested by centrifugation (13,000 rpm, 1 h), resuspended in

153

buffer A (50 mM Tris-HCl, pH 7.5, and 150 mM NaCl) and lysed by sonication. The supernatant

154

fraction was applied to glutathione-Agarose resin (GE Healthcare) for GST fusion proteins and

155

Ni-NTA-Agarose resin (Qiagen) for His-tagged proteins, then eluted in buffer B (50 mM Tris-

156

HCl, pH 8.0, and 10 mM reduced glutathione) or buffer C (50 mM Tris-HCl, pH 7.5, 20 mM

157

imidazole, and 500 mM NaCl). GST and His-tag were cleaved from the fusion proteins using

158

tobacco etch virus protease

159

buffer D (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.5 mM EDTA). The released GST or

160

His-tag was cleared using glutathione-Agarose or Ni-NTA-Agarose resin again. Proteins were

161

further purified on a Superdex-75 size exclusion column (GE HealthCare) pre-equilibrated with

162

the buffer A. Fractions containing pure proteins, judged by SDS-PAGE, were pooled and

163

concentrated by centrifugation through 10 kDa-centrifugal filters (Ambicom Ultra). Protein

164

concentration was determined by measuring absorbance at 280 nm and the Bradford assay (25).

165

Proteins required for synthesizing K63- and K48-linked diubiquitin chains were prepared as

166

follows. GST-HIP2 and GST-Mms2/His-Ubc13 complex were overexpressed in E. coli

167

BL21(DE3) cells and purified as described above. The ubiquitin mutants (77D, K63C and K48C)

168

were overexpressed in E. coli BL21(DE3) cells. After lysis, the supernatant was heated at 70°C

169

for 10 min, and then further purified on a Superdex-75 size exclusion column pre-equilibrated

170

with buffer A, supplemented with 1 mM DTT. His-Uba1 was overexpressed in a baculovirus-

171

infected Sf9 insect cell system (Invitrogen). Cells were harvested by centrifugation, resuspended

(24)

and dialyzing the solution containing the fusion proteins against


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

172

in buffer E (50 mM Tris-HCl, pH 8.0, 200 mM NaCl, 5 mM β-mercaptoethanol, 5% (v/v)

173

glycerol, and 1× protease-inhibitor cocktail (Roche)), and lysed by sonication. His-Uba1 was

174

eluted from the Ni-NTA resin in the buffer E containing 40 mM imidazole. Further purification

175

was done on a Mono Q anion exchange column (GE Healthcare) equilibrated with buffer F (50

176

mM Tris-HCl, pH 8.0, 1 mM DTT, and 5% glycerol). Fractions containing pure His-Uba1 were

177

pooled, concentrated, and stored at -80°C until use.

Page 10 of 49

178 179

K63- and K48-linked diubiquitin synthesis and purification

180

K63- and K48-linked diubiquitin chains were synthesized as reported previously (26, 27). K63-Ub2

181

was synthesized using GST-Mms2/His-Ubc13 complex while GST-HIP2 was used as E2 for

182

K48-Ub2 synthesis. Mammalian E1, His-Uba1, was used in both cases. To synthesize K63-Ub2,

183

5× reaction buffer was prepared (250 mM Tris-HCl, pH 7.5, 50 mM creatine phosphate, 125 mM

184

MgCl2, 15 UI/mL creatine phosphokinase, 15 UI/mL inorganic pyrophosphate, 300 mM ATP,

185

and 2.5 mM DTT). Reaction mixture containing 20 mg/mL Ub-77D and 50 mg/mL Ub-K63C

186

with 1× reaction buffer was supplemented with 8 mM GST-Mms2/His-Ubc13 complex and

187

100 µM His-Uba1. The solution was incubated at 37°C for 5 h with gentle shaking. K48-Ub2

188

was generated similarly. The reaction mixture contained 20 mg/mL Ub-77D, 50 mg/mL

189

Ub-K48C, 50 µM GST-HIP2 and 100 µM His-Uba1 in 1× reaction buffer.

190

Diubiquitin chains were purified by dilution of reaction mixtures in 0.5 M ammonium acetate,

191

pH 4.5, 10-fold. Then, the acidified solutions were applied to a cation exchange Hitrap SH

192

column (GE Healthcare) pre-equilibrated with buffer G. Bound proteins were eluted with a linear

193

gradient of buffers G (50 mM ammonium acetate, pH 4.5) and H (50 mM ammonium acetate,


Page 11 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

194

pH 4.5, and 1 M NaCl). Fractions containing linkage-specific diubiquitin chains were pooled,

195

concentrated and applied to a Superdex-75 size exclusion column (GE Healthcare) pre-

196

equilibrated with buffer A.

197 198

GST pull-down

199

Next, 2 µg of each GST-Rad18 protein was incubated with 40 µL of glutathione-Agarose resin

200

for 30 min at ambient temperature in buffer A, and washed three times with the same buffer to

201

remove unbound proteins. Subsequently, 500 ng of polyubiquitin chains were added to the resin

202

bound with GST-Rad18 protein. The mixture was then incubated for 2 h at 4°C with gentle

203

shaking, and washed with buffer I (50 mM Tris-HCl, pH 7.5, 300 mM NaCl, and 0.5% NP-40).

204

Bound polyubiquitin chains were analyzed by SDS-PAGE and immunoblotting using mouse

205

monoclonal anti-ubiquitin combined with IgG-HRP antibodies or GST-HRP antibodies (Santa

206

Cruz Biotechnology). Protein bands were visualized with an ImageQuant LAS 40 digital system

207

(GE Healthcare).

208 209

Bio-layer interferometry

210

Bio-layer interferometry (BLI) experiments were performed on a Blitz system (ForteBio). For

211

linkage-specific diubiquitin binding to Rad18, either 1 µM linear, K48- or K63-linked Ub2 was

212

immobilized on amine reactive sensors (ForteBio). Binding was monitored by adding varying

213

concentrations of GST-Rad18 proteins (0.263-21 µM) in MES, pH 6.0. To study length-

214

dependent binding of linear polyubiquitin chains to Rad18, 1 µM each of His-Ub, His-Ub2,


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 49

215

His-Ub3, and His-Ub4 was immobilized on Ni-NTA biosensors (ForteBio). Binding of

216

GST-Rad18 proteins (0.263-105 µM) to the immobilized linear polyubiquitin chains was then

217

monitored. GST was used as a negative control. Experiments were repeated at least twice. Kd

218

was determined based on steady-state analysis using PRISM (GraphPad Software).

219 220

Ab initio modeling of Rad18(201-240)

221

Ab initio models for Rad18(201-240) were derived using GalaxyTBM

222

modeling with GalaxyTBM, 100 models that maximally satisfied the restraints were generated

223

within 1 Å r.m.s.d. A representative structure is the nearest model to the largest cluster center.

224

For Rosetta modeling, initially, 20,000 models were generated. Then, 1,044 structural models

225

within 1 Å r.m.s.d. were clustered and the top model from the cluster was selected for further

226

structural analysis. Calculations were performed at the Cassini cluster system at the UCLA-DOE

227

Computer Core facility.

(28)

and Rosetta

(29)

. For

228 229

Small-angle X-ray scattering

230

Small-angle X-ray scattering (SAXS) experiments were carried out at beamline 4C in the Pohang

231

Accelerator Laboratory (South Korea). Data were collected at 3.91, 1.95, and 0.98 mg/mL

232

concentrations of sample in buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and

233

0.5 mM ethylenediaminetetraacetic acid (EDTA). Final scattering data were generated by

234

merging two data sets from 3.91 and 1.95 mg/mL concentrations at 30 s exposure. The radius of

235

gyration (Rg) was determined using the Guinier approximation (30). The pair distribution function


Page 13 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(31)

236

(P(r)) was calculated using Fourier inversion of the scattering intensity I(q) using GNOM

237

Rg was also calculated from the P(r) function. Molecular masses of samples were determined as

238

described previously

239

profile using DAMMIF

240

and SASREF

241

diubiquitin:HOIL1-L-NZF (PDB ID: 3BOA), monoubiquitin:UBZ of Wrnip1 (PDB ID: 3VHT)

242

and ab initio models for Rad18(201-240). The best docked model was chosen using Minimal

243

Ensemble Search (FoXSMES)

244

using PyMOL (37).

(34, 35)

.

(32)

. Low-resolution molecular envelopes were derived from the scattering

(33)

. Rigid body docking and modeling were performed with FoXSDock

using atomic coordinates for monoubiquitin (PDB ID: 1UBQ), linear

(31, 36)

. Three-dimensional graphics representations were prepared

245 246

Cell culture, ionizing radiation treatment and immunofluorescence staining

247

Mammalian 293T cells were maintained in Dulbecco modified Eagle medium (DMEM)

248

supplemented with 10% fetal bovine serum, 5% CO2 (v/v) at 37oC. Cells expressing EGFP

249

tagged Rad18 UBZ-ELRM constructs growing on coverslips and exposed to 10 Gy IR using a JL

250

shepherd 137Cs irradiator. Coverslips were stained and imaged with a Leica confocal microscope

251

as previously described (14, 38). Briefly, coverslips containing grown cells were fixed with 3% (v/v)

252

paraformaldehyde at room temperature for 10 min. Cells were permeabilized and blocked with

253

0.5% (v/v) Triton X-100, 1% (v/v) goat serum and then incubated with primary antibodies for 60

254

min at room temperature. Cells were clearly washed with phosphate buffered saline (PBS) and

255

incubated with an appropriate secondary antibody, rhodamine-conjugated goat anti-rabbit IgG

256

for 30 min. Images were obtained after final washing cells with PBS. Cells exhibiting ≥ 10 foci


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

257

were recognized as positive foci. ≥ 50 cells were counted in each experiment. Nuclei were

258

stained with 4,6-diamidino-2-phenylindole (DAPI) (Invitrogen).

Page 14 of 49

259 260

Competitive ubiquitin binding assay

261

To monitor competitive binding of Rad18 and RAP80 with linear Ub2, GST pull-down and BLI

262

measurements were performed. Rad18 and RAP80 were permutated in the order of addition to

263

linear Ub2. For GST pull-down experiments, 1 µM linear His-Ub2 was immobilized on Ni-NTA

264

agarose resin. In the first experiment, 20 µM Rad18 was added to His-Ub2-loaded Ni-NTA

265

agarose resin and incubated for 2 h at 4°C with gentle shaking, and washed with buffer A to

266

remove unbound proteins. Subsequently, 10 µM GST-RAP80(60-124) was added to the resin

267

and incubated for 2 h in the same condition as above followed by washing with buffer I. Bound

268

GST-RAP80(60-124) and immobilized His-Ub2 were analyzed by immunoblotting using GST-

269

HRP and His-HRP antibodies. In the second experiment, RAP80(60-124) was pre-incubated

270

with Ni-NTA agarose resin with the immobilized linear His-Ub2, then 10 µM GST-Rad18(171-

271

240) was added to the resin and analyzed by immunoblot. For BLI experiments, 1 µM linear His-

272

Ub2 was immobilized on the Ni-NTA sensors. After baseline stabilization with buffer A, 20 µM

273

Rad18(171-240) was loaded until the signal was saturated. Varying concentrations of GST-

274

RAP80(60-124) were then added and analyzed for competitive binding. In reciprocal

275

experiments, 20 µM RAP80 was loaded to Ni-NTA sensors with immobilized linear His-Ub2,

276

then GST-Rad18(171-240) was added. The negative control involved replacing either GST-

277

RAP80(60-124) or GST-Rad18(171-240) by GST.

278


Page 15 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

279

Data analysis

280

All data and statistical analyses were performed using PRISM (GraphPad Software) and

281

SigmaPlot (Systat Software). Differences between data groups were calculated using the two-tail

282

Student’s t-test, and considered significant with a value of P ≤ 0.05. Results are presented as

283

means and standard deviations from 2-4 independent experiments.

284


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 49

285

RESULTS

286

Rad18 recognizes linear and K63-linked diubiquitin marginally better than K48-linked ones

287

Rad18 is known to bind both K63-linked and K48-linked pentaubiquitin chains with similar

288

affinities

289

Rad18 constructs harboring UBZ and LRM as well as a region at the N-terminus of the UBZ

290

domain (Fig. 1A). Because linear polyubiquitin chains are structurally equivalent and function

291

similarly to K63-linked ones

292

linked and K48-linked ones. We asked how GST-Rad18(171-240) differentially recognized

293

linkage-specific diubiquitin chains, considered to be the minimum length for polyubiquitin

294

chains. First, we performed pull-down assays using GST-Rad18(171-240) with either linear Ub2,

295

K63-Ub2, or K48-Ub2. We found that Rad18 bound similarly to linear and K63-Ub2, but less

296

favorably to K48-Ub2 (Fig. 1B), indicating that GST-Rad18(171-240) apparently prefers binding

297

to linear or K63-Ub2. Next, we compared binding affinities of GST-Rad18(171-240) to linkage-

298

specific diubiquitin chains determined by BLI. GST-Rad18(171-240) exhibited marginally

299

enhanced binding affinity to K63-Ub2 (Kd = 5.4±1.5 µM) versus K48-Ub2 (Kd = 19.8±7.5 µM),

300

but similar affinity for linear Ub2 (Kd = 3.6±1.5 µM; Table S1, Fig. 1C, D). Both qualitative and

301

quantitative binding studies showed that GST-Rad18(171-240) bound to linear and K63-linked

302

diubiquitin in preference to a K48-linked one. This observation is in apparent contrast to a

303

previous study where GST-Rad18(186-240) bound K63-linked and K48-linked pentaubiquitin

304

with similar affinities (13). However, the preference observed in our study seemed to be marginal

305

because the ratio of binding affinity of Rad18 with K63-linked polyubiquitin chain to that with

306

K48-linked one was ~2 for pentaubiquitin and ~4 for diubiquitin. The observation that

307

GST-Rad18(171-240) recognizes both linear Ub2 and K63-Ub2 approximately equally is

(13)

. To dissect ubiquitin binding by Rad18 at the domain level, we generated several

(5, 6, 38)

, we included linear polyubiquitin chains as well as K63-


Page 17 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

308

consistent with linear and K63-linked ubiquitin chains have similar structures but both being

309

significantly different from the K48-linked one

310

GST-Rad18(171-240) with linear Ub2 to K63-Ub2, we used linear polyubiquitin chains for

311

further analysis.

(5, 6, 39, 40)

. Given the similar binding affinities of

312 313

The extended LRM at the C-terminus of UBZ is important in polyubiquitin recognition by Rad18

314

To investigate whether Rad18 recognizes linear polyubiquitin in a chain-length dependent

315

manner and the precise roles of UBZ and LRM in polyubiquitin binding by Rad18, we prepared

316

Ub, linear Ub2, Ub3, and Ub4 and four constructs of GST-Rad18 (residues 171-224, 171-234,

317

171-240, and 201-240). Pull-down results showed similar affinities with different lengths of

318

linear polyubiquitins regardless of the GST-Rad18 construct used (Fig. S1). Consistent with the

319

pull-down assays, quantitative BLI results revealed that binding affinities of GST-Rad18

320

constructs with linear polyubiquitin chains (Ub2, Ub3, Ub4) were not statistically significantly

321

different from each other, indicating that chain length may not affect affinity of Rad18 for

322

polyubiquitin chains (Table 1, Figs. 2, S2). However, we observed moderate enhancements in

323

affinities for polyubiquitin chains in comparison to those for monoubiquitin in GST-Rad18

324

constructs containing UBZ and residues at its C-terminus (171-234, 171-240, 201-240). The

325

binding affinity of GST-Rad18(171-224) with monoubiquitin was similar to those with Ub2, Ub3,

326

and Ub4. However, both GST-Rad18(171-234) and GST-Rad18(171-240) exhibited an increase

327

in binding affinities (3.5- or 6-fold, respectively) with linear polyubiquitin chains. By contrast,

328

GST-Rad18(171-200) showed no interaction with ubiquitin at all. These results demonstrate that

329

the region at the C-terminus of UBZ, spanning residues from 225-240, termed the “extended


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 49

330

LRM” (ELRM), enhances linear polyubiquitin binding affinity. The region at the N-terminus of

331

the UBZ domain (residues 171-200) may facilitate polyubiquitin recognition by Rad18 when

332

comparing the binding affinities of GST-Rad18(171-240) and GST-Rad18(201-240). These

333

results demonstrate that the residues between UBZ and LRM are involved in polyubiquitin

334

recognition by Rad18.

335 336

Ab initio modeling of Rad18(201-240)

337

To understand the interaction between Rad18 and polyubiquitin at a molecular level, we derived

338

atomic models for the structure of Rad18(201-240) using GalaxyTMB and Rosetta

339

two programs yielded similar overall structures: the Rad18(201-240) structural model consists of

340

a UBZ domain and a C-terminal α-helix for ELRM (α2, residues 227-237) (Fig. 3). The UBZ

341

domain contains two short antiparallel β-strands (residues 201-204 for β1 and 208-211 for β2)

342

and an α-helix (α1, residues 216-224), which packs against the β-strands with the zinc ion. As

343

with the homologous Wrnip1-UBZ, Rad18-UBZ is classified as a C2HC zinc finger

344

derived structural model for the UBZ domain of Rad18 is similar to a recent NMR structure

345

(residues 198-227)

346

211 and the α-helix spans residues 212-224. Although the ab initio models by the two programs

347

feature the same secondary structural elements, the relative orientation between the UBZ domain

348

and the ELRM helix is slightly different. The ELRM helix from the GalaxyTBM model was

349

formed on the right side of the UBZ domain, while it is on the left side in the Rosetta model

350

(Fig. 3). Another difference between the two ab initio models is that part of LRM (residues 235-

351

240) is modeled to form a loop by GalaxyTBM and an extension of the second helix by Rosetta

(28, 29)

. The

(17)

. The

(20)

. In the NMR structure, two β-strands with a hairpin cover residues 201-


Page 19 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

352

(Fig. 3). Nevertheless, our modeling results suggest that ELRM may form a continuous helix.

353

We used both ab initio models for Rad18(201-240), generated for SAXS modeling.

354 355

Low-resolution solution structure of Rad18:linear Ub2 complex suggests the role of ELRM in

356

proximal ubiquitin recognition

357

To obtain molecular insights into the interaction between UBZ and ELRM of Rad18 and

358

polyubiquitin in solution, we performed SAXS measurements and derived a molecular envelope

359

for the Rad18(201-240):linear Ub2 complex. Prior to the SAXS measurements, the purity and

360

polydispersity of the Rad18(201-240):linear Ub2 complex were checked by size exclusion

361

chromatography, SDS-PAGE, and DLS (Fig. S3). Parameters for SAXS measurements and

362

derivation of low-resolution molecular envelopes for the Rad18(201-240): linear Ub2 complex

363

are summarized in Table 2. The pair distribution functions (P(r)) displayed a narrow range,

364

68.2-68.6 Å (Fig. 4A). The Rg determined from Guinier plots was in good agreement with that

365

from GNOM. The Rg calculated using GNOM ranged from 21.6 to 22.0 Å, supporting the

366

concentration-independency of the sample during SAXS measurements. The results indicated

367

that the complex is monomeric and may not be elongated in solution. For ab initio modeling of

368

the complex, seven envelopes were generated using DAMMIF (Fig. S4). The best model fitting

369

with SAXS data was chosen from the lowest χ2 value (χ2 = 1.43) using FoXSMES (Fig. 4B). The

370

rigid body complex was further modeled with SASREF.

371

Because the FoXSDock complexes generated no clash between residues and fitted the SAXS data

372

with a lower χ2 value, the Rad18(201-240) model derived from GalaxyTMB was used for the

373

SAXS Rad18(201-240): linear Ub2 model. The final molecular envelope with the Rad18(201-


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 49

374

240): linear Ub2 complex docked (Fig. 4C) revealed that Rad18(201-240) contacts linear Ub2

375

mainly by its β1-strand and α1-helix in the UBZ domain and the hydrophobic patch centered on

376

I44 of the distal/donor ubiquitin moiety. This is consistent with Rad18-UBZ using the helix and

377

β1-strand for interaction with the conserved hydrophobic surface on ubiquitin

378

our model suggests that the residues between UBZ and LRM may interact with the

379

proximal/acceptor ubiquitin moiety. This model is consistent with our BLI results showing no

380

difference in binding affinities of GST-Rad18(171-224) (lacking ELRM) with monoubiquitin

381

and linear Ub2.

(20)

. Additionally,

382 383

Mutational analysis of the interaction of Rad18(201-240) with linear Ub2

384

To validate the Rad18(201-240): linear Ub2 complex model derived by SAXS, we prepared nine

385

mutants and assessed their affinities with monoubiquitin and linear Ub2 by BLI: D221A, R226A,

386

E227A, E228A, E231A, V237A, H238A, K239A, and R240A. The D221A mutant, located in

387

the UBZ domain, reportedly abrogates ubiquitin binding (17, 19, 20). Four mutants (R226A, E227A,

388

E228A, and E231A) were positioned in the region between UBZ and LRM. R226, E227, and

389

E228 residues are located close to the proximal ubiquitin moiety in our structural model (Fig.

390

5A). E231 is a moderately conserved among the UBZ domains (Fig. 5A). Finally, we mutated

391

four C-terminal residues (V237A, H238A, K239A, and R240A) belonging to the LRM (Fig. 5A).

392

Quantitative binding analysis showed that all mutants except D221A reduced binding affinities

393

with linear Ub2, but not with monoubiquitin (Fig. 5B, Table S2). The D221A mutation showed

394

virtually no interaction with monoubiquitin or linear Ub2, consistent with previous findings (17, 19,

395

20)

and our rigid body complex model indicates that D221 interacts with R42 on the distal/donor


Page 21 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

396

ubiquitin moiety functioning in complex stabilization (Fig. S5A). E227A, E228A and

397

E227A/E228A exhibited similar affinities with linear Ub2 compared with those with

398

monoubiquitin, consistent with our structural model, where E227 interacts with Q78

399

(corresponding to Q2 in monoubiquitin) of the proximal/acceptor ubiquitin moiety; also, E228

400

may preserve the structure of the domain (Fig. S5B). The LRM mutants (V237A, H238A,

401

K239A, and R240A) showed modest decreases in affinities with linear Ub2 but no statistically

402

significant change in affinities with monoubiquitin. This can be explained in that the flexible

403

loop LMR may interact with the proximal/acceptor ubiquitin moiety where K239 and R240 may

404

form hydrogen bonds with D108 (corresponding to D32 in monoubiquitin; Fig. S6B). This

405

supports previous reports that LRM augments binding affinity with polyubiquitin

406

consistent with our finding that GST-Rad18(171-234), which lacks the LRM, binds linear

407

polyubiquitin chains less tightly than GST-Rad18(171-240) (Fig. 2B). Together, these results

408

establish that ELRM contributes to polyubiquitin recognition by Rad18 and suggest that E227

409

and E228, located between UBZ and LRM, and the second helix, recognize the

410

proximal/acceptor ubiquitin moiety.

(22)

and is

411 412

Polyubiquitin recognition by ELRM is crucial for Rad18 localization in DNA damage sites

413

Since in vitro study revealed that E227 and E228 in ELRM are critical in recognizing linear

414

polyubiquitin (Fig. 5), we tested whether the aforementioned residues functions in localizing

415

Rad18 on DNA damage sites in a cultured cell system. We constructed a plasmid, pEGFP-

416

Rad18(171-240)-SV40NLS, encoding Rad18 UBZ-ELRM (residues 171-240) with EGFP at its

417

N-terminus and SV40 nuclear localization signal at its C-terminus. We introduced the following


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 49

418

mutations in the pEGFP-Rad18(171-240)-SV40NLS plasmid: E227A, E228A or E227A/E228A

419

in ELRM or D221A in UBZ domain. Expression levels of the transfected plasmids harboring the

420

aforementioned mutations were monitored in 293T cells by immunoblotting with a GFP antibody

421

to ensure similar expression levels among the mutant proteins (Fig. S6). The cells were treated

422

with ionizing radiation to test the recruitment of Rad18 UBZ-ELRM to ionizing radiation-

423

induced foci (IRIF). Our results confirmed that D221A mutant, which was known to be critical

424

in both polyubiquitin recognition and recruitment to IRIF

425

recruitment of Rad18 to IRIF (Fig. 6). When compared to D221A mutant, E227A mutant also

426

abrogated the recruitment of Rad18 UBZ-ELRM to IRIF similarly or more while E228A mutant

427

reduced less (Fig. 6). Double mutant E227A/E228A also showed a noticeable reduction in the

428

recruitment to IRIF. These data are consistent with our in vitro polyubiquitin binding data and a

429

structural model in that E227 is likely to make a direct contact with the proximal ubiquitin

430

moiety (Fig. 5). Taken together, these results establish that ELRM as well as UBZ domain is

431

essential in recruitment of Rad18 to IRIF through specific polyubiquitin recognition.

(14, 17)

, significantly reduced the

432 433

Rad18 outcompetes RAP80 in linear polyubiquitin binding

434

In DSB sites, numerous proteins are recruited to initiate the DDR pathway

435

RAP80 play major roles in recruiting other factors during this response by associating with

436

ubiquitylated chromatin through their ubiquitin-binding domains: LRM and two UIMs for

437

RAP80, and UBZ and LRM for Rad18 (22). A previous study reported that overexpressed UBZ of

438

Rad18 interferes with the localization of RAP80 in the nucleus, but not vice versa

439

investigate how Rad18 competes with RAP80 in polyubiquitin binding, we examined whether


(22, 41)

. Rad18 and

(14)

. To

Page 23 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

440

the presence of Rad18 or RAP80 affected binding of the other protein with linear Ub2 in vitro,

441

both qualitatively and quantitatively. We immobilized His-linear Ub2 in Ni-NTA resin and

442

analyzed the binding of GST-RAP80(60-124) to linear Ub2 in the absence and presence of

443

Rad18(171-240) (Fig. 7A). Ubiquitin binding of GST-RAP80(60-124) decreased in the presence

444

of the pre-formed Rad18(171-240):linear Ub2 complex (Lanes 5 vs. 7, Fig. 7A). By contrast, the

445

presence of the pre-formed RAP80(60-124):linear Ub2 did not interfere significantly with

446

ubiquitin binding of GST-Rad18(171-240) (lanes 5 vs. 7, Fig. 7B).

447

Quantitative binding analysis using BLI confirmed the results from the GST pull-down assay.

448

First, we checked whether Rad18(171-240) and RAP80(60-124) interacted with each other in the

449

absence of ubiquitin. When Rad18(171-240) was immobilized on a sensor, GST-RAP80(60-124)

450

did not interact with Rad18(171-240) while GST-linear Ub4 did (Fig. S7). Next, we immobilized

451

linear Ub2 on a sensor and determined binding affinities of GST-RAP80(60-124) or GST-

452

Rad18(171-240) with linear Ub2 in the presence or absence of Rad18(171-240) or RAP80(60-

453

124), respectively, by BLI (Fig. 8A, B). Binding of GST-Rad18(171-240) to linear Ub2 was

454

impaired modestly by the presence of RAP80(60-124), with Kd values being 10.8±4.4 µM in the

455

presence and 2.7±1.3 µM in the absence of RAP80 (Table S3). The reciprocal BLI experiments

456

indicated that Rad18(171-240) also perturbed the interaction of GST-RAP80(60-124) with linear

457

Ub2: the binding affinity decreased from 29.5±7.4 µM to 11.2±6.2 µM in the presence and

458

absence of Rad18(171-240), respectively (Table S3). Although polyubiquitin binding of both

459

Rad18 and RAP80 was affected by the presence of the other protein, Rad18 appears to bind

460

polyubiquitin more tightly (by about 3-fold) than RAP80 in competitive binding. These results

461

are consistent with a previous study in which overexpression of Rad18 lead to blockage of


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

462

recruitment of RAP80, presumably by interfering with the binding of RAP80 to ubiquitylated

463

chromatin (22).

464 465


Page 24 of 49

Page 25 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

466

DISCUSSION

467

In this study, we demonstrated that residues between the UBZ and LRM of Rad18 are also

468

involved in polyubiquitin recognition by biochemical and mutational analyses. We further

469

suggest that the residues between UBZ and LRM are likely to form a continuous helix along with

470

LRM, by SAXS and molecular modeling. Thus, we refer to residues 225-240 as the extended

471

LRM (ELRM). We propose that the ubiquitin-binding domains of Rad18 constitute structurally

472

continuous domains, the UBZ and ELRM. The structural architecture of continuous bipartite

473

ubiquitin binding domains is not unprecedented. For example, the ubiquitin-binding domains of

474

Rabex-5, an A20-type zinc finger plus MIU, include a continuous helix spanning the C-terminal

475

half of the zinc finger and the whole MIU domain

476

pathway, features tandem UIM domains connected by a linker

477

tandem UIMs of RAP80 has been proposed to become part of a long, continuous helix upon

478

ubiquitin binding

479

comprising UMI (UIM- and MIU-related) and MIU1 at the N-terminus

480

continuous bipartite ubiquitin domains in a helix might be a common theme in ubiquitin

481

receptors in the DDR pathway.

482

The accumulation of ubiquitin-binding proteins such as Rad18 in DNA damage sites depends on

483

the linkage-specific polyubiquitin recognition

484

specificity to ubiquitin binding proteins for recruiting their partners as well as an additional

485

source for affinity enhancement towards polyubiquitin (22). We provide compelling evidence that

486

ELRM, covering residues between the C-terminus of UBZ and LRM, contributes to contacting

487

ubiquitin moieties (Figs. 5 and S5). Based on our data we propose differential roles of residues in

488

ELRM in linkage-specific polyubiquitin binding: residues between UBZ and LRM, including

(42, 43)

. RAP80, also involved in the DDR (44, 45)

. The linker between the

(46, 47)

. RNF168 possesses continuous-helix ubiquitin-binding domains (48)

. This architecture of

(2, 49)

. It has been proposed that LRM confers


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 49

489

E227 and E228, contributes to recognition of the proximal ubiquitin moiety while those in LRM

490

augment polyubiquitin recognition by their structural plasticity suggested by ab initio modeling

491

(Fig. 3). However, we cannot rule out the possibility that the interaction between ELRM and

492

ubiquitin may be facilitated by the UBZ domain as well because we could not reliably determine

493

the binding affinity of ELRM alone with monoubiquitin or linear Ub2: D221A mutant in the

494

UBZ doamin totally diminished linear Ub2 binding (Fig. 5B). Our in cell data reveal that E227A,

495

E228A and E227A/E228A mutants in ELRM that abrogated binding to linear Ub2 affected

496

recruitment of Rad18 to DNA damage sites (Fig. 6). Notably, E227A reduced DNA damage foci

497

formation more severely than D221A, thereby uncovering the critical role of E227 in recruitment

498

of Rad18 to DNA damage sites via polyubiquitin recognition. Our in vitro and in cell data

499

collectively identify novel functional roles of the residues in ELRM in addition to the known

500

function in H2A/X recognition (22).

501

A multiple sequence alignment reveals substantial difference in ELRM sequences (Fig. S8),

502

suggesting that the ELRM may be responsible for conferring specificity in ubiquitin recognition

503

(22)

504

architecture (Fig. S8). While the UBZ domains in both proteins exhibit high sequence identity

505

(58%), the ELRMs do not (6%). Such sequence differences in ELRM of Rad18 and Wrnip1 may

506

play a role in differential binding affinities for linkage-specific polyubiquitin recognition: we

507

found Rad18 UBZ-ELRM modestly prefers K63-linked Ub2 to a K48-linked one whereas

508

Wrnip1 recognizes both with similar affinities

509

modeling and SAXS data, we found that linear Ub2 is likely to adopt an extended conformation

510

when binding to Rad18. These results suggest that Rad18-ELRM may preferably interact with

511

linear Ub2 or K63-Ub2 (extended conformations, in general

. For example, Rad18 and Wrnip1 share the same bipartite ubiquitin binding domain

(17)

. Combining Rad18(201-240) ab initio


(5, 6)

) more than K48-linkage

Page 27 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(39, 50)

512

ubiquitin chains (compact conformations

). Our BLI data also revealed that Rad18-ELRM

513

interacts well with linear Ub2 and K63-Ub2, but less well with K48-Ub2 (Fig. 1C vs. D). Taken

514

together, our data suggest that the ELRM acts as a specificity filter for linkage-specific

515

polyubiquitin chains with the UBZ domain being a generic polyubiquitin binder.

516

Our data also provide quantitative and molecular bases for the Rad18-dependent redistribution of

517

DDR factors. Binding analysis revealed that Rad18 UBZ-ELRM binds with linear Ub2 more

518

strongly than RAP80 LRM-UIMs, by about 3-fold (Fig. 8, Table S3). However, the affinity is

519

similar in RAP80 LRM-UIMs and Rad18 UBZ alone. Moreover, Rad18 UBZ-ELRM and

520

RAP80 LRM-UIMs compete for binding with polyubiquitin chains, but the interference seems to

521

be stronger in the Rad18 case. Previous in vivo results also indicated that overexpression of

522

Rad18 UBZ-ELRM can block the recruitment of RAP80 to DSB sites but RAP80 could not

523

These findings demonstrate that Rad18 UBZ-ELRM may downregulate recruitment of RAP80 to

524

DSB sites by favorable binding with polyubiquitin chains. BRCA1 is recruited to DSB sites by

525

interaction with the Abraxas:RAP80 complex

526

Together, our results demonstrate that the continuous bipartite ubiquitin binding domains of

527

Rad18 outcompete RAP80 in polyubiquitin recognition, thereby potentially regulating the DDR

528

pathway.

(51)

; thus, it is abrogated by Rad18 too

529


(14)

.

(14)

.

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

530

ACKNOWLEDGMENTS

531

We thank the staff at beamline 4C at the Pohang Accelerator Laboratory for technical assistance

532

in collecting SAXS data and Prof. David Eisenberg, and Drs. Duilio Cascio and Lukasz

533

Goldschmidt at UCLA for access to the Cassini cluster.

534 535

SUPPORTING INFORMATION

536

Supplemental tables (S1 – S3), figures (S1 – S7) and reference are available. This material is

537

available free of charge via the internet at http://pubs.acs.org.

538 539

CONFLICT OF INTEREST

540

The authors have no competing financial interests.

541


Page 28 of 49

Page 29 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

542

REFERENCES

543

1.

544

3353-3359.

545

2.

546

by Ubiquitin and SUMO, Mol. cell 49, 795-807.

547

3.

548

discovered regulator of cell signalling, Trends in biochemical sciences 38, 94-102.

549

4.

550

signaling, Mol. Cell 33, 275-286.

551

5.

552

biochemistry 81, 203-229.

553

6.

554

D., and Barford, D. (2009) Molecular discrimination of structurally equivalent Lys 63-

555

linked and linear polyubiquitin chains, EMBO Rep. 10, 466-473.

556

7.

557

J. 399, 361-372.

558

8.

559

Yamaizumi, M. (2004) Rad18 guides poleta to replication stalling sites through physical

560

interaction and PCNA monoubiquitination, EMBO J. 23, 3886-3896.

561

9.

Haglund, K., and Dikic, I. (2005) Ubiquitylation and cell signaling, EMBO J. 24,

Jackson, S. P., and Durocher, D. (2013) Regulation of DNA Damage Responses

Rieser, E., Cordier, S. M., and Walczak, H. (2013) Linear ubiquitination: a newly

Chen, Z. J., and Sun, L. J. (2009) Nonproteolytic functions of ubiquitin in cell

Komander, D., and Rape, M. (2012) The ubiquitin code, Annual review of

Komander, D., Reyes-Turcu, F., Licchesi, J. D., Odenwaelder, P., Wilkinson, K.

Hurley, J. H., Lee, S., and Prag, G. (2006) Ubiquitin-binding domains, Biochem.

Watanabe, K., Tateishi, S., Kawasuji, M., Tsurimoto, T., Inoue, H., and

Shiomi, N., Mori, M., Tsuji, H., Imai, T., Inoue, H., Tateishi, S., Yamaizumi, M., 29 Environment ACS Paragon Plus

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

562

and Shiomi, T. (2007) Human RAD18 is involved in S phase-specific single-strand break

563

repair without PCNA monoubiquitination, Nucleic Acids Res. 35.

564

10.

565

in DNA double-strand break repair, DNA Repair (Amst) 9, 1241-1248.

566

11.

567

Odijk, H., Roest, H. P., de Boer, P., and Hoeijmakers, J. H. (2004) Ubiquitin ligase

568

Rad18Sc localizes to the XY body and to other chromosomal regions that are unpaired

569

and transcriptionally silenced during male meiotic prophase, J. Cell Sci. 177.

570

12.

571

(2003) Enhanced genomic instability and defective postreplication repair in RAD18

572

knockout mouse embryonic stem, cells. Mol. Cell. Biol. 23, 474-481.

573

13.

574

transmits DNA damage signalling to elicit homologous recombination repair, Nat Cell

575

Biol. 11, 592-603.

576

14.

577

(2013) A small ubiquitin binding domain inhibits ubiquitin-dependent protein recruitment

578

to DNA repair foci, Cell Cycle 12, 3749-3758.

579

15.

580

M., and Sixma, T. K. (2007) Functional characterization of Rad18 domains for Rad6,

581

ubiquitin, DNA binding and PCNA modification, Nucleic Acids Res. 35, 5819-5830.

Ting, L., Jun, H., and Junjie, C. (2010) RAD18 lives a double life: Its implication

van der Laan, R., Uringa, E. J., Wassenaar, E., Hoogerbrugge, J. W., Sleddens, E.,

Tateishi, S., Niwa,H., Miyazaki,J., Fujimoto,S., Inoue,H. and Yamaizumi,M.

Huang J, H. M., Kim H, Leung CC, Glover JN, Yu X, Chen J. (2009) RAD18

Helchowski, C. M., Skow, L. F., Roberts, K. H., Chute, C. L., and Canman, C. E.

Notenboom, V., Hibbert, R. G., van Rossum-Fikkert, S. E., Olsen, J. V., Mann,


Page 30 of 49

Page 31 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

582

16.

Miyase, S., Tateishi, S., Watanabe, K., Tomita, K., Suzuki, K., Inoue, H., and

583

Yamaizumi, M. (2005) Differential regulation of Rad18 through Rad6-dependent mono-

584

and polyubiquitination, J. Biol. Chem. 280.

585

17.

586

Hofmann, K., Sixma, T. K., and Dikic, I. (2008) Human Wrnip1 is localized in replication

587

factories in a ubiquitin-binding zinc finger-dependent manner, J. Biol. Chem. 283, 35173-

588

35185.

589

18.

590

RNF168 binds and amplifies ubiquitin conjugates on damaged chromosomes to allow

591

accumulation of repair proteins, cell 136, 435-446.

592

19.

593

K., Hoeijmakers, J. H., Grootegoed, J. A., and Baarends, W. M. (2011) Human RAD18

594

interacts with ubiquitylated chromatin components and facilitates RAD9 recruitment to

595

DNA double strand breaks, PloS one 6, e23155.

596

20.

597

structure of the human Rad18 zinc finger in complex with ubiquitin defines a class of

598

UBZ domains in proteins linked to the DNA damage response, Biochemistry 53, 5895-

599

5906.

600

21.

601

the ubiquitin-binding zinc finger domain of human DNA Y-polymerase eta, EMBO Rep.

602

8, 247-251.

Crosetto, N., Bienko, M., Hibbert, R. G., Perica, T., Ambrogio, C., Kensche, T.,

Doil, C., Mailand, N., Bekker-Jensen, S., Menard, P., and Larsen, D. H. (2009)

Inagaki, A., Sleddens-Linkels, E., van Cappellen, W. A., Hibbert, R. G., Sixma, T.

Rizzo, A. A., Salerno, P. E., Bezsonova, I., and Korzhnev, D. M. (2014) NMR

Bomar, M. G., Pai, M. T., Tzeng, S. R., Li, S. S., and Zhou, P. (2007) Structure of


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

603

22.

Panier, S., Ichijima, Y., Fradet-Turcotte, A., Leung, C. C., Kaustov, L.,

604

Arrowsmith, C. H., and Durocher, D. (2012) Tandem protein interaction modules

605

organize the ubiquitin-dependent response to DNA double-strand breaks, Mol. Cell. 47,

606

383-395.

607

23.

608

Purification Problems of RhoGDI Using a Family of “Parallel” Expression Vectors,

609

Protein Expr. Pur. 15, 34-39.

610

24.

611

for High-Level Production of TEV Protease by Superfolder GFP Tag, J. . Biomed.

612

Biotechnol. 2009, 8.

613

25.

614

microgram quantities of protein utilizing the principle of protein-dye binding, Anal.

615

Biochem. 72, 248-254.

616

26.

617

In Methods in Enzymology (Raymond, J. D., Ed.), pp 21-36, Academic Press.

618

27.

619

D. M., Huang, O. W., Fedorova, A. V., Kirkpatrick, D. S., Hymowitz, S. G., and Dueber,

620

E. C. (2011) Preparation of distinct ubiquitin chain reagents of high purity and yield,

621

Structure 19, 1053-1063.

622

28.

623

structure prediction and refinement, Nucleic Acids Res. 40, W294-297.

Sheffield, P., Garrard, S., and Derewenda, Z. (1999) Overcoming Expression and

Wu, X., Wu, D., Lu, Z., Chen, W., Hu, X., and Ding, Y. (2009) A Novel Method

Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of

Pickart, C. M., and Raasi, S. (2005) Controlled Synthesis of Polyubiquitin Chains,

Dong, K. C., Helgason, E., Yu, C., Phu, L., Arnott, D. P., Bosanac, I., Compaan,

Ko, J., Park, H., Heo, L., and Seok, C. (2012) GalaxyWEB server for protein


Page 32 of 49

Page 33 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

624

29.

Raman, S., Vernon, R., Thompson, J., Tyka, M., Sadreyev, R., Pei, J., Kim, D.,

625

Kellogg, E., DiMaio, F., Lange, O., Kinch, L., Shffler, W., Kim, B.-H., Das, R., Grishin,

626

N. V., and Baker, D. (2009) Structure prediction of CASP8 with all-atom refinement

627

using Rosetta, Proteins 77, 89-99.

628

30.

629

York.

630

31.

631

proteins as identified through small angle X-ray scattering, Gen. Physiol. Biophys. 28,

632

174-189.

633

32.

634

F. A. (2010) Determination of the molecular weight of proteins in solution from a single

635

small-angle X-ray scattering measurement on a relative scale, J. Appl. Cryst. 43, 101-109.

636

33.

637

shape determination in small-angle scattering., J. Appl. Cryst. 42, 342-346.

638

34.

639

docking restrained by a small angle X-ray scattering profile., J. Struct Biol. 173, 461-471.

640

35.

641

macromolecular complexes against small-angle scattering data., Biophys. J. 89, 1237-

642

1250.

643

36.

Guinier, A. F., and Fournet, F. (1955) Small Angle X-rays, Wiley Interscience,New

Pelikan, M., Hura, G. L., and Hammel, M. (2009) Structure and flexibility within

Fischer, H., Oliveira Neto de, M., Napolitano, B. H., Polikarpov, I., and Craievich,

Franke, D., and Svergun, D. I. (2009) DAMMIF, a program for rapid ab-initio

Schneidman-Duhovny, D., Hammel, M., and Sali, A. (2010) Macromolecular

Petoukhov, M. V., and Svergun, D. I. (2005) Global rigid body modelling of

Schneidman-Duhovny, D., Hammel, M., Tainer, J. A., and Sali, A. (2013)


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

644

Accurate SAXS profile computation and its assessment by contrast variation

645

experiments, Biophys. J. 105, 962-974.

646

37.

647

Scientific LLC, San Carlos, CA.

648

38.

649

ubiquitin binding domains of RAP80 is critical for lysine 63-linked polyubiquitin-

650

dependent binding activity, BMB reports 42, 764-768.

651

39.

652

and Kato, K. (2010) Crystal structure of cyclic Lys48-linked tetraubiquitin, Biochemical

653

and biophysical research communications 400, 329-333.

654

40.

655

A., Orte, A., Klenerman, D., Jackson, S. E., and Komander, D. (2012) Ubiquitin chain

656

conformation regulates recognition and activity of interacting proteins, Nature 492, 266-

657

270.

658

41.

659

Epigenomics 3, 307-321.

660

42.

661

M., Bonifacino, J. S., and Hurley, J. H. (2006) Structural basis for ubiquitin recognition

662

and autoubiquitination by Rabex-5, Nat. Struct. Mol. Biol. 13, 264-271.

663

43.

DeLano, W. L. (2002) The PyMOL Molecular Graphics System, DeLano

Cho, H. J., Lee, S., and Kim, H. (2009) The linker connecting the tandem

Satoh, T., Sakata, E., Yamamoto, S., Yamaguchi, Y., Sumiyoshi, A., Wakatsuki, S.,

Ye, Y., Blaser, G., Horrocks, M. H., Ruedas-Rama, M. J., Ibrahim, S., Zhukov, A.

Bao, Y. (2011) Chromatin response to DNA double-strand break damage,

Lee, S., Tsai, Y. C., Mattera, R., Smith, W. J., Kostelansky, M. S., Weissman, A.

Penengo, L., Mapelli, M., Murachelli, A. G., Confalonieri, S., Magri, L.,


Page 34 of 49

Page 35 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

664

Musacchio, A., Di Fiore, P. P., Polo, S., and Schneider, T. R. (2006) Crystal structure of

665

the ubiquitin binding domains of rabex-5 reveals two modes of interaction with ubiquitin,

666

Cell 124, 1183-1195.

667

44.

668

(2009) Structural basis for specific recognition of Lys 63-linked polyubiquitin chains by

669

tandem UIMs of RAP80, EMBO J. 28, 2461-2468.

670

45.

671

63-linked polyubiquitin-binding preference of rap80, Mol. Cell 33, 775-783.

672

46.

673

H., and Shirakawa, M. (2012) NMR analysis of Lys63-linked polyubiquitin recognition

674

by the tandem ubiquitin-interacting motifs of Rap80, J. Biomol. NMR 52, 339-350.

675

47.

676

RAP80 on linear polyubiquitin binding probed by calorimetric analysis, Bull. Korean

677

Chem. Soc. 33, 1285–1289.

678

48.

679

UMI, a novel RNF168 ubiquitin binding domain involved in the DNA damage signaling

680

pathway, Mol. Cell. Biol. 31, 118-126.

681

49.

682

inhibition of the DNA double-strand break response, Nat. Rev. Mol. Cell. Biol. 14, 661-

683

672.

Sato, Y., Yoshikawa, A., Mimura, H., Yamashita, M., Yamagata, A., and Fukai, S.

Sims, J. J., and Cohen, R. E. (2009) Linkage-specific avidity defines the lysine

Sekiyama, N., Jee, J., Isogai, S., Akagi, K., Huang, T. H., Ariyoshi, M., Tochio,

Thach, T. T., Jee, J., and Lee, S. (2012) Effects of a phosphomimetic mutant of

Pinato, S., Gatti, M., Scandiuzzi, C., Confalonieri, S., and Penengo, L. (2011)

Panier, S., and Durocher, D. (2013) Push back to respond better: regulatory


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

684

50.

Lai, M. Y., Zhang, D., Laronde-Leblanc, N., and Fushman, D. (2012) Structural

685

and biochemical studies of the open state of Lys48-linked diubiquitin, Biochimica et

686

biophysica acta 1823, 2046-2056.

687

51.

688

and Elledge, S. J. (2007) Abraxas and RAP80 form a BRCA1 protein complex required

689

for the DNA damage response, Science 316, 1194-1198.

Wang, B., Matsuoka, S., Ballif, B. A., Zhang, D., Smogorzewska, A., Gygi, S. P.,

690


Page 36 of 49

Page 37 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

691

Biochemistry

Table 1. Binding affinities of Rad18 constructs with linear polyubiquitin chains

Rad18 a

692 693 694 695

Kd (µM) Linear Ub3 1.5 ± 0.1 6.9 ± 2.3 10.1 ± 0.1 n.d. b 3.0 ± 0.1

Ub Linear Ub2 171-240 9.1 ± 2.0 1.5 ± 0.6 171-234 13.5 ± 2.1 4.7 ± 1.4 171-224 11.2 ± 3.5 14.2 ± 1.1 b 171-200 n.d. n.d. b 4.4 ± 1.1 2.2 ± 0.9 201-240 For all measurements, the goodness of fit, r2 ≥ 0.90. a GST-Rad18 was used. Numbers indicate construct boundaries. b n.d., not determined.

696


Linear Ub4 1.4 ± 0.3 4.3 ± 0.5 13.6 ± 1.5 n.d. b 2.3 ± 0.4

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

697

Table 2. Data-collection and scattering-derived parameters for SAXS measurements Rad18(201-240): linear Ub2

698 699 700

Data-collection parameters Synchrotron beamline Beam geometry Wavelength (Å) Minimum scattering angle q (°) a q range (Å-1) Exposure time (s) Concentration range (mg/mL)

PAL-4C Mica window solid cell/Oscillation capillary 1.24 0.00128 0.009-0.248 30 0.98 - 3.91

Sample parameters Purity of sample (%, by SDS-PAGE) Polydispersity (%, by DLS) Temperature (K)

99 19 293.15

Structural parameters I(0) (cm-1) [from Guinier] Rg (Å) [from Guinier] I(0) (cm-1) [from P(r)] Rg (Å) [from P(r)] Dmax (Å) Porod volume estimate (Å3) Dry volume calculated from sequence (Å3)

0.0194 ± 0.0002 20.92 ± 1.38 0.0201 ± 0.0001 21.79 ± 0.29 68.38 ± 0.28 35592 ± 3659 26198

Molecular-mass determination b Mr [from GNOM file] (Da) Mr [from sequence] (Da) Relative discrepancy (%)

22400 21653 3.5

Software used Primary data reduction RAW Data processing PRIMUS Ab initio analysis DAMMIF Validation and averaging DAMAVER Three-dimensional presentations PyMOL a A scattering vector, q b Determined from the sample at the concentration of 3.91 mg/mL.


Page 38 of 49

Page 39 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

701

FIGURE LEGENDS

702

Figure 1. Rad18 specifically recognizes linkage polyubiquitin chains. (A) Rad18 constructs were

703

used. Residue numbers for the boundaries of domains and the constructs are displayed. UBZ:

704

ubiquitin binding zinc finger domain, LRM: ubiquitylated ligand recognition motif. (B) GST

705

pull-down assays between GST-Rad18(171-240) and K63-linked, K48-linked, and linear Ub2,

706

respectively. (C, D) Binding affinity was determined by steady-state fitting of the response at

707

equilibrium (Req) values from BLI experiments.

708

Figure 2. Binding affinity between Rad18 constructs and linear polyubiquitin chains. (A)

709

Binding affinity was determined by steady-state fitting of Req values from BLI experiments. (B)

710

Quantification of Kd from steady-state fitting. Significant difference was analyzed by Student’s t-

711

test (* P ≤ 0.05, ** P ≤ 0.01).

712

Figure 3. Ab initio models for Rad18(201-240) structure. (Left) The GalaxyTMB and (Right)

713

Rosetta ab initio models for Rad18(201-240) structure with regions like UBZ domain (forest);

714

region between UBZ domain and LR motif (blue); LR motif (orange). The Zn2+ ion is indicated

715

as the magenta sphere.

716

Figure 4. The best ab initio envelop fitting for Rad18(201-240):linear Ub2 complex. (A) Pair

717

distribution function of the Rad18(201-240):linear Ub2 complex. (B) A comparison of

718

experiment scattering profile (black) and minimal ensemble search (red). The minimal ensemble

719

that best fitted the experimental data was identified using FoXSMES. (C) The best ab initio

720

envelop for the Rad18(201-240):linear Ub2 complex.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 49

721

Figure 5. Ubiquitin binding affinity of Rad18 mutants. (A) Proposed distribution of mutant

722

residues in Rad18(201-240):linear Ub2 complex. The I44-centered hydrophobic patch on

723

ubiquitin, which is believed interact with the ubiquitin-binding domain is indicated by the green

724

color. (B) Binding affinity was determined by steady-state fitting of Req values from BLI

725

experiments.

726

Figure 6. ELRM is important in recruitment of Rad18 to IRIF. (A) Localization of EGFP-

727

Rad18(171-240)-SV40NLS constructs in IRIF at 5 hr post 10 Gy IR in mammalian 293T cells.

728

(B) Relative IRIF formation of mutants and WT was quantified. Over 50 cells were counted in

729

each experiment. Standard deviations are shown from duplicate or triplicate experiments.

730

Figure 7. Rad18 UBZ-ELRM and RAP80 LRM-UIMs compete for binding with ubiquitin.

731

Rad18(171-240) and RAP80(60-124) were used. (A) Rad18 UBZ-ELRM interferes with the

732

interaction between RAP80 LRM-UIMs and linear Ub2. GST pull-down (left panels) and

733

percentage of band intensity of ratio between GST-RAP80(60-124) and linear Ub2 in the

734

presence or absence of Rad18(171-240) (right panels). (B) The reciprocal GST pull-down

735

experiments between GST-Rad18(171-240) and linear Ub2 in the presence or absence of

736

RAP80(60-124). Band intensity values were calculated from triplicate experiments.

737

Figure 8. Quantitative analysis of competitive binding of Rad18 and RAP80 to polyubiquitin.

738

Binding affinities of linear Ub2 with Rad18(171-240) and RAP80(60-124) in the absence or

739

presence of RAP80(60-124) or Rad18(171-240), respectively. (A) BLI sensorgrams of

740

competitive binding. (B) Kd values determined by steady-state fitting of the data in (A).

741


Page 41 of149 Figure

A

171

201 ATP

232

224

UBZ

AMP

240

LRM (171-240) (171-234) (171-224) (171-200) (201-240)

124 -R ST G

In p

ST

ad 18 (1 7

124 ut

-R ST G

G

ut In p

ST

ad 18 (1 7

124 ad 18 (1 7 -R ST G

ST G

ut In p

0)

Linear Ub2 0)

K48-Ub2 0)

K63-Ub2

G

B

IB: Ub

17

34 IB: GST

26 (kDa)

C

D 2.0

K63-Ub2 Linear Ub2

Kd (μM)

1.0 0.5 0.0

30

*

K48-Ub2

1.5 Req (nm)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

0

5

10

15

20

25

20

10

0

Linear Ub2

[GST-Rad18(171-240)] (μM)

ACS Paragon Plus Environment

K63-Ub2

K48-Ub2

Figure 2

Page 42 of 49

His-Linear Ub

Req (nm)

Req (nm)

A

Ub2 Ub3 Ub4

[GST-Rad18(171-224)] (μM)

[GST-Rad18(171-234)] (μM)

Req (nm)

Req (nm)

[GST-Rad18(171-240)] (μM)

[GST-Rad18(201-240)] (μM)

B 20

His-Linear Ub

15 Kd (µM)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Ub2 Ub3 Ub4

10

* *

5 0

**

** **** 171-240

n.d. 171-234 171-224 171-200 201-240 GST-Rad18


Page 43 of349 Figure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

C C N


N

Figure 4

A

B

0.06

Page 44 of 49

100

0.04

Exp_data Exp_fit MES

P (r)

10 log I

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

0.02

0.00

1

0

20

40 r (Å)

60

80

0.1

0

0.05

0.10 q (Å-1)

0.15

C

N

I44

C

M77

I120

G76

Distal Ub


Proximal Ub

0.20


A V237 H238

E228

N

C R226

K239

E231 E227

R240

D221

B 15

His-Linear Ub Ub2

10 Kd (µM)

*

5

* *

**

T

0 W

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

1A

2 D2

*

*

*

A A A A A/ A A A A 26 227 228 27 A 231 237 238 239 240 2 K E E2 28 E E V H R R E2


Biochemistry

Figure 6

A

GFP

γH2AX

DAPI

WT

D221A EGFP-Rad18(171-240)-SV40NLS

E227A

E228A

E227A/ E228A

B

Percentage of EGFP-Rad18(171-240)-SVL40NLS DNA foci formation/ γH2AX DNA foci formation

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 46 of 49

60

40

20

0

WT

D221A

E227A

E228A

E227A/E228A


Merge


1

2

3

4

5

6

7

8

9

Rad18(171-240)

- + -

-

+ +

+ -

+ -

+ +

+ +

+ -

GST-RAP80(60-124)

- +

-

-

+

-

+

-

-

GST

- -

+

-

-

+

-

+

-

Linear Ub2

IB:GST

100 GST-RAP80(60-124)

A

50

**

0

R

Li ne ar

2

U ad b 18 2 (1 7 Li 1 ne -2 ar 40 U ): b

IB:His

2

3

4

5

6

7

8

9

RAP80(60-124)

- + -

-

+ +

+ -

+ -

+ +

+ +

+ -

GST-Rad18(171-240)

- +

-

-

+

-

+

-

-

GST

- -

+

-

-

+

-

+

-

IB:GST

50

0

Li

ne

ar

U

2

b

IB:His

2

Linear Ub2

100

Li (60 ne -1 ar 24 U ): b

1

R

AP

80

B

GST-Rad18(171-240)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry


Figure 8

Page 48 of 49

3.0

5.0 4.0 Req (nm)

2.0

1.0

0.0

Linear Ub2

5

10

15

20

2.0 Linear Ub2

1.0

Linear Ub2 Rad18

0

3.0

0.0

25

Linear Ub2 RAP80

0

[GST-RAP80(60-124)] (μM)

B

5

40 30 20 10

(1 Li 71 ne -2 ar 40 U ): b

2

b U R

ad

18

ne

ar

2

80 AP R

GST-Rad18(171-240)

Li

ne

ar

U

2

Li (60 ne -1 ar 24 U ): b

b

0

Li

Kd (μM)

10

15

20

[GST-Rad18(171-240)] (μM)

2

A

Req (nm)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

GST-RAP80(60-124)


25

Page 49 of 49 GRAPHICAL ABSTRACT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

171

Rad18

201 ATP

224

UBZ

AMP

232

240

LRM ELRM

V237 H238

E228

N

C R226

E231 E227

K239 R240

D221

Distal Ubiquitin

Proximal Ubiquitin


Molecular Determinants of Polyubiquitin Recognition by Continuous

Recommend Documents