Combined immunodeficiency caused by a loss-of-function mutation in DNA polymerase delta 1

Journal Pre-proof Combined Immunodeficiency due to a loss of function mutation in DNA Polymerase Delta 1 Ye Cui, PhD, Sevgi Keles, MD, Louis-Marie Charbonnier, PhD, Amélie M. Julé, PhD, Lauren Henderson, MD, MSc., Seyma Celikbilek Celik, BSc, Ismail Reisli, MD, Chen Shen, PhD, Wen Jun Xie, PhD, Klaus Schmitz-Abe, PhD, Hao Wu, PhD, Talal A. Chatila, MD, MSc PII:

S0091-6749(19)31322-3

DOI:

https://doi.org/10.1016/j.jaci.2019.10.004

Reference:

YMAI 14215

To appear in:

Journal of Allergy and Clinical Immunology

Received Date: 5 July 2019 Revised Date:

11 September 2019

Accepted Date: 4 October 2019

Please cite this article as: Cui Y, Keles S, Charbonnier L-M, Julé AM, Henderson L, Celik SC, Reisli I, Shen C, Xie WJ, Schmitz-Abe K, Wu H, Chatila TA, Combined Immunodeficiency due to a loss of function mutation in DNA Polymerase Delta 1, Journal of Allergy and Clinical Immunology (2019), doi: https://doi.org/10.1016/j.jaci.2019.10.004. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of the American Academy of Allergy, Asthma & Immunology.

1

Combined Immunodeficiency due to a loss of function mutation in DNA

2

Polymerase Delta 1

3

Ye Cui, PhD, Sevgi Keles, MD, Louis-Marie Charbonnier, PhD, Amélie M. Julé,

4

PhD,1 Lauren Henderson, MD, MSc. Seyma Celikbilek Celik, BSc, Ismail Reisli, MD,

5

Chen Shen, PhD,

6

Talal A. Chatila, MD, MSc1

7

1

8

Harvard Medical School, Boston;

9

Faculty, Division of Pediatric Allergy and Immunology, Konya; 3Program in Molecular

10

and Cellular Medicine, Department of Pediatrics, Boston Children's Hospital, Harvard

11

Medical School, Boston;

12

Pharmacology,

13

Massachusetts Institute of Technology, Cambridge; 6Division of Newborn Medicine and

14

Neonatal Genomics Program, Boston Children’s Hospital, Harvard Medical School,

15

Boston.

16

Corresponding Author: Talal A. Chatila at the Division of Immunology, Boston

17

Children’s Hospital and the Department of Pediatrics, Harvard Medical School.

18

Address: Karp Family Building, Room 10-214. 1 Blackfan Street, Boston, MA 02115

19

Phone: +1-617-9193529; Fax: 617-3557228; Email: [email protected]

1

2

1

1

3,4

2

5

6

1,3

Wen Jun Xie, PhD, Klaus Schmitz-Abe, PhD, Hao Wu, PhD,

Division of Immunology, The Boston Children's Hospital. Department of Pediatrics,

Harvard

2

Necmettin Erbakan University, Meram Medical

4

Department of Biological Chemistry and Molecular

Medical

School,

Boston;

5

Department

20 21

2

Conflict of Interest Statement: The authors declare no conflict of interest.

of

Chemistry,

22 23

Abstract

24

Background: Mutations affecting DNA polymerases have been implicated in genomic

25

instability and cancer development, but the mechanisms by which they may impact the

26

immune system remain largely unexplored.

27

Objective: To establish the role of POLD1, encoding the DNA polymerase δ 1 catalytic

28

subunit, as the cause of a primary immunodeficiency in an extended kindred.

29

Methods: We performed whole-exome and targeted gene sequencing, lymphocyte

30

characterization, molecular and functional analyses of the DNA polymerase delta (Polδ)

31

complex, and T and B cell antigen receptor repertoire analysis.

32

Results: We identified a missense mutation (c. 3178C>T; p.R1060C) in POLD1 in three

33

related subjects who presented with recurrent, especially herpetic, infections and T cell

34

lymphopenia with impaired T cell but not B cell proliferation. The mutation destabilizes

35

the Polδ complex, leading to ineffective recruitment of replication factor C to initiate DNA

36

replication. Molecular Dynamics simulation revealed that the R1060C mutation disrupts

37

the intramolecular interaction between the POLD1 CysB motif and catalytic domain, and

38

also between POLD1 and the DNA polymerase δ subunit POLD2. The patients

39

exhibited decreased naïve CD4 and especially CD8 T cells in favor of effector memory

40

subpopulations. This skewing was associated with oligoclonality and restricted T cell

41

receptor beta chain V-J pairing in CD8+ but not CD4+ T cells, suggesting that

42

POLD1R1060C differentially impacts peripheral CD8+T cell expansion and possibly thymic

43

selection.

44

Conclusion. These results identify gene defects in POLD1 as a novel cause of T cell

45

immunodeficiency.

46

Capsule Summary

47

A Loss of function mutation in POLD1 causes impaired assembly and function of the

48

DNA polymerase delta (Polδ) complex, resulting in a T cell immunodeficiency.

49

Key Messages:

50

–A POLD1 mutation disrupted the assembly of the Polδ complex and resulted in T cell

51

immunodeficiency.

52

– Patient CD8+ T cells exhibited severe oligoclonality and restricted T cell receptor beta

53

chain V-J pairing.

54

Keywords.

55

Immunodeficiency, Whole Exome Sequencing.

56

Abbreviations. BrdU: Bromodeoxyuridine / 5-bromo-2'-deoxyuridine; LRTI: lower

57

respiratory tract infection; PCNA: Proliferating Cell Nuclear Antigen, Polδ: DNA

58

polymerase delta; POLA1: DNA polymerase alpha 1 catalytic subunit; POLBc: DNA

59

polymerase type B family catalytic domain; POLD1, DNA polymerase delta 1 catalytic

60

subunit; RFC: Replication Factor C; TCR: T cell receptor; IGH: Immunoglobulin heavy

61

chain; Teff: T effector cells, Treg: Regulatory T cells, URTI: upper respiratory tract

62

infection; WES: Whole exome sequencing.

63

DNA

Polymerase

Delta,

POLD1,

Replication

Factor

C,

Primary

64

Introduction

65

DNA replication is a fundamental process for maintaining cellular homeostasis 1. DNA

66

polymerase δ (Polδ), one of the three family B polymerases in eukaryotes, is essential

67

for the leading and lagging strand synthesis

68

tetramer that includes four subunits: POLD1-4 5. POLD1 functions as the catalytic

69

subunit, which is endowed with both polymerase and exonuclease activities and which

70

plays a critical role in several synthetic and DNA-repair processes

71

deficiency is embryonic lethal in mice, while deficiency of POLD1 exonuclease activity in

72

Pold1exo/exo mice (mutator mice), results in a high rate of DNA replication errors 8.

73

POLD2, POLD3, and POLD4 are accessory subunits that interact with other nuclear

74

proteins to regulate the activity and stability of the Polδ complex

75

POLD2 serves as a scaffold by interacting with POLD1 and the other Polδ subunits

76

Additionally, the Polδ complex coordinately interacts with a number of proteins that

77

enable its function, including DNA replication factor C (RFC) and Proliferating Cell

78

Nuclear Antigen (PCNA) 13.

79

Studies have shown that mutations in POLD1 in mice and humans lead to genomic

80

instability, hypermutator phenotype and carcinogenesis

81

mutations in POLD1 proof-reading (exonuclease) domain have been identified in

82

inherited colorectal cancers

83

maps to the catalytic domain and which abrogates the DNA polymerase but not the

84

exonuclease activity has been identified in a developmental disorder of mandibular

85

hypoplasia, sensorineural hearing loss, progeroid features and lipodystrophy with insulin

. In mammals, Polymerase δ is a hetero-

2-4

6, 7

. Total POLD1

9-11

. In particular, 12

.

14-16

. Damaging heterozygous

17

. A POLD1 heterozygous single amino acid deletion that

18, 19

86

resistance

87

distinct disorders and phenotypes.

88

Here we identify a novel POLD1 mutation that affects the stability of the Polδ complex,

89

resulting in a disorder distinct from those caused by other POLD1 mutations. The

90

patients presented with a combined immunodeficiency disorder associated with T cell

91

lymphopenia, CD8+ T cell oligoclonality and repertoire restriction, indicative of a

92

particularly important role for POLD1 in CD8 T cell expansion.

93

. Thus, mutations affecting different domains of POLD1 give rise to

94

Methods

95

Patient Studies. All study participants were recruited after obtaining informed consent

96

at the referring institution (Necmettin Erbakan University, Meram Medical Faculty,

97

Konya, Turkey), and the studies were conducted at the Boston Children’s Hospital

98

under approved protocol #04-09-113R.

99

Whole exome sequencing (WES). WES was performed on genomic DNA of 2 affected

100

siblings (P1 and P2), their parents, and their healthy sister through Axeq (Rockville, Md).

101

The Agilent SureSelect Target enrichment kit was used for exon capture (Agilent

102

Technologies, Santa Clara, Calif). Paired-end sequencing was performed with an

103

Illumina HiSeq2000 instrument (Illumina, San Diego, Calif), which generated 150 base

104

pair reads. Average coverage in WES was 53X, covering 96% of the coding regions.

105

Analysis of WES data was performed using “Variant Explorer Pipeline” (VExP)20 to

106

narrow down potential candidate variants. VExP is a validated comprehensive system

107

that integrates existing methods, genetic information, and probabilistic models into an

108

automated pipeline for the identification of disease genes. Raw data were processed,

109

filtered and

110

information). Candidate genes that passed the criteria for the 2 affected samples were

111

further evaluated by our research team (Table E1 in the Online Repository). The

112

identified POLD1 c. 3178C>T mutation was confirmed by Sanger sequencing by first

113

generating a 404 base pair amplicon from genomic DNA using the following primers:

114

forward

115

GAGAGGCCTTGGAGTCAGAG-3'. The amplicon was then sequenced for the presence

116

of the mutation using the following primer: 5’-GCCTACATGAAGTCGGAGGA-3’. Sanger

analyzed

primer

according

VExP recommendations

5'-AGAAGCTGGGATTGGCAGT-3'

and

(see

reverse

supplementary

primer

5'-

117

sequencing analysis was also used to screen other family members for the POLD1

118

mutation.

119

Antibodies and flow cytometry. Anti-human monoclonal antibodies (mAbs) to the

120

following antigens were used for staining: CD3 (SK7), CD4 (SK3), CD8 (SK1), CD45RA

121

(HI100), CD45RO (UCHL1), CCR7 (150503), CD31 (L133.1),TCR A/B (WT31) and TCR

122

G/D (11F2), CD16+56 (B73,1MY3I), CD19 (SI25C1), IgD (IA6-2), CD27 (L128) (BD

123

Biosciences) and the appropriate isotype controls. The monoclonal antibody against

124

POLD1(sc-17776) was obtained from Santacruz. Antibodies against V5 was purchased

125

from Biolegend (903801), POLD2 (HPA026745) and RFC1 (HPA046116) were

126

purchased from Sigma-Aldrich, POLD3 (A301-244A-M) was from Bethyl Laboratories

127

and β-actin was from Cell Signaling Technology. Whole blood was incubated with mAbs

128

against surface markers for 30 min on ice. Intracellular staining with FOXP3 was

129

performed using eBioscience Fixation/Permeabilization buffer according to the

130

manufacturer’s instructions. Data were collected with a Fortessa cytometer (BD

131

Biosciences) and analyzed with FlowJo software (TreeStar).

132

Cell culture and transfection. HEK293T (American tissue culture collection), and

133

Fibroblast cells were cultured using DMEM (Invitrogen) plus 10% FBS (Gibco),

134

supplemented with 1% penicillin-streptomycin (Invitrogen). Peripheral blood samples (5

135

ml) were collected from each subject and stored in tubes containing EDTA. PBMCs

136

were isolated by Ficoll-Paque PLUS (GE Healthcare) gradient centrifugation, and

137

cultured in RPMI 1640 with 10% FBS and 1% pen-strep (with Non-Essential Amino Acid,

138

Sodium Pyruvate）with anti-CD3/CD28 beads (from Invitrogen). B cell proliferation was

139

quantified by culturing PBMCs after stimulation with anti-CD40 mAb (10 µg/mL; R&D,

140

6245-CL-050) plus IL-21 (30 ng/mL; Peprotech, 200-21) for 5 days. Lipofectamine 2000

141

(Invitrogen) was used for transient transfection of HEK293T Cells according to the

142

manufacturer’s instructions.

143

BrdU cell proliferation assay. Bromodeoxyuridine / 5-bromo-2'-deoxyuridine (BrdU)

144

incorporation assay in patient peripheral blood mononuclear cells (PBMCs) was

145

performed with the Phase-Flow™ FITC BrdU Kit from Biolegend (Catalog No. 370704)

146

following the manufacturer’s instructions. The same kit included 7-aminoactinomycin D

147

(7-AAD) for staining for total DNA.

148

Lentiviral transfections. HEK293T cells plated on 100-mm dishes were transfected

149

with the indicated lentiviral expression plasmid (14ug) together with the GAG (10ug), the

150

REV (5ug) and the VSV (2ug). After 48h, the viral particles were collected, filtered by

151

0.45um membrane filter and used to infect the indicated cells in the presence of

152

polybrene (6 ug/ml). After transfection for 48h, cells were cultured in complete RPMI

153

1640 medium and ready for BrdU assay.

154

Plasmids. POLD1_pLX307 and pLX307 vector were purchased from Addgene. The

155

POLD1 C3178T mutation was introduced into the POLD1 plasmid with the Q5 Site-

156

Directed

157

GCAGTGCCAGtGCTGCCAGGG-3'

158

5'GTCCAGAGGCGCGAGAAGC-3' Reverse primer. POLD1 mutants were confirmed

159

using Sanger sequencing.

160

Immunoprecipitation assay and immunoblot analysis. For immunoprecipitation

161

assay, cells extracts were prepared by using RIPA buffer (50 mM Tris-HCl pH 7.4, 150

162

mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.5% deoxycholate)

Mutagenesis

Kit

(NEB)

using

the

Forward

following primer

primers:

5'and

163

supplemented with a complete protease inhibitor cocktail (Roche), a PhosSTOP

164

phosphatase inhibitor cocktail (Roche). Lysates were incubated with the appropriate

165

antibody for four hours to overnight at 4°C before adding protein A/G agarose for 2 hr.

166

The immunoprecipitates were washed three times with the same buffer and eluted with

167

SDS loading buffer by boiling for 5 min.

168

For immunoblot analysis, the samples were subjected to SDS-PAGE. The resolved

169

proteins were then electrically transferred to a PVDF membrane (Millipore).

170

Immunoblotting was probed with indicated antibodies. The protein bands were

171

visualized by using a SuperSignal West Pico chemiluminescence ECL kit (Pierce).

172

Signal intensities of immunoblot bands were quantified by Image J software.

173

Molecular Modeling and Simulation. We performed molecular dynamics (MD)

174

simulation for the WT and mutant POLD1 to relax the protein structure obtained from

175

homolog modeling. The simulation region starts from the 795th residue. All MD

176

simulations were performed using the AMBER 14 suite of programs

177

was immersed into a simulated water box with the initial structure taken from molecular

178

docking. The force field for protein is AMBER parm10 force field

179

model was adopted for water

180

whose force field is obtained from Joung et al.23 The sizes of the simulation boxes are

181

about 81 Å *65 Å *93 Å. For the WT and mutant POLD1, we carried out NPT ensemble

182

simulation with a 10 ns long trajectory after the initial equilibration. The Berendsen

183

weak-coupling barostat was used to control the pressure at 1 bar

184

was controlled using Langevin dynamics at 300 K. SHAKE algorithm was used to

185

constrain all bonds involving hydrogen atoms

21

. POLD1 protein

21

, and the SPC/E

22

. Chloride ions were added to neutralize the system

25

24

. The temperature

. A cutoff of 10.0 was adopted for

186

nonbonding interactions. For the long-range electrostatic interactions, the Particle Mesh

187

Ewald method was applied.

188

The overall protein secondary and tertiary structure persist during the simulation,

189

suggesting that homolog modeling gives a good prediction of POLD1 structure. A

190

hydrogen bond is considered formed only if the distance between the heavy atom of

191

hydrogen bond donor and acceptor is less than 3.5 Å and the N–H–O angle is greater

192

than 135°.

193

Repertoire analysis. CD4+ T(CD3+CD4+) cells, CD8+ (CD3+CD8+) T cells and B (CD3–

194

CD19+) cells were isolated from fresh PBMCs obtained from POLD1R1060C patients and

195

healthy controls. Genomic DNA was isolated with QIAamp DNA Blood Mini Kit

196

(QIAGEN) and sent to Adaptive Biotechnologies (Seattle, WA) for immune repertoire

197

analysis. A multiplex PCR protocol was used to amplify the CDR3 of the sorted

198

lymphocytes using a standard quantity of DNA as the template. The Illumina platform

199

was used for sequencing of the PCR products. The sequences were aligned to a

200

reference genome, and T cell receptor β chain (TRB) and Immunoglobulin heavy chain

201

(IGH) variable, diversity, and joining (VDJ) gene definitions were based on the

202

International ImMunoGeneTics system (IMGT)26. The data were filtered and clustered

203

using the relative frequency ratio between similar clones and a modified nearest

204

neighbor algorithm to merge closely related sequences and remove both PCR and

205

sequencing errors, as described

206

analyzed for TCR and BCR V, D, and J usage with ImmunoSEQ AnalyzerTM set of

207

online tools and with R (version 3.6.1). Physicochemical properties of amino acid

208

sequences were analyzed using the Alakazam package29.Sequencing of the TRB and

27, 28

. Productive unique and total sequences were

209

IGH rearranged products was completed successfully in all samples (Table E3 in the

210

Online

211

https://clients.adaptivebiotech.com/.(

212

Password: cui-2019-review)

213

Statistical Analysis. Aggregate results are presented as means + S.E.M.. Comparison

214

between groups was carried out with 1-way ANOVA with Bonferroni post-test analysis,

215

as indicated. Differences in mean values were considered significant at a P value of

216

less than 0.05.

217

Repository).

The

primary

sequencing

Username:

data

are

available

at

[email protected]

218

Results

219

Identification of a novel immunodeficiency associated with a novel POLD1

220

mutation. P1 is a 14 years old female, born to consanguineous (first cousins) Turkish

221

parents (Fig 1, A), who suffered from recurrent upper and lower respiratory tract

222

infections (URTI and LRTI, respectively) starting early in life. At 3 years of age she was

223

diagnosed with sensorineural hearing loss following evaluation for language delay. She

224

underwent adenotonsillectomy at 5 1/2 years of age because of recurrent URTI and

225

serous otitis media. She suffered from severe Chickenpox at 6 years of age that

226

eventually resolved. Thereafter, she started experiencing LRTI at a frequency of 4-5

227

times/year, mainly in winter, requiring intravenous antibiotic therapy. Starting age 9

228

years, she had recurrent oral herpes infections every 1-2 months as well as several

229

episodes of recurrent herpes zoster for which she received acyclovir therapy.

230

Investigation into her recurrent herpetic infections revealed marked lymphopenia, for

231

which she was referred for immunological evaluation at 12 years of age. Her

232

immunological workup was particularly notable for profound CD3+CD4+ lymphopenia.

233

She had mildly decreased serum IgG and low IgA and IgM antibody concentrations,

234

while her tetanus vaccine-related antibody responses were absent despite having been

235

fully vaccinated (Table 1). She was started on immunoglobulin replacement therapy,

236

which resulted in the resolution of LRTI and markedly decreased herpetic infections.

237

Patient P2 is the younger sister of P1 and presented at age 3 with history of fever and

238

cough. Her weight (12,3kg; 10th-25th percentile) and height (98 cm; 50th-75th percentile)

239

were in normal range for her age. Her laboratory results revealed marked lymphopenia,

240

hypogammaglobulinemia and low CD3+CD4+ T cell number similar to her older sister

241

(Table 1). She also had an unprotective tetanus antibody response on initial

242

presentation despite her prior vaccination that normalized on booster vaccination. She

243

is currently 5 years of age and still suffers from recurrent croup and oral herpes

244

infections, for which she is treated symptomatically. Auditory testing revealed normal

245

hearing.

246

P3 is a 29 years old paternal aunt of P1 and P2 (Fig 1, A). She had encephalitis

247

following a primary varicella infection at 3 years of age, from which she was left with

248

mental retardation, hearing deficit and speech delay. In the ensuing years, she has had

249

recurrent oral herpetic infections, for which she is treated symptomatically. She also has

250

URTI in winter, requiring oral antibiotic therapy.

251

Pedigree analysis suggested an autosomal recessive inheritance of the disease. To

252

identify the underlying gene defect, we performed whole exome sequencing for the

253

whole family. Knowing that the family is consanguineous, we executed homozygosity

254

mapping using their WES data (see Supplementary Methods for more information). Five

255

regions of homozygosity were identified, and only the largest one found on chromosome

256

19, 14Mb in size, contains mutations with a minor allele frequency <0.01 that segregate

257

within family (Fig E1 and Table E2 in the Online Repository). Of the nine genes thus

258

identified in this region, including six candidate genes with homozygous recessive

259

mutations, the most promising was a homozygous C>T substitution in exon 26 of

260

POLD1 was identified (c.3178C>T; p.R1060C; based on POLD1 isoform 1,

261

NM_001256849.1) (Table E1 in the Online Repository). This missense homozygous

262

mutation has not been previously reported in genomAD, ExAC, dbSNP or 1000Genome

263

and was confirmed by Sanger sequencing (Fig 1, B). Both parents were heterozygous

264

for the mutation. Her sister P2, who also suffered from recurrent infections, was found

265

homozygous for the same mutation by Sanger sequencing, as was a paternal aunt P3,

266

who suffered from an early childhood encephalitis post varicella infection. A number of

267

extended family members had colon and rectal cancers but did not carry the mutant

268

allele.

269

The R1060C substitution localizes to the CysB motif at C-terminus of POLD1, away

270

from the exonuclease and polymerase domains (Fig 1, C). Immunoblot analysis

271

revealed decreased POLD1 expression in P1 and P2 compared to healthy sibling in

272

PBMCs and fibroblasts (Fig 1, D and E). Other components of the Polδ complex are

273

also decreased, including POLD2 and POLD3. Thus, the POLD1R1060C mutation exerted

274

a global effect on the Polδ complex.

275

POLD1R1060C affects the stability of polymerase δ complex. DNA polymerases have

276

been classified into several families based on their amino acid sequences. In

277

polymerase family B, both Polα and Polδ are key enzymes for eukaryotic nuclear DNA

278

replication30. The structure of the family B DNA polymerases has been conserved

279

throughout evolution, although their primary sequence varies. The structure of DNA

280

polymerase alpha 1, catalytic subunit (POLA1) is well studied31. To clarify the role of the

281

R1060C missense substitution in POLD1 from the family with immunodeficiency

282

syndrome, we generated the model of full-length POLD1 using Swiss modeling based

283

on the crystal structure of POLA1, which has a sequence identity of 25.9%. Most of the

284

structural elements aligned well between the POLD1 model and the POLA1 structure,

285

with an average root-mean-square deviation (RMSD) of 0.315 Å (Fig 2, A). Based on

286

the POLD1 model, we introduced the R1060C mutation in PyMol followed by molecular

287

dynamics (MD) simulation for the wild-type and mutant POLD1 to relax the protein

288

structure obtained from homolog modeling with a time scale of 10 ns. The simulation

289

region starts from the residue 795 in consideration of domain of interests and

290

computational cost. After we extract 10-ns snapshots of wild-type POLD1 and the

291

R1060C mutant protein, we found a significance distance change between the CysB

292

motif and the DNA polymerase type B family catalytic subunit (POLBc) delta domain

293

(Fig 2, B). The CysB motif in POLD1 shows an intact interaction with the POLBc delta

294

domain while the C1060 mutant lost the interaction. Detailed interface analysis showed

295

the hydrogen bonds between E830 and R1060, R968 and R1060 backbone, R808 and

296

Q1059, and the hydrophobic interaction between V832 and I1078 dominate the mutual

297

interaction between these two domains (Fig 2, C). By calculating the distance between

298

residue 1060 and 830 in WT and mutant POLD1, we found that the R1060-E830

299

interaction maintains a shorter distance than C1060-E830 after the structural relaxation

300

(Fig E2, A in the Online Repository), indicating that this interaction pair may be critical

301

to maintain the CysB- POLBc_delta interaction. Also, hydrogen bond tracing revealed

302

that the interaction between R1060 and E830 is highly stable during the whole

303

simulation trajectory (Fig E2, B in the Online Repository), which provided further

304

evidence that R1060 is critical for the intramolecular interaction of POLD1. Because of

305

the decrease expression of POLD2 in patient’s polymerase δ, we hypothesized that the

306

POLD1R1060C mutation affects the stability of polymerase δ complex. Considering that

307

the weaker intramolecular interaction in the mutant protein may further affects the

308

recruitment of downstream POLD subunits, we built both the wild-type and mutant

309

model of POLD1 10 ns snapshots/POLD2 complex based on the crystal structure of the

310

POLA1/POLA2 complex (PDB ID: 5EXR). Overall, the docking model clearly indicated

311

the cooperation between the CysB motif and the POLBc_delta domain in recruiting

312

POLD2. This cooperative interaction is weakened by the R1060C mutant, which may

313

disrupt POLD1:POLD2 complex formation (Fig 2, D).

314

To validate the results obtained with molecular modeling, we examined the capacity of

315

recombinant tagged WT and mutant POLD1 expressed in HEK293T cells to interact

316

with the endogenous POLD2. Co-immunoprecipitation studies showed that the R1060C

317

mutation interfered with the formation of the POLD complex, as indicated by the

318

capacity of WT POLD1 but not the POLD1R1060C mutant protein to co-precipitate with

319

POLD2 despite equal expression of the respective POLD1 proteins (Fig 2, E, F).

320

Furthermore, we examined the capacity of POLD1R1060C mutant protein to interact with

321

other components of the DNA replication complex. During DNA replication, Polδ

322

associates with replication factor C (RFC), a step necessary for loading Proliferating

323

Cell Nuclear Antigen (PCNA) onto DNA to initiate DNA replication. However, unlike WT

324

POLD1, POLD1R1060C failed to effectively coprecipitate with RFC, indicative of poor

325

association (Fig 2, E). Overall, these results indicate that that the POLD1R1060C mutation

326

interfered with molecular interactions involved in Polδ assembly and the formation of the

327

overall DNA replication complex.

328

POLD1R1060C results in lower cell proliferation. To determine whether the R1060C

329

mutation in POLD1 impacted DNA replication and cell proliferation, we analyzed the

330

proliferative activities of control and patient T cells upon treatment of PBMC with anti-

331

CD3+anti-CD28 mAb for 3 days followed by a 4-hour stimulation with phorbol myristate

332

acetate (PMA) and ionomycin32. The cells were pulsed with BrdU, which labels newly

333

synthesized DNA characteristic of the cell cycle S phase 33. Results showed that patient

334

T cells, but not B cells, were profoundly deficient in BrdU labeling following stimulation,

335

indicative of poor S phase DNA replication (Fig 3, A-D). To determine whether the poor

336

proliferation of patient T cells was directly related to the POLD1 defect, we undertook

337

rescue experiments in which patient T cells were expanded with anti-CD3+anti-CD28

338

mAb treatment, followed by transfection with a lentiviral vector encoding either wild-type

339

(WT) POLD1 or the POLD1R1060C mutant. Both the WT POLD1 and the POLD1R1060C

340

mutant were equally expressed in the transfected cells (Fig 3, E). Results showed that

341

DNA replication was rescued in WT POLD1–transduced patient cells, with a significant

342

increase in the frequency of cells in the S phase, whereas the POLD1R1060C failed to do

343

so (Fig 3, F, G). Collectively, our studies indicate that the POLD1R1060C mutant severely

344

decreased DNA replication.

345

Increased frequency of memory T cells in POLD1R1060C subjects. In view of their

346

recurrent infections and immune disorder phenotypes, we further examined the cohort

347

of POLD1R1060C subjects for cell abnormalities. Consistently, all three POLD1R1060C

348

subjects showed sharp skewing of peripheral CD4 and CD8 T cell towards a memory

349

(CD45RA−CD45RO+)

350

CD45RA+CCR7+ CD4+T cells and especially CD8+T cells were profoundly decreased in

351

POLD1R1060C subjects, whereas the frequencies of CD45RA–CCR7–CD4+ and CD8+ T

352

cells [CD45RA– effector memory (TEM)] were increased. The frequencies of

353

CD45RA+CCR7– CD8+ [CD45RA+ effector memory (TEMRA)] T cells were similar in

354

patient and control subjects, while the CD4+ TEMRA were marginally increased in patients.

355

The frequencies of CD45RA–CCR7+ CD4+ [central memory (TCM)] T cells and CD8+ T

phenotype

(Fig

4,

A-D).

The

frequencies

of

naïve

356

cells were mildly decreased compared to healthy control group without achieving

357

statistical significance. (Fig 4, E-H).

358

Flow cytometric analysis of the regulatory T cells (Treg) in the peripheral blood of

359

POLD1R1060C patients showed a mild but non-significant increase in the distribution of

360

CD4+CD25hiCD127lo T cells (Fig E3, A and B in the Online Repository). The

361

percentage of induced Treg was found similar in POLD1R1060C patients and controls (Fig

362

E3, C and D in the Online Repository). The expression of several canonical Treg cell

363

markers, including Foxp3, CTLA-4, and Helios, was normal in patients compared to

364

controls (Fig E3, E and F in the Online Repository). Overall, these results establish a

365

predominance of effector memory T cell subpopulations in POLD1R1060C subjects in the

366

context of T cell lymphopenia.

367

Restricted T cell receptor (TCR) V-J pairing in POLD1R1060C CD8+T cells. Given that

368

the POLD1R1060C mutation gave rise to severe T lymphopenia with effector T cell

369

skewing, we examined the patient CD4 and CD8 T cell populations for evidence of

370

abnormalities in their T cell receptor repertoire. Specifically, we asked if POLD1 may

371

play a role in the nonhomologous end joining as reflected in VDJ recombination.

372

Accordingly, we analyzed TRB of cell-sorted CD4+ and CD8+ T cells, and IGH of CD19+

373

B cells for evidence of VDJ recombination defects and oligoclonality. Repertoire

374

analysis showed that the V-J pairing patterns and productive entropy (a measure of

375

clonal diversity) were similar between POLD1R1060C and control CD4+ T cells and B cells

376

(Fig 5, A and B, and Fig E4, A and B in the Online Repository). In contrast, the

377

POLD1R1060C CD8+T cell subset displayed restricted V-J gene combinations and lower

378

productive entropy (Fig 5, A and B). Analysis of the productive frequencies of top 100

379

as well as the total clonotypes further revealed severe oligoclonality among patient

380

CD8+ but not CD4+ T cells (Fig 5, C and D). Altogether, these observations suggested a

381

selective impact of POLD1R1060C on CD8 T cell clonal expansion in the periphery and

382

possibly an additional effect during thymic development.

383

To determine whether POLD1 is involved in somatic hypermutation (SHM), we analyzed

384

SHM patterns in B cells. We found no difference between patients and controls in the

385

IGHV-IGHJ pairing, D gene usage, productive entropy and the proportion of unique IGH

386

rearrangements carrying at least one mutation (Fig E4, A-D in the Online Repository).

387

The rate of SHM in the V segment was similar in both of the productive IGH

388

rearrangements and non-productive IGH rearrangements of POLD1R1060C patients (Fig

389

E4, E in the Online Repository). Finally, the in silico translated amino acid sequences

390

of the IGH CDR3 region no differences in length and tyrosine contents between

391

POLD1R1060C patients and healthy controls (Fig E4, F and G in the Online Repository),

392

whereas the hydrophobicity analysis showed a borderline statistically significant

393

marginal difference between POLD1R1060C patients and healthy controls (p=0.049, Fig

394

E4, F in the Online Repository). These findings, together with the normal B cell

395

proliferation shown earlier, confirmed that the POLD1R1060C mutation spared the B cell

396

compartment.

397

Discussion

398

In this study, we report a homozygous missense mutation in POLD1, encoding the

399

catalytic subunit of the ubiquitous DNA polymerase Polδ, in three patients from an

400

extended consanguineous kindred who presented with recurrent infections with T cell

401

lymphopenia and poor antibody responses. Functional and structural analysis indicated

402

that the mutation impaired the interaction of POLD1 with other Polδ subunits, resulting

403

in poor Polδ complex formation and defective DNA replication. Our results thus

404

establish this mutation as a novel cause of primary combined immunodeficiency.

405

The mutant POLD1R1060C protein failed to associate with Polδ accessory subunit POLD2

406

or with the DNA replication complex component RFC, necessary for loading PCNA onto

407

DNA to initiate DNA replication34. Structural modeling revealed the mutation disrupted a

408

critical molecular interaction between POLD1 and POLD2, in agreement with the co-

409

precipitation studies showing failure of the mutant POLD1R1060C protein to associate with

410

POLD2. While the mutation is located at the C-terminus of POLD1 away from the

411

catalytic and DNA exonuclease domains, a secondary effect of the mutation on the

412

catalytic function of POLD1 cannot be excluded and requires further investigation.

413

Analysis of patient T cells indicated the POLD1R1060C mutation impaired DNA replication

414

and cell proliferation, an effect that was rescued by lentivirus-mediated expression of a

415

WT but not the mutant POLD1R1060C protein. In contrast, B cell proliferation was spared.

416

A key finding in our studies is that the POLD1R1060C mutation not only impaired T cell

417

proliferation but was also associated with restricted TCR V-J pairing and severe

418

oligoclonality in CD8+ T cells. Importantly, these abnormalities appeared limited to the

419

CD8+ T cell compartment, and did not involve the CD4+ T cells or the B cells. These

420

findings suggest that Polδ differentially impact CD8 T cells, severely limiting their

421

peripheral expansion and possibly affecting their thymic development. A differential

422

impact of POLD1R1060C on the CD8+ T cell compartment would be consistent with the

423

heightened susceptibility of the patients to herpetic and respiratory viral infections. In

424

contrast, the patients have so far been spared infections associated with severe CD4+ T

425

cell depletion such as Pneumocystis jiroveci. This sparing may reflect the continued

426

presence of diverse, albeit profoundly lymphopenic, CD4 T cell populations and the

427

absence of concurrent debilitating disease(s).

428

In contrast to the T cell defects noted in the patients, the B cell compartment appeared

429

to be completely spared in terms of B cell proliferation, IgH V-J pairing, D gene usage

430

and somatic hypermutation. Furthermore, indices such as the proportion of tyrosine

431

residue of the IGH-CDR3 region and the IgH hydrophobicity profile, both associated

432

with B cell self-reactivity, were either similar (the former) or marginally affected (the

433

latter). In view of these findings, the presence of mild hypogammaglobulinemia in the

434

patients with waning antibody responses to vaccines that can be rescued by booster

435

immunization is best explained by suboptimal T cell help.

436

Previous studies have identified different heterozygous mutations affecting the catalytic

437

and exonuclease domains of POLD1

438

multisystem developmental characterized by subcutaneous lipodystrophy, deafness,

439

mandibular hypoplasia, while the latter results in familial colorectal cancers. Interestingly,

440

neither of these two types of POLD1 mutations resulted in immunodeficiency. In

441

contrast, the kindred reported herein had very minimal manifestations aside from the

442

immunodeficiency, notably partial sensorineural hearing loss in P1. This sharp

17, 35, 36

. The former gives rise to a distinct

443

segregation of disease phenotypes as a function of the respective domains targeted by

444

mutations suggests the POLD1R1060C mutation, which locates away from the enzymatic

445

domains, may allow for residual Polδ activities to rescue the multisystem and mutator

446

phenotypes associated with the catalytic and proof-reading activities in different tissues.

447

However, the Polδ complex may play a non-redundant role in protein-protein

448

interactions with other components of the double stranded break repair (DSBR) relevant

449

to VDJ recombination in CD8+T cells. The nature of such interactions involving the Polδ

450

complex in DSBR and their role in T cell development requires further investigation.

451

In addition to the POLD1 mutation reported herein, a splice junction mutation in POLE2,

452

encoding the DNA polymerase epsilon subunit 2 (POLE2) has been described in a child

453

with dysmorphic features and combined immunodeficiency and autoimmunity, with

454

absence of circulating B cells, T cell lymphopenia and neutropenia

455

accessory subunit in the Polε complex38, and while the protein expression of mutant

456

POLE2 was unaffected in the child, the mutation may adversely affect the Polε complex

457

assembly and/or its interactions with other proteins relevant to DNA replication. The

458

TRB repertoire in the POLE2 patient appeared largely normal, indicating that the

459

restricted V-J pairing in the TRB of CD8+ T cells in the POLD1R1060C patients was

460

specific to this gene defect. Overall, these results point to an emerging subgroup of

461

primary immunodeficiency disorders caused by genetic lesions in different DNA

462

polymerases. The full spectrum of these disorders and their unique and shared

463

attributes require further investigation.

464

Acknowledgements

37

. POLE2 is an

465

This work was supported by National Institutes of Health grants 5R01AI085090 and

466

5R01AI128976 (to T.A.C.)

467

References

468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506

1. 2.

3.

4.

5.

6.

7. 8. 9.

10. 11. 12.

13. 14.

Coleman WB, Tsongalis GJ. Multiple mechanisms account for genomic instability and molecular mutation in neoplastic transformation. Clin Chem 1995; 41:644-57. Chilkova O, Stenlund P, Isoz I, Stith CM, Grabowski P, Lundstrom EB, et al. The eukaryotic leading and lagging strand DNA polymerases are loaded onto primerends via separate mechanisms but have comparable processivity in the presence of PCNA. Nucleic Acids Res 2007; 35:6588-97. Greenough L, Menin JF, Desai NS, Kelman Z, Gardner AF. Characterization of family D DNA polymerase from Thermococcus sp. 9 degrees N. Extremophiles 2014; 18:653-64. Henneke G, Flament D, Hubscher U, Querellou J, Raffin JP. The hyperthermophilic euryarchaeota Pyrococcus abyssi likely requires the two DNA polymerases D and B for DNA replication. J Mol Biol 2005; 350:53-64. Podust VN, Chang LS, Ott R, Dianov GL, Fanning E. Reconstitution of human DNA polymerase delta using recombinant baculoviruses: the p12 subunit potentiates DNA polymerizing activity of the four-subunit enzyme. J Biol Chem 2002; 277:3894-901. Makarova KS, Krupovic M, Koonin EV. Evolution of replicative DNA polymerases in archaea and their contributions to the eukaryotic replication machinery. Front Microbiol 2014; 5:354. Song J, Hong P, Liu C, Zhang Y, Wang J, Wang P. Human POLD1 modulates cell cycle progression and DNA damage repair. BMC Biochem 2015; 16:14. Uchimura A, Hidaka Y, Hirabayashi T, Hirabayashi M, Yagi T. DNA polymerase delta is required for early mammalian embryogenesis. PLoS One 2009; 4:e4184. Lee MY, Zhang S, Lin SH, Wang X, Darzynkiewicz Z, Zhang Z, et al. The tail that wags the dog: p12, the smallest subunit of DNA polymerase delta, is degraded by ubiquitin ligases in response to DNA damage and during cell cycle progression. Cell Cycle 2014; 13:23-31. Tumini E, Barroso S, Calero CP, Aguilera A. Roles of human POLD1 and POLD3 in genome stability. Sci Rep 2016; 6:38873. Murga M, Lecona E, Kamileri I, Diaz M, Lugli N, Sotiriou SK, et al. POLD3 Is Haploinsufficient for DNA Replication in Mice. Mol Cell 2016; 63:877-83. Zhou Y, Meng X, Zhang S, Lee EY, Lee MY. Characterization of human DNA polymerase delta and its subassemblies reconstituted by expression in the MultiBac system. PLoS One 2012; 7:e39156. Strzalka W, Ziemienowicz A. Proliferating cell nuclear antigen (PCNA): a key factor in DNA replication and cell cycle regulation. Ann Bot 2011; 107:1127-40. Daee DL, Mertz TM, Shcherbakova PV. A cancer-associated DNA polymerase delta variant modeled in yeast causes a catastrophic increase in genomic instability. Proc Natl Acad Sci U S A 2010; 107:157-62.

507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551

15.

16.

17.

18.

19.

20.

21. 22. 23.

24. 25.

26.

27.

28.

29.

30.

Church DN, Briggs SE, Palles C, Domingo E, Kearsey SJ, Grimes JM, et al. DNA polymerase epsilon and delta exonuclease domain mutations in endometrial cancer. Hum Mol Genet 2013; 22:2820-8. Liu D, Frederiksen JH, Liberti SE, Lutzen A, Keijzers G, Pena-Diaz J, et al. Human DNA polymerase delta double-mutant D316A;E318A interferes with DNA mismatch repair in vitro. Nucleic Acids Res 2017; 45:9427-40. Palles C, Cazier JB, Howarth KM, Domingo E, Jones AM, Broderick P, et al. Germline mutations affecting the proofreading domains of POLE and POLD1 predispose to colorectal adenomas and carcinomas. Nat Genet 2013; 45:136-44. Weedon MN, Ellard S, Prindle MJ, Caswell R, Lango Allen H, Oram R, et al. An in-frame deletion at the polymerase active site of POLD1 causes a multisystem disorder with lipodystrophy. Nat Genet 2013; 45:947-50. Sasaki H, Yanagi K, Ugi S, Kobayashi K, Ohkubo K, Tajiri Y, et al. Definitive diagnosis of mandibular hypoplasia, deafness, progeroid features and lipodystrophy (MDPL) syndrome caused by a recurrent de novo mutation in the POLD1 gene. Endocr J 2018; 65:227-38. Schmitz-Abe K, Li Q, Rosen SM, Nori N, Madden JA, Genetti CA, et al. Unique bioinformatic approach and comprehensive reanalysis improve diagnostic yield of clinical exomes. Eur J Hum Genet 2019. Case DAea. AMBER 14. University of California, San Francisco 2014. Berendsen HJC, Grigera JR, Straatsma TP. The Missing Term in Effective Pair Potentials. J Phys Chem 1987; 91:6269-71. Joung IS, Cheatham TE. Determination of alkali and halide monovalent ion parameters for use in explicitly solvated biomolecular simulations. J Phys Chem B 2008; 112:9020-41. Berendsen HJC, Postma JPM, Vangunsteren WF, Dinola A, Haak JR. MolecularDynamics with Coupling to an External Bath. J. Chem. Phys. 1984; 81:3684-90. Ryckaert J-P, Ciccotti G, Berendsen HJC. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of nalkanes. J. Comput. Phys. 1977; 23:327-41. Alamyar E, Duroux P, Lefranc MP, Giudicelli V. IMGT((R)) tools for the nucleotide analysis of immunoglobulin (IG) and T cell receptor (TR) V-(D)-J repertoires, polymorphisms, and IG mutations: IMGT/V-QUEST and IMGT/HighV-QUEST for NGS. Methods Mol Biol 2012; 882:569-604. Carlson CS, Emerson RO, Sherwood AM, Desmarais C, Chung MW, Parsons JM, et al. Using synthetic templates to design an unbiased multiplex PCR assay. Nat Commun 2013; 4:2680. Robins HS, Campregher PV, Srivastava SK, Wacher A, Turtle CJ, Kahsai O, et al. Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells. Blood 2009; 114:4099-107. Gupta NT, Vander Heiden JA, Uduman M, Gadala-Maria D, Yaari G, Kleinstein SH. Change-O: a toolkit for analyzing large-scale B cell immunoglobulin repertoire sequencing data. Bioinformatics 2015; 31:3356-8. Garcia-Diaz M, Bebenek K. Multiple functions of DNA polymerases. CRC Crit Rev Plant Sci 2007; 26:105-22.

552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584

31. 32. 33.

34.

35.

36.

37. 38.

39.

40.

41.

Doublie S, Zahn KE. Structural insights into eukaryotic DNA replication. Front Microbiol 2014; 5:444. Li CJ. Flow Cytometry Analysis of Cell Cycle and Specific Cell Synchronization with Butyrate. Methods Mol Biol 2017; 1524:149-59. Dolbeare F, Gratzner H, Pallavicini MG, Gray JW. Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine. Proc Natl Acad Sci U S A 1983; 80:5573-7. Podust VN, Tiwari N, Stephan S, Fanning E. Replication factor C disengages from proliferating cell nuclear antigen (PCNA) upon sliding clamp formation, and PCNA itself tethers DNA polymerase delta to DNA. J Biol Chem 1998; 273:31992-9. Bertocci B, De Smet A, Berek C, Weill JC, Reynaud CA. Immunoglobulin kappa light chain gene rearrangement is impaired in mice deficient for DNA polymerase mu. Immunity 2003; 19:203-11. Bertocci B, De Smet A, Weill JC, Reynaud CA. Nonoverlapping functions of DNA polymerases mu, lambda, and terminal deoxynucleotidyltransferase during immunoglobulin V(D)J recombination in vivo. Immunity 2006; 25:31-41. Lieber MR. The polymerases for V(D)J recombination. Immunity 2006; 25:7-9. Elsayed FA, Tops CMJ, Nielsen M, Ruano D, Vasen HFA, Morreau H, et al. Low frequency of POLD1 and POLE exonuclease domain variants in patients with multiple colorectal polyps. Mol Genet Genomic Med 2019; 7:e00603. Fiorillo C, D'Apice MR, Trucco F, Murdocca M, Spitalieri P, Assereto S, et al. Characterization of MDPL Fibroblasts Carrying the Recurrent p.Ser605del Mutation in POLD1 Gene. DNA Cell Biol 2018. Frugoni F, Dobbs K, Felgentreff K, Aldhekri H, Al Saud BK, Arnaout R, et al. A novel mutation in the POLE2 gene causing combined immunodeficiency. J Allergy Clin Immunol 2016; 137:635-8 e1. Kraszewska J, Garbacz M, Jonczyk P, Fijalkowska IJ, Jaszczur M. Defect of Dpb2p, a noncatalytic subunit of DNA polymerase varepsilon, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae. Mutat Res 2012; 737:34-42.

585

Figure 1. Characterization of a POLD1 mutation in kindred with combined

586

immunodeficiency. A. Patient Pedigree and familial segregation of the mutant POLD1

587

allele. Double lines connecting parents indicate consanguinity. Probands are indicated

588

as P1-P3. Squares, Male subjects; circles, female subjects; solid symbols, patients;

589

half-filled symbols, heterozygous. B. Sanger sequencing fluorograms of the germline c.

590

3178 C>T POLD1 mutation in patients P1-P3 and the parents of P1 and P2 compared

591

with the equivalent DNA sequences in an unaffected healthy sibling (HS). C. Schematic

592

representation of POLD1 protein. The different domains are depicted as follows: the

593

nuclear localization signal (NLS) in red, the exonuclease in light green, the polymerase

594

domain in light blue and the cysteine-rich metal-binding domains (CysA/B) in orange.

595

The identified R1060C mutation and its position within the CysB polypeptide sequence

596

is indicated by red star. D. Immunoblot analysis of POLD1/2/3 protein expression in

597

PBMCs (left) and primary fibroblasts (right) of P1, P2 and a healthy control subject (HC).

598

E. Quantitation of POLD1/2/3 protein expression in primary fibroblasts of P1 and P2,

599

normalized for β-actin expression and expressed as fold change compared to that of HC

600

fibroblasts (n=3; open circles). Results represent means ± S.E.M.**p<0.01; ***, p<0.001;

601

****, p<0.0001, by one-way ANOVA with Bonferroni post-test analysis.

602

Figure 2. POLD1R1060C mutation disrupts POLD1-intrinsic and POLD1-POLD2

603

molecular interactions. A. The protein structure model of full-length POLD1 using

604

Swiss modeling based on the crystal structure of POLA1. B. Distance between the

605

CysB motif and POLBc_delta domain in WT POLD1 and the POLD1R1060C mutant

606

protein are measured by 10-ns snapshots. C. Detailed interface analysis of POLD1

607

structure. The hydrogen bonds between E830 and R1060, R968 and R1060 backbone,

608

and R808 and Q1059 were in blue and red. D. Docking model showing the cooperation

609

between the CysB motif and the POLBc_delta domain in recruiting POLD2 and its

610

disruption by the R1060C mutation. E. HEK293T cells were transfected with the

611

indicated

612

immunoprecipitated with an anti-V5 antibody and then immunoblotted with the indicated

613

antibodies. F. Quantitation of co-immunoprecipitated POLD2 protein by Empty vector,

614

WT POLD1 and POLD1R1060C in E, expressed as fold change compared to that of WT

615

POLD1 (n=3; open circles). Results represent means ± S.E.M. ***, p<0.001; ****,

616

p<0.0001, by one-way ANOVA with Bonferroni post-test analysis.

plasmids.

Twenty-four

hours

after

transfection,

cell

lysates

were

617 618

Figure 3. Defective proliferation of POLD1R1060C T cells and its rescue by WT

619

POLD1 expression. A. Representative flow cytometric analysis of BrdU+ P1 and P2 T

620

cells versus those of a healthy control (HC) and healthy sibling (HS). B. Frequencies of

621

BrdU+ T cells in healthy control versus P1-P3 cell cultures (n=3; open circles). C.

622

Representative flow cytometric analysis of BrdU+ patient (Pt) B cells versus those of a

623

healthy control (HC). D. Frequencies of BrdU+ B cells in healthy control versus P1-P3

624

cell cultures (n=3; open circles). E. Immunoblot analysis of POLD1 expression in

625

transduced cells. F. Representative dot plot analysis of T cells transfected with either

626

Empty vector, POLD1 WT or POLD1 R1060C lentiviral plasmid. Data are representative

627

of three experiments. G. Frequencies of BrdU+ T transfected with either Empty vector,

628

POLD1 WT or POLD1 R1060C lentiviral plasmid in F. (n=3; open circles). Results are

629

means ± S.E.M. ****, p<0.0001; ns, not significant, by unpaired t test in B and D; *,

630

p<0.05; ns, not significant, by one-way ANOVA with Bonferroni post-test analysis in G.

631

Figure 4. Increased frequency of memory T cells in POLD1R1060C subjects. A. Flow

632

cytometric analysis of circulating CD45RO−CD45RA+ and CD45RO+CD45RA− CD4+ T

633

cells in a control and a POLD1R1060C subject. B. Frequency of activated

634

CD45RO+CD45RA− within the peripheral blood CD4+ T cell pool of controls and

635

POLD1R1060C subjects. C. Flow cytometric analysis of circulating CD45RO−CD45RA+

636

and CD45RO+CD45RA− CD8+ T cells in a control and a POLD1R1060C subject. D.

637

Frequency of activated CD45RO+CD45RA− within the peripheral blood CD8+ T cell pool

638

of controls and POLD1R1060C subjects. E. Flow cytometric analysis of circulating TEMRA

639

CD45RA+CCR7− and naïve CD45RA+CCR7+ CD4+ T cells in a control and a

640

POLD1R1060C subject. F. Frequency of circulating naïve CD45RA+CCR7+, central

641

memory CD45RA-CCR7+, effector memory CD45RA-CCR7- and TEMRA CD45RA+CCR7−

642

within the peripheral blood CD4+ T cell pool of controls and POLD1R1060C subjects. G.

643

Flow

644

CD45RA+CCR7+ CD8+ T cells in a control and a POLD1R1060C subject. H. Frequency of

645

circulating naïve CD45RA+CCR7+, central memory CD45RA-CCR7+, effector memory

646

CD45RA-CCR7- and TEMRA CD45RA+CCR7− within the peripheral blood CD8+ T cell pool

647

of controls and POLD1R1060C subjects. ns, not significant; *, p<0.05; ***, p<0.001 by

648

Student’s unpaired two tailed t test.

649

Figure 5. Restricted usage of Vβ β and Jβ β genes in POLD1R1060C T Cells. A.

650

Frequencies of specific Vβ and Jβ pairing in unique TRB clonotypes of CD8+T cells and

651

CD4+Tcells from healthy controls and patients. White represents the absence of a given

652

Vβ and Jβ pairing. Blue reflects a low frequency while red represents a higher

653

frequency of usage. B. Productive entropy of CD8+ and CD4+ T cells in healthy controls

cytometric

analysis

of

circulating

TEMRA

CD45RA+CCR7−

and

naïve

654

versus patients (n=3; open circles). C. Productive frequencies of top 100 most abundant

655

clonotypes of CD8+ and CD4+ T cells in healthy controls versus patients. D. Tree map of

656

clonality in CD8+ and CD4+ T cells. Each square represents a specific V-J pairing. The

657

area of each square is proportional to the total frequency of the corresponding TRB

658

clonotypes in each sample. ns, not significant; *, p<0.05; **, p<0.01 by Student’s

659

unpaired two tailed t test.

660

Table1 Hematological and immunological parameters in POLD1R1060C patients. Age CBC WBC (cells/µl) Neutrophils(cells/µl) ) Lymphocytes(cells/µl) Eosinophils cells/µl) Hemoglobin(g/dL) Platelets(cells/µl) Immunoglobulins IgA (mg/dL) IgM (mg/dL) IgG (mg/dL) IgE (IU) Specific antibody response Anti-Tetanus IgG (IU/mL) Isohemaglutinin titer (dilutions) Anti Hepatitis B antibodies (IU) Lymphocyte subset analysis + CD3 cells/µl + + CD3 CD4 cells/µl + + CD3 CD8 cells/µl + CD19 cells/µl + + CD16 CD56 cells/µl + + CD3 TCR ߙ/ߚ cells/µl + + CD3 TCR ߛ/ߜ cells/µl + + + CD4 CD45RA CD31 % + + – CD19 IgD CD27 % + – + CD19 IgD CD27 % + + + CD19 IgD CD27 % + + + CD19 CD24 CD38 %

12yrs

P1 P1 13yrs (On IVIG)

P2

P3

Normal range for age

3yrs

4yrs

5yrs

Normal range for age

29yrs

Normal range for age

1830 900 470 130 13 244000

5000-14500 1500-8000 1500-7000 200-600 11.1-17.2 150000-400000

6720 5370 780 70 11.9 152000

4000-10000 1500-7300 800-5500 200-600 12.1-17.2 150000-400000

4300 3100 600 20 12.8 344000

4860 3430 430 260 14.2 348000

4500-13000 1500-7300 1200-5200 200-600 12.1-17.2 150000-400000

3260 1970 765 2 13.0 360000

3070 1740 971 134 13.1 323000

23 49 723 17

23 55 911 19

72-177 63-164 822-1323

26 97 588 18

28 97 568 19

46-129 50-146 722-1195 68

42.1 126 902 18

65-176 86-175 944-1506

0.01 1/16 319

0.65 1/16

>0.1 >1/16 >10

0.01 1/32 38

0.2* 1/64 624

>0.1 >1/16 >10

0.1 1/16 0.06

>0.1 >1/16 >10

330 102 186 120 12 213 54 13.0

177 66 89 151 46

1000-2600 530-1500 330-1100 110-570 70-480 700-2800 39-540 32.9-61.5 51.3-82.5 8.7-25.6 4.6-18.2 5.3-18.9

337 145 176 337 61

437 184 233 456 58 466 39 48.0

1400-6200 700-2200 490-1300 390-1400 130-720 600-4300 27-960 52-67 47.3-77.0 10.9-30.4 5.2-20.4 7.4-23.7

252 101 135 85 402

1000-2600 530-1500 330-1100 110-570 70-480 600-3300 25-200 9.8-43.2 48.4-79.7 8.3-27.8 7.0-23.8 2.2-13.3

7.0

62.6 2.06 0.53 5.48

6.1

100 57 105 239 61

61.8 1.12 0.49 4.43

5 64 1.86 0.47 0.7

* post boo ster vac cina tion.

Table E1. List of candidate variants identified using WES (MAF <=0.01) and segregate within the family (variants in P1, P2 and not their unaffected sibling or parents). Genetic model

Gene name

Exonic function

Amino acid change

Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Recessive Compound-het Compound-het Compound-het Compound-het De novo

NADK FAT4 LY6G5C AGER NOTCH4 C6ORF10 C6ORF10 C6ORF10 HLA-DRB5 HLA-DRB1 HLA-DRB1 HLA-DRB1 HLA-DRB1 HLA-DRB1 HLA-DQA1 TRAF3IP2 TULP4 GLT6D1 PITRM1 PITRM1 C14ORF178 C19ORF33 GGN RASGRP4 MIA PSG6 PSG6 ZNF221 ZNF225 CEACAM16 RSPH6A POLD1 SIGLEC12 LRP1B LRP1B ZFHX3 ZFHX3 MUC12

insertion (6 bases) nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous stopgain nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous insertion (15 bases) nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous nonsynonymous

p.G414delinsRRG:p.G446delinsRRG:p.G591delinsRRG p.Q1257E p.F53L:p.F56L:p.F54L p.G82S:p.G68S:p.G113S p.G294R p.G477V:p.G479V:p.G463V:p.G478V p.L264W:p.L266W:p.L250W:p.L257W:p.L265W p.G143R:p.G122R:p.G145R:p.G129R p.Q220X p.V73M p.V73L p.D70N p.D57Y p.D57N p.M18T p.D10N:p.D19N p.S522N p.P219S p.L64F:p.L441F:p.L785F:p.L883F:p.L884F p.L113V:p.L145V p.G31D:p.G61D p.K90delinsKEGEGQ p.A517V:p.A434V p.G165R p.P16L p.S312F:p.S405F p.I122M:p.I243M p.V165M p.R352H p.S32I p.Q184H:p.Q448H p.R1060C* p.A77T p.E3955K p.V2146F p.Y865C p.K520N p.S1610I:p.S1753I

Homozygosity

14.7 MB 14.7 MB 14.7 MB 14.7 MB 14.7 MB 14.7 MB 14.7 MB 14.7 MB 14.7 MB

PolyPhen

SIFT

na B B P D D D na na D D D P B B D B D D B B na D B na B D D D B D D P B B D B B

na T T T T D D D na D T D D D T D T T D T D na T D na D D D D D D D D T D D T D

*POLD1 R1060C is in reference to POLD1 isoform 1 (NM_001256849.1). PolyPhen and SFIT scores: B: benign; T: tolerant; P: probably deleterious; D: deleterious

Table E2. Homozygous regions identified in autosomal chromosomes using WES and segregating within the family. Chromosome

Start Position

End Position

Size (MB)

SNPs (Homozygous)

3

130,098,639

132,105,588

2.007

262

6

33,032,788

38,650,628

5.618

1732

13

41,567,248

45,563,464

3.996

555

19

38,040,492

38,314,767

0.274

119

19

39,219,780

53,990,002

14.770

8617

Table E3 Summary of immune repertoire analysis. Sample

Cell Input

Total templates

Unique clonotypes

Productive templates

Unique productive clonotypes

HC1-CD8

160,000

7367

4942

5825

3964

HC2-CD8

100,000

2269

2075

1838

1687

S002a-CD8

67,000

1660

1547

1415

1318

P002-CD8

93,000

2568

243

1653

187

S002b-CD8

50,000

1401

496

910

416

P3-CD8

30,000

735

278

588

212

HC1-CD4

100,000

6550

5862

5223

4628

HC2-CD4

100,000

1341

1234

1101

1008

S002a-CD4

100,000

1475

1375

1247

1161

P002-CD4

64,000

4528

3655

3693

2936

S002b-CD4

100,000

2046

1964

1773

1703

P3-CD4

57,000

4905

3358

4085

2772

HC1-B

50,000

1550

1545

1272

719

HC2-B

100,000

2017

2012

1676

1032

S002a-B

199,000

11838

11824

9938

6557

P002-B

28,000

3148

3143

2604

1359

S002b-B

75,000

1652

1643

1294

770

P3-B

54,000

503

497

438

249

HC1, healthy control 1; HC2, healthy control 2; S002a, healthy sibling; P002, patient 1; S002b, patient 2; P3, patient 3.

1

Combined Immunodeficiency due to a loss of function mutation in DNA

2

Polymerase Delta 1

3

Ye Cui, PhD, Sevgi Keles, MD, Louis-Marie Charbonnier, PhD, Amélie M. Julé,

4

PhD,1 Lauren Henderson, MD, MSc. Seyma Celikbilek Celik, BSc, Ismail Reisli, MD,

5

Chen Shen, PhD,

6

Talal A. Chatila, MD, MSc1

7

1

8

Harvard Medical School, Boston;

9

Faculty, Division of Pediatric Allergy and Immunology, Konya; 3Program in Molecular

10

and Cellular Medicine, Department of Pediatrics, Boston Children's Hospital, Harvard

11

Medical School, Boston;

12

Pharmacology,

13

Massachusetts Institute of Technology, Cambridge; 6Division of Newborn Medicine and

14

Neonatal Genomics Program, Boston Children’s Hospital, Harvard Medical School,

15

Boston.

1

2

1

1

3,4

2

5

6

2

1,3

Wen Jun Xie, PhD, Klaus Schmitz-Abe, PhD, Hao Wu, PhD,

Division of Immunology, The Boston Children's Hospital. Department of Pediatrics,

Harvard

2

Necmettin Erbakan University, Meram Medical

4

Department of Biological Chemistry and Molecular

Medical

School,

Boston;

5

Department

of

Chemistry,

16 17

Corresponding Author: Talal A. Chatila at the Division of Immunology, Boston

18

Children’s Hospital and the Department of Pediatrics, Harvard Medical School.

19

Address: Karp Family Building, Room 10-214. 1 Blackfan Street, Boston, MA 02115

20

Email: [email protected]

21

22

23

Supplementary Methods

24

Whole exome sequencing and data analysis: WES data was processed through

25

Variant Explorer Pipeline (VExP) using BWA aligner (version 0.7.17) for mapping reads

26

to the human genome (hg19) and PICARDtools (V2.20.2) to mark/delete duplicate

27

reads. Single Nucleotide Variants (SNVs) and small insertions / deletions (indels) were

28

jointly called across all samples using both GATK (multi-sample variant calling, v4.1)

29

and SAMTools (v1.9). Further, VExP was performed to annotate 21 relevant genetic

30

databases (from Allele frequency and Gene-phenotype consortiums) and 23

31

coding/non-coding variant pathogenicity predictors into the output of the system. Variant

32

analysis was performed using different inheritance models (assuming full penetrance)

33

based on three filtering criteria: first, include variants predicted to have a potential

34

functional coding consequence, including stop gain or loss, splice site disruption, indel,

35

and nonsynonymous. Second, variants were filtered based on allele frequency in control

36

populations (gnomAD, ExAC, EVS, 1000GP and internal data from 2,114 unaffected

37

individuals from BCH). The variants were further prioritized to include those with read

38

depth ≥10X and deleterious prediction (2 or more of 23 softwares, including PolyPhen,

39

SIFT, FATHMM, CADD, etc). For pedigree-consistency analysis, VExP had verified

40

consistency within all family members.

41

Homozygosity mapping: We use Variant Explorer Pipeline “VExP” to determine

42

homozygous regions using whole genome sequencing data (WES). In summary, the

43

method uses a sliding window approach, 100 SNPs, and retained segments with a

44

minimum of 98% homozygosity. Homozygous SNPs cannot be more than 100 Kb away

45

from each other. Next, VExP joins all the homozygous regions using several

46

considerations including regions with no genes or noncoding genes. It retains only

47

segments where observed homozygosity exceeds 3 cM and avoid the effect of residual

48

population homozygosity that is likely innocuous and tolerated by natural selection. We

49

use genetic, as opposed to physical distance, for all calculations. To calculate overall

50

homozygosity for every sample, we sum all segments exceeding 3 cM. Homozygosity

51

Mapping is applied to the results from the whole family, obtaining overlapping

52

homozygous regions between affected individuals with no overlapping with unaffected

53

samples (Fig E1, Table E2).

54

Reference:

55 56 57

1.

Kyte J, Doolittle RF. A simple method for displaying the hydropathic character of a protein. J Mol Biol 1982; 157:105-32.

58

Supplementary Figure Legends

59

Figure E1. Homozygous regions identified in chromosome 19 using WES and

60

segregating within the family (Homozygosity mapping).

61

Figure E2. A. Distance measurement between residue 1060 and 830 in wild-type and

62

mutant POLD1. Longer distance in the mutant POLD1 reflects the interaction lost

63

between CysB and POLBc_delta domains. B. Hydrogen bond (HB) tracing of the

64

interaction between R1060 and E830 in POLD1. 1 means HB exists, 0 means HB

65

disappears.

66

Figure E3. POLD1R1060C patients are normal in regulatory T cell frequency and

67

phenotype. A. Representative dot plot analysis of CD25hi CD127lo CD4+ T cells in

68

patient vs. a control subject. B. Percentages of CD25hi CD127lo CD4+ T cells in the

69

peripheral blood of healthy controls (n=3; open circles) and POLD1R1060C patients (n=3;

70

closed circles). C. Representative dot plot analysis of Helios– CD25hi CD127lo CD4+ T

71

cells in patient vs. a control subject. D. Mean fluorescence intensity of the respective

72

regulatory T cell marker in patient and control subjects. ns, not significant, by unpaired

73

two-tailed Student's t-test.

74

Figure E4. Profiles of IGHV-IGHJ pairing and somatic hypermutation (SHM) analysis in

75

POLD1R1060C B Cells. A. Frequencies of specific IGHV and IGHJ pairing in unique IGH

76

clonotypes of B cells from healthy controls and patients. White represents the absence

77

of a given IGHV and IGHJ pairing. Blue reflects a low frequency while red represents a

78

higher frequency of usage. B. Frequencies of specific IGD gene usage in unique IGH

79

clonotypes of B cells from healthy controls and patients. White represents the lowest

80

gene usage. Blue reflects a higher frequency of usage. C. Productive entropy of B cells

81

in healthy controls versus patients (n=3; open circles). D. Percentage of sequences

82

carrying at least one somatic hypermutation (SHM) among all unique rearrangements

83

(productive and non-productive) with resolved V family, gene or allele. E. Number of

84

mutations per 100 bp within the IGH V segment of unique rearrangements with at least

85

one mutation in non-productive rearrangements and productive rearrangements.

86

Amino acid properties of in silico translated IGH-CDR3 region, for unique productive

87

rearrangements encoding a complete CDR3 region (starting and ending with consensus

88

codons). CDR3 length: control vs. patient group, p>0.1; subject effect, p<10e-6.

89

GRAVY, Grand Average of Hydrophobicity1: control vs. patient group, p= 0.0490;

90

subject effect, p>0.4.). G. Proportion of tyrosine residue in the IGH-CDR3 of unique

91

sequences: control vs. patient group (n=3; open circles). C, D, E and G were analyzed

92

with Student’s unpaired two tailed t test. Results represent means ± S.E.M. ns, not

93

significant. F was analyzed with 2-way ANOVA, to contrast sequences patterns

94

between patients and control while accounting for per-subject variations. ns, not

95

significant; *, p<0.5.

96

F.