PreprintPDF Available

Preexisting memory CD4 T cells in naïve individuals confer robust immunity upon vaccination

August 2020

August 2020

DOI:10.1101/2020.08.22.262568

License
CC BY-NC-ND 4.0

Authors:

Pieter Meysman

University of Antwerp

Esther Bartholomeus

University of Antwerp

Show all 18 authorsHide

Antigen recognition through the T cell receptor (TCR) αβ heterodimer is one of the primary determinants of the adaptive immune response. Vaccines activate naïve T cells with high specificity to expand and differentiate into memory T cells that allow for a quick and robust T cell response upon exposure to the pathogen or re-exposure to the vaccine antigen. This is why the induction of memory T cells is a key feature in vaccine development. However, it has become increasingly evident that antigen-specific memory CD4 and CD8 T cells exist in unexposed antigen-naïve hosts and it is likely that exposure to one antigen might alter the TCR repertoire of memory T cells to a different unrelated antigen. In this study, we utilize high-throughput sequencing to profile the memory CD4 TCRβ repertoire and track vaccine-specific TCRβ clonotypes following the de novo administration of hepatitis B (HepB) vaccine in healthy HepB-naïve individuals and show that vaccinees with preexisting vaccine-specific memory CD4 T cell clonotypes elicited earlier and higher antibody concentrations and mounted a more robust CD4 T cell response to the vaccine. We further identify vaccine-specific TCRβ sequence patterns that can be used to predict which individuals will have an early and more vigorous vaccine-elicited immunity to HepB vaccine. Moreover, we find that an expansion of 4-1BB+ memory TREG is a prominent feature in individuals with delayed and modest vaccine-induced immunity. Our approach shows that modeling pre-vaccination TCRβ repertoire enables prediction of both antibody and CD4 responses to vaccines.

Hepatitis B vaccination (Engerix-B®) study design. 340 a Hepatitis B (Engerix-B®) vaccination and experimental design. (Top) Timeline of vaccination 341 and blood collection. (Bottom) Memory CD4 T cells were magnetically enriched and FACS-342 sorted from two time points (day 0 and day 60) for TCRβ repertoire sequencing. Matrix peptide 343 pools were used to map CD4 T cell epitopes of the vaccine from PBMCs collected at day 60 and 344 to select single peptides. After 7 days of in vitro expansion, single peptide-specific and master 345 peptide pool-specific CFSElow CD4 T cells from PBMCs collected at day 60 were FACS-sorted 346 in two technical replicates for TCRβ repertoire sequencing. PBMCs collected at days 0, 60, 180, 347 and 365 were stimulated with the master peptide pool (HBsAg) and assessed for converse 348 expression of 4-1BB and CD40L by flow cytometry. 349 b Vaccinee cohort can be classified into three groups as determined by anti-Hepatitis B surface 350 (anti-HBs) titer over four times points. 351 Early-converters seroconverted at day 60, late-converters seroconverted at day 180 or day 365 352 and non-converters did not have an anti-HBs titer higher than 10 IU/ml at any of the time points. 353 354 355

…

Overview of the single peptides tested for each vaccinee in the 870 CFSE assay 871

…

Figures - available via license: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Content may be subject to copyright.

Available via license: CC BY-NC-ND 4.0

Content may be subject to copyright.

Preexisting memory CD4 T cells in naïve individuals confer robust

immunity upon vaccination

Running title: Preexisting memory T cells confer immunity to vaccination

George Elias 1,2%, Pieter Meysman 2,3,4%, Esther Bartholomeus 2,5%, Nicolas De Neuter

2,3,4, Nina Keersmaekers 2,6, Arvid Suls 2,5, Hilde Jansens 7, Aisha Souquette 8, Hans De Reu

1,9, Evelien Smits 1,9, Eva Lion1,9, Paul G. Thomas 8, Geert Mortier 2,5, Pierre Van Damme

2,10, Philippe Beutels 2,6, Kris Laukens 2,3,4%, Viggo Van Tendeloo 1,11%, Benson Ogunjimi

2,6,12,13%

% = equal contribution

1Laboratory of Experimental Hematology (LEH), Vaccine and Infectious Disease Institute, University of Antwerp,

Antwerp, Belgium.

2Antwerp Unit for Data Analysis and Computation in Immunology and Sequencing (AUDACIS), Antwerp, Belgium.

3Adrem Data Lab, Department of Mathematics and Computer Science, University of Antwerp, Antwerp, Belgium.

4Biomedical Informatics Research Network Antwerp (biomina), University of Antwerp, Antwerp, Belgium.

5Department of Medical Genetics, University of Antwerp/Antwerp University Hospital, Edegem, Belgium

6Centre for Health Economics Research & Modeling Infectious Diseases (CHERMID), Vaccine & Infectious Disease

Institute (VAXINFECTIO), University of Antwerp, Antwerp, Belgium

7Department of Clinical Microbiology, Antwerp University Hospital

8Department of Immunology, St. Jude Children's Research Hospital, Memphis, Tennessee, USA

9Center for Cell Therapy and Regenerative Medicine, Antwerp University Hospital, Edegem, Belgium

10Centre for the Evaluation of Vaccination (CEV), Vaccine and Infectious Disease Institute, University of Antwerp,

Antwerp, Belgium.

11Janssen Research & Development, Immunosciences WWDA, Johnson & Johnson, Beerse, Belgium

12Antwerp Center for Translational Immunology and Virology (ACTIV), Vaccine and Infectious Disease Institute,

University of Antwerp, Antwerp, Belgium

13Department of Paediatrics, Antwerp University Hospital, Antwerp, Belgium.

.CC-BY-NC-ND 4.0 International license(which was not certified by peer review) is the author/funder. It is made available under a The copyright holder for this preprintthis version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.262568doi: bioRxiv preprint

Corresponding authors: George Elias, Benson Ogunjimi

E-mail address:

george.elias@student.uantwerpen.be, benson.ogunjimi@uantwerp.be

Laboratory of Experimental Hematology (LEH), Vaccine and Infectious Disease Institute,

University of Antwerp, Universiteitsplein 1, 2610 Antwerp, Belgium

Footnotes

This work was funded by

1. University of Antwerp [BOF Concerted Research Action (PS ID 30730) Methusalem

funding; Industrial Research Fund SBO].

2. Research Foundation Flanders (1861219N grant to B. Ogunjimi and FWO SB grant to N.

De Neuter.

3. ALSAC at St. Jude Children’s Research Hospital and by HSN272201400006C and

R01AI107625 from the National Institute of Allergy and Infectious Disease.

Abstract

Antigen recognition through the T cell receptor (TCR) αβ heterodimer is one of the primary

determinants of the adaptive immune response. Vaccines activate naïve T cells with high

specificity to expand and differentiate into memory T cells that allow for a quick and robust T

cell response upon exposure to the pathogen or re-exposure to the vaccine antigen. This is why

the induction of memory T cells is a key feature in vaccine development. However, it has

become increasingly evident that antigen-specific memory CD4 and CD8 T cells exist in

unexposed antigen-naïve hosts and it is likely that exposure to one antigen might alter the TCR

repertoire of memory T cells to a different unrelated antigen. In this study, we utilize high-

throughput sequencing to profile the memory CD4 TCRβ repertoire and track vaccine-specific

TCRβ clonotypes following the de novo administration of hepatitis B (HepB) vaccine in healthy

HepB-naïve individuals and show that vaccinees with preexisting vaccine-specific memory CD4

T cell clonotypes elicited earlier and higher antibody concentrations and mounted a more robust

CD4 T cell response to the vaccine. We further identify vaccine-specific TCRβ sequence patterns

that can be used to predict which individuals will have an early and more vigorous vaccine-

elicited immunity to HepB vaccine. Moreover, we find that an expansion of 4-1BB+ memory

TREG is a prominent feature in individuals with delayed and modest vaccine-induced immunity.

Our approach shows that modeling pre-vaccination TCRβ repertoire enables prediction of both

antibody and CD4 responses to vaccines.

Keywords

CD4 T cell, T cell receptor, immune repertoire sequencing, vaccination

Introduction

Antigen recognition through the T cell receptor (TCR) is one of the key determinants of the

adaptive immune response 1. Antigen presentation via major histocompatibility complex (MHC)

(encoded by HLA genes), together with the right costimulatory and cytokine signals, are

responsible for T cell activation 2,3. In this system, every T cell receptor (TCR) αβ heterodimer

imparts specificity for a peptide-MHC (pMHC) complex. A highly diverse TCR repertoire

ensures that an effective T cell response can be mounted against pathogen-derived peptides 4.

High TCRαβ diversity is generated through V(D)J recombination at the complementary-

determining region 3 (CDR3) of TCRα and TCRβ chains, accompanied with junctional deletions

and insertions of nucleotides, further adding to the diversity 5.

Vaccines activate naïve T cells with high specificity to vaccine-derived peptides and induce their

expansion and differentiation into effective and multifunctional T cells. This is followed by a

contraction phase from which surviving cells constitute a long-lived memory T cell pool that

allows for a quick and robust T cell response upon a second exposure to the pathogen 6.

However, recent work has shown that a prior pathogen encounter is not a prerequisite for the

formation of memory T cells and that CD4 T cells with a memory phenotype can be found in

antigen-naïve individuals 7. The existence of memory-like CD4 T cells in naïve individuals 8 can

be explained by molecular mimicry, as the encounter with environmentally-derived peptides

activates cross-reactive T cells due to the highly degenerate nature of the CD4 T cell recognition

of peptide-MHC complex 9. Indeed, work that attempted to replicate the history of human

pathogen exposure in mice has shown that sequential infections altered the immunological

profile and remodeled the immune response to vaccination 10. The existence of memory CD4 T

cells specific to vaccine-derived peptides in unexposed individuals might confer an advantage in

vaccine-induced immunity. In the present study we used high-throughput sequencing to profile

the memory CD4 TCRβ repertoire of healthy adults before and after administration of a hepatitis

B vaccine to investigate the impact of preexisting memory CD4 T cells on the immune response

to the vaccine. Based on anti-hepatitis B surface (anti-HBs) antibody titers over 365 days,

vaccinees were grouped into early, late and non-converters. Our data reveals that individuals

with preexisting vaccine-specific CD4 T cell clonotypes in the memory CD4 compartment had

earlier emergence of antibodies and mounted a more vigorous CD4 T cell response to the

100

vaccine. Moreover, we identify a set of vaccine-specific TCRβ sequence patterns which can be

101

used to predict which individuals will have an early and more vigorous response to hepatitis B

102

vaccine.

103

104

105

106

107

108

109

110

111

112

113

114

Methods

115

Human study design and clinical samples. A total of 34 healthy individuals (20-29y: 10, 30-

116

39y: 7, 40-49y: 16, 50+y: 1) without a history of HBV infection or previous hepatitis B

117

vaccination were recruited in this study after obtaining written informed consent. Individuals

118

were vaccinated with a hepatitis B vaccine by intramuscular (m. deltoideus) injection (Engerix-

119

B® containing 20 μg dose of alum-adjuvanted hepatitis B surface antigen, GlaxoSmithKline) on

120

days 0 and 30 (and on day 365). At days 0 (pre-vaccination), 60, 180 and 365, peripheral blood

121

samples were collected on spray-coated lithium heparin tubes, spray-coated K2EDTA

122

(dipotassium ethylenediamine tetra-acetic acid) tubes and serum tubes (Becton Dickinson, NJ,

123

USA). The study was approved by the Ethics Committee of Antwerp University Hospital and

124

University of Antwerp (Antwerp, Belgium).

125

126

Peripheral blood mononuclear cells. Peripheral blood mononuclear cells (PBMC) were

127

isolated by Ficoll-Paque Plus gradient separation (GE Healthcare, Chicago, IL, USA). Cells were

128

stored in 10% dimethyl sulfoxide in fetal bovine serum (Thermo Fisher Scientific, Waltham,

129

MA, USA). After thawing and washing cryopreserved PBMC, cells were cultured in AIM-V

130

medium that contained L-glutamine, streptomycin sulfate at 50 µg/ml, and gentamicin sulfate at

131

10 µg/ml (Thermo Fisher Scientific, Waltham, MA, USA) and supplemented with 5% human

132

serum (One Lambda, Canoga Park, CA, USA).

133

134

Serology and complete blood count. Serum was separated and stored immediately at – 80°C

135

until time of analysis. Anti-HBs antibody was titrated in serum from day 0, 60, 180 and 365

136

using Roche Elecsys® Anti-HBs antibody assay on an Elecsys® 2010 analyzer (Roche, Basel,

137

Switzerland). An anti-HBs titer above 10 IU/ml was considered protective 11.

138

Serum IgG antibodies to Cytomegalovirus (CMV), Epstein–Barr virus viral-capsid antigen

139

(EBV-VCA), and Herpes Simplex virus (HSV)-1 and 2 were determined using commercially

140

available sandwich ELISA kits in accordance with the manufacturer’s instructions.

141

A complete blood count including leukocyte differential was run on a hematology analyzer

142

(ABX MICROS 60, Horiba, Kyoto, Japan).

143

144

Sorting of memory CD4 T cells. Total CD4 T cells were isolated by positive selection using

145

CD4 magnetic microbeads (Miltenyi Biotech, Bergisch Gladbach, Germany). Memory CD4 T

146

cells were sorted after gating on single viable CD3+CD4+CD8−CD45RO+ cells. The following

147

fluorochrome-labeled monoclonal antibodies were used for staining: CD3-PerCP (BW264/56)

148

(Miltenyi Biotech), CD4-APC (RPA-T4) and CD45RO-PE (UCHT1) (both from Becton

149

Dickinson, Franklin Lakes, NJ, USA) and CD8-Pacific Orange (3B5) (from Thermo Fisher

150

Scientific, Waltham, MA, USA). Cells were stained at room temperature for 20 min and sorted

151

with FACSAria II (Becton Dickinson). Sytox blue (Thermo Fisher Scientific) was used to

152

exclude non-viable cells.

153

154

Single peptides, matrix peptide pools and epitope mapping. A set of 15-mers peptides with an

155

11-amino acid overlap spanning the 226 amino acids along the small S protein of hepatitis B

156

(HB) surface antigen (HBsAg), also designated as small HBs (SHBs) 12, were synthesized by

157

JPT Peptide Technologies (Berlin, Germany). The set, composed of 54 single peptides (See

158

supplementary table 1), was used in a matrix-based strategy to map epitopes against which the

159

immune response is directed 13. The matrix layout enables efficient identification of epitopes

160

within the antigen using a minimal number of cells. For this purpose, a matrix of 15 pools, 7

161

rows and 8 columns, referred to as matrix peptide pool, was designed so that each peptide is in

162

exactly one row-pool and one column-pool, thereby allowing for the identification of positive

163

peptides at the intersection of positive pools. Matrix peptide pools that induced a CD4 T cell

164

response (as determined by CD40L/CD154 assay described below) which meets the threshold

165

criteria for a positive response were considered in the deconvolution process. Top six single

166

peptides were considered for peptide-specific T cell expansion and sorting. A master peptide

167

pool is composed of all of the 54 single peptides and was used to identify and sort total vaccine-

168

specific CD4 T cells. Each peptide was used at a final concentration of 2 µg/ml.

169

170

171

Ex vivo T cell stimulation (CD40L/CD154 assay). Thawed PBMC from each vaccinee were

172

cultured in AIM-V medium that contained L-glutamine, streptomycin sulphate at 50 µg/ml, and

173

gentamicin sulphate at 10 µg/ml. (GIBCO, Grand Island, NY) and supplemented with 5% human

174

serum (One Lambda, Canoga Park, CA, USA). Cells were stimulated for 6 hours with 2 μg/ml of

175

each of the 15 matrix peptide pools in the presence of 1 µg/ml anti-CD40 antibody (HB14)

176

(purchased from Miltenyi Biotec, Bergisch Gladbach, Germany) and 1 μg/ml anti-CD28

177

antibody (CD28.2) (purchased from BD Biosciences, Franklin Lakes, NJ, USA).

178

Cells were stained using the following fluorochrome-labelled monoclonal antibodies: CD3-

179

PerCP (BW264/56), CD4-APC (REA623), CD8-VioGreen (REA734) and CD40L-PE (5C8)

180

(purchased from Miltenyi Biotec, Bergisch Gladbach, Germany). Viability dye Sytox blue from

181

Invitrogen (Thermo Fisher Scientific, Waltham, MA, USA) was used to exclude non-viable cells.

182

Data was acquired on FACSAria II using Diva Software, both from BD Biosciences (Franklin

183

Lakes, NJ, USA), and analyzed on FlowJo software version 10.5.3 (Tree Star, Inc., Ashland, OR,

184

USA). Fluorescence-minus-one controls were performed in pilot studies. Gates for CD40L+CD4

185

T cells were set using cells left unstimulated.

186

187

In vitro T cell expansion and cell sorting. Thawed PBMC were labelled with

188

carboxyfluorescein succinimidyl ester (CFSE) (Invitrogen, Carlsbad, CA, USA) and cultured in

189

AIM-V medium that contained L-glutamine, streptomycin sulphate at 50 µg/ml, and gentamicin

190

sulphate at 10 µg/ml. (GIBCO, Grand Island, NY) and supplemented with 5% human serum

191

(One Lambda, Canoga Park, CA, USA). Cells were stimulated for 7 days with 2 µg/ml of

192

selected single peptides in addition to the master peptides pool. Cells were stained using the

193

following fluorochrome-labelled monoclonal antibodies: CD3-PerCP (BW264/56), CD4-APC

194

(REA623) and CD8-VioGreen (REA734) (purchased from Miltenyi Biotec, Bergisch Gladbach,

195

Germany). Viability dye Sytox blue from Invitrogen (Thermo Fisher Scientific, Waltham, MA,

196

USA) was used to exclude non-viable cells. Single viable CFSElow CD3+ CD8− CD4+ T cells

197

were sorted into 96-well PCR plates containing DNA/RNA Shield (Zymo Research, Irvine, CA,

198

USA) using FACSAria II and Diva Software (BD Biosciences, Franklin Lakes, NJ, USA). For

199

each of the selected single peptides, 500 cells were sorted in two technical replicates. For the

200

master peptide pool, 1000 cells were sorted in two technical replicates. Plates were immediately

201

centrifuged and kept at – 20°C before TCR cDNA library preparation and sequencing.

202

203

TCRβ cDNA Library Preparation and Sequencing of memory CD4 T cells. DNA was

204

extracted from sorted memory CD4 T cells using Quick-DNA Microprep kit (Zymo Research,

205

Irvine, CA, USA). ImmunoSEQ hsTCRB sequencing kit (Adaptive Biotechnologies, Seattle,

206

WA, USA) was used to profile TCRβrepertoire following the manufacturer’s protocol.

207

After quality control using Fragment Analyzer (Agilent, Santa Clara, CA, USA), libraries were

208

pooled with equal volumes. The concentration of the final pool was measured with the Qubit™

209

dsDNA HS Assay kit (Thermo Fisher Scientific, Waltham, MA, USA). The final pool was

210

processed to be sequenced on the Miseq and NextSeq platforms (Illumina, San Diego, CA,

211

USA).

212

213

TCR cDNA Library Preparation and Sequencing of CFSElow CD4 T cells. RNA was

214

extracted from each of the two technical replicates of sorted CFSElow CD4 T cells using Quick-

215

RNA Microprep kit (Zymo Research, Irvine, CA, USA). Without measuring the resulted RNA

216

concentration, an RNA-based library preparation was used. The QIAseq Immune Repertoire

217

RNA Library kit (Qiagen, Venlo, Netherlands) amplifies TCR alpha, beta, gamma and delta

218

chains. After quality control using Fragment Analyzer (Agilent, Santa Clara, CA, USA),

219

concentration was measured with the Qubit™ dsDNA HS Assay kit (Thermo Fisher Scientific,

220

Waltham, MA, USA) and pools were equimolarly pooled and prepared for sequencing on the

221

Nextseq platform (Illumina, San Diego, CA, USA).

222

223

TCRβ Sequence Analysis. TCRβ clonotypes were identified as previously described 14 where a

224

unique TCRβ clonotype is defined as a unique combination of a V gene, CDR3 amino acid

225

sequence, and J gene. All memory CD4 T cell DNA-based TCRβ sequencing reads were

226

annotated using the immunoSEQ analyzer (v2) from Adaptive Biotechnologies. All small bulk

227

RNA-based TCR sequencing reads were annotated using the MiXCR tool (v3.0.7) from the

228

FASTQ files. As all RNA-based TCR sequencing experiments featured two technical replicates,

229

only those TCR sequences that occurred in both replicates were retained and their counts were

230

summed. Tracking of vaccine-specific TCRβ clonotypes is based on exact TCRβ CDR3 amino

231

acid matches to remove any bias introduced by the different VDJ annotation pipelines. Non-

232

HBsAg TCR annotations were done with the TCRex web tool 15 on the 24th of July, 2019 using

233

version 0.3.0. Inference of similar epitope binding between two TCR sequences is defined

234

according to the Hamming distance (d) calculated on the CDR3 amino acid sequence with a

235

cutoff c, as supported by Meysman et al. 16. All scripts used in this analysis are available via

236

github (https://github.com/pmeysman/HepBTCR).

237

238

Predictive HBs-response model. From the single peptide data generated in the matrix peptide

239

pool experiments, we aimed to create a predictive model to enumerate the HBs response from

240

full TCRβ repertoire data. This approach allows for predictions that are epitope-specific rather

241

than simply vaccine-specific. This model was applied in a leave-one-out cross validation so that

242

vaccine-specific TCRβ sequences from a vaccinee are not used to make predictions for the same

243

vaccinee. While the predictive model is derived from epitope-specific data, it cannot be

244

guaranteed that some of the expanded CD4 T cells detected in the in vitro assay are not due to

245

bystander activation. Vaccine-specific TCRβ sequences of vaccinees who did not respond to the

246

vaccine at day 60 (late-converters and non-converters) are expected to be more enriched in cells

247

triggered to expand due to bystander activation. Indeed, running the set of vaccine-specific

248

TCRβ sequences through the TCRex webtool 15 reveals that there is strong enrichment for the

249

CMV NLVPMVATV epitope (P value = 1.16e-71) and the Mart-1 variant ELAGIGILTV

250

epitope (P value = 2.16e-60), which supports the notion that some of these TCRβ sequences

251

might not be specific to HBsAg. This set of vaccine-specific TCRβ sequences can thus be used

252

to make predictions about possible TCRβ sequences due to bystander activation of CD4 T cells,

253

i.e. common TCRβ sequences that might be present as false positives. The final output of the

254

model is thus a ratio Rhbs for any repertoire repi describing a set of TCRβ sequences trepi:

255

󰇛󰇜     





   

256

with tpep as the set of TCRβ sequences occurring in both biological replicates for a single sample

257

and a single peptide (pep) from the HBsAg matrix peptide pool experiment, and tbystander as the

258

set of TCRβ sequences occurring in both biological replicates of the master peptide pool in any

259

of the non-responding samples. Thus the ratio signifies the number of TCR clonotypes predicted

260

to be reactive against one of the HBsAg peptides, normalized by a count of putative false

261

positive predictions from bystander T-cells.

262

263

Ex vivo T cell phenotyping of vaccine-specific T cells. Thawed PBMC from each vaccinee

264

were cultured in AIM-V medium that contained L-glutamine, streptomycin sulphate at 50 µg/ml,

265

and gentamicin sulphate at 10 µg/ml. (GIBCO, Grand Island, NY) and supplemented with 5%

266

human serum (One Lambda, Canoga Park, CA, USA). Cells were stimulated for 6 hours with 2

267

μg/ml of a master peptide pool representing the full length of the small surface envelope protein

268

of hepatitis B, in the presence of 1 µg/ml anti-CD40 antibody (HB14) (purchased from Miltenyi

269

Biotec, Bergisch Gladbach, Germany) and 1 μg/ml anti-CD28 antibody (CD28.2) (purchased

270

from BD Biosciences, Franklin Lakes, NJ, USA). Cells were stained using the following

271

fluorochrome-labelled monoclonal antibodies:

272

CD3-BV510 (SK7), CD4-PerCP/Cy5.5 (RPA-T4), CD8-APC/Cy7 (SK1), CD45RA-AF488

273

(HI100), CD25-BV421 (M-A251), CD127-BV785 (A019D5) and CD137-PE (4-1BB)

274

(purchased from BioLegend, San Diego, CA, USA), CXCR5 (CD185)-PE-eFluor 610

275

(MU5UBEE) (from eBioscience, Thermo Fisher Scientific, Waltham, MA, USA) and CD40L-

276

APC (5C8) (purchased from Miltenyi Biotec, Bergisch Gladbach, Germany). Fixable viability

277

dye Zombie NIR™ from BioLegend (San Diego, CA, USA) was used to exclude non-viable

278

cells. Data was acquired on FACSAria II using Diva Software, both from BD Biosciences

279

(Franklin Lakes, NJ, USA), and analyzed on FlowJo software version 10.5.3 (Tree Star, Inc.,

280

Ashland, OR, USA) using gating strategy shown in Fig. S1a. Fluorescence-minus-one controls

281

were performed in pilot studies. Gates for CD40L+ and 4-1BB+ CD4 T cells (Fig. S1b) were set

282

using cells left unstimulated (negative control contained DMSO at the same concentration used

283

to solve peptide pools). In order to account for background expression of CD40L and 4-1BB on

284

CD4 T cells, responses in cells left unstimulated were subtracted from the responses to peptides,

285

and when peptides-specific CD40L+ or 4-1BB+ CD4 T cells were not significantly higher than

286

those detected for cells left unstimulated (using one-sided Fisher's exact test), values were

287

mutated to zero.

288

289

Statistical analysis and data visualization. The two-sided Fisher’s exact test was used to

290

evaluate the significance of relationship between early/late-converters and CMV, EBV or HSV

291

seropositivity. For the visualization of marker expression, TCRβ counts and cell frequencies

292

between time points or groups of vaccinees, ggplot2 (V3.3.2) and ggpubr (V0.2.5) packages in R

293

were used. The Wilcoxon signed-rank test was used to compare two or more groups, with

294

unpaired and paired analysis as necessary. The nonparametric Spearman's rank-order correlation

295

was used to test for correlation. We used the following convention for symbols indicating

296

statistical significance; ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001.

297

Statistical analysis. The two-sided Fisher’s exact test was used to evaluate the significance of

298

relationship between early/late-converters and CMV, EBV or HSV seropositivity. For the

299

visualization of marker expression, TCRβ counts and cell frequencies between time points or

300

groups of vaccinees, ggpubr (V0.2.5) package in R was used. The Wilcoxon signed-rank test was

301

used to compare two or more groups, with unpaired and paired analysis as necessary. The

302

nonparametric Spearman's rank-order correlation was used to test for correlation. We used the

303

following convention for symbols indicating statistical significance; ns P > 0.05, * P ≤ 0.05, ** P

304

≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001.

305

306

Data availability

307

The sequencing data that support the findings of this study have been deposited on Zenodo

308

(https://doi.org/10.5281/zenodo.3989144). Flow Cytometry Standard (FCS) data files with

309

associated FlowJo workspaces are deposited at flowrepository.org 17 under the following

310

experiment names: epitope mapping: https://flowrepository.org/id/FR-FCM-Z2TN; in vitro T

311

cell expansion: https://flowrepository.org/id/FR-FCM-Z2TM; ex vivo CD4 T cell assay:

312

https://flowrepository.org/id/FR-FCM-Z2TL.

313

314

315

316

317

318

319

320

Results:

321

Vaccinee cohort can be classified into three groups

322

Out of 34 vaccinees, 21 vaccinees seroconverted (an anti-HBs titer above 10 IU/ml was

323

considered protective 11) at day 60 and were classified as early-converters; 9 vaccinees

324

seroconverted at day 180 or day 365 and were classified as late-converters; remaining 4

325

vaccinees had an anti-HBs antibody titer lower than 10 IU/ml at all time points following

326

vaccination and were classified as non-converters (Fig. 1 and Fig. S2a).

327

Members of Herpesviridae family might alter immune responses to vaccines 18. We found no

328

significant differences in CMV, EBV or HSV seropositivity between the three groups in our

329

cohort (Fig. S2b). Early-converters were slightly younger than late-converters and non-

330

converters were notably younger than both early and late converters (Fig. S2c).

331

332

333

334

335

336

337

338

339

Figure 1. Hepatitis B vaccination (Engerix-B®) study design.

340

a Hepatitis B (Engerix-B®) vaccination and experimental design. (Top) Timeline of vaccination

341

and blood collection. (Bottom) Memory CD4 T cells were magnetically enriched and FACS-

342

sorted from two time points (day 0 and day 60) for TCRβ repertoire sequencing. Matrix peptide

343

pools were used to map CD4 T cell epitopes of the vaccine from PBMCs collected at day 60 and

344

to select single peptides. After 7 days of in vitro expansion, single peptide-specific and master

345

peptide pool-specific CFSElow CD4 T cells from PBMCs collected at day 60 were FACS-sorted

346

in two technical replicates for TCRβ repertoire sequencing. PBMCs collected at days 0, 60, 180,

347

and 365 were stimulated with the master peptide pool (HBsAg) and assessed for converse

348

expression of 4-1BB and CD40L by flow cytometry.

349

b Vaccinee cohort can be classified into three groups as determined by anti-Hepatitis B surface

350

(anti-HBs) titer over four times points.

351

Early-converters seroconverted at day 60, late-converters seroconverted at day 180 or day 365

352

and non–converters did not have an anti-HBs titer higher than 10 IU/ml at any of the time points.

353

354

355

356

a b

Memory CD4 T cell repertoire decreases in clonality following vaccination in early-

357

converters

358

A genomic DNA-based TCRβ sequence dataset of memory CD4 T cells isolated from peripheral

359

blood was generated from a cohort of 34 healthy vaccinees right before vaccination (day 0) and

360

60 days after administration of the first dose of hepatitis B vaccine (30 days after administration

361

of the second vaccine dose).

362

Between 7.32 × 104 and 4.41 × 105 productive TCRβ sequence reads were obtained for each

363

vaccinee at each time point (Fig. S3a). Between 30,000 and 90,000 unique TCRβ sequences

364

were sequenced for each vaccinee at each time point (Fig. S3b). As expected, considering the

365

extremely diverse memory CD4 T cell repertoire 19, only an average of 25% of the TCRβ

366

sequences can be detected at both time points for each vaccinee (Fig. S3c).

367

368

The diversity of the memory CD4 T cell repertoire of each vaccinee at the two time points was

369

explored. Even though the number of unique clones in the memory CD4 TCRβ repertoire

370

remained stable in between the two time points, we detected a significant increase in the TCRβ

371

repertoire Shannon’s entropy for early-converters (Fig. 2a) (P value = 0.042), but not late-

372

converters, suggesting that the memory CD4 T cell repertoires of early-converters have become

373

less clonal, despite the number of distinct TCRβ sequences not changing significantly.

374

375

Unique vaccine-specific TCRβ sequences are trackable within memory CD4 T cell

376

repertoire and increase following vaccination

377

Peripheral blood mononuclear cells from day 60 were labeled with carboxyfluorescein

378

succinimidyl ester (CFSE) and stimulated with a pool of peptides spanning hepatitis B (HB)

379

surface antigen (HBsAg). At day 7 of in vitro expansion, we sorted CFSElow CD4 T cells 20 and

380

extracted mRNA for quantitative assessment of vaccine-specific TCRβ clonotypes by sequencing

381

(see Methods for details), allowing for the tracking of vaccine-specific TCRβ within memory

382

CD4 T cell repertoire over the two times points, based on CDR3β amino acid sequence mapping.

383

We detected a significant increase in the frequency of unique vaccine-specific TCRβ sequences

384

at day 60 post-vaccination compared to pre-vaccination (mean increase = 96.5%, 95% CI = 56.7

385

- 170%) (Fig. 2b). Moreover, this increase was larger for early-converters (mean = 132.1%, 95%

386

CI = 76.4 - 238.2%) than late-converters (mean = 22.1%, 95% CI = 5.9 – 50.1%). For non-

387

converters the mean was 81.6% (95% CI: [42.7% - 110.6%]). A Wilcoxon test shows that the

388

difference between the increase for the early converters and late converters had a P value of

389

0.04909.

390

As vaccine-specific TCRβ sequences were already detected in the memory CD4 T cell repertoire

391

prior to vaccination, we sought to determine whether the vaccination results in an expansion of

392

those sequences. However, using the abundance of vaccine-specific TCRβ sequences within the

393

memory CD4 T cell repertoire, the data does not support a vaccine-induced expansion of

394

preexisting vaccine-specific TCRβ sequences (Fig. S3d). Thus, although we see a rise in the

395

number of vaccine-specific TCRβ clonotypes from day 0 to day 60, this cannot be attributed to

396

an expansion of preexisting TCRβ clonotypes but rather the recruitment of new TCRβ

397

clonotypes (presumably from the naïve T cell compartment), as visualized for one vaccinee in

398

Fig. 2c.

399

It makes sense to not only look at the difference in vaccine-specific TCRβ sequences between

400

time points, but also explore whether there are differences in the proportion of HBsAg-specific

401

clones in the memory repertoire between early-converters, late-converters and non-converters

402

after vaccination. In this case, as we aim for a between-vaccinees comparison (in contrast to the

403

within-vaccinees timepoint comparison), we normalize by the number of HBsAg-specific TCRβ

404

found for each vaccinee. Thus, the values are different from those reported before. From this

405

analysis, it can be concluded that there is a difference in HBsAg-specific TCRβ at day 60

406

between the three groups (Fig. 2d) (ANOVA P value = 0.00238). A Wilcoxon test between

407

early-converters and other vaccinees shows a significant P value of 0.000473, thereby implying

408

that early-converters have a higher relative frequency of vaccine-specific TCRβ sequences

409

present in their memory repertoire at day 60 compared to the rest.

410

It is notable that some of the vaccine-specific T cell clonotypes can be detected in the memory

411

repertoire of vaccinees prior to vaccination (Fig. 2e). However, the relative frequency was not

412

statistically significant different between the categories (ANOVA P value = 0.238).

413

414

415

Figure 2. CD4 T cell memory TCRβ repertoire and vaccine-specific TCRβ clonotypes.

416

a Repertoire diversity and entropy between the two time points

417

b Frequency of unique vaccine-specific TCRβ sequences out of total sequenced TCRβ sequences

418

between two time points for all vaccinees colored by group

419

c Sequenced CD4+ TCR memory repertoire of vaccinee H35 at day 60. Each TCR clonotype is

420

represented by a node. TCRs are connected by an edge if their Hamming distance is one. Only

421

clusters with at least three TCRs are shown. TCR clonotypes in red are the vaccine-specific

422

TCRβ sequences that were not present prior to vaccination.

423

d Frequency of vaccine-specific TCRβ sequences within memory CD4 T cell repertoire

424

normalized by amount of HBsAg-specific TCRβ found for each vaccinee at time point 60 and (e)

425

at time point 0.

426

427

HBsAg single peptide-specific TCRβ identification allows predictive modelling of early

428

converters prior to vaccination

429

To quantify the T-cell response at the level of individual peptides that make up the HBsAg, a

430

matrix peptide pool covering 54 overlapping peptides of the HBsAg was used to extract peptide-

431

specific T-cells using a CD40L/CD154 activation-induced marker (AIM) assay (see Methods for

432

details, Fig. S4). The top 6 peptides for each individual were selected for TCR sequencing after a

433

CFSE assay (Supplementary Table 2). In this manner, TCRβ sequences were identified for T-

434

cells reactive against 44 individual HBsAg peptides. These were not uniformly distributed across

435

the HBsAg amino acid sequence, with the most prominent epitopes covering the regions 1-15,

436

129-144, 149-164, 161-176, 181-200, 213-228. For each of those regions, more than 10

437

individuals had a strong T-cell response and more than 150 unique TCRβ sequences could be

438

identified (Fig. 3a).

439

These peptide-specific TCRβ sequences can be utilized in a peptide-TCR interaction classifier to

440

identify other TCRβ that are likely to react against the same HBsAg epitopes, as it has been

441

shown that similar TCRβ sequences tend to target the same epitopes 16,21. These classifications

442

were integrated into a model which outputs a ratio Rhbs for any TCRβ repertoire representing the

443

amount of HBsAg peptide-specific clonotypes. The ratio Rhbs is based on the frequency of

444

putative peptide-specific TCRβ divided by a normalization term for putative false positive

445

predictions due to bystander activations in the training data set. This model applied to the

446

memory repertoire at day 60 shows that early-converters tend to have a higher frequency of

447

putative HBsAg peptide-specific TCRβ, while late-converters tend to have relatively more false

448

positive hits (Fig. 3b). Thus, the defined ratio Rhbs shows significant difference between the

449

early-converters and the late-converters at day 60 (one-sided Wilcoxon-test P value= 0.0313,

450

Fig. 3c). Furthermore, calculating the Rhbs on the memory repertoires prior to vaccination (day 0)

451

shows a similar difference (one-sided Wilcoxon-test P value= 0.0010, Fig. 3d). In this manner,

452

Rhbs has predictive potential and can be used as a classifier to distinguish early from late-

453

converters prior to vaccination (Fig. 3e), with an AUC of 0.825 (95% CI: 0.657 – 0.994) in a

454

leave-one-out cross validation setting.

455

While Rhbs is able to differentiate between early-and late-converters, it seems to be worse at

456

distinguishing non-converters. This is mainly due to a single non-converter vaccinee (H21) with

457

a high Rhbs, signifying a high number of putative HBsAg peptide-specific TCRβ in their memory

458

repertoire.

459

460

Predictive capacity of TCRβ repertoire holds true for humoral and cellular immune

461

response

462

Response groups used thus far were established based on anti-HBs titers at different time points.

463

However, response groups can be defined differently if one considers the results from the

464

CD40L assay. For the most part, only those with a high Ab titer at day 60, had a positive value

465

for the CD40L assay (Fig S3e). The only exception was H21, who had a low measurable Ab titer

466

but had a measurable CD40L value. Thus, this vaccinee seems to have had a T-cell response

467

against the vaccine, without developing an antibody response. It is interesting to note that this

468

vaccinee is classified as a non-responder in the data set but it is predicted as a possible strong

469

early-converter by the Rhbs predictive model in the leave-one-out cross validation. Redoing the

470

analysis with Rhbs ratio to predict the CD40L assay at day 60, shows that the Rhbs ratio is a good

471

classifier in a leave-one-out cross validation (Fig. 3f). We tested the predictive capacity of Rhbs

472

ratio at different time points and cell types, and found good AUC values for both 4-

473

1BB−CD40L+ (TCON) and 4-1BB+CD40L− (TREG, see further) at day 60, but to a lesser extent at

474

days 180 and 365, thereby indicating that Rhbs predictive model is currently limited to predicting

475

mid-term responses following vaccination (Fig S3f and g).

476

477

478

Figure 3

479

100

200

300

050 100 150 200

HBsAg amino acid position

Unique ep itope−specific TCR .. sequences

MENITSGFLGPLLVLQAGFFLLTRILTIPQSLDSWWTSLNFLGGSPVCLGQNSQSPTSNHSPTSCPPICPGYRWMCLRRFIIFLFILLLCLIFLLVLLDYQGMLPVCPLIPGSTTTNTGPCKTCT TPAQGNSMFPSCCCTKPTDGNCTCIPIPSSWAFAKYLWEWASVRFSWLSLLVPFVQWFVGLSPTVWLSAIWMMWYWGPSLYSIVSPFIPLL

a. Overview of found HBsAg epitope-specific TCRβ sequences. Each bar corresponds to unique

480

TCRβ sequences found against a single 15mer HBsAg peptide, with 11 amino acid overlap to

481

each subsequent peptide. Bars in blue denote those epitopes for which 10 or more volunteers had

482

a strong T-cell reaction. Motif logos on top of bars denote a sampling of the most common

483

TCRβ amino acid sequence motifs for those epitopes.

484

b. Scatter plot with the percentage predicted epitope-specific and bystander TCRβ sequences.

485

Predictions done as a leave-one-out cross-validation. Each dot represents a vaccinee with the

486

color denoting the responder status (blue: early-converter, yellow: late-converter, red: non-

487

converter)

488

c. HBsAg-predictive ratio R(hbs) when calculated on the memory repertoires at day 60

489

d. HBsAg-predictive ratio R(hbs) when calculated on the memory repertoires at day 0

490

e. ROC curve of using epitope-specific / bystander ratio to differentiate between early-converters

491

and late-converters in a leave-one-out cross validation at day 0. Diagonal line denotes a random

492

classifier. Reported is the area under the curve (AUC) and its 95% confidence interval

493

f. ROC curve of using epitope-specific / bystander ratio to differentiate between samples with a

494

positive CD4L value (> 0) in a leave-one-out cross validation. Diagonal line denotes a random

495

classifier. Reported is the area under the curve (AUC) and its 95% confidence interval. Note that

496

the classifier is worse as several samples (EC, LC and NC) had NA values for the CD154 assay,

497

which were all set to zero.

498

499

500

Vaccine-specific conventional and regulatory memory CD4 T cells induced in early-

501

converters

502

After showing evidence for the existence of vaccine-specific TCRβ sequences pre-vaccination

503

and that individuals with a higher number of HBsAg peptide-specific clonotypes had earlier

504

seroconversion, we attempted to link this observation to differences in vaccine-specific CD4 T

505

cells responses using CD4 T cell assays. As TREG cells might suppress vaccine-induced immune

506

responses 22, we used activation markers CD40L (CD154) and 4-1BB (CD137) to help delineate

507

the conventional (TCON) and regulatory (TREG) phenotypes of activated CD4 T cells 23,24. In this

508

scheme, after 6 hours of antigen stimulation, CD40L+4-1BB− can be used as a signature for

509

antigen-specific CD4 TCON cells, as opposed to CD40L−4-1BB+ signature for antigen-specific

510

CD4 TREG cells.

511

Additionally, we added CD25 and CD127 to better identify TREG cells 25,26 and CXCR5 to further

512

distinguish circulatory T follicular helper cells (cTFH) and circulatory T follicular regulatory cells

513

(cTFR) 27,28.

514

Using the converse expression of CD40L and 4-1BB, CD40L+4-1BB− and CD40L−4-1BB+ CD4

515

T cells had a TCON and TREG phenotype, respectively, as shown by the expression of CD25 and

516

CD127 (Fig. S5a and b), and validate their use for the distinction of activated TCON and TREG

517

cells.

518

We detected a significant increase in the frequency of CD40L+4-1BB− and CD40L−4-1BB+

519

memory CD4 T cells at day 60 in our cohort (Fig. 4a) that correlated positively with the increase

520

in antibody titer between day 0 and day 365 (Fig. 4b and Fig. S6). Upon a closer look, the

521

induction of both signatures of vaccine-specific memory CD4 T cells was only true for early-

522

converters (Fig. 4c, see Fig. S7a for non-converters and Fig. S7b for vaccine-specific CD4 T

523

cells) while late-converters did not show a detectable memory CD4 T cell response. Although a

524

subset of both early and late-converters had detectable memory CD4 T cell responses prior to

525

vaccination, we detected no significant difference in the frequency of CD40L+4-1BB− and

526

CD40L−4-1BB+ memory CD4 T cells between the two groups at day 0 (Fig. 4d).

527

Collectively, flow cytometry data reveal that the expression of CD40L and 4-1BB in our ex vivo

528

assay is consistent with our serological data and reflects the lack of seroconversion at day 60 in

529

late-converters. However, it does not support the existence of more vaccine-specific memory

530

CD4 T cells in early-converters prior to vaccination.

531

532

533

Figure 4 Hepatitis B vaccine induces a vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+

534

memory CD4 T cell response in early-converter vaccinees.

535

PBMCs from vaccinees were stimulated with 2 μg/ml of the master peptide pool (HBsAg) and

536

assessed for converse expression of 4-1BB and CD40L by flow cytometry on days 0, 60, 180,

537

and 365. Shown is number of of vaccine-specific memory CD4 T cells out of 1x106 memory

538

CD4 T cells after subtraction of responses in negative control.

539

a Aggregate analysis from vaccinees (including early, late and non-converters) showing a peak

540

of vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell at day 60 (day 60

541

after 1st dose of the vaccine and day 30 after 2nd dose), declining thereafter. Shown are

542

proportions of vaccine-specific memory CD4 T cells out of total memory CD4 t cells.

543

b Correlation between the difference in antibody titer between day 365 and day 0 and vaccine-

544

specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell at day 60

545

c Aggregate analysis from early and late-converter vaccinees showing a significant induction of

546

vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell in early-converters

547

and lack thereof in late-converters.

548

d Aggregate analysis from early and late-converter vaccinees showing no significant differences

549

in vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell at day 0.

550

Wilcoxon signed-rank with unpaired and paired analysis as necessary; statistical significance

551

was indicated with ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001

552

rs, Spearman correlation coefficient, −1 ≤ rs ≤ 1; rs and p value by Spearman’s correlation test

553

a b

An expanded subset of 4-1BB+CD45RA− TREG cells is a prominent feature of late-

554

converters

555

In order to detect any distinct signatures of early and late-converters, we analyzed pre-

556

vaccination flow cytometry data to examine major CD4 T cell subsets; TH, TREG, cTFH and cTFR

557

cells. Using manual gating in which regulatory T cells (TREG) were defined as viable

558

CD3+CD4+CD8−CD25+CD127−CXCR5− and were further divided into CD45RA+ and CD45RA−

559

TREG cells, we identified a significantly higher frequency of 4-1BB+ CD45RA− TREG cells in late-

560

converters compared to early-converters (Fig. 5a and Fig. S8).

561

TREG cells showed higher 4-1BB expression compared to TH, cTFH and cTFR cells (Fig. 5b) and

562

within TREG subset, CD45RA− TREG cells showed significantly higher expression of 4-1BB,

563

accompanied with a higher expression of CD25, compared to CD45RA+ TREG cells (Fig. 5c). In

564

this scheme, CD45RA− TREG can be divided into 4-1BB+CD25high and 4-1BB-CD25int subsets. It

565

is worth noting here that no differences were detected in the frequency of CD45RA− or

566

CD45RA+ TREG cells within CD4 T cell compartment between the two groups (Fig. 5d), and that

567

the composition of TREG compartment that is distinct between the two groups (Fig. 5e).

568

In summary, an expanded subset of 4-1BB+CD45RA− TREG cells pre-vaccination is a prominent

569

feature of a delayed and modest immune response to hepatitis B vaccine in our cohort.

570

571

572

Figure 5 An expanded 4-1BB+CD45RA− TREG cells within TREG compartment is is a

573

prominent feature in late-converters prior to vaccination.

574

PBMCs from vaccinees at day 0 (prior to vaccination) were phenotyped for expression of

575

markers of TREG.

576

a Aggregate analysis of 4-1BB+CD45RA− TREG within CD45RA− TREG CD4 T cells in early and

577

late and non-converter vaccinees before vaccination

578

b Aggregate analysis of the median fluorescence intensity of 4-1BB in TH, cTFH, TREG and cTFR

579

cells before vaccination.

580

c Aggregate analysis of the median fluorescence intensity of 4-1BB (left panel) and CD25 (right

581

panel) in CD45RA− TREG and CD45RA+ TREG cells before vaccination.

582

d Frequency of TREG, CD45RA− TREG and CD45RA+ TREG cells within total CD4 T cells in early,

583

late and non-converter vaccinees before vaccination.

584

e Composition of TREG compartment as determined by expression of 4-1BB and CD45RA in

585

early, late and non-converter vaccinees before vaccination.

586

Wilcoxon signed-rank with unpaired and paired analysis as necessary; statistical significance

587

was indicated with ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001

588

a b c

589

Discussion

590

In this study, we used high-throughput TCRβ repertoire profiling and ex vivo T cell assays to

591

characterize memory CD4 T cell repertoires before and after immunization with hepatitis B

592

vaccine, an adjuvanted subunit vaccine, and tracked vaccine-specific TCRβ clonotypes over two

593

time points. As antigen-naïve adults were found to have an unexpected abundance of memory-

594

phenotype CD4 T cells specific to viral antigens 7,29, we sought to investigate the influence that

595

preexisting memory CD4 T cells can have on vaccine-induced immunity.

596

Commercially available HBV vaccines produces a robust and long-lasting anti-HBs response,

597

and protection is provided by induction of an anti-HBs (antibody against HBV surface antigen)

598

titer higher than 10 mIU/mL after a complete immunization schedule of 3 doses 30. However, 5-

599

10% of healthy adult vaccinees fail to produce protective titers of anti-HBs and can be classified

600

as non-responders 30. In our cohort, 13 vaccinees did not seroconvert by day 60 (30 days

601

following administration of the second vaccine dose), as determined by antibody titer. Out of this

602

group, 9 vaccinees seroconverted by day 180 or day 365, referred to here as late-converters, and

603

4 vaccinees did not seroconvert, referred to here as non-converters.

604

A hallmark of adaptive immunity is a potential for memory immune responses to increase in

605

both magnitude and quality upon repeated exposure to the antigen 31. Our systems immunology

606

data supports the theory that preexisting memory CD4 T cell TCRβ sequences specific to

607

HBsAg, the antigenic component of the current hepatitis B vaccine, predict which individuals

608

will mount an early and more vigorous immune response to the vaccine as evidenced by a higher

609

fold change in anti-HBs antibody titer and a more significant induction of antigen-specific CD4

610

T cells. It is postulated that preexisting memory CD4 T cell clonotypes are generated due to the

611

highly degenerate nature of T cell recognition of antigen/MHC and are cross-reactive to

612

environmental antigens 8. For example, preexisting memory CD4 T cells are well-established in

613

unexposed HIV-seronegative individuals, although at a significantly lower magnitude than HIV-

614

exposed seronegative individuals 32,33, and were likely primed by exposure to environmental

615

triggers or the human microbiome.

616

617

We and others have shown before that the TCRβ repertoire of CD4 T cells encodes the antigen

618

exposure history of each individual and that antigen-specific TCRβ sequences could serve to

619

automatically annotate the infection or exposure history 14,34,35. In this study, we show that

620

similar principles can be used to study vaccine responsiveness. Specifically, the recruitment of

621

novel vaccine-specific T-cell clonotypes into memory compartment following vaccination can be

622

tracked by examining the CD4 memory TCRβ repertoire over time. While we observed no

623

increase in the frequency of the vaccine-specific memory T-cells, as the time point may have

624

missed the peak of the clonal expansion of effector CD4 T cells as was reported before 36–38, a

625

significant rise in the number of unique vaccine-specific T-cell clonotypes was detected.

626

Nevertheless, this observation is consistent with earlier T cell immune repertoire sequencing

627

studies that showed that antigen-specific TCRβ sequences do not always overlap with those

628

TCRβ sequences that increase in frequency after infection or vaccination 39. More interestingly,

629

individuals with the earlier and more robust response against the vaccine, had a telltale antigen-

630

specific signature in their memory TCRβ repertoire prior to vaccination, despite the lack of

631

HBsAg antibodies or prior vaccination history.

632

633

Detection of this vaccine-specific signature was possible due to the development of a novel

634

predictive model that used epitope-specific TCRβ sequences from one set of individuals to make

635

predictions about another. A correction factor was needed to account for the occurrence of

636

bystander activated T-cells within the original epitope-specific TCRβ sequences. Indeed, in those

637

vaccinees without a positive antibody titer at day 60, any putative vaccine-specific T-cells may

638

be considered as bystander vaccination. This was supported by predictions using the TCRex tool

639

15, which matched these TCR sequences to common viral or other epitopes. It is of note that

640

these TCRβ sequences are matched with CD8 T cell epitopes, while they originate from isolated

641

CD4 T cells. This is likely due to the great similarity between the TCRβ sequences of CD4 and

642

CD8 T cells as noted in prior research 16.

643

644

However, our in vitro antigen-specificity data, using an assay that enables discrimination of

645

TCON and TREG cells using the converse expression of the activation markers CD40L and 4-1BB

646

23,40, failed to show a significant difference in preexisting antigen-specific CD4 T cells between

647

early and late-converters prior to vaccine administration. It is plausible that the signal is below

648

the detection limit of the assay and that more sensitive assays that require pre-enrichment of

649

CD40L+ and 4-1BB+ T cells (using magnetic beads) 41 or cultured ELISpot assay 42 are needed to

650

capture preexisting vaccine-specific memory CD4 T cells directly from human peripheral blood.

651

Another plausible explanation is that our activation proteins, CD40L and 4-1BB, might be

652

unsuitable to detect preexisting memory CD4 T cells but this is unlikely as both proteins have

653

been used successfully in similar studies 43,44. A different explanation may be that the diversity of

654

preexisting antigen-specific CD4 T clonotypes as determined by TCRβ sequencing is not

655

reflected in the quantitative measurements of the fractional cell counts.

656

657

TREG cells represent about 5 – 10% of human CD4 T cell compartment and are identified by the

658

constitutive surface expression of CD25, also known as IL-2 receptor α subunit (IL-2Ra), and the

659

nuclear expression of forkhead family transcription factor 3 (Foxp3), a lineage specification

660

factor of TREG cells 45. Regulatory memory T cells play a role in the mitigation of tissue damage

661

induced by effector memory T cells during protective immune responses, resulting in a selective

662

advantage against pathogen-induced immunopathology 46–49. Several studies have identified CD4

663

TREG cells with specificity to pathogen-derived peptides in murine models and showed evidence

664

for an induced expansion of TREG cells followed by an emergence and a long-term persistence of

665

TREG cells with a memory phenotype and potent immunosuppressive properties 47,50. Blom et al.

666

reported a significant and transient activation of TREG cells (identified by upregulation of CD38

667

and Ki67) in humans 10 days after administration of live attenuated yellow fever virus 17D

668

vaccine 37. The induction of vaccine-specific TREG cells in our cohort is unexpected and the role

669

it might play in vaccine-induced immunity warrants further investigation.

670

671

The association of an expanded 4-1BB+ CD45RA− TREG subset with a delayed immune response

672

to hepatitis B vaccine was not described before. Miyara et al. showed that blood contains two

673

distinct subsets of stable and suppressive TREG cells: resting TREG, identified as

674

FOXP3lowCD45RA+ CD4 T cells, and activated TREG, identified as FOXP3highCD45RA− CD4 T

675

cells. They further noted that activated TREG cells constitute a minority subset within cord blood

676

TREG cells and increase gradually with age 51. As activated TREG cells were shown to have an

677

increased expression of proteins indicative of activation, including ICOS and HLA-DR 51–55, it

678

might be the case that an upregulation of 4-1BB is one more feature of this population or a subset

679

thereof. Moreover, TREG cells in mice were shown to modulate TFH formation and GC B cell

680

responses and to diminish antibody production in a CTLA-4 mediated suppression 56.

681

Interestingly, CD45RA− TREG cells were shown to be more rich in preformed CTLA-4 stored in

682

intracellular vesicles compared to CD45RA+ TREG cells 51.

683

4-1BB was shown to be constitutively expressed by TREG cells 57 and that 4-1BB+ TREG cells are

684

functionally superior to 4-1BB− TREG cells in both contact-dependent and contact-independent

685

immunosuppression 58. 4-1BB+ TREG cells are the major producers of the alternatively-spliced

686

and soluble isoform of 4-1BB among T cells 58. 4-1BB was shown before to be preferentially

687

expressed on TREG cells compared with other non-regulatory CD4 T cell subsets 57 and that 4-

688

1BB-costimualtion induces the expansion of TREG cells both in vitro and in vivo 59. Moreover,

689

agonistic anti-4-1BB mAbs have been shown to abrogate T cell-dependent antibody responses in

690

vivo 60 and to ameliorate experimental autoimmune encephalomyelitis by skewing the balance

691

against TH17 differentiation in favor of TREG differentiation 61. It is plausible that the expansion

692

4-1BB+CD45RA− TREG cells in late-converters is involved in the suppression of GC vaccine-

693

specific TFH cells and the ensuing antibody response in our cohort, but this remains speculative

694

and further research is warranted.

695

696

It is enticing to speculate that the preexisting memory CD4 T cells result from the complex

697

interplay between cellular immunity and the human microbiome. A role for the microbiota in

698

modulating immunity to viral infection was suggested in 1960’s 62, and since then we gained

699

better understanding of the impact of the various components of the microbiota including

700

bacteria, fungi, protozoa, archaea and viruses on the murine and human immune systems 63.

701

Viral clearance of hepatitis B virus infection depends on the age of exposure and neonates and

702

young children are less likely to spontaneously clear the virus 64. Han-Hsuan et al. have shown

703

evidence in mice that this age-dependency is mediated by gut microbiota that prepare the liver

704

immunity system to clear HBV, possibly via a TLR4 signaling pathway 65. In this study, young

705

mice that have not reached an equilibrium in the gut microbiota, exhibited prolonged HBsAg

706

persistence, impaired anti-HBs antibody production, and limited Hepatitis B core antigen

707

(HBcAg)-specific IFNγ+ splenocytes. More recently, Tingxin et al. provided evidence for a

708

critical role of the commensal microbiota in supporting the differentiation of GC B cells, through

709

follicular T helper (TFH) cells, to promote the anti-HBV humoral immunity 66.

710

711

Our study bears some intrinsic limitations. A major drawback is the restricted number of days at

712

which TCRβ repertoire was profiled, as vaccine-specific perturbations within the repertoire may

713

occur at different time points for early, late and non-converters. Additionally, more in-depth

714

characterization and functional studies on 4-1BB+CD45RA− TREG cells could have helped shed

715

more light on the role they play in vaccine-induced immunity. Future studies in larger cohorts

716

and with a more comprehensive TCRβ repertoire profiling and CD4 T cells immunophenotyping

717

are required to validate our findings.

718

719

In conclusion, our analysis of the memory CD4 T cell repertoire has uncovered a role for

720

preexisting memory CD4 T cells in naïve individuals in mounting an earlier and more vigorous

721

immune response to hepatitis B vaccine and argue for the utility of pre-vaccination TCRβ

722

repertoire in the prediction of vaccine-induced immunity. Moreover, we identify a subset of 4-

723

1BB+ memory TREG cells that is expanded in individuals with delayed immune response to the

724

vaccine, which might further explain the heterogeneity of response to hepatitis B vaccine.

725

726

727

AUTHOR CONTRIBUTIONS

728

Conception: PM, BO

729

Design: GE, PM, EB, AS, GM, PVD, PB, KL, VVT, BO

730

Experiments: GE, PM, EB, NDN, HJ, AS

731

Data-analysis: GE, PM, NDN, BO

732

Supervision: HDR, EL, PT, GM, PVD, PB, KL, VVT, BO

733

First draft: GE, PM, BO

734

Contributed to the paper: all authors

735

736

COMPETING FINANCIAL INTERESTS

737

Parts of the contents of this manuscript form the topic of patent EPO 19159931.5.

738

VVT is an employee of Johnson & Johnson since 1/11/2019 and remains currently employed at

739

the University of Antwerp.

740

741

References

742

1. Rudolph, M. G., Stanfield, R. L. & Wilson, I. A. HOW TCRS BIND MHCS, PEPTIDES, AND CORECEPTORS. Annu. Rev.

743

Immunol. 24, 419–466 (2006).

744

2. Esensten, J. H., Helou, Y. A., Chopra, G., Weiss, A. & Bluestone, J. A. CD28 Costimulation: From Mechanism to Therapy. Immunity

745

44, 973–988 (2016).

746

3. Curtsinger, J. M. & Mescher, M. F. Inflammatory cytokines as a third signal for T cell activation. Curr. Opin. Immunol. 22, 333–40

747

(2010).

748

4. Turner, S. J., La Gruta, N. L., Kedzierska, K., Thomas, P. G. & Doherty, P. C. Functional implications of T cell receptor diversity.

749

Current Opinion in Immunology 21, 286–290 (2009).

750

5. Krangel, M. S. Mechanics of T cell receptor gene rearrangement. Current Opinion in Immunology 21, 133–139 (2009).

751

6. Farber, D. L., Yudanin, N. A. & Restifo, N. P. Human memory T cells: Generation, compartmentalization and homeostasis. Nature

752

Reviews Immunology 14, 24–35 (2014).

753

7. Su, L. F., Kidd, B. A., Han, A., Kotzin, J. J. & Davis, M. M. Virus-Specific CD4+ Memory-Phenotype T Cells Are Abundant in

754

Unexposed Adults. Immunity 38, 373–383 (2013).

755

8. Sewell, A. K. Why must T cells be cross-reactive? Nature Reviews Immunology 12, 669–677 (2012).

756

9. Wilson, D. B. et al. Specificity and degeneracy of T cells. Mol. Immunol. 40, 1047–1055 (2004).

757

10. Reese, T. A. et al. Sequential Infection with Common Pathogens Promotes Human-like Immune Gene Expression and Altered Vaccine

758

Response. Cell Host Microbe 19, 713–719 (2016).

759

11. Keating, G. M. & Noble, S. Recombinant Hepatitis B Vaccine (Engerix-B). Drugs 63, 1021–1051 (2003).

760

12. Shouval, D. Hepatitis B vaccines. J. Hepatol. 39, 70–76 (2003).

761

13. Precopio, M. L. et al. Optimizing peptide matrices for identifying T-cell antigens. Cytom. Part A 73, 1071–1078 (2008).

762

14. de Neuter, N. et al. Memory CD4+T cell receptor repertoire data mining as a tool for identifying cytomegalovirus serostatus. Genes

763

Immun. (2018). doi:10.1038/s41435-018-0035-y

764

15. Gielis, S. et al. Detection of Enriched T Cell Epitope Specificity in Full T Cell Receptor Sequence Repertoires. Front. Immunol. 10,

765

(2019).

766

16. Meysman, P. et al. On the viability of unsupervised T-cell receptor sequence clustering for epitope preference. Bioinformatics 35,

767

1461–1468 (2018).

768

17. Spidlen, J., Breuer, K., Rosenberg, C., Kotecha, N. & Brinkman, R. R. FlowRepository: A resource of annotated flow cytometry

769

datasets associated with peer-reviewed publications. Cytom. Part A 81 A, 727–731 (2012).

770

18. Furman, D. et al. Cytomegalovirus infection enhances the immune response to influenza. Sci. Transl. Med. 7, 281ra43 (2015).

771

19. Klarenbeek, P. L. et al. Human T-cell memory consists mainly of unexpanded clones. Immunol. Lett. 133, 42–48 (2010).

772

20. Becattini, S. et al. Functional heterogeneity of human memory CD4+ T cell clones primed by pathogens or vaccines. Science (80-. ).

773

347, 400–406 (2015).

774

21. De Neuter, N. et al. On the feasibility of mining CD8+ T cell receptor patterns underlying immunogenic peptide recognition.

775

Immunogenetics 70, 159–168 (2018).

776

22. Brezar, V. et al. T-regulatory cells and vaccination “pay attention and do not neglect them”: Lessons from HIV and cancer vaccine

777

trials. Vaccines 4, (2016).

778

23. Schoenbrunn, A. et al. A converse 4-1BB and CD40 ligand expression pattern delineates activated regulatory T cells (Treg) and

779

conventional T cells enabling direct isolation of alloantigen-reactive natural Foxp3+ Treg. J. Immunol. 189, 5985–94 (2012).

780

24. Elias, G., Ogunjimi, B. & Van Tendeloo, V. Activation-induced surface proteins in the identification of antigen-responsive CD4 T

781

cells. Immunology Letters 219, 1–7 (2020).

782

25. Liu, W. et al. CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ T reg cells. J. Exp. Med.

783

203, 1701–1711 (2006).

784

26. Seddiki, N. et al. Expression of interleukin (IL)-2 and IL-7 receptors discriminates between human regulatory and activated T cells. J.

785

Exp. Med. 203, 1693–1700 (2006).

786

27. Bentebibel, S.-E., Schmitt, N., Banchereau, J. & Ueno, H. Human tonsil B-cell lymphoma 6 (BCL6)-expressing CD4+ T-cell subset

787

specialized for B-cell help outside germinal centers. Proc. Natl. Acad. Sci. 108, E488–E497 (2011).

788

28. Fonseca, V. R. et al. Human blood Tfr cells are indicators of ongoing humoral activity not fully licensed with suppressive function. Sci.

789

Immunol. 2, (2017).

790

29. Su, L. F. & Davis, M. M. Antiviral memory phenotype T cells in unexposed adults. Immunol. Rev. 255, 95–109 (2013).

791

30. Meireles, L. C., Marinho, R. T. & Van Damme, P. Three decades of hepatitis B control with vaccination. World J. Hepatol. 7, 2127–32

792

(2015).

793

31. Sallusto, F., Lanzavecchia, A., Araki, K. & Ahmed, R. From vaccines to memory and back. Immunity 33, 451–463 (2010).

794

32. Ritchie, A. J. et al. Differences in HIV-Specific T Cell Responses between HIV-Exposed and -Unexposed HIV-Seronegative

795

Individuals. J. Virol. 85, 3507–3516 (2011).

796

33. Campion, S. L. et al. Proteome-wide analysis of HIV-specific naive and memory CD4+ T cells in unexposed blood donors. J. Exp.

797

Med. 211, 1273–1280 (2014).

798

34. Emerson, R. O. et al. Immunosequencing identifies signatures of cytomegalovirus exposure history and HLA-mediated effects on the T

799

cell repertoire. Nat. Genet. 49, 659–665 (2017).

800

35. DeWitt, W. S. et al. Human T cell receptor occurrence patterns encode immune history, genetic background, and receptor specificity.

801

Elife 7, (2018).

802

36. Kohler, S. et al. The early cellular signatures of protective immunity induced by live viral vaccination. Eur. J. Immunol. 42, 2363–2373

803

(2012).

804

37. Blom, K. et al. Temporal Dynamics of the Primary Human T Cell Response to Yellow Fever Virus 17D As It Matures from an

805

Effector- to a Memory-Type Response. J. Immunol. 190, 2150–2158 (2013).

806

38. Pogorelyy, M. V. et al. Precise tracking of vaccine-responding T cell clones reveals convergent and personalized response in identical

807

twins. Proc. Natl. Acad. Sci. U. S. A. 115, 12704–12709 (2018).

808

39. DeWitt, W. S. et al. Dynamics of the Cytotoxic T Cell Response to a Model of Acute Viral Infection. J. Virol. 89, 4517–4526 (2015).

809

40. Frentsch, M. et al. Direct access to CD4+ T cells specific for defined antigens according to CD154 expression. Nat. Med. 11, 1118–

810

1124 (2005).

811

41. Bacher, P. et al. Antigen-reactive T cell enrichment for direct, high-resolution analysis of the human naive and memory Th cell

812

repertoire. J. Immunol. 190, 3967–76 (2013).

813

42. Reece, W. H. H. et al. A CD4+ T-cell immune response to a conserved epitope in the circumsporozoite protein correlates with

814

protection from natural Plasmodium falciparum infection and disease. Nat. Med. 10, 406–410 (2004).

815

43. Bacher, P. et al. Identification of Immunogenic Antigens from Aspergillus fumigatus by Direct Multiparameter Characterization of

816

Specific Conventional and Regulatory CD4+ T Cells. J. Immunol. 193, 3332–3343 (2014).

817

44. Bacher, P. et al. Antigen-specific expansion of human regulatory T cells as a major tolerance mechanism against mucosal fungi.

818

Mucosal Immunol. 7, 916–928 (2014).

819

45. Rudensky, A. Y. Regulatory T cells and Foxp3. Immunol. Rev. 241, 260–268 (2011).

820

46. Lovelace, P. & Maecker, H. T. Multiparameter Intracellular Cytokine Staining. in Methods in molecular biology (Clifton, N.J.) 1678,

821

151–166 (2018).

822

47. Lin, P. H. et al. Vaccine-induced antigen-specific regulatory T cells attenuate the antiviral immunity against acute influenza virus

823

infection. Mucosal Immunol. 11, 1239–1253 (2018).

824

48. Garner-Spitzer, E. et al. Tick-Borne Encephalitis (TBE) and Hepatitis B Nonresponders Feature Different Immunologic Mechanisms in

825

Response to TBE and Influenza Vaccination with Involvement of Regulatory T and B Cells and IL-10. J. Immunol. 191, 2426–2436

826

(2013).

827

49. Lanteri, M. C. et al. Tregs control the development of symptomatic West Nile virus infection in humans and mice. J. Clin. Invest. 119,

828

3266–3277 (2009).

829

50. Sanchez, A. M., Zhu, J., Huang, X. & Yang, Y. The Development and Function of Memory Regulatory T Cells after Acute Viral

830

Infections. J. Immunol. 189, 2805–2814 (2012).

831

51. Miyara, M. et al. Functional Delineation and Differentiation Dynamics of Human CD4+ T Cells Expressing the FoxP3 Transcription

832

Factor. Immunity 30, 899–911 (2009).

833

52. Booth, N. J. et al. Different Proliferative Potential and Migratory Characteristics of Human CD4 + Regulatory T Cells That Express

834

either CD45RA or CD45RO. J. Immunol. 184, 4317–4326 (2010).

835

53. Ito, T. et al. Two Functional Subsets of FOXP3+ Regulatory T Cells in Human Thymus and Periphery. Immunity 28, 870–880 (2008).

836

54. Mason, G. M. et al. Phenotypic Complexity of the Human Regulatory T Cell Compartment Revealed by Mass Cytometry. J. Immunol.

837

195, 2030–2037 (2015).

838

55. Mohr, A., Malhotra, R., Mayer, G., Gorochov, G. & Miyara, M. Human FOXP3+ T regulatory cell heterogeneity. Clinical and

839

Translational Immunology 7, (2018).

840

56. Wing, J. B., Ise, W., Kurosaki, T. & Sakaguchi, S. Regulatory T cells control antigen-specific expansion of Tfh cell number and

841

humoral immune responses via the coreceptor CTLA-4. Immunity 41, 1013–1025 (2014).

842

57. McHugh, R. S. et al. CD4(+)CD25(+) immunoregulatory T cells: gene expression analysis reveals a functional role for the

843

glucocorticoid-induced TNF receptor. Immunity 16, 311–23 (2002).

844

58. Kachapati, K. et al. The B10 Idd9.3 locus mediates accumulation of functionally superior CD137(+) regulatory T cells in the nonobese

845

diabetic type 1 diabetes model. J. Immunol. 189, 5001–15 (2012).

846

59. Zheng, G., Wang, B. & Chen, A. The 4-1BB costimulation augments the proliferation of CD4+CD25+ regulatory T cells. J. Immunol.

847

173, 2428–34 (2004).

848

60. Mittler, R. S., Bailey, T. S., Klussman, K., Trailsmith, M. D. & Hoffmann, M. K. Anti-4-1BB monoclonal antibodies abrogate T cell-

849

dependent humoral immune responses in vivo through the induction of helper T cell anergy. J. Exp. Med. 190, 1535–40 (1999).

850

61. Kim, Y. H. et al. 4-1BB Triggering Ameliorates Experimental Autoimmune Encephalomyelitis by Modulating the Balance between

851

Th17 and Regulatory T Cells. J. Immunol. 187, 1120–1128 (2011).

852

62. Robinson, C. M. & Pfeiffer, J. K. Viruses and the Microbiota. Annu. Rev. Virol. 1, 55–69 (2014).

853

63. Winkler, E. S. & Thackray, L. B. A long-distance relationship: the commensal gut microbiota and systemic viruses. Current Opinion in

854

Virology 37, 44–51 (2019).

855

64. Yuen, M.-F. et al. Hepatitis B virus infection. Nat. Rev. Dis. Prim. 4, 18035 (2018).

856

65. Chou, H. H. et al. Age-related immune clearance of hepatitis B virus infection requires the establishment of gut microbiota. Proc. Natl.

857

Acad. Sci. U. S. A. 112, 2175–2180 (2015).

858

66. Wu, T. et al. CD4+ T cells play a critical role in microbiota-maintained Anti-HBV immunity in a mouse model. Front. Immunol. 10,

859

(2019).

860

861

862

Supplemental tables

863

Supplementary Table 1. List of 54 single peptides, each 15 AA long with an 11-amino acid

864

overlap spanning the 226 amino acids along the small S protein of hepatitis B (HB) surface

865

antigen (HBsAg)

866

867

Peptide

No.

15 AA with 11 AA overlap

No. of AA at which the peptide start

MENITSGFLGPLLVL

TSGFLGPLLVLQAGF

LGPLLVLQAGFFLLT

LVLQAGFFLLTRILT

AGFFLLTRILTIPQS

LLTRILTIPQSLDSW

ILTIPQSLDSWWTSL

PQSLDSWWTSLNFLG

DSWWTSLNFLGGSPV

TSLNFLGGSPVCLGQ

FLGGSPVCLGQNSQS

SPVCLGQNSQSPTSN

LGQNSQSPTSNHSPT

SQSPTSNHSPTSCPP

TSNHSPTSCPPICPG

SPTSCPPICPGYRWM

CPPICPGYRWMCLRR

CPGYRWMCLRRFIIF

RWMCLRRFIIFLFIL

LRRFIIFLFILLLCL

IIFLFILLLCLIFLL

FILLLCLIFLLVLLD

LCLIFLLVLLDYQGM

FLLVLLDYQGMLPVC

LLDYQGMLPVCPLIP

QGMLPVCPLIPGSTT

101

PVCPLIPGSTTTNTG

105

LIPGSTTTNTGPCKT

109

STTTNTGPCKTCTTP

113

NTGPCKTCTTPAQGN

117

CKTCTTPAQGNSMFP

121

TTPAQGNSMFPSCCC

125

QGNSMFPSCCCTKPT

129

MFPSCCCTKPTDGNC

133

CCCTKPTDGNCTCIP

137

KPTDGNCTCIPIPSS

141

GNCTCIPIPSSWAFA

145

CIPIPSSWAFAKYLW

149

PSSWAFAKYLWEWAS

153

AFAKYLWEWASVRFS

157

YLWEWASVRFSWLSL

161

WASVRFSWLSLLVPF

165

RFSWLSLLVPFVQWF

169

LSLLVPFVQWFVGLS

173

VPFVQWFVGLSPTVW

177

QWFVGLSPTVWLSAI

181

GLSPTVWLSAIWMMW

185

TVWLSAIWMMWYWGP

189

SAIWMMWYWGPSLYS

193

MMWYWGPSLYSIVSP

197

WGPSLYSIVSPFIPL

201

LYSIVSPFIPLLPIF

205

VSPFIPLLPIFFCLW

209

IPLLPIFFCLWVYI

213

868

869

Supplementary Table 2. Overview of the single peptides tested for each vaccinee in the

870

CFSE assay

871

Vaccinee

Gender

Age

Status

peptides_number

single peptides

43.6

Late-converter

51, 35, 50, 34, 54, 38

48.5

Late-converter

53, 5, 37, 49, 1, 33

47.1

Late-converter

21, 20, 5, 53, 4, 52

44.3

Late-converter

23, 39, 7, 21, 37, 5

38.3

Early-converter

17, 33

43.4

Early-converter

47, 41, 42

33.2

Early-converter

49, 1, 33, 41, 9

Early-converter

53, 50, 5, 2, 8, 21

45.9

Early-converter

24, 8, 16

36.3

Late-converter

45, 47

26.5

Early-converter

10, 42, 9, 41, 14, 46

43.3

Early-converter

6, 38, 5, 37

41.3

Early-converter

40, 16, 8

48.7

Early-converter

6, 1

27.2

Non-converter

2, 1, 3, 42, 41, 43

22.3

Early-converter

10, 13, 12, 16

43.9

Late-converter

51, 54, 48, 43, 47, 46

47.7

Late-converter

22, 24, 23, 20, 18, 46

22.6

Non-converter

29, 30, 28, 21, 22, 20

41.2

Late-converter

23, 47, 17, 41, 7, 49

41.7

Early-converter

54, 38, 52, 36, 40, 39

39.5

Early-converter

7, 47, 6, 1, 23, 46

45.1

Late-converter

22, 46, 54, 19, 20, 43

32.4

Early-converter

54, 38, 49, 33, 53, 37

31.4

Early-converter

7, 4, 1, 6, 31, 28

Early-converter

1, 41, 7, 5, 47, 45

22.5

Early-converter

7, 4, 5, 2, 6

50.2

Early-converter

46, 45, 47, 22, 21, 23

23.2

Early-converter

54, 47, 46, 39, 48, 38

23.2

Early-converter

37, 34, 33, 35, 5, 2

45.8

Early-converter

51, 50, 53, 49, 35, 34

21.3

Early-converter

33, 38, 37, 39, 40

21.6

Non-converter

23, 22, 7, 6, 17, 1

22.7

Non-converter

1, 33, 25, 41

872

Figure S1. Gating strategy of ex vivo T cell phenotyping of vaccine-specific T cells.

873

a Gating strategy started by a lymphocyte gate, followed by gating on viable CD3+CD8− T cells.

874

Doublets were excluded using

875

doublet discrimination (area against the height of forward scatter pulse) before gating on CD4+ T

876

cells. Next, CD45RA, CXCR5, CD25 and CD127 were used to identify main subsets of CD4 T

877

cells using Boolean gates as specified in the accompanying table.

878

b Shown an example of gating for CD154 (CD40L) and CD137 (4-1BB) for cells left

879

unstimulated (left) and cells stimulated with a master peptide pool (right) for an early-converter

880

vaccinee at day 60.

881

882

SSC-A

FSC-A

CD3-Alexa Fluor 430

CD8-Viability-DUMP

CD4-PerCP-Cy5.5

FSC-H

FSC-A

FSC-H

CD137pos

0.049

CD154pos

0.050

CD154-APC

CD137-PE

CD137pos

0.13

CD154pos

0.21

CD154-APC

CD137-PE

CD25-BV421

CD127-BV785

CD137neg_CD154pos

CD137pos_CD154neg

CD45RA-AF488

CXCR5-PE-Cy7

CD25-BV421

CD127-BV785

CD45RACXCR5 CD25 CD127

Naive T

+− −

low + high

Memory T

− − −

low + high

Naive T

+ + −

low + high

Memory T

−+−

low + high

CD45RA

REG

+−+

low

CD45RA

−

REG

− − +

low

CD45RA

+ + +

low

CD45RA

−

−+ +

low

FSC-A

FSC-H

883

Figure S2.

884

a Anti-Hepatitis B surface (anti-HBs) titer of vaccinees over four times points, facetted by

885

groups of early, late and non-converters. An anti-HBs titer above 10 IU/ml was considered

886

protective. Early-converters seroconverted at day 60, late-converters seroconverted at day 180 or

887

day 365 and non–converters did not have an anti-HBs titer higher than 10 IU/ml at any of the

888

time points. b CMV, EBV and HSV seropositivity in the three groups of the cohort as

889

determined by serum IgG antibodies to CMV, EBV-VCA, and HSV-1 and 2 using sandwich

890

ELISA. c Age of vaccinees per group.

891

Wilcoxon signed-rank with unpaired and paired analysis as necessary; statistical significance

892

was indicated with ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001

893

894

b c

895

Figure S3

896

●

0.000

0.001

0.002

0.003

0.004

0.005

Day0 Day60

HBV−specific T−cell frequency

CD40L

4-1BB

memory CD4 T cell

CD40L

4-1BB

memory CD4 T cell

a Scatter plot of the DNA-based TCRβ reads for each vaccinee at each time point

897

b Scatter plot of number of unique TCRβ amino acid sequences for each vaccinee at each time

898

point, where the shape denotes the response as based on antibody titer.

899

c Overview of unique TCRβ amino acid sequences in the memory CD4 T cell repertoire of each

900

vaccinee. The bottom blue bar denotes those TCR sequences that were found at both time points.

901

The green and red bars denote the number of unique TCR sequences at each time point. The total

902

bar height thus represents the total number of unique memory CD4 T cell clonotypes sequences

903

for a specific vaccinee.

904

d Change in frequency of those HBV-specific T-cells present at both time points. The (ns) mark

905

denotes a non-significant paired Wilcoxon signed-rank test (P value = 0.7577).

906

e Antibody-defined categories (EC, LC, NC)) plotted against the values from the CD40L assay

907

f ROC curve for the Rhbs predictive model from day 0 data in a leave-one-out cross-validation

908

compared to CD40L+4-1BB- (TCON) memory CD4 T cell values for each vaccinee at time

909

points 60 (AUC = 0.84), 180 (AUC = 0.56) and 365 (AUC = 0.57).

910

g ROC curve for the Rhbs predictive model from day 0 data in a leave-one-out cross-validation

911

compared to CD40L-4-1BB+ (TREG) memory CD4 T cell values for each vaccinee at time

912

points 60 (AUC = 0.62), 180 (AUC = 0.56) and 365 (AUC = 0.52).

913

914

915

Figure S4. Overview of the in vitro expansion experiments for all vaccinees, vaccinees per

916

group and for each vaccinee.

917

918

Figure S5. CD40L+4-1BB− and CD40L−4-1BB+ CD4 T cells have a TCON and TREG

919

phenotype, respectively.

920

921

CD137pos

0.049

CD154pos

0.050

CD154-APC

CD137-PE

CD137pos

0.13

CD154pos

0.21

CD154-APC

CD137-PE

CD25-BV421

CD127-BV785

CD137neg_CD154pos

CD137pos_CD154neg

CD45RA-AF488

CXCR5-PE-Cy7

CD25-BV421

CD127-BV785

CD40L

4-1BB

−

CD4 T cells

CD40L

−

4-1BB

CD4 T cells

a b

922

Figure S6

923

Correlation between the difference in antibody titer between day 365 and day 0 and vaccine-

924

specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell at day 60 colored by vaccinee

925

group and labeled with vaccinee ID.

926

rs, Spearman correlation coefficient, −1 ≤ rs ≤ 1; rs and p value by Spearman’s correlation test

927

928

929

Figure S7 Hepatitis B vaccine induces a vaccine-specific CD4 T cell response in early-

930

converter vaccinees.

931

PBMCs from vaccinees were stimulated with 2 μg/ml of a pool of peptides of HBsAg and

932

assessed for converse expression of 4-1BB and CD40L by flow cytometry on days 0, 60, 180,

933

and 365.

934

a Aggregate analysis from early, late and non-converter vaccinees showing a significant

935

induction of vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+ memory CD4 T cell in early-

936

converters and lack thereof in late and non-converters. Shown is number of of vaccine-specific

937

memory CD4 T cells out of 1x106 memory CD4 T cells after subtraction of responses in negative

938

control (see Methods for details).

939

b Aggregate analysis from early, late and non-converter vaccinees showing a significant

940

induction of vaccine-specific CD40L+4-1BB− and CD40L−4-1BB+ CD4 T cell in early-

941

converters and lack thereof in late and non-converters. Shown is number of of vaccine-specific

942

CD4 T cells out of 1x106 CD4 T cells after subtraction of responses in negative control (see

943

Methods for details).

944

945

Figure S8

946

Aggregate analysis of 4-1BB+CD45RA− TREG within CD45RA− TREG CD4 T cells in early, late

947

and non-converter vaccinees at days 0, 60, 180 and 365.

948

Wilcoxon signed-rank with unpaired and paired analysis as necessary; statistical significance

949

was indicated with ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001

950

951

Current Status of COVID-19 Vaccine Development: Focusing on Antigen Design and Clinical Trials on Later Stages

Article

Full-text available

Feb 2021

The global outbreak of coronavirus disease 2019 (COVID-19) is still threatening human health, economy, and social life worldwide. As a counteraction for this devastating disease, a number of vaccines are being developed with unprecedented speed combined with new technologies. As COVID-19 vaccines are being developed in the absence of a licensed human coronavirus vaccine, there remain further questions regarding the long-term efficacy and safety of the vaccines, as well as immunological mechanisms in depth. This review article discusses the current status of COVID-19 vaccine development, mainly focusing on antigen design, clinical trials in later stages, and immunological considerations for further study.

Recent advances in T-cell receptor repertoire analysis: Bridging the gap with multimodal single-cell RNA sequencing

Article

Jan 2022

T cells exercise a multitude of functions such as cytotoxicity, secretion of immunomodulating cytokines or regulation of tolerance, collectively resulting in an effective control of immune-related disease. Through the unique mechanism of V(D)J recombination, T cells express a highly specific receptor complex known as the T-cell receptor (TCR). Single-cell sequencing technologies have paved the road for interrogating the transcriptome and the paired αβ TCR repertoire of a single T cell in tandem. In contrast, conventional bulk methods are restricted to only one layer of information. This combination of transcriptomic- and repertoire information can provide novel insight into the functional character of T cell immunity. Recently, single-cell technologies have gained in popularity due to improvements in throughput, decrease in cost and the ability for multi-modal experiments that integrate different information layers. Consequently, this prompts the need for the development of novel computational tools that integrate transcriptomic profiles and corresponding features of the TCR repertoire. Here we discuss the current progress in the field of single-cell T cell sequencing, with a focus on the multimodality of new approaches that allow the paired profiling of cellular phenotype and clonotype information. In addition, this review provides detailed descriptions of recent computational developments for analyzing single-cell TCR sequencing data in an integrative manner using novel computational approaches. Finally, we present an overview of the available software tools that can be used to perform integrative analysis of gene expression and TCR profiles.

Preexisting memory CD4+ T cells contribute to the primary response in an HIV-1 vaccine trial

Article

Full-text available

Dec 2021
J CLIN INVEST

Naive and memory CD4+ T cells reactive with human immunodeficiency virus type 1 (HIV-1) are detectable in unexposed, unimmunized individuals. The contribution of preexisting CD4+ T cells to a primary immune response was investigated in 20 HIV-1-seronegative volunteers vaccinated with an HIV-1 envelope (Env) plasmid DNA prime and recombinant modified vaccinia virus Ankara (MVA) boost in the HVTN 106 vaccine trial (clinicaltrials.gov NCT02296541). Prevaccination naive or memory CD4+ T cell responses directed against peptide epitopes in Env were identified in 14 individuals. After priming with DNA, 40% (8/20) of the elicited responses matched epitopes detected in the corresponding preimmunization memory repertoires, and clonotypes were shared before and after vaccination in 2 representative volunteers. In contrast, there were no shared epitope specificities between the preimmunization memory compartment and responses detected after boosting with recombinant MVA expressing a heterologous Env. Preexisting memory CD4+ T cells therefore shape the early immune response to vaccination with a previously unencountered HIV-1 antigen.

Detection of Enriched T Cell Epitope Specificity in Full T Cell Receptor Sequence Repertoires

Article

Full-text available

Nov 2019

High-throughput T cell receptor (TCR) sequencing allows the characterization of an individual's TCR repertoire and directly queries their immune state. However, it remains a non-trivial task to couple these sequenced TCRs to their antigenic targets. In this paper, we present a novel strategy to annotate full TCR sequence repertoires with their epitope specificities. The strategy is based on a machine learning algorithm to learn the TCR patterns common to the recognition of a specific epitope. These results are then combined with a statistical analysis to evaluate the occurrence of specific epitope-reactive TCR sequences per epitope in repertoire data. In this manner, we can directly study the capacity of full TCR repertoires to target specific epitopes of the relevant vaccines or pathogens. We demonstrate the usability of this approach on three independent datasets related to vaccine monitoring and infectious disease diagnostics by independently identifying the epitopes that are targeted by the TCR repertoire. The developed method is freely available as a web tool for academic use at tcrex.biodatamining.be.

CD4 T Cells Play a Critical Role in Microbiota-Maintained Anti-HBV Immunity in a Mouse Model

Article

Full-text available

Apr 2019

The ability of the host to clear hepatitis B virus (HBV) is closely correlated to the establishment of commensal microbiota. However, how microbiota affects anti-HBV immunity is still unclear. Using a well-known hydrodynamical HBV transfection mouse model and treatment with antibiotics (Atb), we explored the change in adaptive immunity (CD4⁺ cells, germinal center B cells and anti-HBs Ab). In our setting, normal mice exhibited complete clearance of HBV within 6 weeks post-hydrodynamic injection (HDI) of HBV-containing plasmid, whereas Atb-treated mice lost this capacity, showing high serum level of hepatitis B surface antigen (HBsAg) without hepatitis B surface antibodies (anti-HBs), similar as what happened in Rag1−/− mice or CD4−/− mice, suggesting that microbiota may influence the function of CD4⁺ T cells. Furthermore, the numbers of splenic and hepatic effector CD4⁺ T cells (CD44hiCD62L⁻CD4⁺ T cells) both decreased with impaired function (IFN-γ synthesis), resulting in lower frequency of germinal center B cells and CD4⁺ follicular helper T cells, and impaired anti-HBs production. We further tried to find the bacterial species responsible for maintaining anti-HBV immunity, and found that each antibiotic alone could not significantly influence HBV clearance compared to antibiotic combination, suggesting that global commensal microbial load is critical for promoting HBV clearance. We also confirmed that TLRs (e.g., TLR2, 4, 9) are not major players in immune clearance of HBV using their agonists and knock-out mice. These results suggest that commensal microbiota play an important role in maintaining CD4⁺ T cell immunity against HBV infection.

Human T cell receptor occurrence patterns encode immune history, genetic background, and receptor specificity

Article

Full-text available

Aug 2018
eLife

The T cell receptor (TCR) repertoire encodes immune exposure history through the dynamic formation of immunological memory. Statistical analysis of repertoire sequencing data has the potential to decode disease associations from large cohorts with measured phenotypes. However, the repertoire perturbation induced by a given immunological challenge is conditioned on genetic background via major histocompatibility complex (MHC) polymorphism. We explore associations between MHC alleles, immune exposures, and shared TCRs in a large human cohort. Using a previously published repertoire sequencing dataset augmented with high-resolution MHC genotyping, our analysis reveals rich structure: striking imprints of common pathogens, clusters of co-occurring TCRs that may represent markers of shared immune exposures, and substantial variations in TCR-MHC association strength across MHC loci. Guided by atomic contacts in solved TCR:peptide-MHC structures, we identify sequence covariation between TCR and MHC. These insights and our analysis framework lay the groundwork for further explorations into TCR diversity.

Precise tracking of vaccine-responding T-cell clones reveals convergent and personalized response in identical twins

Article

Full-text available

Apr 2018

T-cell receptor (TCR) repertoire data contain information about infections that could be used in disease diagnostics and vaccine development, but extracting that information remains a major challenge. Here we developed a statistical framework to detect TCR clone proliferation and contraction from longitudinal repertoire data. We applied this framework to data from three pairs of identical twins immunized with the yellow fever vaccine. We identified 500-1500 responding TCRs in each donor and validated them using three independent assays. While the responding TCRs were mostly private, albeit with higher overlap between twins, they could be well predicted using a classifier based on sequence similarity. Our method can also be applied to samples obtained post-infection, making it suitable for systematic discovery of new infection-specific TCRs in the clinic.

Vaccine-induced antigen-specific regulatory T cells attenuate the antiviral immunity against acute influenza virus infection

Article

Full-text available

Feb 2018

Peptide-based T cell vaccines targeting the conserved epitopes of influenza virus can provide cross-protection against distantly related strains, but they are generally not immunogenic. Foreign antigen-specific regulatory T (Treg) cells are induced under subimmunogenic conditions peripherally, although their development and role in vaccine-mediated antiviral immunity is unclear. Here, we demonstrated primary vaccination with peptides alone significantly induced antigen-specific Foxp3+ Treg cells, which were further expanded by repeated vaccination with unadjuvanted peptides. Certain adjuvants, including CpG, suppressed the induction and expansion of antigen-specific Treg cells by peptide vaccination. Interestingly, secondary influenza virus infection significantly increased the frequency of preexisting antigen-specific Treg cells, although primary infection barely induced them. Importantly, specific depletion of vaccine-induced antigen-specific Treg cells promoted influenza viral clearance, indicating their inhibitory role in vivo. Immunization with CpG-adjuvanted peptides by the subcutaneous prime–intranasal-boost strategy restricted the recruitment and accumulation of antigen-specific Treg cells in lung, and stimulated robust T cell immunity. Finally, subcutaneous prime–intranasal-boost immunization with CpG-adjuvanted peptides or whole-inactivated influenza vaccines protected mice from heterosubtypic influenza virus infection. In conclusion, antigen-specific Treg cells induced by peptide vaccines attenuate the antiviral immunity against influenza virus infection. CpG-adjuvanted peptide vaccines provide heterosubtypic influenza protection probably by inhibiting Treg development and enhancing T cell immunity.

Human FOXP3 + T regulatory cell heterogeneity

Article

Full-text available

Jan 2018

FOXP3-expressing CD4⁺ T regulatory (Treg) cells are instrumental for the maintenance of self-tolerance. They are also involved in the prevention of allergy, allograft rejection, foetal rejection during pregnancy and of exaggerated immune response towards commensal pathogens in mucosal tissues. They can also prevent immune responses against tumors and promote tumor progression. FOXP3-expressing Treg cells are not a homogenous population. The different subsets of Treg cells can have different functions or roles in the maintenance of immune homeostasis and can therefore be differentially targeted in the management of autoimmune diseases or in cancer. We discuss here how Treg cell subsets can be differentiated phenotypically, functionally and developmentally in humans.

Activation-induced surface proteins in the identification of antigen-responsive CD4 T cells

Article

Dec 2019
IMMUNOL LETT

Identification of antigen specificity of CD4 T cells is instrumental in understanding adaptive immune responses in health and disease. The high diversity of CD4 T cell repertoire combined with the functional heterogeneity of the compartment poses a challenge to the assessment of CD4 T cell responses. In spite of that, multiple technologies allow direct or indirect interrogation of antigen specificity of CD4 T cells. In the last decade, multiple surface proteins have been established as cytokine-independent surrogates of in vitro CD4 T cell activation, and have found applications in the live identification and isolation of antigen-responsive CD4 T cells. Here we review the current knowledge of the surface proteins that permit identification of viable antigen-responsive CD4 T cells with high specificity, including those capable of identifying specialized CD4 T subsets such as germinal center follicular helper T cells and CD4 regulatory T cells.

A long-distance relationship: the commensal gut microbiota and systemic viruses

Article

Aug 2019

Recent advances defining the role of the commensal gut microbiota in the development, education, induction, function, and maintenance of the mammalian immune system inform our understanding of how immune responses govern the outcome of systemic virus infection. While characterization of the impact of the local oral, respiratory, dermal and genitourinary microbiota on host immune responses and systemic virus infection is in its infancy, the gut microbiota interacts with host immunity systemically and at distal non-gastrointestinal tract sites to modulate the pathogenesis of systemic viruses. Gut microbes, microbe-associated molecular patterns, and microbe-derived metabolites engage receptors expressed on the cell surface, in the endosome, or in the cytoplasm to orchestrate optimal innate and adaptive immune responses important for controlling systemic virus infection.

On the viability of unsupervised T-cell receptor sequence clustering for epitope preference

Article

Sep 2018
BIOINFORMATICS

Motivation: The T-cell receptor (TCR) is responsible for recognizing epitopes presented on cell surfaces. Linking TCR sequences to their ability to target specific epitopes is currently an unsolved problem, yet one of great interest. Indeed, it is currently unknown how dissimilar TCR sequences can be before they no longer bind the same epitope. This question is confounded by the fact that there are many ways to define the similarity between two TCR sequences. Here we investigate both issues in the context of TCR sequence unsupervised clustering. Results: We provide an overview of the performance of various distance metrics on two large independent data sets with 412 and 2835 TCR sequences respectively. Our results confirm the presence of structural distinct TCR groups that target identical epitopes. In addition, we put forward several recommendations to perform unsupervised T-cell receptor sequence clustering. Availability and implementation: Source code implemented in Python 3 available at https://github.com/pmeysman/TCRclusteringPaper. Supplementary information: Supplementary data are available at Bioinformatics online.

Hepatitis B virus infection

Article

Jun 2018

Hepatitis B virus (HBV) is a hepatotropic virus that can establish a persistent and chronic infection in humans through immune anergy. Currently, 3.5% of the global population is chronically infected with HBV, although the incidence of HBV infections is decreasing owing to vaccination and, to a lesser extent, the use of antiviral therapy to reduce the viral load of chronically infected individuals. The course of chronic HBV infection typically comprises different clinical phases, each of which potentially lasts for decades. Well-defined and verified serum and liver biopsy diagnostic markers enable the assessment of disease severity, viral replication status, patient risk stratification and treatment decisions. Current therapy includes antiviral agents that directly act on viral replication and immunomodulators, such as interferon therapy. Antiviral agents for HBV include reverse transcriptase inhibitors, which are nucleoside or nucleotide analogues that can profoundly suppress HBV replication but require long-term maintenance therapy. Novel compounds are being actively investigated to achieve the goal of HBV surface antigen seroclearance (functional cure), a serological state that is associated with a higher remission rate (thus, no viral rebound) after treatment cessation and a lower rate of cirrhosis and hepatocellular carcinoma. This Primer addresses several aspects of HBV infection, including epidemiology, immune pathophysiology, diagnosis, prevention and management.

Preexisting memory CD4 T cells in naïve individuals confer robust immunity upon vaccination

Abstract and Figures

Recommended publications

Preexisting memory CD4 T cells in naïve individuals confer robust immunity upon hepatitis B vaccinat...

Memory CD4+ T cell receptor repertoire data mining as a tool for identifying cytomegalovirus serosta...

Transcriptome profiling in blood before and after hepatitis B vaccination shows significant differen...

Transcriptomic profiling of different responder types in adults after a Priorix® vaccination