Bordetella pertussis Proteins Dominating the Major HistocompatibilityComplex Class II-Presented Epitope Repertoire in Human Monocyte-Derived Dendritic Cells

Rachel M. Stenger,a Hugo D. Meiring,b Betsy Kuipers,a Martien Poelen,a Jacqueline A. M. van Gaans-van den Brink,a

Claire J. P. Boog,b Ad P. J. M. de Jong,b Cécile A. C. M. van Elsa

Centre for Immunology of Infectious Diseases and Vaccines, National Institute for Public Health and the Environment, Bilthoven, the Netherlandsa; Institute forTranslational Vaccinology, Bilthoven, the Netherlandsb

Knowledge of naturally processed Bordetella pertussis-specific T cell epitopes may help to increase our understanding of the ba-sis of cell-mediated immune mechanisms to control this reemerging pathogen. Here, we elucidate for the first time the dominantmajor histocompatibility complex (MHC) class II-presented B. pertussis CD4� T cell epitopes, expressed on human monocyte-derived dendritic cells (MDDC) after the processing of whole bacterial cells by use of a platform of immunoproteomics technol-ogy. Pertussis epitopes identified in the context of HLA-DR molecules were derived from two envelope proteins, i.e., putativeperiplasmic protein (PPP) and putative peptidoglycan-associated lipoprotein (PAL), and from two cytosolic proteins, i.e., 10-kDa chaperonin groES protein (groES) and adenylosuccinate synthetase (ASS). No epitopes were detectable from known viru-lence factors. CD4� T cell responsiveness in healthy adults against peptide pools representing epitope regions or full proteinsconfirmed the immunogenicity of PAL, PPP, groES, and ASS. Elevated lymphoproliferative activity to PPP, groES, and ASS insubjects within a year after the diagnosis of symptomatic pertussis suggested immunogenic exposure to these proteins duringclinical infection. The PAL-, PPP-, groES-, and ASS-specific responses were associated with secretion of functional Th1 (tumornecrosis factor alpha [TNF-�] and gamma interferon [IFN-�]) and Th2 (interleukin 5 [IL-5] and IL-13) cytokines. Relative pau-city in the natural B. pertussis epitope display of MDDC, not dominated by epitopes from known protective antigens, can inter-fere with the effectiveness of immune recognition of B. pertussis. A more complete understanding of hallmarks in B. pertussis-specific immunity may advance the design of novel immunological assays and prevention strategies.

Prevention of morbidity and mortality caused by the humanpathogen B. pertussis has effectively relied on national vacci-

nation programs since these were introduced in the 1940s and1950s (1, 2). However, during the last decade an increase ofwhooping cough has been reported (3–9). Pertussis, mostly fearedfor affecting infants too young to be fully vaccinated, is more andmore notified among adolescents and adults, who apparentlygradually lose their vaccine-induced protection to current B. per-tussis strains (2, 10–13). Understanding of protective adaptive im-munity and its weaknesses is essential for us to be able to improvepertussis vaccination. Despite being implicated in protectionagainst severe pertussis (14, 15), levels of preexposure human se-rum antibodies to the major vaccine antigens have never beenexclusively correlated with the efficacy of pertussis vaccines (16–18). In addition to antibodies, CD4� T cells contribute to immu-nological resistance against B. pertussis infection. First, CD4� Tcell responses are essential for avidity maturation of specific B cellresponses. Second, murine (19–25) and human (26–33) B. pertus-sis-specific CD4� T cell immunity has been associated with thesecretion of Th1-, Th2-, and Th17-type cytokines capable of arm-ing different types of antimicrobial effector mechanisms, such asthe activities of macrophages, eosinophils, and neutrophils, re-spectively (34). Unfortunately, information on the fine-specificityand breadth of human B. pertussis-specific CD4� T cell responsesis largely absent, not in the least because of the difficulty of prop-agating B. pertussis-specific CD4� T cells in vitro (R. M. Stenger,M. Poelen, and C. A. C. M. van Els, unpublished data). The fewhuman B. pertussis-specific CD4� T cell epitopes that have beenelucidated were all sought only on the basis of known virulence

factors (29, 35–37). We were interested in gaining nonbiased in-sight into the natural repertoire of B. pertussis-specific major his-tocompatibility complex (MHC) class II-presented epitopes,available for T cell recognition at the cell surface of human pro-fessional antigen-presenting cells after processing of the wholebacterial proteome. We used a dedicated immunoproteomicsmethod to identify naturally processed and HLA-DR-associatedB. pertussis epitopes isolated from human monocyte-derived den-dritic cells (MDDC). Stable isotope labeling of B. pertussis biomassprior to loading on MDDC and nanoscale liquid chromatographyelectrospray ionization mass spectrometry (NanoLC-ESI-MS)were used to sensitively detect pertussis peptide epitopes amongthousands of self-peptides in complex peptide fractions (38). Thisapproach led to the identification of four B. pertussis proteins assources for dominantly selected and functional CD4� T cell targetepitopes. Interestingly, none of the source proteins was a knownvirulence factor currently used in acellular pertussis vaccines.

Understanding peculiarities in the display of the natural B.

Received 26 October 2013 Returned for modification 29 November 2013Accepted 20 February 2014

Published ahead of print 5 March 2014

Editor: T. S. Alexander

Address correspondence to Cécile A. C. M. van Els, cecile.van.els@rivm.nl.

Copyright © 2014, American Society for Microbiology. All Rights Reserved.

doi:10.1128/CVI.00665-13

The authors have paid a fee to allow immediate free access to this article.

May 2014 Volume 21 Number 5 Clinical and Vaccine Immunology p. 641– 650 cvi.asm.org 641

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

pertussis CD4� T cell epitope repertoire on MDDC, such as strongepitope domination, may help to elucidate weaknesses in the hu-man cellular immune response to pertussis and may provide leadson how to design new generations of pertussis assays and vaccines.

MATERIALS AND METHODSSubjects and ethics statement. Blood from volunteers was obtained inaccordance with the Declaration of Helsinki and with Dutch regulations,following approval, respectively, from the Sanquin Ethical AdvisoryBoard (for citrated buffy coat donation from 20 HLA-typed healthy bloodbank donors and for one leukapheresis donation from an HLA-typedhealthy blood bank donor [trial BS03.0015-x]) and from the accreditedReview Board STEG and the review boards of collaborating hospitals forheparinized blood sampling of participants from a clinical study (agerange, 8 to 77 years [median, 40 years]) (30 pertussis patients within 12months after laboratory-confirmed diagnosis of B. pertussis infection and10 healthy household contacts negative for B. pertussis infection based ondiagnostic serology [trial NVI-243, CCMO number NL16334.040.07]).All participants provided written informed consent for the collection ofsamples and subsequent analysis.

Isolation of PBMC. Peripheral blood mononuclear cells (PBMC)from citrated blood samples were separated by centrifugation on a Ficoll-Hypaque gradient (Pharmacia Biotech, Uppsala Sweden) or on a gel inheparinized CPT tubes (BD Biosciences). After washing and counting,cells were used directly or after cryopreservation.

Bacterial strains and growth conditions. The B. pertussis vaccinestrain 509 was grown on Bordet-Gengou (BG) agar plates containing 15%defibrinated sheep blood. Liquid B. pertussis cultures were grown in eithernatural 14N-containing minimal Bioexpress cell growth medium or in98-atom%-enriched 15N stable isotope containing minimal Bioexpresscell growth medium (Cambridge Isotope Laboratories) (both containing0.15% lactic acid [Fluka, Switzerland] adjusted to pH 7.2 with NaOH)until reaching stationary phase and expressing virulence antigens. At thisstage, cultures containing B. pertussis whole cells were heat inactivated for30 min at 56°C and concentrated 5 times in phosphate-buffered saline(PBS) by centrifugation at 2,000 � g for 20 min. Before use, heat-inacti-vated 14N and 15N bacterial biomasses were subjected to quality control.Nanoscale liquid chromatography electrospray ionization tandem massspectrometry (LC-MS/MS) sequence analysis of trypsin-digested singleSDS-PAGE bands of the 14N and 15N biomass confirmed uniform stableisotope labeling of pertussis proteins. The optical densities of both whole-cell biomasses were measured at 600 nm (A600), and based on these A600

values, a 1:1 mixture of the 14N- and 15N-concentrated whole-cell bio-masses was prepared for antigen pulsing of human MDDC (step A in theimmunoproteomics strategy, Fig. 1A).

Proteins and synthetic peptides. Recombinant P.69 pertactin (P.69Prn) was expressed and purified from an Escherichia coli construct aspreviously described (39). Pertussis toxin (Ptx), filamentous hemaggluti-nin (FHA), and fimbriae subtypes 2 and 3 (Fim2/3) were purified in-house according to procedures described in the literature (40–42). Purityof all antigens was confirmed by SDS-PAGE.

Synthetic peptides representing the exact HLA-DR eluted peptide se-quences, as well as 12 amino acid (aa)-overlapping 18-mer peptides rep-resenting the entire protein sequences of 10-kDa chaperonin groES pro-tein (groES) (B. pertussis gene identification number BP3496 [n � 13peptides]), putative periplasmic protein (PPP) (BP3341 [n � 37 pep-tides]), adenylosuccinate synthetase (ASS) (BP2188 [n � 71 peptides]),and putative peptidoglycan-associated lipoprotein (PAL) (BP3342 [n �25 peptides]), respectively, were prepared by solid-phase synthesis usingN�-(9-fluorenylmethoxy carbonyl) (FMOC)-protected amino acids anda SYRO II simultaneous multiple peptide synthesizer (MultiSyntechGmbH, Witten, Germany). The purity and identity of the synthesizedpeptides were assessed by reverse-phase high-performance liquid chro-matography and mass spectrometry. Peptides were used as single pep-

tides, combined adjacent peptides representing a broadened epitope re-gion, or peptide pools representing an entire protein.

Immunoblotting. For evaluation of the relative protein contents ofknown virulence factors, 14N and 15N B. pertussis whole-cell preparationswere run on SDS-PAGE gels (Pierce, Switzerland) and transferred to ni-trocellulose membranes (Schleicher & Schuell). Membranes were washedwith Tris/NaCl buffer (pH 7.4) with 0.5% Tween 80 (WS-T), and thenprobed for 1 h at room temperature (RT) in WS-T with the mouse mono-clonal antibodies anti-FHA (31E2), anti-P.69 Prn (Pem4), anti-Ptx S1(151C1), and anti-Fim2 (136A6) (all from the Dutch National Institutefor Public Health and the Environment [43, 44]). Thereafter, the mem-branes were washed and incubated for 1 h in WS-T containing alkalinephosphatase-labeled goat-anti-mouse IgG (Southern Biotechnology As-sociates, Inc.) and 0.5% Protifar (Nutricia, the Netherlands), respectively.Signal was detected with the ready-to-use alkaline phosphatase (AP) con-jugate substrate kit (Bio-Rad).

Culturing, antigen pulse, and characterization of MDDC. Immaturehuman MDDC were cultured based on a procedure described by Sallustoet al. (34). Briefly, 1 � 109 PBMC obtained from an HLA-DR2-homozy-gous blood donor after leukapheresis were seeded at �5 � 106/ml in150-mm tissue culture dishes (Corning Costar) in Iscove’s modified Dul-becco’s medium (GibcoBRL) supplemented with 1% fetal bovine serum(HyClone) and penicillin-streptomycin-glutamine (GibcoBRL) at 37°Cand 5% CO2 in a humidified incubator for 2 h. After removal of thenonadherent fraction, adherent cells were further cultured for 6 days inmedium containing 500 U/ml recombinant human granulocyte-macro-phage colony-stimulating factor (GM-CSF) (PeproTech) and 250 U/mlrecombinant human IL-4 (Strathmann Biotech GmbH, Germany). Cul-ture medium and growth factors were refreshed on day 3. On day 6, 1.2 �108 MDDC, still immature, were pulsed with a 1:1 wt/wt mixture of 14N-and 15N-concentrated bacterial biomass at an optimized final concentra-tion/A600 of 0.028 and incubated for 6 h. Thereafter, cells were furthercultured and maturated in the continuous presence of antigen, growthfactors, and 20 ng/ml lipopolysaccharide (Salmonella abortus equi;Sigma). On day 8, B. pertussis-pulsed and maturated MDDC were har-vested and washed four times in PBS, counted, pelleted, and snap frozenbefore cell lysis and peptide isolation (Fig. 1B). Maturation of the MDDCwas verified using flow cytometry by comparing the expression of CD83,CD40, CD80, and CD86 cell surface markers on day 6 and day 8 MDDC(data not shown).

Isolation and NanoLC-ESI-MS analysis of HLA-DR-associated pep-tides. B. pertussis-pulsed MDDC were lysed and HLA-DR-peptide com-plexes were isolated essentially as described previously (38), using theHLA-DR-specific monoclonal antibody B8.11.2 bound to CNBr-acti-vated Sepharose 4B beads. Peptides were eluted using 10% acetic acid andspun through a 10-kDa cutoff Spinfilter (Millipore, USA) (Fig. 1C). Thefiltrate was concentrated to �10 �l using a freeze dryer and prefraction-ated using strong cation exchange (SCX) chromatography. Peptide frac-tions were reconstituted in a 5% formic acid and 5% dimethyl sulfoxidesolution in water containing angiotensin-III and oxytocin (each at a con-centration of 250 amol/�l) as internal standard peptides and analyzedusing NanoLC-ESI-MS/MS, as described earlier (45) (Fig. 1D). Charac-teristic heavy and light mass spectral doublets, representing MHC classII-associated peptide epitopes originating from corresponding 14N and15N B. pertussis protein homologues, were allocated in all mass spectrausing the MS-Xelerator mass spectral interpretation software (MsMetrix,the Netherlands) (Fig. 1E). Candidate epitopes were identified by targetedNanoLC-ESI-MS/MS analysis, using identical chromatographic condi-tions on the same quadrupole time of flight (Q-TOF Ultima API) massspectrometer operated at optimized collision energy (Fig. 1F) and data-base searching for a sequence match. The expressed levels of epitopes werequantified based on the relative response factor (RRF) of each naturallyprocessed and presented epitope relative to the two additional standardpeptides in an SCX fraction and the RRF of the corresponding synthetic

Stenger et al.

642 cvi.asm.org Clinical and Vaccine Immunology

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

analogue compared to these two standard peptides when acquired underidentical NanoLC-ESI-MS conditions.

Functional assays to analyze immunogenicity of epitopes. Immuno-genicity of epitopes was tested by incubating PBMC in the absence orpresence of the indicated synthetic peptide or peptide pool (Fig. 1G) (at 1�M per peptide) in complete AIM-V medium (AIM-V medium contain-ing streptomycin, gentamicin, and L-glutamine [GibcoBRL] supple-mented with 2% human AB serum [Harlan]) at 105 cells per well in 150 �l

in 96-well round-bottom plates (Greiner) for 6 or 10 days in 3-, 10-, or40-fold replicated wells per condition, as indicated, at 37°C in a humidi-fied 5% CO2 atmosphere. Plates were visually inspected and either super-natants were harvested for cytokine measurement, 0.5 �Ci (18.5 kBq)[3H]thymidine (Amersham, USA) was added overnight to measure directproliferation, or cultures were expanded on IL-2 and restimulated, asdescribed below, before the addition of [3H]thymidine. Eighteen hoursafter labeling, cells were harvested and the [3H]thymidine incorporation

FIG 1 Immunoproteomics strategy to identify pathogen derived MHC class II-associated peptides. (A) Separate B. pertussis (Bp) cultures are prepared in14N-containing or in 98 atom% 15N-enriched prokaryotic medium. This allows all bacterial proteins synthesized to incorporate heavy or light nitrogen residues.(B) Immature MDDC are loaded with a 1:1 (A600 ratio) mixture of 14N- and 15N-labeled heat-inactivated B. pertussis whole-cell suspensions and MDDC are leftfor 48 h to process and present bacterial antigens while cultured in normal (14N) eukaryotic medium. (C) HLA-DR-epitope complexes are affinity purified usingmonoclonal antibody B.8.11.2 and epitopes are acid eluted and separated from the HLA-DR molecules by size exclusion. (D) After the peptide sample is highlyfractionated by SCX chromatography, a portion of each SCX fraction is subjected to NanoLC-ESI-MS analysis. (E) Pathogen-derived epitopes are easily allocatedby searching mass spectra for 14N- and 15N-ion doublets with similar intensities and retention times, and an average mass difference of 1.2%. Notably,self-epitopes will form only a 14N-ion. (F) Candidate epitopes are identified by targeted NanoLC-ESI-MS/MS sequence analysis and database matching. (G)Functionality of the naturally processed and presented MHC class II epitopes is assessed in T cell assays using in vitro restimulation of human T cells withantigen-presenting cells (APC) and synthetic peptides.

B. pertussis MHC Class II Epitope Repertoire

May 2014 Volume 21 Number 5 cvi.asm.org 643

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

was determined as counts per minute (CPM) using a Wallac 1205 beta-plate liquid scintillation counter. Results are expressed as stimulation in-dex (SI) from triplicate (donors from the NVI-243 study) or decuple wells(healthy blood donors), calculated as follows: average cpm of PBMC inthe presence of peptide(s)/average cpm of PBMC in medium only (directproliferation) or average cpm of stimulated PBMC cultures in the pres-ence of peptide-pulsed antigen-presenting cells/average cpm of stimu-lated PBMC cultures in the presence of mock-pulsed antigen-presentingcells (indirect proliferation). SI values of �1.5 were considered positive.

For cytokine analysis, human Th1/Th2 and Th17 cytokine Bio-Plexkits (Bio-Rad) were used to determine concentrations of IL-2, IL-4, IL-5,IL-10, IL-12(p70), IL-13, IL-17, tumor necrosis factor alpha (TNF-�),and gamma interferon (IFN-�) according to the manufacturer’s instruc-tions in (pooled) supernatants taken, as indicated, from those wells out of40 plated per specific synthetic peptide pool that were visibly activated.

FIG 2 Confirmation of the presence of four virulence factors in the 14N- and15N-labeled B. pertussis whole-cell vaccines. Immunoblots of SDS-PAGE-sep-arated antigen preparations using specific monoclonal antibodies with indi-cated antigen specificities. Lane 1, purified antigen. Lane 2, B. pertussis culturedon Bordet Gengou plates. Lane 3, 14N-labeled B. pertussis culture. Lane 4,15N-labeled B. pertussis culture. FHA, filamentous hemagglutinin; P.69 Prn,P.69 pertactin; Ptx, pertussis toxin; Fim2, fimbriae type 2. The position ofmolecular mass (kDa) of protein markers is indicated.

FIG 3 Mass-tag-assisted identification of naturally processed and presented groES34 –52 peptide. (A) Triply charged mass spectral doublets representing thenative groES34 –52 epitope. (B) The deconvoluted mass spectrum indicating a putative peptide containing 22 nitrogen atoms. The number of pathogen-derivednitrogen atoms contained within an epitope can be deducted from the mass difference between its 14N and 15N isomers. (C) MS/MS fragmentation spectrum ofthe groES34 –52 epitope.

Stenger et al.

644 cvi.asm.org Clinical and Vaccine Immunology

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

Measurements and data analysis were performed with the Bio-Plex systemin combination with Bio-Plex manager software. Results are expressed inpg/ml.

For analysis of cross-reactivity, PBMC from a groES34 –52-reactivehealthy donor were stimulated using groES34 –52 synthetic peptide for 10days and then further expanded on 10 ng/ml IL-2 and fresh medium. Onday 24, expanded T cells were restimulated for 2 days in triplicate wellsusing medium only, synthetic peptide representing B. pertussis groES34 –52,or synthetic peptide representing Homo sapiens hsp1040 –58 in the presenceof autologous antigen-presenting cells. Then [3H]thymidine was addedand incorporation was measured after overnight incubation, as indicatedabove. Results are expressed as stimulation index (SI) from triplicatewells.

Statistics. The statistical significance of differences in ex vivo lym-phoproliferation between groups was calculated with the Student’s t test(two-tailed, unequal variance).

RESULTSImmunoproteomics strategy for sensitive detection of MHCclass II-presented B. pertussis epitopes. Foreign MHC class II-presented epitopes can be isolated from antigen-presenting cells forNanoLC-ESI-MS analysis, but to distinguish them from thousands ofself-derived epitopes in complex peptide mixtures is technicallyhighly challenging. To rapidly recognize B. pertussis-specific peptidemasses in MHC class II eluates, an immunoproteomics strategywas applied, based on metabolic labeling of the bacterial proteome(schematically illustrated in Fig. 1A to F). Antigenically, the 14Nand 15N whole-cell biomasses, harvested in the virulence (bvg�)phase, were comparable, as was confirmed by immunoblotting ofthe known virulence factors P.69 pertactin (P.69 Prn), pertussistoxin subunit 1 (PtxS1), filamentous hemagglutinin (FHA), andfimbriae 2 (Fim2) (Fig. 2). After processing of 1:1 wt/wt mixedbacterial biomasses by MDDC, an HLA-DR-derived peptide sam-ple was obtained, fractionated, and analyzed for the presence ofcandidate B. pertussis-derived epitopes (Fig. 1D to F).

B. pertussis PAL, groES, PPP, and ASS identified as dominantsource proteins of naturally processed and MHC class II-pre-sented epitopes. An example of such a candidate epitope is givenin Fig. 3A and B. Epitope sequencing by targeted nanoscale LC-MS/MS and database matching indeed identified its source pro-tein as a B. pertussis protein, groES (Fig. 3C). In total, seven natu-rally processed and presented and hitherto unknown peptides,representing epitopes from four different B. pertussis proteins, in-cluding length variants, were identified using isotope-tagged iden-

tification (Table 1). A peptide representing aa residues 103 to 118from the envelope PAL protein, PAL103–118, was the most abun-dant epitope, with an estimated MHC-peptide copy number of350 epitopes per MDDC. Lower epitope densities were found for anatural length variant of this epitope and epitopes from groES,PPP, and ASS. No mass spectral doublets were assigned to knownB. pertussis virulence factors, despite the fact that these proteinswere present in the B. pertussis biomass preparations that wereused for MDDC loading (Fig. 2). Also, data-dependent mass se-quencing of peptides present in the eluate did reveal the presenceof a wide repertoire of HLA-DR-associated self-sequences (datanot shown), but this approach did not yield any additional B.pertussis epitopes.

Naturally processed and HLA-DR-presented epitope regionsrepresent functional CD4� T cell targets. To test the immunoge-nicity of the processed and HLA-DR2-presented pertussis pep-tides, PBMC from an HLA-DR-matched blood donor were stim-ulated ex vivo using synthetic peptides representing the elutedgroES34 –52, PPP132–146, ASS166 –179, and PAL103–118 sequences. Inthis donor, direct lymphoproliferation could be measured againstall four epitopes (Fig. 4).

We reasoned that the identified peptide sequences may notonly represent the exact HLA-DR2-presented epitopes, but mayalso indicate a broader protein region liberated by proteolysis andpossibly containing multiple HLA-DR-binding motifs. We alsoassumed, since the epitopes were in vitro derived from an experi-mental whole-cell vaccine, that childhood whole-cell pertussisvaccination could be sufficient to prime CD4� T cell responses tothese epitopes. To evaluate this, PBMC from 20 healthy HLA-DR2-typed adults from a birth cohort associated with whole-cellpertussis vaccination were stimulated with pools of overlappingsynthetic peptides representing the epitope regions, includingflanking sequences, to encompass adjacent motifs that were pre-dicted to bind a wide array of HLA-DR molecules (46, 47) (datanot shown). As summarized in Fig. 5, direct specific proliferationwas observed against all four epitope regions with 30 to 40% over-

TABLE 1 Naturally processed and HLA-DR2-presented epitopes of B.pertussis whole-cell biomass

Source protein Epitope Abundancea

Chaperonin groES protein(groES) (BP3496b)

34KPDQGEVVAVGPGKKTED51 534KPDQGEVVAVGPGKKTEDG52 80

Putative periplasmicprotein (PPP)(BP3341)

135IALYPNSQLAPT146 30132AAFIALYONSQLAPT146 175

Adenylosuccinatesynthetase (ASS)(BP2188)

166LAEVLDYHNFVLTQ179 10

Putative peptidoglycan-associated lipoprotein(PAL) (BP3342)

103GGAEYNLALGQRRA116 10103GGAEYNLALGQRRADA118 350

a Copies/cell.b B. pertussis gene identification number (48).

FIG 4 Recognition of HLA-DR2-presented B. pertussis epitopes groES34 –52,PPP132–146, ASS166 –179, and PAL103–118. Direct lymphoproliferation to the dif-ferent epitopes was measured using PBMC from a HLA-DR2 homozygousblood bank donor using synthetic peptides representing groES34 –52,PPP132–146, ASS166-179-, and PAL103–118 in a proliferation assay. [3H]Thymi-dine incorporation was determined 10 days after in vitro stimulation. Barsrepresent SI SD from decuple wells.

B. pertussis MHC Class II Epitope Repertoire

May 2014 Volume 21 Number 5 cvi.asm.org 645

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

all responsiveness, not limited to HLA-DR2� donors only. Thesedata suggest that the extended epitope regions of groES, PPP, ASS,and PAL are recognized in a broader immunogenetic context.

To investigate the type of helper T cell responses associatedwith epitope immunogenicity, PBMC from responding donors 1,5, 14, and 19 were cocultured in a new set of restimulations withrelevant sets of peptides and cytokine analysis on culture superna-tants of wells with activated cultures. All epitope regions wereassociated with both Th1- (IFN-� and TNF-�) and Th2-type(IL-5 and IL-13) cytokine responsiveness (Fig. 6). No other cyto-kines, such as IL-10 or IL-17, were detected.

Specific proliferation as well as cytokine production of ex-panded CD4� T cell cultures from various donors could beblocked by monoclonal antibodies against HLA-DR molecules,confirming the involvement of HLA-DR as a restriction element(data not shown). Collectively, these findings indicate that thefour HLA-DR2-eluted epitopes represent favorably processed andpresented protein regions immunogenic for CD4� T cells andhave a mixed Th1/Th2 cytokine signature.

Epitope-specific responses observed after clinical pertussisinfection. To test whether epitopes from groES, PPP, ASS, andPAL were naturally exposed during infection, PBMC obtainedfrom pertussis patients within 12 months after their laboratory-confirmed clinical episode (n � 30) and from healthy controls(n � 30) were stimulated with overlapping synthetic 18-mer pep-tides covering the full protein sequences and were assayed forproliferative responses. As illustrated in Fig. 7, a higher percentageof patients had proliferative responses above the SI cutoff of 1.5

than did healthy controls when tested against peptide sets ofgroES, PPP, and ASS, but not of PAL. When tested statistically,differences in lymphoproliferation between groups were not sig-nificant. The trend, however, may implicate a boosting effect ofinfection for CD4� T cell responses of groES, PPP, and ASS.

groES34 –52-specific CD4� T cells do not cross-react to the ho-mologous human hsp10 epitope. The 10-kDa chaperonin groESprotein is a family member of the low molecular weight heat shockproteins, which are conserved not only between bacteria, but alsobetween humans and other mammal species. The identified B.pertussis groES epitope shows 52.5% sequence homology with thecounterpart region on Homo sapiens hsp10 (Fig. 8). To investigatewhether this homology could be immunologically relevant oreven underlie autoimmunity, we compared the capacity ofgroES34 –52 and its human homologue hsp1040 –58 to restimulatean in vitro-expanded B. pertussis groES34 –52-specific CD4� T cellbulk culture in the presence of autologous antigen-presentingcells. As shown in Fig. 9, no recognition of the H. sapiens hsp10epitope was observed, suggesting that priming of human T cellswith the dominantly presented B. pertussis groES region is unlikelyto result in cross-reactivity against endogenous tissue.

DISCUSSION

Using an unconventional immunoproteomics approach, we iden-tified hitherto unknown naturally processed and dominantly pre-sented B. pertussis-specific CD4� T cell epitopes. This approachhas two major advantages over strategies aimed at mappingepitopes on predefined protein targets. First, discovery of epitopesis made independent of their protein source, which is especiallyrelevant for pathogens with large proteomes, such as B. pertussis,with more than 3,800 open reading frames (48). Second, relativeabundances and natural length variants of epitopes can be deter-mined, which may have an immunological impact.

The B. pertussis epitopes discovered in this study to be MHCclass II presented and immunogenic in humans originated fromfour different bacterial source proteins localizing to two differentsubcellular compartments of B. pertussis whole cells. PPP (alter-native name, YbgF) and PAL, encoded by adjacent genes on thesame operon, are both components of the cell envelope Tol-Palsystem, critical for the integrity of the bacterial outer membrane(49). The 10-kDa chaperonin groES, involved in protein foldingand assembly as a homoheptameric ring associated with the chap-eronin cpn60, and ASS, catalyzing the first step in the de novobiosynthesis of AMP, are both located in the cytoplasm (48).

Unexpectedly, none of the dominantly presented epitopesoriginated from known B. pertussis virulence factors, such as FHA,P.69 Prn, PtxS1, or Fim2, while these proteins were abundantlypresent in the digested bacterial biomass (Fig. 2) and some encode

ronod 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 groES19-54 PPP133-162 ASS163-192 PAL103-138

FIG 5 Ex vivo lymphoproliferation to peptides representing naturally MHC class II-presented B. pertussis epitope regions in a panel of healthy HLA-typeddonors. N- and C-terminal amino acid positions of protein region are represented by 12-mer overlapping 18-mer synthetic peptides. HLA-DR2� donors, 1, 11,16, and 17; HLA-DR2 donors, 2 to 9, 12 to 15, and 18 to 20; untyped donor, 10. White cells, mean SI (of decuple wells per stimulation) of �1.4; light-gray cells,mean SI of �1.4 and �3.0; dark-gray cells, mean SI of �3.0.

groES19-54 Donor

PPP133-162 Donor

1 41 5 1 41 5Responsiveness 16 21 1 Responsiveness 8 5 1

Th1 IFN-γ

Th1 IFN-γ

TNF-α TNF-α

Th2 IL-5

Th2 IL-5

IL-13 IL-13

ASS163-192 Donor

PAL103-138 Donor

91 41 5 1 91 5 1Responsiveness 10 9 2 Responsiveness 1 8 2 5

Th1 IFN-γ

Th1 IFN-γ

TNF-α TNF-α

Th2 IL-5

Th2 IL-5

IL-13 IL-13

FIG 6 Cytokines associated with responses against naturally MHC class II-presented B. pertussis epitopes. Shown (as responsiveness) are the numbers ofvisually responding wells out of 40. The concentration of indicated cytokinesin pg/ml was determined in (pooled) supernatant(s) of responding well(s) perdonor per protein. White cells, �25 pg/ml; light-gray cells, �25 and �75pg/ml; dark-gray cells, �75 and �150 pg/ml; black cells, �150 pg/ml.

Stenger et al.

646 cvi.asm.org Clinical and Vaccine Immunology

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

known T cell epitopes (29, 35, 50, 51). In fact, preceding B. pertus-sis harvesting, in-culture gene expression values for groES, PPP,ASS, and PAL were comparably high (groES and PAL) or evenlower (PPP and ASS) than those for FHA, P.69 Prn, PtxS1, or Fim2(52) (B. van der Waterbeemd, personal communication).

Therefore, a high antigen load may be favorable, but it is notsufficient for a source protein to dominate in the MHC class IIligandome of MDDC. It is known that for epitopes to be formed,modest but not destructive proteolysis of source proteins (53) andpeptide affinity to MHC class II molecules (54) are prerequisites.The groES, PPP, ASS, and epitope regions all have very high pre-dicted epitope binding scores for either HLA-DRB1*1501 orHLA-DRB5*0101, the two HLA-DR molecules expressed on theHLA-DR2� MDDC (data not shown). Hence, they can be as-sumed to be strong competitors for binding with epitopes derivedfrom other proteins. Also, other features driving exclusive presen-tation by MDDC could be involved, such as those described for

the Toxoplasma gondii protein profilin, being immunodominantin the CD4� T cell response to the pathogen solely because ofenhanced and selective TLR11-mediated uptake (55). TLR2 couldplay a role in the uptake of the lipoprotein PAL (56–62), possiblyin association with the other Tol-Pal component PPP. TLR2might also sense groES, eventually via associated compounds (63–65). It is unknown whether ASS has innate receptor affinity.

Another unexpected observation was the modest averageepitope abundance of the pertussis MHC class II ligands, withestimated numbers of MHC-peptide complexes per cell of 5 to350. This was dissimilar to the range of MHC class II-presentedepitopes we described previously, using an MDDC model withmeningococcal outer membrane vesicles as the antigen and apply-ing the same degree of sample prefractionation and analytical sen-sitivity. HLA-DR-associated meningococcal epitopes isolatedfrom either HLA-DR1 homozygous or HLA-DR2 homozygousMDDC ranged from 30 to 10,000 copies per cell (reference 38 and

FIG 7 Clinical infection enhances CD4� T cell lymphoproliferation to different B. pertussis epitope regions. PBMC of pertussis patients within 12 months afterlaboratory-confirmed pertussis (n � 30) and of healthy controls (n � 10 noninfected household contacts and n � 20 healthy blood bank donors) were stimulatedwith pools of 18-mer peptides, representing the entire indicated proteins at 1 �M per peptide. [3H]Thymidine incorporation was determined 7 days after in vitrostimulation. Dots represent SI from triplicate wells of different individuals.

Species Amino acid sequence

B. pertussis groES ��������������������� �������� ���

H. sapiens hsp10 ������������� ������������������ ���

FIG 8 Sequence homology between B. pertussis and H. sapiens 10-kDa chaperonins. Gray letters indicate the amino acid sequence of the longest naturallyprocessed B. pertussis groES34 –52 peptide variant and identical amino acid residues in H. sapiens hsp10; dark-gray letters indicate nonhomologous amino-acidsin the H. sapiens hsp10 epitope sequence.

B. pertussis MHC Class II Epitope Repertoire

May 2014 Volume 21 Number 5 cvi.asm.org 647

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

H. Meiring, unpublished data). Although we cannot exclude acertain degree of peptide loss, potentially also from unidentifiedepitopes during the sample isolation procedure, or of undersam-pling of MHC ligand identification at the mass spectrometry level(66), it is unlikely that this fully accounts for the large differencebetween the pertussis and the meningococcal epitope repertoire.MDDC pulsed with B. pertussis whole-cell vaccine had compara-ble levels of HLA-DR molecules at the cell surface, of MHC pro-tein retentate, and of self-peptides in the eluate as did meningo-coccal outer membrane vesicle maturated MDDC (data notshown). Hence, our data indicate a selective and limited breadthand density of B. pertussis MHC class II epitopes on humanMDDC, among which additional epitopes from groES, PPP, ASS,and PAL or epitopes from virulence factors were either absent ortoo low in copy number to be detected by our system. Limitedepitope display, especially from virulence factors, could circum-vent effective immune recognition and mechanisms crucial forprotection. While it has been shown that B. pertussis antigens caninterfere with MDDC maturation and function (67), low andhighly selective epitope expression on MDDC might be anothersuccessful feature of B. pertussis to evade the adaptive immuneresponse. Further immunoproteomics research, extending MHCclass II ligandome analyses to other HLA-DR alleles, antigen-pre-senting cell types, and antigen preparations and to experimentalinfection, is needed to substantiate this hypothesis.

The roles of groES, PPP, ASS, and PAL antigens as targets inhost immunity remain to be elucidated. A direct protective role astargets for CD4� T cell-dependent antibody responses, whichcould be tested in an animal challenge model, is not expected,since the antigens are not surface exposed on the bacterium. How-ever, the antigens could otherwise serve as relevant targets of im-mune mechanisms. The induced specific CD4� T cell populationscould enhance pertussis immunity via the generation of an in-flammatory cytokine milieu, or alternatively, could dampen other

CD4� T cell specificities through competition for space (68) orthrough regulation. The CD4� T cell responses to the identifiedepitope regions, observed in a considerable proportion of testedindividuals, were associated with a mixed Th1/Th2 cytokine pro-file, i.e., secretion of IFN-�, TNF-�, and IL-5 and IL-13, in theabsence of IL-10 and IL-17. IFN-� and TNF-� are both inflam-matory cytokines that have been implicated in controlling B. per-tussis through potentiating bactericidal activity of macrophagesand enhancing phagocytosis by neutrophils (69–71), suggesting arole in protection. Notably, (ex)patients had slightly elevated pro-liferative responses to three of four naturally processed target pro-teins, possibly indicating that the epitopes were exposed duringinfection and could play a role in natural CD4� T cell immunity toB. pertussis. While the data set was possibly too limited to be con-clusive and testing more patients’ samples could help to settle thispoint, it is also possible that elevation of T cell responses is con-fined to a very narrow time span after boosting. The majority ofpatients in this study were adults who were within 2 months aftertheir formal diagnosis. However, since B. pertussis has a relativelylong incubation time (14 to 21 days) and adults are often diag-nosed only after a period of prolonged coughing, an early boostingeffect of the epitopes on T cell levels could have easily been missed.

Responses to the B. pertussis groES epitope were non-cross-reactive to the endogenous human hsp10 homologue. Therefore,dissimilar to the CD4� T cell epitope cross-reactivities observedfor some mycobacterial and human hsp70 protein families, a po-tential role in autoimmunity or immunoregulation for B. pertussisgroES is not envisaged (64).

In conclusion, an unexpected limited epitope breadth andabundance of human MDDC-presented B. pertussis-specificCD4� T cell epitopes was found, with a role for other proteinsthan those that are known virulence factors. Clearly, althoughbeing a nonroutine and highly technical approach, immunopro-teomics can shed light on classes of otherwise elusive T cellepitopes. These results represent a step toward a more completecharacterization of the natural immune response to B. pertussis,involving other specificities than hitherto anticipated. Ultimately,a better understanding of the immunological correlates of protec-tion against whooping cough and their flaws is needed to developnew functional assays and more effective vaccines.

ACKNOWLEDGMENTS

We are grateful to the volunteers and personnel from Sanquin Blood BankNorth West for blood donations (S03.0015-x) and to volunteers and em-ployees from the SKI-study (NVI-243). We thank P. Hoogerhout, H. F.Brugghe, and J. A. M. Timmermans for peptide synthesis, J. A. M. Tim-mermans for preparing figures, B. van de Waterbeemd and P. van der Leyfor discussion, and D. van Baarle for advice.

REFERENCES1. Cohen HH. 1963. Development of pertussis vaccine production and con-

trol in the National Institute of Public Health in the Netherlands duringthe years 1850 –1962. Antonie Van Leeuwenhoek 29:183–201. http://dx.doi.org/10.1007/BF02046052.

2. Hewlett EL, Edwards KM. 2005. Clinical practice. Pertussis—not just for kids. N.Engl. J. Med. 352:1215–1222. http://dx.doi.org/10.1056/NEJMcp041025.

3. De Serres G, Boulianne N, Douville FM, Duval B. 1995. Pertussis inQuebec: ongoing epidemic since the late 1980s. Can. Commun. Dis. Rep.21:45– 48.

4. de Melker HE, Conyn-van Spaendonck MA, Rumke HC, van Wijn-gaarden JK, Mooi FR, Schellekens JF. 1997. Pertussis in the Netherlands:an outbreak despite high levels of immunization with whole-cell vaccine.Emerg. Infect. Dis. 3:175–178. http://dx.doi.org/10.3201/eid0302.970211.

FIG 9 B. pertussis groES34 –52-specific T cells do not recognize the homologoushsp1040 –58 from H. sapiens. A short-term cultured B. pertussis groES34 –52-specific T cell line from a homozygous HLA-DR2 blood donor was restimu-lated with medium, the homologous peptide, or the corresponding H. sapienshsp1040 –58 sequence, in the presence of autologous antigen-presenting cells.[3H]Thymidine incorporation was determined after 3 days of restimulation.Bars represent SI SD from triplicate wells.

Stenger et al.

648 cvi.asm.org Clinical and Vaccine Immunology

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

5. Bass JW, Wittler RR. 1994. Return of epidemic pertussis in the UnitedStates. Pediatr. Infect. Dis. J. 13:343–345. http://dx.doi.org/10.1097/00006454-199405000-00002.

6. Andrews R, Herceg A, Roberts C. 1997. Pertussis notifications in Aus-tralia, 1991 to 1997. Commun. Dis. Intell. 21:145–148.

7. Guris D, Strebel PM, Bardenheier B, Brennan M, Tachdjian R, Finck E,Wharton M, Livengood JR. 1999. Changing epidemiology of pertussis inthe United States: increasing reported incidence among adolescents andadults, 1990 –1996. Clin. Infect. Dis. 28:1230 –1237. http://dx.doi.org/10.1086/514776.

8. Gzyl A, Augustynowicz E, Rabczenko D, Gniadek G, Slusarczyk J. 2004.Pertussis in Poland. Int. J. Epidemiol. 33:358 –365. http://dx.doi.org/10.1093/ije/dyh012.

9. Centres for Disease Control and Prevention. 2013. Notifiable diseasesand mortality tables. MMWR Morb. Mortal. Wkly. Rep. 61:719 –732.

10. Wendelboe AM, Van Rie A, Salmaso S, Englund JA. 2005. Durationof immunity against pertussis after natural infection or vaccination.Pediatr. Infect. Dis. J. 24:S58 –S61. http://dx.doi.org/10.1097/01.inf.0000160914.59160.41.

11. He Q, Mertsola J. 2008. Factors contributing to pertussis resurgence. Future.Microbiol. 3:329–339. http://dx.doi.org/10.2217/17460913.3.3.329.

12. Crowcroft NS, Pebody RG. 2006. Recent developments in pertussis. Lancet367:1926–1936. http://dx.doi.org/10.1016/S0140-6736(06)68848-X.

13. Berbers GA, de Greeff SC, Mooi FR. 2009. Improving pertussis vaccina-tion. Hum. Vaccin. 5:497–503.

14. Cherry JD, Gornbein J, Heininger U, Stehr K. 1998. A search for sero-logic correlates of immunity to Bordetella pertussis cough illnesses. Vac-cine 16:1901–1906. http://dx.doi.org/10.1016/S0264-410X(98)00226-6.

15. Storsaeter J, Hallander HO, Gustafsson L, Olin P. 1998. Levels of anti-pertussis antibodies related to protection after household exposure to Borde-tella pertussis. Vaccine 16:1907–1916. http://dx.doi.org/10.1016/S0264-410X(98)00227-8.

16. Cherry JD. 2007. Immunity to pertussis. Clin. Infect. Dis. 44:1278 –1279.http://dx.doi.org/10.1086/514350.

17. Olin P, Hallander HO, Gustafsson L, Reizenstein E, Storsaeter J. 2001.How to make sense of pertussis immunogenicity data. Clin. Infect. Dis.33(Suppl 4):S288 –S291. http://dx.doi.org/10.1086/322564.

18. Mills KH. 2001. Immunity to Bordetella pertussis. Microbes Infect. 3:655–677. http://dx.doi.org/10.1016/S1286-4579(01)01421-6.

19. Andreasen C, Powell DA, Carbonetti NH. 2009. Pertussis toxin stimu-lates IL-17 production in response to Bordetella pertussis infection in mice.PLoS One 4:e7079. http://dx.doi.org/10.1371/journal.pone.0007079.

20. Ausiello CM, Lande R, Stefanelli P, Fazio C, Fedele G, Palazzo R,Urbani F, Mastrantonio P. 2003. T-cell immune response assessment asa complement to serology and intranasal protection assays in determiningthe protective immunity induced by acellular pertussis vaccines in mice.Clin. Diagn. Lab. Immunol. 10:637– 642. http://dx.doi.org/10.1128/CDLI.10.4.637-642.2003.

21. Higgins SC, Jarnicki AG, Lavelle EC, Mills KH. 2006. TLR4 mediatesvaccine-induced protective cellular immunity to Bordetella pertussis: roleof IL-17-producing T cells. J. Immunol. 177:7980 –7989.

22. Leef M, Elkins KL, Barbic J, Shahin RD. 2000. Protective immunity toBordetella pertussis requires both B cells and CD4(�) T cells for key func-tions other than specific antibody production. J. Exp. Med. 191:1841–1852. http://dx.doi.org/10.1084/jem.191.11.1841.

23. Mills KH, Barnard A, Watkins J, Redhead K. 1993. Cell-mediatedimmunity to Bordetella pertussis: role of Th1 cells in bacterial clearance ina murine respiratory infection model. Infect. Immun. 61:399 – 410.

24. Cahill ES, O’Hagan DT, Illum L, Barnard A, Mills KH, Redhead K.1995. Immune responses and protection against Bordetella pertussis infec-tion after intranasal immunization of mice with filamentous haemagglu-tinin in solution or incorporated in biodegradable microparticles. Vaccine13:455– 462. http://dx.doi.org/10.1016/0264-410X(94)00008-B.

25. van den Berg BM, David S, Beekhuizen H, Mooi FR, van Furth R. 2000.Protection and humoral immune responses against Bordetella pertussisinfection in mice immunized with acellular or cellular pertussis immuno-gens. Vaccine 19:1118 –1128. http://dx.doi.org/10.1016/S0264-410X(00)00329-7.

26. Vermeulen F, Verscheure V, Damis E, Vermeylen D, Leloux G, Dirix V,Locht C, Mascart F. 2009. Cellular immune responses of preterm infantsafter vaccination with whole cell or acellular pertussis vaccines. Clin. Vac-cine Immunol. 17:258 –262. http://dx.doi.org/10.1128/CVI.00328-09.

27. Dirix V, Verscheure V, Goetghebuer T, Hainaut M, Debrie AS, Locht C,

Mascart F. 2009. Monocyte-derived interleukin-10 depresses the Bordetella per-tussis-specific gamma interferon response in vaccinated infants. Clin. Vac-cine Immunol. 16:1816 –1821. http://dx.doi.org/10.1128/CVI.00314-09.

28. Guiso N, Njamkepo E, Vié le Sage F, Zepp F, Meyer CU, Abitbol V, ClytiN, Chevallier S. 2007. Long-term humoral and cell-mediated immunity afteracellular pertussis vaccination compares favourably with whole-cell vaccines 6years after booster vaccination in the second year of life. Vaccine 25:1390–1397. http://dx.doi.org/10.1016/j.vaccine.2006.10.048.

29. Stenger RM, Poelen MC, Moret EE, Kuipers B, Bruijns SC, HoogerhoutP, Hijnen M, King AJ, Mooi FR, Boog CJ, van Els CA. 2009. Immu-nodominance in mouse and human CD4� T-cell responses specific for theBordetella pertussis virulence factor P.69 pertactin. Infect. Immun. 77:896 –903. http://dx.doi.org/10.1128/IAI.00769-08.

30. Ryan M, Murphy G, Gothefors L, Nilsson L, Storsaeter J, Mills KH.1997. Bordetella pertussis respiratory infection in children is associatedwith preferential activation of type 1 T helper cells. J. Infect. Dis. 175:1246 –1250. http://dx.doi.org/10.1086/593682.

31. Ausiello CM, Lande R, la Sala A, Urbani F, Cassone A. 1998. Cell-mediatedimmune response of healthy adults to Bordetella pertussis vaccine antigens. J.Infect. Dis. 178:466 – 470. http://dx.doi.org/10.1086/515628.

32. Mascart F, Verscheure V, Malfroot A, Hainaut M, Pierard D, Temer-man S, Peltier A, Debrie AS, Levy J, Del Giudice G, Locht C. 2003.Bordetella pertussis infection in 2-month-old infants promotes type 1 T cellresponses. J. Immunol. 170:1504 –1509.

33. Mascart F, Hainaut M, Peltier A, Verscheure V, Levy J, Locht C. 2007.Modulation of the infant immune responses by the first pertussis vaccineadministrations. Vaccine 25:391–398. http://dx.doi.org/10.1016/j.vaccine.2006.06.046.

34. Sallusto F, Lanzavecchia A. 1994. Efficient presentation of soluble anti-gen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregu-lated by tumor necrosis factor alpha. J. Exp. Med. 179:1109 –1118. http://dx.doi.org/10.1084/jem.179.4.1109.

35. De Magistris MT, Romano M, Bartoloni A, Rappuoli R, Tagliabue A.1989. Human T cell clones define S1 subunit as the most immunogenicmoiety of pertussis toxin and determine its epitope map. J. Exp. Med.169:1519 –1532. http://dx.doi.org/10.1084/jem.169.5.1519.

36. Petersen JW, Holm A, Ibsen PH, Haslov K, Capiau C, Heron I. 1992.Identification of human T-cell epitopes on the S4 subunit of pertussistoxin. Infect. Immun. 60:3962–3970.

37. Peppoloni S, Nencioni L, Di Tommaso A, Tagliabue A, Parronchi P,Romagnani S, Rappuoli R, De Magistris MT. 1991. Lymphokine secre-tion and cytotoxic activity of human CD4� T-cell clones against Bordetellapertussis. Infect. Immun. 59:3768 –3773.

38. Meiring HD, Kuipers B, van Gaans-van den Brink JA, Poelen MC, Tim-mermans H, Baart G, Brugghe H, van Schie J, Boog CJ, de Jong AP, vanEls CA. 2005. Mass tag-assisted identification of naturally processed HLAclass II-presented meningococcal peptides recognized by CD4� T lympho-cytes. J. Immunol. 174:5636–5643.

39. Hijnen M, van Gageldonk PG, Berbers GA, van Woerkom T, Mooi FR. 2005.The Bordetella pertussis virulence factor P.69 pertactin retains its immuno-logical properties after overproduction in Escherichia coli. Protein Expr.Purif. 41:106 –112. http://dx.doi.org/10.1016/j.pep.2005.01.014.

40. Robinson A, Gorringe AR, Funnell SG, Fernandez M. 1989. Serospecificprotection of mice against intranasal infection with Bordetella pertussis.Vaccine 7:321–324. http://dx.doi.org/10.1016/0264-410X(89)90193-X.

41. Sekura RD, Fish F, Manclark CR, Meade B, Zhang YL. 1983. Pertussistoxin. Affinity purification of a new ADP-ribosyltransferase. J. Biol.Chem. 258:14647–14651.

42. Sato Y, Cowell JL, Sato H, Burstyn DG, Manclark CR. 1983. Separationand purification of the hemagglutinins from Bordetella pertussis. Infect.Immun. 41:313–320.

43. Poolman JT, Kuipers B, Vogel ML, Hamstra HJ, Nagel J. 1990. Descrip-tion of a hybridoma bank towards Bordetella pertussis toxin and surfaceantigens.Microb.Pathog.8:377–382.http://dx.doi.org/10.1016/0882-4010(90)90024-K.

44. Hijnen M, de VR, Mooi FR, Schepp R, Moret EE, van Gageldonk P,Smits G, Berbers GA. 2007. The role of peptide loops of the Bordetellapertussis protein P.69 pertactin in antibody recognition. Vaccine 25:5902–5914. http://dx.doi.org/10.1016/j.vaccine.2007.05.039.

45. Meiring HD, van der Heeft E, ten hove GJ, de Jong APJM. 2002.Nanoscale LC-MS(n): technical design and applications to peptide and

B. pertussis MHC Class II Epitope Repertoire

May 2014 Volume 21 Number 5 cvi.asm.org 649

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from

protein analysis. J. separation science 25:557–568. http://dx.doi.org/10.1002/1615-9314(20020601)25:9�557::AID-JSSC557�3.0.CO;2-F.

46. Bian H, Reidhaar-Olson JF, Hammer J. 2003. The use of bioinformaticsfor identifying class II-restricted T-cell epitopes. Methods 29:299 –309.http://dx.doi.org/10.1016/S1046-2023(02)00352-3.

47. Bian H, Hammer J. 2004. Discovery of promiscuous HLA-II-restricted Tcell epitopes with TEPITOPE. Methods 34:468 – 475. http://dx.doi.org/10.1016/j.ymeth.2004.06.002.

48. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, HoldenMT, Churcher CM, Bentley SD, Mungall KL, Cerdeño-Tárraga AM,Temple L, James K, Harris B, Quail MA, Atkin R, Baker S, Basham D,Bason N, Cherevach I, Chillingworth T, Collins M, Cronin A, Davis P,Doggett J, Feltwell T, Goble A, Hamlin N, Hauser H, Holroyd S, JagelsK, Leather S, Moule S, Norberczak H, O’Neil S, Ormond D, Price C,Rabbinowitsch E, Rutter S, Sanders M, Saunders D, Seeger K, Sharp S,Simmonds M, Skelton J, Squares R, Squares S, Stevens K, Unwin L,Whitehead S, Barrell BG, Maskell DJ. 2003. Comparative analysis of thegenome sequences of Bordetella pertussis, Bordetella parapertussis and Bor-detella bronchiseptica. Nat. Genet. 35:32– 40. http://dx.doi.org/10.1038/ng1227.

49. Godlewska R, Wisniewska K, Pietras Z, Jagusztyn-Krynicka EK. 2009.Peptidoglycan-associated lipoprotein (Pal) of Gram-negative bacteria:function, structure, role in pathogenesis and potential application in im-munoprophylaxis. FEMS Microbiol. Lett. 298:1–11. http://dx.doi.org/10.1111/j.1574-6968.2009.01659.x.

50. De Magistris MT, Romano M, Nuti S, Rappuoli R, Tagliabue A. 1988.Dissecting human T cell responses against Bordetella species. J. Exp. Med.168:1351–1362. http://dx.doi.org/10.1084/jem.168.4.1351.

51. De Magistris MT, Di Tommaso A, Domenighini M, Censini S, TagliabueA, Oksenberg JR, Steinman L, Judd AK, O’Sullivan D, Rappuoli R. 1992.Interaction of the pertussis toxin peptide containing residues 30–42 withDR1 and the T-cell receptors of 12 human T-cell clones. Proc. Natl. Acad. Sci.U. S. A. 89:2990–2994. http://dx.doi.org/10.1073/pnas.89.7.2990.

52. van de Waterbeemd B, Streefland M, Pennings J, van der Pol L, BeuveryC, Tramper J, Martens D. 2009. Gene-expression-based quality scoresindicate optimal harvest point in Bordetella pertussis cultivation for vac-cine production. Biotechnol. Bioeng. 103:900 –908. http://dx.doi.org/10.1002/bit.22326.

53. Watts C. 2012. The endosome-lysosome pathway and information gen-eration in the immune system. Biochim. Biophys. Acta 1824:14 –21. http://dx.doi.org/10.1016/j.bbapap.2011.07.006.

54. Rammensee HG, Friede T, Stevanoviic S. 1995. MHC ligands and pep-tide motifs: first listing. Immunogenetics 41:178 –228. http://dx.doi.org/10.1007/BF00172063.

55. Yarovinsky F, Kanzler H, Hieny S, Coffman RL, Sher A. 2006. Toll-likereceptor recognition regulates immunodominance in an antimicrobialCD4� T cell response. Immunity 25:655– 664. http://dx.doi.org/10.1016/j.immuni.2006.07.015.

56. Brightbill HD, Libraty DH, Krutzik SR, Yang RB, Belisle JT, BleharskiJR, Maitland M, Norgard MV, Plevy SE, Smale ST, Brennan PJ, BloomBR, Godowski PJ, Modlin RL. 1999. Host defense mechanisms triggeredby microbial lipoproteins through toll-like receptors. Science 285:732–736. http://dx.doi.org/10.1126/science.285.5428.732.

57. Aliprantis AO, Yang RB, Mark MR, Suggett S, Devaux B, Radolf JD,Klimpel GR, Godowski P, Zychlinsky A. 1999. Cell activation and apop-tosis by bacterial lipoproteins through toll-like receptor-2. Science 285:736 –739. http://dx.doi.org/10.1126/science.285.5428.736.

58. Hirschfeld M, Kirschning CJ, Schwandner R, Wesche H, Weis JH,Wooten RM, Weiss JJ. 1999. Cutting edge: inflammatory signaling byBorrelia burgdorferi lipoproteins is mediated by toll-like receptor 2. J. Im-munol. 163:2382–2386.

59. Lee HK, Lee J, Tobias PS. 2002. Two lipoproteins extracted from Esch-erichia coli K-12 LCD25 lipopolysaccharide are the major componentsresponsible for Toll-like receptor 2-mediated signaling. J. Immunol. 168:4012– 4017.

60. Liang MD, Bagchi A, Warren HS, Tehan MM, Trigilio JA, Beasley-Topliffe LK, Tesini BL, Lazzaroni JC, Fenton MJ, Hellman J. 2005.Bacterial peptidoglycan-associated lipoprotein: a naturally occurring toll-like receptor 2 agonist that is shed into serum and has synergy with lipo-polysaccharide. J. Infect. Dis. 191:939 –948. http://dx.doi.org/10.1086/427815.

61. Asai Y, Makimura Y, Ogawa T. 2007. Toll-like receptor 2-mediateddendritic cell activation by a Porphyromonas gingivalis synthetic lipopep-tide. J. Med. Microbiol. 56:459 – 465. http://dx.doi.org/10.1099/jmm.0.46991-0.

62. Shim HK, Kim JY, Kim MJ, Sim HS, Park DW, Sohn JW, Kim MJ. 2009.Legionella lipoprotein activates toll-like receptor 2 and induces cytokineproduction and expression of costimulatory molecules in peritoneal mac-rophages. Exp. Mol. Med. 41:687– 694. http://dx.doi.org/10.3858/emm.2009.41.10.075.

63. Kirschning CJ, Schumann RR. 2002. TLR2: cellular sensor for microbialand endogenous molecular patterns. Curr. Top. Microbiol. Immunol.270:121–144.

64. van Eden W, van der Zee R, Prakken B. 2005. Heat-shock proteinsinduce T-cell regulation of chronic inflammation. Nat. Rev. Immunol.5:318 –330. http://dx.doi.org/10.1038/nri1593.

65. Tsan MF, Gao B. 2009. Heat shock proteins and immune system. J.Leukoc. Biol. 85:905–910. http://dx.doi.org/10.1189/jlb.0109005.

66. Hassan C, Kester MG, de Ru AH, Hombrink P, Drijfhout JW, NijveenH, Leunissen JA, Heemskerk MH, Falkenburg JH, van Veelen PA. 2013.The human leukocyte antigen-presented ligandome of B lymphocytes.Mol. Cell. Proteomics 12:1829 –1843. http://dx.doi.org/10.1074/mcp.M112.024810.

67. Stefanelli P, Fazio C, Fedele G, Spensieri F, Ausiello CM, MastrantonioP. 2009. A natural pertactin deficient strain of Bordetella pertussis showsimproved entry in human monocyte-derived dendritic cells. New Micro-biol. 32:159 –166.

68. Kedl RM, Rees WA, Hildeman DA, Schaefer B, Mitchell T, Kappler J,Marrack P. 2000. T cells compete for access to antigen-bearing antigen-presenting cells. J. Exp. Med. 192:1105–1113. http://dx.doi.org/10.1084/jem.192.8.1105.

69. Mobberley-Schuman PS, Weiss AA. 2005. Influence of CR3 (CD11b/CD18) expression on phagocytosis of Bordetella pertussis by human neu-trophils. Infect. Immun. 73:7317–7323. http://dx.doi.org/10.1128/IAI.73.11.7317-7323.2005.

70. Mahon BP, Mills KH. 1999. Interferon-gamma mediated immune effec-tor mechanisms against Bordetella pertussis. Immunol. Lett. 68:213–217.http://dx.doi.org/10.1016/S0165-2478(99)00070-X.

71. Torre D, Ferrario G, Bonetta G, Perversi L, Tambini R, Speranza F.1994. Effects of recombinant human gamma interferon on intracellularsurvival of Bordetella pertussis in human phagocytic cells. FEMS Immunol.Med. Microbiol. 9:183–188. http://dx.doi.org/10.1111/j.1574-695X.1994.tb00492.x.

Stenger et al.

650 cvi.asm.org Clinical and Vaccine Immunology

on Novem

ber 28, 2020 by guesthttp://cvi.asm

.org/D

ownloaded from