Iv(

AAa

b

c

d

a

ARRAA

KHVIP

1

imdmcchith[Lt

c

V

0d

Vaccine 29 (2011) 2797–2802

Contents lists available at ScienceDirect

Vaccine

journa l homepage: www.e lsev ier .com/ locate /vacc ine

mmunogenicity and protection against Haemophilus parasuis infection afteraccination with recombinant virulence associated trimeric autotransportersVtaA)

lex Olveraa,b,∗,1, Sonia Pinab,c,1, Marta Pérez-Simób, Virginia Aragónb,c, Joaquim Segalésb,d,lbert Bensaidb

Departament de Genètica i de Microbiologia, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès (Bellaterra), Barcelona, SpainCentre de Recerca en Sanitat Animal, (CReSA), UAB-IRTA, Campus de la Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, SpainInstitut de Recerca i Tecnologia Agroalimentàries (IRTA), Barcelona, SpainDepartment de Sanitat i Anatomia Animals, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain

r t i c l e i n f o

rticle history:eceived 23 September 2010eceived in revised form 21 January 2011ccepted 30 January 2011

a b s t r a c t

Haemophilus parasuis is the etiological agent of Glässer’s disease in swine, characterized by fibrinouspolyserositis, polyarthritis and meningitis. The lack of a vaccine against a broad spectrum of strains haslimited the control of the disease. Recently, virulence associated trimeric autotransporters (VtaA) weredescribed as antigenic proteins of H. parasuis. In this study 6 VtaA were produced as recombinant proteins

vailable online 12 February 2011

eywords:aemophilus parasuistaA

mmunogenicity

and used to immunize snatch-farrowed, colostrum-deprived piglets. Immunized animals developed spe-cific systemic and mucosal antibodies. The protective capacity of the anti-VtaA antibodies was evaluatedby the inoculation of 3 × 108 or 6 × 106 colony forming units (CFU) of the highly virulent strain Nagasaki.Vaccinated animals had a delayed course of disease and 33 or 57%, respectively, of the animals survivedthe lethal challenge. The partial protection achieved with the recombinant VtaA supports their potential

ed in

rotection as candidates to be includ

. Introduction

Haemophilus parasuis is the etiological agent of Glässer’s diseasen swine, pathologically characterized by fibrinous polyserositis,

eningitis and polyarthritis. In naïve pigs disease onset occurs feways after H. parasuis exposure and is usually lethal. Those ani-als recovering from the acute phase of the disease can develop

hronic arthritis. Historically, it was a sporadic disease of young pigsompromised by stress or immunosuppression. The introduction ofigh health status farms and early weaning resulted in immunolog-

cally naïve populations increasing the morbidity and mortality ofhe disease [1]. To date, 15 serovars of H. parasuis, which typicallyave a range of different virulence potentials, have been described2]. In addition, a high percentage of strains are non-serotypeable.

imited cross-protection among strains has complicated the con-rol of Glässer’s disease by vaccination with bacterins [1,3,4].

The development of antibodies is probably a critical factor toontrol H. parasuis infection [1] and the exposure to a low dose

∗ Corresponding author at: Campus UAB – Edifici CReSA, 08193 Cerdanyola delallès (Bellaterra), Barcelona, Spain. Tel.: +34 93 581 45 67; fax: +34 93 581 44 90.

E-mail address: alex.olvera@cresa.uab.cat (A. Olvera).1 These authors contributed equally to this work.

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2011.01.105

future vaccine formulations against H. parasuis.© 2011 Elsevier Ltd. All rights reserved.

of the bacterium protects against a subsequent challenge [4,5].Maternally derived antibodies also have an important role in pro-tection, allowing the colonization of piglets and the development ofa protective immune response without the induction of disease [6].Animals protected against H. parasuis infection after immunizationwith a bacterin may develop antibodies against outer membraneproteins (OMP). However, animals immunized with purified OMPare only partially protected [7,8]. Immunogens for a vaccine againsta broad range of H. parasuis strains have been sought and severalantigenic or immunogenic H. parasuis proteins have been described(FhuA, TbpB, OmpA, PalA, Omp2, D15, VtaA, OppA, YfeA, CsgG andPlpA and HPS06257). However, no protection data in the naturalhost has been provided, with the exception of TbpB which resultedin non-protection [9–13].

In a previous study, the passenger domains of 6 trimeric auto-transporters (VtaA1, 5, 6, 8, 9, 10) of the virulent serovar 5 Nagasakistrain were expressed and found to be antigenic [11]. In the currentwork, we evaluated the ability of a vaccine containing these 6 VtaAproteins to protect against an intranasal challenge with H. para-

suis Nakasaki. VtaA belong to the trimeric autotransporters (AT-2)protein family of Gram-negative bacteria [15], and were selectedas candidate antigens to be included in future vaccine formulationsdue to three main reasons: (i) VtaA are expressed by H. parasuis andcan stimulate a immune response in pigs [11]; (ii) VtaA passenger

2 cine 29

dr[tirra

2

2

oed(lpvnppaTba2itwcp1LHtwaepla

2

cd8aerbnswiaioctc

798 A. Olvera et al. / Vac

omains show an extensive mosaic structure and serum cross-eactivity among VtaA from different strains has been described11,14]; and (iii) AT-2 are outer membrane multifunctional proteinshat mediate virulence mechanisms [16–19] with demonstratedmmunogenic and protective properties in different proteobacte-ia using animal models [20–26]. Here, the immunigenicity of 6ecombinant VtaA and their capacity to confer protection to pigsgainst a H. parasuis lethal challenge is reported.

. Material and methods

.1. Recombinant production of VtaA passenger domains

In a previous study, the passenger domains of 6 vtaA genesf the serovar 5 Nagasaki strain (vtaA1, 5, 6, 8, 9 and 10) werexpressed and found to be antigenic. The six vtaA passengeromains were cloned [11] into the pASK-IBA33plus plasmid systemIBA–BioTAGnologies, Göttingen, Germany). To overcome prob-ems of low expression, the translocator domains and signaleptides of the 6 vtaA were excluded and the passenger domain oftaA10 was subcloned into two contiguous fragments. For recombi-ant protein production, transformed Escherichia coli Rosetta (DE3)LysS (Novagen–Merck Chemicals Ltd., Nottingham, UK) werelated onto LB agar supplemented with carbenicillin (50 �g/ml)nd chloramphenicol (34 �g/ml) and grown overnight at 37 ◦C.hen, one colony was inoculated into 25 ml of LB broth with car-enicillin (50 �g/ml) and chloramphenicol (34 �g/ml), incubatedt 37 ◦C (225 rpm) up to an OD600 of 0.4–0.5 and left overnight at5 ◦C without shaking. After centrifugation, bacteria were used to

noculate 250 ml of the above mentioned media. When the cul-ure reached an OD600 of 0.4–0.6, recombinant protein productionas induced for 4 h at 25 ◦C by adding 0.2 �g/ml of anhydrotetra-

ycline (AHT). Bacteria were recovered by centrifugation and theellets were solubilized with 8 M Urea, Na2HPO4 100 mM, Tris–HCl0 mM buffer and 100 nM of pefabloc (Fluka – Sigma–Aldrich, St.ouis, MO, USA). Recombinant proteins were purified using theisGraviTrap kit (GE Healthcare, Chalfont St. Giles, UK) following

he manufacturer’s instructions. The purified recombinant proteinsere dialyzed against decreasing urea concentrations (4 and 2 M)

nd then PBS using Slide-A-Lyzer® cassettes (Pierce-Thermo Sci-ntific, Rockford, IL, USA). Proteins were quantified with the BCArotein assay kit (Pierce-Thermo Scientific, Rockford IL, USA) fol-

owing manufacturer’s instructions. Proteins were then aliquotednd stored at −80 ◦C until use.

.2. Experimental design

All procedures involving animals were approved by the Ethi-al and Animal Welfare Committee from the Universitat Autònomae Barcelona and followed EU normative (Council Directive6/609/EEC). Piglets used in this study came from a farm withstandard health status in Spain with no reported Glässer’s dis-

ase outbreaks. Briefly, sows were seropositive against porcineeproductive and respiratory virus (PRRSV) and swine influenza,ut seronegative against Aujeszky’s disease virus. To avoid mater-al antibody interference and colonization by H. parasuis from theows, 39 snatch-farrowed, colostrum-deprived (SF-CD) pigs [27]ere used. At 22 days of life, piglets were randomized and divided

nto immunized and control groups. Groups were balanced for sexnd weights. Groups 1 (N = 15) and 2 (N = 14) were intramuscularly

mmunized, in both sides of the neck, with approximately 50 �gf VtaA 1, 5, 6, 8 and 9. For VtaA10 around 25 �g of 5′ and 3′ sub-lone products were used. Thus, each animal was immunized forhe primo-vaccination with a total of 300 �g recombinant VtaA andomplete Freund’s adjuvant. As negative controls group 3 (N = 4)

(2011) 2797–2802

and group 4 (N = 6) were mock immunized with PBS and completeFreund’s adjuvant. All animals were boosted 14 days later withthe same formulation, but using incomplete Freund’s adjuvant.Three weeks later groups 1 and 3 were intranasally inoculated with3 × 108 CFU/animal and group 2 and 4 were intranasally inoculatedwith 6 × 106 CFU/animal of the Nagasaki strain of H. parasuis. Theinoculum was made in PBS from overnight cultures on chocolateagar plates at 37 ◦C and 5% CO2. Bacterial counts were confirmedby serial dilution and plating onto chocolate agar.

2.3. Sampling and clinico-pathological evaluation

Body weight, rectal temperature and clinical symptoms compat-ible with Glässer’s disease (cough, depression, abdominal breathand joint tumefaction) were recorded weekly from the first immu-nization (days 7, 14, 21 and 28) and daily after the bacterialchallenge (day 35). After challenge, the infection was monitoredfor 7 days. Animals showing signs of suffering before the end ofthe experiment were euthanized by an intravenous sodium pento-barbital overdose (200 mg/kg, Dolethal, Vetoquinol SA). Survivingpiglets were euthanized at 7 days post-infection (dpi) using thesame product. All pigs were necropsied for evaluation of grosslesions with special focus on those potentially attributable to H.parasuis infection. A lesional score (ranging from 0 to 9) was cal-culated as the sum of individual lesions/signs (lack of lesion = 0;presence of lesion = 1): catarrhal rhinitis, pulmonary consolidation,fibrin in abdomen and/or ascites, fibrin in thorax and/or hydrotho-rax, fluid and/or fibrin in the right elbow, in the left elbow, inthe right knee and in the left knee and meningitis. Since evidenceof meningeal damage is poorly visible at necropsy, samples ofbrain (cerebellum and pons) were taken, fixed by immersion in10% buffered formalin, and routinely processed for histopathology.Mortality associated to H. parasuis infection was confirmed by bac-terial isolation from the lesions. Sera and nasal swabs were takenonce a week before challenge, the day of challenge and at post-mortem evaluation for antibody detection: IgG and IgA in serumsamples and IgA in swab-PBS suspensions. Nasal swab-PBS sus-pensions were also used for H. parasuis-specific PCR detection. Inaddition, broncho-alveolar lavages (BAL) were done immediatelyafter post-mortem examination. Nasal antibodies were recoveredby soaking nasal swabs into 1 ml of PBS (PBS-swab) and vortexing.Samples were stored at −80 ◦C until use.

2.4. Bacteriological evaluation

Blood and swabs from thorax, pericardium, joints, meninges,and peritoneum were taken for bacterial isolation. Bacterial isola-tion was performed by plating the swabs obtained from fluids andorgans within 2 h after necropsy onto chocolate agar and incubat-ing them at 37 ◦C with 5% CO2 for 48 h. H. parasuis recovery wassemi-quantified by assigning a score of 3 to samples with conflu-ent growth, 2 to samples yielding isolated colonies (down to 20colonies), 1 to samples yielding 1 to 19 colonies and 0 to sam-ples yielding no bacterial growth. A bacterial score (ranging from0 to 18) was calculated as the sum of the score from blood, tho-rax, pericardium, abdomen, brain and the maximum score of the 4joints.

Randomly selected bacterial isolates recovered from the lesionsof different animals were confirmed as the inoculated Nagasakistrain by ERIC-PCR [28]. Presence of H. parasuis in nasal and BAL

samples of all animals was also detected using PCR. DNA wasextracted from swab-PBS suspensions and BAL using the Nucle-ospin blood kit (Macherey-Nagel) following the manufacturer’sinstructions. H. parasuis-specific PCR was performed as previouslydescribed [29].

cine 29 (2011) 2797–2802 2799

2

dw1IBO(lpsmbswiw(t

2

sb

3

3i

ittaisriikItmgawv0P

bHtPitrn

ui

A. IgG serum 1:1600

1.0

0.5

1.5

***

*** *** ***

3.0

m

B. IgA serum 1:100

******

0.0

1.0

2.0

OD

45

0 n

**

*

***

***

0 0

0.5

1.0

C. IgA nasal swab 1:1

****

0.0

0.0

7 14 21 28

1st immunization 2nd immunization Challenge

0 35

-0.5

Time (days)

Fig. 1. Antibody response to immunization with VtaA of Haemophilus parasuis,kinetics in serum and nasal swabs. IgG and IgA of piglets immunized with VtaA-Freund adjuvant (N = 4, squares) were measured by ELISA with plates coated with the6 VtaA used in the immunization. The antibody response of piglets immunized withPBS-Freund adjuvant (N = 15, diamonds) is also included as negative control. The

A. Olvera et al. / Vac

.5. Antibody response

An in house ELISA to detect antibodies against the VtaA waseveloped as indicated in [30]. High binding plates were coatedith a total of 0.5 �g/well of the 6 antigenic VtaA (1, 5, 6, 8, 9 and

0) and blocked with 5% skim milk. Sera were diluted 1:1600 forgG detection and 1:100 for IgA detection. PBS from nasal swabs andAL were used undiluted. Goat anti-porcine IgG HRP (AbD Serotec,xford, UK) was diluted 1:100,000 and goat anti-porcine IgA HRP

AbD Serotec, Oxford, UK) was diluted 1:10,000. For immunoglobu-in IgA secretory component detection, PBS from nasal swabs wererocessed as above but revealed with a mouse anti-porcine IgAecretory component (AbD Serotec, Oxford, UK) and a goat antiouse IgG (multi species adsorbed) HRP (AbD Serotec, Oxford, UK),

oth diluted 1:100. Antibodies were detected with the HRP sub-trate 3,3′,5,5-tetramethylbenzidine and ELISA positive thresholdsere set as the mean value of negative controls (animals mock

mmunized) plus three times its standard deviation. Antibody titersere calculated at the day of challenge as the last positive dilution

1:1600–1:64,000 for IgG and 1:100–1:6400 for IgA) in ELISA usinghe above positive threshold.

.6. Statistical analysis

Mean values were compared using the Student’s t-test with aignificance threshold of P < 0.05. Correlation between lesion andacterial scores was measured using the correlation coefficient.

. Results

.1. Recombinant VtaA induced both systemic and mucosalmmune responses

Antibody response throughout the experiment was measuredn sera and nasal swabs for groups 1 (vaccinated) and 3 (con-rol). VtaA-immunized pigs induced detectable IgG responses afterhe first vaccine dose (Fig. 1A), reaching the higher levels as soons the first week after the second dose of the vaccine. Interest-ngly, a similar kinetic curve was observed for the induction ofystemic anti-VtaA IgA (Fig. 1B). As expected, no specific antibodyesponses were found in control pigs. The VtaA induced mucosalmmune response was confirmed by the presence of specific IgAn the nasal secretions of vaccinated pigs (Fig. 1C), showing similarinetic responses than in blood. The presence of specific secretorygA was also measured, but only on the day of challenge. At thisime, ELISA (OD 450) values were 0.757 ± 0.704 for immunized ani-

als of groups 1 plus 2 and 0.246 ± 0.092 for negative controls ofroup 3 plus 4 (Student’s t-test, P = 0.0292). Also, the presence ofnti-VtaA IgA in the lower respiratory tract of immunized animalsas confirmed in the BAL obtained at necropsy. ELISA (OD 450 nm)

alues were 1.071 ± 0.393 for immunized animals of group 1 and.132 ± 0.007 for negative controls of group 3 (Student’s t-test,= 0.0002).

It is noteworthy, that there was an increase in nasal IgA afteroth challenges in animals that were protected (N = 13) against. parasuis. ELISA (OD 450 nm) values were 0.383 ± 0.231 before

he challenge and 0.754 ± 0.400 seven days later (Student’s t-test,= 0.0079). This second increase on the specific IgA responses

n nasal swabs after the H. parasuis Nagasaki challenge indicateshat the vaccine could prime a mucosal immune response. Such a

esponse after challenge was not observed in the sera antibodies ofon-protected animals.

To see if protected animals had higher antibody levels thannprotected animals at the day of challenge (5 weeks after the first

mmunization), titers of systemic anti-VtaA IgG and IgA antibod-

discontinuous line indicates the positive threshold of the test. Difference statisticalsignificance (Student’s t-test) between control and vaccinated animals is indicated(*0.05 < P < 0.005; **0.005 < P < 0.0005; ***P < 0.0005). Arrows indicate when the first,second immunization and challenge were performed, respectively.

ies were compared (Table 1). All vaccinated animals (groups 1 and2) had developed a specific antibody response against the VtaA inserum (IgG and IgA) compared to controls (groups 3 and 4). Therewere no significant differences (Student’s t-test) in IgG and IgAtiters in serum between survivors and animals that did not surviveinfection, although animals that were protected had slightly highertiters. Nevertheless, pigs protected against the infection had a min-imum titer (−Log10 sera dilution) of 3.6 for IgG and 2.6 for IgA whilein the non-protected it was of 3.0 for IgG and 2.3 for IgA. Besides, thenon-protected animals had a maximum titer (−Log10 sera dilution)of 4.81 for IgG and 3.81 for IgA. Finally, although not statisticallysignificant (Student’s t-test), the IgG and IgA titers in serum wereslightly higher in the trial with the 3 × 108 CFU compared to the6 × 106 CFU challenge model (Table 1).

3.2. Immunization with VtaA partially protects against H.parasuis lethal challenge

In order to characterize the protective potential of VtaA, groupsof vaccinated and control pigs were challenged with two differ-ent lethal doses of the highly virulent strain Nagasaki. The survivalcurves denoted a dose effect of the inoculum used in the challenge,albeit both bacterial inoculums were lethal and all control pigs diedor were euthanized due to severe Glässer’s disease clinical signs.

Control animals infected with 3 × 108 CFU of the Nagasaki strainof H. parasuis died peracutely or were euthanized within 2 days,while control animals infected with 6 × 106 CFU died acutely orwere euthanized within 4 days post-inoculation. In clear contrast,

2800 A. Olvera et al. / Vaccine 29 (2011) 2797–2802

Table 1Mean IgG and IgA titers (−Log10 sera dilution) in sera at the day of infection. Mean lesion scores, bacterial scores and days of survival after the experimental infectionwith H. parasuis are indicated. Mean temperature values on the day before euthanasia are also included. Means were compared using the Student’s t-test with a statisticalsignificance threshold of P < 0.05. Statistical differences with control groups are indicated with *. Statistical differences between vaccinated protected animals and vaccinatednon-protected animals are indicated by #.

Titer (−Log10 sera dilution)

N IgG IgA Lesion score Bacterial score Survival Temperature

108 CFU infectionControl 4 – – 3.0 ± 0.8 12.3 ± 0.5 2.0 ± 0.0 40.8 ± 1.0Vaccinated 15 4.22 ± 0.49 3.08 ± 0.41 4.3 ± 3.1 9.9 ± 6.2 4.7* ± 1.8 40.5 ± 0.9

Non-protected 10 4.20 ± 0.57 3.05 ± 0.45 6.1 ± 1.8 13.6 ± 3.2 3.6 ± 0.7 40.4 ± 0.9Protected 5 4.26 ± 0.33 3.14 ± 0.33 0.6*,# ± 0.9 2.4 *,# ± 2.9 7.0 *,# ± 0.0 40.5 ± 1.0

106 CFU Infection2.51.94.20.3

3lVtdspN

tpactadtCsvana

FiiaPiN

Control 6 – –Vaccinated 14 3.82 ± 0.28 2.90 ± 0.37

Non-protected 6 3.65 ± 0.23 2.75 ± 0.41Protected 8 3.94 ± 0.25 3.02 ± 0.32

3% (5/15) and 57% (8/14) of VtaA-immunized animals survived theethal challenge with 3 × 108 and 6 × 106 CFU of the bacteria (Fig. 2).taA experimental vaccination significantly delayed the death of

he pigs. While none of the pigs within the vaccinated group (0/15)ied before day 2 after challenge with the higher dose of H. para-uis, all control animals succumbed by that time. Similarly, 12/15igs remained alive by day 5 post-challenge with 6 × 106 CFU ofagasaki while no control pigs remained alive (Fig. 2).

According to the survival data, immunization with VtaA pro-ected against acute Glässer disease development and immunizedrotected animals did not show gross lesions. The inoculum hadstrong effect on the rectal temperature and groups 1 and 3

hallenged with 3 × 108 CFU developed fever (corporal tempera-ure > 40.2 ◦C). However, vaccination had a clear effect in groups 2nd 4 challenged with 6 × 106 CFU, and protected animals did notevelop fever. Besides, there was an effect of the time of survival onhe gross lesions observed in the animals that developed disease.ontrol pigs were more frequently affected by acute lesions likeerositis than VtaA vaccinated pigs (Table 2). On the other hand,

accinated animals challenged with 3 × 108 CFU survived longernd were affected more frequently by arthritis. Interestingly, vacci-ated animals challenged with 6 × 106 CFU showed a reduction ofll systemic lesions compared to their controls. In both challenges

80

100 3 x 108 CFUN= 15

N= 4

0

20

40

60

40

60

80

100

% S

urv

ivin

g a

nim

als

6 x 106 CFUN= 14N= 6

0

20

76543210

Day post-infection

ig. 2. Survival of snatch-farrowed colostrum-deprived piglets after experimentalntranasal inoculation of the Nagasaki strain of Haemophilus parasuis. Mortal-ty associated to H. parasuis infection was studied by post-mortem examinationnd confirmed by bacterial isolation from the lesions. Piglets immunized withBS-Freund adjuvant (diamonds) or with VtaA-Freund adjuvant (squares) werenoculated with 3 × 108 CFU (upper panel) or 6 × 106 CFU (lower panel) of theagasaki strain.

± 2.7 7.3 ± 4.1 2.8 ± 0.8 40.2 ± 1.3± 2.1 4.6 ± 5.6 6.1* ± 1.3 40.1 ± 0.8± 1.0 10.7 ± 1.6 4.8 ± 1.0 40.9 ± 0.3# ± 1.0 0.0*,# ± 0.0 7.0*,# ± 0.0 39.4# ± 0.3

vaccinated animals showed a greater number of lesions in therespiratory tract than control animals; pulmonary craneo-ventralconsolidation was seen only in the vaccinated animals.

Lesion and bacterial scores correlated (correlation coefficientof 80.8%), as expected; the effect of vaccination on these scoreshas been summarized in Table 1. There were no significant dif-ferences (Student’s t-test) when mean lesion or bacterial scoresfrom controls and vaccinated animals were compared, althoughthere was a reduction on the mean scores when the animals werechallenged with 6 × 106 CFU. There were statistically significant dif-ferences (Student’s t-test) between protected (vaccinated) animalsand non-protected animals (from vaccinated or controls groups).As expected, protected vaccinated animals had very low lesionand bacterial scores. In both challenges lesion and bacterial scoresincreased with the time of survival in non-protected vaccinated ani-mals, due to an increase in the frequency of polyarthritis. In fact,15/16 (94%) of the vaccinated, but not protected, animals had bac-teria in at least one joint while in control animal’s bacteria wasrecovered from joints in only 5/10 (50%) of the animals. Bacte-ria were recovered from joints in only 1/13 (8%) of the vaccinatedprotected animals (challenged with 3 × 108 CFU). Interestingly, thechallenge strain was not detected by PCR in nasal or lung samplesof animals protected against Glässer’s disease. Eighteen isolatesrecovered after experimental infection were randomly selected andconfirmed as the Nagasaki strain by their ERIC-PCR pattern.

4. Discussion

The aim of this study was to assess the antibody response andprotective capacity of VtaA against an experimental H. parasuisinfection, using purified recombinant proteins. VtaA are AT-2 ofthe type V bacterial secretion system and, to the authors’ knowl-edge, this is the first study where these molecules are shown tobe immunogenic and protective in the context of a pathogen andits natural host. Besides, this is the first report demonstrating thatrecombinant proteins can protect pigs against Glässer’s disease.

The Nagasaki strain (serovar 5) used in the challenge is wellknown for its high virulence in experimental conditions [31–34].Moreover, several virulence mechanisms, such as serum resistance,phagocytosis resistance and invasion of endothelial cells, have beendescribed for this strain [35–38]. Also, it has been described that theNagasaki strain can cause a severe endotoxic shock [39]. This highvirulence explains why the control animals died in a short period

of time and accounts for the dose effect of the challenge inocu-lums. Generally, animals infected with 3 × 108 CFU survived fewerdays and showed higher bacterial scores than animals infected with6 × 106 CFU. Due to the rapid disease onset and death, control ani-mals from the high dose challenge had less gross lesions, mainly

A. Olvera et al. / Vaccine 29 (2011) 2797–2802 2801

Table 2Numbers of animals showing the different lesions attributable to Haemophilus parasuis infection, percentages are also indicated.

108 CFU infection 106 CFU infection

Control Vaccinated Control Vaccinated

Catarrhal rhinitis 2/4 (50%) 3/15(20%) 0/6 (0%) 1/14 (7%)Pulmonary consolidation 0/4 (0%) 3/15(20%) 0/6 (0%) 3/14 (21%)Fibrin in abdomen/ascites 3/4 (75%) 8/15(53%) 4/6 (67%) 4/14 (29%)

adnGcdoaa

csarbAvittcDpaatdbnepbnsryeait

dgrAtbp1tFicd

[

Fibrin in thorax/Hydrothorax 3/4 (75%)Meningitis 0/4 (0%)Arthritis 2/4 (50%)

rthritis, and lower lesion scores. Proportional to the challengeoses, vaccinated animals had an increased survival. Vaccinatedon-protected animals showed the characteristic clinical signs oflässer’s disease, but their onset was delayed in comparison to theontrol pigs. Also, they were more severely affected by polyarthritisue to longer disease progression. Interestingly, in the conditionsf this experiment, about one third of the animals were protectedgainst a high inoculum (3 × 108 CFU) and a more than a half againstlow inoculum (6 × 106 CFU).

In a previous study, anti-VtaA antibodies were induced in a sub-linical experimental infection with a H. parasuis [11]. In the presenttudy all vaccinated animals developed antibodies, although notll animals were protected against the infection. To explore theeasons for individual differences in protection, antibody titersetween protected and non-protected animals were compared.ll animals immunized with the recombinant VtaAs serocon-erted and developed also a mucosal IgA response after the firstmmunization, indicating that VtaA are good immunogens. In addi-ion, before the challenge, there were no differences in antibodyiters between the two experimental infections or between vac-inated animals that subsequently died or survived the infection.espite the overlap of antibody levels between protected and non-rotected animals, the minimum antibody titer of the protectednimals was higher. This could indicate the existence of a minimalnti-VtaA antibody level needed to have protection using these pro-eins. Interestingly, a good mucosal humoral immune response wasetected in the respiratory tract (nasal cavity and lung). Also, anti-ody memory recall could have been induced since an increase ofasal IgA was detected in protected animals after infection. How-ver, to fully demonstrate the existence antibody memory, strongrimary response to live bacteria infection should be discardedy confirming the absence of nasal anti-VtaA IgA in challengedon-vaccinated controls. Unfortunately, non-vaccinated controlsuccumbed too early to be able to build an immune response. Theseesults also indicate that intramuscular immunization in the neckield a good mucosal humoral immunity in swine, likely due to thexistence of lymphatic nodes connecting different tissues. Thus,lthough antibody response and probably memory recall werenduced in serum and in the respiratory tract it was not enougho protect all the animals.

The partial protection observed could rely on individual geneticifferences in the resistance against bacteria and the capacity toenerate bactericidal or opsonizing antibodies. Also, the cellularesponse was not measured and could play a role in protection.lternatively, other H. parasuis antigens could be important for pro-

ection. Besides, failure to obtain full protection could be explainedy bacterial regulation of the expression of VtaA. It should beointed out that the Nagasaki strain has the capacity to produce3 different VtaA [14] and only the 6 of them, which were found

o be antigenic, were used in the present study as immunogens.ailure to produce cross-reacting antibodies against the remain-ng VtaA might also explain partial protection results. Several vtaAopies per genome (between 9 and 15) with an extensive passengeromain repertoire generated by recombination have been detected,

[

[

6/15(40%) 4/6 (67%) 4/14 (29%)8/15(53%) 2/6 (33%) 1/14 (7%)12/15(67%) 2/6 (33%) 4/14 (29%)

which is compatible with an immune escape by antigenic switch-ing. However, sera of animals subclinically infected with H. parasuisshowed cross-reactivity with VtaA of other strains [11].

Finally, limitations due to the disease model used should be keptin mind. It is difficult to adjust a dose to reproduce the disease andthis change among strains. For the Nagasaki strain a dose of approx-imately 106 CFU did not cause disease, but the slightly higher doseused here (6 × 106 CFU) reproduced an acute form of the disease.

In conclusion, VtaA passenger domains from VtaA1, 5, 6, 8, 9and 10 were immunogenic, but antibodies to them were not fullyprotective against the severe experimental challenge used in thisstudy.

Acknowledgements

This work was funded by the Comisión Interministerial de Cien-cia y Tecnología (AGL2007-60432) of the Spanish government andthe ACC1Ó program of the Generalitat de Catalunya (VALTEC08-1-0005). AO is funded by a Juan de la Cierva fellowship of theMinisterio de Ciencia e Innovación of the Spanish government. Theauthors would like to thank the animal-care tacking staff of CReSAfor its excellent job and Fernando Rodriguez for critically reviewingthis manuscript.

References

[1] Rapp-Gabrielson V, Oliveira S, Pijoan C. Haemophilus parasuis. nineth ed. Iowa:Iowa State University Press; 2006.

[2] Kielstein P, Rapp-Gabrielson VJ. Designation of 15 serovars of Haemophilusparasuis on the basis of immunodiffusion using heat-stable antigen extracts.J Clin Microbiol 1992;30(4):862–5.

[3] Miniats OP, Smart NL, Ewert E. Vaccination of gnotobiotic primary spe-cific pathogen-free pigs against Haemophilus parasuis. Can J Vet Res1991;55(1):33–6.

[4] Nielsen R. Pathogenicity and immunity studies of Haemophilus parasuisserotypes. Acta Vet Scand 1993;34(2):193–8.

[5] Oliveira S, Batista L, Torremorell M, Pijoan C. Experimental colonization ofpiglets and gilts with systemic strains of Haemophilus parasuis and Streptococcussuis to prevent disease. Can J Vet Res 2001;65(3):161–7.

[6] Solano-Aguilar GI, Pijoan C, Rapp-Gabrielson V, Collins J, Carvalho LF, Winkel-man N. Protective role of maternal antibodies against Haemophilus parasuisinfection. Am J Vet Res 1999;60(1):81–7.

[7] Martin de la Fuente AJ, Rodriguez-Ferri EF, Frandoloso R, Martinez S,Tejerina F, Gutierrez-Martin CB. Systemic antibody response in colostrum-deprived pigs experimentally infected with Haemophilus parasuis. Res Vet Sci2009;86(2):248–53.

[8] Miniats OP, Smart NL, Rosendal S. Cross protection among Haemophilus parasuisstrains in immunized gnotobiotic pigs. Can J Vet Res 1991;55(1):37–41.

[9] del Rio ML, Navas J, Martin AJ, Gutierrez CB, Rodriguez-Barbosa JI, RodriguezFerri EF. Molecular characterization of Haemophilus parasuis ferric hydroxam-ate uptake (fhu) genes and constitutive expression of the FhuA receptor. VetRes 2006;37(1):49–59.

10] Hong M, Ahn J, Yoo S, Hong J, Lee E, Yoon I, et al. Identification of novel immuno-genic proteins in pathogenic Haemophilus parasuis based on genome sequenceanalysis. Vet Microbiol 2010;148(1):89–92.

11] Olvera A, Pina S, Perez-Simo M, Oliveira S, Bensaid A. Virulence-associatedtrimeric autotransporters of Haemophilus parasuis are antigenic proteinsexpressed in vivo. Vet Res 2010;41(3):26.

12] Zhang B, Tang C, Yang FL, Yue H. Molecular cloning, sequencing and expres-sion of the outer membrane protein A gene from Haemophilus parasuis. VetMicrobiol 2009;136(3–4):408–10.

2 cine 29

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

802 A. Olvera et al. / Vac

13] Zhou M, Guo Y, Zhao J, Hu Q, Hu Y, Zhang A, et al. Identification and char-acterization of novel immunogenic outer membrane proteins of Haemophilusparasuis serovar 5. Vaccine 2009;27(38):5271–7.

14] Pina S, Olvera A, Barcelo A, Bensaid A. Trimeric autotransporters of Haemophilusparasuis: generation of an extensive passenger domain repertoire specific forpathogenic strains. J Bacteriol 2009;191(2):576–87.

15] Cotter SE, Surana NK, St Geme 3rd JW. Trimeric autotransporters: a distinctsubfamily of autotransporter proteins. Trends Microbiol 2005;13(5):199–205.

16] Leduc I, Olsen B, Elkins C. Localization of the domains of the Haemophilusducreyi trimeric autotransporter DsrA involved in serum resistance and bindingto the extracellular matrix proteins fibronectin and vitronectin. Infect Immun2009;77(2):657–66.

17] Scarselli M, Serruto D, Montanari P, Capecchi B, Adu-Bobie J, Veggi D, et al. Neis-seria meningitidis NhhA is a multifunctional trimeric autotransporter adhesin.Mol Microbiol 2006;61(3):631–44.

18] Serruto D, Spadafina T, Scarselli M, Bambini S, Comanducci M, Hohle S,et al. HadA is an atypical new multifunctional trimeric coiled-coil adhesin ofHaemophilus influenzae biogroup aegyptius, which promotes entry into hostcells. Cell Microbiol 2009;11(7):1044–63.

19] Sjolinder H, Eriksson J, Maudsdotter L, Aro H, Jonsson AB. Meningococcal outermembrane protein NhhA is essential for colonization and disease by preventingphagocytosis and complement attack. Infect Immun 2008;76(11):5412–20.

20] Aebi C, Maciver I, Latimer JL, Cope LD, Stevens MK, Thomas SE, et al. A protectiveepitope of Moraxella catarrhalis is encoded by two different genes. Infect Immun1997;65(11):4367–77.

21] Bowe F, Lavelle EC, McNeela EA, Hale C, Clare S, Arico B, et al. Mucosal vacci-nation against serogroup B meningococci: induction of bactericidal antibodiesand cellular immunity following intranasal immunization with NadA of Neis-seria meningitidis and mutants of Escherichia coli heat-labile enterotoxin. InfectImmun 2004;72(7):4052–60.

22] Durant L, Metais A, Soulama-Mouze C, Genevard JM, Nassif X, Escaich S. Identi-fication of candidates for a subunit vaccine against extraintestinal pathogenicEscherichia coli. Infect Immun 2007;75(4):1916–25.

23] Forsgren A, Brant M, Riesbeck K. Immunization with the truncated adhesinMoraxella catarrhalis immunoglobulin D-binding protein (MID764-913) is pro-tective against M. catarrhalis in a mouse model of pulmonary clearance. J Infect

Dis 2004;190(2):352–5.

24] Giuliani MM, Adu-Bobie J, Comanducci M, Arico B, Savino S, Santini L, et al.A universal vaccine for serogroup B meningococcus. Proc Natl Acad Sci USA2006;103(29):10834–9.

25] Mason KW, Zhu D, Scheuer CA, McMichael JC, Zlotnick GW, Green BA. Reduc-tion of nasal colonization of nontypeable Haemophilus influenzae following

[

[

(2011) 2797–2802

intranasal immunization with rLP4/rLP6/UspA2 proteins combined with aque-ous formulation of RC529. Vaccine 2004;22(25–26):3449–56.

26] Meier PS, Freiburghaus S, Martin A, Heiniger N, Troller R, Aebi C. Mucosalimmune response to specific outer membrane proteins of Moraxella catarrhalisin young children. Pediatr Infect Dis J 2003;22(3):256–62.

27] Blanco I, Galina-Pantoja L, Oliveira S, Pijoan C, Sanchez C, Canals A. Com-parison between Haemophilus parasuis infection in colostrums-deprived andsow-reared piglets. Vet Microbiol 2004;103(1–2):21–7.

28] Oliveira S, Blackall PJ, Pijoan C. Characterization of the diversity of Haemophilusparasuis field isolates by use of serotyping and genotyping. Am J Vet Res2003;64(4):435–42.

29] Olvera A, Pina S, Macedo N, Oliveira S, Aragon V, Bensaid A. Identi-fication of potentially virulent strains of Haemophilus parasuis using amultiplex PCR for virulence-associated autotransporters (vtaA). Vet J 2011,doi:10.1016/j.tvjl.2010.12.014.

30] Ausubel F, Brent R, Kingston R, Moore D, Seidman J, Smith J, et al. Currentprotocols in molecular biology. New York: John Wiley and Sons; 1994.

31] Amano H, Shibata M, Kajio N, Morozumi T. Pathologic observations of pigsintranasally inoculated with serovar 1, 4 and 5 of Haemophilus parasuis usingimmunoperoxidase method. J Vet Med Sci 1994;56(4):639–44.

32] Amano H, Shibata M, Kajio N, Morozumi T. Pathogenicity of Haemophilus para-suis serovars 4 and 5 in contact-exposed pigs. J Vet Med Sci 1996;58(6):559–61.

33] Aragon V, Cerda-Cuellar M, Fraile L, Mombarg M, Nofrarias M, Olvera A, et al.Correlation between clinico-pathological outcome and typing of Haemophilusparasuis field strains. Vet Microbiol 2010;142(3–4):387–93.

34] Rapp-Gabrielson VJ, Gabrielson DA, Schamber GJ. Comparative virulenceof Haemophilus parasuis serovars 1 to 7 in guinea pigs. Am J Vet Res1992;53(6):987–94.

35] Aragon V, Bouchet B, Gottschalk M. Invasion of endothelial cellsby systemic and nasal strains of Haemophilus parasuis. Vet J 2009,doi:10.1016/j.tvjl.2009.08.013.

36] Cerdà-Cuéllar M, Aragon V. Serum-resistance in Haemophilus parasuis is asso-ciated with systemic disease in swine. Vet J 2008;175(3):384–9.

37] Olvera A, Ballester M, Nofrarias M, Sibila M, Aragon V. Differences in phagocy-tosis susceptibility in Haemophilus parasuis strains. Vet Res 2009;40(3):24.

38] Vanier G, Szczotka A, Friedl P, Lacouture S, Jacques M, Gottschalk M.Haemophilus parasuis invades porcine brain microvascular endothelial cells.Microbiology 2006;152(Pt 1):135–42.

39] Amano H, Shibata M, Takahashi K, Sasaki Y. Effects on endotoxin pathogenicityin pigs with acute septicemia of Haemophilus parasuis infection. J Vet Med Sci1997;59(6):451–5.