International Journal of Infectious Diseases 16 (2012) e816–e825

Immunogenicity and protective efficacy of heterologous prime-boost regimenswith mycobacterial vaccines and recombinant adenovirus- and poxvirus-vectoredvaccines against murine tuberculosis

Qingrui You a, Yongge Wu a, Yang Wu a, Wei Wei a, Changyong Wang a, Dehua Jiang a,Xianghui Yu a, Xizhen Zhang a, Yong Wang b, Zhijiao Tang b, Chunlai Jiang a,*, Wei Kong a,*a National Engineering Laboratory for AIDS Vaccine, School of Life Science, Jilin University, Gaoxin District Changchun, Jilin, Chinab ABSL-3 Lab, Wuhan University, 115 Donghu Road, Wuchang District Wuhan, Hubei, 430071, China

A R T I C L E I N F O

Article history:

Received 15 March 2012

Accepted 3 July 2012

Corresponding Editor: William Cameron,Ottawa, Canada

Keywords:

Tuberculosis

MVA (modified vaccinia virus Ankara)

Adenovirus

Boost vaccine

S U M M A R Y

Objectives: To evaluate regimens using bacillus Calmette–Guérin (BCG) or recombinant BCG (rBCG)

overexpressing Ag85B for priming, followed by boosting with a modified vaccinia virus Ankara strain

(MVA) and/or adenovirus vector (AD) expressing an Ag85B–ESAT6 fusion protein.

Methods: Cellular and humoral immune responses were determined after subcutaneous vaccination,

which was employed to trigger systemic immunity against intravenous infection in a mouse model of

tuberculosis (TB). Bacterial loads and lung histology were evaluated.

Results: The relative IgG2a and IgG1 antibody levels indicated that the viral-vectored vaccines generated

a T-helper type 1 (Th1)-biased response after two doses of viral boost vaccinations. Boosting BCG-primed

mice with viral vaccines induced a Th1 immune response that included both CD4 and CD8 T-cells

generating antigen-specific interferon-gamma (IFN-g) and CD8 T cytotoxic activity. Only micevaccinated with two different viral boosters after BCG priming exhibited a significant reduction in

bacterial burden in the lung after challenge. Histology examinations confirmed the attenuation of lung

damage and more compact granulomas. After mycobacteria priming, boosting with AD85B-E6 followed

by MVA85B-E6 afforded better protection than the reverse order of administration of the viral vectors.

Conclusions: This study demonstrates the potential of multiple heterologous viral booster vaccines,

although the exact correlates of protection and optimal regimens should be further investigated for the

rational design of future vaccine strategies.

� 2012 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.

Contents lists available at SciVerse ScienceDirect

International Journal of Infectious Diseases

jou r nal h o mep ag e: w ww .e lsev ier . co m / loc ate / i j id

1. Introduction

Approximately a third of the world’s population is believed tohave been infected with Mycobacterium tuberculosis, the causativebacterial agent of tuberculosis (TB). In 2010, there were anestimated 8.8 million incident cases of TB and almost 1.5 milliondeaths from TB.1 Although this airborne infectious disease can betreated with the directly observed treatment short-course (DOTS)strategy, the cost of the protracted drug treatment may be difficultto afford in some parts of the world. In addition, the emergence ofdrug-resistant M. tuberculosis has further increased the difficultiesof clinical treatment. Hence, effective implementation ofvaccination programs would represent the most cost-effectiveway to curb the global TB epidemic.2

* Corresponding author. Tel.: +86 431 8516 7790; fax: +86 431 8519 5516.

E-mail addresses: jiangcl@jlu.edu.cn (C. Jiang), chysx6@163.com (W. Kong).

1201-9712/$36.00 – see front matter � 2012 International Society for Infectious Diseahttp://dx.doi.org/10.1016/j.ijid.2012.07.008

The only anti-TB vaccine currently licensed, Mycobacteriumbovis bacillus Calmette–Guérin (BCG), has been used in humansworldwide since 1921, but its efficacy is somewhat controversial.Although BCG is effective in protecting from severe forms ofchildhood TB, principally miliary disease and meningitis, it hasfailed to prevent the adult pulmonary TB epidemic, with a widerange of efficacy (0 to 80%). A variety of BCG strains are used indifferent geographical locations, and it is believed that protectioninduced by BCG wanes within 10 to 15 years, along withdiminishing protective T-cell immunity.3 Unlike extracellularbacterial infections which are effectively eliminated byantibody-inducing vaccines, M. tuberculosis is an intracellularpathogen that requires, at least in part, strong cellular immuneresponses for protective immunity. Thus, there is an urgent need todevelop improved TB vaccines and effective vaccination strategiesthat are capable of promoting long-term cellular immunity.

Since BCG revaccination does not provide enhanced protectionand can even be deleterious,4–6 a heterologous prime-boostimmunization strategy represents an ideal way to extend

ses. Published by Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.ijid.2012.07.008mailto:jiangcl@jlu.edu.cnmailto:chysx6@163.comhttp://www.sciencedirect.com/science/journal/12019712http://dx.doi.org/10.1016/j.ijid.2012.07.008

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825 e817

BCG-initiated immunity and improve the protective efficacy.7–9 Todate, several types of TB vaccines, such as protein adjuvantformulations10–12 and viral-based13–16 and plasmid DNAvaccines,17 have been used as boosters, and some combinationshave achieved encouraging results. In particular, recombinant viralvectors, especially adenovirus- and poxvirus-based TB vaccines,which can induce both cellular and humoral responses with strongimmunogenicity and good safety records,18–21 are highly promis-ing candidates. Furthermore, respiratory tract vaccination withrecombinant viral carriers has the potential to defend against thismucosal infectious disease.22

BCG provides thousands of antigens and stimulates the immunesystem for prolonged periods of time as a live-attenuated vaccine,in contrast to subunit vaccines based on one or a limited number ofantigens. Accordingly, attempts have been made to geneticallymodify BCG to improve its immunogenicity.23–25 Among themycobacterial organism-based candidates, recombinant BCG 30(rBCG30), which expresses high levels of Ag85B, has demonstratedimproved immunogenicity and protective efficacy over thetraditional BCG in animals and has already entered clinical trialswith a well-established safety profile.26–28 Thus, a combinationvaccine comprised of an rBCG vaccine followed by subsequentviral-vectored boosters is a logical approach for enhancing theprotective efficacy against TB.

In our previous work, a recombinant bivalent poxvirus-basedvector, MVA85B-E6, and an adenovirus-based vector, AD85B-E6,both of which express the fusion protein Ag85B–ESAT6, wereengineered and evaluated as standalone vaccines.29 In the currentstudy, we investigated whether these recombinant viral vaccinescould boost conventional BCG-induced immunity and provideenhanced protection. Furthermore, we investigated their effects asbooster vaccines after priming with rBCG overexpressing Ag85B.We show that the protective efficacy of mycobacterial vaccine-primed mice was not improved by one dose of either viral booster,while two boosters afforded enhanced protection in the lungs.Furthermore, only the rBCG-primed mice receiving AD85B-E6 andsubsequent MVA85B-E6 exhibited enhanced protection againstM. tuberculosis H37Rv challenge. This study provides support formultiple viral boosting for mycobacterial vaccination and contrib-utes to the future rational design of immunization strategiesagainst TB.

2. Materials and methods

2.1. Recombinant proteins and peptides

Ag85B, ESAT6, and TB10.4 were expressed and purified with theNi-NTA Purification System (Invitrogen) under denaturing condi-tions as recombinant C-terminal 6 � His-tagged proteins, asdescribed previously.30 Based on CD4 and CD8 T-cell epitopemapping, H-2d-restricted class I peptides (9-2 peptide, Ag85B:70–78, MPVGGQSSF; 9-1 peptide, Ag85B: 144–152, IYAGSLSAL) andclass II peptides (18-1 peptide, Ag85B: 100–117, FLTSELPQWLSAN-RAVKP; 18-2 peptide, Ag85B: 262–279, HSWEYWGAQLNAMKGDLQ;20 peptide, ESAT6: 1–20, MTEQQWNFAGIEAA-ASAIQG) wereused.31–35 All proteins and peptides were diluted in phosphate-buffered saline (PBS) at a final concentration of 1 mg/ml and werestored in aliquots at �20 8C until used.

2.2. Preparation of Mycobacterium

M. bovis BCG and M. tuberculosis H37Rv were cultured in DifcoMiddlebrook 7H9 broth supplemented with 10% BBL Middlebrookalbumin–dextrose–catalase (ADC) enrichment (BD Pharmingen) at37 8C. After 3–4 weeks, the bacteria were harvested, washed twicewith PBS, resuspended in PBS containing 10% glycerol, and stored

in aliquots at �80 8C until used. rBCG overexpressing Ag85B ofM. tuberculosis was generated by electrotransformation of BCGwith a plasmid, pSMT3-85B, which carries a hygromycin resistancemarker for selection. The same procedure above was used tohandle rBCG except for additional selection with 50 mg/ml ofhygromycin. Bacterial titers expressed as colony-forming units(CFU) were determined by plating 10-fold serial dilutions ontoMiddlebrook 7H11 agar plates containing 0.5% glycerol andoleic acid–albumin–dextrose–catalase (OADC) enrichment (BDPharmingen).

2.3. Characterization of rBCG

The expression and secretion of Ag85B from rBCG in culturewere verified by Tricine-sodium dodecyl sulfate-polyacrylamidegel electrophoresis (Tricine-SDS-PAGE) and Western blotting. BCGand rBCG were seeded at 1 � 109 CFU and cultured at 37 8C for3 weeks. The bacteria were harvested by centrifugation at 3000 � gfor 10 min for determining the CFU, and then the culturesupernatants were obtained by passage through a 0.45 mm filterto remove the bacteria. The culture filtrate proteins wereprecipitated using ice-cold acetone, and the concentrated proteinsequivalent to those secreted by 1 � 108 bacteria in culture wereseparated by electrophoresis on a polyacrylamide gel. The bands inthe digitized image were analyzed by assigning optical density(OD) values using Bandscan 5.0. Plasmid stability in rBCG wasevaluated by subculturing the strain (1:20) six consecutive timesfor 3 weeks each in the absence of hygromycin. Ag85B-relatedcellular immune responses induced by rBCG were verified byperforming interferon-gamma (IFN-g) enzyme-linked immuno-spot (ELISPOT) assays with Ag85B or an irrelevant antigen TB10.4.

2.4. Tricine-SDS-PAGE and Western blotting

Standard Tricine-SDS-PAGE was performed using a 12%polyacrylamide separation gel and 4% stacking gel. After semi-dry transfer onto nitrocellulose membranes, Western blotting wasperformed with rabbit anti-M. tuberculosis Ag85B polyclonalantibody (Abcam). Alkaline phosphatase-conjugated goat anti-rabbit IgG (Jackson Immunoresearch) was used as the secondaryantibody and staining was carried out with 5-bromo-4-chloro-3-indolylphosphate (BCIP) and nitroblue tetrazolium (NBT)solutions.

2.5. Recombinant viruses

Recombinant MVA85B-E6 and AD85B-E6 were prepared asdescribed previously29 and diluted in PBS for vaccinations.

2.6. Mice, vaccination, and mycobacterial infection

Specific pathogen-free female BALB/c (H-2d) mice (6–8 weeksold) were used. Uninfected mice were kept in animal facilitiesunder specific pathogen-free conditions. Two sets of animalexperiments were carried out in this study, each with six groupsof mice (n = 12). In experiment 1, BCG (designated as ‘B’) wasdiluted in PBS and administered subcutaneously at a dose of1 � 106 CFU, while PBS alone as a control and viral boosters weregiven subcutaneously at a dose of 2 � 107 plaque-forming units(PFU) for MVA85B-E6 (designated as ‘M’) or 1 � 107 PFU forAD85B-E6 (designated as ‘A’), as outlined in Figure 1. Humoralresponses were determined 1 week after the final vaccination. Twoweeks after the final vaccination, five mice in each group weresacrificed for evaluation of cellular-mediated immune responses.After an additional 2 weeks, the remaining mice were challengedby the intravenous route with 5 � 105 CFU of M. tuberculosis in a

Figure 1. Schematic diagram of the experimental design. In experiment 1, mice(12 per group) were vaccinated at the indicated times and sacrificed at 12 weeks

(n = 5) or 22 weeks (n = 5) after the first vaccination. Four weeks after the last

vaccination, mice were challenged with M. tuberculosis by the intravenous route at a

dose of 5 � 105 CFU. BCG (1 � 106 CFU, B), MVA85B-E6 (2 � 107 PFU, M), AD85B-E6(1 � 107 PFU, A), or PBS (naı̈ve) were given as indicated. Humoral responses weredetermined 1 week after the final vaccination, while colony-forming assays and

histological analyses were performed at the time of sacrifice. In experiment 2, the

same procedure was carried out except that rBCG was used for priming instead of BCG.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825e818

biosafety level 3 facility. In experiment 2, rBCG (designated as ‘rB’)was used for priming instead of BCG, with the same dose andprocedure mentioned above for experiment 1. Animal care andhandling was performed by authorized personnel. All procedureswere approved by the Wuhan University Institutional Animal Careand Use Committee (IACUC).

2.7. Humoral responses

Antigen-specific antibodies were detected by standard ELISA1 week after the final vaccination. The ELISA plates were precoatedwith Ag85B or ESAT6 (0.25 mg/well). Sera diluted 1:25 andhorseradish peroxidase (HRP)-conjugated secondary antibodydiluted 1:4000 (goat anti-mouse IgG, IgG1, IgG2a; SouthernBiotech) were used, and the enzyme reaction was developed with3,3,5,5-tetramethylbenzidine (TMB) substrate for 10–15 min andterminated by the addition of 2 M H2SO4. OD values weredetermined with a microplate reader at a wavelength of 450 nm.

2.8. Isolation of whole splenocytes

Immunized mice (n = 5) were sacrificed at 12 weeks after thefirst vaccination. Spleens were removed aseptically and pooledinto complete RPMI 1640 medium supplemented with 10% fetalbovine serum (FBS), 100 U of penicillin/ml, 100 mg of streptomy-cin/ml, and 2 mM L-glutamine. After red blood cell lysis with ACKlysis buffer (NH4Cl 0.15 M, KHCO3 10.0 mM, ethylenediaminete-traacetic acid disodium (Na2-EDTA) 0.1 mM, pH 7.2–7.4), spleno-cytes were filtered and resuspended at 1 � 107 cells/ml.

2.9. IFN-g ELISPOT assays

IFN-g ELISPOT assays were performed as described previous-ly.29 Briefly, splenocytes were seeded into the 96-well PVDFmicroplate (3 � 105 cells/well) precoated overnight with mouseIFN-g antibody. Cells were then incubated with or without thestimulator (final concentrations of 5 mg/ml) for 40 h at 37 8C, andspots were developed in accordance with the manufacturer’sprotocol (BD Biosciences). The results were expressed as thenumber of spot-forming cells (SFC) per 106 splenocytes.

2.10. In vitro cytotoxicity assay

In vitro cytotoxicity was determined as described previously.29

Briefly, P815 cells were loaded with the Ag85B-CD8 peptide

MPVGGQSSF (5 mg/ml; 2 h at 37 8C) and then labeled withdifferent concentrations of the carboxyfluorescein succinimidylester (CFSE) fluorescent dye (5 mM for peptide-loaded, 0.5 mM forunloaded) in RPMI 1640 medium without FBS for 10 min.36

CFSEhigh- and CFSElow-labeled cells were mixed together in a 1:1ratio. Splenocytes (1 � 106 cells) were then incubated with theP815 cells (5 � 104 peptide-loaded, 5 � 104 unloaded) for 6 h at37 8C, after which the co-cultures were analyzed on a BDFACSCalibur for assessment of the percentage of CFSE-labeledP815 cells. Specific killing was calculated as follows: % killing =[1 � (peptide loaded cells/unloaded cells from immunized group) �(peptide loaded cells/unloaded cells from naı̈ve group)] � 100.

2.11. Bacterial colony formation assay

The mice were sacrificed 8 weeks after the challenge. The rightlungs and spleen were removed and homogenized in 0.5% Tween-80–PBS, and the bacterial burden was determined by plating 10-fold serial dilutions in triplicate onto 7H11 medium. Plates wereincubated inside semi-sealed plastic bags at 37 8C in 5% CO2 for 3–4weeks. Colonies were counted, calculated, and presented aslog10CFU per organ.

2.12. Lung histology

The left lungs were fixed in 10% neutral buffered formalin andembedded in paraffin. Sections approximately 5 mm thick werestained with hematoxylin and eosin (H&E) for histologicalexamination and with Ziehl–Neelsen stain for evaluation ofacid-fast bacilli. The histopathological parameters of interstitialinflammation, edema, peribronchiolitis, perivasculitis, and granu-loma formation were each semi-quantitatively scored as absent,slight, moderate, marked, or strong, and noted as 0, 1, 2, 3, and 4,respectively. The cumulative inflammation score was expressed asthe sum of the scores for each parameter, with the maximum being20. Confluent inflammatory infiltrate was quantified separatelyand expressed as a percentage of the lung surface. The results wererecorded by a pathologist without knowledge of the experimentalgroups.

2.13. Statistical analysis

The data were analyzed using one-way analysis of variance(ANOVA). Differences between the groups were assessed forstatistical significance by Tukey’s post-hoc test. The Mann–Whitney U-test was employed for analysis of histologicalevaluations. A p-value of

Figure 2. Humoral responses to BCG priming and viral vaccine boosting. Sera of the mice (n = 12) were collected 1 week after the final vaccination and analyzed individually at1/25 dilution by ELISA for the presence of Ag85B-specific IgG, IgG1, and IgG2a. Absorbance values (measured at 450 nm) are shown, with the vertical solid lines in each panel

representing means. The IgG2a/IgG1 ratio was obtained by dividing the mean OD value of IgG2a by that of IgG1.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825 e819

ratio was clearly increased in these groups compared to those inthe other groups. Due to the low immunogenicity and absence ofESAT6 in BCG, antibody responses to ESAT6 were weaker thanthose to Ag85B after booster vaccination. However, levels of IgG2aantibody to ESAT6 were obviously elevated compared to those ofIgG1 antibodies in the B + A + M and B + M + A groups (data notshown). Together, these data demonstrate that viral boosters couldimprove the humoral responses induced by BCG, and two boostersafter BCG priming considerably enhanced the ratio of IgG2a toIgG1, indicating a Th1-biased immune response.

3.2. IFN-g responses after vaccination

Cell-mediated immunity was assessed by the frequency of IFN-g-producing splenocytes in a sensitive ELISPOT assay inexperiment 1. To determine the relative contribution of each T-cell subset to IFN responses, dominant CD8 and CD4 T-cell epitopeswere employed. As shown in Figure 3, a single vaccination withBCG generated only marginal Ag85B-specific IFN-g producingcells. After boosting with MVA85B-E6 or AD85B-E6, Ag85B-specificIFN-g-secreting cells were significantly increased. While boostervaccination of MVA85B-E6 generated more 18-1p-specific CD4 T-cells, AD85B-E6 as booster induced remarkable levels of both CD8and CD4 IFN-g-producing T-cells, especially 9-2p-specific CD8 T-cells. The B + M and B + A groups also showed enhanced IFN-gresponses as compared to those induced by a single viral vectoredvaccine immunization (data not shown). When BCG-primed micereceived two doses of viral boosters, Ag85B-specific IFN-g-producing cells were not obviously increased as compared with

Figure 3. Frequency of IFN-g-producing splenocytes induced by heterologous prime-boantigens (Ag85B, ESAT6) or peptides (9-2p, 9-1p, 18-2p, 18-1p) at a final concentration of

each peptide was derived are indicated in brackets. Data are presented as the mean � stagroup. Spontaneous IFN-g-producing cells considered as background were subtracted. SFC

those from the B + A and B + M groups. However, IFN-g-secretingCD8 T-cells in the B + A + M and B + M + A groups reached higherlevels than those of CD4 T-cells. In the assay, spots produced bysplenocytes stimulated with CD4 T-cell restricted peptides (18-2pand 18-1p) were larger than those produced by the CD8 T-cellrestricted peptides (9-2p and 9-1p), implying higher IFN-gproduction capacity on a per cell basis. As mentioned above, thelow immunogenicity of ESAT6 resulted in much fewer ESAT6-specific than Ag85B-specific T-cells in the boosted groups. Ingeneral, the prime-boost strategy could significantly augment theantigen-specific IFN-g-producing cells, and the CD8 T-cellresponse was well developed especially in the groups withAD85B-E6 boosting.

3.3. CD8 cytotoxic T-lymphocyte (CTL) responses in vitro

To further investigate the T-cell effector function in experiment1, we detected CD8 cytotoxic T-lymphocyte (CTL) responses thatmay be important for controlling intracellular infection. We co-cultured splenocytes with syngeneic target cells (P815) pulsedwith the dominant Ag85B-specific CD8 T-cell peptide MPVGGQSSF(9-2p) at an optimal 10:1 effector:target cell ratio in all groups(Figure 4). As expected, vaccination with BCG followed byMVA85B-E6 boosting, as with the previous finding29 that neitheralone could trigger CD8 CTLs, did not show significant killing.However, priming with BCG significantly augmented AD85B-E6-induced cytotoxicity from 21.7 � 1.2% to 32.9 � 1.0%. Furthermore,the strongest specific CTL activity was generated by the B + M + Agroup (50.9 � 1.2%), and the enhancement was significant compared

ost vaccination. Mouse splenocytes (3 � 105 cells/well) were incubated with whole 5 mg/ml. Epitope restriction by either CD4 or CD8 T cells and the antigen from whichndard error of the mean of triplicate wells of pooled splenocytes from five mice in each

, spot forming cells; B, BCG; M, MVA85B-E6; A, AD85B-E6.

Figure 4. In vitro CD8 T-cell cytotoxicity assay. P815 cells pulsed with the Ag85B-specific CD8 T-cell peptide MPVGGQSSF (9-2p) were used as target cells. Data are

presented as the mean � standard error of the mean of triplicate co-culture samplesfrom pooled splenocytes of five mice from each group. ***p < 0.001; ns, not statistically

significant.

Figure 5. Immune protection by different vaccination strategies 8 weeks afterM. tuberculosis challenge. The mice were challenged with M. tuberculosis

intravenously 4 weeks after the last vaccination. Mice were sacrificed 8 weeks

later and the M. tuberculosis bacterial titers in infected lungs (A) and spleens

(B) were determined by colony enumeration assays. Data are presented as the mean

� standard error of the mean of triplicate plates from pooled tissues of five mice ineach group. ***p < 0.001; ns, not statistically significant. The comparisons were made

with the BCG group.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825e820

to that of the B + A group. The CTL responses of the B + A + M groupwere weaker than those of the B + A and B + M + A groups, indicatingthat the cytotoxicity may have decreased over time, since adminis-tration of AD85B-E6 in the B + A + M group was 2 weeks earlier thanthat in the B + A and B + M + A groups. In summary, these resultsdemonstrate that the cytolytic activity of AD85B-E6 could beenhanced by BCG or B + M priming, but it was transient and declinedrapidly. BCG and MVA85B-E6 were unable to trigger CTL activitywhen used alone, but they could enhance 9-2p-specific CD8 Tcytotoxic activity tested at 2 weeks after the last immunization ofAD85B-E6.

3.4. Immune protection against TB

In order to determine whether the viral booster improved priorBCG-induced protection, we challenged BALB/c mice with 5 � 105CFU of M. tuberculosis (H37Rv) intravenously 4 weeks after the lastvaccination (Figure 1). Eight weeks later, bacterial burdens in lungsand spleens were determined. Compared with naı̈ve controls (PBSgroup), a single dose of BCG decreased the CFU counts by 1.2 log inthe lungs (Figure 5A) and spleens (Figure 5B). Surprisingly,MVA85B-E6, which previously provided a striking level ofprotection in both organs,29 could not enhance that induced byBCG, indicating that even a vaccine that is efficient by itself maynot necessarily be an appropriate booster to improve protectiveefficacy. Similar CFU counts were found in the B + A group and theBCG group, suggesting that a single dose of AD85B-E6 boostercould not improve protection either. However, in the lungs of BCG-vaccinated mice boosted with M + A and A + M, significantreductions in the bacterial titer were recorded (�0.3 log and�0.6 log, respectively) compared with non-boosted BCG-vaccinatedmice, though both groups conferred splenic protection equivalent tothat by BCG. It should be noted that a better protective effect of0.3 log CFU reduction in the lungs was significant for mice boostedwith A + M compared with M + A-boosted mice. These resultssuggest that combining more than one viral vaccine may be aneffective way to enhance BCG-induced protection, and the outcomemay be influenced by the order in which these boosters areadministered.

3.5. Evaluation of pathological changes after M. tuberculosischallenge

We further assessed the pathology in the lungs of mice post-infection. Based on gross examination of the mouse anatomy,

nearly normal lungs with few tubercles were observed in thevaccinated groups as compared to the lungs with visible tuberclesfrom the naı̈ve control group (PBS). After staining with H&E, thelungs in the PBS group exhibited about 40% multifocal granulo-matous infiltration, which was characterized by numerous centralfoamy macrophages with scattered infiltrates of inflammatorycells (Figure 6, A and C). In BCG-vaccinated mice, less than 20%granulomatous inflammation was observed (Figure 6, A and D).Boosting BCG with AD85B-E6 resulted in diffuse inflammationwith vigorous lymphocytic infiltration (Figure 6F), whereas moreaggregated lymphocytes were noted in the B + M group (Figure 6E)than in the B + A group. Lymphocytes were also frequently found inthe lungs of the B + M + A group, as with those in the B + A group;however, they were more regulated and the inflammatoryinfiltrate was significantly less diffuse (Figure 6, A and H). Thelungs in the B + A + M group appeared nearly normal except fordistinct compact granulomas, which contained and were sur-rounded by abundant lymphocytes. However, the histologicalchanges in the viral vaccine-boosted groups were not significantlyimproved as compared to those in the BCG group in a semi-quantitative evaluation of the pathological response (Figure 6B).

3.6. Characterization of rBCG

We also constructed and characterized a batch of rBCGexpressing Ag85B, which is a promising TB candidate vaccine inclinical development, in order to test the boost potential of viralvaccines to this improved mycobacterial vaccine. It was generatedby electrotransformation of wild-type BCG, and the expression ofAg85B was analyzed in the culture. rBCG expressed a relativelyhigher level of Ag85B (35.8%) as compared with that of the

Figure 6. Histological evaluation in lungs of M. tuberculosis-infected mice. The percentages of confluent inflammatory infiltration in the lung (A) and cumulative inflammationscores (B) were determined. Data are presented as box plots, wherein median values are denoted by a horizontal line, the mean by ‘+’, interquartile range by boxes, and the

maximum and minimum values by whiskers for each group. (C)–(H) show representative histological appearances of the lung tissues stained with H&E (regimens are

indicated in brackets). *p < 0.05.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825 e821

wild-type BCG (12.1%) in the culture filtrate proteins (Figure 7A),and this led to substantially enhanced Ag85B secretion by rBCG ascompared to that by the same number of BCG (Figure 7B). Whensubcultured six times over an 18-week period in the absence ofhygromycin, the cultures showed no decrease in Ag85B expression(Figure 7C), suggesting that rBCG could maintain overexpression ofAg85B without selective pressure. To determine whether theenhanced Ag85B secretion could improve the immunogenicity ofAg85B, mice were injected subcutaneously at a dose of 1 � 106 CFUBCG or rBCG. As expected, rBCG induced significantly more Ag85B-specific IFN-g-producing T-cells than did BCG (Figure 7D), whileresponses to TB10.4, the other immunodominant antigen used asan irrelevant stimulator, showed no difference in these two groups.These results indicate that Ag85B overexpressing rBCG wassuccessfully constructed and that it could trigger increasedantigen-specific cell-mediated immune responses.

3.7. Effects of rBCG priming and viral boosting on protection against

TB challenge

The immune responses and protective efficacy induced by rBCGpriming and viral vaccine boosting in experiment 2 wereevaluated. Immunization regimens here were the same as thosein experiment 1, except that rBCG was used for priming. Asobserved in the BCG priming groups, two sequential viral boostvaccinations enhanced antibody responses that favored the IgG2asubclass and elevated the IgG2a/IgG1 ratio (Figure 8A). Results ofELISPOT and CTL assays showed similar patterns between the

groups primed with rBCG and those with BCG (data not shown).The bacterial burden was evaluated after challenge to determinethe protective efficacy of the various immunization protocols, andonly mice in the rB + A + M group showed significant CFUreductions compared to a single dose of rBCG-vaccinated micein both lungs and spleens (0.4 log and 0.3 log, respectively). Nosignificantly enhanced protective effect over that induced by rBCGwas observed for either organ in the rB + M + A group (Figure 8, Band C). As with the histological changes observed in experiment 1,rBCG followed by sequential AD85B-E6 and MVA85B-E6 boosting,led to compact granulomas with markedly increased lymphocyticinfiltration relative to that induced by rBCG alone, and fewerdiffuse lymphocytes were seen in the surrounding parenchyma ascompared with that induced by rB + M + A vaccination (Figure 8, D,E and F).

4. Discussion

BCG is the only approved vaccine against TB. However, itgenerates limited protection due to extensive attenuation inculture, interference by environmental mycobacteria, and earlyclearance in some populations.38 Furthermore, the protectionafforded by this vaccine declines over time. Thus, there is an urgentneed to develop new anti-TB vaccination strategies. Sincecontinuous antigen exposure provided by a live bacterial vaccineseems necessary for the maintenance of protection against TB, andthe booster should be non-mycobacterial organism-based, aheterologous prime-boost immunization strategy was adopted

Figure 7. Analysis of rBCG. Culture filtrate proteins of BCG and rBCG were analyzedby Tricine-SDS-PAGE (A) and Western blot using a rabbit anti-M. tuberculosis Ag85B

polyclonal antibody (B and C). (A) and (B): lane 1, culture filtrates from BCG; lane 2,

culture filtrates from rBCG; lane M, molecular weight standard. (C) Plasmid stability

in rBCG was evaluated in the absence of hygromycin; lanes 1–6, culture filtrates

from equivalent CFU of rBCG of the first to the sixth passage; lane M, molecular

weight standard. (D) ELISPOT assays were performed with splenocytes from mice in

two independent experiments. Mice receiving a single dose of PBS, BCG, or rBCG

were sacrificed 4 weeks later. Representative wells (3 � 105 cells/well) in eachgroup are shown. Ag85B and TB10.4 were used at a concentration of 5 mg/ml, andthe negative control consisted of unstimulated splenocytes.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825e822

in the current study.8 Here we investigated the immune status andprotective efficacy in mice vaccinated with BCG or rBCG primingfollowed by boosting with viral-vectored vaccines.

During the past few years, viral vectors have been extensivelydeveloped as vaccine platforms.39 Unlike DNA-based and protein-based vaccines, viral vectors with their intense adjuvant activityare effective at carrying and delivering antigen-encoding gene(s) totrigger strong immune responses,40,41 especially cell-mediatedimmune responses, which play a key role in eliminatingM. tuberculosis as an intracellular pathogen.42 Replication-deficientadenovirus and poxvirus vectors with good safety records havebeen developed in the field of TB vaccines.22 In fact, we previouslyengineered the MVA85B-E6 vaccine, which showed a comparableprotective effect to that of BCG, while AD85B-E6 demonstrated noprotective efficacy despite generating apparently high cell-mediated immune responses.29 Those two vaccines were bothconstructed to express Ag85B and ESAT6 as a fusion protein forseveral reasons. First, the two antigens have been demonstrated tobe potent T-cell immunogens and contribute to protective

immunity. Second, since ESAT6 is absent in BCG, the introductionof ESAT6 could broaden the range of epitopes available toovercome major histocompatibility complex (MHC)-relatedrestrictions. Third, the fusion protein processing and presentingmay be more efficient than a multiple component protein cocktail.Furthermore, recent findings indicate that fusion antigens provideimproved immunogenicity and protection against TB, both asprotein-adjuvant formulation12 and as a gene product.43

The levels and IgG subclass distribution of serum antibodies cangive an indication of the nature of the vaccine-induced immuneresponses. In this study, although a single viral booster generatedmoderate effects on humoral responses, two sequential viralbooster vaccinations triggered enhanced antibody responses andshifted the Th1/Th2 balance toward Th1 responses. The viralvaccines induced predominantly the IgG2a isotype, which is quitedifferent from that triggered by protein-adjuvant formulations.Even when formulated with an adjuvant that can induce cell-mediated immunity, a protein antigen will still trigger mainly anIgG1-specific response.44 This difference in favored antibodysubclasses reflects the differential nature of the immune responsesinduced by virus and protein vaccines.

Due to distinct properties such as viral structure, entry,replication and expression of foreign antigens, different viralvector systems possess variable immune-stimulatory features.40,41

Recombinant adenoviruses generally induce robust CD8 T-cellimmunity and CTLs, whereas those responses stimulated bymodified vaccinia virus Ankara (MVA) vectors are relativelyweaker, as we and others have observed.29,39,45 Accordingly, theyalso generate varying boosting effects on mycobacteria-triggeredimmune activation. Hence, it is not surprising that the frequenciesof IFN-g-producing splenocytes and T cytotoxic activity weresubstantially enhanced by a single AD85B-E6 boost, comparedwith MVA85B-E6. Remarkably, the combination of MVA85B-E6and AD85B-E6 further elevated the levels of IFN-g-secreting T-cells, especially CD8 T-cells, suggesting a synergistic effect.Intriguingly, specific cytotoxic T-cell activity in the B + A + Mgroup was lower than that of the B + A and B + M + A groups. Thismay be explained by the kinetics of cytotoxicity triggered byadenovirus, which is the pivotal inducer of CTLs. Although CTLsprovide potent defenses against intracellular pathogens, their lyticmachinery can be deleterious to self tissues.46 To balance thebeneficial and detrimental effects, the host reduces the populationsize of CTLs after destroying the targets in the contractionphase.47,48 Therefore, the decreased CTLs may be due to theearlier administration of AD85B-E6 in the B + A + M groupthan that in the other groups. Furthermore, although BCG andMVA85B-E6 did not induce significant CTL activity when usedalone, their effects could synergize with that of the adenovirusvector, thereby inducing the specific cytotoxicity to a higher level.

Defining immune biomarkers that correlate with protection iscrucial to TB vaccine research and development. Although IFN-g isa potent indicator of Th1 activity and has often been used as anindicator of protective immunity,49 this correlation has recentlybeen questioned, at least in some studies.50–52 Appropriate subsetsof T-cells with multifunctional properties appear to be predictiveof protection in other studies,53–55 suggesting that the magnitudeand quality of T-cell responses would be useful in assessingprotection. Numerous studies have demonstrated that MVAvectors engender more balanced CD4 and CD8 T-cell responsesthan adenovirus vectors, which favor stronger CD8 T-cellresponses. In addition, MVA vectors generate accelerated peaksof antigen expression and T-cell responses, followed by a quickcontraction. In contrast, adenovirus vectors generate persistentantigen expression, with a delayed peak of T-cell responses and aslow contraction.39,45,56 The differences in kinetics of antigenclearance, subsets, and magnitude of T-cell responses will

Figure 8. Immune responses and protective efficacy by combined viral boosting of rBCG. (A) Humoral responses of rBCG priming and viral vaccine boosting. Absorbancevalues of specific anti-Ag85B IgG, IgG1, and IgG2a antibodies were determined by ELISA. Vertical solid lines indicate the means. Bacterial burdens were measured in the lungs

(B) and spleens (C) of mice. Data are presented as the mean � standard error of the mean of triplicates from five mice per group. ***p < 0.001; ns, not statistically significant.Comparisons were made with the BCG group. Representative histological images of lungs are shown for mice that were vaccinated with rB (D), rB + A + M (E), or rB + M + A (F).

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825 e823

eventually result in different characteristics of memory T-cells.57,58

It becomes more complicated when two viral vaccines arecombined, and even the order of vaccination may influencethe outcome.59,60 In the present study, bacterial priming plusAD85B-E6 boosting followed by MVA85B-E6 was shown to providebetter protection than that with MVA85B-E6 boosting followed byAD85B-E6. Furthermore, the observation of more compactgranulomas with controlled lymphocyte recruitment suggeststhat the effectiveness of this regimen may result from theinduction of T-cell responses with appropriate magnitude andquality. Interestingly, the immune protection triggered by anadenovirus vector followed by MVA vector has also been found tobe better than the reverse order of vaccination against malaria.59

Whether the order of administration of adenovirus and MVAvaccines would be important in other vaccine fields and the precisemechanisms underlying different regimens with varying protec-tion levels will require further investigation.

rBCG was engineered as an alternate priming vaccine to BCGfor potential clinical application. Consistent with othersstudies,61 rBCG did not afford better protection than BCG inmice in our study, although it did show enhanced antigenexpression and immunogenicity. However, rBCG has been shownto be highly protective in other animal models.23,26,27 Despitethe different levels of protection in various animal models, BCGoverexpressing Ag85B has entered clinical trials. Accordingly, weinvestigated the efficacy of viral boosters after rBCG priming.

rB + A + M vaccination provided protection to both the lung andthe spleen, in contrast to rB + M + A vaccination. This finding wassomewhat different from the results observed in the B + A + Mand B + M + A groups and suggests that mycobacterial primingwith different strains can also influence the outcome. CD4 T-cell-mediated heterologous immunity between mycobacteria andpoxviruses has been reported.62 Cross-reactivity that influencesthe immune response to an unrelated pathogen may explain, atleast in part, why a single dose of MVA85B-E6 affordingsignificant protective efficacy could not boost that impartedby the mycobacterial vaccine, and why distinct priming regi-mens resulted in various levels of efficacy. This hypothesis issupported by recent findings that the immunogenicity ofMVA85A was reduced by co-administration with othervaccines63 and AD85B-E6 induced serious pathology in the lungsurface with masses of tubercles when used alone,29 while wefound here that mycobacterial vaccine priming plus AD85B-E6boosting resulted in a relatively normal anatomical appearance.This may be due to the immune status primed by themycobacterial vaccine, probably influenced by regulatoryT-cells, which have been proven to be induced by mycobac-teria64 and could be crucial players in balancing harmful tissuedamage and efficient effector immunity.65 Thus, identifying howthe mycobacterial vaccine affects the host response to virusesand the interactions between different types of vaccines isimportant for the design of better vaccines.

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825e824

It has been suggested that the geography of antigen-specific T-cells is important for the protective efficacy against infection,66 i.e.,the functional T-cells should be recruited and retained at the placeof infection. This idea is supported by recent studies showing thatonly mucosal vaccination with viral vaccines can induce orimprove prior BCG-induced protection against aerosol chal-lenge.15,20,67 Although the aerosol challenge more closely mimicsthe natural route of TB infection, it is difficult to carry out, and therelatively imprecise dose limits its application. The intravenousroute of infection with a precise dose of bacteria is potentiallyrelevant for initial screening and comparative studies of multiplevaccines given by parenteral immunization.68 The injectedbacteria infect the organs from the circulation, where they canbe encountered and eliminated by the functional T-cells inducedby the systemic route of vaccine administration. Of course, thisscenario is different from airway-invading bacteria, which wouldrequire destruction by airway luminal T-cells induced byrespiratory mucosal immunization.69,70 The effectiveness of thebacterial priming with two boosters of adenovirus and MVAvectors here provides a foundation for further evaluation bymucosal vaccination against aerosol challenge.

Much effort has been made to improve BCG induced protection.However, the efficacy fluctuates with vaccine type and delivery,antigen expression and duration, route of challenge, and animalmodels. Combinations of different viral systems that can furtherstrengthen the cellular immunity, improve protection, and avoidthe limited efficacy due to pre-existing immunity are increasinglybeing studied for vaccination.60,71,72 The high plasticity of T-cellresponses also prompts the rational design of vaccination regi-mens. Our study here characterized the immune responses andprotective efficacy induced by mycobacteria as a priming vaccineplus virus vectors for boosting. We conclude that AD85B-E6followed by MVA85B-E6 boost vaccination can improve BCG- orrBCG-induced protection. Our findings lend support to futurestudies of mucosal viral vaccine immunization against aerosolinfection. Further work is also needed to identify more relevantbiomarkers for protective immunity and the precise mechanismsof immunological memory formation. Nevertheless, this studyhighlights the superior ability of multiple heterologous viralbooster vaccines and encourages the combined use of AD85A andMVA85A, which will be evaluated in clinical trials in the nearfuture.

Acknowledgements

The authors gratefully acknowledge Marcus A. Horwitz forproviding pSMT3, and Rong Bao, Ming Guo, and Xin Wang (ABSL-3Lab, Wuhan University) for technical assistance. We thank PhuongThi Nguyen Sarkis for editorial assistance on the manuscript.

Ethical approval: Ethical approval for the study was obtainedfrom Wuhan University Institutional Animal Care and UseCommittee (IACUC).

Conflict of interest: No conflict of interest to declare.

References

1. World Health Organization. Global tuberculosis control 2011. Geneva: WHO;2011.

2. Tseng CL, Oxlade O, Menzies D, Aspler A, Schwartzman K. Cost-effectiveness ofnovel vaccines for tuberculosis control: a decision analysis study. BMC PublicHealth 2011;11:55.

3. Brewer TF. Preventing tuberculosis with bacillus Calmette–Guerin vaccine: ameta-analysis of the literature. Clin Infect Dis 2000;31(Suppl 3):S64–7.

4. Tala-Heikkila MM, Tuominen JE, Tala EO. Bacillus Calmette–Guerin revaccina-tion questionable with low tuberculosis incidence. Am J Respir Crit Care Med1998;157:1324–7.

5. Buddle BM, Wedlock DN, Parlane NA, Corner LA, De Lisle GW, Skinner MA.Revaccination of neonatal calves with Mycobacterium bovis BCG reduces the

level of protection against bovine tuberculosis induced by a single vaccination.Infect Immun 2003;71:6411–9.

6. Basaraba RJ, Izzo AA, Brandt L, Orme IM. Decreased survival of guinea pigsinfected with Mycobacterium tuberculosis after multiple BCG vaccinations.Vaccine 2006;24:280–6.

7. Ramshaw IA, Ramsay AJ. The prime-boost strategy: exciting prospects forimproved vaccination. Immunol Today 2000;21:163–5.

8. Xing Z, Charters TJ. Heterologous boost vaccines for bacillus Calmette–Guerinprime immunization against tuberculosis. Expert Rev Vaccines 2007;6:539–46.

9. Rouanet C, Locht C. Boosting BCG to protect against TB. Expert Rev Respir Med2010;4:339–48.

10. Andersen CS, Dietrich J, Agger EM, Lycke NY, Lovgren K, Andersen P. Thecombined CTA1-DD/ISCOMs vector is an effective intranasal adjuvant forboosting prior Mycobacterium bovis BCG immunity to Mycobacterium tubercu-losis. Infect Immun 2007;75:408–16.

11. Palma C, Iona E, Giannoni F, Pardini M, Brunori L, Fattorini L, et al. The LTK63adjuvant improves protection conferred by Ag85B DNA-protein prime-boost-ing vaccination against Mycobacterium tuberculosis infection by dampeningIFN-gamma response. Vaccine 2008;26:4237–43.

12. Aagaard C, Hoang T, Dietrich J, Cardona PJ, Izzo A, Dolganov G, et al. A multistagetuberculosis vaccine that confers efficient protection before and after exposure.Nat Med 2011;17:189–94.

13. McShane H, Pathan AA, Sander CR, Goonetilleke NP, Fletcher HA, Hill AV.Boosting BCG with MVA85A: the first candidate subunit vaccine for tuberculo-sis in clinical trials. Tuberculosis (Edinb) 2005;85:47–52.

14. Pathan AA, Sander CR, Fletcher HA, Poulton I, Alder NC, Beveridge NE, et al.Boosting BCG with recombinant modified vaccinia Ankara expressing antigen85A: different boosting intervals and implications for efficacy trials. PLoS ONE2007;2:e1052.

15. Santosuosso M, McCormick S, Zhang X, Zganiacz A, Xing Z. Intranasal boostingwith an adenovirus-vectored vaccine markedly enhances protection by paren-teral Mycobacterium bovis BCG immunization against pulmonary tuberculosis.Infect Immun 2006;74:4634–43.

16. Wang J, Zganiacz A, Xing Z. Enhanced immunogenicity of BCG vaccine by usinga viral-based GM-CSF transgene adjuvant formulation. Vaccine 2002;20:2887–98.

17. Derrick SC, Yang AL, Morris SL. A polyvalent DNA vaccine expressing anESAT6-Ag85B fusion protein protects mice against a primary infection withMycobacterium tuberculosis and boosts BCG-induced protective immunity.Vaccine 2004;23:780–8.

18. Bennekov T, Dietrich J, Rosenkrands I, Stryhn A, Doherty TM, Andersen P.Alteration of epitope recognition pattern in Ag85B and ESAT-6 has a profoundinfluence on vaccine-induced protection against Mycobacterium tuberculosis.Eur J Immunol 2006;36:3346–55.

19. McShane H, Behboudi S, Goonetilleke N, Brookes R, Hill AV. Protective immu-nity against Mycobacterium tuberculosis induced by dendritic cells pulsed withboth CD8(+)- and CD4(+)-T-cell epitopes from antigen 85A. Infect Immun2002;70:1623–6.

20. Wang J, Thorson L, Stokes RW, Santosuosso M, Huygen K, Zganiacz A, et al. Singlemucosal, but not parenteral, immunization with recombinant adenoviral-basedvaccine provides potent protection from pulmonary tuberculosis. J Immunol2004;173:6357–65.

21. Perera PY, Derrick SC, Kolibab K, Momoi F, Yamamoto M, Morris SL, et al. Amulti-valent vaccinia virus-based tuberculosis vaccine molecularly adjuvantedwith interleukin-15 induces robust immune responses in mice. Vaccine2009;27:2121–7.

22. Xing Z, Santosuosso M, McCormick S, Yang TC, Millar J, Hitt M, et al. Recentadvances in the development of adenovirus- and poxvirus-vectored tubercu-losis vaccines. Curr Gene Ther 2005;5:485–92.

23. Horwitz MA, Harth G, Dillon BJ, Maslesa-Galic S. Recombinant bacillusCalmette–Guerin (BCG) vaccines expressing the Mycobacterium tuberculosis30-kDa major secretory protein induce greater protective immunity againsttuberculosis than conventional BCG vaccines in a highly susceptible animalmodel. Proc Natl Acad Sci U S A 2000;97:13853–8.

24. Grode L, Seiler P, Baumann S, Hess J, Brinkmann V, Nasser Eddine A, et al.Increased vaccine efficacy against tuberculosis of recombinant Mycobacteriumbovis bacille Calmette–Guerin mutants that secrete listeriolysin. J Clin Invest2005;115:2472–9.

25. Sun R, Skeiky YA, Izzo A, Dheenadhayalan V, Imam Z, Penn E, et al. Novelrecombinant BCG expressing perfringolysin O and the over-expression of keyimmunodominant antigens; pre-clinical characterization, safety and protectionagainst challenge with Mycobacterium tuberculosis. Vaccine 2009;27:4412–23.

26. Horwitz MA, Harth G, Dillon BJ, Maslesa-Galic S. A novel live recombinantmycobacterial vaccine against bovine tuberculosis more potent than BCG.Vaccine 2006;24:1593–600.

27. Horwitz MA, Harth G. A new vaccine against tuberculosis affords greatersurvival after challenge than the current vaccine in the guinea pig model ofpulmonary tuberculosis. Infect Immun 2003;71:1672–9.

28. Horwitz MA. Recombinant BCG expressing Mycobacterium tuberculosis majorextracellular proteins. Microbes Infect 2005;7:947–54.

29. You Q, Jiang C, Wu Y, Yu X, Chen Y, Zhang X, et al. Subcutaneous administrationof modified vaccinia virus Ankara expressing an Ag85B–ESAT6 fusion protein,but not an adenovirus-based vaccine, protects mice against intravenous chal-lenge with Mycobacterium tuberculosis. Scand J Immunol 2012;75:77–84.

30. You Q, Wu Y, Jiang D, Wang C, Wei W, Yu X, et al. Immune responses inducedby heterologous boosting of recombinant bacillus Calmette–Guerin with

Q. You et al. / International Journal of Infectious Diseases 16 (2012) e816–e825 e825

Ag85B–ESAT6 fusion protein in levamisole-based adjuvant. Immunol Invest2012;41:412–28.

31. Radosevic K, Wieland CW, Rodriguez A, Weverling GJ, Mintardjo R, Gillissen G,et al. Protective immune responses to a recombinant adenovirus type 35tuberculosis vaccine in two mouse strains: CD4 and CD8 T-cell epitope map-ping and role of gamma interferon. Infect Immun 2007;75:4105–15.

32. D’Souza S, Rosseels V, Romano M, Tanghe A, Denis O, Jurion F, et al. Mapping ofmurine Th1 helper T-cell epitopes of mycolyl transferases Ag85A, Ag85B, andAg85C from Mycobacterium tuberculosis. Infect Immun 2003;71:483–93.

33. Denis O, Tanghe A, Palfliet K, Jurion F, van den Berg TP, Vanonckelen A, et al.Vaccination with plasmid DNA encoding mycobacterial antigen 85A stimulatesa CD4 + and CD8 + T-cell epitopic repertoire broader than that stimulated byMycobacterium tuberculosis H37Rv infection. Infect Immun 1998;66:1527–33.

34. Brandt L, Oettinger T, Holm A, Andersen AB, Andersen P. Key epitopes on theESAT-6 antigen recognized in mice during the recall of protective immunity toMycobacterium tuberculosis. J Immunol 1996;157:3527–33.

35. Olsen AW, Hansen PR, Holm A, Andersen P. Efficient protection againstMycobacterium tuberculosis by vaccination with a single subdominant epitopefrom the ESAT-6 antigen. Eur J Immunol 2000;30:1724–32.

36. Jellison ER, Kim SK, Welsh RM. Cutting edge: MHC class II-restricted killing invivo during viral infection. J Immunol 2005;174:614–8.

37. Stevens TL, Bossie A, Sanders VM, Fernandez-Botran R, Coffman RL, MosmannTR, et al. Regulation of antibody isotype secretion by subsets of antigen-specifichelper T cells. Nature 1988;334:255–8.

38. Russell DG, Barry 3rd CE, Flynn JL. Tuberculosis: what we don’t know can, anddoes, hurt us. Science 2010;328:852–6.

39. Rollier CS, Reyes-Sandoval A, Cottingham MG, Ewer K, Hill AV. Viral vectors asvaccine platforms: deployment in sight. Curr Opin Immunol 2011;23:377–82.

40. Liu MA. Immunologic basis of vaccine vectors. Immunity 2010;33:504–15.41. Brave A, Ljungberg K, Wahren B, Liu MA. Vaccine delivery methods using viral

vectors. Mol Pharm 2007;4:18–32.42. North RJ, Jung YJ. Immunity to tuberculosis. Annu Rev Immunol 2004;22:

599–623.43. Mu J, Jeyanathan M, Small CL, Zhang X, Roediger E, Feng X, et al. Immunization

with a bivalent adenovirus-vectored tuberculosis vaccine provides markedlyimproved protection over its monovalent counterpart against pulmonarytuberculosis. Mol Ther 2009;17:1093–100.

44. Rosenkrands I, Agger EM, Olsen AW, Korsholm KS, Andersen CS, Jensen KT, et al.Cationic liposomes containing mycobacterial lipids: a new powerful Th1 adju-vant system. Infect Immun 2005;73:5817–26.

45. Geiben-Lynn R, Greenland JR, Frimpong-Boateng K, Letvin NL. Kinetics ofrecombinant adenovirus type 5, vaccinia virus, modified vaccinia Ankara virus,and DNA antigen expression in vivo and the induction of memory T-lympho-cyte responses. Clin Vaccine Immunol 2008;15:691–6.

46. Barry M, Bleackley RC. Cytotoxic Tlymphocytes: all roads lead to death. Nat RevImmunol 2002;2:401–9.

47. Sergeev RA, Batorsky RE, Rouzine IM. Model with two types of CTL regulationand experiments on CTL dynamics. J Theor Biol 2010;263:369–84.

48. Williams MA, Bevan MJ. Effector and memory CTL differentiation. Annu RevImmunol 2007;25:171–92.

49. Agger EM, Andersen P. Tuberculosis subunit vaccine development: on the roleof interferon-gamma. Vaccine 2001;19:2298–302.

50. Mittrucker HW, Steinhoff U, Kohler A, Krause M, Lazar D, Mex P, et al. Poorcorrelation between BCG vaccination-induced T cell responses and protectionagainst tuberculosis. Proc Natl Acad Sci U S A 2007;104:12434–9.

51. Goldsack L, Kirman JR. Half-truths and selective memory: interferon gamma,CD4(+) T cells and protective memory against tuberculosis. Tuberculosis (Edinb)2007;87:465–73.

52. Skinner MA, Ramsay AJ, Buchan GS, Keen DL, Ranasinghe C, Slobbe L, et al. ADNA prime-live vaccine boost strategy in mice can augment IFN-gammaresponses to mycobacterial antigens but does not increase the protectiveefficacy of two attenuated strains of Mycobacterium bovis against bovinetuberculosis. Immunology 2003;108:548–55.

53. Lindenstrom T, Agger EM, Korsholm KS, Darrah PA, Aagaard C, Seder RA, et al.Tuberculosis subunit vaccination provides long-term protective immunity

characterized by multifunctional CD4 memory T cells. J Immunol 2009;182:8047–55.

54. Beveridge NE, Price DA, Casazza JP, Pathan AA, Sander CR, Asher TE, et al.Immunisation with BCG and recombinant MVA85A induces long-lasting, poly-functional Mycobacterium tuberculosis-specific CD4 + memory T lymphocytepopulations. Eur J Immunol 2007;37:3089–100.

55. Darrah PA, Patel DT, De Luca PM, Lindsay RW, Davey DF, Flynn BJ, et al.Multifunctional TH1 cells define a correlate of vaccine-mediated protectionagainst Leishmania major. Nat Med 2007;13:843–50.

56. Reyes-Sandoval A, Wyllie DH, Bauza K, Milicic A, Forbes EK, Rollier CS, et al.CD8 + T effector memory cells protect against liver-stage malaria. J Immunol2011;187:1347–57.

57. Yang TC, Millar J, Groves T, Grinshtein N, Parsons R, Takenaka S, et al. TheCD8 + T cell population elicited by recombinant adenovirus displays a novelpartially exhausted phenotype associated with prolonged antigen presentationthat nonetheless provides long-term immunity. J Immunol 2006;176:200–10.

58. Harrington LE, van der Most R, Whitton JL, Ahmed R. Recombinant vacciniavirus-induced T-cell immunity: quantitation of the response to the virus vectorand the foreign epitope. J Virol 2002;76:3329–37.

59. Reyes-Sandoval A, Berthoud T, Alder N, Siani L, Gilbert SC, Nicosia A, et al.Prime-boost immunization with adenoviral and modified vaccinia virus Ankaravectors enhances the durability and polyfunctionality of protective malariaCD8 + T-cell responses. Infect Immun 2010;78:145–53.

60. Tatsis N, Lin SW, Harris-McCoy K, Garber DA, Feinberg MB, Ertl HC. Multipleimmunizations with adenovirus and MVA vectors improve CD8 + T cell func-tionality and mucosal homing. Virology 2007;367:156–67.

61. Palendira U, Spratt JM, Britton WJ, Triccas JA. Expanding the antigenic reper-toire of BCG improves protective efficacy against aerosol Mycobacteriumtuberculosis infection. Vaccine 2005;23:1680–5.

62. Mathurin KS, Martens GW, Kornfeld H, Welsh RM. CD4 T-cell-mediated heter-ologous immunity between mycobacteria and poxviruses. J Virol 2009;83:3528–39.

63. Ota MO, Odutola AA, Owiafe PK, Donkor S, Owolabi OA, Brittain NJ, et al.Immunogenicity of the tuberculosis vaccine MVA85A is reduced by coadmin-istration with EPI vaccines in a randomized controlled trial in Gambian infants.Sci Transl Med 2011;3:88ra56.

64. Joosten SA, van Meijgaarden KE, Savage ND, de Boer T, Triebel F, van der Wal A,et al. Identification of a human CD8 + regulatory T cell subset that mediatessuppression through the chemokine CC chemokine ligand 4. Proc Natl Acad SciU S A 2007;104:8029–34.

65. Belkaid Y, Rouse BT. Natural regulatory T cells in infectious disease. NatImmunol 2005;6:353–60.

66. Santosuosso M, McCormick S, Roediger E, Zhang X, Zganiacz A, Lichty BD, et al.Mucosal luminal manipulation of T cell geography switches on protectiveefficacy by otherwise ineffective parenteral genetic immunization. J Immunol2007;178:2387–95.

67. Goonetilleke NP, McShane H, Hannan CM, Anderson RJ, Brookes RH, Hill AV.Enhanced immunogenicity and protective efficacy against Mycobacteriumtuberculosis of bacille Calmette–Guerin vaccine using mucosal administrationand boosting with a recombinant modified vaccinia virus Ankara. J Immunol2003;171:1602–9.

68. North RJ. Mycobacterium tuberculosis is strikingly more virulent for mice whengiven via the respiratory than via the intravenous route. J Infect Dis1995;172:1550–3.

69. Jeyanathan M, Heriazon A, Xing Z. Airway luminal T cells: a newcomer on thestage of TB vaccination strategies. Trends Immunol 2010;31:247–52.

70. Holt PG, Strickland DH, Wikstrom ME, Jahnsen FL. Regulation of immunologicalhomeostasis in the respiratory tract. Nat Rev Immunol 2008;8:142–52.

71. Roediger EK, Kugathasan K, Zhang X, Lichty BD, Xing Z. Heterologous boostingof recombinant adenoviral prime immunization with a novel vesicular stoma-titis virus-vectored tuberculosis vaccine. Mol Ther 2008;16:1161–9.

72. Capone S, Reyes-Sandoval A, Naddeo M, Siani L, Ammendola V, Rollier CS, et al.Immune responses against a liver-stage malaria antigen induced by simianadenoviral vector AdCh63 and MVA prime-boost immunisation in non-humanprimates. Vaccine 2010;29:256–65.

Immunogenicity and protective efficacy of heterologous prime-boost regimens with mycobacterial vaccines and recombinant adenovirus- and poxvirus-vectored vaccines against murine tuberculosisIntroductionMaterials and methodsRecombinant proteins and peptidesPreparation of MycobacteriumCharacterization of rBCGTricine-SDS-PAGE and Western blottingRecombinant virusesMice, vaccination, and mycobacterial infectionHumoral responsesIsolation of whole splenocytesIFN-γ ELISPOT assaysIn vitro cytotoxicity assayBacterial colony formation assayLung histologyStatistical analysis

ResultsHumoral responses after vaccinationIFN-γ responses after vaccinationCD8 cytotoxic T-lymphocyte (CTL) responses in vitroImmune protection against TBEvaluation of pathological changes after M. tuberculosis challengeCharacterization of rBCGEffects of rBCG priming and viral boosting on protection against TB challenge

DiscussionAcknowledgementsReferences