Immunogenicity of a DNA and Recombinant Protein Vaccine Combining LipL32 and Loa22 for Leptospirosis Using Chitosan as a Delivery System
Immunogenicity of a DNA and Recombinant Protein Vaccine Combining LipL32 and Loa22 for Leptospirosis Using Chitosan as a Delivery System
Journal of Microbiology and Biotechnology. 2015. Apr, 25(4): 526-536
Copyright © 2015, The Korean Society For Microbiology And Biotechnology
  • Received : August 05, 2014
  • Accepted : October 27, 2014
  • Published : April 28, 2015
Export by style
Cited by
About the Authors
Supawadee Umthong
Interdisciplinary Program in Medical Microbiology, Graduate School, Chulalongkorn University, Bangkok 10330, Thailand
Arun Buaklin
Chulalongkorn University Vaccine Research Center, Chulalongkorn University, Bangkok 10330, Thailand
Alain Jacquet
Department of Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok 10330, Thailand
Noppadol Sangjun
Armed Forces Research Institute of Medical Sciences, Bangkok 10400, Thailand
Ruthairat Kerdkaew
Interdisciplinary Program in Medical Microbiology, Graduate School, Chulalongkorn University, Bangkok 10330, Thailand
Kanitha Patarakul
Chulalongkorn University Vaccine Research Center, Chulalongkorn University, Bangkok 10330, Thailand
Tanapat Palaga
Department of Microbiology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand

Leptospirosis is a worldwide zoonotic disease caused by pathogenic Leptospira , a genus of which more than 250 serovars have been identified. Commercial bacterin vaccines are limited in that they lack both cross-protection against heterologous serovars and long-term protection. This study investigated in mice the immunogenicity of an anti-leptospirosis vaccine, using the outer membrane proteins LipL32 and Loa22 as antigens. The immunogenicity of this vaccine formulation was compared with those induced by vaccines based on LipL32 or Loa22 alone. A DNA-encapsulated chitosan nanoparticle was used for in vivo DNA delivery. Using a unique DNA plasmid expressing both lipL32 and loa22 for vaccination, higher antibody responses were induced than when combining plasmids harboring each gene separately. Therefore, this formulation was used to test the immunogenicity when administered by a heterologous prime (DNA)-boost (protein) immunization regimen. The specific antibody responses against LipL32 (total IgG and IgG1) and Loa22 (IgG1) were higher in mice receiving two antigens in combination than in those vaccinated with a single antigen alone. Although no significant difference in splenic CD4 + T cell proliferation was observed among all groups of vaccinated mice, splenocytes from mice vaccinated with two antigens exhibited higher interferon-γ and IL-2 production than when using single antigens alone upon in vitro restimulation. Taken together, the immunogenicity induced by LipL32 and Loa22 antigens in a heterologous primeboost immunization regimen using chitosan as a DNA delivery system induces higher immune response, and may be useful for developing a better vaccine for leptospirosis.
Leptospirosis is a worldwide zoonotic disease [4] caused by pathogenic Leptospira , which are classified into more than 250 serovars [2] . Rodents are the major carriers of the disease. Infections in humans often result from direct contact with the urine of infected animals or indirect contact with a urine-contaminated environment [25] . The clinical manifestations may range from a mild flu-like illness to a severe and fatal disease [20 , 25] . The humoral immune response is considered to be the major protective immune response against Leptospira infection, possibly through antibody neutralization or complement-mediated killing [2] . The transfer of monoclonal anti-lipopolysaccharide (LPS) antibodies to newborn guinea pigs provides passive protection [20] . In contrast, a Th1 response is also essential as a protective response against L. borgpetersenii serovar Hardjo infection in cattle [5 , 28] .
Whole-cell leptospirosis vaccines have been used to protect against Leptospira infection [5] . Nonetheless, they still have limitations such as the inability to induce crossprotection among pathogenic serovars and the generation of short-term immunity. These shortcomings prevent them from being effective and reliable vaccines for different geographic areas that have variations in serovar distribution [18 , 19] . Subunit vaccines, including recombinant protein and DNA vaccines, have been developed to overcome the lack of cross-serovar protection in whole-cell vaccines.
LipL32 and Loa22 represent candidate antigens for subunit vaccine development against Leptospira . LipL32 is the most abundant constituent of the L. interrogans serovar Lai outer-membrane protein [10] and is shared by pathogenic Leptospira genomospecies, but is not present in saprophytic genomospecies [29] . This protein can be recognized by patient sera and expressed both in vitro and in vivo during infection [12 , 13] . However, the immune response induced by LipL32 vaccine only confers partial protection with particular formulations, including DNA vaccines, recombinant BCG vaccines, and adenoviral-mediated vaccines [6 , 7 , 33] . Furthermore, recently emerging evidence suggests that LipL32 locates at the sub-surface region and may not be surface-exposed [31] . Thus, a LipL32-based vaccine may be insufficient to elicit complete protection. Loa22 (also known as Lp0222) is a surface-exposed outer membrane protein containing an OmpA domain [33] . It is expressed during both acute and chronic infections [7] and can be recognized by the sera of human patients [21] . Loa22 is considered a virulence factor in the hamster model of infection [30] . Partial protection by Loa22 vaccine against Leptospira in a hamster model has also been reported [26] . A hamster model of complete protection against a heterologous challenge using a different serovar from that used for vaccination (a live-attenuated LPS mutant of L. interrogans serovar Manilae) revealed that Loa22 is a potential antigen recognized by the serum of immunized protected animals [32] .
DNA plasmids can directly transfect animal cells, and the corresponding proteins can be synthesized and expressed in vivo [34] . Moreover, DNA vaccines carrying genes that encode for candidate antigens have the ability to elicit both humoral and cell-mediated immune responses against the encoded antigens [3 , 38] . The immunogenicity of DNA vaccines, however, is poor in humans, and effective vaccine delivery systems are necessary. Chitosan (CS), a biopolymer containing a polycationic charge from an amine group, is a promising DNA vaccine delivery system. The polycationic chitosan makes it possible to encapsulate negatively charged molecules such as DNA. Chitosan-DNA complexes can be condensed into nanoparticles. This formulation efficiently promotes DNA plasmid uptake into cells [24] .
Using leptospiral outer membrane proteins OmpL1 and LipL41 as antigens, it was shown that two antigens exhibited synergistic immunoprotection [14] . This study prompted us to investigate the immunogenicity induced by a single DNA vaccine encoding lipL32 and loa22 and formulated with chitosan. We evaluated whether the combination of LipL32 and Loa22 generates more robust humoral and cellular immune responses compared with the use of each antigen alone.
Materials and Methods
- Animals and Bacterial Strains
Female BALB/c mice (8 weeks old) were purchased from the National Laboratory Animal Centre, Mahidol University, Thailand. They were housed at the Department of Pathology, Faculty of Medicine, Chulalongkorn University, Thailand. All procedures involving manipulations of animals in this project were approved by the Chulalongkorn University Animal Ethics Committee (review protocol No. 02/55). E. coli strains DH5α and BL21 pLysS (Novagen, USA) were cultivated in Luria-Bertani (LB) broth at 37℃ under appropriate shaking conditions or on LB agar in an incubator at 37℃ with appropriate antibiotics.
- Cell Culture
The human embryonic kidney 293T (HEK293T) cell line (ATCC No. CRL-1573-LGC) was cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin at 37℃ in a 5% CO 2 incubator.
- Construction and Preparation of the Plasmids and Recombinant LipL32 and Loa22 Proteins
The genomic DNA of L. interrogans serovar Pomona was used as a template for the PCR amplification of the full-length lipL32 and loa22 genes encoding the signal sequence and mature protein, from amino acids 1 to 272 for LipL32 and amino acids 1 to 195 for Loa22, respectively. Purified PCR products of lipL32 and loa22 were cloned into pVITRO1-neo-mcs plasmids (InvivoGen, USA) to obtain pVITRO- lipL32 ( lipL32 in MCS1), pVITRO- loa22 ( loa22 in MCS2), and pVITRO- lipL32 - loa22 ( lipL32 in MCS1 and loa22 in MCS2). The sequences of lipL32 and loa22 were confirmed by a commercial sequencing service (First BASE Laboratories, Malaysia). The large-scale preparation of plasmids (pVITRO, pVITRO- lipL32 , pVITRO- loa22 and pVITRO- lipL32 - loa22 ) was performed using endotoxin-free giga Qiagen kits (Qiagen, Germany) according to the manufacturer’s protocol. The plasmid pRSET C- lipL32 was prepared as previously described [13] except using genomic DNA of L. interrogans serovar Pomona as a template. The plasmid p489A4 containing gene loa22 , whose sequence is based on serovar Lai, was a kind gift from Dr. Mathieu Picardeau (Department of Microbiology, Institut Pasteur, Paris, France). Only one amino acid of Loa22 at position 106 is different; that is, G in serovar Pomona and R in serovar Lai. The recombinant LipL32 (rLipL32) and recombinant Loa22 (rLoa22) proteins were produced in E. coli BL21 (DE3) pLysS harboring plasmid pRSETC- lipl32 or p489A4- loa22 by IPTG induction. Both proteins were purified using Hi-Trap chelating columns (Amersham Biosciences, UK) for the first step of purification, followed by HiTrap butyl FF hydrophobic columns (Amersham Biosciences) with FPLC purification systems (GE Healthcare, UK).
- Encapsulation of DNA Vaccine with Chitosan
Acid-soluble low-molecular-weight chitosan from shrimp (average molecular mass 30 kDa and deacetylation rate of 80-85%) was obtained as a kind gift from Dr. Supasorn Wanichwecharungruang (Department of Chemistry, Faculty of Science, Chulalongkorn University). The chitosan was dissolved in 1% acetic acid and adjusted to a pH of 5.5 with NaOH. Chitosan solutions were diluted with 5 mM NaOAc until they reached the desired N/P ratio in the range of 1:1 to 20:1. The DNA plasmid was diluted in 25 mM Na 2 SO 4 and mixed with 100 µl of chitosan with an equal volume of DNA. Nanoparticles were prepared by a complex coacervation technique described by Mao et al . [24] . Briefly, after heating the DNA and chitosan solution at 55℃ for 5 min, equal volumes of the chitosan and the DNA solutions were mixed and immediately vortexed at maximum speed for 60 sec. The nanoparticles containing solution was used immediately after preparation for transfection of the HEK293T cell line for 5 days.
- Transfection and Detection by Western Blot and Immunofluorescence
The HEK293T cell line was transfected with plasmid pVITRO, pVITRO- lipL32 , pVITRO- loa22 , or pVITRO- lipL32 - loa22 at the molar concentration of 0.2 pmol of each plasmid in a 24-well plate using the X-treme GENE HP DNA transfection reagent (Roche Applied Science, USA) according to the manufacturer’s protocol. After 48 h of transfection, the culture supernatant was collected and the protein lysates were harvested from transfected cells. Proteins were evaluated by western blot analysis using mouse anti-LipL32 and mouse anti-Loa22 polyclonal antibodies. To detect LipL32 and Loa22 expression in the same cell, CS/DNA at an amine/phosphate (N/P) ratio of 20:1 of CS/pVITRO- lipL32 - loa22 was transfected into HEK293T cells and immunofluorescent staining was performed. The following antibodies were used: rabbit anti-LipL32 antibody (1:100), mouse anti-Loa22 antibody (1:100), donkey anti-mouse Alexa Fluor488 F(ab’) 2 fragments (1:200), and anti-rabbit Alexa fluor555 F(ab’) 2 fragments (1:200) (Cell Signaling Technology, USA). Cells were directly observed under a fluorescent microscope (Olympus, Japan) using anti-fade mounting media.
- Gel Retarding Analysis
To monitor the encapsulation efficacy, naked pMax-GFP plasmid (Amaxa, USA), a model plasmid, and CS/pMax-GFP nanoparticles prepared at N/P ratios of 1:1, 2:1, 4;1, 8:1, 10:1, and 20:1 were loaded onto 1% agarose gels in 1× TAE buffer. Naked plasmid pMax-GFP was used as a control to compare the efficacy of chitosan in encapsulating DNA. The samples were run on the gel at 100 volts for 40 min.
- Scanning Electron Microscope (SEM) Analysis ofCS/lipl32-loa22DNA Particles
CS, CS/pVITRO, CS/pVITRO- lipL32 , CS/pVITRO- loa22 and CS/pVITRO- lipL32 - loa22 nanoparticles prepared by encapsulation at an N/P ratio of 20:1 were directly observed using SEM to assess their size and morphology. Ten microliters of nanoparticle suspension after encapsulation was placed on copper grids. The grids were air-dried, and the samples were visualized at 15 kV 50,000× setting (Faculty of Science, Chulalongkorn University). The size of the CS/DNA particles from SEM was calculated using ImageJ free software.
- Immunization
To compare the formulation of the DNA vaccine containing two ORFs within a single plasmid with the co-administration of lipL32 and loa22 in different plasmids, mice were immunized three times by intramuscular (i.m.) injection with 10 pmol of CS/pVITRO- lipL32 - loa22 or co-immunized with CS/pVITRO- lipL32 +CS/pVITRO- loa22 at 10 pmol each, on days 1, 30, and 44 ( Fig. 3 ). Sera were collected for the determination of antigen-specific antibody titers, on days 29 37 and 51. To compare the immune responses generated by the combined LipL32-Loa22 vaccine with the LipL32 or Loa22 vaccines, mice were immunized i.m. with 10 pmol of CS/pVITRO- lipL32 , CS-pVITRO- loa22 , or CS/pVITRO- lipL32 - loa22 plasmid on day 1 (prime), and then boosted twice with the recombinant proteins by subcutaneous injection of rLipL32, rLoa22, or both at the concentration of 0.5 µM (11.63 µg of rLoa22, 15.63 µg of rLipL32) in AddaVax (Invivogen, USA) as an adjuvant on days 30 and 44 ( Fig. 4 A). Sera were collected seven days after the second boost. Spleens were isolated for lymphoproliferation assays and the detection of cytokines.
- Determination of Antigen-Specific Total IgG, IgG1, and IgG2a in Sera by ELISA
Blood was collected by using the retro-orbital bleeding method after anesthesia with AERRANE (isoflurane; Baxter, USA). Ninety six-well immunoplates were coated with 100 µl (5 µg/ml) of purified LipL32, Loa22, or coating buffer (0.1 M Na 2 CO 3 , 0.1 M NaHCO 3 , pH 9.5). After washing and blocking, diluted mouse sera in blocking buffer were added to each well at 100 µl/well and the plates were incubated at room temperature for 1 h. Horseradish-peroxidase-conjugated anti-mouse IgG, IgG1, or IgG2a (Invitrogen, USA) was then added. The substrate 3,3’,5, 5’-tetramethylbenzidine (Sigma-Aldrich, USA) was added and incubated for 20 min at room temperature, after which time the reaction was stopped with 2 M sulfuric acid. The absorbance was read at 450 nm with a microplate reader (Anthos Labtec Instruments, UK).
- Lymphocyte Proliferation Assay
Spleens from immunized mice were isolated, and splenocytes were harvested after red blood cell lysis. The resulting cells were resuspended at a final concentration of 1 × 10 6 viable cells/ml in RPMI medium containing 2-mercaptoethanol. A carboxyfluorescein diacetate succinimidyl ester (CFSE) proliferation assay was then performed using a Cell Trace CFSE Cell Proliferation kit (Molecular Probes, USA) according to the manufacturer’s protocol. In the final washing step, cells were plated in 24-well plates at 1 × 10 6 cells/well and were stimulated with 20 µg/ml rLipL32, 20 µg/ml rLoa22 or 5 µg/ml Concanavalin A (ConA; Sigma-Aldrich) as a positive control. After 72 h of stimulation, cells were collected, washed, and stained using mouse anti-CD4-ECD (Beckman Coulter Inc., USA) and were subjected to flow cytometric analysis (FC500; Becman Coulter). The data were analyzed by FlowJo software (Tree Star, Inc., USA).
- Measurement of Cytokine by ELISA
To measure cytokine responses upon restimulation in vitro , splenocytes (2 × 10 6 cells/ml) were seeded in a 12-well plate in complete RPMI medium and were stimulated with 20 µg/ml rLipL32, 20 µg/ml rLoa22, 5 µg/ml ConA (positive control), or medium alone (negative control) for 72 h at 37℃ in a 5% CO 2 incubator. The culture supernatants were then collected and Th1 (IFN γ , IL-2) and Th2 cytokine (IL-4, IL-10) responses were then measured using a cytokine ELISA MAX kit (Biolegend, USA) according to the manufacturer’s protocol.
- Statistical Analysis
All statistical analyses were performed using GraphPad Prism 5 software. The results are expressed as the mean and standard deviation (SD). One-way analysis of variance with Turkey-Kremer’s post-hoc test was used to determine the statistical significance ( p < 0.05). All experiments were performed in duplicate and with five biological samples per group.
- Expression of LipL32 and Loa22 in HEK293T Cell Line
The HEK293T cell line was transfected with pVITRO, pVITRO- lipL32 , pVITRO- loa22 , or pVITRO- lipL32 - loa22 . The expression of LipL32 and Loa22 could be detected both in the total cell lysate and the culture supernatant ( Figs.1 A and 1 B). In both cases, a single major band corresponding to a predicted molecular mass of approximately 34 kDa for LipL32 was detected, whereas a single major band corresponding to Loa22 had a molecular mass of approximately 22 kDa. The absence of bands corresponding to β-actin in the culture supernatant ( Fig. 1 B) indicated that LipL32 and Loa22 were secreted extracellularly rather than released as a result of cell necrosis. Comparison of the expression levels of LipL32 and Loa22 in pVITRO- lipL32 - loa22 -transfected cells with pVITRO- lipL32 and pVITRO- loa22 co-transfected cells revealed that pVITRO- lipL32 - loa22 , which carries both inserts in the same plasmid, resulted in a slightly higher expression of both proteins than the co-transfection method, especially for the level of Loa22 ( Fig. 1 B). Taken together, these results indicated that the cloned genes are correctly expressed both intracellularly and as secreted proteins. Using pVITRO- lipL32 - loa22 resulted in slightly higher expression of LipL32 and Loa22 than the co-transfection of each plasmid at equimolar concentrations.
PPT Slide
Lager Image
Expression of LipL32 and Loa22 in HEK293T, and CS/DNA plasmid encapsulation. (A-B) The HEK293T cell line was transfected with the indicated plasmid or plasmid combinations for 48 h. Cell lysates (A) or culture supernatants (B) were subjected to western blot analysis to detect LipL32 and Loa22 proteins. (C) The encapsulation efficacy of the CS/DNA plasmid was studied using an agarose gel retardation assay using pMax-GFP as a model plasmid. Lane M is the marker. Lanes 1–6 consist of nanoparticles prepared at N/P ratios of 1:1, 2:1, 4:1, 8:1, 10:1, and 20:1, respectively; Lane 7 consists of pMax-GFP without encapsulation. (D-I) Immunofluorescent staining of Lipl32 and Loa22 in the HEK293T cell line transfected with CS/pVITRO-lipl32-loa22. LipL32 (red), Loa22 (green), and DAPI (blue) are shown after transfection with CS/pVITRO-lipl32-loa22 particles at an N/P ratio of 20:1 for 5 days (D-F) or transfection with the X-treme GENE HP DNA transfection reagent (G-I).
- Encapsulation of CS/DNA Plasmids and Transfection Using the CS/DNA Complex
To use chitosan as a DNA vaccine delivery system, the efficiency of DNA encapsulation by CS at different amine/phosphate (N/P) ratios ranging from 1:1 to 20:1 was determined by agarose gel electrophoresis. In this study, pMax-GFP was used as a model plasmid. The encapsulated plasmid, equivalent to 100 ng of DNA, was directly loaded onto a 1% agarose gel and electrophoresed. The naked pMax-GFP DNA migrated at the expected size, whereas the DNA encapsulated in chitosan nanoparticles remained in the loading wells when the N/P ratio was at 8:1 or higher ( Fig. 1 C). When using this CS/DNA complex to transfect HEK293T cells, the highest transfection efficiency of CS/DNA was observed at the N/P ratio of 20:1 (Fig.S2). Based on this result, the efficacy of transfection using CS/DNA particles of pVITRO- lipL32 - loa22 at the N/P ratio of 20:1 was determined by immunofluorescent staining in the HEK293T cell line. The expression of LipL32 and Loa22 could be detected in the same cells ( Fig. 1 D), suggesting that both proteins are expressed in the same cells using CS as a delivery system.
The morphology of the CS, CS/pVITRO, CS/pVITRO- lipL32 , CS/pVITRO- loa22 , CS/pVITRO- lipL32 - loa22 , and CS/pVITRO- lipL32 +CS/pVITRO- loa22 nanoparticles was analyzed by SEM. Particle sizes were in the nanometer range (approximately 100 nm). Different shapes and sizes were observed but most particles exhibited spherical morphology ( Figs. 2 A- 2 F). The average sizes of CS/DNA particles in all conditions were approximately 100-125 nm. There was no significant difference in size among CS/DNA particles ( Fig. 2 G). Therefore, the CS/DNA plasmid complex formulated in this study is referred to as a nanoparticle in this study.
PPT Slide
Lager Image
Size and morphology of CS/DNA plasmid nanoparticles. (A-F) Scanning electron micrographs of the CS/DNA nanoparticles. The encapsulation was carried out at an N/P ratio of 20:1. A, CS only; B, CS/pVITRO; C, CS/pVITRO-lipL32; D, CS/pVITRO-loa22; E, CS/pVITRO-lipL32-loa22; and F, CS/pVITRO-lipL32+pVITRO/loa22 co-encapsulation. (G) The mean diameter of CS/DNA plasmid nanoparticles was analyzed by the ImageJ program from the data obtained by scanning electron microscopy.
- Humoral Immune Response Induced by Vaccination of CS/DNA Nanoparticles
The differences in immunogenicity between two lipL32 -and loa22 -based DNA formulations were evaluated. One formula (CS/pVITRO- lipL32 - loa22 ) contained the two cassettes, lipL32 and loa22 , on the same plasmid, whereas another (CS/pVITRO- lipL32 +CS/pVITRO- loa22 ) entailed the co-administration of CS/pVITRO- lipL32 and CS/pVITRO- loa22 . Specific total sera IgG against LipL32 was not significantly different between the two formulations ( Fig. 3 A). However, total antigen-specific sera IgG against Loa22 in mice immunized with CS/pVITRO- lipL32 - loa22 was higher than in mice co-immunized with CS/pVITRO- lipL32 and CS/pVITRO- loa22 ( Fig. 3 B). The in vivo humoral immune responses correlated well with the level of Loa22 and LipL32 expression in vitro ( Figs.1 A and 1 B). This result possibly reflects the expression level of Loa22 from pVITRO- lipL32 - loa22 -transfected cells, which was higher than in cells co-transfected with CS/pVITRO- lipL32 +CS/pVITRO- loa22 . The CS/pVITRO- lipL32 - loa22 plasmid was therefore selected for further experiments.
PPT Slide
Lager Image
The humoral immune responses against vaccination using CS/DNA plasmid and prime-boost vaccination using CS/DNA plasmid nanoparticles and recombinant LipL32 and Loa22. (A-B) Mice were divided into three groups and immunized with the following vaccines; CS/pVITRO (white bar), CS/pVITRO-lipL32-loa22 (grey bar), and CS/pVITRO-lipL32+CS/pVITRO-loa22 (black bar). Five mice per group were used. Each group was immunized three times with similar doses of 10 pmol of DNA plasmids. The endpoint antibody titers of sera collected on day 29 (four weeks after prime immunization), day 37 (one week after boost 1 vaccination), and day 51 (one week after boost 2 immunization) are shown. The titers of specific total IgG against LipL32 (A) and Loa22 (B) are shown. Data are reported as geometric means ± SD. *p < 0.05 indicates statistical significance. (C-D) Mice were immunized using a heterologous prime-boost regimen with the following vaccines: CS/pVITRO+PBS+PBS (white bar), CS/pVITRO-lipL32+rLipL32 (grey bar), or CS/pVITRO-lipL32-loa22+rLipL32-rLoa22 (black bar). The antibody titers of sera collected seven days after boost 2 are displayed. Each group contained five mice. Mice were immunized with 10 pmol of DNA plasmids and 0.5 µmol of rLipL32, rLoa22, or both. Sera titers of antigen-specific total IgG, IgG1, and IgG2a against LipL32 (C) and Loa22 (D) are shown. Data are shown as geometric means ± SD. *p < 0.05, **p < 0.01 indicate statistical significance when compared with the CS/pVITRO control group or between indicated groups.
- Humoral Immune Responses Induced by Heterologous Prime-Boost Immunization
Using two different formulations of vaccine for priming and boost, respectively, has been shown to yield a higher magnitude of immune response, especially for DNA vaccines [22] . Furthermore, our preliminary experiments found that intramuscular administrations of chitosan-conjugated DNA plasmids induced low humoral immune responses, as measured by low antibody titer (data not shown). Therefore, a heterologous prime-boost regimen containing DNA and recombinant protein vaccines was used for this study. The prime immunization was performed via an intramuscular injection with CS/DNA (CS/pVITRO- lipL32 , CS/pVITRO- loa22 , or CS/pVITRO- lipL32 - loa22 ), followed by two subcutaneous injections for booster immunizations consisting of the recombinant protein in the adjuvant AddaVax( Fig. 4 A). The adjuvant AddaVax drives a balanced Th1/Th2 response when compared with alum [8] . Mice vaccinated with the two antigens exhibited significantly higher sera levels of antigen-specific total IgG and IgG1 than mice receiving only the single LipL32 antigen. However, the total anti-LipL32 IgG and IgG2a levels were not significantly different between the two groups ( Fig. 3 C). For antibody production against Loa22, mice vaccinated with two antigens showed significantly higher levels of anti-Loa22 IgG1 antibodies when compared with those that received Loa22 alone ( Fig. 3 D). The immune response based on produced antibody isotypes appears to be Th2-biased in mice receiving two antigens, as Loa22- and LipL32-specific IgG1 (Th2) levels were higher than those receiving a single antigen ( Figs. 3 C and 3 D). Although we did not perform a side-by-side comparison experiments of vaccination regimens using DNA vaccine alone and heterologous prime boost, the data of Figs. 3 A- 3 B and 3 C- 3 D clearly demonstrated that DNA vaccine alone yielded more than 1,000-fold less specific antibody titer than the heterologous prime boost vaccination.
PPT Slide
Lager Image
Cellular immune response in an in vitro restimulation assay. (A) CFSE-labeled splenocytes from immunized mice were restimulated with recombinant LipL32 or Loa22 (20 µg/ml) for 48 h. The CD4+ T cell population was gated and analyzed for CFSE staining by flow cytometry. The data were analyzed by FlowJo and shown as a proliferation index. The proliferation index is defined as the number of divisions that took place divided by the number of cells of the original population that went into division. Data are reported as the means ± SD for three animals per group. *p < 0.05 indicates statistical significance when compared with the pVITRO-PBS-PBS control immunization group. (B-C) The cytokine responses from splenocytes of indicated immunized mice that were stimulated as described above using recombinant LipL32 (B) or Loa22 (C). The levels of secreted cytokines were measured by ELISA. The experiment was performed in duplicate. The results represent the data from three mice per group. p < 0.05 is considered statistically significant.
- T Cell Responses to LipL32, Loa22, and LipL32-Loa22 in anIn VitroRestimulation
T cell responses to an in vitro restimulation were measured to test whether the two antigens help to improve immunogenicity by better stimulating T cells responses better than LipL32 or Loa22 antigen alone. CFSE was used to monitor cell proliferation after restimulation with recombinant antigens, and gated CD4 + T cells were analyzed for cell proliferation by flow cytometry. Mice immunized with the combination of LipL32 and Loa22 antigens or vaccinated with a single LipL32 or Loa22 antigen showed no significant differences in CD4 + T cell proliferation upon restimulation with 20 µg/ml of LipL32 or Loa22 ( Fig. 4 A).
The cytokine profiles in response to restimulation with LipL32 or Loa22 showed significant increases in all cytokines detected in this study when compared with CS/pVITRO-PBS-PBS control groups. The level of IFN-γ in mice immunized with two antigens was significantly higher than in mice that received only antigen LipL32 ( Fig. 4 B). The levels of Th2 cytokines IL-4 and IL-10 produced by stimulated splenocytes were extremely low, whereas the vaccination with two antigens resulted in stronger anti-Loa22 Th1 responses (IFN-γ and IL-2) than when immunized with Loa22 antigen alone ( Fig. 4 C). Nevertheless, no significant differences were detected in the Th2 cytokine response (IL-4, IL-10) between groups receiving two antigens or one.
Previous investigations showed that LipL32 is a conserved outer membrane protein among pathogenic serovars [15] . The suitability of LipL32 as a good vaccine candidate is still controversial because its protective efficacy is not consistent in the literature [27] . Therefore, the use of LipL32 antigen alone may be insufficient to achieve complete protection. The objective of this study was to therefore test whether a combination of LipL32 with another antigen could promote better humoral and cell-mediated immune responses than when using LipL32 alone.
The antigen used in combination with LipL32 in this study was Loa22, a protein in the OmpA lipoprotein family [21] . Loa22 possesses many characteristics of a good vaccine candidate: (i) it is a surface protein [30] ; (ii) it is confirmed by a mutagenesis study to be a known virulence factor [32] ; (iii) it is expressed during infection and can be recognized by patient sera [29] ; and (iv) it provides partial protection in challenged hamsters [39] . The characteristics of LipL32 and Loa22 could induce a protective response against broad serovars because both of them are conserved among pathogenic serovars of leptospires. A combination of these two antigens may improve the efficacy of the humoral and cellular immune responses and protect against infection by heterologous serovars. A study exploring the combination of LipL32 and Loa22 as a vaccine candidate has never been undertaken before.
Therefore, in this study, a combination of LipL32 and Loa22 was investigated as a candidate vaccine by using biopolymer chitosan as a delivery system for lipL32 and loa22 DNA vaccines. Furthermore, since our unpublished data indicated that DNA vaccines of either lipL32 and loa22 alone induced antibody response poorly, the heterologous prime-boost immunization regimen was carried out using DNA vaccine for priming and recombinant protein vaccine as a booster.
HEK293T cells transfected with DNA plasmids containing full-length lipL32 or loa22 or both expressed the corresponding proteins intracellularly and extracellularly. Both proteins were secreted from the transfected mammalian cells, although the full-length lipL32 and loa22 contained bacterial signal sequences. The secretion of both proteins indicates that the bacterial or native signal peptides of either LipL32 or Loa22 can be recognized in higher eukaryotes such as mammals. This result is consistent with a previous report showing that bacterial signal peptides are functional for direct secretion of proteins in mammalian cells [16] . In the presence of a signal sequence, our DNA vaccines should be subsequently processed not only as endogenous antigens but also as exogenous antigens and therefore induce better antibody production in comparison with those without a signal sequence.
To construct a recombinant DNA plasmid of lipL32 and loa22 , the mammalian expression vector pVITRO1, which possesses two multiple cloning sites, was used. This plasmid made it possible to insert both lipL32 and loa22 into the same plasmid. Both genes were expressed under the control of different promoters that displayed strong activity and yielded similar levels of expression. Therefore, the difference in the levels of expression of both antigens was not due to differences in promoter activity. The encapsulation of the DNA vaccine by chitosan demonstrated an optimal N/P ratio of 20:1, and the efficacy of chitosan in delivering of DNA vaccine of lipL32 and loa22 was consistent with several studies that demonstrated the ability of chitosan to promote DNA uptake by cells [24] and induce an immune response in response to a DNA vaccine [9 , 35] .
The two formulations of the DNA vaccine were compared in terms of in vitro protein expression and their efficacy in inducing in vivo humoral immune responses. The first formulation was CS/pVITRO- lipL32 - loa22 (two ORFs in one plasmid) and the other consisted of a CS/pVITRO- lipL32 +CS/pIVTRO- loa22 co-administration (two ORFs in two plasmids). There was no significant difference in LipL32 protein expression levels in vitro after transfection with the two formulations, in contrast to Loa22 expression levels, in which the transfection of CS/pVITRO- lipL32 - loa22 (two ORFs in one plasmid) resulted in higher levels in both cell lysates and culture supernatants than the cotransfection method. This result was consistent with the in vivo humoral responses observed after vaccination with equimolar concentrations of the DNA vaccines. The levels of total Loa22-specific IgG were higher in mice immunized with CS/pVITRO- lipL32 - loa22 than those receiving co-immunization by the two plasmids. In addition to differences in protein antigen expression, other mechanisms may also contribute to the observed diminished effect in the two-plasmid formulation. The use of CS/pVITRO- lipL32 - loa22 not only results in increased antigen expression and a greater antibody response, but it is more practical owing to the reduced preparation time and cost.
Many vaccine studies in other infectious diseases such as HIV [17 , 36] , tuberculosis [23] , influenza [37] , and leptospirosis [11] have demonstrated that a heterologous prime-boost immunization using two different vaccine formulations is a promising effective strategy for stimulating both the humoral and cellular immune responses than either vaccine formulation alone. In leptospirosis, the humoral immune response is a major protective immunity against infection, but recent evidence also demonstrated the importance of the Th1 response and cell-mediated immune response [28 , 40] . Therefore, a balanced Th1/Th2 response may require to generate a full protective response. Because protein antigens delivered via a DNA vaccine are synthesized and subjected to antigen processing and presentation via the intracellular pathway, it is thought to induce better cell-mediated immune response, whereas protein subunit vaccines are superior in inducing humoral immune responses. Using a heterologous prime ( D NA vaccine)-boost (protein subunit vaccine) may provide better protective responses against leptospirosis. To test this hypothesis, we applied a heterologous prime-boost immunization regimen to promote both humoral and cellular immune responses.
In a heterologous prime-boost immunization using the CS/DNA vaccine and recombinant proteins, all the antibody titers elicited in the protein-boosted groups showed higher levels than the group immunized with only CS/pVITRO- lipL32 - loa22 DNA vaccine. This result indicates that a proteinboost strategy can improve the immunogenicity of DNA vaccines against both LipL32 and Loa22. Priming immunization with the CS/DNA vaccine and boosting twice with recombinant LipL32 and Loa22 proteins in AddaVax drove a Th2 immune response, as evidenced by the greater levels of IgG1 and total IgG compared with IgG2a. AddaVax is an adjuvant described by the manufacturer as a balanced inducer of Th1 and Th2 responses, which is superior to alum, which usually induces a Th2-biased response [8] . However, in this study, the humoral immune response still leaned towards a Th2 response, albeit with slightly higher IgG2a (Th1) levels than those of mice immunized with only DNA vaccines. Another factor that may explain the Th2 response seen in this study is the use of BALB/c inbred mice that are known to mount Th2-biased immune responses. Thus, this vaccination strategy could be valuable for inducing both Th1 and Th2 responses against both antigens.
The antibody responses, including total IgG, IgG1, and IgG2a levels against LipL32 and Loa22, were not different between mice receiving a single LipL32 or Loa22 antigen to mice receiving both LipL32 and Loa22. Likewise, the proliferation of CD4 + T cells upon in vitro restimulation was also the same in both groups, indicating that no changes in humoral and CD4 + T cell proliferative responses were found when the two antigens were used as vaccines.
Cytokine measurements of in vitro restimulated splenocytes revealed that all vaccinated groups showed significantly increased levels of Th1 (IFN-γ and IL-2) but not Th2 (IL-4 and IL-10) cytokines compared with the control groups. More importantly, splenocytes from mice receiving two antigens produced significantly more IFN-γ than those from mice receiving a single antigen. Therefore, the two-antigen vaccination appears to boost Th1 cytokine production better. Using intracellular cytokine staining, however, we could not detect any differences in the percentage of IFN-γ CD4 + T cells or in fluorescent intensities for these cytokines in CD4 + T cells between these two groups (data not shown). Therefore, it is possible that other cell subsets in the spleen may be responsible for the greater production of cytokines observed.
Combining LipL32 and Loa22 antigens did not provide a synergistic effect in stimulating humoral immune responses against LipL32 or Loa22, and the combination of both antigens did not act antagonistically against each other. We observed only a synergistic effect in an in vitro restimulation assay for Th1 cytokines. Various studies have reported that combining two antigens, for example, the ompL1 and lipL41 DNA vaccine, provides a synergistic effect in leptospirosis [14] . A vaccine against Mycobacterium tuberculosis suggests that there is a synergistic effect of combining three antigens [1] . Nevertheless, the result of no synergistic effect in the humoral immune response in our study may be explained by differences in the recombinant LipL32 and Loa22 proteins used. The LipL32 and Loa22 recombinant subunit vaccines were co-administered into mice as separate proteins in our study, whereas the aforementioned studies used fusion proteins that combined two or three antigens into one protein. Antibody and T cell response may react against the overlapping region of each antigen in the peptide. Therefore, this may promote a cross-reaction and provide a synergistic effect. In our study, however, the lipL32 and loa22 DNA plasmids were not constructed as a fusion gene and the LipL32 and Loa22 recombinant proteins did not form a fusion protein. Although the synergistic effect on the humoral immune response between LipL32 and Loa22 may not have been observed, a combination of the two antigens may promote better protective efficacy than using a LipL32 or Loa22 vaccine alone. The protective efficacy of LipL32 and Loa22 in combination needs to be studied further in pathogen-challenged animal models. To determine the precise synergistic protective effects of LipL32-Loa22 vaccine candidates and their protective efficiency, hamster or guinea pig models should be tested in the future.
This work was supported by the Higher Education Research Promotion and the National Research University Project of Thailand, Office of the Higher Education Commission (HR1164A2), the Special Task Force for Activating Research (STAR) from the Centenary Academic Development Project (Chulalongkorn University), and the National Research Council of Thailand (NRCT). The authors would like to acknowledge financial support from the Research, Development and Engineering (RD&E) fund through The National Nanotechnology Center (NANOTEC), the National Science and Technology Development Agency (NSTDA), Thailand (Project No. P-10-10454) to Chulalongkorn University. The equipment used in this study was purchased with funds from the Thai Government Stimulus Package2(TKK2555) under the Project for the Establishment of Innovative Food and Health Products and Agriculture. Arun Buaklin was sponsored by a post-doctoral fellowship (Rachadapisek Sompote Endowment Fund) from Chulalongkorn University Graduate School. Supawadee Umthong was supported by the Development and Promotion of Science and Technology Talents Project (DPST).
Aagaard C , Hoang T , Dietrich J , Cardona PJ , Izzo A , Dolganov G 2011 A multistage tuberculosis vaccine that confers efficient protection before and after exposure. Nat. Med. 17 189 - 194    DOI : 10.1038/nm.2285
Adler B , de la Pena Moctezuma A 2010 Leptospiraand leptospirosis. Vet. Microbiol. 140 287 - 296    DOI : 10.1016/j.vetmic.2009.03.012
Alpar HO , Bramwell VW 2002 Current status of DNA vaccines and their route of administration. Crit. Rev. Ther. Drug Carrier Syst. 19 307 - 383    DOI : 10.1615/CritRevTherDrugCarrierSyst.v19.i45.20
Bharti AR , Nally JE , Ricaldi JN , Matthias MA , Diaz MM , Lovett MA 2003 Leptospirosis: a zoonotic disease of global importance. Lancet Infect. Dis. 3 757 - 771    DOI : 10.1016/S1473-3099(03)00830-2
Bolin CA , Cassells JA , Zuerner RL , Trueba G 1991 Effect of vaccination with a monovalentLeptospira interrogansserovar Hardjo type Hardjo-Bovis vaccine on type Hardjo-Bovis infection of cattle. Am. J. Vet. Res. 52 1639 - 1643
Branger C , Chatrenet B , Gauvrit A , Aviat F , Aubert A , Bach JM , Andre-Fontaine G 2005 Protection againstLeptospira interroganssensu lato challenge by DNA immunization with the gene encoding hemolysin-associated protein 1. Infect. Immun. 73 4062 - 4069    DOI : 10.1128/IAI.73.7.4062-4069.2005
Branger C , Sonrier C , Chatrenet B , Klonjkowski B , Ruvoen-Clouet N , Aubert A 2001 Identification of the hemolysis-associated protein 1 as a cross-protective immunogen ofLeptospira interrogansby adenovirus-mediated vaccination. Infect. Immun. 69 6831 - 6838    DOI : 10.1128/IAI.69.11.6831-6838.2001
Coffman RL , Sher A , Seder RA 2010 Vaccine adjuvants: putting innate immunity to work. Immunity 33 492 - 503    DOI : 10.1016/j.immuni.2010.10.002
Cui Z , Mumper RJ 2001 Chitosan-based nanoparticles for topical genetic immunization. J. Control. Release 75 409 - 419    DOI : 10.1016/S0168-3659(01)00407-2
Cullen PA , Cordwell SJ , Bulach DM , Haake DA , Adler B 2002 Global analysis of outer membrane proteins fromLeptospira interrogansserovar Lai. Infect. Immun. 70 2311 - 2318    DOI : 10.1128/IAI.70.5.2311-2318.2002
Feng CY , Li QT , Zhang XY , Dong K , Hu BY , Guo XK 2009 Immune strategies using single-component LipL32 and multi-component recombinant LipL32-41-OmpL1 vaccines againstLeptospira. Braz. J. Med. Biol. Res. 42 796 - 803    DOI : 10.1590/S0100-879X2009005000013
Guerreiro H , Croda J , Flannery B , Mazel M , Matsunaga J , Galvao Reis M 2001 Leptospiral proteins recognized during the humoral immune response to leptospirosis in humans. Infect. Immun. 69 4958 - 4968    DOI : 10.1128/IAI.69.8.4958-4968.2001
Haake DA , Chao G , Zuerner RL , Barnett JK , Barnett D , Mazel M 2000 The leptospiral major outer membrane protein LipL32 is a lipoprotein expressed during mammalian infection. Infect. Immun. 68 2276 - 2285    DOI : 10.1128/IAI.68.4.2276-2285.2000
Haake DA , Mazel MK , McCoy AM , Milward F , Chao G , Matsunaga J , Wagar EA 1999 Leptospiral outer membrane proteins OmpL1 and LipL41 exhibit synergistic immunoprotection. Infect. Immun. 67 6572 - 6582
Haake DA , Suchard MA , Kelley MM , Dundoo M , Alt DP , Zuerner RL 2004 Molecular evolution and mosaicism of leptospiral outer membrane proteins involves horizontal DNA transfer. J. Bacteriol. 186 2818 - 2828    DOI : 10.1128/JB.186.9.2818-2828.2004
Hall J , Hazlewood GP , Surani MA , Hirst BH , Gilbert HJ 1990 Eukaryotic and prokaryotic signal peptides direct secretion of a bacterial endoglucanase by mammalian cells. J. Biol. Chem. 265 19996 - 19999
Harari A , Bart PA , Stohr W , Tapia G , Garcia M , Medjitna-Rais E 2008 An HIV-1 clade C DNA prime, NYVAC boost vaccine regimen induces reliable, polyfunctional, and long-lasting T cell responses. J. Exp. Med. 205 63 - 77    DOI : 10.1084/jem.20071331
Herrmann JL , Bakoss P , Korver H , Bulu AA , Bellenger E , Terpstra WJ 1994 A new serovar in the Grippotyphosa serogroup comprising leptospiral isolates from different regions. Int. J. Syst. Bacteriol. 44 362 - 364    DOI : 10.1099/00207713-44-2-362
Herrmann JL , Baril C , Bellenger E , Perolat P , Baranton G , Saint Girons I 1991 Genome conservation in isolates ofLeptospira interrogans. J. Bacteriol. 173 7582 - 7588
Jost BH , Adler B , Vinh T , Faine S 1986 A monoclonal antibody reacting with a determinant on leptospiral lipopolysaccharide protects guinea pigs against leptospirosis. J. Med. Microbiol. 22 269 - 275    DOI : 10.1099/00222615-22-3-269
Koizumi N , Watanabe H 2003 Molecular cloning and characterization of a novel leptospiral lipoprotein with OmpA domain. FEMS Microbiol. Lett. 226 215 - 219    DOI : 10.1016/S0378-1097(03)00619-0
Lu S 2009 Heterologous prime-boost vaccination. Curr. Opin. Immunol. 21 346 - 351    DOI : 10.1016/j.coi.2009.05.016
Magalhaes I , Sizemore DR , Ahmed RK , Mueller S , Wehlin L , Scanga C 2008 rBCG induces strong antigen-specific T cell responses in rhesus macaques in a prime-boost setting with an adenovirus 35 tuberculosis vaccine vector. PLoS One 3 e3790 -    DOI : 10.1371/journal.pone.0003790
Mao HQ , Roy K , Troung-Le VL , Janes KA , Lin KY , Wang Y 2001 Chitosan-DNA nanoparticles as gene carriers: synthesis, characterization and transfection efficiency. J. Control. Release 70 399 - 421    DOI : 10.1016/S0168-3659(00)00361-8
McBride AJ , Athanazio DA , Reis MG , Ko AI 2005 Leptospirosis. Curr. Opin. Infect. Dis. 18 376 - 386    DOI : 10.1097/01.qco.0000178824.05715.2c
Monahan AM , Callanan JJ , Nally JE 2008 Proteomic analysis ofLeptospira interrogansshed in urine of chronically infected hosts. Infect. Immun. 76 4952 - 4958    DOI : 10.1128/IAI.00511-08
Murray GL 2013 The lipoprotein LipL32, an enigma of leptospiral biology. Vet. Microbiol. 162 305 - 314    DOI : 10.1016/j.vetmic.2012.11.005
Naiman BM , Alt D , Bolin CA , Zuerner R , Baldwin CL 2001 Protective killedLeptospira borgpeterseniivaccine induces potent Th1 immunity comprising responses by CD4 and gammadelta T lymphocytes. Infect. Immun. 69 7550 - 7558    DOI : 10.1128/IAI.69.12.7550-7558.2001
Nally JE , Whitelegge JP , Bassilian S , Blanco DR , Lovett MA 2007 Characterization of the outer membrane proteome ofLeptospira interrogansexpressed during acute lethal infection. Infect. Immun. 75 766 - 773    DOI : 10.1128/IAI.00741-06
Picardeau M , Bulach DM , Bouchier C , Zuerner RL , Zidane N , Wilson PJ 2008 Genome sequence of the saprophyteLeptospira biflexaprovides insights into the evolution ofLeptospiraand the pathogenesis of leptospirosis. PLoS One 3 e1607 -    DOI : 10.1371/journal.pone.0001607
Pinne M , Haake DA 2013 LipL32 is a subsurface lipoprotein ofLeptospira interrogans: presentation of new data and reevaluation of previous studies. PLoS One 8 e51025 -    DOI : 10.1371/journal.pone.0051025
Ristow P , Bourhy P , da Cruz McBride FW , Figueira CP , Huerre M , Ave P 2007 The OmpA-like protein Loa22 is essential for leptospiral virulence. PLoS Pathog. 3 e97 -    DOI : 10.1371/journal.ppat.0030097
Seixas FK , da Silva EF , Hartwig DD , Cerqueira GM , Amaral M , Fagundes MQ 2007 RecombinantMycobacterium bovisBCG expressing the LipL32 antigen ofLeptospira interrogansprotects hamsters from challenge. Vaccine 26 88 - 95    DOI : 10.1016/j.vaccine.2007.10.052
Srikram A , Zhang K , Bartpho T , Lo M , Hoke DE , Sermswan RW 2011 Cross-protective immunity against leptospirosis elicited by a live, attenuated lipopolysaccharide mutant. J. Infect. Dis. 203 870 - 879    DOI : 10.1093/infdis/jiq127
van der Lubben IM , Verhoef JC , Borchard G , Junginger HE 2001 Chitosan for mucosal vaccination. Adv. Drug Deliv. Rev. 52 139 - 144    DOI : 10.1016/S0169-409X(01)00197-1
Wang S , Pal R , Mascola JR , Chou TH , Mboudjeka I , Shen S 2006 Polyvalent HIV-1 Env vaccine formulations delivered by the DNA priming plus protein boosting approach are effective in generating neutralizing antibodies against primary human immunodeficiency virus type 1 isolates from subtypes A, B, C, D and E. Virology 350 34 - 47    DOI : 10.1016/j.virol.2006.02.032
Wang S , Parker C , Taaffe J , Solorzano A , Garcia-Sastre A , Lu S 2008 Heterologous HA DNA vaccine prime - inactivated influenza vaccine boost is more effective than using DNA or inactivated vaccine alone in eliciting antibody responses against H1 or H3 serotype influenza viruses. Vaccine 26 3626 - 3633    DOI : 10.1016/j.vaccine.2008.04.073
Wolff JA , Malone RW , Williams P , Chong W , Acsadi G , Jani A , Felgner PL 1990 Direct gene transfer into mouse musclein vivo. Science 247 1465 - 1468    DOI : 10.1126/science.1690918
Yan W , Faisal SM , McDonough SP , Chang CF , Pan MJ , Akey B , Chang YF 2010 Identification and characterization of OmpA-like proteins as novel vaccine candidates for leptospirosis. Vaccine 28 2277 - 2283    DOI : 10.1016/j.vaccine.2009.12.071
Zuerner RL , Alt DP , Palmer MV , Thacker TC , Olsen SC 2011 ALeptospira borgpeterseniiserovar Hardjo vaccine induces a Th1 response, activates NK cells, and reduces renal colonization. Clin. Vaccine Immunol. 18 684 - 691    DOI : 10.1128/CVI.00288-10