|Home | About | Journals | Submit | Contact Us | Français|
There is a need for a more efficient vaccine against the bacterium Y. pestis, the agent of pneumonic plague. The F1-LcrV subunit vaccine in alhydrogel is known to induce humoral immunity. In this study, we utilized dendritic cells to investigate cellular immunity. We genetically engineered the LcrV virulence protein into the αDEC-205/CD205 monoclonal antibody and thereby targeted the conjugated protein directly to mouse DEC-205+ DCs in situ. We observed antigen-specific CD4+ T cell immunity measured by intracellular staining for interferon-γ in three different mouse strains (C57BL/6, BALB/c, and C3H/HeJ), while we could not observe such T cell responses with F1-V vaccine in alhydrogel. Using a peptide library for LcrV protein, we identified two or more distinct CD4+ T cell mimetopes in each MHC haplotype, consistent with the induction of broad immunity. When compared to nontargeted standard protein vaccine, DC targeting greatly increased the efficiency for inducing IFN-γ producing T cells. The targeted LcrV protein induced antibody responses to a similar extent as the F1-V subunit vaccine, but Th1-dependent IgG2a and IgG2c isotypes were observed only after αDEC-205:LcrV mAb immunization. This study sets the stage for the analysis of functional roles of IFN-γ producing T cells in Y. pestis infection.
Yersinia pestis (Y. pestis), the causative agent of plague or Black Death, is again drawing attention due to its potential use as a biological weapon [1, 2]. Bubonic plague is the most common clinical form of disease and is characterized by swollen lymph nodes (buboes) following bites by infected fleas. If untreated, it can progress to the life-threatening septicemic and pneumonic forms of plague. The latter form also occurs via contaminated respiratory droplets, or even by artificially generated aerosols containing Y. pestis and is particularly dangerous because of its high contagiousness and mortality . There is a need to develop more effective vaccines for pneumonic plague.
The existing vaccines against Y. pestis that have been used in man are either attenuated or killed organisms, and are insufficiently effective against pneumonic diseases . In a murine model, a subunit vaccine composed of recombinant F1 and LcrV protein (F1-V) has been demonstrated to induce protective responses . F1 antigen is an extracellular capsular antigen, encoded on a 110 kb plasmid (pMT-1) and exerting anti-phagocytic activity . Protective immunity via F1 has been shown in a murine model, and in humans, F1 elicits strong antibody responses following immunization with purified native antigen [6, 7]. The other component of the current subunit vaccine, LcrV, is a secreted multifunctional protein encoded on a 70 kb plasmid (pCD1). LcrV regulates the delivery of other effector molecules of Y. pestis or Yops into host cells upon contact via a type III secretion system (TTSS) [8, 9]. LcrV also is reported to have immunosuppressive potential by stimulating IL-10 secretion . Although LcrV alone has been shown to protect mice against infection with either F1-positive or F1-negative strains, the F1-V combination provides better protection than either subunit vaccine alone, and moreover, it protects mice against pneumonic plague [11–13].
Recent studies have indicated that besides humoral immunity, the induction of cellular immunity might be an important goal for a plague vaccine. CD4+ T cells are able to produce high levels of Th1 cytokines such as IFN-γ, which can activate macrophages to kill intracellular pathogens, and helper T cells also contribute to antibody-based immunity. In addition, CD4+ T cells can exert cytolytic activity on MHC class II-bearing targets . It was first noted that treatment of mice with exogenous IFN-γ plus TNF-α inhibited the multiplication of Y. pestis in vivo, thereby providing protection against intravenous challenge against 10 MLD of LcrV+ Y. pestis KIM . Likewise, LcrV antigen co-encapsulated with IFN-γ induced higher antigen-specific systemic immune responses . Moreover, Stat-4 deficient mice, which have low levels of IFN-γ production, were poorly protected from Y. pestis GB by the s.c. route, despite producing as high levels of serum antibody as wild type controls . A protective role of CD4+ T cells in Y. enterocolitica infection was recently demonstrated by Smiley and colleagues, where three LcrV-specific CD4 epitopes were identified that are presented in the context of the murine I-Ab class II MHC molecule . This group further showed that the transfer of Y. pestis-primed T cells to naive B cell deficient mice (μMT) protected against attenuated Y. pestis (KIM D27) intranasal challenge . In another study, Parent et al concluded that IFN-γ, TNF-α and NOS2 (nitric oxide synthase 2) are key elements of cellular immunity during pulmonary Y. pestis (KIM D27) infection . Therefore, an effective plague vaccine may need to prime not only humoral immunity but also strong Th1 type cellular immunity.
In this study, we targeted the LcrV virulence protein to dendritic cells (DCs), which are potent and specialized antigen-presenting cells, with the aim of generating more effective T helper cells. DCs are known as “nature’s adjuvants”, and various potential strategies to exploit DCs in vaccine design have been suggested . Recent studies provide a new avenue to DC-based vaccines by using an anti-mouse DC monoclonal antibody (mAb), specifically αDEC-205/CD205 mAb, to target vaccine antigens directly to DCs in situ [22–24]. Antigens incorporated within αDEC-205 mAb are efficiently and selectively targeted to DCs, leading to greatly enhanced presentation to T cells when compared to nontargeted antigen. This targeting strategy improves T cell vaccination, e.g. intensified and protective CD4 T cell immunity is induced to HIV gag p24 and p41 proteins by DEC-205-targeting and this provided protection against an airway challenge with recombinant vaccinia-gag virus . Using anti-DEC/LcrV fusion mAb together with DC maturation stimuli, we observed strong and broad antigen-specific Th1 type CD4+ T cell immunity as well as humoral immunity including high titers of Th1 type antibodies, which was not observed with the recombinant subunit F1-V vaccine. This study provides a new way to study the functional roles of Th1 type T cells in plague and suggests a way to the development of vaccines that include strong cell mediated as well as humoral immunity.
To target LcrV protein to DCs directly in vivo, the full length LcrV sequence was first codon optimized to improve expression and cloned in frame into the heavy chain of anti-mouse DEC-205 mAb as described . Due to the insertion of LcrV, which has a mass of 37 kDa, the heavy chain of the chimeric mAb was detected at ~ 97 kDa, following SDS-PAGE and either Coomassie staining or western blotting (Fig. 1A and C). To verify that the chimeric mAb bound properly to mouse DEC-205 receptor, a stable Chinese Hamster Ovary (CHO) cell transfectant, expressing mouse DEC-205 receptor on the surface, was stained with various concentrations of the conjugated- or non-conjugated mAbs. By FACS, the αDEC-205:LcrV mAb bound to DEC-205 receptor as well as the non-conjugated or “empty DEC” mAb (Fig. 1D). In addition, soluble LcrV protein was generated from the stable CHO cell transfectant and purified using an anti-FLAG affinity column (Fig. 1B). To visualize the specificity of targeting in situ, lymph nodes (LNs) sections were prepared after administration of Alexa488-conjugated αDEC-205:LcrV, or isotype control (mouse IgG1) mAb. Staining for the injected mAb was confined to dendritic profiles in T cell area (Fig. 1E, green signal). These profiles were double labeled for anti-DEC-205 applied to the section (Fig. 1E, blue signal). Taken together, these data indicate the successful generation of the chimeric mAb and specific in vivo targeting to DEC-205+ DCs.
To determine whether the chimeric mAb could induce T cell immune responses, groups of C57BL/6 (H-2b), BALB/c (H-2d), and C3H/HeJ (H-2k) mice were injected with a single dose of αDEC-205:LcrV mAb s.c. in the presence of DC maturation stimuli, which was the combination of 50 μg of poly IC and 25 μg of agonisticαCD40 mAb. Fourteen to 21 days later, splenocytes were harvested and restimulated with LcrV peptide pools, and the frequency of IFN-γ secreting antigen-specific CD4+/CD8+ T cells were examined by ELISPOT (data not shown) and intracellular cytokine staining (Fig. 2). The peptide library that we used was divided into 8 pools of 11 overlapping 15-mer peptides staggered every 4 amino acids along the entire LcrV sequence (see methods). For C57BL/6 mice, CD4+ T cell responses were noted to peptides in pool 2 and 3. For BALB/c, peptide pools 3, 5, and 6 were recognized by CD4+ T cells, and for C3H, peptide pools 3 and 6 contained the most reactive epitopes. CD8+ T cell responses were not observed. The reactive peptide pools were then broken down into individual peptides. The reactive mimetope peptides are summarized in Table 1. At least two (C57BL/6, and C3H) and three different mimetopes (BALB/c) were identified. We repeated this experiment twice and summarized the frequency of IFN-γ secreting CD4+ T cells in Table 2. These data indicated that the targeting of LcrV protein to DCs induced a broad immune response by CD4+ T cells in several MHC haplotypes.
To determine the efficiency of targeting strategy, increasing doses of αDEC-205:LcrV mAb in the presence of poly IC and αCD40 were administered into C57BL/6 (data not shown) or BALB/c mice and 14 days later, the frequency of IFN-γ and/or IL-2 secreting LcrV-specific CD4+ T cells was measured. In comparison, soluble nontargeted LcrV protein with poly IC and αCD40 as well as F1-V recombinant protein with alhydrogel as an adjuvant were also included (Fig. 3A). The data showed an increased frequency of IFN-γ and/or IL-2 secreting CD4+ T cells with αDEC-205:LcrV mAb immunization in mice, while non-targeted soluble LcrV protein required at least 100 μg to induce antigen-specific T cell responses. Moreover, similar T cell responses were not observed in mice immunized with F1-V adsorbed in alum adjuvant. This data demonstrated that the targeting strategy increases the efficiency of eliciting IFN-γ producing T cell responses.
We also tested different peptide doses during the immune assay and determined that most T cells responded to antigen at 0.2 μM peptide or higher (Fig. 3B). Based on the data, we decided to use 10 μg of αDEC-205:LcrV mAb, 100 μg of soluble protein, or 10 μg of F1-V recombinant protein for immunization, and 2 μM of reactive peptide for restimulation in this study unless otherwise indicated.
We compared the response to DC-targeted LcrV with other methods of immunization. When we injected soluble LcrV protein, either 10 μg or 100 μg per mouse, with standard adjuvants, complete Freund’s adjuvant (CFA), the immune responses were weak compared with 10 μg of DC-targeted αDEC-205:LcrV mAb in the presence of poly IC and αCD40 as adjuvants (Fig. 4A). Again, F1-V recombinant protein adsorbed in alum adjuvant did not induce IFN-γ secreting CD4+ T cells. In addition, DEC-205 was essential for the enhanced CD4 T cell immunity induced by DC-targeted LcrV protein, because such responses were not observed in DEC-205−/− mice (Fig. 4B). Thus, the data suggested that the targeting of a vaccine protein to DCs improves CD4+ T cell immunization relative to other approaches including standard soluble protein immunization as well as F1-V subunit vaccine.
To study humoral immunity induced by DC-targeted LcrV protein, we tested antibody responses with graded doses of αDEC-205:LcrV mAb, soluble LcrV protein, or F1-V. Antibody titres increased with increasing doses of protein (Fig. 5A). Also the titers induced by 10 μg of F1-V with alhydrogel were comparable to those induced by either 10 μg of αDEC-205:LcrV or 100 μg of soluble LcrV protein in the presence of poly IC and αCD40 as adjuvants (Fig. 5A).
Next, anti-LcrV antibody titers for individual IgG isotypes were examined (Fig. 5B). Again, a single immunization of αDEC-205:LcrV mAb with adjuvants induced serum levels of each isotype that were comparable to F1-V immunization. The IgG1 and 2b isotypes were predominant as described in previous studies . However, the IgG2a (BALB/c) or IgG2c (C57BL/6, data not shown) isotype was observed primarily in mice immunized with αDEC-205:LcrV mAb, with little or no IgG2c/2a in the F1-V immunized group (Fig. 5B). The IgG2c/2a isotypes reflect the influence of Th1 type helper T cells and are consistent with the observation that anti-DEC-205 targeted protein vaccine is a superior way to induce this type of cell-mediated immunity.
It is possible that a successful vaccine for plague needs to induce combined cellular and humoral immune responses against virulence determinants with minimal side effects. The need for cellular immunity was first indicated by the studies of Nakajima and Brubaker who noted protective effects of exogenous IFN-γ and TNF-α . Elvin et al then reported that Stat-4 deficient mice, which are only able to produce very low amounts of IFN-γ, were poorly protected by s.c. infection with Y. pestis even though they produce high levels of antibody similar to controls . A protective role of CD4+ T cells in Y. enterocolitica infection was directly demonstrated by Smiley and colleagues, where three LcrV-specific CD4 epitopes were identified that are presented in the context of the murine I-Ab class II MHC molecule . This group further showed that the transfer of Y. pestis-primed T cells to naive B cell deficient mice (μMT) protected against attenuated Y. pestis (KIM D27) intranasal challenge . Here we have addressed approaches to induce strong T cell immunity of the IFN-γ producing Th1 type in otherwise naive mice.
The existing plague vaccines, such as killed whole-cell vaccine and attenuated vaccine, induce protection against bubonic plague, but the vaccines evoke severe side effects and do not provide protection against pneumonic plague . A subunit vaccine composed of F1 and V proteins has provided the most promising efficiency based on the data obtained from small animal models . Along with evidence showing protection through adoptive serotherapy, the general mechanism by which F1-V subunit vaccine provided protection is considered to be antibody-mediated, and especially, the IgG1 isotype antibody plays an important role [26, 27].
It has been known that Y. pestis targets immune cells during infection, and through its specialized delivery system or TTSS, it has a range of anti-host functions from anti-phagocytic activity to immunosuppression . Therefore, to induce protective immunity, it may be valuable to activate immune cells such as DCs and macrophages that are targeted by Y. pestis. Th1 cytokines might provide a way to bring about this activation and/or overcome the immunosuppression that is induced by Y. pestis. In addition, Th1 type immune responses can provide “help” for antibody-based immunity as well as the generation of memory responses [14, 29]. For these purposes, we have utilized DCs and a DC-targeting strategy to induce immunity. We took a monoclonal antibody against a specific DC receptor, anti-DEC-205/CD205 mAb and genetically engineered the LcrV antigen of Y. pestis into the antibody heavy chain. Our data show the improved quality and quantity of T cell responses induced by DEC-205 targeted antigen.
Recently, LcrV protein epitopes recognized by CD4 T cells were identified in H-2b and H-2d mice in two independent studies [18, 30]. Since these studies depended on standard protein immunization for priming, high concentrations of protein or peptides, or sometimes, prime boost strategies were required to induce detectable CD4 T cell responses. In addition, to detect immune responses, it was necessary to use either a proliferation assay with bone-marrow derived macrophages and T cell hybridomas or ELISA for IL-2 released during in vitro restimulation with splenic antigen presenting cells (APCs). Generating T cell hybridomas is time-consuming and also IL-2 is not the best reflection of Th1 cytokines. In contrast, we assayed the immunity that was induced by in situ targeting of antigen to DCs with intracellular cytokine staining of immune spleen cells challenged for 6 hrs with a peptide library of 15-mers spanning the full length of LcrV protein. In this way we could document the presence of sizeable T cell responses with minimal tissue culture requirements. Our data showed that a single dose of smaller amounts of antigen conjugated to αDEC-205:LcrV mAb induced stronger IFN-γ secreting CD4+ T cell responses when compared to nontargeted soluble protein (Fig. 3A). In addition, relatively low doses (0.2 μM) of the mimetope peptides were used to activate the primed T cells in the assays (Fig. 3B). We also could observe broad T cell immunity induced by a single dose of DC-targeting, since we identified distinct CD4 T cell mimetopes in three different MHC haplotypes that were tested and at least two peptides in each haplotype (Fig. 2 and Table 1 and and2).2). We noted that the DC targeting strategy enhanced CD4+ T cell responses relative to nontargeted soluble protein with standard adjuvants such as CFA or alhydrogel (Fig. 4A). Interestingly the peptides that were recognized following DEC-205 based immunization were identical to those reported by Parent et al and Shim et al.
During primary immune responses, CD4+ T cells control affinity maturation and isotype switching by antigenic-specific B cells . Cytokines secreted from the helper T cells are important to determine isotype switching, such as IFN-γ (Th1) for IgG2a (BALB/c) or 2c(C57BL/6) production, or IL-4 (Th2) for IgG1, respectively . As described in a previous study, DC targeting leads to a broad range of different antibody isotypes, or a combination of Th1 and Th2 responses . In this study, we also observed that LcrV antigen targeted to DCs induced multiple isotypes including IgG1, 2a/2c, or 2b in serum (Fig. 5). However, F1-V recombinant protein vaccine could not induce Th1 isotype antibody as much as αDEC-205:LcrV mAb immunization, even though it generally leads to high titers of IgG1 and 2b isotypes.
Here we have used a combination of TLR3 ligand poly IC and agonistic αCD40 mAb as DC maturation stimuli to overcome T cell tolerance [25, 33–35]. Further study of active adjuvants is required. We observed that poly IC by itself did not lead to a primary immune response that was strong enough to be detected in our assays. We are now investigating the potential of poly IC alone as an optimal DC adjuvant in conjunction with a prime boost strategy, and we are also interested in whether the CD4+ T cell responses observed in the current study translate into a better quality and quantity of humoral immunity, especially memory. The contribution of the cellular immunity observed with our DC-targeting strategy should also be investigated in protection studies, especially against an aerosol challenge of Y. pestis.
C57BL/6 (H-2b), BALB/c (H-2d), and C3H/HeJ (H-2k) mice were purchased from Taconic, and DEC-205−/− mice (B6 background) are available from Jackson labs. All mice were maintained under specific pathogen-free conditions and used at 5–7 wks of age according to the guidelines of our Institutional Animal Care and Use Committee.
The mammalian codon-optimized cDNA encoding LcrV (GenBank accession No. DQ917566) open reading frame of full length was generated by PCR, and was cloned in frame with the carboxyl terminus of the heavy chain of mouse anti-DEC-205 mAb described in our previous study . αDEC-205:empty mAb is a mouse anti-DEC-205 mAb without any antigen insertion to the carboxyl terminus. The fusion αDEC-205:LcrV mAb and a control Ig:LcrV fusion mAb were produced by transient transfection into 293T cells using a calcium-phosphate method as described, and purified on Protein G column (Amersham Pharmacia Biotech) . Soluble LcrV protein was generated by adding a signal peptide and FLAG epitope tag to the NH2-terminus, expressed by stable transfection to Chinese Hamster Ovary (CHO) cells, and purified on anti-FLAG M1 affinity column (Sigma) according to manufacturer’s protocol. A recombinant F1-V protein from U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID, Frederick, MD) has been described .
Mice were injected intraperitoneally (i.p.) or subcutaneously (s.c.) in the hind footpads with fusion mAb, unconjugated αDEC-205 mAb, control Ig, or soluble LcrV protein in the presence or absence of adjuvants, which were 50 μg of poly IC with 25 μg of 1C10 agonistic anti-CD40 mAb (αCD40) per mouse. Soluble LcrV protein was also administered s.c. with complete Freund’s adjuvant (CFA). F1-V recombinant protein was administered with alhydrogel (2.0% Alhydrogel, Superfos Biosector, Vedbaek, Denmark).
15-mer peptides spanning the entire LcrV sequence and overlapping by 11 amino acids (aa) were synthesized by the Proteomics Resource Center (The Rockefeller University). The library was divided into 8 pools of 11 peptides except peptide pool number 8, which was composed of 10 peptides. The peptide pools spanned from aa 1–46 (pool 1), aa 36–86 (pool 2), aa 76–128 (pool 3), aa 118–170 (pool 4), aa 160–210 (pool 5), aa 200–250 (pool 6), aa 240–291 (pool 7), and aa 281–326 (pool 8) of the LcrV protein.
To detect LcrV-specific T cell immunity, spleen or lymph nodes cells were stimulated with pools of peptides (2 μM or indicated concentrations) or medium alone in the presence of a co-stimulatory anti-CD28 mAb (clone 37.51) for 6 hrs. Brefeldin A was added for the last 4 hrs to accumulate intracellular cytokines. Cells were then washed, incubated with anti-mouse-CD3, CD4 or CD8 mAbs for 20 min at 4°C after blocking Fcγ receptor with αCD16/CD23 antibody. Following fixation with Cytofix/Cytoperm Plus ™ (BD PharMingen), cells were stained for intracellular IFN-γ or IL-2 for 15 min at room temperature. All mAbs were purchased from BD PharMingen. Data were collected using FACS Calibur and analyzed by FlowJo (Tree Star).
To detect LcrV specific antibody, high binding ELISA plates (BD Falcon) were coated with 10 μg/ml of LcrV protein overnight at 4°C. Plates were washed 3 times with PBST (PBS and Tween 20, 0.1%) and blocked with PBST-BSA 5% for 1 hr at 37°C. Serial dilutions of serum were added to the plates and incubated for another 1 hr at 37°C. Various secondary goat anti-mouse Fc specific antibodies conjugated with horseradish peroxidase (Southern Biotech) were then added and visualized with OPD (o-phenylenediamine, Sigma) tablet at room temperature for 15–30 min. The reported titers represent the highest dilution of sample showing an OD450 higher than 0.1, and the data was presented as the log antibody titer. Some of the data were shown as OD450 values.
Alexa488-conjugated αDEC-205:LcrV, or isotype control (mouse IgG1) antibodies were prepared using the Alexa Fluor® 488 protein labeling kit (Molecular Probes). A deconvolution microscope (Olympus America, Melville, NY), and one-, two- or three-color fluorescence labeling was used with antibodies and fluorochromes listed in the micrographs.
We thank H. Zebroski of the Rockefeller University Proteomics Facility for synthesizing peptides, and J. Adams for help with the graphics. This work was supported by funds from the National Institutes of Health (#5 U54 AI057158 and AI13013). Y. Do was supported by a Postdoctoral Fellowship through the Career Development Program of the Northeast Biodefense Center.
CONFLICT OF INTEREST STATEMENT
R. M. Steinman is a consultant to Celldex, which is developing human DEC-205-based vaccines. Other authors have no conflict of interest.