Developing subunit immunogens using B and T cell epitopes and their constructs derived from the F1 antigen of Yersinia pestis using novel delivery vehicles

Yersinia pestis is the etiological agent of pneumonic and bubonic plague. As the currently licensed vaccines for plague have their own limitations, there is a need for a rational and more effective form of a subunit vaccine to combat both forms of the disease. Newer methods of antigen delivery coupled with adjuvant offer an alternative approach toward a plague vaccine. In order to develop a new generation vaccine against plague, we chose an immunodominant, outer membrane capsular protein, F1 of Y. pestis . The immunogenicity of the peptide sequences, predicted to possess B (three sequences, B1, B2 and B3) and T (two sequences, T1 and T2) cell determinants, was studied in a murine model with different genetic backgrounds, using alhydrogel and liposomes as delivery vehicles. All the peptide sequences are immunogenic in all mouse strains and showed primary and secondary immune response. B2 peptide was found to be most immunogenic, followed by B1 and B3 peptides. Chimeras made between B and T structures proved highly immunogenic and the antibody levels are comparable with native F1 antigen, thereby proving that T1 and T2 are helper sequences. Interestingly, the liposome mode of immunization was found to be more immunogenic and generated higher affinity antibodies than the alum-based preparation. Immunization using a mixture of all the peptides further proved B2 to be immunodominant. The IgG isotype profile showed predominance of IgG1, IgG2b followed by IgG2a for all the formulations irrespective of mode of antigen delivery. Lymphocyte proliferation of spleen cells primed in vivo with peptides, B-T conjugates and F1 antigen followed by in vitro stimulation with these antigens in soluble (medium) and particulate (liposome) form, showed dose-dependent stimulation of T cells, while B-T constructs showed a higher stimulation index, comparable to F1 antigen. The liposome mode of antigen presentation showed higher lymphoproliferation of spleen cells. Of all the peptides tested, T1 and T2 sequences showed the highest stimulation indices. The pattern of cytokine levels was in the following order: interferon- Q s interleukin-2 s interleukin-4. In vivo protective studies of the B-T conjugates revealed that B1T1 and a mixture of conjugates showed a survival rate of 10 days. Thus, the study highlights the importance of B and T cell epitopes as peptide-based immunogens, being a serious alternative for plague vaccine.


Introduction
Yersinia pestis is the etiological agent of both forms of plague, which is still active in various regions of the world.
Currently, the licensed procedure for immunizing individuals against plague involves parenteral administration of killed whole-cell suspension of Y. pestis [3]. This approach has shown protection against bubonic plague, but failed to prevent infection in the lungs (pneumonic plague). Moreover, the vaccine requires repeated boosters to maintain peak antibody levels [4]. This necessitates the search for acellular vaccines against plague that should consist of immunogenic subunits derived from the parent organism. Two of the most prominent antigens are fraction 1 (F1) and V antigen [1]. F1, the major protein component of the capsule surrounding Y. pestis cells, is only expressed at 37 ‡C and is believed to confer resistance to phagocytosis [2]. High anti-F1 titers have been correlated with survival following plague infection [5]. The protection provided by recombinant and highly puri¢ed F1 antigen is found to be more or less similar. The production of protein-based vaccines, based on isolating these protective antigens from bacteria or on genetically expressing the same, is considered to be tedious and hazardous and needs special laboratory facilities with skilled personnel. The peptide-based vaccines thus provide an alternative approach to conventional vaccines.
For developing peptide-based immunogens, the relevant protective B and T cell determinants have to be mapped on the protein. The amino acid composition, orientation of the epitopes, the adjuvant used and the route of vaccine delivery may in£uence the antibody level and its a⁄nity. Recently, the potential immunogenic epitopes responsible for inducing the immune response have been identi¢ed on the F1 antigen through competitive as well as direct binding assays in our laboratory [6].
It is a generalized observation that B and T cell epitopes cluster within a limited region of the antigen. So emphasis was laid on the carboxy-terminus of the F1 antigen, since the amino-terminus has been shown to share sequence homology with the interleukin (IL)-1 receptor antagonist [7]. When the F1 antigen was subjected to DNAS-TAR software, a high antigenic index and hydrophilic regions with characteristic secondary structures were observed at the carboxy-terminal end of the molecule. Hence, in the present study, the B cell and the T cell recognition motifs were evaluated by testing these predicted sequences for humoral and cellular responses in murine systems with di¡erent genetic backgrounds. The humoral and cellular responses, induced by di¡erent peptide sequences and chimeras made between B and T constructs were studied by delivering the peptide antigen in alhydrogel as well as in liposomes. Liposomes are envisaged as the novel mode of vaccine delivery and rely on controlled release of entrapped antigen [8]. Since alhydrogel is the only licensed adjuvant for humans, the present study emphasized the role of the conventional adjuvant system with a modi¢ed delivery vehicle. Inbred and outbred strains of mice were used to study the in£uence of major MHC genes on each peptide's immunogenicity, since the peptides are known to be genetically restricted. Finally, all the B-T constructs were studied for in vivo protection in mice after challenging them with living bacteria. The outcome of the study based on B-T constructs of F1 antigen in inducing humoral and cellular responses indicates an alternative approach for peptide-based immunogens against human plague. Peptides were selected from the protein regions containing hydrophilic, helical segments as well as L-turns, as judged by the secondary structure predictions. All the sequences were synthesized by solid phase technique, using t-Boc chemistry assembled on Pam resin. All peptides were puri¢ed by semi-preparative high performance liquid chromatography, using a C18 column with a binary aqueous acetonitrile gradient containing 0.1% tri£uoroacetic acid.

Materials and methods
The putative B cell epitopes were chemically linked to putative T cell epitopes using the two-step glutaraldehyde method. At the end of the coupling, peptide stoichiometry was found to be 1:1 as determined by amino acid analysis.
A mixture of all of the above ¢ve peptides except B4 was used to prepare a cocktail for immunization.

Mice
Inbred mice of age group 6^8 weeks [C57BL/6 (H-2 b ), BALB/c (H-2 d ), CBA/J (H-2 k ) and FVB/J (H-2 q )] were procured from the breeding facilities of the National Institute of Immunology (NII), New Delhi. Outbred mice were procured from the breeding facilities of the Central Animal House, AIIMS, New Delhi. All experimental groups consisted of ¢ve or six animals. All animal experiments were conducted in accordance with the guidelines for the care and use of laboratory animals as promulgated by CPCSEA, Ministry of Social Justice, Government of India and adopted by the Ethics Committee on animal experimentation by AIIMS, New Delhi.

Liposomal formulation
Liposomes were prepared according to our protocol reported previously [9]. In brief, multilamellar vesicles (MLVs) were prepared by dissolving phosphatidylcholine, cholesterol and phosphatidylglycerol in a molar ratio of 7:4:1 in chloroform and methanol and subsequently evaporated to dryness. These MLVs thus obtained were suspended in 1 ml of double-distilled water (DDW) and probe-sonicated to form small unilamellar vesicles (SUVs). The SUVs formed were kept at room temperature for 1 h for stabilization and later the solution was centrifuged at 5000 rpm for 15 min to remove the aggregated lipids. The peptide or peptide conjugates (5.0 mg ml 31 ) were added to the SUVs, freeze-dried and subsequently lyophilized. The dry powder obtained was rehydrated with 1.0 ml DDW. After dilution with 5.0 ml phosphate-bu¡ered saline (PBS, 0.01 M, pH 7.4) the suspension was ultracentrifuged at 100 000Ug for 1 h at 4 ‡C. The liposomal pellet was washed twice with PBS and ¢nally suspended in 0.2^0.5 ml PBS and stored at 370 ‡C. The percentage entrapment of peptide in liposomes was estimated in the supernatants by the bicinchoninic acid method [10] and was found to be in the range of 40^45%. The size of the liposomes was determined by scanning electron microscope and they were in the range of 40^60 nm.

Immunization protocol
Mice were immunized with various combinations of peptide and peptide conjugates: 1. peptide alone in alhydrogel: 50 Wg 2. peptide alone in liposome: 25 Wg 3. B-T conjugates in alhydrogel: 50 Wg 4. cocktail of peptides: 50 Wg (10 Wg each) 5. F1 antigen in liposome: 10 Wg The optimum antigen for alhydrogel-adsorbed, liposome-entrapped preparation was standardized by dose^kinetic study. For booster immunization, the amount of antigen was reduced to half the primary dose. The liposome preparation was also adsorbed on alhydrogel before immunization. In the case of alhydrogel and liposome preparation, mice were immunized in the footpad on days 0, 21 and 35. Mice were bled on days 28, 40 and 60. The serum was stored at 320 ‡C until use.

Antibody measurement
In enzyme-linked immunosorbent assay (ELISA), 100 Wl of peptide^bovine serum albumin (BSA) conjugate (100 ng per well) was used to coat the 96-well microtiter plates (Immunlon2, Dynatech, VA, USA). After washing and blocking with skimmed milk powder, a single dilution of the test serum (1:100) or a serial two-fold dilution (1:4001 02 400) of pooled antisera was used for measuring peptide-speci¢c total IgG levels and peak antibody titers, respectively. Titers were assessed as the highest serum dilution giving an absorbance (0.15) higher than that of preimmune sera. The color was developed using OPD as chromogen and H 2 O 2 as the substrate and absorbance was read at 490 nm.

Estimation of IgG isotypes
For estimating IgG isotypes, mice sera (dilution 1:100) from di¡erent immunization groups were incubated for 2 h at 37 ‡C, with di¡erent peptide^BSA conjugates coated on ELISA plates. The IgG isotype binding was detected using secondary goat anti-mouse IgG speci¢c for each subclass (dilution 1:1000, Sigma isotyping kit) for 2 h at 37 ‡C, followed by tertiary antibody (rabbit anti-goat IgGĥ orseradish peroxidase (HRP) conjugate, 1:1000 dilution). The absorbance was read as described above.

Binding a⁄nity of the anti-peptide antibodies (K D )
The binding a⁄nities of the anti-peptide antibodies in various immunizations were determined by measuring the dissociation constant (K D ) [11]. In brief, mice serum (1:200 dilution) was incubated with di¡erent concentrations of the peptide (0.1^10 nM) for 15 h at 20 ‡C, so as to attain antigen^antibody equilibrium. The antigen^antibody complexes were transferred onto the wells of the microtiter plates previously coated with the respective peptide capture antigen (100 ng per well). After washing three times with PBS containing 0.05% Tween 20, the plates were incubated with goat anti-mouse IgG^HRP (1:1000 dilution) for 90 min at 37 ‡C. The color was developed using OPD and hydrogen peroxide as substrate. The reaction was stopped with 8 N H 2 SO 4 and absorbance was read at 492 nm. Dissociation constants were then calculated using regression analysis and a simpli¢cation of the mathematical equation of Scattered and Klotz [11] : where A o is the absorbance without free antigen, A is the absorbance with free antigen and a o the concentration of free antigen. This equation permits the determination of K D even when the concentration of speci¢c antibody is not known.

Antigen-induced splenocyte proliferation
For studying T cell recognition properties, outbred mice were immunized subcutaneously with F1 antigen (10 Wg) and individual peptides/peptide conjugates (50 Wg) adsorbed on alhydrogel. Animals were boosted with half the dose of the antigen on day 8. Mice were killed on day 15. Single cell suspensions of splenocytes (devoid of B cells by panning with anti-mouse immunoglobulins) were prepared and cultured in 96-well plates at 2U10 5 cells per well in RPMI 1640 medium supplemented with gentamicin and streptomycin (50 Wg ml 31 ) and 10% heatinactivated fetal calf serum.
The cells in each well were then stimulated in vitro with peptide/peptide conjugate/F1 antigen dissolved in medium (25, 50 and 100 Wg ml 31 ) or entrapped in liposomes (10, 25 and 50 Wg ml 31 ) in a ¢nal volume of 200 Wl. Phytohemagglutinin (2 Wg ml 31 ) was used as a positive control. Medium and peptide-free liposome served as a negative control. After 72 h of culture, 50 Wl of the culture supernatant was collected and stored at 370 ‡C for cytokine estimation. The cultures were pulsed with [ 3 H]thymidine (speci¢c activity 6.4 Ci mmol 31 , BARC, Trombay, Mumbai, India) at 0.5 WCi per well. The cells were incubated for 18 h and then harvested onto glass ¢ber disks, and thymidine incorporation was measured by a liquid scintillation counter. Proliferative responses were expressed as stimulation index (SI). All cultures were taken in triplicate.

Cytokine measurement
The culture supernatants were centrifuged at 5000 rpm for 15 min, ¢ltered through 0.22-Wm pores and assayed for cytokine levels using sandwich ELISA (RpD systems, Minneapolis, MN, USA) according to the manufacturer's instructions.

Statistical analysis
The data on peptide-speci¢c IgG levels, IgG subclasses, K D measurement, T cell proliferation and cytokine levels were compared by non-parametric, Kruskal^Wallis oneway analysis of variance by ranks. The levels of signi¢cance were compared between peptide(s) in alhydrogel versus peptide(s) in liposomes and between the strains.

In vivo protective studies
BALB/c mice (six in each group) were immunized with respective B-T conjugate adsorbed on alhydrogel (30 Wg per animal) or conjugate mixture (10 Wg of each conjugate) through the intramuscular route in the hind leg on days 1, 7 and 14. On day 21, animals were boosted with half the dose of the conjugate at the same inoculation site. The control group consisted of six mice immunized with adjuvant. F1 antigen (10 Wg) was used under identical experimental conditions as a positive control. On day 28, each group was challenged intraperitoneally with live Y. pestis bacteria (1U10 5 bacteria in 0.1 ml saline per animal, strain 195-p). The survival time was monitored for all the groups for over 10 days. Survivors were culled after this time period.

Humoral response to individual and conjugated peptides
The antibody response in di¡erent mouse strains was observed in terms of peak antibody titer and total/subclass-speci¢c IgG levels. In general, mice bearing haplotype H-2 d and H-2 k showed higher antibody levels as compared to the other two mice strains H-2 q and H-2 b , irrespective of the nature of the peptide used. Low peak antibody titers were observed for all the peptides in all the strains on day 28. Mice bearing the haplotypes H-2 d and H-2 k produced peak titers of 3200^6400, while the other two strains produced peak titers of 1600^3200 on day 28. The peak antibody titers rose to 6400^25 600 on day 42 for most of the peptides. When peptide-speci¢c peak antibody titers were measured on day 60, there was a decline with peak titers ranging from 1600 to 12 800, depending upon the nature of the peptide and strain used. However, for peptides B1, B2, B3 and T1 the peak titers were maintained at the same levels in the mice bearing the haplotypes H-2 b and H-2 d . In general, B2 and T1 peptides showed higher immunogenicity than the other three peptides and the magnitude of the order of peak antibody titers was observed as B2 s T1 s B1WT2 s B3, with the order of responding mouse strains being H-2 d s H-2 k s H-2 b s H-2 q (Fig. 1a^e).
The delivery of the same peptide antigens in liposomes elicited higher peak antibody titers up to 25 600, which was statistically higher (P 6 0.05) as compared to the alhydrogel-based preparations. Though there was a small fall in peak antibody titer on day 60, this fall was insig-ni¢cant on day 42. Interestingly, liposomal delivery produced consistent peak antibody titers with all the peptides in all the mouse strains, which was not the case with the alhydrogel-based preparations (Fig. 2a^e).
When the humoral response was studied for peptide conjugates delivered in alhydrogel, it showed very high peptide-speci¢c peak antibody titers in all the strains and in all the bleeds (Fig. 3a^f). The peak antibody titers for either constituent peptide or the conjugate ranged from 3200 to 12 800 on day 28. By day 42, the antibody titers rose to 51 200 whereas a few peptide conjugates showed peak titers of 102 400. By day 60, the peak antibody titers were maintained at the same level or a marginal fall was observed with the constituent peptides. The B2T1 conjugate showed the highest antibody titers, followed by B1T1 s B2T2 s B1T2 s B3T1 s B3T2.
Outbred strains consistently produced high peak antibody titers in all the bleeds with all the conjugates. The peak titers for F1 antigen in all the strains were 64 000 on day 28 and showed a booster response to 102 400 on day 42. However, by day 60, the titers were the same as at the primary bleed or higher (data not shown). It is important to mention that peak antibody titers exhibited by B-T conjugates are comparable with anti-F1 antibodies.

IgG subclass
Delivery of all the peptides in alhydrogel generated both IgG1 and IgG2a/2b isotypes in all the strains. Initially low levels of these isotypes were observed on day 28. By day 42, there was a concomitant rise in either IgG1 or IgG2a/ 2b isotypes or both the isotypes, which was statistically signi¢cant (P 6 0.01). Incidentally, these isotype levels were maintained at the same level or a marginal fall in their levels was observed on day 60, which was statistically insigni¢cant (data not shown).
The levels of IgG3 isotypes were negligible in all the bleeds with all the peptides. In general, IgG1 and IgG2b are the major isotypes produced followed by IgG2a, irre-  spective of the nature of the peptide or formulation used. In general, liposomal delivery of antigen produced higher isotype levels (P 6 0.05) and these levels were maintained even on day 60 as compared to alhydrogel-based preparations (Fig. 4a^e).
When the IgG isotype pro¢le was observed for peptide conjugates delivered in alhydrogel in outbred as well as inbred strains, the predominant isotypes were either IgG1 or IgG2a or IgG2b or both (Fig. 5a^f). The primary immune response showed low levels of these isotypes while in subsequent bleeds, these isotypes were signi¢cantly higher (P 6 0.01) and the levels were maintained until day 60. All the strains and all the peptide conjugates showed very high IgG1, or IgG2a or IgG2b isotope levels in all the bleeds. Though peptide B3 showed lower isotype levels than other peptides, its conjugate showed similar levels of these isotypes with minimal strain-to-strain variation. The IgG3 levels detected in all bleeds were negligible with all the peptides. The major isotype produced by F1 antisera is predominantly IgG1 and IgG2b in all the four strains. In subsequent bleeds, there was a rise in IgG1 isotype, while levels of IgG2b were maintained at the same level. The levels of IgG2a are almost consistent in all the bleeds (data not shown).

Humoral response to a cocktail of peptides
The humoral response to a mixture of all the peptides was studied in outbred and three inbred strains of mice. High B2 peptide-speci¢c IgG levels were detected when compared to other peptides. Thus, the order of response was B2 s B1 s B3 s T1 s T2 with B2 peptide showing a strong anamnestic response as compared to other peptides in all the strains (data not shown).

Dissociation constant (K D ) of the peptide antisera
The a⁄nity of the antibodies raised against di¡erent peptide formulations (alhydrogel and liposomes) was studied by measuring K D values in low and high respond-ing strains. The serum chosen for each peptide was that on day 42 (Table 1). Alhydrogel-based peptide sera showed K D values in the range of 17.5^60.1 nM, while liposomebased antisera showed K D values of 9.07^18.2 nM, which was statistically signi¢cant (P 6 0.01) when compared to alhydrogel-based preparation. There was a two-to fourfold lower K D value with liposome-based preparations when compared to alhydrogel, indicating that the a⁄nities of the antibodies by this delivery route were higher than those of the alhydrogel-based preparations.

Antigen-induced T cell proliferation
To characterize the prevalence of peptide-speci¢c T lymphocytes, mice were ¢rst primed either with native antigen (F1) or with peptide conjugates of B-T and spleen cells restimulated in vitro, with native antigen/respective peptide with homologous antigen (25 Wg ml 31 ), in soluble and particulate form, respectively (data not shown). When conjugate-primed splenocytes were stimulated in vitro with the respective conjugate or constituent peptide or F1 antigen, high stimulation was observed (SI 10.11 6.9) in soluble form (Table 2). However, there was a marginal increase in SI (10.5^24.1) when these antigens were presented in particulate form. Notably, the constituent T cell peptides always showed a higher SI, irrespective of how they were presented, compared to constituent B cell peptides.

Cytokine levels
The culture supernatants obtained from spleen cells primed and stimulated with F1 antigen showed high levels of cytokines, i.e. IL-2 in soluble (318 pg ml 31 ) and particulate form (355 pg ml 31 ). In culture supernatants obtained from spleen cells primed with F1 antigen and pulsed in vitro with di¡erent peptides, IL-2 levels were maximal (225^300 pg ml 31 ) for T1 and T2 peptides, while the other three peptides B1, B2 and B3 showed low IL-2 levels (1151 85 pg ml 31 ). Similarly, peptide B1-, B2-and B3-primed Table 3 Cytokine levels (pg ml 31 ) in culture supernatants of spleen cells, from outbred mice primed with peptide/F1 antigen and stimulated in vitro with respective peptide/F1 antigen in soluble (medium, M) or particulate form (liposomes, L) and -stimulated cultures showed moderate IL-2 levels (110^238 pg ml 31 ), while T1 and T2 peptide-primed cultures showed high IL-2 levels (250^255 and 220^280 pg ml 31 , respectively). There was no signi¢cant di¡erence between the two formulations, i.e. soluble versus liposome mode of antigen presentation (Table 3). Conjugate-primed and -pulsed cultures showed undoubtedly higher IL-2 levels in all the formulations, which were statistically signi¢cant (P 6 0.05) as compared to B1, B2 and B3 peptide-pulsed cultures. T1-and T2-pulsed cultures always showed higher IL-2 levels, with all the combinations. T1 peptide showed an edge over T2 peptide, in generating higher IL-2 levels. Interestingly, F1primed and -pulsed cells showed higher IL-2 levels and these levels are comparable to T1-and T2-pulsed cultures ( Table 4).
F1-primed and -pulsed spleen cells produced high levels of interferon-Q (IFN-Q) (860^940 pg ml 31 ), irrespective of Table 4 Cytokine levels (pg ml 31 ) in culture supernatants of spleen cells, from outbred mice primed with conjugate and stimulated in vitro with respective conjugate/constituent peptide/F1 antigen in soluble (medium, M) or particulate form (liposomes, L) the way the antigen was presented to the cells, while peptide B1-, B2-and B3-pulsed spleen cells produced low levels of IFN-Q (200^266 pg ml 31 ), whereas T1-and T2pulsed spleen cells produced maximal IFN-Q levels (4006 90 pg ml 31 ). The levels of IFN-Q were statistically higher (P 6 0.01) with T1 and T2 peptides as compared to B1, B2 and B3 peptides, with T1 peptide showing an edge over T2 peptide. In the case of individual peptide-primed and -pulsed spleen cells, T1 and T2 peptides showed signi¢cantly (P 6 0.01) higher IFN-Q levels in the particulate form of presentation (700^800 pg ml 31 ) than in soluble form (440^510 pg ml 31 ), while the other three peptides (B1, B2 and B3) produced low levels of IFN-Q (200^332 pg ml 31 ) ( Table 3).
With conjugate-primed and -pulsed spleen cells, comparatively higher IFN-Q levels (600^1200 pg ml 31 ) were produced, while the particulate form of presentation still produced signi¢cantly higher (P 6 0.01) levels of IFN-Q as compared to the soluble form. Somehow, conjugates of B3 produced low levels of IFN-Q (322^360 pg ml 31 ). Surprisingly, conjugate-primed and -pulsed spleen cells with individual constituent peptides (B1, B2, B3, T1 and T2) showed higher levels of IFN-Q for T1 and T2 peptides than B1, B2 and B3 (Table 4).
Low levels of IL-4 were observed for all the formulations. Spleen cells primed and pulsed with F1 antigen showed low levels of IL-4 (68^102 pg ml 31 ), while peptide-pulsed cultures produced 30^96 pg ml 31 in both soluble and particulate form (Table 3). Peptide-primed and -stimulated spleen cells produced very low levels of IL-4 (35^125 pg ml 31 ). Spleen cells primed with peptides and pulsed with F1 antigen also produced low levels of IL-4 (60^110 pg ml 31 ) in any of the formulations used. Also in the case of conjugates, the spleen cells produced very low levels of IL-4 (40^112 pg ml 31 ), either with conjugates or with constituent peptides (Table 4).

In vivo protective studies
Mice challenged with Y. pestis bacteria in alhydrogel survived until day 4. However, the groups immunized with B1T1 and the mixture of all conjugates showed high protection, comparable with the F1 antigen (Table  5), and the survival rate was nearly 80% on day 10.

Discussion
Human plague infection is associated with exposure to Y. pestis-infected rodents/£eas or by inhalation of infectious aerosols. Even though live and killed vaccines are available to prevent plague, the reactogenicity, long immunization schedule, frequent boosters to maintain peak antibody levels and lack of demonstrable e⁄ciency against pneumonic plague have limited their applicability for humans. Intensive e¡orts have been focused on the development of a subunit vaccine against plague, comprising two major virulence factors of Y. pestis, F1 and V antigen [12]. Several studies have shown that the protective immunity against plague is mediated in part by antibodies directed against the F1 antigen, which is a major envelope protein [13]. Thus, F1 antigen formed the basis of the new generation plague vaccine. Though a great deal of work has been done on native F1 antigen, peptide-based vaccines using the immunogenic potential of antigenic sites have not been reported. The present study is a step forward in this direction for generating humoral and cellular responses to B and T cell antigenic sites on F1 antigen in a murine system.
It is a generalized observation that peptide-based vaccines are weak immunogens and they undergo rapid degradation in the blood stream. Therefore, these vaccines have to be supplemented with potent adjuvants through novel delivery vehicles. Thus, the liposomal mode of antigen delivery was the basis of the present study. Initially, the immunogenicity of F1 antigen was studied in four inbred strains of mice. All strains produced very high antibody titers after immunization with a low dose of antigen in liposomes, indicating that F1 is a highly immunogenic molecule. This observation encouraged us to study the Table 5 In vivo protective study in BALB/c mice immunogenicity of some of the predicted peptide sequences in mice with di¡erent genetic backgrounds, through immunization in liposomes and alhydrogel. Inbred strains are expected to give consistent results, because they lack the genetic variation that characterizes a normal outbred population. This implies that di¡erent haplotypes may present the same antigen di¡erently and may subsequently produce a di¡erent immune response. When the immunogenicity of all the peptides was measured in terms of peak antibody titers, all were immunogenic, both in alhydrogel and in liposomes, and peptide B2 proved to be most immunogenic compared to the other peptides. This is in accordance with our earlier observation that B2 peptide is the immunodominant sequence of F1 antigen [6]. However, liposomal delivery of antigen proved more immunogenic than alum-based immunization in generating high a⁄nity antibody levels and the levels persisted over a longer duration with half the immunizing dose. It appears that liposomal delivery of the antigen may be responsible for internalization, and endosomal transport of the peptides for stimulating high antibody levels and activating high a⁄nity B cell clones. In order to maintain consistent peak antibody levels and memory response, the B cell and T cell determinants should be derived from the same protein antigen. Hence, we made a construct between putative B and T sequences and their immunogenicity was studied in both inbred and outbred strains of mice. The immunogenicity study indicated that all the B-T conjugates produced very high antibody levels by in£uencing B cell activation. The secondary response with the conjugates was better than with the individual peptides, even with alum immunization. This increased antibody level is possible, provided that the length of the B-T constructs containing the T helper sequence creates a conformation that is more e⁄ciently processed and presented than that of the individual peptide. The antibody levels were quite comparable to those with the native F1 antigen. Interestingly, even the constituent peptides showed high antibody levels. This study clearly indicates that the chosen T1 and T2 sequences possess T-helper activity. To our surprise, the conjugates made between T1 peptide and B1, B2 and B3 always showed a higher immune response than the ones made with T2 peptide.
It is generally accepted that the level of antibodies, their subclass and avidity are important factors in preventing disease. F1 antigen entrapped in liposomes generated primarily IgG1 and IgG2b isotypes, followed by IgG2a in all the strains. Similarly, the delivery of individual peptides either in alhydrogel or in liposomes produced IgG1 and IgG2a or IgG2b isotypes in all the strains. The study clearly indicates the activation of both CD4+ TH1 and TH2 phenotypes that may be essential for the clearance of pathogens from the circulation. Again, both IgG2a and IgG2b isotype antibodies are known to be cytophilic in nature. The subclass distribution with conjugates again showed a mixed TH1/TH2 type response, in which both IgG1 and IgG2a or IgG2b isotypes were elevated. Though IgG1 isotype possesses lower protective capacity than IgG2a or IgG2b, in Yersinia infections increased levels of IgG1 isotype have been correlated with protection [14]. A similar isotype distribution as observed in the present study with peptides, peptide conjugates and native F1 antigen probably shows identical antigen-presenting cell (APC) processing pathway and cytokine secretion.
During the study on the immunogenicity of a physical mixture of all the peptides, again the B2 peptide proved more immunogenic than the other peptides. The study further proves that the B2 sequence appears to be highly accessible to the B lymphocytes and may be a surface motif on the F1 antigen. Overall, the study indicates that the predicted sequences are immunogenic with no genetic restriction of the immune response, and outbred strains show an equal or higher immune response as compared to inbred strains.
The a⁄nity of the antibodies was studied for all the peptides in both delivery systems, using sera obtained from high and low responder strains. The K D value is two-to four-fold lower with liposomal delivery as compared to alhydrogel delivery, indicating that the a⁄nity of the antibodies in the former case is superior. This could explain the slow sensitization of high a⁄nity B cell clones using the particulate mode (liposomes) of delivery of peptide antigens.
Cell-mediated immunity (CMI) is important for a plague vaccine, as it eliminates bacilli from intracellular sites (lymph nodes etc.). The ability of the peptides or peptide conjugates of F1 antigen to induce a CMI response was determined by assaying the splenocyte proliferative response and cytokine secretion in outbred mice. When F1 antigen/individual peptide-primed spleen cells were stimulated in vitro with native antigen or respective peptides, high stimulation was observed for T1 and T2 peptides, which showed lymphoproliferation similar to F1 antigen, thereby con¢rming that these predicted sequences are potential T cell epitopes. This is in accordance with the predictive algorithms of the F1 antigen, where these two regions possess folded amphipathic L structure. Similarly, when peptide-primed spleen cells were stimulated in vitro with F1 antigen or the respective peptide, high T cell proliferation was observed with T1 and T2 peptides, thus recon¢rming these as T cell epitopes. Liposomal presentation of antigen in vitro with the primed spleen cells always produced high lymphoproliferation and required half the antigen dose compared with the soluble mode of presentation of the antigen. This observation is further supported by di¡erent conjugates showing, in the presence of T1 and T2 peptides, a lymphoproliferative response similar to that of F1 antigen. Also, the conjugate-primed spleen cells stimulated in vitro with conjugates showed higher SIs than the constituent peptides. This probably explains the generation and expansion of high a⁄nity T cell clones. Thus, the overall observation on humoral and cellular responses conclusively proves B1, B2 and B3 peptides to be B cell epitopes and T1 and T2 to be T cell epitopes.
T lymphocytes and their products are essential both in regulating speci¢c antibody formation and in inducing antibody-independent immunity in various infections. On activation of T cells by the APCs and on interaction with the antigen, T cell subsets (i.e. TH1 and TH2) are activated which are reported to play a crucial role in infectious diseases. CD4+ T cells that produce predominantly TH1 cytokines are in£ammatory T cells and potentiate CMI. The TH1 cytokines activate macrophages and natural killer cells through IL-2 help to expand and activate CD8+ T cells. When F1-primed/peptide-primed spleen cells were stimulated in vitro with the respective peptide or F1 antigen, high levels of IFN-Q were observed. T1 and T2 peptides produced signi¢cantly higher IFN-Q levels compared to B1, B2 and B3 peptides. The particulate mode of peptide antigen presentation consistently produced high IFN-Q and IL-2 levels when compared to the soluble form of antigen presentation. The same holds for the peptide conjugates. This study clearly indicates that the predicted T cell epitopes are indeed T helper epitopes and are able to interact with a wide range of MHC molecules. Low levels of IL-4 were observed with all of the formulations regardless of the mode of antigen presentation to the primed spleen cells. Taking the cytokine data together, a more pronounced CD4+ TH1-driven response was observed, which was re£ected in the generation of IgG2a/IgG2b antibodies. During the in vivo protection studies, mice immunized with F1, B1T1 and a mixture of conjugates showed higher protection and the animals survived for about 10 days. Surprisingly, the conjugates of either B2T1 or B2T2 did not show good protection in vivo. This may be attributed probably to (a) B1T1 showing a higher lympho proliferative response than B2T1 or B2T2, though the di¡erence in humoral response was comparatively not visible, or (b) the fact that in the B1T1 conjugate the two sequences are proximal to each other and they might be positively selected, while in the B2T2 sequence the two sequences are overlapping with a sequence FFVRSIG and thus may be negatively selected during immune activation, or, ¢nally, (c) the fact that T1 peptide consistently showed higher proliferative and antibody responses than T2 peptide.
In conclusion, the study highlights the importance of peptide-based immunogens of F1 antigen as an alternative to conventional vaccines against plague.