|Home | About | Journals | Submit | Contact Us | Français|
Plasmid-encoded DNA vaccines appear to be a safe and effective method for delivering antigen; however, the immunogenicity of such vaccines is often suboptimal. Cytokine adjuvants including interleukin (IL)-12, RANTES, granulocyte-macrophage colony-stimulating factor, IL-15, and others have been used to augment the immune response against DNA vaccines. In particular, IL-15 binds to a unique high-affinity receptor, IL-15Rα; is trans-presented to CD8+ T cells expressing the common βγ chain; and has been shown to play a role in the generation, maintenance, and proliferation of antigen-specific CD8+ T cells. In this study, we took the unique approach of using both a cytokine and its receptor as an adjuvant in an HIV-1 vaccine strategy. To study IL-15Rα expression, a unique monoclonal antibody (KK1.23) was generated to confirm receptor expression in vitro. Coimmunization of IL-15 and IL-15Rα plasmids with HIV-1 antigenic plasmids in mice enhanced the antigen-specific immune response 2-fold over IL-15 immunoadjuvant alone. Furthermore, plasmid-encoded IL-15Rα augments immune responses in the absence of IL-15, suggesting its role as a novel adjuvant. Moreover, pIL-15Rα enhanced the cellular, but not the humoral, immune response as measured by antigen-specific IgG antibody. This is the first report describing that IL-15Rα itself can act as an adjuvant by enhancing an antigen-specific T cell response. Uniquely, pIL-15 and pIL-15Rα adjuvants combined, but not the receptor α chain alone, may be useful as a strategy for generating and maintaining memory CD8+ T cells in a DNA vaccine.
The generation of potent CD8+ T cell responses by DNA vaccine technology is an important goal for this platform. The contribution of CD8+ T cells for controlling viral replication in both human (Koup et al., 1994; Cao et al., 1995; Musey et al., 1997; Ogg et al., 1998; Betts et al., 1999) and nonhuman primate models (Jin et al., 1999; Schmitz et al., 1999; Barouch et al., 2000; Amara et al., 2001; Shiver et al., 2002) in the HIV model as well as other viral infections has underscored the importance of improving the T cell responses induced by this technology. Many unique strategies including improved delivery techniques, enhanced construct design, heterologous prime–boost strategies, and the use of molecular adjuvants have been reported (Gao et al., 2003; Leifert et al., 2004; Fuller et al., 2006; Hokey and Weiner, 2006; Schoenly and Weiner, 2008). Molecular adjuvants including chemokines and cytokines can be incorporated into a vaccine strategy to skew the immune response toward cellular or humoral immunity. Cytokines such as interleukin (IL)-12 and IL-15 have been effective in enhancing the immune response in both murine and nonhuman primate models (Morrow and Weiner, 2008). In particular, IL-15 has been shown to be involved in the expansion of cellular immunity when used in this context.
IL-15 has been shown to play a prominent role in the generation and maintenance of CD8+ T cells (Zhang et al., 1998; Ku et al., 2000; Becker et al., 2002; Moore et al., 2002; Waldmann, 2002; Yajima et al., 2002; Sprent, 2003; Oh et al., 2004, 2008) as it signals through the common βγ chain, which is also used by IL-2 (Giri et al., 1994). IL-15 has a unique high-affinity receptor (Ka,~10−11M−1) chain, termed IL-15Rα (Anderson et al., 1995b). Systemically soluble IL-15 protein is virtually undetectable despite its constitutive mRNA expression by many cell types (Giri et al., 1995) and its expression is tightly controlled at various levels (Tagaya et al., 1996; Onu et al., 1997; Bamford et al., 1998; Waldmann and Tagaya, 1999). Interestingly, IL-15 has been shown to be trans-presented on the surface of antigen-presenting cells via IL-15Rα during the priming of natural killer and CD8+ T cells (Dubois et al., 2002; Koka et al., 2004; Lucas et al., 2007; Sato et al., 2007). IL-15Rα has also been shown to play a prominent role in the regulation of IL-15 secretion (Duitman et al., 2008). This cell surface complex allows IL-15 to signal through the βγ receptor on memory CD8+ T cells, promoting cell division and survival of these cells (Lodolce et al., 1998, 2001; Kennedy et al., 2000; Burkett et al., 2003, 2004; Sandau et al., 2004; Schluns et al., 2004a,b). IL-15 and IL-15Rα together as a complex exhibit enhanced stability and secretion compared with either molecule alone (Bergamaschi et al., 2008).
Regarding its employment in vaccination models, the use of plasmid-encoded IL-15 as an HIV-1 vaccine adjuvant has been previously reported to enhance cytolytic and memory CD8+ T cell responses in mice (Oh et al., 2003; Kutzler et al., 2005; Zhang et al., 2006; Calarota et al., 2008; Li et al., 2008). Studies in rhesus macaques have also shown the ability of IL-15 to enhance effector functions of CD4+ T cells (Picker et al., 2006) and rescue dual interferon (IFN)-γ/tumor necrosis factor (TNF) responses in both effector CD4+ and CD8+ T cells (Halwani et al., 2008). Importantly, addition of pIL-15 with SIV/HIV antigens in rhesus macaques resulted in enhanced protection after SHIV89.6p challenge (Boyer et al., 2007).
However, to our knowledge, no previous vaccination study has attempted to take advantage of the increased stability of the IL-15/IL-15Rα complex as a DNA vaccine adjuvant. To that end, we coimmunized mice with pIL-15 and pIL-15Rα to examine whether we could further enhance the immune response generated by IL-15 in a model using HIV-1 DNA vaccine antigens. Our data show that although the IL-15 and IL-15Rα combination indeed enhanced the overall cellular immune response, surprisingly the IL-15Rα plasmid augmented immune responses in an IL-15-independent manner. Importantly, the induced memory response was maintained only in mice covaccinated with pIL-15 as well as pIL-15Rα, but not with IL-15Rα alone. These studies demonstrate for the first time that the IL-15Rα protein alone can function as an adjuvant with a limited immune expansion phenotype.
Recombinant human IL-15Rα protein was generated by Abgent (San Diego, CA). Briefly, the open reading frame of human IL-15Rα (a generous gift from T. Waldmann, National Cancer Institute, National Institutes of Health, Bethesda, MD) was cloned into the high-expressing bacterial vector pET21a (EMD Biosciences, Gibbstown, NJ). Competent cells were transformed and amplified in Escherichia coli, and recombinant protein was purified with a nickel–nitrilotriacetic acid (Ni–NTA) column. The accuracy of the purified protein was confirmed by direct enzyme-linked immunosorbent assay (ELISA) using anti-human IL-15Rα antibody (R&D Systems, Minneapolis, MN).
Recombinant human IL-15Rα protein was injected into BALB/c mice (n=4) for monoclonal antibody generation. Five micrograms of total protein emulsified in complete Freund's adjuvant (first immunization only) or incomplete Freund's adjuvant (subsequent immunizations) was given per injection (Sigma-Aldrich, St. Louis, MO). Fifty microliters was injected subcutaneously into each flank and 100μl was injected into the peritoneum. Mice were given a final boost of 35μg of protein in sterile phosphate-buffered saline (PBS) intravenously 3 days before fusion. Antibody levels in the sera were determined by direct ELISA using recombinant IL-15Rα protein and horseradish peroxidase-conjugated anti-mouse IgG (IgG–HRP) (Zymed, San Francisco, CA). One mouse with a 1:8000 titer of antibody against IL-15Rα was sacrificed and its spleen removed for fusion with a myeloma cell line. A total of 1500 hybridoma supernatants were screened by ELISA, 8 positive clones were expanded, and 1 was purified by passage through an ammonium sulfate column, yielding antibody KK1.23. Monoclonal antibodies were generated and purified by J. Conicello (Wistar Institute Hybridoma Facility, Philadelphia, PA).
Western blotting analysis was performed according to standard protocols. Three micrograms per well of recombinant IL-15Rα or Vpr protein (Abgent) was run on a sodium dodecyl sulfate (SDS)–polyacrylamide gel (Cambrex, Rockland, ME), blotted on nitrocellulose membrane, and probed with either the R&D Systems or KK1.23 anti-human IL-15Rα antibody. The signal was amplified with anti-mouse IgG–HRP (Zymed) and detected by enhanced chemiluminescence (ECL) (GE Healthcare, Chalfont St. Giles, UK).
DNA vaccine constructs expressing HIV-1 Gag and HIV-1 Pol (Kim et al., 1998) and human IL-15 (Kutzler et al., 2005) were prepared as previously described. The open reading frame of human IL-15Rα was moved into pVAX1 and pTRACER vectors (Invitrogen, Carlsbad, CA). Restriction enzyme digestion using, respectively, EcoRI and BamHI or NheI and EcoRI (New England BioLabs, Beverly, MA) was performed. Positive clones were verified by sequence analysis.
The TnT-T7 quick coupled transcription/translation reticulocyte lysate system (Promega, Madison, WI) and [35S]methionine were used to create labeled IL-15Rα protein product. pVAX vector alone (negative control) or pVAX vector carrying IL-15Rα and [35S]methionine was added to the reaction mix according to the instructions supplied by the manufacturer. The reaction was carried out at 30°C for 1hr. Labeled proteins were immunoprecipitated with 5μg of purified monoclonal anti-IL-15Rα antibody (R&D Systems) or clone KK1.23 at 4°C with rotation overnight in radioimmunoprecipitation assay (RIPA) buffer. Approximately 5mg of protein G–Sepharose beads (GE Healthcare) (50μl of 100-mg/ml stock) was added to each immunoprecipitation reaction, and the samples were incubated at 4°C with rotation for 2hr. The beads were washed three times with binding buffer containing high salt and bovine serum albumin and finally resuspended in 2× sample buffer. Immunoprecipitated protein complexes were eluted from the Sepharose beads by boiling for 5min and were run on an SDS–12% polyacrylamide gel (Cambrex). The gel was fixed and treated with amplifying solution (GE Healthcare) and dried for 2hr in a gel drier (Bio-Rad, Hercules, CA). The dried gel was exposed to X-ray film at −80°C and developed with a Kodak automatic developer (Eastman Kodak, Rochester, NY).
The indirect immunofluorescence assay for confirmation of pIL-15Rα plasmid expression was conducted according to the following protocol as previously described (Ramanathan et al., 2002). HeLa cells (ATCC, Manassas, VA) grown in slide chambers (BD Biosciences, San Jose, CA) at a density of 100,000 cells per chamber in complete Dulbecco's modified Eagle's medium (DMEM) plus 10% fetal bovine serum (FBS) (Hyclone, Logan, UT) and antibiotic–antimycotic (GIBCO; Invitrogen) were allowed to adhere overnight. Cells were transfected with pIL-15Rα pTRACER or pVAX-1 (1μg/well), using FuGENE 6 transfection reagent (Roche Diagnostics, Basel, Switzerland) according to the manufacturer's protocol. Twenty-four hours after transfection, cells were washed with PBS and fixed on slides, using 2% paraformaldehyde (PFA)–PBS, for 1hr at room temperature. Slides were incubated with 5μg of clone KK1.23 mouse anti-human IL-15Rα made in our laboratory or with IgG1 isotype control (R&D Systems) for 90min at 37°C. Rhodamine-conjugated anti-mouse IgG secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was added at 1:200 and the slides were incubated for 45min at room temperature. After 4′,6-diamidino-2-phenylindole (DAPI) (Molecular Probes Invitrogen, Eugene, OR) staining for 10min at room temperature, slides were mounted in Fluoromount G medium (Electron Microscopy Sciences, Hatfield, PA) and analyzed with the phase 3 Image-Pro program for fluorescence microscopy (MediaCybernetics, Bethesda, MD).
The tibialis anterior muscle of 6- to 8-week-old female BALB/c (Jackson Laboratory, Bar Harbor, ME), C57BL/6 (Taconic, Germantown, NY), or IL-15 knockout (Taconic) (Kennedy et al., 2000) mice was injected three times, 2 weeks apart, and electroporated as previously described (Khan et al., 2003; Laddy et al., 2008) with the CELLECTRA adaptive constant current device (VGX Pharmaceuticals, The Woodlands, TX). For all experiments in mice, the animals were immunized with either 35μg of pVAX1, 5μg of HIV-1 antigenic plasmid (Gag, Pol), 10μg of pIL-15, and/or 7.5, 10, or 15μg of pIL-15Rα (n=3–7 per group). Coadministration of various gene plasmids involved mixing the designated DNA plasmids before injection in 0.25% bupivacaine-HCl (Sigma-Aldrich) in isotonic citrate buffer (Kim et al., 1998; Kutzler et al., 2005) to a final volume of 40μl. All animals were housed in a temperature-controlled, light-cycled facility at the University of Pennsylvania, and their care was performed under the guidelines of the National Institutes of Health and the University of Pennsylvania (Philadelphia, PA).
At time points designated in the immunization schedule, the animals were sedated with an analgesic and blood was taken before the animals were sacrificed by cervical dislocation. The spleen from each mouse was harvested and spleens were pooled (per experimental group) into a 15-ml conical tube containing R10 medium (RPMI 1640 plus 10% fetal bovine serum, antibiotic–antimycotic, and 2-mercaptoethanol). In a sterile tissue culture hood, the pooled spleen–medium mixture from each experimental group was crushed into a single-cell suspension, using a stomacher apparatus (Seward 80; Metrohm, Riverview, FL). The cell/tissue stroma were put through a 40-μm cell strainer and washed with R10, pelleted, and incubated for 5min at room temperature in ACK lysing buffer (Lonza, Basel, Switzerland) to lyse red blood cells. The splenocytes were then counted and used in immune assays described subsequently.
An IFN-γ enzyme-linked immunospot assay (ELISpot) was performed as previously described (Kutzler et al., 2005) to determine antigen-specific cytokine secretion from immunized mice. Briefly, ELISpot 96-well plates were coated with anti-mouse IFN-γ capture antibody and incubated for 24hr at 4°C (R&D Systems). Splenocytes (2×105) from immunized mice were added to each well of an ELISpot plate and stimulated overnight at 37°C, 5% CO2, in the presence of R10 (negative control), concanavalin A (positive control), or specific peptide (HIV-1 Gag or Pol) antigens (10μg/ml). HIV-1 consensus clade B subtype HIV-1 Gag and Pol 15-mer peptides spanning the entire respective protein, overlapping by 11 amino acids, were acquired from by the AIDS Reagent and Reference Repository (Frederick, MD). For CD8+ cell depletion experiments, CD8+ T cells were removed from total splenocytes by positive magnetic selection using an anti-CD8a (Ly-2) antibody (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's protocol. After 24hr of stimulation, the plates were washed and incubated at 4°C overnight with biotinylated anti-mouse IFN-γ antibody (R&D Systems). The plates were washed and incubated with streptavidin–alkaline phosphatase (R&D Systems) for 2hr at room temperature. The plates were washed, and 5-bromo-4-chloro-3′-indolyl phosphate p-toluidine salt (BCIP) and nitroblue tetrazolium chloride (NBT) (the chromogen color reagent; R&D Systems) were added. The plates were rinsed with distilled water, and dried at room temperature. Spots were counted with an automated ELISpot reader (Cellular Technology Limited [CTL], Cleveland, OH). Raw values were determined and multiplied by a factor of 5 so that data are represented as spot-forming cells per million splenocytes. Background values in the R10 wells of each group were subtracted from peptide-stimulated wells before graphing.
HIV-1-specific T cell responses were also determined by intracellular cytokine staining, using a Cytofix/Cytoperm kit (BD Biosciences) according to the standard protocol. Splenocytes from immunized mice were stimulated for 5hr in the presence of GolgiPlug (1μl/ml; BD Biosciences) with R10 and dimethyl sulfoxide (DMSO; negative control), phorbol myristate acetate (PMA, 10ng/ml) and ionomycin (250ng/ml; positive control), or HIV-1 consensus clade B Gag or Pol 15-mer peptides. Before surface staining, cells were stained, using a LIVE/DEAD fixable violet dead cell stain kit (Molecular Probes Invitrogen), at 37°C for 10min and Fc Block (BD Biosciences) was added for 15min at 4°C to block Fc receptors. All antibodies were purchased from BD Biosciences and used at 1μl/test. Before permeabilization/fixation cells were stained with CD4–Alexa 700 and CD8–PerCP (peridinin chlorophyll protein) for 30min at 4°C. CD3–PECy5 (phycoerythrin, cyanine 5) and IFN-γ–PECy7 (phycoerythrin, cyanine 7) were included in the intracellular stain for 45min at 4°C. Data from 50,000 live CD3+ lymphocyte-gated events were acquired with an LSRII flow cytometer (BD Biosciences) and analyzed with FlowJo software (Treestar, Ashland, OR). Responses from the negative control wells were subtracted from the antigenic stimulations before graphing.
ELISA was used to determine HIV-1 Gag-specific IgG antibodies in mouse sera as described (Ogawa et al., 1989; Mestecky et al., 2004). EIA/RIA plates (Corning Costar, Cambridge, MA) were coated with a 1-μg/ml concentration of recombinant HIV-1IIIB Gag p24 (Immunodiagnostics, Woburn, MA) diluted in PBS (Mediatech, Herndon, VA) at a final volume of 100μl/well and incubated overnight at 4°C. Plates were washed with PBS–Tween (0.05% Tween 20) and blocked against nonspecific binding with 200μl of blocking buffer–diluent (3% BSA in PBS) for 2hr at room temperature. The plates were washed and dilutions of pooled sera from immunized mice were added in triplicate (100μl/well), at dilutions ranging from 1 to 10 to 1 to 1600 and incubated at room temperature for 2hr. Bound antibodies were detected with horseradish peroxidase-labeled goat anti-mouse IgG(H+L) (Zymed) and developed with substrate 3,3′,5,5′-tetramethylbenzidine (TMB) (Sigma-Aldrich). The color reaction was stopped with 2 N H2SO4, and the absorbance at 450nm was determined with an EL312 Bio-Kinetics microplate reader (Bio-Tek Instruments, Winooski, VT).
To detect expression of the IL-15Rα plasmid (pIL-15Rα) on cells, we first set out to generate a monoclonal antibody against human IL-15Rα as commercially available antibodies are deficient in this activity. Recombinant human IL-15Rα was generated as described in Materials and Methods. To confirm the size of the newly generated IL-15Rα protein, decreasing dilutions of purified protein were run on an SDS–polyacrylamide gel and stained with Coomassie blue dye (Fig. 1A). As shown in Fig. 1A, the generated protein runs at approximately 30kDa, the expected size. We next tested the ability of this protein to bind to commercially available antibody as an indication of its correct integrity. Figure 1B shows an ELISA with plates captured with recombinant IL-15Rα or Vpr protein, a negative control. Vpr was used as it was produced by a similar method as the IL-15Rα protein. Figure 1B shows that the commercially available anti-human IL-15Rα antibody can detect the generated recombinant protein. To generate an antibody against IL-15Rα, recombinant human IL-15Rα protein was injected into BALB/c mice as described in Materials and Methods and in Fig. 1C. After screening approximately 1500 hybridoma supernatants by ELISA, one hybridoma KK1.23 exhibited titers of antibody (≥1 to 12,800) as shown in Fig. 1D. This hybridoma was subsequently cloned, expanded, and purified. Purified antibody KK1.23 is specific for human IL-15Rα as shown by Western blot analysis in Fig. 1E. In addition, KK1.23 appears to bind to human IL-15Rα with higher affinity than does the commercially available antibody (Fig. 1E).
We next set out to create an IL-15Rα expression vector suitable for use in vaccination studies. The human IL-15Rα open reading frame (ORF) was cloned into the pVAX1 expression vector as shown in Fig. 2A, under the control of the cytomegalovirus (CMV) promoter. To assess appropriate expression of the IL-15Rα plasmid, an in vitro translation assay was carried out. The 35S-radiolabeled protein is shown in Fig. 2B and C migrating at roughly 30.0kDa, whereas the control plasmid, pVAX, did not yield any detectable protein product as expected. The commercial R&D Systems (Fig. 2B) or the KK1.23 (Fig. 2C) antibody against human IL-15Rα was used to immunoprecipitate the radiolabeled protein. To confirm expression of the plasmid IL-15Rα, an immunofluorescence assay was also performed with the KK1.23 antibody. For this assay, the ORF of human IL-15Rα was cloned into the pTRACER expression vector, which also encodes the green fluorescent protein (GFP) reporter. Therefore, cells fluorescing green (Fig. 2E–G) also express pIL-15Rα, and the KK1.23 anti-human IL-15Rα is detected with anti-mouse IgG–PE (red). The untransfected control is shown in Fig. 2D, and the isotype control in Fig. 2E. The data illustrate both the ability of the pIL-15Rα plasmid to express as well as the ability of the anti-human pIL-15Rα antibody to detect the translated protein product. Clearly, the pIL-15Rα plasmid encodes a conformationally accurate and surface-localized protein.
To examine the ability of pIL-15Rα to enhance immune responses as compared with pIL-15, BALB/c mice were immunized intramuscularly in the tibialis anterior muscle accompanied by in vivo electroporation, according to the schedule shown in Fig. 3A. Mice were immunized with either pVAX control vector or with 5μg of antigenic construct (HIV-1 Gag, HIV-1 Pol) with 10μg of pIL-15, 15μg of pIL-15Rα, or both pIL-15 and pIL-15Rα in a final volume of 40μl. These doses were predetermined to give optimal responses in preliminary studies (data not shown). As shown in Fig. 3B, immunization with antigenic constructs alone resulted in 2300 spot-forming cells (SFCs) per 106 splenocytes as measured by IFN-γ ELISpot. The addition of pIL-15 enhanced the response to 3800 SFCs, whereas coimmunization with pIL-15Rα and pIL-15 exhibited the most dramatic increase relative to the antigenic group alone, resulting in 5900 SFCs. These results support the idea that formation of the IL-15/IL-15Rα immune complex can serve as a more potent adjuvant than IL-15 alone.
To determine whether this immune complex was truly being formed in vivo, we added another immunization group in which pIL-15 and pIL-15Rα were injected (with antigen) in separate legs. In this split delivery method, plasmid-delivered IL-15 and IL-15Rα would be unable to form an immune complex. We found that coimmunization of pIL-15 and IL-15Rα in separate legs elicited levels of IFN-γ similar to those observed with the same combination delivered in the same leg (4562 vs. 4072 SFCs, respectively).
Somewhat unexpectedly, the immunization group with antigenic construct and pIL-15Rα also augmented antigen-specific IFN-γ secretion to approximately 3500 SFCs (Fig. 3B). To confirm these results, we immunized a new set of mice with increasing doses of pIL-15Rα plasmid in conjunction with antigenic constructs to determine whether pIL-15Rα would induce responses in a dose-dependent fashion. As shown in Fig. 4, the inclusion of pIL-15Rα did enhance the induced IFN-γ secretion in a dose-dependent fashion, in measured responses against pGag (Fig. 4A) or pPol (Fig. 4B). Regardless of the HIV-1 antigenic construct used, coimmunization with pIL-15Rα augmented cellular immune responses by 1.5- to 2-fold at the highest dose used. These results reveal an interesting phenomenon whereby IL-15Rα may enhance antigen-specific immune responses even in the absence of IL-15, which advances its role as a novel adjuvant.
To further confirm the adjuvant properties of pIL-15Rα, we examined the effector functions of CD4+ and CD8+ T cells after vaccination. Accordingly, we depleted CD8+ T cells from splenocytes of mice immunized with each vaccine combination previously mentioned, before carrying out the IFN-γ ELISpot assay. As shown in Fig. 5A, depletion of CD8+ T cells from the splenocytes of mice immunized with either pIL-15, pIL-15Rα, or the combination of both significantly decreased the amount of IFN-γ secretion detected. There was no difference between the CD8− (presumably CD4+ T) cell contribution (Fig. 5A, gray columns) in any of the immunized groups, compared with the total responses observed in whole splenocytes (Fig. 5A, solid columns). Taken together, the combination of pIL-15Rα and pIL-15 in a vaccination strategy greatly enhances the immune response relative to either construct delivered alone. This additive effect acts primarily on CD8+ T cells, as the effect was lost with the depletion of this cell population.
To determine whether pIL-15Rα would also have an effect on humoral immune responses, we also measured antibody responses elicited through each vaccination strategy by ELISA. Sera from immunized mice was assayed to measure the levels of IgG antibodies against the HIV-1 Gag (p24) protein (Fig. 5B). Although the combination of pIL-15 and pIL-15Rα was the best at eliciting cellular immunity, mice immunized with either pIL-15 or pIL-15Rα alone had the highest titers of HIV-1-specific antibodies (1:1600) compared with mice immunized with pVAX (not detected), pGag alone, or the combination (1:800).
As IL-15 has been shown to play a role in the maintenance of CD8+ T cell memory, we were interested in examining the effect of IL-15Rα adjuvant in a similar fashion. To study this, mice were immunized three times as previously mentioned; however, instead of sacrificing these animals 1 week after the third immunization, they were allowed to rest for approximately 30 weeks to be sure the responses observed would be contributed primarily by the memory population. As shown in Fig. 6A, the responses after a significant rest period were still quite robust. Mice immunized with the antigenic construct alone had responses of about 1700 SFCs. The highest responses were clearly in groups of mice coimmunized with pIL-15: ~2800 SFCs for both pIL-15 and the pIL-15/pIL-15Rα combination. In mice coimmunized with pIL-15Rα in the absence of pIL-15, an adjuvant effect was no longer observed (~1700 SFCs). The same trends were also observed by intracellular cytokine staining and flow cytometry (Fig. 6B), in which the level of IFN-γ production by CD8+ T cells was most pronounced in mice coimmunized with pIL-15. The inclusion of pIL-15Rα in our vaccination strategy had little effect on memory responses, whereas pIL-15 was critical. These data support the idea that whereas pIL-15Rα was a robust adjuvant early after vaccination, over time IL-15 is a better inducer of memory CD8+ T cells.
The memory antibody response was similar to that observed during the effector phase. As shown in Fig. 6C, mice immunized with pIL-15 had detectable antibodies against HIV-1 Gag (p24) at dilutions out to 1:1600, whereas all other groups, including the combination of pIL-15/pIL-15Rα, diluted out at 1:400. These data suggest that pIL-15 is an effective adjuvant in the generation of humoral as well as cellular memory responses, whereas pIL-15Rα is not, but instead likely plays a unique role in accelerating the acute immune response to antigen.
We reasoned that it was possible that the transfected and subsequently translated human IL-15Rα protein could be augmenting immune responses in vaccinated mice by forming complexes with endogenous murine IL-15, or independently of IL-15. We had previously studied the sera from IL-15Rα-immunized mice for the induction of anti-IL-15Rα antibodies and observed no induction of such antibody responses, ruling out a more trivial explanation of these results. To determine whether IL-15Rα acts as an adjuvant together with endogenous IL-15 we studied vaccination in IL-15 knockout mice. As a control, we first tested whether the translated human IL-15Rα protein could bind to mouse IL-15. As shown in Fig. 7A, 35S-radiolabeled human IL-15Rα protein incubated with murine IL-15 was able to be immunoprecipitated with an anti-mouse IL-15 antibody, suggesting the ability of murine IL-15 to bind to human IL-15Rα.
We next examined the ability of pIL-15Rα to act as an adjuvant in the absence of murine IL-15. Accordingly, we conducted the same vaccination studies in IL-15 knockout mice, which lack endogenous IL-15 and as a result have a deficiency in natural killer (NK) and memory CD8+ T cells (Kennedy et al., 2000). Figure 7C shows that pIL-15Rα enhances the immune responses in the absence of endogenous murine IL-15 in the knockout mice, as determined by IFN-γ ELISpot. Furthermore, the combination of pIL-15 and pIL-15Rα fails to further enhance the immune response by either adjuvant administered alone as initially observed in BALB/c mice. The IL-15−/− mice were generated on a C57/BL6 black-6 background from Taconic. Therefore, to verify that we would obtain similar responses in the appropriate background control mice as were observed in BALB/c, mice the same experiments were repeated. Figure 7B shows the results from the control mice immunized according to the identical schedule and shows the same trend as BALB/c immunized mice. From these data we conclude that pIL-15Rα is able to act on CD8+ T cells independently of pIL-15, via an unknown and novel mechanism.
Previous studies have shown the effectiveness of codelivering plasmid IL-15 in the DNA vaccine arena to enhance immune responses in both mice (Oh et al., 2003; Kutzler et al., 2005; Zhang et al., 2006; Calarota et al., 2008; Li et al., 2008) and macaques (Picker et al., 2006; Boyer et al., 2007; Halwani et al., 2008). These studies primarily show the effect of IL-15 on the enhanced function and survival of antigen-experienced CD8+ T cells. Because IL-15 has been shown to be a potent inducer of CD8+ T cell function we sought to further enhance these effects with the inclusion of the unique high-affinity receptor α chain, IL-15Rα. Other reports have shown the increased function and stability of this complex (Bergamaschi et al., 2008) and we further extended this concept by codelivering plasmid-encoded IL-15 and IL-15Rα as a potential “super” adjuvant in the vaccine setting.
In this article, we describe the use of a plasmid-encoded human IL-15Rα construct that was first tested for expression by in vitro translation and detectable in transfected cells by immunofluorescence with an antibody generated in our laboratory. Coimmunization of this IL-15Rα construct together with our human IL-15 and HIV-1 antigenic DNA constructs resulted in levels of IFN-γ secretion that were 2.5-fold more potent than were achieved by immunization with the antigenic constructs alone (Fig. 3B). The IFN-γ secretion was attributable to CD8+ T cells, as the depletion of these cells before plating for the ELISpot assay resulted in 10-fold less IFN-γ secretion. The increased potency observed with the combined delivery of IL-15 and IL-15Rα is not likely due to the formation of a stable complex in transfected cells, as injecting pIL-15 and pIL-15Rα into separate legs (with antigen) also elicited immune responses similar to delivery in the same leg. It is therefore more likely that the enhanced response observed with the codelivery of pIL-15 and pIL-15Rα is an additive effect of two independent adjuvants. Codelivery of these two adjuvants did not appear to further enhance humoral immune responses as measured by IgG antibodies in the sera.
Somewhat surprisingly, we found that delivery of the antigenic plasmid with pIL-15Rα also augmented cellular immune responses, equal to those elicited by pIL-15. To be certain, we performed immunizations with increasing amounts of pIL-15Rα and observed dose-dependent responses. We thought it may be possible that our human IL-15Rα protein was able to bind to endogenous murine IL-15 and trans-present it in a similar fashion because murine IL-15 is approximately 73% identical to the human IL-15 (Anderson et al., 1995a). To test this hypothesis, we immunized IL-15 knockout mice, which lack any endogenous IL-15. Interestingly, we observed an approximately 2-fold increase in the IL-15 knockout mice immunized with pIL-15Rα relative to antigen alone. As IL-15−/− female mice are notoriously difficult to acquire in large numbers between the ages of 6–8 weeks, we decided to exclude the pIL-15 group from these experiments. However, the enhanced effect of the combination of pIL-15 and pIL-15Rα was no longer observed. It should be noted that control C57/BL6 mice exhibited the same trends that were seen in BALB/c mice (albeit lower total spot counts and higher background in the pVAX group) and that the IL-15 knockout mice had overall lower responses even compared with the C57BL/6 mice. These mice have been previously described to have somewhat defective host defense responses including the inability to protect against a vaccinia challenge (Kennedy et al., 2000). Regardless of the overall lower immune responses, the adjuvanting effect of pIL-15Rα was still observed in the absence of any endogenous IL-15. Taken together, we propose that IL-15Rα can serve as a novel adjuvant capable of eliciting responses independently of IL-15, which to our knowledge has not been previously reported. This amplification of the immune response appears particularly focused on immune expansion during the acute phase rather than the memory phase of the host T cell response.
We were also interested in examining the long-term effects of the combined pIL-15/pIL-15Rα on the immune response. To look at memory responses we waited 30 weeks after the third immunization before carrying out immune analysis. We found that although the combination of these two adjuvants elicits potent CD8+ T cell responses in the early phase of an immune response, the effect on memory T cells is primarily observed only in mice immunized along with pIL-15. The inclusion of pIL-15 was absolutely necessary for an enhanced memory immune response relative to the antigen alone. Similarly, IL-15Rα initially elicited immune responses equal to or greater than that of IL-15, but it did not help to sustain the memory response. Therefore, although IL-15Rα expanded burst size, burst size in the absence of the IL-15 signal for memory was not enough to sustain a long-term response. This finding opens several questions and possibilities for immune modulation. For example, is it possible that IL-15Rα binds a second molecule? Does the overexpression of IL-15Rα increase formation and subsequent activation through the heterotrimeric complex including the β and γ chains? Alternatively, could overexpression of the IL-15Rα molecule result in dimerization and signaling through the α chain in the absence of IL-15 and the other two receptor chains? As the cytoplasmic tail of the IL-15Rα molecule is extremely short in length, it is typically not thought to play a role in direct signaling (Anderson et al., 1995b). However, reports have shown that IL-15Rα is indeed capable of signaling independently of the other two receptor subunits, specifically through the tyrosine kinase molecule Syk (Anderson et al., 1995b; Bulanova et al., 2001). In addition, Wu et al. (2008) performed an elegant study where they created a chimeric molecule of the IL-15Rα extracellular region and the intracellular region of the IL-2 receptor alpha chain (248). They next created knock-in mice expressing this construct and observed diminished natural killer, natural killer T cells, and CD8+ T cell similar to IL-15 and IL-15Rα knockout mice. These mice expressed wild-type-levels of IL-15Rα and were able to trans-present IL-15 to neighboring cells as well. These data show that the cytoplasmic domain of IL-15α is clearly important for IL-15 function and may have a role in signaling. The exact mechanism by which IL-15Rα enhances the immune response warrants further research, especially in the absence of IL-15.
In summary, we show here that the combination of cytokine IL-15 and its unique receptor α may be useful as a strategy for generating and maintaining memory CD8+ T cells in a model DNA vaccine. It would also be beneficial to test this combination in other vaccine models, such as cancer, in which immune responses against self tumor-associated antigens are often weak and enhancing CD8+ T cell responses is likely important. Further studies using IL-15Rα and IL-15 in various prime–boost strategies to enhance both burst size and memory populations could provide important insight into this important immune expansion receptor system. This cytokine and receptor complex may even work better in other delivery strategies and warrants exploration.
The authors acknowledge helpful discussions with Jean Boyer, Ph.D. (University of Pennsylvania, Philadelphia, PA), and Julia Conicello of the Wistar Institute Hybridoma Facility (Philadelphia, PA) for help in generating and purifying the KK1.23 monoclonal antibody. The authors also thank the Penn Pathology Flow Cytometry and Biomedical Imaging core facilities. K.A.K. acknowledges the support of NIH grants 1-T32-A107632. D.B.W. acknowledges support from the NIH, including HIVRAD funding.
The laboratory notes possible commercial interests associated with this work, which may include the following: Wyeth, VGX, BMS, Virxsys, Ichor, Merck, Althea, and Aldeveron.