ABSTRACT
Most DNA-encoded adjuvants enhance immune responses to DNA vaccines in small animals but are less effective in primates. Here, we characterize the adjuvant activity of the catalytic A1 domain of cholera toxin (CTA1) for human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) antigens in mice and macaques delivered by GeneGun. The inclusion of CTA1 with SIVmac239 Gag dramatically enhanced anti-Gag antibody responses in mice. The adjuvant effects of CTA1 for the secreted antigen HIV gp120 were much less pronounced than those for Gag, as the responses to gp120 were high in the absence of an adjuvant. CTA1 was a stronger adjuvant for Gag than was granulocyte-macrophage colony-stimulating factor (GM-CSF), and it also displayed a wider dose range than GM-CSF in mice. In macaques, CTA1 modestly enhanced the antibody responses to SIV Gag but potently primed for a recombinant Gag protein boost. The results of this study show that CTA1 is a potent adjuvant for SIV Gag when delivered by GeneGun in mice and that CTA1 provides a potent GeneGun-mediated DNA prime for a heterologous protein boost in macaques.
INTRODUCTION
Most DNA vaccines have translated poorly from mice to primates. The exact reasons for this loss of potency are not fully elucidated. This topic is discussed in detail in references 11, 13, 31, and 32. Two strategies to improve the immunogenicity of DNA vaccines for cellular and/or humoral immune responses have evolved in recent years. One utilizes enhanced delivery of the DNA using ballistic particle bombardment with the GeneGun (14) or in vivo electroporation (3). The other main strategy entails the use of plasmid mixtures that combine the DNA vaccine of interest with a plasmid that expresses an adjuvant. The best-studied DNA adjuvants are based on cytokines (9, 25, 29, 30, 48) or chemokines (33, 53, 54). Although several of these adjuvant approaches have shown great promise in small animal models, most have been disappointing in humans and nonhuman primates (11, 13, 32).
Cholera toxin (CT) and the highly related heat-labile enterotoxin of Escherichia coli (LT) have been found to be powerful mucosal immunogens and adjuvants (reviewed in reference 34). In mice, antibody (Ab) responses to CT and bystander antigens last at least 2 years (35, 50). We previously showed that the adjuvant properties of CT can be incorporated into DNA vaccines delivered intramuscularly (i.m.) (7). CT and LT are AB5 enterotoxins produced by Vibrio cholerae and E. coli, respectively. CT and LT consist of catalytic A subunits (separated into A1 and A2 subdomains) anchored in rings of five identical B subunits (43). Their enzymatic active sites reside within the A1 subdomains, while the A2 subdomains anchor the A subunits into the B pentamers (43). The B pentamers of CT and LT bind to gangliosides on cell membranes and facilitate their entry into lysosomes (23). The A1 subdomains are cleaved from the A2 subdomains in the Golgi complex or the endoplasmic reticulum (ER), and then the A1 subdomains exploit host protein retention and degradation pathways to gain access to the cell cytoplasm (reviewed in reference 22).
The A1 subunits of both CT and LT (CTA1 and LTA1) catalyze the transfer of an ADP-ribose from NAD to stimulatory α subunits of G proteins (Gsα). The ADP-ribosylation of Gsα locks it into the GTP-bound form and subsequently leads to the constitutive activation of adenylate cyclase and a sustained increase in intracellular cyclic AMP (cAMP) concentrations (8). The ADP-ribosyltransferase activities of CTA1 and LTA1 are responsible for their toxicity and for most, if not all, of their adjuvant effects (1, 2, 5–7, 15, 36, 46). Although Gsα is the primary target of CTA1 and LTA1, they also ADP-ribosylate other small G proteins (10). It is not known which of the targets of CTA1 and LTA1 are responsible for toxicity and/or adjuvant effects.
Unfortunately, the CT and LT holotoxins are extremely toxic to humans when administered at mucosal sites. Fortunately, the toxic properties of CT and LT can be avoided by immunizing at sites distant from the mucosa such as the skin (16–20). Toxicity can also be avoided by divorcing their A1 domains from their promiscuously binding B pentamers and retargeting the A1 domains using plasmid DNA (7) or other proteins (1, 2).
We undertook the present study to better understand the capabilities of CTA1 as a DNA-based adjuvant by delineating the dose and number of inoculations required to generate durable responses after GeneGun administration. For the first time, we also tested a toxin-based adjuvant head to head against a prevailing DNA vaccine adjuvant (granulocyte-macrophage colony-stimulating factor [GM-CSF]). We further characterized the adjuvant effects of CTA1 for simian immunodeficiency virus (SIV) Gag in macaques and characterized the ability of DNA vaccination by GeneGun with and without CTA1 to prime for a recombinant SIV Gag protein boost. The results indicate that GeneGun immunization requires at least three immunizations to induce maximal antibody responses in the presence or absence of CTA1 and that higher doses of CTA1 yield the best adjuvant effects. The results also show that CTA1 is a more potent adjuvant for the antigen SIV Gag than the “benchmark” adjuvant GM-CSF when delivered by GeneGun in mice. We additionally noted that GeneGun immunization with these antigens induces antibody responses that are strongly Th2 polarized and that the inclusion of the adjuvant CTA1 or GM-CSF does not alter this polarization. Finally, the results indicate that CTA1 potently enhances a GeneGun-delivered DNA prime for a heterologous protein boost in macaques.
MATERIALS AND METHODS
DNA vaccine construction.The pMAX-PRO DNA vaccine vector was created by replacing the yellow fluorescent protein (YFP) gene from pMAX-FP-Yellow (Amaxa, Cologne, Germany) with a multiple cloning site using the NheI and BssHII sites. The DNA sequences of CTA1 and LTA1 (minus the bacterial secretion signals), SIVmac239 Gag, and human immunodeficiency virus (HIV) HIVbal gp120 were codon optimized for expression in human and mouse cells by GeneArt (Regensburg, Germany). The optimized sequences that were provided in the GeneArt shuttle vector were subcloned into pMAX-PRO using the KpnI and XhoI sites. The pRC/CMV-CTA1 vector expressing the native CTA1 sequence was described previously (7).
CREB-Luc reporter assay.Fifty thousand 293 cells/well were plated in 96-well opaque-wall plates (Nunc, Rochester, NY). The next day, individual wells of cells were transfected using 270 nl/well of Fugene 6 transfection reagent (Roche, Basel, Switzerland) mixed with 8 ng of the pCREB-Luc plasmid (Lambda Biotech, St. Louis, MO); the indicated amount of empty pMAX-PRO, pRC/CMVCTA1 (expressing the native CTA1 sequence) (14), pMAX-PRO-CTA1 (expressing the codon-optimized CTA1), or pMAX-PRO-LTA1 (expressing the codon-optimized LTA1); and enough empty pMAX-PRO plasmid to keep the total transfection/well amount at 28 ng of DNA. Twenty-four hours later, all cells were lysed and assessed for luciferase production by light emission using a Bright-Glo luciferase kit (Promega, Madison, WI). Light emission was measured using a DTX 880 multimode detector (Beckman Coulter, Brea, CA).
GeneGun bullet preparation.GeneGun bullets were prepared according to the manufacturer's instructions (Bio-Rad Laboratories Inc., Hercules, CA). Bullets for the mouse studies were loaded with 0.67 mg of 1.0-μm gold beads, 0.5 μg of the pMAX-PRO-SIVmac239-gag plasmid, 0.5 μg of the pMAX-PRO-balgp120 plasmid, and a total of 0.5 μg of pMAX-PRO-CTA1 and/or empty pMAX-PRO. Bullets for the macaque study were loaded with 4.5 mg of 1.0-μm gold beads, 3.75 μg of the pMAX-PRO-SIVmac239-gag plasmid, and 3.75 μg of pMAX-PRO-CTA1 or pMAX-PRO. Ten bullets were given to each macaque at each immunization.
Vaccination procedures.DNA-coated gold beads were delivered to the shaved abdomen of mice or to the shaved upper quadriceps of macaques by using a helium-driven GeneGun (Bio-Rad Laboratories Inc.) with a discharge pressure of 400 lb/in2. Sufficient numbers of bullets were administered to achieve the indicated doses. The SIV Gag protein boost of the macaques was performed as follows: SIVmac251 rp27 Gag was purchased from Immunodiagnostics (Woburn, MA). The rp27 Gag was mixed with injection alum adjuvant according to the manufacturer's instructions (Thermo Scientific, Waltham, MA). Immediately before injection, the rp27 Gag-alum mixture was further mixed with CpG DNA (ODN2006) (Invivogen, Carlsbad, CA). Each macaque was immunized intramuscularly (i.m.) with 100 μg of rp27 Gag and 250 μg of CpG DNA mixed in alum.
Macaque study.Each macaque (4 macaques/group) was anesthetized and then immunized with 37.5 μg of pMAX-PROSIVmac239-gag DNA plus 37.5 μg of empty pMAX-PRO (group 1) or 37.5 μg of pMAX-PROSIVmac239-gag DNA plus 37.5 μg of pMAX-PRO-CTA1 DNA (group 2). The macaques were immunized by GeneGun three times on days 0, 14, and 76. The macaques were then allowed to rest, and then they, along with a third group of four macaques that had not been immunized with DNA, were boosted with recombinant SIV Gag protein (100 μg of rp27 Gag and 250 μg of CpG DNA mixed in alum adjuvant) i.m. on day 407.
Animal housing and handling.Mice were housed in the Advanced BioScience Laboratories (ABL) animal facility in Rockville, MD. The facility is AAALAC-International accredited and U.S. Department of Agriculture (USDA) registered and has a category 1 assurance from the Office of Laboratory Animal Welfare (assurance no. A3467-01). The animal research facility complies with USDA regulations pertaining to primate care (USDA registration no. 51-R-0059) and with the PHS policy on humane care and use of laboratory animals. The facility complies with all policies of the Guide for the Care and Use of Laboratory Animals (National Academy of Sciences, 1996). The GeneGun portion of the macaque study was performed while the animals were housed at the Institute of Human Virology (IHV) (University of Maryland, Baltimore, MD) animal facility. The IHV facility is AAALAC-International accredited and USDA registered and has a category 1 assurance from the Office for Protection from Research Risks (assurance no. A-4181-01). The IHV animal core facility complies with USDA regulations pertaining to primate care (USDA registration no. 51-R-0060) and with the PHS policy on humane care and use of laboratory animals. The IHV animal core facility complies with all policies of the Guide for the Care and Use of Laboratory Animals (National Academy of Sciences, 1996). The macaques were transferred to the animal facility at ABL approximately on day 300 after GeneGun delivery of vaccine preparations. The recombinant Gag booster immunizations were performed at ABL after the quarantine period. All animal studies were conducted in accordance with Institutional Animal Care and Use Committee-approved protocols (no. 004-97 for ABL or no. JB0020-06 for the IHV) and the NIH Guide for the Care and Use of Laboratory Animals (NIH publication no. 85-23, 1985).
Humoral responses to SIV Gag.Solid-phase enzyme-linked immunosorbent assay (ELISA) was used to determine SIV Gag-specific antibody (Ab) titers in the sera of mice or macaques. Briefly, 96-well microtiter plates (Nunc, Rochester, NY) were coated with 100 μl of 10-μg/ml SIVmac251-rp27 Gag (Immunodiagnostics) in phosphate-buffered saline (PBS) overnight at 4°C. Plates were washed three times with Tris-buffered saline (TBS) and then blocked with 100 μl of Blotto (5% [wt/vol] nonfat dried milk in TBS) at room temperature for 30 min. Serially diluted sera were added to the wells, incubated at room temperature for 1 h, and then washed three times with TBS. Peroxidase-conjugated goat anti-mouse IgG or goat anti-monkey IgG (Kirkegaard & Perry, Gaithersburg, MD) diluted 1/1,000 in Blotto (100 μl/well) was added and incubated at room temperature for 1 h. The plates were washed three times with TBS before addition of tetramethylbenzidine (TMB) peroxidase substrate (Kirkegaard & Perry) (100 μl/well) and incubation for 3 to 5 min. The reaction was stopped by adding 50 μl/well of 1 N H2SO4. Absorbance was read at 450 nm using a Beckman Coulter AD 200 plate reader (Brea, CA). Half-maximal serum binding titers were calculated using Sigmaplot 10 software.
Humoral responses to HIV gp120.Solid-phase ELISA was used to determine HIV gp120-specific Ab titers in the sera of mice. Briefly, 96-well microtiter plates were coated with 200 ng/ml of the sheep capture antibody D7324 (Aalto Scientific) (40, 41) (100 μl/well) overnight at 4°C. After 3 washes with TBS, 1 μg/ml of HIVbal isolate gp120 (purified recombinant) in PBS (100 μl/well) was added for 1 h at 37°C. Plates were washed three times with TBS and then blocked with 100-μl/well Blotto at room temperature for 30 min. Serially diluted sera were added to the wells, incubated at room temperature for 1 h, and then washed three times with TBS. Peroxidase-conjugated goat anti-mouse IgG diluted 1/1,000 in Blotto (100 μl/well) was added and incubated at room temperature for 1 h. The plates were washed three times with TBS before addition of TMB peroxidase substrate (100 μl/well) and incubation for 3 to 5 min. The reaction was stopped by adding 50 μl/well of 1 N H2SO4. Absorbance was read at 450 nm using a Beckman Coulter AD 200 plate reader. Half-maximal serum binding titers were calculated using Sigmaplot 10.
Determination of cytokine secretion. (i) Cytokine bead array (CBA).Mouse splenocytes were resuspended in complete RPMI 1640 medium containing either 50 μg/ml phytohemagglutinin M (PHA-M) (Sigma), peptide pools (15-mers overlapping by 11 amino acids; 1 mM [each] final peptide concentration) spanning HIVbal gp120, SIVmac239 Gag, or medium alone and plated in 96-well flat-bottom tissue culture plates (Becton Dickinson). Input cell numbers were 4 × 105 cells/well and were assayed in duplicate wells. Cells were incubated for 22 to 24 h at 37°C, and then supernatants were removed for CBA. CBA was performed on the supernatants per the manufacturer's instructions (BD-Pharmingen), and data were collected using a FACSCalibur flow cytometer (BD Biosciences). CBA was used to determine the secretion of the Th1 cytokines interleukin-2 (IL-2), tumor necrosis factor alpha (TNF-α), and gamma interferon (IFN-γ) and the Th2 cytokines IL-4 and IL-5.
(ii) ELISPOT assay.Ninety-six-well flat-bottomed enzyme-linked immunospot (ELISPOT) plates (Millipore, Bedford, MA) were coated overnight with an anti-mouse IFN-γ monoclonal antibody (BD-Pharmingen) at a concentration of 10 μg/ml, after which the plates were washed three times and then blocked for 2 h with PBS containing 5% heat-inactivated fetal bovine serum (FBS). Mouse splenocytes were resuspended in complete RPMI 1640 medium containing either 50 mg/ml PHA-M (Sigma) or peptide pools (15-mers overlapping by 11 amino acids; 1 mM [each] final peptide concentration) spanning HIVbal gp120, SIVmac239 Gag, or medium alone. Input cell numbers were 4 × 105 cells/well, and cells were assayed in duplicate wells. Cells were incubated for 22 to 24 h at 37°C and then removed from the ELISPOT plates by first being washed with deionized water and then being washed 6 times with PBS containing 0.25% Tween 20 and 3 additional times with PBS. Plates were then treated with a biotinylated anti-mouse IFN-γ detection antibody (0.5 μg/well; BD-Pharmingen) and incubated at room temperature for 2 h. ELISPOT plates were then washed 10 times with PBS containing 0.25% Tween 20, treated with 100 μl per well of streptavidin-horseradish peroxidase conjugate (BD-Pharmingen) diluted 1:100, and incubated for an additional 1 h at room temperature. Unbound conjugate was removed by rinsing the plate 10 times with PBS containing 0.25% Tween 20. Chromogenic substrate (100 μl/well; AEC chromogen [BD-Pharmingen]) was then added for 3 to 5 min before being rinsed away with water, after which the plates were air dried and the resulting spots were counted using an Immunospot reader (CTL Inc., Cleveland, OH). Peptide-specific IFN-γ ELISPOT responses were considered positive if the response (minus medium background) was >3-fold above the medium response and ≥50 spot-forming cells (SFC)/106 cells.
(iii) Intracellular cytokine staining.On days 30, 60, 90, 120, and 150, peripheral blood mononuclear cells (PBMCs) from individual macaques were tested for antigen-specific cellular immune responses. Intracellular cytokine staining by flow cytometry (CFC) assay to detect antigen-specific CD4+ and CD8+ T cell responses was performed essentially using the modifications described in references 28, 42, and 51. Overlapping peptide pools were used to quantify CFC responses in both CD4+ and CD8+ T cell subsets essentially as described for humans (37–39). PBMCs were resuspended in complete RPMI 1640 medium containing either 50 mg/ml PHA-M (Sigma), peptide pool (15-mers overlapping by 11 amino acids; 1 mM [each] final peptide concentration) spanning SIVmac239 Gag, or medium alone. After activation, the cells were processed and stored until analysis on a BD FACSCalibur (4 colors) or BD FACSVantage (6 colors) in the Division of Vaccine Research at the Institute of Human Virology. Data were analyzed by FlowJo software (Treestar, Inc., San Carlos, CA). Antigen-specific responses for IFN-γ, IL-4, and IL-2 for CD4+ and CD8+ T cells were determined.
RESULTS
Selection of the most active toxin A1 subunit.Previous studies indicate that the DNA vaccine adjuvant effects of CT and LT (A + B subunits) are proportional to their enzymatic activities (4, 21). In these studies, LT was found to be more active than CT (4, 21). For our studies, we codon optimized the bacterial A1 sequences. Codon optimization of DNA sequences is a proven way to increase the expression of genes, particularly when prokaryotic genes are expressed by eukaryotic cells. We used a CREB-Luc reporter plasmid assay to assess the abilities of these optimized CTA1 and LTA1 subunits to elevate cAMP levels in transfected 293 cells. Figure 1 shows that codon optimization of CTA1 and LTA1 dramatically increased their activities over that of the native CTA1 sequence. In this regard, the pMAX-PRO-CTA1 plasmid (expressing the optimized CTA1) required 243-fold-less DNA to evoke a similar signal as that evoked by the pRC/CMV-CTA1 plasmid (expressing the native CTA1). Figure 1 also shows that the optimized CTA1 was more active (as measured in this assay) than the optimized LTA1. As the pMAX-PRO-CTA1 plasmid was most active in this assay, it was chosen for the in vivo adjuvant studies below.
CREB-Luc reporter assay. Fifty thousand 293 cells/well were plated in 96-well opaque-wall plates. The next day, individual wells of cells were transfected using Fugene 6 transfection reagent (270 nl/well) mixed with 8 ng of the pCREB-Luc plasmid; the indicated amount of empty pMAX-PRO, pRC/CMV-CTA1, pMAX-PRO-CTA1, or pMAX-PRO-LTA1; and enough empty pMAX-PRO plasmid to keep the total transfection/well amount at 28 ng of DNA. The assay for each condition was run in triplicate on the same plate. Twenty-four hours later, all cells were lysed and assessed for luciferase production by light emission. The results shown are the means and standard errors of the means from the three replicates.
Dose response of the CTA1 adjuvant delivered by GeneGun.To date, a proper dose-effect study of CTA1 delivered by the GeneGun has not been conducted. For this reason, we undertook a study to determine the optimal dose range of CTA1 when delivered to mice by the GeneGun. Groups of 10 BALB/c mice were immunized on days 0, 14, and 28 according to Table 1. Five mice from each group were sacrificed on day 35, and their splenocytes were stimulated with SIV Gag or HIV gp120 peptide pools. Cytometric bead array (CBA) was performed on the supernatants from the peptide-stimulated splenocytes to quantify secreted Th1 (IFN-γ, IL-2, and TNF-α) and Th2 (IL-4 and IL-5) cytokines. Sera were collected from the other five mice/group on days −2, 13, 28, 40, 60, 90, 120, 150, and 180 and were subjected to anti-SIV Gag and anti-HIV gp120 ELISA for the determination of antibody half-maximal titers. Figure 2A shows that the anti-p27 titers never rose above background levels in the absence of the CTA1 adjuvant. In contrast, Fig. 2A shows that the inclusion of the CTA1 adjuvant plasmid with the SIV Gag plasmid resulted in an approximately 100-fold average increase in anti-p27 titers. Decreasing the dose of CTA1 diminished the magnitude of the anti-p27 antibody responses in responding mice but also decreased the number of responders within the group defined as those exhibiting half-maximal titers of >100 (Table 2).
Design of the dose-response study
Humoral responses for the dose escalation study. Mice were immunized according to the description in Table 1. At the indicated times, serum was collected from individual mice and subjected to anti-SIV Gag (A) or anti-HIV gp120 (B) ELISA to determine half-maximal titers as detailed in Materials and Methods. The error bars represent the standard errors of the means as determined using Excel software.
Number of responders
Unlike SIV Gag, HIV gp120 was very efficient at eliciting antibodies when delivered by GeneGun regardless of the presence or absence of CTA1 (Fig. 2B). The inclusion of CTA1 had only a minor positive effect on the anti-gp120 antibody responses, as CTA1 increased half-maximal titers only approximately an average of 2.5-fold at the two highest doses (Fig. 2B). It should be noted that some differences in the glycosylation patterns on gp120 may exist for the vaccine-expressed gp120 (from mouse cells) and the gp120 used for the ELISAs (bacterially produced). These differences may have affected the overall magnitude of titers; however, the differences existed for all groups and therefore should not have influenced the differences between groups.
It is possible that the inclusion of CTA1 could also have a positive effect on the durability of the anti-gp120 antibody response. This was found in our previous study when the vaccine and adjuvant were delivered intramuscularly (7). Unfortunately, the above experiment was not continued long enough for the responses to wane. Together, these results indicate that higher doses of CTA1 yield the best adjuvant effects when delivered by GeneGun. The results also show that CTA1 has a dramatic adjuvant effect on antibody responses to the intracellular antigen SIV Gag but little effect on the antibody responses to the secreted antigen HIV gp120.
Surprisingly, cellular anti-SIV Gag and anti-HIV gp120 immune responses were not detected in any group by CBA. In this regard, gp120- or Gag peptide-stimulated splenocytes failed to secrete measurable amounts of effecter cytokines, including the Th1 cytokine IFN-γ, TNF-α, or IL-2 or the Th2 cytokine IL-4 or IL-5 (data not shown). Importantly, the positive control PHA-M consistently evoked cytokine secretion from the splenocytes, indicating that they were viable. These results are in contrast to previous studies (4, 21) where two similar GeneGun immunizations induced measurable Th1 and Th2 cellular responses in the presence or absence of adjuvants.
Optimizing the number of inoculations.This study was undertaken to determine the number of booster inoculations necessary to induce optimal antibody production to SIV Gag and HIV gp120 when delivered by the GeneGun. Six groups of 6 BALB/c mice were immunized one, two, or three times, on days 0, 14, and/or 28 as shown in Table 3. Sera were collected from all mice on days −2, 13, 28, 40, 60, 90, and 120 and then subjected to anti-SIV Gag and anti-HIV gp120 ELISA. Cellular responses were not monitored here because we did not observe such responses in the previous study.
Design of the optimal inoculation number study
Figures 3A and B show that CTA1 again dramatically enhanced the antibody responses to SIV Gag while only minimally enhancing the responses to gp120 and that at least two booster inoculations in addition to the priming inoculation are required to yield the optimal antibody response. This is true whether CTA1 was included as an adjuvant or not (Fig. 3A and B). The lower responses seen in the groups receiving either one or two inoculations were reflective of both a lower magnitude of responses from responder mice and a lower number of responding mice within the groups (data not shown). Together, the results of this study verify the findings of the above study and show that at least three inoculations with CTA1 are necessary to induce high-titer responses to SIV Gag and gp120 when delivered by GeneGun.
Humoral responses for the optimal inoculation number study. Mice were immunized according to the description in Table 3. At the indicated times, serum was collected from individual mice and subjected to anti-SIV Gag (A) or anti-HIV gp120 (B) ELISA to determine half-maximal titers as detailed in Materials and Methods. The error bars represent the standard errors of the means as determined using Excel software.
Comparison of CTA1 to GM-CSF.The growth factor cytokine GM-CSF is generally considered to be the most potent DNA vaccine adjuvant in the mouse model and was therefore selected as a “benchmark” for comparison to CTA1. Five groups of 5 BALB/c mice were immunized on days 0, 14, and 32 according to Table 4. Sera were collected from the mice on day 47 and subjected to anti-SIV Gag ELISA. All mice were sacrificed on day 47, and splenocytes were then stimulated with an SIV Gag peptide pool. CBA was performed on the supernatants of the peptide-stimulated splenocytes to quantify secreted Th1 and Th2 cytokines. As a control, IFN-γ ELISPOT assays were also performed on the harvested splenocytes.
Design of the study of CTA1 versus GM-CSF
Figure 4A shows that CTA1 boosted the anti-SIV Gag antibody responses an average of 32-fold while GM-CSF boosted the anti-SIV Gag antibody responses an average of 22-fold at the high 2-μg dose. This difference, however, did not reach statistical significance (P = 0.38 by Mann-Whitney test). This result indicates that CTA1 may be marginally more potent than GM-CSF at this dose. In contrast, Fig. 4A shows that CTA1 retained potent adjuvant effects at the lower 0.2-μg dose while GM-CSF lost most of its adjuvant effects at this dose. In fact, the 0.2-μg dose of CTA1 was nearly as potent (77%) as the 2-μg dose of GM-CSF (Fig. 4A). The difference between CTA1 and GM-CSF was significant at this dose (P = 0.016 by Mann-Whitney test). We also found that IgG1 dominates the antibody responses when SIV Gag is delivered by GeneGun (Fig. 4B). This was true whether or not CTA1 or GM-CSF was included as an adjuvant. The IgG1 antibody responses were measurable in the absence of adjuvant and were dramatically enhanced by either CTA1 or GM-CSF. The differences in potency between CTA1 and GM-CSF were more obvious when focused on the IgG1 isotype, where GM-CSF boosted the responses an average of 53-fold and CTA1 boosted them an average of 127-fold (Fig. 4B). IgG2A responses were near background levels in all groups.
Comparison of CTA1 to GM-CSF. Mice were immunized according to the description in Table 4. On day 47, serum was collected from individual mice and subjected to anti-SIV Gag ELISA to determine half-maximal total IgG titers (A) or the IgG1 or IgG2a isotype titers (B) as detailed in Materials and Methods. The error bars represent the standard errors of the means as determined using Excel software.
As was found above, cellular responses were not detected in any group by CBA (data not shown). In agreement, we did not detect any antigen-specific IFN-γ production by peptide-stimulated splenocytes in ELISPOT assays from any group (data not shown). Together, the results of this study indicate that GeneGun delivery of DNA HIV gp120 or SIV Gag in the presence or absence of adjuvants strongly polarizes antibody responses toward Th2. A Th2 polarization is consistent with some earlier reports (12, 24, 44, 52); however, more-balanced Th1/Th2 responses have been recorded using other antigens and in other animal models (4, 21). The results of this study also indicate that CTA1 is more potent and has a broader dose range than the benchmark adjuvant GM-CSF when delivered by GeneGun.
The adjuvant effects of CTA1 in macaques.DNA vaccines and DNA-encoded adjuvants have performed well in small animal studies but typically do not perform as well in humans and nonhuman primates (11–13). Here, we immunized macaques with SIV Gag with or without CTA1 by GeneGun to determine if the adjuvant effects of CTA1 would be retained in nonhuman primates. Two groups of four macaques were immunized by GeneGun on days 0, 14, and 76 as described in Materials and Methods. The first group was immunized with SIV Gag in the absence of CTA1, and the second group was immunized with SIV Gag in the presence of CTA1. Blood was drawn from the macaques once every 30 days and analyzed by ELISA for SIV Gag half-maximal titers. PBMCs were also stimulated with the Gag peptide pool and processed for intracellular cytokine staining.
Figure 5 shows that CTA1 boosted the anti-SIV Gag antibody half-maximal titer to 289. This reflects an increase of approximately 5-fold over the mean half-maximal titer of 56 for the macaques immunized with SIV Gag in the absence of CTA1. As was found in the mouse, no significant antigen-specific cellular immune responses to Gag were detected at any time point tested. After a 331-day resting period, the macaques were boosted with recombinant SIV Gag protein i.m. on day 407 as described in Materials and Methods. A third group of 4 naïve macaques was also immunized with recombinant SIV Gag protein i.m. on day 407 to determine if DNA priming with or without CTA1 would enhance memory responses and would therefore enhance responses induced by a protein boost.
CTA1 macaque study. Three groups of four cynomolgus macaques were immunized three times by GeneGun. On day 407 each macaque was boosted with an i.m. immunization with recombinant SIV Gag mixed with CpG DNA and alum adjuvant as described in Materials and Methods. At the indicated times, sera were collected from individual macaques and subjected to anti-SIV Gag ELISA to determine the half-maximal titers as detailed in Materials and Methods. The error bars represent the standard errors of the means as determined using Excel software.
Figure 5 shows that the Gag DNA prime in the absence of CTA1 increased the average half-maximal titers after the Gag protein boost to 613 on day 421. This reflects a 1.6-fold increase over the average titer of 394 on day 435 from macaques that received the protein immunization in the absence of a DNA prime. In contrast, Fig. 5 shows that the inclusion of CTA1 with the DNA prime increased the average half-maximal titer after the Gag protein boost to 8,520 on day 421. This reflects a 21.6-fold increase over the average titer of 394 on day 435 from macaques that received the protein immunization in the absence of a DNA prime. In agreement with the mouse studies above, Gag-specific cellular responses were not detectable at any time point tested. Cellular responses were not, however, assayed after the protein boost. Together, these data indicate that CTA1 enhances antigen-specific peak and memory antibody responses in macaques when delivered by GeneGun and also enhances DNA priming for a heterologous protein boost.
DISCUSSION
The reduced immunogenicity of DNA vaccines in primates necessitates the development of improved delivery methods and/or adjuvants. Many different DNA-based adjuvant strategies have been tested, and most have been based on cytokines and chemokines (9, 25, 29, 30, 48). Unfortunately, most of the adjuvants tested so far have also been disappointing in human and nonhuman primate studies (11–13). We previously showed that the enzymatically active A1 subdomain of CT is a potent DNA vaccine adjuvant for HIV gp120 when delivered intramuscularly (14).
Others have also shown that CT and LT (containing the A1, A2, and B subunits) are potent adjuvants for a variety of DNA-encoded antigens when delivered by GeneGun (4, 21, 27, 47). In contrast to our results, this group found that the A subunits of CT or LT alone did not posses potent adjuvant effects when measured by immunological assays (4); however, they subsequently found that the A subunit of LT afforded similar protection for a herpes simplex virus (HSV) challenge as did the combination of the A and B subunits when mice were vaccinated with HSV ICP27 (21). It should be noted that there are important differences between the A subunits (containing both the A1 and A2 subdomains) that were tested in their studies (4, 21) and our CTA1 subunit adjuvant. First, our constructs are intended to be retained in transfected cells and therefore lack the A2 subdomain that may have negatively influenced the outcomes of their A-subunit-only studies. In this regard, the A2 subdomains facilitate the binding to the B pentamers (43). The A2 subdomains of both CT and LT also obscure their hydrophobic ARF binding faces (26). For this reason, the A2 subdomains must be proteolytically cleaved from their A1 subunits in order for the A1 subdomains to bind to ARF proteins and become enzymatically active (45). This step normally takes place in the Golgi complex or ER during toxin entry (45).
The expression of the proteins from a plasmid in the absence of a secretion signal (as was the case in their studies) results in direct cytoplasmic targeting (4, 21). Therefore, it is possible that the A2 subdomains of their constructs did not dissociate from the A1 subunits (4, 21). Such a failure to dissociate could have resulted in a failure of those A subunits to become enzymatically active (4, 21).
A second difference between their constructs and ours is that our constructs were codon optimized for increased expression in eukaryotic cells whereas theirs retained the native bacterial sequence. Codon optimization resulted in an approximately 243-fold increase in activity for CTA1 (Fig. 1). Therefore, it can be reasoned that in their study (4), the true doses of CTA and LTA given were much lower than were given in the present study. Such low doses could have negatively influenced their results. For instance, as shown in Fig. 2 and Table 2, although low doses of CTA1 still markedly enhanced antibody responses, the number of responding mice decreased.
The GeneGun offers several advantages over intramuscular administration of DNA vaccines. First, the GeneGun targets the skin, which is considered to be a preferred immune inductive site that is rich in antigen-presenting dendritic cells (49). Second, the amount of DNA necessary to evoke productive immune responses is much lower with the GeneGun (49). Third, the GeneGun applies the vaccine through a needle-free device (49). For these reasons, we sought to better characterize the adjuvant effects of our CTA1 adjuvant for the important retroviral antigens HIV gp120 and SIV Gag when delivered by GeneGun.
We found that the delivery of these DNA vaccine antigens through ballistic particle bombardment of the skin by GeneGun favors the production of humoral immune responses over cellular immune responses and that the humoral responses are strongly polarized toward Th2. These results are consistent with some (but not all) previous reports (12, 24, 44, 52). Surprisingly, we did not detect cellular immune responses to gp120 or Gag in the GeneGun studies above using ELISPOT assays or CBA. In contrast, we routinely measure robust cellular immune responses to these same antigens using CBA and ELISPOT assays after a single intramuscular immunization (unpublished data). Importantly, the positive control PHA-M consistently evoked cytokine secretion in the above GeneGun studies, indicating that the tested splenocytes were viable.
It is most puzzling that Th2 cytokines were also not detected even though the antibody responses were highly biased toward Th2. It is unclear why these responses were not detected, although it is possible that the responses were present but were below the limit of detection. In contrast, humoral antibody responses were easily detected for both gp120 and Gag when delivered by GeneGun. These antibody responses were dominated by the IgG1 isotype whether an adjuvant was included in the vaccinations or not. The results dramatically demonstrate that CTA1 is a potent adjuvant for SIV Gag delivered by GeneGun to mice. CTA1 also enhanced antibody responses to SIV Gag more than those induced by the “gold standard” mouse DNA adjuvant GM-CSF at a high dose, where the adjuvant dose is equal to the antigen dose. CTA1 also proved to have a broader dose range than GM-CSF, as CTA1 remained potent at a lower dose, where the adjuvant dose was only 1/10 of the antigen dose. In contrast, GM-CSF lost nearly all of its activity at this reduced dose. CTA1 also retained activity for boosting Gag antibody responses even at a very low dose where the adjuvant dose was 1/1,000 of the antigen dose, although the number of animals responding at this dose sharply declined.
In contrast, the antibody-boosting effect of CTA1 for gp120 was much less pronounced than that for Gag, showing only modest enhancing effects even at the highest dose. There are at least two possibilities that could explain the discordant results between Gag and gp120. First, gp120 may be a highly immunogenic protein that elicits near-maximal responses in the absence of an adjuvant. If this were the case, an adjuvant may have little influence on further enhancing these responses. The second possibility is that CTA1 may exert most of its effects on GeneGun-delivered antigens that are retained within the same cell as the adjuvant. For instance, CTA1 may enhance the liberation of intracellular antigens through mechanisms such as cell death or an increase in secretion through membrane-bound vesicles such as exosomes. We are currently investigating this possibility.
As has been seen with other adjuvants, the antibody-enhancing effects of CTA1 were not as dramatic in macaques as they were in mice when delivered by GeneGun. CTA1 boosted the anti-SIV Gag antibody responses approximately 5-fold, and the responses reached an average half-maximal titer of only 289 after three immunizations of macaques. In contrast, there was an over 30-fold increase in titers that reached over 1,000 in mice. CTA1 was, however, able to provide dramatic priming for a recombinant Gag protein boost in macaques (>21-fold increase). This indicates that the inclusion of CTA1 induced a significantly higher memory response than did GeneGun vaccination in its absence. In all, the data from this study indicate that CTA1 is a potent adjuvant for enhancing humoral, Th2-polarized immune responses in mice when delivered by GeneGun. The results of this study also indicate that CTA1-adjuvanted DNA vaccines may provide an optimal prime for heterologous boosts in larger animals.
ACKNOWLEDGMENT
This work was supported in part by the Maryland Industrial Partnerships Project grant number 3905 to K.C.B.
FOOTNOTES
- Received 9 March 2011.
- Returned for modification 29 March 2011.
- Accepted 6 April 2011.
- Accepted manuscript posted online 20 April 2011.
- Copyright © 2011, American Society for Microbiology