The plasmid DNA (pDNA) vaccines used in this study, in the VR1020 backbone 58, have been described previously 18. The PfCSP (VCL-2571), PfSSP2/TRAP (VCL-2576), and PfAMA1 (VCL-2577) vaccines encoded full-length proteins; the PfLSA1 vaccine (VCL-2559) encoded the C-terminal 281 amino acid residues (representing 65% of the nonrepeat region of full length PfLSA1); and the PfMSP1 vaccine (VCL-2574) encoded the 42-kDa fragment of the PfMSP1 protein. All vaccines were based on the 3D7 strain of P. falciparum, and the coding sequence for each antigen was modified for mammalian codon usage in order to improve expression of the encoded antigen 59. Expression of the encoded protein was confirmed in vitro by transient transfection of VM92 melanoma cells (kindly provided by Vical Inc.) using Lipofectamine, as described by the manufacturer (Life Technologies, Gaithersburg, MD). Protein expression was quantitated as a chemiluminescent signal by using antigen-specific monoclonal or polyclonal antibodies and a commercially obtained chemiluminescence-linked Western blot kit (Western-Light, Tropix, Bedford, Mass.) according to the manufacturer's directions. Chemiluminescent signals were detected by exposure of the processed membrane to auto radiographic film (Hyperfilm-ECL; Amersham Life Sciences Inc., Cleveland, Ohio). Plasmid DNA for immunization was produced by Vical Inc., checked for physical integrity, expression activity, concentration, and endotoxin levels, and resuspended in phosphate-buffered saline (PBS) at a concentration of 2.5 mg of pDNA/ml (500 μg of each plasmid DNA vaccine). In the PfCSP alone group, 2 mg of plasmid VR1020 without antigen insert was added to 500 μg PfCSP DNA to maintain the total DNA concentration at 2.5 mg/ml.

The poloxamer-based vaccine formulation consisted of the non-ionic block copolymer, CRL1005 (CytRx Corporation, Los Angeles, CA) and a cationic surfactant, benzalkonium chloride (BAK, BTC 50 NF, Stepan Company, Northfield, IL), formulated with pDNA in PBS. To make the formulation, the required concentration of pDNA (to produce a final concentration of 5 mg/ml) in PBS was stirred on ice and the required amount of CRL1005 (to produce a final concentration of 7.5 mg/ml) was added using a positive displacement pipette. The solution was stirred on ice until the poloxamer dissolved and then the required concentration of benzalkonium chloride dissolved in PBS was added (to produce a final concentration of 0.3 mM). The solution was then cycled through the cloud point (4 to 25°C) several times to ensure homogeneity, diluted with PBS to a final pDNA concentration of 2.5 mg/ml, and then filter sterilized at 4°C and stored frozen (-30°C). Prior to injection, the vaccine formulation was thawed at ambient temperature, stored at room temperature, and administered within 6 hours of preparation.

ALVAC-Pf7 (virus # vCP1305, stock # N37) is a highly attenuated canary poxvirus with seven P. falciparum genes (native codon sequence) inserted into its genome: PfCSP and PfSSP2/TRAP (sporozoite stage); PfLSA1 (liver stage); PfAMA1, PfMSP1, and PfSERA (blood stage), and Pfs25 (sexual stage). The recombinant virus was produced by Virogenetics Corporation (Troy, NY).

Recombinant PfCSP 60, PfLSA1 (D. Lanar, WRAIR), PfAMA1 61, and PfMSP1-42 (A. Saul, MVDB) were well characterized Pf antigens (3D7 strain) manufactured at research grade. The recombinant proteins were used at a final concentration of 10 μg/ml as antigen for in vitro ELIspot and intracellular cytokine staining assays, and at a final concentration of 1 μg/ml as capture antigen for antibody ELISAs. The endotoxin levels were <5 EU/ml for recombinant PfAMA1, and PfMSP1; 13 EU/ml for PfCSP; and 74 EU/ml for PfLSA1, as indicated by the LAL testing (Cambrex BioScience Walkersville, Inc, MD).

A series of 15-mer synthetic peptides, with 10-mer overlaps, derived from PfCSP (total 64 peptides), PfSSP2/TRAP (total 138 peptides) and PfLSA1 (total 114 peptides) were synthesized by Mimotopes (Melbourne, Australia) at > 80% purity. Individual peptides were resuspended in DMSO at a concentration of 20 mg/ml and pooled sequentially according to antigen sequence, with 22 peptides or less per pool. Peptides were used at a final concentration of 4 μg/ml for in vitro T cell assays. The final DMSO concentration did not exceed 0.55%.

A total of 20 rhesus monkeys (Macaca mulatta), aged 5 to 10 years and between 3 and 10 kg in weight were screened for anti-PfCSP antibody negativity by ELISA and randomized into four groups with five monkeys per group. At weeks 0, 4, and 8, monkeys were immunized with 500 μg PfCSP plasmid DNA either in PBS (monkey RZc7, RHc7, RLw6, RSf7, Rli6) or CRL-1005 (monkey RLf7, RUm6, ROr6, RAs6, RDr6), or 500 μg of each CSLAM plasmid DNA either in PBS (monkey RDc7, RCf7, RWa7, RYi6, RVi6) or formulated in CRL1005 (PH1019, ROk6, ROa7, RQk6, RBsG) in a total volume of 1 ml, administered intramuscularly in single sites in the rectus femoris muscle using a 22-gauge needle. Subsequent immunizations were administered to the same muscle on alternating sides of the animal. Twelve weeks after the last pDNA immunization, animals were boosted by intramuscular administration of 2 × 10⁸ pfu of ALVAC-Pf7. All immunizations were carried out at the Emory Vaccine Center at the Yerkes National Primate Research Center, Emory University, GA. All experiments were conducted in compliance with the Animal Welfare Act and with Emory University Institutional Animal Care and Use Committee approvals in accordance with the principles set forth in the "Guide for the Care and Use of Laboratory Animals", Institute of Laboratory Animal Resources, National Research Council, and National Academy Press, 1996. Sera and peripheral blood mononuclear cells were collected pre-immunization and at 4 weeks post each immunization for assessment of immunogenicity. The animals were monitored and evaluated daily for general behaviour, level of activity, and visible side effects or adverse reactions. The injection sites were examined closely by the veterinary clinician each time the animals were sedated for designated bleedings and injections. Close observations included analysis of skin for warmth, erythema and edema/swelling, and muscle induration. The clinical veterinarian examining the animals and injection sites was blinded to which vaccine formulation had been given. Analysis of hematology and clinical chemistry of the different animals was performed on blood samples drawn at pre-bleed, immunization and boosting time points. Hematology analysis consisted of an evaluation of white blood cells, erythrocytes, hemoglobin, hematocrit, mean corpuscular volume and platelets values. Analysis of clinical chemistry consisted of glucose, blood urea nitrogen, creatinine, protein, albumin, alkaline phosphatase, serum glutamic pyruvic transaminase, serum glutamic oxaloacetic transaminase, amylase and creatine phosphokinase determinations.

The assay for rhesus IFN-γ was modified from a previously described method 62. In brief, 96-well PVDF plates (Millipore Corporation, Bedford, MA) were coated with 100 μl/well of anti-human IFN-γ mAb (clone GZ-4, BenderMedSystems, Burlingame, CA) at a concentration of 5 μg/ml in 1× PBS and incubated overnight at 4°C. Plates were washed six times with RPMI 1640 medium and blocked with 200 μl/well of 10% FCS-RPMI 1640 medium (R10) for at least 2 hr at 37°C. After blocking, the plates were washed once with R10, and 100 μl of 2 × 10⁶ PBMC (200,000 cells/well) and 100 μl of stimulant in R10 were added per well, in quadruplicate. Plates were incubated for 18 hrs at 37°C in an atmosphere of 5% CO₂. Plates were then washed six times with 1× PBS in the presence of 0.5% Tween 20 (PBS-T) 100 μl/well of 1 μg/ml biotinylated anti-IFN-γ (clone 7-B6-1, MabTech, Sweden) added per well, and the plates incubated for 1 hr at 37°C. Plates were washed six times with 1× PBS-T, and 100 μl/well streptavidin-alkaline phosphatase conjugate (MabTech, Sweden) was added at 1:1000 dilutions in PBS. After 1 hr incubation at room temperature, plates were washed six times with 1× PBS-T followed by three times with PBS, and developed with AP conjugate substrate kit, (BioRad Laboratories, Hercules, CA) according to the manufacturer's instructions. After 15 min, the plates were rinsed extensively with dH₂O to stop the colorimetric reaction, dried and stored in the dark. The IFN-γ spot-forming cells (SFC) were numerated using a high-resolution automated ELIspot reader (Carl Zeiss Vision, Germany). Responses were expressed as the mean number of SFC per million cells in quadruplicate wells. Responses, to both protein and peptide pools, were classified as positive if (1) the net SFC (mean SFC in experiment antigen wells – mean SFC in medium wells) was > 25 SFC per million PBMC, and (2) the stimulation index (ratio of mean SFC in experimental peptide wells to mean SFC in medium wells) was > 2. In some cases, responses in cells collected before immunization to a specific immunogen met the criteria of positivity as defined above. In that case, responders post immunization were classified as positive if the magnitude of net ELIspot was > 2-fold that of the pre-immunization response.

All reagents for the intracellular cytokine staining were purchased from Becton Dickinson ImmunoCytometry Systems (San Jose, CA). A total of 0.5–1.0 × 10⁶ PBMC in 100 μl R10 medium were plated per well in U-bottomed 96-well plates, in the presence of 1 μg/ml anti-human CD28 (Clone CD28.2) and 1 μg/ml anti-human CD49d (clone 9F10) antibodies with or without 100 μl antigen (total 200 μl/well). GolgiPlug was added at 1 μl/well at 2 hrs after the start of incubation, and the plates were then incubated an additional 14 hrs at 37°C in an atmosphere of 5%CO₂. Plates were spun at 1,200 rpm for 5 min, the supernatant flicked, the cell pellet resuspended by gentle vortexing, and the cells washed twice in FACS buffer. The cells were stained with 20 μl anti-CD4-PerCP-Cy5.5 (clone L200), and 5 μl anti-CD8-PE-Cy7 (clone RPA-T8) in a final volume of 100 μl FACS buffer for 45 min on ice in the dark. After the surface staining, cells were washed with FACS buffer twice, gently resuspended, and permeabilized in 100 μl CytoFix/CytoPerm buffer for 20 min. Cells were washed again and then stained with anti-IFN-γ-FITC (clone B27) and anti-IL2-APC (clone MQ1-17H12) for 45 min on ice in the dark. Cells were washed twice with CytoPerm wash buffer, resuspended in 200 μl of FACS buffer, and stored at 4°C prior to analysis. Samples were analysed using the FACSCalibur™ flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, CA) and CellQuest software. The expression level of intracellular cytokines was presented as the percentage of stained cells in gated cell populations of either CD4+ or CD8+ cells, corrected for background responses in the absence of antigen. The non-specific background was typically less than 0.05%.

Pre- and post-immunization serum samples were serially diluted two-fold and assayed in parallel for anti-Plasmodium antibodies. Parasite-specific antibodies were assayed by the indirect fluorescent-antibody test (IFAT) 63 against air-dried P. falciparum (strain NF54 ~ 3D7 clone) sporozoites or parasitized erythrocytes. IFAT results were reported as the endpoint dilution, representing the last serum dilution at which fluorescence was scored as positive. Antigen-specific antibodies were assayed by the enzyme-linked immunosorbent assay (ELISA) 6364 against recombinant PfCSP (0.5 μg/ml), PfSSP2/TRAP (1.0 μg/ml), PfLSA1 (1 μg/ml), PfAMA1(1 μg/ml), or PfMSP1 (1 μg/ml) proteins. ELISA results were reported as OD 0.5 units, representing the reciprocal of the serum dilution at which the mean OD reading was 0.5. ELISA responses were classified as positive if (1) titer of OD 0.5 units was > 25 and (2) the seroconversion index (ratio of titer post-immunization to titer pre-immunization) was > 4.

Experimental outcomes were presented as direct results of ELIspot assays (IFN-γ spot forming cells/million splenocytes), FACS analysis (% responding cells), ELISA (antibody titers) or IFAT (antibody titers). Comparison of ELISA data and intracellular cytokine expression were done by 2-tailed student t test. ELIspot data are reported as mean +/- standard deviation of quadruplicate wells. Responses to all individual antigens, either recombinant proteins (PfCSP, PfLSA1, PfAMA1, and PfMSP1) or peptide pools (i.e. three PfCSP pools, seven PfSSP2/TRAPpools, six PfLSA1 pools), were analysed with an analysis of variance for a repeated measures design with one between-subjects factor (vaccine group at four levels), and one within-subjects factor (time at six levels). All pair-wise comparisons of means were made with Fisher's LSD (Least Significant Difference) test. The sums of the peptide pools were analysed with the same methods. The recombinant proteins were analysed with the same methods except for the addition of one between-subjects factor (called r-protein with 4 and 5 levels, respectively, for analyses with and without PfSSP2/TRAP). Fisher's exact test was used informally to conduct pair-wise vaccine group comparisons of positive responder rates. Trends over time were examined with graphs, but no formal statistical tests of significance were conducted to test for significant trend components (i.e. linear, quadratic or cubic components). The p-value considered significant was p < 0.05.

Animals were evaluated daily for general behaviour and the presence of any adverse reactions or clinical abnormalities, as described in Material and Methods. No unusual behaviour or adverse reactions were observed throughout the course of the trial. Haematology and clinical chemistry values remained within the normal range.

The induction of parasite-specific and antigen-specific antibody responses post vaccination was assessed by IFAT against air-dried Pf sporozoites or parasitized erythrocytes, and by ELISA against each vaccine antigen component, respectively. IFAT responses were evaluated pre-immunization, after the 3^rd DNA immunization, and after virus boost. Positive IFAT responses against Pf sporozoites were detected post DNA immunization in 5/5 animals in the CSLAM/PBS immunized group but 0/5 in the PfCSP/PBS immunized group (data not presented). The titer of IFAT increased approximately five-fold after virus boost (mean, 160 versus 30). No blood-stage parasite specific antibodies were detected by IFAT at any time point.

To determine whether co-immunization of PfCSP with four additional antigens, PfSSP2/TRAP, PfLSA1, PfAMA1 and PfMSP1 in PBS (i.e. the other antigens of the CSLAM/PBS cocktail), adversely affected the PfCSP-antigen specific antibody responses, the frequency of positive responders and the magnitude of ELISA responses were compared between PfCSP/PBS immunized and CSLAM/PBS immunized animals. No antigen-specific antibodies were detected after the first DNA immunization. In monkeys immunized with PfCSP/PBS alone, anti-PfCSP antibodies could not be detected until after third DNA immunization, when only1/5 monkeys met the criteria of positivity and remained positive after virus boost (Figure 1). In monkeys immunized with CSLAM/PBS, however, anti-PfCSP antibodies were detected in 1/5 monkeys after the 2^nd immunization, 3/5 monkeys after 3^rd DNA immunization, and 5/5 monkeys after virus boost. The anti-PfCSP antibody response was significantly higher in the CSLAM/PBS immunized responder monkeys as compared with PfCSP/PBS immunized responders after the viral boost (p = 0.044), but not after DNA immunization (p = 0.20). In the CSLAM/PBS group, anti-PfAMA1 and anti-PfMSP1 antibodies were detected in 0/5 monkeys after the 1^st DNA immunization, 3/5 monkeys after 2^nd DNA immunization, 3/5(PfAMA1) or 4/5 (PfMSP1) after the 3^rd DNA immunization, and 4/5 (PfAMA1) or 5/5 (PfMSP1) monkeys after the viral boost. Anti-PfSSP2/TRAP antibodies were detected in 2/5 monkeys post-viral boost only. No anti-PfLSA1 specific antibody responses were detected at any time point. After virus boosting, antibody titers increased approximately 6-fold for PfCSP, 9-fold for PfAMA1, and 2-fold for PfMSP1 (Figure 1). Overall, the most immunogenic antigen with regard to antibody response was PfAMA1.

Overall, antigen-specific IFN-γ ELIspot responses were observed in 18/20 immunized monkeys regardless of groups and immunogens, and responses were detected as early as 4 weeks after a single DNA immunization. The two non-responder monkeys were monkey RCf7 (CSLAM/CRL10005) and monkey RHc7 (CSP/PBS). The average SFC from positive responders post 1^st DNA immunization (mean 102 SFC/million, range 0–273 SFC/million) was approximately 5 times higher than the average SFC response before immunization (20 SFC/million, p < 0.05) (Figure 2). The ELIspot responses plateaued post 2^nd (mean 95, range 0–286 SFC/million) and 3^rd (mean 83, range 0–401 SFC/million) DNA immunizations but were boosted by ALVAC-Pf7 with magnitudes increasing to an average of 152 SFC/million (range 0–466 SFC/million, p = 0.005 compared to preimmunization) at 4 weeks post virus boost. The response then declined at 12 weeks post viral boost to a level similar to that prior to the boost (mean 79, range 0–268 SFC/million). A similar pattern of PfCSP-specific IFN-γ ELIspot responses was observed to synthetic PfCSP peptide pools, although the overall magnitude of SFC to individual peptide pools was not as great as to the recombinant protein (data not presented). These data confirm previous studies by us 37, demonstrating the immunogenicity of PfCSP plasmid DNA in rhesus monkeys via detection of CTL activity at 3 weeks after the 2^nd DNA immunization. In that study, the PfCSP DNA was native rather than mammalian codon optimized, and CTL responses could not be detected after a single DNA immunization. In other studies in the P. knowlesi/rhesus model, with native P. knowlesi antigens, no response could be detected even after three DNA immunizations 3839.

To determine whether co-immunization of PfCSP with PfSSP2/TRAP, PfLSA1, PfAMA1, and PfMSP1 adversely affected the PfCSP-antigen specific T cell responses, the frequency of positive responders and the magnitude of ELIspot responses were compared between PfCSP/PBS immunized and CSLAM/PBS immunized animals. There was no significant difference between the frequency of PfCSP-specific IFN-γ positive responders in monkeys immunized with PfCSP/PBS alone or with PfCSP in the pentavalent cocktail CSLAM/PBS at anytime point, as determined by responses against both recombinant protein and synthetic peptide immunogens (Table 1). Interestingly, and consistent with the antibody data presented above, there were more PfCSP responders in the group immunized with CSLAM/PBS as compared to the group immunized with PfCSP/PBS alone.

Similarly, there was no significant difference in the magnitude of PfCSP-specific IFN-γ ELIspot responses between the PfCSP/PBS and CSLAM/PBS groups (Figure 3). At four weeks post 3^rd DNA immunization, the SFC/million in responders to recombinant PfCSP protein ranged between 145 and 401 SFC/million (mean 269 SFC/million PBMC) for the CSLAM/PBS group as compared with a range of 40–102 SFC/million (average 71 SFC/million) for the PfCSP/PBS group. At 4 wks post viral boost, the magnitude of IFN-γ ELIspot forming cells was 246 SFC/million (range 51–466 SFC/million PBMC) for the CSLAM/PBS group versus 284 SFC/million (range 162–406 SFC/million PBMC) for the PfCSP/PBS group.

These ELIspot data, together with the antibody data above, demonstrate that co-immunization of PfCSP with two other pre-erythrocytic stage antigens (PfSSP2/TRAP, PfLSA1) and two erythrocytic stage antigens (PfAMA1 and PfMSP1) does not adversely affect the immunogenicity of PfCSP. These data support the down-selection of the CSLAM antigen combination, as determined by studies in the P. yoelii model 18.

Antigen-specific responses to the other components of the CSLAM vaccine were also assessed. As shown in Figure 4A, antigen-specific IFN-γ ELIspot response to all other components of CSLAM/PBS vaccine (PfLSA1 recombinant protein and synthetic peptides, PfSSP2/TRAP synthetic peptide, PfAMA1 recombinant protein, and PfMSP1 recombinant protein) were induced by DNA immunization and were boosted by ALVAC-Pf7 (peptide data other than PfSSP2/TRAP not presented). IFN-γ responses to PfSSP2/TRAP and PfMSP1 were detected in 5/5 CSLAM/PBS immunized monkeys, and responses to PfCSP, PfLSA1 and PfAMA1 were detected in 4/5 monkeys. Positive responses to PfLSA1 and PfAMA1 were observed after the 1^st DNA immunization, as noted for PfCSP, whereas positive responders to PfSSP2/TRAP and PfMSP1 were detected only after the 2^nd DNA immunization. After the ALVAC-Pf7 boost, at least 3/5 monkeys met the defined criteria of positive responses for each immunogen (Figure 4A).

The magnitude of vaccine-induced IFN-γ ELIspot responses in the 5 CSLAM/PBS immunized monkeys to each of the CSLAM components is shown in Figure 4B. The most robust response, at 4 weeks post virus boost, was against PfAMA1 (in 4/5 monkeys); this response was not significantly different from the anti-PfLSA1 response (p > 0.05) but was significantly higher than the PfCSP- and PfMSP1-specific responses (p < 0.05). Of the five antigens tested, PfMSP1 was the least immunogenic for T cell responses at all time points. For all antigens and all responder monkeys (with the exception of one PfCSP immunized monkey), IFN-γ ELIspot responses were boosted byALVAC-Pf7 virus. All boosted responses decreased within 12 weeks post viral boost to a level similar to that prior to the boost.

Consistent with the known genetic restriction of T cell responses to Plasmodium proteins 65, heterogeneity of IFN-γ responses amongst individual monkeys was noted. The same monkey did not respond to all antigens, and different monkeys responded to different antigens (Figures 4A and 4B). For example, at four weeks post virus boost, two monkeys responded to five antigens, one to four antigens, one to two antigens, and one to only one antigen. The most robust IFN-γ responses to all five antigens were detected in monkey RBsG in the CSLAM/PBS group; strong responses to recombinant PfCSP, PfAMA1, PfSSP2/TRAP, and PfLSA1 but not PfMSP1, were detected in monkey RQk6; monkeys PH1019 and ROk6 exhibited strong responses to PfAMA1 but weak response to the other four antigens; and monkey ROa7 had a poor response to PfMSP1 and no response to the other four antigens.

For those antigens for which synthetic peptides spanning the complete antigen were available (PfCSP, PfSSP2/TRAP, and PfLSA1), responses to multiple peptide epitopes were detected. Representative data for responses in monkey RBsG, the most reactive monkey in the CSLAM/PBS group, are presented in Figure 5. Among the three antigens (and consistent with the recombinant protein reactivity reported above), the most robust IFN-γ responses were detected to PfLSA1, with positive responses to five of the six peptide pools. The most N-terminal pool LSA1–1415 (residues 1–84) was the most immunogenic, with an average of 380 SFC/million PBMC at 4 wks post ALVAC-Pf7 boost. The magnitude of IFN-γ responses in three of the other four PfLSA1 peptide pools was comparable, and the least reactive pool was LSA1–1718 (residues 132–236). Similarly, for PfCSP, the most immunogenic peptide pool was the most N-terminal pool CSP-2526 (residues 1–184), although the magnitude of responses to PfCSP peptide pools was lower than that for the other two antigens. The frequency and magnitude of responses to the PfSSP2/TRAP peptide pools were variable; the most immunoreactive pool was residues 44–128, with an average of 277 SFC/million PBMC at 4 wks post viral boost.

These data establish that T cell responses to each component of the CSLAM/PBS vaccine were detected in rhesus monkeys following DNA immunization and were boosted by ALVAC-Pf7.

To investigate the capacity of CEL1000 to enhance the immunogenicity of plasmid DNA, the frequency and magnitude of antigen-specific IFN-γ T cell responses, and parasite-specific and antigen-specific antibody responses, in monkeys administered PfCSP or CSLAM formulated in CRL1005, were evaluated in comparison with those in PBS. The frequencies of responders to PfCSP in the PfCSP/PBS and CSP/CRL1005 immunized monkeys, and the frequencies of responders to each CSLAM antigen in the CSLAM/PBS and CSLAM/CRL1005 group, are presented in Figure 6. The respective magnitudes of responses are presented in Figure 7. Overall, for PfCSP as well as the other CSLAM antigens, there was no measurable effect of CRL1005 formulation on enhancement of either frequency (Figures 6A and 6B) or magnitude (Figure 7) of IFN-γ responses.

Likewise, there was no apparent effect of CRL1005 formulation on the frequency or magnitude of antibody responses to PfCSP in PfCSP immunized animals, nor to any of the CSLAM antigen components in CSLAM immunized monkeys (Additional File 1).

The ELIspot assay as reported here measures the total number of cytokine secreting cells within a bulk cell population and does not allow discrimination between the responding cell phenotype(s). Therefore, to determine the phenotypes of IFN-γ-producing cells, the intracellular expression of IFN-γ and IL-2 for CD4+ and CD8+ T cell subsets, in CSLAM immunized monkeys was measured. Based on cell availability, PBMC from three monkeys (ROa7, RQk6, RBsG), collected at preimmunization and four weeks post virus boost time points, were analysed against peptide pools derived from PfCSP, PfSSP2/TRAP and PfLSA1. Figure 8 shows the result of this analysis from the monkey RBsG, for PfSSP2/TRAP and PfLSA1. Peptide pools PfSSP2/TRAP-1011 (residues 396–495), PfSSP2/TRAP-1213 (residues 484–562), PfLSA1–1415 (residues 1–84) and PfLSA-1516 (residues 44–143) preferentially induced CD4+ IL-2 responses, with low level CD4+IFN-γ responses. All four pools also induced low level CD8+ IFN-γ responses, and two of them induced low level CD8+ IL-2 responses (Figure 8). These data indicate that the CSLAM DNA vaccines preferentially induce CD4+ T cell cytokine responses, rather than CD8+ T cell responses, and that IL-2 responses predominate over IFN-γ responses. The data reported above describing the correlation of summed IFN-γ responses to synthetic peptide overlaps and IFN-γ responses to the recombinant protein also suggested that the IFN-γ responses were mediated by CD4+ T cells, rather than CD8+ T cells. Taken together, these data indicate that CD4 T cells, rather than CD8+ T cells, are primarily responsible for the antigen-specific IFN-γ responses detected by ELIspot, at least under the conditions evaluated herein.