The use of cytokines has shown promise as an approach for amplifying vaccine-elicited immune responses, but the application of these immunomodulatory molecules in this setting has not been systematically explored. In this report we investigate the use of protein- and plasmid-based cytokines to augment immune responses elicited by an HIV-1 gp120 plasmid DNA vaccine (pV1J-gp120) in mice. We demonstrate that immune responses elicited by pV1J-gp120 can be either augmented or suppressed by administration of plasmid cytokines. A dicistronic plasmid expressing both gp120 and IL-2 induced a surprisingly weaker gp120-specific immune response than did the monocistronic pV1J-gp120 plasmid. In contrast, systemic delivery of soluble IL-2/Ig fusion protein following pV1J-gp120 vaccination significantly amplified the gp120-specific immune response as measured by Ab, proliferative, and CTL levels. Administration of plasmid IL-2/Ig had different effects on the DNA vaccine-elicited immune response that depended on the temporal relationship between Ag and cytokine delivery. Injection of plasmid IL-2/Ig either before or coincident with pV1J-gp120 suppressed the gp120-specific immune response, whereas injection of plasmid IL-2/Ig after pV1J-gp120 amplified this immune response. To maximize immune responses elicited by a DNA vaccine, therefore, it appears that the immune system should first be primed with a specific Ag and then amplified with cytokines. The data also show that IL-2/Ig is more effective than native IL-2 as a DNA vaccine adjuvant.
Direct injection of plasmid DNA expressing a gene encoding the protein of a pathogen has proven to be a novel and effective vaccination modality. Intramuscular injection of such “DNA vaccines” leads to uptake of the plasmid by host cells and expression of the protein Ag (1). The protein enters the Ag-processing pathways, resulting in strong and persistent humoral and cellular immune responses (2, 3, 4). Immune responses elicited in this manner have been shown to confer protective immunity against influenza challenge in mice and ferrets (3, 5, 6, 7). Such studies have generated considerable interest in DNA vaccines, since DNA immunization offers the advantage of eliciting Ab and CTL responses without the pathogenic risks inherent in immunization with live vectors (reviewed in Refs. 8–11).
The potential utility of plasmid DNA as a component of an AIDS vaccine is currently an area of active investigation. Although the immune correlates of protection are not fully understood for HIV, the growing consensus is that a strong CTL response as well as a neutralizing Ab response will be necessary to prevent infection or disease (12, 13, 14, 15). A number of reports have shown that DNA vaccines can generate HIV-specific and SIV-specific CTLs, Th cells, and Abs in mice and nonhuman primates (4, 16, 17, 18, 19, 20, 21, 22). In addition, several DNA vaccine trials in nonhuman primates involving viral challenges appear promising. In a macaque SIV DNA vaccine trial, while monkeys were not protected from infection, prior vaccination attenuated the in vivo pathogenicity of the challenge virus (23). A DNA vaccine was also shown to protect chimpanzees against challenge with HIV-1 SF2, although the significance of this finding has been questioned given the ease of protecting chimpanzees from infection with this particular HIV-1 isolate (24, 25). In addition, an HIV-1 env DNA vaccine priming followed by recombinant Env protein boosting protected macaques from a chimeric SIV challenge (26).
The full potential of DNA vaccines has not yet been fully realized. For example, cytokines or other immunoregulatory molecules might be used to enhance or redirect the immune response elicited by a DNA vaccine. Coinoculation of a plasmid expressing GM-CSF3 with a rabies virus DNA vaccine has been shown to increase the rabies-specific Ab response in mice (27). A number of reports have also demonstrated that plasmid IL-12 enhanced the specific CTL response elicited by DNA vaccines encoding HIV or influenza Ags (28, 29, 30, 31). Administration of plasmid IL-2 and GM-CSF has also been reported to augment cellular immune responses generated by a hepatitis C virus DNA vaccine and to reverse the suppressive effects of ethanol (32, 33). Plasmid IL-2 has been reported to augment immune responses induced by a transferrin DNA vaccine, and to increase the proliferative response and transiently augment the Ab response elicited by a hepatitis B virus DNA vaccine (34, 35). Furthermore, immune responses induced by a DNA vaccine encoding carcinoembryonic Ag were augmented by coadministration of plasmid GM-CSF or B7-1 (36).
In this report we investigate the ability of cytokines to boost the Ab, proliferative, and CTL responses elicited by a DNA vaccine encoding HIV-1 gp120 IIIB. For these studies we use both native cytokines and IL-2/Ig, a fusion protein consisting of murine IL-2 and murine Fcγ2a mutated to eliminate its Ab-dependent cell-mediated cytotoxicity and complement binding properties (37, 38, 39). IL-2/Ig has IL-2 functional activity and a longer in vivo half-life. We also investigate the relative efficacy of soluble protein vs DNA-based cytokine administration and examine the effects on the immune response caused by varying the time of administration of cytokine relative to Ag priming.
Materials and Methods
Plasmids were constructed using standard molecular biologic techniques (40). PCRs were conducted using Pfu DNA polymerase (Stratagene, La Jolla, CA), synthetic oligonucleotide primers (Operon Technologies, Alameda, CA), and a Perkin-Elmer temperature cycler (Perkin-Elmer, Norwalk, CT). Reaction conditions included 100 ng of template, 250 ng of each primer, 0.2 mM dNTPs, and 2.5 U Pfu enzyme in a 100-μl volume. Cycling was performed at 95°C for 1 min, at 55°C for 1 min, and at 72°C for 3 min for 25 cycles, followed by a 10-min final extension at 72°C. PCR products were purified by gel electrophoresis and GeneClean (BIO-101, La Jolla, CA). Restriction enzymes, T4 DNA ligase, and bacterial alkaline phosphatase were purchased from Life Technologies (Gaithersburg, MD) and used according to the manufacturer’s protocols. Competent DH5α Escherichia coli were transformed and plated overnight on Luria-Bertoni plates containing 100 μg/ml ampicillin or 50 μg/ml kanamycin (Sigma, St. Louis, MO). Single colonies were picked and grown in 2-ml liquid cultures. Plasmid clones were screened by diagnostic restriction digestion and confirmed by dideoxy sequencing using synthetic oligonucleotide primers (Operon Technologies) at the Beth Israel Deaconess Medical Center Molecular Medicine sequencing facility (Boston, MA).
The IL-2 cDNA was originally obtained from the American Type Culture Collection (ATCC 37553). The IL-2/Ig fusion gene was prepared by PCR amplification of the IL-2 cDNA and cloning into a vector containing a noncytolytic murine Fcγ2a (mutated to eliminate its Fc receptor and C1q binding sites) with a BamHI site spanning the fusion protein hinge region, as described previously (38). The purified IL-2/Ig protein has a molar equivalent biologic function compared with murine IL-2, as determined by functional CTLL stimulation assays.
Inoculated cultures of Luria Bertani broth containing appropriate antibiotics were grown overnight with shaking at 37°C. Minipreparations of plasmids were made using the Wizard DNA Purification Systems (Promega, Madison, WI). Maxipreparations of plasmids were made by standard alkaline lysis followed by double CsCl gradient banding. A 1-L overnight bacterial culture was centrifuged, and the pellet was resuspended in 30 ml of solution I (50 mM glucose, 25 mM Tris-Cl (pH 8), and 10 mM EDTA (pH 8)). The suspension was then lysed using 30 ml of solution II (1% SDS and 0.2 M NaOH), neutralized using 30 ml of solution III (5 M KOAc), and centrifuged at 3000 rpm for 30 min in a Sorvall centrifuge (Sorvall, Braintree, MA). The supernatant was removed and filtered, and 0.6 vol of isopropanol was added. Following a 30-min incubation and centrifugation at 10,000 rpm for 30 min in a Sorvall centrifuge, the supernatants were discarded, and the isopropanol pellets were air-dried and resuspended in 4 ml of TE buffer. Optical grade CsCl (4.7 g; Life Technologies) and 0.3 ml of 10 mg/ml ethidium bromide were added, and the solution was ultracentrifuged at 55,000 rpm overnight at 20°C. The CsCl-banded DNA was removed and then spun on a second CsCl gradient. Following double CsCl banding, the ethidium was extracted five times using water-saturated isobutanol, and the DNA was precipitated with 0.1 vol of NaOAc and 3 vol of ethanol. The DNA was washed with 70% ethanol, resuspended in TE, extracted with phenol/chloroform, extracted with chloroform, reprecipitated with ethanol, washed with 70% ethanol, and then resuspended in sterile 150 mM NaCl. The DNA was then used for diagnostic digestions, in vitro transfections, or injections into mice. The final DNA had an OD 260 nm/280 nm ratio of 1.90 to 1.95.
In vitro transfections
Expression levels of plasmid constructs were tested using transiently transfected COS cells. COS cells were split to a density of 106 cells/100-mm plate, grown for 24 h, and transfected with 10 μg of plasmid with the calcium phosphate method using the CellPhect kit (Pharmacia Biotech, Piscataway, NJ) according to the manufacturer’s protocol. After 2 days, cell supernatants were removed and analyzed for the presence of secreted proteins by ELISA (Endogen, Cambridge, MA).
Mice and immunizations
Eight- to twelve-week-old female BALB/c and C3H mice were purchased from Charles River Laboratories (Wilmington, MA) or Jackson Laboratories (West Grove, PA). Mice were immunized as previously described (19). Briefly, mice were injected i.m. in the quadriceps with 10 to 200 μg of plasmid DNA encoding gp120 or cytokine genes in 100 μl of 150 mM sterile saline with no adjuvant. Half the dose was given in each leg. Soluble IL-2/Ig protein was prepared as previously described (38). Mice receiving IL-2 (BioSource, Camarillo, CA) or IL-2/Ig were given daily i.p. injections of 0.4 to 1 μg protein in 100 μl of PBS. Certain groups of mice were boosted after 2 to 3 mo with 50 μg of pV1J-gp120.
Anti-gp120 ELISA assay
A direct ELISA was used to measure serum titers of murine anti-gp120 Abs. Ninety-six-well Maxisorp ELISA plates (Nunc, Naperville, IL) were coated overnight at 4°C with 100 μl of 1 μg/ml recombinant human gp120 (Intracel, Cambridge, MA) in PBS. The remainder of the ELISA was conducted at room temperature. Following a wash with PBS containing 0.05% Tween-20, the wells were blocked for 2 h with a solution containing 2% BSA (Sigma) and 0.05% Tween-20 in PBS. Sera were prepared from murine blood samples, serially diluted in 2% BSA/0.05% Tween-20, and added to ELISA wells. Following a 1-h incubation, the plate was washed three times and then incubated with a 1/5000 dilution of a peroxidase-conjugated affinity-purified rabbit anti-mouse secondary Ab (Jackson Laboratories) in 2% BSA/0.05% Tween-20 for 1 h. The plate was washed three times, developed with TMB (KPL, Gaithersburg, MD), stopped with 1% HCl, and analyzed at 450 nm with a Dynatech MR5000 ELISA plate reader. Subtyping of Abs was conducted with the Clonotyping System (Southern Biotechnology Associates, Birmingham, AL) using the manufacturer’s protocols.
Preparation and stimulation of murine splenocytes
Spleens from the DNA-vaccinated mice were aseptically removed, and single cell suspensions were prepared using a no. 100 surgical stainless steel mesh. RBCs were removed by treating the spleen cells with NH4Cl-KCl lysis buffer for 5 min at 4°C, followed by two washes in HBSS containing 2% calf serum.
Normal BALB/c splenocytes were incubated with 40 μM gp120 IIIB P18 peptide (RIQRGPGRAFVTIGK, Multiple Peptide Systems, San Diego, CA) for 2 h at 37°C and then irradiated in a GammaCell irradiator. Splenocytes (5 × 107) from DNA-vaccinated mice were stimulated with 5 × 107 peptide-pulsed and irradiated normal syngeneic splenocytes in 12-well tissue culture plates (Falcon, Becton Dickinson, Mountain View, CA) in 2 ml of RPMI 1640 containing 10% FBS (HyClone, Logan, UT), 2 mM l-glutamine, 20 U penicillin/ml, 20 μg streptomycin/ml, and 5 × 10−5 M 2-ME (all from Life Technologies). The splenocytes were incubated at 37°C in 5% CO2 for 6 days. Effector cells were harvested from the culture on the seventh day and used in a 51Cr cytotoxicity assay.
51Cr release cytotoxicity assay
This assay was performed as previously described (17, 19) using the mastocytoma cell line P815 as target cells. P815 cells were pulsed overnight with 40 μM P18 peptide at 37°C in 5% CO2 and labeled with 150 μCi 51Cr (ICN Biomedicals, Irvine, CA) for 90 min at 37°C in 5% CO2. After three washes, the radiolabeled target cells were resuspended in complete RPMI 1640 at a concentration of 1 × 105 cells/ml. The effector cells in a total volume of 100 μl were added in duplicate to the wells of a 96-well, U-bottomed tissue culture plate (Falcon, Lincoln Park, NJ). After a 5-h incubation at 37°C in 5% CO2, 50 μl of supernatants were harvested from each well, mixed with scintillation fluid, and measured using a Wallac 1450 Microbeta liquid scintillation counter (Wallac, Gaithersburg, MD). To measure spontaneous release of 51Cr, target cells were incubated with 100 μl of medium, and for maximum release, target cells were incubated with 100 μl of 10% Triton X-100 in PBS. Spontaneous release in each experiment was approximately 10% of the maximum release. The percent specific cytotoxicity was calculated as: (experimental release − spontaneous release)/(maximum release − spontaneous release).
The [3H]TdR uptake assay was used to measure the proliferation of splenocytes after antigenic stimulation. Splenocytes from DNA-vaccinated animals were resuspended at a concentration of 4 × 106 cells/ml in RPMI 1640 containing 5% FBS and antibiotics as described above. One hundred microliters of the cell suspension was added to each well of a 96-well flat-bottom tissue culture plate. Recombinant HIV-1 gp120 (Intracel) was added at a final concentration of 2.0, 0.4, 0.1, or 0 μg/ml. After 4 days of culture, 1 μCi of [3H]TdR (ICN Biomedicals) was added to each well and incubated overnight at 37°C in 5% CO2. The cells were then harvested on glass filter paper using a Tomtec cell harvester (Tomtec, Orange, CT), and the radioactivity present in the cells was measured in a Wallac 1450 Microbeta liquid scintillation counter.
Cytokine ELISA assays
Splenocytes (4 × 106) from the experimental animals were cultured with 2 μg/ml recombinant gp120 (Intracel) in a total volume of 1 ml of RPMI 1640 containing 5% FBS in a 24-well tissue culture plate for 72 h. The supernatants were harvested and assayed for the presence of cytokines using ELISA kits (Endogen) according to the manufacturer’s protocol.
Immunogenicity of dicistronic DNA vaccines coexpressing gp120 and a cytokine
Studies were initiated to explore the use of plasmid-expressed cytokines as a strategy for amplifying immune responses elicited by plasmid DNA vaccines. pV1J-gp120, a DNA vaccine encoding HXBc2 gp120 IIIB, has previously been shown to elicit potent humoral and cellular immune responses in mice and nonhuman primates (19, 41). This vaccine is derived from pUC19 with a kanamycin resistance gene; a CMV IE1 enhancer, promoter, and intron A; the gene encoding gp120; and a bovine growth hormone polyadenylation sequence (42). To examine the effects of plasmid-expressed cytokines on immune responses to pV1J-gp120, three dicistronic vaccines were constructed from pV1J-gp120 using standard molecular biologic methods (40). These vaccines included the pV1J backbone with both gp120 and a cytokine gene, IL-2, IL-4, or GM-CSF. The gp120 and cytokine genes were separated in these constructs by the encephalomyocarditis virus internal ribosome entry site, which has been shown to promote efficient internal initiation of translation (43).
The pV1J-gp120 control, pV1J (sham), pV1J-gp120/IL-2, pV1J-gp120/IL-4, and pV1J-gp120/GM-CSF vaccines were tested for in vitro protein expression levels. COS cells were transiently transfected with the constructs, and cell supernatants were analyzed after 2 days by ELISA for the presence of gp120 and cytokines. As shown in Table I⇓, the pV1J (sham) negative control plasmid had no detectable expression of gp120, whereas the monocistronic pV1J-gp120 and the dicistronic pV1J-gp120/cytokine plasmids all had comparable high expression levels of gp120. The pV1J-gp120/cytokine constructs also expressed the appropriate cytokines, and the molar ratio of gp120 to cytokine expression for all constructs was 1.5 to 2.0:1.
Groups of BALB/c mice (n = 10/group) were then immunized with 100 or 10 μg of either the monocistronic pV1J-gp120 vaccine or the dicistronic pV1J-gp120/cytokine vaccines. The plasmids, dissolved in sterile saline without adjuvant, were injected in both hind legs in the quadriceps muscle. Three, four, five, and six weeks later, the mice were bled, and sera were tested by ELISA for the presence of anti-gp120 Abs. As shown for the sera from the 4 wk blood samples in Figure 1⇓, a single inoculation of the control pV1J-gp120 vaccine elicited a strong anti-gp120 Ab response. The seroconversion frequency in the mice was >90%. Surprisingly, the mice receiving the dicistronic gp120/IL-2 and gp120/IL-4 vaccines developed Ab responses more than 10-fold lower than those receiving the control gp120 vaccine despite the similar gp120 expression levels of all the constructs in vitro. The mice receiving the dicistronic gp120/GM-CSF vaccine developed Ab responses comparable to those of mice that received the monocistronic gp120 vaccine. In this and all subsequent analyses of Ab responses described in these studies, sera were assessed four times between 3 and 6 wk and showed no significant differences in the kinetics of the immune responses.
Effects of soluble IL-2 protein and soluble IL-2/Ig fusion protein on the anti-gp120 immune responses elicited by pV1J-gp120
IL-2 has previously been characterized as a factor that augments rather than suppresses specific immune responses, and it has been shown to be an effective adjuvant for subunit and inactivated virus vaccines (44, 45, 46). Therefore, additional experiments were conducted to investigate the effects of this cytokine on the immune response elicited by pV1J-gp120. We first examined whether soluble IL-2 protein administered systemically following vaccination would modulate the anti-gp120 Ab response. Groups of BALB/c mice (n = 4/group) were immunized with 50 μg pV1J-gp120 plus daily i.p. injections of either PBS alone or 0.4 μg of IL-2 in PBS. Figure 2⇓A demonstrates that the anti-gp120 Ab response elicited by pV1J-gp120 was not significantly altered by IL-2 administration.
We reasoned that this lack of effect may be explained by the brief circulatory half-life of IL-2. We therefore expressed and purified IL-2/Ig, a fusion protein that has a much longer half-life in vivo and also acts as divalent IL-2 (38, 39). A similar experiment was performed to examine the effects of soluble IL-2/Ig protein on the immune response elicited by pV1J-gp120. Groups of BALB/c mice (n = 8/group) were immunized with either 50 μg of pV1J-gp120 or 50 μg of pV1J (sham) plasmid. Two groups of mice receiving pV1J-gp120 also received daily i.p. injections of either 1 μg of Ig control protein or 1 μg of IL-2/Ig in PBS (1 μg of IL-2/Ig represents a molar equivalent to 0.4 μg of IL-2). Figure 2⇑B demonstrates that the anti-gp120 Ab response elicited by pV1J-gp120 was not altered by injection of the Ig control protein; it was, however, enhanced >10-fold by administration of IL-2/Ig.
The mice were boosted after 3 mo with 50 μg of pV1J-gp120 or 50 μg of pV1J (sham) plasmid without cytokine treatment. Four weeks later the mice were bled, and sera were tested again for anti-gp120 Ab titers. Increased titers were observed, and the IL-2/Ig group maintained a >10-fold higher Ab titer than the control group (data not shown). The mice were sacrificed, and recombinant gp120-specific splenocyte proliferation was assessed by standard thymidine incorporation assays. As shown in Figure 3⇓, the splenocytes of the mice that received IL-2/Ig had higher levels of both specific and nonspecific proliferation than those of the control mice.
CTL activity in the boosted animals was assessed using splenocytes that were cultured with peptide-pulsed irradiated syngeneic APCs. The peptide used in these studies was the H-2d-restricted immunodominant V3 loop epitope of HIV-1 gp120 IIIB (RIQRGPGRAFVTIGK) (47). Figure 4⇓ shows effector cell killing of peptide-pulsed P815 target cells and demonstrates that specific CTL activity in the mice that received IL-2/Ig was significantly greater than that in the control mice.
Table II⇓ shows the cytokine secretion profiles of recombinant gp120-stimulated splenocytes from the same animals. Splenocytes from pV1J (sham)-injected mice demonstrated only low levels of cytokine expression. Splenocytes from the mice that received pV1J-gp120 plus the Ig control protein exhibited high levels of IFN-γ and IL-2 expression and lower levels of IL-4 and IL-10 expression, consistent with the expected Th1 response (22). The splenocytes from the mice that received pV1J-gp120 plus IL-2/Ig showed higher expression of IFN-γ, IL-4, and IL-10, perhaps reflecting cytokine production from both specifically and nonspecifically activated T cells.
Effects of plasmid IL-2 and plasmid IL-2/Ig on anti-gp120 immune responses elicited by pV1J-gp120
To investigate further the effects of IL-2/Ig on DNA vaccine-elicited immune responses, monocistronic plasmids containing either IL-2 or IL-2/Ig in the pV1J backbone were constructed using standard molecular biologic methods (40). Expression of IL-2 and IL-2/Ig was confirmed and quantitated by transient transfection experiments in COS cells followed by ELISA analyses and functional CTLL stimulation analyses using cell supernatants (data not shown). Experiments were then performed 1) to examine whether plasmid-encoded IL-2/Ig had a stimulatory effect on the vaccine-elicited immune response similar to that of soluble IL-2/Ig protein, and 2) to clarify our findings that IL-2 administered as a dicistronic plasmid with gp120 suppressed the vaccine-induced Ab responses (Fig. 1⇑), whereas IL-2/Ig protein administered after vaccination augmented the immune responses ( Figs. 2–4⇑⇑⇑).
Groups of BALB/c mice (n = 6/group) were immunized with 50 μg of pV1J-gp120 on day 0. Four groups of mice were also inoculated with 200 μg of pV1J-IL-2/Ig, but on days −5, 0, 2, or 5 relative to pV1J-gp120 administration. Figure 5⇓A demonstrates that administration of pV1J-IL-2/Ig before or with pV1J-gp120 significantly decreased the anti-gp120 Ab response. More than a 10-fold reduction in specific Ab titers was observed when the cytokine plasmid was administered with the gp120 plasmid on day 0, a result similar to the reduction in the Ab response obtained with the pV1J-gp120/IL-2 dicistronic construct (Fig. 1⇑). In contrast, administration of pV1J-IL-2/Ig after vaccination with pV1J-gp120 amplified the vaccine-elicited anti-gp120 Ab response. Approximately a 5-fold augmentation of specific Ab titers was observed when the cytokine plasmid was administered on day 2 relative to the gp120 plasmid.
Administering plasmid IL-2/Ig at these different times in relation to the vaccine Ag also affected the gp120-specific cellular immune responses. The mice were boosted after 2 mo with 50 μg of pV1J-gp120 without cytokine treatment and were sacrificed 4 wk later. As shown in Figure 5⇑B, the absolute splenocyte proliferation levels were similar in all cases, but administration of pV1J-IL-2/Ig on day −5 or 0 led to significantly higher levels of nonspecific proliferation (solid bars). Figure 5⇑C shows that pV1J-IL-2/Ig injection before Ag administration decreased the vaccine-elicited CTL activity, whereas pV1J-IL-2/Ig injection following the Ag increased CTL activity. Cytokine secretion profiles of cultured splenocytes, depicted in Table III⇓, showed that all groups of mice receiving pV1J-IL-2/Ig had slightly increased expression levels of IFN-γ, IL-4, and IL-10.
To examine whether pV1J-IL-2, a plasmid expressing native IL-2, could also augment the immune response to pV1J-gp120, a similar experiment was performed. Groups of BALB/c mice (n = 6/group) were immunized with 50 μg of pV1J-gp120 on day 0; groups were also inoculated with 200 μg of pV1J (sham) or 200 μg of pV1J-IL-2 on day 2. Neither pV1J (sham) nor pV1J-IL-2 administered on day 2 augmented the vaccine-elicited Ab or CTL responses (data not shown). In addition, pV1J (sham) did not affect the Ab response when administered with pV1J-gp120 on day 0 (data not shown). These results suggest that the increased immune responses observed with the IL-2/Ig plasmid (Fig. 5⇑) are specific and require the use of the IL-2/Ig fusion construct.
Further effects of administration of cytokine plasmids on anti-gp120 immune responses elicited by pV1J-gp120
Additional experiments were then conducted to investigate 1) whether the timing of the administration of other plasmid cytokines relative to pV1J-gp120 is also important, and 2) whether this phenomenon is also observed in other strains of mice. An experiment analogous to that shown in Figure 5⇑A was conducted using pV1J-GM-CSF as the plasmid cytokine. As shown in Figure 6⇓A, the effects of administering pV1J-GM-CSF on the pV1J-gp120-elicited Ab response were less dramatic than those of pV1J-IL-2/Ig, but the overall trend was similar and significant. Administering pV1J-GM-CSF before the pV1J-gp120 Ag suppressed the vaccine-elicited Ab response, whereas administering pV1J-GM-CSF after the plasmid Ag had perhaps a mild augmenting effect. The suppressive effects observed with GM-CSF in Figure 6⇓A were more marked than the results observed with GM-CSF in Figure 1⇑, possibly due to the fourfold higher dose of cytokine administered in the experiment shown in Figure 6⇓A.
A second experiment, similar in design to that shown in Figure 5⇑A, was conducted with pV1J-IL-2/Ig in C3H mice rather than in BALB/c mice. As shown in Figure 6⇑B, administration of the IL-2/Ig plasmid before or at the time of inoculation with the gp120 plasmid suppressed the vaccine-elicited Ab response, whereas administration of the IL-2/Ig plasmid 2 days after the gp120 plasmid augmented the Ab response. Thus, this phenomenon was not strain specific.
In this report we describe the immunologic effects of coadministering protein and plasmid cytokines with an HIV-1 gp120 DNA vaccine in mice. Administering plasmid cytokines before or with the gp120 vaccine decreased gp120-specific Ab titers and T cell functional activity, whereas administering plasmid cytokines after the gp120 vaccine augmented gp120-specific immune responses. These results demonstrate that Ag-cytokine timing is a critical parameter in determining the overall biologic effect of the cytokine. Moreover, IL-2/Ig was significantly more effective than IL-2 in augmenting DNA vaccine-elicited immune responses, indicating that the Ig fusion markedly enhances the adjuvant properties of this cytokine. IL-2/Ig can thus enhance Ab, CTL, and splenocyte cytokine secretory responses to a DNA vaccine. The augmentation of these immune responses would certainly be desirable for a potential vaccine against HIV.
Although administration of plasmid IL-2/Ig before or with pV1J-gp120 led to markedly diminished gp120-specific immune responses, these animals nevertheless showed high levels of nonspecific cellular proliferation (Fig. 5⇑B, solid bars). These data suggest that IL-2/Ig exposure to a naive immune system leads to a high level of nonspecific cellular activation, above which a specific immune response is elicited poorly. In contrast, IL-2/Ig exposure to an immune system that has recently been primed with a specific Ag leads to augmentation of the specific immune response. IL-2/Ig therefore appears to amplify the existing cellular immune repertoire, and the days immediately following the priming of the immune response appear to be the optimal window of time for augmentation of these immune responses. Since a similar result was obtained for GM-CSF, it is likely that immunostimulatory cytokines in general operate in this fashion. In fact, this sequence of events probably recapitulates the immunology of an acute infection; first, the immune system is primed by an Ag, and then a nonspecific cytokine cascade amplifies the specific response to this Ag.
The use of soluble cytokine/Ig fusion proteins to increase the immunomodulatory effects of the cytokines has been reported previously. IL-10/Ig fusion proteins have been shown to alter the clinical course of septic shock and schistosomiasis in animal models (38, 48). IL-2/Ig and IL-4/Ig fusion proteins have also been shown to affect islet cell allograft acceptance (39). In addition, the use of a TNF receptor-Fc fusion protein has recently been demonstrated to be efficacious in large scale human trials for the treatment of rheumatoid arthritis (49). In the present studies, we have shown that IL-2/Ig, delivered either as a protein or on a plasmid, can augment DNA vaccine-induced immune responses and is significantly more active than IL-2 as a DNA vaccine adjuvant. Cytokine/Ig fusion proteins are divalent in avidity and have a longer in vivo half-life than native cytokines. These factors lead to receptor clustering on cell surfaces and high effective circulating concentrations of cytokine. The augmentation of immune responses seen in the mice inoculated with soluble IL-2/Ig protein or plasmid IL-2/Ig probably results from both these factors.
IL-2 has been used clinically in humans to increase CD4 counts in HIV-infected individuals and as adjunctive therapy for metastatic renal cell carcinoma and melanoma (50). The major drawbacks of IL-2 therapy include the high cost of generating large quantities of the cytokine, the complexity of the dosing regimens, and the severe systemic toxicities associated with treatment, including a vascular leak syndrome, fever, malaise, thrombocytopenia, and hypotension. The experiments described in the present report suggest certain strategies that might improve IL-2 therapy. IL-2/Ig protein can be generated rapidly in large quantities using a simple affinity purification protocol. Moreover, the long half-life of IL-2/Ig makes it possible to maintain constant therapeutic levels of cytokine using an infrequent dosing schedule. Avoiding the peak cytokine levels associated with IL-2 therapy would avoid activating intermediate affinity IL-2Rs on NK cells and would target activated CTLs with high affinity IL-2Rs. IL-2/Ig thus could be more efficacious at lower dosages and have fewer associated toxicities than IL-2. Plasmid IL-2/Ig offers an even simpler and more economic means to produce and administer the cytokine and may be associated with low systemic toxicities.
A number of other reports have examined the effects of coadministration of plasmid cytokines with DNA vaccines in mice. It has been shown that GM-CSF augmented and IFN-γ suppressed the specific Ab response to a rabies DNA vaccine (27). A number of reports describe augmentation of vaccine-elicited CTL responses by administration of plasmid IL-12 (28, 29, 30, 31). Plasmid IL-2 has been shown to augment specific CTL responses and seroconversion frequency, but not Ab titers, in response to a hepatitis C virus DNA vaccine (32). In addition, IL-2 expressed as a dicistron or fusion protein with the hepatitis B virus envelope protein was shown to increase proliferation and transiently augment the Ab response (35).
The present study differs from these previous reports in several ways. First, we have used a potent gp120 DNA vaccine that induces seroconversion in >90% of inoculated mice in the absence of cytokine augmentation (19, 41). Most of the previous studies have used less immunogenic DNA vaccine constructs or have used cytokines to augment suboptimal immune responses. It is possible that the differences among reports reflect in part the different potencies of the baseline DNA vaccines and the different levels of responsiveness to cytokines. Second, we have shown simultaneous cytokine-mediated modulation of multiple immune parameters, including Ab, proliferative, CTL, and cytokine secretion activity. Third, we have compared the adjuvant properties of IL-2/Ig and IL-2, both as proteins and as plasmids, and have found significantly more augmentation with the IL-2/Ig fusion construct. Fourth, we have systematically studied the effects of changing the temporal relationship between delivery of Ag and cytokine. Simultaneous administration of plasmids expressing native cytokines with DNA vaccines, as reported previously, may be capable of enhancing the vaccine-elicited immune responses in certain instances. However, the present study suggests that this approach does not optimally harness the use of plasmid cytokines for augmenting immune responses.
We thank Jorge Reyes, Suzanne Robinson, Christine Lekutis, Joern Schmitz, Abul Abbas, Miguel Campanero, and Jack Strominger for generous advice and assistance.
↵1 This work was supported by an American Academy of Allergy, Asthma, and Immunology Fellowship Grant (to D.H.B.) and by National Institutes of Health Grants CA50139 (to N.L.L.) and AI41521 (to T.B.S.).
↵2 Address correspondence and reprint requests to Dr. Norman L. Letvin, Division of Viral Pathogenesis, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215. E-mail address:
↵3 Abbreviation used in this paper: GM-CSF, granulocyte-macrophage CSF.
- Received January 28, 1998.
- Accepted April 14, 1998.
- Copyright © 1998 by The American Association of Immunologists