Aluminum Phosphate Adjuvant Research Articles

Activation of dendritic cells and induction of CD4 + T cell differentiation by aluminum-containing adjuvants

Aluminum-containing adjuvants are widely used in licensed human and veterinary vaccines. However, the mechanism by which these adjuvants enhance the immune response and predominantly stimulate a T H2 humoral immune response is not well understood. In this study, the effects of aluminum hydroxide and aluminum phosphate adjuvants on antigen presentation, expression of costimulatory molecules and cytokines by mouse dendritic cells (DCs) and the ability of DCs to induce T helper cell differentiation were investigated. Dendritic cells pulsed with ovalbumin (OVA) adsorbed to aluminum-containing adjuvants activated antigen-specific T cells more effectively than DCs pulsed with OVA alone. Aluminum hydroxide adjuvant had a significantly stronger effect than aluminum phosphate adjuvant. Both aluminum-containing adjuvants significantly increased the expression of CD86 on DCs but only aluminum hydroxide adjuvant also induced moderate expression of CD80. Aluminum-containing adjuvants stimulated the release of IL-1β and IL-18 from DCs via caspase-1 activation. DCs incubated with LPS and OVA induced T H1 differentiation of naïve CD4 + T cells. In contrast, DCs incubated with aluminum/OVA activated CD4 + T cells to secrete IL-4 and IL-5 as well as IFN-γ. Addition of neutralizing anti-IL-1β antibodies decreased IL-5 production and addition of anti-IL-18 antibodies decreased both IL-4 and IL-5 production. Inhibition of IL-1β and IL-18 secretion by DCs via inhibition of caspase-1 also led to a marked decrease of IL-4 and IL-5 by CD4 + T cells. These results indicate that aluminum-containing adjuvants activate DCs and influence their ability to direct T H1 and T H2 responses through the secretion of IL-1β and IL-18.

Influence of Aluminum-Based Adjuvant on the Immune Response to Multiantigenic Formulation

Several adjuvants have been described and tested in humans. However, the aluminum-based adjuvants remain the most widely used component in vaccines today. Emerging data suggest that aluminum phosphate and aluminum hydroxide adjuvants do not promote a strong commitment to the helper T cell type 2 (Th2) pathway when they are coadministered with some Th1 adjuvants. In this regard, subtle differences between both aluminum-based adjuvants have been demonstrated. We have previously shown that subcutaneous immunization, in aluminum phosphate, of a mixture comprising the surface and core antigens of hepatitis B virus (HBV) and the multiepitopic protein CR3 of human immunodeficiency virus type 1 elicits a CR3-specific Th1 immune response. In these experiments, the antigens were adjuvated at the same time. As the final selection of the best adjuvant should be based on experimental evidence, we asked whether aluminum hydroxide allows a better Th1 immune deviation than aluminum phosphate. We also studied several ways to mix the antigens and the impact on CR3-specific interferon (IFN)-gamma secretion. Our findings indicate that aluminum hydroxide allows better Th1 immunodeviation than aluminum phosphate adjuvant for the mixture of HBV antigens and CR3. In addition, CR3-specific IFN-gamma secretion of the various formulations tested was the same irrespective of the order in which the antigens were combined.

Potentiation of the immune response to non-adsorbed antigens by aluminum-containing adjuvants

The degree of antigen adsorption by aluminum-containing adjuvants is considered an important characteristic of vaccines that is related to immunopotentiation by the adjuvant. This study examined immunopotentiation by aluminum phosphate adjuvant in three model vaccines in which the antigen was not adsorbed in the vaccine formulation nor when mixed in vitro with interstitial fluid. In the first model vaccine, aluminum phosphate adjuvant was pre-treated with 0.5 M KH 2PO 4 to minimize the adsorption of dephosphorylated alpha casein. The second model vaccine was composed of aluminum phosphate adjuvant and ovalbumin that was dephosphorylated by treatment with potato acid phosphatase. The third model vaccine consisted of aluminum phosphate adjuvant and lysozyme (LYS). In order to prevent adsorption of lysozyme, the aluminum phosphate adjuvant was pre-treated with fibrinogen, a protein present in interstitial fluid that binds strongly to aluminum phosphate adjuvant. Immunopotentiation was evaluated by measuring antibody production in mice. It was found that all three model vaccines induced antibody titers that were statistically higher than induced by a solution of antigen without adjuvant and similar to vaccines in which the antigens were adsorbed by aluminum phosphate adjuvant. Confocal microscopy experiments suggested that the antigens used in these experiments, even though not adsorbed to the aluminum phosphate adjuvant, were trapped in void spaces within the adjuvant aggregates, resulting in uptake of antigen by dendritic cells.

Association Between Immunogenicity and Adsorption of a Recombinant Streptococcus pneumoniae Vaccine Antigen by an Aluminum Adjuvant

The impact on immunogenicity of the degree of adsorption of three Streptococcus pneumoniae (Sp) vaccine antigens to aluminum adjuvants was studied. The three antigens evaluated (Sp1, Sp2 and Sp3) were highly adsorbed by aluminum hydroxide adjuvant, but not adsorbed by aluminum phosphate adjuvant. All of the Sp antigens adjuvanted with aluminum hydroxide elicited higher antibody responses in mice than formulations prepared with aluminum phosphate or non-adjuvanted antigen. Varying the percent aluminum-bound Sp antigen in the formulated vaccine affected the observed antibody responses. These observations suggest that the antibody response observed for Sp antigens in this study is stimulated by a depot effect of the antigen bound to an aluminum adjuvant.

Coinoculation with hepatitis B surface and core antigen promotes a Th1 immune response to a multiepitopic protein of HIV‐1

It has been defined that strong and multispecific cellular immune responses correlate with a better prognosis during the course of chronic diseases. A cross-enhancing effect on the resulting immune response obtained by the coadministration of recombinant hepatitis B virus (HBV) surface and core Ag was recently observed. With the objective of studying the effect of such Ag on the immune response to coinoculated heterologous Ag and vice versa, several formulations containing the recombinant HBV Ag and a multiepitopic protein (CR3) composed by CTL and Th epitopes from HIV-1 were evaluated by s.c. and mucosal administration. Combinations of two and three Ag were evaluated for cellular and humoral immune responses. The results showed that the best Ag combination for nasal immunization was the mixture comprising the CR3 recombinant HIV protein and both HBV Ag. Similarly, it was also the best formulation for s.c. immunization in aluminium phosphate adjuvant. In conclusion, it is possible to induce a Th1 stimulation of the cellular immune response specific for a HIV-based recombinant protein by formulating this Ag with the recombinant HBV Ag.

Formulation of botulinum neurotoxin heavy chain fragments for vaccine development: mechanisms of adsorption to an aluminum-containing adjuvant

Heavy chain fragments of botulinum neurotoxin serotypes A and B are being developed as a bivalent vaccine for botulism. To potentiate the immune response, an aluminum containing adjuvant will be formulated with the two antigens. The adsorption mechanisms of each antigen to aluminum phosphate and aluminum hydroxide adjuvants were studied. The adsorption of the serotype A antigen to each adjuvant, and the serotype B antigen to aluminum phosphate adjuvant, is dependent on electrostatic attractive forces. The serotype A antigen is basic, and pretreatment with phosphate anions is required for favorable adsorption conditions to aluminum hydroxide adjuvant. In contrast, the serotype B antigen displays a high affinity to aluminum hydroxide adjuvant even when the two species possess the same charge. It is proposed that the serotype B antigen is adsorbed to aluminum hydroxide adjuvant by a ligand exchange mechanism.

Mechanism of Adsorption of Three Recombinant Streptococcus pneumoniae (Sp) Vaccine Antigens by an Aluminum Adjuvant

The adsorption of three Streptococcus pneumoniae (Sp) vaccine antigens byaluminum-containing adjuvants was studied. The antigens showed high binding affinityisotherms with aluminum hydroxide adjuvant described by the Langmuir equation butvirtually no binding to aluminum phosphate adjuvant. The effects of ionic strength andethylene glycol were evaluated to determine whether electrostatic or hydrophobicinteractions were responsible for the observed binding to aluminum hydroxide, but nosignificant change in the adsorptive capacity was observed at either high ionic strengthnor high concentrations of ethylene glycol for any of the antigens. This indicates thatneither electrostatic nor hydrophobic interactions appear to be responsible for theobserved binding, which means that ligand exchange may be the primary mechanism forthis interaction. Further studies to evaluate the ability to elute a Sp antigen fromaluminum hydroxide using fibrinogen (adsorptive coefficient 2.2 mL/µg) as acompetitive protein resulted in evidence that Sp antigen follows the trend that proteinswith higher adsorptive coefficients are able to displace those with lower adsorptivecoefficients. It was also noted that the Sp antigens and ?-lactalbumin (adsorptivecoefficient 1.8 mL/µg) have similar adsorptive coefficients indicative of high affinitybinding isotherms but do not contain phosphate, which has previously been used toexplain ligand exchange for such proteins as ?-casein and hepatitis B surface antigen(HBsAg). Further investigations using ?-lactalbumin as a model protein may elucidatethe binding interaction between the antigens in this study and aluminum adjuvants.

Effect of phosphorylation of ovalbumin on adsorption by aluminum-containing adjuvants and elution upon exposure to interstitial fluid

The phosphate content of commercial ovalbumin was increased from 1.8 to 3.2 mol PO 4/mol ovalbumin by conjugation of phosphoserine and reduced to 1.2 or 0.14 mol PO 4/mol ovalbumin by treatment with potato acid phosphatase. The four ovalbumin samples were completely adsorbed by aluminum hydroxide adjuvant due to electrostatic attraction of the negatively charged ovalbumin and the positively charged aluminum hydroxide adjuvant as well as by ligand exchange of phosphate groups with surface hydroxyl groups. Elution from aluminum hydroxide adjuvant upon exposure to interstitial fluid was inversely related to the degree of phosphorylation of the ovalbumin. The ovalbumin sample containing 3.2 mol PO 4/mol ovalbumin did not elute while the ovalbumin sample containing 0.14 mol PO 4/mol ovalbumin eluted completely from aluminum hydroxide adjuvant during exposure to interstitial fluid for 30 min. Adsorption of the four ovalbumin samples by aluminum phosphate adjuvant was directly related to the degree of phosphorylation of ovalbumin. Adsorption was due to ligand exchange as an electrostatic repulsive force operated between the negatively charged ovalbumin samples and the negatively charged aluminum phosphate adjuvant. The potential for ligand exchange decreased as the phosphorylation of ovalbumin decreased. Elution upon exposure to interstitial fluid was inversely related to the degree of phosphorylation and was more extensive than observed for aluminum hydroxide adjuvant. Adsorption of ovalbumin by aluminum-containing adjuvants and elution upon exposure to interstitial fluid can be controlled by the degree of phosphorylation of both ovalbumin and the aluminum-containing adjuvant.

Optimization of DNA-based Vaccination in Cows Using Green Fluorescent Protein and Protein A as a Prelude to Immunization Against Staphylococcal Mastitis

Staphylococcus aureus is a contagious pathogen that often results in chronic intramammary infections in dairy cows. Current vaccine formulations are ineffective in preventing this infection. The objective of this study was to stimulate an immune response in dairy cows through injection of plasmid DNA designed to express staphylococcal Protein A in transfected cells. Intramuscular and intradermal vaccination sites were evaluated using a plasmid containing the human cytomegalovirus (CMV) promoter/enhancer directing expression of green fluorescent protein (pcDNA3/GFP). DNA was delivered by needle and syringe, or by high-, intermediate-, or low-pressure jet injections (Ped-o-Jet and LectraJet). Five cows per treatment were injected with 0.5mg of plasmid DNA at 6, 4, and 2 wk prepartum. Serum antibody levels determined by ELISA indicated that intradermal high-pressure jet injection elicited a greater immune response compared to needle and syringe injection. Differences in antibody production among low-pressure and needle and syringe treatment groups were not significant. An expression plasmid containing the CMV promoter/enhancer driving expression of the Fc-binding domain of S. aureus Protein A was coinjected into cows by vulvamucosal vaccination using the high-pressure Ped-o-Jet. Beginning 6 wk prepartum, groups of cows (n = 5) were injected three times at 2-wk intervals with DNA in saline, DNA in aluminum phosphate adjuvant, or served as noninjected controls. A cellular immune response to Protein A was detected in 4 of 10 animals, while cellular responses to GFP were not detected. Humoral responses to Protein A were observed in 6 of 10 animals and to GFP in 2 of 10 animals. Aluminum phosphate adjuvant appeared to enhance antibody production in response to Protein A. In experiment 3, a protein boost injection of Protein A was given to six animals approximately 5 mo postpartum. Three animals were nonvaccinated controls, and three were among those stimulated to produce antibody in response to the DNA-based vaccine. These results showed that Protein A specific antibodies remained elevated as compared to nonvaccinated controls and were stimulated in response to the protein boost. However, the magnitude of the response in animals previously vaccinated with DNA was not different than that observed in the nonvaccinated controls. We have shown that a humoral and cellular immune response to abbreviated Protein A can be raised in dairy cows using intravulvamucosal jet injection of a DNA-based vaccine.

Evaluation of the compatibility of a second generation recombinant anthrax vaccine with aluminum-containing adjuvants

Recombinant protective antigen (rPA) is the active pharmaceutical ingredient in a second generation anthrax vaccine undergoing pre-clinical evaluation. This rPA vaccine differs from the currently licensed vaccine, anthrax vaccine adsorbed (AVA), in that the sole component is a recombinant form of protective antigen (PA). Unlike AVA the rPA vaccine contains no lethal factor (LF) or edema factor (EF), components of the two bipartite toxins, nor many other Bacillus anthracis-related contaminating proteins that are present in AVA. The proposed clinical protocol involves adsorption of the rPA to an aluminum-based adjuvant. The adsorptive characteristics of rPA and two aluminum-containing adjuvants were examined in a physiological buffer with and without EDTA. Based on the p I of rPA (p I=5.6) and the zero charge point of aluminum hydroxide adjuvant (11.5) and aluminum phosphate adjuvant (4.5), it was predicted and demonstrated that rPA bound in a more efficient manner to aluminum hydroxide adjuvant than to aluminum phosphate adjuvant in the physiological buffer. Binding of the rPA to the aluminum hydroxide adjuvant was decreased by increasing amounts of phosphate in the buffer. The adsorptive capacity for rPA onto aluminum hydroxide adjuvant in the physiological buffer and in water were calculated to be 0.46 mg rPA/mg aluminum in DPBS and 0.73 mg rPA/mg aluminum in water. This study also demonstrated that upon desorption from the aluminum hydroxide adjuvant the rPA was physically intact and free of detectable aggregates. Further, the eluted material was biologically active in an in vitro cytotoxicity assay. Desorption was only possible after an overnight incubation of 2–8 °C and not after a room temperature incubation reflecting increased contact with the aluminum hydroxide adjuvant over time. These data suggest that the interaction between rPA and aluminum hydroxide adjuvant is predominantly electrostatic in character.

Using Rate of Acid Neutralization to Characterize Aluminum Phosphate Adjuvant

Five aluminum phosphate adjuvants having P/Al molar ratios ranging from 0.74 to 0.26 were prepared. The adjuvants were characterized by both protein adsorptive capacity and rate of acid neutralization at pH 2.25, 25°C. The protein adsorptive capacity was not a useful parameter to compare the initial properties of the adjuvants, as differences in surface charge of the adjuvants required the use of different proteins. In contrast, the rate of acid neutralization allowed a comparison of the freshly precipitated adjuvants and revealed that the rate of acid neutralization was directly related to the P/Al molar ratio. The protein adsorptive capacity decreased slightly during 39 weeks of aging at room temperature. The changes in the rate of acid neutralization were much greater and indicated that a P/Al molar ratio of at least 0.5 was required to minimize the aging of the adjuvants. Thus, the rate of acid neutralization was found to be the most sensitive parameter to characterize aluminum phosphate adjuvants.

Erratum: Tetanus antigen modulates the gene expression profile of aluminium phosphate adjuvant in spleen lymphocytes in vivo

Erratum: Tetanus antigen modulates the gene expression profile of aluminium phosphate adjuvant in spleen lymphocytes in vivo

Elimination of aluminum adjuvants

In vitro dissolution experiments although perhaps not at typical body concentrations and temperatures demonstrated that the α-hydroxycarboxylic acids present in interstitial fluid (citric acid, lactic acid, and malic acid) are capable of dissolving aluminum-containing adjuvants. Amorphous aluminum phosphate adjuvant dissolved more rapidly than crystalline aluminum hydroxide adjuvant. Intramuscular administration in New Zealand White rabbits of aluminum phosphate and aluminum hydroxide adjuvants, which were labelled with 26Al, revealed that 26Al was present in the first blood sample (1 h) for both adjuvants. The area under the blood level curve for 28 days indicated that three times more aluminum was absorbed from aluminum phosphate adjuvant than aluminum hydroxide adjuvant. In vivo studies using 26Al-labelled adjuvants are relatively safe because accelerator mass spectrometry (AMS) can quantify quantities of 26Al as small as 10 −17 g. A similar study in humans would require a whole-body exposure of 0.7 μSv per year compared to the natural background exposure of 3000 μSv per year. The in vitro dissolution and in vivo absorption studies indicate that aluminum-containing adjuvants which are administered intramuscularly are dissolved by α-hydroxycarboxylic acids in interstitial fluid, absorbed into the blood, distributed to tissues, and eliminated in the urine.

Tetanus antigen modulates the gene expression profile of aluminum phosphate adjuvant in spleen lymphocytes in vivo.

Adjuvants play an important role in stimulation of the immune response to antigens. Very little is known about the molecular mechanisms of this stimulation. Here we address this issue by studying gene expression profiles from spleen lymphocytes after in vivo immunization of mice with a clinically relevant vaccine, tetanus toxoid formulated with aluminum phosphate as adjuvant (TT(ADJ)), or the adjuvant alone (ADJ). The Th1/Th2 response to TT(ADJ) was obtained from a combination of up- and downstream markers to conventional cytokines, which were in good agreement with cytokine protein levels. A clustering algorithm revealed that ADJ elicited expression of 47 genes active in cytotoxic lymphocytes, inflammation, oncogenesis, stress, toxicity and cell cycle regulation. In TT(ADJ) these adjuvant-elicited genes were expressed at lower levels and a compensatory onset of protective and inhibitory genes was observed. We conclude that the antigen, to a larger extent than previously recognized, modulates the molecular mechanism of the aluminum phosphate adjuvant and that the identified genes may serve as predictive biomarkers in the development of new adjuvants and vaccines.

Change in the degree of adsorption of proteins by aluminum-containing adjuvants following exposure to interstitial fluid: freshly prepared and aged model vaccines

The ability of interstitial fluid to change the degree of adsorption of ovalbumin to aluminum hydroxide adjuvant or lysozyme to aluminum phosphate adjuvant was studied. Ovalbumin and lysozyme were almost completely eluted after exposure at 37°C to sheep lymph fluid for 4 h or 15 min, respectively. The ability of sheep lymph fluid to elute lysozyme from aluminum phosphate adjuvant did not change as the model vaccine aged. However, only 60% of the ovalbumin adsorbed to aluminum hydroxide adjuvant was eluted during exposure to sheep lymph fluid for 24 h after the model vaccine aged for 11 weeks at 4°C.

Detoxification of endotoxin by aluminum hydroxide adjuvant

Langmuir adsorption isotherms of endotoxin and aluminum-containing adjuvants at pH 7.4 and 25°C revealed that aluminum hydroxide adjuvant has a greater adsorption capacity (283 μg/mg Al) and adsorption coefficient (1.3×10 4 ml/μg) than aluminum phosphate adjuvant (3.0 μg/mg Al, 0.20 ml/μg). The difference in endotoxin adsorption was related to two adsorption mechanisms: electrostatic attraction and covalent bonding. The isoelectric point (iep) of endotoxin is approximately 2. An electrostatic attractive force will be present with aluminum hydroxide adjuvant (iep=11.4), and an electrostatic repulsive force will operate with aluminum phosphate adjuvant (iep=4.6). Endotoxin contains two phosphate groups in the lipid A portion. Covalent bonding occurs with surface aluminum in aluminum hydroxide adjuvant but is inhibited by surface phosphate in aluminum phosphate adjuvant. In-vitro desorption experiments using components of interstitial fluid showed that endotoxin adsorbed by aluminum hydroxide adjuvant was not desorbed by interstitial anions (5 mM phosphate or 2.7 mM citrate) or interstitial proteins (25 mg albumin/ml). The effect of aluminum-containing adjuvants on the systemic response of Sprague–Dawley rats to a 15 μg/kg subcutaneous dose of endotoxin was determined by measuring the serum concentration of tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6). TNF-α and IL-6 were observed in the group which received an endotoxin solution or endotoxin and aluminum phosphate adjuvant. No TNF-α or IL-6 was detected in the group that received endotoxin and aluminum hydroxide adjuvant. Aluminum hydroxide adjuvant detoxifies endotoxin by adsorbing it in the vaccine and then not releasing it in interstitial fluid upon administration.

Aluminium phosphate adjuvants prepared by precipitation at constant pH. Part I: composition and structure

Aluminium phosphate adjuvant was precipitated under constant pH conditions in an effort to characterize materials formed at defined precipitation conditions. A reaction vessel was designed to provide a continuous steady-state process. An aqueous solution containing aluminium chloride and sodium dihydrogen phosphate was pumped into the reaction vessel at a constant rate. A second pump infused a sodium hydroxide solution at the rate required to maintain the desired pH. Precipitations were performed between pH 3.0 and 7.5, at intervals of pH 0.5. The adjuvants were characterized using 27Al NMR, FTIR, Raman and X-ray diffraction methods along with elemental analysis. The results of this study indicate that a continuum of amorphous aluminium hydroxyphosphates were formed having properties that changed as a continuous function of the precipitation pH. The phosphate content decreased as the pH of precipitation increased. 27Al NMR spectra revealed that the majority of the aluminium was octahedrally coordinated, with a small percentage of tetrahedrally coordinated aluminium. The density of the adjuvants was directly related to the pH of precipitation. The most prominent feature of the IR and Raman spectra is the P–O stretching vibration of the structural PO 4 groups. The positions of these bands decreased linearly as the precipitation pH increased. The results of selective deuteration FTIR experiments are consistent with high surface area materials as most of the OH groups were exposed near the surface of the adjuvant.

Stability of aluminium-containing adjuvants during aging at room temperature

Aluminium phosphate adjuvant and aluminium hydroxide adjuvant became more ordered during aging at room temperature. The increased degree of order was accompanied by a decrease in protein adsorption capacity.

Plasmid DNA Vaccines: Tissue Distribution and Effects of DNA Sequence, Adjuvants and Delivery Method on Integration into Host DNA

A variety of factors could affect the frequency of integration of plasmid DNA vaccines into host cellular DNA, including DNA sequences within the plasmid, the expressed gene product (antigen), the formulation, delivery method, route of administration, and the type of cells exposed to the plasmid. In this report, we examined the tissue distribution and potential integration of plasmid DNA vaccines following intramuscular administration in mice and guinea pigs. We compared needle versus Biojector (needleless jet) delivery, examined the effect of aluminum phosphate adjuvants, compared the results of different plasmid DNA vaccines, and tested a gene (the human papilloma virus E7 gene) whose protein product is known to increase integration frequency in vitro. Six weeks following intramuscular injection, the vast majority of the plasmid was detected in the muscle and skin near the injection site; lower levels of plasmid were also detected in the draining lymph nodes. At early time points (1–7 days) after injection, a low level of systemic exposure could be detected. Occasionally, plasmid was detected in gonads, but it dissipated rapidly and was extrachromosomal – indicating a low risk of germline transmission. Aluminum phosphate adjuvant had no effect on the tissue distribution and did not result in a detectable increase in integration frequency. Biojector delivery, compared with needle injection, greatly increased the uptake of plasmid (particularly in skin at the injection site), but did not result in a detectable increase in integration frequency. Finally, injection of a plasmid DNA vaccine containing the human papilloma virus type 16 E7 gene, known to increase integration in vitro, did not result in detectable integration in mice. These results suggest that the risk of integration following intramuscular injection of plasmid DNA is low under a variety of experimental conditions.

Enhanced type I immune response to a hepatitis B DNA vaccine by formulation with calcium- or aluminum phosphate

DNA vaccines induce protective humoral and cell-mediated immune responses in several animal models. When compared with conventional vaccines, however, DNA vaccines often induce lower antibody titers. We have now found that formulation of a DNA vaccine encoding hepatitis B surface antigen with calcium- or aluminum phosphate adjuvants can increase antibody titers by 10–100-fold and decrease the immunogenic dose of DNA by 10-fold. Furthermore, boosting an HBs protein-primed response with the adjuvanted DNA vaccine resulted in a dramatic increase in the HBs-specific IgG2a response reflecting a shift towards a TH1 response. The mechanism by which aluminum phosphate exerts its adjuvant effect is not through increased expression of HBsAg in vivo; rather, the adjuvant appears to increase the number and affinity of HBs peptide antigen-specific IFN-γ and IL-2 secreting T cells.