Disparate adjuvant properties among three formulations of "alum"

Abstract

Aluminum adjuvants, commonly referred to as "alum," are the most widespread immunostimulants in human vaccines. Although the mechanisms that promote humoral responses to alum-adsorbed antigens are still enigmatic, alum is thought to form antigen depots and induce inflammatory signals that, in turn, promote antibody production. It was recently noted that Imject(®) alum, a commercial aluminum-containing adjuvant commonly used in animal studies, is not the physicochemical equivalent of aluminum adjuvant present in human vaccines. This difference raises concerns about the use of Imject(®) alum in animal research as a model for approved aluminum adjuvants. Here, we compared the capacity of Imject(®) alum, Alhydrogel(®), and a traditional alum-antigen precipitate to induce humoral responses in mice to the hapten-carrier antigen, NP-CGG [(4-hydroxy-3-nitrophenyl)acetyl-chicken γ-globulin]. The magnitude of humoral responses elicited by Alhydrogel(®) and precipitated alum was significantly greater than that induced by Imject(®) alum. The strength of the humoral responses elicited by different alum formulations was correlated with the quantity of pro-inflammatory cytokines induced and the numbers of inflammatory cells at the site of immunization. Moreover, Imject(®) exhibited a severely reduced capacity to adsorb protein antigens compared to Alhydrogel(®) and precipitated alum. These findings reveal substantial differences in the immunostimulatory properties of distinct alum preparations, an important point of consideration for the evaluation of novel adjuvants, the assessment of new alum-based vaccines, and in mechanistic studies of adjuvanticity.

Figure 1

Alhydrogel®, Imject® alum, and precipitated alum differ in their capacity to promote production of antigen-specific IgG. (A) Concentrations (mean+SEM) of NP-specific IgM in serum of C57BL/6 mice after i.p. immunization with 10 μg NP₁₁-CGG precipitated in alum (black square) or adsorbed to Alhydrogel® (gray triangle) or Imject® alum (open diamond) are shown. (B) Concentrations (mean+SEM) of total (NIP₂₅-BSA) and high affinity (NIP₅-BSA) NP-specific IgG in serum after immunization with 10 μg NP₁₁-CGG precipitated in alum (black square) or adsorbed to Alhydrogel® (gray triangle) or Imject® alum (open diamond) are shown. (n=4-6 mice for each alum formulation at days 0, 8, and 16; n=3 for days 25 and 42). (C) Concentrations (mean+SEM) of total (NIP₂₅-BSA) NP-specific IgG in serum after immunization with 10 μg NP₆-OVA (left panel) or 10 μg NP₅-rPA (right panel) precipitated in alum (black square) or adsorbed to Alhydrogel® (gray triangle) or Imject® alum (open diamond) are shown (n=3 mice per data point). (D) Concentrations (mean+SEM) of total (NIP₂₅-BSA) NP-specific IgG in serum after immunization with 10 μg NP₁₁-CGG/Alhydrogel® containing 0.17 mg Al (gray triangle), NP₁₁-CGG/Imject® containing 0.69 mg Al (open diamond), or NP₁₁-CGG/Imject® containing 0.17 mg Al (closed diamond). (n=3-6 mice per data point). * P≤0.05, ** P≤0.01.

Figure 2

Induction of antibody-forming cells and germinal centers by different alum formulations. (A) Mice were immunized i.p. with 10 μg NP₁₁-CGG associated with various alum formulations. Frequencies (mean+SEM) of antigen-specific IgG AFC in spleens of naïve mice and mice immunized with NP₁₁-CGG adjuvanted with precipitated alum (black squares), Alhydrogel® (gray triangles), or Imject® alum (open diamonds), as determined by ELISpot assay, are shown. (B) The frequencies (mean+SEM) of NP-specific IgG AFC in bone marrow are shown on days 0, 8, and 16 after immunization with NP₁₁-CGG adjuvanted with precipitated alum (black squares), Alhydrogel® (gray triangles), or Imject® alum (open diamonds). (n=5-6 mice per data point). (C) Representative FACS plots of splenic germinal center B cells (B220⁺TCRβ^-GL-7⁺Fas⁺ cells) from naïve mice and mice immunized with various alum adjuvants (day 8). Representative histological images of B-cell follicles and GCs from spleens of naïve mice and immunized mice are shown at right (green = GL-7, red = TCRβ, blue = B220; scale bar = 100 μm; 10x objective lens, NA: 0.3). (D) Frequencies (mean±SEM) of GC B cells (GL-7⁺Fas⁺ cells) within the splenic B-cell population (B220⁺TCRβ^- cells) at days 0 (naïve), 8, and 16 after immunization with NP₁₁-CGG adjuvanted with precipitated alum (black squares), Alhydrogel® (gray triangles), or Imject® alum (open diamonds) are shown. (E) Affinity maturation, as measured by the ratio of high affinity NP-specific IgG over total NP-specific IgG (mean+SEM) from Fig. 1, is shown at various days after immunization with 10 μg NP₁₁-CGG precipitated in alum (black square) or adsorbed to Alhydrogel® (gray triangle) or Imject® alum (open diamond). (n=3-6 mice per data point). * P≤0.05, ** P≤0.01.

Figure 3

Cytotoxic and inflammatory effects of Imject® alum, Alhydrogel®, and precipitated alum. Mice were injected i.p. with PBS, 10 μg NP₁₁-CGG without adjuvant (“NPCGG”), or 10 μg NP₁₁-CGG adjuvanted with Imject® alum (“Imject”), Alhydrogel® (“Alhydr”), or precipitated alum (“Prec. alum”). Twenty-four hours later, the peritoneal cavity was lavaged and the acellular fraction was analyzed for DNA (A) and LDH (B). The concentrations (mean+SEM) of DNA (A) and LDH (B) for each treatment are shown (n=3-4 mice per treatment). (C)The cellular fraction of peritoneal lavages was analyzed by flow cytometry to characterize the inflammatory exudate. Macrophages were defined as F4/80^hi, B cells as IgM⁺, neutrophils as Ly6G⁺Ly6B⁺, inflammatory monocytes as Ly6G^-Ly6B⁺, and eosinophils as Siglec F⁺F4/80^low. (D) The numbers (mean+SEM) of total peritoneal lavage cells and each leukocyte type are shown (n=4-6 mice per treatment). * P≤0.05, ** P≤0.01.

Figure 4

Induction of serum cytokines by Imject® alum, Alhydrogel®, and precipitated alum. Mice were injected i.p. with PBS, 10 μg NP₁₁-CGG alone (“NPCGG”), or 10 μg NP₁₁-CGG adjuvanted with Imject® alum (“Imject”), Alhydrogel® (“Alhydr”), or precipitated alum (“Prec. alum”). Six hours after injection, mice were bled and sera were analyzed by multiplex assay for concentrations of 23 cytokines. Concentrations (mean+SEM) of IL-5, IL-6, G-CSF, KC, MCP-1, and RANTES are shown. (n=4-6 mice per treatment). * P≤0.05, ** P≤0.01

Figure 5

Imject® exhibits a poor capacity to adsorb antigen. (A) Following NP₁₁-CGG preparation with precipitated alum, Alhydrogel®, or Imject®, the liquid phase of each vaccine was assayed by ELISA for soluble antigen. Concentrations of soluble NP₁₁-CGG in each vaccine are shown, with each point representing a separate vaccine preparation (n=3). The hashed line represents the antigen concentration in the absence of antigen adsorption (100 μg/ml). (B) The concentrations of soluble CGG (unhaptenated) in vaccines prepared with Alhydrogel® and Imject® are shown (n=2). CGG adsorption to alum formulations took place in PBS (left panel) or physiologic saline (right panel). (C) Concentrations of unadsorbed NP₆-Ova and NP₅-rPA in the various alum-containing vaccines are shown (n=3 preparations per antigen). (D) Two NP₁₁-CGG/Imject® vaccines were prepared, one according to the manufacturer's specifications (“unwashed”, closed diamond) and containing both soluble and adsorbed antigen; the other vaccine was washed of unadsorbed antigen prior to immunization (“washed”, open diamond). Two vaccines of NP₁₁-CGG/Alhydrogel® were prepared similarly (“unwashed,” closed triangle and “washed,” open triangle). Total NP-specific IgG concentrations (mean+SEM) in serum at days 8 and 16 after immunization with washed and unwashed vaccines are shown.