Comparative Immunotoxicity Evaluation of Amphotericin B and ABELCET ®,an Amphotericin B Lipid Complex (ABLC)

Abstract

ABELCET^®, an amphotericin B lipid complex formulation (ABLC) and an aqueous, non–lipid-containing formulation with sodium deoxycholate (AmBd), were evaluated for their potential to induce immunotoxicity in B6C3F1 female mice. ABLC was administered intravenously at doses of 1, 3, and 10 mg/kg daily for 28 days, while AmBd at 1 mg/kg was administered by the same route and duration. The effect of ABLC and AmBd on clinical signs, body weight, and spleen weight was determined. Peritoneal macrophage function was measured by phagocytosis of ⁵¹Cr-labeled chicken red blood cells and generation of hydrogen peroxide during respiratory burst. The ability of natural killer cells to lyse radiolabeled tumor target cells was evaluated in a short-term chromium-release assay. The ability of splenic T and B cells to undergo blastogenesis and of splenic T cells to recognize alloantigens present on foreign cells was assessed in a splenic lymphocyte assay and the ability of mice to generate antibody-forming cells following immunization with sheep red blood cells was measured. Neither ABLC nor AmBd affected the metabolic or functional activity of murine phagocytic cells. These agents also did not cause any biologically significant or dose-related changes in B- or T-cell responses to mitogens, T-cell responses to allogeneic cells in the mixed lymphocyte culture assay, or natural killer cell function. The ability to generate a primary antibody response to a T cell–dependent antigen was also unimpaired. Based on the results of this study, it was concluded that neither ABLC at dose up to 10 mg/kg nor AmBd at dose up to 1.0 mg/kg produce biologically significant immunologic changes in B6C3F1 mice.

Amphotericin B (AmB) is a potent, polyene, broad-spectrum antifungal antibiotic derived from a strain of Streptomyces nodosus. Highly lipophilic and water insoluble, AmB is conventionally formulated with sodium deoxycholate (AmBd; FUNGIZONE^®), a drug formulation that is efficacious but restrictive because of nephrotoxicity, its primary side effect (Bennett 1996). To alleviate the toxicity, amphotericin B lipid complex (ABLC; ABELCET^®) was developed, and its safety evaluation in rats has been described (Zhang, Diener, and Lipman 2006). The recommended human daily dosage for ABLC is 5 mg/kg (Enzon Pharmaceuticals 2002), whereas it is 1 mg/kg for AmBd ( Merck Manual 1992).

The purpose of this study was to determine if ABLC or AmBd specifically targeted the immune system in mice at doses that did not cause overt systemic toxicity. Specific and nonspecific immunity of AmBd and ABLC were evaluated and compared in female B6C3F1 mice. Alteration in humoral immunity was measured by determining the numbers of plaque-forming cells (PFCs) produced after immunization with sheep red blood cells (SRBCs), and the ability of splenic B cells to undergo blastogenesis upon exposure to bacterial lipopolysacchride (LPS). Alterations in cell-mediated immunity were assessed by measuring blastogenesis of splenic T cells to concanavalin A (Con A) or to allogeneic spleen cells. Nonspecific immune function was evaluated by assessing natural killer cell (NK) activity, as well as peritoneal macrophage phagocytosis and hydrogen peroxide generation. These particular assays were selected because they provide a sensitive yet broad overview of the state of immune homeostasis.

MATERIALS AND METHODS

Materials

Amphotericin B lipid complex (ABLC) and Fungizone^® (AmBd) were stored refrigerated, in the dark, in the original containers. Neat ABLC was aliquoted daily into individual containers for subsequent dilution. Mice given the 10 mg/kg dose received ABLC as supplied. The 1 and 3 mg/kg doses were diluted with normal saline (0.5 and 1.5 mg/ml AmB, respectively). AmBd was reconstituted with sterile water once a week. Aliquots were diluted daily with sterile water to achieve AmB concentrations of 0.5 mg/ml. Test articles or saline vehicle were administered intravenously on a mg/kg basis each morning for 28 consecutive days at volumes ranging from 34 to 52 μl. The formulated ABLC was assayed and found to be stable and acceptable in composition, strength and uniformity.

Test System

Four- to 5-week old virus antibody–free female B6C3F1 mice were obtained from Charles River Laboratories (Portage, MI) and quarantined for at least 7 days prior to dosing. Mice were assigned to each dosing group in three experiments by a constrained randomized program based on mean body weight, housed in individual stainless steel suspension cages, exposed to 12-h light/dark cycles and provided with ad libitum pelleted feed (Purina rodent Laboratory Chow 5002) and tap water. All housing conditions and animal husbandry procedures were in compliance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources 1985). Body weights of the mice during the dosing periods ranged between 18 and 27 g in the three experiments (weights were taken twice weekly—Mondays/Fridays).

Animal assignments are summarized in Table 1.

Experiment 1: Cellular Isolation, Phagocytosis, and Hydrogen Peroxide Production

Nonspecific immune function was assessed in the following investigated parameters: body weight; clinical signs; viability, differential count, phagocytosis, and H₂O₂ production of peritoneal macrophages from peritoneal exudates cells (PECs); and NK cell function. PECs were isolated by peritoneal lavage from CO₂-anesthetized mice. Lavage fluid from four mice was combined and the cells were pelleted by centrifugation. Separate samples were counted by hemacytometer or cytocentrifuged onto glass slides for differential counts. Peritoneal macrophages (PMs) were recovered by adherence to plastic for 2 h in RPMI-1640 culture medium supplemented with 20% fetal calf serum. Adherent PMs were then harvested from the culture flasks and resuspended in RPMI medium supplemented with 10% calf serum, prior to testing for phagocytosis and H₂O₂ production. Cellular numbers and viability were counted in a hemacytometer in the presence of trypan blue dye. Differential counts were determined on cytocentrifuge preparations of cell samples stained with Wright’s stain. Receptor-mediated phagocytosis was determined with ⁵¹Cr-chicken red blood cells, using a modification of the method of Smialowicz et al. (1984).

H₂O₂ production was determined in the following manner: Isolated murine PMs were incubated in 96-well tissue culture plates (2 × 10⁵ viable cells/well) with RPMI 1640 and supplemented with 10% fetal calf serum. Recombinant murine gamma interferon (100 IU) was added to half of the wells for each pool to test the ability of interferon to stimulate mouse PMs. After overnight incubation at 37°C in 5% CO₂, the cells were pelleted onto the base of each plate by centrifugation. The supernatant was removed, 100 μl of phenol red solution (Pick and Mizel 1981) and 10 μl of 10 μM phorbol myristate acetate were added to each well and the plates incubated for 2 h at 37°C in 5% CO₂. An H₂O₂ standard curve was prepared in triplicate by serial dilution, and plates were read at a wavelength of 630 nm. H₂O₂concentrations were determined by reverse linear regression.

The natural killer cell assay was performed according to the methods described by Reynolds, Timonen, and Herberman (1981), modified by Thomas et al. (1985a), using effector spleen cell suspensions and suspension cultures of radiolabeled ⁵¹CrO₄ YAC-1 lymphoma target cells. Chromium release was quantitated by a gamma counter.

Experiment 2: Lymphocyte Blastogenesis to Mitogens and Mixed Leukocyte Culture (MLC) Assays

The methodology used to determine lymphocyte blastogenesis to mitogens was from Thomas et al. (1985b) and Oppenheim and Schecter (1976), in which alterations in cell-mediated immunity were assessed by measuring blastogenesis of thymidine-labeled splenic T and B cells to concanavalin A and Salmonella typhosa lipopolysaccharide, respectively. The control group for immunosuppression was injected intraperitoneally (i.p.) with 180 mg/kg cyclophosphamide 48 h prior to assay.

In the mixed leukocyte culture (MLC) assay, lymphocyte blastogenesis was measured in thymidine-labeled splenic T cells exposed to allogeneic spleen cells. The method used was a modification of Hodes and Terry (1974). Stimulator spleen cells were obtained from five untreated DBA/2 male mice (Charles River Laboratories, Portage, MI). Cell suspensions were prepared and adjusted to 5 × 10⁶/ml in supplemented RPMI, irradiated with 2000 rads using a ¹³⁷Ce source to inhibit cell division and then adjusted to 4 × 10⁴/ml. Responder spleen cells were from the same treated B6C3F1 mice used for the mitogen assay. One-tenth milliliter each of responder and stimulator cell suspensions (six replications of 1 × 10⁶ viable cells/ml) were then incubated for 96 h prior to the addition of thymidine. Responder cells also were cultured with medium alone to serve as a control. Prior to incubation, 10 μl 2-mercaptoethanol (5 × 10⁻⁵ M final concentration) was added to each well. In addition, stimulator cells (irradiated and nonirradiated) were cultured with concanavalin A to check the effectiveness of the irradiation in preventing cell division.

Experiment 3: Antibody Plaque Forming Cell (PFC) Assay

Alteration in humoral immunity was measured by determining the numbers of antibody plaque-forming cells (PFCs) produced following immunization with sheep red blood cells (SR-BCs). A modification (Holsapple 1995) of the Jerne plaque assay (Cunningham and Szendberg 1968) was used to quantitate the PFC response to SRBCs. Body weight and clinical signs were also recorded. All mice were injected i.p. with 2 × 10⁸ SRBCs 4 days prior to the end of the 28-day dosing period. This dose of SRBC was empirically determined to produce an optimal immunoglobulin M (IgM) anti-SRBC PFC response 4 days later. Positive-control animals, which were vehicle-treated prior to SRBC injection, were injected i.p. with cyclophosphamide (80 mg/kg) 24 h prior to assay.

Four days post immunization, all mice were weighed and killed by cervical dislocation. Spleens were aseptically removed and weighed, and single-cell suspensions were prepared by gently mashing the tissue in a cell culture dish in RPMI 1640 culture medium supplemented with heat-treated 2% fetal calf serum using a 5-cc plastic syringe plunger. One-tenth milliliter each of diluted spleen cells and SRBC-complement mixture (80% guinea pig complement and 16% washed SRBCs, with pen/strep and glutamine) were then vortexed and added to duplicate slide PFC chambers (Spiral Scientific, Cincinnati, OH), sealed and incubated at 37°C for 1 h. The resulting IgM anti-SRBC plaques were counted. Based on the laboratory’s historical data, production of at least 300 to 1000 PFCs/10⁶ viable spleen cells in vehicle controls constitutes minimum requirements for a valid test.

Data Analyses

Analyses of variance and appropriate post hoc comparisons using Dunnett’s test (Dunnett 1964) were performed on natural logarithmic or logit-transformed data. Individual groups of data were evaluated for outliers according to the method of Dixon (1953) prior to statistical analysis.

RESULTS

In-life Observations in Experiments 1 to 3

No clinical signs of toxicity associated with ABLC were observed in any of the experiments. By contrast, several mice in the AmBd groups had local tissue reactions (desquamation, sloughing of epidermal cells) at the site of the injections in the tail.

Analysis of variance (ANOVA) of mean group body weights revealed slight but significant differences in both body weight measurements made during the first week of dosing in the AmBd group of mice in Experiment 1. ANOVA of body weights at all time points in experiments 2 and 3 were not affected by either ABLC or AmBd treatment (data not shown).

Experiment 1

No effects on viability, numbers, or types of cells recovered by peritoneal lavage were observed when mice were treated with ABLC or AmBd. FC-receptor mediated phagocytosis and H₂O₂ production of peritoneal macrophages also were not affected (Table 2).

Compared to controls, there was no difference in the ability of NK cells from mice treated with ABLC or AmBd to lyse radiolabeled tumor target cells in a short-term chromium-release assay. In contrast, an enhanced response was observed in vehicle-treated mice administered polyinosinic-polycytidylic acid, a known stimulator of NK cell function, 24 h before assay (data not shown).

Experiment 2

Spleen cells from mice treated with ABLC or AmBd responded to T- and B-cell mitogens in a manner roughly equivalent to the saline control group, although the blastogenic responses were somewhat variable (Table 3). As expected, vehicle-treated mice given cyclophosphamide 48 h prior to assay had a significantly reduced response to lipopolysaccharide, but not to concanavalin A, because cyclophosphamide is known to preferentially affect B lymphocyte function (Turk and Parker 1979).

Spleen cells from mice treated with ABLC or AmBd responded equally well to allogeneic spleen cells as the vehicle controls. As expected, vehicle-treated mice given cyclophosphamide 48 h prior to assay did not alter this T-cell response (Table 4).

Experiment 3

In the antibody PFC assay, there was no impairment of specific antibody production (effect on PFC/10⁶ viable spleen cells) in ABLC- or AmBd-treated mice immunized with a T cell–dependent antigen such as SRBCs, whereas cyclophosphamide caused a markedly diminished response. Mice given 1 or 10 mg/kg ABLC or 1 mg/kg AmBd had slightly lower PFC/spleen values than control mice (statistically significant in mice given ABLC at 10 mg/kg); however, the lower response in the ABLC groups was minor compared to the positive-control group and showed no relationship to dose (Table 5).

Absolute and relative spleen weights were slightly increased in mice given ABLC at 3 or 10 mg/kg or AmBd at 1 mg/kg, but statistical significance was limited to absolute weights in the AmBd and positive-control groups. Absolute spleen weights were significantly decreased in the cyclophosphamide controls. Body weight and percent spleen-cell viability in the ABLC and AmBd groups were comparable to controls. Viable cells/spleen were slightly, but significantly, decreased in a non–dose-dependent manner in mice given ABLC at 1 or 10 mg/kg, whereas a marked decrease was present in the cyclophosphamide positive control group.

DISCUSSION AND CONCLUSIONS

Macrophages are the primary cellular defense against microbial infections (Dunn et al. 1985; Hocking and Golde 1979), as they actively phagocytose extracellular materials and are generally more highly exposed to foreign agent than other less active cells (Hocking and Golde 1979). Liposomes containing aminoglycosides enhance the microbicidal activity of macrophages (Dees and Schultz 1990). In addition, immunostimulatory activity has been reported with amphotericin B (Stein et al. 1987; Vecchiarelli, Verducci, and Perito 1986). Highly lipophilic and water insoluble amphoterin B is conventionally formulated with sodium deoxycholate (AmBd; Fungizone^®), a formulation that is efficacious but restrictive because of a number of side effects, including nephrotoxicity (Bennett 1996). To alleviate the toxicity, amphotericin B lipid complex (ABLC; ABELCET^®) was developed. This formulation provides amphotericin B complexed with two phospholipids in a 1:1 drug-to-lipid molar ratio. The two phos-pholipids, l-dimyristoylphosphatidylcholine (DMPC) and l-dimyristoylphosphatylglycerol (DMPG) are present in a 7:3 molar ratio and are suspended in physiological saline at a pH of 5 to 7.

The purpose of this study was to determine if ABLC or AmBd specifically targeted the functional and innate immunity in mice using well established murine test systems and assays validated by the National Toxicology Program (Luster et al. 1988, 1992). Because immunologic changes could be secondary to the systemic toxicity, such as significant body weight lost, and to avoid this situation the high dose used in this study was the maximum dose that did not cause overt systemic toxicity based on our previous studies in mice (data not shown).

Except for some injection-site changes with AmBd, no noteworthy toxic effects were observed with either ABLC or AmBd during the 28-day treatment period. Cell numbers, cell viability, respiratory burst, and macrophage FC-receptor–mediated phagocytosis were unaffected by treatment. Therefore, it can be stated that at the dose levels tested, neither ABLC nor AmBd inhibited or augmented metabolic or functional activity of murine phagocytic cells. The lack of an immunostimulatory effect with amphotericin B may be explained, at least in part, by the fact that the C57B16 parent of the B6C3F1 mouse is a “low responder” to the immunostimulatory activity of AmB (Stein, Little, and Little 1987).

With respect to adaptive and natural immunity, ABLC or AmBd treatment did not result in any biologically significant or dose-related changes in B- or T-cell responses to mitogens, T-cell responses to allogeneic cells in the mixed lymphocyte culture assay, or natural killer cell function. The ability to generate a primary antibody response to a T cell–dependent antigen (which also requires normal macrophage antigen processing) was also unimpaired. Based on the results of this study, it is concluded that neither ABLC at dose up to 10 mg/kg nor AmBd at dose 1.0 mg/kg produce biologically significant immunologic changes in B6C3F1 female mice.