Study of Immunization against Anthrax with the Purified Recombinant Protective Antigen of Bacillus anthracis

Abstract

Protective antigen (PA) of anthrax toxin is the major component of human anthrax vaccine. Currently available human vaccines in the United States and Europe consist of alum-precipitated supernatant material from cultures of toxigenic, nonencapsulated strains of Bacillus anthracis. Immunization with these vaccines requires several boosters and occasionally causes local pain and edema. We previously described the biological activity of a nontoxic mutant of PA expressed in Bacillus subtilis. In the present study, we evaluated the efficacy of the purified mutant PA protein alone or in combination with the lethal factor and edema factor components of anthrax toxin to protect against anthrax. Both mutant and native PA preparations elicited high anti-PA titers in Hartley guinea pigs. Mutant PA alone and in combination with lethal factor and edema factor completely protected the guinea pigs from B. anthracis spore challenge. The results suggest that the mutant PA protein may be used to develop an effective recombinant vaccine against anthrax.

Anthrax is a bacterial disease caused by Bacillus anthracis. The disease is normally associated with domestic livestock such as sheep, goats, and cattle, but humans also get infected due to exposure or consumption of infected animals (12). Although the currently available animal and human vaccines are effective, they have limitations. The veterinary vaccine is a suspension of spores from a nonencapsulated, toxigenic, Sterne strain of B. anthracis (19, 20). The use of the veterinary vaccine occasionally results in necrosis at the inoculation site and occasionally causes death of the animal. The human anthrax vaccine consists of aluminum hydroxide-adsorbed supernatant material from fermentor cultures of toxigenic, nonencapsulated strains of B. anthracis (1, 17). The primary immunogenic component of the human vaccine is protective antigen (PA) (5). Immunization with the human vaccine can induce local pain, edema, and erythema, and frequent boosters are required (2).

The virulence of B. anthracis has been shown to be due to two exotoxins and a poly-d-glutamic acid capsule (4, 15). The two toxins are formed by three proteins, PA, lethal factor (LF), and edema factor (EF). The combination of PA with LF makes lethal toxin, causing death of experimental animals and sensitive macrophages. PA in combination with EF increases cyclic AMP concentrations in cells. EF is known to be a calcium- and calmodulin-dependent adenylate cyclase (10), and LF has been proposed to be a Zn²⁺-dependent metalloprotease (9).

Of the antigens studied, only toxin components have been shown to confer protective immunity (7, 14). Somatic components such as capsule, surface polysaccharides, and proteins have been shown not to provide protection (3, 6). PA has been shown to be an essential component of vaccine. It has been suggested that in addition to PA, LF and EF also play an important role in providing immunity (16). In view of these facts, a mutant PA protein lacking biological activity may be the molecule of choice to develop a recombinant vaccine.

In the process of developing cytotoxicity, PA (83-kDa) protein binds to a cell surface receptor and then is cleaved by cellular proteases to generate a cell-bound COOH-terminal 63-kDa protein. This cleavage is essential for the binding and subsequent internalization of LF or EF into the cytosol. The gene coding for PA was mutagenized to generate a noncleavable PA mutant (18). The mutant PA protein bound to the receptor with an affinity equal to that of native PA but failed to bind LF or EF. The mutant PA protein was nontoxic to anthrax toxin-sensitive macrophage cells (J774A.1) and to rats when administered in combination with LF (18). The objective of this study was to evaluate the protective efficacy of mutant PA protein in combination with other toxin components against anthrax.

PA and mutant PA proteins were purified from a spore-forming, protease-deficient Bacillus subtilis strain, DB104, transformed with plasmid pYS5 or pYS6 as described earlier (18). The LF and EF were purified from the culture supernatant of B. anthracis as described earlier (11). Groups of female Hartley guinea pigs weighing 350 g (Charles River, Kingston, N.Y.) received two intramuscular injections, 4 weeks apart. The injections contained 50 μg of each protein in 500 μl of an adjuvant mixture (Ribi Immunochem Research, Inc., Hamilton, Mont.) containing 10 μl of squalene and 1 μl of Tween 80, along with 0.125 mg each of monophosphoryl lipid A, trehalose dimycolate, and the purified, deproteinized cell wall skeleton of Mycobacterium bovis BCG, a strain of the tubercle bacillus. Four weeks after the second immunization, 2 ml of blood was withdrawn from each animal by cardiac puncture, and the serum was examined for the levels of antibody against PA, LF, and EF by enzyme-linked immunosorbent assay (13). Two days after the guinea pigs were bled, they were challenged intramuscularly with 2 × 10⁵ spores of the virulent B. anthracis Ames strain suspended in 0.2 ml of Dulbecco’s phosphate-buffered saline (PBS) containing 0.1% gelatin. The number of animals that died within 3 weeks after challenge were noted (8).

Immunization with native PA elicited high anti-PA titers and completely protected the guinea pigs. However, immunization with native PA in combination with LF and EF provided variable protection, ranging from 83 to 100% (Table 1). Mutant PA alone and in combination with LF and EF also elicited high anti-PA titers and completely protected the guinea pigs against a lethal challenge with Ames spores. Analysis of the survival data by analysis of variance and Fisher’s exact test indicated that all vaccine preparations were significantly better than the PBS plus adjuvant control (P < 0.05) but were not statistically different from each other. The data presented here indicate that mutant PA is an effective substitute for native PA, since immunization with mutant PA provided substantial protection against a virulent anthrax spore challenge. The inability of mutant PA to interact with either LF or EF to form lethal toxin or edema toxin, respectively, makes the serologically active but biologically inactive mutant PA a particularly attractive alternative to native PA as the primary component of a future anthrax vaccine.

TABLE 1.

Immunization of guinea pigs against virulent anthrax spore challenge^a

Immunization	No. of survivors/total no. (%)	Serologic response (titer)
		Anti-PA	Anti-LF	Anti-EF
PA	10/10 (100)	177,828	3	3
PA + LF	8/9 (89)	56,233	8,912	198
PA + EF	5/6 (83)	129,154	12	27,825
PA + LF + EF	8/8 (100)	59,947	24,483	11,364
Mutant PA	10/10 (100)	79,433	3	3
Mutant PA + LF	9/9 (100)	63,095	14,125	278
Mutant PA + EF	10/10 (100)	100,000	16	5,622
Mutant PA + LF + EF	10/10 (100)	35,480	7,079	5,622
PBS + adjuvant	0/10 (0)	3	3	3

^aGroups of 6 to 10 female Hartley guinea pigs were immunized with antigen preparations (50 μg of each protein/dose) at 0 and 4 weeks. Four weeks after the final injection, the animals received an intramuscular challenge of 200,000 Ames spores (50% lethal dose = 10² spores). The number of animals that died within 3 weeks after challenge were noted. All control animals died 2 to 3 days after challenge. The two immunized animals died 10 and 17 days after challenge.