WO1999047165A1

WO1999047165A1 - Production of purified mutant enterotoxin for use as an adjuvant

Info

Publication number: WO1999047165A1
Application number: PCT/US1999/005625
Authority: WO
Inventors: John D. Clements; Martin Friede
Original assignee: The Administrators Of The Tulane Educational Fund; Smithkline Beecham
Priority date: 1998-03-18
Filing date: 1999-03-17
Publication date: 1999-09-23
Also published as: AU3089599A; CO4970841A1

Abstract

The present invention is directed towards compositions and methods which provide a genetically distinct mutant of E. coli heat-labile enterotoxin (LT). Specifically, the invention relates to formulations and methods for use of a mutant LT designated LT (R192G) which is shown to have adjuvanticity without the toxicity associated with wild type LT, said mutant holotoxin being provided in pure form, substantially free of excess B-subunit. The LR(R192G) is useful when combined with free B-subunit to induce an immune response.

Description

PRODUCTION OF PURIFIED MUTANT ENTEROTOXIΝ FOR USE AS AN ADJUVANT

1. FIELD OF THE INVENTION

The present invention is directed towards 5 compositions and methods which provide a genetically distinct mutant of E. coli heat-labile enterotoxin (LT) .

Specifically, the invention relates to formulations and methods for use of a mutant LT designated LT(R192G), modified by a single amino acid substitution that substantially

10 reduces its inherent toxicity but leaves intact the adjuvant properties of the molecule, provided as a single mutant A- subunit with five B-subunits, i.e., mutant holotoxin, which is shown to have adjuvanticity without the toxicity associated with wild type LT, said mutant holotoxin being

--. provided in pure form, free of excess B-subunit.

2. BACKGROUND OF THE INVENTION

The World Health Organization report of Infectious

Disease deaths for 1995 indicated that there were more than

13 million deaths world-wide during that year. The majority 20 of those deaths were caused by organisms that first make contact with and then either colonize or cross mucosal surfaces to infect the host. The overall morbidity caused by these organisms and other pathogens that interact with mucosal surfaces is impossible to calculate.

25 Traditional vaccine strategies that involve parenteral immunization with inactivated viruses or bacteria or subunits of relevant virulence determinants of those pathogens do not prevent those interactions. In fact, traditional vaccine strategies do not prevent infection but

30 instead resolve infection before disease ensues. In some cases, HIV for example, once the virus crosses the mucosal surface and enters the host cell, be that a dendritic cell, an epithelial cell, or a T-cell, the host-parasite relationship is moved decidedly in favor of the parasite (HIV) . In that case, as in many others, a vaccine strategy that does not prevent the initial infection of the host is unlikely to succeed.

Recently, a great deal of attention has focused on mucosal immunization as a means of inducing secretory IgA (slgA) antibodies directed against specific pathogens of mucosal surfaces. The rationale for this is the recognition that slgA constitutes greater than 80% of all antibodies produced in mucosal -associated lymphoid tissues in humans and that slgA may block attachment of bacteria and viruses, neutralize bacterial toxins, and even inactivate invading viruses inside of epithelial cells. In addition, the existence of a Common Mucosal Immune System permits immunization on or at one mucosal surface to induce secretion of antigen-specific slgA at distant mucosal sites. It is only now being appreciated that mucosal immunization may be an effective means of inducing not only slgA but also systemic antibody and cell-mediated immunity. ^Tne mucosal immune response can be divided into two phases (McGhee and Kiyono, 1993, Infect Agents Dis 1_.2:55-73) .

First, the inductive phase involves antigen presentation and the initiation events which dictate the subsequent immune response. During the initiation events, antigen-specific lymphocytes are primed and migrate from the inductive sites

( e . g. , Peyer's patches in the enteric mucosa) through the regional lymph nodes, into the circulation and back to mucosal effector sites ( e . g. , lamina propria) . Once these effector cells have seeded their effector sites, the second phase, or effector phase, of the mucosal immune response can occur. A significant difference between mucosal immunization and parenteral immunization is that both mucosal and systemic immunity can be induced by mucosal immunization while parenteral immunization generally results only in systemic responses .

Most studies conducted to date have dealt with the secretory antibody component of the mucosal response and the 5 complex regulatory issues involved with induction of slgA following mucosal immunization and not with the systemic antibody response or cellular immunity induced by mucosal immunization. In that regard, it is important to understand the type of helper T lymphocyte response induced by mucosal immunization since the type of helper T lymphocyte stimulated

10 by an antigen is one of the most important factors for defining which type of immune response will follow. Mosmann and colleagues (Cherwinski et al . , 1987, Journal of Experimental Medicine 166 : 1229-1244 ; Mosmann and Coffman, 1989, Annual Revi ews of Immunology 2 : 145 - 113 ) discovered that

^■*-⁵ there are at least two different types of helper T lymphocytes (Th) which can be identified based on cytokine secretion. Thl lymphocytes secrete substantial amounts of IL-2 and INF-gamma and execute cell-mediated immune responses ( e . g. , delayed type hypersensitivity and macrophage

20 activation), whereas Th2 lymphocytes secrete IL-4, IL-5, IL-6 and IL-10 and assist in antibody production for humoral immunity. Theoretically then, antigenic stimulation of one T helper cell subset and not the other would result in production of a particular set of cytokines which would define the resulting immune response.

The presence of IL-2 and INF-gamma coupled with an antigenic stimulus presented by macrophages in the context of Class II MHC molecules can initiate Thl type responses. The ability of Thl cells to secrete IL-2 and INF-gamma further amplifies the response by activating Thl cells in an

30 autocrme fashion and macrophages in a paracrine fashion.

These activated leukocytes can release additional cytokines ( e . g. , IL-6) which may induce the proliferation and

- 3 - differentiation of antigen specific B lymphocytes to secrete antibody (the effector phase) . In this scenario, the predominant isotype secreted by murine B lymphocytes is often

IgG2a. In a second scenario (Urban et al . , 1992, Immunol Rev

5 127 : 205-220) , antigens such as allergens or parasites can effectively stimulate a Th2 lymphocyte response (the inductive phase) . Presentation of such antigens to Th2 cells can result in the production of the lymphokines IL-4 and IL-5 which can induce antigen specific B lymphocytes to secrete

IgE and IgGl or induce eosinophillia, respectively (the 0 effector phase) . Furthermore, stimulated Th2 cells can secrete IL-10 which has the ability to specifically inhibit secretion of IL-2 and INF-gamma by Thl lymphocytes and also to inhibit macrophage function.

While these representations are simplistic, it is 5 obvious that the type of T helper cell stimulated affects the resultant cellular immune response as well as the predominant immunoglobulin isotype secreted. Specifically, IL-4 stimulates switching to the IgE and IgGl isotypes whereas INF-gamma stimulates IgG2a secretion. Numerous studies, 0 predominantly conducted in vitro, have suggested that IL-5, IL-6 and TGF-beta (Th3) can cause isotype switching to IgA.

2.1. MUCOSAL ADJUVANTS

Mucosally administered antigens are frequently not immunogenic. A number of strategies have been developed to facilitate mucosal immunization, including the use of attenuated mutants of bacteria { e . g. , Salmonella spp. ) as carriers of heterologous antigens, encapsulation of antigens into microspheres, gelatin capsules, different formulations of liposomes, adsorption onto nanoparticles, use of 0 lipophilic immune stimulating complexes, and addition of bacterial products with known adjuvant properties. The two bacterial products with the greatest potential to function as

- 4 - mucosal adjuvants are cholera toxin (CT) , produced by various strains of Vibrio cholerae, and the heat-labile enterotoxin (LT) produced by some enterotoxigenic strains of Escherichia coli (Clements et al . , 1988, Vaccine _ : 269-277; Elson, 1989, 5 Immunology Today 146:29-33; Lycke et al . , 1992, European Journal of Immunology 22:2277-2281; Xu-Amano et al . , 1993, Journal of Experimental Medicine 178 : 1309-1320 ; Yamamoto et al . , 1996, Annals of the New York Academy of Sciences 778 : 64- 71) .

Although LT and CT have many features in common, these are clearly distinct molecules with biochemical and immunologic differences which make them unique (see below) . Both LT and CT are synthesized as multisubunit toxins with A and B components. On thiol reduction, the A component dissociates into two smaller polypeptide chains. One of 5 these, the Al piece, catalyzes the ADP-ribosylation of the stimulatory GTP-binding protein (GSa) in the adenylate cyclase enzyme complex on the basolateral surface of the epithelial cell resulting in increasing intracellular levels of cAMP. The resulting increase in cAMP causes secretion of 0 water and electrolytes into the small intestine through interaction with two cAMP-sensitive ion transport mechanisms involving 1) NaCl cotransport across the brush border of villous epithelial cells, and 2) electrogenic Na dependent Cl secretion by crypt cells (Field, 1980, Secretory Diarrhea _ pp21-30) . The B-subunit binds to the host cell membrane receptor (ganglioside GM1) and facilitates the translocation of the A-subunit through the cell membrane.

Recent studies have examined the potential of CT and LT as a mucosal adjuvant against a variety of bacterial and viral pathogens using whole killed organisms or purified 0 subunits of relevant virulence determinants from these organisms. Representative examples include tetanus toxoid (Xu-Amano et al . , 1993, Journal of Experimental Medicine ____78: 1309-1320; Yamamoto et al . , 1996, Annals of the New York

Academy of Sciences 778 : 64-71 ; Xu-Amano et al . , 1994, Vaccine

.12.: 903-911) , inactivated influenza virus (Hashigucci et al . ,

1996, Vaccine _L4: 113-119; Katz et al . , 1996, Options for the

5 control of influenza . III. , pp292-297; Katz et al . , 1997,

Journal of Infectious Diseases 175 :352-363) , recombinant urease from Helicobacter spp . (Lee et al . , 1995, Journal of

Infectious Diseases 1 12 -. 161 - 111 ; Weltzin et al . , 1997,

Vaccine _.4:370-376), pneumococcal surface protein A from

_. _n Streptococcus pneumoniae (Wu et al . , 1997, Journal of

Infectious Diseases 175 : 839-846) , Norwalk virus capsid protein, synthetic peptides from measles virus (Hathaway et al., 1995, Vaccine 13_.: 1495-1500) , and the HIV-1 C4/V3 peptide

T1SP10 MN(A) (Staats et al . , 1996, Journal of Immunology

157 : 462-472) . There are many other examples and it is clear 15 that both LT and CT have significant potential for use as adjuvants for mucosally administered antigens ( see Dickinson and Clements, 1996, Mucosal Vaccines pp73-87 for a recent review) . This raises the possibility of an effective immunization program against a variety of pathogens involving

²⁰ the mucosal administration of killed or attenuated agents or relevant virulence determinants of specific agents in conjunction with LT or CT. However, the fact that these "toxins" can stimulate a net lumenal secretory response may prevent their use. For instance, as little as 5 μg of

25 purified CT was sufficient to induce significant diarrhea in volunteers while 25 μg was shown to elicit a full 20-liter cholera purge (Levine et al . , 1983, Microbiological Reviews 4.7:510-550) . In recently conducted volunteer studies with LT administered alone or in conjunction with the V. cholerae Whole Cell/B-Subunit Vaccine, LT was shown to induce fluid secretion at doses as low as 2.5 μg when administered in conjunction with the vaccine, while 25 μg of LT elicited up to 6-liters of fluid. While the adjuvant effective dose in humans for either of these toxins has not been established, experiments in animals suggest that it may be a comparable to the toxic dose. Taken together, these studies indicate that while LT and CT may be attractive as mucosal adjuvants, c studies in animals do not reflect the full toxic potential of these molecules in humans, and that toxicity will seriously limit their practical use for humans.

A number of attempts have been made to alter the toxicity of LT and CT, most of which have focused on eliminating enzymatic activity of the A-subunit associated 0 with enterotoxicity . The majority of these efforts have involved the use of site-directed mutagenesis to change amino acids associated with the crevice where NAD binding and catalysis is thought to occur. Recently, a model for NAD binding and catalysis was proposed (Domenighini et al . , 1994, ⁵ Molecular Microbiology 14:41-50; Pizza et al . , 1994,

Molecular Microbiology L4: 51-60) based on computer analysis of the crystallographic structure of LT (Sixma et al . , 1991, Nature (London) 3_51: 371-377 ; Sixma et al . , 1993, Journal of Molecular Biology 230 : 890-918) . Replacement of any amino 0 acid in CT or LT involved in NAD-binding and catalysis by site-directed mutagenesis has been shown to alter

ADP-ribosyltransferase activity with a corresponding loss of toxicity in a variety of biological assay systems (Lycke et al . , 1992, European Journal of Immunology 22:2277-2281;

Burnette et al . , 1991, Infection and Immuni ty 59:4266-4270;

Harford et al . , 1989, European Journal of Biochemistry

183 : 311 -316 ; Hase et al . , 1994, Tn-fection and Immuni ty

_5_2 :3051-3057; Lobet et al . , 1991, Infection and Immuni ty

_.59:2870-2879; Merritt et al . , 1995, Nature Structural Biology

2:269-272; Moss et al . , 1993, Journal of Biological Chemistry 0

2 8_.: 6383-6387; Tsuj i et al . , 1991, FEBS Letters 291:319-321;

Tsuj i et al . , 1990, Journal of Biological Chemistry

265 :22520-22525) . In addition, it has been shown that

- 7 - exchanging K for E112 in LT not only removes ADP-ribosylating enzymatic activity, but cAMP activation and adjuvant activity as well (Lycke et al . , 1992, European Journal of Immunology

22_. :2277-2281) . A logical conclusion from the Lycke et al . studies was that ADP-ribosylation and induction of cAMP are essential for the adjuvant activity of these molecules. As a result, a causal linkage was established between adjuvanticity and enterotoxicity . That is, the accumulation of cAMP responsible for net ion and fluid secretion into the gut lumen was thought to be a requisite to adjuvanticity.

Recent studies by a number of laboratories have challenged that linkage.

Dickinson and Clements (Dickinson and Clements,

1995, Infection and Immuni ty 63:1617-1623) (Clements et al . ) explored an alternate approach to dissociation of enterotoxicity from adjuvanticity. LT requires proteolysis of a trypsin sensitive bond to become fully active. In this enterotoxin, that trypsin sensitive peptide is subtended by a disulfide interchange that joins the Al and A2 pieces of the A-subunit. In theory, if the Al and A2 pieces cannot separate, Al will not be able to find its target (adenylate cyclase) on the basolateral surface or assume the conformation necessary to bind or hydrolyze NAD.

The mutant of Clements et al . has been described more fully in PCT Publication WO96/06627, incorporated herein by reference. The mutant LT holotoxin, designated LT(R192G), was constructed using site-directed mutagenesis to create a single amino acid substitution within the disulfide subtended region of the A-subunit separating Al from A2. This single amino acid change altered the proteolytically sensitive site within this region, rendering the mutant insensitive to trypsin activation. The physical characteristics of this mutant were examined by SDS-PAGE, its biological activity was examined on mouse Y-l adrenal tumor cells and Caco-2 cells, its enzymatic properties determined in an in vitro NAD:agmatine ADP-ribosyltransferase assay, and its immunogenicity and immunomodulating capabilities determined by testing for the retention of immunogenicity and adjuvanticity.

PROPERTIES OF LT(R192G)

• 100 - 1,000 fold less active than cholera toxin or native LT in the mouse Y-l adrenal cell assay

• Not sensitive to proteolytic activation

• Does not possess in vitro NAD:agmatine ADP- ribosyltransferase activity

• Does not increase production of cAMP in cultured Caco-2 cells

• Reduced enterotoxicity in the patent mouse intestinal challenge model when compared to native LT

• Promotes the development of both humoral (antibody) and cell -mediated immune responses against co- administered antigens of a pathogenic microorganism in both the systemic and mucosal compartments • Functions as an effective adjuvant when administered mucosally (i.e., orally, intranasally) or parenterally (i.e., subcutaneously)

• Lacks enterotoxicity in humans at adjuvant- effective doses

WO 96/06627 describes plasmid pBD95 which can be used to obtain the mutant LT(R192G) . Although not described in WO 96/06627, it has recently been discovered that when plasmid pBD95 is used to produce the mutant holotoxin,

LT(R192G), by expressing pBD95 in E. coli , varying amounts of free B-subunit can als be recovered as well as the holotoxin.

This phenomenon is well known to those of skill in the art since some excess B-subunit is always present following purification of LT or CT by galactose affinity chromatography . Pizza et al . (Pizza et al . , 1994, Molecular

Microbiology 14:51-61) report mutant and wild-type AB5/AB5+B5 ratios that vary from 40% to 98% depending upon the type of mutation. Such excess B-subunit can be separated from holotoxin by gel filtration chromatography due to the difference in molecular weight between the holotoxin and the free B-subunit pentamer (84 kd vs. 56 kd) .

LT(R192G) has been shown to possess the capability of enhancing an immune response ( e . g. , IgG, IgA) to antigens unrelated to LT or LT(R192G) . Recent experimental evidence shows that LT(R192G) has utility as an adjuvant for mucosally or parenterally administered antigens; such administration results in the production of serum IgG and/or mucosal slgA as well as cell-mediated immune responses against the antigen with which LT(R192G) is delivered and, more importantly, to protect against subsequent challenge with infectious organisms. LT(R192G) has been shown to be an effective mucosal adjuvant and has recently been evaluated in humans in several Phase I safety studies . More recently, Tsuj i et al . (Tsuji et al . , 1997,

Immunology 90_ : 116 - 182 ) demonstrated that a protease-site deletion mutant LT(Δ192-194) also lacks in vitro

ADP-ribosylagmatine activity, has a ten- fold reduction in enterotoxicity in rabbit ligated ileal loops, and a 50% reduction and delayed onset of cAMP induction in cultured myeloma cells. LT(Δ192-194) was shown to have increased adjuvant activity for induction of serum IgG and mucosal IgA against measles virus when compared to native LT, LT-B, or LT(E112K). LT(Δ192-194) was effective when administered intranasally, subcutaneously, intraperitoneally, or orally although mucosal IgA responses were only demonstrated following mucosal administration. These investigators also demonstrated increased adjuvant activity for mucosally

- 10 - administered LT(Δ192-194) in conjunction with KLH, BCG, and Ova. These findings are consistent with the findings with LT(R192G). Komase et al . (Komase et al . , 1998, Vaccine JL6_.(2/3) :248-254) shows inclusion of free B-subunit with a 5 mutant LT, in which the Arg at position 7 was changed to Lys, allows less mutant LT to be used for intranasal immunization.

Citation or identification of any reference in

Section 2 or any section of this application shall not be construed as an admission that such reference is available as 10 prior art to the present invention.

3. SUMMARY OF THE INVENTION

The present invention provides an advantageously easily reproducible means of obtaining pure LT(R192G) free I⁵ from excess B-subunit of LT, by providing a novel plasmid which expresses LT(R192G) free of excess B-subunit.

The present invention also provides for an expressable DNA encoding both a mutant A-subunit, which has arginine at position 192 changed to glycine, and a wild-type 20 B-subunit, which plasmid allows for the expression of LT(R192G) holotoxin with less than 10% excess B-subunit.

It has been observed by that an excess amount of free B-subunit of the heat-labile enterotoxin of E. coli

(LT-B) is able to enhance the adjuvanticity of LT(R192G) . In addition, free B-subunit in combination with LT(R192G) was 25 found to qualitatively enhance the immunological outcome when LT(R192G) was used as an oral adjuvant. See U.S. Patent Application Serial No. 09/040,990, filed concurrently. The present invention provides a method for producing pure LT(R192G), substantially free from excess B-

30 subunit . This can allow for subsequent formulation of adjuvant compositions comprised of a controlled,

11 - reproduceable ratio of LT(R192G) to added free B-subunit which are advantageously useful to induce an immune response.

4. DEFINITIONS As used herein, the term "holotoxin" refers to a complex of five B-subunits and one A-subunit of heat-labile enterotoxin.

As used herein, the term "free B-subunit" refers to the B-subunit of heat-labile enterotoxin substantially free from the A-subunit of heat -labile enterotoxin.

As used herein, the term "excess B-subunit" refers to an amount of B-subunit which results in greater than a 5:1 ratio of B-subunits to A-subunit, 5:1 being the ratio of B:A subunits present in native heat-labile enterotoxin holotoxin.

As used herein, the term "qualitatively enhanced" refers to an immune response which differs from the type of response elicited by adjuvant and immunogen without excess B- subunit . For example, when administered orally, immunogen with LT(R192G) with excess B-subunit elicits an enhanced T- cell response as compared to immunogen and LT(R192G) without excess B-subunit, which elicits a mostly humoral response.

As used herein, the term "quantitatively enhanced" refers to an immune response which is greater than normal, but does not differ in the type of immune response elicited.

In one embodiment, adjuvanticity of LT(R192G) is enhanced four fold, such that only one fourth the amount of LT(R192G) with excess B-subunit is required, as compared to LT(R192G) without excess B-subunit, to elicit a comparable immune response .

5. BRIEF DESCRIPTION OF THE FIGURES

The present invention may be understood more fully by reference to the following detailed description of the

12 invention, examples of specific embodiments of the invention and the appended figures in which:

Figure 1 is a schematic diagram of plasmid pCS95, which encodes both subunits LT A and B under the control of the lac promoter. Figure 1A illustrates the construction of plasmid pCS95 which contains the nucleotide sequence encoding mutant LT(R192G) . Plasmid pCS95 was constructed by replacing the BamHI-Xbal of pBD95 with the BamHI-Xbal fragment of pDF82. Figure IB shows the single amino acid change in

LT(R192G). Plasmid pCS95 provides LT(R192G) which contains the single base substitution at amino acid residue 192 of subunit A, coding for Gly rather than Arg, which preserves the reading frame but eliminates the proteolytic site.

Figure 2 is a graphic illustration of the effect of various ratios of free B-subunit to LT(R192G) in the patent mouse intestinal assay. For these studies, LT(R192G) with no excess B-subunit was admixed with different ratios of B-subunit and examined for enterotoxicity in the patent mouse assay. Groups of mice were orally inoculated with native LT at 5, 25, 50 or 100 μg, or with 25 μg of LT(R192G) admixed with a different amount of free B-subunit. Following a three hour interval, the gut: carcass ratio of each animal was determined. The gut-carcass ratio is defined as the intestinal weight divided by the remaining carcass weight.

There were three animals per group and the means for each data point are shown.

Figure 3 is an additional graphic illustration of the effect of excess B-subunit in the patent mouse intestinal assay. For these studies, groups of mice were orally inoculated with native LT at 5 , 25, or 125 μ , or with 25 μg of LT(R192G) . Other groups received either 25 μg of native T or 25 μg of LT(R192G) admixed with a 3:1 or 10:1 excess of free B-subunit. Following a three hour interval, the gut: carcass ratio of each animal was determined. There were

13 three animals per group and the means for each data point are shown .

Figure 4 is a graphic illustration of the effect of excess B-subunit on the ability of LT(R192G) to function as c an immunologic adjuvant for induction of serum IgG when administered intranasallv. Mice were immunized intranasally with Ovalbumin (Ova) alone or in conjunction with 5 μg of

LT(R192G) or 1.25 μg of LT(R192G) plus 3.75 μg of excess free

B-subunit, designated 1AB5:3B5. Serum anti -Ova IgG was determined by ELISA. There were seven animals per group and 0 the means for each data point are shown.

Figure 5 is a graphical illustration of the effect of excess B-subunit on the ability of LT(R192G) to function as an immunologic adjuvant for production of antigen- specific

Thl-type cytokines, specifically, IFN-gamma, by mononuclear 5 cells from the spleens of animals immunized intranasally. Mice were immunized intranasally with Ovalbumin (Ova) alone or in conjunction with 5 μg of LT(R192G) or 1.25 μg of LT(R192G) plus 3.75 μg of excess free B-subunit, designated 1AB5:3B5. Cytokines were determined by ELISA following a o T-cell restimulation assay.

Figure 6 is a graphic illustration of the effect of excesss B-subunit on the ability of LT(R192G) to function as an immunologic adjuvant for production of antigen-specific

Th2-type cytokines, specifically, IL-10, by mononuclear cells from the spleens of animals immunized intranasally . Mice 5 were immunized intranasally with Ovalbumin (Ova) alone or in conjunction with 5 μg of LT(R192G) or 1.25 μg of LT(R192G) plus 3.75 μg of excess free B-subunit, designated 1AB5:3B5.

Cytokines were determined by ELISA following a T-cell restimulation assay. 0 Figure 7 is a graphical demonstration that excess

B-subunit enhances the ability of LT(R192G) to function as an immunologic adjuvant for induction of serum IgG when

14 - administered orally. Mice were immunized orally with a purified bacterial protein, Colonizing Factor I (CFAI) from enterotoxigenic E. coli , in conjunction with 6.25 μg of

LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg of free B-subunit, designated 1AB5:3B5. Serum anti -CFAI IgG was determined by ELISA. There were seven animals per group and the means for each data point are shown.

Figure 8 is an additional graphic illustration that excess B-subunit enhances the ability of LT(R192G) to function as an immunologic adjuvant for induction of serum

IgG when administered orally. Mice were immunized orally with Ovalbumin (Ova) alone or in conjunction with 25 μg of

LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg of free

B-subunit, designated 1AB5:3B5. Serum anti-Ova IgG was determined by ELISA. There were ten animals per group and the means for each data point are shown.

Figure 9 is a graphic demonstration that excess B-subunit enhances the ability of LT(R192G) to function as an immunologic adjuvant for production of antigen-specific Thl-type cytokines, specifically, IFN-gamma, by mononuclear cells from the spleens of animals immunized orally. Mice were immunized orally with Ovalbumin (Ova) alone or in conjunction with 25 μg of LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg of free B-subunit, designated 1AB5:3B5.

Cytokines were determined by ELISA following a T-cell restimulation assay.

Figure 10 is a graphic demonstration that excess

B-subunit enhances the ability of LT(R192G) to function as an immunologic adjuvant for production of antigen-specific

Th2-type cytokines, specifically, IL-10, by mononuclear cells from the spleens of animals immunized orally. Mice were immunized orally with Ovalbumin (Ova) alone or in conjunction with 25 μg of LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg

15 of free B-subunit, designated 1AB5:3B5. Cytokines were determined by ELISA following a T-cell restimulation assay.

6. DETAILED DESCRIPTION OF THE INVENTION The present invention provides pure LT(R192G) by providing novel compositions of LT(R192G) and methods of using said compositions as adjuvants.

The present invention also provides for an expressable DNA encoding both a mutant A-subunit, which has arginine at position 192 changed to glycine, and a wild-type

B-subunit, which plasmid allows for the expression of

LT(R192G) holotoxin with less than 10% excess B-subunit.

LT is a member of the ADP-ribosylating family of bacterial toxins and, like other bacterial toxins that are members of the A-B toxin family, requires proteolysis of a trypsin sensitive bond to become fully active. That trypsin sensitive peptide is subtended by a disulfide interchange that joins the Al and A2 pieces of the A-subunit. In theory, if the Al and A2 pieces cannot separate, Al will not be able to find its target (adenylate cyclase) on the basolateral surface or assume the conformation necessary to bind or hydrolyze NAD. An unexpected finding was that LT(R192G) has greatly reduced toxicity, but retains the adjuvant properties of wild type LT.

6.1. PRODUCTION OF LT(R192G)

LT(R192G) can be produced by a number of means apparent to those of skill in the art. For example, LT(R192G) can be isolated from E. coli expressing pBD95, a plasmid fully described in PCT Publication WO96/06627.

Subsequent to the effective priority date of WO96/06627, others have had success in isolating LT(R192G) from E. coli expressing other plasmid constructs. (Grant et al . , (1994), Infection and Immuni ty 62_.(10) :4270-4278) . Plasmid pCS95,

- 16 - fully described in Example 6.1 can also be utilized to produce isolated or substantially pure LT(R192G) in E. coli .

LT(R192G) can be isolated by agarose affinity chromatography from bacteria expressing an LT(R192G) encoding plasmid. Alternate methods of purification will be apparent to those skilled in the art.

LT(R192G) produced by any means can be further purified by gel filtration chromatography, which allows for the separation of holotoxin from any free A or B subunits.

6.2. COMPOSITIONS OF LT(R192G)

The present invention provides mutant LT(R192G) useful for compositions and methods to promote the production of serum and/or mucosal antibodies as well as cell -mediated immune responses against an antigen that is simultaneously administered with a genetically modified bacterial toxin, i.e. , LT(R192G) .

Compositions of the present invention are useful for the formulation of adjuvant compositions through the mixing of a substantially pure preparation of LT(R192G), free LT-B subunit, and an antigen.

6.3. MODE OF ADMINISTRATION OF

LT(R192G) AND UNRELATED ANTIGENS

The product of this invention is useful to prepare formulations which can be administered as described below.

LT(R192G) in combination with B-subunit free of holotoxin at any B-subunit to LT(R192G) ratio of 1:1 or greater is administered in conjunction with any biologically relevant antigen and/or vaccine, such that an increased immune response to said antigen and/or vaccine is achieved. LT(R192G) plus free B-subunit and antigen are administered simultaneously in a pharmaceutical composition

- 17 comprising an effective amount of LT(R192G) plus free B-subunit and an effective amount of antigen.

The antigen, the LT(R192G), and the free B-subunit free of holotoxin are administered separately within a short ₅ time of each other.

The antigen is administered separately within a short time of the simultaneous administration of the LT(R192G) and the B-subunit free of holotoxin.

The LT(R192G) administered in combination with free

B-subunit is at a ratio of between 1:1 and 100:1 of B-subunit

10 to LT(R192G). LT(R192G) administered in combination with free B-subunit can be at a weight ratio of 2:1 to 10:1 of B-subunit to LT(R192G). The LT(R192G) administered in combination with free B-subunit can be at a weight ratio of about 3:1 of B-subunit to LT(R192G).

15 The mode of administration is mucosal (i.e., intranasal, oral, rectal) or parenteral (i . e . , subcutaneous, intramuscular, intradermal, intravenous, intraperitoneal) . The respective amounts of LT(R192G) plus free B-subunit and antigen will vary depending upon the identity of the route of

20 administration, antigen employed and the species of animal to be immunized. The initial administration of LT(R192G) plus free B-subunit and antigen can be followed by a boost of the relevant antigen. Alternatively, no boost is given. The timing of boosting may vary, depending on the route, antigen and the species being treated. The modifications in route,

____ O dosage range and timing of boosting for any given species and antigen are readily determinable by routine experimentation. The boost may be of antigen alone or in combination with LT(R192G) plus free B-subunit.

³⁰ 6.4. ANTIGENS

The methods and compositions of the present invention are intended for use both in immature and mature

18 vertebrates, in particular birds, mammals, and humans. Useful antigens, as examples and not by way of limitation, include antigens from pathogenic strains of bacteria ( Streptococcus pyogenes , Streptococcus pneumoniae, Neisseria gonorrhoea, Neisseria meningi tidis , Corynebacterium diphtheriae,

Clostridium botulinum, Clostridium perfringens, Clostridium tetani , Haemophilus influenzae, Klebsiella pneumoniae,

Klebsiella ozaenae, Klebsiella rhinoscleromotis,

Staphylococcus aureus , Vibrio cholerae, Escherichia coli ,

Pseudomonas aeruginosa, Campylobacter jejuni , Aeromonas hydrophila, Bacillus cereus, Edwardsiella tarda, Yersinia enterocoli tica, Yersinia pestis, Yersinia pseudotuberculosis ,

Shigella dysenteriae, Shigella flexneri , Shigella sonnei ,

Salmonella typhimurium, Treponema pallidum, Treponema pertenue, Treponema carateneum, Borrelia vincentii , Borrelia burgdorferi , Leptospira icterohemorrhagiae, Mycobacterium tuberculosis , Toxoplasma gondii , Pneumocystis carinii ,

Francisella tularensis , Brucella abortus, Brucella suis,

Brucella meli tensis , Mycoplasma spp . , Rickettsia prowazeki ,

Rickettsia tsutsugumushi , Chlamydia spp . , Helicobacter pylori) ; pathogenic fungi (Coccidioides immitis, Aspergillus fumigatus, Candida albicans, Blasto yces dermati tidis , Cryptococcus neoformans , Histoplasma capsulatum) ; protozoa (Entomoeba histolytica, Trichomonas tenas, Trichomonas hominis , Trichomonas vaginal is , Trypanosoma gambiense, Trypanosoma rhodesiense, Trypanosoma cruzi , Leishmania donovani , Leishmania tropica, Leishmania braziliensis , Pneumocystis pneumonia, Plasmodium vivax, Plasmodium falciparum, Plasmodium malaria) ; or Helminiths (Enterobius vermicular is , Trichuris trichiura, Ascaris lumbricoides , Trichinella spiralis, Strongyloides stercoralis, Schistosoma japonicum, Schistosoma mansoni , Schistosoma haematobium, and hookworms) either presented to the immune system in whole

19 cell form or in part isolated from media cultures designed to grow said organisms which are well know in the art, or protective antigens from said organisms obtained by genetic engineering techniques or by chemical synthesis. Other relevant antigens include pathogenic viruses

(as examples and not by limitation: Poxviridae, Herpesviridae, Herpes Simplex virus 1, Herpes Simplex virus 2, Adenoviridae, Papovaviridae, Enteroviridae, Picornaviridae, Parvoviridae, Reoviridae, Retroviridae, influenza viruses, parainfluenza viruses, mumps, measles, respiratory syncytial virus, rubella, Arboviridae,

Rhabdoviridae, Arenaviridae, Hepatitis A virus, Hepatitis B virus, Hepatitis C virus, Hepatitis E virus, Non-A/Non-B Hepatitis virus, Rhinoviridae, Coronaviridae, Rotoviridae, and Human Immunodeficiency Virus) either presented to the immune system in whole or in part isolated from media cultures designed to grow such viruses which are well known in the art or protective antigens therefrom obtained by genetic engineering techniques or by chemical synthesis.

Further examples of relevant antigens include, but are not limited to, vaccines. Examples of such vaccines include, but are not limited to, influenza vaccine, pertussis vaccine, diphtheria and tetanus toxoid combined with pertussis vaccine, hepatitis A vaccine, hepatitis B vaccine, hepatitis C vaccine, hepatitis E vaccine, Japanese encephalitis vaccine, herpes vaccine, measles vaccine, rubella vaccine, mumps vaccine, mixed vaccine of measles, mumps and rubella, papillomavirus vaccine, parvovirus vaccine, respiratory syncytial virus vaccine, Lyme disease vaccine, polio vaccine, varicella vaccine, gonorrhea vaccine, schistosomiasis vaccine, rotavirus vaccine, mycoplasma vaccine pneumococcal vaccine, meningococcal vaccine, campylobacter vaccine, helicobacter vaccine, cholera vaccine, enterotoxigenic E. coli vaccine, enterohemmorgagic E. coli

20 vaccine, shigella vaccine, salmonella vaccine and others.

These can be produced by known common processes. In general, such vaccines comprise either the entire organism or virus grown and isolated by techniques well known to the skilled _c artisan or comprise relevant antigens of these organisms or viruses which are produced by genetic engineering techniques or chemical synthesis. Their production is illustrated by, but not limited to, the following:

Influenza vaccine: a vaccine comprising the whole or part of hemagglutinin, neuraminidase, nucleoprotein and 0 matrix protein which are obtainable by purifying a virus, which is grown in embryonated eggs, with ether and detergent, or by genetic engineering techniques or chemical synthesis.

Pertussis vaccine: a vaccine comprising the whole or a part of pertussis toxin, hemagglutinin and K-agglutinin 5 which are obtained from avirulent toxin with formalin which is extracted by salting-out or ultracentrifugation from the culture broth or bacterial cells of Bordetella pertussis, or by genetic engineering techniques or chemical synthesis.

Diphtheria and tetanus toxoid combined with 0 pertussis vaccine: a vaccine mixed with pertussis vaccine, diphtheria and tetanus toxoid.

Japanese encephalitis vaccine: a vaccine comprising the whole or part of an antigenic protein which is obtained by culturing a virus intracerebrally in mice and purifying the virus particles by centrifugation or ethyl 5 alcohol and inactivating the same, or by genetic engineering techniques or chemical synthesis.

Hepatitis B vaccine: a vaccine comprising the whole or part of an antigen protein which is obtained by isolating and purifying the HBs antigen by salting-out or 0 ultracentrifugation, obtained from hepatitis carrying blood, or by genetic engineering techniques or by chemical synthesis.

- 21 Measles vaccine: a vaccine comprising the whole or part of a virus grown in a cultured chick embryo cells or embryonated egg, or a protective antigen obtained by genetic engineering or chemical synthesis. 5 Rubella vaccine: a vaccine comprising the whole or part of a virus grown in cultured chick embryo cells or embryonated egg, or a protective antigen obtained by genetic engineering techniques or chemical synthesis.

Mumps vaccine: a vaccine comprising the whole or part of a virus grown in cultured rabbit cells or embryonated 0 egg, or a protective antigen obtained by genetic engineering techniques or chemical synthesis.

Mixed vaccine of measles, rubella and mumps: a vaccine produced by mixing measles, rubella and mumps vaccines . 5 Rotavirus vaccine: a vaccine comprising the whole or part of a virus grown in cultured MA 104 cells or isolated from the patient's feces, or a protective antigen obtained by genetic engineering techniques or chemical synthesis.

Mycoplasma vaccine: a vaccine comprising the whole o or part of mycoplasma cells grown in a liquid culture medium for mycoplasma or a protective antigen obtained by genetic engineering techniques or chemical synthesis .

Those conditions for which effective prevention may be achieved by the present method will be obvious to the skilled artisan. 5

The vaccine preparation compositions of the present invention can be prepared by mixing the above illustrated antigens and/or vaccines with LT(R192G) at a desired ratio.

Pyrogens or allergens should naturally be removed as completely as possible. The antigen preparation of the 0 present invention can be used by preparing the antigen per se and the LT(R192G) separately or together.

- 22 - Further, the present invention encompasses a kit comprising an effective amount of antigen and an adjuvant effective amount of LT(R192G). In use, the components of the kit can either first be mixed together and then administered _c or the components can be administered separately within a short time of each other.

The vaccine preparation compositions of the present invention can be combined with either a liquid or solid pharmaceutical carrier, and the compositions can be in the form of tablets, capsules, powders, granules, suspensions or 0 solutions. The compositions can also contain suitable preservatives, coloring and flavoring agents, or agents that produce slow release. Potential carriers that can be used in the preparation of the pharmaceutical compositions of this invention include, but are not limited to, gelatin capsules, 5 sugars, cellulose derivations such as sodium carboxymethyl cellulose, gelatin, talc, magnesium stearate, vegetable oil such as peanut oil, etc., glycerin, sorbitol, agar and water. Carriers may also serve as a binder to facilitate tabletting of the compositions for convenient administration. 0

7. EXAMPLES

The following examples are presented for purposes of illustration only and are not intended to limit the scope of the invention in any way.

5

0

23 7.1. EXAMPLE: PRODUCTION OF LT(R192G)

The wild-type LT toxin is encoded on a naturally occurring plasmid found in strains of enterotoxigenic E. coli capable of producing this toxin. Clements et al . had 5 previously cloned the LT gene from a human isolate of E. coli designated H10407. This subclone consists of a 5.2 kb DNA fragment from the enterotoxin plasmid of H10407 inserted into the PstI site of plasmid pBR322 (Clements et al , 1983, Infect . Immun . 4_.0:653). This recombinant plasmid, designated pDF82, has been extensively characterized and expresses LT under control of the native LT promoter. From pDF82, Clements et al . derived plasmid pBD95, which is fully described in PCT Publication WO96/06627.

Figure 1A shows the construction of plasmid pCS95, which was constructed by inserting the native LT-A subunit

15 regulatory region upstream from the LT-A coding region of pBD95. Figure IB shows the Arg to Gly mutation at position

192. The BamHI and Xbal restriction sites referred to in the diagram as "new" were added by site directed mutagenesis, as described in PCT Publication WO96/06627. The new Xbal site

20 was added through a silent mutation, resulting in no alteration of the amino acid sequence of the peptide encoded by the gene .

LT(R192G) was then purified by agarose affinity chromatography from bacteria expressing pCS95. This mutant

__ LT, designated LT(R192G) was then examined by 25

SDS-polyacrylamide gel electrophoresis for modification of the trypsin sensitive bond. Samples were examined with and without exposure to trypsin and compared with native (unmodified) LT. LT(R192G) does not dissociate into A₁ and A₂ when incubated with trypsin, thereby indicating that

30 sensi.ti.vity to protease has been removed.

- 24 7.2. EXAMPLE: LT(R192G) AND FREE B-SUBUNIT

IN THE PATENT MOUSE ENTEROTOXICITY ASSAY

LT(R192G) with no free B-subunit was admixed with different ratios of B-subunit and examined for toxicity in 5 the patent mouse assay. As shown in Figure 2, the addition of excess free B-subunit to LT(R192G) results in a reduction of the gut/carcass ratio as measured in the patent mouse assay. To further evaluate this finding, free B-subunit was admixed with LT(R192G) and also with native LT at a ratio of 0 either 3:1 or 10:1. As shown in Figure 3, free B-subunit suppresses the toxicity of native LT.

7.3. EFFECT OF FREE B-SUBUNIT ON INTRANASAL ADJUVANTICITY

The effect of administration of LT(R192G) with 15 excess free B-subunit on both intranasal and oral adjuvanticity was examined by administration of an illustrative antigen with the adjuvant composition to different mucosal surfaces.

Ovalbumin (Ova) was selected as a representative antigen for these studies. A number of investigations, including our own (Clements et al . , 1988, Vaccine 6_.: 269-277;

Dickinson and Clements, 1996, Mucosal Vaccines -. 13 - 81 ;

Dickinson and Clements, 1995, Infection and Immuni ty 63:1617- 1623; Tsuji et al . , 1997, Immunology 90:176-182; Yamamoto et al., 1997, Journal of Experimental Medicine 185 : 1203-1210;

25

Yamamoto et al . , 1997, Proceedings of the National Academy of

Sciences _9_4: 5267-5272 ; DiTommaso et al . , 1996, Infection and Immuni ty 64:974-979 ; Douce et al . , 1995, Proceedings of the National Academy of Sciences S__ : 1644-1648 ; Douce et al . , 1997, Infection and Immuni ty 6_5: 2821-2828) , have used this ^Jυ protein and it provides a useful reference for comparison to other studies. For those studies, both serum anti -Ova IgG and Ova-specific T-cell responses were examined.

- 25 - A second antigen, Colonizing Factor Antigen I (CFAI) of enterotoxigenic E. coli was included in one set of experiments. In these studies, serum anti -CFAI was examined because anti-CFAI antibodies may be protective against diarrheal disease caused by there organisms.

In the first series of experiments, mice were immunized intranasally with Ova alone or in conjunction with

5 μg of LT(R192G) or 1.25 μg of LT(R192G) plus 3.75 μg of free B-subunit, designated 1AB5:3B5. Serum anti -Ova IgG was determined by ELISA. There were seven animals per group and the means for each data point are shown. As shown in Figure

4, mice immunized intranasally with Ova in conjunction with

LT(R192G) containing excess B-subunit had serum anti-Ova IgG responses indistinguishable from animals immunized with Ova in conjunction with LT(R192G) without excess of B-subunit, even though a significantly lower total amount of LT(R192G) was administered (1.25 μg vs. 5 μg) . This demonstrates that excess free B-subunit is able to enhance the adjuvanticity of LT(R192G). When the Ova-specific T-cell responses from these animals were examined, both the Thl/IFN-gamma (Figure 5) and Th2/IL-10 (Figure 6) anti-Ova responses were equivalent when free B-subunit was included in the adjuvant formulation compared to LT(R192G) without free B-subunit. There is no significant difference in the IFN-gamma and IL-10 responses between these two groups .

7.4. EFFECT OF FREE B-SUBUNIT ON ORAL ADJUVANTICITY

In the next series of experiments, mice were immunized orally with purified Colonizing Factor I (CFAI) from enterotoxigenic E. coli in conjunction with 6.25 μg of LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg of free B-subunit, designated 1AB5:3B5. Serum anti-CFAI IgG was determined by ELISA. There were seven animals per group and the means for each data point are shown. As shown in Figure

- 26 7, mice immunized orally with CFAI in conjunction with LT(R192G) containing excess B-subunit had serum anti-CFAI IgG responses significantly higher that did animals immunized with CFAI in conjunction with LT(R192G) without excess of c B-subunit. This demonstrates that excess B-subunit is able to alter the type of immune response elicited by oral administration of LT(R192G) with an antigen. One possibility was that the CFAI response was unique because of the inherent ability of colonizing factors to bind to epithelial cells.

To further elucidate this unexpected finding, a third series 0 of experiments was performed.

In this series of experiments, mice were immunized orally with Ova alone or in conjunction with 25 μg of LT(R192G) or 6.25 μg of LT(R192G) plus 18.75 μg of free B-subunit, designated 1AB5:3B5. Serum anti-Ova IgG was 5 determined by ELISA. There were ten animals per group and the means for each data point are shown. As shown in Figure

8, mice immunized orally with Ova in conjunction with LT(R192G) containing excess B-subunit had significantly higher serum anti-Ova IgG responses than did animals 0 immunized with Ova in conjunction with LT(R192G) without excess of B-subunit, even though a significantly lower total amount of LT(R192G) was administered (6.25 μg vs. 25 μg) .

This finding was consistent with the increased anti-CFAI serum IgG responses observed when excess B-subunit was included in the oral adjuvant formulation with CFAI as the 5 antigen (Figure 7) . When the Ova-specific T-cell responses from these animals were examined, both the Thl/IFN-gamma

(Figure 9) and Th2/IL-10 (Figure 10) anti-Ova responses were qualitatively different and quantitatively enhanced when free

B-subunit was included in the adjuvant formulation compared 0 to LT(R192G) without free B-subunit.

These findings demonstrate that the presence of free B-subunit elicits an antigen-specific T-cell response, a

27 response which is substantially non-existent when LT(R192G) is used without excess free B-subunit, when LT(R192G) is used as an oral adjuvant. This represents a qualitative change in the type of immune response elicited. These findings also _c illustrate that excess free B-subunit enhances the humoral immune response elicited by LT(R192G) when used as an oral adjuvant; this represents a quantitive change in response.

8. DEPOSIT OF MICROORGANISMS

The following microorganism containing the designated plasmid was deposited with the American Type Culture Collection (ATCC) , (present address: 1081 University Boulevard, Manassas, VA 20110-2209) on March 13, 1998, and has been assigned the indicated accession number:

5 Microorganism Accession Number

E. coli JM83 (pCS95) 98696

Although the invention is described in detail with reference to specific embodiments thereof, it will be 0 understood that variations which are functionally equivalent are within the scope of this invention. Indeed, various modifications of the invention in addition to those shown and described herein will become apparent to those skilled in the art from the foregoing description and accompanying drawings.

Such modifications are intended to fall within the scope of 5 the appended claims.

Various publications are cited herein, the disclosures of which are incorporated by reference in their entireties .

0

28 MICROORGANISMS

Optional Sheet in connection with the microorganism referred to on page 28 , lines 15-16 of the description

A. IDENTIFICATION OF DEPOSIT '

Further deposits are identified on an additional sheet

Name of depositary institution ' American Type Culture Collection

Address of depositary institution (including postal code and country)

10801 University Blvd. anassas, VA 20110-2209 US

Date of deposit * March 13, 1998 Accession Number ' 98696

B. ADDITIONAL INDICATIONS ' (leave blank if not applicable) This information is continued on a separate attached sheet

C. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE ' <_α»

D. SEPARATE FURNISHING OF INDICATIONS ' (leave blank if not applicable)

The indications listed below will be submitted to the International Bureau leter ' (Specify the general nature of the indications e g , "Accession Number of Deposit")

_Aτh uιs sheet was received with the International application when filed (to be checked by the receiving Office)

X l \Cλ(X(7^UXj t 1XA

(Authorized Officer)

D The date of receipt (from the applicant) by the International Bureau '

(Authorized Officer)

Form PCT/RO/134 (January 1981 )

28.1

Claims

WHAT IS CLAIMED IS;

1. A composition comprising an isolated mutant E. coli heat-labile enterotoxin holotoxin, in which arginine at amino acid position 192 is replaced with glycine, which holotoxin has immunologic adjuvant activity and is:

(a) substantially less toxic than native E. coli heat- labile enterotoxin (LT) holotoxin as measured in the Y-1 adrenal cell assay, and

(b) substantially free from excess free B-subunit of

LT.

2. The composition of claim 1 in which the holotoxin is encoded by plasmid pCS95 having ATCC Accession number XXX, which expresses both subunit A and subunit B of the E. coli heat-labile enterotoxin substantially free from excess B-subunit.

3. A composition comprising an isolated mutant E. coli heat-labile enterotoxin holotoxin, in which arginine at amino acid position 192 is replaced with glycine, which holotoxin has immunologic adjuvant activity and:

(a) is substantially less toxic than native E . coli heat-labile enterotoxin (LT) holotoxin as measured in the Y-l adrenal cell assay, and

(b) contains less than 10% free B-subunit of LT.

4. A method for obtaining mutant E. coli heat- labile enterotoxin holotoxin, in which arginine at amino acid position 192 is replaced with glycine, which holotoxin has immunologic adjuvant activity and is:

(a) substantially less toxic than native E. coli heat- labile enterotoxin (LT) holotoxin as measured in the Y-l adrenal cell assay, and

29 (b) substantially free from excess free B-subunit of LT, by expressing plasmid pCS95 in E. coli and harvesting the mutant LT.

- 30