Lipopolysaccharide from Brucella abortus Behaves as a T-Cell ...

2 downloads 0 Views 1MB Size Report
abortus and LPS-BA induce IFN--y secretion from T helper cells. T helper cells have been classified as TH1 and TH2 on the basis of the cytokines they secrete.
Vol. 61, No. 5

INFECTION

AND IMMUNITY, May 1993, p. 1722-1729 0019-9567/93/051722-08$02.00/0

Lipopolysaccharide from Brucella abortus Behaves T-Cell-Independent Type 1 Carrier in Murine Antigen-Specific Antibody Responses

as a

MICHAEL BETTS,' PAUL BEINING,1 MARK BRUNSWICK,' JOHN INMAN,2 R. DALE ANGUS,3 THOMAS HOFFMAN,' AND BASIL GOLDING'*

Laboratory of Cell Biology, Division of Hematology, Center for Biologics Evaluation and Research, U.S. Food and Drug Administration, 1 and Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health,2 Bethesda, Maryland 20892, and National Veterinary Services Laboratories, Science and Technology, Animal and Plant Health Inspection Service, U. S. Department of Agriculture, Ames, Iowa 500103 Received 18 June 1992/Accepted 11 January 1993

In order to determine the carrier nature of lipopolysaccharide from Brucella abortus (LPS-BA) in evoking humoral responses, normal and immunodeficient mice were immunized with trinitrophenyl (TNP)-conjugated LPS-BA (TNP-LPS-BA) and the responses were compared with those to known T-dependent and T-independent antigens. TNP-LPS-BA, like T-independent type 1 (TI-1) antigens such as TNP-BA and TNP-LPS from Escherichia coli (TNP-LPS-EC), generated anti-TNP responses in BALB/c, athymic BALB/c nulnu, and CBAIN mice. In contrast, N-2,4-dinitrophenyl-j8-alanylglycylglycyl-substituted keyhole limpet hemocyanin, a typical T-dependent antigen, was not immunogenic in athymic mice, and TNP-Ficoll (T-independent type 2) was ineffective in eliciting humoral responses in CBA/N mice. These results indicate that LPS from B. abortus acts as a TI-i carrier in generating antibody responses. In C3H/HeJ mice, TNP-LPS-BA generated higher-titer immunoglobulin Gi (IgGI), IgG2a, and IgG2b anti-TNP antibodies than TNP-LPS-EC. Compared with those from BALB/c mice, pure resting B cells isolated from C3H/HeJ mice exhibited a 30-fold lower proliferative response to LPS-EC, whereas the LPS-BA response was reduced to a lesser extent (5-fold). This suggests that the disparity observed in antibody titers was due to different abilities of LPS from B. abortus and E. coli to stimulate C3H/HeJ B cells. The ability of LPS from B. abortus to act as a carrier in generating humoral immune responses indicates that LPS-BA can be substituted for whole B. abortus organisms in vaccine development. It has been shown previously that the trinitrophenyl hapten (TNP) conjugated to Brucella abortus (TNP-BA) induces anti-TNP antibody responses from human B cells (6). The human B-cell response to TNP-BA has been characterized as a T-cell-independent type 1 (TI-1) response because it occurred following T-cell depletion (6) and could be elicited with B cells from Wiscott-Aldrich syndrome patients, who have an X-linked immunodeficiency (8). In normal mice, TI-I antigens, such as TNP-BA and TNPlipopolysaccharide (LPS) from Eschenchia coli (TNP-LPSEC), demonstrate an antibody isotype pattern characterized by high-titer anti-TNP antibodies in all immunoglobulin G (IgG) subclasses. In contrast, T-cell-dependent (TD) antigens induce low titers of IgG3, and T-cell-independent type 2 (TI-2) antigens may induce IgG3 but usually elicit minimal responses in the remaining IgG subclasses (28). Repeated injections with B. abortus result in a predominant polyclonal

polysaccharides such as TNP-Ficoll are TI-2 antigens. Typically, TI-1 and TI-2 antigens are able to induce IgM antibody production in athymic nude mice, whereas a TD antigen, e.g., DNP-KLH, will induce little or no antibody response (28). In CBA/N mice, it is possible to differentiate TI-1 from TI-2 antigens, because these mice have a genetic defect which prevents them from responding to polysaccharide (TI-2) antigens (19, 23, 28). In contrast, TI-1 antigens are able to induce normal isotype responses in these mice, except that IgG3 antibody levels are low or absent. Although T-independent antigens can stimulate B cells and induce antibody responses in the relative absence of T cells (6, 28), they probably require T-cell factors, such as gamma interferon (IFN--y) and interleukin-4 (IL-4), for optimal responses and isotype switching (3, 18). Unlike TD antigens, TI antigens probably do not require cognate (or major histocompatibility complex-restricted) interactions between T and B cells. Not only are T-cell factors required for optimal responses to TI-1 antigens, but evidence from our laboratory (1, 5; also unpublished data) suggests that B. abortus and LPS-BA induce IFN--y secretion from T helper cells. T helper cells have been classified as TH1 and TH2 on the basis of the cytokines they secrete. TH1 cells secrete IFN--y and IL-2, whereas TH2 cells secrete IL-4, IL-5, IL-6, and IL-10 (5). The ability of B. abortus and LPS-BA to induce IFN-y secretion but not IL-4 from T helper cells suggests that they stimulate TH1 cells preferentially. The use of TI-1 antigens as carriers in vaccines could be very useful for patients with deficient T-cell function, such as occurs following human immunodeficiency virus (HIV)

IgG2a response (3). By using mouse strains with various genetic defects, such as athymic nude (BALB/c nu/nu) and CBA/N mice, it is possible to determine whether a particular antigen functions as a TD or TI (type 1 or 2) carrier. Haptenated proteins, such as N-2,4-dinitrophenyl- -alanylglycylglycyl-substituted keyhole limpet hemocyanin (DNP-KLH), function as TD antigens. Examples of prototype TI-1 antigens include TNP-LPS from members of the family Enterobacteriaceae (E. coli and Salmonella typhimurium) and inactivated TNP-BA, whereas

*

Corresponding author. 1722

B. ABORTUS LPS ACTS AS A TI-1 CARRIER

VOL. 61, 1993

infection. During HIV infection, there is a loss of function and then a steady decrease in the number of circulating CD4+ T helper cells (15). A T-independent vaccine should be more effective in boosting antibody responses in HIVinfected persons. In order to test this hypothesis, whole inactivated HIV was conjugated to B. abortus (HIV-BA) and tested for immunogenicity in normal and CD4+ T-celldepleted mice (7). This study demonstrated that HIV-BA was more immunogenic and elicited higher antibody titers than unconjugated HIV. Furthermore, this response was characterized by a high-titer IgG2a antibody response. In anti-L3T4-treated mice (CD4+ T cell depleted), HIV-BA but not HIV alone was able to induce anti-HIV antibodies, mostly of the IgG2a subclass. These sera could neutralize HIV type 1 (HIV-1) envelope-mediated syncytium formation. In order to enhance the feasibility of using B. abortus as a potential carrier in vaccines, we extracted the LPS from B. abortus (LPS-BA) and conjugated it to the trinitrophenyl hapten (TNP-LPS-BA). The potential use of LPS from B. abornus in future vaccines is supported by recent toxicology studies performed in our laboratory (10). These studies demonstrated that LPS-BA is 300 to 10,000 times less potent than LPS-EC in a variety of assays, including the rabbit pyrogen test, mortality of D-galactosamine-sensitized mice, and IL-13 and tumor necrosis factor alpha induction from human monocytes. Although LPS-BA has been shown previously to induce anti-LPS antibody responses (20, 21), it has not been used as a carrier for other determinants. In order to determine whether TNP-LPS-BA behaves as a TI-1 antigen, BALB/c, athymic, and CBA/N mice were studied. In addition, responses in C3HIHeJ mice were assessed to determine whether TNP-LPS-BA could stimulate antibody production in mice known to have defective responses to LPS-EC (31). These studies have implications for further development of LPS-BA as a component of vaccines or immunotherapy protocols, including those for the prevention or treatment of HIV-1 infections. MATERIALS AND METHODS Immunogens. LPS-BA was purified by the n-butanol water extraction procedure (26) and treated with proteinase K to remove proteins. The product was boiled for 1 h on 3 consecutive days to ensure sterility. The ketodeoxyoctonate content was found to be 0.82% by ihe method of Osborn (24). The nucleic acid content was 2.94%, as determined by UV absorption at 260 nm. Protein contamination was assessed with the protein assay kit from Bio-Rad (Richmond, Calif.) and found to be 9% by weight. Two additional batches of LPS-BA were prepared and contained