Antibody response to the lipopolysaccharide and protein antigens of Salmonella typhi during typhoid infection. I. Measurement of serum antibodies by radioimmunoassay.

Serum antibody responses to the lipopolysaccharide and protein antigens of S. typhi in typhoid patients were studied using a solid-phase radioimmunoassay technique. Sera from 24 adult typhoid patients and 20 non-typhoid adult controls were compared. As a group, sera from typhoid patients showed increased IgA, IgG and IgM immunoglobulin levels and gave significantly higher anti-LPS and anti-protein antibody titres in all three major immunoglobulin classes than did non-typhoid controls. Levels of antibodies against LPS or protein in sera of typhoid patients were highly variable with a skew distribution. A good correlation was found between antibody titres to the LPS antigen and those to a protein antigen. No correlation, however, was found between the anti-LPS antibody titres measured by radioimmunoassay and the anti-O antibody titres measured by the Widal agglutination test. Titration of anti-LPS or anti-protein antibodies by radioimmunoassay was found to be more sensitive and specific than Widal test for the serological diagnosis of typhoid fever. The advantages of measuring antibody response by radioimmunoassay over conventional Widal test are discussed.     

Antibodies to porin antigens of Salmonella typhi induced during typhoid infection in humans.

To determine whether lipopolysaccharide (LPS) structures of Campylobacter fetus are related to the three known heat-stable serogroups, proteinase K-treated whole cell lysates obtained from strains of each serogroup were electrophoresed in polyacrylamide gels. All strains had smooth-type LPS with multiple high-molecular-weight repeating units. The profiles of serogroup A from C. fetus subsp. fetus and from C. fetus subsp. venerealis were identical, but they were different from those of C. fetus subsp. fetus serogroups B and AB. When we immunoblotted the LPS of these serogroups with normal or immune rabbit serum we found homologous recognition between serogroups A from C. fetus subsp. fetus and C. fetus subsp. venerealis. Similarly, serogroups AB and B from C. fetus subsp. fetus showed homologous recognition. However, antiserum against serogroup A did not recognize serogroups B and AB and vice versa. Absorption studies confirmed the identity of LPS from all serogroup A C. fetus strains and cross-reactivity of the serogroup B and AB strains with one another. Serogroup A strains were resistant to the bactericidal activity in normal human serum, whereas serogroup B and AB strains generally were susceptible; isolates from humans predominantly belonged to serogroup A. Results of these studies suggest that the LPS composition forms the basis for the heat-stable serotyping system for C. fetus and that the structural and antigenic variants are associated with differential serum susceptibility.     

Determination of antibody from typhoid patients against lipopolysaccharide and protein antigens of Salmonella typhi

Although the Widal test is simple, inexpensive and the most widely used for serodiagnosis of typhoid fever, the sensitivity and specificity of the test is sometimes doubtful. In this study, an enzyme-linked immunosorbent assay (ELISA) was developed for the detection of serum IgG and IgM antibodies to protein and lipopolysaccharide (LPS) antigens of Salmonella typhi which was compared with the Widal test in various groups of subjects. In typhoid patients with hemocultures positive for S. typhi (TP group), ELISA positivity was found on 100% for IgG antiprotein, 94.44% for IgG anti-LPS and 88.89% for IgM to both the protein and LPS antigens. In contrast, the Widal test was positive in only 61.11% for anti-O and 83.33% for anti-H antibodies. In healthy control subjects (HC group), only 5% of serum samples were positive for IgG anti-protein and none was positive for IgG anti-LPS or IgM to either the protein or LPS. In contrast, the Widal test was positive in 7.5% of HC group for anti-O and 17.5% for anti-H antibodies. In blood bank donors (BB group), both ELISA and Widal tests were positive in 23-40% of sera. Since the hospital records of BB group were incomplete. It might be possible that some of these subjects had recently been infected with S. typhi. Our data indicate that the standard Widal test was associated with false negative reactions in 16-39% of blood culture positive subjects.(ABSTRACT TRUNCATED AT 250 WORDS)     

Monoclonal antibodies delineate multiple epitopes on the O antigens of Salmonella typhi lipopolysaccharide.

Fifteen monoclonal antibodies (MAbs) directed against Salmonella typhi were produced and characterized. The specificities of the antibodies were determined by their binding patterns in an enzyme immunoassay, with a panel of lipopolysaccharides isolated from different bacteria. Seven MAbs reacted with S. typhi, Salmonella enteritidis, and Salmonella dublin (all belonging to serogroup D). One MAb also reacted with Salmonella paratyphi A and S. paratyphi B. Five MAbs reacted with S. typhi, S. enteritidis, S. dublin, and S. paratyphi B. Two MAbs did not bind to any lipopolysaccharide but showed reactivity with bacterial sonic extracts isolated from S. typhi, S. paratyphi A, S. paratyphi B, Escherichia coli, and Shigella sonnei. These antibodies would be helpful in studying the complexity of antigenic determinants expressed by S. typhi and the nature of the antibody response during typhoid and paratyphoid fevers and also in the diagnosis of the disease.     

Immune response to the iron-deprivation-induced proteins of Salmonella typhi in typhoid fever.

Iron starvation conditions limited the growth of Salmonella typhi, as evidenced by an increase in the lag phase of a culture and a decrease in the number of bacteria reached in the stationary phase. The analysis of the outer membrane of bacteria grown under these conditions identified new protein components with apparent molecular weights of 83,000, 78,000, and 69,000. The extent of induction of these proteins was regulated by increased iron deprivation. Immunoblot analysis showed that the serum of patients with typhoid fever exhibited an immunoglobulin G response to these iron-deprivation-induced proteins. The results of bioassays and DNA-DNA hybridization experiments indicated that pathogenic strains of S. typhi produced enterochelin but not aerobactin. Immunodetection with an anti-FepA antiserum confirmed that one of the induced proteins is the S. typhi analog of the Escherichia coli fepA gene product. These studies suggest a role for iron uptake in the pathogenesis of typhoid fever and confirm the immunogenicity of some of the outer membrane proteins of this pathogen.     

Antibody response to antigens distinct from smooth lipopolysaccharide complex in Brucella infection.

The smooth lipopolysaccharide complex of the outer surface of smooth Brucella abortus cells is believed to be the antigenic component involved in serological tests routinely used for the diagnosis of brucellosis. Sera from cattle vaccinated or infected with B. abortus generally contain antibody directed toward the smooth lipopolysaccharide complex. The brucella organism contains a large number of other antigenically distinct components. The biological significance of some of these antigens has been demonstrated by showing that sera from infected cattle have precipitins to these components. These sera revealed up to seven distinct lines in immunoelectrophoresis with a protein-rich antigen mixture prepared from rough strain B. abortus 45/20, whereas sera from strain 19-vaccinated cattle did not reveal these lines at 4 or more months after vaccination. Monospecific antisera were prepared against six antigens in this mixture, and the purification of two of them by antibody affinity chromatography is described.     

Monoclonal antibodies to 52-kilodalton protein of Salmonella typhi.

Ten monoclonal antibodies (6 immunoglobulin G1 kappa [IgG1 kappa] and 4 IgG2b kappa) from six hybrid clones specific for Salmonella typhi antigen were produced by immunizing BALB/cJ mice with affinity-purified S. typhi proteins (Bp). The latter were prepared by passing crude S. typhi Bp through an affinity column made from Sepharose conjugated to IgG antibodies against partially purified S. typhi Bp. The eluent was subsequently used as the immunogen for the production of monoclonal antibody. All 10 monoclonal antibodies reacted specifically with a 52-kilodalton (kDa) protein of S. typhi and were species specific. The presence of IgM antibody to the 52-kDa antigen in the sera of a majority of patients with acute typhoid fever suggested that this 52-kDa protein is also a good immunogen for humans. The potential usefulness of this antigen in the early diagnosis of typhoid fever is discussed.     

Diagnosis of typhoid fever: detection of Salmonella typhi porins-specific antibodies by inhibition ELISA.

Porins are highly immunogenic outer membrane proteins of Salmonella. Sera from typhoid patients contained a high level of IgG antibodies directed to porins of Salm. typhi. Since porins are highly conserved proteins, anti-porins antibodies both from typhoid patients and healthy normals reacted with porins from several Gram-negative bacteria. Therefore, in order to improve the specificity of detecting Salm. typhi porins-specific antibodies, an inhibition ELISA was developed using enzyme-conjugated MoAbs (MP1 and MPN4) specific to Salm. typhi porins. Sera from typhoid patients with positive haemoculture (16 out of 17) inhibited the binding of MP1 to porins, thus showing a positive test for typhoid, whereas sera from patients with other Gram-negative bacterial infections (n = 7) and from healthy volunteers (66 out of 67) were found to be negative. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of this assay were 94.1, 98.7, 97.8, 94.1 and 98.7% respectively. The validity of our inhibition ELISA for typhoid was higher than that of the Widal test. The diagnosis of typhoid fever as early as 3 days after the onset of fever, using a single specimen is possible.     

Measurement of antibodies to tubulin by radioimmunoassay.

A solid-phase double antibody radioimmunoassay capable of measuring antibody to tubulin, the principal component of microtubules, is described. This assay is simple, combining sensitivity with specificity and also allowing determination of antibody subclasses.     

Antibody response to Campylobacter coli in children during intestinal infection and carriage.

In the Central African Republic, etiological studies of diarrhea have shown that Campylobacter coli accounts for almost 40% of Campylobacter enteric isolations. This prompted us to investigate the antibody response to C. coli infection in children. As expected from the literature on Campylobacter jejuni infections, our results show that both infection and carriage elicited antibodies against glycine-extracted membrane antigens, flagella, and cholera toxin. The human antibody response to C. coli resembles the response to C. jejuni, and this similarity will allow comparative studies on larger numbers of infections, both symptomatic and asymptomatic. Anti-cholera toxin antibodies were directed against both the A and B subunits.     

Characterization of monoclonal antibodies to protein antigen of Salmonella typhi

Two monoclonal antibodies were produced against protein antigens of Salmonella typhi. One of the antibodies (STP14) belongs to the immunoglobulin G1K subclass, and the other (STP13) was assigned to the immunoglobulin G2a(kappa) subclass. Both antibodies could recognize the 34.0-kilodalton protein antigen from S. typhi. The specificity of these antibodies was tested by immunoblotting with a panel of crude protein antigens from 12 bacteria causing enteric fever and enteric fever-like illness: S. typhi, S. paratyphi A, S. paratyphi B, S. paratyphi C, S. choleraesuis, S. enteritidis, S. krefeld, S. panama, S. typhimurium, Escherichia coli, Pseudomonas pseudomallei, and Yersinia enterocolitica. In a modified double-antibody sandwich enzyme-linked immunosorbent assay they could detect the protein antigen at ca. 0.6 microgram/ml. These monoclonal antibodies should be of great value in the diagnostic test for detecting S. typhi antigen in samples of bodily fluids isolated from patients with typhoid fever and in studies of the chemical structure and other immunological properties of this 34.0-kilodalton protein.     

Measurement of intestinal antibody by radioimmunoassay.

The study of antibody responses in the intestine has been greatly hampered by lack of reproducible sensitive assays. An assay for measuring antibody against bacteria capable of regularly detecting gut antibody in gastroenteritis is described. It is based on absorption of antibody onto bacteria and measurement of the amount of antibody bound using radiolabelled anti-immunoglobulin antibody. Anti-light chain antibody is used to detect all classes of antibody as well as partially degraded antibody which retains the capacity to bind; anti-alpha and anti-gamma antibody is used to measure IgA and IgG antibody. The sensitivity of the assay depends on the use of anti-immunoglobulin antibody purified by affinity chromatography and allows measurement of nanogram amounts of antibody. Its specificity and kinetics are described and the particular advantages it provides in the measurement of antibacterial antibody in the intestine are discussed.     

Development of antibodies to meningococcal protein and lipopolysaccharide serotype antigens in healthy-carriers.

The nasopharyngeal acquisition of meningococci was followed in healthy military recruits during their primary training. The production of antibodies to meningococcal serotype protein antigens and serotype lipopolysaccharide antigens accompanied the carrier state. Bactericidal antibody to protein serotype 2 was generated in response to the carriage of meningococci of low virulence that carried this antigen.     

Diagnosis of typhoid fever by detection of Salmonella typhi antigen in urine.

A monoclonal antibody specific for group D Salmonella antigen 9 was used in an indirect enzyme-linked immunosorbent assay (ELISA) for detecting the antigen in urine specimens collected from patients with clinical typhoid fever in Jakarta, Indonesia. The ELISA had a sensitivity of 95% in identifying patients in whom Salmonella typhi was isolated from hemocultures, 73% in patients in whom S. typhi was isolated from stool specimens, and 40% in patients in whom the organism was isolated from bone marrow cultures. Among patients in whom S. typhi was isolated from blood cultures, the ELISA had a sensitivity of 65% when a single urine specimen was examined and 95% when serially collected urine specimens were examined. A dot blot immunoassay performed on a nitrocellulose filter in parallel had a sensitivity of 85%, versus 83% for the plate ELISA in which S. typhi was isolated from blood, bone marrow, and/or stool specimens. Since S. typhi antigen is intermittently excreted in the urine of patients with typhoid fever, serially collected urine from patients with typhoid should be tested for antigen 9.     

Duodenal isolation of Salmonella typhi by string capsule in acute typhoid fever.

Three of seven volunteers with acute typhoid fever had Salmonella typhi isolated from the duodenum using a string capsule device. The string capsule device provides a simple method for culturing S. typhi from the duodenal contents. Its possible use in typhoid carriers is discussed.     

Effect of parenteral immunization on the intestinal immune response to Salmonella typhi Ty21a.

The effects of parenteral administration of a killed typhoid vaccine on the intestinal immune response to live orally administered Salmonella typhi Ty21a in human subjects was evaluated. Priming with parenteral vaccination neither enhanced nor suppressed the subsequent specific serum and intestinal immunoglobulin A (IgA) immune responses to a booster course of live oral vaccine. Neither a single oral dose of live vaccine nor a single dose of parenteral vaccine had any measurable booster effect on the observed primary intestinal IgA response to the live oral vaccine. Two booster doses of subcutaneously administered killed typhoid vaccine did result in a significant increase in the specific intestinal IgA antibody in those subjects primed with the oral live vaccine. This response was comparable in magnitude to the primary intestinal response. No evidence of this response could be found in serum IgA, although nonsignificant rises in serum IgG were evident. Previous parenteral priming had no effect on secondary immune responses to a live oral vaccine in humans. Serum immune responses were generally found to be of little value as indicators of local intestinal immunity. This study confirmed that parenteral vaccination was only able to induce an intestinal immune response following priming with live, orally administered organisms and that multiple parenteral booster doses were necessary to induce a measurable effect on intestinal immune responses.     

Innate resistance of mice to Salmonella typhi infection.

The basis for the natural resistance of mice to Salmonella typhi was examined. In contrast to Salmonella typhimurium, the virulence of S. typhi for mice was independent of the mouse strain and was not affected by inactivation of murine macrophages with silica. However, mice were more susceptible to S. typhi when given iron alone or iron and an iron chelator. The results suggest that the failure of S. typhi to undergo net growth in murine tissues reflects an inability of the bacterium to multiply rather than rapid killing by resident macrophages.     

Characterization of human antibodies to Salmonella typhi by gel-filtration and antigenic analysis

A normal adult human male was immunized with a heat-killed typhoid vaccine. The appearance of antibodies to `H'' and `O'' antigens was subsequently studied. Antibodies were identified as belonging to specific immunoglobulin classes by gel-filtration on G-200 `Sephadex'', followed by specific immunochemical identification using a two-stage agglutination test. Serum antibodies to `H'' antigen were initially of γ_1M class; later the 7S γ class became predominant. γ_1A antibodies were transient in appearance. Serum antibodies to `O'' antigen were of γ<sub>1M</sub> and γ<sub>1A</sub> classes. Urinary antibodies to `H'' antigen were of γ_1A and 7S γ classes. No antibody activity could be detected in the low molecular weight γ-globulins of urine.