Humoral immune response to Q fever: enzyme-linked immunosorbent assay antibody response to Coxiella burnetii in experimentally infected guinea pigs.

The response of guinea pigs experimentally infected with Coxiella burnetii organisms, the etiologic agents of Q fever, was obtained by the measurement of fever, circulating infectious C. burnetii cells, and anti-C. burnetii antibodies. The detection of antibodies by the enzyme-linked immunosorbent assay (ELISA) and traditional methods against phase I whole cells, phase II whole cells, and phase I lipopolysaccharide (LPS-I) (a virulence marker for phase I cells) antigens in the serum samples of infected animals revealed marked differences between intrastrain phase variants. Animals infected with the phase I Nine Mile strain produced a concomitant increase in temperature, circulating infectious C. burnetii cells, and antibodies against phase II cells, phase I cells, and LPS-I. At 15 weeks, a challenge of phase I-infected animals with viable phase I cells resulted in anamnestic antibody responses to phase I cells and LPS-I but not to phase II cells. Infection of animals with the phase II Nine Mile strain produced antibodies against only phase II cells. The challenge of phase II-infected animals at 15 weeks with viable phase II cells resulted in anamnestic antibody responses to phase I and phase II cells but not to LPS-I. Suppression of anti-phase II responses by the phase I challenge was apparent with only the ELISA, because the immunofluorescence, microagglutination, and complement fixation assays were insensitive to these changes. The sensitivity and specificity of the ELISA with whole-cell and the LPS-I antigens in the detection of phase-specific antibody revealed that avirulent phase II cells induced an immune response to phase I antigenic epitopes. Although the avirulent phase II cells were rapidly cleared by the host immune responses, they were sufficiently infective to induce antibody responses to both phase variants. Thus, in the occurrence of Q fever, any conventional serological technique that uses only phase II antigens may not provide a true incidence of naturally acquired infection with both phase I and II C. burnetii organisms.     

Monoclonal antibodies distinguish phase variants of Coxiella burnetii

Monoclonal antibodies (MAbs) directed against phase I and II variants of Coxiella burnetii were produced by fusing myeloma SP2/O-AG 14 cells with spleen cells from mice immunized with the chloroform-methanol extraction residue of phase I whole cells. Two hybridoma clones which distinguished the phase variants by microimmunofluorescence assay were isolated and characterized. The MAbs showing specificity for phase I cells (MAbI-1, immunoglobulin G, subclass 3 kappa) reacted with the hot phenol-water extract of phase I C. burnetii in immunodiffusion and enzyme-linked immunosorbent assays. MAbI-1 reacted with high-molecular-weight components from phase I phenol-water extract and whole cell in an immunoblot assay. Specificity of MAbI-1 for a carbohydrate epitope in the phenol-water extract was demonstrated by periodic acid inactivation of binding by a competitive enzyme-linked immunosorbent assay. Phase I antigenic sites were apparently well represented on the surface of cells as demonstrated by complete fluorescence and microagglutination. The MAb showing specificity for phase II cells (MAbII-1, immunoglobulin G, subclass 2b kappa) reacted with whole cells in the microimmunofluorescence assay, microagglutination test, complement fixation test, and the enzyme-linked immunosorbent assay. MAbII-1 reacted specifically with a 29,500-dalton surface protein as demonstrated by immunoprecipitation of 125I-surface-labeled cells. Although MAbII-1 reacted with detergent-solubilized protein, it did not react with sodium-dodecyl sulfate-denatured protein by immunoblot assay. This protein was not exposed on the surface of phase I cells, but chloroform-methanol extraction of phase I cells exposed the phase II epitope.     

Isolation and evaluation of Coxiella burnetii O-polysaccharide antigen as an immunodiagnostic reagent.

A lipid A - deprived lipopolysaccharide (LPS) from Coxiella burnetii (C.b.) Priscilla strain in virulent phase I was separated by steric-exclusion chromatography and high performance liquid chromatography (HPLC). The isolated O-specific polysaccharide (PS) fractions were analyzed by different physico-chemical methods and showed noticeable differences in their overall composition. The antigenic potential of the PS fractions was evaluated by ELISA with animal and human sera and in comparison with those of the native C.b. LPSs and phase I and II C.b. Nine Mile stain cells of which the latter are routinely used in diagnosis of Q fever. The results indicate that the high molecular mass PS antigen SG501 could be used in ELISA for a sensitive and specific detection of anti-C.b. antibodies in the examined sera.     

Search for correlates of resistance to virulent challenge in mice immunized with Coxiella burnetii.

Mice immunized with live phase I or phase II Coxiella burnetii, with killed phase I or phase II organisms or with trichloroacetic acid (TCAE) or phenol (PE) extracts were resistant to intraperitoneal infection with phase I C. burnetii irrespective of whether or not they displayed phase I antibody response at the time of virulent challenge. Increased phagocytosis of purified phase I organisms by blood leukocytes or peritoneal exudate cells (PEC) was noticed only in mice with phase I agglutinating antibodies in their sera or peritoneal washings. Passive transfer of resistance was made possible only by sera containing phase I agglutinating antibodies. Adoptive transfer of immunity by spleen cells, but not by PEC, was achieved providing that these cells were taken from mice immunized with live phase I C. burnetii.     

Immunoglobulin responses in acute Q fever

Knowledge of the development of different classes of antibody during the course of acute Q fever is important to the clinician for interpreting a patient's serological test results. In the present study, the appearance of antibodies to Coxiella burnetii phases I and II was determined for a period of 1 year. A total of 683 sera from 191 patients with symptomatic Q fever were evaluated by the complement fixation and indirect immunofluorescence (immunoglobulins M and G [IgM, IgG]) tests. These patients had contracted acute Q fever in the fall of 1983 during an epidemic that resulted in 415 serologically confirmed cases of Q fever. As demonstrated by the complement fixation test, antibodies to C. burnetii phase II remained elevated throughout the entire study period, whereas antibodies to phase I were barely detectable. Although the immunofluorescence test was more sensitive than the complement fixation test, the specific anti-IgG response to C. burnetii to phases I and II gave the same general antibody profiles as did the complement fixation test. IgM anti-phase I and II titers appeared earlier but disappeared after 10 to 12 weeks. During this period, anti-phase II antibody levels were generally much higher than anti-phase I antibody levels.     

Steric hindrance of antibody binding to surface proteins of Coxiella burnetti by phase I lipopolysaccharide.

The exposure of surface protein antigens on virulent phase I Coxiella burnetti was compared with that on avirulent phase II cells. Although anti-phase II antibodies did not bind to the surfaces of native intact phase I cells, they bound to phase I proteins if the proteins were solubilized for sodium dodecyl sulfate-polyacrylamide gel electrophoresis and analyzed by immunoblotting. In addition, removal of the phase I lipopolysaccharide (LPS) by trichloroacetic acid exposed surface proteins for reactivity with anti-phase II antibodies, as shown by immunofluorescence assays, direct antibody binding, and immunoelectron microscopy using protein A-colloidal gold conjugates. Based on these observations, a simple model of phase variation is proposed to explain the apparently conflicting notions of the identity of the phase II antigen(s). The model suggests that the phase I LPS sterically hinders access of anti-phase II antibodies to a multitude of shared protein antigens, any one of which may confer phase II specificity. Exposure of these shared protein antigens through the appearance of a more truncated LPS (phase II) or extraction of the smooth-type phase I LPS allows antibody accessibility and therefore confers apparent phase II serospecificity.     

Red plaque formation of Coxiella burnetii and reduction assay by monoclonal antibodies

A red plaque technique for C. burnetii which utilizes primary chicken embryo cells, is described. Red plaques could be consistently detected as early as 6 days, usually 8 days post inoculation (p.i.), reflecting that C. burnetii proliferated within the phagolysosomes of host cells. Incubation with phase II monoclonal antibodies or inactivated immune sera containing phase I and phase II antibodies or phase II antibodies only, markedly reduced phase II C. burnetii red plaques. On the other hand, red plaques from phase I organisms increased several times when phase I cells were mixed with phase I monoclonal antibodies or inactivated immune sera containing phase I and phase II antibodies. By indirect red plaque reduction assay red plaque production by phase II cells could be reduced as well.     

Analysis of antibody response and immunoglobulins in sera of rabbits and guinea pigs immunized with Coxiella burnetii.

Rabbit and guinea pig sera and their immunoglobulin fractions (IgM and IgG) were examined by complement-fixation (CF), microagglutination (MA), opsonization-phagocytosis (OP) and serum protection (SP) tests at intervals after immunization with live phase I and phase II Coxiella burnetii suspensions. In general, MA antibodies, but also decreased, earlier than CF antibodies. The anamnestic immune response was higher with lower primary doses. Both phase II and phase I CF and MA antibodies as well as phase I opsonins and protective antibodies were found in either immunoglobulin fraction depending on post-immunization intervals. At later intervals and especially after the second immunizing dose the levels of protective antibodies in whole sera, but mainly in the IgG fractions, showed a better agreement with those of phase I opsonins rather than of phase I CF and MA antibodies.     

Immunoblot analysis of antibody response in mice infected with Coxiella burnetii phase I.

Mouse sera collected from day 4 to day 133 postinfection (p.i.) with phase I Coxiella burnetii strain Nine Mile were analysed by immunoblotting with phase I C. burnetii cell lysate. Antibodies of IgG class protein antigens were revealed already on day 10 p.i. (60, 49 and 27 kD proteins), followed by those to 77 kD protein from day 18 p.i. and to further 7 (from 42 to 70 kD) proteins from day 28 p.i. IgG antibody reaction was observed also with 5 antigens in 14-20 kD region corresponding to lipopolysaccharides from day 22 p.i. Surprisingly, antibodies of IgM type appeared later (from day 22 p.i.) and were directed only to protein antigens, most markedly to 60 and 77 kD proteins. Differences in immunoblot patterns observed with the serum collected on day 72 p.i. before and after absorption to phase I and phase II C. burnetii cells, and to phase I cells treated by trichloroacetic acid (TCA) or KIO4, indicate the surface localization of protein phase I antigenic epitopes, which can be destroyed partly by TCA and almost completely by KIO4 treatment.     

Analysis of the cells involved in the lymphoproliferative response to Coxiella burnetii antigens.

Vaccination with an inactivated, whole cell, Q fever vaccine (Q-vax) induces lasting antibody conversion and a positive delayed-type hypersensitivity (DTH) skin reaction in about 60% of recipients but a long-lasting positive lymphoproliferative or mitogenic response to C. burnetii antigens with peripheral blood mononuclear cells (PBMC) in 85-95% of subjects. Analysis of the lymphoproliferative response to C. burnetii antigens has now been made by fractionation-reconstitution experiments with PBMC from vaccines, from past infections, and from healthy controls. The major contributor to the response in immune subjects proved to be the T lymphocyte. T cells were stimulated by both the phase I and phase II antigens of two prototype strains of C. burnetii and responses were greatly amplified by addition of IL-2. Similar T lymphocyte stimulation profiles were obtained with the 'Priscilla' strain of C. burnetii which represents a different biotype of Coxiella isolated from Q fever endocarditis; Q-vax is therefore likely to protect against endocarditis strains. Fractionation-reconstitution experiments with T and B cells from vaccines and subjects infected in the past, using various antigenic or haptenic fractions from C. burnetii indicate that protein, non-lipopolysaccharide components of the organism are responsible for the mitogenic response of immune T cells. However, the role of the lipopolysaccharide in the protective immunogen has still to be defined.     

Efficiency of various serological techniques for diagnosing Coxiella burnetii infection

An indirect immunofluorescence assay (IFA) using a recently developed commercial kit for detecting antibodies against Coxiella burnetii (C.b.), the etiological agent of Q fever, has been evaluated using human field serum samples. The IFA was compared with an ELISA and a complement fixation test (CFT). The IFA was based on the corpuscular C.b. phase I and phase II antigens specific to anti-C.b. phase I and II antibodies, respectively. Fifty sera from persons with symptoms of Q fever were examined in this study. The IFA compared with the ELISA showed the sensitivities of 97.7% and 87.2% for IgG and 66.7% and 60.0% for IgM phase II and I antibodies, respectively and the specificities of 100% and 90.0% for IgG and 75.9% and 64.7% for IgM phase II and phase I antibodies, respectively. Due to a limited number of sera positive in the IgA antibody testing, the data presented should be considered with caution. It appears that the IFA strikes a very good balance between high specificity and sensitivity with phase II and phase I IgG antibodies and a less satisfactory one with IgM antibodies. The CFT failed in one of the above aspects showing a good specificity but a poor sensitivity, especially for phase I antibodies. The study demonstrated that the IFA is suitable for diagnosing Q fever and its therapeutic follow-up and is a good candidate for screening sera in large numbers. A certain limitation, especially in testing early stages of the chronic disease, could be a low fluorescence intensity of the IgA positive control in comparison with the IgA negative one.     

Monoclonal antibodies to Coxiella burnetii that cross-react with strain Nine Mile.

Results are presented to show the binding properties of five monoclonal antibodies directed to Coxiella burnetii Priscilla with cross-reactions to the Nine Mile strain. The monoclonal antibodies preferentially recognize phase I epitopes by ELISA and recognize phase II epitopes by immunoblotting but do not allow differentiation between so-called chronic and acute strains of C. burnetii. The only difference in reactivity was in the staining pattern revealed after reactions with lipopolysaccharide I antigens.     

Lipopolysaccharide variation in Coxiella burnetti: intrastrain heterogeneity in structure and antigenicity.

We isolated lipopolysaccharides (LPSs) from phase variants of Coxiella burnetii Nine Mile and compared the isolated LPS and C. burnetii cells by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting. The LPSs were found to be the predominant component which varied structurally and antigenically between virulent phase I and avirulent phase II. A comparison of techniques historically used to extract the phase I antigenic component revealed that the aqueous phase of phenol-water, trichloroacetic acid, and dimethyl sulfoxide extractions of phase I C. burnettii cells all contained phase I LPS, although the efficiency and specificity of extraction varied. Our studies provide additional evidence that phase variation in C. burnetii is analogous to the smooth-to-rough LPS variation of gram-negative enteric bacteria, with phase I LPS being equivalent to smooth LPS and phase II being equivalent to rough LPS. In addition, we identified a variant with a third LPS chemotype with appears to have a structural complexity intermediate to phase I and II LPSs. All three C. burnetii LPS contain a 2-keto-3-deoxyoctulosonic acid-like substance, heptose, and gel Limulus amoebocyte lysates in subnanogram amounts. The C. burnetii LPSs were nontoxic to chicken embryos at doses of over 80 micrograms per embryo, in contrast to Salmonella typhimurium smooth- and rough-type LPSs, which were toxic in nanogram amounts.     

Studies of the antigenic structure of Coxiella burnetii

The antigenic structure of Coxiella burnetii (C.b.) was studied by absorption of human and animal immune sera with C.b. organisms in the natural phase II (NPh II) or with artificial phase II (ArPh II) organisms prepared by their treatment with KIO4. It was found that immune sera absorbed with one type of phase II organisms still reacted with the antigen of another type of phase II organisms as demonstrated in both microagglutination (MA) and complement-fixation (CF) tests.     

Characterization of a phase I Coxiella burnetii chloroform-methanol residue vaccine that induces active immunity against Q fever in C57BL/10 ScN mice.

The effect of phase I Coxiella burnetii chloroform-methanol residue vaccine (CMRV) on the response of murine splenic lymphocytes to mitogenic and antigenic stimuli was evaluated in C57BL/10 ScN endotoxinnonresponder mice with an in vitro lymphocyte proliferation assay. Intraperitoneal injection of phase I CMRV resulted in antibody production against phases I and II antigens. Lymphocytes were responsive in vitro to concanavalin A, phytohemagglutinin, pokeweed mitogen, and specific recall antigens. Antibodies against phases I and II antigens were not detected after intraperitoneal injection of chloroform-methanol extract (CME). Lymphocytes also were only slightly hyporesponsive to mitogens. Reconstitution of the CMRV with the CME of phase I whole cells restored the immunopathological reactions that were associated with the phase I whole cell vaccine (WCV). The CMRV was more mitogenic than the WCV for lymphocytes from mice injected with saline. Lymphocytes from phase I WCV-injected mice were negatively modulated with nontoxic concentrations of homologous WCV or CMRV. Lymphocytes from phase I CMRV-injected mice were only slightly hyporesponsive to mitogens and were significantly stimulated by antigens of either WCV or CMRV as recall antigens. Vaccination of mice with 100 micrograms of CMRV, CME, or WCV provided 80, 0, or 50% protection, respectively, against a lethal intraperitoneal challenge with viable phase I C. burnetii. The epitopes which induce immunological hyporesponsiveness, negative modulation, and the death of lymphocytes were fractionated into the CMRV and CME. The CMRV provides at least one of the determinants which induce immunosuppression, whereas CME contains specific or nonspecific components or both. Collectively, these results show that the CMRV may be a potential candidate to replace the WCV currently used for human vaccination.     

Antigenicity of chloroform-methanol-treated Coxiella burnetii preparations

Phase I Coxiella burnetii (C.b.) cells untreated (Cb I) or treated with chloroform-methanol (CM) mixture (Cb I-CM) were compared as to their capacity to induce antibodies in laboratory animals and cattle, their ability to elicit delayed type hypersensitivity (DTH) reaction in mice and rabbits and protective effect in mice. In all animal species (mice, guinea pigs, rabbits, cattle) tested, the same doses of Cb I-CM cells induced lower levels of both phase I and phase II microagglutinating (MA) antibodies than Cb I cells at different intervals post-immunization (p.i.). Though for elicitation of DTH reaction in rabbits immunized with different C.b. preparations lower doses of Cb I than of Cb I-C M cells were necessary, C.b. cells caused inflammatory reaction at lower doses also in control rabbits. In mice immunized with Cb I and Cb I-CM cells, but not with trichloracetic acid extract (TCAE) from intact Cb I cells, DTH reaction was elicited by the same doses of Cb I and Cb I-CM cells. Higher immunizing doses of Cb I-CM than of Cb I cells were required, however, to induce DTH reaction (as tested by TCAE) as well as protection to phase I virulent challenge. TCAE from intact Cb I cells was protective in mice also at lower doses than TCAE from Cb I-CM cells (TCAE-CM). In humans who suffered from Q fever one year ago, higher proportion of positive skin test (ST) reactions and antibody recalls with higher mean geometric titres (MGT) of phase II MA antibodies was noticed following intradermal administration of TCAE than of TCAE-CM. When humans with no evidence of Q fever in past were vaccinated with TCAE or TCAE-CM, the former preparation not only caused higher proportion of both local and general post-vaccination reactions, but also of phase II MA antibody response and positive ST reactions as tested by TCAE 3 months post-vaccination in addition to higher proportion of phase II MA antibody recalls.     

Biological and immunological properties of Coxiella burnetii vaccines in C57BL/10ScN endotoxin-nonresponder mice.

Vaccines prepared from Formalin-killed whole cells of Coxiella burnetii (Ohio strain) or from chloroform-methanol residue (CMR) and extract (CME) of such cells were examined for biological and immunological properties in male C57BL/10ScN endotoxin nonresponder mice. Vaccines containing killed whole cells induced a high incidence of gross pathology, as evidenced by liver necrosis, significant splenomegaly, and significant hepatomegaly in mice. The degree and onset of these pathological changes were directly and inversely proportional, respectively, to the dosage of killed organisms administered. Conversely, no splenomegaly, hepatomegaly, or liver necrosis were observed in mice inoculated with CMR or CME. Moreover, killed whole cells were lethal for mice at dosages of 150 to 1,200 micrograms, whereas no deaths were seen in animals given 1,200 micrograms of CMR. In addition, antibodies against phase I and phase II antigens of C. burnetii were detected in the sera of mice inoculated with either whole cells or CMR. Enhanced blastogenic response of splenic lymphocytes was observed when animals were vaccinated with killed whole cells and CMR but not with CME. Moreover, 80 to 90% of mice inoculated with 300 micrograms of the CMR were protected against a lethal challenge of viable rickettsiae, whereas only 50% of the animals given 300 micrograms of killed whole cells were protected. Treatments with CME were essentially without value, since no antibodies were detectable and no significant protection was elicited. Collectively, these results show that, although killed whole cells induced immunity in C57BL/10ScN mice, they induced deleterious tissue reactions, whereas CMR, which also induced immunity, was essentially nondeleterious, based on the parameters employed. These observations suggest that the chloroform-methanol-extractable component(s) is implicated in the deleterious tissue reactions and that the phase I and II antigens may not be involved in the induction of the pathology observed in C57BL/10ScN mice.     

Induction of splenomegaly in mice by killed Coxiella burnetii cells

Splenomegaly induced in mice inoculated intraperitoneally (i.p.) with purified formalin-killed phase I and phase II Coxiella burnetii (C.b.) cells was dose-dependent. The phase I cells induced higher splenomegaly than phase II cells. The splenomegaly-inducing ability of phase I cells was reduced upon incubation with phase I but not with phase II antiserum, whereas the phase II cells preincubated with phase I or phase II immune sera induced higher splenomegaly than the phase II cells alone. Phase I cells caused lower splenomegaly in mice previously immunized with C.b. The splenomegaly-inducing ability of phase I cells was abolished by mild acid hydrolysis, by treatment either with phenol-chloroform-petroleum ether (PCP) or with a chloroform-methanol (CM) mixture. However, either the CM or the PCP-treated phase I cells retained their capacity to protect mice challenged with virulent phase I C.b.     

Specificity of antibodies induced by Streptococcus mutans during immunization against dental caries.

Protection against smooth surface dental caries was investigated in fifteen young rhesus monkeys which were immunized subcutaneously with Streptococcus mutans serotype c in Freund's incomplete adjuvant. Monkeys immunized with killed whole organisms developed significantly less caries than control animals. Monkeys immunized with pronase-treated cell walls developed significantly more caries than control animals while monkeys immunized with untreated cell walls showed no such enhancement of caries. Haemagglutinating and complement-fixing antibodies to cell walls and culture supernatant antigens (SN Ag) of S. mutans developed in the sera of all immunized animals to a similar degree. Antibodies to lipoteichoic acid and to an insoluble dextran preparation were found in all immunized animals and showed no relationship to the prevalence of caries. Antibodies to the serotype c polysaccharide were also found in animals immunized with whole cells and pronase-treated cell walls. However, precipitating antibody levels to partially purified antigens I/II and II, derived from SN Ag, but present also in cells, were related to the development of caries. Animals immunized with whole cells and with untreated cell walls developed a brisk antibody response to antigen I/II, while those immunized with pronase-treated cell walls responded more slowly. The results suggest that immunization may induce both caries reduction and enhancement, depending on the antibody response which is developed.