Lethal effect of Rickettsia rickettsii on its tick vector (Dermacentor andersoni).

Rickettsia rickettsii, the causative agent of Rocky Mountain spotted fever, was lethal for the majority of experimentally and transovarially infected Rocky Mountain wood ticks (Dermacentor andersoni). Overall, 94.1% of nymphs infected as larvae by feeding on rickettsemic guinea pigs died during the molt into adults and 88. 3% of adult female ticks infected as nymphs died prior to feeding. In contrast, only 2.8% of uninfected larvae failed to develop into adults over two generations. Infected female ticks incubated at 4 degreesC had a lower mortality (80.9%) than did those held at 21 degreesC (96.8%). Rickettsiae were vertically transmitted to 39.0% of offspring, and significantly fewer larvae developed from infected ticks. The lethal effect of R. rickettsii may explain the low prevalence of infected ticks in nature and affect its enzootic maintenance.

Characterization of an endosymbiont infecting wood ticks, Dermacentor andersoni, as a member of the genus Francisella.

A microorganism (Dermacantor andersoni symbiont [DAS]) infecting Rocky Mountain wood ticks (D. andersoni) collected in the Bitterroot Mountains of western Montana was characterized as an endosymbiont belonging to the genus Francisella. Previously described as Wolbachia like, the organism's DNA was amplified from both naturally infected tick ovarial tissues and Vero cell cultures by PCR assay with primer sets derived from eubacterial 16S ribosomal DNA (rDNA) and Francisella membrane protein genes. The 16S rDNA gene sequence of the DAS was most similar (95.4%) to that of Francisella tularensis subsp. tularensis. Through a combination of Giménez staining, PCR assay, and restriction fragment length polymorphism analysis, 102 of 108 female ticks collected from 1992 to 1996 were infected. Transovarial transmission to female progeny was 95.6%, but we found no evidence of horizontal transmission.

Chlamydia psittaci IncA is phosphorylated by the host cell and is exposed on the cytoplasmic face of the developing inclusion.

Chlamydiae are obligate intracellular bacteria that replicate within a non-acidified vacuole called an inclusion. Chlamydia psittaci (strain GPIC) produces a 39 kDa protein (IncA) that is localized to the inclusion membrane. While IncA is present as a single 39 kDa species in purified reticulate bodies, two additional higher M(r) forms are found in C. psittaci-infected cells. This finding suggested that IncA may be post-translationally modified in the host cell. Here we present evidence that IncA is a serine/threonine phosphoprotein that is phosphorylated by host cell enzymes. This conclusion is supported by the following experimental findings: (i) treatment of infected cells with inhibitors of host cell phosphatases or kinases altered the electrophoretic migration pattern of IncA; (ii) treatment with calf intestinal alkaline phosphatase eliminated the multiple-banding pattern of IncA, leaving only the protein band with the lowest relative molecular weight; and (iii) radioimmunoprecipitation of lysates of [32P]-orthophosphate-labelled infected HeLa cells with anti-IncA antisera demonstrated that the two highest M(r) IncA bands were phosphorylated. A vaccinia-virus recombinant expressing incA was used to determine if HeLa cells can phosphorylate IncA in the absence of a chlamydial background. IncA in lysates of these cells migrated identically to that seen in C. psittaci-infected cells, indicating the host cell was responsible for the phosphorylation of the protein. Microinjection of fluorescently labelled anti-IncA antibodies into C. psittaci-infected HeLa cells resulted in immunostaining of the outer face of the inclusion membrane. Collectively, these results demonstrate that IncA is phosphorylated by the host cell, and regions of IncA are exposed at the cytoplasmic face of the inclusion.

Improved plaque assays for Rickettsia prowazekii in Vero 76 cells.

Typhus group rickettsiae, including Rickettsia prowazekii and R. typhi, produce visible plaques on primary chick embryo fibroblasts and low-passage mouse embryo fibroblasts but do not form reproducible plaques on continuous cell culture lines. We tested medium overlay modifications for plaque formation of typhus group rickettsiae on the continuous fibroblast cell line Vero76. A procedure involving primary overlay with medium at pH 6.8, which was followed 2 to 3 days later with secondary overlay at neutral pH containing 1 microgram of emetine per ml and 20 micrograms of NaF per ml, resulted in visible plaques at 7 to 10 days postinfection. A single-step procedure involving overlay with medium containing 50 ng of dextran sulfate per ml also resulted in plaque formation within 8 days postinfection. These assays represent reproducible and inexpensive methods for evaluating the infectious titers of typhus group rickettsiae, cloning single plaque isolates, and testing the susceptibilities of rickettsiae to antibiotics.

Directional actin polymerization associated with spotted fever group Rickettsia infection of Vero cells.

Members of the spotted fever group (SFG) of rickettsiae spread rapidly from cell to cell by an unknown mechanism(s). Staining of Rickettsia rickettsii-infected Vero cells with rhodamine phalloidin demonstrated unique actin filaments associated with one pole of intracellular rickettsiae. F-actin tails greater than 70 microns in length were seen extending from rickettsiae. Treatment of infected cells with chloramphenicol eliminated rickettsia-associated F-actin tails, suggesting that de novo protein synthesis of one or more rickettsial proteins is required for tail formation. Rickettsiae were coated with F-actin as early as 15 min postinfection, and tail formation was detected by 30 min. A survey of virulent and avirulent species within the SFG rickettsiae demonstrated that all formed actin tails. Typhus group rickettsiae, which do not spread directly from cell to cell, lacked F-actin tails entirely or exhibited only very short tails. Transmission electron microscopy demonstrated fibrillar material in close association with R. rickettsii but not Rickettsia prowazekii. Biochemical evidence that actin polymerization plays a role in movement was provided by showing that transit of R. rickettsii from infected cells into the cell culture medium was inhibited by treatment of host cells with cytochalasin D. These data suggest that the cell-to-cell transmission of SFG rickettsiae may be aided by induction of actin polymerization in a fashion similar to that described for Shigella flexneri and Listeria monocytogenes.

Immunization of dogs with Q fever vaccines: comparison of phase I, II and phase I CMR Coxiella burnetii vaccines.

Q fever vaccines were tested in mixed breed dogs by vaccinating them with formalin-killed Coxiella burnetii whole cells (WC) in either phase I (WCI) or phase II (WCII), or the chloroform: methanol residue (CMR) subunit of phase I cells. Phase I vaccines mixed (1:1) with Freund's incomplete adjuvant (FIA) induced humoral immune responses to phases I and II antigens as measured by microagglutination assay. The CMR vaccine mixed (1:1) with FIA induced greater antigen-specific antibody levels to both phases I and II antigens than the corresponding WCI vaccine. The WCII vaccine induced antibody responses to only phase II antigens. The time course of erythema and induration after skin testing with C. burnetii antigens were suggestive of cell-mediated immunity (CMI). Although granulomas were observed with only WCI and WCII, none of the skin test antigens induced abscesses at the injection site. In contrast, axillary nodes draining the vaccine injection site developed sterile draining abscesses in all dogs by days 19 to 24 for the WCI and CMR, and day 104 for the WCII vaccines. The abscesses had resolved within 30 days after first appearance. Responses to Con A and PHA and recall antigens of lymphocytes from the blood, axillary and mesenteric nodes, and spleen at 222 days after vaccination were variable among dogs. Lymphocytes from various organs responded to one or more of the recall antigens and to both mitogens in the absence or presence of indomethacin. Although these Q fever vaccines induced humoral and CMI, either the antigens or FIA caused sterile draining abscesses.(ABSTRACT TRUNCATED AT 250 WORDS)

Vaccines against coxiellosis and Q fever. Development of a chloroform:methanol residue subunit of phase I Coxiella burnetti for the immunization of animals.

We have demonstrated the safety, immunogenicity, and efficacy of the WC and CMR vaccines in guinea pigs. Vaccination of guinea pigs with either WC or CMR protects animals against challenge with virulent C. burnetii. A total of 2 micrograms of either WC or CMR vaccine was a significant priming dose. A total of 20 micrograms gave complete protection against lethal challenge. Detection of antibodies to phase II cells by microaglutination, after vaccination with either WC or CMR and before lethal challenge, correlated with the ability of guinea pigs to mount a protective immune response. The PD50 values for WC and CMR vaccines, administered as a single dose, were 0.3 and 1.4 micrograms per animal, respectively. In contrast, the PD50 values for the WC and CMR vaccines, administered as two doses, were 0.83 and 0.72 micrograms per animal, respectively. Although the PD50 values for the two vaccines are similar, the CMR vaccine is preferred over the WC vaccine because it induces significantly fewer adverse reactions, and repeat injections can be given. Unvaccinated guinea pigs do not clear infectious microorganisms after challenge infection. Vaccination before challenge infection reduces the infectious load of C. burnetti in the blood and in various organs of the animals. When vaccinated animals were challenge infected and treated with rifampicin, the microorganisms were not eliminated from various organs. However, the combination of vaccination, challenge, and rifampicin treatment is effective in reducing the number of infectious microorganisms in some of these sites. We have demonstrated the safety and immunogenicity of the CMR vaccine in sheep and goats. Animals that were seropositive for one or more antigens developed significant levels of antibodies to alternate antigens, but no adverse reactions were observed at the site of s.c. injection of the CMR vaccine. This demonstrates that seropositive animals can be successfully immunized with this vaccine. These results also indicate that a long-term vaccination program using the CMR vaccine has the potential for producing animals with significant antibody titers to C. burnetii and perhaps lifelong immunity. The goal of a Q fever vaccination program is to produce immunized animals that are able to clear completely the infectious microorganisms. The appropriate vaccination schedule to render adult animals and their offspring "Q fever-free" should now be thoroughly investigated.

Evidence for proteolytic cleavage of the 120-kilodalton outer membrane protein of rickettsiae: identification of an avirulent mutant deficient in processing.

The 120-kDa rickettsial outer membrane protein (rOmpB) is encoded by a gene with the capacity to encode a protein of approximately 168 kDa. The carboxy-terminal end of the molecule is apparently cleaved to yield 120- and 32-kDa products. Both polypeptides are surface exposed and remain associated with the outer membrane of intact rickettsiae. All species of rickettsiae examined display similar cleavage of rOmpB. Comparison of diverse species of rickettsiae demonstrate a conserved N terminus of the 32-kDa fragment, with a predicted procaryotic secretory signal peptide immediately upstream of the proposed cleavage site. Coprecipitation of the 120-kDa rOmpB protein and the 32-kDa peptide by monoclonal antibodies specific for the 120-kDa portion of the molecule suggests that the two fragments remain noncovalently associated on the surface of rickettsiae. Analysis of an avirulent mutant of Rickettsia rickettsii revealed reduced amounts of the 120- and 32-kDa fragments, but with a correspondingly larger rOmpB protein that displayed properties expected of the putative precursor. This avirulent mutant grows intracellularly but fails to cause the lysis of infected cells that is typical of R. rickettsii. DNA sequence analysis of the region of the gene encoding the cleavage site of the avirulent strain revealed no difference from the sequence obtained from virulent R. rickettsii. The 168-kDa putative precursor of the avirulent strain of R. rickettsii was not extracted from the surface by dilute buffers, as is the 120-kDa protein of virulent R. rickettsii or R. prowazekii. These latter results suggest that the 32-kDa C-terminal region of the molecule may serve as a membrane anchor domain.

Action of pertussigen (pertussis toxin) on serum IgE and on Fc epsilon receptors on lymphocytes.

Pertussigen (pertussis toxin (PT] is one of the most effective stimulators of IgE production in mice and rats. Employing flow microfluorimetric analysis (FMF), we showed that PT increases the percentage of blood and spleen lymphocytes with IgE on their surface. The percentage of IgE-bearing cells in the spleen of normal untreated C57Bl/10SCN mice of various ages varied from 2.2 to 12.2%, with an average value of 6.1 +/- 5.4%. In mice treated with 400 ng of PT and 1 mg of chicken egg albumin (EA), the percentage of these cells increased, 14 days after immunization, to an average value of 31.1 +/- 2.2%. Immunization of mice with PT alone increase the percentage of IgE-bearing cells only slightly (13.1 +/- 2.2% of the splenic lymphocytes) while injection of 1 mg of EA alone did not have any detectable action. As little as 6 ng of PT, when given simultaneously with 1 mg of EA, increased the percentage of IgE-bearing lymphocytes. A booster dose of 10 micrograms of EA given on Day 14 induced a further increase in the percentage of these cells even when as little as 0.039 ng of PT had been given at the time of initial immunization. PT was effective when given 4 days before or 5 days after EA. EA was effective when given 4 days before or 4 days after PT, but not 8 days after. The increase in IgE-bearing cells was mainly due to cytophilic binding of IgE to receptors for the epsilon chain of IgE (Fc epsilon) on the surface of lymphocytes rather than to a greater number of IgE-producing cells. This was shown by removing the IgE from Fc epsilon receptors by acid treatment which reduced the percentage of IgE-bearing cells to nearly normal values. The antibodies of IgE class with specificity to EA were increased dramatically, while antibodies with specificity to PT were not detected.

Expression of pertussis toxin correlates with pathogenesis in Bordetella species.

Pertussis toxin is a principal determinant of virulence produced by Bordetella pertussis in the disease whooping cough and is the primary toxinogenic component of the pertussis vaccine. This toxin is not produced by the closely related species Bordetella parapertussis. Toxinogenic strains of B. parapertussis were constructed by the conjugative introduction of cloned pertussis toxin genes. These and analogous toxinogenic and nontoxinogenic strains of B. pertussis were assayed for their toxicity and reactogenic activities. Expression of active toxin by either Bordetella sp. correlated with the induction of leukocytosis, anaphylaxis, and histamine sensitivity.

Role of pertussigen (pertussis toxin) on the mouse protective activity of vaccines made from Bordetella species.

Pertussigen [pertussis toxin (Ptx)] from Bordetella pertussis, when detoxified, induces protection in mice to intracerebral challenge (ic) with virulent B. pertussis. In its native form, minute nonprotective doses promote the development of immunity induced by other antigens of B. pertussis. As little as 4 ng of Ptx, given with a nonprotective dose of 8 X 10(7) killed cells of the phase III Sakairi strain, promoted detectable protection to ic challenge. Native Ptx in doses of 0.4 to 400 ng did not protect mice, and vaccines made from strains not producing Ptx induced only weak protection. The marked enhancing action of Ptx was also observed with 5 micrograms of purified filamentous hemagglutinin and with vaccines made from other species of the Bordetella genus, such as B. parapertussis and B. bronchiseptica, but it was not observed with B. pertussis endotoxin. In addition, Ptx was still effective when given as late as 7 days after the vaccine. Antibodies to surface antigens of the challenge strain were demonstrated in sera of mice immunized with vaccines prepared with the different Bordetella species tested, but antibodies to Ptx were detected only in the sera of mice immunized with the wild-type B. pertussis strains. Glutaraldehyde detoxified Ptx does not have this action. Pretreatment of normal mice with Ptx, also enhanced the protective action of a mouse antiserum to a wild-type strain of B. pertussis. These observations show that antigens other than Ptx are responsible for the protection, and that Ptx acts non-specifically to enhance the mouse protective action of those antigens.

ADP-ribosyltransferase activity of pertussis toxin and immunomodulation by Bordetella pertussis.

Pertussis toxin is produced by the causative agent of whooping cough, Bordetella pertussis, and is an adenosine diphosphate (ADP)-ribosyltransferase capable of covalently modifying and thereby inactivating many eukaryotic G proteins involved in cellular metabolism. The toxin is a principal determinant of virulence in whooping cough and is a primary candidate for an acellular pertussis vaccine, yet it is unclear whether the ADP-ribosyltransferase activity is required for both pathogenic and immunoprotective activities. A B. pertussis strain that produced an assembled pertussis holotoxin with only 1 percent of the ADP-ribosyltransferase activity of the native toxin was constructed and was found to be deficient in pathogenic activities associated with B. pertussis including induction of leukocytosis, potentiation of anaphylaxis, and stimulation of histamine sensitivity. Moreover, this mutant strain failed to function as an adjuvant and was less effective in protecting mice from intracerebral challenge infection. These data suggest that the ADP-ribosyltransferase activity is necessary for both pathogenicity and optimum immunoprotection. These findings bear directly on the design of a nontoxic pertussis vaccine.

Comparison of enzyme-linked immunosorbent assay and complement fixation and indirect fluorescent-antibody tests for detection of Coxiella burnetii antibody.

An enzyme-linked immunosorbent assay (ELISA) was developed to detect immunoglobulin G to Coxiella burnetii phase II. Serum samples from 213 patients who had had Q fever 1 year previously and from 301 blood donors from six localities in Switzerland were tested by ELISA and by indirect fluorescent-antibody (IFA) and complement fixation (CF) tests. The ELISA and the IFA and CF tests detected antibody to C. burnetii in 202 (94.8%), 193 (90.6%), and 166 (77.8%) of the 213 Q fever patients, respectively. With the serum samples from blood donors, the ELISA yielded a higher percentage of positive sera than did the IFA and CF tests. The high specificity of the three tests was confirmed by analyzing paired serum samples from 36 patients suffering from acute pneumonia of viral or bacterial origin. In these cases, the serological results were negative by the three tests, except for three Q-fever cases included as positive control.

Anaphylaxis or so-called encephalopathy in mice sensitized to an antigen with the aid of pertussigen (pertussis toxin).

Sensitization of mice with 1 mg of bovine serum albumin (BSA) or chicken egg albumin (EA) given intraperitoneally and 300 to 400 ng of pertussigen (pertussis toxin [Ptx]) given intravenously (i.v.) induced a high degree of anaphylactic sensitivity when the mice were challenged i.v. with 1 mg of antigen 14 days later. Regardless of H-2 haplotype, all of the strains tested (CFW, BALB/cJ, DBA/2J, and C3H.SW/SnJ) were susceptible to anaphylaxis. Sensitization of mice by a multiple-dose procedure that has been reported to induce fatal encephalopathy in mice (L. Steinman, A. Weiss, N. Adelman, M. Lim, R. Zuniga, J. Oehlert, E. Hewlett, and S. Falkow, Proc. Natl. Acad. Sci. USA 82, 8733-8736, 1982) (1 mg of BSA on day -1, 100 to 400 ng of Ptx on day zero 1 mg of BSA on day +1, 100 to 400 ng of Ptx on day +2, and 1 mg of BSA on day +6) induced shock in BALB/cJ, DBA/2J, and C3H.SW/SnJ mice, but not in CFW mice. When EA was used instead of BSA, CFW, BALB/cJ, and C3H.SW/SnJ mice did not develop fatal shock, whereas DBA/2J mice did. When dose 3 of antigen (BSA or EA) was postponed to day +21, all mouse strains sensitized by the multiple-dose procedure were found to be susceptible to shock. The fatal shock induced by this procedure, as well as that induced by giving a single sensitizing dose of antigen and Ptx, could be prevented by one to three 1-ml doses of saline given i.v. at the time signs of severe shock appeared. Although only one dose of saline was often sufficient to save the mice, two or three doses were usually needed. Microscopic changes were not found in midsagittal sections of the brains of mice sensitized by either procedure. This was true of mice that died from shock or were saved from shock by injections of saline. From these results, we concluded that the proposed model for encephalopathy induced in mice by Ptx and BSA demonstrates only the well-known anaphylactogenic effect of Ptx or pertussis vaccine. Since there are many other more sensitive methods to detect Ptx, induction of anaphylaxis is not of much value for detection or quantitation of Ptx in pertussis vaccine.

Identification of phase-specific antigenic fractions of Coxiella burnetti by enzyme-linked immunosorbent assay.

Antigenic fractions of Coxiella burnetii phase variants were identified with an enzyme-linked immunosorbent assay (ELISA). Immune sera from guinea pigs immunized with Formalin-inactivated phase I or phase II whole cells were used to measure the antigenic activity of whole cells and various soluble and particulate preparations. Phase-specific antigens of C. burnetii whole cells and fractions were compared by dose-response curves at different (antigen and antibody) dilutions. Water-soluble extracts prepared by meta-periodate, ether, and phenol extraction of phase I whole cells yielded antigenic fractions which reacted with anti-phase I antibodies. The extraction of phase I whole cells with dimethyl sulfoxide, trichloracetic acid, and Formalin yielded antigenic fractions which detected antibodies in both anti-phase I and -phase II sera. Interestingly, the trichloracetic acid extract of phase I whole cells also contained a component which bound nonimmune immunoglobulin. The sera of animals immunized with whole cells of the phase II Australian QD strain reacted with lipopolysaccharides of the phase I and phase II Nine Mile strains. Therefore, variations in lipopolysaccharide structure among phase variants of C. burnetii were detected as cross-reactions with immune sera from an interspecific strain. Comparisons of immunofluorescence, microagglutination, and the complement fixation assays with the ELISA indicated greater sensitivity and specificity of the ELISA for the measurement of phase-specific antigens and antibodies.

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.

Encephalitis caused by Coxiella burnetii.

Acute infection with Coxiella burnetii usually results in a self-limited illness, but it can occasionally cause chronic endocarditis or hepatitis. Headache is a common presenting symptom of acute infection with this agent, but specific neurological abnormalities are rare. We report the case of a patient with acute Q fever that caused frank encephalitis. We also review the literature on central nervous system disease attributable to C. burnetii.

Humoral immune response to Rocky Mountain spotted fever in experimentally infected guinea pigs: immunoprecipitation of lactoperoxidase 125I-labeled proteins and detection of soluble antigens of Rickettsia rickettsii.

Rickettsia rickettsii, the etiologic agent of Rocky Mountain spotted fever, purified from infected L-929 cells by density gradient banding were extrinsically radioiodinated with lactoperoxidase. Immunodominant 125I-labeled antigens were identified by radioimmunoprecipitation of detergent-solubilized antigens with protein A-Sepharose and anti-R. rickettsii sera collected 0, 3, 7, 11, 32, and 163 days after infection of guinea pigs. The average fever greater than or equal to 40 degrees C was detected by days 3 and 4 after infection with 6 X 10(7) and 6 X 10(6) PFU, respectively. By microagglutination and complement fixation assays, anti-R. rickettsii antibodies were detected as early as day 3 after infection, with titers increasing markedly between days 7 and 163. Convalescent sera, collected on day 163, from infected guinea pigs were used to identify seven 125I-labeled antigens with apparent molecular sizes of 186,000 (I), 145,000 (II), 49,000 (III), 32,000 (IV), 27,500 (V), 17,500 (VI), and 16,500 (VII) daltons. Differences in antibody reactivity and specificity against the seven antigens were demonstrated with serially obtained sera. Sera from a guinea pig infected with 6 X 10(7) PFU exhibited antibody-antigen interactions with all seven 125I-labeled antigens by day 7, whereas the same antibody activity required 32 days for an animal infected with 6 X 10(6) PFU. Prominent antibody activities toward proteins II and IV were demonstrated both early and late after infection. The fluids obtained from infected L-929 cells contained three soluble antigens which were detected with the 11-, 32-, and 163-day sera by an immunodiffusion assay. The soluble and 125I-labeled antigens of R. rickettsii identified in this study may be important candidates for vaccines against Rocky Mountain spotted fever.