q fever and coxiella burnetii: a model interactionst · identified in arthropods, fish, birds,...

MICROBIOLOGICAL REVIEWS, June 1983, p. 127-149 Vol. 47, No. 20146-0749/83/020127-23$02.00/0Copyright © 1983, American Society for Microbiology

Q Fever and Coxiella burnetii: a Model for Host-ParasiteInteractionst

0. G. BACA1 AND D. PARETSKY2*Department of Biology, University ofNew Mexico, Albuquerque, New Mexico 87131,1 and Department of

Microbiology, University of Kansas, Lawrence, Kansas 660452

HISTORY .................................................. 127EPIDEMIOLOGY .................................................. 129CULTIVATION AND GROWTH OF C. BURNETII.................................. 130Entry into and Proliferation Within Cultured Cells; Comparison with Other Rickettsiae 130Enumeration .................................................. 131

BIOLOGY OF C. BURNETII.................................................. 131Ultrastructure/Morphology .................................................. 131Genome .................................................. 132Ribosomes .................................................. 133Replication.................................................. 133Phase Variation.................................................. 134LPS.................................................. 135

BIOCHEMISTRY OF C. BURNETII ............................................... 135Glycolytic and Related Enzymes ................................................. 135Anabolic Enzymes of Amino Acids, Proteins, and Nucleic Acids ...................... 136Acidophilic Biochemistry and Intravacuolar Existence............................... 137

PATHOBIOLOGY.................................................. 138Humans .................................................. 138Embryonated Eggs .................................................. 139Guinea Pigs and Other Animals.................................................. 139Cell Cultures.......................................................... 141

IMMUNOLOGY.................................................. 141Vaccines .................................................. 143

SUMMARY AND PROSPECTS .................................................. 143ACKNOWLEDGMENTS .................................................. 144LITERATURE CITED ......................................................... 144

HISTORYIn 1935 an outbreak of a fever of unknown

etiology occurred among abattoir workers inBrisbane, Australia. R. Cilento, Director-Gener-al of Health and Medical Services for Queens-land, requested E. H. Derrick, Director of theLaboratory of Microbiology and Pathology ofthe Queensland Health Department at Brisbane,to investigate the nature of the disease. Der-rick's classic paper (54) provided the name "Q"(for Query) fever for the disease and accuratelydescribed most of the clinical features of theinfection. A subsequent review by Derrick (55)collated data from the time of the first descrip-tion of the disease in 1935.Q fever displayed certain clinical similarities

to typhus fever and to typhoid and paratyphoidfevers, but the absence of a rash and a negativeWeil-Felix reaction ruled out these infections.Although Derrick was unable to identify orisolate the etiological agent, he succeeded in

t This review is dedicated to Cora M. Downs, scientist,teacher, and friend.

transmitting Q fever to guinea pigs with theblood or urine of infected patients (54). As withhumans, guinea pigs developed a long-lived im-munity to the infection. Derrick sent a sample ofinfected guinea pig liver to Burnet, who used itto transmit Q fever to guinea pigs and mice.Burnet identified "typical rickettsiae" in infect-ed mouse spleen impression smears (35); theorganisms were ". . . less than 1 p. in length andabout 0.3 p. across; the shape varied from . . .

rods to coccoid forms" (35). In closely reasonedarguments, Burnet concluded that the agent ofQfever was a rickettsial agent unlike the typhus,scrub typhus, or spotted fever rickettsiae. Thepapers of Derrick and of Burnet and Freemanremain models of careful investigations, criticalanalyses, and conclusions. In 1938 a series ofpapers described a "filter-passing infectiousagent" from Dermacentor andersoni, ticks re-covered near Nine Mile Creek, 32 mile's west ofMissoula, Mont. (48, 52, 61). Guinea pigs infect-ed with the agent developed high fever andenlarged spleens. The organism occurred intra-vacuolarly in infected cells, and the vacuoliza-

127

on July 15, 2020 by guesthttp://m

mbr.asm

.org/D

ownloaded from

http://mmbr.asm.org/

128 BACA AND PARETSKY

tion could be so great as to force the cell nucleustoward the edge of the cell. A minute pleomor-phic organism was isolated which stained deeplywith the Macchiavello and Giemsa techniques(73). A laboratory worker ("case X") becameinfected and developed the symptoms of Q feveras originally described by Derrick and Burnet.Blood from case X successfully infected a guin-ea pig, whose spleen was infectious for otherguinea pigs. Cox concluded that the agent eitherwas a rickettsia or was rickettsia-like. Dyer (62)proposed the identity between the Q fever agentof Derrick and Burnet and the filter-passingagent, or "X virus," and subsequently Cox (48)proposed the name Rickettsia diaporica, incor-porating the rickettsial features of the organismwith its ability to flow through the pores of abacteriological filter. These classic papers estab-lished the presence of Q fever in the UnitedStates and were the background for the subse-quent nomenclature of Coxiella burnetii.Although Derrick and Burnet originally de-

scribed the organism as a rickettsia (Rickettsiaburneti), it is now clear that C. burnetii differsfrom "typical" rickettsiae, such as R. prowa-zeki, R. typhi, R. akari, and R. rickettsii, on thebasis of genome guanine-plus-cytosine content,vector of transmission, nature of intracellularproliferation, size and pleomorphic nature, sen-sitivity to antibiotics, and greater resistance tohigher temperature and lower pH (63, 81, 82).The genome size of C. burnetii is 1.04 x 109daltons, the same magnitude as that ofR. prowa-zeki, R. rickettsii, and R. typhi (138). An earlierestimate of 1.8 x 107 daltons (175) was probablybased on a heterogeneity of sheared DNA frag-ments. The 42.2% guanine-plus-cytosine contentof C. burnetii DNA is strikingly different fromthat of typical rickettsiae; the guanine-plus-cyto-sine compositions of some rickettsial DNAs areas follows: R. prowazekii (Breinl) and R. typhi(Wilmington), 29.0% (138); R. canada, 29.3 to30.3%; R. rickettsii, 32.0 to 33.2%; R. conori,32.9 to 33.3% (196). As described below, C.burnetii proliferates in a wide range of verte-brate and invertebrate hosts and infects humansvia aerosol inhalation, milk and meat ingestion,contact with infected blood, and arthropod andnonarthropod vectors. The agent proliferateswithin lysosomal vacuoles in the cytoplasm (14,38), in contrast to the replication patterns ofmost typical rickettsiae. Based on chicken em-bryo infection and protection assays, C. burnetiiis relatively insensitive to erythromycin, chlor-amphenicol, and thiocymetin as compared withtypical rickettsiae (149). Nevertheless, theGiemsa and Macchiavello staining response andthe obligate intracellular parasitic nature of theorganism justify its taxonomic position in theRickettsiaceae and, as a separate genus, Cox-

iella. A wealth of epidemiological, clinical, andimmunological information on Q fever and onthe biology of C. burnetii has been gathered.Comprehensive reviews are available (10, 11,34, 76, 144, 147, 203, 204, 216). The discussionsof Brezina (31) and Weiss (203, 204) should alsobe consulted.The typical picture of Q fever in humans (35,

54, 55) is an incubation period of 15 days or less,varying with route of exposure, dosage of rick-ettsiae, and age of victim. A febrile onsetreaches a plateau of 40°C within 2 to 4 days, lateraccompanied by malaise, anorexia, muscularpain, weakness, and intense (usually) preorbitalheadache. The headache later becomes general-ized, continuing in intensity throughout the dis-ease. A gradual defervescence occurs over a 1 to2 week period, although the fever may lastlonger in older patients and may display biphasicpeaks. Liver damage accompanied by hepa-tomegaly occurs, with frequent reports of hepat-ic granulomas (see below). A pneumonitis andbronchitis with a dry unproductive cough areoften symptomatic, leading to erroneous diag-noses of influenza or an influenza-like disease(the "Balkan grippe" or "Balkan fever" ofWorld War II and the "Termez fever" in Tash-kent). Derrick (54) had noted that, whereas thecases showed no palpable spleens, they wereprobably "enlarged to some extent in humancases." This prescient comment later provedaccurate, and subsequently both hepatomegalyand splenomegaly were found in human casesand were characteristic of experimentally infect-ed guinea pigs (158). In human cases a rickettse-mia is present in the early stages of fever, andthe rickettsiae may be found in urine at this time.Rickettsiae disappear from the blood with defer-vescence, but kidneys and heart valves mayremain latently infected. Endocarditis, althoughnot common, may be a sequela of infection (seebelow).A long-lived immunity generally results from

the infection, although chronic and latent infec-tions have been reported (180, 181). Q fever ismanaged by tetracycline and chloramphenicoltherapy (72), but C. burnetii is less sensitive tothese antibiotics than typical rickettsioses (149).Intravenous administration of erythromycinmay be an especially effective treatment (51). Qfever remains endemic in Australia where it hascontinued to be a major problem among live-stock workers (79).Reported human cases of Q fever steadily

increased in the United States between 1948 and1977 (50). Q fever had become endemic amongdairy herds throughout the United States by1960 (122). The presence of C. burnetii in milkand the knowledge that humans could becomeinfected by drinking such milk led to a critical

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from


Q FEVER AND C. BURNETII 129

study of conditions for thermal inactivation ofthe rickettsia. The pasteurization procedures ofthe time (143°F [61.7°C], 30 min) were based onthe thermal death point of Mycobacterium tu-berculosis and were inadequate for completekilling of C. burnetii. Heating contaminated milkat 161°F (71.70C) for 15 s killed all rickettsiae,and in 1956 the Assistant Surgeon General of theU.S. Public Health Service wrote to all state andterritorial milk control authorities, recommend-ing these conditions for milk pasteurization;such conditions were then generally adopted(63).More than 4 decades after the first reports of

the disease, Q fever has been found to bedistributed worldwide and remains a potentialhazard in almost any habitable region of theearth.

EPIDEMIOLOGY

Since its first description in Australia, Q feverhas been found throughout the world, except forthe Antarctic regions (96). The agent has beenidentified in arthropods, fish, birds, rodents,marsupials, and livestock. In the United States,Q fever has been reported to the Centers forDisease Control in Atlanta from 31 states (50).Although as of 1979 the disease was not on thelist of nationally notifiable diseases, 26 statesrequired notification of human cases and 5 oth-ers required the reporting of cases in animals(50). In a survey conducted by the Centers forDisease Control of the 1,164 Q fever casesreported between 1948 and 1977, 67% were fromCalifornia (50). No cases had been reported tothe Centers for Disease Control before 1948. It isnot surprising that relatively few cases of Qfever have been reported considering that thisdisease resembles other diseases such as viralinfluenza, the inability to cultivate the agent invitro, and the tendency on the part of cliniciansand public health officials to regard it as anunimportant disease. The actual number of cas-es far exceeds those reported. For example, Bellet al. (19) conducted a serological survey in theLos Angeles area and concluded that more than50,000 persons had been infected with the agentduring a 2-year span in the 1940s. During a 2-year span approximately 20,000 cases had oc-curred in Italy (11).The principal modes of entry of the agent into

humans are via inhalation of contaminated dustparticles and aerosols generated in the milieu ofabattoirs and dairies and the ingestion and han-dling of infected meat and milk (20, 22, 53, 129,187). The agent originates from the major reser-voirs of C. burnetii-dairy cows, sheep, andgoats. Infected cows shed enormous numbers ofrickettsiae in their milk, birth fluids, and colos-trum although appearing healthy (22, 91, 123,

187). In southern California, with up to 98% ofdairy herds seropositive for Q fever, 300 humancases were also reported, and 350 human caseswere associated with Q fever reservoirs in sheepin northern California. Similar correlations werefound in Pennsylvania (130). In 1974 82% oftested dairy cattle in California were seroposi-tive for C. burnetii, and the highest rate oc-curred in southern California. Rickettsiae wereshed in the milk of 51% of a large group of testedcows. The data represented a sevenfold increaseover a 25-year period (22). Similar studies ongoats in California in 1977 showed 24% seroposi-tive reactions, and 7% of the animals shed theorganisms in their milk (173). Q fever has alsobeen contracted by handling infected carcassesor placentas (123, 136, 163, 207) or laboratorymaterial (145, 216). The agent has been found inhuman placentas (69), presenting a hazard formidwives and in obstetric theatres, especially inunderdeveloped countries.

Babudieri (11) reviewed in detail the zoonoticaspects of Q fever. The agent naturally infectsover 40 species (including 12 genera) of ticksfound in five continents (11). Some additionalarthropods have been found infected with C.burnetii (11). Other arthropods, including fleas,lice, and cockroaches, have been experimentallyinfected; their role in the maintenance ofQ feverin nature is unknown. The role of ticks in thetransmission of the rickettsia is also not known,although they probably transmit the parasiteswithin domestic and wild animals and from wildanimals to domestic animals (11). Existence of akangaroo-tick cycle in western Queensland,Australia, was suggested by the discovery ofinfected kangaroos and infective kangaroo ticks,Amblyomma triguttatum, from kangaroos,goats, and sheep (164). Ticks were further impli-cated in outbreaks of Q fever among sheepshearers.

Several wild animals have been reported to beinfected with C. burnetii, including bandicootsin Australia (56, 70) and desert rats, wild rabbits,and mice in North Africa (11). The rickettsia hasbeen isolated from several bird species, includ-ing pigeons and sparrows in Europe and Asia,and in some cases from their ectoparasitic ticksand mites (11). Because most of the infectedbirds reported have close contact with humansand domestic animals, it is thought (84) that theyare not involved in the transmission of thedisease to humans, but are rather indicators ofthe agent in humans.

In none of the animals, wild or domestic, doesit appear that the Q fever agent causes overtdisease. In addition, the role of wild animals andarthropod vectors vis-a-vis disease in humans isprobably of minor significance.Q fever in domestic animals is of major impor-

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



tance and continues to be the principal source ofthe infection in humans. Although cattle, sheep,and goats are the main reservoirs, the agent hasbeen detected as well in dogs, camels, buffaloes,and geese and other fowl (11).Cases of transmission of the agent from hu-

man to human have been reported but are quiterare; the isolation of Q fever patients is notcalled for. In one rare case, Q fever was ac-quired by personnel participating in the necrop-sy of a fatal case of Q fever (129).Recent outbreaks of Q fever at research insti-

tutions underscore the importance of domesticanimals, specifically sheep, in the spread of theagent. In one outbreak involving 19 confirmedcases and 68 presumptive cases, it was suspect-ed that pregnant ewes being used for prenatalresearch were the source of infection (40). In asecond outbreak involving 81 persons, pregnantsheep also being used in prenatal research werethe source of infection (136). Transmission wasdirectly linked to rooms where the sheep wereheld and the corridors through which they weretransported. Affected individuals included per-sonnel working directly with infected placentasand individuals who walked past sheep held in acorridor.

CULTIVATION AND GROWTH OF C.BURNETII

Embryonated hens' eggs serve as excellenthosts for cultivating large numbers of the para-site. Typically, 5- to 7-day-old embryos areinoculated via the yolk sac and incubated at 35°Cfor an additional 10 to 12 days, during whichtime the rickettsiae proliferate luxuriantly withinthe yolk sac and sparsely in the rest of theembryo (146). Multiplication of C. burnetii wasstudied after inoculation into yolk sacs of em-bryonated eggs (121). At 11 days post-inocula-tion, highest numbers of rickettsiae were foundin the yolk sac membrane, and large numberswere also present in the embryo's intestines,chorioallantoic membrane, allantoamnion, andamniotic fluid. C. burnetii was present in lowernumbers in embryonic muscles, liver, spleen,and heart.

Purification of the rickettsiae from host com-ponents is a laborious process involving homog-enization of infected yolk sacs and differentialcentrifugation, followed by passage of the semi-purified organisms through density gradients.The essential steps have involved the use of theadsorbent Celite and trypsin treatment of therickettsial suspensions, followed by passagethrough sucrose gradients (157, 167, 192). Cen-trifugation through a Renografin density gradi-ent has also been used as a final step in purifica-tion (208, 210a). A 1- to 2-g (wet weight) amount

of purified rickettsiae can be obtained from 8dozen eggs.

Entry into and Proliferation Within CulturedCells; Comparison with Other Rickettsiae

C. burnetii has been cultivated in a variety ofanimal cells, both primary and established celllines. Some of these cells are chicken embryocells, L cells, mosquito cells, human embryofibroblasts, green monkey kidney (Vero) cells,tick tissue cultures, and the J774 and P388D1macrophage-like tumor cell lines (14, 38, 166,171, 206, 215).Within several of the cell lines, C. burnetii

establishes a persistent infection, i.e., dividinginfected populations that have been maintainedfor periods exceeding 1 year without the addi-tion of normal cells (14, 38, 168, 215; J. A.Stueckemann, Ph.D. thesis, University of Kan-sas, Lawrence, 1976).

Entry of the parasite into the cells is almostcertainly a passive event on the part of theparasite and occurs by phagocytosis (36). This isunlike other rickettsial species that actively pro-mote their entry into the host cell. The classicwork of Cohn et al. (44) showed that metabolicinhibitors caused a marked reduction in thepenetration of mouse cells by Rickettsia tsutsu-gamushi. Inactivation of R. tsutsugamushi byheat, UV irradiation, and Formalin also prevent-ed cell penetration. Inactivation of C. burnetiby heat or Formalin, however, did not affect itsuptake by mouse L-929 cells (0. G. Baca, per-sonal communication). Internalization of R.prowazeki into L cells required the active partic-ipation of both the rickettsia and the host cell(200). It has been proposed that internalizationof R. prowazeki occurs through a process of"induced phagocytosis": attachment of liverickettsiae to an unidentified site signals the cellto phagocytize (200). Inactive typhus rickettsiaeadhered to the cell membrane but internalizedslowly.

Studies on the nature of the C. burnetii attach-ment site on the L-929 cell surface showed thatonly pronase and subtilisin treatment of the Lcells significantly inhibited attachment of C.burnetii (Baca, personal communication). Thisprovided circumstantial evidence that proteinsare at or near the site of Coxiella attachment.Whether or not there is a specific component towhich the parasite attaches is as yet unknown. Itwas also found that phase II C. burnetii (see"Phase Variation" in "Biology of C. burnetii,"below) attached to the L cell much more readilythan phase I rickettsiae, probably accounting inpart for the ability ofphase II rickettsiae to morequickly infect various cells, including the L-929cell. Because lipopolysaccharide (LPS) is a viru-lence factor ofgram-negative bacteria, it is likely

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



that the LPS, which occurs in greater amounts inphase I than in phase II C. burnetii, plays a rolein hindering the entry of phase I Coxiella. In-deed, adding purified phase I LPS to C. burnetii(either phase) and L cells caused a slight reduc-tion in rickettsial attachment and markedly re-duced their entry. Whether or not the reducedentry was due to toxicity of the LPS is notknown.There are basic differences between C. burne-

tii and other rickettsial species with respect tolocalization within host cells. All members of thegenus Rickettsia grow within the cytoplasm ofcells with no apparent association with vacuolesor closely opposed host-derived membranes (5,183, 205). C. burnetii, however, proliferateswithin vacuoles (23, 36, 206). Electron micro-scopic studies of infected guinea pig liver re-vealed numerous vacuoles containing the rick-ettsiae (83). Occasionally, C. burnetii has beenobserved "free" in the cytoplasm (36, 83). Thepossibility that these free rickettsiae may have aclosely opposed host membrane has not beenexcluded. R. rickettsii and R. tsutsugamushimultiply primarily in the cytoplasm, free of hostmembranes (29, 174). These differences in local-ization may reflect different modes of entry. Therickettsiae mentioned, other than C. burnetii,may be entering by (i) direct penetration into thecytoplasm or (ii) via phagocytosis with subse-quent release into the cytoplasm or (iii) by acombination of the two. Immediately after inte-riorization within vacuoles, they may lyse thephagosomal membrane and escape into the cyto-plasm (for a review on the latter, see 212). Onceentry into the cell is achieved, both phases of C.burnetii proliferate within vacuoles which even-tually fuse and form a single vacuole whichoccupies most of the cell's volume (Fig. 1). Theinfected cell attains a size several diametersgreater than uninfected cells (14, 38); the nucle-us and cytoplasm become displaced to the cell'speriphery (38, 206; Stueckemann, Ph.D. thesis).When white mice were intranasally and intra-peritoneally infected, the rickettsiae grew withinvacuoles of the lung alveolar macrophages andof spleen reticular cells (8), resembling the rick-ettsial growth in guinea pig liver (83) and in Lcells (36, 38). Copelovici et al. (47) found similarintravacuolar growth in fibroblasts of humanembryonic cells and, on the basis of histochemi-cal reactions, showed that RNA content of nu-cleoli of infected cells increased. Ariel et al. (8)proposed that the vacuoles themselves, not therickettsiae, contributed to the pathogenicity andalso supported the phagolysomal origin of thevacuoles, proposing that vacuole formation wasa cellular defensive response to infection.

Cytochemical investigations revealed that therickettsiae-containing vacuoles are phagolyso-

somes (36, 38). Lysosomes labeled with elec-tron-dense thorium dioxide eventually fusedwith rickettsia-containing phagosomes, and acidphosphatase, a lysosomal enzyme, was detectedin the vacuoles. Recently evidence was ob-tained, using fluorescing lysosomotropic dyes,that the pH of the rickettsiae-containing vacu-oles is ca. 5.1, indicative of the presence of theacidic contents of the lysosome (E. Akporiayeand 0. G. Baca, unpublished data).

Enumeration

Predictably for a small obligate intracellularparasite, the enumeration of viable or total rick-ettsiae has presented technical problems. Anumber of methods have been developed, in-cluding the estimation of50% endpoints by usingembryonated eggs, direct microscopic counts ofacridine orange-stained rickettsiae mixed withknown amounts of marker bacteria (182, 184),and plaque assay (133, 151, 209). All of theseprocedures are tedious and time-consuming;those using embryos and cell cultures are diffi-cult to reproduce. The plaque assay technique,which utilizes primary chicken embryo cells forestimating viable C. burnetii, results in detect-able plaques only after 16 days of incubation.

BIOLOGY OF C. BURNETIIUltrastructure/Morphology

The etiological agent of Q fever is highlypleomorphic, coccobacillary in shape with ap-proximate dimensions of 0.3 by 1.0 ,um (35),bounded by an envelope similar to those foundin gram-negative bacteria (37, 143). The enve-lope consists of an outer and inner membrane,each approximately 6.5 nm thick. Sandwichedbetween the membranes and associated with theinner surface of the outer membrane is an elec-tron-dense layer which, although it has the ap-proximate dimensions and location of a typicalgram-negative peptidoglycan layer, is insensi-tive to lysozyme-EDTA treatment, as assessedby electron microscopy (37). Purified "cellwalls" from C. burnetii, however, dissolvewhen exposed to lysozyme (162). That peptido-glycan is a part of the C. burnetii envelope isstrongly suggested by several observations, in-cluding: solubilization of purified cell walls inthe presence of lysozyme and EDTA (162); thedemonstration of muramic acid (3) and diamino-pimelic acid (140) in whole cells, key compo-nents of typical peptidoglycan.Depending on the modification of the Gram

stain used, C. burnetii stains gram negative,gram positive, or gram variable (74). Use ofethyl alcohol-iodine as the mordant results in astrong gram-positive reaction; other rickettsialspecies stain gram negative. Nevertheless, ultra-

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



FIG. 1. L-929 cells in suspension culture (a) continuously infected with C. burnetii phase l, showingcytoplasmic intravacuolar proliferation, with L cell nuclei pushed to the edge of the cell (38). J774 cells inmonolayer culture (b), 27 days p.i. with C. burnetii phase I, showing cytoplasmic intravacuolar proliferation.Marker, 1.18 p.m (13).

structural and chemical analyses clearly indicatethat the agent is morphologically more similar togram-negative than to gram-positive bacteria.

Several workers have examined rutheniumred-stained phase I and phase II cells by electronmicroscopy. Burton et al. (37) detected a"fuzzy" layer approximately 20 nm thick on thesurface of both phases, whereas Ciampor et al.(43) found it only in phase I cells. The nature ofthe host may influence the apearance of such alayer, since in one case (37) the rickettsiae werepropagated in L cells and in the other (43) theywere propagated in yolk sacs.

GenomeRecently the genome size of C. burnetii was

determined by the method of DNA renaturation

to be approximately 1.04 x 109 daltons (138).This is comparable to the genome size of severalother rickettsial species and is well within therange of several free-living bacteria, includingMycoplasma spp. (5 x 108 daltons). Theoretical-ly, there appears to be sufficient coding informa-tion to allow axenic growth of the parasite; itsobligate nature is probably due to a few subtlegenomic lesions.The DNA base composition reported by sev-

eral laboratories (175, 185) is distinctly differentfrom that of other rickettsiae (196). Although wecontinue to refer to C. burnetii as a rickettsia, itis apparent that it differs from the true rickettsi-ae and calling it "rickettsia" is now due tocustom and convenience. We may speculate thatC. burnetii and the "true" rickettsiae, both of

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



FIG. 1 -Continuied

them obligate intracellular procaryote parasites,may actually represent an example of conver-gent evolution.

For some years rickettsiologists have beensearching for the existence of a plasmid in C.burnetii. Such a plasmid, consisting of 36 kilo-bases, has now been discovered by L. Malla-via's group in phase I C. burnetii and is present-ly being fully characterized (personalcommunication).

RibosomesThe first direct proof of the existence of

ribosomes in any of the rickettsial species wasreported in 1973 with the isolation of 70S parti-cles from C. burnetii that dissociated into 50Sand 30S subunits (15). This indicated that C.burnetii had conventional procaryotic proteinsynthesizing machinery, as suggested by earlierreports that cell-free preparations of the agentcould incorporate amino acids into a trichloro-acetic acid-insoluble fraction (125). This waspreceded by the extraction from whole rickettsi-al cells of RNA species comparable in size tothose found in procaryotic ribosomes, 16S, 23S,and 4 to 5S (192). More recently, immunologicalstudies have shown that Escherichia coli ribo-

somal structural proteins L7/L2 and elongationfactor EF-G have their counterparts in C. butr-netii (12). Furthermore, C. burnetii ribosomesand elongation factors are capable of function-ing, in vitro, with complementary componentsfrom E. coli (see "Biochemistry of C. buirnetii,"below). Two-dimensional gel electrophoresis ofCoxiella ribosomal proteins have revealed anarray of proteins with a distinctly different pat-tern than those of E. coli (12). These combinedstudies point out that this parasite's translationapparatus is unremarkably similar to that ofother well-studied procaryotes.

Electron microscopic examination of thin sec-tions of C. burnetii reveals a dense centralnucleoid region composed of fibers 2 nm indiameter (37, 114, 153, 189). The dense core,surrounded by a clear zone, is encompassed by agranular area containing ribosomes (37).

ReplicationAlthough the agent has never been directly

observed to divide, ultrastructural studies pro-vide ample evidence of binary fission: constric-tion of the equatorial region with the concomi-tant appearance of two nucleoid regions (4, 37,

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



143, 153, 186). Occasionally the organism ap-pears in chains, which is additional evidence ofbinary fission. Reports unsubstantiated in otherlaboratories have suggested the presence of avirus-like stage in the life cycle of C. burnetii(113, 114, 171). This claim has been made on thebasis of filtration and infection studies. Untilsubstantiated by others with concrete evidence,these reports must remain open to question.

Recently another, as yet unsubstantiated re-port has suggested the genesis of endospores in a"life cycle" of C. burnetii (132). This ultrastruc-tural electron microscopic report bases its con-clusion solely on the appearance of an electron-dense body at the pole of the rickettsiae. Thisattractive possibility, which would neatly corre-late C. burnetii's ability to resist various physi-cal-chemical treatments and the above-men-tioned "viral" stage, awaits substantive proofand confirmation in other laboratories, includingthe demonstration of viability, isolation and ger-mination, and putative thermodurability, andpossible presence of marker biochemicals suchas dipicolinic acid.

Several reports present electron microscopicevidence suggesting that C. burnetii exhibits twomorphological forms, a large and small form,distinct from the phase variation phenomenon(see below) (87, 132, 199, 208). When centri-fuged to equilibrium in cesium chloride or indensity gradients prepared with sucrose, Reno-grafin, or Ficoll, the cells separated into twodistinct bands, one containing a rod-shaped cellof low density and a second band containinglarger, round or coccobacillary-shaped cells thatbanded at higher density. Two hypotheses havebeen advanced to explain this: the large cellsresult from degeneration of the small cells (199,208), or both cell types represent forms of acomplex life cycle (132, 208). McCaul and Wil-liams (132) further speculated that the "large"form is the progenitor of the putative "spore"mentioned above. The so-called spore then pur-portedly germinates, forming the hardy smallercell which in turn differentiates into the largerless resistant cell form.

Phase VariationBased on immunological tests, C. burnetii

exhibits a phenomenon designated as "phasevariation" (30, 65, 188). The phases were orig-inally distinguished serologically. Injection oflive phase I or II C. burnetii into guinea pigsresulted in antibodies to both phases; phase IIagglutinins and complement-fixing antibodiesappeared first, later followed by phase I anti-body formation (24, 66). In nature or in labora-tory animals the parasite exists in the "phase I"state; repeated passage of phase I organisms

through embryonated eggs results in conversionto the "phase II" state. Injection of phase II C.burnetii into guinea pigs, mice, and hamstersresults in reversion to phase I (100). This phasevariation is apparently due to surface antigenicdifferences (66); the identity of the antigens arenow being identified and include, apparently, atoxic LPS (see below). Most certainly there areadditional phase-specific antigens (198). Phasevariation has been observed in infected cellcultures of L-929 cells and the macrophage-likecell line P388D1 (13, 38; Stueckemann, Ph.D.thesis). In both cell types, phase I C. burneticonverted to phase II (but not in toto) duringprolonged infection of the cell lines (severalmonths to over a year). Cloned phase I and IIrickettsiae were used in the experiments withthe P388D1 cells (13). Within these latter cells,phase II C. burnetii remained in phase II.The fundamental mechanism(s) for phase vari-

ation remains unknown. It may be due to inher-ent genetic differences between phase I and II C.burnetii (i.e., two genetically distinct popula-tions of cells, one of which becomes dominantwithin a host cell). Another possibility is that theorganism's gene expression is influenced by thehost's environment; i.e., factors inherent to thehost cell may suppress or induce expression ofthe organism's phase characteristics. A thirdexplanation for the variation may be due simplyto host lysosomal enzymatic modification of therickettsial surface, resulting in exposure or gen-eration or both of new antigenic determinants.The L-929 and macrophage cell lines may provevaluable in probing the basis of phase variations.

Morphologically the two phases are indistin-guishable, but they do exhibit several biologicaland chemical differences. Equilibrium densitygradient sedimentation in cesium chloride differ-entiates phase I (d = 1.33 g/cm3) and phase II (d= 1.22 g/cm3) organisms (87). Phase I cells aremore virulent for experimental animals than arephase II cells (99). Cytochemical techniquescoupled with electron microscopy indicate sur-face charge differences: phase II rickettsiae dis-play surface anionic binding sites which areabsent from the phase I organisms. The carbo-hydrate compositions and carbohydrate and pro-tein concentrations differ between phase I and IIenvelopes (92). Both envelopes possess glucoseand galactose, but only phase I envelopes con-tain glucuronic acid. Both phases possess agram-negative-type LPS, but there are distinctquantitative and qualitative differences with re-spect to sugar composition and total LPS con-tent (see below). Other differences between thetwo phases include spontaneous clumping of thephase II rickettsiae (68) and differential uptakeby phagocytes; phase II C. burnetii are taken upmore readily (102, 112; Baca, unpublished data).

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



LPS

Endotoxic LPS characteristic of gram-negativebacteria was first demonstrated and partiallycharacterized in phase I C. burnetii by Baca andParetsky (17, 18) and Chan et al. (41) andsubsequently by Schramek and Brezina (176).The LPS of phase II C. burnetii was latercharacterized (16, 177). Other workers had prob-ably extracted endotoxigenic material, but hadnot chemically or physiologically characterizedthe substance as endotoxin. Phase II C. burnetiicontains one-tenth the amount of extractableLPS found in phase I cells (16). Chemical char-acterization of the LPS of both phases revealedthe presence of a variety of sugars and fattyacids, including those characteristic of gram-negative bacterial endotoxic LPS: ,B-hydroxy-myristic acid and heptoses (17, 41, 176). Thepresence of glucosamine and fatty acids, includ-ing ,-hydroxymyristic acid, indicated a lipid Amoiety, the toxic entity of endotoxin. AntilipidA antibodies were elicited in rabbits, using mild-ly hydrolyzed LPS from C. burnetii (178). Bacaet al. (16) compared the chemical composition ofthe LPS's isolated from phases I and II C.burnetii. Phase II LPS was extracted fromcloned as well as noncloned phase II cells whichhad been serially passaged 95 times throughembryonated eggs. Complement fixation andimmunodiffusion tests indicated immunologicalidentity of both LPS species (152). Characteris-tic of other bacterial LPS, phase I and II LPScaused gelation of Limulus lysates; phase I LPSinduced DNA synthesis in guinea pig leuko-cytes, nonspecific resistance in mice to virulentCandida albicans, and the dermal Schwartzmanreaction (152, 176). Thirteen of the 14 sugarsfound in phase I LPS were also present in phaseII LPS, although some of the sugars were pres-ent in different proportions (i.e., glucose com-prised 20% of the total phase I neutral sugars,but only 2 to 3% of the phase II neutral sugars).The fatty acid profiles of phase I and II LPSwere identical. In contrast, Schramek andMayer (179) reported that phase II LPS had onlytwo of the nine sugars found in phase I LPS. Thedifferences between the two reports (16, 179)may be attributed to the passage history of therickettsiae. Schramek and Mayer (179) had iso-lated the LPS from phase II C. burnetii whichhad been serially passaged at least 163 times intheir laboratory and an uncertain number oftimes elsewhere. The compositional differencesbetween phase I and II LPS may reflect thedegree of transition of C. burnetii to phase II.Phase I LPS induced pathophysiological

changes in experimental animals similar to thoseassociated with endotoxins of gram-negativebacteria (18, 176). Some of these effects in

guinea pigs were liver enlargement, hyperther-mia, leukocytosis, increased levels of hepaticand serum cortisol, increased incorporation ofprecursors into hepatic rRNA and plasma pro-tein, loss of body weight, chicken embryo lethal-ity, and hypothermia in rats. Another physiolog-ical effect is that phase I LPS extract, incombination with a mycobacterial glycolipid,induced tumor regression in experimental ani-mals (104). Mice injected intraperitoneally withpurified and killed C. burnetii were then protect-ed from ascites tumor development resultingfrom injection of sarcoma 180 cells (101); therewas the possibility that the protective effect wasmediated by the LPS present in the killed C.burnetii.Tzianabos et al. (197) recently examined the

cellular fatty acid composition of several rickett-sial species and concluded that the fatty acidprofile of C. burnetii was clearly different fromthat of the other rickettsiae examined (R. ricket-sii, R. typhi, R. canada, and Rochalimaea quin-tana). Unlike the other rickettsiae, which con-tained straight-chain saturated and unsaturatedfatty acids, C. burnetii possessed large amountsof iso and anteiso branched-chain fatty acids.Interestingly, the C. biurnetii fatty acid profilewas strikingly similar to that of several Legion-ella species, but DNA hybridization studiesshowed no relatedness between these two or-ganisms.

BIOCHEMISTRY OF C. BURNETIIProblems attendant to a fuller knowledge of

the biochemistry of C. burnetii are largely com-ponents of the general problem of the nature ofhost-obligate parasitic interrelationships. Whatare the intrinsic enzymes of C. burnetii, andwhat are the host's contributions which permitproliferation of the parasite? C. burnetii prolifer-ates within phagolysosomal vacuoles in the ani-mal and in animal cells in culture (14, 36, 83, 105)(Fig. 1). Does the rickettsia use lysozymic prod-ucts within the vacuole as energy sources oranabolic precursors or both? How is the parasiteprotected inside this hostile degradative envi-ronment? Does the organism's cell envelopehave unusual substituents which regulate itspermeability?

Glycolytic and Related EnzymesEarly studies had shown that typhus rickettsi-

ae oxidized glutamate, pyruvate, and succinate(28, 47, 203) and coupled glutamate oxidationwith ATP synthesis (25). These rickettsiae incor-porated [35S]methionine and [14C]glycine froman axenic medium in reactions requiring (26, 27)NAD, ATP, and glutamate; the previously dem-onstrated oxidative phosphorylation explainedthe role of glutamate, and the endergonic syn-

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



thesis of protein accounted for the need of ATP.Probably the first studies on the biochemistry

and physiology of C. burnetii were those relatingto energetics and carbohydrate metabolism.ATPase and ADPase activities were found inintact cells by Paretsky et al. (157). Reactingsuch cells with oxalacetate and acetyl-coenzymeA in the presence of NAD and ATP yieldedsmall amounts of citrate; using cell-free prepara-tions from disrupted cells catalyzed a ninefoldincreased synthesis of citrate (157). The impliedrole of a membrane barrier to substrate permea-bility was strengthened when it was shown thatintact cells could not reduce NAD in the pres-ence of malate and glutamate, but that cell-freepreparations readily reduced NAD (157). Dis-rupted C. burnetii oxidized malate in the pres-ence of NAD. The oxidation was inhibited by p-aminobenzoic acid (PABA), but was reversed byexogenous NAD (80), by a proposed adductformation with PABA (78), but the phenomenonwas unexplained. The proposed adduct formedby PABA and NAD (80) provided a mechanismto explain PABA inhibition of typhus rickettsiaeon the basis of PABA binding endogenous ty-phus NAD, to form a PABA-NAD adduct.

Several laboratories had vainly attempted todemonstrate glucose oxidation by intact typhusrickettsiae, using intact organisms and classicmanometric and isotope techniques (28, 147,203). Weiss (203) attributed the apparent defi-ciency to the organism's lack of permeases orglycolytic enzymes. Intact C. burnetii also failedto oxidize glucose or glucose 6-phosphate, and itwas presumed that this apparent metabolic inac-tivity was due to a permeability problem ratherthan enzyme deficiency. However, disrupted C.burnetii preparations converted glucose to glu-cose 6-phosphate and oxidized the latter to 6-phosphogluconate (45, 156); isocitrate was alsooxidized. The data were not readily accepted;much of the skepticism was based on "unphy-siological conditions." The glycolytic data wereconfirmed and extended in L. P. Mallavia's lab-oratory where it was shown that cell-free prepa-rations of C. burnetii contained glucose 6-phos-phate isomerase, fructose 1,6-diphosphatase,aldolase, glyceraldehyde 3-phosphate dehydrog-enase enolase, and pyruvate kinase (134a). Itwas now clear that C. burnetii possessed most ofthe enzymes of glucose dissimilation, and theresults implied that the C. burnetii cell mem-brane differed in permeability properties fromthose of most free-living procaryotes. In addi-tion to the Meyerhof and Entner-Doudoroff gly-colytic enzymes, the presence of Krebs cycleenzymes was suggested by low-level oxidationsof a-ketoglutarate, succinate, fumarate, malate,oxalacetate, and pyruvate by intact C. burnetii(150).

Because C. burnetii cell walls contain gluco-samine and muramic acid (92), compounds madefrom glucose-derived precursors, and yet glu-cose seemingly could not be metabolized byintact rickettsiae, an apparently anomalous situ-ation existed. This question was partially re-solved (77). When disrupted C. burnetii werereacted with peptidoglycan precursor intermedi-ates, syntheses of acetyl-coenzyme A, UDP-N-acetylglucosamine, and UDP-N-acetylgluco-samine pyruvyl ether were demonstrated, butnot those of glucosamine or N-acetylglucosa-mine synthesis (77). The workers suggested thatif C. burnetii lacked the latter synthases, thenthe rickettsiae would depend on the host forsuch intermediates, explaining one factor ofobligate intracellular parasitism.

Anabolic Enzymes of Amino Acids, Proteins,and Nucleic Acids

The presence of a broad range of intrinsicrickettsial glycolytic enzymes (134) suggestedexistence of additional anabolic enzymes. Thepresence offolic acid in C. burnetii (139) impliedan active metabolic role for this cofactor. Afterdisrupted C. burnetii was incubated with[14C]glycine, ATP, formaldehyde, and exoge-nous tetrahydrofolate, [14C]serine was isolated(141). The folic acids of C. burnetii were isolatedand separated by DEAE-cellulose chromatogra-phy (131). Although several forms of folateswere similar to those of embryonated eggs, afolate unique to C. burnetii was found. No finalcharacterization could be made, but the coen-zyme had properties of a polyglutamate or prefo-lic A form. It could not be determined whetherthe folates were made by the organisms, orwhether they were of host origin and modifiedby the rickettsiae; the general resistance ofrickettsiae to sulfonamides suggested the latteralternative. Additional possible participation forrickettsial folates was found in the synthesis ofcitrulline from ornithine and carbamoyl phos-phate and of the pyrimidine precursors ureido-succinate and orotate from aspartate and car-bamoyl phosphate by disrupted C. burnetii(124). The presence of autonomous rickettsialnucleic acid biosynthesis was substantiated bysubsequent demonstration of RNA polymeraseactivity in C. burnetii (95). The polymerase wasDNA dependent, required the four ribonucleo-side (AGUC) triphosphates and exogenous ener-gy sources, and was inhibited by actinomycin Dand DNase. It was later shown that disrupted C.burnetii could synthesize the needed ribonucleo-side mono- and disphosphates in the presence ofATP or GTP (42). C. burnetii had sufficientendogenous ATP (2 nmol per mg of protein) forits in vitro nucleoside triphosphate synthesis,which would serve as substrates for coordinated

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



RNA synthesis. The workers proposed that dur-ing infection the rickettsiae obtained the neces-sary nucleoside phosphates from the lysosomalvacuoles of the host cell in which the parasiteproliferates. Pyrimidine nucleotide biosynthesishad also been demonstrated by Williams andPeterson (211) in typhus rickettsiae.

Bovarnick's group had reported incorporationof methionine and glycine into protein by intacttyphus rickettsiae (26, 27); the obligate intracel-lular parasitic rickettsiae clearly had endoge-nous cellular machinery for protein synthesis.Additional evidence for autonomous proteinsynthesis by rickettsiae was provided by demon-strating incorporation of amino acids into C.burnetii protein (125). The proposal that therickettsiae had endogenous protein-synthesizingmachinery was strengthened by the direct dem-onstration of rRNA in C. burnetii; this was thefirst clear evidence of this critical RNA class inany of the rickettsiae (192). This work wasfollowed by isolation and characterization of C.burnetii ribosomes (15) and was likewise the firstdirect and unequivocal demonstration of rickett-sial ribosomes. The ribosomes had typical chem-ical and physical procaryotic ribosomal struc-ture. Baca (12) later showed that isolated C.burnetii ribosomes in the presence of the postri-bosomal supernatant (S-100) catalyzed transla-tion of polyuridylic acid to polyphenylalanine.The rickettsial translation system had proteinswhich were antigenically similar to E. coli elon-gation factor G and to E. coli ribosomal proteinsL7 and L12. Autonomous protein synthesis byC. burnetii was more closely investigated byDonahue and Thompson (57-59), who comparedtranslation of polyuridylic acid and Q,3 phageRNA by rickettsial and E. coli cell extracts. C.burnetii had a greater optimum Mg2+ require-ment (17 versus 6 mM) for polyuridylic acidtranslation of Q1 phage RNA translation. Thekinetics of QB RNA translation were the samefor both bacterial systems, with a transit time of3 to 4 min for the coat cistron, and the phagecoat polypeptide product was also the same. Therickettsial extract had no demonstrable mRNA,but the workers proposed that the rickettsiaehad the required complement of aminoacyl-tRNA synthetases, tRNAs, ribosomes, and ribo-somal factors needed for initiation, elongation,and termination of translation. Donahue andThompson (58, 59) proposed that the lowernumber of ribosomes per rickettsia (comparedwith E. coli) could be one of the factors respon-sible for the rickettsial generation time.

Acidophilic Biochemistry and IntravacuolarExistence

The metabolic evidence cited above indicatedthat, although viable C. burnetii was apparently

impermeable to a wide variety of metabolites,the parasite nevertheless maintained itself andproliferated within its host phagolysomal vacu-oles. This perplexing paradox was resolved byHackstadt and Williams (81, 82), who showedthat glucose and certain key intermediates ofglycolysis were actively metabolized by intactC. burnetii at pH 4.5, but were relatively meta-bolically inert at pH 7.0!Analogous to oxidative phosphorylation with

glutamate by typhus rickettsiae (25), C. burnetiioxidized glutamate at pH 4.5, but not at 7.0,under axenic conditions, to generate an intracel-lular ATP pool stable for as long as 96 h. Theadenylate charge increased during glutamate ox-idation with parallel increase in adenylate poolsize. Rickettsial viability was directly correlatedwith the ATP pool, which in turn was morestable at pH 7.0 than at 4.5. Hackstadt andWilliams (81, 82) proposed that rickettsial ATPpool stability at pH 7.0, the host's cytoplasmicpH, enabled the rickettsiae to survive until theyentered the host's phagolysosomal vacuole.Here the acidic pH would permit rickettsialmetabolic activity to begin, thus providing a"biochemical stratagem for obligate parasitism. . .by C. burnetii" (81). The observations rec-onciled previous reports of extremely low levelsof oxidation of these metabolites by intact cellsat "physiological" pH (150) and of glycolysis bydisrupted but not by intact cells (45, 134, 156).The accommodation of C. burnetii to an activeexistence in the phagolysosomal vacuole with apH of 4.5 to 4.8 could now be explained, sub-stantiating earlier suggestions that C. burnetiiutilized lysozymal products in its existence (77).More recently an approach was made to the

question of intraphagolysosomal survival by thefinding that C. burnetii produced superoxideanion and had both superoxide dismutase andcatalase activities (2). In light of the Hackstadtand Williams reports (81, 82) it was of interest tonote that superoxide was produced at pH 4.5(the phagolysosomal environment) but not at pH7.4. On the other hand, catalase activity wasgreater at pH 7.0 than at pH 4.5. The rickettsialdismutase differed from its host L929 cells (2).Indeed the dismutase-catalase system may af-ford protection to the parasite not only from itsown toxic oxygenated metabolites but also fromthose generated by the host cell. Myers et al.(142) reported that R. prowazeki, which prolifer-ated intracytoplasmically, lacks catalase activityand also fails to produce hydrogen peroxide.Unlike C. burnetii, the typhus agent probablydoes not require catalase for survival because ofthe lack of hydrogen peroxide in its intracyto-plasmic environment (142). The host-indepen-dent biochemistry of C. burnetii is shown in Fig.2.

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



Glycine

THFA (141)

Serine

l(*)Amino acids

((125)

Protein

(58,59)RIA, tRNA, rRNA

Ribonucleic acids

(95)(42

ATPe(424) ATP

NDP XQ

(42) suox

NuCleoside mono-P

Orotote

(124)Ureidosuccirnate

(124)

Aspartote

!(124)Citrulline

(157) (157) U) (45)Malote ---0 Oxalocetate -. Citrate ---i--Citrate - aKetogl Itorate

- (45) (134)Glucose 6-P ----6-P Gluconate - - D-Ribulose 5-P

4>156 )Glucose

(134o)

Fructose 6-P

25/ I (1340)E/Fructose l,G-diP (134a), Glyceroldehyde-3-P -(134o)

GIuc Nhh 6-P - AQAc( 5 |

w*) I CoA-OAc -OAcP Pyruvotee (PEP)

1d-OAc-GI ucNH2 6-P1',......H-OAc-GlucNH2 I-P

UTP 1 (77)

UDP-N-OAc-Gl ucrJH2

ibst roteidotions (2)

X 2e2 202

2H+ (2)

H209 + 021 (2)

H20 + 1/2 02

*(77)- -UDP-N-OAc-GlucNH2-oyruvyl ether

(*)(ll4-OAc-Muronzic acid ------PePtidoglIycan )

FIG. 2. Host-independent biochemical activities of C. burnetli. Reactions as described in Literature Cited.*, Reactions not as yet demonstrated. THFA, Tetrahydrofolic acid.

PATHOBIOLOGY

The biology of C. burnetei and the pathobi-ology of Q fever have been studied less inten-sively than other aspects of the disease. Investi-gators have devoted more attention to theecology, immunology, and epidemiology of theinfection because of its public health and eco-nomic significance.The original observations of Derrick (54) and

of Burnet and Freeman (35) on the gross pathol-ogy of the Q fever described splenomegaly andlymphocytosis in humans and hepatomegaly andsplenomegaly in experimentally infected ratsand mice.

Humans

Endocarditis can occur as a sequela ofQ feverin humans, frequently in latent or chronic cases,often with attendant hepatomegaly and spleno-

megaly, and occasionally with cardiomegaly,glomerulonephritis, myocarditis, pericarditis,and cardiovascular lesions (39, 46, 64, 67, 75,107, 127, 169, 172). The rickettsiae themselveshave been isolated from the aortic valve (6, 107,169) as well as from spleen and lungs (6, 169),with cases reported from such diverse locales asEngland, the United States, South Africa,France, and Iran. The mitral valve was involvedas often as the aortic stenoses in endocarditisresulting from Q fever (194). Diagnosticians areagain urged to consider Q fever involvement inendocarditis cases where bacterial isolations inusual media are negative. It is said that where Qfever is endemic among farmers, abattoir work-ers, dairymen and others who work with live-stock, Q fever should be considered in thediagnosis of pericarditis (39). The typical Qfever-associated endocarditis case has been de-scribed (7) as a middle-aged male, with known

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



valvular heart disease (usually aortic) and with ahistory of an influenza-like infection 6 to 16months previously. Latent Q fever in humanshas been shown by the recovery of C. burnetiifrom placentas of women who had experiencedQ fever 3 years earlier and who had recoveredand were apparently normal (191).An analogous situation was found in guinea

pigs, where latent Q fever was reactivated by X-irradiation or cortisone (180, 181), demonstrat-ing that the disease may persist after its overtclinical features have subsided and that latentendemic foci can recrudesce if the carrier popu-lation is appropriately stressed.

Embryonated EggsEarly work on the effects of Q fever at the

cellular and subcellular levels used infectedchicken embryos. On the assumption that infec-tion with an obligate parasite led to utilization ofhost factors by the agent, vitamin concentra-tions were compared between normal and infect-ed embryonated 16-day-old chicken embryosand yolk sacs (139). Infected yolk sacs hadlesser amounts of folic acid (59%), riboflavin(61%), pyridoxine (67%), biotin (73%), and thia-min (85%) material; niacin (93%) and pantothen-ic acid (103%) levels were about similar and B12was 30% greater in infected material. Infectedembryos had lesser amounts of biotin (40%) andpyridoxine (65%); riboflavin (95%), thiamin(118%), and niacin (138%) had similar or greaterlevels than uninfected embryos. Oxidative phos-phorylation of infected embryonic livers wasless active than in uninfected embryos. Oxida-tive phosphorylation of glutamate and citrate byembryonic livers was inhibited by infection.Liver protein synthesis during chick embryogen-esis was measured in infected and uninfectedembryos of similar physiological age (125).Greater protein synthesis occurred in early (8-day) infected embryos, whereas uninfected em-bryos had greater synthesis after 10 days. Whenembryonic age was the parameter, infected em-bryos had greater synthesis to day 14. It shouldbe noted that the infected embryos' physiologi-cal development diminished progressively after10 days of embryogenesis. The work was ex-tended by chromatographic fractionation of thefolic acids of uninfected and infected yolk andyolk sac embryos (131). The folate elution pro-files of uninfected and infected tissues werequalitatively identical but infected tissues con-tained quantitatively less folate, the differencesbecoming larger during embryogenesis.The pleomorphic nature of C. burnetii was

recognized by early workers (52): large andsmall forms were also found in the cytoplasm ofinfected yolk sac cells, as were rarely occurring

"atypical" forms. They failed to react with C.burnetii antibody (4) and might have been the"intermediate" form reported by Rosenberg andKordova (171). Because of their rare occur-rence, it is also possible that the forms werecontaminants. In other studies of infected chick-en embryos (121), C. burnetii was found in allorgans of the embryo including the liver, heart,and muscles by 11 days postinfection (p.i.). Inmore detailed experiments, the infection pat-terns in chicken embryos of the virulent Tabani-dae-Kazakhstan and mildly virulent Apodemusmicrotii-Lauga strains of C. burnetii were com-pared (106). Both strains infected the yolk sacendodermal epithelial cells, and at 13 days p.i.the Tabanidae-Kazakhstan strain had infected100% of endodermal cells whereas the A. micro-tii-Lauga strain infected 60%. The rickettsiaeproliferated within phagolysosomes and in thecytoplasm, and infected cells had lost their gly-cogen granules. The presence of C. burnetiiwithin the cytoplasm was presumed to be due tovacuolar burst. The typical large and smallforms of the rickettsiae were observed, as wellas very small 30-nm forms. The latter could beinferred to have been the atypical forms de-scribed by Anacker et al. (4).

Six-day-old embryonated eggs infected withC. burnetii were assayed for glycolytic enzymesover a 10-day period p.i. (135). Until 6 days p.i.there were few differences between the enzymesof normal and infected embryos, but by 10 daysp.i. (16-day-old embryos) there were increasedspecific activities of aldolase (7-fold), phospho-fructokinase (7-fold), fructose 1,6-diphosphatase(7-fold), glucose isomerase (30-fold), glyceralde-hyde 3-phosphate dehydrogenase (12-fold), eno-lase (4-fold), and pyruvate kinase (10-fold). In-creased enzyme activities were attributed toinfection-stimulated hormonal (notably cortisol)activity (135) analogous to the responses ofguinea pigs to infection by C. burnetii and by itsLPS (18, 153, 154, 193).

Guinea Pigs and Other AnimalsThe ready susceptibility of guinea pigs to

infection by C. burnetii has made the animal auseful model for studying the pathobiology of Qfever. Guinea pigs infected with a low-virulencestrain of C. burnetii (Grita M-44) developed mildendocarditis, hepatitis, and granulomatic andnecrotic livers (93, 94). Animals infected withthe more virulent Nine Mile strain had elevatedblood glucose and greater activities in serum ofalkaline phosphatase, glutamic oxalacetic trans-aminase, and ,B-hydroxybutyric dehydrogenase3 to 5 days p.i.; creatine phosphokinase activityincreased both 3 and 10 days p.i. At 10 days p.i.hypoglycemia, hypophosphemia, splenomegaly,and hepatomegaly developed. Epicarditis with

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



myocardial lesions became visible at 6 days p.i.Experimentally induced respiratory infection inguinea pigs resulted in interstitial pneumoniaaccompanied by lesions of lungs, spleen, liver,and heart, lymphoreticular myocarditis, andsplenomegaly (165). Plasma copper, seromu-coids, and lysozyme concentrations increasedthree- to fourfold 8 to 10 days p.i., coincidentwith pyrexia. Plasma zinc as albumin-boundzinc decreased, probably reflecting a redistribu-tion of tissue zinc rather than an actual loss fromorgans. Plasma copper in the form of ceruloplas-min reflected changes in this protein, and theseromucoid increases were correlated directlywith the serum globulin levels. It was proposedthat changes in proteins and trace metals result-ed from factors released from macrophageswhich had interacted with the rickettsiae.A series of studies identified the nature of

some aspects of the pathobiology of Q fever inguinea pigs. Intraperitoneal infection of guineapigs with C. burnetii produced pyrexia within 24h, with pronounced splenomegaly and hepa-tomegaly by day 3 p.i. (158). Hepatomegaly wasmaximal at 84 to 96 h p.i., due chiefly to fattyinfiltration (158). Liver lipids increased by morethan 300%. The lipid classes were predominant-ly triglycerides, with lesser amounts of choles-terol and unesterified fatty acids (21), but withunchanged concentrations of phospholipids;plasma phospholipids increased. Lipase activityincreased simultaneously in the adipose depots,suggesting fat mobilization from the depots intothe plasma and liver. Hepatomegaly due largelyto steatogenesis is a regular feature of Q fever.Among the causes for fatty liver developmentcould be enhanced synthesis of lipids within theliver or transport of lipids into the liver withfailure to export the lipids, leading to lipidaccumulation. Impaired export in turn couldresult from defective membrang structure ordeficient chylomicrons. Since infected liversshowed no increased lipid synthesis (T. Haneyand D. Paretsky, unpublished data), the liverplasma membrane peptides were examined. Al-though the polypeptide profiles of uninfectedand infected membranes were qualitatively simi-lar, quantitative changes in several peptide spe-cies were found (126). Plasma membranes ofinfected liver incorporated about 50% as muchglucosamine as uninfected membranes, inferringlesser glycoprotein synthesis. Involvement ofmembrane changes in liver pathobiology was notshown, but did raise the possibility of defectivepermeability. Liver glycogen diminished 24 hp.i. and disappeared 48 h p.i. (157); simulta-neously, glycogen synthetase activity decreasedto 50% 84 h p.i. [Glycogen synthetase catalyzesthe transglucosylation reaction: UDP-glucose +(glycogen), = UDP + (glycogen),+1, leading to

glycogen synthesis.] The synthesis of UDP-glu-cose from its precursors (ATP + UDP = UTP;UTP + glucose-l-PO4 = UDP-glucose) was un-inhibited during Q fever (190). This inhibition oftransglucosylation may have been one of thefirst biochemical lesions demonstrated in Q fe-ver pathobiology. The largest loss of glycogensynthetase was in the rough endoplasmic reticu-lum of the liver, accompanied by a shift from theI form (glucose 6-PO4 independent) to the Dform (glucose 6-PO4 dependent) (153, 190, 195).

Additional pathobiochemical sequelae in theguinea pig were stimulated RNA and proteinsyntheses, using [3H]orotate and "4C-labeledamino acid incorporations as indices (190). Stim-ulated protein and RNA syntheses increasedprogressively during 96 h p.i., coincident withincreased hepatic cortisol concentrations (193).Enhanced synthesis of 28S, 18S, and 4S RNAspecies was parallel to increased numbers ofhepatic ribosomes (193), resembling sequelae inregenerating liver (118, 119) and in cortisone-stimulated DNA-dependent RNA polymeraseactivity of rat liver (214). Cyclic AMP increasedmore than 50% in infected liver (126). Furtherparallels between infected guinea pig liver andregenerating liver were found in the increasedconcentrations of polyamines during infection(160). (Polyamine synthesis also accompaniesnucleic acid synthesis in regenerating liver [118];this is but one example of a widespread directcorrelation of increased polyamine levels in rap-idly growing cells and tissues [90, 160].) It wasnow shown that polyamine concentrations in-creased in infected guinea pig liver, togetherwith increased activity of the liver RNA poly-merases (160), and that as infection progressedthere was a shift in the relative proportions ofpolymerases from class II to class I, the poly-merase which synthesizes rRNA. Ornithine de-carboxylase and S-adenosylmethionine decar-boxylase, enzymes which participate inpolyamine synthesis, had increased activitiesduring Q fever, and ornithine decarboxylasedecreased as polyamines increased, again simi-lar to the events in intoxicated and regeneratinglivers (90, 118). It was proposed (154, 159, 160)that polyamines were among the regulatory fac-tors of liver RNA synthesis in Q fever, perhapsby catalyzing nuclear protein phosphorylation(9, 89, 116). Paretsky et al. (160) proposed thatcortisol, which increased during Q fever, acti-vated ornithine decarboxylase, the initial en-zyme of polyamine synthesis, and also stimulat-ed class I polymerase as reported in rat liver(214).

Increased phosphorylation of histone andnonhistone chromatin proteins was subsequent-ly observed in nuclei of infected liver (F. Gon-zales, M. Halevy, and D. Paretsky, Abst. Annu.

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from


Q FEVER AND C. BURNETH 141

Meet. Am. Soc. Microbiol. 1982, H92, p. 128).These observations together with the demon-stration of increased nuclear RNA polymeraseactivity during Q fever (160) or after treatmentwith the rickettsial LPS (154, 155) are compati-ble with the report of increased phosphorylationof nuclear polyadenylic acid polymerase in hep-atoma (170). The findings may aid in identifyingthe mechanisms for stimulated transcription re-sulting from Q fever.When LPS isolated from C. burnetii (17, 41)

was intraperitoneally injected into the guineapig, the animal responded with hyperthermia,weight loss, hepatomegaly, and steatogenesis,elevated plasma and liver cortisol, increasedRNA, increased plasma and liver protein syn-thesis, and leukocytosis (18, 154). The responsesto LPS were characteristic of those resulting inactive Q fever in the guinea pig and stronglysuggested a direct role ofthe rickettsial LPS in Qfever pathobiology. The rickettsial LPS incubat-ed with L-929 cells produced qualitatively simi-larly increased RNA polymerase activities witha similar proportional shift to class I polymerase(155).Tumors were induced in guinea pigs with

hepatocellular carcinoma line 10 cells (104). For-malinized and purified suspensions of C. burne-tii caused tumor regressions in 32 to 42% of thecases, but were less effective than BCG-antitu-morigenic BCG preparations. Enhanced tumorregressions (63%) were obtained with combina-tions of mycobacterial glycolipids with eitherpurified, killed rickettsiae or C. burnetii LPS inoil droplet suspensions. The possible use of therickettsial LPS as an antitumorigen was suggest-ed.Hemolymph cells and organs of ticks infected

in vivo with C. burnetii "filterable particles"became typically vacuolated with cytoplasmicgranulations before appearance of the typicalrickettsial forms (171). In this connection Leykand Krauss (120) demonstrated infectivity of C.burnetii-infected yolk sac suspensions filteredthrough 100-nm pores but were unable to showinfectivity of such homogenates when filteredthrough 40-nm pores.

Cell CulturesMouse fibroblast L-929 cells in monolayer

culture were infected with C. burnetii, whichthen proliferated abundantly within phagolyso-somal vacuoles (36) analogous to the intravacuo-lar localization described for infected guinea pigliver (83). Histochemical demonstration of acidphosphatase and 5'-nucleotidase within the vac-uoles indicated stimulated lysozymal responseto infection. In further studies or cell cultureinfections (38), L-929 cells adapted to suspen-sion culture and Vero cells in monolayer culture

were persistently infected with C. burnetii.(Vero cells purportedly produced little or nointerferon.) The rickettsiae characteristicallyproliferated intravacuolarly. The engorged anddistended vacuoles occupied more than one-halfto two-thirds of the cell volume, with more thanl03 organisms per vacuole not uncommon (38).Vero cell vacuoles showed greater histochemi-cal evidence of acid phosphatase activity thandid the L cells, and the L cells in suspensionculture apparently had lesser activity than didthe infected monolayer L cells (38).Baca et al. (14) observed the fate of phase I

and II C. burnetii in several macrophage-likemurine tumor cells. Phase I rickettsiae estab-lished persistent infections in strain P388D1,J774, and PU-5-IR cells, but not in WEHI-3 andWEHI-274 cells; phase II rickettsiae parasitizedall five strains, and the rickettsiae proliferatedintravacuolarly. Of special interest was the con-version of phase I cells to phase II in P388D1cells similar to the observations in L-929 cells(Stueckemann, Ph.D. thesis).The plasma membrane peptides of L-929 cells

in suspension culture persistently infected withC. burnetii showed greater incorporation of[3H]glucosamine (glycoprotein synthesis?) thanuninfected cells (126). As in the case of guineapig liver membrane proteins, there were noqualitative differences between the peptide pro-files of uninfected and infected L cells, whereasseveral peptide species changed quantitatively.The membranes of infected L cells had 40%greater Na+,K+-ATPase activity. The resultswere consistent with those in infected liverplasma membranes (126).Phase I and II C. burnetii which infected

guinea pig macrophages grew within phagolyso-somal vacuoles (110, 111). The atypical formsseen by Anacker and others were not present.Kishimoto et al. stated that their data failed tosupport a C. burnetii replication cycle as pro-posed by other workers (110).

IMMUNOLOGYThe role of humoral and cellular immunity in

the control of Q fever has been extensivelystudied. It appears that, whereas both play a rolein eliminating C. burnetii from experimentalanimals, cell-mediated immunity is probably ul-timately responsible for eliminating the parasite,with specific antibodies accelerating the process(88).

Probably the first report indicating a role forantibody in the control of Q fever was that ofAbinanti and Marmion (1), who concluded thatmixtures of antibody and C. burnetii organismswere not infectious in experimental animals.Subsequent studies on the efficacy of Formalin-killed phase I and II C. burnetii vaccines in

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



humans and experimental animals indicated thatantibodies were involved in the resistance thatdeveloped against C. burnetii antigens (67, 117,148).Various studies subsequently revealed that

antibody promoted in vitro uptake of C. burnetiiby macrophages and polymorphonuclear leuko-cytes (33, 102, 110-112). Hinrichs and Jerrells(86) have provided convincing evidence thatcell-meditated immunity is involved in the con-trol of C. burnetii infection. Treatment of normalmacrophages with a lymphokine-containing su-pernatant resulted in growth inhibition of phago-cytized C. burnetii. The same laboratory pre-sented evidence that normal guinea pigperitoneal macrophages cultured in vitro withimmune lymphocytes from C. burnetii-infectedguinea pigs inhibited the growth of ingestedrickettsiae. Kelly (103) reported that killed C.burnetii activated, in vivo, guinea pig macro-phages to produce an extracellular factor whichkilled Listeria monocytogenes. Phase I LPS alsoactivated the macrophages in vitro. Kishimotoand Burger (108) reported that peritoneal macro-phages isolated from guinea pigs exposed to C.burnetii exhibited migration inhibition as earlyas 3 days postexposure and at a time when therewas no detectable circulating antibody to phase Iantigen. The same group showed that killedphase I C. burnetii vaccine induced a cell-mediated immune response in guinea pigs, asdetected by lymphocyte transformation and in-hibition of macrophage migration (109). Euthy-mic mice cleared C. burnetii from the peripheralcirculation and spleen within 14 days after expo-sure, whereas in athymic (T-cell-deficient) nudemice the rickettsiae were still present 60 dayspostexposure (109). Both the athymic and euthy-mic mice produced antibody against the rickett-siae. Humphres and Hinrichs (88) found thatantibody indeed altered the course of C. burnetiiinfection in mice by accelerating the initial inter-action of the inductive phase of the cellularimmune response, which subsequently promot-ed a more rapid development of the activatedmacrophages and, ultimately, control of the par-asite. Treatment of athymic mice with immuneserum 24 h before challenge with C. burnetii hadno effect on rickettsial multiplication within theirspleens.

Several studies have compared the uptake andfate of phase I and II C. burnetii by polymorpho-nuclear leukocytes and cells of the monocyticseries. Live or killed phase I rickettsiae werephagocytized to a lesser extent than phase IIorganisms by normal mouse or guinea pig perito-neal macrophages and polymorphonuclear leu-kocytes (33, 102). Phagocytosis was enhancedby rabbit immune serum containing antibodiesversus phase I; phase II antibodies had no such

enhancing effect. A slight increase in phagocyto-sis of phase II C. burnetii was observed withsera containing both antibodies. Similar resultswere obtained with normal and immune humanleukocytes and antibodies, but with no increaseduptake of phase II organisms when antisera wereincluded (213). Downs (60) reported that C.burnetii multiplied in both immune and nonim-mune monocytes, with phase I multiplying to agreater extent. Hinrichs and Jerrels (86) report-ed that C. burnetii was readily phagocytized bynormal guinea pig macrophages and that it wasnot destroyed, but rather proliferated in anddestroyed the macrophage. In the latter study,specific antiserum added to the macrophageculture before or after infection, or reacteddirectly with rickettsiae, failed to prevent intra-cellular infection. Kishimoto's group (110-112)examined the fate of C. burnetii in macrophagesand confirmed the observation by Kazar et al.(102) that phase I C. burnetii is more resistant tophagocytosis by macrophages from normal andimmune animals than are phase II organisms.Pretreatment of phase I and II rickettsiae withnormal serum did not affect their ability tomultiply and destroy macrophages from eithernormal or phase II-immunized guinea pigs. Incontrast, only phase I organisms were destroyedby macrophages from phage I-immunized ani-mals in the presence of normal serum. Apparent-ly the enhanced in vitro rickettsicidal capacity ofmacrophages from phase I-immunized animalsagainst homologous organisms was not attribut-able to enhanced lysosomal hydrolase activitysince no significant differences in the specificactivities of various lysosomal hydrolases wereobserved between macrophages obtained fromuninfected controls and phase I- or II-immu-nized animals. Immune serum potentiated thedestruction of both rickettsial phases in normalmacrophages; these results are opposite to thoseobtained by Hinrichs and Jerrels (86).Some of the apparently contradictory results

obtained by the different laboratories might bepartially reconciled by the use of phagocytesobtained directly from animals. A distinct prob-lem with macrophages derived from variousanimals is the functional heterogeneity exhibitedby such phagocytes (97, 201). Results obtainedwith such cells must be interpreted with caution.Subsequently, one of us (O.G.B.) recently ex-amined the fate and interaction of both phaseson C. burnetii in several macrophage-like tumorcell lines of murine origin. These cell linesexhibit population homegeneity and differentcapacities to interiorize particles and have beenextensively characterized with respect to phago-cytic capabilities (137, 202). With these celllines, it was determined that they exhibiteddifferent capacities to interiorize C. burnetii and

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



furthermore, although phase II C. burnetii in-fected all the cell lines tested (J774, P388D1, PU-5-IR, WEHI-3, WEHI-274) at a multiplicity ofinfection of 500, phase I Coxiella only infectedJ774, P388D1, and PU-5-IR (14). Only at multi-plicities of infection exceeding 3,000 did phase IC. burnetii infect the WEHI-3 cells (Baca, un-published data). Phase I antibody did acceleratethe entry of phase I rickettsiae into the cell lines(14); however, the antibody did not prevent theestablishment of a persistent infection (Baca,unpublished data). Although it appears that anti-bodies play a role in controlling Q fever, theymay also promote infection by accelerating en-try of the parasite into host cells.

VaccinesMarmion (128) and Kazar et al. (98) reviewed

the status of Q fever vaccines. Experimentalvaccines included killed and attenuated C. bur-netii and extracts of the agent. Unfortunatelymost, if not all, of these vaccines induced severelocal skin reactions. Phase I C. burnetii vaccineswere the most effective in providing protectionto challenge with live phase I rickettsiae. For-malin-killed phase I C. burnetii were shown tobe 100 to 300 times more effective in guinea pigsin eliciting antibody and protection to challengewith live organisms. An attenuated C. burnetiistrain designated Grita M44 and apparently inphase II was prepared by Genig (71) and testedin Soviet volunteers with success. However,recent work with the vaccine in guinea pigsindicated that the organisms persisted in theanimals for long periods of time and that theinfection could subsequently reactivate (93, 94).These observations sharply bring into questionthe use of such live vaccines in humans wherereactivation of previous infection has been re-ported (191). Extracts from C. burnetii have alsobeen tested in human volunteers. A trichloro-acetic acid extract of phase I C. burnetii causeda significant rise in phase I antibody titer inpersons tested; reactions to the vaccine weremore severe in those with a previous history ofQ fever (98). Significant progress in solvingproblems of toxic reactogenicity of C. burnetiivaccines was recently reported by Williams andCantrell (210). They demonstrated that vaccinesof killed phase I C. burnetii produced immunitybut also pathological responses in an endotoxinnonresponder strain of mice. However, whenkilled phase I rickettsiae were extracted withchloroform-methanol and the cell-free residue orthe chloroform-methanol extract was injectedinto the mice, no pathological responses wereinduced. Furthermore, the cell-free residue elic-ited antibody production against phases I and IIof C. burnetii and also protected mice againstsubsequent challenge by live rickettsiae. The

chloroform-methanol extract fraction failed toinduce either antibody or protection. The reportproposed that the pathology-producing factorsof whole-cell vaccines could be extracted withchloroform-methanol without impairing the im-munogenicity of the cell-free residue. This inter-esting report should contribute to the productionof an effective, nontoxic reactogenic vaccineagainst Q fever. Further progress in this promis-ing study is awaited to clarify the chemistry andbiological properties of the chloroform-methanolextract and to provide additional immunologicalinformation on the reactions of primates andother experimental animals to the cell-free resi-due. There are no commercial vaccines present-ly available for human use in the United States,and until the problems of pathological reacto-genicity are fully overcome a vaccine is unlikelyto become available for public use.

SUMMARY AND PROSPECTS"It is better to ask some of the questions than

to know all of the answers." (James Thurber,"The Scotty Who Knew Too Much," in Fablesfor Our Time, [Harper & Row, New York,1940]).

Forty-five years after Derrick and Burnet de-scribed Q fever and its rickettsioid etiologicalagent, the disease remains relatively obscure tothe general lay and scientific audiences, lackingthe dramatic appeals of the classic rickettsioses.Even for workers with Q fever, aspects of theimmunology and pathobiology of the infection,the biology of C. burnetii, even the taxonomicposition of the organism, and the molecularbasis of the parasite-host interrelationships re-main inadequately answered questions.

"Obligate intracellular parasitism" by its veryphrasing connotes a complete host dependencyby an organism. As the parasite's biochemical(enzyme, metabolite), physiological (cell enve-lope, intracellular organelles), physical andchemical (environmental pH and redox poten-tial) deficiencies become increasingly under-stood, and when such deficiencies are correctedby appropriate exogenous supplementations, the"parasite" becomes instead "fastidious." Whenmore complete information is obtained, axeniccultivation can be attained. It was said 10 yearsago, "One would be led to believe ... that C.burnetii is ideally suited for the study of meta-bolic reactions and may well serve as the proto-type of obligate intracellular bacteria. Unfortu-nately, nothing is further from the truth [italicsours]. Not only do intact resting cells displayvery little metabolic activity . .. (but) one mustconclude that resting cells of C. burnetii arerelatively inert" (203). A large complement ofanabolic and catabolic enzymes have now beenidentified in C. burnetii; many more will be

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



uncovered. The parasite has been shown to havean acidophilic metabolic permeability, compati-ble with its phagolysosomal existence. Thegenome has been extracted and its molecular.size (109 daltons) is of the same magnitude asthose of free-living procaryotes, opening theprospects of genomic analysis with the tools ofmolecular biology. This will add to our knowl-edge of the organism's biology. Finding a plas-mid in C. burnetii now opens experiments forgene transfer in this rickettsia. It is likely thatthe next decade will witness identification of thenature of C. burnetii's host dependence, fol-lowed by axenic cultivation of the organism.Solving the problem of obligate intracellularparasitism of C. burnetii will prove fruitful forwork with other rickettsiae and other obligateparasites. The biology of C. burnetii will befurther elucidated with the modem techniquesand information of molecular biology and bio-chemistry, including genome analysis and genet-ic recombination. Definition of the chemical,physical, and biological properties of the agent'scell envelope will help explain the basis for theintravacuolar habitat and proliferation of C. bur-netii and, by extrapolation, that of other intra-cellular parasites.The general strategies which procaryotic in-

tracellular parasiieshave evolved to thwart theirdestruction within macrophages and polymor-phonuclear leukocytes include prevention of fu-sion of the phagocytic vesicle with lysosomes(e.g., Mycobacterium tuberculosis, Chlamydiapsittaci) and escape from the phagocytic vacuoleinto the cytoplasm, thereby avoiding lysosomalattack (R. prowazeki). Other procaryotic intra-cellular parasites, including Salmonella spp.,Brucella spp., and C. burnetii, actually thriveand proliferate within the normally hostile pha-golysosomal milieu. The biochemical-biophysi-cal interactions between parasite and host whichresult in survival of the parasite are as yetlargely unknown. It may be speculated that C.burnetii has evolved a hydrophobic cell enve-lope enabling the parasite to resist attack bylysosomal enzymes. Recent demonstrations thatC. burnetii is permeable to metabolites at acidic,but not at physiological pH levels suggest thatrather than serving as a substrate for lysosomalenzymes, C. burnetii can actually utilize theproducts of lysosomal hydrolysis of cell constit-uents as precursor substrates for rickettsial me-tabolism. The host-parasite interactions leadingto cellular entry of C. burnetii will become betterunderstood as a result of continued experimentswith the rickettsia and different host cell lines inculture.

Q-fever-induced phosphorylation-dephos-phorylation of liver chromosomal substituents,polyamine synthesis, and hormone activation

correlate the stimulated hepatic transcriptionand translation of regenerating liver. Togetherwith glycogen depletion and steatogenesis in theinfected liver, the pathobiological events result-ing from Q fever have parallels in several pro-caryotic and eucaryotic infections, in regenerat-ing liver, hepatomas, tissue hypertrophies, andother pathologies. Q fever will continue to serveas an excellent model system for clarifying themolecular basis for many such phenomena andfor the study of general problems of host-para-site interrelationships.Because of the widespread distribution of its

endemic reservoirs and multiple routes of infec-tion, Q fever will likely remain a disease ofeconomic and public health importance in spiteof successful chemotherapy and preventivemeasures such as milk pasteurization. From amedical standpoint, latent infections and sequel-ae such as endocarditis should attract increasedattention as diagnosticians become more awareof the prevalence of Q fever.There are challenging questions to be an-

swered: how does C. burnetii enter the host celland participate in phagolysosomal vacuole for-mation? What nutrilites does the parasite obtainin the vacuole? What are the biochemical andbiophysical conditions necessary for axenic cul-tivation of the organism? What are the proper-ties of the rickettsia's cell envelope which con-tribute on the one hand to its permeability tometabolites at acidic pH values and on the otherhand to its resistance to lysosomal enzymes?What are the mechanisms which trigger develop-ment of host cell and organ pathologies? What isthe mechanism of phase conversion? What role,if any, does the plasmid play in the biology of C.burnetii? Can the parasite's genome be clonedso as to permit development of an effectivevaccine against Q fever, without producingharmful side effects? The next decade shouldwitness answers to some of these questions byusing the information and technology of modernbiochemistry and technology combined with thenecessary adjunct of ingenious experiments.

ACKNOWLEDGMENTS

We gratefully acknowledge support from grants PCM8010633 from the National Science Foundation and RR 08139from the Public Health Service, the Minority BiomedicalSupport Program of the National Institutes of Health (toO.G.B.), and Public Health Service grant Al 16954 from theNational Institutes of Health (to D.P.). Previous support fromthe National Science Foundation, the University of NewMexico, and the University of Kansas is also acknowledged.

LITERATURE CITED

1. Abinanti, F. R., and B. P. Marmbon. 1975. Protective orneutralizing antibody in Q fever. Am. J. Hyg. 66:173-195

2. AkporLaye, E. T., and 0. G. Baca. 1983. Superoxideanion production and superoxide dismutase and catalase

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



activities in Coxiella burnetii. J. Bacteriol. 154:520-523.3. Allison, A. C., and H. R. Perkins. 1960. Presence of cell

walls like those of bacteria in rickettsiae. Nature (Lon-don) 188:796-798.

4. Anacker, R. L., K. Fukusbi, E. G. Pickens, and D. B.lackman. 1964. Electron microscopic observations ofthe development of Coxiella burnetii in the chick yolksac. J. Bacteriol. 88:11301138.

5. Anderson, D. R., H. E. Hopps, M. F. Barle, and B. C.Bernheln. 1965. Comparison of the ultrastructure ofseveral rickettsiae, ornithosis virus, and mycoplasma intissue culture. J. Bacteriol. 90:1387-1403.

6. Andrews, P. S., and B. P. Marmlon. 1959. Morbidanatomical and bacteriological findings in a patient withendocarditis. Br. Med. J. 1:983.

7. Applefeld, M. M., L. M. Bfllngley, H. J. Tucker, and P.Fset. 1977. Q fever endocarditis-a case occurring in theUnited States. Am. Heart J. 93:669-670.

8. Ariel, B. M., T. N. Khavkln, and N. I. n a. 1973.Interaction between Coxiella burnetii and the cells inexperimental Q-rickettsiosis. Pathol. Microbiol. 39:412-423.

9. Atnar, V. J., and G. D. Kuehtn. 1981. Phosphorylation ofornithine decarboxylase by a polyamine-dependent pro-tein kinase. Proc. Natl. Acad. Sci. U.S.A. 78:5518-5522.

10. Babudieri, B. 1953. Epidemiology, diagnosis and prophy-laxis of Q fever, W.H.O. Monogr. Ser. 19:157-174.

11. Babudieri, B. 1959. Q fever: a zoonosis, Adv. Vet. Sci.5:81-154.

12. Baca, 0. G. 1978. Comparison of ribosomes from Cox-iella burnetii and Escherichia coli by gel electrophoresis,protein synthesis, and immunological techniques. J. Bac-teriol. 136:429-432.

13. Baca, 0. G., E. T. Akporiaye, A. S. Aragon, I. L.Marinez, M. V. Robles, and N. L. Warer. 1981. Fate ofphase I and phase II Coxiella burnetii in several macro-phage-like cell lines. Infect. Immun. 33:258-266.

14. Baca, 0. G., A. S. Aragon, E. T. Akporiaye, I. L.Martinez, and N. Warner. 1981. Interaction of Coxiellaburnetii with macrophage-like tumor cell lines, p. 91-101. In W. Burgdorfer and R. L. Anacker (ed.), Rickett-siae and rickettsial diseases. Academic Press, Inc., NewYork.

15. Baca, 0. G., R. T. Hersb, and D. Paretsky. 1973.Ribosomes and ribonucleic acids of Coxiella burneti. J.Bacteriol. 116:441-446.

16. Baca, 0. G., I. L. Martinez, A. S. Aragon, and D.Kase. 1980. Isolation and partial characterization oflipopolysaccharide from phase II Coxiella burnetii. Can.J. Microbiol. 26:819-825.

17. Baca, 0. G., and D. Paretsky. 1974. Partial chemicalcharacterization of a toxic lipopolysaccharide from Cox-ieila burneti. Infect. Immun. 9:959-961.

18. Baca, 0. G., and D. Paretsky. 1974. Some physiologicaland biochemical effects of a Coxiella burneti lipopoly-saccharide preparation on guinea pigs. Infect. jmmun.9:939-945.

19. Bell, J. A., M. D. Beck, and R. J. Huebner. 1950.Epidemiologic studies on Q fever in southern California.J. Am. Med. Assoc. 142:868-872.

20. Benson, W. W., D. W. Brock, and J. Mather. 1963.Serologic analysis of a penitentiary group using raw milkfrom a Q fever infected herd. Public Health Rep. 78:707-710.

21. Bernier, R. D., T. Haney, and D. Paretsky. 1974. Changesin lipids of liver and plasma during Q fever. Acta Virol.18:75-80.

22. Bbersteinb, E. L., D. E. Behymer, R. Bushnell, G. Cren-shaw, H. P. Ripmann, and C. E. Franti. 1974. A survey ofQ fever (Coxiella burnetii) in California dairy cows. Am.J. Vet. Res. 35:1577-1582.

23. Blackford, V. L. 1961. Influence of various metaboliteson growth of Coxiella burnetii in monolayer cultures ofchicken embryo entodermal cells. J. Bacteriol. 81:747-754.

24. Bobb, D., and C. M. Downs. 1962. The phase antigens ofCoxiella burnetii. Can. J. Microbiol. 8:869-700.

25. Bovarnlck, M. R. 1956. Phosphorylation accompanyingthe oxidation of glutamate by the Madrid E strain oftyphus rickettsiae. J. Biol. Chem. 220:353-361.

26. Bovarnk, M. R., and L. Schneider. 1960. The incorpo-ration of glycine-1-"C by typhus rickettsiae. J. Biol.Chem. 235:1727-1731.

27. Bovarnick, M. R., L. Schneider, and H. Walter. 1958.The incorporation of labeled methionine by typhus rick-ettsiae. Biochim. Biophys. Acta 33:414-422.

28. Bovarnlck, M. R., and J. C. Snyder. 1949. Respiration oftyphus rickettsiae. J. Exp. Med. 89:561-565.

29. Bozeman, F. M., H. E. Hopps, J. X. Danauskas, E. B.Jackson, and J. E. Smadel. 1956. Study on the growth ofrickettsiae. I. A tissue culture system for quantitativeestimations of Rickettsia tsutsugamushi. J. Immunol.76:475-487.

30. Brezina, R. 1958. Contribution to the study of phasevariation in Coxiella burnetii. Acta Virol. 2:91-102.

31. Brezina, R. 1969. Advances in rickettsial research. Curr.Top. Microbiol. Immunol. 47:20-39.

32. Brezina, R., andJ. Kazar. 1963. Phagocytosis of Coxiellaburneti and the phase variation phenomenon. Acta Virol.7:476.

33. Brezina, R., and J. Kazar. 1965. Study of the antigenicstructure of Coxiella burneti. IV. Phagocytosis and opso-nization in relation to the phases of C. burnetii. ActaVirol. 9:268-274.

34. Burgdorfer, W., and R. L. Anacker (ed.). 1981. Rickettsi-ae and rickettsial diseaes. Academic Press, Inc., NewYork.

35. Burnet, F. M., and M. Freeman. 1937. Experimentalstudies of the virus of Q-fever. Med. J. Aust. 2:299-305.

36. Burton, P. R., N. Kordova, and D. Paretsky. 1971.Electronmicroscopic studies of the rickettsia Coxiellaburneti: entry, lysosomal response, and fate of rickettsialDNA in L-ells. Can. J. Microbiol. 17:143-158.

37. Burton, P. R., J. Stuckemann, and D. Paretsky. 1975.Electron microscopy studies of the limiting layers of therickettsia Coxiella burnetii. J. Bacteriol. 122:316-324.

38. Burton, P. R., J. Stueckeman_, R. M. Welsh, and D.Paretsky. 1978. Some ultrastructural effects of persistentinfections by the rickettsia Coxiella burnetii in mouse Lcells and green monkey kidney (Vero) cells. Infect.Immun. 21:556-566.

39. Caughey, J. E. 1977. Pleuropericardial lesion in Q fever.Br. Med. J. 1:1447.

40. Centers for Dbease Control. 1979. Q fever at a universityresearch center-California. Morbid. Mortal. WeeklyRep. 28:333-334.

41. Chan, M. L., J. McChesney, and D. Paretsky. 1976.Further characterization of a lipopolysaccharide fromCoxiella burnetii. Infect. Immun. 13:1721-1727.

42. Christian, R. G., and D. Paretsky. 1977. Synthesis ofribonucleotides and their participation in ribonucleic acidsynthesis by Coxiella burnetii. J. Bacteriol. 132:841-846.

43. Ciampor, F., S. Sdcraek, and R. Brezina. 1972. Elec-tron microscopy of ruthenium red-stained phase I and IICoxiella burnetii. Acta Virol. 16:503-506.

44. Cohn, Z. A., F. M. Boenan, J. M. Campbell, J. W.Humphries, and T. K. Sawyer. 1959. Study on growth ofrickettsiae. V. Penetration of Rickettsia tsutsugamushiinto mammalian cells in vitro. J. Exp. Med. 109:271-292.

45. Consigi, R. A., and D. Paretsky. 1962. Oxidation ofglucose 6-phosphate and isocitrate by Coxiella burnetii.J. Bacteriol. 33:206-207.

46. Constanidis, K., and J. P. R. Jenkins. 1979. Chronic Qfever endocarditis. Pract. Med. 222:533-536.

47. Copelovicl, Y., P. Athanasiu, and D. Strulovlic. 1965.Cercetari asupra modificarilor citrochimice din celulefibroblastice de embrion uman inoculate cu Rickettsiaburnetii. Stud. Cercet. Inframicrobiol. 16:311-315.

48. Cox, H. R. 1939. Studies of a filter-passing infectiousagent isolated from ticks. V. Further attempts to culti-

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



vate in cell-free media. Suggested classification. PublicHealth Rep. 54:1822-1827.

49. Cracea, E., S. Dumitrescu-Constantinescu, D. Botez, andL. Icanld. 1977. Q-fever soluble vaccine effects in Cox-iella burneti sensitized humans. Zentralbl. Bakteriol.Parasitenkd. Infektionskr. Hyg. Abt. 1 238:413-418.

50. D'Angelo, L. J., E. F. Baker, and W. Schloer. 1979. Qfever in the United States, 1948-1977. J. Infect. Dis.139:613-615.

51. D'Angelo, J. J., and R. Hetherington. 1979. Q fevertreated with erythromycin. Br. Med. J. 2:305-306.

52. Davs, G. E., and H. R. Cox. 1938. A filter-passinginfectious agent isolated from ticks. I. Isolation fromDermacentor andersoni, reactions in animals, and filtra-tion experiments. Public Health Rep. 53:2259-2282.

53. DeLay, P. D., E. H. Lennette, and K. B. DeOme. 1950. Qfever in California. J. Immunol. 65:211-220.

54. Derrick, E. H. 1937. Q fever, a new fever entity: clinicalfeatures, diagnosis and laboratory investigations. Med.J. Aust. 2:281-305.

55. Derrick, E. H. 1973. The course of infections withCoxiella burneti. Med. J. Aust. 1:1051-1057.

56. DerrIck, E. H., and D. J. W. Smith. 1940. Studies in theepidemiology of Q fever. 2. The isolation of three strainsof Rickettsia burnetii from the bandicoot Isoodon toro-sus. Aust. J. Exp. Biol. Med. Sci. 18:99-102.

57. Donahue, J. P., and H. A. Thompson. 1980. Proteinsynthesis in cell-free extracts of Coxiella burnetii. J.Gen. Microbiol. 121:293-302.

58. Donahue, J. P., and H. A. Thompson. 1981. Subcellularcomponents from Coxiella burnetii synthesize coliphageprotein, p. 441-451. In W. Burgdorfer and R. L. Anacker(ed.), Rickettsiae and rickettsial diseases. AcademicPress, Inc., New York.

59. Donahue, J. P., and H. A. Thompson. 1981. Efficacy oftranslation in the rickettsia Coxiella burnetii. J. Bacte-riol. 146:808-812.

60. Downs, C. M. 1968. Phagocytosis of Coxiella burnetii,phase I and phase II by peritoneal monocytes fromnormal and immune guinea pigs and mice. Zentralbl.Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1206:329-343.

61. Dyer, R. E. 1938. A filter-passing infectious agent isolat-ed from ticks. IV. Human infection. Public Health Rep.53:2277-2282.

62. Dyer, R. E. 1939. Similarity of Australian "Q" fever anda disease caused by an infectious agent isolated fromticks in Montana. Public Health Rep. 54:1229-1238.

63. Enright, J. B., W. W. Sadler, and R. C. Thomas. 1957.Thermal inactivation of Coxiella burnetii and its relationto pasteurization of milk. Public Health Monogr. 47:1-27.

64. Fergson, I. C., J. E. Craik, and N. R. Grist. 1962.Clinical, virological, and pathological findings in a fatalcase of Q Fever endocarditis. Br. J. Clin. Pathol. 15:235-241.

65. Fiet, P. 1957. Phase variation of Rickettsia (Coxiella)Burneti-study of the antibody response in guinea pigsand rabbits. Can. J. Microbiol. 3:435-445.

66. Fet, P. 1964. Serological techniques, p. 225-255. InR. J. C. Harris (ed.), Techniques in experimental virolo-gy. Academic Press, Inc., New York.

67. Fbset, P., and R. Onubee. 1968. The antibody responseto antigens of Coxiella burnetii. Zentralbl. Bakteriol.Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 206:321-329.

68. Fiset, P., A. Ormsbee, R. Silberman, M. Peacock, andS. H. Spelmhan. 1969. A microagglutination technique fordetection and measurement of rickettsial antibodies.Acta Virol. 13:60-6.

69. Flset, P., C. L. W_seman, and Y. El. Batawl. 1975.Immunologic evidence of human fetal infection withCoxiella burneti. Am. J. Epidemiol. 101:65-69.

70. Freeman, M., E. H. Derrick, H. E. Brown, D. J. W.Smith, and D. W. Johnson. 1940. Studies in the epidemi-

ology of Q fever. 5. Surveys of human and animal serafor Rickettsia burnetdi agglutinins. Aust. J. Exp. Biol.Med. 18:193-200.

71. Genig, V. A. 1968. A liver vaccine 1/M-44 against Q-feverfor oral use. J. Hyg. Epidemiol. Microbiol, Immunol.12:265-273.

72. Gilman, A. G., L. S. Goodman, and A. Gilman. 1980. Thepharmacological basis of therapeutics, p. 1092, 6th ed.Macmillan Publishing Co., Inc., New York.

73. Gimenez, D. F. 1964. Staining rickettsiae in yolk saccultures. Stain Technol. 30:135-140.

74. Gimenez, D. F. 1965. Gram staining of Coxiella burnetii.J. Bacteriol. 90:834-835.

75. Giroud, P. 1976. Les lesions cardio-vasculaires des rick-ettsioses et des chlamydioses (Neorickettsioses). An-geioloe 28:175-182.

76. Gbroud, P., and M. Capponi. 1966. La fievre Q. Ed.Medic. Flammarion, Paris.

77. Gonzales, F. R., and D. Paretsky. 1981. Some autono-mous enzymatic reactions of Coxiella burnetdi in theinitial phase of peptidoglycan synthesis, p. 493-502. InW. Burgdorfer and R. L. Anacker (ed.), Rickettsiae andrickettsial diseases. Academic Press, Inc., New York.

78. Greff, D., and H. PInkerton. 1944. Effect of enzymeinhibitors and activators on the multiplication of typhusrickettsiae. I. Penicillin, p-aminobenzoic acid, sodiumfluoride and vitamins of the B group. J. Exp. Mea.80:561-574.

79. Grove, D. I. 1980. Q fever. Austr. Fam. Physician 9:680-682.

80. Guardiola, A. L., and D. Paretsky. 1958. Metabolicreactivation of rickettsiae by diphosphopyridine nucleo-tide. Science 128:141-142.

81. Hacktadt, T., and J. C. Williams. 1981. Biochemicalstratagem for obligate parasitism of eukaryotic cells byCoxiella burnetii. Proc. Natl. Acad. Sci. U.S.A.78:3240-3244.

82. Hackstadt, T., and J. C. Williams. 1981. Stability of theadenosine 5'-triphosphate pool in Coxiella burnetii: influ-ence of pH and substrate. J. Bacteriol. 148:419-425.

83. Handley, J., D. Paretsky, and J. Stueckemann. 1967.Electron microscopic observations of Coxiella burnetii inthe guinea pig. J. Bacteriol. 94:263-267.

84. Hart, R. J. C. 1973. The epidemiology of Q fever.Postgrad. Med. J. 49:535-538.

85. Heggers, J. P., L. H. Billups, D. J. Hlnbchs, and L. P.Malravla. 1975. Pathophysiologic features of Q fever-infected guinea pigs. Am. J. Vet. Res. 36:1047-1052.

86. Hhnrics, D. J., and T. R. Jerrels. 1976. In vitro evalua-tion of immunity to Coxiella burnetdi. J. Immunol.117:996-1003.

87. Hoyer, B. H., R. A. Ormsbee, P. Fbet, and D. B.Lackman. 1963. Differentiation of phase I and phase IICoxiella burnetii by equilibrium density gradient sedi-mentation. Nature (London) 197:573-574.

88. Hu _Pres, R. C., and D. J. H bnrics. 1981. Role forantibody in Coxiella burnetii infection. Infect. Immun.31:641-645.

89. Imad, H., M. Shimoyama, S. Yamamoto, Y. Tanlgwa,and I. Ueda. 1975. Effect of polyamines on phosphoryla-tion of non-histone chromatin proteins from hog liver.Biochem. Biophys. Res. Commun. 66:856-862.

90. Jinne, J., H. P6s6, and A. Raina. 1978. Polyamines inrapid growth and cancer. Biochim. Biophys. Acta473:241-293.

91. JeflIson, W. L., R. J. Huebner, M. D. Beck. R. R. Parker,and E. J. Bell. 1948. Q fever studies in southern Califor-nia. VIII. Recovery of Coxiella burneti from butter madefrom naturally infected and unpasteurized milk. PublicHealth Rep. 63:1712-1713.

92. Jerrels, T. R., D. J. Hlnrhs, and L. P. Maavia. 1974.Cell envelope analysis of Coxiella burneti phase I andphase II. Can. J. Microbiol. 20:1465-1470.

93. Johmon, J. W., G. A. Eddy, and C. E. Pedersen, Jr.1976. Biologic properties of the M-44 strain of Coxiella

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



burneti. J. Infect. Dis. 133:334-338.94. Johnson, J. W., C. G. McLeod, J. L. Stookey, G. A.

Hligbee, and C. E. Pedersen, Jr. 1977. Lesions in guineapigs infected with Coxiella burnetii strain M-44. J. Infect.Dis. 135:995-998.

95. Jones, F., Jr., and D. Paretsky. 1967. Physiology ofrickettsiae. VI. Host-independent synthesis of polyribo-nucleotides by Coxiella burnetii. J. Bacteriol. 93:1063-1068.

96. Kaplan, M. M., and P. Bertagna. 1955. The geographicaldistribution of Q fever. Bull. W. H. 0. 13:829-860.

97. Kavad, M., J. Laczko, and B. Csaba. 1979. Functionalheterogeneity of macrophages. Immunology 36:729-732.

98. Kazar, J., R. Brezina, A. Palanova, B. Tvrda, and S.Schramek. 1982. Immunogenicity and reactogenicity of aQ fever chemovaccine in persons professionally exposedto Q fever in Czechoslovakia. Bull. W.H.O. 60:389-394.

99. Kazar, J., R. Brezdua, S. Schraek, J. Urvo4yi, V.Posplisl, and E. Kovacova. 1974. Virulence, antigenicproperties and physicochemical characteristics of Cox-iella burneti strains with different chick embryo yolk sacpassage history. Acta Virol. 18:434-442.

100. Kazar, J., R. Brezina, E. Skultetyova, E. Kovacova, andS. Schramek. 1975. Phase II to phase I conversion ofCoxiella burneti in immuno-suppressed mice. ActaVirol. 19:359-363.

101. Kazr, J., and S. Sdcramek. 1979. Inhibition by Coxiellaburnetii of ascites tumor formation in mice. Acta Virol.23:267-270.

102. Kazar, J., E. Skultetyova, and R. Brezina. 1975. Phago-cytosis of Coxiella burneti by macrophages. Acta Virol.19:426-431.

103. Kelly, M. T. 1977. Activation of guinea pigs macro-phages by Q fever rickettsiae. Cell. Immunol. 28:198-205.

104. Kely, M. T., D. L. Grangw, E. Ribi, K. C. Mler, S. M.Strain, and H. G. Stoenner. 1976. Tumor regression withQ fever rickettsiae and a mycobacterial glycolipid. Can.Immunol. Immunother. 1:187-191.

105. Khavkin, T. N., B. M. Ariel, N. I. Amosenkova, and F. I.Krasnlk. 1975. Interaction of Rickettsia prowazeki withphagocytes in the course of infectious process. Exp.Mol. Pathol. 22:417-429.

106. Khavkin, T., V. Sukbnin, and N. Anosenkova. 1981.Host-parasite interaction and development of infraformsin chicken embryos infected with Coxiella burnetii viathe yolk sac. Infect. Immun. 32:1281-1291.

107. Klmbrough, R. C., HI, R. A. Ornsbee, and M. G.Peacock. 1981. Q fever endocarditis: a three and one-halfyear follow-up, p. 125-132. In W. Burgdorfer and R. L.Anacker (ed.), Rickettsiae and rickettsial diseases. Aca-demic Press, Inc., New York.

108. Ish , R. A., and G. T. Burger. 1977. Appearance ofcellular and humoral immunity in guinea pigs after infec-tion with Coxiella burnetii administered in small-particleaerosols. Infect. Immun. 16:518-521.

109. Kbno, R. A., H. Rozmlarek, and E. W. Larson.1978. Experimental Q fever infection in congenitallyathymic nude mice. Infect. Immun. 22:69-71.

110. K , R. A., B. J. VetrI, P. G. Cannc, F. G.Shirley, and J. S. Walker. 1976. Electron microscopicstudy on the interaction between normal guinea pigperitoneal macrophages and Coxiella burnetii. Infect.Immun. 14:1087-1096.

111. Klshloto, R. A., B. J. Veltri, F. G. Shirley, P. G.Canonico, and J. S. Walter. 1977. Fate of Coxiellaburnetii in macrophages from immune guinea pigs. In-fect. Immun. 15:601-607.

112. Kshimoto, R. A., and J. S. Walker. 1976. Interactionbetween Coxiella burnetii and guinea pig peritonealmacrophages. Infect. Immun. 14:416-421.

113. Kordova, N. 1959. Filterable particles of Coxiella bur-neti. Acta Virol. 3:25-26.

114. Kordova, N., and J. Rehacek. 1959. Experimental infec-tion of ticks in vivo and their organs in vivo with filterable

particles of Coxiella burneti. Acta Virol. 3:201-209.115. Krauss, H., H. Schewfer, and H. Schmatz. 1977. Ultra-

structural investigations on surface structures involvedin Coxiella burnetii phase variation. Infect. Immun.15:890-8%.

116. Kuehn, G. D., H.-A. Affolter, V. J. Atmar, T. Seebeck. U.Gubler, and R. Braun. 1979. Polyamine-mediated phos-phorylation of a nucleolar protein from Physarum poly-cephalum that stimulates rRNA synthesis. Proc. Natl.Acad. Sci. U.S.A. 76:2541-2545.

117. Lackmnan, D. B., R. N. Philip, E. A. Casper, J. F. Bell,J. B. Enrlght, and E. G. Plckens. 1967. Q fever immunityin man. Health Lab. Sci. 4:236-244.

118. Lesch, R., and W. Reutter (ed.). 1975. Liver regenerationafter experimental injury. Stratton Intercontinental Med-ical Book Corp., New York.

120. Leyk, W., and H. Krauss. 1974. Zum Nachweis vonsubmikroskopischen, vermehrungsfahigen Einheitenvon Coxiella burnetii. Dtsch. Tieraerztl. Wochenschr.81:528-530.

121. Leyk, W., and R. Scheffier. 1975. Multiplication ofCoxiella burned in different tissues and organs of embry-onated eggs. Zentralbl. Bakteriol. Parasitenkd. Infek-tionskr. Hyg. Abt. 1 231:519-524.

122. Luoto, L. 1969. Report on the nationwide occurrence ofQ fever infections in cattle. Public Health Rep. 75:135-140.

123. Luoto, L., and R. J. Huebner. 1950. Q fever studies insouthern California. IX. Isolation of Q fever organismsfrom parturient placentas of naturally infected diarycows. Public Health Rep. 65:541-544.

124. Mallavia, L., and D. Paretsky. 1963. Studies on thephysiology of rickettsiae. V. Metabolism of carbamylphosphate by Coxiella burnetii. J. Bacteriol. 86:232-238.

125. MallavIa, L. P., and D. Paretsky. 1967. Physiology ofrickettsiae. VII. Amino acid incorporation by Coxiellaburnetii and by infected hosts. J. Bacteriol. 93:1479-1483.

126. Mareckli, N., F. Becker, 0. G. Baca, and D. Paretsky.1978. Changes in liver and L-cell plasma membranesduring infection with Coxiella burnetii. Infect. Immun.19:272-280.

127. Marmion, B. P. 1962. Subacute rickettsial endocarditis:an unusual complication of Q fever. J. Hyg. Epidemiol.Microbiol. Immunol. 6:79-84.

128. Marmion, B. P. 1967. Development of Q fever vaccines,1937 to 1967. Med. J. Aust. 54:1074-1078.

129. Marnlon, R. A., and M. G. P. Stoker. 1950. Q fever inBritain. Epidemiology of an outbreak. Lancet iu:611-616.

130. Marshak, R. R., J. Melbin, and M. J. Herman. 1961.Study of Q fever in animals and man in Pennsylvania.Am. J. Public Health 51:1189-1198.

131. Mattbels, M. S., M. Silverman, and D. Paretsky. 1963.Studies on the physiology of rickettsiae. IV. Folic acidsof Coxiella burnetii. J. Bacteriol. 85:37-41.

132. McCaul, T. F., and J. C. Williams. 1981. Developmentalcycle of Coxiella burnetdi: structure and morphogenesisof vegetative and sporogenic differentiations. J. Bacter-iol. 147:1063-1076.

133. McDade, J. E., and P. J. Gerone. 1970. Plaque assay forQ fever and scrub typhus rickettsiae. AppI. Microbiol.19:963-965.

134. McDonald, T. L., and L. Maflava. 1970. Biochemistry ofCoxiella burnetii: 6-Phosphogluconic acid dehydroge-nase. J. Bacteriol. 102:1-5.

134a.McDonald, T. L., and L. Mallavia. 1971. Biochemistry ofCoxiella burnetii: Embden-Meyerhof pathway. J. Bac-teriol. 107:864-860.

135. McDonald, T. L., and L. Malavia. 1971. Host responseto infection by Coxiella burneti. Can. J. Microbiol.21:675-681.

136. Meflkejobn, G., L. G. Reimer, P. S. Graves, and C.Helmick. 1981. Cryptic epidemic of Q fever in a medicalschool. J. Infect. Dist. 144:107-113.

137. Morahan, P. S. 1980. Macrophage nomenclature: where

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from



are we going? RES J. Reticuloendothel. Soc. 27:223-245.138. Myers, W. F., 0. G. Baca, and C. L. Wisseman, Jr. 1980.

Genome size of the rickettsia Coxiella burnetii. J. Bac-teriol. 144:460-461.

139. Myers, W. F., C. M. Downs, and D. Paretsky. 1963.Some biochemical alterations in hosts during Coxiellaburnetii infection. Iowa State J. Sci. 38:69-74.

140. Myers, W. F., R. A. Ormsbee, J. V. Osterman, and C. L.Wisseman, Jr. 1967. The presence of diaminopimelicacid in the rickettsiae. Proc. Soc. Exp. Biol. Med.125:459-462.

141. Myers, W. F., and D. Paretsky. 1961. Synthesis of serineby Coxiella burnetii. J. Bacteriol. 82:761-763.

142. Myers, W. F., L. E. Warfel, and C. L. Wisseman, Jr.1978. Absence of hydrogen peroxide production or cata-lase action by Rickettsia prowazeki. J. Bacteriol.136:452-454.

143. Nermut, M. V., S. Schramek, and R. Brezina. 1968.Electron microscopy of Coxiella burnetii phase I and II.Acta Virol. 12:446-452.

144. Nicolau, S., and N. Constanmtnescu. 1965. Rickettsii sirickettsioze, p. 525-640. Academiei Republicii SocialisteRomania, Bucharest.

145. Oliphant, J. W., D. A. Gordon, A. Meis, and R. R.Parker. 1949. Q fever in laundry workers, presumablytransmitted from contaminated clothing. Am. J. Hyg.49:76-82.

146. Ormnbee, R. A. 1952. The growth of Coxiella burnetii inembryonated eggs. J. Bacteriol. 63:73-86.

147. Ormsbee, R. A. 1969. Rickettsiae (as organisms). Annu.Rev. Microbiol. 23:275-292.

148. Ormsbee, R. A., E. J. Bell, D. B. Lackman, and G.Tallent. 1964. The influence of phase on the protectivepotency of Q fever vaccine. J. Immunol. 92:404-412.

149. Ormsbee, R. A., H. Parker, and E. G. Pickens. 1955. Thecomparative effectiveness of aureomycin, terramycin,chloramphenicol, erythromycin, and thiocymetin in sup-pressing experimental rickettsial infections in chick em-bryos. J. Infect. Dis. 96:162-167.

150. Ormnbee, R. A., and M. G. Peacock. 1964. Metabolicactivity in Coxiella burnetii. J. Bacteriol. 88:1205-1210.

151. Ormsbee, R. A., and M. G. Peacock. 1976. Rickettsialplaque assay and cloning procedures, p. 475-478. InV. J. Evans, V. P. Perry, and M. M. Vincent (ed.),Tissue Culure Association manual, vol. 2. Tissue CultureAssociation, Gaithersburg, Md.

152. Paquet, A., Jr., E. D. Rael, D. Klassen, I. Martinez, and0. G. Baca. 1978. Mitogenic and protective activityassociated with a lipopolysaccharide from Coxiella bur-netii. Can. J. Microbiol. 24:1616-1618.

153. Paretsky, D. 1968. Biochemistry of rickettsiae and theirinfected hosts, with special reference to Coxiella burneti.Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg.Abt. 1 206:284-291.

154. Paretsky, D. 1979. Regulatory mechanisms controllingbiochemical events during infection with Coxiella burne-tii, p. 140-143. In D. Schlessinger (ed.), Microbiology-1979. American Society for Microbiology, Washington,D.C.

155. Paretsky, D., and M. L. Anthes. 1978. Effects of infectionwith Coxiella burnetii on synthesis ofRNA in L cells. J.Infect. Dis. 137:486-489.

156. Paretsky, D., R. A. Consigli, and C. M. Downs. 1962.Studies on the physiology of rickettsiae. III. Glucosephosphorylation and hexokinase activity in Coxiella bur-netii. J. Bacteriol. 83:538-543.

157. Paretsky, D., C. M. Downs, R. A. ConsigU, and B. K.Joyce. 1958. Studies on the physiology of rickettisae. I.Some enzyme systems of Coxiella burnetii. J. Infect.Dis. 103:6-11.

158. Paretsky, D., C. M. Downs, and C. W. Salmon. 1964.Some biochemical changes in the guinea pig duringinfection with Coxiella burnetii. J. Bacteriol. 88:137-152.

159. Paretsky, D., and F. R. Gonzales. 1981. Some modulatingfactors of transcription in Q fever, p. 453-459. In W.Burgdorfer and R. L. Anacker (ed.), Rickettsiae and

rickettsial diseases. Academic Press, Inc., New York.160. Paretsky, D., F. R. Gonzales, and K. Berqulst. 1981.

Temporal studies of factors associated with changes intranscription during Q fever. J. Gen. Microbiol. 122:227-233.

161. Paretsky, D., and J. Stueckemann. 1970. Chemical andbiochemical changes in subcellular fractions of guineapig liver during infection with Coxiella burneti. J. Bacte-riol. 102:334-340.

162. Perkins, H. R., and A. C. Allison. 1963. Cell wallconstituents of rickettisae and psittacosis-lymphogranu-loma organisms. J. Gen. Microbiol. 30:469-480.

163. Philip, C. B. 1948. Observations on experimental Qfever. J. Parasitol. 34:457-464.

164. Pope, J. H., W. Scott, and R. Dwyer. 1960. Coxiellaburneti in kangaroos and kangaroo ticks in westernQueensland. Aust. J. Exp. Biol. Med. Sci. 38:17-28.

165. Powanda, M. C., S. V. Machotka, and R. A. Kishimoto.1978. Metabolic sequelae of respiratory Q fever in theguinea pig. Proc. Soc. Exp. Biol. Med. 158:626-630.

166. Rehacek, J., and R. Brezina. 1964. Propagation of Cox-iella burneti in tick tissue cultures. Acta Virol. 8:380-381.

167. Ribi, E., and B. H. Hoyer. 1960. Purification of Q feverrickettsiae by density-gradient sedimentation. J. Im-munol. 85:314-318.

168. Roberts, A. N., and C. M. Downs. 1959. Study on thegrowth of Coxiella burnetii in the L strain mouse fibro-blast and the chick fibroblast. J. Bacteriol. 77:194-204.

169. Robson, A. O., and C. D. G. L. Shimmin. 1959. ChronicQ fever. 1. Clinical aspects of a patient with endocardi-tis. Br. Med. J. 1:980-983.

170. Rose, K. M., and S. T. Jacob. 1979. Phosphorylation ofnuclear poly(A) polymerase. J. Biol. Chem. 254:10256-10261.

171. Rosenberg, M., and N. Kordova. 1962. Multiplication ofCoxiella burneti in Detroit-6 cell cultures: an electronmicroscope study. Acta Virol. 6:176-180.

172. Rosman, M. S., W. F. Lubbe, M. Hayden, N. Basson, andC. J. Uys. 1977. Q fever endocarditis. S. Afr. Med. J.53:296-300.

173. Ruppaner, R., H. P. Riemann, T. B. Farver, G. West,D. E. Behymer, and C. WlJaysinghe. 1978. Prevalence ofCoxiella burnetii (Q fever) and Toxoplasma gondiiamong dairy goats in California. Am. J. Vet. Res.39:876-870.

174. Schaechter, M., A. J. Tousimns, Z. A. Cohn, H. Rosen, J.Campbell, and F. E. Hahn. 1957. Morphological chemi-cal, and serological studies of the cell walls of Rickettsiamooseri. J. Bacteriol. 74:822-829.

175. Schramek, S. 1968. Isolation and characterization ofdeoxyribonucleic acid from Coxiella burneti. Acta Virol.12:18-22.

176. Schramek, S., and R. Brezina. 1976. Characterization ofan endotoxic lipopolysaccharide from Coxiella burnetii.Acta Virol. 20:152-158.

177. Schramek, S., and R. Brezina. 1979. Isolation of endo-toxic lipopolysaccharide from phase II Coxiella burnetii.Acta Virol. 23:349.

178. Schramek, S., R. Brezina, and J. Kazar. 1978. Influenceof mild acid hydrolysis on the antigenic properties ofphase I Coxiella burnetii. Acta Virol. 22:302-308.

179. Schramek, S., and H. Mayer. 1982. Different sugarcompositions of lipopolysaccharides isolated from phaseI and pure phase II cells of Coxiella burnetii. Infect.Immun. 38:53-57.

180. Sidwell, R. W., B. D. Thorpe, and L. P. Gebhardt. 1964.Studies of latent Q fever infections. I. Effects of wholebody X-irradiation upon latently infected guinea pigs,white mice and deer mice. Am. J. Hyg. 79:113-124.

181. SWdwell, R. W., B. D. Thorpe, and L. P. Gebhardt. 1964.Studies of latent Q fever infections. II. Effects of multi-ple cortisone infections. Am. J. Hyg. 79:320-327.

182. Silberman, R., and P. Fiset. 1968. Method for countingrickettsiae and chlamydiae in purified suspensions. J.Bacteriol. 95:259-261.

MICROBIOL. REV.


mbr.asm

.org/D

ownloaded from



183. Silverman, D. J., P. Fiset, and C. L. Wisseman, Jr. 1979.Simple, differential staining technique for enumeratingricketsiae in yolk sac, tissue culture extracts, or purifiedsuspensions. J. Clin. Microbiol. 9:437-440.

184. Silverman, D. J., and C. L. Wisseman, Jr. 1979. In vitrostudies of rickettsia-host cell interactions: ultrastructuralchanges induced by Rickettsia rickettsii infection ofchicken embryo fibroblasts. Infect. Immun. 26:714-727.

185. Smith, J. D., and M. G. P. Stoker. 1951. The nucleicacids of Rickettsia burneti. Br. J. Exp. Pathol. 32:433-441.

186. Stelzner, A., and W. Linss. 1968. Binary fission in Cox-iella burneti. Nature (London) 218:1070.

187. Stoenner, H. G., D. B. Lacknman, W. W. Benson, J.Mather, M. Casey, and K. A. Harvey. 1961. The role ofdairy cattle in the epidemiology of Q fever in Idaho. J.Infect. Dis. 109:90-97.

188. Stoker, M. G. P., and P. Fiset. 1956. Phase variation ofthe Nine Mile and other strains of Rickettsia burnetiiCan. J. Microbiol. 2:310-321.

189. Stoker, M. G. P., K. M. Smith, and P. Flset. 1956.Internal structure of Rickettsia burnetii as shown byelectron microscopy of thin sections. J. Gen. Microbiol.15:632-635.

190. Stueckemann, J., and D. Paretsky. 1971. Changes inhepatic glycogen, protein, and ribonucleic acid synthe-sis, and some effects of cortisol, during Q fever. J.Bacteriol. 106:920-924.

191. Syrucek, L., 0. Sobeslavsky, and I. Gutvirth. 1958.Isolation of Coxiella burnetii from human placentas. J.Hyg. Epidemiol. Microbiol. Immunobiol. 2:29-35.

192. Thompson, H. A., 0. G. Baca, and D. Paretsky. 1971.Presence of ribosomal ribonucleic acid in the rickettsiaCoxiella burneti. Biochem. J. 125:365-366.

193. Thompson, H. A., and D. Paretsky. 1973. Ribonucleicacid and protein synthesis in guinea pig liver during Qfever. Infect. Immun. 7:718-724.

194. Tobin, M. J., N. Cahill, G. G. Gearty, B. Maurer, S.Blake, and K. Daly. 1982. Q fever endocarditis. Am. J.Med. 72:396-400.

195. Tsung, P. K., and D. Paretsky. 1968. Biochemical andultrastructural changes in liver endoplasmic reticularfractions during Q fever. Acta Virol. 12:49-53.

196. Tyeryar, F. J., Jr., E. Weiss, D. B. Millar, F. M.Bozeman, and R. A. Ormsbee. 1973. DNA base composi-tion of rickettsiae. Science 180:415-417.

197. Tzianabos, T., C. W. Moss, and J. E. McDade. 1981.Fatty acid composition of rickettsiae. J. Clin. Microbiol.13:603-605.

198. Urvolgyi, J., and R. Brezina. 1981. New antigens inCoxiella burnetfii? Acta Virol. 25:169.

199. Wachter, R. F., G. P. Briggs, J. D. Gangemi, and C. E.Pedersen, Jr. 1975. Changes in buoyant density relationships of two cell types of Coxiella burneti phase I. Infect.Immun. 12:433-436.

200. Walker, T. S., and H. H. Winkler. 1978. Penetration ofcultured mouse fibroblasts (L cells) by Rickettsia prowa-zeki. Infect. Immun. 22:200-208.

201. Walker, W. S. 1976. Functional heterogeneity of macro-

phages, p. 91-110. In D. S. Nelson (ed.), Immunobiologyof the macrophage. Academic Press, Inc., New York.

202. Warner, N. L., S. W. Burchiel, E. B. Walker, J. B.Richey, J. F. Leary, and S. McGlaughlin. 1979. Charac-teristics of murine B cell lymphomas and macrophagetumors by flow microfluorometry and functional assays,p. 243. In W. Terry and Y. Yamamura (ed.), Immunobio-logy and immunotherapy of cancer. Elsevier/North-Hol-land Publishing Co., New York.

203. Weiss, E. 1973. Growth and physiology of rickettsiae.Bacteriol. Rev. 37:259-283.

204. Weiss, E. 1982. The biology of rickettsiae. Annu. Rev.Microbiol. 36:345-370.

205. Weiss, E., and H. R. Dressier. 1958. Growth of Rickettsiaprowazeki in irradiated monolayer cultures of chickenembryo entodermal cells. J. Bacteriol. 75:544-552.

206. Weiss, E., and H. C. Pietryk. 1956. Growth of Coxiellaburnetii in monolayer cultures of chicken embryo ento-dermal cells. J. Bacteriol. 72:235-241.

207. Welsh, H. H., E. H. Lennette, F. R. Abinanti, and J.Winn. 1951. Q fever in California. IV. Occurrence ofCoxiella burnetii in the placenta of naturally infectedsheep. Public Health Rep. 66:1473-1477.

208. Wiebe, M. E., P. R. Burton, and D. M. Shankel. 1972.Isolation and characterization of two cell types of Cox-iella burnetii phase I. J. Bacteriol. 110:368-377.

209. Wike, D. A., R. A. Ormsbee, G. Tallent, and M. G.Peacock. 1972. Effects of various suspending media onplaque formation by rickettsia in tissue culture. Infect.Immun. 6:550-556.

210. WilIams, J. C., and J. L. Cantrell. 1982. Biological andimmunological properties of Coxiella burnetii vaccines inC57BL/lOScN endotoxin-nonresponder mice. Infect.Immun. 35:1091-1102.

210a.Williams, J. C., M. G. Peacock, and T. F. McCaul. 1981.Immunological and biological characterization of Cox-iella burnetii, phases I and II, separated from hostcomponents. Infect. Immun. 32:840-851.

211. Williams, J. C., and J. C. Peterson. 1976. Enzymaticactivities leading to pyrimidine nucleotide biosynthesisfrom cell-free extracts of Rickettsia typhi. Infect. Im-mun. 14:430-448.

212. Wissamm, C. L., Jr. 1981. Some biological properties ofrickettsiae pathogenic for man, p. 293-311. In W. Burg-dorfer and R. L. Anacker (ed.), Rickettsiae and rickettsi-al diseases. Academic Press, Inc., New York.

213. Wisseman, C. L., Jr., P. Fiset, and R. A. Ormsbee. 1967.Interaction of rickettsiae and phagocytic host cells. V.Phagocytic and opsonic interactions of phase I and phaseII Coxiella burneti with normal and immune humanleukocytes and antibodies. J. Immunol. 99:669-674.

214. Yu, F. L., and P. Feigelson. 1971. Cortisone stimulationof nucleolar RNA polymerase activity. Proc. Natl. Acad.Sci. U.S.A. 68:2177-2180.

215. Yunker, C. E., R. A. Ormsbee, J. Cory, and M. G.Peacock. 1970. Infection of insect cells with Coxiellaburneti. Acta Virol. 14:383-392.

216. Zdrodovskli, P. F., and H. M. Golinevich. 1960. Therickettsial disease, p. 372-423. Pergamon Press, NewYork.

VOL. 47, 1983


mbr.asm

.org/D

ownloaded from


q fever and coxiella burnetii: a model interactionst · identified in arthropods, fish, birds,...

Documents