in u.s.a. growthand physiology ofrickettsiaethe generic name "rickettsia" (uncapital-ized)...

BACrERIOLOGICAL REVIEWS, Sept. 1973, p. 259-283 Vol. 37, No. 3Copyright © 1973 American Society for Microbiology Printed in U.S.A.

Growth and Physiology of RickettsiaeEMILIO WEISS

Department of Microbiology, Naval Medical Research Institute, Bethesda, Maryland 20014

INTRODUCTION .......................................................... 259STRUCTURE AND CHEMICAL COMPOSITION ......... ....................... 260GROWTH .................. ........................................ 261

Cultivation in Eggs .......................................................... 261Host-Parasite Relationship as Observed in Cell Culture ....... .................. 262General observations ......................................................... 262Plaque formation .......................................................... 262Mechanism of penetration into host cells ........... ........................... 264Intracellular multiplication .................................................... 265Release from cells .......................................................... 266Effect on the host cell.266

Inhibition of Growth and Drug Resistance ............ ........................... 267p-Aminobenzoic acid .......................................................... 267Antibiotics........................................................... 268

METABOLISM........................................................... 269Typhus and Spotted Fever Rickettsiae .......................................... 269

Substrates that stimulate respiration ............. ............................ 269Pathway of glutamate utilization ................. ............................ 270Production of ATP .......................................................... 271Synthetic activities .......................................................... 271Physiology of cell injury ...................................................... 273Reactivation........................................................... 274Metabolism of rickettsiae multiplying in host cells ........ ..................... 275

Q Fever Rickettsiae .......................................................... 275WHY HAS INDEPENDENT CULTIVATION NOT BEEN ACHIEVED? ..... ...... 276Energy Parasitism .......................................................... 276Leakiness .................... ...................................... 277Unidentified Auxotrophy ....................................................... 277Highly Sensitive Regulatory Mechanism ............ ............................ 277Comment ................... ....................................... 278

A FINAL REMARK .......................................................... 278LITERATURE CITED .......................................................... 278

INTRODUCTION

The generic name "rickettsia" (uncapital-ized) is often used very broadly and with littlediscrimination. A rickettsia is regarded as aminuscule organism, intermediate in size andproperties between a bacterium and a virus, orany microorganism larger than a virus that cannot be grown in lifeless media, or any smallorganism seen in insects and in other arthro-pods. It is not difficult to recognize the inade-quacies of the above generalizations. (i) Thedifference between a bacterium and a virus is sogreat that no organism can be truly placed in anintermediate category (107). (ii) Moulder (74)and Hanks (45) have discussed the problemsconfronting, respectively, students of chlamy-diae and of the noncultivable mycobacteria.These problems are quite different from eachother and from those confronting the rickettsi-ologist (81). (iii) Organisms other than rickett-

siae discovered in insects and ticks have seldombeen studied in detail, but one such obligateintracellular bacterium, Wolbachia persica, forwhich there is a fair amount of information,appears to be quite different in morphology andphysiology from rickettsiae (110, 111, 130).What is then a meaningful definition of

rickettsiae? This question can not be properlyanswered at this point, but possible definitionswill emerge in the course of this review. For thepurpose of orientation, it can be stated that theeighth edition of Bergey's Manual of Deter-minative Bacteriology (129) will list 10 speciesin the genus Rickettsia divided into threegroups, namely, typhus, spotted fever, andscrub typhus rickettsiae. Although some inves-tigators may prefer a somewhat different class-ification, there is wide agreement that these 10species are sufficiently similar to be regarded asmembers of the same genus. The term "rick-ettsia" can also be extended to the monogeneric

259

on October 19, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

BACTERIOL. REV.

genus Coxiella, which resembles rickettsiae insome respects and requires similar methodolo-gies for investigation. This is not true of thetrench fever agent (designated Rochalimaeaguintana in the eighth edition of Bergey'sManual), which can be cultivated on relativelysimple bacteriological media (77, 115). Theinclusion of this organism among the rickettsiaeis still of great value to the epidemiologist but oflimited usefulness to the bacterial physiologist.

This review will be concerned only withbacteria belonging to the genera Rickettsia andCoxiella. It will be devoted to the propertieswhich most clearly distinguish them from otherorganisms, namely growth and physiology.Other topics and other bacteria will be dis-cussed only insofar as they contribute to themain purpose of this review.

STRUCTURE AND CHEMICALCOMPOSITION

Recent studies have yielded overwhelmingevidence that the fine structure and chemicalcomposition of rickettsia are entirely similar tothose of other gram-negative bacteria (4, 5, 23,54, 76, 79, 80, 89, 98, 108, 113, 114, 133, 136, 140,143).Members of the genus Rickettsia measure 0.3

to 0.5 kim in width but vary considerably inlength. R. prowazeki is quite pleomorphic andits usual length is 2 to 4 gim, R. rickettsimeasures approximately 2 gm, and R.tsutsugamushi measures 1.5 gim. C. burneti issomewhat smaller, 0.2 to 0.4 ,4m in width by 0.4to 1.0 ,m in length. In contrast to other rickett-siae, C. burneti undergoes an antigenic phasevariation. Freshly isolated rickettsiae are inphase I. When passed serially in chicken em-bryos they are eventually replaced by phase IIorganisms. The cells of the two phases differfrom each other in several respects, as discussedby Fiset and Ormsbee (35). Recently Wiebe etal. (136) isolated two types of C. burneti phase Icells on the basis of buoyant density. One typewas small, compact, rod-shaped, and had adense nucleoid. The other was somewhat larger,rounded, and had disperse nuclear filaments.Both types were viable and produced mixturesof both types when cultured separately.Most electron micrographs of rickettsiae re-

veal a three-layered cell wall, but at highmagnification it has been possible to resolve afive-layer architecture analogous to that de-scribed for Escherichia coli (4). The cell wallcontains, in addition to sugars, amino sugars,and amino acids, muramic acid (89) and diami-nopimelic acid (76, 143). Wisseman (140) hasobtained evidence that rickettsiae have en-

dotoxic activity, which suggests the presence ofa lipopolysaccharide layer, typical of gram-negative bacteria. Teichoic acid, typical ofgram-positive bacteria, was not detected. Cap-sular material surrounding the cell wall can beseen in electron micrographs of fresh prepara-tions (4). There is also morphological as well asphysiological evidence (79) of a cytoplasmicmembrane.The internal structure consists of electron-

dense granules and fine strands which indicatethe presence of ribosomes and deoxyribonucleicacid (DNA), respectively. Both ribonucleic acid(RNA) and DNA have been isolated from rick-ettsiae on numerous occasions. The RNA of C.burneti was subjected to electrophoretic analy-sis and RNA species were isolated with sedi-mentation constants of 4 to 5S, typical oftransfer RNA, and 16 and 23S, typical ofribosomal RNA (113). The size of the DNA iscomparable to that of other microorganisms ofsimilar size (54). Recent studies (114) havepermitted a detailed comparison of the molarpercentages of guanine plus cytosine (% G + C)of the DNA of several species of rickettsiae. Therickettsiae of the typhus group (101, 114) haveapproximately 30% G + C, and those of thespotted fever group have 32.5% G + C. Thisdifference is highly significant and indicates anearly evolutionary separation (114). Previousinvestigators had reported somewhat higher G+ C contents for R. prowazeki and R. rickettsi(91, 144). The % G + C of C. burneti wasdetermined by several investigators and wasfound to be approximately 43 to 45 (70, 99, 100,103).Thus, the obligate intracellular parasitism of

the rickettsiae is not reflected in unusual mor-phologic or macromolecular characteristics. It istrue that some of the observations made are notreadily explained. For example, Anderson et al.(5) described small electron-lucent sphericalstructures in the cytoplasm of R. prowazeki.These are not understood, and at present noparticular significance can be attributed tothem. An unusually high DNA: RNA ratio hasbeen reported in a number of cases, but this isprobably due to loss of RNA from resting ordamaged rickettsiae (28). The low % G + C isbelieved by some (102) to be typical of intracel-lular parasites and other organisms that aresubjected to natural background ratioactivitywhich favors low guanine content, but areprotected from ultraviolet radiation exposurewhich favors low thymine content. This is ahighly speculative view which requires confir-mation. In general, the unbiased investigatorwould be hard put to explain the obligate

260 WEISS


http://mm

br.asm.org/

Dow

nloaded from


GROWTH AND PHYSIOLOGY OF RICKETTSIAE

intracellular parasitism of rickettsiae on thebasis of structure and chemical composition.

GROWTHRickettsiae, despite several elaborate at-

tempts, have not been grown in the absence ofhost cells. They are most commonly cultivatedin fertile hen eggs, tissue cultures, or smalllaboratory animals, and, in certain specializedcases, in arthropods. Eggs have been mostuseful for the production of large harvests forantigenic and metabolic studies and for themanufacture of vaccines. Tissue cultures arebetter suited, however, for detailed observationsof host-parasite relationships. Laboratory ani-mals are still widely used for primary isolations,infectivity assays in some cases, as well as fortests of virulence and immunogenicity. Masscultivation of rickettsiae in arthropods is nowonly of historical interest, but because of theirroles as primary or secondary hosts, arthropodsshould be regarded as potentially useful experi-mental hosts. For the purpose of this reviewonly information derived from eggs and fromcell cultures will be examined.

Cultivation in EggsThe method introduced by Cox of growing

rickettsiae in the yolk sac of chicken embryos isby far the most satisfactory (30, 31). Fertile heneggs are highly susceptible to infection withestablished strains and, in most cases, with newisolates. The incidence of infection of eggsinoculated with the Madrid E strain of R.prowazeki with concentrations covering themean infective dose (ID50) range is in agreementwith the assumption that only one infectiousrickettsia is required for infection (124). Theratio of total particles, identifiable by micros-copy, to particles successful in infecting achicken embryo varies with the species of rick-ettsiae and with the physiological condition ofthe rickettsial preparation. Wike et al. (139)determined the ratio to be 50: 1 for R. typhi and300: 1 for R. rickettsi. The infectivity of C.burneti is so high that the ratio is presumablyclose to 1:1. The instability of scrub typhusrickettsiae has discouraged investigators fromconducting quantitative determinations of thisnature with this organism.

Chicken embryos are usually inoculated dur-ing the fifth or sixth day of embryonic develop-ment and, because they must be harvestedbefore hatching time, they offer a 12- or 13-dayperiod for the growth of rickettsiae. When theinoculum is one or a few infectious units, thisperiod is long enough for maximal growth of

established strains of typhus rickettsiae but notof the other rickettsiae. To make sure thatinfection is not missed in some of the embryos, asecond passage is required for the spotted feverrickettsiae and a third passage for C. burneti.With larger numbers of rickettsiae there ap-pears to be a linear relationship between size ofinoculum and mean survival time of the em-bryos (Fig. 1). For certain strains of typhusrickettsiae this relationship is highly reproduci-ble and has been widely applied for the titrationof rickettsiae by the single dilution method (17,124). With Q fever rickettsiae the range ofuseful concentrations is small (82).The optimal temperature for the growth of

spotted fever rickettsiae is 33.5 C, a compromisebetween the optimum for the rickettsiae, whichis probably about 32 C, and the minimumtemperature generally required by embryos fortheir survival. For reasons that have not been

12

0 lo0

Ii

0

LL

y

I- 8

0

11

(f) e6

zSCI

\ )

\C

0

2 4 6 8 10

INFECTIOUS RCKETTSIAE PER EGG (LOGIO)FIG. 1. Relationship between infectious dose and

mean time of survival of the embryos. This correlationis based on experiments with the Madrid E strain ofR. prowazeki (124) and the Nine Mile strain of C.burneti (82). Both infectious rickettsiae and days ofchicken embryo survival are plotted on logarithmicscales. When the concentration of R. prowazeki isreduced 10-fold the logarithm of the survival time ofchicken embryos is increased by 0.07. By extrapola-tion it can be calculated that embryos inoculated withsingle rickettsiae survive for an average of 13.2 days. A10-fold reduction in the concentration of C. burnetiwill increase the logarithm of the survival time by0.12, but extrapolations to small numbers of infec-tious rickettsiae are not realistic.

261VOL. 37, 1973

6


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

fully explored, the embryos are killed relativelyearly in the course of infection, but if the eggsare further incubated rickettsial multiplicationin the yolk sacs continues for at least 2 days(109). The optimal temperature for the otherrickettsiae is 35 C. In contrast to spotted feverrickettsiae, there is evidence that growth stopspromptly when the embryos die and, in the caseof typhus and scrub typhus rickettsiae, the titerdeclines rapidly after the death of the embryos(142). Optimal yields of infectious particles peregg are about 107 to 109 with most strains ofRickettsia and about 1011 with Coxiella. Interms of dry weight, harvests with members ofthe genus Rickettsia, measured after separationfrom host cell components, average from 0.05 to0.5 mg/gm of yolk sac, R. tsutsugamushi yield-ing the smallest and the typhus rickettsiae thelargest amounts. With C. burneti, yields of 1 to2 mg/gm of yolk sac are not unusual.The growth of rickettsiae in the yolk sac of

chicken embryos has been amazingly useful inso many kinds of investigations (some moreproperly described in other sections) that itsshortcomings have been obscured. Rickettsiaeharvested from yolk sacs can be separated fromhost components only by relatively elaboratetechniques which often result in damage to themicroorganisms. As a model of rickettsial infec-tion the yolk sac has rendered good service, butit has obvious limitations, which in some caseshave been disregarded.

Host-Parasite Relationship as Observed inCell Culture

General observations. Some of the eventsthat take place in the infected yolk sac of thechicken embryo can be visualized by explantingthe entodermal cells (128). The explants can bereadily obtained from the avascular portion ofthe yolk sac of the 4-day-old embryo. The cellsreadily adhere to glass, and although they donot multiply to any extent, they expand intomonolayers composed of a single cell type.These cells are highly vacuolated and havelarge, faintly staining cytoplasms that permitdetailed cytological observation. Examples ofcells infected with R. prowazeki and C. burnetiillustrate the basic difference between thesetwo microorganisms in their relationship to thehost cell (Fig. 2). R. prowazeki is strictly con-fined to the cytoplasm, does not invade thevacuoles, and does not change the morphologyof the host cell (120). C. burneti, on the otherhand, multiplies in the vacuoles and during theprogress of infection converts almost the entirecell into one large vacuole and compresses the

nucleus and remaining cytoplasm towards theperiphery (9, 132).

Spotted fever and scrub typhus rickettsiaehave been studied in numerous cell types. Liketyphus rickettsiae, they multiply primarily inthe cytoplasm and not in the vacuoles, but theydo display some important features of theirown. Spotted fever rickettsiae, unlike otherrickettsiae, are occasionally seen in the nucleus.The frequency of intranuclear invasion is great-est in mammalian cells, but they have beenseen in the nuclei of avian and arthropod cellsas well (24). Scrub typhus rickettsiae arestrictly intracytoplasmic, but they usually ag-gregate in the region adjacent to the nucleus(22). R. canada, which in many respects, in-cluding DNA base composition (114), resemblesthe typhus rickettsiae (66), is occasionally seenin the nucleus (25).

Rickettsiae of each of the major groups havebeen grown in several types of both primarycultures and established cell lines, but com-parative observations have been relatively few.Bozeman et al. (22) noted that most rickettsialspecies eventually formed dense masses in thecytoplasm of MB III mouse lymphosarcomacells, but spotted fever rickettsiae remainedloosely scattered even at the peak of infection.This difference was particularly striking whenR. tsutsugamushi and R. rickettsi were com-pared in cultures of 14 pf rat fibroblasts (97).Besides major differences among rickettsialspecies, cell cultures in some cases reflectedstrain variation. Anderson et al. (5) showed thatthe growth of the Breinl strain of R. prowazekiin the BS-C-1 line of Cercopithecus kidney cellswas more luxuriant than that of the avirulentMadrid E strain. Kordova et al. (58) obtainedexcellent multiplication of phase II C. burnetiin L cells but very poor growth of phase I.Barker et al. (8) obtained equally satisfactorypropagation of the Gilliam, Karp, and Katostrains of R. tsutsugamushi in BS-C-1 cells, butnoted that the three strains produced differentcytopathic effects. The reverse comparison ofthe same rickettsia in different types of cellculture was made by Kenyon et al. (53). Highestyields of R. rickettsi were obtained from duckembryo cells, followed by chicken embryo andVero cells. Titers from L and human diploidWI-38 cells were considerably lower. There is noindication that other rickettsiae have the samedegree of host cell specificity.Plaque formation. Cloning of rickettsiae has

been a goal of several investigators for sometime, but it has become technically feasibleonly recently. Before the development of plaqu-

262 WEISS


http://mm

br.asm.org/

Dow

nloaded from


S4

-IOpsm

FIG. 2. Two entodermal cells of the yolk' sac infected with R. prowazeki (A) and one cell infected with C.burneti (B). Neighboring cells are uninfected. Note that R. prowazeki grows in the cytoplasm and does not alterits vacuolated appearance nor obviously affect the structure of the nucleus. C. burneti, on the other hand, growsin the vacuoles and eventually changes the entire host cell into one large vacuole and compresses the nucleusand remaining cytoplasm toward the periphery. A, from Weiss and Dressler (120); B, from Blackford (9).

263


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

ing methods, it was possible to count infectedcells or foci of infection in monolayer culture atlow magnification (9, 120, 121, 132). Sincerickettsiae migrated relatively slowly from cellto cell in these cultures, the counts reflected theconcentration of the inoculum. This countingmethod, however, required staining of the cul-tures and thus had serious limitations.The plaquing method was first developed by

Kordova (57), who was not entirely satisfiedwith its reproducibility. Weinberg et al. (116)obtained consistent results with R. rickettsi,and McDade' et al. (62, 63) extended the methodto the other major rickettsial species. Primarychicken embryo monolayers were used in mostof the above experiments. The chief technicaldifficulty which had to be surmounted wasmaintaining the monolayer under agar in goodphysiological condition for the required 8 to 17days without stimulating growth to the extentthat it would cover the plaques. Conditionsaffecting plaque formation were discussed byWike et al. (139), who found among other thingsthat suspension of the rickettsial inoculum inbrain heart infusion broth and incubation of thecultures at 32 C was essential for satisfactoryresults.When plaques are formed under optimal

conditions, differences among the major groupsof rickettsiae become apparent. The spottedfever rickettsiae require the shortest period toform plaques. They are first detected at 5 to 6days but are more readily seen 2 or 3 days later,at which time they have a well-delineatedmargin and a diameter of 2 mm. The otherrickettsiae develop smaller plaques, 1 mm indiameter (R. canada 0.75 mm), which some-times have an indistinct perimeter. For bestvisualization, 10 to 12 days are required fortyphus rickettsiae and for R. canada, 17 daysfor R. tsutsugamushi, and 8 to 10 days for C.burneti. The titers of most rickettsial prepara-tions determined by the enumeration of plaqueshave been as high as, and in some cases slightlyhigher than, those obtained in eggs (63, 139).McDade and Gerone (62) reported in a briefstudy that the plaque titer of R. tsutsugamushiwas somewhat higher than that obtained inmice, whereas the reverse was true of C.burneti. If these results can be confirmed, theywould indicate that plaque titration is particu-larly useful for unstable organisms, such as R.tsutsugamushi, which require prompt contactwith their host cells.The usefulness of the plaquing technique for

rickettsiae has been tested in various determi-nations such as antibiotic susceptibility (61),clone isolation (139), direct isolation from tick

hemolymph or from guinea pig blood (137),absorption to host cells, and survival in varioussolutions (138, 139). Its full potential for kineticand genetic studies has not yet been exploited.Mechanism of penetration into host cells.

It is essential to provide more than a casualcontact between rickettsiae and host cells dur-ing the initial stage of infection. In entodermalcell cultures and certain other cell lines this isaccomplished by centrifuging the rickettsiaeonto the cells at 1,500 x g for 1 h at 20 C (121).Suspended cell cultures, such as MB III mouselymphoblasts, are placed in a small volume ofmedium before infection with rickettsiae. Aftera period of absorption, usually 2 h at 37 C,medium is added to reestablish the originalvolume (22). Monolayers are generally overlaidwith inoculum suspended in very small volumesand incubated at room temperature or at 37 Cbefore the medium is added. If the monolayer isto receive an agar medium after the inoculumhas been removed by washing, within limits thenumber of plaques increases with absorptiontime. It was shown by Wike et al. (139) that R.rickettsi continued to absorb at room tempera-ture for as long as 4 h. Absorption time wasgreatly reduced, however, by centrifuging thecultures at 600 x g for 5 min.There is good evidence that in addition to

close contact necessary for absorption, the ac-tive participation of rickettsiae is required forpenetration. This phenomenon has been stud-ied by Cohn et al. (27) with R. tsutsugamushipenetrating into MB III mouse lymphoblasts.There are two other phenomena, studied mostextensively by Bovarnick and Allen and theircollaborators (3, 12-15, 105) with typhus rick-ettsiae, that need not result in penetration intosuitable host cells, but most likely involve thesame mechanism. One of these is mouse toxicitywhich is due to increased permeability of theendothelial cells of mice injected intravenouslywith high concentrations of rickettsiae. Theother is hemolysis of rabbit or sheep erythro-cytes elicited by rickettsiae. Both of thesephenomena, but especially the latter, are welladapted to quantitative studies.

In general, only viable rickettsiae penetrateinto host cells, are toxic for mice, or lyse redblood cells. For example, when R.tsutsugamushi is inactivated by heat (at 56 Cfor 5 min), by incubation with 0.1% Formalin, orby exposure to ultraviolet irradiation, it doesnot penetrate into host cells (27). Likewise, itwas shown with typhus rickettsiae that viabili-ty, toxicity for mice, hemolytic activity, andrespiration stimulated by glutamate are inti-mately associated. It is possible to selectively

264 WEISS


http://mm

br.asm.org/

Dow

nloaded from


GROWTH AND PHYSIOLOCiY OF RICKETTSIAE

destroy infectivity for the chicken embryo byultraviolet irradiation (3), but all other at-tempts at selective inactivation have failed (3,94).Of the factors that affect penetration of

rickettsiae into host cells, some influence themetabolic activity of the rickettsiae directlywhereas others alter the rate of inactivation ofextracellular rickettsiae. With a complete cul-ture medium consisting of balanced salt solu-tion (BSS), beef embryo extract, and horseserum, penetration of R. tsutsugamushi intoMB III cells at 37 C proceeds linearly for 15 minand then the rate gradually declines. About 80%of the penetration takes place within 30 min,and after 120 min there is no further increase.Undoubtedly the stability of rickettsiae plays a

role in the course of penetration, because thereis an appreciable decline in infectivity of ex-

tracellular scrub typhus rickettsiae within 2 h at37 C. When the complete medium is substitutedby BSS alone, penetration of R. tsutsugamushiinto MB III cells is reduced to about one half.When BSS is substituted by various combina-tions of sucrose and electrolytes, it can be shownthat both monovalent and divalent metallicions are required. The effect of the divalent ionsappears to be relatively nonspecific; calciumand magnesium added singly serve as well as

when added jointly, and they can be substitutedby manganese, barium, or cadmium. One per-

cent bovine serum albumin can substitute forbeef extract and horse serum in the completemedium. This suggests that these proteins actprimarily as stabilizing agents (27).Of particular interest are the factors that

most directly affect the metabolism of rickett-siae. Of a large number of substrates tested,only glutamic acid, glutamine, and a combina-tion of a-ketoglutaric acid and aspartic acid(which yield glutamic acid in the presence ofthe proper transaminase) are effective in in-creasing the penetration rate of R.tsutsugamushi (27). Glutamic acid and gluta-mine are important substrates of typhus andspotted fever rickettsiae (21, 44, 91, 133), andthis is probably the best evidence available thatthese compounds play similar roles in scrubtyphus rickettsiae. Although the primary effectis most likely on the metabolism of the rickett-siae leading to cell penetration, it was shown inthe same series of experiments that the survivalof the rickettsiae in the absence of host cells wasincreased by glutamate (27).

Since the oxidation of glutamate by rickett-siae leads to the formation of adenosine triphos-phate (ATP) (10), the above-described resultssuggest that rickettsiae must be capable of

carrying out energy-yielding metabolism inorder to penetrate into host cells. There isconsiderable evidence to support this suggestionfor scrub typhus as well as for other rickettsiae.Penetration of R. tsutsugamushi is greatly re-duced by various metabolic inhibitors such as2, 4-dinitrophenol or cyanide in the presence orabsence of glutamate. Chloramphenicol, whichis an effective inhibitor of the growth of rickett-siae (50), has no effect on penetration (27).Information for the other rickettsiae is of a moreindirect nature. Wike et al. (138) showed thatthe inclusion of glutamate in suspending mediaincreased the number of plaques formed by R.rickettsi and R. typhi. Pyruvate, an importantsubstrate of C. burneti (84), increases the num-ber of foci of infection in entodermal cell cul-tures (9). Glutamate also enhances the lysis oferythrocytes by typhus rickettsiae, provided therickettsiae are not damaged. When the rickett-siae are severely damaged but not completelyinactivated, ATP must be furnished (19).

It is well to remember that an active mode ofpenetration into host cells is not commonamong pathogenic bacteria. Chlamydiae arepassively engulfed into vacuoles where theymultiply (37), and neither viable nor nonviableE. coli or Staphylococcus aureus penetrate orare taken up by MB III cells under conditionsfavorable to scrub typhus rickettsiae (27).

Intracellular multiplication. There is over-whelming evidence from observations (too nu-merous to be cited) of stained smears or sectionsderived from infected yolk sacs, tissue cultures,laboratory animals, and arthropods as well asfrom electron microscopy that members of thegenus Rickettsia multiply by transverse binaryfission. Schaechter et al. (97) examined livingcultures of 14 pf rat fibroblasts infected with R.rickettsi by phase microscopy and recorded thisphenomenon photomicrographically on threeseparate occasions. His observations were con-firmed and extended to other rickettsiae andhost cells by Kokorin (55). Transverse binaryfission appears to be the only mechanism ofmultiplication in the genus Rickettsia. Theexistence of a development cycle or the forma-tion of inclusions, as in the case of chlamydiae,has been suggested in many instances but hasnever been substantiated. These suggestionswere made most frequently with R. tsutsugamu-shi, which has a tendency to form aggregatesnear the host cell nucleus.

Light and electron microscopy observationshave clearly indicated that C. burneti alsomultiplies by transverse binary fission. Kordova(56), however, presented evidence that binaryfission of this organism is preceded by the

265VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

formation of a filterable form. Although Kor-dova's experiments were carefully executed, sheleft a number of questions unanswered, such asthe size, chemical nature, and infectivity titersof these filterable forms. Until her work isconfirmed in another laboratory and the natureof these particles is better defined, this hypothe-sis must be held in abeyance.

Rickettsiae appear to grow best in well-nour-ished cells. When R. tsutsugamushi was grownin MB III or L cells, an increase in titer occurredonly in complete medium containing serum. InEagle medium without serum the titer re-mained approximately constant, but when anyone of the major constituents of Eagle mediumwas omitted, the titer declined rapidly (49).Except for the above findings and other casualobservations, e.g., that in some cases calf serumis more satisfactory than fetal calf serum andthat an acid pH is detrimental (131), theinformation on the nutritional requirements ofrickettsiae is meager. An approach which showsconsiderable promise, involving the use of selec-tive inhibitors of eukaryotic metabolism andlabeled compounds, has not yet been exploitedto study the details of nutrition.

Although rickettsiae require well-nourishedhost cells, the host cells need not multiply. Itwas shown that typhus rickettsiae propagatedin entodermal cells previously exposed to dosesof cobalt-60 irradiation of 100,000 R and in somecases as much as 300,000 R (120). Smaller doses,3000 to 5000 R, which elicit giant cell formationbut do not interfere with the adherence of thecells to the surface of the culture vessel, are usedquite frequently with chlamydiae and rickett-siae (41, 131). Although it is not certain thatirradiation increases the susceptibility of thehost cells to rickettsiae, it simplifies the designof certain quantitative experiments by prevent-ing unlimited host cell multiplication. Some ofthe experiments with R. tsutsugamushi weredone with cells inhibited by colchicine (22). Inthese experiments the host cell is treated simplyas a temporary, convenient microenvironmentfor the growth of rickettsiae.Release from cells. The view that the cell

infected with rickettsiae fills up with organismsand finally bursts and releases them into themedium is an oversimplification. This mayoccur with irradiated cells infected with typhusrickettsiae since, shortly after the peak of infec-tion has been reached, virtually all host meta-bolic activity suddenly ceases (131). This phe-nomenon has not been studied morphologically,however. Careful observations on release of R.rickettsi and R. tsutsugamushi from the 14 pfcell line of rat fibroblasts were made by Scha-

echter et al. (97). Rickettsiae are sometimestrapped in microfibrillar structures protrudingfrom the edge of the cell. When the microfibrilsretract they either carry the organisms backinto the cytoplasm or release them into theextracellular fluid. C. burneti propagates invacuoles which gradually become filled withorganisms. As the cultures age the number oforganisms in the vacuoles decrease, suggestingthat the rickettsiae are released over a variableperiod of time rather than suddenly (132).There is also considerable morphological evi-

dence that the rickettsiae are often released intoneighboring cells. In experiments with monolay-ers of chick entodermal cells (120, 132), thecount of foci of infection with C. burneti and R.prowazeki did not change appreciably as theinfections aged from 5 to 10 days, even thoughthe cells were not overlaid with agar. Each focusof infection consisted of one or more heavilyinfected cells and an increasing number ofneighboring cells that were lightly infected.There was little evidence that new foci ofinfection were established at a distance fromthe older foci.The mechanism of release of rickettsiae from

host cells requires more careful attention. Manytypes of cells can be shown to be susceptible toinfection and to be capable of supporting mod-erate growth, but relatively few cells release therickettsiae in large numbers in a predictablemanner. Good harvests of rickettsiae requiresuch a release and prompt harvest, before theorganisms lose their viability in the extracellu-lar environment.

Effect on the host cell. In a study of theinteraction of L cells and Chlamydia psittaci,Friis (37) showed that chlamydiae inactivatedby heat or neutralized by antiserum were rap-idly digested by the lysosomes. Viable chlamy-diae, on the other hand, remained separatedfrom the lysosomes and eventually multiplied inthe vacuoles. He concluded that chlamydiaemust possess a mechanism which prevents lyso-somal activation. A similar mechanism mayexist in C. burneti, which, like chlamydiae,multiplies in the vacuoles. It is conceivable thatthe disappearance of phase I in L cell and therapid growth of phase II, noted by Kordova etal. (58), is due to a difference in interaction withL cell lysosomes. Members of the genusRickettsia may bypass the action of the lyso-somes by their active penetration into thecytoplasm and avoidance of the vacuoles.Because rickettsial infection can be readily

recognized by ordinary light microscopy, littleattention has been paid to cytopathic effectsdemonstrable at relatively low magnification.

266 WEISS


http://mm

br.asm.org/

Dow

nloaded from



BS-C-1 cells infected with the Gilliam strain ofR. tsutsugamushi were reported to be predomi-nantly rounded, swollen globular in appearance,and those infected with the Karp strain wereshrunken with tendency towards pyknosis andspindling, whereas the Kato strain inducedareas of destruction in the sheet (8). R. typhipropagated in L cells produced little effecton the general appearance of the host cells untilthe cycle of growth was completed. R. akari, onthe other hand, led to a number of obviouschanges early in the course of infection. At 3days, while growth was still proceeding rapidly,about half of the host cells had been releasedfrom the surface of the flask and those remain-ing were rounded or crenated. The abovechanges were reflected in the ability of the hostcells to incorporate thymidine during the periodof most rapid rickettsial growth-an observa-tion facilitated by the failure of rickettsiae totake up this compound. R. typhi reduced hostthymidine incorporation only to a moderateextent, whereas R. akari reduced it almostcompletely (131). The effect of R. akari appearsto be analogous to that of chlamydiae whichwere shown to inhibit the thymidine kinaseactivity of their L host cells (60).The above-described observations just serve

to illustrate that the relationship of rickettsiaeto their host cells may vary. Our knowledge istoo limited to allow us to outline a basic patternof interaction and to attribute an evolutionarysignificance to deviations from this pattern.

Inhibition of Growth and Drug Resistancep-Aminobenzoic acid. Rickettsiae are unaf-

fected by the sulfonamides. Most of them areinhibited, instead, by p-aminobenzoic acid(pAB) (104), which in other microorganismsacts as an antagonist of sulfonamide inhibition.This property is unusual but not unique. It hasbeen encountered in Rochalimaea quintana (J.William Vinson, personal communication),Wolbachia persica (111), and in an isolate of E.coli (33). Mycobacterium tuberculosis is inhib-ited by a related compound, p-aminosalicylicacid, but this compound like the sulfonamidksinterferes with pAB metabolism (59, 145). Theinhibition of rickettsiae by pAB has been dem-onstrated under a variety of conditions withnewly acquired as well as old laboratory strainsand in natural infection of man. Most of thespecies of rickettsiae are affected by pAB, butthe susceptibility of R. tsutsugamushi is of alower order and C. burneti is not affected (104).Although the therapeutic use ofpAB is now onlyof historic interest, the study of this curious

inhibition may contribute to the understandingof the physiology of rickettsiae.The inhibition of rickettsiae by pAB can

easily be demonstrated in eggs. When the inocu-lum of rickettsiae is sufficient to kill untreatedembryos in 6 to 8 days, 1 mg of pAB, injectedthe previous day, delays the logarithmic growthof rickettsiae by 2 days and embryo death by 5days. The rickettsial titer in the eggs at the timeof embryo death is the same as in the controls(124)..pAB has also been used in a number of invitro tests, but in no instance has an effect beendemonstrated on an extracellular activity ofintact rickettsiae (18). Guardiola et al. (42, 43)showed that pAB reacts chemically with nico-tinamide adenine diphosphate (NAD) and pre-vents malic dehydrogenase activity of C.burneti and R. prowazeki cell extracts. Theseresults are thought provoking because R.prowazeki under certain conditions loses en-dogenous NAD (12, 14), but they do not explainwhy only the growth of typhus rickettsiae andvery few other microorganisms is affected bypAB.

Davis (33) isolated a strain of E. coli thatrequired both p-amino- and p-hydroxybenzoicacids, which furnished a useful model for theexplanation of events in rickettsiae. Excess pABinterfered with the utilization of p-hydroxyben-zoic acid (pHB), but the competition betweenthe two metabolites was not symmetrical. Evenlarge concentrations of pHB were not inhibi-tory. J. C. Snyder and B. D. Davis (Fed. Proc.10:419, 1951) applied this information to rick-ettsiae and showed that the inhibition of pABcould be competitively reversed by pHB. Theminimal pHB:pAB ratio required for reversalwas 1:10. Similar results were reported byTakemori and Kitaoka (112). Other explana-tions of the rickettsiostatic effect of pAB havebeen offered, namely that sulfonamides by someunknown mechanism favor the growth of rick-ettsiae or that pAB enhances the respiration ofthe host cells and this is deleterious to therickettsiae, but these explanations have beendisproven.

Weiss et al. (124-126) passed the Madrid Estrain of R. prowazeki serially in eggs in thepresence of concentrations of pAB that wereslightly inhibitory, and they isolated threestrains that were resistant to 1, 3, and 10 mg ofpAB per egg, respectively. The increased re-sistance to pAB was accompanied by an in-creased susceptibility to salicylic and acetyl-salicylic acids. The inhibition of acetylsalicylicacid (aspirin) was competitively reversed bypAB. The role of pAB thus changed from in-hibitor to antagonist of inhibition. These re-

267VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

sults suggest that rickettsiae, like the strain ofE. coli isolated by Davis, required both pHBand pAB. The requirement for pAB, however,was masked by its interference with the me-

tabolism of pHB. When the drug-resistantrickettsiae were exposed to an inhibitory con-

centration of acetylsalicylic acid plus sufficientpAB for complete reversal plus increasingamounts of pHB, inhibition of growth againappeared: pHB reversed the action of pAB, notin its role as inhibitor, but in its role as an an-

tagonist of inhibition. This interaction among

the three compounds can only be explained byassuming that it occurs at two different sitesin the rickettsial cell, presumably at the site ofspecific transport and at the enzymatic site.Only one change needs to have occurred in theresistant strains, namely a shift either in rela-tive permease activity or in relative require-ments for pHB and pAB. A diagram whichattempts to explain the interaction of the threedrugs is presented in Fig. 3. These studieswere not continued because the selection ofmutant strains of rickettsiae is time consumingand the results were not easily applicable to an

understanding of other properties of rickettsiae.Antibiotics. Penicillin and streptomycin in-

hibit the growth of rickettsiae to a small extentand are of no therapeutic value (104). Theireffect can not be discounted, however, whenrickettsiae are grown in cell culture. When themedium contains 60 Mg of penicillin and 20 Mg ofstreptomycin per ml, only R. tsutsugamushi can

be cultivated. The other rickettsiae are materi-ally affected by these concentrations of antibi-otics, and the numbers of rickettsiae that can be

TRANSPORT SITES

pHB pAB ASA|

ABp-ENZYMATIC SITES-. ASA

COMPET TIONS

FIG. 3. Attempt to represent the action of pHB,pAB, and acetylsalicylic acid (ASA) on rickettsiae.Competition may occur between two contiguous sitesat the transport sites or between pHB and pAB orbetween pAB and ASA at the enzymatic sites. Whenthe parent strain (pAB susceptible, ASA resistant)becomes pAB resistant (ASA susceptible), only onechange needs to have occurred. pAB permease activ-ity may be reduced, furnishing less pAB for competi-tion with pHB utilization, but becoming more sus-

ceptible to transport competition with ASA. Con-versely, the permeases may remain the same, but therelative enzymatic requirements may shift in favor ofpAB. Thus less pAB remains to compete with pHBmetabolism, but pAB metabolism becomes more

susceptible to competition with ASA (126).

recognized in infected cells decrease almost tothe vanishing point within a few days (22). Itwas also shown that a mixture of 60 mtg ofpenicillin and 50 ,ug of streptomycin per mlprevents plaque formation by R. rickettsi, ex-cept for pinpoint plaques appearing with veryhigh concentrations of rickettsiae (116).Only with the advent of the "broad-range"

antibiotics did chemotherapy of rickettsial in-fection become eminently efficacious. A de-tailed quantitative study of the effect of fourantibiotics on the growth of four species ofrickettsiae in eggs was carried out by Ormsbeeet al. (83) and is summarized in Table 1. Thetetracycline compounds are the most uniformlyeffective. The inhibitory dose of oxytetracyclineis about one-eighth that of chlortetracycline.There is recent evidence (C. L. Wisseman Jr.,personal communication) that the tetracyclinecompound doxycycline is even more effectivethan oxytetracycline in natural typhus infectionof man. Chloramphenicol is effective againstthe species of the genus Rickettsia in dosesabout three times greater than those needed forchlortetracycline but, surprisingly, it is rela-tively ineffective against C. burneti (Table 1).The action of erythromycin is quite variable: itis highly inhibitory of R. prowazeki, has amoderate effect on two rickettsiae of the spottedfever group, and has only a slight effect on C.burneti. Experiments with R. tsutsugamushi (7,51) indicate that the susceptibility of this rick-ettsia to the tetracycline compounds and tochloramphenicol is comparable to that of theother species of the genus Rickettsia.When rickettsiae are exposed to the tetracy-

cline compounds directly (105) or in tissueculture during adsorption (27), they are inacti-

TABLE 1. Concentration of antibiotic causing a delayof 4 days in mean death time of infected chicken

em bryosa

Antibiotic (tmol/egg)

Organism Chior- Chlor- Oxy- Eryth-amphen- tetra- tetra-royi

icol cycline cycline romycin

Rickettsia rickettsi 0.71 0.25 0.028 0.33R. akari 0.64 0.25 0.038 0.74R. prowazeki 0.83 0.22 0.036 0.032Coxiella burneti 5.90 0.28 0.032 > 3.0

aAdapted from a more extensive presentation byOrmsbee et al. (83). This work confirms and extendsthe results obtained by Jackson (51). The antibioticswere administered 24 h after inoculation of theembryos with the rickettsiae. Slightly different resultscan be expected when the antibiotics are injected justbefore rickettsial inoculation.

268 WEISS


http://mm

br.asm.org/

Dow

nloaded from



vated at a relatively rapid rate. This does nothappen when the tetracycline compounds areadded after infection has been established (7),nor does it happen with chloramphenicol (50).The action of the latter antibiotic more closelyresembles that of a metabolic inhibitor. Hoppset al. (50) exposed L cells infected with R.tsutsugamushi to the highest concentration ofchloramphenicol tolerated by the L cells inlong-term experiments, namely 5 ,g/ml. Whenthe infected cells were incubated at 35 C thenumber of rickettsiae decreased, but completesterilization of the cultures did not occur untilabout 3 weeks of incubation. Up to that time thepresence of rickettsiae could be demonstratedby subinoculation into mice or simply by re-moving the antibiotic from the cultures. Largenumbers of rickettsiae reappeared in the hostcells a week or two after the removal of chloram-phenicol. It is interesting to note that infected Lcells suspended in a nonprotein medium with-out chloramphenicol lose their rickettsiae morerapidly than those that are cultivated in acomplete medium containing 5 ug of chloram-phenicol per ml (49, 50). These results indicatethat R. tsutsugamushi, which is highly unstablein an extracellular environment, is quite stablewithin well-nourished host cells even when it isnot able to multiply.Attempts to isolate antibiotic-resistant mu-

tants in the laboratory were made by Weiss andDressler (117, 122, 123) with the Madrid Estrain of R. prowazeki. After 40 serial passagesin the presence of increasing amounts of chlor-amphenicol and one limit-dilution passage, asubstrain was isolated which appeared to be theresult of a two-step change, each involving atwofold increase in antibiotic requirement for agiven degree of inhibition. Within the limits ofthese experiments it appeared difficult to gobeyond the second step in resistance. Antibioticresistance was not lost after 10 drugless eggpassages at high concentration, which suggeststhat the susceptible strain did not have aselective advantage over the resistant strain asis the case with some bacteria. Resistance tochloramphenicol was accompanied by a corre-sponding increase in resistance to the relatedcompound thiocymetin. In another series ofexperiments an erythromycin-resistant strainwas isolated. The change probably occurredduring the first passage and involved completeresistance to this antibiotic. The parent strainwas highly susceptible to erythromycin: 0.01 mgper egg had a demonstrable inhibitory effect,and although the antibiotic had no in vitroactivity on the rickettsiae, doses of 1 mg per eggsterilized the inoculated eggs. The resistantstrain did not appear to be affected by doses as

large as 2 mg per egg. This level of resistancewas even greater than the one encountered with"wild" strains of C. burneti. Resistance toerythromycin extended to the chemically re-lated antibiotics carbomycin and olean-domycin. Attempts to isolate a tetracycline-resistant strain were unsuccessful. Several at-tempts have been made by using strains resist-ant to p-amino benzoic acid, chloramphenicol,and erythromycin to demonstrate acquisition ofdrug resistance by genetic transfer in eggs orentodermal cell cultures. These experimentshave not been successful (123).

Antibiotic-resistant strains of rickettsiaehave not been isolated in nature.

METABOLISMThe discovery by Bovarnick and Snyder in

1949 (21) that typhus rickettsiae consume oxy-gen and produce carbon dioxide in the presenceof glutamate represented a milestone in thestudy of rickettsiae. The finding has been re-peatedly confirmed and extended. It is difficult,however, to present a single outline of themetabolism of rickettsiae. Comparative studieshave been few and two main approaches havebeen followed. Bovarnick and some of the otherinvestigators who worked primarily with thy-phus rickettsiae used viable cells. This ap-proach has the advantage that it studies thephysiology of the rickettsial cells rather thanjust the presence or absence of enzymes. It hasthe shortcoming that the rickettsiae, subjectedto a long series of steps designed to separatethem from host enzymes, are injured to a lesseror greater extent and are not true representa-tives of intact resting cells. Paretsky and hiscollaborators, who studied C. burneti, chose todisrupt the rickettsial cells and thus eliminatethe problems of transport and possibly reducethe effect of cell injury. The third approach, thestudy of the metabolism of rickettsiae multiply-ing inside host cells by the judicious use ofradioisotopes and metabolic inhibitors, hasbeen used only recently. Ideally all thesemethods should be used on the same rickettsia,but since this has not been done, typhus rickett-siae and C. burneti will be treated separately.The relatively few experiments done with spot-ted fever rickettsiae will be integrated with theformer. The metabolism of scrub typhus rick-ettsiae was not studied except indirectly asdescribed in the section on cell culture.

Typhus and Spotted Fever RickettsiaeSubstrates that stimulate respiration. Res-

piration of rickettsiae in the absence of addedsubstrate is negligible, but it is stimulated by a

269VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

moderate number of substrates, most vigorouslyby glutamate (21). The rate of utilization of thissubstrate is considerably higher in typhus thanin spotted fever rickettsiae and varies widelyfrom preparation to preparation. In terms ofoxygen uptake per milligram of protein at 32 to34 C, it seldom exceeds 1.5 gmol/h and often ismuch lower, and it correlates well with otherbiological activities such as infectivity for thechicken embryo, mouse toxicity, or hemolyticactivity (13, 14, 48, 133).The second most stimulating substrate of the

respiration of both typhus and spotted feverrickettsiae is glutamine (44, 133). The ratiobetween respiration stimulated by glutamateand glutamine is approximately 3:2. It was

shown by Hahn et al. (44) that glutamine isdeamidated by R. typhi to glutamate and thatthis is probably the only pathway by whichglutamine is metabolized by resting (extracellu-lar) rickettsiae. Since glutamine is an impor-tant metabolite of cultivated mammalian cells,required in concentration second only to glu-cose, this reaction deserves further investiga-tion. It is not known, for example, whether themore rapid stimulation of respiration by gluta-mate is due to more rapid transport of thiscompound or to the fact that the deamidation ofglutamine is the rate-limiting reaction.

Other substrates stimulate the respiration ofrickettsiae to a much lesser extent and thesereactions can be demonstrated by conventionalmanometric techniques only with unusuallyactive preparations. Among them are pyruvateand the dicarboxylic acid intermediates of thecitric acid cycle (141). The same is true of thespotted fever rickettsiae (91).

Glucose, glucose-6-phosphate, lactate, su-

crose, and the naturally occurring amino acids,except for glutamate, do not stimulate respira-tion (21). The failure of rickettsiae to catabolizeglucose has been confirmed by sensitive radi-oisotope techniques which measure CO2 pro-duction (85, 134). It is not known whether this isdue to lack of the proper permeases, or one or

two key glycolytic enzymes, or the entire glyco-lytic system (see Q fever rickettsiae below).The evolutionary significance of the ability of

rickettsiae to catabolize glutamate and theirinability to utilize glucose is difficult to assess,but a few cautious comparisons can be madewith other organisms. Both Neisseriameningitidis (69) and Brucella abortus (38, 39)metabolize glutamate more rapidly than anyother substrate, utilizing it for growth and as a

source of energy. The rate of 02 uptake of a

typical suspension of resting cells of N.meningitidis is about 10 ,umol per h per mg of

protein, or 6 to 12 times greater than that of acomparable preparation of rickettsiae (69). Therate for B. abortus is lower than that of N.meningitidis, and it was shown by Dasinger andWilson (32) that a virulent strain of B. abortusoxidizes glutamate at about half the rate(1.5-2.0 pmol of 02 uptake per h per mg ofprotein) of an avirulent strain. This is possiblyrelated to the finding by Freeman et al. (36)that avirulent brucellae can not be cultivated inguinea pig monocytes, not because they do nothave the ability to grow in them, but becausethey destroy them too rapidly. If experiencewith brucellae is applicable to rickettsiae andmanometric experiments reflect intracellularevents, the rate at which rickettsiae removeglutamate plus glutamine from the host cell fortheir own use is probably about as high as canbe expected. A higher rate would lead to thedestruction of the intracellular environment.Pathway of glutamate utilization. Gluta-

mate metabolism of resting rickettsiae does notappear to be different from that of other micro-organisms even though some of the reactionsrequire further elucidation. Three end productshave been recognized: ammonia, carbon diox-ide, and aspartate.The amino group of some of the glutamate is

released as ammonia, but Bovarnick and Miller(16) have shown that most of the amino group istransferred to oxaloacetate with the formationof aspartate. The enzyme involved in this reac-tion, glutamate-oxaloacetate transaminase, isone of the most stable rickettsial enzymes.Rapid freezing and thawing of rickettsial sus-pensions, which destroys most of the oxidativeenzymes, greatly enhances this activity. Theaddition of pyridoxal phosphate is not required,suggesting that the enzyme remains saturatedwith its cofactor (48). The transaminase mayalso catalyze the reverse reaction: respiration inthe presence of a-ketoglutarate plus aspartate isalways much greater than the sum of therespirations that take place when the two sub-strates are added singly. Presumably, there isenough extracellular enzyme in the cell suspen-sions to elicit the formation of glutamate, whichis then rapidly transported into the cell andoxidized. Bovarnick and Miller (16) attributedthe origin of the extracellular enzyme to thedisrupted cell which is always present in arickettsial suspension. The possibility shouldnot be overlooked, however, that the transami-nase is associated with a surface structure.Carbon dioxide is produced from all carbon

atoms of glutamate, although it is not certainthat glutamate is completely oxidized (96).a-Ketoglutarate, succinate, fumarate, malate,

270 WEISS


http://mm

br.asm.org/

Dow

nloaded from



oxaloacetate, and pyruvate are intermediateproducts of the reaction and do not accumulatewithout specific inhibitors (141). Pyruvate isdegraded with production of CO2 from all threecarbons (133). Attempts to demonstrate theformation of citric acid and other tricarboxylicacids have not been successful, but this mighthave been due to technical difficulties ratherthan to the total absence of the correspondingenzymes in rickettsiae (141).

Hayes et al. (46) used the sensitive spectro-photometric method of Chance (26) to studyelectron transport in suspensions of intactrickettsiae. Their studies indicated that R.typhi has a flavine enzyme-iron-cytochrome sys-tem which probably includes cytochromes a,and b. Although the terminal respiratory en-zymes are comparable to those of other bac-teria, the specific activity of the cytochromes isrelatively low.Production of ATP. Bovarnick (10) used an

ingenious method for the demonstration of ATPformation by rickettsiae respiring in the pres-ence of glutamate. The test system consisted ofinorganic phosphate, adenosine diphosphate(ADP), hexokinase, and glucose. Under theseconditions ADP was phosphorylated to ATP,and glucose-6-phosphate was formed. The reac-tion required relatively large amounts of inor-ganic phosphate and ADP and was stimulatedby the addition of small quantities of NAD andcoenzyme A (CoA). Approximately 0.2 to 0.3,gmol of glucose-6-phosphate was formed foreach atom of oxygen consumed. The reactionwas completely inhibited by cyanide, indicatingthat a myokinase type enzyme, i.e., one con-verting two molecules of ADP to one of ATPand one of adenylic acid, was not involved. Di-nitrophenol, as expected, inhibited phosphoryl-ation in concentrations that did not interferewith respiration.

Admittedly, in the above-described experi-ments the rickettsiae were maintained in ahighly artificial environment. In fact, the addi-tion of ADP (or ATP) to glutamate reducedrespiration by 30 to 40%, possibly because itinterfered with the proper utilization of thecorresponding endogenous factors. More satis-factory evidence of oxidative phosphorylationwas obtained by Bovarnick and Allen (13) bythe direct measurement of the ATP content ofrickettsiae incubated under various conditions.Starved rickettsiae, namely rickettsiae in-cubated at 36 C for 3 h without substrate,contained no measurable ATP. When thesestarved rickettsiae were incubated with gluta-mate for 2.5 h at 30 C, the ATP level rose to 1.5to 2.0 gimol per mg of rickettsial protein. When

adenylic acid was also added, somewhat higherlevels of ATP were obtained. The highest ATPlevels demonstrated in rickettsiae were consid-erably lower than those generally encounteredamong viable bacteria. It is possible that rick-ettsiae are not capable of achieving an endoge-nous ATP level sufficiently high to sustain alltheir synthetic functions, and this may be onereason why they are incapable of independentexistence.

Synthetic activities. It was shown by Bovar-nick et al. (11, 18, 20) that rickettsiae arecapable of synthesizing small amounts of pro-tein and lipid. One of the simplest media thathas supported protein synthesis is shown inTable 2. Synthesis was demonstrated by replac-ing one of the amino acids with one that wasradioactively labeled and showing that it wasincorporated into the trichloroacetic acid insol-uble fraction. Incorporation was inhibited bychloramphenicol, which provided further evi-dence that it represented protein synthesis. Toincrease the sensitivity of the test, the labeledamino acid was added at a concentration lowerthan the rest, 0.005 to 0.01 instead of 0.1 mM.Also, low concentrations of rickettsiae (30 to 40,ug of protein per ml) were used because higherspecific activities were obtained with theseamounts.

It is not certain that all of the constituentslisted in Table 2 are essential for optimal aminoacid incorporation. However, the observationsmade on the effect of some of these compoundson incorporation (18) depicted better than anyother experiments the physiology of rickettsiae,or at least of rickettsiae that had been subjectedto the lengthy procedure of purification.Optimal incorporation requires a K+ concen-

tration of at least 0.13 M. When all or asignificant portion of the K salts are substitutedby Na salts or sucrose, incorporation is greatlyreduced. Mg2+ is an absolute requirement,whereas Mn2+ is not essential but stimulatory.There is good evidence that all of the naturalamino acids must be present for the incorpora-tion of one. Methionine-35S, glycine-"4C (orglycine-2-'4C), or valine-1-14C, used individ-ually as the labeled amino acids, yielded com-parable results. When all of the amino acids areomitted except the one that is labeled, incorpo-ration is negligible. The same results are ob-tained when a single amino acid, such as serine,threonine, or valine, is omitted. One aminoacid, glutamine, must be added in far largerconcentrations than the others because it servesthe dual function of providing energy as well asa source of glutamate for protein synthesis.Glutamine and glutamate can be used inter-

VOL. 37, 1973 271


http://mm

br.asm.org/

Dow

nloaded from


TABLE 2. Composition of medium sspporting amino acid incorporation in typhus rickettsiaea

Final Final

Constituent concen- Constituent concen-

tration tration(mM) (mM)

Salts Amino acids-ContinuedKCl .............................. 160 L-isoleucine.0.20K2HPO4.4.5 L-lysine.0.15KH2PO4.0.8 L-methionine.0.05MgC2.1.8 L-phenylalanine.0.08MnC12............. 0.05 L-aspartic acid ........................ 0.10FeC12.0.008 L-glutamine.5.0Sodium acetate.1.2

Nucleotides and cofactors5Amino acids ATP.... .... 1.0

DL-alanine.0.20 CMP.. 0.12L-arginine ............................. 0.10 GMP.. 0.12L-asparagine .......................... 0.10 UMP.. 0.12L-cysteine ............................. 0.04 NAD.. 0.38L-histidine ............................ 0.04 NADP.. 0.03L-proline.0.10 CoA.. 0.038L-hydroxyproline.0.10 Cocarboxylase.. 0.055L-serine.0.10 Glutathione.. 0.5L-threonine.0.15L-tryptophane.0.015 Protein (bovine plasma albumin orL-tyrosine.0.10 yolk sac protein) .......... 0.25%L-valine... 0.15Glycine... 0.10 IndicatorL-leucine.... 0.15 Phenol red...... 0.005%

a Adapted from Bovarnick and Schneider (18) on the basis of personal communication from Bovarnick.bATP, adenosine triphosphate; CMP, cytidine monophosphate; GMP, guanosine monophosphate; UMP,

uridine monophosphate; NAD, nicotinamide adenine dinucleotide; NADP, nicotinamide adenine dinucleotidephosphate; CoA, coenzyme A.

changeably. An early report (20) that glutamatecould not replace glutamine in supportingmethionine-35S incorporation was an error dueto the contamination of glutamate with methio-nine, which reduced the specific activity of theradioactive label (18).

Despite the fact that glutamine, or gluta-mate, must be added as an energy-yieldingsubstrate, rickettsiae must also be providedwith exogenous ATP (18). This need can in partbe met by the addition of ADP. Of course, it isnot known to what extent this requirementreflects cell damage, but in any case it is anunusual bacterial requirement shared withChlamydia (135) and only rarely and inconclu-sively described in other genera. In contrast toChlamydia, rickettsiae need two sources ofATP, one to be generated endogenously fromglutamine and one to be provided exogenously(18). This dual requirement suggests that cer-tain functions are localized at the surface andothers in the cytoplasm.A large number of compounds were tested for

their effect on incorporation. In addition to

ATP, the other ribonucleotides are needed formaximal incorporation, whereas the deoxyribo-nucleotides do not affect it. NAD and reducedglutathione are essential. The effect of glutathi-one is so pronounced that, possibly, it does notact simply as a reducing agent but has cofactorfunction. Nicotinamide adenine dinucleotidephosphate (NADP), CoA, and cocarboxylase donot generally affect incorporation, but they areusually added because it was shown in otherexperiments that they increase rickettsial sta-bility (18). Various mixtures of vitamins ap-peared to be without effect (Bovarnick, personalcommunication).The incorporation medium must also contain

a protein. Its chief function is undoubtedly toincrease the stability of the rickettsiae. It mayalso play a specific role in metabolism byadsorbing critical nutrilites and, in some un-known manner, making them more readilyavailable to the rickettsiae. Many natural pro-teins are toxic and cannot be used. Bovarnickfound that only certain lots of bovine plasmaalbumin were satisfactory and obtained most

272 WEISS BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



consistent results with a fraction of yolk sacprotein extracted with acetone and ammoniumsulfate (20).

In most of Bovarnick's experiments (18) in-corporation was sustained for several hours. Ingeneral, the radioactivity in the trichloroaceticacid fraction was almost twice as high after 24 hof incubation as after 5 h. Maximal incorpora-tion amounted to 10 gmol of amino acid per mgof rickettsial protein. A twofold increase inrickettsial protein would have required incorpo-ration at least 100 times greater.

Bovarnick has also shown that rickettsiaesynthesize lipids (11). Judging from the amountof acetate-i- "4C incorporated, lipid synthesiswas minute, 0.25 umol of acetic acid incorpo-rated per mg of rickettsial protein with the moreactive preparations. However, it is probablethat the total amount of lipid synthesized wasmuch greater and that most of it was derivedfrom acetyl-CoA produced from unlabeled glu-tamate (or glutamine) rather than from labeledacetate.The energy requirements for lipid synthesis

are similar to those described for protein syn-thesis. However, a mixture of amino acids,except for the energy-yielding glutamine orglutamate, is not required. Reduced pyridinenucleotides usually enhance incorporation (11).

It is difficult to assess the significance of theminute synthetic activities demonstrated withpurified rickettsiae. Do they reflect the extent ofthe messenger RNA remaining with the orga-nisms, the residual activities of enzymes whichare derepressed only in an intracellular environ-ment, or activities that can be further stimu-lated by greater wisdom in formulating themedium? Bovarnick compared rickettsiae tocell particulates rather than to intact cells andshe was keenly aware of the fact that therickettsiae used in her study had been injuredduring the process of purification and at-tributed her results, in part, to cell injury (18).Physiology of cell injury. In a classic paper

published in 1950, Bovarnick et al. (17) for-mulated a medium usually designated SPG,which favors the survival of a number of strainsof rickettsiae. Its composition is based on theobservations that rickettsiae are more stable ina sucrose than in a saline solution, in a salinesolution high in K+ rather than one high in Na+,and in a pH close to 7.0 rather than 7.6. Alsoessential is the presence of a moderate amountof glutamate. The stability of rickettsiae isfurther increased by the addition of serumalbumin.This medium has been widely used for the

storage of rickettsial suspensions in the frozenstate, for their dilution before titration, and forother experimental procedures. Some more re-cent evidence indicates, however, that thismedium may not be needed by intact cells, butcompensates for cell damage and that, in somecases, it does not increase the stability ofrickettsiae.Myers et al. (79) studied the permeability of

the cell membrane of R. typhi to sucrose andvarious electrolytes by phase-contrast micros-copy observations of plasmolysis. The resultsclearly indicated that freshly harvested, unpuri-fied rickettsiae had permeability propertiesvery similar to those of E. coli. The cells wereplasmolyzed by high concentrations of sucrose,NaCl, or KC1. This means that these com-pounds penetrated into the spaces between cellwall and cell membrane and did not cross intothe cytoplasm, which shrank because of in-creased osmotic pressure. In contrast, when therickettsiae were frozen and thawed or justsubjected to a lengthy process of purification,the cells remained impermeable to sucrose butbecame permeable to the salts. These observa-tions were confirmed by optical density mea-surements and radioisotope dilution techniques.Thus, it appears that the high K+/Na+ ratiorequired by purified rickettsiae is a reflection oflost ability to control their internal electrolyteratio. Intact cells may not have this require-ment.

Stoenner et al. (109) showed that spottedfever rickettsiae suspended in SPG lost abouthalf of their viability during the brief periodrequired to dilute them and to inject them intoeggs. Rickettsiae suspended in phosphate-buff-ered saline did not suffer such a loss under thesame conditions. These surprising results canbest be interpreted by assuming that enhance-ment of catabolic activity by the glutamate ofSPG does not always increase the stability ofrickettsiae. Rees and Weiss (96) demonstratedthat glutamate stabilized the metabolic activityof spotted fever rickettsiae most effectively inthe presence of reduced glutathione and in a gasphase which contains very low levels of 0,2 It ispossible that in the experiments of Stoenner etal. (109) the redox potential was high enough topermit the accumulation of toxic levels of hy-drogen peroxide. This is probably not an iso-lated incident since instances have been en-countered in which glutamate is detrimental totyphus rickettsiae suspended in an unfavorablemedium (14).The foregoing examples serve to illustrate

that most of the observations of the physiology

VOL. 37, 1973 273


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

of rickettsiae have of necessity been made ondamaged cells or cells maintained under un-physiological conditions. There are two types ofcell damage that have been produced in thelaboratory and studied extensively. One is thedamage caused either by prolonged incubationof the rickettsiae at 0 C in saline solutions or byrepeated freezing and thawing (12, 14). Theother is starvation resulting from incubation ofthe rickettsiae without substrate at 36 C forseveral hours (13). Both types of damage areoften compared to those produced on mito-chondria or protozoa during procedures of isola-tion. As in the case of mitochondria, some ofthis damage can be prevented or repaired.

It was shown by Bovarnick et al. (15) that thesurvival of typhus rickettsiae, as measured byegg infectivity, mouse toxicity, hemolytic activ-ity, and respiration, is increased in the presenceof NAD. This suggests that rickettsiae slowlylose this coenzyme and that its presence in themedium prevents this loss. In fact, it was shown(12, 14) that when rickettsiae are suspended in asaline solution and frozen and thawed or simplymaintained at 0 C for 18 h and then centrifuged,endogenous NAD no longer sediments with thecells but remains in the supernatant fluid, andthe rickettsiae lose biological activity. In addi-tion to NAD, rickettsiae probably lose othercoenzymes such as CoA, and divalent ions,notably Mg2+. If the damage is not too severe,rickettsiae can regain their biological activitywhen they are incubated with NAD, CoA, anddivalent ions for 3 h at 33 C.When rickettsiae are maintained at 36 C for

several hours in the absence of substrate, loss ofbiological activity is accompanied by an almostcomplete disappearance of endogenous ATP(13). This loss can be prevented or restored byincubating the rickettsiae with glutamate orpyruvate. Added ATP, which is very effective inpreventing loss of biological activity, can notalways restore it.The reaction of rickettsiae with exogenous

ATP is a very curious one, as illustrated by theexperiments of Bovarnick and Schneider (19).When the rickettsiae are freshly harvested fromyolk sac, purified, and used immediately, theyhemolyze sheep erythrocytes quite well in thepresence of glutamate, but this reaction isstrongly inhibited by exogenous ATP, and withATP alone there is very little hemolysis. As therickettsiae are progressively damaged, hemol-ysis with glutamate declines, ATP becomes lessinhibitory, and hemolysis with ATP alone in-.creases. Eventually rickettsiae are no longerable to hemolyze sheep erythrocytes in the

presence of glutamate, but only in the presenceof ATP.

It is not known to what extent the mechanismof cell damage and repair in rickettsiae differsfrom that of other bacteria. It is quite possiblethat the inability of rickettsiae to grow withouthost cells facilitated the studies of phenomenathat may easily be overlooked in bacteria inwhich recovery is promptly followed by growth.It is tempting to postulate, however, that rick-ettsiae, unlike most other bacteria, must beable to interact with coenzymes and with otherphosphorylated compounds within the host celland that the above-described phenomena arejust examples of such capabilities. Further-more, there is good evidence that there arenatural fluctuations in rickettsial virulencewhich involve loss and gain in known andunknown factors. This phenomenon has beenencountered in the spotted fever rickettsiae andis usually called "reactivation".

Reactivation. It was shown by several of theearly investigators of Rocky Mountain spottedfever that unfed adult Dermacentor andersoniticks often contain rickettsiae that immunizeguinea pigs but produce no recognizable illness.When ticks from the same batch are allowed tohave a blood meal and the contents are injectedinto guinea pigs, a typical disease follows.Spencer and Parker (106) reproduced the samephenomenon in the laboratory by injecting tickswith virulent rickettsiae and storing them atrefrigerator temperature for several months. Asin the case of the natural infection, the rickett-siae remained immunogenic but were not viru-lent for guinea pigs, and virulence was fullyreestablished when the ticks had a blood mealor were incubated at 37 C for 24 to 48 h.

Price (92, 93) obtained convincing evidencethat this phenomenon could not be attributedto an increase in the number of rickettsiae or toa population shift. It was a true physiologicalchange which Gilford and Price (40) were ableto reproduce in vitro. They incubated infectedyolk sac suspended in SPG at 25 C for 60 h, andshowed that under these conditions rickettsiaelost their infectivity for eggs and guinea pigs.However, when 36 mM pAB was also added tothe suspension, about 1% of the egg infectivitywas retained but the rickettsiae were no longervirulent for guinea pigs. The addition of 15 mMpHB instead of pAB, or combinations of pABand pHB, or pAB and 0.6 mM CoA resulted inretention of guinea pig virulence as well as egginfectivity. The lost guinea pig virulence ofrickettsiae that were still infectious for eggscould be restored by incubation at 25 C for 6 h

274' WEISS


http://mm

br.asm.org/

Dow

nloaded from



with NAD or CoA, but not with pHB. Virulencewas also restored by incubating the rickettsiaeat 37 C for 24 h with suspensions of ground upadult ticks or nymphs that had received a bloodmeal. The factors in tick extracts responsible forthis conversion have not been identified.

It is not known whether reactivation of spot-ted fever rickettsiae is analogous to the previ-ously described reversible inactivation of ty-phus rickettsiae. These two phenomena havepoints of similarity but differ in one importantrespect. Typhus rickettsiae lose and regain all oftheir biological activities, including egg infec-tivity, at about the same rate, whereas spottedfever rickettsiae lose guinea pig virulence morerapidly than egg infectivity. Because a singleegg passage will restore the guinea pig virulenceof R. rickettsi, it is quite possible that the yolksac of the chicken embryo offers to the spottedfever rickettsiae an especially favorable envi-ronment where activity is quickly restored andwhere partial cell damage can not be demon-strated. If this is the case, both phenomenaentail the loss of important cofactors, but thesteps involved in restoration of activity varysomewhat.Metabolism of rickettsiae multiplying in

host cells. Alexander (1, 2) has shown thatcycloheximide, an inhibitor of eukaryotic butnot of prokaryotic protein systems (34, 90), is anexcellent tool for the study of the metabolism ofchlamydiae multiplying in cell culture. Thissame compound has been used recently to studythe independent metabolism of multiplyingrickettsiae (127, 131). Four rickettsial specieshave been studied, namely R. typhi, R. akari, R.rickettsi, and R. tsutsugamushi. The host cellwas the nonmultiplying (irradiated) L cell,except that irradiated duck embryo fibroblastswere used for R. rickettsi. The experimentaldesign was a very simple one: the incorporationof a mixture of 15 "4C-labeled amino acids or of"4C-labeled adenine was measured at intervalsin uninfected and infected cells both in thepresence and absence of cycloheximide. Theresults clearly indicated that rickettsial multi-plication was not affected by cycloheximide andthat rickettsial infection elicited a highly signif-icant increase in cycloheximide-resistant incor-poration of amino acids and adenine into thetrichloroacetic acid insoluble fractions. With R.rickettsi cycloheximide-resistant activity wassmall, possibly because this rickettsia neverachieves a high density even within its mostsusceptible host cell. With the other threerickettsiae, at the peak of the infection cy-cloheximide-resistant activity reached 30 to

60% of the total. The above-described experi-ments leave little doubt that protein and nu-cleic acid syntheses accompanying rickettsialmultiplication are of the prokaryotic type andthat therefore these activities can be attributedto the rickettsiae.

Q Fever RickettsiaeC. burneti is one of the sturdiest of the

nonsporogenic organisms. When suspended indistilled water or sterile milk and stored at 4 Cor room temperature, viability is retained formonths and sometimes years. Flash pasteuriza-tion is not usually effective in sterilizing milkinfected with C. burneti and there have beenreports of incomplete inactivation of organismsmaintained at 63 C for 40 min. The same is trueof chemical sterilization. For example, 1% For-malin will inactivate C. burneti in 72 but not 24h, and viability has been demonstrated after 4days of storage at 4 C in 0.5% Formalin. C.burneti is not immediately inactivated by 5%acetic acid, 5% NaOH, or acetone (6, 95).One would be led to believe, therefore, that C.

burneti is ideally suited for the study of meta-bolic reactions and may well serve as theprototype of obligate intracellular bacteria. Un-fortunately, nothing is further from the truth.Not only do intact resting cells display verylittle metabolic activity, but this activity isgreatly affected by the composition of the sus-pending medium and by added cofactors (R. A.Ormsbee, personal communication). Althoughthe optimal conditions for the demonstration ofindependent metabolic activity have not beenelucidated, one must conclude that resting cellsof C. burneti are relatively inert. Possibly, thischaracteristic, inability to interact with theenvironment unless conditions are favorable forgrowth, may be responsible for their high degreeof stability.The chief substrate of resting cells appears to

be pyruvate, as shown by Ormsbee and Peacock(84). The rate, 0.25 umol of 02 consumed per hper mg of rickettsial protein, is about one-sixthof the rate of respiration of typhus rickettsiae inthe presence of glutamate. Respiration of C.burneti can also be stimulated by oxaloacetate,succinate, serine, and as shown with considera-ble difficulty, by a-ketoglutarate, fumarate,and malate. Glutamate is utilized only in thepresence of added NAD.More rewarding have been the experiments

with disrupted cells carried out by Paretsky andhis collaborators (86). Many of these experi-ments included tests with intact cells whichdisplayed very little, if any, activity. Paretsky,

275VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

on the strength of his investigations, came tothe conclusion that Q fever rickettsiae have theprincipal physiological functions expected ofbacteria. His findings do not explain why C.burneti can not be grown in a host cell-freemedium.Most surprising is the evidence that dis-

rupted preparations of C. burneti have someenzymes of glucose metabolism, especially sincethe most sensitive methods have failed to dem-onstrate these activities with intact cells (85).The existence of intracellular hexokinase, glu-cose-6-phosphate, and 6-phosphogluconatedehydrogenases and more recently of aldolaseactivity and of other enzymes of the Embden-Meyerhof pathway have been demonstrated(29, 64, 65, 87). There is good evidence that Qfever rickettsiae have most, if not all, of theenzymes of the citric acid cycle. Paretsky andhis co-workers (29, 88) have demonstrated thesynthesis of citric acid from oxaloacetate and, toa smaller extent, from acetate and acetyl phos-phate in the presence of NAD, CoA, and ATP.Oxidation of isocitrate, glutamate, and malatewith reduction of their respective cofactors,NADP or NAD, has also been shown. It is notknown whether glucose and citrate metabolismrepresent a true difference from events in ty-phus rickettsiae, since comparable experimentshave not yet been done with the latter orga-nisms.

In addition to the above-described reactions,which can be regarded as indicators of energy-yielding fuctions, Paretsky and his associatesshowed that disrupted Q fever rickettsiae havesynthetic capabilities. Myers (75) demonstratedthe presence of folic acid in C. burneti, andMattheis et al. (71) isolated several types offolates, some with chromatographic elution pat-terns quite different from the folic acid deriva-tives of chicken embryos. These findings led tothe investigation of reactions catalyzed by thefolic acid series. Myers and Paretsky (78)showed that enzyme preparations of C. burnetican form serine from glycine and formaldehydein the presence of tetrahydrofolic acid. Theseresults, in turn, directed the attention of Mal-lavia and Paretsky (67) to two other one-carbontransfers, both involving carbamyl phosphate.They showed that citrulline was formed fromornithine and carbamyl phosphate and thereaction possibly proceeded to the synthesis ofarginine. The pyrimide precursor ureidosucci-nate was produced from aspartate and carba-myl phosphate in the presence of ATP.Demonstration of synthesis of mac-

romolecules was the next step. Mallavia andParetsky (68) showed that cell-free preparations

of C. burneti incorporated minute amounts ofleucine, phenylalanine, and algal protein hy-drolysate into hot trichloroacetic acid insolublefractions. Evidence of a rickettsial DNA-dependent RNA polymerase was provided byJones and Paretsky (52), who incubated rickett-sial extracts with the four ribonucleoside tri-phosphates plus energy-generating compoundssuch as phosphoenolpyruvate and pyruvate ki-nase. When one of the four ribonucleosidetriphosphates was labeled, it could be shownthat it was incorporated into an insoluble tri-chloroacetic acid fraction. Incorporation wasdependent on the presence of all four ribonu-cleosides and was enhanced by either heterolo-gous or homologous DNA. It was inhibited bydeoxyribonuclease or actinomycin D. The syn-thesized compound was destroyed by ribonu-clease or dilute alkaline hydrolysis, and thus itwas shown to have the properties of a polyribo-nucleotide.

Again, there is no reason to believe that C.burneti has enzyme systems not possessed bytyphus rickettsiae. It just happened that earlyfrustrations with disrupted typhus rickettsiaeon the one hand and cell suspensions of Q feverrickettsiae on the other led investigators todifferent approaches and to findings of differentsignificance.

WHY HAS INDEPENDENTCULTIVATION NOT BEEN ACHIEVED?The question of failure to grow rickettsiae in a

cell-free medium can be answered only in ahighly speculative manner. Pertinent informa-tion, as we have seen in previous sections, is atbest incomplete and in some cases misleading.Furthermore, any educated guessing must bebased for the most part on negative results, andthis is very dangerous. A single successful exper-iment may invalidate conclusions based onmany well-planned and well-executed experi-ments in which rickettsiae have not grown.Speculations are necessary, however, becausethey constitute the substrate on which newexperiments must feed. A current textbookincorrectly refers to rickettsiae as "tiny intracel-lular parasites," implying that the genome istoo small to code for all functions necessary forindependent existence. As we have seen, thishypothesis is not tenable (54). Some of the moreobvious reasons why rickettsiae have not yetbeen cultivated in the absence of host cells arebravely discussed below.

Energy ParasitismThere is strong circumstantial evidence from

the investigations carried out in the laboratories

276 WEISS


http://mm

br.asm.org/

Dow

nloaded from



of Weiss and of Moulder (72-74, 118, 135) thatchlamydiae derive energy from the host cell. Insharp contrast, isolated rickettsiae derive ATPfrom the oxidation of glutamate (10), and thisactivity appears to be essential for penetration(27). This finding does not exclude the possibil-ity that rickettsiae, once they are inside theirhost cells, are energy parasites. In fact,, asdiscussed in a previous section, it was shownthat for protein and lipid syntheses exogenousATP must also be supplied (11, 18, 20). Al-though a requirement for exogenous ATP is rareamong bacteria, it is not uncommon to find thatseparate energy-yielding reactions take placefor transport and for synthetic activity (47, 69,119). It is therefore conceivable that rickettsiaeare capable of meeting their endogenous ATPrequirements but depend on the host for energyfor the transport of key metabolites. A criticalexperiment has probably not been performedwhich would exclude the possibility that ATP isa major requirement for the growth of isolatedrickettsiae, since it is not known that all theother requirements have been met, and it can beassumed that in most experiments the cellswere damaged to some extent. Bovarnick andSchneider (19) showed that the hemolytic activ-ity of undamaged rickettsiae is inhibited byconcentrations of ATP as low as 1 mM. Theideal experiment should contain a system gen-erating ATP or other energy-yielding com-pounds at a low constant rate with maximalconcentrations well below the toxic level.

LeakinessThe remarkable effects of the pyridine nu-

cleotides and CoA on maintainance and restora-tion of activity of typhus and spotted feverrickettsiae, discussed in a previous section, wereproduced on damaged cells. It is not certainthat intact cells require cofactors, although thisis a distinct possibility, but it appears that theirinclusion in the medium prevents their leakagefrom the cell. Both Moulder (74) and Hanks(45) have emphasized that a successful intracel-lular parasite must be able to release and acceptnumerous factors and create within the host cella microenvironment in which the required me-tabolites are in proper balance. In an extracellu-lar environment this "leakage" may be a one-way street and may lead to the steady deteriora-tion of the rickettsiae. As we have seen, variousfactors, such as moderately high osmolarity, ahigh K +/Na + ratio, sucrose, and serum al-bumin, retard leakage and stabilize rickettsiae,but these factors may not be adequate when thevolume is disproportionate to the number ofrickettsiae. If "leakiness" is indeed the majorfactor in frustrating extracellular rickettsial

growth, the ideal experiment must take intoconsideration the maximal volume of mediumthat a given number of rickettsiae can tolerate.The extraordinary stability of C. burneti mayindicate that leakage does not occur with thismicroorganism or that the lost compounds canbe efficiently reacquired in an intracellularenvironment. There is no experimental evi-dence yet to distinguish between these twopossibilities.

Unidentified AuxotrophyThe record of the attempts that have been

made to grow rickettsiae on established orexperimental cell-free media is not complete.All of these experiments have been, of course,unsuccessful, and most of them have not beenpublished. For example, the work of Bovarnickon protein and lipid syntheses (11, 18, 20) incomplex media was the result of frustration overattempts to cultivate the organisms in theseand other media. There is always the possibilitythat in each instance a key metabolite was notadded. The search for this metabolite mustcontinue, but one must keep in mind thatcompounds that are not needed are often toxicor antagonistic. This is particularly true ofnatural products such as tissue extracts. Amedium containing a judicious but conservativenumber of defined chemicals is more likely to besuccessful than one that is too generous.Highly Sensitive Regulatory MechanismExtracellular rickettsiae are probably analo-

gous to resting bacteria and display limitedenzymatic activities. A suitable environmentmay be required for the induction of some oftheir synthetic enzymes. It is conceivable thatthe balance between repression, induction, andend-product inhibition is more delicate in rick-ettsiae than in other bacteria and that it isexquisitely adapted to the intracellular environ-ment. The evolutionary advantage of a sensitivebalance is obvious. The rickettsiae would beable to multiply only when conditions are rightin an intracellular environment. As multiplica-tion progresses and some of the host-cell metab-olites become scarcer, induction would stop, orend-product inhibition would start, before thehost cell is irreparably damaged. This is actu-ally the situation in many relationships betweenrickettsiae and their natural host cells. In thelaboratory we strive to produce as many rickett-siae as possible and we disrupt this delicatebalance. We select a host cell that can inducethe synthetic enzymes of the rickettsiae butsends unclear signals for inhibition until infec-tion proceeds to the total destruction of the hostcell. The resulting rickettsiae most likely have

277VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

enzyme systems which have been 'abruptlyrepressed. If this is indeed the reason whyrickettsiae have not been cultivated in hostcell-free medium, it represents a serious obsta-cle. Information on mechanisms of enzymeregulation in rickettsiae is urgently needed.

CommentIt would be foolhardy to rank the above-listed

reasons for failure to cultivate rickettsiae in-dependently of their host cells in order ofprobability. Attention must be paid to all ofthese possibilities. A brilliant new idea for thegrowth of rickettsiae will not bear fruit if all ofthe other requirements are not adequately met.It can not be overemphasized that all workshould be done with rickettsiae damaged aslittle as possible. It might be well to assumethat any rickettsia artificially separated fromhost cells has suffered some damage and shouldbe treated with the care usually accorded toprotoplasts or mitochondria. This applies alsoto C. burneti, whose enormous ability to remaininfectious may belie the damage that mighthave been inflicted to its metabolic capabilitiesin an extracellular environment.The reader may be tempted to ask to what

extent the successful cultivation of Ro-chalimaea quintana on blood agar (115) canpave the way for the independent growth ofrickettsiae. R. quintana is extracellular in itsnatural host vector and possibly in man, and itgrows with difficulty in eggs or tissue culture.Its most notable requirement for growth ishemin (77), which probably is not needed byrickettsiae. With the benefit of hindsight, onemay say that the cultivation of R. quintana onblood agar was not surprising and contributedlittle to the solution of our problem of cultivat-ing rickettsiae without host cells. We can learna great deal from basic studies of many nutri-tionally fastidious bacteria or auxotrophic mu-tants and by keeping faith that the tenets ofcomparative biochemistry apply also to rickett-siae.

A FINAL REMARKA review usually deals with a field that is

rapidly expanding. This review deals with anarea of microbiology which at best has receivedonly moderate attention in recent years. Half ofthe publications quoted here were written morethan 10 years ago. As a result, some of theconcepts and techniques described are obsoleteand much of the information required for acomprehensive discussion of the biology of rick-ettsiae is lacking. This waning interest may in

part stem from a dubious sense of security thatrickettsial infections no longer pose grave publichealth problems, but more importantly it alsoderives from the misconception that rickettsiaehave little to offer to the student of basic biologyor that the efforts needed for meaningful infor-mation are hardly worth the expected results.Neither attitude can be well defended. Moderntechnology is providing numerous new avenuesof approach to the study of these microorga-nisms which have evolved what appears to be amost ingenious mechanism of interaction withtheir microenvironment. Our reticence to workwith rickettsiae is no longer justified.

ACKNOWLEDGMENTS

This work was supported in part by the Bureau ofMedicine and Surgery, Department of the Navy, researchtask MR041.05.01-0007A6GI.

I am indebted to Richard A. Ormsbee, Paul Fiset, andWilly Burgdorfer for their helpful discussions.

LITERATURE CITED

1. Alexander, J. J. 1968. Separation of proteinsynthesis in meningopneumonitis agent fromthat in L cells by differential susceptibility tocycloheximide. J. Bacteriol. 95:327-332.

2. Alexander, J. J. 1969. Effect of infection with themeningopneumonitis agent on deoxyribonu-cleic acid and protein synthesis by its L-cellhost. J. Bacteriol. 97:653-657.

3. Allen, E. A., M. R. Bovarnick, and J. C. Snyder.1964. The effect of irradiation with ultravioletlight on various properties of typhus rickett-siae. J. Bacteriol. 67:718-723.

4. Anacker, R. L., E. G. Pickens, and D. B.Lackman. 1967. Details of the ultrastructureof Rickettsia prowazekii grown in the chickyolk sac. J. Bacteriol. 94:260-262.

5. Anderson, D. R., H. E. Hopps, M. F. Barile, andB. D. Bernheim. 1965. Comparison of theultrastructure of several rickettsiae, ornithosisvirus, and Mycoplasma in tissue culture. J.Bacteriol. 90:1387-1404.

6; Babudieri, B. 1957. Q fever: a zoonosis.Advan. Vet. Sci. 5:81-182.

7. Barker, L. F. 1968. Determination of antibioticsusceptibility of rickettsiae and chlamydiae inBS-C-1 cell cultures. Antimicrob. Ag. Chemo-ther. 8:425-428.

8. Barker, L. F., J. K. Patt, and H. E. Hopps. 1968.Titration and neutralization of Rickettsiatsutsugamushi in tissue culture. J. Immunol.100:825-830.

9. Blackford, V. L. 1961. Influence of various me-tabolites on growth of Coxiella burnetii inmonolayer cultures of chick embryo entoder-mal cells. J. Bacteriol. 81:747-754.

10. Bovarnick, M. R. 1956. Phosphorylation accom-panying the oxidation of glutamate by theMadrid E strain of typhus rickettsiae. J. Biol.Chem. 220:353-361.

278 WEISS


http://mm

br.asm.org/

Dow

nloaded from



11. Bovarnick, M. R. 1960. Incorporation of ace-tate-1-C"4 into lipid by typhus rickettsiae. J.Bacteriol. 80:508-512.

12. Bovarnick, M. R., and E. G. Allen. 1954. Revers-ible inactivation of typhus rickettsiae. J. Gen.Physiol. 38:169-179.

13. Bovarnick, M. R., and E. G. Allen. 1957. Revers-ible inactivation of the toxicity and hemolyticactivity of typhus rickettsiae by starvation. J.Bacteriol. 74:637-645.

14. Bovarnick, M. R., and E. G. Allen. 1957. Revers-ible inactivation of typhus rickettsiae at 0 C.J. Bacteriol. 73:56-62.

15. Bovarnick, M. R., E. G. Allen, and G. Pagan.1953. The influence of diphosphopyridine nu-

cleotide on the stability of typhus rickettsiae.J. Bacteriol. 66:671-675.

16. Bovarnick, M. R., and J. C. Miller. 1950. Oxida-tion and transamination of glutamate by ty-phus rickettsiae. J. Biol. Chem. 184:661-676.

17. Bovarnick, M. R., J. C. Miller, and J. C. Snyder.1950. The influence of certain salts, aminoacids, sugars, and proteins on the stability ofrickettsiae. J. Bacteriol. 59:509-522.

18. Bovarnick, M. R., and L. Schneider. 1960. Theincorporation of glycine-1-C'4 by typhus rick-ettsiae. J. Biol. Chem. 235:1727-1731.

19. Bovarnick, M. R., and L. Schneider. 1960. Roleof adenosine triphosphate in the hemolysis ofsheep erythrocytes by typhus rickettsiae. J.Bacteriol. 80:344-354.

20. Bovarnick, M. R., L. Schneider, and H. Walter.1959. The incorporation of labeled methionineby typhus rickettsiae. Biochim. Biophys. Acta33:414-422.

21. Bovarnick, M. R., and J. C. Snyder. 1949.Respiration of typhus rickettsiae. J. Exp.Med. 89:561-565.

22. Bozeman, F. M., H. E. Hopps, J. X. Danauskas,E. B. Jackson, and J. E. Smadel. 1956. Studyon the growth of rickettsiae. I. A tissue culturesystem for the quantitative estimations ofRickettsia tsutsugamushi. J. Immunol.76:475-488.

23. Brinton, L. P., and W. Burgdorfer. 1971. Finestructure of Rickettsia canada in tissues ofDermacentor andersoni Stiles. J. Bacteriol.105:1149-1159.

24. Burgdorfer, W., R. L. Anacker, R. G. Bird, andD. S. Bertram. 1968. Intranuclear growth ofRickettsia rickettsii. J. Bacteriol.96:1415-1418.

25. Burgdorfer, W., and L. P. Brinton. 1970. In-tranuclear growth of Rickettsia canada, amember of the typhus group. Infect. Im-munity 2:112-114.

26. Chance, B. 1952. Spectra and reaction kinetics ofrespiratory pigments of homogenized and in-tact cells. Nature (London) 169:215-221.

27. Cohn, Z. A., F. M. Bozeman, J. M. Campbell, J.W. Humphries, and T. K. Sawyer. 1959.Study on growth of rickettsiae. V. Penetrationof Rickettsia tsutsugamushi into mammalian

279

cells in vitro. J. Exp. Med. 109:271-292.28. Cohn, Z. A., F. E. Hahn, W. Ceglowski, and F.

M. Bozeman. 1958. Unstable nucleic acids ofRickettsia mooseri. Science 127:282-283.

29. Consigli, R. A., and D. Paretsky. 1962. Oxidationof glucose-6-phosphate and isocitrate by Cox-iella burnetii. J. Bacteriol. 83:206-207.

30. Cox, H. R. 1938. Use of yolk sac of developingchick embryo as medium for growing rickett-siae of Rocky Mountain spotted fever andtyphus groups. Pub. Health Rep.53:2241-2247.

31. Cox, H. R. 1941. Cultivation of rickettsiae ofRocky Mountain spotted fever, typhus and Qfever groups in the embryonic tissues of devel-oping chicks. Science 94:399-403.

32. Dasinger, B. L., and J. B. Wilson. 1962. Gluta-mate metabolism in Brucella abortus strainsof low and high virulence. J. Bacteriol.84:911-915.

33. Davis, B. D. 1951. Inhibition of Escherichia coliby p-aminobenzoic acid and its reversal byp-hydroxybenzoic acid. J. Exp. Med.74:243-254.

34. Ennis, H. L., and M. Lubin. 1964. Cyclohexi-mide: aspects of inhibition of protein synthesisin mammalian cells. Science 146:1474-1476.

35. Fiset, P., and R. A. Ormsbee. 1968. The anti-body response to antigens of Coxiella burneti.Zentralb. Bakteriol Parasitenk. Infektionskr.Hyg. Abt. Orig. 206:321-329.

36. Freeman, B. A., G. R. Pearson, and W. D. Hines.1964. Host-parasite relationships in brucello-sis. III. Behavior of avirulent brucella in tissueculture monocytes. J. Infect. Dis. 114:441-449.

37. Friis, R. R. 1972. Interaction of L cells andChlamydia psittaci: entry of the parasite andhost responses to its development. J. Bacte-riol. 110:706-721.

38. Gerhardt, P., H. B. Levine, and J. B. Wilson.1950. The oxidative dissimilation of aminoacids and related compounds by Brucellaabortus. J. Bacteriol. 60:459-467.

39. Gerhardt, P., and J. B. Wilson. 1948. Thenutrition of brucellae: growth in simple chem-ically defined media. J. Bacteriol. 56:17-24.

40. Gilford, J. H., and W. R. Price. 1955. Virulent-avirulent conversions of Rickettsia rickettsii invitro. Proc. Nat. Acad. Sci. U.S.A.41:870-873.

41. Gordon, F. B., H. R. Dressler, A. L. Quan, W. T.McQuilken, and J. I. Thomas. 1972. Effect ofionizing irradiation on susceptibility ofMcCoy cell cultures to Chlamydiatrachomatis. Appl. Microbiol. 23:123-129.

42. Guardiola, A. L., and D. Paretsky. 1968. Meta-bolic reactivation of rickettsiae by diphos-phopyridine nucleotide. Science 128:141-142.

43. Guardiola, A. L., D. Paretsky, and W. McEwen.1958. The interaction of DPN + with p-aminobenzoic acid and analogous compounds.J. Amer. Chem. Soc. 80:418-421.

44. Hahn, F. E., Z. A. Cohn, and F. M. Bozeman.

VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

1960. Metabolic studies of rickettsiae. V. Me-tabolism of glutamine and asparagine in Rick-ettsia mooseri. J. Bacteriol. 80:400-405.

45. Hanks, J. H. 1966. Host-dependent microbes.Bacteriol. Rev. 30:114-135.

46. Hayes, J. E., F. E. Hahn, Z. A. Cohn, E. B.Jackson, and J. E. Smadel. 1957. Metabolicstudies of rickettsiae. IV. Terminal respiratoryenzymes in Rickettsia mooseri. Biochim. Bio-phys. Acta 26:570-576.

47. Hill, J. C. 1971. Effect of glutamate on exoge-nous citrate catabolism of Neisseriameningitidis and of other species of Neisseria.J. Bacteriol. 106:819-823.

48. Hopps, H. E., F. E. Hahn, C. L. Wisseman, Jr.,E. B. Jackson, and J. E. Smadel. 1956. Meta-bolic studies of rickettsiae. III. Studies oftransamination, oxidative phosphorylationand glutamate-2-C 14 incorporation by purifiedsuspensions of Rickettsia mooseri. J. Bacte-riol. 71:708-716.

49. Hopps, H. E., E. B. Jackson, J. X. Danauskas,and J. E. Smadel. 1959. Study on the growthof rickettsiae. III. Influence of extracellularenvironment on the growth of Rickettsiatsutsugamushi in tissue culture cells. J. Im-munol. 82:161-171.

50. Hopps, H. E., E. B. Jackson, J. X. Danauskas,and J. E. Smadel. 1959. Study on the growthof rickettsiae. IV. Effect of chloramphenicoland several metabolic inhibitors on the multi-plication of Rickettsia tsutsugamushi in tissueculture cells. J. Immunol. 82:172-181.

51. Jackson, E. B. 1951. Comparative efficacy ofseveral antibiotics on experimental rickettsialinfections in embryonated eggs. Antibiot.Chemother. 1:231-241.

52. Jones, F., Jr., and D. Paretsky. 1967. Physiologyof rickettsiae. VI. Host-independent synthesisof polyribonucleotides by Coxiella burnetii. J.Bacteriol. 93:1063-1068.

53. Kenyon, R. H., W. M. Acree, G. G. Wright, andF. W. Melchior, Jr. 1972. Preparation of vac-cines for Rocky Mountain spotted fever fromrickettsiae propagated in cell culture. J. In-fect. Dis. 125:146-152.

54. Kingsbury, D. T. 1969. Estimate of the genomesize of various microorganisms. J. Bacteriol.98:1400-1401.

55. Kokorin, I. N. 1968. Biological peculiarities ofthe development of rickettsiae. Acta Virol.12:31-35.

56. Kordova, N. 1959. Filterable particles of Coxiellaburneti. Acta Virol. 3:25-36.

57. Kordova, N. 1966. Plaque assay of rickettsiae.Acta Virol. 10:278.

58. Kordova, N., P. R. Burton, C. M. Downs, D.Paretsky, and E. Kovacova. 1970. The interac-tion of Coxiella burnetii phase I and phase IIin Earle's cells. Can. J. Microbiol. 16:125-133.

59. Lehman, J. 1946. Para-aminosalicylic acid in thetreatment of tuberculosis. Lancet 250:15-16.

60. Lin, H. 1968. Inhibition of thymidine kinase

activity and deoxyribonucleic acid synthesisin L cells infected with the meningopneumoni-tis agent. J. Bacteriol. 96:2054-2065.

61. McDade, J. E. 1969. Determination of antibioticsusceptibility of Rickettsia by the plaqueassay technique. Appl. Microbiol. 18:133-135.

62. McDade, J. E., and P. J. Gerone. 1970. Plaqueassay for Q fever and scrub typhus rickettsiae.Appl. Microbiol. 19:963-965.

63. McDade, J. E., J. R. Stakebake, and P. J.Gerone. 1969. Plaque assay system for severalspecies of Rickettsia. J. Bacteriol. 99:910-912.

64. McDonald, T. L., and L. Mallavia. 1970. Bio-chemistry of Coxiella burnetii: 6-phosphoglu-conic acid dehydrogenase. J. Bacteriol.102:1-5.

65. McDonald, T. L., and L. Mallavia. 1971. Bio-chemistry of Coxiella burnetii: Embden-Meyerhof pathway. J. Bacteriol. 107:864-869.

66. McKiel, J. A., E. J. Bell, and D. B. Lackman.1967. Rickettsia canada: a new member of thetyphus group of rickettsiae isolated from Hae-maphysalis leporispalustris ticks in Canada.Can. J. Microbiol. 13:503-510.

67. Mallavia, L. P., and D. Paretsky. 1963. Studieson the physiology of rickettsiae. V. Metabo-lism of carbamyl phosphate by Coxiella bur-netii. J. Bacteriol. 86:232-238.

68. Mallavia, L. P., and D. Paretsky. 1967. Physiol-ogy of rickettsiae. VII. Amino acid incorpora-tion by Coxiella burnetii and by infectedhosts. J. Bacteriol. 93:1479-1483.

69. Mallavia, L. P., and E. Weiss. 1970. Catabolicactivities of Neisseria meningitidis: utilizationof glutamate. J. Bacteriol. 101:127-132.

70. Marmur, J., and P. Doty. 1962. Determination ofthe base composition of deoxyribonucleic acidfrom its thermal denaturation temperature. J.Mol. Biol. 5:109-118.

71. Mattheis, M. S., M. Silverman, and D. Pa-retsky. 1963. Studies on the physiology ofrickettsiae. IV. Folic acids of Coxiellaburnetii. J. Bacteriol. 85:37-41.

72. Moulder, J. W. 1962. The biochemistry of in-tracellular parasitism. University of ChicagoPress, Chicago.

73. Moulder, J. W. 1964. The psittacosis group asbacteria. John Wiley & Sons, Inc., New York.

74. Moulder, J. W. 1966. The relation of the psitta-cosis group (chlamydiae) to bacteria and vi-ruses. Annu. Rev. Microbiol. 20:107-130.

75. Myers, W. F. 1958. Host-parasite interrelation-ship involving Coxiella burnetii. Thesis, Uni-versity of Kansas, Lawrence.

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

77. Myers, W. F., J. V. Osterman, and C. L.Wisseman, Jr. 1972. Nutritional studies ofRickettsia quintana: nature of the hematinrequirement. J. Bacteriol. 109:89-95.

78. Myers, W. F., and D. Paretsky. 1961. Synthesis

280 WEISS


http://mm

br.asm.org/

Dow

nloaded from



of serine by Coxiella burnetii. J. Bacteriol.82:761-763.

79. Myers, W. F., P. J. Provost, and C. L. Wisse-man, Jr. 1967. Permeability properties ofRickettsia mooseri. J. Bacteriol. 93:950-960.

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

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

82. Ormsbee, R. A., D. B. Lackman, and E. G.Pickens. 1951. Relationship among comple-ment-fixing values, infectious endpoints, anddeath curves in experimental Coxiella burnetiiinfection. J. Immunol. 67:257-264.

83. Ormsbee, R. A., H. Parker, and E. G. Pickens.1955. The comparative effectiveness of au-

reomycin, terramycin, chloramphenicol,erythromycin and thiocymetin in suppressingexperimental rickettsial infections in chickembryos. J. Infect. Dis. 96:162-167.

84. Ormsbee, R. A., and M. G. Peacock. 1964.Metabolic activity in Coxiella burnetii. J.Bacteriol. 88:1205-1210.

85. Ormsbee, R. A., and E. Weiss. 1963. Trachomaagent: glucose utilization by purified suspen-

sions. Science 142:1077-1078.86. Paretsky, D. 1968. Biochemistry of rickettsiae

and their infected hosts, with special referenceto Coxiella burnetii. Zentralb. Bakteriol. Pa-rasitenk. Infektionskr. Hyg. Abt. Orig.206:284-291.

87. Paretsky, D., R. A. Consigli, and C. M. Downs.1962. Studies on the physiology of rickettsiae.III. Glucose phosphorylation and hexokinaseactivity in Coxiella burnetii. J. Bacteriol.83:538-543.

88. Paretsky, D., C. M. Downs, R. A. Consigli, andB. K. Joyce. 1958. Studies on the physiology ofrickettsiae. I. Some enzyme systems of Cox-iella burnetii. J. Infect. Dis. 103:6-11.

89. Perkins, H. R., and A. C. Allison. 1963. Cell-wallconstituents of rickettsiae and psittacosis-lymphogranuloma organisms. J. Gen. Micro-biol. 30:469-480.

90. Pestka, S. 1971. Inhibitors of ribosome functions.Annu. Rev. Microbiol. 25:487-562.

91. Price, W. H. 1953. The epidemiology of RockyMountain spotted fever. I. The characteriza-tion of strain virulence of Rickettsia rickettsii.Amer. J. Hyg. 58:248-268.

92. Price, W. H. 1953. A quantitative analysis of thefactors involved in the variations in virulenceof rickettsiae. Science 118:49-52.

93. Price, W. H. 1954. Variation in virulence of"Rickettsia rickettsii" under natural and ex-

perimental conditions, p. 164-183. In F. W.Hartman, F. L. Horsfall, Jr., J. G. Kiss (ed.)The dynamics of virus and rickettsial infec-tions. Blakiston, New York.

94. Ramm, L. E., and H. H. Winkler. 1973. Rickett-sial hemolysis: adsorption of rickettsiae toerythrocytes. Infect. Immunity 7:93-99.

281

95. Ransom, S. E., and R. J. Huebner. 1951. Studieson the resistance of Coxiella burneti to physi-cal and chemical agents. Amer. J. Hyg.53:110-119.

96. Rees, H. B., and E. Weiss. 1968. Glutamatecatabolism of Rickettsia rickettsi and factorsaffecting retention of metabolic activity. J.Bacteriol. 95:389-396.

97. Schaechter, M., F. M. Bozeman, and J. E.Smadel. 1957. Study on the growth of rickett-siae. II. Morphologic observations of livingrickettsiae in tissue culture cells. Virology3:160-172.

98. Schaecter, M., A. J. Tousimis, Z. A. Cohn, H.Rosen, J. Campbell, and F. E. Hahn. 1957.Morphological, chemical, and serologicalstudies of the cell walls of Rickettsia mooseri.J. Bacteriol. 74:822-829.

99. Schildkraut, C. L., J. Marmur, and P. Doty.1962. Determination of the base compositionof deoxyribonucleic acid from its buoyantdensity in CsCl. J. Mol. Biol. 4:430-443.

100. Schramek, S. 1968. Isolation and characteriza-tion of deoxyribonucleic acid from Coxiellaburneti. Acta Virol. 12:18-22.

101. Schramek, S. 1972. Deoxyribonucleic acid basecomposition of members of the typhus groupof rickettsiae. Acta Virol. 16:447.

102. Singer, C. E., and B. N. Ames. 1970. Sunlight,ultraviolet and bacterial DNA base ratios.Science 170:822-824.

103. Smith, J. B., and M. G. P. Stoker. 1951. Thenucleic acids of Rickettsia burneti. Brit. J.Exp. Pathol. 32:433-441.

104. Snyder, J. C. 1948. The treatment of the rickett-sial diseases of man p. 169-177. In F. R.Moulton (ed.), Rickettsial diseases of man.

American Association for the Advancement ofScience, Washington, D.C.

105. Snyder, J. C., M. R. Bovarnick, J. C. Miller, andR. S. Chang. 1954. Observations on the hemo-lytic properties of typhus rickettsiae. J. Bacte-riol. 67:724-730.

106. Spencer, R. R., and R. R. Parker. 1930. Studieson Rocky Mountain spotted fever. U. S. Pub-lic Health Serv. Hyg. Lab. Bull. 154:1-116.

107. Stanier, R. Y. 1964. Toward a definition of thebacteria, p. 445-464. In I. C. Gunsalus and R.Y. Stanier (ed.), The bacteria: a treatise on

structure and function, vol. 5. Academic PressInc., New York.

108. Stelzner, A., and W. Linss. 1968. Binary fissionin Coxiella burneti. Nature (London)218:1069-1070.

109. Stoenner, H. G., D. B. Lackman, and E. J. Bell.1962. Factors affecting the growth of rick-ettsias of the spotted fever group in fertilehens' eggs. J. Infect. Dis. 110:121-128.

110. Suitor, E. C., Jr. 1964. Studies on the cellenvelope of Wolbachia persica. J. Infect. Dis.114:125-134.

111. Suitor, E. C., Jr., and E. Weiss. 1961. Isolation ofa rickettsialike microorganism (Wolbachia

VOL. 37, 1973


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

persica, n. sp.) from Argas persicus (Oken). J.Infect. Dis. 108:95-106.

112. Takemori, N., and M. Kitaoka. 1952. Reversal ofp-aminobenzoic acid inhibition of growth ofrickettsiae by p-hydroxybenzoic acid in agar-

slant tissue culture. Science 116:710-711.113. Thompson, H. A., 0. G. Baca, and D. Paretsky.

1971. Presence of ribosomal ribonucleic acid inthe rickettsia Coxiella burneti. Biochem. J.125:365-366.

114. Tyeryar, F. J., Jr., E. Weiss, D. B. Millar, F. M.Bozeman, and R. A. Ormsbee. 1973. DNAbase composition of rickettsiae. Science180:415-417.

115. Vinson, J. W., and H. S. Fuller. 1961. Studies on

trench fever. I. Propagation of rickettsia-likemicroorganisms from a patient's blood. Pa-thol. Microbiol. Suppl. 24:152-166.

116. Weinberg, E. H., J. R. Stakebake, and P. J.Gerone. 1969. Plaque assay for Rickettsiarickettsii. J. Bacteriol. 98:398-402.

117. Weiss, E. 1960. Some aspects of variation inrickettsial virulence. Ann. N. Y. Acad. Sci.88:1287-1297.

118. Weiss, E. 1968. Comparative metabolism ofrickettsiae and other host dependent bacteria.Zentralbl. Bakteriol Parsasitenk. Infectionskr.Hyg. Abt. Orig. 206:292-298.

119. Weiss, E. 1970. Catabolic activities of Neisseriameningitidis: utilization of succinate. J. Bac-teriol. 101:133-137.

120. Weiss, E., and H. R. Dressler. 1958. Growth ofRickettsia prowazeki in irradiated monolayercultures of chick embryo entodermal cells. J.Bacteriol. 75:544-552.

121. Weiss, E., and H. R. Dressler. 1960. Centrifuga-tion of rickettsiae and viruses onto cells and itseffect on infection. Proc. Soc. Exp. Biol. Med.103:691-695.

122. Weiss, E., and H. R. Dressler. 1960. Selection ofan erythromycin-resistant strain of Rickettsiaprowazekii. Amer. J. Hyg. 71:292-298.

123. Weiss, E., and H. R. Dressler. 1962. Increasedresistance to chloramphenicol in Rickettsiaprowazekii with a note on failure to demon-strate genetic interaction among strains. J.Bacteriol. 83:409-414.

124. Weiss, E., H. R. Dressler, and E. C. Suitor, Jr.1957. Selection of a mutant strain of Rick-ettsia prowazeki resistant to p-aminobenzoicacid. J. Bacteriol. 73:421-430.

125. Weiss, E., H. R. Dressler, and E. C. Suitor, Jr.1959. Further studies of drug-resistant strainsof Rickettsia prowazekii. J. Bacteriol.77:91-100.

126. Weiss, E., H. R. Dressler, and E. C. Suitor, Jr.1959. Inhibition by acetylsalicylic acid of rick-ettsial strains resistant to p-aminobenzoicacid. J. Bacteriol. 78:432-440.

127. Weiss, E., A. E. Green, R. Grays, and L. W.Newman. 1973. Metabolism of Rickettsiatsutsugamushi and Rickettsia rickettsi in irra-diated host cells. Infect. Immhunity 8:4-7.

128. Weiss, E., and J. S. Huang. 1954. The infectedcell count method of titration of feline pneu-

monitis virus. J. Infect. Dis. 94:107-125.129. Weiss, E., and J. W. Moulder. 1973. Rickettsia,

Coxiella, Rochalimaea. In R. E. Buchananand N. E. Gibbons, (ed.) Bergey's manual ofdeterminative bacteriology, 8th ed. The Wil-liams & Wilkins Co., Baltimore, in press.

130. Weiss, E., W. F. Myers, E. C. Suitor, Jr., and E.M. Neptune, Jr. 1962. Respiration of a rickett-sialike microorganism, Wolbachia persica. J.Infect. Dis. 110:155-164.

131. Weiss, E., L. W. Newman, R. Grays, and A. E.Green. 1972. Metabolism of Rickettsia typhiand Rickettsia akari in irradiated L cells.Infect. Immunity 6:50-57.

132. Weiss, E., and H. Pietryk. 1956. Growth ofCoxiella burnetii in monolayer cultures ofchick embryo entodermal cells. J. Bacteriol.72:235-241.

133. Weiss, E., H. B. Rees, Jr., and J. R. Hayes. 1967.Metabolic activity of purified suspensions ofRickettsia rickettsi. Nature (London)213:1020-1022.

134. Weiss, E., S. Schramek, N. N. Wilson, and L. W.Newman. 1970. Deoxyribonucleic acid hetero-geneity between human and murine strains ofChlamydia trachomatis. Infect. Immunity2:24-28.

135. Weiss, E., and N. N. Wilson. 1969. Role ofexogenous adenosine triphosphate in catabolicand synthetic activities of Chlamydia psittaci.J. Bacteriol. 97:719-724.

136. Wiebe, M. E., P. R. Burton, and D. M. Shankel.1972. Isolation and characterization of two celltypes of Coxiella burneti phase I. J. Bacteriol.110:368-377.

137. Wike, D. A., and W. Burgdorfer. 1972. Plaqueformation in tissue cultures by Rickettsiarickettsi isolated directly from whole bloodand tick hemolymph. Infect. Immunity6:736-738.

138. Wike, D. A., R. A. Ormsbee, G. Tallent, and M.G. Peacock. 1972. Effects of various suspend-ing media on plaque formation by rickettsiaein tissue culture. Infect. Immunity 6:550-556.

139. Wike, D. A., G. Tallent, M. G. Peacock, and R.A. Ormsbee. 1972. Studies on the rickettsialplaque assay technique. Infect. Immunity5:715-722.

140. Wisseman, C. L., Jr. 1968. Some biologicalproperties of rickettsiae pathogenic for man.

Zentralbl. Bakteriol. Parasitenk. Infektionskr.Hyg. Abt. Orig. 206:299-313.

141. Wisseman, C. L., Jr., F. E. Hahn, E. B. Jackson,F. M. Bozeman, and J. E. Smadel. 1952.Metabolic studies of rickettsiae. II. Studies on

the pathway of glutamate oxidation by puri-fied suspension of Rickettsia mooseri. J. Im-munol. 68:251-264.

142. Wisseman, C. L., Jr., E. B. Jackson, F. E. Hahn,A. C. Ley, and J. E. Smadel. 1951. Metabolicstudies of rickettsiae. I. The effects of antimi-

282 WEISS


http://mm

br.asm.org/

Dow

nloaded from



crobial substances and enzyme inhibitors onthe oxidation of glutamate by purified rickett-siae. J. Immunol. 67:123-136.

143. Wood, W. H., Jr., and C. L. Wisseman, Jr. 1967.The cell wall of Rickettsia mooseri. I. Mor-phology and chemical composition. J. Bacte-riol. 93:1113-1118.

144. Wyatt, G. R., and S. S. Cohen. 1952. Nucleicacids of rickettsiae. Nature (London)170:846-847.

145. Youmans, G. P., G. W. Raleigh, and A. S.Youmans. 1947. The tuberculostatic action ofpara-aminosalicylic acid. J. Bacteriol.54:409-416.

VOL. 37, 1973 283


http://mm

br.asm.org/

Dow

nloaded from


in u.s.a. growthand physiology ofrickettsiaethe generic name "rickettsia" (uncapital-ized)...

Documents