rickettsia felis, from culture to genome sequencing

PDFlib PLOP: PDF Linearization, Optimization, Protection

Page inserted by evaluation versionwww.pdflib.com – [email protected]

26

Ann. N.Y. Acad. Sci. 1063: 26–34 (2005). © 2005 New York Academy of Sciences.doi: 10.1196/annals.1355.004

Rickettsia felis, from Culture to Genome Sequencing

H. OGATA,a C. ROBERT,b S. AUDIC,a S. ROBINEAU,b G. BLANC,a P.E. FOURNIER,a P. RENESTO,b J. M. CLAVERIE,a AND D. RAOULTb

aCNRSIBSM, Information Génomique et Structurale, 13402 Marseille Cedex 20, FrancebUniversité de la Méditerranée, CNRS UMR 6020, Faculté de Médecine, Unité des Rickettsies, 13385 Marseille, France

ABSTRACT: Rickettsia felis has been recently cultured in XTC2 cells. Thisallows production of enough bacteria to create a genomic bank and to sequenceit. The chromosome of R. felis is longer than that of previously sequenced rick-ettsiae and it possess 2 plasmids. Microscopically, this bacterium exhibits twoforms of pili: one resembles a conjugative pilus and another forms hair-likeprojections that may play a role in pathogenicity. R. felis also exhibits severalcopies of ankyrin-repeat genes and tetratricopeptide encoding gene that arespecifically linked to pathogenic host-associated bacteria. It also containstoxin–antitoxin system encoding genes that are extremely rare in intracellularbacteria and may be linked to plasmid maintenance.

KEYWORDS: Rickettsia felis; pili; gene; genome sequencing; plasmid

INTRODUCTION

Rickettsiae are fastidious intracellular bacteria. Their first definition by Wolbachemphasized both these specifications since they were uncultivable on usual mediumand seen with specific staining within cells.1 Rickettsiae were first propagated inanimals (apes, monkeys, guinea pigs), then in embryonnated eggs, and finally in cellcultures.2 The first description of a Rickettsia in the cat flea Ctenocephalides feliswas provided by Sikora in Germany in 1918.1,3 It was named R. ctenocephali and itstransovarial transmission observed in the flea host. This work was largely ignoreduntil now.

In 1990, when C. felis were examined in California as potential vectors ofR. typhi, the agent of murine typhus, a new rickettsia was observed by electronicmicroscopy and named the ELB agent.4 By genomic sequence comparison, based onthe 17-kDa protein gene sequencing, this organism was considered to be a rickettsia.5

It was lately detected by PCR in other fleas such as Pulex irritans in the USA6 andC. felis and C. canis worldwide. R. felis is found in Africa,7 in Oceania,8 in Europe,7

in America5 and in Asia.9 R. felis is a common intracellular parasite of a verycommon pet flea and as many as 15% of C. felis may be infected.10

Address for correspondence: Didier Raoult, Université de la Méditerranée, CNRS UMR 6020,Faculté de Médecine, Unité des Rickettsies, 27 Bd Jean Moulin, 13385 Marseille, France. Voice:+33 4 91 38 55 17; fax: +33 4 91 38 77 72.

[email protected]

27OGATA et al.: R. FELIS, CULTURE TO GENOME SEQUENCING

Cases of human infections were described first by using PCR amplification. Thefirst case was detected in the blood of a patient from Texas when investigating pos-sible cases of murine typhus in 1994.11 Since then, cases were found from France,7

Brazil,7 Mexico,12 Germany,13 Canary Islands and Tunisia (unpublished). Theincidence of R. felis infection in patients is unknown.

CULTURE OF RICKETTSIA FELIS

R. felis culture was a challenge. It was first reported that it grows in cell cultureat 37°C in 1995.14,15 Surprisingly, the authors found that the protein and immuno-genic profiles of R. felis were identical to that of R. typhi.6 The phylogenic positionof R. felis, obtained by the 17-kDa protein-encoding gene, showed that it is clusteredin the spotted fever group rickettsiae. Lately the same group reported, in an erratum,that at some point their cultures were contaminated.16 Their work was never repro-duced, and when we began working with R. felis we were unable to grow it following

FIGURE 1. R. felis growing in XTC2 cells. (a) Gimenez staining magnification ×1000(note the multiplication within the nucleus).

28 ANNALS NEW YORK ACADEMY OF SCIENCES

FIGURE 1—continued. (b) Electronic microscopy showing bacteria in the eukaryotic nucleus.

FIGURE 1—continued. (c) Identification of pili on R. felis by transmission electronicmicroscopy (negative staining).


their report. Our conclusion was that the previous report of culture of R. felis mayhave resulted from a contamination by R. typhi from the very beginning.

We decided to use XTC2 cells to grow R. felis.7 These cells, taken from Xenopus,grow at 28°C and are used for Arbovirus cultures. We were able, using this cell line,to grow, identify, and characterize R. felis17 (FIG. 1). This culture system allowed usto produce large amounts of rickettsiae and therefore to purify the bacteria in orderto clone and sequence its genome.

GENOME SEQUENCING

The sequence of R. felis genome was surprising. We identified early the presenceof a plasmid that was apparently 63 kb long.18 However, we have had problems inclosing our putative plasmid molecule based on the various clones generated by ourbank. Finally, we tested the hypothesis that in fact 2 plasmids were present (one,being of 39 kb is a deleted form of the longer) (see FIG. 2). This was confirmed byPCR, and by pulse field gel electrophoresis coupled to hybridization. Moreover, wefound them in all the R. felis–positive fleas by PCR.18 The longer plasmid apparentlycontains the equipment to allow conjugative plasmd transfer. This is the first rickett-sia harboring a plasmid. Moreover, this potential conjugative capability has beenreported only once in an intracellular bacterium. Greub et al. reported thatParachlamydia acanthamoebae have an integron harboring the essential gene forconjugation.19

The size of the chromosome is longer (1,485 kb) than other previously sequencedrickettsiae (i.e., R. prowazekii, R. conorii, R. sibirica, R. rickettsii, R. akari, andR. typhi). It is colinear with the other rickettsiae but some inversions and transloca-

FIGURE 2. Circular representation of R. felis chromosome and plasmids.


tions were found. They were associated with genes encoding transposases. Weidentified 530 R. felis–specific genes compared to other published genomes(R. prowazekii, R. typhi, R. conorii, R. sibirica). R. felis have more paralogs in severalidentified gene families and unidentified gene families than other rickettsiae.

These families comprise 82 transposases, tnp, 22 ankyrin repeat-containinggenes, ank, 13 surface cell antigens, sca (9 complete and 4 split), 11 tetratricopeptiderepeat-containing protein genes, tpr, 14 spoT genes, and 5 families of toxin-antitoxin(TAT) system genes (16 toxins and 14 antitoxin genes). These specific features maybe associated with the higher genome plasticity oberved in R. felis and with specificprotein interactions mediated by ank and tpr.

R. FELIS SPECIFIC GENES

R. felis have 22 ankyrin repeats which is more than any prokaryotes sequenced sofar. Ankyrin repeats are protein-protein interaction motifs involved in many steps ofcell division. It is found mainly in eukaryotic cells, but also has been identified inpathogenic bacteria and in viruses.20 Tetratricopeptide are also associated withprotein interactions. Eleven copies were found in R. felis, and only host-associatedbacteria exhibit such high number of copies such as Legionella pneumophila,Treponema, and Leptospira.21

As for TAT and spoT a recent review showed that TAT cassettes are abundant inprokaryotes including many pathogenic organisms.22 During starvation, bacteriadownregulate RNA synthesis by upregulating the secretion of the alarmones P4G(ppGpp and pppGpp) under the control of two genes relA and spoT. P4G inhibits sta-ble RNA promoters specifically involving stable RNA promoters with short half-life.The consequence is to increase cell survival during nutritional stress. Toxins of theTAT have been identified first as gene killers present in plasmids to stabilize plasmidin bacteria. A long-time toxin is secreted along with a short-life antitoxin. If the TATloci disappear, the toxin may destroy the bacteria as the antitoxin is no longer tran-scribed. A comparable system uses antisense RNAs. This system is also known asthe plasmid-addiction system. There are 8 families of TAT loci.22 Interestinglybefore our findings in R. felis the TAT system was considered exceptionel in intra-cellular organisms. As a matter of fact R. felis have 16 genes encoding for toxin and14 for antitoxins of 5 of these TAT loci. The fact that rel BE (found in R. felis) hasbeen demonstrated to stabilize plasmid efficiently even when it is chromosomal mayindicate its role in R. felis plasmid maintenance. In general, chromosomally locatedTAT loci may help to stabilize mobile genetic elements including integron. Itspresence in R. felis may indicate that it is associated with plasmid as other sequencedrickettsiae have no plasmid and few TAT copies.23 Moreover, R. felis exhibits 14paralogs. This gene is also critical in controlling the duplication rate.

SURFACE APPENDAGES OF R. FELIS

Two types of surface appendages can be recognized on certain bacterial species:the flagella that provide cell motility and pili, also known as fimbriae. While flagellaoccur on both gram-positive and gram-negative bacteria, pili are found almost exclu-


sively on gram-negative bacteria with only a few gram-positive exceptions.24,25

Analysis of the complete genome sequence of R. felis has revealed that this micro-organism has genes susceptible of involvement in pili biogenesis. To determinewhether pili were expressed or not on the R. felis surface, an electron microscopystudy was carefully performed to avoid disruption of such appendages. R. felis cul-tivated in Xenopus laevis XTC-2 cells for 48 hours at 28°C were collected and cen-trifuged (400 × g, 10 min) before fixation for 1 hour at 4°C in glutaraldehyde (2.5%in PBS). Cells were then washed in PBS and placed on a carbon-formvar-coated 400-mesh copper grid (Electron Microscopy Sciences) for 15 min then negatively stainedwith 2% phosphotungstic acid for 10 s before analysis by electron microscopy(Philips Morgagni 268D). These experiments allowed the existence of two types ofpili to be confirmed. We observed a small number of long pili known as sex pili. Thisform of pili establishes direct contact between bacteria, providing a very typicalfigure of Mpf apparatus (FIG. 1C). These pili may be specialized in conjugation. Theother form of pili forms small hair-like projections emerging out from the cell surfaces,probably involved in the bacteria’s attachment to other cells. Without pili, many

FIGURE 3. Schematic representation of phenotypic R. felis characters highlightedfrom genome sequence. Identification of genes known to confer specific features in other mi-croorganisms prompted us to look for possible similar phenotype in R. felis. Pili and actincomet tails were thus visualized through transmission electronic microscopy (TEM) and im-munofluorescence assay (IF), respectively. Following this strategy, we also evidencedhemolytic and γ-lactamase capacities of these bacteria. These results illustrate the fact thatwhole genome sequencing permit not only to develop classical transcriptomics and proteomicsapproaches but also to gain insights concerning poorly documented bacterial phenotypes.


disease-causing bacteria lose their invasion capability. The latter type appendagesmight be considered as virulence factors, as described for Francisella tularensis.

PHENOTYPIC POST-GENOMICS ANALYSIS

Some post-genomics studies were initiated by using transcriptomics and pro-teomics approaches. Transcription of some R. felis genes has been investigated. Wealso started to analyze the R. felis proteome through bidimensional electrophoresiscoupled with mass spectrometry. Here, we envisaged post-genomics as a way to as-sociate in vivo phenotypes of these bacteria to genomic features. Indeed, the obligateintracellular nature of R. felis hindered progress in the detailed characterization ofits phenotypic diversity. This approach is schematically represented in FIGURE 3.

As previously mentioned, we also found a RickA homologue in R. felis ge-nome.26 Based on this finding, we performed immunofluorescence assays and evi-denced that R. felis can use the actin cytoskeleton to disseminate through eukaryoticcells, as other SFG rickettsiae. Moreover, we identified R. felis multiplying in nucle-us (FIG. 1). This phenomenon is strictly associated with the spotted fever group rick-ettsiae and linked to intracellular mobility provided by actin polymerization.27

Another R. felis phenotypic character suggested from genomic analyses [3 ORFs en-coding for patatin-like proteins and one for phospholipase D (pld)] was its hemolyticcapability. We experimentally confirmed that R. felis lyse erythrocytes, this effectbeing inhibited by DTT.18 Another genome-guided discovery was β-lactam inhibi-tion, which reached 57% and 53% of the concentration and the minimal inhibitoryconcentration (MIC), respectively, following 2 hours of incubation of R. felis withamoxicillin. This was the first betalactamase activity evidenced in a Rickettsia.While preliminary, these results illustrate the fact that whole genome sequencingoffers opportunities to rapidly gain a better understanding of phenotypic charactersof a fastidious microorganism.

PERSPECTIVES AND CONCLUSIONS

The present work showed the usefulness of the genomic approach to rapidly iden-tify specific features in such a bacterium. To date, 4 intracellular bacterial genomeswere entirely sequenced in 7 years or less after their first identification or culture.W. pipientis was first cultured in 1997 and sequenced in 2004,28 which allowed theidentification of several genes involved in host-symbiont interactions. Parachlamy-dia was first identified in amoeba in 1997 and sequenced in 2004,29 which allowedthe unexpected identification of genes for conjugation.30 Tropheryma whipplei wasfirst cultured in eukaryotic cells in 2000. Its genome sequencing in 200331,32

allowed the identification of basic metabolic properties that successfully guided theestablishment of a cell-free culture.33 Moreover, the sequence of the Wolbachiagenome of Brugia malayi was recently released and the bacteria has never been cul-tured.34 The genome sequencing of R. felis provided evidences of the presence ofconjugative plasmids. The genomic findings prompted us to identify two types ofpili, a hemolytic activity, a β-lactamase activity, and intracellular motility. These re-sults illustrate the fact that for such recently identified/cultured fastidious organisms,


complete genome sequencing is a very potent and time-saving strategy facilitatingidentification of unrecognized phenotypic traits.

[Competing interests statement: The authors have no conflicting financialinterests.]

REFERENCES

1. WOLBACH, S.B., J.L. TODD & F.W. PALFREY. 1922. The Etiology and Pathology ofTyphus. Harvard University Press. Cambridge, MA.

2. RAOULT, D. & V. ROUX. 1997. Rickettsioses as paradigms of new or emerginginfectious diseases. Clin. Microbiol. Rev. 10: 694–719.

3. SIKORA, H. 1916. Beiträge zur Anatomie, Physiologie und Biologie der Kleiderlaus(Pediculus vestimenti Nitzsch). I. Anatomie des Verdauungstraktes. Arch. f. Schiffs-undTropen-hyg 20: 5–76.

4. ADAMS, J.R., E.T. SCHMIDTMANN & A.F. AZAD. 1990. Infection of colonized cat fleas,Ctenocephalides felis (Bouché), with a rickettsia-like microorganism. Am. J. Trop.Med. Hyg. 43: 400–409.

5. AZAD, A.F., J.B. SACCI, W.M. NELSON, et al. 1992. Genetic characterization and trans-ovarial transmission of a typhus-like rickettsia found in cat fleas. Proc. Natl. Acad.Sci. USA 89: 43–46.

6. AZAD, A.F., S. RADULOVIC, J.A. HIGGINS, et al. 1997. Flea-borne rickettsioses:ecologic considerations. Emerg. Infect. Dis. 3: 319–327.

7. RAOULT, D., B. LA SCOLA, M. ENEA, et al. 2001. A flea-associated rickettsia patho-genic for humans. Emerg. Infect. Dis. 7: 73–81.

8. KELLY, P.J., N. MEADS, A. THEOBALD, et al. 2004. Rickettsia felis, Bartonella henselae,and B. clarridgiae, New Zealand. Emerg. Infect. Dis. 10: 967–968.

9. PAROLA, P., J.P. CORNET, Y.O. SANOGO, et al. 2003. Detection of Ehrlichia spp., Ana-plasma spp., Rickettsia spp., and other eubacteria in ticks from the Thai-Myanmarborder and Vietnam. J. Clin. Microbiol. 41: 1600–1608.

10. ROLAIN, J.M., M. FRANC, B. DAVOUST & D. RAOULT. 2003. Molecular detection ofBartonella quintana, B. koehlerae, B. henselae, B. clarridgeiae, Rickettsia felis, andWolbachia pipientis in cat fleas, France. Emerg. Infect. Dis. 9: 338–342.

11. SCHRIEFER, M.E., J.B. SACCI, JR., J.S. DUMLER, et al. 1994. Identification of a novelrickettsial infection in a patient diagnosed with murine typhus. J. Clin. Microbiol.32: 949–954.

12. ZAVALA-VELAZQUEZ, J.E., J.A. RUIZ-SOSA, R.A. SANCHEZ-ELIAS, et al. 2000. Rickettsiafelis rickettsiosis in Yucatan. Lancet 356: 1079–1080.

13. RICHTER, J., P.E. FOURNIER, J. PETRIDOU, et al. 2002. Rickettsia felis infection acquired inEurope and documented by polymerase chain reaction. Emerg. Infect. Dis. 8: 207–208.

14. RADULOVIC, S., J.A. HIGGINS, D.C. JAWORSKI, et al. 1995. Isolation, cultivation, andpartial characterization of the ELB agent associated with cat fleas. Infect. Immun 63:4826–4829.

15. RADULOVIC, S., J.A. HIGGINS, D.C. JAWORSKI & A.F. AZAD. 1995. In vitro and in vivoantibiotic susceptibilities of ELB rickettsiae. Antimicrob. Agents Chemother. 39:2564–2566.

16. RADULOVIC, S., J.A. HIGGINS, D.C. JAWORSKI & A.F. AZAD. 1996. In vitro and in vivoantibiotic susceptibilities of ELB rickettsiae. Antimicrob. Agents Chemother. 40:2912.

17. LA SCOLA, B., S. MECONI, F. FENOLLAR, et al. 2002. Emended description of Rickettsiafelis (Bouyer et al. 2001), a temperature-dependent cultured bacterium. Int. J. Syst.Evol. Microbiol. 52: 2035–2041.

18. OGATA, H., P. RENESTO, S. AUDIC, et al. 2005. The genome sequence of R. felis identi-fies the first case of conjugative plasmid in an obligate intracellular parasite. PLoS.Biol. In press.

19. GREUB, G., F. COLLYN, L. GUY & C.A. ROTEN. 2004. A genomic island present alongthe bacterial chromosome of the Parachlamydiaceae UWE25, an obligate amoebal


endosymbiont, encodes a potentially functional F-like conjugative DNA transfersystem. BMC Microbiol. 4: 48.

20. RUBTSOV, A.M. & O.D. LOPINA. 2000. Ankyrins. FEBS Lett. 482: 1–5.21. CAZALET, C., C. RUSNIOK, H. BRUGGEMANN, et al. 2004. Evidence in the Legionella

pneumophila genome for exploitation of host cell functions and high genome plasticity.Nat. Genet. 36: 1165–1173.

22. GERDES, K., S.K. CHRISTENSEN & A. LOBNER-OLESEN. 2005. Prokaryotic toxin-antitoxin stress response loci. Nature Rev. Microbiol. 3: 371–382.

23. PANDEY, D.P. & K. GERDES. 2005. Toxin-antitoxin loci are highly abundant in free-living but lost from host-associated prokaryotes. Nucl. Acids Res. 33: 966–976.

24. YEUNG, M.K. 2000. Actinomyces: surface macromolecules and bacteria-host inter-actions. In Gram-Positive Pathogens. V.A. Fischetti, R.P. Novick, J.J. Ferretti et al.,Eds.: 583–593. American Society for Microbiology. Washington, D.C.

25. WU, H. & P.M. FIVES-TAYLOR. 2001. Molecular strategies for fimbrial expression andassembly. Crit. Rev. Oral Biol. Med. 12: 101–115.

26. CHUNG, S., P. MONDON, Y.C. CHANG & K.J. KWON-CHUNG. 2003. Cryptococcus neofor-mans with a mutation in the tetratricopeptide repeat-containing gene, CCN1, causes sub-cutaneous lesions but fails to cause systemic infection. Infect. Immun. 71: 1988–1994.

27. TEYSSEIRE, N., C. CHICHE-PORTICHE & D. RAOULT. 1992. Intracellular movements ofRickettsia conorii and R. typhi based on actin polymerization. Res. Microbiol. 143:821–829.

28. ALT-MORBE, J., J.L. STRYKER, C. FUQUA, et al. 1996. The conjugal transfer system ofAgrobacterium tumefaciens octopine-type Ti plasmids is closely related to thetransfer system of an IncP plasmid and distantly related to Ti plasmid vir genes. J.Bacteriol. 178: 4248–4257.

29. BERANEK, A., M. ZETTL, K. LORENZONI, et al. 2004. Thirty-eight C-terminal aminoacids of the coupling protein TraD of the F-like conjugative resistance plasmid R1are required and sufficient to confer binding to the substrate selector protein TraM. J.Bacteriol. 186: 6999–7006.

30. KOONIN, E.V. & L. ARAVIND. 2000. The NACHT family—a new group of predictedNTPases implicated in apoptosis and MHC transcription activation. Trends Biochem.Sci. 25: 223–224.

31. QIN, A., A.M. TUCKER, A. HINES & D.O. WOOD. 2004. Transposon mutagenesis of theobligate intracellular pathogen Rickettsia prowazekii. Appl. Environ. Microbiol. 70:2816–2822.

32. COLLINS, N.E., J. LIEBENBERG, E.P. DE VILLIERS, et al. 2005. The genome of the heart-water agent Ehrlichia ruminantium contains multiple tandem repeats of activelyvariable copy number Proc. Natl. Acad. Sci. USA 102: 838–843.

33. BRAYTON, K.A., L.S. KAPPMEYER, D.R. HERNDON, et al. 2005. Complete genomesequencing of Anaplasma marginale reveals that the surface is skewed to two super-families of outer membrane proteins. Proc. Natl. Acad. Sci. USA 102: 844–849.

34. FOSTER, J., M. GANATRA, I. KAMAL, et al. 2005. The Wolbachia genome of Brugiamalayi: endosymbiont evolution within a human pathogenic nematode PLoS. Biol. 3:e121.

rickettsia felis, from culture to genome sequencing

Documents