Transposon Insertion Reveals pRM, a Plasmid of Rickettsia monacensis

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 2007, p. 4984–4995 0099-2240/07/$08.00⫹0 doi:10.1128/AEM.00988-07 Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vol. 73, No. 15

Transposon Insertion Reveals pRM, a Plasmid of Rickettsia monacensis䌤 Gerald D. Baldridge,* Nicole Y. Burkhardt, Roderick F. Felsheim, Timothy J. Kurtti, and Ulrike G. Munderloh Department of Entomology, University of Minnesota, St. Paul, Minnesota 55108 Received 2 May 2007/Accepted 6 June 2007

Until the recent discovery of pRF in Rickettsia felis, the obligate intracellular bacteria of the genus Rickettsia (Rickettsiales: Rickettsiaceae) were thought not to possess plasmids. We describe pRM, a plasmid from Rickettsia monacensis, which was detected by pulsed-field gel electrophoresis and Southern blot analyses of DNA from two independent R. monacensis populations transformed by transposon-mediated insertion of coupled green fluorescent protein and chloramphenicol acetyltransferase marker genes into pRM. Two-dimensional electrophoresis showed that pRM was present in rickettsial cells as circular and linear isomers. The 23,486-nucleotide (31.8% G/C) pRM plasmid was cloned from the transformant populations by chloramphenicol marker rescue of restriction enzyme-digested transformant DNA fragments and PCR using primers derived from sequences of overlapping restriction fragments. The plasmid was sequenced. Based on BLAST searches of the GenBank database, pRM contained 23 predicted genes or pseudogenes and was remarkably similar to the larger pRF plasmid. Two of the 23 genes were unique to pRM and pRF among sequenced rickettsial genomes, and 4 of the genes shared by pRM and pRF were otherwise found only on chromosomes of R. felis or the ancestral group rickettsiae R. bellii and R. canadensis. We obtained pulsed-field gel electrophoresis and Southern blot evidence for a plasmid in R. amblyommii isolate WB-8-2 that contained genes conserved between pRM and pRF. The pRM plasmid may provide a basis for the development of a rickettsial transformation vector. The genus Rickettsia (Rickettsiales: Rickettsiaceae) consists of obligate intracellular gram-negative bacteria that are associated primarily with arthropods. Rickettsia species are generally considered to be endosymbionts in the broad sense of having intimate relationships with their arthropod hosts that are not necessarily mutually beneficial. The better-known Rickettsia species are vertebrate pathogens associated with blood-feeding arthropods and include the etiologic agents of spotted fevers transmitted by ticks, endemic typhus transmitted by fleas, and epidemic typhus transmitted by lice. Those species that have occasionally been associated with human disease may constitute emerging pathogens (6, 34) or may be examples of rickettsiae that circulate in zoonotic transmission cycles and only incidentally infect humans. Other rickettsiae that have never been associated with human disease and have little or no pathogenicity in laboratory animals may exist as strict arthropod endosymbionts (4, 35). The close association and complex interactions between endosymbiotic rickettsiae and arthropods are emphasized by recent evidence that some rickettsial species function as reproductive manipulators of their hosts (19, 35, 36), a role historically associated with Wolbachia (Rickettsiales: Anaplasmataceae). The traditional focus of study on Rickettsia spp. as human pathogens has now expanded to encompass physiological, genetic, and evolutionary aspects of their role as arthropod endosymbionts. Research with rickettsiae has been impeded by severe technical constraints on genetic manipulation imposed by their obligate intracellular growth requirement and a lack of appro-

priate selectable markers and transformation vectors (49, 53). Transformation of rickettsiae with plasmid vectors used to transform free-living bacteria has not been successful, but such transformation has recently been achieved by the electroporation of rickettsiae with Tn5 transposomes containing selectable marker genes under the regulation of rickettsial promoters. The technique was used to obtain gene knockout mutants of the epidemic typhus pathogen, R. prowazekii, by use of a rifampin resistance marker (39, 49) and to transform the spotted fever group (SFG) tick endosymbiont, R. monacensis, to express green fluorescent protein (GFPuv) and a chloramphenicol (CHL) resistance marker (5). The successful transformation of SFG and typhus group rickettsiae using transposomes implies that the technique may be generally applicable to rickettsiae, albeit without the economy and high efficiency typical of plasmid transformation of free-living bacteria. Only a few obligate intracellular bacteria are known to harbor plasmids, including the vertebrate pathogens Coxiella burnetii (44) and various Chlamydiaceae species (37). Among nonpathogenic arthropod endosymbionts, some but not all species of Buchnera carry plasmids encoding proteins involved in the biosynthesis of amino acids that are deficient in the phloemsap diet of the host aphid (50). Genome sequencing of an obligate endosymbiont, Wigglesworthia glossinidia (2), and a secondary symbiont, Sodalis glossinidius (1, 14), of blood-feeding tsetse flies (Diptera: Glossinidae) revealed the presence of plasmids of undefined biological significance. Within the Rickettsiales, a plasmid has been described only for R. felis isolate California 2, a rickettsial endosymbiont of fleas (32). It was detected by genome sequencing as a 63-kbp form and a 39-kbp form that consisted of a subset of sequences from the larger form. They were designated pRF and pRF␦ and suggested to provide a potential basis for development of a rickettsial transformation technology. Demonstration of the existence of plas-

* Corresponding author. Mailing address: Department of Entomology, University of Minnesota, 1980 Folwell Ave., St. Paul, MN 55108. Phone: (612) 624-3688. Fax: (612) 625-5299. E-mail: baldr001@umn .edu. 䌤 Published ahead of print on 15 June 2007. 4984

TRANSPOSITION INTO pRM OF R. MONACENSIS

VOL. 73, 2007

mids in other Rickettsia species would strengthen the concept that rickettsiae in general, rather than R. felis in particular, might be amenable to genetic manipulation with a vector based on a rickettsial plasmid. Pornwiroon and colleagues (38), using PCR assays, confirmed the presence of pRF in R. felis strain LSU but could not detect pRF␦, while PCR assays by Ogata and colleagues (32) did not detect pRF-like plasmids in other Rickettsia species. In this report, we demonstrate the presence of plasmids in R. monacensis and R. amblyommii isolate WB-8-2 by pulsed-field gel electrophoresis (PFGE) and Southern blot analyses. Using transposome technology, we obtained transformant populations of R. monacensis expressing GFPuv and a CHL resistance marker from a transposon inserted within the rickettsial plasmid. The plasmid, designated pRM, was cloned in Escherichia coli as three overlapping fragments and sequenced. Using the R. monacensis pRM plasmid sequence as the query, nucleotide and translated protein BLAST searches of the GenBank database revealed remarkable similarities between pRM and the pRF plasmid of R. felis that included a shared gene complement and a high density of intact and remnant transposase sequences. Southern blot analyses demonstrated that genes conserved on pRM and pRF were also present on the plasmid of R. amblyommii. The presence of related plasmids in SFG Rickettsia species has significant implications for rickettsial biology and the development of technologies for genetic manipulation of rickettsiae. MATERIALS AND METHODS Growth, preparation, and transformation of rickettsiae. All rickettsiae were grown in the Ixodes scapularis ISE6 cell line maintained in L-15B300 medium as described previously (30). They included Rickettsia monacensis, passages 33 to 55 (46), and transformant clones Rmona658 (5), Rmona658A, and Rmona658B, passages 3 to 19 (originally from untransformed passage 19); R. amblyommii isolate WB-8-2, passages 20 to 34 (10, 24); R. peacockii, passages 114 to 116 (45); R. montanensis isolate M5/6, passages 14 to 17 (7); and R. felis isolate LSU, passages 14 to 19 (38) (kindly provided by Kevin Macaluso, Louisiana State University, Baton Rouge, LA). Rickettsiae were purified from infected host cells suspended in L-15B300 medium by passage through a 25-gauge syringe needle six times to lyse the cells, sequential passage of the lysate through 5- and 2-␮m syringe filters to remove cellular debris, and centrifugation of the filtrates at 18,400 ⫻ g at 4°C for 5 min to pellet rickettsiae. Rickettsiae were resuspended and transformed by electroporation with the pMOD658 transposon or used to prepare genomic DNA as described previously (5). Transformant clones of R. monacensis populations were isolated and maintained under CHL (Roche, Indianapolis, IN) selection and then visualized by epifluorescence microscopy as described previously (5). Isolation of rickettsial plasmid DNA was attempted with a QIAGEN midiprep kit (QIAGEN, Valencia, CA) and a Spin Doctor plasmid prep kit (Gerard Biotech, Oxford, OH) according to the manufacturer’s protocols. PFGE. Purified rickettsiae (approximately 2 ⫻ 109 to 6 ⫻ 109 cells) were washed in 0.75 ml of 1 M Tris-HCl (pH 7.6), 5 M NaCl (PETT buffer), collected at 18,400 ⫻ g and 4°C for 5 min in Eppendorf centrifuge tubes, resuspended in 75 to 200 ␮l of PETT buffer, mixed with an equal volume of molten 1% Incert agarose (BMA, Rockland, MD), and cast into blocks. The agarose blocks were digested in 1 ml 10 mM Tris-HCl (pH 8.0), 0.5 M EDTA (pH 8.0), 1% sodium lauroyl sarcosine containing 2 mg/ml proteinase K (Fisher, Pittsburgh, PA) for 24 h at 48°C. Fresh digestion solution was added, and the incubation was continued for 24 h. Proteinase K was inactivated by incubation in 1 ml of 4% phenylmethylsulfonyl fluoride (Sigma-Aldrich, St. Louis, MO) twice for 1 h each at 48°C. The blocks were washed twice in 1.5 ml TE (pH 7.6) and cast into PFGE gels (see below) or stored in 0.5 M EDTA (pH 8.0) at 4°C. Agarose blocks subjected to restriction enzyme digestion with HindIII, PvuI (Promega, Madison, WI), NcoI (New England Biolabs, Beverly, MA), or SmaI (Life Technologies, Rockville, MD) were equilibrated in TE as described above and immersed in 200 ␮l of the manufacturer’s 1⫻ buffer with 10 to 50 units of enzyme and incubated

4985

TABLE 1. PCR oligonucleotide primers Primer acronym

Hsp2F Hsp2R p05F p05R Sca12F Sca12R pRMJunc1 pRMJunc2 pRMGap1 pRMGap2

Sequence (5⬘ to 3⬘)

AAACTAATAGAGCGGGAGAAAG AACC AGGCAAAGGCGAGAGAAATACC TCTTCCGACTATCTTCTTCTTTA CCTG GACAGTTCTATCCACTCTCATCA AACG TGCCTGATTTTGCCAGTTATG ATGACCACACTTCCTCTATCCTCG CAAAGCAGGTGTAACCCAC GGTTTTTCCTGTTCCTGG CGTGGTGATAATGAGGAGTC GCACACCTGCTACTTTAGTGAC

pRM nucleotides

4362–4387 4810–4789 15169–15195 15509–15483 19982–20002 20441–20418 6942–6960 6549–6566 16787–16806 194–173

for 4 h at 37°C (25°C for SmaI). Fresh enzyme was added, and the incubation was continued overnight. Agarose blocks containing rickettsial DNA or 1-kbp (Life Technologies) or 5-kbp (Bio-Rad, Hercules, CA) linear molecular weight marker ladders were equilibrated twice in 1 ml TAFE running buffer (20 mM Tris free base, 5 mM EDTA free acid, 0.00025% glacial acetic acid) for 30 min at 4°C. The blocks were cast into a 1% agarose gel (Geneline LE agarose, PFGE certified; Beckman Instruments, Palo Alto, CA) in TAFE buffer and subjected to PFGE in a Beckman Geneline transverse alternating-field electrophoresis apparatus. Initial electrophoresis conditions were 220 mA for 8 h with 1-s pulses followed by 8 h with 2-s pulses at 5°C. After the initial experiments, current was reduced to 200 mA. Following electrophoresis, the gels were stained in TAFE buffer with 0.4 ␮g/␮l ethidium bromide or SYBR green I as suggested by the manufacturer (Cambrex, Rockland, ME) to visualize DNA. Two-dimensional agarose gel electrophoresis. For second-dimension electrophoresis, PFGE gels electrophoresed as described above were placed in a horizontal electrophoresis apparatus tray at a right angle to the original axis of anode-to-cathode orientation during PFGE and cast in place with fresh 1% agarose in TAFE buffer. The gels were electrophoresed for 80 min under constant current at 6 V/cm in TAFE buffer. Southern blot analyses. Rickettsial DNA, electrophoresed on 1% agaroseTAE gels, was transferred onto Zeta Probe GT genomic membrane (Bio-Rad) as described previously (5). Rickettsial DNA electrophoresed on PFGE gels was depurinated in 0.2 N HCl for 10 min, denatured in 0.4 N NaOH, 1 M NaCl for 35 min, transferred onto membrane in 0.4 M NaOH overnight, neutralized in 0.5 M Tris-HCl (pH 7.2), 1 M NaCl for 15 min, and baked at 80°C for 1 h. The blots were hybridized with digoxigenin-labeled probes prepared by PCR amplification, including a GFPuv gene probe specific for the pMOD658 transposon, as described previously (5) except that the pRM6, pRM16, and pRM21 probes (see below) were hybridized at 55°C. An R. felis pRF plasmid-specific probe was amplified from R. felis genomic DNA by use of the primer pair pRFh-pRFi (32), yielding a 726-bp product spanning the 3⬘ end of the gene encoding the Sca12 cell surface antigen and the 5⬘ end of the gene encoding an ATP-dependent protease. Cycle conditions were as follows: 1 cycle at 95°C for 1 min; 30 cycles at 95°C for 30 s, 48°C for 30 s, and 72°C for 1 min; and a final 5-min cycle at 72°C. Probes specific for the R. monacensis plasmid pRM6, pRM16, and pRM21 genes were prepared using primer pairs Hsp2F/Hsp2R, p05F/p05R, and Sca12F/Sca12R, respectively (Table 1). Cycle conditions were essentially as described above but were extended to 40 cycles, and annealing temperatures were 48°C, 50°C, and 52°C for the pRM6, pRM16, and pRM21 gene primer pairs, respectively. Primers were synthesized by Integrated DNA Technologies (Coralville, IA). Cloning and sequencing of the R. monacensis plasmid. To clone restriction enzyme digestion fragments of the R. monacensis pRM plasmid bearing the pMOD658 transposon carrying a CHL acetyltransferase (CAT) resistance marker, DNA prepared from transformant Rmona658A and Rmona658B populations was digested overnight individually with EcoRI, EcoRV, HindIII, HpaI, NcoI, SpeI, SphI, or StuI (New England Biolabs or Promega) in the manufacturer’s 1⫻ buffers at 37°C. Restriction fragments were treated with a DNA Terminator end repair kit, ligated into the pSMART-LCKan vector with Clone Smart DNA ligase, and electroporated into E. cloni 10G electrocompetent cells according to the manufacturer’s suggested protocols (Lucigen, Madison, WI). Electroporated cells were plated on Luria-Bertani agarose plates containing 50 ␮g per ml of CHL. Resistant colonies growing at 37°C were cultured in LB

4986

BALDRIDGE ET AL.

FIG. 1. GFPuv-fluorescent Rmona658B transformant cells in tick ISE6 cells. Arrows indicate single (bottom right), doublet form (top left), and multicell chain form (bottom center) rickettsiae. Bar, 10 ␮m. Magnification, ⫻1,000, with a fluorescein isothiocyanate filter.

medium with 50 ␮g per ml CHL. LCKan plasmid DNA bearing the rescued pMOD658 transposon was prepared from resistant cultures with a High Pure plasmid isolation kit (Roche). Rickettsial DNA flanking the pMOD658 transposon was sequenced (ABI 377 automated sequencer; Advanced Genetic Analysis Center, University of Minnesota) by a primer walking strategy beginning with the pMOD SqRP and SqFP primers (Epicenter, Madison, WI) followed by a succession of primers designed from each previously obtained sequence. To sequence the complementary strands, the same strategy was used beginning with the LCKan vector SL1 and SR1 primers. Sequences were manually edited as word-processing files to trim ends and aligned to generate two overlapping contigs by use of the MacVector (MacVector Inc., Cary, NC) and Sequencher (Gene Codes Corp., Ann Arbor, MI) programs. To clone PCR amplicons from the R. monacensis pRM plasmid, primer pairs pRmJunc1/pRmJunc2 and pRmGap1/pRmGap2 (Table 1) were designed from the restriction fragment contig sequences and used to amplify 412- and 6,894-bp products, respectively, from Rmona658B genomic DNA by use of an Expand Long template system with the manufacturer’s system 3 cycle parameters (Roche). The PCR products were gel purified with a QIAGEN gel extraction kit (QIAGEN) and sequenced as described above. The larger product was cloned in the pCR4 vector (Invitrogen, Carlsbad, CA). Nucleotide sequence accession number. The R. monacensis pRM plasmid sequence has been deposited at GenBank and assigned accession number EF564599.

RESULTS Transformant R. monacensis populations with multiple pMOD transposon integrations. Electroporation of R. monacensis with the pMODompACAT\GFPuv658 transposon, followed by growth in tick host cells under selection with CHL, resulted in the isolation of the Rmona658 clone (5). We obtained two additional resistant rickettsia populations from independent electroporation cuvettes. Both expressed GFPuv and were designated Rmona658A and Rmona658B (Fig. 1). Southern blot analyses, using a GFPuv gene probe, indicated that the genome of Rmona658 contained one integrated pMOD658 transposon, as shown by single hybridization bands migrating relative to marker DNA at approximately 3.3 kbp in HindIII (Fig. 2, lane 2)- and 7 kbp in EcoRI (lane 6)-digested genomic DNA. The specificity of the probe was confirmed by hybridization to pMOD658 (Fig. 2, lane 10, band at 4.5 kbp) but not to pMOD vector alone (lane 9) or to untransformed R.

APPL. ENVIRON. MICROBIOL.

FIG. 2. Southern blot analysis of pMODompACAT\GFPuv658 transposon insertions in genomic DNA of R. monacensis transformant populations Rmona658, Rmona658A, and Rmona658B. Untransformed R. monacensis (lanes 1 and 5), Rmona658 (lanes 2 and 6), Rmona658A (lanes 3 and 7), Rmona658B (lanes 4 and 8), pMOD2⬍MCS⬎ (lane 9), and pMODompACAT\GFPuv658 (lane 10) DNA was digested with HindIII (lanes 1 to 5) or EcoRI (lanes 6 to 10). The membranes were probed with a GFPuv gene-specific probe (6). Migration positions of linear DNA markers in a 1-kbp ladder are indicated at right (3 to 12 kbp).

monacensis genomic DNA (lanes 1 and 5). In contrast, multiple integrated transposons were detected in DNA from Rmona658A as four or five incompletely resolved HindIIIdigested bands at approximately 2.5 to 3.7 kbp and at 5.4 kbp (Fig. 2, lane 3) and as EcoRI-digested bands at approximately 3.3, 3.9, 4.7, 6.5, and 12 kbp (lane 7). Similarly, two transposon integrations were detected in DNA from Rmona658B as HindIII-digested bands at approximately 3.5 and 5.5 kbp (Fig. 2, lane 4) and as EcoRI-digested bands at approximately 3.9 and 12 kbp (lane 8). We interpreted the results to suggest two alternatives that were not mutually exclusive: (i) that the Rmona658A and 658B populations recovered from single, independent electroporations were derived from multiple transformed founder cells with single transposons integrated at different sites in the rickettsial genome or (ii) that they were derived from single founder cells with multiple transposons integrated within the same genome. The fact that no transformant rickettsiae were obtained from many other electroporation attempts argued against the first alternative as a sole explanation of the results. We reasoned that the recently reported presence of a plasmid in R. felis (32), if also the case for R. monacensis, would provide an additional transposon integration target that could facilitate multiple integrations within the genome of a single founder cell. We confirmed that hypothesis, as described below, by using PFGE and Southern blot analyses to demonstrate the existence of a plasmid containing integrated pMOD658 transposons within the genomes of Rmona658A and Rmona658B transformants. The plasmid was then cloned as three overlapping fragments and sequenced, producing a composite circular plasmid sequence. Probes derived from genes on each of the fragments were used in a final round of Southern blot experiments to physically confirm that each cloned fragment was derived from the plasmid bands observed on PFGE gels and that similar genes were present on a plasmid of another SFG rickettsia, R. amblyommii isolate WB-8-2. Detection of plasmids in rickettsial genomes by Southern blot analyses of PFGE gels. We exploited the presence of a plasmid in the R. felis genome to verify that Southern blot analyses of rickettsial genomic DNAs electrophoresed on

VOL. 73, 2007


4987

FIG. 3. PFGE and Southern blot analyses of R. felis and R. amblyommii isolate WB-8-2 DNA hybridized with the R. felis pRF plasmid probe (32). R. felis DNA (lane 1) digested with PvuI (lane 2) and R. amblyommii DNA (lane 3) digested with PvuI (lane 4). Migration positions of linear DNA markers in a 5-kbp ladder are indicated at right.

PFGE gels could be used as a method to test our hypothesis that multiple transposon integrations in the Rmona658A and Rmona658B populations were at least partially due to integration within an R. monacensis plasmid. Using a probe specific for the pRF plasmid, we applied the method to studying the R. felis isolate LSU, whose genome includes the 63-kbp pRF plasmid containing a single PvuI restriction enzyme site (38). Undigested R. felis genomic DNA that comigrated on PFGE gels with linear 70- to 100-kbp markers hybridized with the pRF probe (Fig. 3, lane 1). The probe also hybridized with PvuI-digested R. felis DNA that migrated just below the 65-kbp marker at the predicted position of linear plasmid (Fig. 3, lane 2). Undigested DNA of R. amblyommii isolate WB-8-2 that migrated in the range of the 25- to 50-kbp markers also hybridized with the pRF probe (Fig. 3, lane 3) as did PvuIdigested WB-8-2 DNA that migrated at the positions of the 20to 25-kbp markers (lane 4). The pRF probe did not hybridize with R. monacensis, R. montanensis, or R. peacockii DNA (not shown). The results confirmed that Southern blot analysis of PFGE gels could detect the R. felis 63-kbp pRF plasmid, indicated the presence of a smaller plasmid in R. amblyommii recognized by the pRF plasmid probe, and confirmed that the method was an appropriate means to analyze the Rmona658A and Rmona658B transformant populations for evidence of transposon integration within a rickettsial plasmid. Southern blot analyses confirm pMOD transposon integration in a plasmid of R. monacensis. Ethidium bromide staining of R. monacensis and R. amblyommii genomic DNA electrophoresed on PFGE gels showed the presence of randomly sheared chromosomal DNA migrating above the position of the linear 100-kbp marker DNA (Fig. 4A, lanes 1, 2, 4, and 6). Potential plasmids were again detected in R. amblyommii as DNA bands (Fig. 4A, lane 1) that migrated in the range of the 25- to 50-kbp DNA markers (lane 8) and in untransformed R. monacensis (lane 2), transformant clone Rmona658 (lane 4), and transformant population Rmona658A (lane 6) as DNA bands that migrated in the range of the 25- to 50-kbp markers.

FIG. 4. PFGE and Southern blot analyses of R. monacensis pMOD658 transformants. (A) Ethidium bromide-stained PFGE gel with DNA from untransformed R. amblyommii isolate WB-8-2 (lane 1), untransformed R. monacensis (lane 2) digested with HindIII (lane 3), Rmona658 (lane 4) digested with HindIII (lane 5), and Rmona658A (lane 6) digested with HindIII (lane 7). Lanes 8 and 9 contain 5- and 1-kbp DNA ladders, respectively, with sizes indicated at right. (B) Southern blot of gel shown in panel A hybridized with a GFPuv gene probe specific for the pMOD658 transposon (6). (C) Ethidium bromide-stained PFGE gel with DNA from untransformed R. monacensis (lane 1), Rmona658 (lane 2), Rmona658B (lane 3) digested with NcoI (lane 4) or SmaI (lane 5), and Rmona658B QIAGEN plasmid prep DNA (lane 6) digested with NcoI (lane 7) or SmaI (lane 8). Lane 9 contains 1- and 5-kbp DNA marker ladders with sizes indicated at right. (D) Southern blot of gel shown in panel C hybridized with the GFPuv gene probe. Arrowheads indicate positions of the rickettsial plasmid containing the pMOD658 transposon (DNA bands migrating in a range between 23 and 100 kbp).

Digestion of the R. monacensis DNAs with HindIII, which cleaves once within the pMOD568 transposon, resulted in the near disappearance of the potential plasmid DNA bands (Fig. 4A, lanes 3, 5, and 7). Southern blot analysis of the PFGE gel, hybridized with a GFPuv gene probe specific for the pMOD658 transposon (5), showed no hybridization to R. amblyommii or untransformed R. monacensis DNA (Fig. 4B, lanes 1, 2, and 3). The probe hybridized to Rmona658 undigested chromosomal DNA that migrated above the position of the 100- to 120-kbp markers (Fig. 4B, lane 4) and to a HindIIIdigested band (lane 5) that comigrated with the 3-kbp marker, as expected of the single chromosomally integrated copy of the pMOD658 transposon in Rmona658 (5). The probe hybridized both to undigested Rmona658A chromosomal DNA and to the potential plasmid bands at 25 to 50 kbp (Fig. 4B, lane 6), indicating pMOD658 transposon integration. The probe hybridized to Rmona658A DNA digested with HindIII that migrated as bands at 2.5, 3, and 3.7 kbp (Fig. 4B, lane 7), but the additional 5.4-kbp band that was originally observed (Fig. 2,

4988

BALDRIDGE ET AL.

lane 3) was absent, suggesting the possible loss of a subpopulation of cells during further passage of the Rmona658A transformant population. The results were consistent with the single-copy chromosomal integration of the pMOD658 transposon near a HindIII site in clone Rmona658 (5) and indicated at least three transposon integrations within Rmona658A chromosomal and potential plasmid DNA. We extended the PFGE and Southern blot analysis to transformant Rmona658B DNA digested with NcoI and SmaI, which cleave once each at G/C-rich recognition sites within the pMOD658 transposon. The sites were expected to occur rarely in A/T-rich rickettsial DNA. On ethidium bromide-stained PFGE gels, undigested chromosomal DNA was visible as a broad band migrating above the position of the 100-kbp linear DNA marker (Fig. 4C, lanes 1, 2, and 3). Potential plasmid DNA bands again migrated in the range of the 25- to 50-kbp markers in lanes containing DNA of untransformed R. monacensis and transformant Rmona658 (Fig. 4C, lanes 1 and 2, respectively) but displayed slightly retarded mobility in the lane containing transformant Rmona658B DNA (lane 3), consistent with insertion of the 2.1-kbp pMOD658 transposon in the plasmid. Digestion of Rmona658B DNA with NcoI and SmaI resulted in complex band patterns (Fig. 4C, lanes 4 and 5, respectively) that indicated much more frequent cleavage of rickettsial DNA by NcoI than by SmaI, but DNA bands were not apparent in lanes that each contained 1.5 ␮g of Rmona658B DNA eluted from a QIAGEN midiprep plasmid purification column and loaded on the gel before digestion (lane 6) or after digestion with NcoI (lane 7) and SmaI (lane 8), suggesting a low content of plasmid in that DNA preparation. Southern blot analysis of the PFGE gel showed that the GFPuv probe did not hybridize to untransformed R. monacensis DNA (Fig. 4C, lane 1) but did hybridize as predicted (5) to Rmona658 chromosomal DNA that migrated above the position of the 100-kbp marker (lane 2). The probe also hybridized to undigested Rmona658B chromosomal DNA and to the potential plasmid DNA bands (Fig. 4C, lane 3) as well as a less abundant and possibly linear form of plasmid that migrated slightly ahead of the 25-kbp marker. In NcoI-digested Rmona658B DNA, the probe hybridized predominantly to 9and 14-kbp bands and less strongly to 7- and 25-kbp bands (Fig. 4C, lane 4). The 50-kbp band seen in Fig. 4C, lane 3, was no longer apparent. In SmaI-digested Rmona658B DNA (Fig. 4C, lane 5), both the 25- and the 50-kbp bands were absent and the probe hybridized to a band at approximately 23 kbp as well as a band of digested chromosomal DNA that migrated at approximately 100 kbp. In undigested Rmona658B plasmid prep DNA, the probe hybridized to bands that migrated at approximately 100, 50, 25, and 23 kbp (Fig. 4C, lane 6). In NcoI- and SmaI-digested Rmona658B plasmid prep DNA (Fig. 4C, lanes 7 and 8, respectively), the probe again hybridized to the previously observed 9- and 14-kbp NcoI digest bands and to a single SmaI-digested band at approximately 23 to 25 kbp. The results showed that a potential plasmid in the Rmona658A and Rmona658B transformant populations that migrated in the range of the 25- to 50-kbp linear DNA markers on PFGE gels contained a copy of the pMOD658 transposon. Digestion of Rmona658B with NcoI and SmaI indicated that the linearized plasmid, including the transposon, was 23 to 25 kbp in length, and we interpreted the 23-, 25-, and 50-kbp


FIG. 5. Two-dimensional gel electrophoresis and Southern blot analysis of the R. monacensis pRM plasmid. (A) Ethidium bromidestained PFGE gel with R. monacensis DNA (lane 1) and linear DNA markers (lane 2) in a 5-kbp ladder with sizes indicated at right. The large arrow indicates the negative-to-positive polarity of current. (B) Constant field electrophoresis (CFE) of PFGE gel shown in panel A. The large arrow indicates the negative-to-positive polarity of current. (C) Southern blot analysis of gel shown in panel B with a probe specific for the pRM6 gene. Putative conformational identities of DNA species present in lanes 1 of all panels are indicated at left as chromosomal (chr), mixed circular and linear dimeric plasmid (c/li), supercoiled plasmid (sc), and linear monomeric plasmid (li).

hybridizing bands in Fig. 4C and D as linear, supercoiled, and open circular plasmid isomers, respectively (22). We confirmed that interpretation by using the two-dimensional electrophoresis procedure employed to distinguish linear from circular plasmids in Borrelia burgdorferi and Borrelia hermsii (17, 18). R. monacensis DNA was electrophoresed on PFGE gels as described above in the first dimension and subsequently stained, before second-dimension constant current electrophoresis, with ethidium bromide at the critical concentration which re-

VOL. 73, 2007


4989

FIG. 6. The R. monacensis pRM plasmid cloning and sequencing strategy. The plasmid is represented as a circle, with the number of base pairs from an arbitrary origin indicated on the inner surface. Lines with double arrowheads indicate the lengths and relative positions of sequenced restriction fragments obtained by CHL marker rescue of Rmona658A and Rmona658B DNA restriction enzyme digests. Lines with single arrowheads indicate the relative positions and orientations of DNA primers used to PCR amplify portions of the plasmid from Rmona658B DNA. The hatched and filled bars represent, respectively, a sequenced PCR product encompassing that portion of the plasmid not present on the Rmona658A and Rmona658B restriction fragment clones and a sequenced product spanning their overlap. Filled circles indicate the relative locations of the pMOD658 transposon within the plasmid in the Rmona658A and Rmona658B transformants.

tards the migration rate of supercoiled plasmid isomer to that of open circular isomer (22). The destained first-dimension PFGE gels again showed a chromosomal DNA band migrating at approximately 100 kbp, the plasmid bands at 25 to 50 kbp corresponding to presumed supercoiled and circular plasmid isomers, previously uncertain bands migrating at approximately 35 and 45 kbp, and a faint band of presumably linear plasmid isomer migrating at approximately 23 kbp (Fig. 5A, lane 1). Second-dimension electrophoresis resulted in the migration of sheared chromosomal DNA and the 23-kbp plasmid band to the right at a rate resembling that of the linear DNA markers (Fig. 5B, lane 1 versus lane 2), confirming the presence of linear DNA in the 23-kbp band. The 25-kbp plasmid band did not migrate, as expected, but the 50-kbp plasmid band resolved into a nonmigrating, presumably circular fraction and a migrating, presumably linear fraction, as did the 35-kbp band. The 45-kbp band migrated at a rate corresponding to that of the linear marker DNAs. Southern blot analysis of the gel shown in Fig. 5B was conducted using a probe specific for a small heat shock protein gene (pRM6) present on the R. monacensis plasmid (see below). The probe hybridized

strongly to the immobile supercoiled plasmid isomer at 25 kbp (Fig. 5C, lane 1) and less strongly to the 23-kbp linear isomer that migrated to the right of lane 1. The probe also hybridized strongly to the mobile 50-kbp linear isomer that migrated to the right of lane 1 but less strongly to the immobile 50-kbp circular isomer and only weakly to the 35- to 45-kbp bands. The results suggested the existence of a circular plasmid in R. monacensis that was also present as a comparatively rare monomeric linear isomer and an abundant dimeric linear isomer. Cloning and sequencing the R. monacensis plasmid. Because we were unable to isolate the intact R. monacensis plasmid by CHL marker rescue of E. coli cells electroporated with Rmona658A or Rmona658B genomic or plasmid prep DNA, we used the same method to rescue restriction enzyme digestion fragments of rickettsial DNA containing the pMOD658 transposon after ligation into pSMART-LCKan, a vector optimized for cloning A/T-rich DNA that is unstable in E. coli. The restriction fragments were sequenced and assembled into composite 7- and 11-kbp DNA sequences from Rmona658A and Rmona658B, respectively, that overlapped by 150 bp, as shown schematically (Fig. 6). Each of the composite sequences

4990

BALDRIDGE ET AL.


FIG. 7. Predicted genes and pseudogenes on pRM. The plasmid is represented as a circle, with the number of base pairs from an arbitrary origin indicated on the inner surface. Black arrows indicate genes identified as BLAST best matches to R. felis pRF plasmid genes. Gray arrows indicate genes (pRM12, pRM13, pRM14, and pRM22) identified as best matches to R. felis or R. bellii chromosomal genes. The hatched arrow indicates a hypothetical protein gene unique to R. monacensis. Unfilled arrows indicate genes identified as BLAST best matches to genes of bacteria other than Rickettsia species.

contained a single pMOD658 transposon encoding the CHL resistance marker, which was flanked by the 9-bp insertion site duplications characteristic of pMOD transposase activity. Using the pRmJunc1 and 2 primers designed to hybridize near the overlapping ends of the two sequences, we PCR amplified and sequenced a 412-bp product from Rmona658B DNA (Fig. 6) which confirmed that the overlapping Rmona658A and Rmona658B restriction fragment sequences were present as directly adjacent sequences on the same DNA template. Using the pRmGap1 and -2 primers hybridizing to the opposite ends of the two composite restriction fragment sequences, we PCR amplified and sequenced a 7-kbp product (Fig. 6) that encompassed the remaining portion of the plasmid, allowing final sequence assembly of the circular R. monacensis plasmid, designated pRM. The 23.5-kbp pRM sequence (excluding the 2.1-kbp pMOD transposons present as single copies in the plasmids found in the Rmona658A and Rmona658B transformant populations) was in close agreement with our estimate of a 23- to 25-kbp plasmid based on PFGE analyses. The numbers and relative positions of EcoRI, HindIII, NcoI, and SmaI restriction enzyme recognition sites within the pRM sequence were consistent with the results of Southern blot analyses that employed those enzymes (Fig. 2 and 4). Finally, marker rescue cloning of restriction enzyme digestion fragments resulted in the recovery of additional clones containing three independent chromosomal integrations of the pMOD658 transposon from

the Rmona658A population and one from the Rmona658B population, consistent with results of our initial Southern blot analyses of transposon integration sites in the rickettsial populations (Fig. 2). pRM plasmid sequence. The R. monacensis pRM plasmid sequence was 23,486 nucleotides in length and had a G/C content of 31.8%. Based on nucleotide and translated protein BLAST searches of the GenBank database with the pRM sequence as the query, we identified 23 predicted genes or pseudogenes (Fig. 7), of which 22 contained open reading frames (ORFs) that encoded hypothetical proteins with significant similarities to proteins encoded by other bacteria (Table 2). Four pRM genes (pRM17 to pRM20) were identified as members of an operon by the FGENESB program (SoftBerry, Inc.), while pRM6 and pRM7 constituted a smaller operon. Three pRM genes encoded hypothetical proteins that were best E value matches to those encoded by R. felis chromosomal genes, and 11 others encoded hypothetical proteins that were best matches to those encoded by R. felis pRF genes. The pRM genes generally had nucleotide similarities of 70 to 90% to the cognate pRF genes, but gene order on the two plasmids was not colinear. The tentative functional identities assigned to hypothetical proteins encoded by pRM (Table 2) indicated a primary emphasis on DNA replication, partitioning, and mobilization and a possible secondary emphasis on host interaction and adaptation. We emphasize that the identities of the


VOL. 73, 2007

4991

TABLE 2. Genes and pseudogenes on the pRM plasmid as determined by BLASTx analysis ORF/positiona

Taxon (annotation)

Accession no.

Score (bits/E value)

R. felis ORFb

pRF41 Split genec pRF37 split Rf/Age partiald pRF38 Partial pRF/pRMf Split partial

pRM1/1523–794 pRM2/2181–1973

R. felis (YhgA-like transposase) R. felis (conjugative transfer protein TraD_Ti)

NC_007110 NC_007110

343/7e-92 66.6/3e-21

pRM3/2854–3417 pRM4/3713–3127 pRM5/3849–4232 pRM6/4726–4250 pRM7/5303–4779 pRM8/7613–5448 pRM9/7869–7641

R. felis (conjugative transfer protein TraA_Ti) Piscirickettsia salmonis (transposase DDE domain)h Photobacterium damselae (transposon resolvase) R. felis (small heat shock protein 2) R. felis (small heat shock protein 1) Nitrobacter hamburgensis (helicase RecD/TraA) R. felis (DNA polymerase III epsilon subunit-like)

NC_007110 AF_184152 DQ_069059 NC_007110 NC_007110 NC_007964 NC_007110

128/3e-27 114/1e-27 373/4e-33 682/6e-69 638/8e-64 787/0 120/4e-25

pRM10/8300–7896 pRM11/8788–8363 pRM12/10214–8964 pRM13/11300–11978 pRM14/13247–11931 pRM15/13638–14396 pRM16/17067–14605 pRM17/17289–17663 pRM18/18255–17665 pRM19/18812–18309 pRM20/19378–18872 pRM21/19518–20538 pRM22/22588–21698 pRM23/23291–22713

R. felis (hypothetical protein pRF65) Sulfitobacter sp. NAS-14.1 (transposase) R. felis (proline/betaine transporter) R. felis (SpoT15 stringent response/adaptation) R. felis (proline/betaine transporter) Ralstonia eutropha (HemK O-methyltransferase) R. felis (DnaA-like replication initiator protein) R. monacensis (unique hypothetical protein) Stappia aggregata (ParA chromosome stability) R. felis (hypothetical protein) R. felis (hypothetical protein) R. felis (cell surface antigen Sca12) R. bellii (chromosome segregation ATPase) Burkholderia thailandensis (transposon resolvase NC_007650)

NC_007110 ZP_00964727 NC_007109 NC_007109 NC_007109 NC_007347 NC_007110

126/1e-27 129/5e-29 438/1.5e-120 239/1e-69 463/3e-128 56.2/3e-06 413/5e-113

NZAAUW01000047 54.3/6e-05 NC_007110 81.6/1e-13 NC_007110 80.5/2e-13 NC_007110 140/8e-31 NZAARC01000001 368/2e-99 254/3e-65

Note

pRF51 pRF52

Rf/pRF/pRMg

pRF34

Split partial pRF/ pRM Rf/pRF/pRM Partial

pRF65 RF0885 RF0384 RF0441

Split gene

pRF05

Rf/Ag

pRF03 pRF02 pRFd02 pRF24

Split partial gene

a

Nucleotide positions of the ORFs on the pRM plasmids. Plasmid or chromosomal ORF of R. felis encoding similar protein. Rickettsial split gene; see references 8 and 31. d The pRM gene contains a partial domain relative to the BLASTx match. e Among Rickettsia species, the gene has previously been found only in R. felis and in ancestral group members R. belli and R. canadensis; see reference 20. f Among Rickettsia species, the gene has previously been found only on the R. felis pRF plasmid (20). g Among Rickettsia species, the gene has previously been found only on the R. felis chromosome and the pRF plasmid (20). h DDE domain, Asp Asp Glu catalytic domain. b c

pRM genes and pseudogenes are based solely on similarity searches and that the translation of functionally active proteins has not been demonstrated. Genes on pRM that contained ORFs for protein domains with known roles in DNA replication and partitioning included pRM16, which encoded a DnaA-like chromosomal replication initiator domain identified as a best match of that encoded by the pRF05 gene, and pRM9, which encoded a partial DNA polymerase III epsilon subunit-like domain identified as a best match of that encoded by pRF34. The pRM18 gene encoded a protein identified as a best match of a Stappia species ParAlike chromosomal stability protein with a probable role in regulation of cytokinesis. A similar ParA-like protein is encoded by pRF23. The pRM22 gene encoded a protein identified as a best match to a chromosome segregation ATPase encoded by an R. bellii gene, while pRM8 encoded a protein identified as a best match to a Nitrobacter species-encoded RecD helicase/TraA protein with DNA replication, recombination, and repair functions. Genes on pRM with ORFs encoding protein domains with known roles in DNA mobilization included pRM2 and pRM3, which encoded Agrobacterium-like TraD_Ti and TraA_Ti conjugative DNA transfer protein domains. They were identified as best matches of the proteins encoded by the pRF plasmid pRF37 and pRF38 genes. The high density of transposase genes on pRF was also characteristic of pRM, which carried a

rickettsial split gene (8, 31), designated pRM1, that encoded a protein identified as a best match of the pRF41-encoded transposase. The pRM4 and pRM11 genes encoded proteins identified as best matches to major portions of transposases encoded by Piscirickettsia and Sulfitobacter species, respectively, while the pRM5 and pRM23 genes encoded transposon resolvases identified as best matches to those of Photobacterium and Burkholderia species, respectively. The pRM plasmid also carried apparent small remnants of several transposase genes found on pRF or in other bacteria (not shown). Several genes on pRM could be involved in interaction and adaptation of R. monacensis with host cells. They included the pRM12 and pRM14 genes, which encoded proline/betaine membrane transporters, identified as best matches to those encoded by the R. felis chromosomal genes RF0885 and RF0441. The transporter genes were located to either side of pRM13, a split gene on the opposite strand that encoded a protein identified as a best match of SpoT15, encoded by the R. felis RF0384 gene, and a member of a family of nutritional stress response transcriptional regulators. The pRM6 and pRM7 genes encoded proteins identified as best matches of the small heat shock proteins Hsp2 and Hsp1 encoded by pRF51 and pRF52, respectively. The pRM19 and pRM20 genes encoded hypothetical proteins that were best matches of the pRF02 and pRFd02 gene products. A helix-turn-helix DNA binding domain typical of some bacterial transcriptional regu-

4992

BALDRIDGE ET AL.

latory proteins was present in the amino-terminal end of the pRM19 protein but was absent from the otherwise very similar pRM20 protein. Lastly, pRM21 encoded a best match of the R. bellii cell surface antigen family protein Sca12. We identified three pRM genes without obvious potential roles in DNA metabolism or host cell interactions. They included the pRM15 gene, which encoded a protein identified as a best match to a Ralstonia species HemK O-methyltransferase probably involved in regulation of protein translation. The pRM10 gene encoded a hypothetical protein that was identified as a best match to a hypothetical protein of unknown function encoded by the pRF65 gene. The pRM17 gene encoded a hypothetical protein unique to R. monacensis. Because bacterial genomes, including those of Rickettsia spp. (16), are known to contain various types of repetitive DNA sequences of potential functional significance, we examined the pRM nucleotide sequence for the presence of repetitive DNA. The sequence contained four perfect palindromes at nucleotide positions 7663 to 7676, 7989 to 8002, 12072 to 12085, and 16069 to 16098. Four imperfect palindromes with 1to 5-bp central mismatches occurred at positions 2659 to 2685, 11161 to 11201, 15646 to 15670, and 15971 to 16004. Two dispersed copies of a 69-bp perfect repeat occurred at positions 17074 to 17142 and 19444 to 19512. Two dispersed copies each of 37- and 44-bp imperfect repeats occurred, respectively, at positions 3792 to 3828 and 20773 to 20809 and at 10192 to 10235 and 13123 to 13166. Tandem arrays of repeats were not present. Southern blots confirm the integrity of the circular pRM sequence and the presence of probable orthologous genes on the plasmid of R. amblyommii. We used Southern blots with probes specific for pRM6, pRM16, and pRM21, genes that were highly similar to the R. felis pRF plasmid genes pRF51, pRF05, and pRF24, respectively (Table 2), to verify the integrity of the deduced pRM sequence and to test for the presence of similar genes on the plasmid of R. amblyommii isolate WB8-2. DNAs of R. monacensis, the Rmona658B transformant, and R. amblyommii were again separated on PFGE gels, and plasmid bands (Fig. 8A, lanes 2 to 9) that migrated in the range of the 25- to 50-kbp markers (lane 1) were visualized by staining with SYBR green. The DNAs were transferred onto hybridization membrane, and lanes 1 to 3 were hybridized with the GFPuv probe specific for the pMOD658 transposon. It again hybridized strongly to the 25- to 50-kbp plasmid bands and the 100-kbp chromosomal DNA band of Rmona658B (Fig. 8B, lane 2) but not to untransformed R. monacensis DNA (lane 3). The probe for pRM16 on the 11-kbp restriction fragment cloned from Rmona658B (Fig. 6 and 7) hybridized strongly to the same 25- to 50-kbp plasmid DNA bands but not to chromosomal DNA of Rmona658B (Fig. 8B, lane 4). It hybridized less strongly to R. amblyommii plasmid bands of similar mobility (Fig. 8B, lane 5), suggesting the presence of a DnaA replication initiator gene with moderate sequence similarity to the pRM16/pRF05 genes. The probe for pRM6 on the 7-kbp restriction fragment cloned from Rmona658A (Fig. 6 and 7) hybridized strongly to the 25- to 50-kbp plasmid DNA bands of Rmona658B (Fig. 8B, lane 8) and to the R. amblyommii plasmid bands (lane 9), suggesting the presence of a small heat shock protein gene with high sequence similarity to the pRM6/ pRF51 genes. The probe for pRM21 on the 7-kbp PCR prod-


FIG. 8. Southern blot confirmation of the circular pRM sequence and evidence for the presence of orthologous genes on the plasmid of R. amblyommii. (A) SYBR green-stained PFGE gel with DNA from the Rmona658B transformant of R. monacensis (lanes 2, 4, 6, and 8), untransformed R. monacensis (lane 3), and R. amblyommii isolate WB-8-2 (lanes 5, 7, and 9). Migration positions of linear DNA markers in a 5-kbp ladder (lane 1) are indicated at left. (B) Southern blot analysis of gel shown in panel A hybridized with probes specific for the GFPuv gene (lanes 1, 2, and 3), the R. monacensis pRM16 gene (lanes 4 and 5), the pRM21 gene (lanes 6 and 7), and the pRM6 gene (lanes 8 and 9).

uct spanning the region of plasmid not contained in the restriction fragment clones (Fig. 6 and 7) hybridized strongly to the 25- to 50-kbp plasmid DNA bands and to chromosomal DNA of Rmona658B (Fig. 8B, lane 6). It hybridized weakly to R. amblyommii plasmid bands of similar size (Fig. 8B, lane 7), suggesting the presence of a probable Sca12 gene with low sequence similarity to the pRM21/pRF24 genes. The integrity of the pRM sequence was confirmed by hybridization of the GFPuv probe and probes specific for genes on each of the three independently cloned and sequenced fragments of pRM to the same 25- to 50-kbp plasmid DNA bands originally detected with the GFPuv probe (Fig. 4). The results also showed that probable orthologs of genes present on both pRM and pRF were present on an R. amblyommii plasmid of similar size. DISCUSSION The pRM plasmid of R. monacensis and the pRF plasmid of R. felis were remarkably similar. Eleven of the 23 predicted genes and pseudogenes on pRM encoded hypothetical proteins that were BLASTx best matches to those encoded by pRF genes. Orthologs of two of those gene pairs, pRM3/pRF38 and pRM9/pRF34 (Table 2), were absent from the chromosomes of 10 sequenced Rickettsia species (20), and the genes were thus unique to the plasmids among rickettsiae. Rickettsial orthologs of the pRM6 and pRM10 genes were present only on pRF and the R. felis chromosome, and orthologs of pRM2 and pRM16 were present only on chromosomes of R. felis and the ancestral group species R. bellii and R. canadensis. The pRF plasmid was originally suggested as possibly unique to R. felis

VOL. 73, 2007

among rickettsiae after a PCR primer pair designed to amplify a region of pRF containing the 3⬘ end of the gene encoding Sca12 and the 5⬘ end of the adjacent gene encoding an ATPdependent protease failed to yield a product from genomic DNA of all available reference Rickettsia species (32). However, subsequent phylogenetic analysis of the pRF genes suggested an origin in the ancestral group (20). That hypothesis was supported by our demonstration of an R. amblyommii plasmid recognized on Southern blots by a probe amplified from R. felis genomic DNA with the above-described pRF primers (Fig. 3) and by the pRM21 probe (Sca12) as well as the pRM6 (Hsp2) and pRM16 (DnaA-like replication initiator) probes (Fig. 8). Our PFGE and Southern blot analyses (data not shown) using the same probes have demonstrated similar plasmids in other SFG rickettsiae from ixodid ticks, including R. helvetica, the MoAa and AaR/SoCarolina isolates of R. amblyommii, and a Rickettsia sp. isolated from Rhipicephalus sanguiness as well as a Rickettsia sp. isolated (submitted) from the argasid tick Carios capensis. Prior PFGE investigations of rickettsial DNA did not detect plasmids, probably because they were focused on estimation of genome size through analysis of restriction enzyme-digested linear DNA (15, 42, 54). Genome sequencing revealed the plasmid in R. felis, but sequencing of nine other rickettsial genomes, including that of R. bellii strain RML 369-C, which has a complete set of conjugation genes and forms apparent pili between cells (33), has not led to further reports of rickettsial plasmids. We suggest that while plasmids may be widespread among SFG rickettsiae, their detection may have been obscured by possible isolate and laboratory passage effects as well as factors related to experimental design and/or genome sequence assembly. While we have not demonstrated that R. monacensis expresses functional proteins from pRM genes, the similarities in the functional natures of the hypothetical proteins encoded by pRM and pRF were nevertheless striking. A majority were predicted to have roles in DNA replication, partitioning, mobilization, and conjugation. Although the observation of apparent pili linking R. felis cells and the presence of chromosomal genes similar to vir genes of A. tumefaciens that encode type IV secretion system proteins involved in later stages of DNA transfer (11) led to the suggestion that pRF was conjugative (33), the conjugation functions encoded by pRF and pRM are incomplete and would be capable only of initiating DNA transfer. We have not attempted to demonstrate the presence of pili in R. monacensis, and neither plasmid has been demonstrated to be mobile. The presence on the plasmids of multiple transposase sequences and genes apparently derived from other bacteria provides a mechanistic explanation, in addition to host selective pressures, for differential gene loss and gain leading to size and sequence variations among rickettsial plasmids with a presumed common origin. A high density of transposase genes is not typical of rickettsial genomes in general but is characteristic of R. felis and R. bellii (32, 33) as well as of Orientia tsutsugamushi (16), the etiologic agent of scrub typhus transmitted by mites and the only member of the sister genus Orientia in the Rickettsiaceae family. Potential host adaptation functions were a secondary emphasis of pRM coding capacity, reminiscent of the plasmids of B. burgdorferi that have gene loci encoding proteins involved in nutrient acquisition and transport functions as well as outer


4993

surface proteins that play essential roles in interactions with hosts (48). Tick feeding, molting, and fasting lead to alternating periods of rickettsial activation, growth, and decline in response to changes in temperature, osmotic pressure, and the nutritional status of the tick (29). Proteins encoded by the spoT and proline/betaine transporter genes on pRM could be involved in the adaptation of R. monacensis to host conditions. In bacteria, SpoT proteins are regulators of the global cellular metabolism or “stringent” response to starvation and stress, which enhances cell survival (12). In B. burgdorferi (9, 13, 41) and R. conorii (43), both of which are tick-borne, spoT mRNAs were differentially expressed following vertebrate-to-arthropod host shifts and changes in nutritional stress. As obligate endosymbionts, rickettsiae do not synthesize many low-molecularweight metabolites, such as proline (40, 52), and rely on membrane transporters for access to host cell metabolites. Proline/ betaine transporters are members of the ABC transporter family and have been classically defined as maintaining cellular osmotic integrity by transporting uncharged solutes across the plasma membrane (21). In rickettsiae, they might perform both a nutritional role and a stress response role by stabilizing protein hydration shells (21). The small heat shock proteins encoded by pRM6 and pRM7 may also enhance rickettsial survival during stress by maintaining protein stability as they do in other bacteria (25, 47). While the pRM21 gene encoded a cell surface antigen, Sca12 (8), that could be involved in interactions with the host cell membrane, other genes with obvious potential for host invasion or virulence functions, such as those encoding a hyaluronidase and a patatin-like phospholipase encoded by pRF genes, were not present on pRM. Regardless of the above inferences drawn from the hypothetical coding capacities of pRM and pRF, the maintenance and role of plasmids in rickettsiae remain obscure. The rickettsial plasmid’s possession of genes with probable plasmid maintenance functions and with significant similarity to those of distantly related bacteria suggests an important role in rickettsial biology. The presence of both linear and circular isomers of pRM in rickettsial cells was suggestive of rolling-circle replication of a circular plasmid (23, 27). However, the multiple linear and circular plasmids of B. burgdorferi replicate independently (48). In addition, it has been suggested that the genesis of linear plasmid dimers in B. burgdorferi may involve a circular intermediate (28), and a linear plasmid of B. hermsii was shown to spontaneously convert to a stably maintained circular form (18). The as-yet-unknown mechanisms of rickettsial plasmid replication, maintenance, and possible conjugative transfer have significant implications for the development of transformation technologies and for the role of plasmids in rickettsial biology. As obligate endosymbionts of blood-feeding arthropods, rickettsiae are under constant selective pressure for reduction in genome size and exist in isolated small populations that may experience successive population bottleneck events (3, 51). In the face of such pressures, plasmids might serve to enhance rickettsial genetic plasticity, possibly through mechanisms involving the transfer of genetic information between chromosomes and plasmids, as has been shown for Buchnera (26). For decades, rickettsiae have been notoriously resistant to genetic manipulation with the techniques and plasmid vectors that have been successfully applied to many other bacteria (49,

4994

BALDRIDGE ET AL.

53). Plasmids that might provide a basis for rickettsial transformation vectors were thought not to exist in Rickettsia until the recent surprising report of pRF in R. felis (32). We have described a closely related plasmid, pRM, for R. monacensis and have obtained additional data indicating that a related plasmid exists in R. amblyommii and in other SFG Rickettsia species (unpublished), consistent with an ancestral group origin of pRF as suggested by Gillespie and colleagues (20). Independent transposon integrations of linked selectable marker and fluorescent reporter genes at two positions in pRM resulted in long-term stable expression of both genes in R. monacensis and allowed us to clone the plasmid as three large overlapping fragments. We have initiated the development of pRM as a transformation vector that may be of general utility in the genetic manipulation of Rickettsia. ACKNOWLEDGMENT This research was supported by NIH grant RO1 AI49424 to U.G.M.


19.

20.

21.

22.

23. 24.

25.

26.

REFERENCES 1. Akman, L., R. V. Rio, C. B. Beard, and S. Aksoy. 2001. Genome size determination and coding capacity of Sodalis glossinidus, an enteric symbiont of tsetse flies, as revealed by hybridization to Escherichia coli gene arrays. J. Bacteriol. 183:4517–4525. 2. Akman, L., A. Yamashita, H. Watanabe, K. Oshima, T. Shiba, M. Hattori, and S. Aksoy. 2002. Genome sequence of the endocellular obligate symbiont of tsetse flies, Wigglesworthia glossinidia. Nat. Genet. 32:402–407. 3. Andersson, S. G., A. Zomorodipour, J. O. Andersson, T. Sicheritz-Ponten, U. C. Alsmark, R. M. Podowski, A. K. Naslund, A. Eriksson, H. H. Winkler, and C. G. Kurland. 1998. The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396:133–140. 4. Azad, A. F., and C. B. Beard. 1998. Rickettsial pathogens and their arthropod vectors. Emerg. Infect. Dis. 4:179–186. 5. Baldridge, G. D., N. Burkhardt, M. J. Herron, T. J. Kurtti, and U. G. Munderloh. 2005. Analysis of fluorescent protein expression in transformants of Rickettsia monacensis, an obligate intracellular tick symbiont. Appl. Environ. Microbiol. 71:2095–2105. 6. Baroqui, P., P. Parola, P. E. Fournier, and D. Raoult. 2007. Spotted fever rickettsioses in southern and eastern Europe. FEMS Immunol. Med. Microbiol. 49:2–12. 7. Bell, E. J., G. M. Kohls, H. G. Stoenner, and D. B. Lackman. 1963. Nonpathogenic rickettsias related to the spotted fever group isolated from ticks, Dermacentor variabilis and Dermacentor andersoni from eastern Montana. J. Immunol. 90:770–781. 8. Blanc, G., M. Ngwamidiba, H. Ogata, P.-E. Fournier, J.-M. Claverie, and D. Raoult. 2005. Molecular evolution of Rickettsia surface antigens: evidence of positive selection. Mol. Biol. Evol. 22:2073–2083. 9. Bugrysheva, J., E. Y. Dobrikova, H. P. Godfrey, M. L. Sartakova, and F. C. Cabello. 2002. Modulation of Borrelia burgdorferi stringent response and gene expression during extracellular growth with tick cells. Infect. Immun. 70:3061–3067. 10. Burgdorfer, W., S. F. Hayes, L. A. Thomas, and J. L. Lancaster. 1981. A new spotted fever group rickettsia from the lone star tick (Amblyomma americanum), p. 595–602. In W. Burgdorfer and R. L. Anacker (ed.), Rickettsiae and rickettsial diseases. Academic Press, Inc., New York, NY. 11. Cascales, F., and P. J. Christie. 2003. The versatile type IV secretion systems. Nat. Rev. Microbiol. 1:137–149. 12. Chatterji, D., and A. K. Ohja. 2001. Revisiting the stringent response; ppGpp and starvation signaling. Curr. Opin. Microbiol. 4:160–165. 13. Concepcion, M. B., and D. R. Nelson. 2003. Expression of spoT in Borrelia burgdorferi during serum starvation. J. Bacteriol. 185:444–452. 14. Darby, A. C., J. Lagnel, C. Z. Matthew, K. Bourtiz, I. Maudlin, and S. C. Welburn. 2005. Extrachromosomal DNA of the symbiont Sodalis glossinidius. J. Bacteriol. 187:5003–5007. 15. Eremeeva, M., N. Balayeva, V. Ignatovich, and D. Raoult. 1995. Genomic study of Rickettsia akari by pulsed-field gel electrophoresis. J. Clin. Microbiol. 33:3022–3024. 16. Eremeeva, M. E., A. Madan, C. D. Shaw, K. Tang, and G. A. Dasch. 2005. New perspectives on rickettsial evolution from new genome sequences of Rickettsia, particularly R. canadensis, and Orientia tsutsugamushi. Ann. N. Y. Acad. Sci. 1063:47–63. 17. Ferdows, M. S., and A. G. Barbour. 1989. Megabase-sized linear DNA in the bacterium Borrelia burgdorferi, the Lyme disease agent. Proc. Natl. Acad. Sci. USA 86:5969–5973. 18. Ferdows, M. S., P. Serwer, G. A. Griess, S. J. Norris, and A. G. Barbour.

27. 28.

29.

30.

31.

32.

33.

34. 35. 36. 37.

38.

39.

40. 41.

42.

43.

44.

1996. Conversion of a linear to a circular plasmid in the relapsing fever agent Borrelia hermsii. J. Bacteriol. 178:793–800. Frati, F., I. Negri, P. P. Fanciulli, M. Pellecchia, and R. Dallai. 2006. Ultrastructural and molecular identification of a new Rickettsia endosymbiont in the springtail Onychiurus sinensis (Hexapoda, Collembola). J. Invertebr. Pathol. 93:150–156. Gillespie, J. J., M. S. Beier, M. S. Rahman, N. C. Ammerman, J. M. Shallom, A. Purkayastha, B. S. Sobral, and A. F. Azad. 2007. Plasmids and rickettsial evolution: insight from Rickettsia felis. PLoS ONE 3:1–17. Horn, C., S. Jenewein, L. Sohn-Bosser, E. Bremer, and L. Schmitt. 2005. Biochemical and structural analysis of the Bacillus subtilis ABC transporter OpuA and its isolated subunits. J. Mol. Microbiol. Biotechnol. 10:76–91. Johnson, P. H., and L. I. Grossman. 1977. Electrophoresis of DNA in agarose gels. Optimizing separations of conformational isomers of doubleand single-stranded DNAs. Biochemistry 16:4217–4225. Khan, S. A. 1997. Rolling-circle replication of bacterial plasmids. Microbiol. Mol. Biol. Rev. 61:442–455. Labruna, M. B., T. Whitworth, D. H. Bouyer, J. McBride, L. Marcelo, A. Camargo, E. P. Camargo, V. Popov, and D. H. Walker. 2004. Rickettsia bellii and Rickettsia amblyommii in Amblyomma ticks from the state of Rondonia, Western Amazon, Brazil. J. Med. Entomol. 41:1073–1081. Laksanalamai, P., D. L. Maeder, and F. T. Robb. 2001. Regulation and mechanism of action of the small heat shock protein from the hyperthermophilic archaeon Pyrococcus furiosus. J. Bacteriol. 183:5198–5202. LaTorre, A., R. Gil, F. J. Silva, and A. Moya. 2005. Chromosomal stasis versus plasmid plasticity in the aphid endosymbiont Buchnera aphidicola. Heredity 95:339–347. Maleszka, R., P. J. Skelly, and G. D. Clark-Walker. 1991. Rolling circle replication of DNA in yeast mitochondria. EMBO J. 10:3923–3929. Marconi, R. T., S. Casjens, U. G. Munderloh, and D. S. Samuels. 1996. Analysis of linear plasmid dimers in Borrelia burgdorferi sensu lato isolates: implications concerning the potential mechanism of linear plasmid replication. J. Bacteriol. 178:3357–3361. Munderloh, U. G., and T. J. Kurtti. 1995. Cellular and molecular interrelationships between ticks and prokaryotic tick-borne pathogens. Annu. Rev. Entomol. 40:221–243. Munderloh, U. G., S. D. Jauron, V. Fingerle, L. Leitritz, S. F. Hayes, J. M. Hautman, C. M. Nelson, B. W. Huberty, T. J. Kurtti, G. G. Ahlstrand, B. Greig, M. A. Mellencamp, and J. L. Goodman. 1999. Invasion and intracellular development of the human granulocytic ehrlichiosis agent in tick cell culture. J. Clin. Microbiol. 37:2518–2524. Ogata, H., S. Audic, P. Renesto-Audiffren, P.-E. Fournier, V. Barbe, D. Samson, V. Roux, P. Cossart, J. Weissenbach, J.-M. Claverie, and D. Raoult. 2001. Mechanisms of evolution in Rickettsia conorii and R. prowazekii. Science 293:2093–2098. Ogata, H., P. Renesto, S. Audic, C. Robert, G. Blanc, P.-E. Fournier, H. Parinello, J.-M. Claverie, and D. Raoult. 2005. The genome sequence of Rickettsia felis identifies the first putative conjugative plasmid in an obligate intracellular parasite. PLoS Biol. 3:e248. Ogata, H., B. La Scola, S. Audic, P. Renesto, G. Blanc, C. Robert, P.-E. Fournier, J.-M. Claverie, and D. Raoult. 2006. Genome sequence of Rickettsia bellii illuminates the role of amoebae in gene exchanges between intracellular pathogens. PLoS Genet. 2:0734–0744. Parola, P., B. Davoust, and D. Raoult. 2005. Tick- and flea-borne rickettsial emerging zoonoses. Vet. Res. 36:469–492. Perlman, S. J., M. S. Hunter, and E. Zchori-Fein. 2006. The emerging diversity of Rickettsia. Proc. R. Soc. B 273:2097–2106. Perotti, M. A., H. K. Clarke, B. D. Turner, and H. R. Braig. 2006. Rickettsia as obligate and mycetomic bacteria. FASEB J. 20:2372–2382. Pickett, M. A., J. S. Everson, P. J. Pead, and I. N. Clarke. 2005. The plasmids of Chlamydia trachomatis and Chlamydophila pneumoniae (N16): accurate determination of copy number and the paradoxical effect of plasmid-curing agents. Microbiology 151:893–903. Pornwiroon, W., S. S. Pourciau, L. D. Foil, and K. R. Macaluso. 2006. Rickettsia felis from colonized cat fleas: isolation and culture in a tick-derived cell line. Appl. Environ. Microbiol. 72:5589–5595. Qin, A., A. M. Tucker, A. Hines, and D. O. Wood. 2004. Transposon mutagenesis of the obligate intracellular pathogen Rickettsia prowazekii. Appl. Environ. Microbiol. 70:2816–2822. Renesto, P., H. Ogata, S. Audic, J.-M. Claverie, and D. Raoult. 2005. Some lessons from Rickettsia genomics. FEMS Microbiol. Rev. 29:99–117. Revel, A. T., A. M. Talaat, and M. V. Norgard. 2002. DNA microarray analysis of differential gene expression in Borrelia burgdorferi, the Lyme disease spirochete. Proc. Natl. Acad. Sci. USA 99:1562–1567. Roux, V., and D. Raoult. 1993. Genotypic identification and phylogenetic analysis of the spotted fever group rickettsiae by pulsed-field gel electrophoresis. J. Bacteriol. 175:4895–4904. Rovery, C., P. Renesto, N. Crapoulet, K. Matsumoto, and P. Parola. 2005. Transcriptional response of Rickettsia conorii exposed to temperature variation and stress starvation. Res. Microbiol. 156:211–218. Samuel, J. E., M. E. Frazier, and L. P. Mallavia. 1985. Correlation of

VOL. 73, 2007

45.

46.

47. 48.

plasmid type and disease caused by Coxiella burnetii. Infect. Immun. 41:488– 493. Simser, J. A., A. T. Palmer, U. G. Munderloh, and T. J. Kurtti. 2001. Isolation of a spotted fever group rickettsia, Rickettsia peacockii, in a Rocky Mountain wood tick, Dermacentor andersoni, cell line. Appl. Environ. Microbiol. 67:546–552. Simser, J. A., A. T. Palmer, V. Fingerle, B. Wilske, T. J. Kurtti, and U. G. Munderloh. 2002. Rickettsia monacensis sp. nov., a spotted fever group rickettsia, from ticks (Ixodes ricinus) collected in a European city park. Appl. Environ. Microbiol. 68:4559–4566. Solow, B. T., and G. A. Somkuti. 2000. Comparison of low-molecular-weight heat stress proteins encoded on plasmids in different strains of Streptococcus thermophilus. Curr. Microbiol. 41:177–181. Stewart, P. E., R. Byram, D. Grimm, K. Tilly, and P. A. Rosa. 2005. The plasmids of Borrelia burgdorferi: essential genetic elements of a pathogen. Plasmid 53:1–13.


4995

49. Tucker, A. M., L. K. Pannell, and D. O. Wood. 2005. Dissecting the Rickettsia prowazekii genome: genetic and proteomic approaches. Ann. N. Y. Acad. Sci. 1063:35–46. 50. Wernegreen, J. J., and N. A. Moran. 2001. Vertical transmission of biosynthetic plasmids in aphid endosymbionts (Buchnera). J. Bacteriol. 183:785– 790. 51. Wernegreen, J. J. 2002. Genome evolution in bacterial endosymbionts of insects. Nat. Rev. Genet. 3:850–861. 52. Winkler, H. H., and R. M. Daugherty. 1984. Proline transport and metabolism in Rickettsia prowazekii. J. Bacteriol. 158:460–463. 53. Wood, D. O., and A. F. Azad. 2000. Genetic manipulation of rickettsiae: a preview. Infect. Immun. 68:6091–6093. 54. Zhang, J. Z., and M. Y. Fan. 2002. Determination of genome size and restriction fragment length polymorphism of four Chinese rickettsial isolates by pulsed-field gel electrophoresis. Acta Virol. 46:25–30.