Examination of Rickettsial Host Range for Shuttle Vectors Based on dnaA and parA Genes from the pRM Plasmid of Rickettsia monacensis

ABSTRACT The genus Rickettsia encompasses a diverse group of obligate intracellular bacteria that are highly virulent disease agents of mankind as well as symbionts of arthropods. Native plasmids of Rickettsia amblyommatis (AaR/SC) have been used as models to construct shuttle vectors for genetic manipulation of several Rickettsia species. Here, we report on the isolation of the complete plasmid (pRM658B) from Rickettsia monacensis IrR/Munich mutant Rmona658B and the construction of shuttle vectors based on pRM. To identify regions essential for replication, we made vectors containing the dnaA and parA genes of pRM with various portions of the region surrounding these genes and a selection reporter cassette conferring resistance to spectinomycin and expression of green fluorescent protein. Rickettsia amblyommatis (AaR/SC), R. monacensis (IrR/Munich), Rickettsia bellii (RML 369-C), Rickettsia parkeri (Tate’s Hell), and Rickettsia montanensis (M5/6) were successfully transformed with shuttle vectors containing pRM parA and dnaA. PCR assays targeting pRM regions not included in the vectors revealed that native pRM was retained in R. monacensis transformants. Determination of native pRM copy number using a plasmid-carried gene (RM_p5) in comparison to chromosomally carried gltA indicated reduced copy numbers in R. monacensis transformants. In transformed R. monacensis strains, native pRM and shuttle vectors with homologous parA and dnaA formed native plasmid-shuttle vector complexes. These studies provide insight on the maintenance of plasmids and shuttle vectors in rickettsiae. IMPORTANCE Rickettsia spp. are found in a diverse array of organisms, from ticks, mites, and fleas to leeches and insects. Many are not pathogenic, but others, such as Rickettsia rickettsii and Rickettsia prowazeckii, can cause severe illness or death. Plasmids are found in a large percentage of nonpathogenic rickettsiae, but not in species that cause severe disease. Studying these plasmids can reveal their role in the biology of these bacteria, as well as the molecular mechanism whereby they are maintained and replicate in rickettsiae. Here, we describe a new series of shuttle plasmids for the transformation of rickettsiae based on parA and dnaA sequences of plasmid pRM from Rickettsia monacensis. These shuttle vectors support transformation of diverse rickettsiae, including the native host of pRM, and are useful for investigating genetic determinants that govern rickettsial virulence or their ability to function as symbionts.


RESULTS
Identification of minimal coding sequences required for pRM shuttle vector replication and partitioning. Because coding sequences that support plasmid replication and partitioning are often clustered together, we identified the minimal region required for rickettsial shuttle vector replication and partitioning by constructing three shuttle vectors with various portions of RM_p16-21 coding sequences: pRMdSGK clone 1 (pRMD1), pRMdSGK clone 2 (pRMD2), and pRMdSGK clone 3 (pRMD3) (Fig. 1). All three clones contained coding sequence for the pRM DnaA-like protein (RM_p16) and ParA (RM_p18), which function in DNA replication and chromosomal stability, as well as RM_p17 and RM_p19, hypothetical proteins (HPs) of unknown function, although RM_p19 contains domains with similarity to those of the HTH_XRE family transcriptional regulators. The smallest test construct, pRMD1, included the gene cluster RM_p16 through RM_p19 and partial coding sequence (359 bp of 507 bp) for RM_p20 (green bar in Fig. 1C), a second likely HTH_XRE family protein with 74% similarity to RM_p19. The pRMD2 construct extended the same dnaA/parA region through an intact RM_p20 coding sequence and an amino-terminal fragment of RM_p21 (orange bar in Fig. 1C). The pRMD3 construct further encoded an intact RM_p21, a likely Sca12 cell surface antigen, and approximately 1 kbp of downstream noncoding sequence (lilac bar in Fig. 1C).
Transformation trials of five SFG Rickettsia spp. The smallest test construct, pRMD1, transformed only R. monacensis (Table 1), which carries pRM as its native plasmid. In contrast, in trials with pRMD2 (encoding the same proteins from RM_p16 through RM_p19 as well as the intact RM_p20 XRE family protein with its upstream region) and pRMD3 (extended to encode the RM_p21 Sca12 cell surface antigen), both constructs transformed all five Rickettsia species tested (Table 1). Species successfully transformed included four spotted fever group (SFG) rickettsiae, notably R. amblyommatis AaR/SC, which contains 3 native plasmids, as well as R. bellii, which occupies a more basal phylogenetic position (24). The cluster from RM_p16 through RM_p19 thus supported transformation of the parental R. monacensis, but inclusion of the RM_p20 locus and/or a short upstream sequence extended the range of a pRM-based shuttle vector to other SFG and ancestral group rickettsiae. In contrast to RM_p20, RM_p21 and the associated noncoding 1-kbp sequence conferred no apparent advantage.
These results indicated that the loci from RM_p16 through RM_p20 of pRM contained the minimum necessary DNA sequence and protein coding capacities for a shuttle vector capable of transforming a wide range of SFG and ancestral group rickettsiae. The promoter prediction program BPROM (25) indicated a single promoter upstream of the RM_p20 locus, but none for RM_p17, -18, or -19, consistent with functional dependence of RM_p18 expression on sequences upstream of RM_p20 and a likely operon extending through the RM_p16 locus (Fig. 1D). The predicted transcription start site was at bp 19395 of pRM, with a 210 box from 19402 to 19410 and a 235 box from 19425 to 19430. A comparison of growth rates of wild-type (WT) R. monacensis and pRMD2-transformed R. monacensis using three separate growth curve analyses indicated no significant differences between the growth rates of WT and transformants (doubling times of 17.5 and 18 h, respectively).
Evaluation of GFP expression in wild-type and transformed R. monacensis using confocal microscopy. The cell-free R. monacensis WT strain and transformants were stained with NucBlue Live Cell Stain ReadyProbes (Thermo Fisher Scientific, Waltham, MA) to visualize all rickettsiae, whether transformed or not, then gently mounted onto slides by using a Cytospin centrifuge (Thermo Fisher) and observed by confocal microscopy. The left column of Fig. 2A with a DAPI (49,6-diamidino-2-phenylindole) filter shows all the rickettsiae present, while the middle column shows only gfp uv -expressing transformed rickettsiae (fluorescein isothiocyanate [FITC] filter). The third column (Fig. 2B) illustrates the overlap of fluorescence in the DAPI and FITC fields. Visually, the degree of colocalization indicated that nearly all the rickettsiae present were transformed. To confirm these observations, we used Pearson's coefficient (PCC) and Manders' colocalization coefficient (MCC) (26,27) to evaluate the colocalization of the DAPI and FITC fluorescence. For the PCC analysis, the correlation coefficient was measured on all pixels in an individual , the R. monacensis plasmid pRM with pMOD658 transposon cloned into the linear plasmid pJAZZ. The two arms formed by digested pJAZZ are indicated by brown double-ended arrows, and pRM is represented by a red double-ended arrow; genes present on pRM are denoted with black arrows. The pMOD658 transposon is shown in green; resistance to chloramphenicol (CAT) and the ability of the clone to express green fluorescent protein (GFP uv ) are conferred by the transposon inserted in pRM. The unique SmaI site with which pRM was linearized for cloning is located in the transposon. Pink balloons indicate the restriction enzyme sites used for subcloning the dnaA/parA region of pRM. (B) The 5.6-kbp selection reporter cassette SGK was ligated with three pRM fragments of various sizes containing dnaA and parA, yielding the R. monacensis pRM shuttle vectors pRMD1, -2, and -3 (C). The numbered black line in panel C represents bp 14400 to 21600 of the 23,486-bp R. monacensis plasmid pRM, and the thick black horizontal arrows beneath it indicate predicted genes and their orientations. Colored vertical arrows indicate the restriction enzyme sites used during pRM subcloning, and colored bars above the numbered black line represent the three fragments of pRM contained in the finished shuttle vectors shown in panel D.  (Fig. 2C).
Presence of shuttle vector in pRM-transformed R. amblyommatis and R. parkeri and conservation of native plasmids in transformed R. amblyommatis. Undigested DNA from R. amblyommatis and Rickettsia parkeri transformed with pRMD2 and pRMD3 was separated by pulsed-field gel electrophoresis (PFGE) (Fig. 3A) and transferred onto Zeta-Probe membranes. Because R. parkeri does not contain native plasmids, hybridization of R. parkeri pRMD2 and pRMD3 transformants with digoxigenin-labeled gfp uv probe (28) identified the unaltered shuttle vectors, with asterisks indicating the nicked linear forms of pRMD2 and pRMD3 at 10.81 and 12.814 kbp, respectively (Fig. 3B). The R. amblyommatis transformants show a band pattern similar to that of R. parkeri; however, extra bands were present in both R. amblyommatis pRMD2 and pRMD3 hybridized with the gfp uv probe (Fig. 3B). Stripped Southern blots were hybridized with recombinase, hsp2, and helicase digoxigenin-labeled probes specific for R. amblyommatis strain AaR/SC plasmids pRAM18, pRAM23, and pRAM32, respectively (12). Bands of the appropriate size were observed for all 3 native pRAM plasmids (pRAM18, 18.344 kbp; pRAM23, 22.852 kbp; and pRAM32, 31.972 kbp) in the transformed R. amblyommatis (marked with asterisks in Fig. 3A, C, D, and E), but not in R. parkeri. Thus, pRM shuttle vector and all three native R. amblyommatis plasmids are conserved in the pRM-transformed R. amblyommatis.
Testing for the presence of native pRM and shuttle vectors in pRMD-transformed R. monacensis. To assess native pRM in shuttle vector transformants, genomic DNA from WT and transformant R. monacensis strains was PCR amplified with native pRMspecific primer sets ( Table 2). All three R. monacensis transformants contained native pRM, as indicated by the presence of amplicons from genes present in native pRM but not in the shuttle vectors (Fig. 4A, B, and C). The dGFPuvF2/R2 primer amplicons confirmed the presence of the shuttle vectors (Fig. 4D). R. amblyommatis AaR/SC pRMD2 DNA was used as a control to confirm primer specificity; as expected, it yielded no amplicon with native pRM primer sets but was positive with dGFPuvF2/R2 primers.
Copy number ratios of native pRM and pRM shuttle vectors in transformed R. monacensis. Quantitative PCR (qPCR) estimation of the relative ratio of the single-copy pRM-encoded resolvase RM_p5 (qRM_p5F/qRM_p5R primer set in Table 2) and chromosome-carried gltA genes (29) indicated 1.5 copies of pRM for each chromosome copy in WT R. monacensis, which decreased to a ratio of 0.5 to 0.8 in pRMD1-, -2-, and -3-transformed R. monacensis (Fig. 5). This ratio was mirrored by the 0.5-to-0.9 ratio of The native plasmid name is given in parentheses, along with the approximate native plasmid size(s). b The construct designation and size are shown, with the pRM genes included in parentheses. An italicized gene number indicates the sequence does not represent the complete gene. 1, transformed; 2, not transformed.
pRM Rickettsial Host Range Applied and Environmental Microbiology shuttle vector (qGFPuvF/qGFPuvR primers from shuttle vector-carried gfp uv in Table 2) to chromosome in the transformant rickettsiae (Fig. 5). There were no further significant changes in copy number ratios during serial passage of the transformant rickettsia (up to 20 passages in the case of pRMD2) (data not shown).
Presence of shuttle vector and native plasmid complexes in pRM-transformed R. monacensis. To confirm the PCR-indicated presence of native pRM in pRMD shuttle vector-transformed R. monacensis, undigested genomic DNA from WT and pRMD1, -2, pRM Rickettsial Host Range Applied and Environmental Microbiology and -3 transformants was separated by pulsed-field gel electrophoresis ( Fig. 6) and transferred onto Zeta-Probe membranes. Nicked linear, circular, and multimeric forms of pRM were present in the WT R. monacensis lanes ( Fig. 6A and C, bands with white asterisks). The presence of the shuttle vector in R. monacensis transformed with pRMD1, pRMD2, and pRMD3 was demonstrated by hybridization with digoxigeninlabeled gfp uv probe (absent in the WT) (Fig. 6B). The gfp uv probe localized to bands in the ;40-kbp region (arrowhead) rather than at 10 to 12 kbp (the size of the shuttle vectors), indicating that there are no detectable monomeric pRMD1, pRMD2, or pRMD3 in transformant populations.  The PFGE gel shown in Fig. 6C was transferred to a membrane and hybridized with a probe containing digoxigenin-labeled RM_p23 (13), a gene present on native pRM but absent on the shuttle vectors (Fig. 6D). In the WT lane, the probe hybridized to the 20-to 30-kbp linear nicked and multimeric forms (black asterisks) of pRM. In the transformant lanes, the RM_p23 probe hybridized in the ;40-kbp region (Fig. 6D, arrowhead) at the same relative positions as the gfp uv probe (Fig. 6B). The strong hybridization of the RM_p23 probe at the 40-kbp position in transformant rickettsia lanes versus weak hybridization at the lowest position (lowest black asterisk in panel D) indicates the likely presence of native pRM that is predominantly complexed with shuttle vector, as transformants lacking pRM would not hybridize with RM_p23. FIG 5 Plasmid-to-chromosome copy number ratios in transformed R. monacensis. Native plasmid (pRM), shuttle vector, and chromosomal copy numbers were quantified by qPCR of RM_p5, gfp uv and gltA, respectively. Passage numbers (e.g., p17) used for this assay were as follows: WT, p17 (all 3 replications); pRMD2, p10 (all 3 replications); pRMD1, one replication with p3 and two replications with p9; pRMD3, one replication with p3 and two replications with p7.

DISCUSSION
Although plasmids have been extensively studied in other bacteria, the discovery of plasmids in rickettsiae is fairly recent (1,2,13), and their biological function is largely unexplored. Genomic analysis has shown that rickettsiae have undergone reductive evolution (30)(31)(32)(33), resulting in such characteristics as AT enrichment, high conservation of genome sequences among species, higher levels of virulence, and variable presence and numbers of plasmids (3). Rickettsial plasmids have likewise undergone reductive evolution and mirror rickettsial genomes in relative size and GC content (34). It has been proposed that a rickettsial ancestor supported a plasmid system that was lost in some species due to their unique obligate intracellular life cycle (3,34), but other species presumably retained plasmids due to an advantage conferred by their presence. Plasmids are known agents of horizontal gene transfer, facilitating host adaptation/virulence, antibiotic and stress resistance, and genetic plasticity (1,(34)(35)(36). Although the origins and functions of many rickettsial plasmid gene sequences have been inferred from similarities to those of other bacteria, the role of plasmids and their interactions in rickettsiae remain to be elucidated. Real-time PCR and whole-genome sequencing showed that these are low-copy-number plasmids (2), while creation of shuttle vectors from pRAM18 and pRAM32 and their subsequent transformation into rickettsiae confirmed that parA and dnaA were essential for plasmid replication and maintenance (12). However, the mechanism by which these genes function is yet to be identified.
We have developed two new Rickettsia plasmid shuttle vectors, pRMD2 and pRMD3, that can be used to transform plasmid-free Rickettsia spp. as well as those carrying native plasmids. In conjunction with pRMD1, they collectively contained various regions of pRM surrounding the dnaA and parA genes. Analysis of their relative efficacies in transformed rickettsiae allowed us to identify plasmid sequences important for the replication and maintenance of the shuttle vectors in rickettsiae. Our data indicated that RM_p20 or its immediate upstream sequence was required for shuttle vector replication in rickettsiae. Specifically, R. parkeri, R. bellii, Rickettsia montanensis (plasmid free), and R. amblyommatis strain AaR/SC (carrying 3 plasmids) were not transformed with shuttle vector pRMD1 (Fig. 1C, containing RM_p16 through RM_p19 and the 39 end of RM_p20) but were transformed with shuttle vectors pRMD2 and D3 (containing RM_p16 through RM_p20 and pRM21, respectively). These data support the prediction (37) that the region of pRM containing RM_p17 to RM_p20 forms an operon. It is possible that RM_p19 and -20 are not themselves required because absence of a single promoter upstream of RM_p20, as predicted by BPROM (25), would prevent expression of RM_p18 (parA) in pRMD1. Without expression of the ParA chromosomal stability protein, the shuttle vector would not be maintained, consistent with the observed phenotype. In contrast, R. monacensis was transformed with pRMD1 when the other species were not, likely due to the presence of native pRM providing the necessary ParA for maintenance of the shuttle vector.
Partition systems usually include three features: a site that acts like a centromere, a centromere-binding protein (CBP) (usually encoded by parB), and an NTPase (parA) (38). As none of the genes in the pRM operon have similarity to known parB genes, the pRM partitioning system could work in one of several ways. The partitioning system could rely on the chromosomal parB to work or represent a new type of system, as in R388, which functions with a single protein and a centromere site (38). Alternatively, the pRM hypothetical genes could act in the capacity of parB as CPBs are not necessarily significantly similar in sequence but are typically dimers of helix-turn-helix (HTH) or ribbon helix-helix DNA-binding proteins (38). Both RM_p19 and -20 have HTH_XRE domains and might function as CPBs.
The pRM operon containing parA appears to be unique among known rickettsial plasmids. A BLASTN search of the genes RM_p17 and -20 showed no rickettsial homology, while RM_p19 only had homology with the rickettsial endosymbiont of Ixodes pacificus plasmid (pREIP). Interestingly, translation of the nucleotide sequence for RM_p19 and -20 and subsequent BLASTP of the amino acid sequence shows a low level of similarity to pREIP (36% identity and 59% positive for RM_p19; 40% identity and 60% positive for RM_p20) and to rickettsial endosymbionts of a variety of arthropods, including beetles such as Platyusa sonomae and Bembidion nr Transversale, bedbugs (Cimex lectularis), and midges (Culicoides impunctatus). These Torix clade Rickettsia spp. (39) have 2 or 3 different regions of similarity for both RM_p19 and -20 and have identities ranging from 30 to 40%, with 53% to 62% positive.
The ParA protein encoded on R. monacensis pRM is sufficiently distinct from those of pRAM18, -23, and -32 (the native plasmids of R. amblyommatis) that it should support uptake and maintenance of the pRM shuttle vector in R. amblyommatis. In their comparison of rickettsial plasmid genes, El Karkouri et al. (3) showed that pRM parA was distinct from the parA of all other sequenced rickettsial plasmids. A BLASTP search of the ParA protein from pRM identified only one rickettsial homolog (R. asembonensis, with 46% identity), while its closest match was from a Mycoplasmataceae bacterium, with 53% identity. Lack of interactions between nonhomologous ParA proteins may explain why R. amblyommatis strain AaR/SC was transformed with pRM shuttle vectors, despite its three native plasmids and their potential for causing plasmid incompatibility. On the other hand, the successful transformation of R. monacensis with its own pRM shuttle vectors was unexpected. It is possible that the presence of identical ParAs from native plasmids and shuttle vectors supports mutual plasmid maintenance in rickettsiae rather than promoting incompatibility. These results also highlight the fact that plasmid transformation of different Rickettsia species is by no means a routine activity with universally predictable results. For example, the construct pRAM18/Rif/GFPuv, containing full-length pRAM18, successfully transformed R. bellii and R. parkeri (12), but for unknown reasons, shuttle vectors that incorporated full-length pRM instead of pRAM18 (data not shown) were unable to transform R. montanensis, R. monacensis, R. peacockii, and R. parkeri or 3 strains of R. amblyommatis. Thus, there is still much to learn about the role of these rickettsial plasmids in the functioning of rickettsiae and the mechanisms by which they operate.
The studies reported here give us a better understanding of the mechanisms of rickettsial plasmid maintenance in diverse rickettsial species. We explored the basis for the ability of pRM-based shuttle vectors to transform R. monacensis. Our PCR results detected the presence of shuttle vector in transformants, which continued to persist through serial transferring. They also indicated a decrease in the ratio of pRM to the chromosome, suggesting that transformants harbored an average of two or more chromosome copies per cell or that plasmids were cured from some of the rickettsiae. Nevertheless, data from Fig. 2 suggest that almost 100% of the rickettsiae contained the shuttle vector. Furthermore, the results showed that the use of homologous rickettsial parA regions leads to the formation and maintenance of complexes between shuttle vectors and native plasmids, suggesting possible defects in partitioning of plasmids carrying the same parA genes. The surprising ability of R. monacensis to be transformed by shuttle vectors containing the parA gene from its native plasmid emphasizes the need for further study of plasmid maintenance and incompatibility in rickettsiae, and further studies are needed to elucidate the molecular basis for the apparent linkage of shuttle vectors with pRM present in R. monacensis.
The R. monacensis plasmid pRM was cloned in E. coli as previously described (37). Briefly, electroporation of R. monacensis with the pMOD658 transposon yielded the Rmona658B transformant, in which the pMOD658 transposon encoding chloramphenicol acetyltransferase (CAT) and carrying a gfp uv fluorescent marker was inserted into pRM (37). The mutated pRM was cloned in its entirety by chloramphenicol marker rescue of Big Easy TSA E. coli cells (Lucigen, Middleton, WI) electroporated with Rmona658B genomic DNA that had been linearized by digestion with SmaI and ligated into the linear vector pJAZZ (Lucigen). The resulting construct was termed pJAZZ[pRM658B] (Fig. 1A).
The following abbreviations are used to designate plasmids and their derivatives: (i) "S" indicates spectinomycin/streptomycin resistance conferred by the aminoglycoside adenyltransferase gene, aadA, driven by a rickettsial ompA promoter, (ii) "G" indicates green fluorescent protein (GFP) encoded by gfp uv under regulation of the rickettsial ompA gene promoter, and (iii) "K" indicates kanamycin resistance from the pET-28a(1) vector (Novagen, EMD Millipore, Bedford, MA), adapted as described below. Specific nucleotide spans from pRM are from accession no. EF564599 (37). Multiple cloning sites (MCSs) are designated "[MCS]." Promoters are designated "p." Construction of pRM-based shuttle vectors. (i) Preparation of selection reporter cassette. We constructed a cassette into which pRM fragments could be cloned. The SGK selection reporter cassette (see Fig. S1C in the supplemental material) was derived from the previously constructed shuttle vector pRAM18dRGA[MCS] (12) (Fig. S1A) and contained genes needed for replication and antibiotic selection in E. coli as well as reporter and antibiotic selection genes for use in rickettsiae (Fig. S1C). Replacement of pGEM with a 3.159-kbp DraIII/PshA fragment of the pET-28a(1) vector was prompted by previous experiments indicating that some genes cloned into the pRAM18 shuttle vector MCS were less stable in pGEM than pET. Because spectinomycin and streptomycin are water soluble and not used to treat rickettsioses, but rifampin may be, we replaced the rpsLp-arr-2 Rp (RIF)/ompAp-gfp uv cassette with an ompAp-aadA/ompAp-gfp uv cassette (see Fig. S2B in the supplemental material).
(ii) Subcloning pRM. To identify which region(s) of pRM yielded the most effective shuttle vector for rickettsial transformation, we cloned specific fragments of pRM to create a family of deletion constructs: pRMD1, pRMD2, and pRMD3. To construct pRMD1 and pRMD3, pJAZZ[pRM658B] was digested with BamHI/PacI or BamHI/PflMI (Fig. 1A and B), and the desired 4,671-bp and 7,177-bp fragments containing either RM_p16 to -20 (pRM bp 14561 to 19231) or RM_p16 through -21 (pRM bp 14561 to 21646) were gel purified (Zymoclean Gel DNA Recovery kit; Zymo Research, Irvine, CA). The fragment ends were blunted (DNATerminator) and ligated into the blunt and dephosphorylated selection reporter cassette (SGK), creating the 10,368-bp pRMD1 and 12,797-bp pRMD3, respectively (Fig. 1D). To construct pRMD2, Q5 DNA polymerase (New England Biolabs) was used to PCR amplify RM_p16 through RM-p20 and into the 59 end of RM_p21 (pRM bp 14512 through 19585) with primers RM_p16 FOR NheI/RM_p21Rev NheI (59-TAT TGC TAG CCG TAA GGA ACA GTT GGT GAG-39 and 59-ATA TGC TAG CGT TAA TAT GCC TCG GGC  TAC -39). The PCR product with NheI sites at both ends (Fig. 1C) was incubated with Taq polymerase to create A9 overhangs and cloned into pCR4 with the TOPO TA Cloning kit (Invitrogen, Carlsbad, CA). Clones were sequenced to verify that they contained the correct pRM fragment, which was then recovered by restriction digest with NheI and ligated it into dephosphorylated, NheI-digested SGK, to yield the 10,881-bp shuttle vector pRMD2 (Fig. 1D).
Preparation of shuttle vector plasmid DNA. Endotoxin-free maxipreps (Qiagen, Valencia, CA) were prepared for all pRMD shuttle vectors as per the manufacturer's recommendations, for use in transforming rickettsiae. The integrity and orientation of inserted genes were reconfirmed by sequencing fragment junctions and by restriction digest analysis. (See Table 3 for the sequencing primers.) Transformation of rickettsiae. Rickettsiae were purified and electroporated as described previously (16). Rickettsial transformants were selected using growth medium containing spectinomycin and streptomycin, each at a final concentration of 100 mg/mL. Cultures were monitored for expression of the GFP uv reporter on a weekly basis by using an inverted microscope (Nikon Diaphot TMD with Y-FL-epifluorescence attachment and a sapphire GFP 31043 filter) or by examining wet mounts on an upright Nikon Eclipse E400 microscope with a B-2E/C FITC filter (Nikon, Melville, NY).
Growth rate analysis of R. monacensis transformed with pRMD2. Evaluation of GFP expression in WT and transformed R. monacensis using confocal microscopy. Cell-free WT and transformed R. monacensis cells were resuspended in complete medium and incubated with NucBlue Live Cell Stain ReadyProbes reagent (1: 50 dilution) (Thermo Fisher Scientific) in the dark for 30 min at room temperature. Fifty-microliter aliquots of cell-free R. monacensis were deposited onto microscope slides (Cytospin centrifuge; Thermo Fisher) at 200 rpm for 3 min. The slides were mounted with 3 mL 1Â phosphate-buffered saline (PBS) and imaged on an Olympus BX61 DSU confocal microscope with a 60Â objective via a double-wavelength filter (DAPI, excitation at 365 nm and emission at 480 nm; FITC, excitation at 495 nm and emission at 519 nm). Colocalization of fluorescence from gfp uv (transformed R. monacensis) and NucBlue (rickettsial DNA) was analyzed by determining signal overlap for each of three random fields of view using Image Fiji (with the JaCoP plugin and Co-localization Threshold plugin), Pearson's coefficient (PCC), and calculation of Manders' colocalization coefficients (MCCs) (26,27).
PCR and sequence confirmation of rickettsial species and copy number estimates of native pRM and pRM shuttle vectors in transformants. To confirm species identities, portions of the ompA gene of spotted fever group rickettsiae were amplified using primers 190-70/190-602 (41) and rickettsial genomic DNA as the template (42). PCR products were purified with the DNA Clean and Concentrator kit (Zymo Research), as per the manufacturer's protocol, for Sanger sequencing on an ABI 3730 Excel automated sequencer (University of Minnesota Genomics Center). The sequences were compared to rickettsial ompA sequences in GenBank by using BLASTN (NCBI, NIH).
Three sets of specific primer pairs (Table 2) were designed for PCR amplification of native pRM, while the dGFPuvF2/R2 primer pair ( Table 2) was used to amplify a product specific to pRM shuttle vectors in rickettsiae. PCRs used 100 ng of template, 1 mM mixed deoxynucleoside triphosphates (dNTPs), 0.5 mM each primer, and 1.25 U of GoTaq polymerase in 50 mL final 1Â PCR buffer (Promega, Madison, WI) and were run in a Robocycler (Stratagene, La Jolla, CA) as follows: 1 cycle at 95°C for 3 min, followed by 40 cycles at 95°C for 15 s, 51°C for 30 s, and 72°C for 1 min, and then one final cycle at 72°C for 7 min. Real-time quantitative PCR (qPCR) was used to estimate relative copy number ratios (43,44) of single-copy genes specific to the R. monacensis chromosome (gltA encoding citrate synthase), the native pRM plasmid (RM_p5 locus for transposon resolvase), and the pRM-derived shuttle vectors (GFP uv ) by using three primer pairs (Table 2). With the exception of a 56°C annealing temperature, reactions and copy number estimates were executed as previously described (2).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.2 MB.