Evaluation of Antibiotic Susceptibilities of Three Rickettsial Species Including Rickettsia felis by a Quantitative PCR DNA Assay

Abstract

Rickettsiae grow only intracellularly, and the antibiotic susceptibilities of these bacteria have been assessed by either plaque, dye uptake, or immunofluorescence assays, which are time-consuming. We used a quantitative PCR (with the LightCycler instrument) to assess the levels of inhibition of Rickettisa felis, R. conorii, and R. typhi DNA synthesis in the presence of various antibiotics. We established the kinetics of rickettsial DNA during growth and showed that R. conorii grows more quickly than R. typhi in cell culture, with maximum replication occurring after 5 and 7 days, respectively. The MICs of the antibiotics tested for R. conorii and R. typhi by the quantitative PCR assay were similar to those previously obtained by plaque and dye uptake assays. We found that R. felis is susceptible to doxycycline, rifampin, thiamphenicol, and fluoroquinolones but not to gentamicin, erythromycin, amoxicillin, or trimethoprim-sulfamethoxazole. The resistance of this new species to erythromycin is consistent with its current taxonomic position within the spotted fever group. We believe that quantitative PCR could be used in the future to simplify and shorten antibiotic susceptibility assays of other rickettsiae and other strict intracellular pathogens.

Rickettsiae are strict intracellular bacteria belonging to the alpha group of the class Proteobacteria (15). Rickettsioses are zoonoses that have geographical distributions similar to those of their vectors. The genus comprises the spotted fever group rickettsiae and the typhus group rickettsiae, which includes Rickettsia prowazekii, the agent of epidemic typhus, and R. typhi, the agent of murine typhus (16). Since they are obligate intracellular bacteria, in vitro studies of their susceptibilities to antibiotics necessitate the use of cell culture systems. Using the plaque assay and the dye uptake assay (18), we have now described the antibiotic susceptibilities of most of the rickettsiae with the exception of R. felis, which, until recently, was not available to us. The two assays depend on the induction of cytopathic effects and plaque formation in cell cultures by the rickettsiae, but some rickettsiae do not normally cause cytopathic effects in primary cultures (18). Recently, Ives et al. (5-7) described a new assay that uses immunofluorescence staining, which avoids the problem of a lack of cytopathic effects. A lack of cytopathic effects occurs with R. felis, the agent of flea-borne spotted fever, which we have now isolated and cultured in a Xenopus laevis cell line (XTC-2 cells) (14) and in Vero cells. To determine the antibiotic susceptibility of this species we developed an original antibiotic assay using PCR with the LightCycler instrument. It is a new, real-time PCR technique which has been used to diagnose several bacterial infections (3) but which has not been used to determine antibiotic susceptibilities. The technique combines PCR amplification and detection of products in a single optically clear glass capillary tube, which enables rapid temperature transitions with hot air and which reduces the PCR time to 45 min (22, 23). In this report we describe the PCR method with the LightCycler instrument that we developed for antibiotic susceptibility testing of Rickettsia isolates. Furthermore, we report on the antibiotic susceptibilities that we obtained for R. felis, R. conorii, and R. typhi and compare these susceptibilities with those obtained by the plaque assay, the reference method (18).

MATERIALS AND METHODS

Rickettsial isolates.

Four isolates of R. felis from fleas were used in the study: Marseille-URRWFXCal2 I-2363 (Cal2) (9), Pete2, Marseille, and Baton Rouge (14). They were cultured in XTC-2 cells with Leibowitz-15 medium (2 mM l-glutamine and amino acids in the l configuration; Gibco) containing 5% fetal calf serum and 2% tryptose phosphate (Gibco), as described previously (14). Also used were R. conorii (ATCC VR-141), the agent of Mediterranean spotted fever, and the R. typhi Wilmington strain (ATCC VR-144), the agent of murine typhus, which were grown in Vero cells in minimum essential medium at 32°C with 2 mM l-glutamine (Gibco) containing 4% fetal calf serum.

Antibiotics.

The antibiotics tested were amoxicillin (Beecham-Sevigne, Paris, France), gentamicin (Dakota Pharm, Creteil, France), ciprofloxacin (Bayer Pharma, Sebs, France), erythromycin (Abbott, Rungis, France), rifampin (Cassenne, Puteaux, France), doxycycline (Pfizer, Neuilly, France), telithromycin (Hoescht Marion Roussel, Romainville, France), levofloxacin (Hoescht Marion Roussel), ofloxacin (Diamant, Puteaux, France), trimethoprim-sulfamethoxazole (Roche, Paris, France), and thiamphenicol (Sanofi Winthrop, Gentilly, France). Stock solutions were prepared by dissolving all antibiotics except telithromycin in sterile distilled water; telithromycin was first dissolved in methanol before being diluted in sterile distilled water. Stock solutions were stored at −20°C in culture medium, and dilutions were made immediately prior to the tests.

The plaque assay for R. conorii and R. typhi was performed as described previously (18) and served as a control for the two other methods that we used, with the results obtained by the plaque assay compared with those obtained by the other two methods. For the plaque assay with R. felis, 24-well microtiter plates with XTC-2 cells were infected with R. felis and incubated at 28°C in Leibowitz-15 medium. Similarly, plates containing Vero cells were infected with R. conorii or R. typhi and incubated at 32°C with minimum essential medium (18). The antibiotics at each dilution were added to rows of infected cells; positive controls consisted of infected cells without antibiotics in the medium, and negative controls consisted of uninfected cells with antibiotics in the medium. From day 0 to the end of the experiment, three wells from each row were harvested each day and stored at −20°C in sterile tubes for the PCR assay. Also, cells from each row were stained with Gimenez stain, and the number of intracellular rickettsiae was counted at ×1,000 magnification. The MIC was defined as the lowest antibiotic concentration that completely inhibited the growth of the rickettsiae. Experiments were performed in duplicate to confirm the results.

Real-time PCR was performed with a LightCycler instrument (Roche Biochemicals, Mannheim, Germany), which is a combined fluorimeter and thermocycler that enables rapid (45 min) PCR thermocycling. Samples and the PCR master mixture are placed in 30-μl glass cuvettes, and sample detection is based on the principle of fluorescence resonance energy transfer (23), with adjacent hybridization probes directed against the intended PCR product. With fluorescein serving as the donor fluorophore and LC-Red 640 (Roche Biochemicals) serving as the acceptor fluorophore, the presence of PCR amplicons can be assessed by detection of the fluorescence of LC-Red 640. Samples can be assayed for the presence of this signal in real time at each PCR cycle, and the cycle number at which the signal is first detected can be correlated to the original concentration of DNA (22). The specificity of amplification can be confirmed by melting curve analysis. Single melting peaks can be generated by depicting the negative derivative of fluorescence versus temperature (−dF/dT) over the course of gradual melting of a PCR product.

Extraction of DNA.

After the harvested cells were thawed, they were centrifuged in a Beckman Avanti 30 centrifuge at 5,000 rpm for 10 min, the supernatant was discarded, and the pellet was washed twice with sterile distilled water before being resuspended in 200 μl of sterile distilled water. Extraction of the DNA was performed with 20% Chelex (biotechnology-grade chelating resin [Chelex 100; Bio-Rad, Richmond, Calif.]) in sterile water (20). Briefly, 500 μl of Chelex was added to each tube of centrifuged and washed cells before the tubes were vortexed and placed in a boiling water bath for 30 min. After centrifugation at 14,000 rpm for 10 min, the supernatants were harvested and stored in sterile tubes at 4°C before PCR was performed.

PCR master mixture.

Master mixtures were prepared by following the instructions of the manufacturer and with primers CS877F (5′-GGG GGC CTG CTC ACG GCG G-3′) and CS1258R (5′-ATT GCA AAA AGT ACA GTG AAC A-3′), whose sequences are specific for the sequence of the citrate synthase gene and which have been described previously (17). The 20-μl sample in each glass cuvette contained 2 μl of LightCycler DNA Master SYBR Green (Roche Biochemicals), 2.4 μl of MgCl₂ at 4 mM, 1 μl of each primer at 0.5 μM, 11.6 μl of sterile distilled water, and 2 μl of extracted DNA.

PCR cycling and melting curve conditions.

After one pulse centrifugation to allow mixing and to drive the mixture into the distal end of each cuvette, the tubes were placed in the LightCycler instrument. The amplification program began with a denaturation step of 95°C for 120 s, which was followed by 40 cycles of denaturation at 95°C for 15 s, annealing at 54°C for 8 s, and extension at 72°C for 15 s, with fluorescence acquisition at 54° in the single mode. Melting curve analysis was done at 45 to 90°C (temperature transition, 20°C/s), with stepwise fluorescence acquisition done by real-time measurement of the fluorescence directly in the clear glass capillary tubes. Sequence-specific standard curves were generated by using 10-fold serial dilutions (10⁵ to 10⁶ copies) of a standard concentration of Rickettsia. The number of copies of each sample transcript was then calculated from the standard curve with the LightCycler software. The MIC was defined as the lowest antibiotic concentration that caused inhibition of growth of the rickettsiae, i.e., that resulted in the number of DNA copies being similar to that on day 0. The experiments were carried out in duplicate.

RESULTS

The Gimenez staining assay was easy to perform with R. felis, but counting of the organisms was laborious and time-consuming, as 1,000 cells were examined for each sample. The results obtained with the reference strains (R. typhi and R. conorii) by the Gimenez staining assay were similar to those obtained by the plaque assay (Table 1). The PCR assay with the LightCycler instrument was rapid and reproducible and allowed the number of DNA copies to be counted precisely. Figures 1 and 2 present the results of the quantitative PCR assay. Figure 1 shows the melting curves obtained with standard concentrations of R. felis and R. conorii, showing the specificity of the PCR product (a single peak was obtained in all PCRs). Figure 2 shows the standard calibration curve obtained with 10-fold serial dilutions of R. felis. The standard curve was determined in each experiment to enable the results of all experiments to be correlated. In Fig. 3 we show the growth kinetics of R. felis, R. conorii, and R. typhi determined by the quantitative PCR assay. In Vero cells, the number of copies of R. conorii DNA increased more rapidly than the number of copies of R. typhi DNA (5 versus 7 days), but the two strains achieved a zenith of about 10⁵ DNA copies after 6 to 7 days. R. felis grown in XTC-2 cells had a maximum number of DNA copies after 5 days.

TABLE 1.

In vitro susceptibilities of rickettsial strains to antibiotics^a

Antibiotic	MIC (μg/ml)
	R. conorii			R. typhi			R. felisb
	Plaque assay	Gim assay	qPCR assay	Plaque assay	Gim assay	qPCR assay	Plaque assay	Gim assay	qPCR assay
Amoxicillin	128	256	256	128	128	256	ND	256	256
Ciprofloxacin	0.5	0.5	0.5	1	1	1	ND	0.5-1	0.5-1
Doxycycline	0.06	0.06	0.06	0.125	0.125	0.125	ND	0.06-0.125	0.06-0.125
Erythromycin	8	8	8	1	1	1	ND	16-32	16
Gentamicin	>32	>32	>32	>32	>32	>32	ND	>32	>32
Levofloxacin	0.5	0.5	0.5	0.5	1	1	ND	0.5-1	0.5-1
Ofloxacin	1	1	1	1	1	1	ND	0.5-1	0.5-1
Rifampin	0.25	0.25	0.06	0.25	0.5	0.25	ND	0.5-1	0.06-0.25
SXT	>16/4	>16/4	>16/4	>16/4	>16/4	>16/4	ND	>16/4	>16/4
Telithromycin	0.5	0.5	1	0.5	1	1	ND	1	1
Thiamphenicol	1	1	1	1	1	2	ND	2-4	1-2

^aThe MICs for the rickettsial strains were determined by Gimenez staining, plaque, and quantitative PCR assays. Abbreviations: Gim Gimenez staining; qPCR, quantitative PCR; SXT, trimethoprim-sulfamethoxazole; ND, not determined.

^bStrains Cal2, Marseille, Pete2, and Baton Rouge.

FIG. 1.

Melting curve obtained with standard concentrations of R. conorii by PCR with the LightCycler instrument. The specificities of the PCR products are shown by detection of a single peak.

FIG. 2.

Standard calibration curve obtained with 10-fold serial dilutions of R. felis (strain Baton Rouge). Quantification was performed by determining the threshold cycle (cycle number). The threshold is defined as the cycle at which the fluorescence exceeds 10 times the standard deviation of the mean baseline emission. One representative curve and its equation are given.

FIG. 3.

Kinetics of R. conorii (▪), R. typhi (•), and R. felis (▴) growth in cell culture determined by quantitative PCR assay.

In the antibiotic susceptibility assays, the results obtained by the Gimenez staining method and the real-time PCR assay were consistent with those obtained previously by the plaque assay, which is regarded as the “gold standard.” The MICs of the antibiotics for the four isolates of R. felis, R. conorii, and R. typhi determined by the different assays are shown in Table 1. The four isolates of R. felis had the same antibiotic susceptibility profiles, being resistant to amoxicillin, gentamicin, and trimethoprim-sulfamethoxazole. They had poor susceptibilities to erythromycin, with MICs ranging from 8 to 32 μg/ml; in contrast, R. typhi was highly sensitive (MIC, 1 μg/ml). Doxycycline was the most effective antibiotic against all isolates tested, with MICs ranging from 0.06 to 0.125 μg/ml. Fluoroquinolone compounds had MICs from 0.5 to 1 μg/ml, whereas the MICs of thiamphenicol were between 1 and 4 μg/ml. Rifampin was active, with MICs ranging from 0.06 to 1 μg/ml. The new ketolide compound, telithromycin, was bacteriostatic at concentrations from 0.5 to 1 μg/ml. In previous studies we have found that plaques caused by R. conorii in cell cultures are larger in the presence of trimethoprim-sulfamethoxazole (unpublished data), and in this study we found the number of DNA copies in cultures with or without trimethoprim-sulfamethoxazole to be almost identical (388,500 and 395,000 copies, respectively) at the time of maximum growth on day 7 of culture.

DISCUSSION

Rickettsiae are gram-negative bacteria that are primarily associated with arthropods that are the vectors of human infections. Since they are obligate intracellular bacteria, in vitro studies of their susceptibilities to antibiotics require the use of cell culture systems (6, 18). In our laboratory we have recently isolated and cultured R. felis, the agent of flea-borne spotted fever. We inoculate crushed cat fleas, in which the DNA of the ELB agent was previously detected, onto XTC-2 cells incubated at 28°C (14). With R. felis no cytopathic effects were observed in XTC-2 cell cultures, and only very small plaques could be seen in Vero cells, with these plaques observed only after 18 days of culture. As it was not possible to use conventional assays to determine the susceptibility of R. felis to antibiotics, we developed an assay using Gimenez staining to determine rickettsial numbers before and after addition of antibiotics to the culture medium. We also developed a new quantitative PCR assay using the LightCycler system (23) to determine the antibiotic susceptibilities of rickettsiae. The Gimenez staining method was too laborious and time-consuming to be a recommended technique. The quantitative PCR assay with the LightCycler instrument has already been shown to be an effective diagnostic tool (11) and has been used to detect antibiotic-resistant strains of Mycobacterium tuberculosis (21). Our experiments by PCR with the LightCycler instrument were, to the best of our knowledge, the first to determine the MICs of antibiotics by that technique.

The quantitative method that we describe enabled us to determine the antibiotic susceptibilities of fastidious Rickettsia spp. rapidly and consistently and may be useful for determination of the antibiotic susceptibilities of many other bacteria. The susceptibility results that we obtained for the reference strains, R. conorii and R. typhi, by the PCR assay were comparable to those obtained by the plaque assay, and for the first time, we were able to describe the kinetics of rickettsial DNA during the rickettsial growth phase. These were consistent with our knowledge of the in vitro culture characteristics of rickettsiae. In cell culture, R. conorii grows rapidly, the majority of cells are infected after 3 to 4 days, and subculturing is generally necessary after 5 days, when all cells have been destroyed. R. typhi, however, grows more slowly and can be maintained for 7 to 9 days before subculturing is necessary. We have observed similar growth kinetics for R. felis in XTC-2 cells, with almost all cells becoming infected after 5 days. The kinetics that we found in our PCR study were consistent with these in vitro culture characteristics of R. felis (14). Finally, the kinetics of growth of each bacterium enabled us to choose the best time to perform the antibiotic assay, i.e., at day 4 for R. conorii and R. felis and at day 6 for R. typhi.

We found that β-lactams, gentamicin, trimethoprim-sulfamethoxazole, and erythromycin are not active against R. felis. These results differ from those reported by Radulovic et al. (12), who found that “R. felis” was susceptible to erythromycin (MICs, 0.5 to 1 μg/ml). These were similar to the MICs for R. typhi (12), and it later appeared that the “R. felis” strain that they studied was in fact a strain of R. typhi (12). The most effective antibiotics in our study were doxycycline, fluoroquinolones, rifampin, thiamphenicol, and telithromycin. Our results also confirm those of previous experiments that showed that doxycycline, rifampin, and fluoroquinolones are the most effective antibiotics in vitro against all strains of rickettsiae (6, 18). The MICs found in our study and those reported previously and obtained by the plaque assay were lower than those obtained by Ives et al. (6) by the immunofluorescence assay, especially for fluoroquinolone compounds. The results obtained by the immunofluorescence assay are not consistent with clinical experience, suggesting that the method should be reconsidered. Patients suffering from Mediterranean spotted fever have been successfully treated with fluoroquinolones (4, 8, 13, 19), and thus, fluoroquinolones may be considered a safe alternative to tetracyclines for the treatment of rickettsial diseases. During pregnancy and childhood, macrolide compounds may represent a safe alternative to doxycycline and fluoroquinolones. There is, however, a high level of heterogeneity in the susceptibilities of spotted fever group rickettsiae to these compounds (18); josamycin is the only antibiotic whose efficacy against rickettsial diseases in this population has been tested (1), whereas erythromycin has not been shown to be effective (10). We have previously shown that it is possible to distinguish the typhus and spotted fever groups of rickettsiae by their susceptibilities to erythromycin (18). The relative resistance of R. felis to erythromycin is consistent with the fact that this rickettsia belongs to the spotted fever rather than the typhus group (2), the members of which are susceptible to erythromycin.

In summary, our report describes for the first time the susceptibilities of R. felis to a wide range of antibiotics determined by a new quantitative PCR assay. Our results show that doxycycline, fluoroquinolones, telithromycin, and rifampin are effective against the organism but that trimethoprim-sulfamethoxazole, β-lactams, and erythromycin are not. This new quantitative PCR assay should be developed and adapted to testing of the antibiotic susceptibilities of other strict intracellular pathogens.