OUP user menu

Widespread use of real-time PCR for rickettsial diagnosis

Aurélie Renvoisé, Jean-Marc Rolain, Cristina Socolovschi, Didier Raoult
DOI: http://dx.doi.org/10.1111/j.1574-695X.2011.00899.x 126-129 First published online: 1 February 2012

Abstract

We report 2 years of experience with rickettsial molecular diagnosis using real-time PCR at the French National Reference Center. All Rickettsia genomes available were compared to discover specific sequences to design new sets of primers and probes. The specificity was verified in silico and against a panel of 30 rickettsial species. Sensitivity was determined using 10-fold serial dilutions. Finally, primers and probes that were both specific and sensitive were routinely used for the diagnosis of rickettsial infections from clinical specimens. We retained sets of primers and probes to detect spotted fever group Rickettsia, typhus group Rickettsia, Rickettsia conorii, Rickettsia slovaca, Rickettsia africae and Rickettsia australis; 643 clinical samples were screened for the presence of Rickettsia DNA. Overall, 45 positive samples were detected, including 15 Rickettsia africae, nine R.conorii, five Rickettsia sibirica mongolitimonae, four R.slovaca, two R.australis, four Rickettsia massiliae, one Rickettsia honei, one Rickettsia typhi and eight Rickettsia sp. Positive samples were detected mainly from cutaneous biopsies and swabs (31/45). Widespread use of real-time PCR is inexpensive and reduces delay in the diagnosis of rickettsial infections. These real-time PCR assays could be implemented easily in laboratories that have molecular facilities and may be added to existing molecular tools as a point-of-care strategy.

Keywords
  • Rickettsia
  • real-time PCR
  • molecular diagnosis
  • genome
  • point-of-care strategy

Members of the genus Rickettsia may be classified into spotted fever group (SFG) Rickettsia, typhus group (TG) Rickettsia, Rickettsia bellii group and Rickettsia canadensis group (Merhej & Raoult, 2011). Rickettsiae can be transmitted to humans by blood-sucking arthropods and are associated with specific diseases termed rickettsioses. For example, Rickettsia conorii is associated with Mediterranean spotted fever (MSF) (Parola et al., 2005), Rickettsia africae with African tick-bite fever (ATBF) (Jensenius et al., 2003), Rickettsia sibirica ssp. mongolitimonae with lymphangitis-associated rickettsiosis (LAR) (Fournier et al., 2005), Rickettsia slovaca with ‘scalp eschar and neck lymphadenopathy after tick bite’ (SENLAT) (Angelakis et al., 2010), Rickettsia australis with Queensland tick typhus (QTT) (Parola et al., 2005), Rickettsia typhi with murine typhus (Civen & Ngo, 2008) and Rickettsia honei with Flinders Island spotted fever (FISF) (Parola et al., 2005). When a rickettsiosis is clinically suspected, biological diagnosis can be obtained using serology, cell culture and/or molecular tools (Parola et al., 2005); among the molecular tools, real-time quantitative PCR (qPCR) is rapid and sensitive (Stenos et al., 2005; Henry et al., 2007; Kidd et al., 2008). Genomic approaches have recently increased our knowledge of Rickettsia sp., and massive amounts of genomic data have become available (Ogata et al., 2001; Fournier et al., 2007; Merhej & Raoult, 2011). We used these sequence data to develop specific qPCR methods to improve the diagnosis of rickettsial infections. Here, we report 2 years of experience with rickettsial molecular diagnosis using qPCR at the French National Reference Center (FNRC).

All rickettsial genomes available were compared to discover sequences that are specific for either SFG or TG or for the identification of Rickettsia spp. at the species level. Specific primers and probes which were selected by genome comparison were designed based on these specific sequences (Supporting information, Table S1). Specificity was verified in silico using blastN analysis on GenBank database. Specificity was also verified in vitro using a local collection panel of 30 rickettsial strains. Sensitivity was determined using 10-fold serial dilutions. Finally, primers and probes that were both specific and sensitive were routinely used for the diagnosis of rickettsial infections from clinical specimens.

As an FNRC for rickettsioses, we routinely receive clinical samples from patients with suspected rickettsiosis. These samples are obtained from both locally hospitalized patients and from outpatients throughout France and the rest of the world. Total DNA was extracted from the samples using a QIAmp DNA Mini kit (Qiagen, Hilden, Germany) as described in the manufacturer's instructions. Master mixtures were prepared with a QuantiTect Probe PCR kit (Qiagen) following the manufacturer's instructions. Sterile human biopsies were used as negative controls; DNA extracted from the cell culture supernatant of Rickettsia montanensis served as a positive control when using the primer and probe set targeting SFG Rickettsia; DNA extracted from the cell-culture supernatant of each Rickettsia species served as a positive control for the corresponding primer and probe set. Appropriate handling and DNA extraction were controlled using qPCR targeting the gene encoding β-actin (Socolovsch et al., 2010). qPCR assays were performed in a LightCycler 3.5 instrument (Roche Diagnostics, Mannheim, Germany). The PCR mixture included a final volume of 20 µL with 10 µL of the Master mixture, 0.5 µL (20 pmol µL−1) of each primer, 2 µL (2 pmol µL−1) of probe, 2 µL of distilled water and 5 µL of extracted DNA. The amplification conditions were as follows: an initial denaturation step at 95 °C for 15 min, followed by 40 cycles of denaturation at 95 °C, annealing and elongation at 60 °C for 60 s, with fluorescence acquisition in single mode.

The first molecular screening was systematically performed with a set of primers and a probe targeting SFG Rickettsia; if clinically and epidemiologically suspected a screening was performed to target TG Rickettsia. Based on clinical and epidemiological investigations and on serological results, if first screening was positive, a second directed step of molecular screening was performed to target Rickettsia spp. at the species level using various sets of primers and probes. When diagnosis at the species level could not be obtained using specific qPCR, conventional PCR followed by sequencing of the gltA, ompA and/or ompB genes was performed (Parola et al., 2005). For the ‘SFG’ set, a mean cycle threshold (Ct) value below 35 indicates the sample is positive, and a Ct value above 35 indicates the sample is positive if another set is positive and/or a sequence is obtained and/or serology is positive. Thus, samples are run in duplicate using sets targeting two different genes.

From January 2009 to December 2009, the set ‘RAF-plasmid’ was used to detect R. africae; its target gene is located on a plasmid of the species. Following recent R. africae genome sequencing, it was reported that this plasmid might be unstable. To avoid false-negative results, we designed a new primer and probe set targeting a non-plasmidic gene. Consequently, the set ‘RAF’ was used to detect R. africae in clinical samples from January 2010 to December 2010.

We retrospectively collected data for the molecular diagnosis of rickettsioses from January 2009 to December 2010 to assess the usefulness of this strategy.

Except for the ‘SFG’ set, which had been previously described (Socolovsch et al., 2010), the sets were found to be specific for the corresponding rickettsial species both in silico and in vitro, when tested against a panel of 30 rickettsial strains (Fig. 1a). Sensitivity was also evaluated using 10-fold serial dilutions (Fig. 1b).

Figure 1

Sensitivity and specificity of the newly designed primers and probes. (a) Panel of 30 rickettsial species used in this study to evaluate specificity. (b) Evaluation of the sensitivity of the different sets of primers and probes used in this study: for the set detecting SFG Rickettsia, the evaluation of sensitivity presented here was performed using Rickettsia africae DNA; for TG Rickettsia, Rickettsia typhi DNA was used; for R. africae, R. africae DNA was used; for Rickettsia conorii, R. conorii DNA was used; for Rickettsia slovaca, R. slovaca DNA was used; for Rickettsia australis, R. australis DNA was used. (c) Determination of the sensitivity for the sets of primers and probes targeting R. africae and R. slovaca with a known inoculum. They present a limit of detection at 103 bacteria mL−1 for the set ‘RAF’ and 1 bacterium mL−1 for the set ‘RSLO’.

A total of 643 clinical specimens corresponding to 465 different patients were received at the FNRC from January 2009 to December 2010. Among these, 204 originated from locally hospitalized patients, 218 from other French hospitals and 43 from international hospitals. Forty-five positive qPCRs were obtained: 31/150 cutaneous biopsies, 8/42 cutaneous swab specimens, 2/223 total blood samples and 4/94 serum samples. The first molecular screening of SFG Rickettsia using the set labelled ‘SFG’ was positive for 44 samples; the 45th sample was positive using the set labelled ‘TG’, which detects TG Rickettsia. Among 45 positive results, 11 were obtained from locally hospitalized patients, 32 from other French hospitals and two from international hospitals.

A final diagnosis of R. africae was obtained for 15 samples (13 cutaneous biopsies, two eschar swabs) corresponding to 15 different patients with a diagnosis of ATBF; five samples were positive for the sets ‘SFG’ and ‘RAF-plasmid’, and 10 samples were positive for the sets ‘SFG’ and ‘RAF’. A final diagnosis of R. conorii was obtained for nine samples corresponding to nine different patients with a diagnosis of MSF; eight samples (cutaneous biopsies) were positive for the sets ‘SFG’ and ‘RCO’. One remaining sample (serum) was positive for the set ‘SFG’ and negative for ‘RCO’; a final diagnosis of R. conorii was obtained using conventional PCR followed by sequencing. A final diagnosis of R. honei was obtained for one sample (serum) corresponding to a patient whose final diagnosis was FISF (Murphy et al., 2011); it was positive for the set ‘SFG’, and a final diagnosis of R. honei was obtained by sequencing because no specific primer and probe set was available in our laboratory (Murphy et al., 2011). A final diagnosis of R. sibirica ssp. mongolitimonae was obtained for five samples corresponding to four different patients with a diagnosis of LAR, including a person returning from Egypt (Socolovschi et al., 2010). The samples (three cutaneous biopsies, two eschar swabs) were positive for the set ‘SFG’. A final diagnosis of R. sibirica ssp. mongolitimonae was obtained using conventional PCR followed by sequencing because no specific primer set was available in our laboratory. A final diagnosis of R. australis was obtained for two samples (cutaneous swabs) corresponding to a single patient with a diagnosis of QTT. The samples were positive for both ‘SFG’ and ‘RAUS’. A final diagnosis of R. slovaca was obtained for four samples (cutaneous biopsies) corresponding to three different patients with a diagnosis of SENLAT. Three samples were positive for both the ‘SFG’ and the ‘RSLO’ sets. One remaining sample (serum) was positive for the set ‘SFG’ and negative for ‘RSLO’; a final diagnosis of R. slovaca was obtained using conventional PCR followed by sequencing. A diagnosis of TG Rickettsia was obtained for one sample (serum) using the set ‘TG’; this sample corresponded to a patient with a diagnosis of murine typhus. Diagnosis at the species level was obtained by Western blot followed by cross-adsorption.

The remaining eight samples (three cutaneous biopsies, two cutaneous swabs, two total blood and one serum) were positive for the set ‘SFG’, but we could not discriminate at the species level using either molecular or serological techniques. These samples corresponded to eight patients with a diagnosis of rickettsiosis. For these eight samples, the Ct obtained using the set ‘SFG’ was significantly higher compared with the positive samples identified at the species level (36.71/31.95, P = 0.0023).

For one diagnosis of R. honei and five diagnoses of LAR, molecular diagnosis was performed by first screening using the ‘SFG’ set and then sequencing because specific primers and probes were not available. The need to resort to sequencing suggests the genomic databases must be updated regularly to develop new systems of primers and probes. Increased genomic data for Rickettsia species will permit the development of accurate qPCR tools.

For eight clinical samples, a diagnosis of rickettsiosis was obtained by systematic screening using the ‘SFG’ set. However, identification at the species level (by different sets of species-specific qPCRs or by conventional PCRs targeting gltA and ompA) remained unsuccessful. We demonstrated that the Ct values for such samples are significantly higher, suggesting that the ‘SFG’ set is more sensitive than conventional PCR (Angelakis et al., 2009); however, molecular tools for diagnosis at the species level are not yet sufficiently sensitive. In such cases, the use of highly sensitive techniques, such as suicide PCR or Western blot followed by cross-adsorption, remains necessary (Fournier & Raoult, 2004; Parola et al., 2005).

Among the positive clinical samples, 68.9% (31/45) were cutaneous biopsies, 17.8% (8/45) were cutaneous swabs, 4.4% (2/45) were total blood samples and 8.9% (4/45) were serum samples. The identification of rickettsial infections using cutaneous swab specimens and PCR testing has recently been reported (Bechah et al., 2011; Mouffok et al., 2011); based on these preliminary results, we collected cutaneous swabs from patients rather than cutaneous biopsies. Our retrospective analysis recovered eight positive cutaneous eschar swabs from different patients, confirming that these provide a rapid and simple means method that can be performed easily without the risk of the side effects related to biopsy collection in patients who display an inoculation eschar and/or a vesicular rash (Mouffok et al., 2011).

In conclusion, the widespread use of qPCR is less expensive than conventional PCR and reduces delay in the diagnosis of rickettsial infections. The development of qPCR strategies in the diagnosis of rickettsioses has previously been proposed (Stenos et al., 2005). Our 2 years of experience of rickettsial diagnosis using qPCR suggests that these molecular tools improve the efficiency of the management of patients with suspected cases of rickettsiosis. These qPCR assays could therefore be easily implemented in laboratories with molecular facilities and may be added to existing molecular tools as a point-of-care strategy (Holland & Kiechle, 2005).

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Table S1.Primers and probes used in this study.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

References

View Abstract