High throughput detection of Coxiella burnetii by real-time PCR with internal control system and automated DNA preparation

Abstract

Background: Coxiella burnetii is the causative agent of Q-fever, a widespread zoonosis. Due to its high environmental stability and infectivity it is regarded as a category B biological weapon agent. In domestic animals infection remains either asymptomatic or presents as infertility or abortion. Clinical presentation in humans can range from mild flu-like illness to acute pneumonia and hepatitis. Endocarditis represents the most common form of chronic Q-fever. In humans serology is the gold standard for diagnosis but is inadequate for early case detection. In order to serve as a diagnostic tool in an eventual biological weapon attack or in local epidemics we developed a real-time 5'nuclease based PCR assay with an internal control system. To facilitate high-throughput an automated extraction procedure was evaluated.

Results: To determine the minimum number of copies that are detectable at 95% chance probit analysis was used. Limit of detection in blood was 2,881 copies/ml [95%CI, 2,188-4,745 copies/ml] with a manual extraction procedure and 4,235 copies/ml [95%CI, 3,143-7,428 copies/ml] with a fully automated extraction procedure, respectively. To demonstrate clinical application a total of 72 specimens of animal origin were compared with respect to manual and automated extraction. A strong correlation between both methods was observed rendering both methods suitable. Testing of 247 follow up specimens of animal origin from a local Q-fever epidemic rendered real-time PCR more sensitive than conventional PCR.

Conclusion: A sensitive and thoroughly evaluated real-time PCR was established. Its high-throughput mode may show a useful approach to rapidly screen samples in local outbreaks for other organisms relevant for humans or animals. Compared to a conventional PCR assay sensitivity of real-time PCR was higher after testing samples from a local Q-fever outbreak.

Figure 1

Determination of detection limits, amplification efficiency of C. burnetii, correlation of automated and manual extraction. Probability of achieving a positive result (y-axis), depending on the DNA input copy number per mL EDTA blood (x-axis). A, Qiagen DNA mini kit; B, Qiagen M48 DNA mini kit, used on a Qiagen M48 automated DNA extraction instrument. Each datum point represents the rate of positive results in six replicate tests per concentration. Limits of detection are comparable with both methods of DNA extraction. C, Threshold cycles (y-axis) as a measure of efficiency of PCR amplification for C. burnetii and internal control. Each reaction contained 15 copies of plasmid-derived C. burnetii target gene and variable numbers of internal control plasmid pCoxmimic, as depicted on the x-axis. Results of eight replicate real-time PCR reactions per setting are shown as a result of box-plot analysis, showing the range of results by whiskers, whereby the two central quartiles of data are represented as a box. Solid line with grey boxes, C. burnetii target gene, broken line with white boxes, internal control. No reduced efficiency in amplification is observed for the C. burnetii target gene in presence of up to 100 copies of internal control. D, Correlation of C. burnetii DNA copies per ml as determined by C. burnetii real-time PCR after automated (x-axis) and manual extraction procedure (y-axis).

Figure 2

Bacterial loads in isolation positive samples, box plot analysis of Ct values. A, bacterial loads and C. burnetii isolation in real-time PCR positive samples (n = 27). Bacterial loads are shown on the y-axis. "+" in "cell isolation" means isolation success as confirmed by detection of inclusion bodies upon microscopy. B, box plot analysis of threshold cycle values in real-time PCR positive/conventional negative (n = 32) and real-time PCR positive/conventional PCR positive (n = 38) samples. Difference in threshold cycle values are significant (p < 0.05).