Multiplexed detection of anthrax-related toxin genes

Abstract

Simultaneous analysis of three targets in three colors on any real-time polymerase chain reaction (PCR) instrument would increase the flexibility of real-time PCR. For the detection of Bacillus strains that can cause inhalation anthrax-related illness, this ability would be valuable because two plasmids confer virulence, and internal positive controls are needed to monitor the testing in cases lacking target-specific signals. Using a real-time PCR platform called MultiCode-RTx, multiple assays were developed that specifically monitor the presence of Bacillus anthracis-specific virulence plasmid-associated genes. In particular for use on LightCycler-1, two triplex RTx systems demonstrated high sensitivity with limits of detection nearing single-copy levels for both plasmids. Specificity was established using a combination of Ct values and correct amplicon melting temperatures. All reactions were further verified by detection of an internal positive control. For these two triplex RTx assays, the analytical detection limit was one to nine plasmid copy equivalents, 100% analytical specificity with a 95% confidence interval (CI) of 9%, and 100% analytical sensitivity with a CI of 2%. Although further testing using clinical or environmental samples will be required to assess diagnostic sensitivity and specificity, the RTx platform achieves similar results to those of probe-based real-time systems.

Figure 1

MultiCode-RTx system schematic. Targets are amplified with a standard reverse primer and a forward primer that contains a single iC nucleotide and a fluorescent reporter. Amplification is performed in the presence of dabcyl-diGTP. Site-specific incorporation places the quencher in close proximity to the reporter, leading to a decrease in fluorescence that can be observed during real-time PCR.

Figure 2

Real-time PCR detection and linear standard curve analysis of anthrax-related toxin genes. The two triplex MultiCode-RTx systems designed using Visual OMP, pagA:capB:IPC (A, pagA; B, capB) and cya:capB:IPC (C, cya; D, capB), were tested for linearity for both corresponding synthetic targets using 10-fold dilution series from 3 to 3 × 10⁵ copies in duplicate on different days. Top panels show the decrease in fluorescence as relative fluorescence units (RFU) versus PCR cycles for all samples as they appear during the real time PCR reactions. As the fluorescence decrease passes the cycle threshold (Ct) line, a cycle number is tabulated and used to generate standard curves. Reactions containing larger target copy numbers yield a detectable decrease in RFU in fewer rounds of PCR corresponding to smaller Ct values. Bottom panels show linear curve analyses of log copy number versus Ct with best fit equation and R² values. Internal positive control is not shown.

Figure 3

Limit of detection (LOD) for B. anthracis Ames genomic DNA. Representative data determining LOD using 10-fold dilution series from 1 pg to 1 fg of B. anthracis Ames total genomic DNA in duplicate. A: Real-time PCR data of decreasing relative fluorescence units (RFU) versus PCR cycle number. B: Postamplification thermal melt analysis of PCR products as the negative first derivative of RFU with respect to temperature (−d(RFU)/dT) versus temperature in °C. Data are from the cya:capB:IPC multiplex. LOD for cya primer set was 10 fg or ∼2 copies.

Figure 4

Detection of nonspecific amplification by T_m. For both A and B, the top amplification curves show real-time PCR data in relative fluorescence units (RFU) versus PCR cycle number, and the bottom panels show postamplification melt data as the negative first derivative of RFU with respect to temperature (−d(RFU)/dT) versus temperature in °C. A representative data set using a series of unrelated organisms along with 1 pg of Ames DNA control was amplified by the pagA:capB:IPC triplex primer set. A: Real-time and amplicon melt data from the pagA read in F1 channel. B: Data from capB read in F2 channel. The Salmonella choleraesius 9150 sample (*) amplified weakly in the F1 channel with a Ct of 42.2 compared with 34.4 for the Ames control (+). The F1 Melt Tm for the S. choleraesius amplicon was significantly different at 76.0°C compared with 79.3°C for the control (A, bottom), indicating that the S. choleraesius amplification product observed during real-time PCR was nonspecific. Triplicate re-testing of the S. choleraesius sample gave no detectable amplification (data not shown).