Seven novel probe systems for real-time PCR provide absolute single-base discrimination, higher signaling, and generic components

Abstract

We have developed novel probe systems for real-time PCR that provide higher specificity, greater sensitivity, and lower cost relative to dual-labeled probes. The seven DNA Detection Switch (DDS)-probe systems reported here employ two interacting polynucleotide components: a fluorescently labeled probe and a quencher antiprobe. High-fidelity detection is achieved with three DDS designs: two internal probes (internal DDS and Flip probes) and a primer probe (ZIPR probe), wherein each probe is combined with a carefully engineered, slightly mismatched, error-checking antiprobe. The antiprobe blocks off-target detection over a wide range of temperatures and facilitates multiplexing. Other designs (Universal probe, Half-Universal probe, and MacMan probe) use generic components that enable low-cost detection. Finally, single-molecule G-Force probes employ guanine-mediated fluorescent quenching by forming a hairpin between adjacent C-rich and G-rich sequences. Examples provided show how these probe technologies discriminate drug-resistant Mycobacterium tuberculosis mutants, Escherichia coli O157:H7, oncogenic EGFR deletion mutations, hepatitis B virus, influenza A/B strains, and single-nucleotide polymorphisms in the human VKORC1 gene.

Figure 1

The iDDS probe system. A and B: iDDS probe/antiprobe design favors probe binding to (in order of decreasing affinity): i) the correct target, ii) the antiprobe, and iii) a mismatched target. The fluorophore and quencher are indicated with white and black circles, respectively. Fluor-quencher alignment minimizes background fluorescence to increase the signal/noise ratio.C: Predicted thermodynamics of an MTB katG mutant 1 (M1) probe hybridizing to the M1 target, an antiprobe, an alternate mutant (M2), and WT katG. These T_m levels run high relative to other common on-line programs. D: Single-base discrimination with iDDS probes for two MTB katG mutants.

Figure 2

Performance of iDDS and TaqMan MGB probes in detecting a diagnostic SNP mutant of O157:H7 E. coli as a function of annealing temperature. Each graph depicts amplification plots with the mutant E. coli template, the WT E. coli template, and a water control, with the annealing temperature during qPCR varying from 63°C to 40°C. The TaqMan minor groove binder (MGB) probe shows increasingly obvious false-positive (FP) curves with the WT template as the annealing temperature decreases, and finally shows false-positives with a water control. The iDDS probe demonstrates effective and exclusive detection of the pathogenic E. coli SNP variant at all annealing temperatures from 63°C to 40°C.

Figure 3

MacMan probes. A: In the absence of target, free-floating MacMan probes are quenched by generic antiprobes that bind generic 3′ tails. Target detection occurs during qPCR as the MacMan probe anneals to the template, separating the probe from the quencher. B: MacMan probe detection of the S gene of hepatitis B virus DNA (BBI Diagnostics) versus two negative controls (HIV-1 pNL4-3 plasmid and an MTB ultramer template). C: Comparison of detection of the hepatitis S gene with MacMan versus TaqMan probes for the same target sequence.

Figure 4

Flip probes. A: In the quenched state, a Flip probe binds to a quencher-labeled antiprobe segment that is joined to a primer. The probe sequence (dashed line) is several bases longer than is the antiprobe sequence. The 5′ antiprobe sequence is structurally connected to the 3′ primer sequence by an S9 spacer (triethylene glycol, black box). The spacer prevents copying of the antiprobe sequence during qPCR. Numbering in the diagram reflects: i) probe bound to antiprobe, ii) probe bound to target, iii) primer linked to antiprobe, and iv) the second primer. B and C: Comparison of the TaqMan and Flip probes in detecting the MTB 16S gene, using MTB genomic DNA that was serially diluted from 10⁷ to 10⁴ copies per reaction. Since TaqMan curves can variably converge at the end point, cycle threshold values must be detected to determine target quantity. In contrast, Flip probe curves are linear so that end point fluorescence also reflects relative target frequency. Linear regression analysis of the MTB samples tested in C supports that end point fluorescence at 40 cycles correlates with copy number, and thus, quantitative end point detection may be achieved (D). R² = 0.985, based on four curves per dilution in C.

Figure 5

ZIPR probes. A: When ZIPR probes incorporate into PCR products, antiprobe binding is prevented and fluorescence signaling is enabled. ZIPR probes label all amplicons with high sensitivity, and antiprobe intervention favors on-target detection. Discriminating influenza strains using ZIPR probes specific for influenza A/New York/18/2009 M1 (B) or influenza B/Ohio/01/05 NS1 (C) transcripts. D: Strategy for detecting EGFR d-19 mutants. The reference ZIPR probe targets a conserved sequence upstream of the EGFR d-19 site, allowing it to detect both WT and deletion mutants with equal efficacy. In contrast, the FAM-labeled iDDS probe detects only the WT EGFR sequence, since all known d-19 mutations (9 to 24 base pairs) disrupt probe binding. Real-time PCR curves observed with samples containing 100% WT template (E), 50%/50% WT/mutant template (F), or 100% mutant template (G). d-19 Mutant frequencies are reflected by the angle of the FAM curve relative to the ROX-labeled reference curve.

Figure 6

Universal probes and Half-Universal probes. A: Universal probe assays involve two-step detection. The linker primer appends the amplicon with a binding site for the Universal probe (red). The linker primer starts amplification in the first cycle, and then the Universal probe takes over and provides signaling. Unincorporated probes are quenched by hybridization with antiprobes. The Universal probe is 7 bases longer than the antiprobe, favoring detection. Schematic representation of an influenza A HA3 segment and four selected Universal probe sites (B), and their use in detecting H3N2 influenza A strain A/Fujian/411/02 (D). C: Discriminating influenza A/California/04/2009 and A/Solomon Islands/03/2006, using viral cDNA and Universal probes. E: Multiplex detection of H5N1 influenza A/Hong Kong/483/97 cDNA at three different HA5 sites with Universal probes. F: Fluorescent Half–Universal-primer probes associate with antiprobes in the absence of target. Amplification separates Half-Universal probes from antiprobes, releasing fluorescence. G: A Half-Universal probe targeting H1N1 influenza A/California/4/2009 cDNA discriminated between H1N1 (A/Solomon Islands/03/2006) and H5N1 (A/Hong Kong/483/97) strains. H: Half-Universal probe-based detection of the influenza A/California/4/2009 HA gene compared with a TaqMan HA assay from the CDC real-time reverse-transcription PCR swine flu panel. I: Multiplex assays with Half-Universal probes targeting the influenza A/California/4/2009 HA gene at three sites. J: Multiplex discrimination of influenza A/California/4/2009 and A/Solomon Islands/03/2006 templates using Half-Universal probes. Parallel detection of serially diluted A/California/4/2009 cDNA with the HA1 and NA3 Half-Universal probes discussed in J and K.

Figure 7

G-Force probes. A: G-Force probes consist of complementary C- and G-rich regions that self-hybridize in the quenched state to juxtapose the fluorophore and the G-rich region, or separate during amplification to activate detection. B: A flanking G-Force probe targeting a conserved site in the human VKORC1 gene shows comparable detection of ultramer templates encoding the WT and SNP variant rs9923231^C. C: iDDS probe detection of the WT VKORC1 template, using mixed WT and SNP variant templates ranging from 100% to 0% WT. D: Multiplex detection of the WT MTB inhA gene with mixed WT/mutant templates. qPCR was performed using a reference G-Force probe and an iDDS probe targeting the WT sequence at a site that confers isoniazid resistance when mutated. With the 75% WT sample, the iDDS signaling curve drops about 25% relative to the G-Force curve, and with the 25% WT sample, the iDDS curve drops another 50% relative to the G-Force curve, providing semiquantitative mutant frequency analysis. w/m, WT/mutant.