Special Section Rapid Genotyping and Quantification on the LightCyclerTM with Hybridization Probes

DEEPIKA DE SILVA1, ASTRID REISER 2, MARK HERRMANN3, KARIM TABITI2, and CARL WITTWER3 1 ARUP Institute for Clinical and Experimental Pathology, 500 Chipeta Way, Salt Lake City, UT 84108 2 Roche Molecular Biochemicals, D-82377, Penzberg, Germany and 3 Department of Pathology, University of Utah School of Medicine, 50 North Medical Drive, Salt Lake City, UT 84132

ey Word: LightCycler™, PCR, hybridization bstract probes, mutation detection, Continuous monitoring of PCR using fluorescently labeled hybridization probes is quantification used to demonstrate rapid PCR quantification and genotyping. Two hybridization probes are designed that recognize adjacent internal sequences within target DNA sequences. Both probes are fluorescently labeled such that, when annealed to the template, the fluorophores are separated by one base. Proximity of the fluors allows resonance energy transfer that is monitored by the LightCycler™. PCR product quantitation is demon- strated using a probe pair for b-globin. Quantification of the initial template copy number over a 106-fold range is illustrated, with the ability to distinguish a single starting template copy from a no template control. A two-probe system is also used for rapid genotyping of the factor V Leiden mutation.

Introduction Methods used for DNA analysis usually involve amplification of target regions by PCR followed by separate product analysis. Sequence-specific monitoring of PCR products is routinely performed by hybridization analysis using blots, gels, or microtiter plates. Hybridization of small probes to template DNA can be visualized with radioactively labeled probes, fluorescently labeled probes, or chemiluminescent tech- niques. These techniques, however, are time-consuming and can involve several handling steps that increase the risk of end-product contamination and sample tracking errors. The LightCycler™ is a microvolume fluorometer integrated with a thermal cycler that com- bines rapid-cycle PCR with real-time fluorescence monitoring (1). Sequence-specific hybridization probes can be designed that allow detection and analysis of PCR products on the LightCycler without the need for any post-PCR sample manipulation (2), allowing high throughput genotyping and product quantification.

Fluorescence offers advantages over other techniques, notably its linear response over a large dynamic range. Fluorescence monitoring of amplification using hybridization probes is based on the concept that a fluorescence signal is generated if fluorescence res- onance energy transfer (FRET) occurs between two adjacent fluorophores, illustrated in Figure 1 (3). Fluorescence-based, single step genotyping methods have been described for the factor V Leiden mutation (4) and the C677T point mutation in the methylenetet- rahydrofolate reductase gene (5) using the LightCycler. Amplification and analysis take approximately 30 min and require no sample handling after loading on the LightCycler. Both these applications employed one labeled primer and one labeled probe with asym-

12 BIOCHEMICA ■ NO. 2 [1998] ➤ CONTENTS Special Section

Figure 1 Fluorescent resonance energy transfer between hybridization probes. A graphical representation of energy transfer between a donor fluorophore and an accep- tor fluorophore on adjacent hybridization probes. Monitoring the fluorescence emission from the acceptor fluorophore after excitation of the donor fluorophore allows highly sensitive product quantification and genotyping. metric amplification of target sequences using a probe pair for b-globin. Fluores- Materials and Methods using a Cy5-labeled primer. Hybridization cence is monitored once each cycle at the Primer and probe design and synthesis of a 3' -labeled probe to the end of the annealing step and increases The factor V amplicon was derived Cy5-labeled target strand results in the two above background at a cycle number that from exon 10 and comprised a 220 base fluorophores being brought into close is dependent on the initial template con- pair fragment. The sense and reverse proximity allowing FRET to occur. After centration. A two probe system is also used amplification primers were a 22-mer and a amplification, a melting curve is generated for genotyping the factor V Leiden muta- 20-mer (antisense intron 10), respectively. that allows rapid genotyping. Mutant tion, showing that both a primer-probe The anchor probe was a 36-mer (antisense alleles are distinguished from wild type by system and a probe-probe system are exon 10), labeled at the 5' end with Light- the melting temperature (Tm) of the probe. equally reliable for mutation detection. Cycler-Red 640 (LC-Red 640) and modi- Continuous fluorescence monitoring of fied at the 3' end by phosphorylation. The the reaction as the temperature is raised from annealing to denaturation, results in a sharp decrease in fluorescence at the tem- perature at which the probe dissociates from the labeled template. A single base change (G1691A) caused by the Leiden mutation results in a decrease in probe Tm of 7oC. A more stable substitution in the MTHFR gene (C677T) gives a 3oC differ- ence in Tm between wild type and mutant alleles. Both changes can easily be distin- guished with the LightCycler, suggesting that this fluorescence method is suitable for all single base mismatches. This paper describes experiments using two singly labeled probes for rapid, sequence-specific mutation detection and quantification. Hybridization probe pairs are designed that recognize adjacent internal sequences within target DNA Figure 2 Binding of the fluorescently-labeled hybridization probes. sequences. Both probes are fluorescently A. Binding of the hybridization probes to the b-globin template. When the probes are hybridized to the template labeled such that, when annealed to the there is one base separation between the 2 fluorophores. B. Binding of the hybridization probes to the Factor V gene. The long anchor probe is labeled with LightCycler-Red template, the fluorophores are separated 640 (LC-Red 640). A shorter hybridization probe is labeled with fluorescein and is complementary to the wild type by one base, allowing a strong FRET signal. sequence. The Leiden mutation would create a single mismatch located in the center of the shorter probe. PCR product quantitation is demonstrated

BIOCHEMICA ■ NO. 2 [1998] 13 ➤ CONTENTS Special Section

(BSA). For the factor V mutation detec- donor fluorophore with an emission spec- tion experiments, 0.2 µM anchor probe trum that overlaps the excitation spectrum Fluorescein LC-Red 640 and 0.2 µM hybridization probe and of an acceptor fluorophore results in non- Cy5 0.4U Taq DNA polymerase per 10 µl radioactive energy transfer to the acceptor. sample was included. For the quantita- The emission spectra of three fluorophores tion experiments with b-globin, 0.2 µM used with the LightCycler are shown in fluorescein labeled probe and 0.4 µM Figure 3. Fluorescein is the donor fluoro- Cy5-labeled probe was used. 0.8 U of phore of choice for the LightCycler, Cy5Ò exonuclease minus polymerase per 10 µl (Amersham), or LightCycler-Red 640 sample was used after preincubation with (Roche Molecular Biochemicals) can be anti-Taq antibody. used as the acceptor fluorophore. Figure 3 Emission spectra of the three dyes Samples were spun into glass capillary used in the experiments. Spectra were collected in cuvettes, capped and placed in the Light- Sequence-specific quantification 3 mM PCR buffer to mimic conditions used with the Cycler. For factor V genotyping, amplifica- Hybridization probes can be used for LightCycler. Fluorescein is used as the donor fluoro- tion was performed for 40 cycles of sequence-specific quantification of PCR phore and either Cy5™ or LC-Red 640 can be used denaturation (94oC, 0 s, ramp rate 20oC/s), products. Quantification of the initial tem- as the acceptor fluorophore. annealing (50oC, 10 s, ramp rate 20oC/s), plate copy number over a 1010-fold range and extension (72oC, 10 s, ramp rate 2oC/ 3' fluorescein-labeled hybridization probe s). After amplification, a melting curve was was a 23-mer (antisense exon 10/intron 10 generated by holding the reaction at 50oC junction). for 30 s, and then heating slowly to 94oC The b-globin amplicon is a 110 base with a ramp rate of 0.2oC per s. Fluores- pair sequence that was amplified using a cence was collected continuously during 20-mer forward primer and a 20-mer the slow temperature ramp. The fluores- reverse primer. The first hybridization cence signal from LC-Red 640 was plotted probe (35-mer) was 3'-labeled with fluor- against temperature to give melting curves escein and the second probe (30-mer) was for each sample. Melting curves were con- 5'-labeled with Cy5. Primers and probes verted to melting peaks by plotting the were synthesized and purified as described negative derivative of the fluorescence with previously (4). The binding of the fluores- respect to temperature (-dF/dT) against cently labeled probes is illustrated in temperature. The entire process takes Figure 2. approximately 30 min with no separate Figure 5 Factor V Leiden genotyping with melt- product manipulation necessary. ing peaks. Each peak represents the melting tem- Fluorescence protocols For the b-globin quantification experi- perature of the short, fluorescein labeled probe off the template. The melting temperature of the probe PCR was performed with 0.2 µM ments, 10-fold serial dilutions of purified off a perfectly matched wild type template is 64oC, 6 primers in a standard PCR reaction, sup- amplicon were made, ranging from 10 the presence of the Leiden mutation causes a 7oC plemented with bovine serum albumin template copies to one copy. Amplification shift in melting temperature to 57oC. Heterozygotes was performed for 45 cycles of denatur- show both peaks representing both the normal and ation (94oC, 0 s, ramp rate 20oC/sec), mutant alleles. no annealing (60oC, 20 s, ramp rate 20oC/s), 1 o o 10 and extension (75 C, 0 s, ramp rate 1 C/s). is possible (2). Figure 4 shows the ampli- 102 A single fluorescence reading for each fication of purified b-globin serial dilu- 103 6 104 sample was taken at the annealing step. At tions, ranging from 10 initial template 105 this stage of the PCR cycle both probes are copies to one copy with fluorescence plot- 6 10 allowed to hybridize and the total amount ted against cycle number. The higher the of product formed was determined by the initial copy number, the earlier the fluores- Figure 4 Fluorescence ratio (Cy5/FITC) vs. cycle log-linear slope intercept of the X-axis. cent signal appears out of the background. number for template quantification using hybrid- As the initial template concentration ization probes. Serial dilutions of purified b-globin Results and Discussion decreases, the curves are shifted to the amplicon were amplified in the presence of fluores- FRET probes right. The sensitivity of the system is such cent hybridization probes. All samples were amplified simultaneously and fluorescence was acquired for Fluorescence monitoring of PCR that a single template copy can be distin- 100 ms for each sample between seconds 13 and 20 amplification is based on the concept of guished from the absence of template. of the annealing phase of the reaction. Data repre- fluorescence resonance energy transfer Fluorescence is acquired once per sample, sents starting template copies of 106, 105, 104, 103, between two adjacent dyes. Excitation of a just before extension when the probes are 102, 10, 1, and a no template control.

14 BIOCHEMICA ■ NO. 2 [1998] ➤ CONTENTS Special Section hybridized, in order to obtain the strongest fluorescence signal since the efficiency of concerns. Rapid cycling combined with signal. energy transfer between dyes decreases sequence-specific fluorescence monitoring with distance. Ideally a spacing of one to on the LightCycler is a powerful tool. Factor V genotyping four bases is optimal with the signal A typical factor V genotyping experi- decreasing as the fluors are separated by Acknowledgments ment using two hybridization probes is greater than ten bases. The fluorophores The authors gratefully acknowledge illustrated below (Figure 5). A long on the probe pairs used in these experi- Klaus Muehlegger (Roche Molecular Bio- anchor probe (a 36-mer) is labeled at the ments were separated by one base to max- chemicals, Penzberg) for synthesis and 5' end with LightCycler Red 640 and a imize the signal. labeling of the LightCycler Red-640 Factor second shorter probe (a 23-mer) is labeled The Leiden mutation represents a G to V hybridization probe. with fluorescein at the 3' end. The probes A substitution within exon 10 of the factor recognize adjacent sequences with the V gene creating an A:C mismatch between References shorter probe lying over the mutation site. the wild type probe and the mutant tem- 1. Wittwer, C., Ririe, K., Andrew, R., David, D., Gundry, TM The greater stability of the anchor probe plate. This mismatch results in a large T R., and Balis, U. (1997) The LightCycler : A micro- m volume multisample fluorimeter with rapid temper- means that loss of fluorescence occurs as shift due to its poor stability and allows the ature control. BioTechniques 22 (1): 176–181. the shorter probe melts off the template. A mutant to be easily distinguished from the 2. Wittwer, C., Herrmann, M., Moss, A., and Rasmussen, single base mismatch under the probe wild type. Other more stable mismatches, R. (1997) Continuous fluorescence monitoring of rapid o cycle DNA amplification. BioTechniques 22 (1): 130– results in a Tm shift of 7 C. Melting curves however, have also been detected using the are converted to melting peaks (Figure 5) same technique (5). One option for 138. 3. Livak, K.J., Flood, S.J.A., Marmaro, J., Giusti, W., allowing easy distinction of wild type from designing probes using this system is to and Deetz, K. (1995) with fluores- mutant. determine which combination (wild type cent dyes at opposite ends provide a quenched Heterozygotes show two distinct melt- probe with mutant template or mutant probe system useful for detecting PCR product and ing peaks representing one mutant and probe with wild type template) is the least hybridization. PCR Methods Appl. 4: one wild type allele. The feasibility of rapid stable and design probes accordingly. If a 357–62. 4. Lay, M. and Wittwer, C. (1997) Real time fluores- genotyping of the factor V Leiden mutation mutant probe with a wild type template is cence genotyping of factor V Leiden during rapid has been previously demonstrated using a selected, the probe will melt off the wild cycle PCR. Clin. Chem. 43 (12): 2262–2267. primer-probe system (4). We show here type allele at a lower Tm than the mutant 5. Bernard, P., Lay, M. and Wittwer. (1998) Integrated that an internal hybridization probe pair allele for which it is a perfect match. amplification and detection of the C677T point works equally well. The probe pair system This paper has dealt with the useful- mutation in the methylenetetrahydrofolate reduc- tase gene by fluorescence resonance energy trans- may be more attractive to many investiga- ness of hybridization probe pairs in fluor- fer and probe melting curves. Anal. Biochem. 255: tors since the synthesis of labeled primers escence monitoring of PCR products using 101–107. is often more difficult. An added advantage the LightCycler. The ability to monitor to the probe-probe system is that the PCR product formation in real-time is Product Cat. No. same amplification primer set can be helpful no matter what fluorescence tech- LightCycler 2 011 468 used with a panel of hybridization nique is employed. The advantage of probes to screen target DNA for multiple hybridization probes is that real-time Cy5 is covered by patents of Amersham. mutation sites. hybridization information is available, The probe design of these experiments allowing mutation detection and quantifi- places both probes on the same strand, as cation in the same reaction. This permits far from the extending primer as possible. product identification, even down to a This delays displacement of the probe by single base change, during amplification. the polymerase and allows maximum Such a “dynamic dot ” removes the hybridization time. The spacing between requirements for processing after amplifi- the fluorophores can be critical for a strong cation and eliminates PCR contamination

BIOCHEMICA ■ NO. 2 [1998] 15 ➤ CONTENTS