AU747662B2 - Fluorescent donor-acceptor pair - Google Patents

Description

AUSTRALIA

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT

ORIGINAL

S.

t Name of Applicant/s: Actual Inventor/s: Address for Service: Invention Title: University Of Utah Research Foundation Carl T. Wittwer BALDWIN SHELSTON WATERS MARGARET STREET SYDNEY NSW 2000 'FLUORESCENT DONOR-ACCEPTOR PAIR' Details of Original Application No. 34812/97 dated 4 June 1997 The following statement is a full description of this invention, including the best method of performing it known to me/us:- File: 25963AUP00 FLUORESCENT DONOR-ACCEPTOR PAIR BACKGROUND OF THE INVENTION This invention relates generally to observing fluorescence signals resulting from hybridization in conjunction with the polymerase chain reaction. More specifically, the present invention relates to observing hybridization with fluorescence during and/or immediately after PCR and using this information for product identification, sequence alteration detection, and quantification.

This application is a divisional application of Australian Patent Application No. 34812/97, the contents of which are incorporated herein by reference.

The polymerase chain reaction (PCR) is fundamental to molecular biology and is the first practical molecular technique for the clinical laboratory. Despite its usefulness and popularity, current understanding of PCR is not highly advanced. Adequate conditions for successful amplifications must be found by trial and error and optimization is empirical. Even those skilled in the art are required to utilize a powerful technique without a comprehensive or predictive theory of the process.

PCR is achieved by temperature cycling of the sample, causing DNA to denature (separate), specific primers to attach (anneal), and replication to occur (extend). One cycle of PCR is usually performed in 2 to 8 min, requiring 1 to 4 hours for a 30-cycle amplification. The S sample temperature response in most PCR instrumentation is very slow compared to the times required for denaturation, annealing, and Sextension. The physical (denaturation and annealing) and enzymatic (extension) reactions in PCR occur very quickly. Amplification times for PCR can be reduced from hours to less than 15 min. Incorporated herein by reference in its entirety is U.S. Serial No. 08/537,612, filed o 30 October 2, 1995, which discloses such a rapid cycling system. Rapid eooe co cycling techniques are made possible by the rapid temperature response *co and temperature homogeneity possible for samples in high surface areaoee to-volume sample containers such as capillary tubes. For further information, see also: C.T. Wittwer, G.B. Reed, and K.M. Ririe, Rapid cycle DNA amplification, in K.B. Mullis, F. Ferre, and R.A. Gibbs, The Spolymerase chain reaction, Birkhauser, Boston, 174-181, (1994).

Improved temperature homogeneity allows the time and temperature requirements of PCR to be better defined and understood. Improved temperature homogeneity also increases the precision of any analytical technique used to monitor PCR during amplification.

Fluorimetry is a sensitive and versatile technique with many applications in molecular biology. Ethidium bromide has been used for many years to visualize the size distribution of nucleic acids separated by gel electrophoresis. The gel is usually transilluminated with ultraviolet light and the red fluorescence of double stranded nucleic acid observed. Specifically, ethidium bromide is commonly used to analyze the products of PCR after amplification is completed.

Furthermore, EPA 0 640 828 Al to Higuchi Watson, hereby incorporated by reference, discloses using ethidium bromide during amplification to monitor the amount of double stranded DNA by measuring the fluorescence each cycle. The fluorescence intensity was noted to rise and fall inversely with temperature, was greatest at the annealing/extension temperature (50 0 and least at the denaturation temperature (94 0

C).

Maximal fluorescence was acquired each cycle as a measure of DNA amount. The Higuchi Watson application does not teach using fluorescence to monitor hybridization events, nor does it suggest acquiring fluorescence over different temperatures to follow the extent of hybridization. Moreover, Higuch Watson fails to teach or suggest using the temperature dependence of PCR product hybridization for identification or quantification of PCR products.

The Higuchi Watson application, however, does mention using other fluorophores, including dual-labeled probe systems that generate flourescence when hydrolyzed by the 5'-exonuclease activity of certain DNA polymerases, as disclosed in US Patent No. 5,210,015 to Gelfand et al. The fluorescence observed from these probes primarily depends on hydrolysis of the probe between its two fluorophores. The amount of PCR product is estimated by acquiring fluorescence once each cycle.

Although hybridization of these probes appears necessary for hydrolysis to occur, the fluorescence signal primarily results from hydrolysis of *e the probes, not hybridization, wherein an oligonucleotide probe with fluorescent dyes at opposite ends thereof provides a quenched probe system useful for detecting PCR product and nucleic acid hybridization, K.J. Livak et al., 4 PCR Meth. Appl. 357-362 (1995). There is no suggestion of following the temperature dependence of probe hybridization with fluorescence to identify sequence alterations in PCR products.

The specific hybridization of nucleic acid to a complementary strand for identification has been exploited in many different formats.

For example, after restriction enzyme digestion, genomic DNA can be size fractionated and hybridized to probes by Southern blotting. As another example, single base mutations can be detected by "dot blots" with allele-specific oligonucleotides. Usually, hybridization is performed for minutes to hours at a single temperature to achieve the necessary discrimination. Alternately, the extent of hybridization can be dynamically monitored while the temperature is changing by using -3fluorescence techniques. For example, fluorescence melting curves have been used to monitor hybridization. L.E. Morrison L.M. Stols, Sensitive fluorescence-based thermodynamic and kinetic measurements of DNA hybridization in solution, 32 Biochemistry 3095-3104, 1993). The temperature scan rates are usually 10°C/hour or less, partly because of the high thermal mass of the fluorimeter cuvette.

Current methods for monitoring hybridization require a lot of time. If hybridization could be followed in seconds rather than hours, hybridization could be monitored during PCR amplification, even during rapid cycle PCR. The many uses of monitoring hybridization during PCR, as will be fully disclosed herein, include, product identification and quantification, sequence alteration detection, and automatic control of temperature cycling parameters by fluorescence feedback.

The prior art, as explained above, carries out temperature cycling slowly and empirically. When analysis of PCR products by hybridization is needed, additional time S-.i consuming steps are required. Thus, it would be a great advance in the art to provide a. methods for monitoring hybridization during PCR and analyzing the reaction while it is taking place, that is, during or immediately after temperature cycling without manipulation of the sample. By monitoring hybridization during PCR, the underlying "principles that allow PCR to work can be followed and used to analyze and optimize the reaction during amplification.

20 BRIEF SUMMARY OF THE INVENTION It is an object of the present invention to overcome or ameliorate at least one of the disadvantages of the prior art, or to provide a useful alternative.

oooo* In a first aspect of the present invention there is provided a detectable complex, comprising a target analyte, a first probe labeled with fluorescein, and a second probe labeled with Cy5 or Cy5.5, wherein the first and said second probe are bound to the target analyte and are in a fluorescence resonance energy transfer relationship.

In a second aspect of the present invention there is provided a detectable complex, comprising a probe system comprising a donor fluorophore having an emission spectrum and an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum Soverlap less than 25%, and wherein the donor and acceptor fluorophore are in a fluorescence resonance energy transfer relationship.

3a- One illustrative fluorescence resonance energy transfer pair described is where the donor fluorophore is fluorescein and the acceptor fluorophore is Cy5 or In a third aspect of the present invention there is provided a detectable complex, comprising: a target analyte, a first probe comprising a donor fluorophore having an emission spectrum, and a second probe comprising an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap less than 25%, and wherein the first and second probes are bound to the target analyte and the first and second probes are in a fluorescence resonance energy transfer relationship.

In a fourth aspect of the present invention there is provided a detectable nucleic acid complex for the detection of a locus of a nucleic acid sample, comprising: A a donor probe hybridised to the locus, the donor probe comprising an 15 oligonucleotide labelled with a donor fluorophore having an emission spectrum, an acceptor probe hybridised to the locus, the acceptor probe comprising an oligonucleotide labelled with an acceptor fluorophore having an absorption spectrum and an extinction coefficient greater than 100,000 M cm 1 wherein the donor fluorophore and the acceptor fluorophore comprise a fluorescence resonance energy transfer pair 20 such that the donor fluorophore emission spectrum and the acceptor fluorophore absorption spectrum overlap by less than 25%, and the donor and acceptor fluorophore's 'are in a resonance energy transfer relationship; wherein the nucleic acid sample further comprises components for amplifying the locus.

In a fifth aspect of the present invention there is provided a kit for amplification and analysis of a locus of a nucleic acid sample, comprising: a donor probe comprising an oligonucleotide labelled with a donor fluorophore having an emission spectrum; an acceptor probe comprising an oligonucleotide labelled with an acceptor fluorophore having an absorption spectrum and an extinction coefficient greater than 100,000 M^cm, wherein the donor fluorophore and the acceptor fluorophore comprise a fluorescence resonance energy transfer pair such that the donor fluorophore emission 3b spectrum and the acceptor fluorophore absorption spectrum overlap by less than and wherein the donor and the acceptor fluorophore's are in a resonance energy transfer relationship when the donor probe and acceptor probe are hybridised to the locus; and components for amplifying the locus; wherein the donor probe and the acceptor probe are provided together in the same container.

In a sixth aspect of the present invention there is provided a detectable nucleic acid complex, comprising: an amplified nucleic acid, wherein the nucleic acid was amplified with first and second primers; and a probe hybridised to the amplified nucleic acid; wherein the first primer and the probe are each labelled with one member of a fluorescence energy transfer pair comprising a donor fluorophore and an acceptor fluorophore such that the donor fluorophore emission spectrum and the acceptor 15 fluorophore absorption spectrum overlap by less than 25%, and wherein the donor and acceptor fluorophore's are in a resonance energy transfer relationship.

a.

a. a a a. a a.

a a.

a a.

a a a a.* a.

*a a a -4- Unless the context clearly requires otherwise, throughout the description and the claims, the words 'comprise', 'comprising', and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in the sense of "including, but not limited to".

Any discussion of the prior art throughout the specification should in no way be considered as an admission that such prior art is widely known or forms part of common general knowledge in the field.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS Figures 1A&B are graphs representing an equilibrium PCR paradigm and a kinetic PCR paradigm Figure 2 illustrates useful temperature v. time segments for fluorescence hybridization monitoring.

Figure 3 is a temperature v. time chart exemplary of rapid temperature cycling for

PCR.

a# Figure 4 shows the results of four different temperature/time profiles and their resultant amplification products after thirty cycles (inset).

Figures 5A-F, illustrate the mechanism of fluorescence generation for three different methods of fluorescence monitoring of PCR double-standard DNA dye ol: hydrolysis probe, and hybridization probes.

20 Figure 6 shows the chemical structure of the monovalent N-hydroxysuccinimide ester of Cy5 T M 0 Figure 7 shows the chemical structure of the monovalent N-hydroxysuccinimide •o••i S•ester of Cy5.5 TM Figure 8 shows the emission spectrum of fluorescein (solid line) and the excitation spectrum of Cy5 (broken line).

Figures 9A, B C show resonance energy transfer occurring between fluoresceinand Cy5-labeled adjacent hybridization probes at each cycle during PCR.

Figure 10 shows the effect of varying the ratio of the Cy5 probe to the fluorescein probe on the resonance energy transfer signal generated during PCR.

Figure 11 shows the effect of varying the probe concentration at a given probe ratio Zon the resonance energy transfer signal generated during PCR.

4a Figure 12 shows the effect of spacing between the labeled oligonucleotides on the resonance energy transfer signal generated during PCR.

Figure 13 shows the time course of adjacent probe hybridization by fluorescence energy transfer immediately after 30 cycles of amplification with Taq DNA polymerase (exo+; solid line) and the Stoffel fragment of Taq DNA polymerase (exo-; dotted line) of temperature cycling and the type of polymerase on fluorescence development during PCR with adjacent hybridization probes; temperature is shown with a bold line.

So a *o Figure 14 is a fluorescence ratio v. cycle number plot for amplification with Taq DNA polymerase (exo+; solid line), Stoffel fragment of Taq DNA polymerase (exo-; broken line), and KlenTaq DNA polymerase (exo-; dotted line): top panel cycling is between 94 0 C and 60 0 C with a 20 second hold at 60 0 C; middle panel cycling is between 94 0 C and 60 0 C with a 120 second hold at 60 0 C; bottom panel cycling is between 94 0 C and 60 0 C with a slow increase from 60 0 C to 75 0

C.

Figure 15 is a fluorescence v. cycle number plot for a number of different initial template copy reactions monitored with Sybr T M Green I: 0, 0, 102, 103, 104, 105, 106, 107, 108, 109, Figure 16 is a fluorescence ratio v. cycle number plot for a number of different initial template copy reactions monitored with a dual-labeled hydrolysis probe: 0, 1, 10, 102, 103, 104, 10 5 106, 107, 108, 109, Figure 17 is a fluorescence ratio v. cycle number plot for a number of different initial template copy reactions monitored with adjacent hybridization probes: 0, 1, 10, 102, 103, 10 105, 106, 107, 108, 10 9 Figure 18 is a fluorescence ratio v. cycle number plot distinguishing two hybridization probe designs monitored by resonance energy transfer: two hybridization probes labeled respectively with fluorescein and Cy5; and primer labeled with Cy5 and a probe S.labeled with fluorescein.

Figures 19A-C provide a comparison of three fluorescence :monitoring techniques for PCR, including the double-strand specific DNA dye SYBR Green I a dual-labeled fluorescein/rhodamine hydrolysis probe and a fluorescein-labeled hybridization probe with a labeled primer Figure 19D shows the coefficient of variation for the three monitoring techniques represented in Figures 19A-C.

Figure 20 shows a typical log fluorescence vs cycle number plot of a standard amplification monitored with SYBR Green I.

Figure 21 shows an expontial curve fit to cycles 20-27 of the data from Figure S 35 Figure 22 shows an exponential fit to an unknown to determine initial copy number from amplification data.

Figure 23 shows a typical fluorescence v. cycle number plot of five standards monitored each cycle with adjacent hybridization probes, wherein initial copy numbers are represented as follows: 103, 104, 105, 106, 107, Figure 24 shows a curve fit to the standard data of Figure 23.

Figure 25 shows a typical fluorescence vs cycle number plot of five standards monitored each cycle with a hydrolysis probe, wherein initial copy numbers are represented as follows: 1.5, 15, 150, 1500, 15,000, Figure 26 shows a curve fit to the standard data of Figure Figure 27 shows a typical log fluorescence vs cycle number plot of three standard amplifications monitored with SYBR Green I, wherein: (A Figure 28 shows different curve fit to the standard data of Figure 27.

Figures 29A&B show plots of time v. fluorescence and time v. temperature demonstrating the inverse relationship between temperature and fluorescence.

Figures 30A, B, C D are charts showing 2D plots of temperature v. time, fluorescence v. time, and fluorescence v. temperature, also shown as a 3D plot, for the amplification of a 180 base pair fragment of the hepatitis B genome in the presence of SYBR Green I.

Figure 31 is a fluorescence v. temperature projection for the amplification of a 536 base pair fragment of the human beta-globin gene in the presence of SYBR Green I.

Figures 32A&B provide a plot showing a linear change in fluorescence ratio with temperature for hydrolysis probes, and a radical change with temperature for hybridization probes.

Figure 33 shows a fluorescence ratio v. temperature plot of 25 amplification with an exo- polymerase in the presence of adjacent hybridization probes.

Figure 34 shows a fluorescence ratio v. temperature plot of amplification with an exo' polymerase in the presence of adjacent hybridization probes.

30 Figure 35 shows a 3-dimensional plot of temperature, time and o:o fluorescence during amplification with an exo- polymerase in the presence of adjacent hybridization probes.

Figure 36 shows a 3-dimensional plot of temperature, time, and fluorescence during amplification with an exo polymerase in the 35 presence of adjacent hybridization probes.

o. Figure 37 shows melting curves for PCR-amplified products of hepatitis B virus 50% GC, 180 bp); beta-globin 53.2% GC, 536 bp); and prostate specific antigen 60.3% GC, 292 bp).

Figure 38 shows melting curves for PCR-amplified product of hepatitis B virus at heating rates of 0.1 0 C to 5.0 0

C.

Figure 39 shows melting curves for PCR-amplified product of hepatitis B virus at various SYBRTM Green I concentrations.

Figures 40A&B show melting curves and electrophoretically fractionated bands of products of a beta-globin fragment amplified with no added template, 106 copies of added template under low stringency conditions, and 106 copies of added template under higher stringency conditions.

Figures 41A&B show melting curves and melting peaks of hepatitis B virus fragment (HBV), P-globin, and a mixture thereof.

Figures 42A-D show a relative fluorescence v. cycle number plot for PCR amplified products from various amounts of P-globin template, melting peaks and electrophoretic bands of the various products, and corrected fluorescence of the data of Figures 43A&B show melting curves and melting peaks from PCR amplified products of a mixture of the cystic fibrosis gene and the c-erbB-2 oncogene.

Figure 44 show melting peaks at various cycle numbers for the cystic fibrosis gene (CFTR) and c-erbB-2 (neu).

Figure 45 shows a graph of integrated melting peaks of CFTR and neu PCT products.

Figures 46A&B show melting curves and melting peaks for PCR products of a person heterozygous for the factor V Leiden mutation (solid line), homozygous for the factor V Leiden mutation (dotted line), homozygous wild type (broken line), and no DNA control (alternating dot and dash).

I•00"0 Figure 47 shows a fluorescence ratio v. temperature plot of 25 continuous monitoring during cycle 40 of PCR products of a sample homozygous for the factor V Leiden mutation (solid line), heterozygous for the factor V Leiden mutation (dotted line), and homozygous wild type (alternating dot and dash) Figure 48 shows melting peaks of a homozygous mutant of the methylenetatrahydrofolate gene (solid line), homozygous wild type (broken line), heterozygous mutant (dotted line), and no DNA control •o.o (alternating dot and dash) i: Figure 49 shows the shape of reannealing curves of amplified Pglobin PCR products from various initial template amounts.

35 Figure 50 shows the determination of a second order rate constant for determining initial template concentration.

Figure 51 shows a block diagram for controlling thermal cycling from fluorescence data.

Figures 52A&B show a temperature v. time plot acquired after 20 cycles and a fluorescence v. time plot acquired after 25 cycles wherein thermal cycling was controlled from fluorescence data.

Figures 53-105 are programming diagrams used in accordance with the present invention.

DETAILED DESCRIPTION Before the present methods for monitoring hybridization during PCR are disclosed and described, it is to be understood that this invention is not limited to the particular configurations, process steps, and materials disclosed herein as such configurations, process steps, and materials may vary somewhat. It is also to be understood that the terminology employed herein is used for the purpose of describing particular embodiments only and is not intended to be limiting since the scope of the present invention will be limited only by the appended claims and equivalents thereof.

It must be noted that, as used in this specification and the appended claims, the singular forms and "the" include plural referents unless the context clearly dictates otherwise.

In describing and claiming the present invention, the following terminology will be used in accordance with the definitions set out below.

As used herein, "nucleic acid," "DNA," and similar terms also include nucleic acid analogs, i.e. analogs having other than a phosphodiester backbone. For example, the so-called "peptide nucleic acids," which are known in the art and have peptide bonds instead of phosphodiester bonds in the backbone, are considered within the scope of the present invention.

As used herein, "continuous monitoring" and similar terms refer .0oo to monitoring multiple times during a cycle of PCR, preferably during S 25 temperature transitions, and more preferably obtaining at least one data point in each temperature transition.

As used herein, "cycle-by-cycle" monitoring means monitoring the o PCR reaction once each cycle, preferably during the annealing phase of

PCR.

30 As used herein, "fluorescence resonance energy transfer relationship" and similar terms refer to adjacent hybridization of an oligonucleotide labeled with a donor fluorophore and another o. .oligonucleotide labeled with an acceptor fluorophore to a target nucleic acid such that the donor fluorophore can transfer resonance energy to the acceptor fluorophore such that the acceptor fluorophore o. produces a measurable fluorescence emission. If the donor fluorophore o and acceptor fluorophore are spaced apart by too great a distance, then the donor fluorophore cannot transfer resonance energy to the acceptor fluorophore such that the acceptor fluorophore emits measurable fluorescence, and hence the donor fluorophore and acceptor fluorophore are not in resonance energy transfer relationship. Preferably, when the two labeled oligonucleotides are both probes and neither functions as a PCR primer ("probe-probe"), then the donor and acceptor fluorophores are within about 0-25 nucleotides, more preferably within about 0-5 nucleotides, and most preferably within about 0-2 nucleotides. A particularly preferred spacing is 1 nucleotide. When one of the labeled oligonucleotides also functions as a PCR primer ("probe-primer"), then the donor and acceptor fluorophores are preferably within about 0-15 nucleotides and more preferably within about 4-6 nucleotides.

As used herein, "effective amount" means an amount sufficient to produce a selected effect. For example, an effective amount of PCR primers is an amount sufficient to amplify a segment of nucleic acid by PCR provided that a DNA polymerase, buffer, template, and other conditions, including temperature conditions, known in the art to be necessary for practicing PCR are also provided.

PCR requires repetitive template denaturation and primer annealing. These hybridization transitionis are temperature- dependent.

The temperature cycles of PCR that drive amplification alternately denature accumulating product at a high temperature and anneal primers to the product at a lower temperature. The transition temperatures of product denaturation and primer annealing depend primarily on GC content and length. If a probe is designed to hybridize internally to the PCR product, the melting temperature of the probe also depends on GC content, length, and degree of complementarity to the target.

Fluorescence probes compatible with PCR can monitor hybridization during amplification.

25 In accordance with the present invention, which is preferably used in connection with rapid cycling, a kinetic paradigm for PCR is appropriate. Instead of thinking about PCR as three reactions S* (denaturation, annealing, extension) occurring at three different .e temperatures for three time periods (Figure 1A), a kinetic paradigm for PCR is more useful (Figure 1B). With a kinetic paradigm, the :O:temperature vs. time curve consists of continuous transitions between se*@ :overlapping reactions. Denaturation and annealing are so rapid that no goes holding time at a particular temperature is necessary. Extension *•occurs over a range of temperatures at varying rates. A complete analysis would require knowledge of all relevant rate constants over eve* all temperatures. If rate constants of all reactions were known, a S"physicochemical description of PCR" could be developed. Determining these rates would require precise sample temperature control and is greatly simplified by reaction monitoring during temperature cycling.

Figure 2 illustrates useful temperature v. time segments for fluorescence hybridization monitoring. Product melting curves are obtained during a.slow temperature increase to denaturation. By quickly lowering the temperature after denaturation to a constant temperature, product, probe, or primer annealing can optionally be followed. Probe melting curves are obtained by slowly heating through temperatures around the probe Tm. The embodiment represented in Figure 2 provides all analysis during temperature cycling with immediate real time display. Fluorescent probes are included as part of the amplification solution for continuous monitoring of primer, probe, or product hybridization during temperature cycling.

The fluorescence hybridization techniques contained herein are based on rapid cycling, with its advantages in speed and specificity.

A sample temperature profile during rapid cycle PCR is shown in Figure 3. Denaturation and annealing appear as temperature "spikes" on these figures, as opposed to the broad plateaus of conventional temperature cycling for PCR, e.g. Figure 1A. Rapid temperature cycling is contrasted to conventional temperature cycling in Figure 4, wherein it is shown that 30 cycles of amplification can be completed in minutes and the resulting PCR products contain many fewer side products.

Thus, with rapid cycling the required times for amplification are reduced approximately 10-fold, and specificity is improved.

Example 1 Figure 4 shows the results of four different temperature/time profiles and their resultant amplification products after thirty cycles The profiles A and B in Figure 4 were obtained using a prior art heating block device using a prior art microfuge tube. As can be seen in Figure 4, the transitions between temperatures are slow and o° a Oi *e 25 many nonspecific bands are present in profiles A and B. Profile B shows improvement in eliminating some of the nonspecific bands (in contrast to profile A) by limiting the time each sample remains at each temperature, thus indicating that shorter times produce more desirable results.

Profiles C and D were obtained using a rapid temperature cycler.

aloe 30 As can be seen in Figure 4, amplification is specific and, even though O eO yield is maximal with 60-second elongation times it is still entirely adequate with 10-second elongation times (D) The optimal times and temperatures for the amplification of a 536 bp fragment of beta-globin from human genomic DNA were also determined.

35 Amplification yields and product specificity were optimal when oh F denaturation (93C) and annealing (55 0 C) were less than 1 second. No advantage was found to longer denaturation or annealing times. The yield increased with longer elongation times at 77C, but there was little change with elongation times longer than 10-20 seconds. These unexpected results indicate that the previously available devices used for DNA amplification are not maximizing the conditions needed to optimize the physical and enzymatic requirements of the reaction.

Further information can be obtained from: C.T. Wittwer et al., Rapid Cycle Allele-Specific Amplification with Cystic Fibrosis delta F(508) Locus, 39 Clinical Chemistry 804 (1993) and C.T. Wittwer et al., Rapid DNA Amplification, THE POLYMERASE CHAIN REACTION 174 (1994), which are both now incorporated herein by this reference. The instrumentation used for fluorescence acquisition and rapid temperature cycling is fully disclosed in Serial No. 08/537,612, supra.

As indicated earlier, the polymerase chain reaction can be performed rapidly. In addition to facilitating rapid heat transfer, the use of optically clear capillary tubes allows for continuous fluorescence monitoring of DNA amplification in accordance with the present invention.

Fluorescent probes can be used to detect and monitor DNA amplification. Useful probes include double-stranded-DNA-specific dyes and sequence-specific probes. Three different fluorescence techniques for following DNA amplification are compared in Figure 5. In Figure fluorescence depends on the hybridization of PCR product as detected with a double-strand-specific DNA dye. In figure 5B, fluorescence depends on the hydrolysis of a 5'-exonuclease quenching probe, which is well known in the art as discussed above. Figure 5C diagrams a hybridization scheme based on resonance energy transfer between fluorophores on two adjacent probes. The method of Figure 5A is not sequence specific, although product specificity can be determined by melting curves, one aspect of the current invention. Both Figures 25 and 5C are sequence specific. However, the hybridization method also allows analysis with melting curves, another aspect of the current S. invention.

In monitoring fluorescence from reactions involving hydrolysis :probes as in Figure 5B and from reactions involving hybridization probes as in Figure 5C, it is advantageous to measure fluorescence emitted by both the donor fluorophore and the acceptor fluorophore. In practice, the majority of the fluorescence emitted by hydrolysis probes is from the donor fluorophore, and the majority of the fluorescence emitted by hybridization probes is from the acceptor fluorophore.

Double-strand-specific DNA dye selection. Those skilled in the art will be familiar with the use of ethidium bromide in fluorescence techniques. When a double strand-specific fluorescent dye is present o during amplification, fluorescence generally increases as more double stranded product is made, see R. Higuchi et al., Simultaneous amplification and detection of specific DNA sequences, 10 Bio/Technology 413-417 (1992). A fluorescence PCR assay for hepatitis C RNA using the intercalater, YO-PRO-1 is also known in the art. See T. Ishiguro et al., Homogeneous quantitative assay of hepatitis C virus RNA by polymerase chain reaction in the presence of a fluorescent intercalater, 229 Anal. Biochem. 207-213 (1995). It is preferred that SYBRTM Green I, which is well known in the art and available from Molecular Probes of Eugene, Oregon, be used as a double-strand-specific dye. The molecular structure of this dye is a trade secret, but it is recommended by the manufacturer as a more sensitive double-strand-specific dye for DNA detection on gels. SYBRTM Green I is heat labile, however, and thus is not useful for fluorescence monitoring of PCR according to conventional methods where the temperature of the reaction mixture is maintained at melting temperatures for extended periods of time. Because of this heat lability, it was unexpected to discover that SYBRTM Green I can be used to monitor PCR reactions when melting temperatures are not maintained for extended periods, i.e. when PCR is carried out by rapid cycling according to the kinetic paradigm described above.

Example 2 Different double-strand-specific DNA dyes were compared by monitoring the amplification of a 110 base pair fragment from the PCO3/PCO4 primer pair of the human beta-globin gene from 10,000 template copies. Primers were synthesized by standard phosphoramidite chemistry as known in the art, namely, using Pharmacia Biotech Gene Assembler Plus (Piscataway, New Jersey). The human beta-globin primers PC03/PC04 (110 base pairs) are described in C.T. Wittwer et al., Automated polymerase chain reaction in capillary tubes with hot air, 17 Nucl. Acids. Res.

4353-4357 (1989), which is now incorporated herein by reference. DNA 25 amplification was performed in 50 mM Tris, pH 8.5 (25 0 3 mM MgCl 2 500 pg/ml bovine serum albumin, 0.5 pM of each primer, 0.2 mM of each deoxynucleoside triphosphate and 0.2 U of Taq polymerase per 5 pi sample unless otherwise stated in the following examples. Purified amplification product was used as DNA template and was obtained by 30 phenol/chloroform extraction and ethanol precipitation, see D.M.

:2 Wallace, Large- and small-scale phenol extractions and precipitation of nucleic acids, 152 Methods in Enzymology 33-48 (1987), followed by o* removal of primers by repeated washing through a Centricon microconcentrator (Amicon, Danvers, Massachusetts). Template 35 concentrations were determined by absorbance at 260 nm. A(260):A(280) t*oo ratios of templates were greater than 1.7.

SYBRTM Green I (Molecular Probes, Eugene, Oregon) was used at a 1:10,000 dilution, ethidium bromide was at 5 pg/ml, and acridine orange was at 3 pg/ml. These concentrations were determined to be optimal concentrations for maximizing the fluorescence signal observed during amplification for each dye. Excitation was through a 450-490 nm interference filter from a xenon arc source, except for ethidium bromide, where a 520-550 nm excitation was used. For SYBRTM Green I, the emmision at 520-550 was monitored. Ethidium bromide fluorescence was observed through a 580-620 nm bandpass. The acridine orange signal was taken as the ratio of green (520-550 nm) to red (>610 nm) fluorescence. The fluorescence of the sample before amplification was compared to the fluorescence after 35 cycles (94 0 C max, 60 0 C for 20 sec) at 60 0 C. The fluorescence increase was 5.3-fold for SYBRTM Green I, 1.7-fold for ethidium bromide, and 1.2-fold for acridine orange. In separate experiments, the fluorescence from SYBRTM Green I was stable for greater than 30 min at 70 0 C. It is also conveniently excited with visable light and is claimed to be less of a mutagen than ethidium bromide. Background fluorescence in all cases arose primarily from the primers.

SYBRTM Green I is a preferred double-strand-specific dye for fluorescence monitoring of PCR, primarily because of superior sensitivity, arising from greater discrimination between double stranded and single stranded nucleic acid. SYBRTM Green I can be used in any amplification and is inexpensive. In addition, product specificity can be obtained by analysis of melting curves, as will be described momentarily.

Resonance energy transfer dye selection for hybridization probes.

Fluorescence resonance energy transfer can occur between 2 fluorophores if they are in physical proximity and the emission spectrum of one S 25 fluorophore overlaps the excitation spectrum of the other. Introductory theory on fluorescence resonance energy transfer can be found in many recent review articles. The rate of resonance energy transfer is: (8.785E-5) (t- 1 (n- 4 (qD) (R- 6 (JDA), where: t excited state lifetime of the donor in the absence of the acceptor; k 2 is an orientation factor between the donor and acceptor; oo" n refractive index of visible light in the intervening medium; qD quantum efficiency of the donor in the absence of the acceptor; R distance between the donor and acceptor (in angstroms); JDA the integral of (FD) (eA) (W 4 with respect to W at all overlapping wavelengths with: FD peak normalized fluorescence spectrum of the donor, eA molar absorption coefficient of the acceptor (M- 1 cm-1), and W wavelength (nm).

For any given donor and acceptor, a distance where 50% resonance energy transfer occurs can be calculated and is abbreviated R 0 Because the rate of resonance energy transfer depends on the 6th power of the distance between donor and acceptor, resonance energy transfer changes rapidly as R varies from R 0 At 2R 0 very little resonance energy transfer occurs, and at 0.5R 0 the efficiency of transfer is nearly complete, unless other forms of de-excitation predominate collisional quenching). R 0 values for many different donor and acceptor pairs have been compiled and vary between 22 and 72 angstroms.

In double helical DNA, 10 bases are separated by about 34 angstroms. By labeling the bases of DNA with donor and acceptor fluorophores, resonance energy transfer has be used as a spectroscopic ruler to observe the helical geometry of DNA and analyze the structure of a four-way DNA junction. Resonance energy transfer can also be used as a monitor of hybridization. If a labeled oligonucleotide is hybridized to a labeled template strand, R can be brought from much greater than R 0 to well below R 0 increasing resonance energy transfer dramatically. Alternately, 2 labeled probes can be hybridized to the same template strand for a similar change in fluorescence energy transfer.

The practical use of resonance energy transfer to monitor hybridization depends on the sensitivity required and how much time is available. Using a competitive hybridization technique with 1 nM labeled probes, PCR-amplified DNA was detected after 15 min at 40 0

C.

25 Faster signal generation is desirable. If only seconds were required for hybridization, PCR products could conveniently be quantified each S° cycle of amplification. Even further, the extent of probe hybridization could be monitored within a temperature cycle.

o Hybridization is a second order process (see B. Young M.

Anderson, Quantitative analysis of solution hybridization, In: Nucleic Acid Hybridization: A Practical Approach 47-71, Hames, S. Higgins eds., 1985). When the concentration of the probe is much greater than ooo° the concentration of the target, the hybridization rate is inversely proportional to concentration of probe. For example, by doubling the probe concentration, the hybridization time is cut in half. High probe concentrations would be necessary for cycle-by-cycle monitoring during .PCR, because hybridization must occur before the hybridization site is covered by polymerase extension.

The high probe concentrations required for hybridization monitoring during PCR require a resonance energy transfer pair with unique characteristics. Consider excitation of a donor and an acceptor pair with light. The number of fluorophores of D and A directly excited will be proportional to the extinction coefficient (e) of each fluorophore at the excitation wavelength, or: Number of D molecules directly excited Number of A molecules directly excited (K)(eA) where K is a proportionality constant. De-excitation of the donor will occur by fluorescence, resonance energy transfer, and other processes summarized as thermal quenching. If PF probability of resonance energy transfer, and PTD probability of donor thermal quenching, then the probability of donor fluorescence is:

I-PF-PTD

and the number of fluorescing donor molecules is: (eD) (I-PF-PTD) If the probability of detecting a donor emission in the donor emission window (for example, a bandpass filter window) is pDD, then the number of observed donor emissions is: (PDD) (eD) (1-PF-PTD) Now, the number of excited acceptor fluorophores is the sum of those directly excited and those excited through resonance energy transfer: (eA) (eD) (PF) 20 If PTA the probability of thermal quenching of the acceptor, then the probability of acceptor fluorescence is: -pTA and the number of fluorescing acceptor molecules is: (eA) (eD) (pF) If the probability of detecting an acceptor emission in the acceptor emission window is pAA, then the number of observed acceptor emissions is: (PAA) (eA) (e (pF) (PTA)] Finally, if the probability of observing a donor emission in the acceptor emission window is PDA, then the number of observed emissions (both D and A) in the acceptor emission window is: 16 (PA) (eA) (eD) (PDA) (eD) (-PF-PTD) Since fluorescence measurements are relative and K is present in all terms, if we remove K and rearrange, the observed intensity at the donor window is proportional to (donor excitation) (energy transfer): 1) (eD) (PDD) (I-PTD) (eD) (PDD) (PF) and the observed intensity at the acceptor window is proportional to (acceptor excitation) (energy transfer) donor emission in the acceptor window): 2) (eA) (PAA) TA) (eD) (PDD) (PF) (I-PTA) (eD) (PDA) (-PTD-PF) As resonance energy transfer increases, the donor signal decreases and the acceptor signal increases. The percent signal change depends on the background light intensity in each window. With high concentrations of probes, this background light intensity is high. During PCR, when varying target (product) concentrations need to be monitored, it is not possible to match the donor concentration to the target concentration.

The excess donor molecules contribute to the background light intensity in both the donor and acceptor windows and partially mask the energy transfer phenomena. Background light in the acceptor window comes from not only donor emission in the acceptor window, but also from direct 20 excitation of the acceptor. This background can be minimized with a low eA and a low pDA.

The fluorescein/rhodamine fluorescence energy transfer pair, commonly used for nucleic acid detection, has high background fluorescence. Both direct acceptor excitation ca. 10% emx) and 25 emission of the donor at wavelengths used to detect acceptor emission (PDA, ca. 20% peak emission) are high. This pair can be used to monitor hybridization if the probe concentration is near to the target concentration and enough time is allowed for complete hybridization. It o.o. is not a useful pair of fluorophores for continuous monitoring of PCR 30 because high probe concentrations are required and the template concentration in PCR is continually changing.

Monitoring product concentration during PCR by hybridization has not been possible in the past because an acceptable resonance energy transfer pair had not been found. There have been few attempts to use resonance energy transfer for direct "noncompetitive" detection of hybridization. For example, U.S. Patent No. 5,565,322 states "the observed energy transfer efficiency in terms of re-emission by the acceptor was relatively low." At probe concentrations that are high enough for significant hybridization to occur in seconds, the background fluorescence is too high.

Fluorescein is perhaps the most widely used fluorophore. Its extinction coefficient and quantum efficiency are high and it is extensively used in microscopy, immunoassays, and flow cytometry. It is the donor in a commonly used resonance energy transfer pair with rhodamine. Cy5 is a popular red-emitting fluorophore with a very high extinction coefficient. The structure of the N-hydroxysuccinimide ester of Cy5 is shown in Figure 6, and the structure of the related dye, Cy5.5, is shown in Figure 7. These dyes are indodicarbocyanine dyes that are used commonly in flow cytometry and automated fluorescence sequencers and are available from Amersham (Pittsburg, PA). Both fluorescein and Cy5 are commercially available as amidites for direct, automated incorporation into oligonucleotides. However, Cy5 has never been reported as a resonance energy transfer pair with fluorescein.

Intuitively, fluorescein emission and Cy5 absorption do not overlap enough for resonance energy transfer to be considered. The emission spectrum of fluorescein and absorption spectrum of Cy5 attached to oligonucleotides are shown in Figure 8. When the areas under the curves are normalized, the overlap from the technical spectra is 19%. excitation is shifted to the red by about 25 nm, further decreasing the overlap with fluorescein emission to about 15%. Working in the red/infrared region of the spectrum is advantageous when choosing optical components for instrumentation. Laser diodes can be used for 25 illumination, photodiode detectors have excellent sensitivity, and most materials have minimal autofluorescence in the pertinent spectral 0region.

Despite low spectral overlap, it has been discovered that fluorescein and either Cy5 or Cy5.5 make an excellent resonance energy transfer pair for hybridization monitoring during PCR.

Example 3 SA 110 bp beta-globin fragment was amplified from 50 ng human genomic DNA according to the procedure of Example 2 with the internal probes CAAACAGACA CCATGGTGCA CCTGACTCCT GAGGA-fluorescein (SEQ ID NO:3) 35 and Cy5-GAAGTCTGCC GTTACTGCCC TGTGGGGCAA G-p (SEQ ID NO:18) at 0.2 pM each and 0.8 U KlenTaql polymerase (a 5'-exonuclease deficient variant of Taq polymerase U.S. Patent No. 5,436, 149) in a 10 p1 reaction.

The probes hybridized internal to the primers on the same strand and were immediately adjacent without any intervening bases.

Probes and primers were synthesized by standard phosphoramidite chemistry as known in the art, using a Pharmacia Biotech Gene Assembler Plus (Piscataway, New Jersey). The fluorescein-labeled probe was synthesized on a fluorescein-labeled CPG cassette (Glen Research, Sterling, VA) with the final trityl-ON to assist with C18 reverse phase HPLC purification. The late eluting peak was collected and the trityl group was removed on a PolyPack (Glen Research). The fluorescein-labeled oligo was eluted with 50% acetonitrile and again purified by C18 reverse phase HPLC. The 5'-Cy5-labeled probe was synthesized with a chemical phosphorylation agent on the 3'-end (Glen Research) and adding a Cy5 amidite (Pharmacia) to the 5'-end during trityl-OFF synthesis. Failure sequences were removed by C18 reverse phase HPLC. Probe purity was checked with polyacrylamide electrophoresis and the absorbance of the dye and the oligo.

HPLC was perfomed on a 4 x 250 mm Hypersil ODS column (Hewlett Packard) with a 0.1 M triethanolamine:acetate mobile phase and an acetonitrile gradient at 1 ml/min. The eluate was monitored for both absorbance (A 260 and fluorescence (490 nm excitation, 520 nm emission for fluorescein and 650 nm excitation, 670 nm emission for Tritylated- and fluorescein-labeled oligonucleotides were eluted with a 10-20% acetonitrile gradient, and Cy5-labeled oligonucleotides eluted over a 10-40% acetonitrile gradient.

Temperature cycling was 94°C for 0 sec with a programmed approach rate of 20 0 C/sec, 60 0 C for 20 sec with an approach rate of 20 0 C/sec, and 750C for 0 sec with an approach rate of l°C/sec in a capillary fluorescence rapid temperature cycler. During temperature cycling, fluorescein and Cy5 fluorescence were acquired each cycle at the end of S 25 the annealing/extension segment. Resonance energy transfer was observed as both a decrease in fluorescein fluorescence, and an increase in fluorescence beginning around cycle 26 of amplification (Figure In general, observing the fluorescence ratio of Cy5 to fluorescein fluorescence is preferred.

The unexpectedly good results with the fluorescein/Cy5 pair can at least partly be rationalized. The overlap integral, JDA depends not Sonly on spectral overlap, but also on the extinction coefficient of the acceptor (Cy5 has an extinction coefficient of 250,000 M- 1 cm' 1 at 650 and on the 4th power of the wavelength. Both of these factors will favor a high JDA for Cy5, even given low spectral overlap. Recently, phycoerythrin and Cy7 were shown to be an effective tandem probe for immunofluorescence, despite low spectral overlap. In a later example, the utility of fluorescein and Cy5.5 as labels on hybridization probes is demonstrated. Fluorescence resonance energy transfer can be used to monitor nucleic acid hybridization even when the interacting dyes have low spectral overlap. The use of fluorescein with Cy5, Cy5.5 and other red or infrared emitting dyes as resonance energy transfer pairs for monitoring hybridization has not been previously recognized.

19 Fluorescein has a long emission "tail" that goes out to 600 nm, 700 nm and beyond that can be used to excite these far red and infrared dyes.

The rate of energy transfer is dependent on the overlap integral, but is also effected by the 6th power of the distance between the fluorophores.

If the probes are designed so that the resonance energy transfer dyes are in close proximity, the transfer rate is high. At least with fluorescein/Cy5.5 and like pairs, resonance energy transfer appears to predominate over collisional quenching and other forms of energy loss when the fluorophores are close together, as in the above example where the fluorophores are attached to adjacent probes with no intervening bases.

The potential usefulness of a resonance energy transfer pair for hybridization probes can be judged by observing the change in the ratio of light intensity in the donor and acceptor windows at minimal and maximal resonance energy transfer. One way to obtain minimal and maximal transfer is to attach both fluorophores to the same oligonucleotide and measure fluorescence ratio before and after digestion with phospodiesterase.

Example 4 The dual-labeled fluorescein/Cy5 probe GCCCTGTGGG GCAAGGp (SEQ ID NO:19) was synthesized by standard phosphoramidite chemistry, where p is a terminal 3'-phosphate (chemical phosphorylation reagent, Glen Research), F is a fluorescein residue -introduced as an amidite with a 2-aminobutyl-l,3-propanediol backbone to S25 maintain the natural 3-carbon internucleotide phosphodiester distance (ClonTech, Palo Alto, CA), and Cy5 is added as the amidite (Pharmacia) The ratio of Cy5 to fluorescein fluorescence in 0.1 M Tris, pH 8.0 was obtained before and after exhastive hydrolysis with phosphodiesterase S (Sigma, St. Louis, MO). The change in the fluorescence ratio was 220-fold after hydrolysis. A dual-labeled fluorescein/rhodamine probe S: F-ATGCCCT*CCC CCATGCCATC CTGCGTp (SEQ ID NO:20) was purchased from o. Perkin Elmer (Foster City, CA), where F is fluorescein and is a rhodamine attached to a modified T residue by an amino-linker arm. The change in the fluorescence ratio (rhodamine to fluorescein) was 22-fold after hydrolysis with phosphodiesterase.

The potential signal from the fluorescein/Cy5 pair was that of the fluorescein/rhodamine pair.

Example The effect of the ratio, concentration, and spacing of fluorescein and Cy5-labeled adjacent hybridization probes during PCR was studied.

Amplification of the beta globin locus and probe pair of Example 3 was used and the maximum change in the fluorescence ratio of Cy5 to fluorescein was observed. The maximal signal occurred when the ratio of to fluorescein-labeled probes was 2:1 (Figure 10). At this 2:1 ratio, the best signal occurred at a fluorescein probe concentration of 0.2 uM and a Cy5-labeled probe concentration of 0.4 pM (Figure 11). The optimal number of intervening bases between adjacent hybridization probes during PCR was also determined. Several probes of the same length but slightly shifted in their hybridization position were synthesized according to Example 3 so that when they hybridized to the beta globin target, 0, 1, 2, 3, 4, or 6 bases remained between the probes. The highest signal during PCR occurred with one intervening base (Figure 12). Although some resonance energy transfer was detected at a spacing of 15 and even 25 bases, much better transfer occurred at bases.

Heller et al. Patent No. 4,996,143), found that energy transfer efficiency decreased as the number of nucleotides between fluorophores decreased from 4 to 0 units. In contrast, the best energy transfer with the fluorescein/Cy5 pair was seen at 0 to 2 intervening nucleotides.

Hybridization probe method. If 2 probes are synthesized that hybridize adjacently on a target and each is labeled with one fluorophore of a resonance energy transfer pair, the resonance energy transfer increases when hybridization occurs (Figure 5C) The fluorescein/rhodamine pair is most commonly used for nucleic acid detection.

One aspect of this invention is to provide a sequence-specific S* 25 homogeneous hybridization method for detection of PCR products. It is not obvious how to achieve this. Using hybridization probes during S* amplification is counterintuitive. It does not seem that both probe hybridization and polymerase extension can occur. To get sequence specific fluorescence, the probes must be hybridized, but the probes cannot be hybridized if the polymerase is to complete primer extension and exponentially amplify DNA.

One solution to this problem is to use a dual-labeled single probe and utilize the 5'-exonuclease activity of common heat stable DNA polymerases to cleave the probe during extension, thereby separating the 35 2 fluorophores. In this case, the fluorescence signal arises from separation of the resonance energy transfer pair upon probe hydrolysis (Figure 5B), rather than approximation of the fluorophores by adjacent hybridization (Figure 5C). However, dual-labeled probes are difficult to make, requiring manual addition of at least one fluorophore to the oligo and usually require extensive purification. The probes are expensive, and two dual-labeled probes are necessary for competitive quantification of a target or for mutation detection. A further concern is that the observed fluorescence depends on the cumulative amount of probe hydrolyzed, not directly on the amount of product present at any given cycle. This results in a continued increase in fluorescence even after the PCR plateau has been reached. Finally and most importantly, probe hydrolysis does not always occur during polymerase extension, an effect that is not well understood. For example, the dual-labeled probe of Example 4 showed very poor hydrolysis during PCR when it was flanked by primers. Indeed, several dual-labeled probes, including those with terminal labels, were made and all showed poor hydrolysis and signal generation during amplification.

Homogeneous detection of PCR products with adjacent hybridization probes would solve many of the problems of the 5'-exonuclease system.

Synthesis of adjacent hybridization probes is relatively simple because amidites for both fluorescein and Cy5 are available for direct incorporation during automated synthesis and dual labeling of one probe is not required. Because their fluorescence results from hybridization, not hydrolysis, the temperature dependence of probe fluorescence could be used for mutation detection and quantification. However, use of adjacent hybridization probes for homogeneous detection of PCR products has not been reported previously. Surprisingly, both hybridization for signal generation and amplification by polymerase extension through the area blocked by the probes can occur.

Example 6 A 110 bp beta-globin fragment was amplified from genomic DNA with Sadjacent fluorescein- and Cy5-labeled probes as described in Example 3.

Either 0.4 U (Taq) or 0.8 U (Stoffel fragment, Perkin Elmer, or KlenTaql) of enzyme was used in 10 pl reactions. Unless indicated otherwise, temperature cycling was 94C for 0 sec with a programmed approach rate of 20 0 C/sec, 60 0 C for 20 sec with an approach rate of 20 0 C/sec, and 75 0 C for 0 sec with an approach rate of l°C/sec. Figure 13 Se shows the development of fluorescence by 2 adjacent hybridization probes S: immediately after the template was amplified for 30 cycles. After a brief denaturation at 94 0 C, the temperature was lowered to 600C and 35 fluorescence increased for about 20 sec. The magnitude of the signal is greater with an exonuclease deficient polymerase (Stoffel fragment) than Swith native Taq polymerase that includes a 5'-exonuclease activity.

After about 20 sec., the fluorescence drops as the polymerase displaces and/or hydrolyzes the probes. The relative decrease in fluorescence is slightly faster when the polymerase has 5'-exonuclease activity (Taq DNA polymerase) then when it lacks this activity (Stoffel fragment).

In Figure 14 (top panel), the temperature is cycled between 94°C and 600C with a 20 sec hold at 600C. Fluorescence is acquired at the end of the 20 sec when fluorescence is maximal. Good amplification occurs with Taq but not with Stoffel fragment (exo-) as verified by both fluorescence development and agarose gels (gels not shown).

However, if the time at 600C is increased from 20 sec to 120 sec (Figure 14, middle panel), the exo- polymerase amplifies well. The slower rate of probe displacement with an exo- polymerase apparently requires more time at 600C for efficient amplification than the exo' polymerase. The time required by exo- polymerases can be reduced by slowly increasing the temperature from 600C to 750C (Figure 14, bottom panel). The polymerase stalls when it reaches the probe. However, at the probe melting temperatures, the probes melt off the template and the polymerase continues unencumbered to complete polymerization of the strand. Polymerization is completed as long as the temperature is not raised too quickly after probe melting. Figure 14 (bottom panel) shows one exo' polymerase (Taq) and two exo- polymerases (Stoffel fragment and KlenTaql).

When exonuclease activity is present, some of the probe is hydrolyzed each cycle as evidenced by an the decrease in fluorescence with extensive amplification. This is observed in Figures 13 and 14 (middle and bottom panels), but is does not occur with exo- polymerases.

Because the fluorescence is stable on extensive amplification, exo polymerases such as KlenTaql are preferred.

25 The success of using adjacent hybridization probes to monitor PCR depends on several factors. Resonance energy transfer is maximized when there is either 0 to 2 intervening bases between adjacent hybridization S* probes. To increase the fraction of strands that hybridize to the probes before the primer extends through the area of probe 30 hybridization, the probe melting temperatures should be greater than the oooo primer melting temperatures (preferably Cycle-by-cycle fluorescence. Conventional endpoint analysis of DNA amplification by gel electrophoresis identifies product size and o.0. estimates purity. However, because amplification is at first 35 stochastic, then exponential, and finally stagnant, the utility of endpoint analysis is limited for quantification. One aspect of the present invention includes cycle-by-cycle monitoring for quantification of initial template copy number with hybridization probes. As will be appreciated by those skilled in the art, once-per-cycle monitoring of multiple samples undergoing DNA amplification is a powerful quantitative tool. Cycle-by-cycle monitoring is achieved by acquiring fluorescence 23 during the extension or combined annealing/extension phase of each cycle and relating the fluorescence to product concentration.

Example 7 Cycle-by-cycle monitoring of PCR was performed by three different fluorescence techniques. Fluorescence was monitored by the double-strand-specific dye SYBRTM Green I, (ii) a decrease in fluorescein quenching by rhodamine after exonuclease cleavage of a dual-labeled hydrolysis probe and (iii) resonance energy transfer of fluorescein to Cy5 by adjacent hybridization probes. Amplification reagents and conditions were as described in Example 2. The human beta-globin primers RS42/KM29 (536 base pairs) and PC03/PC04 (110 base pairs) are described in C.T. Wittwer et al., Automated polymerase chain reaction in capillary tubes with hot air, 17 Nucl. Acids. Res.

4353-4357 (1989), which is now incorporated herein by reference.

Temperature cycling for beta-globin was 95 0 C maximum, 61 0 C minimum, sec at 72 0 C and an average rate between temperatures of 5.2 0 C/sec. The beta-actin primers and fluorescein/rhodamine dual probe were obtained from Perkin Elmer (no. N808-0230). Temperature cycling for beta-actin was 94°C maximum, 60 0 C for 15 sec with an average rate between temperatures of 6.2 0 C/sec. The single labeled probes CCATGGTGCA CCTGACTCCT GAGGA-fluorescein-3' (SEQ ID NO:3) and 5'-Cy5-AAGTCTGCCG TTACTGCCCT GTGGGGCAAGp (SEQ ID NO:4) were synthesized as in Example 3. These adjacent probes hybridize internal to the PC03/PC04 beta-globin primer pair on the same DNA strand and are 0 25 separated by one base pair. Temperature cycling was 94 0 C maximum, 59 0

C

for 20 sec with an average rate between temperatures of 7.0 0 C/sec.

Hybridization probes (beta-actin and beta-globin) were used at 0.2 pM each.

When multiple samples are monitored once each cycle with SYBRTM 30 Green I, a 10'-108 range of initial template concentration can be discerned as represented in Figure 15. This amplification is of a 536 0 base pair fragment of the beta-globin gene, with SYBRTM Green I as the double-strand specific dye. When the data were normalized as the .00. percent maximal fluorescence of each sample, one hundred initial copies 35 were clearly separated from ten copies. However, the difference between one and ten copies was marginal, and no difference was observed between zero and one average copies per sample.

In contrast, sequence-specific probes have a similar dynamic range but, appear to discriminate even a single initial template copy from negative controls. Signal generation with 5'-exonuclease probes (beta-actin fragment, Figure 16) is dependent not only on DNA synthesis, 24 but requires hybridization and hydrolysis between the fluorophores of the dual-labeled probe. This hydrolysis reduces quenching and the fluorescence ratio of fluorescein to rhodamine emission increases.

Whereas the fluorescence from double strand dyes levels off with excess cycling (Figure 15), the signal from exonuclease probes continues to increase with each cycle (Figure 16). Even though no net product is being synthesized, probe hybridization and hydrolysis continue to occur.

As the template copy number decreases below 103, signal intensity decreases, but low copy numbers can still be quantified because the negative control signal is stable.

In Figure 17, amplification is monitored using adjacent hybridization probes and is expressed as a ratio of Cy5 to fluorescein fluorescence. The change in fluorescence ratio is largely due to an increase in Cy5 fluorescence from resonance energy transfer (Figure 9).

In contrast to dual-labeled hydrolysis probes, the fluorescence signal of hybridization probes decreases at high cycle numbers if the polymerase contains an exonuclease activity (see also Figure 14).

The present invention's feasibility using two different methods for resonance energy transfer detection of hybridization during PCR will now be demonstrated. The first method uses two adjacent hybridization probes, one labeled 3' with fluorescein and the other labeled 5' with As product accumulates during PCR, the probes hybridize next to each other during the annealing segment of each cycle. The second method uses a primer labeled with Cy5 and a single hybridization probe.

The labeled primer is incorporated into the PCR product during amplification and only a single hybridization is necessary.

S

Example 8

S.

Cycle-by-cycle monitoring of PCR was performed by resonance energy transfer between a Cy5-labeled primer and a fluorescein-labeled 30 hybridization probe. This was compared to monitoring with adjacent Cy5/fluorescein hybridization probes. The Cy5-labeled primer was CAACTTCATC CACGT*TCACC (SEQ ID NO:21) where T* is a modified T base with attached and the corresponding probe was GTCTGCCGTT ACTGCCCTGT GGGGCAA-fluorescein (SEQ ID NO:22). The adjacent hybridization probes 35 were CCTCAAACAG ACACCATGGT GCACCTGACT CC-fluorescein (SEQ ID NO:23) and GTTACTGCCC TGTGGGGCAAp (SEQ ID NO:24). The hybridizaton probes were synthesized according to Example 3 and used at 0.2 pM. The primer was synthesized in two steps. Automated synthesis was used to incorporate an amino-modifier C6dT (Glen Research) at the desired T position. Then, the monovalent N-hydroxysuccinimide ester of (Figure 6) was manually conjugated to the amino linker according to the manufacturer's instructions (Amersham). HPLC purifification was as described in Example 3.

The Cy5-labeled primer (0.5 pM) was used instead of PC04 to amplify the 110 base pair beta-globin fragent from human genomic DNA as in Example 3, except that 0.4 U of Taq polymerase was used per 10 pl.

The adjacent hybridization probes also monitored amplification of the same beta-globin fragment. Temperature cycling was done at 940C for 0 sec and 600C for 20 sec. The fluorescence was monitored once each cycle at the end of the annealing/extension segment. In both methods, fluorescence energy transfer to Cy5 increases with hybridization and is plotted as a ratio of Cy5 to fluorescein fluorescence (Figure 18).

In additional experiments, the number of bases separating the and the fluorescein label were varied. The best fluorescence resonance energy transfer was observed with about 4-6 bases between the fluorophores, although a signal was detectable up to at least 15 bases.

In contrast to hydrolysis probes, the fluorescence signal from hybridization probes is not cumulative and develops anew during each annealing phase. The fluorescence is a direct measure of product concentration because the hybridization is a pseudo-first order reaction. Because the concentration of probe is much greater than the product, the fraction of product hybridized to probe is independent of product concentration. These characteristics indicate that using a single hybridization probe along with a labeled primer will provide a superior monitor of product accumulation for quantification. The 25 inherent variance of different fluorescence techniques during •cycle-by-cycle monitoring is also important for quantification.

Example 9 DNA amplification was perfomed according to Example 2 for each of three different fluorescence monitoring methods. SYBRTM Green I was 30 used at a 1:10,000 dilution in the amplification of a 205 base pair human beta-globin fragment from primers KM29 and PC04. The hydrolysis g. probe and conditions are those specified in Example 7. The hybridization probe, TCTGCCGTTA CTGCCCTGTG GGGCAAG-fluorescein (SEQ ID was used with KM29 and the Cy5-labeled primer 35 CAACTTCATCCACGTT*CACC (SEQ ID NO:6) where T* was a Cy5-labeled T base synthesized as in example 8. All amplifications were performed in ten replicates with 15,000 template copies (50 ng of human genomic pl). The temperature cycles were 31 sec long (940C maximum, 600C for sec, average rate between temperatures 6.2 0 C/sec). Fluorescence was acquired for each sample between seconds 15 and 20 of the annealing/extension phase.

Figure 19 allows comparison of three fluorescence monitoring techniques for PCR. The fluorescence probes are the dsDNA dye SYBRTM Green I (Figure 19A), a dual-labeled fluorescein/rhodamine hydrolysis probe (Figure 19B), and a fluorescein-labeled hybridization probe with a Cy5-labeled primer (Figure 19C) All probes had nearly the same sensitivity with detectable fluorescence occurring around cycle With extended amplification, the signal continued to increase with the hydrolysis probe, was level with SYBRTM Green I, and slightly decreased with the hybridization probe. The precision of the three fluorescence monitoring techniques are compared in Figure 19D. The mean standard deviations are plotted for each point. The data are plotted as the coefficient of variation (standard deviation/mean) of the fluorescence ratio above baseline (taken as the average of cycles 11-15) Although the change in fluorescence ratio from the hydrolysis probe is greater than that from the hybridization probe (Figures 19B and 19C), the coefficient of variation of fluorescence from the hydrolysis probes is greater (Figure 19D). That is, the fluorescence resulting from the hybridization probe method is more precise than using a hydrolysis probe, even though the absolute signal levels are lower.

This is an unexpected advantage of hybridization probes over the more usual dual-labeled hydrolysis probes.

Quantification of initial template copy number. Quantitative PCR me..

has become an important technique in both biomedical research and in the clinical laboratory. The process of quantification often includes running a standard curve of samples containing known copy numbers of the S target sequence. The copy number of an unknown sample is determined by eo extrapolation between the known values. When a complete amplification curve is monitored cycle-by-cycle using fluorescence, radioactivity or any other method that gives a signal proportional to the amount of DNA present, many data points are available for analysis and it is not •Ill obvious which value to choose to represent a standard or unknown..

me Prior art is to choose a "threshold value" of the signal and then use coo• the cycle number when the standard or unknown crosses that threshold as the representative value (see Higuchi Watson, EPA 0 640 828 Al). This o approach uses a very small amount of the available data in an amplification curve. In addition, the assignment of the threshold value is highly subjective and is subject to conscious or unconscious bias.

More of the available data could be used objectively by applying non-linear curve fitting techniques to the data in an amplification curve. Preferably, equations could be found that describe the shape of the amplification curves by modeling factors of the underlying process.

A number of different equations could be used to fit the data produced during amplification. DNA amplifications typically have a log linear segment and the data in this segment can be fit to an equation that describes an exponential increase like that expected in a DNA amplification. The log-linear portion of a DNA amplification can be described by the equation: y A*[DNA]*(l+E)n wherein A is a scaling factor that converts units of signal to units of DNA; [DNA] is the starting concentration of DNA in the reaction; E is the efficiency of the reaction; and n is the cycle number.

A quantification process would involve: fitting the known standards to this equation allowing the parameters A and E to float, and fitting the unknown samples to the equation using the values of A and E from the standards and allowing [DNA] to float. This technique uses much more of the data and uses the portion of the data, the log-linear portion, that is likely to be most informative. Figures 21 and 22 show an example of this approach. Ten-fold dilutions of a purified PCR product were amplified as a standard curve and an "unknown" human genomic DNA standard was used. Figure 20 shows that the log-linear portion is easily identified either by the user or by software. Figure 21 shows a fit of the equation y=A*[DNA]*(l+E) n to the 104 copy standard. Figure 22 uses average values from several standards for A and E and fits [DNA]. The fit value of 16,700 is very close to the theoretical value for a single copy gene in genomic DNA (15,000 25 copies).

oUsing all the data in an amplification curve would include the background level and the plateau value. While at high copy number the

OS

plateau is uninformative, at low copy number it is often proportional to starting copy number. The background level could be useful in determining the first point that shows a significant increase in signal.

se At this time all the factors involved in the shape of the DNA oo amplification curve are not known, so one approach is to describe the oo o shape of the curve. Figure 23 shows amplification curves using °fluorescent hybridization probes to detect a five order of magnitude 55.5 range of DNA template concentrations. Each curve is fit to the oes equation: 0 O: wherein "as" is the background of the slope line, "ab" is the y intercept of the background line, "ds" is the slope of the plateau line, "db" is the y intercept of the slope line, is cycle number where the reaction is half way from background to plateau (A 50 and is the slope of the log-linear portion of the amplification.

This equation gives good fits to this amplification data, and Figure 24 shows that the value of the A 50 correlates well with the log of the starting copy number across seven orders of magnitude. Figure shows the same equation fit to data from amplifications that used a hydrolysis probe to detect DNA template over a 5 order of magnitude range. This equation gives good fits to this amplification data, and Figure 26 shows that the value of the A 50 correlates well with the log of the starting copy number. This demonstrates the flexibility of the full curve fit approach as the equation has given good fits to both the sharp plateaus of the hybridization probe amplification curves and the steadily increaseing "plateaus" of the hydrolysis probe curves.

Total curve fits are not limited to this equation. Figure 27 shows an amplification of three concentrations of DNA template fit to the equation: y=(((as*x+ab)-(dmax*x/dd+x)) which is similar to the first 6 parameter equations except that the plateau is defined by a hyperbolic curve rather than by a line. Figure 28 shows that the A 50 for this equation correlates well to the starting copy number.

While the A 50 has been used in these examples and level between the background and the plateau could be chosen if a particular technique is more robust lower or higher in the amplification profile. For example a series of amplification standard curves are evaluated for the best correlation between the starting copy number and the A 50 the A 40 the A 30 the A 20 and the A 10 The level of amplification that best correlates with the known starting copy number is determined. This will e oo be different for different detections systems. Figure 19 shows that coefficient of variation for various detection systems. The level of amplification that is the best predictor is likely to be the level with 30 the lowest coefficient of variation. go As the DNA amplification reaction itself is better understood, e other equations that have parameters that reflect physical processes could be used. The plateau of the DNA amplification curve has different ecauses in different reactions. It is often due to the inability of the ee"e* 35 primers to compete with product reannealing in the latter cycles. This effect could be captured with a parameter that is dependent on the square of the concentration of product in the reaction (as reannealing rate is proportional to the square of the product concentration) Another cause of the plateau can be the depletion of the primers.

40 Primer limited reactions have a characteristic shape, they have a very sharp plateau that can be recognized. Primer limited reaction fits will include parameters that define this sharp top. Enzyme limited reactions have a very rounded plateau that can be fit accordingly. Weighting factors can be devised that reflect the known coefficients of variation for the given system to more heavily weight the more reliable data points. By fitting more of the points in an amplification profile, more accurate and robust estimates of starting copy number can be obtained.

One or more of the parameters of these fits can be used to estimate the starting copy number of unknown samples.

Continuous fluorescence monitoring of PCR. The present invention's feature of continuous monitoring, that is, monitoring many times within each PCR cycle, will now be discussed. While fluorescence monitoring during PCR can be done once each cycle at a constant temperature, the present invention provides the important advantage of providing continuous monitoring throughout the PCR cycle. Temperature is important because fluorescence changes as the temperature changes.

Figures 29A&B demonstrate the inverse relationship between temperature and fluorescence for SYBTM Green I. This is a confounding effect during temperature cycling that is usually eliminated by considering fluorescence once per cycle at a constant extension temperature.

However, in accordance with the present invention, monitoring fluorescence during temperature changes is very informative. Prior to the present invention, continuous fluorescence monitoring within each cycle, as opposed to once each cycle, has not been carried out. In accordance with the present invention, time, temperature and fluorescence is acquired every sec, every 200 msec, every 100 msec or even at a greater frequency. Such data can reveal fine details of product denaturation, reannealing and extension, and probe annealing and melting during rapid cycling not available in previously available methods.

Example i A 180-base-pair fragment of the hepatitis B surface antigen gene 30 was amplified from 106 copies of purified PCR product using primers S" 5'-CGTGGTGGAC TTCTCTCAAT-3' (SEQ ID NO:l), and GCATAGCAGC-3' (SEQ ID NO:2)(Genbank sequence HVHEPB). The amplification o:o conditions of Example 2 were followed except that the reaction contained eoo a 1:20,000 dilution of SYBRTM Green I and 2 mM MgCl 2 Each temperature o::e 35 cycle was 27 sec long (92 0 C maximum, 59 0 C minimum, 5 sec at 70 0 C, average ""rate between temperatures 3.0 0 C/sec) Time, temperature, and 2 channels of fluorescence were acquired every 200 msec and continuously displayed *oooo as fluorescence v. cycle number and fluorescence v. temperature plots.

e Figure 30 shows a 3D trace of temperature, time and fluorescence for cycles 10 through 34. This 3D curve is also projected in Figure 30 as 2D plots of temperature v. time, fluorescence v. time, and fluorescence v. temperature. The temperature v. time projection of Figure repeats each cycle and provides essentially the same information as set forth in Figure 3. Because fluorescence varies inversely with temperature, the fluorescence v. time projection shown in Figure 30 at early cycles is a scaled mirror image of the temperature v. time plot (see Figure 29). As product accumulates, the fluorescence increases at all temperatures with double stranded product. However at denaturation temperatures, fluorescence returns to baseline since only single stranded DNA is present. The fluorescence v. temperature projection of double stranded dyes shown in Figure 30 eliminates the time axis and shows the temperature dependence of strand status during DNA amplification.

Example 11 A 536 base pair fragment of the human beta-globin gene was amplified from 25 ng of genomic DNA and a 1:10,000 dilution of SYBRTM Green I in a volume of 5 pI. Each temperature cycle was 28 sec long maximum, 61 0 C minimum, 15 sec at 72 0 C with an average rate between temperatures of 5.2 0 C/sec). Other conditions are the same as those described in Figure 30. Cycles 15-40 are displayed. Thetemperature dependence of product strand status during PCR is revealed by fluorescence v. temperature plots using as shown in Figure 31. Early cycles represented appear identical, with a nonlinear increase in fluorescence at lower temperatures. As amplification proceeds, temperature cycles appear as rising loops between annealing and denaturation temperatures. As the sample is heated, fluorescence is 25 high until denaturation occurs. As the sample cools, fluorescence e o :'..-increases, reflecting product reannealing. When the temperature is Soconstant during extension, increasing fluorescence correlates with S. additional DNA synthesis. As will be appreciated by an understanding of o.

S. this disclosure, continuous monitoring within a cycle can provide 30 insight into DNA amplification mechanics not previously available in the art. Using the present invention, many aspects of DNA amplification that have heretofore been little understood are discernable. For example, rapid cycle amplification claims to denature the product in o rless than one second, while the prior art uses ten seconds to one minute 35 of denaturation. Observing product melting by real time fluorescence monitoring with double strand dyes in accordance with the present *invention (Figures 30 and 31) shows that use of the shorter denaturation times is effective. As another example, many causes of the known "plateau effect" have been proposed, but few data are available to distinguish between alternatives. As shown in Figure 31, product reannealing is very rapid. In fact, during later cycles of amplification, a majority of product is reannealed each cycle during cooling before the primer annealing temperature is reached. This occurs with cooling rates of 5-10 0 C/sec in rapid cycle instrumentation. The product reannealing with slower, prior art temperature cyclers will be more extensive and this undesirable effect will be greater. Product reannealing appears to be a major, and perhaps the sole, cause of the "plateau effect." Now consider continuous monitoring of sequence specific probes.

As will be appreciated by an understanding of this disclosure, continuous monitoring within a cycle can identify the nature of probe fluorescence.

Example 12 Continuous monitoring of amplification every 200 msec was performed with a dual-labeled hydrolysis probe (beta-actin) and adjacent hybridization probes (beta-globin) as in Example 7. In Figure 32A, cycles 20-45 of a reaction monitored with the hydrolysis probe is shown.

Hydrolysis probes show a linear change in fluorescence ratio with temperature and a parallel increase in fluorescence as more probe is hydrolyzed. In contrast, the fluorescence ratio from hybridization probes varies radically with temperature (Figure 32B, cycles 20-40).

During the annealing/extension phase, the probes hybridize to single stranded product and the fluorescence ratio (Cy5/fluorescein) increases.

During heating to product denaturation temperatures, the probes dissociate around 70C, returning the fluorescence ratio to background levels.

25 Example 13 A 110 base pair beta-globin fragment was amplified from 50 ng of genomic DNA in a volume of 10 p1. The amplification conditions and adjacent hybridization probes of Example 3 were followed with either 0.4 U of Taq polymerase or 0.8 U of KlenTaql. Fluorescence was monitored each 100 msec. Fluorescence v. temperature plots using KlenTaql (Figure 33) and Taq (Figure 34) demonstrate melting of the probes at about The maximal signal with KlenTaql is greater than that with Taq, because of the exonuclease activity of the latter. At later cycles with Taq, o:e the fluorescence each cycle begins to decrease as the concentration of intact probe decreases. Three dimensional plot of temperature, time, and fluorescence are shown in Figure 35 (KlenTaql) and Figure 36 (Taq).

The present invention's combination of continuous fluorescence monitoring within each temperature cycle and analysis of the temperature and time dependence of hybridization provides advantages not otherwise obtainable. Figure 2 shows that information that was previously unobtainable can be extracted by continuous monitoring throughout the cycle. Continuous fluorescence monitoring during the product melting phase of the cycle provides useful information on the purity, identity, and quantity of DNA present during that cycle.

As a PCR reaction is heated from the extension temperature to the denaturation temperature, any DNA in the sample is melted to single strands. This denaturation can be observed as a drop in' the fluorescence of SYBRTM Green I. For small PCR products, the melting transition occurs over a narrow temperature range and the midpoint of that melting range is referred to as the Tm. Similar to sizing by gel electrophoresis, melting peak analysis measures a fundamental characteristic of DNA and can be used to identify amplified products.

Unlike gel electrophoresis, melting curve analysis can distinguish products of the same length but different GC/AT ratio. In addition, two products with the same length and GC content, but differing in their GC distribution (for example, equally distributed vs. a GC clamp on one end) would have very different melting curves. Thetemperature at which PCR products melt varies over a large range. Using empirical formulas known in the art, the effect of GC content on the melting temperature (Tm) of DNA predicts that a 0% GC duplex would melt 410 C lower than a 100% GC duplex. Given the same GC content, a 40 base pair primer dimer would melt 120C below a 1000 bp product. Hence, the range of Tm for potential PCR products spans at least 500C. This wide range allows most PCR products to be differentiated by melting curves. Thus, Sthe combination of fluorescence monitoring of PCR with melting curve analysis provides simultaneous amplification, detection, and differentiation of PCR products.

Example 14 30 DNA melting curves for three different PCR products were acquired on a microvolume fluorimeter integrated with a 24-sample thermal cycler with optics for SYBRTM Green I fluorescence (LightCycler LC24, Idaho Technology, Idaho Falls, Idaho). The primers for the 180 base pair hepatitis B surface antigen gene amplification were 35 TTCTCTCAAT-3' (SEQ ID NO:1) and 5'-AGAAGATGAG GCATAGCAGC-3'(SEQ ID NO:2). The primers for the 292 base pair prostate specific antigen (PSA) gene amplification were 5'-CTGTCCGTGA CGTGGATT-3' (SEQ ID NO:7) and 5'-AAGTCCTCCG AGTATAGC-3' (SEQ ID NO:8). The 536 base pair human beta-globin gene amplification was done as in Example 7. PCR was performed as described in Example 2. Amplification products were purified by phenol/chloroform extraction and ethanol precipitation, followed by repeated washing through a Centricon 30 microconcentrator (available from Amicon of Danvers, Massachusetts). Template concentrations were determined by absorbency at 260 nm and had A(260)/A(280) ratios greater than 1.7.

Fifty ng of purified DNA in 50 mM Tris, pH 8.5, 2 mM MgCl 2 and 250 vg/ml bovine serum albumin and a 5 p1 volume were pipetted into the open plastic reservoir of composite glass/plastic reaction tubes, centrifuged at low speed to place the sample at the tip of the glass capillary, and sealed inside with a plastic plug. Fluorescence data for melting curves was acquired by integrating the signal over 0.25 seconds during a linear temperature transition to 95C at 0.1- 10.0 0 C/second. The fluorescence was continuously acquired and displayed at fluorescence v. temperature plots in the LabView programming environment (National Instrument, Austin, TX) Figure 37 shows the melting curves of the three purified PCR products.

The Tm's of three products in Figure 37 span only 6 degrees and two of the curves are separated by only 2 degrees. This small separation is ample to allow easy differentiation of the products. The importance of GC percentage over length on Tm is illustrated by the 292 bp PSA product melting at a higher temperature than the longer 536 bp beta-globin fragment. Melting curves are often obtained at rates of to ensure equilibrium. Moreover, as the heating rate decreases, the melting curve shifts to lower temperatures and becomes sharper (Figure 38, hepatitis B fragment). Note however, that the melting curves of Figure 37 were obtained during a heating rate of 0.2°C/sec (12 0 C/minute) and can differentiate products differing in Tm by 2C or less.

The apparent Tm of PCR products is also dependent on doublestrand-specific DNA dye concentration (Figure 39, hepatitis B fragment) Higher concentrations of dye increase the stability of the DNA duplex 30 and the observed Tm.

For monitoring of melting curves with SYBRTM Green I, the preferred conditions are 1:7,000-1:30,000 fold dilution of SYBR Green I with heating rates of 0.l-0.5°C/second. These conditions allow easy differentiation of products that differ in Tm by 20 C.

35 More precise temperature control and software for melting peak analysis will reduce the detectable difference in Tm to a fraction of a degree. This will allow the differentiation of most PCR products. Not 1 all products can be differentiated by Tm however, just as it is possible to misread electrophoresis results because of comigration of two or more products, it is possible that some of the product melting in the expected range may not be the intended product. However, if no DNA melts in the range of the expected product, it can conclusively be said that none of the expected product is present.

Another form of product differentiation available with melting curve analysis is the distinctive patterns of domain melting seen in longer PCR products. While short products (<300 bp) usually melt in one transition, longer products can have internal melting domains that give melting curves of a complex, distinctive shape. These complex melting curves can be used as a fingerprint for product identification.

Melting curve analysis can be used to differentiate intended product from nonspecific products such as primer dimers. Primer dimers melt over a wide range of low temperatures; very different from the sharp melting curves of specific PCR amplification products. Larger heterogeneous products which resulted from running many cycles at low annealing stringency have lower and broader melting curves when compared with pure PCR product.

Example Amplification of the 536 beta-globin gene fragment was performed as in Example 7 with a 1:30,000 dilution of SYBRTM Green I except that the conditions were varied. In reaction A (Figure 40), no template was added and the reaction was cycled at 94C for 0 sec, 60 0 C for 0 sec, and 72C for 10 sec for 30 cycles to produce small nonspecific amplification products. In B, amplification of 106 initial copies of purified template at low stringency (94C for 0 sec, 50 0 C for 0 sec, and 72 0 C for sec) for 55 cycles showed a broad size range of amplification 25 products on gel electrophoresis and melts across a wide temperature o range. In C, 106 initial copies of purified template were cycled at 94 0 C for 0 sec, 60 0 C for 0 sec, and 72 0 C for 10 sec for 30 times and shows a single bright band and melts in a sharp transition. The e temperature transition rate was 0.2 0 C/sec. A Hind III digest of X phage 30 DNA is used as a marker.

Figure 40 shows how melting curves accurately reflect the :specificity of a PCR reaction. The sharp, high temperature melting curve C corresponds to a single band on a gel. The low temperature, broad melting, curve A comes from analysis of a no template control that o 35 shows only primer dimers. Over-amplification of the product in C gives the intermediate melting curve B, still clearly differentiable from the """specific product.

The melting curves seen, for example, in Figure 37, can be better S" quantified by first taking the derivative of fluorescence with respect to temperature This derivative is plotted as -dF/dT v. T and converts the melting curves to melting peaks.

Example 16 The purified hepatitis B and beta-globin gene fragments of Example 14 were melted individually and together with a temperature transition rate of 0.2°C/sec and other conditions as specified in Example 14 (Figure 41) The somewhat subjective determination of Tm from the melting curves (top) is easily called by eye from the melting peaks (bottom). The area under the melting peaks can also be quantified by integration of the area under the curves. The fluorescence baseline was first subtracted from the -dF/dT v. T plot assuming that the magnitude of the baseline varies as the area under the curve. Then the peaks were fit by nonlinear least squares regression to gaussians with the mean, standard deviation, and height of the peak as the fit parameters. The area under each gaussian was taken as the peak area. All calculations were performed in the LabView programming environment (National Instruments, Austin, TX). Figure 41 shows an example of this conversion of melting curves to melting peaks. The code for these calculations is included as appendix A.

The ability to distinguish specific product from primer dimer and other reaction artifacts improves the use of double-strand-specific DNA dyes in the quantification of low initial copy numbers. Relatively large initial template copy numbers have been quantified using ethidium bromide (Higuchi Watson, supra). However, at low initial copy numbers, the background amplification of primer dimers and other amplification artifacts interferes with the specific amplification signal. With the present invention's ability to differentiate specific products from non-specific artifacts, double-strand-specific DNA dyes can be used to quantify low initial template copy numbers. This is advantageous because of the simplicity of using these dyes. The doublestrand-specific DNA dyes can be used in any amplification and custom fluorescently-labeled oligonucleotides are not necessary.

•Quantification of very low copy numbers with double-strand-specific DNA dyes requires very good amplification specificity or, as provided by the present invention, a means to differentiate the desired product from nonspecific amplification.

a 35 Example 17 The present invention's approach to product purity determination was used to improve quantitative PCR based on once-per-cycle monitoring of double-strand-specific DNA dye fluorescence. Fluorescence was a. acquired once each cycle after polymerase extension of the product for a series of reactions varying in the initial concentration of purified beta-globin template (see Figure 42A). The beta globin template and amplification conditions were as given in Example 7. The log-linear increase above background fluorescence began at a cycle number dependent on initial template concentration. The plots of the five reactions ranging from 106 to 102 copies per reaction were separated by about four cycles. The sample with an average 102 copies per reaction showed a decrease in reaction efficiency, and reactions with initial copy number below 100 gave fluorescence profiles that were less useful. The fluorescence profiles for the reactions containing 10 and 1 (average) copies rise in reverse order, and the negative control showed amplification after about 30 cycles. This is due to amplification of primer dimers and other nonspecific amplification products that cannot be distinguished from the intended product by once-per-cycle fluorescence monitoring of double-strand-specific DNA specific dyes.

Melting peaks were acquired for each sample (Figure 42B) and these were found to correlate well with electrophoresis results (Figure 42C) The reaction containing zero and one average initial template copies produced no discernible electrophoresis band at the expected 536 base pair location. The reactions containing 10 and 100 initial copies of template showed weak electrophoresis bands. This agreed well with the melting peak analysis, which showed no DNA melting in the range of the intended product (90 92C) for the reactions containing zero and one initial copies and small peaks in this temperature range for 10 and 100 copies. Strong electrophoresis bands for the reactions containing 103 106 initial copies correlate well with large melting peaks in the expected 90 92C range.

The ratio of intended product to total product, determined by melting peak integration, ranged from 0.28 for 105 copies to 0.0002 for zero initial template copies. Each fluorescence value in Figure 41A was multiplied by the appropriate ratio to give the corrected plot (designated "corrected fluorescence" in Figure 42D). This procedure extended the effective dynamic range of quantitation to between 10 and 1 .5 initial template copies.

Melting peaks can distinguish specific products from non-specific products (Figure 40) and they can distinguish two purified PCR products mixed together (Figure 41) so they should also be useful for 35 distinguishing two specific products amplified together in a single reaction tube. Melting curves obtained by continuous monitoring of PCR reactions according to the present invention are useful in multiplex Example 18 In this example, two gene fragments were simultaneously amplified from genomic DNA and monitored with SYBRTM Green I fluorescence. During each amplification cycle, different amplification products denature at melting temperatures dependent on the length of the product, GC ratio, and other factors well known in the art. The temperature at which each product melts can be monitored with the double-strand-specific dye,

SYBR

T M Green I. At 81 base pair fragment from the cystic fibrosis gene was amplified using the primers described herein as SEQ ID NO:14 and SEQ ID NO:15 along with a 98 base pair fragment of the c-erbB-2 (HER2/neu) oncogene using the primers described herein as SEQ ID NO:16 and SEQ ID NO:17.

Amplification reactions were comprised of 50 mM Tris-HCl, pH 8.3, 3 mM MgCl 2 500 pg/ml of bovine serum albumin, 200 pM of each dNTP, and 0.5 pM of the cystic fibrosis primers, 0.3 pM of the HER2/neu primers, a 1:30,000 dilution of SYBRTM Green I, 1 U AmpliTaq Gold DNA polymerase (Perkin Elmer, Foster City, CA), and 50 ng of human genomic DNA in 10 p i After activation of the polymerase at 95 0 C for 30 minutes, the samples were cycled at 94 0 C for 0 seconds (slope 20), 55 0 C for 0 seconds (slope 20), and 70 0 C for 10 seconds (slope 20) for cycles. The samples were cooled to 70 0 C, and the fluorescence was continuously acquired during a 0.2 0 C/sec ramp to 94 0 C. Melting curves (Figure 43) clearly showed two distinct products melting at 78 0 C (CFTR) and 88 0 C (neu). The two products differ in Tm by approximately 10 0 C and are easily distinguishable.

Multiplex amplification is useful in cases where an internal 25 control is needed during amplification. For example, many translocations are detectable by PCR by placing primers on each side of the breakpoint. If no amplification occurs, the translocation is not :present as long as the DNA is intact and no inhibitor is present. These possibilities can be ruled out by amplifying a positive control locus in 30 the same reaction mixture. Such control amplifications are best done as internal controls with simultaneous amplification and detection.

Example 19 In this example, the procedure of Example 18 was followed except that after activation of the polymerase at 95 0 C for 30 minutes, the samples were cycled at 94 0 C for 0 seconds (slope 20), 55 0 C for 0 seconds (slope 20), and 70 0 C for 10 seconds (slope 20) for cycles, followed by 94 0 C for 0 seconds (slope 55 0 C for 0 seconds (slope 20), and 70 0 C for 20 seconds (slope 20) for 15 cycles. For cycles 26-31, fluorescence was continuously acquired during each l°C/sec 38 transition from 70 0 C to 94 0 C. The melting curves were converted to melting peaks and displayed (Figure 44) Note that the amplification efficiency of the CFTR fragment appears greater than the neu fragment. The amplification efficiency can be rigorously determined by integrating the melting peak data as in Example 16.

This kind of quantitative data referenced to a control has many applications. For instance, certain oncogenes, such as HER2/neu, are amplified in vivo in many tumors. That is, the genes are replicated in genomic DNA, sometimes many fold. Often, the clinical behavior of the tumor depends on the degree of oncogene replication. Amplification of the oncogene and a control template allows quantitative assessment of the relative copy number. As a further example, quantification of viral load in patients infected with HIV or hepatitis C is important in prognosis and therapy. Using a control template and monitoring the efficiency of amplification of both control and natural templates during amplification, accurate quantification of initial template copy number is achieved.

The present invention's feature of using melting curves for relative quantification will now be explained. In accordance with the present invention, an additional use for melting curves is quantitative PCR. Figure 42 showed there was a correlation between the area under the melting peak and the amount of specific product. Relative quantification of two PCR products would be possible if the two products were amplified with similar efficiency (or if the differing efficiencies were known and compensated for). Relative quantification of two products by integrating melting peak areas (see Example 16) is an aspect of the current invention.

*t Example The cystic fibrosis and HER-2-neu gene fragments of Example 18 were amplified, purified as in Example 2. and adjusted to 175 pg/ml.

The samples were mixed in various ratios (total 8 il) and added to *o buffer (1 p1) and SYBRTM Green I (1 pl). Final concentrations were mM Tris, pH 8.3, 3 mM MgCl 2 250 pg/ml bovine serum albumin, and a 1:30,000 dilution of SYBRTM Green I. Melting curves were acquired at ,o 0.2 0 C/sec, background fluorescence subtracted and the peaks integrated as described in Example 16. The results are displayed in Figure Excellent correlation was found between the relative areas under melting peaks and the relative amounts of the two products.

Relative quantification of two PCR products is important in many quantitative PCR applications. Multiplex amplification of two or more products followed by integration of the areas under the melting peaks will be extremely useful in these areas. mRNA is often quantified relative to the amount of mRNA of a housekeeping gene.

Another important use of relative quantification is in competitive quantitative PCR. Typically a competitor is synthesized that has the same priming sites, but differs in length from the original target sequence. Known amounts of the competitor are spiked into an unknown sample and relative quantitation is performed. Competitors can be made that differ from the target sequence in Tm rather than length. The relative amounts of the products can be quantified by comparing the areas under their melting peaks. As the amount of one of the products is known, the quantity of the original target can be obtained. Using the melting peak method is significantly easier than the currently used methods which involve running multiple tubes for each unknown sample and often pulling tubes at various cycle numbers during the reaction to find the log-linear portion of the reaction. The relative amounts of the two products must then be determined. Usually this is done by labeling one of the dNTPs with a radioisotope and then quantifying the amount of label incorporated into each band after agarose gel electrophoresis. In comparison, the current invention allows the reaction to be monitored continuously so the log-linear portion of the amplification can be easily identified. Relative quantification can be done quickly by integration of melting peaks. An all day process is reduced to less than an hour.

o..,From the foregoing discussion, it will be appreciated that S fluorescence monitoring during DNA amplification is an extraordinarily

SS.":

powerful analytical technique. When sequence-specific detection and 30 quantification are desired, resonance energy transfer probes can be used instead of double-strand-specific DNA dyes. The Tm of hybridization probes shifts about 4-8 0 C if a single base mismatch is present. If a hybridization probe is placed at a mutation site, single base mutations are detectable as a shift in the probe melting temperature.

*.SS

•ore o: 35 Example 21 The factor V Leiden mutation is a single base change (G to A) that substitutes a glutamine residue for an arginine residue at amino acid r :residue 506 (R506Q). For further information, see R.M. Bertina et al., S.e• e Mutation in Blood Coagulation Factor V Associated with Resistance to Activated Protein C, 369 Nature 64-67 (1994) and J. Voorberg et al., Association of Idiopathic Venous Thromboembolism with a Single Point- Mutation at Arg 506 of Factor V, 343 Lancet 1535-36 (1994), both of which are hereby incorporated by reference. As used herein, "factor V Leiden mutation locus" means the nucleotide position in the factor V gene at which a guanine base in the wild type is replaced by an adenine base in the factor V Leiden mutant. SEQ ID NO:9 shows a portion of the wild type factor V gene, and SEQ ID NO:10 shows the corresponding portion of the factor V Leiden gene, with the relevant nucleotide at position 31 in each case. The complete nucleotide sequence of the factor V gene is described at R.J. Jenny et al., Complete cDNA and Derived Amino Acid Sequence of Human Factor V, 84 Proc. Nat'l Acad. Sci. USA 4846-50 (1987), hereby incorporated by reference, and sequences can also be obtained at Genbank locus HUMF10. The amino acid change in the mutant factor V protein makes this clotting factor resistant to degradation and increases the tendency to clotting and thrombosis. As the most common cause of inherited thrombophilia, this mutation is the target of a common laboratory test done in clinical molecular genetics laboratories.

The standard method of analysis for the factor V Leiden mutation is to amplify the gene segment by PCR, digest the resulting amplified products with a restriction endonuclease that cuts the wild type sequence but not the mutant, and distinguish digested wild type and undigested mutant products by gel electrophoresis. R.M. Bertina et al., supra. This is a method well known in the art for analysis for defined mutations. Such a test usually requires about 4 hours, including PCR amplification (2 hours), enzyme digestion (1 hour), and electrophoresis (1 hour). Post-amplification steps include opening the sample tube, adding the enzyme, and transferring the digested sample to the electrophoresis apparatus. Post-amplification processing increases the •ao risk of end product contamination, and manual handling requires care to prevent mislabeling of samples. A method that simultaneously amplifies S• and analyzes for point mutations would eliminate these concerns.

Si 30 A method for complete amplification and analysis of the factor V S: Leiden mutation within 30 min in the same instrument comprises asymmetrically amplifying a portion of a human genomic DNA sample containing the mutation locus, followed by obtaining and analyzing a melting curve for the amplified DNA. Genomic DNA is prepared according 35 to methods well known in the art, e.g. J. Sambrook et al., Molecular Cloning: A Laboratory Manual (2d ed., 1989), hereby incorporated by reference. Preferably, the melting curve is obtained by the resonance Senergy transfer methodology described above with a fluorogenic hybridization probe. Such an assay easily discriminates between 40 homozygous wild type, homozygous mutant, and heterozygous genotypes. In a preferred embodiment, the oligonucleotide probe is 3'-labeled with fluorescein and designed to hybridize on the amplified DNA near to a primer for resonance energy transfer. This method can be applied to any defined mutation.

The probe oligonucleotide is preferably about 15-40 nucleotide residues in length. The probe could conceivably contain as few as about 10 nucleotide residues, however, possible disadvantages of such short oligonucleotides include low specificity, low melting temperature, and increased background. Oligonucleotides larger than 40 residues could also be used, but would be unnecessarily expensive. Thus, the limits on the size of the probe oligonucleotide are only those imposed by functionality. The probe oligonucleotide should span the mutation, but the mutation preferably does not correspond to either the or 3'terminal nucleotide residue of the probe. Since the present invention is based on melting curves, and lack of base pairing at the termini is known to have less of an effect on melting properties than at internal sites, the probe should be designed such that the mutation occurs at an internal position.

The oligonucleotide primers for amplification of the selected mutation locus are preferably about 15 to 30 residues in length.

Primers shorter than the preferred range could be used but may not be as specific as would be desired. Similarly, primers longer than the preferred range could be used, but would be unnecessarily expensive.

Thus, the limits on the sizes of the PCR primers are only those imposed by functionality.

The distance between the resonance energy transfer pair is also important for the proper functioning of the invention. The optimum SI.. distance between the resonance energy transfer pair is about nucleotides. A distance of about 2 to 8 nucleotides is preferred, although a distance of up to about 10-15 nucleotides is functional.

Having the resonance energy transfer pair on adjacent nucleotides is not S 30 necessarily beneficial because the distance between the resonance energy transfer pair is effected by the position on the DNA helix.

In this example, PCR amplification was carried out in 10 1l reaction mixtures comprising 50 mM Tris, pH 8.3, 3 mM MgCl 2 500 pg/ml bovine serum albumin, 200 pM each dNTP, 0.5 pM Cy5-labeled primer (SEQ S 35 ID NO:11), 0.2 pM unlabeled opposing primer (SEQ ID NO: 12), 0.1 pM o fluorescein-labeled probe (SEQ ID NO:13), 0.4 U Taq polymerase, and fifty ng human genomic DNA. Four different samples of DNA were tested: human genomic DNA from an individual homozygous for the factor V Leiden mutation; human genomic DNA from a heterozygous individual; human 40 genomic DNA from an individual homozygous for the wild type factor V allele; and a negative control without DNA. The orientation of the primer, the fluorescein-labeled probe, and the mutation site (identified by asterisk) are shown below: Cy 5'-TAATCTGTAAGAGCAGATCC-3' (SEQ ID NO:11) TAATCTGTAAGAGCAGATCCCTGGACAGGCGAGGAATACAGGTATT (SEQ ID NO:9)

ATTAGACATTCTCGTCTAGGGACCTGTCCGCTCCTTATGTCCATAA

(SEQ ID NO:13) Fluorescein The sequence of the unlabeled opposing primer was TGTTATCACACTGGTGCTAA (SEQ ID NO:12) and the amplified product was 186 base pairs in length.

The CyS-labeled primer was obtained as in Example 8. Cycling conditions were 940C for 0 sec (slope=20), 500C for 10 sec (slope=20), and 720C for 0 sec (slope=l) for 50 cycles, followed by cooling to 450C and continuous fluorescence monitoring at a slope of 0.2 0 C/sec to 940C for the melting curve. The highest quality melting curves were obtained at the end of amplification with a slow temperature transition rate (0.2 0 C/sec Figure 46), although monitoring during each cycle at l°C/sec between 50 0 C and 94C also provided clear genotype identification (Figure 47). The melting curves are easiest to visualize by plotting the negative derivative of fluorescence with respect to temperature vs temperature (-dF/dT vs Such a plot allows facile visual identification of all possible genotypes from the raw fluorescence data.

The closer the Cy5 label is to the primer's 3'-end, the greater the resonance energy transfer signal. However, the 3'-end must have a free 3'-hydroxyl for polymerase extension, and placing the Cy5 too close 30 to the 3'-end (either on the 3' or penultimate base) may inhibit polymerase attachment and extension. The 3'-fluorescein probe should hybridize as close to the primer as possible (minor overlap of 1-3 bases can be tolerated) and the mutation site should be near the middle of the probe. A 5-base separation between the hybridized fluorophores and a mutation at base 8 of a 23-mer probe gave a melting curve shift of between mutant and wild type sequences (Figure 46).

Mutation detection by probe melting can also be performed with 2 oo labeled probes instead of one labeled probe and one labeled primer. In this embodiment, one probe is labeled 5' with Cy5 and the other probe is labeled 3' with fluorescein. Since both these fluorescent probes can be synthesized directly from the amidites, a manual synthesis step is not required as it is in the primer/probe system. The fluorescein-labeled probe should be designed such that the mutation locus is near the center of the fluorescein-labeled probe. The length of the Cy5-labeled probe should be designed such that it melts at a higher temperature than the fluorescein-labeled probe which spans the mutation locus.

43 Because background from fluorescein is more troublesome than that from the concentration of the Cy5-labeled probe should preferably be fold that of the fluorescein-labeled probe. The two probes should hybridize to the same strand of genomic DNA, and the resonance energy transfer pair should be separated by about 0 to 5 nucleotide residues.

Alternately, the probe that spans the mutation site can be labeled with and the other probe labeled with fluorescein.

It will be appreciated that the particular probes and primers disclosed herein for detection of the factor V Leiden mutation are merely illustrative, and that a person of ordinary skill in the art will be able to design other probes and primers for detection of mutations without undue experimentation by following the principles and guidelines set forth herein. It should also be recognized that although the invention is described with respect to detection of a single base mutation in genomic DNA, the same principles can be applied to detection of a mutation in cDNA. Preparation of the cDNA requires extra process steps and time, as is well known in the art, thus it is preferred to use genomic DNA because of the advantages of speed and lower cost. Further, the same technique can be used to detect insertions and deletions by designing the hybridization probe so that it melting temperature changes when the mutation or polymorphism is present. The invention can be used to detect any known mutation where a probe can be designed to differ in melting temperature when hybridized to mutant vs wild type.

Although fluorescein and Cy5 were used as resonance energy transfer labels in the example above, other acceptors, such as can also be used with fluorescein.

4 Example 22 S, The factor V locus of Example 21 was amplified as before except that the primer was labeled with Cy5.5 instead of Cy5. Cy5.5 emission 30 was observed through a 683 nm long pass dichroic and a 683-703 nm bandpass interference filter. The Cy5.5 to fluorescein ratio increased above background at about cycle 30 and the ratio approximately doubled by 50 cycles of asymmetric amplification. When amplified with wild type S: DNA, the probe Tm was 65-66 0 C as judged by melting peaks.

35 Another example for detecting single base mutations will now be given.

Example 23 There is a common point mutation in the methylenetetrahydrofolate reductase (MTHFR) gene (C 677 T) that converts an alanine to a valine residue and results in a thermolabile enzyme. This mutation can reduce MTHFR activity and lead to elevated homocysteine plasma levels which has been implicated as an independent risk factor for early vascular disease and thrombosis as is well known in the art. One of the primers was labeled with Cy5 (TGAAGGAGAAGGTGTCT'GCGGGA) (SEQ ID NO:25) where T* represents a modified T residue linked to Cy5 (see Example 9 for synthesis and purification). The probe sequence was fluorescein- CCTCGGCTAAATAGTAGTGCGTCGA (SEQ ID NO:26) and the other primer was AGGACGGTGCGGTGAGAGTG (SEQ ID NO:27). A 198 base pair fragment of the MTHFR gene was amplified from 50 ng of human genomic DNA in 50 mM Tris, pH 8.3, 2 mM MgCl 2 500 pg/ml bovine serum albumin, 0.2 mM of each dNTP, 0.5 pM of the Cy5-labeled primer, 0.1 pM of the opposing primer, 0.1 pM of the fluorescein-labeled probe, and 0.4 U Taq DNA polymerase per pi. Each cycle was 30 sec long and consisted of denaturation at 94 0

C

followed by a 20 sec combined annealing/extension step at 60 0 C. The temperature transition rate between steps was 20 0 C/sec. After cycles, a melting curve was acquired as follows: heating from 50-65 0 C at 0 C/sec, 65-75 0 C at 0.1 0 C/sec, and 75-94 0 C at 0.5 0 C/sec. After baseline subtraction and conversion to melting peaks, all possible genotypes were easily distinguished (Figure 48).

The discriminatory power of hybridization probes is also used to great advantage in multiplex or competitive PCR. For example, an artificial template is designed with a single internal base change and a hybridization probe designed to cover the base change as in Examples 21 and 23. Relative amplification of the competitor and natural template are determined by acquiring and integrating melting peaks as in Example 25 16. Alternately, if multiple detection probes are used that Ob sequentially melt off different targets at different temperatures, relative quantification is achieved by the same analysis. In general, 9 any quantitative technique described previously for double-strandspecific DNA dyes can be made sequence specific with hybridization 30 probes.

Absolute Product Concentration by Product Reannealing Kinetics. Product S* concentration determinations are also advantageously carried out using the present invention. Continuous monitoring of double stranded DNA oo formation allows DNA quantification at any cycle of amplification by 35 reannealing kinetics. The sample temperature is quickly dropped from the denaturation temperature and held constant at a lower temperature that is still high enough to prevent primer annealing (Figure The rate of product reannealing follows second order kinetics (see B. Young M. Anderson, Quantitative analysis of solution hybridization, In: Nucleic Acid Hybridization: A Practical Approach 47-71 Hames S.

Higgins, eds., (1985), which is now incorporated herein by reference).

For any given PCR product and temperature, a second order rate constant can be measured. Once the rate constant is known, any unknown DNA concentration can be determined from experimental reannealing data.

Cooling is never instantaneous, and some reannealing occurs before a constant temperature is reached. Rapid cooling will maximize the amount of data available for rate constant and DNA concentration determination.

The technique requires pure PCR product, but such can be assured by melting curves also obtained during temperature cycling using the present invention. This method of quantification by the present invention is advantageously independent of any signal intensity variations between samples.

Example 24 A 536 base pair fragment of the beta-globin gene was amplified from human genomic DNA (Example 7) and purified (see Example 2).

Different amounts of the purified DNA were mixed with a 1:30,000 dilution of SYBRTM Green I in 5 pl of 50 mM Tris, pH 8.3 and 3mM MgCl 2 The samples were denatured at 94 0 C and then rapidly cooled to 85 0

C.

The fluorescence at 520-550 nm was monitored at 85 0 C over time. When different concentrations of DNA were tested, the shape of the reannealing curve was characteristic of the DNA concentration (See Figure 49). For any given PCR product and temperature, a second order rate constant can be determined. Figure 50 shows the determination of a second order reannealing rate constant for 100 ng of the 536 base pair fragment in 5 pl at 850C. The curve was fit by non-linear least squares regression with Fmax, Fmin, to and k as the floating parameters 25 using the second order rate equation shown in Figure 50. Analysis V* programs for this kind of curve fitting are well known in the art (for example, the user defined curve fit of Delta Graph, DeltaPoint, Inc, Monteray, CA) Once the rate constant is known, an unknown DNA concentration can be determined from experimental reannealing data.

.9.

With the rate constant defined, DNA concentrations are determined on unknown samples. The fluorescence vs time curves of unknown samples are fit by non-linear least squares regression, e preferably during temperature cycling in real time (for example, using P the nonlinear Levenberg-Marquardt method described in the LabView programming environment, National Instruments, Austin, TX). For this fit, Fmax, Fmin, to and [DNA] are the floating parameters and k is constant.

Since some fluorescent dyes affect reannealing in a concentration dependent manner, the assumption of second order kinetics for product reannealing is checked by determining the rate constant at different standard DNA concentrations. The relationship is defined and alternate formula for fitting incorporated as necessary.

Also within the scope of the present invention is to use probe annealing rates to determine product concentrations. The rate of fluorescence resonance energy transfer is followed over time after a quick drop to a probe annealing temperature that is greater than the primer annealing temperature (Figure For the case of amplification with a labeled primer and one labeled probe, the rate of annealing (and fluorescence generation) is second order. When using two labeled probes, the rate of fluorescence development is third order. These two arrangements are shown in Figure 18. When the concentration of the probe(s) is much greater than the product concentration, pseudo-first order and pseudo-second order equations are adequate to describe the possibilities. The appropriate rate equations for these different conditions are well known in the art (see Young, B. and Anderson, M., supra). For the purposes of this invention, it is adequate that the prior art suggests appropriate rate equations that are tested experimentally and corrected if necessary.

When probe annealing rates are used to determine product concentrations, possible interfering effects include product reannealing (with probe displacement by branch migration) and primer annealing and extension through the probe. The later is minimized when the probe Tm's are higher than the primer Tm's and a probe annealing temperature is chosen to minimize primer annealing. Figure 13 shows that even if extension occurs, the fluorescence increases with time for about 20 sec.

."During this period, the fluorescence increase depends on product 0000 °0°:concentration.

~Probe annealing rates are used to determine product concentration 3 similar to the method described above for determining product S30 concentration by product reannealing. The steps are summarized as e• follows: choosing the appropriate rate equation for the system, (2) running known DNA standards to determine the rate constant, checking the validity of the rate equation by comparing different rate constants oderived from different concentrations, and using the rates constants 0oooo 35 to determine the DNA concentration of unknowns from their probe *0 :°°annealing data.

0000 easeFluorescence Feedback for Control of Temperature Cycling. In contrast goo.. *to endpoint and cycle-by-cycle analysis, the present invention can also So monitor fluorescence throughout each temperature cycle. Continuous fluorescence monitoring can be used to control temperature cycling parameters. The present invention uses fluorescence feedback for real time control and optimization of amplification. Continuous fluorescence monitoring of PCR samples containing a double-strand-specific DNA dye or fluorescently labeled oligonucleotide probes can be used to monitor hybridization and melting during individual amplification cycles. This information can be used by the temperature control algorithms within the temperature cycling apparatus to improve and customize thermal cycling conditions. Conventional PCR is performed by programming all cycling parameters before amplification. With continuous monitoring, determination of temperature cycling requirements can occur during amplification, based on continuous observation of annealing, extension, and denaturation. The potential behefits of using hybridization information to control temperature cycling include: i. Ensuring complete denaturation of the PCR product each cycle while: a. Minimizing exposure to excessively high denaturation temperatures thus avoiding heat-induced damage to the amplification products and polymerase. Limiting the time product is exposed to denaturation temperatures is especially useful for amplification of long products.

b. Increasing reaction specificity by minimizing the denaturation temperature. This selects against products with a Tm higher than the intended amplification product.

e Maximizing the amplification efficiency by ensuring adequate time e for primer annealing each cycle while: a. Minimizing the amount of time required for amplification by 25 allowing no longer than is needed to reach a certain efficiency of primer annealing.

b. Enhancing reaction specificity by minimizing the time at the annealing temperature.

eeoc 3. Maximizing the amplification efficiency by ensuring adequate time 30 for product extension each cycle while: a. Minimizing the amount of time required for amplification by allowing no longer than needed to complete product extension.

b. Enhancing reaction specificity by selecting against products longer than the intended amplification product. These would require longer than the allotted time to complete product extension.

4. Initiating thermal cycling changes dependent on the level of fluorescence obtained or the current efficiency of amplification. For example, over-amplification and nonspecific reaction products can be minimized by terminating thermal cycling when the efficiency drops to a certain level. As another example, temperature cycling can be modified 48 to initiate slower temperature ramps for melting curve acquisition when the fluorescence becomes detectable. This saves time because the slower ramps need not be used on earlier cycles. Other desirable changes may become evident on continued practice of the invention.

Control is based on an estimate of reaction parameters from the fluorescence data. The original fluorescence data is either acquired as a change in fluorescence over time (temperature specific rates of denaturation, annealing, and extension), a change in fluorescence over temperature (product or probe Tm), or a change in extent of amplification (amplification yield and efficiency). These rates, Tm's and their first and second derivatives are used to determine optimal reaction parameters that include denaturation temperature and time, primer annealing temperature and time, probe annealing temperature and time, enzyme extension temperature and time, and number of cycles.

Double-strand-specific DNA dyes are used for the control of denaturation, control of extension, and to initiate thermal cycling changes at a certain amplification level or efficiency. Resonance energy transfer dyes are used for the control of annealing as will be described after the following example.

Example A commercial fluorescence monitoring thermal cycler (LC24 LightCycler, Idaho Technology Inc., Idaho Falls, Idaho) was modified so that the software is no longer programmed with temperature/time setpoints, but is programmed to acquire fluorescence values, then to use 25 these values for thermal cycler control.

As depicted in the Functional Block Diagram (Figure 51), the Run- Time Program communicates through serial and DAQ-board interfaces with the LightCycler. This allows high level access to either temperature or fluorescence data and either can be used by the Board-level Software for 30 temperature control. However, in the current embodiment of the nee* instrument, only the temperature data is converted into digital form at S: the Controller Hardware level. The fluorescence data is sent in analog form through the Digital acquisition board interface, is analyzed by the Run-time Program, and is sent back to the Board-level software via the 35 serial interface.

Product melting control: A melting peak was acquired for the intended PCR product and a baseline fluorescence was acquired for the sample containing the reaction cocktail at the temperature at which the product was completely melted.

Each cycle of the reaction then used this fluorescence value as a target. The approach to product denaturation was made in two stages to 49 overcome the time-lag due to the requirement of sending the fluorescence value to a remote computer for analysis, then returning the instruction that the value had been reached. With each product melting step, the temperature was increased until the fluorescence reached an intermediate value, then the heating power was reduced so that a temperature ramp rate of roughly 3 0 /sec gave the computer time to analyze the fluorescence and signal the thermal cycler that product denaturation had occurred.

The resulting temperature/time plot (Fig 52) shows a characteristic increase in the melting temperature after cycle 20 as the concentration of amplification product increases. Product Tm is a function of product concentration.

Product annealing/extension: During an extended hold at a combined annealing/extension temperature, the fluorescence was monitored and this information was used to ensure that adequate, but not excessive time had been allowed for product extension. The fluorescence was monitored at 10 second intervals, if the fluorescence increased more than a settable ratio (typically 1.00 1.05 then the annealing/extension step was continued. Otherwise, the next product melting step was initiated. The interval of 10 seconds was chosen to give a minimum of 20 seconds at the combined annealing/extension temperature.

The resulting fluorescence/time plot (Fig 52) shows a characteristic increase in the dwell time at the combined 25 annealing/extension temperature as the concentration of amplification S* product grows. As the primer concentration and polymerase become S. limiting, more time is needed to complete product extension with each cycle.

30 Amplification plateau: At the end of each annealing/extension step, the fluorescence value was acquired and stored. When this value had increased to 1.2 times the lowest end-cycle fluorescence value and had subsequently stopped increasing below a user settable ratio (typically 1.00 1.02), the thermal cycling was terminated. Alternately, a melting-curve 35 acquisition step was initiated by entering a slow 0.1 0.2 0 C/second temperature ramp through the product Tm and monitoring the fluorescence of the sample continuously.

The resulting fluorescence/time plot (Fig 52) shows that after twenty-five cycles of amplification the ratio of cycle-by-cycle fluorescence growth fell below 1.00 and the reaction was terminated.

In one embodiment of the present invention, detection of the amplification plateau is used to acquire a high-resolution melting curves for each sample in a multiple sample run at the optimal temperature cycle for each sample. As a sample reaches its amplification plateau, a melting-curve is acquired for that sample, then regular temperature cycling is resumed until another reaction reaches its amplification plateau.

Real time monitoring and control of annealing distinct from extension is also provided by the present invention. If one of the primers is 3'-labeled with Cy5, no extension can occur. However, if labeled primer is mixed with unlabeled primer (90-99%), amplification efficiency will be slightly decreased, but annealing is observable as fluorescence energy transfer from a double-strand-specific dye to Cy5. The primer with the lowest Tm (as determined by nearest neighbor thermodynamics as known in the art) is labeled with Cy5 and SYBRTM Green I is included as a double-strand-specific dye.

Alternately, primer annealing can be monitored indirectly with equivalent complementary oligonucleotides. An oligonucleotide of the same length and Tm as the lowest melting primer is designed with no complementarity to the amplified sequence. This oligonucleotide is labeled with Cy5 and its complement is 3'-labeled with fluorescein or some other resonance energy transfer pair. Hybridization of these oligonucleotides is followed by resonance energy transfer. The concentration of one probe is made the same as the concentration of the lowest Tm primer and the concentration of the other probe is made much less than this in order to obtain pseudo-first-order kinetics that approximates the pseudo-first-order kinetics of primer annealing to product. The efficiency of annealing is monitored and used to control annealing temperature and times by one of these methods.

It is also within the scope of the present invention to entirely replace temperature and time setpoints with fluorescence feedback 30 control. For example, three samples are placed in a fluorescence temperature cycler with feedback capacity. The samples are: i. A non-reacting sample containing amplified product and SYBRTM Green I.

2. A non-reacting sample containing complementary fluorescently 35 labeled primers with a Tm equal to the lowest Tm primer and oooo concentrations as noted above.

3. The sample to be amplified and SYBRTM Green I.

With each cycle of amplification, product denaturation is ensured by monitoring sample 1 as the temperature is increased. A melting curve is determined in real-time and when the sample has denatured, the transition to the annealing step is begun. Primer annealing is monitored indirectly through the hybridization of two complementary primers in sample 2. One of the primers is 3' labeled with fluorescein and the other is 5' labeled with Cy5 or similar dye. The temperature is decreased until sample 2 shows primer hybridization as indicated by an increase in the ratio of fluorescence at 670 nm 540 nm. This ratio increases due to resonance energy transfer between the fluorophores when they are approximated by hybridization. Product extension is followed by monitoring the fluorescence of one or more of the actual samples as demonstrated in Example Summary. From the foregoing discussion, it will be appreciated that continuous fluorescence monitoring during DNA amplification to monitor hybridization is an extraordinarily powerful analytical technique. Using the methods described herein and depending on the number of initial template copies present, product identification and quantification can be achieved in five to twenty minutes after temperature cycling has begun. The present invention achieves several advantages not heretofore available in the art. For example, the present invention provides single-color fluorescence methods to monitor product purity, relative quantitation by multiplex PCR or competitive PCR, absolute product quantification by reannealing kinetics, and an improved method for initial template quantification by fluorescence vs cycle number plots. The present invention also provides dual-color, sequence-specific methods for sequence variation detection, relative quantitation by multiplex PCR or competitive PCR, product quantification by probe annealing kinetics, and initial template quantification by S fluorescence vs cycle number plots.

99 9.

25 The following table compares double-strand-specific DNA dyes, hydrolysis probes, and hybridization probes useful in continuous monitoring of PCR. The fluorescence of double-strand-specific DNA dyes depends on the strand status of the DNA. The dual-labeled hydrolysis probes are quenched while intact and donor fluorescence increases when 30 the probe is hydrolyzed. Hybridization probes depend on increased eeoa resonance energy transfer when hybridization brings 2 fluorophores closer together.

oo Summary of Fluorescent Probes for Continuous Monitoring of PCR Fluorescent Probe 999 dsDNA dye Hydrolysis Hybridization Mechanism Strand status Quenching Transfer Probe Synthesis Unnecessary Difficult Simple Specificity Product Tm Sequence Sequence Melting Analysis Yes No Yes Multicolor Analysis No Yes Yes In accordance with the present invention, time, temperature and fluorescence are acquired 1-10 times every sec and fine details of product and/or probe hybridization are observed during temperature cycling. With double-strand-specific DNA dyes, the hybridization of product with respect to temperature is used to identify products by melting curves. In addition, relative product quantification is achieved by multiplex amplification of two or more different products that differ in Tm. Further, competitive PCR is performed by altering the sequence internal to the common primers so that two or more products have different Tm's. Absolute product quantification is obtained by rapidly cooling the denatured product and observing reannealing kinetics. The sensitivity of initial template quantification with fluorescence vs cycle number plots is increased by analysis of product melting curves to control for nonspecific amplification and curve fitting algorithms. Finally, immediate fluorescence feedback for control of denaturation conditions, elongation times and product yield are obtained by monitoring product strand status with double-strandspecific DNA dyes.

The ability to monitor probe hybridization with fluorescence during temperature cycling is a powerful tool. The present invention provides dual-color fluorescence methods that depend on probe hybridization (not hydrolysis) for sequence-specific detection and quantification during PCR. The annealing kinetics and melting of hybridization probes provides information not available with probes that S 25 rely on exonuclease hydrolysis between fluorophores. Continuous monitoring of sequence-specific probe hybridization can be followed over temperature changes by resonance energy transfer. Probe melting occurs at a characteristic temperature determined by its sequence and complementarity to the product. Two schemes have been detailed by the present invention, two adjacent hybridization probes, and one labeled probe that hybridizes to a single stranded PCR product that incorporates a labeled primer. The melting temperature of sequencespecific probes identifies and discriminates products during PCR. DNA polymorphisms or mutations, including single base mutations, are detected by probe Tm shifts. In addition, relative product oo :quantification is achieved by multiplex amplification of at least two different products with one or more probes that melt from their respective products at different temperatures. Further, competitive PCR is performed by altering the sequence internal to the primers so that one or more probes hybridize to the competitor and the natural template at different Tm's. Alternately, relative or competitive PCR are performed by multicolor analysis with probes labeled with different fluorophores, such as Cy5 and Cy5.5. Absolute product concentration is 53 determined by analysis of probe annealing kinetics. Initial template copy number is determined from fluorescence vs cycle number plots by curve fitting algorithms.

When multiplex analysis in one PCR reaction is desired, it is common practice to use different fluorescent labels with distinguishable emission spectra to identify the multiple products. The analysis is complicated by the limited number of fluorophores available and the overlapping emission spectra of those fluorophores that are available (see HM Shapiro, supra). Analysis of product or probe hybridization with melting curves is another method to distinguish multiple PCR products. By following hybridization during temperature cycling, the number of probes and/or spectral colors needed to distinguish multiple products can be minimized. Thepresent invention may be embodied in other specific forms without departing from its spirit or essential characteristics. The described embodiments are to be considered in all respects only as illustrative and not restrictive. The scope of the invention is, therefore, indicated by the appended claims rather than by the foregoing description. All changes which come within the meaning and range of equivalency of the claims are to be embraced within their scope.

Programming code for carrying out melting curve and other analyses is found in Figures 53-105.

ooo SEQUENCE LISTING GENERAL INFORMATION: APPLICANT: Wittwer, Carl T.

Ririe, Kirk M.

Rasmussen, Randy P.

(ii) TITLE OF INVENTION: Monitoring Hybridization During PCT (iii) NUMBER OF SEQUENCES: 27 30 (iv) CORRESPONDENCE ADDRESS: ADDRESSEE: Thorpe, North Western, L.L.P.

STREET: 9035 South 700 East, Suite 200 S(C) CITY: Sandy 000(D) STATE: Utah •o:o 35 COUNTRY: USA ZIP: 84070 COMPUTER READABLE FORM: MEDIUM TYPE: Diskette, 3.5 inch, 1.44 Mb storage COMPUTER: Toshiba T2150CDS OPERATING SYSTEM: Windows SOFTWARE: Word Perfect (vi) CURRENT APPLICATION DATA: APPLICATION NUMBER: FILING DATE:

CLASSIFICATION:

(vii) PRIOR APPLICATION DATA: APPLICATION NUMBER: US 08/658,993 FILING DATE: 04-JUN-96 (vii) PRIOR APPLICATION DATA: APPLICATION NUMBER: US 08/818,267 FILING DATE: 17-MAR-97 (viii) ATTORNEY/AGENT INFORMATION: NAME: Alan J. Howarth REGISTRATION NUMBER: 36,553 REFERENCE/DOCKET NUMBER: 8616.CIP7 (ix) TELECOMMUNICATION INFORMATION: TELEPHONE: (801)566-6633 TELEFAX: (801)566-0750 INFORMATION FOR SEQ ID NO:1: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:1: CGTGGTGGAC TTCTCTCAAT INFORMATION FOR SEQ ID NO:2: S2 SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:2: 30 AGAAGATGAG GCATAGCAGC 0 INFORMATION FOR SEQ ID NO:3:

O*

SEQUENCE CHARACTERISTICS: LENGTH: 35 base pairs TYPE: nucleic acid 35 STRANDEDNESS: single-stranded TOPOLOGY: linear a (xi) SEQUENCE DESCRIPTION: SEQ ID NO:3: CAAACAGACA CCATGGTGCA CCTGACTCCT GAGGA INFORMATION FOR SEQ ID NO:4: SEQUENCE CHARACTERISTICS: LENGTH: 30 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: AAGTCTGCCG TTACTGCCCT GTGGGGCAAG INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 27 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID TCTGCCGTTA CTGCCCTGTG GGGCAAG 27 INFORMATION FOR SEQ ID NO:6: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: CAACTTCATC CACGTNCACC INFORMATION FOR SEQ ID NO:7: SEQUENCE CHARACTERISTICS: LENGTH: 18 base pairs TYPE: nucleic acid S(C) STRANDEDNESS: single-stranded 25 TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: CTGTCCGTGA CGTGGATT 18 INFORMATION FOR SEQ ID NO:8: SEQUENCE CHARACTERISTICS: 30 LENGTH: 18 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:8: 35 AAGTCCTCCG AGTATAGC 18 INFORMATION FOR SEQ ID NO:9: SEQUENCE CHARACTERISTICS: LENGTH: 46 base pairs TYPE: mucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: 56 TAATCTGTAA GAGCAGATCC CTGGACAGGC GAGGAATACA GGTATT 46 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 46 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID TAATCTGTAA GAGCAGATCC CTGGACAGGC AAGGAATACA GGTATT 46 INFORMATION FOR SEQ ID NO:11: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: TAATCTGTAA GAGCAGATCC INFORMATION FOR SEQ ID NO:12: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear S(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: 25 TGTTATCACA CTGGTGCTAA INFORMATION FOR SEQ ID NO:13: SEQUENCE CHARACTERISTICS: LENGTH: 23 base pairs TYPE: nucleic acid 30 STRANDEDNESS: single-stranded S* TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:13: AATACCTGTA TTCCTCGCCT GTC 23 O INFORMATION FOR SEQ ID NO:14: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:14: ATGCCTGGCA CCATTAAAGA 57 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID GCATGCTTTG ATGACGCTTC INFORMATION FOR SEQ ID NO:16: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:16: CGGATCTTCT GCTGCCGTCG INFORMATION FOR SEQ ID NO:17: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:17: 25 CCTCTGACGT CCATCATCTC INFORMATION FOR SEQ ID NO:18: SEQUENCE CHARACTERISTICS: LENGTH: 31 base pairs TYPE: nucleic acid 30 STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:18: GAAGTCTGCC GTTACTGCCC TGTGGGGCAA G 31 INFORMATION FOR SEQ ID NO:19: SEQUENCE CHARACTERISTICS: LENGTH: 26 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:19: CTGCCGTACT GCCCTGTGGG GCAAGG 26 58 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 26 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID ATGCCCTCCC CCATGCCATC CTGCGT 26 INFORMATION FOR SEQ ID NO:21: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:21: CAACTTCATC CACGTTCACC INFORMATION FOR SEQ ID NO:22: SEQUENCE CHARACTERISTICS: LENGTH: 27 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:22: GTCTGCCGTT ACTGCCCTGT GGGGCAA 27 25 INFORMATION FOR SEQ ID NO:23: SEQUENCE CHARACTERISTICS: LENGTH: 32 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded 30 TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:23: CCTCAAACAG ACACCATGGT GCACCTGACT CC 32 INFORMATION FOR SEQ ID NO:24: 0(i) SEQUENCE CHARACTERISTICS: 35 LENGTH: 30 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:24: GAAGTCTGCC GTTACTGCCC TGTGGGGCAA INFORMATION FOR SEQ ID 59 SEQUENCE CHARACTERISTICS: LENGTH: 23 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID TGAAGGAGAA GGTGTCTGCG GGA 23 INFORMATION FOR SEQ ID NO:26: SEQUENCE CHARACTERISTICS: LENGTH: 25 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:26: CCTCGGCTAA ATAGTAGTGC GTCGA INFORMATION FOR SEQ ID NO:27: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single-stranded TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:27: 9* :AGGACGGTGC GGTGAGAGTG 9 9 9 9 9 f of fttf fttff fte •oo ftooof ft ft ft f ft ft *oo

Claims (21)

1. A detectable complex, comprising: a target analyte, a first probe labeled with fluorescein, and a second probe labeled with Cy5 or wherein the first and said second probe are bound to the target analyte and are in a fluorescence resonance energy transfer relationship.

2. A detectable complex, comprising: a probe system comprising a donor fluorophore having an emission spectrum and an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap less than and wherein the donor and acceptor fluorophore are in a fluorescence resonance energy transfer relationship.

3. The detectable complex of claim 2, wherein the probe system is an oligonucleotide 15 dual-labeled with the donor and acceptor fluorophores.

4. The detectable complex of claim 2 or 3, wherein the donor fluorophore is t fluorescein and the acceptor fluorophore is Cy5 or A detectable complex, comprising: a target analyte, oo 20 a first probe comprising a donor fluorophore having an emission spectrum, and a second probe comprising an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's 4:-6I absorption spectrum overlap less than 25%, and wherein the first and second probes are bound to the target analyte and the first and second probes are in a floroescence resonance energy transfer relationship.

6. The detectable complex of claim 5, wherein the acceptor fluorophore has an extinction coefficient greater than 100,000 M-cm 1

7. The detectable complex of claim 5 or 6, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap less than

8. The detectable complex of any one of claims 5 to 7, wherein the donor 8fluorophore is fluorescein and the acceptor fluorophore is S9. A detectable nucleic acid complex for the detection of a locus of a nucleic acid -61 sample, comprising: a donor probe hybridised to the locus, the donor probe comprising an oligonucleotide labelled with a donor fluorophore having an emission spectrum; an acceptor probe hybridised to the locus, the acceptor probe comprising an oligonucleotide labelled with an acceptor fluorophore having an absorption spectrum and an extinction coefficient greater than 100,000 M 'cm 1 wherein the donor fluorophore and the acceptor fluorophore comprise a fluorescence resonance energy transfer pair such that the donor fluorophore emission spectrum and the acceptor fluorophore absorption spectrum overlap by less than 25%, and the donor and acceptor fluorophore's are in a resonance energy transfer relationship; wherein the nucleic acid sample further comprises components for amplifying the locus.

10. The detectable nucleic acid complex of claim 9, wherein the components are for S conducting nucleic acid amplification by polymerase chain reaction. .r 15 11. The detectable nucleic acid complex of claim 10, wherein the components include STaq polymerase and a pair of primers.

12. The detectable nucleic acid complex of claim 10, wherein the components further comprise probes, and the probes and primers are provided together in a single container.

13. The detectable nucleic acid complex of any one of claims 9 to 12, wherein the donor fluorophore is fluroescein.

14. The detectable nucleic acid complex of anyone of claims 9 to 13, wherein the acceptor fluorophore is

15. The detectable nucleic acid complex of any one of claims 9 to 14, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap by about

16. A kit for amplification and analysis of a locus of a nucleic acid sample, comprising: a donor probe comprising an oligonucleotide labelled with a donor fluorophore having an emission spectrum; an acceptor probe comprising an oligonucleotide labelled with an acceptor S fluorophore having an absorption spectrum and an extinction coefficient greater than -61a- 100,000 wherein the donor fluorophore and the acceptor fluorophore comprise a fluorescence resonance energy transfer pair such that the donor fluorophore emission spectrum and the acceptor fluorophore absorption spectrum overlap by less than and wherein the donor and the acceptor fluorophore's are in a resonance energy transfer relationship when the donor probe and acceptor probe are hybridised to the locus; and components for amplifying the locus; wherein the donor probe and the acceptor probe are provided together in the same container.

17. The kit of claim 16, wherein the components are for conducting nucleic acid amplification by polymerase chain reaction.

18. The kit of claim 16 or 17, wherein the components include Taq polymerase and a pair of primers.

19. The kit of claim 17, wherein the components further comprise probes, and the S• probes and primers are provided together in a single container. 15 20. The kit of any one of claims 16 to 19, wherein the donor fluorophore is fluroescein.

21. The kit of anyone of claims 16 to 20, wherein the acceptor fluorophore is S21. The kit of any one of claims 16 to 20, wherein the acceptor fluorophore's emis

22. The kit of any one of claims 16 to 21, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap by about 2 0 23. A detectable nucleic acid complex, comprising: an amplified nucleic acid, wherein the nucleic acid was amplified with first and second primers, and a probe hybridised to the amplified nucleic acid, wherein the first primer and the probe are each labelled with one member of a fluorescence energy transfer pair comprising a donor fluorophore and an acceptor fluorophore such that the donor fluorophore emission spectrum and the acceptor fluorophore absorption spectrum overlap by less than 25%, and wherein the donor and acceptor fluorophore's are in a resonance energy transfer relationship. -61b-

24. A detectable complex comprising a target analyte, a first probe labelled with fluorescein and a second probe labelled with Cy5 or Cy5.5, and wherein the first and said second probe are bound to the target analyte and are in a fluorescence resonance energy transfer relationship, substantially as hereinbefore described with reference to one or more of the Examples and accompanying drawings. A detectable complex comprising a donor fluorphore having an emission spectrum and an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap less than 25%, and wherein the donor and acceptor fluorophore are in a fluorescence resonance energy transfer relationship, substantially as hereinbefore described with reference to one or more of the Examples and accompanying drawings.

26. A detectable complex comprising a target analyte, a first probe comprising a donor fluorphore having an emission spectrum and a second probe comprising an acceptor .fluorphore having an absorption spectrum, wherein the donor fluorophore's emission 15 spectrum and the acceptor fluorophore's absorption spectrum overlap and the first and S: second probes are bound to a target analyte and are in a fluorescence resonance energy transfer relationship, substantially as hereinbefore described with reference to one or more of the Examples and accompanying drawings.

27. A kit for amplification and analysis of a locus of a nucleic acid sample, comprising 20 a first probe comprising a donor fluorphore having an emission spectrum and a second o* probe comprising an acceptor fluorophore having an absorption spectrum, wherein the donor fluorophore's emission spectrum and the acceptor fluorophore's absorption spectrum overlap, and the first and second probes are provided together in the same container, substantially as hereinbefore described with reference to one or more of the Examples and accompanying drawings. DATED this 13th day of February, 2002 UNIVERSITY OF UTAH RESEARCH FOUNDATION Attorney: DAVID A. ADAMTHWAITE Fellow Institute of Patent and Trade Mark Attorneys of Australia of BALDWIN SHELSTON WATERS