Assay methods and kits for detecting rare sequence variants

Abstract

The present invention relates to methods of analyzing nucleic acids. The present disclosure provides methods of primer-dependent amplification and detection methods capable of amplifying and detecting as few as ten copies of at least one rare target sequence of interest in a sample in the presence of a large number of closely related non-target sequences of interest. Reaction compositions and kits for performing the methods are also provided.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority to U.S. Provisional Application No. 62 / 909,483, filed October 2, 2019, the publication of which is incorporated herein by reference. Background of the Invention

[0004] The ability to detect and quantify the presence of extremely rare nucleic acid sequence variants in samples containing a large number of closely related sequences has long been a goal; for example, detecting rare somatic mutation sequences occurring in tumor cells (10 or fewer per 10,000 cells) in clinical samples (typically 10,000 or 100,000 cells) containing large copies of wild-type sequences from normal cells. Next-generation sequencing analysis has been employed to achieve this goal, but the sensitivity of sequencing methods is limited due to the need for extensive amplification of nucleic acids in the sample, and because nucleic acid polymerases sometimes incorporate incorrect nucleotides (producing mutant sequences not present in the original sample). Furthermore, sequencing analysis requires expensive equipment, is slow (taking several days), and is costly, costing approximately $4,000 per assay.

[0005] A variety of designs of allele-discriminating primers have been used in exponential amplification assays, with the hope of selectively initiating amplification of one or more rare nucleic acid variants while ignoring the large number of closely related wild-type nucleic acids. These designs share in common that they contain at least one nucleotide, which we call an interrogating nucleotide, that is complementary to the target sequence of interest but is mismatched to the closely related non-target sequence. They include: hairpin-shaped primers (published International Patent Application WO 2000 / 71562 (30 November 2000) and corresponding U.S. Patent 6,365,729); Amplification Refractory Mutation System (“ARMS”) primers (Newton et al. (1989) Nucleic Acids Research 17:2503-2516; Kwok et al. (1990) Nucleic Acids Research 18:999-1005); and multi-part primers containing an internal sequence that is not complementary to the target sequence, sandwiched between two target-complementary sequences. Multi-part primers include our own laboratory’s very-highly selective SuperSelective primers against closely related alleles, disclosed in published International Patent Applications WO 2014 / 124290 (14 August 2014) and WO 2017 / 176852 (12 October 2017), and U.S. Patent 9,909,159. A commonality of allele-discriminating primers is that they contain an interrogating nucleotide that is complementary to the target rare sequence of interest but is mismatched to the large number of closely related non-target sequences.

[0006] Superselective primers have a fairly high selectivity (see, e.g., Vargas et al. (2016) PLoS ONE 11 :e0156546 and Vargas et al. (2018) Journal of Molecular Diagnostics 20:415-427) that discriminates among a large number of closely related strands with few copies. It is especially desirable to ensure that when analyzing clinical samples that contain only a few mutant targets (e.g., samples that contain ten or fewer mutant molecules in the presence of 10,000 closely related wild-type molecules), one does not need to average multiple parallel amplifications to determine whether the amplified signal is due to the presence of those few mutant molecules or whether the signal is the result of non-target amplification that occurred on the large number of wild-type molecules. Thus, when only a single amplification is performed, samples that contain only a few mutant targets can sometimes be confused with samples that contain no mutant targets (leading to false negative conclusions), and samples that contain no mutant targets can sometimes be confused with samples that contain only a few mutant targets (false positive conclusions).

[0007] There is a great need for an assay to detect very rare mutations that is extremely sensitive, easy to use on existing spectrofluorometric thermal cyclers, low cost, fast (hours rather than days), and non-invasive. The required assay must be able to distinguish between a sample that contains only a large number of nucleic acid sequences (e.g., wild-type sequences) and a sample that contains, per 10,000 copies of the large number of nucleic acid sequences, as few as ten copies of a closely related rare nucleic acid sequence (e.g., a mutant sequence).

[0008] For assays that use allele-discriminating primers, there is a need for robust assays to detect a very small number of mutant target sequences in the background of a large number of wild-type sequences, e.g., 10 mutant sequences in a mixture containing 10,000 wild-type sequences. Selectivity and sensitivity can be distilled down to a single characteristic that is able to detect as few as ten copies of a rare target sequence in a mixture of 10,000 copies of closely related sequences that differ by one or two nucleotides. SUMMARY

[0009] The present invention addresses the above needs in multiple aspects.

[0010] In one aspect, the present invention provides a primer-dependent amplification and detection method capable of amplifying and detecting as few as ten copies of at least one rare DNA target sequence of interest ("rare target sequence") of a sample in a mixture containing 10,000 copies of a closely related non-target sequence of interest ("closely related sequence" or "non-target sequence") that differs from the rare target sequence by as few as one or two base pairs for each rare target sequence, the method comprising: (a) preparing a primer-dependent amplification reaction mixture comprising the sample, a DNA polymerase, deoxyribonucleotide triphosphates, an amplification buffer, a homogeneous fluorescence detection means for detecting the amplification product, and a primer pair for each rare target sequence consisting of a first primer and a second primer specific for the rare target sequence but mismatched to the closely related sequence, (b) repeating a cycle of the primer-dependent amplification reaction mixture to amplify each rare target sequence present in the sample, and (c) detecting the rare target sequence by measuring the fluorescence intensity from the homogeneous fluorescence detection means. The first primer can be an allele-discriminating multi-part primer comprising, from 5' end to 3' end, a first anchor sequence, a first bridge sequence, and a first foot sequence mismatched to the closely related sequence at least at its 3 '-end or 3 '-second to last nucleotide. The second primer can be an allele-discriminating primer.

[0011] In the method, the first primer, the second primer, or both can be a superselective primer mismatched to the closely related sequence at least at its 3 '-end or 3 '-second to last nucleotide. Each primer can contain a 3 '-end interrogation nucleotide complementary to the rare target sequence but mismatched to the non-target sequence.

[0012] In the method described above, the cycle can comprise a temperature cycle in an asymmetric polymerase chain reaction (PCR) method. In one embodiment, the detecting step can comprise real-time detection. In another embodiment, the PCR method can be a digital PCR method, and the detecting can comprise end-point detection.

[0013] In the methods described above, the at least one rare target sequence in the sample comprises at least two different rare target sequences. In this case, the cognate fluorescence detection tool can comprise at least two different cognate fluorescence detection probes directed to the at least two different rare target sequences, respectively. The at least two rare target sequences can comprise a set of rare target sequences, and the probes directed to the rare target sequences in the set are labeled with the same color. The probes can be color-coded.

[0014] In some embodiments, each different non-target sequence differs from its corresponding rare target sequence by a single base pair, and both the first and second primers are mismatched to the single base pair. Each second primer can be a multipart primer comprising, from 5' end to 3' end, a second anchor sequence, a second bridge sequence, and a second foot sequence. The cognate fluorescence detection tool can comprise a probe directed to each rare target sequence. The first or second primer, or both the first and second primers, directed to each rare target sequence can contain a 5'-tag sequence, and the complement of each 5'-tag sequence is the target of the probe or the probe pair.

[0015] Each probe as described above can comprise a sequence complementary to the complement of the first bridge sequence or the complement of the second bridge sequence. Examples of probes include shared-stem molecular beacons.

[0016] In the methods described above, the at least one rare target sequence differs from its corresponding non-target sequence by a first base pair and a second base pair occurring in cis in the same gene. In this case, the first primer can be complementary to the first base pair, and the second primer can be complementary to the second base pair. In some embodiments, the at least one rare target sequence in the sample can comprise two or more rare target sequences, and the cognate fluorescence detection tool can comprise at least one cognate fluorescence detection probe directed to each rare target sequence. Examples of the cognate fluorescence detection tool directed to each rare target sequence comprise inter-primer specific molecular beacon probes.

[0017] The primer-dependent amplification reaction mixture described above can further comprise an effective amount of a selective enhancement reagent, such as a Hofmeister salt. Examples include tetramethylammonium chloride (TMAC) and bis-tetramethylammonium oxalate.

[0018] The application also provides kits or compositions (e.g., reaction mixtures) of reagents for performing the above-described methods. The kits or compositions can include one, two, or more reagents selected from the above-described primers, nucleic acid polymerase, deoxyribonucleotide triphosphates, and detection agent. Examples of detection agents include homogenous fluorescence detection tools, such as molecular beacon probes for each rare target sequence.

[0019] The method according to the application uses for each of the at least one rare target sequence a pair of first and second allele discriminating primers, at least one of which is a multipartite first primer, preferably a hyperspecific primer. Both primers in each pair of amplification primers are specific for the target sequence but mismatched to the closely related sequence. Each multipartite first primer is mismatched to the closely related sequence at least at its 3'-terminal or 3 '-penultimate nucleotide. In certain preferred embodiments, each allele discriminating second primer can be so as well. Preferably, in both the forward and reverse primer of each primer pair, the interrogating nucleotide is the 3'-terminal nucleotide. The method according to the application optionally comprises amplification and detection of an unrelated wild-type gene sequence for quantification.

[0020] The method can be divided into two general types:

[0021] In the first type, each rare target sequence contains a single difference, such as one mutation (as a result of a deletion, insertion, or nucleotide change such as a single nucleotide polymorphism (SNP)). In the case of an SNP, the rare target sequence contains a single base pair that differs from the closely related sequence, e.g., where the mutant sequence differs from the wild-type sequence by a single base pair. In this case, both primers have an interrogating nucleotide specific for the single base pair that differs in the rare target sequence; i.e., the forward primer has a 3'-terminal or 3 '-penultimate interrogating nucleotide complementary to the single base pair of one of the two strands of the mutant nucleic acid, and the reverse primer has a 3'-terminal or 3 '-penultimate interrogating nucleotide complementary to the other mutant nucleotide of that base pair of the other of the two strands of the mutant nucleic acid, such that one primer binds to the target strand and the other primer binds to the complementary target strand. The amplification reaction mixture containing the sample is subjected to a plurality of cycles of primer-dependent amplification reactions, and fluorescence is detected in real time or after amplification (end-point detection, which is used for digital PCR). During amplification, the nucleic acid polymerase cannot incorporate the incorrect nucleotide into an amplicon produced from a non-target sequence (producing a mutant sequence that does not exist in the original sample) because the primers initiate synthesis only at positions other than where the target mutation would exist.

[0022] The second type addresses the need to sometimes determine whether two different mutations occur on different sister chromosomes (i.e., in trans), in which case there are two different versions of the encoded protein that can be produced, but each version is encoded by only one of the two mutations; or whether the two mutations exist and occur on the same chromosome (in cis), in which case the protein encoded by the gene on the mutant chromosome contains both mutations. In the second type, the rare target sequence contains two base pairs that are different from the wild-type sequence, where the base pairs occur in the same exon or in different exons, and can occur in either cis relationship or in trans relationship. In this target type, the forward primer has an interrogating nucleotide specific to one of the mutations, and the reverse primer has an interrogating nucleotide specific to the other mutation. For example, the occurrence of either a T790M mutation or a C797S mutation in exon 20 of the EGFR gene of a patient's non-small cell lung cancer introduces one amino acid substitution into the encoded EGFR protein, which indicates that first-line therapy (Gefitinib or Erlotinib) will not work, and suggests that the use of Osimertinib will be effective (Lamb and Scott (2017) Targeted Oncology 12:555-562). However, it has recently become apparent that if both the T790M mutation and the C797S mutation occur in the EGFR gene on the same chromosome (i.e., in cis), which results in two amino acid substitutions occurring in the same EGFR protein, then Osimertinib will not kill these cancer cells, and only Brigatinib will be effective (Uchibori et al. (2017) Nature Communications 18:14768). In the method of the present invention to make this determination, one primer in the primer pair interrogates the T790M mutation in one fragment strand, the (+) strand, while the other primer interrogates the C797S mutation in the complementary strand, the (-) strand. Only fragments containing both mutations will be amplified. Although the detection probe can target the complement of the bridge sequence of the limiting primer or the complement of the 5 '-tag sequence of the limiting primer, our preferred embodiment uses a homogenous fluorescence detection probe, preferably a molecular beacon probe, that targets the portion of the amplicon between the sequences to which the two probes bind (i.e., an inter-primer specific probe).

[0023] The method according to the present application is highly selective, highly sensitive and exhibits improved robustness. High selectivity means the ability to detect a rare sequence in a mixture with closely related sequences when the ratio is as low as 1 / 1,000. High sensitivity means the ability to detect as few as ten copies of a rare sequence in such a mixture. Selectivity and sensitivity can be condensed into a single requirement, namely the ability to detect as few as 10 copies of a rare target sequence in a mixture containing 10,000 copies of closely related sequences that differ by one or two nucleotides. The disclosed international patent applications WO 2014 / 124290 (14 August 2014) and WO 2017 / 176852 (12 October 2017) and US patent 9,909,159 describe real-time PCR methods utilizing primer pairs consisting of a superselective forward primer and a conventional reverse primer, including methods capable of detecting as few as 10 copies of a rare target sequence in a mixture containing 10,000 copies of such closely related sequences, wherein the difference (ACt) between the threshold cycle from a sample containing only 10,000 copies of closely related sequences and the threshold cycle from a sample additionally containing ten copies of a rare target sequence typically differs by two or a few cycles, with some variability between replicates. The method according to the present application has improved robustness without sacrificing the aforementioned selectivity and sensitivity. By "improved robustness" is meant that the method meets one of the following two criteria: a sample containing only 10,000 copies of closely related sequences does not reach the threshold fluorescence intensity at least 55 amplification cycles, or the ACt between a sample containing only closely related sequences and a sample additionally containing 10 copies of a rare target sequence is at least 5 cycles greater than when the primer pair contains the same superselective primer and conventional primer.

[0024] As mentioned above, the method and kit according to the present application utilize a pair of allele discriminating primers, where each primer is complementary to the rare target sequence but mismatches with closely related sequences that differ from the rare target sequence by one or two nucleotides. In all cases, the first primer is a multipartite primer, preferably a superselective primer. The second primer is an allele discriminating primer. It can be, for example, a superselective primer or another multipartite primer, an allele discriminating hairpin primer or an ARMS primer. In the first type of embodiment, the second primer, like the first primer, contains one interrogating nucleotide at or near its 3' end. Multipartite primers and ARMS primers are capable of meeting this requirement, but allele discriminating hairpin primers are not. However, the second type of embodiment does not have this requirement for the second primer, so that a multipartite primer (preferably a superselective primer), an ARMS primer or an allele discriminating hairpin primer can be used.

[0025] The assay method according to the present application is a primer-dependent amplification and detection method. The primer-dependent amplification reaction that can be used in the method of the present application can be any suitable exponential amplification method, including polymerase chain reaction (PCR), ligase chain reaction (LCR), nicking enzyme amplification reaction (NEAR), strand-displacement amplification (SDA), nucleic acid sequence-based amplification (NASBA), transcription-mediated amplification (TMA), and rolling circle amplification (RCA). The preferred method uses PCR.

[0026] The primer-dependent amplification and detection method according to the present application can use asymmetric DNA amplification, such as asymmetric PCR. Symmetric DNA amplification can also be used, but we prefer asymmetric amplification. In the asymmetric PCR amplification method, one primer, the limiting primer, is present in limited amounts so as to be exhausted before amplification is complete, preferably at or shortly after the threshold cycle, after which linear amplification occurs, which uses the remaining primer, the excess primer. The asymmetric PCR method that can be used in the present application is LATE-PCR (see, e.g., European Patent EP 1,468,114; and Pierce et al. (2005) Proceedings of the National Academy of Sciences of the United States of America 102:8609-8614).

[0027] The preferred method also includes digital PCR (see, e.g., Vogelstein and Kinzler (1999) Proceedings of the National Academy of Sciences of the United States of America 96:9236-9241), in which a single PCR assay mixture is divided into a very large number of single wells or droplets, such that only one target molecule (or no target molecule) is present in each well or droplet, and thus it requires detection of an amplicon from a single mutant template molecule present in a single well or droplet that can also contain the relevant wild-type molecule.

[0028] If the amplification reaction uses an RNA-dependent DNA polymerase (one example is NASBA), the amplification reaction can be isothermal. We refer to the repeated rounds of amplification product synthesis as "cycles", but they are not thermal cycles for NASBA. For such amplifications, the "target sequence of interest" and "non-target sequence of interest" that are primed by the multipartite primer according to the application are RNA sequences that are present in the original sample and amplification reaction mixture along with the DNA polymerase and the multipartite primer.

[0029] If the amplification reaction uses a DNA-dependent DNA polymerase (one example is PCR), the original sample can contain DNA or RNA targets. For such amplifications, the "target sequence of interest" and "non-target sequence of interest" that are primed by the multipartite primer useful in the methods of the application are DNA sequences that are produced from RNA sequences that are present in the original sample or that occur in the original sample by reverse transcription. If reverse transcription is performed using the multipartite primer, the "target sequence of interest" and "non-target sequence of interest" are RNA as well as cDNA. If reverse transcription is performed using a separate outsider primer, the "target sequence of interest" and "non-target sequence of interest" are cDNA. In either case, the "target sequence of interest" and "non-target sequence of interest" are nucleic acid sequences that are present in the amplification reaction mixture along with the DNA polymerase and the multipartite primer.

[0030] Primer-dependent amplification reactions comprise repeated thermal cycles of primer annealing, primer extension, and strand denaturation (strand melting). Primer annealing can be performed at a temperature lower than the primer extension temperature (e.g., three-temperature PCR), or primer annealing and primer extension can be performed at the same temperature (e.g., two-temperature PCR). The overall thermal profile of the reaction can comprise repetition of specific cycles, or the temperature / time can vary during one or more cycles. For example, once amplification has begun and the priming sequence of the multipartite primer has been extended, a higher annealing temperature appropriate for the longer primer can be used to complete the amplification reaction.

[0031] A preferred method according to the application is a primer-dependent amplification and detection method, most preferably an asymmetric method, that is capable of asymmetric amplification and detection of as few as ten copies of at least one rare DNA target sequence of interest in a sample mixture that contains 10,000 copies of a closely related non-target sequence of interest that differs from the rare target sequence of interest by as few as one or two nucleotides for each rare target sequence, the method comprising:

[0032] (a) preparing a primer-dependent amplification reaction mixture comprising a sample, a DNA polymerase, deoxyribonucleotide triphosphates, an amplification buffer, a homogenous fluorescent probe for detecting an amplification product, and a pair of allele-specific amplification primers specific for each rare target sequence, each primer pair comprising an allele-discriminating moiety first primer and an allele-discriminating second primer that are target gene-specific but mismatched to a closely related sequence at at least the 3 '-end or 3 '-penultimate nucleotide;

[0033] (b) repeating the cycling of the reaction mixture to amplify each rare DNA target sequence present in the sample by the primer-dependent amplification method and detecting the sequence by measuring the fluorescence intensity from the distinguishable labeled probe that targets its amplification product;

[0034] wherein if the method is tested in a first sample containing 10,000 copies of a non-target sequence and a second sample containing 10 copies of a target sequence in a mixture with 10,000 copies of a non-target sequence, then (a) the fluorescence signal from the first sample is suppressed at 55 amplification cycles, or (b) the threshold cycle difference (ACt) between the two amplifications is at least five cycles greater than that obtained in the same test in which the second primer is replaced with a conventional primer.

[0035] If the second primer is a multi-part primer, it is mismatched to the closely related sequence at at least its 3 '-end or 3 '-penultimate nucleotide. If the second primer is an ARMS primer, it is mismatched to the closely related sequence at its 3 '-end nucleotide. If the second primer is a hairpin primer, it is mismatched to the closely related sequence at one nucleotide in its single-stranded loop.

[0036] In the examples described below, we used a typical non-proprietary buffer containing KCl, Tris-HCl (pH 8.0), and MgCl2. Some of the contents of amplification buffers are considered proprietary by the suppliers. Since such buffers are functionally equivalent, they can also be used. In some preferred embodiments, the amplification reaction mixture comprises an effective concentration of a selective enhancement reagent, preferably tetramethylammonium chloride (TMAC).

[0037] In certain preferred methods, each primer contains a 3 '-end interrogating nucleotide that is complementary to the target sequence but mismatched to the non-target sequence. In some embodiments, one of the pair of primers contains a 5 '-tag sequence, and the complement of the 5 '-tag sequence is the target of the probe.

[0038] Detection can be performed using similar detection tools that detect amplification products. Detection tools may contain a single similar detection probe, multiple similar detection probes all labeled with the same color; labeled primers, such as Scorpion primers or LUX primers; or embedded DNA dyes, such as… Green (if there is no need to distinguish amplification products from multiple rare target sequences, such as when there is only one target or a set of targets to be detected, or when it is only necessary to detect the presence of any one of multiple targets). Otherwise, for each target (or group of targets) to be detected, there are similar fluorescent detection probes of different colors. Similar fluorescent detection probes can be, for example... Probes, minor groove binder (MGB) probes, molecular beacon probes, or combinations of MNAzyme / cleavable probes (WO 2013 / 123552). Our preferred detection method is via a molecular beacon probe that targets an amplicon (amplification product) sequence complementary to a 5'-tag sequence contained in a restriction multipart primer sequence, or an amplicon sequence complementary to a bridging sequence in a restriction multipart primer.

[0039] As described above, the functional characteristics of such assays are (a) their ability to detect as few as 10 copies of each rare target sequence in the presence of 10,000 copies of its closely related large sequence, and (b) their improved robustness in distinguishing between (i) only 10,000 copies of closely related sequences and (ii) 10,000 copies of closely related sequences and 10 copies of rare target sequences, according to the following test. If a reaction mixture containing two samples is subjected to real-time PCR using the first and second primers according to the invention, and also performed using the same method except that the conventional primers are replaced with the second allele-discriminating primers, either of the following results is obtained: (1) for the sample containing only closely related sequences, using the primer pairs according to the invention, the fluorescence does not rise above the background after 55 cycles, or (2) the difference in fluorescence intensity above the background between the two samples (ΔCt) is at least five amplification cycles greater when using the primer pairs according to the invention than when the conventional primers are replaced.

[0040] Sensitivity in preferred embodiments can be improved by inclusion of a selectivity enhancing reagent, preferably tetramethylammonium chloride (TMAC) or another Hofmeister salt, in the amplification reaction mixture, disclosed in International Patent Application WO 2017 / 176852 (12 October 2017). The reagent can be included in the amplification reaction mixture to improve selectivity against wild type or other closely related non-target sequences. Such reagents are added at an effective concentration, as determined by trial and error. Certain preferred hyperselective primers used in the methods of the invention have relatively long feet, e.g. 9-11 nucleotides, and also form large bubbles of 32-48 nucleotides, which typically include a long bridge sequence in the range of 15-24 nucleotides. For reaction mixtures containing such hyperselective primers, TMAC can be included at a relatively high concentration, such as 50 mM or 60 mM. For reaction mixtures containing hyperselective primers with shorter feet, such as 5-7 nucleotides, TMAC can be added at a smaller concentration, such as 10 mM, 20 mM or 30 mM, to avoid deleterious effects on the amplification reaction. The assays described in the preceding paragraph can be used to determine whether and how much selectivity enhancing reagent to include in a particular reaction mixture, e.g. by comparing results obtained at different concentrations (e.g. 0, 10 mM, 50 mM and 60 mM) of TMAC. By "effective concentration" is meant a concentration that allows for the assay, and, optimally, a concentration that improves the assay results the most and does not significantly interfere with the amplification reaction.

[0041] The methods according to the present application are particularly suitable for multiplexing, with the ability to amplify multiple rare target sequences that can be present in a sample. Different embodiments have different purposes and features. For example, certain multiplexed embodiments can be designed to be able to detect the presence of each of at least two mutations in a sample containing genomic DNA fragments in the presence of a large amount of wild-type sequence. A different primer pair can be used for each target sequence, with a uniquely colored fluorescent probe, preferably a molecular beacon probe, targeting a sequence of the amplicon that is not present in the probe itself or in the wild-type correct amplification product. For the first type of method, the target of each probe is preferably the complement of the bridge or 5' tag of each different restriction primer. For the second type of method, the target of each probe can also be the complement of the bridge or 5' tag, but we prefer to use inter-primer specific probes. Alternatively, a different primer pair can be used for each of a different set of target sequences, where the presence of one or more mutations in a set is of technical importance, e.g., for the treatment of a cancer patient. The above singleplex and multiplex methods can generally be performed in a spectrofluorometric thermal cycler, which limits the number of distinguishable fluorescent labels to at most seven, or sometimes ten, of the commonly available thermal cyclers are 5-color instruments. The reaction mixture for the assay to detect a different set of rare target sequences contains a different multi-part restriction primer for each mutant target sequence, and optionally an unrelated wild-type gene sequence for quantification purposes, where each multi-part restriction primer in a primer set has the same 5'-tag sequence, and each set has a different 5'-tag sequence.

[0042] The second type of method, i.e., detecting two mutations in cis, can have the possibility that either or even both of the two rare base pairs have multiple possibilities that can even produce the same amino acid change. For example, there can be a situation where one base pair change can be constant, but the other base pair change can be variable, i.e., change X, change Y, or change Z, e.g., at the same position or at a subtly different position. When this occurs, the multiplexed assay method can have allele discriminating primers specific for each possibility; for example, one primer specific for the constant change, but three primers specific for the variable changes (one for X, one for Y, one for Z). In addition, the inter-primer specific probe will signal the presence of a change, but it will not identify what the change is. For this purpose, each of the primers (for X, Y, and Z) can have a unique 5' tag or unique bridge, the complement of which will be the target of a different probe with a unique color.

[0043] Other embodiments of multiplex assays, especially the first type of method, are screening assays whose purpose is to determine which of a list of many different mutant target sequences is present in a sample, or to determine that none of these mutant target sequences is present in the sample. In these assays, whichever mutant target sequence is present is exponentially amplified, preferably in a polymerase chain reaction, and the resulting amplicon, which is produced only if the mutant target sequence is present in the sample, is detected with a fluorescently labeled hybridization probe. In such embodiments, for each possible mutant target sequence, there is a multi-part restriction primer, preferably a hypersselective primer, each with a different 5'-tag sequence whose complement is the target of the hybridization probe. Often, the number of mutant target sequences on the list will exceed the number of different fluorescent colors that the detection instrument can distinguish, and this can occur in other multiplex assays as well. We address this problem in either of two ways. One way is to use color-coded homogenous detection probes, preferably the color-coded molecular beacon probes disclosed in International Patent Publication WO 2004 / 099434 A3, U.S. Patent 7,385,043, and Marras et al. (2019) PloS ONE 14:e0213906. Briefly, a batch of a probe is divided into the number of aliquots in the coding scheme, usually two or three, and then each aliquot is labeled with a different color fluorophore, and the aliquots are then recombined to create a batch that contains two or three colors of coding. The second way to overcome the color limitations of a fluorescent spectroscopic thermal cycler is to use what we call "thermospecific" molecular beacon probes, whose probe-target hybrid has a different melting temperature (Tm). For example, if one wants to test a liquid biopsy sample for the presence of one or more of 35 different target sequences on a five-color fluorescent spectroscopic thermal cycler, one can divide the 35 different multi-part restriction primers each specific for a different target sequence into five groups of seven. All seven of each of the five groups have different 5' tags whose complements are the targets of seven different thermospecific molecular beacon probes, all of which are labeled with the same fluorophore, but all of which produce probe-target hybrids with distinguishable melting temperatures (Tm). Thus, each of the 35 different target sequences, if present, can be identified by the combination of fluorescent color and Tm determined in a post-amplification (end-point) thermal analysis.

[0044] As an alternative to the first type of assay by real-time amplification and detection assay, they can be performed by digital PCR assay methods, including assays performed in many different reaction wells in a thermal cycler, and droplet digital PCR (ddPCR) assays performed in many different droplets in a thermal cycler. In both cases, detection of the resulting amplicons after amplification (end-point detection) is typically performed in a separate detection instrument, such as the Bio-Rad QX200 TM droplet digital PCR system or the Stilla Technologies Naica TM System. The basic principle of the digital PCR assay is that the reaction mixture containing the sample can be diluted to the extent that for each rare target molecule to be detected, there is typically only one target DNA molecule in the well or droplet (rare target of interest or large number of non-target of interest) (or no target molecule in the well or droplet), and there are a large number of wells or droplets; however, some wells or droplets can contain no target molecule, two or three non-target of interest, or one target of interest plus one or two non-target of interest. PCR amplification is then performed simultaneously in each well or droplet. A fluorescently labeled probe present in all wells or droplets binds to the amplicon produced in each well or droplet if it contains the target of interest and emits bright fluorescence in a specific color or color code, indicating that the well or droplet contains the target of interest and determining which one it is. Wells or droplets with fluorescence intensity above a threshold intensity (background) selected at the completion of amplification are considered positive in a specific color or color code. The number of droplets or wells that light up in the same color or color code provides an accurate measure of the number of the corresponding target molecule in the original sample. This method is sensitive to the extent that even a single DNA fragment containing the target sequence in the well or droplet can be detected.

[0045] Classical droplet digital PCR has been used to detect and quantify rare somatic mutations associated with cancer diagnosis, prognosis, and treatment. See Sanmamed et al. (2015) Clinical Chemistry 61 :297-304. In order to separate the rare mutant target molecules from the much larger number of related wild-type molecules, more than one million droplets are required. See, e.g., Hindson et al. (2011) Analytical Chemistry 83:8604-8610. Such a large number of droplets is necessary because the wild-type targets in the sample are much more numerous than the rare related mutant targets, which differ from the wild-type targets usually only by a single nucleotide polymorphism, and because the probes, which are designed to bind to a subsequence within the amplicon containing the mutation, occasionally will bind to the corresponding sequence in an amplicon produced from a related wild-type target. It is desirable to have such a large number of droplets so that a droplet containing a mutant target is extremely unlikely to also contain one or more related wild-type targets. This ensures that there will not be a droplet containing enough wild-type targets to produce a signal strength similar to that which would be produced if the droplet were mistakenly thought to contain a related mutant target. In other words, if the original sample were divided into too few droplets, a droplet containing some wild-type target sequence and no related mutant target sequence could be mistaken for a droplet containing a mutant target.

[0046] However, when using a digital PCR embodiment of the method of the application to detect a rare mutant target molecule in a sample, far fewer droplets or wells are required (e.g., only 10,000 to 30,000 droplets are required) because the primer pair used will not produce a detectable amplicon from the relatively few closely related wild-type DNA molecules that can also be present in the well or droplet. In the digital PCR assay method according to the application, detection can be performed in a thermal cycler, flow cytometer, or microscope, and one of the above detection instruments.

[0047] The application also includes kits for performing the above methods. Such kits can include one or more pairs of allele-specific primers directed to one or more target rare target sequences, dNTPs, primer-dependent polymerase, detection probes directed to each target rare target sequence (or in some embodiments to a group of rare target sequences), and other reagents required for amplification, especially amplification buffer. One primer in each primer pair can be a multipartite primer, and the other primer in each primer pair can be a superselective primer or other multipartite primer, or an ARMS primer.

[0048] The selectivity of a superselective primer can be maximized for a particular target by adjusting the foot length and bulge circumference, with larger bulges and shorter feet generally increasing selectivity. In addition, a superselective primer can be fine-tuned for maximum discrimination by adjusting the length and nucleotide sequence of the bridge sequence, and can be fine-tuned to ensure that a given Ct value for any primer pair in the same assay reflects the same number of nucleic acid targets in the sample.

[0049] BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 A design of the method for detecting rare copy mutations in Example 1 is shown, which uses a pair of superselective primers complementary to the mutation, detected by a molecular beacon probe targeting the complement of the 5'-tag on the superselective forward primer.

[0051] Figure 2 A and 2B show the results of the real-time PCR assay of Example 1, where upper panel A shows amplification and detection using a superselective forward primer whose 3 '-inquiry nucleotide is complementary to the mutant sequence and a regular reverse primer that is complementary to both the mutant sequence and the wild-type sequence closely related to the mutant sequence downstream of its target mutation; and lower panel B shows the results of the same real-time PCR assay using a pair of superselective primers whose 3 '-inquiry nucleotides of both primers are complementary to the target mutation.

[0052] Figure 3 A design of the method for detecting rare copy mutations in Example 2 is shown, which uses a pair of superselective primers complementary to the mutation, detected by a shared stem molecular beacon probe targeting the complement of the bridge sequence of the superselective forward primer.

[0053] Figure 4 A and 4B show the results of the real-time PCR assay of Example 2, where upper panel A shows amplification and detection using a superselective forward primer complementary to the EGFR T790M mutation and a regular reverse primer that is complementary to both the mutant sequence and the wild-type sequence closely related to the mutant sequence downstream of its target mutation; and lower panel B shows the results of the same real-time PCR assay using a pair of superselective primers whose 3 '-inquiry nucleotides of both primers are complementary to the T790M target mutation.

[0054] Figure 5 A design of the method for determining whether two mutations occur on the same chromosome (cis) in Example 3 is shown, which uses a superselective forward primer complementary to one mutation and a superselective reverse primer complementary to the other mutation.

[0055] Figure 6 A, 6B, 6C, 6D, and 6E show the results of the real-time PCR assay of Example 3, in which the reaction mixture contains a superselective forward primer for the EGFR T790M mutation, a superselective forward primer for the EGFR C797S mutation, and a conventional reverse primer. (A) 10 copies of T790M, (B) 10 copies of C797S, (C) 10 copies of T790M and C797S (in cis), (D) 10 copies of T790M and 10 copies of C797S (in trans), and (E) wild type only.

[0056] Figure 7 A, 7B, and 7C show the results of the real-time PCR assay of Example 3, in which the reaction mixture contains a superselective forward primer complementary to one mutation, and a superselective reverse primer complementary to the other mutation. (A) 10 copies of T790M and C797S (in cis), (B) 10 copies of T790M and 10 copies of C797S (in trans), and (C) wild type only.

[0057] Figure 8 The design of the method for determining whether two mutations occur on the same chromosome (in cis) in Example 4 is shown, which uses a superselective forward primer complementary to one mutation and an ARMS reverse primer complementary to the other mutation.

[0058] Figure 9 A, 9B, and 9C show the results of the real-time PCR assay of Example 4, in which the reaction mixture contains a superselective forward primer for the EGFR T790M mutation and an ARMS reverse primer for the EGFR C797S mutation. (A) 10 copies of T790M and C797S (in cis), (B) 10 copies of T790M and 10 copies of C797S (in trans), and (C) wild type only.

[0059] Figure 10 The design of the method for detecting rare copy mutations in Example 5 is shown, which uses an ARMS forward primer complementary to the mutation and a superselective reverse primer complementary to the mutation, detected by a molecular beacon probe targeting the complement of the 5'-tag on the ARMS forward primer.

[0060] Figure 11A, 11B, and 11C show the results of real-time PCR assays of Example 5, where upper panel A shows amplification and detection using a pair of superselective primers; lower left panel B shows amplification and detection using a restriction superselective forward primer and excess ARMS reverse primer; and lower right panel C shows amplification and detection using a restriction ARMS forward primer and excess superselective reverse primer, where in all three cases both primers are complementary to a single mutant base pair in the target sequence.

[0061] Figure 12 The design of the method for detecting rare copy mutant in the presence of large copy normal human genomic DNA in Example 6 is shown, which uses a pair of superselective primers complementary to the mutation, where detection is accomplished using a conventional molecular beacon probe targeting the complement of the 5 '-tag on the target restriction superselective primer labeled with one fluorescent color; and where the assays also contain superselective primers and a conventional reverse primer for the beta-actin reference gene in normal human genomic DNA, where detection of the reference gene is accomplished simultaneously using an inter-primer specific molecular beacon labeled with a different fluorescent color, enabling the relative abundance of the rare mutation to be assessed by comparing the difference in threshold values of the mutation and reference gene.

[0062] Figure 13 A, 13B, 13C, and 13D show the results of real-time PCR assays of Example 6, where all samples contain the same number of large copy whole human genome, and each of the four panels shows the results obtained using samples that also contain different numbers of mutant target DNA, including a sample containing no mutant target DNA. (A) 500 copies of G719C, (B) 50 copies of G719C, (C) 5 copies of G719C, and (D) 0 copies of G719C.

[0063] Definitions and Nomenclature

[0064] The following definitions apply as used in the specification and claims of this patent application:

[0065] An allele discriminating "multi-part primer" means a nucleic acid (e.g., DNA) amplification primer having an internal sequence, which we call the "bridge sequence," that is not sufficiently complementary to a target sequence to hybridize thereto under primer annealing conditions, and which is sandwiched between two target-complementary sequences, which we call the "anchor sequence" and the "foot sequence." The anchor sequence, like a conventional primer, has sufficient complementarity to the target sequence (both the target-of-interest and the non-target-of-interest) to hybridize thereto during the primer annealing step of a primer-dependent amplification reaction for which the primer is designed, typically 15-40 (e.g., 17-35, or 20-30) complementary nucleotides. The foot sequence is sufficiently complementary to the rare target-of-interest (e.g., mutant sequence) to hybridize thereto during primer annealing, but mismatches the abundant, closely related non-target-of-interest (e.g., wild-type sequence) at at least one of its 3' terminal and 3' penultimate nucleotides. We call the nucleotide that is complementary to the target-of-interest but that mismatches the closely related non-target-of-interest the "interrogating nucleotide." The foot sequence can contain intentionally introduced nucleotides near its 3' terminal end that mismatch both the target-of-interest and the non-target-of-interest to destabilize the foot and increase its allele discrimination. The foot sequence can typically have 5-12 (e.g., 5-10, 6-12, 6-9, or most preferably 8-9) nucleotides complementary to the target-of-interest. The bridge sequence can be any length between 1-50 (e.g., 5-40, 10-30, 15-30, 20-30, or most typically 18-22 or 10-14) nucleotides in length. When the multi-part primer is hybridized to its target sequence, there is a region in the target sequence, opposite the bridge sequence, that does not hybridize to the bridge, which we call the "intervening sequence," which can be any length between 1-100 nucleotides in length. Together, the bridge sequence and the intervening sequence form a "bulge" in the primer-target hybrid, whose circumference, ignoring any secondary structure, is the sum of the lengths of the bridge sequence and the intervening sequence plus four nucleotides.

[0066] A "superselective" primer is an allele discriminating multi-part primer that is capable of detecting as few as 10 copies of a rare target sequence in the presence of 10,000 copies of a closely related sequence that differs by as little as a single base pair when the primer is used as a limiting primer in a PCR amplification. A superselective primer has a sequence comprising, in the 5' to 3' direction, the following three consecutive nucleic acid sequences (e.g., DNA sequences) that are copied by extension of another primer:

[0067] an anchor sequence long enough that it can hybridize to the mutant sequence or other closely related DNA target sequence as well as to the relevant wild-type or other abundant DNA target sequence during primer annealing, typically in the range of 15-40 nucleotides, typically 20-30 nucleotides in length;

[0068] a unique bridge sequence at least 6 nucleotides long that does not hybridize to the target sequence of interest of the primer or to any other closely related sequence during primer annealing; and

[0069] a unique foot sequence that can be 6 to 12 nucleotides long and that is perfectly complementary to the target DNA sequence of interest but mismatched at one or more nucleotides (one or more interrogating nucleotides) to the closely related sequence, at least one of the one or more nucleotides being the 3 '-terminal nucleotide or the 3 '-penultimate nucleotide. The foot sequence

[0070] The superselective primer can also have one or more of the following structural and functional features in a polymerase chain reaction (PCR) amplification and detection assay:

[0071] (i) if both the anchor sequence and the foot sequence hybridize to the target sequence of interest of the primer, the primer-target hybrid comprises, in the 5' to 3' direction of the primer: an anchor-target hybrid, a single-stranded bulge, and a foot-target hybrid, the bulge having a circumference of 18-50 nucleotides and being formed by an intervening sequence of at least 8 nucleotides long in the target DNA sequence and that does not hybridize to the bridge sequence during primer annealing;

[0072] (ii) the bulge separates the foot-target hybrid from the anchor-target hybrid, and the separated foot-target hybrid is a weak hybrid that makes it less likely that a copy of the target DNA sequence of interest will be amplified by at least two, preferably at least five cycles, as evidenced by a threshold (Ct) that is delayed compared to the Ct that would occur when using a conventional primer that does not contain any bridge DNA sequence;

[0073] (iii) the multipartite primer will initiate the amplification of a copy of any closely related mutant target DNA sequence or a relevant wild-type target DNA sequence during PCR amplification less likely than it will initiate the amplification of a copy of its target sequence of interest by at least 1,000-fold, as evidenced by a threshold difference (ACt) of at least ten thermal cycles;

[0074] (iv) the multipartite primer that has generated an amplicon strand has a bridge sequence and a foot sequence that are perfectly complementary to the complementary strand of the amplicon strand; and

[0075] (v) the length and sequence of the bridge sequence of each multi-part primer, together with the length of the intervening sequence of its target sequence, results in a threshold (Ct) observed from a sample containing only 10 copies of the target DNA sequence that will occur within 40-65 or preferably 55 exponential amplification cycles and will be at least two cycles less than the Ct observed from a sample containing no copies.

[0076] An allele discriminating "hairpin" primer is a stem-loop oligonucleotide, like a molecular beacon probe, that contains a single-stranded region ("loop") flanked by complementary sequences ("arms") that hybridize to each other to form a double-stranded region ("stem"). The loop and 3' arm of a hairpin primer are sufficiently complementary to the target sequence to hybridize thereto and initiate copying under primer annealing conditions. An allele discriminating hairpin primer contains an interrogating nucleotide at or near the middle of the loop sequence.

[0077] An ARMS primer is an allele discriminating primer because its 3'-terminal nucleotide is an interrogating nucleotide. An ARMS primer can contain an intentionally introduced nucleotide near its 3' terminus that is mismatched to both the target sequence and the non-target sequence, thereby destabilizing the primer and increasing its allele discrimination.

[0078] A "conventional" primer is a single-stranded oligonucleotide 15-40 nucleotides long, more usually 20-30 nucleotides long, and perfectly complementary to the target. Conventional PCR primers are usually designed using any of a variety of computer programs.

[0079] Our convention for describing primer pairs is to refer to the limiting primer as the "forward" primer, which is complementary to the (-) template strand of the target, and to the excess primer as the "reverse" primer, which is complementary to the (+) template strand of the target. We do this for convenience. It should be understood that the limiting primer can be complementary to the (+) strand and the excess primer can be complementary to the (-) strand.

[0080] The limiting superselective primer in Example 1, whose sequence is:

[0081] 5'-ACCTGCCGTCAACACGTGCGCAGTAGACCATC-

[0082] TCTCTTGAGGATCTTGAAGGAAACTGAA-

[0083] CCTCTCCAACGAATCTCGAA-AAGTGCTGT-3' (SEQ ID NO: 1)

[0084] In the 5' to 3' direction, the primer contains four elements, separated by dashes (-). In our nomenclature, the primer is32 -28-20 / 13-8:1:0. Elements 32 represents a 32 nucleotide long 5'-tag sequence; the next element 28 represents a 28 nucleotide long anchor sequence; the next element 20 / 13 represents a 20 nucleotide long bridge sequence that is opposite an intervening sequence in a 13 nucleotide long target sequence; and the last element 8:1:0 represents a 9 nucleotide long (8+1+0) foot sequence with the following properties: 8 nucleotides from the 5' end that are complementary to both the target and non-target sequences of interest, one interrogating nucleotide that is complementary to the target sequence but mismatched to the non-target sequence, and 0 nucleotides 3' of the interrogating nucleotide that are complementary to both the target and non-target sequences (i.e., the interrogating nucleotide is the 3' terminal nucleotide in this primer). If the foot has a nucleotide that destabilizes it, as often occurs in ARMS primers, it will be represented in the sequence in italics and in the characterization by "m". For example, the foot sequence AAGTGCCGT-3' is written as 6:m1:1:1:0, indicating that it has 6 nucleotides at the 5' end that are complementary to the target; then 1 nucleotide, represented by "m", that is mismatched to both the target and non-target sequences; then 1 nucleotide that is complementary to the target; then the interrogating nucleotide; and finally the number of 3' nucleotides that are complementary to the target (here, 0). Because the bridge sequence is characterized not only by its length, but also by the length of the opposite intervening sequence, the size of the bulge circle can be determined as the length of the bridge sequence plus the length of the intervening sequence plus 4, because the bulge contains one hybridized base pair on each side. In the above sequence example, the circumference of the bulge is 37 nucleotides (20+13+4=37). DETAILED DESCRIPTION

[0086] Figure 1An embodiment of the method of the application of the first type is shown, in which the first primer is an allele discriminating multi-part primer and the second primer is an allele discriminating primer, both complementary to a single base pair mutation (SNP) in the rare mutant target sequence. In the upper panel, the primers are shown hybridized to the rare target sequence in a double-stranded template containing a plus (+) strand and a minus (-) strand (indicated by short vertical lines). Although only one primer must be a multi-part primer, preferably a hypersolective primer, the embodiment has a pair of hypersolective primers. As shown in the upper panel, each primer has an anchor sequence, an unhybridized bridge sequence opposite an intervening sequence in the template, and a foot sequence. The target sequence (for purposes of illustration, referred to as the EGFR gene, as described in Example 1) contains a single base pair that is different from the closely related sequence. For purposes of illustration, the base pair is designated an A nucleotide in the (-) template strand of the target sequence and a T nucleotide in the (+) template strand of the target sequence. That is, the mutation to be detected is a single nucleotide polymorphism (EGFR G719C) that occurs in exon 18 of the EGFR gene, which is the subject of Example 1. Each primer has an interrogating nucleotide, here a 3 '-terminal interrogating nucleotide, which is complementary to one of the nucleotides of the base pair. Figure 1 An embodiment is shown in which the amplification reaction is asymmetric. One primer, designated here as the forward primer and also the limiting primer shown in the middle panel, contains a 5 '-tag sequence that is not complementary to the target strand but is copied in the amplification reaction. The other primer, designated here as the reverse primer, and also the excess primer shown in the middle panel. Also shown in the upper panel is a cognate fluorescence detection probe, here a hairpin-shaped molecular beacon probe with a single-stranded loop and a double-stranded stem, in which one arm of the stem is labeled with a fluorophore (o) and the other arm of the stem is labeled with a quencher (·). In the embodiment shown, the molecular beacon is a "regular" molecular beacon probe, that is, one in which only its single-stranded loop is complementary to the probe's target, which in this case is the complement of the 5 '-tag sequence. The lower panel shows detection, which can be real-time detection or end-point detection, as used in digital PCR methods. The probe is shown hybridized to the (-) amplicon (i.e., the amplification product produced by extension of the excess primer), and the probe's target is the complement of the 5 '-tag sequence of the limiting primer. The fluorophore of the probe is separated from the quencher of the probe by hybridization of the probe to its target, and thus fluoresces (Tyagi et al. (1998) Nature Biotechnology 16:49-53).

[0087] Figure 3 An embodiment is shown in which the amplification reaction is asymmetric. One primer, designated here as the forward primer and also the limiting primer shown in the middle panel, contains a 5 '-tag sequence that is not complementary to the target strand but is copied in the amplification reaction. The other primer, designated here as the reverse primer, and also the excess primer shown in the middle panel. Also shown in the upper panel is a cognate fluorescence detection probe, here a hairpin-shaped molecular beacon probe with a single-stranded loop and a double-stranded stem, in which one arm of the stem is labeled with a fluorophore (o) and the other arm of the stem is labeled with a quencher (·). In the embodiment shown, the molecular beacon is a "regular" molecular beacon probe, that is, one in which only its single-stranded loop is complementary to the probe's target, which in this case is the complement of the 5 '-tag sequence. The lower panel shows detection, which can be real-time detection or end-point detection, as used in digital PCR methods. The probe is shown hybridized to the (-) amplicon (i.e., the amplification product produced by extension of the excess primer), and the probe's target is the complement of the 5 '-tag sequence of the limiting primer. The fluorophore of the probe is separated from the quencher of the probe by hybridization of the probe to its target, and thus fluoresces (Tyagi et al. (1998) Nature Biotechnology 16:49-53). Figure 1In the illustrated embodiment, the target of the probe is the complement of the bridge sequence of the restriction primer, not the complement of the 5'-tag sequence of the restriction primer, and thus the restriction primer does not comprise a 5'-tag sequence. In the illustrated embodiment, the molecular beacon probe, sometimes referred to as a "shared stem" molecular beacon (Tsourkas et al. (2002) Nucleic Acids Research 30:4208-4215), has one arm, in this case labeled with a quencher, which is also complementary to the target of the probe. Thus, as shown in the following figure, both the loop and this arm hybridize to the complement of the bridge of the restriction primer. For purposes of illustration, the base pair that occurs in the rare target but not in the closely related sequence is named nucleotide A in the (-) strand of the target sequence of interest and nucleotide T in the (+) strand of the target sequence of interest. That is, the mutation to be detected is a single nucleotide polymorphism (EGFR T790M) that occurs in exon 20 of the EGFR gene, the subject of Example 2.

[0088] In Figure 1 and Figure 3 , one detection probe is shown hybridized to one sequence that is the target of the probe. This does not exclude the inclusion of two sequences that are the target of the probe. For example, if the complement of the bridge sequence of one primer is the target of a first probe, the complement of the 5'-tag sequence of another primer can be the target of a different probe of the same color, in which case twice the number of probe copies can bind and emit fluorescence.

[0089] Example 1 describes an embodiment of the method according to the application as Figure 1 illustrated. The rare target sequence of interest is the mutation G719C in the EGFR gene. The detection of this mutation enables the use of particularly effective targeted therapies (erlotinib or gefitinib) to kill cancer cells containing this mutation present in the non-small cell lung cancer of the patient (Pao et al. (2004) Proceedings of the National Academy of Sciences of the United States of America 101 :13306-13311; Kobayashi and Hagiwara (2013) Targeted Oncology 8:27-33).

[0090] A multi-part first primer, here a superselective restriction primer, and an allele- discriminating second primer, here also a superselective primer, both interrogating a single base pair. The rare target sequence of interest (in this case the mutant G719C in the EGFR gene) differs from the large number of closely related non-target sequences of interest (in this case the wild type sequence) by a single base pair change. In this case, the A:T in the mutant (see Figure 1 ) differs from the C:G in the wild type. The amplification and detection method is a real-time PCR assay. For comparison, in Example 1, the excess of superselective primer (for convenience called reverse primer) is replaced by a regular PCR reverse primer (in this case the sequence of the anchor sequence of the superselective reverse primer) that is complementary to both the target and non-target sequences of interest.

[0091] In Example 1, the foot sequence of the superselective forward primer is 9 nucleotides long and the perimeter of the bulge formed when the primer is hybridized to the target sequence of interest is 37 nucleotides (20+13+4). The foot sequence of the reverse primer is also nine nucleotides long and the perimeter of the bulge formed when this primer is hybridized to the target sequence of interest is 32 nucleotides long (18+10+4). We have found that the method of the application using superselective primers with relatively long (8-12 nucleotides) feet and without destabilizing nucleotides and producing a relatively large bulge perimeter (28-50 nucleotides) benefits from the inclusion of a selectivity enhancing agent. In Example 1, 50 mM tetramethylammonium chloride is included in each amplification reaction mixture as an effective amount of selectivity enhancing agent.

[0092] The samples are subjected to PCR amplification using real-time fluorescence detection. One sample contains only 10,000 copies of the EGFR wild type sequence. The second sample contains 10 copies of the G719C mutant sequence in a mixture containing 10,000 copies of the EGFR wild type sequence. The third sample contains 100 copies of the G719C mutant sequence in a mixture containing 10,000 copies of the EGFR wild type sequence. Each sample is tested in duplicate reactions. The fluorescence intensity curves of the amplification reactions are shown in Figure 2 Table 1. For each primer pair, the average Ct values (of the two replicates) of the sample containing only the wild type and the sample containing 10 copies of the mutant are shown, as are the ΔCt values. By comparison, the lower panel B of Figure 2 shows the results of the method using two superselective primers both selected for a single SNP (Example 1) compared to the same restriction superselective forward primer and a regular reverse primer (Example 2). Figure 2The performance of the method of Figure A) above can be compared to the method of Example 1, which can indeed distinguish 10 mutant from 10,000 wild type and 10,000 wild type, but the performance of the method of the application is much more robust. Table 1 shows that the ΔCt for a primer pair comprising conventional primers is 2.95, while the ΔCt for a primer pair comprising two superselective primers is 12.87. This is an increase of nearly 10 cycles, demonstrating that the method of Example 1, which uses a pair of superselective primers, is a method according to the application.

[0093] Example 2 describes Figure 3 an embodiment of the method according to the application shown in Figure 2. The target sequence mutation is T790M, which is in exon 20 of the EGFR gene. The mutation is a single base pair substitution (SNP) of an A:T base pair for a G:C base pair (in an otherwise identical closely related wild type sequence). Detection of this mutation indicates that a commonly used targeted therapy (erlotinib or gefitinib) that can kill cancer cells containing any of a number of different EGFR mutations, including G719C, G719S, L858R, L861Q, and E746-A750 deletion, will not work, but a different targeted therapy (osimertinib) can be able to kill those cancer cells in a patient's non-small cell lung cancer (Lamb and Scott (2017) Targeted Oncology 12:555-562).

[0094] The amplification and detection method is a real-time PCR method similar to that of Example 1, except that the target mutation is EGFR mutation T790M, and the cognate detection probe is a shared stem molecule beacon that targets the complement of the bridge sequence of the superselective forward primer, which does not contain the 5'-tag sequence. In this case, the A:T in the mutant (see Figure 3 ) differs from the G:C in the wild type. As in Example 1, one set of reactions uses a pair of superselective primers, each of which has an interrogating 3'-terminal nucleotide, and a second set of reactions uses a restrictive superselective forward primer and a conventional reverse primer that is complementary to both the target and non-target target sequences, which in this case has the sequence of the anchor sequence of the superselective reverse primer. Both pairs of primers were tested on three samples. The samples contained 10,000 copies of the wild type closely related non-target sequence plus 0, 10, or 100 copies of the mutant target sequence. 50 mM TMAc was included in each amplification reaction mixture as an effective amount of a selectivity enhancing reagent.

[0095] PCR amplification with real-time fluorescence detection was performed on 5 replicates of each sample. The fluorescence intensity curves of the amplification reactions are shown in Figure 4The average Ct values (of five replicates) for the sample containing only wild type and the sample containing 10 copies of the mutant in the presence of 10,000 wild type are shown in Table 2 for each primer pair, as are the ACt between these average Ct values. Table 2 shows that the method using a superselective forward restriction primer and a conventional reverse primer gives a significant average ACt of 5.35. However, Figure 4 The upper panel A of Figure 2 shows that multiple replicates are required to achieve this result due to variability between replicates. In contrast, the method using two superselective primers performs much more robustly. Figure 4 The lower panel B of Figure 2 shows that fluorescence from the wild type only sample is significantly delayed and three of the five replicates give no Ct after 55 cycles. To calculate and average the Ct, the Ct of these replicates is set to >55. Doing so, the average ACt is >14.00. This is an increase of nearly 9 cycles, demonstrating that the method of Example 2 using a pair of superselective primers is a method according to the invention.

[0096] Figure 5 An embodiment of the invention is shown having a pair of multi-part primers hybridized to a rare target sequence in a double-stranded template containing a positive (+) strand and a negative (-) strand (indicated by short vertical lines). Although only one primer must be a multi-part primer, the embodiment has a pair of superselective primers. As shown in the upper panel, each primer has an anchor sequence, an unhybridized bridge sequence opposite an intervening sequence in the template, and a foot sequence. The target sequence of interest contains two mutant base pairs that are different from closely related sequences. For purposes of illustration, the first mutant base pair is designated as nucleotide A in the (-) template strand of the target of interest and nucleotide T in the (+) template strand of the target of interest. That is, the first mutation to be detected is a single nucleotide polymorphism (EGFR T790M) that occurs in exon 20 of the EGFR gene. The second mutant base pair is designated as G in the (-) template strand of the target of interest and C in the (+) template strand of the target of interest. That is, the second mutation to be detected is a single nucleotide polymorphism (EGFR C797S) that also occurs in exon 20 of the EGFR gene.

[0097] In Figure 5The embodiment of the application shown is directed to identifying the presence of two target mutations in a chromosomal sample, in this case the presence of both EGFR T790M and EGFR C797S. As described in Example 3, the occurrence of both target mutations in the same chromosome in the sample (i.e. in cis) not only distinguishes the sample from one having only the closely related wild type sequence, but also from one containing the same two mutations but each on a different sister chromatid (i.e. in trans). Each primer has a interrogating nucleotide, here a 3 '-terminal interrogating nucleotide, which is complementary to one of the two mutations. One primer, here designated the forward primer, and also the limiting primer as shown in the middle panel, has its interrogating nucleotide complementary to the T790M mutation in the (-) template strand. The other primer, here designated the reverse primer, and the excess primer as shown in the middle panel, has its interrogating nucleotide complementary to the C797S mutation in the (+) template strand. The top panel also shows a cognate fluorescent detection probe, here a molecular beacon probe, having a single stranded loop and a double stranded stem, with one arm of the stem labeled with a fluorophore (o) and the other arm labeled with a quencher (·). In the embodiment, the molecular beacon is a "regular" molecular beacon probe, i.e. one in which only its single stranded loop is complementary to the target of the probe, which in this case is the complement of the region between the primer sequences in the amplified product. The bottom panel shows the detection, which can be real time detection or end point detection, as used in digital PCR methods. This "inter-primer specific" probe is shown hybridized to the (-) amplicon, with the fluorophore of the probe separated from the quencher of the probe by hybridization of the probe.

[0098] Example 3 describes an embodiment of the method according to the application shown in Figure 5 Example 3. The two different target mutations, T790M (which has an A:T base pair in place of a G:C base pair in exon 20 of the EGFR gene) and C797S (which has a G:C base pair in place of a C:G base pair in the same exon), are 20 nucleotides apart from each other if they occur on the same chromosome, i.e. if they occur in cis. The purpose of the assay described in Example 3 is to determine whether the two mutations, if they are both present in a sample, do indeed occur on the same chromosome, or whether they, if they are both present, occur on sister chromatids, i.e. whether they occur in trans. If only one of the two somatic mutations occurs in a sample from a patient with non-small cell lung cancer, or if both occur but in trans on sister chromatids (Vokes and Herbst, 2007), then the patient is not likely to be a candidate for treatment with an EGFR inhibitor, and the patient's sample is not likely to be a candidate for the method of the application. (Oxnard et al. (2017) Journal of Thoracic Oncology 12: 1608-1610), osimertinib would be an effective targeted therapy (Lamb and Scott (2017) Targeted Oncology 12: 555-562). However, if these mutations both occur in cis on the same chromosome, there would be two amino acid substitutions in the resulting EGFR protein, and osimertinib would not be an effective targeted therapy (Wang et al. (2016) Journal of Hematology and Oncology 9: 59). Instead, brigatinib would substitute for it as being effective (Uchibori et al. (2017) Nature Communications 8: 14768).

[0099] To obtain a reference for the evaluation of ACt and to illustrate why the method is needed, a first series of amplifications was performed using a conventional primer as reverse primer, the real-time fluorescence curve of which is as follows Figure 6The reported. Panel E shows the fluorescence curves for four replicates of a sample containing only 10,000 copies of the closely related (wild type) EGFR sequence; upper left panel A shows the curves for four replicates of a sample containing 10,000 copies of the EGFR wild type sequence plus 10 copies of the EGFR T790M sequence; upper right panel B shows the curves for four replicates of a sample containing 10,000 copies of the EGFR wild type sequence plus 10 copies of the EGFR C797S sequence; middle right panel D shows the curves for four replicates of a sample containing 10,000 copies of the EGFR wild type sequence plus 10 copies of the EGFR T790M sequence and 10 copies of the EGFR C797S sequence (i.e., both mutations present in trans); and middle left panel C shows the curves for four replicates of a sample containing 10,000 copies of the EGFR wild type sequence plus 10 copies of a sequence containing both the EGFR T790M mutation and the EGFR C797S mutation (i.e., both mutations present in cis). As shown in Table 3, the average Ct value for the sample containing only 10,000 wild type templates was 45.82, while the average Ct values for the four different types of samples containing, in addition to the 10,000 wild type templates, 10 copies of either or both mutant templates were 38.34, 39.95, 38.44, and 39.49 (with an overall average Ct value of 39.05). The key point here is that, although the average Ct value (an average ΔCt of 6.77) that is lower than the Ct value for the sample containing only wild type templates indicates the presence of one or both mutations in the sample, the average Ct value for the trans configuration (whose replication curves are shown in middle right panel D) is 38.44, which is almost the same as the average Ct value of 39.49 for the cis configuration (whose replication curves are shown in middle left panel C), which gives a very similar average Ct value. Thus, the cis configuration cannot be distinguished from the trans configuration.

[0100] To achieve this distinction, a second series of amplifications was performed using the method shown in Figure 5 Real-time PCR amplification and detection was performed using a pair of superselective primers (as shown in Figure 5 The real-time fluorescence curves for a sample containing both mutations in cis or for a sample containing both mutations in trans are shown in Figure 7The lower panel C shows the fluorescence curves for four replicates of a sample containing only 10,000 copies of the wild-type sequence; the upper left panel A shows the curves for four replicates of a sample containing 10,000 copies of the wild-type sequence plus 10 copies of the two mutations in cis configuration; and the upper right panel B shows the curves for four replicates of a sample containing 10,000 copies of the wild-type sequence plus 10 copies of the two mutations in trans configuration, i.e., in different sequences. The average Ct value for the in cis sample was 41.31. Neither the in trans sample nor the wild-type only sample showed fluorescence above background after 55 amplification cycles. According to our rules, we assign a Ct value of >55 to each, and the ΔCt value is 13.69. This method qualifies as a method according to the invention. It meets the criterion that neither sample without the in cis template (the target target sequence) produced a Ct value above background after 55 amplification cycles. Furthermore, since the ΔCt value is increased by 13.69 cycles relative to the wild-type only sample, it meets the additional criterion of being increased by at least 5 cycles.

[0101] In Example 4, the method of the second series of amplifications in Example 3 was repeated, using ARMS primers instead of superselective reverse primers as the reverse primers. Figure 8 A method for detecting two mutations in cis configuration is shown, and the real-time fluorescence curves are reported in Figure 9 where the lower panel C shows the fluorescence curves for four replicates of a sample containing only 10,000 copies of the wild-type sequence; the upper left panel A shows the curves for four replicates of a sample containing 10,000 copies of the wild-type sequence plus 10 copies of the two mutations in cis configuration; and the upper right panel B shows the curves for four replicates of a sample containing 10,000 copies of the wild-type sequence plus 10 copies of the two mutations in trans configuration, i.e., in different sequences. The average Ct value for the in cis sample was 41.31. Neither the in trans sample nor the wild-type only sample showed fluorescence above background after 55 amplification cycles. According to our rules, we assign a Ct value of >55 to each, and the ΔCt value is 13.69. This method qualifies as a method according to the invention. It meets the criterion that neither sample without the in cis template (the target target sequence) produced a Ct value above background after 55 amplification cycles. Furthermore, since the ΔCt value is increased by 13.69 cycles relative to the wild-type only sample, it meets the additional criterion of being increased by at least 5 cycles.

[0102] Example 5 describes the use of ARMS primers as second primers in a method to detect a single base pair change. Assays were performed to demonstrate the use of ARMS primers as limiting primers or excess primers. Real-time PCR assays using real-time detection were performed to detect 10 copies of a rare KRAS G12D mutant target sequence in a mixture containing 10,000 copies of its closely related non-target wild-type target sequence as described in Example 5, using the following different primer pairs: limiting ARMS forward primer and excess superselective reverse primer, limiting superselective forward primer and excess ARMS reverse primer, and a pair of superselective primers (as a control). Each multipartite primer and each ARMS primer had a 3 '-end interrogating nucleotide complementary to the nucleotide of the mutant base pair.

[0103] Methods in which the ARMS primer is the forward primer are shown in Figure 10 The results of these assays are shown in Figure 11 For all primer pairs containing a superselective primer as the multipartite first primer, the fluorescence intensity of samples containing only the non-target (wild-type) template sequence did not rise above background over 55 amplification cycles. Thus, the methods using each of these primer pairs are methods according to the invention.

[0104] Example 6 describes the selectivity and sensitivity of real-time PCR assays designed to detect the presence of a rare mutant target DNA fragment and determine its relative abundance in a sample containing a large number of DNA fragments from an entire normal human genome. Specifically, this example demonstrates that a very small number of target fragments can be reliably detected using a pair of allele-discriminating multipartite primers each complementary to a single base pair mutation present in a rare DNA fragment being analyzed by a real-time PCR assay in a sample containing a large number of DNA fragments from an entire normal human genome. Specifically, each of 10 samples each purportedly containing 5 mutant DNA fragments in the presence of 10,000 copies of DNA fragments from an entire normal human genome each gave a positive signal for the presence of the mutant DNA fragment. As a control, all 10 samples containing no mutant DNA fragments but containing DNA fragments from 10,000 copies of an entire normal human genome each gave no positive signal for the presence of the mutant DNA fragment.

[0105] These results mean that a positive result in an assay using an allele discriminating primer such as a superselective primer, both specific for the same mutant base pair, can be indicative of a true positive result; and that a negative result in these same assays can be indicative of a true negative result. This is a critical criterion for an extremely sensitive PCR assay, for example one designed to detect the presence of rare mutant fragments in cell-free DNA isolated from the plasma fraction of a 10 mL blood sample obtained from a cancer patient, where the presence of a particular mutation indicates that a particular targeted therapy will be effective (Sabari et al. (2019) Journal of the National Cancer Institute 111 :575-583).

[0106] All assays performed in this example contained DNA fragments from 10,000 copies of an entire normal human genome. This is greater than the amount of cell-free DNA fragments isolated from 1 mL of plasma from a 10 mL blood sample typically obtained from a patient (Meddeb et al. (2019) Scientific Reports 9:5220). However, since the actual number of cell-free DNA fragments in a patient blood sample can vary from hour to hour, it is important to include primers and probes in the real-time PCR assay that detect DNA fragments from a normal reference gene, thereby enabling the amount of DNA in the sample to be determined. The results would then indicate whether there is sufficient DNA in the sample to enable detection of rare mutant DNA fragments. Furthermore, comparison of the threshold cycle (Ct) obtained for the mutant target fragment to the threshold cycle obtained for the reference gene in the PCR assay (ACt) enables the results to be expressed as the relative abundance of the mutation in the patient DNA, a clinically relevant result.

[0107] Figure 12 How a pair of superselective primers directed to the G719C mutation in the EGFR gene were used in the assay of Example 6 is described. The limiting primer directed to the G719C gene contained a 5'-tag sequence, and the FAM-labeled conventional molecular beacon present in the assay emitted a signal for the presence of an amplicon due to the presence of mutant DNA fragments in the sample by binding to the complement of the 5'-tag sequence incorporated into the 3' end of the excess (-) amplicon. All assays performed in Example 6 also contained a limiting concentration of a superselective forward primer directed to the human beta-actin gene, and an excess concentration of a conventional reverse primer directed to the beta-actin gene, and a Quasar 705-labeled inter- primer specific molecular beacon that emitted a signal for the presence of an amplicon produced from the beta-actin reference gene.

[0108] Four sets of ten assays were performed, each containing 10,000 copies of DNA restriction fragments from normal human genomic DNA. In addition, each assay contained, in the first set, 500 copies of a linearized plasmid containing the mutant target sequence, in the second set, 50 copies of the mutant DNA plasmid, in the third set, 5 copies of the mutant DNA plasmid, and in the fourth set, as a negative control, no mutant DNA plasmid.

[0109] The results of these 40 PCR assays are shown in Figure 13 Figure 13 Panel A in the upper left shows the results of each assay containing 500 mutant plasmids; Panel B in the upper right shows the results of assays containing 50 mutant plasmids; Panel C in the lower left shows the results of assays containing only 5 mutant plasmids; and Panel D in the lower right shows the results of assays containing no mutant plasmids.

[0110] All of these assays gave positive FAM signals, and their average Ct value was 43.27. In contrast, all of the reactions containing no mutant plasmid gave no FAM signal above background throughout the 55 amplification cycles. These results demonstrate the exceptional selectivity and sensitivity of the primer pair using allele-specific primers for the exponential amplification assay for detecting single nucleotide polymorphisms, which indicates that assays using these primer pairs will enable extremely sensitive clinical assays to be performed relatively rapidly, at low cost, and in widely available instruments.

[0111] Compositions and kits

[0112] The present application encompasses compositions or reaction mixtures comprising the above-described primers and reagents for performing the above-described methods. For example, the compositions can comprise one or more reagents selected from the group consisting of nucleic acid polymerases, deoxyribonucleotide triphosphates, and detection agents.

[0113] ​The detection agent can be an oligonucleotide probe, such as a molecular beacon probe or a Yin-Yang probe labeled with a fluorophore and a quencher. See, e.g., U.S. Patent Nos. 5925517, 6103476, 6150097, 6270967, 6326145, and 7799522. In addition to the above agents, the composition can also include one or more of the following: salts, such as NaCl, MgCl2, KC1, MgS04; buffers, such as Tris buffer, N-(2-hydroxyethyl)-piperazine-N'-(2-ethanesulfonic acid) (HEPES), 2-(N-morpholino)ethanesulfonic acid (MES), MES sodium salt, 3-(N-morpholino)propanesulfonic acid (MOPS), N-tris-[hydroxymethyl]-methyl-3-aminopropanesulfonic acid (TAPS); solubilizers; detergents, such as non-ionic detergents, such as Tween-20; nuclease inhibitors; and the like.

[0114] Reaction components for amplification and / or detection processes can be provided in a variety of forms. For example, components (e.g., enzymes, deoxyribonucleotide triphosphates, adaptors, blockers, and / or primers) can be suspended in an aqueous solution or as a freeze-dried or lyophilized powder, pellet, or bead. In the latter case, the components form a complete mixture of components for use in an assay upon reconstitution. Examples

[0115] Example 1 : Use of a pair of superselective primers in a real-time PCR assay for detection of rare EGFR G719C mutant template in the presence of a large amount of wild-type template

[0116] The design of this first example is shown in Figure 1 The target mutation (G719C) located in exon 18 of the human epidermal growth factor (EGFR) gene is an A:T base pair in place of a C:G base pair, the latter occurring in an otherwise identical wild-type gene sequence. A PCR assay was performed using mutant and wild-type plasmids containing the target gene sequence, with one superselective primer (which we will refer to as the "forward" primer) having an "anchoring" sequence that binds to all (-) template strands in the sample (rare target of interest and large amounts of non-target of interest).

[0117] In this first example, the limiting forward primer contains a unique "5'-tag sequence". When the forward allele-discriminating primer binds to the mutant (-) template and begins synthesis, the resulting (+) amplicon strand contains the entire forward primer sequence, including the 5'-tag sequence at its 5' end. Subsequently, these (+) amplicons are used as templates for the reverse allele-discriminating primer, or in a control experiment, as the reverse regular (non-discriminating) primer. The resulting (-) amplicons will contain the complement of the 5'-tag sequence at their 3' end. The target for the molecular beacon probe present in these reactions to light up the synthesized amplicon is the 3' complement of the 5' tag sequence. Furthermore, since the forward primer is present in limiting amounts, the single-stranded amplicon is made by extension of the excess reverse primer (either the allele-discriminating primer or the regular primer), which ensures that the molecular beacon probe is able to bind to its target without competition from collapsing amplicon duplexes. The oligonucleotide sequences used in this example are:

[0118] EGFR G719C super-selective forward primer 32 -28-20 / 13-8: 1 :0

[0119] 5'-TCTCTTGAGGATCTTGAAGGAAACTGAA-CCTCTCCAACGAATCTCGAA-AAGTGCTGT-3' (SEQ ID No. 1), where the nucleotides of the 5'-tag are underlined, and the 3'-end interrogation nucleotides are bolded ACCTGCCGTCAACACGTGCGCAGTAGACCATC

[0120] EGFR G719C super-selective reverse primer 24-18 / 10-8: 1 :0

[0121] 5'-CCAGGGACCTTACCTTATACACCG-GATCCTAACTGAGGTCCA-ACCGGAGCA-3' (SEQ ID No. 2), where the 3'-end interrogation nucleotides are bolded

[0122] EGFR exon 18 regular reverse primer

[0123] 5'-CCAGGGACCTTACCTTATACACCG-3' (SEQ ID No. 3)

[0124] Regular molecular beacon (binds to the complement of the 5'-tag sequence)

[0125] 5'-Quasar 670-CCGCCTG- ACCTGCCGTCAACACGTGCGCAGTAGACCATC -CAGGCGG-BHQ2-3' (SEQ ID No. 4), where the nucleotides in the single-stranded loop are underlined​

[0126] Target plasmids containing either the EGFR G719C mutation or the corresponding EGFR wild-type sequence were purchased from Integrated DNA Technologies, Coralville, Iowa (USA) and prepared by inserting a 211 base pair gene fragment into a pIDT Smart Amp vector. Mutant and wild-type plasmid DNA was digested with the restriction enzyme Sca I (New England Biolabs, Ipswich, Massachusetts (USA)). Digestion mixtures contained 10 units of Sca I and 4 μg of either mutant or wild-type plasmid DNA in a 20 μL volume containing 100 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, and 50 mM Tris-HCl (pH 7.9). Reactions were incubated at 37 °C for 120 minutes, then inactivated at 80 °C for 20 minutes to deactivate the endonuclease.

[0127] PCR assays were performed in 30 μL volumes containing 10,000 copies of wild-type template, or 10 copies of mutant template in a mixture containing 10,000 copies of wild-type template, and amplification buffer (50 mM KCl, 10 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2), 50 mM tetramethylammonium chloride (Sigma-Aldrich, St. Louis, Missouri (USA)), 0.5% Tween 20 (Sigma-Aldrich), 1.5 units of Platinum Taq DNA polymerase (Thermo Fisher Scientific, Waltham, Massachusetts (USA)), 250 μM ATP, 250 μM CTP, 250 μM GTP, 250 μM TTP, and 300 nM conventional molecular beacons. One set of reactions contained 60 nM of EGFR G719C superselective forward primer and 300 nM of EGFR exon 18 conventional reverse primer. Another set of reactions contained 60 nM of EGFR G719C superselective forward primer and 300 nM of EGFR G719C superselective reverse primer. Both sets of reactions included amplification in duplicate of 10,000 wild-type copies and amplification in duplicate of 10 mutant copies in a mixture containing 10,000 wild-type copies. Amplification was performed in duplicate using 0.2 ml white polypropylene tubes (USA Scientific, Ocala, Florida (USA)) in a Bio-Rad CFX-96 Touch Real-Time PCR Detection System (Hercules, California (USA)). The thermal cycling program was 95 °C for 2 minutes, followed by 55 cycles of 95 °C for 20 seconds, 60 °C for 20 seconds, and 72 °C for 20 seconds. Molecular beacon fluorescence intensity was measured in real time at the end of the 60 °C annealing phase of each thermal cycle. Threshold cycle (Ct value) was automatically calculated by the thermal cycler.

[0128] Figure 2 Fluorescence intensity readings are shown relative to the number of thermal cycles completed for these real-time PCR amplification and detection assays. Panel A above contains curves for reactions using a superselective (SSP) forward primer and a conventional reverse primer, and panel B below contains curves for reactions using a superselective (SSP) forward primer and a superselective (SSP) reverse primer. The average threshold cycle (Ct value) for amplification in duplicate, and the difference between the average Ct values for these assays (ΔCt value) are listed in Table 1 by comparing assays containing 10 mutant copies in the presence of 10,000 wild-type, with assays containing only 10,000 wild-type.

[0129] Table 1

[0130]

[0131] Example 2: Use of a pair of ultra-selective primers in a real-time PCR assay for detection of rare EGFR T790M mutant templates in the presence of large amounts of wild-type templates

[0132] This example of the method of the invention uses the design shown in Figure 3 which is different from the design used in Figure 1 Example 1. In this example, we used a variant of the molecular beacon probe sometimes referred to as a "shared-stem" molecular beacon probe (Tsourkas et al. (2002) Nucleic Acids Research 30:4208-4215) to target the complement of the bridge sequence of the restrictive ultra-selective primer. Such a molecular beacon probe differs from the conventional molecular beacon probe (Example 1) in that the former has a target sequence of the probe that is complementary to one arm of the stem as well as to the single-stranded loop. The target sequence mutation in exon 20 of the EGFR gene (T790M) is a single base pair substitution (SNP) of an A:T base pair for a G:C base pair, the latter occurring in a closely related wild-type sequence that is otherwise identical. The oligonucleotides used in this example are:

[0133] EGFR T790M ultra-selective forward primer 24-22 / 11-8: 1 :0

[0134] 5'-CGCCTGCTGGGCATCTGCCTCACC-

[0135] CAACACGTGCGCAGTAGACCA C-GCTCATCAT-3' (SEQ ID No. 5), where the 3'-terminal interrogating nucleotides are bolded and their complement is underlined, which is part of the bridge sequence detected by the shared-stem molecular beacon

[0136] EGFR T790M ultra-selective reverse primer 23-10 / 19-7: 1 :0

[0137] 5'-TTTGTGTTCCCGGACATAGTCCA-ATCTTCGGTG-GAGCTGCA-3' (SEQ ID No. 6), where the 3'-terminal interrogating nucleotides are bolded

[0138] EGFR exon 20 conventional reverse primer

[0139] 5'-TTTGTGTTCCCGGACATAGTCCA-3' (SEQ ID No. 7)

[0140] Shared stem molecular beacon (complement of bridge sequence of forward primer)

[0141] 5'-Quasar 670-TGGTCT- CAACACGTGCGCAGT-AGACCA -BHQ2-3' (SEQ ID No. 8), where nucleotides in the loop and arms complementary to the target are underlined

[0142] Target plasmids containing either the EGFR T790M mutation or the corresponding EGFR wild-type sequence were purchased from Integrated DNA Technologies (IDT) and prepared by inserting a 200 base pair gene fragment into a pIDT Smart Amp vector. Mutant and wild-type plasmid DNA was digested with the restriction enzyme Sca I. Digestion mixtures contained 10 units of Sca I and 4 μg of either mutant or wild-type plasmid DNA in a 20 μL volume containing 100 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, and 50 mM Tris-HCl (pH 7.9). Reactions were incubated at 37°C for 120 minutes, then inactivated at 80°C for 20 minutes to deactivate the endonuclease.

[0143] PCR assays were performed in 30 μL volumes containing 50 mM KCl, 10 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2, 50 mM tetramethylammonium chloride, 0.5% Tween 20, 1.5 units of Platinum Taq DNA polymerase, 250 μM ATP, 250 μM CTP, 250 μM GTP, 250 μM TTP, and 300 nM shared stem molecular beacon. One set of reactions contained 60 nM of forward superselective primer and 300 nM of conventional reverse primer. Another set of reactions contained 60 nM of superselective forward primer and 300 nM of superselective reverse primer. Both sets of reactions included five replicates of 10,000 wild-type copies and five replicates of 10 mutant copies in a mixture containing 10,000 wild-type copies. Amplification was performed using 0.2 ml white polypropylene tubes in a Bio-Rad CFX-96 Touch fluorescent spectrothermal cycler. The thermal cycling program was 95°C for 2 minutes, followed by 55 cycles of 95°C for 20 seconds, 60°C for 20 seconds, and 72°C for 20 seconds. Molecular beacon fluorescence intensity was measured at the end of each 60°C annealing phase. Threshold cycles were automatically calculated by the thermal cycler.

[0144] Figure 4Fluorescence intensity readings are shown relative to the thermal cycles completed for the PCR amplification using real-time fluorescence detection. Upper panel A contains the curves for replicate reactions using a restrictive superselective (SSP) forward primer and an excess of a conventional reverse primer. Lower panel B contains the curves for replicate reactions using a restrictive superselective (SSP) forward primer and an excess of a superselective (SSP) reverse primer. The average threshold cycle (Ct value) for the replicate amplifications, as well as the difference between the average Ct values for these assays (ΔCt value) are listed in Table 2 by comparing assays containing 10 mutant types in the presence of 10,000 wild type with assays containing only 10,000 wild type. The fluorescence intensity for three of the five amplifications of the wild type template and superselective primer pair did not rise above background at 55 cycles, so we assigned a Ct value of >55 for each.

[0145] Table 2

[0146]

[0147] Example 3: Use of a pair of superselective primers in a real-time PCR assay to determine whether two different somatic mutations in the same gene occur in cis on the same chromosome or whether they occur in trans on sister chromosomes

[0148] This example of the method of the invention uses the method shown in Figure 5 Two different target mutations, T790M (which has an A:T base pair replacing a G:C base pair in exon 20 of the EGFR gene) and C797S (which has a G:C base pair replacing a C:G base pair in the same exon), are 20 nucleotides apart from each other if they occur on the same chromosome, i.e., if they occur in cis. The purpose of the assay shown in this example is to determine whether these two mutations, if they are both present in a sample, do indeed occur on the same chromosome, or, if they are both present, whether they occur on sister chromosomes, i.e., whether they occur in trans.

[0149] We first performed a series of preliminary assays that illustrate the type of results that would be obtained from a multiplex assay directed to these individual mutations (or from single assays each of which searches for one or the other target mutation), the results of which would require a cis or trans assay to determine an effective targeted therapy. In these assays, there are three primers: a superselective forward primer directed to EGFR T790M, a superselective forward primer directed to EGFR C797S, and a conventional reverse primer that participates in the synthesis of the amplicon whether only one of the two mutations is present in the sample or both of the two mutations are present in the sample. The oligonucleotide sequences used in these experiments are:

[0150] EGFR T790M superselective forward primer 24-22 / 13-8: 1:0

[0151] 5'-GCCGCCTGCTGGGCATCTGCCTCA-AAGAATCAACAAGCTACAACTC- GCTCATCAT-3' (SEQ ID No. 9), with 3 '-end interrogation nucleotides bolded

[0152] EGFR C797S superselective forward primer 21-13 / 20-9: 1:0

[0153] 5'-CTGCCTCACCTCCACCGTGCA-AGCACTCGCAGAA-CCTTCGGCTC-3' (SEQ ID No. 10), with 3 '-end interrogation nucleotides bolded

[0154] EGFR exon 20 regular reverse primer #2

[0155] 5'-CACCAGTTGAGCAGGTACTGG-3' (SEQ ID No. 11)

[0156] Amplifinder-specific molecular beacon

[0157] 5'-FAM-CCGTGG- CTGGACTATGTCCGGGAACACA -CCACGG-BHQl-3' (SEQ ID No. 12), with nucleotides of single-stranded loop underlined

[0158] In a first set of experiments, whose design is not shown in Figure 5 In the first set of experiments, whose design is not shown in

[0159] As in the previous example, three target plasmids were purchased from Integrated DNA Technologies and prepared by inserting 200 base pair gene fragments into pIDTSmart Amp vectors. Each of these plasmid DNAs was digested with the restriction enzyme Sca I. Digestion mixtures contained 10 units of Sca I and 4 μg of mutant or wild-type plasmid DNA in a 20 μL volume containing 100 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, and 50 mM Tris-HCl (pH 7.9). Reactions were incubated at 37°C for 120 minutes, then inactivated at 80°C for 20 minutes to deactivate the endonuclease.

[0160] We prepared five different amplification reaction mixtures in a volume of 30 μΐ^. Each contained 10,000 copies of wild-type DNA template. One set of reactions contained only wild-type template. The other four sets of reactions additionally contained 10 copies of one of the following target plasmids or combinations of plasmids: T790M plasmid; C797S plasmid; T790M plasmid plus C797S plasmid (mimicking the case where the two mutations occur in trans); or T790M-C797S plasmid, which has both mutations present on the same template (in cis). All reaction mixtures contained 60 nM of the superselective T790M forward primer, 60 nM of the C797S superselective forward primer, 300 nM of the conventional reverse primer, 300 nM of the "amplicon-specific" molecular beacon, 50 mM KCl, 10 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2, 50 mM tetramethylammonium chloride (TMAC), 0.5% Tween 20, 1.5 units of Platinum Taq DNA polymerase, 250 μΜ ATP, 250 μΜ CTP, 250 μΜ GTP, and 250 μΜ TTP.

[0161] Four replicate amplifications of each of the five different reaction mixtures were performed using 0.2 ml white polypropylene tubes in a Bio-Rad CFX-96 Touch fluorescent spectrothermal cycler. The thermal cycling program was 95 °C for 2 minutes, followed by 55 cycles of 95 °C for 20 seconds, 60 °C for 20 seconds, and 72 °C for 20 seconds. The molecular beacon fluorescence intensity was measured at the end of each 60 °C annealing phase.

[0162] Figure 6 Fluorescence intensity readings are shown relative to the thermal cycling of PCR amplification using real-time fluorescence detection. As described above, all amplification reaction mixtures contained 10,000 copies of wild-type (non-target template). The upper left panel A shows the results of replicate reactions in which the amplification reaction mixture also contained 10 copies of the T790M mutant target template; the upper right panel B shows the results of replicate reactions in which the amplification reaction mixture also contained 10 copies of the C797S mutant target template; the left middle panel C shows the results of replicate reactions in which the amplification reaction mixture also contained 10 copies of the T790M mutant target template and 10 copies of the T790M-C797S mutant target template; the right middle panel D shows the results of replicate reactions in which the amplification reaction mixture also contained 10 copies of the T790M mutant target template and 10 copies of the C797S mutant target template; and the lower panel E shows the results of replicate reactions in which the amplification reaction mixture contained no copies of mutant target template. The average threshold cycle (Ct value) of these replicate amplifications is listed in Table 3.

[0163] Table 3

[0164]

[0165] Table 3 shows the results of control reactions containing the EGFR T790M superselective forward primer and the EGFR C797S superselective forward primer along with the EGFR exon 20 conventional reverse primer. The average Ct value for the sample containing only the wild type (45.82) can be distinguished from the sample containing the two mutations in cis (39.49). However, this average Ct value is nearly identical to the average Ct value for the reaction containing the two mutations in trans and nearly identical to the average Ct value for the reaction containing only one of the two different mutations.

[0166] To illustrate how one can determine whether the two mutations occurred on the same template (i.e., in cis), according to one embodiment of the application, we performed an additional three sets of assays, designed as shown in Figure 5 These assays contain the limiting EGFR T790M superselective forward primer, excess EGFR C797S reverse primer, and an inter-priming specific molecular beacon probe targeting the region between the (-) amplicon bound by the superselective primers, so that a signal is only emitted when both primers have bound and have been extended on a template molecule containing both the EGFR T790M mutation and the EGFR C797S mutation (i.e., only when the two mutations are present in cis). The oligonucleotides used are:

[0167] EGFR T790M superselective forward primer 24-22 / 13-8: 1 :0

[0168] 5'-GCCGCCTGCTGGGCATCTGCCTCA-AAGAATCAACAAGCTACAACTC- GCTCATCAT-3' (SEQ ID No. 9), with the 3 '-terminal interrogation nucleotide bolded

[0169] EGFR C797S superselective reverse primer 30-20 / 19-8: 1 :0

[0170] 5'-TTGAGCAGGTACTGGGAGCCAATATTGTCT-GTCCTTTACAAGCACGAGTG- CCAGGAGGG-3' (SEQ ID No. 13), with the 3 '-terminal interrogation nucleotide bolded

[0171] Inter-priming specific molecular beacon

[0172] 5'-FAM-CCGTCG- CAGCTCATGCCCTTCGGC- CGACGG-BHQl-3' (SEQ ID No. 14), where the nucleotides of the single-stranded loop are underlined

[0173] The amplification reaction mixtures contained 60 nM of the superselective forward primer, 300 nM of the superselective reverse primer, and 300 nM of the molecular beacon. All reaction mixtures contained 10,000 copies of the EGFR wild-type target template. One set of reactions additionally contained 10 copies of the T790M-C797S target template, a second set of reactions contained 10 copies of the T790M target template and 10 copies of the C797S target template, and a third set of reactions contained neither copy of either mutant target template, i.e., only the wild-type template was present. The other aspects of the amplification reaction mixtures were as described above.

[0174] Four replicate amplifications of each of the three reaction mixtures were performed using real-time detection as described above. Real-time fluorescence readings obtained during the first 55 amplification cycles are shown in the three graphs in Figure 7 Panel C, lower right, shows the results for a reaction mixture containing only 10,000 wild-type templates, where the fluorescence for all four replicates did not rise above background by cycle 55; thus, the average Ct for the four replicates, while not determinable, was at least greater than 55. Panel B, upper right, shows the results for a reaction mixture containing 10,000 wild-type templates plus 10 copies of each of the T790M template and the C797S template (i.e., both mutations present in trans); and the fluorescence for all four replicates did not rise above background by cycle 55; thus, the average Ct for the four replicates, while not determinable, was at least greater than 55. Panel A, upper left, shows the results for a reaction mixture containing 10,000 wild-type templates plus 10 copies of the T790M-C797S template (i.e., both mutations present in cis); and all four of these reactions gave positive signals with an average Ct of 41.31. The ΔCt between the ten cis templates, while not precisely determinable, was at least greater than 13.69 when compared to a reaction containing 10 copies of each of the two different mutant templates present in trans in a reaction also containing 10,000 wild-type templates.

[0175] Example 4: Use of superselective primers as restriction primers and ARMS reverse primers in a real-time PCR assay to determine whether two different somatic mutations in the same gene occur in cis on the same chromosome or whether they occur in trans on two different sister chromatids

[0176] We repeated the method described in Example 3, using ARMS primers as the reverse primers instead of the reverse superselective primers. This alternative arrangement is shown in Figure 8The oligonucleotide in the amplification reaction mixture is:

[0177] EGFR T790M ultra-selective forward primer 24-22 / 13-8: 1:0

[0178] 5'-GCCGCCTGCTGGGCATCTGCCTCA-AAGAATCAACAAGCTACAACTC- GCTCATCAT-3' (SEQ ID No. 9), where the 3 '-end interrogating nucleotides are bolded

[0179] EGFR C797S ARMS reverse primer 19: ml: 1: 1:0

[0180] 5'-TCCCGGACATAGTCCAGGAAGG-3' (SEQ ID No. 15), where the 3 '-end interrogating nucleotides are bolded, and the intentionally introduced mismatched nucleotide is bolded and italicized

[0181] Inter-primer specificity molecular beacon

[0182] 5'-FAM-CCGTCG- CAGCTCATGCCCTTCGGC -CGACGG-BHQl-3' (SEQ ID No. 14), where the nucleotides of the single-stranded loop are underlined

[0183] In the sequence of the ARMS primer, the intentionally introduced nucleotide that is mismatched to both the target (mutant) target sequence and the closely related non-target (wild-type) target sequence is bolded and italicized. This nucleotide is the third nucleotide from the 3 '-end, and it creates an A:C mismatch with respect to both the target sequence and the closely related non-target sequence.

[0184] The reaction mixture was identical to that described in Example 3, as were the thermal cycling conditions and the manner of fluorescence detection, except that the ultra-selective reverse primer was replaced with the ARMS reverse primer. The design of this experiment is shown in Figure 8 Figure 7 Very similar results were obtained for the same series of amplifications as shown in

[0185] The results of these experiments, which included ARMS primers, are shown in Figure 9 ​The fluorescence signal from all four replicates failed to rise above background by cycle 55, so the average Ct value for the four replicates, while not determinable, was at least greater than 55. For the reaction containing 10,000 copies of the wild-type template plus 10 copies of the T790M-C797S template in cis, the average Ct value for the four replicates was 43.31. The ΔCt value between the reaction containing 10 copies of the template in cis and 10,000 copies of the wild-type template, while not precisely determinable, was at least greater than 11.69.

[0186] Example 5: Comparison of using ultra-selective primers as limiting primers or excess primers in a real-time PCR assay comprising ARMS primers for detecting rare KRAS G12D mutant templates in the presence of large amounts of wild-type template

[0187] The experiments in this example used ultra-selective primers as limiting first primers (described here as forward primers) and ARMS primers as excess second primers (described here as reverse primers); and ultra-selective primers as excess first primers (described here as reverse primers) and ARMS primers as limiting second primers (described here as forward primers). For comparison, a pair of ultra-selective primers was also used. In all primer pairs, the forward limiting primer had a 5'-tag sequence whose complement was the target of the molecular beacon probe. The method in which the ultra-selective primer was the reverse primer and the ARMS primer was the forward primer is shown in FIG. 1, where the ARMS forward primer comprises a 5'-tag sequence. The target of interest in this example was a KRAS G12D mutant template, which differs from its closely related wild-type template by a single T:A mutant base pair. Both primers in each primer pair had a 3'-terminal interrogating nucleotide that was complementary to one of the nucleotides of this base pair. The sequences of the oligonucleotides used in this example were: Figure 10

[0188] KRAS G12D ultra-selective forward primer 32 -28-19 / 10-8: 1 : 0

[0189] 5'- ACCTGCCGTCAACACGTGCGCAGTAGACCATC ​-5'-tag sequence in underlined, and 3 '-end interrogation nucleotide in bold

[0190] 5'-tag sequence in underlined, and 3 '-end interrogation nucleotide in bold

[0191] KRAS G12D superselective reverse primer 30-14 / 11-7: 1 :0

[0192] 5'-tag sequence in underlined, and 3 '-end interrogation nucleotide in bold

[0193] Regular molecular beacon (binds to complement of 5'-tag sequence) for use with superselective forward primer

[0194] 5'-Quasar 670-CCGCCTG- ACCTGCCGTCAACACGTGCGCAGTAGACCATC -CAGGCGG-BHQ2-3' (SEQ ID No. 4), where nucleotides in single-stranded loop are underlined

[0195] KRAS G12D ARMS forward primer 21 -21-m1: 1 : 1 :0

[0196] 5'-tag sequence in underlined, and 3 '-end interrogation nucleotide in bold CAACACTGGCGCAGTAGACCA -TAAACTTGTGGTAGTTGGAGCGGA-3' (SEQ ID No. 18), where nucleotides in 5'-tag sequence are underlined, 3 '-end interrogation nucleotide is in bold, and introduced mismatched nucleotide is shown in italics and bold

[0197] KRAS G12D ARMS reverse primer 18-m1: 1 : 1 :0

[0198] 5'-tag sequence in underlined, and 3 '-end interrogation nucleotide in bold, and introduced mismatched nucleotide is shown in italics and bold

[0199] Shared stem molecular beacon (binds to complement of 5'-tag sequence) for use with ARMS forward primer

[0200] 5'-FAM-TGGTCT- CAACACTGGCGCAGT-AGACCA -BHQ1-3' (SEQ ID No. 20), where nucleotides in single-stranded loop are underlined

[0201] The third nucleotide from the 3' end of each ARMS primer is bolded and italicized because it mismatches both the mutant sequence and the wild-type sequence. This mismatch is denoted with an "m" in the sequence designation.

[0202] Target plasmids containing either the KRAS G12D mutation or the corresponding KRAS wild-type sequence were purchased from Integrated DNA Technologies and prepared by inserting a 390 base pair gene fragment into a pUCIDT vector. Mutant and wild-type plasmid DNA was digested with the restriction enzyme Dra I (New England Biolabs). Digestion mixtures contained 10 units of Dra I and 4 μg of either mutant or wild-type plasmid DNA in a 20 μL volume containing 50 mM potassium acetate, 20 mM Tris-acetate (pH 7.9), 10 mM magnesium acetate, and 100 μg / mL bovine serum albumin. Reactions were incubated at 37°C for 120 minutes, then inactivated at 65°C for 20 minutes to deactivate the endonuclease.

[0203] PCR assays were performed in 30 μL volumes containing 50 mM KCl, 10 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2, 50 mM tetramethylammonium chloride (Sigma-Aldrich), 0.5% Tween 20 (Sigma-Aldrich), 1.5 units of Platinum Taq DNA polymerase (Thermo Fisher Scientific), 250 μM ATP, 250 μM CTP, 250 μM GTP, 250 μM TTP, 60 nM forward primer, 300 nM reverse primer, and either 300 nM regular molecular beacon used with the superselective forward primer or 300 nM shared stem molecular beacon used with the ARMS forward primer. Amplification was performed in a Bio-Rad CFX-96 Touch fluorescent spectrothermal cycler using 0.2 ml white polypropylene tubes (USA Scientific). The thermal cycling program was 95°C for 2 minutes, followed by 55 cycles of 95°C for 20 seconds, 60°C for 20 seconds, and 72°C for 20 seconds. Molecular beacon fluorescence intensity was measured at the end of each 60°C annealing phase.

[0204] Reaction mixtures contained 10,000 copies of the wild-type target template and either 10 or 0 copies of the mutant target template. Amplification reactions were run in triplicate. The resulting real-time fluorescence results are shown in Figure 11The results obtained with the superselective primer pair are shown in panel A, the results obtained with the primer pair comprising a superselective forward restriction primer and an excess ARMS reverse primer are shown in panel B, and the results obtained with the primer pair comprising a restriction ARMS forward primer and an excess superselective reverse primer are shown in panel C. In none of the panels did the fluorescence of the wild type sample rise above background after 55 cycles, while the Ct values of all samples with 10 copies of the mutant template were below 45.

[0205] Example 6: Duplex assay: Use of a pair of superselective primers in a real-time PCR assay for detection of rare EGFR G719C mutant templates in the presence of large amounts of normal human genomic DNA template, and simultaneous detection of the beta-actin reference gene using a different superselective primer and a conventional primer

[0206] The design of these PCR assays for detection of G719C is shown in Figure 12 The G719C target mutation, located in exon 21 of the human epidermal growth factor receptor (EGFR) gene, is a T:A base pair in place of a G:C base pair, the latter occurring in an otherwise identical wild type gene sequence. The assay uses a linearized plasmid containing the mutant target sequence, and the assay uses fragmented normal human genomic DNA containing the wild type EGFR gene sequence to mimic cell-free DNA fragments isolated from plasma in a liquid biopsy sample.

[0207] In this example, the restriction forward primer for detection of G719C contains a unique "5'-tag sequence". When the forward allele discriminating primer binds to the mutant (-) template and begins synthesis, the resulting (+) amplicon strand contains the entire forward primer sequence, including the 5'-tag sequence at its 5' end. Subsequently, these (+) amplicons serve as templates for the reverse allele discriminating primer. The resulting (-) amplicon contains the complement of the 5'-tag sequence at its 3' end. It is the 3' complement of the 5'-tag sequence, which is the target for the FAM-labeled conventional molecular beacon probe. The forward superselective primer is present at a limiting concentration, and the reverse superselective primer is present at an excess concentration, which ensures that the molecular beacon probe will be able to bind to the excess (-) amplicon target without significant competition from the limiting concentration of (+) amplicons.

[0208] In addition, the assay also includes detection of the beta-actin reference gene sequence occurring in normal human DNA (containing the wild-type EGFR gene) to provide a reference threshold (Ct) reflecting the amount of DNA in the sample. The amplicon is detected using a Quasar 705 labeled inter-primer specific molecular beacon that binds to excess beta-actin (-) amplicon between the complement of the hyperselective primer sequence and the conventional primer sequence.

[0209] The oligonucleotide sequences used in this example are:

[0210] EGFR G719C hyperselective forward primer 32 -28-20 / 13-8:1:0

[0211] 5'- ACGTGCCCTCAATACGAGCCCCCTTCACCAAC -TCTCTTGAGGATCTTGAAGGAAACTGAA-CCTCTCCAACGAATCTCGAA-AAGTGCTGT-3' (SEQ ID No. 21), where

[0212] 5'-tagged nucleotides are underlined and 3'-end interrogating nucleotides are bold

[0213] EGFR G719C hyperselective reverse primer 24-18 / 10-7:1:0

[0214] 5'-CCAGGGACCTTACCTTATACACCG-GATCCTAACTGAGGTCCA-ACCGGAGCA-3' (SEQ ID No. 2), where 3'-end interrogating nucleotides are bold

[0215] Conventional molecular beacon (binds to complement of 5'-tagged sequence)

[0216] 5'-FAM-CGCCTG- ACGTGCCCTCAATACGAGCCCCCTTCACCAAC -CAGGCG-BHQ1-3', (SEQ ID No. 22), where nucleotides in the single-stranded loop are underlined beta-actin hyperselective forward primer 24-18 / 14-9:0

[0217] 5'-CCAACCGCGAGAAGATGACCCAGG-CATAGCCAGCTAATGACC-CCTCTTCTG-3' (SEQ ID No. 23)

[0218] Beta-actin conventional reverse primer

[0219] 5'-CGGCTA-AGAGAACCAGTGAGAAAGGGC-3' (SEQ ID No. 24) with a 5' tail sequence

[0220] Inter-priming specificity molecular beacon (binds in the inter-priming region of the amplicon)

[0221] 5'-Quasar 705-CCGCTC- CCTCCTTCCTGGCCTCCC -GAGCGG-BHQ2-3' (SEQ ID No. 25), where the nucleotides in the single-stranded loop are underlined

[0222] Target plasmid containing EGFR G719C mutation was purchased from Integrated DNA Technologies and was prepared by inserting a 211 base pair gene fragment into a pUCIDT vector. The mutant plasmid DNA was digested with restriction enzyme Dra I (New England Biolabs). The digestion mixture contained 10 units of Dra I and 4 μg of mutant plasmid DNA in a 20 μL volume containing 50 mM potassium acetate, 20 mM Tris-acetate (pH 7.9), 10 mM magnesium acetate, and 100 μg / mL bovine serum albumin. The reaction was incubated at 37°C for 120 minutes, then inactivated at 65°C for 20 minutes to inactivate the endonuclease.

[0223] Wild-type human DNA (from multiple anonymous donors), catalog number G1521, was purchased from Promega Corporation (Madison, WI). About 9 μg of this DNA was digested at 37°C for 120 minutes in 50 μL of buffer provided by New England Biolabs containing 100 μg / mL bovine serum albumin, 10 mM magnesium acetate, 50 mM potassium acetate, and 20 mM Tris-acetate (pH 7.9) containing 10 units of restriction endonuclease Mse I (New England Biolabs, Ipswich, MA); then inactivated at 65°C for 20 minutes to inactivate the enzyme.

[0224] PCR assays were performed in a 30 μΐ, volume containing 50 mM KCl, 10 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2, 60 mM tetramethylammonium chloride (Sigma-Aldrich), 0.5% Tween 20 (Sigma-Aldrich), 1.5 units of Platinum Taq DNA polymerase (Thermo Fisher Scientific), 250 μΜ ATP, 250 μΜ CTP, 250 μΜ GTP, 250 μΜ TTP, 60 nM each of two different superselective forward primers, 500 nM EGFR G719C superselective reverse primer, 500 nM β-actin regular reverse primer, 300 nM regular molecular beacon for detection of EGFR G719C mutant amplicon, and 500 nM inter-primer specific molecular beacon for detection of β-actin amplicon.

[0225] Amplification was performed in Bio-Rad CFX-96 Touch fluorescent spectrothermal cycler using 0.2 ml white polypropylene tubes (USA Scientific). The thermal cycling program was 95 °C for 2 minutes, followed by 55 cycles of 95 °C for 20 seconds, 60 °C for 20 seconds, and 72 °C for 20 seconds. Molecular beacon fluorescence intensity was measured in FAM channel and Quasar 705 channel at the end of each 60 °C annealing phase.

[0226] Figure 13 Fluorescence intensity readings are shown relative to thermal cycling.

[0227] The first set of 10 reactions, whose results are shown in the upper left panel A, contained 10,000 copies of wild-type genomic DNA plus 500 copies of mutant plasmid DNA; the second set of 10 reactions, whose results are shown in the upper right panel B, contained 10,000 copies of wild-type genomic DNA copies plus 50 copies of mutant plasmid DNA; the third set of 10 reactions, whose results are shown in the lower left panel C, contained 10,000 copies of wild-type genomic DNA plus 5 copies of mutant plasmid DNA; and the fourth set of 10 reactions, whose results are shown in the lower right panel D, contained only 10,000 copies of wild-type genomic DNA and no copies of mutant plasmid DNA.

[0228] The description of the above embodiments and preferred embodiments should be considered as illustrative and not restrictive, and the application as defined in the claims is not to be limited to the embodiments presented herein. It will be apparent to those skilled in the art that various modifications and variations can be made to the methods described herein without departing from the scope of the application as defined in the claims. Such modifications and variations are not to be considered as a departure from the scope of the application, and all such modifications and variations are intended to be included within the scope of the following claims. sequence list <110> Rutgers Biomedical and Health Sciences (PHRI) Fred R. Kramer D.R. Vargas-Gold, Diana R. <120> Assay methods and kits for detecting rare sequence variants <130> 096747.00422 <150> 62 / 909,483 <151> 2019-10-02 <160> 25 <170> PatentIn version 3.5 <210> 1 <211> 89 <212> DNA <213> Artificial <220> <223> Synthetic <400> 1 acctgccgtc aacacgtgcg cagtagacca tctctcttga ggatcttgaa ggaaactgaa 60 cctctccaac gaatctcgaa aagtgctgt 89 <210> 2 <211> 51 <212> DNA <213> Artificial <220> <223> Synthetic <400> 2 ccagggacct taccttatac accggatcct aactgaggtc caaccggagc a 51 <210> 3 <211> twenty four <212> DNA <213> Artificial <220> <223> synthetic <400> 3 ccagggacct taccttatac accg 24 <210> 4 <211> 46 <212> DNA <213> Artificial <220> <223> synthetic <400> 4 ccgcctgacc tgccgtcaac acgtgcgcag tagaccatcc aggcgg 46 <210> 5 <211> 55 <212> DNA <213> Artificial <220> <223> synthetic <400> 5 cgcctgctgg gcatctgcct cacccaacac gtgcgcagta gaccacgctc atcat 55 <210> 6 <211> 41 <212> DNA <213> Artificial <220> <223> synthetic <400> 6 tttgtgttcc cggacatagt ccaatcttcg gtggagctgc a 41 <210> 7 <211> 23 <212> DNA <213> Artificial <220> <223> synthetic <400> 7 tttgtgttcc cggacatagt cca 23 <210> 8 <211> 27 <212> DNA <213> Artificial <220> <223> Synthetic <400> 8 tggtctcaac acgtgcgcag tagacca 27 <210> 9 <211> 55 <212> DNA <213> Artificial <220> <223> Synthetic <400> 9 gccgcctgct gggcatctgc ctcaaagaat caacaagcta caactcgctc atcat 55 <210> 10 <211> 44 <212> DNA <213> Artificial <220> <223> Synthetic <400> 10 ctgcctcacc tccaccgtgc aagcactcgc agaaccttcg gctc 44 <210> 11 <211> 21 <212> DNA <213> Artificial <220> <223> Synthetic <400> 11 caccagttga gcaggtactg g 21 <210> 12 <211> 34 <212> DNA <213> Artificial <220> <223> Synthetic <400> 12 ccgtggctgg actatgtccg ggaacacacc acgg 34 <210> 13 <211> 59 <212> DNA <213> Artificial <220> <223> Synthetic <400> 13 ttgagcaggt actgggagcc aatattgtct gtcctttaca agcacgagtg ccaggaggg 59 <210> 14 <211> 30 <212> DNA <213> Artificial <220> <223> Synthetic <400> 14 ccgtcgcagc tcatgccctt cggccgacgg 30 <210> 15 <211> twenty two <212> DNA <213> Artificial <220> <223> Synthetic <400> 15 tcccggacat agtccaggaa gg 22 <210> 16 <211> 88 <212> DNA <213> Artificial <220> <223> Synthetic <400> 16 acctgccgtc aacacgtgcg cagtagacca tcggcctgct gaaaatgact gaatataaac 60 acacagtctg agcccactct ggagctga 88 <210> 17 <211> 52 <212> DNA <213> Artificial <220> <223> synthetic <400> 17 aaatgattct gaattagctg tatcgtcaag tacccagcta ctaatacgcc at 52 <210> 18 <211> 45 <212> DNA <213> Artificial <220> <223> synthetic <400> 18 caacactggc gcagtagacc ataaacttgt ggtagttgga gcgga 45 <210> 19 <211> 21 <212> DNA <213> Artificial <220> <223> synthetic <400> 19 aggcactctt gcctacgcca t 21 <210> 20 <211> 27 <212> DNA <213> Artificial <220> <223> synthetic <400> 20 tggtctcaac actggcgcag tagacca 27 <210> 21 <211> 89 <212> DNA <213> Artificial <220> <223> synthetic <400> 21 acgtgccctc aatacgagcc cccttcacca actctcttga ggatcttgaa ggaaactgaa 60 cctctccaac gaatctcgaa aagtgctgt 89 <210> twenty two <211> 44 <212> DNA <213> Artificial <220> <223> Synthetic <400> twenty two cgcctgacgt gccctcaata cgagccccct tcaccaacca ggcg 44 <210> twenty three <211> 51 <212> DNA <213> Artificial <220> <223> Synthetic <400> twenty three ccaaccgcga gaagatgacc caggcatagc cagctaatga cccctcttct g 51 <210> twenty four <211> 27 <212> DNA <213> Artificial <220> <223> Synthetic <400> twenty four cggctaagag aaccagtgag aaagggc 27 <210> 25 <211> 30 <212> DNA <213> Artificial <220> <223> Synthetic <400> 25 ccgctccctc cttcctggcc tcccgagcgg 30

Claims

1. Use of a primer pair of a forward primer and a reverse primer specific for a rare DNA target sequence of interest but mismatched to a closely related non-target sequence of interest in the preparation of a kit for use in an asymmetric primer-dependent amplification and detection method, said method comprising: (a) preparing an asymmetric primer-dependent amplification reaction mixture comprising a sample, a DNA polymerase, deoxyribonucleotide triphosphates, an amplification buffer, a homogenous fluorescence detection tool for detecting an amplification product, wherein the homogenous fluorescence detection tool comprises a molecular beacon probe for each rare DNA target sequence of interest, and a primer pair consisting of a forward primer and a reverse primer specific for each rare DNA target sequence of interest but mismatched to a closely related non-target sequence of interest, (b) repeating cycles of the asymmetric primer-dependent amplification reaction mixture by the asymmetric primer-dependent amplification method to amplify each rare DNA target sequence of interest present in the sample, and (c) detecting the rare DNA target sequence of interest by measuring fluorescence intensity from the homogenous fluorescence detection tool; wherein (i) the forward primer is a superselective primer comprising from 5' end to 3' end a first anchor sequence longer than a foot sequence, a first bridge sequence non-complementary to both the rare DNA target sequence of interest and the closely related non-target sequence of interest, and a first foot sequence mismatched to the complementary of the closely related non-target sequence of interest at its 3 '-end or 3 '-second last nucleotide and complementary to the rare DNA target sequence of interest, and (ii) the reverse primer is a superselective primer comprising from 5' end to 3' end a second anchor sequence longer than a foot sequence, a second bridge sequence non-complementary to both the complementary of the rare DNA target sequence of interest and the closely related non-target sequence of interest, and a second foot sequence mismatched to the complementary of the closely related non-target sequence of interest at its 3 '-end or 3 '-second last nucleotide and complementary to the complementary of the closely related non-target sequence of interest; and wherein the method amplifies and detects at least one rare DNA target sequence of interest in a sample of as few as ten copies in a mixture containing 10,000 copies of the closely related non-target sequence of interest different from the rare DNA target sequence of interest by as few as one or two base pairs for each rare DNA target sequence of interest.

2. The use of claim 1, wherein each superselective primer contains a 3 '-end interrogation nucleotide complementary to the rare DNA target sequence of interest but mismatched to the closely related non-target sequence of interest.

3. The use of claim 1, wherein the cycles comprise temperature cycles in an asymmetric polymerase chain reaction method.

4. The use of claim 3, wherein the detection comprises real-time detection.

5. The use of claim 1, wherein the at least one rare DNA target sequence in the sample comprises at least two different rare DNA target sequences, and the homogenous fluorescence detection tool comprises at least two different homogenous fluorescence detection probes for the at least two different rare DNA target sequences, respectively.

6. The use of claim 5, wherein the at least two rare DNA target sequences comprise a set of rare DNA target sequences, and the probes for the rare DNA target sequences in the set are labeled with the same color.

7. The use of claim 1, wherein each different closely related non-target sequence differs from its corresponding rare DNA target sequence by a single base pair, and both the superselective forward primer and the superselective reverse primer are mismatched to the single base pair.

8. The use of claim 1, wherein the superselective forward primer or the superselective reverse primer for each rare DNA target sequence contains a 5 '-tag sequence, and wherein the complement of each 5 '-tag sequence is the target of the probe.

9. The use of claim 5, wherein each probe comprises a sequence complementary to the complement of the first bridge sequence or the second bridge sequence.

10. The use of claim 5, wherein the asymmetric primer-dependent amplification reaction mixture comprises an effective amount of a selectivity enhancing reagent.

11. The use of claim 10, wherein the selectivity enhancing reagent is a Hofmeister salt.

12. The use of claim 11, wherein the Hofmeister salt comprises tetramethylammonium chloride.

Images