Multiplex pyrophosphorolysis activated polymerization to amplify multiple almost-sequence-identical templates in a single reaction

This CIP application claims priority from U.S. non-provisional patent application Ser. No. 15/462,342, filed on Mar. 17, 2017.

This application is being filed along with a Sequence Listing and its electronic format entitled SequenceListing.txt.

The present invention relates to the field of molecular biology and particularly pyrophosphorolysis activated polymerization (PAP) for nucleic acid amplification.

PAP Technology for Nucleic Acid Amplification

Pyrophosphorolysis activated polymerization (PAP) is a method for nucleic acid amplification where pyrophosphorolysis and polymerization are serially coupled by DNA polymerase using 3′ blocked primers (Liu and Sommer, 2000; Liu and Sommer, 2004b). A primer is blocked at the 3′ end with a non-extendable nucleotide (3′ blocker), such as a dideoxynucleotide, and cannot be directly extended by DNA polymerase. When the 3′ blocked primer anneals to its complementary DNA template, DNA polymerase can remove the 3′ blocker from the 3′ blocked primer in the presence of pyrophosphate or its analog, which reaction is called pyrophosphorolysis. The DNA polymerase can then extend the 3′ unblocked primer on the DNA template. In addition to references cited herein, PAP has been described in U.S. Pat. Nos. 6,534,269, 7,033,763, 7,105,298, 7,238,480, 7,504,221, 7,914,995, and 7,919,253.

The serial coupling of pyrophosphorolysis and extension using the 3′ blocked primer in PAP results in an extremely high selectivity (Liu and Sommer, 2004a; Liu and Sommer, 2004b) because a significant nonspecific amplification (Type II error) requires mismatch pyrophosphorolysis followed by mis-incorporation by the DNA polymerase, an event with a frequency estimated to be 3.3×10 −11 .

The bi-directional form of PAP (Bi-PAP) is especially suitable for allele-specific amplification that uses two opposing 3′ blocked primers with a single nucleotide overlap at their 3′ ends (Liu and Sommer, 2004a; Liu and Sommer, 2004b). Bi-PAP can detect one copy of a mutant allele in the presence of 10 9 copies of the wild type DNA without false positive amplifications.

PAP was initially tested with Tfl and Taq polymerases using DNA template of the human dopamine D1 gene, proving the principle that DNA-dependent DNA pyrophosphorolysis and DNA-dependent DNA polymerization can be serially coupled (Liu and Sommer, 2000). The efficiency of PAP was greatly improved using TaqFS, a genetically engineered polymerase comprising a F667Y mutation, which were demonstrated using other DNA templates (Liu and Sommer, 2002).

RNA-PAP was developed that can directly amplify RNA template without additional treatment. RNA-PAP brings in a new mechanism for amplification of RNA template in which RNA-dependent DNA pyrophosphorolysis removes 3′ blocker such as 3′ dideoxynucleotide from a blocked primer when hybridized to RNA template, and then RNA-dependent DNA polymerization extends the activated primer. Due to this serial coupling, RNA-PAP has high selectivity against mismatches on the RNA template, providing highly specific amplification of RNA template (U.S. Pat. No. 9,133,491).

PAP with Acycolonucleotide Blocker and Type II Polymerase

We showed that Type II DNA polymerase efficiently catalyzes template-dependent pyrophosphorolysis to activate primers blocked at their 3′ termini with acyclonucleotides in which a 2-hydroxyethoxymethyl group substitutes for the 2′-deoxyribofuranosyl sugar. Type II DNA polymerases Vent (exo-) and Pfu (exo-) were used for PAP with acyclonucleotide-blocked primers, besides Type I DNA polymerase (Liu and Sommer, 2004c).

Multiplex-PAP at Multiple Loci: Multiple Pairs of Primers Amplify Multiple Templates at Multiple Loci

Advantageous to produce little or no primer-dimer or false priming (Liu and Sommer, 2002), multiple pairs of primers (≥2) were used to amplify multiple potential templates (≥2) located at mutiple loci (≥2) in one reaction (Liu, et al., 2006). In an example, PAP used eight pairs of primers that targeted eight loci in human genome including seven different exons scattered along a 30 Kb sequence of the human factor IX gene and one exon in the human ATM gene.

Inhibitory Interaction in One-Locus-Duplex-PAP: Multiple Pairs of Primers Amplify Multiple Almost-Sequence-Identical Templates at One Locus

We developed many Singleplex-PAP assays for detection of A/T biallelic polymorphisms in human genome, such as Rs4261 and Rs31224 loci. Each polymorphism contains an A or a T nucleotide exactly at the same nucleotide (Table 1).

For the biallelic polymorphism Rs4261, the first pair of blocked primers (SEQ ID No 2 and 3) were regularly designed as 5′-perfect-match primers, which match the A allelic template (SEQ ID 1), but mismatch the T allelic template at the 3′ ends (SEQ ID 4) (Table 1). The Singleplex-PAP amplified the A allelic template, but extremely discriminated against the T allelic template.

The second pair of blocked primers (SEQ ID No 5 and 6) were also regularly designed as 5′-perfect-match primers, which match the T allelic template (SEQ ID 4), but mismatch the A allelic template at the 3′ ends (SEQ ID 1). The Singleplex-PAP amplified the T alleleic template (SEQ ID 4), but did not amplify the A allelic template (SEQ ID 1) at all.

The amplification efficiency of each Singleplex-PAP was measured to be >96% in serial dilution experiments in which the genomic template DNA was 10-fold serially diluted from 10 6 to 10 copies per 20 ul of reaction.

However, when the two pairs of primers (SEQ ID No 2 and 3, 5 and 6), put together in one reaction, to amplify either or both of the two allelic templates (SEQ ID 1 and 4), the One-Locus-Duplex-PAP produced much less corresponding products with the amplification efficiencies 85-87%, leading to 16-fold less products by the end of the 30th cycle, thus indicating inhibitory interaction between the primers.

In another example of the biallelic polymorphism Rs4261, two pairs of blocked primers (SEQ ID No 8 and 9, 11 and 12) were regularly designed as 5′-perfect-match primers for the A and T alleles (SEQ ID 7 and 10) (Table 1). Similar tendencies of the inhibitory interaction were observed in amplification efficiencies, about 10% decreases per cycle between the Singleplex-PAP and One-Locus-Duplex-PAP.

With regularly designed 5′-perfect-match primers, this inhibitory interaction is common in One-Locus-Duplex-PAP. We hypothesize that competitive annealing of multiple almost-sequence-identical primers to their multiple almost-sequence-identical templates leads to the inhibition.

Advantages of the Invention of One-Locus-Multiplex-PAP

In order to amplify multiple potential almost-sequence-identical templates or alleles at one locus, a new design was developed that artificial mutations were introduced into multiple pairs of primers to reduce the inhibitory interaction and thus increase the amplification efficiencies.

A plurality of pairs of forward and reverse blocked primers for pyrophosphorolysis activated polymerization amplify a plurality of potential templates in one reaction, in which the templates are located at one locus in a genome and have at least one nucleotide variance from each other.

The plurality of pairs of forward and reverse blocked primers comprise: 1) a first pair of forward and reverse primers to amplify a first template, in which the forward primer or reverse primer has at least one artificial mutation introduced into its 5′ region, and 2) a second pair of forward and reverse primers to amplify a second template, in which the second forward or reverse primer in the same direction as the above 5′ mutated primer in the first pair has at least one artificial mutation introduced into its 5′ region, and in which the artificial mutations of the 5′ mutated primers in the first pair and in the second pair are located at different nucleotides at the locus of the genome.

A method for multiplex PAP comprises: a) providing a plurality of pairs of forward and reverse blocked primers to amplify a plurality of potential templates in one reaction, in which the templates are located at one locus in a genome and have at least one nucleotide variance from each other, comprising: 1) a first pair of forward and reverse primers to amplify a first template, in which the first forward primer or reverse primer has at least one artificial mutation introduced into its 5′ region, and 2) a second pair of forward and reverse primers to amplify a second template, in which the second forward or reverse primer in the same direction as the above 5′ mutated primer in the first pair has at least one artificial mutation introduced into its 5′ region, in which the artificial mutations of the 5′ mutated primers in the first pair and in the second pair are located at different nucleotides at the locus of the genome, and b) amplifying the templates in one reaction.

The method for multiplex PAP further comprise a step c) sequencing individual molecules of the multiple amplified products in parallel.

Of the method for multiplex PAP, the plurality of pairs of forward and reverse blocked primers further comprise a third pair of forward and reverse primers to amplify a third template, in which the third forward or reverse primer in the same direction as the above 5′ mutated primer in the first and second pairs has at least one artificial mutation introduced into its 5′ region, in which the artificial mutations of the 5′ mutated primers in the first pair, the second pair and the third pair are located at different nucleotides at the locus of the genome.

Of the method for multiplex PAP, the plurality of pairs of forward and reverse blocked primers comprise: 1) the first pair of forward and reverse primers, in which the other primer in the first pair has at least one artificial mutation introduced into the 5′ region, and 2) the second pair of forward and reverse primers, in which the other primer in the second pair has at least one artificial mutation introduced into the 5′ region, in which the artificial mutations of the 5′ mutated primers in the first pair and the second pair are located at different nucleotides at the locus of the genome.

Of the method for multiplex PAP, the 3′ regions of the first pair of primers match the first template but mismatch the second template, and the 3′ regions of the second pair of primers match the second template but mismatch the first template, and in which the first and second templates are located at the same locus but contain at least one nucleotide variance from each other.

Of the method for multiplex PAP, one or two artificial mutations are introduced into the 5′ region of the first forward or reverse primer, whereby the 5′ region substantially but not completely matches its template.

Of the method for multiplex PAP, one or two artificial mutations are introduced into the 5′ region of the second forward or reverse primer, whereby the 5′ region substantially but not completely matches its template.

Of the method for multiplex PAP, the artificial mutation of the 5′ mutated primer in each of the first and second pairs is selected from the group consisting of six types of A to C, C to A, T to G, G to T, A to T, and T to A mutations.

Of the method for multiplex PAP, the artificial mutation of the 5′ mutated primer in each of the first and second pairs result in one of the four types of mismatches of G-A, C-T, A-A, and T-T between the 5′ region of the 5′ mutated primer and the complementary strand of the template.

Of the method for multiplex PAP, the artificial mutation of the 5′ mutated primer in any of the first and second pairs is selected from the group consisting of four types of A to C, C to A, T to G, and G to T mutations.

Of the method for multiplex PAP, the artificial mutation of the 5′ mutated primer in any of the first and second pairs is selected from the group consisting of two types of A to T and T to A mutations.

Of the method for multiplex PAP, the 5′ regions of the 5′ mutated primers in the first and second pairs range from the first to the twelfth nucleotide from the 5′ ends, including the nucleotides at the 5′ ends assigned as the first nucleotides from the 5′ ends.

Of the method for multiplex PAP, the artificial mutations of the 5′ mutated primers in the first and second pairs are at different nucleotides at the locus in the genome.

Of the method for multiplex PAP, the first and second templates are completely or partially overlapped in the locus.

Of the method for multiplex PAP, the first and second templates contain at least one nucleotide variance from each other but are located at the same locus in the genome.

FIG. 1 shows how 5′-artificial-mismatch blocked primers work in One-Locus-Duplex-PAP in exon 19 of the EGFR gene. In the example, two pairs of primers were designed as 5′-artificial-mismatch primers COSM6255 and COSM12369 (SEQ ID No 15 and 16, 19 and 16) (Table 2). Only the 5′ regions of the forward primers (SQE ID 15 and 19) are diagramed with the complementary strands of the starting and duplicated templates COSM6255 (SEQ ID 13 and 14). Underlined and upper cases in the primer sequences are the artificial mutations. Underlined and upper cases in the duplicated templates are artificial mutations duplicated from the 5′-artificial-mismatch primers. The mismatch between the 5′ region of a primer and the complementary strand of a template is indicated by rectangle frame. The number of mismatches is indicated for each case on the right side, showing different levels of complementarities.

FIG. 2 shows comparison of amplification efficiencies in exon 19 of the EGFR gene. In panel A, Singleplex-PAP used a pair of 5′-artificial-mismatch blocked primers COSM6255 (SEQ ID No 15 and 16) to amplify the templates COSM6255 (SEQ ID 13 and 14) form plasmid DNA (Table 2). In panel B, One-Locus-Duplex-PAP used two pairs of primers COSM6255 and COSM12369 (SEQ ID No 15 and 16, 19 and 16) to amplify the templates COSM6255 (SEQ ID 13 and 14) form plasmid DNA. To determine the amplification efficiency, the template was 10-fold serially diluted from 10 6 to 10 copies per 20 ul of reaction. Threshold line is also indicated. X-axis is the cycle number and Y-axis is the net fluorescence signal in arbitrary units. Amplification efficiencies of the Singleplex-PAP and One-Locus-Duplex-PAP were determined together with equations of linear regression and coefficients of determination (R 2 ) by plotting Ct values versus log DNA copies.

FIG. 3 shows how 5′-artificial-mismatch primers work in One-Locus-Triplex-PAP in exon 18 of the EGFR gene. In the example, three pairs of primers were designed as 5′-artificial-mismatch primers COM6252, COSM6253 and COSM6239 (SEQ ID 22 and 23, 26 and 27, 30 and 31) (Table 4). Only the 5′ regions of the forward primers (SEQ ID 22, 26 and 30) are diagramed with the complementary strands of the starting and duplicated templates COM6252 (SEQ ID 20 and 21).

FIG. 4 shows comparison of amplification efficiencies in exon 18 of the EGFR gene. In panel A, Singleplex-PAP used a pair of 5′-artificial-mismatch blocked primers COSM6252 (SEQ ID 22 and 23) to amplify the templates COSM6252 (SEQ ID 20 and 21) form plasmid DNA (Table 4). In panel B, One-Locus-Triplex-PAP used three pairs of primers COSM6252, COSM6253 and COSM6239 (SEQ ID 22 and 23, 26 and 27, 30 and 31) to amplify the templates COSM6252 (SEQ ID 20 and 21) form plasmid DNA.

FIG. 5 shows how 5′-artificial-mismatch primers work in One-Locus-Triplex-PAP in exon 2 of the KRAS gene. In the example, three pairs of primers were designed as 5′-artificial-mismatch primers COSM518, COSM516 and COSM517 (SEQ ID 34 and 35, 38 and 39, 42 and 43) (Table 6). Only the 5′ regions of the forward primers (SEQ ID 34, 38 and 42) are diagramed with the complementary strands of the starting and duplicated templates COSM 518 (SEQ ID 32 and 33).

FIG. 6 shows comparison of amplification efficiencies in exon 2 of the KRAS gene. In panel A, Singleplex-PAP used a pair of 5′-artificial-mismatch blocked primers COSM518 (SEQ ID 34 and 35) to amplify the templates COSM518 (SEQ ID 32 and 33) from plasmid DNA (Table 6). In pane B, One-Locus-Triplex-PAP used three pairs of primers COSM518, COSM516 and COSM517 (SEQ ID 34 and 35, 38 and 39, 42 and 43) to amplify the template COSM518 form plasmid DNA (SEQ ID 32 and 33).

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.

PCR refers to polymerase chain reaction.

Pyrophosphorolysis is the reverse reaction of deoxyribonucleic acid polymerization. In the presence of pyrophosphate, the 3′ nucleotide is removed by a polymerase from duplex DNA to generate a triphosphate nucleotide and a 3′ unblocked duplex DNA: [dNMP] n +PPi→[dNMP] n-1 +dNTP (Deutscher and Kornberg, 1969).

Polymerase or nucleic acid polymerase refers to a polymerase characterized as polymerization or extension of deoxyribonucleic acids.

3′ blocked primer refers to an oligonucleotide with a 3′ non-extendable nucleotide (3′ blocker), such as a dideoxynucleotide or an acycolonucleotide. The 3′ nucleotide could not be directly extended, but it can be removed by pyrophosphorolysis and then the unblocked primer can be extended by polymerase.

PAP refers to pyrophosphorolysis activated polymerization.

Bidirectional-PAP (Bi-PAP) is a form of PAP that uses a pair of opposing blocked primers that overlap by one nucleotide at their 30 termini.

Exponential-PAP is a form of PAP that uses a pair of two opposing forward and reverse primers for exponential product accumulation with cycles. At least one primer is blocked primer.

Sensitivity or detection limit is defined as the smallest copy number of a template that generates a detectable product when the blocked primers match the template at the targeted nucleotide, such as the 3′ end.

Specificity is defined as the largest copy number of a template that generates an undetectable product when the blocked primers mismatch the template at the targeted nucleotide, such as the 3′ end.

Selectivity, the ratio of sensitivity to specificity, is defined as the ability to detect a small number of copies of the matched template in the presence of a large number of copies of mismatched templates without causing false positives.

Thermostable enzyme refers to an enzyme that is heat stable or heat resistant.

TaqFS is a genetic engineered form of Taq polymerase containing G46E and F667Y amino acid changes compared with wild type sequence.

A locus is defined as a short region of nucleotide sequence, such as 40 bp, in genome.

Multiple templates or alleles at one locus mean that at least two templates are located at a short region of nucleotide sequence. The sequence differences among the templates, may be as little as one base substitution, a few base deletion or insertion, and may be located as near as at the same nucleotide. In addition, the alleles may be completely or partially overlapped within the region.

Multiple almost-sequence-identical templates or alleles mean at least two templates, typically located at one locus in genome, among which the sequence differences may be as little as one base substitution, a few base deletion or insertion.

A pair of primers means two opposing forward and reverse primers.

Singleplex-PAP means that one pair of primers amplify one template in a reaction.

Multiplex-PAP means that ≥2 pairs of primers amplify ≥2 potential templates in a reaction.

Multiplex-PAP at multiple loci is a form of multiplex PAP that ≥2 pairs of primers amplify ≥2 potential templates at ≥2 loci in a reaction.

One-Locus-Multiplex-PAP is a form of multiplex PAP that ≥2 pairs of primers amplify ≥2 potential templates at one locus in a reaction.

One-Locus-Duplex-PAP means that 2 pairs of primers amplify 2 potential templates at one locus in a reaction.

One-Locus-Triplex-PAP means that 3 pairs of primers amplify 3 potential templates at one locus in a reaction.

The 5′ region of a primer is the 5′ part of the primer sequence, such as the ten successive nucleotides from the 5′ end.

The 3′ region of a primer is the 3′ part of the primer sequence, such as the ten successive nucleotides from the 3′ end.

Central region of a primer is the middle part of the primer sequence between the 5′ region and the 3′ region.

5′-perfect-match primer: the 5′ region has no artificial mutations and perfectly matches the starting template.

5′-artificial-mismatch primer: artificial mutations are introduced into the 5′ region, resulting to mismatch to the starting template.

5′ mutated primer: i.e., 5′-artificial-mismatch primer, artificial mutations are introduced into the 5′ region.

Artificial mutation means the mutation that is artificially introduced into primer sequences for substitution, typically in the 5′ region.

Artificial mismatch is formed between the artificial mutation in the 5′ region of 5′-artificial-mismatch primer and the template.

Starting template is the original template before amplification starts, such as that from genomic or plasmid DNA template.

Duplicated template is duplicated from the starting template in amplification and can also be taken as template in later cycles.

Terminology of Real-Time Fluorescence Detection

Baseline is the level of fluorescence signal during initial cycles. The low level can be considered as background or “noise” of the reaction.

Threshold is defined as the level of fluorescence signal that is a significant higher than baseline signal and can distinguish amplification signal from the background.

Ct (threshold cycle) is the cycle number at which the fluorescence signal crosses the threshold.

Amplification efficiency is defined as the percent of template that is amplified by the end of a cycle.

Principle of 5′-Artificial-Mismatch Primers for One-Locus-Multiplex-PAP

In order for multiple pairs of blocked primers to amplify multiple almost-sequence-identical templates or alleles at one locus without inhibitory interaction, a novel design of 5′-artificial-mismatch primers was developed that contain artificial mutations in the 5′ regions, as in Examples 2-4.

a) Four Types of Artificial Mismatches Preferred

Mismatches in short DNA duplexes significantly reduce their thermal stabilities, the levels depending on the type of mismatches. The order of thermal stabilities of a total of eight possible mismatches are approximately: G-T>G-G>G-A>C-T>A-A=T-T>A-C=C-C (mismatch G-T=T-G, G-A=A-G, C-T=T-C, and A-C=C-A) (Modrich, 1987) (Aboul-ela, et al., 1985) (Ikuta, et al., 1987).

We prefer four types of mismatches of G-A, C-T, A-A and T-T because 1) their thermal stabilities are medium in the order: they can disrupt the structures of short DNA duplexes, the levels being not too little and not too much, and 2) their thermal stabilities are within a successive range in the order.

b) The Four Types of Mismatches Caused by Six Types of Artificial Mutations

Considering a One-Locus-Multiplex-PAP, at least two pairs of primers are applied to at least two potential templates. We chose six types of artificial mutations of A to C and C to A, T to G and G to T, A to T and T to A in primers, which always lead to the four types of artificial mismatches between the primers and the complementary strands of the starting or duplicated templates. The other six possible types of artificial mutations of A to G and G to A, T to C and C to T, G to C and C to G are not used at all in the design.

The A to C or C to A artificial mutations cause two types of mismatches of C-T and A-G (mismatch C-T=T-C, and A-G=G-A) between the primers and the complementary strands of the starting or duplicated templates. The T to G and G to T artificial mutations cause the same two types of mismatches of G-A and T-C (mismatch G-A=A-G, and T-C=C-T) between the primers and the complementary strands of the starting or duplicated templates. The A to T and T to A artificial mutations cause other two types of mismatches of T-T and A-A between the primers and the complementary strands of the starting or duplicated templates. Thus, a total of four types of mismatches are counted in One-Locus-Multiplex-PAP.

c) The Number of Artificial Mismatches Caused by Artificial Mutations

When artificial mutations of the forward or reverse primers are designed at different nucleotides in the templates, the number of mismatches between the 5′ region of a primer and the complementary strand of a starting or duplicated template depends on the number of artificial mutations of the primer and on the template. In any case, the minimum number of mismatches between the 5′ region of a primer and the complementary strand of a duplicated template is zero, such as in Examples 2-4.

d) Locations of Artificial Mismatches Preferred to Localize in the 5′ Region of Primers

We prefer to localize artificial mismatches in the 5′ regions of primers because 1) besides the types, the locations of mismatches also affect thermal stability of short DNA duplexes (Modrich, 1987) (Piao, et al., 2008), and 2) We found that 28-30mer blocked primers commonly had >90% efficiency of pyrophosphorolysis and extension when mismatches vary from the 1 st to 12 th nucleotides from the 5′ ends, i.e., the 5′ regions. However, they had very low efficiency of pyrophosphorolysis and extension when mismatches are located in the 3′ regions, particularly at the 3′ ends.

Thus, artificial mutations are preferred to localize in the 5′ regions, ranging from the 1 st to the 12 th nucleotide from the 5′ ends, better ranging from the 3 rd to 9 th nucleotide from the 5′ ends.

e) Mechanism by Different Numbers of Mismatches Between Different Primers and Different Templates in One-Locus-Multiplex-PAP

In a One-Locus-Multiplex-PAP, ≥2 pairs of primers amplify ≥2 potential templates at one locus in a reaction. For a 5′-artificial-mismatch primer, such as each of the forward primers of the first pair, the second pair and the third pair, an artificial mutation is designed into the 5′ region. The number of mismatches between the 5′ regions and the complementary strands of the starting or duplicated templates varies, such as from zero to 4, depending on annealing of a specific primer to a specific template. Thus, the different numbers of mismatches between different primers and different templates provide the mechanism to reduce inhibitory interaction: 1) the 5′ region of a 5′-atifitial-mismatch primer matches its corresponding templates more than its competing templates, and 2) a template matches the 5′ region of its corresponding 5′-atifitial-mismatch primer more than its competing 5′-atifitial-mismatch primers, as in Examples 2-4.

Aboul-ela F, Koh D, Tinoco I, Jr., Martin F H. 1985. Base-base mismatches. Thermodynamics of double helix formation for dCA3XA3G+dCT3YT3G (X, Y=A, C, G, T). Nucleic Acids Res 13(13):4811-24.

Deutscher M P, Kornberg A. 1969. Enzymatic synthesis of deoxyribonucleic acid. 28. The pyrophosphate exchange and pyrophosphorolysis reactions of deoxyribonucleic acid polymerase. J Biol Chem 244(11):3019-28.

Ikuta S, Takagi K, Wallace R B, Itakura K. 1987. Dissociation kinetics of 19 base paired oligonucleotide-DNA duplexes containing different single mismatched base pairs. Nucleic Acids Res 15(2):797-811.

Liu Q, Nguyen V Q, Li X, Sommer S S. 2006. Multiplex dosage pyrophosphorolysis-activated polymerization: application to the detection of heterozygous deletions. Biotechniques 40(5):661-8.

Liu Q, Sommer SS. 2000. Pyrophosphorolysis-activated polymerization (PAP): application to allele-specific amplification. Biotechniques 29(5):1072-1080.

Liu Q, Sommer SS. 2002. Pyrophosphorolysis-activatable oligonucleotides may facilitate detection of rare alleles, mutation scanning and analysis of chromatin structures. Nucleic Acids Res 30(2):598-604.

Liu Q, Sommer SS. 2004a. Detection of extremely rare alleles by bidirectional pyrophosphorolysis-activated polymerization allele-specific amplification (Bi-PAP-A): measurement of mutation load in mammalian tissues. Biotechniques 36(1):156-66.

Liu Q, Sommer SS. 2004b. PAP: detection of ultra rare mutations depends on P* oligonucleotides: “sleeping beauties” awakened by the kiss of pyrophosphorolysis. Hum Mutat 23(5):426-36.

Liu Q, Sommer SS. 2004c. Pyrophosphorolysis by Type II DNA polymerases: implications for pyrophosphorolysis-activated polymerization. Anal Biochem 324(1):22-8. Modrich P. 1987. DNA mismatch correction. Annu Rev Biochem 56:435-66.

Piao X, Sun L, Zhang T, Gan Y, Guan Y. 2008. Effects of mismatches and insertions on discrimination accuracy of nucleic acid probes. Acta Biochim Pol 55(4):713-20.