Field of invention

The present invention provides a method for detection of known mutations and differences in nucleotide sequences. The principle of the method is to simultaneously amplify sequences comprising the mutation(s) to be detected and sequences not comprising the mutation(s) to be detected using a three primer system comprising a mutant primer, a non-mutant primer and a common primer. In one embodiment of the invention the mutant primer and the non-mutant primer bind competitively to the same nucleotide sequence. The resulting mutant and non-mutant amplicons are designed to melt differently to enable direct detection by melting analysis. Background of invention

As the fields of cancer diagnostics and treatment move towards the era of personalized medicine, the identification of molecular markers becomes increasingly important. Several well characterized mutations have already been implemented in the clinics as markers of response to different treatment strategies and more are underway.

However, the use of mutations as molecular markers in the clinics presents a challenge to the methodologies employed for their detection. When applying a particular methodology for routine diagnostic testing it must achieve a sufficient sensitivity without compromising specificity, while being convenient and cost-efficient.

Colorectal cancer (CRC) and non-small cell lung cancer (NSCLC) are among the most frequent causes of cancer deaths worldwide. However, new targeted therapies are continuously being developed and may contribute to increased survival rates among these patients in the years to come. The epidermal growth factor Receptor (EGFR) is often overexpressed in CRC and NSCLC, and contributes to cancer development and progression by stimulating proliferation, angiogenesis, invasion, and survival of cancer cells. A subset of NSCLC patients carrying activating somatic mutations in the tyrosine kinase domain of EGFR show excellent response to EGFR tyrosine kinase inhibitors such as gefitinib and erlotinib, and metastatic CRC patients with activating KRAS mutations are unlikely to respond to treatment with monoclonal antibodies against EGFR such as panitumum-ab and cetuxim-ab. Other oncogenic mutations, including the S ¾4F c.1799T>A mutation, and PIK3CA mutations, as well as loss of PTEN expression, may also be predictive markers of resistance to anti-EGFR monoclonal antibodies but require further evaluation before being incorporated in clinical practice. Activating mutations in KRAS and BRAF are found in approximately 40-50% and 10- 15% of CRC patients, respectively, and found to be mutually exclusive. However, the frequency of detected KRAS mutations in clinical samples is influenced by the sensitivity of the method employed for their detection. This may in part be caused by intra tumor heterogeneity and contamination with wild-type DNA from normal cells, which typically are observed in infiltrating cancer types such as pancreatic, colorectal, breast and lung cancer. Other clinical applications also require highly sensitive mutation detection, for instance, the monitoring of minimal residual disease after treatment, monitoring of relapse caused by the emergence of resistance mutations, and identification of somatic mutations in early tumorigenesis. For these reasons, the development of sensitive, reliable, and cost-effective methods for mutation testing is of paramount importance.

Methods based on standard PCR and subsequent assays for mutation detection such as traditional Sanger sequencing suffer from a relatively low sensitivity as mutant alleles must be present in a proportion of at least 10-20% to be reproducibly detected (Vogelstein, B. and Kinzler, K.W. (1999), Proc Natl Acad Sci U S A, 96, 9236-9241). Standard PCR followed by pyrosequencing or high-resolution melting (HRM) is usually more sensitive, the sensitivity being 5-10%. However, the sensitivity limit of standard PCR followed by HRM is not sufficient to identify for example all metastatic CRC patients that carry somatic KRAS mutations (Kristensen, L.S. et al., (2010), Hum Mutat, 31 , 1366-1373). Two kits for the detection of KRAS mutations, the TIB Molbiol KRAS mutation detection kit (LightMix Kit, TIB Molbiol) and the TheraScreen ^® kit (DxS

Diagnostic Innovations), have been approved by the FDA. These kits are more sensitive but also more time-consuming and less cost-effective compared to HRM followed by sequencing of positive samples (Whitehall, V. et al. (2009), J Mol Diagn, 1 1 , 543-552). Therefore, there is a need for more sensitive and cost-effective methods to detect mutations.

Summary of invention

The present invention provides a method for detection of known mutations and differences in nucleotide sequences. The principle of the method is to simultaneously amplify sequences comprising the mutation(s) to be detected and sequences not comprising the mutation(s) to be detected using a three primer system comprising a mutant primer, a non-mutant primer and a common primer. The method is termed Competitive Amplification of Differentially Melting Amplicons (CADMA). In one embodiment of the invention the mutant primer and the non-mutant primer bind competitively to the same nucleotide sequence. The present invention provides a highly sensitive method that allows easy detection of known mutations in nucleotide sequences. In addition the method as described herein enables very sensitive mutation detection regardless of the melting properties of the mutations to be detected.

One aspect of the present invention relates to a method for detecting the absence or presence of at least one mutation in a nucleotide sequence in a sample, said method comprising the steps of

a) providing at least three primers, wherein at least one of said primers is a mutant primer comprising a nucleotide sequence complementary to a region of a mutant nucleotide strand comprising the mutation to be detected and wherein at least one of said primers is a non-mutant primer comprising a sequence that is complementary to a region of a non-mutant nucleotide strand not comprising the mutation to be detected, and wherein at least one of said primers is a common primer comprising a sequence that binds to the nucleotide strand complementary to the nucleotide strands to which the mutant and non-mutant primer bind such that the extension product of said common primer comprises a region complementary to extension products of the mutant and non-mutant primers,

b) contacting the sample with an oligonucleotide system under hybridization

conditions so as to form a reaction mixture, said oligonucleotide system including the at least two competitive primers and the at least one common primer,

c) initiate polymerase chain reaction (PCR) and run at least one cycle of PCR so that if the mutation to be detected is present it will result in simultaneous amplification of mutant and non-mutant nucleotide sequences, wherein the at least one mutant primer and the at least one common primer result in the synthesis of least one mutant amplicon comprising the mutation to be detected and having a melting temperature x, and wherein the at least one wild type primer and the at least one common primer result in the synthesis of at least one non-mutant amplicon not comprising the mutation to be detected and having a melting temperature y different from the temperature x and

detecting the presence or absence of said mutation by melting analysis In one preferred embodiment the at least one mutant primer and the at least one non- mutant primer are competitive primers comprising a sequence that results in competitive binding of said competitive primers to the same nucleotide sequence.

The at least one mutation is in one embodiment a DNA mutation. The DNA mutation may be a point mutation, a deletion of one or more nucleotides or an insertion of one or more nucleotides.

In one embodiment the mutant primer introduces at least one mutation that results in a decrease in the melting temperature of the mutant amplicon. The mutant primer may introduce at least one G>A mutation, at least one C>A mutation, at least one G>T mutation and/or at least one C>T mutation.

In another embodiment the mutant primer introduces at least one mutation that results in an increase in the melting temperature of the mutant amplicon. The mutant primer may introduce at least one A>G mutation, at least one A>C mutation, at least one T>G mutation and/or at least one T>C mutation.

In one embodiment the non-mutant primer introduces at least one mutation that results in a decrease in the melting temperature of the non-mutant amplicon. The non-mutant primer may introduce at least one G>A mutation, at least one C>A mutation, at least one G>T mutation and/or at least one C>T mutation.

In another embodiment the non-mutant primer introduces at least one mutation that results in an increase in the melting temperature of the non-mutant amplicon. The non- mutant primer may introduce at least one A>G mutation, at least one A>C mutation, at least one T>G mutation and/or at least one T>C mutation.

It is appreciated that the 3' end of the mutant primer comprises a sequence

complementary to the mutation to be detected. The sequence of the non-mutant primer is in one embodiment complementary to a region of the non-mutant nucleotide strand that is identical to the mutant nucleotide strand. Thus, annealing of the non-mutant primer to the mutant nucleotide strand may result in a third amplicon comprising the mutation to be detected and having a melting temperature z different from the melting temperature x. It is preferred that the mutant primer comprises a sequence complementary to said third amplicon such that annealing of the mutant primer and the common primer to the third amplicon and at least one cycle of PCR results in an amplicon that is identical to the mutant amplicon. In another embodiment the sequence of the non-mutant primer is complementary to a region of the non-mutant nucleotide strand that is not identical to the mutant nucleotide strand, such that the non-mutant primer has a preference for the non-mutant nucleotide strand. It is preferred that the mutant primer is present in a higher concentration than the non- mutant primer.

The mutant amplicon may in one embodiment be shorter than the non-mutant amplicon. In another embodiment the mutant amplicon is longer than the non-mutant amplicon.

It is appreciated that the difference between the melting temperature x and the melting temperature y is at least 0.2 degrees Celsius. The sample as described herein may be a tissue sample or a body fluid sample. The body fluid sample can for example be selected from the group consisting of blood samples, plasma samples, serum samples, semen samples and urine samples.

In one embodiment the method of the present invention is sensitive to at least 0.25% mutant alleles. In another embodiment the method is sensitive to at least 0.05% mutant alleles. In a preferred embodiment the method is sensitive to at least 0.025% mutant alleles

The at least one mutation to be detected may for example be in the human KRAS gene (SEQ ID NO: 1). The mutation to detect is in one embodiment the human KRAS c.35G>A mutation, the human KRAS c.35G>T mutation or the human KRAS c.35G>C mutation (base pair no. 5571 of SEQ ID NO: 1). In another embodiment the at least one mutation to detect is the human KRAS c.34G>A mutation, the human KRAS c.34G>T mutation or the human KRAS c.34G>C mutation (base pair no. 5570 of SEQ ID NO: 1). In yet another embodiment the at least one mutation to detect is the human KRAS c.34G>A mutation (base pair no. 5574 of SEQ ID NO: 1).

The at least one mutation to be detected can also be the human BRAF gene (SEQ ID NO: 2). This mutation is in one embodiment the human BRAF c.1799T>A mutation (base pair no. 171429 of SEQ ID NO: 2).

The at least one mutation may for example be in the human EGFR gene (SEQ ID NO: 3). In one embodiment the mutation is the human EGFR c.2573T>G mutation (base pair no. 172791 of SEQ ID NO: 3). The at least one mutation to be detected can also be the human PIK3CA gene (SEQ ID NO: 16). This mutation is in one embodiment the human PIK3CA c.3140A>G mutation (base pair no. 86775 of SEQ ID NO: 16).

It is appreciated that the melting analysis as described herein is high resolution melting analysis

The method of the present invention may be combined with COLD-PCR to increase the sensitivity of the method. A second aspect of the present invention pertains to a kit for detecting the absence or presence of a mutant DNA sequence in a sample, said kit comprising at least one mutant primer, at least one non-mutant primer and at least one common primer according to claim 1. The kit may further comprise a temperature resistant DNA polymerase and appropriate substrates, nucleotides and cofactors to initiate amplification of DNA sequences. The kit may also comprise intercalating dyes for HRM analysis. Description of Drawings

Fig. 1. In the present method mutated and wild-type sequences are amplified simultaneously using a three primer system. The first primer, which is the mutant primer is designed to amplify only mutated sequences and to introduce one or more melting temperature decreasing mutations in the amplicon that contains the known mutation. The second primer, which is the non-mutant primer, is in the embodiment illustrated in Fig. 1 designed to amplify both non-mutant and mutant sequences and to anneal in the same region as the mutant primer that results in competition for target binding between the mutant and non-mutant primers. The non-mutant primer can also be designed to give a slightly larger non-mutant amplicon relative to the mutant amplicon thereby contributing to a higher melting temperature of non-mutant amplicons. Finally, a common primer is designed to amplify both non-mutant and mutant sequences. The resulting two amplicons melt differently. The difference in melting temperature can be used to detect low-abundance mutations, for instance, by melting analysis such as for example HRM analysis.

Fig. 2. Heteroduplex formations between non-mutant and mutant sequences can be identified as an early melting peak in the melting curve. If heteroduplexes are identified in samples known to comprise only non-mutant sequences this implies that amplification from the mutant primer has occurred, in spite of the mismatches between the mutant primer and non-mutant sequences. This may give rise to false positive results, and should be avoided by increasing the annealing temperature, optimizing relative primer concentrations, or designing new primers.

Fig. 3. The analytical sensitivity of the method. The sensitivity was defined as the dilution point at which all three replicates could be distinguished from 10 non-mutant replicates when observing the difference graphs. Fifty nanograms of input DNA were used in the experiments. A. The BRAF c.1799T>A assay was sensitive to 0.25% mutant alleles in a non-mutant background. B. The EGFR c.2573T>G assay was sensitive to 0.25% mutant alleles in a non-mutant background. C. The KRAS c.35G>C assay was sensitive to 0.025% mutant alleles in a non-mutant background. D. The PIK3CA C.3140 A>G assay was sensitive to 0.25% mutant alleles in a non-mutant background. Fig. 4. The analytical sensitivity of the method when combined with COLD-PCR. The sensitivity was defined as the dilution point at which all three replicates could be distinguished from 10 non-mutant replicates when looking at the difference graphs. Fifty nanograms of input DNA were used in the experiments. A. The BRAF c.1799T>A assay was sensitive to 0.025% mutant alleles in a non-mutant background. B. The EGFR c.2573T>G assay was sensitive to 0.025% mutant alleles in a non-mutant background. C. The KRAS c.35G>C assay was sensitive to 0.025% mutant alleles in a non-mutant background. Fig. 5. The analytical sensitivity can be increased by increasing the amount of input DNA in the reactions. The KRAS c.35G>C CADMA combined with COLD-PCR assay was repeated using 250 nanograms of DNA instead of 50 nanograms. The use of five times as much DNA increased the analytical sensitivity by a factor of five. This experiment was performed in duplicates.

Fig. 6. Representative results from the screening of colorectal cancer specimens derived from FFPE tissues using the BRAF c.1799T>A CADMA combined with COLD- PCR assay. Fig. 7. The BRAF c.1799T>A CADMA combined with COLD-PCR assay performed on a serial dilution of a colorectal cancer specimen, for which a BRAF mutation was detected, into wild-type DNA. The 0.39% dilution point was readily distinguishable from 10 wild-type replicates. This experiment was performed in duplicates. Fig. 8. A and B: A. Amplification is performed in the absence of non-mutant primer/competitive primer. False amplification, i.e. the mutant primer anneals to and amplifies from non-mutant DNA sequences, is observed in all wild type samples. For example, the diagram shows that amplicons are generated in samples only containing non-mutant DNA sequences (0% mutant alleles), demonstrating that the mutant primer amplifies non-mutant sequences leading to a false positive result. B. Amplification is performed in the presence of a competing non-mutant primer. The diagram shows that when a competitive primer is present (non-mutant primer) no amplification of non- mutant DNA sequences is observed in samples not containing mutant DNA (0% mutant alleles). Further, the mutant DNA sequences are amplified earlier in the PCR process, which is especially evident for samples having a low concentration of mutant DNA sequences. C. Amplification is performed in the absence of non-mutant

primer/competitive primer. Amplification curves for the BRAF CADMA assay performed without the overlapping primer. False-amplification from ten out of ten wild- type replicates is observed. The replicates containing 0.05% mutated alleles amplified after approximately 33 cycles. D. Amplification is performed in the presence of a competing non-mutant primer. Amplification curves for the BRAF CADMA assay. Robust amplification from all dilution points is observed. The replicates containing 0.05% mutated alleles amplified after approximately 23 cycles. Fig. 9. Melting curves of data shown in Figure 8. A. Melting curves for the EGFR

CADMA assay performed in the absence of non-mutant primer/competitive primer. It can be observed that the amplified wild-type replicates melts at the same temperature as the replicates containing the mutation. B. Melting curves for the EGFR CADMA assay. No heteroduplexes, which melt at approximately 77°C, can be observed in the wild-type replicates indicating that no false amplification from the mutant primer has occurred and that the presence of a competitive primer (the non-mutant primer) prevents false amplification. A. Melting curves for the BRAF CADMA assay performed without the non-mutant primer primer. It can be observed that the amplified wild-type replicates melts at the same temperature as the replicates containing the mutation. B. Melting curves for the EGFR CADMA assay. No heteroduplexes, which melt at approximately 74°C, can be observed in the wild-type replicates indicating that no false amplification from the mutant primer has occurred.

Fig. 10. Competition between the mutant primer and the non-mutant primer. A. Melting curves for the EGFR CADMA assay performed using 200 nM of each primer and an annealing temperature of 65°C. No heteroduplexes, which melt at approximately 77°C, can be observed in the wild-type replicates. The sample containing 0.05% mutant alleles could not be distinguished from the samples not containing mutant alleles (wild type samples). B. Melting curves for the EGFR CADMA assay performed using 100 nM of the non-mutant competitive primer and 200 nM of the mutant primer and an annealing temperature of 65°C. Heteroduplexes can now be observed in samples not containing mutant alleles (wild type samples). Thus the use of a higher relative concentration of a competitive primer (the non-mutant primer) can prevent false amplification from wild-type sequence by the mutant primer. C. Melting curves for the EGFR CADMA assay performed using 100 nM of the non-mutant competitive primer and 200 nM of the mutant primer and an annealing temperature of 70°C.

Heteroduplexes in the wild-type replicates have disappeared. The dilution point containing 0.05% mutant alleles can now be distinguished from the wild-type replicates. Fig. 11. Selected CADMA, pyrosequencing, and Sanger sequencing data obtained from testing the cell line dilutions. A. Melting curves from the CADMA experiments. For clarity, not all dilutions are shown for clarity. B. Pyrosequencing results for the dilutions containing a theoretical fraction of 10% and 50% mutant alleles. C. Sanger sequencing results for the dilutions containing a theoretical fraction of 30% and 50% mutant alleles. The position of mutation is indicated by the arrow.

Fig. 12. Selected CADMA results from testing of the FFPE malignant melanoma samples. It can be observed from the shape of the melting curves that CADMA can distinguish low-level mutations from high-level mutations. Sample 19 contained a high fraction of mutated alleles as indicated by the low ACt value (2.0) in the TaqMan assay. Sample 24 contained a low fraction of mutated alleles as indicated by the high ACt value (6.3) in the TaqMan assay. Sample 23 was mutation negative.

Fig. 13. The analytical sensitivity and specificity of the CADMA assays performed using the Rotorgene 6000. Ten wild-type replicates were run together with a standard dilution series of mutant alleles from cell lines carrying the relevant mutations in a wild-type background (50%, 10%, 1 %, and 0.5%) in triplicates. The three replicates of the 0.5% standard could all be distinguished from ten wild-type replicates in all assays. Fig. 14. The analytical sensitivity and specificity of the CADMA assays performed using the Rotorgene Q. Ten wild-type replicates were run together with a standard dilution series of mutant alleles from cell lines carrying the relevant mutations in a wild-type background (50%, 10%, 1 %, and 0.5%) in triplicates. The three replicates of the 0.5% standard could all be distinguished from ten wild-type replicates in all assays.

Fig. 15. Examples from screening of mCRC samples using the c.34 G>T CADMA assay. A. Real-time amplification data. High background fluorescence can be observed for sample ID 69. B. The derivative of the raw melting data (melt curve analysis). Sample ID 56 carry the c.34 G>T mutation. Sample ID 2, 69 and 72 were negative for the c.34 G>T mutation. Sample ID 69 and 72 may carry another KRAS mutation as small deviations from the wild-type replicates can be observed. Sample ID 69, which gave high fluorescence during the PCR amplification, has deviating melt curves from 82 to 88°C. C. Normalized HRM difference graph. Of the mCRC samples shown only one (sample ID 56) deviate more from the wild-type replicates than the standard containing 1 % mutant alleles.

Fig. 16. Examples from screening of mCRC samples using the c.38 G>T CADMA assay. The wild-type mCRC samples showed more variation in the c.38 G>A CADMA assay compared to any of the other CADMA assays. A. The derivative of the raw melting data (melt curve analysis). No heteroduplexes, which melt between 71 and 74°C, can be observed in any of the wild-type mCRC samples. B. Normalized HRM difference graph. Sample ID 2 deviate more from the wild-type replicates than the standard containing 1 % mutant alleles from approximately 79 to 81 °C. This should not be interpreted as a c.38 G>T mutation, since no heteroduplexes are present.

Detailed description of the invention

Definitions The term 'nucleotides' as used herein refers to both natural nucleotides and non- natural nucleotides capable of being incorporated - in a template-directed manner - into an oligonucleotide, preferably by means of an enzyme comprising DNA or RNA polymerase activity, including variants and functional equivalents of natural or recombinant DNA or RNA polymerases. Corresponding binding partners in the form of coding elements and complementing elements comprising a nucleotide part are capable of interacting with each other by means of hydrogen bonds. The interaction is generally termed "base-pairing". Nucleotides may differ from natural nucleotides by having a different phosphate moiety, sugar moiety and/or base moiety. Nucleotides may accordingly be bound to their respective neighbour(s) in a template or a complementing template by a natural bond in the form of a phosphodiester bond, or in the form of a non-natural bond, such as e.g. a peptide bond as in the case of PNA (peptide nucleic acids). Nucleotides according to the invention includes ribonucleotides comprising a nucleobase selected from the group consisting of adenine (A), uracil (U), guanine (G), and cytosine (C), and deoxyribonucleotide comprising a nucleobase selected from the group consisting of adenine (A), thymine (T), guanine (G), and cytosine (C). Nucleobases are capable of associating specifically with one or more other nucleobases via hydrogen bonds. Thus it is an important feature of a nucleobase that it can only form stable hydrogen bonds with one or a few other nucleobases, but that it can not form stable hydrogen bonds with most other nucleobases usually including itself. The specific interaction of one nucleobase with another nucleobase is generally termed "base-pairing". The base pairing results in a specific hybridisation between predetermined and complementary nucleotides. Complementary nucleotides according to the present invention are nucleotides that comprise nucleobases that are capable of base-pairing. Of the naturally occurring nucleobases adenine (A) pairs with thymine (T) or uracil (U); and guanine (G) pairs with cytosine (C). Accordingly, e.g. a nucleotide comprising A is complementary to a nucleotide comprising either T or U, and a nucleotide comprising G is complementary to a nucleotide comprising C.

The nucleotides may also be a locked nucleic acid (LNA), often referred to as inaccessible RNA, is a modified RNA nucleotide. The ribose moiety of an LNA nucleotide is modified with an extra bridge connecting the 2' oxygen and 4' carbon. The bridge "locks" the ribose in the 3'-endo (North) conformation, which is often found in the A-form duplexes.

The term "oligonucleotide primer" as used herein refers to a molecule comprising at least three deoxyribonucleotides or ribonucleotides. The oligonucleotide primer is capable of acting as a point of initiation of nucleotide synthesis when placed under conditions which induce synthesis of a primer extension product complementary to a nucleotide strand. The conditions can include the presence of nucleotides and a polymerase at a suitable temperature and pH. Although the primer preferably is single stranded, it may alternatively be double stranded. If it is double stranded, the primer must first be treated to separate its strands before it is used to produce extension products. In the preferred embodiment, the primer is an oligodeoxyribonucleotide. The primer must be sufficiently long to prime the synthesis of extension products in the presence of the polymerase. The "oligonucleotide primer" may be a 5' primer, which has a sequence that is complementary to the sense strand of a double stranded DNA molecule. The "oligonucleotide primer" may also be a 3' primer, which has a sequence that is complementary to the antisense strand of a double stranded DNA molecule. In a preferred embodiment the oligonucleotide primer is a DNA oligonucleotide primer. The sense strand of a DNA molecule is complementary to the antisense strand. The antisense strand is the strand of DNA transcribed into mRNA during transcription.

The term "common primer" as used herein is an oligonucleotide primer which binds to the nucleotide strand complementary to the strand that the mutant and non-mutant oligonucleotide primers bind and the common primer binds at a site distant from the mutant and non-mutant oligonucleotide primers. The common primer may be a 3' primer or a 5' primer. This distance should be sufficient to allow the synthesis of extension product between the two binding sites, and close enough such that the extension products of the common primer(s) extends beyond the sequence that is complementary to the mutant and non-mutant oligonucleotide primers, and the extension products of the mutant and non-mutant oligonucleotide primers extend beyond the sequence that is complementary the common primer(s). The extension product(s) from the common primer(s) is complementary to the extension products from the mutant and non-mutant primers. The extension products are single stranded. If the common primer binds to the nucleotide strand comprising the mutation to be detected, that is the mutant nucleotide strand, DNA synthesis from the common primer will result in a mutant extension product comprising the mutation to be detected. If the common primer binds to the nucleotide strand not comprising the mutation to be detected, the non-mutant nucleotide strand, DNA synthesis from the common primer will result in a non-mutant extension product not comprising the mutation to be detected.

As used herein the term "base mismatch" refers to a change in the nucleotides, such that when for example a primer anneals to a nucleotide sequence an abnormal base pairing of nucleotides is formed such as for example base pairing between G-G, C-C, A-A, T-T, A-G, A-C, T-G, or T-C. Normally guanine (G) binds to cytosine (C) and adenine (A) binds to thymine (T) in the formation of double stranded nucleic acids. The standard base pairing is A-T or G-C. Thus, base pairing between A-T or G-C is not a base mismatch.

As used herein the term "point mutation" refers to a mutation wherein a single nucleotide is exchanged for another. The point mutation may be an A>G mutation, an A>C mutation, an A>T mutation, a T>G mutation, a T>C mutation, a T>A mutation, a G>T mutation, a G>C mutation, n G>A mutation, a C>G mutation, a C>T mutation or a C>A mutation. A>G means that A is replaced with G.

The term "amplicon" as used herein refers to a nucleotide sequence that is amplified. In the present invention, the nucleotide sequence that is amplified is the sequence between a competitive primer (which may for example be the 5' primer) and a common primer (which may for example be the 3' primer). The amplicon results from annealing of the extension products of a common primer and a competitive primer. Thus, the amplicon is a double stranded nucleotide and comprises the nucleotide sequence of the primers and the nucleotide sequence between the 5' primer and 3' primer. In a preferred embodiment the amplicon is a double stranded DNA molecule.

The term "polymerase" as used herein refers to an enzyme that catalyses the synthesis of a polynucleotide sequence such as RNA or DNA against a nucleotide template strand by adding free nucleotides to the growing polynucleotide sequence using base- pairing interactions. Thus, a polymerase catalyses the polymerization of nucleotides into a polynucleotide sequence using an intact nucleotide strand as a template. In a preferred embodiment the polymerase is a DNA polymerase. A DNA polymerase is an enzyme that catalyzes the polymerization of deoxyribonucleotides using a DNA strand as a template. In a preferred embodiment the DNA polymerase is a Taq polymerase from Thermus aquaticus, which is a thermostable DNA polymerase having an optimum temperature for activity of about 70 to 80°C. Typically a temperature of 72 °C is used.

The term "template" as used herein refers to the nucleotide sequence or the nucleotide region to be amplified by PCR. The template may be a DNA or an RNA template. It is preferred that the template is a DNA template. If the template is an RNA template, the RNA must be converted into DNA using a reverse transcriptase that synthesises single-stranded DNA using RNA as a template. The term "annealing" as used herein refers to the pairing of complementary DNA or RNA sequences by hydrogen bonding to form a double-stranded polynucleotide. The term is for example used to describe the binding of a DNA probe, or the binding of a primer to a DNA strand during a polymerase chain reaction (PCR). The term is also used to describe the reformation (renaturation) of complementary nucleotide strands that were separated by heat (thermally denatured). The term "DNA denaturation" or "denaturation" as used herein refers to the process by which double-stranded deoxyribonucleic acid unwinds and separates into single- stranded DNA through the breaking of hydrogen bonding between the bases. Both terms are used to refer to the process as it occurs when a mixture is heated. DNA denaturation may also be referred to as DNA melting.

The term "reaction mixture" as used herein refers to the PCR reaction mixture comprising components used in the PCR reaction, such as for example

deoxyribonucleotides (dNTPs), cofactors, primers, DNA polymerase, DNA intercalating fluorescent dyes and/or RNA or DNA template at a suitable ionic and pH environment. The reaction mixture may also comprise components, such as for example double stranded DNA intercalating dye, used for melting analysis, such as for example high resolution melting (HRM) analysis.

The term "mutation of interest" as used herein refers to the mutation or the at least one mutation, which is to be detected by the method of the present invention.

The term "wild type DNA", "wild type sequence" or "wild type allele" as used herein refers to DNA sequences or DNA alleles not comprising the mutation to be detected or the mutation of interest. The term "wild type" may be used interchangeably with "non- mutant".

Method for detecting DNA or RNA mutations

The present invention relates to a method for detection of mutations and differences in nucleotide sequences. The principle of the method is to simultaneously amplify sequences comprising the mutation(s) to be detected and sequences not comprising the mutation(s) to be detected using a three primer system. The resulting amplicons are designed to melt differently to enable direct detection by melting analysis.

Thus, the present invention provides a method for detecting the absence or presence of at least one mutation in a nucleotide sequence in a sample, said method comprising the steps of d) providing at least three primers, wherein at least one of said primers is a mutant primer comprising a nucleotide sequence complementary to a region of a mutant nucleotide strand comprising the mutation to be detected and wherein at least one of said primers is a non-mutant primer comprising a sequence that is complementary to a region of a non-mutant nucleotide strand not comprising the mutation to be detected, and wherein at least one of said primers is a common primer comprising a sequence that binds to the nucleotide strand complementary to the nucleotide strands to which the mutant and non-mutant primer bind such that the extension product of said common primer comprises a region complementary to extension products of the mutant and non-mutant primers,

e) contacting the sample with an oligonucleotide system under hybridization

conditions so as to form a reaction mixture, said oligonucleotide system comprising the at least three primers,

f) initiate polymerase chain reaction (PCR) and run at least one cycle of PCR so that if the mutation to be detected is present it will result in simultaneous amplification of mutant and non-mutant nucleotide sequences, wherein the at least one mutant primer and the at least one common primer result in the synthesis of least one mutant amplicon comprising the mutation to be detected and having a melting temperature x, and wherein the at least one non-mutant primer and the at least one common primer result in the synthesis of at least one non-mutant amplicon not comprising the mutation to be detected and having a melting temperature y different from the temperature x and

detecting the presence or absence of said mutation by melting analysis.

The polymerase chain reaction (PCR) is initiated by adding a temperature resistant polymerase with appropriate substrates, deoxynbonucleotides (dNTPs) and cofactors, at a suitable ionic and pH environment, to the reaction mixture under predetermined reaction conditions (see below). The dNTPs include dATP, dCTP, dGTP and dTTP in adequate amounts to provide sufficient substrate for the synthesis of new DNA strands. Each of the four deoxynbonucleotides is typically added in equimolar amounts. The reaction mixture further comprises template and oligonucleotide primers and may further comprise a substance for detection of the mutation by melting analysis such as for example double stranded DNA intercalating fluorescent dye. Double stranded DNA intercalating fluorescent dye may for example include SYTO-9, SYBR Green,

LCGreen, or EvaGreen.

The PCR is commonly carried out in a reaction volume of 10-200 micro-liters (μΙ) in small reaction tubes (0.2-0.5 milliliters (ml) volumes) in a PCR machine called a thermal cycler. The thermal cycler heats and cools the reaction tubes to achieve the temperatures required at each step of the reaction.

Typically, PCR consists of a series of repeated temperature changes, called cycles, with each cycle commonly consisting of 3 temperature steps. The temperatures used and the length of time they are applied in each cycle depend on a variety of parameters. These include the polymerase used for DNA synthesis, the concentration of divalent ions and dNTPs in the reaction, the length of the amplicon, and the melting temperature (Tm) of the primers.

Before the cycling begins, the reaction mixture is typically heated to a temperature of 92 to 98 degress Celcius (°C) for 2 to 20 minutes. In one embodiment the reaction mixture is heated to at least 94°C for at least 10 minutes, such as at least 11 minutes, such as for example at least 12 minutes, such as at least 13 minutes, such as for example at least 14 minutes, such as at least 15 minutes, such as for example at least 16 minutes, such as at least 17 minutes, such as for example at least 18 minutes, such as at least 19 minutes, or such as for example at least 20 minutes. In one preferred embodiment the reaction mixture is heated to at least 94°C for at least 10 minutes. In another embodiment the reaction mixture is heated to at least 95°C for at least 10 minutes, such as at least 11 minutes, such as for example at least 12 minutes, such as at least 13 minutes, such as for example at least 14 minutes, such as at least 15 minutes, such as for example at least 16 minutes, such as at least 17 minutes, such as for example at least 18 minutes, such as at least 19 minutes, or such as for example at least 20 minutes. In a preferred embodiment the reaction mixture is heated to at least 95°C for at least 10 minutes. In a more preferred embodiment the reaction mixture is heated to a temperature of 95 °C for 15 minutes. This is for example preferred when using HotStar Taq polymerase from Qiagen. In a further embodiment the reaction mixture is heated to at least 96°C for at least 10 minutes, such as at least 11 minutes, such as for example at least 12 minutes, such as at least 13 minutes, such as for example at least 14 minutes, such as at least 15 minutes, such as for example at least 16 minutes, such as at least 17 minutes, such as for example at least 18 minutes, such as at least 19 minutes, or such as for example at least 20 minutes. In a preferred embodiment the reaction mixture is heated to at least 96°C for at least 10 minutes.

One PCR cycle may comprise the following steps:

Denaturation step: This step is the first regular cycling event and consists of heating the reaction mixture to a temperature that allows denaturation or melting of the double stranded template by disrupting the hydrogen bonds between complementary bases. The temperature for the denaturation process ranges from 90°C to 100°C and is allowed to proceed for 1 to 30 seconds. In a preferred embodiment the denaturation process is carried out at at least 95°C for at least 5 seconds, such as at least 6 seconds, or at least 7 seconds, such as at least 8 seconds, or at least 9 seconds, such as at least 10 seconds, or at least 11 seconds, such as at least 12 seconds, or at least 13 seconds, such as at least 14 seconds, or at least 15 seconds. In a more preferred embodiment the denaturation process is carried out at 95°C for 10 seconds.

Annealing step: The reaction temperature is lowered to 50 to 80°C for 1 to 40 seconds allowing annealing of the primers to the single-stranded DNA template. Stable DNA- DNA hydrogen bonds are only formed when the primer sequence very closely matches the template sequence. The annealing temperature depends on the primers used. A long primer results in a high melting temperature between primer and template, or a high primer Tm, which allows the use of a high annealing temperature. If a shorter primer is used, the melting temperature between primer and template is lower and consequently the annealing temperature must be lower to allow annealing of the primer to the template. In the present invention base-mismatches may be present in the mutant and non-mutant primers, which also results in a lower annealing temperature (see the "primer section" for primer melting temperatures). When the primer anneals to the template, the polymerase binds to the primer-template hybrid and begins DNA synthesis. In one embodiment the annealing temperature is at least 50°C, such as at least 51 °C, such as for example at least 52°C, such as at least 53°C, such as for example at least 54°C, such as at least 55°C, such as for example at least 56°C, such as at least 57°C, such as for example at least 58°C, such as at least 59°C, such as for example at least 60°C, such as at least 61 °C, such as for example at least 62°C, such as at least 63°C, such as for example at least 64°C, such as at least 65°C, such as for example at least 66°C, such as at least 67°C, such as for example at least 68°C, such as at least 69°C, such as for example at least 70°C, such as at least 71 °C, such as for example at least 72°C, such as at least 73°C, such as for example at least 74°C, such as at least 75°C, such as for example at least 76°C, such as at least 77°C, such as for example at least 78°C, such as at least 79°C, or such as for example at least 80°C.

In one preferred embodiment the annealing temperature is in the range of 65°C to 75°C. In another referred embodiment the annealing temperature is in the range of 55°C to 65°C. In one preferred embodiment the annealing temperature is about 70°C. In another referred embodiment the annealing temperature is about 60°C.

Extension/elongation step: The temperature at this step depends on the DNA polymerase used; Taq polymerase has its optimum activity temperature at 70 to 80 °C, and commonly a temperature of 72 °C is used with this DNA polymerase. At this step the DNA polymerase synthesizes a new DNA strand complementary to the DNA template strand by adding dNTPs that are complementary to the template in a 5' to 3' direction, condensing the 5'-phosphate group of the dNTPs with the 3'-hydroxyl group at the end of the nascent (extending) DNA strand. The extension time depends both on the DNA polymerase used and on the length of the DNA fragment to be amplified. At its optimum temperature, the DNA polymerase will polymerize about a thousand bases per minute. Under optimum conditions, i.e., if there are no limitations due to limiting substrates or reagents, at each extension step, the amount of DNA target is doubled, leading to exponential amplification of the specific DNA fragment.

When one PCR cycle has been completed, each primer pair annealing to a template has generated one extension product. At every cycle the amount of extension products is approximately doubled, resulting in exponential growth of extension products or amplicons. In one embodiment at least one PCR cycle is performed, such as at least two PCR cycles, such as for example at least three PCR cycles, such as at least 4 PCR cycles, such as for example at least 5 PCR cycles, such as at least 6 PCR cycles, such as for example at least 7 PCR cycles, such as at least 8 PCR cycles, such as for example at least 9 PCR cycles. In another embodiment at least 10 PCR cycles are performed, such as for example at least 11 PCR cycles, such as at least 12 PCR cycles, such as for example at least 13 PCR cycles, such as at least 14 PCR cycles, such as for example at least 15 PCR cycles such as least 16 PCR cycles, such as for example at least 17 PCR cycles, such as at least 18 PCR cycles, such as for example at least 19 PCR cycles, such as at least 20 PCR cycles, such as for example at least 21 PCR cycles, such as at least 22 PCR cycles, such as for example at least 23 PCR cycles, such as at least 24 PCR cycles, such as for example at least 25 PCR cycles, such as at least 26 PCR cycles, such as for example at least 27 PCR cycles, such as at least 28 PCR cycles, such as for example at least 29 PCR cycles, or such as at least 30 PCR cycles, such as for example at least 35 PCR cycles.

In one preferred embodiment at least 40 PCR cycles are performed, such as for example at least 50 PCR cycles, or such as at least 60 PCR cycles. In a more preferred embodiment at least 45 PCR cycles are performed

When all the PCR cycles are completed a single step is occasionally performed at a temperature of 70 to 74°C, preferably at 72°C, for 5 to 15 minutes to ensure that any remaining single-stranded DNA is fully extended. Finally a step at 4 to 15 °C for an indefinite time may be employed for short-term storage of the reaction. When immediately performing melting analysis after the PCR the step at 4 to 15°C is omitted and a denaturation step followed by a hybridization step of for instance between 4 to 80 °C depending on the melting temperature of the amplicon is performed. The present method can also be used to determine the concentration or the amount of mutant alleles present in a sample. The method can for example distinguish between the presence of 1 % mutant alleles, 5% mutant alleles or for example 10% mutant alleles. Primers

The primers as used herein are oligonucleotide primers. In a preferred embodiment the oligonucleotide primers is a DNA primer.

The length of the primers used in the method of the present invention may depend on many factors such as the function of the primer, the nature of the mutation to be detected, the annealing temperature and the melting temperature of the amplicons. The primers must be sufficiently long to prime synthesis of extension products in the presence of a polymerase. The length of the primers may typically vary from about 8 nucleotides to 60 nucleotides, such as for example 10 nucleotides to 55 nucleotides, such as 15 nucleotides to 50 nucleotides, such as for example 20 nucleotides to 45 nucleotides, such as 25 nucleotides to 40 nucleotides or such as for example 30 nucleotides to 35 nucleotides. In another embodiment the length of the primers vary from about 8 nucleotides to 30 nucleotides, such as for example 10 nucleotides to 25 nucleotides or such as 15 nucleotides to 20 nucleotides. In yet another embodiment the length of the primers vary from about 30 nucleotides to 60 nucleotides, such as for example 35 nucleotides to 55 nucleotides or such as 40 nucleotides to 50 nucleotides. In a preferred embodiment the length of the primers are from 15 nucleotides to 30 nucleotides.

It is preferred that the at least one mutant primer and the at least one non-mutant primer are competitive primers comprising a nucleotide sequence that results in competitive binding of the mutant and non-mutant primers to the same region of a nucleotide strand or to the same nucleotide sequence.

Thus, in one preferred embodiment the present invention relates to a method for detecting the absence or presence of at least one mutation in a nucleotide sequence in a sample, said method comprising the steps of

g) providing at least two competitive primers and at least one common primer, wherein the sequence identity between said competitive primers results in competitive binding to the same nucleotide sequence and wherein at least one of said competitive primers is a mutant primer comprising a nucleotide sequence complementary to a region of mutant nucleotide strand comprising the mutation to be detected and wherein at least one of said competitive oligonucleotide primers is a non-mutant primer comprising a sequence that is complementary to a region of a non-mutant nucleotide strand not comprising the mutation to be detected, and wherein said common primer comprises a sequence that binds to the nucleotide strand complementary to the nucleotide strands to which the competitive primers bind such that the extension product of said common primer comprises a region complementary to extension products of the mutant and non-mutant primers,

h) contacting the sample with an oligonucleotide system under hybridization

conditions so as to form a reaction mixture, said oligonucleotide system including the at least two competitive primers and the at least one common primer,

i) initiate polymerase chain reaction (PCR) and run at least one cycle of PCR so that if the mutation to be detected is present it will result in simultaneous amplification of mutant and non-mutant nucleotide sequences, wherein the at least one mutant primer and the at least one common primer result in the synthesis of least one mutant amplicon comprising the mutation to be detected and having a melting temperature x, and wherein the at least one wild type primer and the at least one common primer result in the synthesis of at least one non-mutant amplicon not comprising the mutation to be detected and having a melting temperature y different from the temperature x and

detecting the presence or absence of said mutation by melting analysis

The term "competitive oligonucleotide primers" or "competitive primers" as used herein refers to those primers, which differ by at least one base or nucleotide and wherein the sequence difference between said competitive primers results in a differential rate and ability to bind to the same nucleotide sequence and results in a competitive binding to said nucleotide sequence. Thus, the competitive primers bind differentially but competitively to the same nucleotide sequence. The difference in the sequence between the competitive primers must however allow competitive binding to the same nucleotide sequence, i.e. the sequence identity between the competitive primers must be high enough to allow competitive binding to the same nucleotide sequence. A variety of conditions including annealing temperature, ionic strength, the chemical composition of the reaction mixture and relative concentrations between the primers will influence on the ability of the primer to bind the template. When competitive oligonucleotide primers are incubated with a nucleotide template under appropriate conditions such as similar primer concentrations, the oligonucleotide primer, which most nearly matches the known sequence to be hybridized, will bind preferentially over the primer, which has a base mismatch or the most base mismatches.

It is preferred that the competitive primers bind competitively to the same nucleotide sequence during the annealing step in a PCR cycle as defined elsewhere herein. The competitive primers may consist of at least one mutant primer and at least one non- mutant primer.

In one preferred embodiment the mutant primer and the non-mutant primer compete for binding to the non-mutant DNA strand. The presence of the non-mutant primer in the reaction mixture will prevent non-specific binding of the mutant primer to the non- mutant DNA strand, thereby preventing the generation of false positive results. False positive results may arise if the mutant primer anneals to the non-mutant DNA strand and starts amplification. This will result in an amplicon comprising the mutation to be detected although the template used is the non-mutant DNA strand not comprising the mutation.

The dissociation constant between non-mutant DNA and non-mutant primer is defined as

Kd, non-mutant- [Pnon-mutant][T]/[Cnon-mutant]. where [P, non-mutant] is the concentration of non-mutant primer in the reaction mixture, [T] is the concentration of non-mutant template and [C _n0 n-mutant] is the concentration of primer-template complex, i.e. the concentration of non-mutant primer that has bound to the non-mutant template.

Similarly, the dissociation constant between non-mutant DNA and mutant primer is defined as

Kd, mutant- [Pmutant][T]/[Cmutant]. where [P _mu tant] is the concentration of mutant primer in the reaction mixture, [T] is the concentration of non-mutant template and [C _-mu tant] is the concentration of primer- template complex, i.e. the concentration of mutant primer that has bound to the non- mutant template. Thus, if for example equal concentrations of mutant and non-mutant primers are present in the reaction mixture and K _d , _n on-mutant= K _d , _mut ant the concentration of mutant primer that has bound to the non-mutant template is equivalent to the concentration of non-mutant primer that has bound to the non-mutant template. In one preferred embodiment, K _d , _n0 n-mutant < K _d , _mut ant, i.e. the non-mutant primer has a higher binding affinity for the non-mutant DNA template than the mutant primer. Thus, if K _d, non- mutant < _d, mutant, the non-mutant primer binds preferentially over the mutant primer to the non-mutant template.

In one embodiment of the present invention K _{d, m} utant/ _d, non-mutant is equivalent to at least 1 ,1 such as for example at least 1 ,2 or such as at least 1 ,3 such as for example at least 1 ,4 or such as at least 1 ,5 such as for example at least 1 ,6 or such as at least 1 ,7 such as for example at least 1 ,8 or such as at least 1 ,9 such as for example at least 2 or such as at least 2,5 such as for example at least 3 or such as at least 3,5 such as for example at least 4 or such as at least 4,5 such as for example at least 5 or such as at least 6 such as for example at least 8 or such as at least 10 such as for example at least 20, such as at least 30 such as for example at least 40 or such as at least 50.

In one preferred embodiment the non-mutant primer is designed to bind equally to the mutant and non-mutant DNA strand such that if the non-mutant primer anneals to the mutant DNA the resulting extension product will comprise the mutation to be detected. Thereby the amount of non-mutant template is increased, which also prevents false amplification from the mutant primer as it will bind preferentially to the mutant DNA sequences. The term "mutant oligonucleotide primer" or "mutant primer" as used herein refers to a primer comprising a nucleotide sequence complementary to a region of a mutant nucleotide strand comprising the mutation to be detected. The term "mutant primer" may be used interchangeably with "mutation specific primer". It is preferred that the 3' end of the mutant primer comprises the sequence complementary to the mutation to be detected. DNA synthesis from the mutant primer results in a mutant extension product comprising the mutation to be detected. Annealing of the mutant extension product from the mutant primer with the mutant extension product from the common primer results in a mutant amplicon comprising the mutation to be detected and having a melting temperature x. The primer may in an embodiment be used to determine the melting temperature of the amplicons by varying the length of the amplicon and by introducing mutations.

The term "non-mutant oligonucleotide primer" or "non-mutant primer" as used herein refers to a primer comprising a nucleotide sequence complementary to a region of a non-mutant nucleotide strand not comprising the mutation to be detected. The term "non-mutant primer" may be used interchangeably with "wild type primer". DNA synthesis from the non-mutant primer results in a non-mutant extension product not comprising the mutation to be detected. Annealing of the non-mutant extension product from the non-mutant primer with the non-mutant extension product from the common primer results in a non-mutant amplicon not comprising the mutation to be detected and having a melting temperature y. The melting temperature y is different from the melting temperature x.

In one preferred embodiment of the present invention at least two competitive primers and at least one common primer is used. Thus, in one preferred embodiment at least three primers are used. The three primers are used for the detection of at least one mutation. If two mutations, which are located very close on the nucleotide strand, are to be detected, the mutant primer may be designed to anneal to a sequence comprising both mutations. Thus, if the distance between two mutations to be detected is no more than for example 20 nucleotides the mutant primer may comprise a sequence that is complementary to a region of the mutant nucleotide strand comprising both mutations. If the distance between two mutations to be detected is more than for example 20 nucleotides it may be necessary to use two mutant primers each comprising a sequence that is complementary to a region of the mutant nucleotide strand comprising a mutation to be detected.

Thus, if more than one mutation is to be detected it may be necessary to use at least two mutant primers each comprising a sequence that is complementary to a region of the mutant nucleotide strand comprising a mutation to be detected, at least two non- mutant primers and at least two common primers, i.e. at least 6 primers. Thus, in one embodiment at least 2 mutant primers and at least 2 non-mutant primers are used, such as at least 3 mutant primers and at least 3 non-mutant primers, at least 4 mutant primers and at least 4 non-mutant primers or such as at least 5 mutant primers and at least 5 non-mutant primers. Similarly, in one embodiment at least 2 common primers are used, such as at least 3 common primers, such as at least 4 common primers or such as least 5 common primers.

In the method of the present invention, the mutant primer and the non-mutant primer result in the synthesis of two amplicons having different melting temperatures. The mutant primer results in the synthesis of a mutant amplicon having a melting temperature x, whereas the non-mutant primer results in the synthesis of a non-mutant amplicon having a melting temperature y. The melting temperature x is different from the melting temperature y. It is the different melting temperatures of the mutant and non-mutant amplicons that enable direct identification of mutations by melting analysis. A high difference between the melting temperatures x and y facilitates the separation and identification of the mutant and non-mutant amplicons by melting analysis such as for example high resolution melting (HRM) analysis. It is preferred that the difference between the melting temperatures x and y is at least 0,2 °C to allow easy detection by melting analysis.

The melting temperatures x and y of the mutant and non-mutant amplicons are typically in the range of 65 °C to 95 °C.

The difference between the melting temperatures x and y may be in the range of 0, 1 °C to 50 °C. In one embodiment the difference between the melting temperatures x and y is in the range of 5 °C to 20 °C, such as for example 7 °C to 15 °C or such as for example 10 °C to 12 °C. In another embodiment the difference between the melting temperatures x and y is in the range of 20 °C to 50 °C, such as for example 25 °C to 45 °C or such as for example 30 °C to 40 °C.

In one embodiment difference between the melting temperatures x and y is in the range of 0.1 °C to 3 °C, such as for example 0,5 °C to 2 °C. In one preferred

embodiment the difference between the melting temperatures x and y is in the range of 1 °C to 5 °C. In another preferred embodiment the difference between the melting temperatures x and y is in the range of 1 °C to 4 °C. In a further preferred embodiment the difference between the melting temperatures x and y is in the range of 1 °C to 3 °C. In another preferred embodiment the difference between the melting temperatures x and y is in the range of 2 °C to 4 °C. In one embodiment of the present invention the melting temperature x of the mutant amplicon is higher than the melting temperature y of the non-mutant amplicon.

The melting temperature x may be in the range of 0.1 °C to 50 °C higher than the melting temperature y. In one embodiment the melting temperature x is in the range of 5 °C to 20 °C higher than the melting temperature y, such as for example 7 °C to 15 °C or such as for example 10 °C to 12 °C higher than the melting temperature y. In another embodiment the melting temperature x is in the range of 20 °C to 50 °C higher than the melting temperature y, such as for example 25 °C to 45 °C or such as for example 30 °C to 40 °C higher than the melting temperature y.

In one embodiment the melting temperature x is in the range of 0, 1 °C to 3 °C higher than the melting temperature y, such as for example 0,5 °C to 2 °C higher than the melting temperature y. In one preferred embodiment the melting temperature x is in the range of 1 °C to 5 °C higher than the melting temperature y. In another preferred embodiment the melting temperature x is in the range of 1 °C to 4 °C higher than the melting temperature y. In a further preferred embodiment the melting temperature x is in the range of 1 °C to 3 °C higher than the melting temperature y. In another preferred embodiment the melting temperature x is in the range of 2 °C to 4 °C higher than the melting temperature y.

In another embodiment the melting temperature x of the mutant amplicon is lower that the melting temperature y of the non-mutant amplicon.

The melting temperature x may be in the range of 0,1 °C to 50 °C lower than the melting temperature y. In one embodiment the melting temperature x is in the range of 5 °C to 20 °C lower than the melting temperature y, such as for example 7 °C to 15 °C or such as for example 10 °C to 12 °C lower than the melting temperature y. In another embodiment the melting temperature x is in the range of 20 °C to 50 °C lower than the melting temperature y, such as for example 25 °C to 45 °C or such as for example 30 °C to 40 °C lower than the melting temperature y.

In one embodiment the melting temperature x is in the range of 0, 1 °C to 3 °C lower than the melting temperature y, such as for example 0,5 °C to 2 °C lower than the melting temperature y. In one preferred embodiment the melting temperature x is in the range of 1 °C to 5 °C lower than the melting temperature y. In another preferred embodiment the melting temperature x is in the range of 1 °C to 4 °C lower than the melting temperature y. In a further preferred embodiment the melting temperature x is in the range of 1 °C to 3 °C lower than the melting temperature y. In another preferred embodiment the melting temperature x is in the range of 2 °C to 4 °C lower than the melting temperature y. ₀

o

The difference in the melting temperatures between the mutant and non-mutant amplicon can be obtained by designing amplicons of different length. This can be achieved by designing the mutant and non-mutant primers such that they anneal at different regions on the nucleotide strand and/or by designing mutant and non-mutant primers of different length.

In one embodiment the mutant amplicon is shorter that the non-mutant amplicon. The mutant amplicon may be at least one nucleotide shorter that the non-mutant amplicon, such as for example at least 2 nucleotides, at least 3 nucleotides, such as at least 4 nucleotide, such as for example at least 5 nucleotides, at least 6 nucleotides, such as at least 8 nucleotide, such as for example at least 10 nucleotides, at least 12 nucleotides, such as at least 15 nucleotide or such as at least 20 nucleotides shorter that the non-mutant amplicon.

In another embodiment the mutant amplicon is longer than the non-mutant amplicon. The mutant amplicon may be at least one nucleotide longer that the non-mutant amplicon, such as for example at least 2 nucleotides, at least 3 nucleotides, such as at least 4 nucleotide, such as for example at least 5 nucleotides, at least 6 nucleotides, such as at least 8 nucleotide, such as for example at least 10 nucleotides, at least 12 nucleotides, such as at least 15 nucleotide or such as at least 20 nucleotides longer that the non-mutant amplicon.

A difference in the melting temperatures between the mutant and non-mutant amplicon may also be achieved by introducing one or more point-mutations in the mutant or non- mutant amplicon that result in a decrease or increase in the melting temperature. In Watson-Crick DNA base pairing, adenine (A) forms a base pair with thymine (T) and guanine (G) forms a base pair with cytosine (C). The two nucleotides are connected via hydrogen bonds. A G-C base pair is connected via three hydrogen bonds, whereas an A-T base pair is connected via two hydrogen bonds. Consequently, a double stranded DNA molecule or an amplicon with a high G-C content or a low A-T content has a higher melting temperature that an amplicon with a low G-C content or a high A-T content. The mutation can be introduced in the amplicon by designing a non-mutant primer and/or a mutant primer comprising one or more base-mismatches, such that when the primer anneals to the nucleotide strand an abnormal nucleotide base-pairing is formed. If for example the primer comprises an adenine (A) instead of a cytosine (C), wherein the cytosine would result in normal base-pairing, then an abnormal base-pairing between adenine (A) and guanine (G) will be formed. This will lead to the synthesis of an amplicon product comprising an A-T base pair instead of a C-G base pair. The change of C-G with A-T results in a decrease in the melting temperature of the amplicon.

The mutant primer may introduce at least one, such as at least 2, such as for example at least 3, such as at least 4, such as for example at least 5, such as at least 6 or at least 8 point mutations that results in a decrease or in an increase in the melting temperature of the resulting mutant amplicon.

The mutant primer may in one preferred embodiment introduce at least one point mutation that results in a decrease in the melting temperature of the mutant amplicon. Thus, in one embodiment the mutant primer introduces at least one G>A mutation, at least one C>A mutation, at least one G>T mutation and/or at least one C>T mutation.

The mutant primer may in another preferred embodiment introduce at least one mutation that results in an increase in the melting temperature of the mutant amplicon. Thus, in one embodiment the mutant primer introduces at least one A>G mutation, at least one A>C mutation, at least one T>G mutation and/or at least one T>C mutation.

The non-mutant primer may introduce at least one, such as at least 2, such as for example at least 3, such as at least 4, such as for example at least 5, such as at least 6 or at least 8 point mutations that results in a decrease or in an increase in the melting temperature of the resulting non-mutant amplicon.

The non-mutant primer may in one embodiment introduce at least one mutation that results in a decrease in the melting temperature of the non-mutant amplicon. Thus, in one embodiment the non-mutant primer introduces at least one G>A mutation, at least one C>A mutation, at least one G>T mutation and/or at least one C>T mutation. In another embodiment the non-mutant primer introduces at least one mutation that results in an increase in the melting temperature of the non-mutant amplicon. Thus, in one embodiment the non-mutant primer introduces at least one A>G mutation, at least one A>C mutation, at least one T>G mutation and/or at least one T>C mutation.

In one embodiment of the present invention the sequence of the non-mutant primer is complementary to a region of the non-mutant nucleotide strand that is identical to the mutant nucleotide strand. Thus, in one embodiment the non-mutant primer binds the non-mutant nucleotide strand and the mutant nucleotide strand equally.

Annealing of the non-mutant primer to the mutant nucleotide strand and at least one cycle of PCR results in a third amplicon comprising the mutation to be detected and having a melting temperature z different from the melting temperature x. The melting temperature z of the third amplicon is typically in the range of 70 °C to 95 °C as described above for the melting temperatures x and y.

The mutation, which is present in the mutant strand, may result in a change in the melting temperature z of the third amplicon when compared to melting temperature y of the non-mutant amplicon. Thus, the melting temperature z of the third amplicon may in one embodiment be different from the melting temperature y of the non-mutant amplicon.

If the mutation, which is present in the mutant strand, does not result in a change in the melting temperature z of the third amplicon when compared to melting temperature y of the non-mutant amplicon, then the melting temperature z is equal to the melting temperature y. Thus, in another embodiment the melting temperature z of the third amplicon is equal to the melting temperature y of the non-mutant amplicon.

In one embodiment the mutant primer comprises a sequence complementary to said third amplicon such that annealing of the mutant primer and the common primer to the third amplicon followed by at least one cycle of PCR results in the synthesis of a mutant amplicon having the melting temperature x.

In another embodiment the sequence of the non-mutant primer is complementary to a region of the non-mutant nucleotide strand that is not identical to the mutant nucleotide strand, such that the non-mutant primer has a preference for the non-mutant nucleotide strand.

Heteroduplex formation

If non-mutant and mutant extension products anneal, this will lead to the formation of heteroduplexes, which have a lower melting temperature than the corresponding homo-duplexes due to one or more base mismatches. Heteroduplex formations between non-mutant and mutant extension products can be identified as an early melting peak in the melting curve (see Fig. 2). If heteroduplexes are identified in samples known to comprise only non-mutant sequences this implies that amplification from the mutant primer has occurred, in spite of the mismatches between the mutant primer and non-mutant sequences. This may give rise to false positive results, which can be avoided by increasing the annealing temperature, optimizing relative primer concentrations, or designing new primers.

Primer concentrations

The primer concentration of each individual primer, i.e. the concentration of the common primer, the mutant primer or the non-mutant primer may typically be in the range of 20 mM to 500 mM.

In one embodiment the concentration of mutant primer in the reaction mixture is equal to the concentration of non-mutant primer. In another embodiment the concentration of mutant primer in the reaction mixture is lower than the concentration of non-mutant primer.

The presence of non-mutant primer prevents the amplification of false positive results false amplification from mutant primer which may lead to false positive results. False positive results arise if the mutant primer anneals to the non-mutant DNA strand and starts amplification. Thus, in one preferred embodiment the concentration of mutant primer in the reaction mixture is lower than the concentration of non-mutant primer.

The concentration of mutant primer in the reaction mixture may in one embodiment be in the range of 20 mM to 100 mM, such as for example 20 mM to 80 mM, such as 20 mM to 60 mM, such as for example 20 mM to 50 mM, such as 20 mM to 40 mM, such as for example 20 mM to 30 mM or such as 40 mM to 100 mM, such as for example 60 mM to 100 mM, such as 70 mM to 100 mM, such as for example 80 mM to 100 mM or such as 90 mM to 100 mM.

In another embodiment the concentration of mutant primer in the reaction mixture is in the range of 100 mM to 500 mM, such as for example 100 mM to 400 mM, such as 100 mM to 300 mM or such as for example 200 mM to 500 mM or such as 300 mM to 500 mM or such as for example 400 mM to 500 mM.

In one preferred embodiment the concentration of mutant primer in the reaction mixture is in the range of 100 mM to 200 mM. In another preferred embodiment the

concentration of mutant primer in the reaction mixture is about 100 mM. In a further preferred embodiment the concentration of mutant primer in the reaction mixture is about 150 mM. In yet another preferred embodiment the concentration of mutant primer in the reaction mixture is about 200 mM.

The concentration of non-mutant primer in the reaction mixture may in one embodiment be in the range of 20 mM to 100 mM, such as for example 20 mM to 80 mM, such as 20 mM to 60 mM, such as for example 20 mM to 50 mM, such as 20 mM to 40 mM, such as for example 20 mM to 30 mM or such as 40 mM to 100 mM, such as for example 60 mM to 100 mM, such as 70 mM to 100 mM, such as for example 80 mM to 100 mM or such as 90 mM to 100 mM.

In another embodiment the concentration of non-mutant primer in the reaction mixture is in the range of 50 mM to 500 mM, such as for example 50 mM to 400 mM, such as 50 mM to 300 mM, or such as 70 mM to 500 mM, such as for example 80 mM to 500 mM, or such as 100 mM to 500 mM, such as for example 150 mM to 500 mM, such as 200 mM to 500 mM, such as for example 300 mM to 500 mM, or such as 400 mM to 500 mM. In one preferred embodiment the concentration of non-mutant primer in the reaction mixture is in the range of 50 mM to 150 mM. In another preferred embodiment the concentration of non-mutant primer in the reaction mixture is in the range of 50 mM to 100 mM. In yet another preferred embodiment the concentration of mutant primer in the reaction mixture is about 50 mM. In a further preferred embodiment the concentration of mutant primer in the reaction mixture is about 100 mM. In yet another preferred embodiment the concentration of mutant primer in the reaction mixture is about 150 mM.

The concentration of common primer in the reaction mixture may in one embodiment be in the range of 20 mM to 100 mM, such as for example 20 mM to 80 mM, such as 20 mM to 60 mM, such as for example 20 mM to 50 mM, such as 20 mM to 40 mM, such as for example 20 mM to 30 mM or such as 40 mM to 100 mM, such as for example 60 mM to 100 mM, such as 70 mM to 100 mM, such as for example 80 mM to 100 mM or such as 90 mM to 100 mM.

In another embodiment the concentration of common primer in the reaction mixture is in the range of 100 mM to 500 mM, such as for example 100 mM to 400 mM, such as 100 mM to 200 mM or such as for example 200 mM to 500 mM or such as 300 mM to 500 mM or such as for example 400 mM to 500 mM.

In a preferred embodiment the concentration of common primer in the reaction mixture is in the range of 100 mM to 300 mM, such as for example 150 mM to 250 mM. In another preferred embodiment the concentration of common primer in the reaction mixture is about 200 mM.

In a preferred embodiment the concentration of mutant primer in the reaction mixture is in the range of 100 mM to 200 mM, the concentration of non-mutant primer in the reaction mixture is in the range of 50 mM to 150 mM and the concentration of common primer in the reaction mixture is in the range of 100 mM to 300 mM.

In one particular preferred embodiment the concentration of mutant primer in the reaction mixture is about 200 nM, the concentration of non-mutant primer in the reaction mixture is about 100 nM and the concentration of common primer in the reaction mixture is about 200 nM.

In another particular preferred embodiment the concentration of mutant primer in the reaction mixture is about 150 nM, the concentration of non-mutant primer in the reaction mixture is about 100 nM and the concentration of common primer in the reaction mixture is about 200 nM. In yet another particular preferred embodiment the concentration of mutant primer in the reaction mixture is about 200 nM, the concentration of non-mutant primer in the reaction mixture is about 50 nM and the concentration of common primer in the reaction mixture is about 200 nM.

The primer melting temperature is the temperature at which the primer dissociates from the nucleotide strand or the template. The primer melting temperature depends on the length of the primer, the G-C and A-T content. A long primer and a high G-C content result in a high melting temperature. The primer melting temperature also depends on the number of mispaired bases between primer and template. In the present invention the mutant primer and/or the non-mutant primer may be designed to contain one or more base-mismatches that result in base-mispairing between primer and template. Such base-mismatches result in a decrease in the primer melting temperature. During the annealing step of a PCR cycle it is important that the annealing temperature is lower than the primer melting temperature to allow annealing of the primer to the template. However, if the annealing temperature is too low this may result in unspecific binding of the primer to the template.

In the present invention the primer melting temperature is typically in the range of 50 °C to 75 °C or more preferably in the range of 50 °C to 72 °C.

Sensitivity

The present invention provides a highly sensitive method for determining the presence or absence of mutations in a DNA sample. The use of a three primer PCR method as described herein, wherein the mutant primer and the non-mutant primer may bind competitively to the same nucleotide sequence, combined with melting analysis such as for example HRM analysis provides a highly sensitive method for detection of known mutations.

The sensitivity of the method refers to the minimum fraction or percentage of mutant alleles, which can be detected in a sample. If for example the sensitivity of the method is 0.05 %, the method is sensitive to 0.05 % mutant alleles in a background of wild-type alleles, i.e. the method can detect a mutation, which is present in at least 0.05 % of the alleles in the sample. Methods based on standard PCR and subsequent assays for mutation detection such as traditional Sanger sequencing suffer from a relatively low sensitivity as mutant alleles must be present in a proportion of at least 10-20% to be reproducibly detected. Standard PCR followed by pyrosequencing or high-resolution melting (HRM) is usually more sensitive, the sensitivity being 5-10%. However, the sensitivity limit of standard PCR followed by HRM is not sufficient to identify for example all metastatic colorectal cancer (CRC) patients that carry somatic KRAS mutations. The present invention provides a method for detecting the presence or absence of a known mutation, wherein said method has a sensitivity which is lower than 5%, such as for example lower than 4%, such as lower than 3%, such as for example lower than 2%, such as lower than 1 % or such as for example lower than 0.5%. In one embodiment of the present invention the method is sensitive to at least 1 % mutant alleles. In another preferred embodiment the method is sensitive to at least 0.5%.

In one preferred embodiment the method according to the present invention is sensitive to at least 0.25% mutant alleles. In another preferred embodiment the method according to the present invention is sensitive to at least 0.05% mutant alleles. In yet another preferred embodiment the method according to the present invention is sensitive to at least 0.025% mutant alleles or such as at least 0.01 % mutant alleles The sensitivity is dependent on the DNA concentration ro the amount of DNA used in the assay. In the Example section 50 ng DNA is for example used. However, increased concentration of DNA may increase the sensitivity of the assay.

Mutations

The present invention refers to a method for detecting the absence or the presence of a mutation in a nucleotide sequence. The term "nucleotide sequence" as used herein refers to a nucleotide strand or a polynucleotide. The nucleotide sequence may be an RNA molecule or a DNA molecule and the nucleotide sequence may be either single stranded or double stranded. In one preferred embodiment the nucleotide sequence is a DNA molecule. In another preferred embodiment the nucleotide sequence is a double stranded DNA molecule.

Thus, in one embodiment the at least one mutation is an RNA mutation. In another preferred embodiment the at least one mutation is a DNA mutation. It is appreciated that the mutation to be detected is a known mutation present in a known nucleotide sequence.

The method of the present invention may also be used for detecting the absence or the presence of mutated alleles. Thus, in one preferred embodiment the mutation to be detected is present in genomic DNA.

The mutation to be detected may include all mutation types found in nucleotide sequences.

In one embodiment the at least one mutation is a point mutation. In a point mutation a single nucleotide is exchanged for another. The point mutation may be an A>G mutation, an A>C mutation, an A>T mutation, a T>G mutation, a T>C mutation, a T>A mutation, a G>T mutation, a G>C mutation, n G>A mutation, a C>G mutation, a C>T mutation or a C>A mutation. The point mutation may for example be a silent mutation that code for the same amino acid, a missence mutation that code for another amino acid or non-sense mutation that code for a stop codon and can result in truncation of the protein. The method of the invention may be used for the detection of one or more point mutations or a combination of one or more of the point mutations as listed above.

In another embodiment the at least one mutation is a deletion of one or more nucleotides, such as for example at least 1 nucleotide, such as at least 2 nucleotides, at least 3 nucleotides, such as for example at least 4 nucleotides, such as at least 5 nucleotides, at least 6 nucleotides, such as for example at least 7 nucleotides, such as at least 8 nucleotides, at least 9 nucleotides, such as for example at least 10 nucleotides, such as at least 1 1 nucleotides, at least 12 nucleotides, such as for example at least 13 nucleotides, such as at least 14 nucleotides, at least 15 nucleotides or such as for example at least 20 nucleotides or even more. In a further embodiment the at least one mutation is an insertion of one or more nucleotides, such as for example at least 1 nucleotide, such as at least 2 nucleotides, at least 3 nucleotides, such as for example at least 4 nucleotides, such as at least 5 nucleotides, at least 6 nucleotides, such as for example at least 7 nucleotides, such as at least 8 nucleotides, at least 9 nucleotides, such as for example at least 10 nucleotides, such as at least 1 1 nucleotides, at least 12 nucleotides, such as for example at least 13 nucleotides, such as at least 14 nucleotides, at least 15

nucleotides or such as for example at least 20 nucleotides or even more. The method of the present invention may also be used for detecting the absence or the presence of mutated alleles as exemplified in the following section, where the method is used for detection of mutations in the human KRAS, BRAF and EGFR genes.

Thus, in one embodiment the method of the present invention also relates to determining the presence or absence of at least one mutation in the KRAS gene (SEQ ID NO: 1), the BRAF gene (SEQ ID NO:2) and/or the EGFR gene (SEQ ID NO: 3).

Activating mutations in KRAS and BRAF are found in approximately 40-50% and 10- 15% of all CRC patients, respectively and are found to be mutually exclusive. The BRAF gene encodes the serine/threonine-protein kinase B-Raf. Acquired mutations in this gene have also been found in cancers, including non-Hodgkin lymphoma, colorectal cancer, malignant melanoma, papillary thyroid carcinoma, non-small cell lung carcinoma, and adenocarcinoma of lung.

The human KRAS gene encodes a GTPase that performs an essential function in normal tissue signaling, and mutation of the KRAS gene is an essential step in the development of many cancers. A single amino acid substitution, and in particular a single nucleotide substitution can be responsible for an "activating" mutation that results in a mutated protein which is implicated in various malignancies, including lung adenocarcinoma, mucinous adenoma, ductal carcinoma of the pancreas and colorectal carcinoma. KRAS mutations are predictive of a very poor response to EGFR-inhibiting drugs such as the anti-EGFR monoclonal antibodies panitumumab (Vectibix) and cetuximab (Erbitux) used in the treatment of colorectal cancer. Currently, the most reliable way to predict whether a colorectal cancer patient will respond to one of the EGFR-inhibiting drugs is to test for certain activating mutations in the gene that encodes KRAS. Mutations that lead to EGFR overexpression (known as upregulation) or overactivity have been associated with a number of cancers, including lung cancer, anal cancers and glioblastoma multiforme. The epidermal growth factor Receptor (EGFR) is often overexpressed in CRC and non-small cell lung cancer (NSCLC), and contributes to cancer development and progression by stimulating proliferation, angiogenesis, invasion, and survival of cancer cells. A subset of NSCLC patients carrying activating somatic mutations in the tyrosine kinase domain of EGFR show excellent response to EGFR tyrosine kinase inhibitors such as gefitinib and erlotinib.

The human PIK3CA gene encoding phosphoinositide-3-kinase has also been shown to be mutated in diverse human cancers such as for example colorectal cancer.

Thus, in one embodiment of the present invention the at least one mutation is in the human KRAS gene (SEQ ID NO: 1). The at least one mutation to be detected may be any kind of mutation such as for example a point mutation, a deletion mutation and/or an insertion mutation. In one specific embodiment the at least one mutation is the human KRAS c.35G>C mutation (base pair no. 5571 of SEQ ID NO: 1). In another embodiment of the present invention the at least one mutation is in the human EGFR gene (SEQ ID NO: 3). The at least one mutation to be detected may be any kind of mutation such as for example a point mutation, a deletion mutation and/or an insertion mutation. In one specific embodiment the at least one mutation is the human EGFR c.2573T>G mutation (base pair no. 172791 of SEQ ID NO: 3).

In a further embodiment of the present invention the at least one mutation is in the human BRAF gene (SEQ ID NO: 2). The at least one mutation to be detected may be any kind of mutation such as for example a point mutation, a deletion mutation and/or an insertion mutation. In one specific embodiment the at least one mutation is the human BRAF c.1799T>A mutation (base pair no. 171429 of SEQ ID NO: 2).

In yet another embodiment of the present invention the at least one mutation to be detected is in the human PIK3CA gene (SEQ ID NO: 16). The at least one mutation to be detected may be any kind of mutation such as for example a point mutation, a deletion mutation and/or an insertion mutation. In one specific embodiment the at least one mutation is the human PIK3CA c.3140A>G mutation (base pair no. 86775 of SEQ ID NO: 16).

Determining the presence or the absence of a mutation

When the PCR reaction has been completed the presence or the absence of the at least one mutation must be determined by melting analysis. Melting analysis refer to methods, wherein the presence or the absence of mutations are determined based on temperature differences between the amplicons. If the mutation of interest is not present in the sample only the non-mutant amplicon having a melting temperature y will be generated. If the mutation of interest is present in the sample this may result in the generation of a non-mutant amplicon having a melting temperature x and a non-mutant amplicon having a melting temperature. The presence of two or more amplicons having different melting temperature can be determined by melting analysis.

Melting analysis can be performed using Temperature Gradient Gel Electrophoresis (TGGE) or Denaturing Gradient Gel Electrophoresis (DGGE). DNA is a negatively charged molecule and in the presence of an electric field, will move towards the positive electrode in a gel. The DNA moves through the gel with a speed that is roughly proportional to the length of the DNA molecule. When the temperature is raised, the two strands of the DNA start to separate, that is the DNA denatures or melts. At a specific temperature that depends on the melting temperature of the DNA molecule, the two DNA strands will be partly separated, and consequently the movement of the DNA through the gel decreases drastically when these partially melted structures are formed. The exact temperature at which this occurs depends on the sequence of the DNA. A high G-C content results in a high melting temperature, whereas a low G-C content results in a low melting temperature.

A similar method for detecting the mutation is Single-strand conformation

polymorphism (SSCP) analysis. Single-strand conformation polymorphism (SSCP) is defined as conformational difference of single-stranded nucleotide sequences of identical length as induced by differences in the sequences under certain experimental conditions. This property allows distinguishing the sequences by gel electrophoresis, which separates the different conformations.

Another method that can be used for detecting the mutation is Denaturing high pressure liquid chromatography (DHPLC), which uses heteroduplex formation between wild-type and mutated DNA strands to identify mutations. Heteroduplex molecules are separated from homoduplex molecules by ion-pair, reverse-phase liquid

chromatography on a special column matrix with partial heat denaturation of the DNA strands.

In a preferred embodiment of the present invention the melting analysis is high resolution melting analysis. High Resolution Melting (HRM) analysis is performed on double stranded DNA samples. In the present invention PCR is used prior to HRM analysis to amplify the nucleotide region in which the mutation of interest lies.

High Resolution Melting (HRM) analysis is a post-PCR analysis method used to identify variations in nucleic acid sequences. The method is based on detecting small differences in the melting temperature of PCR generated amplicons. The process is simply a precise warming of the sample or the reaction mixture from around 50°C up to around 95°C. At some point during this process, the melting temperature of the amplicon is reached and the two strands of DNA separate, i.e. the amplicon denatures into single stranded DNA.

In HRM analysis a fluorescent dye is used that allows to monitor the process in realtime. The dyes that are used for HRM are known as intercalating dyes that bind specifically to double-stranded DNA and when they are bound they fluoresce brightly. When the double stranded DNA denatures into single stranded DNA the intercalating dyes dissociates from the DNA. When the intercalating dyes are not bound to DNA they only fluoresce at a low level. At the beginning of the HRM analysis there is a high level of fluorescence in the sample because of the double stranded amplicons. But as the temperature increases the two strands of the amplicon melt apart, the

concentration of double stranded DNA decreases and the fluorescence is consequently reduced. The HRM machine has a camera that measures the fluorescence and the machine then plots the data as a graph known as a melt curve, showing the level of fluorescence versus temperature. When for example two amplicons with different melting temperatures are present in the sample this will give rise to two different melting curves.

The HRM analysis may be performed by increasing the temperature of the sample from at least 50°C to at least 70°C, such as for example at least 50°C to at least 75°C, such as at least 50°C to at least 80°C, such as at least 50°C to at least 85°C such as for example at least 50°C to at least 90°C or such as at least 50°C to at least 95°C.

HRM analysis can also be performed by increasing the temperature of the sample from at least 55°C to at least 70°C, such as for example at least 55°C to at least 75°C, such as at least 55°C to at least 80°C, such as at least 55°C to at least 85°C, such as for example at least 55°C to at least 90°C or such as at least 55°C to at least 95°C.

Alternatively, HRM analysis can be performed by increasing the temperature of the sample from at least 60°C to at least 70°C, such as for example at least 60°C to at least 75°C, such as at least 60°C to at least 80°C, such as at least 60°C to at least 85°C, such as for example at least 60°C to at least 90°C or such as at least 60°C to at least 95°C. HRM analysis can also be performed by increasing the temperature of the sample from at least 65°C to at least 70°C, such as for example at least 65°C to at least 75°C, such as at least 65°C to at least 80°C, such as at least 65°C to at least 85°C, such as for example at least 65°C to at least 90°C or such as at least 65°C to at least 95°C. In one preferred embodiment of the invention HRM analysis is performed by increasing the temperature of the sample from 55°C to 95°C. In another preferred embodiment of the invention HRM analysis is performed by increasing the temperature of the sample from 60°C to 95°C. In yet another preferred embodiment of the invention HRM analysis is performed by increasing the temperature of the sample from 65°C to 95°C. In a further preferred embodiment of the invention HRM analysis is performed by increasing the temperature of the sample from 55°C to 95°C. In another preferred embodiment of the invention HRM analysis is performed by increasing the temperature of the sample from 55°C to 90°C. The rate at which the temperature is increased during the HRM analysis may be in the range of 0.01 °C/s to 1 °C/s.

In a preferred embodiment the temperature during the HRM analysis is increased by o.rc/s.

CO-amplification at Lower Denaturation-PCR (COLD-PCR)

The method according to the present invention may be combined with co-amplification of lower denaturation temperature PCR (COLD-PCR), which selectively denatures the mutation containing amplicons. The method can however only be applied if the temperature of the mutant amplicon is lower than the temperature of the non-mutant amplicon. The underlying principle of COLD-PCR is that single nucleotide mutations may slightly alter the melting temperature of the double-stranded DNA if the mutation implies that number of hydrogen bonds in the amplicon is decreased, for instance, G>A mutations, C>T mutations, G>T mutations, or C>A mutations (melting temperature decreasing mutations). A single nucleotide melting temperature decreasing mutation anywhere along a double-stranded DNA sequence generates a small change in the melting temperature for that sequence, with mutated sequences melting at a lower temperature than wild-type sequences. COLD-PCR uses a critical temperature during the PCR process in order to enrich mutations of the amplified sequence. During the denaturation step in the PCR reaction, the temperature is set to this critical temperature that results in denaturation of only the mutant amplicon. Thereby, mutation-containing sequences are preferentially denatured and available for primer binding during the annealing step and subsequent amplification.

Thus, mutant amplicons generated by the method of the present invention may be further enriched by combining the method of the present invention with COLD-PCR, which may increase the sensitivity limit. This, however, requires that the melting temperature x of the mutant amplicon is lower than the melting temperature y of the non-mutant amplicon. Samples

The sample that is used in the method of the present invention may be in a form suitable to allow analysis by the skilled artisan. The samples according to the present invention may be selected from a tissue sample, or from body fluid samples such as samples selected from the group consisting of blood, plasma, serum, semen and urine.

In one embodiment of the present invention the sample is a tissue sample.

In another particular embodiment of the present invention, the sample is a tissue sample, such as a biopsy of the tissue, or a superficial sample scraped from the tissue. The tissue samples may also be in the form of Formalin-Fixed Paraffin Embedded blocks from biopsies (see the "Example" section for further details).

In another embodiment the sample may be prepared by forming a suspension of cells made from the tissue. The sample may, however, also be an extract obtained from the tissue or obtained from a cell suspension made from the tissue.

In another embodiment of the present invention the sample is a body fluid sample. The body fluid sample may be selected from the group consisting of blood samples, plasma samples, serum samples, semen samples and urine samples.

Working with tumor material in general requires biopsies or body fluids suspected to comprise relevant cells. Working with RNA in general requires freshly frozen or immediately processed biopsies, or chemical pre-treatment of the biopsy. Apart from the cancer tissue, biopsies do normally contain many different cell types, such as cells present in the blood, connective and muscle tissue, endothelium etc. In the case of DNA studies, microdissection or laser capture are methods of choice, however the time-dependent degradation of RNA may make it difficult to perform manipulation of the tissue for more than a few minutes. The sample may be fresh or frozen, or treated with chemicals.

Kit

The present invention also pertains to a kit for detecting the absence or presence of at least one mutation in a nucleotide sequence in a sample, said kit comprising at least one mutant primer, at least one non-mutant primer and at least one common primer as described herein.

In one embodiment the kit further comprises a temperature resistant DNA polymerase and appropriate substrates, nucleotides and cofactors to initiate amplification of DNA sequences.

The kit may also comprise fluorescently labelled oligonucleotide or hybridization probes, such as for example TaqMan probes for melting analysis. The fluorescently labelled oligonucleotide or hybridization probe may for example be DNA minor groove binding probes.

In one embodiment the kit comprises intercalating dyes for HRM analysis. The kit can for example be used for detecting mutations that may lead to cancer. In one embodiment the at least one mutation to be detected by the kit is in the human KRAS gene (SEQ ID NO: 1). The at least one mutation to be detected by the kit may be the human KRAS c.35G>C mutation (base pair no. 5571 of SEQ ID NO: 1). In another embodiment the at least one mutation to be detected by the kit is in the human BRAF gene (SEQ ID NO: 2). The at least one mutation to be detected by the kit may be the human BRAF c.1799T>A mutation (base pair no. 171429 of SEQ ID NO: 2). In another embodiment the at least one mutation to be detected by the kit is in the human EGFR gene (SEQ ID NO: 3). The at least one mutation to be detected by the kit may be the human EGFR c.2573T>G mutation (base pair no. 172791 of SEQ ID NO: 3) In yet another embodiment of the present invention the at least one mutation to be detected is in the human PIK3CA gene (SEQ ID NO: 16). The at least one mutation to be detected by the kit may be the human PIK3CA c.3140A>G mutation (base pair no. 86775 of SEQ ID NO: 16). EXAMPLES

Methods

Samples and DNA Extraction

Formalin-Fixed Paraffin Embedded (FFPE) blocks from surgical biopsies of 60 patients diagnosed with adenocarcinoma in colon were selected from the archives at the Institute of Pathology, Aarhus University Hospital. The specimens were between 2 and 10 years old. Specimens containing vital tumor cells were chosen by an experienced pathologist and no micro-dissections were performed prior to DNA extraction. For each sample, six tissue sections of 10 μηι were used for DNA extraction. Deparaffinization and DNA extraction were performed as previously described (Kristensen, L.S. et al., (2010); Hum Mutat, 31 , 1366-1373).

DNA from peripheral blood (PB), obtained from medical students of both sexes in their first year of Medical school, was extracted following a modified salt precipitation protocol as previously described (Hansen, L.L. et al., (1998); APMIS, 106, 371-377).

Cell lines and dilution series

Four different cell lines, FM82, CRL-5908, CRL-5883, RPMI 8226 and HCT 116 containing the BRAF c.1799T>A, EGFR c.2573T>G, KRAS c.35G>C mutations and PIK3CA C.3140 A>G mutations respectively, were used in this study. The cell lines were cultured in RPMI 1640 culture media (Invitrogen, Breda, The Netherlands) and supplemented with 10% fetal bovine serum, 2mM L-glutamine, and 10,000 lU/ml penicillin, and 0.1 mg/ml streptomycin (both Leo Pharma, Copenhagen, Denmark) antibiotics. The cell cultures were maintained at 37 °C in a humidified atmosphere of 5% C0 ₂ . The cells were harvested by scraping, and DNA extracted as described for PB with slight modifications; the centrifugation steps were carried out for 5 min at 13000 rpm instead of 15 min at 3000 rpm, and 1 volume of isopropanol was added instead of two volumes of ice cold ethanol.

DNA from each of the cell lines was quantified using a NanoDrop ND-1000

spectrophotometer (NanoDrop Technologies, Wilmington, DE) and serially diluted into wild-type (non-mutant) DNA obtained from PB from medical students to the following fractions of mutated alleles in a wild-type (non-mutant) background (50%, 25%, 5%, 1 %, 0.5%, 0.25%, 0.05% and 0.025%). Primer design For each mutation a primer was designed to selectively amplify mutation containing sequences. The following mutations were tested for detection by the method of the present invention:

BRAF c.1799T>A (base pair no. 171429 of SEQ ID NO: 2)

EGFR c.2573T>G (base pair no. 172791 of SEQ ID NO: 3)

KRAS c.35G>C (base pair no. 5571 of SEQ ID NO: 1)

PIK3CA .3140 A>G (base pair no. 86775 of SEQ ID NO: 16)

For each mutation a primer was designed to selectively amplify mutation containing sequences. This was done by including a 3' terminal mismatch to the wild-type sequences in the mutation specific primer. Each mutation specific primer was also designed to introduce one or two melting temperature decreasing mutations in the amplicon to enable direct detection by HRM analysis and selective amplification by COLD-PCR. A competitive primer that amplifies both mutant and wild-type (non- mutant) sequences was designed for each assay. This primer facilitates robust amplification of samples containing low-abundance mutations and competes with the mutant primer for target binding, thereby limiting false amplification from wild-type (non- mutant) sequences by the mutant primer. Finally, a common primer was designed for each assay to allow amplification of both mutant and wild-type (non-mutant) sequences.. In addition, the wild-type (non-mutant) amplicons were designed to be slightly longer than the mutant amplicons, which contributes to a higher melting temperature of wild-type amplicons versus mutant amplicons.

The primers were designed to target BRAF (SEQ ID NO: 2), EGFR (SEQ ID NO: 3), KRAS (SEQ ID NO: 1) and PIK3CA (SEQ ID NO: 16) sequences obtained from

GenBank [http://www.ncbi.nlm.nih.gov/GenBank/] (BRAF GenBank accession number NM_004333.4, EGFR GenBank accession number NM_005228.3, KRAS GenBank accession number NM_033360.2 and PIK3CA GenBank accession number

NM_006218). Primer sequences can be found in Table 1. The BRAF, KRAS and PIK3CA primers were designed to avoid amplification from their respective

pseudogenes (see Table 1).

PCR and HRM Conditions

PCR cycling and HRM analysis were performed on the Rotor-Gene 6000™ (Corbett Research, Sydney, Australia). SYTO ^® 9 (Invitrogen) was used as the intercalating dye. The final reaction mixtures consisted of 50 ng of DNA, 1x PCR buffer, 2.5 mmol/L MgCI ₂ , optimized relative primer concentrations (Table 1), 200 μΓΤΐοΙ/L of each dNTP, 5 mol/L of SYTO ^® 9, 0.5U of HotStarTaq (QIAGEN) (51Ι/μΙ_) in a volume of 20 μΙ_. The CADMA cycling protocol was initiated by one cycle at 95°C for 15 min, followed by 45 cycles of 95°C for 10 s, annealing temperature (T _A ) for 20 s (Table 1), 72°C for 20 s, and one cycle at 95°C for 1 min. When combining CADMA with COLD-PCR 10 standard PCR cycles were followed by 35 cycles of COLD-PCR in which the critical denaturation temperatures (T _c 's) were experimentally determined to ensure strong mutation enrichment while maintaining adequate amplification of all dilutions (Table 1). The COLD-PCR cycling protocol was T _c for 10 s, T _A for 20 s, 72°C for 20 s, and one cycle at 95°C for 1 min.

HRM was performed from 60°C to 95°C, with a temperature increase at 0.1 °C/s.

Samples were analyzed in triplicates (cell line experiments) or in duplicates (colorectal cancer specimens). 20 of water was added to the tubes in the rotor that were not in use, as it was observed that the actual temperature in the chamber is influenced by the number of empty tubes.

Data analysis

The Rotor-Gene 6000 Series Software version 1.7.87 supplied with the instrument was used to analyze the data. The derivative of the raw data (Melt curve analysis), and the normalized HRM and difference graphs (High resolution melting analysis), were used. For the difference graphs a wild type sample was selected as reference.

Example 1 : The analytical sensitivity of CADMA for the detection of BRAF, EGFR, KRAS and PIK3CA mutations

The present invention concerns a method for the detection of known mutations that has a very high sensitivity and specificity, while being convenient and cost-effective. The use of for example HRM as the detection platform presents several advantages. First, HRM can be coupled directly with the PCR and, therefore, the whole process can be performed in a closed tube system. This limits the risk of PCR contamination and allows for rapid and convenient analysis. Second, HRM is based on double stranded DNA (dsDNA) intercalating fluorescent dyes, which are cost-effective relative to the use of fluorescently labeled oligonucleotides. Finally, relative high-throughput is possible depending on the equipment used. The method of the present invention takes in one embodiment advantage of HRM and the high sensitivity provided by using a mutant primer that specifically amplifies mutant sequences, i.e. the mutant primer comprises a nucleotide sequence that is complementary to a nucleotide sequence comprising the mutation to be detected. Furthermore, a non-mutant primer, which anneals in the same region of the gene as the mutation specific primer, but does not comprise a sequence complementary to the mutation to be detected, was designed to amplify wild-type sequences and mutant sequences. This facilitates a competitive amplification between the mutant and non- mutant primers, which limits false amplification from wild-type sequences by the mutant primer. Furthermore, the two resulting amplicons was designed to melt differently by introducing one or more melting temperature decreasing mutations in the mutant amplicon to allow direct detection of the mutation by HRM analysis. Also, this design allowed us to successfully combine the assay with COLD-PCR to further increase the sensitivity.

The method, wherein the mutant primer and non-mutant primes are competitive primers may be referred to as Competitive Amplification of Differentially Melting Amplicons (CADMA).

The CADMA assays were optimized to avoid false amplification from wild-type sequences by the mutation specific primer, while maintaining a high sensitivity. False amplification can be identified as heteroduplex formation in wild-type samples when observing the melt curves. Heteroduplexes between non-mutant and mutant sequences have a lower melting temperature than the corresponding homoduplexes and thus melt earlier relative to the respective homoduplexes (Figure 2). Increasing the concentration of the mutant primer relative to the non-mutant primer that amplifies both mutant and non-mutant sequences increases the sensitivity of the assay but may also lead to false amplification of mutant amplicons in case the mutant primer anneals to the non-mutant nucleotide sequence. However, this may be circumvented by using a higher annealing temperature.

The BRAF c.1799T>A assay was sensitive to 0.25% mutant alleles. However, two out of three replicates containing 0.05% mutant alleles could be distinguished from 10 wild- type (non-mutant) replicates. The EGFR c.2573T>G was sensitive to 0.25% mutant alleles. However, two out of three replicates containing 0.05% mutant alleles, and three out of three replicates containing 0.025% mutant alleles could be distinguished from 10 wild-type (non-mutant) replicates. The KRAS c.35G>C assay was sensitive to 0.025% mutant alleles, and the PIK3CA c.3140 A>G assay was sensitive to 0.25% mutant alleles (Figure 3). The sensitivity limit was defined as the dilution point at which all three replicates could be distinguished from 10 wild-type replicates.

Example 2: Combining CADMA with COLD-PCR

The CADMA assays were designed to ensure that the mutant amplicons melt at lower temperatures relative to the wild-type (non-mutant) amplicons (Figure 1). This did not only allow for direct identification of low-abundance mutations by HRM analysis, it also implies that the sensitivity can be further improved by combining the assay with COLD- PCR, which is a new form of PCR that selectively amplifies mutation-containing templates based on the lower melting temperature of mutant homoduplexes versus wild type homoduplexes by using a critical denaturation temperature (T _c ) (Li, J., Wang, et al., (2008); Nat Med, 14, 579-584). The T _c selected for each assay further directed the PCR bias towards the amplification of mutation containing sequences (data not shown). Thus, the use of COLD-PCR may increase the sensitivity of the CADMA assay (Compare Figure 3 and 4).

The BRAF c.1799T>A, EGFR c.2573T>G, and KRAS c.35G>C assays were sensitive to 0.025% mutant alleles (Figure 4).

The sensitivity limit was defined as the dilution point at which all three replicates could be distinguished from 10 wild-type replicates.

Example 3: The use of more DNA increases the analytical sensitivity of the KRAS assay In order to assess whether the sensitivity limit was determined by the magnitude of the wild-type DNA background or by available template, the KRAS c.35G>C assay (CADMA combined with COLD-PCR) was performed using five times as much DNA in the reactions (250 ng). Using five times as much DNA increased the sensitivity of the assay 2.5-fold so that the 0.01 % standard could be easily distinguished from ten wild- type samples (Figure 5). Thus, for this assay the sensitivity seems to be determined by available template and not by the amount of background wild-type DNA.

Example 4: Detection of BRAF mutations in clinical specimens derived from FFPE tissues 58 colorectal cancer specimens derived from FFPE tissues for BRAF mutations have been analyzed using the CADMA and COLD-PCR assay. However, five additional standard PCR cycles was performed before switching to COLD-PCR in order to get adequate amplification from all samples when using the BRAF assay. When analyzing DNA derived from FFPE tissues more variation of the melting curves can be expected. This was also the case in the present study (Figure 6). However, the wild-type colorectal cancer samples were still very easily distinguished from the standards containing 0.25% mutated DNA in a wild-type background. In total 6 samples (10%) were found to carry a BRAF mutation.

To determine whether CADMA can detect low-abundance mutations in a large wild- type background when the DNA is of poor quality (derived from FFPE tissues) one of the clinical specimens for which a BRAF mutation was detected into wild-type DNA derived from PB were serially diluted in the following fractions; 100%, 50%, 25%, 12.5%, 6.25%, 3.125%, 1.56%, 0.78%, and 0.39%. When using the CADMA assay combined with COLD-PCR the 0.39% dilution point could be readily distinguished from 10 wild-type replicates (Figure 7). When the assay was performed as a traditional PCR and HRM assay using 200nM each of the BRAF forward and reverse primers without the mutation specific primer, only the 50% dilution point could be distinguished from the wild-type replicates (data not shown). Thus, the sensitivity was increased by at least a factor of 128 when using CADMA as opposed to standard PCR and HRM analysis.

For comparison, the same samples have been analyzed by a TaqMan based assay, which has been published recently (Lang, et al., 2011), and by sequencing as described in the Materials and Methods section. The TaqMan assay determines mutation status using a predetermined cutoff ACt value (Ct (allele-specific assay) - Ct (reference assay)) as described in (Lang, et al., 2011). The sensitivity limit of the assay was reported to be 1 % mutant alleles in a wild-type (non-mutant) background. The sequencing assay was sensitive to 25% mutant alleles in a wild-type (non-mutant) background (data not shown).

When analyzing the 58 colorectal cancer specimens a 100% concordance between the CADMA combined with COLD-PCR assay and the TaqMan assay was found. However, when using the sequencing assay only three samples were found to carry a BRAF mutation. In the CADMA combined with COLD-PCR assay these samples had melting curves which were close to the melting curves of the standard containing 50% mutant alleles, whereas the samples for which the mutations could not be detected by sequencing had melting curves which more resembled the melting curves of the standard containing 5% mutant alleles. In the TaqMan assay the samples that tested mutation positive by sequencing had ACt values between 0.74 and 1.15, whereas the samples for which the mutation only were detected by the TaqMan assay and the CADMA combined with COLD-PCR assay showed ACt values between 3.19 and 5.84, which is consistent with the mutation being present in less than 25% of the alleles.

Example 5: False positive results are prevented by the competitive primer

False positive results may arise if the mutant primer anneals to the non-mutant DNA strand and starts amplification. This will result in an amplicon comprising the mutation to be detected although the template used is the non-mutant DNA strand not comprising the mutation.

The mutant primer and the competitive primer that amplifies both mutant and wild-type DNA are likely to compete for target binding. This may limit the potential for nonspecific binding of the mutant primer to the wild type (non-mutant) DNA strand and thereby contribute to an increased specificity of the assays. This issue has been assessed by performing the CADMA assays without the competitive primer, otherwise using the exact same PCR conditions. False amplification from wild-type (non-mutant) sequence was observed in all assays when the competitive primer was excluded (Fig.

1 1) , and the melting curves showed that this amplification was indeed a result of nonspecific binding of the mutant primer to the wild type (non-mutant) DNA strand (Fig.

12) . In the EGFR and BRAF assays ten out of ten wild-type replicates amplified, and in the KRAS and PIK3CA assays one out of ten and five out of ten wild-type replicates amplified, respectively. However, when the non-mutant primer was included no heteroduplexes could be observed in wild-type samples (Figure 12), and data not shown) indicating that non-specific binding by the mutant primer was prevented by adding a competitive primer. Further evidence of a competition between the competitive primer and the mutant primer was provided by adjusting relative primer concentrations. When decreasing the concentration of the competitive primer relative to the mutant primer heteroduplexes could be observed in wild-type replicates. Therefore, it is important to optimize relative primer concentrations and annealing temperatures to ensure a high analytical sensitivity and specificity. This applied to all of the CADMA assays. An example is shown in Fig. 13.

Thus, presence of the non-mutant primer in the reaction mixture prevents non-specific binding of the mutant primer to the non-mutant DNA strand, thereby preventing the generation of false positive results.

Example 6: BRAF Mutation Testing in melanoma

A number of different detection methods are being used in melanoma research for detection of the BRAF V600E mutation, including Sanger sequencing, pyrosequencing, high-resolution melting (HRM) analysis, and various allele-specific PCR-based methods. Sanger sequencing, pyrosequencing, and HRM analysis generally suffer from a relatively low sensitivity, and allele-specific PCR assays may be susceptible to false positive results if false amplification of wild-type DNA occurs in early PCR cycles despite mutation-specific primers. COLD-PCR is capable of selectively amplifying mutant alleles based on melting temperature differences between mutant and wild-type amplicons. This greatly improves the sensitivity of downstream detection methods (Li J et al., Nat Med 2008, 14:579-584.). However, when the mutation in question is melting- temperature retaining mutation (such as the BRAF V600E mutation) or a melting- temperature increasing mutation, only a modest increase in sensitivity may be observed (Stadelmeyer E et al., J Mol Diagn 2011 , 13:243). As described herein, CADMA enables sensitive and direct detection of all mutation types by high-resolution melting analysis.

In the present example it is shown that the frequency of detected BRAF V600E mutation in cutaneous melanoma is influenced directly by the analytical sensitivity of the applied method. For this purpose five different methods, the Cobas® 4800 BRAF V600 Mutation Test, Sanger sequencing, pyrosequencing, TaqMan-based allele- specific PCR, and CADMA were used. The analytical sensitivity of each method was determined using a serial dilution of mutated cell line DNA in a wild-type background. It was also investigated how the percentages of tumor cells in primary cutaneous melanoma DNA samples derived from Formalin-Fixed Paraffin-Embedded (FFPE) tissues influenced the detection capabilities of the methods applied. Finally, it was evaluated whether tumors may be BRAF mutation-positive as a result of only a small fraction of the tumor cells carrying the BRAF mutation, as such tumors may not respond to Vemurafenib. Materials and Methods

Tissue samples and cell line standard dilution

A total of twenty-eight FFPE cutaneous melanoma tissue samples (collected from 2009 to 2010) were obtained from the pathology archives at Aarhus University Hospital. The clinicopathological features are shown in Table 2.

A standard dilution of mutant alleles in a wild-type background was prepared using DNA extracted from the cell line FM82, which is heterozygous for the BRAF V600E mutation. DNA extracted from peripheral blood, obtained from Danish medical students in their first year of Medical school at Aarhus University, was used as wild-type DNA. Using pyrosequencing it was estimated that the cell line does indeed contain 50% mutant alleles, however, if the fraction of BRAF copies relative to the overall amount of genomic material in the cell line is not the same in the cell line and blood sample, this will result in a bias between the methods which amplify both mutated and wild-type DNA and the methods which only or preferably amplify mutated DNA. To correct for this potential bias three dilutions of the cell line DNA into wild-type DNA (40%, 30%, and 20% mutant alleles) were carried out and the allele frequencies using

pyrosequencing were measured. These measurements were then used to calculate which concentration the wild-type DNA sample should have to avoid a bias, before serially diluting the cell line DNA into wild-type DNA to the following fractions (50%,

20%, 10%, 5%, 2.5%, 1.25%, 0.625%, 0.3125%, 0.15625% and 0.078125%). The DNA was quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop

Technologies, Wilmington, DE). Tumor tissue fraction estimation

The percentage of tumor tissue was estimated semi-quantitatively on H&E stained slides with two different methods and subdivided into the following intervals: <10%, 10- 49%, and >50%.

In the first method the tumor tissue area was estimated as a fraction of all visible tissue at the slide. In the second method the tumor tissue area was estimated as a fraction of cell dense tissue only (fatty tissue and cell deprived tissue was excluded). DNA extraction

Each FFPE sample was sectioned using a microtome (five 10 μηι thick slides) and DNA was extracted using a commercial kit (Qiamp DNA FFPE Tissue Kit from Qiagen, Qiagen AB, Sweden) following the manufacturer's protocol.

Pyrosequencing

Primers for the pyrosequencing assay were designed using the PyroMark Assay Design Software (Qiagen). The amplification forward primer 5'- TTCATGAAGACCTCACAGTAAAAA-3' (SEQ I D NO: 17) and reverse primer 5'-

GGCCAAAAATTTAATCAGTGGAA-3' (SEQ ID NO: 18) amplified a 152-bp region of BRAF containing the c.1799T>A variant. The forward primer was 5' biotin-labeled to enable preparation of a single stranded template. Amplification was performed in 25-μΙ reactions containing 200 nmol/l of each primer, 0.1 mmol/l each dNTP, 1 unit of HotStarTaq DNA Polymerase (Qiagen), and 25 ng of DNA. Reactions were started by denaturation at 95°C for 15 min, followed by 45 cycles of 95°C for 10 sec, 56°C for 20 sec and 72°C for 20 sec. Pyrosequencing was performed on a PyroMark Q24 platform (Qiagen), using PyroMark Gold Q24 Reagents and the sequencing primer 5'- GCCAGGTCTTGATGTACT-3' (SEQ ID NO: 19), with the following dispensation order: GCATCTGT. Data analysis was performed with the PyroMark Q24 Software.

Sequencing was performed in the reverse direction. Samples which were borderline positive (6-7%) were repeated. If the second run gave 6% or more, the sample was scored as positive. Allele-specific PCR (TaqMan)

The TaqMan based allele-specific PCR assay used herein has been published recently (Lang AH, Drexel H, Geller-Rhomberg S, Stark N, Winder T, Geiger K, Muendlein A: Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF. J Mol Diagn 201 1 , 13:23-28.). PCR conditions and real-time PCR instrument as described by Lang et al. (2011) were used. When analyzing the samples 50 ng of each sample were used. This assay determines mutation status using a predetermined cutoff ACt value (Ct [allele-specific assay] - Ct [reference assay]) of nine as described. The analytical sensitivity of the assay was reported to be 1 % mutant alleles in a wild-type background (Lang AH, Drexel H, Geller-Rhomberg S, Stark N, Winder T, Geiger K, Muendlein A: Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF. J Mol Diagn 2011 , 13:23-28). Samples were analyzed in duplicates for all TaqMan experiments.

Competitive Amplification of Differentially Melting Amplicons (CADMA)

PCR cycling and HRM analysis were performed on the LightCycler® 480 (Roche

Applied Science, Mannheim, Germany). The reaction mixtures consisted of 25 ng DNA using a 1x final concentration of the LC480 HRM Scanning Master (Roche), and a final MgCI2 concentration of 2.5 mmol/l. Primer concentrations were; 200 nmol/l of the reverse non-mutant primer, 5'-TGATGGGACCCACTCCATCG-3' (SEQ ID NO: 5), 400 nmol/l of the mutation specific reverse primer 5'-

TGAGACCCACTCTATCGAGATTTCT-3' (SEQ ID NO: ???), and 400 nmol/l of the common forward primer 5'-AGGTGATTTTGGTCTAGCTACAG-3' (SEQ ID NO:4). The cycling protocol was initiated by one cycle at 95°C for 10 min, followed by 15 standard PCR cycles of 95°C for 10 sec, 69°C for 20 sec, 72°C for 20 sec, followed by 35 fast COLD-PCR cycles of 78°C for 10 sec, 69°C for 20 sec, 72°C for 20 sec, and a final denaturation step at 95°C for 1 min. The HRM step was performed from 65°C to 95°C using 30 acquisitions per °C. Samples were analyzed in duplicates for all CADMA experiments.

The analytical sensitivity of the CADMA assay was determined as the dilution point at which both replicates could be distinguished from 10 wild-type replicates (Fig. 11 A).

Cutaneous melanoma samples were scored as mutation positive if the melting profiles deviated more from the wild-type melting curves than the standard containing approximately 0.15625% mutant alleles in a wild-type background. Sanger sequencing

PCR amplicons were generated using the following primers; forward: 5'- AGGTGATTTTGGTCTAGCTACAG-3' (SEQ ID NO:4) and reverse: 5'- GTTGAGACCTTCAATGACTTTCTAG-3' (SEQ ID NO: 20) and analyzed on the ABI Genetic Analyzer 3130 XL (Applied Biosystems, Foster City, California, US), using the BigDye terminator kit v1.1 (Applied Biosystems) according to the manufactures' instructions with slight modifications; the single-stranded PCR was performed using 1 μΙ of the BigDye terminator in a final volume of 10 μΙ. Sequencing was performed in the reverse direction only.

The Cobas® 4800 BRAF V600 Mutation Test The Cobas® 4800 BRAF V600 Mutation Test (Molecular Diagnostics, Roche

Diagnostics A/S, Hvidovre, Denmark) is a real-time PCR analysis using TaqMan probes and it is designed for detecting the presence of the V600E mutation. Results are binary (mutation detected or mutation not detected). Except for the preparation of the tissue (cutting the paraffin blocks and DNA extraction), all analyses were done according to the manufacturer's protocol including dilution and standardization of the samples so that at least 125 ng DNA was used from each sample.

Statistics

Statistical analyses were done using Stata/IC 1 1 (StataCorp LP, College Station, Texas, US).

Inter-and intra observer variation was calculated using un-weighted kappa statistics and interpreted as poor, slight, fair, moderate, substantial or almost perfect, according to previously defined groups (Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics 1977, 33: 159-174). The zero hypothesis of no difference in tumor size (Breslow thickness) was estimated using a two-sample student's t-test on a logarithmically transformed scale. Direct estimates of diagnostic sensitivity and specificity of each method were not possible, as no reliable

predetermined gold standard exists.

Results

Serial dilution of mutated cell line DNA and the relative analytical sensitivity

The analytical sensitivities of all assays were compared by analyzing a serial dilution of

DNA extracted from the cell line FM82, which is heterozygous for the BRAF V600E mutation, into wild-type DNA. The analytical sensitivity of the CADMA assay was determined as the dilution point at which both replicates could be distinguished from 10 wild-type replicates (Figure 11 A). The analytical sensitivity of the TaqMan assay was determined as the dilution point which had a ACt value below nine. Each dilution was analyzed once when using Sanger sequencing, pyrosequencing, and the Cobas Test. The results have been summarized in Table 3.

Pyrosequencing is a quantitatively accurate method, and from the values in the pyrograms it was observed that the actual allele frequencies in the dilutions were a little lower than expected. Sanger sequencing and the Cobas test were the least sensitive (theoretical fraction of 30% and 20% mutant alleles respectively). Pyrosequencing was more sensitive and capable of detecting the mutation in the dilution having a theoretical fraction of 10% mutant alleles (Fig. 11 B). The TaqMan assay was sensitive to 2.5% mutant alleles and the CADMA assay to 0.078% mutant alleles, however, one of the two replicates of this dilution point was only barely distinguishable from the wild-type replicates (Fig. 11 A).

Detection of the BRAF V600E mutation in clinical samples (n=28)

Twenty-eight primary cutaneous melanoma samples derived from FFPE tissues were screened for the BRAF V600E mutation using the Sanger sequencing,

pyrosequencing, Taqman, and CADMA assays, and the Cobas Test. The results of all mutation detection methods are summarized in Table 4.

The frequency of detected mutations varied substantially depending on the analytical sensitivity of the method used for their detection and the fraction of tumor cells in the samples (Table 5). The overall mutation detection frequency was 29% by Sanger sequencing, 36% by the Cobas test, 43% by pyrosequencing, 46% by CADMA, and 50% by TaqMan. Sample 13 was positive by TaqMan and not by CADMA, and was therefore repeated twice by TaqMan and once by CADMA and found to be negative in all of these runs. The TaqMan result for sample 13 in Table 4 may therefore be a false- positive result.

Using pyrosequencing, other mutations than the V600E mutation were detected in two samples but were not detected by Sanger sequencing. To verify these two mutations, the samples were tested using denaturing gradient gel electrophoresis (DGGE), which confirmed the presence of other mutations than the V600E mutation (data not shown).

Pyrosequencing and TaqMan are quantitative methods. There was a good correlation between the values from the pyrograms and the ACt values from the TaqMan assay. High values in the pyrograms (15% - 24%) were observed in the samples having low ACt values (2.0 - 2.8). Whereas low values in the pyrograms (6% - 7%) were observed in the samples having high ACt values (4.5 - 6.3). CADMA is, as indicated in Fig. 11 A, semi-quantitative. Therefore, samples which had low ACt values in the TaqMan assay could be distinguished from those having higher ACt values (Fig. 12).

Consensus between all five assays was observed in 22 out of 28 samples (79%).

Consensus between the two most sensitive methods, TaqMan and CADMA, was observed in 27 out of 28 (96%). Consensus between Cobas and TaqMan was observed in 24 out of 28 samples (86%). Consensus between Cobas and CADMA was observed in 25 out of 28 samples (89%). Consensus between Cobas and

pyrosequencing was observed in 27 out of 28 samples (93%). Finally, consensus between Cobas and Sanger sequencing was observed in 26 out of 28 samples (96%). The consensus was 100% between all five assays in samples containing more than 10% tumor tissue (of all tissue) or more than 50% tumor tissue (of cell dense tissue).

Finally, it was evaluated whether some tumors were scored as mutation positive as a result of only a small fraction of the tumor cells carrying the BRAF mutation. For this purpose, the ACt values of the semi-quantitative TaqMan assay were compared with the estimates of percentages of tumor cells in the samples. ACt values of more than 3.9 indicate that the mutation is present in less than 10% of the alleles (Lang AH et al.: Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF. J Mol Diagn 201 1 , 13:23-28.). This was the case for five of the tumor samples (Table 4), however, these were all estimated to contain less than 10% tumor cells (of all tissue). Thus, no evidence for tumors being mutation positive as a result of only a fraction of the tumor cells carrying the BRAF mutation were observed. Tumor fraction estimation methods

When estimating tumor as a fraction of all tissue on the slide a moderate inter observer agreement (79%, Kappa=0.58) and a substantial intra observer agreement (89%, Kappa=0.77) were found. Using this counting method a cut-off point of 10% tumor cells was found, above which complete consensus between the five mutation detection methodologies was observed (Table 4).

When estimating tumor tissue as a fraction of cell dense tissue only, a moderate inter- and intra observer agreement (68%, Kappa=0.49, and 61 %, Kappa=0.43, respectively) were found. Using this counting method the best cut-off point was 50% tumor cells, above which complete consensus between the five mutation detection methodologies was observed.

Tumor size

There was no difference in Breslow thickness between BRAF V600E positive samples compared to BRAF V600E negative samples when excluding samples with less than 10% tumor tissue (as a fraction of all tissue; p=0.7194) or <50% tumor tissue (as a fraction of cell dense tissue only; p=0.9852). Conclusion

In conclusion, the CADMA assay proved to have the highest sensitivity corresponding to 0.078% mutant alleles, i.e. the method can detect a mutation, which is present in at least 0.078 % of the alleles in the sample. The TaqMan assay had a sensitivity of 2.5% mutant alleles, whereas Sanger sequencing and pyrosequencing were less sensitive. In particular, Sanger sequencing and the Cobas test failed to detect mutations in a significant proportion of the samples, which contained small fractions of tumor cells. For this reason, our results underscore the notion that it is important to evaluate the percentage of tumor tissue relative to non-tumor tissue in the samples prior to performing mutation analysis when using less sensitive methods such as Sanger sequencing and the Cobas test.

So far the only FDA-approved BRAF V600E mutation detection test is the Cobas® 4800 BRAF V600 Mutation Test. However, the present study shows that this test lacks sufficient analytical sensitivity to detect mutations in small tumors, unless macro- dissections are performed before DNA is extracted. In addition to this time-consuming procedure, specialized equipment and expensive reagents are required to perform the test which is non-quantitative. The use of more sensitive and quantitative methods such as CADMA which are less time-consuming and less expensive may therefore have a future in testing for BRAF mutations in clinical settings. Example 7: KRAS Hotspot Mutation Testing in Metastatic Colorectal Cancer Patients.

The CADMA assay was further optimized for the seven most common KRAS exon 2 hotspot mutations. The sensitivity and specificity of each assay were evaluated using serial dilutions of cell line DNA containing the relevant mutations in a wild-type background. The potential of these assays were evaluated for the detection of KRAS mutations in CRC samples derived from formalin fixed paraffin embedded (FFPE) tissues. In total 100 samples were tested using the CADMA assays, and these results were compared with results obtained using the TheraScreen ^® KRAS mutation kit, which tests for the same seven mutations. Samples which did not give the same result when using CADMA and the TheraScreen kit were tested using a previously published highly sensitive TaqMan based assay. MATERIALS AND METHODS

Samples and DNA Extraction

Formalin-Fixed Paraffin Embedded (FFPE) blocks from surgical biopsies from 100 patients diagnosed with adenocarcinoma in colon were selected from the archives at the Department of Pathology, Aarhus University Hospital. The specimens were up to 10 years old. Specimens containing vital tumor cells were chosen by an experienced pathologist and no micro-dissections were performed prior to DNA extraction. For each sample, six tissue sections of 10 μηι were used for DNA extraction. Deparaffinization and DNA extraction were performed as previously described (Kristensen, L.S. et al., (2010); Hum Mutat, 31 , 1366-1373).

DNA from peripheral blood (PB) was obtained from medical students of both sexes in their first year of Medical school, and used as wild-type controls. The DNA was extracted following a modified salt precipitation protocol as described in Hansen LL et al., APMIS 1998, 106(3):371-377. The Local Ethical Committee, Aarhus County, Denmark, approved this study.

Cell lines and dilution series

Seven different cell lines each containing different KRAS mutations were used in this study; A549 (c.34 G>A, codon 12), DLD-1 (c.38 G>A, codon 13), LS174T (c.35 G>A, codon 12), NCI-H23 (c.34 G>T, codon 12) PSN-1 (c.34 G>C, codon 12), RPMI 8226 (c.35 G>C, codon 12), and SW480 (c.35 G>T, codon 12). The cell lines were cultured and harvested, and the DNA was extracted as described (Kristensen, L.S. et al., (2010); Hum Mutat, 31 , 1366-1373), with the exception of PSN-1 for which extracted DNA was purchased from Health Protection Agency Culture Collection, UK, and NCI- H23 for which extracted DNA was kindly donated by Professor Dmitri Loukinov, NIAID/NIH.

DNA from each cell line was quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE) and serially diluted into wild-type DNA obtained from PB from medical students to the following fractions of mutated alleles in a wild-type background; 50%, 10%, 1 %, and 0.5% (assuming no pipetting errors and that all cell lines are monoclonal). CADMA primer design

The primers were designed for the seven most common KRAS exon 2 mutations. Each mutant primer introduces two melting temperature decreasing mutations in the mutated amplicon to enable direct detection by HRM analysis.

The primer sequences were designed to target the KRAS sequence obtained from

GenBank [http://www.ncbi.nlm.nih.gov/GenBank/] (KRAS GenBank accession number NM_033360.2). The primers were designed to avoid pseudogene amplification.

Each mutant primer was designed to selectively amplify mutation containing sequences. In this example the following mutations were tested for detection by the method of the present invention:

KRAS c.34G>A (base pair no. 5570 of SEQ ID NO: 1)

KRAS c.34G>T (base pair no. 5570 of SEQ ID NO: 1)

KRAS c.34G>C (base pair no. 5570 of SEQ ID NO: 1)

KRAS c.35G>A (base pair no. 5571 of SEQ ID NO: 1)

KRAS c.35G>T (base pair no. 5571 of SEQ ID NO: 1)

KRAS c.35G>C (base pair no. 5571 of SEQ ID NO: 1)

KRAS c.38G>A (base pair no. 5574 of SEQ ID NO: 1) Primer sequences can be found in Table 6.

PCR and HRM Conditions for the CADMA assays

PCR cycling and HRM analysis were performed on the Rotor-Gene 6000™ (Corbett Research, Sydney, Australia) or the Rotorgene Q (Qiagen, Hilden, Germany). SYTO® 9 (Invitrogen) was used as the intercalating dye. The final reaction mixtures consisted of 50 ng of DNA, 1x PCR buffer, 2.5 mmol/L MgCI2, optimized relative primer concentrations (Table 6), 200 μηιοΙ/Ι_ of each dNTP, 5 μηιοΙ/Ι_ of SYTO® 9, 0.5U of HotStarTaq (Qiagen) (51Ι/μΙ_) in a volume of 20 μΙ_. The CADMA cycling protocol was initiated by one cycle at 95°C for 15 min, followed by 45 cycles of 95°C for 10 s, annealing temperature (TA) for 20 s (Table 6), 72°C for 20 s, and one cycle at 95°C for 1 min. HRM was performed from 65°C to 95°C with a temperature increase of 0.1 °C/s. Samples were analyzed in triplicates (cell line experiments) or in duplicates (colorectal cancer specimens). CADMA data analysis

The Rotorgene 6000 Series Software version 1.7.87 supplied with the instrument was used to analyze the data. The derivative of the raw data (melt curve analysis), and the normalized HRM and difference graphs (high resolution melting analysis), were used. For the difference graphs a wild type sample was selected as reference. The mCRC samples were tested for one mutation at a time. When a sample was scored as mutation positive it was not tested using the remaining CADMA assays unless the result was in disagreement with the result provided by the TheraScreen kit.

Mutation analysis using the TheraScreen® KRAS mutation Kit

The mCRC samples were analyzed using the TheraScreen® KRAS mutation kit

(Qiagen). This kit analyzes the mutation status for the seven most commonly found KRAS exon 2 mutations by a technology that combines ARMS® (allele specific PCR) with Scorpions® real-time PCR. The manufacturer has reported the sensitivity to be 1 % mutant alleles in a wild type background if sufficient DNA input is used.

Alelle-specific PCR (TaqMan)

The TaqMan based allele-specific PCR assay used herein has been published recently in Lang et al., (2011) (Lang AH et al., Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF", The Journal of molecular diagnostics : JMD 201 1 , 13(1):23-28). PCR conditions and real-time PCR instrument as described in Lang et al. (201 1) were used. This assay determines mutation status using a predetermined cutoff ACt value (Ct [allele-specific assay] - Ct [reference assay]) as described in Lang et al. (2011). The analytical sensitivity of the assays was reported to be 1 % mutant alleles in a wild-type background (Lang et al. (2011)).

Samples were analyzed in duplicates for all TaqMan experiments and the average Ct value of the duplicates was used to calculate ACt values.

RESULTS The analytical sensitivity of the CADMA assays for the detection of KRAS hotspot mutations

The CADMA assays were optimized to avoid false amplification from wild-type sequences by the mutation specific primer, while maintaining a high sensitivity. False amplification can be identified as either heteroduplex formation between mutant and wild-type sequences or mutant homoduplex formation in wild-type samples when evaluating the melt curves. When false amplification was seen the concentration of the non-mutant primer, which amplifies both mutant and wild-type sequences, was increased to prevent false amplification by the mutation specific primer. After successful optimization of the assays, the sensitivity and specificity were evaluated by running ten wild-type replicates together with a standard dilution series of mutant alleles into wild-type alleles (50%, 10%, 1 %, and 0.5%) in triplicates. All of the three replicates of the 0.5% standard could be distinguished from the ten wild-type replicates in all assays (Fig. 13).

To assess in between run variation, each of these experiments were repeated using the Rotorgene Q (Qiagen, Hilden, Germany) in a different laboratory. Again, the three replicates of the 0.5% standard could all be distinguished from the ten wild-type replicates in all assays, and no false amplification in wild-type replicates was observed (Fig. 14). However, the separation between the standard dilutions and the wild-types were more pronounced in some runs than others, indicating that some in between run variation may occur (compare Fig. 13 and Fig. 14).

KRAS hotspot mutation analysis in CRC samples using CADMA and the

TheraScreen® KRAS mutation Kit

100 mCRC samples derived from FFPE tissues for KRAS mutations were analyzed using the CADMA assays and the TheraScreen® kit. The standards containing 1 % mutant alleles were used as cut-off point in the CADMA assays to facilitate direct comparison with the TheraScreen® kit. When using CADMA, the samples were scored manually by visual inspection of the derivative of the raw data (melt curve analysis) and the normalized HRM and difference graphs (high resolution melting analysis).

Examples from the KRAS c.34 G>T CADMA assay are shown in Fig. 15. Since the non-mutant primer in the present case amplifies both mutant and non-mutant (wild- type) sequences each CADMA assay may detect KRAS mutations other than the one targeted by the mutant primer albeit at a lower sensitivity. The shape of the melting curves could easily be used to distinguish the mutation, targeted by each CADMA assay, from other mutations detected by the non-mutant primer, as the resulting amplicons have different melting properties, due to the two additional mismatches incorporated by the mutant primer. Examples of this are also shown in Fig. 15. When samples amplify late this may cause the melting curves to be shifted, and other abnormalities, such as the one shown for sample ID 69 in Fig. 15, may also result in deviations of the melting curves. Shifted melting curves may result in differences in the normalized HRM and difference graphs, which could lead to wrong interpretation of the results, if the melting curves are not inspected. When testing DNA samples derived from FFPE tissues more variation in the melt curves is likely to be observed compared to DNA samples of high quality. For this reason, it is also important to analyze the samples of unknown mutation status relative to standards of known ratios of wild-type to mutant alleles. Generally, the wild-type samples showed more variation in the c.38 G>A CADMA assay compared to any of the other CADMA assays (Fig. 16). However, this did not give rise to misclassification of any of the samples.

When using the TheraScreen kit 45/98 (45.9%) of the samples were found to carry a KRAS exon 2 mutation. When using CADMA 44/99 (44.4%) of the samples were mutation positive. The results have been summarized in Table 2. Consensus between the two methods were found in 93/97 (95.9%) of the samples. Samples which did not give the same result when using CADMA and the TheraScreen kit were tested using a previously published TaqMan based assay, which was reported to be sensitive to 1 % mutant alleles (Lang AH et al., Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF. The Journal of molecular diagnostics: JMD 201 1 , 13(1):23-28). This was the case for four of the samples (Table 7). For one of these samples (Sample ID 73) the TheraScreen kit was positive for the c.35 G>A mutation, while being negative by CADMA. Another sample (Sample ID 65) was positive for the c.34 G>A mutation and the c.35 G>C mutation by the TheraScreen kit, while only being positive for the c.35 G>C mutation by CADMA. Sample ID 91 was positive for the c.35 G>A mutation by the TheraScreen kit and the c.38 G>A by CADMA. Finally, sample ID 96 was positive for the c.34 G>T mutation by the

TheraScreen kit and the c.34 G>T mutation and the c.35 G>A mutation by CADMA. The CADMA results were confirmed by the TaqMan assay for all four samples.

In the present example, 100 mCRC samples using the TheraScreen kit and the CADMA assays were analyzed. To facilitate direct comparison between the two methods, the samples were analyzed at 1 % mutant level when using CADMA. Overall, consensus between the two methods was very high (95.9%), and in four out of four cases where different results were observed, the CADMA result could be confirmed by a previously published TaqMan based assay, which is also sensitive to about 1 % mutant alleles in a wild-type background (Lang AH et al., Optimized allele-specific real-time PCR assays for the detection of common mutations in KRAS and BRAF", The Journal of molecular diagnostics : JMD 201 1 , 13(1):23-28). Therefore, it is likely that the TheraScreen kit gave false positive results in three cases (c.34 G>A in sample ID 65 and c.35 G>A in samples ID 73 and 91) and false negative results in two cases (c.38 G>A in sample ID 91 and c.35 G>A in sample ID 96). Nevertheless, mCRC patients are pro tern only classified as mutation positive or negative for selection of treatment groups. Therefore, only one out of the 100 patients studied is likely to have been misclassified. False positive and negative results have previously been observed when using the TheraScreen kit in the order 1-2% (Tol J et al., "High sensitivity of both sequencing and real-time PCR analysis of KRAS mutations in colorectal cancer tissue", Journal of cellular and molecular medicine 2010 , 14(8) : 2122-2131 ) .

In conclusion, CADMA improves screening for KRAS mutations in mCRC. The use of an overlapping primer, which competes with the mutation specific primer for target binding reduces or eliminates false amplification, which is often observed in allele- specific PCR. In addition, the robust amplification by the overlapping primer of samples containing low abundance mutations prevents false negative results.

„

66

TABLES

Table 1. Details of the method. Primer sequences are given in 5'→ 3' directions.

Mutation Primers Primer Annealing Critical concentrations temperature temperature

BRAF SEQ ID NO: 4 200 nM 70°C 77.5°C c.1799T>A Common primer forward:

aggtgattttggtctagctacag

SEQ ID NO: 5 100 nM

Non-mutant primer reverse:

tgatgggacccactccatcg

SEQ ID NO: 6 200 nM

Mutant primer reverse:

tgggacccactctatcgagatttct

EGFR SEQ ID NO: 8 100 nM 70°C 82°C c.2573T>G Non-mutant primer forward:

cgcagcatgtcaagatcacagat

SEQ ID NO: 9 200 nM

Mutant primer forward:

g catgtca ag atcacag attttagg eg

SEQ ID NO: 7 200 nM

Common primer reverse:

tctcttccgcacccagcagt

KRAS SEQ ID NO: 14 50 nM 58°C 80°C c.35G>C Non-mutant primer forward:

atgactgaatataaacttgtggtagttg

SEQ ID NO: 15 200 nM

Mutant primer forward:

gaatataaacttgtagtaattggagctgc

SEQ ID NO: 13 200 nM

Common primer reverse:

tgtategtcaagg ca ctcttg cc

PIK3CA SEQ ID NO: 1 1 500 nM 68°C N/A c.3140 A>G Non-mutant primer forward:

ggagtatttcatgaaacaaatgaatgatg

SEQ ID NO: 12 600 nM

Mutant primer forward:

agtatttaatgaaacaaatgaatgatacacg

SEQ ID NO: 10 600 nM

Common primer reverse:

catttttgttgtacagccatcatgac Table 2. The clinocopathological features of the 28 study patients.

Feature Number (%)

Gender

Males 16 (57)

Females 12 (43)

Breslow (mm)

Mean (95% CI) 1.14 (1.03; 1.27)

Range (min-max) 0.60-1.89

Median 1.12

Clark level

3 9 (32)

4 19 (68)

Ulceration

Yes 2 (7)

No 26 (93)

Regression

Yes 2 (7)

No 26 (93)

Histological subtype

Superficial spreading 23 (82)

Nodular 5 (18)

Total 28 (100)

Table 3. The relative Analytical Sensitivities of the methods used in this study.

Each dilution was analyzed in duplicates by the semi-quantitative/quantitative assays (TaqMan, pyrosequencing, CADMA) and once by the qualitative assays (Sanger sequencing and the Cobas test).

Theoretica Sanger Pyrosequencing TaqMan CADMA Cobas I fraction sequencing (Values from the (Average ACt test of mutant pyrograms are values of the two

alleles* shown in the technical replicates

brackets) are shown in the

brackets)

50% Mutant Mutant (53%, 49%) Mutant (1 .53) Mutant/Mutant Mutant

40% Mutant Mutant (33%, 31 %) Mutant (1 .67) Mutant/Mutant Mutant

30% Mutant Mutant (21 , 20%) Mutant (2.44) Mutant/Mutant Mutant

20% Wild-type Mutant (13%, 13%) Mutant (3.31) Mutant/Mutant Mutant

10% Wild-type Mutant (7%, 7%) Mutant (4.75) Mutant/Mutant Wild-type

5% Wild-type Wild-type (5%, 5%) Mutant (5.64) Mutant/Mutant Wild-type 2.5% - - Mutant (7.32) Mutant/Mutant -

1 .25% - - Wild-type (9.12) Mutant/Mutant -

0.625% - - Wild-type (10.77) Mutant/Mutant -

0.3125% - - - Mutant/Mutant -

0.15625% - - - Mutant/Mutant -

0.078125% - - - Mutant/Mutant -

Theoretical fractions assuming no pipetting error and that the cell line is 100% monoclonal.

Table 4. Detection of the BRAF V600E mutation in clinical samples (n=28).

The samples are ordered according to fraction of tumor tissue. Detected mutations are indicated by the value 1 , whereas the value 0 indicates that the V600E mutation was not detected.

mutation detected. Mutation not found in subsequent runs.

Table 5. The Frequencies of the BRAF V600E mutation depend on the analytical sensitivity of the method and the fraction of tumor cells in the samples.

Sanger The Cobas Pyro-

CADMA TaqMan sequencing test sequencing

All samples 29% 36% 43% 46% 50%

Tumor content as a fraction of

all tissue

Samples <10% tumor 21 % 32% 42% 47% 53%

Samples >10% tumor 44% 44% 44% 44% 44%

Tumor content as a fraction of cell

dense tissue only

Samples <10% tumor 0% 14% 43% 57% 57%

Samples 10% - 49% tumor 36% 43% 43% 43% 50%

Samples >50% tumor 43% 43% 43% 43% 43%

Table 6. Details of the CADMA assays. Primer sequences are given in 5'

directions.

Mutation Cell Primers (introduced mutations are Primer Annealing line underlined) concentempetrations rature c.34 G>A A549 SEQ ID NO: 21 400 nM 57°C

Mutant primer forward:

G AATATAAACTTATG GTAGTTG GAG ATA

SEQ ID NO: 22 100 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 23 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.34 G>T NCI-H23 SEQ ID NO: 24 400 nM 60°C

Mutant primer forward:

GAATATAAACTTGTAGTAATTGGAGCTT

SEQ ID NO: 25 150 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 26 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.34 G>C PSN1 SEQ ID NO: 27 400 nM 60°C

Mutant primer forward:

GAATATAAACTTGTAGTAATTGGAGCTC

SEQ ID NO: 28 100 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 29 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.35 G>A LS174T SEQ ID NO: 30 400 nM 60°C

Mutant primer forward:

GAATATAAACTTGTGGTAATTGGAGATGA

SEQ ID NO: 31 150 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 32 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.35 G>T SW480 SEQ ID NO: 33 400 nM 64°C

Mutant primer forward:

GAATATAAACTTGTAGTAATTGGAGCTGT SEQ ID NO: 34 150 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 35 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.35 G>C RPMI82 SEQ ID NO: 36 400 nM 58°C

26 Mutant primer forward:

GAATATAAACTTGTAGTAATTGGAGCTGC

SEQ ID NO: 37 100 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 38 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

c.38 G>A DLD-1 SEQ ID NO: 39 400 nM 62°C

Mutant primer forward:

AAACTTGTGGTAGTTGGAGATGGTTA

SEQ ID NO: 40 150 nM

Non-mutant primer forward:

ATGACTGAATATAAACTTGTGGTAGTTG

SEQ ID NO: 41 400 nM

Common primer reverse:

ACTGTCAAGGCACTCTTGCCTAC

Table 7. Overview of the results from screening 100 mCRC samples using the TheraScreen kit and CADMA. A TaqMan based assay was used to test samples for which the TheraScreen kit and CADMA did not give the same result (Sample IDs highlighted with orange).

Sample ID TheraScreen result CADMA result TaqMan result Sample ID TheraScreen result CADMA result TaqMan result

1 W Id-type Wild-type Not performed 51 Wild-type Wild-type Not performed

2 W Id-type Wild-type Not performed 52 Wild-type Wild-type Not performed

3 w Id-type Wild-type Not performed 53 Wild-type Wild-type Not performed

4 w Id-type Wild-type Not performed 54 Wild-type Wild-type Not performed

5 w Id-type Wild-type Not performed 55 c.35 G>A c.35 G>A Not performed

6 w Id-type Wild-type Not performed 56 c.34 G>T c.34 G>T Not performed

7 w Id-type Wild-type Not performed 57 c.35 G>A c.35 G>A Not performed

8 w Id-type Wild-type Not performed 58 c.35 G>T c.35 G>T Not performed

9 w Id-type Wild-type Not performed 59 c.35 G>T c.35 G>T Not performed

10 w Id-type Wild-type Not performed 60 c.34 G>T c.34 G>T Not performed

11 w Id-type Wild-type Not performed 61 c.35 G>A c.35 G>A Not performed

12 w Id-type Wild-type Not performed 62 c.35 G>A c.35 G>A Not performed

13 w Id-type Wild-type Not performed 63 c.35 G>T c.35 G>T Not performed

14 w Id-type Wild-type Not performed 64 c.35 G>A c.35 G>A Not performed

15 w Id-type Wild-type Not performed 65 c.35 G>C/c.34 G>A c.35 G>C c.35 G>C

16 w Id-type No data Not performed 66 c.35 G>A c.35 G>A Not performed

17 w Id-type W Id-type Not performed 67 c.35 G>T c.35 G>T Not performed

18 w Id-type W Id-type Not performed 68 c.35 G>C c.35 G>C Not performed

19 w Id-type W Id-type Not performed 69 c.38 G>A c.38 G>A Not performed

20 w Id-type W Id-type Not performed 70 c.38 G>A c.38 G>A Not performed

21 w Id-type W Id-type Not performed 71 c.35 G>A c.35 G>A Not performed

22 w Id-type W Id-type Not performed 72 No data c.35 G>A Not performed

23 w Id-type W Id-type Not performed 73 c.35 G>A Wild-type Wild-type

24 w Id-type W Id-type Not performed 74 c.35 G>C c.35 G>C Not performed

25 w Id-type W Id-type Not performed 75 c.35 G>T c.35 G>T Not performed

26 w Id-type W Id-type Not performed 76 c.35 G>T c.35 G>T Not performed

27 w Id-type W Id-type Not performed 77 c.35 G>T c.35 G>T Not performed

28 w Id-type W Id-type Not performed 78 c.35 G>T c.35 G>T Not performed

29 w Id-type W Id-type Not performed 79 c.35 G>A c.35 G>A Not performed

30 w Id-type W Id-type Not performed 80 c.35 G>C c.35 G>C Not performed

31 w Id-type W Id-type Not performed 81 c.35 G>T c.35 G>T Not performed

32 w Id-type W Id-type Not performed 82 c.35 G>C c.35 G>C Not performed

33 w Id-type W Id-type Not performed 83 c.35 G>A c.35 G>A Not performed

34 w Id-type W Id-type Not performed 84 c.35 G>T c.35 G>T Not performed

35 w Id-type W Id-type Not performed 85 c.35 G>A c.35 G>A Not performed

36 w Id-type W Id-type Not performed 86 c.35 G>T c.35 G>T Not performed

37 w Id-type W Id-type Not performed 87 c.35 G>A c.35 G>A Not performed

38 No data W Id-type Not performed 88 c.38 G>A c.38 G>A Not performed

39 Wild-type W Id-type Not performed 89 c.34 G>T c.34 G>T Not performed

40 Wild-type W Id-type Not performed 90 c.34 G>A c.34 G>A Not performed

41 Wild-type W Id-type Not performed 91 c.35 G>A c.38 G>A c.38 G>A

42 Wild-type W Id-type Not performed 92 c.35 G>C c.35 G>C Not performed

43 Wild-type w Id-type Not performed 93 c.35 G>T c.35 G>T Not performed

44 Wild-type w Id-type Not performed 94 c.35 G>T c.35 G>T Not performed

45 Wild-type w Id-type Not performed 95 c.35 G>A/c.38 G>A c.35 G>A/c.38 G>A Not performed

46 Wild-type w Id-type Not performed 96 c.34 G>T c.35 G>A/c.34 G>T c.35 G>A/c.34 G>T

47 Wild-type w Id-type Not performed 97 c.35 G>T c.35 G>T Not performed

48 Wild-type w Id-type Not performed 98 c.35 G>T c.35 G>T Not performed

49 Wild-type w Id-type Not performed 99 c.38 G>A c.38 G>A Not performed

50 Wild-type w Id-type Not performed 100 c.38 G>A c.38 G>A Not performed SEQUENCES

SEQ ID NO: 1

Human KRAS gene

GenBank accession number NM_033360.2

Homo sapiens chromosome 12:25278036-25295130

The sequence according to SEQ ID NO: 1 is shown as sequence number 1 in the sequence listing.

SEQ ID NO: 2

human BRAF gene

GenBank accession number NM_004333.4

Homo sapiens chromosome 7: 140080282-140271033

The sequence according to SEQ ID NO: 2 is shown as sequence number 2 in the sequence listing.

SEQ ID NO: 3

human EGFR gene

GenBank accession number NM_005228.3

Homo sapiens chromosome 7:55054219-55238262

The sequence according to SEQ ID NO: 3 is shown as sequence number 3 in the sequence listing. SEQ ID NO: 4

Forward common primer used in the BRAF c.1799T>A assay

Aggtgattttggtctagctacag

The sequence according to SEQ ID NO: 4 is also shown as sequence number 4 in the sequence listing.

SEQ ID NO: 5

Reverse non-mutant primer used in the BRAF c.1799T>A assay

Tgatgggacccactccatcg

The sequence according to SEQ ID NO: 5 is also shown as sequence number 5 in the sequence listing. SEQ ID NO: 6

Reverse mutant primer used in the BRAF c.1799T>A assay

Tgggacccactctatcgagatttct

The sequence according to SEQ ID NO: 6 is also shown as sequence number 6 in the sequence listing.

SEQ ID NO: 7

Reverse common primer used in the EGFR c.2573T>G assay

Tctcttccgcacccagcagt

The sequence according to SEQ ID NO: 7 is also shown as sequence number 7 in the sequence listing.

SEQ ID NO: 8

Forward non-mutant primer used in the EGFR c.2573T>G assay

Cgcagcatgtcaagatcacagat

The sequence according to SEQ ID NO: 8 is also shown as sequence number 8 in the sequence listing. SEQ ID NO: 9

Forward mutant primer used in the EGFR c.2573T>G assay

Gcatgtcaagatcacagattttaggcg

The sequence according to SEQ ID NO: 9 is also shown as sequence number 9 in the sequence listing.

SEQ ID NO: 10

Reverse common primer used in the PIK3CA c.3140 A>G assay

catttttgttgtacagccatcatgac

The sequence according to SEQ ID NO: 10 is also shown as sequence number 10 in the sequence listing.

SEQ ID NO: 11

Forward non-mutant primer used in the PIK3CA c.3140 A>G assay

Ggagtatttcatgaaacaaatgaatgatg

The sequence according to SEQ ID NO: 1 1 is also shown as sequence number 1 1 in the sequence listing. SEQ ID NO: 12

Forward mutant primer used in the PIK3CA c.3140 A>G assay

Agtatttaatgaaacaaatgaatgatacacg

The sequence according to SEQ ID NO: 12 is also shown as sequence number 12 in the sequence listing.

SEQ ID NO: 13

Reverse common primer used in the KRAS c.35G>C assay

Tgtatcgtcaaggcactcttgcc

The sequence according to SEQ ID NO: 13 is also shown as sequence number 13 in the sequence listing.

SEQ ID NO: 14

Forward non-mutant primer used in the KRAS c.35G>C assay

Atgactgaatataaacttgtggtagttg

The sequence according to SEQ ID NO: 14 is also shown as sequence number 14 in the sequence listing. SEQ ID NO: 15

Forward mutant primer used in the KRAS c.35G>C assay

Gaatataaacttgtagtaattggagctgc

The sequence according to SEQ ID NO: 15 is also shown as sequence number 15 in the sequence listing.

SEQ ID NO: 16

Human PIK3CA gene

GenBank accession number NM_006218

Homo sapiens chromosome 3: 180349005-180435191

The sequence according to SEQ ID NO: 16 is shown as sequence number 15 in the sequence listing.

SEQ ID NO: 17 Forward primer used for BRAF pyrosequencing.

TTCATGAAGACCTCACAGTAAAAA

The sequence according to SEQ ID NO: 17 is also shown as sequence number 17 in the sequence listing.

SEQ ID NO: 18

Reverse primer used for BRAF pyrosequencing

GGCCAAAAATTTAATCAGTGGAA

The sequence according to SEQ ID NO: 18 is also shown as sequence number 18 in the sequence listing.

SEQ ID NO: 19 Sequencing primer used for BRAF pyrosequencing

GCCAGGTCTTGATGTACT

The sequence according to SEQ ID NO: 19 is also shown as sequence number 19 in the sequence listing. SEQ ID NO: 20

Reverse primer used for Sanger sequencing

GTTGAGACCTTCAATGACTTTCTAG

The sequence according to SEQ ID NO: 20 is also shown as sequence number 20 in the sequence listing.

SEQ ID NO: 21

Reverse mutant primer used in the BRAF c.1796T>A assay

TGAGACCCACTCTATCGAGATTTCT

The sequence according to SEQ ID NO: 21 is also shown as sequence number 20 in the sequence listing.