Classifications

• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
  • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
  • C12Q1/6813—Hybridisation assays
  • C12Q1/6827—Hybridisation assays for detection of mutation or polymorphism
• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
  • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
  • C12Q1/6813—Hybridisation assays
  • C12Q1/6832—Enhancement of hybridisation reaction

Definitions

• the present invention relates generally to methods for improving fidelity in nucleic acid hybridisation assays and PCR priming reactions.
• the invention provides methods for reducing mis-pairing in hybridisation between sample DNA and an oligonucleotide.
• the invention therefore finds application, inter alia, in micro-array DNA analysis, for example in DNA methylation analysis, and in the detection of point mutations and single nucleotide polymorphisms.
• Nucleic acid hybridisation is a fundamental tool in molecular genetics.
• the inherent ability of single stranded DNA to form a double helix with complementary DNA or RNA is the basis for all nucleic acid hybridisation assays.
• the accurate determination of complementarity between two nucleic acid strands relies on correct Watson-Crick base pairing between the strands, ie the inherent base pairing between adenine and thymine (A-T) nucleotides and guanine and cytosine (G-C) nucleotides, and is an important condition for the implementation of any nucleic acid hybridisation assay.
• A-T adenine and thymine
• G-C guanine and cytosine
• cytosine in the sample DNA sequence is represented by the presence of a thymine base in the amplified DNA sequence, whereas methylated cytosine in the sample DNA remains as cytosine in the amplified DNA.
• methylated cytosine in the sample DNA remains as cytosine in the amplified DNA.
• cytosine methylation is mostly found at CpG dinucleotides; however cytosines at other positions may be methylated, especially, for example, at CpNpG trinucleotides in plant DNA.
• accurate sequencing of methylation patterns is limited by the available technology and the need to amplify and sequence a large number of amplicons using conventional sequencing technology. A higher throughput technology with potential for multiplexing would significantly expand capacity in this fast growing experimental and clinical area.
• improvements to DNA sequencing chip technology to allow accurate reading and quantification of cytosine methylation is preferred to other sequencing methods .
• Such a chip preferentially uses pairs of probes for all sequences to be read.
• the pairs of probes are identical in sequence except for the base, C or T, which interrogates the methylation status of C at CpG sites in the original DNA sample. If the target binds to the probe containing C(pG), the C was methylated in the original sample; if the target binds to the probe containing T(pG), the C was unmethylated in the original DNA sample.
• the complements of these probe pairs, containing CpG and CpA at the interrogating sites may also be used.
• the present invention relates to methods for improving hybridisation fidelity, for example, on oligonucleotide arrays, and advantageous applications of these methods.
• Guanine to thymine (G-T) mis-pairing frequently occurs in target DNA hybridisation to probes in oligonucleotide arrays.
• the present inventors have surprisingly found that G- T mis-pairing may be controlled such that its incidence is reduced, resulting in increased fidelity of target DNA hybridisation to probe sequences.
• the present invention provides a method of reducing nucleotide mis-pairing in hybridisation between a probe oligonucleotide and a target DNA, the method comprising:
• the hybridisation may take place in solution or with one or both of the DNA and the at least one oligonucleotide probe attached to a solid support.
• the sample of step (b) may comprise bisulphite-treated DNA.
• the bisulphite-treated DNA may undergo a further step of amplification.
• the present invention provides a method of reducing nucleotide mis-pairing in hybridisation between a probe oligonucleotide and a target DNA, the method comprising:
• the hybridisation may take place in solution or with one or both of the DNA and the at least one oligonucleotide probe attached to a solid support.
• the sample of step (b) may comprise bisulphite-treated DNA.
• the present invention provides a method of increasing fidelity of hybridisation between a probe oligonucleotide and a target DNA, the method comprising: (a) providing at least one oligonucleotide probe of about 7 to about 25 nucleotides;
• the sample of step (b) may comprise bisulphite-treated DNA.
• the present invention provides a method of oligonucleotide array-based analysis of DNA, the method comprising: (a) providing a sample comprising target DNA;
• the method prior to step (b) also comprises treating the DNA of the sample with bisulphite to convert unmethylated cytosine bases to uracil.
• the present invention provides a method of oligonucleotide array-based analysis of DNA methylation, the method comprising:
• the oligonucleotide probe may entirely comprise standard DNA nucleotides or may comprise one or more modified nucleotides, such as nucleotides having modifications in the base and/or sugar and/or phosphate moieties.
• target molecules may entirely comprise standard DNA nucleotides or may comprise one or more modified nucleotides, such as nucleotides having modifications in the base and/or sugar and/or phosphate moieties.
• the modified nucleotides may be selected from 2'-O-methyl nucleotides, such as 2'-O-methyl adenosine to replace 2'-deoxy-adenosine, 2'-O-methyl-uridine or 2'-O- methyl-thymidine to replace 2'-deoxy-thymidine, and, 2-amino-adenosine to replace adenine.
• 2'-O-methyl nucleotides such as 2'-O-methyl adenosine to replace 2'-deoxy-adenosine, 2'-O-methyl-uridine or 2'-O- methyl-thymidine to replace 2'-deoxy-thymidine, and, 2-amino-adenosine to replace adenine.
• the oligonucleotide probe may be of any desired length from about 7 to about 25 nucleotides, in one embodiment, the oligonucleotide probe is about 7 nucleotides in length.
• the oligonucleotide probe maybe about 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 or 25 nucleotides long.
• the at least one oligonucleotide probe may optionally comprise one or more universal bases or a mixture of normal or modified nucleotides at one or both ends to increase the melting temperature of the oligonucleotide-target double-helices.
• Universal bases include 5-nitro-indole.
• the oligonucleotide probes may be immobilised on a solid support.
• Sample may comprise genomic DNA, cDNA or synthetic DNA.
• the DNA of the sample is subjected to a fragmentation step to generate fragments of up to about 100 nucleotides. Fragmentation of DNA may be achieved by any means such that fragments of up to about 100 base pairs are generated comprising target sequences.
• Such methods include DNAaseI digestion of DNA of the sample or the incorporation of uracil in amplified DNA of the sample followed by treatment with uracil-DNA glycosylase to create abasic sites at every uracil base incorporated, and either alkali cleavage or endonuclease VIII cleavage of abasic sites.
• endonuclease V may be used to introduce nicks at the second or third phosphodiester bond 3' to the uracil site.
• DNA fragments having up to about 100 nucleotides may also be prepared by amplification techniques or restriction enzyme digestion.
• the conditions suitable to enable hybridisation between probe and target DNA encourage the formation of short hybrids with probe melting temperature (Tm) independent of base sequence.
• Detection of target DNA hybridised to the oligonucleotide probe may be achieved through the detection of an optional detectable label associated with the DNA of the sample, or by other means such as a labelling dye that only binds duplex DNA or through the measurement of one or more physical properties indicative of the presence of duplex DNA.
• Suitable detectable labels include biotin, digoxygenin, radioactive labels, and fluorescent labels such as Cy3, Cy5, Quasar570 (QSR570), Quasar670 (QSR670), and Qdots, and other labels that function through a similar labelling mechanism.
• Labelling dyes that only bind duplex DNA include chromomycin, SybrGreen, and YOYO-I.
• the one or more physical properties that can be measured to detect duplex DNA include surface plasmon resonance.
• both the amplified DNA and the at least one oligonucleotide probe may comprise one or more modified nucleotides.
• the present invention provides a method of reducing nucleotide mis-pairing in hybridisation between a probe oligonucleotide and a target DNA, the method comprising: (a) providing at least one oligonucleotide probe of about 7 to about 25 nucleotides comprising one or more modified nucleotides;
• the sample of step (b) may comprise bisulphite-treated DNA.
• the bisulphite-treated DNA may undergo a further step of amplification.
• the present invention provides a method of reducing nucleotide mis-pairing in hybridisation between a probe oligonucleotide and a target DNA, the method comprising:
• the sample of step (b) may comprise bisulphite-treated DNA.
• the present invention provides a method of increasing fidelity of hybridisation between a probe oligonucleotide and a target DNA, the method comprising:
• the hybridisation may take place in solution or with one or both of the DNA and the at least one oligonucleotide probe attached to a solid support.
• the sample of step (b) may comprise bisulphite-treated DNA.
• the present invention provides a method of oligonucleotide array-based analysis of DNA, the method comprising:
• the method prior to step (b) also comprises treating the DNA of the sample with bisulphite to convert unmethylated cytosine bases to uracil.
• the present invention provides a method of oligonucleotide array-based analysis of DNAmethylation, the method comprising: (a) providing a sample comprising target DNA;
• the present invention provides a method, of oligonucleotide array-based analysis of DNA methylation, the method comprising:
• Figure 1 DNA sequence of single stranded LTM19, LTU19, and LTM55 short DNA targets. The sequence of LTM 19 within LTM55 is underlined.
• Figure 2 DNA sequence of one strand of double-stranded 85 base pair DNA target, LTM85. The sequence of LTMl 9 within LTM85 is underlined.
• Figure 3 Sequences of cloned DNA, LINE5, LINE8 and LINE9 used as target controls. The locations of the sequences of LTM19, within the clones LINE8 and LINE9, and LTUl 9, within the clone LINE5, are underlined.
• Figure 4 Fragmentation of 150 nucleotide DNA targets, (a) DNA target, internally labelled using d Cy5 CTP, was treated with DNAse 1 for 0, 1, 3 or 10 minutes, (b) DNA labelled internally with d Cy5 CTP and prepared with differing ratios of dTTP:dUTP [no dUTP (lane A), or with dTTP:dUTP ratios of 9:1 (lane B), 3:1 (lane C), or 1 : 1 (lane D)] were fragmented using UDG followed by alkali cleavage, (c) DNA labelled internally with d Cy5 CTP and prepared with differing ratios of 1 : 1 or 3 : 1 dTTP:dUTP and either dATP (A) or d(2-amino)ATP (2AA) as shown was fragmented with USER enzyme mix. Fragmented DNA samples were electrophoresed on 4% agarose gels and imaged with a phosphorimager.
• Figure 5 Long, 150-nt, single-stranded DNA targets bind weakly to short, 9-nt, DNA probes on the microarray, with strength depending on the methylation status of the target DNA.
• Hybridisation was overnight at 23°C, in 3M TMACl, 1OmM Tris-Cl pH8, ImM EDTA, 5% sarkosyl, after which slides were washed in hybridisation solution containing no target DNA, and scanned on a Phosphorimager. Each row bears five probes, each probe spotted as adjacent duplicates.
• the methylated target should bind to all five probes in the top row and one probe in the third row, first column, as indicated by ovals in (a), and the unmethylated target should bind to all five probes in the second row, and one probe in the last row, first column, as indicated by the ovals in (b).
• Methylated target shows almost no binding, while unmethylated target (LINE5) binds. All probes are 9mers of normal DNA.
• Figure 6 Length of DNA target affects properties of binding to oligonucleotide probes on chip.
• the targets represent methylated DNA controls, (a) Long target, 150-nt LINE9, does not bind to any probe, (b) Target of intermediate length, 85-nt LTM85, binds well to probes, forming G-C base pairs as expected (circled spots in Figure), but also forming strong G-T mispairs (all spots not circled), (c) Short target, 19-nt LTM19, shows correct binding in general to form G-C base pairs (circled spots); there are also numerous G-T mis-pairs (all spots not circled), but fewer than for the 85-nt target (compare the number of spots not circled in (c) and (b)). All probes are 9-mers of normal DNA.
• Figure 7 Mis-paired bases form between short DNA targets and DNA probes, with strength depending on base sequence.
• Hybridisation of target oligonucleotide LTM19 with array containing normal DNA probes was overnight at 23 0 C in 0.5M sodium phosphate, pH 7.2, 7% sarkosyl, with subsequent washing in 3x SSC for 5 minutes, and twice in 0.5x SSC for 5 minutes,
• Circled spots indicate correct binding of target LTMl 9 with normal DNA probes; numbers indicate the specific probe sequence, eg.
• Figure 8 Effect of shortening probe length on the strength of binding to target LTM85.
• Figure 9 Effect of lengthening a probe with a central section of 9 defined bases by adding a random mix of bases at each end.
• Fragmented, methylated target LINE9 was hybridised to the probes overnight at room temperature in 1.9M TMA-formate, 1OmM Tris-Cl, ImM EDTA, 0.002% Tween-20; the slides were washed in hybridisation solution (minus DNA) and scanned.
• the target binds to the probes LP 13Xn on the left, Watson-Crick base pairs are formed (a); when it binds to the probes LP 15Xn on the right, a G-T mis-pair is formed (b).
• Base sequences of probes. X represents an equal mix of the normal bases A 3 G 3 C 5 T.
• Figure 10 Effect of 2'-O-methyl-nucleotides in 9-nt probes on the specificity and strength of target binding, (a) Sequences of the probe sets, (b) Probes in column "DNA” have normal DNA 3 "OU” have all T replaced by 2'-O-methyl-U, "OA” have all A replaced by 2'-O-methyl-A, "OAU” have all A and all T replaced by 2'-0-methyl-A and 2'-O-methyl-U, respectively, and "OaIl” have all nucleotides replaced by 2'-O-methyl- nucleotides. Generic probe names are on the left of the image. Target was LTM85.
• Hybridisation solution was 2.0M TMA formate, 0.056M MES, 5% DMSO 3 2.5X Denhardt's solution, 5.8mM EDTA 3 0.0115% Tween-20. Circles indicate locations of probes to which the methylated target should bind with no mismatches.
• Figure 11 Effect of 2-amino-adenine and 2'-O-methyl-2-amino-adenine in 9-nt probes on the specificity and strength of target binding,
• Probes in column "DNA” have normal DNA 3 "D” have all A replaced by 2-amino-A
• “M” have all A replaced by 2'-O-methyl-2-amino-A.
• Generic probe names are on the left of the image.
• Target was LTM85.
• Hybridisation solution was 2.7M TMACl 3 0.056M MES, 5% DMSO 3 2.5X Denhardt's solution, 5.8mM EDTA, 0.0115% Tween-20.
• Circles indicate locations of probes to which the methylated target should bind with no mismatches
• Figure 12 Effect on the specificity and strength of target binding to probes containing nucleotides with 2'-O-methyl modifications in combination with universal bases at the termini.
• Probes in column "DNA" have 9-nt of normal DNA, "V2” have 9- nt of DNA with a universal base at each end, "V4" have two universal bases at each end, "OU” have all T replaced by 2'-O-methyl-U, "OA” have all A replaced by 2'-O-methyl- A, “OAU” have all A and all T replaced by 2'-O-methyl-A and 2'-O-methyl-U respectively, "0UV4" have all T replaced by 2'-O-methyl-U and two universal bases at each end, "OA V4" have all A replaced by 2'-O-methyl-A and two universal bases at each end, "0AUV4" have all A and all T replaced by 2'-O-methyl-A and 2'-O-methyl-U respectively and two universal bases at each end.
• the universal base was 5-nitro-indole.
• Target was LTM85.
• Hybridisation solution was 2.7M TMACl 3 0.056M MES 3 5%
• Figure 13 Effect of 5-propynyl-C and 5-propynyl-U in 9-nt probes on the specificity and strength of target binding.
• Probes in column "DNA” have normal DNA 5
• P have each interrogating C and T replaced by 5-propynyl-C and 5-propynyl-U, respectively, and "Pall” have all C and T replaced by 5-propynyl-C and 5-propynyl-U, respectively.
• Generic probe names are on the left of the image.
• Target was LTM85.
• Hybridisation solution was 2.7M TMACl, 0.056M MES, 5% DMSO, 2.5X Denhardt's solution, 5.8mM EDTA, 0.0115% Tween-20. Circles indicate locations of probes to which the methylated target should bind with no mismatches.
• Figure 14(a) Names of probes, in the order in which they appear, as adjacent pairs, on the DNA chip. Probes within a row have the same base sequence, but contain different modifications as indicated by the column headings (all probes in the "DNAX2" column have normal nucleotides; all probes in the "DX2” column have every A as 2- amino-A; all probes in the "0AX2” column have every A as 2'-O-methyl-A; all probes in the "OATX2” column have every A as 2'-O-methyl-A and the interrogating base T, if present, as 2'-O-methyl-T.
• Probes highlighted in yellow form correct base pairs with the fully methylated target LINES', and probes highlighted in green form correct base pairs with the fully unmethylated target LINE5. All probes have 11 nucleotides, with 9 central nucleotides of defined base sequence, and one nucleotide at each end being an equal mix ofA, G, C and T.
• Figure 14(b) Image of DNA chip after hybridisation with the fragmented methylated target LINE9.
• the generic names of probes in each row are on the left of the image, and the heading of each column indicates the type(s) of modified nucleotides in probes in that column. Circles indicate locations of probes which form correct base pairs with target LINE9, and which are highlighted in yellow in Figure 14(a). Spots not circled indicate target binding to probes through G-T mis-pairs.
• Figure 14(c) Averaged intensities for adjacent spot pairs shown in Figure 14(b), for binding by methylated target LINE9. The intensity of each spot was corrected for local background, and then background-corrected intensities for identical probes were averaged, to give the tabulated value for a particular probe, with error in parenthesis.
• Figure 14(e) Image of DNA chip after hybridisation with the unmethylated target LINE5.
• the generic names of probes in each row are on the left of the image, and the heading of each column indicates the type of modified nucleotides in probes in that column. Circles indicate locations of probes which form correct base pairs with target LINE5, and which are highlighted in green in Figure 14(a). Spots not circled indicate target binding to probes through A-C mis-pairs.
• Figure 14(f) Averaged intensities for adjacent spot pairs shown in Figure 14(e), for binding by unmethylated target LINE5. The intensity of each spot has been calculated as described in Figure 14(c). The probe order is the same as in Figure 14(a).
• Figure 14(g) Relative intensities reveal extent of mis-pairing by unmethylated control target, LINE5, to probes with various modifications. The relative intensity is the intensity for target binding to the probe relative to the intensity for target binding to the probe's partner with correct base pairing. The relative intensities were calculated from data in Figure 14(f).
• Figure 15 Improved specificity from incorporation of 2-amino adenine in target
• probes in the the upper rows contain normal DNA, those in the second rows (OA) contained 2'0-methyl adenosine at the sites shown in bold in the probe sequences and those in the third row (OAU) contain 2'O-methyl adenosines and also 2'O-methyl uridine at the interrogated base position.
• Figure 16 Nucleotide sequence of the human TPEF gene following bisulphite treatment. Target sites for primers are underlined. CpG sites are labelled A to I and shown in bold with cytosines that may be C or T after conversion shown as Y.
• Figure 17 (a) Layout of probes listed in Example 13 on arrays shown in Figure 17(b) and (c). Results of hybridisation with methylated target (b) and unmethylated target (c) are shown.
• Figure 18 Hybridisation of modified probes with TPEF target sites A, B and C. Probe sequences and array layout for sites A and B (a) and site C (c). Chemically modified bases are shown in bold and discriminating bases are underlined. Results of hybridisation with unmethylated and methylated target for sites A and B (b) and unmethylated and methylated target for site C (d) are shown.
• FIG. 19 Micro-array probes containing modified nucleotides improve the discrimination between matched normal and tumour samples in the analysis of DNA methylation of the highly-repeated sequence, LINE, taken from a human patient with colorectal cancer.
• LINE DNA extracted from normal tissue (left) and from tumour tissue (right) is bound to "DNA" probes (containing only normal DNA), "OU” probes (normal DNA with interrogating bases of normal C (upper) or 2'- O-methyl-U (lower)), "OA” probes (all A are 2'-O-methyl-A), “OAT” probes (all A are 2' ⁇ O ⁇ methyl-A, and interrogating bases are normal C (upper) or 2'-O-methyl-T (lower)).
• each panel the upper probes are specific for methylated target DNA, and lower probes are specific for unmethylated target DNA; each probe has been spotted on the micro-array as an adjacent pair.
• the number below each panel is the ratio of intensities (background-corrected) for target binding to the upper probes compared with the lower probes (intensities have been averaged over all probes of the same type spotted on the chip, including those appearing in the panel).
• the single number in bold italics below each pair of panels is the ratio of intensities of methylated:unmethylated (uppe ⁇ lower) probes for the normal sample compared with the tumour sample, and represents the discrimination achieved by each probe type in binding normal versus tumour DNA.
• Figure 20 Methylation of TPEF gene in clinical samples, (a) Gel electrophoresis of restriction enzyme digested (HpyCH4TV and BstU ⁇ ) DNA from normal (N) and tumour (T) tissue from five human patients following bisulphite treatment and PCR amplification, (b) Probe sequences and array layout for hybridisation. Chemically modified bases are shown in bold and discriminating bases are underlined, (c) Hybridisation of probes to patient DNA.
• an element means one element or more than one element.
• probe refers to an oligonucleotide.
• the probe may be attached to a solid support.
• solid support refers to any surface on which probe oligonucleotides and/or target DNA may be immobilised. Such surfaces include nitrocellulose or vinyl membranes, glass, plastic or silicon slides, microbeads and other similar surfaces known in the art.
• attachment encompasses both direct and indirect means of attachment of a nucleic acid molecule to the solid support. For example, indirect attachment may involve attaching, by any suitable means, the nucleic acid molecule to one or more linker molecules or compounds which in turn are bound to the surface of the solid support.
• DNA chip As used in the context of the present invention, the terms “DNA chip”, “DNA microarray”, “microarray” or “hybridisation array” refers to a solid support typically having oligonucleotide probes arrayed on its surface.
• the solid support utilised in the preparation of a chip or microarray is a nitrocellulose or nylon membrane, or a glass, plastic or silicon slide, or a bead.
• fidelity refers to accuracy in nucleotide base pairing.
• Accurate base pairing equates with Watson-Crick base pairing, ie preferential base pairing between standard 2'-deoxy adenosine and T- deoxy-thymidine nucleosides, and standard 2'-deoxy-cytidine and 2'-deoxy-guanosine nucleosides, and includes base pairing between standard and modified nucleosides and base pairing between modified nucleosides, where the modified nucleosides are capable of substituting for the appropriate standard nucleosides according to the Watson-Crick pairing.
• an increase in fidelity refers to an increase in Watson-Crick pairing over non- Watson-Crick pairing.
• oligonucleotide refers to a nucleic acid comprising deoxyribonucleotides, and optionally one or more modified nucleotides, and/or universal bases. Modified nucleotides include ribonucleotides.
• modified nucleotide or “modified base” refers to a nucleotide that differs in structure from the standard or “unmodified” nucleotides 2 '-deoxy- adenosine, 2'-deoxy-thymidine, 2'-deoxy-cytidine and 2'-deoxy-guanosine, and that is capable of pairing with an unmodified nucleotide or a modified nucleotide.
• Modified nucleotides include ribonucleotides.
• the term "target” refers to a DNA molecule that is added in hybridisation buffer for binding to the probes.
• the target is either internally labelled or end-labelled with a fluorophore or other detectable moiety.
• Sample may be known to contain target DNA, as applies when interrogating the methylation status of a target DNA, or, the sample may be 'searched' by an oligonucleotide array for target DNA sequence (as determined by probe sequence), such as in a gene expression array.
• sample refers to a biological sample that comprises DNA that in turn typically comprises, or is suspected to comprise, target DNA. DNA of the sample maybe subjected to amplification and/or fragmentation.
• methylated target refers to DNA of the sample and the target DNA it comprises that originates from methylated genomic DNA, enzymatically methylated DNA fragments or enzymatically methylated synthetic DNA, and that have undergone treatment with bisulphite, a reaction that converts unmethylated cytosines in the DNA to uracils but does not convert methylated cytosines. Following PCR amplification of the bisulphite treated DNA, methylated cytosines in the starting DNA appear as cytosines in the amplified DNA.
• methylated target also refers to chemically synthesised DNA with cytosines at sites representing 5-methyl-cytosine.
• unmethylated target refers to sample DNA and the target DNA it comprises originating from unmethylated genomic DNA, unmethylated DNA fragments or unmethylated synthetic DNA that has undergone treatment with bisulphite such that unmethylated cytosines in the DNA have been converted to uracils. Following PCR amplification of the bisulphite treated DNA, the uracils are replaced with thymines, and therefore unmethylated cytosines in the starting DNA are represented by thymines in. the amplified DNA. Unmethylated target also includes chemically synthesised DNA with thymine at sites representing unmethylated cytosine. As used in the context of the present invention, the term "target DNA region", or
• target region in addition to referring to a single target region may also refer to multiple, independent target regions. Thus, multiple, independent target regions may be amplified simultaneously.
• PCR may refer to linear, non-exponential amplification of DNA in addition to exponential amplification of DNA, where the person skilled in the art would recognise that either form of amplification is appropriate for the purpose of the invention.
• the present invention relates to methods for improving fidelity in hybridisation assays, and advantageous applications of these methods.
• methods in accordance with embodiments of the invention for improving hybridisation fidelity are shown to substantially improve oligonucleotide array-based sequencing of DNA methylation and allow the use of very short (7 to 11 base) oligonucleotides.
• the methods demonstrated herewith for improving hybridisation fidelity are equally applicable to any nucleic acid hybridisation assay, a reduction in nucleotide mis-pairing in hybridisation being a universally recognisable and desirable advantage for most, if not all, DNA hybridisation-based assays.
• methods of the invention can also be used in the detection of point mutations and single nucleotide polymorphisms (SNPs) and more generally in a variety of PCR-related applications.
• Methods carried out in accordance with the invention find application in the diagnosis of, or predictor of susceptibility to, a disease or condition in a subject, which disease or condition is characterised by or associated with a variant genetic sequence. It will also be appreciated that methods of the invention may be employed in liquid or solid phase assay systems. Thus the methods are equally applicable to assays conducted in solution, such as in standard PCR and related reactions and to array-based assays in which one or more of the nucleic acid molecules to be hybridised is attached to a solid support.
• the inventors have discovered a surprising relationship between the length of the target DNA hybridising to the probes and the fidelity of the hybridisation reaction. Reduction in the length of the target DNA is expected to increase the rate of hybridisation, but it has also been found that the extent of G-T mis-pairing decreases with decreasing length of the target DNA. G-T mis-pairing is also observed to be sequence dependent.
• a first aspect of the present invention provides a method of reducing nucleotide mis-pairing in hybridisation between a probe oligonucleotide and a target DNA, the method comprising:
• the conditions suitable to enable hybridisation between probe and target DNA encourage the formation of short.hybrids with probe melting temperature independent of base sequence. This maybe encouraged in the presence of TMA salts of singly-charged anions such as chloride, formate, and acetate.
• hybridisation and wash solutions comprise TMA-formate at a concentration of between 0.8M and 2.0M, or TMA-chloride at a concentration of between 1.7M and 2.3M. More preferably, TMA-formate is present at 1.7M, or TMA-chloride is present at 2.0M.
• Hybridisation and wash solutions may further comprise one or more additional reagents.
• additional reagents include Tween 20 and SDS.
• modified nucleotides may be present in either the target DNA or the oligonucleotide probe, or both.
• Suitable modified nucleotides include, but are not limited to, 2'-O-methyl nucleotides, such as 2'-O-methyl adenosine, 2 ?
• -O-methyl-uridine 2'-O-methyl-thymidine, 2'-O-methyl-cytidine, 2'-O-methyl-guanosine, 2'-O-methyl-2- amino-adenosine; 2-amino-adenosine, 2-amino-purine, inosine; propynyl nucleotides such as 5-propynyl uridine and 5-propynyl cytidine; 2-thio-thymidine; universal bases such as 5-nitro-indole; locked nucleic acid (LNA), and peptide nucleic acid (PNA).
• LNA locked nucleic acid
• PNA peptide nucleic acid
• An exemplary application of the present invention relates to the analysis of DNA methylation status, often used in the diagnosis of, or determination of susceptibility to, disease.
• Methylation of cytosine bases at CpG sites in genomic DNA can be detected by treating the DNA with bisulphite, a reaction that converts unmethylated cytosines to uracils but does not change methylated cytosines. After amplifying the bisulphite-treated DNA using PCR, the uracils are converted to thymines, and the methylated cytosines become normal cytosines (Frommer, et al. 1992; Clark, et al. 1994).
• methylation status of genomic DNA can be determined by sequencing the DNA after bisulphite/PCR treatment.
• oligonucleotide microarrays for determining the methylation status of cytosines in an amplified PCR product, it is necessary to be able to clearly distinguish hybridisation of methylated and unmethylated target DNAs to oligonucleotide probes of the type nnniiCGnnn and nnnnTGnnn where n are variable nucleotides; these short probes may contain one or more CG and/or TG dinucleotide steps.
• Shorter probes have two advantages - firstly, the relative effect of single base mismatches on hybrid stability increases with decreasing length of the probe and secondly, the number of probes needed to include all possible sequence variations decreases substantially as probe length decreases. Conversely, hybrids with shorter probes are of lower stability and their stability is more strongly influenced by base composition.
• oligonucleotides containing 9 specified bases if the central base is designated as the cytosine to be interrogated (i.e. all oligonucleotides are of the type nnnCGnnn or nnnnTGnnn) all possible sequence variants, including multiple CpG sites, are contained in a set of 9374 oligonucleotides.
• oligonucleotide probes are needed to cover all combinations of CpG sites that could arise from bisulphite-treated DNA. Similar oligonucleotide sets could be prepared to interrogate other contexts of cytosine methylation, such as CpNpG sites in plant DNA. Alternatively, methylation sequencing chips with fewer probes may be used to interrogate defined sequences of interest, such as for diagnostic purposes.
• Target DNAs for hybridisation to arrayed probes were either synthesised as oligonucleotides representing sequences that would be obtained after bisulphite conversion of DNA or were amplified by PCR from bisulphite treated DNA or clones of PCR amplicons. All oligonucleotides were purchased from Geneworks Pty Ltd (Adelaide, South Australia). DNA was bisulphite treated using standard methods (Clark et al., 1994) or a kit either from Qiagen (Epitect) or from Human Genetic Signatures (MethylEasy), and following the manufacturer's instructions. 1.1. Preparation of target DNA molecules with defined DNA sequences.
• Methylated and unmethylated target DNA molecules with various numbers of nucleotides and of known sequence were prepared as controls to test the effect of target length on the binding of targets to the microarray.
• Short DNA targets having 19 or 55 nucleotides Short, single-stranded, DNA target molecules, containing 19 nucleotides, were synthesised with a fluorophore, QSR570 (similar characteristics to Cy3) or QSR670 (similar to Cy5), on the 3' end, and a single-stranded DNA target molecule of 55 nucleotides was synthesised with Cy5 on the 5' end.
• LTM19 and LTM55 represented the methylated form of the target of interest, and LTU 19 the unmethylated form ( Figure 1).
• the underlined bases in LTM55 are the same bases as in the 19-mer methylated target, LTMl 9.
• a double-helix of 85 base pairs was made by annealing two partially overlapping synthetic oligonucleotides (a 45-nucleotide molecule and a 55-nucleotide molecule 5'- end labelled with Cy5 (LTM55)) and filling with the Klenow fragment of DNA polymerase I.
• the Cy5 -labelled strand produced by this method, LTM85 is a longer version of LTM55 and includes the same bases as LTM19 (underlined, see Figure 2).
• the complementary strand is not labelled.
• Standard PCR in a 25 ⁇ l volume contained IX Taq DNA polymerase buffer (GoTaq, Promega), with 2mM MgCl 2 , 0.3 ⁇ M forward primer F3, 0.3 ⁇ M reverse primer R3, 200 ⁇ M each of dGTP, dATP, dTTP and dCTP and 1 unit of Taq polymerase.
• the mixture was heated at 95°C for 1 minute, and then subjected to 30 cycles of 95°C for 10 seconds, 50°C for 20 seconds, 72°C for 60 seconds; finally the mixture was held at 72°C for 2 minutes.
• Primers F3 and R3 have the sequences 5' GGGRGAGGAGTTAAGATGGT and 5' CCCACTATCTAACACTCCCTAATA respectively.
• the PCR amplicon (from either Sssl or whole-genome-amplification) was cloned into the pGEM T-easy vector (Promega). Inserts of the desired sequence (typically all CpG or all TpG/CpA) were kept as target controls for further work using DNA chips.
• Figure 3 shows the base sequences of cloned targets LINE5, LINE8 and LINE9.
• Underlined bases in LINE5 are the same as those in the unmethylated short target, LTU19.
• Underlined bases in LINE8 and LINE9 are the same as those in the methylated short target, LTMl 9.
• Cy5- or Cy3-containing long-target controls were each amplified from a plasmid DNA template of desired sequence (either Sssl or whole-genome-amplification, as described above, diluted 1/50) in 25 ⁇ l under standard PCR conditions except that 200 ⁇ M dCTP was substituted with a mix of lO ⁇ M dCTP and lO ⁇ M d Cy5 CTP (in a molar ratio of 1:1). Each completed PCR was checked by electrophoresing l ⁇ L on a 2.5% agarose mini-gel.
• the PCR mixture contained lO ⁇ M dCTP and lO ⁇ M d biotin CTP (in a molar ratio of 1 : 1), for subsequent detection of the target (after the hybridisation step) using streptavidin conjugates.
• PCR amplifications were performed in duplicate or quadruplicate (to make more material), and the identical reactions were pooled at this stage.
• the completed PCR could be purified (to remove Taq polymerase, buffer, unreacted primers and dNTPs) using a silica-based spin column (Rapid PCR Purification System, Marligen Biosciences).
• this extra step was not essential, and may generally be omitted to save time and effort.
• Each completed (and unpurified) PCR reaction contains sufficient labelled target DNA for two to four hybridisation experiments.
• Example 2 Fragmentation of single-stranded and double-stranded DNAs for targets having normal or modified nucleotides.
• Double-stranded, internally labelled, DNA products of 150 bp were prepared by PCR (1.1.3.1) and products were purified over Marligen columns to remove unreacted dNTPs and excess primers.
• the purified products in lOO ⁇ l were adjusted to 0.5X NEB buffer 2 (New England Biolabs) containing 5mM MgCl 2 .
• About 0.5 units of DNAaseI (Affymetrix) were added, and the DNA was digested for 1, 3 or 10 minutes. The reaction was stopped in excess EDTA. 2 ⁇ l from each reaction was analysed on a 4% agarose/Nu- sieve gel to determine the sizes of the fragmented DNA (Figure 4a).
• the solution was then neutralised to pH 6 using an equal amount of KOAc, pH 5.
• the sizes of the fragmented targets produced for various T:U ratios are shown in Figure 4(b).
• the uncleaved target, shown in lane A is 150-nt. Increasing the amount of dUTP, relative to dTTP, produced shorter targets.
• dATP was substituted with d(2- amino-adenine)TP. Pooled, duplicated reactions were purified over Marligen columns. Mixtures were then heated at 95°C for 1 minute, cooled briefly to room temperature, and 1 unit of USER enzyme was added (NewEngland Biolabs).
• USER enzyme a mixture of UDG (Uracil DNA glycosylase) and Endonuclease VIII, a DNA glycosylase-lyase
• UDG Uracil DNA glycosylase
• Endonuclease VIII a DNA glycosylase-lyase
• target fragment lengths may be controlled as desired by adjusting the T:U ratio (eg the target DNA may be fragmented to shorter lengths by lowering the T:U ratio from 3:1 to 1:1).
• endonuclease V may be used to fragment uracil-containing DNA molecules. Endonuclease V nicks DNA containing uracil at the second or third phosphodiester bond 3' to the uracil site (Miyazaki, K. (2002).
• Pairs of probes were designed for binding with Watson-Crick base pairing to sites in the target DNA which contained CpG! Where the C in CpG sites is methylated, the target will retain the sequence CpG after bisulphite treatment and PCR. Where the C in CpG sites is unmethylated, the target will have the sequence TpG after bisulphite treatment and PCR (with CpA in the complementary strand).
• the pairs of probes have identical base sequences except for the interrogating bases G or A (for interrogating CpG or TpG, respectively, in the target) or, alternatively, they have the interrogating bases C or T (for interrogating CpG or CpA in the complementary strand of the target). Most probe pairs were tested with C and T as the interrogating bases.
• probe pairs are required to determine the methylation status of closely- spaced CpG sites.
• sequence 5 'NNCGNCGNN in a target DNA requires the four probes, 5 'NNCGNCGNN, 5 'NNTGNTGNN, 5 'NNTGNCGNN and 5 'NNCGNTGNN, to determine the status of cytosine methylation in these closely spaced sites.
• Oligonucleotides were synthesised by Gene Works Pty Ltd (Adelaide, South Australia). All probes contained an amine linker on one end, mostly the 3' end. The concentration of each probe was calculated by measuring the absorbance at
• each oligonucleotide was analysed by recording the absorbance spectrum between 220 and 300nm.
• AU oligonucleotides with 3' amine linkers were also labelled on their 5' ends with P 32 -phosphate, electrophoresed on 20% polyacrylamide gels containing 7M urea, and scanned with a Phosphorlmager (Molecular Dynamics); a probe was subsequently gel-purified if this analysis showed that shorter fragments were present in significant amounts (>20%) together with the desired full-length molecule.
• Probe solutions were prepared for printing on Codelink slides typically as 25 ⁇ M solutions in 5OmM sodium phosphate buffer, pH 8.5.
• Codelink slides (Amersham Biosciences) consist of standard glass microscope slides coated with a hydrophilic polymer containing N-hydroxysuccinimide ester reactive groups. Probes were arrayed on the slides using a Floating Pin Replicator (VP478A), and a Scientific Glass Slide Indexing System (VP470), from V&P Scientific. The oligonucleotide probes were bound through their amine linkers to activated Codelink slides following the manufacturer's instructions. Areas on the slide not printed with oligonucleotide probes were subsequently blocked with ethanolamine. Slides arrayed with the oligonucleotide probes were stored in a desiccator until ready for use.
• hybridisation solution consisted of 2M TMACl (tetra-methyl-ammonium- chloride, Sigma) or 1.7M TMA-Formate (tetra-methyl-ammonium-formate, Sigma), 5OmM Tris-HCl pH 7.6 (Sigma), 5mM EDTA (Sigma), 0.01% Tween-20 (Pierce), 0.05% SDS (Sigma), and 10OnM short synthetic DNA, or 25 ⁇ L target DNA solution prepared as described above in Example 2.
• TMA concentrations were chosen for a given TMA salt to give approximately equal Tms for target binding to probes with a range of base sequences.
• any TMA salt with a singly charged anion is suitable, in particular, chloride, formate or acetate.
• 50-100 ⁇ L of hybridisation buffer containing target DNA was placed on a DNA chip, and covered with a glass cover slip. The slide was placed in a sealed box, humidified by paper soaked in water, and left to sit at room temperature for 15-30 minutes (or longer, 1 hr to overnight, if convenient). The slide was then washed briefly at room temperature three times with 150 ⁇ L hybridisation buffer containing no target.
• the target was labelled with R-phyco after the hybridisation and washing steps.
• 50 ⁇ l of Ix hybridisation solution containing 0.1 ⁇ g Streptavidin R-phycoerythrin conjugate (SAPE, Molecular Probes (Invitrogen)) was squirted onto the chip, covered with a clean glass slip, and left to sit for 15 minutes at 2O 0 C. The cover slip was removed, and unbound material was washed off three times with 150 ul of Ix hybridisation solution. Fluorescence on the slide was read on a Phosphorimager using a green 532 nm laser and TAMRA 580 nm filter.
• oligonucleotide probes may be re-used by rinsing briefly under distilled water at 20°C, and then soaking in distilled water at 65°C for 10 minutes (twice optionally), and air drying.
• the length of a normal DNA target influences its extent and specificity of binding to oligonucleotide probes on a chip.
• Target DNAs of differing lengths were tested to determine whether target length influenced binding to the microarrays.
• the targets represent methylated DNA.
• Hybridisation of targets to probes was overnight at 23°C in a solution containing 2.7M TMACl, 0.056M MES, 5% DMSO, 2.5X Denhardt's solution, 5.8mM EDTA, 0.0115% Tween-20. The slides were washed in hybridisation solution without target, and scanned. All probes were 9-mers of normal DNA.
• a 150-nt long methylated DNA target, LINE9 showed no binding to the 9-nt probes on the chip ( Figure 6(a)).
• LINE9 was prepared from methylated genomic DNA as described in Example 7.1; it is a different clone from LINE8 which was used in Example 7.1, but it has the same sequence as LINE8 in the region which binds to the probes printed on the DNA chip. Thus, poor binding of long 150-nt methylated DNA targets to the DNA probes was confirmed.
• Probes LPl and LP3 each contain the same single interrogating site stepped by 2 bases; LTMl 9 binds strongly to the correct probes (circled) forming a G-C base pair, and weakly (with a G-T mis-pair) to the paired probes LP2 and LP4 (immediately underneath the circled probes). Sequences LP5, LP9, LP13 and LP17 each have two interrogated sites and hence require a total of four probes each to determine their methylation states.
• LTM19 will form two G-C base pairs at the interrogated sites with probes in the top line (circled) in Figure 7(a), two G-T mis-pairs with probes in the second line (under 5, 9, 13, and 17), one G-C base pair and one G-T mis-pair with probes in the third line, and one G-T mis-pair and one G-C base pair with probes in the fourth line (where the locations of the G-C and G-T are swapped compared with the third line).
• the relative intensities of the uncircled spots for the probe sets LP5, LP9, LP 13, and LP 17 show that G-T mis-pairing with methylated target LTM 19 is strongest for sequences LP5 and LP9, less strong for sequence LPl 3, and weak for sequence LP 17, indicating that sequences flanking the interrogated sites influence the strength of the G-T mis-pair.
• the interrogated sites in the LP 13 and LP 17 sets of probes are the same, except that their position in the probe is displaced by two bases. Likewise the probe pairs LP1/LP2, LP3/LP4 and LP5/LP7 all interrogate the same CpG site.
• G-T mis-pairing occurs in a sequence and position-dependent manner for short targets binding to short probes, with both target and probe consisting of normal DNA, and the extent of G-T mis-pairing increases as the target lengthens.
• Example 8 Investigating the effect of various design parameters for normal DNA probes on the strength and specificity of target binding. A number of parameters were investigated to determine the optimal design of the oligonucleotide probes in order to achieve strong and highly specific binding by the target DNA, with Tms approximately equalised over probes of widely varying base sequence.
• the Tm of each short probe may be increased by effectively lengthening the probe with "undefined" bases. This may be done in a number of ways. Li one method, we have made a series of probes in which the base sequence of the central 9 nucleotides are defined, and have added at each end either 1, 2, 3 or 4 nucleotides taken from an equal mix of the four normal bases, A, G, C or T; this effectively makes probes with 11, 13, 15 and 17 nucleotides, respectively.
• Probes LP 13 and LP 15 were chosen as the base probes for the series because of the significant level of mis-pairing of the 9-mer probes.
• Methylated target LINE9 prepared and fragmented as described in Example 2.2, was hybridised to the probes, the slides washed in hybridisation solution at room temperature, and scanned.
• Figure 9(a) shows correct binding of the fragmented, methylated target to the series of probes, LP 13Xn, of increasing length, where a G-C base-pair is formed at the interrogated site.Increasing the probe length from 11 (X2) to 17 (X8) nucleotides does not appear to substantially affect the amount of target bound.
• Figure 9(b) shows binding of the target to the series of probes LP 15Xn, of increasing length, in which a single, central, G-T mispair is formed. Optimal specificity of binding is observed for the LP13X2/LP15X2 probe pair with hybridisation and washing at room temperature.
• Such 1 lmer probes comprising central defined 9 bases flanked by one random nucleotide at each end were found to work well under a variety of conditions for various probe sequences (data not shown).
• Target was found to bind with equal preference to probe connected to the chip through either the 5' end or the 3' end (data not shown), and switching the target sequence to the probe and probe sequence to the target had no effect on target binding (data not shown).
• Example 9 Investigating the effect of modified nucleotides in probes on increasing the strength and specificity of target binding.
• probes in the same line have the same base sequence but have modifications introduced as indicated by the label at the top of each column.
• the probe in line LPl and column DNA consists of normal DNA and has the base sequence 5' TTTTAGCGT, while the probe in line LPl and column OU is probe LPlOU which has the same sequence as LPl but with all T replaced by 2'-O-methyl-U, 5' UUUUAGCGU.
• Positions of probes which should bind correctly to the methylated target LTM85 to form G-C pair(s) at the interrogated site(s) are circled. Spots appearing in other places indicate G-T mis-pairing on target binding. Each probe has been spotted as an adjacent pair.
• the Tms for correct binding of target to probes were affected for modifications of the types "OU” and "OAU” to varying degrees depending on the sequence.
• the Tm for probe LPlOU (with five 2'-0-methyl-U, including four consecutive Us) was greatly reduced relative to LPl with normal DNA, while the Tm for probe LP30U (with three 2.'-O-methyl-U, including two consecutive Us) was reduced to a slightly lesser extent compared with LP3.
• probes "OA”, with all A replaced by 2'-O- methyl-A produced the desired effect of improving the specificity of binding while not affecting Tm.
• the replacement of A by 2'-O-methyl-A in the 9-nt probes conferred correct binding with a target of 85-nt, while probes with normal DNA mis- paired with the same target.
• longer targets may be used on DNA chips containing probes with 2'-0-methyl-A, than on DNA chips containing probes with normal DNA.
• Probe LPlD, LP3D and LP17D have, respectively, one, two and three 2-amino- A nucleotides.
• the relative increase in Tm for the "D" probes compared with their equivalent unmodified "DNA” probes was greater for LP 17 and LPl than for LP3 (compare spot intensities for LP17D with LP 17, LPlD with LPl, and LP3D with LP3, in Figure 1 l(a)).
• an increase in Tm is dependent either on the sequence flanking the 2-amino-A and/or on the location of the 2-amino-A in the probe (in LP17D there are 2- amino-A at each end of the probe) more than on the number of 2-amino-As in the probe.
• AIl three probe sets containing 2-amino-A also showed reduced G-T mis-pairing with the 85-nt target, compared with all-DNA probes (in Figure 1 l(a), compare spots not circled in column "D” with spots not circled in column "DNA” on the same row).
• probes with A replaced by 2-amino-A show both increased binding strength and increased discrimination with normal DNA targets, compared with probes of normal DNA.
• a universal base is a moiety that can bind equally well to all four bases of DNA.
• 5-nitro-indole was selected to test if a universal base placed at each end of a probe could effectively increase the Tm of a probe, while preserving the number of the defined bases.
• Example 8.3 5-nitro-indole was shown to be effective in increasing the Tm of probes with normal DNA, but without affecting specificity of target binding.
• 5-nitro- indole is used in combination with 2 '-O-methyl-nucleotides to test if both Tm and specificity for probes may increase.
• DNA probes with normal DNA nucleotides LP3 5' TTAGCGTGA-NH 2 3'
• V2 probes with normal DNA nucleotides, and one universal base at both the 5' and 3' ends, indicated by V. LP3V2 5' VTTAGCGTGAV-NH 2 3'
• V4 probes with normal DNA nucleotides, and two universal bases at both the 5' and 3' ends, indicated by VV. LP3V4 5' VVTTAGCGTGAVV-NH 2 3'
• UV4 probes with two universal bases at both the 5' and 3' ends, indicated by VV, and in which all T are replaced by 2'-O-methyl-U, as indicated by U LP13OUV4 5' VVUGAGCGACGVV-NH ? 3' LP 15OUV4 5 ' WUGAGUGACGVV-NH 7 3 '
• 0AV4 probes with two universal bases at both the 5' and 3' ends, indicated by VV, and in which all A are replaced by 2'-O-methyl-A, as indicated by A LP13OAV4 5' VVTGAGCGACGVV-NH 2 3' LP 15OAV4 5 ' VVTGAGTGACGVV-NH 2 3 '
• OAUV4 probes with two universal bases at both the 5' and 3' ends, indicated by VV, and in which all A and all U are replaced by 2'-O-methyl-A and 2'-O-methyl-U, as indicated by A and U, respectively.
• a set of probes (“Pall") was made in which all C and T were replaced by 5-propynyl-C and 5-propynyl-U, respectively, and a second set (“P") contained 5-propynyl-C and 5-propynyl-U only at the interrogation sites.
• the effect of target LTM85 binding to these probes with modified nucleotides was compared with the equivalent normal DNA probes. All probes had 9 nucleotides.
• P probes in which the interrogating C and T are replaced by 5-propynyl-C and 5- propynyl-U, as indicated by C and U, respectively LP3P 5' TTAGCGTGA-NH2 3' LP4P 5' TTAGUGTGA-NH2 3' LP 17P 5 ' AGCGACGT A-NH2 3 '
• 5-propynyl groups on the interrogating bases of probes can assist the fidelity of binding by the target in a sequence dependent manner, but inclusion in all pyrimidine bases of the probe is disadvantageous.
• Example 10 Quantification of intensities on the DNA chips
• the human eye and brain can very effectively integrate the information contained in images of microarrays, as shown in Example 9.
• the intensities of spots in the images have been quantified and presented in tables for numerical scrutiny.
• Fragmented methylated target LINE9 and unmethylated target LINE5 were hybridised for 2 hours at room temperature to probes on the DNA chip in 1.7M TMA-formate, 5OmM Tris-HCl, pH 7.5, 5mM EDTA, 0.02% Tween-20, 0.04% SDS, and washed twice at room temperature in the same solution (without targets). The slides were scanned on a Phosphorlmager.
• the amount of target bound to each probe was determined as follows. The raw fluorescence at each probe spot was measured on the Phosphorlmager by recording the intensity within a circle drawn tightly around the spot. The background was measured at four positions adjacent to the probe spot. The background counts were averaged, and subtracted from the raw intensity. The background-corrected intensities for identical probes were averaged, and these are reported in Figures 14(c) and (f). Standard deviations are given in parentheses. The arrangement in which the probes were spotted on the DNA chip is indicated in Figure 14(a). Probes LP3 and LP4 interrogate the methylation status of a single CpG site, where the methylated target should bind to LP3 and the unmethylated target should bind to LP4.
• LP13, LP14, LP15 and LP16 interrogate the methylation status of two closely-spaced CpG sites, where the fully methylated target should bind to LP 13, the fully unmethylated target should bind to LP 14, and the target in which one CpG is methylated and the other not should bind sequence-specifically to LP 15 or, for the alternative combination, to LP 16.
• probes LP21, LP22, LP23 and LP24 interrogate two closely-spaced CpG sites, with LP21 being specific for the fully methylated target, and LP22 for the fully unmethylated target.
• LP61 and LP62 interrogate a single CpG site, where the methylated target should bind to LP61 and the unmethylated target to LP62.
• the probes which bind fully methylated targets are colour-coded yellow, and those which bind fully unmethylated targets are colour-coded green, in Figure 14(a).
• Probe sequences in which X is an equal mix of A, G, C and T; D is 2-amino-A; A is T- O-methyl-A; T is 2'-O-methyl-T:
• LINE9 the methylated target, LINE9
• LP3 generic name forming a central G-C base pair.
• all LP3 probes LP3X2, LP3DX2, LP3OAX2, and LP3OAX2 repeated
• the averaged intensities for each of these circled pairs of spots, corrected for background, are given in the top row of Figure 14(c).
• Probe LP4 has the same sequence as LP3 except for a central base change from
• any binding by the fully methylated control target, LINE9, to any probe LP4 indicates the extent of G-T mis-pairing for this LP3/LP4 probe pair.
• fragmented LINE9 binds strongly to LP4 when it contains normal DNA nucleotides (first column), indicating a strong G-T mis-pair is formed.
• the intensity for target LINE9 incorrectly binding the normal DNA probe LP4X2 is 372 (first column), for incorrectly binding LP4DX2 is 101, for LP4OAX2 is 50, and for LP4OAT1X2 is 8.
• the intensity of each incorrectly bound LP4 probe, relative to the intensity of the correctly bound LP3 probe with the same modifications, is given in the second row of Figure 14(d).
• Intensities in the fourth (LP 14), fifth (LP 15) and sixth (LP 16) rows of Figure 14(c) are compared to those in the third row (LP 13) to calculate the relative intensities of the G-T mis-pairs given in Figure 14(d).
• probes of the type 0AX2 all A are 2'-O- methyl- A
• O ATX2 all A are 2 ' -O-methyl- A with interrogated T as 2 ' -O-methyl-T
• the seventh row in each of Figures 14(b), (c) and (d) shows correct binding of methylated target LINE9 to probes with the generic name and sequence of LP21.
• the set of probes LP21/LP22/LP23/LP24 interrogates two closely-spaced CpG sites, and rows eight, nine and ten in Figure 14(d) indicate the level of mis-pairing by LINE9 to probes LP22, LP23 and LP24, relative to correct binding to probe LP22.
• all probes with modified nucleotides DX2, 0AX2, and OATX2
• DX2, 0AX2, and OATX2 modified nucleotides
• Unmethylated target, LINE5 The unmethylated target, LINE5, should bind to probes LP4, forming a central A-T base pair. As shown by the circled spots in the second line of Figure 14(e), all LP4 probes (LP4X2, LP4DX2, LP4OAX2 and LP4OAT1X2) are bound by the unmethylated target, and there is minimal mis-pairing to analogous probes (LP3) in the row immediately above; in this case any mis-pairs of LINE5 with LP3 probes would be of the A-C type.
• LP3 analogous probes
• the averaged intensities for each LP3 and LP4 probes are given in the first and second rows, respectively, of Figure 14(f), and the first row of Figure 14(g) shows the intensities of the LP3 probes relative to those of their matched LP4 probes.
• the extent of A-C mis-pairing by the unmethylated target to LP3 probes, relative to correct binding to the LP4 probes, is less than 5% for all probes.
• the set of probes LP 13/LP 14/LP 15/LP 16 interrogates two closely spaced CpG sites, and correct binding by the unmethylated target is to the LP 14 probes (in the fourth rows of Figures 14(e) and (f)).
• LP 14 probes in the fourth rows of Figures 14(e) and (f)
• incorrect binding of the unmethylated target LINE5 to probes LP 13, LP 15 and LP 16 is given in rows three (immediately above LP 14), five and six (the two rows below LP 14) in Figures 14(e), (f) and (g).
• the quantified data in Figure 14(g) support the visual results in Figure 14(e) indicating that probes of the type 0AX2 and OATX2 confer fidelity of binding (for this set all probes, including normal DNA, show less than 3% mis-pairing) by unmethylated targets to probes bearing the sequence of this set. Similarly, the intensities of binding by unmethylated target LINE5 to probes
• LP22 (row eight) with A-C mis-pairing to probes LP21, LP23, and LP24 (rows seven, nine and ten), and correct binding to probes LP62 (bottom row) with mis-pairing to LP61 (second bottom row) can be seen in Figures 14(e) and (f), with relative bindings in Figure 14(g).
• all probes confer good discrimination, but the Tm for correct binding by LINE5 to LP22OAT2X2 is drastically reduced when the interrogating T are 2'-O-methyl-T (compare intensity for LP22OAT2X2 with LP22OAX2 in Figures 14(e) or (f)).
• probes containing either normal DNA, or all A as 2-amino-A, or all A as 2'-O-methyl-A confer good discrimination and strength in binding to the unmethylated target LINE5.
• Probes with all A as 2'-O-metliyl-A are preferred for probes with sequences similar to LP22.
• Example 11 Solution melting temperatures of duplexes containing modified nucleotides 11-mer or 13-mer DNA target oligonucleotides corresponding to 11 or 13-mer probe DNAs were annealed (at 2 ⁇ M each) and the melting temperatures of duplexes determined by following SYBR Green fluorescence with increasing temperature in an ABI7700 real time PCR machine.
• Melting temperatures were measured in two buffers: "NaCl” buffer contained 0.1 M NaCl, 40 mM Tris, 10 mM EDTA, pH 8, 0.02% Tween 20, 0.04% SDS and 0.5 ⁇ M SYBR Green; "TMACl” contained 2.0 M tetramethylammonium chloride, 40 mM Tris, 10 mM EDTA, pH 8, 0.02% Tween 20, 0.04% SDS and 32 ⁇ M SYBR Green (a higher concentration of SYBR Green being required under the high salt conditions).
• Probes contained normal bases (N columns in Table 1), two 2 '-O-methyl- adenosines (OA or VA columns in Table 1), or two 2'-O-methyl-adenosines and one 2'-O-methyl-thymidine or 2'-O- methyl-uridine at their central step (OAT or VAU columns in Table 1).
• the VAU probes contained two 2'-O-methyl-adenosines and three or four 2'-O-methyl-uridines for 13-C- G partners, or one/two 2'-O-methyl-uridines for 13-ED-GG and 13-ED-GA partners.
• each cell is shown the melting temperature for the perfectly matched duplex (G/C or A/T) and for duplexes containing mismatches (G/T or A/C) in each buffer, NaCl or TMAC. Melting temperature differences between paired and mismatched sequences are shown ( ⁇ ).
• 2'0-methyl-adenosine along with 2'0-methyl- thymidine or 2'0-methyl-uridine typically increases the difference in melting temperature of the fully matched and mismatched duplexes.
• site C the inclusion of three 2'-O- methyl nucleotides increases the mismatch discrimination from about 13 0 C for normal DNA to about 22 0 C for the substituted DNA.
• a more modest difference is seen for the duplexes containing two mismatches at sites D and E with base mixtures at the termini, but the differences are significantly greater when the termini comprise the "universal" base 5-nitro-indole.
• Example 12 Incorporation of modified bases in target DNA.
• DNA polymerases are able to incorporate a range of modified bases, including propynyl derivatives and 2-amino adenine, into DNA. Since improved specificity from incorporation of 2-amino adenine in probes was observed, the inventors investigated whether improved hybridisation specificity could be obtained through its incorporation during PCR into the target DNA.
• Target DNA was prepared from fully methylated DNA after bisulphite treatment by two rounds of PCR using primers targeted to the hMLHl gene ( Figure 15(a)). In the bisulphite-treated sequence potential methylation sites are designated YG and those within the final amplicon are shown in bold and denoted A through to H.
• Probes in the second and third rows of each set of probes contain 2'0-methyl adenosine at the locations shown in bold and those in the third row additionally contain 2'0-methyl uridine in place of thymidine at the discriminating nucleotide position (underlined).
• each set of probes contained normal DNA , with the left hand pair of spots corresponding to correct hybridisation with the methylated sequence and the right pair of spots G-T mispairing.
• target DNA contains 2-amino adenine.
• Reduced mispairing is seen for sites B and E for both normal and 2-amino adenine substituted target DNA.
• the additional inclusion of 2'O-methyl uridine at the interrogation site reduced mispairing somewhat for sites F and G.
• Example 13 Hybridisation of modified probes with representing CpG sites in the TPEF gene.
• the human TPEF gene (also known as TMEFF2) is frequently methylated in colorectal cancer. Nested PCR primers were designed to amplify a region of the TPEF promoter shown in Figure 16 (target sites for primers underlined). The CpG sites A to I within the amplicon are indicated and the cytosines that could be C or T after conversion labelled Y. TPEF sequences were amplified from fully methylated and unmethylated DNA by two rounds of PCR. Second round synthesis included dUTP and either Cy5- dUTP or biotin-dUTP and fragmented target DNA was prepared by USER cleavage as described in Example 2.
• Targets were applied to microarrays at 2O 0 C followed by brief washing at 37 0 C as in Example 6. Results obtained using Cy5 fluorescence were very similar to those using biotin and strepatvidin R-phycoerythrin (Example 6).
• TPEF site H was studied using four kinds of chemically-modified nucleotide as follows (modified nucleotides are shown in bold and discriminating nucleotides are underlined):
• any probe tells both its PCR target site and also its kind of modification.
• TH means "TPEF site H
• 1 or 2 means "cytosine or thymidine at CpG position”
• THlOA means "TPEF site H-I, 2'-0-methyl- adenosine”.
• Chemically-modified bases are shown in bold: “A” for 2'-O-methyl- adenosine, "C” for 2'-0-methyl-cytidine, "T” for 2'-O-methyl-thymidine, or "U” for T- O-methyl-uridine.
• Layouts of the probes on the arrays are shown in Figure 17(a). Correct base-pairing locations of the probes iii the hybridisation dot blots shown in Figure 17(b) and (c) have been outlined with thin rectangles, while incorrect base- pairing locations remain unlabelled.
• Figure 17(b) shows results for hybridisation of the methylated TPEF target and quantification by densitometry. It can be seen that one 2'-O-methyl T produces a slight reduction of G-T mispairing from 0.70-0.76 for normal DNA to 0.36-0.38, while two T- O-methyl T produce a more substantial reduction to 0.07-0.22, and three a further reduction down to 0.01-0.13. Results for the unmethylated TPEF target are shown in Figure 17(c). A-C mispairing is less of a problem than for G-T.
• Probes sequences and their layouts on the arrays of Figure 18(b) are shown in Figure 18(a).
• Probes in the second row of the arrays fully match the unmethylated target DNA; probes in the top row match the methylated target and will have two mismatches with the unmethylated target.
• Chemically modified bases are shown in bold, and the discriminating bases are underlined. The bottom two rows of probes contain one mismatch with either the methylated or unmethylated probe.
• Mismatch hybridisation is substantially reduced by the inclusion of 2'O-methyl nucleotides with a fifty-fold reduction for the probe containing three 2'O-methyl Ts; the relative specificity compared to binding to the correct unmethylated probes is also substantially improved (6-fold).
• site C Figure 18 (c) and (d)
• Example 14 Testing the microarray using DNA from a human patient with colorectal cancer
• the microarray with oligonucleotide probes of the type "X-9-X” was used to determine if differences in methylation could be seen in DNA samples taken from an individual with colorectal cancer.
• the methylation status of DNA from the tumour, and from adjacent normal tissue, was compared for a multiply-repeated LINE sequence (long interspersed nuclear element). Demethylation of Ll (LINE) promoter sequences has been reported in tumour tissue, and also to a degree in normal adjacent tissue, from patients with colorectal cancer (Suter, et al., 2004).
• the DNA was treated with bisulphite using a kit from Human Genetic Signatures. The reaction was for 4 hours at 75 0 C with heat pulses for 1 minute at 95°C every hour. 20 ng of bisulphite-converted sample was PCR-amplified incorporating a biotin label and with a 1 :3 ratio of dUTP:dTTP and then fragmented with USER enzyme as in Example 2.3. Fragmented DNA was denatured by heating and hybridisation to probes was done in 5OmM Tris HCl, pH 7.5, 5mM EDTA, pH 8, 2.0M TMA-Cl, 0.01% Tween 20, 0.05% SDS at room temperature for one hour. After 3 washes at room temperature, bound target was detected using streptavidin-R-phyco (Example 6)
• Spot intensities were quantified by measuring the intensity within a circle drawn tightly around each spot. Background was measured at four positions adjacent to each probe spot. The background counts were averaged, and subtracted from the spot intensity. Background corrected intensities for identical probes were averaged. The averaged intensity for target DNA binding to fully-methylated probes was divided by the averaged intensity for target DNA binding to the matched fully-unmethylated probes to obtain the ratio of methylated:unmethylated binding for each probe pair; this was done for DNA extracted from both the normal tissue and the tumour tissue samples.
• the ratio of methylated:unmethylated probe binding by DNA extracted from the normal tissue was divided by the ratio of methylated:unmethylated probe binding by DNA extracted from the tumour tissue, for each probe pair, to determine the relative differences in probe-pair binding by bisulphite-treated DNA extracted from normal compared with tumour tissue.
• Figure 19 shows the binding of the bisulphite-treated, PCR-amplified, and fragmented LINE target to LINE-specific probes on the microarray, where the probes in the top line of each image are specific for fully-methylated target DNA, and probes in the bottom line are specific for fuUy-unmethylated DNA; specifically, the probes have the generic base sequence of (a) LP 13 (top) and LP 14 (bottom), (b) LP21 (top) and LP22 (bottom), and (c) LPBl (top) and LPB2 (bottom).
• LPB1/LPB2 (c) all indicate that LINE DNA extracted from the normal sample is more methylated than LINE DNA extracted from the tumour sample, for the selected sequence region in this patient.
• Our microarray data showing an increase in demethylation of C at CpG sites in LINE in tumour tissue compared with matched normal tissue, are consistent with previous reports.
• Example 15 Detecting methylation of the TPEF gene in colorectal cancer DNA.
• Cy5 and biotin labelled PCR products' from bisulphite treated DNAs were prepared from control methylated and unmethylated DNAs as well as from matched tumour and normal DNA from patients with colorectal cancer. Methylation at specific sites within these DNAs were analysed using the restriction enzymes HpyCH4IV and BstUI that cut at ACGT and CGCG respectively — these correspond to sites I and A/B respectively in the TPEF sequence ( Figure 16). Cutting will occur at these sites only if the CpG site was methylated in the original DNA and the C is retained after bisulphite treatment. Digests of Cy5-labelled amplicons in Figure 20(a) show the detection of methylated DNA in tumour samples 2, 3 and 4 in comparison to the unmethylated DNA from adjacent normal tissue.
• Biotin-labelled Target DNA prepared from four patients was hybridised to an array of probes for detection of methylation at CpG sites A and B.
• the sequences and layout of these probes, that contained various modified nucleotides are shown in Figure 20(b).
• the positions of the interrogated sites are underlined and positions of modifications shown in bold.
• the left hand panel of Figure 20(c) shows hybridisation with control unmethylated DNA and with the DNA from non-diseased tissue of four patients (patients 2 to 5 from Figure 20(a)). Probes that match fully with unmethylated target DNA are boxed. It can be seen that the pattern of hybridisation is essentially the same for unmethylated DNA and DNAs and DNA from normal tissue. Mispairing to probes of normal DNA containing single mismatches is evident in the lower two rows of the left columns. This is essentially eliminated for probes containing 2'O-methyl uridine or thymidine.
• the right hand panel of Figure 20(c) shows hybridisation with a fully methylated control target or with tumour DNA from four patients. Probes on the top row and right hand side of the array should match with fully methylated DNA. As well as normal DNA bases, these contain a range of modified nucleotides as indicated. Hybridisation is similar in all cases and there is minimal mispairing of these probes with the unmethylated target DNA (left side). The boxed probes on the lower left contain single mismatches to fully methylated DNA and show considerable mispairing with unmethylated target. Incorporation of 2'O-methyl nucleotides, particularly three 2'O- methyl thymidines substantially reduces this mispairing.

Landscapes

• Chemical & Material Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Organic Chemistry (AREA)
• Engineering & Computer Science (AREA)
• Zoology (AREA)
• Wood Science & Technology (AREA)
• Proteomics, Peptides & Aminoacids (AREA)
• Health & Medical Sciences (AREA)
• Biophysics (AREA)
• General Engineering & Computer Science (AREA)
• Immunology (AREA)
• Microbiology (AREA)
• Molecular Biology (AREA)
• Analytical Chemistry (AREA)
• Physics & Mathematics (AREA)
• Genetics & Genomics (AREA)
• Biochemistry (AREA)
• Bioinformatics & Cheminformatics (AREA)
• Biotechnology (AREA)
• General Health & Medical Sciences (AREA)
• Chemical Kinetics & Catalysis (AREA)
• Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Priority Applications (2)

Applications Claiming Priority (2)

Publications (2)

Family

ID=39430076

Family Applications (1)

Country Status (3)

Cited By (6)

Family Cites Families (5)

• 2007
  • 2007-11-22 AU AU2007324273A patent/AU2007324273A1/en not_active Abandoned
  • 2007-11-22 WO PCT/AU2007/001797 patent/WO2008061311A2/en active Application Filing
  • 2007-11-22 EP EP07815599A patent/EP2087133A4/de not_active Withdrawn

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events