Single-molecule phenotyping and sequencing of nucleic acid
A cost-effective method using a single fluorescently labelled reporter probe and unlabelled competitive probes on a single-stranded gap section addresses the limitations of existing sequencing technologies, enabling high-throughput sequencing and phenotyping of nucleic acids, linking functional properties with DNA sequences.
Patent Information
- Application Number
- PCT/GB2026/050861
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-28
- Filing Date
- 2026-05-18
- Publication Date
- 2026-11-19
AI Technical Summary
Existing single-molecule sequencing methods, such as Gap-Seq, are costly and limited by high concentrations of labelled seal DNAs, requiring complex equipment and failing to directly link single-molecule phenotype with DNA sequence.
A method involving a single fluorescently labelled oligonucleotide reporter probe and unlabelled competitive probes to sequence nucleic acids, using a single-stranded gap section on a surface, allowing for cost-effective and high-throughput sequencing and phenotyping by detecting binding interactions.
Enables cost-effective, high-throughput sequencing and phenotyping of nucleic acids, linking functional properties with DNA sequences, and identifying sequence-dependent processes on DNA or RNA substrates, with the ability to introduce randomised sequences for interaction studies.
Smart Images

Figure GB2026050861_19112026_PF_FP_ABST
Abstract
Description
[0001] SINGLE-MOLECULE PHENOTYPING AND SEQUENCING OF NUCLEIC ACID MOLECULES
[0002] The present invention relates to a method of sequencing of a nucleic acid molecule and a method of single-molecule phenotyping and sequencing of a nucleic acid molecule.
[0003] Single-molecule DNA sequencing methods have made important contributions to genomics, diagnostics and functional analysis of genetic processes. Despite the fact that there are a large variety of single-molecule sequencing methods, there are a limited number of high-throughput methods to connect the functional properties of a single DNA molecule (single-molecule phenotype) with DNA sequence.
[0004] Within the last decade, single-molecule sequencing has become established in the commercial long-read sequencing market due to two companies, Pacific Biosciences and Oxford Nanopore Technologies, both with differing approaches. Pacific Biosciences use a single immobilised polymerase within a zero-mode waveguide to incorporate fluorescently labelled nucleotides (nts) during DNA synthesis allowing single bases to be read in real-time. Whereas, Oxford Nanopore Technologies' DNA sequencing method utilises the potential difference change across a protein pore as nucleotides pass through. Both major single-molecule sequencing methods require specialist equipment and cannot directly link single-molecule phenotype with DNA sequence.
[0005] Patent application publication No. W02019053430A1 (herein incorporated by reference) made significant progress in providing the ability to sequence single nucleic acid molecules and provide single molecule phenotyping. The technology introduced the concept of “Gap-seq” or “gap sequencing”, which uses fluorescent signals from the binding of different “seal” DNAs to surface-immobilised DNAs that feature a ssDNA gap. For the interrogation of a single position in the gap, a single-color, sequential seal DNA addition approach was implemented, which requires 4 labelled seal DNAs. If a 5-nt sequence is to be sequenced, a collection of 20 labelled DNAs is needed. If the fluorescence of the unbound labelled DNAs needs to be suppressed (desirable, since it allows use of higher unbound labelled DNA concentrations, which in turn leads to more frequent binding, and allows faster calling of a base), then these 20 labelled DNAs need to be doubly labelled. Such a requirement comes with a considerable cost per oligo(compared to unlabelled versions of the same sequences), while still placing a concentration limit for the use of Anorogenic doubly labelled seal DNAs. Therefore, further improvements are desirable to develop the usability of Gap-Seq and overcome these issues.
[0006] The aim of the present invention is to provide alternative, and preferably improved, methods for sequencing single nucleic acid molecules and molecule phenotyping and sequencing of a nucleic acid molecule.
[0007] According to the first aspect of the invention, there is provided a method of sequencing a nucleic acid molecule, the method comprising:
[0008] (a) providing the nucleic acid immobilised on a surface;
[0009] (b) forming a single -stranded gap section by partially duplexing the nucleic acid such that the single-stranded gap section is Aanked by duplex sections, wherein the sequence to be sequenced is a sequence of the single-stranded gap section;
[0010] (c) providing a Auorescently labelled oligonucleotide reporter probe arranged to fit into the gap section of the immobilised nucleic acid, and comprising universal and / or degenerate bases arranged to oppose the nucleotide positions of the gap section,
[0011] (d) providing a set of at least four competitive probes, wherein each competitive probe is arranged to fit in the gap section of the immobilised nucleic acid, and wherein each competitive probe comprises:
[0012] (i) a nucleotide X at a position arranged to oppose a nucleotide of the gap section of the immobilised nucleic acid, wherein such nucleotide is an interrogated nucleotide, and
[0013] (ii) fully complementary nucleotides for the remaining nucleotide positions of the competitive probe, which are arranged to oppose a complementary nucleotide of the gap section, or comprising complementary nucleotides combined with universal and / or degenerate nucleotides arranged to oppose some or all of the remaining nucleotide positions of the gap section (i.e. part complementary),
[0014] wherein the at least four competitive probes comprise:
[0015] - a first competitive probe wherein nucleotide X is a cytosine; - a second competitive probe wherein nucleotide X is a guanine;- a third competitive probe wherein nucleotide X is an adenine; and
[0016] - a fourth competitive probe wherein nucleotide X is a thymine or uracil;
[0017] wherein the competitive probes are arranged to compete with the fluorescently labelled oligonucleotide reporter probe for binding to the gap section;
[0018] (e) detecting binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probe with the gap section of the immobilised nucleic acid in the presence of one of the four competitive binding probes;
[0019] (f) repeating step (e) for each of the remaining competitive binding probes, wherein the identity of the interrogated nucleotide of the gap section of the immobilised nucleic acid is identified as the complementary base of the nucleotide X of the competitive probe that has the highest incidence of disrupting the binding of the fluorescently labelled oligonucleotide reporter probe; and optionally
[0020] (g) repeating steps (d) to (f) for interrogating subsequent nucleotide positions of the gap section of the immobilised nucleic acid until sufficient nucleotides of the gap section have been identified to be able to determine a sequence.
[0021] The invention advantageously provides the ability to sequence a nucleic acid at a singlemolecule level and in a form that can be linked to functional studies of the same nucleic acid molecule, or molecules linked thereto. In particular, due to the immobilisation of the nucleic acid on a surface, the observed functional / phenotypic affects associated with the same surface location can be attributed to the nucleic acid sequence. Using large libraries of nucleic acids (such as DNA and RNA oligonucleotides), this capability allows the identification of the sequence-dependence of processes that operate on DNA or RNA substrates, the discovery of DNA / RNA aptamers for desirable binding / reaction profiles versus various targets (such as proteins, nucleic acids, small molecules etc.), and the reading of bar-codes that can be introduced in biological molecules and cells.
[0022] The invention further advantageously allows the ability to introduce 4AN randomised sequences into nucleic acids, which are then fixed on a surface, and then studied for interactions with individual biomolecules such as proteins, and then sequenced. For example, a skilled person can create a library of DNA molecules with 4AN randomized sequence members (e.g. 4A5 = 1024 sequences) using automated DNA synthesis and degenerate nucleotides for the section to be randomized. DNAs can then converted intoa double-stranded form and, if needed, can be labelled or modified using suitably modified primers and PCR. The library can then be immobilised on a solid support such as a glass coverslip for functional interrogation at the single-molecule level (to obtain the single-molecule phenotype).
[0023] The randomised sequences can also be introduced in modified genes, that are introduced to cells, which are then fixed on a surface and then interrogated for functionality, and DNA sequence. For example, a skilled person can create a library with 4AN randomized sequence members (e.g. 4A5), which can then be integrated into a plasmid library. The randomised region is flanked from either side with two sequences that are long enough and unique relative to the genome of an organism of interest, allowing the generation of a single-stranded gap section for sequencing after phenotyping at a cellular level. Transformation of these fragments can lead to populations of cells (e.g., E. coli), with each cell making one of many permutations of a certain protein according to the introduced 4AN randomised sequences. The protein may be any protein that has a cellular phenotype, such as a protein that kills bacteria or changes their expression. Fixing the cells to a surface (thereby immobilising their nucleic acid content to a surface) followed by sequencing according to the present invention allows the link to be established between a cell having a certain phenotype and the nucleic acid sequence associated with that phenotype. The fixed cell(s) may be permeabilised prior to the sequencing in order to allow sequencing reagent access to the immobilised nucleic acid. Vvedenskaya et al. (Molecular Cell 60, 953-965, 2017 - incorporated herein by reference) is a publication that makes such libraries on plasmid, transforms them in bacteria, and then sequences the nucleic acid by a different method using nextgeneration sequencing. However, such a method does not provide phenotypic analysis and sequencing at the single-cell level, or the single-molecule level, by fixing the cells to a surface prior to sequencing their nucleic acid content.
[0024] Furthermore, the method of this invention is a significant improvement on the method disclosed in prior Patent Application No. W02019053430A1, because the method relies on only a single reporter probe, which needs to be provided with fluorescence labelling. The competitive probes can remain unlabelled and be produced economically in large quantities with multiple sequence variations. There is a considerable cost saving per oligo relative to the previous method, and there is less constraint on the concentration limit when using a fluorescently labelled oligonucleotide probe.The skilled person will recognise that the steps of the method herein may be in a different order, for example where it is possible or appropriate to carry out one step before another despite the order of presentation. In one embodiment, the steps are in the order presented.
[0025] In one embodiment, steps (d) to (f) are repeated until the gap section of the immobilised nucleic acid is fully sequenced.
[0026] Interrogating subsequent nucleotide positions of the gap section of the immobilised nucleic acid may comprise providing a further set of at least four competitive binding probes, wherein the position of nucleotide X is changed to oppose a different / subsequent nucleotide to be interrogated of the immobilised nucleic acid. The identity of nucleotide X in the competitive probe(s) may be known / pre-determined.
[0027] In an embodiment wherein the competitive probes comprise complementary nucleotides combined with universal and / or degenerate nucleotides arranged to oppose some or all of the remaining nucleotide positions of the gap section, there may be up to 7 nucleotides that are universal and / or degenerate nucleotides (not including the X position). For embodiments comprising competitive probes of up to 7 nucleotides in length, all of the nucleotides may be degenerate and / or universal. The universal and / or degenerate nucleotides may be immediately adjacent to the X position. The universal and / or degenerate nucleotides may flank the X position (i.e. with complementary nucleotides therebetween). The competitive probe may comprise the motif X (for a single position interrogation), the motif DX, XD (for a 2-nt sequence interrogation), the motif XDD, DXD, DDX (for a 3 -nt sequence interrogation), and so on (wherein X can be in any position), wherein D = a degenerate or universal base, and X is the herein defined X position (i.e. A, C, T, G, or U). The herein described motif sequence of the competitive probes may be flanked by fully complementary nucleotides. The herein described motif sequence of the competitive probes may be flanked by 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 fully complementary nucleotides on each side (5’ and 3’). The 5’ and 3’ ends of the competitive probe may comprise at least 1, 2, or 3 fully complementary nucleotides. For embodiments of competitive probes that are more than 7 nucleotides in length, the herein described motif sequence of the competitive probes may be flanked by 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, or more fully complementary nucleotides on each side(5’ and 3’). For embodiments of competitive probes that are more than 7 nucleotides in length, the 5’ and 3’ ends of the competitive probe may comprise at least 1, 2, or 3 fully complementary nucleotides.
[0028] Detecting binding, or absence thereof, by the fluorescently labelled oligonucleotide reporter probe in the presence of a competitive binding probe may be done sequentially one after the other (i.e. the use of each competitive binding probe may be done in the absence of other competitive binding probes of the competitive binding probe set).
[0029] Detecting the binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probes with the gap section of the immobilised nucleic acid may comprise detecting the transient binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probes with the gap section of the immobilised nucleic acid. Detecting the binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probes with the gap section of the immobilised nucleic acid may comprise detecting the specific binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probes with the gap section of the immobilised nucleic acid.
[0030] The identity of the interrogated nucleotide of the gap section of the immobilised nucleic acid may be identified as the complementary base of the nucleotide X of the competitive binding probe that has the highest level of disruption of the binding of the fluorescently labelled oligonucleotide reporter probe (i.e. the competitive probe that most outcompetes the fluorescently labelled oligonucleotide reporter probe), or highest level of disruption of transient binding of the fluorescently labelled oligonucleotide reporter probes as appropriate. There are many ways to detect the binding or absence of binding of the fluorescently labelled oligonucleotide probe into the gap, e.g., the fraction of time that the fluorescence from the oligonucleotide probe shows higher or lower intensity, the intensity distribution above or below a threshold, the number of binding events, or detecting a temporal pattern, for example that can be learnt using an Al algorithm. The binding of the fluorescently labelled oligonucleotide reporter probe in the presence of one competitive probe may be compared with the binding of the fluorescently labelled oligonucleotide reporter probe in the presence of another competitive probe (i.e. having a different X base). The lowest incidence of binding may be indicative of the fluorescently labelled oligonucleotide reporter probe being out competed by a competitive probe with the most complementary sequence, e.g. thesequence with a base X complementary to the interrogated nucleotide of the immobilised nucleic acid.
[0031] The method may comprise a wash step between steps e) and f) and / or between steps f) and g). In one embodiment, a competitive probe is removed (e.g. washed away) before introducing a subsequent different competitive probe. Similarly, a fluorescently labelled oligonucleotide reporter probe may be removed (e.g. washed away) prior to repeating the interrogation with a different fluorescently labelled oligonucleotide reporter probe. The wash step may comprise rinsing the immobilised nucleic acid with a wash solution to wash away any unbound probes. Due to the nature of the transient binding in the gap, probes (labelled probes and competitive probes) may be washed away without the need for denaturing solutions / buffers. The wash may be capable of de-hybridising and clearing any probes, but not capable of denaturing the duplexed sections of the immobilised nucleic acid. The wash solution may comprise a buffer or salt.
[0032] The method may further comprise a photobleaching step between interrogation of different probes. The method may comprise a photobleaching step between steps e) and f) and / or between steps f) and g). The photobleaching and wash steps may be combined.
[0033] In one embodiment, the single-stranded gap section is flanked by a pair of duplex sections. In one embodiment, partially duplexing the nucleic acid comprises hybridising a pair of gap-flanking nucleic acid strands that are each complementary to a different section of the nucleic acid to be duplexed, wherein the single-stranded gap section is formed therebetween. In another embodiment, the partially duplexed nucleic acid comprising a single stranded gap-section is formed by circularising the nucleic acid with a splint nucleic acid that is arranged to hybridise to each end of the nucleic acid (i.e. forming a bridge between each end to form a circularised partially duplexed nucleic acid molecule comprising a single stranded gap-section).
[0034] In another embodiment, partially duplexing the nucleic acid comprises forming a hairpin loop in the nucleic acid, such that the nucleic acid partially hybridises with itself. A pair of hairpin loops may be formed at opposing ends of the nucleic acid, wherein the single-stranded gap section is formed therebetween. Alternatively, a hairpin loop may be formed at one end / section of the nucleic acid, and a gap-flanking nucleic acid strand that is complementary to a section of the nucleic acid to be duplexed may be hybridisedto the nucleic acid, wherein the single-stranded gap section is formed between the duplex formed by the hairpin, and the duplex formed by the hybridised gap-flanking nucleic acid strand.
[0035] The gap-flanking nucleic acid strands may have a predetermined / known sequence. The sequences of the nucleic acid to be duplexed may determine the complementary sequences of the gap-flanking nucleic acid strand(s), or in an embodiment requiring a hairpin loop, determine the section of the nucleic acid to be looped back and hybridise to form the hairpin loop.
[0036] The sequences of the nucleic acid to be duplexed may be designed or built into the nucleic acid, such that they flank a sequence of unknown / random nucleotides to be interrogated. For example, the sequences of the nucleic acid to be duplexed may be designed or built into the nucleic acid during the synthesis of the nucleic acid. In another example, the sequences of the nucleic acid to be duplexed may be built into the nucleic acid during recombinant assembly of the nucleic acid.
[0037] The duplexing of the nucleic acid may be arranged to create the single-stranded gap section. The single-stranded gap section can be placed in the middle of the nucleic acid sequence, or closer to the ends of the nucleic acid. The single-stranded gap section can also be formed in circular nucleic acid molecules. The positioning of the single-stranded gap section (i.e. middle or near to an end) may depend on the nature of a functional assay that may be performed prior to the sequencing of the nucleic acid.
[0038] The sequencing may be done at, or around, room temperature (i.e. between 18°C and 28°C, such as about 24°C). The sequencing may be done at a temperature of between 5 °C and 45 °C. Different temperatures will allow the utilisation of gaps of different length.
[0039] The transient binding detection may be carried out in a buffer. Different buffer components can be provided to facilitate the sequencing. In one embodiment, a monovalent ion concentration (Na+) can be provided at a concentration of between about 0.01 and about 3 M. Adjusting the monovalent ion concentration (Na+) can change the binding affinity of the sequencing probes to the single-stranded gap section; high saltconcentration screen electrostatic repulsions and therefore enhance the stability of hybridised nucleic acids. An additive TMAC can be used, for example at 3M concentrations, to make A-T rich sequences equivalent in stability to G-C rich sequences. The buffer may comprise dextran sulphate (such as 5-20%) and / or formamide (such as 5-20%). Additionally or alternatively, the buffer may comprise spermidine (e.g. about 250 pM), for example to increase the on-rate of binding.
[0040] During phenotypic and sequencing assays, buffer components are chosen by the skilled person to enhance the observation of the binding or reaction in question. To allow extended observations, oxygen scavengers (such as glucose oxidase / catalase / glucose, or PCA / PCD) and triplet-state quenchers (such as TROLOX and mercaptoethylamine) can be used to extend the photobleaching lifetime of the fluorophores, and to minimise complicating photophysics. Therefore, in one embodiment, the sequencing reaction is provided with an oxygen scavenger and / or triplet-state quencher.
[0041] In one embodiment, the fluorescently labelled oligonucleotide reporter probe comprises universal bases arranged to oppose some or all of the nucleotide positions of the gap section. In one embodiment, the fluorescently labelled oligonucleotide reporter probe comprises universal bases arranged to oppose all of the nucleotide positions of the gap section (i.e. fills the gap). The skilled person will recognise that a universal base can be described as a base analog without hydrogen bonding groups that can pack efficiently in duplex DNA, but shows no or insignificant selectivity in pairing with any particular native base, such as A, C, G, and T. Examples of universal bases are inosine, deoxyinosine, 5’ nitroindole, and 3 -nitropyrrole. Several universal bases are also described by Berger et al. (Nucleic Acids Res. 2000 Aug 1; 28(15): 2911-2914), which is incorporated herein by reference. In one embodiment, the universal bases may comprise or consist of inosine.
[0042] In another embodiment, the fluorescently labelled oligonucleotide reporter probe comprise degenerate bases arranged to oppose some or all of the nucleotide positions of the gap section. In another embodiment, the fluorescently labelled oligonucleotide reporter probe comprise degenerate bases arranged to oppose all of the nucleotide positions of the gap section (i.e. fills the gap). The term “degenerate bases” used herein means any base, such as any of A, C, T, or G. The degenerate base may or may not match the opposing base of the immobilised nucleic acid in the gap section.The fluorescently labelled oligonucleotide reporter probe comprise a mixture of degenerate bases and universal bases.
[0043] The fluorescently labelled oligonucleotide reporter probe may be equal in length to the gap section of the immobilised nucleic acid. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be shorter or longer in length to the gap section of the immobilised nucleic acid. The fluorescently labelled oligonucleotide reporter probe may be at least one nucleotide in length (i.e. may consist of a single labelled nucleotide). The fluorescently labelled oligonucleotide reporter probe may be at least 2 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be at least 3 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be at least 4 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be at least 5 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be at least 6 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be 1-12 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be 2-12 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be 3-12 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be 6-12 nucleotides in length. The fluorescently labelled oligonucleotide reporter probe may be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 nucleotides in length. Use of conditions that destabilise nucleic acid hybrids (e.g., low Na+concentration) will also allow longer gaps (e.g. up to 20nts) to be used. Therefore, in another embodiment, the fluorescently labelled oligonucleotide reporter probe is 13, 14, 15, 16, 17, 18, 19, 20, or more nucleotides in length.
[0044] In embodiments wherein the fluorescently labelled oligonucleotide reporter probe is greater than 7 nucleotides in length, up to 1-3 nucleotides of the 5’ end and up to 1-3 nucleotides of the 3’ end may be complementary to the immobilised nucleic acid sequence.
[0045] In one embodiment, the fluorescently labelled oligonucleotide reporter probe comprises or consists of DNA. In another embodiment, the fluorescently labelled oligonucleotide reporter probe comprises or consists of RNA. In another embodiment, the fluorescently labelled oligonucleotide reporter probe comprises or consists of nucleic acid analoguesuch as PNA, LNA or PMO, or combinations thereof. The fluorescently labelled oligonucleotide reporter probe may comprise or consist of nucleic acid selected from DNA, RNA, and nucleic acid analogues, or combinations thereof.
[0046] The fluorescently labelled oligonucleotide reporter probe may comprise further structural or identification sequences (e.g. barcode sequence) that is not arranged to hybridise to, oppose, or complement the sequence of the gap section of the immobilised nucleic acid.
[0047] The fluorescent labels of the fluorescently labelled oligonucleotide reporter probe may be selected from any of the group comprising 5-carboxyfluorescein (FAM), tetramethylrhodamine (TMR), Alexa-Fluor fluorophores (such as Alexa488, Alexa532, Alexa546, Alexa555, Alexa568, Alexa594, and Alexa647; available from Molecular Probes / Invitrogen), BODIPY dyes, ATTO dyes (such as ATTO532, ATTO568, ATTO594, ATTO647n, and ATTO655; available from Attotech) and cyanine dyes (such as Cy2, Cy3, Cy3B, Cy3.5, Cy5, Cy5.5, and Cy7; available from GE Healthcare), or combinations thereof.
[0048] In one embodiment, the binding, such as transient binding, of a fluorescently labeled oligonucleotide reporter probe to the gap-section of the nucleic acid is detected by measurement of Forster resonance energy transfer (FRET) between the fluorophore on the fluorescently labeled oligonucleotide reporter probe and a second fluorophore. The second fluorophore may be tagged on the immobilized nucleic acid molecule or duplex thereof. For example, the second fluorophore may be tagged on a gap-flanking nucleic acid strand that is hybridized to the immobilized nucleic acid molecule. For example, the fluorophores Cy3B and ATTO647N can be used together as a donor and an acceptor pair for FRET. Therefore, the fluorophore on the fluorescently labeled oligonucleotide reporter probe may comprise Cy3B, and the second fluorophore may comprise ATTO647N, or vice versa.
[0049] A second fluorophore or any other label of the probes / seals can also be employed. In one embodiment, the 2ndfluorophore allows FRET where both fluorophores are on the reporter probe (also known as “labelled-seal” herein) and the quenching state is dependent on being duplexed and far apart (thus resulting in low FRET efficiency), or non-duplexed and fluorophores brought together (thus resulting in contact-mediatedquenching). The FRET acceptor can either be a high-fluorescent fluorophore (such as ATTO647N); in this case, signals in both FRET donor and FRET acceptor emission channels will be used to monitor binding. The FRET acceptor can also be a dark (non-fluorescent) quencher (such as QSY7, QSY21, BHQ2 or BHQ3); in this case, the signals in the FRET donor emission channels will be used to monitor binding. Where there are two fluorophores tagged on an oligonucleotide reporter probe, the fluorophores may be at opposite ends of the probe, or at least sufficiently far apart that the length of nucleic acid therebetween eliminates contact-mediated quenching upon binding, while allows efficient FRET between donor and acceptor upon binding of the distance between the fluorophores may be at least 5 or more nucleotides. The skilled person can provide an appropriate distance depending on the exact pair of fluorophores used (their structure, dynamics, and spectral properties). The fluorophores may be bound to the 5’ and 3’ nucleotides respectively, or the immediately adjacent nucleotides.
[0050] The fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 0.1 nM and about 100 pM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 0.1 nM and about 10 pM. The fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of about 10 nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may also be provided at a concentration of about 500 nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 5 nM and about lOOOnM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 10 nM and about 800nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 100 nM and about 800nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 300 nM and about 800 nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 400 nM and about 600 nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 5 nM and about 50nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 5 nM and about 20nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probemay be provided at a concentration of between about 8 nM and about 15nM. In another embodiment, the fluorescently labelled oligonucleotide reporter probe may be provided at a concentration of between about 10 nM and about 20nM.
[0051] In another embodiment, the fluorescently labelled oligonucleotide probes may be provided at a concentration of greater than lOOnM, or greater than 200nM. The provided concentrations of fluorescently labelled oligonucleotide probes refers to the concentration of total probe species unless stated otherwise i.e. not the concentration of a singly fluorescently labelled oligonucleotide probe.
[0052] The concentration of fluorescently labelled oligonucleotide reporter probe determines the rate of binding of the reporter probe to the immobilised DNA. Higher concentrations (e.g., >100 nM) enhance the rate of binding, but contribute to high fluorescence background, and prevent or make difficult the detection of single immobilised molecules. Lower concentrations are associated with low background, but lead to increased waiting times for collecting sufficient binding events for base calling. Other techniques minimised this background either by suppressing background fluorescence via confined function / detection (such as with zero-mode waveguides) or by adopting label-free approaches (such as nanopores). The present invention advantageously can be used with a high concentration range that allows rapid detection without the need for specialised structures such as zero-mode waveguides.
[0053] In one embodiment, the competitive probes comprise universal bases arranged to oppose some or all of the nucleotide positions of the gap section, except for position X. In one embodiment, the competitive probes comprise universal bases arranged to oppose all of the nucleotide positions of the gap section (i.e. fills the gap), except for position X. In one embodiment, the universal bases may comprise or consist of inosine.
[0054] In another embodiment, the competitive probes comprise degenerate bases arranged to oppose some or all of the nucleotide positions of the gap section, except for position X. In another embodiment, the competitive probes comprise degenerate bases arranged to oppose all of the nucleotide positions of the gap section (i.e. fills the gap), except for position X. The degenerate base may or may not match the opposing base of the immobilised nucleic acid in the gap section.The competitive probes may comprise a mixture of degenerate bases and universal bases.
[0055] The competitive probes may be equal in length to the gap section of the immobilised nucleic acid. In another embodiment, the competitive probes may be shorter or longer in length to the gap section of the immobilised nucleic acid. The competitive probes may be at least one nucleotide in length (i.e. may consist of a single labelled nucleotide). The competitive probes may be at least 2 nucleotides in length. The competitive probes may be at least 3 nucleotides in length. The competitive probes may be at least 4 nucleotides in length. The competitive probes may be at least 5 nucleotides in length. The competitive probes may be at least 6 nucleotides in length. The competitive probes may be 1-12 nucleotides in length. The competitive probes may be 2-12 nucleotides in length. The competitive probes may be 3-12 nucleotides in length. The competitive probes may be 6-12 nucleotides in length. The competitive probes may be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 nucleotides in length. Use of conditions that destabilise nucleic acid hybrids (e.g., low Na+concentration) will also allow longer gaps (e.g. up to 20nts) to be used. Therefore, in another embodiment, the competitive probes are 13, 14, 15, 16, 17, 18, 19, 20, or more nucleotides in length. In one embodiment, the competitive probes are equal in length to the fluorescently labelled oligonucleotide reporter probe.
[0056] In an embodiment wherein the competitive probes comprise complementary nucleotides combined with universal and / or degenerate nucleotides arranged to oppose some or all of the remaining nucleotide positions of the gap section, the competitive probe may be 8 or more nucleotides in length. In an embodiment wherein the competitive probes comprise complementary nucleotides combined with universal and / or degenerate nucleotides arranged to oppose some or all of the remaining nucleotide positions of the gap section, the competitive probe may from about 8 to 20 nucleotides in length.
[0057] Where the length of the competitive probe is defined, decreased or increased, the corresponding length of the gap section and / or fluorescently labelled oligonucleotide reporter probe may also be adjusted to match the competitive probe length. Where the length of the fluorescently labelled oligonucleotide reporter probe is defined, decreased or increased, the corresponding length of the gap section and / or competitive probe may also be adjusted to match the competitive probe length. Where the length of the gap section is defined, decreased or increased, the corresponding length of the fluorescentlylabelled oligonucleotide reporter probe and / or competitive probe may also be adjusted to match the competitive probe length.
[0058] In one embodiment, the competitive probes comprise or consist of DNA. In another embodiment, the competitive probes comprise or consist of RNA. In another embodiment, the competitive probes comprise or consist of nucleic acid analogue such as PNA, LNA or PMO, or combinations thereof. The competitive probes may comprise or consist of nucleic acid selected from DNA, RNA, and nucleic acid analogues, or combinations thereof.
[0059] The competitive probes may comprise a structural or identification sequences (e.g. barcode sequence) that is not arranged to hybridise to, oppose, or complement the sequence of the gap section of the immobilised nucleic acid. The competitive probes may or may not be labelled, such as fluorescently labelled. If labelled, the fluorescent label may be different to / distinguishable from the fluorescent label(s) of the fluorescently labelled reporter probes. Preferably, the competitive probes are not fluorescently labelled.
[0060] Each competitive probe may be provided at a concentration of between about 0.1 nM and about 500 pM. In another embodiment, each competitive probe may be provided at a concentration of between about 0.1 nM and about 100 pM. In another embodiment, each competitive probe may be provided at a concentration of between about 5 nM and about 500 pM. Each competitive probe may be provided at a concentration of about 10 nM. In another embodiment, each competitive probe may also be provided at a concentration of about 500 nM. In another embodiment, each competitive probe may be provided at a concentration of between about 5 nM and about lOOOnM. In another embodiment, the fluorescently labelled oligonucleotide reporter pr each competitive probe may be provided at a concentration of between about 10 nM and about 800nM. In another embodiment, each competitive probe may be provided at a concentration of between about 100 nM and about 800nM. In another embodiment, each competitive probe may be provided at a concentration of between about 300 nM and about 800 nM. In another embodiment, each competitive probe may be provided at a concentration of between about 400 nM and about 600 nM. In another embodiment, each competitive probe may be provided at a concentration of between about 5 nM and about 50nM. In another embodiment, each competitive probe may be provided at a concentration ofbetween about 5 nM and about 20nM. In another embodiment, each competitive probe may be provided at a concentration of between about 8 nM and about 15nM. In another embodiment, each competitive probe may be provided at a concentration of between about 10 nM and about 20nM. In another embodiment, each competitive probe may be provided at a concentration of greater than lOOnM, or greater than 200nM.
[0061] The concentrations or relative amounts of the at least four competitive probes may be substantially equal to each other.
[0062] The fluorescently labelled reporter probe may be substantially equal in concentration to the competitor probe. The ratio of the number of fluorescently labelled reporter probes to competitor probes may be from about 10:1 to 1:10. In another embodiment, the ratio of the number of fluorescently labelled reporter probes to competitor probes may be from about 2: 1 to 1:2. The ratio of the number of fluorescently labelled reporter probes to competitor probes may be about 1:1.
[0063] The skilled person will understand that the concentration of the probes (reporter or competitive) and / or the assay time may be adjusted depending on the nature of the probe (i.e., the degree of base pairing, the G / C content, the length of the single stranded gapsection, and the degree of quenching / dequenching achieved upon binding of the probe to its target). The concentration and / or assay time adjustment may also depend on the use of universal or degenerate bases, since in the case of universal bases, a large fraction of the probes (25%) will be able to bind preferentially to the interrogated sequence, whereas in the case of degenerate bases, a much smaller fraction of the probes (1 out of 4Nbases) will be able to bind preferentially to the interrogated sequence. Therefore, in an embodiment wherein universal bases are provided in the probes, the concentration of the probes may be between 10 nM and 500 pM (total probe concentration). In another embodiment wherein universal bases are provided in the probes, the concentration of the probes may be between 10 nM and 1000 nM (total probe concentration). In an embodiment wherein degenerate bases are provided in the probes, the concentration of the competitive probes may be between 10 nM and 500 pM (total probe concentration), and the assay may be performed for a longer time to account for the less frequent binding of the probe to the immobilised target. In another embodiment wherein degenerate bases are provided in the probes, the concentration of the competitive probes may be between 10 nM and 1000 nM (total probe concentration), and the assay may be performed for alonger time to account for the less frequent binding of the probe to the immobilised target.
[0064] In one embodiment, the nucleic acid is synthetically produced. Additionally the nucleic acid may be non-natural / synthetic in origin (i.e. comprises non-wild-type sequence). In another embodiment, the nucleic acid may be recombinantly produced in vivo, for example in a cell. In another embodiment, the nucleic acid may be naturally occurring in vivo, for example in a cell.
[0065] In one embodiment, the nucleic acid comprises a combination of pre-determined sequence and unknown / randomised sequence. The nucleic acid may comprise a pair of pre-determined sequences flanking an unknown / randomised sequence therebetween. The predetermined sequences may be arranged to form a duplex with gap-flanking nucleic acid strands.
[0066] The nucleic acid may be provided in the form of a library of nucleic acids comprising a randomised nucleotide region. In one embodiment, the library has 4Nmembers, where N is the number of randomised bases, for example, a randomised 5-nucleotide region will be represented by 1024 possible DNA sequences and a randomised library of 9-nucleotide region will be represented by 2.6xl05different sequences. In one embodiment, A is 5. In another embodiment, A is 9. In another embodiment, A is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12. In addition to the randomised nucleotide region, the library members may be provided with known / pre-determined sequence. The randomised nucleotide region may be flanked by known / pre-determined sequences. The known / pre-determined sequences may be identical for each library member. The known / pre-determined sequences may comprise the sequences that are arranged to form a duplex with the gap-flanking nucleic acid strands. The library members may be labelled, such as fluorescently, preferably the same label for each member.
[0067] Advantageously, the use of small (e.g. 9 nts) gap-sections can require libraries of short (e.g. 9 nts) competitive probes (otherwise known as “competitive seals” or “unlabelled seals”); the short size can prevent stable hybridization between unbound probes in solution. The formation of the gap also removes secondary structures that may be present on the template, addressing a significant issue with templates such as RNA. Thegap also allows simple DNA (i.e., not only DNA nanostructures) to be interrogated for binding.
[0068] The transient binding can also work with short single-stranded DNAs that can be prepared after the functional assay via restriction digestion and removal of the nonsurface-immobilised strand of the short dsDNA fragment. The assay can also work with template-primer hybrids that leave a short overhang (with a size similar to the preferred gap size, i.e., 6-12 nts) on one of the ends.
[0069] In one embodiment, the nucleic acid comprises or consists of DNA. In another embodiment, the nucleic acid comprises or consists of RNA. In another embodiment, the nucleic acid comprises or consists of nucleic acid analogue such as PNA, LNA or PMO, or combinations thereof. The nucleic acid may be selected from DNA, RNA, and nucleic acid analogues, or combinations thereof.
[0070] The nucleic acid (for example each of the library members) may comprise a functionality (such as 5’ or 3 ’-biotinylation of the DNA) that allows its surfaceimmobilisation on a surface. For example, the functionality may comprise a 5’ or 3’-biotinylation of the nucleic acid to allow surface immobilisation on neutravidin-coated glass surfaces. In one embodiment, the nucleic acid is biotinylated. The skilled person will recognise that there are other methods to immobilise the nucleic acid, such as crosslinking it to surfaces decorated with reactive groups (e.g. NFU-modified DNA covalently linked to NHS-modified surfaces; SH-modified DNA covalently linked to maleimide-modified surfaces; and azido-modified DNA covalently linked to alkyne-modified surface via clink chemistry, any of which may be used in the method according to the invention), or using antibody / antigen interactions (e.g. digoxygenin-modified DNA covalently linked to anti-dig antibody-coated surfaces).
[0071] The biotin, or other functional group for immobilisation, may be provided on a linker. The linker may be sufficiently long enough to avoid steric hindrance. Polyethylene glycol (PEG) linkers may be used to provide a hydrophilic layer that can prevent / minimise non-specific adsorption to surfaces. PEG spacers of different sizes (e.g. PEG333, PEG5000) and different functionalities (e.g. lipid-PEG, poly-lysine PEG) can be used to further minimise non-specific binding.The nucleic acid may be immobilised on a solid surface. The surface may comprise or consist of glass. The surface may be a plate, such as a flat plate. In one embodiment, the surface is a glass coverslip. The surface may be treated to facilitate immobilisation of the nucleic acid. In one embodiment, the surface, such as a glass surface, is neutravidin-coated.
[0072] In one embodiment, immobilised nucleic acids are spatially arranged onto the surface, such as glass. For example, single nucleic acid molecules or groups of the same species of nucleic acid molecules may be spaced apart from other nucleic acid molecules or groups. The spacing may be sufficient that they can be spatially resolved from each other, for example when detecting a fluorescent oligonucleotide probe emission. The fluorescent oligonucleotide probe emission may be detected and visualised by a singlemolecule fluorescence microscope. Crowded-field analysis of immobilised fluorescent molecules may be used to detect and / or visualise the fluorescent oligonucleotide probe emissions. High-precision localisation of the binding events may be detected via identification of the centre of the point-spread function of the bound fluorophores.
[0073] The spatial arrangement may be random or patterned, for example into an array. The nucleic acid molecules may be sparsely immobilised on the surface to allow their detection on the surface at the single-molecule level, for example using a singlemolecule fluorescence microscope.
[0074] In one embodiment, the nucleic acid molecules may be immobilised on the surface at a surface density of about 1 molecule per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of about 10 molecules per pm2, for example when crowded-field analysis of immobilised fluorescent molecules is used. In another embodiment, the nucleic acid molecules may be immobilised on the surface at a surface density of about 1000 molecules per pm2, for example when employing high-precision localisation of the binding events (via identification of the centre of the point-spread function of the bound fluorophores). In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of between about 1 and about 1500 molecules per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of between about 1 and about 1100 molecules per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density ofbetween about 1 and about 1000 molecules per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of between about 1 and about 500 molecules per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of between about 1 and about 100 molecules per pm2. In another embodiment the nucleic acid molecules may be immobilised on the surface at a surface density of between about 1 and about 50 molecules per pm2.
[0075] The use of a regular array for the deposition of DNA will increase the throughput and decrease the computational costs of sequencing, since it will provide spacing that is optimised for the microscope, fluorophores and binding times involved.
[0076] Advantageously, imaging on a single-molecule fluorescence microscope with a large field-of-view, such as with sCMOS-based wide-field imaging, allows parallel detection of 500-2000 single molecules per field-of-view. Instrumentation with lower magnification on a sCMOS camera allows 15,000 molecules to be visualised per field of view. Implementation on sCMOS-based wide-field imaging combined with high-precision localisation of the binding events allows ~1.5xl07molecules to be visualised per field of view, providing extensive sequence coverage without the need for scanning, and opening a large number of potential applications.
[0077] In one embodiment, the immobilised nucleic acid may be immobilised to the bottom of a well, such as an array of zero-mode waveguides (for example, comprising subwavelength apertures of about 10-100 nm in diameter, about 100 nm in depth). This allows a higher concentration of probe to be used, providing faster binding to the gap DNA and faster readout of sequences.
[0078] In one embodiment, the immobilised nucleic acid may be immobilised by way of being in situ in a cell and the cell is chemically fixed (e.g., using 1-4% paraformaldehyde in phosphate-buffer saline), permeabilised (e.g., using treatment with ice-cold 90% methanol) to allow access of the sequencing reagents to the nucleic acid of the cell, and then attached to the surface (e.g., to a polylysine-coated surface or to an agarose pad). Alternatively, the cells can be immobilised by capture in microfluidic channels.In another embodiment the nucleic acid or groups of identical nucleic acid molecules may be immobilised on a bead or a nanoparticle (such as a quantum dot or a fluorescent nanodiamond). The nucleic acid may be immobilised on a glass bead. For sequencing according to the invention, the bead may be fixed into position on a surface. Beads comprising immobilised nucleic acid on their surface may be spatially arranged, such that beads are spaced apart. The spacing may be sufficient to resolve the beads for example when detecting a fluorescent probe emission.
[0079] The method of the invention may further provide the step of amplifying the immobilised nucleic acid. The amplification may comprise a rolling circle amplification (RCA) of the immobilised nucleic acid. The surface-immobilised nucleic acid may be circularised using a splint nucleic acid molecule (for example a DNA molecule), to bring together and close the ends, followed by ligation. The splint nucleic may also be used as a primer for the RCA.
[0080] Amplification of the immobilised nucleic acid can increase the speed of the assay and decrease the concentration of fluorescent seal DNAs needed is to amplify the gapsection. RCA can produce >100 copies of the randomised sequences within an hour (Mohsen and Coll, Acc Chem Res, 2016); removal of the DNA polymerase and annealing of a single gap-flanking oligo (which essentially combines the two gapflanking DNAs) provides >100 amplified gap-sections, but still maintaining the link between any initial phenotype analysis and the sequence. The abundance of targets (>100) will decrease the time needed for binding or the concentration of fluorescence seals needed for binding by a similar fold (~100-fold).
[0081] According to another aspect of the invention, there is provided a method of singlemolecule phenotyping and sequencing of a nucleic acid molecule, the method comprising:
[0082] a) providing the nucleic acid molecule immobilised to a surface;
[0083] b) testing for a phenotype / function of the immobilised nucleic acid molecule; and
[0084] c) determining the sequence of the immobilised nucleic acid molecule such that an observed phenotype / function is linked to the sequence, wherein the nucleic acid is sequenced by the method of sequencing according to the invention described herein.In one embodiment, testing for a phenotype / function of the immobilised nucleic acid molecule may comprise testing for binding of the nucleic acid to a target molecule. The target molecule may comprise a small molecule, an oligopeptide, a polypeptide (such as an enzyme or a protein), a nucleic acid, or a lipid molecule.
[0085] The target molecule may be labelled for detection, such as fluorescently labelled. The fluorescent label of target molecule may be selected from any of the group comprising 5 -carboxyfluorescein (FAM), tetramethylrhodamine (TMR), Alexa-Fluor fluorophores (such as Alexa488, Alexa532, Alexa546, Alexa555, Alexa568, Alexa594, and Alexa647; available from Molecular Probes / Invitrogen), BODIPY dyes, ATTO dyes (such as ATTO532, ATTO568, ATTO594, ATTO647n, and ATTO655; available from Attotech) and cyanine dyes (such as Cy2, Cy3, Cy3B, Cy3.5, Cy5, Cy5.5, and Cy7; available from GE Healthcare), or combinations thereof.
[0086] In one embodiment, the binding, such as transient binding, of a fluorescently labeled target molecule to the immobilised nucleic acid is detected by measurement of Forster resonance energy transfer (FRET) between the fluorophore on the fluorescently labeled target molecule and a second fluorophore. The second fluorophore may be tagged on the immobilized nucleic acid molecule or duplex thereof. For example, the second fluorophore may be tagged on a gap-flanking nucleic acid strand that is hybridized to the immobilized nucleic acid molecule. For example, the fluorophores Cy3B and ATTO647N can be used together as a donor and an acceptor pair for FRET. Therefore, the fluorophore on the fluorescently labeled target molecule may comprise Cy3B, and the second fluorophore may comprise ATTO647N, or vice versa.
[0087] The method of the invention can be particularly useful for discovering DNA / RNA aptamers, and for investigating desirable binding / reaction profiles versus various targets (such as proteins, nucleic acids, small molecules etc.).
[0088] In one embodiment, the phenotyping / functional assay may comprise the determination of the interaction of the immobilised nucleic acid with a sequence-dependent DNA / RNA binding protein. The sequence-dependent DNA / RNA binding protein may comprise a transcription factor or a viral protein. The sequence-dependent processing of the nucleic acid by a nucleic-acid polymerase may be determined, such as by RNA polymerase or DNA polymerase. The interaction of the nucleic acid with another nucleic acid may bedetermined, e.g., another DNA or RNA strand. In an embodiment wherein the immobilised nucleic acid is an aptamer, the binding to a target protein, biomolecule or particle may be determined.
[0089] An example of functional assay that can be carried out according to the invention is the binding and subsequent dissociation of a DNA-binding protein. The protein may be fluorescent; in such a case, binding of the protein to the DNA results in a clear observable fluorescence, which is repeated binding events of the protein at the same location of the surface.
[0090] The functional / phenotypical analysis may be on the double-stranded form of the nucleic acid, i.e. the fully double-stranded form (and not the partially duplexed form), which may be provided for sequencing. In one embodiment, the nucleic acid may be provided double-stranded. In one embodiment, the nucleic acid to be interrogated for function / phenotype is the gap section of the nucleic acid to be sequenced.
[0091] Providing the immobilised nucleic acid may comprise providing a library of nucleic acids, such as DNA, which is prepared using automated DNA synthesis with degenerate nucleotides inserted at the gap section of the nucleic acid to be interrogated for function / phenotype and to be sequenced. Providing the immobilised nucleic acid may comprise the same provision as for the method of sequencing according to the invention herein.
[0092] Following the phenotypic analysis, the method may further comprise a denaturing and / or wash step, for example to remove any interacting or non-interacting molecules, prior to sequencing the nucleic acid. The wash step may comprise any suitable method for weakening binding between a nucleic acid and a bound molecule. In another embodiment, the wash step may comprise washing away unbound probes and / or molecules from the immobilised nucleic acid. In one embodiment, the wash step comprises washing the immobilised nucleic acid and its interacting partner in a high salt buffer.
[0093] The nucleic acid may be denatured to remove the top strand (i.e., the strand that is not directly immobilised / linked to the surface). Denaturing may be done chemically (e.g.using treatment with high NaOH concentration), thermally (e.g. by melting the strands by increasing temperature), or biochemically (e.g. using specialised nucleases).
[0094] An advantageous application of the method of the invention is the ability to directly read the sequence of protein-DNA / RNA (ribonucleic acid) interactions and reactions. There are many different types of sequence dependent protein-DNA / RNA interactions. Some proteins interact with DNA / RNA by simply binding such as catabolite activator protein (CAP) and carbon storage regulator A protein (CsrA). Other proteins bind to the DNA and instigate reactions such as RNA polymerase initiating transcription. Other proteins bind to DNA as complex with an RNA and instigate reactions such as CRISPR-Cas9 initiating DNA cleavage and degradation. DNA / RNA also can be used to bind proteins, for example, aptamers (three-dimensional DNA / RNA), which bind and capture specific proteins, with many biomedical applications. There is a vast array of protein-DNA / RNA sequence dependent interactions that can be investigated.
[0095] According to another aspect of the invention, there is provided a method of molecule phenotyping and identification of a molecule, wherein the molecule is tagged with a nucleic acid and identified by sequencing of the tagged nucleic acid, the method comprising:
[0096] a) providing the molecule immobilised to a surface;
[0097] b) testing for a phenotype / function of the immobilised molecule; and c) determining the sequence of the tagged nucleic acid molecule such that an observed phenotype / function is linked to the tagged nucleic acid sequence, wherein the tagged nucleic acid is sequenced by the method of sequencing according to the invention described herein.
[0098] The immobilised molecule may comprise a protein that is fused (i.e. physically linked) to a nucleic acid. The nucleic acid may comprise nucleic acid that encodes the protein, such as its encoding mRNA. Such tagging of the nucleic acid is described in Roberts et al Current Opinion in Chem Bio 1999, which is herein incorporated by reference. Tagging the encoding mRNA of the protein may be performed using an in vitro translation system where the mRNA comprises an oligo linker of about 19-30 nt and a puromycin molecule at its 3 ’end. Upon reaching the oligo linker, the puromycin forms a covalent bond between the mRNA and its corresponding protein. Such tagging allowsa direct link between measuring protein function in large libraries, and determination of the sequences associated with the desire function.
[0099] Testing for a phenotype / function of the immobilised molecule may comprise detection of an interaction, such as binding between the immobilised molecule and a target molecule.
[0100] The immobilised molecule may comprise a small molecule, an oligopeptide, a polypeptide (such as an enzyme or a protein), a nucleic acid, or a lipid molecule. The target molecule may comprise a small molecule, an oligopeptide, a polypeptide (such as an enzyme or a protein), a nucleic acid, or a lipid molecule.
[0101] The immobilised molecule and / or target molecule may be labelled for detection, such as fluorescently labelled. The fluorescent label of immobilised molecule and / or target molecule may be selected from any of the group comprising 5 -carboxyfluorescein (FAM), tetramethylrhodamine (TMR), Alexa-Fluor fluorophores (such as Alexa488, Alexa532, Alexa546, Alexa555, Alexa568, Alexa594, and Alexa647; available from Molecular Probes / Invitrogen), BODIPY dyes, ATTO dyes (such as ATTO532, ATTO568, ATTO594, ATTO647n, and ATTO655; available from Attotech) and cyanine dyes (such as Cy2, Cy3, Cy3B, Cy3.5, Cy5, Cy5.5, and Cy7; available from GE Healthcare), or combinations thereof.
[0102] In one embodiment, the binding, such as transient binding, of a fluorescently labeled target molecule to the immobilised molecule is detected by measurement of Forster resonance energy transfer (FRET) between the fluorophore on the fluorescently labeled target molecule and a second fluorophore. The second fluorophore may be tagged on the immobilized molecule. For example, the fluorophores Cy3B and ATTO647N can be used together as a donor and an acceptor pair for FRET. Therefore, the fluorophore on the fluorescently labeled immobilised molecule may comprise Cy3B, and the second fluorophore on the target molecule may comprise ATTO647N, or vice versa. The FRET pair may comprise a fluorophore and a quencher.
[0103] According to another aspect of the invention, there is provided a method of single-cell phenotyping and sequencing of a nucleic acid molecule associated with the phenotype of the cell, the method comprising:a) providing a population of cells comprising modifications in a nucleic acid sequence of interest;
[0104] b) observing a phenotype of individual cells, wherein the cells are fixed to a surface before, during or after the phenotypic observation, such that the nucleic acid sequence is also immobilised;
[0105] c) sequencing the nucleic acid sequence by the method of sequencing according to the invention herein, such that an observed phenotype / function of the cell is linked to the nucleic acid sequence.
[0106] According to all aspects, and in embodiments where a phenotype or function of the immobilised nucleic acid or associated molecule is tested or observed, determination of the nucleic acid sequence may be carried out either after or before the phenotype or function testing or observation. In particular, the sequencing step may be performed prior to the phenotype or function testing (for example, prior to step (b) in embodiments comprising testing followed by sequencing). Performing sequencing prior to phenotype or function testing may be advantageous in embodiments where the nucleic acid is subject to chemical alteration during testing or observation of the phenotype or function, for example cleavage, modification, or degradation of the nucleic acid. In such embodiments, the nucleic acid sequence determined prior to phenotype or function testing may be retained in association with the immobilised molecule or its spatial location on the surface, such that subsequently observed phenotype or function can be correlated with the previously determined sequence.
[0107] The population of cell may comprise nucleic acid sequence from a randomised library of nucleic acid modifications. The nucleic acid sequence may be a gene sequence that is modified. The library may comprise 4AN randomized sequence members (e.g. 4A5). The cells may be modified by transformation of a plasmid or nucleic acid fragment library of the modified nucleic acid sequence. Transformation of the plasmids / nucleic acid fragments may provide the population of cells, with each cell making one of many permutations of a certain protein according to the introduced 4AN randomised sequences.
[0108] The observed phenotype may comprise any of cell survival, cell killing (such as in the presence of a toxin or antibiotic), cell growth, cell activity, cell membrane transport, cell motility, chemotaxis, cell morphology, cell biochemistry, cell protein expression,cell regulation, cell surface display, cell immunogenicity, cell division; or combinations thereof.
[0109] The population of cells may comprise prokaryote or eukaryote cells. In one embodiment, the cells are prokaryote, such as E. coli.
[0110] The cell may be chemically fixed (e.g., using 1-4% paraformaldehyde), permeabilised (e.g., using treatment with ethanol) to allow access of the sequencing reagents to the nucleic acid of the cell, and then attached to the surface (e.g., to a polylysine-coated surface or to an agarose pad). Alternatively, the cells can be immobilised by capture in microfluidic channels.
[0111] According to another aspect of the invention, there is provided a nucleic acid molecule immobilised on a surface, wherein the immobilised nucleic acid molecule is partially duplexed such that it comprises a single-stranded gap-section flanked by duplex sections, and wherein the immobilised nucleic acid is fluorescently labelled.
[0112] The immobilised nucleic acid may comprise features as described herein according to the method of the invention. In one embodiment, the immobilised nucleic acid molecule is immobilised on a glass surface. The gaps section may be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 nucleotides in length. The gap section may be 20 nucleotides in length or less. In another embodiment, the gap section may be 15 nucleotides in length or less. In another embodiment, the gap section may be 12 nucleotides in length or less. In another embodiment, the gap section may be 10 nucleotides in length or less. In another embodiment, the gap section may be 9 nucleotides in length or less. In another embodiment, the gap section may be 6 nucleotides in length or less. In another embodiment, the gap section may be 5 nucleotides in length or less. The immobilised nucleic acid may comprise known / predetermined sequence and unknown or randomised sequence. The unknown or randomised sequences may be located with the gap section (non-duplexed section).
[0113] The immobilised nucleic acid may be isolated (e.g. immobilised apart from other nucleic acid molecules).According to another aspect of the invention, there is provided an array of surface-immobilised nucleic acids according to the invention herein. The array may be arranged in a regular pattern or may be irregular. The immobilised nucleic acid may be spatially separated from each other.
[0114] According to another aspect of the invention, there is provided a composition or a kit comprising a fluorescently labelled oligonucleotide reporter probe and a competitive probe arranged to compete for binding with the fluorescently labelled oligonucleotide reporter probe,
[0115] wherein the competitive probe comprises one of at least four competitive probes comprising:
[0116] (i) a nucleotide X at one position in the sequence of the competitive probes, wherein the position is the same for each of the at least four fluorescently labelled oligonucleotide probes, and
[0117] (ii) universal or degenerate bases at the remaining nucleotide positions, and wherein the at least four competitive probes comprise:
[0118] - a first competitive probe wherein nucleotide X is a cytosine;
[0119] - a second competitive probe wherein nucleotide X is a guanine;
[0120] - a third competitive probe wherein nucleotide X is an adenine; and
[0121] - a fourth competitive probe wherein nucleotide X is a thymine or uracil.
[0122] In one embodiment, the competitive probe(s) are not fluorescently labelled. In another embodiment, the competitive probe(s) are not labelled. The competitive probe(s) may be as herein described.
[0123] The composition or kit may further comprise a nucleic acid molecule (target to be sequenced / interrogated) as described herein. The nucleic acid molecule may be immobilised, or arranged to be immobilised, on a surface. The composition or kit may further comprise a nucleic acid molecule immobilised, or arranged to be immobilised, on a surface, wherein the nucleic acid molecule is partially duplexed such that it comprises a single-stranded gap-section flanked by duplex sections, and optionally wherein the immobilised nucleic acid is fluorescently labelled.
[0124] The fluorescently labelled oligonucleotide reporter probe may comprise or consist of degenerate and / or universal bases.The degenerate bases may form a randomised nucleic acid sequence. For example, the degenerate bases may not form a predetermined sequence. The skilled person will recognise that the provision of a degenerate base containing probe is the provision of an ensemble of probes having different nucleotides at the degenerate base position(s).
[0125] The fluorescently labelled oligonucleotide reporter probe may comprise features as described herein according to the method of the invention. The fluorescently labelled oligonucleotide reporter probe may comprise a second fluorophore label. Two fluorophores of a fluorescently labelled oligonucleotide reporter probe may be a FRET pair. The FRET pair on the same oligonucleotide probe may comprise a fluorophore and a quencher. The fluorescently labelled oligonucleotide reporter probe may be single stranded, such as ssDNA or analogues thereof.
[0126] The fluorescently labelled oligonucleotide reporter probe may not be a molecular beacon (e.g. having a secondary hairpin structure comprising self-complementary nucleotides).
[0127] The term “single-molecule phenotyping” used herein is understood to mean the analysis or measurement of an observable characteristic or trait of a single molecule, such as the ability of a single nucleic acid molecule to bind or process or to be processed by another molecule, such as a polypeptide.
[0128] The term “highest level of transient binding” used herein is understood to mean the highest overall time spent in the bound state for the seal DNA provided.
[0129] The skilled person will understand that optional features of one embodiment or aspect of the invention may be applicable, where appropriate, to other embodiments or aspects of the invention.
[0130] Embodiments of the invention will now be described in more detail, by way of example only, with reference to the accompanying drawings.
[0131] Figure 1: A single-molecule method that links protein-DNA interactions to DNA sequence.Left
[0132]
[0133] : a library of DNA molecules (e.g., with 45different DNA sequences, reflecting a 5 -nt randomised sequence) is deposited randomly and immobilised on a surface. Field-of-view shows diffraction limited images of the deposited single DNA molecules as imaged using a widefield fluorescence microscope. Different molecules contained different DNA sequences (e.g., green- and red-circled molecules).
[0134] Middle
[0135]
[0136] fluorescently labelled proteins of interest are added, and the single-molecule kinetics of protein-DNA interactions (its “single-molecule kinetic phenotype”) are recorded.
[0137] Right panel: After the dsDNA is denatured, an immobilised ssDNA strand remains on the surface, and is converted into a gapped DNA molecule in situ. Short oligonucleotides (DNA “seals”) are added and bind transiently to the gap in the gapped-DNA. The kinetics and extent of binding discriminate between fully complementary seals and those with mismatches, forming the basis for a short-read single-molecule sequencing assay.
[0138] Figure 2. Detecting single base-pair mismatches via transient DNA binding to gapped DNA.
[0139] A. Schematic of a fully complementary 8-nucleotide (nt) seal binding in 8-nt gapped-DNA.
[0140] B. Top: fluorescence intensity time-traces (red) for ATTO647N-labelled DNA seal binding to gapped DNA. Movies recorded using 100-ms exposures. Traces are fitted (black) using Hidden Markov Modelling (HMM), identifying bound states and measuring the corresponding dwell times. Bottom: dwell-time distribution for the bound states. The data fit best to double-exponential distributions that correspond to a long and short dwell time, n: number of binding events.
[0141] C-D. As in panels A-B, but for a G-G mismatch at position 3 in the 8-nt seal binding to an 8-nt gapped-DNA.
[0142] E. An 8-nt DNA gap on an immobilised DNA molecule is sequentially interrogated at position 3 (where the base is a G) of the gap using 4 seals thatdiffer only on the base corresponding to the interrogated position. Only the seal featuring C in position 3 shows significant binding to the gap; the remaining seals show negligible binding. Right panel: map of localisations clustering around the interrogated DNA gap molecule.
[0143] F. As in E, but for gaps featuring a T, an A, and a C at position 3 of the gap. Only the seal with the complementary base at position 3 shows appreciable binding.
[0144] G. A sequential deposition approach (that uses cycles of deposition, localisation, and bleaching) of 4 different DNA gap molecules of known sequence.
[0145] H. Time-traces of sequential interrogation of the deposited DNA (panel G) by DNA seals specific for position 3 of the gap. The extent of binding allows the identification of a C and an A base in the top and bottom molecules, respectively. I. Field of view (size 50 x 100 pm) of molecules sequenced for position 3 of the DNA gap. Circled molecules (red circle for GapA, green for GapT, blue for GapG, cyan for GapC) were identified correctly; in-square molecules were assigned incorrectly. Bright white spots: fiducial markers.
[0146] J. Base-calling accuracies for individual bases for interrogation of single sequences (i.e., using a surface that carries only a single sequence, as in panels E-F) and for interrogation of mixture of 4 sequences (as in panels G-H).
[0147] Figure 3. Single-molecule sequencing via competitive inhibition of hybridizations in DNA gaps.
[0148] A. Concept of sequencing by competitive inhibition. Left: A labelled seal (R-seal) containing degenerate bases is designed to bind well to a DNA gap and to give a fluorescence signal. Middle: in the presence of a fully complementary unlabelled seal, R-seal binding is reduced, and fluorescence signals are suppressed. Right: in the presence of a unlabelled seal with one or more misatches, R-seal binding is maintained, and fluorescence signals are observed.
[0149] B. Representative traces for identification of single nucleotide difference using the competitive-inhibition strategy. Interrogation of position 3 of the gap for a G3-gap (top) and a A3-gap (bottom) using 200 nM of labelled R-seal and the set of 4 unlabelled U-seals (all used at 600 nM). Complementary U-seals lower theextent of R-seal binding (see highlighted columns of timetraces), enabling base identification.
[0150] C. Map of the spatial locations of seal-binding localizations. Mismatch seals are associated with a large number of localisations that cluster at the same point in space; in contrast, very few localisations appear in the presence of the complementary seal. Left panels: localisations from an interrogation with a single U-seal. Right panel: pooled localisations from interrogations from all 4 seals.
[0151] D. Base-calling accuracies for individual bases for interrogation of single sequences.
[0152] E. DNA design for sequencing a 3-base sequence using competitive inhibition. F. A representative traces of 3-base sequencing using competitive inhibition. Interrogated sequence on the template DNA: GGC. In all cases, the R-seal (D5D6D7) was kept at 500 nM. A series of 4 interrogations with the 4 U-seals (added at 2 pM) “reads” a single base on the template. Interrogation with the complementary U-seal led to decreased binding (highlighted traces).
[0153] G. Base-calling accuracies for individual bases during 3-base sequencing. H. DNA design for sequencing a 5-base sequence using competitive inhibition. I. A representative traces of 5-base sequencing using competitive inhibition. Interrogated sequence on the template DNA: GGCAG. In all cases, the R-seal (D5D6D7D8D9) was kept at 2 pM. A series of 4 interrogations with the 4 U-seals (added at 2 pM) “reads” a single base on the template. Interrogation with the complementary U-seal led to decreased binding (highlighted traces).
[0154] G. Base-calling accuracies for individual bases during 5-base sequencing.
[0155] Figure 4. Linking the kinetics of protein-DNA interactions to DNA sequence at the single-molecule level: interaction of a transcriptional activator with its cognate DNA.
[0156] A. Schematic of fluorescently labelled catabolite activator protein (CAP) binding to dsDNA with a 22-bp binding sequence for CAP (yellow shading). Bottom: symmetric 22-bp consensus CAP sequence binding sequence (CAPcons).B. Fluorescence intensity time-traces (red) for Alexa647-CAP binding to CAPcons DNA. HMM analysis identifies two different states in the traces: high intensity due to CAP binding, and low intensity due to absence of CAP.
[0157] C. Dwell-time histogram for CAP binding to CAPcons; the data fit best to double-exponential distributions that correspond to a long and short dwell time. D. Three-step assay for linking kinetic phenotype to sequence. Step 1: CAP binds to surface-immobilised CAPcons dsDNA. Step 2: dsDNA denatured by NaOH wash leaving only surface-immobilised bottom DNA strands, followed by gapped-DNA formation via in situ hybridisation of left-hand (fluorescently labelled) and right-hand flanking strands. Step 3: Fully complementary or single mismatch seal interrogates the gap sequence in the gapped-DNA.
[0158] E. CAP binding trace (grey) in step 1 and complementary seal binding trace (red) in step 3 for a single DNA molecule on the surface.
[0159] F. Dwell time histograms for seal binding in step 3 of the three-step assay. Top: Complementary seal. Bottom: Single mismatch seal (C-C mismatch at position 5).
[0160] Figure 5. Linking the kinetics of protein-DNA interactions to DNA sequence at the single-molecule level: sequence-dependence of transcription initiation.
[0161] A. Schematic of the 3 main paths for the short RNAs synthesized during transcription initiation. Green path: productive extensive of RNA to enter elongation. Pink path: abortive release of RNA and restart of RNA synthesis. Cyan path: futile cycling between a stably scrunched and RNA-backtracked state (fig taken from Dulin... Kapanidis NatComm 2018 - to be redrawn). RNAP tends to pause at the ITC6 state (initially transcribing complex with 6-nt RNA). B. Formation of transcription complexes on a promoter library containing 16 variants of the lacCONS promoter after randomizing positions +6 and +7 from the transcription start site. After addition of the NTPs, transcription leads to FRET signal changes due to movement of the DNA downstream of the transcription start site. After recording transcription-associated FRET changes, the protein and template strand are removed, and the sequence at positions +6 and +7 is read using the competitive inhibition assay.C. Example of anticorrelated donor (green) and acceptor (red) fluorescence changes, consistent with single-molecule FRET changes. Different patterns of changes are seen for different molecules.
[0162] D. Examples of traces of molecules belonging to the main 4 classes of behavior observed by the active transcription complexes (see main text for descriptions). E. Distribution of initiation pause lifetimes, and single-exponential decay fitting to obtain the pause lifetime for each of the 16 variants. Right: sequence logo of the DNA element associated with initiation pausing.
[0163] F. Heat map of the pause lifetimes for all 16 sequences, based on the results in panel E.
[0164] G-E Heat map of propensity for stable-scrunching (panel G), cycling (panel H) and single-pause (panel I), shown as fraction of molecules in the population belonging to this class of behaviour.
[0165] Figure SI. Effect of base-stacking of transient binding of short DNA to immobilised gap DNA strands. The seal DNA is stabilised in the DNA gap via H-bonded interactions between complementary base pairs and by stacking between adjacent bases. To evaluate the effect of stacking on the hybridization kinetics, we constructed immobilised DNAs with 2, 1 and 0 stacking interactions (see top schematics; sequence of bottoms strand is as in Fig. 2A), and evaluated the lifetime of the bound seal DNAs. Localization of the gap has been performed using 532 nm laser and binding-unbinding of the complementary ATTO647-labelled seal was tracked at a concentration of 10 nM using 638 nm laser at an exposure time of 300 ms.
[0166] A. Top, time-traces of 8-nt seal binding to a surface-immobilised DNA featuring a DNA gap; when the seal DNA binds, it forms two stacking interactions with the bases flanking the DNA gap. Bottom, dwell-time histograms for the bound state of the seal. A single exponential fit gives a mean lifetime of ~ 5 s.
[0167] B. Top, time-traces of 8-nt seal binding to a surface-immobilised DNA featuring a DNA overhang; when the seal DNA binds, it forms one stacking interactions with the 3’ base of the shorter oligo, leading to less stable gap binding. Bottom, dwell-time histograms for the bound state of the seal. A single exponential fit gives a mean lifetime of ~ 3.4 s.
[0168] C. Top, time-traces of 8-nt seal binding to a surface-immobilised ssDNA; when the seal DNA binds, it forms no stacking interactions, leading to even less stablegap binding. Bottom, dwell-time histograms for the bound state of the seal. A single exponential fit gives a mean lifetime of ~ 2.9 s. The comparison of the results shows that stacking interactions between adjacent bases provided stability to the hybridized seal and increase its dwell on the DNA gap.
[0169] Figure. S2. Kinetic rates for binding of 8 nt seal in the corresponding GAP for varying combinations of nucleotides
[0170] A. Dwell times of binding between 8-nt seal varying all four nucleotides at position 3 and an 8-nt gap with four varying nucleotides at the corresponding anti -position on the gap. Complementary nucleotides exhibit longer dwell times compared to the non-complementary ones.
[0171] B. Binding on-rates for the same 8-nt gap DNA with 8-nt seals of varying base compositions. Binding on-rates are also higher for fully complementary seals compared to 1-nt mismatch seals.
[0172] C. The dwell times and on-rates were used to create a similar heat map for equilibrium dissociation constants (Kd values).
[0173] Figure S3. Quantification of seal binding via localization counting as a means of base calling. Representative bar plots for localization counting for seal binding to different gap molecules featuring a different base at position 3 of the gap (Gap-3G, Gap-3A, Gap-3C, and Gap-3T).
[0174] A. Examples of three Gap-3G molecules sequentially interrogated with 4 labelled seals. We observe much higher number of localizations for the complementary seal (seal 3C). Inset, map of the spatial locations of seal-binding localizations; localisations due to the complementary seal are clustered in nearcircular pattern; occasionally, mismatch seals bind to a proximal (but different) location (e.g., in the first example, seal 3T binds -100 nm away from the main cluster), leading to background; these results also point to the possibility of using spatial filtering to improve base-calling accuracy.
[0175] B. Similar examples for gap molecules Gap-3A, Gap-3C, and Gap-3T.
[0176] Figure S4: Standardization of conditions for single base sequencing using competitive approach.
[0177] (a) Dwell time distributions of binding of 250 nM of 8-nt R-seal with a single degenerate base in presence of 100 nM, 250 nM and 500 nMcomplementary unlabelled seal in imaging buffer supplemented with 10 % Dextran Sulphate. Histograms are fitted with bi-exponential decay model (red lines).
[0178] (b) Dwell time distributions of binding of 250 nM of 8-nt R-seal with a single degenerate base in presence of 100 nM, 250 nM and 500 nM unlabelled seal with single nucleotide mismatch in an imaging buffer supplemented with 10 % Dextran Sulphate. Histograms are fitted with bi-exponential decay model (red lines).
[0179] (c) Average number of bindings per trace obtained for the above measurements. In all cases, the R-seal is at 250 nM.
[0180] (d) Average dwell times obtained from above measurements with varying concentration of unlabelled seals. In all cases, the R-seal is at 250 nM.
[0181] Figure S5. Representative traces for identification of single nucleotide difference using the competitive-inhibition strategy. Interrogation of each of the 4 gaps used 200 nM of labelled R-seal and the set of 4 unlabelled U-seals (all used at 600 nM). Complementary U-seals lower the extent of R-seal binding (see highlighted columns of timetraces), enabling base identification.
[0182] Figure S6. Verification of identifying single nucleotide in a mixed population with prior information. All four gaps were deposited sequentially, followed by registration of their location on the surface, and photobleaching. The DNA gaps were then interrogated using the competitive-inhibition assay using 200 nM of 8 nt degenerate R-seal and four cycles of interrogation with the competing U-seals with four different bases, each at 600 nM.
[0183] A. Representative traces for interrogation depicting traces featuring varying propensity of binding at different interrogation cycles with different GAP molecules.
[0184] B. Circled molecules (red circle for GapA, green for GapT, blue for GapG, cyan for GapC) were identified correctly; in-square molecules were assigned incorrectly. The base-calling accuracy was ~98 %.
[0185] C. Table with the accuracies for individual bases.
[0186] Figure S7. Optimizing formamide concentration for 13-nt seal binding.
[0187] Binding of a 13-nt degenerate seal (with 3 degenerate bases) to a 13-nt gap leadsto longer binding than the 8-mer (Fig. 1), yielding a small number of on-off cycles, which is key for sequencing. We thus increased formamide concentration in the imaging buffer to destabilise the seal binding and examined formamide concentrations of 10%-25%.
[0188] A. Representative traces from hybridization of 500 nM 13 -nt labelled degenerate R-seal in 10% dextran sulphate and formamide concentrations of 10% (black), 15% (red), 20% (blue), and 25% (green). Increasing formamide shortens dwell times in the bound (high-intensity) state.
[0189] B. Dwell-time histograms for the bound states at different formamide concentrations.
[0190] C. Mean dwell times in the bound states (top) and number of binding -unbinding cycles per gap molecule (bottom) as a function of formamide concentration. Formamide concentrations of 15% and 20% provide an optimal combination of dwell time and number of binding -unbinding cycles.
[0191] Figure S8. Optimization of the complementary U-seal concentration for 3-base sequencing using the competitive-inhibition assay. In the presence of 500 nM R-seal (D5D6D7), the unlabelled complementary seal (U-seal C5D6D7) concentration was varied from 0.5 to 3.5 pM to identify the concentration where sufficient competition exists to suppress R-seal binding.
[0192] A. Representative traces for the binding of the R-seal to the gap in the presence of increasing complementary U-seal. Increasing the U-seal concentration substantially decreases the R-seal binding.
[0193] B. Dwell-time histograms for the bound states at different complementary U-seal concentrations.
[0194] C. Mean dwell times in the bound states (top) and number of binding -unbinding cycles per gap molecule (bottom) as a function of complementary U-seal concentration; at a U-seal concentration of >2 pM, there is substantial decrease in both R-seal binding indicators, providing an operational regime for the competitive-inhibition assay.
[0195] Figure S9. Optimization of the mismatch U-seal concentration for for 3-base sequencing using the competitive-inhibition assay. In the presence of 500 nM R-seal (DDD), the unlabelled 1-nt mismatch seal (U-seal T5D6D7) concentration was varied from 0.5 to 3.5 pM to identify the concentration where sufficientcontrast exists between the competition exerted by the complementary and non-complementary U-seals.
[0196] A. Representative traces for R-seal binding to the gap in the presence of increasing mismatch U-seal. Increasing the U-seal concentration also slightly decreases R-seal binding, but to a much lower degree than the complementary U-seal.
[0197] B. Dwell-time histograms for the bound states at different mismatch U-seal concentrations.
[0198] C. Mean dwell times in the bound states (top) and number of binding -unbinding cycles per gap molecule (bottom) as a function of mismatch U-seal concentration; at a U-seal concentration of >2 pM, there is substantial contrast between the complementary and mismatch U-seals, providing an operational regime for the competitive-inhibition assay.
[0199] Figure S10. Representative traces of 3-base sequencing using competitive inhibition. Additional example traces for 3-base sequencing are provided for gap positions 5 (panel A), 6 (panel B), and 7 (panel B). Conditions and descriptions as in Fig. 3F. In all cases, the R-seal (D5D6D7) was kept at 500 nM. Interrogation with the complementary U-seal led to decrease binding (highlighted traces).
[0200] Figure Sil. Optimization of the complementary U-seal concentration for 5-base sequencing using the competitive-inhibition assay. In the presence of 1 pM R-seal (D5D6D7D8D9), the unlabelled complementary seal (U-seal D5C6D D8D9) concentration was varied from 1 to 5 pM to identify the concentration where sufficient competition exists to suppress R-seal binding.
[0201] A. Representative traces for the binding of the R-seal to the gap in the presence of increasing complementary U-seal. Increasing the U-seal concentration substantially decreases R-seal binding.
[0202] B. Dwell-time histograms for the bound states at different complementary U-seal concentrations. The fit was done using a single-exponential decay model.
[0203] C. Mean dwell times in the bound states (top) and number of binding -unbinding cycles per gap molecule (bottom) as a function of complementary U-seal concentration; at a U-seal concentration of >2 pM, there is substantial decreasein both R-seal binding indicators, providing an operational regime for the competitive-inhibition assay.
[0204] Figure S12. Optimization of the mismatch U-seal concentration for 5-base sequencing using the competitive-inhibition assay. In the presence of 1 pM R-seal (D5D6D7D8D9), the unlabelled 1-nt mismatch seal (U-seal D5T6D7D8D9) concentration was varied from 1 to 6 pM to identify the concentration where sufficient competition exists to suppress R-seal binding.
[0205] A. Representative traces for R-seal binding to the gap in the presence of increasing mismatch U-seal. Increasing the U-seal concentration decreases R-seal binding, but to a much lower degree than the complementary U-seal.
[0206] B. Dwell-time histograms for the bound states at different mismatch U-seal concentrations. The histograms were fitted using a bi-exponential decay model, indicating the presence of multiple binding kinetics.
[0207] C. Mean dwell times in the bound states (top) and number of binding -unbinding cycles per gap molecule (bottom) as a function of mismatch U-seal concentration. The average dwell time of the R-seal initially decreases sharply with increasing concentrations of the mismatch U-seal, followed by a slight decrease at higher concentrations. The average number of binding-unbinding events per molecule initially increases with rising mismatch U-seal concentrations (promoted by reduction in dwell time), followed by a decline at higher concentrations. Along with Fig. Sil, the results suggest using an equimolar ratio of R-seal to U-seals to achieve good contrast between the complementary and mismatch U-seals.
[0208] Figure S13. Representative traces of 5-base sequencing using competitive inhibition. Additional examples for 5-base sequencing are provided for gap positions 5 to 9 (panels A-E). Conditions and descriptions as in Fig. 31. In all cases, the R-seal (D5D6D7D8D9) was kept at 2 pM, and the U-seals at 2 pM. Interrogation with the complementary U-seal led to decrease binding (highlighted traces).
[0209] Figure S14. CAP-DNA interactions: dwell-time distribution of unbound states. Off-time distributions for CAP-DNA interactions with the consensus sequence(panel A). Data fit best to double-exponential distributions that correspond to a long and short dwell time, n: number of binding events.
[0210] Figure S15. Single-molecule analysis of CAP binding to mutant sequences CAPcons sequences.
[0211] A. Top, three representative time-traces showing binding of CAP to theCAPmutl sequence. Bottom, dwell-time distributions in the bound and unbound states, along with fitting to double-exponential decays.
[0212] B. Top, three representative time-traces showing binding of CAP to the CAPmut2 sequence. Bottom, dwell-time distributions in the bound and unbound states, along with fitting to double-exponential decays.
[0213] Figure S16. Preparation of library of 16 labelled lacCONS promoter DNA variants. The library was prepared in 3 steps. Step 1 involves a primer extension of a Cy3B-modified strand on a template strand with degenerate bases at sites corresponding to positions +6 and +7 from the transcription start site; the extended product is then purified. Step 2 involves preparation of another duplex fragment, wherein the bottom strand is labelled with ATTO647N at the site that corresponding to position +20 form the transcription start site. In Step 3, the two dsDNA fragments are ligated using a high-performance T4 DNA ligase.
[0214] Figure S17. Time evolution of FRET efficiencies for transcription complexes at different conditions. FRET time-traces have been superimposed to produce heat maps that show the time evolution of the FRET signals.
[0215] A. Open complexes (RPo) prepared using the labelled promoter library and RNAP were immobilized on surface and imaged using alternating-laser excitation to obtain their EFRET. The mean EFRET was -0.22 and does not change during the recording time.
[0216] B. Upon adding 500 pM ApA and 200 pM of all NTPs to the surface-immobilized RPo complexes, a large fraction of the immobilised RPo show an increase in their FRET efficiency in a matter of a few seconds due to transcription reaction.
[0217] C. Blocking transcription initiation using rifampicin leads to a profile and mean EFRET similar to RPo, confirming that the FRET change in the presence of NTPs is due to the transcription reaction.Figure S18. Example traces obtained from SPIN-seq analysis of the promoter library.
[0218] A. Left, Donor and acceptor intensities upon donor excitation exhibiting anticorrelated changes characteristic of FRET interactions between the donor and acceptor fluorophores. Right, FRET efficiency variation with time showing transition between lower FRET value (-0.25) and paused state (-0.38).
[0219] B. Sequencing +6 position using a labelled R-seal (D+6D+7) and a set of competing U-seals (A+6D+7, T+6D7, G+6D7, C+6D7). The interrogation step with the T+6D+7seal shows the least R-seal binding, showing that the interrogated molecule contains an A at position +6.
[0220] C. Similar interrogation at position +7 for the same molecule finds that the interrogation step with the D+6G+7seal shows the least R-seal binding, showing that the +7thposition contains a C at position +7.
[0221] D. A second example of a transcription complex showing dynamics, and the sequencing of its DNA promoter variant (A+6G+7).
[0222] Figure S19. Verification of sequence-dependence effects on transcription initiation. Analysis of two lacCONS promoter variants labelled at -15 and +20 position with Cy3B and Atto647N, respectively (as for the promoter library). Variants lacCONS-CG and lacCONS-AT were associated with long and short pause duration, respectively, as seen using SPIN-Seq.
[0223] A. Pause-duration distribution (top) and distribution between classes (bottom) for variant lacCONS-CG (having C and G at positions +6 and +7, respectively). Single-exponential fitting analysis shows of pause lifetime of -38 s. The variant is associated with tendency for stable scrunching and cycling.
[0224] B. Pause-duration distribution (top) and distribution between classes (bottom) for variant lacCONS-AT (having A and T at positions +6 and +7, respectively). Single-exponential fitting analysis shows of pause lifetime of -6.5 s. The variant is associated with tendency for short pausing.
[0225] Figure S20. Linking kinetics of protein-DNA interactions to DNA sequence: transcription factor binding to its cognate DNA. (A) Structure of Escherichia coli catabolite activator protein (CAP) bound to its consensus DNA site (1CGP); red segments denote the 5-bp sequences conferring sequence specificity. (B)CAPcons sequence. (C) Labelled CAP binds to DNA variants transiently, yielding on- and off-rates. (D) Left: CAP binding to CAPcons DNA. HMM analysis identifies a high-intensity state (CAP-DNA complex), and a low-intensity state (free DNA). Right: dwell-time and off-time distributions fit best to double-exponential distributions; the longer lifetimes (t2) identify the lifetimes for stable binding and the frequency of stable complex formation. (E) As for (D), but for CAPcons variant with C:G at position 5. (F) Linking kinetic phenotype to sequence. After evaluating CAP binding to surface-immobilised CAPcons variants, the DNA is denatured, and gapped-DNA is formed via in situ hybridisation of two flanking strands; gap position 5 is then read. (G) CAP binding to variant CAPcons library. (H) Gap interrogation of the two DNA molecules in panel (G) using competitive inhibition; the variants are identified as 5G (top) and 5C (bottom). (I) Dwell-time histograms for bound CAP for CAPcons variants. (J) Dwell-time histograms for unbound CAP for CAPcons variants. (K) Dwell times, association rate constants kon, and Kd values for CAPcons variants; error bars: S.E.M. from 2 experiments.
[0226] Fig. S21. Sequencing an 8-nt gap-DNA containing a stretch of 8 A bases. A.
[0227] Representative traces obtained from four sequential interrogations using four seals possessing A, T, G and C at gap position 3, and all T bases at all other positions. We observed many more binding-unbinding cycles for seal T3(complementary seal) compared to the other three 1-nt mismatched seals. B. We performed localization counting and base calling for all gapped DNA molecules, showing an -97.3% accuracy at a confidence of 0.3, calling -850 molecules.
[0228] Fig. S22. Efficiency and functionality of in-situ DNA gap formation.
[0229] A. Localisation image of surface-immobilised 8-nt gapped DNA molecules. B.
[0230] Localisation image of in-situ formed surface-immobilised 8-nt gapped DNA molecules (same field-of-view). C. Superimposed images of panels A and B; colocalised molecules appear yellow. D. Cross-colocalization of flanking strands; cross-colocalised pairs of flanking strands on a single gap molecule appear yellow. For details, see SI text. E. In black: time-traces of transient seal hybridization to pre-formed surface-immobilised gapped DNA. In red: timetraces of transient seal hybridization to the in-situ formed gapped DNA. The binding patterns pre-formed and in-situ formed DNA gaps are similar.Fig. S23. Assay performance as a function of interrogation times. A. Basecalling accuracy as a function of interrogation time. Accuracy was evaluated for detecting base 3 on the Gap8-G3template using the non-competitive mode. The time refers to the duration of each interrogation with one of the 4 seals. B-C.
[0231] Binding capability of DNA molecules over repeated interrogations. Gap-A3molecules were probed using an ATTO647N-labeled complementary strand (seal T3) across eight interrogation cycles, each lasting 5 min, followed by a washing step; 877 molecules (-97%) exhibited characteristic binding-unbinding events across the interrogation traces (examples in panel B), while 29 molecules (-3%) lost the binding-unbinding signature during one of the interrogation cycles (examples in panel C). D. Number of lost molecules per interrogation cycle.
[0232] Example 1 - Single-molecule phenotyping and in-situ sequencing for mechanistic analysis of protein-DNA interactions and reactions
[0233] Summary
[0234] DNA sequence orchestrates and regulates the mechanisms of complex reactions and interactions on DNA; yet, sequence effects are not well understood due to the lack of direct approaches to study sequence-dependence at high throughput at the levels of individual molecules. Here, we introduce Single-molecule Phenotyping and In-Situ Sequencing (SPIN-Seq), a method that links functional and structural properties of a single DNA molecule (its “single-molecule phenotype”) with the sequence of the same molecule, on the same surface and instrument. Our approach uses DNA libraries and requires no amplification. SPIN-Seq starts by recording the dynamics of a protein-DNA interaction using single-molecule imaging, followed by a new sequencing-by-hybridization approach that reads the sequence of each immobilised DNA molecule, and links it to its single-molecule phenotype to identify patterns of behaviours and gain mechanistic insight. Our assay is demonstrated to read sequences of up to 5 nucleotides in length, covering libraries of up to 1024 different DNA sequences. To demonstrate the capabilities of SPIN-Seq, we have shown its ability to characterise the interaction of a transcription factor with its target sequence, as well as the sequence-dependence of transcriptional pausing and related reaction paths during transcription initiation kinetics. The method can be further extended to other nucleic acids and interaction,providing a powerful new way to dissect molecular mechanisms and their sequence dependence.
[0235] Introduction
[0236] Protein-DNA interactions are key for genetic processes and biotechnological applications. Analysis of such interactions relies heavily on structural, biochemical, and biophysical approaches; in particular, single-molecule methods offer key insight on kinetic and structural aspects of protein-DNA interactions, often capturing them in realtime1-3’7’8. However, typical single-molecule experiments study one sequence at a time, and cannot offer a global view of the multitude of ways DNA sequence affects protein-DNA interactions. Such a comprehensive view can arise from sequence-multiplexed single-molecule approaches, which may also decipher effects of individual bases on specific steps of a process, and identify sequences for specific applications.
[0237] An attractive route to explore DNA-sequence-dependence at single-molecule level involves coupling commercial single-molecule DNA sequencing (nanopore- or fluorescence-based) to analysis of protein-driven processes on DNA strands being sequenced9“1112; however, these approaches are confined to proteins translocating on nucleic acidsn, and may require special surfaces9. Recently, two powerful singlemolecule approaches enabled multiplexing on DNA-DNA and protein-DNA interactions14 15; both involved immobilisation of DNA libraries on Illumina chips, single-molecule fluorescence imaging on custom-built microscopes, DNA amplification, Illumina sequencing, and linking of the single-molecule and sequence readouts. These methods offered insight on how DNA controls the structure and dynamics of folded DNA and protein-DNA complexes. However, these methods have not addressed labelled proteins or non-equilibrium reactions and their kinetics, mainly due to the need for DNA amplification and their dependence on a solid support (MiSeq chips) suboptimal for single-molecule imaging.
[0238] Here, we build on our early work13to introduce Single-molecule Phenotyping and InSitu Sequencing (SPIN-Seq; Fig. 1), a method that links the functional properties of a single DNA molecule (its “single-molecule phenotype”) with the sequence of the same molecule, on the same surface and instrument. The single-molecule phenotype involves structural and dynamical aspects, e.g., kinetics of biochemical reactions, reversible binding, or distances within a protein-DNA complex. Our approach uses DNA librariesand requires no amplification. SPIN-Seq starts by immobilising DNA libraries randomly on glass coverslips optimal for single-molecule imaging; the immobilised DNAs interact with proteins and the dynamics are recorded using single-molecule fluorescence; finally, proteins are removed, and a new sequencing -by-hybridization approach reads the sequence of each immobilised DNA molecule on the same instrument, and links it to its single-molecule phenotype to identify patterns of behaviours and gain mechanistic insight. Our assay has been shown to read sequences of up to 5 nucleotides in length, covering libraries of up to 1024 different DNA sequences and opening a new avenue for exploring sequence-dependent protein-DNA interactions.
[0239] Results
[0240] Single-molecule hybridization on DNA gaps. To enable in situ single-molecule sequencing, we introduced Gap-Seq, a method that reads short DNA sequences using transient binding of fluorescent single-stranded (ss) DNA oligos (“seals”) to surface-immobilised DNAs featuring short ssDNA gaps (Fig. 2A). Such transient binding is also used in other high-throughput endeavours, such as super-resolution imaging using DNA-PAINT (15), which uses ssDNA as a binding template. Here, we selected gapped-DNA templates instead of ssDNA, since the coaxial stacking provided by bases at the two gap ends stabilises the bound seal; the gap also defines clearly the short segment to be sequenced. Base-calling relies on identifying gap binding by fully complementary DNA seals versus gap binding by DNA seals with one or more mismatches. The short gap length (6-13 nt) allows bound seals to thermally dissociate and be replaced by seal molecules from the surrounding solution; repeated seal binding increases statistics and interrogation accuracy, while eliminating photobleaching via abundant supply of unbleached seals.
[0241] To test our ability to deduce sequence information in gapped-DNA substrates by transient seal binding, we examined the binding between a surface-immobilised DNA featuring an 8-nt gap, and labelled 8-nt seals fully complementary to the gap (Fig. 2A). Upon seal binding to the gap, the local fluorescence intensity jumps to a high value lasting until a seal dissociates, at which point the signal returns to background levels (Er rpr!, Rfefcrence source not
[0242]
[0243] top). The dwell-time distribution for the complementary seal binding (Fig. 2B, bottom) fits well to a double-exponential decay with dwell times ti ~ 0.95 s and t2 ~ 11.9 s (due to partial and complete seal binding,respectively; Fig. SI). Similar experiments with a labelled 8-nt seal with a 1-nt mismatch to the position 3 of the gap (introducing a G-G mismatch, Fig. 2C) showed >5-fold less stable binding (t2 of -2.2 s; Fig. 2D) than the complementary seal, establishing that a single-nt difference can drastically affect seal binding (as shown for different short DNA hybrids; Cisse NSMB), and offering a basis for short-read singlemolecule DNA sequencing.
[0244] Sequencing single bases via sequential transient DNA hybridization. We then explored reading one base at a specific position on a single immobilised gapped-DNA molecule. Using the approach in Fig. 2B with minor modifications (use of 5% dextran, 10% formamide, and 0.05% Tween20; Methods), we observed frequent binding of an 8-nt seal to a fully complementary 8-nt gap (seal3C, forming a G-C pair with a G on position 3 of the gap; green trace, Fig. 2E). Sequential interrogation of the same molecule by seals with a 1-nt mismatch at position 3 (seal3A, then seal3T, then seal3G) showed substantially reduced binding (Fig. 2E), due to reduced dwell times and reduced on-rates (as seen by analysing many molecules; Fig. S2). We obtained similar results for all remaining bases on position 3 of the gap (Fig. 2F, Fig. S2), establishing the generality of the approach.
[0245] To identify the complementary seal within the set of 4 seals, and “read” the interrogated base on a single immobilised molecule, we quantified the extent of binding for each seal by counting the total number of single-molecule localizations of a given seal to a single gap molecule (“localisation counting” or LC; Methods), and selecting as the complementary seal the one leading to the highest number of localizations (Fig, 2F, right; Fig. S3C). This analysis showed 1247 correct and 6 incorrect identifications of a G-base at position 3 of the gap. We performed the same analysis for the remaining 3 bases for position 3 in the gap; as before, the time-traces for the complementary sequences (A-T, T-A, C-G) showed the strongest binding and identified the correct base with -98.5% accuracy (Fig. 2F, Fig. S3B-D; Table 2J). These results establish that we can use sequential hybridization of unknown bases and subsequent time-trace analysis (hereafter, Gap-Seq) to read the base at a single interrogated position with high accuracy.
[0246] To test the Gap-seq performance on a mixed population of sequences, we deposited all four single-nucleotide variants (for position 3 of the gap) at an equimolar ratio on asingle surface. To evaluate our base-calling accuracy, we surface-immobilised the 4 variants sequentially in 4 rounds (with each round comprising gap deposition, singlemolecule localisation, and photobleaching); this approach provides the ground truth on the location of each molecule of each variant (Fig. 2G). We then interrogated the surface sequentially using the four labelled 8-nt seals, recorded transient DNA binding for each interrogation (Fig. 2H), analysed traces using localization counting to deduce base identity, and mapped the predicted sequences to the pre-registered sequence identities (Fig. 21). We identified 343 out of 348 gaps correctly (98.6% accuracy; Fig. 2J), establishing that we can read a base at a single interrogated position in a mixed population with high accuracy.
[0247] Sequencing single bases using competitive inhibition of hybridization. To extend Gap-seq to multiple bases required overcoming two challenges. First, we need to account for the unknown sequence context around each interrogated position (e.g., 2 bases upstream and 2 bases downstream of position 3 in an unknown 5-nt sequence). Second, sequencing 5 bases would require 20 labelled seal DNAs, which may also need to be Anorogenic; such a requirement is costly, while placing a concentration limit on the use of labelled seal DNAs.
[0248] To overcome these challenges, we introduced a competitive-inhibition Gap-seq variant that uses mixtures of DNAs containing degenerate bases. Each interrogation step uses a mixture of two seal DNAs that compete for gap binding: the first is always a reporter seal (R-seal; either singly labelled and Auorescent prior to binding, or Anorogenic and quenched prior to binding) that binds to the gap well on its own (Fig. 3A, left). The R-seal contains degenerate bases at all interrogated base positions; use of a degenerate base denotes a 25% probability for each of the 4 natural bases to be present in every position. The second seal is unlabelled (U-seal), and is used at a concentration that, when fully complementary to the gap sequence, outcompetes the R-seal, and suppresses R-seal binding and its associated Auorescence (Fig. 3A, middle). However, if a U-seal contains even a single mismatch to the gap sequence, then the R-seal outcompetes the U-seal, leading to high-Auorescence dwells (Fig. 3A, right). For every sequence position examined, we use one R-seal and four U-seals, each containing one of the 4 standard bases at the position corresponding to the interrogated site. Each sequence position is examined via a set of 4 sequential interrogations, each involving a mix of the R-seal with one of the U-seals (e.g., R+U(A), then R+U(C), then R+U(G), thenR+U(T)); for each set of 4 interrogations, we expect to see one trace with reduced R-seal binding (corresponding to the complementary U-seal) and three traces with more prominent R-seal binding (corresponding to the mismatched U-seals). The interrogation showing reduced binding identifies the interrogated base.
[0249] To test our approach, we examined an 8-nt gap using a set of an 8-nt R-seal with a degenerate base corresponding to position 3 in the gap, and four 8-nt U-seals covering all possible bases at position 3 of the gap. We first optimised R-seal binding; using 200 nM R-seal and a buffer with 10% dextran and 10% formamide (Methods), we observed an average of 6 binding-unbinding cycles per 5 min, with dwell times (in the bound state) of 2.4 s and 8 s (Fig. S4). We also established that a mixture of 250 nM R-seal and 500 nM U-seals (Fig. S4C-D) maximises the “contrast” for the competitive assay, since these concentrations allow the complementary U-seal to prevent R-seal binding, while not allowing the mismatch U-seals to compete well and prevent R-seal binding. We then interrogated the 8-nt gap sequence Gap3G using competitive-inhibition by sequentially tracking R-seal binding while in a mixture with each of the four U-seals. For Gap3G, we observed more R-seal binding for the mixtures containing U-seals for 3A, 3T, and 3G, and little R-seal binding for the mixture containing U-seal for 3C (Fig.
[0250] 3B-C, Fig. S5). Using UC analysis, we detected 624 out of 628 molecules exhibiting more than 5-fold higher R-seal binding for the U(A), U(T), and U(G)-seals compared to that for the U(C)-seal, yielding an -99% base-calling accuracy. We also obtained high accuracy (-98.5%) for the remaining 3 bases (Fig. 3D, Fig. S5).
[0251] We repeated the experiment with mixed sequences using competitive inhibition and achieved an overall accuracy of -97% (Fig. S6). These experiments clearly establish that we can use competition inhibition to identify single bases for all four different bases with high accuracy that matches that of the original approach.
[0252] Reading sequences of up to 5 bases. We then set out to sequence 3 consecutive bases using competitive inhibition; such a length requires increasing the number of degenerate bases in the R-seal, and increasing the gap length (to compensate for some loss of complementarity due to the additional degenerate bases). We thus generated a 13 nt gap (Fig. 3E) and a 13-nt R-seal with three degenerate bases corresponding to positions 5, 6, and 7 of the gap. Increasing the number of degenerate bases decreases the effective concentration of the fully complementary seal (e.g., for seals with 3 degenerate bases,only one in 64 sequences is fully complementary), lowering the hybridization on-rate. To maintain frequent but short R-seal binding to the gap, we increased the R-seal concentration in the solution, while quenching the ATTO647N fluorescence in the unbound R-seal via a BHQ1 dark quencher at the R-seal 3’ end (Kummerlin 2025) and modified our hybridization buffer (by adding 250 pM spermidine to increase the on-rate of binding and increasing formamide to 15% to shorten the dwell time of binding; Fig. S7 and Methods). We modified our U-seal strategy to include degenerate bases at the two remaining unknown positions; e.g., for interrogating position 5 within a gap with unknown sequence at 5, 6 and 7, we used a set of 4 U-seals: U(A5D6D7), U(C5D6D7), U(G5D6D7), and U(T5D6D7), where D is a degenerate base, and the superscript denotes positions in the gap. Finally, we optimised the R-seal and U-seal concentrations to maximise the contrast between complementary and mismatch U-seals (Fig. S8-9).
[0253] To read the base at position 5 on the 13 -nt gap, we performed four sequential competitive-inhibition interrogations using mixtures of the 13 -nt R-seal and one of the four U-seals above (Fig. 3F). We observed markedly decreased binding only with the U(C5D6D7) seal (green trace), identifying the correct base (G) in the gap for 129 out of 132 molecules (-98% accuracy). Similar interrogation for gap positions 6 and 7 (“basewalking”) using the same R-seal and two new sets of 4 U-seals led to similarly high accuracies (98-99%; 142 / 145 for position 6 and 152 / 154 for position 7; Fig. 3F-G, Fig. S10); in total, the 3-nt sequencing required 12 interrogation steps, each lasting 5 min.
[0254] To test the limits of our sequencing approach, we explored reading a 5-base sequence. For this, we utilized the same 13 -nt gap, and increased the number of degenerate bases in the R-seal and the U-seals to 5 and 4, respectively (Fig. 3H). Due to presence of 5 degenerate bases in the labelled seal, the probability of a fully complementary strand in the set of degenerate sequences becomes 1 in 1024, which decreases the on-rate drastically. We thus increased the R-seal concentration (to 2 pM) and adjusted the concentration of the U-seals (to 2 pM; Fig. SI 1-12) to achieve both sufficient R-seal binding and sufficient contrast in the competition by different U-seals; our assay was performed in an imaging buffer similar to that used for 3 -base sequencing (using 10% dextran, 15% formamide, and 250 pM spermidine). Using these conditions, we interrogated position 5 in the gap using sequential addition of R-seal / U-seal mixtures. We found that the U(C5D6D7D8D9) seal decreased R-seal binding substantially in comparison to the other 3 U-seals (Fig. 31; Fig. S13), correctly detecting the G base atposition 5 of the gap in 166 out of 167 molecules (-99% accuracy). We made similar observations for the remaining four positions, achieving accuracies of 96-98% (Fig. 3J). Our results establish that we can sequence up to 5 bases with high accuracy, paving the way for sequencing >1,000 different sequences on the same surface at the singlemolecule level, and in a manner fully compatible to functional analysis of sequences at the single-molecule level.
[0255] Linking real-time protein-DNA interactions to the sequence of single DNA molecules. To link a single-molecule phenotype to DNA sequence, we first examined the sequencedependent binding of transcriptional activator (catabolite activator protein, CAP) to DNA and connected the affinity of CAP-DNA binding to the DNA sequence at the single-molecule level. In the presence of cyclic AMP, CAP binds with high affinity to a 22-bp consensus sequence (5’- AAATGTGATCTAGATCACATTT-3’, CAPcons DNA; Fig. 4A). To observe CAP binding to CAPcons DNA at the single-molecule level, we immobilised a biotinylated Cy3B-labelled CAPcons DNA on neutravidin-coated PEGylated slides and localised the DNA using 532-nm excitation. We then added 1.5 nM Alexa647-labelled CAP and observed CAP binding to the DNA using 640-nm excitation (Fig. 4B). The dwell-times for CAP on the DNA fit to a bi-exponential distribution (Fig. 4C), with the long lifetime t2 - 8.5 s reflecting the stable CAP binding to the full CAPcons DNA site (kOff -0.12 s’1). Analysis of the frequency of binding (Fig. S14) yielded an association rate constant kon-9.73 x 106M-1s-1, which corresponds to a binding rate of 0.012 s’1and a Kd of -12 nM.
[0256] To examine the effect of base substitutions on CAP binding, we examined two CAPcons variants (Fig. S15). A single base substitution (a G / C-to-C / G pair conversion at position 5 of the CAPcons site, which is known to contact CAP) led similar transient events (Fig. S15A), but with a shorter dwell time t2 - 5.2 s, and a Kd of - 22 nM; a double substitution (symmetric conversions at positions 5 and 18 of CAPcons) further reduced the CAP dwell time to a t2 - 3.3 s (Fig. S15B) and a Kd of - 30 nM. These results showed that the effects of substitutions can be observed directly and quantitatively.
[0257] We then linked the single-molecule phenotype (here, the CAP-binding kinetics) for a single DNA molecule to the sequence of the same single DNA molecule (Fig. 4D). In step 1, we evaluated the CAP-DNA binding as above. For step 2, the dsDNA was denatured and washed using 25 mM NaOH, removing the protein and top DNA strand,and leaving only the biotinylated bottom DNA strand on the surface; subsequently, two DNA strands were hybridised in-situ to the bottom strand to form an immobilised gapped DNA with a 9-nt ssDNA gap, which mapped to the first 9 bases of the CAPcons sequence.
[0258] In step 3, we performed transient DNA binding measurements for the same field-of-view used in step 1. We added 100 nM complementary 9-nt seal DNA, recorded timetraces, and linked CAP-DNA binding events (from step 1) to seal-binding events for the same molecule (Fig. 4E). The complementary seal bound with a t2 of -3.3 s, whereas a single-nt mismatch seal (with a C-C mismatch at position 5) showed a reduced dwell time t2 of ~1.7 s; this large difference in seal binding should be sufficient for sequencing DNA libraries. These experiments established our ability to be able to work with labelled proteins while also connecting protein-DNA interactions with DNA sequence at the single-molecule level.
[0259] Sequence-dependence of transcription initiation mechanisms. To test the ability of SPIN-Seq to address a complex non-equilibrium protein-driven reaction on DNA, we applied it to bacterial transcription initiation, largely controlled by sequence-specific interactions between the DNA promoter region and RNA polymerase (RNAP). Singlemolecule studies (1-3) showed that, during initiation, RNAP can pause for several seconds after synthesizing a 6-mer RNA, and then enter a brunched kinetic pathway (Fig. 5A) towards either an abortive path (where the short RNA dissociates), or a productive path (where the short RNA is extended to a full-length transcript; Dulin NComm2018). The initial transcription sequence affects initiation pausing (2); NGS-based work (4) identified a sequence motif, (C / T)+6G+7(with numbering denoting a position relative to the transcription start site) which resembles the G-10Y-1G+1elemental-pause sequence responsible for pausing during transcription elongation (5, 6). However, the sequence-dependence of initiation pausing kinetics, and their association to other paths during initial transcription are unknown.
[0260] To examine how DNA sequence affects pausing on the lacCONS promoter (rate-limited during initial transcription), we generated a library of 16 promoter variants (using all base variations at positions +6 and +7) using synthesis of a short ssDNA library, primer extension, ligation, and gel purification (Fig. S16). To monitor RNA synthesis and the onset of pausing, we used a smFRET strategy we established before on the samepromoter17; we incorporated a FRET pair with a Cy3B donor at the -15 position on the non-template strand, and an ATTO647N acceptor at +20 site of the template strand (Fig.
[0261] 5B). During initial transcription, FRET increases as the acceptor moves closer to the donor due to DNA scrunching (Kapanidis 2006; Duchi 2016) and reaches an apparent FRET efficiency EFRET of -0.38 after synthesis of a 6-mer RNA, where pausing occurs. Further synthesis past the pause and escape from the promoter gives rise to a transient EFRET of -0.6 (Duchi 2016).
[0262] To capture the initial transcription kinetics, we incubated the promoter library with RNAP to form RNAP-DNA open complexes, immobilised them on PEGylated surfaces, and imaged them at the single-molecule level (Methods),' we observed a mean EFRET -0.2, as in our previous work (Fig. S17A). After adding 500 pM ApA dinucleotide and 200 pM of all NTPs, most molecules show FRET changes indicating transcription activity, with several patterns of FRET changes over time for different molecules (Fig.
[0263] 5C; Fig. S17B). A control experiment where rifampicin (an antibiotic that blocks RNA extension beyond 3 -nt in length) was added to the open complexes in the presence of NTPs showed no FRET change from the E*~0.2 state, confirming that the observed changes are due to transcriptional activity (Fig. S15C).
[0264] After monitoring transcription for 5 min, we removed the protein and the DNA template strand (by incubating with 30 mM NaOH for 5 min) and annealed two strands to create an 8-nt gap centered around positions +6 and +7. We then performed sequential interrogation via competitive inhibition by using an 8-mer R-seal with two degenerate bases corresponding to positions +6 and +7, and the corresponding sets of U-seals (Fig. S18).
[0265] After selecting traces for further analysis (see Methods), we found that 91% (1142 out of 1255) of the molecules exhibited a pause at E*~0.4. These pausing molecules displayed 4 main classes of behaviour (Fig. 5D). Class I molecules (8%) showed stepwise E* changes up to -0.7 and decrease to a low FRET state, consistent with promoter escape, and completion of transcription (“run-off’) on the entire template (Duchi 2016). Class II molecules (18%) displayed extensive cycling between the E* -0.2 and -0.4 states, suggesting involvement in several abortive initiation cycles, during which a short RNA (4-6-mer) is released from the initial transcribing complex (ITC), allowing RNAP to restart synthesis of a new RNA chain. Class III molecules(25%) showed only a single paused state at E*~0.4, after which E* returns to ~0.2 and stayed there for the remainder of the trace; these molecules are likely inactive ITCs with backtracked RNA (Dulin 2018). Finally, Class IV molecules (39%) reached the E* -0.4 state and stayed there for the remainder of the trace (-5 min); these molecules are likely to be stably scrunched complexes, previous seen as major intermediates during / acCONS initial transcription (Duchi 2016; Dulin 2018).
[0266] To determine the pause lifetimes of all 16 promoter variants, we measured the duration of each first pause at E*~0.4 after NTP addition for Classes I, II and III (Fig. 5D), plotted the first-pause dwell-time distributions for each sequence (as read by Gap-seq) and determined the mean pause lifetime for each sequence (Fig. 5E). Consistent with our earlier work (Duchi 2016; Dulin 2018), the T+6G7sequence, found on both the native lac promoter and the synthetic / acCONS promoter, is associated with a long pause (-17 s, n=73). Further, the C+6G+7sequence exhibited an even longer pause (-23 s, n=87), while in contrast, the A+6T+7sequence exhibited the shortest pause (-4.3 s, n=56); hence, just a 2-nt sequence can change the pause lifetime by >7-fold. In general, we observed that promoters with pyrimidines (Y; C or T) at +6, and a G at +7 exhibited long pausing, supporting an initiation pause element of (C / T)+6G+7or Y+6G+7(Fig. 5F). This sequence element matches the sequence consensus for pausing in transcription elongation (Vvendeskaya 2015; Larson 2015), pausing in initial transcription (Winkelman 2020), and o-dependent pausing in early elongation (Pukhrambam PNAS 2022) as determined using NGS approaches; in all cases, the pause-inducing sequence preferences were due to the tendency of specific nucleotides to disfavour the translocation of RNAP from its pre-translocated to its post-translocated state, with the Y+6preference reflecting RNAP active-site binding preferences for specific nucleotides, and the G+7preference reflecting effects of duplex stability and active-site binding preferences (Vendeskaya 2015; Hein 2011 (Biochemistry 50, 7002 (2011)).
[0267] To test our findings, we prepared and analysed two variants of the library: one associated with long pausing and stable scrunching (C+6G+7), and one associated with short pausing (A+6T+7). After analysing the variants separately, we obtained results that recapitulated the findings from the variant library, validating our approach (Fig. SI 9). We also examined how the sequence at positions +6 and +7 impacts the overall transcription profile by looking at the sequence-dependence of the tendency to belong to a given class. We found that the T+6G+7, C+6G+7and C+6C+7sequences (linked to longpauses in Classes I-III) also formed the most abundant stably scrunched complexes (Class IV; Fig. 5G). This result strongly suggests that the Y+6G+7sequences favour formation of a common reaction intermediate that precedes entry into either a long (but transient) paused state, or an exceptionally long paused state; this intermediate is very likely to be the pre-translocated state of ITC6 (Fig. 5A), which is the branchpoint entry to form Class II (abortive complexes), Class III (stable RNA-backtracked complexes), and Class IV molecules (stably scrunched complexes). Further, the strong DNA:RNA hybrid at the 3’ end of the RNA prevents backtracking, and thus keeps the 3’ end close to the active site (entering thus the long-lived inactive state (half-translocated state). In contrast, cycling molecules (Class II events) show a preference for A / T-rich sequences vs G / C sequences (e.g., 37% of the A+6A+7traces cycle, where only 7% of the G+6G+7traces cycle; Fig. 5H). Cycling molecules leave the paused state via RNA backtracking, followed by RNA entry in the secondary channel, RNA release and restart of RNA synthesis. The observed preference for A / T base pairs implies low duplex stability for the RNA-DNA hybrid at the 3’ end of the transcript and matches a well-known determinant for RNA backtracking due to easier opening of the RNA-DNA duplex at the 3 ’-end of the RNA (Nudler et al., 1997, Tadigotla et al., 2006). It is also likely that a significant fraction of these complexes synthesize a 7-mer (since it is easier to melt the A-T base pair at +7), and therefore have a 2 A-T bp at the 3’ end, while still encountering the o70finger block (an RNAP structural element found in the path of the growing RNA), which we have shown is not fully displaced on the lacCONS promoter until a 10-nt RNA is synthesized (Wang biorxiv 2023).
[0268] Finally, sequences G+6G+7and A+6A+7are over-represented in single-pause events (Fig.
[0269] 51), where the RNA backtracks, enters the NTP channel and remains arrested there for a long time. A possible explanation of the preference for purines is some sequence specificity that favours stable binding of the 3 ’-end of the backtracked RNA to structural elements in the secondary channel of the RNAP. Such binding can lead to transcription arrest; consistent with this hypothesis, elongation complexes with RNA backtracked by at least 2 nt block trigger loop folding and favour swivelled states, which suppress productive elongation (Weixlbauer paper Mol Cell 2019).
[0270] Discussion
[0271] Our work establishes the ability to sequence DNA via detection of transient hybridization to single surface-immobilised DNA molecules featuring short gaps. Weshow how to link DNA-sequence information to protein-DNA interaction kinetics at the single-molecule level by looking at the sequence-dependence of a complex nonequilibrium reaction (initial transcription) on DNA. Our work currently paves the way for an assay that links single-molecule phenotypes to short randomised sequences (up to ~6 nt) represented in moderate-size libraries (up to 4096 sequences). Such complexity will offer substantial insight on how DNA sequence impacts the structure, interactions, and reaction kinetics of many systems involving nucleic acids. Our method uses standard surfaces for single-molecule fluorescence assays with labelled proteins or other biomolecules, thus opening a wide repertoire of potential target systems.
[0272] Currently, the optimum length of the DNA gap to be sealed by transiently bound DNA is 7-13 nt, depending on the length to be sequenced. If needed, seals can be engineered using modified nucleotides which will increase affinity for the complementary target (e.g., locked nucleic acids, modified bases or spermine monomers) and improve mismatch discrimination. As an alternative to degenerate bases, the assay may involve “universal bases” (such as nitroindole (35)), arranged to oppose some or all of the remaining nucleotide positions of the gap section. Assay extensions will involve faster data acquisition and analysis to cover larger libraries. Integration of automated microfluidics will minimise the tedium associated with manual washing and sampleloading steps and allow many fluid-delivery events to surface-immobilised samples.
[0273] There is also scope to reduce the acquisition time for a single base interrogation, which will increase the rate of sequencing and the overall assay. Faster acquisition will also be helped by working with higher densities of gapped molecules, which in turn can be achieved by resolving closely spaced molecules via high-precision localisation (15), which has already been used in our analysis (Fig. 2E-F). To increase throughput and sequence coverage, we envision using large single-FOV instruments, encoding methods that can read one base in a single interrogation, and use of faster acquisition to accelerate the discrimination between a complementary and mismatch seals.
[0274] The method can also operate by sequencing only functionally important molecules of a large libraries (e.g., scan the surface rapidly to identify the top 3-5 molecules showing activity, and sequence only these molecules, and use any of their features to further engineer the library). For large libraries, information on sequence dependence can beobtained by pooling some of the sequences - i.e., one does not need to have 100 molecules per each sequence to be able to identify sequence dependence.
[0275] The SPIN-seq approach will allow detailed study of the sequence-dependence of processes involving proteins that process DNA in multiple steps (RNA polymerases, gene-editing enzymes), as well as proteins that bind and deform DNA (transcription factors, proteins detecting DNA damage, nucleoid-associated proteins). Extension to RNA-binding proteins should also be straight-forward; since libraries can comprise RNA or modified DNA constructs that fold into aptamers (38), our approach offers a way to determine the sequence-specificity for protein-aptamer binding, and generate improved aptamers for biomolecular capture, labelling, and detection, as well as for drug discovery. Our approach, along with other single-molecule multiplexing methods performed on MiSeq chips (REFS), offer an exciting new avenue of experimentation that will transform our understanding of protein-nucleic interactions.
[0276] SUPPLEMENTAL INFORMATION METHODS
[0277] Preparation of DNA fragments. Oligodeoxyribonucleotides (“oligos”) were prepared both in house (see below for detailed description) and sourced from commercial sources (IBA Life Sciences; IDT; Biomer; metabion); in the latter case, the oligos were dissolved at a final concentration of 100 pM and stored at -20°C.
[0278] Unmodified and modified oligonucleotides prepared in house were synthesized on an Applied Biosystems 394 automated DNA / RNA synthesizer using a standard 1.0 pmole phosphoramidite cycle of acid-catalyzed detritylation, coupling, capping, and iodine oxidation. Stepwise coupling efficiencies and overall yields were determined by the automated trityl cation conductivity monitoring facility and was >98.0%. Pre-packed SynBase™ CPG 1000 / 110 (Link Technologies) 3'-amino C7, 3'-biotin, 3'-dabcyl and 3'-Cy3 resins were used for the synthesis. All -cyanoethyl phosphoramidite monomers were dissolved in anhydrous acetonitrile to a concentration of 0.1 M immediately prior to use with coupling time of 50 s for normal A, G, C, and T monomers and was extended to 600 s for 5 ’-amino, 5’-Cy3, 5 ’-biotin and 5 -amino C6 dT phosphoramidite monomers (from Link Technologies Ltd.).5’-Amino, 3’-amino and internal 5-amino C6 dT modified oligonucleotides on resin were treated with a solution of diethylamine (10% in acetonitrile) for 20 min in order to selectively remove the cyanoethyl protecting groups. Cleavage and deprotection of oligonucleotides were achieved by exposure to concentrated aqueous ammonia solution for 60 min at room temperature followed by heating in a sealed tube for 5 hrs at 55 °C. Oligonucleotides labelled with either: 1) two ATTO647N fluorophores; 2) an ATTO647N fluorophore and a DABCYL quencher; or 3) Cy3B were prepared by NHS-ester coupling. The unlabelled pre-cursor oligonucleotide (40 nmol) was freeze-dried and resuspended in sodium carbonate buffer (18 pl, 0.5 M, pH 8.5). Separately, Cy3B or ATTO647N NHS-ester (6 pl, 20 nmol / pl in DMSO) was mixed with DMSO (6 pL). The NHS ester solution was then added to the oligonucleotide solution and mixed vigorously before shaking (800 rpm, room temperature). After 3 hrs, reactions were diluted with water (970 pl) and purified by desalting on a disposable gel-filtration column (NAP10, GE Healthcare Life Sciences, cat. no. GE17-0854-02) according to the manufacturer’s instruction to remove the majority of the free dye. Labelled oligonucleotides were further purified by reverse-phase HPLC chromatography using a Gilson HPLC system with ACE® C8 column (10 mm x 250 mm, pore size 100 A, particle size 10 pm) with a gradient of buffer A (0.1 M TEAB, pH 7.5, where TEAB = triethylammonium bicarbonate) to buffer B (0.1 M TEAB, pH 7.5 containing 50% v / v acetonitrile) and flow rate of 4 ml / min over a total run time of 17.5 min. For all ATTO647N labelled oligonucleotides, a gradient of 70-100% buffer B was used. For Cy3B-labelled oligonucleotides, a gradient of 20-35% buffer B was used. For other oligonucleotides, a gradient of 10-25% buffer B was used. The TEAB buffer and acetonitrile in the collected fractions were removed in vacuo or by freeze drying.
[0279] All oligonucleotides were characterised by negative-mode electrospray using a UPLC-MS Waters XEVO G2-QTOF mass spectrometer and an Acquity UPLC system with a BEH C18 1.7 pm column (Waters). A gradient of methanol in triethylamine (TEA) and hexafluoroisopropanol (HFIP) was used (buffer A, 8.6 mM TEA, 200 mM HFIP in 5% methanol / water (v / v); buffer B, 20% v / v buffer A in methanol). Buffer B was increased from 0-70% over 7.5 min or 15-30% over 12.5 min for normal oligonucleotides and 50-100% over 7.5 min for hydrophobic oligonucleotides. The flow rate was set to 0.2 mL / min. Raw data were processed and deconvoluted using the deconvolution software MassLynx v4.1 and in all cases confirmed the integrity of the sequences.Oligonucleotides with quencher pair Atto647N and BHQ1 was procured from Biomers. All the oligos used in competition experiments was bought from metabion and Biomer corporation.
[0280] With regards to our DNA nomenclature, in a strand named GSeq(l-27)-TCy3B,P18, GSeq refers to Gap sequencing experiments, (1-27) refers to the length of the strand and the position within the construct with respect to the 5' end, T denotes the top strand, and the superscript (Cy3B, Pl 8) notes the DNA modification and its position.
[0281] To form the Gap-containing DNA substrates, as well as the fully duplexed DNA fragments, oligos were annealed by mixing two complementary strands in a ratio of 1: 1 in hybridization buffer (200 mM Tris-HCl pH 8.0, 500 mM NaCl, 1 mM EDTA) and by heating for 5 min at 95°C, followed by cooling to 25°C at a rate of 2°C per minute and placed on ice before being used.
[0282] Sample preparation for real-time transient DNA binding and sequencing. The biotinylated DNA constructs were immobilised at a density of ~0.2 molecules per pm2via neutravidin inside the wells of silicon gaskets, on plasma-cleaned coverslips coated with polyethylene-glycol (PEG; (4, 17)). We then added 20 pl of 50 pM Gapped-DNA constructs in Milli Q water, incubated for 10 s, and washed three times with 200 pl PBS. Subsequently, we added 20 pl of DNA imaging buffer (0.5 M NaCl, 50 mM HEPES pH 7.3, 6 mM BSA, 1 mM TROLOX, 1% Glucose, 40 pg / ml catalase and 0.1 mg / ml glucose oxidase) containing 10 nM (unquenched) or 100 nM (quenched) DNA seals in the well. Sequencing by competitive inhibition. For the single-base sequencing using competitive inhibition, we used as an R-seal an 8-nucleotide ssDNA having degenerate base at position 3 of the gap, and labelled with Atto647N at the 5’ end. Binding of this R-seal was competed with four different unlabelled seals of 8 nt length having A,T,G,C variants at position 3 of the gap. To standardize the competition assay, we varied the concentration of degenerate seal from 100 nM, 200 nM and 250 nM. Competing unlabelled seals concentrations also varied from IX, 2X, 3X and 5X of that of the degenerate labelled seal. In every round of imaging a mixture of labelled and one of the unlabelled seals was employed.
[0283] In the mixed single-base sequencing experiment, all the four different gaps were mixed at an equimolar concentration of 50 pM. Competing labelled and unlabelled seals werekept at a concentration of 250 nM and 500 nM, respectively. Imaging buffer consisted of 20 mM HEPES, 200 mM NaCl, 80 mM MgC12, 10 % formamide, pH 7.7.
[0284] In the mixed single-base sequencing with prior information, each DNA gap was immobilized on the surface at a concentration of 50 pM sequentially, followed by photobleaching after each deposition.
[0285] For 3-base sequencing, a 13-nt DNA gap was formed using 1-65 biotinylated sequence and two flanking strands of 1-27 Cy3 strand and 41-65 base sequence. Binding of 13-nt labelled seal containing 3 degenerate bases was standardized by varying its concentration from 250 nM to 1 pM respectively. Optimum concentration of dextran sulphate and formamide in the binding buffer was selected by titrating 5%, 10%, 15% and 20% of dextran sulphate, as well as by fixing the dextran sulphate concentration and varying the formamide concentration from 5% to 20%. The concentration of unlabelled seal was optimised by varying it from lx to 3x. Finally, the 3-base sequencing experiment was performed using 500 nM labelled degenerate seal (R-seal) and 500 nM of unlabelled competing seals, in the buffer composed of 20 mM HEPES (pH 7.7), 200 mM NaCl, 80 mM MgC12, 10% Dextran Sulphate, and 15 % formamide. For 5-base sequencing, a DNA gap was prepared as for 3-base sequencing. A 13-nt labelled seal with 5 degenerate bases at positions 5, 6, 7, 8, 9 labelled with ATTO647N and BHQ1 and 5’ and 3’ end respectively was used. Competition condition in terms of concentration of labelled, unlabelled seals and buffer composition were standardized as in the case of 3-base sequencing. Finally, 5-base sequencing was performed using 2 pM labelled degenerate seal and an equimolar concentration of competing seals, in an imaging buffer composed of 20 mM HEPES (pH 7.7), 200 mM NaCl, 80 mM MgC12, 15% Dextran Sulphate, and 20 % formamide.
[0286] Sequence identification analysis pipeline. Drift correction is performed by first identifying localizations. The AIM algorithm (doi: 10.1126 / sciadv.adm7765) is then applied to calculate the drift, and the movies are shifted in the opposite direction to compensate for it.
[0287] Both the acceptor and donor channels in the movies are aligned to the donor channel of localization movie using the pystackreg (doi: 10.1109 / 83.650848) library with a rigid-body transformation. The transformation matrix between donor channels, as well as thetransformation between the donor and acceptor channels, is computed using fiducial markers from the initial frames. Background subtraction is carried out using a median filter with a window size of 21 pixels. The resulting images are scaled and translated to bring them into the same intensity range before performing the registration.
[0288] Localizations in the donor channel are identified using Picasso (10.1038 / nprot.2017.024) with a gradient threshold of 400. Localizations in the acceptor channel are found using a gradient threshold of 1000, with the following camera parameters: baseline = 400, box size = 5, EM gain = 1, quantum efficiency = 0.82, sensitivity = 2.5, and pixel size = 117 nm. For each localization in the donor channel, the number of localizations in all acceptor channels within a radius of 2 pixels is counted. Stronger binding results in a higher number of localizations, allowing for the selection of the complementary nucleotide in the non-competition sequencing approach. In the competitive-inhibition approach, we look for the trace (out of 4) that displays the lowest number of localisations.
[0289] Preparation of RNA polymerase holoenzyme. Escherichia coli RNAP holoenzyme was expressed and purified by co-expressing genes of RNAP p, ’, a, co and o70. A primary culture of 20 mL LB broth was inoculated using a single colony of BL21 DE3 E. coli strain transformed with plasmid pVIO (Belogurov et al., 2007) and plasmid pRSFduet-sigma (Hudson et al., 2009). Ampicillin (100 pg / ml) and kanamycin (50 pg / ml) were used as selection antibiotics. A second culture (2* IL) was started using 10 mL of primary culture using the same antibiotic level and was grown at 37°C until O.D. reached 0.6. IPTG was added at 1 mM and the cultures were further incubated at 37°C for 3 hrs. Cells were harvested and lysed followed by precipitation of cell lysate using polyethyleneimine and ammonium sulphate. Purification of the protein was performed using Ni-NTA agarose affinity column with elution using 200 mM imidazole in buffer A. Affinity purified sample was passed through anion-exchange chromatography equipped with Mono-Q 10 / 100 GL. Fractions containing the protein of interest were pooled together and concentrated using a 30 kDa MWCO Amicon Ultra-15 centrifugal ultrafilters.
[0290] Library preparation for initiation pause Gap-seq assay. LacCONS promoter with all possible sequences at +6 and +7 position has been generated using primer extension followed by ligation. Biotinylated non-templated strand up to -10 position having Cy3Bat -15 position and another strand from -30 to +10 position having degenerate bases at +6 and +7 position was mixed at 1:2 ratio. Primer extension has been performed using Q5 DNA polymerase using a thermocycler. Melting temperature, annealing temperature, extension temperature were set at 95 °C, 72 °C and 62 °C respectively. Another two strands having sequences from position +10 to +25 were ligated using T4 DNA ligase having ATTO647N at position +20. To check for successful ligation, the products were run on 20% Bis-acrylamide-acrylamide native gel at a voltage of 70 mV. Gel imaging was performed using Typhoon gel doc. Gel extraction and ethanol precipitation was performed to purify the double stranded, fully ligated DNA library.
[0291] Transcription assay. Open complex was formed by incubating 25 nM of DNA library with 50 nM of RNAP holoenzyme in KG7 buffer (40 mM HEPES, 100 mM Potassium glutamate, 10 mM MgC12, 1 mM DTT, 100 pg / mL BSA, 5% glycerol) at 32°C for half an hour. Subsequently, 1 pL of 1 mg / mL heparin was incubated in the solution for one minute, followed by centrifuging for 2 minutes at 2000 RPM. The supernatant was collected in a separate centrifuge tube, and 0.5 pL of the open complex mixture was incubated on a neutravidin-coated slide having 30 pL of KG7 as prepared for singlemolecule DNA imaging. After washing unbound open complex from the surface, we added 500 pM ApA dinucleotide for 10 minutes. We then started image collection using ALEX excitation, and 200 pM NTP was added in real-time using a fluidic system.
[0292] In-situ DNA gap preparation and sequencing. After the transcription assay, the surface was washed using PBS buffer. To de-hybridization the promoter fragments, 20 mM NaOH was incubated on the surface for 5 minutes, followed by three washed using annealing buffer. To create a 8-nt gap, flanking sequences were incubated on the surface for 15 minutes at a final concentration of 500 nM. The sequencing assay has been carried out using an 8-nt degenerate seal labelled with Atto647N at the 5’ end and having 2 degenerate bases at positions 4 and 5 from the 5’ end, respectively. Sequential interrogations were performed using competing unlabelled U-seals at a concentration of 500 nM.
[0293] References
[0294] All references herein may be incorporated by reference.1. A. M. Malinen, J. Bakermans, E. Aalto-Setala, M. Blessing, D. L. V. Bauer, O. Parilova, G. A. Belogurov, D. Dulin, A. N. Kapanidis, Real-Time Single-Molecule Studies of RNA Polymerase-Promoter Open Complex Formation Reveal Substantial Heterogeneity Along the Promoter-Opening Pathway. J Mol Biol 434 (2022).
[0295] 2. A. Mazumder, R. H. Ebright, A. N. Kapanidis, Transcription initiation at a consensus bacterial promoter proceeds via a ‘bindunwindload-and-lock’ mechanism. Elife 10 (2021).
[0296] 3. T. Ha, Single-Molecule Techniques: A Laboratory Manual (Cold Spring Harbor Press, 2008).
[0297] 4. K. M. Herbert, A. La Porta, B. J. Wong, R. A. Mooney, K. C. Neuman, R. Landick, S. M. Block, Sequence-Resolved Detection of Pausing by Single RNA Polymerase Molecules. Cell 125, 1083-1094 (2006).
[0298] 5. J. Eid, A. Fehr, J. Gray, K. Luong, J. Lyle, G. Otto, P. Peluso, D. Rank, P. Baybayan, B. Bettman, A. Bibillo, K. Bjornson, B. Chaudhuri, F. Christians, R. Cicero, 5. Clark, R. Dalal, A. DeWinter, J. Dixon, M. Foquet, A. Gaertner, P. Hardenbol, C. Heiner, K. Hester, D. Holden, G. Kearns, X. Kong, R. Kuse, Y. Lacroix, S. Lin, P. Lundquist, C. Ma, P. Marks, M. Maxham, D. Murphy, I. Park, T. Pham, M. Phillips, J. Roy, R. Sebra, G. Shen, J. Sorenson, A. Tomaney, K. Travers, M. Trulson, J. Vieceli, J. Wegener, D. Wu, A. Yang, D. Zaccarin, P. Zhao, F. Zhong, J. Korlach, S. Turner, Real-time DNA sequencing from single polymerase molecules. Science 323, 5910, 133-138 (2009).
[0299] 6. J. Clarke, H. C. Wu, L. Jayasinghe, A. Patel, S. Reid, H. Bayley, Continuous base identification for single-molecule nanopore DNA sequencing. Nat Nanotechnol 4 (2009).
[0300] 7. A. H. Laszlo, I. M. Derrington, J. H. Gundlach, MspA nanopore as a singlemolecule tool: From sequencing to SPRNT. Methods 105, 75-89 (2016).
[0301] 8. J. Aguirre Rivera, G. Mao, A. Sabantsev, M. Panfilov, Q. Hou, M. Lindell, C. Chanez, F. Ritort, M. Jinek, S. Deindl, “Massively parallel analysis of single-molecule dynamics on next-generation sequencing chips. Science 385, 6711, 892-898 (2024). 9. I. Severins, C. Bastiaanssen, S. H. Kim, R. B. Simons, J. Van Noort, C. Joo, “Single-molecule structural and kinetic studies across sequence space. Science 385, 6711, 898-904 (2024).
[0302] 10. M. Kummerlin, Q. Zhao, J. Hazra, C. Hepp, A. Farrar, P. Turner, A. N. Kapanidis, Tunable Anorogenic DNA probes drive fast and high-resolution single-molecule fluorescence imaging. [Preprint] (2025). https: / / doi.org / 10.1101 / 2025.01.21.634148.
[0303] 11. D. L. V. Bauer, D. Duchi, A. N. Kapanidis, E. Coli RNA Polymerase Pauses during Initial Transcription. Biophysical Journal 110 (2016).
[0304] 12. D. Dulin, D. L. V. Bauer, A. M. Malinen, J. J. W. Bakermans, M. Kaller, Z. Morichaud, I. Petushkov, M. Depken, K. Brodolin, A. Kulbachinskiy, A. N. Kapanidis, Pausing controls branching between productive and non-productive pathways during initial transcription in bacteria. Nat Commun 9 (2018).
[0305] 13. E. Lernera, S. Y. Chung, B. L. Allen, S. Wang, J. Lee, S. W. Lu, L. W. Grimaud, A. Ingargiola, X. Michalet, Y. Alhadid, S. Borukhov, T. R. Strick, D. J. Taatjes, S. Weiss, Backtracked and paused transcription initiation intermediate of Escherichia coli RNA polymerase. Proceedings of the National Academy of Sciences of the United States of America 113 (2016).
[0306] 14. J. T. Winkelman, C. Pukhrambam, I. O. Vvedenskaya, Y. Zhang, D. M. Taylor, P. Shah, R. H. Ebright, B. E. Nickels, XACT-Seq Comprehensively Defines the Promoter-Position and Promoter-Sequence Determinants for Initial-Transcription Pausing. Molecular Cell 79 (2020).
[0307] 15. I. O. Vvedenskaya, J. S. Sharp, S. R. Goldman, P. N. Kanabar, J. Livny, S. L. Dove, B. E. Nickels, Growth phase-dependent control of transcription start site selection and gene expression by nanoRNAs. Genes & development 26, 1498-507 (2012).
[0308] 16. M. H. Larson, R. a Mooney, J. M. Peters, T. Windgassen, D. Nayak, C. a Gross, S. M. Block, W. J. Greenleaf, R. Landick, J. S. Weissman, A pause sequence enriched at translation start sites drives transcription dynamics in vivo. Science (New York, N.Y.) 344, 1042-7 (2014).
[0309] 17. D. Duchi, D. L. V. Bauer, L. Fernandez, G. Evans, N. Robb, L. C. Hwang, K. Gryte, A. Tomescu, P. Zawadzki, Z. Morichaud, K. Brodolin, A. N. Kapanidis, RNA Polymerase Pausing during Initial Transcription. Mol Cell 63, 939-950 (2016).
[0310] 18. A. N. Kapanidis, E. Margeat, S. O. Ho, E. Kortkhonjia, S. Weiss, f Richard, H. Ebright, Initial Transcription by RNA Polymerase Proceeds Through a DNA-Scrunching Mechanism. Science 314, 5802, 1144-1147
[0311] 19. D. Dulin, D. L. V. Bauer, A. M. Malinen, J. J. W. Bakermans, M. Kaller, Z. Morichaud, I. Petushkov, M. Depken, K. Brodolin, A. Kulbachinskiy, A. N. Kapanidis, Pausing controls branching between productive and non-productive pathways during initial transcription in bacteria. Nat Commun 9 (2018).20. I. O. Vvedenskaya, Y. Zhang, S. R. Goldman, A. Valenti, V. Visone, D. M. Taylor, R. H. Ebright, B. E. Nickels, Massively Systematic Transcript End Readout, “MASTER”: Transcription Start Site Selection, Transcriptional Slippage, and Transcript Yields. Mol Cell 60, 953-965 (2015).
[0312] 21. M. H. Larson, R. A. Mooney, J. M. Peters, D. Nayak, C. A. Gross, S. M. Block, W. J. Greenleaf, R. Landick, J. S. Weissman, A pause sequence enriched at translation start sites drives transcription dynamics in vivo. Science (1979) 344, 1042-1047 (2014).
[0313] 22. J. T. Winkelman, C. Pukhrambam, I. O. Vvedenskaya, Y. Zhang, D. M. Taylor, P. Shah, R. H. Ebright, B. E. Nickels, XACT-Seq Comprehensively Defines the Promoter-Position and Promoter-Sequence Determinants for Initial-Transcription Pausing. Mol Cell 79, 797-811 (2020).
[0314] 23. C. Pukhrambam, V. Molodtsov, M. Kooshkbaghi, A. Tareen, H. Vu, K. S. Skalenko, M. Su, Z. Yin, J. T. Winkelman, J. B. Kinney, R. H. Ebright, B. E. Nickels, Structural and mechanistic basis of o-dependent transcriptional pausing. PNAS 119 (23) e2201301119 (2022)
[0315] 24. P. P. Hein, M. Palangat, R. Landick, RNA transcript 3-2-proximal sequence affects translocation bias of RNA polymerase. Biochemistry 50, 7002-7014 (2011). 25. E. Nudler, A. Mustaev, E. Lukhtanov, A. Goldfarb, The RNA-DNA Hybrid Maintains the Register of Transcription by Preventing Backtracking of RNA Polymerase. Cell 4, 89, 33-41 (1997)
[0316] 26. V. R. Tadigotla, D. Ibhid'oibhid' Ibhid'o Maoile Idigh* f J, A. M. Sengupta, V. Epshtein, R. H. Ebright, E. Nudler, A. E. Ruckenstein, Thermodynamic and kinetic modeling of transcriptional pausing. PNAS 103, 12, 4439-4444 (2006)
[0317] 27. A. Wanga, A. Mazumdera, A. N. Kapanidis, The displacement of the o 70 finger in initial transcription is highly heterogeneous and promoter-dependent. [Preprint] (2023). https: / / doi.org / 10.1101 / 2023.06.10.544452.
[0318] 28. M. Abdelkareem, C. Saint-Andre, M. Takacs, G. Papai, C. Crucifix, X. Guo, J. Ortiz, A. Weixlbaumer, Structural Basis of Transcription: RNA Polymerase Backtracking and Its Reactivation. Mol Cell 75, 298-309. e4 (2019).
[0319] 29. D. Loakes, D. M. Brown, 5-nitroindole as an universal base analogue. Nucleic Acids Research 22 (1994).
[0320] 30. R. Jungmann, C. Steinhauer, M. Scheible, A. Kuzyk, P. Tinnefeld, F. C. Simmel, Single-molecule kinetics and super-resolution microscopy by fluorescence imaging of transient binding on DNA origami. Nano Letters 10, 4756-4761 (2010).31. M. R. Dunn, R. M. Jimenez, J. C. Chaput, Analysis of aptamer discovery and technology. Nature Reviews Chemistry 1 (2017).
[0321] Example 2 - Further examples
[0322] Linking protein-DNA interactions to DNA sequence for single molecules.
[0323] To link singlemolecule phenotypes to sequence, we examined DNA interactions of transcription factor catabolite activator protein (CAP18; Fig. S20A). In the presence of cyclic adenosine 30 monophosphate (cAMP), CAP binds with high affinity to consensus sequence 5’-AAATGTGATCTAGATCACATTT-3 ’ (CAPcons19), with the underlined sequences controlling specificity (Fig. S20B). To observe CAP-DNA binding, we surface-immobilized CAPcons DNA, added labelled CAP in the presence of cAMP, and observed transient CAP binding to DNA (Fig. S20C and Fig. S20D, left). CAP dwell times on DNA (Fig. S20E, middle) showed a lifetime t2 -8.7 s, which reflects stable CAP binding to CAPcons DNA; this assignment was supported by control experiments in the absence of DNA or in the absence of cAMP. Binding -frequency analysis (using unbound-time distributions; Fig. S20D, right) yielded an association rate constant kon ~1.3 x 107Ml s1and a Kd -9.1 nM. A variant featuring a C:G pair at position 5 (as opposed to the consensus G:C) showed weakened CAP-DNA interactions (t2 -6.4 s; kon-7.7 x 106M’1s ’ Fig. S20E), and 2-fold reduced affinity (Kd -20 nM). We then linked CAP-DNA binding for a single DNA molecule to the sequence of the same DNA molecule (Fig. S20F) for a library of surface-immobilized CAPcons DNA molecules, where position 5 of CAPcons was randomized. In step 1, we evaluated CAP-DNA binding as above. In step 2, the dsDNA was denatured, leaving only the biotinylated bottom DNA strand on the surface; subsequently, two DNA strands were hybridized efficiently to the surface-immobilized strand in situ (Fig. S22) to form a surface-immobilized gapped DNA with a 9-nt gap matching the first 9 bases of CAPcons (5’-AAATGTGAT-3 ’). 5 In step 3, we analyzed the same field-of-view and sequenced position 5 using competitive inhibition. Using the surface locations of each protein-DNA binding and sequencing event, we linked CAP-DNA binding events (from step 1; Fig. S20G) to sequencing traces for single molecules (Fig. 6H). Binding analysis showed that G:C (known to form 2 hydrogen bonds from guanosine N7 and 06 atoms to CAP Arg 180 side chain18,2°) leads to a slower off-rate 10 and a faster on-rate relative to the other variants (Fig. S20I-K), leading to a Kd 1.9- to 2.6-fold lower than the non-consensus variants, corresponding to a -0.6-0.9 kT (-0.35-0.55 Kcal / mole) stabilization of the complex versus non-consensus variants. Our results reproduced those from pure sequences (Fig. 6D-E) and established our ability to work with labelled proteins and to link protein-DNA interactions with DNA sequence.A / T-rich sequences
[0324] We examined how well our sequencing approach works on A / T-rich sequences, which may be challenging due to weaker base-pairing potential. We constructed an 8-nt-gapped DNA where all gap bases were adenines (AAAAAAAA, or “8A”), and interrogated it using the same sequential approach we used for mixed sequences. We observed much higher binding for the fully complementary seal (i.e having T at seal position 3), and little binding for the three 1-nt mismatched seals (fig. S21), identifying the A base at position 3 with -97% accuracy, and demonstrating our ability to read A / T-rich sequences.
Claims
CLAIMS1. A method of sequencing a nucleic acid molecule, the method comprising:(a) providing the nucleic acid immobilised on a surface;(b) forming a single -stranded gap section by partially duplexing the nucleic acid such that the single-stranded gap section is flanked by duplex sections, wherein the sequence to be sequenced is a sequence of the single-stranded gap section;(c) providing a fluorescently labelled oligonucleotide reporter probe arranged to fit into the gap section of the immobilised nucleic acid, and comprising universal and / or degenerate bases arranged to oppose the nucleotide positions of the gap section, (d) providing a set of at least four competitive probes, wherein each competitive probe is arranged to fit in the gap section of the immobilised nucleic acid, and wherein each competitive probe comprises:(i) a nucleotide X at a position arranged to oppose a nucleotide of the gap section of the immobilised nucleic acid, wherein such nucleotide is an interrogated nucleotide, and(ii) fully complementary nucleotides for the remaining nucleotide positions of the competitive probe, which are arranged to oppose a complementary nucleotide of the gap section, or comprising complementary nucleotides combined with universal and / or degenerate nucleotides arranged to oppose some or all of the remaining nucleotide positions of the gap section (i.e. part complementary),wherein the at least four competitive probes comprise:- a first competitive probe wherein nucleotide X is a cytosine;- a second competitive probe wherein nucleotide X is a guanine;- a third competitive probe wherein nucleotide X is an adenine; and- a fourth competitive probe wherein nucleotide X is a thymine or uracil; wherein the competitive probes are arranged to compete with the fluorescently labelled oligonucleotide reporter probe for binding to the gap section;(e) detecting binding, or absence thereof, of the fluorescently labelled oligonucleotide reporter probe with the gap section of the immobilised nucleic acid in the presence of one of the four competitive binding probes;(f) repeating step (e) for each of the remaining competitive binding probes,wherein the identity of the interrogated nucleotide of the gap section of the immobilised nucleic acid is identified as the complementary base of the nucleotide X of the competitive probe that has the highest incidence of disrupting the binding of the fluorescently labelled oligonucleotide reporter probe; and optionally(g) repeating steps (d) to (f) for interrogating subsequent nucleotide positions of the gap section of the immobilised nucleic acid until sufficient nucleotides of the gap section have been identified to be able to determine a sequence.
2. The method according to claim 1, wherein the fluorescently labelled oligonucleotide reporter probe comprises universal or degenerate bases arranged to oppose all of the nucleotide positions of the gap section.
3. The method according to claim 1 or claim 2, wherein the fluorescently labelled oligonucleotide reporter probe is equal in length to the gap section of the immobilised nucleic acid.
4. The method according to any preceding claim, wherein the fluorescently labelled oligonucleotide reporter probe and / or the gap section is 2-20 nucleotides in length.
5. The method according to any preceding claim, wherein the competitive probes are 2-20 nucleotides in length.
6. The method according to any preceding claim, wherein the competitive probes are equal in length to the gap section of the immobilised nucleic acid and the fluorescently labelled oligonucleotide reporter probe.
7. The method according to any preceding claim, wherein the binding of the fluorescently labeled oligonucleotide reporter probe to the gap-section of the nucleic acid is detected by measurement of Forster resonance energy transfer (FRET) between the fluorophore on the fluorescently labeled oligonucleotide reporter probe and a second fluorophore, wherein the second fluorophore is tagged on the immobilized nucleic acid molecule or duplex thereof, or wherein the fluorescently labelled oligonucleotide reporter probe comprises a second fluorophore or quencher.
8. The method according to any preceding claim, wherein the immobilised nucleic acids are spatially arranged onto the surface.
9. The method according to any preceding claim, wherein the nucleic acid molecules are immobilised on the surface at a surface density of between about 1 and about 1500 molecules per pm2.
10. The method according to any preceding claim, wherein the immobilised nucleic acid are immobilised by way of being in situ in a cell and the cell is attached to a surface or the cell is immobilised by capture, such as in microfluidic channels.
11. The method according to any preceding claim, further comprising the step of amplifying the immobilised nucleic acid.
12. A method of single-molecule phenotyping and sequencing of a nucleic acid molecule, the method comprising:a) providing the nucleic acid molecule immobilised to a surface;b) testing for a phenotype / function of the immobilised nucleic acid molecule; andc) determining the sequence of the immobilised nucleic acid molecule such that an observed phenotype / function is linked to the sequence, wherein the nucleic acid is sequenced by the method of sequencing according to any preceding claim; optionally wherein testing for a phenotype / function of the immobilised nucleic acid molecule is before or after determining the sequence of the immobilised nucleic acid molecule.
13. The method according to claim 12, wherein the testing for a phenotype / function of the immobilised nucleic acid molecule comprises testing for binding of the nucleic acid to a target molecule.
14. A method of molecule phenotyping and identification of a molecule, wherein the molecule is tagged with a nucleic acid and identified by sequencing of the tagged nucleic acid, the method comprising:a) providing the molecule immobilised to a surface;b) testing for a phenotype / function of the immobilised molecule; andc) determining the sequence of the tagged nucleic acid molecule such that an observed phenotype / function is linked to the tagged nucleic acid sequence, wherein the tagged nucleic acid is sequenced by the method of sequencing according to any of claims 1-11; optionally wherein testing for a phenotype / function of the tagged nucleic acid molecule is before or after determining the sequence of the tagged nucleic acid molecule.15 A method of single-cell phenotyping and sequencing of a nucleic acid molecule associated with the phenotype of the cell, the method comprising:a) providing a population of cells comprising modifications in a nucleic acid sequence of interest;b) observing a phenotype of individual cells, wherein the cells are fixed to a surface before, during or after the phenotypic observation, such that the nucleic acid sequence is also immobilised;c) sequencing the nucleic acid sequence by the method of sequencing according to any of claims 1-11, such that an observed phenotype / function of the cell is linked to the nucleic acid sequence.
16. A composition or a kit comprising a fluorescently labelled oligonucleotide reporter probe and a competitive probe arranged to compete for binding with the fluorescently labelled oligonucleotide reporter probe,wherein the competitive probe comprises one or more of at least four competitive probes comprising:(i) a nucleotide X at one position in the sequence of the competitive probes, wherein the position is the same for each of the at least four fluorescently labelled oligonucleotide probes, and(ii) universal or degenerate bases at the remaining nucleotide positions, and wherein the at least four competitive probes comprise:- a first competitive probe wherein nucleotide X is a cytosine;- a second competitive probe wherein nucleotide X is a guanine;- a third competitive probe wherein nucleotide X is an adenine; and- a fourth competitive probe wherein nucleotide X is a thymine or uracil.
