Patent Application
20100130368
This invention relates to fabricated arrays of molecules, and to their analytical applications. In particular, this invention relates to the use of fabricated arrays in methods for obtaining genetic sequence information.
Advances in the study of molecules have been led, in part, by improvement in technologies used to characterise the molecules or their biological reactions. In particular, the study of nucleic acids, DNA and RNA, has benefited from developing technologies used for sequence analysis and the study of hybridisation events.
An example of the technologies that have improved the study of nucleic acids, is the development of fabricated arrays of immobilised nucleic acids. These arrays typically consist of a high-density matrix of polynucleotides immobilised onto a solid support material. Fodor et al., Trends in Biotechnology (1994) 12:19-26, describes ways of assembling the nucleic acid arrays using a chemically sensitised glass surface protected by a mask, but exposed at defined areas to allow attachment of suitably modified nucleotides. Typically, these arrays may be described as “many molecule” arrays, as distinct regions are formed on the solid support comprising a high density of one specific type of polynucleotide.
An alternative approach is described by Schena et al., Science (1995) 270:467-470, where samples of DNA are positioned at predetermined sites on a glass microscope slide by robotic micropipetting techniques. The DNA is attached to the glass surface along its entire length by non-covalent electrostatic interactions. However, although hybridisation with complementary DNA sequences can occur, this approach may not permit the DNA to be freely available for interacting with other components such as polymerase enzymes, DNA-binding proteins etc.
Recently, the Human Genome Project determined the entire sequence of the human genome—all 3×109 bases. The sequence information represents that of an average human. However, there is still considerable interest in identifying differences in the genetic sequence between different individuals. The most common form of genetic variation is single nucleotide polymorphisms (SNPs). On average one base in 1000 is a SNP, which means that there are 3 million SNPs for any individual. Some of the SNPs are in coding regions and produce proteins with different binding affinities or properties. Some are in regulatory regions and result in a different response to changes in levels of metabolites or messengers. SNPs are also found in non-coding regions, and these are also important as they may correlate with SNPs in coding or regulatory regions. The key problem is to develop a low cost way of determining one or more of the SNPs for an individual.
The nucleic acid arrays may be used to determine SNPs, and they have been used to study hybridisation events (Mirzabekov, Trends in Biotechnology (1994) 12:27-32). Many of these hybridisation events are detected using fluorescent labels attached to nucleotides, the labels being detected using a sensitive fluorescent detector, e.g. a charge-coupled detector (CCD). The major disadvantages of these methods are that it is not possible to sequence long stretches of DNA, and that repeat sequences can lead to ambiguity in the results. These problems are recognised in Automation Technologies for Genome Characterisation, Wiley-Interscience (1997), ed. T. J. Beugelsdijk, Chapter 10: 205-225.
In addition, the use of high-density arrays in a multi-step analysis procedure can lead to problems with phasing. Phasing problems result from a loss in the synchronisation of a reaction step occurring on different molecules of the array. If some of the arrayed molecules fail to undergo a step in the procedure, subsequent results obtained for these molecules will no longer be in step with results obtained for the other arrayed molecules. The proportion of molecules out of phase will increase through successive steps and consequently the results detected will become ambiguous. This problem is recognised in the sequencing procedure described in U.S. Pat. No. 5,302,509.
An alternative sequencing approach is disclosed in EP-A-0381693, which comprises hybridising a fluorescently-labelled strand of DNA to a target DNA sample suspended in a flowing sample stream, and then using an exonuclease to cleave repeatedly the end base from the hybridised DNA. The cleaved bases are detected in sequential passage through a detector, allowing reconstruction of the base sequence of the DNA. Each of the different nucleotides has a distinct fluorescent label attached, which is detected by laser-induced fluorescence. This is a complex method, primarily because it is difficult to ensure that every nucleotide of the DNA strand is labelled and that this has been achieved with high fidelity to the original sequence.
WO-A-96/27025 is a general disclosure of single molecule arrays. Although sequencing procedures are disclosed, there is little description of the applications to which the arrays can be applied. There is also only a general discussion on how to prepare the arrays.
According to the present invention, a device comprises a high density array of molecules capable of interrogation and immobilised on a solid generally planar surface, wherein the array allows the molecules to be individually resolved by optical microscopy, and wherein each molecule is immobilised by covalent bonding to the surface, other than at that part of each molecule that can be interrogated.
According to a second aspect of the invention, a device comprises a high density array of relatively short molecules and relatively long polynucleotides immobilised on the surface of a solid support, wherein the polynucleotides are at a density that permits individual resolution of those parts that extend beyond the relatively short molecules. In this aspect, the shorter molecules can prevent non-specific binding of reagents to the solid support, and therefore reduce background interference.
According to a third aspect of the invention, a device comprises an array of polynucleotide molecules immobilised on a solid surface, wherein each molecule comprises a polynucleotide duplex linked via a covalent bond to form a hairpin loop structure, one end of which comprises a target polynucleotide, and the array has a surface density which allows the target polynucleotides to be individually resolved. In this aspect, the hairpin structures act to tether the target to a primer polynucleotide. This prevents loss of the primer-target during the washing steps of a sequencing procedure. The hairpins may therefore improve the efficiency of the sequencing procedures.
The arrays of the present invention comprise what are effectively single molecules. This has many important benefits for the study of the molecules and their interaction with other biological molecules. In particular, fluorescence events occurring on each molecule can be detected using an optical microscope linked to a sensitive detector, resulting in a distinct signal for each molecule.
When used in a multi-step analysis of a population of single molecules, the phasing problems that are encountered using high density (multi-molecule) arrays of the prior art, can be reduced or removed. Therefore, the arrays also permit a massively parallel approach to monitoring fluorescent or other events on the molecules. Such massively parallel data acquisition makes the arrays extremely useful in a wide range of analysis procedures which involve the screening/characterising of heterogeneous mixtures of molecules. The arrays can be used to characterise a particular synthetic chemical or biological moiety, for example in screening for particular molecules produced in combinatorial synthesis reactions.
The arrays of the present invention are particularly suitable for use with polynucleotides as the molecular species. The preparation of the arrays requires only small amounts of polynucleotide sample and other reagents, and can be carried out by simple means. Polynucleotide arrays according to the invention permit massively parallel sequencing chemistries to be performed. For example, the arrays permit simultaneous chemical reactions on and analysis of many individual polynucleotide molecules. The arrays are therefore very suitable for determining polynucleotide sequences.
An array of the invention may also be used to generate a spatially addressable array of single polynucleotide molecules. This is the simple consequence of sequencing the array. Particular advantages of such a spatially addressable array include the following:
1) Polynucleotide molecules on the array may act as identifier tags and may only need to be 10-20 bases long, and the efficiency required in the sequencing steps may only need to be better than 50%, as there will be no phasing problems.
2) The arrays may be reusable for screening once created and sequenced. All possible sequences can be produced in a very simple way, e.g. compared to a high density multi-molecule DNA chip made using photolithography.
According to the present invention, the single molecules immobilised onto the surface of a solid support should be capable of being resolved by optical means. This means that, within the resolvable area of the particular imaging device used, there must be one or more distinct images each representing one molecule. Typically, the molecules of the array are resolved using a single molecule fluorescence microscope equipped with a sensitive detector, e.g. a charge-coupled detector (CCD). Each molecule of the array may be analysed simultaneously or, by scanning the array, a fast sequential analysis can be performed.
The molecules of the array are typically DNA, RNA or nucleic acid mimics, e.g. PNA or 2′-O-Meth-RNA. However, any other biomolecules, including peptides, polypeptides and other organic molecules, may be used. The molecules are formed on the array to allow interaction with other “cognate” molecules. It is therefore important to immobilise the molecules so that the portion of the molecule not physically attached to solid support is capable of being interrogated by a cognate. In some applications all the molecules in the single array will be the same, and may be used to interrogate molecules that are largely distinct. In other applications, the molecules on the array may all, or substantially all, be different, e.g. less than 50%, preferably less than 30% of the molecules will be the same.
The term “single molecule” is used herein to distinguish from high density multi-molecule arrays in the prior art, which may comprise distinct clusters of many molecules of the same type.
The term “individually resolved” is used herein to indicate that, when visualised, it is possible to distinguish one molecule on the array from its neighbouring molecules. Visualisation may be effected by the use of reporter labels, e.g. fluorophores, the signal of which is individually resolved.
The term “cognate molecule” is used herein to refer to any molecule capable of interacting, or interrogating, the arrayed molecule. The cognate may be a molecule that binds specifically to the arrayed molecule, for example a complementary polynucleotide, in a hybridisation reaction.
The term “interrogate” is used herein to refer to any interaction of the arrayed molecule with any other molecule. The interaction may be covalent or non-covalent.
The terms “arrayed polynucleotides” and “polynucleotide arrays” are used herein to define a plurality of single molecules that are characterised by comprising a polynucleotide. The term is intended to include the attachment of other molecules to a solid surface, the molecules having a polynucleotide attached that can be further interrogated. For example, the arrays may comprise protein molecules immobilised on a solid surface, the protein molecules being conjugated or otherwise bound to a short polynucleotide molecule that may be interrogated, to address the array.
The density of the arrays is not critical. However, the present invention can make use of a high density of single molecules, and these are preferable. For example, arrays with a density of 106-109 molecules per cm2 may be used. Preferably, the density is at least 107/cm2 and typically up to 108/cm2. These high density arrays are in contrast to other arrays which may be described in the art as “high density” but which are not necessarily as high and/or which do not allow single molecule resolution.
Using the methods and apparatus of the present invention, it may be possible to image at least 107 or 108 molecules simultaneously. Fast sequential imaging may be achieved using a scanning apparatus; shifting and transfer between images may allow higher numbers of molecules to be imaged.
The extent of separation between the individual molecules on the array will be determined, in part, by the particular technique used to resolve the individual molecule. Apparatus used to image molecular arrays are known to those skilled in the art. For example, a confocal scanning microscope may be used to scan the surface of the array with a laser to image directly a fluorophore incorporated on the individual molecule by fluorescence. This may be achieved using the apparatus illustrated in
Resolving single molecules on the array with a 2-D detector can be done if, at 100× magnification, adjacent molecules are separated by a distance of approximately at least 250 nm, preferably at least 300 nm and more preferably at least 350 nm. It will be appreciated that these distances are dependent on magnification, and that other values can be determined accordingly, by one of ordinary skill in the art.
Other techniques such as scanning near-field optical microscopy (SNOM) are available which are capable of greater optical resolution, thereby permitting more dense arrays to be used. For example, using SNOM, adjacent molecules may be separated by a distance of less than 100 nm, e.g. 10 nm. For a description of scanning near-field optical microscopy, see Moyer et al., Laser Focus World (1993) 29(10).
An additional technique that may be used is surface-specific total internal reflection fluorescence microscopy (T FM); see, for example, Vale et al., Nature, (1996) 380: 451-453). Using this technique, it is possible to achieve wide-field imaging (up to 100 μm×100 μm) with single molecule sensitivity. This may allow arrays of greater than 107 resolvable molecules per cm2 to be used.
Additionally, the techniques of scanning tunnelling microscopy (Binnig et al., Helvetica Physica Acta (1982) 55:726-735) and atomic force microscopy (Hansma et al., Ann. Rev. Biophys. Biomol. Struct. (1994) 23:115-139) are suitable for imaging the arrays of the present invention. Other devices which do not rely on microscopy may also be used, provided that they are capable of imaging within discrete areas on a solid support.
Single molecules may be arrayed by immobilisation to the surface of a solid support. This may be carried out by any known technique, provided that suitable conditions are used to ensure adequate separation of the molecules. Generally the array is produced by dispensing small volumes of a sample containing a mixture of molecules onto a suitably prepared solid surface, or by applying a dilute solution to the solid surface to generate a random array. In this manner, a mixture of different molecules may be arrayed by simple means. The formation of the single molecule array then permits interrogation of each arrayed molecule to be carried out.
Suitable solid supports are available commercially, and will be apparent to the skilled person. The supports may be manufactured from materials such as glass, ceramics, silica and silicon. The supports usually comprise a flat (planar) surface, or at least an array in which the molecules to be interrogated are in the same plane. Any suitable size may be used. For example, the supports might be of the order of 1-10 cm in each direction.
It is important to prepare the solid support under conditions which minimise or avoid the presence of contaminants. The solid support must be cleaned thoroughly, preferably with a suitable detergent, e.g. Decon-90, to remove dust and other contaminants.
Immobilisation may be by specific covalent or non-covalent interactions. Covalent attachment is preferred. If the molecule is a polynucleotide, immobilisation win preferably be at either the 5′ or 3′ position, so that the polynucleotide is attached to the solid support at one end only. However, the polynucleotide may be attached to the solid support at any position along its length, the attachment acting to tether the polynucleotide to the solid support. The immobilised polynucleotide is then able to undergo interactions with other molecules or cognates at positions distant from the solid support. Typically the interaction will be such that it is possible to remove any molecules bound to the solid support through non-specific interactions, e.g. by washing. Immobilisation in this manner results in well separated single molecules. The advantage of this is that it prevents interaction between neighbouring molecules on the array, which may hinder interrogation of the array.
In one embodiment of the invention, the surface of a solid support is first coated with streptavidin or avidin, and then a dilute solution of a biotinylated molecule is added at discrete sites on the surface using, for example, a nanolitre dispenser to deliver one molecule on average to each site.
In a preferred embodiment of the invention, the solid surface is coated with an epoxide and the molecules are coupled via an amine linkage. It is also preferable to avoid or reduce salt present in the solution containing the molecule to be arrayed. Reducing the salt concentration minimises the possibility of the molecules aggregating in the solution, which may affect the positioning on the array.
If the molecule is a polynucleotide, then immobilisation may be via hybridisation to a complementary nucleic acid molecule previously attached to a solid support. For example, the surface of a solid support may be first coated with a primer polynucleotide at discrete sites on the surface. Single-stranded polynucleotides are then brought into contact with the arrayed primers under hybridising conditions and allowed to “self-sort” onto the array. In this way, the arrays may be used to separate the desired polynucleotides from a heterogeneous sample of polynucleotides.
Alternatively, the arrayed primers may be composed of double-stranded polynucleotides with a single-stranded overhang (“sticky-ends”). Hybridisation with target polynucleotides is then allowed to occur and a DNA ligase used to covalently link the target DNA to the primer. The second DNA strand can then be removed under melting conditions to leave an arrayed polynucleotide.
In an embodiment of the invention, the target molecules are immobilised onto non-fluorescent streptavidin or avidin-functionalised polystyrene latex microspheres, as shown in
In one aspect of the present invention, the devices comprise arrayed polynucleotides, each polynucleotide comprising a hairpin loop structure, one end of which comprises a target polynucleotide, the other end comprising a relatively short polynucleotide capable of acting as a primer in the polymerase reaction. This ensures that the primer is able to perform its priming function during a polymerase-based sequencing procedure, and is not removed during any washing step in the procedure. The target polynucleotide is capable of being interrogated.
The term “hairpin loop structure” refers to a molecular stem and loop structure formed from the hybridisation of complementary polynucleotides that are covalently linked. The stem comprises the hybridised polynucleotides and the loop is the region that covalently links the two complementary polynucleotides. Anything from a 10 to 20 (or more) base pair double-stranded (duplex) region may be used to form the stem. In one embodiment, the structure may be formed from a single-stranded polynucleotide having complementary regions. The loop in this embodiment may be anything from 2 or more non-hybridised nucleotides. In a second embodiment, the structure is formed from two separate polynucleotides with complementary regions, the two polynucleotides being linked (and the loop being at least partially formed) by a linker moiety. The linker moiety forms a covalent attachment between the ends of the two polynucleotides. Linker moieties suitable for use in this embodiment will be apparent to the skilled person. For example, the linker moiety may be polyethylene glycol (PEG).
There are many different ways of forming the hairpin structure to incorporate the target polynucleotide. However, a preferred method is to form a first molecule capable of forming a hairpin structure, and ligate the target polynucleotide to this. Ligation may be carried out either prior to or after immobilisation to the solid support. The resulting structure comprises the single-stranded target polynucleotide at one end of the hairpin and a primer polynucleotide at the other end.
In one embodiment, the target polynucleotide is genomic DNA purified using conventional methods. The genomic DNA may be PCR-amplified or used directly to generate fragments of DNA using either restriction endonucleases, other suitable enzymes, a mechanical form of fragmentation or a non-enzymatic chemical fragmentation method. In the case of fragments generated by restriction endonucleases, hairpin structures bearing a complementary restriction site at the end of the first hairpin may be used, and selective ligation of one strand of the DNA sample fragments may be achieved by one of two methods.
Method 1 uses a first hairpin whose restriction site contains a phosphorylated 5′ end. Using this method, it may be necessary to first de-phosphorylate the restriction-cleaved genomic or other DNA fragments prior to ligation such that only one sample strand is covalently ligated to the hairpin.
Method 2: in the design of the hairpin, a single (or more) base gap can be incorporated at the 3′ end (the receded strand) such that upon ligation of the DNA fragments only one strand is covalently joined to the hairpin. The base gap can be formed by hybridising a further separate polynucleotide to the 5′-end of the first hairpin structure. On ligation, the DNA fragment has one strand joined to the 5′-end of the first hairpin, and the other strand joined to the 3′-end of the further polynucleotide. The further polynucleotide (and the other strand of the DNA fragment) may then be removed by disrupting hybridisation.
In either case, the net result should be covalent ligation of only one strand of a DNA fragment of genomic or other DNA, to the hairpin. Such ligation reactions may be carried out in solution at optimised concentrations based on conventional ligation chemistry, for example, carried out by DNA ligases or non-enzymatic chemical ligation. Should the fragmented DNA be generated by random shearing of genomic DNA or polymerase, then the ends can be filled in with Klenow fragment to generate blunt-ended fragments which may be blunt-end-ligated onto blunt-ended hairpins. Alternatively, the blunt-ended DNA fragments may be ligated to oligonucleotide adapters which are designed to allow compatible ligation with the sticky-end hairpins, in the manner described previously.
The hairpin-ligated DNA constructs may then be covalently attached to the surface of a solid support to generate a single molecule array (SMA), or ligation may follow attachment to form the array.
The arrays may then be used in procedures to determine the sequence of the target polynucleotide. If the target fragments are generated via restriction digest of genomic DNA, the recognition sequence of the restriction or other nuclease enzyme will provide 4, 6, 8 bases or more of known sequence (dependent on the enzyme). Further sequencing of between 10 and 20 bases on the SMA should provide sufficient overall sequence information to place that stretch of DNA into unique context with a total human genome sequence, thus enabling the sequence information to be used for genotyping and more specifically single nucleotide polymorphism (SNP) scoring.
Simple calculations have suggested the following based on sequencing a 107 molecule SMA prepared from hairpin ligation: for a 6 base pair recognition sequence, a single restriction enzyme will generate approximately 106 ends of DNA If a stretch of 13 bases is sequenced on the SMA (i.e. 13×106 bases), approximately 13,000 SNPs will be detected. One application of such a sample preparation and sequencing format would in general be for SNP discovery in pharmaco-genetic analysis. The approach is therefore suitable for forensic analysis or any other system which requires unambiguous identification of individuals to a level as low 103 SNPs.
It is of course possible to sequence the complete target polynucleotide, if required.
In a separate aspect of the invention, the devices may comprise immobilised polynucleotides and other immobilised molecules. The other molecules are relatively short compared to the polynucleotides and are intended to prevent non-specific attachment of reagents, e.g. fluorophores, with the solid support, thereby reducing background interference. In one embodiment, the other molecules are relatively short polynucleotides. However, many different molecules may be used, e.g. peptides, proteins, polymers and synthetic chemicals, as will be apparent to the skilled person. Preparation of the devices may be carried out by first preparing a mixture of the relatively long polynucleotides and of the relatively short molecules. Usually, the concentration of the latter will be in excess of that of the long polynucleotides. The mixture is then placed in contact with a suitably prepared solid support, to allow immobilisation to occur.
The single molecule arrays have many applications in methods which rely on the detection of biological or chemical interactions with arrayed molecules. For example, the arrays may be used to determine the properties or identities of cognate molecules. Typically, interaction of biological or chemical molecules with the arrays are carried out in solution.
In particular, the arrays may be used in conventional assays which rely on the detection of fluorescent labels to obtain information on the arrayed molecules. The arrays are particularly suitable for use in multi-step assays where the loss of synchronisation in the steps was previously regarded as a limitation to the use of arrays. When the arrays are composed of polynucleotides they may be used in conventional techniques for obtaining genetic sequence information. Many of these techniques rely on the stepwise identification of suitably labelled nucleotides, referred to in U.S. Pat. No. 5,634,413 as “single base” sequencing methods.
In an embodiment of the invention, the sequence of a target polynucleotide is determined in a similar manner to that described in U.S. Pat. No. 5,634,413, by detecting the incorporation of nucleotides into the nascent strand through the detection of a fluorescent label attached to the incorporated nucleotide. The target polynucleotide is primed with a suitable primer (or prepared as a hairpin construct which will contain the primer as part of the hairpin), and the nascent chain is extended in a stepwise manner by the polymerase reaction. Each of the different nucleotides (A, T, G and C) incorporates a unique fluorophore at the 3′ position which acts as a blocking group to prevent uncontrolled polymerisation. The polymerase enzyme incorporates a nucleotide into the nascent chain complementary to the target, and the blocking group prevents further incorporation of nucleotides. The array surface is then cleared of unincorporated nucleotides and each incorporated nucleotide is “read” optically by a charge-coupled detector using laser excitation and filters. The 3′-blocking group is then removed (deprotected), to expose the nascent chain for further nucleotide incorporation.
Because the array consists of distinct optically resolvable polynucleotides, each target polynucleotide will generate a series of distinct signals as the fluorescent events are detected. Details of the full sequence are then determined.
The number of cycles that can be achieved is governed principally by the yield of the deprotection cycle. If deprotection fails in one cycle, it is possible that later deprotection and continued incorporation of nucleotides can be detected during the next cycle. Because the sequencing is performed at the single molecule level, the sequencing can be carried out on different polynucleotide sequences at one time without the necessity for separation of the different sample fragments prior to sequencing. This sequencing also avoids the phasing problems associated with prior art methods.
Deprotection may be carried out by chemical, photochemical or enzymatic reactions.
A similar, and equally applicable, sequencing method is disclosed in EP-A-0640146.
Other suitable sequencing procedures will be apparent to the skilled person. In particular, the sequencing method may rely on the degradation of the arrayed polynucleotides, the degradation products being characterised to determine the sequence.
An example of a suitable degradation technique is disclosed in WO-A-95/20053, whereby bases on a polynucleotide are removed sequentially, a predetermined number at a time, through the use of labelled adaptors specific for the bases, and a defined exonuclease cleavage.
A consequence of sequencing using non-destructive methods is that it is possible to form a spatially addressable array for further characterisation studies, and therefore non-destructive sequencing may be preferred. In this context, term “spatially addressable” is used herein to describe how different molecules may be identified on the basis of their position on an array.
Once sequenced, the spatially addressed arrays may be used in a variety of procedures which require the characterisation of individual molecules from heterogeneous populations.
One application is to use the arrays to characterise products synthesised in combinatorial chemistry reactions. During combinatorial synthesis reactions, it is usual for a tag or label to be incorporated onto a beaded support or reaction product for the subsequent characterisation of the product. This is adapted in the present invention by using polynucleotide molecules as the tags, each polynucleotide being specific for a particular product, and using the tags to hybridise onto a spatially addressed array. Because the sequence of each arrayed polynucleotide has been determined previously, the detection of an hybridisation event on the array reveals the sequence of the complementary tag on the product. Having identified the tag, it is then possible to confirm which product this relates to. The complete process is therefore quick and simple, and the arrays may be reused for high through-put screening. Detection may be carried out by attaching a suitable label to the product, e.g. a fluorophore.
Combinatorial chemistry reactions may be used to synthesise a diverse range of different molecules, each of which may be identified using the addressed arrays of the present invention. For example, combinatorial chemistry may be used to produce therapeutic proteins or peptides that can be bound to the arrays to produce an addressed array of target proteins. The targets may then be screened for activity, and those proteins exhibiting activity may be identified by their position on the array as outlined above.
Similar principles apply to other products of combinatorial chemistry, for example the synthesis of non-polymeric molecules of m.wt.<1000. Methods for generating peptides/proteins by combinatorial methods are disclosed in U.S. Pat. No. 5,643,768 and U.S. Pat. No. 5,658,754. Split-and-mix approaches may also be used, as described in Nielsen et al., J. Am. Chem. Soc. (1993) 115:9812-9813.
In an alternative approach, the products of the combinatorial chemistry reactions may comprise a second polynucleotide tag not involved in the hybridisation to the array. After formation by hybridisation, the array may be subjected to repeated polynucleotide sequencing to identify the second tag which remains free. The sequencing may be carried out as described previously.
Therefore, in this application, it is the tag that provides the spatial address on the array. The tag may then be removed from the product by, for example, a cleavable linker, to leave an untagged spatially addressed array.
A further application is to display proteins via an immobilised polysome containing trapped polynucleotides and protein in a complex, as described in U.S. Pat. No. 5,643,768 and U.S. Pat. No. 5,658,754.
In a separate embodiment of the invention, the arrays may be used to characterise an organism. For example, an organism's genomic DNA may be screened using the arrays, to reveal discrete hybridisation patterns that are unique to an individual. This embodiment may therefore be likened to a “bar code” for each organism. The organism's genomic DNA may be first fragmented and detectably-labelled, for example with a fluorophore. The fragmented DNA is then applied to the array under hybridising conditions and any hybridisation events monitored.
Alternatively, hybridisation may be detected using an in-built fluorescence based detection system in the arrayed molecule, for example using the “molecular beacons” described in Nature Biotechnology (1996) 14:303-308.
It is possible to design the arrays so that the hybridisation pattern generated is unique to the organism and so could be used to provide valuable information on the genetic character of an individual. This may have many useful applications in forensic science. Alternatively, the methods may be carried out for the detection of mutations or allelic variants within the genomic DNA of an organism.
For genotyping, it is desirable to identify if a particular sequence is present in the genome. The smallest possible unique oligomer is a 16-mer (assuming randomness of the genome sequence), i.e. statistically there is a probability of any given 16-base sequence occurring only once in the human genome (which has 3×109 bases). There are ˜4×109 possible 16-mers which would fit within a region of 2 cm×2 cm (assuming a single copy at a density of 1 molecule per 250 nm×250 nm square). It is therefore necessary to determine only if a particular 16-mer is present or not, and so quantitative measurements are unnecessary. Identifying a mutation in a particular region and what the mutation is can be carried out using the 16-mer library. Mapping back onto the human genome would be possible using published data and would not be a problem once the entire genome has been determined. There is built-in self-check, by looking at the hybridisation to particular 16-mers so that if there is a single point mutation, this will show up in 16 different 16-mers, identifying a region of 32 bases in the genome (the mutation would occur at the top of one 16-mer and then at the second base in a related 16-mer etc). Thus, a single point mutation would result in 16 of the 16-mers not showing hybridisation and a new set of 16 showing hybridisation plus the same thing for the complementary strand. In summary, considering both strands of DNA, a single point mutation would result in 32 of the 16-mers not showing hybridisation and 32 new 16-mers showing hybridisation, i.e. quite large changes on the hybridisation pattern to the array.
By way of example, a sample of human genomic DNA may be restriction-digested to generate short fragments, then labelled using a fluorescently-labelled monomer and a DNA polymerase or a terminal transferase enzyme. This produces short lengths of sample DNA with a fluorophore at one end. The melted fragments may then be exposed to the array and the pixels where hybridisation occurs or not would be identified. This produces a genetic bar code for the individual with (if oligonucleotides of length 16 were used) ˜4×109 binary coding elements. This would uniquely define a person's genotype for pharmagenomic applications. Since the arrays should be reusable, the same process could be repeated on a different individual.
In one embodiment of the invention, a method for determining a single nucleotide polymorphism (SNP) present in a genome comprises immobilising fragments of the genome onto the surface of a solid support to form an array as defined above, identifying nucleotides at selected positions in the genome, and comparing the results with a known consensus sequence to identify any differences between the consensus sequence and the genome. Identifying the nucleotides at selected positions in the genome may be carried out by contacting the array sequentially with each of the bases A, T, G and C, under conditions that permit the polymerase reaction to proceed, and monitoring the incorporation of a base at selected positions in the complementary sequence.
The fragments of the genome may be unamplified DNA obtained from several cells from an individual, which is treated with a restriction enzyme. As indicated above, it is not necessary to determine the sequence of the full fragment. For example, it may be preferable to determine the sequence of 16-30 specific bases, which is sufficient to identify the DNA fragment by comparison to a consensus sequence, e.g. to that known from the Human Genome Project. Any SNP occurring within the sequenced region can then be identified. The specific bases do not have to be contiguous. For example, the procedure may be carried out by the incorporation of non-labelled bases followed, at pre-determined positions, by the incorporation of a labelled base. Provided that the sequence of sufficient bases is determined, it should be possible to identify the fragment. Again, any SNPs occurring at the determined base positions, can be identified. For example, the method may be used to identify SNPs that occur after cytosine. Template DNA (genomic fragments) can be contacted with each of the bases A, T and G, added sequentially or together, so that the complementary strand is extended up to a position that requires C. Non-incorporated bases can then be removed from the array, followed by the addition of C. The addition of C is followed by monitoring the next base incorporation (using a labelled base). By repeating this process a sufficient number of times, a partial sequence is generated where each base immediately following a C is known. It will then be possible to identify the full sequence, by comparison of the partial sequence to a reference sequence. It will then also be possible to determine whether there are any SNPs occurring after any C.
To further illustrate this, a device may comprise 107 restriction fragments per cm2. If 30 bases are determined for each fragment, this means 3×108 bases are identified. Statistically, this should determine 3×105 SNPs for the experiment. If the fragments each comprise 1000 nucleotides, it is possible to have 1010 nucleotides per cm2, or three copies of the human genome. The approach therefore permits large sequence or SNP analysis to be performed.
Viral and bacterial organisms may also be studied, and screening nucleic acid samples may reveal pathogens present in a disease, or identify microorganisms in analytical techniques. For example, pathogenic or other bacteria may be identified using a series of single molecule DNA chips produced from different strains of bacteria. Again, these chips are simple to make and reusable.
In a further example, double-stranded arrays may be used to screen protein libraries for binding, using fluorescently labelled proteins. This may determine proteins that bind to a particular DNA sequence, i.e. proteins that control transcription. Once the short sequence that the protein binds to has been determined, it may be made and affinity purification used to isolate and identify the protein. Such a method could find all the transcription-controlling proteins. One such method is disclosed in Nature Biotechnology (1999) 17:573-577.
Another use is in expression monitoring. For this, a label is required for each gene. There are approximately 100,000 genes in the human genome. There are 262,144 possible 9-mers, so this is the minimum length of oligomer needed to have a unique tag for each gene. This 9-mer label needs to be at a specific point in the DNA and the best point is probably immediately after the poly-A tail in the mRNA (i.e. a 9-mer linked to a poly-T guide sequence). Multiple copies of these 9-mers should be present, to permit quantitation of gene expression. 100 copies would allow determination of relative expression from 1-100%. 10,000 copies would allow determination of relative gene expression from 0.01-100%. 10,000 copies of 262,144 9-mers would fit inside 1 cm×1 cm at close to maximum density.
The use of nanovials in conjunction with any of the above methods may allow a molecule to be cleaved from the surface, yet retain its spatial integrity. This permits the generation of spatially addressable arrays of single molecules in free solution, which may have advantages where the surface attachment impedes the analysis (e.g. drug screening). A nanovial is a small cavity in a flat glass surface, e.g. approx 20 μm in diameter and 10 μm deep. They can be placed every 50 μm, and so the array would be less dense than a surface-attached array; however, this could be compensated for by appropriate adjustment in the imaging optics.
The following Examples illustrate the invention, with reference to the accompanying drawings.
The microscope set-up used in the following Example was based on a modified confocal fluorescence system using a photon detector as shown in
The target sample was a 5′-biotin-modified 13-mer primer oligonucleotide prepared using conventional phosphoramidite chemistry, and having SEQ ID No. 1 (see listing, below). The oligonucleotide was post-synthetically modified by reaction of the uridine base with the succinimdyl ester of tetramethylrhodamine (TMR).
Glass coverslips were prepared by cleaning with acetone and drying under nitrogen. A 50 μl aliquot of biotin-BSA (Sigma) redissolved in PBS buffer (0.01 M, pH 7.4) at 1 mg/ml concentration was deposited on the clean coverslip and incubated for 8 hours at 30° C. Excess biotin-BSA was removed by washing 5 times with MilliQ water and drying under nitrogen. Non-fluorescent streptavidin functionalised polystyrene latex microspheres of diameter 500 nm (Polysciences Inc.) were diluted in 100 mM NaCl to 0.1 solids and deposited as a 1 μl drop on the biotinylated coverslip surface. The spheres were allowed to dry for one hour and unbound beads removed by washing 5 times with MilliQ water. This procedure resulted in a surface coverage of approximately 1 sphere/100 μm×100 μm.
The non-fluorescent microspheres were found to have a broad residual fluorescence at excitation wavelength 514 nm, probably arising from small quantities of photoactive constituents used in the colloidal preparation of the microspheres. The microspheres were therefore photobleached by treating the prepared coverslip in a laser beam of a frequency doubled (532 nm) Nd:YAG pulsed dye laser, for 1 hour.
The biotinylated 13-TMR ssDNA was coupled to the streptavidin functionalised microspheres by incubating a 50 μl sample of 0.1 pM DNA (diluted in 100 mM NaCl, 100 mM Tris) deposited over the microspheres. Unbound DNA was removed by washing the coverslip surface 5-times with MilliQ water.
Low light level illumination from the microscope condenser was used to position visually a microsphere at 10× magnification so that when the laser was switched on the sphere was located in the centre of the diffraction limited focus. The condenser was then turned off and the light path switched to the fluorescence detection port. The MCS was initiated and the fluorescence omitted from the latex sphere recorded on one or both channels. The sample was excited at 514 nm and detection was made on the 600 nm channel.
This Example illustrates the preparation of single molecule arrays by direct covalent attachment to glass followed by a demonstration of hybridisation to the array.
Covalently modified slides were prepared as follows. Spectrosil-2000 slides (TSL, UK) were rinsed in milli-Q to remove any dust and placed wet in a bottle containing neat Decon-90 and left for 12 h at room temperature. The slides were rinsed with milli-Q and placed in a bottle containing a solution of 1.5% glycidoxypropyltrimethoxy-silane in milli-Q and magnetically stirred for 4 h at room temperature rinsed with milli-Q and dried under N2 to liberate an epoxide coated surface.
The DNA used was that shown in SEQ ID No. 2 (see sequence listing below), where n represents a 5-methyl cytosine (Cy5) with a TMR group coupled via a linker to the n4 position.
A sample of this (5 μl, 450 pM) was applied as a solution in neat milli-Q.
The DNA reaction was left for 12 h at room temperature in a humid atmosphere to couple to the epoxide surface. The slide was then rinsed with milli-Q and dried under N2.
The prepared slides can be stored wrapped in foil in a desiccator for at least a week without any noticeable contamination or loss of bound material. Control DNA of the same sequences and fluorophore but without the 5′-amino group shows little stable coverage when applied at the same concentration.
The TMR labelled slides were then treated with a solution of complementary DNA (SEQ ID No. 3) (5 μM, 10 μl) in 100 mM PBS. The complementary DNA has the sequence shown in SEQ ID No. 3, where n represents a methylcytosine group.
After 1 hour at room temperature the slides were cooled to 4° C. and left for 24 hours. Finally, the slides were washed in PBS (100 mM, 1 mL) and dried under N2.
A chamber was constructed on the slide by sealing a coverslip (No. 0, 22×22 mm, Chance Propper Ltd, UK) over the sample area on two sides only with prehardened microscope mounting medium (Eukitt, O. Kindler GmbH & Co., Freiburg, Germany) whilst maintaining a gap of less than 200 μm between slide and coverslip. The chamber was flushed 3× with 100 μl PBS (100 nM NaCl) and allowed to stabilise for 5 minutes before analysing on a fluorescence microscope.
The slide was inverted so that the chamber coverslip contacted the objective lens of an inverted microscope (Nikon TE200) via an immersion oil interface. A 60° fused silica dispersion prism was optically coupled to the back of the slide through a thin film of glycerol. Laser light was directed at the prism such that at the glass/sample interface it subtends an angle of approximately 68° to the normal of the slide and subsequently undergoes Total Internal Reflection (TIR). The critical angle for glass/water interface is 66°.
Fluorescence from single molecules of DNA-TM or DNA-Cy5 produced by excitation with the surface specific evanescent wave following TIR is collected by the objective lens of the microscope and imaged onto an Intensified Charge Coupled Device (ICCD) camera (Pentamax, Princeton Instruments, NJ). Two images were recorded using a combination of 1) 532 nm excitation (frequency doubled solid state Nd:YAG, Antares, Coherent) with a 580 nm fluorescence (580DF30, Omega Optics, USA) filter for TMR and 2) 630 nm excitation (nd:YAG pumped dye laser, Coherent 700) with a 670 nm filter (670DF40, Omega Optics, USA) for Cy5. Images were recorded with an exposure time of 500 ms at the maximum gain of 10 on the ICCD. Laser powers incident at the prism were 50 mW and 40 mW at 532 nm and 630 nm respectively. A third image was taken with 532 nm excitation and detection at 670 nm to determine the level of cross-talk from TMR on the Cy5 channel.
Single molecules were identified by single points of fluorescence with average intensities greater than 3× that of the background. Fluorescence from a single molecule is confined to a few pixels, typically a 3×3 matrix at 100× magnification, and has a narrow Gaussian-like intensity profile. Single molecule fluorescence is also characterised by a one-step photobleaching process in the time course of the intensity and was used to distinguish single molecules from pixel regions containing two or more molecules, which exhibited multi-step processes.
To count molecules, a threshold for fluorescence intensities is first set to exclude background noise. For a control sample, the background is essentially the thermal noise of the ICCD measured to be 76 counts with a standard deviation of only 6 counts. A threshold is arbitrarily chosen as a linear combination of the background, the average counts over an image and the standard deviation over an image. In general, the latter two quantities provide a measure of the number of pixels and range of intensities above background. This method gives rise to threshold levels which are at least 12 standard deviations above the background with a probability of less than 1 in 144 pixels contributing from noise. By defining a single molecule fluorescent point as being at least a 2×2 matrix of pixels and no larger than a 7×7, the probability of a single background pixel contributing to the counting is eliminated and clusters are ignored.
In this manner, the surface density of single molecules of DNA-TMR is measured at 2.9×106 molecules/cm2 (238 molecules in
Hybridisation was identified by the co-localisation of discreet points of fluorescence from single molecules of TMR and Cy-5 following the superposition of two images.
This Example demonstrates that single molecule arrays can be formed, and hybridisation events detected according to the invention. It is expected that the skilled person will realise that modifications may be made to improve the efficiency of the process. For example, improved washing steps, e.g. using a flow cell, would reduce background noise and permit more concentrated solutions to be used, and hybridisation protocols could be adapted by varying the parameters of temperature, buffer, time etc.
This experiment demonstrates the possibility of performing enzymatic incorporation on a single molecule array. In summary, primer DNA was attached to the surface of a solid support, and template DNA hybridised thereto. Two cycles of incorporation of fluorophore-labelled nucleotides was then completed. This was compared against a reference experiment where the immobilised DNA was pre-labelled with the same two fluorophores prior to attachment to the surface, and control experiments performed under adverse conditions for nucleotide incorporation.
The primer DNA sequence and the template DNA sequence used in this experiment are shown in SEQ ID NOS. 4 and 5, respectively.
The buffer used contained 4 mM MgCl2, 2 mM DTT, 50 mM Tris. HCl (pH 7.6) 10 mM NaCl and 1 mm K22PO3 (100 μl).
Preparation of Slides
Silica slides were treated with decon for at least 24 hours and rinsed in water and EtOH directly before use. The dried slides were placed in a 50 ml solution of 2% glycidoxypropyltrimethoxysilane in EtOH/H2SO4 (2 drops/500 ml) at room temperature for 2 hours. The slides were then rinsed in EtOH from a spray bottle and dried under N2.
The DNA samples (SEQ ID NO. 4) were applied either as a 40-100 pM solution (5 μl) in 10 mM K2PO3 pH 7.6 (allowed to dry overnight), or at least 1 μM concentration over a sealed slide. The slides allowed to dry overnight were left over a layer of water for 18 hours at room temperature and then rinsed with milli-q (approx. 30 ml from a spray bottle) and dried under N2. The sealed slides were simply flushed with 50 ml buffer prior to use. Control slides with no coupled DNA were simply left under the buffer for identical time periods.
Enzyme Extensions on a Surface
For the first incorporation cycle, samples were prepared with the buffer containing BSA (to 0.2 mg/ml), the triphosphate (Cy3dUTP; to 20 μM) and the polymerase enzyme (T4 exo-; to 500 nM). In certain experiments, the template DNA was also added at 2 μM. The mixture was flowed into cells which were incubated at 37° C. for 2 hours and flushed with 500 ml buffer. The second incorporation cycle with Cy5dCTP (20 μM), dATP (100 μm) and dGTP (100 μM) was performed in the same way. The cells were flushed with 50 ml buffer and left for 12 hours prior to imaging. Control reactions were performed as above with: a) no DNA coupled prior to extension; b) DNA attached but no polymerase in the extension buffer; and c) DNA attached, but the polymerase denatured by boiling.
Reference Sample
A reference sample, not immobilised to the surface, was prepared in the following way.
Buffer containing 1 μM of the sample DNA, BSA (0.2 mg/ml), TMR-labelled dUTP (20 μM) and the polymerase enzyme (T4 exo-; 500 nM; 100 μl) was prepared.
The reaction was analysed and purified by reverse phase HPLC (5-30% acetonitrile in ammonium acetate over 30 min.) with UV and fluorescence detection. In all cases, the labelled DNA was clearly separate from both the unlabelled DNA and the labelled dNTP's. The material was concentrated and dissolved in 10 mM K2PO3 for analysis by A260 and fluorescence. The material purified by HPLC was further extended with labelled dCTP (20 μM), dATP (100 μM) and dGTP (100 μM) and HPLC purified again. Surface coupling was then performed dry, at 100 pM concentrations.
Microscopic Analysis
Following the single molecule DNA attachment procedure and extension reactions, the sample cells were analysed on a single molecule total internal reflection fluorescence microscope (TIRFM) in the following manner. A 60° fused silica dispersion prism was coupled optically to the slide through an aperture in the cell via a thin film of glycerol. Laser light was directed at the prism such that at the glass/sample interface it subtends an angle of approximately 68° to the normal of the slide and subsequently undergoes total internal reflection. The critical angle for a glass/water interface is 66°. An evanescent field is generated at the interface which penetrates only ˜150 nm into the aqueous phase. Fluorescence from single molecules excited within this evanescent field is collected by a 100× objective lens of an inverted microscope, filtered spectrally from the laser light and imaged onto an Intensified Charge Coupled Device (ICCD) camera.
Two 90 μm×90 μm images were recorded using a combination of: 1) 532 nm excitation (frequently doubled Nd:YAG) with a 580 nm interference filter for Cy3 detection; and 2) 630 nm excitation (Nd:YAG pumped DCM dye laser) with a 670 nm filter for Cy5 detection. Images were recorded with an exposure time of 500 ms at the maximum ICCD gain of 5.75 counts/photoelectron. Laser powers incident at the prism were 30 mW and 30 mW at 532 nm and 630 nm respectively. Two colour fluorophore labelled nucleotide incorporations are identified by the co-localisation of discreet points of fluorescence from single molecules of Cy3 and Cy5 following superimposing the two images. Molecules are considered co-localised when fluorescent points are within a pixel separation of each other. For a 90 μm×90 μm field, projected onto a CCD array of 512×512 pixels, the pixel size dimension is 0.176 μm.
Results
The results of the experiment are shown in Table 1. The values shown are an average of the number of molecules imaged (Cy3 and Cy5) over all frames (100 in each) compiled in each experiment and the number of those molecules which are co-localised. The final column represents the number of co-localised molecules expected if the two fluorophores were randomly dispersed across the sample slide (N˜πΔr where n is the surface density of molecules and Δr=0.176 μm is the minimum measurable separation). The number in brackets indicates the magnitude by which the level of co-locations in each experiment is greater than random.
The percentage of co-localisation observed on this sample represents the maximum measurable for a dual labelled system, i.e. there is a detection ceiling due to photophysical effects which means the level is not 100%. These effects may arise from interactions of the fluorophores with the DNA or the surface or both.
There is a statistically higher level of co-localisation in the incorporation experiments compared to the controls (8% versus 2% respectively). This shows that it is possible to perform enzymatic incorporation on the SMA and the level of incorporation is close to that of the reference sequence. Improvements in the surface attachment and the nature of the surface are required to increase the level of co-localisation in the reference and to increase the detection efficiency of the enzymatic incorporation.
This Example illustrates the preparation of single molecule arrays by direct covalent attachment of hairpin loop structures to glass.
A solution of 1% glycidoxypropyltrimethoxy-silane in 95% ethanol/5% water with 2 drops H2SO4 per 500 ml was stirred for 5 minutes at room temperature. Clean, dry Spectrosil-2000 slides (TSL, UK) were placed in the solution and the stirring stopped. After 1 hour the slides were removed, rinsed with ethanol, dried under N2 and oven-cured for 30 min. at 100° C. These ‘epoxide’ modified slides were then treated with 1 μM of labelled DNA (5′-Cy3-CTGCTGAAGCGTCGGCAGGT-heg-ami-nodT-heg-ACCTGCCGACGCT-3′) (SEQ ID NOS. 6 and 7) in 50 mM potassium phosphate buffer, pH 7.4 for 18 hours at room temperature and, prior to analysis, flushed with 50 mM potassium phosphate, 1 mM EDTA, pH 7.4. The coupling reactions were performed in sealed teflon blocks under a pre-mounted coverslip to prevent evaporation of the sample and allow direct imaging.
The DNA structure was designed as a self-priming template system with an internal amino group attached as an amino deoxy-thymidine held by two 18 atom hexaethylene glycol (heg) spacers, and was synthesised by conventional DNA synthesis techniques using phosphoramidite monomers.
For imaging, one slide was inverted so that the chamber coverslip contacted the objective lens of an inverted microscope (Nikon TE200) via an immersion oil interface. A 60° fused silica dispersion prism was coupled optically to the back of the slide through a thin film of glycerol. Laser light was directed at the prism such that at the glass/sample interface it subtends an angle of approximately 68° to the normal of the slide and subsequently undergoes Total Internal Reflection (TIR). The critical angle for glass/water interface is 66°.
Fluorescence from single molecules of DNA-Cy3, produced by excitation with the surface-specific evanescent wave following TIR, was collected by the objective lens of the microscope and imaged onto an Intensified Charge Coupled Device (ICCD) camera (Pentamax, Princeton Instruments, NJ). The image was recorded using a 532 nm excitation (frequency-doubled solid-state Nd:YAG, Antares, Coherent) with a 580 nm fluorescence (580DF30, Omega Optics, USA) filter for Cy3. Images were recorded with an exposure time of 500 ms at the maximum gain of 10 on the ICCD. Laser powers incident at the prism were 50 mW at 532 nm.
Single molecules were identified as described in Example 2.
The surface density of single molecules of DNA-Cy3 was measured at approximately 500 per 100 μm×100 μm image or 5×106 cm2.
Section B (from U.S. Ser. No. 10/153,267)
This invention relates to fabricated arrays of polynucleotides, and to their analytical applications.
Advances in the study of molecules have been led, in part, by improvement in technologies used to characterise the molecules or their biological reactions. In particular, the study of nucleic acids, DNA and RNA, has benefited from developing technologies used for sequence analysis and the study of hybridisation events.
An example of the technologies that have improved the study of nucleic acids, is the development of fabricated arrays of immobilised nucleic acids. These arrays typically consist of a high-density matrix of polynucleotides immobilised onto a solid support material. Fodor et al., Trends in Biotechnology (1994) 12:19-26, describes ways of assembling the nucleic acid arrays using a chemically sensitised glass surface protected by a mask, but exposed at defined areas to allow attachment of suitably modified nucleotides. Typically, these arrays may be described as “many molecule” arrays, as distinct regions are formed on the solid support comprising a high density of one specific type of polynucleotide.
An alternative approach is described by Schena et al., Science (1995) 270:467-470, where samples of DNA are positioned at predetermined sites on a glass microscope slide by robotic micropipetting techniques. The DNA is attached to the glass surface along its entire length by non-covalent electrostatic interactions. However, although hybridisation with complementary DNA sequences can occur, this approach may not permit the DNA to be freely available for interacting with other components such as polymerase enzymes, DNA-binding proteins etc.
WO-A-96/27025 is a general disclosure of single molecule arrays. Although sequencing procedures are disclosed, there is little description of the applications to which the arrays can be applied. There is also only a general discussion on how to prepare the arrays.
According to the present invention, a device comprises a high density array of single polynucleotide molecules, comprising relatively short molecules and relatively long polynucleotides immobilised on the surface of a solid support, where the relatively long polynucleotides are at a density that permits individual resolution and/or interrogation of those parts that extend beyond the relatively short molecules. The device can be any device that comprises this array, including, but not limited to, a sequencing machine or genetic analysis machine. In this aspect, the relatively short molecules help to control the density of the relatively long polynucleotides, providing a more uniform array of single polynucleotide molecules, thereby improving imaging. The relatively short molecules can also prevent non-specific binding of reagents to the solid support, and therefore reduce background interference. For example, in the context of a polymerase reaction to incorporate nucleoside triphosphates onto a strand complementary to a relatively long polynucleotide, the relatively short molecules prevent the polymerase and nucleosides from attaching to the solid support surface, which may otherwise interfere with the imaging process.
The relatively short molecules can also ensure that each relatively long polynucleotide is maintained upright, preventing the polynucleotides from interacting lengthwise with the solid support, which may otherwise prevent efficient interaction with a reagent, e.g., a polymerase. This can also prevent the fluorophore being quenched by the surface and therefore lead to more accurate imaging of the single polynucleotide molecules.
As used herein, the term “array” refers to a population of polynucleotide molecules that are distributed over a solid support; preferably, these polynucleotides are spaced at a distance from one another sufficient to permit the individual resolution of the polynucleotides.
“Relatively long polynucleotides”, “long polynucleotides”, “and single polynucleotide molecules”, are used interchangably herein. “Relatively short molecules”, “short molecules”, “relatively small molecules” and “small molecules”, are also used interchangably herein. In the context of the present invention, the terms “relatively short” and “relatively long” should be interpreted to mean that the portion of at least a subset of the “relatively long” polynucleotides that is not used for attachment to the substrate or to a linker molecule(s) attached to the substrate, is physically longer than that of the “relatively short” molecules when the relatively long polynucleotides and the relatively short molecules are arrayed. In general, the relatively long polynucleotides can be one nucleotide (or one nucleotide pair, if the polynucleotide is double stranded) or greater in length than the relatively short molecules. That is, the relatively long polynucleotides are longer, with respect to the distance from the planar surface of the solid support, than the relatively short molecules. The length of the long polynucleotides can be 50 to 10,000 nucleotides in length, preferably 100 to 1000 nucleotides in length. If the relatively short molecules are not polynucleotides, then the relatively long polynucleotides are at least the equivalent physical distance of one nucleotide longer (or one nucleotide pair, if the polynucleotide is double stranded) than the relatively short molecules. The term “relatively long” also encompasses polynucleotides which extend above the relatively short molecules in an array format where the relatively long polynucleotides are distributed on the solid support at a density of about 106 to about 109 polynucleotides per cm2, and where the relatively short molecules are distributed at a density greater than about 108 to about 1014 molecules per cm2. In general, the surface of the substrate is engineered so that the short molecules display a hydrophilic group from the surface. The relatively short molecules can therefore be silanes, amino acids, an acid, phosphate, thiophosphate, sulfate, thiol, hydroxyl or polyol, etc. and may include polyethers such as PEG. The types of molecules used will also depend on the surface chemistry used to attach the long molecules to the surface.
As used herein, the term “single polynucleotide molecule” refers to one polymeric molecule of a nucleic acid sequence. Thus, an array feature or address corresponding to a single relatively long polynucleotide consists of one polynucleotide molecule immobilized onto a solid support. The immobilized single polynucleotide molecule can be single- or double-stranded, or have both single-stranded portions and double-stranded portions. For example, it can include a hairpin. In one embodiment, the single polynucleotide molecule is both single-stranded and double-stranded. This is in contrast to the arrays of the prior art, in which a given address typically comprises a plurality of copies (e.g., 10 or more) of a given nucleic acid molecule, often thousands of copies or more. The term “single molecule” is also used herein to distinguish from high density multi-molecule (polynucleotide) arrays in the prior art, which may comprise distinct clusters of many polynucleotides of the same type. As used herein, at least some (e.g., 10 or more) of the addresses in the array are intended to be populated by only one polynucleotide molecule.
“Solid support”, as used herein, refers to the material to which the relatively long polynucleotides and relatively short molecules are attached. Suitable solid supports are available commercially, and will be apparent to the skilled person. The supports can be manufactured from materials such as glass, ceramics, silica and silicon. Supports with a gold surface may also be used. The supports usually comprise a flat (planar) surface, or at least a structure in which the polynucleotides to be interrogated are in approximately the same plane. Alternatively, the solid support can be non-planar, e.g., a microbead. Any suitable size may be used. For example, the supports might be on the order of 1-10 cm in each direction.
The term “individually resolved by optical microscopy” is used herein to indicate that, when visualised, it is possible to distinguish at least one polynucleotide on the array from its neighbouring polynucleotides using optical microscopy methods available in the art. Visualisation may be effected by the use of reporter labels, e.g., fluorophores, the signal of which is individually resolved. As used herein, the term “interrogate” means contacting one or more of the relatively long polynucleotides with another molecule, e.g., a polymerase, a nucleoside triphosphate, a complementary nucleic acid sequence, wherein the physical interaction provides information regarding a characteristic of the arrayed polynucleotide. The contacting can involve covalent or non-covalent interactions with the other molecule. As used herein, “information regarding a characteristic” means information regarding the sequence of one or more nucleotides in the polynucleotide, the length of the polynucleotide, the base composition of the polynucleotide, the Tm of the polynucleotide, the presence of a specific binding site for a polypeptide or other molecule, the presence of an adduct or modified nucleotide, or the three-dimensional structure of the polynucleotide.
As used herein, the term “portion that is immobilized by bonding to the surface” refers to the nucleotide or nucleotides of an immobilized single polynucleotide molecule that is or are either directly involved in linkage to the solid substrate or an intermediate linker molecule (which is then bound to the substrate), or, because of their proximity to the point of immobilization, are not physically accessible to be capable of interrogation (e.g., to serve as a template or substrate for the primer extension activity of a nucleic acid polymerase enzyme). It is preferred that polynucleotides be immobilized by either their 5′ end or their 3′ end, but polynucleotides can also be immobilized via one or more internal nucleotides.
As used herein, the term “portion that is capable of interrogation” refers to that portion of an immobilized polynucleotide molecule that is physically accessible to a physical interaction with another molecule or molecules, the interaction of which provides information regarding a characteristic of the arrayed polynucleotide as defined herein. Generally, the “portion of an immobilized single polynucleotide molecule that is capable of interrogation” is that part which is not the “portion that is immobilized by covalent bonding to the surface” as that term is defined herein.
In one aspect of the invention, the device comprises a high density array of a plurality of first molecules, i.e., the relatively short molecules, and a plurality of second polynucleotides, i.e., the relatively long polynucleotides, immobilised on the surface of a solid support, where each molecule of at least a subset of the plurality of first molecules is shorter in length than the length of each of the second polynucleotide of at least a subset of the plurality of second polynucleotides such that the second polynucleotides are of a length and at a density that permits individual resolution of at least two of the second polynucleotides of the subset. “Plurality” is used to mean that multiple short molecules and multiple long polynucleotides are placed on the array. The short molecules can be of all the same type, or of multiple, i.e., different, types. The long polynucleotides will also generally be of multiple types, and can all be different from each other. The long polynucleotides can also be of different lengths relative to each other, e.g., some of the polynucleotides may be 100 nucleotides in length, while others may be 120 nucleotides in length. By saying that each molecule of “at least a subset” of the plurality of first molecules is shorter in length than the length of each of the second polynucleotide of “at least a subset” of the plurality of second polynucleotides, is meant that one practicing the invention has arrayed polynucleotides that are intended to be physically longer (in that portion of the relatively long polynucleotide that is not used for attachment to the substrate or to a linker molecule(s) attached to the substrate) than the short molecules, but due to breakage of the polynucleotides or binding of short molecules to each other, or some other occurrence, not every individual polynucleotide may be longer than every short molecule.
According to a second aspect of the invention, a method for the production of an array of polynucleotides which are at a density that permits individual resolution, comprises arraying on the surface of a solid support, a mixture of relatively short molecules and relatively long polynucleotides, wherein the short molecules are arrayed in an amount in excess of the polynucleotides. By “in excess” is meant that, in such an embodiment, the small molecules are at a density of from 108 to 1014 molecules/cm2, more preferably greater than 1012 molecules/cm2, whereas the long polynucleotides are at a density of 106 to 109 polynucleotides per cm2, preferably 107 to 109 polynucleotides per cm2.
In another aspect, only a minor proportion of the short molecules that are arrayed at high density on the solid support comprise a group that reacts with the polynucleotides; the majority are non-reactive. In general “a minor proportion” means that reactive and non-reactive molecules exist on the substrate in a ratio of about 1/10 to about 1/1,000,000, preferably about 1/10 to about 1/10,000.
For example, the short molecules can be mixed silanes, a minor proportion of which are reactive with a functional group on the polynucleotides, and the remaining silanes are unreactive and form the array of short molecules on the device. Therefore, controlling the concentration of the minor proportion of short molecules also controls the density of the polynucleotides.
The arrays of the present invention comprise what are effectively single analysable polynucleotides. This has many important benefits for the study of the polynucleotides and their interaction with other biological molecules. In particular, fluorescence events occurring on each polynucleotide can be detected using an optical microscope linked to a sensitive detector, resulting in a distinct signal for each polynucleotide.
When used in a multi-step analysis of a population of single polynucleotides, the phasing problems (loss of syncronization) that are encountered using high density (multi-molecule) arrays of the prior art, can be reduced or removed. Therefore, the arrays also permit a massively parallel approach to monitoring fluorescent or other events on the polynucleotides. Such massively parallel data acquisition makes the arrays extremely useful in a wide range of analysis procedures which involve the screening/characterising of heterogeneous mixtures of polynucleotides.
The preparation of the arrays requires only small amounts of polynucleotide sample and other reagents, and can be carried out by simple means.
a and 6b are images of a single polynucleotide array, where single polynucleotides are indicated by the detection of a fluorescent signal generated on the array.
The single polynucleotide array devices of the present invention are fabricated to include a “monolayer” of relatively short molecules that coat the surface of a solid support material and provide a flexible means to control the density of the single polynucleotides and optionally to prevent non-specific binding of reagents to the solid support.
The single polynucleotides immobilised onto the surface of a solid support should be capable of being resolved by optical means. This means that, within the resolvable area of the particular imaging device used, there must be one or more distinct signals, each representing one polynucleotide. Typically, the polynucleotides of the array are resolved using a single molecule fluorescence microscope equipped with a sensitive detector, e.g., a charge-coupled device (CCD). Each polynucleotide of the array may be imaged simultaneously or, by scanning the array, a fast sequential analysis can be performed.
The long polynucleotides of the array are typically DNA or RNA, although nucleic acid mimics, e.g., PNA or 2′-O-methyl-RNA, are within the scope of the invention. The long polynucleotides are formed on the array to allow interaction with other molecules. It is therefore important to immobilise the long polynucleotides so that the portion of the long polynucleotide not physically attached to solid support is capable of being interrogated. In some applications all the long polynucleotides in the single array will be the same, and may be used to capture molecules that are largely distinct. In other applications, the long polynucleotides on the array may all, or substantially all, be different, e.g., less than 50%, preferably less than 30% of the long polynucleotides will be the same.
The term “interrogate” is used herein to refer to any interaction of the arrayed long polynucleotide with any other molecule, e.g., with a polymerase or nucleoside triphosphate or a complementary nucleic acid sequence.
The density of the arrays is not critical. However, the present invention can make use of a high density of single long polynucleotides, and these are preferable. For example, arrays with a density of 106-109 long polynucleotides per cm2 may be used. Preferably, the density is at least 107/cm2 and typically up to 109/cm2. These high density arrays are in contrast to other arrays which may be described in the art as “high density” but which are not necessarily as high and/or which do not allow single molecule resolution.
The shorter molecules will typically be present on the array at much higher density than the relatively long polynucleotides, to coat the surface of the solid support not occupied by the relatively long polynucleotides. The shorter molecules may therefore be brought into contact with the solid support at an excess concentration. Preferably, the small molecules are at a density of from 108 to 1014 molecules/cm2, more preferably greater than 1012 molecules/cm2.
Using the methods and device of the present invention, it may be possible to image at least 106-108, preferably 107 or 108 long polynucleotides/cm2. Fast sequential imaging may be achieved using a scanning apparatus; shifting and transfer between images may allow higher numbers of polynucleotides to be imaged.
The extent of separation between the individual polynucleotides on the array will be determined, in part, by the particular technique used to resolve the individual polynucleotide.
Apparatus used to image molecular arrays are known to those skilled in the art. For example, a confocal scanning microscope may be used to scan the surface of the array with a laser to image directly a fluorophore incorporated on the individual polynucleotide by fluorescence. Alternatively, a sensitive 2-D detector, such as a charge-coupled device, can be used to provide a 2-D image representing the individual polynucleotides on the array. “Resolving” single polynucleotides on the array with a 2-D detector can be done if, at 100× magnification, adjacent polynucleotides are separated by a distance of approximately at least 250 nm, preferably at least 300 nm and more preferably at least 350 nm. It will be appreciated that these distances are dependent on magnification, and that other values can be determined accordingly, by one of ordinary skill in the art.
Other techniques such as scanning near-field optical microscopy (SNOM) are available which are capable of greater optical resolution, thereby permitting more dense arrays to be used. For example, using SNOM, adjacent polynucleotides may be separated by a distance of less than 100 nm, e.g., 10 nm. For a description of scanning near-field optical microscopy, see Moyer et al., Laser Focus World (1993) 29(10).
An additional technique that may be used is surface-specific total internal reflection fluorescence microscopy (TIRFM); see, for example, Vale et al., Nature (1996) 380:451-453). Using this technique, it is possible to achieve wide-field imaging (up to 100 μm×100 μm) with single molecule sensitivity. This may allow arrays of greater than 107 resolvable polynucleotides per cm2 to be used.
Additionally, the techniques of scanning tunnelling microscopy (Binnig et al., Helvetica Physica Acta (1982) 55:726-735) and atomic force microscopy (Hansma et al., Ann. Rev. Biophys. Biomol. Struct. (1994) 23:115-139) are suitable for imaging the arrays of the present invention. Other devices which do not rely on microscopy may also be used, provided that they are capable of imaging within discrete areas on a solid support.
The devices according to the invention comprise immobilised polynucleotides and other immobilised molecules. The other molecules are relatively short compared to the polynucleotides and are used to control the density of the polynucleotides. They may also prevent non-specific attachment of reagents, e.g., nucleoside triphosphates, with the solid support, thereby reducing background interference. In one embodiment, the shorter molecules are also polynucleotides. However, other molecules may be used, e.g., peptides, proteins, polymers or synthetic chemicals, as will be apparent to the skilled person and depending on the application to which the array will be used. The preferred molecules are organic molecules that contain groups that can react with the surface of a solid support.
Preparation of the devices may be carried out by first preparing a mixture of the relatively long polynucleotides and of the relatively short molecules. Usually, the concentration of the latter will be in excess of that of the long polynucleotides. By “in excess” is meant that the short molecules are at least 100-fold in excess of the long molecules. The mixture is then placed in contact with a suitably prepared solid support, to allow immobilisation to occur.
Single polynucleotides may be immobilised to the surface of a solid support by any known technique, provided that suitable conditions are used to ensure adequate separation. Density of the polynucleotide molecules may be controlled by dilution. The gaps between the polynucleotides can be filled in with short molecules (capping groups) that may be small organic molecules or may be polynucleotides of different composition. The formation of the array of individually resolvable “longer” polynucleotides permits interrogation of those polynucleotides that are different from the bulk of the molecules.
Immobilisation may be by specific covalent or non-covalent interactions. Covalent attachment is preferred. Immobilisation of a polynucleotide will be carried out at either the 5′ or 3′ position, so that the polynucleotide is attached to the solid support at one end only. However, the polynucleotide may be attached to the solid support at any position along its length, the attachment acting to tether the polynucleotide to the solid support; this is shown for the hairpin constructs, described below. The immobilised (relatively long) polynucleotide is then able to undergo interactions with other molecules or cognates at positions distant from the solid support. Immobilisation in this manner results in well separated long polynucleotides. The advantage of this is that it prevents interaction between neighbouring long polynucleotides on the array, which may hinder interrogation of the array.
Suitable methods for forming the devices with relatively short molecules and relatively long polynucleotides will be apparent to the skilled person, based on conventional chemistries. The aim is to produce a highly dense layer of the relatively short molecules, interspersed with the relatively large polynucleotides which are at a density that permits resolution of each single polynucleotide.
A first step in the fabrication of the arrays will usually be to functionalise the surface of the solid support, making it suitable for attachment of the molecules/polynucleotides. For example, silanes are known functional groups that have been used to attach molecules to a solid support material, usually a glass slide. The relatively short molecules and relatively long polynucleotides can then be brought into contact with the functionalised solid support, at suitable concentrations and in either separate or combined samples, to form the arrays.
In one preferred embodiment, the long polynucleotides and the short molecules each have the same reactive group that attaches to the solid support, or to an intermediary molecule.
In an alternative embodiment, the support surface may be treated with different functional groups, one of which is to react specifically with the relatively short molecules, and the other with the relatively long polynucleotides. Controlling the concentration of each functional group provides a convenient way to control the densities of the molecules/polynucleotides.
In a still further embodiment, the relatively short molecules are immobilised at high density onto the surface of the solid support. The molecules are capable of reacting with the polynucleotides (either directly or through an intermediate functional group) which can be brought into contact with the molecules at a suitable concentration to provide the required density. “Intermediate functional group” means any homo- or heterobifunctional crosslinking agent. The polynucleotides are therefore immobilised on top of the monolayer of molecules.
Those molecules that are not in contact with a polynucleotide may be reacted with a further molecule to block (or cap) the reactive site. This may be carried out before, during or after arraying the polynucleotides. The blocking (capping) group may itself be a relatively short polynucleotide.
In another embodiment, only a minor proportion of the short molecules that are arrayed at high density on the solid support comprise a group that reacts with the polynucleotides; the majority, e.g., 90% or greater, are non-reactive. For example, the short molecules can be mixed silanes, a minor proportion of which are reactive with a functional group on the polynucleotides, and the remaining silanes are unreactive and form the array of short molecules on the device. Therefore, controlling the concentration of the minor proportion of short molecules also controls the density of the polynucleotides.
In another embodiment, the short molecules may have been modified in solution prior to immobilisation on the array so that only a minor proportion contain a functional group that is capable of undergoing covalent attachment to a complementary functional group on the polynucleotides.
In a related embodiment, the relatively short molecules are polynucleotides, and appropriate concentrations of both relatively long and relatively short polynucleotides are reacted with a functional group and then arrayed on the solid support, or to an intermediate molecule bound to the solid support.
Suitable functional groups will be apparent to the skilled person. For example, suitable groups include: amines, acids, esters, activated acids, acid halides, alcohols, thiols, disulfides, olefins, dienes, halogenated electrophiles, thiophosphates and phosphorothioates. It is preferred if the group contains a silane.
The relatively small molecules may be any molecule that can provide a barrier against non-specific binding to the solid support.
Suitable small molecules may be selected based on the required properties of the surface and the existing functionality.
In a preferred embodiment, the molecules are silanes of type RnSiX(4−n) (where R is an inert moiety that is displayed on the surface of the solid support and X is a reactive leaving group of type Cl or O-alkyl). The silanes include tetraethoxysilane, triethoxymethylsilane, diethoxydimethylsilane or glycidoxypropyltriethoxy-silane, although many other suitable examples will be apparent to the skilled person.
In an embodiment of the invention, the short molecules act as surface blocks to prevent random polynucleotide association with the surface of the solid support. Molecules therefore require a group to react with the surface (which will preferably be the same functionality as used to attach the polynucleotide to the surface) and an inert group that will be defined by the properties required on the surface. In an embodiment, the surface is functionalised with an epoxide and the small molecule is glycine, although other compounds containing an amine group would suffice.
It is also preferred if the small molecule is hydrophilic and repels binding of anions. The molecule therefore may be acid, phosphate, sulfate, hydroxyl or polyol and may include polyethers such as PEG.
In one embodiment, the relatively short molecules are polynucleotides. These may be prepared using any suitable technique, including synthetic techniques known in the art. It may be preferable to use short polynucleotides that are immobilised to the solid support at one end and comprise, at the other end, a non-reactive group, e.g., a dideoxynucleotide incapable of incorporating further nucleotides. The short polynucleotide may also be a hairpin construct, provided that it does not interact with a polymerase.
In one embodiment of the present invention, each relatively long polynucleotide of the array comprises a hairpin loop structure, one end of which comprises a target polynucleotide, the other end comprising a relatively short polynucleotide capable of acting as a primer in a polymerase reaction. This ensures that the primer is able to perform its priming function during a polymerase-based sequencing procedure, and is not removed during any washing step in the procedure. The target polynucleotide is capable of being interrogated.
The term “hairpin loop structure” refers to a molecular stem and loop structure formed from the hybridisation of complementary polynucleotides that are covalently linked. The stem comprises the hybridised polynucleotides and the loop is the region that covalently links the two complementary polynucleotides. Anything from a 5 to 25 (or more) base pair double-stranded (duplex) region may be used to form the stem. In one embodiment, the structure may be formed from a single-stranded polynucleotide having complementary regions. The loop in this embodiment may be anything from 2 or more non-hybridised nucleotides. In a second embodiment, the structure is formed from two separate polynucleotides with complementary regions, the two polynucleotides being linked (and the loop being at least partially formed) by a linker moiety. The linker moiety forms a covalent attachment between the ends of the two polynucleotides. Linker moieties suitable for use in this embodiment will be apparent to the skilled person. For example, the linker moiety may be polyethylene glycol (PEG).
If the short molecules are polynucleotides in a hairpin construct, it is possible to ligate the relatively long polynucleotides to a minor proportion of the hairpins either prior to or after arraying the hairpins on the solid support.
The arrays have many applications in methods which rely on the detection of biological or chemical interactions with polynucleotides. For example, the arrays may be used to determine the properties or identities of cognate molecules. Typically, interaction of biological or chemical molecules with the arrays are carried out in solution.
In particular, the arrays may be used in conventional assays which rely on the detection of fluorescent labels to obtain information on the arrayed polynucleotides. The arrays are particularly suitable for use in multi-step assays where the loss of synchronisation in the steps was previously regarded as a limitation to the use of arrays. The arrays may be used in conventional techniques for obtaining genetic sequence information. Many of these techniques rely on the stepwise identification of suitably labelled nucleotides, referred to in U.S. Pat. No. 5,654,413 as “single base” sequencing methods.
In an embodiment of the invention, the sequence of a target polynucleotide is determined in a similar manner to that described in U.S. Pat. No. 5,654,413, by detecting the incorporation of nucleotides into the nascent strand through the detection of a fluorescent label attached to the incorporated nucleotide. The target polynucleotide is primed with a suitable primer (or prepared as a hairpin construct which will contain the primer as part of the hairpin), and the nascent chain is extended in a stepwise manner by the polymerase reaction. Each of the different nucleotides (A, T, G and C) incorporates a unique fluorophore at the 3′ position which acts as a blocking group to prevent uncontrolled polymerisation. The polymerase enzyme incorporates a nucleotide into the nascent chain complementary to the target, and the blocking group prevents further incorporation of nucleotides. The array surface is then cleared of unincorporated nucleotides and each incorporated nucleotide is “read” optically by a charge-coupled device using laser excitation and filters. The 3′-blocking group is then removed (deprotected), to expose the nascent chain for further nucleotide incorporation.
Because the array consists of distinct optically resolvable polynucleotides, each target polynucleotide will generate a series of distinct signals as the fluorescent events are detected. Details of the full sequence are then determined.
Other suitable sequencing procedures will be apparent to the skilled person. In particular, the sequencing method may rely on the degradation of the arrayed polynucleotides, the degradation products being characterised to determine the sequence.
An example of a suitable degradation technique is disclosed in WO-A-95/20053, whereby bases on a polynucleotide are removed sequentially, a predetermined number at a time, through the use of labelled adaptors specific for the bases, and a defined exonuclease cleavage.
A consequence of sequencing using non-destructive methods is that it is possible to form a spatially addressable array for further characterisation studies, and therefore non-destructive sequencing may be preferred. In this context, the term “spatially addressable” is used herein to describe how different molecules may be identified on the basis of their position on an array.
Once sequenced, the spatially addressed arrays may be used in a variety of procedures which require the characterisation of individual molecules from heterogeneous populations.
The following Examples illustrate the invention, with reference to the accompanying drawings.
Glass slides were cleaned with decon 90 for 12 hours at room temperature prior to use, rinsed with water, EtOH and dried. A solution of glycidoxypropyltrimethoxysilane (0.5 mL) and mercaptopropyltrimethoxysilane (0.0005 mL) in acidified 95% EtOH (50 mL) was mixed for 5 min. The clean, dried slides were added to this mixture and left for 1 hour at room temperature rinsed with EtOH, dried and cured for 1 hour at 100° C. Maleimide modified DNA was prepared from a solution of amino-DNA (5′-Cy3-CtgCTgAAgCgTCggCAggT-heg-aminodT-heg-ACCTgCCgACgCT; SEQ ID NO:8) (10 μM, 100 μL) and N-[g-Maleimidobutryloxy]succinimide ester (GMBS); (Pierce) (1 mM) in DMF/diisopropylethylamine (DIPEA)/water (89/1/10) for 1 hour at room temperature. The excess cross-linker was removed using a size exclusion cartridge (NAPS) and the eluted DNA freeze-dried in aliquots and freshly diluted prior to use. An aliquot of the maleimide-GMBS-DNA (100 nM) was placed on the thiol surface in 50 mM potassium phosphate/1 mM EDTA (pH 7.6) and left for 12 hours at room temperature prior to washing with the same buffer.
The slide was inverted so that the chamber coverslip contacted the objective lens of an inverted microscope (Nikon TE200) via an immersion oil interface. A 60° fused silica dispersion prism was optically coupled to the back of the slide through a thin film of glycerol. Laser light was directed at the prism such that at the glass/sample interface it subtended an angle of approximately 68° to the normal of the slide and subsequently underwent Total Internal Reflection (TIR). Fluorescence from the surface produced by excitation with the surface specific evanescent wave generated by TIR was collected by the objective lens of the microscope and imaged onto an intensified charged coupled device (ICCD) camera (Pentamax, Princeton Instruments).
Images were recorded using a combination of a 532 Nd:YAG laser with a 580DF30 emission filter (Omega optics), with an exposure of 500 ms and maximum camera gain and a laser power of 50 mW at the prism.
The presence of glycidoxypropyltrimethoxysilane gave improved results (
Slides were cleaned with decon 90 for 12 hours prior to use and rinsed with water, EtOH and dried. A solution of tetraethoxysilane (0.7 mL) and N-(3-triethoxysilylpropyl)bromoacetamide (0.0007 mL) in acidified 95% EtOH (35 mL) was mixed for 5 minutes. The clean, dried slides were added to this mixture and left for 1 hour at room temperature, rinsed with EtOH, dried and cured for 1 hour at 100° C. Phosphorothioate modified DNA (5′-TMR-TACCgTCgACgTCgACgCTggCgAgCgTgCTgCggTTsTsTsTsT ACCgCAgCACgCTCgCCAgCg; SEQ ID NO:9) where s=phosphorothioate (100 pM, 100 μL) in sodium acetate (30 mM, pH 4.5) was added to the surface and left for 1 hour at room temperature. The slide was washed with a buffer containing 50 mM Tris/1 mM EDTA.
Imaging was performed as described in Example 1 and a good dispersion of single molecules was seen (
Slides were cleaned with decon 90 for 12 hours prior to use and rinsed with water, EtOH and dried. A solution of glycidoxypropyltrimethox-ysilane (0.5 mL) in acidified 95% EtOH was prepared and the cleaned slides placed in the solution for 1 hour, rinsed with EtOH and dried. Amino modified DNA (5′-Cy3-CTgCTgAAgCgTCggCAggT-heg-aminodT-heg-ACCTgCCgACgCT; SEQ ID NO:8) (1 μM, 100 μL) was placed on the surface and left for 12 hours at room temperature. The slide was washed with a solution of 1 mM glycine at pH 9 for 1 hour and flushed with 50 mM potassium phosphate/1 mM EDTA (pH 7.6). A good dispersion of coupled single molecules was seen by TIR microscopy, as described in Example 1.
The slide was then exposed to a mixture containing Cy5-dUTP (20 μM) and T4 exo-polymerase (250 nM) and Tris (40 mM), NaCl (10 mM), MgCl2 (4 mM), DTT (2 mM), potassium phosphate (1 mM), BSA (0.2 mgs/ml) 100 μL) at room temperature for 10 minutes and then flushed with Tris/EDTA buffer.
Imaging was performed using a pumped dye laser at 630 nm with a 670DF40 emission filter at 40 mW laser power using the TIR setup as described. A lower level of non-specific triphosphate binding was seen in the case using glycine, than in a control not treated with glycine.
All patents, patent applications, and published references cited herein are hereby incorporated by reference in their entirety. While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.
Section C (from U.S. Ser. No. 10/153,240)
This invention relates to fabricated arrays of polynucleotides, and to their analytical applications. In particular, this invention relates to the use of fabricated polynucleotide arrays in methods for obtaining genetic sequence information.
Advances in the study of molecules have been led, in part, by improvement in technologies used to characterise the molecules or their biological reactions. In particular, the study of nucleic acids, DNA and RNA, has benefited from developing technologies used for sequence analysis and the study of hybridisation events.
An example of the technologies that have improved the study of nucleic acids, is the development of fabricated arrays of immobilised nucleic acids. These arrays typically consist of a high-density matrix of polynucleotides immobilised onto a solid support material. Fodor et al., Trends in Biotechnology (1994) 12:19-26, describes ways of assembling the nucleic acid arrays using a chemically sensitised glass surface protected by a mask, but exposed at defined areas to allow attachment of suitably modified nucleotides. Typically, these arrays can be described as “many molecule” arrays, as distinct regions are formed on the solid support comprising a high density of one specific type of polynucleotide.
An alternative approach is described by Schena et al., Science (1995) 270:467-470, where samples of DNA are positioned at predetermined sites on a glass microscope slide by robotic micropipetting techniques. The DNA is attached to the glass surface along its entire length by non-covalent electrostatic interactions. However, although hybridisation with complementary DNA sequences can occur, this approach may not permit the DNA to be freely available for interacting with other components such as polymerase enzymes, DNA-binding proteins etc.
Recently, the Human Genome Project generated a draft of the entire sequence of the human genome—all 3×109bases. The sequence information represents that of an average human. However, there is still considerable interest in identifying differences in the genetic sequence between different individuals. The most common form of genetic variation is single nucleotide polymorphisms (SNPs). On average one base in 1000 is a SNP, which means that there are 3 million SNPs for any individual. Some of the SNPs are in coding regions and produce proteins with different binding affinities or properties. Some are in regulatory regions and result in a different response to changes in levels of metabolites or messengers. SNPs are also found in non-coding regions, and these are also important as they may correlate with SNPs in coding or regulatory regions. The key problem is to develop a low cost way of determining one or more of the SNPs for an individual.
The nucleic acid arrays can be used to determine SNPs, and they have been used to study hybridisation events (Mirzabekov, Trends in Biotechnology (1994) 12:27-32). Many of these hybridisation events are detected using fluorescent labels attached to nucleotides, the labels being detected using a sensitive fluorescent detector, e.g. a charge-coupled detector (CCD). The major disadvantages of these methods are that it is not possible to sequence long stretches of DNA, and that repeat sequences can lead to ambiguity in the results. These problems are recognised in Automation Technologies for Genome Characterisation, Wiley-Interscience (1997), ed. T. J. Beugelsdijk, Chapter 10: 205-225.
In addition, the use of high-density arrays in a multi-step analysis procedure can lead to problems with phasing. Phasing problems result from a loss in the synchronisation of a reaction step occurring on different molecules of the array. If some of the arrayed molecules fail to undergo a step in the procedure, subsequent results obtained for these molecules will no longer be in step with results obtained for the other arrayed molecules. The proportion of molecules out of phase will increase through successive steps and consequently the results detected will become ambiguous. This problem is recognised in the sequencing procedure described in U.S. Pat. No. 5,302,509. This method is therefore not suitable for the determination of SNPs, where the precise identification of a particular sequence is required.
WO-A-96/27025 is a general disclosure of single molecule arrays. Although sequencing procedures are disclosed, there is little description of the applications to which the arrays can be applied. There is also only a general discussion on how to prepare the arrays.
The invention encompasses a method for determining a single nucleotide polymorphism present in a genome, comprising: (a) immobilizing polynucleotide molecules onto the surface of a solid support to form an array comprising polynucleotides located at addresses capable of interrogation, wherein each address of at least a subset of addresses on the array corresponds to a single polynucleotide molecule, and the array permits the subset of addresses to be individually resolved by optical microscopy, and wherein each such single polynucleotide molecule comprises a first portion that is immobilized by covalent bonding to the surface and a second portion that is capable of interrogation; (b) interrogating an address that corresponds to a single polynucleotide molecule to identify nucleotides of a sequence in the single polynucleotide molecule on the array; and (c) comparing the nucleotides identified in step (b) with a known consensus sequence, and thereby determining differences between the consensus sequence and the sequence of the single polynucleotide molecule.
In one embodiment, the polynucleotide molecules comprise fragments of a genome.
In another embodiment, the interrogating step comprises identifying nucleotides of a sequence in the second portion of the single polynucleotide molecule.
In another embodiment, step (b) comprises: (i) contacting the array with each of the nucleotides dATP, dTTP, dGTP and dCTP, under conditions that permit a nucleic acid polymerase reaction to proceed and thereby form sequences complementary to the polynucleotides immobilized on said array; (ii) determining the incorporation of a nucleotide in the complementary sequences formed in step (i); and (iii) optionally repeating the steps (i) and (ii).
In a preferred embodiment, each nucleotide contains a removable fluorescent label.
In another preferred embodiment, each nucleotide contains a removable blocking group that prevents further nucleotide incorporation, and the blocking group is removed after each step of determining nucleotide incorporation.
In another embodiment, step (i) is carried out by first contacting the array with three of the four nucleotides dATP, dTTP, dCTP and dGTP under conditions that permit a nucleic acid polymerase reaction to proceed and thereby form sequences complementary to those in the array, then removing unincorporated nucleotides from the array, and then contacting the array with the remaining nucleotide under conditions that permit a nucleic acid polymerase reaction to proceed and thereby form sequences complementary to those in the array, so that step (ii) proceeds only after incorporation of said remaining nucleotide.
In another embodiment, adjacent single polynucleotides of the array are separated by a distance of at least 10 nm.
In another embodiment, the adjacent single polynucleotides are separated by a distance of at least 100 nm.
In another embodiment, the adjacent single polynucleotides are separated by a distance of at least 250 nm.
In another embodiment, the array has a density of from 106 to 109 single polynucleotides per
In another embodiment, the array density is from 107 to 109 single polynucleotides per cm2.
In another embodiment the polynucleotides are immobilised to the solid support via the 5′ terminus, the 3′ terminus or via an internal nucleotide.
According to one aspect of the invention, a method for determining a single nucleotide polymorphism present in a genome comprises the steps of: (i) immobilising fragments of the genome onto the surface of a solid support to form an array of polynucleotide molecules capable of interrogation, wherein the array allows the molecules to be individually resolved by optical microscopy, and wherein each molecule is immobilised by covalent bonding to the surface, other than at that part of each molecule that can be interrogated; (ii) identifying nucleotides at selected positions in the genome; and (iii) comparing the results of step (ii) with a known consensus sequence, and identifying any differences between the consensus sequence and the genome.
The features or addresses of the arrays of the present invention comprise what are effectively single molecules. This has many important benefits for the study of the molecules and their interaction with other biological molecules. In particular, fluorescent labels can be used in interactions with the single polynucleotide molecules and can be detected using an optical microscope linked to a sensitive detector, resulting in a distinct signal for each polynucleotide.
The arrays permit a massively parallel approach to monitoring fluorescent or other events on the polynucleotides. Such massively parallel data acquisition makes the arrays extremely useful in the detection and characterisation of single nucleotide polymorphisms.
As used herein, the term “feature,” or the equivalent term “address,” refers to each nucleic acid molecule occupying a discrete physical location on an array; if a given sequence is represented at more than one such site, each site is classified as a feature. It is preferred that a subset of the features on an array according to the invention comprise a single polynucleotide molecule only. It is more preferred that substantially all of the features on an array according to the invention comprise a single polynucleotide molecule only. As used herein, “substantially all of the features” means at least 50%, and preferably at least 60%, 70%, 80%, 85%, 90%, 92%, 94%, 96%, 98%, 99% or more of the features.
As used herein, the term “array” refers to a population of nucleic acid molecules that is distributed over a solid support; preferably, these molecules differing in sequence are spaced at a distance from one another sufficient to permit the identification of discrete addresses or features of the array. The population can be a heterogeneous mixture of nucleic acid molecules.
“Solid support”, as used herein, refers to the material to which a nucleic acid sample is attached. Suitable solid supports are available commercially, and will be apparent to the skilled person. The supports can be manufactured from materials such as glass, ceramics, silica and silicon. Supports with a gold surface may also be used. The supports usually comprise a flat (planar) surface, or at least a structure in which the polynucleotides to be interrogated are in approximately the same plane. Alternatively, the solid support can be non-planar, e.g., a microbead. Any suitable size may be used. For example, the supports might be on the order of 1-10 cm in each direction.
As used herein, the term “interrogate” means contacting the arrayed polynucleotide molecule with any other molecule, wherein the physical interaction provides information regarding a characteristic of the arrayed polynucleotide. The contacting can involve covalent or non-covalent interactions with the other molecule. As used herein, “information regarding a characteristic” means information regarding the sequence of one or more nucleotides in the polynucleotide, the length of the polynucleotide, the base composition of the polynucleotide, the Tm of the polynucleotide, the presence of a specific binding site for a polypeptide or other molecule, the presence of an adduct or modified nucleotide, or the three-dimensional structure of the polynucleotide.
As used herein, the term “features capable of interrogation” or “addresses capable of interrogation” refers to array features or addresses in which the immobilized single polynucleotide comprises at least a portion that is accessible for a physical interaction with another molecule or molecules, wherein the interaction provides information regarding a characteristic of the arrayed polynucleotide. For example, when nucleic acid sequence information is the characteristic sought to be determined, features capable of interrogation include those features wherein at least a portion of the immobilized single polynucleotide molecule is physically accessible to and can serve as a functional substrate for a nucleic acid polymerase enzyme. By “functional substrate” is meant that the immobilized polynucleotide itself, or a primer annealed to it, can be extended by the template-dependent polymerase activity of such enzyme.
As used herein, the term “single polynucleotide molecule” refers to one molecule of a nucleic acid sequence. Thus, an array feature or address corresponding to a single polynucleotide molecule consists of one polynucleotide molecule immobilized at that location on a solid support. This is in contrast to the array features of the prior art, in which a given feature or address typically comprises a plurality of copies of a given nucleic acid molecule, often thousands of copies or more.
“Single polynucleotide molecules” according to the invention can be single- or double-stranded. In one embodiment, the single polynucleotide molecule is single stranded. In another embodiment, the single polynucleotide molecule to be interrogated is a single nucleic acid strand attached to the array by hybridization to a covalently immobilized oligonucleotide; in this embodiment, the molecule to be interrogated is still considered to be a “single polynucleotide molecule.” In another embodiment, single polynucleotide molecules on the array are single stranded, yet form a hairpin at the immobilized end.
As used herein, the term “individually resolved” is used to indicate that, when visualised, it is possible to distinguish one polynucleotide on the array from its neighbouring polynucleotides. Visualisation may be effected by the use of reporter labels, e.g. fluorophores, the signal of which is individually resolved. Visualisation can be accomplished through the use of optical microscopy methods known in the art.
The terms “arrayed polynucleotides” and “polynucleotide arrays” are used herein to define a plurality of single polynucleotides. The term is intended to include the attachment of other molecules to a solid surface, the molecules having a polynucleotide attached that can be further interrogated during the SNP analysis. For example, the arrays can comprise linker molecules immobilised on a solid surface, the linker molecules being conjugated or otherwise bound to a polynucleotide that can be interrogated, to determine the presence of a SNP.
As used herein, the term “portion that is immobilized by bonding to the surface” refers to the nucleotide or nucleotides of an immobilized single polynucleotide molecule that is or are either directly involved in linkage to the solid substrate, or, because of their proximity to the point of immobilization, are not physically accessible to be capable of interrogation (e.g., to serve as a template or substrate for the primer extension activity of a nucleic acid polymerase enzyme). Depending upon the means of immobilization (e.g., direct immobilization, immobilization through a linker, etc.), the portion of a polynucleotide that is immobilized by bonding to a surface can be as small as one nucleotide or as large as 100 nucleotides or more, as long as there remains at least a portion of the immobilized polynucleotide molecule that is capable of interrogation. It is preferred that polynucleotides be immobilized by either their 5′ end or their 3′ end, but polynucleotides can also be immobilized via an internal nucleotide.
As used herein, the term “portion that is capable of interrogation” refers to that portion of an immobilized single polynucleotide molecule that is physically accessible to a physical interaction with another molecule or molecules, the interaction of which provides information regarding a characteristic of the arrayed polynucleotide as defined herein. Generally, the “portion of an immobilized single polynucleotide molecule that is capable of interrogation” is that part which is not the “portion that is immobilized by bonding to the surface” as that term is defined herein.
As used herein, the term “blocking group” refers to a moiety attached to a nucleotide which, while not interfering substantially with template-dependent enzymatic incorporation of the nucleotide into a polynucleotide chain, abrogates the ability of the incorporated nucleotide to serve as a substrate for further nucleotide addition. A “removable blocking group” is a blocking group that can be removed by a specific treatment that results in the cleavage of the covalent bond between the nucleotide and the blocking group. Specific treatments can be, for example, a photochemical, chemical or enzymatic treatment that results in the cleavage of the covalent bond between the nucleotide and the fluorescent label. Removal of the blocking group will restore the ability of the incorporated, formerly blocked nucleotide to serve as a substrate for further enzymatic nucleotide additions.
As used herein, the term “removable fluorescent label” refers to a covalently linked fluorescent label on a nucleotide, which label can be removed by a specific treatment of the nucleotide or a polynucleotide comprising the nucleotide. Specific treatments can be, for example, a photochemical, chemical or enzymatic treatment that results in the cleavage of the covalent bond between the nucleotide and the fluorescent label. In those instances where the fluorescent label blocks further nucleotide incorporation, removal of the fluorescent label after incorporation of the labeled nucleotide restores the ability of the formerly labeled nucleotide to serve as a substrate for further enzymatic nucleotide additions.
As used herein, the phrase “conditions that permit a nucleic acid polymerase reaction to proceed and thereby form sequences complementary to the polynucleotides immobilized on the array” refers to those refers to those conditions of salt concentration (metallic and non-metallic salts), pH, temperature, and necessary cofactor concentration under which a given polymerase enzyme catalyzes the extension of an annealed primer. Conditions for the primer extension activity of a wide range of polymerase enzymes are known in the art. As one example, conditions permitting the extension of a nucleic acid primer by Klenow exo-polymerase include the following: 50 mM Tris. HCl, 1 mM EDTA, 5 mM MgCl2, 10 mM NaCl (pH 7.4), 2 μM dNTPs, 1 mM DTT, Klenow exo- (10 units in 100 μl final volume) at 37° C. A chain terminator can be included, depending upon the type of primer extension or sequencing being performed.
According to the present invention, the single polynucleotides immobilised onto the surface of a solid support should be capable of being resolved by optical means. This means that, within the resolvable area of the particular imaging device used, there must be one or more distinct signals, each representing one polynucleotide. Typically, the polynucleotides of the array are resolved using a single molecule fluorescence microscope equipped with a sensitive detector, e.g. a charge-coupled device (CCD). Each polynucleotide of the array can be analysed simultaneously or, by scanning the array, a fast sequential analysis can be performed.
The polynucleotides of the array are preferably derived from fragments of genomic DNA.
The density of the array is not critical. However, the present invention can make use of a high density of single molecules (polynucleotides), and these are preferable. For example, arrays with a density of 106 to 109 single polynucleotides per cm2 can be used. Preferably, the density is at least 107/cm2 to 109/cm2. These high density arrays are in contrast to other arrays which may be described in the art as “high density” but which are not necessarily as high and/or which do not allow single molecule resolution. On a given array, it is the number of single polynucleotides, rather than the number of features, that is important. The concentration of nucleic acid molecules applied to the support can be adjusted in order to achieve the highest density of addressable single polynucleotide molecules. At lower application concentrations, the resulting array will have a high proportion of addressable single polynucleotide molecules at a relatively low density per unit area. As the concentration of nucleic acid molecules is increased, the density of addressable single polynucleotide molecules will increase, but the proportion of single polynucleotide molecules capable of being addressed will actually decrease. One skilled in the art will therefore recognize that the highest density of addressable single polynucleotide molecules can be achieved on an array with a lower proportion or percentage of single polynucleotide molecules relative to an array with a high proportion of single polynucleotide molecules but a lower physical density of those molecules.
Using the methods and apparatus of the present invention, it can be possible to image at least 107 or 108 polynucleotides. Fast sequential imaging can be achieved using a scanning apparatus; shifting and transfer between images can allow higher numbers of molecules to be imaged.
The extent of separation between the individual polynucleotides on the array will be determined, in part, by the particular technique used to resolve the individual polynucleotide. Apparatus used to image molecular arrays are known to those skilled in the art. For example, a confocal scanning microscope can be used to scan the surface of the array with a laser to image directly a fluorophore incorporated on the individual molecule by fluorescence. Alternatively, a sensitive 2-D detector, such as a charge-coupled device, can be used to provide a 2-D image representing the individual polynucleotides on the array.
Resolving single polynucleotides on the array with a 2-D detector can be done if, at 100× magnification, adjacent polynucleotides are separated by a distance of approximately at least 250 nm, preferably at least 300 nm and more preferably at least 350 nm. It will be appreciated that these distances are dependent on magnification, and that other values can be determined accordingly, by one of ordinary skill in the art.
Other techniques such as scanning near-field optical microscopy (SNOM) are available which are capable of greater optical resolution, thereby permitting more dense arrays to be used. For example, using SNOM, adjacent polynucleotides can be separated by a distance of less than 100 nm, e.g. 10 nm. For a description of scanning near-field optical microscopy, see Moyer et al., Laser Focus World (1993) 29(10).
An additional technique that can be used is surface-specific total internal reflection fluorescence microscopy (TIRFM); see, for example, Vale et al., Nature, (1996) 380: 451-453). Using this technique, it is possible to achieve wide-field imaging (up to 100 μm×100 μm) with single molecule sensitivity. This can allow arrays of greater than 107 resolvable polynucleotides per cm2 to be used.
Additionally, the techniques of scanning tunnelling microscopy (Binnig et al., Helvetica Physica Acta (1982) 55:726-735) and atomic force microscopy (Hansma et al., Ann. Rev. Biophys. Biomol. Struct. (1994) 23:115-139) are suitable for imaging the arrays of the present invention. Other devices which do not rely on microscopy can also be used, provided that they are capable of imaging within discrete areas on a solid support.
Single polynucleotides can be arrayed by immobilisation to the surface of a solid support. This can be carried out by any known technique, provided that suitable conditions are used to ensure adequate separation. Generally the array is produced by dispensing small volumes of a sample containing a mixture of the fragmented genomic DNA onto a suitably prepared solid surface, or by applying a dilute solution to the solid surface to generate a random array. The formation of the array then permits interrogation of each arrayed polynucleotide to be carried out.
Suitable solid supports are available commercially, and will be apparent to the skilled person. The supports can be manufactured from materials such as glass, ceramics, silica and silicon. The supports usually comprise a flat (planar) surface, or an array in which the polynucleotides to be interrogated are in the same plane. However, “solid supports” as the term is used herein can also encompass non-planar supports, for example, a microbead. Any suitable size can be used. For example, the supports might be of the order of 1-10 cm in each direction.
Immobilisation can be by specific covalent or non-covalent interactions. Covalent attachment is preferred. Immobilisation can be at an internal position or at either the 5′ or 3′ position. However, the polynucleotide can be attached to the solid support at any position along its length, the attachment acting to tether the polynucleotide to the solid support. The immobilised polynucleotide is then able to undergo interactions at positions distant from the solid support. Typically the interaction will be such that it is possible to remove any molecules bound to the solid support through non-specific interactions, e.g. by washing. Immobilisation in this manner results in well separated single polynucleotides.
In one embodiment, the array comprises polynucleotides with a hairpin loop structure, one end of which comprises the target polynucleotide derived from the genomic DNA sample.
The term “hairpin loop structure” refers to a molecular stem and loop structure formed from the hybridisation of complementary polynucleotides that are covalently linked. The stem comprises the hybridised polynucleotides and the loop is the region that covalently links the two complementary polynucleotides. Anything from a 5 to 25 (or more) base pair double-stranded (duplex) region can be used to form the stem. In one embodiment, the structure can be formed from a single-stranded polynucleotide having complementary regions. The loop in this embodiment can be anything from 2 or more non-hybridised nucleotides. In a second embodiment, the structure is formed from two separate polynucleotides with complementary regions, the two polynucleotides being linked (and the loop being at least partially formed) by a linker moiety. The linker moiety forms a covalent attachment between the ends of the two polynucleotides. Linker moieties suitable for use in this embodiment will be apparent to the skilled person. For example, the linker moiety can be polyethylene glycol (PEG).
There are many different ways of forming the hairpin structure to incorporate the target polynucleotide. However, a preferred method is to form a first molecule capable of forming a hairpin structure, and ligate the target polynucleotide to this. Ligation can be carried out either prior to or after immobilisation to the solid support. The resulting structure comprises the target polynucleotide at one end of the hairpin and a primer polynucleotide at the other end. The target polynucleotide can be either single stranded or double stranded as long as the 3′-end of the hairpin contains a free hydroxyl amenable to further polymerase extension.
The DNA to be analyzed can be PCR-amplified or used directly to generate fragments of DNA using either restriction endonucleases, other suitable enzymes, a mechanical form of fragmentation or a non-enzymatic chemical fragmentation method or a combination thereof. The DNA can be genomic DNA. The fragments can be of any suitable length, preferably from 20 to 2000 bases, more preferably 20 to 1000 bases, most preferably 20 to 200 bases. In the case of fragments generated by restriction endonucleases, hairpin structures bearing a complementary restriction site at the end of the first hairpin can be used. In the case of non-selective fragmentation, ligation of one strand of the DNA sample fragments can be achieved by various methods.
Method 1: The fragments are ligated to a hairpin made, for example, with a 3′ overhang containing all possible sequences of a few nucleotides (preferably 3-20 bases long, more preferably 5-9 bases long), a 3′ hydroxyl and a 5′ phosphate. Ligation creates a 5′ overhang that is capable of being sequenced from the 3′ hydroxyl of the hairpin using the newly ligated genomic fragment as a template by the methods described.
Method 2: in the design of the hairpin, a single (or more) base gap can be incorporated at the 3′ end (the receded strand) such that upon ligation of the DNA fragment only one strand is covalently joined to the hairpin. The base gap can be formed by hybridising a further separate polynucleotide to the 5′-end of the first hairpin structure. On ligation, the DNA fragment has one strand joined to the 5′-end of the first hairpin, and the other strand joined to the 3′-end of the further polynucleotide. The further polynucleotide (and the other strand of the DNA fragment) can then be removed by disrupting hybridisation.
Method 3: Genomic fragments are left in their double stranded-form or are made to be double stranded and blunt ended by conventional means and are phosphatased to produce 3′ and 5′ hydroxyls as is known in the art. The fragments are ligated to a hairpin made for example with a blunt end, a 3′ hydroxy and a 5′ phosphate. Ligation of only one strand creates a 5′ overhang that is capable of being sequenced from the 3′ hydroxyl of the hairpin using the newly ligated genomic fragment as a template by the methods described.
The net result should be covalent ligation of only one strand of a DNA fragment of genomic DNA, to the hairpin, the DNA fragment being then in the form of a 5′ overhang that is capable of being sequenced. Such ligation reactions can be carried out in solution at optimised concentrations based on conventional ligation chemistry, for example, carried out by DNA ligases or non-enzymatic chemical ligation. Should the fragmented DNA be generated by random shearing of genomic DNA, then the ends can be filled in with any polymerase to generate blunt-ended fragments which can be blunt-end-ligated onto blunt-ended hairpins. Alternatively, the blunt-ended DNA fragments can be ligated to oligonucleotide adapters which are designed to allow compatible ligation with the sticky-end hairpins, in the manner described previously.
The hairpin-ligated DNA constructs can then be covalently attached to the surface of a solid support to generate the single molecule array, or ligation can follow attachment to form the array.
The arrays can then be used in procedures to determine the presence of a SNP. In the case of random fragmentation of the DNA sample, cycles of sequencing can be performed to place the fragment in a unique context within the sample from which it originated. If the target fragments are generated via restriction digest of genomic DNA, the recognition sequence of the restriction or other nuclease enzyme will provide 4, 6, 8 bases or more of known sequence (dependent on the enzyme). Further sequencing of at least 4 bases and preferably between 10 and 30 bases on the array should provide sufficient overall sequence information to place that stretch of DNA into unique context with a total human genome sequence, thus enabling the sequence information to be used for genotyping and more specifically single nucleotide polymorphism (SNP) scoring.
Simple calculations have suggested the following based on sequencing a 107 molecule array prepared from hairpin ligation: for a 6 base pair recognition sequence, a single restriction enzyme will generate approximately 106 ends of DNA. If a stretch of 13 bases is sequenced on the array (i.e. 13×106 bases), approximately 13,000 SNPs will be detected. The approach is therefore suitable for forensic analysis or any other system which requires unambiguous identification of individuals to a level as low as 103 SNPs.
It is of course possible to sequence the complete target polynucleotide, if required.
Sequencing can be carried out by the stepwise identification of suitably labelled nucleotides, referred to in U.S. Pat. No. 5,654,413 as “single base” sequencing methods. The target polynucleotide is primed with a suitable primer (or prepared as a hairpin construct which will contain the primer as part of the hairpin), and the nascent chain is extended in a stepwise manner by the polymerase reaction. Each of the different nucleotides (A, T, G and C) incorporates a unique fluorophore which can be located at the 3′ position to act as a blocking group to prevent uncontrolled polymerisation. The polymerase enzyme incorporates a nucleotide into the nascent chain complementary to the target, and the blocking group prevents further incorporation of nucleotides. The array surface is then cleared of unincorporated nucleotides and each incorporated nucleotide is “read” optically by a charge-coupled detector using laser excitation and filters. The 3′-blocking group is then removed (deprotected), to expose the nascent chain for further nucleotide incorporation.
Because the array consists of distinct optically resolvable polynucleotides, each target polynucleotide will generate a series of distinct signals as the fluorescent events are detected. Details of the sequence are then determined and can be compared with known sequence information to identify SNPs.
The number of cycles that can be achieved is governed principally by the yield of the deprotection cycle. If deprotection fails in one cycle, it is possible that later deprotection and continued incorporation of nucleotides can be detected during the next cycle. Because the sequencing is performed at the single molecule level, the sequencing can be carried out on different polynucleotide sequences at one time without the necessity for separation of the different sample fragments prior to sequencing. This sequencing also avoids the phasing problems associated with prior art methods.
The labelled nucleotides can comprise a separate label and removable blocking group, as will be appreciated by those skilled in the art. In this context, it will usually be necessary to remove both the blocking group and the label prior to further incorporation.
Deprotection can be carried out by chemical, photochemical or enzymatic reactions. A similar, and equally applicable, sequencing method is disclosed in EP-A-0640146. Other suitable sequencing procedures will be apparent to the skilled person.
It is not necessary to determine the sequence of the full polynucleotide fragment. For example, it can be preferable to determine the sequence of 16-30 specific bases, which is sufficient to identify the DNA fragment by comparison to a consensus sequence, e.g. to that known from the Human Genome Project. Any SNP occurring within the sequenced region can then be identified. The specific bases do not have to be contiguous. For example, the procedure can be carried out by the incorporation of non-labelled bases followed, at pre-determined positions, by the incorporation of a labelled base. Provided that the sequence of sufficient bases is determined, it should be possible to identify the fragment. Again, any SNPs occurring at the determined base positions, can be identified. For example, the method can be used to identify SNPs that occur after cytosine. Template DNA (genomic fragments) can be contacted with each of the bases A, T and G, added sequentially or together, so that the complementary strand is extended up to a position that requires C. Non-incorporated bases can then be removed from the array, followed by the addition of C. The addition of C is followed by monitoring the next base incorporation (using a labelled base). By repeating this process a sufficient number of times, a partial sequence is generated where each base immediately following a C is known. It will then be possible to identify the full sequence, by comparison of the partial sequence to a reference sequence. It will then also be possible to determine whether there are any SNPs occurring after any C.
To further illustrate this, a device can comprise 107 restriction fragments per cm2. If 30 bases are determined for each fragment, this means 3×108 bases are identified. Statistically, this should determine 3×105 SNPs for the experiment. The approach therefore permits analysis of large amounts of sequence for SNPs.
The images and other information about the arrays, e.g. positional information, etc. are processed by a computer program which can perform image processing to reduce noise and increase signal or contrast, as is known in the art. The computer program can perform an optional alignment between images and/or cycles, extract the single molecule data from the images, correlate the data between images and cycles and specify the DNA sequence from the patterns of signal produced from the individual molecules.
The individual DNA sequence reads of at least 4 bases, and more preferably at least 16 bases in the case of human genomic DNA, and more preferably 16-30 bases, are aligned and compared with a genomic sequence. The methods for performing this alignment are based upon techniques known to those skilled in the art. The individual DNA sequence reads are aligned with respect to the reference sequence by finding the best match between the individual DNA sequence reads and the reference sequence. Using the known alignments, one or many individual DNA sequence reads covering a given region of the genomic DNA sequence are obtained. All the aligned individual DNA sequence reads are interpreted at each nucleotide position in the reference sequence as either containing the identical sequence to the reference sequence, or containing an error in some of the individual DNA sequence reads, or containing a known or novel mutation, SNP, deletion, insertion, etc. at that position. Furthermore, for most chromosomes, at each position in the reference sequence, the individual can contain one (homozygous) or two (heterozygous) different nucleotides corresponding to the two copies of each chromosome. The sum total of all the individual variations in the reference sequence corresponding to a given individual sample is collectively referred to as a “total genotype”.
The following Example illustrates the invention.
Preparation of hairpin single molecule array (unlabelled DNA): A 10 μM solution of oligonucleotide (5′-TCgACTgCTgAAAAgCgTCggCTggT-HEG-amin-odT-HEG-ACCAgCCgACGCTTT; SEQ ID NO. 8) in DMF containing 10% water and 1% diisopropylethylamine (DIPEA) was prepared. To this, a stock solution of the GMBS crosslinker was added to give a final concentration of 1 mM N-[γ-Maleimidobutryloxy]succinimide ester (GMBS) (100 eqvs.). The reaction was left for 1 h at room temperature, purified using a NAP size exclusion column and freeze-dried in aliquots that were re-dissolved immediately prior to use.
A fused silica slide was treated with decon for 12 h then rinsed with water, EtOH, dried and placed in a flow cell. A solution of the GMBS DNA (150 nM) and mercaptopropyltrimethoxysilane (3 μM) in 9:1 sodium acetate (30 mM, pH 4.3): isopropanol was placed over the slide for 30 min. at 65° C. The cell was flushed first with 50 mM Tris. HCl, 1 mM EDTA, pH 7.4 and then 50 mM Tris.HCl, 1 mM EDTA, 5 mM MgCl2, 10 mM NaCl (pH 7.4) (10 mL) at 37° C. (TKF buffer). The cell was filled with 100 μL of 2 μM Cy5-dCTP, 2 μM dTTP, 2 μm dATP, 1 mM DTT, Klenow exo- (10 units) in TKF buffer and incubated at 37° C. for 10 mins. than flushed with TKF buffer (20 mL) and TKF buffer containing NaCl (1 M) which removes bound protein. A second cycle consisting of 100 μL of 2 μM Cy3-dCTP, 2 μM dGTP, 2μM dATP, 1 mM DTT, Klenow exo-(10 units) in TKF buffer was incubated at 37° C. for 10 mins. then flushed with TKF buffer (20 mL) and TKF buffer containing NaCl (1 M).
The flowcell was inverted so that the chamber coverslip contacts the objective lens of an inverted microscope (Nikon TE200) via an immersion oil interface. A 60° fused silica dispersion prism was optically coupled to the back of the slide through a thin film of glycerol. Laser light was directed at the prism such that at the glass/sample interface subtended an angle of approximately 68° to the normal of the slide and subsequently underwent Total Internal Reflection (TIR). Fluorescence from the surface produced by excitation with the surface specific evanescent wave generated by TIR was collected by the 100× objective lens of the microscope and imaged onto an intensified charged coupled device (ICCD) camera (Pentamax, Princeton Instruments).
Images were recorded using a combination of a 532 Nd:YAG laser with a 580DF30 emission filter (Omega optics) and a pumped dye laser at 630 nm with a 670DF40 emission filter. Images were recorded with an exposure of 500 ms and maximum camera gain and a laser power of 50 mW (green) and 40 mW (red) at the prism.
Two colour fluorophore labelled nucleotide incorporations were identified by the co-localisation of discreet points of fluorescence from single molecules of Cy3 and Cy5 following superimposing the two images. Molecules were considered co-localised when fluorescent points were within a pixel separation of each other. For a 90 μm and 90 pm field projected onto a CCD array of 512×512 pixels the pixel size dimension is 176 nm.
An average 46.2% of Cy3 and 57.5% of Cy5 were colocalized; showing >50% of the molecules that underwent the Cy5 incorporation underwent a second cycle of Cy3 incorporation. In the absence of enzyme in the second cycle the level of Cy3 was greatly reduced and the colocalisation was <2%. Polymerase fidelity controls, whereby the dATP or dGTP was omitted from the cycles, gave colocalisation levels of approximately 4%.
This demonstrates that sequence determination at the single molecule level can be achieved and makes it possible to extend this to genomic fragments to identify SNPs.
Those skilled in the art should appreciate that they can readily use the disclosed conception and specific embodiments as a basis for designing or modifying other methods for carrying out the same purposes of the present invention without departing from the spirit and scope of the invention as defined by the appended claims. All literature and patent references referred to herein are hereby incorporated by reference in their entirety.
| Number | Date | Country | Kind |
|---|---|---|---|
| 98306094.8 | Jul 1998 | EP | regional |
| 9822670.7 | Oct 1998 | GB | national |
| 0002310.1 | Feb 2000 | GB | national |
The present application is a Continuation-In-Part of co-pending U.S. Ser. No. 10/153,267, filed May 22, 2002, which is a Continuation-In-Part of PCT/GB02/00438, filed Jan. 30, 2002, and a Continuation-In-Part of U.S. application Ser. No. 09/771,708, filed Jan. 30, 2001 [now U.S. Pat. No. (USPN) 6,787,308], which in turn, is a Continuation-In-Part of United Kingdom App. No. 0002310.1, filed Feb. 1, 2000, and a Continuation-In-Part of PCT/GB99/02487, filed Jul. 30, 1999, which in turn claims benefit of United Kingdom App. No. 9822670.7, filed Oct. 16, 1998, and European App. No. 98306094.8, filed Jul. 30, 1998. The present application is also a Continuation-In-Part of co-pending U.S. Ser. No. 10/153,240, filed May 22, 2002, which is a Continuation-In-Part of PCT/GB02/00439, filed Jan. 30, 2002, and a Continuation-In-Part of U.S. application Ser. No. 09/771,708, filed Jan. 30, 2001 (now U.S. Pat. No. 6,787,308), which in turn, is a Continuation-In-Part of United Kingdom App. No. 0002310.1, filed Feb. 1, 2000, and a Continuation-In-Part of PCT/GB99/02487, filed Jul. 30, 1999, which in turn claims benefit of United Kingdom App. No. 9822670.7, filed Oct. 16, 1998, and European App. No. 98306094.8, filed Jul. 30, 1998. The present application is also a Continuation-In-Part of co-pending U.S. Ser. No. 10/864,887, filed Jun. 9, 2004, which is a Continuation of U.S. application Ser. No. 09/771,708, filed Jan. 30, 2001 (now U.S. Pat. No. 6,787,308), which in turn, is a Continuation-In-Part of United Kingdom App. No. 0002310.1, filed Feb. 1, 2000 and a Continuation-In-Part of PCT/GB99/02487, filed Jul. 30, 1999, which in turn claims benefit of United Kingdom App. No. 9822670.7, filed Oct. 16, 1998, and European App. No. 98306094.8, filed Jul. 30, 1998. The entire teachings of the above-identified applications are incorporated by reference. The specifications and Figures of U.S. Ser. Nos. 10/864,887; 10/153,240 and 10/153,267 are reproduced herein. Section A (from U.S. Ser. No. 10/864,887)
| Number | Date | Country | |
|---|---|---|---|
| Parent | 09771708 | Jan 2001 | US |
| Child | 10864887 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10153267 | May 2002 | US |
| Child | 12583658 | US | |
| Parent | PCT/GB02/00438 | Jan 2002 | US |
| Child | 10153267 | US | |
| Parent | 09771708 | Jan 2001 | US |
| Child | PCT/GB02/00438 | US | |
| Parent | PCT/GB99/02487 | Jul 1999 | US |
| Child | 09771708 | US | |
| Parent | 10153240 | May 2002 | US |
| Child | PCT/GB99/02487 | US | |
| Parent | PCT/GB02/00439 | Jan 2002 | US |
| Child | 10153240 | US | |
| Parent | 09771708 | Jan 2001 | US |
| Child | PCT/GB02/00439 | US | |
| Parent | 10864887 | Jun 2004 | US |
| Child | 09771708 | US | |
| Parent | PCT/GB99/02487 | Jul 1999 | US |
| Child | 09771708 | US |
