WO2011102802A1 - Method for reducing mismatches in double-stranded dna molecules - Google Patents

Abstract

There is presently provided a method of reducing DNA mismatches in a pool of double-stranded DNA molecules, the pool comprising 5'-protected double-stranded DNA molecules, by treating the pool with a mismatch endonuclease and a DNA ligase, followed by treating the pool with one or more exonucleases, the one or more exonucleases together having 5' to 3' activity on double-stranded DNA and 3' to 5' activity on single-stranded DNA.

Description

METHOD FOR REDUCING MISMATCHES IN DOUBLE- STRANDED DNA MOLECULES

FIELD OF THE INVENTION

[0001] The present invention relates to methods for improving the fidelity of double stranded DNA by reducing mismatches in a pool of double-stranded DNA molecules.

BACKGROUND OF THE INVENTION

[0002] In many current molecular techniques that involve the use of DNA, it is necessary to generate a supply of DNA molecules by synthetic methods. Common methods of DNA synthesis include polymerase chain reaction methods and ligation chain reaction methods. Usually, ensuring that the DNA contains the correct base sequence is important, if not essential, for the success of the molecular technique in which the synthesized DNA is to be used. For example, generation of a DNA coding sequence for use in gene expression of functioning proteins requires a precise DNA sequence; even one base substitution, insertion or deletion can have significant consequences for the protein that is ultimately produced. Thus, the process of removal of DNA molecules having incorrect DNA sequences from a synthetic DNA pool is essential in providing error-free synthetic DNA produced by a de novo gene synthesis method.

[0003] To date, there have been various methods described for removing errors in DNA sequence from a pool of synthetic DNA molecules. Such methods include: use of MutS mismatch binding enzyme to bind mismatched DNA [1, 2]; use of a second DNA microarray chip to purify synthetic oligonucleotides [3]; and use of

endonuclease enzymes, such as T4 Endonuclease VII (T4E7), Endonuclease V (Endo V), and T7 Endonuclease I (T7EI), to digest DNA molecules containing a mismatch into smaller segments [4,5].

[0004] In particular, solid-phase error filtering using immobilized Thermus aquaticus MutS [1], and gel electrophoretic separation of MutS/DNA complex [2] have been reported for removal of incorrect DNA sequences with error rate of -0.3 errors per kb, and ~0.1 errors per kb respectively, demonstrated using de novo synthesized GFPwv coding sequence. Although the MutS enzyme can significantly reduce the error rate in synthetic DNA, a MutS. -DNA ratio > 10 is typically required, since any given DNA molecule could contain more than one mismatch and since MutS non-specifically binds to DNA that contains no errors [6, 7]. The MutS dynamically screens the DNA sequence, displaying higher binding affinity for mismatched DNA over perfect match DNA [8].

[0005] Techniques using a second DNA microchip were reported to effectively reduce the error rate within a pool of DNA molecules from 6.3 errors/kb to 0.72 errors/kb [3]. These methods utilize the hybridization efficiency difference of match and mismatch DNA hybridization to purify synthesized oligonucleotides. However, the requirement for a second DNA microarray increases both cost and process time.

[0006] Enzymatic techniques involving mismatch cleavage by endonucleases, including T4E7, Endo V and T7EI, have also been used to reduce error rates within a pool of DNA molecules, with subsequent re-assembly of the cleaved DNA segments [4]. An endonuclease is used that cleaves the double-stranded DNA, releasing DNA segments containing the mismatch at a 3 '-overhang. The 3 '-overhangs are then removed using a single-strand specific 3 ' exonuclease. These "repaired" fragments are then re-assembled to full-length DNA in a subsequent PGR method. With this approach, error frequency was reduced from 5.8 errors per kbase to 1.62 and to 1.98 errors per kbp for treatment with T4E7 and E. coli endonuclease V, respectively. However, the endonucleases used possess strong non-specific nicking and 5'- exonuclease activity [9-11]. Another method utilizing Endo V for LCR-based gene, synthesis has also been reported having a reduced error rate of from 1.8 to 0.4 errors/kb [5].

[0007] There exists a need for alternative methods for reducing mismatches within a pool of DNA molecules.

SUMMARY OF THE INVENTION

[0008] The present invention provides methods of reducing DNA mismatches in a pool of double-stranded DNA molecules. The methods involve protection of the 5' end of the full-length double-stranded DNA molecules. An endonuclease that recognizes and cleaves mismatches from DNA is used to remove errors hi the double- stranded DNA, with concomitant or subsequent use of a ligase to repair any nicks that may occur due to non-specific activity of the endonuclease. This endonuclease/ligase treatment is combined with exonuclease treatment to remove the products of the endonuclease digestion from the pool of DNA. Protection of the 5' end of the full- length double-stranded DNA means that 5'-protected DNA that is not digested by the endonuclease remains intact, surviving both the endonuclease and exonuclease treatments. Thus, the resulting pool of double-stranded DNA molecules following performance of the present methods should have a reduced mismatch frequency, and thus have improved sequence fidelity for use in subsequent molecular techniques.

[0009] The protection of the 5 ' end of the double-stranded DNA molecules plays two roles in the present method: protecting the 5' end of full-length DNA molecules from 5 'exonuclease activity possessed by a mismatch endonuclease, and to selectively protect full-length DNA that has survived the endonuclease treatment from

subsequent digestion during exonuclease treatment.

[0010] Thus, the present methods may involve the following features: enzymatic combinations designed to reduce the effect of non-specific DNA nicking activity of mismatch endonucleases; 5' protection that selectively protects full-length DNA from exonuclease activity of mismatch endonucleases; and selective protection of full- length DNA during exonuclease digestion of fragments generated by endonuclease cleavage. The methods may include heat inactivation and/or buffer exchange between enzymatic steps, if desired, which may result in increased yield of full-length DNA.

[0011 ] De novo synthesis of double-stranded DNA molecules by reaction-based methods (as opposed to cellular replication methods) produces DNA having a relatively high error rate, typically about 1-11 errors per kilobase-pair of DNA. This high error rate can result in introduction of errors into molecular applications of DNA synthesized by reaction-based techniques. Reducing the error rate using the present methods allows for greater fidelity in subsequent molecular applications, including protein expression, protein engineering, artificial gene networks, synthetic genome construction, DNA vaccines, drug development, DNA and protein microarrays, and polymorphism (or other mutation) detection and screening.

[0012] When compared to other error filter methods for reducing errors in synthesized double-stranded DNA [1-5], the methods of the present invention may reduce or eliminate steps of isolating and extracting full-length DNA, which steps can be tedious and time-consuming. Instead, the present methods remove undesired digestion fragments from the target full-length DNA using exonucleases, which may render the present methods amenable for automation.

[0013] Thus, in one aspect the present invention provides a method comprising, in sequence, treating a pool of double-stranded DNA molecules with a mismatch endonuclease and a DNA ligase, the pool of double-stranded DNA molecules comprising 5 '-protected double-stranded DNA molecules; and treating the pool of double-stranded DNA molecules with one or more exonucleases, the one or more exonucleases together having 5' to 3' activity on double-stranded DNA and 3' to 5' activity on single-stranded DNA.

[0014] The 5'-protected double-stranded DNA molecules-may have a protecting group attached at each 5' end. The protecting group may be a biotin group, a locked nucleic acid group or a peptide nucleic acid group.

[0015] Alternatively, the 5'-protected double-stranded DNA molecules may be circular.

[0016] The mismatch endonuclease may be endonuclease V, T7 endonuclease I, Surveyor nuclease, T4 endonuclease VII, or CEL I nuclease. In one embodiment, the mismatch endonuclease is T7 endonuclease I. In another embodiment, the mismatch endonuclease is Surveyor nuclease.

[0017] The DNA ligase may be Ampligase, 9°N ligase, T4 DNA ligase, Taq DNA ligase, or E. coli DNA ligase. In one embodiment, the DNA ligase is Ampligase. In another embodiment, the DNA ligase is T4 DNA ligase.

[0018] The one or more exonucleases may comprise one of Lambda exonuclease, RecJ and T7 exonuclease and one of exonuclease I and exonuclease T. In one embodiment, the one or more exonucleases comprise Lambda exonuclease and exonuclease I.

[0019] The pool of double-stranded DNA molecules may be treated with the mismatch endonuclease prior to treatment with the DNA ligase, or may be treated with the mismatch endonuclease together with the DNA ligase.

[0020] In one embodiment, the mismatch endonuclease is T7 endonuclease I, the DNA ligase is Ampligase and the one or more exonucleases comprise Lambda exonuclease and exonuclease I, and the pool of double-stranded DNA molecules is treated with the mismatch endonuclease together with the DNA ligase.

[0021] In a different embodiment, the mismatch endonuclease is Surveyor, the DNA ligase is T4 DNA ligase and the one or more exonucleases comprise Lambda exonuclease and exonuclease I, and the the pool of double-stranded DNA molecules is treated with the mismatch endonuclease prior to treatment with the DNA ligase.

[0022] In another aspect, the present invention provides a kit comprising a mismatch endonuclease, a DNA ligase and one or more exonucleases together having combined 5' to 3' activity on double-stranded DNA and 3' to 5' activity on single- stranded DNA.

[0023] Other aspects and features of the present invention will become apparent to those of ordinary skill in the art upon review of the following description of specific embodiments of the invention in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The figures illustrate, by way of example only, embodiments of the present invention, are as described below.

[0025] Figure 1: Schematic illustration of the method of the invention. Synthetic DNAs was synthesized with 5'-biotin or 5' -locked nucleic acid (LNA) primers to protect the 5' ends of full-length DNA being digested by the 5' exonuclease activity of T7EI and Surveyor™ endonucleases. The non-specific nicking and cutting of T7EI and Surveyor™ on perfect match DNA was repaired with DNA ligase (1). The cleaved DNA molecules with 5'-phosphate were digested by Lambda Exonuclease and Exonuclease I (2). The 5'-biotin or 5'-LNA was removed by restriction digestion (3), and the 3'-underhang ends were re-filled by Taq DNA polymerase (4). A heat inactivation or buffer exchange was performed between each process step to control the activity of enzymes.

[0026] Figure 2: Schematic illustration of GFPwv circularization. Two double- stranded DNA species (760 bp and 720 bp) were amplified in separated pools, using one 5'-phosphorylated primer and a 5'-hydroxylated primer (1). Single-stranded DNAs were generated by Lambda Exonuclease digest of the 5'-phosphorylated strand (2). The two single-stranded DNAs were annealed together (3) and then annealed with a 40mer linker (4). After phosphorylation, the DNA was circularized by ligation (5).

[0027] Figure 3: 69 base pair oligonucleotides, each having one of all possible mutation mismatches (AA, AG, AC, TG, CT, TT, GG and CC), and 1-, 2-, 3-, 5- and 10-base deletion mismatches (Id, 2d, 3d, 5d and lOd) were treated with Surveyor™ Nuclease, T7E1, EndoV, T4E7 and MutS. (A) 20 pmol DNA with 1 unit Surveyor™ Nuclease for 60 min at 42 °C. (B) 20 pmol DNA with 10 units T7E1 for 4 h at 37 °C. (C) 20 pmol DNA with 10 units EndoV for 4 h at 37 °C. (D) 1 pmol DNA with 500 units T4E7 for 4 h at 37 °C. (E) 2 pmol DNA with 20 pmol MutS for 4 h at 37 °C. (M: 5 bp DNA ladder, PM: perfect match).

[0028] Figure 4: Time course experiment of control DNA containing 50% homoduplex and 50% heteroduplex DNAs (GG or CC mismatch at position 417). DNA mixture was incubated with T7EI for 0, 20, 25 and 30 min at 37 °C. M, 100 bp Plus DNA Ladder.

[0029] Figure 5: Study the repairing of non-specific DNA nicking using

Ampligase. Panel A. Control DNA was treated with T7EI without Ampligase™ in NEBuffer2 (B2) or Ampligase™ buffer (AB). (AB+Ligase): Control DNA was treated with T7EI and Ampligase™ mixture in a Ampligase™ buffer. T7EI digested samples were further denaturized and re-hybridized to confirm the ligation (Re- hybridization). Panel B, re-hybridized DNAs were subjected to Lambda Exonuclease and Exonuclease I treatment. GC, control DNA mixture. M, 100 bp Plus DNA Ladder.

[0030] Figure 6: Study of exonuclease activity of T7EI. GFPwv gene was synthesized with outer primers containing either 5'-hydroxyl or 5'-biotin (lanes 1 and 5) and subjected to T7EI (lanes 2 and 6). After digestion, the product was incubated with Lambda Exonuclease and Exonuclease I (lanes 3 and 7). As controls, the substrates were also treated with exonucleases without prior T7EI incubation (lanes 4 and 8). M, 100 bp DNA ladder.

[0031 ] Figure 7: Study the temperature effect of T7EI treatment. Panel A, DNA was incubated with both of T7EI and Ampligase™ at 37°C or 45°C for 25 min. Controls were performed analogous without Ampligase. Panel B, the digested products were re-hybridized. PGR, original GFPuv. M, 100 bp DNA Ladder.

[0032] Figure 8 : Incubation of synthetic GFPwv with T7EI and either 5 or 15 units Ampligase™ at 37 °C for 25 min. (lanes 2 and 3). The products were re- hybridized (lanes 3 and 4) and subjected to Lambda Exonuclease and Exonuclease I (lanes 5 and 6). GFP, re-hybridized GFPuv. M, 100 bp Plus DNA Ladder.

[0033] Figure 9: Incubation of synthetic GFPuv with T7EI and 15 units of Ampligase™ for either 30 or 40 min (lanes 2 and 3) at 37 °C. After re-hybridization (lanes 4 and 5), the products were subjected to Lambda Exonuclease and Exonuclease I (lanes 6 and 7). M, 100 bp Plus DNA Ladder. Lane 1 : Synthetic GFPuv.

[0034] Figure 10: Reaction scheme for post-T7EI process and preparation of GFPuv for cloning. Mismatch is represented as a triangle. GFPuv is treated with T7EI and Ampligase. Cleaved DNA fragments are degraded by exonucleases, and 5'-Biotin (shaded circles at 5' DNA ends) is removed by restriction enzymes (BamHI and EcoRI). Heat inactivation or solid-phase buffer exchange is also conducted between each step to inactivate or remove the T7E1, exonucleases, and restriction enzymes.

[0035] Figure 11: Panel A, enzymes were heat inactivated after every reaction. GFPuv was treated with T7EI and 15 units Ampligase™ for 40 min at 37 °C (lane 2). The sample was heat inactivated (95 °C for 30 min), re-hybridized, and further treated with Lambda Exonuclease and Exonuclease I for 60 min at 37 °C (lane 3).

Exonucleases were inactivated at 80 °C for 20 min, followed by restriction digest to remove the biotin-tag (lane 4). Panel B, enzyme was removed by buffer exchange method. GFPuv was treated by T7EI and Ampligase™ as in Panel A (lane2), and then immediately subjected to Lambda Exonuclease and Exonuclease I for 5 min at 37 °C. Subsequently, the reaction mixture was purified by a buffer exchange to remove the T7EI and exonucleases (lane 3) from the assembled product, followed by the biotin- tag removal (lane 4). Lane 1 : re-hybridized GFPuv. M, 100 bp DNA Ladder.

[0036] Figure 12: Synthetic GFPwv was treated with T7EI and Ampligase™ for 40 min at 37 °C (lane 2). Mixture contained truncated DNA was digested by λ Exo and Exo I, and then inactivated at 80°C for 20 min to inactivate enzymes (lane 3). Biotin-tag was removed by restriction enzyme cleavage followed by a buffer exchange (lane 4). Lane 1 : Intact GFPwv.

[0037] Figure 13: Simultaneous and consecutive incubation of control DNA with Surveyor™ Nuclease and 9°N™ DNA Ligase. The DNA substrate contained 50% heteroduplex at position 417 and 50% homoduplex DNAs (632 bp). DNA was incubated with nuclease at 42 °C for 20 min, followed by: 1) another 30 min at 45 °C without additional ligase (lane 2); and 2) added with 9°N™ DNA Ligase and incubated for either another 30 min (lane 8) or 60 min (lane 10) at 45 °C. For simultaneous incubation, DNA was incubated with nuclease and 9°N™ DNA Ligase together at 42 °C for 20 min followed by either another 30 min (lane 4) or 60 min (lane 6) at 45°C. Lanes 3, 5, 7, 9, and 11 were the re-hybridized products for each corresponding samples. M, 100 bp Plus DNA ladder. Lane 1, Control DNA.

[0038] Figure 14: Control experiments for studying the Surveyor™ Nuclease activity at various temperatures. Control DNA was first treated with Surveyor™ Nuclease at 42 °C for 20 min, followed by an additional incubation at 16-85 °C for 30 min. Control, sample with additional incubation. M, 100 bp Plus DNA ladder.

[0039] Figure 15: T4 DNA Ligase repair of control DNA nicked by Surveyor™ Nuclease. DNA was incubated with Surveyor™ Nuclease at 42 °C for 20 min, followed by ligase repair for either 30 min (L30), 60 min (L60) or 120 min (LI 20) at either 16°C or 25°C. The samples were then denaturized and re-hybridized. A control was also conducted with only the Surveyor™™ treatment (42 °C for 20 min) and re- hybridized (lane S). Panel A, ligation at 16 °C. Panel B, ligation at 25 °C.

[0040] Figure 16: Reaction scheme for Surveyor™-based error filtering. Linear DNA remaining in the circularization mixture is cleanup by exonucleases (step 1). Circularized DNA with mismatch is cleaved by Surveyor™ (step 2), followed by a nick repair by T4 DNA ligase (step 3).

[0041] Figure 17: Synthetic GFPuv was circularized, error filtered using

Surveyor™ Nuclease and repaired by T4 DNA ligase. GFPwv was circularized by annealing a 760-bp single-stranded GF?uv with a 720-bp complement single-stranded GFPwv and a 40mer linker (lane 1). The product was incubated with Lambda

Exonuclease, Exonuclease I and Surveyor™ Nuclease, followed by T4 DNA Ligase at either 25 °C (lane 2) or 16 °C (lane 4). The quantity of remaining GFPwv was amplified with a PCR (lanes 3 and 5). M, 100 bp Plus DNA ladder.

DETAILED DESCRIPTION

[0042] The methods described herein use a combination of 5' protecting groups and the various activities of nuclease enzymes to reduce mismatches within a pool of double-stranded DNA, in order to provide a population of DNA molecules for use in subsequent applications. After treatment in accordance with these methods, the resulting treated pool of double-stranded DNA molecules may have a reduced error rate in comparison to the error rate prior to treatment.

[0043] Thus, a method is provided comprising treating a pool of double-stranded DNA molecules that contains 5 '-protected double-stranded DNA molecules with a series of nucleases and ligases in order to reduce occurrences of mismatches within the pool of double-stranded DNA molecules. The pool of double-stranded DNA molecules that comprises 5 '-protected double-stranded DNA molecules is treated with . a mismatch endonuclease and a ligase, as well as with one or more exonucleases together having combined 5' to 3' double-stranded digestion activity and 3' to 5' single-stranded digestion activity.

[0044] As will be appreciated, the method is performed in vitro, in a cell-free system.

[0045] As used herein, the term mismatch in reference to double-stranded DNA refers to one or more bases in one strand of the DNA that is not properly paired with an opposing Watson-Crick base in the complementary strand. As will be appreciated, the Watson-Crick base pairs are G-C and A-T base pairs. Thus, mismatches include mispairing, which occurs when the incorrect base appears in the opposite strand, for example a T has been replaced with a G, resulting in an A-G mismatch. For example, a mismatch occurs when an A base in one strand does not have a complementary T base in the opposite strand with which to form a base pair. Possible single base mispairing mismatches include A-G, A-C, A-A, T-G, T-C, T-T, G-G and C-C.

Mismatches also include unpaired bases that form loops due to insertions or deletions within one strand of the double-stranded DNA. Thus, a mismatch may be one or more unpaired nucleotides that were incorrectly inserted into a sequence and that do not have a corresponding base with which to pair on the opposite strand or one or more unpaired bases for which the opposing nucleotide or nucleotides have been deleted from the opposing strand. Such unpaired mismatches result in the one or more unpaired nucleotides forming a loop that projects from the paired double- stranded DNA that flanks the mismatch. An unpaired mismatch may be one or more, two or more, three or more, four or more, five or more, ten or more, 15 or more or 20 or more unpaired nucleotides within one strand of a double-stranded DNA.

[0046] As used herein, the terms match or perfect match when used in reference to double-stranded DNA refers to double-stranded DNA in which all the bases in one strand form a Watson-Crick base pair with a corresponding base in the opposite, complementary strand.

[0047] Double-stranded DNA having one or more mismatches is referred to as mismatch DNA or heteroduplex DNA. Double-stranded DNA that has no

mismatches is referred to as match DNA, perfect match DNA or homoduplex DNA.

[0048] A pool of DNA molecules refers to a population of DNA molecules, which may or may not all have the same sequence and structure, including modifications. Thus, the population of DNA molecules within the pool may be heterogeneous or homogeneous. In the methods described herein, a pool of DNA molecules refers to a population of DNA molecules, at least some of which molecules where at least generated based on a common DNA sequence or template, but which may include molecules in which one or more errors was incorporated during synthesis of the DNA. Such molecules, when in double-stranded form, may be mismatch DNA (i.e. in which a strand containing an error is paired with an error-free strand, forming a heteroduplex DNA). [0049] The pool of double-stranded DNA comprises any DNA population for which it is desired to have mismatch occurrence within the double-stranded DNA - reduced. The double-stranded DNA may be DNA that is synthesized by chemical or enzymatic methods, or a combination thereof, for example by polymerase chain reaction (PGR), ligation chain reaction (LCR) or polymerase cycling assembly (PCA). Alternatively, the double-stranded DNA may be DNA created using molecular biology methods, such as replication within a cell or within a cell-free system.

[0050] The double-stranded DNA for which mismatch occurrence is to be reduced is 5 '-protected, which means that the 5' end of each strand of the DNA does not have a free phosphate group, but rather has a protecting group on the 5' end of each strand or is circular and therefore does not to have any 5' end. Thus, reference to 5'- protected DNA is reference to DNA that has a protecting group attached to at the 5' end of each strand, or to DNA that has no 5 ' end.

[0051] If the 5'-protected DNA has an attached protecting group, the protecting group may be any group that, when attached to a free 5' end of a DNA strand, protects the DNA from digestion by an exonuclease that would otherwise digest the 5' end of the DNA if it were unprotected. Thus, the protecting group protects the 5' end of single-stranded DNA from acting as a substrate for an exonuclease that has 5' to 3' activity for single-stranded DNA and also protects the 5' end of double-stranded DNA from acting as a substrate for an exonuclease that has 5' to 3' activity for double- stranded DNA.

[0052] Protecting groups are known in the art, and include, without limitation, biotin, locked nucleic acids (LNA™), peptide nucleic acids, or any other known group that may be attached to the 5' end of a DNA strand and which serve to protect the DNA strand from 5' to 3' exonuclease digestion.

[0053] The protecting group may be added during synthesis of DNA, for example as a modified nucleotide added to synthesis of a primer used for PGR or LCR.

Alternatively, the protecting group may be added after synthesis of the DNA, for example by treatment of a pool of DNA with a reactive protecting group under conditions which allow for the protecting group to react and attach to the 5' end of the DNA. [0054] As stated above, instead of having a protecting group on the 5' end of the DNA strands, the double-stranded DNA may be circular DNA. The DNA may be synthesized in a cellular replication system from a circular DNA. Alternatively, the DNA may be synthesized as linear DNA and then subsequently circularized. One such method of creating circular double-stranded DNA is described in the Examples set out below.

[0055] The pool of double-stranded DNA molecules may be isolated or purified, or may have impurities or contaminates removed that may interfere with subsequent enzyme reactions, prior to use in the methods described herein.

[0056] The pool of double-stranded DNA molecules that comprises 5 '-protected DNA molecules is treated with a mismatch endonuclease in order to cut double- stranded DNA molecules that contain one or more mismatches.

[0057] As will be appreciated, an endonuclease catalyses cleavage of a nucleic acid strand at a position within the strand, in contrast to an exonuclease, which cleaves a nucleic acid strand at an end to remove one nucleotide at a time. A mismatch endonuclease is an endonuclease that recognizes mismatches within double-stranded DNA, including mispairing and unpaired mismatches, and cleaves the DNA (cuts both strands of the double-stranded DNA) at the site of the mismatch in order to excise the mismatch f om the DNA. Depending on the mismatch endonuclease used, the endonuclease will cut the DNA either 5' or 3' to the mismatch.

[0058] One or more mismatch endonucleases may be used to treat the double- stranded DNA, including for example endonuclease V, T7 endonuclease I,

Surveyor™ nuclease (Transgenomic, Omaha, Nebraska, USA), T4 endonuclease VII, and CEL I nuclease. In particular, the endonuclease used may be T7 endonuclease I (T7E1), or the endonuclease used maybe Surveyor™ nuclease. T7 endonuclease I and Surveyor™ both exhibit high efficiency and can recognize all of the single-base mismatches, as well as multi-base deletions and insertions. Mismatch endonucleases are generally readily commercially available. Surveyor™ is a mismatch-specific endonuclease derived from celery that recognizes and cleaves all types of single base mispairing mismatches as well as small insertion or deletion mismatches. [0059] A sufficient amount of mismatc endonuclease is used to treat the pool of double-stranded DNA to allow for reduction of the mismatch occurrence within the pool of DNA over a given time period. The amount of mismatch endonuclease used will depend on various factors, including the amount and concentration of the double- stranded DNA within the reaction mixture. A skilled person can easily determine an appropriate amount of a particular mismatch endonuclease to use under certain reaction conditions by conducting a time course experiment for various amounts of mismatch endonuclease.

[0060] For example, in a reaction volume of 10 to 20 μΐ containing from 1 to 50 pmole double-stranded DNA, from about 1 to about 500 units of a particular mismatch endonuclease may be used. The particular conditions, amounts, concentrations and times to use can readily be optimized using routine laboratory methods for a given DNA pool/mismatch endonuclease reaction, as described in the Examples below.

[0061 ] The treating of the pool of double-stranded DNA with the mismatch endonuclease is performed in a suitable buffer for the mismatch endonuclease that contains any coenzymes or counterions that may be required for mismatch

endonuclease activity. If the endonuclease is purchased commercially, the supplier will typically provide a suitable buffer or buffers. The reaction is conducted at a suitable temperature and under conditions that allow for the mismatch endonuclease to cleave mismatch DNA.

[0062] Thus, treating the pool of double-stranded DNA with the mismatch endonuclease is performed under suitable conditions to allow the mismatch endonuclease to recognize and cleave mismatch DNA present in the pool of double- stranded DNA. The mismatch endonuclease cleaves mismatch DNA at the site of the mismatch, yielding at least two DNA fragments, each generated DNA fragment having at least one unprotected 5' end. Alternatively, if the DNA is circular, the mismatch endonuclease will cleave mismatch DNA to yield linear DNA (one or more fragments from a single circular DNA molecule), with each newly generated 5' end being unprotected.

[0063] However, as is known, mismatch endonucleases may often have additional nuclease activities, including 5' exonuclease activity and non- specific single-strand nicking activity. The protection of the 5' end of perfect match DNA will protect such DNA molecules within the pool of double-stranded DNA from any 5' exonuclease activity possessed by the one or more mismatch endonucleases used. If the 5' exonuclease activity of the mismatch endonuclease is strong, 5' protection in the form of circularization may be more effective than protection using a protecting group attached at the 5' end of each strand.

[0064] In order to repair any nicking in single strands of double-stranded DNA that may occur at random, non-specific sites as a result of the mismatch endonuclease " activity, the methods involve treatment of the pool of double-stranded DNA molecules that have been treated with the mismatch endonuclease with a DNA ligase enzyme. The DNA ligase may be used simultaneously with, subsequently to, or in overlap with the mismatch endonuclease. For example, if the mismatch endonuclease has a tendency for non-specific nicking, but does not tend to recognize the site of a nick as a mismatch, the DNA ligase may be used subsequent to the treatment with the mismatch endonuclease. However, for mismatch endonucleases that have a tendency for non-specific nicking and that may tend to recognize a nick site as a mismatch, the DNA ligase may be used simultaneously with, or overlapping with, the mismatch endonuclease in order to reduce non-specific cutting of both strands of the double- stranded DNA.

[0065] A DNA ligase is an enzyme that catalyses bond formation between a free 5' end of one DNA strand with a free 3' end of a second DNA strand, thus joining two DNA segments to create one continuous DNA strand.

[0066] One or more DNA ligases may be used. The DNA ligase is any DNA ligase that repairs nicks in a single strand of a double-stranded DNA. Suitable DNA ligases include, without limitation, Ampligase™, 9°N™ ligase, T4 DNA ligase, Taq

DNA ligase, and E. coli DNA ligase. hi particular, the DNA ligase used may be T4

DNA ligase, may be Ampligase™, or may be 9°N™ ligase. DNA ligases are generally readily commercially available. Ampligase™ (Epicentre Biotechnologies,

Madison, Wisconsin, USA) is a thermostable DNA ligase derived from a thermophilic bacterium and catalyzes NAD-dependent ligation of adjacent 3 '-hydroxylated and 5^r- phosphorylated termini in duplex DNA. 9°N™ (New England Biolabs, Ipswich, Maine, USA) is a DNA ligase active at elevated temperatures (45-90°C) that is isolated from a thermophilic archaea Thermococcus sp.

[0067] A sufficient amount of DNA ligase is used to treat the pool of double- stranded DNA molecules to allow for repair of non-specific nicks within the double- stranded DNA molecules over a given time period. The precise amount of DNA ligase used will depend on various factors, including the amount and concentration of double-stranded DNA, the type and amount of mismatch endonuclease used to cleave mismatch DNA molecules, and the length of time for treatment with the DNA ligase. A skilled person can easily determine an appropriate amount of a particular DNA ligase to use under certain reaction conditions by conducting a time course experiment for various amounts of DNA ligase.

[0068] For example, in a reaction volume of 10 to 20 μΐ containing from 1 to 50 pmole double-stranded DNA treated with a mismatch endonuclease, from about 1 to about 500 units of a particular DNA ligase may be used. The particular conditions, amounts, concentrations and times to use can readily be optimized using routine laboratory methods for a given DNA pool/ligase reaction, as described in the

Examples below.

[0069] The DNA ligase may be added to the pool of double-stranded DNA at the same time as the mismatch endonuclease, allowing the ligase to act for the whole period during which the mismatch endonuclease is acting on the pool of double- stranded DNA. Alternatively, the mismatch endonuclease may be added for a period of time and then the DNA ligase may be added, thus being allowed to act together with the mismatch endonuclease for some of the total reaction time of the mismatch endonuclease. In another alternative, the DNA ligase may be added after the treatment with the mismatch endonuclease is completed. Chemical or heat inactivation of the mismatch endonuclease may be used to ensure the endonuclease reaction is completed, or the buffer containing the mismatch endonuclease may be exchanged, thus removing the mismatch endonuclease from the pool of double- stranded DNA.

[0070] The treating of the pool of double-stranded DNA with the DNA ligase is performed in a suitable buffer for the DNA ligase that contains any coenzymes or counterions that may be required for DNA ligase activity. If the DNA ligase is purchased commercially, the supplier will typically provide a suitable buffer or buffers. The reaction is conducted at a suitable temperature and under conditions that allow for the DNA ligase to ligate nicked DNA. The conditions under which the pool of double-stranded DNA is treated with the DNA ligase should be conditions compatible with DNA ligase activity, but will also depend on the timing of addition of the DNA ligase. If the DNA ligase is added to act together with the mismatch endonuclease, the buffer used and the temperature at which the reaction is conducted should be suitable for both the DNA ligase and the mismatch endonuclease.

Particular conditions used will of course depend on various factors, but can be readily determined with routine laboratory techniques, for example as demonstrated in the following Examples. For example, the pool of double-stranded DNA may be treated with the ligase for from about 10 minutes to about 1 hour.

[0071] In the method, the pool of double-stranded DNA is treated with one or more exonucleases. The one or more exonucleases together have a combination of a 5' to 3' activity for double-stranded DNA and a 3' to 5' activity for single-stranded DNA, meaning that these activities are provided by the one or more exonucleases in total, rather than each exonuclease having a combination of these activities. As stated above, an exonuclease cleaves a nucleic acid strand at an end to sequentially remove one or more nucleotides. 5' to 3 ' activity for double-stranded DNA means that the exonuclease recognizes a free 5' end of a nucleic acid strand and progressively cleaves nucleotides, moving in the direction of the 3' end. Since the 5' to 3' exonuclease activity acts on double-stranded DNA, the 5' to 3' exonuclease recognizes a free 5' end of double-stranded DNA and cleaves through both strands in a double-stranded DNA. Conversely, 3' to 5' activity for single-stranded DNA means that the exonuclease recognizes a free 3 ' end of DNA and progressively cleaves nucleotides, moving in the direction of the 5' end. The 3' to 5' activity is on single- stranded DNA, and thus the 3' to 5' activity recognizes and cleaves a 3' overhang of double-stranded DNA, removing the overhang to yield a blunt ended double-stranded DNA, which would have a free 5' end available for digestion by the 5' to 3' exonuclease activity on double-stranded DNA.

[0072] The exonuclease activities are used in the present method to remove any DNA that is not 5' protected. Thus, any mismatch DNA that has been cleaved by the mismatch endonuclease will have an unprotected end that is available for digestion by either the 5' to 3' exonuclease activity or 3' to 5' exonuclease activity. Any perfect match DNA that is intact will be 5 '-protected, either by a protecting group or due to circularization, and thus will not be an available substrate for the exonuclease activity.

[0073] The 5' to 3 ' exonuclease activity and the 3 ' to 5' exonuclease activity may be contained within the same exonuclease, or may be provided by a combination of exonucleases.

[0074] Suitable 5' to 3' double-stranded exonucleases include, without limitation, Lambda exonuclease, RecJ and T7 exonuclease. Suitable 3' to 5' single-stranded exonucleases include, without limitation, exonuclease I and exonuclease T. In certain embodiments, the combination of suitable 5' to 3' double-stranded exonuclease activity and the 3' to 5' single-stranded exonuclease activity comprises Lambda exonuclease and exonuclease I.

[0075] For example, Lambda exonuclease has 5' to 3 ' digestion activity for both double-stranded and single-stranded DNA, while exonuclease I has 3' to 5' digestion activity only on single-stranded DNA. Since the 5' ends of the full-length double- stranded DNA that has not been cut by the mismatch endonuclease are protected, these ends will not be act as a substrate for either Lambda exonuclease or exonuclease I. Only mismatch DNA cut by the mismatch endonuclease will have a 5' end with a free phosphate group or a 3' overhang and thus be available for digestion by Lambda exonuclease or exonuclease I, and thus can be selectively digested and thus removed from the pool of double-stranded DNA molecules.

[0076] Where the 5'-protected DNA is linear with a protecting group on the 5' end, or is circular, the treatment of the pool of double-stranded DNA with the one or more exonucleases may be performed following treatment with the mismatch endonuclease and the DNA ligase, in order to digest any DNA molecules that were cleaved by the mismatch endonuclease. Alternatively, or in addition, if the 5'- protected double-stranded DNA molecules in the pool are circular, the treatment with the one or more exonucleases may be performed prior to treatment with the mismatch endonuclease in order to remove linear DNA molecules prior to the mismatch endonuclease treatment.

[0077] A sufficient amount of the one or more exonucleases is used to treat the pool of double-stranded DNA molecules to allow for digestion of double-stranded DNA molecules lacking 5' protection at both ends of the double-strand, over a given time period. The precise amount of the one or more exonucleases used will depend on various factors, including the amount and concentration of double-stranded DNA, the length of time for treatment with the one or more exonucleases. A skilled person can easily determine an appropriate amount of one or more exonucleases to use under certain reaction conditions by conducting a time course experiment for various amounts of particular one or more exonucleases.

[0078] For example, in a reaction volume of 10 to 20 μΐ containing from 1 to 50 pmole double-stranded DNA treated with a mismatch endonuclease and a DNA ligase, from about 1 to about 50 units of each of an exonuclease having 5' to 3' double-stranded activity and an exonuclease having 3' to 5' single-stranded activity may be used. The particular conditions, amounts, concentrations and times to use can readily be optimized using routine laboratory methods for a given DNA

pool/exonuclease digestion reaction, as described in the Examples below.

[0079] As with the other enzyme reactions above, treatment with the one or more exonucleases is performed in a suitable buffer for one or more exonucleases, the buffer containing any coenzymes or ions that may be required for each of the exonuclease activity. If the one or more exonucleases are purchased commercially, the supplier will typically provide a suitable buffer or buffers. The reaction is conducted at a suitable temperature and under conditions that allow for the exonucleases to digest substrate DNA, either double-stranded with an unprotected 5' end or single-stranded 3' overhang. The conditions under which the pool of double- stranded DNA is treated with the one or more exonucleases should be conditions compatible with exonuclease activity. Particular conditions used will of course depend on various factors, but can be readily determined with routine laboratory techniques, for example as demonstrated in the following Examples. For example, the pool of double-stranded DNA may be treated with the ligase for from about 2 minutes to about 15 minutes. [0080] As stated above, the various types of enzymes used in the method may be commercially available, and suitable buffers in which the enzymes are active are typically available. Thus, the conditions for activity of such enzymes tend to be known, including requisite salts, ions, co-enzymes, temperature and pH. A skilled person could adjust conditions based on the known parameters for the enzymes to determine appropriate conditions for each of the above enzyme reactions involved in the described method, without undue experimentation.

[0081] Optionally, procedures may be taken to remove or inactivate enzymes prior to performing a subsequent step in the method. For example, once the mismatch endonuclease treatment is completed, prior to treatment with the DNA ligase, the mismatch endonuclease may be inactivated. Inactivation may be by heat or by addition of a chemical that inhibits the activity of the mismatch endonuclease but which does not interfere with the DNA ligase activity. In another example, following mismatch endonuclease and DNA ligase treatment, the mismatch endonuclease and DNA ligase may be removed prior to addition of the one or more exonuclease.

Enzymes may be removed by buffer exchange methods, which are generally known in molecular biology techniques, including by precipitation methods, spin column methods or solid phase methods such as microbeads.

[0082] Embodiments of the described method are depicted in the schematic diagram shown in Figure 1. The schematic diagram also includes subsequent treatment steps to pair the pool of double-stranded DNA molecules for use in subsequent cloning procedures.

[0083] In one embodiment, the mismatch endonuclease used is T7E1 , the DNA ligase is Ampligase™, and the exonucleases are Lambda exonuclease and

exonuclease I. The double-stranded DNA is 5 '-protected with a biotin group or an LNA group. The pool of double-stranded DNA is treated with the T7E1 and the Ampligase™ together in the same reaction for about 30 to about 40 minutes at about .37 °C. For this combination of enzymes, elevated temperatures may increase the non-specific nicking activity of the T7EI, while lower temperatures may impair the T7EI cleavage efficiency on heteroduplex DNA. The simultaneous treatment with T7E1 and Ampligase™ allows for repair of any non-specific nicking that results from the T7EI. [0084] In another embodiment, the mismatch endonuclease used is Surveyor , the DNA ligase is 9°N™ or T4 DNA ligase, and the exonucleases are Lambda exonuclease and exonuclease I. The double-stranded DNA may be 5 '-protected by circularization, or may be 5' protected with a protecting group such as biotin or an LNA group. The pool of double-stranded DNA is first treated with the Surveyor™ endonuclease, for example for about 20 minutes or longer at about 42 °C followed subsequent treatment with either DNA ligase, for about 30 to 120 minutes at about 45 °C (9°N™) or about 16 °C (T4 DNA ligase). The kinetics of Surveyor™ nuclease are different from those of T7EI. Surveyor has shown high non-specific nicking and low non-specific cutting activities. The nicked DNA segments remain hybridized at the endonuclease treatment temperature (45 °C). This approach may allow Surveyor™ incubation at a temperature that is optimal for the endonuclease activity in order to fully cleave any mismatch DNA, followed by an extended ligation at ligase-optimal temperature in order to repair the nicked DNA.

[0085] If desired, the pool of double-stranded DNA that results following the above method may be amplified using known amplification methods, such as PGR, LCR, PCA methods, and if applicable depending on the method of amplification, re- protected at the 5' ends, and then the above steps of treating with a mismatch endonuclease and a DNA ligase and treating with one or more exonucleases together having combined 5' to 3' double-stranded activity and 3' to 5' single-stranded activity may be repeated. Thus, the double-stranded DNA produced from the method having reduced mismatches may then be used as a template for amplification, and the amplified product may then again be treated in order to again reduce occurrence of mismatches that may arise during the amplification process.

[0086] Following completion of the method, the final resulting pool of double- stranded DNA in which the occurrence of mismatches has been reduced may be used in subsequent applications. To prepare the double-stranded DNA, the protecting group may be removed, for example by digestion with a suitable restriction endonuclease to yield a desired portion of double-stranded DNA molecule ready for cloning or other applications. Similarly, a circular DNA may be linearized using a suitable restriction endonuclease. Restriction endonucleases are well known and are commercially available. The double-stranded DNA molecules can be readily designed to include recognition sequences for a particular endonuclease near each end of a linear double stranded DNA or at relevant positions within a circular DNA to assist with manipulation of the DNA following the described method. For example, a particular DNA sequence that is to be amplified can be screened using available computer programs to identify existing restriction endonuclease recognition sites within the sequence, and then appropriate restriction sites may be chosen and included within primers used in the amplification process.

[0087] Also provided is a commercial package or kit for performing the described method. The kit may comprise the various enzymes, and optionally suitable buffers, for performing the various enzyme reactions in the method. Thus, in one embodiment the kit comprises one or more mismatch endonuclease, one or more DNA ligase, and one or more exonuclease having combined activity of 5' to 3' activity on double- stranded DNA and 3' to 5' activity on single-stranded DNA. The kit may further comprise a buffer suitable for use with the one or more endonuclease, alone or together with the one or more DNA ligase, a buffer suitable for use with the one or more DNA ligase, and/or a buffer suitable for use with the one or more exonucleases. The kit may also include instructions for performance of the described method on a pool of double-stranded DNA molecules, the pool comprising 5 '-protected double- stranded DNA molecules.

[0088] The described methods are further exemplified by way of the following non-limiting examples.

EXAMPLES

[0089] 1. MATERIALS AND METHODS [0090] 1.1 Preparation of DNA Substrates

[0091] Activity assay of mismatch-specific enzymes: Short control DNAs (69 bp) containing perfect match, one of all possible single-base mutations or from 1-10 base deletion mismatch at a central position were generated by annealing single- stranded oligonucleotides (sequences listed in Table 1 ; mismatch bases (underlined) are located in the center of each single-strand oligonucleotide) as follows: 10 μΜ forward and reverse HPLC-purified oligonucleotides (69mers, Sigma-Aldrich, Singapore) in 100 μΐ, l TE Buffer (10 mM Tris-HCl, 1 mM EDTA) were heated at 95 °C for five minutes, then cooled to 25 °C with cooling rate of -3.6 °C/min.

Double-stranded DNAs having one of all possible mismatches (AA, AG, AC, TG, CT, TT, GG and CC), or a deletion mismatch of from 1 to 10 bases were generated by the hybridization of two oligonucleotides with the combination as shown in Table 2. Perfect match DNA contains C/G at the central position.

Table 1. Primers used to created double stranded DNA pools

Table 2. Double-stranded 69mers

Mismatch type Primer Pair

A/A Mis_f69A /Mis_r69A

A G Mis_f69A / Mis_r69G

A/C Mis_f69A / Mis_r69C

T/G Mis_f69T /Mis_r69G

T/C Mis_f69T / Mis_r69C

T/T Mis_f69T / Mis_r69T

G/G Mis_f69G / Mis_r69G

c/c Mis_f69C / Mis_r69C

1 deletion Mis f69G / Mis ID 2 deletion Mis f69G /Mis 2D

3 deletion Mis f69G/Mis 3D

5 deletion Mis r69A/Mis 5D

10 deletion Mis r69A/Mis 10D

Perfect match Mis f69C / Mis r69G

[0092] The DNA 69mers (20 pmole) were treated either with Surveyor™, T7 Endonuclease I (T7EI), Endonuclease V (Endo V), T4 Endonuclease VII (T4E7), or MutS. A total reaction volume of 10 \iL^~ contained either: 1) 1 unit Surveyor™ Nuclease and 1 μϊ_^ Surveyor™ Nuclease enhancer (Transgenomic^®, USA) in 0.5x KOD Hot Start DNA Polymerase Buffer (Novagen, USA) at 42 °C for 60 min, 2) 10 units T7E1 (NEB, USA) in 50 mM NaCl, 10 mM Tris-HCl, 10 mM MgCl₂ and 1 mM dithiothreitol (pH 7.9, NEB, USA) at 37 °C for 4 h, 3) 10 units Endo V (NEB, USA) in 20 mM HEPES-NaOH (pH 7.4), 100 mM KC1, 2 mM MnCl₂ and 0.1 mg/mL BSA (Trevigen, USA) at 37 °C for 4 h, 4) 500 units T4E7 (USB, USA) in 50 mM Tris-HCl (pH 8.0), 10 mM MgS0₄, 10 mM 10x dithiothreitol and 100 ng/niL BSA at 37 °C for 4 h, 5) 21 pmol MutS enzyme (USB, USA) with 2 pmol DNA in T7EI buffer.

Treatment was conducted at 37 °C for 4 h.

[0093] Amplification of long DNA: Control DNA (632 bp, Transgenomic) with a GG or CC mismatch at position 417 was used for optimization the performances of T7EI and Surveyor™ Nuclease. Two templates, one with cytosine, the other with guanine at position 417 were amplified in a 50 \iL reaction with 10 ng DNA template, 250 nM primers (Table 3), 2 mM dNTPs, 2 mM MgS0₄ and 1 unit KOD Hot Start DNA Polymerase (Novagen) in 1 x KOD Hot Start buffer. The thermal cycling was conducted at 95 °C for 2 min, 30 cycles of 95 °C for 30 s, 65 °C for 30 s, 72 °C for 30 s and 72 °C for 10 min. For mismatch formation, equal volumes of both amplicons were combined, denatured at 95 °C for 5 min and re-hybridized by decreasing the temperature from 95 °C to 50 °C with -1 °C/min and then cooled down to room temperature. This process generated a mixture contained 50% single-base mismatch DNA at position 417, and 50 % homoduplex DNA.

Table 3. Primers for long DNA amplification containing a single base mismatch

Primer Sequence SEQ ID NO P_Eorward ACACCTGATCAAGCCTGTTCATTTGATTAC 14

P_Reverse AAGCTCCGCAGATCATTCTTTGGCG 15

[0094] GFPi/v DNA synthesis: DNA encoding GFPwv (760bp) was chosen as a model DNA. Oligonucleotides for GFPwv were designed according to Binkowski et al. [1] (Table 4). All inner oligonucleotides comprise a 5'-phosphate while the outer primers contain either 5'-biotin or 5'-hydroxyl. Desalted oligonucleotides were

obtained from Research Biolabs (Singapore) without additional purification.

Table 4. Primers GUPnv DNA circularization

Oligomer Sequence Seq ID No T_m (°C) Length

OH-F_Pr₇₆₀ OH-AGAGGATCCCCGGGTACCGG 16 62.5 20

P-R_Pr₇₆₀ P-CTCAGTTGGAATTCATTATT 17 53.0 20

P-F_Pr₇₂₀ P-TAGAAAAAATGAGTAAAGGA 18 51.0 20

OH-R_Pr₇₂₀ OH-TGTAGAGCTCATCCATGCCA 19 64.0 20

Linker OH-CCGGTACCCGGGGATCCTCTCTCAGTTGGAATTCATTATT 20 82.0 40

[0095] GFPwv DNA (760 bp) was assembled by a two-step PCR-based process, performed with 50 μΐ of reaction mixture including 1 x PGR buffer (Novagen), 2 mM of MgS0₄, 0.5 mM of each of deoxynucleotide precursor (dNTP) (Stratagene), 500 g/ml of bovine serum albumin (BSA), 10 nM of oligonucleotides, 400 nM of

forward and reverse primers, and 1 U of OD Hot Start (Novagen). The polymerase chain assembly (PCA) was conducted under the following conditions: 2 min of initial denaturation at 95°C, 20 cycles of 95 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s and a last extension of 72°C for 10 min. For gene amplification, the PGR protocol used the same conditions, using 1 μΙ_. of assembly product diluted in an amplification reaction mixture of 20 μΐ with primers concentration of 400 nM (F_Pr and R_Pr primers in

Table 5). The assembled product mixture contained a pool of DNAs having various sequences, PGR amplified from the original DNA templates. The assembled DNA was denatured and re-hybridized to generate assorted mismatch DNAs. Table 5 lists the oligonucleotides used for the two-step PCR-based gene synthesis of E. coli codon- optimized GFPuv gene. F_Pr and R_Pr are the outer forward and reverse

amplification primers, respectively. Oligonucleotides were synthesized with 5'- hydroxyl while outer primers contain either 5'-hydroxyl, 5 '-phosphate or 5'-biotin- tag.

Table 5. Oligonucleotides gene synthesis of GFP¾v gene

Oligomer Oligonucleotide sequence (5' to 3') Seq ID T_m Overlap Length

No (°C) (bp) (nt)

F0 AGAGGATCCCCGGGTACCGGTAGAAAAAATGAGTAAAGGA 21 44.5 20 40

R0 ACTCCAGTGAAAAGTTCTTCTCCTTTACTCATTTTTTCTA 22 50.7 20 40

Fl GAAGAACTTTTCACTGGAGTTGTCCCAATTCTTGTTGAAT 23 50.5 20 40

RI CCCGTTAACATCACCATCTAATTCAACAAGAATTGGGACA 24 52.2 20 40

F2 TAGATGGTGATGTTAACGGGCACAAATTTTCTGTCAGTGG 25 50.7 20 40

R2 TTGCATCACCTTCACCCTCTCCACTGACAGAAAATTTGTG 26 56.6 20 40

F3 AGAGGGTGAAGGTGATGCAACATACGGAAAACTTACCCTT 27 50.5 20 40

R3 CCAGTAGTGCAAATAAATTTAAGGGTAAGTTTTCCGTATG 28 47.1 20 40

F4 AAATTTATTTGCACTACTGGAAAACTACCTGTTCCATGGC 29 53.6 20 40

R4 GAAAGTAGT GACAAGT GTT GG CCAT GGAACAGGT GT T T T 30 49.8 20 40

F5 CAACACTTGTCACTACTTTCTCTTATGGTGTTCAATGCTT 31 50.4 20 40

R5 TATGATCCGGATAACGGGAAAAGCATTGAACACCATAAGA 32 53.2 20 40

F6 TTCCCGTTATCCGGATCATATGAAACGGCATGACTTTTTC 33 52.8 20 40

R6 CCTTCGGGCATGGCACTCTTGAAAAAGTCATGCCGTTTCA 34 60.5 20 40

F7 AAGAGTGCCATGCCCGAAGGTTATGTACAGGAACGCACTA 35 51.5 20 40

R7 CCCGTCATCTTTGAAAGATATAGTGCGTTCCTGTACATAA 36 50.4 20 40

F8 TATCTTTCAAAGATGACGGGAACTACAAGACGCGTGCTGA 37 57.7 20 40

R8 TATCACCTTCAAACTTGACTTCAGCACGCGTCTTGTAGTT 38 49.3 20 40

F9 AGTCAAGTTTGAAGGTGATACCCTTGTTAATCGTATCGAG 39 51.1 20 40

R9 TTAAAATCAATACCTTTTAACTCGATACGATTAACAAGGG 40 40.6 20 40

F10 TTAAAAGGTATTGATTTTAAAGAAGATGGAAACATTCTCG 41 50.7 20 40

RIO GTTGTACTCGAGTTTGTGTCCGAGAATGTTTCCATCTTCT 42 52.1 20 40

Fl l GACACAAACTCGAGTACAACTATAACTCACACAATGTATA 43 43.6 20 40

Rl l TTTGTTTGTCTGCCGTGATGTATACATTGTGTGAGTTATA 44 54.9 20 40

F12 CATCACGGCAGACAAACAAAAGAATGGAATCAAAGCTAAC 45 48.7 20 40

R12 ATGTTGTGGCGAATTTTGAAGTTAGCTTTGATTCCATTCT 46 52.3 20 40

F13 TTCAAAATTCGCCACAACATTGAAGATGGAAGCGTTCAAC 47 54.1 20 40

R13 TTGTTGATAATGGTCTGCTAGTTGAACGCTTCCATCTTCA 48 49.5 20 40

F14 TAGCAGACCATTATCAACAAAATACTCCAATTGGCGATGG 49 53.5 20 40

R14 TGTCTGGTAAAAGGACAGGGCCATCGCCAATTGGAGTATT 50 54.4 20 40

F15 CCCTGTCCTTTTACCAGACAACCATTACCTGTCGACACAA 51 54.5 20 40

R15 GGATCTTTCGAAAGGGCAGATTGTGTCGACAGGTAATGGT 52 55.3 20 40

F16 TCTGCCCTTTCGAAAGATCCCAACGAAAAGCGTGACCACA 53 57.4 20 40

R16 TACAAACTCAAGAAGGACCATGTGGTCACGCTTTTCGTTG 54 51.4 20 40

F17 TGGTCCTTCTTGAGTTTGTAACTGCTGCTGGGATTACACA 55 57.0 20 40 K17 TGTAGAGCTCATCCATGCCATGTGTAATCCCAGCAGCAGT 56 56.2 20 40

F18 TGGCATGGATGAGCTCTACAAATAATGAATTCCAACTGAG 57 46.2 20 40

R18 CTCAGTTGGAATTCATTATT 58 53.0 20

F_Pr AGAGGATCCCCGGGTACCGG 59 62.5 20

R_Pr CTCAGTTGGAATTCATTATT 60 53.0 20

Average T_m 51.8

[0096] Synthesis of circularized GFPwv: The synthesis scheme for GFPwv

circularization (760 bp) is given in Figure 2. DNA was assembled with 25 nM inner oligonucleotides in a reaction composed of 2 mM dNTPs, 2 niM MgS0₄, lx KOD

Hot Start buffer and 1 unit KOD Hot Start DNA Polymerase. The programme used was 95 °C for 2 min, 30 cycles of 95 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s and 72 °C for 10 min. The assembled product (black arrows) divided into two tubes, and separate PGR amplification reactions were performed with two different sets of

primers (Table 2). The programme used was 95 °C for 2 min, 25 cycles of 95 °C for 30 s, 60 °C (for OH-F_Pr₇₆₀ andP-RJ^>r₇₆₀) or 57 °C (for OH-R_Pr₇₂₀ andP-F_Pr₇₂₀) for 30 s and 72 °C for 30 s, followed by 72 °C for 10 min. The first primer set

contained 5'-hydroxyl forward primer (OH-F_Pr₇₆₀) and 5 '-phosphate reverse primer (P-R_Pr₇₆₀) generated a 760-bp full-length GFPwv, while the second set of primers

(5 '-phosphate forward primer (P-FJPr₇₂₀) beginning at position 21 of GFP sequence and 5'-hydroxyl reverse primer (OH-R_Pr₇₂₀) beginning at position 21 of the

complement strand) generated a 720-bp GFPwv segment. 5 \L of each amplicon were subsequently digested with 10 units Lambda Exonuclease in a mixture of 12 xL at 37 °C for 30 min which generated two complementary single-stranded DNAs (step 2).

Both ssDNAs were annealed by heating up to 95 °C for 2 min and cooled down to room temperature with cooling rate of -1 °C/min (step 3). In step 4, 5 μΜ of the linker oligonucleotide (40mer) was added with the reaction mixture, incubated at 70 °C for 10 min and rapidly cooled down on ice. Finally (step 5), 13 xL of the DNA mixture was incubated with 20 units T4 Polynucleotide Kinase (2 iL), lx Ampligase™ buffer (2 μΐ,), lx T4 DNA ligase buffer (2 xL) 2 and 10 units Ampligase™ (2 iL) in a

reaction of 21 μΐ,, for 4 hours at 37 °C (phosphorylation) in lx T4 DNA ligase buffer.

The temperature was increased to 60 °C for 4 hours for Ampligase™ ligation. Note, the reaction mixture contained both Ampliase ligase and T4 DNA ligase buffers to provide proper buffer conditions for T4 Polynuclease Kinase and Ampligase™ simultaneously.

[0097] 1.2 T7El-based Error Filtering

[0098] Optimization of digestion using control DNA: A time course experiment for T7EI digestion was carried out with 4 \iL of control DNA (632 bp), 10 units T7EI in 1 NEBuffer2 with a total volume of 20 μΐ, at 37 °C for 20-30 min. Nicked DNA was repaired by Ampligase™ DNA Ligase contained 4 xL of control DNA, 10 units T7EI together with 5 units Ampligase™ DNA Ligase in 1 x Ampligase™ buffer instead of NEBuffer2, incubated at 37 °C for 25 min. Two control experiments were also conducted with only T7EI in 1 NEBuffer2 or 1 Ampligase™ buffer to investigate the buffer compatibility of T7EI. Afterwards all reactions were heated to 95 °C for 5 min and re-hybridized by cooling down to 50 °C with cooling rate of -1 °C/min and then incubation at room temperature. To remove cleaved DNA fragments, 5 iL of each three re-hybridized mixtures were incubated with 5 units Lambda Exonuclease (1 μΕ) and 20 units Exonuclease I (1 at 37 °C for 5 min. Agarose gel electrophoresis were performed for all of these samples (5 μΐ per sample).

[0099] Optimization of digestion using synthetic GFP v: All experiments contained 8 iL of synthetic GFPuv, 10 units T7EI in 1 x Ampligase™ buffer (total volume of 20 μΕ) unless stated otherwise. Four sets of experiment were performed to optimize the reaction: 1) Protecting the 5' terminus of full-length DNA using 5'-biotin primers; 2) Optimizating incubation temperature (37°C versus 45°C); 3) Optimizing the quantity of Ampligase™ (5-15 units); and 4) Optimizing the duration of incubation (30 min versus 40 min). The protocol details for each optimization is as follows. Optimization 1: GFPwv was synthesized with either 5'-hydroxyl or 5'-biotin primers, and then treated with T7EI without Ampligase™. 5 \L of reaction mixture was then incubated with 5 units of Lambda Exonuclease and 20 units of Exonuclease

I in 7 μΕ for 5 min at 37 °C. Controls were also conducted with analogous protocol without T7EI. Optimization 2: The 5'-biotin GFPHV was treated with T7EI with or without co-added 5 units of Ampligase™ DNA Ligase at either 37 °C or 45 °C for 25 min. Samples were heated to 95 °C and cooled down to 50°C with cooling rate of -1

°C/min and then room temperature. Optimization 3 : The 5'-biotin GFPwv was incubated with T7EI and with either 5 or 15 units of Ampligase™ at 37 °C for 25 min. Optimization 4: The 5'-biotin GFPwv was with T7EI in the present of 15 units of Ampligase™ at 37 °C for either 30 min or 40 min. Samples of optimizations 3 and 4 were re-hybridized and cleaved fragments were digested by exonucleases.

[00100] DNA preparation for cloning: To check the error rate and to confirm the fidelity of synthetic DNA, endonuclease treated DNA (either T7EI or Surveyor™) need to be sequenced. Herein is the process in preparing the synthetic GFPwv for DNA cloning. The whole process included three steps: Step 1) simultaneously mismatch cleavage and nicking repair using T7EI and Ampligase; Step 2) Cleanup the non-full-length DNA with Lambda Exonuclease and Exonuclease I; and Step 3) Remove the 5'-biotin tag by simultaneously EcoRI and BamHI cleavage. Heat inactivation of T7E1, exonucleases, and restriction enzymes or solid-phase buffer exchange was also conducted between each step. The followings are the detail of each process step. Step 1): reaction mixture contained 8 μΐ, re-hybridized GFVuv (with 5'- biotin), 10 units T7EI, 15 units Ampligase™ DNA Ligase, lx Ampligase™ buffer in 20 μΐ, for 40 min at 37 °C. Step 2) 10 units Lambda Exonuclease and 40 units Exonuclease I were added to 15 μΐ. DNA product from step 1 and incubated for either 5 min (product from step 1 was buffer exchanged) or 60 min (product from step 1 was heat inactivated) at 37 °C. Step 3) Biotin-tag was removed in a 20 iL reaction containing 10 or 13 xL DNA product of step 2 with 10 units BamHI , 10 units EcoRI, 100 g μL bovine serum albumin and lx EcoRI buffer for 60 min at 37 °C. For enzymatic heat inactivation, T7EI was inactivated for 30 min at 95 °C and

exonucleases and restriction enzymes were heat treated for 20 min at 80 °C after each process step. Sample was cooled down to 30 °C with cooling rate of -l°C/min after each inactivation step. For buffer exchange method, reaction mixture was purified after each process step by a magnetic beads based DNA purification (ChargeSwitch PCR clean-up kit, Invitrogen), aimed to remove the enzymes in the reaction mixture. The sample was incubated with 40 xL magnetic beads and 50 \iL binding buffer for 2 min at room temperature in a PCR tube. The magnetic beads were captured by a magnet to remove the surfactant. Then, the beads were resuspended in 150

wash buffer, and washed twice with wash buffer and with the addition of a magnet. Finally, the DNA was eluted in 10 \xL elution buffer at 60 °C for 3 min. In another

experiment, combined T7EI and exonucleases heat inactivation were combined into one step (at 80 °C for 20 min) with buffer exchange to remove the restriction enzymes. [00101] 1.3 Surveyor™-based Error Filtering

[00102] Repair of non-specific DNA nicking with 9°N™ DNA ligase:

Control DNA was used as substrate for repair of nicked DNA by 9°N™ DNA ligase. A 35 xL master mixture was prepared containing 14 yiL of deionized- water, 14 yiL control DNA, 3.5 units Surveyor™ Nuclease and 3.5 iL Surveyor™ Nuclease Enhancer. The master mixture was aliquot into three tubes, where the first tube (12.5 iL, with Surveyor™ only) was incubated at 42 °C for 20 min and at 45 °C for another 30 min. The second tube (12.5 μΐ,, first Surveyor™ and then adding in 9°N™ DNA Ligase) was incubated at 42 °C for 20 min, and then added with 2 μΐ, 9°N™ DNA Ligase and 1.5 μL 9°N DNA Ligase reaction buffer and then further incubated at 45 °C for another 30-60 min. The third tube (Surveyor™ added together with 9°N™ DNA Ligase) was conducted where the Surveyor™ and 9°N™ DNA Ligase was added together with the same buffer conditions as in second tube and incubated at 42 °C for 20 min and then further incubated at 45 °C for another 30-60 min. All samples were denatured at 95 °C for 5 min and re-hybridized by cooling to 50 °C with -1 °C/min and then to room temperature.

[00103] Repair of non-specific DNA nicking with T4 DNA ligase: An activity assay of Surveyor™ Nuclease was conducted in a 5^L reaction with 2 μL control DNA, 0.5 units Surveyor™ and 0.5 μL Surveyor™ Nuclease Enhancer. All samples were incubated at 42 °C for 20 min before additional incubation for 30 min at various temperatures ranging from 16 °C to 85 °C. A second experiment was conducted to optimize nicking repair with T4 DNA ligase. A master mixture (40 μΕ) contained 16 μΐ, οΐ control DNA, 16 xL deionized- water, 4 units (check) Surveyor™ Nuclease and 4 iL Surveyor™ Nuclease Enhancer was prepared and incubated at 42 °C for 20 min. Then, the master mixture was aliquot into 8 tubes where each tube contained 5 iL of the master mixture, 400 units T4 DNA Ligase (1 μΐ,), and 0.6 μΐ_^ of 10x T4 DNA Ligase Buffer. Afterward, these tubes were incubated at either 16 °C or 25 °C for 30 min, 60 min and 120 min. The reaction was heated up to 95 °C for 5 min and cooled down to 50 °C with -1 °C/min and then to room temperature.

[00104] Digestion of circularized GFP«v DNA by Surveyor™ Nuclease:

Circularized DNA was purified from remaining single-stranded DNA (ssDNA) and from linear DNA simultaneously with Surveyor™ Nuclease digestion according to the following process. 6 μ∑ of the ligation product described above in T7E1 -based Error Filtering were incubated with 5 units Lambda Exonuclease, 20 units

Exonuclease I, 1 unit Surveyor™ Nuclease and 1 μΕ Surveyor™ Nuclease Enhancer in 10 μL for 10 min at 37 °C, for exonuclease cleanup of the ssDNA and linear DNA. Then, the temperature was increased to 42 °C for 20 min to enhance the Surveyor™ mismatch cleavage. Non-specifically nicked DNA was repaired in an additional incubation by adding in 600 units T4 DNA Ligase and 1 x T4 DNA Ligase buffer for 30 min at either 16 °C or 25 °C. The repaired DNA (ideally now only circularized GFPuv) was further amplified in a 20 μΐ reaction containing 1 iL product, 2 mM dNTP, 2 mM MgS0₄, 400 nM of primers F_Pr and R_Pr, 1 x KOD Hot Start buffer and 1 unit KOD Hot Start DNA Polymerase with 95 °C for 2 min, 20 cycles of 95 °C for 30 s, 55 °C for 30 s and 72 °C for 30 s, followed by 72 °C for 10 min.

[00105] 1.4 DNA characterization

[00106] Agarose gel electrophoresis: fn general, 5 μΐ of treated samples were analyzed by gel electrophoresis in either a 4% agarose gel (for 69-bp control DNA) or 1.5% agarose gel (NuSieve® GTG®, Cambrex Corporation) and 0.5* TBE buffer at 60 V for 60 min, containing either 0.1

ethidium bromide or post-stained with ethidium bromide. Either O'rangeRuler™ 5 bp or GeneRuler™ 100 bp Plus DNA Ladder (Fermentas, USA) was used as standard. Gel images were processed using Typhoon 9410 (Amersham Biosciences) and rmageQuant® 5.2.

[00107] Cloning and sequencing: Both of the Surveyor™ Nuclease and T7EI treated GFPwv genes were sequenced to check the error rate. For Surveyor™ Nuclease experiment, a 5 μΕ of Surveyor™ Nuclease treated GFPwv was incubated with 5 units Taq Polymerase, 10 mM dATP, lx ThermoPol buffer in 10 μϊ, at 72 °C for 10 min. For T7EI experiment, the restriction enzymes (BarnHI and EcoRI) digested full- length DNA contained overhang/underhang DNA, thus a T7EI treated GFPwv (5 μ&) was incubated likewise in the presence of 1 mM dNTP instead of 10 mM dATP. DNAs were cloned into One Shot® TOP 10 Competent R coli with TOPO TA cloning® Kit (Invitrogen) and spread out on LB plates containing 1 μg/mL ampicillin and 1.2 mg X-Gal Solution. In total, 50 white clones of each sample were picked and incubated in 6 mL LB medium containing 1 μg mL ampicillin at 37 °C overnight.

Plasmid DNA was extracted using Qiagen Mini preparation Kit (Qiagen), and sequenced by Aitbiotech PTE LTD, Singapore. All sequence results were analyzed using sequence analysis tool Vector NIT, and the errors were verified by visual confirmation of the electropherograms.

[00108] 2. COMPARISON OF MISMATCH-SPECIFIC ENZYMES WITH SHORT CONTROL DNA

[00109] Oligonucleotides of 69 bp with one of all possible mutation mismatches (AA, AG, AC, TG, CT, TT, GG and CC) or 1-, 2-, 3-, 5- and 10-base deletion mismatches (Id, 2d, 3d, 5d and lOd) were generated to study the cleavage activity of Surveyor™ Nuclease, T7E1, EndoV, T4E7 and MutS. Perfect match oligonucleotides (PM) were treated at the same condition with mismatch DNAs to study the non-specific cleavage of these enzymes.

[00110] Among all studied enzymes, the Surveyor™ Nuclease and T7EI were selected for continued study (see gel results shown in Figure 3). Surveyor™ showed high activity for all single-base mutations and multi-base deletion mismatches following a 1-hr treatment (Figure 3A). The 69-bp DNA was digested into short fragments of ~ 35 bp. T7E1 demonstrated the highest efficiency (Figure 3B) for single-base mutations and multi-base deletions, since only small amount of the 69-bp DNA remained after 4-hr incubation. The smear gel results of perfect match DNA indicated both these endonucleases exhibit non-specific cleavage activity.

[00111 ] Endo V (10 units) and T4E7 (500 units) both demonstrated no obvious digestion on all possible mismatches (Figures 3C, 3D) after 4-h treatment. A faint smear band with DNA sizes > 69 bp was observed for the Endo V treatment, indicating formation of a enzyme-substrate complex. The enzyme-substrate complex was stronger for the T4E7, as indicated by the disappearance of the 69-bp band and of a band with apparent DNA length of ~ 100 bp. The 30-bp difference corresponds to a molecular weight of 19.8 kDa (660 Da per bp), indicative of a DNA-T4E7 complex.

[001 2] A DNA-MutS complex formation was observed, as indicated by a second gel band with molecular weight > 69 bp, when the molar enzyme:DNA ratio was increased to 10:1 for DNAs with single-base mutation and 1-2 base deletions (Figure 3E). This second gel band was also observed with MutS treatment of perfect match DNA, suggesting non-specific binding of MutS. No obvious second gel band was observed for multi-deletion > 3 bp, which implies that the MutS may not effectively recognize sequences having multi-base deletion mismatches.

[00113] 3. T7EI-BASED ERROR FILTERING

[00114] Optimization of T7EI digestion with control long DNA: The effect of incubation time of T7EI (0.5 unit/μΐ) on a control DNA (632 bp) mixture containing 50% heteroduplex (25% of GG and CC mismatches at position 417) and 50% homoduplex DNAs was examined by varying the incubation from 0 min to 30 min at 37 °C (Figure 4). The gel results clearly indicate the capability of the mismatch digestion of T7E1. Control DNAs were digested into 4 smaller DNAs with lengths of 417 bp, 215 bp, -300 bp, and -360 bp. The first two bands (417 bp and 215 bp) correspond to the GG and CC mismatches, while the other two bands (~ 300 bp and - 360 bp) are unidentified. The band corresponding to full-length DNA decreased with prolonged incubation, indicating that the T7EI digestion progresses with longer incubation time. This implies that the T7EI may possess non-specific nicking and cleavage activity on homoduplex DNAs [12,13].

[00115] To repair the non-specific nicked DNA of T7EI, Ampligase™ (0.25 unit/μΐ) was added together with the T7EI and control DNA mixture, incubated at 37 °C for 25 min (Figure 5A). Ampligase™ DNA Ligase was chosen for its high efficiency at elevated temperature (37 °C). Control experiments were also conducted in either NEBuffer2 or Ampligase™ buffer with T7EI only. The gel results indicated that both of NEBuffer2 and Ampligase™ buffer are compatible with T7EI. No obvious gel intensity difference was observed for T7EI in these two buffers (Figure 5 A). In addition, more full-length DNA remained with Ampligase™ addition, indicating that the Ampligase™ can repair the non-specific nicked DNA.

[00116] To further verify the presence of nicked DNA, all three samples were denatured and re-annealed (Figure 5A, Re-hybridized) by heating the samples to 95°C for 5 min and then cooling to 50 °C with cooling rate of -1 °C/min and then to room temperature. This re-hybridization process frees the nicked DNA segments (single-stranded), and generates random DNAs of various lengths. As indicated by the gel results (the intensity of full-length band), the Ampligase™ repairs nicked DNA. However, some cleavage of nicked DNA by T7EI was still observed, as indicated by the smear of DNA shorter than full-length DNA (632 bp). The gels were significantly more smeared in control experiments without Ampligase™ (Figure 5A, lanes B2 and AB of Re-hybridization).

[00117] The re-hybridized samples were further treated with Lambda

Exonuclease and Exonuclease I, in order to clean up the truncated and T7E1 -cleaved DNAs. Both truncated and T7EI-cleaved DNAs contain an exposed 5'-phosphate group, and thus are substrates for Lambda exonulcease, while the full-length DNA bearing a 5'-hydroxyl group should not be a preferred substrate for Lambda exonuclease [14]. Interestingly, the exonuclease treatment degraded all samples

(Figure 5B). This implies that T7EI may have a putative 5' exonuclease activity which removes the 5 '-end of full-length DNAs, converting the end group from 5'- hydroxyl to 5 '-phosphate. Thus, it appears that Lambda Exonuclease recognized as a substrate the DNA originally having a 5'-hydroxyl group. This observation was confirmed with other synthetic DNAs including the promoter of human calcium binding protein A4 (S100A4, 752 bp), protein kinase B2 gene (1446 bp) and mOrange gene (711 bp, GenBank AY678265).

[00118] Optimization of T7EI digestion with GFPMV gene: GFPMV DNA was synthesized using the two-step gene synthesis process with primers containing either 5'-hydroxyl or 5'-biotin group. The DNAs were treated with T7EI only (37 °C, 25 min), and then subsequently incubated with Lambda exonuclease and Exonuclease I (37 °C, 5 min). A control experiment was also conducted with the same synthetic GFPMV treated only with the exonucleases (i.e. without T7EI digestion). The gel results indicate (Figure 6) that the 5'-hydroxyl and 5'-biotin groups both effectively protect the DNAs (lanes 4 and 8) from exonuclease digestion, with the quantity of full-length DNA similar to that of untreated DNA (lanes 1 and 5),. In addition, T7E1 effectively digests the synthetic GFPMV either through the non-specific nicking or mismatch cleavage (lanes 2 and 6). The 5'-biotin protected full-length DNA remains intact following T7EI and exonuclease treatment (lane 7), while the DNA with a 5'- hydroxyl terminus failed to withstand the 5' exonuclease of T7EI (lane 3). Based on these results, it appears that a protecting group such as the 5'-biotin-tag is able to prevent the undesired 5 '-end digestion of T7EI. [00119] As only the biotin-protected DNA withstood the exonuclease digest of T7EI, further studies were conducted with GFPuv synthesized with 5 '-biotin primers. Next, the effects of incubation temperature and Ampligase™ concentration were examined in order to balance the non-specific nicking of T7EI and ligation repair of Ampligase™. The manufacturer suggests the optimal temperature for T7EI is 37 °C while the Ampligase™ has better activity at temperature > 45°C. Figure 7 shows the gel results of the temperature effect, incubated at either 37°C or 45°C for 25 min. Nicked DNA was partially repaired by Ampligase™ at both temperatures of 37 °C and 45 °C as indicated by the higher intensity of full-length gel band than that without the ligase (Figure 7A). More full-length DNA remained after treatmen at 37°C than that at 45°C, even though Ampligase™ has improved activity at 45°C. One possibility is that non-specific nicking and cleavage by T7EI is promoted at 45°C, counteracting the Ampligase™ activity. The re-hybridized samples did not contain much more DNA of the desired length (as compared to non- Ampligase™ control), indicating that the quantity of Ampligase™ may not be enough to effectively repair all nicked sites (Figure 7B).

[00 20] The effect of increasing the Ampligase™ quantity is shown in Figure 8. The yield of full-length DNA was increased when the Ampligase™ was increase from 5 units (lane 1) to 15 units (lane 3), which is readily apparent following

.exonuclease treatment (lanes 5 and 6).

[00121 ] Thus, in the above experiments, most full-length DNA was obtained with 15 units of Ampligase™ incubated at 37 °C. Based on this result, digestion time was extended from 30 min to 40 min to investigate the stability and robustness of this protocol (Figure 9). The quantity of full-length DNA decreased slightly as the incubation was increased from 30 min to 40 min. To ensure the mismatch sites were fully cleaved, prolonged digestion of 40 min with 10 units of T7EI and 15 unit of Ampligase™ was adopted. Quantity of full-length DNA decreased significantly after the T7EI error filtering process, as indicated by the gel band intensity difference of lane 1 (original) and lane 7 (afterwards).

[00122] Summary for T7EI error filtering: The above experiments illustrate approaches to identify favourable conditions for mismatch cleavage by T7EI endonuclease. Side-effects of the enzyme can be partially corrected/avoided and desired DNA can be successfully obtained by addition of Lambda Exonuclease and Exonuclease I treatment. No visible by-products of truncated DNAs were observed in gel results. Since the DNA maybe protected by a biotin-tag at the 5' terminus, the products can subsequently be modified/prepared to allow for cloning and sequencing, and to confirm the fidelity of the synthesis product.

[00123] In a control experiment, the biotin-tag was successfully removed by restriction digestion using BamHI and EcoRI, following elimination of undesired

DNA by exonucleases. Restriction sites of BamHI and EcoRI are at positions 4 and 747 of GFPwv. The molecular process of the entire T7EI error filtering is illustrated in

Figure 10, including the following three steps: Step 1) mismatch cleavage and

nicking repair using T7EI and Ampligase™; Step 2) digestion of the non-full-length DNA with Lambda Exonuclease and Exonuclease I; and Step 3) removal of the 5'- biotin tag by EcoRI and BamHI restriction cleavage. Heat inactivation or solid-phase buffer exchange was also conducted between each step to inactivate or remove the

T7E1, exonucleases, and restriction enzymes (see also methods 1-3 in Table 6

below).

Table 6: Various methods for post-T7EI processing and preparation of GFPwv for

Procedure Method 1 Method 2 Method 3

1 T7EI + Ampligase 40 rain 40 min 40 min la Heat inactivation 95 °C; 30 min — — lb Buffer exchange — — —

2 λ Exo and Exo I 60 min 5 min 5 min

2a Heat inactivation 80 °C; 20 rrtin — 80 °C; 20 min

2b Buffer exchange ~ —

3 Restriction digest 60 min 60 min 60 min

3a Heat inactivation 80 °C; 20 min 80 ^"°C; 20 min —

3b Buffer exchange ~ ~ ✓

[00124] The gel results with heat inactivation (method 1 in Table 6) are shown in Figure 11 A. T7EI was successfully heat inactivated at 95°C within 30 min as indicated by comparison of full-length intensity before (lane 2) and after (lane 3) prolonged Lambda Exonuclease and Exonuclease I treatment (60 min at 37 °C). The exonucleases were also heat inactivated (80 °C for 20 min) before the restriction digestion. An added heat inactivation between each enzymatic process is very robust. The gel results indicated there is no observed loss of the full-length DNA after each procedure (lane 4). The smeared band with DNA length ~100 bp may represent primer-dimers, which would not affect the cloning reaction since these molecules are still protected by 5'-biotin-tags.

[00125] A reaction buffer exchange approach was also included (method 2 in Table 6) to avoid repetitive exposure of DNA to high temperature during multiple heat inactivation steps. Reaction buffer was exchanged using magnetic beads to extract the DNA, a wash step to remove impurities such as enzymes, and then elution of DNA into suitable buffer for a subsequent procedure (Figure 11B). In this approach, GFPwv was treated with T7EI and Ampligase™ for 40 min at 37 °C (lane 2), then treated directly with Lambda exonuclease and Exonuclease I, and incubated for 5 min at 37 °C. The exonuclease digestion step was reduced from 60 min (method 1) to 5 min (method 2) to limit non-specific cleavage of T7EI, which may still be active in method 2. Subsequently, the reaction mixture was purified by buffer exchange to remove T7EI and exonucleases (lane 3), followed by biotin-tag removal (lane 4). The quantity of full-length DNA slightly decreased, possibly due to loss resulting from the buffer exchange steps, which has extraction efficiency of ~ 80%.

[00126] In the combination approach (method 3 in Table 6), the buffer exchange of method 2 following exonuclease treatment was replaced with a heat inactivation (80 °C for 20 min) (Figure 12, lane 3), with an additional buffer exchange after the restriction digestion (lane 4). This approach decreases the yield of full-length DNA in comparison to methods 1 and 2. Thus, method 1 with heat inactivation between each process step results in higher yield than that observed for methods with a corresponding buffer exchange step.

[00127] These results demonstrate use of T7EI to remove mismatch DNA sequences from a pool of synthetic DNA. The T7EI cleaves the mismatch DNA, but it may also produce undesired side-effects of non-specific cleavage and 5' exonuclease digestion. To limit the non-specific cleavage activity, a ligase such as Ampligase™ may be added with the T7EI to repair the non-specifically nicked DNA. To protect the 5' terminus of full-length DNA from undesired exonuclease activity, a 5' protecting group such as a 5'-biotin may be used. Additionally, exonucleases such as Lambda Exonuclease and Exonuclease I may be employed to digest unwanted DNA fragments, and restriction digestion may be used to deprotect DNA in preparation for subsequent uses such as cloning and sequencing. Optional approaches such as heat inactivation and/or buffer exchange may be used.

[00128] 4. SURVEYOR™ NUCLEASE-BASED ERROR FILTER

[00129] Repair of DNA nicked by Surveyor™ Nuclease: Similar to T7EL Surveyor™ Nuclease also has non-specific nicking property. Thus, the effectiveness of ligase enzyme in repairing the nicked DNA was examined using control DNA. A series of experiments were conducted with: 1) Surveyor™ treatment (without ligase) for a total of 50 min (20 min at 42 °C + 30 min at 45°C); 2) simultaneous Surveyor™ and 9°N™ DNA Ligase (a thermostable ligase) treatment for 30-60 min at 42°C- 45 °C; and 3) consecutive endonuclease and ligase treatments, in which control DNA was first treated with Surveyor™ for 20 min at 42°C, and then the 9°N™ DNA Ligase was added and further incubated for 30-60 min at 45°C (Figure 13).

[00130] As seen in Figure 13, Surveyor™ effectively digested the control DNA into two smaller DNAs (417 bp and 215 bp) with limited non-specific cleavage even after 50 min incubation (lane 2). The smeared band of re-hybridized Surveyor™ treated DNA (lane 3) indicated the Surveyor™ does possess non-specific nicking activity. The full-length gel band may contain randomly nicked DNAs which are still hybridized at 37°C, but which become denatured and randomized during the re- hybridization step. The 9°N™ DNA Ligase failed to repair nicked DNAs in either the simultaneous reaction (lanes 5 and 7), or in the consecutive reaction (lanes 9 and 11). The re-hybridization of all treated samples contained DNAs of random lengths (lanes 3, 5, 7, 9, 11).

[00131] To repair the non-specific nicking of Surveyor™, a low temperature ligase was tested (T4 DNA ligase). Optimal reaction temperatures for T4 DNA ligase and Surveyor™ are described as 16°C and 42°C, respectively. Thus, to balance the activities of ligase and mistmatch digestion, Surveyor™ activity was examined at various temperatures ranging from 16°C to 85°C using control DNA as substrate (Figure 14). Control DNA was incubated with Surveyor™ at 42 °C for 20 min, followed by a further incubation at various temperatures of 16°C - 85°C for 30 min. As indicated by the results in Figure 14 (the gel band intensities), Surveyor™ displays effective mismatch digestion with limited non-specific cleavage at temperature of 16°C to 42°C. The non-specific nicking and cleavage increases at temperature > 55°C.

[00132] Next, control DNA was treated with Surveyor™ Nuclease at 42 °C for 20 min and T4 DNA ligase was subsequently added and incubated at either 16 °C or 25 °C for another 30-120 min (Figure 15). Samples were then denatured and re- hybridized to verify the repair of nicked DNA. A control experiment was also performed with only Surveyor™ treatment (42 °C for 20 min) and DNA re- hybridization (lane S). The intact full-length bands (L30-L120) clearly indicate the efficient mismatch digestion capability of the Surveyor™/T4 DNA ligase

combination. Without the T4 DNA ligase, no full-length DNA was observed in the re- hybridized DNA mixture (lane S). The gel results of re-hybridized DNA mixture is smeared, implying that the Surveyor™ may have high nicking activity, and the Surveyor™ may not recognize the nicking site as mismatch such that the nicked DNA is not fully cleaved.

[00133] Digestion of circularized DNA: Surveyor™ endonuclease has strong 5' exonuclease activity and appears to digest even DNA having a 5'-biotin protected terminus. Thus, to protect the perfect match DNA from exonuclease digestion, an approach to circularize synthetic DNA was used, as illustrated in Figure 16. Linear DNA would have an exposed 5'-phosphorylated terminus and is thus subjected to Lambda Exonuclease digestion. Figure 17 illustrates the results of Surveyor™

Nuclease mismatch removal using circularized DNA. The gel results of GFPwv corresponding to the each step depicted in Figure 16 are shown in Figure 17. The predominant product after DNA circularization was full-length DNA (lane 1). The faint smear with molecular sizes larger than the full-length DNA suggests the mixture may also contain end-to-end extended linear DNAs. No detectable band was observed following co-treatment with Lambda Exonuclease, Surveyor™ Nuclease and T4 DNA ligase at either 25 °C (lane 3) or 16 °C (lane 5), which may indicate low efficiency of the circularization process. PGR amplification was conducted in order to amplify circularized DNA. The bands corresponding to full-length DNA (lanes 4 and 6) indicate the presence of circularized DNA. The PGR products were cloned directly without further purification using PCR^®2.1-TOPO^® cloning vector (Invitrogen). In total, 50 clones were sequenced for the error rate calculation. Surprisingly, the error rate (10 errors/kbp) was lower than that of untreated GFPwv (4.11 errors/kbp). One possibility for this result is low efficiency of DNA circularization. The full-length DNAs (lanes 3 and 5) may result from re-assembly of non-fully digestion linear DNA segments in the reaction mixture, given the high efficiency of PGR amplification.

[00134] All publications and patent applications cited in this specification are herein incorporated by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. The citation of any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention.

[00135] As used in this specification and the appended claims, the singular forms "a", "an" and "the" include plural reference unless the context clearly dictates otherwise. As used in this specification and the appended claims, the terms

"comprise", "comprising", "comprises" and other forms of these terms are intended in the non-limiting inclusive sense, that is, to include particular recited elements or components without excluding any other element or component. Unless defined otherwise all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this invention belongs.

[00136] All lists and/or ranges provided herein are intended to include any sub- list and/or narrower range falling within the recited list and/or range.

[00137] Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it is readily apparent to those of ordinary skill in the art in light of the teachings of this invention that certain changes and modifications maybe made thereto without departing from the spirit or scope of the appended claims. References

1. Binkowski,B.F., Richmond,K.E.₅ Kaysen,J., Sussman,M.R. and Belshaw,PJ.

(2005) Correcting errors in synthetic DNA through consensus shuffling. Nucleic Acids Res., 33, e55.

2. Carr,P.A., Park,J.S., Lee,Y.J., Yu,T., Zhang,S. and JacobsonJ.M. (2004) Protein- mediated error correction for de novo DNA synthesis. Nucleic Acids Res., 32, el62.

3. Tian,J., Gong,H., Sheng,N., Zhou,X., Gulari,E. Gao,X. and Church,G. (2004) Accurate multiplex gene synthesis from programmable DNA microchips. Nature, 432, 1050-1054.

4. Fuhrmann,M., Oertel,W., Berthold,P., Hegemann,P. (2005). Removal of mismatched bases from synthetic genes by enzymatic mismatch cleavage. Nucleic Acids Res., 33, e58.

5. Bang, D., and Church, G.M., (2008). Gene synthesis by circular assembly amplification. Nature Methods, 5, 37-39.

6. Brown, J., Brown, T., and Fox, K.R., (2001) Affinity of mismatch-binding protein MutS for heteroduplexes containing different mismatches. Biochem. J., 354, 627- 633.

. Biswas, L, and Hsieh, P., (1997) Interaction of MutS Protein with the Major and Minor Grooves of a Heteroduplex DNA. J. Biol. Chem., 272, 13355-13364.

. Selmane, T., Schofield, M. J., Nayak, S., Du, C, and Hsieh, P., (2003) Formation of a DNA mismatch repair complex mediated by ATP. J. Mol. Biol, 334, 949- 965.

. Pincas, H., Pingle, M. R, Huang, J., Lao, K., Paty, P. B., Friedman, A. M., and Barany, F., (2004) High sensitivity EndoV mutation scanning through real-time ligase proofreading. Nucleic Acids Res., 32, el48.

10. Youil, R., Kemper, B. W., and Cotton, R. G., (1995) Screening for mutations by enzyme mismatch cleavage with T4 endonuclease VII. Proc. Natl.Acad. Sci. USA, 92, 87-91.

1. Babon, J. J., McKenzie, M., and Cotton, R. G., (1999) Mutation detection using fluorescent enzyme mismatch cleavage with T4 endonuclease VII. Electrophoresis, 20, 1162-1170.

2. Guan, C, Kumar, S., Kucera, R., Ewel, A., (2004) Changing the enzymatic activity of T7 endonuclease I by mutations at the beta-bridge site: alteration of substrate specificity profile and metal ion requirements by mutation distant from the catalytic domain. Biochemistry, 43, 4313-4322.

3. Tsuji, T., and Niida, Y., (2008) Development of a simple and highly sensitive mutation screening system by enzyme mismatch cleavage with optimized conditions for standard laboratories. Electrophoresis, 29, 1473-1483.

4. Subramanian, K, Rutvisuttinunt,W., Scott,W., and Myers,R.S., (2003) The Enzymeatic basis of processivity in λ exonuclease. Nucleic Acids Research, 31, 1585-1596.

5. Rarany,F. (1991) Genetic disease detection and DNA amplification using cloned thermostable ligase. Proc. Natl. Acad. Sci., 88, 189-193.

6. Nilsson,S. and Magnusson,G. (1982). Sealing of gaps in duplex DNA by T4 DNA ligase. Nucleic Acids Research, 10, 1425- 1437. Qiu,P., Shandilya,H.₅ D'Alessio,J.M., 0'Connor,K., Durocher,J. and Gerard,G.F. (2004). Mutation detection using Surveyor™™ nuclease. BioTechniques, 36, 702- 707.

Li, J., Berbeco,R., Distel,R.J., Janne, P. A., Wang, L., and Makrigiorgos, G.M., (2007). S-RT-MELT for rapid mutation scanning using enzymatic selection and real time DNA-melting: new potential for multiplex genetic analysis. Nucleic Acids Res., 35, e84.

Cross, M. J., Waters, D. L. E., Lee, L. S. and Henry, R. J., (2008) Endonucleolytic Mutation Analysis by Internal Labeling (EMAIL), Electrophoresis, 29, 1291- 1301.

Tsuji, T. and Niida, Y., (2008) Development of a simple and highly sensitive mutation screening system by enzyme mismatch cleavage with optimized conditions for standard laboratories. Electrophoresis, 29, 1473-1483.

Claims

WHAT IS CLAIMED IS:

1. A method comprising, in sequence: treating a pool of double-stranded DNA molecules with a mismatch endonuclease and a DNA ligase, the pool of double-stranded DNA molecules comprising 5 '-protected double-stranded DNA molecules; and treating the pool of double-stranded DNA molecules with one or more exonucleases, the one or more exonucleases together having 5' to 3' activity on double-stranded DNA and 3' to 5' activity on single-stranded DNA.

2. The method of claim 1, wherein the 5 '-protected double-stranded DNA

molecules have a protecting group attached at each 5' end.

3. The method of claim 2, wherein the protecting group is a biotin group, a locked nucleic acid group or a peptide nucleic acid group.

4. The method of claim 1 , wherein the 5 '-protected double-stranded DNA

molecules are circular.

5. The method of any one of claims 1 to 4, wherein the mismatch endonuclease is endonuclease V, T7 endonuclease I, Surveyor nuclease, T4 endonuclease VII, or CEL I nuclease.

6. The method of claim 5, wherein the mismatch endonuclease is T7

endonuclease I.

7. The method of claim 5, wherein the mismatch endonuclease is Surveyor nuclease.

8. The method of any one of claims 1 to 7, wherein the DNA ligase is

Ampligase, 9°N ligase, T4 DNA ligase, Taq DNA ligase, or E. coli DNA ligase.

9. The method of claim 8, wherein the DNA ligase is Ampligase.

10. The method of claim 8, wherein the DNA ligase is T4 DNA ligase.

11. The method of any one of claims 1 to 9, wherein the one or more exonucleases comprise one of Lambda exonuclease, RecJ and T7 exonuclease and one of exonuclease I and exonuclease T.

12. The method of claim 11, wherein the one or more exonucleases comprise Lambda exonuclease and exonuclease I.

13. The method of any one of claims 1 to 12, wherein the pool of double-stranded DNA molecules is treated with the mismatch endonuclease prior to treatment with the DNA ligase.

14. The method of any one of claims 1 to 12, wherein the pool of double-stranded DNA molecules is treated with the mismatch endonuclease together with the DNA ligase.

15. The method of any one of claims 1 to 4, wherein the mismatch endonuclease is T7 endonuclease I, the DNA ligase is Ampligase and the one or more exonucleases comprise Lambda exonuclease and exonuclease I, wherein the pool of double-stranded DNA molecules is treated with the mismatch endonuclease together with the DNA ligase.

16. The method of any one of claims 1 to 4, wherein the mismatch endonuclease is Surveyor, the DNA ligase is T4 DNA ligase and the one or more

exonucleases comprise Lambda exonuclease and exonuclease I, wherein the the pool of double-stranded DNA molecules is treated with the mismatch endonuclease prior to treatment with the DNA ligase.

17. A kit comprising a mismatch endonuclease, a DNA ligase and one or more exonucleases together having combined 5' to 3' activity on double-stranded DNA and 3' to 5' activity on single-stranded DNA.