US20100167942A1

US20100167942A1 - Compositions, Methods and Related Uses for Cleaving Modified DNA

Info

Publication number: US20100167942A1
Application number: US12/644,666
Authority: US
Inventors: Yu Zheng; Richard J. Roberts
Original assignee: New England Biolabs Inc
Current assignee: New England Biolabs Inc
Priority date: 2008-12-23
Filing date: 2009-12-22
Publication date: 2010-07-01
Also published as: CN102264900A; US20120202712A1; EP2376632A1; US9200316B2; US20140106350A1; US8653007B2; US9029087B2; JP2013514758A; US20140194299A1; EP2376632B1; WO2010075375A1

Abstract

Compositions, methods and related uses are provided relating to cleaving modified DNA. For example, a set of DNA fragments obtainable by enzymatic cleavage of a large DNA is described where at least 50% are similarly sized and have a centrally positioned modified nucleotide. In addition, an enzyme preparation is provided that includes one or more enzymes that recognize a modified nucleotide in a DNA and cleave the DNA at a site that is at a non-random distance from the modified nucleotide. The one or more enzymes are further characterized by an N-terminal conserved domain with greater than 90% amino acid sequence homology to WXD(X)₁₀YXGD. The related uses include creating a methylome, methods of purifying DNA fragments containing a modified nucleotide and diagnostic applications.

Description

CROSS REFERENCE

This application claims priority from U.S. Provisional Applications Ser. No. 61/140,586 filed Dec. 23, 2008 and Ser. No. 61/267,617 filed Dec. 8, 2009, herein incorporated by reference.

BACKGROUND

DNA cleaving enzymes associated with methyltransferases are widely present in the prokaryotic genomes. The DNA cleaving enzymes typically consist of restriction endonucleases, which protect host cells from invading DNA (e.g., bacteriophages) by cleaving DNA at defined sites, and DNA methyltransferases, which protect host DNA from being degraded by methylating a specific base within the restriction endonuclease sites (Roberts, et al. Nucleic Acids Res 35: D269-270 (2007)). Hence, these restriction endonucleases are termed methylation-sensitive.
While modified bases in prokaryotes and phage DNA play a role in protecting the genome against cleavage by restriction endonucleases, methylated cytosine (m5C) is involved in gene expression of the mammalian genome. Techniques for identifying methylated DNA are cumbersome and experimentally difficult to implement in a reproducible fashion. Two approaches are commonly used. One involves the use of restriction enzymes like HpaII and MspI which are differently sensitive to cytosine methylation. For example, HpaII endonuclease is blocked by methylation of either of the two cytosines within the CCGG recognition site, but its isoschozimer, MspI, is blocked only when the outer C is methylated. It will cleave DNA when the inner cytosine is modified. The second method involves bisulfite modification of the unmethylated cytosine residues followed by selective amplification and sequencing of the remaining DNA. In this method, methylated cytosines are resistant to the treatment. This method is not easy to optimize and involves a complicated chemical modification step followed by amplification using a complicated set of primers. The method is widely used in the absence of simpler alternative approaches.

SUMMARY

In an embodiment of the invention, a set of double-stranded oligonucleotide fragments are provided that are obtainable by enzymatic cleavage of a large DNA wherein the large DNA contains one or more modified nucleotides and may be derived from mammalian cells, more specifically human cells. At least 50% of the fragments in the set should preferably be of a similar size and preferably contain a centrally positioned modified nucleotide. One or more fragments may be isolated from the set. The large DNA may be at least 100 nucleotides in length; the modified nucleotide is, for example, a modified cytosine such as a methylated cytosine or a hydroxymethylated cytosine and a modified cytosine may be proximate to a guanine to form a CpG or a CNG. However, a modified cytosine may be alternatively located next to another cytosine, an adenine or a thymidine. Oligonucleotide fragments in the set may preferably be less than 60 nucleotides long, for example 28-36 nucleotides; and/or the modified nucleotide, in particular, cytosine may be located within 30 nucleotides from either end of the fragment.
In an embodiment of the invention, an enzyme preparation is provided that is characterized by one or more enzymes that recognize a modified nucleotide in a DNA such that each enzyme is capable of cleaving the DNA at a site that is a non-random distance from the modified nucleotide. More particularly, the non-random distance between the cleavage site and the modified nucleotide may be characteristic for the enzyme so as to generate a set of fragments of the type described above. The one or more enzymes are further characterized by an N-terminal conserved domain with greater than 90% amino acid sequence homology with WXD(X)₁₀YXGD, more particularly with greater than 90% amino acid sequence homology with WXD(X)₆G(X)₃YXGD(X)_10-15GN(X)₂L X_10-20PX₃F.
In an embodiment of the invention, the one or more enzymes in the enzyme preparation are further defined by a recognition domain and a cleavage domain within a single open reading frame. The cleavage domain may have an amino acid sequence which has greater than 90% amino acid sequence homology to FEX_20-30DX_2-4DX_19-22(Q/E)XK. In addition, at least one of the enzymes may have an amino acid sequence homology of greater than 90% to any of the sequences identified as SEQ ID NOS:7-22. Additionally, one or more of the enzymes may be covalently or non-covalently linked or fused to a protein affinity tag or other tag. Examples of suitable affinity tags include a chitin-binding domain, maltose-binding domain, an antibody and a His tag. In addition, the one or more enzymes may be recognized by an antibody with binding specificity for an amino acid sequence comprising WXD(X)₁₀YXGD. Additionally, the preparation may include an activator DNA.
In an embodiment of the invention, an enzyme preparation is provided that includes one or more enzymes that recognize a modified nucleotide in a DNA such that each enzyme is capable of cleaving the double-stranded DNA at a site that is at a non-random distance from the modified nucleotide, more particularly where the distance between the cleavage site and the modified nucleotide is characteristic for the enzyme, thereby generating a set of fragments. The set of fragments may be of similar size if the DNA contains a modified nucleotide on each strand of the duplex at approximately opposing positions or may be of varying size for hemi-modified DNA. The one or more enzymes may be further characterized by an N-terminal conserved domain with greater than 90% amino acid sequence homology to WXD(X)₁₀YXGD.
In an embodiment of the invention, an antibody is provided that is capable of recognizing and binding to an N-terminal domain of an enzyme described above.
In an embodiment of the invention, a method is provided which comprises cleaving a large DNA containing one or more modified nucleotides with a composition described above and obtaining a set of oligonucleotide fragments. The method may further include separating the set of oligonucleotide fragments from uncleaved DNA and additionally may include sequencing from the separated set of fragments at least one fragment to determine the location of one or more modified nucleotides contained within at least one fragment. The method may include analyzing some or all of the oligonucleotide fragments for the presence and location of one or more modified nucleotides in the large DNA by sequencing or other means and mapping the sequences onto a genome or methylome map to determine the location of modified nucleotides.
In an embodiment of the invention, a method is provided for identifying an enzyme such as described above that includes searching a sequence database using a sequence selected from the group consisting of SEQ ID NO:7-22 and variants thereof, and identifying additional sequences having an N-terminal region characterized by a consensus sequence of WXD(X)₁₀YXGD. The method may include the further step of identifying a C-terminal end comprising a catalytic domain with a consensus sequence of FEX_20-30DX_2-4DX_19-22(Q/E)XK, more particularly FE(X)₂A(X)_15-18T/SX₄DGGXDX₂G/LX_15-20E/QAK.
In an embodiment of the invention, a method is provided for isolating from a mixture of DNA fragments those DNA fragments containing one or more modified nucleotides, the mixture resulting from enzyme cleavage of a large DNA containing at least one modified nucleotide. The method may include adding to the mixture an immobilized or labeled affinity-binding molecule that is capable of binding selectively those fragments containing a modified nucleotide. Alternatively, those fragments containing a modified nucleotide may be size-separated from those fragments that do not contain a modified nucleotide. An example of an affinity-binding molecule is an enzyme preparation described above, wherein the one or more enzymes in the enzyme preparation have been mutated so as to lack enzyme cleavage activity and wherein the mutated enzyme is immobilized on a solid surface so as to bind the DNA fragments containing one or more modified nucleotides. Other examples of affinity-binding molecules include antibodies, inactivated T4 glucosyltransferase and the methyl-binding domain of a cell protein such as DNMT1. These molecules may in turn be fused to any of a chitin-binding domain, a maltose-binding domain or a biotin molecule for example and hence, bind to a suitable column.
In another embodiment of the invention, a method is provided for identifying a present or future phenotypic property in a cell preparation or tissue sample from a pattern of modified nucleotides. The method includes cleaving into fragments a large DNA from a cell preparation or tissue by means of an enzyme preparation described above; and comparing a location for modified nucleotides in the fragments with a pattern of modified nucleotides in a control DNA so as to determine a present or future phenotypic property.
In another embodiment of the invention, the above method further comprises contacting the cleavage fragments with an affinity-binding molecule capable of binding the modified nucleotide or by means of electrophoresis or other means known in the art capable of effecting size separation. The binding moiety may include an enzyme preparation as described above, wherein the enzyme cleavage activity has been inactivated by conventional means. Thus, fragments with a modified nucleotide may be separated from fragments lacking a modified nucleotide. The above method may additionally include identifying on a methylome or a genome a location for the one or more modified nucleotides in the immobilized cleavage fragments. The location may be determined by sequencing the separated fragments.
In another embodiment of the invention, a method is provided for determining the location of at least one modified nucleotide in a large DNA. The method includes: cleaving a large DNA with an enzyme preparation described above; obtaining a set of oligonucleotide cleavage products containing at least one modified nucleotide; and determining the location of the at least one modified nucleotide in a sequence of the large DNA by for example sequencing the set of oligonucleotide cleavage products. The number of oligonucleotide fragments for sequencing in the set may depend on whether the set is derived from cloned DNA or from repeats in which it may be sufficient to sequence a subset of fragments or on whether the set is expected to contain unique sequences in which it may be desirable to sequence substantially all the fragments in the set.
In an embodiment of the invention, a method is provided for obtaining a purified preparation of fragments containing one or more modified nucleotides that includes contacting a mixture of DNA fragments in which one or more the fragments contain at least one modified nucleotide with an immobilized affinity-binding protein capable of binding covalently or non-covalently to the DNA fragment. An example of an affinity-binding protein is a mutated enzyme in the enzyme preparation described above, wherein the enzyme cleavage activity has been inactivated. The method may further include binding the one or more fragments containing at least one modified nucleotide to the binding protein; and obtaining a purified preparation of fragments containing one or more modified nucleotides.
In an embodiment of the invention, a kit for generating oligonucleotide fragments containing a modified nucleotide is provided that includes an enzyme preparation described above in a container with instructions for use. The kit may further include an activator molecule.

BRIEF DESCRIPTION OF THE FIGURES

FIGS. 1A-1B show the MspJI RM system.

FIG. 1A shows the genomic segment of Mycobacterium sp. JLS encoding the MspJI RM system. NCBI annotations for open reading frames are: MjIs0821, putative helicase; MjIs0822 (MspJI), restriction endonuclease; MjIs0823 (V.MspJIP), DNA mismatch endonuclease vsr; MjIs0824 (M.MspJI), DNA cytosine methyltransferase.

FIG. 1B shows the schematic domain structure of the MspJI enzyme family. The N-terminal domain is defined here as about 50% of the protein sequence upstream of the remaining C-terminal domain.

FIG. 1C shows the conserved motif in the N-terminal domain.

FIG. 1D shows the conserved motif in the C-terminal domain.

FIG. 2A shows modification-dependent enzyme activity for MspJI.

Lane

1, 1 μg of pBR322(dcm+) DNA only;

Lane

2, 1 μg of pBR322(dcm+)+0.8 μg MspJI;

Lane

3, 1 μg of pBR322(dcm+)+0.8 μg MspJI+10 units of BstNI;

Lane

4, 1 μg of pBR322(dcm+)+10 units of BstNI only;

Lane

5, 1 μg of pBR322(dcm−) DNA only; and

Lane

6, 1 μg of pBR322(dcm−)+0.8 μg MspJI.

All reactions were incubated at 37° C. for 1 hour and resolved on a 1% agarose gel.

Lanes

3 and 4 show that MspJI does not cut any sites not cut by BstNI. Thus, on this substrate which is methylated at CmC A/T GG, only modified Dcm sites are cut by MspJI. All Dcm sites are cut by BstNI which is not sensitive to Dcm methylation.

FIG. 2B shows MspJI digestion on pBR322(dcm−) methylated at other sites. All reactions were done in 50 μl volume at 37° C. for 2 hours and resolved on 1% agarose gel. 0.8 μg of MspJI were used in all reactions.

Lane

1, 1 μg of M.MspI (C^mCGG) modified pBR322;

Lane

2, 1 μg of M.HaeIII (GG^mCC) modified pBR322;

Lane

3, 1 μg of M.HpaII (C^mCGG) modified pBR322;

Lane

4, 1 μg of M.HhaI (G^mCGC) modified pBR322; and

Lane

5, 1 μg of M.AluI (AG^mCT) modified pBR322.

FIG. 2C shows MspJI digestion on hydroxymethylcytosine-containing DNA. All reactions were carried out in 50 μl volume at 37° C. for 1 hour.

Lane

1, 1 μg of T4 wild-type (wt) DNA with glucosylated hydroxymethylcytosine;

Lane

2, 1 μg of T4 gt DNA with hydroxymethylcytosine;

Lane

3, 1 μg of T4 wt DNA+10 units of McrBC;

Lane

4, 1 μg of T4 gt DNA+10 units of McrBC;

Lane

5, 1 μg of T4 wt DNA+20 units of MspI;

Lane

6, 1 μg of T4 gt DNA+20 units of MspI;

Lane

7, 1 μg of T4 wt DNA+0.8 μg of MspJI; and

Lanes 8 through 11, 1 μg of T4 gt DNA with 2-fold serially-diluted MspJI starting at 1.6 μg (Lane 8).

FIG. 2D shows the effect of different amounts of a DNA activator on MspJI activity. From Lanes 1 to 5, each reaction contains 1 μg (0.35 pmol) pBR322 and 1.6 pmol MspJI. Lanes 1-4 show a titration (40, 20, 10, 5 pmol) of the DNA activator containing methylated CCWGG sites. Lane 5 shows pBR322-digestion using MspJI without DNA activator. Lane 6 shows pBR322-digestion using BstNI (CCWGG).

FIG. 3 shows a schematic diagram of MspJI's cleavage activity on fully-methylated DNA. The double-stranded cleavage can happen at either side of the methylated site. The cleavage is on the 3′ side of the recognized methylated base. In this figure, when the top strand methylated cytosine is recognized, MspJI cleaves on the right side; when the bottom strand methylated cytosine is recognized, MspJI cleaves on the left side. The distances from the cleavage sites to the recognized methylated cytosine are fixed. For example, when top strand methylated cytosine is recognized, the bottom strand nicking site is 16 nucleotides away from it and the top strand nicking site is 12 nucleotides away from it.

FIGS. 4A-4D show MspJI cleavage on fully-methylated and hemi-methylated oligo substrates, fractionated on a denaturing gel.

FIG. 4A shows the expected enzyme cleavage sites (designated Rt, Rb, Lt, Lb) in a synthetic double-stranded oligonucleotide TGGTAATAATAAGGTTGAGGACTTTTTCCGGATGCCCGGAATGGGTTCAAA GG (SEQ ID NO:1). The 3′ end of the top strand or the 5′ end of the bottom strand is labeled with FAM as indicated in 4B.

FIG. 4B shows MspJI-digestion on fully and hemi-methylated oligo substrates.

Lane 1, no methylation, top strand labeled;

Lane 2, no methylation, bottom strand labeled;

Lane 3, both top and bottom methylated, top strand labeled;

Lane 4, both top and bottom strand methylated, bottom strand labeled; cleaved products with sizes of 8 nt and 7 nt suggest wobbling cuts in the bottom strand;

Lane 5, top strand methylated, top strand labeled;

Lane 6, top strand methylated, bottom strand labeled; as in Lane 4, cleaved products with sizes of 8 nt and 7 nt suggest wobbling cuts in the bottom strand;

Lane 7, bottom strand methylated, top strand labeled; and

Lane 8, bottom strand methylated, bottom strand labeled.

As a control, markers are run on the right side of the gel.

FIG. 4C shows an oligonucleotide sequence having a CpG and enzymatic cleavage sites that would yield a fragment with the CpG in a central location.

FIG. 4D shows digestion of the oligonucleotide shown in FIG. 4C using MspJI in the presence or absence of activator DNA. Reactions were done in 10 μl at 37° C. for 1 hour. The oligonucleotide is 1 pmol and MspJI is 0.4 μg in each reaction. In reactions with activator, 1 μl of stock (15 μM) was added into the 10 μl reaction. 5 μl of reactions were taken out and stopped at different time points and resolved on 20% native polyacrylamide gel.

Lane 1, DNA only;

Lane 2, digestion reaction without activator at 30 min;

Lane 3, digestion reaction with activator at 30 min;

Lane 4, digestion reaction without activator at 1.5 hour; and

Lane 5, digestion reaction with activator at 1.5 hour.

FIG. 5 shows the sequence analysis of MspJI cleavage site positions on different m5C methylated sites. pBR322 DNA (dcm−) was methylated using various methyltransferases. The methyltransferases are shown under the column heading “methylase”. The run-off sequencing pattern is shown under the column “examples of sequencing chromatogram” (SEQ ID NOS:2-6). The deduced cleavage patterns are shown in column “MspJI cleavage site”.

FIG. 6A shows a double-stranded DNA with a centrally located modified cytosine, which is a representative fragment of the set of oligonucleotides obtained when a large DNA is cleaved with a member of the novel enzyme family.

FIG. 6B shows a DNA sample of human genomic DNA on a polyacrylamide gel in which the set of oligonucleotide fragments shown in FIG. 6A appear as a coherent band. The 32 bp band represents the pool of short fragments containing methylated CpG sites from the genome. This pool can be purified and directly put into the Next-Generation sequencing platforms for methylome analysis.

FIG. 6C shows a comparison of cleavage products generated by various member enzymes in the MspJI family. The first lane contains DNA markers. All subsequent lanes contain the digests of CpG-methyated Hela genomic DNA, each with a different MspJI family member enzyme. Lanes 1-3 and 5 show a band that corresponds to about 32 nucleotides in length (arrow). In lane 4, RlaI recognizes CCWGG and cleaves on either side of the recognition site to provide a centrally located modified cytosine. However, this sequence is not found in Hela genomic DNA.

Lane 1 shows the cleavage product from MspJI;

Lane 2 shows the cleavage product from Frankia 5336;

Lane 3 shows the cleavage product from Lpg 1234;

Lane 4 shows the cleavage product from RlaI;

Lane 5 shows the cleavage product from AspBHI; and

Lane 6 shows DNA only.

FIGS. 7-1 to 7-7 shows the amino acid sequence alignment of representative members of the MspJI family of enzymes. Residues conserved in 5 or more members of the alignment are indicated in the top line (“Conservation”). Secondary structure prediction is listed at the bottom (“Consensus_ss”). Secondary structural elements are: e, β sheet; h, α helix.

FIGS. 8-1 to 8-2 shows a bioinformatics analysis of CG-centered sequences of different lengths in three organisms. Total number of CG sites are listed for each organism for human, mouse and Arabidopsis genomes. Columns report the number of CG-centered sequences that are distinct (differ in sequence), the number that are unique (occur in single copy), the fraction of the total such CG-centered sequences that are unique and the fraction of distinct CG-centered sequences that are unique (single copy).

DETAILED DESCRIPTION OF THE EMBODIMENTS

A novel family of modification-specific DNA cleavage enzymes has been found where the members of the family recognize a modified nucleotide in double-stranded DNA and then cut at a non-random distance downstream (3′ direction) from the modified nucleotide. One of the unique properties of these enzymes is that they are capable of releasing short DNA fragments containing a modified nucleotide directly from large DNA including genomic DNA. These enzymes are capable of generating double-stranded breaks in the DNA on both strands when a modified nucleotide is present in opposing positions on each strand. When DNA contains a modified nucleotide on one strand only, double strand breaks occur on one side of the modified nucleotide. The location of the modified nucleotide in a large DNA can thus be deduced by cloning the cleavage products and/or by sequencing. Using ultra high throughput sequencing platforms, it is possible to identify and map modified nucleotides such as methylated cytosines or hydroxyl-methylated cytosines in a reliable and quick manner.
“Modified” nucleotide is intended to refer to any nucleotides that contain an extra chemical group such as a 5-hydroxymethyl or 5-methyl group. For example, a “modified” cytosine generally arises in mammalian genomes as a CpG and in plant genomes as a CNG which, because of the symmetry, make possible methylation on both strands at the same position. Hydroxymethylcytosine has been recognized as a constituent of human DNA (Tahiliani et al., Science 324(5929): 930-5 (2009); Kriaucionis and Heintz, Science 324(5929): 929-30 (2009)).
“Large DNA” is intended to refer to any naturally occurring or synthetic DNA having a size greater than 100 nucleotides up to a size of a genome.
“Similar size” with reference to a “set” of oligonucleotide fragments is intended to refer to fragments that vary no more than about ±5 nucleotides in length. However, different “sets” of fragments may have a size range of 5-50 nucleotides.
“Centrally positioned” is intended to correspond to a location of a modified nucleotide on one strand which is approximately centered in the same strand of a double-stranded fragment. The location is generally within 5 nucleotides of the center determined by counting the nucleotides from either end of the fragment.
“N-terminal domain” refers to a region extending to about 50% of the amino acid sequence of the protein. In an embodiment of the invention, a conserved region within the N-terminal domain corresponds to amino acids 81 to 224 of SEQ ID:NO:22 (MspJI) and a conserved region within the C-terminal domain corresponds to amino acids 300 to the C-terminus of the protein (SEQ ID NO: 22 (MspJI)).
A “set of oligonucleotide fragments” of similar size obtained by cleavage of a large double-stranded DNA refers to the fragments resulting from cleavage of the large DNA on both sides of a modified nucleotide when the modified nucleotide is located on one strand approximately opposite to another modified nucleotide on the second complementary strand.
An “enzyme preparation” is intended to refer to a reagent and not something occurring in its natural state in vivo.
If a genome that consists of multiple large DNAs (e.g., chromosomes) is cleaved, each large DNA will give rise to a set of oligonucleotide fragments of similar size. A mixture of fragments obtained from cleavage of an entire human genome can be considered as a plurality of sets of oligonucleotide fragments, each set derived from a chromosome or as a single set of fragments depending on the context. In an embodiment, the set of oligonucleotides comprises at least 6 oligonucleotide fragments with different DNA sequences. For example, the set of oligonucleotides may comprise at least 10 oligonucleotides with different sequences or at least 20 oligonucleotides with different sequences. In one embodiment, a cloned double-stranded DNA can be enzymatically modified at a target nucleotide at one site for example, modification of a cytosine at a CpG on both strands. In this example, double strand cleavage by a member of the MspJI enzyme family will occur on both sides of the modified cytosine at a non-random distance. The set of oligonucleotides will consist of similarly sized fragments with a centrally located modified nucleotide.
Members of the newly described MspJI family of enzymes have been identified from microbial sources although enzymes in the family are not limited to those found in microbes. A BLAST search has shown that the number of sequences identified in the DNA database containing genomes from all living sources that encode proteins in the family as defined herein is relatively small. Sixteen homologs are shown in FIGS. 1C, 1D and 7-1 to 7-7 and their % sequence identity and % homology (similarity) are shown in Table 2.
Enzymes capable of recognizing a modified nucleotide and cleaving the DNA at a non-random distance from the modified nucleotide have been found to share sequence motifs in the N-terminal domain. These enzymes have been found to be capable of uniquely cleaving the DNA on both sides of the modified nucleotide to produce a fragment of a non-random size. A modification on the 5 position of cytosine (m5C) in a eukaryotic genome is most commonly associated with regulation of gene expression. Embodiments of the invention may encompass enzymes capable of recognizing a modified cytosine at a CpG site, the exocyclic N4 position of cytosine (mN4C) or a modified nucleotide other than cytosine such as adenine, for example, the exocyclic N6 position of adenine (mN6A) where such enzymes cleave on either side of the modified recognition sequence.
The family of enzymes defined herein by a conserved sequence domain and certain functional features include derivative enzymes or variants that have sequence modifications outside or inside the recognition and/or catalytic domains. Additionally, recombinant derivative enzymes or enzyme variants are included in the family that may be fused to a second protein which serves as a label, tag or marker (U.S. Pat. No. 5,643,758) or may contain a substitution which acts as a label such as occurs with a selenocysteine substitution (see U.S. Pat. No. 7,141,366). In addition to the above described family of enzymes, derivative enzymes and enzyme variants are contemplated in which the catalytic domain is modified or absent such that the N-terminal domain acts as a methylated DNA or hydroxymethylated DNA binding domain.
The use of the MspJI family of enzymes to generate a set of oligonucleotide fragments may rely on a single enzyme or may include a plurality of enzymes where some or all of the enzymes are members of the MspJI family or are derived from members of the MspJI family.
The members of the newly defined family, before chemical modification or mutation, can be defined structurally by one or more of the features listed below.

- (a) non-heteromeric;
- (b) recognition and cleavage functions in a single open reading frame;
- (c) the coding sequence and protein sequence do not contain methyltransferase motifs;
- (d) at least 90% sequence homology to a conserved motif in the N-terminal domain which is WXD(X)₁₀YXGD; and
- (e) common secondary structure elements that encompass the conserved motif.

FIG. 1B and FIGS. 7-1 to 7-7 show an embodiment of the enzyme family in which the overall order of the secondary structure elements that build the catalytic core is helix(H1)-helix(H2)-sheet(S1)-sheet(S2)-sheet(S3)-helix(H3)-sheet(S4)-helix(H4) (see, for example, FIGS. 7-1 to 7-7 where h=helix and a series of e represents a β-pleated sheet). The conserved FE is in an α-helix H2; the first conserved aspartic acid (D) is in a hinge region between two β sheets S1 and S2; the second conserved aspartic acid is in β sheet S2; the conserved (Q/E)XK is in a β sheet S3.
Members of the family may be identified by a BLAST search using a sequence selected from SEQ ID NO:7-22 or a related sequence. The hits are then further searched for the above specified consensus sequence in the N-terminal domain and also optionally searched for at least 90% sequence homology or identity in the C-terminal domain of a consensus sequence FEX_20-30DX_2-4DX_19-22(Q/E)XK. Optionally, the conserved sequence in the N-terminal domain can be extended to greater than 90% sequence homology to WXD(X)₆G (X)₃YXGD(X)_10-15GN(X)₂L X_10-20PX₃F and/or greater than 90% sequence homology with FE(X)₂A(X)_15-18T/SX₄DGGXDX₂G/LX_15-20E/QAK.
The selected sequences may then be expressed by techniques known in the art, for example in vitro transcription-translation (PURExpress™, New England Biolabs, Inc. (NEB), Ipswich, Mass.) or by cloning into a microbial host lacking modified bases such as #ER2655 (NEB Express, #C2523, NEB, Ipswich, Mass.) and assayed for cleavage of DNA containing modified nucleotides to produce oligonucleotide fragments of a defined size and/or containing a centrally located modified nucleotide.
Antibodies may be raised to members of the newly defined family of enzymes (MspJI enzyme family) using standard techniques for generating monoclonal or polyclonal antibodies. These antibodies or fragments thereof may be used for in situ-labeling of a member of an MspJI enzyme family bound to the modified large DNA. The enzyme may be mutated so that the cleavage function is inactivated or removed. In this context, the fragments may then be separated by binding to affinity matrices capable of antibody binding.
Functionally, MspJI was identified as a DNA sequence in the database adjacent to a methylase gene sequence and was hence named an endonuclease gene. However, when it was expressed as a protein, it was found to be inactive using standard assays for determining restriction endonuclease activity. This would have normally terminated any further study but for the fortuitous discovery described here that when incubated with DNA from a Dcm+ strain of E. coli, the enzyme was active, while it was inactive when tested on DNA from a Dcm− strain of E. coli. When the enzyme was incubated with eukaryotic DNA that is known to contain modified cytosines such as human genomic DNA, a smear of high molecular weight DNA was observed on polyacrylamide gels together with a clearly visible band containing fragments that correspond to a size of about 32 base pairs (see FIG. 6B).
A family of related enzymes were identified (FIGS. 7-1 to 7-7) and DNA cleavage by representative examples of these enzymes were tested with human genomic DNA. The tested enzymes yielded a set of similarly sized oligonucleotide fragments of about 32 nucleotides as can be observed on the gel in FIG. 6C.
The newly defined family of enzymes described here is of particular interest for reasons that include their ability to recognize nucleotide residues modified at the 5 position and to produce a set of oligonucleotide fragments where cleavage occurs at a substantially fixed distance downstream of the enzyme recognition site on the DNA (see FIGS. 6A, 6B and 6C). In embodiments of the invention, the cleavage distance from the modified sites conform to the following rules:
(1) For double-stranded DNA with a palindromic m5 CpG or other modified nucleotide on both strands in close proximity, a double-stranded break may be generated on each side of the modified nucleotide to generate fragments of similar size. In one embodiment, the distance between the cleavage site on one strand with a modified CpG was found to be 12 bases and the distance to the cleavage site on the opposite strand from the m5C was found to be 16 bases (MspJI) including a 4 base overhang resulting in oligonucleotide fragments of 32 bases in length.
(2) For hemi-modified double-stranded DNA, a double-stranded break occurs at a position which is 3′ downstream from the modified nucleotide. The distance from the cleavage site on the same strand to the modified nucleotide is constant (for example for MspJI the distance is 12 bases and the distance from the cleavage site on the other strand to the modified nucleotide is 16 bases). Sites of hemi-modification in the DNA can be detected by ligating an oligonucleotide containing a site recognized by an MmeI-like enzyme (such as MmeI, see U.S. Pat. No. 7,115,407) to a hemi-modified DNA cleaved at one side at a site 16 nucleotides from the modified nucleotide. The oligonucleotide may include four degenerate nucleotides at the 5′ end of the MmeI-site oligonucleotide to allow annealing to the 4 base extension on the bottom strand. Alternatively, a blunt-ended oligonucleotide might be used such that the single strand region at the 4-base extension is filled in using standard molecular biology techniques. The MmeI-like enzyme will cleave 18 or 19 nucleotides upstream which for MspJI cleavage fragments is about 2 nucleotides upstream of the modified fragment. Fragments produced in this manner can be sequenced and the position of the hemi-modified nucleotide in the DNA determined.
The number of CpG sites in genomes from human, mouse and Arabidopsis has been determined using bioinformatics. If the genomes of these organisms are then cleaved into various length fragments from 24 bases to 60 bases containing a centrally positioned CpG, then the fragments with unique sequences represent between 71% and 91% of the total unique sequence in humans according to increasing size, 83% to 90% in mice and 89% to 95% in Arabidopsis by the same criteria.
If those sequences of a defined length which have a distinct sequence are distinguished from the total sequence and separately analyzed, then 96%-98% of fragments of size 24-60 nucleotides will match a single locus in the human genome. In the bioinformatic analysis provided in FIGS. 8-1 to 8-2, if the fragment length is 60 nucleotides, then there are 26,185,493 fragments that contain a centrally positioned CpG of which 25,538,480 have distinct sequences and 98% of these match a single locus in the genome (see FIGS. 8-1 to 8-2).
Hence, there is significant informational value in a set of oligonucleotide fragments generated by an enzyme in the newly defined family where the enzyme recognizes a modified nucleotide and cleaves the DNA at a distance from the modified nucleotide to preferably generate fragments of a similar size. The data shows that a large fraction of such fragments are highly likely to map to a single locus in the genome. This makes possible for the first time a simple and efficient method for creating a methylome. Consequently, high throughput sequence analysis can yield the location of the majority if not all of the actual modified nucleotides in the genome rapidly and easily.
In an embodiment of the invention, screening assays are described for determining modification-specific cleavage activity of an enzyme (see Example 1). These assays are not intended to be limiting. In one embodiment, selected host cells were transformed with a plasmid containing a specific DNA methyltransferase gene. The expressed methyltransferase then methylated the host genome at specific sites. Hundreds of methyltransferases with varying defined sequence specificities have been described in the literature (see for example REBASE®, a publicly available online database maintained by New England Biolabs, Ipswich, Mass.). Any of these methyltransferases with different methylation specificities can be used for screening purpose. Introduction of a compatible plasmid expressing a gene with modification-dependent cleavage activity able to act on the host's modification pattern would reduce or eliminate the viability of these transformed cells, leading to a low transformation plating efficiency. Non-methylated hosts would show high plating efficiency in a parallel transformation with the methylation-specific endonuclease gene. Thus, this test would confirm the modification-dependent cleavage property of the encoded gene product.
It was found that the activity of the enzymes in the MspJI family can be enhanced in the presence of a double strand DNA activator preferably having a length of less than 16 by and containing a modified dcm site (for example, dcm(C^5mCWGG) site). A 30 bp cleavage-resistant DNA activator containing phosphorothioate linkages at the cleavage site also stimulated the enzyme reaction.
Determining the level of methylation or hydroxymethylation of DNA samples is important for epigenetic studies. Epigenetic regulation of the genome includes chromatin remodeling which may be accomplished by the addition of methyl groups to the DNA, mostly at CpG sites to convert cytosine to 5-methylcytosine, and its reversal possibly via hydroxymethylcytosine. Methylation of cytosines in the eukaryotic genome may persist from the germline of one parent into the zygote marking the chromosome as being inherited from this parent (genetic imprinting). Moreover, large changes in methylation occur following zygosis and in tissues of the developing organism (Morgan et al. Hum Mol Genet 14 Spec No. 1:R47-58 (2005)). In addition, methylation in some regions of the genome may vary in response to environmental factors (Li, et al Cell 69 (6): 915-926 (1992)). Differences in methylation pattern may be critical indicators of inappropriate developmental processes for example for embryonic stem cells (Brunner et al. Genome Research 19:1044-1056 (2009)).
Certain enzymes (such as DNMT1) have a high affinity for the m5C. If this enzyme reaches a “hemi-methylated” portion of DNA (where methylcytosine is in only one of the two DNA strands), the enzyme will methylate the other half.
DNA methylation occurs in repeated sequences, and helps to suppress the expression and mobility of transposable elements (Slotkin, et al. Nat Rev Genet. 8(4):272-85 (2007)). Because of spontaneous deamination, 5-methylcytosine can be converted to thymidine; hence, CpG sites are frequently mutated and thus become rare in the genome, except at CpG islands where they remain unmethylated. Deamination in this situation converts cytosine to uracil. Diagnostic changes in methylation pattern have the potential to detect increased frequencies of permanent genetic mutation. Methylation in the human genome has been studied in cancer cells for purposes of exploring therapies (see for example, Gargiulo, et al. The International Journal of Biochemistry & Cell Biology 41:127-35 (2009); and Gronbaek, et al. Basic Clin Pharmacol Toxicol 103:389-96 (2008)).
Embodiments of the invention significantly advance the ability to map modified nucleotides in a genome to generate a map (methylome). A human methylome would facilitate studies on interpersonal phenotypic variations in whole organisms and in individual cells and can yield useful information on development, aging and disease. From this information, it may be possible to determine susceptibility for diseases such as cancer even before a pathology appears and to design appropriate treatments with the possibility of providing powerful diagnostic tests and therapeutic agents.
The identification of a family of enzymes with novel properties and the creation of novel oligonucleotide fragments permits a description of the status of the human methylome by allowing the isolation of a set or sets of oligonucleotide fragments that provide a concentration of modified bases found in the human genome. Isolation of the set(s) of fragments can be facilitated by gel electrophoresis, solid phase affinity-binding or other means. Methylome analysis can be aided by the addition of a control that may include treating the genome with M.SssI, which methylates substantially all CpG dinucleotides in the genome. (Yegnasubramanian et al. Nucleic Acids Res 34:e19 (2006)).
The set(s) of oligonucleotide fragments resulting from enzyme cleavage can be sequenced using high throughput sequencing methods of the sort that are currently available using NextGen sequencing methods to identify and map modified cytosine nucleotides in a DNA. This approach greatly simplifies the generation of a methylome for any large DNA or genome such as a mammalian genome. Selection of specific oligonucleotide cleavage products for rapid diagnostic methods to particular regions of the genome can determine the abnormal presence or absence of modified cytosines correlated with a disease such as cancer. Specific oligonucleotides may be used to determine a particular phenotype for an individual. For example, hybridization of a set of fragments to a defined sequence or set of sequences presented on a solid surface (array hybridization) or tagged in a solution (or visa versa) can reveal discrepancies with a standard set of fragments that characterize the methylome. qPCR or array hybridization may also be used to interrogate one or more known locations of interest for abundance. The modified nucleotide or binding molecule may be labelled with a fluorescent or chemiluminescent tag or other labelling methods known in the art to facilitate detection.
Modified nucleotides in the genome may be identified in situ using a mutant enzyme member of the newly defined family having an inactivated cleavage site. By visualizing the binding sites of the mutant enzyme, the location of the modified nucleotides can be determined.
The members of the newly defined family of enzymes may be genetically engineered so as to form recombinant proteins for large scale production. The purification of recombinant proteins may be facilitated by the formation of fusion proteins such that the enzyme is fused to an affinity tag which has an additional use. For example, if the tag is biotin, a His peptide, chitin-binding domain or maltose-binding protein or another substrate-binding domain, the member of the family may be isolated on an affinity matrix either directly itself acting as a methyl-binding domain or by binding to an antibody affinity matrix or by means of the affinity tag. The recombinant protein either alone, modified or fused to a tag may be fluorescently labeled for imaging purposes.
Where the enzyme kinetics are single turnover, low turnover or the enzyme lacks catalytic activity altogether, the enzyme or enzyme fusion protein while bound to a modified nucleotide-containing fragment may also be bound directly to an affinity matrix to separate the oligonucleotide fragments containing modified nucleotides from the remaining fragments for sequencing or for diagnostic tests.
The experimental protocols provided below in large part for MspJI are not intended to be limiting. One of ordinary skill in the art could employ the experimental design as provided below to any additional member of the newly defined family.

EXAMPLES

Example 1

Methylation-Specific DNA Cleavage Activity of MspJI Enzyme Family

Production of the Enzymes

Recombinant members of the MspJI enzyme family were expressed in dcm− strain ER2566 and purified until substantially homogeneous using multiple chromatography steps. The enzymes which had an N-terminal 8×His Tag were first purified on a HiTrap Heparin HP column (GE, Piscataway, N.J.), then a HisTrap HP column (GE, Piscataway, N.J.), and finally a HiTrap SP column (GE, Piscataway, N.J.). The purification procedure followed the manufacturer's recommendation. The cleavage activity of the enzyme fractions were assayed on lambda DNA (which is partially dcm− methylated). To further improve expression levels, the DNA encoding the enzymes can be codon-optimized.

Determining Cleavage Patterns of MspJI Enzyme Family by a Screening Assay

The assay may include one or more of the following steps:

- 1. Methylate a target nucleotide in a synthetic or naturally occurring large DNA optionally having a known sequence. For example, lambda DNA can be used which is partially dcm− methylated at CmCWGG sites and XP-12 phage genomic DNA whose cytosines are completely replaced by 5mC cytosines.
- 2. React the large DNA with an MspJI enzyme family.
- 3. Size-separate the cleavage products for example using a polyacrylamide gel.
- 4. Sequence a set of oligonucleotide fragments of similar size to determine the position of the modified nucleotide; and
- 5. Optionally map the fragment sequence on the large DNA sequence.

An elaboration of step (1) includes using different large DNA preparations which have been reacted with different methyltransferases for modifying the DNA in vitro. These substrates are used to identify substrate specificities.
The products are analyzed by 1% agarose gel electrophoresis and can be visualized by ethidium bromide. For example, M.HpaII (NEB, Ipswich, Mass.) can produce CmCGG-modified DNA. Plasmid DNA-digestion can be monitored by 1% agarose gel electrophoresis and visualization using ethidium bromide staining. Alternatively, synthetic double-stranded oligonucleotide containing a modified site can be used in which any methylated sites can be easily created, independent of the availability of the methyltransferases. Modified nucleotides in palindromic sites of interest include for example, NmCGN, mCNG, NGmCN, GNmC etc., where N is A, T, G or C. In addition to fully-methylated oligonucleotides, oligonucleotides with hemi-methylated sites can be tested in this way. Other types of modification, such as 5-hydroxymethylated cytosine and 5-glucosylated-hydroxymethylated cytosine, can be either directly incorporated into the oligonucleotides during synthesis or by further modification of hydroxymethylated cytosine residues with bacteriophage T4 glucosyltransferase.
To determine cleavage sites, substrate oligonucleotides are labeled either at their 5′ end or 3′ end with ³³P. Cleavage products are run on a 7M-urea 20% polyacrylamide denaturing gel to single nucleotide resolution and analyzed.

Characterizing MspJI Enzyme Family Members by an In Vivo Screening Assay

The ER1992 strain with endogenous methylase gene dcm, which methylates the inner cytosine in CCWGG sites to CmCWGG and serves as a target substrate for an enzyme with the desired cleavage activity, and ER2566, with a dcm− genotype with no 5-methyl cytosine and not subject to cleavage by a methylation-specific enzyme, were used to screen for methylation-specific activity of a novel recombinant restriction endonuclease.

Measuring the Activity of MspJI Enzyme Family Members by an In Vitro Assay

A plasmid is used which contains only two methyl-C's separated by 1 kb of intervening sequence. This cleaves leaving three fragments, the plasmid backbone 3 kb, the insert 1 kb and the two 32 bp fragments. When this digest goes to completion, the uncut plasmid disappears and subsequent appearance of the 1 kb and backbone bands is easily quantifiable on an agarose gel. This plasmid is transformed into a dam-dcm− strain and purified as an assay substrate. Such plasmids are described in Stewart, F., et al. Biological Chemistry 379:611-616 (1998).

Determining Specificity of MspJI Enzyme Family Members for Modified Binding Sequences Versus Unmodified Binding Sequences Using an In Vitro Assay

The in vitro activity of MspJI was assessed on a variety of methylated and non-methylated DNA substrates, as shown in FIG. 2A for the dcm− methylated plasmid DNA pBR322. MspJI showed endonuclease activity (FIG. 2A, Lanes 1 and 2) where this endonuclease activity was DNA methylation-dependent. In contrast, MspJI did not act on pBR322 without dcm− modification (FIG. 2A, Lanes 5 and 6). By using the restriction enzyme BstNI (CC↓WGG), which is insensitive to m5C methylation, in a double-digestion assay, cleavage sites on pBR322(dcm+) by MspJI were shown to be at or close to the dcm sites (FIG. 2A, Lanes 2, 3 and 4). The double digest did not alter the BstNI pattern, suggesting that MspJI did not cleave at non-BstNI sites.
In addition to the m5C-modified DNA tested above, MspJI did not exhibit endonuclease activity on M.TaqI-(TCGmA) or dam-(GmATC) methylated pBR322(dcm−) DNA. This confirmed that MspJI did not target m⁶-adenine methylated DNA, consistent with the fact that the MspJI gene can be maintained and expressed in a dam+ strain (ER2566, NEB, Ipswich, Mass.). Moreover, MspJI does not apparently act on N⁴-methylcytosine-containing plasmid DNA, as can be determined by using M.BstNI (CCWGG, a N4-cytosine methylase) methylated DNA.
Assaying Activity of MspJI Enzyme Family Members on DNA Substrates which Contain 5-Hydroxymethylcytosine or 5-Glucosyl-Hydroxymethylcytosine
Wild-type T4 phage DNA with glucosylated cytosines and the DNA from a T4 α gt57 β gt14 (a mutant which has defective glucosyltransferases and therefore contains hydroxymethylated cytosines in DNA hereafter T4gt) were used as substrates (FIG. 2C, Lanes 1 and 2). MspJI was able to degrade T4 gt DNA (FIG. 2C, Lanes 8-11) and was inactive on glucosylated DNA (FIG. 2C, Lane 7). For comparison, activity of another modification-dependent endonuclease, McrBC (FIG. 2C, Lanes 3 and 4) and a typical type IIP restriction enzyme MspI (FIG. 2C, Lanes 5 and 6) with these modified DNA substrates were shown. McrBC, which recognizes pairs of (A/G)^mC separated by 40-3000 base pairs, also exhibited nuclease activity on the hydroxymethylcytosine-containing DNA but not on T4 wild-type DNA, while MspI was inactive with respect to both substrates. Note that MspJI was able to degrade T4 gt DNA to a greater extent than McrBC, which can be explained by its broader recognition sequence than McrBC. Overall, it appears that MspJI specifically targets cytosine-modified DNA with 5-CH₃or 5-CH₂OH addition on the pyrimidine ring.

Determining Substrate Sequences Around a Cleavage Site

MspJI-digested DNA samples with different methylated sites were subjected to capillary sequencing and the cleavage sites were deduced from the location at which peaks are reduced in height near the methylated sites in the sequencing chromatograms (examples shown in FIG. 5). Positions of cleavage occur at locations where the sequence signals (peaks) are reduced in height. In many cases, a non-templated adenine is added as the polymerase runs off the DNA, and location of such a “runoff peak” adenine is additional evidence for the location of cleavage. One observation on the sequencing chromatogram data was that cleavage sites occur at a site distant from the methylated sites. FIG. 5 also displays the deduced cleavage pattern of MspJI on different methylated sites. Another observation was that the height reduction in sequencing peaks and addition of adenine not present in the substrate were generally present on both sides of the methylated sites. The response in the chromatogram on both sides demonstrated that MspJI cleaved the DNA each side of the methylated binding sequence. This is consistent with the symmetry of the methylated binding sites. The presence of two run-off peaks is evidence for two independent cleavage events on the same strand.
It was concluded that MspJI recognized the m5C on one strand and then cut 12 nucleotides 3′ downstream on the same strand and 16 nucleotides downstream on the complementary strand to leave a 4-base 5′ overhang. Similarly, when the m5C on the complementary strand was recognized, the same pattern of cleavage was observed demonstrating that two double-stranded breaks around the same recognition site released the fragment with the methylated site in the middle. The exact length of that fragment depended on the distance between the methyl groups on the two strands. In the case of HpaII methylated sites (Cm5CGG) or HhaI methylated sites (Gm5CGC), the length of the fragment excised from the DNA substrate was expected to be 32 nucleotides including the two 4-base 5′ overhangs.

Comparing the Activity of an MspJI Enzyme Family Members on Fully-Methylated and Hemi-Methylated DNA

To investigate whether MspJI is active on hemi-methylated DNA substrates, which can arise during replication, FAM-labeled synthetic substrates were used in a digestion assay (FIG. 4A). FIG. 4A indicates the expected cleavage sites and product sizes and FIG. 4B shows the digestion reactions resolved on a 7M urea 20% polyacrylamide denaturing gel. The m5C for interrogation is at an M.HpaII site (CmCGG) in the oligos. Null-methylation, full-methylation, hemi-methylation on the top strand or bottom strand wer tested and the cleavage events on the top strand and bottom strands were observed by labeling them individually, as shown in FIG. 4B.
On fully-methylated DNA, MspJI makes cuts on both sides of the methylated site. On the top strand, it cleaves on either side of the methylated site, resulting in 40 by (from cut Lt) and 11 or 12 by (from cut Rt) fragments (FIG. 4B, Lane 3). Symmetrically, on the bottom strand, it cleaves twice and generates a long fragment of 36 nt (from cut Lb) and a shorter fragment of 7 or 8 nt (from cut Rb) (FIG. 4B, Lane 4).
On the hemi-methylated substrates, strand methylation status dictates the side of the cleavage, so that double-stranded breaks only occur on the 3′ side of the strand containing the methylated base. For example, for substrate with only top strand methylation, each cleavage event is at the 3′ side of the 5mC so that only shorter fragments are observed (FIG. 4B, Lanes 5 and 6). The same applies to the substrate with bottom strand methylation where only longer fragments are seen (FIG. 4B, Lanes 7 and 8). The results show that each m5C is associated with two cuts on the same side and such association is symmetrical. Thus, while not wishing to be bound by theory, it is proposed that MspJI recognizes each half of the methylated site separately in a fully-methylated site, either the top or the bottom strand, and that the half site then dictates the directionality of the cleavage.
Characterizing the MspJI Enzyme Family Members with Shared Conserved DNA Sequences, Secondary Sequence Motifs, and Binding and Cleavage Properties
By using the amino acid sequence of a member of the MspJI family such as MspJI as the query sequence, a PSI-BLAST search (Altschul et al., Nucleic Acids Res 25:3389-3402 (1997)) against GenBank retrieved more than 100 hits with significant sequence homology. Sixteen genes among the top hits had significant similarity to MspJI throughout the sequence length. In FIG. 1D, a partial multiple sequence alignment is provided around the conserved catalytic motif inside the MspJI subfamily. The significance of the conserved catalytic motif is shown by the site-directed mutagenesis experiments, in which both D334A and Q355A mutations completely abolish the catalytic activity of MspJI.
The predicted secondary structure elements of the MspJI family were determined using multiple sequence alignment created by PROMALS webserver (Pei et al. Nucleic Acids Res 35:W649-652 (2007)) (FIG. 1B shows a schematic and FIGS. 7-1 to 7-7 the full alignment). The structure core of the catalytic C-terminal domain has three consecutive strands (β1β2β3 in FIG. 1B), with the motif (Q/E)xK at the end of β3 and the conserved residue D at the beginning of β2 (FIGS. 7-1 to 7-7) (Wah et al. Proc Natl Acad Sci USA 95:10564-10569 (1998)). The two helices and strands in the order of α4-β4-α5-β5 after β1β2β3 forms an interaction interface between monomers (FIG. 1B).

Determining the Role of an Activator in Improving Cleaving Activity of Members of the Newly Defined Family of Enzymes

An activator dimer containing double-stranded 5-methyl cytosine (e.g., 11mer, 15-mer, 19mer and 23mer) are tested to determine whether digestion by members of the MspJI enzyme family can be enhanced. These dimers are constructed by annealing two single strand oligonucleotides or by hairpin formation of a single oligonucleotide.
The assay for the activator includes constructing self-complementary oligonucleotides containing 5-methyl-C's at the center of various lengths. Oligonucleotides are biotinylated at the 5′ end for subsequent removal from reactions and 3′ amino-modified such that they cannot be ligated or extended. The activators are then assayed for their abilities to enhance cleavage before and after streptavidin bead removal for interference in sequencing.

Example 2

Demonstration of the Application of Enzymes in Mapping the Methylome

To analyze the methylome analysis of a mouse or a human genome, 1-2 μgs of human or mouse genomic DNA is used for the methylome analysis at single nucleotide resolution. The genome is digested with a member of the MspJI family optionally in the presence of biotin-containing activator molecules, followed by removal of activator molecules using streptavidin magnetic affinity beads. The digested DNA is end-repaired using the NEBNext™ (NEB, Ipswich) end-repair module, ethanol-precipitated and dissolved in a suitable volume of water. The digested genomic DNA is ligated to bar-coded SOLiD primer and P1 primer using NEBNext™ quick ligation module (NEB, Ipswich, Mass.). The ligated product is separated on a 10% TBE polyacrylamide gel and the ligated product of ˜110 bps (between 100-130 bp) is excised after visualization by ethidium bromide staining. A crush and soak or a suitable elution method is used to isolate DNA for SOLiD sequencing (Applied Biosystems, Inc., Life Technologies, Inc., Carlsbad, Calif.). MspJI for example does not distinguish between methylated and hydroxymethylated cytosine residues for cleavage, thus the sequencing data will result in analysis of the whole methylome.

Determination of the Biological Role of 5-Hydroxy Methylcytosine in Mammals

The dynamic changes of DNA methylation during mouse embryonic stem cell (ES) differentiation can be identified using the newly defined family of enzymes. Previous reports suggest that as much as 10% of the modified cytosines are in the form of 5-hydroxymethylcytosine. These will have been missed using current methodologies involving bisulfite. This modified adduct is complementary to guanine and is read as cytosine in polymerase-based amplification.

Exploration of Methylomes of Other Model Organisms.

The MspJI enzyme family not only acts on mCpG, but is capable of recognizing and cleaving other types of methylated sites. For example, mCNG, which is present in the genomic DNA of Arabidopsis, is a natural substrate for MspJI. This provides a simple method of assaying for the presence of modified bases in any organism. For instance, digestion of total genomic DNA with MspJI gives a 32 by fragment that is easy to isolate from a polyacrylamide gel. It can then be digested to mononucleosides using a standard cocktail of enzymes and the total digest examined by HPLC and/or mass spectrometry to identify the modified bases. A variety of organisms such as Arabidopsis, Xenopus, zebrafish, chicken, Neurospora crassa as well as genomes known to contain unusual modifications like Base J found in kinetoplastid protozoans such as Trypanosomes (Cross et al. EMBO J. 18:6573-6581 (1999)) is studied by the methods described herein. Once an epigenome is confirmed, the digested bands can be sent for high throughput sequencing using the established protocols for human.

TABLE 1

Genomic context analysis of the MspJI subfamily

			Close to			Activity on
#	Gene ID	Genbank ID	methylase?	Species	Additional notes	5mC DNA*

1	MspJI	YP_001069123	Y	Mycobacterium sp. JLS	close to M and V genes	Y
2	Sbal195_0369	YP_001552810	Y	Shewanella baltica OS195	2 ORFs from M gene	N/T
3	PE36_01892	ZP_01896882	Y	Moritella sp. PE36	SAM-dependent	N/T
					methylase

4	Spea_3849	YP_001503694	N	Shewanella pealeana ATCC		N
				700345
5	Xcc3577	NP_638923	N	Xanthomonas campestris	Among transposase	N
				ATCC 33913	islands
6	Lhv_0031	YP_001576608	N	Lactobacillus helveticus DPC	Among transposase	N
				4571	islands
7	Lpg1234	YP_095265	Y	Legionella pneumophila strain	Close to R and M	Y
				Philadelphia
1	genes; M gene is
					active.
8	Franean1_5336	YP_001509600	N	Frankia sp. EAN1pec	Standalone	Y
9	CATMIT_00196	ZP_03681584	N/A	Catenibacterium mitsuokai	Unfinished genome	N/T
				DSM 15897	sequence
10	Rmet_0004	YP_582159	N	Ralstonia metallidurans CH34	Standalone	N
11	Bcenmc03_0011	YP_001763314	Y	Burkholderia cecepacia MC0-3	Close to M and V	N/T
					genes

12	AspBH1	YP_931859	Y	Azoarcus sp. BH72	Close to M and V	Y
					genes

13	RlaI	ZP_03168528	N	Ruminococcus lactaris ATCC	Close to V gene	Y
				29176
14	Xccb100_0619	YP_001902024	N	Xanthomonas campestris strain	Among transposase	N/T
				B100	islands

15	ZP_03855940	ZP_03855940	Y	Veillonella parvula DSM 2008	Close to M gene	N

TABLE 2

Similarity and identity of sequences in the family

similarity

identity	AspBHI	Bcenmc03_0011	CATMIT_00196	Franean1_5336	MspJI	PE36_01892	RlaI	Rmet_0004

AspBHI	100/100	36.5/47.4	41.3/57.7	40.1/52.4	25.6/38.0	22.8/38.3	42.2/58.0	38.6/52.7
Bcenmc03_0011	36.5/47.4	100/100	30.8/43.6	27.9/40.6	18.9/28.7	16.7/29.8	32.8/45.9	33.4/47.4
CATMIT_00196	41.3/57.7	30.8/43.6	100/100	35.9/51.1	20.2/37.1	24.7/41.2	40.9/61.8	33.1/47.3
Franean1_5336	40.1/52.4	27.9/40.6	35.9/51.1	100/100	22.3/34.5	23.2/37.0	36.6/51.7	35.9/46.7
MspJI	25.6/38.0	18.9/28.7	20.2/37.1	22.3/34.5	100/100	36.5/53.4	23.1/40.2	22.6/35.3
PE36_01892	22.8/38.3	16.5/29.1	24.7/41.2	23.2/37.0	36.5/53.4	100/100	21.6/36.8	24.1/36.7
RlaI	42.3/58.2	32.8/45.9	40.9/61.8	36.6/51.7	23.1/40.2	21.6/36.8	100/100	38.0/53.4
Rmet_0004	38.6/52.7	32.3/48.5	33.1/47.3	35.5/46.5	22.5/36.6	24.0/37.7	38.0/53.4	100/100
Sbal195_0369	22.6/36.7	16.4/28.8	24.7/42.0	22.8/35.0	36.3/54.4	73.8/83.3	24.3/41.3	24.1/35.1
SgriT_16873	45.7/60.9	30.2/42.8	37.0/55.4	52.1/64.2	25.4/39.8	26.4/40.6	41.3/57.7	37.4/50.8
Spea_3849	21.5/35.2	17.5/29.7	24.8/43.5	23.6/36.2	34.3/53.9	70.6/81.7	23.9/42.0	24.2/35.2
Xcc3577	24.2/37.9	19.6/30.2	25.7/38.7	22.4/32.0	38.2/51.5	40.9/60.0	21.5/39.5	22.6/36.7
gi\|227372459\|	39.1/58.0	29.6/44.8	42.3/60.6	31.9/47.7	22.2/41.2	24.8/39.7	42.3/59.9	34.8/49.2
ref\|ZP_03855940.1\|
V.par
gi\|260101829	22.2/36.1	19.4/34.6	22.3/39.1	21.4/30.3	29.2/43.6	34.6/50.9	24.1/38.0	23.8/39.2
DSM 20075
lhv_0031	21.3/34.8	17.3/30.8	23.2/40.7	21.7/31.0	30.4/46.0	36.3/53.5	24.8/40.0	21.5/34.9
lpg1234	42.6/62.9	28.4/43.8	37.8/57.0	37.0/52.4	21.3/37.8	26.1/40.9	41.6/58.7	49.3/62.1

similarity

					gi\|227372459\|
					ref\|
					ZP_03855940.1\|	gi\|260101829
identity	Sbal195_0369	SgriT_16873	Spea_3849	Xcc3577	V.par	DSM 20075	lhv_0031	lpg1234

AspBHI	22.6/36.7	45.7/60.9	21.5/35.2	24.2/37.9	39.1/58.0	22.2/36.1	21.3/34.8	42.6/62.9
Bcenmc03_0011	16.4/28.8	30.2/42.8	17.5/29.7	19.6/30.2	29.6/44.8	19.4/34.6	17.3/30.8	28.4/43.8
CATMIT_00196	24.7/42.0	37.0/55.4	24.8/43.5	25.7/38.7	42.3/60.6	22.3/39.1	23.2/40.7	37.8/57.0
Franean1_5336	22.8/35.0	52.1/64.2	24.0/36.3	22.4/32.0	31.9/47.7	21.4/30.3	21.7/31.0	37.0/52.4
MspJI	36.3/54.4	25.4/39.8	34.3/53.9	38.2/51.5	22.2/41.2	29.2/43.6	30.4/46.0	21.3/37.8
PE36_01892	73.8/83.3	26.2/40.4	70.6/81.7	40.9/60.0	24.8/39.7	34.6/50.9	36.3/53.5	26.1/40.9
RlaI	24.3/41.3	41.3/57.7	23.9/42.0	22.2/40.2	42.5/60.1	24.1/38.0	24.8/40.0	41.6/58.7
Rmet_0004	24.1/35.1	37.1/50.4	24.2/35.2	22.6/36.7	34.8/49.2	23.8/39.2	21.5/34.9	49.3/62.1
Sbal195_0369	100/100	24.0/36.5	86.5/93.1	40.6/60.7	22.9/40.7	36.0/50.1	38.6/54.9	25.1/41.3
SgriT_16873	24.0/36.5	100/100	25.4/38.2	23.2/32.7	38.8/56.1	23.7/36.4	25.3/38.4	44.3/60.1
Spea_3849	86.5/93.1	25.4/38.2	100/100	39.9/58.9	24.7/42.7	35.1/49.8	36.5/53.0	23.5/38.0
Xcc3577	40.6/60.7	23.2/32.7	39.9/58.9	100/100	21.2/38.2	31.5/48.5	33.1/51.6	19.8/39.0
gi\|227372459\|ref\|	22.9/40.7	38.8/56.1	24.7/42.7	21.2/38.2	100/100	100/100	23.9/40.7	42.0/60.9
ZP_03855940.1\|
V.par
gi\|260101829	36.0/50.1	23.9/36.6	35.1/49.8	31.5/48.5	100/100	100/100	88.2/88.2	21.3/35.6
DSM 20075
lhv_0031	38.6/54.9	25.5/38.5	36.5/53.0	33.1/51.6	23.9/40.7	88.2/88.2	100/100	23.2/38.7
lpg1234	25.1/41.3	44.3/60.1	23.5/38.0	19.8/40.4	42.0/60.9	21.3/35.6	23.2/38.7	100/100

Claims

1. A set of double-stranded oligonucleotide fragments obtainable by enzymatic cleavage of a large DNA, the large DNA containing one or more modified nucleotides, the set comprising fragments wherein at least 50% are of a similar size and have a centrally positioned modified nucleotide.

2. A set according to claim 1, wherein the one or more of the fragments are isolated from the set.

3. A set of oligonucleotide fragments according to claim 1, wherein the large DNA is at least 100 nucleotides in length.

4. A set of oligonucleotide fragments according to claim 1, wherein the large DNA is a mammalian genomic DNA.

5. A set of oligonucleotide fragments according to claim 1, wherein the large DNA is human genomic DNA.

6. A set of oligonucleotide fragments according to claim 1, wherein the centrally located modified nucleotide is a cytosine.

7. A set of oligonucleotide fragments according to claim 6, wherein the centrally located modified cytosine is proximate to a guanine.

8. A set of oligonucleotide fragments according to claim 7, wherein the modified cytosine is a methylated or hydroxymethylated cytosine.

9. A set of oligonucleotide fragments according to claim 1, wherein the fragments are less than 60 nucleotides in size.

10. A set of oligonucleotide fragments according to claim 9, wherein the fragments have a similar size in the range of 28-36 nucleotides.

11. A set of oligonucleotide fragments according to claim 1, wherein at least one of the modified nucleotides is located within 30 nucleotides from one end of the fragment.

12. An enzyme preparation, comprising: at least one enzyme that recognizes a modified nucleotide in a DNA and cleaves the DNA at a site that is distant from the modified nucleotide thereby generating a set of fragments as claimed in claim 1, the at least one enzyme further characterized by an N-terminal conserved domain with greater than 90% amino acid sequence homology to WXD(X)₁₀YXGD.

13. An enzyme preparation according to claim 12, wherein the at least one enzyme cleaves the DNA at a non-random distance from the modified nucleotide.

14. An enzyme preparation according to claim 12, wherein the at least one enzyme has an N-terminal conserved domain with greater than 90% sequence homology with WXD(X)₆G(X)₃YXGD(X)_10-15GN(X)₂L X_10-20PX₃F.

15. An enzyme preparation according to claim 12, wherein the at least one enzyme comprises a recognition domain and a cleavage domain encoded by a single open reading frame.

16. An enzyme preparation according to claim 12, wherein the at least one enzyme has a C-terminal conserved domain with greater than 90% amino acid sequence homology to FEX_20-30DX_2-4DX_19-22(Q/E) XK.

17. An enzyme preparation according to claim 12, wherein the at least one enzyme has an amino acid sequence with greater than 90% sequence homology to a protein sequence selected from SEQ ID NOS: 7-22.

18. An enzyme preparation according to claim 12, wherein the at least one enzyme is fused to an affinity tag.

19. An enzyme preparation according to claim 17, wherein the affinity tag is selected from the group consisting of a chitin-binding domain, maltose-binding domain and a His tag.

20. An enzyme preparation according to claim 12, further comprising an activator DNA.

21. An enzyme preparation according to claim 12, wherein the N-terminal domain is capable of being recognized by an antibody.

22. An antibody as defined in claim 21.

23. An enzyme preparation, comprising: one or more enzymes that recognize a modified nucleotide in a DNA and cleave the DNA at a site that is at a non-random distance from the modified nucleotide, the one or more enzymes further characterized by an N-terminal conserved domain with greater than 90% amino acid sequence homology with WXD(X)₁₀YXGD.

24. A method for obtaining a set of oligonucleotide fragments as claimed in claim 1, comprising:

a. enzymatically cleaving a large DNA containing one or more modified nucleotides; and

b. obtaining the set of oligonucleotide fragments.

25. A method according to claim 24, further comprising separating the set of oligonucleotide fragments from uncleaved DNA.

26. A method according to claim 25, further comprising: sequencing from the separated set of fragments at least one fragment in the set of fragments to determine the location of one or more modified nucleotides contained within the at least one fragment.

27. A method according to claim 24, further comprising: analyzing some target oligonucleotide fragments for the presence and location of one or more modified nucleotides in the large DNA.

28. A method according to claim 24, further comprising: sequencing substantially all the fragments in the set of oligonucleotide fragments and mapping the sequences onto a genome sequence map to determine the location of modified cytosines.

29. A method for identifying one or more isolated enzymes in the enzyme preparation according to claim 12, comprising:

a. searching a sequence database using a sequence selected from the group consisting of SEQ ID. NO: 7-22 and variants thereof; and

b. identifying additional sequences having an N-terminal region characterized by a consensus sequence of WXD(X)₆G (X)₃YXGD(X)_10-15GN(X)₂L X_10-20PX₃F.

30. A method according to claim 29, wherein the identified additional sequences have a C-terminal end comprising a catalytic domain with a consensus sequence of FEX_20-30DX_2-4DX_19-22(Q/E)XK.

31. A method for isolating from a mixture, DNA fragments containing one or more modified nucleotides, comprising:

a. adding to the mixture an enzyme preparation according to claim 12, wherein the at least one enzyme has been mutated so as to lack enzyme cleavage activity, wherein the mutant enzyme is immobilized on a solid surface; and

b. separating the DNA fragments bound to the immobilized enzyme from the mixture.

32. A method to determine the location of at least one modified nucleotide in a large DNA, comprising:

a. cleaving a large DNA with an enzyme preparation according to claim 12;

b. obtaining a set of oligonucleotide fragments, each fragment containing at least one modified nucleotide; and

c. determining the location of the at least one modified nucleotide in a sequence map of the large DNA by sequencing one or more oligonucleotides in the set of oligonucleotide cleavage products.

33. A method of identifying a present or future phenotypic property in a cell preparation or tissue sample from a pattern of modified nucleotides; comprising:

a. cleaving into fragments a large DNA from a cell preparation or tissue sample by means of an enzyme preparation described in claim 12; and

b. comparing a location for modified nucleotides in the fragments with a pattern of modified nucleotides in a control DNA so as to determine a present or future phenotypic property.

34. A method according to claim 33, wherein (a) further comprises: separating fragments with one or more modified nucleotides from fragments lacking a modified nucleotide by:

(i) contacting the cleavage fragments with a molecule that binds to the fragments containing the one or more modified nucleotides with an immobilized preparation of an affinity-binding protein; or

(ii) size separation.

35. A method according to claim 34, wherein the affinity-binding protein is derived from an enzyme preparation according to claim 12, wherein the enzyme cleavage activity of the at least one enzyme has been inactivated.

36. A method according to claim 33, where (a) further comprises: identifying on a methylome or a genome a location for the one or more modified nucleotides in the immobilized cleavage fragments.

37. A method for obtaining a purified preparation of fragments containing one or more modified nucleotides, comprising:

a. contacting a mixture of DNA fragments in which one or more of the fragments contain at least one modified nucleotide with an immobilized preparation of an affinity-binding molecule;

b. binding the one or more fragments containing at least one modified nucleotide to the affinity-binding molecule; and

c. obtaining a purified preparation of fragments containing one or more modified nucleotides.

38. A method according to claim 37, wherein the affinity-binding molecule is an enzyme preparation according to claim 12, wherein the enzyme cleavage activity has been inactivated.

39. A method according to claim 38, wherein the isolated enzymes in the enzyme preparation are associated with a binding moiety.

40. A kit, comprising: an enzyme preparation according to claim 12, in a container and instructions for use.

41. A kit according to claim 40, further comprising an activator molecule.