WO 94/21663 215 9 fl 81 PCT~S94103246 DINUCLEOTIDE RESTRICTION ENDONUCLEASE PREPARATIONS AND METHODS OF
USE.
MELD OF THE INVENTION
The present invention relates generally to isolated purified polynucleotides which encode restriction enzymes and to methods of expressing the restriction enzymes from such polynucleotides. More particularly this invention relates to isolated purified polynucleotides which encode CviJI and related methods for the production of this enzyme.
Other aspects of the invention relate to methods for partially or completely digesting DNA at a dinucleotide sequence. More particularly, this aspect of the invention relates to methods of generating quasi-random fragments of DNA, and methods of cloning, labeling, and sequencing DNA, as well as epitope mapping of proteins. The invention also relates to methods for generating sequence-specific oligonucleotides from DNA, without prior knowledge of the nucleic acid sequence of such DNA, and to methods for cloning and labeling DNA after restriction digestion by a two base recognition endonuclease reagent.
This invention also relates to methods for cloning, labeling, and detecting nucleic acids using two base restriction endonuclease reagents, such as CviJ I, BsuR
I, Aci I or CGase I. Further the invention relates to labeling DNA by taking advantage of certain properties of the holo-enzyme of thermostable DNA
polymerases.
BACKGROUND OF THE IIVVENTION
Restriction endonucleases are a group of enzymes originally found to be expressed in a wide variety of prokaryotic organisms. More recently they have also been found to be encoded in viral genomes. These enzymes catalyze the selective cleavage of DNA at generally short sequences, often unique to the individual enzyme. This ability to cleave makes restriction endonucleases indispensible tools in recombinant DNA technology. The increased commercial availability of the isolated enzymes has contributed in large part to the enormous expansion in the field of recombinant DNA technology over the last few years.
These enzymes have been classified into three groups. Because of properties of the type I and type III enzymes, they have not been widely used in molecular biology applications, and will not be discussed further. Type II
enzymes are part of a binary system known as a restriction modification system consisting of a restriction endonuclease that cleaves a s~cific sequence of nucleotides and a separate DNA modifying enzyme that. modifies the same recognition sequence and thereby prevents cleavage by the: cognate endonuclease.
A total of about 2103 restriction enzymes are known, encompassing 179 different type II specificities (Roberts, et al. , Nucl. Acids Res. 20:2167-2180 (1992)).
Although there are more than 1200 type II restriction enzymes, many of them are members of groups which recognize the same sequence. Restriction enzymes that recognize the same sequence are said to be isoschizomers.
The vast majority of type II restriction enzymes recognize specific double-stranded sequences which are four, five, or six nucleotides in length and which display twofold (palindromic) symmetry. A few enzymes recognize longer sequences or degenerate sequences.
The location of cleavage sites within a palindrome differs from enzyme to enzyme. Some enzymes cleave both strands exactly at the axis of symmetry generating fragments of DNA that carry blunt ends, while others cleave each strand at similar sequences on opposite sides of the axis of symmetry, creating fragments of DNA that carry protruding, single-stranded termini.
Restriction endonucleases with shorter recognition sequences cut DNA more frequently than those with longer recognition sequences. For example, assuming a 50% G-C content, a restriction endonuclease with a 4-base recognition sequence will cleave, on average, every 44 (256) bases compared to every 46 (4096) bases for a restriction endonuclease with a 6-base recognition sequence. Under certain conditions some restriction endonucleases are capable of cleaving sequences which are similar but not identical to their defined recognition sequence. This altered specificity has been termed "star" (*) activity and is observed only under certain non-standard reaction conditions. The manner in which an enzyme's specificity is altered depends on the particular enzyme and on the conditions employed to induce the star activity. Conditions that contribute to star activity include high glycerol concentration, high ratio of enzyme to DNA, low ionic strength, high pH, the presence of organic solvents, and the substitution of Mg+'~ with other divalent rations. The most common types of star activity involve cutting at a .recognition sequence having a single base substitution, cutting at sites having truncation of the outer bases of the recognition sequence, and single-strand nicking. The following restriction endonucleases show star activity:
Ase I, BamH I, BssH II, BsuR I, CviJ I, LcoR I, EcoR V, Hind III, Hinf I, Kpn I, Pst I, Pw II, Sal I, Sca I, Taq I, and Xmn I. Star activity is generally viewed as undesirable, and of little intrinsic value.
Of the 179 unique type II restriction endonucleases, 31 have a 4-base recognition sequence, 11 have a S-base recognition sequence, 127 have a 6-base recognition sequence, and 10 which have recognition sequences of greater than 6 bases. In two cases, a restriction endonuclease has a recognition sequence of less than 4 bases.
The restriction enzyme CviJ I has a three base recognition sequence or a two-base recognition sequence, depending on the reaction conditions.
Under normal reaction conditions CviJ I recognizes the sequence PuGCPy (wherein Pu=purine and Py=pyrimidine) and cleaves between the G and C to leave blunt ends (Xia et al., 1987. Nucleic Acids ReS. 15:6075-6090). Under "relaxed" or "star" conditions (in the presence of 1 mM ATP and 20 mM DTT) the specificity of CviJ I may be altered to cleave DNA more frequently. This activity is referred to as CviJ I*, for star or altered specificity. However, CviJ I* activity is not observed under conditions which favor star activity of other restriction endonucleases.
~l~9pg~.
The restriction enzyme BsuR I normally recognizes the sequence GGCC and cleaves between the G and C to leave blunt ends. (Heininger, et al. , Gene 1:291-303 (1977)). Under relaxed conditions (high pH, low ionic strength, and high glycerol concentration) the specificity of Bsu RI may be altered to cleave DNA more frequently. An isoschizomer of this enzyme, Hae III, does not display this star activity.
In bacteria, the restriction endonuciease provides a mechanism of defense against foreign DNA molecules (e.g., bacteriophage DNA) by virtue of its ability to distinguish and cleave only exogenous DNA, leaving endogenous bacterial DNA unaffected. Viral endonucleases possess the same discerning capabilities, but rather than providing a means for defense, this activity has presumably evolved to cripple the host's ability to replicate its own DNA and allows the virus to assume control of the host's replication machinery.
Bacteria and viruses which express restriction endonucleases necessarily possess the inherent ability to protect their own genome from cleavage by their endogenous endonuclease. The primary mechanism by which this is accomplished is by modifying the organisms own DNA by, for example methylating a base in the recognition sequence which prevents binding and cleavage by the endonuclease. Therefore, to insure viability, the genome of an organism which expresses a restriction endonuclease is almost always heavily modified, usually by methylation of cytosine or adenosine bases. The methylase enzyme which modifies the genome (itself a useful tool in molecular biology) acts in tandem with the endonuclease, either as part of an enzyme complex (restriction/modification complex) or as two distinct entities. Therefore, recognizing that an organism expresses an enzyme with endonuclease activity strongly suggests the expression of an associated modifying methylase enzyme (and vice versa) and this association has led to isolation and cloning of a number of commercially available restriction/modification enzymes for use in the laboratory as discussed below.
One of the limitations iz~ the use of restriction endonucleases exists when cleavage of a given sequence is required and no known endonuclease exists which is specific for that particular sequence. Therefore, the continued identification and isolation of unique restriction endonucleases and altered reaction conditions will allow for even more sophisticated manipulation of DNA in vitro.
A number of publications and patents describe the cloning of DNAs encoding restriction endonucleases. Included among theses publications is Kiss.
A. , et al. , Nucleic Acid Research 13:6403-6421 (1985), which describes the cloned nucleotide sequence of the BsuRI restriction-modification system isolated from Bacillus subtillis. This system is specific for the sequence 5 '-GGCC-3 ' and is defined by two gene products which are transcribed by different promoters.
The methylase component of the system shows homology to the methylase from the BspRI and SPR restriction-modification systems.
Nwanko, D.O. and Wilson, G.G. Gene 64:1-8 (1988), describe the cloning and expression of the MspI restriction and modification genes isolated from Moraxella sp. This system recognizes the sequence 5 '-CCGG-3 ' and both enzymes are functional in E. coli. Evidence indicates that these genes are transcribed in opposite directions, thus are probably under the control of different promoters.
Ashok, K.D., et al., Nucleic Acids Research 20:1579-1585 (1992), describe the purification and characterization of cloned MspI
methyltransferase, over-expressed in E. coli. At low concentrations the enzyme exists as a monomer, but at higher concentrations it exists mainly as a dimer. Polyclonal antibodies to the enzyme cross-react with methyltransferase genes of other modification systems.
Brooks, J.E., et al. Nucleic Acids Research 19:841-850 (1991), characterizes the cloned BamHI restriction modification system from Bacillus subtilis. The two genes are divergently oriented and separated by an open reading - frame which may serve as a transcriptional regulator in the native bacteria.
Slatko, B.E., et al. Nucleic Acids Research 15:9781-9796 (1987), describe tl:~ cloning, sequencing and expression of the TaqI restriction-modification system. These genes have the same transcriptional orientation, with the methylase gene 5 ' to the endonuclease gene. E. coli clones which carry only the endonuclease gene are viable even in the absence of the methylase gene.
This is an unusual case possibly explained by the 65°C optimal temperature for TaqI
restriction and the 37°C optimal temperature for E. coli growth.
Howard, K. A. , et al. , Nucleic Acids Research 14:7939-7951 (1986), describe the cloning of the DdeI restriction modification system from Desulfovibrio desu~ricans by a two step method wherein the methylase gene is first cloned and transformed into E. coli, followed by the cloning of the endonuclease gene and transformation of this second gene into the methylase expressing bacteria. In order to maintain cell viability, high levels of methylase expression are required before the endonuclease gene can be introduced into the bacteria.
Ito, H., et al., Nucleic Acids Research 18:3903-3911 (1990), describe the cloning, nucleotide sequence and expression of the HincII
restriction-modification system. The DNA was isolated from H. influenzae Rc, with the two genes positioned in the same transcriptional orientation.
Shields, S.L., et al., urology 76:16-24 (1990), describe the cloning and sequencing of the cytosine methyltransferase gene M. GwiJI from the Chlorella virus IL-3A. The methylase recognizes the sequence (G/A)GC(T/C/G) and shows amino acid sequence homology with 5-methylcytosine methylases isolated from bacteria. DNA encoding the methylase was obtained from the viral genome which was propagated in the green alga host Chlorella.
Xia, Y., et al., Nucleic Acids Research 15:6075-6090 (1987), discovered that IL-3A virus infection of Chlorella-like green alga induces the expression of the DNA restriction endonuclease CviJI which has novel sequence specificity. This endonuclease recognizes the sequence PuGCPy (wherein Pu =
~159Q~~_ _7_ purine and Py = pyrimidine) but does not cut the sequence PuGmCPy, where mC
is 5-methylcytosine.
U.S. Patent 5,137,823, issued August 11, 1992, to Brooks, 1.E., describes a two step method for cloning the BamHI restriction modification system wherein the methylase is cloned first and then introduced into a bacterial host. The endonuclease is then cloned and introduced into the methylase expressing bacteria. This two step procedure provides the host DNA protection from cleavage of the subsequently introduced endonuclease.
U.S. Patent 5,200,333, ('333) issued April 6, 1993, to Wilson, G.G., describes a method for cloning restriction and modification genes.
Specifically this reference describes the cloning of the TaqI and HaeII
systems from Thermos aquaxicus and Haemophilus aegypticus, respectively. In this method, bacterial DNA was initially purified and digested, and the fragments were then cloned into a vector to produce a bacterial DNA library. The library was then transformed into E. coli and the cells were plated. Colonies were then scraped from the plate to form a primary cell library. Plasmid DNA from this cell library was purified and digested with the endonuclease of the two gene system. Bacteria which expressed the methylase gene had modified plasmid DNA
which was protected from endonuclease activity, while plasmids from bacteria which lacked the intact methylase gene were digested. The resulting, undigested plasmid DNA was then transformed into another bacterial strain and the bacteria were plated. Surviving colonies were again harvested to give a secondary cell library and the entire procedure repeated. Plasmids which code for the complete restriction-modification system presumably survived each round of purification and were enriched. Bacteria which survive several rounds of enrichment were subsequently assayed for both methylase and endonuclease activity.
U.S. Patent 5,196,331, ('331) issued March 23, 1993, to Wilson, G.G. and Nwanko, D., describes a method for cloning the MspI restriction and modification genes. This patent describes a method identical to that of U.S.
wo 9ani663 ,,.., Pt"T/US9a/03246 -8- ~ 2 1 5 9 0 8 '~
Patent 5,200,333 ('333).
As mentioned above, Chlorella virus IL-3A encodes a unique restriction endonuclease called GwiJI (Xia et al. Nucleic Acids Res. 15:6075-S (1987)). IL-3A is a large, polyhedral, plaque-forming phycodnavirus (Francld, R.LB., et al. Arch. Ytrol. suppl.2. Springer-Verlag, Vienna (1991)) that replicates in unicellular, eukaryotic green algae, Chlorella strain NC64A (Schuster, A.M., et al. Yrology 150:170-177 (1986)). The double-stranded DNA genome of IL-3A
is approximately 330 kbp (Rohozinsld et al., Yrology 168:363-369 (1989)) and contains 9.7% methyiated cytidine (Van Etten, J.L. et al., Nucleic Acids Res.
13:3471-3478 (1985)). The cognate methyltransferase of CviJI, M.G~iJI, methylates (A/G)GC(T!C/G) sequences and, has been cloned and sequenced (Shields, S.L. et al., urology 176:16-24 (1990)).
The use of a two/three base recognition endonuclease, such as GwiJI, to improve numerous conventional molecular biology applications as well as permitting novel applications has been described in U.S. Patent No.
5,472,872.
The application discloses methods for generating sequence-specific oligonucleotides from DNA without prior knowledge of the nucleic acid sequence of such DNA, and to methods for cloning and labeling DNA after restriction digestion by a two base recognition endonuclease. The application also teaches methods for generating quasi-random fragments of DNA, methods for cloning, labeling, and sequencing DNA, as well as epitope mapping of proteins. The ability to generate numerous oligonucleotides with perfect sequence specificity or quasi-random distributions of DNA fragments such as is possible with CviJI*
~ ~Port~t implications for a number of conventional and novel molecular .
biology procedures.
Infection of Chlorella species NC64A with the IL-3A virus produces sufficient G~iJI restriction endonuclease (CviJI) for research purposes.
However, production of commercially useful amounts of G~iJI is limited with this B
system due to the slow growth of Chlorella algae, the large number of contaminating Nucleases associated with the virus, and the small yield of enzyme obtained after purification. In addition, biochemical and biophysical characterization of the enzyme, such as molecular weight determination, are difficult from the native source. Because of these limitations it would be useful to clone the gene for G'viJI in order to provide an adequate large scale source of enzyme for use as a molecular biological reagent.
._ SUMMARY OF THE INVENTION.;.' In one of its aspects, the present invention provides purified and isolated polynucleotides (e.g., DNA sequences and RNA transcripts thereof) encoding a unique restriction endonuclease, G'viJI, as well as polypeptides and variants thereof which display activities characteristic of CviJI. Activities of G'viJI
include the recognition of specific DNA sequences, binding to these sequences and cleaving the bound DNA into fragments. Preferred DNA sequences of the invention include viral genomic sequences as well as wholly or partially chemically synthesized DNA sequences. Replicas (i.e., copies of the isolated DNA sequences made in vivo or in vitro) of DNA sequences of the invention are also contemplated. A preferred DNA sequence is set forth in SEQ 1D NO: 2 herein and is contained as an insert in the plasmid pCJHl.4. In another of its aspects, the invention provides purified isolated DNA encoding a G'viJI
polypeptide by means of degenerate colons.
Also provided are autonomously replicating recombinant constructions such as plasmid DNA vectors incorporating CviJI sequences and especially vectors wherein DNA encoding CviJI or a G'viJI variant is operatively linked to an endogenous or exogenous expression control DNA sequence.
According to another aspect of the invention, host cells such as prokaryotic and eukaryotic cells, are stably transformed with DNA sequences of the invention in a manner allowing the desired polypeptides to be expressed 21~ga81 - to -therein. Host cells expressing G'viJI and CviJI variant products are u~ful in methods for the large scale production of G~ilI and G'viJI variants whe:ein the cells are grown in a suitable culture medium and the desired polypeptide products are isolated from the host cells or from the medium in which the cells are grown.
A preferred host cell is E. coli. Still another aspect of the invention is a recombinant G'viJI polypeptide.
The present invention is also directed .to''a method for the digestion of DNA with a restriction endonuclease reagent under conditions wherein said ,, DNA is 'pieaved at a dinucleotide sequence selected from the group consisting of PyGCPy, PuGCPy, PuGCPu, and wherein Pu = purine and Py = pyrimidine.
The present invention is also directed to a method for restriction endonuclease digestion of DNA comprising the step of digesting DNA with a restriction endonuclease reagent under conditions wherein said DNA is digested at 11 of 16 possible dinucleotide sequences and wherein said dinucleotide sequences are selected from the group consisting of PuCGPu, PuCGPy, and PyCGPu, and wherein Pu = purine and Py = pyrimidine.
The present invention is directed to shotgun cloning of DNA, epitope mapping, and for labeling DNA using the digestion methods of the present invention. The present invention provides methods for quasi-random fragmenting of DNA using the digestion methods of the present invention under conditions wherein the DNA is only partially cleaved and the site preference of the restriction endonuclease reagent is greatly reduced. By quasi-random is meant an overlapping population of DNA fragments produced by digesting DNA using the methods of the present inventions without apparent site-preference and which appears as a smear upon electrophoresis in a 1-2 wt. ~ agarose gel. The present invention is also directed to the shotgun cloning and sequencing of quasi-random fragments of DNA produced by the methods of the present invention. Quasi-random fragments in the shotgun cloning method of the present invention are produced by partial digestion of DNA with a restriction endonuclease reagent 21~~08~
according to the methods of the present invention. More particularly, quasi-random fragments of DNA useful in the cloning method of the present invention are produced by the parCial digestion of the DNA to be cloned with CviJ I, BsuR
I or with a restriction endonuclease reagent termed CGase I comprising Taq I
and Hpa II. Quasi-random fragments having a length of between about 100 and about 10,000 nucleotides are preferred. More preferred are quasi-random fragments of about 500 to ab~~ut 10,000 nucleotides in length. The present invention is also directed to the generation of quasi-random fragmentation of DNA using the method of the present invention for the purposes of epitope mapping and gene cloning. These quasi-random fragments are expressed either in vitro or in vivo and the smallest fragment containing the desired function is identified by screening assays well known in the art.
The present invention is also directed to the production of anonymous primers from any DNA without prior knowledge of the nucleotide sequence. The present invention provides methods for anonymous primer cloning and sequencing after complete digestion of DNA utilizing CviJ I, BsuR I or CGase I using th.e methais of the present invention.
Additionally, the present invention is directed to methods of labeling and detecting DrIA comprising the complete digestion of DNA using the methods of the present invention, followed by a heat denaturation step, to yield sequence specific oligonucleoddes. In particular, an aspect of the present invention involves labeling DNA with sequence specific oligonucleotides of about 20 to about 200 bases in length (with an average size of between 20-60 bases) generated by Cvi1 I, BsuR I or CGase I digestion of the template DNA.
More particularly, the invention is directed to restriction generated oligonucleotide lsibeling (RGOL) of DNA which comprises the digestion of an aliquot of template DNA with CviJ I followed by a simple heat denaturation step, thereby generating numerous sequence specific oligonucleotides, which can then be utilized for laheling nucleic acids by a number of methods, including primer 2 15 908' extension type reactions with a DNA polymerise and various labels, isotopic or non-isotopic (RGOL-PEL); 5' end labeling with polynucleotide kinase: 3' end labeling using terminal transferase and various labels,isotopic or non-isotopic.
Labeling at the 3' end, also referred to as tailing, adds numerous labels per oligonucleotide (1-200), depending on the labeling conditions. The addition of 10-S00 oligonucleotides generated per template, results in a significant signal amplification not obtainable by conventional methods.
The invention is also directed to thermal cycle labeling (TCL) which comprises the simultaneous labeling and amplification of probes utilizing CviJ I or CGase I restriction generated oligonucleotides as the starting material.
In this method, natural DNA of unlrnown sequence is digested with CviJ I to generate numerous double-stranded fragments which are then heat denatured to yield oligonucleotides. These oligonucleotides are combined with the intact template and subjected to repeated cycles of denaturation, annealing, and extension in the presence of a thermostable DNA polymerise or functional fragment thereof which maintains polymerise activity, deoxynucleotide triphosphates and the appropriate buffer. Alpha 32P-dATp (or any of the other three deoxynucleotide triphosphates), biotin-dUTP, fluorescein-dUTP, or digoxigenin-dUTP is incorporated during the extension step for subsequent detection purposes. Thermal cycle labeling efficiently labels DNA while simultaneously amplifying large amounts of the labeled probe. In addition, TCL
probes exhibit a 10 fold improvement in detection sensitivity compared to conventional probes.
The present invention is also directed to TCL in which the thermostable DNA polymerise supplies endogenous primers for enrymatic extension. This method is referred to as Universal Thermal Cycle Labeling (LTTCL). In this method natural DNA of unlrnown sequence is combined intact with the holo-enzyme of a thermostable DNA polymerise, deoxyribonucleotide triphosphates, and the appropriate buffer: The holo-enzyme and its associated WO 94/21663 ~ PCTIUS94/03246 endogenous primers are then combined with intact template and subjected to repeated cycles of denaturation annealing and extension. Alpha 32P-dATP, 32P-dTTP, 32P-dGTP, 32P-dCTP, biotin-dUTP, fluorescein-dUTP, or digoxigenin-dUTP is also included in the extension step for subsequent detection purposes.
Isotopic labels useful in the practice of the present invention include but are not limited to 32p, 33p~ 355 14C ~d 3H. Non-isotopic labels useful in the present invention include but are not limited to fluorescein.,biotin, dinitrophenol and digoxigenin.
The present invention is also directed to an improved method for purifying CviJ I from the algae Chlorella infected with the virus IL-3A.
In addition the present invention is directed to restriction endonuclease reagents which, under conditions which relax the sequence specificity of one or more restriction endonucleases, cleave DNA at the dinucleotide sequences AT or TA.
The present invention is also directed to a restriction endonuclease reagent comprising in combination, Taq I and Hpa II, which is capable of digesting DNA at 11 of 16 possible dinucleotide sequences, said sequences selected from the group consisting of PuCGPu, PuCGPy, PyCGPy and PyCGPu, and wherein Pu = purine and Py = pyrimidine.
The following examples are intended to be illustrative of the several aspects of the present invention and are not intended in any way to limit the scope of any aspect of the present invention.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a map of the plasmid p710 which contains DNA
sequences encoding for the IL-3A viral methyltransferase M. CwiJI;
Figure 2 is the nucleotide sequence of 5497 by of cloned IL-3A
viral DNA;
,~ WO 94121663 PCTIUS941031,46 Figure 3 is a restriction map of the cloned IL-3A viral DNA, including the identified open reading frames;
Figure 4 is the DNA sequence of the G~iJI gene with its flanking regions. The predicted amino acid sequence is provided below the nuckeotide sequences;
Figure SA depicts the theoretical frequency and distribution of CvilI * restriction generated oligomers of individual lengths; Figure SB shows the actual frequency and distribution of CviJI* restriction generated oligomers of various lengths;
Figure 6 is a flow chart depicting anonymous primer cloning;
Figure 7 is a photographic reproduction of a gel depicting CbiJI
restriction digests of pUC 19;
Figure 8 is a photographic reproduction of a gel depicting comparisons of sonicated versus G~ilI* partially digested DNAs;
Figure 9A is a photographic reproduction of an agarose gel ekectrophoresis analysis of size-fractionated DNA by microcolumn chromatography compared to fractionation by agarose gel electroelution; In Figure 9A: Lane M, k kb DNA ladder; lane 7v,, untreated ~. DNA (0.25 pg); lane l, unFractionated (UF) CviJI** partially-digested ~, DNA (1.0 pg); lane 2, column-fractionated (CF) CviJI* * partially-digested ~, DNA ( 1.0 pg); lane 3, gel fractionated (GF) CviJI * * partially-digested ~, DNA ( 1.0 pg);
Figure 9B-E illustrates additional trials of the same procedures used in Figure 9A;
Figure l0A illustrates the size distribution of DNA fragments produced by partial digestion of DNA by G~iJI and fractionated by microcolumn chromatography;
Figure lOB-C illustrates the size distribution of DNA fragments produced by partial digestion of DNA by C~ilI and fractionated by agarose gel electrophoresis;
- 14a -Figure 11 is a schematic depiction of the distributibn of G~iJI sites in pUC 19; and Figure 12 is a graph of the rate of sequence accumulation by G~iJ;I** shotgun cloning and sequencing.
f WO 94/21663 ~ 15 9 0 81 PCT~S94103246 DETAILED DESCRIPTION
The gene for the restriction endonuclease R. CviJI was cloned into E. coli so as to provide an adequate source of R. G'viJI for use as a molecular biological reagent. Biologically active G'viJI has been purified from E. coli to apparent homogeneity. The molecular weight of E. coli derived R. CviJI is 32.5 kD by SDS gel electrophoresis. N-terminal amino acid sequence analysis of this protein and comparison to the nucleotide sequence of the gene revealed that the translation of this enzyme is probably initiated with a GTG start colon, instead of the usual ATG initiation colon. The structural gene is 834 nucleotides in length coding for a protein of 278 amino acids (31.6 kD). A second peak of R. G~iJI activity which elutes separately from the 32.5 kD form can be seen in the initial stages of enzyme purification. Trace amounts of a larger molecular weight form have not been observed to date. However, the R. G~iJI gene does possess an in-frame upstream ATG colon which if translated would yield a predicted 41.4 kD protein. The structural gene for this potentially larger product is 1074 nucleotides in length coding for a putative protein of 358 amino acids.
The present invention is also directed to a method for the fragmentation and cloning of DNA using the restriction endonuclease CviJ I
under conditions which allow the enzyme to cleave DNA at the dinucleotide sequence GC. In addition, the present invention is also directed to the cloning of quasi-random fragments of DNA digested using the fragmentation method of the present invention.
As an alternative to the methods for constructing random clone libraries described above, methods were devised for the construction of such libraries which require fewer steps and reagents, which require smaller amounts of DNA, which have relatively high cloning efficiencies and which takes less time to complete. These methods relate to the recognition that a partial digest with a two or three base recognition endonuclease cleaves DNA frequently enough to be functionally random with respect to the rate at which sequence data may be WO 94/21663 ~ PCT/US94/03246 accumulated from a shotgun clone bank. The restriction enzyme CviJ I normally recognizes the sequence PuGCPy and cleaves between the G and C to leave blunt ends (Xia et al., Nucl. Acids Res. 15:6075-6090 (1987)). Under "relaxed"
conditions (in the presence of 1 mM ATP and '20 MM DTT) the specificity of CviJ I can be altered to cleave DNA more frequently and perhaps as frequently as at every GC. This activity is referred to as CviJ I*. Because of the high frequency of the dinucleotide GC in all DNA (16 by average fragment size for random DNA), quasi-random libraries may be constructed by partial digestion of DNA with CviJ I*. A DNA degradation method with low levels of sequence specificity produces a smear of the target DNA when analyzed by agarose gel electrophoresis. Digestion of the plasmid pUC 19 under partial CviJ I*
conditions does not result in a non-discrete smear; rather, a number of discrete bands are found superimposed upon a light background of smearing, suggesting that CviJ
I* has some site preference. Atypical reaction conditions according to the present invention eliminate this apparent site preference of CviJ I* to produce an activity (termed Cvi~ I**) in combination with a rapid gel filtration size exclusion step, streamlines a number of aspects involved in shotgun cloning.
One aspect of the present invention involves the use of the two/three base recognition endonuclease Cvi1 I, in conjunction with a simple spin column method to produce libraries equivalent in final form to those generated by the combination of sonication and agarose gel electroelution. However, the method of the present invention requires fewer steps, a shorter time period, and significantly less substrate (nanogram amounts) when compared to conventional procedures. Both small and large sequencing projects using the methods described herein are within the scope of the present invention.
Current sequencing paradigms require the generation of a new template for each 350-500 nucleotides sequenced. On this basis, sequencing both strands of the human genome would require at least 12 million templates 500 nucleotides long, assuming no overlap between templates.
2 15 908 ~
A random approach, such as shotgun sequencing, would require 30 to 50 million templates, assuming the entire genome were randomly subcloned.
As many as 250,000 libraries may be needed to generate the requisite templates from a subcloned and ordered array of this genome, depending on the type of vector utilized, and the degree of overlap between such clones. The ability to generate shotgun libraries in a semi-automated, microtiter plate format would greatly simplify such large scale projects.
The development of methods for cloning large DNA molecules in yeast artificial chromosomes (Burke et al., Science 236:806-812 (1987), or in bacteriophage P1-derived vectors (Sternberg, Proc. Natl. Acad. Sci. LISA
87:103 107 (1990)), simplifies the subdivision and analysis of very large genomes.
However, the large size of the resulting subclones (100 - 1000 kbp) presents additional challenges for subsequent sequencing efforts. A report of the sequencing of a 134 kbp genome by random shotgun cloning directly into a bacteriophage M13 vector indicates that numerous intermediate stages of subcloning, mapping, and overlapping such clones may be eliminated (Davison, J. DNA Seq. and Mapping 1:389-394 (1992). An order of magnitude reduction in the amount of DNA required for shotgun cloning would substantially simplify efforts to directly sequence 100,000 by sized molecules and beyond.
The ability to generate an overlapping- population of randomly fragmented DNA molecules is considered essential for minimizing the closure of nucleotide sequence gaps by the shotgun cloning method. The use of a very frequent-cutting restriction enzyme, such as CviJ I, is an approach which has not been utilized. Reaction conditions according to the present invention result in the quasirandom restriction of pUCl9 and lambda DNA, as judged by the degree of smearing observed.
The randomness of this CviJ I** reaction was quantified by sequence analysis of 76 such partially-fragmented pUCl9 subclones. The analysis showed that CviJ I* * partial digestion (limiting enzyme and time) restricts DNA
WO 94/21663 PCTIUS94/03246 .,..,,~
at PyGCPy, PuGCPu, and PuGCPy (but not PyGCPu), and is thus a hybrid reaction which combines the three base recognition specifity of CviJ I with the "two" base recognition specifity of CviJ I*. Interestingly, most of the "relaxed"
cleavage observed under CviJ I** conditions occurred in those portions of the sequence which were deficient in "normal" restriction sites. CviJ I**
treatment produces a relatively uniform size distribution of DNA fragments, permitting sequence information to be accumulated in a statistically random fashion.
Shotgun cloning with CviJ I** digested DNA is efficient partly because the resulting fragments are blunt ended. Other methods currently used to randomly-fragment DNA, including sonication, DNAse I treatment, and low pressure shearing, leave ragged ends which must be converted to blunt ends for efficient vector ligation. Other than a heat denaturation step to inactivate the endonuclease, no additional treatments are required for cloning CviJ I**
restricted DNA. In addition, the preligation step required to equalize representation of the ends of a DNA molecule prior to sonicadon or DNAse I treatment is not necessary with CviJ I** fragmentation. CviJ I* cleaves its cognate recognition site very close to the ends of a linear molecule, as judged by the very small fragments resulting from complete digestion of pUC 19 as depicted in Figure 2, lane 1.
The overall efficiency of shotgun cloning depends not only on the fragmentation process, but also upon the size fractionation procedure used to remove small DNA fragments. The efficiency of cloning agarose gel fractionated DNA was found to be unexpectedly variable. Numerous experiments produced an erratic distribution of sized material and the resulting cloned inserts were uniformly small (70% < 500 by in one trial, 100 % < 500 by in another). The method of the present invention includes a simple and rapid micro-column fractionation method, which has resulted in three to thirteen times more transformants than agarose gel fractionation. More importantly, the size distribution of the cloned inserts from column-fractionated DNA was skewed toward larger fragments (88 % > 500 bp). Micro-column fractionation also eliminates the chemical extraction steps required for agarose fractionated DNA.
After the target DNA has been column-fractionated, no further treatments are required for cloning. Combining CviJ I** partial restriction with micro-column fractionation permits the construction of useful libraries from as little as 200 ng of substrate, an order of magnitude less starting material than recommended for sonication/end-repair and agarose gel fractionation procedures.
The CviJ I** reaction represents a unique alternative for controlling the partial digestion of DNA, a technique which is fundamental to the construction of genomic libraries (Maniatis et al. Cell 15:687-701 (1978), and restriction site mapping of recombinant clones (Smith, et al. Nucl. Acids Res. 3:2387-2398 (1976). Partial DNA digests are notably variable and are strongly dependent on the concentration and purity of the DNA, the amount of enzyme used, the incubation time, and the batch of enzyme. Partial digestions may also be variable with respect to the rate at which a particular recognition sequence is cleaved throughout the substrate. Optimal reaction conditions, such as those which render such partial digests independent of one or more of these variables, allows more precise control of the end product. Several controlling schemes may be employed, including: the addition of a constant amount of carrier DNA (Kohara et al., Cell 50:495-508 (1987)), the use of limiting amounts of Mg2+
(Albertson et al. Nucl. Acids Res. 17:808 (1989)), ultraviolet irradiation (Whitaker, et al.
Gene 41:129-134), and the combination of a restriction enzyme and a sequence complementary DNA methylase (Hoheisel et al., Nucl. Acids Res. 17:9571-9582 (1989)). Utilizing three different batches of CviJ I, and three different DNA
templates from five separate preparations, a uniform CviJ I** partial digestion pattern was obtained that was primarily time-dependent when a constant ratio of 0.3 units of enzyme per ~cg of DNA was used.
The rate at which a particular restriction site is cleaved at different locations in a substrate is variable for many endonucleases (Brooks, et al., WO 94121663 PCTIUS94/03246 ..."
'~159~$
Methods in Enrymol, 152:113-129 (1987)). Reaction conditions for CviJ I may be optimized to substantially reduce the site preferences of this enzyme during partial digestion (see Figure 2, lanes 3 and 4). Normally, "star" reaction conditions result in cleavage at new sites. The use of star reaction conditions according to the present invention (dimethyl sulfoxide [DMSO] and lowered ionic strength) to affect the partial digestion activity o~~CviJ I** does not result in an altered restriction site cleavage as assayed by sequencing the products of 76 digestion reactions. Instead, the relative rate of cleavage of individual sites appears to be more uniform under these conditions. A 3-5 fold increase in the rate of normal Cvi1 I restriction with the standard buffer and DMSO further substantiates this approach. All of these results indicate that, under the appropriate reaction conditions, CviJ I is useful for a number of other applications, such as high resolution restriction mapping and fingerprinting, diagnostic restriction of small PCR fragments, and construction of genomic DNA
libraries.
Another aspect of the present invention involves quasi-random fragmentation of DNA using the method of the present invention for epitope mapping and cloning intact genes. The same method as described above for shotgun cloning is utilized, except that an expression vector is used to generate functional proteins from the DNA.
Another aspect of the present invention involves fragmenting DNA
using the present invention to generate multiple oligonucleotides from any double-stranded DNA template. Restriction-generated oligonucleotides (RGO) are sequence specific oligonucleotides generated from any DNA according to the present invention. CviJ I* presumably cleaves the recognition sequence GC
between the G and C to leave blunt ends (Xia et al., Nucl. Acids Res. 15:6075-6090, (1987)). Because of the high frequency of dinucleotide GC in all DNA
(l6bp average fragment size for random DNA), a complete CviJ I* restriction results in numerous fragments which are about 20-200 by in size. These 2i~908~.
restriction fragments ire generated from an aliquot of the template itself and are heat-denatured to yield numerous single-stranded oligonucleotides which are of variable length but which are specific for the cognate template. Complete CviJ
I* restriction of the small plasmid pUCl9 (2689 bp) theoretically yields 314 oligonucleotides after a heat-denaturation step. The ability to generate numerous oligonucleotides with perfect sequence specificity is an unusual result of the use of this class of enzyme according to the present invention. Such oligonucleotides are uniquely suited for purposes of labeling DNA, as described below.
One application of CviJ I* restriction-generated oligonucleotides is to directly label them using conventional methods. There are several important advantages in using Cvi1 I* restriction-generated oligonucleotides.
Conventional methods employing synthetic oligonucleotides for detection purposes generally use one oligonucleotide containing one or a few labels. A complete CviJ I* digest generates hundreds of oligonucleotides from a given template, depending on the size of the template, and thus makes hundreds of sites available for labeling, regardless of the labeling scheme utilized. These hundreds of sequence specific restriction-generated oligonucleotides have two important advantages over conventional probes used in nucleic acid detection methods. First, the generation of multiple oligonucleotide probes directed at multiple sites in a given target (theoretically, 314 sites in pUC 19) provides enhanced detection sensitivities compared to synthetic oligonucleotides which are directed at 1 or a few sites in a target. The numerous labeled restriction-generated oligonucleotides represent a 10-100 fold amplification of the signal for detection compared to the use of a single oligonucleotide. Second, the short length of the restriction-generated oligonucleotides permits more efficient hybridization. This is important for two reasons. First, hybridization times using restriction-generated oligonucleotides is reduced to 1 hr as opposed to an overnight incubation with conventional probes hundreds of nucleotides in length. This is a very important advantage when using oligonucleotide probes in clinical settings. Second, the penetration of probes into ~1~908~.
permeabilized cells is a critical issue for in situ hybridization procedures.
The smaller the probe, the easier the entry into the cell. Thus, the use of multiple oligonucleotide probes generated by the two base cutters greatly improves the sensitivity of in situ hybridization, a technique's of considerable importance in research and clinical labs. Finally, whenusing membrane-based hybridization procedures, only small sections of a target nucleic acid are exposed and available for hybridization. Multiple oligonucleotides derived from a cognate template exhibit better detection sensitivities compared to long probes.
Another application of restriction-generated oligonucleotides for labeling is to employ them as primers in a polymerase extension labeling reaction in conjunction with a repetitive thermal cycling regimen of denaturation, annealing, and extension. Thermal Cycle Labeling (TCL) is a method for efficiently labeling double-stranded DNA while simultaneously amplifying large amounts of the labeled probe. The TCL system employs the two base recognition endonuclease CviJ I* to generate sequence-specific oligonucleotides from the template DNA itself. These oligonucleotides are combined with the intact template and subjected to repeated cycles of denaturation, annealing, and extension by a thermostable DNA polymerase from, for example, Thermus,flavus.
A radioactive- or non-isotopically-labeled deoxynucleotide triphosphate is incorporated during the extension step for subsequent detection purposes. The amplified, labeled probes represent a very heterogeneous mixture of fragments, which appears as a large molecular weight smear when analyzed by agarose gel electrophoresis. Primer-primer amplification, a side product of this reaction (produced by leaving out the intact template in the TCL reaction), may result in enhanced detection sensitivity, perhaps by forming branched structures. Biotin-labeled probes generated by the TCL protocol detect as little as 25 zeptomoles (2.5 x 10-20 moles) of a target sequence. A 50 ~.1 TCL reaction yields as much as 25 ~cg of labeled DNA, enough to probe 25 to 50 Southern blots. After 20 cycles of denaturation and extension, biotin-dUTP-incorporated TCL probes may WO 94/21663 ~ PCT/US94/03246 be routinely detected at a 1:106 dilution, which is 1000 fold more sensitive than RPL, and indicates that a significant degree of net synthesis or amplification of the probe is occurring. In addition, non-isotopically-labeled TCL probes exhibit a 10-fold improvement in detection sensitivity when compared to RPL-generated probes. 32P-labeled probes generated by the TCL protocol may also detect as little as 50 zeptomoles (2.5 x10-20 moles) of a target sequence. As little as pg of template DNA is enough to synthesize 5-10 ng of radioactive version of TCL generates probes having extremely high specific activities, e.g. (about 5 x 109 cpm/~cg DNA), which permits 5 to 10-fold lower detection limits than conventional labeling protocols.
There are several advantages to using restriction-generated oligonucleotides for primer extension labeling of DNA. One advantage is the specificity of the primers. All of the oligonucleotides generated by the TCL
system are specific for the template utilized, unlike random primer labeling (RPL) which utilizes synthetic oligonucleotides 6-9 bases in length having a random sequence: The amount of primer required for efficient labeling with the TCL
system is only 10 ng, compared to the 10 ~cg of random primers utilized for RPL.
Due to their short length, random primers anneal very inefficiently above 25-37°C, thus RPL is limited to DNA polymerases such as Klenow or T7. The size of the restriction-generated oligonucleotides are longer than the random primers, which extends the hybridization and extension conditions to include a wide variety of temperatures and polymerases. Thus, the use of the restriction-generated sequence-specific oligonucleotides results in more efficient hybridization and extension as compared to RPL. The TCL system has been optimized for labeling with a thermostable DNA polymerase which allows the option of temperature cycling. After 20 cycles of denaturation and extension, a significant amount of amplified TCL probes can be generated. Most importantly, TCL-labeled probes exhibit a 10 fold improvement in defections sensitivity when compared to RPL-generated probes.
Another aspect of the present invention involves a variation of TCL
called Universal Thermal Cycle Labelling (LJTCL) in which the extension primers are not supplied by CviJI restriction, but rather, are found endogenously in the enzyme preparations of thermostable DNA polyrnerases. Random sequence DNA
is usually co-purified along with the holo-e~yme preparation of the thermostable DNA polymerises, regardless of the source of the enzyme, i.e. native or cloned.
However, only the holo-enzyme, and not the exonuclease minus deletion variants, contain the endogenous DNA. Typically, when the bolo-enzymes of thermostable polymerises are used in protocols such as the polymerise chain reaction, the presence of such primers can create spurious results. Methods for circumventing the problems of endogenous DNA are described in PCR Protocols: A Guide to Methods and Applications, Eds. M. Innis, et al. , Academic Press, 1990.
This residual DNA is rather short (approximately 5-25 bases), as assayed by end-labeling with y32P[ATP] and polynucleotide ldnase and acts as endogenous "random" primers in a TCL-type reaction. UTCL combines the holo enzyme of a thermostable polymerise from, for example, Thermos , flavus, with the intact DNA template and is subjected to repeated cycles of denaturation, annealing, and extension. A radioactive- or non-isotopically-labeled deoxynucleotide triphosphate is incorporated during the extension step for subsequent detection purposes. The amplified, labeled probe represents a very heterogenous mixture of fragments, which appears as a large molecular weight smear when analyzed by agarose gel electrophoresis. Biotin-labeled probes generated by the UTCL protocol detect as little as 25 zeptomoles (2.5 x 10-20 moles) of a target sequence. A 15 ul UTCL reaction yields as much as 5-10 ~cg of labeled DNA, enough to probe 5 to 10 Southern blots. After 20 cycles of denaturation and extension, biotin-dUTP-incorporated UTCL probes may be routinely detected at a 1:106 dilution, which is 1000 fold more sensitive than RPL, and indicates that a significant degree of net synthesis or amplification of the probe is occurring. In addition, non-isotopically-labeled UTCL probes exhibit a 10-fold improvement in detection sensitivity when compared to RPL-generated probes. 32P-labeled probes generated by the LTTCL protocol may also detect as little as 50 zeptomoles (2.5 x10-20 moles) of a target sequence. The radioactive version of UTCL generates probes having extremely high specific activities, e.g.
(about 5 x 109 cpm/~cg DNA), which permits 5 to 10-fold lower detection limits than conventional labeling protocols.
The present invention is illustrated by the following examples relating to the isolation of a full length viral DNA clone encoding R. CviJI, to the expression of R. G~iJI DNA in E. coli strain DI~iSaF 'MCR and to purification of R. G'vilI from this bacterial stain. More particularly, Example 1 provides for the propagation of IL-3A virus and isolation of viral genomic DNA. Example 2 addresses the improved expression of a clone for the viral methylase M. CviJI
.
Example 3 describes the strategy for isolating and cloning the viral R. G'viJI
gene by a forced co-cloning strategy of the M. GwiJI gene. Example 4 describes the sequencing of cloned IL-3A genomic DNA and identification of the R. G~iJI
gene.
Example 5 relates the methods for purification of C~iJI to homogeneity from an E.coli strain, DHSaF'MCR, transformed with a plasmid which encodes the R. G~iJI enzyme. Example 6 details the amino acid sequence analysis of the purified R. G'viJI enzyme. Example 7 describes the analysis of G~iJI*
recognition sequences. Example 8 relates to a technique for producing restriction generated oligonucleotides using CviJI. Example 9 relates the generation of anonymous primers using CviJI. Example 10 describes end-labeling of G'viJI restriction generated oligonucleotides. Example 11 describes primer extension labeling of DNA using restriction generated oligonucleotides. Example 12 relates the use of CviJI in thermal cycle labeling of DNA as well as the method of universal thermal cycle labelling. Example 13 provides a method for generation of quasi-random DNA fragments using CviJI. Example 14 describes fractionation of G~iJI
digested DNA by size using spin column chromatography. Example 15 details the relative cloning efficiency of G'viJI digested, size-fractionated DNA by gel elution and ,",.,. WO 94121663 PC'T/US94103246 chromatographic methods. Example 16 describes the ~.omparison of cloning efficiency using lambda DNA fragmented by both sonication and G~iJI**
techniques. Example 17 details the use of G~iJI** fragmentation for shotgun cloning and sequencing. Example 18 describes the shotgun cloning of lambda DNA using f~iJI. Example 19 describes the use of CviJI in epitope mapping techniques. Example 20 describes the restriction endonuclease reagent CGase I.
Example 1 Propagation of IL-3A Virus The exsymbiotic Chlorella-like alga, NC64A, originally isolated from Paramecium bursaria (Karakashian, S.J. and Karakashian, M.W., Evolution and Symbiosis in the Genus Chlorella and Related Algae. Evolution 19:368-377 ( 1965)), , was grown and maintained in Bolds basal medium (BBM), (Nichols, H.W. and Bold, H.C. J. Phycol. 1:34-38 (1965)) modified by the addition of 0.5 ~6 sucrose, 0.1 % protease peptone, and 20 ug/ml tetracycline (IvIBBM).
Cultures were innoculated with 1 X 106 algae cells/ml and grown at 25°C
in 250 ml of MBBM in 500 ml Erlenmeyer flasks on a rotary shaker (150 rpm) in continuous light (ca. 30 ~Ei, m-2,sec 1). Growth was monitored by light scattering measured as A~~m and/or by direct cell counts with a hemocytometer.
When the cultures reached approximately 1 X 107 algae cells/ml they were innoculated with filter sterilized (0.4 ~cm nitrocellulose filter, Nucleopore, Pleasanton; California) IL-3A virus at a multiplicity of infection of 0.01 and incubated for an additional 48 - 72 hours at 25°C. The crude lysate was then centrifuged at 3000 rpm (2000 xg) for 10 minutes to remove cellular debris.
Nonidet P-40TM was then added to 1 % {v/v) and the virus was pelleted from the supernatant by centrifuging at 15,000 rpm at 4°C for 75 minutes in a Beckman No. 30 rotor. The viral pellet was gently resuspended in O.OS M Tris-HCI, pH
a WO 94/21663 21 ~ g ~ ~ ~ PCT/US94/03246 ",...
7.8, and the sample was layered on linear 10 - 40% sucrose gradients equilibrated with 0.05 M Tris-HCl, pH 7.$, and centrifuged for 20 minutes at 20,000 rpm at 4°C in a Beckman SW28 rotor. The viral band, which was present in the center of the gradient as an opaque band, was removed, diluted with 0.05 M Tris-HCI, pH 7.8, and pelleted by centrifugation at 15,000 rpm at 4°C for 120 minutes in a Beckman No. 80 rotor. The virus was resuspended in a small volume (lOml) of 0.05 M Tris-HCI, pH 7.8, and stored at 4°C.
IL-3A viral DNA was purified from the viral particles using a modification of the protocol described by (Miller, S.A., Dykes, D.D., and Polesky, H.L, Nucleic Acids Res. 16:1215 (1988)). Briefly, 100 ~1 of IL-3A
virus (9.8 X 1011 plaque forming units/ml) was diluted with 400 ~cl of water and then mixed with 10 ~,l TEN (0.5 M Tris-HCI, pH 9.0, 20 n~Ivl EDTA, 10 mM
NaCI) and 10 ~1 of 10% SDS. After incubating at 70°C for 30 minutes the solution was extracted twice with phenol-chloroform-isoamyl alcohol, extracted once with chloroform, and precipitated with ice-cold ethanol using methods well known in the art and resuspended in 500 ~d of H20. (Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (Fds.) (1987) Current Protocols in Molecular Biology, Wiley, New York; Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989), Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York).
Example 2 CviJI Methyltransferase Clone The CviJI methyltransferase gene (M. CviJI) from Chlorella virus IL-3A was cloned and sequenced by Shields et al., urology 176:16-24 (1990).
Briefly, Sau3A partial digest of Chlorella virus IL-3A was ligated to BamHI
digested pUCl9 and transformed into E. coli strain RR1. This library of plasmids was restricted with HindIII (AAGCTT) and SstI (GAGCTC), both of which are ~.~5gp81 inhibited by 5-methylcytidine (5mC) in the AGCT portion of their recognition sequences, and transformed again into RR1 cells. M.CviJI methylates the internal cytidine in (G/A)GC(T/C/G) sequences. If the M. CviJI gene is cloned and expressed appropriately, the plasmid DNA would be expected to be resistant to HindlB and SstI restriction.
The CviJI methyltransferase gene was originally cloned as a 7.2 kb insert, termed pIL-3A.22. Plasmid pIL-3A.22 was only partially resistant to CviJI
digestion. Partial digestion is most likely due to the inefficient expression of the M. CviJI gene and the numerous CviJI sites in both the vector (pUC 19 has 45 CviJI sites) and in the insert DNA. The M. CviJI gene was eventually sublocalized to a region of 3.7 kb by subcloning using methods well known in the art (Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (Eds.) (1987) Current Protocols in Molecular Biology, Wiley, New York; Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989), Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York ) and testing the subcloned DNA for sensitivity/resistance to HindIII, SstI, and CviJI. (Shields et al. , supra) The entire sequence was determined and three open reading frames which could code for polypeptides 161, 367, and 162 amino acids, respectively, were identified.
The 367 amino acid open reading frame (ORF) was identified as the M. CviJI
gene by three criteria: (i) it is the only ORF located in the region identified by transposon mutagenesis; (ii) it has amino acid motifs similar to those of other cytosine methyltransferases; and (iii) a 1.6 kb DraI fragment containing the amino acid ORF (1101 bp) produces the methyltransferase. This 1.6 kb M.CviJI
encoding fragment was subcloned into the EcoRV site of pBluescript KS(-) (Stratagene, LaJolla, CA), in the same transladonal orientation as the IacZ' gene of this vector. A physical map of the resulting plasmid termed p710 is shown in Figure 1.
~".., w 2~5908~
The plasmid p710 was digested with several endonucleases to indirectly test the efficiency of M. G'viJI expression. Fully active methylase should render the plasmid DNA completely resistant to digestion by the following enzymes: HaeIII (which recognizes the sequence GGCC), SacI (which recognizes S the sequence GAGCTC), and HindIII (which recognizes the sequence AAGCTT).
The plasmid was partially resistant to HaeIII (90 % ) and SacI (90 % ), and even less resistant to HindIll (25 ~O) digestion. This lack of complete protection of the plasmid DNA made it impractical to attempt cloning the three/two base restriction endonuclease encoded by the R. G~iJI gene. Thus, improvements in the efficiency of M. G'viJI expression were required before attempting to clone the R. G'viJI
gene.
The translation efficiency of the M. CviJI gene was improved by removing extraneous 5 ' open reading frames, creating a perfect fusion of the lacZ' Shine-Delgarno sequence with the methyltransferase start colon (see Figure 1). This was achieved by site-specific oligonucleotide mutagenesis, using the oligomer 5 '-CAATTTCACACAGGAAACAGCTATGTCTIZ'TCGCACGTTAGAAC-3 ' (SEQ ID NO: 1) to precisely remove the intervening lacZ' DNA. The relevant DNA sequences are indicated in Figure 1 (SEQ ID N0:12). The mutagenesis was facilitated by converting the double stranded plasmid DNA of p710 to single-stranded DNA by co-infecting the E. coli host strain with the helper phage (Russel, M., Kidd, S. and Kelly, M.R. Gene 45:333-338), using methods well lmown in the art. The mutagenesis reaction was completed using a commercially available kit according to the manufacturer's instruction (Mutagene, Bio-Rad, Hercules, California). The oligonucleotide was annealed to the single-stranded plasmid, extended in the presence of T4 DNA polymerise, ligated using T4 DNA
ligase, and transformed into competent SURE'" cells (Stratagene, La Jolla, California). Transformed cells were then grown overnight as a pool, the DNA
isolated and purified.
2~5908~.
Enrichment for the mutagenized plasmids was made possible by virtue of the loss of an XhoI site located in the sequence that was deleted by mutagenesis. Enrichment was accomplished by digesting the isolated, purified plasmid DNA with XhoI, followed by dephosphorylation with calf intestinal alkaline phosphatase (CIAP), and transformed into SURE cells. Plasmid DNA
was isolated from 18 individual colonies and the DNA tested for resistance to XhoI. Plasmid DNA from 11 colonies were resistant to XhoI digestion, indicating that they lacked the deleted sequence. Five of these plasmids were restricted with HaeIII, HindIII, PvuII (which recognizes the sequence CAGCTG), and CviJI. All five appeared 100% resistant to these enzymes. Four of the plasmids were sequenced and the deletion was confirmed as being correct. One of these, pBMCS, was chosen for further modification.
Example 3 Forced Co-Cloning of R.Cvi,TI
The location of the R. CviJI gene on the IL-3A virus genome was inferred as being 3' to the M.CviJI gene for two reasons: 1) the cloned DNA
sequence S ' to the M. CvilI gene did not produce a restriction activity; and 2) several attempts to clone the DNA 3 ' to the M. CviJI gene resulted in deletions/rearrangements of this downstream region. This information permitted a forced co-cloning strategy to obtain the restriction endonuclease gene. This strategy uses a deletion derivative of pBMCS lacking the 3 ' half of the M.
CviJI
gene. Digestion of the IL-3A genome with the same enzyme used to create the M. CviJI deletion, followed by ligation of the respective DNAs, transformation, and digestion with enzymes incapable of recognizing methylated DNA (e. g. , HaeIII, HindIII, PvuII, CviJI, etc.) should force the selection of clones which have a restored M. CviJI gene (and thus active methylase enzyme), as well as downstream DNA. Thus, if a clone is found to be CviJI resistant, the 3 ' half of _ 21~90~~
M. CviJI must hove been restored, and downstream DNA containing the R. CviJI
gene, at least in part, would presumably be cloned.
Tile details of this cloning strategy are as follows. pBMCS has two EcoRI sites, one approxirnately in the middle of the M. CviJI gene, while the other site lies in the vextor DNA, 3 ' to the M. CviJI gene (see Figure 1). pBMCS
was restricted with EcoRI andl ligated at a dilute concentration (10-50 ng/~cl) to favor circularization without the 3 ' M. CviJI fragment. The reaction mixture was then transformed into compevtent SURE cells and plated on TY agar containing ampicillin. Plas:mid DNA from the resulting colonies was tested for the lack of this EcoRI fragment by dligestion with EcoRI. One of these clones, pBMCSRI, was used for the subsequent co-cloning work. Plasmid pBMCSRI was digested with EcoRI and dephosphorylated using CIAP. IL-3A genomic DNA was then digested to comFrletion with EcoRI. The EcoRI digested pBMCSRI and IL-3A
DNAs were combined at a ratio of 1:3 in a ligation reaction using T4 DNA
ligase, and the products of the ligation reaction were subsequen.e. used to transform competent SURE cells. The pBMCSRI/IL-3A transforman« were not plated, but rather grown overnight in culture as a library or pool of cells.
The cells were harvested the next day and DNA was isolated and purified. Isolated, p-~rified DNA was dige:>ted with HaeIII, dephosphorylated with CIAP, and tr: ~sformed into competent SURE cells. The cells were then plated and grown overnight. Six colonies grew, of which only one containing the plasmid, pCJHl.4, was resistant to HaeIII. The plasmid pCJHl.4 was found to encode CviJI restriction activity. Plasmid pCJHl.4 was further characterized to localize the gene for CviTf by deletion analysis, subcloning experiments, and sequencing.
The plasmid pCJ131.4 was deposited with the American Type Culture Collection on June 30, 1993 under Accession Number 69341.
~,~ WO 94121663 PCTIUS94/03246 2 15 908 ~
Fxample 4 Sequencing of Cloned II~3A DNA Containing CviJ: Gene The LcoRI fragment cloned into pCJHl.4 jas described in Example 3) is 4901 by in length. Except for the 519 by corresponding to the 3 ' portion of the M. CwiJI gene, the remainder of the 4901 by EcoR I fragment cloned into pCJHl.4 was sequenced using the SEQUALTM DNA Sequencing System (CHIMERx, Madison, WI) by methods well known in the art. Sequencing was accomplished using three approaches: 1) primer walking on pCHJl.4, 2) cloning various restriction endonuclease digests of pCHJl.4 into an M13 type sequencing i0 vector; and 3) sequencing various restriction endonuclease deletion derivatives of pCHJl.4. The nucleotide sequence of 5497 by of IL-3A viral DNA is shown in Figure 2 and set forth in SEQ ID NO.: 2.
Six open reading frames (ORF) of 1155 by (ORF1), 468 by (ORF2), 555 by (ORF3), 1086 by (ORF4), 397 by (ORFS) and 580 by (ORF6) which could code for polypeptides containing 358 (41.4 kD), 156 (19.4 kD), 185 (20.3 kD), 362 (38.9 kD), 132 (14.5 kD) and 193 (21.9 kD) amino acids, respectively, were identified (see Figure 3). ORFs 4-6 do not code for the R. G'viJI gene, as the deletion derivative pCdAl2, which lacks the DNA between the AvaI and BamHI sites (see Figure 3), does produce C~iJI restriction endonuclease activity. In addition, the deletion derivative pCdEB7, lacking the DNA between the EcoRI and BamHI sites, did not produce Gl~iJI activity. Thus ORF 1 or ORF3 were the most likely candidates for encoding the R. G~iJI gene.
The sequence of the 1155 by ORF1 (SEQ ID NO: 3), its deduced amino acid sequence (SEQ ID NO: 4) (as shown in capital letters), plus flanking bases, is presented in Figure 4. The vertical line in Figure 4 and the associated arrow indicate where the DNA sequence from pJCHl.4 diverges from that of pIL-3A.22-8 (Shields, S.L., et al., Yrology 76:16-24, 1990). This open reading (ORF1) frame is believed to represent the C~iJI gene because 14 out of 15 N-z~
terminal amino acids from the protein sequence (see Example 6) matched the predicted translation product of the nucleic acid sequence (Figure 4). Also, the 32.5 kD molecular weil;ht of the homogeneously purified enzyme described in Example 5 matched the predicted translation product of the nucleic acid sequence (31.6 kD) if the encodE:d protein was translated beginning at the GTG colon located at nucle~~tides 299 - 301 (Figure 4), instead of the 5 ' ATG colon located at nucleotides 59 - 61. This possibility is not surprising in light of the fact that approximately 10 % of prokaryotic and eukaryotic gene products begin translation with a GTG start colon, rather than the usual ATG colon (Kozak, M., Microbiol.
Rev. 47:1-45 (1983); K:ozak, M. J. Cell.Biol. 108:229 (1989); Gold, L. et al., Annu.Rev.Micrc~biol. 35:365-403 (1981)). The structural gene was identified to be 834 nucleotides in length, coding for a protein of 278 amino acids (31.6 kD) and is set forth in SEQ ID NO: 4. It is also interesting to note that the CwiJI gene was shown to possess an in-frame, upstream ATG colon which if translated could yield a protein v~rith a predicted molecular weight of 41.4 kD (Figure 4). A
larger molecular weight form pbssessing CviJI restriction activity has not been detected by SDS gel ela~trophorcais. However, a second peak of C'viJI activity which eluted separately from the 32.5 kD form was detected in the initial stages of enzyme purification. The DNA sequence which could theoretically code for a larger form of C~iJI wound be approximately 1074 nucleotides in length (assuming it starts at the upstream ATG colon) and would code for a protein of 358 amino acids.
Example 5 Purification of Recombinant CviJI Restriction Endonuclease Inutially, f.0 ml of LB medium (plus 100 ~cg/ml ampicillin) were inoculated with a 1 ml stock of E. coli transformed with the plasmid pCJHl.4 described above and grovwn overnight at 37°C with shaking. The next day, 20 ml __.. , __.__.~...-_.,..-._ __.__._.___.
.-. W0 94/21663 PCT/US94103246 of this initial overnight culture was used to inoculate another 1 liter of LB
medium and grown overnight. The following day, 50 liters of TB medium (12 g Bacto-Tryptone, 24 g Bacto Yeast Extract, 4 ml glycerol, 2.31 g KH2P04, 12.54 g K2HP04, 0.1 g MgS04, 100 ~cg/ml ampicillin, and water to 1 liter) were inoculated with an aliquot of the secondary overnight culture and grown at 37°C
with 20 liters/min aeration at 200 RPM, until the OD595nm ached 1.0 unit.
Vigorous aeration was essential for G'viJI expression and a typical yield contained 70 g of cell paste after centrifugation.
The cell pellet was immediately resuspended in lysis buffer A
(30 mM Tris-HCI, pH 7.9 at 4°C, 2 mM EDTA, 10 mM beta-mercaptoethanol, SO ~cg/ml phenylmethylsulfonyl fluoride (PMSF), 20 ~cg/ml benzamidine, 2 ~g/ml 0-phenanthroline, 0.7 p,g/ml pepstatin) at a volume of 3 ml of buffer A per 1 g of cells. The cell suspension was then passed through a Manton-Gaulin cell disrupter (Gaulin Corporation, Everett, MA) twice and centrifuged for 1 hr (8000 RPM, Sorvall GS3 Rotor) at 4°C. To the supernatant, solid NaCI was added to a final concentration of 200 mM, and 10% polyethyleneimine (PEI) solution slowly added to a final concentration of 196. The mixture was stirred for 3 hr, and then centrifuged 30 min, at 4°C, 8000 RPM (Sorvall GS3 Rotor).
Solid ammonium sulfate was then added to the supernatant at 0.5 glml and the mixture was stirred overnight at 4°C. The precipitated pioteins were centrifuged for 1 hr.
(8000 RPM, Sorvall GS3 Rotor) at 4°C and the resulting pellet dissolved in 100 ml of buffer B (10 mM K/P04, pH 7.2, 0.5 mM EDTA, 1G mM beta-mercaptoethanol, 50 mM NaCI, 10% glycerol, 0.05% Triton X-100TM, 50 pg/ml 1 PMFS, 20 ~cg/ml benzamidine, 2 ~.g/ml o-phenanthroline, 0.7 ~cg/ml pepstatin).
The dissolved protein solution was then dialysed (l4kD cut-off) for 12 hours against three 1 liter changes of buffer B. The dialyzed solution was then diluted to 600 ml with buffer B and applied to a 5 x 20 cm phosphocellulose P11TM
(Whatman) column (flow rate 100 ml/hr).
;, ~"'~'WO 94121663 PC'T/US94103246 3'hc column was then washed with 1.5 liter of buffer B followed by a 0 - 1.5 M NaCI gradient in buffer B (5 liters). R. G~iJI eluted at approximately 600 mM NaCI. The active fractions were then pooled and concentrated to 50 ml with a 76 mm Amicon YM 1 OTM membrane. The resulting solution was then diluted to 300 ml with buffer C (20 mM Tris-acetate, pH 7.4 at 4°C, 2 mM EDTA, 10 mM beta-mercaptoethanol, 50 mM NaCI, 10~
glycerol, 0.01 ~ Triton X-100, 50 ~cg/ml PMFS, 20 ~cg/ml benzamidine, 2 ~cg/ml o-phenanthroline, 0.7 ~g/ml pepstatin) and applied to 2.5 x 7 cm Heparin-Sepharose column at a flow rate of 25 ml/hr.
After a 400 ml wash with buffer B, R. CviJI was eluted with a 1.5 liter gradient of 0 - 1.3 M NaCI in buffer C. G~iJI eluted at approximately 400 mM NaCI. The most active fractions were pooled and applied to a 2.5 x 7 cm Blue-agarose column equilibrated in buffer D (20 mM Tris-acetate pH
8.0, 1 mM EDTA, 7 mM beta-mercaptoethanol, 30 mM NaCl, 10°.~ glycerol, 0.0196 Triton X-100, - 50 ~cg/ml PMFS, 20 ~cg/ml benzamidine, 2 ~cg/ml o-phenanthroline, 0.7 ~cg/ml pepstatin). After a 500 ml wash with buffer D, CviJI
was eluted with a 0 - 1.5 M. NaCI gradient (1.5 1) in buffer D. Active fractions were dialyzed against buffer G (10 mM K/P04 pH 7.0 (4°C), 10 mM beta-mercaptoethanol, 50 mM NaCI, 10% glycerol, 0.01 % Triton X-100, 50 ~cg/ml PMFS, 20 ~cg/ml benzamidine, 2 ~cg/ml o-phenanthroline, 0.7 ~cg/ml pepstatin) and loaded (20 ml/h) onto a ceramic HTPTM column (American International Chemical, Natick MA) (1.5 x 3 cm), equilibrated in buffer F (20 mM Tris-HCI
pH 8.0, 0.5 mM EDTA, 3 mM DTT, 50 mM K-acetate, 5 mM Mg acetate, SO ~
glycerol). After washing with 100 ml of buffer F, a 400 ml gradient 0 - 0.9 M
K/P04 in buffer F was run. The HTP column was washed with buffer G, containing 3 mg/ml BSA, then with 1 M phosphate buffer and reequilibrated in buffer G. The active fractions were then pooled and concentrated using a TM10 membrane to a final volume of 3 - 4 ml. This concentrate was then applied to a 2.5 x 95 cm SephadexTM G-100 column, equilibrated in buffer E (20 mM Tris-"",., WO 94121663 PC'TIUS94/03246 pH 7.5 (4°C), 5 mM Mg-Acetate, 2 mM EDTA, 10 mM beta-mercaptoethanol, 100 mM NaCI, 5 ~6 glycerol, O.OI % Triton X-100, 50 ~cg/ml PMFS, 20 ~cg/ml benzamidine, 2 ~cg/ml o-phenanthroline, 0.7 ~cg/ml pepstatin) at a flow rate of 6 ml/hr, and 3 ml fractions collected. Active fractions were dialyzed against storage buffer F.
The molecular weight of the purified CviJI was determined by comparison to la~own protein standards on a denaturing IOR~o SDS
polyacrylamide gel and a single band migrating with an apparent molecular weight of 32.5 ldlodaltons was seen indicating that by these criteria, G~iJI was purified to homogeneity.
Example 6 N-Terminal Amino Acid Sequence of R.CviJI
To confirm that the restriction endonuclease encoded by the insert in pCJHl.4 was Cv~I the sequence of the first 15 N-terminal amino acids of purified C~ilI was determined by the Edman degradation method using an Applied Biosystems (Foster City, CA) 477ATM Liquid Phase Protein Sequencer with an on-line 120A PT'HTM Analyser. The results of that analysis are shown in Table 1.
B
""" WO 94/21663 ~ PCT/US94/03246 Table 1 N-Terminal Amino Acid Analysis of CviJI
Amino Retentionpmol Pmol Pmol Pmol Amino Acid m Acid # Time (Raw) (-bkgd)(+lag) Ratio (min) 1 9.17 6.11 3.86 5.10 34.53 THR, MET, ARG, OR LYS
2 10.32 3.92 1.54 1.82 9.96 GLU
3 10.33 4.28 2.22 2.18 11.96 GLU
4 27.37 2.23 1.49 1.72 7.64 LYS
5 27.35 2.37 1.66 1.67 7.39 LYS
6 17.95 3.37 2.76 2.81 9.48 ARG
7 28.10 3.19 1.73 2.08 6.09 LEU
8 13.58 3.58 2.11 2.49 12.08 ALA
9 28.10 3.23 1.68 1.58 4.63 LEU
10 18.17 0.71 0.78 0.36 1.21 ILE
11 10.30 1.65 0.78 0.96 5.26 GLU
12 9.72 8.03 0.41 1.31 3.25 LYS
13 8.53 1.54 0.53 0.55 2.97 GLN
14 18.18 2.19 1.74 1.67 5.63 ARG
15 26.80 3.33 0.43 - 0.89 ILE
Abbreviations used:
threonine rTHR), methionine (MET), arginine (ARG), lysine (LYS), tamic LU), glu acid leucine (G (LEU), alanine (ALA).
isoleucine (>LE) and glutamine (GLN).
The results of this analysis confirm that the protein encoded by the DNA insert in pCJHl.4 (ORF1) is G'viJI.
2~~,90~~-The following Examples illustrate some of the unique properties of and important uses for CviJI.
Example 7 Analysis of CviJI* Recognition Sequences The G'viJI* recognition sequence (see Xia, et al., Nuc.Acids Res.
15: 6025-6090, 1987) was deduced by cloning and sequencing G~iJI* digested pUC 19 DNA fragments. A complete G~iJI * digest of pUC 19 was ligated to an M13mp18 cloning derivative for nucleotide sequence analysis. The sequence of the entire insert was read in order to determine which sites were or were not utilized. A total of 100 clones were sequenced, resulting in 200 G'viJI*
restricted junctions, the data for which are compiled in Table 2.
WO 94/21663 3 g~ ~ ~ PCT/US94/03246 a, ~' °'a, ~ b' d.' ~ '~ ~ ~ cL
V
O
~ .. ~ ~ .. ~ ~ .~ ~ ~ ~ ~~ -. .. ~ ,.
O~ O~ O O ~ h ~ M O M ~D Q~ ~ ~ M ~ ~O
O ~O !'rf ~.i O O O O V v ~ ~ 1p ~ v ~ ~ yr ~ ~ N ~
~O
~~ ~ O O N ~ ~ v7 N 1n cn ~ 00 ~ ~ N en v O ~j ~n ~n W ~n O V~ W i V~7. O O ~
r, b C~ U M et emn v~ .. ~ p t~ en .-. ~c N N ~-n t'n Ov ..r .~ y~ ,~., ,~ ,.N., ~ ~ O O O O O N
_O
V
~
~ ~ a M ~ O a ~ ~ p ~ N v c~ Ov M
.O ~v~~ O ~~v~ O v~vO 0~O ~~v~
~~ N O N ~O N
E"n ~' ""' O ~ ~ ~' ~ '-n ~~,, .r ~'~ N a .r ~ ~w N v ~ ..M.n ~ ..fir ~ h O
U U U ~.. ~ U U ~-. H ~ ~ C7 C7 ~ ~ a C7 C5 ~ N
~U(~~~~ CU,7V~~ CU,7UCU7U CU,7~CU7CU7 0 a C7 C7 a U E-~ t-~ ~ ~ ~ ~ ~ U E-~ U E- E
r V
w C
O
O
b O y A ~y' ~~~ U~H
U
C7 U E-~ ~ C7 U
o° .. '~ ~
Z ~, en r~
m U Z
~~~9~8~.
The dinucleotide GC is found at 205 sites in pUCl9. These GC
sites (shown in Table 2) can be divided into four classes based on their flanking Pu/Py structure, the normal recognition sequence (l~ and three potential classes of relaxed sites (R2 and R3). As seen in Table 2, the fraction of such NGCN
sites which belong to each classification is roughly equal (22.0%-27.8%). A
total of 200 CviJI* restricted junctions were analyzed by sequencing 100 cloned inserts.
If G'viJI* cleaved at all NGCN sites without sequence preferences, it would be expected that the fraction of each classification should be restricted approximately equally. Instead, most of the sites cleaved by this treatment were found to be normal, or PuGCPy sites (47.5 % ). R1 (PyGCPy) and R2 (PuGCPu) restricted sites were found at nearly the same frequency (25.5% and 27.0%, respectively).
Out of 200 G'viJI* junctions, no R3 (PyGCPu) restricted sites were found.
Thus, G~iJI* cleaves all NGCN sites except for PyGCPu. As G'viJI* cleaves 12 out of 16 possible NGCN sites, it may be referred to as a 2.25-base recognition endonuclease.
In addition to the restricted sites, those sites which were not cleaved by CviJI* conditions were also compiled for analysis, as shown in Table 2. A
total of 116 non-cleaved NGCN sites were found in the 100 inserts which were sequenced. PyGCPu sites represented the largest class of non-cleaved sites (52.6 % ). In only two cases were PuGCPy sites found not to be cleaved. An approximately equal fraction of Rl and R2 sites were not cleaved as were found cleaved (22.4 % versus 25.5 % for R1 and 23.3 % versus 27.0 % for R2). Based on the frequency of cleavage, or lack thereof, a hierarchy of restriction under CviJI* conditions is evident, where PuGCPy > > PuGCPu = PyGCPy.
WO 94/21663 ~ ~ ~ ~ PCT/US94/03246 Example 8 CviJI* Restriction Generated Oligonucleotides Due to the high frequency of CviJI or CviJI* restriction, it is possible to generate useful oligonucleotides by digestion and a heat denaturation step as described above. The size and number of the resulting oligonucleotides are important for subsequent applications such as those described above. If for example, an oligonucleotide is to be used with a large genome, it has to be long enough so that the sequence detected has a probability of occuring only once in the genome. This minimum length has been calculated to be 17 nucleotides for the human genome (Thomas, C.A., Jr. Prog. Nucl. Acid Res. Mol. Biol., 5:315 (1966)). Oligonucleotides used for sequencing or PCR amplification are generally 17-24 bases in length. Oligomers of shorter length will often bind at multiple positions, even with small genomes, and thus will generate spurious extension products. Thus, an enzymatic method for generating oligomers should ideally result in polymers greater than 18 bases in length.
The theoretical number of pUCl9 CviJI* restriction-generated oligomers is 314 (157 CviJI* restriction fragments x 2 oligomers/fragment), the size distribution of which is shown in panel A of Figure 5. Most of the expected CviJI* restriction-generated oligomers (about 7590 are smaller than 20 bp.
This assumes that CviJI is capable of restricting DNA to very small fragments, the shortest of which would be 2 bp. However, in practice, about 93 ~ of the cloned CviJI* fragments were 20-56 by in size, and 3~ of the fragments generated by CviJI* were smaller than 20 by (panel B of Figure 5). This suggests that CviJI*
is not able to bind or restrict those fragments below a certain threshold length.
Since the smallest observed fragment is 18 bp, it may be assumed that this length is the minimal size which can be generated from a given larger fragment.
Whatever the reason for this phenomenon, CviJI* treatment of DNA produces a _ 2 15 908' relatively small range of oligomers, (mostl;~ 20-60 bases in length), most of which are a perfect size class for molecular biology applications.
Example 9 Anonymous Primer Cloning Primers are critical tools in many molecular biology applications such as PCR, sequencing, and as probes. Anonymous primers are useful as sequencing primers for genomic sequencing projects, as probes for mapping chromosomes, or to generate oligonucleotides for PCR amplification.
The Anonymous Primer Cloning (APC) method is a variation of shotgun cloning in that unknown sequences of DNA are being randomly cloned.
However, unlike GwilI shotgun cloning, wherein a partial G~ilI** digest of DNA
is cloned, anonymous primer cloning utilizes'a complete G~iJI* digest to restrict large DNAs into small fragments 20-200 by in size. These small fragments are cloned into a unique vector designed for excising the anonymous DNA as labeled primers. The strategy for this method is illustrated in Figure 6.
As illustrated in Figure 6, the APC strategy reduces large DNAs to small fragments, which are cloned and excised for use as primers. Plasmid pFEM has a unique an~angement of the restriction sites for MboII and FokI, which permits DNA cloned into the EcoRV site to be excised without associated vector DNA. This is possible because FokI cleaves 9/13 bases to the left of the recognition site shown in pFF.M and MboII cleaves 8/7 bases to the right of the recognition site shown in pFEM, which is well into the cloned anonymous sequence. After MboII or FokI restriction, a known flanlang primer is annealed (primer 1 or 2) and extended using a DNA polymerise and dNTPs. The primer is previously end-labeled, or alternatively, one or more of the dNTPs is radioactive.
After denaturation of the newly synthesized DNA and separation from its cognate template, the labeled anonymous primer is ready for use in sequencing the original template from which it was subcloned. The presence of the pFEM vector sequence fused to the anonymous sequence does not influence the enzymatic extension of this primer from its unique binding site, as the vector DNA is at the 5' end and the unique sequence is located at the 3' end (all polymerases extend 5' to 3'). Both the top and bottom strand primers may be excised from pFEM due to the symmetrical placement of restriction sites and flanking primer binding sites. Thus, two primers may be derived from each cloning event. APC is particularly well suited to the genomic sequencing strategy of Church and Gilbert Proc Natl. Acad Sci. USA 81:1991-1995 (1984), although its utility is not limited thereto.
Example 10 End Labeling of Restriction-Generated Oligonucleotides As is clear from the foregoing examples, digesting DNA with CviJI* provides the ability to generate sequence-specific oligonucleotides ranging in size from 20-200 bases in length with an average length of 20-60 bases.
Sequence specific oligonucleotides generated by CviJI* digestion may be labeled directly at the S'-end or at the 3'-end using techniques well known in that art.
For example, 5'-end labeling may be accomplished by either a forward reaction or an exchange reaction using the enzyme T4 polynucleotide kinase. In the forward reaction, 32P from [y32P]ATP is added to a 5' end of an oligonucleotide which has been dephosphorylated with alkaline phosphatase using standard techniques widely known in the art and described in detail in Sambrook et al. , Molecular Cloning: A Laboratory Manual, 2nd Edition. Cold Spring Harbor Laboratory Press (1989). In an exchange reaction, an excess of ADP
(adenosine diphosphate) is used to drive an exchange of a 5'-terminal phosphate from the sequence specific oligonucleotide to ADP which is followed by the transfer of 32P from y32P-ATP to the 5'-End of the oligonucleotide. This reaction is also catalyzed by T4 polynucleotide kinase and is decribed in Sambrook et al. , Molecular Cloning: A Laboratory Manual, 2nd Edition. Cold Spring Harbor Laboratory Press (1989).
Homopolymeric tailing is another standard labeling technique useful in the labeling of CviJI*-generated sequence specific oligonucleotides. lfiis reaction involves the addition of 32P-labeled nucleotides to the 3'-end of the sequence specific oligonucleotides using a terminal deoxynucleotide transferase.
(Sambrook et al. , Molecular Cloning: A Laboratory Manual, Znd Edition. Cold Spring Harbor Laboratory Press (1989)).
Commonly used labeling techniques typically employ a single oligonucleotide directed to a single site on the target DNA and containing one or a few labels. Oligonucleotides generated by the method of the present invention are directed to many sites of a target DNA by virtue of the fact that they are generated from a sample of the target sequence. Thus, the hybridization of multiple oligonucleotides (labeled by the methods described above) allows a significantly enhanced sensitivity in the detection of target sequences. In addition, the short length of the labeled oligonucleotides used in the methods of the present invention allows a reduction in hybridization time from overnight (as is used in conventional methods) to 60 mins.
Although labeling sequence specific oligonucleotides with 32P is described above, labeling with other radionucleotides, and non-radioactive labels is also within the scope of the present invention.
WO 94/21663 215 9 0 81 pCT~S94/03246 Example 11 Primer Extension Labeling of DNA Using Restriction-Generated Oligonucleotides (PEL_RGO) Another aspect of the present invention includes methods for labeling DNA which include the generation of oligonucleotide primers by complete digestion with CviJI*, followed by heat denaturation. PEL-RGO
requires three steps: 1) generating the sequence-specific oligonucleotides by CviJI*
restriction of the template DNA; 2) denaturation of the template and primer;
and 3) primer extension in the presence of labeled nucleotide triphosphates.
Plasmid DNA may be prepared by methods known in the art such as the alkaline lysis or rapid boiling methods (Sambrook et al., Molecular Cloning: A Laboratory Manual, 2ru1 Edition). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York (1989)). In addition, the vector should be linearized to ensure effective denaturation. A restriction fragment may be labeled after separation on low melting point agarose gels by methods well known in the art.
In PEL-RGO labeling, template DNA to be labeled is divided into two aliquots; one is used to generate the sequence specific oligonucleotide primers and the other aliquot is saved for the primer annealing and extension reaction.
A typical reaction mix for generating sequence-specific oligonucleotides is assembled in a microcentrifuge tube and includes: 100 ng DNA; 2 ~cl Sx CviJI*
buffer; 0.5 ~,1 CviJI (lu/~1); sterile distilled water to 10 ~cl final volume.
CviJI*
SX restriction buffer includes: 100 mM glycylglycine (Sigma, St. Louis, Missouri, Cat. No. 62265) pH adjusted to 8.5 with KOH, 50 mM magnesium acetate (Amresco, Solon, Ohio, Cat. No. P0013119), 35 mM ~B-mercaptoethanol (Mallinckrodt, Paris, Kentucky, Cat. No. 60-24-2), 5 mM ATP, 100 mM
dithiothreitol (Sigma, St. Lous, Missouri, Cat. No. D9779) and 25 gb v/v DMSO, (Mallinckrodt Cat. No. 67-68-5). CviJI is obtained from CHIMERx (Madison, Wisconsin). The reaction mix is incubated at 37°C for 30 min, followed by the 2~~gp81 inactivation of CviJI by heating at 65°C for 10 min. The CviJI*-restricted DNA
may be used directly without further purification, or it may be stored at -20°C for several months for subsequent labeling reactions.
After heat-inactivating , E'viJI, 0.2 ~cg of the digested and undigested DNA are electrophoresed on a 1.5 % agarose gel, using a suitable molecular weight marker for comparison. The CviJI restriction fragments appear as a low molecular weight smear in the 20-200 by range.
By way of example, 1-10 ng of linearized pUCl9 was labeled under the conditions described below. A template-primer cocktail was prepared by mixing 10 ng of linearized pUC 19 DNA template with 20 ng pUC 19 sequence-specific oligonucleotides (prepared as described above) and the mixture is brought to a final volume of 17 ~,1 with sterile distilled water. The template-primer mixture is denatured in a boiling water bath for 2 minutes and immediately placed on ice.
The following labeling mixture is then added to the template-primer mix:2.5 ~,l lOX labeling buffer (500 mM Tris HCl at pH 9.0, 30 mM MgCl2, 200 mM (NH4)2S04, 20~cM dATP, 20~cM dTTP, 20~cM dGTP, 0.4 % NP-40);
5.0 ~1 [a-32P] dCTP (3000Ci/mmol, lO~Ci/~cl New England Nuclear, Catalog No. NEG013H); 0.5 ~cl Thermos flavus DNA polymerase (5u/~cl) (Molecular Biology Resources, Milwaukee, Wisconsin); up to 25 ~cl final volume with distilled water. The reaction was incubated at 70°C for 30 min and then stopped by adding 2~c1 of 0.5M EDTA at pH 8.0 to the reaction mix.
The efficiency of the labeling reaction is gauged by the percentage of radioisotope incorporated into labeled DNA. One microliter of the labeling reaction is added to 99 ~cl of IOmM EDTA in a microcentrifuge tube. This serves as the source of diluted probe for total and trichloroacetic acid (TCA)-precipitable counts. 2 ~cl of diluted probe is spotted onto the center of a glass fiber filter disc (Whatman number 934-AH). The disc is then allowed to dry and is then placed in a vial containing scintillation cocktail for counting total radioactivity in a liquid scintillation counter. Another 2 ~cl aliquot from the diluted probe is added to 1 ml of 10 k ice cold TCA followed by the addition of 2 ~cl of carrier bovine serum albumin (BSA). This mixture was then placed on ice for 10 minutes. The precipitate is then collected on a glass filter disc (Whatman No. 934-AI-~ by vacuum filtration. The filter is then washed with 20m1 of ice cold 10 % TCA, allowed to dry and is placed in a vial containing scintillation cocktail and counted.
Because primer extension oligonucleotide labeling results in net DNA synthesis, the specific activity of labeled DNA is calculated using the following guidelines.
Total cpm incorporated = TCA cpm X 50 X 27 Wherein the factor 50 is derived from using 2 ~cl of a 1:100 dilution for TCA
precipitation. The number 27 converts this back to the total reaction volume (which is the reaction volume plus 2 ~,1 of stop solution).
Synthesized DNA (ng of DNA synthesized) _ theoretical yield X fraction of radioactivity incorporated.
Theoretical yield (ng of DNA)= uCi dNTPs added x 4 X 330ng/nmole specific activity dNTP(Ci/mmole=~cCi/nmole) Fraction of incorporated label = TCA precipitated cpm/ total cpm.
Specific activity (cpm/~cg of DNA) = totem incom9rated x 1000 synthesized DNA + input DNA
Wherein 1000 is the factor converting nanograms to micrograms.
~15908~.
By way of example, the following represents the calculation of specific activity for an aliquot of pUC 19 DNA labeled using this method.
Using 50 ~Ci of [a- 32P)dCTP in a 25 ~cl reaction, and if the TCA precipitated cpm is 26192 and total cpm is 102047;
Total cpm incorporated = 26192 X 50 X 27 =3.27 x 107cpm Synthesized DNA (ng of DNA synthesized) _ Theoretical yield X fraction of radioactivity incorporated.
Theoretical yield = ~.cCi of dNTPs x 4 x 330 3000 ~Ci/nmole =50 pCi x 4 X 330 = 22 ng Fraction of label incorporated = TCA precipitated cram = 26192 - 0.256 Total cpm 102047 Synthesized DNA = 22 X 0.256 = 5.6 ng Specific activity (cpm /~.g) = Total cpm incorporated x 1000 Synthesized DNA +input DNA
Input DNA = 10 ng Specific activity = 3.27 x 107x 1000 5.6+10 =2.09 x 1()9 cpm/~cg Unincorporated radioactive label may be removed using standard methods well lrnown in the art.
_2159pgI
Comparisons were made between PEL-RGO vs RPL under similar conditions, and it was observed that a detection limit of 100 fg was seen using PEL-RGO labeled DNA compared to a detection limit of 500 fg with RPL, using a radiolabeled probe.
Example 12 Thermal Cycle Labeling and Universal Thermal Cycle Labeling Thermal Cycle Labeling (TCL) is a method according to the present invention for efficiently labeling double-stranded DNA while simultaneously amplifying large amounts of the labeled probe. TCL of DNA requires two general steps: 1) generation of the sequence-specific oligonucleotides by G'viJI*
restriction of the template DNA; and 2) repeated cycles of denaturation, annealing, and extension in the presence of a thermostable DNA polymerise or a functional fragment thereof which maintains polymerise activity. Optimal results are obtained after 20 such cycles, which is best performed in an automated thermal cycling instrument such as a Perkin-Elmer Model 480 thermocycler. In conjunction with such an instrument, about 1.5 hr. is required to complete this protocol. If a thermal cycler is not available these reactions may be performed using heat blocks. As few as 5 cycles may yield probes with acceptable detection sensitivities. The generation of sequence specific oligonucleotides for use in this method may also be accomplished using the restriction endonuclease reagent CGase I described in Example 20 or the restriction endonuclease Aci I which has as a recognition sequence CCGC.
Non-radioactive labeling of DNA using TCL is accomplished by mixing: 10 pg - 100 ng linearized template, 50 ng CviJI*-digested primers (prepared as described above), 1.5 ~cl lOX labeling buffer, 0.5 ~cl Thermos flavus DNA polymerise (Su/~cl) (Molecular Biology Resources, Inc. , Milwaukee, ~lr~gpgl - -so-Wisconsin), 1 ~1 of 1mM Biotin-11-dUTP (Enzo Diagnostics, New York, New York), l.s ~cl each of dAT':P, dCTP, and dGTP (2 mM), and 1.0 ~cl 2mM dTTP.
Radioactive labeling of DNA using TCL was accomplished by mixing 10 pg - 100 ng of CviJI generated primers, 10 pg-2s ng of linearized s template, l.s ~,1 of lOX labeling buffer;' 5 ~1 of 32P-dCTP (3000 Ci/mmole, ~cCi/~,1 or 40 ~Ci/yl), O.s ~l of Thermus,flavus DNA polymerise (su/~1), and O.s ~cl each of dATP, dGTP, and dTTP (1 mM) was added. The reaction mix was brought to a volume of is ~cl with deionized H20, overlaid with mineral oil and cycled through 20 rounds of denaturation, annealing and extension. A typical cycling regimen ernployed :ZO cycles of denaturation at 91°C for s sec, annealing at s0°C for s se<; and exaension at 72°C for 30 sec. The reaction is then terminated by adding 1 ~1 of O.sM EDTA, pH 8Ø The amplified, labeled probe is a very heterogeneous mixture of fragments, which appears as a smear when analyzed by agaro:~e gel electrophoresis.
is Uni,~ersal thermal cycle labeling (UTCL) is a method according to the present invention for efficiently labeling double-stranded DNA while simultaneously amplifying large amounts of labeled probe. UTCL is unique in that no sequence information is required regarding the template. The extension primers are suppleri endogenously via the holo-enzyme of the thermostable DNA
polymerise and amy anonymous DNA template can be labeled by repeated cycles of denaturation, annealing, and extension in the presence of a labeled deoxynucleotide triphosphat:e. Optimal results are obtained after 20 such cycles, which is best performed in an automated thermal cycling instrument such as a Perkin-Elmer Modal 480 th~ermocycler. In conjunction with such an instrument, 2s about l.s hr are rE;quired t~o complete this protocol. If a thermal cycler is not available these reactions may be performed using heat blocks. As a few as s cycles may yield probes wio acceptable detection sensitivies.
Nom-radioactive labeling of DNA using UTCL is accomplished by mixing: 10 ng lint;arized template, l.s ~,l lOX labeling buffer, O.s ~cl Therntus ,A..., WO 94121663 PCT/US9~1032.~6 ,flavus DNA polymerise (5u/~cl) (Molecular Biology Resources, Inc.,'Milwaukee, Wisconsin), 1 ~cl of 1mM Biotin-11-dU'TP (Enzo Diagnostics, New York, New York), 1.5 ~l each of dATP, dCTP, and dGTP (2 mM), and 1.0 ~cl 2mM dTTP.
Radioactive labeling of DNA using UTCL was accomplished by mixing: 10 pg-100 ng of linearized template, 1.5 ~cl of lOX labeling buffer, 5 ~cl of 32P-dCTP (3000 Ci/mmole, 10 ~cCi/~,cl or 40 ~Ci/~cl), 0.5 ~cl of The»~eus Jlavus DNA polymerise (5u/~cl), and 0.5 ~d each of dATP, dGTP, and dTTP (1 mM) was added. The reaction mix was brought to a volume of 15 ~cl with deionized H20, overlaid with mineral oil and cycled through 20 rounds of denaturation, annealing and extension. A typical cycling regimen employed 20 cycles of denaturation at 91°C for 5 sec, annealing at 50°C for 5 sec and extension at 72°C
for 30 sec. The reaction is then terminated by adding 1 ~d of O.SM EDTA, pH
8Ø The amplified, labeled probe is a very heterogeneous mixture of fragments, which appears as a smear when analyzed by agarose gel electrophoresis.
Estimation of Bio-l l d incomoration:
In order to estimate the level of incorporation of biotin-11-dUTP
into DNA, a serial dilution from 1:10 to 1:108 of the labeled probe (free of unincorporated biotin-11-dLTTP) is made in TE (IOmM Tris, 1mM EDTA, pH 8).
A microliter of each dilution is placed on a neutral nylon membrane, and the DNA sample is bound to the membrane either by L1V cross linking for 3 min or by baking at 80°C for 2 hr.
The unbound sites on the membrane are blocked using a blocking buffer for 15 min at 25°C. Streptavidin-alkaline phosphatase (Gibco-BRL
Gaithersburg, Maryland, Cat. No. 9545A) is added to the blocking buffer (0.058 M Na2HP04, 0.017 M NaH2P04, 0.068 M NaCl, 0.02 % sodium azide, 0.5 °k casein hydrolysate, 0.1 % Tween-20) at a 1:5000 dilution and incubated for a min., and the membrane is rinsed 3 times for 10 min. each with wash buffer (lx PBS [0.058 M NazHP04, 0.017 M NaH2P04, 0.068 M NaCI], 0.3% TweenTM, 215908.
0.2~ sodium azide), rinsed briefly (5 minutes) with AP buffer (100 mM NaCI, mM MgCl2, 100 mM Tris-Cl pH 9.5) aid then enough AP buffer containing 4.0 ~,1/ml vitro blue tetrazolium (NBT) (Sigma Cat. No. N6639), (Sigma Cat.
No.
B6777), and 3.5 ~l/ml of S-bromo-4-chloro-3-indolyl phosphate (BCIP) was added 5 in order to cover the membrane. The membrane is left in the dark for approximately 30 minutes or until the reaction is complete. The reaction is stopped by rinsing in 1 X PBS.
Detection Sensitivities 32p_labeled probes generated by the protocol above described labelling detect as little as 25 zeptomoles (2.5 x 10-20 moles) of a target sequence. As little as 10 pg of template DNA is enough to synthesize 5-10 ng of radiolabeled probe, which is sufficient for screening 5 Southern blots. The radioactive versions of TCL and UTCL facilitate extremely high specific activities of labeled probe (about 5 x 109 cpm/~cg DNA), which permits 5-10 fold lower detection limits than conventional labeling protocols. The synthesis of higher specific activity probes is probably the net result of the sequence-specific oligonucleotide primers and their increased length when compared to the short random primers used in other labeling methods. In addition, the thermal cycling permits probe amplification.
Biotin-labeled probes generated by the TCL and UTCL protocols detect as little as 25 zeptomoles (2.5 x 10-20 moles) of a target sequence. A
~cl TCL or ITTCL reaction yields as much as 5-10 ~g of labeled DNA, enough to probe 5 to 10 Southern blots. Biotin-labeled TCL and UTCL probes provide a 10 fold greater detection sensitivity when compared to RPL biotin probes. In addition, the thermal cycling permits probe amplification.
Non-radioactive, biotinylated probes labeled by the TCL and UTCL
methods were shown to have detection limits that are identical to the radioactive probes. These methods have the advantage of eliminating the need to work with ,~.... WO 94/21663 hazardous radioactive materials without sacrificing sensitivity. In addition, results Ore obtained from non-isotopic probes in 3-4 hours compared to 3-4 days for radiolabeled probes. The ability to substitute non-radioactive probes for radioactive probes may be very useful to clinical laboratories, which do not use radioisotopes but do need greater detection- sensitivities. Research laboratories favor the use of non-isotopic systems if detection sensitivity is not an issue. The non-isotopic labeling version of the TCL and UTCL systems represent a major improvement in labeling DNA probes. Non-radioactive probes generated by the methods of the present invention are also useful in the detection of RNA in situ.
An advantage of this system is that labeling protocols of the present invention yield highly sensitive non-radioactive probes, and the size of the probes are predominantly in the small molecular weight range and can therefore penetrate the tissue easily, unlike RPL. Because non-radioactive probes labeled using the labeling protocols of the present invention have the same detection limits as do radioactive probes similarly labeled, it is within the scope of this invention to use either radioactive or non-radioactive probes for probing, for example, Southern blots, Northern blots, for in situ hybridization for the detection of mRNA or DNA
in cells or tissue directly, and for colony or plaque lifts.
Example 13 Quasi-Random fragmentation of DNA
Shotgun cloning and sequencing requires the generation of an overlapping population of DNA fragments. Therefore, conditions were established for the partial digestion of DNA with CviJI to produce an apparently random pattern, or smear, of fragments in the appropriate size range.
Conventional methods for obtaining partially restricted DNA include limiting the incubation time or limiting the amount of enzyme used in the digestion.
Initially, ~~.~9~81 agarose gel electrophoresis and ethidium bromide staininb of the treated DNA
were utilized to assess the randomness and size distribution of the fragments.
CviJI was obtained from CHIMERx (Madison, Wisconsin).
Digestion of pUCl9 DNA for limited time periods, or with limiting amounts of CviJI under normal or relaxed conditions, did not produce a quasi-random restriction pattern, or smear. Instead, a number of discrete bands were observed, as shown in Figure 7, lane 3 for the CviJI* partial digestion of pUCl9.
Complete digests of pUCl9 under normal and GwilI* buffer conditions are shown in lanes 1 and 2 respectively. These results show that, under these relaxed conditions, G'viJI has a strong restriction site preference.
To eliminate the apparent restriction site preferences observed under the partial restriction conditions described above, a series of altered reaction conditions were explored. Conditions of high pH, low ionic strength, addition of solvents such as glycerol or dimethylsulfoxide, and/or substitution of Mn2+
for Mg2+ were systematically tested with CviJI endonuclease using the plasmid pUC 19. Figure 7 shows the results of these tests. In Lane M, a 100 by DNA
ladder was run. In Lanes 1-4, pUCl9 DNA (1.0 ~cg) was run after digestion at 37°C in a 20 ~1 volume for the following times and conditions: Lane 1, complete G'viJI digest (1 unit of enzyme for 90 min in 50 mM Tris-HCI, pH 8.0, 10 mM
MgCl2, 50 mM NaCI); Lane 2, complete G~iJI* digest (1 unit of enzyme for 90 min in 50 mM Tris-HCI, pH 8.0,10 mM MgCl2, 50 mM NaCI, 1 mM ATP, 20 mM DTT); Lane 3, partial CviJI* digest (0.25 units of enzyme for 30 min in 50 mM Tris-HCI, pH 8.0, 10 mM MgCl2, 50 mM NaCl, 1 mM ATP, 20 mM
DTT); Lane 4, partial G'viJI** digest (0.5 units of enzyme for 60 min in 10 mM
Tris-HCI, pH 8.0,10 mM MgCl2, 10 mM NaCI, 1 mM ATP, 20 mM DTT, 20 v/v DMSO); and Lane 5, uncut pUCl9 (1.0 ~.g).
The digestion condition which yielded the best "smearing" pattern was obtained when the ionic strength of the relaxed reaction buffer was lowered and an organic solvent was added (Figure 7, lane 4). Plasmid pUCl9 partially ".~~~,. WO 94/21663 PCTIUS94I0324b digested under these conditions yields a relatively r~on-discrete smear. This activity is referred to as C~~iJI** to differentiate it from the originally-characterized star activity described in Xia et al. , Nucl. Acids Res. 15:
(1987). The appearance of diffuse, faint bands overlying a background smear generated from this 2686 by molecule indicates that some weakly preferred or resistant restriction sites may bias the results of subsequent cloning experiments.
DNA was mechanically sheared by sonication utilizing a Heat Systems Ultrasonics (Farmingdale, New York) W-375 cup horn sonicator as specified by Bankier et al., Methods in Fruymology 155:51-93 (1987). DNA
fragmented by this method has random single-stranded overhanging ends (ragged ends).
G~iJI* digested, and sonicated samples were size fractionated by agarose gel electrophoresis and elactroelution, or by spin columns packed with the size exclusion gel matrix, Sephacryl S-SOOT"s (Pharmacia LKB, Piscataway N.J.) to eliminate small DNA fragments. Spin columns (0.4 crn in diameter) were packed to a height of 1.3 cm by adding 1 ml of Sephacryl S-500 slurry and centrifuging at 2000 RPM for 5 minutes in a Beckman CPR centrifuge. The columns were rinsed 3 times with 1 ml aliquots of 100 mM Tris-HCl (pH 8.0) by centrifugation at 2000 RPM for 2 min. Typically, 0.2-2.0 pg of fragmented DNA
in a total volume of 30 ~,l was applied to the column. The void volume, containing those DNA fragments larger than S00 bp, was recovered in the column eluant after spinning at 2000 RPM for 5 minutes. The capacity of this micro-column procedure is 2 pg of DNA. Agarose gel electrophoresis and electroelution are described in detail by Sambrook et al. Molecular Cloning: A
Laboratory Manual, Second Edition Cold Spring Harbor Laboratory Press, Cold Spring Harbor N.Y. (1989) and is well known to those skilled in the art. In these experiments, 5 pg of sample was pipetted into a 2 cm-wide slot on a 1 %
agarose gel. Electrophoresis was halted after the bromophenol blue tracking dye had migrated WO 94/21663 PCT/US94/03246 "",~
2Z~90~~.
6 cm. Fragments larger than 750 bp, as judged by molecular size markers, were separated from smaller sizes and electrophoresed onto dialysis tubing (1000 MW
cutoff). The fractionated material was extracted with phenol-chloroform and precipitated using ice cold ethanol (50 ~ final volume) and ammonium acetate (2.5 M final concentration).
The ragged ends of the sonicated DNA were rendered blunt utilizing two different end repair reactions. In one end repair reaction (ER
1) sonicated DNA was treated according to the procedure outlined by Bankier et al.
Methods in Enzymology 155:51-93 (1987), where 2.0 ~g of sonicated lambda DNA is combined with 10 units of the Klenow fragment of DNA polymerase I, 10 units T4 DNA polymerase, 0.1 mM dNTPs, (deoxynucleotide triphosphates=deoxyadenosine triphosphate, deoxthymidine triphosphate, deoxycytosine triphosphate, and deoxyguanosine triphosphate) and reaction buffer (50 mM Tris-HCI, pH 7.5,10 mM MgCl2, 10 mM DTT). This mixture was incubated at room temperature for 30 min followed by heat denaturation of the enzymes at 65°C for 15 minutes. In a second end repair reaction (ER 2), an excess of the reagents and enzymes described above were utilized to ensure a more efficient conversion to blunt ends. In this reaction, 0.2 ~g of the sonicated lambda DNA sample was treated under the same reaction conditions described above.
Figure 8 shows comparisons of the size distributions of sonicated DNA versus DNA that was partially digested with CviJI**. In Lanes M, a 1 kb DNA ladder was run. In Lanes 1-3, untreated ~ DNA (0.25 ~cg), sonicated ~
DNA (1.0 ~cg), and CviJI** partially-digested ~ DNA (1.0 ~cg) were run, respectively. In Lanes 4-6, untreated pUC 19 (0.25 fig), sonicated pUC 19 ( 1.0 fig), and CviJI** partially-digested pUCl9 (1.0 ~cg) were run, respectively.
Fragmentation of a large substrate such as lambda DNA (45 kb) revealed essentially no banding differences between the CviJI** method and sonication, as demonstrated in Figure 8, lanes 2 and 3. In addition, pUCl9 DNA
21~~0~~.
-5,-that was partially digested with G~iJI** gave a size distribution or "smear"
that closely resembled that achieved with sonication (Figure 8, lanes 5 and 6). As expected, the minor bias evident with a small molecule such as pUC 19 was not detectable with a larger substrate such as lambda DNA.
The intensity and duration of sonic treatment affects the size distribution of the resulting DNA fragments. The results obtained from the sonication of lambda and pUCl9 samples (Figure 8) were obtained from three 20 second pulses at a power setting of 60 watts. Sonication-generated smears are similar, although the size distribution of fragments is consistently greater with G'viJI** fragmentation. This result favors the cloning of larger inserts, which facilitates the efficiency of end-closure strategies (Edwards et al. , Genome 6:593-608 (1990)). The size distribution of the DNA fragmented by CviJI** is controlled by incubation time and amount of enzyme, variables which are readily optimized by routine analysis. An excess of enzyme or a long incubation time will completely digest pUCl9 DNA, resulting in fragments which range in size from approximately 20 by to approximately 150 by (Figure 7, lanes 1 and 2).
The results shown in Figure 8 were obtained by incubating pUCl9 for 40 minutes and lambda DNA for 60 minutes with 0.33 units of CviJI/~cg substrate. The efficiencies of the two methods for randomly fragmenting DNA were quantitatively analyzed for use in molecular cloning, as described below.
Example 14 Rapid DNA Size Fractionation Utilizing Spin Column Chromatography The amount of data obtained by the shotgun sequencing approach is substantially increased if fragments of less than 500 by are eliminated prior to the cloning step. Small fragments yield only a portion of the sequence data which may be collected from polyacrylamide gel based separations and, thus, such small fragments lower the efficiency of this strategy. Agarose gel electrophoresis ~15908~.
-s8-followed by electroelution is commonly used to size fractionate DNA prior to shotgun cloning (Bankier et dl., Methods in Enzymol. 155a1-93 (198'0.
Approximately three hours are required to prepare the agarose gel, electrophorese the sample, electroelute fragments larger thar~a00 bp, perform phenol-chloroform s extractions, and precipitate the resulting material.
The results of s out of 9 independent trials size-fractionating CviJI**-fragmented lambda DNA by agarose gel electrophoresis are shown in Figures 9A-E. Figures 9A-D illustrate the following. In Figure 9A: Lane M, 1 kb DNA ladder; lane ~, untreated ~ DNA (0.2s ~cg); lane 1, unfractionated (UF) G'viJI** partially-digested ~ DNA (1.0 ~.g); lane 2, column-fractionated (CF) CviJI** partially-digested ~ DNA (1.0 ~cg); lane 3, gel-fractionated (GF) G'viJI**
partially-digested ~ DNA (1.0 ug); and in Figures 9B-E are additional trials of the same treatments as in the lanes of Figure 9A which have the same label.
Small DNA fragments may also be removed by passing the sample is through a short column of Sephacryl S-s00. Approximately is min. are needed to prepare the column and s min. to fractionate the DNA by this method.
The results of three out of nine trials using a Sephacryl S-s00 column are shown in Figures 9A-C. The efficiency of eliminating small DNA
fragments ( < s00 bp) by spin column chromatography appears high, and the reproducibility was excellent. This result is in contrast to the agarose gel electrophoresis and electroelution data presented in Figures 9A-E wherein nine replicate trials of this method yielded nine differently sized products, regardless of the source of the agarose. Both methods yielded 30-40 °Ib recoveries as measured by UV spectrophotometry. To quantitate the relative efficiencies of the 2s two fractionation methods, the lambda DNA size fractionated in Figure 9A
lanes 2 and 3, and Figure 9B lane 3 were analyzed for cloning efficiency and insert size, as described below.
,~,» WO 94/21663 PCT/US94/03246 Example 15 Cloning Efficiencies of Gel Elution and Chromatography Fractionation Methods The efficacy of size selection was quantified by two criteria: 1) by comparing the relative cloning efficiency of CviJI** partially-digested lambda DNA fragments fractionated either by agarose gel electrophoresis and electroelution or micro-column chromatography, and 2) determining the size distribution of the resulting cloned inserts. To reduce potential variables, large quantities of the cloning vector and ligation cocktail were prepared, ligation reactions and transformation of competent E, coli were performed on the same day, numerous redundant controls were performed, and all cloning experiments were repeated twice. Ligation reactions were carried out overnight at 12°C in 20 ~cl mixtures using the following conditions: 25 mM Tris-HCl (pH 7.8), 10 mM
MgCl2, 1 mM DTT, 1 mM ATP, DNA, and 2000 units of T4 DNA ligase. For unfractionated samples, 10 ng of fragments and 100 ng of HincII-restricted, dephosphorylated pUCl9 were combined under the above conditions. For Sephacryl S-500 fractionated samples, 50 ng of size-selected fragments were ligated with 100 ng of HincII-restricted, dephosphorylated pUCl9. This increase in fractionated DNA was determined empirically to compensate for the lower concentration of "ends" resulting from the fractionation procedure and/or the lowered efficiency of cloning larger fragments. Ligation reaction products were added to competent E. coli DHSaF' (~80d1acZOMIS 0(IacZYA-argF)U169 deoR
gyrA96 recAl relAl endAl thi-1 hsdRl7(rK-,mK+) supF~44 ~-) in a transformation mixture as specified by the manufacturer (Life Technologies, Bethesda, Maryland) and aliquots of the transformation mixture were plated on T agar (Messing, Methods in Enzymol. 101:20-78 (1983)) containing 20 ~,g/ml ampicillin, 25 ~cl of a 2 % solution of isopropylthiogalactoside (IPTG) and 25 ~cl of a 2% solution of 5-dibromo-4-chloro-3-indolylgalactoside (X-GAL). The cloning efficiencies reported are the average of triplicate platings of each ligation reaction. The concentration of the fractionated material was checked spectrophotometrically so that 50 ng was added to all ligation reactions. This material was ligated to HincII-digested and dephosphorylated pUCl9. This cloning vector was chosen because it permits a simple blue to white visual assay to indicate whether a DNA fragment was cloned (white) or not (blue) (Messing, Methods in Enzymol. 101:20-78 (1983)).
A summary of the cloning efficiencies calculated from two independent trials is given in Table 3.
WO 94/21663 ~ ~ ~ PCT/US94/03246 'FABLE 3 Cloning Efficiencies of CviJI** Partially Digested Lambda DNA
Fractionated by Microcolumn Versus AgaroseGel Chromatography Electroelution.
Trial I Trial II
Colony Phenot~
DNA/treatment Blue White Blue Supercoiled pUC 19 55000 < 10 50000 <
pUC 19/HincII/CIAP 210 < 1 320 1 pUC 19/HincII/CIAP/ 150 4 210 7 T4 DNA ligase ~/CviJI** partial/CF 140 240 210 240 + pUC 19 ~/CvilI** partial/GFEl 98 49 200 18 + pUC 19 ~/CviJI** partial/GFE2 82 54 95 74 + pUC 19 Cloning efficiencies reflect the number of ampicillin-resistant colonies/ng pUC 19 DNA. CIAP represents treatment with calf intestinal alkaline phosphatase used to dephosphorylate HincII-digested pUCl9 to minimize self ligation. CF refers to DNA that was fractionated on Sephacryl S-500 columns as described above. GFEl and GFE2 refer to two runs wherein DNA was fractionated by agarose gel electrophoresis and electroeluted. ~ refers to bacteriophage ~ DNA.
These trials represent repeated experiments in which ~ DNA
fragments generated by G'viJI** partial digestion were ligated to HincII-linearized, dephosphorylated pUCl9 and transformed into DHSa F' competent cells described above.. The first three rows in Table 2 show controls performed to establish a baseline to better evaluate the various treatments. Supercoiled pUC 19 transforms E. coli 10 times more efficiently than the HincII-digested plasmid and 150-260 times more efficiently than the HincII-digested and dephosphorylated plasmid.
The number of blue and white colonies which resulted from transforming HincII-cut and dephosphorylated pUCl9 was determined both before and after treatment with T4 DNA ligase in order to differentiate these background events from cloning inserts. The background of blue colonies (which represent the uncut and/or non-dephosphorylated population of molecules) averaged 0.4 % , compared to supercoiled plasmid. The background of white colonies (which presumably results from contaminating nucleases in the enzyme treatments or genomic DNA
in the plasmid preparations) after HincII-digestion, dephosphorylation, and ligation of pUC 19 averaged 0.014 R'o as compared to the supercoiled plasmid.
The number of white colonies obtained when micro-column fractionated DNA was cloned into pUC 19 was 240/ng vector in both trials. The efficiency of cloning gel fractionated and electroeluted DNA ranged from 18-74 white colonies/ng vector. The data show that column fractionated DNA results in three to thirteen times the number of white colonies, and presumably recombinant inserts, as gel fractionated and electroeluted DNA. The size distribution of the inserts present in these white colonies is depicted in Figures l0A-C. In Figure 10A, a CviJI** partial digest of 2~,g of ~ DNA was size fractionated on a 4 mm by 13 mm column of Sephacryl S-500 at 2,000 x g for 5 minutes. The void volume containing partially digested DNA was directly ligated to linear, dephosphorylated pUC 19 and 43 resulting clones were analyzed for insert size. The DNA for this experiment is the same as that shown in Figure 9A, lane 2. In Figure 10B, a CvJI** partial digest of 5 ~cg of ~ DNA was size fractionated by agarose gel electroelution. The eluted DNA was phenol-extracted and ligated to linear, dephosphorylated pUC 19, and the resulting 40 clones were ~15908~.
analyzed for insert size. The DNA for this experiment is the same as that shown in Figure 9A, lane 3. In Figure lOC, the procedure is the same as in Figure 9B, except the DNA for this experiment came from Figure 9B, lane 3.
A total of 43 random clones obtained from micro-column chromatography fractionation were analyzed for insert size (as shown in Figure l0A). Most of these inserts were larger than 500 by (37/43 or 866), 11.6%
(5/43) were smaller than 500 bp, and one clone (2.3 % ) was smaller than 250 bp.
The average insert size was 1630 bp. These results are in contrast to those obtained by agarose gel fractionation (as shown in Figures lOB and lOC). In the first trial (Figure lOB) most of the inserts were smaller than 500 by (26/37 or 70. 3 9'0 ) and only 29.7 % ( 11 /37) were larger than 500 by in size. In the second trial (Figure lOC) all of the inserts (40 total) were smaller than 500 bp.
Thus, the use of agarose gel electroelution for the size fractionation of DNA
results in unexpectedly variable and low cloning efficiencies.
Example 16 Cloning Sonicated and CviJI**-Digested Lambda DNA
To compare the cloning efficiencies of sonicated and CviJI**
digested nucleic acid, ~ DNA was fragmented by each of these methods and ligated to pUCl9 which was linearized with HincII and dephosphorylated to minimize self ligation.
DNA fragmented by CviJI** digestion and sonication was cloned both before and after Sephacryl S-500 size fractionation. Sonicated lambda DNA
was subjected to an end repair treatment prior to ligation. Ligations were performed as described in Example 11. One-tenth of the ligation reaction (2 ~,1) was utilized in the transformation procedure, and the fraction of nonrecombinant (blue) versus recombinant (white) colonies was used to calculate the efficiency of this process.
~1~9~81 The efficacy of the methods was quantified by comparing the cloning efficiency of lambda DNA fragments generated either by sonication or G'viJI** partial digestion. To reduce potential cloning differences based on si~E
preference, the size distribution of the DNA generated by these two methods was closely matched. Other experimental details were designed to reduce potential variables, as described above. Certain variables were unavoidable, however.
For example, the sonicated DNA fragments required an enzymatic step to repair the ragged ends as described in Example 1 prior to ligation, whereas the C'viJI**
digests were heat-denatured and directly ligated to HincII digested pUCl9.
A summary of the cloning efficiencies calculated from two independent trials is given in Table 4, section A (unfractionated samples), and Section B (fractionated samples).
..
V
C
E~
.-.v~ cnt~ ..
a1 ..en ... n ov -.~
A.
A U
.c "
o V V ~ '~ o ...~ r, D
D
m ~
t ~f 'C o~
o U
a ' ~ ~ U
a U U ;~ ~
..
~
...~~..~ .~~ +
a a r_ o ~. ~ ~ ~ ov ~ a a , ~ ~ ~ V v +
- c . ~
~
_ _ A .~~ o o o.o U U a.
+
A n o o ~
~ . , E- ~ ~ ~ ~ ~ +
.. ..
v v v 'C
M ~
f '1 O
A
d .C
GL
O
_O
O
V
:_:
V V V
o ~ C
it~ E"'~
u M
C~t ~ N en N N
rf o a _ ao a w H
ai o, a a a a v v °~ V ~~
a ~~o~p°°" +W W
o_v ~ ~ ov ~ ov ,uo_vo_vz ~~V ~ ~U ~(j ~y V V ~1 .v > p" .o .o c, .o a.
A rn o. o. H ~ ~ + ~ ~ + ~ +
"r,r WO 94/21663 2moo~~
Cloning efficiencies represent the number of ampicillin-resistant colonies/ng pUCl9 DNA. CIAP indicates treatment with calf intestinal alkaline phosphatase. ER 1 and ER 2 are end repair methods described in Example 13.
~ refers to bacteriophage lambda.
The indicated trials represent repeated experiments in which two identical sets of lambda DNA fragments generated by AIuI complete digestion, G'viJI** partial digestion, or sonication were each ligated to HincII-linearized, dephosphorylated pUCl9 and transformed into DHSaF' competent cells. The cloning efficiencies reported are the average of triplicate platings of each ligation reaction. In case the Sephacryl S-500 size fractionation step introduced inhibitors of ligation or transformation or resulted in differences attributable to the size of the material, the sonicated and CviJI**-digested samples were ligated with pUCl9 both prior to (A) and after (B) the fractionation steps. The first three rows in Table 4, sections A and B, are controls performed to establish a baseline to better evaluate the various treatments. These data show that supercoiled pUCl9 transforms E. coli 200-1000 times more efficiently than the HincII-restricted and dephosphorylated plasmid. Without this dephosphorylation step, the cloning efficiency is 10% that of the supercoiled molecule (data not presented). The background of blue colonies averaged 0.5 % in these experiments, compared to supercoiled plasmid, while the background of white colonies averaged 0.005 % .
A comparison of the data from unfractionated versus fractionated samples in Table 4, sections A and B, reveals a general decline in the number of white and blue colonies obtained after sizing. This decrease is primarily due to the fact that cloning efficiencies are dependent upon the size of the fragment, favoring smaller fragments and thus giving higher efficiencies for the unfractionated material. This is illustrated by comparing the efficiency of cloning unfractionated and fractionated ~ DNA which was completely restricted with AIuI.
This four base recognition endonuclease produces blunt ends and cuts ~ DNA
(48,502 bp) at 143 sites. Only 25 of the resulting 144 fiagments (17% ) are larger 21.59081 than 500 bp. The number of white colonies obtained when unfractionated ~
DNA, completely restricted with AIuI, was cloned into pUCl9 ranged from 250-400/ng vector, versus 23-48/ng vector for the fractionated material. This ten fold decrease was only noticed for the ~ Alu I digests, and probably reflects the large portion of small molecular weight fragments (approximately 75 ~) which is excluded from the fractionated ligation reactions.
The number of white colonies obtained when unfractionated CviJI**
treated ~ DNA was cloned into pUCl9 ranged from 160-340/ng vector, versus 68-90 white colonies/ng vector if the same material was fractionated.
Unfractionated ~ DNA, completely digested with AIuI, results in cloning efficiencies very similar to unfractionated G~iJI** treated DNA. Sonicated ~ DNA is a poor substrate for ligation, compared to C~iJI** treatment, as indicated by the roughly ten-fold reduced cloning efficiencies.
Enzymatic repair of the ragged ends produced by sonication results in an increased cloning efficiency. Using conditions described in Example 13 for the first end repair treatment (ER 1), 10-44 (fractionated) and 19-32 (unfractionated) white colonies/ng vector were observed. However, ER 1 conditions may not be optimal, as an alternate end repair reaction (ER 2) (as described in Example 13) resulted in greater numbers of white colonies (63 and 100/ng vector for fractionated and unfractionated DNA, respectively). In this reaction, a ten-fold excess of reagents and enzymes were utilized to repair the sonicated DNA, which apparently improved the efficiency of cloning such molecules by two to three fold. The data collected from multiple cloning trials in Table 3, sections A and B, show that CviJI** partial digestion results in three to sixteen times the number of white colonies than sonicated ER 1-treated DNA.
Even with an optimal end repair reaction for the sonicated fragments, DNA
treated with G'viJI** yielded three times more white colonies.
~15~0~~.
Example 17 Analysis of CviJI** Fragmentation for Shotgun Cloning and Sequencing The ability of CviJI** partial digestion to create uniformly representative clone libraries for DNA sequencing was tested on pUC 19 DNA.
pUCl9 DNA was digested under CviJI** conditions and size fractionated as described above. The fractionated DNA was cloned into the EcoRV site of M13SPSI, a lacZ minus vector constructed by adding an EcoRV restriction site to wild type M13 at position 5605. M13SPSI lacks a genetic cloning selection trait, therefore after ligation of the pUC 19 fragments into the vector the sample was restricted with EcoRV to reduce the background of nonrecombinant plaques.
l3acteriophage M13 plaques were picked at random and grown for 5-7 hours in 2 ml of 2XTY broth containing 20 ~cl of a DHS«F' overnight culture. After centrifugation to remove the cells, single-stranded phage DNA was purified using Sephaglass'"' as specified by the manufacturer (Pharmacia LKB, Piscataway New Jersey). The single-stranded DNA was sequenced by the dideoxy chain termination method using a radiolabeled M13-specific primer and Bst DNA
polymerise (Mead et al., Biotechniques 11:76-87 (1991)). The first 100 bases of 76 randomly chosen clones were sequenced to determine which CviJI recognition site was utilized, the orientation of each insert and how effectively the cloned fragments covered the entire molecule, as shown in Figure 11. The positions of the 45 normal CviJI sites (PuGCPy) in pUCl9 are indicated beneath the line labeled "NORMAL" in the Figure 11. Similarly, the 160 CviJI* sites (GC) are indicated beneath the line labeled "RELAXED" in Figure 11. The marks above these lines indicate the CviJI** pUCl9 sites which were found in the set of 76 sequenced random clones. The frequency of cloning a particular site is indicated by the height of the line, and the left or right orientation of each clone is also indicated at the top of each mark. There are a total of 205 CviJI and CviJI*
sites in pUCl9.
The data presented in Figure 11 demonstrate that, under G'viJI**
partial conditions, normal G'viJI sites are preferentially restricted over relaxed (CviJI*) sites. Of the 76 clones that were analyzed, only 13%, or 1 in 7, had sequence junctions corresponding to a relaxed G'viJI* site. Thirty-five of the forty-five possible normal restriction sites'-were cloned, as compared to eight of the possible one hundred sixty relaxed sites. If the enzyme had exhibited no preference for normal or relaxed sites under the G~iJI** partial conditions utilized here, then 78 % of the sequence junctions analyzed should have been generated by cleavage at a relaxed G'viJI* site. It may be noted that the relaxed C'viJI*
restriction sites that were found appear to be clustered in two regions of the plasmid that are deficient in normal G'viJI sites. In addition, the combined distribution of the normal and relaxed sites which were restricted to generate the 76 clones appears to be quasi-random. That is, the longest gap between cloned restriction sites was no greater than 250 by and no one particular site is over-utilized.
A detailed analysis of the distribution of CvilI** sequence junctions found from cloning pUCl9 is presented in Table 5.
*
*~
N
~
a -. ri ri ~O M
.) m O~ ~D ~f1 ~ O~
~
11 V ., !~ ri ~
a 1(1 N v 10 ~
v ~ vv~~ ~ ..
.
[A
~'~1 ~O O N ~
N llfy0 N~ Il1 'i rl rl N
~O
O la U
Ca ,o a .,, _____ __ p ~srav~o aro~
o~no~
i ~ r i r i ~
.i u u do c u ~ a o vo ... .-. .r ...
r. .. ....1 N N
b ~ -- r...
..i .a rn a x m ~ o~ .~ .1 N
o Inl o NI
u, In rl ~-1 v-i 'i e-i d' ri N
H a~
.~
~ ~
C H H
~
b b O
.i U
w a O
'~~ o d ..1 U
U U
H H
C U U
U U
p O ~ U
O
C
O
-..1 ro z --U
-.1 .d O
w -a G1 O
.a ro x s~
m ro ~
m s~ .., ro o a x a U
~9 8~
2~
.
a, a W
' w w ~ n' w _ __ H
h C ~"~ N vo 'H O
D ..1 rl m0 r. ..
.1O Od~pn OO OOIO
C f.~
O ~
..r m U ~.-1 A
t1 Ov d' Q~ t~
r CD O~ t'~I
~
~ ri ui W c a o ~ r~ ri .r ..
.r .r .. .a b -a ~
o.
a p ~o ao .i O e~f N n O m I aD
r-1 rl rl '-1 N e-1 lf1 ri d' te UU UU UU U
~
..
a N U H
N
C
O C!
-.a U ,~ U C~ ~ U C9 ~ d O ~ U C9 U U N
O
O N
U t/!
4l C
a a ro U 'd 'O
N
x x ro m c~
m a a U
215~08.~
The GC sites in pUCl9 may be divided into four classes based on their flanking Pu/Py structure. The fraction of GC sits observed in pUCl9 which belong to each classification is roughly equal (22.0-27.8%). A striking difference was found between the observed distribution in pUCl9 of normal and relaxed (R1, R2, R3) G~iJI recognition sites and the distribution revealed by shotgun cloning and sequence analysis of CviJI**-treated DNA. While most of the sites cleaved by this treatment were found to be PuGCPy (about 87 % ), or "normal"
restriction sites, a significant fraction of the cleavage occurred at PyGCPy (about 6.5%) and PuGCPu (about 6.6 % ) sites, considering the short incubation times and limiting enzyme concentrations. The latter two categories of sites, and presumably the PyGCPu sites as well, are completely restricted under "relaxed" conditions, provided an excess of enzyme is present and sufficient time is allowed (see Figure 7, and Xia et al., Nucleic Acids Res. 15:6075-6090 (1987)).
Digestion using G'viJI** treatment results in a relatively even distribution of breakage points across the length of the molecule (as shown in Figure 11). As described above, Figure 11 depicts a linear map of pUCl9 showing the relative position of the lacZ' gene (a peptide of ~-galactosidase gene) and ampicillin resistance gene (Amp). The marks extending beneath the top line (labeled "NORMAL") show the relative position of the 45 normal CvilI sites (PuGCPy) present in pUCk9. The marks above the line are the cleavage sites found from sequencing the G~iJI** partial library. The height of the line indicates the number of clones obtained from cleavage at that site, and the orientation of the flag designates the right or left orientation of the respective clone. The marks extending beneath the second line (labeled "RELAXED") show the relative positions of the 160 G'viJI* sites (GC) present in pUCl9. Those marks above the line were found from sequencing the G'viJI** partial library. The bottom portion of Figure 11 shows the relative position and orientation of the first 20 clones sequenced, assuming a 350 by read per clone. CviJI** cleavage at relaxed sites appears to be important in "filling gaps" left by normal restriction.
.~ PCTIUS94I03246 _.
The primary goal of this effort was to determine the efficacy of these methods for rapid shotgun cloning and sequencing. For these purposes, only 100 bases of sequence data were acquired per clone. However, if 350 bases of sequence had been determined from each clone, then the entire sequence of pUCl9 would have been assembled from the overlap of the first 20 clones (Figure 11). In this sequencing simulation 75% of pUCl9 would have been sequenced at least 2 times from the first 20 clones. The highest degree of overfold sequencing would have been 6, and only involved 2.2 % of the DNA. Figure 11 also shows that most of the lx sequencing coverage occurred in a region of the plasmid with a very low density of normal and relaxed G'vilI restriction sites.
Most of the single coverage occurs in a 240 by region of the plasmid between 1490 by and 1730 by where there are only 4 G~iJI relaxed sites. It should also be noted that by the 27th randomly picked clone most of this region would have been covered a second time.
Shotgun sequencing strategies are efficient for accumulating the first 80-95 % of the sequence data. However, the random nature of the method means that the rate at which new sequence is accumulated decreases as more clones are analyzed. In Figure 12 the total amount of unique pUCl9 sequence accumulated was plotted as a function of the number of clones sequenced. The points represent a plot of the total amount of determined pUCl9 sequence versus the total number of clones sequenced. The horizontal dashed line demarcates the 2686 by length of pUCl9. The smooth curve represents a continuous plot of the discrete function S(I~=NIx-cs[((ecs_1)/c)+(1-s)). The theoretical accumulation curve expected for a process in which sequence information is acquired in a totally random fashion is also shown. The smooth curve is a continuous plot of the discrete function S(1~ where S(N)=NI-a c~(((ec~_1)/c+(1-Q)).
This equation is based upon the results developed by Lander et al. , Genomics 2:231-239 (1988) for the progress of contig generation in genetic mapping. In the 2~.5~081 equation: N is the number of clones sequenced, L is the length of clone insert in bp, c is the redundancy of coverage or LN/G (where G is length of fragment being sequenced in bp), and Q = 1-8, where 8 is the fraction of length that two clones must share. The curve in Figure 12 was calculated with G = 2686 bp, L
= 350 bp, and Q = 1. The plotted points lie close to the theoretical curve, and it thus appears that the sequence of pUCl9 was accumulated in an apparent random fashion utilizing CviJI** fragmentation and column fractionation.
Example 18 Shotgun Cloning Utilizing 200 ng of Lambda DNA
Generally, 2-5 ~cg of DNA are needed for the sonication and agarose gel fractionation method of shotgun cloning in order to provide the several hundred colonies or plaques required for sequence analysis (Banlaer et al.
Methods in Enzymol. 155:51-93 (198'0. A ten-fold reduction in the amount of substrate required greatly ' 'fies the construction of such libraries, especially from large genomes, (Davi~son, J. DNA Sequencing and Mapping 1:389-394 (1991)). The efficiency of constructing a large shotgun library from nanogram amounts of substrate was tested utilizing 200 ng of CviJI**-digested lambda DNA.
This material was column-fractionated as described previously. In this case, of the column eluant (15 ~cl containing 50 ng of DNA) was ligated to 100 ng of HincII-digested and dephosphorylated pUC 19 as described in Example 15. The cloning efficiencies of the control DNAs were similar to those reported in Tables 2 and 3. The 50 ng cloning experiment yielded 230 white colonies per ligation reaction in one trial, and 410 white colonies per ligation reaction in a second trial.
Thus, it should be possible to routinely construct useful quasi-random shotgun libraries from as little as 0.2 - 0.5 ~,g of starting material.
WO 94/21663 PCT/L1S94103246 _ 21~J~1g1 Example 19 Epitope Mapping G'viJI* recognizes the sequence GC (except for PyGCPu) in the target DNA. Under partial restriction conditions the length of fragment may be controlled by incubation time. Epitope mapping using G'viJI** partial digests involves generating DNA fragments of 100-300 by from a cDNA coding for the protein of interest, by methods described in Example 13, inserting them into an M13 expression vector, plating out on solid media, lifting plaques onto a membrane, screening for binding to the ligand of interest, and picking the positive plaques for isolation of the DNA, which is then sequenced to identify the epitope.
Thus, the same epitope may be expressed as a small fragment or a larger fragment. This approach allows one to determine the smallest fragment containing the epitope of interest using functional assays such as binding to an antibody or other ligand, or using a direct assay for activity. For insertion into an M13 vector, linkers may be added to the fragments or the insert may be dephosphorylated to ensure that each fragment is cloned alone without ligation of multiple inserts.
The expression vectors recommended for subcloning of the CviJI
fragments are Lambda Zap (Stratagene, LaJolla, California) or bacteriophage M13-epitope display vectors. An advantage of using an M13-based vector is that the peptide or protein of interest may be displayed along with the M13 coat protein and does not require host cell lysis in order to analyze the protein of interest. The lambda-based vectors yield plaques and hence the protein can be directly bound to a membrane filter.
215908.
_77_ Example 20 CGase I
CGase I as used herein, refers to a restriction endonuclease reagent which cleaves DNA at the dinucleotide CG. CGase I activity is based on the combined star activities of the restriction endonucleases Hpa II and Taq I. Under normal reaction conditions (10 mM Bis Tris Propane-HCl pH 7.0, 10 mM MgCl2, 1 mM
DTT; 1 unit of enzyme/~cg DNA, 37°C for 1 hr), Hpa II recognizes CCGG and cleaves after the first C to leave a 2-base 5' overhang. Under normal reaction conditions (100 mM NaCI, 10 mM Tris-HCl pH 8.4, 10 mM MgCl2, 10 mM 2-mercaptoethanol, 1 unit of enzyme/~,g DNA, 65°C for 1 hr) the restriction endonuclease Taq I recognizes TOGA and cleaves after the T to leave a 2-base 5' overhang.
Reaction conditions have been described for Taq I* activity which decrease the cleavage specificity of Taq I (10 mM Tris-HCl pH 9.0, 5 mM MgCl2, 6 mM
2-mercaptoethanol, 20% DMSO; 2000 units of enzyme/~cg DNA, 65°C for 1 hr) (Barany, Gene, 65:149-165 (1988)). These reaction conditions allow Taq I* to cleave DNA at the following sequences:
Taq I* TCGA
CCGA (TCGG) ACGA (TCGT) TCTA (TAGA) TCAA (TTGA) GCGA (TCGC) We are unaware of any literature descriptions of Hpa II* conditions.
However, the following conditions were established to promote Hpa II* activity which are also compatible with Taq I* activity: 5 mM KCI, 10 mM Tris-HCl pH
8.5, 10 mM MgCl2, 1 mM DTT, 15 % DMSO, 100 ug/ml BSA (CGase buffer);
50 units of enzyme/~g DNA 50°C for 1 hr. The Hpa II* recognition sites were WO 94/21663 PCT/US94/03246 _ _78_ determined by cloning and sequencing Hpa II* restricted fragments. The characterized Hpa II* recognition sequences are as follows:
Hpa II* CCGG
CCGC (GCGG) CCGA (TCGG) ACGG (CCGT) Taq I (400 units/~cg DNA) and Hpa II (50 units/~,g DNA) were then combined (CGase I) in CGase I buffer and the following recognition sites were identified by cloning and sequencing restricted pUCl9 fragments.
CGase I GCGC
TCGA
CCGG
GCGT
ACGA
ACGG (CCGT) GCGG (CCGC) CCGA (TCGG) CGase I restriction of natural DNA, (i.e. pUCl9, lambda), results in fragments ranging from 20-200 by in length (average 20-60 bp). Heat denaturation of these fragments generates numerous oligonucleotides of variable length but precise specificity for the cognate template as was the case with CviJ I* digestion.
CGase I restriction of the small plasmid pUC 19 (2689 bp) theoretically yields 174 restriction fragments, or 384 oligonucleotides after a heat denaturation step.
The "two-cutter" activity of CviJ I* and CGase I represent a unique class of restriction endonuclease activity in that no other known restriction endonucleases will generate this size range of oligonucleotides. The ability to generate numerous oligonucleotides with perfect sequence specificity from any DNA, without regard to sequence composition, genetic origin, or prior sequence knowledge is one of the properties that CGase I shares with CviJ I*. In addition, the generation of numerous oligonucleotides by CviJ I or CGase I results in a form of probe or primer amplification not practical using conventional means of organic synthesis.
Based on ability to recognize a dinucleotide sequence, the present invention contemplates the interchangeability of CGase I with CviJ I* in all of the applications described herein.
Example 21 Purification of Cv~ I Restriction Endonuclease from IIr3A-Infected Chlorella Cells CviJ I was prepared by a modification of the method described by Xia et al., Nucl. Acids Res. 15:6025-6090 (1987). Chlorella NC64A cells (ATCC Accession No. 75399 deposited on January 21, 1993, American Type Culture Collection, Rockville, Maryland) were infected with the virus IL-3A
(ATCC Accession No. 75354 deposited November 6, 1992, American Type Culture Collection, Rockville, Maryland) according to Van Etten et al. , Yrology 126:117-125 (1983). Five grams of IL-3A infected Chlorella NC64A cells were suspended in a glass homogenization flask with 15 g of 0.3 mm glass beads in buffer A (10 mM Tris-HCl pH 7.9, 10 mM 2-mercaptoethanol, 50 ~cg/ml phenylmethylsulfonyl fluoride (PMSF), 20 ug/ml benzamidine, 2 ~g/ml o-phenanthroline). Cell lysis was carried out at 4000 rpm for 90 sec in a Braun MSK mechanical homogenizer (Allentown, PA) with cooling from a C02 tank.
After lysis 2 M NaCI was added to a final concentration of 200 mM, after which 10 % polyethyleneimine (PEn (Life Technologies, Bethesda, MD) (pH 7.5) was added to a final concentration of 0.3 % . The mixture was then stirred for 2 hrs.
at 4°C then centrifuged for 1 hr. at 50,000 g. Ammonium sulfate was added to the supernatant to 70 % saturation and stirred overnight. A protein pellet was recovered by centrifugation for 1 hr. at 50,000 g. The resulting pellet was dissolved in 20 ml of buffer B (20 mM Tris-acetate,pH 7.5, 0.5 mM EDTA, 10 ~ w0 94121663 PCT/US94/03246 - so -mM 2-mercaptoethanol, lpSo glycerol, 30 mM KCI, 50 ug/ml PMSF, 20 ~cg/m1 benzamidine [Sigma, St. Louis, Missouri), 2 ~cg/ml o-phenanthroline [Sigma]) and dialysed against 500 ml of buffer B with 3 changes. The dialysed solution was then applied to 1 x 6 cm Heparin-SepharoseTM (pharmacia LKB, Piscataway, New Jersey) column. After a 50 ml wash with buffer B, a 100 ml gradient of 0 to 0.7 M KCl in buffer B was run. Fractions having CviJ I activity as measured by digestion of pUC 19 DNA and agarose gel electrophoresis, were pooled, diluted in 5 volumes of buffer C ( 10 mM K/P04 pH 7.4, 0.5 mM EDTA, 10 mM
2-mercaptoethanol, 75 mM NaCI, 0.05% Triton X-100, 10% glycerol, SO p.g/ml PMSF, 20 ~cg/ml benzamidine, 2 ~cg/ml o-phenanthroline) and applied to a 1 x 7 cm Phosphocellulose P11 (Whatman) column equilibrated in buffer C. After washing with 30 ml of buffer C, CviJ I was eluted by a 100 ml gradient of 0 to 0.7 M NaCI in buffer C. At this step Cvi1 I activity separated from non-specific nucleases. CviJ I containing fractions were pooled and diluted in 4 volumes of i5 buffer C and applied to a 1 x 4 cm hydroxyapatite HTP column (BioRad, Hercules, CA). After washing with 30 ml of buffer C, CviJ I was eluted by a 0 to 0.7 M potasium phosphate (pH 7.4) gradient in buffer C. Active fractions containing CviJ I activity and lacking non-specific nuclease activity were pooled and were dialysed overnight against storage buffer (50 mM potassium phosphate 200 mM KCI, 0.5 mM EDTA, 50~ glycerol, 20 ug/ml PMSF were pooled) and stored at -20°C.
Although the present invention has been described in types of preferred embodiments, it is intended that the present invention encompass all modifications and variations which occur to those skilled in the art upon consideration of the disclosure herein, and in particular those embodiments which are within the broadest proper interpretation of the claims and their requirements.
~~ -a SEQUENCE LISTING
(1) GENERAL INFORMATION:
(i) APPLICANT: Molecular Biology Resources, Inc.
(ii) TITLE OF INVENTION: Materials and Methods for Restriction Endonuclease Applications (iii) NUMBER OF SEQUENCES: 13 (iv) CORRESPONDENCE ADDRESS:
(A) ADDRESSEE: Marshall, O~Toole, Gerstein, Murray & Borun (B) STREET: 6300 Sears Tower, 233 South blacker Drive (C) CITY: Chicago (D) STATE: Illinois (E) COUNTRY: United States of America (F) ZIP: 60606-6402 (v) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Floppy disk (B) COMPUTER: IBM PC compatible (C) OPERATING SYSTEM: PC-DOS/MS-DOS
(D) SOFTWARE: PatentIn Release ,~1.0, Version #1.25 (vi) CURRENT APPLICATION DATA:
(A) APPLICATION NUMBER:
(B) FILING DATE:
(C) CLASSIFICATION:
(viii) ATTORNEY/AGENT INFORMATION:
(A) NAME: Clough, David W.
(B) REGISTRATION NUMBER: 36,107 (C) REFERENCE/DOCKET NUMBER: 28003/31967/PCT
(ix) TELECOMMUNICATION INFORMATION:
(A) TELEPHONE: 312/474-6300 (B) TELEFAX: 312/474-0448 (C) TELEX: 25-3856 (2) INFORMATION FOR SEQ ID NO: l:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 44 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID NO: l:
~~g~~~ _ g2 (2) INFORMATION
FOR SEQ
ID N0:2:
(i) S EQUENCE CHARACTERISTICS:
(A) LENGTH: 5496 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE
TYPE:
DNA (genomic) (xi) S EQUENCE DESCRIPTION: SEQ
ID N0:2:
AGTTATAAAT
TAAATCCAGA
ATCCCCGACC
AGACATAACC
ATATCTAAAA
ATCGTTCAAG
GACCGTATGG
TTACGGTAAG
AACAAAAAGA
CTTGGAGAAT
TAGCGGGAGA
TATACATCCA
CGAGAGACTT
TAGAAGCAGG
TGGTGATATT
CTTAGATTTT
TTCAGCTTTG
TTCTGCTAAT
ATCATCGTCA
ATCTCGTGTG
ATGTCCATCC
ATAACTATAA
TTCAGATCTA
ATCATCCGAT
CATCTCTGAA
ACGTTTCATA
TAATGGATAA
TGAAGATGAT
GAAATACACC
TTGACGCAGG
CGGTTTTGAA
AATTT~GGATGGATACCGTTACATTACGCGGCTTTTAATGGTAATGATGCGATTTTGAGG3420 GGTGGTTTCG
GGTGGTTTCG
GGTGGTTTCG
21~948~.
-ss-ATCCGTTAAATTCCCGCATTCACCTAATGATGTACTCCATAAAGAACCGGG:~~GCGCATTG5280 (2) INFORMATION FOR SEQ ID N0:3:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 1225 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (ix) FEATURE:
(A) NAME/KEY: CDS
(B) LOCATION: join(1..33, 55..1128) (xi) SEQUENCE DESCRIPTION: SEQ ID N0:3:
GAA TTG TAT TGATACACTA
TAT
CTT
GGT
GlnGlu Tyr Leu Tyr Val Gln Asp Gly Leu Tyr ATG
GAT
ATA
AGA
AGA
AAA
CGT
TTT
ACA
ATA
GAA
GGG
GCT
AAA
CGT
Met Asp Ile Arg Arg Lys Arg Phe Thr Ile Glu Gly Ala Lys Arg AAA AGA GAG GCG
IleIle Leu Glu Lys Leu Glu Lye Lys Arg Ile Glu Lys Arg Glu Ala ATT CTT AAA GAA
GluLys Lye Arg Ala Ile Glu Gln Arg Ile Ala Glu Ile Leu Lye Glu GCG GAG CGA GAG
LysLye Arg Ile Glu Lys Lys Phe Ala Leu Glu Lys Ala Glu Arg Glu GAA AAA ATC AAA
LyeArg Ile Ala Glu Lys Arg Ala Glu Glu Lys Arg Glu Lys Ile Lye AAA AGA CTT CGA
IleVal Glu Glu Lys Leu Ala Ile Glu Lye Gln Ile Lys Arg Leu Arg ATT TCG AAA ATT
AGA AAG
AGG ATC
AlaGlu Glu Lys Ala Gly Arg Ile Arg Lys Arg Ser Ile Ser Lys Ile 15908.
ACA GAA AAA
AAA GTT
ThrAsn AlaThrLyeHis GluArgGl.uPhe ValLysValIle AanSer MetPhe ValGlyProAla ThrPheValPhe ValAapIleLye GlyAan LysSer ArgGluIleHis AsnValValArg PheArgGlnLeu GlnGly SerLya AlaLyeSerPro ThrAlaTyrVal AapArgGluTyr AanLys ProLys AlaAepIleAla AlaValAspIle ThrGlyLyaAsp ValAla TrpIle SerHiaLyeAla SerGluGlyTyr GlnGlnTyrLeu LysIle SerGly LysAanLeuLye PheThrGlyLys GluLeuGluGlu ValLeu SerPhe LyaArgLyaVal ValSerMetAla ProValSerLya IleTrp ProAla AsnLyaThrVal TrpSerProIle LyaSerAanLeu IleLya AanGln AlaIlePheGly PheAspTyrGly LyeLyeProGly ArgAsp AanVal AspIleIleGly GlnGlyArgPro IleIleThrLys ArgGly SerIle LeuTyrLeuThr PheThrGlyPhe SerAlaLeuAan GlyHie LeuGlu AanPheThrGly LysHisGluPro ValPheTyrVal ArgThr GluArg SerSerSerGly ArgSerIleThr ThrValValAan GlyVal ThrTyr LysAanLeuArg PhePheIleHis ProTyrAsnPhe ValSer _87_ Ser Lye Thr Gln Arg Ile Met (2) INFORMATION FOR SEQ ID N0:4:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 369 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID N0:4:
Gln Glu Tyr Leu Gly Tyr Leu Val Gln Tyr Aep Met Asp Ile Arg Arg Lye Arg Phe Thr Ile Glu Gly Ala Lys Arg Ile Ile Leu Glu Lye Lys Arg Leu Glu Glu Lys Lys Arg Ile Ala Glu Glu Lys Lys Arg Ile Ala Leu Ile Glu Lys Gln Arg Ile Ala Glu Glu Lys Lys Arg Ile Ala Glu Glu Lys Lys Arg Phe Ala Leu Glu Glu Lys Lys Arg Ile Ala Glu Glu Lys Lys Arg Ile Ala Glu Glu Lys Lys Arg Ile Val Glu Glu Lys Lys Arg Leu Ala Leu Ile Glu Lys Gln Arg Ile Ala Glu Glu Lys Ile Ala Ser Gly Arg Lys Ile Arg Lye Arg Ile Ser Thr Asn Ala Thr Lye His Glu Arg Glu Phe Val Lys Val Ile Asn Ser Met Phe Val Gly Pro Ala Thr Phe Val Phe Val Asp Ile Lye Gly Asn Lys Ser Arg Glu Ile His Aen Val Val Arg Phe Arg Gln Leu Gln Gly Ser Lys Ala Lye Ser Pro Thr Ala Tyr Val Asp Arg Glu Tyr Asn Lye Pro Lys Ala Asp Ile Ala Ala Val Asp Ile Thr Gly Lys Asp Val Ala Trp Ile Ser His Lys Ala Ser Glu Gly Tyr Gln Gln Tyr Leu Lys Ile Ser Gly Lys Asn Leu Lye WO 94/21663 PCT/US94I03246 _ _88_ Phe Thr Gly Lye Glu Leu Glu Glu Val Leu Ser Phe Lye Arg Lys Val Val Ser Met Ala Pro Val Ser Lys Ile Trp Pro Ala Asn Lys Thr Val Trp Ser Pro Ile Lys Ser Asn Leu Ile Lys Asn Gln Ala Ile Phe Gly Phe Asp Tyr Gly Lys Lye Pro Gly Arg Aep Asn Val Asp Ile Ile Gly Gln Gly Arg Pro Ile Ile Thr Lys Arg Gly Ser Ile Leu Tyr Leu Thr Phe Thr Gly Phe Ser Ala Leu Asn Gly His Leu Glu Asn Phe Thr Gly Lys His Glu Pro Val Phe Tyr Val Arg Thr Glu Arg Ser Ser Ser Gly Arg Ser Ile Thr Thr Val Val Aen Gly Val Thr Tyr Lys Asn Leu Arg Phe Phe Ile His Pro Tyr Asn Phe Val Ser Ser Lys Thr Gln Arg Ile Met (2) INFORMATION FOR SEQ ID N0:5:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 17 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:5:
(2) INFORMATION FOR SEQ ID N0:6:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 16 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:6:
21~908~
(2) INFORMATION FOR SEQ ID N0:7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 31 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:7:
(2) INFORMATION FOR SEQ ID N0:8:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 14 baee pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:8:
(2) INFORMATION FOR SEQ ID N0:9:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 33 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:9:
(2) INFORMATION FOR SEQ ID NO:10:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 81 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10:
(2) INFORMATION FOR SEQ ID NO:11:
(i) SRQUENCE CHARACTERISTICS:
(A) LENGTH: 81 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11:
(2) INFORMATION FOR SEQ ID N0:12:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 270 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (ix) FEATURE:
(A) NAME/KEY: CDS
(B) LOCATION: join(26..148, 190..207, 244..270) (xi)SEQUENCE DESCRIPTION:
SEQ ID N0:12:
ACACAGGAAA ATG
CAGCT
Met Thr IleThrProSer SerLye Met AAC TGG
LeuThr Leu Thr Lys Gly Lye Ser TyrArgGlyPro ProSer Asn Trp AGC AAC
ArgSer Thr Val Ser Ile Leu Ile HisLeuTyrAsn LysArg Ser Asn TTGTATATAC ATG
GTCATTTCGT TTA
TATATCAACA TCA
TAT
Met Leu Ser Tyr ACGAGGTGTA ACTATA
TyrThr MetSer PheArg ThrLeu Glu Leu Phe (2) INFORMATION FOR SEQ ID N0:13:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 56 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID N0:13:
Met Thr Met Ile Thr Pro Ser Ser Lys Leu Thr Leu Thr Lys Gly Asn Lys Ser Trp Tyr Arg Gly Pro Pro Ser Arg Ser Thr Val Ser Ile Ser Leu Ile Asn His Leu Tyr Aen Lys Arg Met Leu Ser Tyr Tyr Thr Met Ser Phe Arg Thr Leu Glu Leu Phe '~~9a~1 INDICATIONS RELATING TO A DEPOSITED MICROORGANISM
(PCT Rulc l3bis) relate to the mtaoorgatusm referred to in the description A. The indications made bel ~9 on page , line 13 .
B. IDENTIFICATION OF DEPOSIT ~
, Further deposits are identified on an additional sheet XQ
Name of depositary institution American Type Culture Collection Address of depositary institution (iecLrdin; posro! cods and cornrry) 12301 Parklawn Drive Rockville, Maryland 20852 UNITED STATES OF AMERICA
Date of deposit Accession Number November 6, 1992 A.T.C.C. 75354 C. ADD1'TIONAL INDICATIONS (learbla~i/eoropplicable) 'Ibis infortnstion is oontioued on an additional sheet a "In respect of those designations in which a European patent is sought, a sample of the deposited microorganism will be made available until the publication of the mention of the grant of the European patent or until the date on which the application has been refused or withdrawn or is deemed to be withdrawn, only by the issue of such a sample to an expert nominated by the person requesting the sample (Rule 23(4) EPC)."
D. DESIGNATED STATES FOR WHICH
INDICATIONS ARE MADE (ijtJ~ciedicW
owrmaat fmUJai~cdStatesl E. SEPARA'Tlr FIJRN1SSING OF INDICATIONS
IJ~ blob i/aat applicrbk) IbeindicationslistedbelowwillbesubttaittedtotbelntetmtiomlBureaulater(spai/rthe ~Jnetyrco/tJriwlicoLO~se;., Acceuiae Nrwba o/Depoait7 For receiving Office use only For International Bureau use only 'Ibis sheet was received with the ' ternational application a 'ibis sheet was received by the international Bureau on:
~~~~~ / ~~~~
Authorized officer Form PCT/ROI134 (July 1992) WO 94/21663 9 3 ~ 215 9 0 81 pCT~S94/03246 INDICATIONS REL~ITIIYG TO A DEPOSITED MICROORGANISM
(PCT Rule l3bis) A. the indicauoru made below relate to the mtaoorgantsm referred to in the description on page 79 ~ line 10 B. IDENTIFICATION OF DEPOSIT ~
Further deposits sre identified on an sdditionsl sheet QX
Name of depositary institution American Type Culture Collection Address of deposiary institution /ieclrJias pasral code and coynt~y) 12301 Parklawn Drive Rockville, Maryland 20852 UNITED STATES OF AMERICA
Date of deposit January 21, 1993 Accession Number A.T.C.C. 75399 C. ADDITIONAL IIYDICAT10NS U~bLnJ~ijna opplicabk) Tbis information is continued on an additional sheet "In respect of those designations in which a European patent 1s sought, a sample of the deposited microorganism will be made available until the publication of the mention of the grant of the European patent or uatil the date on which the application has been refused or withdrawn or is deemed to be withdrawn, only by the issue of such a sample to an expert nominated by the person requesting the sample (Rule 23(4) EPC)."
D. DESIGNATED STATES FOR WHICH
INDICATIONS ARE MADE (i(tluidiccoamnaor/omOloi~atdStata) E. SEPARATE FIJRNISSING OF INDICATIONS
liaeve blad iywr apphcrbk) 'Ibe indicatiotm listedbelowwiU
be submitted to the Internatioml Bureau later (:pal/yrlu~lsamrc/thsulinrio~aej., 'Acccsioe Nrariw ojDepocit7 For receiving Office use oaly For international Buresu use only 'Ibis sheet was re ived with the interstational a lication a 'This sheet was received by the international Bureau on:
»ri~~'~~~~~
Authorized officer Fotm PCr/R0/134 (July 1992) ~~~,9p81 (PCT Rule l3bis) A. 'Ihe indicauotu made below relate to the mtaoorgantsm referred to in the description on page 31 , line 25 B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional :beet Q
Name of depositary itutitution American Type Culture Collection Address of depositary institution (i~rl~di~a pnsrol code and co~nay/
12301 Parklawn Drive Rockville, Maryland 20852 UNITED STATES OF AMERICA
Date of deposit Accession Number June 30, 1994 A.T.C.C. 69341 C. ADDITIONAL INDICATIONS (laswrbloni;
ilea opplicoblr/ This information is continued on an additional :beet Q
"In respect of those designations in which a European patent is sought, a sample of the deposited microorganism will be made available until the publication of the mention of the grant of the European patent or until the date on which the application has been refused or withdrawn or is deemed to be withdrawn, only by the issue of such a sample to an expert nominated by the person requesting the sample (Rule 23(4) EPC)."
D. DESIGNATED STATES FOR WHICH INDICATIONS
ARE MADE (i/tlrcindicodaaren~wr/orilJlaignarcdStata/
E. SEPARATE FIJRMSHING OF INDICATIONS
(laove blank i/aot oppliabk) The indications listed beiowwill be submitted to the international Bureau later (Jpai/ytheaotaol aeuinof tl~sielicrrioas e.a., 'Accesnoe Nawb~s o/Dep~t~
For receiving Office use only ~ ~ For International Bureau use only Tbis:beet was rcce ved itb the international a ~ ~ Q This :beet was received by the international Bureau on:
>/?r~i:i~~ / /'.
Authorized officer Form PCT/R0/l34 (July 1992)