METHOD FOR CLONING AND EXPRESSION OF TSPRI RESTRICTION ENDONUCLEASE AND TSPRI METHYLASE IN E. COLI BACKGROUND OF THE INVENTION The present invention relates to recombinant DNA that encodes the TspRI restriction endonuclease (TspRI endonuclease or TspRI) as well as TspRI methyltransferase (TspRI methylase or M. TspRI), expression of TspRI endonuclease and methylase in E. coli cells containing the recombinant DNA.

TspRI endonuclease is found in the strain of Thermus species R (New England Biolabs'strain collection). It recognizes the double-stranded, palindromic DNA sequence 5'NNCASTGNN3' (SEQ ID NO : 1) (S = C or G, ffi indicates the cleavage position) and cleaves on both sides of the recognition sequence, generating a 9-base 3'overhang. TspRI methylase (M. TspRI) is also found in the same strain. It recognizes the double-stranded DNA sequence 5'CASTG 3' (SEQ ID NO : 2) and presumably modifies the cytosine at the C5 position on hemi-methylated or non-methylated TspRI sites.

Type II restriction endonucleases are a class of enzymes that occur naturally in bacteria and in some viruses. When they are purified away from other bacterial/viral proteins, restriction endonucleases can be used in the laboratory to cleave DNA molecules into small fragments for molecular cloning and gene characterization.

Restriction endonucleases recognize and bind particular sequences of nucleotides (the recognition sequence') along the DNA molecules. Once bound, they

cleave the molecule within (e. g. BamHI), to one side of (e. g. SapI), or to both sides (e. g. TspRI) of the recognition sequence. Different restriction endonucleases have affinity for different recognition sequences. Over two hundred and eleven restriction endonucleases with unique specificities have been identified among the many hundreds of bacterial species that have been examined to date (Roberts and Macelis, Nucl. Acids Res. 27: 312-313 (1999)).

Restriction endonucleases typically are named according to the bacteria from which they are discovered. Thus, the species Deinococcus radiophilus for example, produces three different restriction endonucleases, named DraI, DraII and DraIII. These enzymes recognize and cleave the sequences 5'TTTAAA 3' (SEQ ID NO : 3), 5'PuGGNCCPy 3' (SEQ ID NO : 4) and 5'CACNNNGTG 3' (SEQ ID NO : 5) respectively.

Escherichia coli RY13, on the other hand, produces only one enzyme, EcoRI, which recognizes the sequence 5' G ; AATTC 3' (SEQ ID NO : 6).

A second component of bacterial/viral restriction- modification (R-M) systems are the methylase. These enzymes co-exist with restriction endonucleases and they provide the means by which bacteria are able to protect their own DNA and distinguish it from foreign DNA.

Modification methylases recognize and bind to the same recognition sequence as the corresponding restriction endonuclease, but instead of cleaving the DNA, they chemically modify one particular nucleotide within the sequence by the addition of a methyl group (C5 methyl cytosine, N4 methyl cytosine, or N6 methyl adenine).

Following methylation, the recognition sequence is no longer cleaved by the cognate restriction endonuclease.

The DNA of a bacterial cell is always fully modified by the activity of its modification methylase. It is therefore completely insensitive to the presence of the endogenous restriction endonuclease. Only unmodified, and therefore identifiably foreign DNA, is sensitive to restriction endonuclease recognition and cleavage.

During and after DNA replication, usually the hemi- methylated DNA (DNA methylated on one strand) is also resistant to the cognate restriction digestion.

With the advancement of recombinant DNA technology, it is now possible to clone genes and overproduce the enzymes in large quantities. The key to isolating clones of restriction endonuclease genes is to develop an efficient method to identify such clones within genomic DNA libraries, i. e. populations of clones derived by shotgun'procedures, when they occur at frequencies as low as 10-3 to 10-4. Preferably, the method should be selective, such that the unwanted clones with non-methylase inserts are destroyed while the desirable rare clones survive.

A large number of type II restriction-modification systems have been cloned. The first cloning method used bacteriophage infection as a means of identifying or selecting restriction endonuclease clones (EcoRII: Kosykh et al. , Mol. Gen. Genet. 178: 717-719, (1980) ; HhaII: Mann et al., Gene 3: 97-112, (1978) ; PstI : Walder et al. , Proc. Nat. Acad. Sci. 78: 1503-1507, (1981)).

Since the expression of restriction-modification systems in bacteria enables them to resist infection by bacteriophages, cells that carry cloned restriction- modification genes can, in principle, be selectively isolated as survivors from genomic DNA libraries that have been exposed to phage. However, this method has

been found to have only a limited success rate.

Specifically, it has been found that cloned restriction- modification genes do not always confer sufficient phage resistance to achieve selective survival.

Another cloning approach involves transferring systems initially characterized as plasmid-borne into E. coli cloning vectors (EcoRV: Bougueleret et al., Nucl.

Acids. Res. 12: 3659-3676, (1984) ; PaeR7: Gingeras and Brooks, Proc. Natl. Acad. Sci. USA 80: 402-406, (1983) ; Theriault and Roy, Gene 19: 355-359 (1982) ; PvuII : Blumenthal et al. , J. Bacteriol. 164: 501-509, (1985) ; Tsp45I : Wayne et al. Gene 202: 83-88, (1997)).

A third approach is to select for active expression of methylase genes (methylase selection) (U. S. Patent No. 5,200, 333 and BsuRI: Kiss et al. , Nucl. Acids. Res.

13: 6403-6421, (1985) ). Since restriction-modification genes are often closely linked together, both genes can often be cloned simultaneously. This selection does not always yield a complete restriction system however, but instead yields only the methylase gene (BspRI: Szomolanyi et al., Gene 10: 219-225, (1980) ; BcnI : Janulaitis et al., Gene 20: 197-204 (1982) ; BsuRI: Kiss and Baldauf, Gene 21: 111-119, (1983) ; and MspI : Walder et al. , J. Biol. Chem. 258: 1235-1241, (1983)).

A more recent method, the"endo-blue"method, has been described for direct cloning of thermostable restriction endonuclease genes into E. coli based on the indicator strain of E. coli containing the dinD :: lacZ fusion (Fomenkov et al. , U. S. Patent No. 5,498, 535; Fomenkov et al. , Nucl. Acids Res. 22: 2399-2403, (1994)).

This method utilizes the E. coli SOS response signals following DNA damage caused by restriction endonucleases

or non-specific nucleases. A number of thermostable nuclease genes (TaqI, TthlllI, BsoBI, Tf nuclease) have been cloned by this method (U. S. Patent No. 5,498, 535, 1996). The disadvantage of this method is that some positive blue clones containing a restriction endonuclease gene are difficult to culture due to the lack of the cognate methylase gene.

There are three major groups of DNA methylases based on the position and the base that is modified (C5 cytosine methylases, N4 cytosine methylases, and N6 adenine methylases). N4 cytosine and N6 adenine methylases are amino-methyltransferases (Malone et al.

J. Mol. Biol. 253: 618-632, (1995) ). When a restriction site on DNA is modified (methylated) by the methylase, it is resistant to digestion by the cognate restriction endonuclease. Sometimes methylation by a non-cognate methylase can also confer the DNA site resistant to restriction digestion. For example, Dcm methylase modification of 5'CCWGG 3' (SEQ ID NO : 7) (W = A or T) can also make the DNA resistant to PspGI restriction digestion. Another example is that CpM methylase can modify the C in CG dinucloetide and make the NotI site (5'GCGGCCGC 3' (SEQ ID NO : 8) ) refractory to NotI digestion (New England Biolabs'Catalog, 2000-01, page 220). Therefore methylases can be used as a tool to modify certain DNA sequences and make them uncleavable by restriction enzymes.

Because purified restriction endonucleases and modification methylases are useful tools for creating recombinant molecules in the laboratory, there is a strong commercial interest to obtain bacterial strains through recombinant DNA techniques that produce large quantities of restriction enzymes. Such over-expression

strains should also simplify the task of enzyme purification.

SUMMARY OF THE INVENTION The present invention relates to isolated DNA coding for the TspRI restriction endonuclease as well as to a method for cloning the TspRI restriction gene, tspRIR, from Thermus species R into E. coli by direct PCR from genomic DNA using degenerate primers based on the N-terminus and internal amino acid sequences.

It proved extremely difficult to clone TspRI endonuclease gene by conventional method. At first, a Sau3AI partial genomic DNA library was constructed.

After TspRI digestion of the plasmids in the library, methylase positive clones were identified among the surviving transformants. The entire tspRIM gene was sequenced and adjacent DNA sequences beyond tspRIM gene were derived by inverse PCR. Four open reading frames (ORF1-ORF4) were found upstream and one ORF (ORF5) was found downstream. These ORFs were expressed in M. TspRI pre-modified host, but no TspRI activity was detected in cell extracts prepared from the clones with inserts of ORF1-ORF4 or ORF5.

Since methylase selection and inverse PCR cloning did not yield any TspRI positive clones, another cloning method, the"endo-blue"method was used to screen clones containing nuclease genes. More than 40 blue colonies were found from the Sau3AI partial library using the dinD :: lack indicator strain. However, no apparent TspRI endonuclease activity was detected in the cell extracts of blue colonies.

To find out whether TspRI methylase is a multi- specific methylase, the plasmid pUC-TspRIM was digested with many restriction enzymes that would be blocked by C5 methylation of their cognate C5 methylases. The plasmid pUC-TspRIM can be cleaved by all the restriction enzymes tested except TspRI endonuclease, suggesting that TspRI methylase is not a multi-specific methylase.

In order to obtain the N-terminus and internal amino acid sequences, major efforts were made to purify the native TspRI endonuclease to homogeneity. The successful cloning strategy was to design degenerate primers based on the N-terminus and internal amino acid sequences and to amplify TspRI coding sequence directly from genomic DNA by PCR. TspRI endonuclease was purified from the native strain Thermus cell extract by chromatography through Heparin hyper D, Source 15Q, Heparin tsk gel, Source 15S, Heparin tsk columns, and gel filtration column Sephadex 75. The purified homogeneous TspRI protein has an apparent molecular mass of 58 kDa, which was subjected to sequential degradation to obtain the N-terminus amino acid sequence. TspRI protein was also digested partially with CNBr, resulting three peptides with apparent molecular mass 6 kDa, 14 kDa, and 26 kDa. They were electro-blotted and sequenced to obtain the internal amino acid sequence of TspRI protein. Degenerate primers were made and a-260 bp PCR product was found in a PCR reaction using a forward primer (designed from TspRI N-terminus amino acid sequence) and a reverse primer (designed from the internal 6 kDa amino acid sequence). The PCR product was cloned, sequenced and proved to be the bona fide N- terminal TspRI coding sequence. The C-terminus coding sequence of TspRI was identified from the partial ORF (355 bp) downstream of the putative T-G mismatch repair

gene in that the predicted amino acid sequence matches the actual amino acid sequence of the CNBr-derived 14 kDa peptide. The entire tspRIR gene was amplified by PCR and ligated to a T7 expression vector pET21at and transformed into pre-modified expression host ER2566 [pACYC-TspRIM]. However, no desired insert was detected among the ApR CmR transformants. Therefore, the tspRIR gene wan cloned and expressed in a low-copy-number T7 expression vector pACYC-T7ter. After clones with inserts were identified, the recombinant TspRI activity in cell extracts was detected by digestion of X DNA. Both the tspRIR PCR product and the insert in pACYC-T7ter were sequenced and confirmed to encode the wild type amino acid sequence.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1. TspRI restriction and modification system and adjacent open reading frames. Open reading frames 1-5 are shown as ORF1-ORF5. tspRIR, TspRI restriction endonuclease gene; tspRIM, TspRI methylase gene. ORF1 overlaps with tspRIM gene, ORF2,3, and 4 are located upstream of tspRIM gene. ORF5, tmr gene, encoding the putative T-G mismatch repair protein.

Figure 2. TspRI methylase gene sequence (tspRIM, 1296 bp (SEQ ID NO : 9) ) and the encoded amino acid sequence (SEQ ID NO : 10).

Figure 3. TspRI endonuclease gene sequence (tspRIR, 1566 bp (SEQ ID NO : 11)) and the encoded amino acid sequence (SEQ ID NO : 12).

Figure 4. T-G mismatch repair gene sequence (tmr gene, 678 bp (SEQ ID NO : 13) ) and the encoded amino acid sequence (SEQ ID NO : 14).

Figure 5. SDS-PAGE analysis of the purified native TspRI endonuclease protein. Lane 1, protein size marker, lanes 2 to 8, purified TspRI protein.

Figure 6. Protein expression profiles of TspRI- producing clones on SDS-PAG gel. Lane 1, protein size marker; lanes 2 and 4, non-induced cell extract; lanes 3 and 5, IPTG-induced cell extract.

Figure 7. Recombinant TspRI endonuclease activity in cell extract. Lanes 1 to 5, X DNA incubated with diluted cell extract containing recombinant TspRI. The dilution factors in lanes 1 to 5 were : 1/10,1/30, 1/50, 1/100,1/1000. Lane 6, X DNA ; lane 7, X DNA digested with native TspRI.

DETAILED DESCRIPTION OF THE INVENTION The tspRIM gene was cloned by methylase selection from a Sau3AI partial genomic DNA library. However, cloning of tspRIR gene proved to be extremely difficult using conventional methods such as methylase selection, "endo-blue"or inverse PCR of adjacent DNA beyond the methylase gene.

Since R-M genes in a particular R-M system are usually located in close proximity to each other, initial efforts were made to clone the adjacent DNA sequences by inverse PCR.

In the first round of inverse PCR walking toward the upstream, an EagI PCR fragment was sequenced, generating-700 bp of new sequence. In a second round of inverse PCR, an ApaI fragment PCR product was sequenced, giving rise to-370 bp of new sequence upstream. A total of 1433 bp DNA was found before the methylase gene start codon, with-1070 bp sequence derived by inverse PCR.

Four ORFs were found in a segment of 1892 bp, part of which overlaps with tspRIM gene (see Figure 1). The predicted amino acid sequences from ORF3 and ORF4 indicated that they have low homology to DNA metabolic enzymes (DNaseI, Integrase, and recombinase).

In order to express the four ORFs together in the same cell, ORF1 to ORF4 were amplified in PCR. Following restriction digestion with NdeI and HindIII, the PCR DNA was ligated to pET21a with compatible ends. The ligated DNA was transformed into pre-modified host ER2566 [pACYC-TspRIM]. Clones with PCR inserts were found and cell extracts prepared and assayed for TspRI activity.

No restriction activity was detected. It was concluded that ORF1 to ORF4 do not encode TspRI endonuclease.

Since the upstream sequences (ORF1 to ORF4) did not yield any TspRI activity, efforts were directed to clone DNA sequence downstream of the M gene. PCR products were found in AatII, BsaHI, Mspl, and RsaI templates. The PCR products were gel-purified and sequenced, generating 516 bp of new sequence downstream. One complete ORF (ORF5) and a partial ORF (ORF6,-355 bp) were found. The predicted amino acid sequence from ORF5 has high homology to T-G mismatch DNA repair protein and endonuclease III. ORF5 was renamed as tmr gene (T-G mismatch repair), which is 675 bp, encoding a 225-amino

acid protein with predicted molecular mass of 26 kDa.

ORF5 was amplified by PCR and cloned into a low-copy- number T7 expression vector pACYC-T7ter and transformed into pre-modified host ER2566 [pBR-TspRIM]. Cell extracts were prepared and assayed for TspRI endonuclease activity. However, no apparent TspRI activity was detected. It was concluded that ORF5 (tmr gene) is not TspRI endonuclease gene.

Since methylase selection and inverse PCR cloning of DNA beyond M gene did not yield any positive results, attempt was made to use the"endo-bluen method to clone nuclease genes from TspRI genomic DNA Sau3AI partial genomic DNA was ligated to BamHI- digested and CIP treated pUCl9 and the ligated DNA was used to transform E. coli indicator strain ER1992 (dinD: : lacZ). The DNA damage inducible promoter was fused with lacZ and any DNA damage on the host genomic DNA will induce SOS response and also increase ß-galactosidase expression. The colony turns blue on X- gal indicator plate if the colony contains a plasmid with nuclease gene insert. The nuclease gene can be non- specific endo/exo nuclease or type II restriction endonucleases. Forty blue colonies were identified and individual blue colony was amplified in 10 ml onvernight cultures and cell extracts prepared and assayed for restriction activity. No site-specific restriction activity was detected among 40 cell extracts.

There have been reports of multi-specific phage methylases that modify many restriction sites. For example, the BssHII phage methylase was shown to harbor at least five specificities (Schuman J. et al, (1996), J. Mol. Biol. , 257: 949-959). The phage-encoded

methylases are usually orphan methylases that no endonucleases have been found next to them. To find out whether TspRI methylase is a multi-specific methylase, the plasmid pUC-TspRIM was digested with many restriction enzymes that would be blocked by C5 methylation of their cognate C5 methylases. However, the plasmid pUC-TspRIM can be cleaved by all the restriction enzymes tested except TspRI endonuclease, indicating that TspRI methylase is not a multi-specific methylase.

The successful cloning strategy involved sequencing the native TspRI endonuclease protein to obtain the N- terminus and internal amino acid sequences. Degenerate PCR primers were used to amplify the coding sequence directly.

The preferred method described herein by which the TspRI methylase gene and the TspRI restriction endonuclease genes were cloned and expressed in E. coli included the following steps: 1. Genomic DNA preparation and genomic library construction Genomic DNA was prepared from Thermus species R cells and digested partially with Sau3AI. The Sau3AI fragments were cloned into pUC19 vector with multiple TspRI sites. Approximately 30,000 ApR transformants were amplified in 1 L culture and plasmid DNA prepared.

2. Cloning of tspRIM gene by methylase selection Varying amount of the plasmid library DNA was challenged with TspRI and the challenged DNA was transferred into ER2502 and ER2688 competent cells.

Plasmid DNA was prepared from survivor transformants and

screened for resistance to TspRI digestion. Two resistant clones were identified and the insert from one clone was completely sequenced. The sequence revealed that the TspRI methylase is a C5 methylase that shows extensive homology to other C5 methylases.

3. Subcloning of tspRIM gene in pACYC184 and pBR322 In order to express tspRIM gene in E. coli, the tspRIM gene was amplified from the genomic DNA by PCR.

The PCR DNA was purified, digested with SphI and SalI and ligated to pACYC184 or pBR322 with compatible ends.

Plasmid pACYC-TspRIM showed partial resistance to TspRI digestion whereas plasmid pBR-TspRIM showed full resistance. The pre-modified hosts ER2566 [pACYC-TspRIM] and ER2566 [pBR-TspRIM] were used for expression of TspRI endonuclease.

4. Purification of TspRI endonuclease from the native cells TspRI endonuclease was purified to homogeneity from the native strain Thermus cell extract by chromatography through Heparin hyper D, Source 15Q, Heparin tsk gel, Source 15S, Heparin tsk columns, and gel filtration column Sephadex 75. The purified homogeneous TspRI protein has an apparent molecular mass of 58 kDa.

5. Protein sequencing of TspRI protein and direct PCR of tspRIR coding sequence The purified proteins were subjected to electrophoresis and electro-blotted to a membrane (Matsudaira, P., J. Biol. Chem. , 262: 10035-10038, (1987).

Waite-Rees, P. A. et al. J. Bacteriology, 173: 5207-5219, (1991) ). The membrane was then stained with Commassie

blue R-250 and the 58 kDa bands was excised and subjected to sequential degradation in an automated Precise 494 Protein/Peptide Sequencer (Applied Biosystems). A forward degenerate primer was designed from the N-terminus 8 amino acid residues (MKRSEIEE) An additional sample of the TspRI endonuclease was treated with cyanogen bromide (CNBr). The partially digested peptides were subjected to electrophoresis and electro-blotted. The three major peptide bands 6 kDa, 14 kDa, and 26 kDa were cut out and subjected to sequential degradation.

The 6 kDa peptide contained the following amino acid sequence (SEQ ID NO : 15): KGDFLFFFQADPQDPELGSRRGIRGVYTVKG.

The amino acid sequence FFFQADPQDP (SEQ ID NO : 16) was used to design PCR reverse primers. A-260 bp PCR fragment was amplified in the PCR reaction which was blunted and ligated to HincII or SmaI digested and CIP treated pUC19. Clones with inserts were identified and sequenced. The DNA coding sequence was derived and translated into amino acid sequence which was compared to the actual amino acid sequence obtained by protein sequencing of the native purified TspRI protein. Among the N-terminus 33 amino acid residues, only two discrepancies were found. The inserts from multiple clones were sequenced and the accuracy of DNA and amino acid sequences were further confirmed.

6. Cloning of the entire tspRIR gene by PCR There was a 355-bp partial ORF downstream of the T- G mismatch repair gene. The predicted amino acid sequence derived from the partial ORF matches perfectly with the amino acid sequence derived from the internal 14 kDa peptide of TspRI protein. It was concluded that the partial ORF was part of tspRIR gene, encoding the C- terminal part of the protein. Apparently, the TspRI R-M genes are not immediately next to each other. Instead, the R-M genes were separated from each other by a putative T-G mismatch repair gene.

7. Expression of tspRIR gene in E. coli So far, the N-terminal part coding sequence (-260 bp) and the C-terminal part coding sequence (355 bp) had been sequenced. In order to obtain the remaining coding sequence and to construct a stable expression clone, the strain ER2566 [pBR-TspRIM] was used as the expression host. It was difficult to express tspRIR gene in pET21at vector in the host ER2566 [pACYC-TspRIM], probably due to under-methylation of TspRI sites.

The tspRIR gene was amplified by PCR and the PCR product was completely sequenced to obtain the wild type reference sequence. It was also digested with NdeI and ligated to NdeI-cut and CIP treated pACYC-t7ter. The ligated DNA was used to transform ER2566 [pBR-TspRIM], selecting ApR and CmR colonies. Postive clones with the correct size insert and orientation were found. IPTG- induced cell extracts were prepared after 3 h of IPTG induction of late log phase 10 ml cell cultures. The cell extracts were assayed for TspRI activity on ADNA.

Four cell extracts display recombinant TspRI activity,

with #6 and #17 displaying highest TspRI activity. The proteins in uninduced and IPTG-induced cell extracts were analyzed on SDS-PAGE and an induced protein band of approximately 58 kDa was detected in the IPTG-induced cell extract, but absent in the non-induced cell extract. The cell extracts were also heat-treated at 65°C and 75°C for 30 min and denatured proteins were removed by centrifugation at room temperature for 15 min. The clarified supernatant was then assayed for TspRI activity. Both samples displayed high TspRI activity at 65°C, indicating that like the native enzyme, the recombinant TspRI is also thermostable at 65°C and 75°C.

The plasmid DNA pACYC-T7ter-TspRIR clone #17 was prepared by Qiagen tip-20 column and the entire insert was sequenced. It was found that the insert contained the wild type sequence except one base silent mutation that still encodes the wild type amino acid.

The present invention is further illustrated by the following Example. This Example is provided to aid in the understanding of the invention and is not construed as a limitation thereof.

The references cited above and below are herein incorporated by reference.

EXAMPLE 1 Cloning of TspRI restriction-modification system in E. coli 1. Preparation of genomic DNA Genomic DNA was prepared from 4 g of Thermus species R (New England Biolabs collection) by the standard procedure consisting of the following steps: (a) Cell lysis by addition of lysozyme (2 mg/ml final), sucrose (1% final), and 50 mM Tris-HCl, pH 8.0.

(b) Further cell lysis by addition of SDS at a final concentration of 0. 1%.

(c) Further cell lysis by addition of 1% Triton X-100,62 mM EDTA, 50 mM Tris-HCl, pH 8.0.

(d) Removal of proteins by phenol-CHC13 extraction of DNA 3 times (equal volume) and CHC13 extraction once.

(e) DNA dialysis in 4 liters of TE buffer, change 3 times.

(f) RNase A treatment to remove RNA, genomic DNA precipitation in 95% ethanol, centrifuged, washed, dried and resupended in TE buffer.

2. Restriction digestion of genomic DNA and construction of genomic DNA library Restriction enzyme Sau3AI was diluted by 2-fold serial dilutions (4,2, 1,0. 5 units). Ten g Thermus genomic DNA were digested partially with Sau3AI and partial digestion was achieved by 1 and 0.5 units of

Sau3AI. The Sau3AI genomic fragments were ligated to BamHI digested and CIP treated pUC19 vector that contains multiple TspRI sites. The ligated DNA was used to transform E. coli RR1 competent cells (ER2502, DnaseI-) by electroporation. Approximately 30,000 ApR transformants were obtained for the Sau3AI library. All the colonies were pooled and amplified in 1 liter LB + Ap overnight. Plasmid DNA was prepared by Qiagen Maxi- prep columns.

3. Cloning of TspRI methylase gene by methylase selection Varying amount of the plasmid library DNA (0.25 pg, 0.5 pg, 1 Ag) was challenged with TspRI for 2 h at 65°C.

The TspRI-digested DNA was used to transform ER2502 and ER2688 competent cells. Plasmid DNA was prepared from 1.5 ml cell cultures inoculated from the transformants and screened for resistance to TspRI digestion. Eighteen plasmid mini-preparations were made and two plasmids were found to be resistant to TspRI digestion. The insert of two clones (&num 1 and #8) showed identical restriction patterns and the insert from &num 1 was completely sequenced using pUC universal primers, custom-made primers, and GPS insertion primers (NEB).

The TspRI methylase is a C5 methylase that shows extensive homology to other C5 methylases. M. TspRI presumably methylates the C5 position of cytosine in the recognition sequence 5'NNCASTGNN 3' (S = C OR G) (SEQ ID NO : 1) on hemi-methylated or non-methylated DNA.

4. Subcloning of tspRIM gene in pACYC184 and pBR322 In order to express tspRIM gene in E. coli, two primers were made with the following sequences:

5'tcagcagcatgcggaggtttaaaaatgtgtcccgcaagcgcctggagg 3' (202-51, underlined bases = SphI site) (SEQ ID NO : 17) 5'cgacgagtcgactcagaagagtgccacgcgagtaac 3' (202-40, underlined bases = SalI site) (SEQ ID NO : 18) The tspRIM gene was amplified from the genomic DNA using primers 202-51 and 202-40 under PCR condition of 95°C 1 min, 60°C 1 min, 72°C 2 min for 20 cycles. The PCR DNA was purified by phenol-CH3Cl extraction and CH3C1 extraction, precipitated with ethanol, dried, and resuspended in TE buffer. Following restriction digestion with SphI and SalI and purification through Qiagen spin column, the PCR DNA was ligated to pACYC184 and pBR322 with compatible ends. After screening 18 CmR transformants in pACYC, 15 clones contained tspRIM gene insert. However, all 15 pACYC-TspRIM plasmids showed partial resistance to TspRI digestion following 1.5 h incubation with TspRI endonuclease at 65°C. Eighteen ApR transformants in pBR322 were screened for inserts and their resistance to TspRI digestion. Three plasmids pBR- TspRIM (#4, #10, #11) showed full resistance to TspRI digestion. The premodified hosts ER2566 [pACYC-TspRIM] and ER2566 [pBR-TspRIM] were used for expression of TspRI endonuclease (see section 13, Expression of tspRIR gene in E. coli).

5. Cloning of DNA upstream of TspRI methylase Since R-M genes in a particular R-M system are usually located in close proximity to each other, efforts were made to clone the upstream sequence.

Two primers were synthesized with the following sequences: 5'cggcccagcgggccctgcaccagt 3' (199-48) (SEQ ID NO : 19) 5'gaggaccaccacccgctcctttcc 3' (199-49) (SEQ ID NO : 20) The genomic DNA was digested with AfiIII, AluI, ApoI, AvrII, BamHI, BsaWI, BssHI, BstEII, EagI, HaeII, HhaI, HincII, NcoI, NlaIII, NsiI, NspI, PstI, PvuII, SacII, StyI, TfiI, and TseI, respectively. The digested DNA was purified through Qiagen spin columns. Self- ligation was set up at a low DNA concentration at 2 Zg/ml overnight at 16°C. T4 DNA ligase was inactivated at 65°C for 30 min and the circular DNA was precipitated by ethanol. Five to 10 Rl of the ligated products were used as the templates for inverse PCR. PCR conditions were 95°C 30 sec, 60°C 30 sec, 72°C 2 min for 30 cycles.

PCR products were found in EagI, HaeII, HhaI, NlaIII, PstI, and TfiI templates. The PCR DNA products were purified from a low-melting agarose gel and sequenced with primers 199-48 and 49. The entire Eagl fragment was sequenced, generating-700 bp of new sequence upstream.

A second set of inverse PCR primers were synthesized with the following sequences: 5'cgaatcttttgcgaatgctatact 3' (SEQ ID NO : 21) 5'gagggaagcccagaccgaggaaga 3' (SEQ ID NO : 22) The genomic DNA was digested with ApaI, BsrFI, HaeII, HhaI, SpnI, NlaIII, SacI, Sau3AI, TaXI, and XhoI,

respectively. The digested DNA was purified through Qiagen spin columns. Self-ligation was set up at a low DNA concentration at 2 Ag/ml overnight at 16°C. T4 DNA ligase was inactivated at 65°C for 30 min and the circular DNA was precipitated by ethanol. Five to 10 all of the ligated products were used as the template for inverse PCR. PCR conditions were 95°C 30 sec, 55°C 1 min, 72°C 2 min for 30 cycles. PCR products were found in ApaI and Sau3AI templates. The PCR DNA products were purified from a low-melting agarose gel and sequenced with primers 205-54 and 55. The entire ApaI fragment was sequenced, generating-370 bp of new sequence upstream.

A total of 1433 bp of DNA sequence was found upstream of the M gene in which four ORFs were found in a segment of 1892 bp, part of which overlaps with tspRIM gene. The predicted amino acid from ORF1 has less than 20% homology to a transcription factor. The predicted amino acid sequence from ORF2 shows 33% amino acid sequence identity to a collagen protein. ORF3 amino acid sequence demonstrated low homology to a human herpevirus DNase. ORF4 amino acid sequence has high homology to DNA integrase/invertase/recombinase.

6. Expression of upstream sequence (ORF1 to ORF4) in E. coli In order to express these four ORFs (ORF1 to ORF4) together in the same cell, two PCR primers were made with the following sequences: 5'ctcatcattcatatgtctggtggtcaaggaaaagccgtg 3' (205-153, underlined bases = NdeI site) (SEQ ID NO : 23)

5'gcttgggccaagcttttgatggtcagcaggagcttgcct 3' (206-119, underlined bases = HindIII site) (SEQ ID NO : 24) ORF1 to ORF4 were amplified in PCR using 205-153 and 206-119 under PCR condition of 95°C 1 min, 60°C 1 min, 72°C 2 min for 20 cycles. The PCR DNA was purified by phenol-CH3C1 extraction and CH3C1 extraction, precipitated with ethanol, dried and resuspended in TE buffer. Following restriction digestion with NdeI and HindIII and purification through Qiagen spin column, the PCR DNA was ligated to pET21a with compatible ends. The ligated DNA was transformed into pre-modified host ER2566 [pACYC-TspRIM]. Clones with PCR inserts were found and ten cell extracts were prepared and assayed for TspRI endonuclease activity. The activity result was negative. It was concluded that ORF1 to ORF4 did not encode TspRI endonuclease.

7. Cloning of downstream sequence of tspRIM gene Since the upstream sequences (ORF1 to ORF4) did not yield any TspRI endonuclease activity, efforts were directed to clone DNA sequence downstream of the M gene.

Inverse PCR primers were made with the following sequences: 5'gtgtcccctttgtcccgtgcgtg 3' (200-42) (SEQ ID NO : 25) 5'ctaggtttggtcgggtgacaaatt 3' (200-43) (SEQ ID NO : 26) The genomic DNA was digested with AatII, ApoI, BsaHI, DraI, EaeI, EcoRI, Fspl, HaeIII, MspI, RsaI, and

Tsp509I, respectively. The digested DNA was purified through Qiagen spin columns. Self-ligation was set up at a low DNA concentration at 2 pg/ml overnight at 16°C.

T4 DNA ligase was inactivated at 65°C for 30 min and the circular DNA was precipitated by ethanol. Five to 10 1 of the ligated products were used as the template for inverse PCR. PCR conditions were 95°C 30 sec, 55°C 1 min, 72°C 2 min for 30 cycles. PCR products were found in AatII, BsaHI, MspI, and RsaI templates. The PCR DNA products were purified from a low-melting agarose gel and sequenced with primers 200-42 and 43, generating 516 bp of new sequence downstream. One complete ORF (ORF5) and a partial ORF (ORF6,-355 bp) downstream of tspRIM were found. The predicted amino acid sequence from ORF5 has high homology to T-G mismatch DNA repair protein and endonuclease III. ORF5 was renamed as tmr gene (T-G mismatch repair), which is 675 bp, encoding a 225-amino acid protein with predicted molecular mass of 26 kDa.

ORF5 was amplified by PCR and cloned into a low-copy- number T7 expression vector pACYC-T7ter and transformed into ER2566 and pre-modified host ER2566 [pBR-TspRIM].

Cell extracts were prepared and assayed for TspRI endonuclease activity. However, no apparent TspRI endonuclease activity was detected. It was concluded that ORF5 (tmr gene) is not TspRI endonuclease gene.

8. Attempt to use"endo-blue"method to clone nuclease gene from TspRI genomic DNA Sau3AI partial genomic DNA was ligated to BamHI- digested and CIP treated pUCl9 and the ligated DNA was used to transform E. coli indicator strain ER1992 (dinD :: lacZ). The DNA damage inducible promoter is fused with lacZ and any DNA damage on the host genomic DNA will induce SOS response and also increase ß-galactosidase expression. The colony turns blue on X-

gal indicator plate if the colony contains a plamid with nuclease gene insert. The nuclease gene can be non- specific endo/exo nuclease or type II restriction endonucleases. Forty blue colonies were identified from two transformation experiments. Individual blue colony was inoculated into 10 ml LB plus Ap and cultured overnight at 30°C in a shaker. Cells was harvested by centrifugation and resupended in a sonication buffer (50 mM Tris-HCl, pH 7.8, 10 mM P-mercaptoethanol) and lysed by sonication. Five l of Cell extracts was incubated with 1 Zg ADNA at 65°C for one h and the digested DNA was then analyzed by agarose gel electrophoresis. No site-specific restriction endonuclease activity was detected among 40 cell extracts.

9. TspRI methylase is a mono-specific methylase There have been reports of multi-specific phage methylases that modify many restriction sites. For example, the BssHII phage methylase was shown to harbor at least five specificities (Schuman J. et al, (1996), J. Mol. Biol. , 257: 949-959). The phage-encoded methylases are usually orphan methylases that no endonucleases have been found next to them. To find out whether TspRI methylase is a multi-specific methylase, the plasmid pUC-TspRIM was digested with many restriction enzymes that would be blocked by C5 methylation via their cognate C5 methylases. Plasmid pUC-TspRIM was digested with the following restriction enzymes: AciI, AclI, AluI, ApaI, ApaLI, AvaII, BanI, BbvI, BlpI, BsaAI, BsrFI, BssHII, EagI, HaeIII, HgaI, HhaI, HinPlI, HpaII, MspI, NspI, SacI, Sau96I, Sau3AI, TseI, and TspRI, respectively.

The digested DNA was then analyzed by agarose gel electrophoresis. The plasmid pUC-TspRIM can be cleaved by all the restriction enzymes tested except TspRI endonuclease, indicating that TspRI methylase is not a multi-specific methylase. This experiment demonstrated that TspRI methylase specificity does not overlap with the 24 enzymes tested here. Although it cannot be completely ruled out at this stage, TspRI methylase is not likely a prophage-encoded orphan multi-specific methylase.

10. Purification of TspRI endonuclease from the native cells Four hundred and ten grams of Thermus sp. R cells were resuspended in a 820 ml of SB buffer (20 mM KPO4, pH 6.9, 0.1 mM EDTA, 7 mM a-mercaptoethanol) plus 0.1 M NaCl, 5% glycerol. Cell lysis was completed by passing through a Gaulin Press four times. Complete cell lysis was achieved by monitoring the maximum level of proteins released into the buffer. The clarified supernatant was loaded into a Heparin hyper D column (392 ml). After extensive washing with low salt SB buffer, the proteins were eluted with a salt gradient of 0. 1-1 M NaCl in buffer SB plus 5% glycerol. Fractions were assayed for TspRI endonuclease activity on X DNA. Active fractions 40 to 72 (-800 ml) were pooled and dialyzed against a TESH buffer (20 mM Tris-HCl, pH 8.0, 0.1 mM EDTA, 7 mM P-mercaptoethanol) plus 50 mM NaCl and 5% glycerol.

The dialyzed pool was loaded into a Source 15Q column (70 ml). TspRI endonuclease activity was found in the flow-through and washing fractions. Nevertheless, some contaminating proteins bound to the column and were separated from TspRI endonuclease. The active fractions were pooled and loaded into a Heparin tsk gel AF column

(20 ml). After extensive washing with TESH buffer, proteins were eluted with a salt gradient of 50 mM to 1 M NaCl in TESH buffer plus 5% glycerol. TspRI endonuclease activity was identified in fractions 25-33, but only fractions 27 to 32 were pooled and diluted by the addition of SB buffer.

The proteins were loaded into a Resource 15S column. Some TspRI protein bound to the column and was eluted with a salt gradient of 50 mM to 1 M NaCl in SB buffer plus 5% glycerol. The eluted active fractions 11- 13 were pooled and dialyzed against a storage buffer (0.2 M NaCl, 20 mM Tris-HC1, pH 7.4, 0.1 mM EDTA, 1 mM DTT, 50% glycerol). This preparation was called purified TspRI"pool 1"which contained 1-1.2 million units of TspRI. During the chromatography using Resource 15S column, some TspRI activity was detected in the flow- through fractions. These fractions were pooled and loaded into a Heparin tsk column (AF HR 10/10) using SB buffer. After washing, TspRI protein was eluted with a salt gradient of 50 mM to 1 M NaCl in SB buffer plus 5% glycerol. The active fractions 27-29 were identified and pooled and dialyzed in a storage buffer. This preparation was called TspRI"pool 2"which contained about 0.9 million units of TspRI endonuclease. SDS-PAGE analysis of the proteins in"pool 2"indicated that there was a major protein band of 58 kDa TspRI protein.

However, there were still two minor contaminating proteins (-12 and-17 kDa, respectively). They were removed by chromatography through a Sephadex 75 column.

Proteins were eluted in TESH buffer plus 0.5 M NaCl and 5% glycerol. Fraction 14-24 contained the homogeneous TspRI protein (Figure 5). TspRI protein in fraction 18 was used for protein sequencing.

11. Amino acid sequence analysis of TspRI endonuclease The purified proteins were subjected to electrophoresis and electro-blotted to a membrane (Matsudaira P., J. Biol. Chem. , 262: 10035-10038, (1987).

Waite-Rees P. A. et al. , J. Bacteriology, 173: 5207-5219, (1991). The membrane was then stained with Commassie blue R-250 and the 58 kDa bands was excised and subjected to sequential degradation in an automated Precise 494 Protein/Peptide Sequrncer (Applied Biosystems). The 58 kDa protein contained the following N-terminal amino acid sequence: MKRSEIEELLEIFRxSLLSIPSGdF (x) RRVHQFT (x = unknown amino acid, d = erroneous calling, (x) = omitted amino acid calling) (SEQ ID NO : 27) Arg and Ser codons are very degenerate. To reduce the degeneracy, four forward PCR primers were made based on the amino acid sequence MKRSEIEE (SEQ ID NO : 28).

Primer 5'cgcggatccatgaaragrtcngaratcgarga 3' (234-276) (SEQ ID NO : 29) Protein M K R S E I E E (SEQ ID NO: Primer 5' cgcggatccatgaarcgg#ctcngaratcgarga 3' (234-277) (SEQ ID NO : 30) Primer 5'cgcggatccatgaaragragygaratcgarga 3' (234-278) (SEQ ID NO : 31) Primer 5'cgcqgatccatgaarcggscagygaratcgarga 3' (234-279) (SEQ ID NO : 32)

The reverse PCR primer was made based on the following amino acid sequence: PSGdFRR (SEQ ID NO : 33) 5'cgncgraartcnccrctngg 3' (237-95) (SEQ ID NO : 34) (The complementary sequence of 5'ccntcnggngayt tycgncg 3' (SEQ ID NO : 35)).

Four sets of PCR reactions were set up using primers 234-276 and 237-95; 234-277 and 237-95; 234-278 and 237-95 ; 234-279 and 237-95, respectively. PCR conditions were 95°C 5 min for 1 cycle, 95°C 1 min, 40°C 1 min, 72°C 30 sec for 35 cycles. PCR products between 72 to 86 bp were gel purified and sequenced using the forward and reverse primers. No readable sequence was obtained, partly because of possible mixed PCR products or random-amplified products. The failure to amplify the N-terminus 80 bp coding sequence was probably due to the reverse primer that based on the erroneous amino acid calling. Later (described in section 12) it was discovered that the amino acid sequence PSGdFRR (SEQ ID NO : 36) used for reverse primer design contained two mis- callings. The correct amino acid sequence should be: PSGPFARR (SEQ ID NO : 37).

12. Amino acid sequencing of CNBr-derived fragements of TspRI protein An additional sample of the TspRI endonuclease, 5 , ug in 20 gel, was treated with 1 jug of cyanogen bromide (CNBr, Sigma) dissolved in 200 ul of 88% distilled formic acid for 24 hours in the dark at room temperature. This reaction mixture was evaporated to dryness and the sample was resuspended in 100 ul of SDS- PAGE loading buffer and subjected to electrophoresis and western blotted to PVDF. The three major peptide bands 6

kDa, 14 kDa, and 26 kDa were cut out and subjected to sequential degradation.

The 6 kDa peptide contained the following amino acid sequence: KGDFLFFFQADPQDPELGSRRGIRGVYTVKG (SEQ ID NO : 38) The amino acid sequence FFFQADPQDP (SEQ ID NO : 39) was used to design reverse primers for PCR.

The 14 kDa peptide contained the following amino acid sequence : HLGNLVGQPGRLVEVHLTPVLVGARLVGRGQNRIHVLPRGYDRTVxYYN (x = unknown amino acid (SEQ ID NO : 40)) The 26 kDa peptide contained the following amino acid sequence: MGAGKGSSVRQLLPEEALGIYK (SEQ ID NO : 41) Two reverse PCR primers were made based on the 6 kDa amino acid sequence: 5'ggrtcytgnggrtcngcytg 3' (249-198) (SEQ ID NO : 42) 5'ggrtcngcytgraaraaraa 3' (249-199) (SEQ ID NO : 43) Four PCR reactions were set up using primers 234- 277 (F) and 249-198 (R); 234-277 (F) and 249-199 (R); 234-279 (F) and 249-198 (R); 234-279 (F) and 249-199 (R); (forward primer, R=reverse primer). A ~260 bp PCR fragment was found in the PCR reaction of 234-279 and 249-199 under PCR conditions of 95°C 5 min for 1 cycle,

95°C 1 min, 40°C 1 min, 72°C 2 min for 25 cycles. The PCR product was blunted by treatment with T4 polynucleotide kinase in the presence of ATP and ligated to HincII or SmaI digested and CIP treated pUC19. The ligated DNA was used to transform ER2688 and plated on LB agar, X-gal, Ap plates. After screening 18 white colonies, clones with PCR insert were identified and sequenced with pUC universal primers. The DNA coding sequence was derived and translated into amino acid sequence which was compared to the actual amino acid sequence obtained by protein sequencing of the native purified TspRI protein.

Among the N-terminus 33 amino acid residues, only two discrepancies were found. The inserts from multiple active TspRI clones were sequenced and the correct DNA and amino acid sequences were further confirmed. The correct N-terminus amino acid sequence of TspRI is as follows: MKRSEIEELLEIFRCSLLSIPSGPFARRVHQFT (SEQ ID NO : 44) (bold and underlined residues, different from the original amino acid sequence derived from the native protein) There was a 355-bp partial ORF downstream of the T- G mismatch repair gene. The predicted amino acid sequence derived from the partial ORF matches perfectly with the amino acid sequence derived from the internal 14 kDa peptide. It was concluded that the partial ORF was part of tspRIR gene, encoding the C-terminus part of the protein. Apparently, the TspRI R-M genes are not immediately next to each other. Instead, they were separated from each other by a T-G mismatch repair gene.

13. Expression of tspRIR gene in E. coli So far, the N-terminal part coding sequence (-260 bp) and the C-terminal part coding sequence (355 bp) had been sequenced. In order to obtain the remaining coding sequence, two PCR primers were made with following sequences: 5 cgcggatcccatatgaaacggagcgagatcgaggaacttctagaa 3' (250-102, underlined bases, Ndel site) (SEQ ID NO : 45) 5'tgggtcgacgagetcttaaaggagggggattcccatagagag 3' (250-287, underlined bases, Sacl site) (SEQ ID NO : 46) The tspRI. R gene was amplified by PCR using primers 250-102 and 250-287 under PCR conditions of 95°C 2 min for 1 cycle, 95°C 1 min, 60°C 1 min, 72°C 2 min for 20 cycles. PCR product was digested with NdeI and SacI and gel-purified from a low-melting agarose gel. It was ligated to the T7 expression vector pET21at with compatible ends and the ligated DNA was used to transform ER2566 [pACYC-TspRIM]. After screening 18 transformants, no correct size insert was found. The failure to find the positive clones was probably due to the under-methylation of TspRI sites on the choromosomal DNA because tspRIM gene was expressed from a low-copy- number plasmid. In order to construct a stable expression clone, the tspRIM gene was cloned in pBR322 first and the strain ER2566 [pBR-TspRIM] used as the pre-modified host. This cloning strategy proved to be successful. Two new primers were made for PCR of tspRIR gene:

5'tggccccaccatatgttaaaggagggggattcccatagagag 3' (253-90, underlined bases = NdeI site) (SEQ ID NO : 47) 5'cgcgtaggccatatgaaacggagcgagatcgaggaacttcta 3' (253-91, underlined bases = NdeI site) (SEQ ID NO : 48) The tspRIR gene was amplified by PCR using primers 253-90 and 253-91,2 units of Vent DNA polymerase under PCR conditions of 95°C 2 min for 1 cycle, 95°C 1 min, 60°C 1 min, 72°C 2 min for 22 cycles (3 to 5 mM MgSO4).

PCR product was digested with NdeI and ligated to NdeI- cut and CIP treated pACYC-t7ter. The ligation condition was 0.2 Rg pACYC-T7ter, 0.5 Rg PCR DNA (tspRIR gene), 3 1 lOx ligation buffer, 1200 units T4 DNA ligase, sdH2O to 30 Rl at 16°C overnight. The ligated DNA was used to transform ER2566 [pBR-TspRIM]. After screening 36 ApR CmR transformants, 7 positive clones with the correct size insert were found. IPTG-induced cell extracts were prepared after 3 h of IPTG induction of late log phase 10 ml cell cultures. The cell extracts were assayed for TspRI activity on X DNA. Four cell extracts display recombinant TspRI activity, with #6 and #17 displaying highest TspRI activity. The proteins in uninduced and IPTG-induced cell extracts were analyzed on SDS-PAGE and an induced protein band of approximately 58 kDa was detected in the IPTG-induced cell extract, but absent in the non-induced cell extract. The cell extracts were also heat-treated at 65°C and 75°C for 30 min and denatured proteins were removed by centrifugation at room temperature for 15 min. The clarified supernatant was then assayed for TspRI activity. Both samples displayed high TspRI activity at 65°C, indicating that

like the native enzyme, the recombinant TspRI is also thermostable at 65°C and 75°C.

The plasmid DNA pACYC-T7ter-TspRIR clone #17 was prepared by Qiagen column tip-20 and the entire insert was sequenced. It was found that the insert contained the wild type sequence except one base silent mutation that still encodes the wild type amino acid.

The strain NEB#1346, ER2566 [pBR-TspRIM, pACYC- T7ter-TspRI] has been deposited under the terms and conditions of the Budapest Treaty with the American Type Culture Collection on October 11,2001 and received ATCC Accession No. PTA-3779.