High-Level Expression, Purification, and Thermus Aquatlcus DNA

Home , Klenow fragment, Pfu DNA polymerase, Taq polymerase, Thermus aquaticus

Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

High-level Expression, Purification, and Enzymatic Characterization of Full-length Thermus aquatlcus DNA Polymerase and a Truncated Form Deficient in 5' to 3' Exonuclease Activity

Frances C. Lawyer, 1 Susanne Stoffel, 1 Randall K. Saiki, 2 Sheng-Yung Chang, 3 Phoebe A. Landre, ~ Richard D. Abrarnson, 1 and David H. Gelfand 1

1Program in Core Research and Departments of 2Human Genetics and 3Infectious Disease, Roche Molecular Systems, Alameda, California 94501

The Thermus aquaticus DNA poly- life of 9 min at 97.5~ The Stoffel I in the native host is quite low (0.01- merase I (Taq Pol I) gene was cloned fragment has a half-life of 21 min at 0.02% of total protein). The cloning and into a plasmid expression vector that 97.5~ Taq Pol I contains a polymer- expression of full-length 94-kD Taq Pol I utilizes the strong bacteriophage ization-dependent 5' to 3' exonu- in E. coli under control of the E. coli lac PL promoter. A truncated form of Taq clease activity whereas the Stoffel promoter r or the tac promoter (7~ has Pol I was also constructed. The two fragment, deleted for the 5' to 3' ex- been reported. Because polymerase constructs made it possible to com- onuclease domain, does not possess yields in these constructs were low pare the full-length 832-amino-acid that activity. A comparison is made (-0.01% of total protein in our initial Taq Pol I and a deletion derivative among thermostable DNA poly- construct; see ref. 5), we sought to im- encoding a 544-amino-acid transla- merases that have been character- prove the expression level of the enzyme tion product, the Stoffel fragment. ized; specific activities of 292,000 by mutagenizing the 5' and 3' ends of Upon heat induction, the 832-amino- units/mg for Taq Pol I and 369,000 the gene and cloning the mutagenized acid construct produced 1-2% of to- units/mg for the Stoffel fragment are gene into a more suitable expression vec- tal protein as Taq Pol I. The induced the highest reported. tor. We also constructed a truncated Taq 544-amino-acid construct produced Pol I gene, deleted for the first 867 bp of 3% of total protein as Stoffel frag- the gene, which yields a predicted 61-kD ment. Enzyme purification included translation product. The 61-kD deriva- cell lysis, heat treatment followed by tive, Taq Pol I Stoffel fragment, is active Polymin P precipitation of nucleic ac- The thermostable DNA polymerase I in polymerase assays and PCR and is de- ids, phenyl sepharose column chro- (Taq Pol I) from Thermus aquaticus (Taq) void of the inherent 5' to 3' exonuclease matography, and heparin-Sepharose greatly improves the yield, specificity, activity of Taq Pol I. A similarly trun- column chromatography. For full- automation, and utility of the poly- cated gene deleted for the first 705 bp, length 94-kD Taq Pol I, yield was merase chain reaction (PCR) ~l'z~ method has recently been constructed, yielding 3.26x 107 units of activity from 165 of amplifying DNA fragments. ~3~ Fur- an approximately 67-kD translation grams wet weight cell paste. For the thermore, because of the high turnover product, KlenTaq DNA polymerase. (8~ 61-kD Taq Pol I Stoffel fragment, the number, lack of proofreading activity, Protein produced by the two constructs yield was 1.03 • 106 units of activity high temperature optimum, and ability described here provides a plentiful sup- from 15.6 grams wet weight cell to incorporate 7-deaza-2-deoxyguano- ply of the two forms of Taq Pol I enzyme, paste. The two enzymes have maxi- sine efficiently, Taq Pol I yields long thereby enabling their biochemical mal activity at 75~ to 80~ 2-4 mM stretches of readable DNA sequence that properties to be investigated. The study MgClz and 10-55 mM KCI. The nature are uniform in intensity and free of of these enzymes not only provides key of the substrate determines the pre- background. ~4~ insights into nucleic acid metabolism, cise conditions for maximal enzyme A 94-kD Taq Pol I has been purified but also provides useful information to activity. For both proteins, MgCI2 is from Taq, but growing the organism is those investigators who exploit DNA the preferred cofactor compared to more difficult than growing Escherichia polymerases in specialized molecular bi- MnCI z, CoCIz, and NiCI 2. The full- coll. Although the activity yield is high ology techniques, including the PCR. length Taq Pol I has an activity half- (40--60%), the expression level of Taq Pol Advancements in PCR methodology

2:275-2879 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/93 $3.00 PCR Methods and Applications 275 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

depend in part on an increased under- (pH 7.2), 5 mM MgC12.(17~ Single- tion products were transformed into standing of the biochemistry of thermo- stranded template DNA was prepared by DG98. Nitrocellulose filters containing stable DNA polymerases, as such conventional methods r from M13 ampicillin-resistant transformants were information aids substantially in the phage TSYC657 (7.359 kb), a derivative hybridized according to the method of constant refinement of this and other re- of M13mpl0 containing a 129-nucle- Woods et al. (23) with 3zp-labeled DG26 search techniques. otide insert derived from a Streptornyces probe. Positive colonies were picked and limosus a-amylase gene. (~8~ The 24-mer screened by restriction analysis of alka- synthetic oligodeoxynucleotide primer line-SDS miniptasmid preparations. A MATERIALS AND METHODS SC64 (Table 1) is complementary to the representative correct candidate was des- Bacterial Strains and Plasmids S. limosus-derived insert in TSYC657 and ignated pLSG2 (Fig. 1). has a calculated melting temperature of For mutagenesis of pLSG2, 0.5 pmole E. coli strain DG98 (thi-1 endA1 hsdR17 95~ at 0.12 I~M in 100 mM NaCI. Primer of single-stranded pLSG2, 0.5 pmole of lacI Q IacZAM15 proC::TnlO supE44/ F'la- and template were annealed at $83 nM PvuII-digested, CIAP-treated pBS § and cIQ IacZAM15 proC +) has been de- SC64 and 256 nM TSYC657(2.2:l primer: 2.5 pmoles of oligonucleotide SC107 scribed. (9~ E. coli strain DGl16 [thi-1 template) in 10 mM Tris-HC1 (pH 8.0) 6 (Table 1) were annealed. Extension, endA1 hsdR17 supE44 (hcI857 bloT76 mM MgCl2, and 50 mM KCI. The mixture transformation, and screening were per- M-/1)] is a derivative of strain MM294 (1~ was heated at 95~ for 4 rain, incubated formed as described above, using 32p-la- containing a defective h prophage. at 70~ for 10 min, and cooled to room beled SC107 as the probe. A representa- DGl16 was prepared using bacterioph- temperature. tive correct candidate was designated age P1 transduction ~1~ and E. coli strain pSYC1578 (Fig. 1). N6590 [C600 rK- mK- thr leu pro IacZXA21::Tn10 (hcI857 bloT76 M-/1), Cloning Procedures Assembly of a Full-length Taq Pol I provided by M. Gottesman, Columbia For oligonucleotide site-directed mu- Gene in High-level Expression University] as the source of the defective tagenesis, single-stranded DNA was pre- Vectors prophage. Plasmid pBS § was purchased pared from pBS § plasmid derivatives from Stratagene. Plasmid pFC54.T has pLSG1 and pLSG2 as described. (s~ Com- An SphI and BglII-digest of pSYC1578 been described. (~z~ Plasmid pDG160, a petent cells were prepared and transfor- DNA was ligated with HindIII- and derivative of pFC54.T, contains a conve- mations were carried out by the method BamHI-digested vector pFC54.T and an nient restriction site polylinker between of Hanahan. (19~ Alkaline-SDS miniplas- annealed duplex oligonucleotide, DG271 the PL promoter and retroregulator cas- mid DNA preparations were carried out 28 (Table 1). The oligonucleotide duplex settes. by the method of Birnboim and Doly. (2~ provided HindIII and SphI cohesive ends Enzymatic dideoxynucleotide DNA se- and codons one, two, and two of three Reagents quence analysis r confirmed all DNA nucleotides for codon three of Taq Pol I. sequence alterations. The resulting plasmid was designated Restriction endonucleases [New England pLSG5 (Fig. 1). Biolabs (NEB)], E. coli DNA polymerase I, Oligonucleotide Site-directed large fragment (Klenow) (U.S. Biochem- Mutagenesis (22) Assembly of a Truncated Taq Pol I icals), T4 DNA ligase, (~3~ calf intestine al- Gene kaline phosphatase (CIAP) (Boehringer- For pLSG1 mutagenesis, 0.24 pmole of Mannheim), and polynucleotide kinase single-stranded pLSG1, 0.36 pmole of A BstXI digest of pSYC1578 was treated (kinase) (NEB) were used according to PvuII-digested CIAP-treated vector pBS +, with Klenow to generate blunt ends. The standard procedures. Oligonucleotides and 0.48 pmole of primer DG26 (Table DNA was then treated with BglII and a were phosphorylated as described. (~4~ 1) were annealed. The mixture was put 1619-bp BstXI-blunt-BgllI fragment (en- Molecular weight standards for SDS- on ice and TTP, dGTP, dATP, [3H]dCTP coding Taq Pol I codons 294-832) was polyacrylamide gel electrophoresis were each added to 200 p.M along with 1 isolated via gel electrophoresis and elec- (PAGE) were purchased from Pharma- unit of Klenow fragment. The extension troelution. The isolated DNA fragment cia [~/-3zp]ATP (3000 Ci/mmole), mixture was incubated at 0~ for 30 min was ligated to HindlII- and BamHI-di- [R-32p]dCTP (800 Ci/mmole), and and then at 30~ for 30 rain. The reac- gested vector pFC54.T and an annealed [3H]dCTP (22.8 Ci/mmole), were purchased from New England Nuclear. Oli- gonucleotide synthesis was performed TABLE 1 Oligonucleotides Used in These Studies on a Biosearch Model 8750 DNA synthe- sizer using controlled-pore glass and DG26 5'-CCCTTGGGCTCAAAAAGTGGAAGCATGCCTCTCATAGCTGTTTCCTG O-cyanoethyl-N, N-diisopropyl phos- DG27 5'-AGCTTATGAGAGGCATG phoramidites. (ls'16~ Reagents and syn- DG28 ATACTCTCC-5' thesis protocols were obtained from Mil- DG29 5'-AGCTTATGTCTCCNNAAGCT ligen/Biosearch (Novato, CA). DG30 ATACAGAGGTITTCGA-S' Crude salmon sperm DNA was ob- DG48 5'-GGGAAGGGCGATCGGTGCGGGCCTCTFCGC tained from Sigma (#D-1626, Type III) DG67 5'-CCCGGGCGGCGCCGCAGCGGCGGGp SC64 5'-CCCGGGCGGCGCCGCAGCGGCGGG and "activated" by incubation at 4~ for SC107 S'-GCATGGGGTGGTAGATCTCACTCCTFGGC 24 hr at 1.3 mg/ml in 10 mM Tris-HCl

276 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

b Plasmid Promotera Taq Poll

1 2 3 4 5 6 7 832 Met Arg Gly Met Leu Pro Leu Glu pLSG1 lac ...... AGCT ATG ACC-160bp-TAAC ATG AGG GGG ATG CTG CCC CTC---GAG TGA TAC (Taq Pol I) 1 2 3 4 5 6 7 832 Met Arg Gly Met Leu Pro Leu Glu pLSG2 lac ...... AGCT ATG AGA GGC ATG CTT CCA CTT---GAG TGA TAC (Taq Pol I) Sph l

1 2 3 4 5 6 7 832 pSYC1578 Met Arg Gly Met Leu Pro Leu Glu (Taq Pol I) lac ...... AGCT ATG AGA GGC ATG CTT CCA CTT---GAG TGA GATCTAC Sph I Bgi ii 1 2 3 4 5 6 7 832 pLSG5 Met Arg Gly Met Leu Pro Leu Glu P .... AAGCTT ATG AGA GGC ATG CTT CCA CTT---GAG TGA GATCC--ter L (Taq Poll Hind III Sph I Bgl II/ Batrfrll high expression construct) 1 2 3 4 5 6 7 544 (290) (291) (292) (293) (294) (295) (832) Met Ser Pro Lys Ala Leu Glu Glu pLSG8 PL .... AAGCTT ATG TCT CCA AAA GCT CTG GAG---GAG TGA GATCC--ter (Stoffel Hind III Bgl II/BamHI fragment) FIGURE 1 Taq Pol I expression constructions. Plasmids are described in the text. apromoter-ribosome binding sites are: lac, [3-galactosidase; PL, h PL promoter/gene N ribosome binding site. ~ Pol I codons are numbered above each sequence to show the amino-terminal codons and carboxy- terminal codons and how they were altered in the mutagenesis and recloning described in the text. Restriction endonuclease cleavage sites are underlined and labeled. The B. thuringiensis B-toxin retroregulator is indicated by ter. Numbers appearing in parentheses in pLSG8 refer to the corresponding codon numbering of the full-length gene.

duplex oligonucleotide DG29/DG30 PCR Using Heat-treated Cell Extracts were lysed in an Aminco French Pressure (which provides a Met start codon as Heat-treated extracts from 37~ 9-hr Cell (20,000 psi), sonicated to reduce vis- well as codons 290--293 of Taq Pol I; see heat inductions of cultures of pLSG5 cosity, and diluted to 935 ml with lx TE Table 1). The truncated gene assembled (Taq Pol I) and pLSG8 (Stoffel fragment) containing 2.4 mM PMSF and 0.5 lag/ml in HindIII/BamHI-digested vector were assayed for enzyme activity. The ex- leupeptin. This lysate (Fraction I; see Fig. pFC54.T was designated pLSG8 (Fig. 1). tracts were diluted to 4 units/~l in 1• 2A) contained 18.4 grams of protein and storage buffer (20 mM Tris-HC1, pH 7.5, 45,000,000 units of DNA polymerase ac- 100 mM KCI, 0.1 mM EDTA, 1 mM dithio- tivity. Inductions threitol, 0.2% Tween 20 [Pierce, "Sur- Ammonium sulfate was added to 0.2 Isopropyl-13-D-thiogalactopyranoside fact-Amps"], and 50% glycerol).The dilu- M and the lysate centrifuged at 25,000g (IPTG)-inductions of pLSG1 and pLSG2 tions were used in the PCR under for 20 min. The supernatant, Fraction II, were as described. (s~ For heat inductions standard conditions ~3~ at 4, 2, 1, 0.5, or was incubated at 75~ for 15 min to de- of plasmids pDG160, pLSG5, and pLSG8, 0.25 units per reaction. Reactions were nature E. coli host proteins, then cooled cultures were inoculated to OD6o O = subjected to 30 cycles of PCR amplifica- to 0~ and adjusted to 0.6% Polymin P, 0.05 from a fresh 30~ overnight culture tion using human globin gene primers, precipitating approximately 90% of the in Bonner-Vogel salts (24) plus 0.25% PC03 and PC04, and template DNA iso- Az6o absorbing material. The mixture casamino acids, 0.2% glucose, 10 p.g/ml lated from the human cell line Molt 4. C3) was stirred at 0~ for 3 hr and centri- thiamine, and 100 p.g/ml ampicillin. fuged at 25,000g or 30 min. The heat- Cultures were incubated at 30~ until inactivation/Polymin P step removed the OD6oo of the culture reached 0.7, Purification of Recombinant Taq 93% of the total soluble protein and then induced at 37~ or 41~ for 2.5-21 DNA Polymerase yielded a greater than 14-fold purifica- hr. For enzyme assays of small-scale in- All operations were carried out at 0-4~ tion (Fraction III, Fig. ZA). ductions, frozen pellets were thawed, unless otherwise specified. All glassware Fraction III was loaded onto a 2.2 x suspended in 50 mM Tris-HCl (pH 7.4), was baked and all solutions were auto- 11-cm (42 ml) phenyl-Sepharose CL-4B 10 mM EDTA, 0.5 p.g/ml leupeptin, 2.5 claved where possible. Induced pLSG5/ (Pharmacia-LKB, Lot #MI 02547) col- mM phenylmethylsulfonyl flouride DGl16 cells (165 grams wet cell weight) umn (equilibrated in TE containing 0.2 (PMSF), and sonicated. Heat-treated ex- were thawed in 165 ml of 2• TE buffer M ammonium sulfate) at 40 ml/hr. The tracts were prepared by adding (100 mM Tris-HC1, pH 7.5, 2 mM EDTA), column was washed with 400 ml of the (NH4)2SO 4 to 0.2 M, heating at 75~ for containing 2.4 mM PMSF and 0.5 i~g/ml same buffer (Azso to baseline), followed 20 min, chilling on ice, and clarifying by leupeptin, and homogenized at low by 180 ml of TE at 50 ml/hr, and finally centrifugation. speed in a blender. The thawed cells 105 ml of 20% (wt/vol) ethylene glycol

PCR Methods and Applications 277 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

between 275 mM and 325 mM KCI were pooled (98 ml total) and the pooled fractions concentrated/diafiltered on an Arn- icon YM30 membrane to a final volume of 20 ml in a 2.5 x storage buffer preparation lacking glycerol. Thirty milliliters of 80% glycerol were added to the concentrated and diafiltered sample forming Fraction V (Fig. 2A), which was stored at - 20~

Purification of 61-kD Taq Pol I, Stoffel Fragment Purification of the Stoffel fragment from induced pLSG8/DG116 cells proceeded as the purification of full-length Taq Pol I (above) with some modifications: In- duced pLSG8/DG116 cells (15.6 grams) were homogenized and lysed (Fraction I; see Fig. 2B), as above, yielding 1.87 grams of protein and 1,046,000 units of DNA polymerase activity. Following heat treatment (Fraction II, Fig. 2B), as above, Polymin P (pH 7.5) was added slowly to 0.7%. The heat-inactivation/Polymin P step removed 92% of the total soluble protein and yielded a greater than 14- fold purification. The supernatant, Frac- tion III (Fig. 2B), was loaded onto a 1.15 x 3.1-cm (3.2 ml) phenyl-Sepharose column at 10 ml/hr. All of the applied activity was retained on the column. The column was washed with 15 ml of the equilibration buffer and then with 5 ml of 100 mM KC1 in 50 mM Tris-HC1 (pH 7.5), 1 mM EDTA (TE). The polyrnerase activity was eluted with 2 M urea in TE containing 20% ethylene glycol. Frac- FIGURE 2 Purifications of full-length of Taq Pol I and Stoffel fragment. (A) Full-length Taq Pol I purification fractions analyzed via SDS-PAGE on a 10% running gel, 4.75% stacking gel, stained tions (0.5 ml each) with polymerase ac- with Coomassie brilliant blue. (Lane 1) Molecular weight standard, molecular weights (in thou- tivity were pooled (8.5 mt) and dialyzed sands) shown at left; (lane 2) 20 ~g of Fraction I; (lane 3) 20 t~g of Fraction II; (lane 4) 3 i~g of into heparin sepharose buffer contain- Fraction III; (lane 5) 3 I~g of early phenyl-Sepharose flow-through; (lane 6) 3 i~g of late phenyl- ing 100 mM KCI, forming Fraction IV Sepharose flowthrough; (lane 7) 3 i~g of phenyl-Sepharose low-salt eluate; (lane 8) 1 i~g of (Fig. 2B). This fraction contained 87% of Fraction IV; (lane 9) control from a previous purification; (lane I0) 1 I~g (230 units) of Fraction the applied activity and 3.6% of the ap- V. (B) Stoffel fragment purification fractions analyzed via SDS-PAGE on a 4-20% gradient gel and plied protein, yielding a 24-fold purifica- stained with silver (Integrated Separation Systems). (Lane I) Molecular weight standard, molec- tion. Fraction IV was loaded onto a 1.0- ular weights (in thousands) shown at left; (lane 2) 1.1 t~g of Fraction I; (lane 3) 0.9 i~g of Fraction ml bed volume heparin-Sepharose II; (lane 4) 2.4 p~g of Fraction III; (lane 5) 0.1 I~g of Fraction IV; (lane 6) 0.1 I~g (22.5 units) of column equilibrated as above. The col- Fraction V. umn was washed with 6 ml of the same buffer (A28o to baseline) and eluted with a 15-ml linear gradient of 100-500 mM KCI in the same buffer. Fractions (0.15 in TE, removing most of the residual nu- Fraction IV was loaded onto a 3.2 x ml) eluting between 165 and 255 mM cleic acid. The Taq Pol I activity was 9-cm (72 rnl) heparin-Sepharose column KCI were pooled (2.5 ml) and diafiltered eluted with a 600-ml linear gradient of (equilibrated with 100 mM KC1, 50 mM on a Centricon 30 membrane into a 2.5x 0-4 M urea in TE containing 20% ethyl- Tris-HCl, pH 7.5, 0.1 mM EDTA, and 0.2% storage buffer preparation lacking glyc- ene glycol. Fractions (5.5 ml) eluting be- Tween 20) at 70 ml/hr. The column was erol, forming Fraction V (Fig. 2B). tween 0.25 and 2.7 M urea were pooled washed with 170 ml of the same buffer forming Fraction IV (Fig. 2A), which containing 150 mM KC1 and eluted with Enzyme Purification Activity Assays contained 73% of the applied activity a 700-ml linear gradient of 150-650 mM and 12% of the applied protein. KC1 in the same buffer. Fractions eluting DNA polymerase assays were performed

278 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

in a 50-p.1 volume at 75~ essential- tions were either 2 mM MgCI2, 12 fmoles ixg/ml) of Taq Pol I, or 2 pmoles (4.1 ixg! ly as described) s) For full-length Taq polymerase (22.5 ng/ml), and 250 txg/ml ml) of Stoffel fragment. Reactions were Pol I, assay conditions were 25 mM activated template or 3 mM MgCl 2, 74.4 stopped by the addition of 6 Ixl of 60 mM N-t ris [hydroxym ethyl] methyl-3-amino- fmoles polymerase (140 ng/ml), and 60 EDTA. The reaction products were frac- propanesulfonic acid (TAPS) (pH 9.5), 1 nM primer-template, in an otherwise tionated electrophoretically on a 20% mM [3-mercaptoethanol, 2 mM MgCl2, 50 standard 50-1xl TAPS, pH 9.4 reaction. As- polyacrylamide gel and analyzed by au- mM KCI, 200 IXM each dATP, TTP, and say conditions of the thermal activity of toradiography. Radioactivity was quanti- dGTP, 100 IXM [o~-3ep]dCTP (0.05 Ci/ the Stoffel fragment were 25 mM Tris- tated by liquid scintillation spectrometry mmole), and 250 ixg/ml activated HCl (pH 8.3), 25 mM KCl, and either 4.0 of excised reaction products. salmon sperm DNA template. For the mM MgCl2, 11.2 fmoles polymerase (13.7 Stoffel fragment, assay conditions at ng/ml), and 250 txg/ml activated DNA 75~ were 25 mM Tris-HCl (pH 8.3), 1 mM template or 4.5 mM MgCla, 74.4 fmoles RESULTS [3-mercaptoethanol, 4 mM MgC12, 15 mM polymerase (90.8 ng/ml), and 60 nM DNA Sequence Manipulations KCI, 200 IXM each dATP, TTP, and dGTP, primer-template. 100 IxM [~-32p]dCTP (0.1--0.2 Ci/mmole), Figure 1 shows a summary of the plas- and 250 ixg/ml of activated salmon raids described below and the pertinent Thermal Inactivation Assay sperm DNA template. DNA sequence changes among them. As Thermal inactivation assays of Taq Pol I previously described, ~s) plasmid pLSG1 contains the full-length Taq Pol I gene Divalent Metal Ion Optimization and Stoffel fragment were carried out in a standard PCR mixture: 10 mM Tris-HC1 under transcriptional control of the E. A Titrisol standard solution of MgCl 2 (pH 8.3), 0.5 ng ~, DNA template, 200 p.M coli lac promoter. IPTG-induced cultures (VWR Scientific) was used to titrate a so- each of dATP, TTP, and dGTP, 0.2 IxM harboring pLSG1 produce 14-37 units of lution of EDTA colorimetrically. (es) The each of two oligonucleotide primers, Taq Pol I per milligram of total crude EDTA was then used to determine the and either 50 mM KC1, 2 mM MgC12, and extract protein (Table 2 and ref. 5) or actual concentration of previously pre- 1.25 units (4.3 ng) of Taq Pol I or 10 mM -0.005-O.015% of total protein. To in- pared stocks of MgCl2, MnCl 2, CoCl 2, KC1, 3 mM MgCl 2, and 5 units (13.5 ng) crease expression levels of Taq Pol I, sev- and NiCl 2. All optimization assays used of Stoffel fragment in a 50-1xl reaction eral manipulations were performed. Oli- activated salmon sperm DNA template volume. Reaction mixtures were overlaid gonucleotide mutagenesis of the 5' end (250 p.g/ml). Optimizations were per- with paraffin oil and incubated at either of the gene resulted in a 170-bp deletion formed in 25 mM Tris-HC1 (pH 8.3) (25 95~ or 97.5~ for 0, 2, 5, 10, 30, or 60 of Taq and vector DNA, moving the ATG mM TAPS, pH 9.4, for Taq Pol I MgC12 min. Aliquots (5 p.l) were held on ice un- of Taq Pol I into the position of the ATG optimizations), 55 mM KC1, and 11.2 til assayed in the presence of 100 ~M codon of [3-galactosidase. In addition, a fmoles polymerase (21 ng/ml for Taq Pol [e~-32p]dCTP for polymerase enzyme ac- SphI site (in italics, below) was intro- I and 13.6 ng/ml for the Stoffel frag- tivity on activated salmon sperm DNA in duced near the Taq Pol I ATG start codon ment) per reaction, in otherwise stan- the enzyme purification activity assay by changing a G to a C in the third dard 50-txl assays) s) described above. codon. Also, codons two, five, six, and seven were altered in their third posi- tions to be more AT rich, changing the KCI Optimization 5' to 3' Exonuclease Activity Assay DNA sequence from ATG AGG GGG ATG Optimization reactions (50 txl) for KC1 Five picomoles of 5'-32p-labeled, 3'- CTG CCC CTC to ATG AGA GGC ATG concentration with Taq Pol I (11.2 phosphorylated oligonucleotide DG67 CIT CCA CTT without altering the fmoles enzyme; 21 ng/ml) using both (Table 1) was annealed to 6 pmoles of amino acid sequence of the translated the M13 primer-template (53 riM) and ac- single-stranded TSYC657 DNA in the protein. The resulting plasmid was des- tivated DNA template (250 ixg/ml) con- presence (to form a 133-nucleotide gap ignated pLSG2. An IPTG-induced pLSG2 tained 2.5 mM MgC12 in an otherwise between the 3' end of DG48 and the 5' culture has Taq Pol I activity of 639 standard TAPS, pH 9.4 buffer) s) For KC1 end of DG67) or absence of 15 pmoles of units/mg crude extract protein, a 38-fold optimization in assays using the Stoffel the synthetic oligonucleotide primer increase over an uninduced culture (Ta- fragment, activated DNA template (250 DG48 (Table 1) in a 60-txl reaction con- ble 2). The 3' end of the Taq Pol I gene in txg/ml) assays (50 ixl) were performed in taining 10 mM Tris-HCl (pH 8.3), 50 mM pLSG2 was mutagenized to introduce a 25 mM Tris-HCl (pH 8.3), 4 mM MgC12 KC1, and 0.1 mM EDTA. The annealing BglII site (in italics, below) at the stop with 11.2 fmoles enzyme (13.7 ng/ml), mixture was heated to 95~ for 4 rain, codon, changing that sequence from whereas M13 primer-template (53 riM) incubated at 70~ for 10 min, and then TGATAC to TGAGATCTAC. The resulting assays (50 Ixl) were performed in 25 mM cooled to room temperature. Exonu- plasmid, with the Taq Pol I gene altered Tris-HC1 (pH 8.3), 4.5 mM MgC12 with clease activity was assayed at 70~ for 6 at both the 5' and 3' ends was designated 27.9 fmoles polymerase (34 ng/ml). KC1 min in a 30-1xl reaction containing 1 mM pSYC1578. concentration was varied from 0 to 200 [3-mercaptoethanol, 25 mM TAPS (pH mM. 9.4), 50 mM KCI, 2 mM MgCI 2, 200 ~M Cloning and Expression of Taq Pol I each TTP, dATP, dGTP, and dCTP, 6 Ixl in PL Expression Vector Thermal Activity Profile Assay (0.6 pmole single-strand DNA) of the above annealing mixture, and either The plasmid expression vector into Taq Pol I thermal activity assay condi- 0.05 pmole (157 ng/ml) or I pmole (3.1 which we cloned the Taq Pol I gene uti-

PCR Methods and Applications 279 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

TABLE 2 Taq Polymerase Activity in E. coli Extracts polymerase domain of E. coli Pol I. Be- cause the E. coli Pol I Klenow fragment is Experiment Extract Uninduced a ( - )/Induced ( + ) Specific activity,b active as a polymerase, we made a dele- I pDG160 + <~0.06c tion of Taq Pol I that we hoped would pLSG1 - 1.1 express a thermostable polymerase mol- + 14.3 ecule analogous to Klenow fragment. pLSG2 - 16.7 Unlike its full-length counterpart, the + 639 Stoffel fragment activity is not lost upon pLSG5 - 97.6 heat-induction at 41~ (Table 2). After 21 + (37~ 2790 hr at 41~ a heat-treated crude extract + (41~ 362.7 from a culture harboring pLSG8 had II pDG160 + ~<0.07 12,310 units of heat-stable DNA poly- pLSG8 - (5 hr, 30~ 61.9 merase activity per milligram of crude + (5 hr, 37~ 1037.4 extract protein, a 24-fold increase over + (5 hr, 41~ 3163.7 an uninduced culture. A heat-treated ex- - (21 hr, 30~ 506.7 tract from a 21-hr, 37~ pLSG8 + (21 hr, 37~ 9503.4 culture had 9503 units of activity per + (21 hr, 41~ 12,310.1 milligram of crude extract protein. Table aExperiment I: pDG160 and pLSG5 were heat-induced as described in Materials and Methods for 2 shows the difference in accumulated 9 hr. Experiment II: pDG160 and pLSG8 heat inductions were 5 or 21 hr. Uninduced cultures levels of polymerase activity between 5 were incubated at 30~ for the same amount of time. 1PTG induction of pLSG1 and pLSG2 is also hr and 21 hr of induction--a ninefold described in Materials and Methods. increase between 5 hr and 21 hr at 37~ bSpecific activity in units/mg total crude extract protein when assayed, as described ~s~ on clari- and a nearly fourfold increase between 5 fied, heat-treated extract, for Experiment I and as described in Materials and Methods for Exper- iment II. hr and 21 hr at 41~ CA background of 0.01% input counts has been subtracted. The specific activity represents less than two times background. Activity of Crude Extracts in PCR To determine whether both the full- length and Stoffel fragment forms of Taq lizes the bacteriophage k PL promoter from the 37~ induction (Table 2). In ad- Pol I (encoded by pLSG5 and pLSG8, re- and gene N ribosome binding site. The dition, we observed more degraded spectively) were active in PCR, we per- Taq Pol I cloning preserved the natural forms of Taq Pol I in a Western blot of formed PCR assays using heat-treated spacing between the gene N ribosome total protein from the 41~ cul- crude extracts from induced cultures of binding site and ATG start codon. In ad- ture compared to that from the 37~ - pLSG5 and pLSG8. Heat-treated extracts dition, the vector provides the Bacillus duced culture (data not shown). Based containing approximately 4, 2, 1, 0.5, or thuringiensis positive retroregulator and on the specific activity of the polymerase 0.25 units of enzyme activity were used point mutations in RNA II, rendering the and protein determination on total in otherwise standard PCR amplifica- plasmid temperature sensitive for copy crude lysate, the induced activity in the tions. ~3~ The results, shown in Figure 3, number as described. (12~ A culture har- pLSG5/DG116 culture at 37~ repre- indicated that both the full-length (lanes boring pLSG5 was heat-induced, as de- sented 1-2% of total protein being ex- 1-5) and Stoffel fragment (lanes 7-11) scribed in Materials and Methods. Taq pressed as Taq Pol I. Taq Pol 1 enzymes were thermostable Pol I activity in a heat-treated crude ex- and functioned well in PCR. The full- tract of induced pLSG5 was approxi- length enzyme functioned well in PCR at Cloning and Expression of mately four-fold higher than in pLSG2 0.5-2 units of enzyme. Faint PCR prod- Truncated Taq Pol I (Table 2). A corresponding increase in ucts were visible from reactions carried induced Taq Pol I protein was also seen We had observed immunoreactivity and out with 0.25 unit of enzyme. Four units upon SDS-PAGE analysis (data not enzymatic activity in induced cultures of enzyme in the reactions yielded high- shown). harboring pFC84 and pFC85, which were molecular-weight nonspecific products, Others have observed (26,27~ that pro- deleted for the first 615 nucleotides of as expected/28~ In contrast, PCR product teins produced upon heat induction of the Taq Pol I gene/s~ Thus, there was rea- was only observed in the reactions uti- PL promoter constructs at 42~ are insol- son to believe that other truncated forms lizing 4 units and 2 units of Stoffel frag- uble. By lowering the induction temper- of Taq Pol I could be active. The BstXI ment (lanes 7 and 8). Interestingly, a sig- ature to 37~ or 41~ these investigators recognition site in the Taq Pol I gene at nificantly higher yield of intended found they could increase the solubility nucleotide 872 (amino acid 291) is di- specific product was obtained with 4 of induced proteins significantly. We rectly upstream of the region corre- units of Stoffel fragment. In addition, compared heat-treated crude extracts of sponding to the end of E. coli Pol I small nonspecific higher-molecular-weight pLSG5 that had been induced at 37~ or fragment. In addition, the BstXl recogni- products were not visible. 41~ Enzyme assays show that we recov- tion site is approximately 570 bp (190 ered 8- to 13-fold less DNA polymerase amino acids) proximal to the beginning Enzyme Purification activity in the heat-treated extract from of the region that, when translated, the 41~ induction compared to that shows high amino acid similarity to the The purification of full-length Taq DNA

280 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

which had the amino-terminal Met, the other of which had a Pro at its amino terminus.

Optimization Studies Figure 4 shows divalent metal ion titration data for Taq Pol I (A) and Stoffel fragment (B) enzymes using MgC12, MnCI2, and CoC12. Assays using NiCI 2 yielded little, if any, detectable polymerase activity for either enzyme (data not shown). For Taq Pol I, magnesium chloride stimulated enzyme activity maximally at 2-3 mM (Fig. 4A) and 10 mM MgC12 yielded only 60% of the maximum activity, at 200 ~M each TTP, dATP, and dGTP and 100 ~M dCTP. Max- imal stimulation of activity with MnC12 (0.6 mM) and CoCI 2 (0.6 mM) was approximately 80% of the maximal activity stimulated by MgCI2. Similar to Taq Pol I, the Stoffel fragment showed a prefer-

FIGURE 3 PCR comparing the two forms of Taq Pol I. PCR using heat-treated extracts of induced cultures is described in Materials and Methods. (Lanes 1-5) Taq Pol I from pLSG5; (lanes 7-11) Stoffel fragment from pLSG8; (lane 6) molecular weight marker; (lanes 1 and 7) 4 units of enzyme; (lanes 2 and 8) 2 units; (lanes 3 and 9) 1 unit; (lanes 4 and I0) 0.5 units; (lanes 5 and 11) 0.25 units. 100

80 polymerase (Fraction V, Fig. 2A) resulted yielding a calculated molecular turnover 60 in the recovery of 32,600,000 units number of 130 nucleotides per second at 40 (73%) of polymerase activity. Between 75~ For comparison, the specific activ- 11 mg (protein concentration deter- ity of E. coli DNA polymerase I Klenow 20 mined by amino acid composition) and Fragment is between 8,000 and 30,000 0 , 14 mg (protein concentration deter- units/mg, (29-31) and the specific activity 0 2 4 6 8 mined by Lowry assay using BSA as a of T7 DNA polymerase (T7 gene 5 pro- [DIVALENT ION ] (mM) standard) of protein were obtained, re- tein and E. coli thioredoxin) is 10,000- sulting in a calculated specific activity 55,000 units/mg. (29'32) The ranges are a between 230,000 and 290,000 units/mg reflection of the dependence of the en- protein, and representing a 97- to 122- zyme activity on the nature of the DNA fold purification. The predicted molecu- substrate in the assay. 100 lar weight of Taq Pol I (based on the de- Based on amino acid analysis (data eo termined DNA sequence of the 832- not shown), recombinant Taq Pol I (in codon open reading frame) is 93,920 contrast to native Taq DNA Polymerase) 6o daltons. A specific activity of 290,000 was not blocked at the amino terminus units/mg protein yields a calculated mo- and retained the initiating methionine o 40 z_ lecular turnover number of 150 nucle- residue, as would be predicted from the 20 otides per second at 75~ properties of E. coli methionine amino The purification of the Stoffel frag- peptidase. (33) Amino acid analysis of the 0 0 5 10 15 ment (Fraction V, Fig. 2B) yielded purified Stoffel fragment yielded a com- [DIVALENT ION ] (mM) 1,291,588 units of DNA polymerase ac- plex result. The expected amino acid se- tivity and 2.8 mg of protein. The calcu- quence at the amino terminus of the FIGURE 4 Divalent metal ion titration. The influence of MgCI2(Q ), MnC12 (i), and lated specific activity of the Stoffel frag- protein is Met-Ser-Pro-Lys-Ala. Unlike CoCI 2 (C)) on Taq Pol I(A), and Stoffel fragment is 369,000 units/mg protein, Taq Pol I, the majority of the purified ment (B) enzyme activities. Assays and titra- representing a 451-fold purification. The Stoffel fragment did not retain the initi- tion are as described in Materials and Meth- predicted molecular weight of the Stoffel ating methionine residue. The amino ods and results are plotted as the percent of fragment (based on the determined DNA terminus of the major species was Ser- maximum dCMP incorporation with respect sequence of the 543- or 544- codon open Pro-Lys-Ala. However, there were two to increasing concentration of divalent metal reading frame, see below) is 61.3 kD, minor species in the analysis, one of ion.

PCR Methods and Applications 281 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

ence for MgCI z. Unlike the full-length mately 58% activity at 0 and 150 mM. In A form, Stoffel fragment enzyme activity contrast, with Taq Pol I and the M13 120 was maximally stimulated over a broad primer-template, maximum activity was concentration range of MgCl2, with a achieved at 10 mM KC1 (activity in the lOO maximum of 4 mM (Fig. 4B). Similarly, absence of KCI was not determined) with 80 Stoffel fragment activity was stimulated 48% activity at 50 mM KCI. For the Stoffel over a broad concentration range of fragment with activated DNA template, 6o MnC1 z, with a maximum of 0.92 mM maximum activity was achieved with 0 o 40 (60% of the maximal activity achieved KC1, although polymerase activity re- z with MgC12). Cobalt chloride stimulated mained above 90% of maximum up to o~ 20

Stoffel fragment activity maximally at 45 mM KCI and remained over S0% at o 0.88 mM (27% of the maximal activity 100 mM KCI. In contrast, with M13 40 50 60 70 80 90 achieved with MgClz). primer-template and the Stoffel frag- TEMPERATURE(C) Figure 5 shows that the influence of ment, activity declined rapidly at greater KCI concentration differs between Taq than 10 mM KCl. Assaying the M13 Pol I and the Stoffel fragment and also primer-template with 40 mM KCI yielded 120 depends on the template used in the as- approximately 50% maximum activity 10o say. For Taq Pol I assayed with activated and activity declined to near 0 by 100 z DNA template, the optimum KC1 con- o_ mM KCI. I- 8o ,,r centration was 55 mM with approxi- n- O 80 n- o 40 _z Thermoactivity and Thermal A o~ 20 120 [ Inactivation 0 - _o ,oo ~ Thermal activity profiles of Taq Pol I and 30 40 50 60 70 80 90 I- the Stoffel fragment on M13 primer-tem- TEMPERATURE (C) ,,r 80 n- plate and activated salmon sperm DNA O FIGURE 6 Thermal activity profiles of purified were compared. Figure 6 shows that the recombinant Taq DNA polymerases. Activity optimal temperature depends on the na- is represented as percent of maximum dCMP ture of the DNA template. DNA poly- incorporation for each template and enzyme merase activity increased faster, initially, (assays performed under previously deter- with increasing temperature on the M13 mined optimum conditions). (A) Taq Pol I o ~ I = I with: (O) activated template (100% represents o loo 2oo primer-template than on the activated 114 pmoles of dCMP incorporated); (C)) M13 [KCI] (ma) DNA. For both forms of the enzyme, the primer-template substrate is primarily primer-template (100% represents 80 pmoles of dCMP incorporated). (B) Stoffel fragment single-stranded DNA (approximately 390 with: (0) activated template (100% represents }.LM total single-stranded template nucle- 150 pmoles of dCMP incorporated); (C)) M13 100 otide and 53 nM primer duplex). In con- primer-template (100% represents 130 pmoles trast, the activated DNA (approximately of dCMP incorporated). 80 750 p,M total DNA nucleotide) substrate

60 is primarily duplex DNA containing an undefined, but probably high, concentration of nicks and short gaps. (17) The tivity on salmon sperm DNA at 70-75~ pronounced difference in thermal activ- could reflect denaturation of the tem- ity with inherently different substrates plate blocking strand.

. i , o i o | , could reflect differences either in DNA Figure 7 shows the steady-state ther- 20 40 60 80 1O0 binding constants or stability constants mal inactivation of recombinant Taq Pol [KCl] (mM) for single-stranded or double-stranded I and the Stoffel fragment at 95~ and FIGURE 5 KC1 titration. The influence of KCI DNA as a function of temperature. Alter- 97.S~ under PCR buffer conditions. Taq on Taq Pol I,(A) and Stoffel fragment (B) en- natively, the relative increase in activity Pol I has an activity half-life of approxi- zyme activities using either activated salmon on M13 primer-template DNA at 65~ mately 45-50 min at 95~ and 9 min at sperm DNA template (Q) or the defined M13 and lower may accurately reflect the dif- 97.5~ Native Taq Pol I was inactivated primer-template (0), as described in Materials ference in extension rate for the poly- with similar kinetics (data not shown). and Methods. One-hundred percent incorpo- merase when there is no blocking The half-life of the Stoffel fragment at ration represents 155 pmoles of dCMP incor- strand. If the extension rate is limited by 97.5~ is approximately 21 min. The porated for Taq Pol I with the activated template, 106 pmoles for Taq Pol I with the M13 the presence of a blocking strand (e.g., thermal inactivation of the Stoffel frag- primer-template, 110 pmoles for Stoffel frag- limited 5' to 3' exonuclease activity dur- ment was unaffected by KC1 concentra- ment with the activated template, and 92 ing synthesis or limited ability to cata- tion over a range of O-50 mM (data not pmoles for Stoffel fragment with the M13 lyze strand-displacement synthesis), the shown). These results do not measure primer-template. sharp and pronounced transition in ac- thermoresistance of the enzymes in ther-

282 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

however, it is only 60% as effective rela- 100 phosphorylated oligonucleotide. The activity was moderate on the 5'-phos- tive to MgCI z. 80 phorylated recessed end of double- Based on the physical and biochemi- >- stranded DNA, which was a 5'-phospho- cal characterization of the polymerases I- 60 rylated oligonucleotide annealed to purified by Chien et al. (34) and Kaledin et I-- single-stranded circular DNA in this as- al., (3s) it is unclear whether these pro- say. It is only in the presence of an up- teins are the products of a distinct gene 20 stream primer that the enzyme pos- (or genes) from that which encodes the sessed substantial 5' to 3' exonuclease DNA polymerase described here, or are i i i i l i 0 20 40 60 activity, yielding 78% release of 5' nucle- partially purified proteolytic degrada- TIME (min) otides compared to 18% in the absence tion fragments of the same translation FIGURE 7 Thermoresistance of purified re- of a competent upstream primer. Thus, product. The Taq DNA polymerase de- combinant Taq DNA polymerases. Percent re- this activity is a polymerization-depen- scribed by Engelke et al. (7) is the transla- maining enzyme activity as a function of dent 5' to 3' exonuclease. In contrast, as tion product of the gene initially identi- time. (0) Taq Pol I, 97.5~ (C)) Taq Pol I, predicted from the amino acid sequence, fied (s) and further modified as reported 95~ (11) Stoffel fragment, 97.5~ The inter- the Stoffel fragment was completely de- in this manuscript. In the expressed pro- section of each solid line with the dashed line void of exonuclease activity (Table 3). tein isolated by Engelke, et al., (7) the first indicates the approximate t 1/2 for each condi- two amino acids encoded for in the na- tion. tive gene (Met-Arg-) were replaced with three vector-encoded amino acids Met- DISCUSSION Asn-Ser. Expression levels were approxi- mal cycling or otherwise repeated short DNA polymerases isolated from meso- mately 3.5-fold lower than those exposure to high temperatures. philic microorganisms have been exten- achieved with pLSG5 (Table 2), as deter- sively characterized, but much less is mined by activity assays of clarified heat- known about the properties of thermo- treated extracts. The purification re- 5' to 3' Exonuclease Activity stable DNA polymerases. With the intro- sulted in an enzyme preparation that The ability of Taq Pol I and the Stoffel duction of the PCR method of DNA am- contained trace contamination from E. fragment to excise 5'-terminal nucle- plification, considerable interest has coli proteins and had a specific activity of otides from a variety of DNA substrates been focused on the DNA polymerases of 5263 units/mg. A clone expressing a Taq was examined (Table 3). The 5' to 3' ex- thermophilic organisms. A DNA poly- DNA polymerase gene has also been onuclease activity of Taq Pol I was neg- merase isolated from T. aquaticus has identified by Sagner et al. (6~ This clone ligible on single-stranded DNA, i.e., a 5'- been previously described. (34'3s~ The en- expresses a 94-kD Taq DNA polymerase zyme is reported to have an approximate that is presumably identical to the poly- molecular weight of 62,000-68,000, a merase described here. An expression temperature optimum of 70-80~ and a clone producing an active amino-termi- TABLE 3 5' to 3' Exonuclease Activity pH optimum in the range of 7.8-8.3. Op- nally truncated Taq DNA polymerase, Percent timal activity is obtained with 60--200 KlenTaq, has also been described, (8) with Enzyme Substrate release a mM KCI and 10 mM Mg 2+ . Manganese is translational initiation occurring at the only 20-50% effective and shows an op- corresponding Met 236 codon of the na- Taq Pol I timum of 1-2 mM. The enzyme as iso- tive gene. (0.05 pmole) ss DNAb 1.7 ds DNAc ND lated by Chien et al. (34~ has a specific ac- The DNA polymerases from a number ds DNAd 64.4 tivity ~36) greater than 474.8 units/mg, of other thermophilic eubacteria have with primer whereas the enzyme isolated by Kaledin also been isolated and partially charac- et al. (3s) has a specific activity between terized (Table 4). These include Bacillus Taq Pol I 1612 units/mg and 5200 units/mg, de- stearothermophilus (Bst), (37'38) Thermus (1 pmole) ss 6.4 ruber (Tru), (39) T. flavus (Tfl), (4~ T. ther- ds 18.0 pending on the nature of the substrate. ds with primer 77.6 In contrast, the T. aquaticus DNA poly- mophilus (Tth) HB-8, (a1'42~ and Thermo- merase we describe has a molecular toga sp. (Tsp) strain FjSS3-B.1. (43) In ad- Stoffel fragment weight of 94,000, and a specific activity dition to the eubacteria, thermophilic (2 pmoles) ss ND of 292,000 units/mg. Optimal polymer- archaeal DNA polymerases have also ds ND been characterized. The Archaea consti- ds with primer 0.5 ization activity is achieved at 75-80~ in the presence of 10-55 mM KC1 and 2-3 tute a group of prokaryotes with an in- aRelease is percent of 5'-terminal nucleotides mM Mg 2+. Manganese is approximately termediate phylogenetic position be- converted to mono- or dinucleotides, quanti- 80% as effective and shows an optimum tween eukaryotes and eubacteria. They tated by liquid scintillation spectrometry of of 0.6 mM. The Stoffel fragment behaves comprise the most extreme thermo- excised reaction products previously sepa- somewhat differently than the full- philic organisms known. A DNA poly- rated via PAGE. merase has been purified and character- bOligonucleotide DG67, 32p-phosphorylated length enzyme in activity assays. It has a ized from the thermoacidophilic at its 5' end. preference for low ionic strength regard- CDG67 (above) annealed with TSYC657. less of template, and has a broader and archaeon Sulfolobus acidocaldarius dTSYC657/DG67 complex (above) addition- higher (-4 mM) Mg z+ optimum. The (Sac). (44--46) Similarly, DNA polymerases ally annealed with upstream DG48. optimum for manganese is also broader; from Sulfolobus solfataricus (Sso),(47)

PCR Methods and Applications 28J Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

TABLE 4 Properties of Thermophilic DNA Polymerases

Optima Molecular Exonuclease Temperature weight Specific activity Enzyme MgC12 (mM) pH KCI (mM) (~ (x 10 3) activity~ 5' to 3' 3' to 5'

Taq Pol I 2 9.4 55 80 94 292,000 + Stoffel fragment 4 8.3 10 80 61.3 369,000 - Taq Pol (34) 10 7.8 60 80 63-68 >475 - Taq Pol (35) 10 8.3 100-200 70 60-62 1,600-5,200 - Bst Pol (37) 30 9.0 ND 65 ND 424 - Bst Pol (38) 20 8-9 270 60 76 16,000 ___ Tru Pol (39) 2.5 9.0 15 70 70 8,000 - Tfl Pol (40) 10-40 10.0 50 70 66 12,000 - Tth Pol (41) A ND ND ND 50 110-150 58 - B ND ND ND 63 - 110 922 - C ND ND ND 63 - 110 1,380 - D Tth Pol (42) ND ND ND ND 67 4,000 - ND Tsp Pol (43) 10 7.5-8.0 0 >80 85 16 ND ND Sac Pol (44) 0.1-1 6.0-8.0 0 65 100 71,500 - + Sac Pol (45) ND ND ND 70 100 60,000 - Sso Pol (47) 3 6.8 0 75 110 42,360 - Tac Pol (48) 4 8.0 5 b 65 88 17,500 - + Mth Pol (52) 10-20 8.0 100 65 72 4,720 + + aSpecific activities have been normalized such that one unit equals the amount of activity which will incorporate 10 nmoles of dNTPs into product in 30 rain. bNH4C1.

Thermoplasma acidophilum (Tac), (4s) only Mth DNA polymerase resembles E. grows optimally at 80~ ~43~ Sac grows at Therrnococcus litoralis (T/i), (49'50) and Py- coli DNA polymerase I, in that it contains temperatures up to 85~ (4s) and Sso rococcus furiosus (Pfu) (sl) have been iso- both a 5' to 3' exonuclease and a 3' to 5' grows optimally at 87~ ~aT~ whereas Taq lated. The purification and characteriza- exonuclease. ~s2~ The lack of any 5' to 3' has an optimum temperature for growth tion of a DNA polymerase from a exonucleolytic activity associated with of only 70~ ~s6) Yet for those enzymes methanogenic archaeon Methanobacte- the majority of the thermostable DNA whose thermostabilities are reported in rium thermoautotrophicum (Mth) has also polymerases may reflect a real difference the literature, Taq Pol I, with an activity been reported (sz) among the enzymes, differences in the half-life of 45--50 min at 95~ and 9 rain Table 4 shows that considerable vari- sensitivity of the assay procedures, or, al- at 97.5~ is only second in thermosta- ation exists among those thermostable ternatively, may be the consequence of bility to that of the truncated Stoffel DNA polymerases that have been charac- proteolytic degradation resulting in an fragment, with an activity half-life of 21 terized in the literature. A distinct fea- amino-terminal truncation of the pro- rain at 97.5~ In comparison, Tsp strain ture of all these DNA polymerases is that tein. Considerable care in purification is FjSS3-B.1 has an activity half-life of 7 they all appear to be monomeric. required, as DNA polymerases appear to min at 95~ (43~ Sac has an activity half- Whether or not a multimeric DNA poly- be particularly susceptibl.e to proteolytic life of less than 15 min at 90~ ~as) and merase similar to E. coli DNA polymerase cleavage. (44,47'4s's2~ Preliminary evi- Sso has an activity half-life of 5 rain at III exists in thermophilic bacteria is un- dence suggests that a 94,000-dalton DNA 90~ ~47~ This extraordinary thermosta- known. It is also unknown whether polymerase purified from Tth contains bility makes Taq Pol I especially suited to these polymerases are involved in repli- an inherent 5' to 3' exonuclease identi- those applications where repeated incu- cation or repair. A substantial difference cal with that of Taq Pol I. (ss~ The absence bation at very high temperatures is de- among the enzymes is the presence or of an inherent 3' to 5' exonuclease activ- sired. Subsequent to completion of this absence of associated exonuclease activity in a majority of the polymerases may manuscript, the thermostability of Tli ities. Only Taq Pol I (53,54) (Table 3), and again reflect a real difference between DNA polymerase has been described. ~sT~ Mth DNA polymerase (sz) have been species; alternatively a 3' to 5' exonu- The polymerase, derived from a hyper- shown to contain an inherent 5' to 3' clease may exist as a separate subunit or thermophilic Archae which grows at exonuclease activity, whereas only Tac auxiliary protein. temperatures up to 98~ has an activity DNA polymerase, (48~ Mth DNA poly- A unique feature of Taq Pol I is its half-life of about 2 hr at 100~ merase, (sz~ Tli DNA polymerase, ~49,s~ exceptional thermostability. This is sur- Taq Pol I has many features in com- and Pfu DNA polymerase r have been prising considering that several of the mon with E. coli DNA polymerase I. Con- shown to contain an inherent 3' to 5' above-mentioned organisms grow opti- siderable amino acid similarity exists be- exonuclease activity. Of the character- mally at temperatures considerably tween Taq Pol I and E. coli DNA ized thermostable DNA polymerases, higher than Taq; Tsp strain FjSS3-B.1 polymerase I. ~s'sS's9) The amino-termi-

284 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

hal region of Taq Pol I corresponds to the replication of short stretches of DNA merase at the site of the hairpin. (s3'6s'66) the amino-terminal domain of E. coli like those presumably seen with acti- Extension with Taq Pol I could result in DNA polymerase I shown to contain the vated salmon sperm DNA template, the exonucleolytic cleavage of the template 5' to 3' exonuclease activity. ~6~ The car- decreased dissociation rate of the poly- strand at the hairpin structure, rather boxy-terminal region of Taq Pol I corre- merase-replicated template complex than strand displacement, creating DNA sponds to the E. coli DNA polymerase I may cancel the gains from increased pro- fragments that are incapable of serving domain shown to contain DNA poly- cessivity, and result in a somewhat lower as templates in later cycles of the PCR. merase activity. (61~ For the 3' to 5' exo- synthesis rate (see Fig. 5A). For the Stof- The Stoffel fragment may be useful in nuclease domain of E. coli DNA poly- fel fragment, however, with a much PCR where a large amount of product is merase I, (62~ no meaningful alignment lower processivity, the increase in pro- desired. Standard PCRs with Taq Pol I with Taq Pol I was obtained. The domain cessivity dominates, and maximal activ- reach a plateau stage in late cycle PCR, structure of E. coli DNA polymerase I is ity is seen at low ionic strength with where product accumulation is no conserved in Taq Pol I. Similar to the both templates (see Fig. 5B). An in- longer exponential. An effect contribut- Klenow fragment of E. coli DNA poly- creased enzyme-substrate stability in ing to plateau may be the renaturation of merase I, (6~ Stoffel fragment retains low salt may increase the activity of the product strands during extension. This all of the DNA polymerase activity of the Stoffel fragment on activated salmon could result in a substrate for the 5' to 3' full-length protein, but is completely de- sperm template by favoring strand-dis- exonuclease activity of Taq Pol I, where void of any 5' to 3' exonuclease activity. placement synthesis. Alternatively, the the renatured strand can be cleaved, and Characterization of the 5' to 3' exonu- differences between full-length Taq Pol I synthesis may proceed under nick-trans- clease activity of the two full-length pro- and the Stoffel fragment observed on ac- lation conditions, resulting in no net teins reveals that the mechanism of ex- tivated template may be the result of the gain in product. The Stoffel fragment, on onuclease action is very similar for the effect of ionic strength on the 5' to 3' the other hand, may proceed under two enzymes. (64~ Like E. coil DNA poly- exonuclease activity of Taq Pol I, for strand-displacement conditions, with no merase I, Taq Pol I requires a duplex which nick-translation synthesis pre- destruction of product. The increased structure for 5' to 3' exonuclease activity sumably predominates over strand-dis- thermostability of the Stoffel fragment and is stimulated by concurrent poly- placement synthesis. Differences are also allows for the amplification of excep- merization. ~6s'66~ The major difference apparent in the response of the two en- tionally G/C-rich targets, where high de- between the two enzymes, aside from zymes to divalent cations. The Stoffel naturation temperatures are required. the exceptional thermostability and fragment is optimally active over a The broader Mg 2+ optimum for activity thermoactivity of Taq Pol I, appears to be broader range of Mg 2§ concentrations, of the Stoffel fragment may prove useful the lack of an inherent 3' to 5' exonu- but is less active when either Mn 2§ or in multiplex PCR, where multiple tem- clease activity in Taq Pol I. Although this Co 2 + is substituted for Mg 2+ (see Fig. 4). plates are being amplified simulta- is not a unique feature of all thermo- The truncated Taq DNA polymerase, neously. (67) Alternatively, in applica- stable DNA polymerases, as discussed KlenTaq, was measured to have a two- tions such as random mutagenesis above, it appears to be a common feature fold lower mutation rate as compared to PCR C68~ and reverse transcription, (60) among the DNA polymerases isolated to the full-length enzyme. (8) The model for where Mn 2+ is desired as the divalent date from thermophilic eubacteria. this comparative increase in fidelity de- cation cofactor, full-length Taq Pol I may Subtle differences exist between the rives from a consideration of the relative prove superior. The 5' to 3' exonuclease polymerase activity of the Stoffel frag- processivities of the two forms of the en- activity of Taq Pol I can be used in a PCR- ment and full-length Taq Pol I. Whereas zyme, and their subsequent abilities to based detection system to generate a spe- the Stoffel fragment is more thermo- extend a mispaired nucleotide (i.e., cific detectable signal concomitantly stable than Taq Pol I, it is not as active at lower processivity results in suppression with amplification. (s3) Taq Pol I also temperatures greater than 80~ (see Fig. of mismatch extension, thereby decreas- may be preferable in the amplification of 6). This may be due to a difference in the ing the misincorporation rate of the en- long templates, where increased proces- ability of the enzyme to bind to DNA at zyme). Analogously, the Stoffel fragment sivity may be an advantage, whereas the high temperatures. That a difference in may also possess a lower comparative Stoffel fragment, if shown to have DNA binding exists between the two en- mutation rate. higher fidelity than Taq Pol I, may be zymes is reflected in the approximately The differences between the two en- preferable for high-fidelity amplifica- 10-fold lower processivity of the Stoffel zymes can be exploited for a variety of tions. The lower processivity of the Stof- fragment. ~ss~ This difference in proces- applications. The Stoffel fragment may fel fragment also makes the enzyme use- sivity also may explain the observed dif- be preferred for performing dideoxy nu- ful for the amplification of rare mutant ference in sensitivity to ionic strength cleotide sequencing, where the 5' to 3' alleles in a background of normal DNA, between the enzymes. A low salt concen- exonuclease activity of Taq Pol I may using allele-specific primers where 3' tration increases the stability of the pro- produce unwanted artifacts. The Stoffel mismatch extension is suppressed rela- tein-nucleic acid complex and decreases fragment also may prove superior in per- tive to that of full-length Taq Pol I. ~7~ the rate of dissociation of the poly- forming the PCR with certain templates merase from the replicated template, re- that contain stable secondary structure. ACKNOWLEDGMENTS sulting in an increase in processivity. Extension of a primer on a template Thus, in low salt the higher processivity strand possessing a hairpin structure cre- We gratefully acknowledge Max Gottes- results in increased activity on long tem- ates an ideal substrate for the 5' to 3' man for providing E. coli strain N6590; plates (i.e., primed M13 template). For exonuclease activity of Taq DNA poly- Kirk Dakis for technical assistance; Lauri

PCR Methods and Applications 285 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

Goda, Dragan Spasic, and Corey Leven- 10. Meselson, M. and R. Yuan. 1968. DNA re- cleic Acids Res. 16: 6987-6999. son for preparation of oligonucleotides; striction enzyme from E. coll. Nature 23. Woods, D.E., A.F. Markham, A.T. Ricker, Ken Myambo for DNA sequence analy- 217: 1110-1114. G. Goldberger, and H.R. Colten. 1982. Iso- sis; Keith Bauer and Jodi Switzer for fer- ll. Miller, J.H. 1972. Generalized transduc- lation of cDNA clones for the human mentation support; Ken Watt and Frank tion; use of P1 in strain construction. In complement protein factor B, a class III Experiments in molecular genetics (ed. J.H. major histocompatibility complex gene Buschman for amino-terminal sequence Miller), pp. 201-205. Cold Spring Harbor product. Proc. Natl. Acad. Sci. 79: 5661- analysis and analysis of amino acid com- Laboratory, Cold Spring Harbor, New 5665. position; Robert Watson for advice and York. 24. Vogel, H.J. and D.M. Bonner. 1956. Acety- support; Tom Myers, Tom White, John 12. Wang, A.M., A.A. Creasey, M.B. Ladner, lornithinase of Escherichia coil: Partial pu- Sninsky, Hamilton Smith, Norman Arn- L.S. Lin, J. Strickler, J.N. Van Arsdell, R. rification and some properties. J. Biol. helm, and Ellen Daniell for advice and Yamamoto, and D F. Mark. 1985. Molec- Chem. 218: 97-102. critical review of the manuscript; Sharon ular cloning of the complementary DNA 25. Skoog, D.A. and D.M. West. 1976. Funda- Nilson and Eric Ladner for preparation for human tumor necrosis factor. Science mentals of analytical biochemistry, 3rd edi- of figures; and Robin Kurka for prepara- 228: 149-154. tion, pp. 737-739. tion of the manuscript. 13. Moore, S.K. and E. James. 1976. Purifica- 26. Piatak, M., J.A. Lane, W. Laird, M.J. Bjorn, tion and electrophoretic assay of T4-in- A. Wang, and M. Williams. 1988. Expres- duced polynucleotide ligase for the in sion of soluble and fully functional ricin REFERENCES vitro construction of recombinant DNA A chain in Escherichia coli is temperature- molecules. Anal. Biochem. 75: 545-554. sensitive. J. Biol. Chem. 263: 4837-4843. 1. Saiki, R.K., S. Scharf, F. Faloona, K.B. Mul- 14. Sambrook, J., E.F. Fritch, and T. Maniatis. 27. Lin, T.-C., J. Rush, E.K. Spicer, and W.H. lis, G.T. Horn, H.A. Erlich, and N. Am- 1989. Molecular cloning: A laboratory man- Konigsberg. 1987. Cloning and expres- helm. 1985. Enzymatic amplification of ual, 2nd edition. Cold Spring Harbor Lab- sion of T4 DNA polymerase. Proc. Natl. ~-globin genomic sequences and restric- oratory Press, Cold Spring Harbor, New Acad. Sci. 84: 7000-7004. tion site analysis for diagnosis of sickle York. 28. Innis, M.A. and D.H. Gelfand. 1990. Op- cell anemia. Science 230: 1350-1354. 15. Beaucage, S.L. and M.G. Caruthers. 1981. timization of PCRs. In PCR protocols: A 2. Mullis, K.B. and F. Faloona. 1987. Specific Deoxynucleoside phosphoramidites-a guide to methods and applications (ed. M.A. synthesis of DNA in vitro via a poly- new class of key intermediates for deoxy- Innis, D.H. Gelfand, J.J. Sninsky, and T.J. meraseocatalyzed chain reaction. Methods polynucleotide synthesis. Tetrahedron White), pp. 3-12 Academic, San Diego, Enzymol. 155: 335-350. Lett. 22: 1859-1862. California. 3. Saiki, R.K., D.H. Gelfand, S. Stoffel, S. 16. Sinha, N.D., J. Biernat, J. McManus, and 29. Tabor, S., H.E. Huber, and C.C. Richard- Scharf, R. Higuchi, G.T. Horn, K.B. Mullis, H. K6ster. 1984. Polymer support son. 1987. Escherichia coli thioredoxin and H.A. Erlich. 1988. Primer-directed en- oligonucleotide synthesis XVlII: Use confers processivity on the DNA poly- zymatic amplification of DNA with a ther- of ~-cyanoethyl-N,N-dialkylamino-/N- merase activity of the gene 5 protein of mostable DNA polymerase. Science morpholino phosphoramidite of deoxy- bacteriophage T7. J. Biol. Chem. 262: 239: 487-491. nucleosides for the synthesis of DNA frag- 16212-16223. 4. Innis, M.A., K.B. Myambo, D.H. Gelfand, ments simplifying deprotection and 30. Englund, P.T., J.A. Huberman, T.M. Jovin, and M.A.D. Brow. 1988. DNA sequencing isolation of the final product. Nucleic Ac- and A. Kornberg. 1969. Enzymatic synthe- with Thermus aquaticus DNA polymerase ids Res. 12: 4539-4557. sis of deoxyribonucleic acid: XXX. Bind- and direct sequencing of polymerase 17. Livingston, D.M., D.C. Hinkel, and C.C. ing of triphosphates to deoxyribonucleic chain reaction-amplified DNA. Proc. Natl. Richardson. 1975. Deoxyribonucleic acid acid polymerase. J. Biol. Chem. 244: Acad. Sci. 85: 9436-9440. polymerase Ill of Escherichia coli: Purifica- 3038-3044. 5. Lawyer, F.C., S. Stoffel, R.K. Saiki, K. tion and properties. J. Biol. Chem. 31. Kelley, W.S. and K.H. Stump. 1979. A Myambo, R. Drummond, and D.H. Gel- 250: 461-469. rapid procedure for isolation of large fand. 1989. Isolation, characterization, 18. Long, C.M., M.-J. Virolle, S.-Y. Chang, S. quantities of Escherichia coli DNA poly- and expression in Escherichia coil of the Chang, and M.J. Bibb. 1987. a-Amylase merase I utilizing a hpolA transducing DNA polymerase gene from Thermus gene of Streptomyces limosus: Nucleotide phage. ]. Biol. Chem. 254: 3206-3210. aquaticus. J. Biol. Chem. 264: 6247-6437. sequence, expression motifs, and amino 32. Hori, K., D.F. Mark, and C.C. Richardson. 6. Sagner, G., R. Rfiger, and C. Kessler. 1991. acid sequence homology to mammalian 1979. Deoxyribonucleic acid polymerase Rapid filter assay for the detection of DNA and invertebrate a-amylases. J. Bacteriol. of bacteriophage T7: Purification and polymerase activity: Direct identification 169: 5745-5754. properties of the phage-encoded subunit, of the gene for the DNA polymerase from 19. Hanahan, D. 1983. Studies on transforma- the gene 5 protein. L Biol. Chem. Thermus aquaticus. Gene 97: 119-123. tion of Escherichia coli with plasmids. J. 254: 11591-11597. 7. Engelke, D.R., A. Krikos, M.E. Bruck, and MoL Biol. 166: 557-580. 33. Ben-Bassat, A., K. Bauer, S.-Y. Chang, K. D. Ginsburg. 1990. Purification of Ther- 20. Birnboim, H.C. andJ. Doly. 1979. A rapid Myambo, A. Boosman, and S. Chang. mus aquaticus DNA polymerase expressed alkaline extraction procedure for screen- 1987. Processing of the initiation methio- in Escherichia coll. Anal. Biochem. ing recombinant plasmid DNA. Nucleic nine from proteins: Properties of the Es- 191: 396-400. Acids Res. 7: 1513-1523. cherichia coil methionine peptidase and 8. Barnes, W.M. 1992. The fidelity of Taq 21. Sanger, F., S. Nicklen, and A.R. Coulson. its gene structure. J. Bacteriol. 169: 751- polymerase catalyzing PCR is improved 1977. DNA sequencing with chain-termi- 757. by an N-terminal deletion. Gene 112: 29- nating inhibitors. Proc. Natl. Acad. Sci. 34. Chien, A., D.B. Edgar, and J.M. Trela. 35. 74: 5463-5467. 1976. Deoxribonucleic acid polymerase 9. Cole, G.E., P.C. McCabe, D. lnlow, D.H. 22. Fritz, H.-J., J. Hohlmaier, W. Kramer, A. from the extreme thermophile Thermus Gelfand, A. Ben-Bassat, and M.A. Innis. Ohmayer, and J. Wippler. 1988. Oligonu- aquaticus. J. Bacteriol. 127: 1550-1557. 1988. Stable expression of Aspergillus cleotide-directed construction of muta- 35. Kaledin, A.S., A.G. Slyusarenko, and S.I. awamori glucoamylase in distiller's yeast. tions: A gapped duplex DNA procedure Gorodetskii. 1980. Isolation and proper- Biotechnology 6: 417-421. without enzymatic reactions in vitro. Nu- ties of DNA polymerase from extremely

286 PCR Methods and Applications Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

thermophilic bacterium Thermus aquati- rium Sulfolobus solfataricus, ltal. J. Bio- 61. Freemont, P.S., D.L. Ollis, T.A. Steitz, and cus YT1. Biokhimiya 45: 644-651. chem. 39: 83-99. C.M. Joyce. 1986. A domain of the Kle- 36. Specific activities have all been normal- 48. Hamal, A., P. Forterre, and C. Elie. 1990. now fragment of Escherichia coli DNA ized such that one unit is defined as that Purification and characterization of a polymerase I has polymerase but no exo- amount of activity that will incorporate DNA polymerase from the archaebacte- nuclease activity. Proteins: Struct. Funct. 10 nmoles of dNTPs into product in 30 rium Thermoplasma acidophilum. Eur. J. Genet. 1: 66-73. min. Biochem. 190: 517-521. 62. Derbyshire, V., P.S. Freemont, M.R. Sand- 37. Stenesh, J. and B.A. Roe. 1972. DNA poly- 49. Matilla, P., J. Korpela, T. Tenkanen, and K. erson, L. Beese, J.M. Friedman, C.M. merase from mesophilic and thermo- Pitk~nen. 1991. Fidelity of DNA synthesis Joyce, and T.A. Steitz. 1988. Genetic and philic bacteria: I. Purification and proper- by the Thermococcus litoralis DNA poly- crystallographic studies of the 3',5'-exo- ties of DNA polymerase from Bacillus merase-an extremely heat stable enzyme nucleolytic site of DNA polymerase I. Sci- licheniformis and Bacillus stearothermo- with proofreading activity. Nucleic Acids ence 240: 199-201. philus. Biochim. Biophys. Acta 272: 156- Res. 19: 4967-4973. 63. Klenow, H and I. Henningsen. 1970. Se- 166. 50. Perler, F.B., D.G. Comb, W.E. Jack, L.S. lective elimination of the exonuclease ac~ 38. Kaboev, O.K., L.A. Luchkina, A.T. Akhme- Moran, B. Qiang, R.B. Kucera, J. Benner, tivity of the deoxyribonucleic acid poly- dov, and M.L. Bekker. 1981. Purification B.E. Slatko, D.O. Nwankwo, S.K. Hemp- merase from Escherichia coli B by limited and properties of deoxyribonucleic acid stead, C.K.S. Carlow, and H. Jannasch. proteolysis. Proc. Natl. Acad. Sci. 65: 168- polymerase from Bacillus stearothermo- 1992. Intervening sequences in an Ar- 175. philus. ]. Bacteriol. 145: 21-26. chaea DNA polymerase gene. Proc. Natl. 64. R.D. Abramson and D.H. Gelfand, in 39. Kaledin, A.S., A.G. Slyusarenko, and S.I. Acad. Sci. 89: 5577-5581. preparation. Gorodetskii. 1982. Isolation and proper- 51. Lundberg, K.S., D.D. Shoemaker, M.W.W. 65. Deutscher, M.P. and A. Kornberg. 1969. ties of DNA polymerase from the ex- Adams, J.M. Short, J.A. Sorge, and E.J. Enzymatic synthesis of deoxyribonucleic tremely thermophilic bacterium Thermus Mathur. 1991. High-fidelity amplification acid: XXIX. Hydrolysis of deoxyribonu- tuber. Biokhimiya 47: 1785-1791. using a thermostable DNA polymerase cleic acid from the 5' terminus by an ex- 40. Kaledin, A.S., A.G. Slyusarenko, and S.I. isolated from Pyrococcus furiosus. Gene onuclease function of deoxyribonucleic Gorodetskii. 1981. Isolation and proper- 108: 1-6. acid polymerase. J. Biol. Chem. 244: ties of DNA polymerase from the ex- 52. Klimczak, L.J., F. Grummt, and K.J. 3029-3037. tremely thermophilic bacterium Thermus Burger. 1986. Purification and character- 66. Kelly, R.B., N.R. Cozzarelli, M.P. Deut- flavus. Biokhimiya 46: 1576-1584. ization of DNA polymerase from the ar- scher, I.R. Lehman, and A. Kornberg. 41. Rfittimann, C., M. Cotor~s, J. Zaldivar, chaebacterium Methanobacterium ther- 1970. Enzymatic synthesis of deoxyribo- and R. Vicufia. 1985. DNA polymerases moautotrophicum. Biochemistry 25: 4850- nucleic acid: XXXII. Replication of duplex from the extremely thermophilic bacte- 4855. deoxyribonucleic acid by polymerase at a rium Thermus thermophilus HB~ Eur. J. 53. Holland, P.M., R.D. Abramson, R. Wat- single strand break. L Biol. Chem. 245: Biochem. 149: 41-46. son, and D.H. Gelfand. 1991. Detection of 39-45. 42. Carballeira, N., M. Nazabal, J. Brito, and specific polymerase chain reaction prod- 67. Chamberlain, J.S, R.A. Gibbs, J.E. Ranier, O. Garcia. 1990. Purification of a thermo- uct by utilizing the 5' to 3' exonuclease P.N. Nguyen, and C.T. Caskey. 1988. De- stable DNA polymerase from Thermus activity of Thermus aquaticus DNA poly- letion screening of the Duchenne muscu- thermophilus HB8, useful in the polymerase. Proc. Natl. Acad. Sci. 88: 7276- lar dystrophy locus via multiplex DNA merase chain reaction. BioTechniques 7280. amplification. Nucleic Acids Res. 16: 9: 276-281. 54. Longley, M.J., S.E. Bennett, and D.W. 11141-11156. 43. Simpson, H.D., T. Coolbear, M. Vermue, Mosbaugh. 1990. Characterization of the 68. Leung, D.W., E. Chen, and D.V. Goeddel. and R.M. Daniel. 1990. Purification and 5' to 3' exonuclease associated with Ther- 1989. A method for random mutagenesis some properties of a thermostable DNA mus aquaticus DNA polymerase. Nucleic of a defined DNA segment using a modi- polymerase from a Thermotoga species. Acids Res. 18: 7317-7322. fied polymerase chain reaction. Technique Biochem. Cell Biol. 68: 1292-1296. 55. R.D. Abramson, unpublished data. 1: 11-15. 44. Klimczak, L.J., F. Grummt, and K.J. 56. Brock, T.D. and H. Freeze. 1969. Thermus 69. Myers, T.W. and D.H. Gelfand. 1991. Re- Burger. 1985. Purification and character- aquaticus gen. n. and sp. n., a non-sporu- verse transcription and DNA amplifica- ization of DNA polymerase from the ar- lating extreme thermophile. J. Bacteriol. tion by a Thermus thermophilus DNA poly- chaebacterium Sulfolobus acidocaldarius. 98: 289-297. merase. Biochemistry 30: 7661-7666. Nucleic Acids Res. 13: 5269-5282. 57. Kong, H., R.B. Kucera, and W.E. Jack. 70. Huang, M-M., N. Arnheim, and M. Good- 45. Elie, C., S. Salhi, J.-M. Rossignol, P. Fort- 1993. Characterization of a DNA poly- man. 1992. Extension of base mispairs by erre, and A.-M. de Recondo. 1988. A DNA merase from the hyperthermophile Ar- Taq DNA polymerase: Implications for polymerase from a thermoacidophilic archaea Thermococcus litoralis. J. Biol. Chem. single nucleotide discrimination in PCR. chaebacterium: Evolutionary and techno- 268: 1965-1975. Nucleic Acids Res. 20: 4567-4573. logical interests. Biochem. Biophys. Acta 58. Delarue, M., O. Poch, N. Tordo, D. Moras, 71. Tada, M., M. Omata, S. Kawai, H. Saisho, 951: 261-267. and P. Argos. 1990. An attempt to unify M. Ohto, R.K. Saiki, and J.J. Sninsky, in 46. Salhi, S., C. Elie, P. Forterre, A.-M. de Re- the structure of polymerases. Prot. Eng. preparation. condo, and J.-M. Rossignol. 1989. DNA 3: 461-467. polymerase from Sulfolobus acidocaldar- 59. Blanco, L., A. Bernad, M.A. Blasco, and M. ius: Replication at high temperature of Salas. 1991. A general structure for DNA- Received January 29, 1993; accepted in long stretches of single-stranded DNA. J. dependent DNA polymerases. Gene revised form March 25, 1993. Mol. Biol. 209: 635-644. 1OO: 27-38. 47. Rella, R., C.A. Raia, F.M. Pisani, S. D'Auria, 60. Brutlag, D., M.R. Atkinson, P. Setlow, and R. Nucci, A Gambacorta, M. De Rosa, and A. Kornberg. 1969. An active fragment of M. Rossi. 1990. Purification and proper- DNA polymerase produced by proteolytic ties of a thermophilic and thermostable cleavage. Biochem. Biophys. Res. Commun. DNA polymerase from the archaebacte- 37: 982-989.

PCR Methods and Applications 287 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press

High-level expression, purification, and enzymatic characterization of full-length Thermus aquaticus DNA polymerase and a truncated form deficient in 5' to 3' exonuclease activity.

F C Lawyer, S Stoffel, R K Saiki, et al.

Genome Res. 1993 2: 275-287 Access the most recent version at doi:10.1101/gr.2.4.275

References This article cites 63 articles, 26 of which can be accessed free at: http://genome.cshlp.org/content/2/4/275.full.html#ref-list-1

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the Service top right corner of the article or click here.

To subscribe to Genome Research go to: https://genome.cshlp.org/subscriptions