CA1337405C - Purified thermostable enzyme - Google Patents

Abstract

A purified thermostable enzyme is obtained that has unique characteristics. Preferably the enzyme is isolated from the Thermus aquaticus species and has a molecular weight of about 86,000-90,000 daltons. The thermostable enzyme may be native or recombinant and may be used in a temperature-cycling chain reaction wherein at least one nucleic acid sequence is amplified in quantity from an existing sequence with the aid of selected primers and nucleotide triphosphates. The amplification process comprises treating separate complementary strands of the nucleic acid with a molar excess of two oligonucleotide primers, extending the primers with a thermostable enzyme to form complementary primer extension products which act as templates for synthesizing the desired nucleic acid sequence, and detecting the sequence so amplified. The steps of the reaction can be repeated as often as desired and involve temperature cycling to effect hybridization, promotion of activity of the enzyme, and denaturation of the hybrids formed. The enzyme is preferably stored in a buffer of non-ionic detergents that lends stability to the enzyme.

Description

337~05 `
This application is a divisional of application S.N. 542,406 filed 17 July 1987 from which there are two other divisionals as well. This application and supplementary disclosure have claLms directed to a purified th~rm~table enzyme and a method for puri~ying such enzyme only.

The present invention relates to a purified thermostable enzyme. In one embodiment the enzyme is DNA polymerase purified from Thenmus aquaticus and has a molecular weight of about 86,000-90,000.
The present invention relates to a process for amplifying existing nucleic acid sequences if they are present in a test sample and detecting the~ if present by using a probe. More specifically, it relates to a process for producing any particular nucleic acid sequence from a given sequence of DNA or RNA in amounts which are large compared to the amount initially present so as to facilitate detection of the sequences, using a thennostable enzyme to catalyze the reaction. The DNA or RNA may be single- or double-stranded, and may be a relatively pure species or a component of a mixture of nucleic acids. The process of the invention utilizes a repetitive reaction to accomplish the amplification of the desired nucleic acid sequence.
Extensive research has been conducted on the isolation of DNA polymerases from mesophilic microorganisms such as E. coli. See, for example, Bessman et al., J. Biol. Chem. (1957) 233:171-177 and Buttin and Kornberg (1966) J. Biol. Chem. 241:5419-5427.
In contrast, relatively little investigation has been made on the isolation and purification of DNA polymerases from thenmophiles, such as Thenmus aquaticus. Kaledin et al., Biokhymiya (1980) 45:644-651 discloses a six-step isolation and purification procedure of DNA polymerase from cells of T. aquaticus YT1 strain.
These steps involve isolation of crude extract, DEAE-cellulose ~ 30 chromatography, fractionation on hydroxyapatite, fractionation onDEAE-cellulose, and chromatography on single-strand DNA-cellulose.
The pools from each stage were not screened for contaminating endo-and exonuclease(s). The molecular weight of the purified enzyme is reported as 62,000 daltons per monomeric unit.
A second purification scheme for a polymerase from T.
aquaticus is described by A. Chien et al., J. Bacteriol. (1976) 127:1550-1557. In this process, the crude extract is applied to a DEAE-Sephadex* column. The dialyzed pooled fractions are then subjected to treatment on a phosphocellulose column. The pooled fractions are dialyzed and bovine serum albumin (BSA) is added to prevent loss of polymerase activity. The resulting mixture is loaded on a DNA-cellulose column. The pooléd material from the column is dialyzed and analyzed by gel filtration to have a molecular weight of about 63,000 daltons, and, by sucrose gradient centrifugation of about 68,000 daltons.
The use of a thermostable enzyme to amplify existing nucleic acid sequences in amounts that are large compared to the amount initially present has been sùggested in European Pat. Pub. No. 200,362 published December 10, 1986. Primers, nucleoside triphosphates, and a polymerase are used in the process, which involves denaturation, synthesis of template strands and hybridization. The extension product of each primer becomes a template for the production of the desired nucleic acid sequence. The application discloses that if the polymerase employed is a thermostable enzyme, it need not be added after every denaturation step, because the heat will not destroy its activity. No other advantages or details are provided on the use of a purified thermostable DNA
polymerase. The amplification and detection process is also described by Saiki et al., Science, 230:1350-1354 (1985), and by Saiki et al., Bio/Technolo~Y, 3:1008-1012 (1985).
Accordingly, there is a desire in the art to produce a purified, stable thermostable enzyme that may be used to improve the diagnostic amplification process described above.
Accordingly, the present invention provides a purified thermostable enzyme that catalyzes combination of nucleoside triphosphates to form a nucleic acid strand complementary to a nucleic acid template strand. Preferably the purified enzyme is DNA polymerase from Thermus aquaticus and has an estimated molecular weight of about 86,000-90,000 daltons, when compared with a phosphorylase B standard assigned a molecular weight of 92,500 daltons, or about 86,000-95,000 daltons, when compared with a phosphorylase B standard assigned a molecular weight of 97,400, the enzyme migrating on a denaturing polyacrylamide gel faster than phosphorylase B and moLe * Trade Mark _3_ 1 3 3 7 4 05 slowly than does bovine serum albumin. This purified material may be used in a temperature-cycling amplification reaction wherein nucleic acid sequences are produced from a given nucleic acid sequence in amounts that are large compared to the amount initially present so that they can be detected easily.
The gene encoding the enzyme from DNA polymerase from Thermus aquaticus has also been identified and provides yet another means to retrieve the thermostable enzyme of the present invention.
In addition to the gene encoding the approximately 86-000-90,000 dalton enzyme, gene derivatives encoding DNA polymerase activity are also presented.
Finally, the invention also encompasses a stable enzyme composition comprising a purified, thermostable enzyme as described above in a buffer containing one or more non-ionic polymeric detergents.
The purified enzyme, as well as the enzymes produced by recombinant DNA techniques, provide much more specificity than the Klenow fragment, which is not thermostable. In addition, the purified enzyme and the recombinantly produced eny mes exhibit the appropriate activity expected when dTTP or other nucleotide triphosphates are not present in the incubation mixture with the DNA template. Also, the enzymes herein have a broader pH profile than that of the thenmostable enzyme from Thermus aquaticus described in the literature, with more than 50g of the activity at pH 7 and at least 50% of the activity 2S at pH 6.4 as at pH 8. In addition, the thermostable enzyme herein can be stored in a buffer with non-ionic detergents so that it is stable, not losing activity over a period of time.
The present invention resides in a process for amplifying one or more specific nucleic acid sequences present in a nucleic acid or mixture thereof using primers and a thermostable enzyme. The extension product of one primer when hybridized to the other becomes a template for the production of the desired specific nucleic acid sequence, and vice versa, and the process is repeated as often as is necessary to produce the desired amount of the sequence. The method 4 l 3 3 7 4 0 5 herein improves the specificity of the amplification reaction, resulting in a very distinct signal of amplified nucleic acid. In addition, the method herein eliminates the need for transferring reagents from one vessel to another after each amplification cycle.
5 Such transferring is not required because the thermostable enzyme will withstand the high temperatures required to denature the nucleic acid strands and therefore does not need replacement. The temperature cycling may, in addition, be automated for further reduction in manpower and steps required to effectuate the amplification reaction.
More specifically, the present invention provides a process for amplifying at least one specific nucleic acid sequence contained in a nucleic acid or a mixture of nucleic acids, wherein if the nucleic acid is double-stranded, it consists of two separated complementary strands of equal or unequal length, which process comprises:
(a) contacting each nucleic acid strand with four different nucleotide triphosphates and one oligonucleotide primer for each different specific sequence being amplified, wherein each primer is selected to be substantially complementary to different strands of 20 each specific sequence, such that the extension product synthesized from one primer, when it is separated from its complement, can serve as a template for synthesis of the extension product of the other primer, said contacting being at a temperature which promotes hybridization of each primer to its complementary nucleic acid strand;
(b) contacting each nucleic acid strand, at the same time as or after step (a), with a thermostable enzyme which enables combination of the nucleotide triphosphates to form primer extension products complementary to each strand of each nucleic acid;
(c) maintaining the mixture from step (b) at an effective temperature for an effective time to activate the enzyme, and to synthesize, for each different sequence being amplified, an extension product of each primer which is complementary to each nucleic acid strand template, but not so high (a temperature) as to separate each extension product from its complementary strand template;

(d) heating the mixture from step (c) for an effective time and at an effective temperature to separate the primer extension products from the templates on which they were synthesized to produce single-stranded molecules, but not so high (a temperature) as to denature irreversibly the enzyme;
(e) cooling the mixture from step (d) at an effective temperature for an effective time to promote hybridization of each primer to each of the single-stranded molecules produced in step (d);
and (f) maintaining the mixture from step (e) at an effective temperature for an effective time to promote the activity of the enzyme and to synthesize, for each different sequence being amplified, an extension product of each primer which is complementary to each nucleic acid strand template produced in step (d), but not so high (a temperature) as to separate each extension product from its complementary strand template, wherein steps (e) and (f) are carried out simultaneously or sequentially.
The steps (d), (e) and (f) may be repeated until the desired level of sequence amplification is obtained. The preferred thermostable enzyme herein is a polymerase extracted from Thenmus aquaticus (Taq polymerase). Most preferably, if the enzyme is Taq polymerase, in step (a) the nucleic acid strands are contacted with a buffer comprising about 1.5-2 mM of a magnesium salt, 150-200 ~M each of the nucleotides, and 1 ~M of each primer, steps (a), (e) and (f) are carried out at about 45-58C, and step (d) is carried out at about 90-100C.
In a preferred embodiment, the nucleic acid(s) are double-stranded and step (a) is accomplished by (i) heating each nucleic acid in the presence of four different nucleotide triphosphates and one oligonucleotide primer for each different specific sequence being amplified, for an effective time and at an effective temperature to denature each nucleic acid, wherein each primer is selected to be substantially complementary to different strands of each specific sequence, such that the extension product synthesized from one primer, -6- l 3 3 7 4 ~ 5 when it is separated from its complement, can serve as a template for synthesis of the extension product of the other primer; and (ii) cooling the denatured nucleic acids to a temperature which promotes hybridization of each primer to its complementary nucleic acid strand.
In other embodiments the invention relates to a process for detecting the presence or absence of at least one specific nucleic acid sequence in a sample containing a nucleic acid or mixture of nucleic acids, or distinguishing between two different sequences in said sample, wherein the sample is suspected of containing said sequence or sequences, and wherein if the nucleic acid(s) are double-stranded, they each consist of two separated complementary strands of equal or unequal length, which process comprises steps (a) to (f) mentioned above, resulting in amplification in quantity of the specific nucleic acid sequence(s), if present;
(9) adding to the product of step (f) a labeled oligonucleotide probe, for each sequence being detected, capable of hybridizing to said sequence or to a mutation thereof; and (h) determining whether said hybridization has occurred.
In yet another embodiment, the invention relates to a process for detecting the presence or absence of at least one nucleotide variation in sequence in one or more nucleic acids contained in a sample, wherein if the nucleic acid is double-stranded it consists of two separated complementary strands of equal or unequal length, which process comprises steps (a)-(f) mentioned above, wherein steps (d), (e) and (f) are repeated a sufficient number of times to result in detectable amplification of the nucleic acid containing the sequence, if present;
(g) affixing the product of step (f) to a membrane;
(h) treating the membrane under hybridization conditions 30 with a labeled sequence-specific oligonucleotide probe capable of hybridizing with the amplified nucleic acid sequence only if a sequence of the probe is complementary to a region of the amplified sequence; and _ -7-(i) detecting whether the probe has hybridized to an amplified sequence in the nucleic acid sample.
If the sample comprises cells, preferably they are heated before step (a) to expose the nucleic acids therein to the reagents.
This step avoids extraction of the nucleic acids prior to reagent addition.
In a variation of this process, the primer(s) and/or nucleotide triphosphates are labeled so that the resulting amplified seguence is labeled. The labeled primer(s) and/or nucleotide triphosphate(s) can be present in the reaction mixture initially or added during a later cycle. The sequence-specific oligonucleotide (unlabeled) is affixed to a membrane and treated under hybridization conditions with the labeled amplification product so that hybridization will occur only if the membrane-bound sequence is present in the amplification product.
In yet another embodiment, the invention herein relates to a process for cloning into a cloning vector one or more specific nucleic acid sequences contained in a nucleic acid or a mixture of nucleic acids, which nucleic acid(s) when double-stranded consist of two separated complementary strands, and which nucleic acid(s) are amplified in quantity before cloning, which process comprises steps (a)-(f) mentioned above, with steps (d), (e) and (f) being repeated a sufficient number of times to result in detectable amplification of the nucleic acid(s) containing the sequence(s);
(9) adding to the product of step (f) a restriction enzyme for each of said restriction sites to obtain cleaved products in a restriction digest; and (h) ligating the cleaved product(s) of step (g) containing the specific sequence(s) to be cloned into one or more cloning vectors containing a promoter and a selectable marker.
In a final embodiment, the invention herein relates to a process for cloning into a cloning vector one or more specific nucleic acid sequences contained in a nucleic acid or mixture of nucleic acids, which nucleic acid(s), when double-stranded, consist of two separated complementary strands of equal or unequal length which nucleic acid(s) are amplified in quantity before cloning, which process comprises steps (a)-(f) mentioned above, with steps (d), (e) and (f) being repeated a sufficient number of times to result in effective amplification of -the nucleic acid(s) containing the sequence(s) for blunt-end ligation into one or more cloning vectors;
and (g) ligating the amplified specific sequence(s) to be cloned obtained from step (f) into one or more of said cloning vectors in the presence of a ligase, said amplified sequence(s) and vector(s) being present in sufficient amounts to effect the ligation.
In a product embodiment, the invention provides a composition of matter useful in amplifying at least one specific nucleic acid sequence contained in a nucleic acid or a mixture of nucleic acids, comprising four different nucleotide triphosphates and one oligonucleotide primer for each different specific sequence being amplified, wherein each primer is selected to be substantially complementary to different strands of each specific sequence, such that the extension product synthes;zed from one primer, when it is separated from its complement, can serve as a template for synthesis of the extension product of the other primer.
In another product embodiment, the invention provides a sample of one or more nucleic acids comprising multiple strands of a specific nucleic acid sequence contained in the nucleic acid(s). The sample may comprise about 10-100 of the strands, about 100-1000 of the strands, or over about 1000 of the strands.
In a final product embodiment, the invention provides an amplified nucleic acid sequence from a nucleic acid or mixture of nucleic acids comprising multiple copies of the sequence produced by the amplification processes herein.
Figure 1 is a restriction site map of plasmid pFC83 that contains the ~4.5 kb HindIII T. aquaticus DNA insert subcloned into plasmid BSM13+.

1 33740~
g Figure 2 is a restriction site map of plasmid pFC85 that contains the ~2.8 kb HindIII to ~ 718 T. aquaticus DNA insert subcloned into plasmid BSM13+.
As used herein, "cell", "cell line", and "cell culture" can be used interchangeably and all such designations include progeny.
Thus, the words "transformants" or "transformed cells" includes the primary subject cell and cultures derived therefrom without regard for the number of transfers. It is also understood that all progeny may not be precisely identical in DNA content, due to deliberate or inadvertent mutations. Mutant progeny that have the same functionality as screened for in the originally transformed cell are included.
The term "control sequences" refers to DNA sequences necessary for the expression of an operably linked coding sequence in a particular host organism. The control sequences that are suitable for procaryotes, for example, include a promoter, optionally an operator sequence, a ribosome binding site, and possibly, other as yet poorly understood sequences. Eucaryotic cells are known to utilize promoters, polyadenylation signals, and enhancers.
The term "expression system" refers to DNA sequences containing a desired coding sequence and control sequences in operable linkage, so that hosts transformed with these sequences are capable of producing the encoded proteins. In order to effect transformation, the expression system may be included on a vector; however, the relevant DNA may then also be integrated into the host chromosome.
The term "gene" as used herein refers to a DNA sequence that encodes a recoverable bioactive polypeptide or precursor. The polypeptide can be encoded by a full-length gene sequence or any portion of the coding sequence so long as the enzymatic activity is retained.
"Operably linked" refers to juxtaposition such that the normal function of the components can be performed. Thus, a coding sequence "operably linked" to control sequences refers to a configuration wherein the coding sequences can be expressed under the control of the sequences.

"Non-ionic polymeric detergents" refers to surface-active agents that have no ionic charge and that are characterized, for purposes of this invention, by their ability to stabilize the enzyme herein at a pH range of from about 3.5 to about 9.5, preferably from 4 to 8.5.
The term "oligonucleotide" as used herein is defined as a molecule comprised of two or more deoxyribonucleotides or ribonucleotides, preferably more than three. Its exact size will depend on many factors, which in turn depend on the ultimate function or use of the oligonucleotide. The oligonucleotide may be derived synthetically or by cloning.
The term "primer" as used herein refers to an oligonucleotide, whether occurring naturally as in a purified restriction digest or produced synthetically, which is capable of acting as a point of initiation of synthesis when placed under conditions in which synthesis of a primer extension product which is complementary to a nucleic acid strand is induced, i.e., in the presence of four different nucleotide triphosphates and thermostable enzyme in an appropriate buffer ("buffer" includes pH, ionic strength, cofactors, etc.) and at a suitable temperature. For Taq polymerase the buffer herein preferably contains 1.5-2 mM of a magnesium salt, preferably MgCl2, 150-200 ~M of each nucleotide, and 1 ~M of each primer, along with preferably 50 mM KCl, 10 mM Tris buffer, pH 8-8.4, and 100 ~g/ml gelatin.
The primer is preferably single-stranded for maximum efficiency in amplification, but may alternatively be double-stranded. If double-stranded, the primer is first treated to separate its strands before being used to prepare extension products.
Preferably, the primer is an oligodeoxyribonucleotide. The primer must be sufficiently long to prime the synthesis of extension products in the presence of the thermostable enzyme. The exact lengths of the primers will depend on many factors, including temperature, source of primer and use of the method. For example, depending on the complexity of the target sequence, the oligonucleotide primer 1 33740~

typically contains 15-25 nucleotides, although it may contain more or fewer nucleotides. Short primer molecules generally require cooler temperatures to form sufficiently stable hybrid complexes with template.
The primers herein are selected to be "substantially"
complementary to the different strands of each specific sequence to be amplified. This means that the primers must be sufficiently complementary to hybridize with their respective strands. Therefore, the primer sequence need not reflect the exact sequence of the template. For example, a non-complementary nucleotide fragment may be attached to the 5' end of the primer, with the remainder of the primer sequence being complementary to the strand. Alternatively, non-complementary bases or longer sequences can be interspersed into the primer, provided that the primer sequence has sufficient complementarity with the sequence of the strand to be amplified to hybridize therewith and thereby fonm a template for synthesis of the extension product of the other primer. However, for detection purposes, particulary using labeled sequence-specific probes, the primers typically have exact complementarity to obtain the best results.
As used herein, the terms "restriction endonucleases" and "restriction enzymes" refer to bacterial enzymes each of which cut double-stranded DNA at or near a specific nucleotide sequence.
As used herein, the tenm "DNA polymorphism" refers to the 25 condition in which two or more different nucleotide sequences can exist at a particular site in DNA.
As used herein, the tenm "nucleotide variation in sequence"
refers to any single or multiple nucleotide substitutions, deletions or insertions. These nucleotide variations may be mutant or 30 polymorphic allele variations. Therefore, the process herein can detect single nucleotide changes in nucleic acids such as occur in ~-globin genetic diseases caused by single-base mutations, additions or deletions (some ~-thalassemias, sickle cell anemia, hemoglobin C
disease, etc.), as well as multiple-base variations such as are involved with -thalassemia or some ~-thalassemias. In addition, the process herein can detect polymorphisms, which are not necessarily associated with a disease, but are merely a condition in which two or more different nucleotide sequences (whether having substituted, 5 deleted or inserted nucleotide base pairs) can exist at a particular site in the nucleic acid in the population, as with HLA regions of the human genome and random polymorphisms such as mitochondrial DNA. The polymorphic sequence-specific oligonucleotide probes described in detail hereinafter may be used to detect genetic markers linked to a disease such as insulin-dependent diabetis mellitus or in forensic applications. If the nucleic acid is double-stranded, the nucleotide variation in sequence becomes a base pair variation in sequence.
The term "sequence-specific oligonucleotides" refers to oligonucleotides which will hybridize to specific sequences whether or not contained on alleles, which sequences span the base pair variation being detected and are specific for the sequence variation being detected. Depending on the sequences being analyzed, one or more sequence-specific nucleotides may be employed for each sequence, as described further hereinbelow.
As used herein, the the term "restriction fragment length polymorphism" ("RFLP") refers to the differences among individuals in the lengths of restriction fragments formed by digestion with a particular restriction endonuclease.
As used herein, the term "thermostable enzyme" refers to an 25 enzyme which is stable to heat and is heat resistant and catalyzes (facilitates) combination of the nucleotides in the proper manner to form the primer extension products that are complementary to each nucleic acid strand. 6enerally, the synthesis will be initiated at the 3' end of each primer and will proceed in the 5' direction along the template strand, until synthesis tenrinates, producing molecules of different lengths. There may be a thermostable enzyme, however, which initiates synthesis at the 5' end and proceeds in the other direction, using the same process as described above.

_ -13- l 337405 The thermostable enzyme herein must satisfy a single criterion to be effective for the amplification reaction, i.e., the enzyme must not become irreversibly denatured (inactivated) when subjected to the elevated temperatures for the time necessary to effect denaturation of double-stranded nucleic acids. Irreversible denaturation for purposes herein refers to permanent and complete loss of enzymatic activity. The heating conditions necessary for denaturation will depend, e.g., on the buffer salt concentration and the length and nucleotide composition of the nucleic acids being denatured, but typically range from about 90 to about 105C for a time depending mainly on the temperature and the nucleic acid length, typically about 0.5 to four minutes. Higher temperatures may be tolerated as the buffer salt concentration and/or GC composition of the nucleic acid is increased. Preferably, the enzyme will not become irreversibly denatured at about 90-100C.
The thermostable enzyme herein preferably has an optimum temperature at which it functions that is higher than about 40C, which is the temperature below which hybridization of primer to template is promoted, although, depending on (1) magnesium and salt 20 concentrations and (2) composition and length of primer, hybridization can occur at higher temperature (e.g., 45-70C). The higher the temperature optimum for the enzyme, the greater the specificity and/or selectivity of the primer-directed extension process. However, enzymes that are active below 40C, e.g., at 37C, are also within the 25 scope of this invention provided they are heat-stable. Preferably, the optimum temperature ranges from about 50 to 90C, more preferably 60-80C .
The thermostable enzyme herein may be obtained from any source and may be a native or recombinant protein. Examples of enzymes that have been reported in the literature as being resistant to heat include heat-stable polymerases, such as, e.g., polymerases extracted from the thermophilic bacteria Thermus flavus, Thermus ruber, Thenmus thermophilus, Bacillus stearothermophilus (which has a somewhat lower temperature optimum than the others listed), Thermus aquaticus, Thermus lacteus, Thermus rubens, and Methanothermus fervidus.

The preferred thermostable enzyme herein is a DNA polymerase isolated from Thermus aquaticus. Various strains thereof are available from the American Type Culture Collection, Rockville, Maryland, and is described by T.D. Brock, J. Bact. (1969) 98:289-297, and by T. Oshima, Arch. Microbiol. (1978) 117: 189-196. One of these preferred strains is strain YT-1.
For recovering the native protein the cells are grown using any suitable technique. One such technique is described by Kaledin et al., Biokhimiya (1980), supra. Briefly, the cells are grown on a medium, in one liter, of nitrilotriacetic acid (100 mg), tryptone (3 9), yeast extract (3 9), succinic acid (5 9), sodium sulfite (50 mg), riboflavin (1 mg), K2HP04 (522 mg), MgS04(480 mg), CaCl2 (222 mg), NaCl (20 mg), and trace elements. The pH of the medium is adjusted to 8.0 ~ 0.2 with KOH. The yield is increased if cultivated with vigorous aeration up to 20 g/liter of cells at a temperature of 70C. Cells in the late logarithmic stage (determined by absorbance at 550 nm) are collected by centrifugation, washed with a buffer and stored frozen at -20C.
In another method for growing the cells, described in Chien et al., J. Bacteriol. (1976), supra, a defined mineral salts medium containing 0.3~ glutamic acid supplemented with 0.1 mg/l biotin, 0.1 mg/l thiamine, and 0.05 mg/l nicotinic acid is employed. The salts include nitrilotriacetic acid, CaS04, MgS04, NaCl, KN03, NaN03, ZnS04, H3B03, CuS04, NaMoO4, CoCl2, FeCl3, MnS04, and Na2HP04. The pH of the 25 medium is adjusted to 8.0 with NaOH.
In the Chien et al. technique, the cells are grown initially at 75C in a water bath shaker. On reaching a certain density, 1 liter of these cells is transferred to 16-liter carboys which are placed in hot-air incubators. Sterile air is bubbled through the 30 cultures and the temperature maintained at 75C. The cells are allowed to grow for 20 hours before being collected by centrifuge.
After cell growth, the isolation and purification of the enzyme take place in six stages, each of which is carried out at a temperature below room temperature, preferably about 4C.

-1~- 1 337405 In the first stage or step, the cells, if frozen, are thawed, disintegrated by ultrasound, suspended in a buffer at about pH
7.5, and centrifuged.
In the second stage, the supernatant is collected and then fractionated by adding a salt such as dry ammonium sulfate. The appropriate fraction (typically 45-75~ of saturation) is collected, dissolved in a 0.2 potassium phosphate buffer preferably at pH 6.5, and dialyzed against the same buffer.
The third step removes nucleic acids and some protein. The fraction from the second stage is applied to a DEAE-cellulose column equilibrated with the same buffer as used above. Then the column is washed with the same buffer and the flow-through protein-containing fractions, determined by absorbance at 280 nm, are collected and dialyzed against a 10 mM potassium phosphate buffer, preferably with the same ingredients as the first buffer, but at a pH of 7.5.
In the fourth step, the fraction so collected is applied to a hydroxyapatite column equilibrated with the buffer used for dialysis in the third step. The column is then washed and the enzyme eluted with a linear gradient of a buffer such as 0.01 M to 0.5 M potassium phosphate buffer at pH 7.5 containing 10 mM 2-mercaptoethanol and 5%
glycerine. The pooled fractions containing thermostable enzyme (e.g., DNA polymerase) activity are dialyzed against the same buffer used for dialysis in the third step.
In the fifth stage, the dialyzed fraction is applied to a 25 DEAE-cellulose column, equilibrated with the buffer used for dialysis in the third step. The column is then washed and the enzyme eluted with a linear gradient of a buffer such as 0.01 to 0.6 M KCl in the buffer used for dialysis in the third step. Fractions with thermostable enzyme activity are then tested for contaminating 30 deoxyribonucleases (endo- and exonucleases) using any suitable procedure. For example, the endonuclease activity may be determined electrophoretically from the change in molecular weight of phage ~ DNA
or supercoiled plasmid DNA after incubation with an excess of DNA
polymerase. Similarly, exonuclease activity may be determined _- 1 3 3 7 4 05 electrophoretically from the change in molecular weight of DnA after treatment with a restriction enzyme that cleaves at several sites.
The fractions determined to have no deoxyribonuclease activity are pooled and dialyzed against the same buffer used in the third step.
In the sixth step, the pooled fractions are placed on a phosphocellulose column with a set bed volume. The column is washed and the enzyme eluted with a linear gradient of a buffer such as 0.01 to 0.4 M
KCl in a potassium phosphate buffer at pH 7.5. The pooled fractions having thermostable polymerase activity and no deoxyribonuclease activity are dialyzed against a buffer at pH 8Ø
The molecular weight of the dialyzed product may be determined by any technique, for example, by SDS PAGE using protein molecular weight markers. The molecular weight of one of the preferred enzymes herein, the DNA polymerase purified from Thermus aquaticus, is determined by the above 15 method to be about 86,000-90,000 daltons, when compared with a phosphorylase B standard assigned a molecular weight of 92,500 daltons, or about 86,000-95,000 daltons, when compared with a phosphorylase B standard assigned a molecular weight of 97,400, the enzyme migrating on a denaturing polyacrylamide gel faster than phosphorylase B and more slowly than does bovine serum albumin.
The thermostable enzyme of this invention may also be produced by recombinant DNA techniques, as the gene encoding this enzyme has been cloned from Thenmus aquaticus genomic DNA. The complete coding sequence for the Thermus aquaticus (Taq) polymerase can be derived from 25 bacteriophage CH35:Taq#4-2 on an approximately 3.5 kilobase (kb) B~lII-AsP718 (partial) restriction fragment contained within an ~18 kb genomic DNA insert fragment. This bacteriophage was deposited with the American Type Culture Collection (ATCC) on May 29, 1987 and has accession no. 40,336. Alternatively, the gene can be constructed by ligating an 30 ~750 base pair (bp) B~ HindIII restriction fragment isolated from plasmid pFC83 (ATCC 67,422 deposited May 29, 1987) to an ~2.8 kb HindIII-Asp718 restriction fragment isolated from plasmid pFC85 (ATCC
67,421 deposited May 29, 1987). The pFC83 restriction fragment comprises the amino-terminus of the Taq polymerase gene while the restriction fragment from pFC85 comprises the carboxyl-terminus. Thus, ligation of these two fragments into a correspondingly digested vector with appropriate , 1 3 3 7 4 0 ~

control sequences will result in the translation of a full-length Taq polymerase.
It has also been found that the entire coding sequence of the Taq polymerase gene is not required to recover a biologically active gene product with the desired enzymatic activity. Amino-terminal deletions wherein approximately one-third of the coding sequence is absent has resulted in producing a gene product that is quire active in polymerase assays.
In addition to the N-terminal deletions, individual amino acid residues in the peptide chain comprising Taq polymerase may be modified by oxidation, reduction, or other derivatization, and the protein may be cleaved to obtain fragments that retain activity. Such alterations that do not destroy activity do not remove the DNA sequence encoding such protein from the definition of gene.
lS Thus, modifications to the primary structure itself by deletion, addition, or alternation of the amino acids incorporated into the sequence during translation can be made without destroying the activity of the protein. Such substitutions or other alternations result in proteins having an amino acid sequence encoded by DNA falling within the contemplated scope of the present invention.
Polyclonal antiserum from rabbits immunized with the purified estimated 86,000-90,000 or 86,000-95,000 daltons (as previously described) polymerase of this invention was used to probe a Thermus aquaticus partial genomic expression library to obtain the appropriate coding sequence as described below. The cloned genomic sequence can be expressed as a fusion polypeptide, expressed directly using its own control sequences, or expressed by constructions using control sequences appropriate to the particular host used for expression of the enzyme.
Of course, the availability of DNA encoding these sequences provides the opportunity to modify the codon sequence so as to generate mutein forms also having DNA polymerase activity.
Thus, these tools can provide the complete coding sequence for Taq DNA polymerase from which expression vectors applicable to a variety of host systems can be constructed and the coding sequence expressed. It is also evident from the foregoing that portions of the Taq polymerase-encoding sequence are useful as probes to retrieve other thermostable polymerase-encoding sequences in a variety of species. Accordingly, portions of the genomic DNA encoding at least six amino acids can be replicated in E. coli and the denatured forms used as probes or oligodeoxyribonucleotide probes can be synthesized which encode at least six amino acids and used to retrieve additional DNAs encoding a thermostable polymerase. Because there may not be a precisely exact match between the nucleotide sequence in the Thermus aquaticus form and that in the corresponding portion of other species, oligomers containing approximately 18 nucleotides (encoding the six amino acid stretch) are probably necessary to obtain hybridization under conditions of sufficient stringency to eliminate false positives. The sequences encoding six amino acids would supply information sufficient for such probes.

Suitable Hosts, Control Systems and Methods In general terms, the production of a recombinant form of Taq polymerase typically involves the following:
First, a DNA is obtained that encodes the mature (used here to include all muteins) enzyme or a fusion of the Taq polymerase to an additional sequence that does not destroy its activity or to an additional sequence cleavable under controlled conditions (such as treatment with peptidase) to give an active protein. If the sequence is uninterrupted by introns it is suitable for expression in any host. This sequence should be in an excisable and recoverable form.
The excised or recovered coding sequence is then preferably placed in operable linkage with suitable control sequences in a replicable expression vector. The vector is used to transform a 30 suitable host and the transformed host cultured under favorable conditions to effect the production of the recombinant Taq polymerase. Optionally the Taq polymerase is isolated from the medium or from the cells; recovery and purification of the protein may not be necessary in some instances, where some impurities may be tolerated.

`_ -19- 1 337405 Each of the foregoing steps can be done in a variety of ways. For example, the desired coding sequences may be obtained from genomic fragments and used directly in appropriate hosts. The constructions for expression vectors operable in a variety of hosts s are made using appropriate replicons and control sequences, as set forth below. Suitable restriction sites can, if not nonmally available, be added to the ends of the coding sequence so as to provide an excisable gene to insert into these vectors.
The control sequences, expression vectors, and transformation methods are dependent on the type of host cell used to express the gene. Generally, procaryotic, yeast, insect or mammalian cells are presently useful as hosts. Procaryotic hosts are in general the most efficient and convenient for the production of recombinant proteins and therefore preferred for the expression of Taq polymerase.
In the particular case of Taq polymerase, evidence indicates that considerable deletion at the N-terminus of the protein may occur under both recombinant and native conditions, and that the activity of the protein is still retained. It appears that the native proteins isolated may be the result of proteolytic degradation, and not translation of a truncated gene. The mutein produced from the truncated gene of plasmid pFC85 is, however, fully active in assays for DNA polymerase, as is that produced from DNA encoding the full-length sequence. Since it is clear that certain N-terminal shortened forms are active, the gene constructs used or expression of the 25 polymerase may also include the corresponding shortened forms of the coding sequence.

Control Sequences and Corresponding Hosts Procaryotes most frequently are represented by various strains of E. coli. However, other microbial strains may also be 30 used, such as bacilli, for example, Bacillus subtilis, various species of Pseudomonas, or other bacterial strains. In such procaryotic systems, plasmid vectors that contain replication sites and control sequences derived from a species compatible with the host are used.

For example, E. coli is typically transformed using derivatives of pBR322, a plasmid derived from an E. coli species by Bolivar, et al., Gene (1977) 2:95. pBR322 contains genes for ampicillin and tetra-cycline resistance, and thus provides additional markers that can be either retained or destroyed in constructing the desired vector.
Commonly used procaryotic control sequences, which are defined herein to include promoters for transcription initiation, optionally with an operator, along with ribosome binding site sequences, include such commonly used promoters as the ~-lactamase (penicillinase) and lactose (lac) promoter systems (Chang, et al., Nature (1977) 198:1056), the tryptophan (trp) promoter system (Goeddel, et al., Nucleic Acids Res.
(1980) 8:4057) and the lambda-derived PL promoter (Shimatake, et al., Nature (1981) 292:128) and N-gene ribosome binding site, which has been made useful as a portable control cassette, which comprises a first DNA sequence that is the PL promoter operably linked to a second DNA sequence corresponding to NRBS upstream of a third DNA sequence having at least one restriction site that permits cleavage within six bp 3' of the NRBS sequence. Also useful is the phosphatase A (phoA) system described by Chang, et al. in European Patent Publication No.
196,864 published October 8, 1986, assigned to the same assignee.
However, any available promoter system compatible with procaryotes can be used.
In addition to bacteria, eucaryotic microbes, such as yeast, may also be used as hosts. Laboratory strains of Saccharomyces cerevisiae, Baker's yeast, are most used, although a number of other strains are commonly available. While vectors employing the 2 micron origin of replication are illustrated (Broach, J. R., Meth. Enz.
(1983) 101:307), other plasmid vectors suitable for yeast expression are known (see, for example, Stinchcomb, et al., Nature (1979) 282:39, Tschempe, et al., Gene (1980) 10:157 and Clarke, L., et al., Meth.
Enz. (1983l 101: 300). Control sequences for yeast vectors include promoters for the synthesis of glycolytic enzymes (Hess, et al., J.
Adv. Enzyme Reg. (1968) 7:149; Holland, et al., Biotechnology (1978) 17:4900).

_ -21- l 3 3 7 4 0 5 Additional promoters known in the art include the promoter for 3-phosphoglycerate kinase (Hitzeman, et al., J. Biol. Chem. (1980) 255:2073), and those for other glycolytic enzymes, such as glyceraldehyde-3-phosphate dehydrogenase, hexokinase, pyruvate decarboxylase, phosphofructokinase, glucose-6-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, triosephosphate isomerase, phosphoglucose isomerase, and glucokinase. Other promoters that have the additional advantage of transcription controlled by growth conditions are the promoter regions for alcohol dehydrogenase 2, isocytochrome C, acid phosphatase, degradative enzymes associated with nitrogen metabolism, and enzymes responsible for maltose and galactose ultilization (Holland, supra).
It is also believed that tenminator sequences are desirable at the 3' end of the coding sequences. Such terminators are found in the 3' untranslated region following the coding sequences in yeast-derived genes. Many of the vectors illustrated contain control sequences derived from the enolase gene containing plasmid peno46 (Holland, M. J., et al., J. Biol. Chem. (1981) 256:1385) or the LEU2 gene obtained from YEpl3 (Broach, J., et al., Gene (1978) 8:121);
however, any vector containing a yeast-compatible promoter, origin of replication, and other control sequences is suitable.
It is also, of course, possible to express genes encoding polypeptides in eucaryotic host cell cultures derived from multicellular organisms. See, for example, Tissue Culture, Academic Press, Cruz and Patterson, editors (1973). Useful host cell lines include murine myelomas N51, VERO and HeLa cells, and Chinese hamster ovary (CHO) cells. Expression vectors for such cells ordinarily include promoters and control sequences compatible with mammalian cells such as, for example, the commonly used early and late promoters from Simian Virus 40 (SV 40) (Fiers, et al., Nature (1978) 273:113), or other viral promoters such as those derived from polyoma, Adenovirus 2, bovine papiloma virus, or avian sarcoma viruses, or immunoglobulin promoters and heat shock promoters. A system for expressing DNA in mammalian systems using the BPV as a vector is 35 disclosed in U.S. Patent 4,419,446. A modification of this system is described in U.S. Patent 4,601,978. General aspects of mammalian cell 1 33740~
_ -22-host system transformations have been described by Axel, U.S. Patent No. 4,399,216. It now appears, also, that "enhancer" regions are important in optimizing expression; these are, generally, sequences found upstream of the promoter region. Origins of replication may be obtained, if needed, from viral sources. However, integration into the chromosome is a common mechanism for DNA replication in eucaryotes.
Plant cells are also now available as hosts, and control sequences compatible with plant cells such as the nopaline synthase promoter and polyadenylation signal sequences (Depicker, A., et al., J. Mol. Appl. Gen. (1982) 1:561) are available.
Recently, in addition, expression systems employing insect cells utilizing the control systems provided by baculovirus vectors have been described (Miller, D. W., et al., in Genetic Engineering (1986) Setlow, J. K. et al., eds., Plenum Publishing, Vol. 8, pp. 277-297). These systems are also successful in producing Taq polymerase.

Transformations Depending on the host cell used, transformation is done using standard techniques appropriate to such cells. The calcium 20 treatment employing calcium chloride, as described by Cohen, S. N., Proc. Natl. Acad. Sci. (USA) (1972) 69:2110 is used for procaryotes or other cells that contain substantial cell wall barriers. Infection with Agrobacterium tumefaciens (Shaw, C. H., et al., Gene (1983) 23:315) is used for certain plant cells. ~or mammalian cells without 25 such cell walls, the calcium phosphate precipitation method of Graham and van der Eb, Virology (1978) 52:546 is preferred. Transformations into yeast are carried out according to the method of Van Solingen, P., et al., J. Bact. (1977) 130:946 and Hsiao, C. L., et al., Proc.
Natl. Acad. Sci. (USA) (1979) 76:3829.

Construction of a ~gtll Expression Library The strategy for isolating DNA encoding desired proteins, such as the Taq polymerase encoding DNA, using the bacteriophage 1 33740~

vector lambda gtll, is as follows. A library can be constructed of EcoRI-flanked AluI fragments, generated by complete digestion of Thenmus aquaticus DNA, inserted at the EcoRI site in the lambda gtll phage (Young and Davis, Proc. Natl. Acad. Sci USA (1983) 80:1194-1198). Because the unique EcoRI site in this bacteriophage is locatedin the carboxyl-terminus of the R-galactosidase gene, inserted DNA (in the appropriate frame and orientation) is expressed as protein fused with ~-galactosidase under the control of the lactose operon promter/operator.
Genomic expression libraries are then screened using the antibody plaque hybridization procedure. A modification of this procedure, referred to as "epitope selection," uses antiserum against the fusion protein sequence encoded by the phage, to confirm the identification of hybridized plaques. Thus, this library of 15 recombinant phages could be screened with antibodies that recognize the 86,000-90,000 dalton Taq polymerase in order to identify phage that carry DNA segments encoding the antigenic determinants of this protein.
Approximately 2 x 105 recombinant phage are screened using 20 total rabbit Taq polymerase antiserum. In this primary screen, positive signals are detected and one or more of these plaques are purified from candidate plaques which failed to react with preimmune serum and reacted with immune serum and analyzed in some detail. To examine the fusion proteins produced by the recombinant phage, 25 lysogens of the phage in the host Y1089 are produced. Upon induction of the lysogens and gel electrophoresis of the resulting proteins, each lysogen may be observed to produce a new protein, not found in the other lysogens, or duplicate sequences may result. Phage containing positive signals are picked; in this case, one positive plaque was picked for further identification and replated at lower densities to purify recombinants and the purified clones were analyzed by size class via digestion with EcoRI restriction en y me. Probes can then be made of the isolated DNA insert sequences and labeled appropriately and these probes can be used in conventional colony or plaque hybridization assays described in Maniatis et al., Molecular Cloning: A Laboratory Manual (1982).

-24- l 3 3 7 4 0 5 The labeled probe was used to probe a second genomic library constructed in a Charon 35 bacteriophage (Wilhelmine, A. M. et al., Gene (1983) 26:171-179). This library was made from Sau3A partial digestions of genomic Thermus aquaticus DNA and size fractionated fragments tl5-20 kb) were cloned into the BamHI site of the Charon 35 phage. The probe was used to isolate phage containing DNA encoding the Taq polymerase. One of the resulting phage, designated CH35:Taq#4-2, was found to contain the entire gene sequence. Partial sequences encoding portions of the gene were also isolated.

Vector Construction Construction of suitable vectors containing the desired coding and control sequences employs standard ligation and restriction techniques that are well understood in the art. Isolated plasmids, DNA sequences, or synthesized oligonucleotides are cleaved, tailored, and religated in the form desired.
Site-specific DNA cleavage is performed- by treating with the suitable restriction enzyme (or en y mes) under conditions that are generally understood in the art, and the particulars of which are specified by the manufacturer of these commercially available restriction enzymes. See, e.g., New England Biolabs, Product Catalog. In general, about 1 ~9 of plasmid or DNA sequence is cleaved by one unit of enzyme in about 20 ~l of buffer solution; in the examples herein, typically an excess of restriction enzyme is used to ensure complete digestion of the DNA substrate. Incubation times of about one hour to two hours at about 37C are workable, although variations can be tolerated. After each incubation, protein is removed by extraction with phenol/chlorofonm, and may be followed by ether extraction, and the nucleic acid recovered from aqueous fractions by precipitation with ethanol. If desired, size separation 30 of the cleaved fragments may be performed by polyacrylamide gel or agarose gel electrophoresis using standard techniques. A general description of size separations is found in Methods in Enzymology (1980) 65.499-560.

Restriction-cleaved fragments may be blunt-ended by treating with the large fragment of E. coli DNA polymerase I (Klenow) in the presence of the four deoxynucleotide triphosphates (dNTPs) using incubation times of about 15 to 25 minutes at 20 to 25C in 50 mM Tris pH 7.6, 50 mM NaCl, 10 mM MgCl2, 10 mM DTT and 50-100 ~M dNTPs. The Klenow fragment fills in at 5' sticky ends, but chews back protruding 3' single strands, even though the four dNTPs are present. If desired, selective repair can be performed by supplying only one of the, or selected, dNTPs within the limitations dictated by the nature of the sticky ends. After treatment with Klenow, the mixture is extracted with phenol/chloroform and ethanol precipitated. Treatment under appropriate conditions with S1 nuclease results in hydrolysis of any single-stranded portion.
Synthetic oligonucleotides may be prepared using the triester method of Matteucci, et al., (J. Am. Chem. Soc. (1981) 103:3185-3191) or using automated synthesis methods. Kinasing of single strands prior to annealing or for labeling is achieved using an excess, e.g., approximately 10 units of polynucleotide kinase to 1 nM
substrate in the presence of 50 mM Tris, pH 7.6, 10 mM MgCl2, 5 mM
dithiothreitol, 1-2 mM ATP. If kinasing is for labeling of probe, the ATP will contain high specific activity y_32p~
Ligations are performed in 15-30 ~l volumes under the following standard conditions and temperatures: 20 mM Tris-Cl pH 7.5, 10 mM MgCl2, 10 mM DTT, 33 ~g/ml 8SA, 10 mM-50 mM NaCl, and either 40 25 ~M ATP, 0.01-0.02 (Weiss) units T4 DNA ligase at 0C (for "sticky end"
ligation) or 1 mM ATP, 0.3-0.6 (Weiss) units T4 DNA ligase at 14C
(for "blunt end" ligation). Intermolecular "sticky end" ligations are usually performed at 33-100 ~g/ml total DNA concentrations (5-100 nM
total end concentration). Intermolecular blunt end ligations (usually 30 employing a 10-30 fold molar excess of linkers) are performed at 1 ~M
total ends concentration.
In vector construction employing "vector fragments", the vector fragment is commonly treated with bacterial alkaline phosphatase (BAP) in order to remove the 5' phosphate and prevent ~ -26- l 3 3 7 4 0 5 religation of the vector. BAP digestions are conducted at pH 8 in approximately 150 mM Tris, in the presence of Na+ and Mg+2 using about 1 unit of BAP per mg of vector at 60C for about one hour. In order to recover the nucleic acid fragments, the preparation is extracted with phenol/chloroform and ethanol precipitated. Alternatively, religation can be prevented in vectors that have been double digested by additional restriction enzyme digestion of the unwanted fragments.

Modification of DNA Sequences For portions of vectors derived from cDNA or genomic DNA
that require sequence modifications, site-specific primer-directed mutagenesis is used. This technique is now standard in the art, and is conducted using a primer synthetic oligonucleotide co~plementary to a single-stranded phage DNA to be mutagenized except for limited mismatching, representing the desired mutation. Briefly, the 15 synthetic oligonucleotide is used as a primer to direct synthesis of a strand complementary to the phage, and the resulting double-stranded DNA is transformed into a phage-supporting host bacterium. Cultures of the transformed bacteria are plated in top agar, permitting plaque formation from single cells that harbor the phage.
Theoretically, 50% of the new plaques will contain the phage having, as a single strand, the mutated form; 50% will have the original sequence. The plaques are transferred to nitrocellulose filters and the "lifts" hybridized with kinased synthetic primer at a temperature that permits hybridization of an exact match, but at which 25 the mismatches with the original strand are sufficient to prevent hybridization. Plaques that hybridize with the probe are then picked and cultured, and the DNA is recovered.

Verification of Construction In the constructions set forth below, correct ligations for plasmid construction are confirmed by first transforming E. coli strain MM294, or other suitable host, with the ligation mixture.
Successful transformants are selected by ampicillin, tetracycline or -27- l 3 3 7 4 o 5 other antibiotic resistance or using other markers, depending on the mode of plasmid construction, as is understood in the art. Plasmids from the transformants are then prepared according to the method of Clewell, D.B., et al., Proc. Natl. Acad. Sci. (USA) (1969) 62.1159, optionally following chloramphenicol amplification (Clewell, D.B., J.
Bacteriol. (1972) 110: 667). The isolated DNA is analyzed by restriction and/or sequenced by the dideoxy method of Sanger, F., et al., Proc. Natl. Acad. Sci. (USA) (1977) 74:5463 as further described by Messing, et al., Nucleic Acids Res. (1981) 9:309, or by the method of Maxam, et al., Methods in Enzymology (1980) 65:499.

Host Exemplified Host strains used in cloning and expression herein are as follows:
For cloning and sequencing, and for expression of constructions under control of most bacterial promoters, E. coli strain MM294 obtained from E. coli Genetic Stock Center GCSC #6135, was used as the host. For expression under control of the PLNpBS
promoter, E. coli strain K12 MC1000 lambda lysogen, N7N53cI857 SusP80, ATCC 39531 may be used. Used herein is E. coli DG116, which was deposited with ATCC (ATCC 53606) on April 7, 1987.
For M13 phage recombinants, E. coli strains susceptible to phage infection, such as E. coli K12 strain DG98, are employed. The DG98 strain has been deposited with ATCC July 13, 1984 and has accession number 39768.
Mammalian expression can be accomplished in COS-7 COS-A2, CV-1, and murine cells, and insect cell based expression in Spodoptera frugipeida).

Stabilization of Enzyme Activity For long-term stability, the enzyme herein must be stored in 30 a buffer that contains one or more non-ionic polymeric detergents.
Such detergents are generally those that have a molecular weight in the range of approximately 100 to 250,000, preferably about 4,000 to ~ 1 337405 200, 000 dal tons and stabi 1 i ze the enzyme at a pH of from about 3.5 to about 9.5 preferably from about 4 to 8 5 Examples of such detergents include those specified on pages 295-298 of McCutcheon's Emulsifiers & Detergents North American edition (1983), published by the McCutcheon Division of MC Publishing Co., 175 Rock Road, Glen Rock, NJ;(USA). Preferably, the detergents are selected from the-group comprising ethoxylated fatty alcohol ethers and lauryl ethers, ethoxylated alkyl phenols, octylphenoxy polyethoxy ethanol compounds, modified oxyethylated and/or oxypropylated straight-chain alcohols, polyethylene glycol monooleate compounds, polysorbate compounds, and phenolic fatty alcohol ethers. More particularly preferred are Tween 20, from ICI Americas Inc., Wilmington, DE, which is a polyoxyethylated (20) sorbitan monolaurate, and Iconol~ NP-40, from BASF Wyandotte Corp. Parsippany, NJ, which is an ethoxylated alkyl phenol (nonyl~.
The thermostable enzyme of this invention may be used for any purpose in which such eny me is necessary or desirable. In 2 particularly preferred embodiment, the enzyme herein is employed in the amplification protocol set forth below.

Amplification Protocol The amplification protocol using the enzyme of the present invention may be the process for amplifying existing nucleic acid sequences that is disclosed and claimed in European Pat. Pub. No.
200,362, supra. Preferably, however, the enzyme herein is used fn the amplification process described below.
In general, the amplification process involves a chain reaction for producing, in exponential quantities relative to the number of reaction steps involved, at least one specific nucleic acid sequence given (a) that the ends of the required sequence are known in sufficient detail that oligonucleotides can be synthesized which will hybridize to them, and (b) that a small amount of the sequence is available to initiate the chain reaction. The product of the chain r-:i 29 l 3374Gs reaction will be a discrete nucleic acid duplex with termini corresponding to the ends of the specific primers employed.
Any nucleic acid sequence, in purified or nonpurified form, can be utilized as the starting nucleic acid(s), provided it contains or is suspected to contain the specific nucleic acid sequence desired. Thus, the process may employ, for example, DNA or RNA, including messenger RNA, which DNA or RNA may be single-stranded or double-stranded. In addition, a DNA-RNA hybrid which contains one strand of each may be utilized. A mixture of any of these nucleic acids may also be employed, or the nucleic acids produced from a previous amplification reaction herein using the same or different primers may be so utilized. The specific nucleic acid sequence to be amplified may be only a fraction of a larger molecule or can be present initially as a discrete molecule, so that the specific 15 sequence constitutes the entire nucleic acid.
It is not necessary that the sequence to be amplified be present initially in a pure form; it may be a minor fraction of a complex mixture, such as a portion of the ~-globin gene contained in whole human DNA (as exemplified in Saiki et al., Science, 230, 1530-20 1534 (1985)) or a portion of a nucleic acid sequence due to aparticular microorganism which organism might constitute only a very minor fraction of a particular biological sample. The starti/ng nucleic acid sequence may contain more than one desired specific nucleic acid sequence which may be the same or different. Therefore, 25 the amplification process is useful not only for producing large amounts of one specific nucleic acid sequence, but also for amplifying simultaneously more than one different specific nucleic acid sequence located on the same or different nucleic acid molecules.
The nucleic acid(s) may be obtained from any source, for example, from plasmids such as pBR322, from cloned DNA or RNA, or from natural DNA or RNA from any source, including bacteria, yeast, viruses, organelles, and higher organisms such as plants or animals.
DNA or RNA may be extracted from blood, tissue material such as chorionic villi, or amniotic cells by a variety of techniques such as 35 that described by Maniatis et al., supra, p. 280-281.

_30 l 3 3 7 4 0 5 If probes are used which are specific to a sequence being amplified and thereafter detected, the cells may be directly used without extraction of the nucleic acid if they are suspended in hypotonic buffer and heated to about 90-100C, until cell lysis and S dispersion of intracellular components occur, generally 1 to 15 minutes. After the heating step the amplification reagents may be added directly to the lysed cells.
Any specific nucleic acid sequence can be produced by the amplification process. It is only necessary that a sufficient number of bases at both ends of the sequence be known in sufficient detail so that two oligonucleotide primers can be prepared which will hybridize to different strands of the desired sequence and at relative positions along the sequence such that an extension product synthesized from one primer, when it is separated from its template (complement), can serve as a template for extension of the other primer into a nucleic acid sequence of defined length. The greater the knowledge about the bases at both ends of the sequence, the greater can be the specificity of the primers for the target nucleic acid sequence, and thus the greater the efficiency of the process.
It will be understood that the word "primer" as used hereinafter may refer to more than one primer, particularly in the case where there is some ambiguity in the information regarding the terminal sequence(s) of the fragment to be amplified. For instance, in the case where a nucleic acid sequence is inferred from protein sequence information, a collection of primers containing sequences representing all possible codon variations based on degeneracy of the genetic code will be used for each strand. One primer from this collection will be homologous with the end of the desired sequence to be amplified.
The oligonucleotide primers may be prepared using any suitable method, such as, for example, the phosphotriester and phosphodiester methods described above, or automated embodiments thereof. In one such automated embodiment, diethylphosphoramidites are used as starting materials and may be synthesized as described by _ -31-Beaucage et al., Tetrahedron Letters (1981), 22:1859-1862. One method for synthesizing oligonucleotides on a modified solid support is described in U.S. Patent No. 4,458,066. It is also possible to use a primer which has been isolated from a biological source (such as a restriction endonuclease digest).
The specific nucleic acid sequence is produced by using the nucleic acid containing that sequence as a template. The first step involves contacting each nucleic acid strand with four different nucleotide triphosphates and one oligonucleotide primer for each different nucleic acid sequence being amplified or detected. If the nucleic acids to be amplified or detected are DNA, then the nucleotide triphosphates are dATP, dCTP, dGTP and TTP.
The nucleic acid strands are used as a template for the synthesis of additional nucleic acid strands. This synthesis can be performed using any suitable method. Generally it occurs in a buffered aqueous solution, preferably at a pH of 7-9, most preferably about 8. Preferably, a molar excess (for cloned nucleic acid, usually about 1000:1 primer:template, and for genomic nucleic acid, usually about 106:1 primer:template) of the two oligonucleotide primers is added to the buffer containing the separated template strands. It is understood, however, that the amount of complementary strand may not be known if the process herein is used for diagnostic applications, so that the amount of primer relative to the amount of complementary strand cannot be determined with certainty. As a practical matter, however, the amount of primer added will generally be in molar excess over the amount of complementary strand (template) when the sequence to be amplified is contained in a mixture of complicated long-chain nucleic acid strands. A large molar excess is preferred to improve the efficiency of the process.
Preferably the concentration of nucleotide triphosphates is 150-200 ~M each in the buffer for amplification and MgCl2 is present in the buffer in an amount of 1.5-2 mM to increase the efficiency and specificity of the reaction.

The resulting solution is then treated according to whether the nucleic acids being amplified or detected are double or single-stranded. If the nucleic acids are single-stranded, then no denaturation step need by employed, and the reaction mixture is held at a temperature which promotes hybridization of the primer to its complementary- target (template) sequence. Such temperature is generally from about 35C to 65C or more, preferably about 37-60C
for an effective time, generally one-half to five minutes, preferably one-three minutes. Preferably, 45-58C is used for Taq polymerase and >15-mer primers to increase the specificity of primer hybridization.
Shorter primers need lower temperatures.
The complement to the original single-stranded nucleic acid may be synthesized by adding one or two oligonucleotide primers thereto. If an appropriate single primer is added, a primer extension product is synthesized in the presence of the primer, the thermostable enzyme and the nucleotide triphosphates. The product will be partially complementary to the single-stranded nucleic acid and will hybridize with the nucleic acid strand to form a duplex of strands of unequal length which may then be separated into single strands as described above to produce two single separated complementary strands. Alternatively, two appropriate primers may be added to the single-stranded nucleic acid and the reaction carried out.
If the nucleic acid contains two strands, it is necessary to separate the strands of the nucleic acid before it can be used as the template. This strand separation can be accomplished by any suitable denaturing method including physical, chemical or enzymatic means.
One preferred physical method of separating the strands of the nucleic acid involves heating the nucleic acid until it is completely (>99%) denatured. Typical heat denaturation involves temperatures ranging from about 90 to 105C for times generally ranging from about 0.5 to 5 minutes. Preferably the effective denaturing temperature is 90-100C
for 0.5 to 3 minutes. Strand separation may also be induced by an enzyme from the class of enzymes known as helicases or the enzyme RecA, which has helicase activity and in the presence of riboATP is 35 known to denature DNA. The reaction conditions suitable for separating the strands of nucleic acids with helicases are described by Kuhn Hoffmann-Berling, CSH-Quantitative Biology, 43:63 (1978), and techniques for using RecA are reviewed in C. Radding, Ann. Rev.
Genetics, 16:405-37 (1982). The denaturation produces two separated 5 complementary strands of equal or unequal length.
If the double-stranded nucleic acid is denatured by heat, the reaction mixture is allowed to cool to a temperature which promotes hybridization of each primer present to its complementary target (template) sequence. This temperature is usually from about 35C to 65C or more, depending on reagents, preferably 37-60C, maintained for an effective time, generally 0.5 to 5 minutes, and preferably 1-3 minutes. In practical terms, the temperature is simply lowered from about 95C to as low as 37C, preferably to about 45-58C
for Taq polymerase, and hybridization occurs at a temperature within 15 this range-Whether the nucleic acid is single- or double-stranded, the thermostable enzyme may be added at the denaturation step or when the temperature is being reduced to or is in the range for promoting hybridization. The reaction mixture is then heated to a temperature 20 at which the activity of the enzyme is promoted or optimized, i.e., a temperature sufficient to increase the activity of the enzyme in facilitating synthesis of the primer extension products from the hybridized primer and template. The temperature must actually be sufficient to synthesize an extension product of each primer which is 25 complementary to each nucleic acid template, but must not be so high as to denature each extension product from its complementary template (i.e., the temperature is generally less than about 80C-90C).
Depending mainly on the types of enzyme and nucleic acid(s) employed, the typical temperature effective for this synthesis reaction generally ranges from about 40 to 80C, preferably 50-75C.
The temperature more preferably ranges from about 65-75C when a DNA
polymerase from Thermus aquaticus is employed. The period of time required for this synthesis may range from about 0.5 to 40 minutes or more, depending mainly on the temperature, the length of the nucleic _ 34_ 1337405 acid, the enzyme and the complexity o~ the nucleic acid mixture, preferably one to three minutes. If the nucleic acid is longer, a longer time period is generally required. The presence of dimethylsulfoxide (DMS0) is not necessary or recommended because DMS0 was found to inhibit Taq polymerase enzyme activity.
The newly synthesized strand and its complementary nucleic acid strand form a double-stranded molecule which is used in the succeeding steps of the process. In the next step, the strands of the double-stranded molecule are separated by heat denaturation at a temperature effective to denatuTe the molecule, but not so high that the thermostable enzyme is completely and irreversibly denatured or inactivated. Depending mainly on the type of enzyme and the length of nucleic acid, this temperature generally ranges from about 90 to 105C, more preferably 90-100C, and the time for denaturation typically ranges from 0.5 to four minutes, depending mainly on the temperature and nucleic acid length.
After this time, the temperature is decreased to a level which promotes hybridization of ~he primer to its complementary single-stranded molecule (template) produced from the previous step.
Such temperature is described above.
After this hybridization step, or in lieu of (or concurrently with) the hybridization s~ep, the temperature is adjusted to a temperature that is effective to promote the activity of the thermostable enzyme to enable synthesis of a primer extension product 25 using as template the newly synthesized strand from the previous step. The temperature again must not be so high as to separate (denature) the extension product from its template, as previously described (usually from 40 to 80C for 0.5 to 40 minutes, preferably 50 to 70C for one-three minutes). Hybridization may occur during 30 this step, so that the previous step of cooling after denaturation is not required. In such a case, ~sing simultaneous steps, the preferred temperature range is 50-70C.
The heating and cooling steps of strand separation, hybridization, and extension product synthesis can be repeated as _ _35 l 3 3 7 4 0 5 often as needed to produce the desired quantity of the specific nucleic acid sequence, depending on the ultimate use. The only limitation is the amount of the primers, thermostable enzyme and nucleotide triphosphates present. Preferably, the steps are repeated at least twice. For use in detection, the number of cycles will depend, e.g., on the nature of the sample. For example, fewer cycles will be required if the sample being amplified is pure. If the sample is a complex mixture of nucleic acids, more cycles will be required to amplify the signal sufficiently for its detection. For general amplification and detection, preferably the process is repeated at least 20 times.
When labeled sequence-specific probes are employed as described below, preferably the steps are repeated at least five times. When human genomic DNA is employed with such probes, the process is repeated preferably 15-30 times to amplify the sequence sufficiently that a clearly detectable signal is produced, i.e., so that background noise does not interfere with detection.
As will be described in further detail below, the amount of the specific nucleic acid sequence produced will accumulate in an exponential fashion.
No additional nucleotides, primers, or thermostable enzyme need be added after the initial addition, provided that the enzyme has not become denatured or inactivated irreversibly, in which case it is necessary to replenish the enzyme after each denaturing step.
25 Addition of such materials at each step, however, will not adversely affect the reaction.
When it is desired to produce more than one specific nucleic acid sequence from the first nucleic acid or mixture of nucleic acids, the appropriate number of different oligonucleotide primers are 30 utilized. For example, if two different specific nucleic acid sequences are to be produced, four primers are utilized. Two of the primers are specific for one of the specific nucleic acid sequences and the other two primers are specific for the second specific nucleic acid sequence. In this manner, each of the two different specific 35 sequences can be produced exponentially by the present process.

-36- l 3 3 7 4 0 5 After the appropriate length of time has passed to produce the desired amount of the specific nucleic acid sequence, the reaction may be halted by inactivating the enzyme in any known manner (e.g., by adding EDTA, phenol, SDS or CHCl3) or by separating the components of the reaction.
The amplification process may be conducted continuously. In one embodiment of an automated process, the reaction mixture may be temperature cycled such that the temperature is programmed to be controlled at a certain level for a certain time.
One such instrument for this purpose is an automated machine for handling the amplification reaction of this invention. This instrument utilizes a liquid handling system under computer control to make liquid transfers of enzyme stored at a controlled temperature in a first receptacle into a second receptacle whose temperature is controlled by the computer to conform to a certain incubation profile. The second receptacle stores the nucleic acid sequence(s) to be amplified plus the nucleotide triphosphates and primers. The computer includes a user interface through which a user can enter process parameters that control the characteristics of the various steps in the amplification sequence such as the times and temperatures of incubation, the amount of enzyme to transfer, etc.
A preferred machine that may be employed utilizes temperature cycling without a liquid handling system because the en y me need not be transferred at every cycle. Such a machine consists of the following systems:
1. A heat-conducting container for holding a given number of tubes, preferably 500 ~l tubes, which contain the reaction mixture of nucleotide triphosphates, primers, nucleic acid sequences, and enzyme.
2. A means to heat, cool, and maintain the heat-conducting container above and below room temperature, which means has an input for receiving a control signal for controlling which of the temperatures at or to which the container is heated, cooled or maintained. (These may be Peltier heat pumps available from Materials Electronics Products Corporation in Trenton, NJ or a water heat exchanger.) 3. A computer means (e.g., a microprocessor controller), coupled to the input of said means, to generate the signals that control automatically the amplification sequence, the temperature levels, and the temperature ramping and timing.
In another embodiment, the enzyme used for the synthesis of primer extension products can be immobilized in a column. The other reaction components can be continuously circulated by a pump through the column and a heating coil in series. Thus, the nucleic acids produced can be repeatedly denatured without inactivating the enzyme.
The amplification protocol is demonstrated diagrammatically below, where double-stranded DNA containing the desired sequence [S]
comprised of complementary strands [S+3 and [S~] is utilized as the nucleic acid. During the first and each subsequent reaction cycle, extension of each oligonucleotide primer on the original template will produce one new ssDNA molecule product of indefinite length that terminates with only one of the primers. These products, hereafter referred to as "long products," will accumulate in a linear fashion;
that is, the amount present after any number of cycles will be proportional to the number of cycles.
The long products thu~s produced will act as templates for one or the other of the oligonucleotide primers during subsequent cycles and will produce molecules of the desired sequence [S+] or [S].
25 These molecules will also function as templates for one or the other of the oligonucleotide primers, producing further [S+] and [S~], and thus a chain reaction can be sustained that will result in the -~ accumulation of [S] at an exponential rate relative to the number of cycles.
By-products formed by oligonucleotide hybridizations other than those intended are not self-catalytic (except in rare instances) and thus accumulate at a linear rate.
The specific sequence to be amplified, [S], can be depicted diagrammatically as:

~ - 38- 1 3 3 7 4 0 ~

[S+] 5 ' MMMMMAXXXXXXXXXXCCCCCCCCCC 3 ' [S-] 3 ' 1111111111 YYYYYYYYYYGGGGGGGGGG 5 ' The appropriate oligonucleotide primers would be:
Primer 1: 3 ' CCCCCCCCCC 5 ' Primer 2: 5' AAMAAMM 3' so that if DNA containing [S]
.... zzzzzzzzzzzzzzzzAAAMAAAAAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzzzzzzz ....
....zzzzzzzzz777zz7zl l l l l l l l l I~YYYYYYYYYCCCCCCCCCCzzzzzzzzzzzzzzzz....
is separated into single strands and its single strands are hybridized to Primers 1 and 2, the following extension reactions can be catalyzed by a thenmostable polymerase in the presence of the four nucleotide triphosphates:
3' 5' extends < CCCCCCCCCC Primer 1 .... zzzzzzzzzzzzzzzzAAMAAAAAAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzzzzzzz ....original template strand+
original template strand~
....zzzzzzzzzzzzzzzzlIIIIIIIII~YYYYYYYYYCCCCCCCCCCzzzzzzzzzzzzzzzz....
Primer 2 MAMMMM > extends 5' 3' On denaturation of the two duplexes formed, the products are:
3' 5' ....ZZZZZZZZZZZZZZZZIIIIIIIIII~YYYYYYYYYCCCCCCCCCC
newly synthesized long product 1 25 5' 3' .... z777777z77777777AAAMMMAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzzzzzzz ....
original template strand+
3' 5' ....zz7777zzzzzzzzzzl l l l l l l l l I~YYYYYYYYYCCCCCCCCCCzzzzzzzzzzzzzzzz....
original template strand~
5' 3' MMMMAAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzzzzzzz ....
newly synthesized long product 2 _ 39 1 337405 If these four strands are allowed to rehybridize with Primers 1 and 2 in the next cycle, the thermostable polymerase will catalyze the following reactions:
Primer 2 5' MAMMAM ) extends to here 3'... zzzzzzzzzz7777zz771111111111~YYYYYYYYYCCCCCCCCCC 5' newly synthesized long product 1 extends < - CCCCCCCCCC 5' Primer 1 5'....zzzzzzz~zz77z7PAAAAAMMXXXXXXXXXXCCCCCCCCCCzzzzz7777z7zz7....3' original template strand+
Primer 2 5' AAAAAAAAAA ) extends 3'....zzzzzzzzzzzzzzzzzz1111111111YYYYYYYYYCCCCCCCCCCzzzzzzzzzz....5' original template strand~
extends to here ~ CCCCCCCCCC 5' Primer 1 5' MAAMAMAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzzzzzzz..3' newly synthesized long product 2 If the strands of the above four duplexes are separated, the following strands are found:
5' MAAMAAAAXXXXXXXXXX+CCCCCCCCC 3' newly synthesized [S ]
20 3'....zzzzzzzzzzzzzzzzzzzllllllllll~YYYYYYYYYCCCCCCCCCC 5' first cycle synthesized long product 1 3'....zzzzzzzzzzz7z7z727zll111111II)'YYYYYYYYYCCCCCCCCCC 5' newly synthesized long product 1 5'....7zz7z7777z77z7zz777AAAAMMAAXXXXXXXXXXCCCCCCCCCCzzzzzzzzz....3' 25 original template strand+
5' MMAMAl~AXXXXXXXXXXCCCCCCCCCC.7zz77zz777zzzzzz...3' newly synthesized long product 2 3'..z7zzz777777z7771111111111~'YYYYYYYYYCCCCCCCCCCzzzzzz7777zz7zzz...5' original template strand~
3' Illlll~ YYYYYYYYYCCCCCCCCCC S' newly synthesized ~S~]
5' MAAAMMAXXXXXXXXXXCCCCCCCCCCzzzzzzzzzzz7zzz3' first cycle synthesized long product 2 It is seen that each strand which terminates with the oligonucleotide sequence of one primer and the complementary sequence of the other is the specific nucleic acid sequence [S] that is desired to be produced.
The amount of original nucleic acid remains constant in the entire process, because it is not replicated. The amount of the long products increases linearly because they are produced only from the original nucleic acid. The amount of the specific sequence increases exponentially. Thus, the specific sequence will become the predominant species. This is illustrated in the following table, which indicates the relative amounts of the species theoretically present after n cycles, assuming 100~ efficiency at each cycle:

Number of Double Strands After 0 to n Cycles Long Specific Cycle Number Template ProductsSequence [S]

lO 10 1 10 1013 1 15 32,752 1 20 1,048,555 n 1 n (2n-n-1) When a single-stranded nucleic acid is utilized as the template, only 15 one long product is formed per cycle.
A sequence within a given sequence can be amplified after a given number of amplifications to obtain greater specificity of the reaction by adding after at least one cycle of amplification a set of primers that are complementary to internal sequences (that are not on 20 the ends) of the sequence to be amplified. Such primers may be added at any stage and will provide a shorter amplified fragment.
Alternatively, a longer fragment can be prepared by using primers with non-complementary ends but having some overlap with the primers previously utilized in the amplification.
The method herein may be utilized to clone a particular nucleic acid sequence for insertion into a suitable expression vector. The vector may then be used to transform an appropriate host organism to produce the gene product of the sequence by standard methods of recombinant DNA technology.
The amplification process herein may yield a mixture of nucleic acids, resulting from the original template nucleic acid, the expected target amplified products, and various background non-target products. The amplified product can also be a mixture if the original template DNA contains multiple target sequences, such as in a heterozygous diploid genome or when there is a family of related genes.
The primers herein may be modified to assist the rapid and specific cloning of the mixture of DNAs produced by the amplification reaction. In one such modification, a restriction site is contained in each of the primers or in the sequence to be amplified and cloned. Preferably, the same or different restriction sites are incorporated at the 5' ends of the primers to result in restriction sites at the two ends of the amplified product. When cut with the appropriate enzymes, the amplified product can then be easily inserted into plasmid or vial vectors and cloned. This cloning allows the analysis or expression of individual amplified products, not a mixture.
If the primers have restriction sites incorporated therein, the same restriction site can be used for both primers. The use, however, of different sites allows the insertion of the product into the vector with a specific orientation and suppresses multiple insertions as well as insertions arising from amplifications based on only one of the two primers. The specific orientation is useful when cloning into single-strand sequencing vectors, when single-strand hybridization probes are used, or when the cloned product is being expressed.
One method to prepare the primers is to choose a primer sequence which differs minimally from the target sequence. Regions in which each of the primers is to be located are screened for homology to restriction sites appropriate to the desired vector. For example, the target sequence "CAGTATCCGA..." differs by only one base from one containing a BamHI site. A primer sequence is chosen to match the target exactly at its 3' end, and to contain the altered sequence and restriction site near its 5' end (for example, "CAGgATCCGA...", where the lower case letter symbolizes a mismatch with the target sequence). This minimally altered sequence will not interfere with the ability of the primer to hybridize to the original target sequence and to initiate polymerization. After the first amplification cycle the primer is copied, becomes the target, and matches exactly with new primers.
After the amplification process, the products are cleaved with the appropriate restriction enzymes, the restriction digest is optionally separated from inhibitors of ligation such as the nucleotide triphosphates and salts by, for example, passing over a desalting column, or a molecular weight chromatography column, or through a membrane, and the digestion product(s) containing the amplified sequence to be cloned is/are inserted by ligation into a cloning vector such as bacteriophage M13. The cloning vector generally has a selectable marker and may optionally also have a promoter. The gene may then be sequenced and/or expressed, if it codes for a protein, using well known techniques. The gene may also be sequenced by adding an appropriate primer during the amplification process which primer is complementary to the desired portion which is to be sequenced. The primer will form an extension product, and the extent of amplification with such extension product will provide sequence information.
Another method for preparing the primers involves taking the 3' end of the primers from the target sequence and adding the desired restriction site(s) to the 5' end of the primer. For the above example, a H~ndIII site could be added to make the sequence "cgaagcttCAGTATCCGA...", where lower case letters are as described above. The added bases would not contribute to the hybridization in the first cycle of amplification, but would match in subsequent cycles. The final amplified products are then cut with restriction enzyme(s) and cloned and expressed as described above. The gene being amplified may be, for example, human beta-hemoglobin or the human HLA
DQ, DR or DP-a and -~ genes.
In an alternative, but less preferred and less efficient, method of cloning wherein blunt-end ligation is employed rather than sticky-end ligation (using restriction enzymes), the basic amplification procedure is employed without concern for restriction enzymes in the primers or sequence(s) to be cloned. The steps must be repeated sufficiently, however, to produce enough amplified sequence(s) to effect ligation. Blunt-end ligation requires greater concentrations of sequence(s) and cloning vector(s) to be present than sticky-end ligation. In addition, the ligation must take place in the presence of a ligase, such as T4 ligase, E. coli ligase and ligase.
Once the amplified product is obtained, the ligation procedure is a standard procedure using conditions well known to those skilled in the art.
The cloning method which does not involve blunt end ligation controls the orientation or multiplicity of insertion of the amplified product into the cloning vector.
In addition, the process herein can be used for ln vitro mutagenesis. The oligonucleotide primers need not be exactly complementary to the nucleic acid sequence which is being amplified.
It is only necessary that they be able to hybridize to the sequence sufficiently well to be extended by the thermostable enzyme. The product of an amplification reaction wherein the primers employed are not exactly complementary to the original template will contain the sequence of the primer rather than the template, thereby introducing an ln vitro mutation. In further cycles this mutation will be amplified with an undiminished efficiency because no further mispaired primings are required. The mutant thus produced may be inserted into an appropriate vector by standard molecular biological techniques and might confer mutant properties on this vector such as the potential for production of an altered protein.
The process of making an altered DNA sequence as described 30 above could be repeated on the altered DNA using different primers to induce further sequence changes. In this way a series of mutated sequences could gradually be produced wherein each new addition to the series could differ from the last in a minor way, but from the original DNA source sequence in an increasingly major way. In this manner changes could be made ultimately which were not feasible in a single step due to the inability of a very seriously mismatched primer to function.
In addition, the primer can contain as part of its sequence a non-complementary sequence, provided that a sufficient amount of the primer contains a sequence which is complementary to- the strand to be amplified. For example, a nucleotide sequence which is not complementary to the template sequence (such as, e.g., a promoter, linker, coding sequence, etc.) may be attached at the 5' end of one or both of the primers, and thereby appended to the product of the amplification process. After the extension primer is added, sufficient cycles are run to achieve the desired amount of new template containing the non-complementary nucleotide insert. This allows production of large quantities of the combined fragments in a relatively short period of time (e.g., two hours or less) using a simple technique.
The method herein may also be used to enable detection and/or characterization of specific nucleic acid sequences associated with infectious diseases, genetic disorders or cellular disorders such as cancer, e.g., oncogenes. Amplification is useful when the amount of nucleic acid available for analysis is very small, as, for example, in the prenatal diagnosis of sickle cell anemia using DNA obtained from fetal cells. Amplification is particularly useful if such an analysis is to be done on a small sample using non-radioactive detection techniques which may be inherently insensitive, or where radioactive techniques are being employed but where rapid detection is desirable.
For purposes of this invention genetic diseases may include specific deletions and/or mutations in genomic DNA from any organism, such as, e.g., sickle cell anemia, a-thalassemia, ~-thalassemia, and the like. Sickle cell anemia can be readily detected via oligomer restriction analysis as described in EP Patent Publication 164,054 published December 11, 198~, or via a RFLP-like analysis following amplification of the appropriate DNA sequence by the present method.

_ -46- l 3 3 7 4 0 5 ~-Thalassemia can be detected by the absence of a sequence, and ~-thalassemia can be detected by the presence of a polymorphic restriction site closely linked to a mutation that causes the disease.
All of these genetic diseases may be detected by amplifying the appropriate sequence and analyzing it by Southern blots without using radioactive probes. In such a process, for example, a small sample of DNA from, e.g., amniotic fluid containing a very low level of the desired sequence is amplified, cut with a restriction enzyme, and analyzed via a Southern blotting technique. The use of non-radioactive probes is facilitated by the high level of the amplifiedsignal.
In another embodiment, a small sample of DNA may be amplified to a convenient level and then a further cycle of extension reactions perfonmed wherein nucleotide derivatives which are readily 15 detectable (such as 32p labeled or biotin-labeled nucleotide triphosphates) are incorporated directly into the final DNA product, which may be analyzed by restriction and electrophoretic separation or any other appropriate method.
In a further embodiment, the nucleic acid may be exposed to 20 a particular restriction endonuclease prior to amplification. Since a sequence which has been cut cannot be amplified, the appearance of an amplified fragment, despite prior restriction of the DNA sample, implies the absence of a site for the endonuclease within the amplified sequence. The presence or absence of an amplified sequence 25 can be detected by an appropriate method.
A practical application of this technique can be illustrated by its use in facilitating the detection of sickle cell anemia via the oligomer restriction technique described herein and in EP 164,054, supra, and Saiki et al., Bio/Technology, 3, pp 1008-1012 (1985).
30 Sickle cell anemia is a hemoglobin disease which is caused by a single base pair change in the sixth codon of the ~-globin gene.
The method of this invention may also be used-to detect directly single-base pair variations in nucleic acid sequence (such as genomic DNA) using sequence-specific oligonucleotides. In this ~ -47-method, the sequence variation, whether resulting from cancer, an infectious disease, or a genetic disease, e.g., a genetic lesion, is directly detected, eliminating the need for restriction digestion, electrophoresis, and gel manipulations otherwise required. The use of sequence-specific oligonucleotides in a dot blot format after amplification, as described herein, provides for improved specificity and sensitivity of the probe; an interpretable signal can be obtained with a 0.04 ~9 sample in six hours. Also, if the amount of sample spotted on a membrane is increased to 0.1-0.5 ~9, non-isotopically labeled oligonucleotides may be utilized rather than the radioactive probes used in previous methods. Furthermore, the process described hereinbelow is applicable to use of sequence-specific oligonucleotides less than 19-mers in size, thus allowing use of more discriminatory sequence-specific oligonucleotides.
Regarding genetic diseases, while RFLP requires a polymorphic restriction site to be associated with the disease, sequence-specific oligonucleotides directly detect the genetic lesion and are generally more useful for the analysis of such diseases as hemoglobin C disease, ~-1-antitrypsin and R-thalassemia, which result from single-base mutations. In addition, the oligonucleotides can be used to distinguish between genetic variants which represent different alleles (e.g., HLA typing), indicating the feasibility of a sequence-specific oligonucleotide-based HLA typing kit that includes a thermostable enzyme.
In one embodiment of the invention herein wherein a nucleotide variation in sequence is to be detected, the sample, amplified as described above using one primer for each strand of each nucleic acid suspected of containing the nucleotide variation, is spotted directly on a series of membranes and each membrane is 30 hybridized with a different labeled sequence-specific oligonucleotide probe. One procedure for spotting the sample on a membrane is descibed by Kafotos et al., Nucleic Acids Research, 7:1541-1552 (1979).

Briefly, the DNA sample affixed to the membrane may be pretreated with a prehybridization solution containing sodium dodecyl sulfate, Ficoll, serum albumin and various salts prior to the probe being added. Then, a labeled oligonucleotide probe which is specific to each sequence variation to be detected is added to a hybridization solution similar to the prehybridization solution. The hybridization solution is applied to the membrane and the membrane is subjected to hybridization conditions that will depend on the probe type and length, the type and concentration of ingredientsJ etc. Generally, hybridization is carriéd out at about 2~ to 75C, preferably 35 to 65C, for 0.25-50 hours, preferably less than three hours. The greater the stringency of conditions, the greater the required complementarity for hybridization between the probe and sample. If the background level is high, stringency may be increased 15 accordingly- The stringent conditions can also be incorporated in the wash.
After the hybridization, the sample is washed of unhybridized probe using any suitable means such as by washing one or more times with varying concentrations of standard saline phosphate 20 EDTA (SSPE) (180 mM NaCl, 10 mM NaHP04 and 1 M EDTA, pH 7.4) solutions at 25-75C for about 10 minutes to one hour, depending on the temperature. The label is then detected by using any appropriate detection technique.
The sequence-specific oligonucleotide employed herein is an 25 oligonucleotide which is generally prepared and selected as described above for preparing and selecting the primers. As described above, the sequence-specific oligonucleotide must encompass the region of the sequence which spans the nucleotide variation being detected and must be specific for the nucleotide variation being detected. For example, 30 if it is desired to detect whether a sample contains the mutation for sickle cell anemia, one oligonucleotide will be prepared which contains the nucleotide sequence site characteristic of the normal ~-globin gene and one oligonucleotide will be prepared which contains the nucleotide sequence characteristic of the sickle cell allele.
35 Each oligonucleotide would be hybridized to duplicates of the same sample to determine whether the sample contains the mutation.
* Trad~ Mark The polymorphic areas of HLA class II genes are localized to specific regions of the first exon and are flanked by conserved sequences, so that oligonucleotide primers complementary to opposite strands of the conserved 5' and 3' ends of the first exon can be prepared.
The number of oligonucleotides employed for detection of the polymorphic areas of the HLA class II genes will vary depending on the type of gene, which has regions of base pair variation which may be clustered or spread apart. If the regions are clustered, as in the case with HLA-DQ-, then one oligonucleotide is employed for each allele. If the regions are spread apart, as in the case with HLA-DQ-~and HLA-DR-~, then more than one probe, each encompassing an allelic variant, will be used for each allele.. In the case of HLA-DQ-~ and HLA-DR-~, three probes are employed for the three regions of the locus lS where allelic variations may occur. For detection of insulin-dependent diabetes mellitus (IDDM) four probes for the HLA-DR~ second exon are employed.
Haplotypes can be inferred from segregation analysis in families or, in some cases, by direct analysis of the individual DNA
sample. Specific allelic combinations (haplotypes) of sequence-specific oligonucleotide reactivities can be identified in heterozygous cells by using restriction enzyme digestion of the genomic DNA prior to amplification.
For example, if in DQ~ one finds three highly variable subregions A, B, and C within a single amplified region, and if there are six different sequences at each region (A1-6, B1-6, C1-6, then an individual could be typed in the DQR locus by sequence-specific oligonucleotide probe analysis as containing Al, A2; B2, B3; C1, C4, with the possible haplotype combinations of A1, B2, C1; A1, B2, C4;
A2, B2, Cl; A2, B2, C4; Al, B3, C1; Al, B3, C4; Al, B2, Cl; and Al, B2, C4.
If the genomic DNA is digested with a polymorphic restriction en y me prior to amplification, and if the enzyme cuts both alleles between the primers, there is no reactivity with the sequence-specific probes due to lack of amplification, and the result is uninformative. If the enzyme cuts neither allele, the probe results with digested and undigested genomic DNA are the same and the result is uninformative. If the enzyme cuts only one allele, however, then one can infer both haplotypes by comparing the probe reactivity patterns on digested and undigested DNA.
The haplotypes can be deduced by comparing sequence-specific oligonucleotide reactivities with uncut genomic DNA and genomic DNA
cut with one or several enzymes known to be polymorphic and to recognize sites between the primers.
The length of the sequence-specific oligonucleotide will depend on many factors, including the particular target molecule being detected, the source of oligonucleotide, and the nucleotide composition. For purposes herein, the sequence-specific oligonucleotide typically contains 15-25 nucleotides, although it may contain more or fewer nucleotides. While oligonucleotides which are at least 19 mers in length may enhance specificity and/or sensitivity, probes which are less than 19 mers, e.g., 16-mers, may show more sequence-specific discrimination presumably because a single mismatch is more destabilizing. Because amplification increases specificity so that a longer length is less critical, and hybridization and washing temperatures can be lower for the same salt concentration, it is preferred to use oligonucleotides which are less than 19 mers in length.
Where the sample is first placed on the membrane and then detected with the oligonucleotide, the oligonucleotide must be labeled with a suitable label moiety, which may be detected by spectroscopic, photochemical, biochemical, immunochemical or chemical means.
Immunochemical means include antibodies which are capable of forming a 30 complex with the oligonucleotide under suitable conditions, and biochemical means include polypeptides or lectins capable of forming a complex with the oligonucleotide under the appropriate conditions.
Examples include fluorescent dyes, electron-dense reagents, enzymes capable of depositing insoluble reaction products or being detected 35 chromogenically, such as horseradish peroxidase, alkaline phosphatase, ~ -51- l 3 3 7 4 0 5 a radioactive label such as 32p, or biotin. If biotin is employed, a spacer arm may be utilized to attach it to the oligonucleotide.
Preferably, the label moiety is horseradish peroxidase.
Alternatively, in one "reverse" dot blot format, at least one of the primers and/or at least one of the four nucleotide triphosphates is labeled with a detectable label, so that the resulting amplified sequence is labeled. These labeled moieties may be present initially in the reaction mixture or added during a later cycle of the amplification to introduce the label to the amplification product. Then an unlabeled sequence-specific oligonucleotide capable of hybridizing with the amplified nucleic acid sequence, if the sequence variation(s) (whether normal or mutant) is/are present, is spotted on (affixed to) the membrane under prehybridization conditions as described above. The amplified sample is then added to the pretreated membrane under hybridization conditions as described above. Finally, detection means are used to determine if an amplified sequence in the nucleic acid sample has hybridized to the oligonucleotide affixed to the membrane. Hybridization will occur only if the membrane-bound sequence containing the variation is 20 present in the amplification product, i.e., only if a sequence of the probe is complementary to a region of the amplified sequence.
In another version of the "reverse" dot blot format, the amplification is carried out without employing a label as with the "forward" dot blot format described above, and a labeled sequence-25 specific oligonucleotide probe capable of hybridizing with theamplified nucleic acid sequence containing the variation, if present, is spotted on (affixed to) the membrane under prehybridization conditions as described above. The amplified sample is then added to the pretreated membrane under hybridization conditions as described above. Then the labeled oligonucleotide or a fragment thereof is released from the membrane in such a way that a detection means can be used to determine if an amplified sequence in the sample hybridized to the labeled oligonucleotide. The release may take place, for example, by adding a restriction enzyme to the membrane which recognizes a 35 restriction site in the probe. This procedure, known as oligomer -52- l 3 3 7 4 0 5 restriction, is described more fully in EP Patent Publication 164,054 published December 11, 1985.
In both the forward and reverse dot blot methods, the genetic diseases which may be detected include specific deletions, insertions and/or substitutions in any base pair mutation or polymorphism in nucleic acids, for example, genomic DNA, from any organism. Examples of diseases in which base pair variation is known include sickle cell anemia, hemoglobin C disease, -thalassemia, ~-thalassemia, and the like. Other diseases that may be detected include cancerous diseases such as those involving the RAS oncogenes, e.g., the n-RAS oncogene, and infectious diseases.
A dot blot process may also be used for HLA typing in the areas of tissue transplantation, disease susceptibility, and paternity determination. The HLA class II genes, consisting of the and ~
genes from the HLA-DR, HLA-DQ and HLA-DP regions, are highly polymorphic; their genetic complexity at the DNA level is significantly greater than the polymorphism currently defined by serological typing. In addition, the process herein may be employed to detect four DNA sequences coding for HLA class II ~ proteins (e.g.
20 DR~) associated with insulin-dependent diabetes mellitus (IDDM).
Briefly, the four DNA sequences associated with IDDM are selected from the group consisting of:
1) 5'-GAGCTGCGTAAGTCTGAG-3', 2) 5'-GAGGAGTTCCTGCGCTTC-3', 2 5 3) 5'-CCTGTCGCCGAGTCCTGG-3', and 4) 5'-GACATCCTGGAAGACGAGAGA-3', or the DNA strands that are complementary thereto. Sequence-specific probes may be prepared that will hybridize to one or more of these sequences.
Various infectious diseases can be diagnosed by the presence in clinical samples of specific DNA sequences characteristic of the causative microorganism. These include bacteria, such as Salmonella, Chlamydia, Neisseria; viruses, such as the hepatitis viruses, and 1 33740~

parasites, such as the Plasmodium responsible for malaria. U.S.
Patent Reexamination Certificate B14,358,535 issued on May 13, 1986 to Falkow et al. describes the use of specific DNA hybridization probes for the diagnosis of infectious diseases. A relatively small number of pathogenic organisms may be present in a clinical sample from an infected patient and the DNA extracted from these may constitute only a very small fraction of the total DNA in the sample. Specific amplification of suspected sequences prior to immobilization and hybridization detection of the DNA samples could greatly improve the sensitivity and specificity of traditional procedures.
Routine clinical use of DNA probes for the diagnosis of infectious diseases would be simplified considerably if non-radioactively labeled probes could be employed as described in EP
63,879 to Ward. In this procedure biotin-containing DNA probes are detected by chromogenic enzymes linked to avidin or biotin-specific antibodies. This type of detection is convenient, but relatively insensitive. The combination of specific DNA amplification by the present method and the use of stably labeled probes could provide the convenience and sensitivity required to make the Falkow and Ward procedures useful in a routine clinical setting.
A specific use of the amplification technology for detecting or monitoring for the AIDS virus is described as follows. The amplification and detection process is used with primers and probes which are designed to amplify and detect, respectively, nucleic acid sequences which are substantially conserved among the nucleic acids in AIDS viruses and specific to the nucleic acids in AIDS viruses. Thus, the sequence to be detected must be sufficiently complementary to the nucleic acids in AIDS viruses to initiate polymerization, preferably at room temperature, in the presence of the enzyme and nucleotide triphosphates.
In addition, the probe may be a biotinylated probe in which the biotin is attached to a spacer arm of the formula:

_ 54 1337405 H
-Y-(CH2)2-0-[(CH2)xO]y-cH2cH2-N
where Y is 0, NH or N-CHO, x is a number from 1 to 4, and y is a number from 2 to 4. The spacer arm is in turn attached to a psoralen moiety of the formula:

O ~ CH3 The psoralen moiety intercalates into and crosslinks a "gapped circle"
probe as described by Courage-Tebbe et al., Biochim. Biophys. Acta, 6 (1982) 1-5, wherein the single-stranded hybridization region of the gapped circle spans the region contained in the primers. The details of this biotinylation and dot blot procedure are described more fully in commonly assigned U.S. Patent No. 4,582,789 issued April 15, 1986 and U.S. Patent No. 4,617,261 issued October 14, 1986.
The amplification process can also be utilized to produce sufficient quantities of DNA from a single copy human gene such that detection by a simple non-specific DNA stain such as ethidium bromide can be employed to diagnose DNA directly.
In addition to detecting infectious diseases and pathological abnormalities in the genome of organisms, the amplification process can also be used to detect DNA polymorphisms which may not be associated with any pathological state.
In summary, the amplification process is seen to provide a process for amplifying one or more specific nucleic acid sequences using a chain reaction and a thermostable enzyme, in which reaction primer extension products are produced which can subsequently act as templates for further primer extension reactions. The process is especially useful in detecting nucleic acid sequences which are initially present in only very small amounts.
The following examples are offered by way of illustration only and are by no means intended to limit the scope of the claimed _55 l 3 3 7 4 0 5 invention. In these examples, all percentages are by weight if for solids and by volume if for liquids, unless otherwise noted, and all temperatures are given in degrees Celsius.

EXAMPLE I
I. Synthesis of the Primers The following two oligonucleotide primers were prepared by the method described below:

5'-ACACAACTGTGTTCACTAGC-3' (PC03) 5'-CAACTTCATCCACGTTCACC-3' (PC04) These primers, both 20-mers, anneal to opposite strands of the genomic DNA with their 5' ends separated by a distance of 110 base pairs.
A. Automated Synthesis Procedures: The diethylphosphoramidites, synthesized according to Beaucage and Caruthers (Tetrahedron Letters (1981) 22:1859-1862) were sequentially condensed to a nucleoside derivatized controlled pore glass support.
The procedure included detritylation with trichloroacetic acid in dichloromethane, condensation using benzotriazole as activating proton donor, and capping with acetic anhydride and dimethylaminopyridine in tetrahydrofuran and pyridine. Cycle time was approximately 30 minutes. Yields at each step were essentially quantitative and were determined by collection and spectroscopic examination of the dimethoxytrityl alcohol released during detritylation.
B. Oligodeoxyribonucleotide Deprotection and Purification Procedures: The solid support was removed from the column and exposed to 1 ml concentrated ammonium hydroxide at room temperature for four hours in a closed tube. The support was then removed by filtration and the solution containing the partially protected oligodeoxynucleotide was brought to 55C for five hours. Ammonia was removed and the residue was applied to a preparative polyacrylamide gel. Electrophoresis was carried out at 30 volts/cm for 90 minutes after which the band containing the product was identified by UV
shadowing of a fluorescent plate. The band was excised and eluted with 1 ml distilled water overnight at 4C. This solution was applied -56- l 3 3 7 4 0 5 to a RP-HPLC column and eluted with a 7-13~ gradient of acetonitrile in 1~ ammonium acetate buffer at pH 6Ø The elution was monitored by UV absorbance at 260 nm and the appropriate fraction collected, quantitated by UV absorbance in a fixed volume and evaporated to dryness at room temperature in a vacuum centrifuge.
C. Characterization of Oligodeoxyribonucleotides: Test aliquots of the purified oligonucleotides were 32p labeled with polynucleotide kinase and y-32P-ATP. - The labeled compounds were examined by autoradiography of 14-20~ polyacrylamide gels after electrophoresis for 45 minutes at 50 volts/cm. This procedure verifies the molecular weight. Base composition was determined by digestion of the oligodeoxyribonucleotide to nucleosides by use of venom diesterase and bacterial alkaline phosphatase and subsequent separation and quantitation of the derived nucleosides using a reverse phase HPLC column and a 10~ acetonitrile, 1~ ammonium acetate mobile phase.

II. Isolation of Human Genomic DNA from Cell Line High molecular weight genomic DNA was isolated from a T cell line, Molt 4, homozygous for normal B-globin available from the Human Genetic Mutant Cell Depository, Camden, NJ as GM2219C using essentially the method of Maniatis et al., supra, p. 280-281.

III. Purification of a Polymerase From Thenmus aquaticus Thermus aquaticus strain YT1, available without restriction from the American Type Culture Collection, 12301 Parklawn Drive, Rockville, MD, as ATCC No. 25,104 was grown in flasks in the following medium:
Sodium Citrate 1 mM
Potassium Phosphate, pH 7.9 5 mM
Ammonium Chloride 10 ~l Magnesium Sulfate 0.2 mM
Calcium Chloride 0.1 mM
Sodium Chloride 1 g/l Yeast Extract 1 g/l Tryptone 1 g/l Glucose 2 g/l Ferrous Sulfate 0.01 mM
(The pH was adjusted to 8.0 prior to autoclaving.) - 1 3 3 7 4 G ~

A 10-liter fermentor was inoculated from a seed flask cultured overnight in the above medium at 70C. A total of 600 ml from the seed fl ask was used to inoculate 10 liters of the same medium. The pH was controlled at 8.0 with ammonium hydroxide with the 5 dissolved oxygen at 40~, with the temperature at 70C, and with the stirring rate at 400 rpm.
After growth of the cells, they were purified using the protocol (with slight modification) of Kaledin et al., supra, through the fi rst five stages and using a different protocol for the sixth 10 stage. All six steps were conducted at 4C. The rate of fractionation on columns was 0.5 columns/hour and the volumes of gradients during elution were 10 column volumes. An alternative and preferred purification protocol is provided in Example VI below.
Briefly, the above culture of the T. aquaticus cells was 15 harvested by centrifugation after nine hours of cultivation, in late log phase, at a cell density of 1.4 9 dry weight/l. Twenty grams of cells were resuspended in 80 ml of a buffer consisting of 50 mM
Tris HCl pH 7.5, 0.1 mM EDTA. Cells were lysed and the lysate was centrifuged for two hours at 35,000 rpm in a rotor at 4C. The 20 supernatant was collected (fraction A) and the protein fraction precipitating between 45 and 75% saturation of ammonium sul fate was coll ected, dissol ved i n a buffer consisting of 0.2 M potassium phosphate buffer, pH 6.5, 10 mM 2-mercaptoethanol, and 5% glycerine, and finally dialyzed against the same buffer to yield fraction B.
Fraction B was applied to a 2.2 x 30-cm column of DEAE-cellulose, equilibrated with the above described buffer. The column was then washed with the same buffer and the fractions containing protein (determined by absorbance at 280 nm) were collected. The combined protein fraction was dialyzed against a second buffer, 30 containing 0.01 M potassium phosphate buffer, pH 7.5, 10 mM 2-mercaptoethanol, and 5% glycerine, to yield fraction C.
Fraction C was applied to a 2.6 x 21-cm column of hydroxyapatite, equilibrated with a second buffer. The column was then washed and the enzyme was eluted with a linear gradient of 0.01--58- l 3 3 7 4 0 5 0.5 M potassium phosphate buffer, pH 7.5, containing 10 mM 2-mercaptoethanol and 5% glycerine. Fractions containing DNA polymerase activity (90-180 mM potassium phosphate) were combined, concentrated four-fold using an Amicon stirred cell and YM10 membrane, and dialyzed against the second buffer to yield fraction D.
Fraction D was applied to a 1.6 x 28-cm column of DEAE-cellulose, equilibrated with the second buffer. The column was washed and the polymerase was eluted with a linear gradient of 0.01-0.5 M
potassium phosphate in the second buffer. The fractions were assayed for contaminating endonuclease(s) and exonuclease(s) by electrophoretically detecting the change in molecular weight of phage A DNA or supercoiled plasmid DNA after incubation with an excess of DNA polymerase (for endonuclease) and after treatment with a restriction enzyme that cleaves the DNA into several fragments (for exonuclease). Only those DNA polymerase fractions (65-95 mM potassium phosphate) having minimal nuclease contamination were pooled. To the pool was added autoclaved gelatin in an amount of 250 ~g/ml, and dialysis was conducted against the second buffer to yield Fraction E.
Fraction E was applied to a phosphocellulose column and 20 eluted with a 100 ml gradient (0.01-0.4 M KCl gradient in 20 mM
potassium phosphate buffer pH 7.5). The fractions were assayed for contaminating endo/exonuclease(s) as described above as well as for polymerase activity (by the method of Kaledin et al.) and then pooled. The pooled fractions were dialyzed against the second buffer, 25 then concentrated by dialysis against 50% glycerine and the second buffer.
The molecular weight of the polymerase was determined by SDS
PAGE. Marker proteins were phosphorylase B (92,500), bovine serum albumin (66,200), ovalbumin (45,000), carbonic anhydrase (31,000), 30 soybean trypsin inhibitor (21,500), and lysozyme (14,400).
Preliminary data suggest that the polymerase has a molecular weight of about 86,000-90,000 daltons, not 62,000-63,000 daltons reported in the literature (e.g., by Kaledin et al.).

_59_ l 3 3 7 4 0 5 The polymerase was incubated in 50 ~l of a mixture containing 25 mM Tris-HCl pH 6.4 and pH 8.0, 0.1 M KCl, lO mM MgC12, 1 mM 2-mercaptoethanol, 10 nmoles each of dGTP, dATP, and TTP, and 0.5 ~Ci (3H) dCTP, 8 ~9 "activated" calf thymus DNA, and 0.5-5 units of the polymerase. "Activated" DNA is a native preparation of DNA after partial hydrolysis with DNase I until 5% of the DNA was transferred to the acid-soluble fraction. The reaction was conducted at 70C for 30 minutes, and stopped by adding 50 ~l of a saturated aqueous solution of sodium pyrophosphate containing 0.125 M EDTA-Na2. Samples were processed and activity was determined as described by Kaledin et al., supra.
The results showed that at pH 6.4 the polymerase was more than one-half as active as at pH 8Ø In contrast, Kaledin et al.
found that at pH about 7.0, the enzyme therein had 8% of the activity at pH 8.3. Therefore, the pH profile for the thermostable enzyme herein is broader than that for the Kaledin et al. enzyme.
Finally, when only one or more nucleotide triphosphates were eliminated from a DNA polymerase assay reaction mixture, very little, if any, activity was observed using the enzyme herein, and the 20 activity was consistent with the expected value, and with an enzyme exhibiting high fidelity. In contrast, the activity observed using the Kaledin et al. (supra) enzyme is not consistent with the expected value, and suggests misincorporation of nucleotide triphosphate(s).

IY. Amplification Reaction One microgram of the genomic DNA described above was diluted in an initial 100 ~l aqueous reaction volume containing 25 mM Tris HCl buffer (pH 8.0), 50 mM KCl, 10 mM MgCl2, 5 mM dithiothreitol, 200 ~g/ml gelatin, 1 ~M of primer PC03, 1 ~M of primer PC04, 1.5 mM dATP, 1.5 mM dCTP, 1.5 mM dGTP and 1.5 mM TTP. The sample was heated for 10 30 minutes at 98C to denature the genomic DNA, then cooled to room temperature. Four microliters of the polymerase from Thermus aquaticus was added to the reaction mixture and overlaid with a 100 ~l mineral oil cap. The sample was then placed in the aluminum heating block of the liquid handling and heating instrument described above employing pipettes programmed to deliver liquids and a temperature-control device to effect temperature changes.
The DNA sample underwent 20 cycles of amplification in the 5 machine, repeating the following program cycle:
1) heating from 37C to 98C in heating block over a period of 2.5 minutes; and 2) cooling from 98C to 37C over a period of three minutes to allow the primers and DNA to anneal.
After the last cycle, the sample was incubated for an additional 10 minutes at 55C to complete the final extension reaction.
Y. Synthesis and Phosphorylation of Oligodeoxyribonucleotide Probes A labeled DNA probe, designated RS24, of the following 15 sequence was prepared:
5' -*CCCACAGGGCAGTMCGGCAGACTTCTCCTCAGGAGTCAG-3 ' ( RS24) where * indicates the label. This probe is 40 bases long, spans the fourth through seventeenth codons of the gene, and is complementary to the normal ~-globin allele (~A). The schematic diagram of primers and 20 probes is given below:
110 bp < ~-globin >

This probe was synthesized according to the procedures 25 described in Section I of Example I. The probe was labeled by contacting 20 pmole thereof with 4 units of T4 polynucleotide kinase and about 40 pmole y-32P-ATP (about 7000 Ci/mmole) in a 40 ~l reaction volume containing 70 mM Tris buffer (pH 7.6), 10 mM MgC12, 1.5 mM
spenmine, and 10 mM dithiothreitol for 60 minutes at 37C. The total 30 volume was then adjusted to 100 ~l with 25 mM EDTA and purified according to the procedure of Maniatis et al., supra, p. 466-467 over a 1 ml spin dialysis column equilibrated with Tris-EDTA (TE) buffer (10 mM Tris buffer, 0.1 mM EDTA, pH 8.0). TCA precipitation of the reaction product indicated that for RS24 the specific activity was 4.3 ~Ci/pmole and the final concentration was 0.118 pmole/~l.

VI. Dot Blot Hybridizations Four microliters of the amplified sample from Section IV and 5.6 ~l of appropriate dilutions of ~-globin plasmid DNA calculated to represent amplification efficiencies of 70, 75, 80, 85, 90, 95 and 100~ were diluted with 200 ~l 0.4 N NaOH, 25 mM EDTA and spotted onto a nylon filter by first wetting the filter with water, placing it in an apparatus for preparing dot blots which holds the filters in place, applying the samples, and rinsing each well with 0.1 ml of 20 x SSPE
(3.6 M NaCl, 200 mM NaH2P04, 20 mM EDTA), as disclosed by Reed and Mann, Nucleic Acids Research, 13, 7202-7221 (1985). The filters were then removed, rinsed in 20 x SSPE, and baked for 30 minutes at 80C in a vacuum oven.
After baking, each filter was then contacted with 16 ml of a hybridization solution consisting of 3 x SSPE, 5 x Denhardt's solution (1 x = 0.02~ polyvinylpyrrolidone, 0.02~ Ficoll,* 0.02X bovine serum albumin,, 0.2 mM Tris, 0.2 mM EDTA, pH 8.0), 0.~ SDS and 30 formamide, and incubated for two hours at 42C. Then 2 pmole of probe 20 RS24 was added to the hybridization solution and the filter was incubated for two minutes at 42C.
Finally, each hybridized filter was washed twice with 100 ml of 2 x SSPE and 0.1X SDS for 10 minutes at room temperature. Then the filters were treated once with 100 ml of 2 x SSPE, 0.1X SDS at 60C
25 for 10 minutes.
Each filter was then autoradiographed, with the signal readily apparent after two hours.

VII. Discussion of Autoradiogram The autoradiogram of the dot blots was analyzed after two 30 hours and compared in intensity to standard serial dilution ~-globin reconstructions prepared with HaeIII/MaeI-digested pBR:~A, where BA jS
the wild-type allele, as described in Saiki et al., Science, supra.
* Trade Mark , , ~, ~, -62- l 3 3 7 4 ~ ~
Analysis of the reaction product indicated that the overall amplification efficiency was about 95%, corresponding to a 630,000-fold increase in the ~-globin target sequence.

EXAMPLE II
I. Amplification Reaction Two 1 ~9 samples of genomic DNA extracted from the Molt 4 cell line as described in Example I were each diluted in a 100 ~l reaction volume containing 50 mM KCl, 25 mM Tris HCl buffer pH 8.0, 10 mM MgCl2, 1 ~M of primer PC03, 1 ~M of primer PC04, 200 ~g/ml gelatin, 10~ dimethylsulfoxide tby volume), and 1.5 m~l each of dATP, dCTP, dGTP
and TTP. After this mixture was heated for 10 minutes at 98C to denature the genomic DNA, the samples were cooled to room temperature and 4 ~l of the polymerase from Thermus aquaticus described in Example I was added to each sample. The samples were overlaid with mineral oil to prevent condensation and evaporative loss.
One of the samples was placed in the heating block of the machine described in Example I and subjected to 25 cycles of amplification, repeating the following program cycle:
(1) heating from 37 to 93C over a period of 2.5 minutes;
20(2) cooling from 93C to 37C over a period of three minutes to allow the primers and DNA to anneal; and (3) maintaining at 37C for two minutes.
After the last cycle the sample was incubated for an additional 10 minutes at 60C to complete the final extension reaction.
The second sample was placed in the heat-conducting container of a temperature cycling (heating and cooling) machine. In this machine, the heat-conducting container is attached to Peltier heat pumps which adjust the temperature upwards or downwards and a microprocessor controller to control automatically the amplification sequence, the temperature levels, the temperature ramping and the timing of the temperature.

The second sample was subjected to 25 cycles of amplification, repeating the following program cycle:
(1) heating from 37 to 95C over a period of three minutes;
(2) maintaining at 95C for 0.5 minutes to allow s denaturation to occur;
(3) cooling fro~ 95 to 37C over a period of one minute;
and (4) maintaining at 37C for one minute.

II. Analysis Two tests were done for analysis, a dot blot and an agarose gel analysis.
For the dot blot analysis, a labeled DNA probe, designated RS18, of the following sequence was prepared.
5'-*CTCCTGAGGAGAAGTCTGC-3' (RS18) 15 where * indicates the label. This probe is 19 bases long, spans the fourth through seventeenth codons of the gene, and is complementary to the normal R-globin allele (~A). The schematic diagram of primers and probes is given below:
< 110 bp B-9lobin >

This probe was synthesized according to the procedures described in Section I of Example I. The probe was labeled by contacting 10 pmole thereof with 4 units of T4 polynucleotide kinase 25 and about 40 pmole y~32P-ATP (about 7000 Ci/mmole) in a 40 ~l reaction volume containing 70 mM Tris HCl buffer (pH 7.6), 10 mM MgCl2, 1.5 mM
spermine and 10 mM dithiothreitol for 60 minutes at 37C. The total volume was then adjusted to 100 ~l with 25 mM EDTA and purified according to the procedure of Maniatis et al., supra, p. 466-467 over 30 a 1 ml spin dialysis column equilibrated with Tris-EDTA (TE) buffer ~ 1 337405 (10 mM Tris HCl buffer, 0.1 mM EDTA, pH 8.0). TCA precipitation of the reaction product indicated that for RSlB the specific activity was 4.6 ~Ci/pmole and the final concentration was 0.114 pmole/~l.
Five microliters of the amplified sample fro0 Section I and of a sample amplified as described above except using the Klenow fragment of _ coli DNA Polymerase I instead of the thermostable enzyme were diluted with 195 ~l 0.4 N NaOH, 25 ~M EDTA and spotted onto two replicate nylon filters by first wetting the filters with water, placing them in an apparatus for preparing dot blots which holds the filters in place, apptying the samples, and rinsing each well with 0.4 ml of 20 x SSPE (3.6 M NaCl, 200 mM NaH2P04, 20 mM
EDTA), as disclosed by Reed and Mann, supra. The filters were then removed, rinsed in 20 x SSPE, and baked for 30 minutes at 80C in a vacuum oven.
After baking, each filter was then contacted with 6 ml of a hybridization solution consisting of 5 x SSPE, 5 ~ Denhardt's solution (1 x = 0.02~ polyvinylpyrrolidone, 0.02X Ficoll, 0.02~ bovine serum albumin, 0.2 mM Tris, 0.2 mM EDTA, pH 8.0) and 0.5~ SOS, and incubated for 60 minutes at 55C. Then 5 ~l of probe RS18 was added to the hybridization solution and the filter was incubated for 60 minutes at 55C.
Finally, each hybridized filter was washed twice with 100 ml of 2 x SSPE and 0.1X SDS for 10 minutes at room temperature. Then the filters were treated twice more with 100 ml of 5 x SSPE, 0.1~ SDS at 25 60C for 1) one minute and 2) three minutes, respectively.
Each filter was then autoradiographed, with the signal readily apparent after 90 minutes.
In the agarose gel analysis, 5 ~l each amplification reaction was loaded onto 4X NuSiev~/0.5~ agarose gel in 1 x T8E buffer (0.089 M Tris, 0.089 M boric acia; and 2 mM EDTA) and electrophoresed for 60 minutes at 100V. After staining with ethidium bromide, DNA was visualized by UY fluorescence.

* Tr ad e Mark -The results show that the machines used in Example I and this example were equally effective in amplifying the DNA, showing discrete high-intensity 110-base pair bands of similar intensity, corresponding to the desired sequence, as well as a few other discrete bands of much lower intensity. In contrast, the amplification method that involves reagent transfer after each cycle using the Klenow fragment of E. coli Polymerase I gave a DNA smear resulting from the non-specific amplification of many unrelated DNA sequences.
It is expected that similar improvements in amplification and detection would be achieved in evaluating HLA-DQ, DR and DP
regions.
If in the above experiments the amplification reaction buffer contains 2 mM MgCl2 instead of 10 mM MgCl2 and 150-200 ~M of each nucleotide rather than 1.5 mM of each, and if the lower temperature of 37C is raised to 45-58C during amplification, better specificity and efficiency of amplification occurs. Also, DMS0 was found not necessary or preferred for amplification.

EXAMPLE III
Amplification and Cloning For amplification of a 119-base pair fragment on the human ~-globin gene, a total of 1 microgram each of human genomic DNA
isolated from the Molt 4 cell line or from the GM2064 cell line (representing a homozygous deletion of the ~- and ~-hemoglobin region and available from the Human Genetic Mutant Cell Depository, Camden, NJ) as described above was amplified in a 100 ~l reaction volume containing 50 mM KCl, 25 mM Tris HCl pH 8, 10 mM MgCl2, 200 ~g/ml gelatin, 5 mM 2-mercaptoethanol, 1.5 mM each of dATP, dCTP, TTP, and dGTP, and 1 ~M of each of the following primers:
5'-CTTCTGcagC M CTGTGTTCACTAGC-3' (GH18) 5'-CACaAgCTTCATCCACGTTCACC-3' (GH19) where lower case letters denote mismatches from wild-type sequence to create restriction enzyme sites. GH18 is a 26-base oligonucleotide complementary to the negative strand and contains an internal PstI

~ 337405 site. GH19 is a 29-base oligonucleotide complementary to the plus strand and contains an internal HindIII recognition sequence. These primers were selected by first screening the regions of the gene for homology to the PstI and HlndIII restriction sites. The primers were then prepared as described in Example I.
The above reaction mixtures were heated for 10 minutes at 95C and then cooled to room temperature. A total of 4 ~l of the polymerase described in Example I was added to each reaction mixture, and then each mixture was overlayed with mineral oil. The reaction mixtures were subjected to 30 cycles of amplification with the following program:
2.5 min. ramp, 37 to 98C
3 min. ramp, 98 to 37C
2 min. soak, 37C
After the last cycle, the reaction mixtures were incubated for 20 minutes at 65C to complete the final extension. The mineral oil was extracted with chloroform and the mixtures were stored at A total of 10 ~l of the amplified product was digested with 20 0.5 ~g M13mplO cloning vector, which is publicly available, in a 50 ~lvolume containing 50 mM NaCl, 10 mM Tris HCl, pH 7.8, 10 mM MgCl2, 20 units P I and 26 units HindIII for 90 minutes at 37C. The reaction was stopped by freezing at -20C. The volume was adjusted to 110 ~1 with TE buffer and loaded (100 ~l) onto a 1 ml BioGel P-4 spin 25 dialysis column. One 0.1 ml fraction was collected and ethanol precipitated.
(At this point it was discovered that there was ~-globin amplification product in the GM2064 sample. Subsequent experiments traced the source of contamination to the primers, either GH18 or GH19. Because no other source of primers was available, the experiment was continued with the understanding that some cloned sequences would be derived from the contaminating DNA in the primers.) The ethanol pellet was resuspended in 15 ~l water, then adjusted to 20 ~l volume containing 50 mM Tris HCl, pH 7.8, 10 mM

_ -67- l 3 3 7 4 0 5 MgC12, 0.5 mM ATP, 10 mM dithiothreitol, and 400 units ligase. [One unit is the amount of enzyme required to give 50~ ligation of HindIII
digested ~DNA in 30 minutes at 16C in 20 ~l at a 5' termini concentration of 0.12 mM (about 330 ~l/ml)]. This mixture was incubated for three hours at 16C.
Ten microliters of ligation reaction mixture containing Molt 4 DNA was transformed into E. coli strain JM103 competent cells, which are publicly available. The procedure followed for preparing the transformed strain is described in Messing, J. (1981) Third Cleveland Symposium on Macromolecules:Recombinant DNA, ed. A. Walton, Elsevier, Amsterdam, 143-163. A total of 651 colorless plaques (and O blue plaques) were obtained. Of these, 119 had a (+l- strand insert (18%) and 19 had a (-)- strand insert (3%). This is an increase of almost 20-fold over the percentage of ~-globin positive plaques among the primer-positive plaques from the amplification technique using Klenow fragment of E. coli Polymerase I, where the reaction proceeded for two minutes at 25C, after which the steps of heating to 100C for two minutes, cooling, adding Klenow fragment, and reacting were repeated nine times. These results confirm the improved specificity of the amplification reaction employing the thermostable enzyme herein.
In a later cloning experiment with GM2064 and the contaminated primers, 43 out of 510 colorless plaques (8~) had the (+)- strand insert. This suggests that approximately one-half of the 119 clones from Molt 4 contain the contaminant sequence.
Ten of the (+)- strand clones from Molt 4 were sequenced.
Five were normal wild-type sequence and five had a single C to T
mutation in the third position of the second codon of the gene (CAC to CAT). Four of the contaminant clones from GM2064 were sequenced and all four were normal.
Restriction site-modified primers may also be used to amplify and clone and partially sequence the human N-ras oncogene and to clone base pair segments of the HLA DQ-a, DQ-~ and DR-~ genes using the above technique.

` 1 337405 Again, if the concentrations of MgCl2 and nucleotides are reduced to 2 mM and 150-200 ~M, respectively, and the minimum cycling temperature is increased from 37C to 45-58C, the specificity and efficiency of the amplification reaction can be increased.

EXAMPLE IV
Gene Retrieval A. IDENTIFICATION OF A DNA SEQUENCE PROBE FOR THE TAQ POLYMERASE GENE
A specific DNA sequence probe for the Taq pol gene was obtained following immunological screening of a Agt11 expression library. T. aquaticus DNA was digested to completion with AluI, ligated with EcoRI 12-mer linkers (CCGGAATTCCGG, New England Biolabs), digested with EcoRI and ligated with dephosphorylated, EcoRI-digested Agt11 DNA (Promega Biotech). The ligated DNA was packaged (Gigapack Plus, Strategene) and transfected into E. coli K-12 strain Y1090 (provided by R. Young).
The initial library of 2 x 105 plaques was screened (Young, R.A., and R.W. Davis (1983) Science, 222:778-782) with a 1:2000 dilution of a rabbit polyclonal antiserum raised to purified Taq polymerase (see Examples I and VI). Candidate plaques were replated at limiting dilution and rescreened until homogeneous (~3 cycles).
Phage were purified from candidate plaques which failed to react with preimmune serum and reacted with immune serum.
Candidate phage were used to lysogenize E. coli K-12 strain Y1089 (R. Young). Lysogens were screened for the production of an IPTG inducible fusion protein (larger than B-galactosidase) which reacted with the Taq polymerase antiserum. Solid phase, size-fractionated fusion proteins were used to affinity purify epitope-specific antibodies from the total polyclonal antiserum (Goldstein, L.S.B., et al. (1986) J. Cell Biol. 102:2076-2087).
The "fished", epitope-selected antibodies were used, in turn, in a Western analysis to identify which Agt11 phage candidates encoded DNA sequences uniquely specific to Taq polymerase. One Agt11 _ -69-phage candidate, designated ~gt:1, specifically selected antibodies from the total rabbit polyclonal Taq polymerase antiserum which uniquely reacted with both purified Taq polymerase and crude extract fractions containing Taq polymerase. This phage, Agt:1, was used for further study.
The ~115 bp EcoRI-adapted AluI fragment of Thermus aquaticus DNA was labeled (Maniatis et al., supra) to generate a Taq polymerase-specific probe. The probe was used in Southern analyses and to screen a T. aguaticus DNA random genomic library.

B. CONSTRUCTION AND SCREENING OF A THERMUS AQUATICUS RANDOM GENOMIC
LIBRARY
Lambda phage Charon 35 (Wilhelmine, A. M. et al., supra) was annealed and ligated via its cohesive ends, digested to completion with BamHI, and the annealed arms were purified from the "stuffer"
fragments by potassium acetate density gradient ultracentrifugation (Maniatis, et al., supra). T. aquaticus DNA was partially digested with Sau3A and the 15-20 kb size fraction purified by sucrose density gradient ultracentrifugation. The random genomic library was constructed by ligating the target and vector DNA fragments at a 1:1 20 molar ratio. The DNA was packaged and transfected into E. coli K-12 strains LE392 or K802. A library of >20,000 initial phage containing ~99% recombinants was amplified on E. coli K-12 strain LE392.
The CH35 Taq genomic phage library was screened (Maniatis et al., supra) with the radiolabeled EcoRI insert of ~gtll:1.
2 5 Specifically hybridizing candidate phage plaques were purified and further analyzed. One phage, designated Ch35::4-2, released > four T.
aquaticus DNA fragments upon digestion with HlndIII (~8.0, 4.5, 0.8, 0.58 kb) The four HindIII T. aquaticus DNA fragments were ligated 30 with HindIII digested plasmid BSM13+ (3.2 kb, Vector Cloning Systems, San Diego) and individually cloned following transformation of E. coli K-12 strain DG98.

_ -70- l 3 3 7 4 0 5 The ~8.0 kb HindIII DNA fragment from CH35::4-2 was isolated in plasmid pFC82 (11.2 kb), while the 4.5 kb HindIII DNA fragment from CH35::4-2 was isolated in plasmid pFC83 (7.7 kb).
E. coli strain DG98 harboring pFC82 was shown to contain a thermostable, high temperature DNA polymerase activity (Table 1). In addition, these cells synthesize a new ~60 kd molecular weight polypeptide which is immunologically related to Taq DNA polymerase.
The Taq polymerase coding region of the 8.0 kb HindIII DNA
fragment was further localized to the lac-promoter proximal 2.8 kb HindIII to Asp718 portion of the 8.0 kb HindIII fragment. This region was subcloned to yield plasmid pFC85 (6.0 kb). Upon induction with IPTG, E. coli DG98 cells harboring plasmid pFC85 synthesize up to 100-fold more thermostable, Taq polymerase-related activity (Table 1) than the original parent clone (pFC82/DG98). While cells harboring pFC85 synthesize a significant amount of a thermostable DNA polymerase activity, only a portion of the Taq pol DNA sequence is translated, resulting in the accumulation of a ~60 kd Taq polymerase-related polypeptide.

Expression of a Thermostable DNA Polymerase Activity in E. coli#

Sample Units*/ml IPTG +IPTG
BSM13/DG98- - 0.02 pFC82/DG98 2.2 2.7 pFC85/DG98 11.9 643.8 #cells were grown to late log phase (~/- IPTG, 10 mM), harvested, sonicated, heated at 75C for 20 minutes, centrifuged and the clarified supernatant assayed at 70C for DNA polymerase activity.
30 * 1 unit = 1 nM dCTP incorporated in 30 minutes.

~_ -71-EXAMPLE V
Expression of Taq Polymerase The thermostable gene of the present invention can be expressed in any of a variety of bacterial expression vectors including DG141 (ATCC 39588) and pPLNRBSATG. Both of these host vectors are pBR322 derivatives that have either a sequence containing a tryptophan promoter-operator and ribosome binding site with an operably linked ATG start codon (DG141) or a sequence containing the lambda PL promoter and gene N ribosome binding site operably linked to an ATG start codon (pPLNRBSATG). Either one of these host vectors may be restricted with SacI, and blunt ended with Klenow or Sl nuclease to construct a convenient restriction site for subsequent insertion of the Taq polymerase gene.
The full-length Taq polymerase gene was constructed from the DNA insert fragments subcloned into plasmids pFC83 and pFC85 as follows. Vector BSM13+ (commercially available from Yector Cloning Systems, San Diego, CA) was digested at the unique HlndIII site, repaired with Klenow and dNTPs, and ligated with T4 DNA ligase to a BglII octanucleotide linker, 5'-CAGATCTG-3', and transformed into E.
coli strain DG98. Plasmids were isolated from AmpR lacZa+
transformants. One of the clones was digested with Bgl II and Asp718 restriction enzymes, and the large vector fragment purified by gel electrophoresis.
Next, plasmid pFC83 was digested with Bgl II and HlndIII and the ~750 base pair fragment was isolated. Plasmid pFC85 was digested with HindIII and Asp718 and the ~2.8 kb fragment isolated and joined in a three-piece ligation to the ~750 base pair BglII-HlndIII fragment from pFC83 and the BglII-Asp718 vector fragment of BSM13+. This ligation mixture was used to transform E. coli strain DG98 (ATCC
39,768 deposited July 13, 1984) from which AmpR colonies were selected and an ~6.75 kilobase plasmid (pLSG1) was isolated. Isopropyl-g-D-thiogalactoside (IPTG)-induced DG98 cells harboring pLSG1 synthesized Taq DNA polymerase indistinguishable in size from the native enzyme isolated from _ aquaticus. Plasmid pLSG1 can then be used to ~ -72- l 3 3 7 4 o 5 generate a single strand DNA template according to the procedure recommended by Yector Cloning Systems.
Oligonucleotide-directed mutagenesis (see Zoller and Smith, Nuc. Acids Res. (1982) 10:6487-6500) can then be used to introduce an SphI restriction site as part of the ATG start codon (upstream of the internal HindIII site in the coding sequence of the Taq polymerase gene). Similarly, a Bgl II site can be introduced after the carboxyl-terminus of the gene (~0.7 kb upstream from the Asp718 site) to facilitate subcloning of the Taq polymerase gene into an expression vector. After the site-directed mutagenesis is performed, the gene can be isolated from the BSM13+ vector on an ~3.2 kb SphI-BstEII
restriction fragment, treated with Klenow fragment and all four dNTPs, and inserted with T4 DNA ligase (blunt-end conditions) into either one of the aforementioned expression vectors, which have been digested with SacI, repaired with Klenow and dNTPs, and treated with calf intestine phosphatase to generate dephosphorylated blunt ends. This ligation mixture is used to transform E. coli DG116 and the resulting transformants are screened for production of Taq polymerase.
Expression of the enzyme can be confirmed by Western immunoblot analysis and activity analysis.
A greater proportion of the Taq polymerase gene contained within the ~2.8 kb HindIII-Asp718 fragment of plasmid pFC85 can be expressed using, for example, plasmid pPLNpBsATG~ by operably linking the amino-terminal HlndIII restriction site encoding the Taq pol gene to an ATG initiation codon. The product of this fusion upon expression will yield an ~66,000-68,000 dalton truncated polymerase.
This specific construction can be made by digesting plasmid pFC85 with HindIII and treating with Klenow fragment in the presence of dATP, dGTP and dCTP. The resulting fragment is treated further B0 with Sl nuclease to remove any single-stranded extensions, and the resulting DNA digested with Asp718 and treated with Klenow fragment in the presence of all four dNTPs. The recovered fragment can be ligated using T4 DNA ligase to dephosphorylated plasmid pPLNpBsATG~ which had been digested with SacI and treated with Klenow fragment in the 73_ l 3 3 7 4 o 5 presence of dGTP to construct an ATG blunt end. This ligation mixture can then be used to transfonmed E. coli DG116 and the transfonmants screened for production of Taq polymerase. Again, expression can be confinmed by Western immunoblot analysis and activity analysis.

5, EXAMPLE VI
Purification The the mostable polymerase may be purified directly from a culture of Thenmus aquaticus -following the example disclosed below or, alternatively, from a bacterial culture containing the recombinantly produced enzyme with only minor modifications necessary in the preparation of the crude extract.
After harvesting by centrifugation, 60 grams of cells were resuspended in 75 ml of a buffer consisting of 50 mM Tris-Cl pH 8, 1 mM EDTA. Cells were lysed in a French Press at 14,000-16,000 PSI
after which 4 volumes (300 ml) of additional Tris-EDTA were added.
Buffer A (~-mercaptoethanol to 5 ~M and NP-40 and Tween 20 to 0.5X
(v/v) each) was added and the solution was sonicated thoroughly while cooling. The resultant homogeneous suspension was diluted further with 8uffer A such that the final volume was 7.5-8 times the starting cell weight; this was designated Fraction I.
The polymerase act~vity in Fraction I and subsequent fractions was detenm~ned in a 50 ~l ~ixture conta~ning 0.025 M TAPS-HCl pH 9.4 (20C), 0.002 M MgC12, 0.05 M KCl, 1 mM 2-mercaptoethanol, 0.2 mM each dGTP, dATP, TTP, 0.1 mM dCTP ~a-32P, .05 Ci/mM], 12.5 ~9 activatedN salmon spenm DNA and 0.01-0.2 units of the polymerase (diluted in 10 mM Tris-HCl, pH 8, 50 mM KCl, 1 mg/ml autoclaved gelatin, 0.5X NP-40, 0.5X Tween 20, and 1 mM 2-mercaptoethanol). One unit corresponds to 10 nM product in 30 minutes. ''ActivatedU DNA is a native preparation of DNA after partial hydrolysis with DNase I until 5X of the DNA was transferred to the acid-soluble fraction. The reaction was conducted at 74C for 10 minutes and then 40 ~l was transferred to 1.0 ml of 50 ~g/ml carrier DNA in 2 mM EDTA at 0C. An equal volume (1.0 ml) of 20X TCA, 2X sodium pyrophosphate was added.
* Trade Mark ~f ~ _74_ l 3 3 7 4 0 5 After 15-20 minutes at 0C the samples were filtered through Whatman GF/C discs and extensively washed with cold 5X TCA-1X pyrophosphate, followed by cold 95~ ethanol, dried and counted.
-- Fraction I was centrifuged for two hours at 35,000 rpm in a 5 Beckman TI 45 rotor at 2C and the collected supernatant was designated Fraction II.
The Taq polymerase activity was precipitated with Polymin P
(BRL, Gaithersburg, MD) (lOX, w/v, adJusted to pH 7.5 and autoclaved) after the minimum amount of Po~ymin P necessary to precipitate 90-95X
of the activity was determined, which amount was generally found to be between 0.25X and 0.3~ final volume.
An appropriate level of Polymin P was added slowly to Fraction II while stirring for 15 minutes at 0C. This solution was centrifuged at 13,000 rpm for 20 minutes in a Beckman JA 14 rotor at 15 2C. The supernatant was assayed for activity and the pellet was resuspended in 1/5 volume of 0.5X Buffer A (di~uted 1:2 with H20~.
This suspension was recentrifuged and the pellet resuspended in 1/4 volume of Buffer A containing 0.4 M KCl. This suspension was homogenized thoroughly and left overnight at 4C. The homogenate was 20 centrifuged as above and the collected supernatant designated Fraction III .
The protein fraction was collected by ~precipitationU at 75X
saturation of ammonium sulfate, centrifuged (at 27,000 rpm, SW27 rotor, 30 minutes) and the floating pellicle was resuspended in 50 mM
25 Tris-Cl pH 8, 1 mM EDTA. These steps were repeated and the protein suspension was dialyzed extensively with P-cell buffer (20 mM KP04 pH
- 7.5, 0.5 mM EDTA, 5 mM g-mercaptoethanol,- 5X (w/v) glycerol, O.SX
(v/v) NP-40 and Twee~ 20) containing 80 ~M KCl.
The dialysate was transferred to a centrifuge bottle to 30 which was added any recovered protein from sacks rinsed with the P-cell buffer containing 80 mM KCl. Centrifugation was performed at 20,000 x 9 and the time was reduced to 15 minutes. The supernatant was saved and any pellet remaining was washed, extracted with P-cell buffer and 80 mM KCl, and recentrifuged. The supernatants were then 35 combined to fonm Fraction IV.
* Trade Mark l 3 3 7 4 o 5 Fraction IV was applied to a 2.2 x 22-cm column of phosphocellulose, equilibrated with the P-cell buffer containing 80 mM
KCl. The column was washed (2.5-3 column volumes) with the same buffer and the protein eluted using a linear gradient of 80 to 400 mM
KCl in P-cell buffer. Fractions containing DNA polymerase activity (~0.18-0.20 M KCl) were pooled and concentrated 3-4 fold on an Amicon stirred cell and YM30 membrane. The cell was rinsed with the P-cel, buffer without KCl and added to the fraction concentrate (0.15 M KCl adjusted final volume~ to fonm Fraction Y.
Fraction Y was applied to a 5 ml Heparin Sepharose CL-6 column (Pharmacia) equilibrated with P-cell buffer and 0.15 M KCl.
The column was washed with 0.15 M KCl buffer (3-4 column volumes) and the protein eluted with a linear gradient from 0.15 to 0.65 M KCl in P-cell buffer. A 1:10 dilution into diluent without gelatin was made for SDS-PAGE analysis and a subsequent 1:20 dilution into diluent with 1 mg/ml gelatin was made for use in enzyme assays. The activity fractions (eluting at ~0.3 M KCl) were assayed on supercoiled DNA
template for specific and non-specific endonucleases/topoisomerase by electrophoretically detecting the change in molecular weight of 20 supercoiled plasmid DNA after incubation with an excess of D~A
polymerase. Exonuclease contamination was detected following incubation with small linear DNA fragments. In peak fractions, an ~88 - kd protein was found to be the major band. The major pool, designated Fraction YI, had the highest polymerase activity with minimal 25 detectable endonuclease activity when this pool was assayed for 30 minutes at 55C with ~3-5 polymerase units/6Q0 ng DNA.
Fraction YI was dialyzed against 10 mM KP04 pH 7.5, 5 mM B-mercaptoethanol, 5X glycerol, 0.2~ NP-40, and 0.2X Tween* 2Q (HA
buffer). The dialyzed sample was applied to a 3 ml column of 30 hydroxyapatite and the enzyme eluted with a linear ~radient of 10 to 250 mM KP04 pH 7.5, HA buffer. DNA polymerase activity began to elute at 75 mM KP04 with the peak at 100 mM KP04. Active peak fractions were assayed at 1:100-1:300 dilution. As in the prior chromatography step, a 1:10 dilution in diluent was prepared without gelatin for SDS-3 5 PAGE analysis. Fractions with no significant endonuclease or double-*Trade Mark -76- l 3 3 7 4 o 5 strand exonuclease when assayed at 55C with 5 polymerase units were pooled and designated Fraction YII.
Fraction YII was dialyzed against a solution of 25 mM sodium acetate pH 5.2, 5X glycerol, 5 mM ~-mercaptoethanol, O.l mM EDTA, 0.1 NP-40, and 0.1~ Tween 20, adjusted to pH S at room temperature. The dialyzed sample was applied to a 2 ml DEAE-Tris-Acryl-M (LK~) column pre-equilibrated and subsequently washed with the same buffer. The fraction containing polymerase activity that did not adhere to the column was pooled and adjusted to 50 mM NaCl in the same buffer to yield Fraction YIII.
Fraction YIII was applied to a 2 ml CM-Tris-Acryl M (LKB) column equilibrated with the same buffer (25 mM sodium acetate, 50 mM
NaCl, 5~ glycerol, 0.1 mM EDTA, 0.1~ NP-40, and 0.1X Tween 20). The column was washed with 4-5 column volumes of the same buffer and the eny me eluted with a linear gradient from 50 to 400 mM NaCl in sodium acetate buffer. The polymerase activity peak eluted ~0.15-0.20 M
NaCl. The polymerase activity was assayed at 1:300 to 1:500 dilution with the first dilution 1:10 into diluent without gelatin for the SDS-PAGE analysis. An assay across the activity peak on supercoiled DNA
templates for specific and non-specific endonuclease/topoisomerase using DNA polymerase assay salts (25 mM TAPS-HCl pH 9.4, 2.0 mM MgC12 and 50 mM KCl) at 74C was performed, as well as assays for nucleases on M13 ss DNA and pBR322 fragments. Active fractions with no detectable nuclease(s) were pooled and run on a silver stained SDS-PAGE mini gel. The results show a single ~88 kd band with a specificactivity of ~250,000 units/mg.
This specific activity is more than an order of magnitude higher than that claimed for the previously isolated Taq polymerase and is at least an order of magnitude higher than that for E. coli polymerase I.

EXAMPLE VII
The Taq polymerase purified as described above in Example YI
was found to be free of any contaminating Taq endonuclease and ~, *Trade Mark 77 l 3 3 7 ~ o 5 exonuclease activities. In addition, the Taq polymerase is preferably stored in storage buffer containing from about 0.1 to about 0.5 volume/volume of each non-ionic polymeric detergent employed. More preferably the storage buffer consists of ~OX (v/v) glycerol, 100 mM
KCl, 20 mM Tris-Cl pH 8.0, 0.1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol, O.SX v/v NP-40, 0.~ v/v Tween 20, and 200 ~g/ml gelatin, and is preferably stored at -20C.
The stored Taq polymerase was diluted in a buffer consisting of 25 mM Tris Cl pH 8.0, 20 mM KCl, 1 mM ~-mercaptoethanol, 0.5~ NP-40, 0.5~ Tween-20, and 500 ~g/ml gelatin. A reaction buffer was then prepared containing S0 mM KCl, 10 mM Tris-Cl, pH 8.3, 1.5 mM MgC12, 0.01~ (w/v) gelatin, 200 ~M each dNTP, 1 ~M each of the primers that define a 500 base pair target sequence on a control template from bacteriophage A, and 2.0-2.5 units Taq polymerase/assay in a final volume of 100 ~l. Template was added to the reaction buffer, the sample placed in a 0.5 ml polypropylene tube, and the sample topped with 100 ~l of heavy white mineral oil to prevent evaporation.
At least a 105-fold amplification was achieved when the following conditions were employed, using 1 ng of control template (bacteriophage ~ DNA) where the target sequence represented approximately 1~ of the starting mass of DNA.
First the template mixture was denatured for one minute, 30 seconds at 94C by placing the tube in a heat bath. Then the tube was placed in a heat bath at 37C for two minutes. Then the tube was 25 placed in a heat bath at 72C for three minutes, and then in the heat bath at 94C for one minute. This cycle was repeated for a total of 25 cycles. At the end of the 25th cycle, the heat denaturation step at 94C was omitted and replaced by extending the 72C incubation step by an additional three minutes. Following tenmination of the assay, 30 the samples were allowed to cool to room temperature and analyzed as described in preYious examples.
The template may be optimally amplified with a different concentration of dNTPs and a different amount of Taq polymerase.
Also, the size of the target sequence in the DNA sample will directly *Trade Mark -78- l 3 3 7 4 0 5 impact the minimum time required for proper extension (72C incubation step). An optimization of the temperature cycling profile should be perfonmed for each individual template to be amplified, to obtain maximum efficiency.

EXAMPLE YIII
Taq polymerase purified as described above in Example I was fonmulated for storage as described in the previous example, but without the non-ionic polymeric detergents. When assayed for activity as described in that example, the eny me storage mixture was found to be inactive. When the NP-40 and Tween 20 were added to the storage buffer, the full enzyme activity was restored, indicating that the presence of the non-ionic detergents is necessary to the stability of the enzyme fonmulation.

EXAMPLE IX
Several 1 ~9 samples of human genomic DNA were subjected to 20-3~ cycles of amplification as described in Example Y, with equivalent units of either Klenow fragment or Taq polymerase, and analyzed by agarose gel electrophoresis and Southern blot. The primers used in these reactions, PC03 and PC04, direct the synthesis of a 110-bp segment of the human beta-globin gene. The Klenow polymerase amplifications exhibited the smear of DNA typically observed with this enzyme, the apparent cause of which is the non-specific annealing and extension of primers to unrelated genomic sequences under what were essentially non-stringent hybridization conditions (1x Klenow salts at 37C). Nevertheless, by Southern blot a specific 110-bp beta-globin target fragment was detected in all lanes. A substantially different electrophoretic pattern was seen in the amplifications done with Taq polynerase where the single major band ;s the 110-bp target sequence. This remarkable specificity was undoubtedly due to the temperature at which the primers were extended.
Although, like Klenow fragment amplifications, the annealing step was performed at 37C, the temperature of Taq-catalyzed reactions * Trade Mark ~ _79_ l 3 3 7 4 0 5 had to be raised to about 70C before the enzyme exhibited significant activity. During this transition from 37 to 70C, poorly matched primer-template hybrids (which formed at 37C) disassociated so that by the time the reaction reached an enzyme-activating temperature, only highly complementary substrate was available for extension. This specificity also results in a greater yield of target sequence than similar amplifications done with Klenow fragment because the non-specific extension products effectively compete for the polymerase, thereby reducing the amount of 110-mer that can be made by the Klenow fragment.

EXAMPLE X
Amplification was carried out of a sample containing 1 ~9 Molt 4 DNA, 50 mM KCl, 10 mM Tris pH 8.3, 10 mM MgC12, 0.01~ gelatin, 1 ~M of each of the following primers (to amplify a 150 bp region):
5'-CATGCCTCTTTGCACCATTC-3'(RS79) and 5'-TGGTAGCTGGATTGTAGCTG-3'(RS80) 1.5 mM of each dNTP, and 5.0 units of Taq polymerase per 100 ~l reaction volume. Three additional samples were prepared containing 2.5, 1.3, or 0.6 units of Taq polymerase. The amplification was carried out in the temperature cycling machine described above using the following cycle, for 30 cycles:
from 70 to 98C for 1 minute hold at 98C for 1 minute from 98C to 35, 45 or 55C for 1 minute hold at 35, 45 or 55C for 1 minute from 35, 45 or 55C to 70C for 1 minute hold at 70C for 30 seconds At 35C annealing temperature, the 2.5 units/100 ~l Taq enzyme dilution gave the best-signal-to noise ratio by agarose gel electrophoresis over all other Taq polymerase concentrations. At 45C, the 5 units/100 ~l Taq enzyme gave the best signal-to-noise ratio over the other concentrations. At 55C, the 5 units/100 ~l Taq enzyme gave the best signal-to-noise ratio over -the other concentrations and over the 45C annealing and improved yield. The Taq polymerase has more specificity and better yield at 55C.

In a separate experiment the Molt 4 DNA was 10-fold serially diluted into the cell line GM2064 DNA, containing no ~- or ~-globin sequences, available from the Human Genetic Mutant Cell Depository, Camden, New Jersey, at various concentrations representing varying copies per cell, and amplification was carried out on these samples as described in this example at annealing temperatures of 35C and 55C. At 35C, the best that can be seen by agarose gel electrophoresis is 1 copy in 50 cells. At 55C, the best that can be seen is 1/5,000 cells (a 100-fold improvement over the lower temperature), illustrating the importance of increased annealing temperature for Taq polymerase specificity under these conditions.
In a third experiment, DNA from a cell line 368H containing HIV-positive DNA, available from B. Poiesz, State University of New York, Syracuse, NY, was similarly diluted into the DNA from the SC1 cell line (deposited with ATCC on March 19, 1985, an EBY-transformed ~
cell line homozygous for the sickle cell allele and lacking any HIY
sequences) at various concentrations representing varying copies per cell, and amplification was carried out as described in this Example at annealing temperatures of 35C and 55C, using the primers SK38 and SK39, which amplify a 115 bp region of the HIV sequence:
5'-ATAATCCACCTATCCCAGTAGGAGA M T-3'(SK38) and 5'-TTTGGTCCTTGTCTTATGTCCAG M TGC-3'(SK39) The results by agarose gel electrophoresis showed that only the undiluted 368H sample could be detected with the annealing temperature at 35C, whereas at least a 10-2 dilution can be detected with the annealing temperature at 55C, giving a 100-fold improvement in detection.
EXAMPLE XI
cDNA was made from 1 ~9 of rabbit reticulocyte mRNA
(Bethesda Research Laboratories) in a 100 ~l reaction volume containing 150 mM KC1, 50 mM Tris HC1 (pH 8.3), 10 mM MgC12, 5 mM DTT, 0.5 mM dATP, 0.5 mM dCTP, 0.5 mM TTP, 0.5 mM dGTP, 0.2 ~9 oligo(dT)12-18 (Pharmacia), 40 units RNasin (Promega Biotec), and 5 units AMV
reverse transcriptase (BRL) and incubated for 30 minutes at 42C. The -81- l 3 3 7 4 0 5 reaction was stopped by heating for 10 minutes at 95C. Two ~g RNase A was added to the sample (2 ~l of a 2 mg/ml solution in water) and incubated for 10 minutes at 37C.
Three amplification reactions were done with the Klenow fragment using different pairs of primers. The primer pair PC03/PC04 define a 110-bp product. The primer pair RS45/oligo(dT)25-30 define an about 370-bp product, and the primer pair PC03/oligo(dT)25-30 an about 600-bp product. PC03, PC04, and RS45 are complementary to the human ~-globin gene and each has two mismatches with the rabbit gene. PC03 and PC04 are described in Example I. RS45 has the sequence: 5'-C M GAAGGTGCTAGGTGCC-3'.
The amplification reactions were performed with 1/20th (5 ~l of the cDNA described above in a 100 ~l reaction volume containing 50 mM NaCl, 10 mM Tris HCl (pH 7.6), 10 mM MgC12, 200 ~g/ml gelatin, 10 15 DMSO, 1 ~M PC03 or RS45, 1 ~M PC04 or oligo(dT)25-30, 1.5 mM dATP, 1.5 mM dCTP, 1.5 mM TTP and 1.5 mM dGTP. The samples were heated for five minutes at 98C, then cooled to room temperature and overlayed with 100 ~l mineral oil.
The samples were subjected to 10 cycles of automated 20 amplification using the machine described in Example I and using the following program:
1) heating from 37C to 98C in a heating block over 2.5 minutes (denature);
2) cooling from 98C to 37C over 3.0 minutes (anneal);
3) adding 1 unit Klenow fragment; and 4) maintaining at 37C for 20 minutes (extend).
The final volume of each sample was about 140 ~l.
One-twentieth (7 ~l) of each sample was analyzed by electrophoresis on a 2~ agarose gel. After staining with ethidium bromide, discrete bands were seen in the PC03/PC04 and RS45/oligo(dT) samples. The sizes of the bands were consistent with the expected lengths: 110-bp for the former, about 370-bp for the latter. No , t" 1 ~ -82- l 3 3 7 4 0 5 evidence of amplification of an about 600-bp fragment with the PC03/oligo(dT) primer pair was observed.
The contents of the gel were Southern blotted onto a Genatran nylon membrane and hybridized with a nick-translated human ~-globin probe, pBR328:betaA, described in Saiki et al., Science, supra,using standard techniques. The resulting autoradiogram extended the conclusions reached previously - the 110 and about 370-bp fragments were ~-globin specific amplification products and no significant amplification of the about 600-bp band was detected.
Three additional samples were amplified with the Taq polymerase obtained as described above using the same primer pairs described previously. Five microliter portions of cDNA were amplified in lOO ~1 reaction volumes containing 50 mM KCl, 25 mM Tris HCl (pH
8.0), 10 mM MgCl2, 200 ~g/ml gelatin, 10~ DMSO, 1 ~M PC03 or RS45, 1 ~M PC04 or oligo-(dT)25-30, 1.5 mM dATP, 1.5 mM dCTP, 1.5 mM TTP and 1.5 mM dGTP. The samples were heated for five minutes at 98C, then cooled to room temperature. One microliter of Taq polymerase (1/8 dilution of lot 2) was added to each and overlayed with about 100 ~l mineral oil.
The samples were subjected to 9 cycles of amplification in the Peltier device described in the previous example using the following program:
1) 1 min., 35 to 60C ramp;
2) 12 min., 60 to 70C ramp (extend);
3) 1 min., 70-95C ramp tdenature);
4) 30 sec., 95C soak;
5) l min., 95 to 35C ramp (anneal); and 6) 30 sec., 35C soak.
After the last cycle, the samples were incubated an additional lO minutes at 70C to complete the final (1Oth cycle) extension. The final volume of each was about lOO ~l.

-83- l 3 3 7 4 0 ~
As before, 1/20th (10 ~l) of each sample was analyzed on a 2~ agarose gel. In this gel, amplification products were present in all three samples: 110-bp for PC03/PC04, about 370-bp for RS45/oligo(dT), and about 600-bp for PC03/oligo(dT). These results were confinmed by Southern transfer and hybridization with the pBR328:betaA probe.
The production of the 600-bp product with Taq polymerase but not with the Klenow fragment is significant, and suggests that Taq polymerase is capable of producing longer DNA than the Klenow fragment.
The following bacteriophage and bacterial strains were deposited with the Cetus Master Culture Collection, 1400 Fifty-Third Street, Emeryville, California, USA (CMCC) and with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Maryland, USA
(ATCC). These deposits were made under the provisions of the Budapest Treaty on the International Recognition of the Deposit of Microorganisms for purposes of Patent Procedure and the Regulations thereunder (Budapest Treaty). This assures maintenance of a viable culture for 30 years from the date of deposit. The organisms will be made available by ATCC under the terms of the Budapest Treaty, and subject to an agreement between applicants and ATCC that assures unrestricted availability upon issuance of the pertinent U.S.
patent. Availability of the deposited strains is not to be construed as a license to practice the invention in contravention of the rights granted under the authority of any government in accordance with its patent laws.

Deposit Designation CMCC No. ATCC No. Deposit CH35:Taq#4-2 3125 40,336 5/29/87 30_ coli DG98/pFC83 3128 67,421 5/29/87 E. coli DG98/pFC85 3127 67,422 5/29/87 1 33740~
-83a-In summary, the present invention is seen to provide a process for amplifying one or more specific nucleic acid sequences using a temperature-cycled chain reaction and a thermostable enzyme, in which reaction primer extension products are produced that can subsequently act as templates for further primer extension reactions. The process is especially useful in detecting nucleic acid sequences that are initially present in only very small amounts and in detecting nucleotide variations using sequence-specific oligonucleotides. Also, the amplification process herein can be used for molecular cloning.
The process herein results in increased yields of amplified product, greater specificity, and fewer steps necessary to carry out the amplification procedure, over what has been previously disclosed.

~_ 1 3 3 7 4 0 ~

Supplementary Disclosure The DNA sequence of the gene for and the predicted amino acid sequence of the Taq polymerase of the Principal Disclosure have now been determined.

Figure 3 of the drawings shows these sequences. The amino acid sequence corresponding to the deduced primary translation product is numbered 1-832.

For convenience, the amino acid sequence of this Taq polymerase will be used as a reference and other forms of the thermostable enzyme will be designated by referring to the sequence shown in Figure 3. Since the N-terminal methionine may or may not be present, both forms are included in all cases wherein the thermostable enzyme is produced in bacteria.

In addition, any thermostable polymerase cont~ining at least 50% homology to any contiguous stretch of nine or more amino acids presented therein is also intended to be within the scope of the invention. This homology can be determined using commercially available data banks such as the European Molecular Biology Laboratory (EMBL) or Genh~nk. Moreover, as new thermostable polymerases are identified, specific regions of homology between the newly identified sequences and the Taq polymerase sequence may be determined using, for example, the Sequence Analysis Software Package of the Genetics Computer Group of the University of Wisconsin. Specific regions of homology include the following sequences (numbered according to the numbering of amino acids in Figure 3): residues 190-204, 262-270, 569-587, 718-732, 743-759, and 778-790.

-_ . 1 337405 As mentioned in the Principal Disclosure the thermostable enzyme may be obtained from any source and may be a native or recombinant protein. In addition to the examples of the enzymes mentioned in the Principal Disclosure, thermostable polymerases isolated from the thermophilic archaebacteria include, for example, Sulfolobus solfataricus, Sulfolobus acidocaldarius, Thermoplasma acidophilum, Methanobacterium thermoautotrophicum, and Desulfurococcus mobilis.

A method of purification has also been developed for the thermostable polymerase of the Principal Disclosure which is in addition to the purification-procedures described in the Principal Disclosure and comprises treating an aqueous mixture cont~ining the thermostable polymerase with a hydrophobic interaction chromatographic support under conditions which promote hydrophobic interactions and eluting the bound thermostable polymerase from said support with a solvent which attenuates hydrophobic interactions. This hydrophobic interaction chromatographic technique is a separation technique in which substances are separated on the basis of differing strengths of hydrophobic interaction with an uncharged bed material contAining hydrophobic groups. Typically, the column is first equilibrated under conditions favorable to hydrophobic bonding, e.g., high ionic strength. A descending salt gradient may be used to elute the sample.

According to the invention, thé aqueous mixture cont~ining either native or recombinant polymerase) is loaded onto a column contA i n i ng a relatively strong hydrophobic gel such as Phenyl Sepharose*(manufactured by Pharmacia) or Phenyl TSK*
(manufactured by Toyo Soda). To promote hydrophobic interaction with a Phenyl Sepharose column, a solvent is used which contains, for example, greater than or equal to 0.2 M ammonium sulfate, with * Trade Malk . ~, i 1 33740~

O.2 M being preferred. Thus the column and the sample are adjusted to 0.2 M ammonium sulfate in 50 mM Tris-lmM EDTA buffer and the sample applied to the column. The column is washed with the 0.2 M ammonium sulfate buffer. The enzyme may then be eluted with solvents which attenuate hydrophobic interactions such as, for example, decreasing salt gradients, ethylene or propylene glycol, or urea. For the recombinant Taq polymerase, a preferred embodiment involves washing the column sequentially with the Tris-EDTA buffer and the Tris-EDTA buffer cont~ining 20% ethylene glycol. The Taq polymerase is subsequently eluted from the column with a 0.4 M urea gradient in the Tris-EDTA ethylene glycol buffer.

These and yet further aspects are described in the following additional examples which are not to be taken as limiting to the scope of the claims.

Example XII
In Example V of the Principal Disclosure and on page 16, reference is made to an approximately 750 base pair fragment isolated from plasmid pFC83, an approximately 2.8 Rb fragment isolated from plasmid pFC85 and an approximately 6.75 Kbase plasmid pLSGl. On further determinations these values were shown to be approximately 730, approximately 2.68 and approximately 6.58 respectively.

The oligonucleotide-directed mutagenesis referred to in Example V was specifically used to simultaneously 1) introduce an Sph I site within codons 3 to 5 of the Taq DNA polymerase gene sequence (see Figure 3, nt 8-13); 2J increase the A/T content of four of the first seven codons without effecting a change in the encoded amino acids (within codons 2-7 in Figure 3); 3) delete 170 nucleotides of the lac Z DNA and T. aquaticus DNA 5' to the DNA
polymerase gene initiation codon.

Then, bacteriophage R408 (Russel, M., et al., Gene, (1986) 45:333-338) was used to infect pLSGl/DG98 cells and direct the synthesis of the single-stranded D~A (ss) form (plus s~rand) of pLSG1. Purified pLSG1 ssDNA was ~nne~led with purified PvuII-digested BSM13~ BGIII vector fragments and the 47-mer mutagenic oligonucleotide DG26 (5'-CCCTTGGGCTCA AAA ~GTGGAAGCATGCCTCTCATAGCl~lllCCTG). Followingextension with E. coli DNA polymerase I Klenow fragment, transformation of DG98 cells, and selection of AmpR transformants, the colonies ~ere screened with 5' 32P-labeled DG26. Hybridizing candidates were screened for loss of the BglII restriction site, deletion of approximately 170 base pairs of lacZ:T. aquaticus DNA, and introduction o~ a unique SphI site. One candidate, designated pLSG2, was sequenced and shown to encode the desired sequence.

pLSGl sequence:
S.D. 47bp BqlII 10~bp CA~ 4CAGCT ATG ACC ATG ......... AGATCT
... M C ATG AGG GGG ATG CTG CCC CTC
pLSG2 sequence:
S.D. SphI
CAGGAAACAGCTATG AGA GGC ATG CTT CCA CTT m Oligonucleotide-directed mutagenesis was used to introduce a unique BqlII site in plasmid pLSG2 im~ediatety following the TGA stop codon for the Taq polymerase gene (following nucleotide 2499 in Figure 3J. As above. bacteriophage R408 was used to generate the single-stranded (plus) fon~ of plas~id pLS62. Purified pLSG2 ssDNA was annea~ed with purified PvuII-digested BSM13~ 8~ ector fra~ment and the 29-mer mutagenic oligonucleotide SC107 (5'-GCATGGGGTGGTAGATCTCACTC~IlGGC~. Following extension with Klenow fragment (SO mM each dNTP), transfonmation of DG98 cells and selection for AmpR transfor~ants. colonies were sc~eened with 5' 32P-labeled SC107. Hybridi~ing candidates were screened for acquisition of a unique 8qlII site. One candidate, designated pSYC15~8, was sequenced and shown to contain the desired sequence.

._ 1 337405 pLSG2 seq~ence:
... GCC MG GAG T~A TAC CAC CCC AT~
pSYC15~8 sequence:
BqlII
... GCC AAG GAG TGA GATC TAC CAC CCC ATG C ...

EXAMPLE XTII
Construction of expression Yectors pDG160 and pDG161 The AmpR or TetR lambdaPL promoter, gene N ribosome binding site, polylinker,BT cry PRE (Bl~ (positive retroregulatory element, described in U.S. Patent No. 4,666,848, issued May 19, 1987), in a ColE1 coptS vector were constructed from previously descr~bed p1asmids and the duplex synthetic otigonucleotide linkers DG31 and DG32. The DG31/32 duplex linker encodes a 5' HlndIII cohesive end followed by SacI, NcoI, KpnI/Asp718, XmaI/SmaI recognition sites and a 3' ~am~I
cohesive end.
Asp718 SacI ~coI XmaI
DG31 5' AGCTTATGAGCTCCATCGTAGCCCCGG
ATACTOGAGGTACCATGGGGCCCCTAG-5' DG32 A. Construction of AmpR plasmid pDG160 Plasmid pFC54.t, a 5.96 kb plasmid described in U.S. Patent 4,666,848, supra, was digested with HlndIII and Ba~HI and the isolated vector fragnent was ligated with a S-fold molar excess of nonphosphorylated and annealed DG31132 duplex. Following ligation, the DNA was digested with XbaI (to inactivate the parent vector IL-2 DNA frag~ent) and used to transform E. coli K12 strain DG116 to ampicitlin resistance. Colonies were sc.eened for loss of the des-ala-serl25 IL-2 ~utein sequence and acquisition of the DG31t32 potylinker se~uence by restriction enzyme digestion. The polylinker region in one candidate, designated pDG160, was sequenced and shown to encode the desired polylinker DNA sequence.

. Construction of TetR plasmid pDG161 Plasmid pA~74~HB ( ATCC 67605~. the source of a modi fied tetracycline resistance gene wherein the BamHI and HindIII restriction sites were eliminated, and which contains the lambdaPL promoter~ gene N ribosome binding site, cry PRE in a ColEl coptS vector, was digested to completion with HlndIII and BamHI and the 4.19 kb vector fragment purified by agarose gel electrophoresis. The purified vector DNA
fragment was ligated with a 5-fold molar excess of nonphosphorylated annealed DG31/32 duplex. E. coli K12 strain DG116 was transfo~med with a portion of the DNA, and TetR colonies screened for presence of 4.2 kb plasmids. Several candidates were further screened by restriction enzyme digestion and the polylinker region sequenced by the Sanger method. One of the candidates with the desired sequence was designated pDG161.

EXAMPLE xIV
A. Construction of an AmpR PL promoter, qene N ribosome binding site, (NR8S) Taq pol~erase (832) BT cry PRE, coptS expression Yector To express the full-length (832 amino acid) mutated Taq pol~merase sequence encoded by plasmid pSYC1578 under the control of the lambdaPL promoter and gene N ribosome binding site, plasmids pSrC1578 and pFC54.t were used. Plasmid pSYC1578 was di gested with Sph I and Bql II and the resul ting approximate 2.5 kb Taq polymerase gene fragrnent purified by agarose gel electrophoresis and electroelution. Plasmid pFCS4.t was digested to completion with HlndIII and BamHI and the vector frag~ent purified by agarose gel electrophoresis. The synthetic oligonucleotides DG27 (S'-A6CTTATCI~C~ECCAT6) and DG28 (5'-CCTCTCAlA) were synthesized and annealed. P~lrified pFC54.t fragment (0.085 pmoles), purified Taq pol~merase gene fra~aent (0.25 pnloles) and annealed nonphosphorylated DG27/28 duplex adaptor (0.43 p~oles) were combined in 30_~1 and ligated at-14 C. A portion of the ligated DNA was heated- to 75C (15 oinutes) to inacti~rate the DnA ligase ~n the sa~ples and treated with XbaI to tinearize (inactivate) any ~L-2 ~utein containing ligation `

products. The liqated and digested DNA (approximately 100 ng) was used to transform E. coli K12 strain DG116 to ampicillin resistance.
AmpR colonies were screened for the presence of an approximate 8 kb plasmid which ~ielded the expected digestion products with HindIII
(621 bp ~ J,410 bp), EcoRI (3.250 bp + 4,781 bp) and SphI (8,031 bp), Asp718 (8,031 bp~. BamHI (8,031 bp~ and PYuII (4,0gO bp + 3,477 bp +
464 bp). Several candidates were subjected to DNA sequence analysis at the 5' lambdaPL:TaqPol junction and the 3' TaqPol:BT junction. One of the candidates was also screened with an anti-Taq polymerase antibody for the synthesis of an approximate gO kd immunoreactive antigen. Single colonies were transferred from a 3oc culture plate to a 41C culture plate for two hours. The colonies were scraped with a toothpick from both the 30C and 41C plates, boiled in SDS loading buffer. subjected to SDS-PAGE electrophoresis and the separated proteins transferred to a nitrocellulose membrane. The membranes were probed with a 1:6,000 dilution of a polyclonal anti-Taq antibody and developed with a goat anti-rabbit HRP conjugate. All of the candidates tested showed evidence of temperature inducible approximate 90 kd Taq polymerase-related protein. One of the several plasmid candidates which directed the synthesis of Taq polymerase in E. coli and contained the expected DNA seguence was designated pLSG5.

B. Construction of a TetR PL promoter. gene N ribosome binding site, Taq polymerase (832) BT cry PRE coptS expression vector To express the full length (832 amino acid) mutated Taq poly~erase sequence encoded by plasmid pSrC1578 under control of the lambdaPL promoter and gene N ribosome binding site in a TetR vector, we used plasmids pSrC1578 and pAW740CHB. Plasmid pSYC1578 was digested with SphI and B~lII and the resulting approximate 2.5 kb Taq polymerase gene fragment was purified by agarose gel electrophoresis and electroelution. Plasmid pA~740CHB was digested to completion with HindIII and 8amHI and the resulting 4.19 kb Yector fragment purified by agarose gel electrophoresis and electroelution. The synthetic oligonucleotides D627 and DG28 (described previously) were annealed.
Purified pA~740CHB vector fragment (0.12 pmoles) was 1igated with "_ 1 337405 purffied Taq polymerase gene frag~ent (0.24 pmoles~ and annealed nonphosphorylated DG27128 duplex adaptor 10.24 pmoles~ in 30~u1 at 14 C. A portion of the l igated DNA (100 ng) was used to transfor~ E.
coli K12 strain DG116 to tetracycline resistance. TetR candidates were screened for the presence of an approximate 6.7 kb plasmid which yielded the expected digestion products with HindIII (621 bp ~ 6,074 bp), EcoRI (3,445 bp ~ 3,250 bp). Asp718 (6,695 bp), SphI (3,445 bp t 3,250 bp~, BamHI (6,695 bp) and PvuII (3,477 bp ~ 2,754 bp ~ 464 bp). Several candidates were subjected to DNA sequence analysis at the 5' laTIbdaPl :TaqPol junction and the 3' TaqPol :BT junction.
Candidates were also screened by single colony i~unoblot as described above for the temperature inducible synthesis of Taq polymerase. One of the plasmid candidates which directed the synthesis of Taq polymerase in E. coli and contained the expected DNA sequence was desi gnated pLSG6.

EXAMPLE xv Construction of a Met4 (~3) 829 amino acid form of Taq polymerase The predicted fourth codon of native Taq pol~merase directs the incorporation of a methionine residue (see pLSG1 and pLSG2 S' sequences above). To obtai n a further mutated form of the Taq polymerase gene that would direct the synthesis of an 829 amino acid primary transl ation product~ we used pl asmids pSYC1578 and pDG161.
Plasmid pSYC1578 was d;gested with SphI. treated with E. coli DNA
polymerase I Klenow frag~ent in the presence of dGTP to remove the four-base 3' cohesive end and generate a CTT (leucine, 5th codon) blunt end. Following inactivation of the DNA polymerase and concentration of the sample. the DNA was digested with B~lII and the approximate 2.5 kb Taq pol~merase gene frag~ent purified by agarose gel electrophoresis and electroelution. Plasmid pDG161 was digested to completion with SacI, repaired with E. coli DNA polymerase I Klenow fragment i~ the presence of dGTP to remove the four base 3' cohesive end and generate an AT6 terminated duplex blunt end. Following inactivation of the polymerase, the sample was digested with BamHI.

Digested pDG161 (0.146 pmole) and purified Taq poly~erase fragnent (0.295 pmole) were ligated at 30~g/ml under sticky end conditions oYernight. The partially ligated DNA sample (BamHIJB~lII
ends~ was diluted to 15Jug/ml and ligated for fiYe hours under blunt end conditions. The DNA ligase was inactiYated (75C. 10 minutes~ and the sample digested with NcoI to linearize any ligation products containing the pDG161 polylinker sequence. Sixty nanograms of the ligated and digested DNA was used to transform E. coli K12 strain DG116 to tetracycline resistance. TetR candidates were screened for the presence of an approximate 6.7 kb plasmid which yielded the expected digestion products when treated with HlndIII (612 bp ~ 6,074 bp). EcoRI t3,44~ bp ~ 3,241 bp) and SphI (6,686 bp). Colonies were screened as aboYe by single colony immunoblot for the temperature inducible synthesis of an approximate 90 kd Taq polymerase-related polypeptide. One of the plasmids, designated pLSG7, that directed the synthesis of a Taq polymerase-related polypeptide was subjected to Sanger sequence determination at the S' lamodaPL promoter:Taq polymerase junction and the 3' Taq polymerase:BT junction. Analysis of the DNA sequence at the 5' junction confin~ed the restriction enzyme analysis (loss of one of the SphI sites and a 612 bp HlndIII
fragment, slightly smaller than the 621 bp HindIII fragment in pLS66~
and indicated the derivation of a plasmid encoding an 829 amino acid form of Taq polymerase.

EXAMPLE- XVI
Construction of Met289 kJ289) 544 amino acid form of Taq polymerase During purification of native Taq polymerase (Example XIII) we obtained an altered fonm of Taq polymerase that catalyzed the template dependent incorporation of d~TP at 70C. ~his altered fonm of Taq pol~erase was i~unolo~ically related to the approximate 90 kd form described in Example vI ~ut was of lower molecular weight.
Based on mobility. relative to BSA and ovalbumin following SDS-PAGE
electrophoresis. the apparent molecular weight of this form is approxi~ately 61 kd. ~his altered fo m of the enzyme is not present in carefully prepared crude extracts of Thermus aquaticus cells as ~ 33740S

determined by SDS-PAGE Western blot analysis or in situ DNA polymerase activity detef~ination (Spanos, A., and Hubscher, U. (1983) Meth. Enz.
9I:263-277) following SDS-PAG gel electrophoresis. This form appears to be proteolytic artifact that may arise during sample handling.
This la~er molecular weight form was purified to homogeneity and sub~ected to N-terminal sequence determination on an ABI automated gas phase sequencer. Comparison of the obtained N-terminal sequence with the pr~dicted amino acid sequence of the Taq pol~erase gene (see Figure ~) indicates this shorter form arose as a result of proteolytic cleavage between glu289 and ser2gO.
To obtain a further truncated form of a Taq polymerase gene thit would direct the synthesis of a 544 amino acid pri~ary translation product we used plas~ids pFC54.t, pSYC1578 and the complementary synthetic oligonucleotides DG29 (5'-AGCTTATGTCTCCAAAAGCT) and DG30 (5'-AGCTIlIGGAGACATA). Plasmid pFC54.t was digested to completion with HindlII and BamHI. Plasmid pSYC1~78 was digested with BstXI and treated with E. coli DNA polymerase I
Klenow fragn~ent in the presence of all 4 dNTPs to remove the 4 nucleotide 3' cohesive end and generate a Cr~terminated duplex blunt end encoding leu294 in the Taq polymerase sequence (see pLSG1, nucleotide 880). The DNA sample was digested to completion with BglII
and the approximate 1.6 kb Bst%I (repaired~/B~lII Taq DNA fragment was purified by agarose gel electrophoresis and electroelution. The pFC54.t plasmid digest (0.1 pmole) was ligated with the Taq polymerase gene fra~ent (0.3 pmole) and annealed nonphosphorylated DG29/DG30 dup~ex adaptor ~0.5 pmole) under sticky ligase condltions at 30~g/ml, 15C overnight. The DNA was diluted to approximately 10 microgram per ~l and ligation continued under blunt end conditions. The ligated DNA
sample was digested with %baI to linearize linactivate) any IL-2 mutein encoding ligation products. 80 nanograms of the ligated and digested DNA was used to transfoml E. coli K12 strain D6116 to ampicillin resistance. A~pR candidates were sc~eened for the presence of an approxi~ate 7.17 kb p~asmid which yielded the expected digestion products with EcoRI (4,781 bp 1 2.386 bp), PstI (4.138 bp ~ 3,029 bp), ApaI (7,167 bp) and HindIII/PstI (3,400 bp ~ 3.029 bp ~ 738 bp). E.

1 33740~

coli colonies harboring candidate plasmids were screened as above by single colony immunoblot for the temperature-inducible synthesis of an approximate 61 kd Taq pol~merase related polypeptide. In addition, candidate plasmids were subjected to ONA sequence determination at the 5' lambdaPL pro~oter:Taq DNA junction and the 3' Taq DNA:~T cry PRE
junction. One of the plasmids encoding the intended DNA se~uence and directing the synthesis of a temperature-inducible 61 kd Taq polymerase related polypeptide was designated pLSG8.
Yet another truncated Taq polymerase gene contained within the ~J2.68 kb HlndIII-Asp718 fragment of plasmid pFC85 can be expressed using, for example, plasmid pPL~R8SATG, by operably linkinq the amino-terminal HindIII restriction site encoding the Taq pol gene to an ATG
initiation codon. The product of this fusion upon expression will yield an ~70,000-72.000 dalton truncated polymerase.
This specific construction can be made by digestinq plasmid pFC85 with HlndIII and treating with Klenow fragment in the presence of dATP and dGTP. The resulting fragnent is treated further with Sl nuclease to remo~e any single-stranded extensions and the resulting DNA digested with Asp718 and treated with Klenow fragment in the presence of all four dNTPs. The recoYered fragment can be ligated using T4 DNA ligase to dephosphorylated plasmid pPL~RBSAT6, which had been digested with SacI and treated with Klenow fragment in the presence of d6TP to construct an ATG blunt end. This ligation mixture can then be used to transform E. coli DG116 and the transfonmants screened for production of Taq polymerase. Expression can be confinmed by Western immunoblot analysis and activity analysis.

EXAMPLE ~VII
Construction of AmpR trp promoter operator. trpL riboso~e binding site. Taq polymerase (832l BT cry PRE coptS expression Yector ~ substitute the E. coli trp operon promoter/operator and leader peptide riboso~e binding site. we used plasmids pLSG5 and pFC52. pFCS2 was the source of the trp promoter, coptS and ampicillin resistant detenminants. Howe~er. plasmid pCS4, described in U.S.

Patent No. 4,711,845, supra, may be used -to provide the identical fragment_ . .
Plasmid pLSG5 was ~igested to completion with SphI. The Sph~ was inactivated (70 C, 10 minutes) and the digested DNA was ligated overnight at 15C with an excess of annealed nonphosphorylated DG27/28 duplex adaptor (see aboYe~. The T4 DNA ligase was inactiYated (70C, 10 minutes) and the DNA digested to completion with M1uI. The DNA
sample ~as sequentially extracted with phenol and ether, ethanol precipitated and finally resuspended in 10 mM Tris chloride pH 8, 1 mM
EDTA. Plasmid pFC52 (or pCS4) was digested to completion with M1uI
and extracted with phenol, ether and concentrated as aboYe. The D~A
sample was digested to completion with H7ndIII and the HlndIII
inactiYated (75 C, 15 minutes). The pLSG5 and pFC52 samples were ligated overnight in equal ~olar ratio and at 30 ~ug/ml under sticky end conditions. The T4 ligase was inactiYated (70C, 10 minutes) and the ligated DNA was digested with %baI to linearize (inactivate) any IL-2 encoding ligation products (from the pFC52 unwanted, 1.65 kb HlndI1I/MluI DNA fragment). E. coli K12 strain DG116 was transfonmed to ampicillin resistance with 30 nanogram of the ligated DNA. AmpR
colonies were screened for the presence of approximate 7.78 kb plasmids which yielded the expected digestion products with EcoRI
(4.781 bp + 3,002 bp), SphI (7,783 bp), H~ndIII (7,162 bp + 621 bp~, ClaI (7,783 bp) and C1aI/MluI (3,gO5 bp + 3,878 bp~. Candidate colonies were further screened for expression of an approximate 90 kd Ta~ polymerase related protein by single colony SDS-PAGE
i~munoblotting (as aboYe~. Plasmids from two of the candidates sho~ing the intended properties were transformed into E. coli Kt2 strain KB2 (ATCC No. 53075).
~ y Uestern im~unoblot, both plas~ids in both hosts were shown to direct the s~nthesis of an approximate gO kd Taq polymerase-related polypeptide upon trp limitation. By Comassie staining of SDS-PAG fractionated ~hole cell extract proteins, the trp promoter/Taq polymerase plasmids in E. coli KL2 strain KB2 -direct the accumulation of significant1y more Taq polymerase than in E. coli K12 strain D6116. One of the plasmids was designated pLSG10.

-- 1 3 3 7 4 0 ~
EXAMPL~- XVIII
Synthesis of Recombinant ~aq DNA PDlymerase Activity in E. coli E. coli K12 (DG116) strains harboring plasmids pDG160, or pLSGS, or pLSG6 were gro~n at 32C in Bonner-Vogel minimal salts media containing O.SS glucose, 10 ~g/ml thiamine, 0.25X (w/v) Difco casamino acids and ampicillin (100 ~g/~t) or tetracycline (10 ~g/ml) as appropriate. Cells were grown to ~ 0O of about 0.8 and-shifted to 37 C to si~ultaneously derepress the la~bda PL promoter ~inactivation of cI857 repressor~ and increase the copy number of the Col a Copts plas~id vector. After six-nine hours of growth at 37C, aliquots of the cells were harvested. the cells centrifuged and the pellets stored at -70~C.
Alternatively. E. coli K12 strain KB2 harboring plasmid pLSG10 was grown for eight hours at 32C in Bonner-Yogel minimal salts media containjng 0.5X glucose. 5~g/ml tryptophan. 10~/ml thiamine.
0.25S Difco casamino acids and 100 ~lg/ml ampicillin to an ~ 0O of 3Ø Cells were harvested 2S aSove.
Cell pellets were resuspended to about 62.5 ~00/ml ~V150-160~l9 total protein/ml~ in 50 mM Tris-Cl. pH 7.5. 1 mM EDTA. 2.4 mM
PMSf and 0.5Jlgtml lepeptin and lysed by sonication. Aliquots of the sonicated extracts were subjected to SDS-PAGE and analyzed by Coomassie staining and Uestern immunoblotting with rabbit polyclonal anti-Taq polymerase afftibody. In addition. portions of the extracts were ass~yed in a high temperature (74C) DNA polymerase assay (see f~a~ple- VI).
Uestern immunoblotting showed significant induction and synthesis of an approximately 94 kd Taq DNA polymerase related polypeptide in induced strafns harboring ptasmids pLS65. 6. and 10.
Coomassie blue staining -of SD5-PAG-separated total cell protein revaled the presence of a new predominant protein at rv94 kd in these induced str~ins. Finall~. high te~perature activity assays confirmed the significant level of recombinant Taq DNA poly~erase synthesis in these E. coli strains (see table, below).

-9~6-1 3374~5 Uninduced(-) or Plasmid Host Taq Pol 6ene Pronoter Induced(~) Units*/OD6 pDG160/DG116 - PL -or+ C l.O
pLSG5/DG116 ~ PL ~ 23 pLSG5/DG116 f PL ~ 308 pLS~6/DG116 + PL
pLSG6/DG116 ~ PL ~ 170 pLSGlOlKB2 ~ Trp ~ 300 * 1 unit = 10 nmole total nucleotide incorporated at 74C/30 minutes.
EXAMPLE XIX
Purification of Recombinant Taq DNA Polymerase E. coli strain DG116 harboring plasmid pLSG5 was grown in a 10 ' feralentor. The medi~ was 10 mM (NH4)2S04, -25 mM KH2P04, 4 mM
Na3Citrate, 400 ,~M FeC13, 28,M ZnC12, 34f,M CoC12, 33,c.M NaMoO4, 27 ,c~M CaCl2, 30,~M CUC12, and 32",M H3B03. The medit~m was adjusted to pH
~.5 with NaOH, 15 aM, and sterilized. The following sterile components were added: 20 mq/l thiamine HCl, 3 mM MgS04. 10 g/l glucose and 12.5 mg/l ampicillin. The pH was adjusted to 6.8 and held there using NH40~1. Glucose was fed to the culture in conjunction with the alkali demand, to maintain a qlucose concentration at 40~ of air saturation, by autonatic increases in rpm (350 to 1000) and airflow (2 to 5 l/min~. Foaming ~as cont~rolled on de~and using polypropylene glycol .
The fer~entor was inoculated ~ith cells and grown to A680 =
5.0 (14.25 hours ). The temperature ~as raised to 37~C to induce synthesis of recombinant Taq poly~erase and qrowth continued for fiYe hours to A6~30 of 16.5.
Unless other~ise indicated. all purification steps were con~ucted ~t 4C. Tt~enty grams (wet weight) of induced fro~en E. coli K12 strain D6116 harboring plasmid pLSG5 ~as thawed in 3 volumes of 50 mM Tris-Cl. pH 7.5. 1 mM EDTA, 3 ~lM PMSF. 0.64h/~a1 leupeptin and disrupted in a French Press at 20,000 psi. The lysate was adjusted to 5.5X cell volume with additional buffer and sonicated (4 x 30 seconds) to reduce viscosity (Fraction I). The crude total cell lysate was adjusted to 0.2 M (NH4)2S04 (26.43 g/l) and centrifuged for 15 minutes at 20,000XG. The supernatant (Fraction II) was heated to 75°C (in a 100 C water bath) and maintained at 72-75°C for 15 minutes to denature E. Coli host proteins. The sample was rapidly cooled to 4°C by swirling in an ice water bath. After 20 minutes of 0°C, the sample was centrifuged at 20,000XG for 15 minutes to precipitate the denatured proteins. The supernatant (Fraction III) was applied at 4 ml/hr to 61 ml Phenyl-Sepharose? Cl-4B (Pharmacia) column equilibrated with 50 mM Tris-Cl, pH. 7.5, 1 mM EDTA (Buffer A) containing 0.2 M
volumes of a) the same buffer, b) buffer A, c) Buffer A containing 20%
proteins. Taq DNA polymerase activity was eluted with 60 ml linear gradient of 0-4 M urea in Buffer A containing 20% ethylene glycol.
The active fraction (?2 M urea) were pooled (Fraction IV) and applied at 3 ml/hr to a 12 ml (1.5 X 6.0 cm) Heparin-Sepharose? Cl-6B
(Phanmacia) column equilibrated in 50 mM Tris-Cl, ph 7.5. 0.1 mM EDTA, 0.2% Tween? 20 (Buffer B) containing 0.1 M KCI. The column was washed with 2 column volumes of Buffer B containing 0.15 M KCI. The taq polmerasw was eluted with a 120 ml linear gradient of 0.15-0.65 M KCI
in Buffer B. The taq polmerase eluted as a single A280 and activity peak at ?0.29 M KC1.
Purified recombinant and native Taq polmerase proteins comigrate following electrophoresis on SDS-PAGE and staining with Coomassie blue. The purified Taq Polymerase proteins migrate slightly faster than purified Phosphorylase B (Pharmacia). consistent with a moleculat weight predicted from the DNA sequence (of pLSG5) of 93,920 daltons.
The peak activity fractions were pooled and a portion subjected to N-terminal amino acid sequence determination on an Applied Biosystems gas phase sequencer. In contrast native Taq polymerase which has a blocked amino terminus, the sequence of the *Trade Mark .

purified recombinant Taq polymerase and the individual cycle yields were consistent with the sequence predicted for the amino terminus of the Taq polymerase protein encoded by plasmid pLSG5.
The recombinant Taq polymerase encoded by plasmid pLSG5 and purified as described could amplify a human ~single copy~ sequence.
Usin~ a low temperature limit of 55C, extension temperature of 72C, upper temperature limit of 94 C and a 2-2.5 minute cycle time, comparable yields and efficiency were noted for native and recombinant Taq polymerase using 1^2 units/lOO~l PCR.

_99_

Claims (25)

1. A thermostable nucleic acid polymerase enzyme that catalyzes a combination of nucleoside triphosphates to form a nucleic acid strand complementary to a nucleic acid template strand, the enzyme being obtainable from an organism of the genus Thermus and having an estimated molecular weight of about 86,000-95,000 daltons when compared with a phosphorylase B standard assigned a molecular weight of 97,400 daltons, the enzyme migrating on a denaturing polyacrylamide gel faster than phosphorylase B and more slowly than does bovine serum albumin, and the enzyme having substantially no activity in the absence of one of the nucleoside triphosphates required for the formation of a nucleic acid strand which is fully complementary to the nucleic acid template strand, indicating high fidelity; and fragments of the enzyme having the same activity.

2. The enzyme of claim 1 that is DNA polymerase.

3. The enzyme of claim 2 from Thermus aquaticus.

4. The enzyme of claim 1 or 2 having at least 50% of the activity at pH 6.4 that it has at pH 8Ø

5. The enzyme of claim 3 having at least 50% of the activity at pH 6.4 that it has at pH 8Ø

6. The enzyme of claim 1 that is in its native form.

7. The enzyme of claim 2 that is in its native form.

8. The enzyme of claim 3 that is in its native form.

9. The enzyme of claim 4 that is in its native form.

10. A purified native thermostable DNA polymerase from Thermus aquaticus having an estimated molecular weight of 86-95,000 daltons when compared with a phosphorylase B
standard assigned a molecular weight of 97,400 daltons, said enzyme migrating on a denaturing polyacrylamide gel faster than phosphorylase B and more slowly than does bovine serum albumin and having at least half of the activity at pH 6.4 that it has at pH
8Ø

11. A thermostable enzyme that catalyzes the template-dependent combination of nucleoside triphosphates to form a nucleic acid strand complementary to a nucleic acid template strand, has an estimated molecular weight of about 86,000 to 95,000 daltons when compared with a phosphorylase B standard assigned a molecular weight of 97,400 daltons, the enzyme migrating on a denaturing polyacrylamide gel faster than phosphorylase B and more slowly than does bovine serum albumin, and has substantially no activity in the absence of one of the nucleoside triphosphates required for the formation of a nucleic acid strand which is fully complementary to the nucleic acid template strand, the enzyme being obtainable from a thermophilic bacterium of the genus Thermus by a process comprising the following steps:
(a) recovering cells of the thermophilic bacterium from a medium and lysing the cells.
(b) collecting the protein fraction by ammonium sulfate precipitation, (c) subjecting the recovered fractions to chromatography using a DEAE-cellulose column and collecting protein-containing fractions, (d) applying the collected fractions to a hydroxyapatite column and combining eluted fractions containing DNA polymerase activity, (e) subjecting the combined fractions to chromatography using a second DEAE-cellulose column equilibrated with a second buffer and pooling DNA polymerase containing fractions having minimum nuclease contamination, and (f) applying the pooled fractions to a chromatography support using a phosphocellulose column and performing elution with a KCl gradient and assaying the fractions for contaminating endo/exonucleases and for polymerase activity and pooling fractions with polymerase activity.

12. A composition of matter for use in the amplification of at least one nucleic acid sequence contained in a nucleic acid or a mixture of nucleic acids, comprising athermostable enzyme as defined in claim 1 and a carrier which does not reduce the polymerase activity of the enzyme.

13. The composition of claim 12, wherein the enzyme is a DNA polymerase.

14. The composition of claim 13, wherein the enzyme is a DNA polymerase from Thermus aquaticus.

CLAIMS SUPPORTED BY SUPPLEMENTARY DISCLOSURE

15. A method for purifying a thermostable polymerase which comprises treating anaqueous mixture containing the thermostable polymerase with a hydrophobic interaction support under conditions which promote hydrophobic interactions and eluting saidthermostable polymerase from said support with a solvent which attenuates hydrophobic interactions.

16. The method of claim 15 wherein the hydrophobic chromatographic support is Phenyl Sepharose.

17. The method of claim 15 wherein said hydrophobic interactions are provided by a buffer with an ionic strength corresponding to greater than or equal to 0.05 m NaCl.

18. The method of claim 17 wherein said hydrophobic interaction promotion conditions are provided using a buffer containing greater than or equal to 0.2 M ammonium sulfate.

19. The method of claim 15 wherein said elution solvent uses a 0-4 M urea gradient.

20. The method of claim 15 wherein the thermostable polymerase is DNA polymeraseisolated from Thermus aquaticus.

21. The method of claim 15 wherein said thermostable polymerase is a recombinantenzyme.

22. The method of claim 21 wherein the aqueous mixture has previously been enriched in thermostable polymerase activity by treatment.

23. The method of claim 22 wherein the heat treatment is conducted at temperatures in the range of at least 45°C to about 90°C.

24. A method for purifying a recombinant thermostable polymerase produced in a heat labile host cell which method comprises treating the cell lysate with temperatures in the range of at least 45°C to about 90°C to denature heat labile proteins in the cell lysate and recovering the active thermostable polymerase.

25. The method of claim 24 wherein said thermostable polymerase is from Thermus aquaticus.