CA2067991A1 - A thermostable ligase mediated dna amplification system for the detection of genetic diseases - Google Patents

Abstract

A THERMOSTABLE LIGASE MEDIATED DNA AMPLIFICATION
SYSTEM FOR THE DETECTION OF GENETIC DISEASES

ABSTRACT OF THE DISCLOSURE:
The present invention relates to the cloning of the gene of a thermophilic DNA ligase, from Thermus aquaticus strain HB8, and the use of this ligase for the detection of specific sequences of nucleotides in a variety of nucleic acid samples, and more particularly in those samples containing a DNA sequence characterized by a difference in the nucleic acid sequence from a standard sequence including single nucleic acid base pair changes, deletions, insertions or translocations.

Description

~7~91 A THERMOSTABLE LIGASE lUiEDIATED DNA AMPLIFICATION
SYSTE~A FOR THE C)ET CTION OF GENETIC DiSEASES

More than 2,000 conditions have been identified as singls-gene 5 defects for which the risk of producing affected offspring can be mathematically predicted. Among these conditions in man include Huntington's chorea, cystic fibrosis, alpha1 antitrypsin deficiency, muscular dystrophy, Hunter's syndrome, Lesch-Nyhan syndrome, Down's syndrome, Tay-Sachs disease, hemophilias, phenylketonuria, 10 thalasemias, and sickle-cell anemia.
Three important techniques have been developed recently for directly detecting these single nucleic acid base pair changes, deletions, insertions, translocations or olther mutations. However, two of these techniques cannot be easily automated. In the first 15 such technique, the presence or absence of the mutation in a patient's clinical sample is detected by analysis of a restriction digest of the patient's DNA using Southern blotting [see Journal of Molecular Biology 98:503 (1975)]. However, the Southern blotting technique cannot be used for genetic diseases where the mutation 20 does not alter a rsstriction site as, for example in alpha1 antitrypsin deficiency. The second technique is by the use of DNA
probes which inYolves the synthesis of an oligonucleotide of about 19 base pairs that is complementary to the normal DNA sequence around the mutation site. The probe is labelled and used to ~9 ~ 9 distinguish normal from mutant genes by raising the strin~ency of hybridization to a level which the probe will hybridize stably to the normal gene, but not to the mutant gene with which it has a single base pair mismatch [see Proc. Natl. Acad. Sci. USA 80:278 (1983)].
5 The original method has been modified by immobili~ing the oligonucleotide and probing with a labelled PCR arnplified sample. In this modification, the sample is allowed to hybridize to an immobilized oligonucleotide and then washed off by raising the strin~ency of hybridization as described above [see Proc. Natl, Acad~
10 Sci. USA 86:6230 (1989)]. Other methods have been developed which use fluorescent PCR primers to specifically amplify only one mutation or allele [see Proc. Natl. Acad. Sci. USA 86:9178 (1989)].
This method requires the separation of products from primers by spin columns or gel electrophoresis and hence is not amenable to 15 large scale automation. The third technique utilizes the presence of both diagnostic and contiguous probes under conditions wherein the dia~nostic probe remains substantialiy covalently bound to the contiguous probe only in the case wherein the sample nucleic acid contains the exact target sequence. In addition, the diagnostic 20 oligonucleotide probe may contain a "hook" (for example, a biotinylated oligonucleotide) which is captured (for example, by streptavidin) as a means of increasing the efficiency of the technique, and the contiguous probe may contain a detectable rnoiety or label [see Science 241:1077 (1988) and U.S. Patent 4,883,750].

2~79~

Although it is not always necessary, the detection of sin~le base pair mutations in DNA is usually preceded by techniques to increase or amplify the amount of DNA sample rnaterial. A number of techniques exist to perform nucleic acid amplification, among 5 which are: (1) polymerase chain reaction which can amplify DNA a million fold from a single copy in a matter of hours using Taq polymerase and running 20 to 30 reaction cycles on a temperature cycling instrument [see Science 239:487 (1988), and United States Patents 4,683,195, 4,683,202, and 4,800,159]; (2~ self-sustained 10 sequence replication or 3SR can amplify DNA or RNA 10 million fold from a single copy in less than an hour using reverse transcriptase, T7 RNA polymerase, and RNase H under isothermal conditions at 37C
[see Proc. Natl. Acad. Sci. USA 87:1874 (1990)]; and (3) Q Beta Replicase can replicate a few thousand RNA molecules containing a 15 special 300bp recognition sequence a billion fold in 30 minutes.
Additional techniques are available, and one, the ligase chain reaction, is discussed in the following description of the cloned thermophilic ligase according to the present invention.
In addition to various genetic diseases which may be diagnosed 2 0 utilizing the present invention, various infectious diseases can be diagnosed by the presence in a clinical sample of a specific DNA
sequence characteristic of the causative microorganism. These include bacteria, viruses, and parasites. In such procedures, a relatively small number of pathogenic organisms may be present in a 2~7~9~

clinical sample from an infected patient and the DNA extracted from these organisms may constitute only a very small fraction of the total DNA in the sample. However, specific amplification of suspected pathogen-specific sequences prior to immobilization and 5 detection by hybridization of the DNA samples should greatly improve the sensitivity and specificity of traditional procedures.
In addition, amplification is particularly useful if such an analysis is to be done on a small sample using nonradioactive detection techniques which may be inherently insensitive, or where l O radioactive techniques are employed, but where rapid detection i5 desirable.
Qlthough techniques such as these are available, the search for other techniques for determining single base pair mutations continues. The present invention, that is DNA amplification and/or 15 detection by a ligase detection reaction ~LDR) or ligase chain reaction (LCR) utilizing the thermophilic DNA ligase from Thermus aquaticus to detect a target DNA se~uence is part of that continuing effo rt.
Although other techniques utilizing E. coli or T4 DNA ligase 20 for DNA amplification have been attempted, these have been found to be unacceptable because of a high background "noise" levels (after as few as 10 cycles), a condition which does not exist in the ligase chain reaction according to the present invention.

2 ~ 9 1 DNA amplification and/or detection has also been atternpted utilizing specific ligases. For example, a ligase amplification reaction has been reported ~see Gene 76:245 (1989)~ that can amplify DNA starting with 500,000 copies in 95 hours, using 75 cycles and S replenishing the T4 DNA ligase used after each cycle. However, this reported technique is slow and requires the addition of fresh T4 ligase at each step, both of which requirements make this reported technique unacceptable for automation. The ligase chain reaction according to the present invention allows for amplification of DNA
10 from 200 copies in 3 hours using 30 cycles and does not require the addition of ligase following each cycle.
Throughout the following description of the present invention, terminology specific to the technology field will be used. In order to avoid any misunderstandings as to what is being referenced, and to 15 provide the reader with a clear understanding of what is being described, the foilowing definitions will be used:
"Amplification" refers to the increase in the number of copies of a particular nucleic acid fragrnent resulting either from an enzymatic chain reaction (such as a polymerase chain reaction, a 20 ligase chain reaction, or a self-sustained sequence replication). or from the replication of the vector into which it has been cloned.
"Blunt end ligation" refers to the covalent linkage of two ends of DNA that are completely flush, i.e. have no cohesive end overhangs.

2~S799~

"Cell", "cell line", and "cell culture" may 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 reyard for the S number of transfers. It is also understood that all progeny may not be precisely identical in DNA content due to deliberate or inadvertent mutations. However, all mutant progeny having the same functionality as screened for in the originally transformed cell are included.
"Clone" refers to a group of genetically identical molecules, cells or organisms asexually descended from a common ancestor.
"Cloning" is the process of propagating such identical molecules, cells or organisms. Recombinant DNA tachniques make it possible to clon~ individual genes; this is referred to as "molecular cloning".
l S "Covalently attaching" refers to forming a covalent chemical bond between two substances.
"Cycle" refers to a single melting and cooling of DNA. For example, at very high temperatures such as 94C, virtually all double stranded DNA (independent of length) unwinds and melts. If 20 one cools the temperature (to 45-65C) in the presence of complementary oligonucleotides, they can hybridize to the correct sequences of the unwound melted DNA. DNA that has been meltecl and cooled in the presence of complementary oligonucleotides is now a substrate for the DNA ligase reaction. See I'Tm".

2~67~91 "Diagnostic portion" refers to that portion of the target sequence which contains the nucleotide change, the presence or absence of which is to be detected. "ContiguQus portion" refers to a sequence of DNA which is a continuation of the nucleotide sequence 5 of that portion of the sequence chosen as diagnostic. The continuation can be in either direction .
It will be recognized, based on the following description, that the precise position of the selected oligonucleotide containing the diagnostic portion is arbitrary, except that it must contain the 10 nucleotide(s) which differentiate the presence or absence of the target sequence at one of its ends. Thus, the oligonucleotide containing the contiguous portion continues the sequence of this arbitrarily chosen oligonucleotide containing the diagnostic portion such that the diagnostic nucleotide(s) is at the junction of the two 1 5 oligonucleotides.
"Endonuclease" refers to an enzyme (e.g., restriotion endonuclease, DNase 1) that cuts DNA at sites within the molecule.
~ Expression system" refers to DNA sequences containing a desired coding sequence and control sequence in operable linkage in 20 such a manner that hosts transformed with these sequences are capable of producin~ the encoded proteins. In order to effect transformation, the expression system may be included on a vector, or the transformed vector DNA rnay also be integrated into the hnst chromosome.

2~799~

"Gene" refers to a DNA sequence which 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.
"Gene library" or "library" refers to a collection of randomly-cloned fragments that encompass substantially the entire genome of a given species. This is also referred to as a clone bank or shotgun collection .
"Genome" refers to the entire DNA of an organism.
"Hook" refers to a modification of a probe that enables the user to rapidly and conveniently isolate probes containing this modification by "catching" the hook. The interaction between hook and catching mechanism can be, for exarnple, covalent bonding or ligand/receptor binding of sufficient affinity. Such hooks may include antigens which can be recovered by antibody, biotin which can be recovered by avidin or streptavidin, specific DNA sequences which can be recovered by compiementary nucleic acid, or DNA
binding proteins (repressors), and specific reactive chemical functionalities which can be recovered by other appropriate reactive 2 0 groups.
"Hybridization" and ~'binding" in the context of probes and denatured melted DNA are used interchangeably. Probes which are hybridized or bound to denatured DNA are base paired or "aggregated"
to complementary sequences in the polynucleotide. Whether or not a ~$7g9~

particular probe remains base paired or aggregated with the polynucleotide depends on the degree of complementarity, the length of the probe, and the stringency of the binding conditions. The higher the stringency, the higher must be the degree of complementarity, and/or the longer the probe.
"Klenow fragment" refers to a 76,00û dalton polypeptide obtained by partial proteolytic digestion of DNA polymerase 1. This enzyme possesses the 5'--->3' polymerase and 3'--->5' exonuclease activities, but not the 5'---~3' exonuclease activity of DNA
polymerase i.
"Label" refers to a modification to the probe nucleic acid which enables the user to identify the labelled nucleic acid in the presence of unlabelled nuclelc acid. Most commonly, this is the replacement of one or more atoms with radioactive isotopas.
However, other labels may be substituted for the isotopes as, for example, covalently attached chromophores, fluorescent moieties, enzymes, antigens, groups with specific reactivity, chemiluminescent moieties, and electrochemically detectable moieties.
"Ligase" refers to an enzyme which catalyses the formation of a phosphodiester bond at the site of a single-stranded break in duplex DNA. The iigase enzyme also catalyses the covalent linkage of duplex DNA; blunt end to blunt end, or one cohesive end to another complementary cohesive end.

2~7991 "Ligase Chain Reaction (LCR)" refers to the amplification of a oligonucleotide ligation product. For example, if oligonucleo$ides are designed such tha~ the DNA products of one cycle can become the DNA substrates of the next cycle, repeating such cycles will cause S an exponential amplification of the DNA (a "chain reaction"). As a thermophilic ligase enzyme is capable of remaining active during many DNA melting and cooling cycles, this allows a DNA
amplification to occur rapidly and automatically in a single reaction vessel subject to many thermal cycles in which the oligonucleotide 10 ligation product is amplified.
"Ligase detection reaction (LDR)" refers to the use of two adjacent oligonucleotides for the detection of specific sequences with the aid of a thermophilic ligase with linear product amplific~tion.
"Ligase DNA sequence" refers to the DNA sequence in Thermus aqauticus HB8 for the thermophilic ligase of the present invention which comprises, at the amino terminus of the ligase protein, the following nucleic acid sequence:
TCGrG~TAGG GGATGCGCCC CTAGTCC~AG GG~AGTATA GCCC~AGGTA
C~CTAGGGCC
ATG ACC CTG G~A G~G GCG AGG AAG CGG GTA AAC GAG TTA CGG GAC
CTC ATC CGC TAC CAC AAC TAC CGC TAC TAC GTC CTG GCG GAC CCG
GAG ATC TCC GAC GCC GAG TAC GAC CGG CTT CTT AGG GAG CTC AAG
GAG CTT GAG GAG CCC TTC CCC GAG CTC A~A AGC CCG GAC TCC CCC
ACC CTT CAG GTG GGG GCG AGG CCT TTG GAG GCC ACC TTC CGC CCC

~7~gl GTC CGC CAC CCC ACC CGC ATG TAC TCC TTG GAC AAC GCC TTT AAC

GGG CGG AAG GSC CCC TTC GCC TAC ACC GTG GAG CAC AAG GTG GAC
GGG CTT TCC GTG AAC CTC TAC TAC GAG GAG GGG GTC CTG GTC TAC
GrG GCC ACC GCC GSG GAC GSG GAG GTG GGG GAG GAG GTC ACC CAG
AAC CTC CTC ACC ATC CCC ACC ATC CCG AGG AGG CTC AAG GSG GTG
CCG GAG CGC CTC GAG GTC CGG GSG GAG GTC TAC ATG CCC ATA GAG
GCC TTC CTC CGG CTC AAC GAG GAG CTG GAG GAG CGG GrG GAG AGG
ATC TTC A~A AAC CCT AGG AAT GCG GCG GCG GGT TCC TTA AGG CAA
AAA GAC CCC CGC ATC ACC GCC A~G CGG GGC CTC AGG GCC ACC TTC
TAC GCC TTA GSG CTT GSG CTG GAG GAG GTG GAG AGG C~A GSG GTG
GCG ACC CAG TTT GCC CTC CTC CAC TGG CTC AAG G~A A~A GSC TTC
CCC GTG GAG CAC GGC TAC GCC CGG GCC GTG GGG GCG G~A GSG GTG
GAG GCG GTC l'AC CAG GAC TGG CTC AAG AAG CGG CGG GCG CTT CCC
TTT GAG GCG GAC GSG GTG GTG GTG AAG CTG GAC GAG CTT GCC CTT
TGG CGG GAG CTC GGC TAC ACC GCC CGC GCC CCC CGG TTC GCC ATC

GTG GTC TTC CAG GTG G~rG CGC ACC C~G CGG GTG ACC CCC GTG GGG
ATC CTC GAG CCC GTC TTC CTA GAG GGC AGC GAG GTC TCC CGG GTC
ACC CTG CAC AAC GAG AGC TAC ATA GAG GAG TTG GAC ATC CGC ATC
GSG GAC TGG GTT TTG GTG CAC AAG GCG GSC GGG GTC ATC CCC GAG
GTC CTC CGG GTC CTC AAG GAG AGG CGC ACG GSG GAG GAA AGG CCC
ATT CGC TGG CCC GAG ACC TGC CCC GAG TGC GGC CAC CGC CTC CTC
AAG GAG GGG AAG GTC CAC CGC TGC CCC AAC CCC TTG TGC CCC GCC
A~G CG~ TTT GAG GCC ATC CGC CAC TTC GCC TCC CGC AAG GCC ATG
GAC ATC QG GGC CTG GSG G~A AAG CTC ATT GAG AGG CTT TTG G~A
AAG GSG CTG GTC AAG GAC GTG GCC GAC CTC TAC CGC TTG AGA AAG
GAA GAC CTG GTG GSC CTG GAG CGC ATG GSG GAG AAG AGC GCC CAA
AAC CTC CTC CGC GAG ATA GAG GAG AGC AAG A~A ACA GSC CTG GAG
CGC CTC CTC TAC GCC TTG GSG CTT CCC GGG GTG GGG GAG GTC TTG
GCC CGG AAC CTG GCG GCC CGC TTC GGG AAC ATG GAC CGC CTC CTC
GAG GCC AGC CTG GAG GAG CTC CTG GAG GTG GAG GAG GTG GSG GAG

2~7~

CTC ACG GCG AGG GCC ATC CTG GAG ACC TTG AAG GAC CCC GCC TTC
CGC GAC CTG GTA CGG AGG CTC A~G GAG GCG G5G GTG GAG ATG GAG
GCC AAG GAG A~G GGC G5G GAG GCC CTT A~A GGG CTC ACC TCC GTG
ATC ACC GGG GAG CTT TCC CGC CCC CGG G~ GAG GTG AAG GCC CTC
S CTA AGG CGC CTC GSG GCC AAG GTG ACG GAC TCC GTG AGC CGG A~G
ACG AGC TAC CTC GTG GTG GGG GAG A~C CCG GGG GAG AAC CCG GGG
AGC AAG CTG CAG AAG GCC AGG GCC CTC GGG GTC CCC ACC CTC ACG

GCG GAG GAG CTC GTC TAA AG5CTTCC.
The corresponding amino acids are:
Met Thr Leu Glu Glu Ala Arg Lys Arg Val Asn Glu Leu Arg Asp Leu Ile Arg l'yr His Asn Tyr Arg Tyr Tyr Val Leu Ala Asp Pro Glu Ile Ser Asp Ala Glu Tyr Asp Arg Leu Leu Arg Glu Leu Lys Glu Leu Glu Glu Arg Phe Pro Glu Leu Lys Ser Pro Asp Ser Pro Thr Leu Gln Val Gly Ala Arg Pro Leu Glu Ala Thr Phe Arg Pro Val Arg His Pro m r Arg Met Tyr Ser Leu Asp Asn Ala Phe Asn Leu Asp Glu Leu Lys Ala Phe Glu Glu Arg Ile Glu Arg Ala Leu Gly Arg Lys Gly Pro Phe Pla Tyr Thr Val Glu His Lys Val Asp Gly Leu Ser Val Asn Leu Tyr Tyr Glu Glu Gly Val Leu Val Tyr Gly Ala Thr Arg Gly C-lu Gly Glu Val Gly Glu Glu Val Thr Gln Asn Leu Leu Thr Ile Pro l'hr Ile Pro Arg Arg Leu Lys Gly Val Pro Glu Arg Leu Glu Val Arg Gly Glu Val Tyr M~t Pro Ile Glu Ala Phe Leu Arg Leu Asn Glu Glu Leu Glu Glu Arg Gly Glu Arg Ile Phe Lys Asn Pro Arg Asn Ala Ala Ala Gly Ser Leu Arg Gln Lys Asp Pro Arg Ile Thr Ala Lys Arg Gly Leu ~rg Ala Thr Phe Tyr Ala Leu Gly Leu Gly Leu Glu Glu Val Glu Arg Glu Gly Val Ala m r Gln Phe Ala Leu Leu His Trp Leu Lys Glu Lys Gly Phe Pro Val Glu His Gly Tyr Ala Arg Ala Val Gly Ala Glu Gly Val Glu Ala Val Tyr Gln Asp Trp Leu Lys Lys Arg Arg Ala Leu Pro Phe Glu Ala Asp Gly Val Val Val Lys Leu Asp Glu Leu Ala Leu Try Arg Glu Leu Gly Tyr Thr Ala Arg Ala Pro Arg Phe Ala Ile 2~67~9~

Ala Tyr Lys Phe Pro Ala Glu Glu Lys Glu Thr Arg Leu Leu Asp Val Val Phe Gln Val Gly Arg Thr Gly Arg Val Thr Pro Val Gly Ile Leu Glu Pro Val Phe Leu Glu Gly Ser Glu Val Ser Arg Val Thr Leu His Asn Glu Ser Tyr Ile Glu Glu Leu Asp Ile Arg Ile Gly Asp Trp Val Leu Val His Lys Ala Gly Gly Val Ile Pro Glu Val Leu Arg Val Leu L~s Glu Arg Arg Thr Gly Glu Glu Arg Pro Ile Arg Trp Pro Glu Thr Cys Pro Glu Cys Gly Hi.s Arg Leu Leu Lys Glu Gly Lys Val His Arg Cys Pro Asn Pro Leu Cys Pro Ala Lys Arg Phe Glu Ala Ile Arg His Phe Ala Ser Arg Lys Ala Met Asp Ile Gln Gly Leu Gly Glu Lys Leu Ile Glu Arg Leu Leu Glu Lys Gly Leu Val Lys Asp Val Ala Asp Leu Tyr Arg Leu Arg Lys Glu Asp Leu Val Gly Leu Glu Arg Met Gly Glu Lys Ser Ala Gln Asn Leu Leu Arg Gln Ile Glu Glu Ser Lys Lys Arg Gly Leu Glu Arg Leu Leu Tyr Ala Leu Gly leu Pro Gly Val Gly Glu Val Leu Ala Arg Asn Leu Ala Ala Arg Phe Gly Asn Met Asp Arg Leu Leu Glu Ala Ser Leu Glu Glu Leu Leu Glu Val Glu Glu Val Gly Glu Le.u Thr Ala Arg Ala Ile Leu Glu Thr Leu Lys Asp Pro Ala Phe Arg Asp Leu Val Arg Arg Leu Lys Glu Ala Gly Val Glu Met Glu Ala Lys Glu Lys Gly Gly Glu Ala Leu Lys Gly Leu Thr Phe Val Ile Thr Gly Glu Leu Ser Arg Pro Arg Glu Glu Val Lys Ala Leu Leu Arg Arg Leu Gly Ala Lys Val Thr Asp Ser Val Ser Arg Lys Thr Ser Tyr Leu Val Val Gly Glu Asn Pro Gly Ser Lys Leu Glu Lys Ala Arg Ala Leu Gly Val Pro Thr Leu Thr Glu Glu Glu Leu Tyr Arg Leu Leu Glu Ala Arg Thr Gly Lys Lys Ala Glu Glu Leu Val "Ligating" refers to covalently attaching polynucleotide sequences together to form a single sequence. This is typically performed by treatment with a li~ase which catalyzes the formation of a phosphodiester bond between the 5' end of one sequence and the 3 0 3' end of the other. I lowever, in the context of the invention, the I ~

2~7991 term "ligating" is also intended to encompass other methods of covalently attaching such sequences, e.g., by chemical means. The terms "covalently attaching" and "ligating" may be used interchangeably.
"Nick closing activity" refers to covalent linkage of adjacent strands of DNA. It may be used to assay for ligase activity by virtue of converting open circular DNA (OCDNA) to covaiently ciosed circular DNA (CCCDNA) and determining the speed at which the specimen l:)NA migrates on an ethidium bromide stained agarose ~el (OCDNA migrates slower than CCCDNA).
"Oli~onucleotide" refers to a molecule comprised of two or more deoxyribonucleotides or ribonucleotides, preferably more than three. Its exact size will depend on the ultimate function or use of the oli~onucleotide. The oligonucleotide may be derived synthetically or by cloning.
"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 control sequences.
"Overproducer strain" refers to a strain of bacteria or other host cell that may be induced to overproduce a particular enzyme or chemical substance.

2~7g~

"Polymerase" refers to enzymes which catalyze the assembly of deoxyribonucleotides into DNA.
"Polymerase chain reaction (PCR)" refers to a patented process (described in United States Patents 4,683,202 and 4,683,195) for S the exponential amplification of a specific DNA fragment by utilizing two oligonucleotide primers that hybridize to opposite strands and flank the region of interest in a target DNA. The process consists of a repetitive series of cycles involving template denaturation, primer annealing, and the extension of tha annealed primers by Taq DNA polymerase.
"Probe" refers to an oligonucleotide designed to be sufficiently complementary to a sequence in a denatured nucleic acid to be probed (in relation to its length) to be bound under selectcd stringency conditions. "Contiguous probe" describes a probe which is complementary to the contiguous portion. "Diagnostic probe"
describes a probe which is complementary to the diagnostic portion.
"Target probe" describes a probe which is complementary to the target sequence and is made by covalently attaching (ligating) the diagnostic probe and the contiguous probe.
"Reporter group" refers to a group that signifies the presence of a particular moiety (see "label`').
"Restriction endonucleases" refers to those enzymes which cut DNA by recognizing specific sequences internal to the molecule and ~7~

subsequently cutting the DNA in both strands at sites either within or outside of the recognition sequence.
"Sticky end ligation" refers to the covalent linkage of two ends of DNA that contain complementary 5' or 3' single strand overhangs 5 which are usually, but not limited to, one to five nucleotides in length .
"Stringency" refers to the combination of conditions to which nucleic acids are subjec~ that cause the double stranded DNA to dissociate into component single strands; among these are pH
1 û extremes, high temperature, and salt concentration. "High string~ncy" refers to the conditions, specifically hybridization and washing, which are sufficient to allow for the detection of unique sequences usin~ an oligonuclaotide probe or closely related sequence under standard Southern hybridization protocols [as described in J.
15 Mol. Biol. 98:503 (1 975)].
"Tm" refers to the temperature at which two complementary strands of DNA unwind and separate. This is a function of the single stranded DNA length and its base composition - for small fragments, an approximate value of Tm in C is equal to 4(G+C) ~ 2(A+T). For 20 example, an oligonucleotide which has 5G, 7C, 5A, and 4T bases has a temperature of 4(5+7) + 2(5+4~ or 66C.
"Tar~et sequence" refers to a nucleic acid sequence, the presence or absence of which is desired to be detected. In the context of a preferred application of the method according to the 7~

present invention, it is a sequence which forms part of a coding region in a gene associated with a genetic disease, such as sickle-cell anemia. In many such diseases, the presence of the genetic aberration is characterized by small changes in the coding sequence;
5 most frequently, normal individuals have sequences which differ by one nucleotide from the corresponding sequences present in individuals with the genetic "deficiency." In the method according to the present invention, either the normal or altered sequence can be used as the target sequence.
"Thermophilic enzyrne" refers to an enzyme which functions at high temperatures of 50 to 90C; some rnay survive brief exposure to temperatures of 94 to 100C at which normai enzymes denature and thus become inactive.
"Thermostable ligase" refers to an enzyme which is stable to 15 heat, is heat rasistant, and catalyzes (tacilitates) ligation, at high temperatures of 50 to 90C, of adjacent oligonucleotides in the proper manner to form a product which is complementary to the target nucleic aoid strand. Generally, the enzyme activates the 5' end of one oligonucleotide and links this to the 3' strand of an 20 adjacent DNA molecules. There rnay, however, be thermostable enzymes which use other mechanisms to covalently attach adjacent oligonucleotides. Thermostable ligase can, under the proper conditions, covalently link a number of different nucleic acid 2~579~

substrates at high temperatures of 50 to 90C, such as closing "nioks" in DNA, and sticky end and blun~ end ligations.
The therrnostable enzyme according to the present invention must satisfy a single criterion to be effective for the amplification reaction, i.e., the enzyme must not become irrev0rsibly denatured (inactivated) when subjected to the elevated temperatures for the time necessary to effect denaturation of double-stranded nucleic acids. By "irreversible denaturation" as used in this connection, is meant a process bringing about a permanent and complete loss of 10 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 85C, for shorter oligonucleotides, to about 105C for a tirne depending mainly on the 15 temperature and the nucleic acid length, typically from about 0.25 minutes for shorter oligonucleotides, to 4.0 minutes for longer pieces of DNA. 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 20 denatured at about 90 to 100C. The thermostable enzyme according to the present invention has an optimum temperature at which it functions that is greater than about 45C, probably between 50 and 90C, and optimally between 60 and 80C.

I ~
~6~9~L

A more thorough and complete understanding of the cloning of the therrnophilic ligase sequencs and the use of this enzyme in the thermophilic ligase mediated DNA amplification procedure for the detection of single base pair sequence differences in genetic 5 diseases can be obtained by reference to the following figures and examples which are presented by way of illustration only and are not intended, nor should they be considered, to limit the scope of the claimed invention.
With specific reference to the figures, FIG. 1 is a depiction of plasmids pDZ1 and pDZ7;
FIG. 2 is a flow chart of the Ligase Chain Reaction (LCR) according to the present invention;
FIG. 3 is an autoradiogram demonstrating the specificity of T.
aquaticus thermophilic ligase under both LDR and LCR arnplification conditions according to the present invention;
FiG. 4 is an autoradiogram demonstrating LCR amplification at different target concentrations;
FIG. 5 is an autoradiogram demonstrating the detection of B
globin alleles using human genomic DNA.
FIG. 6 is an overview of an ELISA based oligonucleotide ligation assay according to the present invention.
FIG. 7 is a photographic representation of SDS-10%
polyacrylamide gel electrophoresis of the thermostable lisase, 2Q~7~9:~

according to the pr~sent invention, at different stages of purification .
FIG. 8 is a second photographic representation of SDS-10%
polyacrylamide gel electrophoresis of the thermostable ligase, 5 according to the present invention, at different stages of purification .
FIG. 9 is a depiction of three clones prepared in accordance with the present invention.
In Fig. 7, lanes A and G represent marker proteins (molecular 10 weights are given in kd); B represents whole cells after induction; C
represents crude supernatant after sonication; D represents pooled DEAE flow-through after heat treatment; and F and F represent fractions 23 and 24 after phosphocellulose chromatography. In Fig.
8, lanes A and H represent marker proteins (molecular weights are 15 given in kd); B represents whole cells after induction; C represents crude supernatant after sonication; D represents pooled DEAE flow-through after heat treatment; E represents fraction 23 after phosphocellulose chromatography; F represent fraction 23 incubated with nieked DNA in ligase buffer in the absence of NAD; and G
20 represents fraction 23 incubated with NAD in ligase buffer in the absence of nicked DNA. In Fig. 8, the higher molecular weight ligase (approximately 81 kd) is the adenylated form, while lower molecular wsight ligase (approximately 78 kd) is non-adenylated.

~7~

The plasmids depicted in Fig. 1 have been deposited with, and accepted by, a collection agency under the Budapest Treaty deposit rules. Plasmid pDZ1 has been incorporated within a host bacteria (E.
coli strain AK53), deposited with the American Type Culture Collection, and granted the collection number ATCC No. 68307.
Plasmid pDZ7 has been incorporated within a host bacteria (E. co/i strain AK53), deposited with the American Type Cultllre Collection, and granted the collection number ATCC No. 68308.
While other methods may be used, in general, the production of 10 the thermophilic ligase according to the present invention will be by recombinant means which typically involve the following:
First, a DNA is obtained which encodes the mature (as used herein the term includes all muteins) enzyme or a fusion of the thermophilic ligase to an additional sequence that does not destroy 15 its activity or to an additional sequence cleavable under controlled conditions to give an active protein. if the sequence is uninterrupted by introns, it is suitable for expression in any host.
However, the sequence should be in an excisable and recov~rable form. Using PCR technology, for example, most DNA sequences 20 coding for enzymes may be amplified and hence recovered in an 'excisedll form.
The excised or recavered coding sequence is then placed in operable linkage with suitable control sequences in a replicable expression vector which is used to transform a suitable host. The 2~7991 transformed host is then cultured under suitable conditions to effect the production of the recombinant thermophilic ligase, and the ligase isolated and purified by known rneans.
Each of the above procedures may be accomplished in a variety S 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 are made using appropriate replicons and control sequences; and suitable restriction sites may, if not normally available, be added to 10 the ends of the coding sequence so as to provide an excisable gene to insert into the appropriate vector.
The control sequences, expression vectors, and transformation methods are dependent on the type of host cell used to express the gene. Generally, bacterial hosts are the most efficient and 15 convenient for the production of recombinant proteins and therefore preferred for the expression of the thermophilic ligase according to the pres~nt invention. However, other hosts such as yeast, plant, and insect or mammalian cells may also be used if convenient. For the purposes of the present invention, one source of the host cell is 20 considered to be equivalent to any other available and suitable host cell source.

2~7991 EXAMPLE I
(growth of T. aqvaticus strain HB8 and isolation of DNA) DNA was isoiated from Thermus thermophilus strain HB8 (ATCC No. 27634). This strain has recently been reclassified as S Thermus aqauticus strain HB8 [see Arch. Microbiol 117:189 (1978)l.
Cells were grown overnight at 75C in a water bath shaker in TAB broth [see Nuc. Acids Res., pgs 6795-6804 (1981)~ (which contains per liter, 5 g BactoTM-tryptone, 3 g yeast extract, 2 ~ NaCI, and 1 g dextrose) adjusted to pH 7.2 - 7.5 with NaOH, and harvested by centrifugation to yield 3.1 g wet weight from 800 ml of media.
Cells were resuspended in 15 ml of 50 mM Tris pH 8.0 buffer containing 50 mM EDTA and 15 mg egg white Iysozyme. The resuspended cells were Iysed by the addition of 2 ml of 10%
(weight/volume) sodium dodecyl sulfate followed by incubation at 37C for 15 minutes and two repeated c:ycles of ~reezing at -50C
and thawing at 37C. The aqueous solution was extracted sequentially with equal volumes of aqueous phenol (preequilibrated to pH 7.5 with sodium borate), followed by phenol/chloroform, and finally chloroform.
Nucleic acids were precipitated by mixing with 2 volumes of 95% ethanol, chilling to -50C for 15 rnin., and pelleted by centrifugation. After removal of the supernatant and drying the pellet, nucleic acids were resuspended in 1 ml TE buffer (10 mM Tris HCI, pH 8.0, containing 1 mM EDTA). RNA was digested by the 2~S799:L

addition of 100 ,ug RNase A to each ml of suspension, and ~he mixture incubated at 37C for 1 hr. DNA was precipitated by adding 1/10th vol. of 3 M sodium acetate and 3 vol. of 100% ethanol, chilled to -50C for 15 min., pelleted by centrifugation, washed with 70%
S ethanol, and finally resuspended in TE buffer at a final concentration of 2 mg/ml.

~7~

Although DNA utilized in the example given above was isolated from Thermus aquaticus, the resultant thermophilic ligase having the necessary properties according to the present invention may have as its initial source DNA isolated from other Thermus species 5 or other thermophilic bacteria, phages, or viruses.
DNA isolated from T. aquaticus strain HB8 cannot be cleaved by the restriction endon~3cleases Taq I (whose recognition sequence is TCGA) or EcoRI (whose recognition sequence is GAATTC~. The inability to cleave certain sequences is a consequence of protective 10 methylation [see H. O. Smith and S. V. Kelly, DNA Methylati~n:
Biochemistry and Biological Significance, eds. Razin, Cadar and Riggs, p 39 - 71, Springer-Verlag Inc., New York (1987)] at the N6 position of adenine residues. Previous investigators [see J. Bact.
169:3243 (1987)] have shown that there is a gene, termed mrr, 15 which restricts adenine methylated DNA of the form G-6MeANTC and CTGC-6MeAG. In the cioning of the Taq I restriction endonuclease and methylase, several E. coli strains were found te restrict the TCGA methylated DNA, an affect originally (but incorrectly) attributed to the mrr gene [see Gene ~6:13 ~1987) and Nuc. Acid Res.
20 15:9781 (1987)]. Recent work conducted at the Cornell University Medical College has shown the presence of an additional gene, besides mrr which encodes a protein that restricts TCGA
methylated DNA. Briefly, strains containing a Tn5 (KmR) transposon disrupting the mrr gene were [see J. Bact. 169:3243 (1987)] used for ~679~

transduction [aecording to J. H. Miller in Experiments in Molecular Genetics, Cold Spring Harbor Laboratory, pp 201-205 (1972)] of the KmR marker into several strains of Escherichia coli that resulted in strain converts to a mrr~ (defective mrr protein) genotype. None of 5 these transduced strains could tolerate the Taq methylase gene, indicating there is a second gene responsible for the restri~tion of TCGA methylated DNA. Thus, one of the first necessary requirements (which prior to the present invention had not been apparent) for the making of the present invention was the selection 10 of an E. coli strain which would not heavily restrict TCGA
methylated DNA.
In the present invention, a derivative of the RRI strain of E.
coli which could tolarate the Taq methylase gene and which contained a Tn10 (TcR) transposon was transduced to a ligts7 strain 1~ ~N3098, see Wilson and Murray, J. Mol. Biol. (1979) and J. Mol. Biol.
77:531 (1973)] to create E. coli strain AK76. This strain has been deposited in the American Type Culture Collection, and has been granted the collection number ATCC No. ~5032. This s~rain con~ains a temperature sensitive ligase gene, such that at 42C the strain 20 cannot grow. This strain can tolerate the Taq methylase gene, and other methylated DNA, especially the DNA isolated from T. aquaticus.
Since it also has a temperature sensitive ligase gene, it could be used as a host for the cloning of a functional T. aquaticus ligase gene by selecting for growth at 42C.

2~799:~

Cloning of the T. aquaticus ligase gene was based on a positive selection scheme similar to that described by Wilson and Murray.
The approach was to construct libraries of J. aquaticus DNA inserted into a suitable vector. These libraries were then introduced vi~
S transformation into a lis~ts7 E. coli strain that did not restrict methylated T. aquaticus DNA, such as strain AK76. These cells were then grown at the nonpermissive temperature, that is a~ 42C. Any survivors could be (i) revertants to a ~ phenotype; (ii) second site revertants that increase expression of the defective E. coli ligase 10 gene product; (iii) a cloned piece of T. aquaticus DNA that increases expression of the defective E. coli ligase gene product; or (iv) a cloned piece of T. aquaticus DNA that contains the T. aquaticus ligase gene.
For the desired last alternative to work, it is necessary that 15 (i) the entirP ligase gene is cloned; (ii) that either the endogenous control sequences for T. aquaticus ligase expression function in E.
coli, or that exogenous vector control sequences are sufficiently close to the amino terminus and the ligase gene is cloned in the correct orientation to allow for proper expression in E. coli; (iii) the 20 T. aquaticus ribosorne binding site works in E. coli; and (iv) the T.
aquaticus ligase is active enough at 42C, and the amount synthesized is sufficisnt to complement ligase function in E. coli without intsrfering with other processes.

20~ 799 1 Construction of the suitable libraries used in the present invention utilized conventional vectors containing desired control sequences, and standard restriction endonuclease and ligation techniques. Purified plasmid DNA, T. aq~aticus DNA sequences, or 5 synthesized oligonucleotides for use in the present invention, were cleaved, tallored, and religated in the form desired also by conventional techniques.
The selec~ion of a suitable vector for use in the present invention is more than a mere matter of selecting a vector among 10 the many which exist and have been usecl in the past. High copy number derivatives of pUC plasmids [see for example, C. Yanisch-Peron et al., Gene 33:103 (1985), or J. \~'ieira et al., Gene 19:2~9 (1982)] are actually somewhat unstable when grown at 42C. Low copy plasmids such as pBR322 derivatives pFBI 1, 2, 13, 14 and 15 15 [see F~ Barany, Proc. Natl. Acad. Sci. USA 82:4202 (1g85)] may not produce enough enzyme to complement the ligase defect. In making the present invention, 18 different libraries using 3 different sets of vectors were constructed. The successful clone was derived from the vector pTZ18R [see D. A. Mead et al., Protein Engineering 1:67 20 (1986)], although other vectors may also be utilizable.
Generally, site-specific DNA cleavage, as more particularly described in the following example, is perforrned by treating the [~NA with a suitable restriction enzyme under conditions which are generally understood in the art, and the particulars of which are 2~67~

specified by the manufacturers of these commercially avaiiable restrictions enzymes. In general, about 1 ~g of plasmid or DNA
sequence is cleaved by two to ten units of enzyme in about 20 1ll of buffer solution. Incubation times of about one to two hours at about 5 37C are preferable, although variations in both the tirne and temperature can be tolerated. After each incubation, protein is removed by extraction with phenol/chloroform, and may be followed by a further extraction. The nucleic acids are recovered by precipitation with ethanol. If desired, size separations of the 10 cleaved fragments may be performed by polyacrylamide or agarose gel electrophoresis using standard techniques.

~$7~1 EXAMPLE ll (site specific cleavage) Site-specific cleavage of both plasmid and T. aquaticus DNA
was performed using commercially available restriction 5 endonucleases in standard buffers.
In general, about 10 ~Lg or plasmid or T. aquaticus DNA was cleaved in 100 ~11 of buffer solution by the addition of 20 to 100 units of the appropriate restriction endonuclease, and incubating the mix~ure at 37C for I to 2 hrs.
After each incubation, protein was removed by sequential extractions with phenol (2x), n-butanol (2x), and the nucleic acid was recovered by precipitation with ethanol.

2~991 Construction of suitable vectors containing the desired coding and control sequences employs conventional ligation and res~riction techniques. Briefly, isolated plasmids, DNA sequences, or synthesized oligonucleotides are cleaved, tailored, and religated in 5 the form desired.
The restriction endonucleases utilized for cleavage of the specific libraries used in accordar:ce with the procedure outlined in Example ll were BamHI, ~1, Kpnl (~718), ~I, I lindlll, and Smal, however, other endonucleases or partial digests with ~aulllA, for 10 example, could have been used. Due to adenosine methylation, the commonly utilized restriction endonucleases EcoRI, Sall or XhQI
were no~ used since DNA from ~. aquaticus strain HB8 could not be cleaved by these enzymes.
Restriction fragments resulting from the procedure outlined in 15 Example ll containing 5' overhangs may be blunt ended by fiiling in with DNA polymerase I large (Klenow fragment) in the presence o~
the four deoxynucleotide triphosphates using incubation times of about 15 to 30 minutes at 37C In 50 mM Tris pH 7.6 buffer containing 50 mM NaCI, 10 mM MgC12, 10 mM DTT, and 50 -100 ,uM
2 0 deoxynucleotide triphosphates. The Klenow fragment will fill in at 5' sticky ends. If 3' overhangs are generated, they may be chewed back with mung bean nuclease. After treatment with Kl~now, the mixture is extracted with phenol/chloroform and precipitated with ethanol. Subsequent treatment under appropriate conditions with S1 ~7~ ~

nuclease results in hydrolysis of any single stranded portion. These conventional procedures may be used for cloning any fragment into a (blunt end) site within the vector.

EXAMPLE ill (vector construction) In vector constructions, the linearized vector is commonly treated with a phosphatase enzyme (or alternatively with a second nearby restriction endonuclease) to prevent recircularization of ths vector in the absence of insert DNA. For example, a sample of ~mHI
(5' overhang) or Sacl (3' overhang) DNA (9 ~g) in 150 ~1150 mM Tris HCI buffer at pH 8.0 and containing 10 mM MgCI2 and 6 mM
mercaptoethanol in the presence of Na+ may be treated with Calf Intestine Alkaline Phosphatase (CIAP, 22 units) at 37C for 15 min., followed by incubation at 50C for 30 min. to remove phosphate groups from either 5' or 3' overhangs. Alternatively, Bacterial Alkaline Phosphatase (BAP, 10 units) may be used in 150 ~l 10 ml Tris HCI in the presence of Na+ and M~-~t- and incubating at 60C for about 1 hr. CIAP may be subsequently denahJred by the addition of EDTA and EGTA to chelate divalent cations, and heating to 85C for 15 min. Either CIAP or BAP protein is them removed by sequential extractions with phenol (2x), n-butanol (2x), and nucleic acid recovered by precipitation with ethanol.
2Q The effectiveness of the phosphatase step is assayed by comparison of the number of transformants generated when vector is religated in the absence or presence of insert DNA. Typical results of from 10 to 100 fold more transformations when insert 2~99~

DNA is present is indicative that the vector DNA has been properly phosphatased .

3s 2~679~

EXAMPLE IV
(ligations) Ligations were performed in 30-100 1ll volumes using 1-2 ~Lg linearized and phosphatased vector made as previously describetl. 2-5 4,ug T. aquatlcus DNA cut with a restriction endonucleasegenerating the same ends as the vector, in 50 mM Tris HCI buffer at pH 8.0 and containing 10 mM MgCI2, 1 mM EDTA, 1 mM ATP, 6 mM
mercaptoethanol and from 3 to 7 (Weiss) units of T4 ligase, by incubating at either 4 or 15C overnight. After ligation, EDTA was 10 added, the T4 ligase inactivated by heating the solution to 65C for 15 min., and nucleic acids recovered by ethanol precipitation.
Ligation mixtures were introduced into a suitable host such as E. coti strains RR1, AK53 or AK76 - the last ona suitable for immediate positive selection of the lig~ phenotype - via 15 conventional transformation proc~dures [see Hanahan, J. Mol. Biol.
166:3243 ~1987)]. Transformants were selected by plating on ampicillin (or other drugs such as tetracycline or kanamycin depending upon the plasmid used) containing plates. For positive selection of the lig~ phenotype, AK76 transformants were plated 20 onto SOB plates (made by autoclaving 20 g BactoTM-tryptone, 5 g BactorM-yeast extract, 0.5 g NaC~I, 16 9 BactorM-agar in 1 liter of distilled water adjusted to pH 7.5 with NaOH prior to autoclaving, then adding 20 ml 1 M MgSO4) containing 0.2% maltose, 0.2 mg/ml IPT(; (to induce the lac promoter), and 50 ~g/ml ampicillin (to ~7~9~L

seleot the plasmid-containing cells), and grown overnight at 42C to 4~.5C.

~7991 Libraries ran0ed in size from about 5,000 to 27,000 clones.
Given the general estimate that the bacterial chromosome contains about 2,000 to 4,000 kilobases, and the average insert consisted of 5 to 10 kb, it was apparent that several libraries contained 5 redundant clones.
Mixed plasmid preparations were made from six libraries using conventional techniques [see Methods En ymol. 100:243 (1983)], and introduced into fresh AK76 cells. Transformants from each library were plated on 6 SOB plates (each plate receiving between 30,000 10 and 70,000 clones) and incubated at 42C. One library produced from 11 to 19 exceedingly small colonies per plate; the remaining libraries produced an occasionai lar~e colony.
Inclividual clones were picked, plasmid DNA prepared using conventional techniques [see Anal. Biochem. 114:193 (1981)], and 15 analyzed by restriction digestion. All 12 small clones prnduced a 6.8 kb plasmid containing two ~mHI fragments (1.8 and 2.1 kb respectively~ cloned within the ~HI site of pTZ18R. One such plasmid has been designated pDZ1 as depicted in Figure 1. By calculating back to the original library, (of ~,200 clones), it appears 20 that all pD:Z1 plasmids derived from a single clone. The large colonies contained plasmids close to the size of the original vector.
Therefore, these large colonies are probably revertants of the chromosomal li~ts7 gene which contained any plasmid solely to confer resistance to ampicillin.

2~7991 Retransforming plasmid pDZ1 into AK76 cells, and selecting at 42C on SOB plates containing maltose, IPTG, and ampicillin as described in Example IV, again yielded small colonies. Plating fresh transformants on tryptone yeast agar containing ampicillin did not 5 produce colonies. This result suggests that induction of the lac promoter during plasmid establishment is necessary for production of sufficient quantities of T. aquaticus ligase to complement the genetic defect. Once the plasmid has become established in AK76 cells, such clones will give exceedingly small colonies when 10 streaked and allowed to grow on tryptone yeast plates containing ampicillin at 42~C.
Digestion of pDZ1 with ~HI, followed by religation would scramble the fragments. Transformation of such a ligation mix into AK76, followed by plating at 37C, i.a. under non-selective 15 conditions, compared to plating at 42C, i.e. under selective conditions, yielded 1,000 fold more colonies under non-selective conditions. The starting pDZ1 plasmid yielded only 2 foid more colonies under non-selective than selective conditions. This finding strongly suggests that the presence of both fragments, and the 20 orientation they are cloned, is necessary for proper expression of T.
aquaticlls ligase.
Although pDZ1 contains several Sa~l and Smal sites, it only contains a single (vector derived) ~I, ~.Qnl, or ~ dlll site. Thus, it would have been expected that a number of !igase clones would have ~7~9~

been isolated from the ~I, J~l, or ~liDdlll digest libraries.
However, the only ligase clone was derived from the partial ~mHI
digest library. Although it is not clear why this happened, one conceivable explanation is that other clones did not bring the lac S promoter controlling element sufficiently close to the start of the ligase gene to adequately express the ligase protein during plasmid establishment.
The cloning of T. aquaticvs ligase as described above will now enable those skilled in the art to clone any thermophilic or 10 thermostable ligase, whether of procaryotic, archebacterial, eukaryotic or phage origin by additional approaches. Accordingly the cloning of such ligases are within the scope of the present invention.
Such additional approaches to cloning may include, for example, (i) cloning T. aquaticus DNA into a red- lambda vsctor and 15 screening for the ability of recombinant phage lambda to form pla~ues at 39C on a liÇLts7 strain such a AK76 [essentially as generally described in J. Mol. Biol. 132:471 (1979)]; (ii) use of the lambda gtl1 phage to express portions of the ligase gene, and subsequently screening with antibodies raised to purified T.
20 aquaticus ligase - the positive lambda gtl1 clone may then be used to identify the full length gene by hybridization to other plasmid or phage libraries, essentially as described in the cloning of T.
aquaticus polymerase [see J. Biol. Chem 264:6427 (1989~]; 90(iii) based upon th~ ligase DNA sequence, probes can be made that would ~7~9 ~

hybridize to and therefore help to identify and retrieve other thermostable ligase encoding sequences in a variety of species.
Accordingly, portions of the DNA encoding at least five amino acids ~rom T. aquaticlls ligase can be replicated, or amplffied using PCR
5 techniques, and the denatured or single stranded forms may be use~l as probes to retrieve additional DNAs encoding a thermophilic or thermostable ligase. Alternatively, oli~odeoxyribonucleotide probes can be synth0si7ed which encode at least five amino acids, and these may be used to retrieve additional DNAs encoding a thermophilic or 1 û thermostable ligase.
The selection of a portion of DNA encoding for at least five amino acids is based upon the portion containing fifteen nucleic acid bases which is more than the statistical minimum length that an oligonucleotide should have in order to find a single complementary 15 sequence in a genome. However, portions slightly smaller (the minimum number in E. coli is, for example 12, indicating a portion as small as that encoding for four amino acids may be acceptable) or larger (the minimum number for higher animals is as high as 19, indicating that a portion encoding for at least seven amino acids 20 may be necessary) [see Oligonucleotides: Antisense Inhibitors of Gene Expression, vol. 12, pages 137- 140, Macmillan Press Ltd., London (1989)] may be used to obtain similar results However, because there may not be a precise match between the nucleotide sequence in the corresponding portions between species, oli~omers 2al~7~1 containing approximately 15 nucleotides are a preferred minimum in order to achieve hybriclization under conditions of sufficient stringency to eliminate false positives; the s~quence encoding 5 amino acids would supply information sufficient for the generation 5 of such probes.
By way of example, a comparison of the T. aquaticus ligase and E. coli amino acid sequences reveals an identity betwe~n amino acids 34-40 (Asp-Ala-Glu-Tyr-Asp-Arg-Leu) at statistically acceptable levels. Using the preferred six amino acid sequence, a 10 degenerate probe of the form GA(C/T)-GC(G/A/T/C)-GA(G/A)-TA(C/T)-GA(C/T)-(C/A)G(G/A/T/C)-(C/T)T could be used to identify and retrieve either of the above ligases. The areas of sequence identities between the Thermophilvs li~ase according to the present invention and E. coli ligase include the amino acids at the following 1 S positions:
mino Acid Po~sitions CotLsecutiv~ identical aa's 34 to 40 7 57 to 61 137 to 142 6 2û 168 to 175 8 199 to 210 12 212 to 219 8 308 to 312 5 333 to 339 7 2 5 485 to 490 492 to 4g6 5 2~1~7991 513 to 517 5 620 to 6~4 5 Overall, of the 676 amino acids contained in the ligase according to the present invention, the percent similarity between the Thermophilus ligase and E. coli ligase is 66%; the percent identity is 47%.
The construction of an overproducer strain from a cloned and properly oriented gene may be achieved by using procedures which are conventional in the art. The general principle of such construction is to bring an enabling sequence into close proximity to the starting codon of the gene to affect efficient transcription and translation of that gene. There are many promoter systems (including a ribosome binding site [see Proc. Natl. Acad. Sci. USA
78:5543 (1981)]) that have been successfully used to turn on genes, including the lac promoter, the trp prornc~ter [see Gene 20:231 (1982)~, the lambda phage PL promoter lsee Nature 292:128 (1981)], the tac fusion promoter [see Proc. Natl. Acad. Sci. USA 80:21 (1983~, and the T7 phage promoters [see Proc. Natl. Acad. Sci. USA 82:1074 (1 985~].
Plasmid pDZ1 contains the T. aquaticus ligase gene downstream from both lac and T7 promoters present in the starting vector. There are several methods for removing excess DNA
sequences from between the promoters and the gene, including use of Bal31 [see Nucl. Acids Res. 5:1445 (1978)] and Exolll and Mung 2~7~91 Bean or S1 Nuclease ~see Meth. Enzymol. 155:156 (1987)]. However, a somewhat simpler method as described in Example V was used to bring th~ amino terminus of the T. aquaticus ligase gene oloser to the two promoters in the present instance.

2~7g~

E)CAMPLE V
(removal of excess DNA from between promoter and gene) Plasmid pDZ1 was randomly linearized with the restriction endonuclease ~ PI (G CGC) and blunt ended with Klenow or 5 alternatively with CviJi (PuG CPy) ~see DNA and Protein Engineering Techniques 1 :29 (1988)].
DNA was purified by sequential extractions with phenol (2x), n-butanol (2x), and the nucleic acid recovered by precipitation with ethanol. These randomly linearized plasmids were then treated with 1 0 ~.718 which cleaves the polylinker site directly downstream of the two promoters, and blunt ended with Klenow. The resulting fragments were separated via electrophoresis in low melting agarose, sequential slices ~including full length linear and progressively smaller DNA fragments) excised, and the DNA
15 recovered. The DNA fragments were subsequently recircularized by blunt end ligation. This involved overnight incubation at 4C in 1ûG
1 in 50 mM Tris HCI pH 8.0 buffer containing 10 mM MgC12, 1 mM
EDTA, 1 mM ATP, 6 mM mercaptoethanol, and from 3 to 7 Weiss units of T4 ligase. After ligations, EDTA was added, the T4 ligase 20 inactivated by heat (for 15 min at 65C), and nucleic acicls recovered by ethanol precipitation.
The ligation mixes prepared were introduced into AK76 cells using conventional techniques, and the lig~ phenotype was selected 2~799~

at 42C on SOE~ plates containing maltose, IPTG, and ampicillin as described previously.

2~67~

Based upon previous work, plasmids containing deletions between the promoters and the start of the T. aquaticus ligase ~ene would be expected to confer viability under these conditions.
Deletions of the vector (promoter regions), or of an essential portion 5 of the ligase gene should nct confer viability. There~ore, individual clones were picked, plasmid DNA prepared using conventional methods [see Anal. Biochem. 114:193 (1981)], and analyzed by restriction digestion. Results from this testing found that plasmid pDZ2, pDZ3, pDZ6 and pDZ7 lacked the 1.8 kb ~mHI
10 fragment, and contained instead a 1.3, 1.4, 1.2, or 1.2 kb fragment, respectively. All these plasmids re-created the Asp718 site as would be expected with proper blunt end fill-ins and ligations.
Single stranded DNA was prepared from these plasmids using conventional techniques [see Nucl. Acids Research 13:1103 (1985), 15 and Protein Ençlineering 1:64 (1986)], anci these were sequenced usin~ the universal "reverse primer" oligonucleotide 5'd(AGCGGATAACAATTTCAGACAGGA)3' and T7 DNA polymerase [see Proc~ Natl, Acad~ Sci~ USA 84:4767 (1987)].
Analysis of the DNA sequence reveals two ATG start codons, 20 the first open reading frame being three codons in length and the second, the ligase DNA sequence, giving a long readin~ frame. In conjunction with Figure 1, this sequence (including the partial ligase DNA sequence) derived from plasmids pDZ6 and pDZ7 is:

2~7~

pTZ18R
GGCTCGTATG TTGTGTGGAA TTGTGAGCGG ATAACAATTT
La~Z' T7 Promoter CACACAGGAA ACAGCTATGA CCATGATTAC GAATTTAATA
pDZ6,7 CGACTCACTA TAGG~a~Ç GAGCTCGGTA CCCC~AGGTA
EcoRI SacI ~I
CACTAGGGCC
thermophilic ligase ATG ACC CTG G~A GAG GCG AGG AAG CGG GTA AAC G~G TTA CGG GAC
CTC ATC CGC TAC CAC AAC TAC CGC TAC TAC GTC CTG GCG GAC CCG
GAG ATC TCC GAC GCC GAG TAC GAC CGG CTT CTT AGG GAG CTC AAG
GAG CTT GAG GAG CGC TTC CCC GAG CTC A~A AGC CCG GAC TCC CCC
ACC CTT CAG GTG GGG GCG AGG CCT TTG GAG GCC ACC TTC CGC CCC
GTC CGC CAC CCC ACC CGC ATG TAC TCC TTG GAC ARC C,CC TTT AAC
CTT GAC GAG CTC AAG GCC TTT GAG GAG CGG ATA GAA CGG GCC CTG
GGG CGG AAG GGC CCC TTC GCC TAC ACC GTG GAG CAC AAG GTG GAC
GGG CTT TCC GTG AAC CTC TAC TAC GAG GAG GGG GTC CTG GTC TAC
GGG GCC ACC GCC GGG GAC GGG GAG GTG GGG GAG GAG GTC ACC CAG
AAC CTC CTC ACC ATC CCC ACC ATC CCG AGG AGG CTC AAG GGG GTG
CCG GAG CGC CTC GAG GTC CGG GGG GAG GTC TAC ATG CCC ATA GAG
GGC TTC CTC CGG CTC AAC GAG GAG CTG GAG GAG CGG GGG GAG AGG
ATC TTC A~A AAC CCT AGG AAT GCG GCG GCG GGT TCC TTA AGG C~A
A~A GAC CCC CGC ATC ACC GCC AAG CGG GGC CTC AGG GCC ACC TTC

GCG ACC CAG TTT GCC CTC CTC CAC TGG CTC AAG GAA A~A GGC TTC
CCC GTG GAG CAC GGC TAC GCC CGG GCC GTG GGG GCG G~A GGG GTG
GAG GCG GTC TAC CAG GAC TGG CTC ARG AAG CGG CGG GCG CTT CCC
TTT GAG GCG GAC GGG GTG GTG GTG AAG CTG GAC GAG CTT GCC CTT
TGG CGG GAG CTC GGC TAC ACC GCC CGC GCC CCC CGG TTC GCC ATC
GCC TAC AAG TTC CCC GCC GAG GAG AAG GAG ACC CGG CTT TTG GAC
GTG GTC TTC CAG GTG GGG CGC ACC GGG CGG GTG ACC CCC GTG GGG

~8 2~7~91 ATC CTC GAG CCC GTC TTC CTA G~G GGC AGC GAG GTC TCC CGG GTC
ACC CTG CAC A~C GAG AGC TAC ATA G~G GAG TTG GAC ATC CGC ATC
GGG GAC TGG GTT TTG GTG CAC A~G GCG GGC GSG GTC ATC CCC GAG
GTC CTC CGG GTC CTC AAG GAG AGG CGC ACG GSG GAG GA~ AGG CCC
S ATT CGC TGG CCC GAG ACC TGC CCC GAG TGC GGC CAC CGC CTC CTC
AAG GAG GSG A~G GTC CAC CGC TGC CCC AAC CCC TTG TGC CCC GCC
AAG CGC TTT GAG GCC ATC CGC CAC TTC GCC TCC CGC AAG GCC ATG
GAC ATC CAG G~C CTG GSG GA~ AAG CTC ATT GAG AGG CTT TTG GAA
AAG GGG CTG GTC AAG GAC GTG GCC GAC CTC TAC CGC TTG AGA AAG
GAA GAC CTG GTG GSC CTG GAG CGC ATG GGG GAG AAG AGC GCC CAA
AAC CTC CTC CGC GAG ATA GAG GAG AGC AAG A~A AGA GSC CTG GAG
CSC CTC CTC TAC GCC TTG GGG CTT CCC GGG GTG GSG GAG GTC TTG
GCC CGG AAC CTG GCG GCC CGC TTC GGG AAC ATG GAC CGC CTC CTC
GAG GCC AGC CTG GAG GAG CTC CTG GAG GTG GAG GAG GTG GGG GAG
CTC ACG GCG AGG GCC ATC CTG GAG ACC TTG AAG GAC CCC GCC TTC
CGC GAC CTG GTA CGG AGG CTC A~G GAG GCG GSG GTG GAG ATG GAG
GCC AAG GAG AhG GSC GSG GAG GCC CTT A~A GGG CTC ACC TCC GTG
ATC ACC GGG G~G CTT TCC CGC CCC CGG G~A GAG GTG AAG GCC CTC
CTA AGG CGC CTC GSG GCC AAG GTG ACG GAC TCC GTG AGC CGG AAG
ACG AGC TAC crc GTG GTG GGG GAG AAC CCG GGG GAG AAC CCG GGG
AGC AAG CTG GAG AAG GCC AGG GCC CTC GGG GTC CCC ACC CTC ACG
GAG GAG GAG CTC TAC CGG CTC CTG GAG GCG CGG ACG GGG AAG A~G
GCG GAG GAG CTC GTC T.~A AGGCTTCC
The nucleic acid sequence for the thermophilic ligase 25 according to the present invention correspcnds to the amino acid sequence:
Met Thr Leu Glu Glu Ala Arg Lys Arg Val Asn Glu Leu Arg Asp Leu Ile Arg Tyr His Asn Tyr Arg Tyr Tyr Val Leu Ala Asp Pro Glu Ile Ser Asp Ala Glu Tyr Asp Arg Leu Leu Arg Glu Leu Lys Glu I~u Glu Glu Arg Phe Pro Glu Leu Lys Ser Pro Asp Ser Pro ~Q~79~

Thr Leu Gln Val Gly ALa Arg Pro Leu Glu ALa Thr Phe Arg Pro Val Arg His Pro Thr Arg Met Tyr Ser Leu Asp Asn ALa Phe Asn Leu Asp Glu Leu Lys ALa Phe Glu Glu Arg Ile Glu Arg Ala Leu Gly Arg Lys Gly Pro Phe Ala Tyr Thr Val Glu His Lys Val Asp S Gly Leu Ser Val Asn Leu Tyr Tyr Glu Glu Gly Val Leu Val Tyr Gly Ala Thr Arg Gly Glu Gly Glu Val Gly Glu Glu Val Thr Gln Asn Leu Leu Thr Ile Pro Thr Ile Pro Arg Arg Leu Lys Gly Val Pro Glu Arg Leu Glu Val Arg Gly Glu Val Tyr Met Pro Ile Glu ALa Phe Leu Arg Leu Asn Glu Glu Leu Glu Glu Arg Gly Glu Arg Ile Phe Lys Asn Pro Arg Asn ALa ALa ALa Gly Ser Leu Arg Gln Lys Asp Pro Arg Ile Thr Ala Lys Arg Gly Leu Arg Ala Thr Phe Tyr ALa Leu Gly Leu Gly Leu Glu Glu Val Glu Arg Glu Gly Val ALa Thr Gln Phe ALa Leu Leu E~is Trp Leu Lys Glu Lys Gly Phe Pro Val Glu His Gly Tyr Ala Arg Ala Val Gly Ala Glu Gly Val Glu Ala Val Tyr Gln Asp Trp Leu Lys Lys Arg Arg Ala Leu Pro Phe Glu Ala Asp Gly Val Val Val Lys Leu Asp Glu Leu Ala Leu Try Arg Glu L,eu Gly Tyr Thr Ala Arg .Ala Pro Arg Phe Ala Ile Ala Tyr Lys Phe Pro Al.a Glu Glu Lys Glu m r Arg Leu Leu Asp Val Val Phe Gln Val Gly Arg rrhr Gly .Arg Val m r Pro Val Gly Ile Leu Glu Pro Val Phe Leu Glu Gly Ser Glu Val Ser Arg Val m r Leu His Asn Glu Ser Tyr Ile Glu Glu Leu Asp Ile Arg Ile Gly Asp Trp Val Leu Val His Lys Ala Gly Gly Val Ile Pro Glu Val Leu Arg Val Leu Lys Glu Arg Arg Thr Gly Glu Glu Arg Pro Ile Arg Trp Pro Glu Thr Cys Pro Glu Cys Gly His Arg Leu Leu Lys Glu Gly Lys Val His Arg Cys Pro Asn Pro Leu Cys Pro Ala Lys Arg Phe Glu Ala Ile Arg His Phe Ala Ser Arg Lys Ala Met Asp Ile Gln Gly Leu Gly Glu Lys Leu Ile Glu Arg Leu Leu Glu Lys Gly Leu Val Lys Asp Val Ala Asp Leu Tyr Arg Leu Arg Lys Glu Asp Leu Val Gly Leu Glu Arg Met Gly Glu Lys Ser Ala Gln 3 0 Asn Leu Leu Arg Gln Ile Glu Glu Ser Lys Lys Arg Gly Leu Glu Arg Leu Leu Tyr Ala Leu Gly Leu Pro Gly Val Gly Glu Val Leu Ala Arg Asn Leu Pla Ala Arg Phe Gly Asn Met Asp Arg Leu Leu 2~7~91 Glu Ala Ser Leu Glu Glu LRU Leu Glu Val Glu Glu Val Gly Glu Leu Thr Ala Arg Ala Ile Leu Glu Thr Leu Lys Asp Pro Ala Phe Arg Asp LRU Val Arg Arg Leu Lys Glu Ala Gly Val Glu M~t Glu ~la Lys Glu Lys Gly Gly Glu Ala Leu Lys Gly Leu Thr Phe Val S Ile Thr Gly Glu Leu Ser Arg Pro Arg Glu Glu Val Lys Ala Leu Leu Arg ~rg Leu Gly Ala Lys Val Thr Asp Ser Val Ser Arg Lys Thr Ser Tyr Leu Val Val Gly Glu Asn Pro Gly Ser Lys Leu Glu Lys Ala Arg Ala Leu Gly Val Pro Thr Leu Thr Glu Glu Glu Leu Tyr Arg Leu LRU Glu Ala Arg Thr Gly Lys Lys Ala Glu Glu Leu 1 0 Val Translation of the first 60 amino acids of this open reading frame (the thermophilic ligase) shows better than 50% homology to E. coli iigase [see Mol. Gcn. Genet. 204:1 (1986)] suggesting that this long open reading frame represents the start of the T. aquaticus 15 gene. From the genetic results with the !3amHI fragments, one can conclude that the size of this ligase is between 400 and 1,100 amino acids in length. The purified protein has been reported to have a molecular w~ight of about 79,000 [see J. Biol. Chem. 259:10041 (1984)] which is within the limits of the genetic results found for 20 the present invsntion. Given that clone pDZ7 produces functional ~.
aquaticus li~ase (that is it encodes the ~ene in its entirety)~ and given the DNA sequence of the amino terminus, the entire DNA
sequence of the gene was determined using either manual or automated methods as described in the literature [ses, for example, 25 Proc. Natl. Acad. Sci. 84:4767 (1987); Proc. Natl. Acad. Sci. 86:4076 (1989); Science 239:487 (1987); Nature 321:674 (1986);

Biotechniques 8:184 (1990); Proc. Natl. Acad. Sci. USA 85:5610 (1988); and Proc. Natl. Acad. Sci. USA 85:9436 (1988)].
Plasmids pDZ2, pDZ3, pDZ6 or pDZ7 may be used to construct further overproduction vectors using methods common to those 5 skilled in biotechnology studies. This may include using promoters and ribosome binding sites as described above. For example, plasmid pDZ7 (see figure 1) may be linearized at its unique As~718 site, and excess nucleotides in front of the ~. aquaticus ligase gene trimmed close to the ATG start codon by the use of Bal31 or a combination of 10 .EXQIII and Mung Bean or S1 I\luclease as described above. This may then be blunt end ligated to a natural enabling se~uence (a promoter and translation start sequence) generatecl in a similar manner, or by a synthetic enabling sequence manufactured for this purpose. In addition, sequences external or internal to the T. aquaticus gene 15 may be modified to remove potential RNIA structures that may inhibit transcription OT translation. Thes~e methods have been reported previously to a~fect overproduction of the thermophilic restriction endonuclease Taq I to greater than 30% of soluble E. coli proteins [see Gene 65:166 (198~)]. Alternatively, synthetic 20 oligonucleotides may be synthesized such that the start of the T.
aquaticus ligase gene is fused directly to an enabling sequence using PCR methods [see, for exarnple, Biotechniques 8:17~ (1990);
Gene 77:51 (1989); and Nucl. Acids Res. 17:723 (1989)].

2~g79~

From the preceeding sequences, it can be seen that there is a Bgl ll site corresponding to the nucleotides that code for amino acid residues 31-33. With this information, a strong prornoter with an optimal Shine-Dalgarno sequence could be inserted in front of this 5 gene using PCR. Two minor caveats need to be considered (1) attempts to PCR copy the entire gene (3 kb, high GC content) were not always successful, and (2) plasmid pDZ7 had two Bam Hi and Bgl Il sites, one each within the ligase gene.
Plasmid pDZ7 was partially digested with both Bam Hl and Bgl 10 Il, the correct size smaller linear fragment separated from full length linear by electrophoresis, excised, and purified as described previously. Since Bam Hl and B~l ll produce the same overhang (5' G~TC), the linear fragment could be recircularized with T4 ligase, and introduced into E. coli strain AK53 via transformation. Several 15 clones had deleted the 0.5 kb Bam Hl/Bgl ll fragment resulting in a 5.7 kb plasmid, and one such clone w~s designate~ pDZ12. Synthetic oli~onucleotides #66, #78, #85l and #94 were synthesized, to allow for fusion of pho A
promoter [from plasmid pFBT64; see Gene 56:13 (1987)] and 20 ribosome binding sequence to the start of the ligas~ gene using PCR
[see Biotechniques 8:178 (1990); Gene 77:51 (1989~; Gene 77:61 (198g); and Nuol. Acids Res.17:723 ~1989)]. These clones are depicted in Fig. 9, and are:

2~79~1 #66 19 mer; Pvu ll site to T7 promoter through phoA
promoter, top strand of plasmid pFBT64 (direction of Taql endonuclease gene):
5' CTG GCT TAT CGA AAT TAA T 3' #78 32 mer; 5' end complementary to start of Thermus ligase gene; 3' end complementary to Shine-Dalgarno side of pho A promoter, bottom strand of plasmid pFBT64:
5' CCA GGG TCA TTT TAT TTT CTC CAT GTA CAA AT 3' #85 33 mer; 5' end complementary to Shine-Daigarno side of pho A promoter; 3' end cornplementary to start of r hermus ligase gene, top strand of plasmid pDZ7 (direction of ligase gene):
5' CAT GGA GAA AAT AAA ATG ACC CTG GAA GAG GCG 3' #94 18 mer; bottom strand of plasmid pDZ7 corresponding to non-translated strand of amino acid residues 40 to 35 of ligase gene, downstream of Bgl ll site at amino acid residues 33 to 31:
5' MG CCG GTC GTA CTC GGC 3' Briefly, this was accomplished in a single reaction tube in which 400 ng of primers #66 and ~78 were added to 200 ng of Pst i/Pvu li digested pFBT64 containing 50 ~rnoles of clATP, cCTP, cGTP, and dTTP each, and 2.5 units Amplitaq in 100 1ll PCR buffer and cycled at 94C for 1 min, 55C for 2 min, 7~C for 3 min with 3 2~7~91 sec. extension per cycle for 25 cycles as per the manufacturer's (Cetus, Emoryviile, California) protocol. A second reaction tube contained 400 ng of primers #85 and #941 200 ng of Eco Rl/Bam Hl digested pDZ7, in the same reaction buffer and 5 enzyme, and incubated as above. The products of these reactions were shown to be the correct length as analyzed by gel electrophoresis. A third reaction tube contained 2~11 from each product, 400 ng primers #66 and #94 in the sarne reaction buffer and en~yrne, and incubated as above. Primers were designed such 10 that overlap between the two products would allow for PCR
synthesis of the combined length fused product. The resultant fragment was extracted with phenol, n-butanol, and ethanol precipitated to rernove Taq polymerase. The product PCR fragment was treated with Bgl 11 and Eco Rl, electrophoresed in low melting 1~ agarose, and purified as described above. Meanwhile, the 2.7 kb Pst l-Bgl 11 ligase ~ene containing fragment from pDZ12 and the 2.4 kb Pst l-Eco Rl 13-lactamase gene and origin containing fragment from pFBT64 were purified. All three fragments were combined in a three way ligation and introduced into E. coli strain AK53 via 20 transformation. Several clones contained a 5.5 kb plasmid which overproduced ligase under pho A promoter control. One such plasmid has been designated pDZ13.
In reported studies in overproduction of the thermophilic restriction endonuclease Taq I to greater than 30% of soluble E. coli 9 ~ ~

proteins [see Gene 6~:166 (1988)], it was noticed that endonuclease yields were somewhat better if the ~-lactamase gene was reversed, and hence transcribing in the opposite direction as the pho A
promoter. To make a similar construction with the ligase gene S accordin~ to the present invention, the 2.3 kb Pst l-Pvu ll fragment from plasmid pFBLT69 (which contains the B-lactamase in reverse orientation) was ligated to the 3.2 kb Pst l-Pvu ll ligase ~ene containing fragment of plasmid pDZ13. The ligation mix was transformed into E. coli strain AK53, and several 10 transformants were analyzed by restriction digests to confirm the orientation of û-lactamase gene. One such clone has been designated pDZ15. Production of ligase in pDZ15 is as ~ood as, if not slightly better than, pDZ13. The ligase enzyme appears to be somewhat sensitive to proteases, and thle cells should be grown for 15 no more than 9 hours after induction. Proteolytic products of the ligase gene may still have thermostable ligase activity (this has been demonstrated for Taq polymerase).
- Thermophilic proteins may be subs~antially rnodified and stili retain sufficient activity for use in the present invention. For 20 example, it has been shown that deletion of approximately one-third of the coding sequence at the amino-terminus of ~aq polymerase still produces a gene product that is active in polymerase activity [see J. Biol. Chem. 264:6427 (1989)]. Alternatively, another thermophilic protein, the restriction endonuclease Taq 3, was shown 2~7~

to retain essentially full activity when amino acids were added to the amino-terminus (+7), the carboxy-terminus (~38~, or at certain positions internally ~from +2 to ~34) [see Gene 65:166 (1988)].
Thus, modification of the primary structure by deletion, n-terminus 5 addition, c-terminus addition, internal addition or duplication, or alteration of the amino acids incorporated into the sequence during translation can be made without destroying the activity or thermostable nature of the protein. In addition, the availability of DNA encoding these sequences provides the opportunity to modify 10 the codon sequence so as to generate mutein forms also having ligase activity. Such substitutions or other alterations result in novel proteins having amino acid sequence encoded by DNA falling within the scope of the present invention.
It will also be appreciated that other ligating proteins may be 15 isolated by the process according to the present invention as exemplified in these examples. Different cell lines may be expected to produce ligases having different physical prop~rties to that isolated from the T. aquaticus HB8 strain used in the making of the present invention. Additionally, variations may exist due to genetic 2 0 polymorphisms or cell-mediated modifications of the enzyme or its precursors. Furthermore, the amino acid sequence of a ligase so isolated may be modified by genetic techniques to produce ligases with altered biological activities and properties. The resultant DNA
sequence may then be able to encode a protein having substantia!ly ~7~91 the same amino acid sequence as T. aquaticus HB8 ligase, but exhibiting a higher or lower level of activity. Such ligating proteins should also be considered to be within the scope of the present invention .

s~

2~7~9~

EXAMPLE Vl (purification of ligase enzyme) E. coli cells AK53 containing plasmids pDZ6 and pGP1-2 (containing the T7 RNA polymerase gene behind the lambda PL
S promoter and under control of the temperature sensitive lambda repressor Cl~87) [see Proc. Natl. Acad. Sci. USA 82:1074 (1985) and United States Patent 4,795,699], were grown overnight at 32C on TY plates containing ampicillin at 50 llg/ml and kanamycin at 50 ~g/ml to ensure maintenance of both plasmids. Fresh colonies were 10 resuspended in 1 liter of sterile 50 mM Tris HCI buffer at pH 7.~ and containing 6 g NaCI, 25 ~ BactoTM-tryptone, 7.5 g yeast extract, 1 g glucose, 1.6 g casein amino acid hydrolysate, 50 ~l~/ml kanamycin and 50 ,ug/ml ampicillin, and grown at 32C in a 2 liter flask shaking at 200 rpm. When the O.D.sso reach~d between 0.8 and 1.0, 15 synthesis of the T7 polymerase was induced by shifting the cells to 42C for 30 to 40 minutes. Further synthesis of E. coli proteins were inhibited by the addition of 5 ml of 20 mg/ml rifampicin dissolved in methanol to a final concentration of 100 ,ug/ml. Under these conditions, only genes behind the T7 promoter should be 20 transcribed and hence translated. Cells were incubated for an additional 5 hours at 42C.
Alternatively, E. coli cells AK53 containing plasmids pDZ15 (ligase under pho A promoter control) were grown overnight at 37C
on TY plates containing ampicillin at 50 ~lg/ml. Fresh coionies were s9 2~7~

resuspended in 50 ml of fortified broth containing 50 ~lg/ml ampicillin and grown at 37C in a 500 ml flask shaking at 200 rpm in a G76 benchtop shaker. When the O.D.soo reached between 0.65 and 0.85, 20 ml was diluted into 1 liter of MOPS media containing 0.2 mM K2HPO4 ~see J. Bacteriology 119:736 (1974)] to induce the phoA promoter. Cells wer0 grown at 37C in a 2 liter flask shaking at 200 rpm in a G25 floor shaker for an additional 9 hours.
Following incubation, the oells were chilled in ice, harvested by centrifugation (5,000 rpm for 15 min), resuspended in 20 ml of water, transferred to 35 ml centrifuge tubes, recentrifuged (7,000 rpm for 6 min), and the pellet frozen until ready for protein isolation. After thawing, the pellet was resuspended in 20 ml of buffer A (20 mM Tris ~ICI buffer at pH 7.6 containing 1 mM EDTA) containing 10 mM 2-marcaptoethanol and 0.15 mM PMSF. After sonication (5 x 1 min at 50% power at 4C), the solution was centrifuged at 39,000 x g for 60 min.
The en~yme has an estimated molecular weight of from 75,000 to 85,000 daltons when compared with a phosphorylase B standard assigned a molecular weight of 92,500 daltons.
2 0 Alternatively, 2 liters of pDZ15 induoed cells were harvested, sonicated, and debris cleared by centrifugation as clescribed above.
The supernatant ~40 ml) was brought to 300 mM KCI and passed through a 5 ml DEAE sephacel column to remove extraneous DNA
using 70 ml buffer A containing 0.3 M KCI. The flowthrough 2~7991 fractions containing the ligase were combined, and treated at 65C
for 20 minutes to irreversably heat denature many E. coli enzymes including endo or exonucleases. Denatured proteins were then removed by centrifugation at 39,0ûO x g for 15 minutes, and the ligase enzyme precipitated ~rom the supernatant by adding an equal volume of saturated (NH4)2SO4 at room temperature for 30 minutes.
The ammonium sulfate precipitate was harvested by centrifugation at 8,000 rpm in a clinical centrifuge, and resuspended in 4 ml of distilled water. Samples were dialyzed against buffer A, followed 10 by buffer A containing 50 mM KCI. The dialized protein solution was applied to a 40 ml phosphocellulose column equilibrated with buffer A containing 50 mM KCI. After washing with 80 ml of the same buffer, the column was eluted with a 121) ml linear gradient of KCI
(0.05 to 0.5 M) in buffer A. The enzyme e~luted as a sharper peak 15 from 0.25 to 0.35 M KCI. The protein migrates as two bands of apparent molecular weight approximately 81,000 (adenylated form) and 78,000 (non-adenylated form) and is about 98-99% pure as monitored by SDS-10% polyacrylamide gel electrophoresis. One can convert between the two forms by incubating 150 ~Lg protein in 20 ligase buffer containing either 25~g nicked Salmon sperm DNA
without NAD (resulting in the non-adenylated form), or in ligase buffer with 10 mM NAD (resulting in the adenylated form) for 30 min at 65C. An equal volume of 20 mM Tris HCI pH 8.0 in 100% glycerol containing 1 mM EDTA, 2 mM dithiothreitol (DTT), and 200 ~g/ml 6 l ~79~

Bovine Serum Albumin (Fraction V) is added (final glycerol concentration is 50%), and enzyme stored at either -70C or -20C.
From 2 liters of cells, a final yield of 6 mg ligase in 16 ml storage buffer, at 625 nick closing units per microliter. This corresponds to a total of 10,000,000 units of enzyme, and a specific activity of 1,666,667 units/mg.

~7~

Since it is known that thermophilic proteins tend to be somewhat more hydrophobic than their mesophilic counterparts, addition of non-ionic detergents or other stabili~ing agents may help in long term storage. Storage buffers may therefore include additional components such as glycerol ~50%), sucrose (25%~, protease inhibitors (0.5-1.0 mM PMSF, 10-7 M pepstatin A), salt (KCI, preferably at 100-500 mM), EDTA (0.1-1.0 mM) bovine serum albumin (100-500 ~g/ml), gelatin, dithiothreitol ~1-10 mM), and mercaptoethanol (1-10 mM). In addition, it is preferable that the 10 storage buffer contain at least one non-ionic polymeric detergent. A
partial listing of such detergents would include ethoxylated fatty alcohol ethers and lauryl ethers, ethoxylated alkyl phenols, polyethylene ~Iycol monooleate compounds, and more particularly Triton X-100, NP-40, and Tween 20 at 0.1-0.5% vol/vol.
To assay for ligase activity, it is irnportant to use a method that is not skewed by the melting temperature (Tm) of the substrates. For example, a 4 base cohesive end ligation is most efficient at a low temperature such as 4C, well below the temperature optimum for T4 ligase (which is 37C), and certainly 20 below the temperature optimum of a thermophilic ligase. C)ne assay method that should be consistent is the nick-closing assay in which circular plasmid DNA is randomly nicked in several places by DNasek The ability of ligase to close all these nicks and ~enerate covalently closed circular DNA can be assayed by separating nicked circle from ~679~1 open circle DNA via electrophoresis in an agarose gel containing ethidium bromide. For example, the covalently closed circular form of plasmid pUC4KiXX [see Gene 37:111 (1985)~ migrates faster than the linear form, and considerably faster than the nicked form on a 1% agarose gel containing 0.2 Al glycine NaOH pH 8.5 0.1 mM EDTA, and 1 llg/ml ethidium bromide and run at 150V for 1.5 hr in ths same buffer.

206799~

EXAMPLE Vll (thermophilic ligase assay) Nicked pUC4KlXX DNA was generated by adding 3 ~11 of freshly diluted 1 llg/ml DNasel to 5 ~ DNA in 50 ~l of 50 mM Tris HCI ph 8.0 buffer containing 10 mM MgCI2, 1 mM EDTA, and 6 mM
mercaptoethanol. The mixture was incubated at room temperature for 5 min, the DNase heat killed at 65C for 10 min, and the sample stored until used be freezing at -20C. Under these conditions, about 90% of the DNA was in the nicked circular form, with about 5%
in the linear and 5% in the covalently closed circular Form.
Thermophilic li~ase prepared as above was assayed by adding seriai dilutions of ligase to 0.5 llg nicked pUC4KlXX in 20 ,ul of 20 rnM Tris HCI pH 7.6 buffer containing 5û mM KCI, 10 mM MgC12, 1 mM
EDTA, 10 ml\l NAD, 10 mM dithiothreitolt overlaying with a drop of rnineral oil, and incubating at 65C for 15 min. As a control, T4 ligase was assayed by adding serial dilutions of ligase to 0.5 ~g nicked pUC4KlXX in 20 ~11 of 50 mM Tris HCI pH 8.0 buffer containing 10 mM MgCI2, 1 mM EDTA, 1 mM ATP, 6 mM mercaptoethanol, and incubating at 37C for 15 min.
Reactions were terminated by the addition of 4 1ll stop buffer containing 0.2 M EDTA, 50% glycerol, 1% SDS and û.1% bromphenol blue, and the products were analyzed by gel electrophoresis as described above.

6s 2 ~ 9 1 One nick closing unit of ligase is defined as the amount of ligase that circularizes 0.5 ,ug of nicked pUC4KlXX DNA under the buffer and time conditions set forth in the preceding example, such that addition of further li~ase does not circularize additional DNA.
As a mini-prep procedure, E. coli cells AK53 containing plasmids pD~15 (ligase underphoA promoter control) were grown overnight at 37C on TY plates containing ampicillin at 50,ug/ml.
Fresh colonies were resuspended in 5 ml of fortified broth containing 50 ~g/ml ampicillin, and grown at 37C. When the O.D.sso reached between 0.65 and 0.85, 0.12 ml was diluted into 6 rnl of MOPS media containing n.2 mM K2HPO~ to induce the pho A
promoter. Cells were incubated overnight at 37C (some proteolysis that occurs after prolonged incubation, SO caution is advised in overgrowing induced cells). Cells were harvested in 1.5 ml microcentrifuge tubes, resuspended in 0.3 ml of 20 mM Tris HCL pH
7.6 containing 1 mM ~DTA and 10 mM 2-mercaptoethanol, and sonicated 2 x 10 seconds. After clearing debris by centrifugation (12,0Q0 rpm for 2 min.), the supernatant was treated at 65C for 20 min to irreversably heat denature many E. oo/i enzymes including the endo and exonucleaseses [see Gene 56:13 (1987)]. The denatured debris was removed by centrifugation and the supernatant assayed as described above. One microliter of this supernatant contained approximately f~25 nick closing units of activity.

2~7~

The r aquaticus ligase preparation described in the preceding examples, as well as commercially available T4 ligase, were shown to contain approximately 125 nick closing units per microliter.
Thus, from 1 liter of E. coli cells overproducing T. aquaticus ligase, S the process according to the present invention has purified approximately (800 x 125) 100,000 nick closing units of enzyme.
The thermophilic ligase prepared according to the precedin~
description has a number of valuable properties which makes it especially useful as an assay that both amplifies DNA and allows it 10 to discriminate a single base substitution in a DNA sequence. The single most important property of this ligase allowing for these uses is that the ligase retains activity during repeated thermal denaturation/renaturation cycles thus allowing for the amplification of DNA without necessitating repeated addition of l S ligase. In addition, the ligase accordiny to the present invention will ligate oligonucleotides of a length which is sufficient to assure their uniqueness in compiex genomes at or near the Tm temperatures of 65C, and will also accurately discriminate between exactly complementary and single based mismatched oligonucleotide 2 0 sequences.
In the simpler of the two procedures developed as a result of cloning the thermophilic ligase DNA sequence, termed a ligase deteotion reaction (LDR~, two oligonucleotide probes are allowed to hybridize to denatured DNA such that the 3' end of one is ~7~

immediately adjacent to the 5' end of the othsr. The oligonucleotides are chosen to be sufficiently long (20 to 25 nucleotides) such that each will preferentially hybridize to its unique position in the human genome. A thermophilic ligase can then 5 form a covalent phosphodiester bond between the two oligonucleotides, provided that the nucleotides at the junction are perfectly complementary to the target. The specificity of this nick-closing reaction is particularly enhanced by virtue of performing the ligation at or near the Tm of the two oligonucleotides for their 10 target. Thus, a single base mismatch at the junction not only forms an imperfect double helix, but also destabilizes the hybrid at the higher temperature. Consequently, thermophilic ligase will efficiently link correctly base paired oligonucleotides and give near zero background ligation in the presence of the imperfectly matched 15 sequences. Using LDR, the amount of product obtained in the ligation reaction can be increased in a linear fashi3n by repeated thermal cyciing .
in the thermophilic ligase chain reaction according to the present invention, both strands serve as targets for oligonuclentide 2 0 hybridization. By using an additional two oligonucleotides complementary to the opposite strand, the ligation products of one cycle become the targets for the next cycle of ligation as generally depicted in figure 2. For each adjacent oligonucleotide pair, the diagnostic nucleotide is on the 3' side of the junction. Thus, 2~799~

aberrant target independent ligation of complementary oligonucleotides is avoided by use of temperatures near the Tm, and by taking advantage or the poor ligation efficiency of single base 3' overhangs. Using ligase chain reaction, the amount of product can be S increased in an exponential fashion by repeated thermal cycling.
In order to test the potential of the thermophilic ligase chain reaction (LCR), the gene encoding human B globin was selected as an initial model system to test the technique of the present invention.
Previous work has determined that the normal û~ allele and sickle 10 13S ailele differ by a single A-~T transversion of the second nucleotide in the sixth codon of the 13 globin gene, changing a glutamic acid residue into a valine in the hemoglobin 13 chain according to the following Table 1:

TABLE I

~D ~b ~ ~ ~ ~ ~ # ~ U '2 0 o o o :, ~ p,~ n.
~ n ~ n " ~-X
,~ , ~ ~ . . .
n ~
C ~ C~ . . .
M ~
r . c~ t~. . .

,~ . n ~
o ~ ~ -@ ~P~
C~ ~ C~ -. . . c~ Q .
:~ n ct -p . ~ . C~ ~ .
Q -J

p ~ ~ C'l ~_ >~
~ C~
,_ _ n ~ ~ ~ ~ ~ W V~ -- ~
~ B ~ a ~ a ~ N

a O : O
O O ~ O ~ O O O Ea r~ n c~ c~n ~ n ~

2~7~

In the following continuation of Table 1, presents the oligonucleotide sequences listed in the preceding portion in their conventional ~' --> 3' orientation:

Sequence Sequence size Tm no.5' ~ 3' (mer) (C3 101C ;T C ATG GTG CAC CTG ACT CCT GA 23 6 6 1û 103GTTI I T C ATGGTGCACCTGACTCCTGG 27 6 4 106 CTl~TT GC AGT AAC GGC AGA CTT c rc cc 2 8 6 6 109 C AGG AGT CAG GTG CAC CAT G~T 2 2 7 0 Oligonucleotides containing the 3' nucleotide unique to each allele were synthesi~ed with different length 5' tails (see Table 1).
Upon ligation to the invariant 32p radiolabelled adjacent oligonucleotide, the individual products could be separated on a polyacrylamide denaturing gel and detected by autoradiography.
2 5 Based upon these initial findings with autoradiography, subsequent assays were preformed using an automated, non-radioactive detection scheme in which the allele specific oligonucleotides were 20~991 5'-biotinylated for capture, and the invariant oligonucleotides 3'-tailed with digoxygenin. The label was then visualized in an ELISA
format using anti-digoxigenin conjugated to alkaline phosphatase, and a colorimetric substrate for the enzyme.
S As depicted in Table 1, the nucleotide sequence and corresponding translated sequence of the oligonucleotides used in detecting 13A and 13S globin genes are depicted. Oligonucleotides 101 and 104 detect the 13A target, while 102 and 105 detect the 13S
target when ligated to labelled oligonucleotides 107 and 104, respectively. Oligonucleotides 103 and 106 were designed to assay the efficiency of ligation of G:T or G:A and C:A or C:T mismatches using BA or 13S globin gene targets respectively. Oligonucleotides were designed with slightly different length tails to facilitate discrimination of various products when separated on a denaturing polyacrylamide gel. The tails which were not complementary to the target sequence, may be considered as being "reporter groups" for the individual sequence. Consequently, ligation of oligonucleotides 101, 102, or 103 to 107 gives lengths of 45, 47, or 49 nucleotides, respectively. For the complementary strand, ligation of oligonucleotides 104, 105, or 106 to 109 gives lengths of 46, 48, or 50 nucleotides, respectively. The oligonucleotides were also designed to have calculated Tm values of 66 to 70C, which is just at or slightly above the ligation temperature.

2~79~

In order to detect the ligation produ~ts, oiigonucleotides 107 and iO9 were 5'-end labelled with 32p using T4 polynucleotide kinase and 32p according to the following example.

~P1~9'1 EXAMPLE Vlll (radioactive labelling) Oligonucleotide 107 (0.1 ~lg) was 5' end labelled in 20 ~l 3û
mM Tris HCI buffer at pH 8.0 containing 20 mM Tricine, 10 mM MgCI2, 5 0.5 mM EDTA, 5 mM dithiothreitol, and 400 ,uCi of [32P]ATP, by the addition of 15 units of T4 polynucleotide kinase. After incubation at 37C for 45 min, unlabelled ATP was added to 1 mM, and incubation was continued an additional 2 min at 37C. The reaction was terminated by the addition of 0.5 ~l 0.5 M EDTA, and kinase heat 1() inactivated at 65C for 10 min. Unincorporatsd 32p label was removed by chromatography with Sephadex G-25 pre-equilibrated with TE buffer. Specific activity ranged from 7 x 1û3 to 10 x 108 cpm/~lg of oligonucleotide.

2~79~1 The specificity of the J. aquaticus thermophllic li~ase according to the present invention for complementary vs.
misrnatched target was compared under both LDR and LCR conditions (see figure 3 and the following Table ll). In the LDR series, two S adjacent oli~onucleotides were incubated with denatured target DNA
and ligase, where the last nucleotide of the unlabelled oligonucleotide was either compiemented or mismatched the target DNA. The oligonucleotides were designed with slightly different length tails to facilitate discrimination of various products by 10 allowing them to be separated on a denaturing gel. Consequently, as disclosed earlier, ligation of oligonucleotide 101 (13A allele), l 02 (13S allele), or 103 to labelled 107 gives lengths of 45, 47 or 49 nucleotides, respectively. For the complementary strand, ligation of oligonucleotides 104 (l3A allele), 105 (13S allele), or 106 to labelled 15 109 gives lengths of 46, 48 or 50 nucleotides, respectively. The oligonucleotides were also designed to have a calculated Tm values of 66C to 70C, that is just at or slightly above the ligation temperature. Thus, the specificity of ligating two oligonucleotides hybridized to target DNA with perfect complementarity (A:T~ could 20 be directly compared to each possible mismatch (A:A, T:T, G:A, ~:T, C:A, or C:T). The methodology for determining specificity of ligation of these oligonucleotides in the presence of BA or BS globin gene target was determined as in the following example:

7s 2~7~

EXAMPLE IX
(determination of specificity of thermophilic ligase) Labelled oligonucleotide (200,000 cpm; 0.28 ng; 40 fmoles) and unlabelled oligonucleotide (.27 ng; 40 fmoles) were incubated in the 5 presence of target DNA (1 fmole = 6 x 10 ~ molecules Taq I digested 13A or BS globin plasmid) in 10 lli 20 mM Tris HCI buffer at pH 7.6 and containing 100 mM KCI, 10 mM MgC12, 1 mM EDTA, 10 mM NAD, 10 mM dithiothreitol, 4 ~g Salmon sperm DNA, and 15 nick-closing units of the thermophilic ligase, and overlaid with a drop of mineral 10 oil. The reactions were incubated at 94C for 1 min followed by 65C for 4 min, and this cycle was repeated between 5 and 30 times.
The reactions were terminated by the addition of 8 ~I formamide containing EDTA (10 mM), xylene cyanol ~0.2,~/o), and bromphenol blue (0.2%). Sarnples (4 ~11) were denatured by boiling for 3 min prior to 15 loading (40,000 cpm/lane~ into the gel.
Products were separated by electrophoresis in which samples were loaded in groups of eight, run into the gel, and then the next set loaded, thereby accounting for the slightly slower mobility of the bands on the right side of the autoradiogram of figure 3.
20 Electrophoresis was in a 10% polyacrylamide gel containing 7 M urea in a buffer of 100 mM Tris borate pH 8.9 and 1 mM EDTA, for 2 hrs at 60 W constant power.
After removing the urea by soaking for 10 min in 10% acetic acid followed by a second soak of 5 min in water, the gels were ~7~9~

dried onto Whatman 3 mm paper and autoradiographed overnight at -70~C on Kodak XAR-5 film (with or without Du Pont Cronex lightin3 plus intensifying screen). Bands from 20 cycles were excised from the gels and assayed for radioactivity. The results are given in 5 Table 11.

2 ~

TABLE ll Quantitation of complementary and mismatched LDR and LCR bands from 20 cycle LDR and 30 cycle LCR experiments described in 5 Example IX and depicted in Fig. 3 were excised from gels and assayed for radioactivity. Percentage product formed = cpm in product band/cpm in starting oligonucleotide band. Percentage mismatched/complementary _ cpm in band of mismatched oligonucleotides/cpm in band of complementary oligonucleotide 10 using the same target DNA, and gives an indication of the noise to signal ratio. LDR amplification was performed using 6 x to8 target molecules or 1 femtomole; LCR amplification was performed using 6 x 106 target molecules or 10 attomoles.
LDR Oligo base: Productmismatched/
tarç~et bas~ ~m~complem~ntary (%!
A: T 21.~
T:A 13.2 T :A 17.9 A : T 1 2.4 A: A <0.1 <Q.4 T:T 0.12 0.7 T: T 0.16 1.0 A: A <0.1 c0.4 G :T 0.30 1.4 C :T c0.1 c0.4 a :A <0.1 <0.4 ~: A <0.1 <0.4 A:T, T:A 41.4 T:A, A:T 10.4 A:A, T:T 0.45 1.1 2~79~

T:T, A:A ~0.05 <0.2 G:T, C:A û.51 1.3 G:A, C:T <0.05 <0.2 Thus, the thermophilic T. aquaticus ligase was shown to 5 discriminate complementary from mismatched oligonucleotide sequences for all possible mismatched base pairs in LDR assays.
Under both competition and individual ligation experiments (at varying salt concentrations), the worst case mismatch ligations were 1.5 to 1.0% (see Table ll, G:T and T:T), while others were 0.4%
10 to ~0.1% (see Table ll, A:A, C:T, G:A and C:A) of the products formed with complementary base pairs (A:T). This is substantially better than reported (usin~ radioactive detection) for the mesophilic T4 ligase of E. coli [see Gene 76:245 (1989)].
In the LCR amplification/detection series of experiments, two 15 adjacent oligonucleotides were incubated with denatured target DNA
and ligase, as well as with the complementary set of oligonucleotides. Under these conditionc;, the 3' nuoleotide of the unlabelled diagnostic oligonucleotide either complemented or mismatched the target DNA, but always complemented its unlabelled 20 counterpart, i.e. A:T for 101 and 104, T:A for 102 and 105, and G:C
for 103 and 106. Thus, an initial "incorrect" ligation of a mismatched oligonucleotide would subsequently be amplified with the same efficiency as a correct ligation. Samples contained pairs of unlabelled oligonucleotides (BA allele specific 101 and 104, BS

, . . ~

20~7~91 allele specific 102 and 105, or 103 and 106) with the complementary and adjacent pairs of labeJled oligonucleotides, 107 and 109. These labelled and unlabelled oligonucleotides were incubated in the presence of li~ase and 10 attonnoles of target DNA
S (100 fold less target DNA than for LDR) for 20 or 30 cycles as in Example IX,. The resulting bands are depicted in the left portion of figure 3 and the lower half of Table ll.
As can be seen in figure 3 and Table li, the thermophilic ligase according to the present invention was capable of discriminating complementary from mismatched oligonucleotide sequences for all possible mismatched base pairs in LCR assays. Under both competition and individual ligation experiments the worse case mismatch ligations were from 1.3% tc 0.6% (G:T, C:A and A:A, T:T), while others were <0.2% (T:T, A:A and G:A, C:T) of the products formed with complementary base pairs (A:T, T:A). LCR, using thermophilic ligase according to the present inven~ion, is thus the only method which can both amplify and detect single base mismatches with hi~h signal to noise ratios [see Genomics 4:560 (1989)]. Thus, by utilizing LCR one can detect the difference between a single base mismatch such as occurs between 13A and BS, and use the results of this assay as a diagnostic for the normal, the carrier, or the diseased patient.
When the entire set of experiments described above were repeated using buffer containing 150 mM instead of 100 rnM KCI, the ~7~

results were essentially the same as in figure 3 and tabulated in Table il, with ligation of mismatch oligonucleotides for LDR ranging from 0.6% to <0.3% and for LCR ranging from 1.7% to <0.3% of the exactly complementary products.. Thus, the exquisite discrimination S between matched and mismatched oligonucleotides appears not to be critically dependent upon salt conditions.
Alternatively, a different procedure based on phosphatase may also be used. The LCR or LDR reaction may be performed in a 10 ~11 volume under mineral oil. To this is added 50,u1 of 10 mM Tris HCI
10 pH 7.6 containing 0.5 units of Bacterial Alkaline Phosphatase (BAP), and 10 mM MgCI2, and the incubation continued at 65C for 2 hrs (note that the ligase enzyme is not kille:l under these conditions).
The ~' end label on an oligonucleotide that has become covalentl link~d is no lon~er susceptible to BAP. Ligated product is separated 15 from monophosphate by the addition of 20~11 of 10 mg/ml sonicated salmon sperm DNA as a carrier and precipitated with 20 ~ll of 50%
TCA. After centrifugation for 5 min at 12,000 rpm, the supernatant is removed, and the ration of pellet to pellet + supernatant gives the percenta~e of product formed. A similar assay has been used with 20 Taql endonuclease, and the experimental error for positive and negative controls is around 1-2%.
Use of the thermophilic ligase according to the present invention obviates the need to carefully titrate both salt and enzyme 2~7~91 concerltration as required for mesophilic ligases. The data from this series of experiments is tabulated in the following Table 111.

~799~

TABLE l l I

Quantitation of complementary and mismatched LDR and LCR bands, at 100 and 150 mM KCI concentrations, from 20 cycle LDR and 30 S cycle LCR experiments described in Example IX and depicted in Fig. 3.
LDR amplification was per~ormed using 6 x 108 target molecules or 1 femtomole; LCR amplification was performed using 6 x 1û6 tar~at molecules or 10 attomoles. The mismatched/complementary gives an indication of the noise to signal ratio.
LDR
Oligo base:Product mismatched/
tar~et baseformed (%) complementary (%) [KCI] (mM) [KCI] (mM) lS 100 150 100 150 A: T 21.5 23.2 T: A 13.2 17.2 T:A 17.9 12.8 A:T 12.4 11.7 A: A <0.1 <0.2 <0.4 ~0.3 T: T 0.12 0.21 0.7 0.3 T: T 0.16 0.30 1.0 0.6 A: A <0.1 <0.2 cO.4 <0.3 2 5 G: T 0.30 0.25 1.4 0.4 C :T <0.1 <0.2 <0.4 <0.3 G :A <0.1 0.25 <0.4 0.4 C: A <0.1 0.20 <0.4 0.3 LCR
A:T, T:A41.4 14.2 T:A, A:T1Q.4 18.5 A:A, T:T0.45 0.09 1.1 0.6 T:T, A:A<0.05 <0.05 c0.2 0.3 2~7g~1 G:T, C:A 0.51 0.24 1.3 1.7 G:A, C:T <0.05 ~0.1 <0.2 <0.7 LCR and LDR specificity was tested usin~ both BA and 13S
specific oligonucleotides in direct competition for ligation to the S invarient labelled oli~onucleotides. Using target DNA (B~, ~S, and an equimolar ratio of ûA and BS) ranging from 1 femtomole to 1 attomole, thermophilic ligase specifically formed the correct product(s) in each case; no background incorrect ligation product was observed when only one target allele was present). Howev~r, 10 the efficiency of forming the 13~ specific products is somewhat less than forming the BA products, and after 20 cycles of arnplification, the 13S specific products were approximately one-third of the 13A
specific prodwts as quantitated by assaying excised products for radioactivity. Hence a direct competition assay, wherein two 15 oligonucleotides are differentially labelled (for example with fluorescent groups) to quantitate the ralative initial concentrations of each target sequence allele will require careful titrations for each allele.
The specificity of LCR DNA amplification with sub-attomole 20 quantities of target DNA was also examined. The extent of LCR DNA
amplification was determined in the presence of target DNA ranging from 100 attomoles (6 x 107 molecules~ to less than one molecule per tube. Reactions were incubated for 20 or 30 cycles, and 9 ~ ~l products separated and quantitated as depicted in figure 4 and the following table IV.

~5 ~6~9~1 TABLE IV

Quantitation of LCR amplification. Bands from 30 cycle LCR
5 experiments were excised from the gels and assayed for radioactivity. At higher target concentration, DNA amplification was essentially complete after 20 cycles; slightly impreciss excision of 30 cycle bands from this portion of the gel probably accounts for product formed values in excess of 100%. Percentage 10 product formed = cpm in product band/cpm in startin~
oligonucleotide band; Amplification = No. of product molecules formed/No. of target molecules Target Product formed Amplification 1 5 ~L~ ~%~
6x 107 134 2 x 107 96 x ~o6 107 2 x 10~ 78 20 6 x 105 85 2 x 105 48 5.8 x 104 6 x 104 25 1.0 x 105 2 x 104 4.5 5.4 x 104 6 x 103 2.3 9.2 x 1û4 25 2 x 103 0.36 4.3 x 104 6 x 102 0.18 7.2 x 104 2 x 102 0.14 1.7 x 105 ~0 <0.05 <0.05 3 û 6 <0.05 2 <0.~5 0 <0.05 2~7~9 l In th~ absence of target, no background signal was detected when carrier salmon sperm DNA (4 ~ug) was present as seen in figure 4. At higher initial target concentrations, DNA amplification was essentially complete after 2û cycles, while at lower initial target S concentrations substantially more product is formed with additional arnplification cycles. Under these conditions, 200 molecules of initial target DNA could easily be detected after 30 cycles.
The thermostable nature of the enzyme is readily apparent in figure 4. By c:omparing the amount of product formed after 20 l O cycl0s to that formed after 30 cycles, it is apparent that at the lower target DNA concentrations additional product is formed after more cycles (see especially 2 x 104 to 2 x 10~ tar~et DNA
molecules). In other words, the enzyme still has activity after 20 cycles of 94C for 1 minute followed by 65C for 4 minutes.
Thus, T. aquaticus ligase retains the ability to catalyze formation of a phosphodiester bond between two adjacent oligonucleotides hybridized to a complementary strand of DNA at a temperature in the range of about 50C to about 85C after repeated exposure to temperatures that denature DNA, namely in the range of 20 about 105C for about 0.25 minutes to about 4 minutes.
Hence, the specific amplification of a nucleic acid test substance of known nucleotide sequence using LCP~ requires: (1) twc adjacent oligonucleotides complementary to and in molar excess of the target sequence nucleic acid, and having no mismatch to the target sequence nucleic acid at the junction of the adjacent oligonucleotides; (2) a second set of adjacent oligonucleotides cornplementary to the first set of adjacent oligonucleotides, complementary to and in molar excess of the target sequence nucleic 5 acid, and having no mismatch to the target sequence nucleic acid at the junction of this second set of adjacent oligonucleotides; (3) a therrnostable ligase which does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures of from about 5ûC to about 105C; and (4) subjecting this ligase 10 mixture to repeated temperature cycles which comprises a first temperature to denature the DNA (in a range of about 90C to about 1 05C), and a second temperature to allow for hybridization/ligation (in a range of about 50C to about 85C). In the amplification of BA
globin allele described above, the components were (1) 15 oligonucleotides 1 01 and 1 07; (2) oligonucleotides 1 04 and 1 09; (3) T. aquaticus ligase; and (4) 30 temperature cycles of 94C for 1 minute followed by 65C for 4 minutes.
in figure 4, bands of 45 and 46 nucleotides correspond to ligation products of the coding and complementary 13A globin 20 oligonucleotides~ Lower molecular weight products correspond to ligation of deletion oligonucleotides present in the initial ligation reaction. Since samples were loaded in groups of eight, the right siàe of the autoradiogram gives the appearance of slower rnigration.

J ~ ~ ~

To further test the ability of ligase to discriminate between complementary and mismatched oligonucleotides, an LCR experiment was performed in the presence and absence of oligonucleotides which would give G-T and C-A mismatches in accordance with the S following example which not only shows DNA amplification, but also supports the thermostable nature of the enzyme found in Example IX.

~79~1 EXAMPLE X
One set of experiments contained 40 fmoles each of unlabelled 101 and 104 oligonucleotides, while the second set had in addition 40 fmoles of unlabelled 103 and 106 oligonucleotides. Both sets 5 contained 40 frnoles each of labelled 107 and 109. Labelled oligonucleotides (200,000 cpm; .28 ng; 40 fmoles) and unlabelled oligonucleotides (.27 ng; 40 fmoles) were incubated in the presence of target DNA, ranging from 100 attomoles (6 x 107 molecules) to 0.01 attomoles (6 x 103 molecules) of Taq I digested 13A or 13S
10 globin plasmid. Incubation was carried out in 10 ,ul 20 mM Tris-HCI, ph 7.6 buffer containing 100 mM M~CI21 1 mM EDTA, 10 mM NAD, 10 mM dithiothreitol, 4 ',l9 Salmon sperm DNA, and 15 nick-closing units of T. aquaticus ligase, and overlaid with a drop oF mineral oil.
Reactions were incubated at 94C for 1 min followed by 65C for 4 15 min, and this cycle was repeated 20 or '30 times.
The resulting samples were electrophoresed, gel autoradiographed overnight with the aid of a Cronex intensifying screen and the bands counted. The bands from the autoradiographed gel are depicted in figure 4, and the quantitation of LCR
2 0 amplification tabulated in the following Table V.

9o ~7~9~

TABLE V

Quantitation of LCR amplification the presence or absence of mismatched competltor molecules.
S Complementary Complementary & Mismatched Oligonucleotides Oligonucleotides (101, 104) (101, 104 & 103, 106) (A:T, T:A) (A:T, T:A & G:T, C:A) Target ProductAmpli- Product Ampli- Mismatchad/
mQleGulesfQrme~fication formedfi~ation Complementary 6 x 107 (~A) 114 93 1.0 2 x 107 93 95 1.8 6 x 106 102 93 0.5 2 x 1o6 90 67 0.5 1 S 6 x 105 51 46 2 x 105 313.7 x 104 23 2.8 x 104 6 x 104 176.8 x 104 9.3 3.7 x 104 2 x 104 8.61.0 x 105 2.9 3.5 x 104 6 x 103 3.21.3 x 105 0.8 3.4 x 104 2 0 0 ~0.1 ~0.1 6x107(BS) 2.1 1.5 At high target concentrations, sufficient mismatched product was produced to be visualized (as in figure 4), the quantity of 25 mismatched product ranging from 1.8% to 0.5% of the complementary product. Use of an excess of mismatched target DNA (BS instead of 13A globin DNA at 6 x 107 molecules per tube) gave only 2.1% and 1.5% product The same amount of product may be formed when using three to ten thousand folcl less complementary target DNA.

2~79~

Based upon this, the signal from correctly paired ligation products is 50 to 500 fold higher than mismatched products under competition or individual LCR ligation eonditions.
At low target concentrations, the extent of DNA amplification ranged from 3.7 x 104 to 1.7 x 105 (see Tables IV and V). Assuming the efficiency of ligation is the same in each cycle, the average arnplification per cycle is between 40 and 50%.
The efficiency per cycle could, of course, be potentially enhanced by altering buffer conditions, enzyme concentration, or thermal cycling times and temperatures - all within the capabilities of those skilled in the art. It has, for example, been shouvn that the ligation efficiency of thermophilic ligase (and other ligases) may be enhanced by altering buffer compositions, such as using NH4CI, HEPES, polyamines such as spermidine, or polyethylene glycols [see 1~ J. Biol. Chem 259:10û41 (19843, and J. E3iochem. 100:123 (1986)].
Varying the amounts of each component in the currently used buffer and either supplementing or exchan~ing one or more components with, but not limited to, the chemical and biologicai components listed above, are amony the methods of improving LCR that are straight forward for those skilled in the art. One skilled in the art can also easily vary the cycling times and temperatures. For example, at later time points, the majority of target present is oligonucleotide product from a previous LCR reaction. These oligonucleotides are short (preferably but not limited to 40-60 2~79~

mers) and may melt more rapidly, allowing more rapid cycling. in the present invention, successful ligase chain reactions have been completed for 30 and 40 cycles under cycling conditions of 94C for 0.5 minutes followed by 65C for 2 minutes (half the time of the 1 5 minute at 94C and 4 minutes at 65C cycle time for the preferred ligase chain reaction conditions). Both the ligation temperature and the DNA denaturing temperatures may be varied with respect to actual degree, duration, and number of repeated cycles. Optimal conditions must maximize the amount of product formed in the 10 presence of perfectly complementary target DNA, while minimizing the amount of incorrect product formed in the presence of mismatched target DNA or in the absence of complementary target DNA.
Utilizing these findings, a method for the detection of 15 specific sequences of oligonucleotides in clinical samples was developed. The source of the sample may be any material or substance which comprises nucleic acid. The nucleic acid need not be a naturally occurring nucleic acid, but may be synthesiz~d by chemical, enzymatic, or biological means and may have other than 20 naturally occurring purines and pyrimidines. The source of the ciinical sample may be cel!ular or non-cellular, and may be derived from such physiological media as blood, serum, plasma, breast milk, stool, pus, tissue scrapings, washings, urine, or the like.
Furthermore, the sample may be associated with a set or subset of 2~799~

cells, such as neoplastic cells, Iymphocytes (for example, T-cells or B-cells, monocytes, neutrophils, etc3; may include pathogens including viruses, bacteria, mycoplasma, fungi, protozoa, etc.; may include constructs, etc. or RNA, such as messenger RNA, transfer 5 RNA, ribosomal RNA, viruses, or the like; and it may involve structural genes, untranslated regions, regulatory regions, introns, exons, or the like. In addition, the detection may be for a wide variety of purposes such as, for example, the diagnosis of a potential or actual disease state in plant or animal species, as weli 10 as the detection of sets or subsets of pathogens, the monitoring of genetic engineering, or the like.
One such method for which the present invention may be use~
(and whioh clearly demonstrates the feasibility of direct LCR allelic detection from blood samples without the need for prior PCR
15 amplification) is embodied, for example, in the detection of 13-globin alleles in human genomic DNA. Based upon the high level of DNA
amplification, the allele specific LCR detection of DNA was examined from blood collected from normal (13ABA), carrier ~I~A~S), and sickle cell (13S13S) individuals as more fully described in the 2 0 following example:

2~67991 EXAMPLE Xl (detection of ~-globin alleles in human genomic DNA) Human genomic DNA was isolated from 0.5 ml whole blood ~see PCR 7~chnology, H. A. Erlich editor, 5tockton Press (1989) pg 36].
S Whole blood (0.5 ml) was mixed with an equal volume of Iysis buffer (10 mM Tris-HCI, pH 7.6, containing 5 mM MgCI2 and 0.32 M sucrose).
After a brief centrifugation (1 min at 12,000 rpm in an eppendorf desktop eentrifuge~, the supernatant was very carefully rernoved, leaving 0.15 to 0.? ml of supernatant and loosely pelleted nuclei.
10 The pellet was resuspended with vortexing in an additional 0.5 ml Iysis buffer, nuclei pelloted and the supernatant removed as above.
This step was repeated three or four times until the supernatant was clear or just barely pink. After rernoval of the final sup~rnatant (again leaving about 0.15 to t).2 ml), 0.25 ml of LCR DNA
15 Buffer containing non-ionic detergents (20 rnM Tris-HCI, pH 7.6, containing 2 mM ~DTA and 0.45% each of non-ionic detergents NP40 and Tween 20) was added. Any excess RNA was digested by the addition of 2 ~11 of 4 mg/ml heat treated IRNase A for 15 min at 37C.
Any proteins were digested by the addition of 5 ~ul of 10 rng/ml 20 freshly made Proteinase K and incubation at 50C for 1 to 2 hours.
Proteinase K and RNase A were removed by sequential extractions with phenol, phenol/chloroform, chloroform, n-butanol (2X) and the nucleic acid recovered by precipitation with ethanol. Samples were boiled for 5 min prior to use in LCR assays.

9s 2~g~

Each isolated human genomic DNA was tested in two reaction mixtures, the first testing for the presence of the normal 13A ailele, and the second testing for $he presence of the sickle 13S allele. The first reaction mixture contained BA test oligonucleotides 101 and 104 (0.27 ng or 40 fmoles each), labelled oligonucleotides (107 and 109; 200,000 cpm (0.28 ng or 40 fmoles each), genomic DNA
(correspondin~ to 10 ,ul of blood, or about 6 x 104 nucleated cells) in 10 ~l 20 mM Tris-HCI buffer, pH 7.6, containing 100 mM KCI, 10 mM
MgC12, 1 mM EDTA, 10 mM NAD, 10 mM dithiothreitol, and 15 nick-closing units of T. aquaticvs ligase, and overlaid with a drop of mineral oil. The second reaction mixture containad 13S test oli~onucleotides 102 and 105 (0.27 ng or 40 fmoles each), labelled oligonucleotides 107 and 109 (200,000 cpm or 0.28 ng or 40 fmoles each), genomic DNA corresponding to 10 ~ll of blood or about 6 x 104 nucleated cells) in 10 ~LI 20 mM Tris-HCI buffer, pH 7.6 and containing 100 mM KCI, 10 mM MgCI2, 1 mM EDTA, 10 mM NAD, 10 mM
dithiothreitol, and 15 nick-closing units of T. aquaticus ligase, and overlaid with a drop of mineral oil.
Both reaction mixtures were incubated at 94C for 1 min followed by 65C for 4 min, and this cycle was repeated 20 tc 30 times. Reactions were terminated by the addition of 8 ~11 formamide containing EDTA (10 mM), xylene cyanol ~0.2%), and bromphenol blue (0.2%).

~79~

Samples (4 ~11) were denatured by boiiing for three min prior to loading (40,000 cpm/lane). Electrophoresis was in a 10%
polyacrylamide gel containing 7 M urea in a buffer of 100 rnM Tris borate at ph 8.9 and 1 mM EDTA, for 2 hours at 60 Watt constant 5 power. After removing the urea (10 min soak in 10% acetic acid, followed by 5 min soak in H2O). Gels were then dried onto Whatman 3 mm paper and autoradiographed overnight at -70C on Kodak XAR-5 film with a DuPon~ Cronex intensifying screen. Ligation products of 45 and 46, or 47 and 48 nucleotides indicate the presence of the BA
1`0 or BS globin gene, respectively. As noted with plasmid derivad target DNA, the efficiency of ligation (and hence detection) is somewhat less for the BS than the BA specific oligonucleotides.

~7~

Figure 5 is an autoradiogram showing th~ detection of 13-globin alleles in human genomic DNA made in accordance with the proceeding example. Ligation products of 45 and 46, or 47 and 48 nucleotides indicate the presence of the BA or BS globin ~ene, S respectively. Thus, with target DNA corresponding to 10 ,ul of blood, BA and BS alleles could be readily detected using allele specific LCR.
Hence, the successful detection of a biologically derived nucleic acid test substance, which has a known normal nucleotide 10 sequence and a known possible mutation at at least one target nucleotide position in the sequence, requires (1) a first reaction mixture comprising two sets of adjacent oligonucleotides complementary to each other, complementary to the target sequ~nce nucleic acid, wherein there is at least one mismatched base pair to 15 the mutant target sequence nucleic acid, but not to the normal target sequence nucleic acid at the junction of the adjacent oligonucleotides; (2) a second reaction mixture comprising two sets of adjacent oligonucleotides complementary to each other, complementary to the target sequence nucleic acid, wherein there is 20 at least one mismatched base pair to the normal target sequence DNA, but not to the mutant target sequence nucleic acid at the junction of the adjacent oligonucleotides; (3) a therrnostable ligase which does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures of from about 50C to about 2~799 ~

105C; and (4) subjecting these ligase mixtures to repeated ~emperature cyole which comprises a first temperature to denature the DNA (in a range of about 90C to about lQ5C), and a second temperature to allow for hybridization/ligation (in the range of S about 50C to about 85C) - this also allows adjacent oligonucleotides in each reaction mixture to become possibly covalently linked; (5) separating the test substance and any unlinked test oligonucleotides from covalently linked oligonucleotide product (if formed); and (6) detecting the presence or absence of covalently 1 û linked oligonucleotides in each reaction mixture whereby the presence of covalently linked oligonucleotide product in the first reaction mixtura indicates the presence of normal target sequence and the presence of covalently linked oligonucleotide product in the second reaction mixture indicates the presence of mutant target l S sequence. In the detection of BA and BS globin alleles described above, th~ components were (1) oligonucleotides 10~, 104, 107 and 109; (2) oligonucleotides 102, 105, 107 and 109; (3) T. aquaticus ligase; (4) 30 temperature cycles of 94(: for 1 min followed by 65C for 4 min; (5) denaturing nucleic acids by boiling in 45%
20 formamide and separating on a sequencing gel; and (6) autoradiographing of the gel.
This clearly demonstrates the feasibility of direct L.CR allelic detection from blood samples according to the present invention without the need for PCR amplification.

~7~9~

As noted with plasmid derived target DNA, the efficiency of ligation (and hence detection) is somewhat less for the BS than the ~A specific oligonucleotides. After 30 cycles of amplification, ~S
specific products were approximately one-third of 13A specific 5 products, as quantitated by assaying excised products for radioactivity. These differences may be a function of the exact nucleotide sequence at the ligation junction, or the particular oligonucleotides (with differing 5' tails) used in the LCR
experiments. However, the present invention still allows for a 10 direct competition assay where two oligonucleotides are differentially labelled (for example with fluorescent groups or, in ~his case, with different length tails) to determine the presence or absence of either allele in a reaction mixture. In the generalized form, the method according to the pres~nt invention allows one to 15 assay two alleles in the same vessel, pl~oviding the sets of oligonucleotides containing at least one mismatched base pair to the mutant target sequence nucleic acid, but not to the normal target sequence nucleic acid at the junction of the adjacent oligonucleotides, are labelled with one set of labels, and the 20 oligonucleotides containing at least one mismatched base pair to the normal target sequence nucleic acid, but not to the mutant target sequence nucleic acid at the junction of the adjacent oligonucleotides, are labelled with a different label.

799~

In a comparable non-radioactive assay, as depicted in figure 6, a minimum of two oligonucleotide probes are synthesized and modified for particular functions in ~he ligation assay. One probe contains a hook that permits the capture of the oligonucleotide S following ligation. An example of such a hook is biotin which can be captured by streptavidin or avidin bound to appropriate supports.
The other probe has a reporter group. Although a variety of reporter groups, both radioisotopic and non-radioactive, are available and can be used with the assay according to the present invention, such as 10 fluorophores or luminescent moieties, th~ currently preferred reporter is one which may participate in an ELISA (enzyme-linked immuno sorbent assay). More specifically, figure 6 depicts a schematic diagram of an ELISA based oligonucleotide ligation assay in which biotinylated (B) and digoxigenin-labelled (D) 15 oligonucleotides are hybridized with a DNA target in the presence of ligase (arrcw). Biotinylated oligonucleotides are captured on streptavidin ~SA) coated within the wells of microtiter plates. The wells are washed to remove unbound oligonucleotides, and alkaline phosphatase (AP) conjugat~d anti-digoxigenin antibodies ( D) are 20 added to the wells. Following an incubation and wash cycle, alkaline phosphatase substrate (S) is add~d, and digoxigenin detected by the production of a color product.
The non-radiolabelled assay according to the present invsntion consists of several steps: (1) preparation of the DNA target; (2) 2~799~

denaturation and hybridization of th~ modified oligonucleotide probes; ~3) ligation; ~4~ capture of the biotinylated probe; (5) washing to remove free nonbiotinylated oligonucleotides and target;
(6) addition of alkaline phosphatase conjugated anti-digoxigenin S antibodies; (7) washing to removed unbound antibody; (8) addition of alkaline phosphatase substrate; and (9) spectrophotometric analysis.
The following flow chart details the general procedure (which has automated on a modified Biomek 1000 workstation instrument) by which a non-radiolabelled assay according to the present invention 10 can be conducted:

~7~

Amplified Target DNA
-T4 Ligase Dete~;~;~ ~aq Ligase Detection S ~ ~
Denature residual Taq polymerase by adding:

45ll10fO.3NN~H jS,ulofO.1 NKOH

Renature target DNA by adding:

45 ~LI of 0.3 N HCI 45 ul of 0.1 N HCI

Distribute amplified ~tb m~er plates at ~iO ,ul per well Add biotinylated and reporter oligonucleotides to DNA targets (200 fmoles of each oligonucleotide in lO ,ul of 2 X ligation mix) 2 5 Ligation Mix: Ligation Mix:
200 fmole biotinylated oligo 200 fmole biotinylated olign 200 fmole reporter oligo 200 fmole reportsr oligo 100 mM Tris-HCI, pH 7.5 100 mM Tris-HCI, pH 7.5 20 mM MgCI2 20 mM MgC12 3 0 10 mM DTT 10 mM DTT
2mMATP 2mMATP
2 mM Spermidine 2 mM Spermidine 50% Formamide 2 mM NAD
100 mM KCI
3 5 \~ Taq Ligase Denature target oligonucieotide mix at 93C for 2 minutes .

2~7~

Cool to room temperature and add 5 ~,11 T4 ligase in Cool to 60-68C and ligate 200 mM NaCI for 15 minutes ~repeat 50 mM Tris-HCI, pH 7.5 denaturation and ligation 10 mM MgC~2 step to amplify) 5 mM DTT
1 mM ATP
1 mM Spermidine Ligate at room temperature (25C) for 15 mlnutes Stop ligation reaction and denature products by adding 10 ~LI of 0.3 N NaOH
Neutralize reactions by adding 4 ,LI of 3 M sodium acetate Transfer reactions to an avidin coated and blocked microtiter plate (avidin coating - 60 1l9 of avidin/well in 60 111 of PBS, pH 7.0 for 60 min at 37C;
blocking - remove avidin from the plate and add 200 ,ul/well of 100 mM Tris-HCI, 150 mM NaCI, 0.05/O Tween, 0.5% dry milk, and 100 ~,lg/ml of salmon sperm DNA) Capture biotinylated oligonucleotide at room temperature for 30 minutes J, Wash plate to remove unbound oligonucleotides and targets with ~1) 100 mM Tris-HCI, pH 7.~, in 150 mM NaCI in 0.05% Tween; (2) 0.01 N NaOH in 0.05% Tween; and (3) 100 mM TRIS-HCI, pH 7.5 in 150 mM NaCI In 0.05% Tween ~, Add alkaline phosphatase conjugated antibody to the reporter oligonucleotide; 30 111 per well in 100 mM TRIS-HCI, pH 7.5, 150 mM NaCI, 0.5% dry milk and 0.05% Tween Incubate plates for 30 min at room temperature for antibody binding to the reporter Wash the plate with 100 mM TRIS-HCI, pH 7.5, 150 mM NaCI in 0.05% Tween to remove 4 0 unbound antibody Add substrate Read plate for appropriate colormetric, chemiluminescent, or fluorescent product 2~7991 Genomic sequences required to begin this assay can be amplified by a number of different methods, including LCR, 3SR, and PCR. We have used PCR amplification to obtain DNA targets listed on the following Table Vl for litigation assay primers:
TABLE Vl (sequences of amplification primer sets) TarQ~t (~i~ Amplification Primers 13 - g l o b i n CMCTTCATCCACGTTCACNTGCC
1 0 AGGGCAGGAGCCA~GGCTGGGG
alpha1-antitrypsin TCAGCCTTACMCGTGTCTGTGCTT
GTATGGCCTCTAMMCATGGCCCC
cystic fibrosis CAGTGGMGAATGGCATTCTGTT
GGCATGCmGAT~3ACGCTTCTG

DNA amplification was performed using 5 ~l of DNA (2 ng/,ul for genomic DNA or 5 ~LI of treated material from an alternative source) is mixed with a pair of prirner oligonucleotides (0.5 IlM
eaoh) specific for the region of DNA to be amplified in a PCR buffer 20 containing 0.05 U/~ll of Taq polymerase, 50 mM KCI, ~5 mM Tris HCI
buffer at pH 8.3, 10 mM MgC12, 200 ,ug/ml gelatin, 0.1% Triton X-100, and 1.5 mM each of dATP, dCTP, dGTP and dTTP. The sample was overlaid with 60~1 of li~ht mineral oil, denatured at 93C for 5 min target, and subjected to 40 cycles consisting of 20 sec at 93C, 25 40 sec at 55C, and 1 min at 72C. Followin~ temperature cyciing, ~7~

the sample was subjected to 10 min at 72C to complete extension of the DNA sample.
Oli~onucleotides are synthesized and modified for particular functions in the iigation assay. The assay requires a minimum of 5 two modified oligonucleotides. One oligonucleotide has a hook that permits capture of the oligonucleotide following iigation. An example of this is a biotinylated oli~onucleotide which can be captured on streptavidin or avidin supports. The other oligonucleotide has a reporter group which, in the case of a 10 fluorophore reporter, multiple reporters with different emission spectra could easily be incorporated into a single assay.
For an ELISA based system, probes which discriminate allelic forms of a gene are synthesized with a 5' biotin group. Reporter probes are enzymatically or chemically 5'-phosphorylated and 15 labellsd with the hapten digoxigenin. The hapten is added to the 3' end of the reporter probe by tailing 500 pM of oligonucleotide at 37G for 1 hour in 10 mM potassium cacodylate, pH 7.û, 1 mM CoC12, 0.1 mM DTT, ~ nM of digoxigenin dUTP, 0.05 ,uM of dATP, and 100 units of the enzyme terminal transferase in a total volume of 20 ~I.
2û After labelling, 2 ,ul of 3 M sodium acetate and 1 ,ul of yeast t-RNA
(1 mg/ml) and 60 ,ul of 95% ethanol is added. The oligonucleotide is precipitated at 4C for 5 min and then collected by centrifugation at 6500 x g for 5 minutes. The pellet is resuspended in 20 1ll of distilled water and the process repeated. This precipitation removes unconjugated excess digoxigenin from the labeiled probe.
Example of oligonucleotides which discriminate alleles for three pathologic states are given in the following table V17:
TABLE Vll (sequences of example oligonucleotides for ELISA detection) Target Gene Form of Gene Biolinylated Labelled (L) Detected Primer Primer 1 0 13-globin I~A B1~rGGr~TccTGA
BS B2-ATGGTGCACCTG~CTCCTGT
alpha1 M B1~1 (~I~I~CGACG
anti-trypsin ~1 Z ~a~
cystic non-508 B1-ATTAMGAMATATCATCTT
fibrosis ~\ G~
508 B2-ACCAT^rAMGMM-rATCAT

Utilizing the procedure contained in the previous flow chart, a number of experiments were run and, after color development, data were obtained spectrometricaily at a wavelength of 490 mN.
Typical results for such tests have been tabulated in the following table Vill:

g 9 ~

TABLE Vll!

(spectrophotometric data from automated ligation reactions using Taq ligase) ~igation Primer Mix Amplified Genomic DNA Tar~çt Fr~m: _B1 + L B2 + L
~ - globin ~A 1.27 + 0.060.01 + 0.01 13S 0.04 + O.û3 1.85 + 0.03 alpha1 -antitrypsin M 1.8~ ~ 0.1~0.03 ~ 0.01 Z 0.03 + 0.031.47 ~ 0.07 cystic fibrosis:
non-508 1.33 + 0.200.0~ + 0.01 508 0.01 + 0.011.66 ~ 0.16 Comparable levels of detection were achieved with either T4 or Taq ligase. In addition, a number of ligation reactions have been performed for several other disease associat~d polymorphisms with comparable results. Additionally, eight different polymorphisrns in the human T csll receptor loci have been examined with similar deteçtion results. The present invention, therefore, appears to be generally applicable in the analysis of DNA polymorphisms consisting of single base substitutions, DNA deletion or insertions, or DNA translations.

9 ~ ~

In addition, a number of alkaline phosphatase substrates can be employed in the ELISA assay of the present invention including sensitive chemiluminescent substrates (10 attomole detection). The format of the assay is easiiy adapted to other reporter formats such as fluoropores which can be read in the appropriate microtiter format. Incorporation of the appropriate fluorophore format would, for example, permit multiplex analysis by ligation. In this scheme, oligonucleotides discriminating different alleles and/or different genes could be evaluated in a single assay. Furthermore, it is also possible that tandem ligation assays (ligation of oligonucleotides in chains) could be employed to assess closely spaced DNA
polymorphisms such as those which exist in the major histocompatibility complex ~enes. Such modifications to the assay specifically depicted above are considered to be well within the scope of the present invention The present invention can be used in a wide variety of DNA
diagnostic screening. For example) and not intending to limit the scope of the present invention, such DNA diagnostic screens may include those according to the following summary:
2 0 A - INFECTIOUS DISEASES:
1. Vira! Diseases: HIV, EBV, HPV, HSV, CMV, Hepatitis (non-A, non-B) (i3 blood and tissue screening ( i i ) rapid identification (iii) distinguish chronic infection from past exposure 10~

2~7~
(iv) distinguish resistant strains in mixed infection 2. Bacterial Diseases: Mycobacteria, Syphilis, Clamydia, Legionella, Campylobacter; Pneumonocystis, Lysteria Lyme, Leprosy (i) rapid identification of slow growing microbes ( i i ) identification in immuno-deficient patients ( i i i ) testing food for contamination 3. Parasitic Diseases: Malaria, Trypanosomes, Leishmania ( i ) rapid identification of "third world" blood 1 0 diseases ( i i ) screening travelers a

Claims (17)

1. The cell line AK75 designated as ATCC 55032.

2. A plasmid selected from pDZ1 and designated as ATCC
68307, and pDZ7 and designated as ATCC 68308.

3. A purified isolated DNA fragment or nucleic acid sequence that hybridizes to such sequence under high stringency consisting essentially of a DNA sequence coding for a thermostable Thermus aquaticus strain HB8 ligase enzyme.

4. A purified isolated DNA fragment or nucleic acid sequence that hybridizes to such fragment under high stringency which comprises the partial sequence coding for a thermostable ligase having the nucleic acid sequence .

5. An expression vector comprising a DNA sequence encoding for a thermostable ligase selected from the group consisting of (1) Thermus aquaticus HB8 ligase; (2) a thermostable ligase having at least 6 sequential amino acid residues corresponding to 6 sequential amino acid residues in Thermus aquaticus HB8 ligase; and (3) a ligase active mutant of Thermus aquaticus HB8 ligase or a fragment thereof wherein an amino acid residue has been inserted, substituted or deleted in or from the amino acid sequence of the ligase or its fragment.

6. A purified isolated polypepticle which catalyses the formation of a phosphodiester bond at the site of a single-stranded break in duplex DNA at temperatures of about 50°C to about 85°C, and which does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures of from 90°C to about 105°C.

7. A purified isolated polypeptide having the amino acid sequence:

.

8. A purified polypeptide isolated from a recombinant organism transformed with a vector that codes for the expression of Thermus aquaticus ligase which polypeptide catalyses the formation of a phosphodiester bond at the site of a single-stranded break in duplex DNA at a temperature of about 50°C to about 85°C, and which polypeptide does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures in the range of from about 90°C to about 105°C.

9. A purified isolated ligase which catalyses the formation of a phosphodiester bond between two adjacent oligonucleotides hybridized to a complementary strand of DNA at a temperature in the range of about 50°C to about 85°C.

10. A purified isolated ligase which catalyses ligation of two adjacent oligonucleotides hybridized to a complementary target sequence of DNA at a temperature in the range of about 50°C to about 85°C, wherein the products formed of the ligation are about 50 to about 500 more than if a single base mismatch is present at the junction of the adjacent oligonucleotides.

11. A purified isolated ligase which retains the ability to catalyze formation of a phosphodiester bond between two adjacent oligonucleotides hybridized to a complementary strand of DNA at a temperature of about 50°C to about 85°C after repeated prior exposure to temperatures of from about 90°C to about 105°C for about 0.25 min to about 4 minutes.

12. A purified isolated ligase which retains the ability to catalyze formation of a phosphodiester bond between two adjacent oligonucleotides hybridized to a complementary strand of DNA at a temperature range of about 50°C to about 85°C and wherein the products formed of the ligation are about 50 to about 500 more than if a single base mismatch is present at the junction of the adjacent oligonucleotides, after repeated prior exposures to a temperature range of from about 90°C to about 105°C for about 0.25 min to about 4 minutes.

13. A method for amplifying a nucleic acid test substance of known nucleotide sequence comprising:
(1 ) providing a reaction mixture comprising a first set of two adjacent oligonucleotides complementary to and in molar excess of a target sequence nucleic acid and further having no mismatch to the target sequence DNA at the junction of the adjacent oligonucleotides;
(2) providing a thermostable ligase which does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures of from about 50°C to about 105°C; and (3) subjecting the ligase-mixture to at least two temperature cycles which comprises a first temperature range of about 90°C to 105°C and a second temperature range of about 50°C to about 85°C.

14. A method for detecting a biologically derived nucleic acid test substance having a known normal nucleotide sequence and a known possible mutation at at least one target nucleotide position in the sequence which comprises:
(1 ) providing a reaction mixture comprising two adjacent oligonucleotides complementary to the target sequence nucleic acid and wherein the oligonucleotides have at least one mismatched base pair to the mutant target sequence nucleic acid but not to the normal nucleotide sequence nucleic acid at the junction of the adjacent nucleotides;
(2) providing a second reaction mixture comprising two adjacent oligonucleotides complementary to the target sequence nucleic acid and having at least one mismatched base pair to the normal target sequence nucleic acid, but no mismatch to the mutant target sequence nucleic acid at the junction of the adjacent oligonucleotides;
(3) providing a thermostable ligase which does not become irreversibly denatured and lose catalytic activity when subjected to temperatures of from 50°C to about 105°C to each of the first and second reaction mixtures;
(4) subjecting each of the ligase-mixture to at least one temperature cycle which comprises a first temperature of from about 90°C to about 105°C and a secondtemperature of about 50°C to about 85°C;
(5) allowing the adjacent oligonucleotides in each reaction mixture to become possibly covalently linked;
(6) separating the test substance and unlinked oligonucleotides in each reaction mixture from possible covalently linked oligonucleotides; and (7) detecting the presence or absence of covalently linked oligonucleotide product in each reaction mixture wherein the presence of covalently linked oligonucleotides in the first reaction mixture indicates the presence of normal sequence and the presence of covalently linked oligonucleotides in the second mixture indicates the presence of mutant sequences.

15. A kit for assaying a biologically derived DNA or RNA test substance which has a known normal nucleotide sequence and a known possible mutation at at least one target nucleotide position in the sequence, the kit comprising:
(1) a container holding a first reaction mixture comprising two adjacent oligonucleotides complementary to the target sequence nucleic acid, wherein one oligonucleotide is labelled and there is at least one mismatched base pair to the mutant target sequence nucleic acid, but not to the normal target sequence nucleic acid at the junction of the adjacent oligonucleotides, (2) a container holding a second reaction mixture comprising two adjacent oligonucleotides complementary to the target sequence nucleic acid, wherein one oligonucleotide is labelled, wherein there is at least one mismatched base pair to the normal target sequence nucleic acid, but not to the mutant target sequence nucleic acid at the junction of the adjacent oligonucleotides; and (3) a thermostable ligase which does not become irreversibly denatured and lose its catalytic ability when subjected to temperatures of from 50°C to about 105°C.

16. A method according to Claim 13 which further comprises providing a second set of two adjacent oligonucleotides wherein the adjacent oligonucleotides of the second set are complementary to the first set, complementary to and in molar excess of the target sequence nucleic acid, and has no mismatch to the target sequence nucleic acid at the junction of the second set of adjacent oligonucleotides.

17. A method according to Claim 14 which further comprises:
(1) providing a second set of two adjacent oligonucleotides to the first reaction mixture which set is complementary to the first set of adjacent oligonucleotides in the mixture and complementary to the target sequence nucleic acid, and wherein there is at least one mismatched base pair to the mutant target sequence nucleic acid but not to the normal target sequence nucleic acid at the junction of the second set of adjacent oligonucleotides; and (2) providing a second set of two adjacent oligonucleotides to the second reaction mixture which set is complementary to the first set of adjacent oligonucleotides in the mixture and complementary to the target sequence nucleic acid, and wherein there is at least one mismatched base pair to the normal target sequence nucleic acid but not to the mutant target sequence nucleic acid at the junction of the second set of adjacent oligonucleotides.