AU2007202518A1 - Methods and compositions utilizing nucleic acids - Google Patents

Description

P/00/011 Regulation 3.2

AUSTRALIA

Patents Act 1990 COMPLETE SPECIFICATION FOR A DIVISIONAL PATENT

ORIGINAL

TO BE COMPLETED BY APPLICANT Name of Applicant: Actual Inventor: Address for Service: Invention Title: INVITROGEN CORPORATION James L. HARTLEY; Michael A. BRASCH, Gary F. TEMPLE; Donna K. FOX CALLINAN LAWRIE, 711 High Street, Kew, Victoria 3101, Australia METHODS AND COMPOSITIONS UTILIZING NUCLEIC

ACIDS

The following statement is a full description of this invention, including the best method of performing it known to us:- 30/05/07.at16620.front pg,1 la- SMETHODS AND COMPOSITIONS UTILIZING NUCLEIC ACIDS Jt BACKGROUND OF THE INVENTION Field of the Invention 00 The present invention relates to recombinant DNA technology and to t methods and compositions for producing ligated nucleic acid molecules.

r" Related Art Site-specific recombinases. Site-specific recombinases are proteins that are present in many organisms viruses and bacteria) and have been characterised to have both endonuclease and ligase properties. These recombinases (along with associated proteins in some cases) recognise specific sequences of bases in DNA and exchange the DNA segments flanking those segments. The recombinases and associated proteins are collectively referred to as "recombination proteins" (see, e.g., Landy, Current Opinion in Biotechnology 3:699-707 (1993)).

Numerous recombination systems from various organisms have been described. See, Hoess et al, Nucleic Acids Research 14(6):2287 (1986); Abremski et al., J. Biol. Chem. 261(1):391 (1986); Campbell, J. Bacteriol.

174(23):7495 (1992); Qian et al., J. Biol. Chem. 267(11):7794 (1992); Araki et al., J.

Mol. Biol. 225(1):25 (1992); Maeser and Kahnmann, Mol. Gen. Genet. 230:170-176 (1991); Esposito et al., Nucl. Acids Res. 25(18):3605 (1997).

Many of these belong to the integrase family of recombinases (Argos et al., EMBO J. 5:433-440 (1986). Perhaps the best studied of these are the 30/05/07.at 6620.specipgs.2 U Integrase/att system from bacteriophage X (Landy, A. Current Opinions in ct Genetics andDevel. 3:699-707 (1993)), the Cre/loxP system from bacteriophage P 1 (Hoess and Abremski (1990) In Nucleic Acids and Molecular Biology, vol. 4.

mC Eds.: Eckstein and Lilley, Berlin-Heidelberg: Springer-Verlag; pp. 90-109), and the FLP/FRT system from the Saccharomyces cerevisiae 2 t circle plasmid oO S(Broach et al. Cell 29:227-234 (1982)).

C While these recombination systems have been characterized for particular organisms, the related art has only taught using recombinant DNA flanked by Srecombination sites, for in vivo recombination.

I Backman Patent No. 4,673,640) discloses the in vivo use of A recombinase to recombine a protein producing DNA segment by enzymatic sitespecific recombination using wild-type recombination sites attB and attP.

Hasan and Szybalski (Gene 56:145-151 (1987)) discloses the use of Int recombinase in vivo for intramolecular recombination between wild type attP and i attB sites which flank a promoter. Because the orientations of these sites are inverted relative to each other, this causes an irreversible flipping of the promoter region relative to the gene of interest.

Palazzolo el al. Gene 88:25-36 (1990), discloses phage lambda vectors having bacteriophage X arms that contain restriction sites positioned outside a cloned DNA sequence and between wild-type loxP sites. Infection ofE. coli cells that express the Cre recombinase with these phage vectors results in recombination between the loxP sites and the in vivo excision of the plasmid replicon, including the cloned cDNA.

P6sfai et al. (Nucl. Acids Res. 22:2392-2398 (1994)) discloses a method for inserting into genomic DNA partial expression vectors having a selectable marker, flanked by two wild-type FRT recognition sequences. FLP site-specific recombinase as present in the cells is used to integrate the vectors into the genome at predetermined sites. Under conditions where the replicon is functional, this cloned genomic DNA can be amplified.

Bebee et al. Patent No. 5,434,066) discloses the use of site-specific recombinases such as Cre for DNA containing two loxP sites is used for in vivo recombination between the sites.

N Boyd (Nucl. Acids Res. 21:817-821 (1993)) discloses a method to facilitate the cloning of blunt-ended DNA using conditions that encourage Sintermolecular ligation to a dephosphorylated vector that contains a wild-type loxP Cc site acted upon by a Cre site-specific recombinase present in E. coli host cells.

Waterhouse et al. (PCT No. 93/19172 and Nucleic Acids Res. 21 0_ (9):2265 (1993)) disclose an in vivo method where light and heavy chains of a particular antibody were cloned in different phage vectors between loxP and loxP 511 sites and used to transfect new E. coli cells. Cre, acting in the host cells Son the two parental molecules (one plasmid, one phage), produced four products N 10 in equilibrium: two different cointegrates (produced by recombination at either loxP or loxP 511 sites), and two daughter molecules, one of which was the desired product.

In contrast to the other related art, Schlake Bode (Biochemistry 33:12746-12751 (1994)) discloses an in vivo method to exchange expression cassettes at defined chromosomal locations, each flanked by a wild type and a spacer-mutated FRT recombination site. A double-reciprocal crossover was mediated in cultured mammalian cells by using this FLP/FRT system for sitespecific recombination.

Transposases. The family of enzymes, the transposases, has also been used to transfer genetic information between replicons. Transposons are structurally variable, being described as simple or compound, but typically encode the recombinase gene flanked by DNA sequences organized in inverted orientations. Integration of transposons can be random or highly specific.

Representatives such as Tn7, which are highly site-specific, have been applied to the in vivo movement of DNA segments between replicons (Lucklow et al., J. Virol. 67:4566-4579 (1993)).

Devine and Boeke Nucl. Acids Res. 22:3765-3772 (1994), discloses the construction of artificial transposons for the insertion of DNA segments, in vitro, into recipient DNA molecules. The system makes use of the integrase of yeast TY 1 virus-like particles. The DNA segment of interest is cloned, using standard methods, between the ends of the transposon-like element TY1. In the presence

C

I of the TYI integrase, the resulting element integrates randomly into a second c target DNA molecule.

DNA cloning. The cloning of DNA segments currently occurs as a daily mC routine in many research labs and as a prerequisite step in many genetic analyses.

The purpose of these clonings is various, however, two general purposes can be oO Sconsidered: the initial cloning of DNA from large DNA or RNA segments C (chromosomes, YACs, PCR fragments, mRNA, etc.), done in a relative handful CN of known vectors such as pUC, pGem, pBlueScript, and the subcloning of Sthese DNA segments into specialized vectors for functional analysis. A great deal

C

I of time and effort is expended both in the transfer of DNA segments from the initial cloning vectors to the more specialized vectors. This transfer is called subcloning.

The basic methods for cloning have been known for many years and have changed little during that time. A typical cloning protocol is as follows: digest the DNA of interest with one or two restriction enzymes; gel purify the DNA segment of interest when known; prepare the vector by cutting with appropriate restriction enzymes, treating with alkaline phosphatase, gel purify etc., as appropriate; ligate the DNA segment to the vector, with appropriate controls to eliminate background of uncut and self-ligated vector; introduce the resulting vector into an E. coli host cell; (6).pick selected colonies and grow small cultures overnight; make DNA minipreps; and analyze the isolated plasmid on agarose gels (often after diagnostic restriction enzyme digestions) or by PCR.

The specialized vectors used for subcloning DNA segments are functionally diverse. These include but are not limited to: vectors for expressing genes in various organisms; for regulating gene expression; for providing tags to aid in protein purification or to allow tracking of proteins in cells; for modifying the cloned DNA segment generating deletions); for the synthesis of probes riboprobes); for the preparation of templates for DNA sequencing; for the Ci identification of protein coding regions; for the fusion of various protein-coding c regions; to provide large amounts of the DNA of interest, etc. It is common that Sa particular investigation will involve subcloning the DNA segment of interest into c several different specialized vectors.

As known in the art, simple subclonings can be done in one day the 00 00 DNA segment is not large and the restriction sites are compatible with those of the subcloning vector). However, many other subclonings can take several weeks, especially those involving unknown sequences, long fragments, toxic genes, unsuitable placement of restriction sites, high backgrounds, impure enzymes, etc.

C1 Subcloning DNA fragments is thus often viewed as a chore to be done as few times as possible. Several methods for facilitating the cloning of DNA segments have been described, as in the following references.

Ferguson, et al. Gene 16:191 (1981), discloses a family of vectors for subcloning fragments of yeast DNA. The vectors encode kanamycin resistance.

Clones of longer yeast DNA segments can be partially digested and ligated into the subcloning vectors. If the original cloning vector conveys resistance to ampicillin, no purification is necessary prior to transformation, since the selection will be for kanamycin.

Hashimoto-Gotoh, et al. Gene 41:125 (1986), discloses a subcloning vector with unique cloning sites within a streptomycin sensitivity gene; in a streptomycin-resistant host, only plasmids with inserts or deletions in the dominant sensitivity gene will survive streptomycin selection.

Accordingly, traditional subcloning methods, using restriction enzymes and ligase, are time consuming and relatively unreliable. Considerable labor is expended, and if two or more days later the desired subclone can not be found among the candidate plasmids, the entire process must then be repeated with alternative conditions attempted. Although site specific recombinases have been used to recombine DNA in vivo, the successful use of such enzymes in vitro was expected to suffer from several problems. For example, the site specificities and efficiencies were expected to differ in vitro; topologically-linked products were expected; and the topology of the DNA substrates and recombination proteins was expected to differ significantly in vitro (see, Adams et al, J. Mol.

Biol. 226.61-73 (1992)). Reactions that could go on for many hours in vivo were expected to occur in significantly less time in vitro before the enzymes became inactive. Multiple DNA recombination products were expected in the biological host used, resulting in unsatisfactory reliability, specificity or efficiency of subcloning.

Thus, in vitro recombination reactions were not expected to be sufficiently efficient 00 to yield the desired levels of product.

t Accordingly, there is a long felt need to provide alternative methods which confer advantages over the known use of restriction enzymes and ligases.

SUMMARY OF THE INVENTION The present invention is a divisional application of Australian Patent Application No. 2002325588 which is itself a divisional of Application No. 752704 (11995/99), (the "parent" applications), the specifications of which are herein incorporated by reference.

The invention provides a method for producing a ligated nucleic acid molecule which comprises ligating a first nucleic acid molecule comprising a recombination site to a second nucleic acid molecule, the method comprising contacting the first nucleic acid molecule and the second nucleic acid molecules in the presence of a topoisomerase under conditions which allow for ligation of the first nucleic acid molecule to the second nucleic acid molecule.

The invention also provides a method for preparing a ligated nucleic acid molecule which contains at least one recombination site, the method comprising: (a) generating a first nucleic acid molecule by polymerase chain reaction, contacting the first nucleic acid molecule with a topoisomerase and a second nucleic acid molecule which contains a recombination site and under conditions which allow for ligation of the first nucleic acid molecule to the second nucleic acid molecule to form said ligated nucleic acid molecule.

The invention further provides a method for preparing a ligated nucleic acid molecule which contains at least one recombination site, the method comprising: mixing a first nucleic acid molecule with one or more adapters 31/05107.atl 16620.specipgs.6 Q comprising one or more recombination sites in the presence of a topoisomerase, wherein the adapter is a second nucleic acid molecule; and incubating the mixture under conditions sufficient to add one or more of the adapters to one or more termini of the first nucleic acid molecule to produce the ligated nucleic acid molecule.

In addition the invention provides a composition comprising: an isolated 00 nucleic acid molecule comprising one or more recombination sites; and one or In more topoisomerases associated with said isolated nucleic acid molecule, wherein 0 one said one or more recombination sites are selected from the group consisting of Slox sites, attL sites, attR sites, attB 1 (SEQ ID NO:6), attB2 (SEQ ID NO:7), attB3 (SEQ ID NO:8), attPl (SEQ ID NO:15) and attP2 (SEQ ID NO:16).

Further the invention provides a composition comprising: an isolated nucleic acid molecule comprising two or more recombination sites; and one or more topoisomerases associated with said isolated nucleic acid molecule, wherein said two or more recombination sites do not recombine with each other.

The invention provides a composition, outside of a host cell, comprising a nucleic acid molecule which is a product of a polymerase chain reaction, and an isolated nucleic acid molecule having one or more recombination sites, said composition further comprising one or more topoisomerases.

The invention also provides an isolated nucleic acid molecule comprising: one or more recombination sites; and one or more topoisomerase recognition sites, wherein one said one or more recombination sites are selected from the group consisting of lox sites, attL sites, attR sites, attB 1 (SEQ ID NO:6), attB2 (SEQ ID NO:7), attB3 (SEQ ID NO:8), attPI (SEQ ID NO:15) and attP2 (SEQ ID NO:16).

The invention also provides an isolated nucleic acid molecule comprising: two or more recombination sites that do not recombine with each other; and (b) one or more topoisomerase recognition sites.

Preferably the recombination site(s) are selected from the group consisting of lox sites, attL sites, attR sites, attB 1 (SEQ ID NO:6), attB2 (SEQ ID NO:7), attB3 (SEQ ID NO:8), attP1 (SEQ ID NO: 15) and attP2 (SEQ ID NO: 16).

The compositions may comprise one or more Insert Donor molecules (as defined in the parent specifications), one or more Vector Donor molecules (as defined in the parent specifications) and one or more recombination proteins (or 305/07.ai 16620.specipgs,7 combinations thereof). The compositions may also comprise one or more cointegrate molecules, one or more Product molecules and one or more Byproduct molecule (or combinations thereof) (as defined in the parent specifications).

SCompositions related to preparing Insert Donor molecules may vary depending on the particular method utilized in preparing the desired Insert Donor 00 molecules. Compositions for preparing such molecules by amplification may In comprise one or more polypeptides having polymerase activity, one or more primers (Ni comprising one or more recombination sites, one or more nucleotides and one or (Ni more nucleic acid molecule to be amplified (or combinations thereof). Specifically compositions intended for inserting or adding recombination sites in a desired nucleic acid molecule may comprise one or more nucleic acid molecules or adapters comprising one or more recombination sites, one or more ligases, one or more restriction endonucleases, one or more topoisomerases, and one or more nucleic acid molecules desired to contain such recombination sites (or combinations thereof).

Compositions related to integration of recombination sites in a desired nucleic acid molecule may comprise one or more integration sequences comprising one or more recombination sites and one or more nucleic acid molecules desired to contain the recombination sites.

In a particularly preferred aspect, libraries populations of genomic DNA or cDNA, or populations of nucleic acid molecules, produced by de novo synthesis such as random sequences or degenerate oligonucleotides) are utilized in accordance with the present invention. By inserting or adding recombination sites to such populations of nucleic acid molecules, a population of Insert Donor molecules are produced. By the recombination methods of the invention, the library may be easily moved into different vectors (or combinations or vectors) and thus into different host systems (prokaryotic and eukaryotic) to evaluate and analyse the library or a particular sequences or clones derived from the library. Alternatively, the vectors containing the desired molecule may be used in vitro systems such as in vitro expression systems for production of RNA and/or protein.

Insert Donor molecules may comprise at least two recombination sites and where the Insert Donor molecules are linear, such two or more recombination sites are preferably located at or near both termini of the molecules. In accordance with the 30/05/07.at I 6620.specipgs.8 -9invention, the use of additional recombination sites more than two) may be used to facilitate subcloning of different inserts within the Insert Donor molecule, depending on the type and placement of such recombination sites.

Other preferred embodiments of the present invention will be apparent to one of ordinary skill in light of what is known in the art, in light of the following drawings and 00 description of the invention, and in light of the claims.

BRIEF DESCRIPTION OF THE DRAWINGS c Figure I depicts one general method of the parent invention, wherein the starting (parent) DNA molecules can be circular or linear. The goal is to exchange the new subcloning vector D for the original cloning vector B. It is desirable in one embodiment

C

(I to select for AD and against all the other molecules, including the Cointegrate. The square and circle are sites of recombination: loxP sites, att sites, etc. For example, segment D can contain expression signals, new drug markers, new origins of replication, or specialized functions for mapping or sequencing DNA.

Figure 2A depicts an in vitro method of recombining an Insert Donor plasmid (here, pEZC705) with a Vector Donor plasmid (here, pEZC726), and obtaining Product DNA and Byproduct daughter molecules. The two recombination sites are attP and loxP on the Vector Donor. On one segment defined by these sites is a kanamycin resistance gene whose promoter has been replaced by the tetOP operator/promoter from transposon Tn 10. See, Sizemore et al., Nucl. Acids Res. 18(10):2875 (1990). In the absence of tet repressor protein, E. coli RNA polymerase transcribes the kanamycin resistance gene from the tetOP. If tet repressor is present, it binds to tetOP and blocks transcription of the kanamycin resistance gene. The other segment of pEZC726 has the tet repressor gene expressed by a constitutive promoter. Thus cells transformed by pEZC726 are resistant to chloramphenicol, because of the chloramphenicol acetyl transferase gene on the same segment as tetR, but are sensitive to kanamycin. The recombinase-mediated reactions result in separation of the tetR gene from the regulated kanamycin resistance gene. This separation results in kanamycin resistance only in cells receiving the desired recombination product. The first recombination reaction is driven by the addition of the recombinase called Integrase. The second recombination reaction is driven by adding the recombinase Cre to the Cointegrate (here, pEZC7 Cointegr).

Figure 2B depicts a restriction map of pEZC705.

Figure 2C depicts a restriction map of pEZC726.

Figure 2D depicts a restriction map of pEZC7 Coint.

30/05/07.at l6620.specipgs.9 Figure 2E depicts a restriction map of Intprod.

Figure 2F depicts a restriction map of Intbypro.

SFigure 3A depicts an in vitro method of recombining an Insert Donor r c plasmid (here, pEZC602) with a Vector Donor plasmid (here, pEZC629), and i obtaining Product (here, EZC6prod) and Byproduct (here, EZC6Bypr) daughter 00 Smolecules. The two recombination sites are loxP and loxP 511. One segment of pEZC629 defined by these sites is a kanamycin resistance gene whose promoter has been replaced by the tetOP operator/promoter from transposon Tn 10. In the absence of tet repressor protein, E. coli RNA polymerase transcribes the kanamycin resistance gene from the tetOP. Iftet repressor is present, it binds to letOP and blocks transcription of the kanamycin resistance gene. The other segment of pEZC629 has the let repressor gene expressed by a constitutive promoter. Thus cells transformed by pEZC629 are resistant to chloramphenicol, because of the chloramphenicol acetyl transferase gene on the same segment as tetR, but are sensitive to kanamycin. The reactions result in separation of the tetR gene from the regulated kanamycin resistance gene. This separation results in kanamycin resistance in cells receiving the desired recombination product. The first and the second recombination events are driven by the addition of the same recombinase, Cre.

Figure 3B depicts a restriction map of EZC6Bypr.

Figure 3C depicts a restriction map of EZC6prod.

Figure 3D depicts a restriction map of pEZC602.

Figure 3E-depicts a restriction map of pEZC629.

Figure 3F depicts a restriction map of EZC6coint.

Figure 4A depicts an application of the in vitro method ofrecombinational cloning to subclone the chloramphenicol acetyl transferase gene into a vector for expression in eukaryotic cells. The Insert Donor plasmid, pEZC843, is comprised of the chloramphenicol acetyl transferase gene ofE. coli, cloned between loxP and atuB sites such that the loxP site is positioned at the 5'-end of the gene. The Vector Donor plasmid, pEZC1003, contains the cytomegalovirus eukaryotic promoter apposed to a loxP site. The supercoiled plasmids were combined with 1 t lambda Integrase and Cre recombinase in vitro. After incubation, competent SE. coli cells were transformed with the recombinational reaction solution.

Aliquots of transformations were spread on agar plates containing kanamycin to c select for the Product molecule (here CMVProd).

Figure 4B depicts a restriction map of pEZC843.

00 Figure 4C depicts a restriction map ofpEZC1003.

Figure 4D depicts a restriction map of CMVBpro.

N Figure 4E depicts a restriction map of CMVrod.

Figure 4E depicts a restriction map of CMVcoint.

N Figure 5A depicts a vector diagram of pEZC3 t.

Figure 5B depicts a vector diagram of pEZC1305.

Figure 5C depicts a vector diagram of pEZC 1309.

Figure 5D depicts a vector diagram ofpEZC 1313.

Figure 5E depicts a vector diagram of pEZC1317.

Figure 5F depicts a vector diagram of pEZC1321.

Figure 5G depicts a vector diagram ofpEZC1405.

Figure 5H depicts a vector diagram of pEZC1502.

Figure 6A depicts a vector diagram of pEZC1603.

Figure 6B depicts a vector diagram of pEZC 1706.

Figure 7A depicts a vector diagram of pEZC29 1.

Figure 7B depicts a vector diagram ofpEZC2931 Figure 7C depicts a vector diagram ofpEZC3101.

Figure 7C depicts a vector diagram of pEZC1802.

Figure 8A depicts a vector diagram of pGEX-2TK.

Figure 8B depicts a vector diagram ofpEZC3501.

Figure 8C depicts a vector diagram of pEZC3601.

Figure 8D depicts a vector diagram ofpEZC3609.

Figure 8E depicts a vector diagram ofpEZC3607.

Figure 8E depicts a vector diagram of pEZC3616.

Figure 8G depicts a vector diagram of pEZC3663.

Figure 8H depicts a vector diagram of pEZC3621.

12- SFIG. 81 depicts a vector diagram of GST-CAT.

FIG. 8J depicts a vector diagram of GST-phoA.

FIG. 8K depicts a vector diagram of pEZC3201.

FIG. 9A depicts a diagram of 5.2 kb PCR prod.

FIG. 9B depicts a vector diagram of pEZC1202.

00 FIG. 9C depicts a vector diagram of 5.2 kb clone.

ttn FIG. 10A depicts a vector diagram of pEZC5601.

O FIG. 10B depicts a vector diagram of pEZC6701.

FIG. 10C depicts a vector diagram of attL product.

FIG. 10D depicts attR product.

FIG. A depicts a vector diagram of pEZC7102.

FIG. 11A depicts a vector diagram of pEZC7510.

FIG. 11B depicts a vector diagram of pEZC7510.

FIG. 11C depicts the attL product.

FIG. 12A depicts an amp PCR product with terminal attB sites.

FIG. 12B depicts a tet PCR product with terminal attB sites.

FIG. 12C depicts a restriction map of amp7102.

FIG. 12D depicts a restriction map of tet 7102.

DETAILED DESCRIPTION OF THE INVENTION In the present inventions it was unexpectedly discovered that reversible and/or repeatable cloning and subcloning reactions can be used to manipulate nucleic acids to form chimeric nucleic acids using recombination proteins and recombination sites. Recombinational cloning according to the parent invention thus uses recombination proteins with recombinant nucleic acid molecules having at least one selected recombination site for moving or exchanging segments of nucleic acid molecules, in vitro and in vivo.

These methods use recombination reactions to generate chimeric DNA or RNA molecules that have the desired characteristic(s) and/or nucleic acid segment(s). The methods of the invention provide a means in which nucleic acid molecule of interest may be moved or transferred into any number of vector systems.

In accordance with the parent inventions, such transfer to various vector systems transfer to various vector systems may be accomplished separately, sequentially or in mass into any number of different vectors in one step). The improved 31 05/07.a 16620.specipgs. 12 13 specificity, speed and/or yields of the parent inventions facilitates DNA or RNA cloning, subcloning, regulation or exchange useful for any related purpose. Such purposes include in vitro recombination of DNA or RNA segments and in vitro or in vivo insertion or modification of transcribed, replicated, isolated or genomic DNA or

RNA.

00 In Definitions: are the same as used in the parent specifications, herein incorporated by reference. For clarity the meaning of a number of the terms are repeated here.

In the description that follows, a number of terms used in recombinant DNA 10 technology are utilized extensively. In order to provide a clear and consistent (,i understanding of the specification and claims, including the scope to be given such terms, the following definitions are provided.

Byproduct: is a daughter molecule (a new clone produced after the second recombination event during the recombinational cloning process) lacking the segment which is desired to be cloned or subcloned.

Cointegrate: is at least one recombination intermediate nucleic acid molecule of the present invention that contains both parental (starting) molecules. It will usually be circular. In some embodiments it can be linear.

Insert or Inserts: include the desired nucleic acid segment or a population of nucleic acid segments (segment A of FIG. 1) which may be manipulated by the methods of the parent invention. Thus, the terms Insert(s) are meant to include a particular nucleic acid (preferably DNA) segment or a population of segments. Such Insert(s) can comprise one or more genes.

Insert Donor: is one of the two parental nucleic acid molecules RNA or DNA) of the present invention which carries the Insert. The Insert Donor molecule comprises the Insert flanked on both sides with recombination sites. The Insert Donor can be linear or circular. In one embodiment of the parent invention, the Insert Donor is a circular DNA molecule and further comprises a cloning vector sequence outside of the recombination signals (see FIG. When a population of Inserts or population of nucleic acid segments are used to make the Insert Donor, a population of Insert Donors result and may be used in accordance with the parent invention.

30/05/07.at 16620.specipgs. 13 14- Product: is one the desired daughter molecules comprising the A and D sequences which is produced after the second recombination event during the recombinational cloning process (see FIG. The Product contains the nucleic acid which was to be cloned or subcloned. In accordance with the parent invention, when a population of Insert Donors are used, the resulting population of Product molecules 00 will contain all or a portion of the population of Inserts of the Insert Donors and t) preferably will contain a representative population of the original molecules of the 0 Insert Donors.

Recognition sequence: Recognition sequences are particular sequences which a protein, chemical compound, DNA, or RNA molecule restriction endonuclease, a modification methylase, or a recombinase) recognizes and binds. In the present invention, a recognition sequence will usually refer to a recombination site. For example, the recognition sequence for Cre recombinase is loxP which is a 34 base pair sequence comprised of two 13 base pair inverted repeats (serving as the recombinase binding sites) flanking an 8 base pair core sequence. See FIG. 1 of Sauer, Current Opinion in Biotechnology 5:521-527 (1994). Other examples of recognition sequences are the attB, attP, attL, and attR sequences which are recognized by the recombinase enzyme X Integrase. attB is an approximately 25 base pair sequence containing two 9 base pair core-type Int binding sites and a 7 base pair overlap region. attP is an approximately 240 base pair sequence containing coretype Int binding sites and arm-type Int binding sites as well as sites for auxiliary proteins integration host factor (IHF), FIS and excisionase (Xis). See Landy, Current Opinion in Biotechnology 3:699-707 (1993). Such sites may also be engineered according to the present invention to enhance production of products in the methods of the invention. When such engineered sites lack the P1 or HI domains to make the recombination reactions irreversible attR or attP), such sites may be designated attR' or attP' to show that the domains of these sites have been modified in some way.

Recombinase: is an enzyme which catalyzes the exchange of DNA segments at specific recombination sites.

Recombinational Cloning: is a method described herein, whereby segments of nucleic acid molecules or populations of such molecules are exchanged, inserted, 30/05/07.at 16620.specipgs. 14 F, replaced, substituted or modified, in vitro or in vivo.

Recombination proteins: include excisive or integrative proteins, enzymes, co-factors or associated proteins that are involved in recombination reactions involving one or more recombination sites. See, Landy (1994), infra.

Repression cassette: is a nucleic acid segment that contains a repressor of a 00Selectable marker present in the subcloning vector.

In Selection scheme: is any method which allows selection, enrichment, or identification of a desired Product or Product(s) from a mixture containing the Insert r Donor, Vector Donor, any intermediates a Cointegrate), and/or Byproducts as described in the parent specifications.

In one embodiment of the parent specifications, the selection schemes (which can be carried out in reverse) will take one of three forms, which will be discussed in terms of FIG. 1. The first, exemplified herein with a Selectable marker and a repressor therefore, selects for molecules having segment D and lacking segment C.

The second selects against molecules having segment C and for molecules having segment D. Possible embodiments of the second form would have a DNA segment carrying a gene toxic to cells into which the in vitro reaction products are to be introduced. A toxic gene can be a DNA that is expressed as a toxic gene product (a toxic protein or RNA), or can be toxic in and of itself. (In the latter case, the toxic gene is understood to carry its classical definition of "heritable trait".) Examples of such toxic gene products are well known in the art and are described in the parent specifications.

Many genes coding for restriction endonucleases operably linked to inducible promoters are known and are described in the parent specification.

In the second form, segment D carries a Selectable marker. The toxic gene would eliminate transformants harboring the Vector Donor, Cointegrate, and Byproduct molecules, while the Selectable marker can be used to select for cells containing the Product and against cells harboring only the Insert Donor.

The third form selects for cells that have both segments A and D in cis on the same molecule, but not for cells that have both segments in trans on different molecules. This could be embodied by a Selectable marker that is split into two inactive fragments, one each on segments A and D.

31/05/07.at 16620.specipgs. -16- SThe fragments are so arranged relative to the recombination sites that when the segments are brought together by the recombination event, they reconstitute a functional Selectable marker. For example, the recombinational event can link a promoter with a structural gene, can link two fragments of a structural gene, or can link genes that encode a heterodimeric gene product needed for survival, or can link 00 portions of a replicon.

SVector Donor: is one of the two parental nucleic acid molecules RNA or DNA) of the present invention which carries the DNA segments comprising the DNA vector which is to become part of the desired Product. The Vector Donor comprises a subcloning vector D (or it can be called the cloning vector if the Insert Donor does not already contain a cloning vector) and a segment C flanked by recombination sites (see FIG. Segments C and/or D can contain elements that contribute to selection for the desired Product daughter molecule, as described above for selection schemes. The recombination signals can be the same or different, and can be acted upon by the same or different recombinases. In addition, the Vector Donor can be linear or circular.

Adapter: is an oligonucleotide or nucleic acid fragment or segment (preferably DNA) which comprises one or more recombination sites (or portions of such recombination sites) which in accordance with the invention can be added to a circular or linear Insert Donor molecule as well as other nucleic acid molecules described herein. When using portions of recombination sites, the missing portion may be provided by the Insert Donor molecule. Such adapters may be added at any location within a circular or linear molecule, although the adapters are preferably added at or near one or both termini of a linear molecule. Preferably, adapters are positioned to be located on both sides (flanking) a particularly nucleic acid molecule of interest. In accordance with the invention, adapters may be added to nucleic acid molecules of interest by standard recombinant techniques restriction digest and ligation). For example, adapters may be added to a circular molecule by first digesting the molecule with an appropriate restriction enzyme, adding the adapter at the cleavage site and reforming the circular molecule which contains the adapter(s) at the site of cleavage. Alternatively, adapters may be ligated directly to one or more and preferably both termini of a linear molecule thereby resulting in linear 30/05/07.at 16620.specipgs. 16 -17- O molecule(s) having adapters at one or both termini. In one aspect of the invention, adapters may be added to a population of linear molecules, a cDNA library or genomic DNA which has been cleaved or digested) to form a population of linear molecules containing adapters at one and preferably both termini of all or substantial portion of said population.

00 The terms host, promoter, selectable, marker, site-specific recombinase, In subcloning vector, vector, primer, template. Library amplification, oligonucleotide, (,i nucleotide and hybridization are as defined in the parent specification.

(Ni r Other terms used in the fields of recombinant DNA technology and molecular and cell biology as used herein will be generally understood by one of ordinary skill in the applicable arts.

Recombination Schemes One general scheme for an in vitro or in vivo method of the parent invention is shown in FIG. 1, where the Insert Donor and the Vector Donor can be either circular or linear DNA, but is shown as circular. Vector D is exchanged for the original cloning vector B. The Insert Donor need not comprise a vector. The method of the parent invention allows the Inserts A to be transferred into any number of vectors. According to the parent invention, the Inserts may be transferred to a particular Vector or may be transferred to a number of vectors in one step.

Additionally, the Inserts may be transferred to any number of vectors sequentially, for example, by using the Product DNA molecule as the Insert Donor in combination with a different Vector Donor. The nucleic acid molecule of interest may be transferred into a new vector thereby producing a new Product DNA molecule. The new Product DNA molecule may then be used as starting material to transfer the nucleic acid molecule of interest into a new vector. Such sequential transfers can be performed a number of times in any number of different vectors. Thus the parent invention allows for cloning or subcloning nucleic acid molecules and because of the ease and simplicity, these methods are particularly suited for high through-put applications. In accordance with the parent invention, it is desirable to select for the daughter molecule containing elements A and D and against other molecules, including one or more Cointegrate(s). The square and circle are different sets of 30O5S/07.ai 16620.specipgs. 17 18recombination sites lox sites or att sites). Segment A or D can contain at least one Selection Marker, expression signals, origins of replication, or specialized functions for detecting, selecting, expressing, mapping or sequencing DNA, where D is used in this example. This scheme can also be reversed according to the parent invention, as described herein. The resulting product of the reverse reaction the 00 Insert Donor) may then be used in combination with one or a number of vectors to t produce new product molecules in which the Inserts are contained by any number of

(N

vectors Examples of desired DNA segments that can be part of Element A or D include, but are not limited to, PCR products, large DNA segments, genomic clones or fragments, cDNA clones or fragments, functional elements, etc., and genes or partial genes, which encode useful nucleic acids or proteins. Moreover, the recombinational cloning of the present invention can be used to make ex vivo and in vivo gene transfer vehicles for protein expression (native or fusion proteins) and/or gene therapy.

In FIG. 1, the scheme provides the desired Product as containing A and Vector D, as follows. The Insert Donor (containing A and B) is first recombined at the square recombination sites by recombination proteins, with the Vector Donor (containing C and to form a Co-integrate having each of A-D-C-B. Next, recombination occurs at the circle recombination sites to form Product DNA (A and D) and Byproduct DNA (C and However, if desired, two or more different Cointegrates can be formed to generate two or more Products.

In one embodiment of the parent inventions in vitro or in vivo recombinational cloning method, a method for selecting at least one desired Product DNA is provided. This can be understood by consideration of the map of plasmid pEZC726 depicted in FIG. 2. The two exemplary recombination sites are attP and loxP. On one segment defined by these sites is a kanamycin resistance gene whose promoter has been replaced by the tetOP operator/promoter from transposon TnlO.

In the absence of tet repressor protein, E. coli RNA polymerase transcribes the kanamycin resistance gene from the tetOP. If tet repressor is present, it binds to tetOP and blocks transcription of the kanamycin resistance gene. The other segment of pEZC726 has the tet repressor gene expressed by a constitutive promoter. Thus 30/05/07.al 6620.specipgs. 18 18acells transformed by pEZC726 are resistant to chloramphenicol, because of the chloramphenicol acetyl transferase gene on the same segment as tetR, but are sensitive to kanamycin. The recombination reactions result in separation of the tetR Sgene from the regulated kanamycin resistance gene. This separation results in kanamycin resistance in cells receiving the desired recombination Product.

00 Two different sets of plasmids were constructed to demonstrate the in vitro t) method. One set, for use with Cre recombinase only (cloning vector 602 and O subcloning vector 629 (FIG. contained loxP and loxP 511 sites. A second set, for use with Cre and integrase (cloning vector 705 and subcloning vector 726 (FIG. 2)) contained loxP and att sites. The efficiency of production of the desired daughter plasmid was about 60 fold higher using both enzymes than using Cre alone.

Nineteen of twenty four colonies from the Cre-only reaction contained the desired product, while thirty eight of thirty eight colonies from the integrase plus Cre reaction contained the desired product plasmid.

A variety of other selection schemes can be used that are known in the art as they can suit a particular purpose for which the recombinational cloning is carried out as described in the parent specification.

In vitro method of selection can be devised for the Insert Donor and Vector Donor DNA molecules. Such scheme can involve engineering a rare restriction site in the starting circular vectors in such a way that after the recombination events the rare cutting sites end up in the Byproduct. Hence, when the restriction enzyme which binds and cuts at the rare restriction site is added to the reaction mixture in vitro, all of the DNA molecules carrying the rare cutting site, the starting DNA molecules, the Cointegrate, and the Byproduct, will be cut and rendered nonreplicable in the intended host cell. For example, cutting sites in segments B and C (see FIG. 1) can be used to select against all molecules 30/05/07.at 16620.specipgs. 12 19except the Product. Alternatively, only a cutting site in Cis needed if one is able Sto select for segment D, by a drug resistance gene not found on B.

SSimilarly, an in vitro selection method can be devised when dealing with linear DNA molecules. DNA sequences complementary to a PCR primer oo sequence can be so engineered that they are transferred, through the recombinational cloning method, only to the Product molecule. After the reactions are completed, the appropriate primers are added to the reaction solution and the sample is subjected to PCR. Hence, all or part of the Product molecule Sis amplified.

Other in vivo selection schemes can be used with a variety of host cells, particularly E. coli lines. One is to put a repressor gene on one segment of the subcloning plasmid, and a drug marker controlled by that repressor on the other segment of the same plasmid. Another is to put a killer gene on segment C of the subcloning plasmid (Figure Of course a way must exist for growing such a plasmid, there must exist circumstances under which the killer gene will not kill. There are a number of these genes known which require particular strains of E. coli. One such scheme is to use the restriction enzyme DpnI, which will not cleave unless its recognition sequence GATC is methylated. Many popular common E. coli strains methylate GATC sequences, but there are mutants in which cloned DpnI can be expressed without harm. Other restriction enzyme genes may also be used as a toxic gene for selection. In such cases, a host containing a gem encoding the corresponding methylase gene provides protected host for use in the invention. Similarly, the ccdB protein is a potent poison of DNA gyrase, efficiently trapping gyrase molecules in a cleavable complex, resulting in DNA strand breakage and cell death. Mutations in the gyrA subunit of DNA gyrase, specifically the gyrA462 mutation, confers resistance to ccdB (Bernard and Couturier, J. Mol. Bio. 226 (1992) 735-745). An E. coli strain, DB2, has been constructed that contains the gyrA462 mutation. DB2 cells containing plasmids that express the ccdB gene are not killed by ccdB. This strain is available from Life Technologies and has been deposited on October 14, 1997 with the Collection, Agricultural Research Culture Collection (NRRL), 1815 North University Street, Peoria, IL 61604 USA as deposit number NRRL B-21852.

SOf course analogous selection schemes can be devised for other host organisms. For example, the tet repressor/operator of TnO has been adapted to 00 control gene expression in eukaryotes (Gossen, and Bujard, Proc. Natl.

Acad. Sci. USA 89:5547-5551 (1992)). Thus the same control of drug resistance by the tet repressor exemplified herein or other selection schemes described herein can be applied to select for Product in eukaryotic cells.

Recombination Proteins In the present invention, the exchange ofDNA segments is achieved by the use of recombination proteins, including recombinases and associated co-factors and proteins. Various recombination proteins are described in the art. Examples of such recombinases include: Cre: A protein from bacteriophage PI (Abremski and Hoess, J. Biol.

Chem. 259(3):1509-1514 (1984)) catalyzes the exchange causes recombination) between 34 bp DNA sequences called loxP (locus of crossover) sites (See Hoess et al., Nucl. Acids Res. 14(5):2287 (1986)). Cre is available commercially (Novagen, Catalog No. 69247-1). Recombination mediated by Cre is freely reversible. From thermodynamic considerations it is not surprising that Cre-mediated integration (recombination between two molecules to form one molecule) is much less efficient than Cre-mediated excision (recombination between two loxP. sites in-the same molecule to form two daughter molecules).

Cre works in simple buffers with either magnesium or spermidine as a cofactor, as is well known in the art. The DNA substrates can be either linear or supercoiled. A number of mutant loxP sites have been described (Hoess et al., supra). One of these, loxP 511, recombines with another loxP 511 site, but will not recombine with a loxP site.

Integrase: A protein from bacteriophage lambda that mediates the integration of the lambda genome into the E. coli chromosome, The bacteriophage A Int recombinational proteins promote recombination between its substrate att sites as part of the formation or induction of a lysogenic state.

-21 Reversibility of the recombination reactions results from two independent pathways for integrative and excisive recombination. Each pathway uses a unique, Sbut overlapping, set of the 15 protein binding sites that comprise aft site DNAs.

cc Cooperative and competitive interactions involving four proteins (Int, Xis, IHF and FIS) determine the direction of recombination.

SIntegrative recombination involves the Int and IHF proteins and sites attP (240 bp) and attB (25 bp). Recombination results in the formation of two new I sites: attL and attR. Excisive recombination requires Int, IHF, and Xis, and sites attL and attR to generate attP and attB. Under certain conditions, FIS stimulates excisive recombination. In addition to these normal reactions, it should be appreciated that attP and attB, when placed on the same molecule, can promote excisive recombination to generate two excision products, one with attL and one with attR. Similarly, intermolecular recombination between molecules containing attL and attR, in the presence of Int, IHF and Xis, can result in integrative recombination and the generation ofattP and attB. Hence, by flanking

DNA

segments with appropriate combinations of engineered att sites, in the presence of the appropriate recombination proteins, one can direct excisive or integrative recombination, as reverse reactions of each other.

Each of the att sites contains a 15 bp core sequence; individual sequence elements of functional significance lie within, outside, and across the boundaries of this common core (Landy, Ann. Rev. Biochem. 58:913 (1989)). Efficient recombination between the various att sites requires that the sequence of the central common region be identical between the reconibining partners, however, the exact sequence is now found to be modifiable. Consequently, derivatives of the art site with changes within the core are now discovered to recombine as least as efficiently as the native core sequences.

Integrase acts to recombine the attP site on bacteriophage lambda (about 240 bp) with the attB site on the E. coli genome (about 25 bp) (Weisberg,

R.A.

and Landy, A. in Lambdall, p. 211 (1983), Cold Spring Harbor Laboratory)), to produce the integrated lambda genome flanked by attL (about 100 bp) and attR (about 160 bp) sites. In the absence of Xis (see below), this reaction is essentially irreversible. The integration reaction mediated by integrase and IHF works o -22in vitro, with simple buffer containing spermidine. Integrase can be obtained as described by Nash, Methods ofEnzymology 100:210-216 (1983). IHF can Sbe obtained as described by Filutowicz, et al., Gene 147:149-150 (1994).

Numerous recombination systems from various organisms can also be oo used, based on the teaching and guidance provided herein. See, Hoess etal, Nucleic Acids Research 14(6):2287 (1986); Abremski et al, J. Biol.

Chem.261(1):391 (1986); Campbell, J. Bacteriol. 174(23):7495 (1992); Qian et Sa., J. Biol. Chem. 267(11):7794 (1992); Araki et al., J. Mol. Biol. 225(1):25 (1992)). Many of these belong to the integrase family of recombinases (Argos et al. EMBO J. 5:433-440 (1986)). Perhaps the best studied of these are the Integrase/att system from bacteriophage A (Landy, A. (1993) Current Opinions in Genetics andDevel. 3:699-707), the Cre/loxP system from bacteriophage

PI

(Hoess and Abremski (1990) In Nucleic Acids and Molecular Biology, vol. 4.

Eds.: Eckstein and Lilley, Berlin-Heidelberg: Springer-Verlag; pp. 90-109), and the FLP/FRT system from the Saccharomyces cerevisiae 2 g circle plasmid (Broach et al. Cell 29:227-234 (1982)).

Members of a second family of site-specific recombinases, the resolvase family y6, Tn3 resolvase, Hin, Gin, and Cin) are also known. Members of this highly related family of recombinases are typically constrained to intramolecular reactions inversions and excisions) and can require hostencoded factors. Mutants have been isolated that relieve some of the requirements for host factors (Maeser and Kahnmann (1991)Mol Gen. Genet. 230:170-176), as well as some of-the constraints of intramolecular recombination.

Other site-specific recombinases similar to X Int and similar to P Cre can be substituted for Int and Cre. Such recombinases are known. In many cases the purification of such other recombinases has been described in the art. In cases when they are not known, cell extracts can be used or the enzymes can be partially purified using procedures described for Cre and Int.

While Cre and Int are described in detail for reasons of example, many related recombinase systems exist and their application to the described invention is also provided according to the present invention. The integrase family of sitespecific recombinases can be used to provide alternative recombination proteins -23 and recombination sites for the present invention, as site-specific recombination Sproteins encoded by, for example bacteriophage lambda, phi 80, P22, P2, 186, P4 Sand P1. This group of proteins exhibits an unexpectedly large diversity of sequences. Despite this diversity, all of the recombinases can be aligned in their Co terminal halves. A 4 0-residue region near the C terminus is particularly well 0 conserved in all the proteins and is homologous to a region near the C terminus N of the yeast 2 mu plasmid Flp protein. Three positions are perfectly conserved within this family: histidine, arginine and tyrosine are found at respective alignment positions 396, 399 and 433 within the well-conserved C-terminal region. These residues contribute to the active site of this family of recombinases, and suggest that tyrosine-433 forms a transient covalent linkage to DNA during strand cleavage and rejoining. See, Argos, P. et al., EMBO J. 5:433-40 (1986).

The recombinases of some transposons, such as those of conjugative transposons Tn916) (Scott and Churchward. 1995. Ann Rev Microbiol 49:367; Taylor and Churchward, 1997. J Bacteriol 179:1837) belong to the integrase family of recombinases and in some cases show strong preferences for specific integration sites (Ike et al 1992. J Bacteriol 174:1801; Trieu-Cuot et al, 1993. Mol. Microbiol 8:179).

Alternatively, IS231 and other Bacillus thuringiensis transposable elements could be used as recombination proteins and recombination sites.

Bacillus thuringiensis is an entomopathogenic bacterium whose toxicity is due to the presence in the sporangia ofdelta-endotoxin crystals active against agricultural pests and vectors of human and animal diseases. Most of the genes coding for these toxin proteins are plasmid-borne and are generally structurally associated with insertion sequences (IS231, IS232, IS240, ISBTI and ISBT2) and transposons (Tn4430 and Tn5401). Several of these mobile elements have been shown to be active and participate in the crystal gene mobility, thereby contributing to the variation of bacterial toxicity.

Structural analysis of the iso-IS23 1 elements indicates that they are related to IS 1151 from Clostridium perfringens and distantly related to IS4 and IS186 from Escherichia coli. Like the other IS4 family members, they contain a conserved transposase-integrase motif found in other IS families and retroviruses.

-24- Moreover, functional data gathered from IS23 IA in Escherichia coli indicate a Snon-replicative mode of transposition, with a preference for specific targets.

Similar results were also obtained in Bacillus subtilis and B. thuringiensis. See, Mahillon, J. et al., Genetica 93:13-26 (1994); Campbell, J. Bacteriol. 7495- 00 7499 (1992).

SAn unrelated family ofrecombinases, the transposases, have also been used O to transfer genetic information between replicons. Transposons are structurally variable, being described as simple or compound, but typically encode the Srecombinase gene flanked by DNA sequences organized in inverted orientations.

Integration oftransposons can be random or highly specific. Representatives such as Tn7, which are highly site-specific, have been applied to the efficient movement of DNA segments between replicons (Lucklow et al. 1993. J. Virol 67:4566- 4579).

A related element, the integron, are also translocatable-promoting movement of drug resistance cassettes from one replicon to another. Often these elements are defective transposon derivatives. Transposon Tn21 contains a class I integron called In2. The integrase (Intll) from In2 is common to all integrons in this class and mediates recombination between two 59-bp elements or between a 59-bp element and an attl site that can lead to insertion into a recipient integron.

The integrase also catalyzes excisive recombination. (Hall, 1997. Ciba Found Symp 207:192; Francia et al., 1997. J Bacteriol 179:4419).

Group II introns are mobile genetic elements encoding a catalytic

RNA

and protein. The protein coinponent possesses reverse franscriptase, maturase and an endonuclease activity, while the RNA possesses endonuclease activity and determines the sequence of the target site into which the intron integrates. By modifying portions of the RNA sequence, the integration sites into which the element integrates can be defined. Foreign DNA sequences can be incorporated between the ends of the intron, allowing targeting to specific sites. This process, termed retrohoming, occurs via a DNA:RNA intermediate, which is copied into 0 cDNA and ultimately into double stranded DNA (Matsuura et al., Genes and Dev 1997; Guo et al, EMBO J, 1997). Numerous intron-encoded homing endonucleases have been identified (Belfort and Roberts, 1997.

NAR

2 5 3 3 79).Such systems can be easily adopted for application to the described subcloning methods.

The amount of recombinase which is added to drive the recombination n reaction can be determined by using known assays. Specifically, titration assay is used to determine the appropriate amount of a purified recombinase enzyme, or the appropriate amount of an extract.

I SEngineered Recombination Sites SThe above recombinases and corresponding recombinase sites are suitable for use in recombination cloning according to the present invention. However, wild-type recombination sites may contain sequences that reduce the efficiency or specificity of recombination reactions or the function of the Product molecules as applied in methods of the present invention. For example, multiple stop codons in attB, attR, attP, attL and loxP recombination sites occur in multiple reading frames on both strands, so translation efficiencies are reduced, where the coding sequence must cross the recombination sites, (only one reading frame is available on each strand of loxP and attB sites) or impossible (in attP, attR or attL).

Accordingly, the present invention also provides engineered recombination sites that overcome these problems. For example, att sites can be engineered to have one or multiple mutations to enhance specificity or efficiency of the recombination reaction and the properties of Product DNAs attl, att2, and att3 sites); to decrease reverse reaction removing PI and HI from attR).

The testing of these mutants determines which mutants yield sufficient recombinational activity to be suitable for recombination subcloning according to the present invention.

Mutations can therefore be introduced into recombination sites for enhancing site specific recombination. Such mutations include, but are not limited to: recombination sites without translation stop codons that allow fusion proteins to be encoded; recombination sites recognized by the same proteins but differing in base sequence such that they react largely or exclusively with their homologous partners allowing multiple reactions to be contemplated; and mutations that -26prevent hairpin formation of recombination sites. Which particular reactions take place can be specified by which particular partners are present in the reaction mixture. For example, a tripartite protein fusion could be accomplished with parental plasmids containing recombination sites attRl and attLl; and attB3; attRl; attP3 and 10xP; and/or attR3 and 10xP; and/or attR3 and attL2.

c There are well known procedures for introducing specific mutations into nucleic acid sequences as described in the parent specification.

00 SThe following non-limiting methods can be used to modify or mutate a core region Sof a given recombination site to provide mutated sites that can be used in the parent S 10 invention: 1. By recombination of two parental DNA sequences by site-specific attL and attR to give attB) or other homologous) recombination mechanisms where the parental DNA segments contain one or more base alterations resulting in the final mutated core sequence; 2. By mutation or mutagenesis (site-specific, PCR, random, spontaneous, etc) directly of the desired core sequence; 3. By mutagenesis (site-specific, PCR, random, spontaneous, etc) of parental DNA sequences, which are recombined to generate a desired core sequence; 4. By reverse transcription of an RNA encoding the desired core sequence; and By de novo synthesis (chemical synthesis) of a sequence having the desired base changes.

2 3 /2/02.gc 3 120 spe.26 -27- The functionality of the mutant recombination sites can be demonstrated Sin ways that depend on the particular characteristic that is desired. For example, the lack of translation stop codons in a recombination site can be demonstrated by expressing the appropriate fusion proteins. Specificity ofrecombination between homologous partners can be demonstrated by introducing the appropriate Smolecules into in vitro reactions, and assaying for recombination products as described herein or known in the art. Other desired mutations in recombination sites might include the presence or absence of restriction sites, translation or Stranscription start signals, protein binding sites, and other known functionalities iN ofnucleic acid base sequences. Genetic selection schemes for particular functional attributes in the recombination sites can be used according to known method steps. For example, the modification of sites to provide (from a pair of sites that do not interact) partners that do interact could be achieved by requiring deletion, via recombination between the sites, of a DNA sequence encoding a toxic substance. Similarly, selection for sites that remove translation stop sequences, the presence or absence of protein binding sites, etc., can be easily devised by those skilled in the art.

Accordingly, the present invention provides a nucleic acid molecule, comprising at least one DNA segment having at least two engineered recombination sites flanking a Selectable marker and/or a desired DNA segment, wherein at least one of said recombination sites comprises a core region having at least one engineered mutation that enhances recombination in vitro in the formation of a Cointegrate DNA or a Product

DNA.

While in the preferred embodiment the recombination sites differ in sequence and do not interact with each other, it is recognized that sites comprising the same sequence can be manipulated to inhibit recombination with each other.

Such conceptions are considered and incorporated herein. For example, a protein binding site can be engineered adjacent to one of the sites. In the presence of the protein that recognizes said site, the recombinase fails to access the site and the other site is therefore used preferentially. In the cointegrate this site can no longer react since it has been changed e.g. from attB to attL. In resolution of the -28cointegrate, the protein can be inactivated by antibody, heat or a change of buffer) and the second site can undergo recombination.

The nucleic acid molecule can have at least one mutation that confers at least one enhancement ofsaid recombination, said enhancement selected from the 00group consisting of substantially favoring integration; (ii) favoring recombination; (ii) relieving the requirement for host factors; (iii) increasing the Nefficiency of said Cointegrate DNA or Product DNA formation; and (iv) increasing the specificity of said Cointegrate DNA or Product DNA formation.

The nucleic acid molecule preferably comprises at least one recombination site derived from attB, attP, attL or attR, such as attR' or attP'. More preferably the att site is selected from attl, att2, or att3, as described herein.

In a preferred embodiment, the core region comprises a DNA sequence selected from the group consisting of: RKYCWGCTTTYKTRTACNAASTSGB (m-att) (SEQ ID NO:1 I); AGCCWGCTTTYKTRTACNAACTSGB (m-attB) (SEQ ID NO:2); GTTCAGCTTTCKTRTACNAACTSGB (m-attR) (SEQ ID NO:3); AGCCWGCTTTCKTRTACNAAGTSGB (m-attL) (SEQ ID NO:4); GTTCAGCTTTYKTRTACNAAGTSGB(m-attP1) (SEQ ID RBYCW GCTTTYTTRTACWVAA STKGD (n-att) (SEQ ID NO:39); ASCCW GCTTTYTTRTACWAA STKGW (n-attB) (SEQ ID ASCCW GCTTTYTTRTACWAA GTTGG (n-attL) (SEQ ID NO:4 1); GTTCA GCTTTYTTRTACWAA STKGW (n-attR) (SEQ ID NO:42); GTTCA GCTTTYTTRTACWAA GTTGG (n-attP) (SEQ ID NO:43); -29or a corresponding or complementary DNA or RNA sequence, wherein R=A or G; K=G or T/U; Y=C or T/U; W=A or T/U; N=A or C or G or T/U; S=C or G; and B=C or G or T/U, as presented in 37 C.F.R. §1.822, which is entirely incorporated herein by reference, wherein the core region does not contain a stop codon in one or more reading frames.

00 The core region also preferably comprises a DNA sequence selected from the It) group consisting of: O AGCCTGCTTTTTTGTACAAACTTGT (attB1) (SEQ ID NO:6); AGCCTGCTTTCTTGTACAAACTTGT (attB2) (SEQ ID NO:7); ACCCAGCTTTCTTGTACAAAGTGGT (attB3) (SEQ ID NO:8); GTTCAGCTTTTTTGTACAAACTTGT (attR (SEQ ID NO:9); GTTCAGCTTTCTTGTACAAACTTGT (attR2) (SEQ ID NO: GTTCAGCTTTCTTGTACAAAGTGGT (attR3) (SEQ ID NO: 11); AGCCTGCTTTTTTGTACAAAGTTGG (attL1) (SEQ ID NO: 12); AGCCTGCTTTCTTGTACAAAGTTGG (attL2) (SEQ ID NO: 13); ACCCAGCTTTCTTGTACAAAGTTGG (attL3) (SEQ ID NO: 14); GTTCAGCTTTTTTGTACAAAGTTGG (attPl) (SEQ ID GTTCAGCTTTCTTGTACAAAGTTGG (attP2,P3) (SEQ ID NO:16); or a corresponding or complementary DNA or RNA sequence.

The parent inventions thus also provides a method for making a nucleic acid molecule, comprising providing a nucleic acid molecule having at least one engineered recombination site comprising at least one DNA sequence having at least 80-99% homology (or any range or value therein) to at least one of the above sequences, or any suitable recombination site, or which hybridizes under stringent conditions thereto, as known in the art.

Clearly, there are various types and permutations of such well-known in vitro and in vivo selection methods, each of which are not described herein for the sake of brevity. However, such variations and permutations are contemplated and considered to be the different embodiments of the present invention.

It is important to note that as a result of the preferred embodiment being in vitro recombination reactions, non-biological molecules such as PCR products can be manipulated via the present recombinational cloning method. In one example, it is 30/05/07.at 16620.specipgs. 12 possible to clone linear molecules into circular vectors.

There are a number of applications for the present invention. These uses include, but are not limited to, changing vectors, apposing promoters with genes, constructing genes for fusion proteins, changing copy number, changing replicons, cloning into phages, and cloning, PCR products (with an attB site 00 at one end and a loxP site at the other end), genomic DNAs, and cDNAs.

Vector Donors SIn accordance with the invention, any vector may be used to construct the S 10 Vector Donors of the invention as described in the parent specification.

30/0507 .atl 6620.specipgs.12 -31 Polymerases SPreferred polypeptides having reverse transcriptase activity those polypeptides able to catalyze the synthesis of a DNA molecule from an RNA template) include, but are not limited to Moloney Murine Leukemia Virus (M- MLV) reverse transcriptase, Rous Sarcoma Virus (RSV) reverse transcriptase Avian Myeloblastosis Virus (AMV) reverse transcriptase, Rous Associated Virus (R-AV) reverse transcriptase, Myeloblastosis Associated Virus (MAV) reverse transcriptase, Human Immunodeficiency Virus (HIV) reverse transcriptase retroviral reverse transcriptase, retrotransposon reverse transcriptase, hepatitis

B

reverse transcriptase, cauliflower mosaic virus reverse transcriptase and bacterial reverse transcriptase. Particularly preferred are those polypeptides having reverse transcriptase activity that are also substantially reduced in RNAse H activity (ie., "RNAse polypeptides). By a polypeptide that is "substantially reduced in RNase H activity" is meant that the polypeptide has less than about 20%, more preferably less than about 15%, 10% or and most preferably less than about of the RNase H activity of a wildtype or RNase H" enzyme such as wildtype M-MLV reverse transcriptase. The RNase H activity may be determined by a variety of assays, such as those described, for example, in U.S. Patent No.

5,244,797, in Kotewicz, M.L. e al., Nucl. Acids Res. 16:265 (1988) and in Gerard, e al., FOCUS 14(5):91 (1992), the disclosures of all ofwhich are fully incorporated herein by reference. Suitable RNAse H- polypeptides for use in the present invention include, but are not limited to, M-MLV H- reverse transcriptase, RSV H- reverse transcriptase, AMV H- reverse transcriptase,

RAV

H reverse transcriptase, MAV H- reverse transcriptase, HIV H- reverse transcriptase, and SUPERSCRIPTTM I reverse transcriptase and SUPERSCRIPTrM

II

reverse transcriptase which are available commercially, for example from Life Technologies, Inc. (Rockville, Maryland).

-32- SOther polypeptides having nucleic acid polymerase activity suitable for use Sin the present methods include thermophilic DNA polymerases such as DNA Spolymerase I, DNA polymerase III, Klenow fragment, T7 polymerase, and polymerase, and thermostable DNA polymerases including, but not limited to, Thermus thermophilus (Th) DNA polymerase, Thermus aquaticus (Taq)

DNA

polymerase, Thermotoga neopolitana (Tne) DNA polymerase, Thermotoga Smaritima (Tma) DNA polymerase, Thermococcus litoralis (Ti or VENT®) DNA polymerase, Pyrococcusfurios (Pfu or DEEPVENT®) DNA polymerase, Pyrococcus woosii (Pwo) DNA polymerase, Bacillus sterothermophilus (Bst) 1 DNA polymerase, Sulfolobus acidocaldarius (Sac) DNA polymerase, Thermoplasma acidophilum (Tac) DNA polymerase, Thermus favus (77T/T7b) DNA polymerase, Thermus ruber (Tru) DNA polymerase, Thermus brockianus (DYNAZYME®) DNA polymerase, Methanobacteriun thermoautotrophicium (Mth) DNA polymerase, and mutants, variants and derivatives thereof It will be understood by one of ordinary skill in the relevant arts that other suitable modifications and adaptations to the methods and applications described herein are readily apparent and may be made without departing from the scope of the invention or any embodiment thereof. Having now described the present invention in detail, the same will be more clearly understood by reference to the following examples, which are included herewith for purposes of illustration only and are not intended to be limiting of the invention.

Examples 2s The present recombinational cloning method accomplishes the exchange of nucleic acid segments to render something useful to the user, such as a change of cloning vectors. These segments must be flanked on both sides by recombination signals that are in the proper orientation with respect to one another. In the examples below the two parental nucleic acid molecules (e.g.

plasmids) are called the Insert Donor and the Vector Donor. The Insert Donor contains a segment that will become joined to a new vector contributed by the Vector Donor. The recombination intermediate(s) that contain(s) both starting -33 Smolecules is called the Cointegrate(s). The second recombinant event produces two daughter molecules, called the Product (the desired new clone) and the Byproduct.

Buffers Various known buffers can be used in the reactions of the present invention.

00 For restriction enzymes, it is advisable to use the buffers recommended by the in manufacturer. Alternative buffers can be readily found in the literature or can be (Ni O devised by those of ordinary skill in the art.

Examples 1-3. One exemplary buffer for lambda integrase is comprised of 50 mM Tris-HC1, at pH 7.5-7.8, 70 mM KC1 5 mM spermidine, 0.5 mM EDTA, and 0.25 mg/ml bovine serum albumin, and optionally, 10% glycerol.

One preferred buffer for PI Cre recombinase is comprised of 50 mM Tris- HC1 at pH 7.5, 33 mM NaC1, 5 mM spermidine, and 0.5 mg/ml bovine serum albumin.

The buffer for other site-specific recombinases which are similar to lambda Int and P1 Cre are either known in the art or can be determined empirically by the skilled artisans, particularly in light of the above-described buffers.

Example 1: Use of Topoisomerase to Stimulate Recombination The stimulation of the recombinant reaction by making one or the parental plasmids linear was not expected. If the stimulation resulted from relief of some conformation constraint arising during the two recombination reactions (formation of the Cointegrate and resolution to the two daughter molecules), then unwinding of the plasmids with a topoisomerase might also be stimulatory when one or both parental plasmids were circular.

The Insert Donor was Pezc2901 (Figure 7A), and the Vector Donor was pECZ3101 (Figure 7B). A portion of pEZC101 was linearized with Mlu I. 20 ng of pEZC2901 and/or pECZ3101 were used in each 10 pl reaction (29 ng Int, 2.9 ng Xis, 5.4 ng IHF in 50 mM Tris HCI pH about 7.8, 16.5 mM NaC1, 35 mM KC1, 5 mM spermidine, 0.375 mg/ml BSA, 0.25 mM EDTA, 2% glycerol).

30/05/07.a 16620.specipgs. 12 34-

OO

cr

O

in

(N

0", Topoisomerase I (from calf Thymus; Life Technologes, Inc.) was diluted from units/pl to the concentrations indicated in Table 1 in 1 X EZC buffer.

Table 1 1 2 3 4 5 6 7 8 9 Circular 3101 2 2 2 2 2 Linear 3101 2 2 2 2 2 Circular 2901 2 2 2 2 2 2 2 2 Recombinase 2 2 2 2 2 2 2 2 2 2 TE 2 2 Topoisomerase, 1:60 2 2 Topoisomerase, 1:20 2 2 Topoisomerase, 1:6 2 2 3 X Buffer 2 2 2 2 2 2 2 2 2 2 1 X Buffer 2 2 2 2 These reactions were assembled in the following order: buffer; TE; DNAs; Clonase; Topoisomerase. The reactions were incubated at 22* 28° for minutes, then at 70* for 5 minutes. I gl aliquots were transformed into UltraMax competentE. coli (Life Technologies, Inc.). Following expression, aliquots were plated on 100 pg/ml kanamycin and incubated at 30* for 48 hours. Results: see Table 2.

Table 2 Reaction Colonies Vector Donor Insert Recombinase Topo- Donor isomerase 1 0 linear 3101 2 245 linear 3101 circular 2901 3 221 linear 3101 circular 0.5 units 2901 4 290 linear 3101 circular 1.6 units 2901 355 linear 3101 circular 5 units 2901 6 0 circular 3101 7 23 circular 3101 circular 2901 8 209 circular 3101 circular 0.5 units 2901 9 119 circular 3101 circular 1.6 units 2901 195 circular 3101 circular 5 units 2901 30/05107.at 16620.specipgs. 12 Analysis Linearizing the Vector Donor increased the number of colonies about 10 fold (reaction 2 vs. reaction Addition of 0.5 to 5 units of topoisomerase I to reactions containing circular Insert Donor and linear Vector Donor had little or no effect on the number of colonies (reaction 2 compared to reactions 3, 4, and 5; maximum 1.4 00 fold). In contrast, if both parental plasmids were circular (reaction 7-10), the In addition of topoisomerase stimulated the number of colonies 5 to 9 fold. Thus, O addition of topoisomerase I to reactions in which both parental plasmids were circular stimulated the recombination reactions nearly as much as linearizing the 0 10 Vector Donor parent. Topoisomerase I was active when used in combination with the three recombination proteins, in recombination buffer. The addition of topoisomerase I to the recombination reaction relieves the necessity to linearize the Vector Donor to achieve stimulation of the recombination reactions.

Example 2: Recombination Cloning Using Cre and Cre Int Two pairs of plasmids were constructed to do the in vitro recombinational cloning method in two different ways. One pair, pEZC705 and pEZC726 (Figure 2A), was constructed with loxP and att sites, to be used with Cre and X integrase.

The other pair, pEZC602 and pEZC629 (Figure 3A), contained the loxP (wild type) site for Cre, and a second mutant lox site, loxP 511, which differs from loxP in one base (out of 34 total). The minimum requirement for recombinational cloning of the present invention is two recombination sites in each plasmid, in general X and Y, and X' and Recombinational cloning takes place if either or both types of site can recombine to form a Cointegrate X and and if either or both can recombine to excise the Product and Byproduct plasmids from the Cointegrate Y and It is important that the recombination sites on the same plasmid do not recombine. It was found that the present recombinational cloning could be done with Cre alone.

30/05/07.at 16620.specipgs. 12 36 00 Cre-Only C Two plasmids were constructed to demonstrate this conception (see

O

C Figure 3A). pEZC629 was the Vector Donor plasmid. It contained a constitutive 0drug marker (chloramphenicol resistance), an origin of replication, loxP and C loxP 511 sites, a conditional drug marker (kanamycin resistance whose expression is controlled by the operator/promoter of the tetracycline resistance operon of transposon Tn10), and a constitutively expressed gene for the tet repressor protein, tetR. E. coli cells containing pEZC629 were resistant to chloramphenicol at 30 pg/ml, but sensitive to kanamycin at 100 pg/ml. pEZC602 was the Insert Donor plasmid, which contained a different drug marker (ampicillin resistance), an origin, and loxP and loxP 511 sites flanking a multiple cloning site.

This experiment was comprised of two parts as follows: Part I: About 75 ng each of pEZC602 and pEZC629 were mixed in a total volume of 30 il of Cre buffer (50 mMTris-HCI pH 7.5, 33 mM NaCI, 5 mM spermidine-HCI, 500 g/ml bovine serum albumin). Two 10 pl aliquots were transferred to new tubes. One tube received 0.5 1 of Cre protein (approx. 4 units per p1; partially purified according to Abremski and Hoess, J. Biol. Chem.

259:1509 (1984)). Both tubes were incubated at 37C for 30 minutes, then for 10 minutes. Aliquots of each reaction were diluted and transformed into Following expression, aliquots were plated on 30 pg/ml chloramphenicol; 100 pg/ml ampicillin plus 200 pg/ml methicillin; or 100 pg/ml kanamycin.

Results: See Table 3. The reaction without Cre gave 1.1 x 106 ampicillin resistant colonies (from the Insert Donor plasmid pEZC602); 7.8x10' chloramphenicol resistant colonies (from the Vector Donor plasmid pEZC629); and 140 kanamycin resistant colonies (background). The reaction with added Cre gave 7.5xl0' ampicillin resistant colonies (from the Insert Donor plasmid pEZC602); 6. chloramphenicol resistant colonies (from the Vector Donor plasmid pEZC629); -37-

O

O

and 760 kanamycin resistant colonies (mixture of background colonies and colonies from the recombinational cloning Product plasmid). Analysis: Because the number of colonies on the kanamycin plates was much higher in the presence o of Cre, many or most of them were predicted to contain the desired Product plasmid.

Table 3 Enzyme Ampicillin Chloramphenicol Kanamycin Efficiency None l.lxl0 7.8x10 s 140 140/7.8x10 0.02% Part II: Twenty four colonies from the Cre" kanamycin plates were picked and inoculated into medium containing 100 jg/ml kanamycin. Minipreps were done, and the miniprep DNAs, uncut or cut with Sina or HindIII, were electrophoresed. Results: 19 of the 24 minipreps showed supercoiled plasmid of the size predicted for the Product plasmid. All 19 showed the predicted Smal and HindlII restriction fragments. Analysis: The Cre only scheme was demonstrated.

Specifically, it was determined to have yielded about 70% (19 of 24) Product clones. The efficiency was about 0.1% (760 kanamycin resistant clones resulted from 6. lx 10 chloramphenicol resistant colonies).

Cre Plus Integrase The plasmids used to demonstrate this method are exactly analogous to those used above, except that pEZC726, the Vector Donor plasmid, contained an attP site in place of loxP 511, and pEZC705, the Insert Donor plasmid, contained an attB site in place ofloxP 511 (Figure 2A).

This experiment was comprised of three parts as follows: Part I: About 500 ng of pEZC705 (the Insert Donor plasmid) was cut with Scal, which linearized the plasmid within the ampicillin resistance gene. (This was done because the integrase reaction has been historically done with the altB plasmid in a linear state Nash, personal communication). However, it was.

-38- Sfound later that the integrase reaction proceeds well with both plasmids supercoiled.) Then, the linear plasmid was ethanol precipitated and dissolved in

C

20 pl of A integrase buffer (50 mM Tris-HCI, about pH 7.8, 70 mM KCI, 5 mM spermidine-HCI, 0.5 mM EDTA, 250 pg/ml bovine serum albumin). Also, about 00 500 ng of the Vector Donor plasmid pEZC726 was ethanol precipitated and Ci dissolved in 20 pl A integrase buffer. Just before use, integrase (2 pl, Ci 393 pg/ml) was thawed and diluted by adding 18 pl cold A integrase buffer.

0 One pl IHF (integration host factor, 2.4 mg/ml, an accessory protein) was diluted Ci into 150 pl cold A integrase buffer. Aliquots (2 pl) of each DNA were mixed with integrase buffer, with or without 1 pl each X integrase and IHF, in a total of il. The mixture was incubated at 25C for 45 minutes, then at 70'C for minutes. Half of each reaction was applied to an agarose gel. Results: In the presence of integrase and IHF, about 5% of the total DNA was converted to a linear Cointegrate form. Analysis: Activity of integrase and IHF was confirmed.

Part II: Three microliters of each reaction with or without integrase and IHF) were diluted into 27 pl of Cre buffer (above), then each reaction was split into two 10 il aliquots (four altogether). To two of these reactions, 0.5 Pl of Cre protein (above) were added, and all reactions were incubated at 37'C for minutes, then at 70'C for 10 minutes. TE buffer (90 gl; TE: 10 mM Tris-HC1, pH 7.5, 1 mM EDTA) was added to each reaction, and 1 pl each was transformed into E. coli DH5a. The transformation mixtures were plated on 100 lig/ml ampicillin plus 200 pg/ml methicillin; 30 gg/ml chloramphenicol; or 100 Pg/ml kanamycin. Results: See Table 4.

39 00 4 Enym Ainpicillin Chioranphenicol Kanamycin Efficiency None 990 20000 4 41/2xl 04 =0.02% Cre only 280 3640 0 0 Integrase 1040 27000 9 9 /2.7x10'= only 0.03% Integrase 110 1110 76 76 1.1x1 3 6 9%/ *Integrase reactions also contained 11V.

Analysis: The Gre protein impaired transformation. When adjusted for this effect, the number of kanainycin resistant colonies, compared to the control reactions, increased more than 100 fold when both Gre and Integrase were used.

This suggests a specificity of greater than 99%/.

Part 111. 38 colonies were picked from the Integrase plus Gre plates, mniniprep DNAs were made and cut with HindIII to give diagnostic mapping information. Result: All 38 had precisely the expected fragment sizes. Analysis: The Cre plus X integrase method was observed to have much higher specificity than Gre-alone. Conclusion: The Gre plus I integrase method was demonstrated. Efficiency and specificity were much higher than for Gre only.

Example 3: Using in vitro Recombinaltional Cloning to Subclone the Chiamphenicol Acetyl Transferase Gene into a Vector for Exvpression in Eukaiyotic Cells (Figure 4A) An Insert Donor plasmid, pEZC843, was constructed, comprising the chioramphenicol acetyl transferase gene of E coi, cloned between loxP and aitB sites such that the loxP site was positioned at the 5'-end of the gene (Figure 4B).

A Vector Donor plasmid, pEZG 1003, was constructed, which contained the cytomnegalovirus eukaryotic promoter apposed to a IoxP site (Figure 4Q). One microliter aliquots of each supercoiled plasmid (about 50 ng crude miniprep DNA) wvere combined in a ten microliter reaction containing equal parts of lambda integrase buffer (50 mM Tris-HCI, pH 7.8, 70 mM KCI, 5 mM spermidine, mM EDTA, 0.25 mg/ml bovine serum albumin) and Cre recombinase buffer c (50 mM Tris-HCI, pH 7.5, 33 mM NaCI, 5 mM spermidine, 0.5 mg/ml bovine serum albumin), two units of Cre recombinase, 16 ng integration host factor, and 00 32 ng lambda integrase. After incubation at 30'C for 30 minutes and 75'C for C' 10 minutes, one microliter was transformed into competent E. coli strain

O

C (Life Technologies, Inc.). Aliquots of transformations were spread on agar plates 0 containing 200 pg/ml kanamycin and incubated at 37C overnight. An otherwise C1 identical control reaction contained the Vector Donor plasmid only. The plate receiving 10% of the control reaction transformation gave one colony; the plate receiving 10% of the recombinational cloning reaction gave 144 colonies. These numbers suggested that greater than 99% of the recombinational cloning colonies contained the desired product plasmid. Miniprep DNA made from six recombinational cloning colonies gave the predicted size plasmid (5026 base pairs), CMVProd. Restriction digestion with NcoI gave the fragments predicted for the chloramphenicol acetyl transferase cloned downstream of the CMV promoter for all six plasmids.

Example 4: Subcloned DNA Segments Flanked by attB Sites Without Stop Codons Part I: Background The above examples are suitable for transcriptional fusions, in which transcription crosses recombination sites. However, both anR and loxP sites contain multiple stop codons on both strands, so translational fusions can be difficult, where the coding sequence must cross the recombination sites, (only one reading frame is available on each strand of loxP sites) or impossible (in attR or attL).

A principal reason for subcloning is to fuse protein domains. For example, fusion of the glutathione S-transferase (GST) domain to a protein of interest allows the fusion protein to be purified by affinity chromatography on glutathione agarose (Pharmacia, Inc., 1995 catalog). If the protein of interest is fused to runs -41 of consecutive histidines (for example His6), the fusion protein can be purified by affinity chromatography on chelating resins containing metal ions (Qiagen, Inc.).

C It is often desirable to compare amino terminal and carboxy terminal fusions for activity, solubility, stability, and the like.

00 The attB sites of the bacteriophage I integration system were examined r as an alternative to loxP sites, because they are small (25 bp) and have some sequence flexibility (Nash, H.A. et al., Proc. Natl. Acad Sci. USA 84:4049-4053 (1987). It was not previously suggested that multiple mutations to remove all stop Scodes would result in useful recombination sites for recombinational subcloning.

Using standard nomenclature for site specific recombination in lambda bacteriophage (Weisber, in Lambda III, Hendrix, et al., eds., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1989)), the nucleotide regions that participate in the recombination reaction in an E. coli host cell are represented as follows: attP attB Int, IHF 0t Xis, Int, IHF attR attL where: O represents the 15 bp core DNA sequence found in both the phage and E. coli genomes; B and B' represent approximately 5 bases adjacent to the core in the E. coli genome; and P H P2, X, H2, C, P'1, P'2, and P'3 represent known DNA sequences encoding protein binding domains in the bacteriophage A genome.

The reaction is reversible in the presence of the protein Xis (excisionase); recombination between attL and attR precisely excise the genome from its integrated state, regenerating the circular 1 genome containing attP and the linear E. coli genome containing attB.

-42- Part II: Construction and Testing of Plasmids Containing Mutant att Sites Mutant attL and attR sites were constructed. Importantly, Landy et al.

m (Ann. Rev. Biochem. 58:913 (1989)) observed that deletion of the PI and HI domains of attP facilitated the excision reaction and eliminated the integration 5 reaction, thereby making the excision reaction irreversible. Therefore, as t mutations were introduced in attR, the PI and HI domains were also deleted.

attR' sites in the present example lack the P1 and H1 regions and have the Ndel site removed (base 27630 changed from C to and contain sequences Scorresponding to bacteriophage A coordinates 27619-27738 (GenBank release 92.0, bg:LAMCG, "Complete Sequence of Bacteriophage Lambda").

The sequence of attB produced by recombination of wild type attL and attR sites is: B 0 B' attBwt: 5' AGCCT GCTTTTTTATACTAA CTTGA 3' (SEQ.

ID

NO: 3' TCGGA CGAAAAATATf7AT GAACT 5' (SEQ.

ID

NO: 44) The stop codons are italicized and underlined. Note that sequences of attL, attR, and attP can be derived from the attB sequence and the boundaries of bacteriophage A contained within attL and attR (coordinates 27619 to 27818).

When mutant attRl (attR') and attL1 sites were recombined the sequence anB I was produced (mutations in bold, large font): B 0 B' attBl: 5' AGCCT GCTrTTTTGTACAAA CTTGT3 (SEQ.

ID

NO: 6) 3' TCGGA CGAAAAAACATGTTT GAACA5 (SEQ.

ID

NO: Note that the four stop codons are gone.

-43.- When an additional mutation was introduced in the attR 1 (attR') and attL I sequences (bold), attR2 (attR') and attL2 sites resulted. Recombination of attR2 and attL2 produced the attB2 site: 00 B 0 B' attB2: 5' AGCCT GCTICTTGTACAAA CTTGT 3, (SEQ.

ID

3' TCGGA cGAAAGAACATGTTT GAC 5i (SQ c-i

ID

NO: 46) The recombination activities of the above attL and attR' sites were assayed as follows. The attB site of plasmid pEZC7O5 (Figure 2B) was replaced with attLwt, attLI, or attL2. The atiP site of plasmid pEZC726 (Figure 2C) was replaced with attRwt, attRl (attR', lacking regions P1I and HI1) or attR2 (attR', lacking regions P1 and HI). Thus, the resulting plasmids could recombine via their loxP sites, mediated by Ore, and via their attR' and attL sites, mediated by Int, Xis, and fIF. Pairs of plasmids were mixed and reacted with Cre, bIt, Xis, and IHU, transformed into E. coli competent cells, and plated on agar containing kanamycin. The results are presented in Table -44 00 ~33 Table Vector donor att site Gene donor att site of kanamycin resistant colonies" attRkI wt None 1 (background) (pEZC13OI) attLwt(pEZC1313) 147 attLi (pEZC1317) 47 attL2 (pEZC1321) 0 attR' I (pEZC 1305) None I (background) attbwt (EZC 1313) 4 attLI (pEZC1317) 128 attL2 (pEZCI321) 0 attR' 2 (pEZC13O9) None 0 (background) attLwt (pEZC 1313) 0 attLi (pEZC1317) 0 209 I 1% of each transformation was spread on a kanamycin plate.) The above data show that whereas the wild type att and atti sites recombine to a small extent, the atti and att2 sites do not recombine detectably with each other.

Pad HLI Recombination was demonstrated when the core region of both attB sites flanking the DNA segment of interest did not contain stop codons. The physical state of the participating plasmids was discovered to influence recombination efficiency.

The appropriate att sites were moved into pEZC7OS and pEZC726 to make the plasmids pEZCI1405 (Figure 5G) (attR I I and attR 1 2) and pEZC 1502 (Figure 5H) (attLi and attL2). The desired DNA segment in this experiment was a copy of the chioramphenicol resistance gene cloned between the two attL sites of pEZC 1502. Pairs of plasmids were recombined in vitro using Int, Xis, and IF (no Cre because no lox? sites were present). 100 ng of each plasmid were incubated in 10 pi reactions of 50 mM Tris HCI pH about 7.8, 16.5 nm NaCI, 3 mM KCI, 5 mM sperinidine, 0.25 mM EDTA, 0.375 mg/mI BSA, 3% glycerol that contained 8.1 ng Il-F, 43 ng Int, 4.3 ng Xis, and 2 units Cre. Reactions were incubated at 25 0 C for 45 min., 65-C for 10 min, and 1 p1 aliquots were transformed into DH4cc cells, and spread on kanamycin plates. The yield of desired kanamycin resistant colonies was determined when both parental plasmids were circular, or when one plasmid was circular and the other linear as presented in Table 6.

5 Table 6 00 Vector donor' Insert donor' Kanamycin resistant colonies' Circular pEZC1405 None Circular pEZC1405 Circular pEZC1502 2680 Linear pEZC1405 None Linear pEZC1405 Circular pEZC1502 172000 Circular UEZC1405 Linear pEZC1502 73000 DNAs were purified with Qiagen columns, concentrations determined by A260, and linearized with Xbal (pEZC1405) or AlwNI (pEZC1502). Each reaction contained 100 ng of the indicated DNA. All reactions (10 pl total) contained 3 pl of enzyme mix (Xis, Int, and IHF). After incubation (45 minutes at 25', minutes at 65 one pl was used to transform E. coli DH5a cells.

2 Number of colonies expected if the entire transformation reaction (1 ml) had been plated. Either 100 tl or I pl of the transformations were actually plated.

Analysis: Recombinational cloning using mutant attR and attL sites was confirmed. The desired DNA segment is subcloned between attB sites that do not contain any stop codons in either strand. The enhanced yield of Product DNA (when one parent was linear) was unexpected because of earlier observations that the excision reaction was more efficient when both participating molecules were supercoiled and proteins were limiting (Nunes-Duby et al., Cell 50:779-788 (1987).

Example 5: Demonstration of Recombinational Cloning Without Inverted Repeats Part I: Rationale The above Example 4 showed that plasmids containing inverted repeats of the appropriate recombination sites (for example, attL1 and attL2 in plasmid pEZC1502) (Figure 5H) could recombine to give the desired DNA segment flanked by attB sites without stop codons, also in inverted orientation. A concern -46was the in vivo and in vitro influence of the inverted repeats. For example, transcription of a desired DNA segment flanked by attB sites in inverted c orientation could yield a single stranded RNA molecule that might form a hairpin structure, thereby inhibiting translation.

00 Inverted orientation of similar recombination sites can be avoided by C' placing the sites in direct repeat arrangement att sites. If parental plasmids each C' have a wild type attL and wild type attR site, in direct repeat the Int, Xis, and IHF Sproteins will simply remove the DNA segment flanked by those sites in an C' intramolecular reaction. However, the mutant sites described in the above Example 4 suggested that it might be possible to inhibit the intramolecular reaction while allowing the intermolecular recombination to proceed as desired.

Part II: Structure of Plasmids Without Inverted Repeatsfor Recombinational Cloning The attR2 sequence in plasmid pEZC1405 (Figure 5G) was replaced with attL2, in the opposite orientation, to make pEZC1603 (Figure 6A). The attL2 sequence of pEZC1502 (Figure 5H) was replaced with attR2, in the opposite orientation, to make pEZCI706 (Figure 6B). Each of these plasmids contained mutations in the core region that make intramolecular reactions between attl and att2 cores very inefficient (see Example 3, above).

Plasmids pEZC1405, pEZC 1502, pEZC 1603 and pEZC 1706 were purified on Qiagen columns (Qiagen, Inc.). Aliquots of plasmids pEZC1405 and pEZC1603 were linearized with Xbal. Aliquots of plasmids pEZC1502 and pEZC 706 were linearized with AlwNI. One hundred ng of plasmids were mixed in buffer (50 mM Tris HCL pH about 7.8, 16.5 mM NaCI, 35 mM KCI, 5 mM spermidine, 0.25 mM EDTA, 0.375 mg/ml BSA, 3% glycerol) containing Int (43.5 ng), Xis (4.3 ng) and IHF (8.1 ng) in a final volume of 10 pl. Reactions were incubated for 45 minutes at 25"C, 10 minutes at 65*C, and 1 pl was transformed into E. coli DH5a. After expression, aliquots were spread on agar plates containing 200 pg/ml kanamycin and incubated at 37 0

C.

Results, expressed as the number of colonies per 1 [l of recombination reaction are presented in Table 7.

-47- Table 7 Vector Donor Gene Donor Colonies Predicted product Circular 1405 100 Circular 1405 Circular 1502 3740 3640/3740 97% Linear 1405 90 Linear 1405 Circular 1502 172,000 171,910/172,000 99.9% Circular 1405 Linear 1502 73,000 72,900/73,000 99.9% Circular 1603 Circular 1603 Circular 1706 410 330/410 Linear 1603 270 Linear 1603 Circular 1706 7000 6730/7000 96% Cirular 1603 Linear 1706 10-800 10-530/10 -00 97% Analysis. In all configurations, circular or linear, the pEZC1405 x pEZC 1502 pair (with att sites in inverted repeat configuration) was more efficient than pEZC1603 x pEZC1706 pair (with att sites mutated to avoid hairpin formation). The pEZC 1603 x pEZC 1706 pair gave higher backgrounds and lower efficiencies than the pEZC1405 x pEZC 502 pair. While less efficient, 80% or more of the colonies from the pEZC1603 x pEZC1706 reactions were expected to contain the desired plasmid product. Making one partner linear stimulated the reactions in all cases.

Part III: Confirmation of Product Plasmids'Structure Six colonies each from the linear pEZC1405 (Figure 5G) x circular pEZC 1502 (Figure SH), circular pEZC 1405 x linear pEZC 1502, linear pEZC 1603 (Figure 6A) x circular pEZC1706 (Figure 6B), and circular pEZC1603 x linear pEZC1706 reactions were picked into rich medium and miniprep DNAs were prepared. Diagnostic cuts with Ssp I gave the predicted restriction fragments for all 24 colonies.

-48-

O

O

Analysis. Recombination reactions between plasmids with mutant attL and attR sites on the same molecules gave the desired plasmid products with a c high degree of specificity.

00 C' Example 6: Recombinational Cloning with a Toxic Gene 0 0Part I: Background C Restriction enzyme DpnI recognizes the sequence GATC and cuts that sequence only if the A is methylated by the dam methylase. Most commonly used E. coli strains are dam'. Expression of DpnI in dam' strains of E. coil is lethal because the chromosome of the cell is chopped into many pieces. However, in danm cells expression of DpnI is innocuous because the chromosome is immune to DpnI cutting.

In the general recombinational cloning scheme, in which the vector donor contains two segments C and D separated by recombination sites, selection for the desired product depends upon selection for the presence of segment D, and the absence of segment C. In the original Example segment D contained a drug resistance gene (Km) that was negatively controlled by a repressor gene found on segment C. When C was present, cells containing D were not resistant to kanamycin because the resistance gene was turned off.

The DpnI gene is an example of a toxic gene that can replace the repressor gene of the above-embodifnent. If segment C expresses the DpnI gene product, transforming plasmid CD into a dam* host kills the cell. If segment D is transferred to a new plasmid, for example by recombinational cloning, then selecting for the drug marker will be successful because the toxic gene is no longer present.

Part II: Construction of a Vector Donor Using Dpnl as a Toxic Gene The gene encoding DpnI endonuclease was amplified by PCR using primers S'CCA CCA CAA ACG CGT CCA TGG AAT TAC ACT TTA ATT TAG3' (SEQ. ID NO: 17) and 5'CCA CCA CAA GTC GAC GCA TGC CGA -49u CAG CCT TCC AAA TGT3' (SEQ ID NO: 18) and a plasmid containing the Dpnl gene (derived from plasmids obtained from Sanford A. Lacks, Brookhaven c National Laboratory, Upton, New York; also available from American Type Culture Collection as ATCC 67494) as the template.

00 Additional mutations were introduced into the B and B' regions ofattL and SattR respectively, by amplifying existing attL and attR' domains with primers Scontaining the desired base changes. Recombination of the mutant attL3 (made 0with oligo Xis115) and attR' 3 (attR', made with oligo Xis 12) yielded attB3 C with the following sequence (differences from attBI in bold): B 0 B' ACCCA GCTTTCTTGTACAAA GTGGT (SEQ ID NO: 8) TGGGT CGAAAGAACATGTTT CACCA (SEQ ID NO:47) The attL3 sequence was cloned in place of attL2 of an existing Gene Donor plasmid to give the plasmid pEZC2901 (Figure 7A). The attR' 3 sequence was cloned in place of attR' 2 in an existing Vector Donor plasmid to give plasmid pEZC2913 (Figure 71). The DpnI gene was cloned into plasmid pEZC2913 to replace the tet repressor gene. The resulting Vector Donor plasmid was named pEZC3101 (Figure 7C). When pEZC3101 was transformed into the dam strain SCS 110 (Stratagene), hundreds of colonies resulted. When the same plasmid was transformed into the dam+ strain DH5a, only one colony was produced, even though the DH5 a cells were about 20 fold more competent than the SCS 110 cells.

When a related plasmid that did not contain the DpnI gene was transformed into the same two cell lines, 28 colonies were produced from the SCS 110 cells, while 448 colonies resulted from the DH5a cells. This is evidence that the Dpn I gene is being expressed on plasmid pEZC3101 (Figure 7C), and that it is killing the dam' DH5a cells but not the dam' SCSI 10 cells.

50 Part III. Demonstration of Recombinalionol Coning Using DpnJ Selection A pair of plasmaids was used to demonstrate recombinational cloning with selection for Product dependent upon the toxic gene DpnI. Plasmid pEZC3 101 (Figure 7Q) was linearized with Milul and reacted with circular plasrnid pEZC290I (Figure 7A). A second pair of plasmids using selection based on control of drug resistance by a repressor gene was used as a control: plasmid pEZC 1802 (Figure 7D) was linearized with Xbal and reacted with circular plasmid pE-ZC 1502 (Figure SH). Eight microliter reactions containing buffer (50 mM Tris HCI pH about 7.8, 16.5 mM NaCl, 35 mM KCI, 5 mM spermidine, 0.375 mg/mI BSA, 0.25 mM EDTA, 2.5% glycerol) and proteins Xis (2.9 ng), mnt (29 ng), and IHF (5.4 ng) were incubated for 45 minutes at 25*C, then 10 minutes at 75*C, and I Al aliquots were transformed into DI-I~c darn+) competent cells, as presented in Table 8.

Table 8 Reaction Vector donor Basis of selection hisert donor Colonies 201pEZC31OI/Mlu Dpn Itoxicity -3 2 pEZC3 l~l/Mlu Dpn I toxicity Circular 4000 3 pEZC 1 802/Xba Tet repressor -0 4 pEZC I 8O2Xba Tet repressor Circular 12100 1502 Miniprep DNAs were prepared from four colonies from reaction and cut with restriction enzyme Ssp I. All gave the predicted fragments.

Analysis: Subcloning using selection with a toxic gene was demonstrated.

Plasmids of the predicted structure were produced.

51 Example 7: Cloning of Genes with Uracil DNA Glycosylase and Subcloning of the Genes with Recombinational Cloning to Make Fusion SProteins Part I: Converting an Existing Expression Vector to a Vector Donor for 00 Recombinational Cloning A cassette useful for converting existing vectors into functional Vector

V')

C' Donors was made as follows. Plasmid pEZC3101 (Figure 7C) was digested with C Apal and Kpnl, treated with T4 DNA polymerase and dNTPs to render the ends S 10 blunt, further digested with Smal, Hpal, and AlwNI to render the undesirable C DNA fragments small, and the 2.6 kb cassette containing the attR' 1 Cm Dpn I attR domains was gel purified. The concentration of the purified cassette was estimated to be about 75 ng DNA/pl.

Plasmid pGEX-2TK (Figure 8A) (Pharmacia) allows fusions between the protein glutathione S transferase and any second coding sequence that can be inserted in frame in its multiple cloning site. pGEX-2TK DNA was digested with Smal and treated with alkaline phosphatase. About 75 ng of the above purified DNA cassette was ligated with about 100 ng of the pGEX-2TK vector for hours in a 5 pl ligation, then 1 ll was transformed into competent E. coli BRL 3056 cells (a dam' derivative of DHIOB; dam' strains commercially available include DMI from Life Technologies, Inc., and SCS 110 from Stratagene).

Aliquots of the transformation mixture were plated on LB agar containing 100 pg/ml ampicillin (resistance gene present on pGEX-2TK) and 30 pg/ml chloramphenicol (resistance gene present on the DNA cassette). Colonies were picked and miniprep DNAs were made. The orientation of the cassette in pGEX- 2TK was determined by diagnostic cuts with EcoRI. A plasmid with the desired orientation was named pEZC3501 (Figure 8B).

Part II: Cloning Reporter Genes Into an Recombinational Cloning Gene Donor Plasmid in Three Reading Frames Uracil DNA glycosylase (UDG) cloning is a method for cloning PCR amplification products into cloning vectors patent No. 5,334,515, entirely incorporated herein by reference). Briefly, PCR amplification of the desired DNA -52segment is performed with primers that contain uracil bases in place ofthymidine bases in their 5' ends. When such PCR products are incubated with the enzyme UDG, the uracil bases are specifically removed. The loss of these bases weakens base pairing in the ends of the PCR product DNA, and when incubated at a 0 5 suitable temperature 37*C), the ends of such products are largely single t stranded. If such incubations are done in the presence of linear cloning vectors Scontaining protruding 3' tails that are complementary to the 3' ends of the PCR products, base pairing efficiently anneals the PCR products to the cloning vector.

When the annealed product is introduced into E. coli cells by transformation, in vivo processes efficiently convert it into a recombinant plasmid.

UDG cloning vectors that enable cloning of any PCR product in all three reading frames were prepared from pEZC3201 (Figure 8K) as follows. Eight oligonucleotides were obtained from Life Technologies, Inc. (all written 5' 3': rfl top (GGCC GAT TAC GAT ATC CCA ACG ACC GAA AAC CTG TAT TTT CAG GGT) (SEQ. ID NO:19), rfl bottom (CAG GTT TTC GGT CGT TGG GAT ATC GTA ATC)(SEQ. ID NO:20), rf2 top (GGCCA GAT TAC GAT ATC CCA ACG ACC GAA AAC CTG TAT TTT CAG GGT)(SEQ. ID NO:21), rf2 bottom (CAG GTT TTC GGT CGT TGG GAT ATC GTA ATC T)(SEQ. ID NO:22), rf3 top (GGCCAA GAT TAC GAT ATC CCA ACG ACC GAA AAC CTG TAT TTT CAG GGT)(SEQ. ID NO:23), rf3 bottom (CAG GTT TTC GGT CGT TGG GAT ATC GTA ATC TT)(SEQ. ID NO:24), carboxy top (ACC GTT TAC GTG GAC)(SEQ. ID NO:25) and carboxy bottom (TCGA GTC CAC GTA AAC GGT TCC CAC TTA TTA)(SEQ. ID NO:26). The rfl, 2, and 3 top strands and the carboxy bottom strand were phosphorylated on their 5' ends with T4 polynucleotide kinase, and then the complementary strands of each pair were hybridized. Plasmid pEZC3201 (Figure 8K) was cut with NotI and Sail, and aliquots of cut plasmid were mixed with the carboxy-oligo duplex (Sal I end) and either the rfl, rf2, or rf3 duplexes (Notl ends) (10 pg cut plasmid (about 5 pmol) mixed with 250 pmol carboxy oligo duplex, split into three 20 l volumes, added 5 Il (250 pmol) of rfl, rf2, or rf3 duplex and 2 pl 2 units T4 DNA ligase to each reaction). After 90 minutes -53of ligation at room temperature, each reaction was applied to a preparative agarose gel and the 2.1 kb vector bands were eluted and dissolved in 50 pl ofTE.

Part III: PCR of CAT andphoA Genes Primers were obtained from Life Technologies, Inc., to amplify the chloramphenicol acetyl transferase (CAT) gene from plasmid pACYC184, and O phoA, the alkaline phosphatase gene from E. coli. The primers had 12-base extensions containing uracil bases, so that treatment of PCR products with uracil DNA glycosylase (UDG) would weaken base pairing at each end ofthe DNAs and allow the 3' strands to anneal with the protruding 3' ends of the rfl, 2, and 3 vectors described above. The sequences of the primers (all written 5' 3') were: CAT left, UAU UUU CAG GGU ATG GAG AAA AAA ATC ACT GGA TAT ACC (SEQ. ID NO:27); CAT right, UCC CAC UUA UUA CGC CCC GCC CTG CCA CTC ATC (SEQ. ID NO:28); phoA left, UAU UUU CAG GGU ATG CCT GTT CTG GAA AAC CGG (SEQ. ID NO:29); and phoA right, UCC CAC UUA UUA TTT CAG CCC CAG GGC GGC TTT C (SEQ. ID The primers were then used for PCR reactions using known method steps (see.

U.S. patent No. 5,334,515, entirely incorporated herein by reference), and the polymerase chain reaction amplification products obtained with these primers comprised the CAT or phoA genes with the initiating ATGs but without any transcriptional signals. In addition, the uracil-containing sequences on the amino termini encoded the cleavage site for TEV protease (Life Technologies, Inc.), and those on the carboxy terminal encoded consecutive TAA nonsense codons.

Unpurified PCR products (about 30 ng) were mixed with the gel purified, linear rfl, rf2, or rf3 cloning vectors (about 50 ng) in a 10 pl reaction containing IX REact 4 buffer (LTI) and 1 unit UDG (LTI). After 30 minutes at 37 0 C, 1 pl aliquots of each reaction were transformed into competent E. coli DH5a cells (LTI) and plated on agar containing 50 gg/ml kanamycin. Colonies were picked and analysis of miniprep DNA showed that the CAT gene had been cloned in reading frame 1 (pEZC3601) (Figure 8C), reading frame 2 (pEZC3609) (Figure 8D) and reading frame 3 (pEZC3617) (Figure 8E), and that the phoA gene -54had been cloned in reading frame 1 (pEZC3606) (Figure 8F), reading frame 2 (pEZC3613) (Figure 8G) and reading frame 3 (pEZC3621) (Figure 8H).

Part IV: Subcloning of CAT orphoAfrom UDG Cloning Vectors into a 00 5 GST Fusion Vector Plasmids encoding fusions between GST and either CAT or phoA in all three reading frames were constructed by recombinational cloning as follows.

Miniprep DNA of GST vector donor pEZC3501(Figure 8B) (derived from S 10 Pharmacia plasmid pGEX-2TK as described above) was linearized with Clal.

About 5 ng of vector donor were mixed with about 10 ng each of the appropriate circular gene donor vectors containing CAT or phoA in 8 ul reactions containing buffer and recombination proteins Int, Xis, and IHF (Example After incubation, 1 pl of each reaction was transformed into E. coli strain DH5a and plated on ampicillin, as presented in Table 9.

Table 9 DNA Colonies (10% of each transformation) Linear vector donor (pEZC3501/Cla) 0 Vector donor CAT rfl 110 Vector donor CAT rf2 71 Vector donor CAT rf3 148 Vector donor phoA rfl 121 Vector donor phoA rf2 128 Vector dnor phnA rf3 31 Part V: Expression of Fusion Proteins Two colonies from each transformation were picked into 2 ml of rich medium (CircleGrow, BiolOl Inc.) in 17 x 100 mm plastic tubes (Falcon 2059, Becton Dickinson) containing 100 pg/ml ampicillin and shaken vigorously for about 4 hours at 37*C, at which time the cultures were visibly turbid. One ml of each culture was transferred to a new tube containing 10 p1 of 10% IPTG to induce expression ofGST. After 2 hours additional incubation, all cultures had about the same turbidity; the A600 of one culture was 1.5. Cells from 0.35 ml Ceach culture were harvested and treated with sample buffer (containing SDS and P-mercaptoethanol) and aliquots equivalent to about 0.15 A600 units of cells were 00 applied to a Novex 4-20% gradient polyacrylamide gel. Following electrophoresis C the gel was stained with Coomassie blue.

C Results: Enhanced expression of single protein bands was seen for all 12 cultures. The observed sizes of these proteins correlated well with the sizes predicted for GST being fused (through attB recombination sites without stop codons) to CAT (Figure 81) or phoA (Figure 8J) in three reading frames: CAT rfl 269 amino acids; CAT rf2 303 amino acids; CAT rf3 478 amino acids; phoA rfl 282 amino acids; phoA rf2 280 amino acids; and phoA rf3 705 amino acids.

Analysis: Both CAT and phoA genes were subcloned into a GST fusion vector in all three reading frames, and expression of the six fusion proteins was demonstrated.

Example 8: Reverse Recombination and Subcloning by Recombination Two plasmids were constructed to demonstrate reverse recombination according to the present invention. The vector pEZC5601 (Figure containing attB recombination sites and termed the attB parent plasmid (this vector may correspond to the Product DNA), further contained an ampicillin resistance gene, an origin of replication, an attB2 site, a tetracycline resistance gene, and an attB0 site, as described above. Plasmid pEZC6701 (Figure containing attP recombination sites and termed the attP parent plasmid (this vector may correspond to the Byproduct DNA or may correspond to a different Vector Donor DNA), also contained a kanamycin resistance gene, an origin of replication, an attP2 site, a gene encoding the toxic protein ccdB, and an attPO site. Integrase buffer at 10 X concentration comprised 0.25 M Tris HCI pH 0.25 M Tris HC pH 8.0, 0.7 M potassium chloride, 50 mM spermidine HCI, EDTA, and 2.5 mg /ml BSA. Note that attPO and attP2 contained the PI -56and H1 domains. Integrase (1.5 pl of 435 ng /pl) and IHF (1.5 p1 of 16 ng /pl in I X Integrase buffer) were mixed with 13.5 tl of 1 X Int buffer to make the C recombinase mixture.

Two 8 pl reactions were assembled. Reaction A contained 300 ng pEZC6701 plasmid and 2 pl of recombinase mixture in 1 X Integrase buffer.

Reaction B contained 300 ng pEZC5601, 300 ng pEZC6701, and 2 pl of recombinase mixture in 1 X Integrase buffer. Both reactions were incubated at for 45 minutes, then at 70*C for 5 minutes, and then cooled. TE buffer C (792 pl of 10 mM Tris HC1 pH 7.5, 1 mM EDTA) was added to each reaction, and 1 pl of this diluted reaction was transformed into DH5a UltraMax competent E. coli cells (Life Technologies, Inc., Rockville, MD). After 1 hour of expression in non-selective medium, one tenth (100 pl) of each transformation was spread onto agar plates containing 100 pg/ml kanamycin.

After overnight incubation at 37*C, the plate from reaction A contained 1 colony, while the plate from reaction B contained 392 colonies. Twelve colonies were picked from the reaction B plate into rich liquid medium and grown overnight. Miniprep DNAs prepared from these cultures were run uncut on an agarose gel and all 12 contained a plasmid of about 3.8 kb. Six of the miniprep DNAs were cut with restriction enzyme Clal and run along with pEZC6701 (the kanamycin resistant parental plasmid) also cut with Clal. Plasmid pEZC6701 was cut once with Clal to give a fragment of about 3.8 kb. The six miniprep DNAs cut twice with Clal to give fragments of about 900 base pairs and about 2900 base pairs.

Analysis: Recombination between the attP sites on pEZC6701 and the attB sites on pEZC5601 resulted in the production of two daughter plasmids, the attL product plasmid (Figure 10C) (which may correspond to the Vector Donor DNA or a new Byproduct DNA) that contained the ampicillin resistance and ccdB genes, and the attRproduct plasmid (Figure 1 OD) (which may also correspond to the Insert Donor DNA or a new Product DNA) that contained the kanamycin and tetracycline resistance genes. Competent E. coli cells that received the attL product plasmid, the attP parent plasmid pEZC6701, or recombination intermediates, were killed by the toxic ccdB gene product. Competent E. coli cells -57that received the attB parent plasmid pEZC5601 were killed by the kanamycin selection. Only competent E. coli cells that received the desired attR product Cc plasmid, comprising the kanamycin and tetracycline resistance genes, survived to form colonies. The success of the selection strategy was indicated by the large 00 number of colonies from the reaction that contained both parental plasmids, C< compared to the reaction that contained only one parental plasmid. The reaction C1 mechanism predicted that the desired product plasmid would contain two Clal Srestriction sites, one in the kanamycin resistance gene from the pEZC6701 attP 1 parent plasmid and one in the tetracycline resistance gene from the pEZC5601 attBparent plasmid. The presence of the two sites and the sizes of the fragments resulting from the Clal digestion confirmed the reaction mechanism.

Thus, the present invention relates to reversal of the recombination reaction shown in Figure 1, in which the Product DNA and Byproduct DNA may be combined to produce the Insert Donor DNA and the Vector Donor DNA.

Additionally, the invention provides for subcloning recombinations, in which a Product DNA (produced according to Figure 1) may be combined with a new Vector Donor DNA to produce a new Product DNA (in a different Vector background) and a new Byproduct.

Example 9: Subcloning of Linearized Fragments Plasmid pEZC7102 (Figure 11 the attP parent plasmid (which may correspond to the -Vector -Donor DNA), contained segments attP origin of replication, kanamycin resistance, attP3, chloramphenicol resistance, and the toxic gene ccdB, and in the experiment described here was supercoiled. Plasmid pEZC7501 (Figure 11 the attB parent plasmid (which may correspond to the Insert Donor DNA or the Product DNA), contained the GFP gene cloned between attB 1 and attB3 sites in a vector that comprised the functional domains of (Life Technologies, Inc.). The attP sites contained the P1 and HI domains. Plasmid pEZC7501 was used uncut, or was linearized within the ampicillin resistance gene with Scal, or was cut with XbaI and Sail, to yield a -58- 4c fragment comprising the Sall end, 22bp, the attB I site, the GFP gene, the attB3 Ssite, and 14 bp to the Xbal end: SalI end 22bp attB1 GFP attB3 14bp Xbal end 00 Reactions (8 pl final volume) contained about 40 ng of each DNA, 1 X Int 0 buffer (25 mM Tris HCI pH 7.5, 25 mM Tris HCI pH 8.0, 70 mM KCI, 5 mM spermidine HCI, 0.5 mM EDTA, and 0.25 mg/ml BSA), 12.5% glycerol, 8 ng SIHF, and 43 ng lambda integrase. Reactions were incubated at 25 *C for minutes, then at 70"C for 5 minutes, and then cooled. Duplicate 1 pl aliquots of each reaction were transformed into DH5a UltraMax cells and plated in duplicate on kanamycin agar plates.

The reaction that contained only (supercoiled) pEZC7102 gave an average of 2 colonies (range 1 to The reaction that contained both pEZC7102 and supercoiled pEZC7501 gave an average of 612 colonies (range 482 762). The reaction that contained pEZC7102 and linear (Scal-cut) pEZC7501 gave an average of 360 colonies (range 127-605). The reaction that contained pEZC7102 and the GFP gene on a fragment with attB sites and 22bp and 14 bp beyond the attB sites (pEZC7501 cut with Sail and Xbal) gave an average of 274 colonies (range 243-308).

Miniprep DNAs were prepared from 4 colonies from the pEZC7102 x supercoiled pEZC7510 reaction, and from 10 colonies from the pEZC7102 x pEZC7501/SalI XbaI reaction. All 14 DNAs were rUn uncut on an agarose gel, and the 10 DNAs from the pEZC7102 x pEZC7501/Sail XbaI reaction were cut with a mixture of Ncol and PstI and run on an agarose gel. All the uncut plasmids were about 2.8 kb in size. All ten plasmids cut with the mixture of NcoI and PstI gave fragments of about 700 and 2100 bp.

The results are presented in Table 00 S

IS

59 Table attP attB Colonies Minipreps Uncut Fragment sizes, Parent Parent (average of done product Nco Pst digest 4 plates) plasmid size sc7102 2 sc7102 sc 7501 612 4 2.8 kb sc7102 7501/ScaI 360 sc7102 7501/Sall 274 10 2.8 kb ca. 2100 bp, 700 bp L+ al I Analysis: It was expected that the integrative reaction between the attB sites on plasmid pEZC7501 and the attP sites on plasmid pEZC7102 would produce the attL product plasmid (Figure 11C) (corresponding to the Insert Donor DNA) containing the GFP segment from pEZC7501, and the kanamycin origin segment from pEZC7102. The presence of the toxic gene ccdB on the attP parent plasmid pEZC7102 (corresponding to the Byproduct DNA) was predicted to kill all the cells that received this plasmid. The large increase in the number of colonies when pEZC7501 was present indicated that the desired reaction was occurring, and that the efficiency of the reaction was adequate even if the attB parent plasmid (corresponding to the Product DNA) was linear (Scal cut), or if the attB sites and the GFP gene were present an a fragment that contained little additional sequence beyond the attB sites.

These results show that linear fragments can be suitably subcloned into a different vector by the method of the invention.

Example10: Cloning Long PCR Fragments A PCR product was designed to have an attB0 (wild type) site at one end and a loxP site at the other end. The rationale was that the attPO x attBO reaction would go well with the attB0 molecule (the PCR product) linear. (since it involves a normal lambda integration reaction), and that the 1 oxP x 1 oxP excision from the cointegrate would also be efficient (the unimolecular excision reaction is efficient, Sthe bimolecular integration reaction is inefficient with Cre).

cThe sequence of the attB-containing PCR primer was 5'-TCC GTT GAA GCC TGC TTT TTT ATA CTA ACT TGA GCG AAG CCT CGG GGT CAG 0 0 5 CAT AAG G-3' (SEQ ID NO:3 The sequence of the loxP primer was n ATA ACT TCG TAT AGC ATA CAT TAT ACG AAG TTA TTG CCC CTT O GGT GAC ATA CTC G-3' (SEQ ID NO:32). These primers amplify a part of the human myosin heavy chain. Polymerase chain reactions were performed using O ELONGASE T and K562 human DNA as template. Polymerase chain reactions were performed as follows. Reactions (50 microliters) contained 100 ng K562 human DNA (Life Technologies, Inc.), 0.2 jM of each primer, and 0.2 mM of each dNTP, in ELONGASE T SuperMix (Life Technologies, Inc.). Reactions in thin wall tubes under mineral oil were denatured at 94"C for 1 minute, then cycled times at 94"C for 30 seconds, 65"C for 30 seconds, and 68"C for 8 minutes 30 seconds. Following thermal cycling, reactions were maintained at 4*C. The 5.2 kb PCR product (Figure 9A) was gel purified.

Plasmid pEZC1202 (Figure 9B) contained a wild-type attP site, a chloramphenicol resistance gene, a gene encoding the tet repressor, a wild-type loxP site, an origin of replication, and a tet operator/promoter transcribing and controlling the transcription of a kanamycin resistance gene. This plasmid conferred chloramphenicol resistance but not kanamycin resistance, because the tet repressor made by one element of the plasmid kept the kanamycin resistance gene turned off. The pEZC1202 DNA used in this experiment was a miniprep whose concentration was estimated to be about 50 ng per microliter.

About 40 ng of the gel purified 5.2 kb PCR product were included in a Vl reaction that contained about 50 ng of supercoiled pEZC1202, 0.2 units of Cre recombinase, 3.6 ng IHF, and 11 ng of Int in 50 mM Tris HCI pH about 7.8, 16 mM NaCI, 35 mM KCI, 0.25 mM EDTA, 0.375 mg/ml bovine serum albumin.

A second reaction that did not contain the PCR product was included as a control.

After incubating at 27* for 45 min and then 70" for 5 minutes, 1 pl aliquots were transformed into DHSa UltraMax competent E. coli cells (Life Technologies, Inc.). One fifth of each expression mix was plated on agar that contaihed -61 100 g/ml kanamycin and the plates were incubated overnight at 37* C. The reaction that contained the PCR product gave 34 colonies, while the reaction that c lacked the PCR product gave 31 colonies. After the plates sat at room temperature for four days, 26 additional small colonies were seen on the plate 00 from the positive PCR product) reaction, while only one additional small colony was seen on the plate from the negative (no PCR product) reaction.

STwelve of the 26 small colonies were grown overnight in rich broth (CircleGrow) that contained 25 pg/ml kanamycin, and miniprep DNAs were C prepared from these cultures. All twelve miniprep plasmids were about 8 kb in size, which corresponded to the size expected for replacement of the choramphenicol resistance and tet repressor genes in pEZC1202 with the 5.2 kb PCR product. The predicted recombinant product is shown in Figure 9C. Two of these plasmids were cut with Aval (8 sites predicted) and BamHI (4 sites predicted). All the predicted Aval fragments appeared to be present. One of the BamH I sites predicted in the PCR product (the one closest to the attB end) was absent from both minipreps, but the other BamHI fragments were consistent with the expected structure of the cloned 5.2 kb PCR product.

Analysis: The replacement of the choramphenicol resistance and tet repressor genes in pEZC1202 with the 5.2 kb PCR product (part of the human myosin heavy chain) conferred a moderate resistance of the host E. coli cells to kanamycin, but this resistance was not sufficient to allow colonies to appear after overnight incubation. Thus, colonies containing the desired recombination product grew on kanamycin plates, but were not seen after overnight incubation, but only after an additional room temperature incubation. Of the 12 Aval and BamHI restriction sites predicted from the nucleotide sequence, 11 were confirmed experimentally. -Thus the following three observations support the conclusion that the 5.2 kb PCR product was cloned by recombination: small, slow growing colonies appeared only on the plate from the reaction that contained the PCR product; the miniprep plasmids from these colonies were the expected size; and diagnostic restriction cuts gave the expected fragments (with the one above noted exception).

62 Example 11: Cloning of PCR Fragments Three sets of pairs of PCR primers (Table 9) were designed to amplify an 830 bp sequence within plasmid pEZC75O1 (Figure I I B) comprising: attB I1-- 00 GFP--attB3, with or without additional nucleotides at the outer ends of the 25 bp attB I and attB3 recombination sites. (Here "outer" refers to the end of the attB sequence that is not adjacent to the GFP gene sequence.) Primer set A added 17 nucleotides upstream of attBlI and 15 nucleotides downstream of attB3; primer set B added 5 and 8 nucleotides to attB 1 and attB3, respectively; and primer set C added no additional nucleotides to either attB recombination sequence.

The primer sequences are provided in Table 11: Table 11 upper GFP A 5'-TCA CTA GTC GGC GGC CCA CA (SEQ ID NO:33) lower GFP A 5'-GAG CGG CCC CCG CGG ACC AC (SEQ ID NO:34) upper GFP B 5'-GGC CCA CAA (iTt TGT ACA AAA (SEQ ID lowe.,r GFP B 5'-CCC CGC GGA CCA CIT TGT AC (SEQ ID NO:36) upper GFP C 5 '-ACA AGT TITG TAC AAA AAA (iCA (SEQ ID NO:37) Inve T~Fl S'-ACT' AflT TrGTAr AAfl AAA GT (SRO ID NO1R9) PCR Reactions Primer sets A and C were used first wtth the followingp PCR reactions, in VI, in duplicate. Final concentrations were: mM TfisHCl, pH 8.4 mM KCI 0.2mM of all four deoxynucleotide triphosphates (dNTPs) 400 ng/ml pEZC7501 supercoiled DNA template pLM of each primer Recombinant Taq DNA polymerase (BRL-GIBCO) 100 U/mI A duplicate set of the above reactions contained I M betaine.

-63- The reactions were first heated for to 94*C for then cycled 25 times at 94*C for 45", 55C for 30", and 72*C for 1'.

c The size of the PCR reaction products was analyzed on a 1% agarose gel in TAE buffer containing 0.5 gg/ml ethidium bromide. All reactions yielded 00 products of the expected size, thus duplicate reactions were pooled. As the C corresponding reactions with and without betaine were not significantly different, these also were pooled, giving a final pooled volume for reactions with primer sets A and C of 200 pl each.

Primer set B was then used with identical reactions to those above performed, except that the reaction volumes were increased to 100tl. After duplicate reactions and reactions plus and minus betaine were pooled, the final volume of the reactions with primer set B was 400 pl.

The three pooled primer reaction products were stored at -20"C for 4 weeks.

PCR Product Purification Each of the three pooled PCR products was extracted once with an equal volume of a mixture ofTris-buffered phenol, isoamyl alcohol and chloroform. The aqueous supernatant then was extracted twice with an equal volume ofisobutanol, and the aqueous layer ethanol precipitated with two volumes of ethanol, 0.1 M sodium acetate, pH 6.2. The ethanol precipitates were recovered by centrifugation at 13,000 rpm for 10' at room temperature, and the supernatant discarded. The dried pellets were dissolved in TE: 100l for reactions prepared with primer sets A and C; 200 pl for the reactions with primer set B.

To remove PCR primers and extraneous small PCR products, the PCR products were precipitated with polyethylene glycol (PEG) by adding /2 volume of a solution of 30% PEG 8000 (Sigma), 30 mM MgCI,. mixing well, and centrifuging at 13,000 rpm for 10', all at room temperature. The supernatant was discarded, and the pellets were dissolved in their previous volume ofTE buffer.

1 u 1 aliquots of each of the three PCR products were checked on a 1% agarose gel to quantitate the recovery, which was estimated to be over 90%. The concentration of each PCR product was adjusted using TE to 40 ng/pl.

-64f Recombination Reaction with the PCR Products of Primer sets A, B, and C Five 8 pl reactions were assembled in 1 X Integrase buffer (25 mM Tris

C

c HCI pH 7.5, 25 mM Tris HCI pH 8.0, 80 mM KCI, 5 mM spermidine, 0.5 mM EDTA, 0.25 mg/ml BSA) containing: 40 ng of pEZC7102 DNA, 2 pl of 00 recombinase mixture (8 ng/pl- IHF, 22 ng/pl Int in 1 X Int Buffer, 50% glycerol) C' the reactions differed by the addition of either the PCR product of primer set A C' (reaction primer set B (reaction or primer set C (reaction the addition Sof no PCR product (reaction or the addition of 40 ng of pEZC7501 SC C' (supercoiled) DNA(reaction E) as a positive control. All reactions were performed in duplicate.

The reactions were incubated for 45' at 25"C, for 10' at 70*C, then held at 0-5"C. 2 pi aliquots of each reaction were transformed into Max Efficiency in a 50 pL transformation reaction, and following expression in medium, 1/5 (100 pl) and 4/5 (400 pl) of the reactions were plated on kanamycincontaining (50 pg/ml) LB culture plates. The results of the duplicate reactions are shown in Table 12.

Table 12 Transfection No. Colonies A 100 pl 464, 668 A 400 pl >1000, >1300 B 100 pl 980, 1292 B 400 pl >3000, >3000 C 100 pl 2, 8 C 400 pl 13,20 D 100 pl 0, 0 D 400 pl 0, 0 F 100 tt 56 Analysis of the colonies obtained Miniprep DNA was prepared from 8 colonies of each of the Recombination reactions with primer sets A, B. or C. The supercoiled DNA obtained was analyzed on a 1% agarose gel: all eight of colonies from the recombination products of primer sets A and B were of the predicted size (2791 Sbp) for correct recombination between the PCR products (about 824 bp) and the c attBl-ori-kan'-attB3 sequence donated by pEZC 7102 (1967 bp). Three of the eight reaction products of primer set C were of the predicted size; the other five 00 5 all were slightly larger than 4 kb.

t Further analysis of the reaction products was performed using two O different restriction enzymes, Aval and PvuII, each of which cleaves twice (but at different locations) within the predicted recombinant product, once within the PCR product sequence and once within the sequence contributed by pEZC7102.

Both of these enzymes should cleave the intact pEZC7102 recombination partner plasmid at two sites, to give fragments easily distinguished from those of the expected recombination products.

The two restriction enzyme digests yielded the expected sizes offragments (2373 and 430 bp for Aval; 2473 and 330 bp for Pvull) from the colonies generated from the recombination reactions with primer sets A and B, as well as for the three colonies from primer set C that displayed the expected size of supercoiled DNA. For the other five colonies from primer set C that yielded larger SC DNA, however, the Pvull digest revealed fragments of approximate size to those predicted from a digestion of pEZC7102, Whereas the Aval digest revealed only a single fragment, approximately the size of linearized pEZC7102 (4161 bp).

Analysis These results indicate that PCR products generated from templates containing a gene flanked by attB sites can serve as efficient substrates for the reverse recombination reaction. The addition of even short DNA sequences to the ends of the attBl and attB3 sites or core regions 5bp and 8bp, respectively, in primer set B) stimulated this reaction by 100 fold or more. Surprisingly, reverse recombination reactions with PCR products containing additional sequence beyond the attB sites appeared in these reactions to be more efficient recombination partners than the supercoiled positive control plasmid, pEZC7501.

-66- All the recombination products were generated faithfully. A low level of background colonies emerged from the relatively inefficient recombination C reactions with primer set C, which lacked additional sequence beyond the 25 bp attB sites. This background appeared to be due to a largely intact pEZC7102 (which encodes kanamycin resistance) lacking an active ccdB death gene, allowing it to survive. Consistent with this interpretation is that one of the two restriction sites for Aval in this plasmid was also altered. One of the Aval sites is present within the ccdB region of pEZC7102. It is likely therefore that the alteration of this site was secondary to mutational inactivation of the ccdB gene.

Example 12: Further Cloning of PCR Fragments Two sets of 6 primers for preparing PCR products from the plasmid pBR322 as template were used. One set (Table 1 anneals to sequences flanking the TetR gene, including the TetR promoter. The other set (Table 14) anneals to sequences flanking the AmpR gene, including its promoter. The "tet" and "amp" primers used contain no attB sequences, only sequences inherent to the pBR322 plasmid; the "attB" primers contain, in addition to the pBR322 sequences, the bp of attB 1 or attB3 sequences; the "attB+4" primers contain the pBR322specific sequences, plus the 25 bp attB 1 or attB3 sequences, each with four Gs at the outer end. (Here "outer" refers to the end of the attB sequence not adjacent to the template-specific primer sequence.) Preparation of pBR322 template To improve the efficiency of the PCR reaction, the supercoiled pBR322 DNA was linearized by incubating 3.5 pg of Suerpcoiled (SC) pBR322 DNA in a 200 gl reaction with 15 units of the restriction enzyme NdeI and final concentration of 50mM Tris-HCI, pH8.0, 10mM MgCl,, and 50mM NaCI, for one hour at 37C.

The digested pBR322 DNA was extracted once with phenol, isoamyl alcohol, and chloroform, extracted twice with isobutanol, and precipitated by adding two volumes of ethanol plus 0.15M sodium acetate. The precipitate was 67 washed once with 100% ethanol, dried, then dissolved in TE buffer. Recovery of DNA, quantitated on a I agarose gel in TAE buffer, 0.5 jig/ml ethdium bromide, was estimated as greater than Table 13 tet Primer Primer Sequence SE

NO:

tet-L AAT TCT CAT GIT TGA GAG CTT ATC 48 tet-R CGA TOG ATA TOT TCT 0CC AAG 49 attBI-tetL ACAAG TTHGTA CAAAAA AGGA GGCT- AAT TCT CAT GTT TGA CAG CTT ATC attB3-tetR ACCAC TTTGTA CAAGAA AGCT GGGT- 51 CGA TGG ATA TOT TCT 0CC AAG attBl1+4-tetL 0000 ACAAG 1TTTGTA CAAAAA AGGA- 52 GOCT AAT TCT CAT GTT TGA GAG CTT-

ATC

attB3+4-tetR GGGG ACCAC ITTGTA CAAGAA AGCT- 53 GGGT CGA TGG ATA TGT TCT GCC AAG Table 14 amp Primer Primer Sequence SEIU amp-L AAT ACA HGC AAA TAT GTA TGC GC 54 amp-R HTA CCA ATG CUT AAT-CAG TGA G attB]-anipL ACAAG~MGTA CAAAAA AGCA GGCT 56 AAT ACA UTC AAA TAT OTA TCC GC artB3-ampR ACCAC TOTA CAAGAA AGCT GGGT- 57 HTA CCA ATG CUT AAT CAG TGA 0 arB 1+4- G000 ACAAG TITTOTA CAAAAA AGGA- 58 anipL GGCT AAT AGA TTC AAA TAT OTA TCCanB3+4- 0000 ACCAC TGTA CAAGAA AOCT- 59 ampR GT TTA CCA ATG CUT AAT GQAG TGAG I__ -68i PCR amplification of tet and amp gene sequences Six PCR reactions were performed, in 100pl, consisting of 20mM Trisc HCI, pH 8.4, 50mM KC1, 1.5 mM MgCl, 0.2mM dNTPs, 2 ng linearized pBR322, 2.5 units of Taq DNA polymerase (GIBCO-BRL), and 0.5 uM of each 00 pair of PCR primers listed in Tables 5 and 6. The reactions were first heated to 94 C for then subjected to 25 cycles of 94*C for 45 seconds, 55C for seconds, and 72"C for 1 minute. Based on 1% agarose gel analysis, all the reactions generated products of the expected size, in reasonable yields.

Purification ofPCRproducts The products from duplicate reactions were pooled; extracted with an equal volume of phenol, isoamyl alcohol, and chloroform; extracted twice with an equal volume of isobutanol; and precipitated with two volumes of ethanol, as above. The six precipitates were washed once with 100% ethanol, dried and dissolved in 100 pl TE. 1 pl aliquots were taken for gel analysis of the product before PEG precipitation.

To each tube was added 50 ll of 30% PEG 8000, 30 mM MgCl,. The solution was mixed well and centrifuged at 13,000 rpm for 10', at room temperature. The supernatant was carefully removed, and the precipitate dissolved in 100 pl TE. Recovery was quantitated on a 1% agarose and estimated to be over 90%. The gel analysis also revealed that nucleic acid products smaller than about 300 nucleotides had been effectively removed by the PEG precipitation step.

Recombination Reactions Seven recombination reactions were performed, each in a total volume of 8 Vl, containing 1 X integrase buffer, 40 ng pEZC7102 (Figure 11A), and 2 pl recombinase mixture (see above, Example 11). Each of the reactions also contained approximately 40 ng of one of the six above PCR products or, as a positive control, 40 ng of pEZC7501 (Figure 11B). The amp and tet PCR products with attB sites at their termini are shown in Figures 12A and 12B. The -69- 00 In 1t 0~ 0 (Nq reactions were incubated at 25 C for 45', at 70° C for 10', then held at 0-5 C for 1-2 hours until used to transform E. coli.

E. coli transformation with recombination reaction products I !l of each of the recombination reactions was transformed into Max Efficiency DHSa in a 50 l transformation reaction, and following expression in SOC medium, 1/5 (100 il) and 4/5 (400 pl) of each reaction were plated on culture plates containing 50 pg/ml kanamycin. The plates were incubated overnight and colonies were counted. The number of colonies obtained from each set of duplicate reactions are displayed in Table Table No. Colonies Recombination Reactions tet 100 (100 pI) 6, tet 400 (400 pl) 27,32 attB-tet 100 9, 6 attB-tet 400 27,36 attB+4-tet 100 824, 1064 attB+4-tet 400 >2000, >4000 amp 100 7, 13 amp 400 59 attB-amp 100 18, 22 attB-amp 400 66, 66 attB+4-amp 100 3020, 3540 attB+4-amp 400 >5000, >5000 pEZC7501 100 320, 394 pEZC7501 400 1188, 1400 Analysis of the colonies obtained As a rapid phenotypic screen, 10 of the colonies from the tet EZC reactions and 33 of the colonies from the attB+4-tet EZC reactions were streaked onto an LB culture plate containing tetracycline (15 pg/ml). As a control for the potency of the tetracycline, 3 colonies of pUC 19-transformed cells, lacking a TetR gene, were also streaked onto the plate. All colonies from the attB+4-tet EZC Sreactions grew well; colonies from the let EZC reactions grew only very slightly, and the pUC19 colonies grew not at all.

C Analogous results were obtained by streaking colonies from the amp PCR reactions on culture plates containing ampicillin (100 ug/ml All 21 colonies 00 generated from the attB+4-amp recombination reactions grew well, whereas only one of 13 colonies from the attB-amp reactions grew in the presence of ampicillin.

No growth was seen with any of the 15 colonies from the recombination reaction with amp PCR products.

STo characterize plasmid DNA, eight colonies generated from the six EZC reactions with PCR products were picked into LB broth containing 50 pg/ml kanamycin and grown overnight at 37 C. Miniprep DNA was prepared from 0.9 ml of each culture, and the size of the supercoiled DNA was analyzed on a 1% agarose gel in TAE buffer containing 0.5 gg/ml ethidium bromide. The results are displayed in Table 16. The predicted structures of the recombination products are shown in Figure 12C and 12D.

Table 16 Recombination DNA Predicted Size (bp) Number with Reactions Predicted Size tet SC 3386 0/8 (supercoiled) attB-tet SC 3386 1/8 attB+4-tet SC 3386 7/7 Aval+Bam 485, 2901 3/3 amp SC 2931 0/8 attB-amp SC 2931 3/8 attB+4-amp SC 2931 8/8 Pst 429, 2502 3/3 -71 SAnalysis These results, based on t he amplification of two different gene sequences, tet and amp, within the plasmid pBR322, clearly demonstrate that PCR products Sgenerated using primers containing the 25 bp attB 1 and attB3 recombination sequence serve as highly efficient substrates for the recombination reaction.

00 Addition of a short sequence to the outside of each 25 bp attB site stimulates the In recombination reaction by over 100 fold, as also observed in the experiments of Example 11. Also similar to Example 11, the efficiency of the recombination reactions using linear PCR products with attB sites exceeded the efficiency obtained with the positive control SC DNA plasmid, pEZC7501.

Further, a high percentage of the reaction products are as predicted, since all 33 colonies tested from the attB+4-tet reactions displayed functional tetracycline resistance, and all 21 of the colonies from the attB+4-amp reactions displayed ampicillin resistance. All 16 of the miniprep DNAs, examined from the recombination reactions of either attB+4-tet or attB+4-amp PCR products with pEZC7102, generated supercoiled DNA and restriction digest fragments of the correct sizes.

Having now fully described the present invention in some detail by way of illustration and example for purposes of clarity of understanding, it will be obvious to one of ordinary skill in the art that the same can be performed by modifying or changing the invention within a wide and equivalent range of conditions, formulations and other parameters without affecting the scope of the invention or any specific embodiment thereof, and that such modifications or changes are intended to be encompassed within the scope of the appended claims.

All publications, patents and patent applications mentioned in this specification are indicative of the level of skill of those skilled in the art to which this invention pertains, and are herein incorporated by reference to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference.

Where the terms "comprise", "comprises", "comprised" or "comprising" are used in this specification, they are to be interpreted as specifying the presence of the stated features, integers, steps or components referred to, but not to preclude the 30/05,07.at 16620.specipgs. 12 72 presence or addition of one or more other feature, integer, step, component or group thereof.

00 30105/07.ai 6620.specipgs. 12

Claims (22)

1. A method for producing a ligated nucleic acid molecule which comprises ligating a first nucleic acid molecule comprising a recombination site to a second nucleic acid molecule, the method comprising contacting the first nucleic 00 acid molecule and the second nucleic acid molecules in the presence of a t topoisomerase under conditions which allow for ligation of the first nucleic acid molecule to the second nucleic acid molecule.

2. A method for preparing a ligated nucleic acid molecule which contains at least one recombination site, the method comprising: generating a first nucleic acid molecule by polymerase chain reaction, contacting the first nucleic acid molecule with a topoisomerase and a second nucleic acid molecule which contains a recombination site and under conditions which allow for ligation of the first nucleic acid molecule to the second nucleic acid molecule to form said ligated nucleic acid molecule.

3. A method for preparing a ligated nucleic acid molecule which contains at least one recombination site, the method comprising: mixing a first nucleic acid molecule with one or more adapters comprising one or more recombination sites in the presence of a topoisomerase, wherein the adapter is a second nucleic acid molecule; and incubating the mixture under conditions sufficient to add one or more of the adapters to one or more termini of the first nucleic acid molecule to produce the ligated nucleic acid molecule.

4. The method of any one of claims 1 to 3, wherein the first nucleic acid molecule is blunt ended. The method of claim 1, wherein the first nucleic acid molecule also contains a selectable marker or an origin of replication.

6. The method of any one of claims I to 3, wherein the second nucleic acid molecule comprises a selectable marker or an origin of replication.

7. The method of any one of claims 1 to 6, wherein the recombination site(s) are selected from the group consisting of lox sites, attL sites, attR sites, attB 1 30/05107.a I 6620-specipgs.73 -74- S(SEQ ID NO:6), attB2 (SEQ ID NO:7), attB3 (SEQ ID NO:8), attPI (SEQ ID NO: 15) and attP2 (SEQ ID NO: 16).

8. A composition comprising: an isolated nucleic acid molecule comprising one or more recombination sites; and 00 one or more topoisomerases associated with said isolated nucleic acid Vt molecule, wherein one of said one or more recombination sites are selected from the (N O group consisting of lox sites, attL sites, attR sites, attB 1 (SEQ ID NO:6), attB2 (SEQ r- ID NO:7), attB3 (SEQ ID NO:8), attP1 (SEQ ID NO:15) and attP2 (SEQ ID NO:16).

9. A composition comprising: an isolated nucleic acid molecule comprising two or more recombination sites; and one or more topoisomerases associated with said isolated nucleic acid molecule, wherein said two or more recombination sites do not recombine with each other. A composition, outside of a host cell comprising an isolated nucleic acid molecule having one or more recombination sites, said composition further comprising one or more topoisomerases.

11. A composition, outside of a host cell, comprising a nucleic acid molecule which is a product of a polymerase chain reaction, and an isolated nucleic acid molecule having one or more recombination sites, said composition further comprising one or more topoisomerases.

12. An isolated nucleic acid molecule comprising: one or more recombination sites; and one or more topoisomerase recognition sites, wherein one said one or more recombination sites are selected from the group consisting of lox sites, attL sites, attR sites, attBl (SEQ ID NO:6), attB2 (SEQ ID NO:7), attB3 (SEQ ID NO:8), attP1 (SEQ ID NO:15) and attP2 (SEQ ID NO:16).

13. An isolated nucleic acid molecule comprising: two or more recombination sites that do not recombine with each other; and 31/05/07,at 16620.specipgs.74 one or more topoisomerase recognition sites.

14. The composiiton of any one of claims 8 to 11 or the isolated nucleic acid molecule of claim 12 or claim 13, wherein said nucleic acid molecule is circular. The composition of any one of claims 8 to 11 and 14, or the isolated nucleic acid molecule of claim 12, claim 13 or claim 14, wherein the nucleic acid 00 molecule further comprises an origin of replication. n 16. The composition or isolated nucleic acid molecule of claim 0 wherein the nucleic acid molecule further comprises a selectable marker.

17. The composition or isolated nucleic acid molecule of claim 16, 0 10 wherein the selectable marker is an antibiotic resistance gene.

18. The composition of any one of claims 8 to 11 and 14 to 17 or the nucleic acid molecule of claim 12, 13 or 14 to 17, wherein the recombination site is a lox site.

19. The composition or nucleic acid molecule of claim 18, wherein said lox sites are loxP sites. The composition or nucleic acid molecule of claim 19, wherein said loxP sites are selected from the group consisting of loxP and loxP511.

21. The composition of any one of claims 8 to 11 and 14 to 17 or the nucleic acid molecule of claim 12 or claim 13, wherein the recombination sites are att sites.

22. The composition or the nucleic acid of claim 21, wherein said att sites are attL sites or attR sites selected from the group consisting of attL1 (SEQ ID NO:12), attL2 (SEQ ID NO:13) and attL3 (SEQ ID NO:14) and wherein said attR sites are selected from the group consisting of attR (SEQ ID NO:9), attR2 (SEQ ID NO: 10) and attR3 (SEQ ID NO:11).

23. The composition of claim 8 or the isolated nucleic acid molecule of claim 12 comprising two or more recombination sites.

24. The isolated nucleic acid molecule of claim 23, wherein at least one of said two or more recombination sites flanks each of said topoisomerase recognition sites. The composition of any one of claims 8 to 11, or the isolated nucleic acid of claim 12 or claim 13, wherein said topoisomerase is a type I topoisomerase. 30/05/07.a 16620.specipgs. 7 -76-

26. The isolated nucleic acid molecule of claim 12 or claim 13, wherein said topoisomerase recognition site is recognized and bound by a type I topoisomerase.

27. A vector comprising the isolated nucleic molecule of claim 12 or claim 13. 00 28. A host cell comprising the vector of claim 27. t) 29. A kit comprising the isolated nucleic acid molecule of claim 12 or O claim 13. (N The kit of claim 29, further comprising one or more components 0 10 selected from the group consisting of one or more topoisomerases, one or more recombination proteins, one or more vectors, one or more polypeptides having polymerase activity, and one or more host cells.

31. The kit of claim 29, further comprising one or more components selected from the group consisting of one or more topoisomerases, one or more recombination proteins, one or more vectors, one or more polypeptides having polymerase activity, and one or more host cells. DATED: 31 (VLq a-S INVITROGEN CORPORATION By their Patent Attorneys: A CALLINAN LAWRIE W 30/05/07.ar 16620.specipgs.76