METHOD FOR AMPLIFICATION OF LIGATION REACTIONS

FIELD OF INVENTION

The present invention relates to methods for preparing libraries of DNA molecules. In particular, the present invention relates to methods for preparing libraries of DNA molecules by amplifying a library of nucleic acids and transforming said library into host cells. The invention employs rolling circle replication using a strand displacing polymerase, preferably phi29 polymerase.

BACKGROUND TO THE INVENTION

Introduction

The assembly of self-replicating, circular DNA molecules and their transformation into bacteria, as first reported by Berg and co-workers in 1972 (1), lies at heart of recombinant DNA technology. However the assembly (by ligation) of circular DNA molecules from cohesive linear fragments is inefficient especially with multiple fragments (2), as is the transformation of bacteria (3). Less than 1 in 10 ² molecules of circular supercoiled plasmid can be successfully transformed (4), and for linear molecules the efficiency is even lower (less than 1 in 10 ^s molecules) (5). To obtain large numbers of recombinant clones, it is therefore usually necessary to use large amounts of DNA for the ligations, and to undertake multiple transformations.

The generation of large numbers of recombinant clones is particularly useful for generating libraries of host cells expressing a library or repertoire of polypeptides. This usually requires the generation of a large recombinant DNA library of variants for cloning into a phage or plasmid vector, and the transformation of a host organism for expression of the variant proteins. The expression library can then be screened for polypeptides of interest.

However library size is often limited by the low yields of circular DNA and the poor transformation efficiencies of linear DNA. Moreover, molecular diversity of the recombinant DNA library can be lost.

Accordingly, there is a need for improved methods of generating libraries of host cells.

Rolling-circle amplification (RCA), also known as rolling-circle replication, is a procedure in which small single-stranded DNA (ssDNA) circles serve as templates for DNA polymerases (or sometimes RNA polymerases) generating numerous concatemerized copies of the circle. When a single primer complementary to the circle is employed, the accumulation of product proceeds in most cases linearly over time. An exponential (geometric) amplification via a so- called hyperbranched RCA (HRCA), also known as ramification or cascade RCA, is achieved by using second or further primers with a sequence identical to a part of the DNA circle. The RCA technology is promising for molecular diagnostic and pharmacogenomic use because of its simplicity, high sensitivity, large multiplex potential, immunity to false positives/cross- contamination and easy compatibility with other detection/imaging techniques.

DNA replication by strand displacement is an isothermic process, unlike PCR ₅ and has applications in whole genome amplification techniques. A number of polymerases are available for strand displacement techniques. Phi29 polymerase specifically amplifies DNA circles (6-8) at the expense of short linear DNA molecules. Phi29 polymerase has the unique property of catalyzing strand displacement synthesis with high processivity (9) and low error rate (10), and unlike PCR introduces little bias into the amplified population, as shown for whole genome amplifications (11). Several applications of Phi29 polymerase for the amplification of DNA molecules have been reported (6,12-17). These include the amplification of DNA circles such as plasmid and phage genomes (6), as well as extended linear molecules such as human chromosomes (17). Most approaches are based on the work of Lasken and co-workers (6), and rely on random priming using high concentrations of short synthetic oligonucleotides. The amplification is based on a rolling-circle mechanism and leads to linear concatamers containing multiple template repeats (6). This process is facilitated by the strongly strand displacing properties of the polymerase and high primer

concentrations. The tandem repeats are subsequently used as templates for further amplification, thereby producing extensive amounts of linear DNA suitable for sequencing and hybridization.

However, amplification of DNA does not necessarily lead to improved libraries, since the inefficiencies associated with transformation and the loss of diversity observed remain, even in PCR-amplifϊed libraries. A need remains, therefore, for a method for making libraries which does not sacrifice molecular diversity and which improves transformation efficiency.

SUMMARY OF THE INVENTION

The present invention relates to an improved method for generating libraries of host cells through amplifying circular DNA prior to transformation of host cells with recircularised amplified DNA.

Accordingly in a first aspect of the invention there is provided a method for preparing an amplified library of transformed host cells said method comprising the steps of: a) taking a library of closed circular DNA molecules; b) amplifying said library of closed circular DNA molecules through a rolling circle mechanism in the presence of a strand displacement polymerase to generate a library of linear concatamers; c) cleaving said linear concatamers to generate a library of linear DNA molecules; d) re-circularising said library of linear DNA molecules to generate an amplified library of circular DNA molecules; e) transforming said amplified library of circular DNA molecules into host cells.

It has been found that recircularising amplified DNA and using such DNA in a transformation protocol greatly improves transformation efficiency, and leads to the maintenance of molecular diversity in the library.

The strand displacement polymerase used in the invention has high strand displacement activity, as well as a proofreading activity. Preferred polymerases are phi29 polymerase and a combination of Bst polymerase and the T4 gene 32 protein. Advantageously, the polymerase is Phi29 polymerase.

The library of closed circular DNA molecules subjected to amplification in the method of the invention may be a subset of a library of DNA molecules prepared, such as in bacteriophage, from natural or synthetic sources. The library of closed circular DNA molecules may moreover be or be derived from an entire library. Suitable such libraries include a library encoding polypeptides, including a library encoding antibody domains. A number of libraries encoding antibody domains including libraries encoding antibody variable domains will be familiar to those skilled in the art and include, for example, the libraries referred to in the detailed description and examples sections herein.

The library of closed circular DNA molecules advantageously comprises at least 10 ² different members; preferably at least 10 ³ , 10 ⁴ , 10 ⁵ , 10 ⁶ , 10 ⁷ , 10 ⁸ , 10 ⁹ or 10 ¹⁰ different members.

In one embodiment, the circular DNA molecules comprise two or more segments of DNA. Suitably, one of said segments comprises an origin of replication. In one embodiment the origin of replication is the origin of replication sequence for the host cell of interest. Suitably, the origin of replication is E.coli origin of replication. At least one of the other said segments may be a polypeptide reading frame.

As used herein, a "segment" of DNA may be any DNA sequence, but preferably refers to a DNA fragment comprising a desired sequence and advantageously encoding a desired polypeptide. For example, a segment of DNA encoding one or more CDRs may be combined with another segment of DNA encoding other CDRs or framework regions, to construct a DNA molecule encoding an antibody variable domain.

Alternatively, segments encoding heavy chain variable domains (VH) and light chain variable domains (VL) may be combined to construct a DNA molecule encoding a

monoclonal antibody. Suitable monoclonal antibody formats include IgG, IgE ₅ IgM, IgA, scFv and Fab fragments.

Other DNA segments encoding promoters and enhancers may also be included.

In another embodiment, the library of closed circular DNA molecules may be obtained by circularisation of linear DNAs. In this embodiment, the method further comprises circularisation of linear DNAs prior to step a). Circularisation may be achieved by ligating in the presence of a ligase. Suitable ligases include T4 ligase.

In a further embodiment, the amplification step b) includes adding hexamer primers to the template circular DNA and incubating in the presence of the strand displacement polymerase. Suitable incubation conditions include those conditions allowing circular DNA to be amplified through a rolling circle mechanism such as those conditions described herein. In one embodiment, amplification is achieved by incubating isothermally at 3O ⁰ C overnight. The hexamer primers are for example random hexamer primers.

In one embodiment, primers are added at regular intervals during the amplification in order to maintain primer concentrations. Suitably the primers do not have phosphothioates incorporated.

Preferably the DNA obtained through amplification is non-methylated.

The resulting concatamers are cut, for example with a restriction endonuclease. Suitable restriction endonucleases are described herein.

Re-circularisation is preferably through self-ligation by incubating at low DNA concentrations in the presence of a ligase such as T4 ligase. Preferably, dilute DNA concentrations are less than lng/μl.

Transformation of host cells can be, for example, through electroporation.

Advantageously, the amplification by, for example, bacteriophage Phi29 polymerase in accordance with the method of the present invention increases the number of transformants. In a preferred embodiment, template DNA is amplified >10 ³ fold while the number of transformants is increased >10 ⁵ fold.

In addition, the method of the invention advantageously requires only minute quantities of DNA as a starting template. In one example, described herein, a nanogram-scale ligation of DNA fragments comprising two sub-libraries of variant antibody domains was used and a highly diverse and large combinatorial phage antibody library (>10 ⁹ transformants in E. coli and 10 ⁵ -fold more than without amplification) was amplified. From the amplified library, but not from the smaller un-amplified library, several antibody fragments against a target antigen were isolated. Therefore, the method of the present invention also enables clones to be recovered such that molecular diversity, which would otherwise be lost in the transformation step, is maintained. A further and convenient feature of the method is the option of using PCR amplified inserts and vectors for ligations.

The method of the invention also provides for the generation of large amounts of non- methylated DNA.

Furthermore, advantageously, the method of the present invention gives no detectable bias in the amplification procedure such that constructs with differences in sequences and lengths of DNA are amplified to a similar extent. Molecular diversity otherwise lost in conventional transformation process is maintained.

The method can be applied, for example, to the generation of an antibody repertoire.

In one embodiment, the library produced by the method in accordance with the invention can be used in a method for selecting antibodies with a desired binding affinity.

In one embodiment the amplified library produced by the method in accordance with the invention can be used in a method for molecular evolution, i.e. the generation of genetic diversity followed by the selection of those nucleic acids which result in beneficial characteristics.

In another aspect of the invention there is provided a repertoire of polypeptides produced from the amplified library obtained by the method of the first aspect of the invention.

In a further aspect of the invention there is provided a method for selecting a polypeptide by function comprising: a) transforming host cells with an amplified library produced in accordance with the method of the first aspect of the invention; b) culturing the host cells under conditions under which they express polypeptides encoded by the DNA they contain; c) selecting by function.

For example, the function is binding to a target ligand. In one embodiment, the polypeptide is an antibody.

In another aspect, the invention provides a kit for carrying out the methods of the present invention. Suitably said kit is for preparing a library of host cells and comprises a Phi29 polymerase, hexamer primers, a restriction enzyme and a ligase.

DESCRIPTION OF THE FIGURES

Figure 1

In vitro amplification of ligation reactions. Linear fragments were joined into recombinant, circular units by treatment with DNA ligase. Hexamer primers were annealed and Phi29 polymerase added. This causes extensive amplification of circular species through rolling- circle replication and the formation of extended linear concatamers. The concatamers were cleaved by restriction digestion and re-circularized using DNA ligase.

Figure 2

Amplification of dilutions of a model ligation reaction. Products from ligation reactions of ampicillin resistance gene into vector, were diluted, amplified, and used to transform E. coli, plating in the presence of ampicillin. The number of colonies (before and after amplification, grey and black respectively) was plotted as a function of the dilution of the ligation reaction.

Figure 3

Amplification of mixtures of two model ligation reactions. Products from ligation reactions of ampicillin and chloramphenicol resistance genes into vector were mixed in different ratios, amplified, and used to transform E. coli in the presence of ampicillin or chloramphenicol. The ratio of the number of colonies (ampicillin/chloramphenicol) growing under the two different antibiotic selections (before and after amplification, grey and black respectively) was plotted as a function of the ratios of the original mixtures (ampicillin/chloramphenicol) .

Figure 4

Antibody repertoire by combinatorial ligation. Regions corresponding to CDRs ¹ A and

CDR 3 were PCR-amplified and recombined by ligation into a PCR amplified phagemid vector backbone. The ligation reactions were either directly electro orated or amplified using Phi29 polymerase followed by electroporation into E. coli bacteria.

DETAILED DESCRIPTION OF THE INVENTION

LIBRARIES

The term library refers to a mixture of heterogeneous polypeptides or nucleic acids. The library is composed of members, each of which has a unique polypeptide or nucleic acid sequence. To this extent, library is synonymous with repertoire. Sequence differences between library members are responsible for the diversity present in the library. The library may take the form of a simple mixture of polypeptides or nucleic acids, or may be in the form

organisms or cells, for example bacteria, viruses, animal or plant cells and the like, transformed with a library of nucleic acids, or microencapsulated, for instance as described herein. Preferably, each individual organism or cell or microcapsule contains only one member of the library.

Advantageously, a library is a library of nucleic acids and encodes a repertoire of polypeptides.

Advantageously, the nucleic acids are incorporated into expression vectors, in order to allow expression of the polypeptides encoded by the nucleic acids. In a preferred aspect, therefore, a library may take the form of a population of host organisms, each organism containing one or more copies of an expression vector containing a single member of the library in nucleic acid form which can be expressed to produce its corresponding polypeptide member. Thus, the population of host organisms has the potential to encode a large repertoire of genetically diverse polypeptide variants.

DNA POLYMERASES

DNA polymerases useful in the method of the invention must be capable of displacing, either alone or in combination with a compatible strand displacement factor, a hybridized strand encountered during replication. Such polymerases are referred to herein as strand displacement DNA polymerases.

Strand displacement polymerases are described in detail in US Patent Application publication no. 2004/0180372. Such polymerases include:

Preferred strand displacement DNA polymerases include Bst large fragment DNA polymerase (Exo(-) Bst; Aliotta et al., Genet. Anal. (Netherlands) 12:185-195 (1996)), exo(-)Bca DNA polymerase (Walker and Linn, Clinical Chemistry 42:1604-1608 (1996)), phi29 DNA polymerase (U.S. Pat. Nos. 5,198,543 and 5,001,050), phage M2 DNA polymerase (Matsumoto et al., Gene 84:247 (1989)), phage phiPRDl DNA polymerase (Jung et al., Proc.

Natl. Acad. Sci. USA 84:8287 (1987)), exo(-)VENT DNA polymerase (Kong et al., J. Biol. Chem. 268:1965-1975 (1993)), Klenow fragment of DNA polymerase I (Jacobsen et al., Eur. J. Biochem. 45:623-627 (1974)), T5 DNA polymerase (Chatterjee et al., Gene 97:13-19 (1991)), Sequenase (U.S. Biochemicals), PRDl DNA polymerase (Zhu and Ito, Biochim. Biophys. Acta. 1219:267-276 (1994)), and T4 DNA polymerase holoenzyme (Kaboord and Benkovic, Curr. Biol. 5:149-157 (1995)).

Certain of the above polymerases require the use of a strand displacement factor, such as helicase. Strand displacement factors useful in strand displacement replication include BMRFl polymerase accessory subunit (Tsurumi et al., J. Virology 67(12):7648-7653 (1993)), adenovirus DNA-binding protein (Zijderveld and van der Vliet, J. Virology 68(2): 1158-1164 (1994)), herpes simplex viral protein ICP8 (Boehmer and Lehman, J. Virology 67(2):711-715 (1993); Skaliter and Lehman, Proc. Natl. Acad. Sci. USA 91(22):10665-10669 (1994)); single-stranded DNA binding proteins (SSB; Rigler and Romano, J. Biol. Chem. 270:8910- 8919 (1995)); phage T4 gene 32 protein (Villemain and Giedroc, Biochemistry 35:14395- 14404 (1996); and calf thymus helicase (Siegel et al., J. Biol. Chem. 267:13629-13635 (1992)).

Phi29 polymerase has been found to possess particularly advantageous properties in the context of the present invention and its use is preferred.

POLYPEPTIDES

As used herein, the term "polypeptide" refers to a molecule or molecular construct that comprises a sequence of amino acids. Advantageously, the amino acids are part of a polypeptide domain. A domain, as used herein, is a polypeptide or protein sequence that is capable of folding to a stable three-dimensional structure in isolation. In a preferred embodiment, the polypeptide domain is an antibody domain. Examples of antibody domains include VH or VL domains.

A typical antibody is a multi-subunit protein comprising four polypeptide chains; two "heavy" chains and two "light" chains. The heavy chain has four domains, the light chain has two domains. All of the domains are classified as either variable or constant.

The antigen binding domain of an antibody comprises two separate regions: a heavy chain variable domain (VH) and a light chain variable domain (VL: which can be either V _κ or Vλ).

The antigen-binding site itself is formed by six polypeptide loops: three from the VH domain (Hl, H2 and H3) and three from the V _L domain (Ll, L2 and L3).

The V _H gene is produced by the recombination of three gene segments, V _H , D and JH. In humans, there are approximately 51 functional V _H segments (Cook and Tomlinson (1995) Immunol Today, 16: 237), 25 functional D segments (Corbett et al. (1997) J. MoI. Biol, 268: 69) and 6 functional JH segments (Ravetch et al (1981) Cell, 27: 583), depending on the haplotype. The V _H segment encodes the region of the polypeptide chain which forms the first and second antigen binding loops of the V _H domain (Hl and H2), whilst the V _H , D and JH segments combine to form the third antigen binding loop of the V _H domain (H3).

The V _L gene is produced by the recombination of two gene segments, V _L and JL. In humans, there are approximately 40 functional V _K segments (Schable and Zachau (1993) Biol. Chem. Hoppe-Seyler, 374: 1001), 31 functional Vλ segments (Williams et al. (1996) J. MoI. Biol, 264: 220; Kawasaki et al (1997) Genome Res., 7: 250), 5 functional J _κ segments (Hieter et al (1982) J. Biol. Chem., 257: 1516) and 4 functional Jχ segments (Vasicek and Leder (1990) J. Exp. Med., 172: 609), depending on the haplotype. The V _L segment encodes the region of the polypeptide chain which forms the first and second antigen binding loops of the V _L domain (Ll and L2), whilst the V _L and JL segments combine to form the third antigen binding loop of the V _L domain (L3). Antibodies selected from this primary repertoire are believed to be sufficiently diverse to bind almost all antigens with at least moderate affinity. High affinity antibodies are produced by "affinity maturation" of the rearranged genes, in

which point mutations are generated and selected by the immune system on the basis of improved binding.

A repertoire of polypeptide domains may be encoded in the form of a library of nucleic acids.

Typically, the antibody domains will be provided encoded by a nucleic acid library, which will in most cases require the screening of a large number of nucleic acids encoding variant antibody domains. Libraries of antibody domains may be created in a variety of different ways, including the following.

Repertoires of naturally occurring antibody domain genes may be cloned from genomic DNA or cDNA (Sambrook et al., 1989); for example, phage antibody libraries, made by PCR amplification repertoires of antibody genes from immunised or unimmunised donors have proved very effective sources of functional antibody fragments (Whiter et al., 1994; Hoogenboom, 1997). Libraries of genes encoding antibody domains may also be made by encoding all (see for example Smith, 1985; Parmley and Smith, 1988) or part of genes (see for example Lowman et al., 1991) or pools of genes (see for example Nissim et al., 1994) by a randomised or doped synthetic oligonucleotide. Libraries may also be made by introducing mutations into an antibody domain or pool of antibody domains 'randomly' by a variety of techniques in vivo, including; using 'mutator strains', of bacteria such as E. coli mutD5 (Liao et al., 1986; Yamagishi et al., 1990; Low et al., 1996); and using the antibody hypermutation system of B-lymphocytes (Yelamos et al., 1995). Random mutations can also be introduced both in vivo and in vitro by chemical mutagens, and ionising or UV irradiation (see Friedberg et al., 1995), or incorporation of mutagenic base analogues (Freese, 1959; Zaccolo et al., 1996). 'Random' mutations can also be introduced into antibody domains genes in vitro during polymerisation for example by using error-prone polymerases (Leung et al., 1989).

Further diversification may be introduced by using homologous recombination either in vivo (see Kowalczykowski et al., 1994 or in vitro (Stemmer, 1994a; Stemmer, 1994b)).

Preferably, the antibody domain is a V _H or a V _L antibody domain.

The antibody domain may be a Camelid VHH domain (i.e. a V domain derived or derivable from a Camelid antibody consisting of two heavy chains).

The antibody domain may be part of a monoclonal antibody (mAb), e.g. IgG, IgE IgM, IgA ₅ scFv and Fab fargments or a dAb. dAbs are described in Ward et al. (1989) Nature 341, p544-546.

Preferably, the antibody VL domain is VK-

NUCLEOTIDE SEQUENCES

Nucleotide sequences according to the present invention may comprise any nucleic acid (for example, DNA, RNA or any analogue, natural or artificial, thereof).

The DNA or RNA may be of genomic or synthetic or of recombinant origin (e.g. cDNA), or combinations thereof.

The nucleotide sequence may be double-stranded or single-stranded whether representing the sense strand or the antisense strand or combinations thereof. The nucleotide sequence may be a gene.

Preferably, the nucleotide sequence is selected from the group consisting of a DNA molecule, an RNA molecule, a partially or wholly artificial nucleic acid molecule consisting of exclusively synthetic or a mixture of naturally-occurring and synthetic bases, any one of the foregoing linked to a polypeptide, and any one of the foregoing linked to any other molecular group or construct.

AMPLIFICATION

The method according to the invention uses strand displacement replication of the nucleic acid sequences by multiple primers. The method can be used to amplify circularized DNA molecules in DNA libraries. The method generally involves hybridization of primers to a target nucleic acid sequence and replication of the target sequence primed by the hybridized primers such that replication of the target sequence results in replicated strands complementary to the target sequence. During replication, the growing replicated strands displace other replicated strands from the target sequence (or from another replicated strand) via strand displacement replication. Examples of such displacement of replicated strands are illustrated in U.S. Pat. No. 6,323,009. As used herein, a replicated strand is a nucleic acid strand resulting from elongation of a primer hybridized to a target sequence or to another replicated strand. Strand displacement replication refers to DNA replication where a growing end of a replicated strand encounters and displaces another strand from the template strand (or from another replicated strand).

Displacement of replicated strands by other replicated allows multiple copies of a target sequence to be made in a single, isothermal reaction. This makes amplification less complicated and much more consistent in output. Strand displacement allows rapid generation of multiple copies of a nucleic acid sequence or sample in a single, continuous, isothermal reaction.

It is preferred that the set of primers used for SDA have a sequence composition and be used at concentrations that allow the primers to hybridize at desired intervals within the nucleic acid sample. For example, by using a set of primers at a concentration that allows them to hybridize, on average, every 4000 to 8000 bases, DNA replication initiated at these sites will extend to and displace strands being replicated from adjacent sites. It should be noted that the primers are not expected to hybridize to the target sequence at regular intervals. Rather, the average interval will be a general function of primer concentration.

Displacement of an adjacent strand makes it available for hybridization to another primer and subsequent initiation of another round of replication. The interval at which primers in the set of primers hybridize to the target sequence determines the level of amplification. For example, if the average interval is short, adjacent strands will be displaced quickly and frequently. If the average interval is long, adjacent strands will be displaced only after long runs of replication.

EXPRESSION

Expression, as used herein, is used in its broadest meaning, to signify that a nucleotide sequence is converted into its gene product.

Thus, where the nucleic acid is DNA, expression refers to the transcription of the DNA into RNA; where this RNA codes for protein, expression may also refer to the translation of the RNA into protein. Where the nucleic acid is RNA, expression may refer to the replication of this RNA into further RNA copies, the reverse transcription of the RNA into DNA and optionally the transcription of this DNA into further RNA molecule(s), as well as optionally the translation of any of the RNA species produced into protein.

Preferably, therefore, expression is performed by one or more processes selected from the group consisting of transcription, reverse transcription, replication and translation.

Expression of the nucleotide sequence may thus be directed into either DNA, RNA or protein, or a nucleic acid or protein containing unnatural bases or amino acids (the gene product).

ISOLATING/SORTING/SELECTING

The terms "isolating", "sorting" and "selecting", as well as variations thereof, are used herein.

"Isolation", according to the present invention, refers to the process of separating an polypeptide domain with a desired specificity from a repertoire of polypeptide domains having a different specificity.

In a preferred embodiment, isolation refers to purification of an polypeptide domain essentially to homogeneity.

"Sorting" of a polypeptide domain refers to the process of preferentially isolating desired polypeptide domains over undesired polypeptide domains. In as far as this relates to isolation of the desired polypeptide domains, the terms "isolating" and "sorting" are equivalent. The method of the present invention permits the sorting of desired nucleotide sequences from pools (libraries or repertoires) of nucleotide sequences which contain the desired nucleotide sequence.

"Selecting" is used to refer to the process (including the sorting process) of isolating a polypeptide domain according to a particular property thereof.

In a highly preferred application, the method of the present invention is useful for sorting libraries of polypeptide (e.g. antibody) domain nucleotide sequences. The invention accordingly provides a method, wherein the polypeptide domain nucleotide sequences are isolated from a library of nucleotide sequences encoding a repertoire of polypeptide domains, for example, antibody domains. Herein, the terms "library", "repertoire" and "pool" are used according to their ordinary signification in the art, such that a library of nucleotide sequences encode a repertoire of gene products. In general, libraries are constructed from pools of nucleotide sequences and have properties, which facilitate sorting.

METHOD OF IN VITRO EVOLUTION

According to a further aspect of the present invention, therefore, there is provided a method of in vitro evolution comprising the steps of: (a) selecting one or more polypeptide domains from a library according to the present invention; (b) mutating the selected polypeptide domain(s) in order to generate a further library of nucleotide sequences encoding a repertoire of gene products; and (c) iteratively repeating steps (a) and (b) in order to obtain a polypeptide domain with enhanced specificity.

Mutations may be introduced into the nucleotide sequences using various methods that are familiar to a person skilled in the art - such as the polymerase chain reaction (PCR). PCR used

for the amplification of DNA sequences between rounds of selection is known to introduce, for example, point mutations, deletions, insertions and recombinations.

In a preferred aspect, the invention permits the identification and isolation of clinically or industrially useful polypeptide domains. Ih a further aspect of the invention, there is provided a polypeptide domain when isolated, obtained or obtainable by the method of the invention.

Evolution may be performed in a variety of systems, including cellular systems, for example using phage or phagemid vectors, as well as in vitro systems such as microencapsulation. Such systems are described in detail in Tawfik, D.S. and Griffiths, A.D. (1998) Nat. Biotechnol, 16, 652-656, as well as European Patent Applications EP1019496 and EPl 141272. The selection of suitable encapsulation conditions is desirable. Depending on the complexity and size of the library to be screened, it may be beneficial to set up the encapsulation procedure such that 1 or less than 1 nucleotide sequence is encapsulated per microcapsule. This will provide the greatest power of resolution. Where the library is larger and/or more complex, however, this may be impracticable; it may be preferable to encapsulate nucleotide sequences together and rely on repeated application of the method of the invention to achieve sorting of the desired activity. A combination of encapsulation procedures may be used to obtain the desired enrichment.

Theoretical studies indicate that the larger the number of nucleotide sequence variants created the more likely it is that a molecule will be created with the properties desired (see Perelson and Oster, 1979 for a description of how this applies to repertoires of antibodies). Recently it has also been confirmed practically that larger phage-antibody repertoires do indeed give rise to more antibodies with better binding affinities than smaller repertoires (Griffiths et al., 1994). To ensure that rare variants are generated and thus are capable of being selected, a large library size is desirable. Thus, the use of optimally small microcapsules is beneficial.

In addition to the nucleotide sequences described above, the artificial microcapsules will comprise further components required for the sorting process to take place. Other components of the system will for example comprise those necessary for transcription and/or translation of

the nucleotide sequence. These are selected for the requirements of a specific system from the following; a suitable buffer, an in vitro transcription/replication system and/or an in vitro translation system containing all the necessary ingredients, enzymes and cofactors, RNA polymerase, nucleotides, nucleic acids (natural or synthetic), transfer RNAs, ribosomes and amino acids, to allow selection of the modified gene product.

A suitable buffer will be one in which all of the desired components of the biological system are active and will therefore depend upon the requirements of each specific reaction system. Buffers suitable for biological and/or chemical reactions are known in the art and recipes provided in various laboratory texts, such as Sambrook et al, 1989.

The in vitro translation system will usually comprise a cell extract, typically from bacteria (Zubay, 1973; Zubay, 1980; Lesley et al, 1991; Lesley, 1995), rabbit reticulocytes (Pelham and Jackson, 1976), or wheat germ (Anderson et al, 1983). Many suitable systems are commercially available (for example from Promega) including some which will allow coupled transcription/translation (all the bacterial systems and the reticulocyte and wheat germ TNT™ extract systems from Promega). The mixture of amino acids used may include synthetic amino acids if desired, to increase the possible number or variety of proteins produced in the library. This can be accomplished by charging tRNAs with artificial amino acids and using these tRNAs for the in vitro translation of the proteins to be selected (Ellman et al, 1991; Benner, 1994; Mendel et al, 1995).

In a preferred embodiment, the in vitro transcription reaction is performed for 1 hour or less at room temperature.

After each round of selection the enrichment of the pool of nucleotide sequences for those encoding the molecules of interest can be assayed by non-compartmentalised in vitro transcription/replication or coupled transcription-translation reactions. The selected pool is cloned into a suitable plasmid vector and RNA or recombinant protein is produced from the individual clones for further purification and assay.

Nucleotide sequences encoding polypeptide (e.g. antibody) domains that exhibit the desired binding - such as the native binding - can be selected by various methods in the art - such as affinity purification using a molecule that specifically binds to, or reacts specifically with, the polypeptide domain.

Sorting by affinity is dependent on the presence of two members of a binding pair in such conditions that binding may occur.

Ih accordance with the present invention, binding pairs that may be used hi the present invention include an antigen capable of binding specifically to the polypeptide (e.g. antibody) domain. The antigen may be a polypeptide, protein, nucleic acid or other molecule.

The term "binding specifically" means that the interaction between the polypeptide (e.g. antibody) domain and the antigen are specific, that is, hi the event that a number of molecules are presented to the polypeptide domain, the latter will only bind to one or a few of those molecules presented. Advantageously, the polypeptide domain-antigen interaction will be of high affinity.

Using affinity purification, a solid phase immunoabsorbant is used — such as an antigen covalently coupled to an inert support (e.g. cross linked dextran beads). The immunoabsorbant is placed in a column and the polypeptide domain is run in. Antibody to the antigen binds to the column while unbound antibody washes through. In the second step, the column is eluted to obtain the bound antibody using a suitable elution buffer, which dissociates the antigen- antibody bound.

Suitably, streptavidin-coated paramagnetic microbeads (e.g. Dynabeads, Dynal, Norway), coated with biotinylated target protein, are used as the solid phase support to capture those protein-DNA complexes which display desired activity.

Various immunoabsorbants for affinity purification are known in the art, for example, protein A, protein L, protein G.

Preferably, for model selection purposes, the immunoabsorbant is protein L.

Protein L exhibits a unique combination of species-specific, immunoglobulin-binding characteristics and high affinity for many classes of antibodies and antibody fragments. Protein L is a recombinant form of a Peptostreptococcus magnus cell wall protein that binds immunoglobulins (Ig) through light-chain interactions that do not interfere with the Ig antigen-binding site. A majority of Ig sub- classes, including IgG, IgM, IgA, IgD, IgE, and IgY, from human, mouse, rat, rabbit, and chicken possess light chains and can thus be bound with high affinity by Protein L. Protein L also binds Ig fragments, including scFv and Fab.

Commercially available kits can be obtained from, for example, Clonetech and SigmaAldrich.

Polypeptide domains binding to other molecules of interest - such as proteins, haptens, oligomers and polymers - can be isolated by coating them onto the chosen solid supports instead of protein L.

VECTORS

The nucleotide sequences of the present invention may be present in a vector.

The term "vector" includes expression vectors and transformation vectors and shuttle vectors.

The term "expression vector" means a construct capable of in vivo or in vitro expression.

The term "transformation vector" means a construct capable of being transferred from one entity to another entity - which may be of the species or may be of a different species. If the construct is capable of being transferred from one species to another - such as from an E. coli plasmid to a bacterium, such as of the genus Bacillus, then the transformation vector is sometimes called a

"shuttle vector". It may even be a construct capable of being transferred from an E. coli plasmid to an Agrobacterium to a plant.

The vectors may be transformed into a suitable host cell to provide for expression of a polypeptide.

The vectors may be for example, plasmid, virus or phage vectors provided with an origin of replication, optionally a promoter for the expression of the said polynucleotide and optionally a regulator of the promoter.

The vectors may contain one or more selectable marker nucleotide sequences. The most suitable selection systems for industrial micro-organisms are those formed by the group of selection markers which do not require a mutation in the host organism. Examples of fungal selection markers are the nucleotide sequences for acetamidase (amdS), ATP synthetase, subunit 9 (oliC), orotidine-S'-phosphate-decarboxylase (pvrA), phleomycin and benomyl resistance (benA). Examples of non-fungal selection markers are the bacterial G418 resistance nucleotide sequence (this may also be used in yeast, but not in filamentous fungi), the ampicillin resistance nucleotide sequence (E. coli), the neomycin resistance nucleotide sequence (Bacillus) and the E. coli uidA nucleotide sequence, coding for β-glucuronidase (GUS).

Vectors may be used in vitro, for example for the production of RNA or used to transfect or transform a host cell.

Thus, polynucleotides may be incorporated into a recombinant vector (typically a replicable vector), for example a cloning or expression vector. The vector may be used to replicate the nucleic acid in a compatible host cell.

Genetically engineered host cells may be used for expressing an amino acid sequence (or variant, homologue, fragment or derivative thereof).

EXPRESSION VECTORS

The nucleotide sequences of the present invention may be incorporated into a recombinant replicable vector. The vector may be used to replicate and express the nucleotide sequence in and/or from a compatible host cell. Expression may be controlled using control sequences, which include promoters/enhancers and other expression regulation signals. Prokaryotic promoters and promoters functional in eukaryotic cells may be used. Chimaeric promoters may also be used comprising sequence elements from two or more different promoters described above.

The protein produced by a host recombinant cell by expression of the nucleotide sequence may be secreted or may be contained intracellularly depending on the sequence and/or the vector used. The coding sequences can be designed with signal sequences, which direct secretion of the substance coding sequences through a particular prokaryotic or eukaryotic cell membrane.

HOST CELLS

As used herein, the term "host cell" refers to any cell that may comprise the nucleotide sequence of the present invention and may be used to express the nucleotide sequence.

Thus, in a further embodiment the present invention provides host cells transformed or transfected with a library according to the present invention. Preferably, said library is carried in vectors for the replication and expression of polynucleotides. The cells will be chosen to be compatible with the said vectors and may for example be prokaryotic (for example bacterial), or eukaryotic, such as mammalian, fungal, yeast or plant cells.

The gram-negative bacterium E. coli is widely used as a host for heterologous nucleotide sequence expression. However, large amounts of heterologous protein tend to accumulate

inside the cell. Subsequent purification of the desired protein from the bulk of E. coli intracellular proteins can sometimes be difficult.

In contrast to E. coli, bacteria from the genus Bacillus are very suitable as heterologous hosts because of their capability to secrete proteins into the culture medium. Other bacteria suitable as hosts are those from Streptomyces and Pseudomonas.

Depending on the nature of the polynucleotide and/or the desirability for further processing of the expressed protein, eukaryotic hosts such as yeasts or other fungi may be preferred.

The use of host cells - such as yeast, fungal and plant host cells - may provide for post- translational modifications (e.g. myristoylation, glycosylation, truncation, lapidation and tyrosine, serine or threonine phosphorylation) as may be needed to confer optimal biological activity on recombinant expression products of the present invention.

REGULATORY SEQUENCES

In some applications, polynucleotides may be linked to a regulatory sequence, which is capable of providing for the expression of the nucleotide sequence, such as by a chosen host cell. By way of example, the present invention covers a vector comprising the nucleotide sequence of the present invention operably linked to such a regulatory sequence, i.e. the vector is an expression vector.

The term "regulatory sequences" includes promoters and enhancers and other expression regulation signals.

The term "promoter" is used in the normal sense of the art, e.g. an RNA polymerase binding site.

Enhanced expression of polypeptides may be achieved by the selection of heterologous regulatory regions, e.g. promoter, secretion leader and terminator regions, which serve to

increase expression and, if desired, secretion levels of the protein of interest from the chosen expression host and/or to provide for the inducible control of expression.

Aside from the promoter native to the nucleotide sequence encoding the polypeptide, other promoters may be used to direct expression of the polypeptide. The promoter may be selected for its efficiency in directing the expression of the polypeptide in the desired expression host.

In another embodiment, a constitutive promoter may be selected to direct the expression of the polypeptide. Such an expression construct may provide additional advantages since it circumvents the need to culture the expression hosts on a medium containing an inducing substrate.

Examples of strong constitutive and/or inducible promoters which are preferred for use in fungal expression hosts are those which are obtainable from the fungal nucleotide sequences for xylanase (xlnA), phytase, ATP-synthetase, subunit 9 (oliC), triose phosphate isomerase (tpi), alcohol dehydrogenase (AdhA), α-amylase (amy), amyloglucosidase (AG - from the glaA nucleotide sequence), acetamidase (amdS) and glyceraldehyde-3-phosphate dehydrogenase (gpd) promoters.

Examples of strong yeast promoters are those obtainable from the nucleotide sequences for alcohol dehydrogenase, lactase, 3-phosphoglycerate kinase and triosephosphate isomerase.

Examples of strong bacterial promoters are the α-amylase and SP02 promoters as well as promoters from extracellular protease nucleotide sequences.

Hybrid promoters may also be used to improve inducible regulation of the expression construct.

The promoter can additionally include features to ensure or to increase expression in a suitable host. For example, the features can be conserved regions such as a Pribnow Box , a TATA box

or T7 transcription terminator. The promoter may even contain other sequences to affect (such as to maintain, enhance, decrease) the levels of expression of a nucleotide sequence. Suitable other sequences include the Shl-intron or an ADH intron. Other sequences include inducible elements - such as temperature, chemical, light or stress inducible elements. Also, suitable elements to enhance transcription or translation may be present. An example of the latter element is the TMV 5' signal sequence (see Sleat Gene 217 [1987] 217-225; and Dawson Plant MoI. Biol. 23 [1993] 97).

GENERAL RECOMBINANT DNA METHODOLOGY TECHNIQUES

The present invention employs, unless otherwise indicated, conventional techniques of chemistry, molecular biology, microbiology, recombinant DNA and immunology, which are within the capabilities of a person of ordinary skill in the art. Such techniques are explained in the literature. See, for example, J. Sambrook, E. F. Fritsch, and T. Maniatis, 1989, Molecular Cloning: A Laboratory Manual, Second Edition, Books 1-3, Cold Spring Harbor Laboratory Press; Ausubel, F. M. et al. (1995 and periodic supplements; Current Protocols in Molecular Biology, ch. 9, 13, and 16, John Wiley & Sons, New York, N.Y.); B. Roe, J. Crabtree, and A. Kahn, 1996, DNA Isolation and Sequencing: Essential Techniques, John Wiley & Sons; M. J. Gait (Editor), 1984, Oligonucleotide Synthesis: A Practical Approach, M Press; and, D. M. J. Lilley and J. E. Dahlberg, 1992, Methods of Enzymology: DNA Structure Part A: Synthesis and Physical Analysis of DNA Methods in Enzymology, Academic Press. Each of these general texts is herein incorporated by reference.

The invention will now be further described by way of Examples, which are meant to serve to assist one of ordinary skill in the art in carrying out the invention and are not intended in any way to limit the scope of the invention.

EXAMPLES

Material and Methods

Model amplifications

Regions of the pUC19 plasmid (20) were amplified by PCR using primers 5'-GAA ATT GCG GCC GCA TTT TTA ATT TAA AAG GAT CTA GGT G-3' and 5'-TGC ATT CTC GAG CAT TTC CCC GAA AAG TGC CAC CTG-3'. A fragment encoding the β-lactamase (hid) gene (ampicillin resistance marker) was amplified from pUC 19 using primers 5'-ACT ATT GCG GCC GCG AAG TTT TAA ATC AAT CTA AAG-3' and 5'-CCA AGT CTC GAG TTA CTT ACT TAT GTG CGC GGA ACC CCT ATT TG-3', while the chloramphenicol-acetyl-transferase (cat) gene (chloramphenicol resistance marker) was amplified from plasmid pBCSK ⁺ (Invitrogen) using primers 5'-AGG TCA CTC GAG TAA GTA AGT AAG GCA CGT AAG AGG TTC CAA C-3' and 5'-GTA GCA GCG GCC GCC GAA TTT CTG CCA TTC ATC-3'. AU PCR amplifications were performed using Expand HighFidelity polymerase (Roche). The amplified fragments were digested with Xhol and Notl (New England Biolabs) and purified by gel electrophoresis. Phosphatase treatment of DNA is preferably avoided, as the introduction of "nicks" into the circular product can prevent rolling-circle amplification. DNA concentrations were determined by measuring absorbance at 260 nm. After purification, ligations were performed over night at 16 ⁰ C at a concentration of 40 ng/μl vector and 40 ng/μl insert using 16 units/μl of T4 DNA ligase in ligase buffer (New England Biolabs). Ligation reactions were purified by phenol/chloroform extraction and diafiltration. Dilutions were performed by serial dilution of bla ligations into cat ligations (5- fold dilution steps were used).

Phi29 amplifications were performed in a 50 μl volume using 10 ng of purified ligation reaction as a template and 10 units of Phi29 polymerase (New England Biolabs). Reaction conditions were as follows: 1 mM dNTPs, 50 μM random hexamer primer, 0.1 mg/ml BSA, 50 mM Tris-HCl pH 7.4, 10 mM (NEU) ₂ SO ₄ , 10 mM MgCl ₂ , 4 mM DTT. The reaction was incubated over night at 30 ⁰ C and purified by phenol/chloroform extraction and diafiltration. The amplified concatemer was digested over night with Notl and the restriction digest purified using a PCR purification kit (Qiagen). Plasmids were re-circularized by self-ligation at dilute DNA concentrations (<1 ng/μl) using 4 units/μl of T4 DNA ligase in ligase buffer.

After two hours at room temperature the ligase was inactivated by phenol/chloroform extraction and the reaction concentrated by diafiltration. The reaction was transformed into E. coli TGl (21) by electroporation and plated on agar plates containing 4% glucose and either 100 μg/ml ampicillin or 15 μg/ml chloramphenicol. Plates were incubated at 37 ⁰ C over night.

Library construction

As a template we used a previously reported synthetic domain antibody repertoire, "library 1" (18), based on the human DP47 heavy chain framework. The library incorporates diversity in all three CDR regions and is cloned in a phage format. The repertoire was pre-selected to enrich for antibodies that resisted aggregation upon heating (data not shown).

Regions of the library comprising to CDRs 1/2 and CDR 3 (all with attached framework regions) were amplified by PCR from a plasmid preparation. Primer pairs 5'-ACG TCA GAA GAC ATC AGG TGC AGC TGT TGG AGT C-3', 5'-TGG ACT GAA GAC AGT CAC GGA GTC TGC GTA GTA TGT G-3' (CDRl/2) and 5'-GTA ACT GAA GAC TAG TGA AGG GCC GGT TCA CCA TC-3', 5'-TCA GTT GAA GAC CTC GAA TTC AGA TCC TCT TCT GAG ATG-3' (CDR3) were used in the amplification. Phagemid pR2 was amplified using the primer pair 5'-ATA GCT GAA GAC ATT TCG GCC GCA CAT TAT ACA GAC ATA GAG ATG AAC-3' and 5'-GAT TAC GAA GAC ACC CTG GGC CAT GGC CGG CTG GGC CGC ATA G-3'. The phagemid is derived from pHENl (22) and encodes a HiS ₈ -VSV tag upstream of the phage gene IH. Amplifications were performed using Expand HighFidelity polymerase (Roche). DNA concentrations were determined by measuring absorbance at 260 run. The amplified fragments were digested with Bbsl (New England Biolabs), gel purified and ligated at 16°C for six hours in a three-way ligation using 40 units/μl of T4 DNA ligase in ligase buffer (New England Biolabs). Vector concentrations of 40 ng/μl and insert concentrations of 22 ng/μl (CDRl/2) and 33 ng/μl (CDR3) were used in the ligation. After heat inactivation (5 minutes at 7O ⁰ C) and diafiltration, 125 ng of the reaction was either used as template for Phi29 amplification or directly transformed into E. coli XL-I Blue (Stratagene) by electroporation and plated on agar plates containing 4% glucose and 100 μg/ml ampicillin and incubated at 37°C over night.

Phi29 amplification of antibody library

Amplifications were performed using a 500 μl starting volume and 125 ng of purified ligation reaction as a template (reaction conditions as above). Due to the increased reaction volume the concentration of Phi29 polymerase was reduced to 0.025 units/μl in order to minimize the amount of enzyme required (as was the concentration of ligase, as below). The reaction was heated at 7O ⁰ C for 5 minutes before adding the polymerase and incubated at 30°C. The reaction volume was doubled every 3 hours by adding reaction mixture and polymerase (no additional template was added) up to a volume of 8 ml. After an additional 3 hours at 3O ⁰ C the reaction was stopped by heating to 70 ⁰ C for 20 minutes.

The amplified concatemer was cleaved by digestion with Hindiπ (New England Biolabs), the restriction endonuclease was heat inactivated and the reaction concentrated by diafiltration. Excess random hexamer primers were removed by gel filtration on a Sephacryl S300 column (Amersham). Plasmids were re-circularized by self-ligation at dilute DNA concentrations (<3 ng/μl) using 0.7 units/μl of T4 DNA ligase in ligase buffer (New England Biolabs). After incubation for two hours at room temperature the ligase was heat inactivated (20 minutes at 70 ⁰ C) and the reaction concentrated by dialfiltration. The reaction was transformed into E. coli XL-I Blue (Stratagene) by electroporation of lμl in twenty separate transformations, plated on agar plates containing 4% glucose and 100 μg/ml ampicillin and incubated at 37°C over night. For analyses of transformation efficiencies, 125 ng portions were transformed in five separate reactions into E. coli XL-I Blue bacteria by electroporation, plated on agar plates containing 4% glucose and 100 μg/ml ampicillin and incubated at 37 ⁰ C over night.

Selection of binders

Libraries were re-transformed into E. coli TGl (21) for selection. Phages were produced by co-infection with helper phage KM13 (23). Binding selections were performed directly from culture supernatant using strepavidin-sepharose (Amersham) and biotinylated β-galactosidase

(Sigma- Aldrich). Marvel milk powder and Tween-20 were present during the binding steps (5% (w/w) and 2% (v/v) respectively). Washes were performed with phosphate buffered saline supplemented with 0.05% (v/v) Tween-20. Phages were eluted with 10 μg/ml trypsin and used to infect TGl bacteria. For the second round of selection β-galactosidase (Sigma- Aldrich) was directly immobilized on MaxiSorb plates (Nunc).

Analysis of selected antibodies

Binders were detected after two rounds of selection by ELISA using β-galactosidase coated MaxiSorb plates (as above), anti-M13 HRP-conjugate (Amersham) and 3,3', 5,5'- tetramethylbenzidine as substrate. For expression, phagemids were transferred into E. coli HB2151 by phage infection. Proteins were purified using a Streamline rProteinA column (Amersham), followed by acid elution and gel filtration on a Sephadex HR75 column (Amersham). Surface plasmon resonance measurements were performed on a Biacore 2000 instrument using biotinylated β-galactosidase (Sigma-Aldrich), streptavidin (SA) chips and HBS-EP running buffer (Biacore).

Results

Amplification of ligation reactions

In a first step we tested whether our approach would amplify circular DNA from ligation reactions. As a model system we ligated an antibiotic resistance marker into a bacterial plasmid; we used PCR-amplified fragments comprising the origin of replication of the E. coli plasmid pUC19 and the ampicillin resistance marker from the same plasmid. After restriction digestion the resulting PCR products were joined by ligation.

A dilution series was prepared from the ligation reaction and used as a template for Phi29 amplification. Hexamer primers and Phi29 polymerase were added and the reactions incubated isothermally at 30 ⁰ C o/n, allowing circular ligation products to be amplified

through a rolling-circle mechanism (Figure 1). After amplification the resulting linear concatamers were cut with restriction endonuclease, and the plasmids were recircularised by self-ligation at low DNA concentrations. Both amplified and un-amplified reactions were electroporated into E. coli bacteria and plated on agar plates containing ampicillin.

Phi29 amplification significantly increased the number of transformants (about 50-fold), with the number of colonies directly proportional to the amount of template DNA used (Figure 2). We were able to obtain colonies from the amplified ligations with template dilutions that yielded no colonies from the un-amplified ligations (Figure 2; 5-6 and 5-7 dilution steps).

Amplification of mixtures

We then analyzed whether our method was also capable of amplifying mixtures of DNA molecules without detectable bias. For this purpose, the template dilution series (as above) had been "spiked" with a constant amount of a ligation reaction incorporating a different, selectable resistance marker (a PCR-amplified chloramphenicol gene ligated into the pU19 vector backbone). From the ratio of the colonies before and after amplification (by counting colonies on ampicillin or chloramphenicol plates), it appears that these two constructs are amplified to a similar extent (Figure 3), despite the differences in sequences and lengths of the DNAs encoding the two resistance markers.

Generation of a large antibody repertoire from nanogram-scale ligation reactions

We generated a repertoire of antibody variable heavy chains by recombining gene segments from a library (18) with diversity in all three complementary determining regions (CDRs) (Figure 4). Thus, we PCR-amplified a segment comprising CDRs 1 and 2, and also a segment comprising CDR 3 from the library. The two segments of amplified DNAs were cut with restriction endonuclease and recombined by ligation into a phagemid vector, itself generated by PCR, in order to create a combinatorial library. As three-way ligations are inefficient (2), we tried to optimize the process. For this purpose restriction sites cleavable by the type Es restriction endonuclease Bbsl were introduced into the PCR primers, enabling "sticky" but

non-palindromic overhangs and non-ambiguous assembly. Nevertheless when a portion of the ligation reaction (comprising 125 ng vector backbone, >10 ^π DNA molecules) was directly transformed into E. coli bacteria, only 2x10 ⁴ colonies were obtained (and 10 ⁴ fold less than with the same amounts of circular supercoiled plasmid).

An identical portion of the ligation was then used as a template for Phi29 amplification. This polymerase exhibits a strong exonuclease activity and primers incorporating phosphothioate linkages (that are resistant to such exonucleolytic attack) have been reported to substantially increase DNA yields (6). However, we had noted that the incorporation of phosphothioates also reduced transformation efficiencies (data not shown). We therefore adopted a different strategy; we added primer at regular intervals during the amplification in an attempt to maintain the primer concentration (see Material and Methods for details). In addition, the final reaction volume was significantly increased (160-fold) compared to the model amplifications (as above). The amplified concatemers were then purified by gel filtration chromatography, cleaved by restriction digestion and the plasmids re-circularized by self- ligation. The process yielded approximately 0.1 mg of amplified DNA (corresponding to a 10 ³ -fold amplification of the template).

We compared transformation efficiencies of the amplified and un-amplified DNA by transforming multiple 125 ng portions of each into E. coli bacteria. This revealed that the transformation efficiency of the amplified DNA was significantly higher than that of the un- amplified DNA, with the average number of colonies increasing by a factor of 4x10 ³ . Clones from the transformed repertoires were analyzed by DNA sequencing; of these 31/36 (amplified) and 36/46 (un-amplified) were free of mutations outside the CDR regions and encoded full-length antibody domains. Thus, no detectable increase of point mutations or rearrangements was observed for the amplified repertoire.

Based on these initial experiments we then generated a large repertoire in bacteria by transforming 1 μg portions of the amplified DNA into E. coli. Each transformation yielded approximately 1.5x10 ⁸ colonies; twenty were pooled, providing a library of 3x10 ⁹ transformants. The Phi29 amplification had therefore increased the number of transformants

by more than 10 ⁵ -fold, due to both the greater amounts of DNA and its higher transformation efficiency.

Analysis of library and selection of binders

To characterize the libraries we selected binders from both the amplified and un-amplified antibody library (encompassing 3x10 ⁹ and 2x10 ⁴ transformants respectively) by phage display. We used β-galactosidase as the antigen, performed two rounds of binding selections and then screened 18 clones of each repertoire by phage ELISA (19). Among these, seven binders comprising four unique DNA sequences were observed for the amplified repertoire but none for the un-amplified repertoire (Table 1). The proteins were expressed in E. coli and purified by affinity chromatography using protein A-Sepharose. All bound to β-galactosidase as determined by ELISA and by surface plasmon resonance, where we observed dissociation rates in the range of lO^-lO ^"2 s ^"1 for three of the four proteins. The off-rates for these single antibody domains are therefore comparable to those reported for single chain Fv fragments isolated from a large repertoire (>10 ¹⁰ clones) (18).

Table 1

Selection of antibody repertoire. Beta-galactosidase was used as an antigen in phage display selections. Antigen-binding clones were identified after two selection rounds by phage ELISA and their DNA sequences were determined.

Conclusions

Our results demonstrate that Phi29 polymerase can amplify ligation reactions in-vitro without detectable biases, yielding sub-milligram quantities of DNA and potentially increasing the number of transformants in excess of 10 ⁶ fold. This enabled us to rapidly generate a library of antibody heavy chains from which binders could be readily selected. We did not observe any binders in the un-amplified repertoire, strongly indicating that the amplification process had allowed us to tap molecular diversity otherwise lost in the transformation process. The methodology should also be suitable for the generation of large amounts of non-methylated DNA for the transformation of eukaryotic species. While for many host organisms re- circularization of the amplified concatemer will increase transformation efficiencies, other organisms (such as yeast) may be directly transformable with linear DNA.

Our method requires only minute quantities of DNA as a starting template. This allowed us to assemble transformable constructs in a single step from PCR products, removing the need for plasmid preparations or stepwise cloning. It may also facilitate the cloning of rare DNAs from fossil, environmental or forensic samples, that are frequently not accessible by direct cloning or amplification. Although we have so far exclusively used the method for the amplification of ligation reactions, it should also be suitable for other applications in which only limited amounts of circular DNA are available, such as the amplification of BAC libraries. We suggest that amplification of circular DNA by Phi29 polymerase in-vitro has the potential to improve a wide range of cloning techniques that are currently limited by the transformation efficiency of the host organism.

REFERENCES

1. Jackson, D.A., Symons, R.H. and Berg, P. (1972) Biochemical method for inserting new genetic information into DNA of Simian Virus 40: circular SV40 DNA molecules containing lambda phage genes and the galactose operon of Escherichia coli. Proc Natl Acad Sci USA, 69, 2904-2909.

2. Legerski, RJ. and Robberson, D.L. (1985) Analysis and optimization of recombinant DNA joining reactions. J MoI Biol, 181, 297-312.

3. Miller, J.F. (1994) Bacterial transformation by electroporation. Methods Enzymol, 235, 375-385.

4. Dower, WJ., Miller, J.F. and Ragsdale, CW. (1988) High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res, 16, 6127-6145.

5. Conley, E.C. and Saunders, J.R. (1984) Recombination-dependent recircularization of linearized pBR322 plasmid DNA following transformation of Escherichia coli. MoI Gen Genet, 194, 211-218.

6. Dean, F.B., Nelson, J.R., Giesler, T.L. and Lasken, R.S. (2001) Rapid amplification of plasmid and phage DNA using Phi 29 DNA polymerase and multiply-primed rolling circle amplification. Genome Res, 11, 1095-1099.

7. Lizardi, P.M., Huang, X., Zhu, Z., Bray- Ward, P., Thomas, D.C. and Ward, D.C. (1998) Mutation detection and single-molecule counting using isothermal rollingcircle amplification. Nat Genet, 19, 225-232.

8. Fire, A. and Xu, S.Q. (1995) Rolling replication of short DNA circles. Proc Natl Acad Sci USA, 92, 4641-4645.

9. Blanco, L. and Salas, M. (1985) Replication of phage phi 29 DNA with purified terminal protein and DNA polymerase: synthesis of full-length phi 29 DNA. Proc Natl Acad Sci U S A, 82, 6404-6408.

10. Esteban, J.A., Salas, M. and Blanco, L. (1993) Fidelity of phi 29 DNA polymerase. Comparison between protein-primed initiation and DNA polymerization. J Biol Chem, 268, 2719-2726.

11. Paez, J.G., Lin, M., Beroukhim, R., Lee, J.C., Zhao, X., Richter, D.J., Gabriel, S., Herman, P., Sasaki, H., Altshuler, D. et al. (2004) Genome coverage and sequence fidelity of phi29 polymerase-based multiple strand displacement whole genome amplification. Nucleic Acids Res, 32, e71.

12. Holbrook, J.F., Stabley, D. and Sol-Church, K. (2005) Exploring whole genome amplification as a DNA recovery tool for molecular genetic studies. J Biomol Tech, 16, 125- 133.

13. Hutchison, C.A., 3rd, Smith, H.O., Pfannkoch, C. and Venter, J.C. (2005) Cellfree cloning using phi29 DNA polymerase. Proc Natl Acad Sci USA, 102, 17332-17336.

14. Nelson, J.R., Cai, Y.C., Giesler, T.L., Farchaus, J. W., Sundaram, S.T., Ortiz- Rivera, M., Hosta, L.P., Hewitt, P.L., Mamone, J.A., Palaniappan, C. et al. (2002) TempliPhi, phi29 DNA polymerase based rolling circle amplification of templates for DNA sequencing. Biotechniques, Suppl, 44-47.

15. Reagin, MJ., Giesler, T.L., Merla, A.L., Resetar-Gerke, J.M., Kapolka, K.M. and Mamone, J.A. (2003) TempliPhi: A sequencing template preparation procedure that eliminates overnight cultures and DNA purification. J Biomol Tech, 14, 143- 148.

16. Sato, M., Ohtsuka, M. and Ohmi, Y. (2005) Usefulness of repeated GenomiPhi, a phi29 DNA polymerase-based rolling circle amplification kit, for generation of large amounts of plasmid DNA. Biomol Eng, 22, 129-132.

17. Buckley, P.G., Mantripragada, K.K., Benetkiewicz, M., Tapia-Paez, L, Diaz De Stahl, T., Rosenquist, M., AIi, H., Jarbo, C, De Bustos, C, Hirvela, C. et al. (2002) A full-coverage, high-resolution human chromosome 22 genomic microarray for clinical and research applications. Hum MoI Genet, 11, 3221-3229.

18. Jespers, L., Schon, O., Famm, K. and Winter, G. (2004) Aggregation-resistant domain antibodies selected on phage by heat denaturation. Nat Biotechnol, 22, 1161-1165.

19. McCafferty, J., Griffiths, A.D., Winter, G. and Chiswell, DJ. (1990) Phage antibodies: filamentous phage displaying antibody variable domains. Nature, 348,

552-554.

20. Yanisch-Perron, C, Vieira, J. and Messing, J. (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene, 33, 103-119.

21. Gibson, TJ. (1984) PhD thesis, University of Cambridge, Cambridge.

22. Hoogenboom, H.R., Griffiths, A.D., Johnson, K.S., Chiswell, DJ., Hudson, P. and Winter, G. (1991) Multi-subunit proteins on the surface of filamentous phage: methodologies for displaying antibody (Fab) heavy and light chains. Nucleic Acids Res, 19, 4133-4137.

23. Kristensen, P. and Winter, G. (1998) Proteolytic selection for protein folding using filamentous bacteriophages. Fold Des, 3, 321 -328.

AU publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are obvious to those skilled in molecular biology or related fields are intended to be within the scope of the following claims.