WO1999025739A1 - Variable region fusion peptides that form effector complexes in the presence of antigen - Google Patents

Abstract

The fusion polypeptides of this invention contain a variable region sequence linked to an effector sequence. The polypeptides do not form stable complexes in solution, except in the presence of an antigen. Upon contacting with the antigen, the two variable region sequences bind together, which in turn drives the effector sequences into juxtaposition. Complementation of one effector sequence with the other provides an effector function, such as enzyme activation, which has a number of therapeutic or diagnostic applications.

Description

VARIABLE REGION FUSION PEPTIDES THAT FORM EFFECTOR COMPLEXES

IN THE PRESENCE OF ANTIGEN

REFERENCE TO RELATED APPLICATIONS:

This application claims the priority benefit of U.S. provisional application 60/065,719, filed November 14, 1997, pending. The priority application is hereby incorporated herein by reference in its entirety.

TECHNICAL FIELD

This invention relates generally to the fields of immunochemistry and peptide association. More specifically, it provides a system for obtaining fusion polypeptides with variable regions that drive dimerization in the presence of antigen.

BACKGROUND

Antibody molecules have been designed by evolution to direct a relatively non-specific effector function on to a specific target. The antibody repertory of an individual can be primed against a limitless variety of foreign antigens. Upon revisitation of a previously encountered antigen, the induced antibody will bind and bring into play elements of the complement cascade, or Fc receptor bearing cells with all their capabilities.

The contemporary biomolecular chemist has capitalized on the targeting specificity of the antibody for diagnostic and therapeutic purposes. Attaching the antibody with a label permits the detection or quantitation of antigen in a test solution. Attaching the antibody to a drug permits targeting to certain cells or tissues. New ways of delivering an effector function by way of an antibody are clearly of benefit.

Immunoassays used in routine clinical measurement involve an antibody specific for an analyte of interest in a biological sample. In separation based assays, the detecting of the complex involves a process wherein the complex formed is physically separated from either unreacted analyte, unreacted antibody, or both (U.S. Patent No. 3,646,346). The complex can be first formed in the fluid phase, and then subsequently captured by a solid phase reagent or separated on the basis of an altered physical or chemical property, such as by gel filtration or precipitation.

Alternatively, one of the reagents can be attached to a solid phase before contacting with other reagents, and then the complex may be recovered by washing the solid phase free of unreacted reagents.

In homogeneous assays, the presence of the complex is detected by a property which at least one of the reactants acquires or loses as a result of being incorporated into the complex. Homogeneous assays known in the art include systems involving fluorochrome and fluorochrome quenching pairs on different reagents (U.S. Patent Nos. 3,996,345, 4,161 ,515, 4,256,834, and 4,261 ,968); enzyme and enzyme inhibitor pairs on different reagents (U.S. Patent Nos. 4,208,479 and 4,233,401 ); and chromophore and chromophore modifier pairs on different reagents (U.S. Patent No. 4,208,479). A particularly powerful homogeneous assay system is the cloned enzyme donor immunoassay (U.S. Patent No. 4,708,929). Two subunits of the enzyme β-galactosidase associate to provide the detectable signal, which is quantitatively affected by analyte-specific antibody except in the presence of a sample containing free analyte.

Recent advances in antibody engineering have produced various artificially engineered antibodies and chimeras. Many of these molecules are superior to the natural antibody in aspects such as stability, size, low production cost, higher affinity, or have additional functions such as bi- specificity.

The isolated heavy and light chain variable domains (V_H and V_L) of an antibody constitute a heterodimer known as the Fv fragment, which contains a single antigen binding pocket. Fv fragments may dissociate at low protein concentrations. Klein et al. measured the equilibrium and kinetic aspects of the interaction of isolated variable and constant domains of IgG, using ultraviolet difference spectroscopy. The equilibriuim binding curve between the light chain variable domain and Fd' (a heavy chain fragment containing the heavy chain variable domain) was 1.2 x 10⁶ M^"1 at pH 5.4. Subsequently, Hamel et al. found that the association between V_H and V_L did not depend on antigen specificity, and some variable domains associated better with a counterpart from another antibody molecule.

Isolated Fv fragments are expected to have better properties for penetration of solid tumor tissue, lower antigenicity, and improved pharmacokinetics. To prevent dissociation of the V_H and V_L, a single chain variable region (scFv) can be constructed in which the two variable domains are part of the same poly peptide chain, interconnected by a peptide linker (Tsumoto et al.). A comparison of strategies to stabilize immunoglobulin Fv fragments has been described by Glockshuber et al.

Various other constructs of antibody molecules have been prepared. Monoclonal antibodies of a non-human species can be humanized by placing the three antigen-binding CDR regions of each V_H and V_L of the specific antibody into the framework of human V_H and V_L. See, for example, EP 0329400.

Constructs have also been prepared in which antibody binding sites are part of a molecular chimera. Maeda et al. proposed preparing a chimeric molecule in which an antibody binding monodomain was bioengineered onto Vargula luciferase. Ueda et al. (1992) constructed artificial chimeric cell-surface receptors, combining murine IgM with the cytoplasmic portion of the human EGF receptor. The chimeric receptor showed both antigen binding and protein tyrosine kinase activity, but the kinase activity was constitutive and independent of antigen binding. With IgM lacking the CH2 domain, autophorphorylation increased with increasing concentrations of hapten- BSA conjugate Monovalent hapten could not induce phosphorylation, but inhibited stimulation by the conjugate

U S Patent No 4,859,609 (Dull et al ) constructed hybrid receptors that comprise the ligand binding domain of a predetermined receptor, and a heterologous reporter polypeptide The hybrid receptors are said to be useful for performing assays The ligand binding domain (something other than an immunoglobulin) undergoes a conformational change upon binding of the ligand, which in turn affects the reporter peptide attached on the C-terminal end The model reporter molecule is a phosphorylkinase An assay method is claimed, in which the hybrid receptor is incubated with a test sample, and then a conformational change is correlated with the presence of ligand in the sample

SUMMARY OF THE INVENTION

The fusion polypeptides of this invention contain a variable region sequence linked to an effector sequence The polypeptides do not form stable complexes in solution, except in the presence of an antigen for which the combined variable region is specific The antigen brings the variable region sequences on two polypeptides together, which in turn drives the effector sequences into juxtaposition Complementation of one effector sequence with the other provides an effector function of therapeutic or diagnostic importance

Embodiments of the invention include product embodiments In particular, the invention includes a pair of fusion polypeptides that complex with each other in the presence of an antigen, consisting of a first fusion polypeptide comprising a first variable domain sequence linked to a first effector sequence, and a second fusion polypeptide comprising a second variable domain sequence linked to a second effector sequence, wherein complexing between the first and second variable domain sequences in a solution is stabilized if the solution contains the antigen, wherein the first and second effector sequences do not complex with each other in a solution containing antigen when not attached to the first and second variable domain sequences, respectively, and wherein complexing between the variable domains in the first and second fusion polypeptides in the presence of the antigen results in complexing between the effector sequences The polypeptides of the invention are freely soluble in solution, and are not membrane proteins In certain embodiments, the polypeptides are covalently tethered The invention also includes one or other of the fusion polypeptides adapted to perform as a member of the polypeptide

Exemplary effector sequences are enzyme fragments and toxin fragments Where the effector sequences are enzyme fragments, the polypeptides preferably have one or more of the following features 1) conversion of the substrate to the product occurs more rapidly in a solution containing the two fusion polypeptides and the antigen, than in a solution containing the two fusion polypeptides but no antigen, 2) the first and second fusion peptides do not have the catalytic activity except when complexed with each other in the presence of the antigen; 3) the substrate does not promote complexing between the two enzyme fragments.

Also embodied is a method of preparing a pair of fusion polypeptides, comprising the steps of: a) selecting a first variable domain sequence and a second variable domain sequence that form a complex that is stabilized in a solution if the solution contains the antigen; b) selecting a first effector sequence and a second effector sequence that do not complex with each other in a solution containing the antigen; c) preparing a first fusion polypeptide in which the first variable domain sequence is linked to the first effector sequence, and a second fusion polypeptide in which the second variable domain sequence is linked to a second effector sequence; and d) confirming that the first fusion polypeptide forms a complex with the second fusion polypeptide that is stabilized in a solution if the solution contains the antigen, and that upon binding between the polypeptides, enzyme activity is reconstituted.

Further embodiments relate to a method of converting a substrate to a product in a manner that depends on the presence of an antigen, comprising the step of creating an environment that contains the antigen, the substrate, and a pair of fusion polypeptides. The environment can be, for example, the microenvironment inside an individual being treated with a therapeutic or pharmaceutical composition according of the invention, or an in vitro environment present in a reaction mixture for an assay.

Diagnostic embodiments include a method of measuring the amount of an antigen in a sample, comprising the steps of preparing a reaction mixture containing the sample, a pair of fusion polypeptides, and a substrate for the effector enzyme of the fusion polypeptides, and then measuring any product formed in the reaction mixture.

Other embodiments of the invention will be apparent from the description that follows.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram illustrating the procedure for measuring an antigen concentration in a sample according to the present invention.

FIG. 2 is a bar graph indicating the amount of V_H phage bound to the biotinylated V_L or biotinylated hen egg lysozyme (HEL), in the presence of variable analysis.

FIG. 3 is a calibration curve for the concentration of HEL prepared on the basis of the amount of V_H phage bound to the biotinylated V_L which is immobilized on solid-phase in a plate.

FIG. 4 is a graph plotting the HEL concentration vs. the absorption of samples which shows the amount of alkaline phosphatase-labeled V_H bound to the biotinylated V_L which is immobilized on solid-phase in a plate.

FIG. 5 is a graph plotting the HEL concentration vs. the increase in fluorescence intensity ratio between fluorescein-labeled V_Hand Rhodamine X-labeled V_L. FIG. 6 is a line drawing representing two fusion polypeptides of this invention interacting in the presence of an antigen. The variable region sequences (indicated by the solid lines) drive interaction of the effector sequences (indicated by the dotted lines), which I reconstitutes enzymatic activity. In this example, the combined variable region is specific for the model antigen hen egg lysozyme, and the effector sequences are monomer subunits of mitochondrial malate dehydrogenase.

FIG. 7 is a half-tone reproduction of a gel showing the size of the cloned encoding region for mitochondrial malate dehydrogenase.

DETAILED DESCRIPTION

This invention provides fusion peptide pairs with the property that they associate with each other when a particular substance (termed the "antigen") is present. As a result of the antigen- induced association, effector sequences come together in a way that can create a useful chemical or biological effect, or constitute a completed labeling complex that reflects the presence of the antigen

While the fusion peptide of the invention can be part of a larger protein or molecular complex, each fusion peptide minimally comprises the following elements:

• a driver of the complexation reaction, which is a variable domain sequence, with the property that it forms a stable complex with the opposing variable domain on the opposing peptide in the presence of the antigen;

• a reporter of the complexation reaction, termed the effector sequence, which does not substantially associate with the effector on the opposing peptide unless the variable sequences are complexed, and which has an enzymatic, chemical or biological property of interest. • a covalent linkage between the variable domain sequence and the effector sequence, which can be a peptide bond, a polypeptide linker sequence, or any other type of chemical structure covalently connecting the variable domain and the effector in a manner that permits the fusion peptide to have the required functional activity.

One example is shown in FIG. 6. The drawing shows the predicted three-dimensional structure of the polypeptide backbone of a fusion polypeptide pair which is in the complexed configuration. The two solid lines show V_H and V_L domains (left and right) of a monoclonal antibody specific for the antigen hen egg lysozyme. In the presence of the antigen, the domains associate along an interface of opposing β-pleated sheets, which form an antigen binding pocket oriented towards the bottom of the drawing. Each variable domain is coupled at the C-terminal end to a monomer subunit of malate dehydrogenase, shown by dashed lines. The monomer subunits of this enzyme normally self-associate along an interface to form an active enzyme complex. When adapted for use in the invention, the subunit interface is modified to prevent self-association, but permit association when driven by the variable region domains. The presence of antigen in the mixture can be detected by providing a substrate for the active enzyme, and measuring product formation.

What follows is a full description for the making and using of the invention. These fusion protein pairs have a number of diagnostic and therapeutic applications, which are described in a later section.

In addition to the techniques outlined in this disclosure, the practice of the invention will employ conventional techniques of molecular biology, genetic engineering, microbiology, cell biology, biochemistry and immunology, which are within the skill of the art. These techniques are explained fully standard textbooks, such as, "Molecular Cloning: A Laboratory Manual", second edition (Sambrook et al., 1989); "Oligonucleotide Synthesis" (M.J. Gait, ed., 1984); "Animal Cell Culture" (R.I. Freshney, ed., 1987); "Methods in Enzymology" (Academic Press, Inc.); "Handbook of Experimental Immunology" (D.M. Weir & C.C. Blackwell, eds.); "Gene Transfer Vectors for Mammalian Cells" (J.M. Miller & M.P. Calos, eds., 1987); "Current Protocols in Molecular Biology" (F.M. Ausubel et al., eds., 1987); "PCR: The Polymerase Chain Reaction", (Mullis et al., eds., 1994); "Current Protocols in Immunology" (J.E. Coligan et al., eds., 1991). Various polypeptide crosslinking agents are described in Hermanson, G.T., "Bioconjugate Techniques", Academic Press: New York, 1996; and in "Chemistry of Protein Conjugation and Cross-linking" by S.S. Wong, CRC Press, 1993.

The skilled artisan has several strategic choices for the preparation and assembly of the polypeptides of the invention. Generally, the polypeptides are obtained by a method that includes the following steps, which will each be discussed in turn.

• selecting two variable domain sequences that form a complex that is stabilized in a solution if the solution contains the antigen of interest;

• selecting two identical or non-identical effector sequences that complement to provide the effector function, and do not substantially complex with each other in a solution containing the antigen;

• preparing fusion peptides each containing one variable domain sequence and one effector sequence; and

• confirming that the two fusion peptides have the desired functional properties.

The two variable domain sequences are usually V_H and V_L domains, although other combinations are possible (for example, homologous or heterologous V_L-V_L pairs in the Bence Jones configuration, and T cell variable region pairs). The variable domain sequences may correspond to a complete intact variable region domain, or may be longer and shorter in length, or incorporate amino acid changes, inserts, or deletions. Typically but not invariably, each variable domain will have the three CDR regions found in intact variable region. Sensitive to alterations are segments that make up the antigen binding site and the interface between the variable region pair, and changes should be made so as not to impair the required binding properties. The variable region domains can be of human origin, mouse origin, or of any other species, or they can be artificial sequences designed as a chimera or consensus of multiple species. Variable regions of human origin (or having human framework residues) are of interest for therapeutic applications, in order to minimize unwanted immunogenicity. Also of interest are variable regions of camel origin, or variable regions modified to incorporate camelid mutations which decrease the affinity between variable regions.

The "antigen" to which the variable region pairs bind can be a small molecule drug or hapten, protein, nucleic acid, carbohydrate, proteoglycan, glycolipid, or any structure which can be used to select the variable region pairs or binds the variable region pair with sufficient affinity and specificity. An antigen which induces the dimerization of a variable region pair is also referred to as a "driver antigen".

Raising and selecting variable regions with the specificity for a particular antigen is standard practice in the art. General techniques used in raising, purifying and modifying antibodies, and the design and execution of immunoassays, are found in Handbook of Experimental Immunology (D.M. Weir & C.C. Blackwell, eds.); Current Protocols in Immunology (J.E. Coligan et al., eds., 1991); David Wild, ed., The Immunoassay Handbook (Stockton Press NY, 1994); and R. Masseyeff, W.H. Albert, and N.A. Staines, eds., Methods of Immunological Analysis (Weinheim: VCH Verlags gesellschaft mbH, 1993). For hybridoma technology, the reader is referred generally to Harrow & Lane (1988), U.S. Patent Nos. 4,491 ,632, 4,472,500, and 4,444,887, and Methods in Enzymology, 73B:3 (1981). Briefly, the immunogen is optionally modified to enhance immunogenicity, for example, by aggregating with glutaraldehyde or coupling to a carrier like KLH, and then mixed with an adjuvant, preferably Freund's complete adjuvant for the first administration, and Freund's incomplete adjuvant for booster doses. The most common way to produce monoclonal antibodies is to immortalize and clone a splenocyte or other antibody-producing cell recovered from an animal that has been immunized. The cione is immortalized by a procedure such as fusion with a non- producing myeloma, by transfecting with Epstein Barr Virus, or transforming with oncogenic DNA. The treated cells are cloned and cultured, and clones are selected that produce antibody of the desired specificity. Specificity testing is performed on clone supernatants usually by immunoassay. Other methods for obtaining specific variable regions from antibodies or T cells involve contacting a library of immunocompetent cells or viral particles with the target antigen, and growing out positively selected clones. Immunocompetent phage can be constructed to express immunoglobulin variable region segments on their surface. See Marks et al., New Engl. J. Med. 335:730, 1996; WO patent applications 94/13804, 92/01047, 90/02809; and McGuinness et al., Nature Biotechnol. 14:1149, 1996. Phage of the desired specificity are selected by adherence to antigen attached to a solid phase, and then amplified in E. coli.

Screening variable regions with the property of antigen-dependent association involves assaying the association of one of the variable regions with the other in the presence and absence of antigen. Solid phase enzyme or fluorescein labeled association tests are quite appropriate, and fully described in USSN 08/663,922 by Ueda et al., which is hereby incorporated herein in its entirely. The association between heavy and light chains is due in large part to association between C_H1 and C_L. It is estimated that about 1 in 10 variable region pairs have sufficiently low association constant when detached from the constant regions for use in this invention without further association. The association constant is predicted to be a function of interacting residues along the interface. Accordingly, variable domains that have an antigen-dependent association can be obtained for any antigen, using this selection strategy.

The selected variable region pair should have an association constant of one variable region for the other should be at least 10-fold higher in the presence of antigen, and is progressively more preferred if it is at least about 10², 10³, 10", or 10⁵ fold higher. Association in the absence of antigen is generally less than 10⁸ M^"1 and preferably less than 10⁶ M^"1. Association of the variable regions for each other in the presence of antigen, and association of antigen for the variable region complex, is generally over 10⁸ M^"\ preferably above 10¹⁰ M^~1, and more preferably above 10¹² M^"1. Association constants can be modified, if desired, by altering amino acids along the interface. It is not necessary to measure the affinities to practice the invention, as long as a sufficient difference is observed in the presence or absence of antigen in the intended context.

The effector sequences have the property that they do not associate with each other when in their fusion polypeptides, except when driven together by the variable region domains. In most instances, the effector sequences will also not associate with each other when not connected with the variable region domains, regardless of whether antigen is present. When driven together in the fusion polypeptides, the effector sequences will interact with each other at an interface typically separated by less than 10 Angstroms. Association of the effector sequences for each other when not driven by antigen is generally less than 10⁸ M^"1 and preferably less than 10⁶ M^"1 in the environment of its intended use. Where the effector sequences are complementing enzyme fragments, the enzyme substrate (or other component of the reaction mixture) should not be able to induce association of the effector sequences. Low levels of undriven association are more tolerable and may even be of assistance when the association of the variable regions for each other in the absence of antigen is negligible. The degree of association can be measured by techniques known in the art, such as gel filtration, blotting techniques, and quantitative solid-phase separation assays. A particularly convenient method for measuring the association constant is using a BIAcore™ SPR biosensors made by Pharmacia (Uppsala, Sweden) according to manufacturers directions (see in particular the BIAtechnology handbook). While effector sequences with these properties may be prepared by any known technique, such as de novo computer modeling, it is more convenient to start with self-associating fragments or subunits of a protein with the desired effector function. The protein is then modified to prevent the association but maintain a surface that can act as an interface when the units are driven together. There are several types of modification that can be used. If subunit association is stabilized by disulfide bonds, then the cysteines can be blocked or preferably replaced with another amino acid. If there is strong subunit association due to noncovalent forces, then it may be possible to reduce the polypeptide length of each subunit until the association constant is sufficiently low. In a third approach, the interface residues are mutated in various test combinations. Where

X-ray crystallographic data is available, some predictions can be made as to the residues likely to be involved in interaction — such as opposing hydrophobic side chains or opposing negative and positively charged side chains. The importance of subunit or fragment interfaces in assembling enzymatic activity is well established: See, for example, Jones et al. (1985, Biochemistry 24:5852, 18), Ward et al. (1987, Biochemistry 26:4131), and Babe et al. (1992, Protein Science 1:1244). In another example, Lu et al. demonstrated the importance of the dimer-dimer interface for fructose- 1 ,6-bisphosphatase. The general approach of site directed mutagenesis as a tool for enzyme mechanism dissection is reviewed by Wagner et al. in Trends Biotechnol. 8:263, 1990. The expression of candidate effector sequences in a phage display library may assist in the screening of different mutation candidates.

Following fragmentation or adaptation of the interface, and following optional alteration of substrate specificity characteristics as described below, the effector substance will have a degree of identity with the native enzyme subunit or fragment, generally of the order of at least about 70%, and possibly at least about 80% or 90%. Identity is calculated as the percent of amino acids in the consecutive sequence of the native molecule that are preserved in the same order (with no penalty for gaps or inserts) in the adapted form, and is independent of enzyme specificity. Where the native enzyme is made up of non-identical fragments or subunits, the polypeptide pair will usually also have non-identical effector sequences. Where the native enzyme is made up of identical fragments or subunits, the polypeptide pair can have identical or non-identical effector sequences, independently of the variable region sequences, which are usually non-identical. For instances where the effector sequence are non-identical, the choice of which effector sequence to attach to which variable domain sequence is determined empirically.

The interacting effector sequences complement each other to provide an effector function of some chemical or biological interest, such as formation of an epitope, a ligand for a receptor, an enzyme, or a toxin. Contexts where each of these embodiments are relevant are described in a later section.

A number of enzymes have complementing subunits or fragments. Amongst them are dihydrofolate reductase (DHFR), which regenerates tetrahydrofolate from dihydrofolate, using NADPH as the reductant. Roles of several amino acids in DHFR function have been elucidated, and the crystal structure has been obtained. Dicker et al., 1993, Proc. Natl. Acad. Sci. U.S.A. 90; Lee et al., 1996, Biochemistry 35: 7012-20. The X-ray crystal structure of the beta 2 homodimer of luciferase from V. harveyi has been determined and the active site partly described. Fisher et al., 1996, J. Biol. Chem. 271 : 21956-68.. Thoden et al., 1997, Protein Sc/ 6: 13-23; Tanner et al., 1997, Biochemistry 36: 665-72; Baldwin et al., 1995, Curr. Opin. Struct. Biol. 5: 798-809. The refined crystal structure of alkaline phosphatase from several organisms has been obtained. Hulett et al., 1991 , J. Biol. Chem. 266: 1077-84. The hydrophobic core of the E co// alkaline phosphatase has been successfully replaced with the core from the signal of the maltose-binding protein OmpA. Laforet et al., 1989, J. Biol. Chem. 264: 14478-85. Other enzyme structures that have been extensively studied include penicillin amidase, cytochrome c oxidase, lactose permease, maltose permease, cytochrome b5 reductase, staphylococcal nuclease, thioredoxin, barnase, \and beta- galactosidase. Another example is ribonuclease. Bovine Rnase-A consists of a single chain of 124 amino acids (13,683 mol wt). When subjected to limited digestion by subtilisin, the peptide bond between residues Ala 20 and Ser 21 is cleaved. The two fragments, S-peptide (residues 1-20) and S-protein (residues 21-124) can be separated and reconstituted to give the fully active complex ribonuclease S. Simonson et al., 1992, Biochemistry 31:8661; Varadarajan et al., 1992, Biochemistry 31:12315; Kim et al., 1992, Biochemistry 31 :12304.

Effector peptides with a cytotoxic function include bacterial toxins, such as cholera toxin and cholera toxin B subunit, E. coli heat-labile enterotoxin and its B subunit, Bordetella pertussis toxin and the subunits S2, S3, S4, and S5 (in any combination), diphtheria toxin and the β toxin fragment, shiga and shiga-like toxins, staphylococcal α-hemolysin, vibrio thermostable direct hemolysin, alpha- sarcin, ricin, and abrin.

Other cytotoxic effector peptides are RNases, such as colicins; T1 ribonucleases (including fungal ribonucleases); T2 ribonucleases (both plant style and seed RNases and ribosome- inactivating proteins). Also of interest are members of the RNase A superfamily, including RNase A, seminal RNase, RNase dimers, eosinophil-derived neurotoxin, eosinophil cationic protein, onconase, frog ribonucleases, and angiogenin. Also contemplated for use in this invention are RNases that are engineered to be cell-type selective by coupling to ligands for cell-surface receptors (cytotoxic ribonuclease chimeras). For a critical review of RNase chimeras, the reader is referred to Youle et al., 1993, Crit. Rev. Ther. Drug Carrier Syst. 10:1-18. Prior et al. (1996, Bioconjugate Chem. 7:23-29) have described a chimeric molecule of pseudomonas exotoxin, conjoined to barnase. This protein is toxic to cells due to its RNase activity, which is delivered to the cell by way of the endotoxin delivery pathway.

A critical consideration in choosing the effector sequences is the position of the terminating residues. Since it is necessary that the variable region sequences be able to drive the two enzyme effector sequences together, the three-dimensional distance between terminating residues of the interacting effectors optimally matches that of the interacting variable regions. The most usual configuration of the fusion peptides is for the C-terminus of each variable region to be linked to the N-terminus of each effector, although other configurations are possible. It is also possible to trim a few residues from the variable regions or the effectors, or both, to enhance the match of the spread. The opposite approach — that is, adding a linker sequence between the variable sequence and the effector sequence on one or both chains — becomes increasingly more difficult with increasing length of the linker. Precedents for conformational shifts through a connector between neighboring domains certainly exists, however, most notably represented by the immunoglobulins themselves. Where a linker is necessary, it is appropriate to begin with candidates that form a rigid bridge, such as a sequence predicted to form a helix. However, the problem is best avoided, and it is worth choosing effector sequences deliberately to match the span on the variable regions, adjusting the effector function if necessary to the intended purpose. The variable regions and effectors are then prepared as various fusion polypeptide candidates for empirical testing. As used in this disclosure, the term "fusion polypeptide" is a polypeptide made up of two or more amino acid sequences that do not normally occur together (or at least in the same configuration) in a naturally occurring protein. Fusion polypeptides are typically prepared by expressing a recombinant polynucleotide encoding it, either by PCR-type amplification or using a suitable expression vector, but polypeptide synthesis or conjugation of separate polypeptides using a cross-linking agent can also be used. The fusion proteins of this invention are designed to be freely soluble in solution, and are not membrane proteins.

The testing of candidate fusion polypeptides involves verifying that the polypeptide chains assemble under the proper conditions, and that the effector sequences interact upon assembly. The polypeptides should preferentially complex with each other in the presence of the analyte, but complexation should not be inducible by other possible compounds in the reacting environment, such as substrate or receptor for the effector sequences. The ability of the effector sequences to "complex" or interface with each other can be measured by techniques such as diference spectroscopy or circular dichroism. More usually, complexing of the subunits is inferred from the ability of the assembled polypeptide pair to provide the activity of the combined effector sequences: catalytic activity of the expected specificity, if the effector is an enzyme, or cytotoxic activity, if the effector is a toxin.

This invention also includes "tethered" compositions. In this embodiment, the polypeptide pair is interconnected through an interchain covalent bond or bridge. The bridge can be a disulfide bond, a peptide bond between amino acid side chains, or a chemical moiety created by treatment of the polypeptides with a crosslinking agent. The bridge is selected and positioned so as not to stabilize dimerization of the variable domain or of the effector domain, permitting the polypeptides to dissociate and pivot around the connection point. As a result, the effector will still not be in the fully active form until driven into the correct position by association of the variable domain sequences with antigen. When antigen is present, however, the reaction is nearly bimolecular. As a result of tethering, the activity of the enzyme will be higher in the presence of antigen, but the background in the absence of antigen will also be higher. Candidate polypeptide pairs that show low levels of activity can be adapted in this fashion by incorporating an additional cysteine into each of the opposing chains along the interface. One example of suitable position to create the tethering effect without causing the effector to assume a permanently activated form is near the base of the variable domain sequences.

Once a pair of fusion polypeptides with all the desired properties has been obtained, it can serve as a prototype for other polypeptides. The fold of immunoglobulin variable region domains is consistent between variable regions of different specificity. Accordingly, one variable region domain can be substituted for another atop the effector sequence. Alternatively, the CDRs of a variable region of a new specificity can be casetted into the framework of a proven fusion polypeptide, taking care to avoid disturbing the association properties of the variable domain interface. It is also possible to modify the features of the effector sequences by substitution or mutation. For example, where the effector is an enzyme, mutations near the catalytic site can be used to change the substrate specificity. Hogan et al. (1995, Biochemistry 34:4225) describe this process for the enzyme L-lactate dehydrogenase. Enzymes can also be displayed on filamentous phage (Soumillion et al., Appl. Biochem. Biotechnol.), which may allow for efficent screening of mutants against a new substrate.

The fusion polypeptide pairs of this invention have a number of applications in both clinical medicine and research. Two applications of particular interest are as biopharmaceuticals and as assay reagents. The use of the fusion polypeptides will generally involve converting a substrate to a product in a manner that depends on the presence of an antigen, by creating an environment that contains the antigen, the substrate, and a pair of fusion polypeptides. The environment can be, for example, the microenvironment inside an individual being treated with a therapeutic or pharmaceutical composition according of the invention, or an in vitro environment present in a reaction mixture for an assay.

When adapted for use as biopharmaceuticals for human therapy, the variable region sequences, the effector sequences, and the linker sequences (if used) will typically be chosen to resemble human sequences as much as possible, to avoid immunogenicity. The specificity of the variable region and the function of the effector will depend on the nature of the embodiment. Veterinary and ex vivo therapeutic use is also contemplated.

One therapeutic embodiment relates to prodrug activation. In this case, the variable region of the combined peptide pair is specific for a small artificial molecule, termed the activator. The effector sequences assemble to form an enzyme which is capable of converting a prodrug into the active form. Both the prodrug and the activator are in the general circulation, with the prodrug in excess, acting as a drug reservoir. The activator is typically inert in its biological effect, except for its ability to assemble the polypeptide pair, which then activates the prodrug. Since the activator is small, it can potentially be administered orally, nasally, or by inhalation. As a result, the prodrug may be administered only on an occasional basis, and then titrated to the effective dose on an ongoing basis using the activator

This is of interest when the active form of the prodrug itself is a large molecule that is administered by injection or by some other invasive procedure Examples of this include any polypeptide drug, including growth factors such as GM-CSF, EPO, and insulin Polypeptide drugs can be converted into a prodrug according to the strategy outlined in USSN 60/[pendιng, attorney docket 33746-30011 00] The strategy involves using a cross-linking agent to form the prodrug into an inactive loop configuration The loop contains either a protease recognition sequence in the ammo acid sequence, or else an enzyme cleavable group within the cross-linker Examples of enzyme cleavable cross-linkers are outlined in USSN 08/883,632, and include those that are cleavable by glycosidase, phosphatase, amidase or esterase The combined effector sequences of the polypeptide pair mediating the prodrug activation would have the corresponding catabo c activity for either the peptide recognition sequence or the cross-linker

Another therapeutic embodiment relates to drug targeting In this case, the variable region of the combined peptide pair is specific for a target substance in a particular tissue or on a particular cell Suitable target substances include tissue specific antigens, such as one of the CD markers for certain cells in the hematopoietic line, the asialoglycoprotein receptor on hepatocytes, integπns on blood vessel walls, or IgE on mast cells and IgE-secreting lymphocytes Other target substances include malignancy markers like prostate-specific antigen, carcinoembπonic antigen, and gangliosides (enriched on melanoma cells), or mfectivity markers, such as viral core proteins presented by some infected cells The corresponding effector can take several forms In one form, it is a toxin, which assembles only on cells having the marker, inducing cell-specific lysis In another form, it is a prodrug activator The prodrug will be in the general circulation, but will only be activated near the surface of the target cell, by virtue of the targeted assembly of the polypeptide pair Suitable prodrugs include those described earlier, and also include small molecule compounds in an activatable form Of interest are those activatable by addition or removal of a phosphate, acetyl or amide group

The preparation of pharmaceutical compositions is conducted in accordance with generally accepted procedures for the preparation of pharmaceutical preparations See, for example, Remington's Pharmaceutical Sciences 18th Edition (1990), E W Martin ed , Mack Publishing Co , PA Pharmaceutical preparations suitable for human use are sterile and substantially free of mycobacteπa For systemic distribution, administration is typically intravenous or intramuscular, although other routes are possible, or the composition can be administered locally near the site of the intended effect Formulation of the composition typically includes both fusion polypeptides of the associating pair, although they can also be administered at separate sites or at separate times As an alternative, one or both polypeptides can be substituted with naked DNA or an expression vector having the corresponding encoding sequence, permitting expression of the polypeptides in situ When intended for use as an assay reagent, the polypeptide pair will have a variable region specificity for an antigen to be detected or quantified in a sample. The effector sequences may simply assemble to form an epitope, which can then be labeled with a secondary reagent, but more typically will directly form an active enzyme. One assay embodiment is a reagent for immunohistochemistry of tissue sections. The variable region will be specific for an antigen in the target tissue, and the enzyme will catalyze the precipitation of a stain or an electron-dense particle. In this fashion, the tissue can be developed with a single reagent containing the polypeptide pair, either mixed with or followed by substrate, avoiding the multiple incubation and washing steps of indirect immunostaining. Another assay embodiment is a separation type assays. Any of the classic immunoassays using solid phase capture or other types of separation can be adapted and simplified using the polypeptide pairs of this invention. In one example, a plastic surface is coated with an antigen- specific capture antibody, the surface is contacted with the sample, and then the surface is contacted with the polypeptide pair. Presence of antigen in the sample is revealed by the immediate conversion of substrate to product, mediated by antigen-induced assembly of the polypeptides. There is no washing step required between contacting the bound antigen with the polypeptides, and supplying the substrate.

The fusion polypeptides of this invention are particularly well adapted for homogeneous immunoassays. A quantitative assay simply involves preparing a reaction mixture containing the sample, the polypeptide pair, and the substrate. The rate of conversion of substrate to product will be directly related to the number of assembled polypeptide pairs, which in turn will be directly related to the amount of antigen in the sample. Where the dimerization constant of the polypeptides in the absence of antigen is low, the components of the reaction mixture can in principle be mixed in any order. For convenience, the two polypeptides and the substrate can be precombined, and then added to the sample, whereupon the development reaction will begin immediately. The product formed (or substrate consumed) can be measured at a certain time after mixing, or the rate can be measured following initiation of the reaction, either manually or in an automated procedure.

It will be recognized, however, that transient association of the polypeptides will invariably occur, even in the absence of antigen, and the polypeptide-substrate mixture will decay if not used within a reasonable period after mixing. This is more of a problem when the de-dimerization time in the absence of analyte is more than a few seconds. In this case, a two-step reaction may be preferable, in which a reaction mixture is first prepared with the sample and the polypeptide pair, and the substrate is added later. Another possible option is to premix the substrate with the polypeptide pair near the time of use, and maintain the mixture at about 0 to 4°C. The reaction mixture with the sample and the reagents is first incubated at 0 to 4°C, and then warmed to 37°C. Antigen binding by variable regions is often fairly independent of temperature, whereas conversion of product is often temperature-sensitive, depending on the activation energy of the conversion reaction. By manipulating the temperature in the manner described, the conversion of substrate will begin only upon warming, by which time the binding reaction may be close to equilibrium.

Any of the assay embodiments of this invention are suitable for routine clinical chemistry analysis. Suitable biological or clinical samples include but are not limited to urine, plasma, serum, and histological sections. Suitable antigens for detecting or quantifying include those that correlate with certain clinical conditions, such as ferritin, prostate specific antigen, alpha fetoprotein, carcinoembrionic antigen, hCG, prolactin, thyroid stimulating hormone, progesterone, T3 and T4, free T3 and T4, aldosterone, insulin, and so on. Also suitable are drugs administered in therapy or drugs of abuse. Drugs of abuse include LSD and other halucinogens, amphetamines, barbiturates, cannabinols, and the like. To optimize accuracy and precision, the enzyme formed upon assembly of the polypeptides will preferably a reaction which involves a deep color change in the solution. A substrate of particular interest for homogeneous assays of this invention is X-gal.

Reagents used in the assays of this invention (including each member of the polypeptide pair, the enzyme substrate, and antigen standards) can be packaged separately or in any combination into kit form to facilitate distribution. The reagents are provided in suitable containers, and typically provided in a package along with written instructions relating to assay procedures.

All patents, patent applications, articles and publications mentioned herein, both supra and infra, are hereby incorporated herein by reference. Further illustration of the development and use of fusion polypeptides according to this invention are provided in the Example section below. The examples are provided as a further guide to a practitioner of ordinary skill in the art, and are not meant to be limiting in any way.

EXAMPLES

Example 1 : Antigen-dependent association of V_k and Vμ

This example describes binding experiments conducted using variable region sequences from anti-hen egg lysozyme (anti-HEL) monoclonal antibody with the designation HyHEL-10. The Fv fragment was previously known to form a trimolecular complex of 39 kDa in size, as measured by exclusion chromatography. Its association constant (K_a) measured by titration microcalorimetry was previously reported as high as 4.2 x 10⁸ M^"1 at 30°C.

Fv fragments were expressed by BL21(DE3) transformed with pKTN2, essentially as described in Tsumoto et al. (J. Biol. Chem 269:28777, 1994). V_L and V_H were separated in an FPLC MonoQ HR5/5 anion exchange column (Pharmacia) using a buffer gradient (20 mM Tris buffer pH

8.8 containing 0 to 2 M NaCI), the V_L being essentially non-adsorbing under these conditions.

Alternatively, the variable region sequences were expressed separately, precipitated with 65% ammonium sulfate, dissolved in water, dialized into 0.1 M phosphate buffer, and optionally further purified using a DEAE resin. For further details, the reader is referred to Ueda et al., Nature Biotechnology 14:1714, 1996.

The binding kinetics were analyzed using the SPR biosensors BIAcore™ or Biacore™ 2000 (Pharmacia) at room temperature. Measurements were generally performed with a continuous flow of 5 μl/min of HBS buffer (10 mM HEPES, pH 7.44, 150 mM NaCI, 3.4 mM EDTA, 0.05% Tween™ 20). To measure binding of V_H fragment or HEL to wild-type or mutant V_L fragment, a minimum amount of the purified protein was immobilized by amine coupling to a CM5 sensor chip using 5mM formate pH 3.2 as buffer. Samples with various combinations of V_H and HEL concentrations were diluted in HBS and injected. To avoid denaturation as much as possible, stock V_H solution was thawed, diluted, and centrifuged at 15 krpm for 5 min at 4 degrees immediately before use.

HEL showed considerable binding to the V_L-immobilized sensor chip with fast on and off rates. On the other hand, the V_H fragment alone did not show a time-dependent increase in resonance units (RU), except for the slight increment in baseline. At concentrations up to 1.84 μM, V_H did not bind to the V_L chip. The equilibrium constant of HyHEL-10 V_H with the corresponding V_L, was below detection (<10⁵/M).

Biphasic dissociation was observed that might be due to the orientation of a portion of the immobilized V_L that disabled the association of V_H. To test this possibility, a mutant V_L, having the one lysine residue (Lys 47) located at the V_H. interface mutated to threonine, was made to exclude possible fragment association. The monoclonal antibody (Mab) with this mutation (V_LK49T), which is analogous to HyHEL-8 V_L retains antigen binding affinity (Lavoie et al.). The mutant V_L was expressed, purified, and immobilized on the sensor chip. When analytes containing a fixed concentration of V_H (1.9 μM) and various concentrations of HEL (0 to 1.4 μM) were applied to the sensor chip, very slow dissociation was observed. The r , calculated from RU of 1.4 μM HEL (600s to 800s) was as low as 2.73 x 10^~5 ± 1.43 x lO^/sec. The RU levels during dissociation phase correlated with the HEL concentration. As in the case of the wild-type V_L-chip, a very small amount of V_H fragment was observed to bind the VL(K49T) sensor chip in the absence of HEL. The result showed HEL-induced Fv stabilization on the sensor chip.

The association rate constant obtained by the curve fitting based on simple bimolecular model varied from 6.36 x 10³/M/sec (at 1.4 μM) to 9.28 x 10⁴/M/sec (at 88 nM) depending on HEL concentration. The apparent equilibrium association constant K.₂ was calculated to be 2.3 x 10⁸ to 3.4 x 10⁹/M, which is close to the value obtained by calorimetry.

When excessively biotinylated wild-type V_L fragment was used for V_L plate preparation, the absorbance response was severely inhibited. This was thought to be cause by biotinylation of Lys49 because no other lysines or N terminal NH₂ group were in the range of van der Waals contact with HEL or V_H.

Without intending to be limited by theory, one possible mechanism of antigen-dependent Fv stabilization is that simultaneous binding of V_H and V_L to HEL induces some conformational change of V_L.V_H interface residues, leading to increased stability of the complex. Example 2 Testing V_L and V_H binding using filamentous phage

Preparation of V_L-Domaιn Polypeptide of HyHEL-10 From the vector plasmid pKTN2 (Tsumoto, K et al , J Biol Chem 69, 28777-28782, 1994) which encodes pel B signal peptide sequence upstream o the structural genes of V_H and V_L of the antibody HyHEL-10 which is specific to HEL, the 670 bp portion thereof encoding the pe1B, V_L and ssi transcription termination sequence were cleaved by the restriction enzymes, Nhel and EcoRI, and purified in agarose electrophoresis This DNA fragment was ligated with the DNA fragment obtained by digesting the vector pET20b having a T7 promoter (Novagen Inc ) by using the restriction enzymes Xbal and EcoRI, to prepare V_L-expressιon vector pETVLhel Then, E coll BL21 (DE3) having T7 polymerase gene on its genome was transformed with this vector, and the transformant cells were cultivated at 30°C in LB medium When the cells reached a saturation density in 1 liter medium, they were collected with centnfugation, and re-suspended in 1 liter of fresh LB medium containing 0 5 m MPTG and then cultured continuously for further 24 hours The cells were again collected with centnfugation, and then ammonium sulfate was added to the supernatant thereof until 66% saturation, thereby precipitating the protein in the supernatant The precipitates were collected by centnfugation and dissolved in a small amount of water The aqueous solution was dialyzed into 10 mM phosphate buffer (pH 7 0), and subsequently, % volume of DEAE cellulose was added to absorb the protein other than V_L in the medium Repeating this process two more times could give V_L-domaιn polypeptide with a purity of 95% or higher

Preparation of the V_H-Domaιn Polypeptide of HyHEL-10, Displayed on M13 Phage

Firstly, a vector plasmid pluck 2001 was prepared by altering the vector portion pTZ18U of pluck 2000 (Japanese Patent Publication No 129516/1995) to pTZ19U This pluck 2001 possesses pe1 B signal peptide sequence, the part of V_L and V_H structural genes of HyHEL-10, c-myc tag, C- terminal of M13 phage gene 3 protein (g3p), M13 origin ampicillin resistance gene, and an origin derived from plasmid pUC series The following operations were performed to remove the V_L structural gene from the vector plasmid pluck 2001 That is, in order to combine the pel B signal sequence and V_H structural gene with the downstream c-myc tag and g3c C-terminus in matching reading frame of their codon, PCR reaction was conducted by using 1 ng of pluck 2001 as template and 50 pmol each of primers identified in SEQ ID NOS 1 and 2, respectively, in a 100 μl of reaction liquid containing 2 5-unιt Pfu DNA polymerase (Stratagene), dNTPs with final concentration of 0 2 nM and 10 μl of 10 X reaction liquid The thus obtained 380 bp DNA fragments were then digested with the restriction enzymes Sfil and NotI, and were ligated with fragments of pluck2001 vector having been digested with Sfil and NotI From the restriction enzyme analyses of the resultant plasmid, it was confirmed at the target one and was name pluck2010 E coll XL-1 Blue (Δ(lac), end A1 , gyrA96, hsdR17(rk , mk⁺), recA1 , re1A1, supE44, thil, [F', laclq, lacZ Δ M15, proAB, TM10(tet')j) was transformed with the pluck2010, and the colony was cultivated overnight at 37°C in 5 ml of LB medium containing 12.5 mg/l of tetracylin and 1% glucose. From this culture, 50 μl was taken out and mixed with 25 μl of M13VCS phage (Stratagene SC200251 , >10¹ ' pfu/ml), at 37°C for 20 minutes. Then, the mixture was transferred into a 5 ml of LB medium containing 100 mg/l of ampicillin and 70 mg/l of kanamycin, and was cultured with various shaking at 37°C for 16 hours. The culture was applied to a centnfugation to remove the cells, and the supernatant thereof was preserved at 4°C till it was used as a phage solution.

Immobilizing of V_L and HEL to a Microtiter Plate, and ELISA using the V_H-Phage: In order to immobilize V_Ls and HEL to microtite plate, they were biotinylated with biotin NHS

(Pierce) according to manufacturer's instructions. A PBS (I0 mM phosphate buffer: pH7.2, I50 mM NaCI) solution containing 10 μg/mi of streptavidin (Wako) was poured into a microliter plate (Falcon 3912) at a rate of 100 μ I per well and preserved at 4°C overnight to adsorb the streptavidin to the plate. After the solution was removed, the plate was blocked with 200 μ 1 of binding buffer (2% skimmed milk/PBS) for one hour at room temperature. The streptavidin plate thus obtained was washed twice with PBS containing 0.1% Tween™ 20 (PBS-T), and in succession, the biotinylated lysozyme which was diluted with PBS so that the concentration thereof was 10 mg/l. was poured by 100 μ 1 into the plate and preserved at room temperature for one hour. After removing the solution, the plate was rinsed two times using PBS-T. Into this plate were added 10 μ 1 of sample in which 0, 0.1 mg/ml or 10 mg/l of V_Ls were contained in PBS, and 90 μ I of V_H-phage solution which had been mixed with the equivalent amount of binding buffer 30 minutes before, and the resulting mixture was incubated at 37°C for one hour. After further two times of washing, 100 μ I of 1/5000 diluted peroxidase-labeled anti-MI3 antibody (Pharmacia) in binding buffer was added. The plate was washed five times after one hour at 37°C, and then the sample was measured for absorbance at 490 nm by the ordinary color development method using orthopenyl enediamine for the quantitative determination of the M13 phage fixed on the plate.

Another experiment was also conducted using biotinylated V_Ls instead of biotinylated lysozyme and lysozyme instead of the V_Ls as a sample, and the result obtained were compared with that of the first experiment. The results are indicated in FIG. 2. In the figure, V_L1-0 means that the batch 1 of biotinylated V_Ls was immobilized onto solid-phase in a plate, and V_H-phages and O μg/ml of HEL were added thereto for incubation. In this case, the batch 1 of biotinylated V_Ls was prepared by mixing 500 μg of purified V_L in 350 μ1 of 0.1 mo1/1 NaHC0₃ : pH8. 3, 150 mmo1/1 NaCI and 100 μg of biotin-NHS in 3.5 μl of DMSO were mixed, reacting at room temperature for 30 minutes, and thereafter, dialyzing into PBS containing 0.02% sodium azide. For the V_L1 , measurements were made with samples having HEL concentrations of final 0.1 and 10 μg/ml, respectively. When the batch 2 of biotinylated V_L, i.e., V_L2 (prepared using I μl of biotin-NHS solution instead of 3.5 μ1 of biotin-NHS solutions of V_L1) was immobilized onto solid-phase, measurements were also conducted in a similar way. Furthermore,~HEL-0 means that biotinylated HEL was immobilized onto solid- phase in a plate, V_Ls with a concentration of 0 μg/ml, together with V_H-phages, was added thereto for incubation. Tested concentrations of V_Ls were 0 μg/ml, 1 μg/ml and 10 μg/ml. The V_H-phages were prepared from three kinds of phage samples of independent colonies (phage 1 , 2 and 3). As shown in FIG. 2, it was confirmed that the amount of bounded M13 phages increased with increasing concentrations of the co-existing protein (HEL or V_L) in a sample, in both cases where V_Ls and HEL were immobilized onto solid-phase in a plate.

Measurement of the HEL Concentrations in a Sample by ELISA using VH-Phages: Similarly as in the procedures (c) above, by immobilizing biotinylated V_Ls on solid-phase in a plate, and by varying HEL concentrations in a sample, the amount of bounded VH-phages were measured, and a calibration curve was drawn based on the measurements.

The results are shown in FIG. 3. which indicates that the method of the instant invention is highly sensitive and reproducible enough to measure antigen of which final concentration is 0.015 μg (15 ng)/ml or higher. This sensitivity is essentially the same as the one provided by the conventional sandwich ELISA method.

Example 3: Separation assays using enzyme reporter molecules.

Preparation of the Expression Plasmid for E. coli alkaline phosphatase:

The chromosome DNA of E. coli XL1-blue was extracted by the known method (Sambrook, Fritsh, Maniatis, "Molecular Cloning, Ver. 2", 1989), and the 1450 bp portion encoding alkaline phosphatase (the gene: PhoA, EC3.1.3.1) was amplified by PCR. To add to the terminals of the PCR products the NotI site, the oligonucleotides of SEQ ID NOS. 3 and 4 were used as PCR primers. More specifically, 35 cycles of PCR processes were conducted with 1 ng chromosome DNA of E. coli XL1-Blue as template DNA in a 100 μ1 of reaction liquid containing 50 pmol each of the primers, and 2.5 unit of Tag DNA polymerase (Perkin Elmers). The alkaline phophatase gene fragments obtained by the PCR were digested with the restriction enzyme NotI, and purified in agarose electrophoresis. The expression vector pET20 b (Novagen Inc.) was digested with NotI, treated with the phosphatase from bovine small intestine (Takara) and ligated with the alkaline phosphotase DNA fragment. From the restriction enzyme analyses of the resultant plasmid, it was confirmed as the target plasmid and was named pAP.

Preparation of V_H-Alkaline Phosphatase-Expression Plasmid: From the vector plasmid p'CTN2, the 480 bp portion encoding peIBb signal sequence and V_H was amplified by PCR. To add 3'-terminal Hindlll site to the PCR products, the oligonucleotides of SEQ ID NOS. 5 and 6 were used as PCR primers. More specifically, PCR reactions were conducted by using pKTN2 plasmid as template DNA in a I00 μ1 of reaction liquid containing the primers of each 50 pmol, and 2.5 unit Pfu DNA polymerase. The DNA fragments obtained were digested with EcoRV and Hindlll, and purified by agarose electrophoresis. The resultant products were ligated with the pAP having been digested with EcoRV and Hindlll, and purified by agarose electrophoresis, thereby preparing a V_H-alkaline phosphatase (V_H-AP)-expression plasmid. From the restriction analyses of the plasmid, it was confirmed as the target plasmid and was named as pV_HAP.

Expression and Purification of VH-AP Chimeric Protein in E. coli.:

The E. coli BL21(DE3)LysS having T7 polymerase on its genome was transformed with pV_HAP by the calcium chloride method, and was cultivated at 37°C overnight in an LB medium containing 1.5% agar and antibiotics (50 mg/l ampicillin and 34 mg/l chloramphenicol), to form colonies. These colonies were transferred to a 5 ml of LB medium containing antibiotics (50 mg/l ampicillin and 34 mg/l chloramphenicol), and were cultured at 28°C overnight again. The cells reached to saturation density were collected by centnfugation, and were suspended in a 50 ml of fresh LB medium containing antibiotics (50 mg/l carbenicillin and 34 mg/l chloramphenicol). Following this, they were cultured at 28°C for three to four hours and were then transferred to one liter medium. When the cell density reached OD_6oo - 0.3 - 0.4, final 0.I mM of IPTG was added to induce the expression of V_H-AP chimeric proteins. The cells were cultured overnight, and using the 20 μ1 of culture liquid directly, they were subjected to SDS-polyacrylamide electrophoresis to confirm that the target proteins had been expressed. The cells were collected by centnfugation and dissolved in sonication buffer (59 mM NaH₂P0₄, I0 mM Tris-HC1:pH8.0). Then the cell wall was disrupted with a french press and the insoluble fractions were removed by ultracent fugation using SW50.1 rotor (at 3200 rpm for one hour) to prepare lysate.

The obtained V_H-AP chimeric protein was purified in two stages, one by TALON™ Metal Chelating Column, using 6 x His sequence encoded at the C-terminus of the protein, and the second by the negative ion exchange column. First, the TALON™ metal chelating resin (by Clontech Laboratories Inc.; 2 ml for the column head) and 20 ml of protein solution were stirred gently in a 50 ml tube at 4°C for one hour to adsorb the protein having a polyhistidine sequence onto the column. The resin was transferred to a 5ml column, where it was washed and eluted with 20 mM MES-Na buffer. Of the fractions obtained, those containing protein were collected and applied to a dialysis tube. They were condensed by externally covering them with an adequate amount of PEG6000, then were dialyzed into 20 mM Tris-CI solution (pHδ.O) and applied to a negative ion exchange column (MonoQ HR5/5, Pharmacia). The resultant fractions were confirmed with the SDS polyacrylamide electrophoresis. Those containing the V_H-AP chimeric proteins alone were collected and condensed, then dialyzed into, Tris solution (0.1 M Tris-CI , 0.1 M NaCI: pH8.0). These successive purification operations proceeded up to a point where the target chimeric protein could be identified as a single band on SDS polyacrylamide electrophoresis. The proteins obtained were measured for concentration by means of BCA protein assay (Pierce), and preserved at 4°C until they were used for measurement.

Measurement of HEL Concentration Using V_H-Alkaline Phosphatase: Using a V_L plate prepared similarly as in Example 1 , 10 μl of sample containing various concentrations of HEL in PBS and 90 μ1 of V_H-AP chimeric protein solution (4 μg/ml in O.IM Tris HCI: pHδ.O) were added to their respective wells, and incubated at room temperature for two hours. Each well was washed four times using PBS containing 0.1% Tween (PBS-T), a substrate (I mM 4-nitro phenylphosphate, Wako) was added at a rate of 100 μl per well, and 30 minutes and one hour thereafter, color development (yellow) was measured at absorbance of 410 nm.

The results are shown in FIG. 4. This figure plots the absorbance against the final concentrations of HEL at one hour after the reaction, with the plotted figures being the means of the two measurements. As is evident from this figure, it was confirmed that the method of this invention assure highly sensitive and reproducible measurement of antigen substance with a final concentration of I ng/ml. The measurement sensitivity achieved by using this alkaline phosphatase as a reporter molecule is approximately tenfold of that recorded in Example 1 , wherein filamentous phages were employed.

Example 4: Homogeneous assays using fluorochromes.

Preparation of V_H and V_L:

Using -the structural-.genes of V_H- and V_L-domain of the antibody HyHEL-10 and the vector plasmid pKTN2, and also using the known procedure, Fv fragments of the HyHEL-10 were prepared. E. coli BL21 (DE3) was transformed with pKTN2, and cultured at 30°C in a 5 ml of LB medium containing 50 m/1 ampicillin followed by successive culture in 50 ml and one liter of medium, until the cell density reached saturation. After collecting the cells, they were cultured in the same medium containing 0.4 mM IPTG for 24 hours, and the supernatant from centnfugation was subjected to salting out using 66% saturated ammonium sulfate, and the precipitate was dialyzed into 50 mM phosphoric acid (pH7.2) and 0.2M NaCI. After one cycle of centnfugation (10 k/rpm for 10 min.), the sample was adsorbed to the HEL affinity column equilibrated with the same buffer (in which 10 mg/ml HEL is immobilized to CNBR activated Sepharose™ 4B made by Pharmacia Co. in accordance with the manual), washed with 100 ml of 0.I M Tris-HCI (pH8.5) and 0.5 M NaCI, and eluted with 10 ml of O.I M glycine buffer (pH2.0). Immediately after elution, the sample was neutralized with the equivalent amount of 1 M Tris-HCI (pH7.5). By this process, approximately 10 mg of Fv fragment was recovered. It was dialyzed into

50 mM Tris-HCI (pH8.8), and separated into V_H and V_L polypeptides in a negative ion exchange chromatography column (MonoQ HR5/5, Phamarcia Co.). Since few of the V_L are adsorbed to the column, the V_H and V_L could be separated relatively easily, with the non-binding fractions taken as V_L and the binding fractions as V_H. The purity of both polypeptides after separation was found to be 90% or higher when confirmed by SDS polyacrylamide electrophoresis.

Labeling of V_H and V_L with Fluorochrome: 285 μg of purified V_H (0.5 ml ) was dialyzed with 0.2 M sodium phosphate (pH7.0) and 0.1 M

NaCI. 4.5 μ1 of I0 mM (about 5mg/ml) fluorescein succinimide ester (Molecular Probes Inc., Eugene, USA) in dimethyl sulfoxide was added to this dialyzed substance. The mixture was stirred sufficiently and allowed to react at 4°C for 10 hours. Concurrently, 4 μ1 of I0 mM (about 5mg/ml Rhodamine X succinimide ester (Molecular Probes Inc., Eugene, USA) in dimethyl sulfoxide was added to 235 μg of purified V_L (0.5ml) dialyzed into the same buffer. The mixture was stirred sufficiently and allowed to react at 4°C for 10 hours. The molar ratio of protein to the dye was 1 :2. The equivalent volume of 1 M Tris-HCL (pH7. 5) was added to terminate the reaction. The mixture was placed to gel filtration in PD-10 column (Pharmacia Co. ) equilibrated with 0.2 M sodium phosphate (pH7.0) and 0 1 M NaCI to remove the dyes which had not been reacted.

Measurement of HEL Concentrations:

20 μ1 of the fluorescein-labeled V_H and 25 μ1 of the Rhodamine X-labeled V_L were added to a solution composed of 0.2 M sodium phosphate (pH7.0), O.I M NaCI. and 1% bovine serum albumin. With the mixture placed into a cuvette and by using HITACHI Type 850 Spectrofluorophotometer, emission spectrum was measured at 4°C for the fluorescence wavelength ranging from 500 nm to 650 nm with 490 nm excitation. With continuous stirring, HEL was added to the mixture so that its final concentration was increased successively from I0 ng/ml to I mg/ml, and the changes in emission spectra were recorded.

As a result, it was observed that with increasing HEL concentrations, the fluorescence peak intensity at 530 nm decreased, and the energy transfer-derived fluorescence peak intensity at 603 nm increased. The time required for the change to occur was within a few seconds. Three independent experiments indicated that, as illustrated in FIG. 5, it became possible to measure antigen HEL concentrations reproducibly by taking the changed values in the ratio of fluorescence peak intensity between 620 nm and 530 nm.

Example 5: Construction of fusion peptides with a malate dehydrogenase effector

In this example, a pair of fusion polypeptides is obtained that have enzymatic effector sequences based on mitochondrial malate dehydrogenase. Three-dimensional computer modeling was performed using known amino acid sequences, and X-ray crystallographic data available from the Brookhaven database. The sequences of the heavy and light chain variable regions of monoclonal antibody HyHEL-10 was imposed on the crystal structure of the intact Fv fragment. Various candidate enzymes with homologous or heterologous subunits were reviewed to determine whether the distance spanned by the ends of the polypeptide chains matched the distance spanned by the two C-terminal amino acids of the Fv fragment. Results are shown in Table 1 :

TABLE 1: inxyme ca ttj es

ΕϊiZfϊϊϊβ Insufficient data for 34> prediction: bβtweβt* ter fei unsuitable

Triose phosphate tsomerase

Liver ateeSio? dehydrogenase X iscciϊrate δhy rogeriase X

Superoκlde dls utask x

TftymJ yϊate syni>as$

HIV protease

Hexok ase

Glucose αxidass X (80 Angstroms)

Isopropyt a atδ defiyc&ogenas©

Alkaline phosphatase

Horse radish peroxϊdase

DihydroSpc Sa te dehydrøgenase

8-phøsρhogftJccnate reduetase X 3ltita$hto*ι® rerjuclase

Trimefftyiamine tiehyrjrogenase

W& Ki s®

Enoiase X

Aspartatø <tehydrogenase

Preferred candidate

UHttβWttUftβ

Malate dehydrogenase has a number of advantages for this type of construct. The 3D structure is known to 1.87 Angstroms; it is a homodimer with a distance between termini suitable for fusing to V_H and V_L. The fused proteins can optionally be brought closer together, if necessary, by deletions of a few amino acids at the N termini of the enzyme or the C termini of the V_H and V_L, without any predicted loss in activity. Malate dehydrogenase can be used for sensitive assays with a detection limit in the picomolar range. It is not present in plasma or other biological fluids likely to be tested in a standard clinical assay. It is a proven label in other clinical chemistry technologies, and is stable. Mitochondrial malate dehydrogenase is allosterically regulated. Moreover, the mechanism of catalysis is understood, which should facilitate adaptation to other substrates where desirable.

FIG. 6 shows the predicted three-dimensional structure of the polypeptide backbone of a fusion polypeptide pair which is in the complexed configuration. The two solid lines show V_H and V_L domains (left and right) of the anti-HEL antibody. In the presence of the antigen (hen egg lysozyme), the domains are predicted to associate in the manner shown. The malate dehydrogenase homodimer structure is shown by the dashed lines. The structure has been rotated so that the two N-termini of the enzyme correspond with the C-termini of the variable region sequences. Fusion of the separate variable region domains each to a malate dehydrogenase subunit (with the possible removal of a few amino acids) can be done without distorting either structure or disturbing the interaction between subunits.

FIG. 7 shows the successful amplification of the mitochondrial malate dehydrogenase (MDH) encoding region from a cDNA library. PCR primers were prepared that hybridize to flanking sequences in the cloning vector. Track 1 (no band): cDNA prepared with cytoplasmic MDH-specific primers, amplified with mitochondrial MDH specific primers. Track 2 (~1 kb band): cDNA prepared with cytoplasmic MDH-specific primers, amplified with cytoplasmic MDH specific primers. Track 3 (no band): cDNA prepared with mitochondrial MDH-specific primers, amplified with cytoplasmic MDH specific primers. Track 4 (~1 kb band): cDNA prepared with mitochondrial MDH-specific primers, amplified with mitochondrial MDH specific primers. Tracks 6-8 (no bands): controls. Track 9 (ladder): molecular weight standards.

SEQ. ID NOS:7 and 8 provide the amino acid sequence and partial nucleic acid sequence of the heavy chain of HyHEL-10. SEQ. ID NOS:9 and 10 provide the amino acid sequence and nucleic acid sequence of the light chain of HyHEL-10. SEQ. ID NOS:11 and 12 provide the mouse MDH amino acid sequence and nucleic acid sequence. SEQ. ID NOS:13 and 14 provide the pig MDH amino acid sequence and nucleic acid sequence.

MDH variants are designed in which various amino acids at the MDH subunit interface are substituted so as to lessen the dimerization constant. The interface is readily identified from the structure shown in FIG. 6, and about 3 residues are changed in various combinations. Degree of association is determined by either direct molecular weight analysis, or by BIAcore™ binding, as outlined in Example 1. Candidates that do not dimerize on their own are then rescreened for dimerization in the presence of the substrate malate.

Recombinant polynucleotides are then prepared, in which L 108 of the light chain or His 116 of the heavy chain are attached to the N-terminal of candidate modified MDH sequences. The expressed fusion polypeptides are tested for the criteria of antigen-driven but not substrate-driven association, and the antigen-dependent ability of the fusion polypeptides to catalyze conversion of malate. Further iterations of sequence alteration and testing is undertaken as necessary that adjust the amino acids at the effector subunit interface or the linkage between the variable domain sequences and the effector sequences to optimize the properties of the polypeptide pair. REFERENCES

1. Davies, J. and Riechmann, L. 1994. "Camelising" human antibody fragments NMR studies on V_H domains. FEBS Letters 339:285-290.

2. Davies, J. and Riechmann, L. 1995. Antibody V_H domains as small recognition units. Bio/Technology 13:475-479.

3. Figlini, M., Marks, J.D., Winter, G., and Griffiths, A.D. 1994. In vitro assembly of repertoires of antibody chains on the surface of phage by renaturation. J. Mol. Biol. 239:68-78. 4. Glockshuber, R., Malia, M., Pfitzinger, I., and Plϋckthun, A. 1990. A comparison of strategies to stabilize immunoglobulin Fv-fragments. Biochemistry 29:1362-1367. 5. Hamel, P.A., Klein, M.H., Smith-Gill, S.J., and Dorrington, K.J. 1987. Relative noncovalent association constant between immunoglobulin H and L chains is unrelated to their expression of antigen-binding activity. J. Immunol. 139:3012-3020. 6. Hogan, J.K. et al. 1995. Improved specificity toward substrates with positively charged side chains by site-directed mutagenesis of the L-lactate dehydrogenase of Bacillus stearothermophilus. Biochemistry 34:4225-4230.

7. Huston, J.S. et al. 1988. Protein engineering of antibody binding sites: recovery of specific activity in an anti-digoxin single-chain Fv analogue produced in Escherichia coli. Proc. Natl. Acad. Sci USA 85:5879-5883.

8. Joh T.H. et al. 1987. Cloning and sequence analysis of cDNAs encoding mammalian mitochondrial malate dehydrogenase. Biochemistry 26:2515-2520.

9. Klein, M. 1979. Equilibrium and kinetic aspects of the interaction of isolated variable and constant domains of light chain with the Fd' fragment of immunoglobulin G. Biochemistry 18:1473-1481.

10. Lavoie, T.B., Drohan, W.N., and Smith-Gill, S.J. 1992. Experimental analysis by site-directed mutagenesis of somatic mutation effects on affinity and fine specificity in antibodies specific for lysozyme. J. Immunol. 148:503-513.

11. Lu, G. et al. 1997. Importance of the dimer-dimer interface for allosteric signal transduction and AMP cooperativity of pig kidney fructose-1 ,6-bisphosphatase. Site-specific mutagenesis studies of Glu-102 and Asp-187 residues on the 190's loop. J. Biol. Chem. 272:5076-5081.

12. Maekawa, M. et al. 1993. Detection and characterization of new genetic mutations in individuals heterozygous for lactate dehydrogenase-B(H) deficiency using DNA conformation polymorphism analysis and silver staining. Hum. Genet. 91:163-168. 13. Maeda, Y. et al. 1996. Chimeric antibody binding domain-Vargu/a luciferase engineered for immunological purposes. Protein Engineering 9:811-812. 14. Maenaka, K. et al. 1996. A stable phage-display system using a phagemid vector: phage display of hen egg-white lysozyme (HEL), Escherichia coli alkaline phosphatase, and anti-HEL monoclonal antibody, HyHEL-10. Biochem. Biophys. Res. Comm. 218:682-687.

15. Mainhart, S.G.S.J. et al. 1987. A three-dimensional model of an anti-lysozyme antibody. J. Mol. Biol. 194:713-724.

16. Miyazaki K. et al. 1994. Chemical modification and site-directed mutagenesis of Tyr36 of 3- isopropylmalate dehydrogenase from Thermus termophilus HB8. Protein Eng. 7:99-102, 1994.

17. Muyldermans, S., Atarhouch, T., Saldanha, J., Barbosa, J.A.R.G., and Hamers, R. 1994. Sequence and structure of V_H domain from naturally occuring camel heavy chain immunoglobulins lacking light chains. Protein Engineering 7:1129-1135.

18. Padlan, E.A. et al. 1989. Structure of an antibody-antigen complex: crystal structure of the HyHEL-10 Fab-lysozyme complex. Proc. Natl. Acad. Sci. USA 86:5938-5942.

19. Parsons H.L. et al. 1996. Directing phage selections towards specific epitopes. Protein Engineering 9:1043-1049. 20. Pharmacia Biosensor AB. 1994. BIAtechnology handbook. Merk AB, Uppsala, Sweden.

21. Truong, H.T. et al. 1991. Inactive and temperature-sensitive folding mutants generated by tryptophan substitutions in the membrane-bound d-lactate dehydrogenase of Escherichia coli. Biochemistry 30: 10722-10729.

22. Tsumoto, K. et al. 1994. Effect of the order of antibody variable regions on the expression of the single chain HyHELIO Fv fragment in E coli and the thermodynamic analysis of its antigen- binding properties. Biochem. Biophys. Res. Commun. 201:546-551.

23. Tyutyulkova, S. and Paul, S. 1994. Selection of functional human immunoglobulin light chains from a phage-display library. Appl. Biochem. Biotechnol. 47:191-198.

24. Ueda, H. et al. 1992. Antigen responsive antibody-receptor kinase chimera. Bio/Technology 10:430-433.

25. Ueda, H. et al. 1996. Analysis of intramolecular interaction of an antibody Fv region and its application to immunoassay [abstract]. Protein Engineering 9:819.

26. Ueda, Y., Tsumoto, K., Watanabe, K., and Kumagai, I. 1993. Synthesis and expression of a DNA encoding the Fv domain of an anti-lysozyme monoclonal antibody, HyHEL-10, in Streptomyces lividans, Gene 129129-134.

27. Wigley, D.B. et al. 1992. Structure of a ternary complex of an allosteric lactate dehydrogenase from Bacillus stearothermophilus at 2.5 A resolution. J. Mol. Biol. 223:317-335.

28. Wagner, C.R. et al. 1990. Site directed mutagenesis: a tool for enzyme mechanism detection. Trends Biotechnol. 1:263-270, 1990. 29. Ward, E.S. 1992 Expression and purification of antibody fragments using Escherichia coli as a host, pp. 121-138 in Antibody engineering: A practical guide. Borrebaeck, C.A.K. (ed.) W.H. Freeman & Co., New York. U.S. Patent 4,859,609 Dull et al. Hybrid receptors convenient for assay

U.S. Patent 5,030,576 Dull et al. Hybrid receptors convenient for assay

Additional references can be found at various places throughout the disclosure.

CONDENSED SEQUENCE LISTING

SEQ. ID N0:1 TTTGGCCCAGCCGGCCATGGCC

SEQ. ID N0:2 TTTGCGGCCGCCGCCGAGACGGTGACGAGGGT

SEQ. ID NO:3 CCGCGGCCGCGGGTACCCCAGAAATGCCTGTTCTAGAAA SEQ. ID N0:4

AAGCGGCCGCCTTAAGCCCCAGAGCGGC

SEQ. ID NO:5 TTAATACGACTCACTAT

SEQ. ID N0:6 TTTAAGCTTGGACTCACCCGCCGAGACGGTGACGAC

SEQ. ID N0:7

SEQ. ID N0:8

SEQ. ID N0:9 SEQ. ID N0:10

SEQ. ID NO: 11

SOURCE Mouse (strain C3H/He) liver, cDNA to mRNA, clone pmmHDH-1. HLSALARPAGAALRRSFSTSAQNNAKVAVLGASGGIGQPLSLLL

KNSPLVSRLTLYDIAHTPGVAADLSHIETRAKVKGYLGPEQLPDCLKGCDVVVIPAGV

PRKPGHTRDDLFNTNATIVATLTAACAQHCPEAMVCIIANPVNSTIPITAEVFKKHGV

YNPNKI FGVTTLDIVRANTFVAELKGLDPARVN PVIGGHAGKT11PLISQCTP VDF

PQDQLATLTGRIQEAGTEVVKAKAGAGSATLSMAYAGARFVFSLVDAMNGLEGWECS FVQSKETECTYFSTPLLLGKKGLEKNLGIGKITPFEEKHIAEAIPELKASIKKGEDFV NHK

SEQ. ID NO:12 1 ttcttgtagc tcctgccagt agctccgtgt cccgcccgcc ctagccatgc tgtccgctct

61 cgcccgtcct gccggcgccg ctctccgccg cagcttcagc acttcggccc agaacaatgc

121 taaagtggct gtcctgggag cttctggggg cattgggcaa cccctttcac tcctgctgaa

181 gaacagcccc ctagtgagcc gcctgaccct ctacgatatc gctcacacac ctggtgtggc

241 agcagatctg agtcacattg agaccagagc aaaggtgaaa ggctaccttg gaccggagca 301 gttgccagat tgcctcaaag gttgtgatgt ggtggtcatc ccagccggag tgcccaggaa

361 accaggaatg acacgggatg acctgttcaa caccaacgct accattgtgg ccaccctgac

421 ggctgcctgt gcccagcact gtcctgaagc catggtttgc atcattgcca acccagtgaa

481 ctccaccatc cccatcacag cagaagtttt caagaagcac ggtgtgtaca accctaacaa

541 gatcttcggt gtgacaaccc ttgacatcgt cagagcgaac acgtttgtgg cagagctaaa 601 gggtttggat ccagctcgag tcaacgtgcc tgtcattggc ggccacgccg ggaagacgat

661 catccccctg atctctcagt gtaccccgaa ggttgacttt ccccaagacc agctggccac

721 actcaccggg aggatccagg aggctggcac agaagtcgtg aaggccaagg ctggagcagg

781 ttctgccact ctgtccatgg cttatgctgg agcccgcttt gtcttctccc tcgtggacgc

841 catgaacggg ttggaaggag tcgttgagtg ttcttttgtt cagtccaaag agacggaatg 901 cacttacttc tctacgccct tgctcttggg gaaaaagggc ctggagaaga acctgggcat

961 tggcaagatc actccttttg aggaaaaaat gattgccgag gctatccctg agctgaaagc

1021 ctccatcaag aaaggcgagg actttgtcaa gaacatgaag tgagaggtgt gagcctcgag

1081 cagcagcagc agcagcatcc taacttattc agcatcatgt ctttggaacc acttgagaat

1141 ctagtttgcg ttgatggagg gtgttgagtc agcatcagca tctcttccaa attatgtctg 1201 gtctgttgat aatgacagta aagcaggctc tgattttctt tttc SEQ. ID NO: 13

SOURCE Pig liver, cDNA to mRNA, clone ppn+IDH-1.

SLLLKNSPLVSRLTLYDIAHTPGVAADLSHIETRATVKGYLGPE QLPDCLKGCDVWIPAGVPRKPGMTRDDLFNTNATMVATLTVACAQHCPDAMICIISN PVNSTIPMTAEVFKKHGVYNPNKI FGVTTLDIVRANAFVAELKGLDPARVSVPVIGGH AGKT11PLIS^QCTPKVDFPQDQLSTLTGRIQEAGTE VKAKAGAGSATLSHAYAGARF VFSLVDAMNGKEGVVECSFVKSQETDCPYFSTPLLLGKKGIEKNLRIGKISPFEE MI AEAIPELKASIKKGEEFVKNTK

SEQ. ID NO:14 tttcgcttct cctgaaaaac agccccctgg tgagccgcct gaccctctac gatatcgcgc

6 acacacccgg agtggccgcg gatctgagcc acatcgagac cagagcgact gtgaaaggct

12 acctcggacc tgagcagctg ccagactgcc tgaagggctg cgatgtggtg gttattccag

18 ccggagtccc aagaaagcca ggcatgacac gggatgacct gttcaacacc aacgccacca

24 tggtggccac cctgacggtc gcctgcgccc agcactgccc cgacgccatg atctgcatca

30 tttccaaccc ggttaactcc accatcccaa tgacggcgga ggtcttcaag aaacacggcg

36 tgtacaaccc caataaaatc ttcggggtga cgaccctgga cattgtccga gccaacgctt

42 ttgttgcaga gctgaagggt ttggacccgg ctcgagtcag cgttcccgtc attggcggcc

48 acgccgggaa gaccatcatc ccgctcatct ctcagtgcac cccgaaggtg gactttccgc

54 aggaccagct ctccaccctc accgggcgca tccaggaggc cggcaccgag gtggtcaagg

60 ctaaggccgg agcaggctct gccaccctgt ccatggcata tgccggagcc cggtttgtct

66 tctccctcgt ggatgcaatg aacgggaagg aaggcgttgt cgagtgttcc tttgtcaagt

72 cccaggaaac ggactgtccg tatttctcca cgccattgct gctggggaaa aagggcatcg

78 agaagaatct acgcatcggc aaaatctccc cttttgaaga gaagatgatc gccgaggcca

84 ttcctgagct gaaagcctcc atcaagaaag gagaggagtt tgtcaagaac acgaaatgag

90 caggcggttg gcgagcagtc cgcttcctta acttattcgg gcatcatgtc actgtaaagc

96 catttcagac ccctctgtct cctccctttg ctttggtgat gagtgtcctg tttacaaagc

102 accccttcca aatctcgggg catctctcgg tgcatttgta aagcaggctc tggtcttgtt

108 tttgagagtc cctccggttg aatactcggc tttcttcccc aacagagctg actgcagagt

114 gtccatgctg attcggaact agatgtgttt ccaaaataga cggaagcatg atggtctgtc

120 agtagcttca gatctcacac ctttatcagt aactgcttct cccaccctgc cctctgctcc

126 gtcgagtccc tcgggtgtgt gggctgaagg aagggctggc tgcatccgag agggtggcag

132 cccggtggcc aacgccaggc tgccgtctta tgttccaagc ttgtttctgt gtgtgttttc

138 tggccaaaca aacagcacca ttggctggat ttcgttgtct aattctgatc aagccctgaa

144 attcagtggc agtgacaatc cacagccaga agtgggagga aagctgggca gggcaggaca

150 gctgcaccaa aacaccagct gccaccgccg ggctgaagac atggggaagg ggtttcacgg

156 cctcatcagt ggcctgaggg gcattcgttg cattagctgg gcttacggta aatgtgagga

162 attacaggag acctgactcc cttagtgtca ggtgaaaaca gctccatact ggggaagaac

168 tttctggcca cgggtgatga ttga

Claims

CLAIMSWhat is claimed as the invention is:

1. A pair of fusion polypeptides that complex with each other in the presence of an antigen, consisting of: a) a first fusion polypeptide comprising a first variable domain sequence linked to a first effector sequence; b) a second fusion polypeptide comprising a second variable domain sequence linked to a second effector sequence; wherein presence of the antigen in a solution containing the fusion polypeptides promotes complexing between the first and second variable domain sequences; wherein the first and second effector sequences do not complex with each other in a solution containing antigen when not attached to the first and second variable domain sequences, respectively; and wherein complexing between the first and second variable domain sequences in the fusion polypeptides in the presence of the antigen results in complexing between the first and second effector sequences.

2. The pair of fusion polypeptides of claim 1 , wherein the first and second variable domain sequences are V_H and V_L sequences, respectively.

3. The pair of fusion polypeptides of claim 1 or claim 2, wherein the first and second effector sequences are a pair of enzyme fragments or a pair of toxin fragments.

4. The pair of fusion polypeptides of any of claims 1 to 3, wherein the first and second effector sequences are two identical or nonidentical fragments of an enzyme that complement to provide catalytic activity for conversion of a substrate to a product; and wherein conversion of the substrate to the product occurs more rapidly in a solution containing the two fusion polypeptides and the antigen, than in a solution containing the two fusion polypeptides but no antigen.

5. The pair of fusion polypeptides of any of claims 1 to 3, wherein the first and second effector sequences are two identical or nonidentical fragments of an enzyme that complement to provide catalytic activity for conversion of a substrate to a product; and wherein the first and second fusion peptides do not have the catalytic activity except when complexed with each other in the presence of the antigen.

6. The pair of fusion polypeptides of any of claims 1 to 3, wherein the first and second effector sequences are two identical or nonidentical fragments of an enzyme that complement to provide catalytic activity for conversion of a substrate to a product; and wherein the substrate does not promote complexing between the two enzyme fragments.

7. The pair of fusion polypeptides of any of claims 4 to 6, wherein the catalytic activity is selected from the group consisting of NDP kinase, enolase, aspartate dehydrogenase, and malate dehydrogenase.

8. The pair of fusion polypeptides of any preceding claim, wherein the first and second effector sequences are each independently at least about 80% identical to the monomer subunit of mitochondrial malate dehydrogenase.

9. The pair of fusion polypeptides of any of claims 1 to 8, wherein the first and second fusion polypeptides remain essentially separate in solution in the absence of analyte.

10. The pair of fusion polypeptides of any of claims 1 to 8, wherein the first and second fusion polypeptides are covalently tethered.

11. A method of preparing a pair of fusion polypeptides according to any preceding claim, comprising the steps of: a) selecting a first variable domain sequence and a second variable domain sequence wherein presence of the antigen in a solution promotes complexing between the variable domain sequences; b) selecting a first effector sequence and a second effector sequence that do not complex with each other in a solution containing the antigen; c) preparing a first fusion polypeptide in which the first variable domain sequence is linked to the first effector sequence, and a second fusion polypeptide in which the second variable domain sequence is linked to a second effector sequence; and d) confirming that the first fusion polypeptide forms a complex with the second fusion polypeptide that is stabilized in a solution if the solution contains the antigen.

12. A method of converting a substrate to a product in a manner that depends on the presence of an antigen, comprising the step of creating an environment that contains the antigen, the substrate, and a pair of fusion polypeptides according to any of claims 4 to 8.

13. A method of determining an antigen in a sample, comprising the steps of: a) preparing a reaction mixture containing the sample and a pair of fusion polypeptides according to any of claims 1 to 10; b) measuring complexes formed between the first effector sequence and the second effector sequence; and c) correlating the complexes measured in step b) with the presence, absence or amount of the antigen in the sample.

14. The method according to claim 13, wherein the first and second effector sequences are two fragments of an enzyme that complement to provide catalytic activity for conversion of a substrate to a product, wherein the reaction mixture prepared in step a) further comprises the substrate, and wherein the measuring in step b) comprises measuring the product formed in the reaction mixture.

15. A kit for determining an antigen in a sample according to claim 13 or claim 14, comprising a package containing each of the pair of fusion polypeptides according to any of claims 1 to 10.