WO1999012974A1 - Monoclonal antibodies that bind testosterone - Google Patents

Abstract

The present invention provides monoclonal antibodies and derivatives thereof suitable for use in immunodiagnostic assay of testosterone. In particular, such reagents are disclosed having affinity for testosterone of at least about 109 M-1 and exhibiting less than 0.05 % cross-reactivity with dehydroepiandrosterone sulphate (DHEAS), less than 10 % cross-reactivity with 5α-dihydrotestosterone (5α-DHT) and less than 5 % cross-reactivity with androstenedione determined on the basis of ED¿50? values.

Description

MONOCLONAL ANTIBODIES THAT BIND TESTOSTERONE

This invention relates to monoclonal antibodies and antibody engineering technology. More particularly, the present invention relates to monoclonal antibodies and derivatives thereof that bind testosterone with high affinity and specificity. The present invention also relates to processes for making and engineering such testosterone-binding monoclonal antibodies and to methods for using these antibodies and derivatives thereof in the field of immunodiagnostics enabling qualitative and quantitative determination of testosterone in biological samples.

Background of the invention

Determination of serum testosterone (TES) is important in the evaluation of hirsutism, virilism, infertility and other conditions associated with hyperandrogenism in women such as polycystic ovarian syndrome (Wheeler 1994, Wilke and Utley 1987, Osborn and Yannone, 1973). In men TES levels are used to diagnose reasons for delayed or premature puberty, hypogonadism, impotence and problems in spermatogenesis (Wang and SwerlofF 1992, Ismail et al, 1986, Wu et al, 1981, Odel and SwerdlofF, 1978). It has also been used to monitor hormone-replacement therapy (Ismail et al., 1986) and the treatment of patients with congenital adrenal hyperplasia (Korth-Schutz et al, 1978). Measurement of serum TES levels is also important to control the forbidden use of steroid hormones, e.g. TES, to increse the performance of athletes, both female and male, in sports.

The great number of closely related steroid structures and significant variation in their relative concentrations even between normal healthy individuals necessitates great qualitative performance of any measurement technique. This, together with high conservation of steroid structures in the animal kingdom, has made it extremely difficult to produce monoclonal anti- steroid antibodies with sufficiently high specificity and affinity for clinically useful routine immunoassays. The majority of commercially available diagnostic test kits for steroid hormones utilise polyclonal antibodies, not because of optimal performance and product quality, but because of apparent lack of monoclonal antibodies good enough to fulfil the clinical requirements. Even in a good antiserum, over 90% of the Ig molecules have little or no affinity for the antigen, and the "specific antibodies" themselves represent a whole spectrum of molecules with different affinities directed against different determinants on the antigen. The large amount of non-specific relative to antigen-specific Ig in an antiserum means that unspecific binding in any given immunological test may be uncomfortably high The continuous supply of good polyclonal antibody reagents of uniform quality is a severe problem for the immunodiagnostic industry and requires continuous immunisation of large numbers of laboratory animals These drawbacks of the currently used polyclonal anti-testosterone antibodies have motivated the search for more specific reagents for quantitative TES determination

In contrast to polyclonal antiserum, all antibody molecules produced by a given hybridoma are identical having identical structure and binding properties In vitro cell culture production of these monoclonal antibodies provides an unending supply of a single standard material Today, monoclonal antibodies and binding fragments can also be produced by recombinant methods Microbial expression offers the means for efficient production and rapid engineering of antibody fragments for different specific applications (Better et al, 1988, Skerra and Plϋckthun 1988) The possibility of displaying active antibody fragments on the surface of a filamentous bacteriophage provides a powerful tool for selection of specific or improved binders from a heterogeneous mixture of antibody fragments (McCafferty et al , 1990, Barbas et al, 1991, Marks et al, 1991). In the absence of detailed structural data of the interactions involved in antigen or hapten binding, the in vitro maturation strategy, where an initially selected antibody fragment is subjected to mutagenesis and improved variants are selected from large mutant libraries, has proven to be a very efficient method Random mutagenesis of the whole variable regions is possible by a number of different approaches such as chemical mutagenesis (Myers et al , 1985), polymerase-induced mutagenesis (Leung et al , 1989) and in vivo mutagenesis using mutator strains of E.coh (Schaaper, 1988) Focused mutagenesis, in which several residues are targeted, allows all possible mutations in a defined region to be explored For antibodies, this strategy involves targeting all or some of the three complementarity determining regions (CDRs) present both in the light and heavy chain and residues immediately flanking these regions to mutagenesis since these most likely have effect on the binding properties (Berek and Milstein, 1987) Numerous recent publications describe successful use of random mutagenesis of CDRs or of whole variable domain encoding gene regions of the heavy and light chain combined to phage display library selection to isolate antibodies having greatly improved affinity towards different protein antigens (Barbas et al , 1994, Yang et al , 1995, Schier et al , 1996a,b,c), haptens or carbohydrates (Hawkins et al , 1992, Deng et al , 1995, Short et al , 1995, Yelton et α/ , 1995)

Interactions of the CDR3 hypervariable loops of the light and heavy chains with small antigens are known to dominate the overall binding interactions (Wilson and Stanfield, 1994) The molecular basis of steroid antibody interactions has been studied in detail by determining the 3D structures of a progesterone binding antibody (DB3) in its free and complex form and with different high- affinity cross-reactive progesterone derivatives (Arevalo et al , 1993a,b, Arevalo et al, 1994) The progesterone molecule binding site is a hydrophobic pocket in which binding specificity is determined by the van der Waals interactions and the few hydrogen bonds formed with the ligand (Arevalo et al , 1993a) The high-affinity binding mode of a number of progesterone derivatives is due to two alternative binding sites for the A ring of the steroid skeleton (Arevalo et al , 1993 b, 1994) The restricted number or complete lack of direct interactions between antibodies and small, hydrophobic and rigid ligand molecules, is further demonstrated by the 3D structure of the anti-digoxin antibody (26-10) in the complex form (Jeffrey et al , 1993) The high-affinity binding mode to digoxin (1 x 10¹⁰ M^"1 ) is solely determined by hydrophobic shape complementarity, no direct interactions are formed between the antibody and the ligand (Jeffrey et al , 1993, Schildbach et al , 1994) The limited number of the direct interactions in the combining sites of the steroid hormone binding antibodies makes the improvement of these binding interactions highly demanding

Another possible approach to developing high-affinity, high-specificity steroid hormone binding antibodies in vitro has been published recently Pope et al. (1996) isolated from a large, naive human antibody phage display library a single chain antibody (scFv) binding oestradiol with relatively high specificity The affinity of this antibody to oestradiol was 3 7 nM and it had over 1000-fold selectivity in binding to other steroids However, the performance of this scFv antibody in a clinical oestradiol immunoassay was not evaluated in the publication

The present invention makes use of the unique property of monoclonal antibodies to bind their antigens with high affinity and specificity Monoclonal antibodies binding TES with high affinity and fine specificity were discovered in a stepwise process of engineering anti-testosterone antibody originally derived from a hybridoma cell line by random mutagenesis of the CDRs and phage display selection We have previously published preliminary data on the binding properties of some moderately improved anti-testosterone Fab fragments selected at the early stage of this work from the randomised CDR-libraries of the TES-binding 3-C^ Fab fragment. (Abstract: Fine tuning of an antibody binding site by mutagenesis and phage display, Exploring and Exploiting Antibody and Ig Superfamily Combining Sites, Keystone Symposia 5 on Molecular and Cellular Biology, Taos, New Mexico February 22-28, 1996 and Seventh Annual International Conference on Antibody Engineering, December 2-4, 1996, Coronado, CA; Abstract: Can man improve evolution ? In vitro affinity and specificity maturation of an antibody binding site, Ke ian Paivat, November 12-14, 1996, Helsinki; also poster presentation at INPEC-97, International Network of Protein Engineering Centres Annual lθ Meeting, Lillenhammer, Norway, January 26-28, 1997). These Fab fragments, with too low specificity and affinity for diagnostic immunoassay, were selected using panning methods still under optimization and also the sequence diversity of the CDR mutant libraries, especially the CDR3 libraries, was improved upon later. These Fab fragments are described in greater detail in Examples 1,2 and 3. It is shown in specific examples of the present invention that the

15 monoclonal antibodies and derivatives thereof produced by genetic engineering techniques bind TES with high affinity and specificity and can be successfully used as reagents in clinical immunoassays designed for the quantitative measurement of TES. To our knowledge, monoclonal antibody fragments herein described are the first shown to fulfil the tight qualitative requirements of a diagnostic TES immunoassay for clinical use. Because of these favourable

20 effects, new, economical and high-sensitivity diagnostic assay methods of uniform quality can be set up.

Objects of the Invention

25 One object of the present invention is to provide monoclonal antibodies or fragments or other derivatives thereof, that bind testosterone with affinity and specificity high enough to allow qualitative and quantitative measurement of testosterone in clinical samples. The antibodies of the present invention demonstrate an affinity of at least about 10⁹M^_I for testosterone, less than 0.05% cross-reactivity with dehydroepiandrosterone sulfate (DHEAS), less than 10% cross-

30 reactivity with 5 -dihydrotestosterone (5α-DHT) and less than 5% cross-reactivity with androstenedione. Peferably, such antibodies demonstrate less than 0.01% cross-reactivity with DHEAS. It is another object of the present invention to provide cDNA clones encoding testosterone- specific antibody chains, as well as constructs and methods for expression of such clones to produce testosterone-binding monoclonal antibodies or fragments or other derivatives thereof.

It is a fiαrther object of this invention to provide methods of using such testosterone-binding monoclonal antibodies or fragments or other derivatives thereof or combinations of them for qualitative and quantitative measurement of testosterone in clinical samples.

Other objects, features and advantages of the present invention will become apparent from the following figures and detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

Brief Description of the Figures

The figures of the constructions are not in scale.

Fig. 1. shows the schematic presentation of an intact murine IgGi subclass antibody, Fab fragment and single chain antibody (scFv). The antigen binding site is indicated by a triangle.

Fig. 2. shows the deduced amino acid sequence of the light chain variable region of the 3-C F5 antibody. CDRs are underlined. Numbering is according to Kabat (Kabat et al, 1991).

Fig. 3. shows the deduced amino acid sequence of the heavy chain variable region of the 3-CφF5 antibody. CDRs are underlined. Numbering is according to Kabat (Kabat et al, 1991).

Fig. 4. shows the schematic presentation of the E.coli expression vector pKKtoc used for the production of soluble TES-binding Fab and scFv fragments.

Fig. 5. shows schematically the competitive specificity panning procedure.

Fig. 6. shows the deduced amino acid sequences of the light and heavy chain CDR3 loops of the wild-type and selected mutant clones (numbering is according to Kabat et al, 1991). Only amino acid changes are indicated for the mutants.

Fig. 7. shows schematically the affinity panning procedure. Fig 8 shows the deduced amino acid sequences of the LCDR1, LCDR2 and HCDR1 loops of the monclonal antibody 3-G_tFs and selected mutant clones (numbeπng is according to Kabat et al , 1991) Only amino acid changes are indicated for the mutants Fig 9 shows the standard curve of a competitive one-step fluoroimmunoassay for TES utilising monoclonal TES-binding antibody Fab fragment The standard curve of a commercial DELFIA® Testosterone kit is shown for compaπsion

Fig 10 shows the correlation of the results obtained by using TES-binding Fab (A60/HCDR1/LCDR2) in a competitive one-step fluoroimmunoassay with the results obtained by gas chromatography - mass spectrometry analysis Fig 11 shows the deduced amino acid sequences of the re-optimised CDR3 loops (numbeπng is according to Kabayt et al , 1991) Only amino acid changes are indicated for the mutants Fig 12 shows the amino acid sequences of the V_L regions of Fab fragments S73, S77, S83 and S88 according to the invention as referred to in Example 4 aligned with the amino acid sequence of the V_L region of monclonal antibody 3-C^s (WT-V_L) [character to show that a position in the alignment is prefectly conserved ^*, character to show that a position is well- conserved *, CDRs underlined in the WT-V_L sequence ]

Fig 13 shows the amino acid sequences of the V_H regions of Fab fragments S73, S77, S83 and S88 aligned with the amino acid sequence of the V_H region of monclonal antibody 3-C4F5 (WT- N_H) [character to show that a position in the alignment is perfectly conserved ^*, character to show that a position is well-conserved *, CDRs underlined in the WR-V_H sequence ]

Fig 14 shows the values obtained from 48 female patient samples using either Fab fragments in the competitive one-step immunoassay protocol or a commercial kit designed for in vitro measurement of the concentration of TES in human serum or plasma Fig 15 shows the values obtained from 32 male patient samples using either Fab fragments in the competitive one-step immunoassay or a commercial TES kit

Fig 16 shows schematically the stepwise in vitro affinity and specificity maturation process of the 3-C F₅ Fab fragment

Summary of the Invention

Therre is now described the development and characterisation of monoclonal antibodies that bind TES with affinity and specificity high enough to be utilised as reagents in immunoassays designed for the quantitative measurement of TES in clinical samples Specifically, the present invention describes the development of hybridoma cell line 3-C4F5 , the cloning and random mutagenesis of the 3-QF₅ antibody Fab fragment encoding genes, the selection of highly improved variants of the 3-QF₅ by the phage display technique and the characterisation of the binding properties of the engineered Fab fragments produced in E. coli.

The usefulness of the developed TΕS-binding monoclonal recombinant Fab fragments was shown by comparing the optimised Fab fragments in the context of competitive immunoassay to a currently used commercial diagnostic kit, in which rabbit polyclonal antibodies are used as TΕS-specific reagents. By taking account of simultaneous improvements in affinity and specificity the overall functioning of the recombinant antibodies was improved considerably and achieved the utmost demands for cross-reactivities to fulfil the diagnostic requirements. This invention thus provides new reagents to be utilised in different kinds of immunoassay protocols. The invention also permits guaranteed continuous supply of these highly specific reagents of uniform quality eliminating inherent batch-to-batch variation of polyclonal antisera. These advantageous effects permit the manufacture of new, highly sensitive and economical immunodiagnostic assays of uniform quality.

Detailed Description

The following definitions are provided for some terms used throughout this specification.

The terms, "immunoglobulin", "heavy chain", "light chain" and "Fab" are used in the same way as in Εur. Pat. Appln. No. 0125023, of Cabilly et a , published November 14, 1984, corresponding to U.S. Serial No. 483,457, filed April 8 1983.

"Antibody" in its various grammatical forms is used herein as a collective noun that refers to a population of immunoglobulin molecules and/or immunologically active portions of immunoglobulin molecules, i.e., molecules that contain an "antigen binding site" or paratope.

An antigen binding site is the structural portion of an antibody molecule that specifically binds an antigen. o

Exemplary antibodies are intact immunoglobulin molecules, substantially intact immunoglobulin molecules and those portions of an immunoglobulin molecule that contain the paratope, including those portions known as Fab, Fab', F(ab')₂ and F_v.

Fab, and F(ab')₂ portions of antibodies can be prepared by the proteolytic reaction of papain and pepsin, respectively, on substantially intact antibodies by methods that are well known. See for example, U.S. Patent No. 4, 342,566 to Theofilopolous and Dixon. Fab' antibody portions are also well known and are produced from F(ab')₂ portions by reduction of the disulphide bonds linking the two heavy chain portions, as with mercaptoethanol, followed by alkylation of the resulting protein mercaptan with a reagent such as iodoacetamide. Antibodies and binding fragments can also be produced by recombinant methods, which are well known to those skilled in the art. See, for example, U.S. Patent 4, 949, 778 to Lander et al.

"Domain" is used to describe an independently folding part of a protein. General structural definitions for domain borders in natural proteins are given in Argos, 1988 (Argos, Protein Engineering, 2:101-113,1988).

A "variable domain" or "Fv" is used to describe those regions of the immunoglobulin molecule which are responsible for antigen or hapten binding. Usually these consist of approximately the first 100 amino acids of the N-termini of the light and the heavy chain of the immunoglobulin molecule.

"Single chain antibody" (scFv) is used to define a molecule in which the variable domains of the heavy and light chain of an antibody are joined together via a linker peptide to form a con- tinuous amino acid chain synthesised from a single rnRNA molecule (transcript).

"Linker" or "linker peptide" is used to describe an amino acid sequence that extends between adjacent domains in a natural or engineered protein.

A "TES-binding antibody" is an antibody which specifically recognises and binds to TES due to interactions mediated by its variable domains. As hereinbefore indicated, antibodies of the present invention demonstrate an affinity for testosterone of at least about 10⁹M', less than 0.05% cross-reactivity with dehydroepiandrosterone sulfate (DHEAS), less than 10% cross-reactivity with 5a- dihydrotestosterone (5 -DHT) and less than 5% cross-reactivity with androstenedione Preferably, such antibodies demonstrate less than 0.01% cross-reactivity with DHEAS, for example, less than 0.01% cross-reactivity with DHEAS, less than 8% cross-reactivity with 5α- DHT and no more than 2.5% cross-reactivity with androstenedione.

As examples of fragments of such antibodies falling within the scope of the invention, there are herein disclosed Fab fragments S73, S77, S83 and S88 derived from the 3-C4F5 Fab fragment by substitution of the heavy chain constant domain and mutation of the V_L and V_H regions as shown in Figures 12 and 13. In one preferred embodiment, the present invention thus provides testosterone-binding monoclonal antibodies and derivatives thereof, e.g. Fab fragments or ScFv antibody derivatives, wherein the CDRs of the V_L region are selected from the V_L region CDRs of the Fab fragments S73, S77, S83 and S88 and the CDRs of the V_H region are selected from the V_H region CDRs of the same antibody fragments. Such a monoclonal antibody or antibody derivative may have V_L and V_H regions corresponding to complete V_L and V_H regions selected from those regions of the Fab fragments S73, S77 S83 and S88 and additionally the V_H region of Fab fragment S77 in which the C-terminal Serine is substituted by Alanine. It will be appreciated that mutant versions of the same CDR sequences or complete V_L and V_H sequences having one or more conservative substitutions which do no substantially affect binding capability may alternatively be employed. The invention also encompasses fragments and extensions of the Fab fragments S73, S77 S83 and S88 which retain high affinity and good specificity for testosterone as hereinbefore defined.

An antibody or derivative thereof according to the invention may be labelled with any detectable label known in the art. This may include for example a radiochemical label, an enzyme label, a fluorescent label, a chemical label or a lanthanide label. The enzyme label may be alkaline phosphatase; the flourescent label may be fluorescein or a rhodamine; the chemical label may be biotin (which may be detected by avidin or streptavidin conjugated to peroxidase); the lanthanide label may be europium. These labels are given by way of example only. Furthermore, the antibodies and derivatives thereof according to the invention may be attached to or coated upon a solid support for example the walls of a micro-titre plate or filter paper.

Antibodies and derivatives thereof according to the invention have a variety of applications. For example, use of antibodies for in vivo diagnosis, imaging and therapy currently attracts considerable interest from the pharmaceutical industry. In in vivo therapeutic use low immunogenicity (small size) and high affinity for antigen are typically the main requirements for the development of antibody characteristics, thus an antibody or derivative thereof according to the invention is especially suitable for in vivo therapeutic applications. Another potential application is in immunoassays, both non-competitive and competitive, especially for the qualitative or quantitative determination of testosterone in clinical samples. Immunoassay methods include radio immunoassay, immunoradiometric assay, enzyme immunoassay, immunoenzymometric assay, time-resolved fluoroimmunoassay, immunofluorometric assay, chemiluminescence immunoassay, anodic or cathodic electrochemiluminescene immunoassays, homogenic immunoassays such as proximity assays with two different labels and particle immunoassays.

The reagents needed to perform the above assays according to the invention can conveniently be packaged into a kit. Thus, the invention extends to a kit comprising an antibody or derivative according to the invention in a suitable container for storage and transport. The kit preferably comprises the antibody or derivative thereof according to the invention attached to or coated upon a solid support.

In another aspect, the present invention also provides DNAs encoding an antibody or antibody derivative of the invention and fragments of such DNAs which encode the CDRs of the V_L and/or V_H region. Such a DNA may be in the form of a vector, more particularly, for example, an expression vector which is capable of directing expression of an antibody or antibody derivative of the invention or at least one component antibody chain.

In a further aspect of the invention, there are provided host cells, which may be any of bacterial cells, yeast cells, fungal cells, insect cells, plant cells, and mammalian cells containing a DNA of the invention, including host cells capable of expressing an antibody or antibody derivative of the invention. Thus, an antibody or antibody derivative of the invention may be prepared by culturing a host cell or host cells of the invention expressing the required antibody chain(s) and either directly recovering the desired protein or, if necessary, initially recovering and combining component chains

As hereinbefore indicated, Fab fragments S73, S77 S83 and S88 were obtained by an in vitro maturation procedure starting from the anti-testosterone monoclonal antibody 3-QF5 The hybridoma cell line producing monoclonal antibody 3-QFs was developed by hyperimmunizing mice with a TES-3-carboxymethyloxime(CMO)-thyroglobulin conjugate 3-GtF₅ has a relatively high affinity (0 3xl0⁹ M^"1) for TES and has a reasonable low cross-reactions (~11 6% and ~1%) with 5 -dihydrotestosterone (5α-DHT) and dehydroepiandrosterone sulfate (DHEAS)

The murine monoclonal anti-testosterone antibody, clone 3-QFs, IgG_2b, K, was purified from the hybridoma cell culture supernatant on a Protein A-Sepharose (Pharmacia, Sweden) The proteolytic production of the Fab fragment was carried out using a minor modification of the method described by Parham et al. (1982) The TES affinity and DHEAS cross-reactivities of the proteolytic Fab fragment were comparable to those of the original monoclonal 3-QFs antibody.

The N-terminal amino acid sequences of clone 3-G_tF₅ were determined and complementary DNAs of the light and heavy chains were synthesised from mRNAs isolated from hybridoma cells The light and heavy chain cDNAs were amplified by PCR, cloned using the restriction sites introduced into the PCR primers and sequenced (Figs 2 and 3) The cloned antibody genes were modified by PCR and the assembled Fab fragment expression unit was cloned under the tac promoter in the pKKtαc vector (Takkinen et al , 1991, Alfthan et al , 1993) The first constant domain of the heavy chain (IgG _b) was replaced by a gene fragment encoding the first constant domain of the IgGi subclass heavy chain to promote secretion and folding of the Fab fragment in E.coh (Alfthan et al , 1993, MacKenzie et al , 1994)

The Fab fragment expression unit in the pKKtαc vector was transformed into the E.coh RV308 strain The Fab fragment was secreted as an active and soluble protein into the periplasmic space and was released to the culture medium during the overnight inductions The 3-C4F5 Fab fragment was purified from the culture supernatant in a single step by a TES affinity column Binding characteristics of the parental mouse monoclonal anti-testosterone antibody 3- F₅ (IgG_2b), its proteolytic Fab fragment and E.co/z-derived recombinant Fab fragment for TES and DHEAS were characterised by a competitive two-step time-resolved fluoroimmunoassay as described in Example 1 The TES affinity and DHEAS cross-reactivity of the recombinant Fab fragment were comparable to those of the original monoclonal 3-QFs antibody and proteolytic Fab fragment. The heavy chain constant domain subclass replacement did not affect the binding properties of the recombinant Fab fragment

A molecular model was constructed for the 3-CtFs Fab fragment as described in Example 2. According to the model most of the contacts between TES and the antibody originate from the third CDR loops of the heavy and the light chains Thus, these CDRs were targeted for mutagenesis by using spiked PCR primers (Hermes et al, 1989; 1990) The randomness of PCR primers was adjusted by using a defined, biased mixture of nucleotides during the oligonucleotide synthesis for CDRs (80% wild-type and 20 % equal mixture of three other nucleotides). The CDR3 mutant libraries were constructed by the overlapping PCR method and cloned in the separate pComb3 phagemid vectors (Barbas et al.,1991).

The mutant libraries were selected by phage display using a competitive panning procedure. The concentration of DHEAS was increased stepwise from 0.1 μM to 0.3 μM during seven panning cycles to achieve at least 50% inhibition of binding onto immobilised TES on each round of panning. By performing the last panning cycle without amplifying the sixth phage eluate πvE.coli cells, and eluting the binders with 100 μM soluble TES, the number of different clones was limited and the elution was done in a more specific way. After 7 rounds of panning, soluble Fab fragments were produced from isolated phages. Relative affinities and cross- reactivities of the selected mutant anti-testosterone Fab fragments were determined by the competitive time-resolved fluoroimmunoassay and compared to the recombinant wild-type Fab fragments produced either by the pKKtαc or the phagemid pComb3 vectors. In one light chain mutant the DHEAS cross-reactivity was decreased to 0.52% and in one heavy chain mutant to 0.09% By combining the light and heavy chain CDR3 mutations, the TES affinity was preserved on the wild-type level but the DHEAS cross-reactivity was further decreased to 0.03%) A surprising finding was that by the competitive panning procedure the overall binding specificity of the 3-CφFs antibody was refined, since the cross-reactivities to 5α-DHT and androstenedione were also significantly decreased in the combined mutant The combined CDR3 mutant clone was used as a template for a subsequent round of modification by targetting parallely the remaining CDRs. A higher frequency of mutations was allowed for these generally less important loops. The V_L and V_H CDRl and CDR2 loops (as defined in Kabat et al, 1991) of the combined CDR3 mutant Fab fragment of the 3-C4F5 antibody were targeted for mutagenesis by using spiked PCR primers. To restrict the frequency of amino acid mutations in single clones, the randomness of PCR primers was adjusted by nucleotide doping during the oligonucleotide synthesis for CDRs (using 62.5% wild-type and 12.5%) each three other nucleotides in oligo synthesis).

The mutant libraries were selected by the phage display technique. To improve the TES affinity the new mutant libraries were selected by using limiting, decreasing concentrations of biotinylated TES in solution and capturing the binders on streptavidin. To improve the specificity, mutant libraries were selected by preincubating the libraries in solution with high concentration of the cross-reacting steroid, DHEAS, and then isolating TES binders using microtiter plate wells coated with a TES conjugate. In two different light chain CDRl mutant clones isolated from the affinity pannings, the TES affinity was increased over tenfold while the cross-reactivities to related steroids were preserved on the same level as in the parental combined CDR3 mutant clone. New heavy chain CDRl and light chain CDR2 sequences were isolated from specificity selections showing decreased cross-reactivity to related steroids. By combining compatible mutant CDRs together a Fab fragment with 18-fold lower cross- reactivity to DHEAS and over 11-fold higher TES affinity compared to original monoclonal antibody (3-CφFs) was created. Evaluation of the two best mutant Fab fragments in a competitive one-step research immunoassay protocol with TES standards in serum showed good correlation with gas chromatography - mass spectrometry analysis.

Next the CDR3 loops of the combined mutant clone were retargetted in parallel by oligo- directed PCR mutagenesis. The bias of the oligonucleotides for the parental sequence was weaker than in the previous CDR3 mutant libraries (62.5%) parental type and 12.5%> each three other nucleotides) and the mutant libraries were cloned by utilising unique restriction sites created by site-directed mutagenesis nearby the CDRs thus improving the quality and size of the libraries. A combined selection procedure allowing simultaneous selection with respect to affinity and specificity was used in addition to the previously utilized panning technique. A high concentration of cross-reacting steroid as a protein conjugate (DHEA-3-SucH-BSA) was used in phage solutions to inhibit binding of cross-reacting phages to limiting concentrations of biotinylated TES. The contact time of phages and biotinylated TES was decreased to 10 min and the biotinylated TES was immobilised in the streptavidin wells beforehand to ensure complete immobilisation. The elution of phages was done using 50 mM NaOH, pH 12.6 for 15 min and immediate neutralisation with 1 M Tris, pH 7.5. The mutant libraries were also affinity selected without any competing steroid by using limiting, gradually decreasing concentrations of biotinylated TES to catch the high-affinity binders. The concentration of TES used in this approach was clearly lower (1 nM to 10 pM) compared to approaches where cross-reacting steroid was used.

After panning steps, about 50 individual clones were grown up from each library to produce soluble Fab fragments, which were analysed first on a competitive ELISA test in order to select clones having either decreased DHEAS cross-reactivity and/or improved TES affinity. The selected clones were cloned into the pKKtαc expression vector and transformed into the E.coli strain RV308 for small scale production and the Fab fragments were characterised by the competitive time-resolved fluoroimmunoassay. In the case of the best light chain CDR3 mutant clone isolated from the combined affinity/specificity panning approach, the DHEAS and 5α- DHT cross-reactivities were clearly decreased (3 -fold and 2-fold, respectively) while the TES affinity was preserved at the same level as in the parental mutant clone. The TES affinity of one light chain CDR3 mutant clone selected by affinity panning was improved 3 -fold but all cross- reactivities were increased.

To further improve the overall binding properties of the of the TES-binding antibody, new light chain CDR3 sequences were created by combining mutations from different light chain CDR3 mutant clones. Finally, a number of new CDR combinations were created by combining different CDR loop sequences selected during the stepwise optimisation of the CDRs of the 3- C^s antibody. The optimisation of the light chain CDR3 sequence and the CDR combination resulted in clones having an excellent overall binding profile. A TES-binding Fab fragment having mutations in five CDRs and showing 32-fold higher affinity to TES, 60-fold lower cross-reactivity to DHEAS, 2-fold lower cross-reactivity to 5α-DHT and over 9-fold lower cross-reactivity to androstenedione compared to the original monoclonal antibody was created. The binding properties of many mutant Fab fragments were now comparable to those of a rabbit polyclonal anti-testosterone antiserum good enough for diagnostic immunoassay. As described herein, the stepwise optimisation of the CDRs of a reasonably good hybridoma cell-derived monoclonal TES-binding antibody using random mutagenesis and phage display techniques and in the final step analysis of the additivity of different mutations and different CDRs is an efficient and feasible approach to develop recombinant monoclonal anti- testosterone antibodies for sensitive diagnostic applications.

While one successful in vitro maturation strategy for obtaining antibody fragments of the invention has been described, numerous variations whereby antibodies or antibody fragments of the invention may be obtained will be apparent to those skilled in the art. Depending upon the starting TES-binding antibody either one or both of the V_L and V_H regions may be mutated. It may prove possible to select an antibody fragment or ScFv of the invention directly from a phage or microbial display library of antibody fragments and/or ScFv antibody derivatives. A phage or microbial cell which presents an antibody fragment or ScFv or the invetnion as a fusion protein with a surface protein represents a still further aspect of the invention.

While microbial expression of antibodies and antibody derivatives of the invention offers means for efficient and economical production of uniform quality highly specific reagents suitable for use in immunodiagnostic assays, alternatively it may prove possible to produce such a reagent, or at least a portion thereof, synthetically. By applying conventional genetic engineering techniques, initially obtained antibodies or antibody fragments of the invention may be altered, e.g. new sequences linked, without substantially altering the binding characteristics. Such techniques may be employed to produce chimeric, humanised and CDR-grafted anti- testosterone antibodies as well as novel testosterone-binding hybrid proteins which retain both affinity and high specificity for testosterone as hereinbefore defined.

The development of the original TES-binding monoclonal antibody and the hybridoma cell line producing it, the development and characterisation of the highly improved TES-binding recombinant monoclonal antibodies, and their usefulness in immunoassays is now described in more detail in the following examples. EXAMPLE 1

AN ANTIBODY AND FAB FRAGMENTS SPECIFICALLY BINDING TO TESTOSTERONE

The anti-testosterone monoclonal antibody 3-C F₅ was developed by hyperimmunizing mice with a TES-3-carboxymethyloxime(CMO)-thyroglobulin conjugate. The 3-GtF₅ was purified from the hybridoma cell culture and proteolytic production of the Fab fragment was carried out. The genes of the 3-CφFs were cloned and the recombinant Fab fragment was produced in E.coh. Binding characteristics of the parental monoclonal antibody and Fab fragments were analysed by the competitive two-step time-resolved fluoroimmunoassay. Immunisation and screening procedures, cloning and characterisation of the anti -testosterone monoclonal antibody 3-C F₅ are described in this example.

I. Immunisation of mice and cell fusion

Male mice (8 weeks old) of BALB/c were used for immunisation. Cell fusion was according to the method of Galfre & Milstein (Galfre and Milstein, 1981).

The mice used in a B-cell fusion, leading to clone 3 -C₊Fs were immunised with TES-3- carboxymethyloxime(CMO)-thyroglobulin conjugate (70 μg) emulsified in the Complete Freund's Adjuvant (Difco, USA). The immunisation was administered intraperitoneally. The mice were then reimmunized six times with TES-3-CMO-thyroglobulin conjugate (70 μg) emulsified in the Incomplete Freund's Adjuvant (Difco) at four weeks' intervals. A few days after the last immunisation, the antibody response was tested. The mouse with the highest antibody response was chosen for cell fusion and given a booster injection (70 μg in saline) intravenously into the tail.

Three days after the booster injection, the spleen cells of the mouse were fused with mouse myeloma cells (X63-Ag8.653) by treating the cells with 50 % PEG 4000 in DMEM (Gibco, USA). Fusion cells were then grown in HAT (Gibco) medium with 20 % Fetal Calf Serum (Gibco). About three weeks after the fusion, the hybridoma cells had grown enough to be screened. A competitive RIA was used at the beginning as a fast screening method. The anti- body-producing cell lines were further cultured until they reached a cell density of 1 5 x IO⁶ cells/ml (estimated with hematocytometer and Trypan Blue stain) The cells were then cloned at least two times by the limiting dilution method in 96-well plates and between and after the clonings the cultures were tested with the competitive RIA The antibody-producing cell lines were further cultured, frozen, and stored in a liquid nitrogen.

II Purification of antibody binding to testosterone and preparation of proteolytic Fab fragment

The monoclonal 3-G_tF₅ antibody, IgG₂b, K, was purified from the hybridoma cell culture supernatant on a Protein A-Sepharose (Pharmacia, Sweden) column using the protocol of the manufacturer The proteolytic production of the Fab fragment was carried out using a minor modification of the method described by Parham et al (1982) Papain (Pharmacia) was preactivated before use with 50 mM cysteine in 100 mM Tris-HCl, 50 mM NaCl, 2 mM EDTA, pH 7 5 for 30 minutes at 37° C and was added to the purified antibody equilibrated in the same buffer without cysteine to yield an antibody. enzyme ratio (w/w) of 33 1. The reaction was stopped after 2 hours incubation at 37 °C by adding iodoacetamide to the final concentration of 5 mM. The Fc fragment and undigested complete antibody were immobilised on a Protein A- Sepharose column and the unbound Fab fragment in the effluent was gel filtrated and equilibrated in the phosphate-buffered-saline, pH 7 4 (PBS). The purity of the Fab fragment was analysed by SDS-polyacrylamide gel electrophoresis (Laemmli, 1970). The N-terminal amino acid sequences of the purified 3- Fs antibody were determined on an Applied Biosystems 477 A on-line 120 A modified pulsed liquid phase/gas phase sequencer as described by Baumann (1990)

III. Cloning and sequencing of the 3-CφFs Fd and L chain cDNAs

All basic recombinant DNA methods were done essentially as described (Sambrook et al , 1990) mRNA was isolated from ~5 x IO⁷ hybridoma cells using the FastTrack™ mRNA Isolation Kit (Invitrogen Co , San Diego, CA) according to manufacturer's protocol The first strand cDNA was synthesised from poly(A) RNA either with oligo-dT priming (for the light (L) chain) or with a primer complementary to the heavy (H) chain C_H2 domain coding region Polymerase chain reaction (PCR) was performed after the first strand cDNA synthesis (Orlandi et al , 1989) The V_H-5' and V_L-5' PCR primers were synthesised according to the N-terminal amino acid sequences of the 3-C^s H and L chains, respectively The C_Hl-3' and C_L-3' PCR primers were complementary to the conserved sequences in the 3' end of the CHI and CL regions of the heavy and light chains, respectively The restriction sites, introduced to the 5' end of oligonucleotides Xba I (V_L-5*), Hmdlll (C_L-3 Xba I (V_H-5') and Xho I (C_Hl-3') were used for cloning of the light and heavy chain cDNA fragments into the pSP73 vector (Promega Co , Madison, Wisconsin) The nucleotide sequences of the cloned L and Fd chains were determined from both strands of the double-stranded plasmid DNA by dideoxy DNA sequencing (Sanger et al , 1977) The deduced amino acid sequences are shown in Figs. 2 and 3

The cDNAs of the L and Fd chains were modified with PCR (Saiki et al , 1988) to provide them with restriction sites allowing precise in frame fusions in a Fab fragment expression unit in the pKKtαc vector as described earlier (Alfthan et al, 1993, Takkinen et al, 1991) The first constant domain of the heavy chain (IgG_2b) was replaced by a gene fragment encoding the first constant domain of the murine IgGi subclass heavy chain to promote secretion and folding of the Fab fragment in E.coh (Alfthan et al, 1993, MacKenzie et al, 1994) This replacement was done by utilising the unique BsaA I restriction site near the V_H-3 ' end and did not alter the amino acid sequence of the V_H The nucleotide sequence of the final expression construct (Fig 4) was verified by dideoxy sequencing.

IV Production and purification of the recombinant Fab fragment

The expression vector was transformed into the E.coh strain RV308 Overnight culture was diluted 1 50 and grown to an Asoo_nm of 0 7-1 5, after which isopropyl β-D- thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM and the growth was continued overnight at 30°C The cells were harvested by centrifugation (5000 g, 10 min ) and the culture supernatant was analysed on SDS-PAGE (Laemmli, 1970) followed by electrophoretic transfer of proteins onto a nitrocellulose filter (Towbin et al , 1979) The produced Fab fragment was reacted with a commercial alkaline phosphatase-conjugated goat anti-mouse kappa IgG (Southern Biotechnology Associates, Inc , USA) and detected using BCIP/NBT as substrate (Promega) For further characterisation, the E.coΛ-produced Fab fragment was purified from the culture supernatant in a single step by a TΕS affinity column Briefly, TΕS-3-O-(6-aminohexyl)oxime was coupled to vinylsulfone agarose (Kem-En-Tec, Denmark) according to the manufacturer's protocol and packed into a column The gel was saturated with BSA and washed with elution buffer (0.1 M citrate, pH 2.5) and equilibrated with PBS. The overnight supernatant of an induced culture was concentrated using the Pellicon Cassette System (Millipore, USA), applied to the column (1.2 cm x 3.0 cm) and the affinity chromatography was performed as described (Harlow and Lane, 1988). The eluted Fab was immediately neutralised with 1 M Tris-HCl buffer, pH 8.5. The pooled fractions were concentrated and then gel filtrated on an Econo-Pac 10DG column (Bio-Rad Laboratories, Richmond, CA) to change the buffer to PBS. The purified Fab fragment was analysed by SDS- PAGE. Protein concentrations were determined from the absorbance at 280 nm based on the Tip and Tyr content of the Fab protein sequence (Wetlaufer, 1962).

V. Characterisation of testosterone-binding antibodies

The relative affinities and cross-reactivities of the purified parental monoclonal antibody and its Fab fragments were determined by a competitive two-step time-resolved fluoroimmunoassay. Briefly, goat anti-mouse IgG (Fab-specific, Southern Biotechnology Associates, Inc., USA) was coated on microtiter plate wells. The wells were blocked and samples were incubated in the wells for 2 hours. After a washing step, dilution series of different steroids (in PBS) were added together with TES-3-CMO-polylysine labelled with an Europium-chelate (Wallac, Finland), and incubated for 1 hour. After fluorescense development with the DELFIA® enhancement solution (Wallac), the amount of bound label was measured with a 1230 Arcus fluorometer (Wallac). Dilution series of TES and DHEAS (in PBS) were used to analyse the relative affinities (ED-50 %> concentrations) for the corresponding steroids. DHEAS cross- reactivity was calculated from these values. The TES affinity and the DHEAS cross-reactivity of the proteolytic and recombinant Fab fragments were comparable to those of the original monoclonal 3-CφFs antibody (Table I). The heavy chain constant domain subclass replacement did not affect the binding properties of the recombinant Fab fragment.

Table I. Relative affinities and cross-reactivities of the anti-testosterone 3-C_AF_S antibody and its proteolytic and recombinant Fab fragments. ED₅₀ is the concentration of TES required to reduce binding by 50 percent. ED₅₀ (nM) Cross-reaction %

Antibody # TES DHEAS

IgG_2b (3-C₄F₅) 13 1.6

Fab prot. 14 2.0

Fab rec. 14 1.2

The heavy chain constant domain subclass replacement did not affect the binding properties of the recombinant Fab fragment.

Before the competitive immunoassay described above, the samples were titrated without soluble steroid in the same assay system in order to choose suitable dilutions for the antibody samples.

EXAMPLE 2

SPECIFICITY IMPROVEMENT OF THE RECOMBINANT ANTI-TESTOSTERONE FAB FRAGMENT BY CDR3 MUTAGENESIS AND PHAGE DISPLAY SELECTION

The constructed molecular model within the ligand TES was docked revealed that several amino acid residues in the third CDRs of the heavy and light chains were in close proximity to the ligand. The CDR3 loops were randomised using spiked synthetic oligonucleotides and mutant Fab libraries were selected by a competitive panning procedure. After panning Fab fragments with lowered DHEAS cross-reaction were isolated. The cross-reactivity was further decreased by combining the light and heavy chain CDR3 mutations.

I. Molecular modelling

The crystal structures of the immunoglobulin Fab fragments were extracted from the

Brookhaven Databank (Bernstein et al, 1977). For the identification of the best conserved sequences, the light and heavy chain sequences of 3-C4F₅ were aligned with the databank sequences using the computer program MALIGN (Johnson et al, 1993; Johnson and Overington, 1993). The two 3-D structures with the highest sequence similarity to either the H or the L chain of 3-C_tFs were selected for model building. The anti-oxazolone antibody (Alzari et al, 1990) has 72 % sequence identity with the heavy chain of 3-GtF₅ and the immunoglobulin 4-4-20 Fab fragment (Herron et al, 1989) has 90 % sequence identity with the light chain of the 3-GtF₅. Optimal sequence alignment of the complete Fab fragments of the anti-oxazolone antibody, the 4-4-20 and the 3-C₄F5 was obtained by MALIGN.

The computer program COMPOSER (Sali et al, 1990; Blundell et al, 1990) was used in building the structural model for the Fab fragment of 3- F5. Five out of the six CDRs were built in as canonical conformations. The CDR3 of the heavy chain does not have any canonical conformation available and therefore fragment that fitted best to the structural model and had highest sequence similarity with the sequence of the CDR3 of the V_H was selected from the database for that CDR3. First, the selected 3-D structures of the anti-oxazolone antibody and the 4-4-20 antibodies were simultaneously aligned and a rigid body superposition of their C_a backbones was performed in order to define the structurally conserved regions (SCRs) for the model. Next, the backbone of the 3-QF₅ SCRs was built and the side chain atoms were added using rules derived from the known 3-D structures and from the library of the most probable conformations of amino acid side chains (Sutcliffe et al, 1987). Finally, the structurally variable regions (SVRs) were built by selecting fragments from a database containing C_a - C_a distances generated from known protein structures of resolution better than 3.0 A. The final model was refined by energy minimising using CHARMm (Brooks et al, 1988).

The crystal structure of the ligand, TES, was extracted from the Cambridge Structural Database (Allen et al, 1991) and it was energy minimised. TES was then docked into the same position as the oxazolone hapten in the anti-oxazolone antibody. The initial position was selected from the basis that the size of the oxazolone hapten is very similar to that of TES and the anti-oxazolone antibody was one of the crystal structures used as a template for the model. The A-ring of TES must be solvent-exposed, because conjugated derivative of TES (conjugated through the A-ring C3 carbonyl oxygen atom) was used for immunisation. However, the TES molecule can be accommodated within the binding site in a number of different orientations of the ring system. Four different starting conformations of TES were used. Beside the initial conformation, where the methyl groups point towards the antibody light chain, three other starting conformations were used. In these cases, the ring system was rotated by 90, -90 and 180 degrees with respect to the initial conformation. Each complex was energy minimised and simulated 100 picoseconds at the room temperature by using a stochastic boundary molecular dynamics method, SBMD, (Brooks and Karplus, 1989). This method allows detailed study of interactions between TES and the antibody including water molecules and explicit hydrogen atoms into the simulation. The model of the complex in which TES had lowest interaction energy with the antibody was selected for further studies. Our earlier experience has shown that longer simulation time improves the quality of the model (Hoffren et al, 1992). Thus the low energy model was further simulated 400ps by SBMD in order to optimise the structure of the complex.

The molecular model was inspected by molecular graphics to identify the regions of the variable domains of the heavy and light chains which are in close proximity of the antigen binding site. The identified contact residues (<3.5 A distance from TES), most of them locating in the third hypervariable loops of the heavy and light chains, were in the light chain Gly91, His93, Val94, Pro96 and Phe98, and in the heavy chain Ser35, Trp47 Ser50, Tyr58, Glu95, Tyr99, Vail 00, GlylOOJ and LeulOOK (numbering is according to Kabat et al, 1991).

II. Construction and selection of the mutant libraries of the 3-C^^ antibody

The CDR3 loops of the light and heavy chains were targeted for oligonucleotide-directed random mutagenesis by using spiked nucleotides in PCR primer synthesis (Hermes et al, 1989; 1990). To restrict the number of amino acid mutations in single clones and thus to limit the number of the clones not retaining original structural features of the CDRs biased mixtures of nucleotides in the oligonucleotide synthesis for CDR3 loops (80% wild-type and 20 % equal mixture of three other nucleotides) were used. The CDR3 region mutant libraries of the light and heavy chains were cloned separately by the overlapping PCR method (Ho et al, 1989) into the wild-type Fab fragment expression unit cloned in the phagemid vector pComb3, kindly provided by Dr. C. Barbas (Barbas et al, 1991).

The light and heavy chain CDR3 mutant libraries were separately selected by the phage display technique (McCafferty et al, 1990, Barbas et al, 1991) using a competitive specificity panning procedure (Fig. 5). Briefly, the mutant Fab fragment libraries displayed on the surface of the bacteriophage were first incubated with soluble DHEAS for at least 30 minutes to inactivate those Fab phages having high cross-reactivity to DHEAS Thereafter, the phage pools were transferred to microtiter plate wells coated with the probe, TES-4-mercaptopropionic acid- BSA conjugate (1 μg/well), to catch the TES binders from the libraries After 1 hour incubation, the wells were washed 10 times during 1 hour period with washing solution (TBS- 1% BSA-0.5% Tween 20) and the binders were eluted with acidic buffer (0.1 M HCl - 1 mg/ml BSA, pH adjusted to 2.2. with glycine), and immediately neutralised with 2 M Tris solution. For the next panning rounds the eluted phage pools were amplified by infecting E.coh XL1- Blue cells. The inhibition effect of the soluble DHEAS was analysed during each round of selection by panning the same phage pools with and without DHEAS and titering eluted binders on selective plates. The concentration of DHEAS was adjusted to achieve at least 50% inhibition on each round of panning. Seven rounds of panning were performed, and during the cycles the concentration of DHEAS was increased stepwise from 0.1 μM to 0 3 μM. The seventh panning cycle was performed without amplifying the sixth phage eluate in E.coli XL1- Blue cells and on this cycle the binders were eluted with 100 μM TES in TBS-1% BSA for 1 hour.

rV. Characterisation of the mutant Fab fragments

After 7 rounds of panning soluble Fab fragments were produced from isolated clones as described (Barbas et al, 1991). A number of individual clones were picked and the binding properties of the recombinant Fab fragments were preliminarily characterised by a competitive ELISA. Briefly, microtiter plate wells were coated overnight with a TES-3-CMO-BSA- conjugate (2 μg/ml) and blocked by incubating for 1 hour at room temperature with 0.5%> BSA-PBS. Samples were added with dilution series of TES and DHEAS (in PBS), and after 2 hours incubation and a washing step (three times with PBS), goat anti-mouse IgG (κ-specific) labelled with alkaline phosphatase (Southern Biotechnology Associates, Inc., USA) was added and incubated for 1 hour. After washing, ?-nitrophenylphosphate in diethanolamine-MgCl₂ - buffer (Orion, Finland) was added and after an incubation period the absorbance in each well was read at 405nm.

Clones having decreased DHEAS cross-reactivity but still retaining the TES binding activity on the wild-type level were selected for further characterisation The most interesting ones were purified using the HiTrap protein G column (Pharmacia Biotech) using the protocol recommended by the manufacturer for purification of IgG molecules. Purified Fab fragments were characterised using microtiter plates coated with goat anti-mouse IgG (Fab-specific) and Europium-labelled TES-3-CMO-polylysine as fluorescent label (as described in Example 1). Dilution series of TES, DHEAS, 5α-DHT and androstenedione (all in PBS) were used to analyse the relative affinities (ED-50 % concentrations) for the corresponding steroids. Cross- reactivities were calculated from these values (Table I). The amino acid sequences of the variable regions of the interesting clones were deduced by dideoxy sequencing.

Table I. Affinities and cross-reactivities of the mutant anti-testosterone Fab fragments compared to the recombinant wild-type Fab fragments produced either by pKKtαc or pComb3 vectors.

ED₅₀ (nM) Cross-reaction %

rFab TES DHEAS 5α-DHT Androstenedione

wt Fab (pKKtαc) 6.7 1.214 11.9 8.3 wt Fab (pComb3) 13.8 1.463 15.0 6.7 mutant 44 9.7 0.085 5.4 1.8 mutant 28 18.3 0.524 2.5 12.9 combined (44+28) 9.7 0.033 2.4 1.4

Mutant clone 44 had two mutations in the heavy chain CDR3 (Fig. 6): E95A and A101V (Kabat et al, 1991). In this mutant clone the DHEAS cross-reactivity was decreased to 0.085 % while the TES affinity was retained almost at the same level as in the wild-type (wt) Fab fragment. Mutant clone 28 had one mutation in the light chain CDR3 (Fig. 6): H93T (Kabat et al, 1991). The DHEAS cross-reactivity was decreased to 0.524 %> but this mutation also reduced the TES binding activity three-fold compared to the wild-type produced from the pKKtαc vector. The ED₅₀ (TES) value of the wild-type Fab fragment produced from the pKKtαc vector was almost two times lower than that of the wild-type Fab fragment produced from the pComb3 vector (Table I). In the pComb3 vector the cloning sites of the cDNAs have been created within the very N-terminal regions of the heavy and light chain genes, resulting in changes of these sequences (two amino acid changes in each chain). This is the probable reason for the decreased affinity. The same phenomenon has also been noticed earlier with an anti- digoxin Fab fragment (Short et al, 1995). When the individual CDR3 mutations identified in the heavy chain (clone 44) and light chain (clone 28) were combined into the same Fab expression unit in the pComb3 vector, the specificity of the anti-testosterone Fab fragment was still improved by a factor of 2.6 (combined (44+28)). In this mutant the cross-reactivity of DHEAS was 0.033 %. Surprisingly, also the cross-reactivities to 5α-DHT and androstenedione were much lower than in the wild-type although these steroids were not used in the panning to inhibit cross-reactive phages. The TES ED₅₀ value of this combined mutant was only a bit higher than in the wild-type. When expressed from the pKKtαc vector with authentic N-terminal amino acids in the heavy and light chains, the affinity of the (44+28) mutant Fab was restored to the wild-type level.

EXAMPLE 3

CONSTRUCTION AND SELECTION OF THE CDRl AND CDR2 MUTANT LIBRARIES

In this example the mutagenesis strategy of the TES-binding Fab fragment was extended by mutating the other CDRs and selecting the mutant libraries using two different approaches in order to further improve the specificity and affinity. To improve the TES affinity, the new mutant libraries were selected by using limiting, decreasing concentrations of a biotinylated TES in solution and capturing the binders on streptavidin. To improve the specificity, the mutant libraries were selected by preincubating the libraries in solution with a high concentration of the cross-reacting steroid, DHEAS, and then isolating the TES binders using microtiter plate wells coated with a TES conjugate. By combining compatible mutant CDRs together, a Fab fragment with lowered cross-reactivity to DHEAS and raised TES affinity was created.

I. Construction of the mutant libraries

The V_L and V_H CDRl and CDR2 regions (as defined in Kabat et al, 1991) of the combined CDR3 mutant Fab fragment (mutant 44+28) (Example 2) of the 3-C4F5 antibody were targeted to mutagenesis by using spiked PCR primers (Hermes et al, 1989; 1990). To restrict the frequency of amino acid mutations in single clones and thus to limit the number of the clones not retaining original structural features of the CDR regions, the randomness of PCR primers was adjusted by nucleotide doping during the oligonucleotide synthesis for CDR regions (62.5%o wild-type and 12.5% each three other nucleotides).

The heavy chain CDR2 mutant library was cloned by the overlapping PCR method into the pComb3 vector containing the mutant Fab fragment (44+28) (Ho et al, 1989, Barbas et al, 1992). The three other libraries were constructed into the pComb3 vector (Barbas et /.,1991) containing the mutant Fab fragment (44+28) by utilising unique restriction sites nearby the CDRs. For this reason, the two amino terminal codons of the light chain CDRl containing a Bgl II restriction site used for the cloning of the light chain CDRl region were not randomised.

II. Selection of the mutant libraries

The mutant libraries were selected by the phage display technique (McCafferty et al, 1990, Barbas et a , 1991). To improve specificity, the libraries were panned using the same method as described earlier for the CDR3 mutant libraries of this 3-GjF₅ antibody (Example 2). Briefly, phage libraries were first incubated with a high concentration of soluble DHEAS in order to inactivate Fab phages with cross-reactivity to DHEAS. The phage pools were then transferred to microtiter plate wells coated with 1 μg of TES-3-CMO-BSA and the TES binders were eluted with an acidic buffer after washes. For the next panning rounds the eluted phage pools were amplified by infecting E.coli XL-1 Blue cells. The inhibition effect of the soluble DHEAS was analysed during each round of selection by panning the same phage pools with and without DHEAS inhibition and titering the eluted binders on selective plates. The concentration of DHEAS was adjusted to achieve at least 50%> inhibition on each round of panning. A total of six cycles were performed and during the cycles the concentration of DHEAS was increased stepwise from 0.15 mM to 10 mM.

To increase affinity, the libraries were selected by using limiting, gradually decreasing concentrations of a biotinylated hapten to catch the high-affinity binders (Fig. 7; Hawkins et al, 1992; Schier et al, 1996b). The libraries were first incubated with a biotinylated TES (TES -3 -biotin) for at least 30 minutes, then the phage pools were transferred into streptavidin- coated microtiter plate wells (Boehringer Mannheim GmbH, Germany) and incubated for 15 minutes. After washing the plate 10 times during a 1 hour period with the washing solution (TBS-1% BSA-0.5% Tween 20) the TES binders were eluted by adding the XLl-Blue cells (200 μl) into the wells together with high concentration of soluble TES (100 μM final). After 15 minutes infection the cells were transferred from the plate and amplified as usually. The amount of unspecific binding was analysed on each panning cycle by panning the same phage pools without adding the biotinylated TES in the first step, the binders eluted from streptavidin plates were titered on selective plates. This control panning was done because we noticed that the amount of unspecific binders eluted from the streptavidin-coated wells was quite high. The difference between the specific and unspecific binding was used to monitor the efficiency of the panning and to adjust the concentration of the biotinylated TES during the panning steps. Totally four rounds of panning were performed, and during the cycles the concentration of biotinylated TES was decreased gradually from 176 nM to 100 pM.

III. Characterisation of the mutants

After the last panning step phagemid DNA was isolated and the pin gene region removed by Spe I and Nhe I digestion from the pComb 3 vector (Barbas et al., 1991). The re-ligated vector DNA was transformed and about 100 individual clones were grown in a small scale to produce soluble Fab fragments for preliminary characterisation. Clones were analysed on a competitive ELISA test using TES-3-CMO-BSA-coated wells (0.1 μg) to catch the TES binders and soluble TES and DHEAS to achieve inhibition of binding (as described earlier, Example 2). The most promising clones having either decreased DHEAS cross-reactivity and/or improved TES binding affinity were sequenced (Sanger et al, 1977) and selected for further characterisation. These selected clones were cloned into the pKKtac expression vector (Takkinen et al, 1991). All possible different combinations of the isolated CDR regions were also collected into the pKKtαc vector, and transformed into the E.coli strain RV308 for small scale production. Clones were characterised using goat anti-mouse IgG (Fab-specific)-coated microtiter wells and TES-3-CMO-polylysine labelled with Europium-chelate (from Wallac) was used as the fluorescent label (Example 1). Dilution series of TES, DHEAS, 5 -DHT and androstenedione (all in PBS) were used to analyse the relative affinities (ED-50 %> concentrations) for the corresponding steroids. Cross-reactivities were calculated from these values (Tables I, II and III). The most interesting clones were produced in larger scale and purified using the HiTrap protein G columns (Pharmacia Biotech) following the same protocol as recommended by the manufacturer for purification of IgG samples.

TABLE I . Affinities and cross-reactivities of the anti-testosterone Fab fragments after specificity panning.

ED50 (nM) Cross-reaction %>

rFab TES DHEAS 5α-DHT Androstenedione

combined (44+28) 5.9 0.044 2.6 1.4

HCDRl (2B5) 7.6 0.031 1.4 1.2

LCDR2 (1D7) 3.1 0.021 3.4 1.6

HCDR1/LCDR2 4.8 0.036 2.5 1.4

The specificity panning resulted in moderate improvements (Table I). In the light chain CDR2 mutant clone 1D7, the DHEAS cross-reactivity was decreased from 0.044 % to 0.021 %> while the TES affinity was improved almost 2-fold. In the LCDR2 1D7 cross-reactivities to 5α-DHT and androstenedione were slightly increased. The heavy chain CDRl mutant 2B5 bound TES with slightly lower affinity than the parental combined CDR3 mutant clone 44+28 but still had decreased cross-reactivities to all other three steroids. Interestingly, when LCDR2 and HCDRl mutations were combined into the same Fab expression unit in the pKKtαc expression vector, the affinity and cross-reactivities of the Fab fragment were essentially the same as in the clone 44+28.

The light chain CDRl mutant clones A58 and A60 selected by affinity panning procedure showed about 10-fold improved affinity to TES while cross-reactivities to related steroids remained essentially unchanged (Table II). TABLE II Affinities and cross-reactivities of the anti-testosterone Fab fragments after affinity panning.

ED₅₀ (nM) Cross-reaction %>

rFab TES DHEAS 5α-DHT Androstenedione

combined (44+28) 6.4 0.037 4.2 1.7

LCDR1 (A58) 0.7 0.054 4.3 2.6 LCDR1 (A60) 0.5 0.040 3.4 2.1

By combining the affinity-selected CDRs (LCDR1: A58 and A60) with specificity-selected CDRs (HCDRl 2B5 and LCDR2 1D7) the specificity and affinity of the anti-testosterone Fab were further improved (Table III). In the best combined mutant, clone A60/HCDR1/LCDR2, the cross-reactivity to DHEAS was 0.024 % and also the cross-reactivities to 5α-DHT and androstenedione retained on the level of the parental Fab clone (44+28). The TES ED₅₀ value of this combined mutant was 11.6-fold lower than in the wild-type.

TABLE HI Affinities and cross-reactivities of the anti-testosterone Fab fragments after combining the best mutations found in specificity and affinity pannings.

ED₅₀ (nlv f> Cross-reaction %

rFab TES DHEAS 5α-DHT Androstenedione

wt Fab (pKKtαc) 5.8 0.435 11.0 8.1

A58 0.5 0.047 2.4 1.4

A58/HCDR1 0.9 0.041 5.0 2.1

A58/HCDR1/LCDR2 1.3 0.043 6.1 2.0

A60 0.6 0.055 6.3 2.1

A60/HCDR1 0.8 0.026 4.7 2.3

A60/HCDR1/LCDR2 0.5 0.024 4.8 1.7 The deduced protein sequences of the LCDR1 mutant clones A58 and A60 showed significant differences compared to the wild-type sequence (Fig. 8). In the LCDR2 mutant (1D7) four out of the seven amino acid residues have been mutated (V51A, S52R, N53Y and S56P) whereas the HCDRl mutant (2B5) had only one amino acid substitution (T31R).

IV. Kinetic measurements and affinity constant determinations by BIAcore™

The kinetics of binding of the purified mutant Fab fragments to the TES-3-CMO-BSA immobilised on a dextran-coated sensor chip were determined from the dependence of the surface plasmon resonance response upon the concentration of the purified Fab fragments injected into the biosensor (BIAcore™ , Pharmacia Biosensor AB, Sweden) (Karlsson et al, 1991; Jόnsson et al, 1992). The antigen, TES-3-CMO-BSA, was coupled to the CM5 sensor chip through its amine groups using the Amine Coupling Kit (Pharmacia Biosensor AB)(Johnsson et al, 1991). The antigen was diluted in 10 mM acetate buffer (pH 4.0) to the concentration of 6.25 μg/ml and 398 resonance units (RU) was immobilised using automated immobilisation cycle at a constant flow rate of 5 μl min as follows: 70 μl of 0.4 MN-ethyl-N'- (dimethylamino-propyl)carbodiimide (EDC) and 0.1 M N-hydroxysuccinimide (NHS) were transferred to the mixing vial, whereafter 35 μl of the mixed solution was injected over the surface. The antigen solution, 35 μl, was then injected, followed by injection of 35 μl of 1 M ethanolamine hydrochloride solution. Kinetic measurements were carried out at 25 °C in 10 mM HEPES complete buffer, pH 7.4 (HBS)(containing 3.4 mM Titriplex III, 0.15 M sodium chloride, and 0.05% surfactant P20 (Pharmacia)) with a constant flow rate of 5 μl/min and injected sample volume of 30 μl. In the measurements six different concentrations (between 12.5 - 400 nM) of the purified Fab fragments were used, and the antigen was regenerated with 10 μl of 50 mM NaOH, pH 12.6, between samples. The kinetic evaluation of data was performed using the BIAevaluation 2.1 software (Pharmacia Biosensor AB).

As shown in table IV, the most significant change in the rate constants was the about 20-fold slower dissociation rate of the best mutants compared to the wild-type. Changes in the association rate were smaller but together with dissociation rate improvements resulted in over 40-fold higher affinities of the best mutants to the TES conjugate compared to the wild-type Fab. TABLE IV. Binding kinetics and affinities of the wild-type and three CDR mutant library- derived, combined mutant Fab fragments. The k_<,_n and ko_ff rates for the Fab fragments were determined by BIAcore™. Affinity constants were calculated from these values.

Fab "on oir Iv_a— Kon'koff

M Vio⁵ s^'Vio-⁴ MVIO⁹

wt Fab 1.29 21.10 0.06 combined (44 + 28) 1.53 9.03 0.17

A58 HCDR1 3.12 1.14 2.74

A60/HCDR1/LCDR2 2.13 0.87 2.45

V. Evaluation of the A60/HCDR1/LCDR2 mutant Fab fragment in a competitive one-step fluoroimmunoassay

The improved sensitivity of the best mutant Fab fragment (A60/HCDR1/LCDR2) allowed use of serum based standards and a competitive one-step immunoassay protocol to test the performance of the mutant Fab in a clinically relevant TES concentration area. The developed competitive one-step immunoassay protocol was compared to a currently used commercial kit, in which rabbit polyclonal antibodies are used as TES specific reagents, designed for in vitro measurement of the concentration of TES in human serum or plasma (DELFIA® Testosterone kit, Time- resolved fluoroimmunoassay kit 1244-050 (Wallac) and to gas chromatography - mass spectrometry analysis.

First, suitable dilution for the A60/HCDR1/LCDR2 Fab (in 0.5% BSA/PBS) was titrated.

Briefly, 50 μl of TES-free serum, 100 μl of TES-Eu tracer (from the DELFIA® Testosterone kit, diluted 1:50 to Testosterone Assay Buffer from the same kit) and 100 μl of different dilutions of the Fab (in 0.5% BSA/PBS) were pipetted to microtiter plate wells precoated with goat anti-mouse IgG. After 2 hours incubation at room temperature with shaking, the wells were washed three times with DELFIA® wash solution and then 200 μl of DELFIA® Enhancement Solution were added to the wells. After fluorescence development at room temperature the fluorescence was measured by a 1230 Arcus fluorometer (Wallac). Suitable dilution of the Fab for subsequent TES measurements was chosen from a linear part of a curve representing the fluorescence as a function of a dilution of the Fab.

With the suitable dilution of the Fab and using different concentrations (0-50 nmol/1) of TES standards (in TES-free serum) unknown patient serum specimens were analysed using the same one-step assay protocol as described. After fluorescence measurement a standard curve, representing B/B_max (%>) as a function of TES concentration, was generated by the Multicalc program (Wallac) and TES concentrations of unknown samples were read from the curve using the same program. The same unknown patient serum specimens were analysed by using gas chromatography - mass spectrometry analysis and by the currently used commercial kit, in which rabbit polyclonal antibodies are used as TES specific reagents ( DELFIA® Testosterone kit, Time-resolved fluoroimmunoassay kit 1244-050, Wallac).

The standard curve for the modified research assay method utilizing the developed Fab fragment was essentially identical to that of the kit and covered the clinically relevant concentration range (Fig. 9). The lowest standard (0.5 nmol/1) gave 85.4%o and 84.5% (B/B_max %) binding for the A60/HCDR1/LCDR2 Fab and the kit, respectively. The highest standard (50 nmol/1) gave 11.8% and 10.2% binding for the A60/HCDR1/LCDR2 and the kit, respectively. The determined TES ED-50% values were 5.6 nmol 1 for the A60/HCDR1/LCDR2 and 4.3 nmol/1 for the kit. No proper binding inhibition of the parental wild-type monoclonal was achieved with clinically relevant standard concentrations in the modified research assay system used.

Figure 10. shows the correlation of the results obtained for unknown serum samples using the A60/HCDR1 LCDR2 Fab-based immunoassay to the results determined by gas chromatography - mass spectrometry analysis. Totally 28 samples were analysed in the range 0.501 to 31.851 ιτmol/1 with the mean value 9. A good correlation was found (r = 0.983, slope = 1.11). However, the obtained intercept value (= 5) revealed that although the overall binding specificity of the A60/HCDR1/LCDR2 Fab is significantly improved compared to the wild- type, it still has to be improved in order to allow determination of low serum TES values. The correlation to the DELFIA® Testosterone kit results confirmed the same (r = 0.989, slope = 1.17, intercept 3.95).

The use of the A60/HCDR1/LCDR2 Fab with patient serum specimens revealed that the developed Fab fragments do work in complex assay medium containing human serum and, e.g., components intended to facilitate the release of TES from binding proteins in the sample. However, particularly the low TES concentrations deviate significantly from the gas chromatography - mass spectrometry analysis. This positive bias is most probably as a consequence of cross-reactions with related steroids, notably DHEAS, normal concentration in female samples being 300 -10 000 times higher than TES concentration.

EXAMPLE 4

REOPTIMIZATION OF CDR3 LOOPS AND CDR COMBINATION

In this example the CDR3 loops of the combined mutant clone were retargetted to oligo- directed PCR mutagenesis. The bias of the oligonucleotides for the parental sequence was weaker than in the previous CDR3 mutant libraries and the mutant libraries were cloned by utilising unique restriction sites created by site-directed mutagenesis nearby the CDRs thus improving the quality and size of the libraries. Combined selection procedure allowing simultaneous selection with respect to affinity and specificity was used in addition to the previously utilized affinity panning approach.

I. Reoptimization (Mutagenesis) of the CDR3 loops of the A60/HCDR1/LCDR2 antibody

The V_L and V_H CDR3 loops (as defined in Kabat et al, 1991) of the mutant Fab fragment (after CDRl, 2 and 3 metagenesis, A60/HCDR1/LCDR2, Example 3) of the 3-C^ antibody (Example 1) were retargetted to mutagenesis by using spiked PCR primers (Hermes et al, 1989; 1990). To restrict the frequency of new amino acid mutations in single clones, and thus to limit the number of the clones not retaining previously optimised structural features of the CDRs the randomness of PCR primers were adjusted by nucleotide doping during the oligonucleotide synthesis for CDR loops (62.5% parental-type (A60 HCDR1/LCDR2) and 12.5%) each three other nucleotides). Both mutant libraries were constructed into the pComb3 vector (Barbas et al, 1991) containing the mutant Fab fragment (A60/HCDR1/LCDR2) by utilising unique restriction sites created before cloning by site-directed mutagenesis sites nearby the CDR3s thus improving the quality and size of the libraries. II. Selection of the mutant libraries

The mutant libraries were selected by the antibody phage display technique (McCafferty et al, 1990, Barbas et al, 1991). Briefly, on every panning round mutant libraries were first incubated with a high concentration of soluble DHEA-BSA conjugate (DHEA-3-SucH-BSA, 17-20 mol DHEA mol BSA, 0.4 mg/ml) for at least 30 min at 37°C in order to inactivate Fab phages with cross-reactivity to DHEAS. On two first rounds the phage pools were then transferred into microtiter plate wells coated with 1 μg of TES-3-CMO-BSA and on next three rounds to streptavidin-coated microtiter plate wells (Boehringer Mannheim GmbH, Germany) preincubated for 1 h with TES-3-biotin. After a 10 min incubation at 37 °C, the wells were washed (TBS-1% BSA-0.05% Tween 20) 15 times during an 1 h period before elution of binders with 50 mM NaOH, pH 12.6 for 15 min and immediate neutralisation with 1 M Tris, pH 7.5. Between panning rounds the eluted phage pools were amplified by infecting E.coli XL-1 Blue cells. A total of five panning cycles were performed and during the last three cycles the concentration of TES-3-biotin used for preincubation of streptavidin-coated wells was decreased stepwise from 22 nM to 0.2 nM.

The mutant libraries were also affinity selected without any competing steroid by using limiting, gradually decreasing concentrations of biotinylated TES to catch the high-affinity binders (Hawkins et al, 1992; Schier et al, 1996b). The libraries were first incubated with a biotinylated TES (TES-3 -biotin) for 5 min, then the phage pools were transferred to streptavidin-coated microtiter plate wells (Boehringer Mannheim GmbH, Germany) and incubated for 25 min at 37 °C. After washing the plate 15 times during an 1 h period with the washing solution (TBS-1% BSA-0.05% Tween 20) the TES binders were eluted with 50 mM NaOH, pH 12.6 for 15 min and then immediately neutralised with 1 M Tris, pH 7.5. A total of five rounds of panning were performed, and during the cycles the concentration of biotinylated TES was decreased gradually from 1 nM to 10 pM.

HI. Characterisation of the mutants

After the last panning step phagemid DNA was isolated and the pill gene region was removed by Spe I and Nhe I digestion from the pComb 3 vector (Barbas et al, 1991). The re-ligated vector DNA was transformed and about 50 individual clones were grown in a small scale to produce soluble Fab fragments for preliminary characterisation. Clones were analysed on a competitive ELISA test using TES-3-CMO-BSA-coated wells (0 1 μg) to catch the TES- binders and soluble TES and DHEAS to achieve inhibition of binding (as described earlier, Example 2). The most promising clones having either decreased DHEAS cross-reactivity and/or improved TES-binding affinity were sequenced (Sanger et al, 1977) and selected for further characterisation.

These selected clones were cloned into the pKKtαc expression vector (Takkinen et al , 1991) and transformed into the E.coli strain RV308 for small scale production Clones were characterised using goat anti-mouse IgG-coated microtiter wells, and TES-3-CMO-polylysine labelled with Europium-chelate (from Wallac) was used as the fluorescent label. A one incubation step assay protocol was used for characterisation of these clones i e. the label, antibody sample and competing steroid were all added in the same time onto the wells (Example 3/N). Dilution series of TES, DHEAS, 5α-DHT and androstenedione (all in serum) were used to analyse the relative affinities (ED-50 % concentrations) for the corresponding steroids. Cross-reactivities were calculated from these values (Table I )

The most interesting clones were produced in a larger scale and purified using the HiTrap protein G columns (Pharmacia Biotech) following the same protocol as recommended by the manufacturer for purification of IgG samples

TABLE I. Affinities and cross-reactivities of the best anti-testosterone Fab fragments after combined affinity/specificity and affinity panning procedures

ED50 (nM) Cross-reaction %

rFab TES DHEAS 5 -DHT Androstenedione

A60/HCDR1/LCDR2 6 1 0 020 14.9 2.0

LCDR3 A4 8 7 0 007 6.3 2 4

LCDR3 A46 2 3 0 030 22.6 3 1

HCDR3 A52 / LCDR3 A4 8 4 0 009 7 8 3 4 In the best light chain CDR3 mutant clone isolated from the combined affinity/specificity panning approach, A4, the DHEAS and 5 -DHT cross-reactivities were clearly decreased (3- fold and 2-fold, respectively) while the TES affinity was preserved on the same level as in the parental mutant clone. The TES affinity of the light chain CDR3 mutant clone A46, selected by affinity panning, was improved 3-fold but all cross-reactivities were increased. The binding properties of the affinity-selected heavy chain CDR3 mutant clone A52, combined to new light chain CDR3 mutant A4, were essentially the same as those of the clone A4 alone, although the same CDR combination showed positive effect when expressed from the pComb3 vector. The deduced amino acid sequences of the re-optimised CDR3 loops are shown in Figure 11.

To further improve the overall binding properties of the TES antibody, new light chain CDR3 sequences were created by combining mutations from three different LCDR3 mutant clones (28, A4 and A46). Finally, a number of new CDR combinations were created by combining different CDR loop sequences selected during the stepwise optimisation of the CDRs of the 3- G^ antibody.

TABLE II. Affinities and cross-reactivities of the best anti-testosterone Fab fragments after combining the LCDR3 mutations from different mutant clones and by optimisation of the CDR combination.

ED50 (nlv 1) Cross-reaction %>

rFab TES DHEAS 5α-DHT Androstenedione

A60/HCDR1/LCDR2 6.1 0.020 14.9 2.0

S58 2.7 0.009 11.3 2.3

S73 4.8 0.006 7.2 2.3

S77 4.8 0.007 7.9 2.2

S83 2.4 0.005 6.9 1.8

S88 4.6 0.007 5.4 2.5 As shown in Table II, the optimisation of LCDR3 sequence and CDR combination resulted in clones having excellent overall binding profile. No proper inhibition of the original monoclonal antibody was achieved in the modified research assay system with the same clinically relevant standard concentrations. The extrapolated TES ED₅₀ value for the original antibody was 76.5 nM and the cross-reactivities 0.3% (DHEAS), 13.7% (5 -DHT) and 16.6%

(androstenedione). Based on the extrapolated values, we were able to produce a TES-binding Fab fragment having mutations in five CDRs and showing 32-fold higher affinity to TES, 60- fold lower cross-reactivity to DHEAS, 2-fold lower cross-reactivity to 5 -DHT and over 9- fold lower cross-reactivity to androstenedione compared to the original monoclonal antibody. The amino acid sequences of the V_L and V_H regions of Fab fragments S73, S77, S83 and S88 aligned with the V_L and V_H sequences of the 3-C4F5 antibody are shown in Figures 12 and 13 respectively.

EXAMPLE 5

PERFORMANCE OF SELECTED RECOMBINANT FAB FRAGMENTS IN A CLINICAL IMMUNOASSAY

In this example the performance of the best mutant Fab fragments were analysed in the context of competitive immunoassay protocol and compared to currently used commercial kit, in which rabbit polyclonal antibodies are used as TES specific reagents, designed for in vitro measurement of the concentration of TES in human serum or plasma (DELFIA® Testosterone kit, Time-resolved fluoroimmunoassay kit 1244-050, Wallac).

I. The immunoassay protocol

First, suitable dilutions for the Fab fragments (in 0.5% BSA/PBS) were titrated as described in Example 3/V. Next, different TES standards (0-50 nmol/1 in human TES-free serum) and unknown patient serum specimens were analysed using the same assay protocol as described in Example 3/N. The unknown patient serum specimens were also analysed using the DELFIA® Testosterone kit as recommended by the manufacturer. II. Performance of the selected recombinant mutant Fab fragments in the clinical immunoassay

The standard curves for all the analysed Fab fragments were in the clinically relevant concentration range. The lowest standard (0.49 nmol/1) gave 86.5% - 93.1% (B/Bmax %) binding for the Fab fragments and 85.9%) for the kit. The highest standard (50.9 nmol 1) gave 8.6%o - 15.6%) and 10.4%> binding for the Fab fragments and the kit, respectively. The determined TES ED-50%> values were 3.4 - 8.8 nmol/1 for the Fab fragments and 4.3 nmol 1 for the kit.

Figures 14 and 15 show the results obtained for 48 female and for 32 male patient samples, respectively, using either Fab fragments in the one-step immunoassay or the commercial DELFIA® testosterone kit designed for in vitro measurement of the concentration of TES in human serum or plasma. Table I shows the comparision of correlation parameters between different Fab fragments in the one-step immunoassay and the commercial DELFIA® Testosterone kit when 80 patient serum samples were measured.

TABLE I. The comparison between Fab fragments and commercial DELFIA® Testosterone kit used for measuring 80 patient serum samples.

A4 A52/A4 S73 S77 S83 S88 DELFIA

TES ED-50 8,6 8,8 5,1 5,6 3,4 4,8 4,3

Sample min 0,98 1 ,09 0,47 0,1 0,69 0,34 0,13

Sample max 35,62 33,89 30,8 36,45 29,53 30,77 26,57

Slope 1.161 1.187 1.094 1.105 1.059 1.043

Intercept 1.765 1.217 0,372 0,005 0,877 0,059 r-value 0,971 0,971 0,964 0,943 0,962 0,968

A good correlation was found for all the Fab fragments (r = 0.943 - 0.971). However, the obtained intercept values of the A4 and A52/A4 Fab fragments (1.765 and 1.217, respectively) revealed that the analysis of the additivity of different mutations and different optimised CDRs was a necessary and efficient step to further fine-tune functioning of the Fab fragments. After the final optimisation the binding properties of the mutant Fab fragments, notably S77 and S88 Fabs, in this modified research assay system were comparable to those of a rabbit polyclonal anti-testosterone antiserum good enough for diagnostic immunoassay. By taking account of simultaneous improvements in affinity and specificity we were able to improve the overall functioning of the recombinant antibody in the context of competitive immunoassay considerably and have now achieved the utmost demands for cross-reactivities to fulfil the diagnostic requirements.

References

Alfthan, K , Takkinen, K , Sizmann, D , Seppala, I , Immonen, T , Vanne, L , Keranen, S ,

Kaartinen, M , Knowles, J K C and Teeri, T T (1993) Gene. 128, 203-209 Allen, F H , Davies, J E , Galloy, J J , Johnson, O , Kennard, O , Macrae, C F , Mitchell, E M ,

Mitchell, G F , Smith, J M and Watson, D G (1991) J. Chem. Inf. Corny. Sci , 31, 187-204

Alzari, P M , Spinelli, S , Mariuzza, R A , Boulot, G , Poljak, R J , Jarvis, J M and Milstein,

C (1990) EMBO Journal 9. 3807-3814

Arevalo, J H , Stura, E A , Taussig, M J and Wilson, I A (1993a) J. Mol Biol , 231, 103- 118

Arevalo, J H , Taussig, M J and Wilson, I A (1993b) Nature. 365, 859-863

Arevalo, J H , Hassig, C A , Stura, E A , Sims, M J , Taussig, M J and Wilson, I A (1994) J_

Mol. Biol ,_241, 663-690

Argos, P (1988) Protein Engineering. 2. 101-113 Barbas III, C F , Bain, J D , Hoekstra, D M , and Lerner, R A (1992 Proc. Nat Acad. Sci ,

USA , 89, 4457-4461

Barbas III, C F , Hu, D , Dunlop, N , Sawyer, L , Cababa, D , Hendry, R M , Nara, P L and

Burton, D R (1994) Proc. Natl Acad. Sci USA, 91, 3809-3813

Barbas III, C F , Kang, A S , Lerner, R A and Benkovic, S J (1991) Proc. Natl Acad. Sci USA, 88, 7978-7982

Baumann, M (1990) Anal Bwchem , 190, 198-208

Berek, C and Milstein, C (1987) Immunol. Rev . 105. 5-23

Bernstein, F C , Koetzle, T F , Willliams, G J B , Meyer, E F , Jr, Brice, M D , Rodgers,

J R ,Kennard, O , Shimanouchi, T and Tasumi, Y (1977) J. Mol. Biol , 112, 535-542 Better, M , Chang, C P , Robinson, R R and Horwitz, AH (1988) Science. 240, 1041-1043

Blundell, T L , Johnson, M S , Oveπngton, J P , and Sali, A (1990) In Hook, J B and Poste,

G (eds ). Protein Design and the Development of New Therapeutics and Vaccines, Plenum

Press, USA, pp 209-227

Brooks, B R , Bruccoleri, R E , Olafson, B D , Sates, D J , Swaminathan, S and Karplus, M (1983) J. Corny. Chem , 4, 187-217

Brooks, C L , III, Karplus, M and Pettit, B M (1988) Proteins A Theoretical Perspective of

Dynamics, Structure, and Thermodynamics, Advances in Chemical Physics, Vol LXXI,

Wiley, New York, pp 33-58 Deng, S -J , MacKenzie, C R , Himara, T , Brousseau, R , Lowary, T L , Young, N M , Bundle, D R and Narang, S A (1995) Proc. Natl Acad. Sci USA, 92, 4992-4996 Galfre, G and Milstein, C (1981) Methods Enzymol . 73, 3-46

Harlow, E and Lane, D (1988) Antibodies. A Laboratory Manual Cold Spπng Harbor Laboratory Press, Cold Spring Harbor, NY

Hawkins, R E and Winter, G (1992) Eur. J Immunol . 22. 867-870

Hermes, J D , Parekh, S M , Blacklow, S C , Koster, H and Knowles, J R (1989) Gene. 84,

143-151

Hermes, J D , Blacklow, S C and Knowles, J R (1990) ) Proc. Natl Acad. Sci. USA, 87, 696- 700

Herron, J N , He, X , Mason, M L , Voss, Jr , W and Edmundson, A B (1989) PROTEINS

5, 271

Ho, S N , Hunt, H D , Horton, R M , Pullen, J K and Pease, L R (1989) Gene. 11, 51-59

Hoffren, A-M , Holm, L , Laaksonen, L , Teeri, T T and Teleman, O (1992) ImmunoMeihods, 1, 80-90

Ismail, A A A , Astley, P , Cawood, M , Short, F , Wakelin, K and Wheeler, M (1986) Ann. Clin. Biochem , 23, 135-145

Jeffrey, P D , Strong, R K , Sieker, L C , Chang, C Y Y , Cambell, R L , Petsko, G A , Haber, E , Margolies, M N and Sheriff, S (1993) Proc. Natl Acad. Sci. USA, 90, 10310-10314 Johnson, M S , Overington, J P and Blundell, T L (1993) J. Mol Biol , 231, 735-752 Johnson, M S and Overington, J P (1993) J. Mol. Biol , 233, 716-738 Johnsson, B , Lofas, S , and Lindqvist, G (1991) Anal. Biochem , 198, 268-277 Jonsson, U , Fagerstam, L , Ivarsson, B , Johnsson, B , Karlsson, R , Lundh, K , Lofas, S , Persson, B , Roos, H , Ronnberg, I , Sjolander, S , Stenberg, E , Stahlberg, R , Urbaniczky, C , Ostlin, H , and Malmqvist, M (1991) B oTechn ques. 11, 620-627

Kabat, E A , Wu, T T , Reid-Miller, M , Perry, H M and Gottesman, K S (1991) Sequences of Proteins of Immulosical Interest. 4th Ed , U S Dept of Health and Human Services, Bethesda, MD Karlsson, R , Michaelsson, A , and Mattson, A (1991 J. Immunol Methods, 145, 229-240 Korth-Schutz, S , Virdis, R , Saenger, P , Chow, D M , Levine, L S and New, M I (1978) J Clin Endocrinol Metab , 46, 452-457 Laemmli, U K (1970) Nature. 227, 680-685 Leung, D W , Chen, E and Goeddel, D V , (1989) Tech J Methods Cell Mol Biol , 1, 11 MacKenzie, C R , Sharma, V , Brummell, D , Bilous, D , Dubuc, G , Sadowska, J , Young, N M , Bundle, D R and Narang, S A (1994) Bio/Technoloev, 12, 390-395 Marks, J D , Hoogenboom, H R , Bonnert, T P , McCafferty, J , Griffiths, A D and Winter, G ( 1991) J. Mol. Biol . 222. 581-597 McCafferty, J , Griffiths, A D , Winter, G and Chiswell, F J (1990) Nature. 348, 552-554 Myers, R M , Lerman, L S and Maniatis, T (1985) Science. 229, 242-247 Odel, W D and Swerdloff, R S (1978) Clin. Endocrinol , 8, 149-180

Orlandi, R , Gussow, D H , Jones, P T and Winter, G (1989) Proc. Natl Acad. Sci USA, 86, 3833-3837 Osborn, R H and Yannone, M E (1973) Gvnecol. Surv , 26, 195-228

Parham, P , Androlewicz, M J , Brodsky, F M , Holmes, N J and Ways, J P (1982) J_ Immunol. Methods, 53, 133-173

Pope, A , Pritchard, K , Williams, A , Roberts, A , Hackett, J R , Mandecki, W and Johnson, K S (1996) Immunotechnolosv. 2. 209-217 Saiki, R K , Gelfand, D H , Stoffel, S , Scharf, S J , Higuchi, R , Horn, G T , Mullis, K B and Erlich, H A (1988) Science. 239, 487-491

Sali, A , Overington, J P , Johnson, M S and Blundell, T L (1990) TIBS. 15, 235-240 Sambrook, J , Fritsch, E F and Maniatis, T (1990) Molecular Cloning: A Laboratory Manual 2nd Ed , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Sanger, F , Nicklen, S and Coulson, A R (1977) Proc. Natl Acad. Sci USA, 74, 5463-5467 Schaaper, R M (1988) Proc. Natl Acad. Sci USA, 85, 8126-8130

Schier, R , Balint, R F , McCall, A , Apell, G , Larπck, J W and Marks, J D (1996b) Gene, 169, 147-155 Schier, R , Bye, J M , Apell, G , McCall, A , Adams, G P , Malmqvist, M , Weiner, L M and Marks, J D (1996a) J. Mol Biol , 225, 28-43

Schier, R , McCall, A , Adams, G P , Marshall, K W , Merritt, H , Yim, M , Crawford, R S , Weiner, L M , Marks, C and Marks, J D (1996c) J. Mol Biol , 263, 551-567 Schildbach, J F , Shaw, S Y , Bruccoleπ, R E , Haber, E , Herzenberg, L A , Jager, G C , Jeffrey, P D , Panka, D J , Parks, D R , Near, R I , Novotny, J , Sheriff, S and Margolies, M N (1994) Protein Sci . 3. 737-749

Short, M K , Jeffrey, P D , Kwong, R -F and Margolies, M N (1995) J Biol Chem , 270,

28541-28550

Skerra, A and Pluckthun, A , (1988) Science, 240, 1038-1041 Sutcliffe, M.J., Hayes, F.R.F. and Blundell, T.L. (1987) Protein Ensin.. 1, 385-392.

Takkinen, K., Laukkanen, M-L., Sizmann, D., Alfthan, K., Immonen, T., Vanne, L., Kaartinen,

M., Knowles, J.K.C. and Teeri, T.T. (1991) Protein Ensin., 4, 837-841.

Towbin, H., Staehlin, T. and Gordon, J. (1979) Proc. Natl Acad. Sci. USA, 76, 4350-4354. Wang, C. and Swerloff, R.S. (1992) Baillieres Clin. Endocrinol. Metab., 6, 405-435.

Wetlaufer, D. (1962) Adv. Protein Chem., 17, 303-390.

Wheeler, M.J. (1994) In: Wild, D. (ed.) The immunoassay handbook. Macmillan Press, Hants,

UK, pp. 365-78.

Wilke .J. and Utley, D.J. (1987) Clin. Chem.. 33, 1372-1375. Wilson, LA. and Stanfield, R.L. (1994) Current Qyinion in Structural Biolosv. 4, 857-867.

Wu, F.C.W., Edmond, P., Raap, G. and Hunter, W.M. (1981) Clin. Endocrinol. 14, 493-507.

Yang, W.-P., Green, K., Pinz-Sweeney, S., Briones, A.T., Burton, D.R. and Barbas, C.F.

(1995) J. Mol Biol, 254, 392-403.

Yelton, D.E., Rosok, M.J., Cruz, G., Cosand, W.L., Bajorath, J., Hellstrόm, I., Hellstrόm, K.E., Huse, W.D. and Glaser, S.M. (1995) J. Immunol, 155, 1994-2004.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT:

(A) NAME: Orion-yhtyma Oyj Orion Diagnostica

(B) STREET: Orionintie 1

(C) CITY: Espoo

(E) COUNTRY: Finland

(F) POSTAL CODE (ZIP) : FIN-02200

(A) NAME: Wallac Oy

(B) STREET: P.O. Box 10

(C) CITY: Turku

(E) COUNTRY: Finland

(F) POSTAL CODE (ZIP) : FIN-20101

(ii) TITLE OF INVENTION: Monoclonal Antibodies (iii) NUMBER OF SEQUENCES: 26

(iv) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: Patentin Release #1.0, Version #1.25 (EPO)

(2) INFORMATION FOR SEQ ID NO: 1:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 112 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain variable region of the monoclonal 3-C₄F₅ antibody)

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 24..39

(D) OTHER INFORMATION: /label= CDR

/note= "CDRl"

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 55..61

(D) OTHER INFORMATION: /label= CDR /note= "CDR2"

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 94..102

(D) OTHER INFORMATION: /label= CDR /note= "CDR3"

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 1:

Asp Val Val Val Thr Gin Thr Pro Leu Ser Leu Pro Val Ser Leu Gly 1 5 10 15 Asp Gin Ala Ser lie Ser Cys Arg Ser Ser Gin Ser lie Val His Ser 20 25 30

Asn Gly Asn Ser Tyr Leu Glu Trp Tyr Leu Gin Lys Pro Gly Gin Ser 35 40 45

Pro Lys Leu Leu lie Tyr Lys Val Ser Asn Arg Phe Ser Gly Val Pro 50 55 60

Asp Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr Leu Lys lie 65 70 75 80

Ser Arg Val Glu Ala Glu Asp Leu Gly Val Tyr Tyr Cys Phe Gin Gly 85 90 95

Ser His Val Pro Pro Thr Phe Gly Gly Gly Thr Lys Leu Glu lie Lys 100 105 110

(2) INFORMATION FOR SEQ ID NO : 2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 118 amino acids

(B) TYPE: amino acid

(C) STRANDΞDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain variable region of the monoclonal 3-C₄F₅ antibody)

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 31..35

(D) OTHER INFORMATION: /label= CDR

/note= "CDRl"

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 50..65

(D) OTHER INFORMATION: /label= CDR /note= "CDR2"

(ix) FEATURE:

(A) NAME/KEY: Region

(B) LOCATION: 98..107

(D) OTHER INFORMATION: /label= CDR /note= "CDR3"

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2:

Glu Val Lys Leu Val Glu Ser Gly Gly Gly Leu Val Lys Pro Gly Gly 1 5 10 15

Ser Leu Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Thr Tyr 20 25 30

Ala Leu Ser Trp Val Arg Gin Thr Ala Asp Lys Arg Leu Glu Trp Val 35 40 45

Ala Ser lie Val Ser Gly Gly Asn Thr Tyr Tyr Ser Gly Ser Val Lys 50 55 60 Gly Arg Phe Thr lie Ser Arg Asp He Ala Arg Asn He Leu Tyr Leu 65 70 75 80

Gin Met Ser Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala 85 90 95

Arg Glu Tyr Tyr Gly Tyr Val Gly Leu Ala Tyr Trp Gly Gin Gly Thr 100 105 110

Leu Val Thr Val Ser Ala 115

(2) INFORMATION FOR SEQ ID NO : 3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 9 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDR3-loop of the wild-type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 3:

Phe Gin Gly Ser His Val Pro Pro Thr 1 5

(2) INFORMATION FOR SEQ ID NO : 4:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 9 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDR3-loop of the mutant clone 28)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 4:

Phe Gin Gly Ser Thr Val Pro Pro Thr 1 5

(2) INFORMATION FOR SEQ ID NO : 5:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 10 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain CDR3-loop of the wild-type clone) (xi) SEQUENCE DESCRIPTION: SEQ ID NO : 5:

Glu Tyr Tyr Gly Tyr Val Gly Leu Ala Tyr 1 5 10

(2) INFORMATION FOR SEQ ID NO : 6:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 10 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain CDR3-loop of the mutant clone 44)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 6:

Ala Tyr Tyr Gly Tyr Val Gly Leu Val Tyr 1 5 10

(2) INFORMATION FOR SEQ ID NO : 7:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDRl-loop of the wild type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 7:

Arg Ser Ser Gin Ser He Val His Ser Asn Gly Asn Ser Tyr Leu Glu 1 5 10 15

(2) INFORMATION FOR SEQ ID NO : 8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDRl-loop of the mutant clone A60)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 8:

Arg Ser Ser Glu Val He Val Thr Arg Asn Gly Tyr Thr Pro He Glu 1 5 10 15 (2) INFORMATION FOR SEQ ID NO : 9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDRl-loop of the mutant clone A58)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 9:

Arg Ser Ser Gin Arg Met Val Gin Arg Asn Gly His Thr Pro Leu Glu 1 5 10 15

(2) INFORMATION FOR SEQ ID NO: 10:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 7 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDR2-loop of the wild type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 10:

Lys Val Ser Asn Arg Phe Ser 1 5

(2) INFORMATION FOR SEQ ID NO: 11:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 7 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain CDR2-loop of the mutant clone 1D7)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 11:

Lys Ala Arg Tyr Arg Phe Pro 1 5

(2) INFORMATION FOR SEQ ID NO: 12:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 5 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain CDRl-loop of the wild-type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 12:

Thr Tyr Ala Leu Ser 1 5

(2) INFORMATION FOR SEQ ID NO: 13:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 5 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain CDRl-loop of the mutant clone 2B5)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 13:

Arg Tyr Ala Leu Ser 1 5

(2) INFORMATION FOR SEQ ID NO: 14:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 9 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain re-optimised CDR3-loop of the wild-type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 14:

Phe Gin Gly Ser His Val Pro Pro Thr 1 5

(2) INFORMATION FOR SEQ ID NO: 15:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 9 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain re-optimised CDR3-loop of the mutant clone A4) (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 15:

Phe Asp Gly Ser His Val Pro Pro Lys 1 5

(2) INFORMATION FOR SEQ ID NO: 16:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 9 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (light chain re-optimised CDR3-loop of the mutant clone A46)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 16:

Phe Glu Gly Ser Gin Val Pro Pro Thr 1 5

(2) INFORMATION FOR SEQ ID NO: 17:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 10 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain re-optimised CDR3-loop of the wild-type clone)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 17:

Glu Tyr Tyr Gly Tyr Val Gly Leu Ala Tyr 1 5 10

(2) INFORMATION FOR SEQ ID NO: 18:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 10 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (heavy chain re-optimised CDR3-loop of the mutant clone A52)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 18:

Ala Tyr Tyr Gly Tyr Val Gly Leu Val His 1 5 10

(2) INFORMATION FOR SEQ ID NO : 19: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 112 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_L region of the Fab fragment of S73)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 19:

Asp Val Val Val Thr Gin Thr Pro Leu Ser Leu Pro Val Ser Leu Gly 1 5 10 15

Asp Gin Ala Ser He Ser Cys Arg Ser Ser Glu Val He Val Thr Arg 20 25 30

Asn Gly Tyr Thr Pro He Glu Trp Tyr Leu Gin Lys Pro Gly Gin Ser 35 40 45

Pro Lys Leu Leu He Tyr Lys Ala Tyr Lys Arg Phe Pro Gly Val Pro 50 55 60

Asp Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr Leu Lys He 65 70 75 80

Ser Arg Val Glu Ala Glu Asp Leu Gly Val Tyr Tyr Cys Phe Asp Gly 85 90 95

Ser Thr Val Pro Pro Lys Phe Gly Gly Gly Thr Lys Leu Glu He Lys 100 105 110

(2) INFORMATION FOR SEQ ID NO : 20:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 112 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_L region of the Fab fragment of S77)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 20:

Asp Val Val Val Thr Gin Thr Pro Leu Ser Leu Pro Val Ser Leu Gly 1 5 10 15

Asp Gin Ala Ser He Ser Cys Arg Ser Ser Glu Val He Val Thr Arg 20 25 30

Asn Gly Tyr Thr Pro He Glu Trp Tyr Leu Gin Lys Pro Gly Gin Ser 35 40 45

Pro Lys Leu Leu He Tyr Lys Ala Tyr Lys Arg Phe Pro Gly Val Pro 50 55 60

Asp Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr Leu Lys He 65 70 75 80 Ser Arg Val Glu Ala Glu Asp Leu Gly Val Tyr Tyr Cys Phe Asp Gly 85 90 95

Ser Thr Val Pro Pro Lys Phe Gly Gly Gly Thr Lys Leu Glu He Lys 100 105 110

(2) INFORMATION FOR SEQ ID NO : 21:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 112 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_L region of the Fab fragment of S83)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 21:

Asp Val Val Val Thr Gin Thr Pro Leu Ser Leu Pro Val Ser Leu Gly 1 5 10 15

Asp Gin Ala Ser He Ser Cys Arg Ser Ser Glu Val He Val Thr Arg 20 25 30

Asn Gly Tyr Thr Pro He Glu Trp Tyr Leu Gin Lys Pro Gly Gin Ser 35 40 45

Pro Lys Leu Leu He Tyr Lys Ala Tyr Lys Arg Phe Pro Gly Val Pro 50 55 60

Asp Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr Leu Lys He 65 70 75 80

Ser Arg Val Glu Ala Glu Asp Leu Gly Val Tyr Tyr Cys Phe Asp Gly 85 90 95

Ser Gin Val Pro Pro Lys Phe Gly Gly Gly Thr Lys Leu Glu He Lys 100 105 110

(2) INFORMATION FOR SEQ ID NO: 22:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 112 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_L region of the Fab fragment of S88)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 22:

Asp Val Val Val Thr Gin Thr Pro Leu Ser Leu Pro Val Ser Leu Gly 1 5 10 15

Asp Gin Ala Ser He Ser Cys Arg Ser Ser Glu Val He Val Thr Arg 20 25 30 Asn Gly Tyr Thr Pro He Glu Trp Tyr Leu Gin Lys Pro Gly Gin Ser 35 40 45

Pro Lys Leu Leu He Tyr Lys Ala Tyr Lys Arg Phe Pro Gly Val Pro 50 55 60

Asp Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr Leu Lys He 65 70 75 80

Ser Arg Val Glu Ala Glu Asp Leu Gly Val Tyr Tyr Cys Phe Asp Gly 85 90 95

Ser His Val Pro Pro Lys Phe Gly Gly Gly Thr Lys Leu Glu He Lys 100 105 110

(2) INFORMATION FOR SEQ ID NO : 23:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 118 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V„ region of the Fab fragment of S73)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 23:

Glu Val Lys Leu Val Glu Ser Gly Gly Gly Leu Val Lys Pro Gly Gly 1 5 10 15

Ser Leu Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Arg Tyr 20 25 30

Ala Leu Ser Trp Val Arg Gin Thr Ala Asp Lys Arg Leu Glu Trp Val 35 40 45

Ala Ser He Val Ser Gly Gly Asn Thr Tyr Tyr Ser Gly Ser Val Lys 50 55 60

Gly Arg Phe Thr He Ser Arg Asp He Ala Arg Asn He Leu Tyr Leu 65 70 75 80

Gin Met Ser Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala 85 90 95

Arg Ala Tyr Tyr Gly Tyr Val Gly Leu Val Tyr Trp Gly Gin Gly Thr 100 105 110

Leu Val Thr Val Ser Ala 115

(2) INFORMATION FOR SEQ ID NO: 24:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 118 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide (vi) ORIGINAL SOURCE: (C) INDIVIDUAL ISOLATE: (V_H region of the Fab fragment of S83)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 24:

Glu Val Lys Leu Val Glu Ser Gly Gly Gly Leu Val Lys Pro Gly Gly 1 5 10 15

Ser Leu Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Arg Tyr 20 25 30

Ala Leu Ser Trp Val Arg Gin Thr Ala Asp Lys Arg Leu Glu Trp Val 35 40 45

Ala Ser He Val Ser Gly Gly Asn Thr Tyr Tyr Ser Gly Ser Val Lys 50 55 60

Gly Arg Phe Thr He Ser Arg Asp He Ala Arg Asn He Leu Tyr Leu 65 70 75 80

Gin Met Ser Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala 85 90 95

Arg Ala Tyr Tyr Gly Tyr Val Gly Leu Val Tyr Trp Gly Gin Gly Thr 100 105 110

Leu Val Thr Val Ser Ala 115

(2) INFORMATION FOR SEQ ID NO : 25:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 118 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_H region of the Fab fragment of S88)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 25:

Glu Val Lys Leu Val Glu Ser Gly Gly Gly Leu Val Lys Pro Gly Gly 1 5 10 15

Ser Leu Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Arg Tyr 20 25 30

Ala Leu Ser Trp Val Arg Gin Thr Ala Asp Lys Arg Leu Glu Trp Val 35 40 45

Ala Ser He Val Ser Gly Gly Asn Thr Tyr Tyr Ser Gly Ser Val Lys 50 55 60

Gly Arg Phe Thr He Ser Arg Asp He Ala Arg Asn He Leu Tyr Leu 65 70 75 80

Gin Met Ser Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala 85 90 95

Arg Ala Tyr Tyr Gly Tyr Val Gly Leu Val Tyr Trp Gly Gin Gly Thr 100 105 110 Leu Val Thr Val Ser Ala 115

(2) INFORMATION FOR SEQ ID NO: 26:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 118 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(vi) ORIGINAL SOURCE:

(C) INDIVIDUAL ISOLATE: (V_H region of the Fab fragment of S77)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 26:

Glu Val Lys Leu Val Glu Ser Gly Gly Gly Leu Val Lys Pro Gly Gly 1 5 10 15

Ser Leu Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Arg Tyr 20 25 30

Ala Leu Ser Trp Val Arg Gin Thr Ala Asp Lys Arg Leu Glu Trp Val 35 40 45

Ala Ser He Val Ser Gly Gly Asn Thr Tyr Tyr Ser Gly Ser Val Lys 50 55 60

Gly Arg Phe Thr He Ser Arg Asp He Ala Arg Asn He Leu Tyr Leu 65 70 75 80

Gin Met Ser Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala 85 90 95

Arg Ala Tyr Tyr Gly Tyr Val Gly Leu Val His Trp Gly Gin Gly Thr 100 105 110

Leu Val Thr Val Ser Ser 115

Claims

Claims:

1. A monoclonal antibody or derivative thereof which binds testosterone with affinity of at least about IO⁹ M¹ and exhibits less than 0.05% cross-reactivity with dehydroepiandrosterone sulphate (DHEAS), less than 10% cross-reactivity with 5α- dihydrotestosterone (5α-DHT) and less than 5% cross-reactivity with androstenedione determined on the basis of ED50 values.

2. A monoclonal antibody or derivative thereof as claimed in claim 1 which exhibits less than 0.01% cross-reactivity with DHEAS.

3. A monoclonal antibody or derivative thereof as claimed in claim 2 which exhibits less than 0.01% cross-reactivity with DHEAS, less than 8% cross-reactivity with 5α- DHT and no more than 2.5% cross-reactivity with androstenedione.

4. A monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 3 wherein the complementarity determining regions (CDRs) of the V_L region are selected from the V_L region CDRs of the Fab fragments S73, S77, S83 and S88 as shown in Figure 12 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability and the CDRs of the V_H region are selected from the V_H region CDRs of the Fab fragments S73, S77, S83 and S88 as shown in Figure 13 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability.

5. A monoclonal antibody or derivative thereof as claimed in claim 4 wherein the V_L region is selected from the V_L regions of the Fab fragments S73, S77, S83 and S88 as shown in Figure 12 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability and the V_H region is selected from the V_H regions of the Fab fragments S73, S77, S83 and S88 as shown in Figure 13, the V_H region of Fab fragment S77 in which the C-terminal Serine is substituted by Alanine, and mutants of said V_H regions having one or more conservative substitutions which do not substantially affect binding capability.

6. A testosterone-binding antibody fragment according to any one of claims 1 to 5.

7. An antibody fragment as claimed in claim 6 selected from the Fab fragments S73, S77, S83 and S88, fragments thereof and extensions thereof which retain binding specificity and affinity for testosterone as defined in claim 1.

8. A derivative according to any one of claims 1 to 5 which is a ScFv.

9. A monoclonal antibody or derivative thereof according to any one of claims 1 to 8 which is labelled.

10. A DNA encoding a monoclonal antibody or derivative thereof according to any one of claims 1 to 8 and fragments thereof which encode the CDRs of the V_L and/or V_H region.

11. A DNA as claimed in claim 10 in the form of a vector.

12. A DNA as claimed in claim 11 which is an expression vector capable of expressing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 8 or at least one antibody chain of said antibody or antibody derivative.

13. A host cell containing a DNA according to any one of claims 10 to 12.

14. A host cell as claimed in claim 13 capable of expressing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 8 or at least one antibody chain of said antibody or antibody derivative.

15. A method of preparing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 8 which comprises culturing a host cell or host cells according to claim 14 capable of expressing the required antibody chain(s) and recovering said antibody or antibody derivative, if necessary after initially recovering and combining component chains.

16. A method of preparing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 8 wherein at least a portion of said antibody or antibody derivative is produced synthetically.

17. A method as claimed in claim 15 or claim 16 which further comprises the step of labelling said antibody or antibody derivative and/or attachment of said antibody or antibody derivative to a solid support.

18. A phage or microbial cell which presents an antibody fragment as claimed in claim 6 or a ScFv as claimed in claim 8 as a fusion protein with a surface protein.

19. A method of selecting an antibody fragment as claimed in claim 6 or a ScFv as claimed in claim 8 wherein said antibody fragment or ScFv is selected from a display library of antibody fragments and/or ScFv antibody derivatives containing a phage or cell according to claim 18.

20. A method of obtaining a monoclonal antibody or derivative thereof according to any one of claims 1 to 8 which comprises selecting a testosterone-binding antibody or antibody fragment, mutating the V_L and/or V_H regions of said antibody or fragment and selecting from the mutants thus obtained an antibody or antibody fragment having the required affinity and specificity for testosterone.

21. A method of assaying testosterone wherein a monoclonal antibody or derivative thereof according to any one of claims 1 to 9 is employed.

22. A kit comprising an antibody or derivative thereof according to any one of claims 1 to 9 in a suitable container for transport and storage.

23. A monoclonal antibody or derivative thereof according to any one of claim 1 to 9 which is attached to or coated upon a solid support.

24. A monoclonal antibody or derivative thereof according to any one of claims 1 to 9 for use in in vivo diagnosis.

5. A monoclonal antibody or derivative thereof according to any one of claims 1 se in in vivo treatment.

AMENDED CLAIMS

[received by the International Bureau on 5 February 1999 (05.02.99); original claims 1-25 replaced by new claims 1-26 (4 pages)]

1. A monoclonal antibody or derivative thereof which binds testosterone with affinity of at least about IO⁹ M^"1 and exhibits less than 0.05% cross -reactivity with dehydroepiandrosterone sulphate (DHEAS), determined on the basis of ED₅₀ values.

2. A monoclonal antibody or derivative thereof as claimed in claim 1 which exhibits less than 0.01% cross-reactivity with DHEAS.

3. A genetically modified derivative of the monoclonal antibody as claimed in claim 1 or 2, which further exhibits less than 10% cross-reactivity with

5α-dihydrotestosterone (5α-DHT) and less than 5% cross-reactivity with androstenedione.

4. The genetically modified derivative as claimed in claim 3, which exhibits less than 8% cross-reactivity with 5α-DHT and no more than 2.5% cross- reactivity with androstenedione.

5. A monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 4 wherein the complementarity determining regions (CDRs) of the V_L region are selected from the V_L region CDRs of the Fab fragments S73, S77, S83 and S88 as shown in Figure 12 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability, and the CDRs of the V_H region are selected from the V_H region CDRs of the Fab fragments S73, S77, S83 and S88 as shown in Figure 13 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability.

6. A monoclonal antibody or derivative thereof as claimed in claim 5 wherein the V_L region is selected from the V_L regions of the Fab fragments S73, S77, S83 and S88 as shown in Figure 12 and mutants thereof having one or more conservative substitutions which do not substantially affect binding capability, and the V_H region is selected from the V_H regions of the Fab fragments S73, S77, S83 and S88 as shown in Figure 13, the V_H region of Fab fragment S77 in which the C- terminal Serine is substituted by Alanine, and mutants of said V_H regions having one or more conservative substitutions which do not substantially affect binding capabil- ity.

7. A testosterone-binding antibody fragment according to any one of claims 1 to 6.

8. An antibody fragment as claimed in claim 7 selected from the Fab fragments S73, S77, S83 and S88, fragments thereof and extensions thereof which retain binding specificity and affinity for testosterone as defined in claim 1.

9. A derivative according to any one of claims 1 to 6 which is a ScFv.

10. A monoclonal antibody or derivative thereof according to any one of claims 1 to 9 which is labelled.

11. A DNA encoding a monoclonal antibody or derivative thereof according to any one of claims 1 to 9 and fragments thereof which encode the CDRs of the V_L and/or V_H region.

12. A DNA as claimed in claim 11 in the form of a vector.

13. A DNA as claimed in claim 12 which is an expression vector capable of expressing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 9 or at least one antibody chain of said antibody or antibody derivative.

14. A host cell containing a DNA according to any one of claims 11 to 13.

15. A host cell as claimed in claim 14 capable of expressing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 9 or at least one antibody chain of said antibody or antibody derivative.

16. A method of preparing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 9 which comprises culturing a host cell or host cells according to claim 15 capable of expressing the required antibody chain(s) and recovering said antibody or antibody derivative, if necessary after initially recovering and combining component chains.

17. A method of preparing a monoclonal antibody or derivative thereof as claimed in any one of claims 1 to 9 wherein at least a portion of said antibody or antibody derivative is produced synthetically.

18. A method as claimed in claim 16 or claim 17 which further comprises the step of labelling said antibody or antibody derivative and/or attachment of said antibody or antibody derivative to a solid support.

19. A phage or microbial cell which presents an antibody fragment as claimed in claim 7 or a ScFv as claimed in claim 9 as a fusion protein with a surface protein.

20. A method of selecting an antibody fragment as claimed in claim 7 or a ScFv as claimed in claim 9 wherein said antibody fragment or ScFv is selected from a display library of antibody fragments and/or ScFv antibody derivatives containing a phage or cell according to claim 19.

21. A method of obtaining a monoclonal antibody or derivative thereof according to any one of claims 1 to 9 which comprises selecting a testosterone- binding antibody or antibody fragment, mutating the V_L and/or V_H regions of said antibody or fragment and selecting from the mutants thus obtained an antibody or antibody fragment having the required affinity and specificity for testosterone.

22. A method of assaying testosterone wherein a monoclonal antibody or derivative thereof according to any one of claims 1 to 10 is employed.

23. A kit comprising an antibody or derivative thereof according to any one of claims 1 to 10 in a suitable container for transport and storage.

24. A monoclonal antibody or derivative thereof according to any one of claims 1 to 10 which is attached to or coated upon a solid support.

25. A monoclonal antibody or derivative thereof according to any one of claims 1 to 10 for use in in vivo diagnosis.

26. A monoclonal antibody or derivative thereof according to any one of claims 1 to 10 for use in in vivo treatment.