A Single-Step “Breeding” Generated a Diagnostic Anti-cortisol Antibody Fragment with Over 30-Fold Enhanced Affinity

Abstract

Cortisol levels in bodily fluids represent a useful index for pituitary−adrenal function, and thus practical anti-cortisol antibodies are required. We have studied “antibody-breeding” approaches, which involve in vitro evolution of antibodies to improve their antigen-binding performances. Here, we produced an antibody fragment to measure serum cortisol levels with over 30-fold enhanced affinity after single mutagenesis and selection steps. A mouse anti-cortisol antibody, Ab-CS#3, with insufficient affinity for practical use, was chosen as the prototype antibody. A “wild-type” single-chain Fv fragment (wt-scFv; K_a, 3.4×10⁸ M⁻¹) was prepared by bacterial expression of a fusion gene combining the V_H and V_L genes for this antibody. Then, random point mutations were generated separately in V_H or V_L by error-prone PCR, and the resulting products were used to assemble scFv genes, which were displayed on filamentous phages. Repeated panning of the phage library identified a mutant scFv (scFv#m1-L10) with an over 30-fold enhanced affinity (K_a 1.2×10¹⁰ M⁻¹). Three amino acid substitutions (Cys49Ser, Leu54Pro, and Ser63Gly) were observed in its V_L sequence. In a competitive enzyme-linked immunosorbent assay (ELISA), the mutant scFv generated dose–response curves with measuring range ca. 0.03–0.6 ng/assay cortisol, midpoint of which (0.15 ng/assay) was 7.3-fold lower than that of wt-scFv. Although cortisone, 11-deoxycortisol, and prednisolone showed considerable cross-reactivity, the mutant scFv should enable sensitive routine cortisol assays, except for measurement after metyrapone or high-dose of prednisolone administrations. Actually, cortisol levels of control sera obtained with the scFv-based ELISA were in the reference range.

Immunoassays are essential tools for monitoring various biomarkers in bodily fluids because of the excellent specificity conferred by antigen−antibody reactions.¹⁾ Currently, most diagnostic antibodies are produced by B-cell hybridoma technology,^2,3) which generates “native” antibodies (in vivo antibodies) induced in animals by immunization as cloned products ensuring constant binding abilities. However, the limited B-cell clone repertoire in mammals often prevents the generation of antibodies with practical performance. In particular, the equilibrium affinity constant (K_a) of antibodies against small biomarkers (i.e., haptenic compounds) rarely exceeds the 10¹⁰ M⁻¹ range.^4,5)

We have employed an “antibody-breeding” approach to overcome this limitation inherent to native antibodies and to enable subfemtomole detection of small molecules.⁵⁾ Using this approach, the antigen-binding affinity can be enhanced by in vitro mutagenesis and subsequent selection of improved species (i.e., in vitro affinity maturation).^3,5–9) Typically, the antibody of interest is converted to the corresponding single-chain Fv fragment (scFv)^10–12) or Fab fragment (Fab) by gene manipulation. Then, random or site-directed mutations are introduced into the genes encoding the variable heavy and light domains (V_H and V_L) to generate a diverse “antibody library,” from which “improved binders” (fragment mutants with improved functions) are selected. Based on these strategies, we have succeeded in enhancing the affinity of anti-estradiol-17β^13,14) and anti-cotinine¹⁵⁾ antibody fragments by >150-fold and >40-fold, respectively. We also showed that these improved scFv molecules could be used in clinical applications.

Cortisol is the major glucocorticoid in humans and is used as a biomarker for the functions of hypothalamic–pituitary–adrenal axis,¹⁶⁾ and thus practical anti-cortisol antibodies have been in great demand as diagnostic reagents. However, only a few publications have demonstrated the production of monoclonal antibodies capable of targeting serum or urinary cortisol.^17–19) We recently established several hybridoma clones producing a mouse anti-cortisol antibody.²⁰⁾ As part of our research mentioned above, we planned to generate anti-cortisol antibody fragments with practical sensitivity. To evaluate the potential of the antibody-breeding approach, we chose an antibody Ab-CS#3 as the prototype antibody, which showed somewhat insufficient affinity (K_a, 4.7×10⁷ M⁻¹) for clinical use, from the panel of monoclonal anti-cortisol antibodies.²⁰⁾ This antibody was converted to the scFv form and improved to a mutant species with over 30-fold greater K_a, which reached 10¹⁰ M⁻¹ range, via single step mutagenesis and subsequent selection.

MATERIALS AND METHODS

Steroids and Steroid-Protein Conjugate

Steroids, including cortisol 3-(O-carboxymethyl)oxime (CS-3-CMO) (Fig. 1), were purchased from Sigma (St. Louis, MO, U.S.A.). [1,2,6,7-³H]-cortisol (3.4 TBq/mmol) was obtained from PerkinElmer, Inc. Japan (Yokohama, Kanagawa, Japan). CS-3-CMO was linked with bovine serum albumin (BSA; Sigma) to afford a CS−BSA conjugate (average of cortisol residues per BSA molecule was 12)²⁰⁾ (Fig. 1).

Fig. 1. Structures of Cortisol, Cortisol 3-(O-Carboxymethyl)oxime (CS-3-CMO), and Cortisol–BSA Conjugate

Part of numbering of carbon atoms for the C-21 steroids (pregnanes: including cortisol) is also shown.

Primers

The single-stranded oligo-DNAs used for PCR amplification were synthesized and purified by Tsukuba Oligo Service (Ushiku, Ibaraki, Japan). The nucleotide sequences of the primers used are listed in Table 1.

Table 1. Nucleotide Sequences of PCR Primers Used in This Study

Primer	Sequence (5′→3′; Restriction sitea))
CS#3VH-Rev	CTCGCGGCCCAGCCGGCCATGGCCCAGGTCCAACTGCAGCAGCCTG (NcoI)
CS#3VH-For	TGAACCGCCTCCACCGCTCGAGACTGCAGAGACAGTGACCAGAGTC (XhoI)
CS#3VL-Rev	GGATCCGGCGGTGGCGGGTCGACGGACATTGTGCTGACACAGTCTC (SalI)
CS#3VL-For	GGGCTCAACTTTCTTTGCGGCCGCAGCCCGTTTTATTTCCAGCTTG (NotI)

a) Underlined in the nucleotide sequences.

Buffers

PB: 0.050 M sodium phosphate buffer (pH 7.3); PBS: PB containing 9.0 g/L NaCl; G-PBS: PBS containing 1.0 g/L gelatin; M-PBS: PBS containing 20 g/L skim milk; T-PBS: PBS containing 0.050% (v/v) Tween 20; PVG-PBS: G-PBS containing 1.0 g/L polyvinyl alcohol with an average polymerization degree of 500; PBS-2: 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, 0.14 M NaCl, 2.7 mM KCl (pH 7.4); M-PBS-2: PBS-2 containing 20 g/L skim milk; and T-PBS-2: PBS-2 containing 0.10% (v/v) Tween 20.

Hybridoma

The hybridoma clone secreting anti-cortisol antibody Ab-CS#3 (denoted as Ab#3 in our previous article)²⁰⁾ was established previously in our laboratory.²⁰⁾

Assembly and Expression of the Wild-Type scFv Gene

The V_H and V_L DNA fragments for Ab-CS#3, which were cloned in pBluescript II plasmid (Agilent Technologies; Hachioji, Tokyo, Japan) previously,²⁰⁾ were amplified separately to add restriction sites for subcloning into the phagemid vector pEXmide 7, a variant of pEXmide 5²¹⁾ having cloning sites for V_H and V_L separately, each of which is directly adjacent (upstream and downstream) to a linker sequence present in the vector (Fig. 2; preparation not published previously). Each reaction mixture consisted of a 100-µL solution containing the recombinant pBluescript II vector (50 ng), the corresponding reverse and forward primers (50 pmol each), and Ex Taq DNA polymerase (TaKaRa Bio; Kusatsu, Shiga, Japan) (2.5 U). V_H and V_L were amplified with the CS#3V_H-Rev and CS#3V_H-For (V_H) or CS#3V_L-Rev and CS#3V_L-For (V_L) primers (Table 1), respectively. The thermocycling profile was as follows: 35 cycles each of 95°C (1 min), 64°C (1 min), and 72°C (2 min), followed by a 10-min incubation at 72°C. The amplified V_H and V_L gene fragments were gel-purified and digested with NcoI and XhoI (V_H) or SalI and NotI (V_L), respectively, subcloned into the relevant sites in the pEXmide 7 vector, and propagated in Escherichia coli (E. coli) XL1-Blue cells (Agilent Technologies).

Fig. 2. Schematic Illustration of (A) the Strategy Used to Generate the V_H-Randomized and V_L-Randomized scFv Gene Libraries and Their Phage Display Using the pEXmide 7 Phagemid Vector, and (B) the Primary Structures of wt-scFv and the Improved scFvs

CDRs, including V_H-CDR1 (H1), -CDR2 (H2), -CDR3 (H3), V_L-CDR1 (L1), -CDR2 (L2), and -CDR3 (L3), as well as the linker sequence and amino acid substitutions are shown with 1-letter notation.

Construction of an scFv Mutant Library and Phage Display

scFv mutants were generated by error-prone PCR,²²⁾ as previously described.²³⁾ The pEXmide 7 vector containing the wt-scFv gene (0.1 ng) was prepared as described above, and mixed in a buffer solution (100 µL) with (0.10 or 0.20 mM) or without MnCl₂, with the CS#3V_H-Rev and CS#3V_H-For (or CS#3V_L-Rev and CS#3V_L-For) primers (Table 1) (0.10 nmol each), AmpliTaq DNA polymerase (Thermo Fisher Scientific; Waltham, MA, U.S.A.) (5 U), and an unbalanced mix of deoxyribonucleotide triphosphates (dNTPs), i.e., 0.10 µmol of each dNTP except for deoxyadenosine triphosphate (dATP) (0.020 µmol). This mixture was amplified for 35 cycles at 95°C (1 min), 50°C (1 min), and 72°C (3 min), followed by a 10-min extension at 72°C. The mutant V_H or mutant V_L genes generated using different MnCl₂ concentrations were combined and ligated into pEXmide 7, and transformed into XL1-Blue cells by electroporation. The resulting 2 bacterial libraries, V_H-mutated and V_L-mutated, were separately used to rescue phage particles, which were partially purified, combined as shown in Fig. 2A, and examined as follows.

Selection of Cortisol-Binding scFv-Phage Clones

MaxiSorp 12×75 mm polystyrene tubes (Thermo Fisher Scientific) were coated overnight at room temperature with 2.0 mL CS–BSA in 0.10 M carbonate buffer (pH 8.6) (0.50 µg/mL), and blocked with M-PBS-2, as described previously.^13,24) During each round of panning, antigen-coated tubes were incubated with phages (ca. 1–10×10¹¹ colony-forming units) in M-PBS-2 (2.0 mL) for 1.0 h at 37°C with continuous tumbling. Tubes were then washed 3 times with T-PBS-2 (4 mL), and bound phages were eluted by adding 1.0 mL of PVG-PBS containing 0.10 ng/mL cortisol. The recovered phage suspension was added to log-phase XL1-Blue cells in 2×YT medium containing 10 µg/mL tetracycline (9 mL), and then incubated at 37°C for 30 min. After centrifugation (10000×g, 10 min), the pellet was suspended in 2×YT medium (100 µL) and spread on a 90ϕ plate containing 2×YT agar supplemented with 100 µg/mL ampicillin, 10 µg/mL tetracycline, and 10 g/L glucose, and incubated overnight at 37°C. Colonies were then scraped into 2×YT medium containing 15% glycerol (1.5 mL) and a small aliquot was used for bacteriophage rescue with the aid of VCSM13 helper phage (Agilent Technologies), as described previously.^13,23,24) The resulting phages were used in the next round of selection.

Preparation of Soluble (Non-phage-Linked) scFvs

Recombinant plasmids were extracted from infected bacterial clones and digested with NcoI and NotI to cut down scFv genes therein, which were then ligated into a variant of pEXmide 7 vector with TAA TGA double stop codons after the FLAG sequence (pEXmide 7′ vector). The resulting plasmids were transformed into XL1-Blue cells, and transformants were grown and induced with isopropyl β-D-thiogalactopyranoside and sucrose.^13,24) Periplasmic extracts containing soluble scFv proteins were then prepared^13,24) and used for enzyme-linked immunosorbent assays (ELISAs) without further purification.

Competitive ELISA Using scFv-Phages

Ninety six-well RIA/EIA #3590 microplates (Corning, Corning, NY, U.S.A.) were coated overnight at room temperature with CS−BSA in 0.10 M carbonate buffer (pH 8.6) (1.0 µg/mL) and blocked with M-PBS at 37°C for 2.0 h.¹³⁾ Wells were then washed three times with T-PBS, and incubated at 37°C for 1.0 h with a mixture of cortisol dissolved in G-PBS (50.0 µL/well) and scFv-phages diluted with M-PBS (100 µL/well). Subsequently, wells were washed and probed using an anti-M13 antibody labeled with peroxidase (POD) (GE Healthcare UK Ltd., Amersham Place, Little Chalfont, Buckinghamshire, U.K.) diluted in PVG-PBS (0.50 µg/mL) (100 µL/well). After incubation at 37°C for 30 min, the wells were washed and captured POD activity was determined colorimetrically at 490 nm, using o-phenylenediamine as the hydrogen donor.^24,25) In these assays, the concentrations of scFv-phages were adjusted to give bound enzyme activities at B₀ (the reactions without unlabeled CS or analogs) of ca. 1.0–1.5 absorbance after a 30-min enzyme reaction.

Competitive ELISA Using Soluble scFv Proteins

Ninety six-well RIA/EIA #3590 microplates (Corning) were coated with CS−BSA and blocked as described above. Wells were incubated at 4°C for 2.0 h with a mixture of cortisol (or related compounds) dissolved in PVG-PBS (50.0 µL/well) and soluble scFv protein diluted with G-PBS (100 µL/well). Subsequently, wells were washed and probed with POD-labeled anti-FLAG M2 antibody (Sigma) diluted in PVG-PBS (0.20 µg/mL) (100 µL/well). After incubation at 37°C for 30 min, wells were washed and captured POD activity was determined as described above. In these assays, concentrations of scFvs were adjusted to give bound enzyme activities at B₀ of ca. 1.0–1.5 absorbance units after a 30-min enzyme reaction.

Measurement of Serum Cortisol Levels with Improved scFv

Commercially available control sera (Bio-Rad Lyphocheck Immunoassay Plus Control, Levels 1, 2, and 3) and serum samples obtained from healthy volunteers (5 males, aged 20–52 years old and 12 females, aged 18–24 years old) were each diluted with G-PBS, and submitted to the competitive ELISA (see above) using the improved scFv (scFv#m1-L10; soluble proteins). For the measurement of serum specimens from the volunteers, protocol approval was obtained from an institutional review board.

RESULTS AND DISCUSSION

Generation of the Wild-Type Anti-cortisol scFv

Improvement of antibody molecules by constructing genetically-engineered antibody libraries requires conversion of target antibodies into the relevant smaller fragments, i.e., scFv or Fab fragments, to facilitate their bacteria expression. Thus, we first generated the wild-type scFv derived from mouse anti-cortisol antibody Ab-CS#3 by subcloning the relevant V_H and V_L genes in the pEXmide 7 phagemid vector (Fig. 2A), in which the V_H and V_L genes were combined via the linker sequence encoding (GlyGlyGlyGlySer)₃ to assemble scFv genes. DNA encoding FLAG tag²⁶⁾ was added to the 3′-end to facilitate purification and detection. This wt-scFv gene, with the structure 5′-V_H-linker-V_L-FLAG-3′, was expressed in E. coli cells. Typically, a 1-L culture of the transformed bacteria yielded ca. 200 µg scFv protein, the characteristics of which are described below.

In Vitro Affinity Maturation of scFv

Previously, we generated several improved scFvs against estradiol-17β,^13,14) cotinine,¹⁵⁾ and Δ⁹-tetrahydrocannabinol²⁷⁾ with higher affinities by the antibody breeding (in vitro molecular evolution) strategy. In these studies, the entire V_H and V_L genes were separately randomized by error-prone PCR, and then combined in a shuffling manner to assemble scFv genes before being expressed as phage-displayed proteins. In the present study, we randomly mutated either the V_H or V_L gene to investigate which domain is more effective as target of mutagenesis based on improved antigen-binding affinities in the scFvs.

Thus, the V_H and V_L regions of the wt-scFv gene were amplified separately by single error-prone PCR under the “dATP-diminished” condition,¹⁴⁾ and the resulting V_H or V_L randomized genes were spliced with the wild-type gene encoding the other domain to create 2 diverse sets of scFv genes (Fig. 2A). Transformation of E. coli cells with these scFv mutants generated “V_H-randomized” and “V_L-randomized” bacterial libraries, each of which contained ca. 10⁶ transformants with scFv genes. From these libraries, phage particles were separately rescued, combined, and then subjected to 3 rounds of panning against polystyrene tubes coated with CS–BSA. After the third panning, 96 phage clones were randomly selected from the bacterial libraries infected with recovered phages, and screened for cortisol-binding ability. Consequently, 3 improved clones, each displaying an scFv mutant, named scFv#m1-L7 scFv#m1-L9, and scFv#m1-L10, were isolated. As shown below, these scFvs had amino acid substitution in the V_L region, thus were generated from the V_L-randomized library.

Characterization of Wild-Type and Mutant scFvs

In the ELISAs, involving competitive reactions between immobilized cortisol residues and standard cortisol, the 3 improved phage clones generated dose–response curves with >10-fold higher sensitivity based on comparison of the midpoint (i.e., the amount of standard cortisol required for 50% binding) than with the wild-type phage (Fig. 3A). The primary structures of the wild-type and improved mutant scFvs are illustrated in Fig. 2B, with the sequences of the complementarity-determining regions (CDRs) shown.²⁸⁾ The V_H and V_L sequences of these scFvs belonged to subgroup IIB and III, respectively, based on the Kabat definition²⁸⁾: thus, their structures should be rather different from that of the anti-cortisol scFv derived from Ab#10 (its V_H and V_L sequences belonged to IB and II, respectively) that we previously reported.²⁰⁾ We found that 5, 2, and 3 amino acids were substituted in scFv#m1-L7, scFv#m1-L9, and scFv#m1-L10, respectively. Interestingly, the Cys^L49Ser substitution in the second framework region in the V_L domain (defined based on the Kabat definition)²⁸⁾ occurred commonly in these mutants. In the antibody variable domains, Cys residues generally appear only at fixed positions (for V_H, 22 and 92; for V_L, 23 and 88) and are strongly conserved to form intra-chain disulfide bonds necessary for adequate protein folding.²⁸⁾ Thus, the Cys at the position 49 in V_L, where highly conserved Tyr residue appears for native antibodies, is quite an unusual residue and may have reflected an obstacle towards higher affinity with the wild-type sequence.

Fig. 3. Dose–Response Curves for Cortisol ELISAs (A) Performed Using Phages Displaying wt-scFv (●), scFv#m1-L7 (▲), scFv#m1-L9 (■), and scFv#m1-L10 (◆), or (B) Using Soluble wt-scFv (●) and scFv#m1-L10 (◆) Proteins and the Parent Mouse Antibody Ab-CS#3 (×)

The midpoints of the dose–response curves (ng/assay) are shown in parentheses. The vertical bars indicate the standard deviation (n=4).

Because scFv#m1-L10 showed the highest sensitivity among the phage-displayed scFvs (Fig. 3A), it was then prepared as a soluble protein and further characterized in comparison with the soluble form of wt-scFv. On sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, affinity-purified wt-scFv and scFv#m1-L10 proteins migrated as single bands with nearly the expected relative molecular mass (M_r) values of 27289 and 27227 (calculated based on their primary amino acid sequences), respectively (Fig. 4A). The K_a was determined at 4°C by Scatchard analysis²⁹⁾ using tritium-labeled cortisol as a tracer.¹⁹⁾ The mutant exhibited a K_a of 1.2×10¹⁰ M⁻¹, which was >30-fold greater than the K_a of wt-scFv (3.4×10⁸ M⁻¹) (Fig. 4B). Protein modeling suggested that 4 CDRs, i.e., V_H-CDR3, V_L-CDR1, 2, and 3 may be located within close proximity to the antigen (cortisol). Although no apparent alteration in paratope conformation was observed, cortisol docked with the paratope from different angle between the before and after the amino acid substitution (Fig. 4C).

Fig. 4. Characterization of the Soluble scFv Proteins

(A) SDS-PAGE with Coomassie brilliant blue staining after purification by affinity chromatography: lane 1, M_r marker; 2, wt-scFv; 3, scFv#m1-L10. (B) Scatchard analysis²⁹⁾ for wt-scFv (●) and scFv#m1-L10 (◆). A constant amount of each scFv was incubated at 4°C for 2.0 h with increasing amounts of cortisol, with a trace quantity of [³H]-cortisol (ca. 250 Bq). The bound and free fractions were separated by a dextran-coated charcoal method. The vertical bars indicate the standard deviation (n=4). (C) scFv protein modeling. Protein ribbon structures for wt-scFv and scFv#m1-L10 were constructed using the SWISS-MODEL Protein Modeling Server,³⁴⁾ and their conformations when docked to cortisol were predicted using SwissDock.³⁵⁾ In the ribbon representation of the scFv backbones, V_H-CDR1 (H1; yellow), -CDR2 (H2; orange), -CDR3 (H3; magenta), V_L-CDR1 (L1; dark blue), -CDR2 (L2; light green), and -CDR3 (L3; light blue) are shown with β-sheet structures (bold gray arrows). The 3 amino acids changed after the mutation are indicated with arrows. The backbone of the cortisol molecule is shown in light purple.

In the ELISA experiments, soluble wt-scFv showed a dramatically more sensitive dose-response curve with a lower midpoint (1.1 ng/assay) (Fig. 3B) than the phage-displayed wt-scFv and parent antibody Ab-CS#3 (midpoints: 100 and 28 ng/assay, respectively; Figs. 3A, B). This increase in assay sensitivity, when wt-scFv was used, could be explained based on not only the greater K_a of wt-scFv than the parent IgG-form antibody (K_a, 4.7×10⁷ M⁻¹) (it is unclear why the scFv form showed stronger affinity than the relevant IgG form),²⁰⁾ but also its monovalent structure, which is less susceptible to avidity effects that should occur in ELISA systems based on the binding of multiply immobilized antigens on microplates and bivalent IgG-form antibodies.²⁵⁾ Although it is unlikely that wt-scFv-displaying phages contained a considerable fraction of virions displaying more than 2 scFvs, several virions each having a single scFv might aggregate to form multi-valent complexes.

The scFv#m1-L10 mutant showed a dose-response curve with 7.3-fold higher sensitivity than that of wt-scFv based on comparison of midpoints (0.15, 1.1 ng/assay for the mutant and wild-type, respectively), measuring range of which was ca. 0.03–0.6 ng/assay (Fig. 3B). The limit of detection (0.014 ng/assay) was determined as the amount of cortisol required for a bound absorbance 2 standard deviations below the average (n=10) of the B₀ absorbance.

The cross-reactivities of wt-scFv and scFv#m1-L10 were determined for 8 endogenous and 3 synthetic steroids (Fig. 5). Although entire recognition patterns were similar for both scFvs, the mutant exhibited higher specificity as shown with decreased cross-reactivities with cortisone, prednisolone, and prednisone. Almost negligible cross-reactivity with corticosterone (<0.2%) and aldosterone (<0.01%) indicated that both scFvs recognized modifications at the C-17 and C-18 positions of cortisol. Regarding the C-11 position, however, these scFvs lacked satisfactory recognition. 11-Deoxycortisol, the biosynthetic precursor of cortisol lacking the 11-hydroxy group, showed rather stronger reactivity than cortisol (>100%). Cortisone, an active metabolite with a ketone group at the C-11 position, also cross-reacted for wt-scFv (45%), but this was significantly improved after the mutation (17%). Among the exogenous steroids, prednisolone with an additional carbon–carbon double bond exhibited 18% cross-reactivity for wt-scFv, which was dramatically improved for the mutant scFv (3.3%).

Fig. 5. Percent Cross-Reactivity of scFvs in ELISA Determined by the 50% Displacement Method^27,36) According to the Following Equation

Cross reactivity (%)=(X/Y)×100; where X is the midpoint (ng/assay) of cortisol, and Y is the midpoint (ng/assay) of a related steroid cross-reactivity of which is to be determined. The competitive reactions in ELISA were performed at 4°C for 2.0 h.

Considering these recognition patterns, we expected that the mutant scFv could be applicable for clinical use, except in case where patients were under metyrapone-stimulation tests or administered high-dose of prednisolone: in the former conditions, serum 11-deoxycortisol levels dramatically increase.³⁰⁾ In normal subjects, however, the 11-deoxycortisol levels are usually much lower (<10 ng/mL)³¹⁾ than cortisol levels (10−250 ng/mL),³¹⁾ and thus this precursor should not cause substantial overestimation. Recently, the average cortisone/cortisol concentration ratio in serum was reported to be 0.225.³²⁾ Taking into account of the cross-reactivity of the mutant scFv (see above), overestimation due to cortisone, if any, would not cause serious problem.

In fact, for commercially available control sera, we obtained acceptable assay values with a reasonable parallelism between the sample dilution rates (Fig. 6). Average serum levels for healthy male and female subjects were 23±5.5 (n=5) and 23±5.6 (n=12) µg/dL (total levels, 23±5.4 µg/dL), respectively, which agreed with the reference ranges.

Fig. 6. Serial Dilution Study for Measuring Serum Specimens

Control sera (levels 1, 2, 3; see text) were serially diluted and analyzed by ELISA using scFv#m1-L10. The competitive reactions were performed at 4°C for 2.0 h. The vertical bars indicate the standard deviation (n=4).

CONCLUSION

Previously, we generated a mutant scFv clone against the nicotine metabolite, cotinine, with >40-fold enhanced antigen-binding affinity (calculated as the K_a).¹⁵⁾ This was achieved via “1-shot” in vitro evolution (i.e., via single-step mutagenesis and subsequent selection) and resulted in 5 amino acid substitutions. Here, we succeeded again in improving the K_a by >30-fold by single-step mutagenesis and selection with an scFv targeting another small biomarker, cortisol, with only 3 amino acid substitutions concentrated on the V_L domain. In this study, we selected a mouse anti-cortisol antibody (Ab-CS#3) with a practically insufficient affinity as the prototype. As expected, the antibody-breeding approach improved the affinity to produce a novel species with practical binding performance.

A limitation of this study is that cannot explain why the V_H-randomized library did not generate any improved scFv clones. We are unsure whether no improved clones were actually generated, or whether this outcome simply depended on inadequate experimental parameters. Next, we will explore affinity-maturation of the present V_L-substituted mutants by randomizing its V_H domain. The phagemid vector pEXmide 7 used in this study, which allows for independent subcloning of randomized V_H and V_L genes, is suited for such a chain-shuffling strategy.³³⁾

Acknowledgments

This work was supported in part by Grants from the Japan Society for the Promotion of Science (JSPS) (JSPS KAKENHI Grant Number JP16K08954) and the Science Research Promotion Fund of the Japan Private School Promotion Foundation. We would like to thank Dr. Eskil Söderlind (Avena Partners AB, Sweden) and Professor Carl A. K. Borrebaeck (Lund University, Sweden) for providing the pEXmide 5 vector and allowing us to modify it to generate pEXmide 7 and pEXmide 7′ vectors.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) The Immunoassay Handbook: Theory and applications of ligand binding, ELISA and related techniques, 4th Edition. (Wild D ed.) Elsevier, the Netherlands (2013).
• 2) Köhler G, Milstein C. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature, 256, 495–497 (1975).
• 3) Therapeutic Monoclonal Antibodies: From bench to clinic. (Zhiqiang A ed.) Wiley, Hoboken, NJ, U.S.A. (2009).
• 4) Kobayashi N, Goto J. Noncompetitive immunoassays for small molecules with high sensitivity and specificity. Adv. Clin. Chem., 36, 139–170 (2001).
• 5) Kobayashi N, Oyama H. Antibody engineering toward high-sensitivity high-throughput immunosensing of small molecules. Analyst, 136, 642–651 (2011).
• 6) Antibody Engineering: Methods and protocols. (Lo BKC ed.) Humana Press, Totowa, NJ (2004).
• 7) Handbook of therapeutic antibodies. (Dübel S ed.) Wiley-Blackwell, Hoboken, NJ (2010).
• 8) Bradbury ARM, Sidhu S, Dübel S, McCafferty J. Beyond natural antibodies: the power of in vitro display technologies. Nat. Biotechnol., 29, 245–254 (2011).
• 9) Chiu ML, Gilliland GL. Engineering antibody therapeutics. Curr. Opin. Struct. Biol., 38, 163–173 (2016).
• 10) Skerra A, Plückthun A. Assembly of a functional immunoglobulin Fv fragment in Escherichia coli. Science, 240, 1038–1041 (1988).
• 11) Bird RE, Hardman KD, Jacobson JW, Johnson S, Kaufman BM, Lee SM, Lee T, Pope SH, Riordan GS, Whitlow M. Single-chain antigen-binding proteins. Science, 242, 423–426 (1988).
• 12) Ahmad ZA, Yeap SK, Ali AM, Ho WY, Alitheen NBM, Hamid M. scFv antibody: Principles and clinical application. Clin. Dev. Immunol., 2012, 980250 (2012).
• 13) Kobayashi N, Oyama H, Kato Y, Goto J, Söderlind E, Borrebaeck CAK. Two-step in vitro antibody affinity maturation enables estradiol-17β assays with more than 10-fold higher sensitivity. Anal. Chem., 82, 1027–1038 (2010).
• 14) Oyama H, Yamaguchi S, Nakata S, Niwa T, Kobayashi N. “Breeding” diagnostic antibodies for higher assay performance: A 250-fold affinity-matured antibody mutant targeting a small biomarker. Anal. Chem., 85, 4930–4937 (2013).
• 15) Oyama H, Morita I, Kiguchi Y, Banzono E, Ishii K, Kubo S, Watanabe Y, Hirai A, Kaede C, Ohta M, Kobayashi N. One-shot in vitro evolution generated an antibody fragment for testing urinary cotinine with more than 40-fold enhanced affinity. Anal. Chem., 89, 988–995 (2017).
• 16) Chrousos GP, Kino T, Charmandari E. Evaluation of the hypothalamic–pituitary–adrenal axis function in childhood and adolescence. Neuroimmunomodulation, 16, 272–283 (2009).
• 17) Crichton D, Grattage L, McDonald A, Corrie JET, Steel CM, Hubbard AL, Al-Dujaili EAS, Edwards CRW. The production and assessment of monoclonal antibodies to cortisol. Steroids, 45, 503–517 (1985).
• 18) Lewis JG, Manley L, Whitlow JC, Elder PA. Production of a monoclonal antibody to cortisol: application to a direct enzyme-linked immunosorbent assay of plasma. Steroids, 57, 82–85 (1992).
• 19) Kobayashi N, Sun P, Fujimaki Y, Niwa T, Nishio T, Goto J, Hosoda H. Generation of a novel monoclonal antibody against cortisol-[C-4]-bovine serum albumin conjugate: application to enzyme-linked immunosorbent assay for urinary and serum cortisol. Anal. Sci., 18, 1309–1314 (2002).
• 20) Oyama H, Morita I, Kiguchi Y, Miyake S, Moriuchi A, Akisada T, Niwa T, Kobayashi N. Gaussia luciferase as a genetic fusion partner with antibody fragments for sensitive immunoassay monitoring of clinical biomarkers. Anal. Chem., 87, 12387–12395 (2015).
• 21) Jirholt P, Ohlin M, Borrebaeck CAK, Söderlind E. Exploiting sequence space: shuffling in vivo formed complementarity determining regions into a master framework. Gene, 215, 471–476 (1998).
• 22) Leung DW, Chen E, Goeddel DV. A method for random mutagenesis of a defined DNA segment using a modified polymerase chain reaction. Technique, 1, 11–15 (1989).
• 23) Kobayashi N, Kato Y, Oyama H, Taga S, Niwa T, Sun P, Ohtoyo M, Goto J. Anti-estradiol-17β single-chain Fv fragments: generation, characterization, gene randomization, and optimized phage display. Steroids, 73, 1485–1499 (2008).
• 24) Kobayashi N, Shibahara K, Ikegashira K, Shibusawa K, Goto J. Single-chain Fv fragments derived from an anti-11-deoxycortisol antibody: affinity, specificity, and idiotype analysis. Steroids, 67, 733–742 (2002).
• 25) Kobayashi N, Banzono E, Shimoda Y, Oyama H, Kunihiro T, Morita I, Ohta M. A monoclonal antibody-based enzyme-linked immunosorbent assay for human urinary cotinine to monitor tobacco smoke exposure. Anal. Methods, 3, 1995–2002 (2011).
• 26) Knappik A, Plückthun A. An improved affinity tag based on the flag^® peptide for the detection and purification of recombinant antibody fragments. Biotechniques, 17, 754–761 (1994).
• 27) Morita I, Oyama H, Yasuo M, Matsuda K, Katagi K, Ito A, Tatsuda H, Tanaka H, Morimoto S, Kobayashi N. Antibody fragments for on-site testing of cannabinoids generated via in vitro affinity maturation. Biol. Pharm. Bull., 40, 174–181 (2017).
• 28) Kabat EA, Wu TT, Perry HM, Gottesman KS, Foeller C. Sequences of Proteins of Immunological Interest; U.S. Department of Health and Human Services, National Institutes of Health, U.S. Government Printing Office: Washington, DC (1991).
• 29) Scatchard G. The attractions of proteins for small molecules and ions. Ann. N. Y. Acad. Sci., 51, 660–672 (1949).
• 30) “Metyrapone stimulation tests.”: ‹https://www.uptodate.com/contents/metyrapone-stimulation-tests›
• 31) Oyama H, Tanaka E, Kawanaka T, Morita I, Niwa T, Kobayashi N. Anti-idiotype scFv−enzyme fusion proteins: A clonable analyte-mimicking probe for standardized immunoassays targeting small biomarkers. Anal. Chem., 85, 11553–11559 (2013).
• 32) Lee S, Lim H-S, Shin H-J, Kim S-A, Park J, Kim H-C, Kim H, Kim HJ, Kim Y-T, Lee K-R, Kim Y-J. Simultaneous determination of cortisol and cortisone from human serum by liquid chromatography-tandem mass spectrometry. J. Anal. Methods Chem., 2014, 787483 (2014).
• 33) Marks JD, Griffiths AD, Malmqvist M, Clackson TP, Bye JM, Winter G. By-passing immunization: building high affinity human antibodies by chain shuffling. Biotechnology, 10, 779–783 (1992).
• 34) Guex N, Diemand A, Peitsch MC. Protein modelling for all. Trends Biochem. Sci., 24, 364–367 (1999).
• 35) Grosdidier A, Zoete V, Michielin O. SwissDock, a protein-small molecule docking web service based on EADock DSS. Nucleic Acids Res., 39 (Suppl.), W270–W277 (2011).
• 36) Abraham GE. Solid-phase radioimmunoassay of estradiol-17β. J. Clin. Endocrinol. Metab., 29, 866–870 (1969).