Retrieving Dissociation-Resistant Antibody Mutants: An Efficient Strategy for Developing Immunoassays with Improved Sensitivities

Abstract

Previously, we generated high-affinity antibody mutants that enabled sensitive immunoassays by exploring diverse libraries of single-chain Fv fragments (scFvs) displayed on bacteriophage. To isolate rarely-occurring desirable clones, “panning” has commonly been performed but is often unsuccessful. Therefore, we previously developed a clonal array profiling (CAP) method, wherein scFv-displaying phage (scFv-Ph) clones in a library were examined individually regarding their ability to target antigens immobilized on microwells. Clones that showed strong reactivity were recovered via dissociation using an acidic treatment. The CAP successfully discovered cortisol-specific scFvs showing 17–31-fold improved K_a from libraries generated via site-directed insertions in a prototype anti-cortisol scFv (wt-scFv; K_a, 3.6 × 10⁸ M⁻¹), but their K_a did not exceed 1.1 × 10¹⁰ M⁻¹. In this study, to break this possible affinity ceiling, we devised a new system employing a dissociation-independent recovery. scFv-Phs were individually reacted to target antigen (cortisol) immobilized on microwells via a linker containing a disulfide bond. Following acidic and basic treatments to eliminate scFv-Phs with “ordinary affinities,” dissociation-resistant scFv-Phs remaining on the microwells were retrieved via reductive cleavage of the disulfide bonds. This system allowed for a straightforward and efficient discovery of scFv mutants with 33–56-fold increased K_a (1.2–2.0 × 10¹⁰ M⁻¹), exceeding the previous affinity ceiling. These scFvs enabled an enzyme-linked immunosorbent assay for cortisol with 18–51-fold higher sensitivity than the assay performed using wt-scFv.

INTRODUCTION

Genetic engineering of antibody molecules to produce evolved species has enabled the development of more sensitive immunoassays than those depending on conventional in vivo antibodies.^1–9) In our previous study,^3–9) a prototype antibody was converted to single-chain Fv fragments (scFvs) and then randomly mutagenized to generate a diverse library, each member of which was displayed on M13 bacteriophage particles.^10,11) The scFv-displaying phages (scFv-Phs) with improved antigen-binding characteristics were isolated via panning: all the members are simultaneously reacted with a limited amount of immobilized antigens (Fig. 1A). Desirable species, obtained as the antigen-bound fraction, were proliferated using infecting bacteria for further investigations.

Fig. 1. Principle of Conventional Panning and the Original CAP System¹²⁾

Both methods begin with (i) the transformation of E. coli cells with mutagenized scFv genes. Next, (ii) transformants are grown overnight on agar to serve as the initial library with “undisturbed” diversity. (A) Conventional panning: (iii) The bacterial library is scraped from the agar, and re-cultivated in liquid medium for (iv) phage rescue, increasing clonal bias mainly due to differences in bacterial and/or phage propagation abilities. (v) The resulting polyclonal scFv-Phs are collectively reacted with a limited number of immobilized antigens. (B) CAP system: (iii′) The bacterial clones on the agar are individually cultivated in microwells containing the helper phage. Subsequently, each clone propagates and produces scFv-Phs without competing with different clones. (iv′) The scFv-Phs specific to target antigen bind to pre-immobilized antigen, (v') and are detected via a bioluminescence assay using fusion proteins linking an anti-phage scFv with luciferases. (vi′) The signal-intensity data obtained from the microwells are ranked, and scFv-Phs that generated stronger luminescence are retrieved from microwells and propagated for subsequent characterization of displayed scFvs.

However, the panning did not always function as expected, and often failed to provide improved scFv-Phs. We previously estimated that these problems might be due to the interference of excess scFv-Phs with weaker affinities that competitively prevent the binding of desirable scFv-Phs to the immobilized antigens.¹²⁾ To overcome these panning-inherent limitations, we developed a “clonal array profiling (CAP)” method that enabled efficient recovery of affinity-matured scFv-Phs¹²⁾ (Fig. 1B). Therein, phage clones generated from the initial bacterial libraries were examined individually in microwells, without disturbing the original diversity. Namely, scFv-Ph progenies generated from each bacterial clone were, if they showed sufficient affinity, captured via antigen pre-immobilized on the microwell and detected via a phage-specific bioluminescent assay. Subsequently, promising scFv-Phs that generated high luminescence were recovered for further investigation following dissociation from the immobilized antigens using an acidic treatment.

The CAP successfully isolated 14 unique scFvs that exhibited satisfactorily improved affinity to cortisol from libraries composed of scFvs having insertions of randomized amino acids at their V_H domain¹³⁾ (Figs. 2A, B). These scFvs showed 17–31-fold higher equilibrium affinity constants (K_a) over the prototype scFv (wt-scFv; K_a, 3.6 × 10⁸ M⁻¹),¹³⁾ but none of them showed a K_a exceeding 1.1 × 10¹⁰ M⁻¹ (Fig. 2C). We assumed that there might be an affinity ceiling that is associated with the stringency of the acidic dissociation conditions that we employed. Therefore, we devised a CAP employing a dissociation-independent recovery (DIR) step based on reductive cleavage of disulfide (SS) bonds^14,15) (Fig. 3). According to the original CAP system, the generated monoclonal scFv-Phs were individually reacted to a target antigen (cortisol) that had been immobilized via a linker containing an SS bond. Next, scFv-Phs showing ordinarily-high, moderate, and low affinities were eliminated using serial acidic and basic treatments. Subsequently, dissociation-resistant scFv-Phs that remained bound to cortisol moieties were recovered via an SS-bond cleavage using dithiothreitol (DTT). This CAP/DIR system allowed for a more straightforward and efficient discovery of satisfactory affinity-matured scFv mutants that enabled more sensitive immunoassays than the original CAP.

Fig. 2. Summary of Our Previous Results That Prompted Us to Perform the Current Study

(A) Structure and affinity of anti-cortisol scFv (wt-scFv) used as the prototype for mutagenesis.⁷⁾ Sequence of amino acids at the 1–10 positions of the V_H domain is shown. The V_H and V_L domains were linked with a linker sequence VSS(GGGGS)₃T. (B) The design of sub-libraries sl-1–6 comprising randomized amino acid(s) between the 6 and 7 positions of the V_H domain. The letter “X” means that any one of the 20 proteinogenic amino acid can appear.¹³⁾ (C) Inserted amino acid(s) and affinity of the improved scFvs showing K_a values higher than 9.0 × 10⁹ M⁻¹, which were selected from the 14 affinity-matured scFvs (see text).¹³⁾

Fig. 3. Principle and Procedure of the CAP/DIR System

CS-SS-bio (A) (500 pg/well) was incubated at 37 °C for 60 min in 96-well Costar#3922 white microplates coated with streptavidin. After washing with T-PBS buffer,^4–8) a 2 × YT medium containing KM13 helper phage (approx. 5 × 10⁸ pfu/mL), ampicillin, and kanamycin (“CAP medium”) was added (200 µL/well), and bacterial colonies were picked and dipped individually in each well (i). After continuous shaking (800 rpm) at 25 °C for 45 h, the microplates were washed with T-PBS-2 buffer^4–8) and, without checking the binding profile using the bioluminescent assay, incubated serially with 0.10 M glycine-HCl (pH 2.2) followed by 0.10 M triethylamine (pH 12) at room temperature for 10 min. After washing, the microplates were probed with an anti-phage scFv²²⁾ fused with NanoLuc luciferase (anti-Ph-NLuc),²³⁾ for profiling luminescence intensities (C) (ii). The microwells exhibited a strong luminescence (i.e., Nos. 1, 5, 16) were washed and incubated with 50 mM DTT (100 µL/well) at room temperature for 60 min (iii). An aliquot of the solution in the microwells (10 µL) was mixed with a log-phase culture of E. coli TG1 cells (90 µL) and incubated at 37 °C for 30 min to propagate eluted scFv-Phs (i.e., Nos. 1, 5, 16) in the Costar#3922 microplates pre-coated with CS-BSA (B) and the microplates pre-coated with BSA in a parallel manner. Following incubation, CAP medium was added (100 µL/well), and the microplates were shaken (800 rpm) at 25 °C for 45 h. Finally, all microplates were washed again and scFv-Phs specific to the cortisol residues (i.e., No. 5) were monitored using anti-Ph-NLuc (iv). The assay results for several scFv-Phs are shown in Supplementary Fig. S3. The scFv-Ph clone (No. 5) is evaluated to be promising for developing sensitive immunoassays. It should be mentioned that the indirectly immobilized CS-SS-bio on microplates retained its initial reactivity against an anti-cortisol scFv, suggesting that no significant deterioration in its chemical structure occurred after the 45 h incubation at 25 °C regardless of the presence or absence of growing TG1 cells generating scFv-Phs (Supplementary Fig. S5).

MATERIALS AND METHODS

Preparation of the Library of scFv Mutants with Amino-Acid Insertions

The scFv sub-library 4/5 (sl-4/5), where each member has an insertion of serial 4 or 5 randomized amino acids at the framework region (FR) 1 of the V_H domain (V_H-FR1) (Figs. 2A, B), was prepared.¹³⁾ The anti-cortisol wt-scFv gene⁷⁾ was PCR-amplified using reverse primer NNS-4 or NNS-5 comprising serial 4 or 5 NNS [i.e., (A/C/G/T)(A/C/G/T)(C/G)] degenerated codons and forward primer CS#3V_L-For⁷⁾ (Supplementary Table S1). The resulting scFv gene libraries (sl-4 and sl-5) were subcloned in the pEXmide 7 vector⁷⁾ and transformed into Escherichia coli (E. coli) TG1 cells by electroporation. The transformant suspension was spread on agar plates, followed by incubation at 37 °C overnight. Colonies grown were subjected to CAP/DIR.

CAP/DIR System for Isolating Affinity-Matured scFv-Phs

scFvs exhibiting satisfactorily-high affinities were isolated as shown in Fig. 3. The key reagent that enabled DIR system, i.e., cortisol-biotin conjugate containing a linker with an SS bond (CS-SS-bio; Fig. 3A) was synthesized, and the reductively-cleavable nature of its linker was confirmed (Supplementary Fig. S1).

Preparation and Characterization of Soluble scFvs

Selected scFv-Phs were transformed into soluble (non-phage-linked) scFv proteins each having a FLAG tag at their C-terminus, via subcloning of the corresponding scFv genes in pEXmide 7′ vector.⁷⁾ The K_a values of these scFvs were determined by the Scatchard analysis¹⁶⁾ using a [³H]-labeled cortisol^12,13) (Supplementary Fig. S2). Competitive enzyme-linked immunosorbent assays (ELISAs) were performed using 96-well microplates coated with a conjugate of cortisol and bovine serum albumin (CS-BSA)⁷⁾ (Fig. 3B). In the microwells, cortisol standard and the soluble scFv protein were incubated, and the bound scFvs were monitored colorimetrically with the aid of anti-FLAG M2 antibody,^7,12,13) as referred to in Fig. 4.

Fig. 4. Characterization of Selected scFv Mutants

(A) The inserted amino acids, K_a values, and midpoints of dose-response curves in the following ELISA are listed for the selected scFvs. scFv#n5-260 and n5-363 were found separately but had an identical inserted amino acids. a) scFv#n4-177, n4-285, n5-116, n5-195, and n5-220 did not show the affinity that could be determined. b) We previously reported that the midpoint observed with wt-scFv as 706 pg/assay.¹³⁾ The difference from the current value (512 pg/assay) is owing to differences in production lot of CS-BSA and incubation conditions (see below). (B) Typical dose-response curves for cortisol in a competitive ELISA format. The upper group is for scFvs having 4 inserted amino acids, while the lower for scFvs having 5 inserted amino acids. The ELISA was performed in 96-well microplates (Costar#3590) coated with CS-BSA (Fig. 3B) and blocked with Block Ace (KAC).¹²⁾ Cortisol standard and the soluble scFv protein, both dissolved in G-PBS buffer,^4–8) were added to the microwells (50 and 100 µL/well, respectively) and incubated at 4 °C for 120 min. After washing with T-PBS buffer, the bound scFvs were probed with anti-FLAG M2 antibody labeled with peroxidase (POD) and the captured POD activity was determined colorimetrically using o-phenylenediamine as chromogen.^4–8) The unit “X g/assay” was used in the abscissa, which refers to the total mass (X g) of the analyte added to each microwell for the antigen-antibody reactions. The vertical bars indicate the standard deviation (n = 4). The scFv concentrations were adjusted to give bound enzyme activities at B₀ (the reaction without cortisol standard) of approximately 0.50 absorbance after a 30-min enzyme reaction (cf., in our previous study,¹³⁾ this was adjusted to give the B₀ enzyme activities of approximately 1.0–1.5 absorbance). The background absorbance (observed without addition of scFvs) was lower than 5.0% of the B₀ absorbance. (C) Cross-reactivity of scFv#n4-335 and n5-325 that enabled the most and the second most sensitive ELISA of cortisol is shown in comparison with the data of wt-scFv (which was reported previously). Chemical structures of cortisol and related endogenous steroids tested here are shown in Supplementary Fig. S6.

RESULTS

Background of the Present Study

We previously explored the sub-libraries sl-1–3, sl-4/5, sl-6, where the scFv members had an insertion between the positions 6 and 7¹⁷⁾ in the V_H-FR1 of wt-scFv with a single or 2–6 serial amino acid(s), randomized to present any of the 20 proteinogenic amino acids¹³⁾ (Figs. 2A, B). The CAP system successfully isolated affinity-matured species (Fig. 2C), but there seemed to be an upper limit of the K_a (i.e., 1.1 × 10¹⁰ M⁻¹) for these scFvs. In fact, using a different anti-cortisol scFv library with patternized amino-acid substitutions in their V_H-FR1, the CAP also provided three mutants showing a K_a of 1.1 × 10¹⁰ M⁻¹ as the most improved species.¹³⁾ In these studies, we recovered bound scFv-Phs using treatment with 0.10 M glycine–HCl buffer (pH 2.2), a common method for dissociating antigen-antibody complexes. Considering this as a probable cause, we devised the CAP/DIR system in which high-affinity scFv-Phs were “plucked off” together with complexed cortisol residues by cleaving the linker connecting cortisol and microwells. In the previous study, scFvs showing a K_a in the 10¹⁰-range were isolated from the sub-libraries sl-1–3 and sl-4/5¹³⁾ (Fig. 2C). Here we applied the CAP/DIR system for mining the sub-library sl-4/5.

Discovery of Affinity-Matured scFvs via the CAP/DIR System

Randomly-selected colonies (n = 4700; only 0.09% from the initial transformants that contained 5.1 × 10⁶ colony-forming unit) were subjected to the CAP/DIR (Fig. 3). After the acidic and basic treatments, dissociation-resistant scFv-Phs were detected via the bioluminescent assay. The 374 clones that generated >100000 arbitrary units of luminescence (Fig. 3C) were retrieved with the DTT treatment and propagated, and their binding specificity to cortisol residues in the CS-BSA-coated microplates was assessed (Fig. 3, Supplementary Fig. S3). This step was necessary because a part of retrieved clones was unexpectedly exhibited similar or even higher binding to BSA, compared with the binding to cortisol residues: this property crucially deteriorated the sensitivity of the cortisol ELISA. The top 20 scFv-Phs that showed stronger cortisol-binding signals were converted to soluble scFvs and characterized in detail. We confirmed that these scFv-Phs were actually resistant to the previous dissociation condition with 0.10 M glycine–HCl (pH 2.2): >70% cortisol-binding activity was observed when considering the activity under pH 7.0 as 100% (data not shown).

Analytical Performances of the Isolated scFv Mutants

As shown in Fig. 4A, 15 among the 20 soluble scFvs exhibited K_a values to cortisol ranging 1.2–2.0 × 10¹⁰ M⁻¹, thereby exceeding the previous “ceiling” of 1.1 × 10¹⁰ M⁻¹. Regarding insertion, 5 among the 15 scFvs had 4 amino acids, and the remaining 10 had 5 amino acids, but no common motif was found. None of the 15 scFvs were identical with those found in the previous study¹³⁾ (Fig. 2C). Three scFvs among the 5 “inactive” species were the products of genes having single or double nonsense codons. In the case of scFv-Ph generation, sufficient amount of intact scFv-pIII fusion proteins might have accumulated, via partial read through of the nonsense codon(s)¹²⁾ during the 45-h incubation to generate cortisol-binding phage particles. No explanation is currently available for the remaining 2 inactive scFvs.

In the ELISA, all 15 scFvs showing binding activity exhibited significantly more sensitive dose-response curves than the wt-scFv (Fig. 4B), generating 18–51-fold lower midpoints (Fig. 4A). scFv#n5-325 (K_a, 1.3 × 10¹⁰ M⁻¹) provided dose-response curves with the lowest midpoint (10.1 pg/assay). scFv#n4-335, which showed the highest K_a, resulted in the second lowest midpoint (11.7 pg/assay). Figure 4A indicated the fact that scFvs showing higher K_as did not always exhibit lower midpoints (e.g., scFv#n4-265, n5-100, and n5-260). This might be attributed to the possible recognition of the bridge (linker) structure connecting cortisol and BSA in the conjugate coated on the microplates.¹⁸⁾ Cross-reactivity with endogenous steroids showed that scFv#n4-335 and n5-325 have been improved for recognition of 11-modified corticosteroids (i.e., cortisone and 11-deoxycortisol), compared with the parental wt-scFv (Fig. 4C); thus these mutants were expected to be applicable for practical use. These improved mutants demonstrated reasonable migration on sodium dodecyl sulfate-polyacrylamide electrophoresis (SDS-PAGE) (Supplementary Fig. S4).

DISCUSSION

In our original CAP, stronger luminescence did not always indicate generation of higher-affinity scFv-Phs; this might be attributed to a high copy number of scFv-Phs with moderately-high affinities.¹²⁾ To distinguish these situations, we devised off-rate-dependent (ORD) selection.¹²⁾ Promising scFv-Phs, found in the initial luminescence profiling, were captured again using the immobilized antigen, and free antigen was subsequently added and incubated to remove scFv-Phs with fast off-rates from the solid phase. The scFv-Phs remaining on the solid phase were estimated to show desirably high affinity owing to the slow off-rates. Although the ORD actually improved the efficiency for isolating affinity-matured scFv-Phs, it required careful optimizations and somewhat laborious operations containing a repeated cycle of incubation with free antigen, removal of the eluted scFv-Phs, and the luminescence assay.

In the current CAP/DIR, we employed acidic and basic treatments, both with the standard stringency used in the panning for recovering desirable clones; we expected that this step would eliminate not only scFv-Phs with low- and moderate affinities but also scFv-Phs with “ordinarily-high” affinities. Next, still-bound scFv-Phs were retrieved in a dissociation-independent manner, i.e., via reductive cleavage of the SS bonds.^14,15) This CAP/DIR system enabled more straightforward and efficient discovery of satisfactory affinity-matured scFvs than our original CAP, which exceeded our expectations: 15 clones among the top 20 promising candidates (i.e., 75% hit without trial-and-error process), selected from a single survey of 4700 transformants (i.e., only 0.09% of the entire library), exhibited a K_a of 1.2–2.0 × 10¹⁰ M⁻¹, which exceeded the affinity ceiling observed in the previous study.¹³⁾ These scFvs consequently enabled ELISAs that can determine cortisol with 18–51-fold improved sensitivity (based on the comparison of midpoints) than the assay with wt-scFv. The mutant scFv#n5-325 enabled the most sensitive ELISA of cortisol, showing approx. 5–50 pg/assay as the measurable range and promising specificity for practical use. We should mention that conventional hybridoma methods have rarely generated anti-cortisol antibodies showing practical affinity and specificity so far.^19–21)

We expect that the CAP/DIR system is a highly versatile and robust strategy for generating antibody mutants with higher affinity and consequently for developing more sensitive immunoassays than those based on conventional antibodies. Automation of the procedure will enable a complete survey of the initial library, thereby discovering much more satisfactory scFv species. Universality of this system, i.e., whether the present system widely results in successful isolation of affinity-matured scFvs for various antigens, will be examined in the future.

Acknowledgments

This work was supported by JSPS KAKENHI [Grant No. JP19K07021]. We would like to thank Dr. Eskil Söderlind (Avena Partners AB, Sweden) and Professor Carl A. K. Borrebaeck (Lund University, Sweden) for permitting us to use the pEXmide 7 and 7′ vectors.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Kramer K, Hock B. Recombinant antibodies for environmental analysis. Anal. Bioanal. Chem., 377, 417–426 (2003).
• 2) Siegel RW, Baugher W, Rahn T, Drengler S, Tyner J. Affinity maturation of tacrolimus antibody for improved immunoassay performance. Clin. Chem., 54, 1008–1017 (2008).
• 3) Kobayashi N, Oyama H. Antibody engineering toward high-sensitivity high-throughput immunosensing of small molecules. Analyst, 136, 642–651 (2011).
• 4) 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).
• 5) 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).
• 6) 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).
• 7) Oyama H, Morita I, Kiguchi Y, Morishita T, Fukushima S, Nishimori Y, Niwa T, Kobayashi N. A single-step “breeding” generated a diagnostic anti-cortisol antibody fragment with over 30-fold enhanced affinity. Biol. Pharm. Bull., 40, 2191–2198 (2017).
• 8) Oyama H, Kiguchi Y, Morita I, Yamamoto C, Higashi Y, Taguchi M, Tagawa T, Enami Y, Takamine Y, Hasegawa H, Takeuchi A, Kobayashi N. Seeking high-priority mutations enabling successful antibody-breeding: systematic analysis of a mutant that gained over 100-fold enhanced affinity. Sci. Rep., 10, 4807 (2020).
• 9) Morita I, Kiguchi Y, Nakamura S, Yoshida A, Kubo H, Ishida M, Oyama H, Kobayashi N. More than 370-fold increase in antibody affinity to estradiol-17β by exploring substitutions in the V_H-CDR3. Biol. Pharm. Bull., 45, 851–855 (2022).
• 10) Smith GP, Petrenko VA. Phage display. Chem. Rev., 97, 391–410 (1997).
• 11) Introduction to Antibody Engineering. (Rüker F, Wozniac-Knopp G ed.) Springer, Switzerland (2021).
• 12) Kiguchi Y, Oyama H, Morita I, Morikawa M, Nakano A, Fujihara W, Inoue Y, Sasaki M, Saijo Y, Kanemoto Y, Murayama K, Baba Y, Takeuchi A, Kobayashi N. Clonal array profiling of scFv-displaying phages for high-throughput discovery of affinity-matured antibody mutants. Sci. Rep., 10, 14103 (2020).
• 13) Kiguchi Y, Oyama H, Morita I, Nagata Y, Umezawa N, Kobayashi N. The V_H framework region 1 as a target of efficient mutagenesis for generating a variety of affinity-matured scFv mutants. Sci. Rep., 11, 8201 (2021).
• 14) Kobayashi N, Karibe T, Goto J. Dissociation-independent selection of high-affinity anti-hapten phage antibodies using cleavable biotin-conjugated haptens. Anal. Biochem., 347, 287–296 (2005).
• 15) Kobayashi N, Oyama H, Nakano M, Kanda T, Banzono E, Kato Y, Karibe T, Nishio T, Goto J. “Cleavable” hapten-biotin conjugates: preparation and use for the generation of anti-steroid single-domain antibody fragments. Anal. Biochem., 387, 257–266 (2009).
• 16) Scatchard G. The attractions of proteins for small molecules and ions. Ann. N. Y. Acad. Sci., 51, 660–672 (1949).
• 17) 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).
• 18) Hosoda H, Kobayashi N, Ishii N, Nambara T. Bridging phanomena in steroid immunoassays. The effect of bridge length on sensitivity in enzyme immunoassay. Chem. Pharm. Bull., 34, 2105–2111 (1986).
• 19) 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).
• 20) 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).
• 21) 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).
• 22) Kiguchi Y, Oyama H, Morita I, Katayama E, Fujita M, Narasaki M, Yokoyama A, Kobayashi N. Antibodies and engineered antibody fragments against M13 filamentous phage to facilitate phage-display-based molecular breeding. Biol. Pharm. Bull., 41, 1062–1070 (2018).
• 23) Oyama H, Kiguchi Y, Morita I, Miyashita T, Ichimura A, Miyaoka H, Izumi A, Terasawa S, Osumi N, Tanaka H, Niwa T, Kobayashi N. NanoLuc luciferase as a suitable fusion partner of recombinant antibody fragments for developing sensitive luminescent immunoassays. Anal. Chim. Acta, 1161, 238180 (2021).