Comparative analysis of anti-MICA scFv affinities: Insights from three label-free biophysical methods and biological validation

TL;DR

This study compares three label-free methods to measure antibody fragment binding to MICA, finding that the wild-type variant is a better therapeutic candidate.

Contribution

The study provides a benchmark for selecting affinity assays using orthogonal biophysical methods and validates the superiority of the wild-type scFv.

Findings

SPR provided the most precise and discriminative affinity measurements among the scFv variants.

The wild-type scFv showed superior binding, stability, and functionality compared to the Beta mutant.

Combined biophysical and cellular data support the wild-type scFv as a promising therapeutic candidate.

Abstract

•Three label-free methods and ELISA were compared to assess scFv–MICA binding affinity for therapeutic candidate selection.•Surface plasmon resonance (SPR) yielded the most precise and discriminative affinity measurements of the scFv variants.•The wild-type scFv showed superior binding, stability, and functionality compared to the Beta mutant.•Combined biophysical, in silico, and cellular data support the wild-type scFv as a promising therapeutic candidate.

Three label-free methods and ELISA were compared to assess scFv–MICA binding affinity for therapeutic candidate selection.

Surface plasmon resonance (SPR) yielded the most precise and discriminative affinity measurements of the scFv variants.

The wild-type scFv showed superior binding, stability, and functionality compared to the Beta mutant.

Combined biophysical, in silico, and cellular data support the wild-type scFv as a promising therapeutic candidate.

Antibody–antigen affinity is a key factor in the development of therapeutic antibodies, influencing candidate selection and their clinical efficacy. We used a combination of three label-free biophysical techniques: SPR, ITC, and fluorescence quenching, to rank anti-MICA single-chain variable fragments (scFvs), complemented by ELISA. We evaluated a wild-type (WT) scFv and its Beta mutant across the platforms, complementing this with molecular dynamics simulations and functional assays. The affinity constants (KD) varied depending on the method used, but SPR provided the most precise discrimination. Flow cytometry confirmed stronger binding of the WT scFv to MICA-expressing gastric cell lines, which was consistent with the results obtained using SPR. Simulations indicated greater conformational stability and more favourable antigen contacts for the WT scFv. Together, these results highlight the value of orthogonal measurements for robust affinity assessment and support WT as a promising lead for further development. This benchmark provides practical guidance for selecting affinity assays in discovery pipelines.

Image, graphical abstract

Figures

C](#fig0001)). After purification and refolding, all proteins exhibited purity levels above 90%, which was confirmed by SDS-PAGE and western blotting, with molecular weights of approximately 30 kDa for WT and Beta mutant scFvs ([Fig. 1D](#fig0001)) and 27 kDa for MICA ([Fig. 1E](#fig0001)). These results validate the successful production and characterization of the proteins for downstream analyses.Fig. 1Characterization of Recombinant Proteins: MICA and anti-MICA scFvs. (A) Molecular model, shown as a ribbon representation, of the variable fragment of the anti-MICA scFvs. The framework is dis

A](#fig0003)). For the mathematical analysis, we considered the response obtained during an equilibrium period of 200 seconds. We then built a curve fitting the Langmuir isotherm model with one binding site and used Equation 2 to determine the K_D_.Fig. 3Kinetic binding analysis of scFvs to MICA by SPR. Top. Sensorgram showing the interactions between MICA (at concentrations ranging from 1 to 80 nM) and immobilized WT scFv (A) and Beta mutant scFv (B) during 600 s at a constant temperature of 37 °C. (Low. Langmuir Isotherm Model of WT (A) and Beta mutant (B) scFvs (10 µg/mL) with MICA at incre

A](#fig0004)). The formation of both complexes was exothermic. The released heat, normalized by the number of moles added in each injection, was plotted as a function of the ratio of the total scFV and MICA concentrations in the cell ([Fig. 4B](#fig0004)). Titrations of both antibodies with MICA displayed sigmoid binding isotherms, with inflection points at the ligand/receptor molar ratio ∼1, consistent with a 1:1 interaction stoichiometry. A simple site-independent binding model was fitted to the binding isotherms, revealing similar thermodynamic parameter values for both scFv-MICA complexes

A](#fig0005)).Fig. 5Fluorescence quenching titration for scFv-MICA affinity determination. (A) Illustration of the analytical technique of tryptophan fluorescence quenching used to measure the binding interactions of biomolecules. Various concentrations of anti-MICA scFvs were incubated with a fixed concentration of MICA (1 μM) in a quartz cell. Titrations were performed by adding aliquots of the titrant to the solution containing MICA, continuing until a molar ratio of [titrant]/[MICA] of ten was achieved. The fluorescence emission spectra were collected using an FS5 Spectrofluorometer with b

A](#fig0006), the RMSD profiles for both complexes stabilize after approximately 20 ns. To assess local flexibility in greater detail, the root mean square fluctuation (RMSF) profiles for the CDRs were analyzed, as shown in [Fig. 6C](#fig0006). Notably, the Beta mutant displayed substantial decreases in RMSF, around residues P43^FW^, S59^L2^, S96^L3^, G143^FW^, E238^H3^, and D243^H3^, compared to the WT scFv. These pronounced changes in flexibility suggest that the Beta mutations within the paratope, potentially compromise antigen recognition dynamics.Fig. 6Molecular dynamics simulations of th

Full text

Introduction

1

The accurate characterization of antibody-antigen affinity is crucial for comprehending molecular interactions within biological, therapeutic, and biotechnological frameworks. Affinity, measured by the dissociation constant (K_D_), represents the strength of the reversible interaction between an antibody's paratope and an antigen's epitope [1,23,32]. Affinities in the nanomolar range are particularly significant in clinical applications, as these interactions often lead to greater specificity, reduced off-target effects, and improved therapeutic efficacy in complex biological environments [52]. High-affinity antibodies display enhanced binding probabilities in challenging conditions and reduced adverse effects by minimizing non-specific interactions [50].

Affinity is fundamentally governed by a dynamic balance of molecular forces, including hydrophobic interactions, hydrogen bonds, van der Waals forces, and electrostatic interactions [50,58]. These physical interactions govern the thermodynamic determinants of Gibbs free energy — enthalpy (ΔH) and entropy (ΔS) — which, in turn, provide deeper insights into the antigen-antibody binding mechanism, offering a more comprehensive understanding of the molecular forces that define affinity [56].

Several experimental methods have been developed to measure affinity, each offering distinct advantages and limitations. Surface-based methods, such as enzyme-linked immunosorbent assays (ELISA) [3,15], radioimmunoassays [22], and optical biosensor technologies [9], including biolayer interferometry (BLI) [19] and surface plasmon resonance (SPR) [47], provide high sensitivity and real-time interaction analysis. However, the immobilization of antigens can affect their native conformation and, in some cases, introduce artifacts. Conversely, solution-based methods, such as isothermal titration calorimetry (ITC) and fluorescence quenching titration, enable the study of interactions in near-physiological environments, although they may be limited by sensitivity and the complexity of data interpretation [5,62].

In addition to experimental techniques, in silico approaches, such as molecular dynamics simulations, offer valuable insights into antigen-antibody complexes' structural and energetic properties. These simulations complement experimental findings by identifying molecular patterns that correlate with affinity. Furthermore, assays with antigen-expressing cell lines to validate in vitro biophysical measurements and in silico analyses provide a more accurate assessment of antibody functionality in its physiological environment [63].

MHC class I polypeptide-related sequence A (MICA), a stress-induced protein expressed on tumor cell surface, is a key player in immune surveillance by interacting with the cellular cytotoxicity-inducing NKG2D receptor on natural killer (NK) cells and subsets of T lymphocytes [12,21,44,53]. However, MICA overexpression on the tumor cell membrane, followed by its cleavage into a soluble form by metalloproteases, facilitates immune evasion [41,48,61], thus highlighting MICA’s potential as a promising target for therapeutic antibody development [4,14,42,54].

In this context, single-chain variable fragments (scFvs) emerge as valuable tools because of their structural simplicity, consisting of variable regions of heavy and light chains joined by a flexible linker. This format includes the antibody's complementary determining regions (CDR), which are directly involved with antigen interaction [40]. scFvs can also be produced quickly using a prokaryotic expression system (such as E. coli) [57], which facilitates the evaluation of their recognition and affinity to the antigen. In addition, its rapid production enables the assessment of affinity mutants proposed by in silico affinity maturation studies. Thus, tools that provide the screening of antibodies and the selection of the best candidates with high binding affinity are essential for biotechnological development and research.

In this study, we selected and characterized an anti-MICA scFv (WT) from a phage display library. We then engineered a mutant (Beta-ScFv) to increase its affinity for MICA and improve its functional activity in relevant cellular assays, which supports its potential as a therapeutic candidate to counter tumor immune evasion. To advance in the development of scFvs as potential biologicals, it is essential to accurately determine their binding affinity, which is critical for the subsequent steps in developability. To achieve this, we compared the affinities of the anti-MICA scFv and its mutant using three label-free methods: SPR, ITC, and fluorescence quenching titration and ELISA. Our experimental results were further supported by in silico simulations and in vitro assays to identify the most suitable method for characterizing antigen-antibody interactions and validating their biological relevance. This comprehensive approach aims to deepen our understanding of how different methodologies determine affinity measurements and guide the development of high-affinity therapeutic antibodies. We believe this work bridges the gap between molecular-level analyses and their therapeutic implications by integrating in silico simulations, in vitro assays, and label-free techniques.

Materials and methods

2

Production of recombinant MICA and scFv in E. coli BL21 (DE3)

2.1

The wild-type (WT) anti-MICA scFv and its Beta mutant variant, which contains four mutations (I32Y^L1^, S164F^H1^, P188W^H2^ and G190W^H2^), were expressed in Escherichia coli BL21(DE3) as C-terminal 6xHis-tagged fusion proteins. Untagged MICA was expressed as a truncated extracellular form containing the α1 and α2 domains (residues 1-183) of the MICA*001 variant (PDB ID: 1HYR) [6,38]. Protein production was performed following previously described protocols [26].

Inocula were prepared using 25 mL of 2x YT (BD Biosciences, USA) supplemented with 100 μg/mL ampicillin (United States Biological, USA) and 10 μL of glycerol stocks. These cultures were then grown overnight at 37 °C at 200 rpm. Subsequently, 25 mL of culture was transferred to 500 mL of fresh 2x YT medium containing ampicillin and grown until the mid-log phase. Protein expression was induced with 1 mM IPTG for 3 h for scFvs and 5 h for MICA.

After induction, bacterial cultures were harvested by centrifugation at 3200 x g for 10 min at 4 °C. Thereafter, the cultures were washed with ice-cold PBS and resuspended in lysis buffer containing 25 mM Tris (pH 8.0), 100 mM NaCl, 5 mM imidazole, 1% Triton X-100, lysozyme and a protease inhibitor cocktail (Merck KGaA, Germany). Lysis was performed by sonication on ice, with ten cycles of 20 s of sonication and 40 s of rest. Subsequently, the lysates were centrifuged at 10000 x g for 10 min at 4 °C. The soluble fractions (SF) were stored at -20 °C, while the inclusion bodies (IB) were washed with 20 mL of washing buffer (50 mM Tris, 100 mM NaCl, 1% Triton X-100, 0.1% sodium deoxycholate (DOC), pH 8.0). The IB were treated overnight at 4 °C with 20 mL of equilibrium/denaturation buffer (50 mM Tris, 500 mM NaCl, 6 M guanidine hydrochloride, 25 mM imidazole, pH 8.0) under gentle shaking.

MICA purification was carried out under denaturing conditions. Samples in denaturation buffer were centrifuged at 10000 x g for 10 min at 4 °C. Purification of scFvs was performed using affinity chromatography with a Ni-NTA matrix. To achieve this, 10 mL of denatured samples were incubated with 1.5 mL Ni-NTA-matrix at 4 °C for 2 h under gentle agitation. The matrix was then transferred to gravity columns, and the unbound proteins (UBP) were collected. The matrix was washed twice with 8 mL of equilibrium buffer (50 mM Tris, 500 mM NaCl, 25 mM imidazole, 6M guanidine hydrochloride, pH 7.4), after which bound proteins were eluted with 1.5 mL fractions of elution buffer (50 mM Tris, 500 mM NaCl, 300 mM imidazole, pH 8.0). The eluates were stored at -20 °C. The protein concentration was determined using the Bradford method, and the purity was assessed via SDS-PAGE (12% acrylamide gel). The identity of MICA and scFvs proteins was confirmed by western blot using a HRP-conjugated anti-His tag monoclonal antibody (200-303-382, Rockland, USA).

The refolding of scFvs was performed by dissolving 1 mg of protein in 100 mL of renaturation buffer (50 mM Tris, 500 mM L-arginine, 3 mM reduced glutathione (GSH), and 0.3 mM oxidized glutathione (GSSG)). A buffer composed of 50 mM Tris, 500 mM NaCl, 3 mM GSH and 0.3 mM GSSG was used to refold MICA. Eluted fractions were gradually added to the renaturation buffer in aliquots of 20 to 30 μL every 3 min with gentle agitation. After overnight incubation at 4 °C, the samples were centrifuged at 10000 x g for 10 min at 4 °C to remove insoluble aggregates. The renatured proteins were then concentrated using a centrifugal filter unit (Centricon Plus-70 10 kDa MWCO, Merck Millipore, USA) at 3000 x g for 10 min at 4 °C. Subsequently, buffer exchange was performed to replace the renaturation buffer with 10 mM HEPES buffer, pH 7.4. The protein concentrations were determined using the Bradford method.

Affinity constants determination

2.2

The affinity constants of the scFvs (WT and Beta mutant) were determined using enzyme immunoassay (ELISA) and three label-free methods: surface plasmon resonance (SPR), isothermal titration calorimetry (ITC), and fluorescence quenching titration.

Enzyme immunoassay (ELISA)

2.2.1

A 96-well plate (Maxisorp, Thermo Scientific™ Nunc™, USA) was coated with 100 µL of recombinant MICA at concentrations of 1.5, 3 and 5 µg/mL in PBS buffer. The plate was incubated overnight at 4 °C. After incubation, the content was discarded and a blocking solution containing 1% bovine serum albumin (BSA) was added, followed by incubation at 37 °C for 2 h. The plate was washed five times between each step with PBS containing 0.05% Tween-20. Next, varying concentrations of renatured scFvs were added, ranging from 1.05 to 537.63 nM, and the plate was incubated at 37 °C for 1.5 h. After washing, HRP-conjugated anti-His tag monoclonal antibody was added as a detection antibody at a dilution of 1:20000 and incubated at 37 °C for 1.5. After incubation with the anti-His tag antibody, TMB ultra solution (1-Step™, Thermo Fisher™, USA) was added and allowed to develop for 10 min at room temperature. The reaction was stopped by adding 2 M sulfuric acid. The absorbance was measured using a spectrophotometer (Biotek, Synergy, USA) at 450 nm.

The affinity constants (K_aff_) for each scFv were determined using Equation 1, described by Beatty J.D and cols [3].

$$[eqn]$$

Ag corresponds to the concentration of MICA used to build a curve with a higher optical density (OD) at 450 nm, while Ag’ refers to the concentration of MICA used to generate a curve with a lower signal than the previous one. Ab and Ab’correspond to the concentrations of scFv that reach fifty percent of the maximum O.D. in the curves with Ag and Ag’ concentrations, respectively.

Surface plasmon resonance (SPR)

2.2.2

SPR utilizes the optical properties of gold to generate surface plasmons upon the incidence of a plane-polarized light, with changes in the reflection index detected by a prism [7]. SPR was performed at 25 °C using the Reichert Dual channel SPR 7500DC model (Reichter, Inc, USA), which includes an autosampler system. To prepare the biosensing surface, the gold chip surface was modified with a self-assembled monolayer of 4-aminobenzoic acid (4ABA > 99% w/w; Merck) via drop coating outside the instrument. Specifically, 150 µL of a 1.0 mM solution in ethanol was placed in a closed petri dish for 30 min at room temperature. The gold chip was then washed with ethanol to remove the excess of 4-mercaptobenzoic acid (4MBA) and dried using a stream of nitrogen gas. The 4MBA/gold chip was placed on a drop of immersion oil (7 µL) in the SPR instrument.

The SPR setup has two channels: the working and reference channels. In both channels, the carboxyl groups were activated using N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC, ≥ 98%, 191.70 g/mol; Merck) and N-hydroxysuccinimide (NHS, 98%, 115.09 g/mol; Merck). Aqueous solutions of EDC (0.4 M) and NHS (0.1 M) were prepared and mixed in equal volumes immediately before the experiments. Next, 800 µL of the EDC/NHS mixture (final concentration of 0.2 M EDC and 0.05 M NHS) was injected twice at a 20 µL/min flow rate for 720 seconds. The working channel was immediately modified with the scFv WT solution by injecting 500 µL (10 µg/mL in 10 mM HEPES) at a flow rate of 10 µL/min for 1200 seconds while the reference channel remained closed. Finally, 750 µL of ethanolamine (≥ 98%, 61.08 g/mol; Merck) (0.1 M to pH 8.5) was injected three times at a flow rate of 10 µL/min for 720 seconds to block the free sites in both channels.

Solutions of MICA recombinant protein were injected at various concentrations for 12 min at a flow rate of 10 µL/min, using HEPES buffer as the running solution. After each interaction assay, the biosensor surface was regenerated using HCl (10 mM) for 12 min at a 10 µL/min flow rate. Surface regeneration was conducted at 37 °C. Data acquisition was performed using integrated SPR Autolink from Reichert Technologies model (Reichter, Inc, USA), and all data were processed using TraceDrawer 1.6.1 and OriginPro 8.0 software (OriginLab, USA).

The composition of each solution, the flow rates used, and the concentration range evaluated were selected to ensure reliable measurement of the interaction between MICA and scFvs, avoid artifacts associated with mass transport, and prevent precipitation at high concentrations.

The association curves were corrected using a double reference approach, which consists of subtracting the response obtained from the reference channel and that of an injected buffer solution. The average signal of refractive index unit (sRIU) in the equilibrium stage (Req), at different MICA concentrations, was graphed and analyzed using the Langmuir isotherm model with a single binding site, obtaining the corresponding K_D_ and the maximum response (Rmax) (Equation 2), while Ka corresponds to the inverse of K_D_.

$$[eqn]$$

Where A corresponds to the MICA concentrations and K_D_ is constant.

Isothermal titration calorimetry (ITC)

2.2.3

ITC measurements were carried out using a MicroCal™ iTC200 instrument (GE Healthcare, USA). The titration schedule consisted of 12 to 14 consecutive injections (1.2 µL each) of the ligand (MICA) into a low volume of each scFv (300 µL in the sample cell), with a 5-min interval between injections. Titrations were conducted in a 10 mM HEPES buffer at pH 7.4 and 37 °C. The solution was stirred at 350 rpm., and the calorimeter measured the resulting heat of the reaction. All samples were degassed for 10 min before the experiment. The ligand's dilution heat was obtained by adding MICA to the buffer solution (10 mM HEPES, pH 7.4) under identical conditions, using the same injection schedule as with the protein sample. The binding reaction was monitored by recording the heat release upon small additions of the MICA solution to each scFv solution. The variability observed in ITC measurements was addressed by performing multiple replicates and applying statistical corrections to ensure reliable thermodynamic parameters. Binding parameters were determined by using an identical and independent binding site model [24] ((3), (4)):

$$[eqn]$$

where Q is the normalized heat evolved per mol of ligand at the end of the i^th^ injection, ΔH_a_ is the enthalpy change, n is the stoichiometry, V_o_ is the working volume of the cell, and L_t_ and M_t_ are the total ligand and macromolecule concentrations, respectively. The heat released in the i^th^ injection is:

$$[eqn]$$

where v_i_ is the aliquot volume added at injection i and q_dil_ is a fitting term introduced to account for experimentally uncorrected dilution heat effects. All non-linear regressions were carried out using the MicroCal Origin v7. (Origin Lab, USA)

Fluorescence quenching titration

2.2.4

Stock solutions of MICA, WT, and Beta scFvs were prepared in 10 mM HEPES buffer at pH 7.4 to measure K_aff_ using fluorescence quenching titration. A 1 μM solution of MICA was prepared by dilution and placed in a quartz cell. Titrant solutions of anti-MICA scFvs were ready at 35 μM, containing 1 μM of MICA to avoid dilution during titration. Titrations were performed by adding aliquots of the titrant to the solution containing MICA, continuing the titration until a molar ratio [titrant]/[MICA] of ten was achieved. Additionally, control titrations were conducted by adding aliquots of the same titrant solution (35 μM, without MICA) to the HEPES buffer to establish a reference.

The fluorescence emission spectra were collected using an FS5 Spectrofluorometer (Edinburgh Instruments, Scotland) with 5 nm bandwidths for excitation and emission. Samples were excited at 295 nm. The fluorescence signal from the buffer was subtracted from all analyses. Since both the antigen and antibody contain a tryptophan residue that exhibits fluorescence, the fluorescence signal from the titrant was subtracted from the MICA titration fluorescence data before data analysis. Non-linear fit of the fluorescence data of MICA with scFvs, plotting the area of fluorescence as a function of molarity. The association constant (K) was calculated from the binding curves using OriginPro 8.0 software (OriginLab, USA).

In silico simulations

2.3

Molecular dynamics simulations were performed to explore the structural and energetic interactions between the scFvs and MICA. The heavy chain and light chain sequences of the scFv were modeled using RosettaAntibody [60], producing 4,000 distinct models. The model with the most favorable Rosetta total score was then selected. Additionally, the MICA structure was obtained from its crystal structure (PDB: 1HYR) [38].

To develop a model of the complex, both structures were subjected to global and local docking. Initially, ClusPro2.0 [34] was employed for global protein-protein docking in antibody mode, and the largest resulting cluster was subjected to local docking using SnugDock [49].The system was constructed by inserting the resulting MICA-scFv complex in a TIP3P water box, using the CHARMM-GUI server [31]. A 0.15 M NaCl concentration was added, and the titratable side chains were kept at their dominant protonation state at pH 7.0. The ff19SB force field [51] was used for all molecular dynamics (MD) simulations. All MD simulations were conducted using AMBER20 [8]. The protocol consisted of an initial minimization step consisting of up to 5,000 conjugate gradient steps and 2,500 steps of steepest descent. The system was then heated from 0 K to 310 K under NPT conditions for 100 ps, employing a harmonic restraint of 10.0 kcal·mol^−1^·Å^−2^ on all protein atoms. Subsequently, a 1 ns equilibration was performed without restraints to allow the system to reach equilibrium. A 300 ns production run followed, during which binding energies were computed using MMPBSA.py [39] using the Generalized Born implicit-solvent model (igb=8) and mbondi3 atomic radii. The MMGBSA calculations were performed using 1500 frames selected from the production trajectory. The trajectory analysis was conducted using the cpptraj program. The backbone RMSD was calculated for the Cα atoms of each scFv relative to the initial structure. Additionally, RMSF profiles were generated to assess local flexibility. To further elucidate the contribution of each CDR residue in the MICA-Fv complex, energy decomposition analysis was performed with MMPBSA.py [39] on these residues.

Analysis of scFvs binding to MICA expressed on gastric cells by flow cytometry

2.4

Human gastric cell lines GES-1 and MKN-45 were cultured in RPMI-1640 medium (ThermoFisher scientific, USA) supplemented with 2 mM L-glutamine, 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (GIBCO, USA). To assess cell viability, cells were incubated with Zombie NIR dye (Biolegend, USA) for 15 min at room temperature. After staining, cells were washed with FACS buffer and centrifuged at 400 x g for 5 min at 4 °C. The resulting cell pellets were incubated with 5 μg of WT scFv or its Beta mutant variant for 45 min at 4 °C, followed by incubation with an anti-6xHis tag secondary antibody conjugated to APC fluorochrome (Clone: 1:500 dilution) to detect scFv binding. After additional washes, the cell pellets were resuspended in FACS buffer and analyzed using a BD LSR Fortessa X-20 cytometer (BD Biosciences, USA). Data analysis was performed using FlowJo v10 software (FlowJo, LLC).

Results

3

Characterization of recombinant proteins

3.1

In this study, we generated three recombinant proteins: two scFvs antibodies (WT and its Beta mutant) and their immuno-target, the MICA protein. Figs. 1A and 1B show the vector design and the mutation sites introduced into the anti-MICA WT scFv to generate the Beta mutant. The Beta mutant contains four specific mutations: I32Y^L1^, S164F^H1^, P188W^H2^ and G190W^H2^. All proteins were produced in E. coli BL21(DE3) as inclusion bodies (Fig. 1C). After purification and refolding, all proteins exhibited purity levels above 90%, which was confirmed by SDS-PAGE and western blotting, with molecular weights of approximately 30 kDa for WT and Beta mutant scFvs (Fig. 1D) and 27 kDa for MICA (Fig. 1E). These results validate the successful production and characterization of the proteins for downstream analyses.Fig. 1Characterization of Recombinant Proteins: MICA and anti-MICA scFvs. (A) Molecular model, shown as a ribbon representation, of the variable fragment of the anti-MICA scFvs. The framework is displayed in white, the light chain CDRs are shown in cyan, and the heavy chain CDRs are shown in yellow. The residues with mutations are shown as magenta spheres [residues 32 (CDR L1), 164 (CDR H1), and 188/190 (CDR H2)]. (B) Schematic diagram of the scFv gene. The modified pET-15b vector was used for the expression of the WT and Beta mutant scFvs, each carrying four mutations: I32Y in CDR1 of the VL, and S164F, P188W, and G190W in CDR1, CDR2, and CDR2 of the VH, respectively. Recombinant proteins were expressed in E. coli BL21(DE3). (C) SDS-PAGE analysis showing the purity of recombinant proteins: WT scFv, Beta mutant scFv, and MICA. Proteins were resolved on a 12% acrylamide gel under reducing conditions. SDS-PAGE results show the soluble fraction (SF), unbound protein (UBP), elution of purified scFv (E), renatured proteins (R) and inclusion bodies (IB). MW, molecular weight. (D-E) Western blot analysis confirming the identity of scFvs and MICA using an anti-HisTag antibody. For the identification of the WT and Beta mutant scFvs, Anti-6xHis Epitope Tag mouse monoclonal antibody conjugated with peroxidase (200-303-382) was used at a dilution of 1:1000. For the identification of MICA, a biotinylated Anti-MICA antibody (BAMO3 (BAFI300, BamOmaB)) and Streptavidin were used at a dilution of 1:2000. A total of 2 μg of purified protein was loaded. The negative control (Ctrl -) for MICA detection was WT scFv and MICA protein was used for scFv detection. Original gel is presented in Fig. S1, Supplementary information.Fig 1 dummy alt text

Determination of affinity constants of anti-MICA scFvs by ELISA

3.2.1

To determine the affinity constants of anti-MICA scFvs, we used the non-competitive ELISA described by Beatty et al. [3]. Three titration curves were generated, and the scFvs concentration corresponding to 50% of the maximum optical density (OD) was determined (Fig. 2). Using this value in Equation 1, the calculated K_D_ for WT scFv was 68.1 ± 12.4 nM, while the Beta mutant scFv exhibited a K_D_ of 21.3 ± 4.0 nM. The data provided by this method indicate a significantly higher apparent affinity for the Beta mutant compared to the WT scFv (p = 0.0034).Fig. 2Affinity determination of anti-MICA scFvs by ELISA. Representative titration curves for wild-type (WT) (A) and (B) Beta mutant single-chain variable fragments (scFvs) binding to MICA. A 96-well plate was coated with varying concentrations of MICA (1.5, 3, and 6 µg/mL). The plate was then incubated with ten different dilutions of each scFv, ranging from 1.05 to 537.63 nM. The OD_450_ nm readings were plotted against scFv concentration, and the K_D_ was calculated using the concentration of scFvs that reached 50% of the maximum OD. The results demonstrate significant differences in binding affinity between WT and Beta mutant scFvs, with Beta mutant exhibiting higher affinity (K_D_ = 21.3 ± 4.0 nM) compared to WT (K_D_ = 68.1 ± 12.4 nM). Data are presented as mean ± SD from three independent experiments (p = 0.0034).Fig 2 dummy alt text

Determination of affinity constants of anti-MICA scFvs by SPR

3.2.2

SPR is one of the most widely used techniques to determine affinity constants and characterize the interaction of an antibody with its target in the biopharmaceutical industry. International regulatory agencies currently accept it for requesting approval for novel or biosimilar biological molecules for human application. In this context, we used this method to asses antibody-antigen affinity constants and compare the values with those obtained from ELISA. To immobilize scFv antibodies through their NH_2_ sites, we employed a chemical technique that involves activation of a 4MBA gold chip with EDC-NHS, resulting in the formation of amide bonds, which allows for a randomized immobilization of the proteins. After standardization of the method, we determined the binding response, at different time points, upon injection of MICA at different concentrations (1 to 80 nM). Kinetic binding analysis showed that the WT scFv and MICA interaction equilibrium was reached within 120 to 180 seconds irrespective of MICA concentration (Fig. 3A). For the mathematical analysis, we considered the response obtained during an equilibrium period of 200 seconds. We then built a curve fitting the Langmuir isotherm model with one binding site and used Equation 2 to determine the K_D_.Fig. 3Kinetic binding analysis of scFvs to MICA by SPR. Top. Sensorgram showing the interactions between MICA (at concentrations ranging from 1 to 80 nM) and immobilized WT scFv (A) and Beta mutant scFv (B) during 600 s at a constant temperature of 37 °C. (Low. Langmuir Isotherm Model of WT (A) and Beta mutant (B) scFvs (10 µg/mL) with MICA at increasing concentrations (1 to 80 nM) built with an equilibrium range of 200 s of response. K_D_ values obtained were 3.57 ± 0.01 nM for WT scFv and 128 ± 6.10 nM for Beta mutant scFv, indicating stronger binding affinity for WT. Error bars represent mean ± SD from three replicates.Fig 3 dummy alt text

The affinity constants of scFvs binding to MICA, expressed as K_D_, were 3.57 ± 0.01 nM for WT scFv and 128 ± 6.10 nM for the Beta mutant scFv (Fig. 3, Fig. 3). Contrary to data obtained by ELISA, the WT scFv displayed higher affinity to MICA compared to Beta scFv, as assessed by SPR.

Determination of affinity constants of anti-MICA scFvs by ITC

3.2.3

ITC was employed to determine the thermodynamic binding parameters of the scFvs, providing a deeper insight into the forces governing the antigen-antibody interactions. The calorimeter’s sample cell was loaded with a solution of scFv (∼5 μM) and progressively saturated through stepwise additions of MICA (∼ 50 μM) (Fig. 4A). The formation of both complexes was exothermic. The released heat, normalized by the number of moles added in each injection, was plotted as a function of the ratio of the total scFV and MICA concentrations in the cell (Fig. 4B). Titrations of both antibodies with MICA displayed sigmoid binding isotherms, with inflection points at the ligand/receptor molar ratio ∼1, consistent with a 1:1 interaction stoichiometry. A simple site-independent binding model was fitted to the binding isotherms, revealing similar thermodynamic parameter values for both scFv-MICA complexes (Fig. 4C). The affinity constants, expressed as K_D_, were comparable for both anti-MICA scFvs, with K_D_ values of 0.247 ± 0.048 µM for WT scFv and 0.274 ± 0.086 μM for the Beta mutant. These data did not show a statistically significant difference (p=0.6597). K_D_ values, in the high nanomolar range, were driven by favorable enthalpic and entropic contributions to ΔG_b_. At the molecular level, these thermodynamic signatures indicate that the favorable energy from the direct MICA-scFv interactions outweighs the energy cost associated with the desolvation of the interacting surfaces. Additionally, the entropic gain resulting from the release of water molecules upon binding was larger than the loss of degrees of freedom caused by the freezing of rotatable bonds and the reduced number of particles in the system [18].Fig. 4Thermodynamic analysis of scFv-MICA recognition by isothermal titration calorimetry (ITC). The Figure illustrates the ITC analytical technique used to measure the binding interactions of biomolecules by quantifying heat changes during the binding of scFv proteins to MICA. (A) A solution of 50 µM of MICA was injected into the reaction cell containing a solution of 5 µM of each scFv at a constant temperature of 37 °C. Heat was recorded and normalized by the number of MICA moles added in each injection. (B) Binding isotherms fitted to an identical and independent binding site model (red line). (C) Thermodynamic parameters of MICA-scFv recognition. Both scFvs exhibited similar thermodynamic profiles, with binding driven by both enthalpic and entropic contributions. Error bars represent variability across three replicates*.*Fig 4 dummy alt text

Determination of affinity constants of anti-MICA scFvs by fluorescence quenching titration

3.2.4

Here, we also employed fluorescence quenching titration to determine scFvs affinity to MICA. The critical requirement for this technique is the presence of an amino acid capable of emitting fluorescence when excited at a specific wavelength, such as tryptophan or tyrosine, in at least one of the molecules at their interaction sites. This approach allowed us to track fluorescence spectra changes or quenching upon adding the titrant molecule (anti-MICA scFv). In this case, we initially added a solution of MICA to the quartz cell and measured the basal fluorescence. We then added a titrant solution of each scFv until the molar ratio [scFv]/[MICA] reached ten (Fig. 5A).Fig. 5Fluorescence quenching titration for scFv-MICA affinity determination. (A) Illustration of the analytical technique of tryptophan fluorescence quenching used to measure the binding interactions of biomolecules. Various concentrations of anti-MICA scFvs were incubated with a fixed concentration of MICA (1 μM) in a quartz cell. Titrations were performed by adding aliquots of the titrant to the solution containing MICA, continuing until a molar ratio of [titrant]/[MICA] of ten was achieved. The fluorescence emission spectra were collected using an FS5 Spectrofluorometer with bandwidths of 5.0 nm for both, excitation and emission. Samples were excited at 295 nm. The fluorescence signal from the buffer was subtracted from all analyses. (B) Representative quenching curves resulting from the interaction of scFvs with MICA. The curves represent the reduction in tryptophan fluorescence intensity observed upon binding of WT or Beta mutant scFvs at specific antibody concentrations. Non-linear fit of the fluorescence data of scFvs binding to MICA. The area of fluorescence was plotted as a function of molarity. The affinity constant (K_D_) was calculated based on the binding curves, which provided information on scFvs affinity to MICA. Data are presented as mean ± SD from three independent experiments.Fig 5 dummy alt text

The affinity constant of the scFvs for MICA, expressed as K_D_, was 429 ± 114 nM for WT and 176 ± 29.2 nM for the Beta mutant, indicating a stronger interaction between Beta scFv and MICA, compared to WT anti-MICA scFv (Fig. 5B). These findings are consistent with our ELISA results but in contrast to SPR and ITC data.

Molecular modeling of anti-MICA scFvs

3.3

Molecular dynamics simulations of the complexes between MICA and the WT anti-MICA scFv and its Beta mutant revealed distinct structural and binding energy profiles. As shown in Fig. 6A, the RMSD profiles for both complexes stabilize after approximately 20 ns. To assess local flexibility in greater detail, the root mean square fluctuation (RMSF) profiles for the CDRs were analyzed, as shown in Fig. 6C. Notably, the Beta mutant displayed substantial decreases in RMSF, around residues P43^FW^, S59^L2^, S96^L3^, G143^FW^, E238^H3^, and D243^H3^, compared to the WT scFv. These pronounced changes in flexibility suggest that the Beta mutations within the paratope, potentially compromise antigen recognition dynamics.Fig. 6. Molecular dynamics simulations of the structural and energetic interactions between the scFvs and MICA. (A) RMSD comparison of WT and Beta mutant scFvs. RMSD profiles over the simulation time for the WT scFv (red) and the Beta mutant scFv (blue). The RMSD was calculated relative to the initial structure over all Cα atoms. (B) RMSF profiles for both scFvs. (C) Calculated binding energies for both scFvs complexes. (D-E) Per-residue free energy decomposition of WT and Beta mutant scFvs. Representative models of the molecular dynamics trajectory of WT (D) and Beta mutant (E) scFvs are depicted, with the framework in white cartoon, common hotspots residues are green, WT and Beta-exclusive hotspots are yellow. Individual residues that provide favorable energy contributions (kcal/mol) are shown in spheres and listed on the tables on the right.Fig 6 dummy alt text

The per-residue energy decomposition further supported these observations. In the WT scFv, twelve residues contributed substantially to the binding free energy of this complex (-62.3 ± 8.4 kcal/mol) (Fig. 6C-D). In contrast, the Beta mutant showed a different set of hotspots, including S161^FW^, W190^H2^, and V235^H3^ to mediate binding interactions. While these residues contributed favorably to the interaction with MICA (Fig. 6E), the binding free energy for the Beta mutant was markedly higher (-48.4 ± 10.8 kcal/mol) than that of the WT scFv (Fig. 6C), indicating a distinct binding energy profile between the two constructs.

Collectively, these findings demonstrate that the Beta substitutions disrupt the CDR loop dynamics and the interaction network within the complex, leading to a reduction in binding affinity. The integration of structural and energetic analyses delineates the specific residues and conformational features underlying the functional disparities between the WT and Beta scFvs, a finding consistent with the higher K_D_ observed experimentally via SPR.

Binding capacity of scFvs by flow cytometry

3.4

Flow cytometry (FC) is a powerful tool for studying antibody interactions with proteins in their native conformation on the cell surface. In this study, we evaluated scFv binding to MICA expressed on two human gastric cell lines, GES-1 and MKN-45 [27]. MICA expression level had been previously validated using a murine anti-human MICA monoclonal antibody (Fig.S2, Supplementary information). The WT and Beta mutant scFvs binding capacity to MICA was assessed at equal concentrations and experimental conditions (pH, buffers, temperature, and time). We observed that 5 μg of WT scFv bound to 89.8% of GES-1 cells, whereas the same amount of the Beta mutant scFv bound to only 34.8%. Similarly, WT scFv bound to 66% of MKN-45 cells, a markedly higher union rate compared to the 7.5% observed for the Beta mutant scFv (Fig. 7). These results revealed a significantly higher binding capacity for WT scFv compared to the Beta mutant scFv in both cell lines, which is in line with our previous SPR data, thus confirming the superior binding capacity of WT scFv in a native biological context.Fig. 7Binding capacity of WT and Beta mutant scFvs to MICA-expressing human gastric cell lines evaluated by flow cytometry. Representative histograms showing the binding of WT and Beta mutant scFvs to GES-1 and MKN-45 cells. WT scFv demonstrated a significantly higher binding capacity to both cell lines compared to the Beta mutant scFv. Specifically, WT scFv bound to 89.8% of GES-1 cells and 66% of MKN-45 cells, while the Beta mutant scFv bound to 34.8% and 7.5% of cells, respectively. The same amount of each antibody (5 μg) was used to assess MICA binding. The negative control corresponded to cells incubated with a PE-conjugated anti-6xHis tag secondary antibody alone. MFI: Mean fluorescence intensity.Fig 7 dummy alt text

Comparison of scFv affinity constants by different label-free methods

3.5

Table 1 summarizes the scFv affinity constants expressed as K_D_ obtained using different label-free methods employed in this work. While the methods are not directly comparable due to their inherent characteristics, each enables differentiation of the binding affinity between the WT and Beta mutant scFv antibodies. Both ELISA and fluorescence quenching titration indicated that the Beta mutant scFv displays higher affinity than the WT scFv. However, SPR data showed stronger binding for the WT scFv, which is consistent with in silico simulations. ITC revealed no differences between the two molecules. These differences underscore the importance of using orthogonal methods for comprehensive antibody-antigen affinity characterization.Table 1. Summary of the affinity constants (K_D_) of scFvs determined by three label-free methods and ELISA, and binding free energies (ΔG) estimated from molecular dynamics simulations.Table 1: dummy alt textAnti-MICA scFvELISA (nM)SPR (nM)ITC (μM)Fluorescence (nM)ΔG_MMGBSA_(kcal/mol)WT68.10 ± 12.43.57 ± 0.010.247 ± 0.048429 ± 114-62.3 ± 8.4Beta21.30 ± 4.0128 ± 6.100.274 ± 0.086176 ± 29.2-48.4 ± 10.8MMGBSA: Molecular Mechanics / Generalized Born Surface Area

Discussion

4

The results obtained in this study highlight the importance of using complementary techniques to comprehensively assess antibody-antigen interactions, especially regarding therapeutic antibody development. By integrating three label-free methods —SPR, ITC, and fluorescence quenching — with ELISA, in silico simulations, and flow cytometry assays, we robustly characterized the binding properties of a WT and its Beta mutant anti-MICA scFvs. SPR yielded the most precise and discriminative affinity measurements of the scFv variants. The WT scFv demonstrated a significantly higher affinity, more stable interactions, and enhanced binding to gastric cells expressing MICA in its native conformation. Our findings validate the WT anti-MICA scFv as a promising candidate for targeted immunotherapies and emphasize the value of combining multiple methodologies to optimize antibody development.

SPR is recognized as a pivotal technique in drug discovery and therapeutic antibody development because of its capacity for label-free and real-time analysis of biomolecular interactions with exceptional sensitivity and specificity [1]. In this study, the K_D_ between scFvs and MICA was accurately measured using SPR. The K_D_ values obtained, at nanomolar concentrations, for WT scFv revealed its superior binding affinity compared to the Beta variant. These results align with both, in silico predictions and flow cytometry data, further validating the potential of WT anti-MICA scFv for therapeutic development. Recent advancements in SPR platforms, such as enhanced sensitivity and high-throughput screening capabilities, have expanded their applications in drug discovery and therapeutic antibody development, enabling robust characterization of molecular interactions [1]. However, as with any surface-based technique, SPR can present challenges in ultra-high-affinity interactions due to potential artifacts, which include prolonged saturation times [45]. While not directly addressed in this study, complementary methods, such as KinExA, which excels in the femtomolar affinity range, may offer an alternative for overcoming these limitations [13]. Nevertheless, our findings collectively highlight SPR’s status as the gold standard for evaluating binding kinetics and affinities in near-physiological conditions, making it indispensable for the characterization of therapeutic antibody candidates.

Although ELISA provides high-throughput assessment, it can introduce artifacts due to antigen immobilization, which may distort apparent affinities by altering the antigen’s native conformation [15,36]^.^ In contrast, oriented antibody immobilization, as demonstrated by Tsekenis et al. [55], significantly enhances antigen-binding efficiency, which proposes opportunities to optimize ELISA methodologies. In this study, the Beta mutant scFv showed higher affinity values by ELISA, compared to the WT antibody, possibly as a result of multivalent effects caused by antigen immobilization. However, data from SPR and flow cytometry showed contrasting results with ELISA, indicating a higher affinity for the WT scFv under physiologically relevant conditions. These discrepancies suggest the limitations of ELISA in assessing interactions in the antigen’s native conformation and underscore the importance of integrating complementary approaches, such as SPR and flow cytometry, which validated the WT scFv’s superior binding affinity to MICA in its native state. We propose that ELISA is useful as an initial screening tool, but emphasize the need for orthogonal methods to achieve a more robust and accurate antigen-antibody binding characterization.

ITC provided crucial thermodynamic insights into the binding mechanisms of both scFvs, revealing entropy-driven interactions suggestive of conformational rearrangements during antigen-antibody binding. Although the K_D_ values obtained for both scFvs were similar, they differed from those measured by SPR. This discrepancy highlights the inherent differences in sensitivity and detection ranges between these techniques. Differences in affinity values obtained by ITC and SPR are well documented [10]. ITC measures binding in free solution through heat release, whereas SPR quantifies kinetics on a surface where one partner is immobilized. Subtle processes not easily detected by either technique, including ancillary interactions or conformational rearrangements coupled to or independent of complex formation, can differentially influence the thermodynamic and kinetic readouts. ITC may also yield higher K_D_ values when the enthalpy change is small or when experimental conditions fall outside the optimal Wiseman c range ( $[eqn]$ , optimal values fall within 10 < c < 100). In our experiments, the conditions yielded $[eqn]$ , ensuring a well-defined isotherm. Overall, the lower affinity measured by ITC is consistent with known methodological and mechanistic factors intrinsic to each technique [10]. While ITC provides invaluable thermodynamic insights into the binding mechanisms, SPR offers detailed kinetic data, allowing for a more comprehensive understanding of antigen-antibody interactions [7,35]. The agreement between these methods suggests that the WT scFv forms a more stable and energetically favorable complex with MICA, reinforcing its potential as a promising candidate for therapeutic development. As emphasized by Frostell et al. [16], integrating multiple biophysical techniques enhances the reliability of affinity measurements and provides a holistic view of molecular interactions.

Although fluorescence quenching is a straightforward and rapid method for evaluating protein-ligand interactions, the K_D_ values of scFv derived from this method were inconsistent with those obtained using SPR and flow cytometry. These discrepancies are likely attributable to solvent effects on tryptophan fluorescence, as described by Yammine et al [62], including the inner filter effect and the environmental sensitivity of tryptophan residues. Changes in local polarity or conformational shifts near the interaction site may also contribute to quenching artifacts, making data interpretation more complex.

Despite these challenges, fluorescence quenching provided complementary insights by emphasizing solvent-dependent quenching effects and allowing for a relative comparison of binding strengths. Its rapidity, simplicity, and suitability for solution-based measurements highlight its potential as a preliminary screening tool, especially in resource-limited settings.

Flow cytometry, in contrast, offered valuable confirmation of binding under biologically relevant conditions, demonstrating significantly higher binding of WT scFv to MICA-expressing cells compared to the Beta variant antibody. These results aligned closely with SPR and in silico findings, further substantiating the superior binding characteristics of the WT scFv. The integration of flow cytometry with other orthogonal techniques, such as SPR, enhances the reliability of affinity measurements by providing complementary perspectives on antigen-antibody interactions. This comprehensive approach ensures a more accurate characterization of binding dynamics and highlights the importance of validating preliminary findings with physiologically relevant methods.

The in silico analysis performed in this study complements the experimental observations by providing structural and energetic insights that help elucidate the molecular mechanisms behind the scFv-MICA interaction. Computation simulations revealed greater stability in the WT scFv-MICA complex compared to the Beta variant, consistent with experimental affinity data obtained from SPR and flow cytometry. These findings align with previous studies emphasizing the ability of computational tools to predict affinities and guide targeted mutations to enhance antigen-antibody binding [30,43]. Additionally, the results confirm that the WT scFv exhibits structural features positioning it as a promising candidate for therapeutic applications, consistent with recent research underscoring the importance of conformational stability and specific molecular interactions in designing targeted immunotherapies [4,59]. This integration of in silico and experimental data reinforces the value of combined approaches for the rational development of therapeutic antibodies.

The data obtained herein position the WT scFv as a promising candidate for therapeutic applications targeting MICA, a key protein in tumor immune evasion. SPR and molecular simulations confirmed its superior stability and affinity compared to the Beta variant. Additionally, flow cytometry assays demonstrated significantly higher binding capacity of WT scFv to cells expressing MICA in its native conformation. These findings suggest that the WT scFv could restore immune surveillance by facilitating the activation of NK cells and other cytotoxic immune cells in the tumor microenvironment. Flow cytometry validated the physiological relevance of the WT scFv’s superior binding characteristics, demonstrating high levels of interaction with MICA-expressing gastric cell lines. The strong correlation between flow cytometry and SPR data reinforces the robustness of WT scFv’s binding profile. These results highlight the translational potential of WT scFv for therapeutic applications, as high binding levels correlate strongly with therapeutic efficacy [2,28].

Advanced techniques such as mass cytometry (CyTOF) could further enhance analyses of molecular interaction by enabling multiplexed evaluations of surface markers and intracellular signaling pathways, as demonstrated by Gadalla et al. [17].

The clinical application of the fully human WT scFv holds significant promise, particularly when considering potential strategies to enhance its efficacy. Various optimization approaches, such as conjugation with cytokines, antitumor drugs, or chimeric antigen receptor (CAR) T cells, could further expand its therapeutic potential [20,25]. In this context, studies have shown that antibodies targeting MICA can prevent proteolytic shedding of its α3 domain, a key mechanism of immune evasion, and enhance NK cell-mediated cytotoxicity [11,14,33]. Exploring this approach in combination with the WT scFv could significantly boost its therapeutic potential by binding to the α1 domain of MICA in both its soluble and membrane-bound forms. Depending on the tumor characteristics, this interaction may further promote the binding of the NKG2D receptor to other ligands, such as ULBPs and MICB, which also trigger target-cell death [29]. These proposals are particularly relevant in immunotherapy, where restoring immune surveillance in tumor microenvironments remains a critical challenge [4].

Future preclinical studies in animal models will allow us to evaluate the biodistribution and efficacy of the WT scFv. Emerging technologies such as mass photometry and single-cell analysis could complement current characterizations by enabling immobilization-free measurements and multiplexed evaluations of protein interactions. Advances in artificial intelligence (AI) and machine learning (ML) also present promising opportunities for optimizing antibody development. Recent breakthroughs, such as DG-Affinity, allow for efficient estimation of antigen-antibody affinity directly from sequences, bypassing the need for structural data [64]. Additionally, ML approaches have facilitated the creation of highly diverse scFv libraries with subnanomolar affinities, significantly accelerating the discovery process [37]. These innovations complement traditional methods by reducing experimental timelines and improving developability predictions [46,65].

Finally, our findings about the relative performance and discriminative power of the different affinity determination methods are derived from a detailed case study of one anti-MICA scFv system. While the methodological insights are likely to be informative for other antibody discovery campaigns, the exact quantitative relationships between methods may differ for other scFvs, antigens, or experimental conditions. Future studies involving larger panels of antibody fragments and antigens will be essential to validate, extend, and refine these observations, and to establish more general guidelines for selecting and integrating affinity determination methods in biotherapeutic development.

Conclusions

5

In conclusion, this work highlights the importance of integrating multiple analytical approaches to characterize antigen-antibody interactions, which are crucial in the early stages of development. The findings obtained for the WT scFv validate its therapeutic potential and provide a guide for the rational design of antibodies with high affinity and functionality in biological contexts. These advances could significantly contribute to developing innovative immunotherapeutic strategies against cancer.

Funding

This work was supported by 10.13039/501100005304National Agency of Research and Development ANID/FONDECYT GRANTS 1221031 (MCM), 3230454 (KT-S), 11221289 (RS) and 3240175 (FG); ANID Scholarship 21221729 (IC) and 21231772 (ST); FONDEF IDEA I+D ID23i10018 (MCM), ANID-FONDEQUIP EQM GRANT 140112 (SB) and Anillo Regular de Ciencia y/o Tecnología 2021 ACT210068 (CA, MV, GZ, MCM); Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT) of the Universidad Nacional Autónoma de México UNAM, IV201220 and IN210822 (NAV-C) and Consejo Nacional de Humanidades, Ciencias y Tecnologías CF-2023-I-1549 (NAV-C). The funders had no role in data collection and analysis, publication decisions, or manuscript preparation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work the authors used ChatGPT-5. in order to review and edit the content as required, taking full responsibility for the published article.

CRediT authorship contribution statement

Karen Toledo-Stuardo: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Nicolás Fehring: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Homero Gómez-Velasco: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Rodrigo Sierpe: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Daniel Guerra: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Douglas J. Matthies: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Yuneisy Guerra: Investigation, Writing – review & editing. Fabiola González-Herrera: Conceptualization, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Mauricio González: Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Ivo Campos: Investigation, Methodology, Visualization, Writing – review & editing. Samantha Tello: Investigation, Methodology, Visualization, Writing – review & editing. María José Garrido: Investigation, Methodology, Visualization, Writing – review & editing. Gonzalo Vásquez: Investigation, Methodology, Visualization, Writing – review & editing. Jose Rodríguez-Siza: Investigation, Methodology, Visualization, Writing – review & editing. Mauricio Vergara: Investigation, Writing – review & editing. Carolina H. Ribeiro: Conceptualization, Investigation, Visualization, Writing – original draft, Writing – review & editing. Gerald Zapata-Torres: Conceptualization, Formal analysis, Resources, Software, Visualization, Writing – original draft, Writing – review & editing. Soledad Bollo: Funding acquisition, Investigation, Methodology, Visualization, Writing – review & editing. Enrique García-Hernández: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. Denis Fuentealba: Conceptualization, Data curation, Investigation, Methodology, Resources, Visualization, Writing – review & editing. Norma A. Valdez-Cruz: Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Claudia Altamirano: Funding acquisition, Investigation, Visualization, Writing – original draft, Writing – review & editing. María Carmen Molina: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Maria Carmen Molina reports financial support was provided by National Agency for Research and Development. Claudia Altamirano reports financial support was provided by National Agency for Research and Development. Norma A. Valdez-Cruz reports financial support was provided by Consejo Nacional de Humanidades, Ciencias y Tecnologías (CONAHCYT). Soledad Bollo reports financial support was provided by National Agency for Research and Development. Maria Carmen Molina has patent #WO2019126895A1 pending to EP. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.