The genes for V_H and V_L of the anti-CGRP antibody were designed based on previously reported sequences (Benschop et al., 2014), and synthesis was carried out by Shanghai Sangon Biotechnology (Shanghai, China). Plasmid extraction reagents, including the DiaSpin column-based plasmid DNA miniprep kit, DipSpin column-based DNA gel recovery kit, and EZ-10 column-based polymerase chain reaction (PCR) product purification kit, were purchased from Shanghai Sangon Biotechnology (Shanghai, China). Escherichia coli DH5α was used for gene cloning and plasmid amplification, since it lacks immunity to exogenous DNA and therefore does not cleave exogenous DNA, allowing it to stably maintain exogenous DNA in its cells. E. coli TG-1, obtained from GE Healthcare (Tokyo, Japan), was used for phage display, which can suppress amber mutations, thus serving as a host for plasmids and bacteriophages carrying amber codons. E. coli BL21 (DE3) was used for protein expression, since it lacks the genes for the proteases lon and ompT, which facilitates the purification of exogenous proteins. The DH5α and BL21 (DE3) were purchased from Kangti Life Technology (Guangzhou, China). Restriction enzymes were purchased from New England BioLabs (NEB, Beijing, China). The primers used in this study were obtained from Shanghai Sangon Biotechnology. Biotin-CGRP was synthesized by LifeTein, LLC (Hillsborough, NJ, United States). Unless otherwise specified, all other reagents were purchased from Sigma-Aldrich or GenScript Biotechnology (Shanghai, China).

The anti-CGRP antibody gene was synthesized and inserted into a pUC57 vector. Using the pUC57 vector as a template, the V_H and V_L genes were amplified by PCR with polymerase KOD-Plus-Neo (TOYOBO, Japan) using the primer pairs NcoICGRPVHback/XhoICGRPVHfor and SalCGRPVLback/NotCGRPVLforII. The amplification program used three steps repeated in 30 cycles: denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 60 s. The PCR products were recovered after being separated by agarose gel electrophoresis. The primers used in this study are listed in Table 1.

The pDong1 vector (Dong et al., 2009) and recovered V_H genes were digested with NcoI and XhoI at 37°C for 5 h. After digestion, agarose gel electrophoresis was performed, and the target fragments were recovered using a gel recovery kit. The recovered V_H fragment was then ligated with the pDong1 vector using Ligation high Ver.2 (TOYOBO Bio) at 16°C for 2 h. The ligation product was transformed into E. coli DH5α and spread on Luria-Bertani agar (LB) plates containing 100 μg/mL of ampicillin (LBA). The plates were incubated at 37°C for 12 h, and single colonies were selected after 12 h. These colonies were subjected to colony PCR using the primers M13rv, pHENseq, and Easy-Load PCR Master Mix (Beyotime Biotechnology, Shanghai, China). Colonies with successful PCR results were selected and cultured in 4 mL of an LB liquid medium at 37°C for 12 h. The plasmid DNA was extracted and used for downstream experiments.

Plasmid pDong1, including the V_H gene (pDong1-V_H) and the pre-amplified V_L gene, was subjected to enzymatic digestion using SalI and NotI and recovered after agarose gel electrophoresis. The V_L gene was then ligated into the recovered pDong1-V_H vector using the same method. The ligation product was transformed into E. coli DH5α and spread on LBA plates. After incubation at 37°C for 12 h, single colonies were picked, and colony PCR was performed using the primers SalCGRPVLback and NotCGRPVLforII. The colonies with the right DNA insert were cultured in 4 mL of an LB liquid medium at 37°C for 12 h, and plasmid DNA extraction was carried out. The plasmids were sequenced, and those with accurate sequences were designated as pDong1/Fab-CGRP and used for phage display.

The successfully constructed recombinant plasmid pDong1/Fab-CGRP was transformed into E. coli TG-1 competent cells for phage display. A TG-1 colony with pDong1/Fab-CGRP was picked and transferred to 2 YT medium with 100 μg/mL of ampicillin and 1% glucose (2YTAG) medium, where they were left to grow overnight. The next day, 40 μL of the overnight culture was transferred to 4 mL of a fresh 2YTAG medium. When the bacterial culture reached an optical density at 600 nm of approximately 0.4, 10⁹ PFU of the KM13 helper phage (Kristensen and Winter 1998) were added. The mixture was incubated at 37°C for 30 min, followed by incubation at 25°C for 10 min. After centrifugation at 3,300 g for 10 min at 4°C, the supernatant was removed. A total of 4 mL of the 2YTAG medium containing 50 μg/mL kanamycin (2YTAGK) was added to the remaining pellet. The culture was then incubated at 30°C for 20 h with shaking at 220 rpm. Following incubation, the bacterial culture was centrifuged at 4°C for 15 min at 3,300 g, and the supernatant was collected. To 4 mL of the supernatant, 1 mL of polyethylene glycol 6,000/2.5 M NaCl solution (PEG/NaCl) was added, and the mixture was thoroughly mixed before being placed on ice for 1 h. After the incubation, the mixture was centrifuged at 5,500 g for 1 h at 4°C. The supernatant was removed, and the pellet was centrifuged again to eliminate the remaining supernatant. The pellet was suspended in 200 μL of phosphate buffer saline (PBS), and after centrifugation at 11,600 g for 10 min at 4°C, the supernatant was collected and stored at −20°C.

Streptavidin (SA) was coated on microplates at 5 μg/mL and left at 4°C overnight. The next day, 2 μg/mL of biotin-labeled CGRP was added and incubated for 30 min. After blocking with PBS containing 2% skim milk (MPBS) for 2 h, the plates were washed twice with PBS containing 0.5% TWEEN 20 (PBST). Each well was incubated with 10⁹ CFU of phage-Fab at 25°C for 1 h. The microplates were washed six times with PBST, followed by the addition of a 5000-fold diluted anti-M13/HRP antibody in MPBS and incubation at 25°C for 1 h. After six washes with PBST, the 3, 3, 5, 5-tetramethylbenzidine (TMBZ) substrate was added for color development at 37°C. The reaction was terminated with 20% sulfuric acid, and the absorbance at 450 and 630 nm was measured using an iMark™ microplate reader (Bio-Rad).

Competitive ELISA was conducted to detect CGRP. SA was coated on microplates at 5 μg/mL, and biotin-labeled CGRP was added the next day at 20 μg/mL, with an incubation time of 30 min. After blocking with MPBS for 2 h, the plates were washed twice with PBST. In each well, CGRP solutions at 0, 0.05 nM, 0.5 nM, 1 nM, 2.5 nM, 5 nM, 10 nM, 25 nM, 50 nM, and 100 nM were added along with phage-Fab, and each well was brought to a total volume of 100 μL with MPBS. The plates were then incubated for 1 h. After six washes with PBST, the assay was performed according to the protocol described in Section 2.5.

The C_H1 gene of pDong1/Fab-CGRP had SgrAI restriction enzyme sites at both the 5′and 3′ends. The pDong1/Fab-CGRP vector was digested with SgrAI at 37°C for 5 h. After digestion, the DNA fragments were recovered and self-ligated with Ligation high Ver.2 at 16°C for 2 h. The ligation product was then transformed into E. coli DH5α and plated, followed by incubation at 37°C for 12 h. After 12 h, colony PCR was performed using primers M13rv and pHENseq. Colonies with bands of the correct size were picked after agarose gel electrophoresis and cultured in 4 mL of the LBA medium overnight. The next day, plasmids were extracted and sequenced, and the plasmid with the deleted C_H1 gene was collected and named pDong1/OS-CGRP.

pDong1-OS/CGRP was transformed into E. coli TG-1 cells, which were then plated. The following day, well-growing single colonies were picked and cultured in 4 mL of the 2YTAG medium at 37°C overnight. The following day, 40 μL of the culture was added to 4 mL of fresh medium and cultured at 37°C. When the OD₆₀₀ reached approximately 0.4, a KM13 helper phage was added, and infection was introduced. The culture was then centrifuged, and the pellet was resuspended in 2YTAGK and cultivated at 30°C for 20 h with shaking at 200 rpm. The culture was taken out and centrifuged at 5,500 g for 30 min at 4°C to obtain the supernatant, which contained the phage-V_H and L chains. The co-expression supernatant was subjected to phage titer testing.

Anti-L chain (κ) antibodies (MBL, Beijing) were diluted to a concentration of 2 μg/mL and coated on microplates at 4°C overnight. The following day, the plates were blocked with 3% MPBS for 2 h. After blocking, the plates were washed thrice with PBST. In each well, final concentrations of 0, 0.005, 0.05, 0.5, 5, 50, and 500 nM CGRP were added along with the phage-V_H and light chain co-expression supernatant. Instead of CGRP, the C-terminal peptide of human bone Gla protein (BGP-C7; NH₂-RRFYGPV-COOH) was added to the assay as a negative control. The volume in each well was brought to 100 μL with MPBS, and the plates were left at 25°C for 1 h. After incubation, the plates were washed six times with PBST. Then, a 1:5,000 diluted anti-M13/HRP antibody was added and incubated at 25°C for 1 h. The plate was washed six times with PBST, and the assay was developed using the methods described in Section 2.5.

To detect CGRP without relying on anti-L-chain antibodies, the vector pMAL-V_L (CGRP) was constructed to express V_L fused with maltose binding protein (MBP-V_L). The V_L gene was amplified by PCR using primers pMALCGRPVLSalFor and pMALCGRPVLNotRev under conditions same to those used for the V_L gene amplification. The PCR product was recovered, digested with SalI and NotI, ligated to the pMAL-DON vector (unpublished data), which was treated with the same enzymes. The ligation mixture was used to transform E. coli DH5α and plated on LBAG plates. After culture, colony PCR was performed using the primers pMALCGRPVLSalFor and pMALCGRPVLNotRev to screen for successful clones. The plasmid with the correct sequence was selected and designated as pMAL-V_L (CGRP).

The pMAL-V_L (CGRP) vector was transformed into E. coli BL21 (DE3). A single colony was picked and inoculated into 4 mL of an LBA medium, followed by incubation at 37°C for 12 h. A total of 3 μL of this culture was then inoculated into 300 mL of LB liquid medium and cultured at 37°C with shaking at 150 rpm until the OD₆₀₀ reached 0.6. Isopropyl β-D-1-thiogalactopyranoside was then added at a concentration of 0.4 mM, and the culture temperature was reduced to 16°C with shaking at 110 rpm for 20 h. The culture medium was collected and centrifuged at 9,500 g for 10 min to remove the supernatant. The bacterial pellet was resuspended in 10 mL of TALON washing buffer. After centrifugation to remove the supernatant, the cells were thoroughly washed with washing buffer. Subsequently, 10 mL of TALON washing buffer was added to completely resuspend the cells, followed by cell disruption. The bacterial lysate was centrifuged at 9,500 g for 20 min, and the supernatant was collected. A total of 300 μL of Ni-NTA Sefinose™ Resin was washed with TALON washing buffer, added to the supernatant of the disrupted cells, and rotated at 4°C for 2 h. The supernatant and Ni-NTA resin were transferred to a purification column. The column was washed with TALON washing buffer to remove impurities, and the target protein was eluted using an elution buffer and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

A total of 4 mL of a solution containing phage-V_H and L chain was mixed with 1 mL of PEG/NaCl and incubated on ice for 1 h, followed by centrifugation at 5,500 g for 1 h at 4°C, and the supernatant was removed. The pellet was suspended in 200 μL PBS, and then centrifugation at 11,600 g for 10 min at 4°C was carried out to recover the supernatant containing phage-V_H.

MBP-V_L was coated on microplates overnight at 20 μg/mL and blocked with 3% MPBS for 2 h after being washed thrice with PBST. After blocking, an MPBS solution containing 10⁹ CFU of phage-V_H and CGRP at a series of concentrations (0.005, 0.05, 0.5, 5, 50, and 500 nM) was added to the wells and incubated for 1 h. The microplates were washed six times with PBST, and an anti-M13/-HRP antibody diluted 1:5,000 in MPBS was added and incubated for 1 h. After washing six times, TMBZ was added, and the assay was performed as described in Section 2.5.

Molecular docking utilized the multimer model in Alphafold v2.3.2 to predict the multimeric structure of the antibody. The parameters used were all default settings. The database versions used were as follows: uniport and uniref90 as of 2023-03-01, pdb_mmcif and pdb_seqres as of 2023-03-03, and the remaining database versions were the default.

In the experiment of obtaining dose-response curves, three samples were set for each concentration, and three parallel experiments were conducted simultaneously to ensure the accuracy of the experiment. Dose-response curves were fitted to a four-parameter equation as shown following using Prism 9 (GraphPad Software, San Diego, CA, United States) and the goodness of fit was evaluated with R squared value. y = a + b − a 1 + x c d

In the competitive assays, the detection limit (LOD) was defined as the antigen concentration corresponding to the absorbance measured in the absence of antigen, minus three times the error value. In the OS-ELISA method, the LOD was defined as the antigen concentration corresponding to the absorbance measured in the blank plus three times the standard deviation. The half-maximal inhibitory concentration (IC₅₀) in the competitive assay was defined as the concentration at which half of the maximum binding was inhibited. The half-maximal effective concentration (EC₅₀) of the OS-ELISA method was defined as the antigen concentration at which the maximum value was halved.

The phagemid pDong1/Fab-CGRP was constructed to display the antigen-binding fragment (Fab) of the anti-CGRP antibody. As shown in Figures 1B,C, the V_H-C_H1 gene of the antibody was linked to the gIII gene of the M13 phage with a His-tag and C-myc-tag. This resulted in the V_H being displayed on the phage, and His-tag and C-myc-tag could be used for protein detection. The light chain of Fab contained V_L and IgG kappa chains with a FLAG tag on the N terminus. Kappa chain and C_H1 dimerize and result in Fab formation.

The V_H and V_L genes were amplified using PCR and confirmed by agarose gel electrophoresis. As shown in Figure 2A, two distinct bands at the 300-400 bp range were observed, indicating successful amplification of V_H and V_L. The PCR products were recovered and cloned into the pDong1 vector to obtain pDong1-V_H. V_L was cloned into pDong1-V_H using a similar approach, resulting in the phagemid vector pDong1/Fab-CGRP, which allows the generation of Fab-displaying phages via E. coli TG-1, a suppressor strain.

To validate the binding activity of Fab to the CGRP antigen, ELISA was performed following the principle illustrated in Figure 2B. Because CGRP is a small molecule that is difficult to immobilize directly onto a plate, an immobilization step with streptavidin was used. The streptavidin-coated plates were further used to immobilize biotinylated CGRP (Biotin-CGRP), and detection was achieved by the addition of MPBS containing 10⁹ CFU/mL of phage-Fab. The results are shown in Figure 2C. The absorbance of samples containing CGRP was 3.15, whereas that of samples without CGRP was 0.02, indicating that the Fab displayed on the phage surfaces had a high affinity for the CGRP antigen.

Indirect competitive ELISA was performed on plates containing immobilized streptavidin and biotin-labeled CGRP to detect CGRP, as shown in Figure 3A. When the free CGRP concentration in the solution was low, Fab was displayed on the phage bound to the immobilized Biotin-CGRP, resulting in a higher absorbance. As the free Biotin-CGRP concentration increased, phage-Fab was competitively bound to free Biotin-CGRP, leading to reduced binding to the plate-immobilized Biotin-CGRP and decreased absorbance. Using different concentrations of CGRP for competitive ELISA experiments, based on the measured data, we obtained a dose-response curve with an R ² value of 0.984. The LOD for CGRP was calculated to be 0.75 nM, with a detection range of 0.75–10 nM, and the IC₅₀ was 2.07 nM (Figure 3B).

The OS-ELISA employs separate V_H and V_L. To prepare elements for performing the OS-ELISA, the phagemid pDong1/OS-CGRP was prepared. We used the pDong1 system to delete the C_H1 open reading frame (ORF), resulting in the display of only the V_H on the phage. In the Fab display system, Fab is formed through the interaction between C_H1 and C_L of the light chain. Because there is no C_H1 gene in pDong1/OS-CGRP, we co-secreted the light chain and phage-V_H into the culture supernatant separately, allowing for OS-ELISA. Self-ligation was performed after processing with SgrAI (Figure 4A). The colony PCR results indicated that the sizes of the PCR products for clones 1, 2, and three were 650 bp, which was smaller than those of the PCR products obtained without removing the C_H1 plasmid, confirming a successful construction (Supplementary Figure S2). This streamlined approach enabled rapid assessment of the interaction between V_H and V_L and its antigen dependency without the need for recloning segments. One single colony was picked and cultured in 4 mL of liquid LBA medium for 12 h, followed by plasmid extraction and sequencing. Sequence results without errors were designated as pDong1/OS-CGRP.

An OS-ELISA with phage-V_H, light chain antibody, and anti-L chain antibody was performed, as illustrated in Figure 4B. A mixture of co-expression solutions with different CGRP concentration gradients was used. In the presence of the CGRP antigen, the light chain and phage-V_H exhibited apparent antigen dependency, resulting in stronger signals as the antigen concentration increased. Therefore, the final absorbance increased with increasing antigen concentration (Figure 4C). As the CGRP concentration in the samples increased, the signal gradually increased and stabilized. In contrast, the detection signal of the negative control sample using BGP-C7 did not increase, indicating that the specificity of the detection technique and the antibody displayed a clear antigen-dependent signal increase. The LOD and EC₅₀ for CGRP detection were 0.05 nM and 0.89 nM, respectively, with a detection range of 0.05–10 nM, determined using a dose-response curve with an R ² value of 0.993. OS-ELISA exhibited higher sensitivity and a broader detection range than competitive ELISA. From these results, we conclude that removing the C_H1 region of antibodies did not affect their specificity.

To simplify the OS detection technique, another OS-ELISA assay platform with phage-V_H and MBP-V_L was evaluated.

First, plamid pMAL-V_L(CGRP) for the expression of MBP-V_L was constructed. The V_L gene was amplified by PCR from pDong1/Fab-CGRP. The PCR product was subjected to agarose gel electrophoresis, and a distinct band at the 300–400 bp range (Supplementary Figure S3A) was observed, indicating successful amplification of the V_L gene. The gene was cloned into the pMAL vector, and the presence of V_L was confirmed by colony PCR (Supplementary Figure S3B) and sequence analysis. The successfully constructed pMAL-V_L (CGRP) (Figure 5A) was used to transform E. coli BL21 (DE3) cells, and protein expression was induced with IPTG. MBP-V_L was purified using a Ni²⁺ column and analyzed using SDS-PAGE. As shown in Figure 5B clear band was observed with a molecular weight of 55 kDa, corresponding to MBP-V_L with a theoretical molecular weight of 55 kDa. The SDS-PAGE results also demonstrated the purity of MBP-V_L of 93%, which was further utilized in OS-ELISA.

After coating with MBP-V_L, a mixture of phage-V_H and different CGRP concentrations was added. In the presence of the CGRP antigen, MBP-V_L and phage-V_H would bind together. The higher the antigen concentration, the higher the absorbance values (Figure 5C). As the CGRP concentration in the sample increased, the signal gradually increased and tended to stabilize, generating a dose-response curve as shown in Figure 5D. The LOD and EC₅₀ for this assay were 0.15 nM and 14.6 nM, respectively, with a wide detection range of 0.15 nM–500 nM, determined using a dose-response curve with an R ² value of 0.996. The detection signal for the negative control sample BGP-C7 did not increase, indicating the specificity of the assay.

The predicted multimeric structures were analyzed and identified amino acid pairs between proteins that can interact to form hydrogen bonds as shown in Figure 6. Specifically, isoleucine at position 16 (LEU-16), serine at position 19 (SER-19), and valine at position 23 (VAL-23) of CGRP formed hydrogen bonds with lysine at position 57 (LYS-57) of the heavy chain variable region. Asparagine at position 3 (ASN-3) of CGRP formed a hydrogen bond with tyrosine at position 32 (TYR-32) of the light chain variable region. The formation of these hydrogen bonds enhances the interaction between the antibody V_H and V_L.