Site-Specific Encoding of Photoactivity in Antibodies Enables Light-Mediated Antibody-Antigen Binding on Live Cells

Abstract

Antibodies have found applications in several fields, including, medicine, diagnostics, and nanotechnology, yet methods to modulate antibody-antigen binding using an external agent remain limited. Here, we have developed photoactive antibody fragments by genetic site-specific replacement of single tyrosine residues with photocaged tyrosine, in an antibody fragment, 7D12. A simple and robust assay is adopted to evaluate the light-mediated binding of 7D12 mutants to its target, epidermal growth factor receptor (EGFR), on the surface of cancer cells. Presence of photocaged tyrosine reduces 7D12-EGFR binding affinity by over 20-fold in two out of three 7D12 mutants studied, and binding is restored upon exposure to 365 nm light. Molecular dynamics simulations explain the difference in effect of photocaging on 7D12-EGFR interaction among the mutants. Finally, we demonstrate the application of photoactive antibodies in delivering fluorophores to EGFR-positive live cancer cells in a light-dependent manner.

Keywords: antibodies; cancer; protein design; synthetic biology; unnatural amino acids.

Figure 1

Genetic site‐specific incorporation of pcY in 7D12. A) Crystal structure of 7D12 (grey)–EGFR domain III (yellow) complex (PDB ID: 4KRL)23 showing Y32, Y109, and Y113 (pink) in the antigen binding pocket of 7D12, that were replaced with pcY. B) The expression of three amber mutants of 7D12, viz. 32TAG, 109TAG and 113TAG only occurs in the presence of pcY. Comparison of band intensities for amber mutants with wt7D12 shows efficient incorporation of pcY.

Figure 2

Assessing the binding of photocaged mutants of antibody fragment to EGFR on cell surface. A) Schematic representation of procedure followed for measurement of 7D12–EGFR binding on the surface of A431 cancer cells. 1. 40 000 cells were seeded into each well of a 96‐well plate. 2. These cells were incubated with the complete media containing the antibody fragment. 3. The antibody solution was replaced with 3.7 % formaldehyde solution for fixing the cells. 4. Incubation with blocking solution. 5. Incubation with primary antibody specific for hexa‐histidine tag. 6. incubation with HRP‐linked secondary antibody. 7. The substrate for HRP was added and the cells were imaged for chemiluminescence (Page S5). B) Comparison of ESI‐MS of photocaged mutants of 7D12 before and after irradiation with 365 nm light confirms light‐mediated decaging. See Page S6 for decaging conditions. C) The on‐cell binding assay demonstrates that the presence of pcY at positions 32 and 113 in 7D12 inhibits its binding to EGFR. 7D12pcY109 mutant does not show inhibition in binding to EGFR. Binding of 7D12pcY32 and 7D12pcY113 is restored upon irradiation with 365 nm light. These experiments were performed in triplicate (Figure S8). D) Chemiluminescence intensity was quantified using a CLARIOstar plate reader and plotted against concentration of 7D12, where X‐axis is in log scale; the data was fitted to a sigmoidal nonlinear curve using GraphPad (Figure S9). Some error bars are too small to be clearly visible.

Figure 3

MD simulations of wt7D12 and its photocaged mutants show that wt7D12 and 7D12pcY109 form more stable complexes with EGFR domain III as compared to 7D12pcY32 and 7D12pcY113. A) Simulation snapshots taken at intervals of 30 ns during 300 ns simulations for each system, (EGFR: grey for all, wt7D12: black, 7D12pcY32: red, 7D12pcY109: green, 7D12pcY113: blue) highlight the extent of motion of 7D12. B) Left: RMSDs from starting structure for protein Cα atoms during simulations show large deviations for 7D12pcY32 and 7D12pcY113. Right: The R30–D355 salt‐bridge (residues shown in (A) wt7D12 snapshot, in yellow), monitored as the distance between the R30 guanidine C and the D355 carboxyl C, breaks frequently in the 7D12pcY32 and 7D12pcY113 systems. These observations suggest that the presence of pcY at positions 32 and 113 destabilizes the 7D12–EGFR domain III complex.

Figure 4

Light‐mediated delivery of fluorophores by photoactive antibodies on live A431 cells. A) Labeled 7D12pcY32 is injected at 5 min. Near‐background fluorescence is observed 1.5 min after passing labeled 7D12pcY32 over live A431 cells demonstrating that due to the presence of caged group, 7D12pcY32 does not bind to the cell surface. B) Background fluorescence before re‐injecting labeled 7D12pcY32. C) Labeled 7D12pcY32 was injected at 17 min and the irradiated with 365 nm light at 18 min (1 min after injecting 7D12pcY32) for 1 min. Significant fluorescence was observed 1.5 min after stopping the injection of labeled 7D12pcY32, demonstrating light‐dependent localisation of 7D12 on the surface of A431 cells. D) Fluorescence from 7D12 reduces to background level. E) Labeled wt7D12 was injected at 29 min. Significant fluorescence observed 1.5 min after stopping the injection of labeled wt7D12 due to localisation of labeled wt7D12 on the surface of A431 cells. F) Fluorescence from wt7D12 slowly reduces to near background level. See Movie S1 in the Supporting Information.