Rapid Covalent Labeling of Membrane Proteins on Living Cells Using a Nanobody-Epitope Tag Pair

Abstract

Synthetic molecules that form a covalent bond upon binding to a targeted biomolecule (proximity-induced reactivity) are the subject of intense biomedical interest for the unique pharmacological properties imparted by irreversible binding. However, off-target covalent labeling and the lack of molecules with sufficient specificity limit more widespread applications. We describe the first example of a cross-linking platform that uses a synthetic peptide epitope and a single domain antibody (or nanobody) pair to form a covalent linkage rapidly and specifically. The rate of the cross-linking reaction between peptide and nanobody is faster than most other biocompatible cross-linking reactions, and it can be used to label live cells expressing receptor-nanobody fusions. The rapid kinetics of this system allowed us to probe the consequences on signaling for ligand cross-linking to the A2A-adenosine receptor. Our method may be generally useful to site-specifically link synthetic molecules to receptors on mammalian cell surfaces.

Figure 1.. Design, synthesis, and preliminary evaluation of the Nb_6E-6E epitope crosslinking pair.

A) Summary of structure-activity relationship studies of the interaction between Nb_6E and the 6E peptide epitope. Amino acid residues are mapped onto a hypothetical α-helical wheel diagram and a putative interface for binding to Nb_6E is illustrated. Orange arrows indicate sites evaluated for crosslinker installation. B) Synthetic scheme for preparation of crosslinking peptide-electrophile conjugates using cysteine-maleimide chemistry. Sites with mutations relative to the original peptide sequence are shown in colored font. Mass spectrometry data from LC/MS is shown at right. The full set of MS data is shown in Table S1. C) Schematic reaction between Nb_6E and Ac-1(NO₂). Deconvoluted MS data for Nb_6E (5 μM) before and after incubation with Ac-1(NO₂) (20 μM) for 1 hour is shown at right.

Figure 2.. Evaluation of epitope-nanobody crosslinking kinetics.

A) Measurement of the rate of reaction between Nb_6E and crosslinking peptides with varying ester groups (see Figure S1 for full chemical structures). Nb_6E (5 μM) was mixed with an excess of the indicated crosslinking peptides (20 μM). The abundance of remaining non-crosslinked Nb_6E was quantified using LC/MS. B) Assessment of crosslinking kinetics using in-gel fluorescence. Nb_6E (100 nM) was mixed with an excess of FAM-1(NO₂) (5 μM). Crosslinking was quenched at the indicated time points by addition of an excess of unlabeled competitor 6E peptide. Samples were run on SDS-PAGE and formation of the crosslinked product was assessed by in-gel fluorescence with quantitation by densitometry. Additional data from all in-gel fluorescence experiments is found in Figure S2E. C) Comparison of second-order rate constants for different types of chemoselective reactions. Rate constant ranges are adapted from previous work. SPAAC: strain-promoted azide alkyne cycloaddition. CuAAC: copper-catalyzed azide alkyne cycloaddition reaction. LDBB: ligand-directed dibromophenyl benzoate reaction. Tetrazine and TCO: transcyclooctene and tetrazine reaction. LDNASA: ligand-directed N-acyl-N-alkyl sulfonamide reaction. Enzymatic labeling: SNAP tag and Halo tag reactions. The rate of the reaction characterized in panel B (97,000 M⁻¹⁻s⁻¹) is marked with the green star. See Figure S2F for an estimate of the model-associated uncertainty for this rate constant.

Figure 3.. Evaluation of epitope-nanobody crosslinking on mammalian cells.

A) Schematic of A2AR(Nb_6E) binding to labeled 6E peptide (cyan) or unlabeled competitor peptide (red). The model of receptor is hypothetical and prepared using Pymol and a published crystal structure (PDB: 4UG2). See Supporting Methods for sequence information of this receptor. B) Comparison of cells expressing A2AR(Nb_6E) treated with FAM-1 (left) or FAM-1(NO₂) (right). Cells were treated with 100 nM of fluorescein-labeled peptides on ice for 30 m, washed, and then exposed to unlabeled competitor peptide (6E) at the indicated concentrations for an additional 30 m. After washing, FAM-labeled peptides were detected with Alexafluor647 (AF647) modified anti-FAM antibody and analyzed by flow cytometry. C) Kinetics of covalent labeling of cells expressing A2AR(Nb_6E). Cells were labeled with 100 nM of FAM-1(NO₂) for the indicated durations at which time unlabeled competitor peptide was added to prevent further labeling. Positive control cells were incubated a FAM-labelled peptide (FAM-6E-C14) shown in previous work to label this receptor with high affinity. Cells were washed and analyzed as in panel B. “No stain” indicates no FAM-labeled peptide was added. “No competitor” indicates no unlabeled competitor peptide was added. D) Quantitation of data from panel C. Data correspond to mean ± SD from three independent experiments. E) Analysis of probe labeling specificity using Western Blotting. Lane 1: No probe added; Lane 2: 100 nM FAM-1(NO₂); Lane 3: Pretreatment with 10 μM unlabeled competitor peptide (6E) followed by treatment with 100 nM FAM-1(NO₂). Samples were resolved by SDS-PAGE and transferred to PDVF membranes. Labeled proteins were detected using horseradish peroxidase (HRP) conjugated anti-FAM antibody or HRP-anti-GAPDH. F) Analysis of crosslinking peptide labeling using fluorescence microscopy. HEK293 cells expressing A2AR(Nb_6E) were stained with FAM-1(NO₂) (100 nM) and Nb_Alfa-TMR (300 nM) either with (top) or without (bottom) pre-treatment with unlabeled competitor 6E peptide. Cells were washed, fixed, stained with DAPI and imaged. Left and middle panels show identical images with filter sets applied as indicated. Right panels show the boxed inset from the middle panel.

Figure 4.. Assessment of biological consequences of epitope tag-mediated crosslinking on signaling properties.

A) Synthetic scheme for synthesis of conjugates consisting of a 6E peptide analogue with or without crosslinking capacity and an A2AR ligand (CGS21680) using copper catalyzed azide alkyne cycloaddition chemistry. B) Schematic comparing association between a crosslinking peptide-ligand conjugate and a non-crosslinking peptide-ligand conjugate. C) Dose-response assays for the induction of cAMP responses in cells expressing A2AR(Nb_6E). Curves are generated from a sigmoidal dose-response model with variable slope. Replicate data and tabulation of results are shown in Figures S14–15. D) Analysis of kinetics of ligand-derived signaling following the removal of free ligand (at 12 m) in a washout assay. For panels C-D, the data points show mean ± SD from a single experiment.