A Single Bioorthogonal Reaction for Multiplex Cell Surface Protein Labeling

Abstract

Small-molecule fluorophores are invaluable tools for fluorescence imaging. However, means for their covalent conjugation to the target proteins limit applications in multicolor imaging. Here, we identify 2-[(alkylthio)(aryl)methylene]malononitrile (TAMM) molecules reacting with 1,2-aminothiol at a labeling rate over 10⁴ M^-1 s^-1 through detailed mechanistic investigation. The fast TAMM molecules and mild reaction conditions enable site-specific labeling of surface proteins in not only cell lines but also primary neurons and living mice. The combination of genetic code expansion and sequence-specific proteolytic cleavage enables selective modification of three different cell surface proteins through iterative TAMM condensation. TAMM condensation is also compatible with Cu-catalyzed azide-alkyne cycloaddition and tetrazine ligation for four-color fluorescent labeling, reaching the maximum available colors of conventional confocal microscopes. Thus, bioconjugation chemistry is no longer the limiting factor for multiplex cell surface protein imaging.

Figure 1

Mechanistic study of TAMM condensation. (A) TAMM condensation mechanism elucidated in this work with two identifiable intermediates Int-1 and Int-2. (B) TAMM molecules employed in mechanistic and kinetic studies. The observed consumption rate constants () at 25 °C and pH 7.4 correspond to reaction with 2x. The stability of TAMM is indicated by the half-life (t_1/2) in PBS (10 mM, pH 7.4) at 37 °C. n.d. = not determined. (C) 1,2-Aminothiol molecules employed in mechanistic and kinetic studies. (D) A representative TAMM condensation between 1a and model peptide 2x to form dihydrothiazole 3ax. Left: HPLC chromatograms of the reaction mixture analyzed at the indicated time point (λ_monitor = 280 nm). Right: time evolution of [1a] and [2x] (empty circles: data points, solid lines: fitting curves to the model of second-order reaction). (E) Effect of TAMM para substitution (R¹) on reaction kinetics. Hammett plot of rate constants for the consumption of 1 and 2x versus the Hammett substituent constants σ_p at 25 °C; slope ρ = 0.93. (F) Formation of an intermediate in TAMM condensation. HPLC chromatograms for reactions between 1a–1d and 2x at 42 min. Peak of the intermediates is indicated with a black circle with the m/z value of the quasi molecular ion (either as [M + H]⁺ or [M–H]⁻) shown. (G) Formation of dihydrothiazole 3ay from disulfide 9ay by TCEP reduction. Left: HPLC chromatograms for the reaction mixture analyzed at the time specified. Right: time course of the concentrations of 7ay and 3ay (empty circles: data points, solid lines: fitting curves to the model of first-order kinetics). (H) Reaction of TAMM 1i with peptide 2x to form dihydrothiazole 3ax. Left: HPLC chromatograms at the time specified. Right: time course of the HPLC peak intensities. The areas under the peaks, as opposed to the concentrations, are shown here as the calibration of Int-1, an unstable intermediate, is not available. The dashed lines connecting data points serve as a guide to the eye. (I) Formation of Int-2 (10) for fast-reacting TAMM supported by mass spectrum of 10ax for reactions involving 1i with peptide 2x.

Figure 2

Comparison of ethanethiol or 2,2,2-trifluoroethanethiol as the leaving group in TAMM conjugates for fluorescent labeling of a cell-surface protein on live mammalian cells. (A) Structure of Cy5-TAMM-SEt and Cy5-TAMM-SCH₂CF₃. (B) In-gel fluorescence analysis of HEK293T cells overexpressing C-HA-Nlgn3 labeled with Cy5-TAMM-SEt or Cy5-TAMM-SCH₂CF₃. FL: fluorescence; low: low exposure (0.5 s); high: high exposure (10 s); IB: immunoblotting. Quantification of the fluorescence intensity is shown in Figure S21B. (C) Representative confocal microscopy images of HEK293T cells overexpressing C-HA-Nlgn3-mCherry labeled with 2 μM Cy5-TAMM-SEt or Cy5-TAMM-SCH₂CF₃ for 60 min. Merged images and two other sets of data are shown in Figure S22. (D) Representative confocal microscopy images of primary mouse cortical neurons overexpressing C-HA-Nlgn3-mCherry labeled with Cy5-TAMM-SCH₂CF₃. Transfected and nontransfected neurons are shown in solid and hollow yellow arrows. The zoom-in section shows labeling of the transfected neuron (i.e., mCherry positive) by TAMM labeling and α-HA immunostaining. Scale bars of all images = 20 μm.

Figure 3

Iterative TAMM condensation for labeling three different target proteins. (A) Schematic presentation of the experimental flow. HEK293T cells overexpressing CysK-HA-Nlgn3 were first labeled with BODIPY-TAMM-SCH₂CF₃. FXa protease treatment then generated proteins with N-terminal cysteine on cells overexpressing IDGR-C-HA-Nlgn3 for labeling with Cy5-TAMM-SCH₂CF₃. Lastly, enterokinase treatment removed DDDDK of DDDDK-C-HA-Nlgn3 for labeling with Cy3-TAMM-SCH₂CF₃. (B) Structure of BODIPY-TAMM-SCH₂CF₃ and Cy3-TAMM-SCH₂CF₃. (C) Representative confocal microscopy images. Scale bar = 20 μm.

Figure 4

Using TAMM condensation, CuAAC, and tetrazine ligation for multicolor fluorescent labeling. (A) Structure of unnatural amino acids CysK, PPA, BCNK, and HexC. (B) Substrate scope of the orthogonal aminoacyl-tRNA synthetases. (C) Structure of FITC-N₃ and Pyranine-tetrazine. (D) Using three unnatural amino acids to control the bioorthogonal reactions for labeling three target proteins on different HEK293T cells. (E) Using two sequence specific proteases for iterative TAMM condensation in combination with tetrazine ligation and CuAAC for labeling four different target proteins on HEK293T cells. (D,E) Schematic presentation of the experimental flow is shown at the top. Representative confocal microscopy images are shown at the bottom. Two other sets of data are shown in Figures S32 and S33. Scale bar = 20 μm.

Figure 5

Using TAMM condensation and tetrazine ligation for dual-color labeling in living mice. (A) Schematic presentation of the experimental flow. (B) Structure of the functional groups on 293T-Cys and 293T-BCN. (C) Mice treated with Cy5-TAMM-SCH₂CF₃. (D) Mice treated with Cy7-tetrazine. (E) Cy5 and Cy7 fluorescence of the living mice treated with two dye conjugates and their major organs. Maximum and minimum epi fluorescence values are shown at the top and bottom of the scale bars.