Influence of nanobody binding on fluorescence emission, mobility, and organization of GFP-tagged proteins

Abstract

Advanced fluorescence microscopy studies require specific and monovalent molecular labeling with bright and photostable fluorophores. This necessity led to the widespread use of fluorescently labeled nanobodies against commonly employed fluorescent proteins (FPs). However, very little is known how these nanobodies influence their target molecules. Here, we tested commercially available nanobodies and observed clear changes of the fluorescence properties, mobility and organization of green fluorescent protein (GFP) tagged proteins after labeling with the anti-GFP nanobody. Intriguingly, we did not observe any co-diffusion of fluorescently labeled nanobodies with the GFP-labeled proteins. Our results suggest significant binding of the nanobodies to a non-emissive, likely oligomerized, form of the FPs, promoting disassembly into monomeric form after binding. Our findings have significant implications on the application of nanobodies and GFP labeling for studying dynamic and quantitative protein organization in the plasma membrane of living cells using advanced imaging techniques.

Keywords: Biochemistry; Biochemistry Methods; Biophysical Chemsitry; Biophysics; Optical Imaging.

Graphical abstract

Figure 1

Change in excitation and emission spectra of recombinant GFP and EGFP in solution upon addition of unlabeled Nb (A–F) Excitation spectra for fluorescence detection at 510 - 520 nm (A and B) and emission spectra following 488 nm (C and D) and 405 nm excitation (E and F) of GFP-His (A, C, and E) and EGFP-His (B, D, and F) without Nb (green dashed line) and with excess of unlabeled Nb (solid black line). All spectra are averages of three measurements acquired at 2.5 μg/mL (87 nM) fluorescent protein and 10 μg/mL (720 nM) nanobody in PBS.

Figure 2

Effect of GFP-nanobody binding on GUV-anchored (E)GFP Data for His-tagged (E)GFP anchored to GUVs (98 mol% DOPC and 2 mol% DGS-NTA) before and after addition of unlabeled Nb as marked. (A) Representative confocal fluorescence microscopy images of the equatorial plane of GUVs decorated with GFP. Scale bar 10 μm. (A–E) (B) Normalized fluorescence intensity over time obtained from subsequently recorded confocal image frames at the equatorial plane of a GFP-tagged GUV, where arrow marks the time of Nb addition. Relative change in (C) molecular fluorescence brightness (cpm) and (D) in number of particles (N) of GFP before and after Nb addition as marked. Relative change in (E) cpm and (D) N of EGFP before and after Nb addition as marked. Values were determined from FCS experiments on individual (E)GFP-tagged GUVs. p-values were determined using the Kolmogorov–Smirnov non-parametric test. Number of data points is indicated on each graph.

Figure 3

Effect of Nb-binding on (E)GFP in the plasma membrane of live PtK2 cells GPI-anchored proteins GFP-LYPD6 (left panels) and GPI-EGFP (right panels). (A–D) Representative confocal images before and after addition of Nb for (A) GFP-LYPD6 and (B) GPI-EGFP in photon counting mode. Scale bars are 10 μm. Normalized fluorescence intensity traces for (C) GFP-LYPD6 and (D) GPI-EGFP for the cells as indicated in panel a and b, respectively (BG = background). Arrows show the time point when Nb was added. The intensities per frame represent mean values over each cell (see Transparent methods for details). (E and F) Enhancement of fluorescence was ≈1.1-fold for GFP and ≈1.5-fold for EGFP. Change in τ_D, N and cpm for (E) GFP-LYPD6 and (F) GPI-EGFP upon nanobody addition (values after Nb addition divided by values before). Change in average transit time (τ_D i.e. mobility), average fluorescing particle number (i.e. concentration, N), and molecular fluorescence brightness (cpm) upon Nb addition are determined from FCS experiments (one dot = one cell, for each cell 6–9 single FCS measurements were averaged, data was pooled from three different days). The values on top of ratios in e,f indicate p-values obtained from Wilcoxon sign-rank non-parametric tests with hypothetical median values of 1 (ratio of 1 would indicate no change upon Nb addition).

Figure 4

Effect of unlabeled Nb binding on mobility and brightness of GFP-LYPD6 and GPI-EGFP as probed by large sFCS data sets Analysis of diffusion dynamics and molecular brightness (cpm) of GFP-LYPD6 (top panels) and GPI-EGFP (bottom panels) expressed on PtK2 cells and the effect of unlabeled nanobody. (A and B) Transit time histograms for protein without (green) and with (magenta) presence of nanobody. It contains data from >2000 single FCS curves for GFP-LYPD6 and >700 curves for GPI-EGFP from >10 cells each. (C–F) (C and D) Two-dimensional pair value histograms (bivariate histograms) of transit times and cpms for control (without Nb) and with addition of nanobody (E, F) for GFP-LYPD6 (top panels) and GPI-EGFP (bottom panels).

Figure 5

Diffusion of labeled Nb (AbStar635P-Nb) on GPI-EGFP expressing PtK2 Cells (A) Representative normalized sFCS autocorrelation carpets for (A) GPI-EGFP before and after addition of AbStar635P-Nb (x-axis: correlation lag time, y-axis: line pixels (space), color scale: normalized correlation from zero (blue) to one (red)), revealing a shift of average transit time (yellow region, average transit time highlighted by the dashed line) toward shorter times after addition of labeled Nb. (B) τ_D values for GPI-EGFP before and after AbStar635P-Nb addition including mean values and standard deviations. (C) Normalized autocorrelation carpet for AbStar635P-Nb bound to PtK2 cells expressing GPI-EGFP. (D) Histogram of τ_D for GPI-EGFP (with and without Nb) and AbStar635P-Nb. The p-value given in panel B was calculated using the Kolmogorov–Smirnov non-parametric test.

Figure 6

Missing co-diffusion of labeled Nb and (E)GFP-tagged surface proteins PtK2 cells expressing GPI-EGFP or GFP-LYPD6 were treated with labeled Nb (AbStar635P-Nb) and FCCS data acquired. Positive cross correlation (CC) indicates interaction, i.e. co-diffusion. (A) point FCCS of GPI-EGFP and Nb. Autocorrelation EGFP is green, autocorrelation of AbberiorStar635P-labeled Nb is magenta and cross correlation (CC) is blue. (B) Representative dual-color intensity trace showing that the detection events for EGFP and AbStar635P-Nb rarely overlap in time. (C–E) Scanning FCCS data of GFP-LYPD6 expressed on the surface of PtK2 cells and Nb, with representative normalized auto-correlation data for (C) GFP-LYPD6, (D) AbStar635P-Nb and (E) normalized cross-correlation data of these two. The dashed black line indicates the average transit times. The temporal cross-correlation of these two dataset does not show any positive cross-correlation, i.e. no co-diffusion.