A quantitative comparison of antibodies against epitope tags for immunofluorescence detection

Abstract

Epitope tags recognized by specific antibodies have been widely used over the last few decades, notably to localize tagged proteins within cells by immunofluorescence. The diversity of tags and antibodies usually prevents a side-by-side comparison of the efficiency with which each antibody recognizes its cognate tag. We expressed chimeric proteins, each composed of an invariant domain (IL2Ra) associated with a specific epitope tag. Double immunofluorescence allowed us to quantify in parallel the reference signal generated by the anti-IL2Ra antibody and the signal generated by the anti-epitope tag antibody. Since all antibodies used in this study were recombinant antibodies fused to the same mouse Fc domain, the generated signals were directly comparable. Three groups of tags/antibodies were revealed: 'good' antibodies generated high signals even when used at a low concentration (50 ng·mL^-1 ), 'fair' antibodies generated a high signal only at high concentrations (5000 ng·mL^-1 ), and 'mediocre' antibodies generated positive but weak signals. Except for an anti-myc antibody, similar results were obtained when cells were fixed in paraformaldehyde or methanol. These results provide a side-by-side quantitative evaluation of different tag/antibody pairs. This information will be useful to optimize the choice of epitope tags and to choose optimal antibodies.

Keywords: 6xHis; ABCD database; Epitope tags; FLAG; HA; Myc.

Fig. 1

Strategy to determine the relative efficiency with which antibodies recognize their cognate tag. (A) Plasmids encoding a fusion protein composed of the IL2Ra extracellular and transmembrane domains fused to a cytosolic tag (e.g., HA) were transfected in HEK cells. (B) The IL2Ra‐tag fusion protein was detected with an anti‐IL2Ra antibody (AJ519, red) and an anti‐tag antibody (e.g., AF291, blue). The fluorescence levels were determined over a number of small areas (white boxes). Scale bar: 20 μm. (C) The fluorescence levels were plotted for each antibody, and linear regression was used to calculate the relative efficiency with which the anti‐tag recognized its epitope. (D) For each antibody, the relative binding efficiency was compared with the value obtained for the AF291 antibody (set to 100), included in all experiments.

Fig. 2

Quantification of the signals generated by different antibodies. The signal generated by each antibody against its cognate tag was determined as described in Fig. 1 (mean ± SEM; n = 3 to 5 independent experiments). (A) In paraformaldehyde‐fixed cells, three antibodies generated weak specific signals (AI177, AF371, and AI179). (B) In methanol‐fixed cells, the specific signals generated by different antibodies were highly similar to those obtained in paraformaldehyde‐fixed cells except for anti‐myc antibodies, which generated much weaker specific signals.

Fig. 3

Amount of secondary but not primary antibody is limiting in our experimental setup. To assess whether primary or secondary antibodies were present in limiting amounts in our immunofluorescence procedure, we tested four dilutions of the primary (AF291) or of the secondary antibodies in parallel. The experiment was carried out as described in the legend to Fig. 2, with paraformaldehyde‐fixed cells (mean ± SEM; n = 50 cells). The signal decreased strongly when the concentration of the secondary antibody was diminished, but more gradually when the concentration of the primary anti‐tag antibody was decreased. This indicates that an excess of antibody is present when primary antibodies are used at a concentration of 5 μg·mL⁻¹.

Fig. 4

Detection of epitope tags in stringent conditions. In order to test antibodies in more stringent conditions, the efficiency with which each antibody bound its cognate tag in paraformaldehyde‐fixed cells was tested at low antibody concentration (50 ng·mL⁻¹). In these conditions, all tested antibodies still exhibited specific binding, except the TA002 anti‐myc antibody (mean ± SEM; n = 3 independent experiments).