Chen et al. have developed a new method, based on chemoselective ligation of probes to a keto-biotin analog, for labeling cell surface proteins in living cells.

Methods for protein labeling have enabled significant advances in understanding protein function. However, currently available methods have disadvantages that limit their utility. Protein fusions, such as those with green fluorescent protein (GFP), provide a high level of specificity, but the large size of the tag can interfere with protein function. Peptide labels, including aptamers, are generally less disruptive, but the peptide tags now available either have lower specificity or rely on relatively unstable noncovalent interactions. In the February issue of Nature Methods, Chen et al. describe a method based on the specific, covalent labeling of peptide tags that combines the advantages of both approaches.

A biotin ligase, BirA, was known to biotinylate a specific lysyl side chain within a 15-residue acceptor peptide, AP. To adapt this reaction for detecting AP-tagged proteins, the authors synthesized a ketone-containing biotin analog (ketone 1, Fig. 1) to serve as a modification ‘handle’. In vitro, BirA efficiently catalyzed the transfer of ketone 1 to the acceptor lysine of AP that had been fused to cyan fluorescent protein fusion (CFP-AP). Chemoselective reaction of ketone 1 with fluorescein hydrazide and reduction of the resulting hydrazone led to fluorescent labeling of the fusion protein. Expressed CFP-AP was specifically detected in mammalian cell lysates with greater sensitivity than is provided by antibody-based methods.

Figure 1: Strategy for labeling cell surface proteins by acceptor peptide (AP) tagging with ketone 1 by BirA, followed by ligation of hydrazine derivatives to ketone 1 (Figure courtesy of Nature Methods).

Full figure and legend (16K) Figures & Tables index

Taking advantage of the absence of endogenous ketones on cell surfaces, the authors visualized tagged cell surface proteins in living cells. In cells with CFP-AP targeted to cell surfaces by fusion to a transmembrane domain (TM-CFP-AP), specific membrane-localized labeling was observed in transfected cells within 20 minutes with a sensitivity of ~10⁶ copies of the tagged receptors per cell (Fig. 2). GFP fusion to the epidermal growth factor (EGF) receptor (EGFR) is known to disrupt receptor function. In contrast, AP-EGFR showed a cellular distribution and a response to EGF similar to those of wild-type EGFR. Thus this approach may permit new insights into receptor trafficking.

Figure 2: Fluorescent labeling of cell-surface proteins in living cells.

HeLa cells expressing cyan fluorescent protein (CFP)-acceptor peptide (AP) fused to a transmembrane (TM) domain were labeled with Ketone 1, followed by a biotinylated hydrazine probe. (a) Difference interference contrast (DIC) image demonstrates localization of TM-CFP-AP constructs in two cells. (b) Membrane labeling of cells expressing TM-CFP-AP by hydrazine probe (Figure courtesy of Nature Methods).

Full figure and legend (81K) Figures & Tables index

As the authors discuss, this method has limitations. Labeling intracellular proteins may prove challenging because of ketone-containing small molecules inside cells. In addition, the time scale of labeling may limit the biological processes that can be studied. Nonetheless, this method may open important new avenues of research into fundamental biological processes such as cell-cell interactions.

References

1. Chen, I. Howarth, M. Lin, W. & Ting, A.Y. Site-specific labeling of cell surface proteins with biophysical probes using biotin ligase. Nature Methods 2, 99−104 (2005). | Article |