Background

Fluorescence microscopy (FM), electron microscopy (EM), and correlative light and EM (CLEM) enable investigations into the multi-protein complexes and macromolecular interactions that mediate normal and disease-associated cellular functions. However, multiscale microscopy is restricted by the shortage of methods for attaching FM-, EM-, and CLEM-compatible reporter chemistries to target proteins. Additionally, there are few methods for protein labeling that facilitate switching between imaging systems. As a result, most multiscale imaging studies obtain protein-specific contrast with immunolabeling. However, there are known drawbacks to immunolabeling. The large size of antibodies reduces localization precision, and labeling protocols can disrupt cellular ultra-structure (Schnell et al., 2012). Scarce proteins and rare interactions can elude detection unless immunolabeling is efficient (Griffiths and Hoppeler, 1986; Schnell et al., 2012; Griffiths and Lucocq, 2014). Many antibodies have poor target specificity and cross-reactivity (Berglund et al., 2008; Bordeaux et al., 2010; Baker, 2015; Bradbury and Pluckthun, 2015), which can result in misleading observations. 

The central obstacle that has limited progress in multiscale microscopy is the shortage of genetic tags for labeling proteins. Most tags were developed for FM (Liu et al., 2015), with the most commonly used tags being fluorescent proteins [e.g., GFP] (Tsien, 1998; Cranfill et al., 2016; Rodriguez et al., 2017). By comparison, there are few genetic tags for EM or CLEM (Ellisman et al., 2012). We saw this as an opportunity to create a new class of genetically-encoded peptide tags for multiscale microscopy (Zane et al., 2017; Doh et al., 2018). We named this technology Versatile Interacting Peptide (VIP) tags (Figure 1). VIP tags consist of a heterodimeric coiled-coil between a genetically-encoded peptide tag and a reporter-conjugated peptide (“probe peptide”). Binding is driven by a hydrophobic interface and inter-strand salt bridges between the two coils. Initially we reported VIP Y/Z, which was used to label cellular proteins with fluorophores and Qdots (Zane et al., 2017). This pair consists of a heterodimeric CoilY-CoilZ pair with a reported dissociation constant (K_D) of less than 15 nM (Reinke et al., 2010). Either CoilY or CoilZ could serve as the genetically-encoded tag. In 2018, we reported the VIPER tag, which enables high-affinity labeling of proteins for imaging by FM and CLEM (Doh et al., 2018). Binding between the CoilE tag and the CoilR probe peptide to form VIPER is specific and nearly irreversible [K_D ~10^-11 M] (Moll et al., 2001).


Figure 1. Versatile interacting peptide (VIP) tags are a new technology for imaging proteins by FM, EM or CLEM. VIPER labeling of transferrin receptor 1 (TfR1) is mediated by heterodimer formation between the CoilE tag and a fluorescent CoilR probe peptide. Fluorescent micrograph: VIPER-tagged TfR1 labeled with CoilR-Cy5 (magenta) and colocalized with fluorescent transferrin (Tf-AF488; green) at the cell surface of transfected CHO TRVb cells (63x magnification). Magenta-green signal overlap appears white and nuclei are blue.


Figure 2. VIP tags are a versatile technology for multi-scale microscopy. After a target protein is tagged, it can be labeled using a variety of probe peptides selected for the particular application.

For VIP tags, the versatility is imparted by the customizable probe peptide. After introduction of the CoilE tag onto a target protein, the protein can be labeled with one of many different reporters attached to CoilR (Figure 2). For example, we imaged the transmembrane receptor, TfR1-CoilE, with CoilR-BODIPY, CoilR-Cy5 (see Figure 1), and CoilR-biotin (Doh et al., 2018). In other words, the probe peptide can be customized for different studies or imaging systems without changing the genetic tag. This is possible because CoilR encodes a single cysteine residue for site-specific modification via thiol-maleimide chemistry. The CoilR probe peptide can be bioconjugated to a variety of probes, including fluorophores, small molecules (e.g., biotin), or nanoparticles. Many companies sell thiol-reactive probes, which makes this conjugation reaction accessible to labs without synthetic chemistry expertise. For more information on bioconjugation reactions, we recommend reading Hermanson’s Bioconjugate Techniques (Hermanson, 2013).

In this Bio-Protocol, we provide methods for making CoilR probe peptides that can be used for VIPER-labeling of cellular proteins for imaging by FM or EM. The CoilR peptide and the CoilE tag sequences are provided in Table 1. As described in prior work (Doh et al., 2018), we used gene assembly PCR to enable the recombinant expression of probe peptides in E. coli. The method for peptide expression is described in Procedure A. CoilR was designed to interact with CoilE via an optimized alpha-helical coil-coil, as originally described by Vinson and coworkers (Moll et al., 2001). We included a hexahistidine tag at the C-terminus of CoilR for purification by immobilized metal affinity chromatography (IMAC) (Hochuli et al., 1987); this is described in Procedure B. 

Table 1. Sequences of CoilR and CoilE

^§Italics: Linker sequence; Bold: Peptide coil; C: Cysteine (conjugation site).
^‡Heptad position: Residues a and d form a hydrophobic interface, residues at e and g form inter-strand salt bridges. 

Procedures C and D describe thiol-maleimide reactions to label CoilR with a small molecule reporter. In Procedure C, we describe the method that we used to generate probe peptides in our prior work (Doh et al., 2018). In Procedure D, we adapted a method described by Weiss and coworkers for solid state-based labeling of peptides (Kim et al., 2008). Lastly, we include methods for purifying fluorophore-labeled (Procedure E) or biotinylated (Procedure F) probe peptide. This Bio-Protocol is accompanied by two companion articles, which include detailed methods for imaging VIPER-labeled cellular proteins by FM (Doh et al., 2019a) and CLEM (Doh et al., 2019b) (Figure 3). 


Figure 3. A decision tree for implementing VIPER. Procedures are color-coded by the publication in which they appear. Methods in this publication are color-coded purple. Methods in Doh et al., 2019a are yellow and methods in Doh et al., 2019b are orange. Publication 1: this article; Publication 2: Doh et al., 2019a; Publication 3: Doh et al., 2019b).