|Home | About | Journals | Submit | Contact Us | Français|
One of the more useful protein tags for a protein in biochemical experiments is biotin, due to its femtomolar dissociation constant with streptavidin or avidin. Robust methodologies have been developed for the in vivo addition of a single biotin to recombinant protein or either in vitro enzymatic or chemical addition of biotin to a protein. Such modified proteins can be used in a variety of experiments, such as affinity selection of phage-displayed peptides or antibodies, pull-down of interacting proteins from cell lysates, or arraying proteins on arrays. We present three complementary approaches for biotinylating proteins in vivo in Escherichia coli, and biotinylating proteins in vitro either chemically or enzymatically that can be scaled up to tag large numbers of proteins in parallel.
Recombinant proteins are typically overexpressed in heterologous hosts (i.e., bacteria, insect cells, mammalian cells, plants) with “fusion tags”, which are short peptides, protein domains, or entire proteins which can be fused to proteins of interest, with the goal of imparting the biochemical properties of the fusion tag to the protein of interest. This is done at the genetic level by fusing the gene of interest to the gene encoding the fusion tag of interest, resulting in the expression of a single protein fused to the tag. In general, the type of fusion tag used is dictated by its application. Short peptide tags (e.g., six-histi-dine, epitopes, StrepTag, calmodulin-binding peptide) regularly serve to permit facile purification of the recombinant protein, permit detection of the fusion protein, or to direct interaction of the recombinant protein with other proteins or inert surfaces. Larger fusion partners, such as protein domains (e.g., chitin-binding domain) or proteins (e.g., cu-tinase, green fluorescent protein (GFP), glutathione-S-transferase (GST), intein, maltose binding protein (MBP), are commonly used to promote folding, solubility, purification, labeling, chemical ligation, or immobilization of the recombinant protein. If desired, the fusion tag can be detached from the protein of interest by cleavage of a linker region with a site-specific protease, which does not cleave the protein of interest.
One popular tag for detecting recombinant and native proteins is the small molecule biotin, which is a component of the vitamin B2 complex. It binds with high affinity to the chicken egg white protein, avidin, and the fungal protein, streptavidin. (Note that a degly-cosylated, recombinant form of avidin, with near-neutral isoelectric point (i.e., pI = 6.3) that minimizes nonspecific interactions, is commercially distributed as "neutravidin".) Avidin and streptavidin are tetrameric proteins that bind four molecules of D-biotin extremely tightly (i.e., dissociation constant of ~10−15 M) (1, 2). Proteins can be modified (i.e., biotinylated) with biotin very easily in vitro with chemical reagents, which are linked to biotin, under relatively mild condition that do not affect protein stability, three-dimensional structure, or function.
However, two drawbacks of in vitro chemically biotinylated target protein is that the number of biotin added is not uniform and the modification of certain lysine residues may lead to inactivation of the binding site(s). In E. coli, the biotin carboxy carrier protein (BCCP) is biotinylated by BirA (3), a biotin ligase which covalently attaches a biotin to the amino group of a lysine residue present within the recognition sequence within BCCP (4). A minimal biotinylation sequence has been found from screens of combinatorial peptide libraries; this 13 amino acid peptide (5), along with a 15 amino acid long variant (6), termed the AviTag™, have been identified as effective in vivo and in vitro substrates for the BirA enzyme. When targets proteins are fused to the AviTag and co-expressed in vivo along with BirA, they can be biotinylated in bacteria (7-9), yeast (10-12), insect (13), or mammalian cells (14, 15). Furthermore, when recombinant proteins are fused to the AviTag and incubated in vitro with purified BirA, they can be bi-otinylated efficiently on the central lysine residue in the AviTag (16, 17).
To generate biotinylated proteins in the laboratory, one has several options. If the protein is not available but is found to express well in E. coli, then it may be expedient to construct recombinant DNA in which its coding region is fused with the AviTag at its N- or C-terminus. (While there is a single biotin attached per protein molecule, the efficiency of biotinylation typically ranges between 50 and 80%.) Alternatively, if the protein is already available in sufficient amounts, then one can chemically biotinylate the protein prior to affinity selection experiments. (Typically, 100% of the molecules will be labeled, with one or more biotins one of the lysine residues.) Finally, one can construct recombinant DNA with the AviTag fused at the protein's N- or C-terminus, express it and purify it from E. coli, and then biotinylate the protein in vitro with purified BirA. (Typically, 80–100% of the target protein is biotinylated in vitro.) All three approaches are described herein.
To generate biotinylated proteins for affinity selection experiments, one can transfer the open reading frame (ORF) of a protein of interest into plasmids that contain both the AviTag biotinylation sequence and a six-histidine tag, at either the N- or C- terminus of the ORF. The AviTag encodes the peptide sequence, GLNDIFEAQKIEWHE, where the underlined lysine residue is biotinylated by BirA. The two bacterial expression vectors, pMCSG16 and pMCSG17, also contain a ligation independent cloning (LIC) site for efficient cloning of the ORFs, which allows high-throughput cloning, expression, in vivo bi-otinylation, purification, and streptavidin/avidin immobilization of target proteins for affinity selection of phage-displayed libraries (9). Expression of the target-AviTag fusion protein is under the control of the T7 RNA polymerase promoter, which is under the transcriptional control of the LacZ promoter in E. coli strain BL21 (DE3). To produce enough BirA in the bacterial cells, they also contain the pBirA Cmr plasmid (16), which carries resistance to chloramphenicol and a compatible origin of replication. A protocol for generating the recombinants in pMCSG16 and pMCSG17 is briefly described below, with more extensive protocols found elsewhere (9), Dr. Frank Collart's publication in this book).
To biotinylate of target proteins in E. coli, it is important to grow the expression plasmids in bacteria that contain the pBirA Cmr biotinylation plasmid (16). Without this plasmid, the levels of endogenous BirA enzyme are inadequate to get more than 5% of the overex-pressed protein biotinylated. It is also important to add biotin to the culture medium at the time of overexpression of the recombinant protein, to ensure that sufficient amounts of this molecule are available for post-translational modification of the target protein. A typical protocol for in vivo labeling of AviTagged proteins in bacteria (Fig. 1) consists of the following.
Pierce has developed a Slide-A-Lyzer™ dialysis cassette that can conveniently remove low molecular weight contaminants and salts. Determine the molecular weight and available volume of your protein to choose an appropriate dialysis cassette size and membrane capacity for the molecular weight cut-off (MWCO). Avoid choosing a membrane with a MWCO that is too close to the size of your protein; this can cause some loss of sample. To protect the physical integrity of the dialysis membrane, use precaution when handling the dialysis cassette and touch only the plastic frame.
Determine the molecular weight of your protein to choose an appropriate filter membrane capacity for the nominal molecular weight limit (NMWL). Avoid choosing a membrane with a NMWL that is too close to the size of your protein; this can cause some loss of sample. It should be noted, that generally some of the protein sample is lost due to sticking to the membrane.
Biotinylation of proteins can be monitored by three different means:
The authors acknowledge intellectual contributions by Mr. Michael Scholle and Dr. Frank Collart, and financial support from the National Institutes of Health (1R01 GM079096, P01 GM075913, 1U54 CA119343).