How to Choose a Protein Expression Tag: Purification and Cleavage

A protein expression tag is a peptide or fusion domain that is added to a recombinant protein so the product can be captured, solubilized, detected, or later cleaved. How to choose a tag for protein expression and purification is a trade-off among affinity chemistry, solubility rescue, antibody detection, and whether the tag must be removed.

N-terminal versus C-terminal placement and the protease site belong to the same decision. A tag that purifies well can still ruin folding or leave scar residues that affect activity.

Decide What the Tag Must Do Before You Name One

Tags are asked to do four different jobs, and one peptide rarely does all four well. Purification wants a capture chemistry that works in your buffer. Solubility wants a large, well-folded partner that can pull a difficult ORF into the soluble fraction. Detection wants an epitope an antibody or resin can still see on a blot or a pull-down. Cleavage wants a protease site that can restore a nearly native terminus without destroying the protein in the process.

Write the primary job first. If the protein is for enzymology or a structure series, purification and later cleavage usually outrank a fluorescent fusion. If the protein is a bait in a lysate, detection and mild elution may outrank yield. If the ORF crashes into inclusion bodies in every trial, a solubility partner such as MBP or SUMO is the actual first decision, and the affinity handle can sit on that partner or in tandem.

Secondary jobs create dual-tag designs (for example His plus FLAG, or MBP plus His). Dual tags add junctions, extra proteases, and extra ways to go out of frame. They are justified when one chemistry captures and the second proves identity, not when the lab has not chosen a primary job.

Tag Trade-Offs: His, GST, MBP, FLAG, Strep, and SUMO

The six options below cover most bacterial and many eukaryotic purification designs. Sizes are approximate; always read the actual amino acid sequence on your map. The table is a trade-off sheet, not a vendor ranking.

His remains the default when the protein is reasonably soluble and the next step is IMAC plus a polish. GST is a reasonable default when glutathione chemistry is already in the lab and the protein tolerates a dimerization-prone fusion. MBP is the usual first solubility attempt for bacterial expression of stubborn eukaryotic domains. FLAG and Strep are small affinity or detection handles when you do not want 26 to 40 kDa glued to the product. SUMO is a solubility strategy with a built-in cleavage philosophy rather than a stand-alone capture resin story.

Solubility Fusions Versus Small Affinity Peptides

Small tags (His, FLAG, Strep) rarely rescue folding. They are the right size when the untagged protein is already soluble, when the tag must stay out of a crystal lattice after cleavage, or when a blot or pull-down is the real assay. They fail when the ORF never leaves the pellet; a better IMAC wash will not fix an inclusion body.

Large fusions (GST, MBP, SUMO) change the expression problem. They can raise soluble yield, and they can also create a soluble fusion whose passenger domain is inactive, aggregated after cleavage, or still unfolded once the partner is removed. The evaluation is therefore two-step: is the fusion soluble, and does the target remain soluble and active after the partner is cleaved? If the answer to the second question is no, the fusion was a purification convenience, not a folding solution.

Labs sometimes stack a small handle on a large fusion (His-MBP, His-SUMO, His-GST) so capture does not depend on amylose or glutathione alone. That layout is coherent when each part has a job. It is incoherent when three tags are added because previous constructs failed and nobody retired the extras. Every extra junction is another place to lose frame or to leave a scar.

N-Terminal Versus C-Terminal Placement

The terminus is not a cosmetic choice. The N-terminus is translated first. A tag there can interfere with a signal peptide, a modification, or co-translational folding. A tag at the C-terminus can sit inside a folded bundle, get clipped by cellular proteases, or block a C-terminal motif the protein needs. Secreted proteins usually need the signal peptide at the extreme N-terminus, with the affinity tag after processing if the mature protein must still bind resin.

His tags are often tried at both ends because they are small; many ORFs still prefer one orientation. MBP and GST are commonly N-terminal in bacterial vectors because that is how most solubility cassettes were built, not because C-terminal MBP is forbidden. SUMO protease cleavage is typically used on N-terminal SUMO fusions so the protease can restore a defined N-terminus on the passenger. FLAG and Strep need surface exposure; burying either peptide in a domain junction will make an antibody or Strep-Tactin resin look as if it failed.

When both orientations are biologically plausible, a two-construct pilot is cheaper than a large failed prep. Design both maps, keep the linker and protease site constant, and let a small expression and capture test decide. Folding is not fully predictable from hydropathy plots.

Protease Sites, Scar Residues, and When to Cleave

Cleave when the tag will interfere with activity, oligomeric state, crystallization, immunization, or a downstream assay that assumes a native terminus. Do not cleave when the tag is the detection handle you still need, or when the fusion is more stable than the isolated protein and the tag is tolerated. Cleavage is a unit operation with its own yield loss, not a default checkbox on every ORF.

TEV, SUMO protease (Ulp1 family), 3C-type proteases, thrombin, and Factor Xa are common. They are not interchangeable. TEV commonly leaves a short scar (often an extra Gly or similar residue depending on the site design). SUMO protease can leave a native N-terminus if the passenger starts immediately after the SUMO fold. Thrombin and Factor Xa have more off-target risk in some proteins. Place the protease site between the tag and the passenger, not inside a structured domain, and give it a short flexible linker so the protease can dock.

After cleavage, the large fusion partner should be removable by a second capture (the tag binds, the passenger flows through, or the reverse). If you cannot separate the cleaved tag from the target, you have only changed the apparent molecular weight on a gel. Design that second step at the same time as the tag, especially for GST, MBP, and His fusions destined for structural work.

Frame, Linker, Host, and Map Checks Before You Clone

A tag that is one nucleotide out of frame is not a purification problem, it is a different peptide. Translate the fusion on the map: start codon context, tag, linker, protease site, passenger, and stop. In bacteria, a His tag that replaces the native N-terminus still needs a ribosome-binding site and an ATG. In eukaryotes, Kozak context around the first ATG still applies when the tag is N-terminal.

Linkers are not decorative. A short Gly-Ser linker (for example a GGGGS motif) often lets a small peptide remain accessible. Too short a linker can jam a large fusion against the passenger. Too long a linker can add proteolysis. Annotate the linker as its own feature so a later editor does not treat it as leftover MCS junk.

Host matching still applies. Rare codons, a T7 promoter in a non-DE3 strain, or a mammalian tag cassette dropped into a bacterial plasmid will not be saved by a clever epitope. Sequence visualization, translation view, and plasmid construction in ZettaGene exist to catch those junctions before synthesis. Record the tag, terminus, protease, host, and the small-scale result in ZettaNote so the next construct does not repeat a failed orientation. Keeping the map beside the purification note in a cloud-based R&D workspace is a documentation choice, not a substitute for a solubility test.

FAQ

Which tag should I use if I only need IMAC purification?

A polyhistidine tag is the usual IMAC handle. His6 is enough for many soluble proteins; His10 or a tandem His can raise affinity when a single hexahistidine peptide washes off too easily, at the cost of a slightly larger N- or C-terminal addition. IMAC works under denaturing conditions, which is why His is still used when the protein must be purified from inclusion bodies and refolded. The limitation is specificity: histidine-rich host proteins bind the same metal, so a second polish (ion exchange, size exclusion, or a second epitope) is common. If the protein is insoluble even as a His fusion, switching resin grades will not fix folding. In that case the decision shifts to a solubility partner, with His retained as a capture handle if needed.

When should I use MBP or GST instead of a His tag?

Use MBP or GST when soluble yield is the bottleneck, not when IMAC resin is inconvenient. MBP is the more common solubility engine for difficult bacterial expression. GST can also improve solubility and brings a mature glutathione workflow, but it dimerizes and is a large passenger on activity assays. Neither fusion is a small, invisible handle; plan cleavage if the downstream assay assumes a native protein. Many vectors place His on the MBP or GST fusion so you are not limited to amylose or glutathione capture. If the untagged protein is already soluble and you only need capture, a His, FLAG, or Strep peptide is usually a better size match than a 26 or 40 kDa domain.

Should I put the tag on the N-terminus or the C-terminus?

Put the tag where it remains accessible and where it does not block a signal peptide, active site, or interaction motif. N-terminal tags are the default on many bacterial solubility cassettes and on SUMO fusions that will be cleaved to a defined N-terminus. C-terminal tags are safer when the N-terminus must stay native, when a leader sequence occupies the N-terminus, or when N-terminal fusions have already failed. His, FLAG, and Strep are small enough to try at both ends. Large fusions are harder to move because the vector architecture and the protease geometry were built for one orientation. When both ends are plausible, build both and compare a small expression and capture test rather than arguing from sequence alone.

Do I need a protease cleavage site in a tagged protein construct?

You need a cleavage site when the tag will interfere with activity, structure, oligomeric state, or an assay that assumes a native terminus. You do not need one when the tag is also the detection handle, when the fusion is tolerated, or when you will never remove it. TEV is a common general protease; SUMO protease is the reason many people choose SUMO in the first place, because it can leave a native N-terminus. Factor Xa and thrombin are older options with more off-target caution. Design the scar on purpose: TEV does not restore a fully native N-terminus in most layouts. Also design how you will separate the cleaved tag from the target, or cleavage only changes a band on a gel.

What is the difference between FLAG and Strep tags for purification?

Both are short peptides. FLAG is an epitope tag detected and captured with antibodies; it is strong for immunoprecipitation, Western blots, and some affinity elutions, and it depends on the peptide being exposed. Strep-tag II binds Strep-Tactin or related streptavidin-class resins and is often eluted under mild conditions with a biotin analog, which is useful when antibody elution is too harsh. Strep chemistry can be competed by biotin in rich media or lysates, so growth and buffer details matter. Neither tag is a solubility domain. Choose FLAG when detection and IP are central. Choose Strep when you want a small peptide and a mild, high-purity native elution. Some constructs carry both, with His as a first capture, if each tag has a distinct job.

How do I keep a tagged ORF in frame on the expression plasmid?

Translate the entire open reading frame on the map, from the start codon through tag, linker, protease site, passenger, and stop. A single-base insertion at a restriction site or a Golden Gate scar is enough to produce a nonsense peptide that still looks like a correct plasmid cartoon. Check that N-terminal tags did not overwrite a signal peptide you still need, and that C-terminal tags did not displace the stop codon. Then sequence the actual clone across every junction, because synthesis and PCR introduce errors the map cannot see. In silico translation in ZettaGene is the cheap version of this check; the expensive version is a purification that yields no band after two weeks of culture.

Conclusion

How to choose a tag for protein expression and purification is a job match: capture, solubility, detection, or cleavage, then a chemistry (His, GST, MBP, FLAG, Strep, SUMO), then a terminus and a protease plan that leaves an acceptable scar. Small peptides purify what is already folded; large fusions buy solubility at the cost of bulk and extra steps. Check frame and linkers on the map before you clone, and record which orientation actually captured protein. To review tagged ORFs, termini, and cleavage sites on the plasmid before expression, use ZettaGene in Zettalab's molecular biology tools.