|Home | About | Journals | Submit | Contact Us | Français|
Chimeric proteins, including bi-specific antibodies, are biological tools with therapeutic applications. Genetic fusion and ligation methods allow the creation of N-to-C and C-toN fused recombinant proteins, but not the majority of non-template encoded fusions. The present protocol describes a simple procedure for the production of unnaturally linked N-to-N and C-to-C chimeric proteins. Equipping the N-terminus or C-terminus of the proteins of interest with a set of click handles using sortase A, followed by a click reaction, establishes unnatural N-to-N and C-to-C (hetero)dimer linked fusions. If the peptides, sortase A, and the proteins of interest are in hand, the unnaturally fused proteins can be obtained in 3–4 days.
Chimeric proteins are useful research tools as well as interesting therapeutic possibilities. Fusing a protein of interest with a fluorescent protein, such as GFP, allows the study of its intracellular localization and trafficking 1. Fusions of toxins with cytokines or antibodies have been used to yield effective cancer treatments 2.
Genetic fusion of protein-encoding genes of interest is the most common method for the production of chimeric proteins. While the genetic engineering required to produce such fusions is straightforward, the resulting product may not express or fold properly, causing it to lose its function. Sortase A-catalyzed transacylation reactions 3–5 can prevent such pitfalls and covalently link two already correctly folded proteins. In this strategy, the N-terminus of one of the proteins of interest is equipped with an oligoglycine sequence, while the C-terminus of the other protein is engineered with an LPXTG sortase-recognition motif. Sortase A from Staphylococcus aureus cleaves the LPXTG sequence between the threonine and the glycine residues to yield an acyl-enzyme intermediate. The protein with an N-terminal oligoglycine sequence can act as a nucleophile and resolve the acyl-enzyme intermediate to create the fusion protein. (For detailed information on sortase A, the reaction and references see the C-terminal sortase A labeling protocol [ref]). This strategy is site-specific and versatile and has been used to prepare a wide array of protein fusions 6–9. A disadvantage of sortase-mediated reactions and similar conjugation strategies, such as intein chemistry 10, split intein 11, and native chemical ligation 12 is that they exclusively afford N-to-C and C-to-N fused proteins. However, many proteins, including antibodies, require one of their termini to remain unmodified for retention of activity. Thus, a standard genetic fusion of two proteins of this type would diminish or abolish the activity of one of the proteins. A fusion of the same termini (i.e. N-to-N, or C-to-C) might then be preferable. In this protocol we describe a strategy that combines sortase A-catalyzed transacylation with a strain-promoted click reaction, to allow production of such unnaturally linked proteins (Fig 1) 13.
Production of unnaturally linked fusion proteins requires: proteins equipped with a sortase A recognition sequence, sortase A from S. aureus and probes equipped with click handles. The preparation of sortase A from S. aureus is described in “Site-specific C-terminal labeling of proteins using sortase-mediated reactions” [ref]. The design of proteins containing a sortase A recognition sequence is described in detail in “Site-specific C-terminal labeling of proteins using sortase-mediated reactions” and “Site-specific N-terminal labeling of proteins using sortase-mediated reactions” [ref].
Proteins that require their N-terminus for activity must be sortagged via their C-terminus with a click handle. Using standard molecular cloning techniques, the protein of interest is engineered with a C-terminal LPXTG sequence, followed by an optional His tag. The His tag is lost in the course of the sortase reaction and simplifies analysis of the sortase catalyzed transacylation of the click probe and purification of the modified protein. Both sortase A and input substrate can be His-tagged, and removed from the final reaction mixture by adsorption onto Ni-NTA agarose. For fusion proteins that require free C-termini for their function, the click handles are introduced at the N-terminus. For that purpose, one to five glycine residues are engineered at the protein’s N-terminus, using standard molecular cloning techniques.
The accessibility of the N-terminal oligoglycine sequence and the C-terminal LPXTG sequence has a dramatic influence on the reaction rate of the transpeptidation. Since this parameter is different for every protein, we recommend testing the substrates empirically, using fluorescently labeled or biotinylated peptides, prior to performing the reaction with the click handles. Design of a linker between the sortase A sequence (Glyn and/or LPXTG) and the protein may be required to enhance the extent of conversion and reaction rate. In cases where the initiator methionine is not removed by methionyl-aminopeptidase in the course of protein expression, a thrombin cleavage site can be cloned directly upstream of the Glyn nucleophile to ensure a clean and active N-terminus [ref]. The optimized protein substrates and conditions can be used with no further modification in transpeptidation reactions with click-handle containing peptides. After purification of the modified proteins, mixing of the azide-containing protein with the cyclooctyne-containing protein in a suitable buffer is sufficient to initiate dimerization of the proteins.
To prepare N-to-N linked proteins, we synthesize peptides containing the LPXTGG sortase A recognition sequence at their C-terminus (X can be any residue but we prefer a polar residue, such as a glutamic acid) and an azido or a cyclooctyne (DIBAC)14 group at the N-terminus of the probe (Figure 2). While the azido group is stable under basic and acidic conditions and can be introduced in the course of synthesis on a solid support, the DIBAC is labile under strong acidic and alkaline conditions14 and is best introduced after cleaving the peptide from the resin.
Peptides for creating C-to-C linked proteins are synthesized with an N-terminal triglycine motif and an azide or cyclooctyne at the C-terminus (Figure 2). Again, the azido group is introduced on the resin by coupling Fmoc-azidolysine-OH or by modifying a lysine side-chain with azidohexanoic acid. The DIBAC group must be introduced after cleaving the peptide off the resin, using cysteine maleimide chemistry to selectively link the cyclooctyne in the presence of a free N-terminus. A fluorescent dye, such as the TAMRA group in peptide 1, can be incorporated for visualization purposes.
Perform a Kaiser test to monitor the extent of coupling 15.
Protocols for the preparation of sortase A from S. aureus and for the N- and C-terminal labeling of proteins are described in detail in other protocols [ref]. Below follows a description for the N- and C-terminal labeling of proteins with the click handles.
|Incomplete conversion for C-terminal protein modification||Wrong buffer||The buffer pH should be around pH 7.5. If S. aureus Sortase A is used the buffer should contain 1 mM CaCl2. Avoid using phosphate-based buffers when using S. aureus sortase.|
|The ratio between the C-terminal protein-LPXTGG and GGG-label is not optimal.||Increase the amount of nucleophile. The conditions need to be determined empirically.|
|Sortase is degraded||Prepare a fresh batch of Sortase|
|Sortase is inactive||Test sortase on GFP containing a C-terminal LPETGG motif.|
|C-terminus is not adequately exposed||Extend the C-terminus of the protein by introducing a linker (Gly4Ser)n immediately upstream of the sortase motif.|
|Incomplete conversion for N-terminal protein modification||Wrong buffer||The pH of the reaction mixture should be ~ pH 7.5. If S. aureus Sortase A is used, the buffer should contain 1mM CaCl2. Avoid using phosphate-based buffers when using S. aureus sortase.|
|The ratio between the N-terminal protein-LPETGG and GGG-label is not optimal.||Decrease the amount of protein or increase the amount of LPETGG probe. These conditions need to be determined empirically.|
|Sortase is not active||Test activity of sortase with oligo-glycine GFP|
|N-terminus is not exposed||Increase the length of the Glyn N-terminus|
|Incomplete dimerization||Excess probe from the sortase A reaction remains present.||Purify the reaction mixture by size exclusion column chromatography or an alternative method to resolve protein from click handles.|
|Concentration of proteins is not sufficient.||Check concentration of the starting proteins. Concentrations typically used for dimerization are around 10–80 μM.|
|The ratio of azide and cyclooctyne is not correct.||Check the concentrations and make sure azide and cyclooctyne containing proteins are mixed 1:1.|
|The cyclooctyne has been modified.||Make sure that in the course of peptide synthesis, the cyclooctyne is introduced in the final step to prevent ring rearrangements.|
Exclude the cyclooctyne from azides and/or thiols to prevent unwanted side reactions