Electrophiles and Nucleophiles: Basic Bioconjugation Strategies
When I began working at Molecular Probes in technical services, my experience was cell biology-centric. The only chemistry knowledge I had back then came from basic courses in university. But, Molecular Probes was at its core a chemical company, albeit fluorescence chemistry. The reagents might be used on cells, but the chemistry was our business. In tech services, our bible was called Bioconjugate Techniques by Greg T. Hermanson. I still have my torn up, rabbit eared copy sitting next to me on my desk at home. It’s like a safety blanket.
In tech services, I was expected to be able to look at a chemical structure, be able to make educated guesses about the function, the weaknesses, the likely basic applications of the structures of the products we sold. When gathering enough courage to ask for help on a query, inevitably the answer would be ‘look at the structure. It will tell you everything you need.’ Though not helpful in the short term (more like hazing to be accurate), within a year I could assess function and troubleshoot problems from what the chemical structure told me. The most basic principal you needed to master was bioconjugation methods.
At the most basic level, to create a covalent bond, the bonded pair needs to be receptive to the introduction of electrons into their orbital. There are two possibilities, electrophilic or nucleophilic substitution reactions when it comes to donating an electrophile like a protonated hydrogen atom or a nucleophile like the reduced oxygen in a carboxylic acid side chain. Depending on the reaction, one of them needs to donate electrons to the other orbital to obtain a covalent bond. For bioconjugation this involves three basic methods targeting either primary amine, thiol, or carboxylic acid side chains.
Primary Amine Conjugation Chemistry
The most common method most people have experience with is the conjugation of an NHS-ester (N-hydroxysuccinimide) or succinimidyl ester to a primary amine residue (-NH2). The popularity is due to the fact that on average, -NH2 containing lysine amino acids comprise about 7% of all protein or antibody amino acids. Also, every protein or peptide structurally has a terminal -NH2 group. When the succinimidyl ester is >95% purity, this method of conjugating antibodies or middle range molecular weight proteins (>30kD) can be >90% efficient within 30 minutes under the right conditions. The biggest threats to the efficiency and yield of this conjugation method is oxygen, temperature, pH and other amines to compete with the conjugation site.
The most difficult condition to control is protection of the NHS ester from oxygen and temperature. This is why the reagent is resuspended in DMSO or DMF that is >99% anhydrous. We’d routinely have people call tech services that had resuspended their reagent in DMSO that was left on the lab bench intended for freezing down cells.
Needless to say, it was no longer sufficiently anhydrous and you could see the resultant hydrolysis of the NHS ester represented in the low degree of labeling of the protein. It is also always best to buy reactive dye in smaller aliquots in order to be able to freeze them down in their powder form. Often, reactive dye is also packaged with a layer of argon gas, which displaces the lighter oxygen from the packaging vial and thus prolongs the reactive lifetime of the reagent. These protective methods are relevant for any reagent containing a succinimidyl ester, including common reagents like CFDA-SE (CFSE) or Fixable Live/Dead reagents.
On the reaction side, to ensure high efficiency, the reaction buffer needs to be adjusted to a lower pH, ideally around 7.2 using sodium bicarbonate or sodium borohydride and needs to be free of any additional primary amines. However, whatever buffer you use should not be so weak as to significantly dilute the concentration of the antibody solution. If this isn’t possible, you’ll have to do a buffer exchange like dialysis to obtain the optimal concentration for efficient conjugation. If the solution of unconjugated antibody contains serum, BSA or Tris buffer, all of these will compete with the antibody for the reactive dye. This is why most pure antibodies will be sold free of BSA as a packaging stabilizer.
Isothiocyante groups can also directly conjugate primary amines, however the reaction is much slower, less efficient than succinimidyl esters and it’s been suggested that the covalent bond isn’t as stable over time. However, this reactive group is how we got the colloquial names of FITC and TRITC, which are fluorescein isothiocyanate and tetramethylrhodamine isothiocyanate respectively. People using the term FITC in flow cytometry always left me a little perplexed since the isothiocyanate (-N=C=S) is no longer intact after conjugation.
However, all things considered the primary amine is your most reliable, most abundant and generally your best shot at achieving the conjugation or cross-linking required, even if it is completely indiscriminate to any primary amine side chain comprising the protein or antibody.
The lysine residue and a terminal amine do have different likelihoods of achieving conjugation depending on the pH and microenvironment of the residue. Lysine has a pKa value of 10.5 whereas a terminal amine residue has a pKa value of 8. What this means is that first of all, neither is a particularly strong base and you can control whether the two amine groups on a lysine residue are individually protonated or deprotonated by controlling the pH of the reaction. Needless to say for our purposes, at physiological pH and at the ideal pH (7.2) to conjugate a protein or antibody, both side chains are protonated and equally available for conjugating a succinimidyl ester reactive group.
Sulfhydrl and Disulfide Bond (Thiol) Conjugation Chemistry
When you have a fluorophore that is REALLY big, like the KIRAVIA or the Horizon Brilliant polymerized fluors, the randomness of conjugating primary amines can be undesirable. The risk of conjugating too close to the antigen recognition site or changing the folding and function of a protein can be too much. Instead, a more reliable, predictable target can be a cysteine, thiol-containing side chain or the disulfide bond of the antibody hinge region. To directly conjugate thiols, you employ a maleimide reactive chemistry.
Maleimide reactive groups are more stable to oxygen and temperature than succinimidyl esters, however, they also don’t have quite the level of reaction efficiency or speed. But that can be helpful if you want to limit the number of fluors you are conjugating to a biomolecule. OR, you can thiolate an alcohol or carboxylate group in order to cross-link it to a primary amine.
Carboxylate (and Alcohol) Conjugation Chemistry
Which leads us to the more complicated conjugation strategies common for biomolecules, the carboxylate conjugation chemistry. Just like a primary amine will always be one terminal end of a peptide or protein, you can always be sure a carboxylate group will comprise the other end. This can be very, very helpful for conjugating small peptides especially, where the size of the peptide only will allow for a single fluor anyway and the application might be to track the localization or uptake of that peptide. This reaction is called an EDAC or EDC reaction. There are two possibilities here. The first would be to use a zero-length cross linker called a carbodiimide which facilitates the cross linking without leaving any part of itself remaining in the final product. This is a great way to label the alcohol residues of polysaccharides and the terminal carboxylate of fatty acid chains. The first step of this reaction is to activate the carboxyl group. Although this occurs best the more acidic the reaction buffer, a pH around 4 may not be ideal for the protein or biomolecule you want to conjugate. The higher the pH, the less efficient the reaction albeit necessary for an active end product to result. The low pH creates an intermediate state with the carbodiimide that is very susceptible to a strong nucleophile, like a primary amine. I would call this kind of chemistry a last resort for an atypical chemical structure or again, more selective labeling of only the terminal carboxylate.
Hetero-bifunctional Crosslinking
To bring all three of these common conjugation methods together, the most difficult of all the bioconjugations you might do is a large protein to another large protein, for example a phycoerythrin fluorescent protein that is 240kD to an antibody that is 150kD. Because of the size of the phycoerythrin, you may want to localize the fluor to the hinge region of the antibody (thiol-mediated) to reduce the risk of losing binding affinity of the antibody. In this instance the goal is to activate the primary amines on the PE molecule with an SMCC cross linker.
SMCC cross linkers are hetero-bifunctional in that one side (the NHS ester) of the cross linker will conjugate the amine-containing protein (PE in this case) and then conjugate directly to the reduced disulfide bond of the antibody via the maleimide end of the SMCC crosslinker. It does not leave much of a bridge between the two conjugation sites. This method, with the introduction of a linker, can also be useful for conjugating quantum dots to biomolecules.
The second way to conjugate a carboxylate residue involves a two step process involving the SMCC as well. If you wanted to conjugate a primary amine residue of the antibody and not target the disulfide, you would have to employ this method. First step is to convert the carboxylate residues of the PE protein to thiol groups (thiolate the protein). You can do this with a linker called SPDP. The reverse reaction with SMCC is now possible, the amines of the antibody linked to the newly converted sulfhydrls of the PE molecule.
I won’t go into Click Chemistry or other engineered site-specific chemistry here. What I want to show how conserved reactive chemistry is for biomolecules. You only have so many side chains that are accessible, useful nucleophiles or electrophiles at biologically reasonable pH. Some combination of amines, carboxylates or thiols will usually enable a solution to the strangest peptide, saccharide, IgG, IgM, lipid or even small organic molecule to which you want to attach a fluorophore.