Glen Report 33-13: Application Note – Protein Labeling with NHS Esters

The use of NHS esters is an effective and selective method for labeling primary aliphatic amines. The chemistry works very well for amine-labeled oligonucleotides, as demonstrated in the previous article, but it perhaps works even better for proteins that naturally contain reactive amine residues (courtesy of the N-terminus) and lysine side chains. The addition of biotin, small organic fluorophores or other small molecule labels to proteins is very common, and labeled proteins have been routinely used for applications ranging from immunofluorescence1 to proteomics.2

Although the chemistry is the same, there are notable differences between oligonucleotide and protein NHS ester labeling. On an oligonucleotide, there is typically one single reactive amine and the goal is to achieve quantitative conjugation yield, while for a protein, there are many reactive amines. According to the UniProt Database, lysines make up 5.6 % of the amino acids found in proteins.3 If the average human protein is 500 amino acids in length, that would equate to 28 lysines. Typically, to minimize the possibility of disrupting the protein function or structure, only a fraction of the amines is converted. As such, the location of conjugation is random rather than fixed. A pure single protein, upon labeling, could be converted to a mixture of differentially labeled versions of the same protein. If a specific lysine was absolutely critical to the protein’s function, random lysine labeling ensures that the protein sample as a whole is not significantly affected. 

General Labeling Procedure

  1. Prepare a 5-20 mg/mL protein solution in 0.1 M bicarbonate buffer (pH 8-9) or phosphate buffered saline (PBS).
  2. Dissolve NHS ester in a small/minimal amount of DMF or DMSO.
  3. Add NHS ester solution to protein solution
  4. Agitate the mixture by pipetting and incubate at room temperature for 1-4 hr.
  5. Separate the protein-conjugate from salts and excess label by size exclusion on a Glen Gel-PakTM desalting column or equivalent.

In planning these experiments, the first step is to understand how much NHS ester is required. This will be dependent on the protein, protein concentration and the degree of labeling desired. For the latter, some applications require minimal labeling, whereas others may require multiple labels per protein molecule. For a more predictable outcome, one may choose to perform a series of small-scale labeling experiments prior to scaling up. As with any NHS ester coupling, a non-amine-based buffer needs to be used. This would typically be a bicarbonate buffer (pH 8-9), but for very pH sensitive proteins, PBS, which mimics physiological conditions, might be a better choice. At pH 7.4, the NHS ester coupling reaction is much slower, but NHS ester hydrolysis is slower as well. As a result, reactions in PBS will also work well but will require longer periods of incubation time.

To further illustrate the similarities and differences of labeling between oligonucleotides and proteins, we performed a couple of bovine serum albumin (BSA) labeling experiments with Fluorescein NHS Ester. BSA is one of the most commonly used proteins in science. The mature protein has a length of 583 amino acids, a molecular weight of 66,463 Da and a total of 59 lysine residues. Of these lysines, some are glycosylated or succinylated and others may simply be inaccessible, but there are still many available for conjugation. 6.5 equivalents of FAM NHS ester in DMSO was added to a 10 mg/mL solution of BSA in 0.1 M sodium bicarbonate (pH 9.0). After incubating the reaction for 1 hour at room temperature, the reaction was desalted on a Glen Gel-PakTM column. UV/VIS absorbance analysis of the labeled BSA revealed a degree of labeling of 1.1. The same exact reaction was also performed in PBS. This second labeling experiment was incubated for 4 hours at room temperature and gave a somewhat lower degree of labeling of 0.9. 

For perspective, these BSA labeling reactions were performed at approximately 2.5-fold less concentration than the initial amino oligonucleotide reactions in the previous article, and we know how important reagent concentrations can be in terms of reaction rates. BSA is clearly more readily labeled than 5’-amino oligonucleotide, but BSA does not appear to be anywhere close to 60-fold (59 lysine residues and the N-terminus) more reactive. This is partly due to the fact that lysine side chains have notably higher pKa’s than a typical aliphatic alkyl amine, and partly due to the fact that, as mentioned earlier, some of the lysines may not be accessible.

Traditionally, Glen Research has been providing high quality reagents for the oligonucleotide synthesis community, but as we have discussed here, a subset of the products can also be applied to other areas, such as protein labeling. All the NHS ester and azide (not discussed) labels we offer will work with proteins, and our Glen Gel-Pak columns will desalt proteins just as well as oligonucleotides.  


  1. A.H. Coons, and M.H. Kaplan, J Exp Med, 1950, 91, 1-13.
  2. M. Unlu, M.E. Morgan, and J.S. Minden, Electrophoresis, 1997, 18, 2071-7.
  3. UniProt release statistics 2021_01 results.


Product Information

NHS Esters