How to tell the difference between lysine and the N-terminal amino acid in DNP labeling?

Question

The N-terminal amino acid of a protein can be identified by reacting the N-terminal amine with FDNB (1-fluoro-2,4-dinitrobenzene), or Dansyl chloride, or Dabsyl chloride, then digesting the protein with acid.

This question points out that the epsilon amino group of lysine can also react with the labeling reagents, creating a side-chain labeled lysine. Dinitrophenyl (DNP) derivatives of amino acids (specifically epsilon-DNP-lysine)

But if we're just looking for the labeled amino acid, how do we tell the difference between the N-terminal amino acid and lysine?

I suppose that if we ran the amino acid mixture on HPLC or TLC that the epsilon labeled lysine would have different retention times, and we could learn to ignore those peaks once we knew what lysine looked like.

Answer 1

Actually DNP-labeling of amino groups was first introduced in 1923 when 2,4-dinitrochlorobenzene (DNCB) was used for the identification of the terminal groups of a partial hydrolysate of silk fibroin (Ref.1). Then, Sanger has introduced 1-fluoro-2,4-dinitrobenzene (DNFB) in 1945, which bears his name, the Sanger’s reagent. DNFB was first used to detect free amino acids of insulin (Ref.2). In a nutshell, DNFB undergoes nucleophilic aromatic substitution with the $\ce{N}$-terminal amino group of a peptide or protein. DNP-group is stable to acid hydrolysis. After the DNP-labeled peptide or protein is hydrolyzed, the individual amino acids can be separated and only the DNP-labeled $\ce{N}$-terminal amino acid can be detected by a colorimetric detection at specific wavelength. Hence, DNFB can be used in protein sequencing to determine $\ce{N}$-terminal amino acids.

Not only the $\ce{N}$-terminal amino acid of the protein is labeled by Sanger’s reagent, but also the backbone amino acids with an active group to DNFB such as $\epsilon$-$\ce{NH2}$ in lysine, $\ce{SH}$ in cysteine, $\ce{OH}$ in tyrosine, and $\ce{NH}$ in histidine would be labeled as well. The labeled protein is then hydrolyzed (DNP derivatives are stable to acid hydrolysis), and extracted with ether. If lysine, cysteine, tyrosine, histidine are terminal amino acids, they should be di-DNP-labeled amino acids. If any of them are in the backbone (within $\alpha$-helix), they should have free amino group after hydrolysis. These mono-DNP-labelled amino acids with free amine group are soluble in acid phase and won't come to ether layer when extracted (Ref.2).

The concentrated ether layer is loaded on silica gel chromatography column and the elution profile is matched with the standard profile obtained from individual DNP derivative of all the amino acids. The reagents produce yellow colored derivatives which can be easily detected by absorbance (Ref.2). This method can be applied to any protein (best works with less than 50-70 amino acids). For larger proteins, before DNFB treatment, the protein should be hydrolyzed to small fragments by enzymes such as trypsin or pepsin, or by chemicals like cyanogen bromide.

Note: Extraction and separation methods are detaied in Ref.2, hence I attached the PDF file. For instance, in insulin (MW: ~$\pu{12000 Da}$), there are four terminals (two glycines and two phenylalanines), and two $\epsilon$-DNP-lysines.

References:

1. E. Abderhalden, W. Stix, Hoppe-Seyl. Z. 1923, 129, 143 (Reference 2 of Ref.2).
2. F. Sanger, "The free amino groups of insulin," The Biochemical journal 1945, 39(5), 507–515 (doi:10.1042/j0390507)(PDF).