A Facile Method for Preferential Modification of the N-Terminal Amino Group of Peptides Using Triazine-Based Coupling Reagents

It has been shown that chemical modification of the peptide N-terminus with a charged tag greatly affects the fragmentation process caused by collision-induced dissociation to obtain more interpretable product ion spectra. In this study, we examined the selective introduction of a charged tag, 4-(guanidinomethyl)-benzoic acid (Gmb), into the peptide N-terminus. After optimization of the reaction conditions, we found that the most effective conversion in terms of the reaction rate and selectivity was achieved by reacting the peptide with the active ester of Gmb, prepared using 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpho-linium chloride (DMT-MM) at pH 7. This method is applicable to the introduction of various carboxylic acid-containing compounds into the N-terminus of peptides, which will be useful not only for improvement of MS/MS fragmentation but also for various biochemical studies of peptides and proteins.


INTRODUCTION
Chemical modi cation of peptides and proteins has been utilized for the modulation of their function and development of biotherapeutics. 1) Introduction of various chemical tags also provides many useful applications in the eld of analytical sciences including proteomics. e isotope-coded tags, for example, are essential for the precise quantitation of protein expression by mass spectrometry. 2) We have been interested in the charged tags that a ect the fragmentation process caused by collision-induced dissociation (CID) in tandem mass spectrometers, because it provides more easily interpretable product ion spectra to facilitate the de novo sequencing of peptides. 3) ese modi cations are usually performed at the Nterminal amino group of peptides, because it readily forms an amide bond with carboxyl group-containing compounds. However, due to the slight di erence in reactivity between N-terminal αand Lys ε-amino groups, both types of amino groups are modi ed unless special attention is paid for the reaction conditions. is is particularly disadvantageous in the case of MS-based de novo sequencing, where N-terminal-speci c modi cation is essential to facilitate peptide fragmentation in a more interpretable fashion. 3) To achieve N-terminal-speci c modi cation, various methods have been developed. One of the most commonly employed approaches is the precise control of pH values of the reaction medium, which takes advantage of the slight di erence in pK a values between the αand ε-amino groups. 4) To date, N-hydroxysuccinimide (NHS) esters of carboxylic acids have been widely used for this purpose, because of its ease of preparation and relatively high stability. However, even under judiciously controlled pH conditions, the selectivity is o en unsatisfactory, accompanying a considerable level of modi cation at the Lys side chains. 5) As an alternative approach, utilization of transamination reaction was developed, 6) in which the N-terminal amino acid residue is speci cally converted into α-ketoacid using aldehyde reagents for the subsequent introduction of various components having a hydroxylamine moiety. While this method is demonstrated to be e ective in terms of selectivity, it is incompatible with peptides having the particular amino acid residues, such as Ser, r, Cys, and Trp, at the N-terminus because of their side reactions with aldehydes, limiting the range of applications. us, to overcome this problem, the same research group revised the method by utilizing 2-pyridinecarboxyaldehyde to form an imidazolizinone structure speci cally at the N-terminus through an imine intermedi-ate. 7) Likewise, Chan et al. also reported a method for selective N-terminal modi cation using ketene-containing compounds. 5) Although above-cited methods allow highly selective modi cations of the peptide N-terminus regardless of the amino acid structure, they are disadvantageous in that they require laborious and time-consuming preparation of reagents containing a 2-pyridinecarboxyaldehyde or ketene moiety. Moreover, since ketene is highly reactive, care must be taken to avoid degradation of the reagents during storage. us, a simple and facile method for the N-terminal selective modi cation using commercially available reagents would be of great value.
4 -(4 , 6 -Di met hox y-1, 3, 5 -t r ia zi n-2-yl)-4 -met hylmorpholinium chloride (DMT-MM, Fig. 1) is known to be a powerful condensation reagent for formation of amide bonds even in water as a solvent. 8) is property is suitable for chemical modi cation of peptides and proteins, because water is the best reaction solvent for these molecules in most cases. Although DMT-MM has been widely employed for peptide and protein modi cations, its capability for selective modi cation at the N-terminal amino group of peptides has never been investigated. In this study, we examined the e ectiveness of DMT-MM on N-terminal-speci c introduction of 4-(guanidinomethyl) benzoic acid (Gmb, Fig. 1) as a charged tag, which is known to improve the MS/MS fragmentation of peptides. 9)

EXPERIMENTAL Chemicals
Gmb and nicotinoyloxysuccinimide (Nic-NHS) were synthesized as previously described. 9) DMT-MM was obtained from Kokusan Chemical (Tokyo, Japan). e peptides used in this study were synthesized using Fmoc solid-phase chemistry in our laboratory. Endoproteinase Lys-C (mass spectrometry grade) was purchased from Wako Pure Chemical Industries, Ltd.

Derivatization of model peptides
e Gmb (10 nmol) was mixed with the DMT-MM (10 nmol) in H 2 O (11 µL) for 20 h to obtain the active ester of Gmb (Gmb-DMT). A er lyophilizing the reaction mixture, the peptide solution (1 nmol) in 0.2 M sodium phosphate bu er (pH 7.0, 14 µL) was added to Gmb-DMT (10 nmol). A er reaction for 30 min at room temperature, the progress of the reaction was monitored by HPLC and LC/MS. Conversion rates were calculated by dividing the UV peak area (absorbance at 215 nm) of both mono-and di-modi ed peptides by the total area of unmodi ed, mono-and di-modi ed peptides. e N-terminal selectivity was determined based on the UV peak area of the peptides modi ed at the N-terminus, Lys side chains or both.

Derivatization of enzymatic digests of LaIT1
LaIT1 (40 nmol) was treated with 45 mM DTT in 0.2 M sodium phosphate bu er (pH 7.8, 80 µL) for 30 min at 50°C. en, 100 mM iodoacetic acid in 0.2 M sodium phosphate bu er (pH 7.8, 80 µL) was added to this reaction mixture. A er reaction for 30 min at room temperature, the reagents were removed by a C18 spin column (MonoSpin C18, GL Sciences, Tokyo, Japan). e resultant peptide was subjected to enzymatic digestion with endoproteinase Lys-C at a peptide/enzyme ratio of 50 : 1 (w/w) in 0.2 M sodium phosphate bu er (pH 7.8) for 18 h at 37°C. e peptides were partially puri ed by a C18 spin column. e digested peptides (0.3 nmol each) in 0.2 M phosphate bu er (pH 7.0, 10 µL) were mixed with Gmb-DMT (20 nmol) prepared as described above. A er reaction for 30 min at room temperature, the progress of the reaction was monitored by LC/MS.

HPLC analysis
HPLC analysis was performed using a Vydac Everest C18 column (1×250 mm, Grace, Columbia, MD). e column was eluted at a ow rate of 0.05 mL/min using a mixture of 0.1% TFA in water (solvent A) and 0.1% TFA in acetonitrile (solvent B) with a 20-min linear gradient from 5 to 40% solvent B. e amount of each peptide was estimated based on peak areas of UV chromatograms monitored at 215 nm absorbance.

Mass spectrometric analysis
LC/MS/MS analysis was performed on an LCMS-IT-TOF mass spectrometer (Shimadzu, Kyoto, Japan) equipped with an electrospray ionization probe. e HPLC separations were performed with a TSK-ODS V100 column (1×150 mm, TOSOH, Tokyo, Japan). e column was eluted at a ow rate of 0.05 mL/min using a mixture of 0.1% formic acid in water (solvent C) and 0.1% formic acid in acetonitrile (solvent D) with a 30-min linear gradient from 10 to 50% solvent D. All mass spectra were obtained in the positive mode, and precursor ions for MS/MS analysis were automatically selected based on the ion intensity. e amount of each peptide was estimated based on peak areas in total ion current (TIC) chromatograms.

RESULTS AND DISCUSSION
We previously demonstrated that modi cation of the peptide N-terminus with Gmb is e ective for improving the backbone fragmentation by low-energy CID, 9) as well as for enhancing the side chain fragmentation by high-energy CID due to its high proton a nity. 10) Since Gmb contains a carboxyl group, its introduction can be achieved by formation of an amide bond with the amino group of peptides using condensation reagents. DMT-MM has been widely used for amide bond formation between various compounds, due to its high reactivity superior to other reagents such as carbodiimide or NHS. 11) erefore, we investigated the capability of DMT-MM for N-terminal-speci c introduction of Gmb into peptides.
First, introduction of Gmb into the peptide N-terminus was evaluated using a model peptide I (AAGLAIASAKDD), which contains a Lys residue in the C-terminal region. e peptide was mixed with Gmb in the presence of DMT-MM in the bu er at pH 7.0, but none of modi ed peptides was observed a er a reaction for 30 min (Entry 1 in Table 1, Fig. 2a). A er incubation for another 20 h, formation of the modi ed peptide was observed although its conversion rate was relatively low. Based on the molecular mass measured by LC/MS, this peptide was postulated to be a peptide containing a single Gmb moiety (Entry 2 in Table  1, Fig. 2b). MS/MS analysis of the modi ed peptide (Fig.  3) indicated the existence of b 1 -b 9 ions in the product ion spectrum, which con rmed the Gmb modi cation at the N-terminal amino group. en, to improve the conversion rate, we optimized the reaction conditions. Mechanistically, it is known that DMT-MM forms active esters with carboxylic acids, which subsequently react with the amino group(s) of the peptide to be modi ed. In this regard, the procedure employed in Entries 1 and 2 was considered to be unfavorable, in which formation of the active ester could not proceed e ciently in the presence of the peptide or/and the bu er salts, to lower the overall reaction yield. erefore, we divided the procedure into two steps; rstly, Gmb was reacted with DMT-MM to obtain the active ester (Gmb-DMT) in H 2 O without the peptide, and the reaction mixture was lyophilized to remove H 2 O. Secondly, to this concentrated Gmb-DMT was added the peptide dissolved in the bu er suitable for selective amide bond formation at the N-terminus to complete the reaction. is procedure dramatically improved the conversion rate from 9 to 70%, without a ecting the reaction selectivity (Entry 3 in Table 1, Fig. 2c). e formation of the di-modi ed peptide, in which both the N-terminus and the Lys side chain were modi ed, was also observed, but its proportion was relatively low compared to the N-terminally modi ed peptide.
Next, the e ect of the pH condition on the N-terminalspeci c modi cation was examined. When the reaction was performed at pH 8.0 instead of pH 7.0 using the preformed  Gmb-DMT, the conversion rate was improved from 70 to 83% (Entry 4 in Table 1, Fig. 2d). However, the ratio of the di-modi ed peptide to the N-terminally modi ed peptide increased more than twice as much as that observed at pH 7.0. On the other hand, the reaction at pH 6.0 showed the lower conversion rate (35%) than that observed at pH 7.0, although formation of the di-modi ed peptide was signicantly suppressed (Entry 5 in Table 1, Fig. 2e). Interestingly, the pH values of the reaction media did not a ect the selectivity between the N-terminus and the Lys side chains (Nterminus : Lys=100 : 4-6). is tendency was also observed for the ketene-mediated reactions, 5) in which formation of the di-modi ed peptides increases with increasing conversion rates, but without a ecting the selectivity. is implies that, under the conditions that lead to high conversion rates, the reaction occurs rst at the N-terminal amino group of most of the peptides, followed by the modi cation at the Lys side chains to generate the di-modi ed peptides. Considering the balance between the conversion rates and the ratio of the di-modi ed peptide formation, the reaction at pH 7.0 was judged to be the most practical for the N-terminal selective modi cation using DMT-MM under current conditions. To examine the general applicability of this method, modi cation was performed using a peptide with di erent sequence (peptide II, Entry 6 in Table 1, Fig. S1) or a di er-ent tag (Nic, Entry 7 in Table 1, Fig. S1). Even using the peptide II, its N-terminal amino group was selectively modi ed. Likewise, high N-terminal selectivity was observed using Nic as a tag, suggesting that the method shown in this study is applicable to other carboxyl group-containing compounds. In the case of Nic, the conversion rate was lower (11%) than that using Gmb, but it would be improved by optimizing the reaction conditions, such as pH values. When this result was compared with that obtained by the NHS ester method (Entry 8 in Table 1), the latter showed a much lower conversion rate and lower selectivity. is clearly indicates that the method developed in this study is superior to the NHS ester method. To further evaluate the e ectiveness of this method, the modi cation was carried out for the peptides obtained by enzymatic digestion. e scorpion toxin LaIT1 was used as a model peptides, which contains several Lys residues. 12) A er reduction and alkylation of cysteine residues, LaIT1 were digested with endoproteinase Lys-C, and resultant peptides were subjected to the modi cation. As shown in Table 2, conversion rates were satisfactory for all peptides (78-93%). Preferential introduction of Gmb into the Nterminal amino group was also achieved for all peptides (N-terminus : Lys=100 : 0-14), which were con rmed based on the observation of a series of b-ions in product ion spectra (Fig. S2). Relatively high ratios of di-modi ed peptide formation were observed for peptides 1 and 2, compared to peptides 3 and 4. is situation in peptides 1 and 2 seems similar to that observed for the modi cation of peptide I at pH 8.0. It is possible that the reactions of peptides 1 and 2 proceed faster than those of peptides 3 and 4 even at pH 7.0. e presence of acidic and/or aromatic residues in the N-terminal region in peptides 1 and 2 may a ect the microenvironment of the terminal amino group to increase the reactivity. e use of lower pH values or shorter reaction times could reduce the di-modi ed peptide formation in peptides 1 and 2, although the conversion rates of peptides 3 and 4 would decrease. Furthermore, the di-modi ed peptide  formation does not practically interfere with the following interpretation of the product ion spectra, because they are easily distinguishable from the singly modi ed peptides based on the di erence in the molecular mass. is is totally not the case when the peptides singly modi ed at the Nterminus and at the Lys side chain are mixed to give the ions having the same molecular mass with di erent structures. In consequence, this method is practically useful for obtaining N-terminally modi ed peptides without special preparation of reagents, which will facilitate the de novo sequencing analysis of bioactive peptides.

CONCLUSION
In this study, we demonstrated that a commercially available reagent DMT-MM can be used for N-terminal speci c introduction of a charged tag in a facile manner. Unlike other methods reported to date, the reaction procedure used in this study is relatively simple and requires no special preparation of reagents. is modi cation method is applicable to introduction of various carboxyl group-containing compounds to the N-terminus of peptides and proteins, which will be useful not only for improvement of peptide fragmentation but also for various biochemical studies of peptides and proteins.