Separation and Identification of Isoleucine Enantiomers and Diastereomers Using an Original Chiral Resolution Labeling Reagent

Makoto Ozaki; Motoshi Shimotsuma; Takefumi Kuranaga; Hideaki Kakeya; Tsunehisa Hirose

doi:10.1248/cpb.c23-00439

Abstract

D-Amino acids, which are present in small amounts in living organisms, are responsible for a variety of physiological functions. Some bioactive/biomolecular peptides also contain D-amino acids in their sequences; such peptides express different functions than peptides composed only of L-form amino acids. Among the 20 amino acids that make up proteins, threonine (Thr) and isoleucine (Ile) have two chiral carbons and thus have two enantiomers and diastereomers. These stereoisomers have been previously analyzed through HPLC using chiral columns or chiral resolution labeling reagents. However, the separation and identification of these stereoisomers are highly laborious and complicated. Herein, we propose an analytical method for the separation and identification of Ile stereoisomers through LC–MS using our original chiral resolution labeling reagent, 1-fluoro-2,4-dinitrophenyl-5-L-valine-N,N-dimethylethylenediamine-amide (L-FDVDA) and a PBr column packed with pentabromobenzyl-modified silica gel. Twenty DL-amino acids including Thr stereoisomers (41 amino acids including glycine) were separated and identified using C₁₈ column. Ile stereoisomers could be separated using not a C₁₈ column but a PBr column. Additionally, we showed that peptides containing Thr and Ile stereoisomers can be accurately detected through labeling with L-FDVDA.

Introduction

Amino acids are essential for protein synthesis and energy production in all organisms. Remarkable advances in analytical technology have revealed that organisms contain not only L-amino acids but also a variety of D-amino acids to perform specific functions.^1–4) D-Amino acids have also been detected in disease-causing and bioactive/biomolecular peptides.^5–8) Among the 20 amino acids that are the building blocks of proteins, threonine (Thr) and isoleucine (Ile) possess two chiral centers on α-carbon and β-carbon, and thus have two enantiomers and two diastereomers (D, L-Thr, D, L-allo-Thr, D, L-Ile, and D, L-allo-Ile, Fig. 1). DL-Amino acids can be separated and identified through HPLC and LC–MS using chiral columns and labeling with 4-(N,N-dimethylaminosulfonyl)-7-(3-isothiocyanatopyrrolidin-1-yl)-2,1,3-benzoxadiazole (DBD-PyNCS), N-(4-nitrophenoxycarbonyl)-L-phenylalanine 2-methoxyethyl ester (NIFE), and Marfey’s reagents such as Nα-(5-fluoro-2,4-dinitrophenyl)-L-alaninamide (L-FDAA) and Nα-(5-fluoro-2,4-dinitrophenyl)-L-leucinamide (L-FDLA).^9–15) Additionally, highly sensitive analytical techniques were developed for the detection of DL-amino acids by reversed phase (RP)–LC after labeling with two reagents, O-phthalaldehyde (OPA) and N-acetyl-L-cysteine, or by two dimensional (2D)–HPLC combining an RP–column and a chiral column after labeling with achiral labeling reagent 4-fluoro-7-nitro-2,1,3-benzoxadiazole (NBD-F).^16–19) However, the separation and identification of DL-Ile and DL-allo-Ile are complicated, and has not been achieved yet. The DL-allo-Ile is present in some bioactive peptides that show antibacterial activity and cancer cell growth inhibition.^8,20–23)Allo-Ile is also present as a free amino acid in vivo, and L-allo-Ile concentration in plasma is 1.9 ± 0.6 µM in healthy adults and 1.6 ± 0.4 µM in children aged 3–11 years.²⁴⁾ Free allo-Ile is involved in the induction of DNA damage.²⁵⁾ Schadewaldt et al. have reported that in patients with maple syrup urine disease, the concentration of L-allo-Ile in plasma is >5 µM.²⁴⁾ Therefore, DL-allo-Ile is a potential biomarker for the maple syrup urine disease. For the above reasons, a practical method must be developed for the separation of DL-Ile, DL-allo-Ile, and DL-Leu with the same molecular weights using HPLC and LC–MS.

Fig. 1. Chemical Structure of DL-Ile and DL-allo-Ile

We previously developed an original chiral resolution labeling reagent for the separation and highly sensitive detection of DL-amino acids.^26–28) Additionally, through the use of our labeling reagent, 1-fluoro-2,4-dinitrophenyl-5-D-leucine-N,N-dimethylethylenediamine-amide (D-FDLDA), we have succeeded in separating and identifying substances such as peptide fragments of amyloid β containing racemized and isomerized amino acids that are difficult to separate using conventional chiral resolution labeling reagents such as Marfey’s reagents.^29,30) Herein, we propose a method for the separation and identification of DL-Ile and DL-allo-Ile using HPLC with our original chiral resolution labeling reagent, 1-fluoro-2,4-dinitrophenyl-5-L-valine-N,N-dimethylethylenediamine-amide (L-FDVDA) (Fig. 2a). Additionally, we compared the chiral resolutions of Ile stereoisomers obtained through labeling with our reagent and through labeling with conventional reagents such as L-FDLA and Nα-(2,4-dinitro-5-fluorophenyl)-L-valinamide (L-FDVA). Furthermore, we established a method for the separation and identification of Thr and Ile stereoisomers in the peptides by labeling with L-FDVDA after hydrolysis.

Fig. 2. Chemical Structure of Labeling Reagent and Labeling Scheme

(a) Chemical structure of L-FDVDA. (b) Schematic of the labeling protocol for amino acids using L-FDVDA.

Experimental

Materials

DL-Alanine (Ala), DL-arginine (Arg), DL-asparagine (Asn), DL-aspartic acid (Asp), DL-cysteine (Cys), DL-glutamic acid (Glu), glycine (Gly), DL-histidine (His), DL-isoleucine, DL-leucine (Leu), DL-phenylalanine (Phe), DL-proline (Pro), DL-serine (Ser), DL-threonine, DL-tryptophan (Trp), DL-tyrosine (Tyr), DL-valine (Val), DL-lysine (Lyn), DL-methionine (Met), DL-glutamine (Gln), DL-amino acid labeling kit, methanol (HPLC grade), acetonitrile (HPLC grade), formic acid, m-cresol, trifluoroacetic acid (TFA), and diethyl ether were obtained from Nacalai Tesque, Inc. (Kyoto, Japan). D-allo-Isoleucine, L-allo-isoleucine, D-allo-threonine, L-allo-threonine, L-FDLA, N,N-diisopropylethylamine, and thioanisole were purchased from Tokyo Chemical Industries (Tokyo, Japan). L-FDVA was purchased from Merck (Darmstadt, Germany). 1-[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate (HATU) was purchased from Oakwood Products, Inc. (Estill, SC, U.S.A.).

Labeling Protocol Using L-FDVDA

The DL-amino acid labeling reaction was performed using solutions (start solution, side chain delabeling reagent solution, and stop solution) from the DL-amino acid labeling kit (Nacalai Tesque). The labeling method for protocol 1 is as follows. Labeling solution {100 µL L-FDVDA solution, enantiomeric excess (ee); >99.9%, Supplementary Fig. S1 and Table S1} and 100 µL of the start solution (labeling initiator) were added to 100 µL of the sample solution (0.04–1.0 mg/mL), and the mixture was incubated for 2 h at 50 °C. Next, 100 µL of the side chain delabeling reagent solution containing 6-mercapto-1-hexanol was added to the labeled sample solution, mixed using a vortex mixer for 5 s, and then reacted at 50 °C for 15 min. Stop solution (reaction stopping reagent, 100 µL) and acetonitrile (500 µL) were added to the sample solution, and the obtained mixture was analyzed using HPLC (LC–UV) and LC–MS. The labeling method in protocol 2 is the same as that in protocol 1 before the delabeling reaction. After the labeling reaction, 100 µL of stop solution and acetonitrile (600 µL) were added to the sample solution.

HPLC Measurements

HPLC analysis was performed using an LC2060-C HPLC system (Shimadzu, Kyoto, Japan). Separations were conducted on COSMOSIL 3C₁₈-AR-II (3.0 mm I.D. × 150 mm, Nacalai Tesque, particle size; 3 µm), COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, Nacalai Tesque, particle size; 3 µm), COSMOSIL 3πNAP (3.0 mm I.D. × 150 mm, Nacalai Tesque, particle size; 3 µm), and COSMOSIL CHiRAL 3C (4.6 mm I.D. × 250 mm, Nacalai Tesque, particle size; 3 µm) at 30 or 40 °C. The optical purity of the labeling reagent was analyzed using triethylamine/ethanol/n-hexane (0.1/25/75, vol/vol/vol) as the mobile phase in isocratic elution mode at a flow rate of 1.0 mL/min with UV detection at 340 nm. In the analysis of DL-amino acids, acetonitrile containing 0.1% formic acid or methanol containing 0.1% formic acid was used as mobile phase in linear gradient elution mode at a flow rate of 0.4 mL/min with UV detection at 340 nm. The resolutions (Rs) of peaks A and B were calculated as follows; Rs = 1.18 × (t_r of peak B − t_r of peak A)/(W_1/2 of peak A + W_1/2 of peak B), where t_r is the retention time, and W_1/2 is the half width.

LC–MS Measurements

LC–MS analysis was performed using a Nexera Lite HPLC system (Shimadzu) and LCMS-2050 mass spectrometer (Shimadzu). The 39 amino acids labeled with L-FDVDA, excluding the Ile stereoisomers, were separated using a COSMOSIL 3C₁₈-AR-II column (3.0 mm I.D. × 150 mm, particle size; 3 µm) with 10% acetonitrile in ultrapure water (containing 0.1% formic acid) as solvent A using a gradient from 10 to 100% (0–60 min) and 50% acetonitrile in ultrapure water (containing 0.1% formic acid) as solvent B at a flow rate 0.4 mL/min at 40 °C. The Ile stereoisomers labeled with L-FDVDA were separated using a COSMOSIL 3PBr column (3.0 mm I.D. × 150 mm, particle size; 3 µm) with 70% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a gradient from 0 to 0 to 30% (0–10–30 min) and 100% methanol (containing 0.1% formic acid) as solvent B at a flow rate 0.4 mL/min at 40 °C. In the LC–MS measurements, the nebulizing gas flow was set to 0.2 L/min, the drying gas flow was set to 5.0 L/min, the heating gas flow was set to 7.0 L/min, and the desolvation temperature was set to 450 °C. The m/z range of 100–1000 was covered with a scan time of 0.5 s; the data were collected in the positive ion mode at a detector voltage of 1.0 kV. Amino acids were detected at m/z 507.5 (DL-His), m/z 526.6 (DL-arginine (DL-Arg)), m/z 484.5 (DL-Asn), m/z 471.5 (DL-Thr and DL-allo-Thr), m/z 457.5 (DL-Ser), m/z 498.5 (DL-Gln), m/z 485.5 (DL-Asp), m/z 499.5 (DL-Glu), m/z 441.5 (DL-Ala), m/z 467.5 (DL-Pro), m/z 427.4 (Gly), m/z 473.5 (DL-Cys), m/z 533.6 (DL-Tyr), m/z 469.5 (DL-Val), m/z 501.6 (DL-Met), m/z 483.5 (DL-Ile, DL-allo-Ile, and DL-Leu), m/z 556.6 (DL-Trp), m/z 517.6 (DL-Phe), and m/z 425.5 (DL-Lys, divalent ion peak).

Peptide Synthesis

The designed peptides were synthesized manually on 9-fluorenylmethyloxycarbonyl (Fmoc)-NH-SAL-PEG resin (Watanabe Chemical Industries, Hiroshima, Japan) using Fmoc chemistry³¹⁾ with Fmoc-AA-OH (4 equivalents (equiv.)) according to the HATU method. The side chain protecting groups used were 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl (Pbf) for Arg, trityl (Trt) for Cys, t-butyloxycarbonyl (Boc) for Lys. The peptides were cleaved from the resins, and side chain protection was removed by incubating the peptides for 1 h in m-cresol/ethanedithiol/thioanisole/TFA (1/3/3/40, vol/vol/vol/vol). The peptides were then precipitated by adding cold diethyl ether and collected through centrifugation. The peptides were dissolved in ultrapure water.

Amino Acid Analysis of Hydrolyzed Peptides

Synthesized peptides (20 µL) were added to a microtube and hydrolyzed in 6 M HCl at 110 °C for 15 h. Ethanol/ultrapure water/triethylamine (2/2/1 = vol/vol/vol, 20 µL) was added to the hydrolyzed samples and dried in vacuo. Ultrapure water (100 µL) was added to each dried sample, and the sample was labeled with L-FDVDA. Amino acids were analyzed by LC–MS using COSMOSIL 3C₁₈-AR-II (3.0 mm I.D. × 150 mm, particle size; 3 µm) and COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm) columns.

Results and Discussion

L-FDVDA reacts with amino groups and also with some side chains, such as the phenolic hydroxy group of tyrosine (Tyr), the thiol group of cysteine (Cys), and the imidazole group of histidine (His). Figure 2b shows the scheme of labeling reaction for amino acids using L-FDVDA. First, the labeling reagent solution and starting solution (reaction initiator) were added to the sample solution containing the amino acids; then, L-FDVDA was bound to the amino acids through incubation for 2 h at 50 °C. Amino acids with a functional group in the side chain, such as Tyr, Cys, and His, were additionally labeled on the side chain. L-FDVDA bound to the side chain was desorbed through the addition of the delabeling reagent solution followed by incubation for 15 min at 50 °C (protocol 1 in Fig. 2b). Amino acids whose side chains reacted with L-FDVDA were immediately analyzed by adding the stop solution without delabeling (protocol 2 in Fig. 2b).²⁹⁾ However, L-FDVDA has a low labeling ability on the imidazole group of His, and when labeled with protocol 2, peaks derived from mono-form (with one labeling reagent) and di-form (with two labeling reagents) were detected. Therefore, when all proteinogenic amino acids are analyzed simultaneously, labeling with protocol 1 should be used.

We confirmed ee of L-FDVDA through HPLC. We first measured the retention factor (k), enantioselectivity factor (α), and resolution (Rs) of L-FDVDA. The retention factors for D, L-FDVDA were 1.80 (D-FDVDA) and 2.25 (L-FDVDA), the enantioselectivity factor was 1.25, and the resolution was 4.11 (Supplementary Table S1). These results indicate that the ee of L-FDVDA can be accurately measured. The ee of L-FDVDA was calculated as follows: (peak area of L-FDVDA − peak area of D-FDVDA)/(peak area of D-FDVDA + peak area of L-FDVDA) × 100%. The ee of L-FDVDA was confirmed by HPLC measurements using a chiral column to be >99.9% (Supplementary Figs. S1a, b and Table S1).

Subsequently, we confirmed that the DL-amino acids labeled with L-FDVDA using protocol 1 (Fig. 2b), which was established in a previous study, could be separated.²⁹⁾ LC–MS was performed with acetonitrile containing 0.1% formic acid (solvent A: 10% acetonitrile in ultrapure water containing 0.1% formic acid; solvent B: 50% acetonitrile in ultrapure water containing 0.1% formic acid) as the mobile phase. DL-Forms of 19 amino acids except for glycine (Gly) were separated and identified using labeling with L-FDVDA followed by LC–MS (Fig. 3 and Supplementary Table S2). Several studies have reported that lysine (Lys), glutamic acid (Glu), and the DL-form of proline (Pro), which contains a secondary amine, are difficult to separate using a chiral column.¹⁴⁾ In contrast, DL-amino acids labeled with L-FDVDA were readily separated even using C₁₈ columns. Furthermore, DL-Thr and DL-allo-Thr labeled with L-FDVDA were separated under the same conditions using also a C₁₈ column (Fig. 3 and Supplementary Table S2). However, DL-Ile and DL-allo-Ile labeled with L-FDVDA could not be separated (Supplementary Fig. S2). Furthermore, changing the mobile phase from acetonitrile to methanol (containing 0.1% formic acid) did not improve the separation performance of DL-Ile and DL-allo-Ile labeled with L-FDVDA (Supplementary Fig. S3). These results show that the DL-forms of amino acids labeled with L-FDVDA, except for DL-Ile and DL-allo-Ile, were separated and identified on a C₁₈ column.

Fig. 3. LC–MS Chromatograms of DL-Amino Acids Labeled with L-FDVDA Using Protocol 1

LC–MS was performed using a COSMOSIL 3C₁₈-AR-II column (3.0 mm I.D. × 150 mm, particle size; 3 µm) with 10% acetonitrile in ultrapure water (containing 0.1% formic acid) as solvent A using a gradient from 10 to 70% (0–60 min) with 50% acetonitrile in ultrapure water (containing 0.1% formic acid) as solvent B at a flow rate of 0.4 mL/min at 40 °C.

We aimed to separate DL-Ile, DL-allo-Ile, and DL-Leu using columns with multiple interactions, such as dispersion force and π–π stacking, instead of a C₁₈ column that depends on the strength of hydrophobic interactions. Therefore, we compared the separation patterns of DL-Ile, DL-allo-Ile, and DL-Leu using a PBr column packed with pentabromobenzyl-modified silica gel and a πNAP column packed with naphthylethyl-modified silica gel^32–34) (Fig. 4a). Methanol (containing 0.1% formic acid) was used as the mobile phase for the PBr and πNAP columns to prevent the interaction between π-electrons in acetonitrile and the stationary phases. Although the C₁₈ column could not separate and identify the Ile stereoisomers labeled with L-FDVDA (Fig. 4b and Table 1), six peaks of the labeled Ile stereoisomers containing DL-Leu were separated and identified using the PBr column (Fig. 4c). However, the Rs value for L-allo-Ile (peak 1) and L-Ile (peak 2) labeled with L-FDVDA analyzed using the PBr column is low (Table 1). Ile stereoisomers containing DL-Leu were well separated and identified if the labeling reagent was appropriately selected between the L-form (L-FDVDA) and the D-form (D-FDVDA) based on the target Ile stereoisomers (Figs. 5b, c and Table 2). In contrast, L-Ile and L-Leu labeled with L-FDVDA were not separated using a πNAP column (Fig. 4d and Table 1). These results indicate that the PBr column can be used to completely separate and identify Ile stereoisomers labeled with D- and L-FDVDA.

Fig. 4. Structures of the Stationary Phase of Various Columns and HPLC Chromatograms of Ile Stereoisomers Including DL-Leu Labeled with L-FDVDA Using Various Columns

(a) Structures of the stationary phase of C₁₈, PBr, and πNAP columns. HPLC chromatograms of DL-allo-Ile, DL-Ile, and DL-Leu labeled with L-FDVDA using protocol 2. HPLC was performed using (b) COSMOSIL 3C₁₈-AR-II (3.0 mm I.D. × 150 mm, particle size; 3 µm), (c) COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm), and (d) COSMOSIL 3πNAP (3.0 mm I.D. × 150 mm, particle size; 3 µm) columns with (b) 20% acetonitrile in ultrapure water (containing 0.1% formic acid), (c) 70% methanol in ultrapure water (containing 0.1% formic acid), and (d) 65% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a linear gradient (b) from 10 to 60% (0–30 min), (c) from 0 to 0 to 30% (0–10–30 min), and (d) from 0 to 0 to 10% (0–15–30 min) with (b) 50% acetonitrile (containing 0.1% formic acid) and (c, d) 100% methanol (containing 0.1% formic acid) as solvent B for 30 min at a flow rate of 0.4 mL/min at 40 °C.

Table 1. Resolution (Rs) of DL-allo-Ile, DL-Ile, and DL-Leu Labeled with L-FDVDA Obtained via HPLC

Fig. 5. Chemical Structures of Labeling Reagents and HPLC Chromatograms of Ile Stereoisomers Including DL-Leu Labeled with Various Labeling Reagents

(a) Chemical structures of D-FDVDA, D-FDLDA, L-FDVA, L-FDLA. HPLC chromatograms of DL-allo-Ile, DL-Ile, and DL-Leu labeled with (b) L-FDVDA, (c) D-FDVDA, (d) D-FDLDA, (e) L-FDVA, and (f) L-FDLA using protocol 2. HPLC was performed using a COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm) column with 70% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a linear gradient (b) and (c) from 0 to 0 to 30% (0–10–30 min), (d) from 5 to 5 to 45% (0–10–30 min), (e) from 10 to 20% (0–30 min), and (f) from 20 to 30% (0–30 min) with 100% methanol (containing 0.1% formic acid) as solvent B for 30 min at a flow rate of 0.4 mL/min at 40 °C.

Table 2. Resolution (Rs) of DL-allo-Ile, DL-Ile, and DL-Leu Labeled with Various Chiral Resolution Labeling Reagents Obtained via HPLC Using a PBr Column

Next, we compared the separation patterns of Ile stereoisomers labeled with other conventional chiral resolution labeling reagents (D-FDLDA, L-FDVA, and L-FDLA; Fig. 5a) using a PBr column. Samples labeled with our previously developed D-FDLDA show separation patterns similar to that of the samples labeled with L-FDVDA. The elution order of the peaks of D- and L-forms of Ile, allo-Ile, and Leu in the samples labeled with D-FDLDA are reversed compared to those labeled with L-FDVDA because of the different chirality of the labeling reagent itself. The separation performance for the L-allo-Ile, L-Ile, D-allo-Ile, and D-Ile using D-FDLDA is not better than that using L-FDVDA (Figs. 5b, d and Table 2). Samples labeled with L-FDVA and L-FDLA show separate DL-Ile and DL-allo-Ile peaks (Figs. 5e, f). However, the peaks of labeled DL-Ile and DL-Leu completely overlap. These results indicate that all Ile stereoisomers containing DL-Leu can be separated using only L-FDVDA with the PBr column.

To quantitatively analyze the concentration of Ile stereoisomers derived from free amino acids in plasma and foods and accurately quantify the Ile stereoisomers in proteins and peptides, the labeling reagent and each amino acid must be completely bound. Therefore, the peak areas of labeled Ile stereoisomers at each time point were plotted, and changes in the binding efficiency of the labeling reagent were observed for each Ile stereoisomer. Reactivity of all Ile stereoisomers containing DL-Leu and the labeling reagents plateau upon incubation for >1 h at 50 °C (Fig. 6). Next, we constructed calibration curves based on peak areas of Ile stereoisomers labeled with L-FDVDA at each concentration and confirmed that each Ile stereoisomer can be measured quantitatively (Figs. 7a–d). Furthermore, we investigated the occurrence of racemization during the labeling reaction. When 1 mM L-Ile alone was labeled with L-FDVDA, only the L-Ile peak was detected (Fig. 8a). In contrast, when 1 mM L-Ile was added to 1 µM D-Ile (amounts of 1/1000 of D-Ile relative to L-Ile) and labeled with L-FDVDA, both L-Ile and D-Ile peaks were identified (Fig. 8b). These results suggest that the racemization of amino acids did not occur during the labeling of amino acids. The proposed labeling reagents are indicated to guarantee a quantitative analysis of racemization and epimerization.

Fig. 6. Reactivity of Labeling Reagent and Ile Stereoisomers Including DL-Leu

Changes in the amount of L-FDVDA bound to DL-allo-Ile, DL-Ile, and DL-Leu. HPLC was performed using a COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm) column with 70% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a linear gradient from 0 to 0 to 30% (0–10–30 min) and 100% methanol (containing 0.1% formic acid) as solvent B for 30 min at a flow rate of 0.4 mL/min at 40 °C. UV detection was performed at 340 nm.

Fig. 7. Calibration Curves for (a) L-allo-Ile, (b) D-allo-Ile, (c) L-Ile, and (d) D-Ile Labeled with L-FDVDA

LC–MS was performed using a COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm) column with 70% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a linear gradient from 0 to 0 to 30% (0–10–30 min) and 100% methanol (containing 0.1% formic acid) as solvent B for 30 min at a flow rate of 0.4 mL/min at 40 °C.

Fig. 8. LC–MS Chromatograms of (a) 1 mM L-Ile and (b) Mixture of 1 mM L-Ile and 1 µM D-Ile That Were Labeled with L-FDVDA

Finally, we attempted to separate and identify peptides containing Thr and Ile stereoisomers using C₁₈ and PBr columns by labeling with L-FDVDA. We studied the fragment peptides containing Thr and Ile stereoisomers of the ShK toxin, which binds to the Kv1.3 ion channel protein with high affinity, as model peptides for confirming sequence analysis²¹⁾ (Table 3). The activity of this peptide is highly dependent on whether Thr and Ile in the sequence are in DL- or DL-allo-forms. As residues 5 and 12 of the toxin are Thr and residues 3 and 6 are Ile, four peptides with Thr and Ile at residues 5 and 3, respectively, in different combinations of D- and/or D-allo-forms were used to demonstrate sequence analysis (Table 3). Synthesized peptides were hydrolyzed to amino acids through the reaction with 6 M HCl at 110 °C overnight. The amino acids were then labeled with L-FDVDA and analyzed through LC–MS using C₁₈ and PBr columns. All amino acids except for Ile stereoisomers in the hydrolyzed peptides were separated and identified using a C₁₈ column and L-FDVDA labeling (Supplementary Figs. S4a–d). As shown in Figs. 9a and 9b, Thr and Ile stereoisomers were precisely separated and identified using C₁₈ and PBr columns, respectively. These results indicate that all DL-amino acids, even those containing Thr and Ile stereoisomers in the peptide chain, were separated and identified using C₁₈ and PBr columns after labeling with L-FDVDA.

Table 3. Fragment Peptides Containing Thr and Ile Stereoisomers of the Toxin

Fig. 9. LC–MS Chromatograms of Hydrolyzed Peptides Containing D-allo-Thr, DL-Thr, D-allo-Ile, and DL-Ile Labeled with L-FDVDA Using Protocol 1

LC–MS was performed using (a) COSMOSIL 3C₁₈-AR-II (3.0 mm I.D. × 150 mm, particle size; 3 µm) and (b) COSMOSIL 3PBr (3.0 mm I.D. × 150 mm, particle size; 3 µm) columns with (a) 10% acetonitrile in ultrapure water (containing 0.1% formic acid) and (b) 70% methanol in ultrapure water (containing 0.1% formic acid) as solvent A using a linear gradient (a) from 10 to 70% (0–60 min) and (b) from 0 to 0 to 30% (0–10–30 min) with (a) 50% acetonitrile in ultrapure water (containing 0.1% formic acid) and (b) 100% methanol (containing 0.1% formic acid) as solvent B for (a) 60 min and (b) 30 min at a flow rate of 0.4 mL/min and 40 °C. * Compound different from DL-Thr and DL-allo-Thr.

Conclusion

We developed a method for the separation and identification of proteinogenic amino acids containing Thr and Ile stereoisomers using C₁₈ and PBr columns and labeling with L-FDVDA. Additionally, the reactivity of all Ile stereoisomers containing DL-Leu and the labeling reagents plateaued upon incubation for >1 h at 50 °C (Fig. 6). Furthermore, we could precisely separate and identify proteinogenic DL-amino acids containing Thr and Ile stereoisomers in peptides. Our method is expected to detect not only racemization and epimerization of free amino acids in plasma and foods but also DL-amino acids containing Thr and Ile stereoisomers in proteins and peptides with physiological activities. This analytical method is expected to contribute to the studies on maple syrup urine disease, its diagnostics using Ile stereoisomers as biomarkers, and highly sensitive sequence analysis of trace amounts of novel bioactive proteins and peptides.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

1) Suzuki M., Sujino T., Chiba S., Harada Y., Goto M., Takahashi R., Mita M., Hamase K., Kanai T., Ito M., Waldor M. K., Yasui M., Sasabe J., Sci. Adv., 7, eabd6480 (2021).
2) Wolosker H., Mol. Neurobiol., 36, 152–164 (2007).
3) Fiore M. M. D., Santillo A., Falvo S., Longobardi S., Baccari G. C., Int. J. Mol. Sci., 17, 1127 (2016).
4) Kakegawa W., Miyoshi Y., Hamase K., Matsuda S., Matsuda K., Kohda K., Emi K., Motohashi J., Konno R., Zaitsu K., Yuzaki M., Nat. Neurosci., 14, 603–611 (2011).
5) Saido T. C., Yamoto-Harigaya W., Iwatsubo T., Kawashima S., Neuro. Sci., 215, 173–176 (1996).
6) Takata T., Fujii N., FEBS J., 283, 850–859 (2016).
7) Jimenéz E. C., Olivera B. M., Gray W. R., Cruz L. J., J. Biol. Chem., 271, 28002–28005 (1996).
8) Donalle G. C., Martorell M. M., Siless G. E., Ruberto L., Cabrera G. M., Nat. Prod. Bioprospect., 12, 26 (2022).
9) Min J. Z., Hatanaka S., Yu H., Higashi T., Inagaki S., Toyo’oka T., J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 879, 3220–3228 (2011).
10) Visser W. F., Verhoeven-Duif N. M., Ophoff R., Bakker S., Klomp L. W., Berger R., de Koning T. J., J. Chromatogr. A, 1218, 7130–7136 (2011).
11) Fujii K., Ikai Y., Mayumi T., Oka H., Suzuki M., Harada K., Anal. Chem., 69, 3346–3352 (1997).
12) Fujii K., Ikai Y., Oka H., Suzuki M., Harada K., Anal. Chem., 69, 5146–5151 (1997).
13) Ayon N. J., Sharma A. D., Gutheil W. G., J. Am. Soc. Mass Spectrom., 30, 448–458 (2019).
14) Konya Y., Taniguchi M., Fukusaki E., J. Biosci. Bioeng., 123, 126–133 (2017).
15) Kimura R., Tsujimura H., Tsuchiya M., Soga S., Ota N., Tanaka A., Kim H., Sci. Rep., 10, 804 (2020).
16) Yokoyama T., Tokuda M., Amano M., Mikami K., Biosci. Biotechnol. Biochem., 81, 1681–1686 (2017).
17) Hamase K., Morikawa A., Ohgusu T., Lindner W., Zaitsu K., J. Chromatogr. A, 1143, 105–111 (2007).
18) Hamase K., Miyoshi Y., Ueno K., Han H., Hirano J., Morikawa A., Mita M., Kaneko T., Lindner W., Zaitsu K., J. Chromatogr. A, 1217, 1056–1062 (2010).
19) Miyoshi Y., Oyama T., Itoh Y., Hamase K., Chromatography, 35, 49–57 (2014).
20) Mijiddorj B., Kaneda S., Sato H., Kitahashi Y., Javkhlantugs N., Naito A., Ueda K., Kawamura I., Biochim. Biophys. Acta. Proteins Proteomics, 1866, 789–798 (2018).
21) Dang B., Chhabra S., Pennington M. W., Norton R. S., Kent S. B. H., J. Biol. Chem., 292, 12599–12605 (2017).
22) Dewapriya P., Prasad P., Damodar R., Salim A. A., Capon R. J., Org. Lett., 19, 2046–2049 (2017).
23) Matsuda K., Kuranaga T., Sano A., Ninomiya A., Takada K., Wakimoto T., Chem. Pharm. Bull., 67, 476–480 (2019).
24) Schadewaldt P., Bodner-Leidecker A., Hammen H. W., Wendel U., Clin. Chem., 45, 1734–1740 (1999).
25) Hauschild T. C., Guerreiro G., Mescka C. P., Coelho D. M., Steffens L., Moura D. J., Manfredini V., Vargas C. R., Toxicol. In Vitro, 57, 194–202 (2019).
26) Kuranaga T., Minote M., Morimoto R., Pan C., Ogawa H., Kakeya H., ACS Chem. Biol., 15, 2499–2506 (2020).
27) Morimoto R., Matsumoto T., Minote M., Yanagisawa M., Yamada R., Kuranaga T., Kakeya H., Chem. Pharm. Bull., 69, 265–270 (2021).
28) Kuranaga T., Kakeya H., Methods Enzymol., 665, 105–133 (2022).
29) Ozaki M., Kuwayama T., Hirose T., Shimotsuma M., Hashimoto A., Kuranaga T., Kakeya H., Anal. Bioanal. Chem., 414, 4039–4046 (2022).
30) Ozaki M., Shimotsuma M., Kuranaga T., Kakeya H., Hirose T., Analyst, 148, 1209–1213 (2023).
31) Chan W. C., White P. D., “Fmoc Solid Phase Peptide Synthesis; A Practical Approach,” Oxford University Press, New York, 2000.
32) Ozaki M., Shimotsuma M., Hirose T., Anal. Biochem., 655, 114837 (2022).
33) Ozaki M., Shimotsuma M., Hirose T., MethodsX, 10, 102061 (2023).
34) Kato R., Sato T., Kanamori M., Miyake M., Fujimoto A., Ogawa K., Kobata D., Fujikawa T., Wada Y., Mitsuishi R., Takahashi K., Imano H., Ijiri Y., Mino Y., Chikuma M., Tanaka K., Hayashi T., Biol. Pharm. Bull., 40, 290–296 (2017).

Corresponding author

Register with J-STAGE for free!