- Email: [email protected]
l Configuration Determination of Peptides with Fluorogenic Edman Reagent 7-[(N,N-Dimethylamino)sulfonyl]-2,1,3- benzoxadiazol-4-yl Isothiocyanate
Analytical Biochemistry 270, 257–267 (1999) Article ID abio.1999.4031, available online at http://www.idealibrary.com on
Detection of DBD-Carbamoyl Amino Acids in Amino Acid Sequence and D/L Configuration Determination of Peptides with Fluorogenic Edman Reagent 7-[(N,N-Dimethylamino)sulfonyl]-2,1,3benzoxadiazol-4-yl Isothiocyanate Yong Huang, Hirokazu Matsunaga, Akira Toriba, Tomofumi Santa, Takeshi Fukushima, and Kazuhiro Imai 1 Graduate School of Pharmaceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan
Received November 16, 1998
A method for amino acid sequence and D/L configuration identification of peptides by using fluorogenic Edman reagent 7-[(N,N-dimethylamino)sulfonyl]-2,1,3benzoxadiazol-4-yl isothiocyanate (DBD-NCS) has been developed. This method was based on the Edman degradation principle with some modifications. A peptide or protein was coupled with DBD-NCS under basic conditions and then cyclized/cleaved to produce DBD-thiazolinone (TZ) derivative by BF 3, a Lewis acid, which could significantly suppress the amino acid racemization. The liberated DBD-TZ amino acid was hydrolyzed to DBD-thiocarbamoyl (TC) amino acid under a weakly acidic condition and then oxidized by NaNO 2/H 1 to DBD-carbamoyl (CA) amino acid which was a stable and had a strong fluorescence intensity. The individual DBD-CA amino acids were separated on a reversed-phase high-performance liquid chromatography (RP-HPLC) for amino acid sequencing and their enantiomers were resolved on a chiral stationary-phase HPLC for identifying their D/L configurations. Combination of the two HPLC systems, the amino acid sequence and D/L configuration of peptides could be determined. This method will be useful for searching D-amino-acid-containing peptides in animals. © 1999 Academic Press Key Words: amino acid sequence; D-amino-acid-containing peptide; fluorogenic reagent; Edman method; chiral separation; HPLC; oxidation.
1 To whom correspondence should be addressed. Fax: 81-3-58023339.
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Proteins and peptides in nature are generally believed to consist of L-type amino acids exclusively. The D-amino-acid-containing peptides are found only in the peptidoglycan of bacteria cell wall and some antibiotic peptides, which are synthesized by multienzyme complexes (1). However, recent studies have indicated that the ribosomal-origin peptides in animals also contain D-amino acid residues (2). Furthermore, the D-amino acids in these peptides are essential for their biological activities because the change of D-amino acid to L-amino acid would lose their biological activities (2, 3). At present, the D-amino acid at the specific position of the mature peptides is thought to be due to the conversion of L-isomer to D-isomer by a posttranslational mechanism, since the gene-encoded peptide precursors consisted of only L-amino acids (2), and some isomerase involving the conversion has recently been purified (4). Although the racemization occurs in aged human proteins such as a-crystallin and amyloid (5, 6), it is not certain whether the physiologically active peptides containing D-amino acid residue exist in mammalian tissues. To identify D-amino acid residues in peptides, a method in which an amino acid sequence as well as its stereo configuration could be determined has been developed. However, there are very few reports on such studies (7–11). The Edman degradation method, the most widely used method for N-terminal sequencing of proteins, exhibited significant racemization of amino acids during conventional Edman degradation procedures making identification of the D/L configuration difficult (12). In previous papers, we have developed an 257
258
HUANG ET AL.
Edman fluorogenic reagent containing the benzofurazan structure, 7-[(N,N-dimethylamino)sulfonyl]2,1,3-benzoxadiazol-4-yl isothiocyanate (DBD-NCS) 2 (13), and established a sequential method by utilizing DBD-thiazolinone (TZ) derivatives for amino acid identification (14). We also elucidated that the amino acid racemization occurring at the cleavage stage using trifluoroacetic acid (TFA) was due to the replacement of the proton at the a-carbon position by the acid hydrogen of TFA, and found that BF 3, a Lewis acid, could significantly suppress the racemization (15). To obtain a more stable derivative than DBD-TZ amino acid for sequential analysis, here we examine the conversion of DBD-TZ amino acid to DBD-carbamoyl (CA) amino acid mediated by DBD-thiocarbamoyl (TC) amino acid and establish a practical method for amino acid sequence and D/L configuration identification of peptides. MATERIALS AND METHODS
Material. DBD-NCS was synthesized as previously described (13). Standard D- or L-amino acids, dipeptides, insulin chain B, oxidized and (D-Ala 2)-deltorphin II were obtained from Sigma Chemical Co. (St. Louis, MO). Boron trifluoride diethyl etherate complex (BF 3 z Et 2O) obtained from Tokyo Chemical Industry Co., Ltd was distilled and stored as before (16). Pyridine and TFA were of amino acid sequencing grade (Wako Pure Chemicals, Japan). All other reagents and solvents were of analytical or guaranteed reagent grade. High-performance liquid chromatography conditions. A Hitachi HPLC system consisting of an L-6200 intelligent pump, a 7161 injector (Rheodyne, Cotati, CA), an F-1000 fluorometric detector, a D-2500 Chromato Integrator, an L-4000 UV–vis detector, and a Shimadzu CR-3A Chromatopac was used. The analytes were monitored simultaneously by fluorometric and UV–vis detectors, those connected with different recorders. Fluorometric detection was made at 520 nm with excitation at 384 nm. The UV–vis was monitored at 267 or 384 nm. The reversed phase (RP)-HPLC column was TSK gel ODS 80Ts (250 mm 3 4.6 mm i.d., Tosoh, Tokyo). For separation of individual DBD-CA amino acids, the following mobile phase was used: Eluent A, CH 3CN/H 2O (1/9) containing 0.1% TFA and eluent B, CH 3CN/H 2O (9/1) containing 0.1% TFA; gradient program: 0–2 min, B, 10%; 2 Abbreviations used: DBD-NCS, 7-[(N,N-dimethylamino)sulfonyl]2,1,3-benzoxadiazol-4-yl isothiocyanate; DBD-TZ, DBD-thiazolinone; DBD-TC, DBD-thiocarbamoyl; DBD-CA, DBD-carbamoyl; DBD-TH, DBD-thiohydantoin; RP-HPLC, reversed-phase HPLC; b-CD, b-cyclodextrin; Ph-CD, phenylcarbamoylated b-cyclodextrin; TFA, trifluoroacetic acid; TEA, triethylamine.
2–32 min, B, 10–30%; 32–37 min, B, 30–50%; 37–55 min, B, 50%. Flow rate was kept at 1.0 ml/min. The column temperature was maintained at 40°C with a 655A52 column oven (Hitachi). For isolation of DBD-TZ, DBD-TC, and DBD-CA amino acids, the mobile phase was 10 mM formic acid in CH 3CN/H 2O solutions (the ratio of CH 3CN to H 2O was changed according to the amino acids). The other HPLC conditions were the same as the above. For chiral separation of DBD-CA amino acids, the following HPLC conditions were used—Column, Ultron ES-CD (150 3 6.0 mm i.d., Shinwa Chemical Industries, Ltd., Japan) at room temperature; Eluents: A, H 2O/CH 3OH/ CH 3CN (75/5/20) containing 10 mM acetic acid; B, H 2O/ CH 3OH/CH 3CN (10/70/20) containing 10 mM acetic acid; gradient program, 0 – 8 min, B, 0 – 0%; 8 –25 min, B, 0–100%; 25–35 min, B, 100–100%; flow rate, 1.0 ml/min. Fluorometric detection was the same as described above. Preparation of DBD-TZ, DBD-TC, and DBD-CA amino acids. DBD-TZ amino acids were prepared according to the method described previously (14). DBD-TC amino acids were prepared by coupling DBD-NCS (20 mM 3 10 ml) with free amino acids in 50% pyridine at 50°C for 15 min. After washing with 200 ml of n-heptane/dichloromethane (9/1) for 3 times, the mixture was dried by a centrifugal evaporator at 50°C for 15 min. DBD-CA amino acids were prepared by oxidation of DBD-TC amino acids with NaNO 2/H 1 as described below. Hydrolysis of DBD-TZ amino acids to DBD-TC amino acids. The DBD-TZ amino acid isolated from HPLC was dried by centrifugal evaporator, dissolved in 2 ml of CH 3CN, and added with 8 ml of 0.4 M HCl. The mixture was reacted at 50°C for different times and analyzed by HPLC. Oxidation of DBD-TC amino acids to DBD-CA amino acids. The DBD-TC amino acid (200 pmol) prepared as described above was dissolved in 10 ml of CH 3CN and added with 5 ml of 4 M HCl and 5 ml of 0.5 M NaNO 2 in water and reacted at room temperature(about 22°C) for different times. The reaction mixture was then neutralized with 1.0 M NaHCO 3 to stop the reaction and an aliquot was analyzed immediately by HPLC. Identification of DBD-CA amino acids. An M-1200H LC/APCI-MS (Hitachi) system connected to a Rheodyne injector and an intelligent pump (L-6200, Hitachi) was used. The correspondence of DBD-CA amino acid prepared as described above was collected by HPLC and then dried by centrifugal evaporator. The residue was dissolved with methanol and analyzed by LC–MS. The mass spectrometer was operated in the APCI mode with a drift, focus, and multiplier voltages respectively set at 20, 120, and 1800 V. The temperature of the vaporizer and dessolvation regions were set at 200 and 400°C, respectively. The flow injection analysis (FIA) mode was
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
259
FIG. 1. Scheme of Edman method.
used at a flow rate of 1.0 ml/min of methanol. Intelligent fluorometric spectrometer (F-4010, Hitachi) and UV–vis spectrophotometer (Ubset 50, JASCO) were used for fluorometric and UV–vis spectra measurements, respectively. Effects of pH and solvents on the fluorescence intensity and the stability of DBD-CA amino acid. DBDCA-Leu collected by HPLC was diluted with 0.1 M phosphate buffers of different pH or various solvents such as methanol, ethanol, acetonitrile, and water. The fluorescence spectrum of the solution was immediately measured on the fluorescence spectrophotometer. To examine the stability of DBD-CA amino acid, the DBDCA-Leu collected by HPLC was divided into several equal parts and dried by centrifugal evaporator; the residue was dissolved in 0.1 M phosphate buffers of different pH or solvents such as methanol, acetonitrile, and water, and the solution was kept at room temperature. An aliquot was withdrawn at appropriate time
intervals and determined by HPLC. The degradation of DBD-CA-Leu was obeyed pseudo-first-order kinetics, ln[A] 5 2k obst 1 C, where [A] is the residual content of DBD-CA-Leu, while t and C are the time and intercept values, respectively. The k obs is the observed degradation rate constant. Procedures for manual sequencing of peptides. A peptide (insulin chain B (500 pmol) or deltorphin II (2 nmol) ) dissolved in 20 ml of 50% pyridine/water solution was added with 5 ml 1% triethylamine (TEA) in acetonitrile and 10 ml of 20 mM DBD-NCS in pyridine. The mixture was reacted in an inert atmosphere at 50°C for 15 min. After the coupling reaction, the mixture was washed with 200 ml of n-heptane/dichloromethane (6/4) three times and dried by a centrifugal evaporator at 50°C for 15 min. The DBD-thiocarbamoyl-peptide was cyclized/cleaved by adding 30 ml of 1% BF 3 z Et 2O in acetonitrile and heated at 50°C for 5 min, and then dried by a stream of N 2. The residue was
260
HUANG ET AL.
for 5 min for hydrolysis. The reaction mixture was then oxidized at room temperature (22°C) for 10 min after adding 5 ml of 4 M HCl and 5 ml of 0.5 M NaNO 2 in water. The reaction was stopped by neutralizing with 1.0 M NaHCO 3 (23 ml) and added with 20 ml of 0.15 M methionine for reducing the excessive oxidizing agent. An aliquot (20 ml) of the mixture was subjected to HPLC for analysis. RESULTS AND DISCUSSION
FIG. 2. Time course of hydrolysis of DBD-TZ amino acids to DBD-TC amino acids. Reaction conditions: DBD-TZ amino acid in acetonitrile (2 ml) was added with 0.4 M HCl ( 8 ml) and reacted at 50°C for different times.
added with 20 ml of water and extracted two times with 100 ml of benzene/ethyl acetate (1/4) for insulin chain B or n-hepatane/ethyl acetate (1/1) for deltorphin II. The aqueous phase containing residual peptide was dried by centrifugal evaporator and subjected to the next cycle. The combined organic phase was dried by N 2 and dissolved with 2 ml of CH 3CN, and added with 8 ml of 0.4 M aqueous HCl. The mixture was heated at 50°C
The conventional Edman degradation method was carried out by three steps as shown in Fig. 1 (17). A peptide or protein was first coupled with arylisothiocynate (aryl:phenyl or DBD-) to form the thiocarbamoyl peptide, which was then cyclized/cleaved to produce the TZ derivative by strong acid. The released TZ amino acid was further converted into a stable thiohydantoin (TH) amino acid, which was identified by HPLC. As previously reported, the DBD-TH amino acids lack the fluorescence to obtain a highly sensitive detection; one way is to detect DBD-TZ amino acids directly, omitting the conversion step (14). The other is to convert the DBD-TZ derivative to another derivative. Edman previously observed that the phenylthiocarbamoyl moiety showed a tendency toward oxidative desulfuration with formation of phenyl-CA amino acids (18). We found that DBD-CA amino acids were stable and had a strong fluorescence intensity, although DBD-TC amino acids afforded very little fluorescence intensity. We therefore attempted the conversion of
FIG. 3. Time course of oxidation of DBD-TC amino acids to DBD-CA amino acids. Reaction conditions: DBD-TC amino acid in acetonitrile (10 ml) was added with 4 M HCl (5 ml) and 0.5 M NaNO 2 (5 ml) and reacted at room temperature for different times.
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
261
FIG. 4. Transformation of DBD-TZ-Leu into DBD-CA-Leu by two steps. DBD-TZ-Leu (A) was hydrolyzed to DBD-TC-Leu (B) and further oxidized to DBD-CA-Leu (C). Hydrolysis conditions: DBD-TZ-Leu in CH 3CN (2 ml) was added with 0.4 M HCl (8 ml) and reacted at 50°C for 5 min. Oxidative conditions: The above hydrolyzed mixture was further added with 4 M HCl (5 ml) and 0.5 M NaNO 2 (5 ml) and reacted at room temperature for 10 min. HPLC conditions: Column, TSK gel ODS 80Ts (250 3 4.6 mm i.d.) at room temperature; eluents, CH 3CN/H 2O (55/45) containing 10 mM formic acid; flow rate, 1.0 ml/min. Detected simultaneously by fluorescence (upper panel) and UV (lower panel) with wavelength at Ex 5 384 nm; Em 5 520 and 267 nm, respectively.
DBD-TZ amino acids to DBD-CA amino acids for sequential analysis. Hydrolysis of DBD-TZ Amino Acids to DBD-TC Amino Acids DBD-TZ amino acids were unstable in protic solvents and easily hydrolyzed to DBD-TC amino acids under weakly acidic conditions. The reaction rate was dependent on the ratio of CH 3CN and water. The lower CH 3CN concentration favored a higher rate of hydrolysis (14). Figure 2 shows the time course of hydrolysis of several DBD-TZ amino acids (Leu, Phe, Arg, and Asn) compared to their corresponding DBD-TC amino acids under a 2/8 CH 3CN/0.4 M aqueous HCl solution at 50°C. The reaction was completed within 5 min. Ten minutes were necessary to complete hydrolysis when the CH 3CN/0.4 M aqueous HCl solution ratio was 5/5
at 50°C (data not shown). Under the above hydrolysis conditions, DBD-TZ amino acids could be quantitatively transformed to DBD-TC amino acids, whereas no DBD-TH amino acids were formed. Oxidation of DBD-TC Amino Acids to DBD-CA Amino Acids The transformation of thiocarbonyl compounds to their corresponding oxo-analogues has been reported by utilizing NaNO 2/H 1 (19), H 2O 2 (20), MnO 2 (21), m-chloroperbenzoic acid (22), etc. To investigate the oxidative desulfuration of DBD-TC amino acid to DBD-CA amino acid, various oxidizing agents were examined. We found that DBD-TC-Leu could be rapidly and specifically oxidized to DBD-CA-Leu by NaNO 2/H 1 as shown in Figs. 3, 4B, and 4C. The formed DBD-CA-Leu was identified by LC–MS and UV spec-
262
HUANG ET AL.
FIG. 5. UV spectra of DBD-TC-Leu (A) and DBD-CA-Leu (B) in acetonitrile.
tra. The mass number (M 1 H) 1 of the peak at Rt of 6.2 min in Fig. 4C was 400 which corresponds to the molecular weight of DBD-CA-Leu (399). The DBD-CA-Leu has an absorption at 244 nm, whereas it loses the absorption at around 267 nm (Fig. 5B), which is similar to that of phenylcarbamoyl amino acid as previously reported by Edman (18). The produced DBD-CA derivatives of other amino acids by NaNO 2/H 1 oxidation were also identified by LC–MS and UV spectra. The time course of transformation of several DBD-TC amino acids to their corresponding DBD-CA amino acids is shown in Fig. 3. The reactions for all the amino acids were completed within 5 min except Lys, for which the maximal yield was reached at 30 min. This might be due to Lys being labeled by two thiocarbamoyl groups at a and e positions; both were desulfurated by oxidation. Most DBD-CA amino acids formed were stable in NaNO 2/H 1 medium for 60 min except Tyr, Met, and Trp, whose side chains were sensitive to NaNO 2/H 1 oxidation. Tyrosine was further nitrated at the 3 position of hydroxyphenyl group if the reaction time was longer. From these results, we decided that 10 min was an optimal time in which the nitration of Tyr was less than 15%. Methionine was very sensitive to oxidation, and most of the methionine was oxidized into DBD-CA-methionine sulfoxide under the above conditions. No further oxidation to DBD-CA-methio-
nine sulfone occurred even after 1 h. Therefore, DBDCA-methionine sulfoxide was used for identifying the methionine residue during sequencing. Tryptophan was also sensitive to oxidation. The oxidation of DBDTC-Trp by NaNO 2/H 1 produced two major fluorescence peaks, which presumably correspond to DBD-CA-oxindole derivatives (23–25). DBD-CA-Trp has little fluorescence intensity due to fluorescence quenching of two fluorophores of benzofurazan structure and aminoindole moiety of Trp. The oxidation of the Trp group makes it detectable by fluorescence, although its intensity is several times lower than that of other DBD-CA AAs. DBD-TC-Pro was difficult to oxidize to DBD-CAPro under oxidative conditions, instead, facilely converting to another fluorescence derivative, which showed a longer wavelength than that of the DBD-CA derivatives. The former’s maximal excitation and emission in CH 3CN/H 2O solution were at 440 and 575 nm, respectively, while the latter’s were at 384 and 520 nm, respectively. Proline was a secondary amine, which differed from other amino acids. Considering that the facile dimerization of thiocarbamates or thioureas by oxidation was previously reported (26, 27), we speculated that DBD-TC-Pro might form a dimeric derivative through disulfide linkage under the oxidation. It was convenient that the resulting hydrolysis samples could directly be followed by oxidative procedures
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
263
under the same chromatographic conditions, indicating that the DBD-CA amino acid was superior to the DBD-TZ amino acid for detection.
TABLE 1
Stability of DBD-CA-Leu in Different Solvents at Room Temperature Solvent
K obs (h 21)
t 0.9 (h)*
t 0.5 (h) a
H 2O Methanol Acetonitrile 0.1% TFA,CH 3CN/H 2O (1/9)
0.0073 0.0037 0.0023 0.0012
14.4 28.4 45.7 87.5
94.9 187.3 301.3 577.5
a
Time at which the residual amount of the derivative is 90 or 50% of the initial amount. K obs, observed degradation rate constant as described under Materials and Methods.
without any treatment because both hydrolysis and oxidation reactions were performed in HCl solutions. The quantitative transformation of DBD-TZ-Leu to DBD-CA-Leu by two-step reactions is shown in Fig. 4. No byproducts were observed in the chromatograms. The fluorescence intensity (peak area) of DBD-CA-Leu was about 3 times higher than that of DBD-TZ-Leu
Effects of pH and Solvents on the Fluorescence Intensity and the Stability of DBD-CA Amino Acids DBD-CA amino acids showed a stronger fluorescence intensity under acidic conditions and in organic solvents. The fluorescence intensity of DBD-CA-Leu in organic solvents such as CH 3CN, MeOH, and EtOH was about 7–11 times higher than that in H 2O, while the fluorescence intensity of DBD-CA-Leu in 0.1 M phosphate buffer at pH 2.0 was 1.3 times higher than that at pH 8.5. In comparison with DBD-TZ amino acids (14), DBD-CA amino acids were relatively stable in various solvents (Table 1). The k obs of DBD-CA-Leu in 0.1 M phosphate buffer at pH 2.0, 4.0, and 8.5 was 0.004, 0.009, and 0.014 (h 21), respectively, which indicated that the DBD-CA derivative was more stable under
FIG. 6. Separation of DBD-CA amino acids on RP-HPLC. Peaks: 1, cysteic acid, 2, His, 3, Asn, 4, Gln, 5, Arg, 6, Met-O, 7, Ser, 8, Gly, 9, Asp, 10, Glu, 11, Thr, 12, Ala, 13, Tyr, 14, Val, 15, Pro-X, 16, Trp-X, 17, Ile, 18, Leu, 19, Phe, 20, Lys. Met-O, Met-sulfoxide; Pro-X, unidentified DBD-TC-Pro oxidative; Trp-X, unidentified DBD-TC-Trp oxidative. *The shoulder peak appearing just before Val was a degradation product of DBD-NCS formed in coupling reaction. The amounts of these derivatives corresponding to each amino acids were 20 pmol except Trp which was 50 pmol. HPLC conditions: Column, TSK gel ODS 80Ts (250 3 4.6 mm i.d.) at 40°C. Eluents: A, CH 3CN/H 2O (10/90) containing 0.1% TFA; B, CH 3CN/H 2O (90/10) containing 0.1% TFA; Gradient program, 0 –2 min, B, 10 –10%, 2–32 min, B, 10 –30%, 32–37 min, B, 30 –50%, 37–50 min, B, 50 –50%; flow rate, 1.0 ml/min. Ex 5 384 nm; Em 5 520 nm.
264
HUANG ET AL.
acidic conditions. The DBD-CA derivatives of other amino acids also showed a stability similar to that of DBD-CA-Leu. These results revealed that DBD-CA amino acids were suitable for detection under various chromatographic conditions.
TABLE 2
Chiral Separation of DBD-CA Amino Acids on Native b-CD Column k9 DBD-CA AA
Separation of DBD-CA Amino Acids on RP-HPLC Since the fluorescence intensity and the stability of DBD-CA amino acids were favored under acidic conditions, we selected CH 3CN/H 2O containing 0.1% TFA as a mobile phase. Under acidic conditions, the ionization of the COOH group was less than that under neutral conditions, which resulted in increasing the retention of DBD-CA amino acid on a column because it decreased its polarity and, alternatively, increased the fluorescence intensity since a higher organic solvent ratio in solution was needed to elute the analytes. All DBD-CA amino acids could be separated on RP-HPLC under the conditions described (Fig. 6). The column temperature kept at 40°C was necessary for the separations of DBD-CA-Gly, DBD-CA-Asp, and DBD-CAThr. This was also of benefit for reproducing peak retention times. Under acidic conditions, the fluorescence intensity of Pro derivative at Ex 5 384 nm with Em 5 520 nm was 2.2 times lower than at its optimal wavelength of Ex 5 440 nm with Em 5 575 nm. Still the Pro derivative had an intensity similar to other DBD-CA AAs at Ex 5 384 nm with Em 5 520 nm. The detection limits for DBD-CA amino acids under HPLC conditions were 100 fmol to subpicomole (signal/noise 5 3). Enantiomeric Separation of DBD-CA Amino Acids on Chiral Stationary Phases To separate DBD-CA amino acid enantiomers, we examined various types chiral HPLC columns and found that DBD-CA amino acid enantiomers were well separated on a native b-cyclodextrin (CD)-bonded chiral stationary phase with an acidic mobile phase. The mobile phase for the b-CD series chiral columns were usually carried out by water with organic modifiers (16, 28). Although most of the DBD-CA amino acid enantiomers could also be resolved on native b-CD or phenylcarbamoylated b-CD (Ph-CD) or ES-1/4 Ph-CD columns (16) with neutral mobile phases such as ammonium acetate/MeOH (data not shown), the peaks were broad and one or two DBD-CA amino acid enantiomers could not be separated under these conditions. Furthermore, the neutral eluents would decrease DBD-CA amino acid retention on columns compared with acidic eluents. We therefore chose MeOH/H 2O containing 10 mM acetic acid as a mobile phase. The separation factors for DBD-CA amino acid enanti-
Arg His Ala Val Ser Ile Gly Leu Asn Thr Glu Met Gln Asp Phe Lys Met-O a
D
L
1.35 1.94 8.56 8.64 8.73 8.89
a
1.64 2.16 8.96 9.01 9.27 9.26
1.21 1.11 1.05 1.04 1.06 1.04
9.63 9.92 10.19 10.42 10.49 10.56 11.37 11.64 12.04 13.19 13.49
1.02 1.05 1.04 1.04 1.04 1.05 1.03 1.04 1.04 1.06 1.06
9.42 9.47 9.48 9.76 10.01 10.05 10.08 11.00 11.15 11.59 12.40 12.72
Note. k9 and a are capacity factor and separation factor, respectively. HPLC conditions: Column, Ultron ES-CD(150 3 6.0 mm i.d.) at room temperature; Eluents: A, H 2O/MeOH/CH 3CN (75/5/20) containing 10 mM acetic acid; B, H 2O/MeOH/CH 3CN (10/70/20) containing 10 mM acetic acid; Gradient program, 0 – 8 min, B, 0 – 0%; 8 –25 min, B, 0 –100%; 25–35 min, B, 100 –100%; Flow rate, 1.0 ml/min. Ex 5 384 nm; Em 5 520 nm. a Met-O, Met-sulfoxide, which has two asymmetry centers at a-C and d-S.
omers did not decrease with this eluent, while it improved DBD-CA amino acid detection limits. In acidic mobile phase, DBD-CA amino acids revealed longer retention on native b-CD column than that on Ph-CD and ES-1/4 Ph-CD columns; we therefore selected native b-CD column for separation of DBD-CA amino acids. In the native b-CD column with MeOH/H 2O containing 10 mM acetic acid as mobile phase, DBD-CA-Leu enantiomers could only be partially resolved. When acetonitrile was added into the mobile phase, however, not only was the chiral separation of DBD-CA-Leu improved but also DBD-CA amino acid peaks were sharpened. The ionized forms of DBD-CA-His and DBD-CA-Arg under acidic conditions make them become eluted much faster than other amino acid derivatives. To overcome this problem, a gradient elution was performed, which also sharpened the peaks. Table 2 lists the capacity factors and separation factors for each DBD-CA amino acid under the established HPLC conditions. DBD-CA-Met-sulfoxide had two asymmetry centers at a-C and d-S. All the DBD-CA amino acids eluted D form faster than L type. The enantiomeric
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
265
FIG. 7. Chromatograms of manual sequencing of insulin chain B, oxidized (500 pmol). HPLC conditions were the same as those described in the legend to Fig. 6.
resolutions of all amino acids except Pro and Trp were achieved; however, the separations of individual DBD-CA amino acids were not successful. The detection limits for DBD-CA amino acids under the chiral separation conditions were at subpicomole to picomole levels (signal/noise 5 3).
Amino Acid Sequencing of Peptides A manual procedure described under Materials and Methods was applied to sequencing of peptides. The coupling condition was modified by further addition with 1% TEA to the reaction mixture to increase pH,
266
HUANG ET AL.
FIG. 8. Amino acid sequencing and D/L configuration identification of (D-Ala 2)-deltorphin II. (D-Ala 2)-deltorphin II (2 nmol), Tyr-D-Ala-PheGlu-Val-Val-Gly-amide. *The peaks at Rt of 26.5 and 27.5 min of cycle 1 were DBD-CA-(D)-3-nitro-Tyr and DBD-CA-(L)-3-nitro-Tyr, respectively. HPLC conditions were the same as those described in Table 2.
which was necessary to complete the coupling reaction, although it also increased byproduct formation (14). The DBD-TC peptides were cyclized/cleaved by BF 3 instead of TFA, to suppress the racemization as reported previously (15). The released DBD-TZ amino acid was further converted into DBD-CA amino acid by hydrolysis and oxidative reactions as described above. Figure 7 shows the chromatograms of the results of first six cycles obtained from the sequencing of insulin chain B (500 pmol). The amino acid residues were easily identified due to few interfering peaks appearing on the chromatograms. The major byproduct peak of DBD-NH 2 formed during coupling reaction disappeared on the chromatograms because it was destroyed by NaNO 2/H 1 oxidation and could not be detected fluorometrically. The amino acid residue in the former cycle was not observed in the next cycle indicated that the coupling was completed. The present sequential results are comparable to the other fluorescence Edman reagents such as fluorescein isothiocyanate (29) and dansylamino isothiocyanate (30). Amino Acid Sequencing and D/L Identification of Peptide The manual sequencing procedures for D-amino-acidcontaining peptides were the same as those described above. Figure 8 shows the first four cycles for amino acid sequencing and D/L identification of (D-Ala 2)-deltorphin II (2 nmol), which was a neuroactive D-aminoacid-containing peptide analogous to the dermorphin family (31). Although amino acid racemizations were
partially occurring at the proposed method, the D or L configuration of each amino acid residue could still be discriminated. The nitration of tyrosine by NaNO 2/H 1 oxidation at the first cycle was less than 15%. The second amino acid residue of the peptide was easily identified as D-Ala. The partial racemization was thought to occur at the hydrolysis step because no or little racemization occurred at the oxidation and cleavage steps, respectively (15). The opening of the TZ ring during the hydrolysis in HCl solution would cause some racemizations. Because DBD-CA amino acid enantiomers were eluted closely with exclusively D-type amino acids faster than the L-type under the HPLC conditions, the partial racemizations would produce asymmetric dual peaks on the chromatogram, in which a lower former peak compared to the latter peak would indicate an L-amino acid residue and otherwise a D-amino acid residue. Since peptides consist mainly of L-amino acids, the occurrence of a D-amino acid residue in sequence was easily identified by the chromatograms. Under HPLC conditions, the individual DBD-CA amino acid enantiomers have not been completely separated in a single run; the sequencing of amino acid and D/L configuration identification therefore needed to be performed under two HPLC conditions. One portion of a sample was first applied to RP-HPLC for amino acid sequencing; the other part of the sample was applied for chiral column HPLC system for D/L configuration identification. The amino acid residues could be reconfirmed by the chiral stationary-phase HPLC system.
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
CONCLUSIONS
A method for amino acid sequence and D/L configuration determination using the fluorogenic Edman reagent DBD-NCS has been developed. This method was a modified Edman degradation procedure by conversion of DBD-TZ amino acids to DBD-CA amino acids instead to DBD-TH amino acids for amino acid identification. The facile conversion of DBD-TZ amino acid to DBD-CA amino acid and the stable and fluorescence properties of DBD-CA amino acid indicated that it was suitable for sequential analysis. Although the exact structures of the oxidative products of DBD-TC-Pro and DBD-TC-Trp have not been elucidated as yet, these two amino acid residues could still be identified by the present method. The characteristics of the new procedure were that there were few interfering peaks on chromatograms, which makes amino acid identification easy. Despite a partial racemization occurring at the hydrolysis step, the D/L configurations of amino acid residue could still be identified.The proposed method will be useful for searching the D-amino acid in animal peptides. Further studies are currently in progress to solve the remaining problems and to automate these procedures. ACKNOWLEDGMENTS The authors thank Shinwa Chemical Industries, Ltd. for the generous gifts of b-CD series chiral columns and Tosoh Co. for supplying TSK gel ODS 80Ts column. This work was financially supported by the Japan Society for Promotion of Sciences and a Grant-in-Aid for Science Research from the Ministry of Education, Science, and Culture of Japan. Y. Huang is a recipient of a foreign postdoctoral fellowship from the Japan Society for Promotion of Sciences.
REFERENCES 1. Kleinkauf, H., and von Dohren, H. (1996) Eur. J. Biochem. 236, 335–351. 2. Kreil, G. (1997) Annu. Rev. Biochem. 66, 337–345. 3. Montecucchi, P. C., de Castiglione, R., Piani, S., Gozzini, L., and Erspamer, V. (1981) Int. J. Peptide Protein Res. 17, 275–283. 4. Shikata, Y., Watanabe, T., Teramoto, T., Inoue,A., Kawakami, Y., Nishizawa, Y., Katayama, K., and Kuwada, M. (1995) J. Biol. Chem. 270, 16719 –16723. 5. Fujii, N., Muraoka, S., Satoh, K., Hori, H., and Harada, K. (1991) Biomed. Res. 12, 315–321. 6. Roher, A. E., Lowenson, J. D.,Clarke, S., Wolkow, C., Wang, R., Cotter, R. J., Reardon, I. M., Zu¨rcher-Neely, H. A., Heinrikson, R. L., Ball., M. J., and Greenberg, B. D. (1993) J. Biol. Chem. 268, 3072–3083.
PEPTIDES
267
7. Scaloni, A., Simmaco, M., and Bossa, F. (1991) Anal.Biochem. 197, 305–310. 8. Toyo’oka, T., Suzuki, T., Watanabe, T., and Liu, Y. M. (1996) Anal. Sci. 12, 779 –782. 9. Suzuki, T., Watanabe, T., and Toyo’oka, T. (1997) Anal. Chim. Acta 352, 357–363. 10. Kurosu, Y., Murayama, K., Shindo, N., Shisa, Y., Satou, Y., and Ishioka, N. (1996) J. Chromatogr. 752, 279 –286. 11. Kurosu, Y., Murayama, K., Shindo, N., Shisa, Y., Satou, Y., and Ishioka, N. (1997) J. Chromatogr. 771, 311–317. 12. Edman, P. (1950) Acta Chem. Scand. 4, 283–293. 13. Uzu, S., Imai, K., Nakashima, K., and Akiyama, S. (1993) Biomed. Chromatogr. 9, 152–153. 14. Matsunaga, H., Santa, T., Hagiwara, K., Homma, H., Imai, K., Uzu, S., Nakashima, K., and Akiyama, S. (1995) Anal. Chem. 67, 4276 – 4282. 15. Matsunaga, H., Santa, T., Iida, H., Fukushima, T., Homma, H., and Imai, K. (1996) Anal. Chem. 68, 2850 –2856. 16. Iida, T., Matsunaga, H., Fukushima, T., Santa, T., Homma, H., and Imai, K. (1997) Anal. Chem. 69, 4463– 4468. 17. Edman, P. (1956) Acta Chem. Scand. 10, 761–768. 18. Ilse, D., and Edman, P. (1963) Aust. J. Chem. 16, 411– 416. 19. Jorgensen, K. A., Ghattas, A.-B. A. G., and Lawesson, S.-O. (1982) Tetrahedron 38, 1163–1168. 20. Walter, W., and Randau, G. (1969) Liebigs Ann. Chem., 722, 57–72. 21. Sharma, T. C., Sachni, N. S., and Lal, A. (1978) Bull. Chem. Soc. Jpn. 51, 1245–1246. 22. Kochar, K. S., Cottrell, D. A., and Pinnick, H. W. (1983) Tetrahedron Lett. 24, 1323–1326. 23. Kato,Y., Kawakishi, S., Aoki, T., Itakura, K., and Osawa, T. (1997) Biochem. Biophy. Res. Commun. 234, 82– 84. 24. Watanabe, N., Toyo’oka, T., and Imai, K. (1987) Biomed. Chromatogr. 2, 99 –103. 25. Imai, K., Ueda, E., and Toyo’oka, T. (1988) Anal. Chim. Acta 205, 7–14. 26. Blankespoor, R. L., Doyle, M. P. Hedstrand, D. M. Tamblyn, W. H., and Van Dyke, D. A. (1981) J. Am. Chem. Soc. 103, 7096 –7101. 27. Olah, G. A., Arvanghi, M., Ohannesian, L. and Surya Prakash, G. K. (1984) Synthesis 1984, 785–786. 28. Rizzi, A. M., Cladrowa-Runge, S., Jonsson, H., and Osla, S. (1995) J. Chromatogr. 710, 287–295. 29. Muramoto, K., Kamiya, H., and Kawauchi, H. (1984) Anal. Biochem. 141, 446 – 450. 30. Hirano, H., and Wittmann-Liebold, B. (1986) Biol. Chem. 367, 1259 –1265. 31. Erspamer, V., Melchiorri, P., Falconieri-Erspamer, G., Negri, L., Corsi, R., Severini, C., Barra, D., Simmmaco, M., and Kreil, G. (1989) Proc. Natl. Acad. Sci. USA 86, 5188 –5192.
Detection of DBD-Carbamoyl Amino Acids in Amino Acid Sequence and D/L Configuration Determination of Peptides with Fluorogenic Edman Reagent 7-[(N,N-Dimethylamino)sulfonyl]-2,1,3benzoxadiazol-4-yl Isothiocyanate Yong Huang, Hirokazu Matsunaga, Akira Toriba, Tomofumi Santa, Takeshi Fukushima, and Kazuhiro Imai 1 Graduate School of Pharmaceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan
Received November 16, 1998
A method for amino acid sequence and D/L configuration identification of peptides by using fluorogenic Edman reagent 7-[(N,N-dimethylamino)sulfonyl]-2,1,3benzoxadiazol-4-yl isothiocyanate (DBD-NCS) has been developed. This method was based on the Edman degradation principle with some modifications. A peptide or protein was coupled with DBD-NCS under basic conditions and then cyclized/cleaved to produce DBD-thiazolinone (TZ) derivative by BF 3, a Lewis acid, which could significantly suppress the amino acid racemization. The liberated DBD-TZ amino acid was hydrolyzed to DBD-thiocarbamoyl (TC) amino acid under a weakly acidic condition and then oxidized by NaNO 2/H 1 to DBD-carbamoyl (CA) amino acid which was a stable and had a strong fluorescence intensity. The individual DBD-CA amino acids were separated on a reversed-phase high-performance liquid chromatography (RP-HPLC) for amino acid sequencing and their enantiomers were resolved on a chiral stationary-phase HPLC for identifying their D/L configurations. Combination of the two HPLC systems, the amino acid sequence and D/L configuration of peptides could be determined. This method will be useful for searching D-amino-acid-containing peptides in animals. © 1999 Academic Press Key Words: amino acid sequence; D-amino-acid-containing peptide; fluorogenic reagent; Edman method; chiral separation; HPLC; oxidation.
1 To whom correspondence should be addressed. Fax: 81-3-58023339.
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Proteins and peptides in nature are generally believed to consist of L-type amino acids exclusively. The D-amino-acid-containing peptides are found only in the peptidoglycan of bacteria cell wall and some antibiotic peptides, which are synthesized by multienzyme complexes (1). However, recent studies have indicated that the ribosomal-origin peptides in animals also contain D-amino acid residues (2). Furthermore, the D-amino acids in these peptides are essential for their biological activities because the change of D-amino acid to L-amino acid would lose their biological activities (2, 3). At present, the D-amino acid at the specific position of the mature peptides is thought to be due to the conversion of L-isomer to D-isomer by a posttranslational mechanism, since the gene-encoded peptide precursors consisted of only L-amino acids (2), and some isomerase involving the conversion has recently been purified (4). Although the racemization occurs in aged human proteins such as a-crystallin and amyloid (5, 6), it is not certain whether the physiologically active peptides containing D-amino acid residue exist in mammalian tissues. To identify D-amino acid residues in peptides, a method in which an amino acid sequence as well as its stereo configuration could be determined has been developed. However, there are very few reports on such studies (7–11). The Edman degradation method, the most widely used method for N-terminal sequencing of proteins, exhibited significant racemization of amino acids during conventional Edman degradation procedures making identification of the D/L configuration difficult (12). In previous papers, we have developed an 257
258
HUANG ET AL.
Edman fluorogenic reagent containing the benzofurazan structure, 7-[(N,N-dimethylamino)sulfonyl]2,1,3-benzoxadiazol-4-yl isothiocyanate (DBD-NCS) 2 (13), and established a sequential method by utilizing DBD-thiazolinone (TZ) derivatives for amino acid identification (14). We also elucidated that the amino acid racemization occurring at the cleavage stage using trifluoroacetic acid (TFA) was due to the replacement of the proton at the a-carbon position by the acid hydrogen of TFA, and found that BF 3, a Lewis acid, could significantly suppress the racemization (15). To obtain a more stable derivative than DBD-TZ amino acid for sequential analysis, here we examine the conversion of DBD-TZ amino acid to DBD-carbamoyl (CA) amino acid mediated by DBD-thiocarbamoyl (TC) amino acid and establish a practical method for amino acid sequence and D/L configuration identification of peptides. MATERIALS AND METHODS
Material. DBD-NCS was synthesized as previously described (13). Standard D- or L-amino acids, dipeptides, insulin chain B, oxidized and (D-Ala 2)-deltorphin II were obtained from Sigma Chemical Co. (St. Louis, MO). Boron trifluoride diethyl etherate complex (BF 3 z Et 2O) obtained from Tokyo Chemical Industry Co., Ltd was distilled and stored as before (16). Pyridine and TFA were of amino acid sequencing grade (Wako Pure Chemicals, Japan). All other reagents and solvents were of analytical or guaranteed reagent grade. High-performance liquid chromatography conditions. A Hitachi HPLC system consisting of an L-6200 intelligent pump, a 7161 injector (Rheodyne, Cotati, CA), an F-1000 fluorometric detector, a D-2500 Chromato Integrator, an L-4000 UV–vis detector, and a Shimadzu CR-3A Chromatopac was used. The analytes were monitored simultaneously by fluorometric and UV–vis detectors, those connected with different recorders. Fluorometric detection was made at 520 nm with excitation at 384 nm. The UV–vis was monitored at 267 or 384 nm. The reversed phase (RP)-HPLC column was TSK gel ODS 80Ts (250 mm 3 4.6 mm i.d., Tosoh, Tokyo). For separation of individual DBD-CA amino acids, the following mobile phase was used: Eluent A, CH 3CN/H 2O (1/9) containing 0.1% TFA and eluent B, CH 3CN/H 2O (9/1) containing 0.1% TFA; gradient program: 0–2 min, B, 10%; 2 Abbreviations used: DBD-NCS, 7-[(N,N-dimethylamino)sulfonyl]2,1,3-benzoxadiazol-4-yl isothiocyanate; DBD-TZ, DBD-thiazolinone; DBD-TC, DBD-thiocarbamoyl; DBD-CA, DBD-carbamoyl; DBD-TH, DBD-thiohydantoin; RP-HPLC, reversed-phase HPLC; b-CD, b-cyclodextrin; Ph-CD, phenylcarbamoylated b-cyclodextrin; TFA, trifluoroacetic acid; TEA, triethylamine.
2–32 min, B, 10–30%; 32–37 min, B, 30–50%; 37–55 min, B, 50%. Flow rate was kept at 1.0 ml/min. The column temperature was maintained at 40°C with a 655A52 column oven (Hitachi). For isolation of DBD-TZ, DBD-TC, and DBD-CA amino acids, the mobile phase was 10 mM formic acid in CH 3CN/H 2O solutions (the ratio of CH 3CN to H 2O was changed according to the amino acids). The other HPLC conditions were the same as the above. For chiral separation of DBD-CA amino acids, the following HPLC conditions were used—Column, Ultron ES-CD (150 3 6.0 mm i.d., Shinwa Chemical Industries, Ltd., Japan) at room temperature; Eluents: A, H 2O/CH 3OH/ CH 3CN (75/5/20) containing 10 mM acetic acid; B, H 2O/ CH 3OH/CH 3CN (10/70/20) containing 10 mM acetic acid; gradient program, 0 – 8 min, B, 0 – 0%; 8 –25 min, B, 0–100%; 25–35 min, B, 100–100%; flow rate, 1.0 ml/min. Fluorometric detection was the same as described above. Preparation of DBD-TZ, DBD-TC, and DBD-CA amino acids. DBD-TZ amino acids were prepared according to the method described previously (14). DBD-TC amino acids were prepared by coupling DBD-NCS (20 mM 3 10 ml) with free amino acids in 50% pyridine at 50°C for 15 min. After washing with 200 ml of n-heptane/dichloromethane (9/1) for 3 times, the mixture was dried by a centrifugal evaporator at 50°C for 15 min. DBD-CA amino acids were prepared by oxidation of DBD-TC amino acids with NaNO 2/H 1 as described below. Hydrolysis of DBD-TZ amino acids to DBD-TC amino acids. The DBD-TZ amino acid isolated from HPLC was dried by centrifugal evaporator, dissolved in 2 ml of CH 3CN, and added with 8 ml of 0.4 M HCl. The mixture was reacted at 50°C for different times and analyzed by HPLC. Oxidation of DBD-TC amino acids to DBD-CA amino acids. The DBD-TC amino acid (200 pmol) prepared as described above was dissolved in 10 ml of CH 3CN and added with 5 ml of 4 M HCl and 5 ml of 0.5 M NaNO 2 in water and reacted at room temperature(about 22°C) for different times. The reaction mixture was then neutralized with 1.0 M NaHCO 3 to stop the reaction and an aliquot was analyzed immediately by HPLC. Identification of DBD-CA amino acids. An M-1200H LC/APCI-MS (Hitachi) system connected to a Rheodyne injector and an intelligent pump (L-6200, Hitachi) was used. The correspondence of DBD-CA amino acid prepared as described above was collected by HPLC and then dried by centrifugal evaporator. The residue was dissolved with methanol and analyzed by LC–MS. The mass spectrometer was operated in the APCI mode with a drift, focus, and multiplier voltages respectively set at 20, 120, and 1800 V. The temperature of the vaporizer and dessolvation regions were set at 200 and 400°C, respectively. The flow injection analysis (FIA) mode was
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
259
FIG. 1. Scheme of Edman method.
used at a flow rate of 1.0 ml/min of methanol. Intelligent fluorometric spectrometer (F-4010, Hitachi) and UV–vis spectrophotometer (Ubset 50, JASCO) were used for fluorometric and UV–vis spectra measurements, respectively. Effects of pH and solvents on the fluorescence intensity and the stability of DBD-CA amino acid. DBDCA-Leu collected by HPLC was diluted with 0.1 M phosphate buffers of different pH or various solvents such as methanol, ethanol, acetonitrile, and water. The fluorescence spectrum of the solution was immediately measured on the fluorescence spectrophotometer. To examine the stability of DBD-CA amino acid, the DBDCA-Leu collected by HPLC was divided into several equal parts and dried by centrifugal evaporator; the residue was dissolved in 0.1 M phosphate buffers of different pH or solvents such as methanol, acetonitrile, and water, and the solution was kept at room temperature. An aliquot was withdrawn at appropriate time
intervals and determined by HPLC. The degradation of DBD-CA-Leu was obeyed pseudo-first-order kinetics, ln[A] 5 2k obst 1 C, where [A] is the residual content of DBD-CA-Leu, while t and C are the time and intercept values, respectively. The k obs is the observed degradation rate constant. Procedures for manual sequencing of peptides. A peptide (insulin chain B (500 pmol) or deltorphin II (2 nmol) ) dissolved in 20 ml of 50% pyridine/water solution was added with 5 ml 1% triethylamine (TEA) in acetonitrile and 10 ml of 20 mM DBD-NCS in pyridine. The mixture was reacted in an inert atmosphere at 50°C for 15 min. After the coupling reaction, the mixture was washed with 200 ml of n-heptane/dichloromethane (6/4) three times and dried by a centrifugal evaporator at 50°C for 15 min. The DBD-thiocarbamoyl-peptide was cyclized/cleaved by adding 30 ml of 1% BF 3 z Et 2O in acetonitrile and heated at 50°C for 5 min, and then dried by a stream of N 2. The residue was
260
HUANG ET AL.
for 5 min for hydrolysis. The reaction mixture was then oxidized at room temperature (22°C) for 10 min after adding 5 ml of 4 M HCl and 5 ml of 0.5 M NaNO 2 in water. The reaction was stopped by neutralizing with 1.0 M NaHCO 3 (23 ml) and added with 20 ml of 0.15 M methionine for reducing the excessive oxidizing agent. An aliquot (20 ml) of the mixture was subjected to HPLC for analysis. RESULTS AND DISCUSSION
FIG. 2. Time course of hydrolysis of DBD-TZ amino acids to DBD-TC amino acids. Reaction conditions: DBD-TZ amino acid in acetonitrile (2 ml) was added with 0.4 M HCl ( 8 ml) and reacted at 50°C for different times.
added with 20 ml of water and extracted two times with 100 ml of benzene/ethyl acetate (1/4) for insulin chain B or n-hepatane/ethyl acetate (1/1) for deltorphin II. The aqueous phase containing residual peptide was dried by centrifugal evaporator and subjected to the next cycle. The combined organic phase was dried by N 2 and dissolved with 2 ml of CH 3CN, and added with 8 ml of 0.4 M aqueous HCl. The mixture was heated at 50°C
The conventional Edman degradation method was carried out by three steps as shown in Fig. 1 (17). A peptide or protein was first coupled with arylisothiocynate (aryl:phenyl or DBD-) to form the thiocarbamoyl peptide, which was then cyclized/cleaved to produce the TZ derivative by strong acid. The released TZ amino acid was further converted into a stable thiohydantoin (TH) amino acid, which was identified by HPLC. As previously reported, the DBD-TH amino acids lack the fluorescence to obtain a highly sensitive detection; one way is to detect DBD-TZ amino acids directly, omitting the conversion step (14). The other is to convert the DBD-TZ derivative to another derivative. Edman previously observed that the phenylthiocarbamoyl moiety showed a tendency toward oxidative desulfuration with formation of phenyl-CA amino acids (18). We found that DBD-CA amino acids were stable and had a strong fluorescence intensity, although DBD-TC amino acids afforded very little fluorescence intensity. We therefore attempted the conversion of
FIG. 3. Time course of oxidation of DBD-TC amino acids to DBD-CA amino acids. Reaction conditions: DBD-TC amino acid in acetonitrile (10 ml) was added with 4 M HCl (5 ml) and 0.5 M NaNO 2 (5 ml) and reacted at room temperature for different times.
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
261
FIG. 4. Transformation of DBD-TZ-Leu into DBD-CA-Leu by two steps. DBD-TZ-Leu (A) was hydrolyzed to DBD-TC-Leu (B) and further oxidized to DBD-CA-Leu (C). Hydrolysis conditions: DBD-TZ-Leu in CH 3CN (2 ml) was added with 0.4 M HCl (8 ml) and reacted at 50°C for 5 min. Oxidative conditions: The above hydrolyzed mixture was further added with 4 M HCl (5 ml) and 0.5 M NaNO 2 (5 ml) and reacted at room temperature for 10 min. HPLC conditions: Column, TSK gel ODS 80Ts (250 3 4.6 mm i.d.) at room temperature; eluents, CH 3CN/H 2O (55/45) containing 10 mM formic acid; flow rate, 1.0 ml/min. Detected simultaneously by fluorescence (upper panel) and UV (lower panel) with wavelength at Ex 5 384 nm; Em 5 520 and 267 nm, respectively.
DBD-TZ amino acids to DBD-CA amino acids for sequential analysis. Hydrolysis of DBD-TZ Amino Acids to DBD-TC Amino Acids DBD-TZ amino acids were unstable in protic solvents and easily hydrolyzed to DBD-TC amino acids under weakly acidic conditions. The reaction rate was dependent on the ratio of CH 3CN and water. The lower CH 3CN concentration favored a higher rate of hydrolysis (14). Figure 2 shows the time course of hydrolysis of several DBD-TZ amino acids (Leu, Phe, Arg, and Asn) compared to their corresponding DBD-TC amino acids under a 2/8 CH 3CN/0.4 M aqueous HCl solution at 50°C. The reaction was completed within 5 min. Ten minutes were necessary to complete hydrolysis when the CH 3CN/0.4 M aqueous HCl solution ratio was 5/5
at 50°C (data not shown). Under the above hydrolysis conditions, DBD-TZ amino acids could be quantitatively transformed to DBD-TC amino acids, whereas no DBD-TH amino acids were formed. Oxidation of DBD-TC Amino Acids to DBD-CA Amino Acids The transformation of thiocarbonyl compounds to their corresponding oxo-analogues has been reported by utilizing NaNO 2/H 1 (19), H 2O 2 (20), MnO 2 (21), m-chloroperbenzoic acid (22), etc. To investigate the oxidative desulfuration of DBD-TC amino acid to DBD-CA amino acid, various oxidizing agents were examined. We found that DBD-TC-Leu could be rapidly and specifically oxidized to DBD-CA-Leu by NaNO 2/H 1 as shown in Figs. 3, 4B, and 4C. The formed DBD-CA-Leu was identified by LC–MS and UV spec-
262
HUANG ET AL.
FIG. 5. UV spectra of DBD-TC-Leu (A) and DBD-CA-Leu (B) in acetonitrile.
tra. The mass number (M 1 H) 1 of the peak at Rt of 6.2 min in Fig. 4C was 400 which corresponds to the molecular weight of DBD-CA-Leu (399). The DBD-CA-Leu has an absorption at 244 nm, whereas it loses the absorption at around 267 nm (Fig. 5B), which is similar to that of phenylcarbamoyl amino acid as previously reported by Edman (18). The produced DBD-CA derivatives of other amino acids by NaNO 2/H 1 oxidation were also identified by LC–MS and UV spectra. The time course of transformation of several DBD-TC amino acids to their corresponding DBD-CA amino acids is shown in Fig. 3. The reactions for all the amino acids were completed within 5 min except Lys, for which the maximal yield was reached at 30 min. This might be due to Lys being labeled by two thiocarbamoyl groups at a and e positions; both were desulfurated by oxidation. Most DBD-CA amino acids formed were stable in NaNO 2/H 1 medium for 60 min except Tyr, Met, and Trp, whose side chains were sensitive to NaNO 2/H 1 oxidation. Tyrosine was further nitrated at the 3 position of hydroxyphenyl group if the reaction time was longer. From these results, we decided that 10 min was an optimal time in which the nitration of Tyr was less than 15%. Methionine was very sensitive to oxidation, and most of the methionine was oxidized into DBD-CA-methionine sulfoxide under the above conditions. No further oxidation to DBD-CA-methio-
nine sulfone occurred even after 1 h. Therefore, DBDCA-methionine sulfoxide was used for identifying the methionine residue during sequencing. Tryptophan was also sensitive to oxidation. The oxidation of DBDTC-Trp by NaNO 2/H 1 produced two major fluorescence peaks, which presumably correspond to DBD-CA-oxindole derivatives (23–25). DBD-CA-Trp has little fluorescence intensity due to fluorescence quenching of two fluorophores of benzofurazan structure and aminoindole moiety of Trp. The oxidation of the Trp group makes it detectable by fluorescence, although its intensity is several times lower than that of other DBD-CA AAs. DBD-TC-Pro was difficult to oxidize to DBD-CAPro under oxidative conditions, instead, facilely converting to another fluorescence derivative, which showed a longer wavelength than that of the DBD-CA derivatives. The former’s maximal excitation and emission in CH 3CN/H 2O solution were at 440 and 575 nm, respectively, while the latter’s were at 384 and 520 nm, respectively. Proline was a secondary amine, which differed from other amino acids. Considering that the facile dimerization of thiocarbamates or thioureas by oxidation was previously reported (26, 27), we speculated that DBD-TC-Pro might form a dimeric derivative through disulfide linkage under the oxidation. It was convenient that the resulting hydrolysis samples could directly be followed by oxidative procedures
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
263
under the same chromatographic conditions, indicating that the DBD-CA amino acid was superior to the DBD-TZ amino acid for detection.
TABLE 1
Stability of DBD-CA-Leu in Different Solvents at Room Temperature Solvent
K obs (h 21)
t 0.9 (h)*
t 0.5 (h) a
H 2O Methanol Acetonitrile 0.1% TFA,CH 3CN/H 2O (1/9)
0.0073 0.0037 0.0023 0.0012
14.4 28.4 45.7 87.5
94.9 187.3 301.3 577.5
a
Time at which the residual amount of the derivative is 90 or 50% of the initial amount. K obs, observed degradation rate constant as described under Materials and Methods.
without any treatment because both hydrolysis and oxidation reactions were performed in HCl solutions. The quantitative transformation of DBD-TZ-Leu to DBD-CA-Leu by two-step reactions is shown in Fig. 4. No byproducts were observed in the chromatograms. The fluorescence intensity (peak area) of DBD-CA-Leu was about 3 times higher than that of DBD-TZ-Leu
Effects of pH and Solvents on the Fluorescence Intensity and the Stability of DBD-CA Amino Acids DBD-CA amino acids showed a stronger fluorescence intensity under acidic conditions and in organic solvents. The fluorescence intensity of DBD-CA-Leu in organic solvents such as CH 3CN, MeOH, and EtOH was about 7–11 times higher than that in H 2O, while the fluorescence intensity of DBD-CA-Leu in 0.1 M phosphate buffer at pH 2.0 was 1.3 times higher than that at pH 8.5. In comparison with DBD-TZ amino acids (14), DBD-CA amino acids were relatively stable in various solvents (Table 1). The k obs of DBD-CA-Leu in 0.1 M phosphate buffer at pH 2.0, 4.0, and 8.5 was 0.004, 0.009, and 0.014 (h 21), respectively, which indicated that the DBD-CA derivative was more stable under
FIG. 6. Separation of DBD-CA amino acids on RP-HPLC. Peaks: 1, cysteic acid, 2, His, 3, Asn, 4, Gln, 5, Arg, 6, Met-O, 7, Ser, 8, Gly, 9, Asp, 10, Glu, 11, Thr, 12, Ala, 13, Tyr, 14, Val, 15, Pro-X, 16, Trp-X, 17, Ile, 18, Leu, 19, Phe, 20, Lys. Met-O, Met-sulfoxide; Pro-X, unidentified DBD-TC-Pro oxidative; Trp-X, unidentified DBD-TC-Trp oxidative. *The shoulder peak appearing just before Val was a degradation product of DBD-NCS formed in coupling reaction. The amounts of these derivatives corresponding to each amino acids were 20 pmol except Trp which was 50 pmol. HPLC conditions: Column, TSK gel ODS 80Ts (250 3 4.6 mm i.d.) at 40°C. Eluents: A, CH 3CN/H 2O (10/90) containing 0.1% TFA; B, CH 3CN/H 2O (90/10) containing 0.1% TFA; Gradient program, 0 –2 min, B, 10 –10%, 2–32 min, B, 10 –30%, 32–37 min, B, 30 –50%, 37–50 min, B, 50 –50%; flow rate, 1.0 ml/min. Ex 5 384 nm; Em 5 520 nm.
264
HUANG ET AL.
acidic conditions. The DBD-CA derivatives of other amino acids also showed a stability similar to that of DBD-CA-Leu. These results revealed that DBD-CA amino acids were suitable for detection under various chromatographic conditions.
TABLE 2
Chiral Separation of DBD-CA Amino Acids on Native b-CD Column k9 DBD-CA AA
Separation of DBD-CA Amino Acids on RP-HPLC Since the fluorescence intensity and the stability of DBD-CA amino acids were favored under acidic conditions, we selected CH 3CN/H 2O containing 0.1% TFA as a mobile phase. Under acidic conditions, the ionization of the COOH group was less than that under neutral conditions, which resulted in increasing the retention of DBD-CA amino acid on a column because it decreased its polarity and, alternatively, increased the fluorescence intensity since a higher organic solvent ratio in solution was needed to elute the analytes. All DBD-CA amino acids could be separated on RP-HPLC under the conditions described (Fig. 6). The column temperature kept at 40°C was necessary for the separations of DBD-CA-Gly, DBD-CA-Asp, and DBD-CAThr. This was also of benefit for reproducing peak retention times. Under acidic conditions, the fluorescence intensity of Pro derivative at Ex 5 384 nm with Em 5 520 nm was 2.2 times lower than at its optimal wavelength of Ex 5 440 nm with Em 5 575 nm. Still the Pro derivative had an intensity similar to other DBD-CA AAs at Ex 5 384 nm with Em 5 520 nm. The detection limits for DBD-CA amino acids under HPLC conditions were 100 fmol to subpicomole (signal/noise 5 3). Enantiomeric Separation of DBD-CA Amino Acids on Chiral Stationary Phases To separate DBD-CA amino acid enantiomers, we examined various types chiral HPLC columns and found that DBD-CA amino acid enantiomers were well separated on a native b-cyclodextrin (CD)-bonded chiral stationary phase with an acidic mobile phase. The mobile phase for the b-CD series chiral columns were usually carried out by water with organic modifiers (16, 28). Although most of the DBD-CA amino acid enantiomers could also be resolved on native b-CD or phenylcarbamoylated b-CD (Ph-CD) or ES-1/4 Ph-CD columns (16) with neutral mobile phases such as ammonium acetate/MeOH (data not shown), the peaks were broad and one or two DBD-CA amino acid enantiomers could not be separated under these conditions. Furthermore, the neutral eluents would decrease DBD-CA amino acid retention on columns compared with acidic eluents. We therefore chose MeOH/H 2O containing 10 mM acetic acid as a mobile phase. The separation factors for DBD-CA amino acid enanti-
Arg His Ala Val Ser Ile Gly Leu Asn Thr Glu Met Gln Asp Phe Lys Met-O a
D
L
1.35 1.94 8.56 8.64 8.73 8.89
a
1.64 2.16 8.96 9.01 9.27 9.26
1.21 1.11 1.05 1.04 1.06 1.04
9.63 9.92 10.19 10.42 10.49 10.56 11.37 11.64 12.04 13.19 13.49
1.02 1.05 1.04 1.04 1.04 1.05 1.03 1.04 1.04 1.06 1.06
9.42 9.47 9.48 9.76 10.01 10.05 10.08 11.00 11.15 11.59 12.40 12.72
Note. k9 and a are capacity factor and separation factor, respectively. HPLC conditions: Column, Ultron ES-CD(150 3 6.0 mm i.d.) at room temperature; Eluents: A, H 2O/MeOH/CH 3CN (75/5/20) containing 10 mM acetic acid; B, H 2O/MeOH/CH 3CN (10/70/20) containing 10 mM acetic acid; Gradient program, 0 – 8 min, B, 0 – 0%; 8 –25 min, B, 0 –100%; 25–35 min, B, 100 –100%; Flow rate, 1.0 ml/min. Ex 5 384 nm; Em 5 520 nm. a Met-O, Met-sulfoxide, which has two asymmetry centers at a-C and d-S.
omers did not decrease with this eluent, while it improved DBD-CA amino acid detection limits. In acidic mobile phase, DBD-CA amino acids revealed longer retention on native b-CD column than that on Ph-CD and ES-1/4 Ph-CD columns; we therefore selected native b-CD column for separation of DBD-CA amino acids. In the native b-CD column with MeOH/H 2O containing 10 mM acetic acid as mobile phase, DBD-CA-Leu enantiomers could only be partially resolved. When acetonitrile was added into the mobile phase, however, not only was the chiral separation of DBD-CA-Leu improved but also DBD-CA amino acid peaks were sharpened. The ionized forms of DBD-CA-His and DBD-CA-Arg under acidic conditions make them become eluted much faster than other amino acid derivatives. To overcome this problem, a gradient elution was performed, which also sharpened the peaks. Table 2 lists the capacity factors and separation factors for each DBD-CA amino acid under the established HPLC conditions. DBD-CA-Met-sulfoxide had two asymmetry centers at a-C and d-S. All the DBD-CA amino acids eluted D form faster than L type. The enantiomeric
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
PEPTIDES
265
FIG. 7. Chromatograms of manual sequencing of insulin chain B, oxidized (500 pmol). HPLC conditions were the same as those described in the legend to Fig. 6.
resolutions of all amino acids except Pro and Trp were achieved; however, the separations of individual DBD-CA amino acids were not successful. The detection limits for DBD-CA amino acids under the chiral separation conditions were at subpicomole to picomole levels (signal/noise 5 3).
Amino Acid Sequencing of Peptides A manual procedure described under Materials and Methods was applied to sequencing of peptides. The coupling condition was modified by further addition with 1% TEA to the reaction mixture to increase pH,
266
HUANG ET AL.
FIG. 8. Amino acid sequencing and D/L configuration identification of (D-Ala 2)-deltorphin II. (D-Ala 2)-deltorphin II (2 nmol), Tyr-D-Ala-PheGlu-Val-Val-Gly-amide. *The peaks at Rt of 26.5 and 27.5 min of cycle 1 were DBD-CA-(D)-3-nitro-Tyr and DBD-CA-(L)-3-nitro-Tyr, respectively. HPLC conditions were the same as those described in Table 2.
which was necessary to complete the coupling reaction, although it also increased byproduct formation (14). The DBD-TC peptides were cyclized/cleaved by BF 3 instead of TFA, to suppress the racemization as reported previously (15). The released DBD-TZ amino acid was further converted into DBD-CA amino acid by hydrolysis and oxidative reactions as described above. Figure 7 shows the chromatograms of the results of first six cycles obtained from the sequencing of insulin chain B (500 pmol). The amino acid residues were easily identified due to few interfering peaks appearing on the chromatograms. The major byproduct peak of DBD-NH 2 formed during coupling reaction disappeared on the chromatograms because it was destroyed by NaNO 2/H 1 oxidation and could not be detected fluorometrically. The amino acid residue in the former cycle was not observed in the next cycle indicated that the coupling was completed. The present sequential results are comparable to the other fluorescence Edman reagents such as fluorescein isothiocyanate (29) and dansylamino isothiocyanate (30). Amino Acid Sequencing and D/L Identification of Peptide The manual sequencing procedures for D-amino-acidcontaining peptides were the same as those described above. Figure 8 shows the first four cycles for amino acid sequencing and D/L identification of (D-Ala 2)-deltorphin II (2 nmol), which was a neuroactive D-aminoacid-containing peptide analogous to the dermorphin family (31). Although amino acid racemizations were
partially occurring at the proposed method, the D or L configuration of each amino acid residue could still be discriminated. The nitration of tyrosine by NaNO 2/H 1 oxidation at the first cycle was less than 15%. The second amino acid residue of the peptide was easily identified as D-Ala. The partial racemization was thought to occur at the hydrolysis step because no or little racemization occurred at the oxidation and cleavage steps, respectively (15). The opening of the TZ ring during the hydrolysis in HCl solution would cause some racemizations. Because DBD-CA amino acid enantiomers were eluted closely with exclusively D-type amino acids faster than the L-type under the HPLC conditions, the partial racemizations would produce asymmetric dual peaks on the chromatogram, in which a lower former peak compared to the latter peak would indicate an L-amino acid residue and otherwise a D-amino acid residue. Since peptides consist mainly of L-amino acids, the occurrence of a D-amino acid residue in sequence was easily identified by the chromatograms. Under HPLC conditions, the individual DBD-CA amino acid enantiomers have not been completely separated in a single run; the sequencing of amino acid and D/L configuration identification therefore needed to be performed under two HPLC conditions. One portion of a sample was first applied to RP-HPLC for amino acid sequencing; the other part of the sample was applied for chiral column HPLC system for D/L configuration identification. The amino acid residues could be reconfirmed by the chiral stationary-phase HPLC system.
SEQUENTIAL ANALYSIS OF
D-AMINO-ACID-CONTAINING
CONCLUSIONS
A method for amino acid sequence and D/L configuration determination using the fluorogenic Edman reagent DBD-NCS has been developed. This method was a modified Edman degradation procedure by conversion of DBD-TZ amino acids to DBD-CA amino acids instead to DBD-TH amino acids for amino acid identification. The facile conversion of DBD-TZ amino acid to DBD-CA amino acid and the stable and fluorescence properties of DBD-CA amino acid indicated that it was suitable for sequential analysis. Although the exact structures of the oxidative products of DBD-TC-Pro and DBD-TC-Trp have not been elucidated as yet, these two amino acid residues could still be identified by the present method. The characteristics of the new procedure were that there were few interfering peaks on chromatograms, which makes amino acid identification easy. Despite a partial racemization occurring at the hydrolysis step, the D/L configurations of amino acid residue could still be identified.The proposed method will be useful for searching the D-amino acid in animal peptides. Further studies are currently in progress to solve the remaining problems and to automate these procedures. ACKNOWLEDGMENTS The authors thank Shinwa Chemical Industries, Ltd. for the generous gifts of b-CD series chiral columns and Tosoh Co. for supplying TSK gel ODS 80Ts column. This work was financially supported by the Japan Society for Promotion of Sciences and a Grant-in-Aid for Science Research from the Ministry of Education, Science, and Culture of Japan. Y. Huang is a recipient of a foreign postdoctoral fellowship from the Japan Society for Promotion of Sciences.
REFERENCES 1. Kleinkauf, H., and von Dohren, H. (1996) Eur. J. Biochem. 236, 335–351. 2. Kreil, G. (1997) Annu. Rev. Biochem. 66, 337–345. 3. Montecucchi, P. C., de Castiglione, R., Piani, S., Gozzini, L., and Erspamer, V. (1981) Int. J. Peptide Protein Res. 17, 275–283. 4. Shikata, Y., Watanabe, T., Teramoto, T., Inoue,A., Kawakami, Y., Nishizawa, Y., Katayama, K., and Kuwada, M. (1995) J. Biol. Chem. 270, 16719 –16723. 5. Fujii, N., Muraoka, S., Satoh, K., Hori, H., and Harada, K. (1991) Biomed. Res. 12, 315–321. 6. Roher, A. E., Lowenson, J. D.,Clarke, S., Wolkow, C., Wang, R., Cotter, R. J., Reardon, I. M., Zu¨rcher-Neely, H. A., Heinrikson, R. L., Ball., M. J., and Greenberg, B. D. (1993) J. Biol. Chem. 268, 3072–3083.
PEPTIDES
267
7. Scaloni, A., Simmaco, M., and Bossa, F. (1991) Anal.Biochem. 197, 305–310. 8. Toyo’oka, T., Suzuki, T., Watanabe, T., and Liu, Y. M. (1996) Anal. Sci. 12, 779 –782. 9. Suzuki, T., Watanabe, T., and Toyo’oka, T. (1997) Anal. Chim. Acta 352, 357–363. 10. Kurosu, Y., Murayama, K., Shindo, N., Shisa, Y., Satou, Y., and Ishioka, N. (1996) J. Chromatogr. 752, 279 –286. 11. Kurosu, Y., Murayama, K., Shindo, N., Shisa, Y., Satou, Y., and Ishioka, N. (1997) J. Chromatogr. 771, 311–317. 12. Edman, P. (1950) Acta Chem. Scand. 4, 283–293. 13. Uzu, S., Imai, K., Nakashima, K., and Akiyama, S. (1993) Biomed. Chromatogr. 9, 152–153. 14. Matsunaga, H., Santa, T., Hagiwara, K., Homma, H., Imai, K., Uzu, S., Nakashima, K., and Akiyama, S. (1995) Anal. Chem. 67, 4276 – 4282. 15. Matsunaga, H., Santa, T., Iida, H., Fukushima, T., Homma, H., and Imai, K. (1996) Anal. Chem. 68, 2850 –2856. 16. Iida, T., Matsunaga, H., Fukushima, T., Santa, T., Homma, H., and Imai, K. (1997) Anal. Chem. 69, 4463– 4468. 17. Edman, P. (1956) Acta Chem. Scand. 10, 761–768. 18. Ilse, D., and Edman, P. (1963) Aust. J. Chem. 16, 411– 416. 19. Jorgensen, K. A., Ghattas, A.-B. A. G., and Lawesson, S.-O. (1982) Tetrahedron 38, 1163–1168. 20. Walter, W., and Randau, G. (1969) Liebigs Ann. Chem., 722, 57–72. 21. Sharma, T. C., Sachni, N. S., and Lal, A. (1978) Bull. Chem. Soc. Jpn. 51, 1245–1246. 22. Kochar, K. S., Cottrell, D. A., and Pinnick, H. W. (1983) Tetrahedron Lett. 24, 1323–1326. 23. Kato,Y., Kawakishi, S., Aoki, T., Itakura, K., and Osawa, T. (1997) Biochem. Biophy. Res. Commun. 234, 82– 84. 24. Watanabe, N., Toyo’oka, T., and Imai, K. (1987) Biomed. Chromatogr. 2, 99 –103. 25. Imai, K., Ueda, E., and Toyo’oka, T. (1988) Anal. Chim. Acta 205, 7–14. 26. Blankespoor, R. L., Doyle, M. P. Hedstrand, D. M. Tamblyn, W. H., and Van Dyke, D. A. (1981) J. Am. Chem. Soc. 103, 7096 –7101. 27. Olah, G. A., Arvanghi, M., Ohannesian, L. and Surya Prakash, G. K. (1984) Synthesis 1984, 785–786. 28. Rizzi, A. M., Cladrowa-Runge, S., Jonsson, H., and Osla, S. (1995) J. Chromatogr. 710, 287–295. 29. Muramoto, K., Kamiya, H., and Kawauchi, H. (1984) Anal. Biochem. 141, 446 – 450. 30. Hirano, H., and Wittmann-Liebold, B. (1986) Biol. Chem. 367, 1259 –1265. 31. Erspamer, V., Melchiorri, P., Falconieri-Erspamer, G., Negri, L., Corsi, R., Severini, C., Barra, D., Simmmaco, M., and Kreil, G. (1989) Proc. Natl. Acad. Sci. USA 86, 5188 –5192.