Separation by high-performance liquid chromatography of (3R)- and (3S)-β-leucine as diastereomeric derivatives

ANALYTICAL

BIOCHEMISTRY

Separation

151,88-91

(1985)

by High-Performance Liquid Chromatography (3S)-P-Leucine as Diastereomeric Derivatives’

D. JOHN ABERHART,~ JACQUES-ALAIN COTTING,AND

of (3R)- and

HORNG-JAU LINT

Worcester Foundation for Experimental Biology, Shrewsbury, Massachusetts 01545 Received July 8, 1985 For studies on the coenzyme B,,-dependent enzyme, leucine-2,3-aminomutase, (3R)- and (3S)&leucines were synthesized. The 10-camphorsulfonamide pnitrobenzyl esters could be resolved by normal-phase HPLC. A much better separation was obtained by reversed-phase HPLC of the diastereomeric derivatives obtained by treatment of.-leucine with Marfey’s reagent (N*-(5-fluoro2,4dinitrophenyl)+alaninamide). Q 1985 Academic press, Inc. KEY WORDS: 6-leucine; HPLC techniques; Marfey’s reagent; leucine 2,3-aminomutase.

Nearly 10 years ago, Poston first reported the detection of the interconversion of a- and @-leucine by a coenzyme Bi2-dependent enzyme named leucine 2,3-aminomutase (1). The presence of this enzyme was first detected in Clostridia (where it was most active), in the livers of several animal species, and in human leukocytes (1) and hair roots (2). The enzyme was also found to be present in bean seedlings (3), potatoes (4), a yeast (5), and various animal organs (6,7). The metabolism in various rat tissues of L-leucine via ,&leucine (p-keto pathway) was found to be ca. l-5% of the total except in testicular tissue where the &keto pathway constituted ca. one-third of the total (8). Overton et al. (9) have also detected leutine 2,3-aminomutase activity in tissue cultures of Andrographis paniculata, but their enzyme was not found to be coenzyme B12dependent. For studies on the steric course of leucine 2,3-aminomutase, we required a method of separating (3R)- and (3S)-/3-leucine. We now report two methods for the separation by HPLC of these enantiomers as diastereomeric derivatives.

MATERIALS

Equipment. A Waters Associates (Milford, Mass.) liquid chromatograph was used, with a Model 6000A solvent delivery system, Model 450 variable-wavelength detector, and a Model U6K injector. The columns used are indicated in the sections which follow. All solvents used were filtered through 0.45-pm filters and degassed by pumping. NMR spectra were run in CDCls solutions in 5-mm tubes with internal TMS4 (6 = 0) as reference using a Varian EM360 instrument. Chemicals. [RhCl(COD)]2 was obtained from Strem Chemicals, Inc., Newburyport, Massachusetts. (-)DIOP and BPPM were obtained from Reaction Design Corporation, Hillside, New Jersey. Marfey’s reagent (N2-(5fluoro-2,4-dinitrophenyl)-L-alaninamide) was obtained from Pierce Chemical Company, Rockford, Illinois. All other organic chemicals were obtained from Aldrich Chemical Company, Milwaukee, Wisconsin. 4 Abbreviations used: Eu(htb&, tris[3-(heptafluoropropylhydroxymethylene)-d-camphoratoleuropium (III) derivative; [RhCI(COD)],, chloro(l,5-cycIooctadiene) rhodium (I) dimer; (-)DIOP, (-)-2,3-o-ipropylidene-2,3-dihydroxy- 1,4-bis(diphenylphosphino)butane; BPPM, (2$4S)-I-t-butoxycarbonyl-2diphenylphosphino methyl-4-diphenylphosphinopyrrolidine; AspcHex, L-aspartylcyclohexylamide; TMS, tetramethylsilane.

’ This investigation was supported by Grant GM 259 19 from the National Institutes of Health. 2 To whom correspondence should be addressed. 3 Present address: Chemical Division, PPG Industries, Barberton, Ohio 44203. 0003-2697/85 $3.00 Copyright 0 1985 by Academic Press. Inc. All rights of reproduction in any form reserved.

AND METHODS

88

CHROMATOGRAPHIC

SEPARATION

Synthesis of (3RS)-fl-leucine. This was prepared by treatment of 4-methyl-2-pentenoic acid with ammonia as reported by Poston (I), except that the reaction was run in a 450-ml Parr stirred pressure reactor, instead of in sealed glass bottles or tubes which occasionally exploded when heated. Synthesis of (3R)- and (-IS)-P-leucine. (3R)fl-Leucine was synthesized from L-valine essentially by following the method of Balenovic and Dvornik (IO). The optical purity of the product was shown as follows: a portion of the product was acetylated (acetic anhydrideacetic acid, reflux 30 min). After evaporation of the solvents, the residue was treated with an excess of diazoethane (11) in ether. The NMR spectrum (CDC& solution) of the resultant N-acetate ethyl ester, in the presence of Eu(hfbch (ca. 0.5 eq) showed a single Nacetyl resonance. In contrast, (3RS)-/I-leucine N-acetate ethyl ester shows two well-separated IV-acetyl resonances in the region ofS 5-7 (depending on how much Eu(hfbcb is used). (The IV-acetyl of the (3R)-P-leucine derivative appears at higher field than that of the (3s) isomer). (3S)$Leucine was synthesized from D-V& line in an analogous manner. An attempt to synthesize chiral P-leucines by asymmetric hydrogenation failed. Hydrogenation of ethyl [Z]3-acetamido-4-methyl-2pentenoate ( 12) in benzene-ethanol solution in the presence of catalysts formed in situ either from [RhCl(COD)]2 plus (-)DIOP (13) or from [RhCl(COD)], plus BPPM (14) gave a nearly equal mixture of the (3R) and (3s) products. Attempted HPLC separations of (3RS)-pleucine by use of chiral cosolvents. (a) (3RS)/3-Leucine dansyl derivative (15) was run on a PBondapak C8 column (Waters Associates) using as eluant 0.004 M Na(OAc),/0.004 M Lproline octylamide (16)/0.088 M acetic acid, pH 8.5, in methanol (60%)-H20 (40%). The P-leucine dansyl derivative was not resolved, whereas the dansyl derivative of DL-a-leucine was well separated. (b) Underivatized (3RS)fi-leucine was chromatographed on a PBondapak Cl8 column, using as eluant 1 mM

89

OF (3RS)-&LEUCINE

AspcHex (17) plus 1 mM CuSO4 in H20. Whereas DL-a-leucine was separable by this method, (3RS)+leucine was not resolved. HPLC of (3RS)-fi-leucine N-d-lO-camphorsulfonamide p-nitrobenzyl ester, 1. The derivative was prepared from (3RS)-P-leucine according to the general method of Furukawa and co-workers ( 18,19). A portion of the crude product was purified by preparative TLC (30% ethyl acetate-hexane; silica gel HF 254 + 366). NMR (CDC&) 6 0.93 (6H, s), 1.05 (6H, s) (camphor methyl singlets superimposed on isopropyl doublet), 1.2-3.9 (13H, m), 5.18 (lH,s),5.23(1H,S), 5.58(lH,m,-NH),7.49 and 8.16 (2H, AB, JAB = 8 Hz). A 1% solution of 1 was prepared in 30% ethyl acetate-hexane and filtered through an 0.45 pm Gelman Acre LC-13 filter. For the chromatogram in Fig. 1, a 1O-p.1injection was made. The eluting solvent was 15% ethyl acetate-isooctane at 2.0 ml/min. A 3.9-mm X 30-cm PPorasil column (Waters Associates) was used. The detector was set at 254 nm, 1.O AUFS. The (3R)-fl-leucine derivative was eluted first as shown by separate injections of the derivative prepared from optically pure (3R)+leucine and (3S)-P-leucine. HPLC of (3RS)-P-leucine derivatized with Marfey’s reagent. (3RS)-/3-Leucine (0.7 mg, 0.5 Fmol) was treated with 20 ~1 of a 1% solution of Marfey’s reagent (20,2 1) in acetone, plus 40 ~1 of 1 M NaHCOj at 40°C for 1 h. After cooling to 25°C and addition of 20 ~12 M HCl, the solution was filtered through a 3R 9

3s 1

b

I

;

;

l'0

1'5

2;

2;

WIN

FIG. 1. HPLC chromatograms of DL-p-k?UCine N-d- locamphorsulfonamide pnitrobenzyl ester. (conditions, see Materials and Methods).

90

ABERHART,

COTTING,

AND LIN

I

1

,

I

I

I

I

I

0

5

10

15

20

25

30

35

,

40 MIN

FIG. 2. HPLC chromatogram of DL-fl-leucine derivatized with Marfey’s reagent (conditions, see Materials and Methods). The elution times of (ZS)-cu-leucine (L-leucine) and (2R)-cy-leucine (D-leucine) under the same conditions are also shown on the chromatogram.

Gelman Acre LC-13 filter and injected directly. For the chromatogram in Fig. 2, a lo~1 injection was made. The eluting solvent was 30% acetonitrile-0.05 M triethylamine phosphate, pH 3.0, at 2.0 ml/min. A 3.9-mm X 15cm Nova-Pak C i8 column (Waters Associates) was used. The detector was set at 340 nm, 0.4 AUFS. RESULTS AND DISCUSSION

(3R)- and (3S)-/3-leucine have been previously separated by gas chromatography as the methyl ester (-)camphanamide derivatives (9). For our work, however, we preferred to develop an HPLC method for separating these enantiomers, due (in part) to the inaccessibility of radio-GC equipment and also in the expectation that larger quantities of labeled pleucine could be separated by HPLC. Since at the outset, we did not wish to attempt to develop a new reagent specifically for this purpose, our efforts were restricted to a limited examination of several methods reported to be useful for the HPLC resolution of DL-cramino acids. Initially, we examined two methods ( 16,17) which involved the use of chiral cosolvents (see Materials and Methods for details). Although DL-cY-leucine was separated by these methods as reported, no separation of (3RS)-/3-leucine was observed.

On the other hand, two methods were found for the separation of (3RS)-@-leucine as diastereomers. First, the 1O-camphorsulfonamide pnitrobenzyl esters (1, Scheme 1) of &leucine were resolved by normal-phase HPLC (l&19), Fig. 1. Unfortunately the method is of limited utility, due to the difficulty of preparing this derivative on a small scale and to the marginal extent of separation observed. However, use of the recently reported Marfey’s reagent (20,21), N*-(5-fluoro-2,4dinitrophenyl)-Lalaninamide, 56 2, gave an exeellent separation, Fig. 2, of the corresponding diastereomers, 3. The (3R) diastereomer was eluted before ‘the (3s) diastereomer, as shown by injecting derivatives prepared from (3R)- and (3S)+leutine. The fact that the observed peaks are indeed due to the (3R)- and (3S)-/3-leucines was supported by separating the derivative prepared from (3RS)-[2,3-3H2(N)]-@-leucine (22). Virtually all of the radioactivity was confined to the two peaks (essentially equal amounts) labeled (3R) and (3S), Fig. 2. It should be ’ In the earlier publications (20,2 I), this compound is called I-fluoro-2,4-dinitrophenyl-5+alanine amide. The nomenclature used in this paper is consistent with IUPAC rules of organic chemical nomenclature. We thank Dr. Kurt L. Loening (Chemical Abstracts Service) for assistance with this nomenclature. 6 Current Chemical Abstracts index name of 2 = (S)2-[(5-fluoro-2,4-dinitrophenyl)amino]propanamide.

CHROMATOCRAPHIC

SEPARATION

91

OF (3RS)+LEUCINE

NH2 0 ‘J2N

F

0

NO2

ii NH2 H

CH3

O2N

NH

UC

C02H

3 SCHEME

noted that the diastereomers obtained from (2RS)-a-leucine are also well separated by this method (their elution times are noted on the chromatogram, Fig. 2). Although the free (3R)- and (3S)-@-leucines are, of course, not readily recoverable from the derivative prepared with Marfey’s reagent, nevertheless the excellent separation should allow the ready determination of the chirality of the product formed by leucine 2,3-aminomutase and, considering the high sensitivity of detection and ease of preparation, may be useful in assays of this enzyme. REFERENCES 1. Poston, J. M. (1976)J. Biol. Chem. 251, 1859-1863. 2. Poston, J. M. (198O)J. Biol. Chem. 255, 10067-10072. 3. Poston, J. M. (1977) Science (Washington, D. C.) 195, 30 I-302. 4.

5. 6. 7. 8. 9.

Poston, J. M. (1978) Phytochemistry 17,401-402. Poston, J. M., and Hennings, B. A. (1979) J. Bucferiol. 140, 1013-1016. Poston, J. M. (1980) Biochem. Biophys. Res. Commun. %,838-843. Poston, J. M. (1981) Dev. B&hem. l&401-404. Poston, J. M. (1984) J. Biol. Chem. 259,2059-206 1. Freer, I., Pedrocchi-Fantoni, G., Picken, D. J., and

I Overton, K. H. (198 1) J. Chem. Sot. Chem. Commun., 80-82. IO. BalenoviC, K., and Dvornik, D. (1954) J. Chem. Sot., 2976.

I I. Redeman, C. E., Rice, F. O., Roberts, R., and Ward, H. P. (1955) Organic Syntheses, Coil. Vol. 3, pp. 245-247, Wiley, New York. 12. Aberhart, D. J., and Lin, H-J. (1981) J. Org. Chem. 46,3749-375

1.

13. Meyer, D., Poulin, J-C., Kagan, H. B., Levine-Pinto, H., Morgat, J-L., and Fromageot, P. (1980) J. Org. Chem. 45,4680-4682. 14. Achiwa, K., and Soga, T. (1978) Tetrahedron Lett., 1119-1120. 15. Niederwieser, A. (1972) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., eds.), Vol. 25, Part B, pp. 88-90, Academic Press, New York. 16. Tapuhi, Y., Miller, N., and Karger, B. L. (1981) J. Chromatogr. 205, 325-337. 17. Gilon, C., Leshem, R., and Grushka, E. (I 980) Anal. Chem. 52, 1206- 1209. 18. Furukawa, H., Sakakibara, E., Kamei, A., and Ito, K. ( 1975) Chem. Pharm. Bull. 23, 1625-1626. 19. Furukawa, H., Mori, Y., Takeuchi, Y., and Ito, K. (1977) J. Chromatogr. 136,428-43 I. 20. Marfey, P. (1984) Carlsberg Res. Commun. 49,591596. 2 1.

Pierce Chemical Company “Previews,” April, 1985, p. 9. 22. Aberhart, D. J., and Lin, H-J. (1983) J. Labelled Camp. Radiopharm. 20,6 1 l-6 17.