Journal of Chromatography,259 (1983) 49U98Elsevier Science Publishers B.V.. Amsterdam - Printed in The Netherlands

CHROM. t5. f f i2

Note

Resolut ion by aff ini ty chromatography: stereoselective binding of ra-cemic oxazepam esters to human serum albumin

0021-9673183/$03.00 O 1983 Elsevier Science Publishers B.V.

ILONA FITOS, MIKLOS SIMONYI*, ZSUZSANNA TEGYEY ANd TASZT-6 OTVOS

Central Research Institutefor Chemistry, Hungarian Academy of Sciences, P.O. Box 17, H-1525 Budapest

( Hungary)

and

JUDIT KAJTAR ANd MARTON KAJTAR

Institute of Organic Chemisty, Eiitvds University, Budapest (Hungary)

(Received December lst, 1982)

The detection of stereoselective binding of racemic drugs to serum proteins isof great importance for the elucidation of chiral drug action. Applying resolved enan-tiomers of oxazepam hemisuccinate, Mtiller and Wollertr found that the binding ofthe S( * )-enantiomer to human serum albumin (HSA) is about 30 times stronger thanthat of the R(-)-antipode. Ultrafiltration techniques developed specifically for thestudy of racemic compounds2-a have been successfully applied for the quantitativedetermination of stereoselective binding of oxazepam acetate. Increasing the hydro-phobicity of the acyl moiety in oxazepam esters:

o

.eqt: ,_-rAO\2

however, causes systematic errors due to non-specific binding of the ligand to ultra-filtration membranes.

Proteins immobilized in hydrophylic gels usually maintain a specific bindingaffinity towards small moleculess. Hence, affinity chromatography can be used todetect the stereoselectivity of binding to serum albumin6-8. This method even pro-vides quantitative data on binding affinity if certain column parameters are knowne.As non-specific binding of the drugs can be avoided and very small amounts ofracemic samples are applicable, the chromatographic technique seems superior toultrafiltration in most instances. A micro-preparative resolution of racemic oxazepamacetate on an HSA-sepharose column has been performed, which allowed the chi-roptical parameters for the enantiomers to be determineda.

In this paper we compare different methods for studying the stereoselectivebinding of a number of oxazepam esters to HSA.

NOTES 495

EXPERIMENTAL

Racemic oxazepam esters were prepared from oxazepam with acyl chloridesusing pyridine-catalysed acylationlo. Lyophilized HSA was obtained from the"Human" Serum and Vaccine Institute, Budapest, Hungary. Ringer buffer (pH 7.a)with 0-2oh of ethanol was used.

Affinity chromatographic resolution of racemic samples (5 pg in 20 pl ofethanol) was performed on a column containing HSA immobilized oq cyanogenbromide-activated Sepharose 48 (Pharmacia, Uppsala, Sweden) using UV detection.The HSA concentration was about 10-4 M.The eluent contained 0.02% of sodiumazide. The elution of ethanol was indicated by a small negative peak.

Preparative ultrafiltration was carried out in an Amicon 202 cell using PM-10membranes. Partially resolved esters were extracted with chloroform from both thefiltrate and the retentate and purified by thin-layer chromatography. Concentrationswere determined from UV spectra. Circular dichroism spectra were taken in ethanolsolution on a Roussell-Jouan (Jobin-Yvon, Longjumeau, France) Dichrograph No.III instrument.

RESULTS

All racemic oxazepam esters studied could be eluted as two equivalent peaks.Fig. I shows the chromatographic resolution of racemic oxazepam propionate, andalso that of the samples partially resolved by ultrafiltration. The filtrate containsmore of the w-eakly bound enantiomer, while the ligand obtained from thE retQntateis enriched with the strongly bound stereoisomer. Table I shows the elution volumesof the enantiomers obtained by the resolution of different racemic oxazepam esters.Elution volumes over 50 ml could not be determined accurately owing to the broadand flat character of the peaks (see Fig. 1). The elution volume for diazepam is alsogiven which corresponds to a binding constant (K)tt of about 1.8 . 10s M-r. Theratio given in Table I, which includes the elution volumes of corresponding €rtoo-tiomers corrected for the elution volume of a ligand with no specific interaction (e.g.,solvent), is characteristic of the stereoselectivitye. Considering separately the varia-tion in the elution volumes Zp (first peak) and Vs (second peak) due to the increasinghydrophobicity of the acyl moiety from acetate to phenylacetate, it can be observedthat the binding affinity increases for both enantiomers, leaving the stereoselectivityalmost unchanged. Methylsuccinate ester shows a low binding affinity, in accordancewith its low hydrophobicity. The relative decrease is, however, more pronounced forthe (S)-enantiomer (cf., acetate), resulting in a smaller stereoselectivity. The hemi-succinate with an anionic nature exhibits a strikingly higher stereoselectivity than theother esters, because the binding affinity is small for the (R)-enantiomer, in accord-ance with its polar character, whereas it is relatively high for the (S)-enantiomer.

These results can be compared with those obtained by a combinedultrafiltration-circular dichroism methoda (Table II). Because of non-specific bindingto the membrane, which is increasingly disturbing with increasing hydrophobicityand the decreasing water solubility of the ligand, the binding constants of the enrantiomers could not be calculated separately and only their ratio, i.e., their stereo.selectivity, is indicated. The stereoselectivity is the lowest for methylsuccinate ester

Eco|rlGI

o

o(,coooUIII

496 NOTES

0 50 looElution volume Iml]

Fig. 1. Chromatographic resolution of oxazepam propionate on an HSA-Sepharose column. Elutionprofiles for racemate and for samples partially resolved by ultrafiltration.

TABLE I .

RESOLUTION OF OXAZEPAM ESTERS ON AN HSA_SEPHAROSE COLUMN

Elution volumes for R(-)-enantiomers (Zd and ,S(+ )-enantiomers (Zs) and their application for charac-terizing stereoselectivity.

Oxazepamester

Va @l) Vs @l) Vs - 11vR- l I

(Solvent)Aetatehopionaten-ButyrateIsobutyratePivaliiatec-Ethylbutyrate .Plenllacetate ,lvtethylsUccinate.Hemisuccinatg@iiiepam)

l923282837657320l6

4.43.54.44.43.3

2.611.8

4653858596

> 150> 150

3470

65

NOTES 497

TABLE II

STEREOSELECTIVITY FOR THE BINDING OF OXAZEPAM ESTERS TO HSA

Optical purities for the ligand in the filtrate (6r) and retentate (() were measured by circular dichroismafter ultrafiltrating solutions of racemic oxazepam esters and HSA.

Oxazepantester

C""rn

(M x 10-5)

Cnst *(M x lq-s)

KsKx

AcetatePropionaten-ButyrateIsobutyratePivaloateMethylsuccinateHemisuccinate*

8.13.03.22.9l . l5.1

15.04.53.03.81.54.5

-0.565 + 0.01-0.41 + 0.02-0.31 + 0.01-0.35 + 0.01-0.39 + 0.02-0.26 + 0.02

0.265 + 0.01 6.50.58 + 0.03 l00.67 + 0.03 120.56 + 0.03 80.4 + 0.01 60.37 + 0.02 4

30**

+l+3+3+3t2+2

* Hydrolysed during purificationn From ref. l.

. t , '

and the largest for hemisuccinate, in agreement with the chromatographic results.The variation of the stereoselectivity in going from acetate to pivaloate does not see4to be significant owing to the large uncertainty of the Ks/Ka values.

Oxazepam, the parent 3-hydroxy compound, was also investigated. It cannotbe resolved into enantiomers because of its very rapid racemization in solutionr2. Itspeculiar elution profile, shown in Fig. 2, is presumably a manifestation of its stereo-selective binding and tends to indicate the lack of racemization for the bound enan-tiomer.

Ec

or.r)c\I

(d

oIJ

co-ooo-o

0.02

Fig. 2. Elution profile of oxazepam on an HSA-Sepharose column. Elution volume of the solvent, I I ml.

NOTES

EXPERIMENTAL

Racemic oxazepam esters were prepared from oxazepam with acyl chloridesusing pyridine-catalysed acylationlo. Lyophilized HSA was obtained from the"Human" Serum and Vaccine Institute, Budapest, Hungary. Ringer buffer (pH 7.a)with 0-2o/o of ethanol was used.

Affinity chromatographic resolution of racemic samples (5 pg in 20 pl ofethanol) was performed on a column containing HSA immobilized on cyanogenbromide-activated Sepharose 48 (Pharmacia, Uppsala, Sweden) using UV detection.The HSA concentration was about l0-4 M.The eluent contained 0.02% of sodiumazide. The elution of ethanol was indicated by a small negative peak.

Preparative ultrafiltration was carried out in an Amicon 202 cell using PM-10membranes. Partially resolved esters were extracted with chloroform from both thefiltrate and the retentate and purified by thin-layer chromatography. Concentrationswere determined from UV spectra. Circular dichroism spectra were taken in ethanolsolution on a Roussell-Jouan (Jobin-Yvon, Longjumeau, France) Dichrograph No.III instrument.

RESULTS

All racemic oxazepam esters studied could be eluted as two equivalent peaks.Fig. I shows the chromatographic resolution of racemic oxazepam propionate, andalso that of the samples partially resolved by ultrafiltration. The filtrate containsmore of the w-eakly bound enantiomer, while the ligand obtained from the retqntateis enriched with the strongly bound stereoisomer. Table I shows the elution volumesof the enantiomers obtained by the resolution of different racemic oxazepam esters.Elution volumes over 50 ml could not be determined accurately owing to the broadand flat character of the peaks (see Fig. 1). The elution volume for diazepam is alsogiven which corresponds to a binding constant (K)tt of about 1.8 . lOs M-r. Theratio given in Table I, which includes the elution volumes of corresponding enan-tiomers corrected for the elution volume of a ligand with no specific interaction (e.g.,solvent), is characteristic of the stereoselectivitye. Considering separately the varia-tion in the elution volumes Zp (first peak) and Z5 (second peak) due to the increasinghydrophobicity of the acyl moiety from acetate to phenylacetate, it can be observedthat the binding affinity increases for both enantiomers, leaving the stereoselectivityalmost unchanged. Methylsuccinate ester shows a low binding affinity, in accordancewith its low hydrophobicity. The relative decrease is, however, more pronounced forthe (S)-enantiomer (cf., acetate), resulting in a smaller stereoselectivity. The hemi-succinate with an anionic nature exhibits a strikingly higher stereoselectivity than theother esters, because the binding affinity is small for the (R)-enantiomer, in accord-ance with its polar charactero whereas it is relatively high for the (S)-enantiomer.

These results can be compared with those obtained by a combinedultrafiltration-circular dichroism methoda (Table II). Because of non-specific bindingto the membrane, which is increasingly disturbing with increasing hydrophobicityand the decreasing water solubility of the ligand, the binding constants of the en:antiomers could not be calculated separately and only their ratio, i.e., their stereo.selectivity, is indicated. The stereoselectivity is the lowest for methylsuccinate ester

495

DISCUSSION

Summarizing the changes obtained by varying the acyl moiety, the followingconclusions can be drawn. i ;

(l) Increasing hydrophobicity of the acyl group increases the binding of bothenantiomers without a significant change in the stereoselectivity. This is in accordancewith the observationlo that the binding of racemic oxazepam esters to HSA gives agood correlation with the hydrophobicity. On the other hand, our findings seem tobe in contradiction with the results found for 3-alkyl-substituted compounds. Thestereoselectivity factor of 7 for the 3-isopropyl derivativel3 compared with the factorof 40 for the 3-methyl compoundla was explainedls by steric hindrance that occursonly with the (S)-enantiomer.

(2) The presence of polar atoms in the acyl moiety (cf., methylsuccinate ester)decreases both the binding affinities of the enantiomers and the stereoselectivity.

(3) A negative charge on the acyl moiety (cf., hemisuccinate) decreases thebinding of the (R)- and enhances the binding of the (S)-enantiomer, resulting in ahigh stereoselectivity.

Elution volumes obtained by affinity chromatography are characteristic of theoverall binding affinity of the strongest binding site at the protein. Hence, it cannotsubstitute techniques for detailed studies on binding. Nevertheless, it is a rapid, usefuland economic method when relative values are to be investigated. The results withoxazepam showed that phenomena that cannot be observed in solution might bedetected for the bound molecule by chromatography.

REFERENCES

I W. E. Miiller and U. Wollert, Mol. Pharmacol.,ll (1975) 52.2 M. Simonyi, I. Fitos andZs. Tegyey, J. Chem. Soc.,Chem. Commun., (1980) 1105.3 M. Simohyi, I. Fitos, Zs.Tegyey and L. Otvos, Biochem. Biophys. Res. Commun,,97 (1980) l.4 M. Simonyi, I. Fitos, J. Kajt6r and M. Kajt6r, Biochem. Biophys. Res. Commun., 109 (1982) 851.

5 J. Porath, J. Chromatogr., 218 (1981) 241.6 K. K. Stewart and R. F. Doherty, Proc. Nst. Acad. Sci. U.S., 70 (1973) 2850.7 C.Lagercrantz, T. Larsson and I. Denfors, Comp. Biochem. Physiol.,69C (1981) 375.8 S. Allenmark, B. Bomgren and H. Bor6n, J. Chromatogr.,237 (1982)473.

9 C. Lagercrantz, T. Larsson and H. Karlsson, Anal. Biochem.,99 (1979) 352.l0 G. Maksay, Zs. Tegyey and L. Otvos, J. Med. Chem.,22(1979) 1443.I I I. Sjoholm, B. Ekman, A. Kober, I. Ljungstedt-Pihlman, B. Seiving and T. Sjodin, Mol. Pharmacol.,

t6 (t979) 767.l2 M. Stromat, V. 5unji6, T. Kovad, L. Klasinc and F. KajfeZ, Croat. Chem. Acta,46 (1974) 265.

l3 G. Gratton, S. Rendi6, V. Sunji6 and F. KajfeL, Acta Pharm. Jugosl., 29 (1979) ll9.14 T. Alebi6-Kolbah, F. KajfeZ, S. Rendi6, V. Sunji6, A. Konowal and G. Snatzke, Biochem. Pharmacol,

28 (1979) 24s7.