Synthesis and Characteristics of the Human SerumAlbumin-Triazine Chiral Stationary Phase

QIANG ZHANG, HANFA ZOU,* XIAOMING CHEN, HAILIN WANG, QUANZHOU LUO, AND JIANYI NINational Chromatography R&A Center, Dalian Institute of Chemical Physics, Chinese Academy of Science,

Dalian, China

ABSTRACT Human serum albumin (HSA) was successfully bonded to silica withs-triazine as activator. The coupling reaction by this method was rapid and effective. Thetriazine-activated silica is relatively stable and can be installed for at least 1 monthwithout obvious loss of reactivity when stored below 30°C, pH below 7. It was observedthat the amount of bound HSA reached 120 mg/g silica calculated from the UV absor-bance difference of the HSA solution. d,l-tryptophan was selected as the probe solute tocharacterize the properties of HSA bonded s-triazine chiral stationary phase, and sepa-ration factor of 9.4 was obtained for d,l-tryptophan. Furthermore, the amount of effectiveHSA on silica was measured by high-performance frontal analysis, and only 16.8 mg/gsilica was responsible for the resolution of d,l-tryptophan. These results indicate that theamount of both the bound and effective HSA on silica with triazine as activator was muchhigher than those by the Schiff base coupling method. Different kinds of enantiomerswere resolved successfully on the aminopropylsilica-bonded HSA s-triazine chiral sta-tionary phase. Chirality 12:714–719, 2000. © 2000 Wiley-Liss, Inc.

KEY WORDS: HSA chiral stationary phase; s-triazine activator; high-performance frontalanalysis; resolution of enantiomers

Enantiomer separation is an area of increasing impor-tance in chemistry and the pharmaceutical industry. It isalso one of the most important accesses to elucidate thenature of molecular recognition. Several methods havebeen used successfully to realize chiral separation. Thechromatographic method is one of the most practical pro-cesses for this feasibility and effectiveness. In the past de-cades there have been several kinds of chiral stationaryphases (CSPs) developed1–5 based on the various chiralselectors such as celluloses, cyclodextrins, synthetic enan-tiomers, synthetic polymers, molecular imprinting phases(MIPs), and so on. Proteins are natural high-molecularweight polymers composed of L-amino acids, which havebeen shown to stereoselectively bind chiral molecules,6–11

and it has been shown that the protein-chiral stationaryphases (protein-CSPs) usually show better enantiomer dis-crimination, in general because proteins possess complexand changeable conformations. This property has beenused for the synthesis of a number of CSPs in HPLC, in-cluding the phases based on bovine serum albumin (BSA),human serum albumin (HSA), a1-acid glycoprotein (AGP),avidin, cellulase, ovoglycoprotein, and enzymes such astrypsin, a-chymotrypsin, and pepsin.7,12–17 Since holdingthe natural conformation for proteins is essential for chiralseparation, the methods for immobilization of proteinsmust be mild and quick.

HSA is a 66.5-kDa protein responsible for the transportof a wide variety of biological and pharmaceutical com-pounds in blood. The interaction of these compounds withHSA is believed to occur at a series of well-defined bindingsites on the surface of the protein, with the strength and

number of sites to which a given molecule binds varyingfrom different compounds or even between different chiralforms of the same compound. This ability to provide chiraldiscrimination, plus the inherent biological importance ofHSA, has made this protein attractive for examining thechiral separation and drug–protein interactions.18,19

Dominici et al.20 bonded HSA on silica by 1,1-carbonyldiimidazole as the activator; the amount immobi-lized on the support was reported to be ∼95 mg per gramof silica gel. As we know, 1,1-carbonyldiimidazole-activatedsilica is not stable enough at room temperature and is bet-ter used immediately after activation.21 S-triazine was usedas the activator to bond BSA successfully on silica.22 Ex-perimental results indicated the bonding of BSA was rapidand effective, and since the bonding could be carried outby simply pumping the protein solution through the acti-vated column, there is no stirring shortcoming that coulddestroy the natural conformation of the proteins.

In this article, we report for the first time a synthesis ofHSA chiral stationary phase using s-triazine as the activa-tor. In comparison with the 1,1-carbonyldiimidazole-activated method, triazine-silica is very stable when storedat a temperature below 30°C, pH <7. The triazine-activated

Contract grant sponsor: the National Natural Science Foundation of China;Contract grant numbers: 29635010, 29725512.*Correspondence author: Dr. Hanfa Zou, National Chromatography R&ACenter, Dalian Institute of Chemical Physics, Chinese Academy of Science,Dalian 116011, China. E-mail: zouhfa@mail.dlptt.ln.cnReceived for Publication 28 April 2000; Accepted 13 June 2000

CHIRALITY 12:714–719 (2000)

© 2000 Wiley-Liss, Inc.

matrix can be stored at 2–4°C for more than 1 month with-out obvious lose of activity for protein immobilization.There is no problem of diminishing the immobilizationamount of HSA and the activated matrix can be conve-niently preserved under normal conditions for a long time.

EXPERIMENTALMaterials

HPLC grade of acetonitrile was purchased from theShanghai Bioengineering Center, China. (3-amino-propyl)triethoxysilane, atropine, dopa, dns-d,l-threonine,

dnp-d,l-glutaric acid, dnp-d,l-methionine, d,l-tryptophan,warfarin, ketoprofen, fenoprofen, ibuprofen, and naproxenwere purchased from Sigma Chemical Co. (St. Louis, MO,USA). HSA 20% (w/v) solution was purchased from theBehring Co. (Behringwerke AG, Marburg, Germany);1-propanol, 2-propanol, toluene, acetone, sodium dihydro-gen phosphate, disodium hydrogen phosphate, and otherreagents were of analytical-reagent grade. Silica (5 µm,300Å) was purchased from Chrom Expert Co. (Sacra-mento, CA, USA). 2,4,6-trichloro-1,3,5-triazine was suppliedby San Zheng (Yingkou, Liaoning, China). The deionizedwater was purified in-house using a Milli-Q system (Milli-pore, Bedford, MA, USA).

Synthesis of HSA-Triazine Chiral Stationary Phase

The procedure for synthesis of HSA bonded s-triazinechiral stationary phase from silica gel has been describedpreviously.21 In brief, the amino-silica is obtained by react-ing silica with (3-aminopropyl)triethoxysilane, after dryingthe aminopropylsilica gel (2 g) was added to 40 ml acetoneand the mixture was cooled to 0°C by ice-bath. Underagitation 0.36 g of 2,4,6-tricholoro-1,3,5-triazine was addedslowly; the temperature was never above 5°C. Two hourslater the mixture was filtered and the activated silica waswashed several times by acetone to guarantee no excess2,4,6-tricholoro-1,3,5-triazine was absorbed on the silica-gel.

The activated silica was then packed into 150 × 4.6 mmi.d. stainless-steel column following a slurry packing tech-nique. After that the column was connected to the HPLCinstrument. A solution of HSA in 50 mM phosphate buffer(pH 7.0) was delivered through the column at 0.5 ml/min,and UV adsorbance at 215 nm and 225 nm of protein solu-tion was monitored during the immobilization. The amountof HSA immobilized on silica (C) can be calculated by:

C ~mg/L! = 144 × ~A215 − A225! (1)

About 2 h, later a solution of glycine ethyl ester (1% w/v,pH 6.7) was pumped through the column at 25°C for an-other 2 h. Then the column was washed successively with50 ml of phosphate buffer (50 mM, pH 7.0). The entireprocess of synthesis with an HSA chiral column can beachieved within 6 h.

Chromatography

The liquid chromatography system consisted of twoP200 pumps controlled by a WDL-95 workstation, a UV-200multiple wavelength detector (National ChromatographicR&A Center, Dalian, China), and a model 7125i Rheodyne

TABLE 1. Stability test of triazine-activated silica by bonding HSA on silica immediately (batch 1) and 1 month late(batch 2) after the activation

Batch t1 (min) t2 (min) k1 k2 a Rs l2

1 2.15 12.95 1.47 13.88 9.44 6.46 1.782 1.71 10.17 1.44 13.53 9.37 6.36 1.78

t1,t2: Retention time for the first and second peak of the enantiomer. k1,k2: Retention factor which was calculated by equation k = (t−t0)/t0, where t0 wasdetermined by the first turbulence of baseline. a: Separation factor which was calculated by equation a = (t2−t0)/(t1−t0). Rs: Resolution factor which wascalculated by equation Rs = 2 × (t2−t1)/(W1 1/2 + W2 1/2), where W1/2 is the peak with at half height for a solute. l2: Asymmetric factor.

Fig. 1. Chiral resolution of tryptophan on two 50 × 4.6mm i.d. s-triazineHSA-CSPs. Experimental conditions: mobile phase, 50 mM phosphatebuffer (pH 7.4); flow rate, 1.0 mL/min; column temperature, 25°C; UVdetection wavelength, 280 nm. HSA bonded to silica (a) immediately and(b) 1 month after activation with s-triazine.

ALBUMIN-TRIAZINE CHIRAL STATIONARY PHASE 715

injection valve with a 20 µL loop. The column temperaturewas controlled by a ZW-Isotemp Controller (National Chro-matographic R&A Center). The phosphate buffer was pre-pared from mixing 50 mM Na2HPO4 solution with 50 mMNaH2PO4 solution to adjust the pH of the buffer to thedesired value. The mobile phases were obtained by theaddition of the organic modifier into the buffer in propor-tion. The mobile phases were filtered and degassed beforeuse.

RESULTS AND DISCUSSION

There are two important factors that affect chiral dis-crimination ability for a protein chiral stationary phase,including the bonded amount and the holding of naturalconformations of a protein in the process of immobiliza-tion. There are two disadvantages for most of the tradi-tional methods to prepare protein CSPs: 1) the reactivematrix is usually used immediately to react with proteinafter activation to preserve its reactivity (it is difficult tohold its reactivity at room temperature for a long time); 2)the procedure of bonding is somewhat long, mostly at least12 h—it is an especially delicate question for enzyme im-mobilization. However, in the case of using s-triazine as theactivator the reactivated silica is very stable when the tem-perature is below 30°C and pH below 7. In our experi-

ments, s-triazine-activated silica was packed into two col-umns and one was HSA-bonded immediately, another wasHSA-bonded 1 month later. Tryptophan was selected as theprobe solute to be resolved at the same separation condi-tions to examine the difference between the two HSA-CSPs; the results are shown in Figure 1 and Table 1, re-spectively. It can be seen that, although one column withthe s-triazine-activated silica was stored for 1 month, thedifference in the chiral resolution of tryptophan is not sig-nificant. The separation factor for tryptophan on these twoHSA-CSPs are 9.44 and 9.37, which is much larger thanthat on any other HSA-CSPs reported previously as far aswe know; the resolution was 6.5 and 6.4, respectively. Highseparation factor on our HSA-CSP probably resulted fromthe high amount of HSA bonded on silica and better hold-ing of HSA conformations by using s-triazine as an activa-tor.

The amount of bound HSA on triazine silica can be mea-sured in two ways: 1) measure the difference in UV adsor-bance of HSA solution during bonding, as described byDominici et al.,20 then the amount of bound HSA can becalculated by Eq. (1) as 120 mg/g silica; 2) determine theeffective amount of bound protein by high performancefrontal analysis (HPFA). In this work, frontal analysis wasadopted to characterize the effective number of moles ofHSA bound on the column by using tryptophan as theprobe solute. Since d,l-tryptophan were bonded to singlebut distinct sites on HSA, the experiment was carried outon an HSA column with a dimension of 50 × 4.6mm i.d. Theexperimental data was treated with the following equa-tions24:

1V − V0

=@A#0

Bt+

Kd

Bt(2)

1~V − V0!@A#0

=Kd

Bt@A#0+

1Bt

(3)

where V and V0 are tryptophan elution volume and deadvolume, respectively, [A]0 is the concentration of trypto-phan applied to the column, Kd is the dissociation constantbetween tryptophan and HSA, and Bt is the effective num-ber of moles of binding sites on the column, which is re-sponsible for the chiral separation of tryptophan on triazineHSA-CSP.22 The typical chromatogram obtained by thefrontal analysis of d-tryptophan on this column is shown inFigure 2 and the results of data treatment according to

Fig. 2. HPFA profiles of d-tryptophan on s-triazine-HSA column. Ex-perimental conditions: column, 50 × 4.6mm i.d. packed with s-triazine-HSAstationary phase; other conditions are the same as in Figure 1. d-Tryptophan concentration of (1) 5 µM, (2) 11 µM, (3) 56 µM, (4) 111 µM,and (5) 161 µM.

Fig. 3. HPFA experimental data treated according to Eq. (2) for tryp-tophan enantiomers. Solute: (1) l-tryptophan and (2) d-tryptophan.

Fig. 4. HPFA experimental data treated according to Eq. (3) for tryp-tophan enantiomers. Solute: (1) l-tryptophan and (2) d-tryptophan.

716 ZHANG ET AL.

Eqs. (2) and (3) obtained for d- and l-tryptophan are givenin Figures 3 and 4, respectively. It can be seen that theplots of 1/(V-V0) vs. [A]0 and 1/(V-V0)[A]0 vs. 1/[A]0 gavegood linear relationships over the entire concentrationstudied. The correlation coefficients for these plots were0.998 and 0.999, respectively. The amount of HSA immo-bilized on the silica gel calculated from the change in UVabsorbance during the circulation of HSA solution wasabout 1.8 mM HSA per gram of silica, but the effectiveamount of HSA for chiral separation of tryptophan calcu-lated from Eqs. (2) and (3) are in the range 253–292 nM/gsilica, which means that only about 15% of the bound HSAwas responsible for the resolution of tryptophan enantio-mers. Such a relatively low activity by the immobilization ofHSA on silica has been reported previously and is probablya result of hindrance, improper orientation of protein at-tached to the matrix.25 However, in contrast to the immo-bilization of HSA by the Schiff base method with a bindingcapacity of 170 nM/g silica at maximum, the active sites ofHSA on triazine HSA-CSP is large enough to reach at least253 nM/g silica. The dissociation constant and bindingcapacity of d,l-tryptophan on triazine HSA-CSP are listed inTable 2.

CSPs based on the proteins have been used to resolve awide number of chiral compounds of pharmacological in-terest26,27 and to probe the binding of a drug to pro-teins.28,29 Effects of the various factors, such as mobilephase composition, eluent pH value, and column tempera-ture on the separation of chiral compounds have been in-vestigated extensively.21,30 In this study, the separations ofnine chiral compounds on the s-triazine-bonded HSA col-

umn were performed. Table 3 lists the retention factors forthe first eluted enantiomers, separation factors (a), reso-lution factor (Rs), column temperature, and eluents usedfor chromatography of various chiral compounds.

In chiral resolution of enantiomers on triazine-HSA chi-ral stationary phase, two kinds of eluent composites as thephosphate buffer-hexanoic acid and phosphate buffer-acetonitrile-propanol were adopted. The affinity of arylpro-pionic acids to serum albumin is usually high, and thechiral resolution cannot be realized in a reasonable time.Because it is known that the long-chain fatty acids such asoleate, stearate, and palmitate are among the most tightlybound organic ligands and the affinity for some chiral siteson albumin protein increases with chain length,31 n-hexanoic acid was added to the mobile phase to produce anallostic effect.28 Under these conditions, ibuprofen, feno-profen, and naproxin can be separated within 20 min withsatisfactory resolutions. For most of the other enantiomers,use of phosphate buffer-acetonitrile-propanol composite asthe mobile phase is commonly recommended. The se-lected chromatograms for separation of some chiral com-pounds are shown in Fig. 5.

From the above experimental results the following con-clusions can be drawn: 1) The s-triazine-activated silica isrelatively stable, and can be stored for at least 1 monthwithout obvious loss of bonding activity when the tempera-ture is below 30°C and pH is below 7; 2) The amount ofHSA bonded on silica with s-triazine as activator is higherthan that immobilized by 1,1-carbonydiimidazole and theSchiff base formation method, and the active sites of HSAon s-triazine silica for chiral separation of tryptophan is also

TABLE 2. Quantitative data obtained for the interaction of triazine-HSA chiral stationary phase withtryptophan enantiomers

Stationary phase Solute Bt (nM/gSi) Kd (mM) R Method

Triazine-HSA l-Tryptophan 285.7 0.077 0.998 Eq.2Triazine-HSA d-Tryptophan 292.4 0.353 0.999 Eq.2Triazine-HSA l-Tryptophan 253.8 0.064 0.999 Eq.3Triazine-HSA d-Tryptophan 269.5 0.323 0.999 Eq.3

Bt: The effective number of moles of binding sites per gram silica gel using tryptophan as the probe solute. Kd: The dissociation constant betweentryptophan and HSA. R: Correlation coefficient.

TABLE 3. Chiral resolution of enantiomers on triazine-HSA column

No. SampleFlow rate(ml/min) k2 a Rs l2

Temperature(°C)

1 Tryptophan 1.0a 13.88 9.44 6.46 1.78 25.02 Fenoprofen 0.8b 3.99 1.55 1.12 1.30 40.03 Ibuprofen 0.8b 6.86 1.41 0.96 1.24 40.04 Naproxen 0.8b 11.41 2.33 0.95 1.23 40.05 Ketoprofen 0.8c 2.74 1.76 1.46 1.18 25.06 Atropin 0.8d 5.28 3.08 2.49 1.19 ambient7 Dopa 0.8d 0.53 2.04 0.54 — ambient8 Dnp-glutamic acid 0.8d 6.69 1.23 1.36 1.55 25.09 Warfarin 0.8e 15.42 1.96 3.95 1.70 ambient

l2: Asymmetric factor, the resolution of tryptophan was on 50 × 4.6 mm i.d. column, others were on 150 × 4.6 mm i.d. column. Mobile phases: (a) 50 mMphosphate buffer (pH 7.4) containing (b) 15 mM hexoic acid and 15% acetonitrile, (c) 15% acetonitrile and 3% 1-propanol, (d) 15% acetonitrile and 2%1-propanol, and (e) 15% acetonitrile, respectively.

ALBUMIN-TRIAZINE CHIRAL STATIONARY PHASE 717

the highest among the three methods; 3) The methodfor synthesis of protein-CSPs reported here is a reliableand general procedure for the immobilization of bio-polymers on the solid matrices that may be useful to co-valently bond biopolymers in situ for preparation of biore-actors, biosensors, and affinity chromatography supports,etc.

CONCLUSION

HSA was successfully bonded to silica with s-triazine asan activator for the first time. The coupling reaction in this

method was rapid and effective. The amount of bound HSAper gram silica by this method and the binding capacitywith d,l-tryptophan as probe solute on this matrix deter-mined by HPFA are about 120 mg/g silica and at least 16.8mg/g silica, respectively, which are much higher thanthose by the Schiff base coupling method. Different kindsof enantiomers have been resolved successfully on the ami-nopropylsilica-bonded HSA s-triazine CSP. The triazine-activated silica is relatively stable and can be stored for atleast 1 month without obvious reactivity loss when storedbelow 30°C, pH below 7.

Fig. 5. Selected chromatograms for resolution of enantiomers on the s-triazine HSA-CSP. Experimental conditions as in Table 3. Solute: (a)dnp-glutamic acid, (b) ketoprofen, (c) atropine, (d) warfarin, (e) naproxen, and (f) fenoprofen.

718 ZHANG ET AL.

ACKNOWLEDGMENT

Dr. Hanfa Zou is a recipient of the Excellent Young Sci-entist award from the National Natural Science Foundationof China (29725512).

LITERATURE CITED1. Allenmark SG, Schurig V. Chromatography on chiral stationary phase.

J Mater Chem 1997;7:1955.2. Armstrong DW. Bonded phase material for chromatographic separa-

tions. US Patent 4,539,399, 1985.3. Pirkle WH, Pochapasky CT. Considerations of chiral recognition rel-

evant to the liquid chromatographic separation of enantiomers. ChemRev 1989;89:347.

4. Remcho VT, Tan ZT. MIP as chromatographic stationary phases formolecular recognitions. Anal Chem 1999;71:248A.

5. Yuki H, Okamoto Y, Okamoto I. Resolution of racemic compoundsoptically active poly(triphenylmethyl methacrylate). J Am Chem Soc1980;102:6358.

6. Marle I, Erlandsson P, Hasson L, Isaksson R, Pettersson C, PetterssonG. Separation of enantiomers using cellulase (CBH I) silica as a chiralstationary phase. J Chromtogr 1991;586:233.

7. Soglowek SM, Geisslinger G, Brune K. Stereoselective high-performance liquid chromatographic determination of ketoprofen, ibu-profen and fenoprofen in plasma using a chiral a-acid glycoproteincolumn. J Chromatogr 1990;532:295.

8. Allenmark SG, Bomgren B, Boren H. Direct resolution of enantiomersby liquid affinity chromatography on albumin-agarose under isocraticconditions. J Chromatogr 1982;273:473.

9. Allenmark SG, Andeersson S. Direct liquid chromatographic separa-tion of enantiomers on immobilized protein stationary phases. V. Op-tical resolution of N-(2,4-Dinitrophenyl)- and Dansyl-D,L-amino acids. JChromatogr 1986;351:231.

10. Gubitz G. Separation of drug enantiomers by high performance liquidchromatography using chiral stationary phases—a selective review.Chromtographia 1990;30:555.

11. Divne C, Stahlberg J, Reinikainen T, Ruohonen L, Pettersson G,Knowles JKC, Teeri TT, Jones TA. The three-dimensional crystal struc-ture of the catalytic core of cellobiohydrolase I from trichoderma re-esei. Science 1994;265:524.

12. Miva T, Miyacawa T, Miyake T, Miyake Y. Characteristics of an avi-dine-conjugated column in direct liquid chromatographic resolution ofracemic compounds. J Chromatogr 1988;457:277.

13. Erlandsson P, Marle I, Hansson L, Isaksson R, Pettersson C, Petters-son G. Immobilized cellulase (CBH I) as a chiral stationary phase fordirect resolution of enantiomers. J Am Chem Soc 1990;112:4573.

14. Haginaka J, Murashima T, Seyama C. Separation of enantiomers on alysozyme-bonded silica column. J Chromatogr A 1994;666:203.

15. Thelohan S, Jadaud P, Wainer IW. Immobilized enzymes as chromato-graphic phases for HPLC: the chromatography of free and derivatizedamino acids on immobilized trypsin. Chromatographia 1989;28:551.

16. Wainer IW, Jadaud P, Schombaum GR, Kadodkar SV, Henry MP. En-zymes as HPLC stationary phases for chiral resolutions: initial inves-tigations with a-chymotrypsin. Chromatographia 1988;25:903.

17. Haginaka J, Miyano Y, Saizen Y, Seyama C, Murashima T. Separationof enantiomers on a pepsin-bonded column. J Chromatogr A 1995;708:161.

18. He X, Carter DC. Atomic structure and chemistry of human serumalbumin. Nature 1992;358:209.

19. Zou HF, Wang HL, Zhang YK. Drug protein interaction studied bytechnique of microdialysis combined with HPLC. Review. Anal Chem1999;18:383.

20. Domenici E, Bertucci C, Salvadori P, Felix G, Cahangne I, Motellier S,Wainer W. Synthesis and chromatographic properties of an HPLCchira stationary phase based upon human serum albumin. Chromto-graphia 1990;29:170.

21. Zhang Q, Zou HF, Wang HL, Ni JY. Synthesis of a silica-bonded bovineserum albumin s-triazine chiral stationary phase for high-performanceliquid chromatographic resolution of enantiomers. J Chromatogr A2000;866:173.

22. Bethell GS, Ayers JS, Hancock WS. A novel method of activation ofcross-linked agaroses with 1,18-carbonyldiimidazole with gives a ma-trix for affinity chromatography devoid of additional charged groups. JBiol Chem 1979;254:2572.

23. Yang J, Hage DS. Characterization of the binding and chiral separationof D- and L-tryptophan on a high-performance immobilized humanserum albumin column. J Chromatogr 1993;645:241.

24. Kasai K-I, Oda Y, Nishikata M, Ishii S-I. Frontal affinity chromatogra-phy: theory for its application to studies on specific interactions ofbiomolecules. J Chromatogr 1986;376:33.

25. Tukova J. Affinity chromatography. Amsterdam: Elsevier; 1978.26. Hermansson J. Direct liquid chromatographic resolution of racemic

drugs using a-acid glycoprotein as the chiral stationary phase. J Chro-matogr 1983;269:71.

27. Davies NM. Methods of analysis of chiral non-steroidal anti-inflammatory drugs. J Chromatogr B 1997;691:229.

28. Oravcova J, Bohs B, Lindner W. Drug-protein binding studies, newtrends in analytical and experimental methodology. J Chromatogr B1996;677:1.

29. Loun B, Hage DS. Chiral separation mechanisms in protein-basedHPLC columns. 1. Thermodynamic studies of R- and S-warfarin bind-ing to immobilized human serum albumin. Anal Chem 1994;66:3814.

30. Zou HF, Wang HL, Zhang YK. Stereoselective binding of warfarin andketoprofen to human serum albumin determined by microdialysiscombined with HPLC. J Liq Chromatogr 1998;21:2663.

31. Peters T Jr. Serum albumin. Adv Protein Chem 1985;37:161.

ALBUMIN-TRIAZINE CHIRAL STATIONARY PHASE 719