Edited by G. Subramanian
Copyright ©2001 Wiley-VCH Verlag GmbH
ISBNs:3-527-29875-4 (Hardcover); 3-527-60036-1 (Electronic)
11Electrophoretically Driven PreparativeChiral Separations using Cyclodextrins
Apryll M. Stalcup
11.1Introduction
The separation of enantiomers is a very important topic to the pharmaceutical indus-try. It is well recognized that the biological activities and bioavailabilities of enan-tiomers often differ [1]. To further complicate matters,the pharmacokinetic profileof the racemate is often not just the sum of the profiles of the individual enantiomers.In many cases,one enantiomer has the desired pharmacological activity,whereas theother enantiomer may be responsible for undesirable side-effects. What often getslost however is the fact that,in some cases,one enantiomer may be inert and,inmany cases,both enantiomers may have therapeutic value,though not for the samedisease state. It is also possible for one enantiomer to mediate the harmful effects ofthe other enantiomer. For instance,in the case of indacrinone,one enantiomer is adiuretic but causes uric acid retention,whereas the other enantiomer causes uric acidelimination. Thus,administration of a mixture of enantiomers,although not neces-sarily racemic,may have therapeutic value.
Despite tremendous advances in stereospecific synthesis,chiral separations willcontinue to be important because of possible racemization along the synthetic path-way [2],during storage or in vivo (e.g.,ibuprofen [3]). While analytical methods arenecessary and have become almost routine,economical methods for preparative andsemipreparative scale-chiral separation remains largely unexplored. Yet,preparativechiral separations may be particularly important in an R&D setting where only smallamounts of material may be required to initiate screening prior to developing apotentially more costly stereospecific synthetic strategy. In addition,the pharmaceu-tical industry has a critical need for methods which produce pure enantiomers forreference materials.
During the past two decades,significant progress has been made in chromato-graphic chiral separation technology. However,the bioavailability of drug sub-stances dictates that the compounds be water-soluble,and many are ionized at phys-iological pH. The pKas of many drugs (see Table 11-1) are well outside the safeoperating range for silica-based media,and almost all high-performance liquid chro-matography (HPLC) chiral stationary phases currently available commercially areon silica substrates. In addition,most preparative liquid chromatography (LC)
288
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
Table 11-1.Examples of pKas for various chiral drugs.DrugAlbuterolBupivacaineChloroquineChlorpheniramineClassBronchodilatorAnestheticAntimalarialAntihistaminepKa9.38.110.8,8.49.2chiral chromatographic separations are performed using organic mobile phases inwhich many drug substances have limited solubility. Hence,there is a critical needfor alternative methods for preparative chiral separations. Chiral additives have been shown to be very effective for chiral separations bycapillary electrophoresis (CE) [4,5]. Indeed,it may be argued that there has beenconsiderably more research activity in chiral separations by CE than by LC methodssince the introduction of the former technique. Chiral additives in CE have severaladvantages,some of which are highlighted in Table 11-2. Table 11-2.Advantages of chiral additives in Ce.GGGGGGGGAdditive can be readily changedVariety of chiral selectors availableRapid sreening of chiral selectors/conditions/analytesSmall amounts of background electrolyte requiredSmall amounts of chiral additive requiredCan use “counter current”processesMultiple complexation possibleNo pre equilibrationWhile most discussion of resolution in CE focuses on the tremendous efficiencies(e.g.,narrow peak widths) achievable with capillary columns,it should be noted thatresolution is also a function of differences in selectivity. Unlike HPLC,where flowis unidirectional,CE using chiral additives can exploit true countercurrent migrationof oppositely charged analytes and additives. For instance,in Fig. 11-1,an electro-pherogram [6] obtained with minimal cathodic electro-osmotic flow directed awayfrom the detector,the cationic analytes only reach the detector through complexationwith the anionic cyclodextrin. Inhibition of complexation through the addition ofmethanol amplifies chiral recognition because the analyte effectively experiences a“longer”column as its own intrinsic electrophoretic mobility carries it back up thecolumn when it is in the uncomplexed state.CE is generally more suited to analytical separations than to preparative-scaleseparations. However,given the success of CE methods for chiral separations,itseems reasonable to explore the utility of preparative electrophoretic methods to chi-ral separations. Thus,the purpose of this work is to highlight some of the develop-ments in the application of preparative electrophoresis to chiral separations. Bothbatch and continuous processes will be examined.Important in all of this preparative electrophoretic work is the recognition that CE has been used in the method development of these preparative electrophoretic 11.2Classical Electrophoretic Chiral Separation:Batch Processes
289
Fig. 11-1.Effect of the addition of methanol on the enantiomeric separation of terbutaline using 2 %sulfated cyclodextrin in 25 mMphosphate buffer (pH 3).
methods. An analogous relationship may be seen in the use of chiral mobile phaseadditives in thin-layer chromatography (TLC) for screening potential chiral selectorsfor immobilization in chiral stationary phases for HPLC.
In considering the applicability of preparative classical electrophoretic methods tochiral separations,it should be noted that practitioners in the art of classical elec-trophoresis have been particularly inventive in designing novel separation strategies.For instance,pH,ionic strength and density gradients have all been used. Isoelectricfocusing and isotachophoresis are well-established separation modes in classicalelectrophoresis and are also being implemented in CE separations [7,8]. Thesetrends are also reflected in the preparative electrophoretic approaches discussedhere.
11.2Classical Electrophoretic Chiral Separations:
Batch Processes
Classical gel electrophoresis has been used extensively for protein and nucleic acidpurification and characterization [9,10],but has not been used routinely for smallmolecule separations,other than for polypeptides. A comparison between TLC andelectrophoresis reveals that while detection is usually accomplished off-line in bothelectrophoretic and TLC methods,the analyte remains localized in the TLC bed andthe mobile phase is immediately removed subsequent to chromatographic develop-ment. In contrast,in gel electrophoresis,the gel matrix serves primarily as an anti-
290
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
convective medium and is usually designed to minimize interactions with the solute,excluding molecular sieving effects. In addition,the presence of bulk liquid in thepost-run gel no doubt contributes to the solute diffusivity problem,thereby reducingefficiency and complicating detection. Hence,separation of small molecules by clas-sical gel electrophoresis is generally not done. However,solute diffusion may bereduced through complexation with bulky additives.
Righetti and co-workers [11] were one of the first to demonstrate the utility ofclassical isoelectric focusing for the chiral separation of small molecules in a slabgel configuration. In their system,dansylated amino acids were resolved enan-tiomerically through complexation with β-cyclodextrin. Preferential complexationbetween the cyclodextrin and the derivatized amino acid induced as much as a 0.1pH unit difference in the pKbs of the dansyl group.
Stalcupet al. [12] also demonstrated that chiral analytes complexed with a bulkychiral additive (e.g. sulfated cyclodextrin,MW ~2000–2500 Da,depending ondegree of substitution,DS),with reasonably large binding constants (~103 m–1) [13]could be resolved enantiomerically using classical gel electrophoresis. Initial workused a tube agarose gel containing sulfated cyclodextrin as the chiral additive.Mechanical support and cooling for the gel was provided by a condenser from anorganic synthetic glassware kit (Fig. 11-2). Although 10 mg of racemate were loadedonto the gel and significant enantiomeric enrichment was obtained,recovery of theanalyte required extrusion and slicing of the gel with subsequent extraction of theindividual slices followed by chiral analysis of the extracts,a fairly labor-intensiveprocess.
Fig. 11-2.Schematic for preparative gel electrophoresis using a condenser for mechanical support andcooling.
11.2Classical Electrophoretic Chiral Separation:Batch Processes
291
Fig. 11-3.Mini-prep continuous elution electrophoretic cell.
Fig. 11-4.UV trace of piperoxan enantiomers eluting from mini-prepelectrophoresis cell.
Stalcup and co-workers [14] adapted this method to a continuous elution mini-prep electrophoresis apparatus shown in Fig. 11-3. In this apparatus,the end of theelectrophoretic gel is continuously washed with elution buffer. The eluent can thenbe monitored using an HPLC detector (Fig. 11-4) and sent to a fraction collectorwhere the purified enantiomers,as well as the chiral additive,may be recovered. Inthis system,the gel configuration was approximately 100 mm ×7 mm,and was air-cooled. The number of theoretical plates obtained for 0.5 mg of piperoxan with thisgel was approximately 200. A larger,water-cooled gel was able to handle 15 mg of
292
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
terbutaline. However,run times were on the order of 20 h and thus represents theprobable limitations of this approach.In contrast to most classical electrophoretic separations of biologically derivedsamples,the racemic mixtures separated on these gels should be relatively pure,andsample degradation should not contaminate the gels. Hence,a significant advantageto the apparatus used in Fig. 11-3 is that the gel can be used several times. Indeed,Table 11-3 illustrates the results for four consecutive runs obtained on the same gel.Thus,in essence,the gel may serve as a surrogate column,suggesting that classicalelectrophoresis may provide a low-cost alternative to chiral chromatography. Table 11-3.Migration times for piperoxan enantiomers forfour consecutive runs on the same agarose gel using sulfatedcyclodextrin as the chiral selector.Run1234AverageSDT1 (min.)1331361291371343.6T2 (min.)1671721681791725.4Ultimately,however,it should be noted that these examples of classical gel elec-trophoretic separations are batch processes and therefore limited in sample through-put. To achieve true preparative-scale separations by electrophoresis,it becomesnecessary to convert to continuous processes.11.3Classical Electrophoretic Chiral Separations:Continuous ProcessesPreparative continuous free flow electrophoresis was first reported in 1958 [15]. Asin the case of classical gel electrophoresis,most of the work done in this area hasbeen primarily in the purification of biopolymers. Continuous free flow elec-trophoresis for the separation of small molecules has remained relatively unexplored[16],although this is beginning to change.In preparative continuous free flow electrophoresis,continuous buffer and samplefeed are introduced at one end of a thin,rectangular electrophoresis chamber. Aschematic is presented in Fig. 11-5. The sample stream is usually introduced througha single port while buffer is introduced through several ports,essentially producinga buffer “curtain”. Because the buffer streams are introduced independently,it isfairly easy to establish a variety of gradients (e.g.,pH,density,ionic strength) acrossthe buffer “curtain”. 11.2Classical Electrophoretic Chiral Separation:Continuous Processes
293
Fig. 11-5.Schematic of continuous free flowelectrophoresis apparatus.When an electric field is imposed perpendicular to the flow,differential interac-tion between the various solutes and the electric field produce a lateral displacementof individual analyte streams between the two electrodes (Fig. 11-5). Thus,the sep-arations are accomplished in free solution. Individual fractions are collected throughan array of closely spaced ports evenly placed across the other end of the chamber. In the presence of a buffer with constant composition across the electrophoreticchamber,the angle of deflection (Θ) of the solute in the electric field is dependentupon the intrinsic electrophoretic mobility of the solute (µi),the linear velocity of thebuffer (ν) and the current through the chamber (Ι) and can be described as [17]:tanΘ=µiiqκν(1)where qis the cross-section of the separation chamber and κis the specific conduc-tance of the buffer.For the separation of enantiomers,we are interested in Θ1–Θ2. Substituting a =I/qκν,using the expression relating the apparent mobility of an analyte to its bind-ing constant with a chiral additiveΘ1−Θ2≈a(µ1−µ2)(2)and the concentration of the additive,and using a series expansion of tan Θ,to a firstapproximation,the difference in the angle of deflection for the enantiomers can beexpressed as [µf−µc][K1−K2][CA](3)Θ1−Θ2≈a[](1+K1[CA])(1+K2[CA])where the subscripts f and c refer to the free and complexed analyte,respectively,and the numbered subscripts refer to the two enantiomers,1 and 2. Because themobilities of the free enantiomers are the same and assuming,to a first approxima-tion,that the mobilities of the complexes formed by each of the enantiomers with thecyclodextrin are the same,Equation (3) predicts that,as in CE,separation dependsupon differences in the mobilities of the free and complexed state and differences in294
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
the binding constants,mediated by the dimensions of the chamber as well as the spe-cific conductance and linear velocity of the buffer.
Despite the use of density and pH gradients,cooling and performance in micro-gravitational environments (e.g. the space shuttle) [18],convection and heat dissipa-tion contributed to flow stream instability which was parasitic to the desired separa-tions and limited the utility of this approach.
Recent innovations [19] have circumvented the heat dissipation and samplestream distortion inherent in most of the previous designs. In one apparatus,devel-oped by R&S Technologies,Inc. (Wakefield,RI,USA),Teflon capillary tubes arealigned close to each other in the electrophoretic chamber. Coolant is pumpedthrough the Teflon capillary tubes during the electrophoretic run while the elec-trophoretic separation is accomplished in the interstitial volume between the Teflontubes.
Continuous free flow electrophoresis has been used for the separation of biopoly-mers (e.g. ovalbumin and lysozyme) [20] as well as smaller inorganic species (e.g.[CoIII(sepulchrate)]3+and [CoIII(CN)6]3-) [21]. Sample processing rates of 15 mg h–1were reported for a mixture of Amaranth (MW:804) and Patent Blue VF (MW:1159) [22].
Three basic approaches have been used for chiral separations by continuous freeflow electrophoresis. Thormann and co-workers [23] used 2-hydroxypropyl-b-cyclodextrin as an additive for the enantiomeric enrichment of methadone in anOctopus continuous free flow electrophoresis apparatus. In this work,both zone and isotachophoretic electrophoresis was used. Processing rates were on the order of 10–20 mg h–1,which represents a significant improvement in sample throughputrelative to CE or the earlier gel work. The authors realized higher enantiomericpurities with interrupted buffer flow than with continuous buffer flow,and sugges-ted the potential of multistage continuous free flow for achieving even higher puri-ties.
Glukhovskiy and Vigh [24] also used 2-hydroxypropyl-β-cyclodextrin as an addi-tive,but their strategy involved isoelectric focusing. These authors developed thetheoretical framework and effectively demonstrated the synergism between CE andcontinuous free flow electrophoresis. In this work,also using an Octopus continuousfree flow apparatus,they were able to establish a pH gradient between 3.5 to 3.6across the electrophoretic chamber by using polydisperse ampholytes or Bier’s ser-ine-propionic acid binary buffers in the buffer stream. As in Righetti’s earlier work,complexation with the cyclodextrin additive induced sufficient differences in the pIof various dansylated amino acid enantiomers that complete enantioresolution wasobtained. Although production rates were somewhat lower (~1.3 mg h–1) thanachieved by Thormann and co-workers,the enantiomeric purity was significantlyhigher.
In a different approach,Stalcup and co-workers [25] used sulfated β-cyclodextrinfor the enantioseparation of piperoxan in work directly derived from earlier CE andclassical gel results. Their results were obtained using a continuous free flow appa-ratus developed by R&S Technologies,Inc. Processing rates on the order of 4.5 mgh–1were reported.
11.2Classical Electrophoretic Chiral Separation:Continuous Processes
295
Several issues important from a processing standpoint were addressed in thiswork. With the exception of the single isomer derivatized cyclodextrins developedby Vigh [26],almost all commercially available derivatized cyclodextrins are com-plex mixtures of homologues and isomers. Each component,in all probability,hasdifferent affinities for the two enantiomers. In the case of neutral cyclodextrins,eachcyclodextrin component has the same electrophoretic mobility (e.g. migrates onlywith the electro-osmostic flow) and the potential complexes should also have fairlysimilar mobilities. Therefore,chiral additive polydispersity should not contributesignificantly to sample band dispersion. However,in the case of cyclodextrins func-tionalized with ionizable moieties,different degrees of substitution should produceions with significantly different electrophoretic mobilities. In addition,for analytesinteracting through electrostatic attraction,substitution patterns may also signifi-cantly impact affinity. Thus,solute sample stream dispersion may be significantlyaggravated by chiral additive polydispersity. Figure 11-6 shows the distributionobtained for piperoxan in the presence and the absence of sulfated cyclodextrin. Ascan be seen from the figure,the number of vials across which the individual piper-oxan enantiomers are distributed is about the same number as the piperoxan race-mate. Thus,polydispersity of the cyclodextrin does not appear to be an issue withregard to bandwidth. Table 11-4.Distribution of piperoxan enantiomers in CFFE vials.Day1222222222V:Vial:W:i:V200200200200200160180Vial112121212121312Vial215151515151615W189899810W27897788i186185186189189151169R0.40.350.350.380.380.350.33voltage.vial containing max concentration.number of vials containing enantiomer.current.With respect to method robustness,Table 11-4 shows results obtained on severaldifferent days during which a variety of buffer conditions were used. As can be seenfrom the table,the vial corresponding to the maximum concentration of the individ-ual enantiomers as well as the number of vials containing piperoxan is fairly con-stant. As in CE,changing system variables (e.g.,pH,ionic strength,additive concentra-tion) is very easy in any of the continuous free flow electrophoresis systems reportedhere because all the interactions take place in free solution. Indeed,changing systemvariables may be easier in continuous free flow electrophoresis systems than in a CEsystem because there are essentially no wall effects. Of course,changing systemvariables in the continuous free flow electrophoresis apparatus may also be easier296
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
Histogram for CFFE Run 3 05/07/99
Absorbance(a)
Vial number
Histogram for CFFE Run 3 05/07/99
Absorbance(b)
Vial number
Fig. 11-6.Histograms showing the distribution of piperoxan enantiomers in the absence (a)and pre-sence (b) of sulfated cyclodextrin in continuous free flow electrophoresis.
than in a chromatographic system because there is no solid sorbent that is subject to degradation or that needs to be pre-equilibrated with the mobilephase.
11.4Conclusions
297
11.4Conclusions
Clearly,chiral separations,particularly preparative,present such a challenging prob-lem that no single technology can provide complete satisfaction. Much of the activ-ity in chiral separations by CE may be attributed to the advantages of CE relative toliquid chromatography (e.g.,efficiency,rapid method development,ease of chang-ing chiral selector,etc.). However,it must be noted that the true countercurrent pro-cesses possible in CE also allow selectivity to be manipulated to a much greaterextent and with greater ease than is generally feasible in liquid chromatographyusing immobilized chiral selectors. In principle,any of the chiral entities enan-tiomerically resolved by CE should be amenable to preparative electrophoreticmethods.
An important consideration for the ultimate economic viability of any preparativeelectrophoretic approach is the potential recovery of the chiral additive. Because theelectrophoretic separation depends only upon the stability of the chiral additive itselfand not the combined stability of an immobilized chiral selector,a spacer and anunderlying substrate,as in the case of cyclodextrins immobilized on a silica chro-matographic support,preparative electrophoretic separations have the potential to bemore robust than analogous chromatographic methods. Although still in its infancy,preparative chiral electrophoresis represents an important technological advance inchiral separations,and has the potential to complement preparative chiral chromato-graphic methods as well as chiral CE complements chiral analytical chromato-graphy.
Acknowledgments
The author would like to acknowledge R& S Technologies,Inc. (Wakefield,RI,USA) for the loan of the continuous free flow electrophoresis system,and Cerestar,Inc. for the donation of the sulfated cyclodextrin. The author would also like to thankDrs. Chris Welch and Prabha Painuly for helpful discussions.
References
[1]W.D. Hooer and M.S. Qing. Clin Pharmacol. Ther.,48(1990) 633.
[2]D. W. Armstrong,L. F. He,T. Yu,J. T. Lee and Y. S. Liu. Tet. Assym.,10(1999) 37.
[3]W. J. Wechter,D. G. Loughhead,R.J. Reischer,G. J. Van Giessen and D. G. Kaiser. Biochem. Bio-phys. Res. Comm.,61(1974) 833.
[4]T. J. Ward. Anal. Chem.,66(1994) 632A.
[5]H. Nishi and S. Terabe. J. Chromatogr. A,694(1995) 245.[6]S. R. Gratz and A. M. Stalcup. Anal. Chem.,70(1998) 5166.[7]X. W. Yao and F. E. Regnier. J. Chromatogr.,632(1993) 185.[8]E. Kenndler. Chromatographia,30(1990) 713.
298
11Electrophoretically Driven Preparative Chiral Separations using Cyclodextrins
[9]R. C. Allen and B. Budowle.Gel electrophoresis of proteins and nucleic acids:selected techniques.Walter de Gruyter,Berlin (1994).
[10]T. J. Bruno. Chromatographic and electrophoretic methods. Prentice Hall,NJ (1991) pp. 141-169.
[11]P. G. Righetti,C. Ettori,P. Chafey and J. P. Wahrmann. Electrophoresis,11(1990) 1.
[12] A. M. Stalcup,K. H. Gahm,S. R. Gratz and R. M. C. Sutton. Anal. Chem..,70(1998) 144.[13]N. M. Agyei. Master’s Thesis,University of Hawaii,1994.
[14]R. M. C. Sutton,S. R. Gratz and A. M. Stalcup. Analyst,123(1998) 1477.[15]M. C. Roman and P. R. Brown. Anal. Chem.,68(1994) 86A.
[16]R. M. C. Sutton and A. M. Stalcup. Preparative Electrophoresis. Encyclopedia of Separation Sci-ence,Academic Press Ltd (1999) in press.
[17]M. C. Roman and P. R. Brown. Anal. Chem.,66(1994) A86.
[18]R. E. Allen,P. H. Rhodes and R. Snyder. Sep. Purif. Meth.,6(1977) 1.[19]Y. Tarnopolsky. U. S. Patent 5,104,505 (1992).
[20]P. Painuly and M. C. Roman. Appl. Theor. Electroph.,3(1993) 119.
[21]M. E. Ketterer,G. E. Kozerski,R. Ritacco and P. Painuly. Sep. Sci. Tech.,32(1997) 641.[22]Y. Tarnopolsky,M. Roman and P. R. Brown. Sep. Sci. Tech.,28(1993) 719.
[23]P. Hoffmann,H. Wagner,G. Weber,M. Lanz,J. Caslavska and W. Thormann. Anal. Chem.,71
(1999) 1840.
[24]P. Glukhovskiy and G. Vigh. Anal. Chem.,71(1999) 3814.
[25]A. M. Stalcup,R. M. C. Sutton and J. Vela-Rodrigo. Unpublished results.
[26]J. B. Vincent,D. M. Kirby,T. V. Nguyen and G. Vigh. Anal. Chem.,69(1997) 4419.
因篇幅问题不能全部显示，请点此查看更多更全内容