Opinion Polls & Market Research  

CHIRAL CHROMATOGRAPHY

(There are plenty of new developments since this review was written!)  

The biological activity of chiral substances often depends upon their stereochemistry, since the living body is a highly chiral environment. A large percentage of commercial and investigational pharmaceutical compounds are enantiomers, and many of them show significant enantioselective differences in their pharmacokinetics and pharmacodynamics. The importance of chirality of drugs has been increasingly recognized, and the consequences of using them as racemates or as enantiomers has been frequently discussed in the pharmaceutical literature during recent years.  With increasing evidence of problems related to stereoselectivity in drug action, enantioselective analysis by chromatographic methods has become the focus of intensive research of separation scientists.  Most of the pharmaceutical and pharmacological studies of stereoselectivity of chiral drugs before the mid eighties involved pre-column derivatization of the enantiomers with chiral reagents, forming diastereomers. The diastereomers were subsequently separated in the normal or reversed phase mode of chromatography.
 
 

Great efforts have been devoted to the development of better methodology for enantioselective chromatography during the past decade, and have resulted in new chiral stationary phases, pioneered by Pirkle [1 and references therein]. Chiral agents were derivatized and immobilized on the surface of the support (silica gel mostly), and served as the in situ chiral discriminators during the chromatographic process. The preference of chiral stationary phases lies in the inherent advantages of any chromatographic separation, such as the speed of the analysis, the possibility to analyze or purify the enantiomers in complex mixtures, the reproducibility of the analysis and its flexibility. Moreover, analytical chromatographic systems can be adapted to preparative separations, in which pure enantiomers can be collected.
In addition to their distinct practical applicability, chiral stationary phases can uniquely contribute to studies of the nature of molecular recognition. Since the differential retention of enantiomers in the chromatographic system employing chiral stationary phases, can be attributed only to chiral discrimination by the chiral sites, these interactions can be isolated and explored.  It has been shown that chromatographic parameters obtained by chiral stationary phases can be sensitive to very subtle differences between the enantiomers. Moreover, chiral stationary phases can be tailor-made to accommodate specific studies of chiral recognition between molecules.
A review by Taylor and Maher [2] describes in detail the principles underlying chiral discrimination by the various chiral chromatographic systems, utilizing chiral agents either in the mobile or the stationary phase. Another review by Gubitz [3] describes the application of chiral stationary phases to chiral drugs, emphasizing the main principles of chiral discrimination of the various categories of stationary phases known so far. A conventional classification of types of chiral stationary phases is used here:

A. Chiral affinity by proteins (serum albumin, a1-acid glycoprotein, ovomucoid and chymotrypsin).
B. Stereoselective access to helical chiral polymers (derivatized or free polysaccharides).
C. Steric interactions between p-Donor p-acceptor type of chiral aromatic amide groups (Pirkle).
D. Host-guest interactions inside chiral cavities (cyclodextrins, crown ethers and imprinted polymers).
E. Ligand exchange (copper ions complexed with chiral moieties).

Most of the analytical methods for pharmaceutical compounds in biological samples use types A-D of the aforementioned stationary phases, and therefore, the discussion will be focused on them. The parameters of importance in chiral recognition by the chromatographic stationary phases will be discussed in each section. It may be generalized that in most cases the difference in steric fit, anchored by hydrogen bonding of the solutes into the chiral environment in the specific discriminating sites, is responsible for the resolution.
The various biological sources from which samples were taken for analysis are specified in the tables, listing compounds of pharmaceutical interest analyzed by the various stationary phases.  The purpose of listing the sample source, namely the biological fluid or tissue, is to portray the type of pharmaceutical research that benefits from the availability of the enantioselective analysis.  In general, whenever a method describes the enantioselective analysis of drug in plasma, it is being used for therapeutic monitoring of drug levels in the blood, or for studies of enantioselectivity of pharmacokinetic parameters.  If the method is applied to plasma and urine, both being analyzed simultaneously, the purpose is probably to track the enantioselective metabolism and/or clearance of drug.  If chiral pharmaceutical compounds are analyzed in various tissues, the aim is to study enantioselective absorption and distribution in the various organs (disposition).  On many occasions the users develop the analytical procedure to fit their specific needs for stereoselective analysis.  The tables can give an indication of the types of problems encountered in the pharmaceutical research that require enantioselective analysis.

A.  PROTEIN IMMOBILIZED ON SILICA GEL

One of the most appealing types of chiral stationary phases for pharmaceutical analysis involves the use of protein immobilized to the surface of silica gel, or other support, as the chiral discriminator.  Many small chiral biomolecules have shown stereoselective affinity to serum albumin and to a1-acid glycoprotein, and consequently, the two proteins have been chosen as chiral selectors for the analysis of these molecules.  Naturally, the mobile phases are mostly aqueous buffers containing a limited percentage of organic modifiers.  When the protein stationary phases are efficient, even very small differences in binding affinity of the enantiomers to the protein give rise to resolution between them.  Further more, if the immobilized protein maintains its native binding ability and the mobile phase composition does not affect its chiral binding properties, valuable information of drug-protein interaction can be deduced from chromatographic parameters.

a.  Serum Albumin

The most abundant protein in the blood is serum albumin, which is regarded as a non-specific binder and carrier.  The bioavailability of plasma protein-bound molecules exceeds that of the free molecules, since they have temporary protection and slower metabolism and excretion.  Therefore, enantioselective binding of drugs by blood proteins is a vital function in their action, and can be explored clinically and pharmacologically by using enantioselective analysis using immobilized proteins.
Immobilization of serum albumin on the silica gel requires sophisticated chemistry, using the appropriate spacers and anchoring agents.  Chromatographic stationary phases consisting of immobilized serum albumin on silica gel are commercially available, enabling enantioselective analysis of pharmaceutical compounds, as well as studies of drug-protein binding by chromatographic parameters.  The availability of commercial stationary phases with serum albumin immobilized on silica gel have triggered many studies of drug-protein interactions. Derivatives of 2-aryl propionic acid (profens), the widely used anti-inflammatory agents, were also separated on the serum albumin column.  This group provides an interesting example of enantioselective metabolism.  The enantiomers undergo a unidirectional stereochemical inversion in-vivo, namely, the (-)-R enantiomer is converted into the active (+)-S form by an enzymatic mechanism.  Therefore, the (-)-R form should be considered as a pro-drug of the (-)-S enantiomer, and a mixture may be used in therapy.

b.  a1-acid glycoprotein (AGP)

The concentration of AGP in blood is much lower than HSA, and consequently, its binding capacity is lower.  Nevertheless, basic drugs have significant affinity to this protein, and the stereoselectivity of their binding may considerably affect their pharmacokinetic and pharmacodynamic behavior.  Not all the properties of the binding sites in AGP are known.  It is anticipated that binding studies using chromatographic parameters will shed light on the mechanism of chiral recognition by the protein.  An efficient analytical column can be constructed with the AGP stationary phase, for an increasing number of applications in pharmaceutical analysis.  The mechanism of the stereoselective affinity of this protein cannot be easily deduced, until the structural features are fully established.
The anti-inflammatory 2-arylpropionic acid derivatives, some of which were previously mentioned, were separated on AGP columns as well as on albumin columns.  Efficient separations were obtained on the AGP, and it was demonstrated that it is applicable for stereoselective pharmacokinetic studies of ketoprofen, ibuprofen and fenoprofen after administration under clinical conditions.

c.  Ovomucoid and a-chymotrypsin

Another type of chiral affinity stationary phase is ovomucoid, immobilized on silica gel, which has also proven effective in chiral discrimination of various pharmaceutical compounds.  Also in use is immobilized a-chymotrypsin, which has a known recognition site for specific chiral substrates.  The a-chymotrypsin stationary phase was developed mainly by Wainer and co-workers.  They were able to resolve a number of the enantiomeric D,D- and L,L-dipeptides as well as the diastereomeric D,D-/L,L- and L,D-/D,L-dipeptides.  Another attempt to explore the binding of dipeptides to immobilized a-chymotrypsin indicated that the observed enantioselectivity of the stationary phase to the particular dipeptides is a measure of the difference in the binding affinities at two sites rather than differential affinities at a single site.

B.  POLYSACCHARIDE DERIVATIVES

Polysaccharides such as cellulose and amylose consist of D glucose units linked by 1-4 glucosidic bonds, forming the natural polymers with a highly ordered helical structure.  The three hydroxyls on each glucose unit can be derivatized to form strands around the chiral glucose.  The derivatized glucose unit can in principle act as a chiral site discriminating between enantiomers that interact differently with the strands.  Resolution can sometimes be achieved with unsupported natural cellulose, but the immobilized version has proven far better.  The acetate ester, bezoate ester, or phenylcarbamate derivatives of glucose, have shown better performance.  Mobile phases are usually organic, normal phase type solvents, however, aqueous solvents can also be used in many versions of the stationary phase.
Figure 2 illustrates the structure of a glucose unit of the amylose based stationary phase, derivatized with dimethylphenylcarbamate.
 
 

The structure provides the possibilities of p-p interaction of aromatic groups with the aromatic amide at the chiral site with the anchoring effect of hydrogen bonding with the amide groups.  Wainer studied aromatic alcohols on cellulose tribenzoate and suggested that both insertion of the aromatic group and hydrogen bonding stabilize the enantiomers inside the chiral cavity.  The discrimination is affected by the steric fit in the cavity.

An example of the separation of enantiomers of cannabidiol, one of the substances in Marijuana is shown in the following Figure.
 
 

C.  CHIRAL CAVITY

Another general strategy for chiral discrimination on a stationary phase is creation of chiral cavities, in which stereoselective guest-host interactions govern the resolution.  The first important consideration for retention and chiral recognition in such stationary phases is the proper fit of the molecule to the chiral cavity in terms of size and shape.  This category of stationary phases includes crown ethers, imprinted polymers and cyclodextrins.  A majority of pharmaceutical applications were accomplished using cyclodextrins, and therefore, the discussion is concentrated on them.

Cyclodextrins
Cyclodextrins are macrocyclic molecules containing 6, 7 and 8 glucopyranose units  (a-, b-, g- cyclodextrin respectively), as shown in Figure 3.

The monomers are arranged so that a shape of a hollow truncated cone is obtained.  A relatively hydrophobic chiral cavity is formed, comprised of essentially methylene and 1,4 glucoside bonds, with which the intercalated solute interacts.  In contrast to the interior, the exterior surface is hydrophilic, surrounded by hydroxyls.  Mobile phases are usually aqueous solutions mixed with organic solvents, however, normal phase type solvents can also be used.  When cyclodextrin stationary phases are used with aqueous mobile phases, the mechanism of retention is based on inclusion complexation.   This mechanism represents the attraction of the apolar molecular segment to the apolar cavity.  When an aromatic group is present, the orientation in the cavity will be stereoselective due to the interactions with the glucoside oxygens.  Linear or acyclic hydrocarbons can occupy positions in the cavity in a random fashion.  It is therefore essential that the solute has at least one aromatic ring, if a chiral separation is attempted in the reversed phase mode.  The high density of secondary hydroxyls at the larger opening of the torroid is responsible for the preferential hydrogen bonding.  Amines and carboxyl groups react strongly with these hydroxyl groups, as a function of the pK of the solute and pH of the aqueous mobile phase.

D.  p-DONOR p-ACCEPTOR - PIRKLE TYPE

Historically, this type of chiral stationary phase preceded all the others described here [4].  The pioneering work of Pirkle had such an impact on the field that the whole category of donor-acceptor type stationary phases was named after him.  The structure of these type of stationary phases is based on single strands of chiral selectors, connected via amidic linkage onto aminopropyl silica as shown in Figure 4.

The strands possess either p-donor or p-acceptor aromatic fragments as well as a hydrogen bonding agent and dipole stacking inducing structure.  The preliminary work used p-donor type anthryl groups, which were subsequently changed to p-acceptor dinitrobenzoylphenyl (DNBP) derivatives of amino acids.  Their success marked the beginning of the proliferation of the field of chiral liquid chromatographic separations using chiral stationary phases.  Since the DNBP group is a p-acceptor, solutes should possess a p-donor group such as an aromatic ring with alkyl, ether or amino substituents, in order to be separated.  Moreover, solutes should be able to form hydrogen bonds or enter into dipole-stacking with the amide group attached to an aromatic system on the stationary phase.

CONCLUSION
The mechanism of chiral discrimination by the various chiral stationary phases described here becomes more apparent, however, there is still a long way to go until  the prospects for resolution of specific solutes on specific stationary phases can be easily predicted.  Optimization of the separation is still not quite understood, and systematic comparative approaches are pursued.

REFERENCES

[1] Pirkle W.H. and Pochapsky T.C., "Advances in Chromatography" eds. Giddings J.C., Grushka E. and Brown P.R., Marcel Dekker Inc. NY, vol 27, 73-127, 1987.

[2] D. R. Taylor and K. Maher, "Chiral Separations by High-Performance Liquid Chromatography", J. Chromatogr. Sci., 30, 67-85 (1992).

[3] G. Gubitz, "Separation of Drug Enantiomers by HPLC Using Chiral Stationary Phases- A Selective Review", Chromatographia, 30, 555-564 (1990).