Review: Journal of Antimicrobial Chemotherapy (1996) 37, 7-32
Review: Journal of Antimicrobial Chemotherapy (1996) 37, 7-32
Review: Journal of Antimicrobial Chemotherapy (1996) 37, 7-32
Review
"Department of Pharmacy, King's College London, Manresa Road, London SW3 6LX;
h
Daiichi Pharmaceuticals UK Ltd, 76 Shoe Lane, London EC4A 3JB, UK
Introduction
Over the last ten years drug chirality has become a 'big issue', not only within the
scientific and medical communities but also in the 'quality' lay press (Hawkes, 1993;
Moran, 1993) and popular scientific press (Mason, 1984; Matteson, 1991). This interest
in chirality has arisen as a result of recent advances in the areas of stereoselective
synthesis and stereospecific analysis of chiral drug molecules. As a result of these
advances, and the increasing realisation of the significance of the pharmacodynamic and
pharmacokinetic differences between the enantiomers of chiral drugs, there has been
increasing concern over the use of racemates, and other stereoisomeric mixtures, in
therapeutics. The use of such mixtures may present problems, particularly if the adverse
affects, or toxicity, of the administered agent is associated with the less active, or
inactive, isomer or does not show stereoselectivity.
7
0305-7453/96/010007 + 26 $12.00/0 % 1996 The British Society for Antimicrobial Chemotherapy
8 A. J. Hntt and J. O'Grady
A survey of 1675 drugs carried out in the early 1980s, indicated that 1200 (71.6%)
could be classified as synthetic and 475 (28.4%) as natural products or semisynthetic
agents. Four hundred and eighty (28.7%) of the synthetic compounds were chiral and of
these 58 (3.5%) were marketed as single isomers, the remainder (25.2%) were marketed
as racemates. In contrast 469 (28%) of the natural or semisynthetic products were chiral
and of these 98.3% (461) were marketed as single isomers (Aliens, Wuis & Veringa,
1988). More recent surveys have indicated that the position with respect to natural/
semisynthetic products has not changed greatly but that the proportion of synthetic
single isomer drugs increased considerably up to 1991 (Millership & Fitzpatrick, 1993).
It should be obvious from the above figures that drug chirality is not a problem
restricted to a single therapeutic group of agents but an 'across-the-board' problem;
mixtures of stereoisomers are found in the majority of therapeutic groups. As many of
the agents used in antimicrobial chemotherapy are natural, or semisynthetic, the reader
may wonder why this issue is being addressed in this journal. The problems associated
with drug stereochemistry are complex, many of the semisynthetic agents are marketed
as mixtures of diastereoisomers and a number of the synthetic agents are used as
racemates. Such mixtures are regarded by some as 'compounds containing 50%
impurity' and their use is essentially 'polypharmacy' with the proportions in the mixture
being determined by chemical properties rather than therapeutic need. As a result of
this increased concern, drug stereochemistry has become an issue for both the
pharmaceutical industry and all the major regulatory authorities (De Camp, 1989;
Cayen, 1991; Nation, 1994; Rauws & Groen, 1994). At present there is no absolute
requirement from any authority for the development of drugs as single isomers but in
the future the introduction of mixtures will require scientific justification. Indeed, several
compounds currently marketed as racemates are undergoing re-evaluation as single
isomer products and while relatively few, e.g. dexfenfluramine, have been remarketed
to date, several such compounds are in an advanced stage of development.
A compound frequently cited, particularly in the popular press, to support arguments
for the development of single isomer drugs is the teratogen thalidomide. Recent
investigations have indicated that the /?-enantiomer of thalidomide has hypnotic
properties while (iS)-thalidomide is both an hypnotic and a teratogen in SWS mice
(Blaschke et al., 1979). Thus, if the drug had been used as a single isomer then the
tragedy of the early 1960s could have been avoided. However, some older data obtained
with a more sensitive test species, New Zealand White rabbits, indicates that both
enantiomers of the drug are teratogenic (Fabro, Smith & Williams, 1967). An additional
problem with the compound is its facile racemization in biological media (Testa,
Carrupt & Gal, 1993 and references therein). Taken together these data indicate that
the situation with thalidomide is by no means as clear as some of the secondary
literature implies.
In this review when the structure of a molecule is specifically referred to the name
of the molecule is followed by a number in parentheses. These structures can be located
within the figures using this numerical identification.
HO
H HO y\%
"NHCOCHClj C12CHCOHN H
O-.N'' ^" 'NO,
(la) (lb)
HO
NHCOCHC12 CljCHCOHN
(lc) (Id)
Figure 1. Stereoisomers of chloramphenicol. The active stercoisomer of chloramphenicol (la) has the
R,/{-absolute configuration. The British Pharmacopoeia (1993) designates I a as 2,2-dichloro-A4(ot-/?, /)-/?)-
/J-hydroxy-/?-hydroxymethyl-4-nitrophenethyl]acetamide. Using this nomenclature the remaining three
stereoisomers may be designated as follows: lb, (a-S, /J-S); lc, (a-R,fi-S); Id, (a-S, fl-R). Thus, in this
diagram those compounds which are related horizontally (i.e. la with lb; lc with Id) are enantiomeric, while
those which are related vertically (i.e. la with both lc and Id; lb with both lc and Id) are diastereomenc.
As there are only two chiral centres in the molecule the diasteroisomers are also epimeric to one another,
but at different centres. The active isomer la being an a-epimer of Id and a /)-epimer of lc.
10 A. J. Hutt and J. O'Grady
B I D K B A *, I
I)' I)
Receptor
(a) (b)
Figure 2. Biological discrimination between a pair of enantiomers. The enantiomer on the left (a) takes
part in three complementary interactions with the active site, whereas that on the right (b) interacts at two
sites only. Alternative orientations of the enantiomer (right) to the active site are possible, but only two
interactions may take place at any one time. The vertical line represents the mirror plane where the centre
structure is the reflection of that on the left.
both isomers with different therapeutic indications. Both enantiomers of the drug
propoxyphene are available, one as the analgesic dextropropoxyphene, with the
(2R, 3S)-configuration, and the other laevopropoxyphene ((2S, 3^)-configuration) as
an antitussive. In the case of this example not only are the molecules mirror image
related but so are their trade names (Darvon/Novrad).
Stereoselectivity is also observed in drug disposition particularly for those pro-
cesses which depend on an interaction with a chiral biological macromolecule, e.g.
active transport processes, binding to plasma proteins, and drug metabolism (Williams
& Lee, 1985; Caldwell, Winter & Hutt, 1988; Tucker & Lennard, 1990). The passage
of the majority of drugs through biological membranes depends on their physico-
chemical properties, e.g. lipid solubility, pKa, size. In such cases differences between
enantiomers would not be expected but differences between diastereoisomers may well
occur as a result of differences in their solubility. For example the water solubility of
the D-diastereoisomer of ampicillin (5, Figure 3) is greater than that of the
L-diastereoisomer (Doyle et ah, 1962). However, if a chiral drug molecule is a substrate
for an active transport process then differences between both enantiomers and
diastereoisomers would be expected with preferential absorption of the stereoisomer
with a spatial arrangement similar to that of the natural substrate. In theory such
processes may be expected to increase the rate rather than the extent of absorption. In
fact the bioavailability of D-methotrexate is only 2.5% that of the L-isomer (Hendel
& Brodthagen, 1984). Similarly, selective transport processes may influence drug
distribution by selective tissue uptake and renal excretion, as a result of active secretion
and/or reabsorption. Plasma protein binding may also influence drug stereoisomer
distribution and renal excretion. In metabolism, a process resulting from a direct
interaction between a drug and a chiral macromolecule, stereodifferentiation is the rule
rather than the exception and stereoselective metabolism is probably responsible for the
majority of the differences observed in enantioselective drug disposition (Caldwell et al.,
1988).
As a result of the above processes the pharmacokinetic profiles of the enantiomers
of a drug administered as a racemate may differ markedly. Pharmacokinetic parameters,
e.g. clearance, volume of distribution, half-life etc., based on the determination of'total'
drug substance present in biological samples is essentially meaningless data and
potentially highly misleading or "sophisticated nonsense" (Ariens, 1984).
As pointed out above many of the agents used in antimicrobial chemotherapy are
natural or semisynthetic products and frequently single isomers are used. However,
mixtures of diastereoisomers and enantiomers do occur and the remainder of this article
will examine such cases using the /J-lactams and quinolone derivatives as representative
examples.
/J-Lactams
Within the /J-lactam group of compounds the stereochemistry of the 6-aminopenicil-
lanic acid (6-APA) and 7-aminocephalosporanic acid nucleii are thought to be absolute
requirements. Alteration of, for example, the configuration of any of the chiral centres
in 6-APA results in a marked or total loss of activity (Naylor, 1973). This perhaps
should not be surprising considering the mode of action of these compounds. The
introduction of an ot-substituent and thus an additional chiral centre in the side chain
however, results in the formation of two epimeric diastereoisomers. In the case of
, H3CO H H H
CH CONH.. I ! ^O RCONH^l |
COOH
COOH
COOH
4
CHjR' to
COOH 3
COOH
(6fl, 7fl)-7-acylsubstituted-7- Ampicillin (5)
aminocephalosporanic acid (4)
H NH2
w + H H H H
// s\ *
CH3
CONH^J |^S // \N CH C O N H ^ ! 1
CH3
COOH
O COOH
H
COOH
ampicillin (5, Figure 3) the two epimers differ in aqueous solubility (see above) and
activity, the ratio (D/L) in activity varying between 2 to 5-fold depending on the test
microorganism (Naylor, 1973). In the case of ampicillin the official preparation is the
epimer of the D-absolute configuration (which corresponds to the R configuration using
the Cahn-Ingold-Prelog system).
The introduction of a carboxyl group in the a-position yields carbenicillin (7,
Figure 3) a compound used as a mixture of epimers. The individual epimers of this
compound reportedly display only slight differences in activity and are stereochem-
ically unstable undergoing rapid epimerization in solution (Naylor, 1973; Hoover &
Dunn, 1979). In the case of this compound, separation of the individual epimers for
therapeutic use would appear to be a futile exercise.
The absorption of a number of /?-lactam antibiotics is mediated by the intestinal
dipeptide transport system and as such their absorption would be expected to be
stereoselective. The influence of configuration of the a-substituent in the side chain on
the absorption of the epimers of cephalexin has been investigated in the rat (Tamai
et al., 1988). Following the administration of L-cephalexin the unchanged drug could
not be detected in either serum or urine. In contrast the D-isomer was found to be well
absorbed. In-vitro studies indicated that both epimers were substrates for a
carrier-mediated transport system with the L-epimer showing a higher affinity than, and
acting as a competitive inhibitor for, D-cephalexin (6, Figure 3) transport. The L-epimer
was also more susceptible to the hydrolytic enzymes present in the tissues and the
unchanged drug could not be detected in the analytical samples (Tamai et al., 1988).
Latamoxef (moxalactam) (2, Figure 3) is a mixture of two epimeric forms, designated
as R and S (see above; Yamada et al., 1981), the antimicrobial activity of the ./?-epimer
being ca twice that of the S depending on the test system used (Wise et al., 1981). The
two isomers are stereochemically unstable undergoing epimerization to yield
equilibrium mixtures in the ratio R:S of 50:50 and 45:55 in buffer and serum
respectively. The rates of epimerization depending on the environment and epimeric
form (Wise et al., 1981). However, at 37C in serum the half-life of epimerization is the
same for both compounds at 1.5 h, compared to a pharmacokinetic apparent serum
elimination half-life of 2.3 h for 'total drug' following intravenous infusion of the
epimeric mixture to man (Liithy et al., 1981; Wise et al., 1981). In man the serum
concentrations of the less active S-epimer are approximately twice those of the /?-epimer
within 4 h and the ratio (R:S) in renal clearance is 1.5 (Liithy et al., 1981; Yamada et al.,
1981). In addition to the facile epimerization in serum the pharmacokinetics of
latamoxef are complicated by stereoselectivity in protein binding, the fractions unbound
being 0.47 and 0.33 for (R)- and (S)-latamoxef respectively, resulting in similar unbound
renal clearances of 140 and 132 mL/min/1.48 m2 for the R and S-epimers respectively
(Yamada et al., 1981). It would, therefore, appear that the epimeric composition of
latamoxef in plasma may be explained by a combination of epimerization and
stereoselectivity in plasma protein binding resulting in preferential renal clearance of the
/?-epimer (Yamada et al., 1981).
It is of interest to note that in the pharmacokinetic study by Luthy et al. (1981) the
serum concentrations of latamoxef were determined by both stereospecific
high-performance liquid chromatography (HPLC) and a bioassay method. The serum
concentrations determined using the bioassay methodology were consistently lower
than those obtained using the HPLC method, the difference in values increasing
progressively in samples obtained up to 2 h post drug administration. This difference
Chirality of antimicrobial agents 15
presumably reflects the more rapid excretion of the more active epimer and illustrates
the potential problems involved using bioassay methodology for the determination of
isomeric mixtures (Hutt, 1990).
OH
SCH2CH2NHR
R
Thienamycin (8) H
Imipenem (9) CH=NH
CO2H
H
H. tCH2S(CH2)4
NHCCK i
HO2C NH2
Me
Me
Cilastatin CIO)
COOH
SCH2CH2NHCH=NH2
SCH2CH2NHCH=NH2
HOOC HN
coo-
OH
H H
SCH2CH2NHCH=NH2
HOOC N
coo-
Flgure 5. Hydrolysis of imipenem (9) by dehydropeptidase-I (DHP-I) to yield a pair of diastereoisomeric
1-pyrroline derivatives. Adapted from RatclifTe el al. (1989).
COOH
Figure 6. Influence of stereochemistry on the penicillinase resistance, relative potency and susceptibility to porcine dehydropeptidase-1
of oirbapenem antibiotics (adapted from Birnbaum et al (1985) and Kropp et al. (1982)); DHP-I, porcine dehydropeptidase-I, penicillinase
resistance, hydrolysis by Bacillus ceretis penicillinase, relative potency was determined against a panel of 35 bacterial species.
18 A. J. Hutt and J. O'Grady
Prodrugs
Esterification of the carboxyl group, to yield lipophilic ester prodrugs, has been used
extensively within the /Mactams in order to improve their absorption following oral
administration. A number of these derivatives involve the formation of either an
acyloxymethyl or acyloxyethyl function which undergo rapid enzymatic hydrolysis
in vivo to yield the corresponding hydroxymethyl or hydroxyethyl esters which, being
hemiacetal derivatives, spontaneously cleave with liberation of the active /J-lactam and
the corresponding aldehyde.
The introduction of the hydroxyethyl function into the promoiety results in the
introduction of an additional chiral centre into the molecule and therefore the possibility
of a pair of diastereoisomeric compounds, e.g. for cefuroxime axetil (14) and
cefdaloxime pentexil (16). As pointed out above diastereoisomers may differ in their
physicochemical properties, e.g. solubility, and also in their susceptibility with respect
to in-vivo enzymatic hydrolysis (for structures see Figure 7).
Cefuroxime axetil (14) is the 1-acetoxyethyl ester prodrug of cefuroxime (13) and
undergoes hydrolysis in vivo to yield cefuroxime, acetaldehyde and acetic acid. The drug
material consists of an equal parts mixture of the two possible diastereoisomers of the
\'S,6R,7R (14a) and l'R,6R,7R (14b) absolute configurations. Following adminis-
tration to man the prodrug undergoes rapid hydrolysis and cannot be detected in the
systemic circulation (Harding, Williams & Ayrton, 1984) and shows a bioavailability
with respect to cefuroxime (13) of between 30 to 50% in the fasted and fed states
(Harding et al., 1984; Finn et al., 1987). Similar values for bioavailability have been
reported following administration of the prodrug to the rat and may be due to an
esterase, isolated from intestinal washings, which converts the ester to the unabsorbed
drug (Campbell, Chantrell & Eastmond, 1987). A more recent investigation has
examined the stereoselectivity of the hydrolysis using both serum and intestinal mucosal
esterases isolated from both rat and dog tissue preparations (Mosher, McBee &
Shaw, 1992). These workers found that the hydrolysis was stereoselective for the
l'S,6^,7./?-diastereoisomer (14a) but that the stereoselectivity varied with both tissue
source and species, the ratio I'S/l'R (14a/14b) being 14 and 2.5 for dog serum and
intestinal esterases respectively. The corresponding values for rat tissue preparations
being 13 and 3.4, the rat tissue esterases being faster in both cases (Mosher et al., 1992).
The possible contribution of stereoselectivity in the intestinal enzymatic hydrolysis of
the prodrug in man is not known, but such selectivity may contribute to the observed
bioavailability of between 30-50%. In human blood the mixture of diastereoisomers has
a half-life of 3.5 min (Harding et al., 1984), i.e. is rapid, and thus stereoselectivity in
hydrolysis is not a problem (for structures, see Figure 7).
20 A. J. Hutt and J. O'Grady
OCH 3
NT
II
/O Ji H H
C0NH
\J v
CH2OCONH2
COOR
R
Cefuroxime (13) H
H,;\L N
// CONH^J J^
CH2OCH3
COOR
R
Cfdaloxime (15) H
Me
(l'S, 6R, 7fl )-Cefdaloxime pentexil, HR916 K(16a)
OCOCMe3
H Me
(I'R, 6R, 7fl)-Cefdaloximepentexil; HR916 J (16b)
OCOCMe3
Figure 7. Cefuroxime axetil and cefdaloxime pentexil.
for all three stereoisomeric forms of the prodrug (Defossa et al., 1992). A more extensive
pharmacokinetic study in mice, rats and dogs, however, yielded some species differences
(Isert et al., 1992). In the mouse all three forms of the prodrug showed rapid and
essentially complete absorption; in the rat bioavailability of the drug was reduced but
no differences were observed between the individual diastereoisomers. However, in the
dog the r5,6^,77?-diastereoisomer (HR 916 K; 16a) showed approximately three times
the bioavailability of the r^,6^,77?-diastereoisomer (HR 916 J; 16b) as determined by
comparison of the areas under the cefdaloxime serum concentration time curves and
urinary drug recovery (Isert et al., 1992). Defossa et al. (1992) have also stated that the
absorption of the l'S,6.fl,77?-diastereoisomer (16a) was significantly higher in man,
but no experimental data were presented. This diastereoisomer, HR 916 K, has been
selected for evaluation and the pharmacokinetic properties of cefdaloxime following
administration of the prodrug to man have been reported (Mendes et al., 1992).
Stereoselectivity in the absorption of diastereoisomeric prodrugs, together with the
subsequent availability of the drug, may arise as a result of differential solubility at the
absorption site, rates of diffusion through the gut wall and enzymatic activity in the gut
contents, mucosa, liver and blood, and as such the potential problems associated with
the introduction of a chiral promoiety into a molecule need to be taken into
consideration at the compound design stage.
Quinolone derivatives
As pointed out above the problem of chirality is of greater significance for synthetic
agents than with natural or semisynthetic agents. One group of compounds where the
significance of stereochemistry in relation to activity has been addressed in some detail
are the substituted l,4-dihydro-4-oxopyridine-3-carboxylic acid derivatives (17),
collectively known as the quinolones. In terms of structure activity relationships within
this series the substituted oxopyridine ring system with the carboxyl group at the
3-position and the 4-carbonyl group being coplaner, is regarded by some as being
essential for activity (Shen, 1994) although useful activity has been observed with
alternative functionalities in the 3-position (Chu el al., 1989). The fused ring system may
be either aromatic or heteroaromatic with substituents at positions 6 and 7. In the
majority of compounds in this series the elements of chirality have been introduced at
positions 1 and 7 of structure 17 (Mitscher, Sharma & Zavod, 1989) (for structures, see
Figure 8).
Substituents at Nl
An important subgroup of the quinolones are those with a fused tricyclic ring system
involving attachment at positions 1 and 8 on the bicyclic ring structure (17). This ring
fusion imparts a degree of rigidity to the substituent and several of this series possess
a chiral centre in the third ring adjacent to the nitrogen atom at position 1 (e.g.
structures 18 to 20). The antibacterial activity of several members of this group has been
shown to reside in the enantiomers of the 5-absolute configuration (18a, 18c, 19a, 20a),
the 7?-enantiomers being considerably less active than the racemic mixture and the
S-enantiomers having ca twice the activity of the racemate (Hayakawa et al., 1986;
Atarashi et al., 1987; Gerster et al., 1987, 1989; Mitscher et al., 1987; Une et al., 1988).
The difference in the in-vitro enantiomeric activity ranges from 4- to 250-fold against
COOH COOH
C0011
(21)
(17)
R1 R2 R1 R2
S-(21a) CH 3 H Ciprofloxacin (22) H H
R-(2\b) H CH, Methyl analogue (23) CH3 CH3
Phenyl analogue (24) C6Hfi CH3 X
S
F C2"S
9H HN
(25) (27)
K1 R2 R1 R2 R1 R2
Call5NHCH2 H (S (-Tfemafloxacin (26a) H CH3 S-(27a) CH2OH H
K-(25b) H (fi)-Temarioxacin(26b) CH3 H -(27b) H CH2OH
Figure 8. Quinolone derivatives
Chirality of antimicrobial agents 23
O O
COOH F
R2
Compound R1 R2 R
(S)-Flumequine(18a) CH3 H H
(fl )-Flumequme (18b) H CH 3 H
(S)-Methylflumequine (18c) CH 3 H CH 3
(R )-Methylflumequine (18d) H CH 3 CH 3
O
O
COOH
R1 R2
S-(28a) NH 2 H
R-(28b) H NH 2
Figure 10. Quinolone derivatives continued.
of the antimicrobial activity of all four stereoisomers, of both 23 and 24, indicated
stereoselectivity in action but in all cases the activity was less than that of ciprofloxacin
(22) (Mitscher et al., 1989). It is of interest to note that the most potent isomer of the
phenyl series, the \'S, 2'/?-stereoisomer, was about four-fold more potent in a DNA
gyrase assay, derived from Micrococcus luleus, than its enantiomer and ciprofloxacin,
whereas against an E. coli enzyme system the above stereoisomer was equiactive with
its enantiomer and about 12-fold less active than ciprofloxacin (Mitscher et al., 1989)
(for structures see Figure 8).
Subslituenls at C7
A number of quinolones have been developed substituted at position 7 of the bicyclic
nucleus (17) with a heterocyclic ring system containing a chiral centre (structures 25-28,
Figures 8 and 10). In comparison to the tricyclic systems (18-20, Figure 9) differences
in enantiomeric activity appear to be of relatively minor significance. This is presumably
due to the centre of chirality being in a position remote from the critical binding region
of the molecules. However, stereoselectivity is observed in this series and appears to vary
with the position of substitution on the attached heterocylic ring. Thus in the case of
compound 27, the chiral centre being adjacent to the heterocyclic nitrogen attached to
the bicyclic nucleus, the /?-enantiomer (27b) is ca. 50 and 30 times more potent against
E. coli and S. aureus than its 5-antipode (27a) (Mitscher et al., 1989). In the case of
compounds 25, 28 and temafloxacin (26), compounds substituted fi to the nitrogen
atom, the in-vitro activities of the individual enantiomers are either similar or show only
relatively minor differences (Mitscher et al., 1989; Rosen et al., 1988; Chu et al., 1991).
In the case of temafloxacin (26) the enantiomers were found, within experimental
error, to possess similar activities against DNA gyrase but (S)-temafloxacin (26a)
showed slightly greater in-vivo potency in a mouse protection test (Chu et al., 1991),
which may be a result of a better pharmacokinetic profile compared with the
.R-enantiomer (26b).
A number of novel compounds substituted with chiral functionalities at both
positions 1 and 7 of the bicyclic ring system are currently under development. One such
Chirality of antimicrobial agents 25
IC (mg/L)
Compound DNA gyrase topoisomerase II selectivity"
agent DU-6859 (29), a compound with three chiral centres, is ca. 8-64 times more active
than ofloxacin against Gram-positive and Gram-negative bacteria (Sato et al., 19926).
DU-6859 shows the highest potency and high selectivity (~9000) for DNA gyrase
compared to topoisomerase II of the four stereoisomers examined (Hayakawa et al.,
1991; Hoshino et al., 1991a).
Ofloxacin
The stereoselectivity of the in-vitro antimicrobial action of the enantiomers of ofloxacin
(19) has been referred to above. The S-enantiomer (19a) being between 8- to 128-fold
more active against both Gram-positive and Gram-negative bacteria than the
fl-antipode (19b) (Hayakawa et al., 1986; Atarashi et al., 1987) (for structures, see
Figure 9).
The target enzyme of the quinolone derivatives is believed to be DNA gyrase
(bacterial topoisomerase II) (Sato et al., 1993) and a good correlation between
26 A. J. Hutt and J. O'Grady
the model proposed by Shen et al. (1990) this difference may be due to an unfavourable
binding orientation of the /?-enantiomer (19b) such that the molecular self-association
cannot take place.
The metabolism and pharmacokinetics of the enantiomers of ofloxacin (19 see
Figure 9) have been investigated following their administration as such and as both
racemic and non-racemic mixtures, to the rat, dog and cynomolgus monkey. Following
their individual administration to rats the serum concentrations of (,/?)-ofloxacin (19b)
were significantly higher than those of the S-enantiomer (19a) with a corresponding
greater area under the serum concentration time curve (AUC) and a longer apparent
serum elimination half-life. These pharmacokinetic differences arise due to
stereoselective conjugation of (S')-ofloxacin with glucuronic acid, together with
preferential biliary excretion of (S^ofloxacin, and its glucuronide, and urinary excretion
of (7?)-ofloxacin (Okazaki, Kurata & Tachizawa, 1989). In addition, in-vitro studies,
using rat hepatic microsomal preparations have indicated relatively minor differences
in the apparent Km for glucuronide formation of the two enantiomers (1.43 and 1.14 mM
for (R)- and (S)-ofloxacin, respectively) but a 6.5-fold difference in V^, the ratio of
VmMx/Km, an index of intrinsic hepatic clearance, S/R being 8.1 (Okazaki et al., 19916).
Further studies indicated that the ./?-enantiomer is a competitive inhibitor of the
glucuronidation of (S')-ofloxacin with a A", value of 2.92 mM. As a result of this
enantiomeric interaction in metabolism the serum concentrations of (5^-ofloxacin are
markedly increased following administration of the racemic drug compared with those
observed following an equivalent dose of the single enantiomer, resulting in a 1.7-fold
increase in the AUC (Okazaki et al., 19916).
Following administration of racemic ofloxacin (19, see Figure 9) to cynomolgus
monkeys significant differences were observed between the two enantiomers in AUC
(S > R), mean residence time (5 > R) and total clearance (R > S) (Okazaki et al.,
1992). Interestingly, administration of the 5-enantiomer with increasing amounts of the
/?-isomer resulted in an increase in AUC, a decrease in volume of distribution and a
decrease in both total and renal clearance of (S)-ofloxacin (19a). As the drug undergoes
minimal metabolism in this species these differences cannot be rationalised by metabolic
interactions. The renal excretion of ofloxacin is believed to involve both glomerular
filtration and tubular secretion, mediated by the organic cation transport system. Thus
the enantiomer-enantiomer interaction in the case of the monkey may be explained by
competition for the secretion or reabsorption process (Okazaki et al., 1992). In contrast
to the above two species, no differences were observed in the pharmacokinetic
parameters of the two enantiomers in the dog (Okazaki et al., 1992).
The dispositional properties of the enantiomers of ofloxacin illustrate the potential
problems which may arise when dealing with racemic mixtures, i.e. stereoselectivity in
metabolism resulting in stereoselectivity in routes of excretion; enantiomeric
interactions in both metabolism and active transport processes; species variability in
enantiomeric metabolism and excretion together with the associated difficulty of species
selection for toxicological evaluation.
Following the oral administration of racemic ofloxacin to healthy volunteers the
serum concentration time profiles of the individual enantiomers are similar to those
obtained following determination of total drug concentrations (Okazaki et al., 1991a).
Small, but statistically significant, differences were observed between the enantiomers
in AUC (S > R), mean residence time (S > R) and both total and renal clearance
(R > S) but not in plasma protein binding or volume of distribution. As the drug
28 A. J. Hutt and J. O'Grady
Concluding comment
The above discussion has attempted to highlight the significance of stereochemical
considerations in the area of antimicrobial agents. In the current regulatory climate all
the components present in a medicinal product require justification and as has been
observed in other therapeutic areas, the introduction of single stereoisomers of both new
and existing chiral drugs is likely to increase (the so-called racemic switches). However,
such introductions are not without problems and may provide unexpected results.
Labetalol, an established combined a- and j3-blocking drug used in the treatment of
cardiovascular disease, contains two chiral centres and the marketed material is a
mixture of all four stereoisomeric forms. Of these stereoisomers the ^-blocking activity
resides in the R,R-isomer, the a-blocking activity in the S,/?-isomer and the remaining
pair are essentially inactive. Clinical trials with the single ^-blocking, R,R-\somtr,
named dilevalol, resulted in elevated liver function tests in a small number of patients.
This toxicity had not been observed with labetalol and resulted in the withdrawal of
the single isomer. Why such toxicity is not observed with the isomeric mixture is not
clear, but this example does illustrate that removal of the isomeric 'impurity' may not
be a trivial matter.
The decision to market a racemate, nonracemic isomeric mixture or single
stereoisomer depends on a number of factors, including technical feasibility, i.e.
production on an industrial scale, stereochemical stability, toxicological profile and the
clinical significance of the agent, i.e. the risk-benefit ratio. There are no simple answers
to the single stereoisomer versus isomeric mixture debate and each example must be
examined on a case-by-case basis. Several of the compounds cited above indicate the
potential problems that may arise during drug development. For example, the
epimerization of carbenicillin appears to be so rapid as to preclude the use of a single
isomer. Whereas, in the case of latamoxef, a compound with a half-life of epimerization
under physiological conditions only slightly shorter than the apparent serum
elimination half-life, the single isomer or mixture question is more difficult to answer.
In the case of the quinolones, particularly with respect to ofloxacin and its derivatives,
there can be little doubt of the significance of stereochemical considerations, particularly
in terms of providing an insight into the mechanism of action at a molecular level,
potency and selectivity.
References
Albers-Schonberg, G., Anson, B. H., Hensens, O. D., Hirschfield, J., Hoogsteen, K.., Kaczka,
E. A. et al. (1978). Structure and absolute configuration of thienamycin. Journal of the
American Chemical Society 100, 6491-9.
Chirality of antimicrobial agents 29
Fabro, S., Smith, R. L. & Williams, R. T. (1967). Toxicity and teratogenicity of optical isomers
of thalidomide. Nature 215, 269.
Finn, A., Straughn, A., Meyer, M. & Chubb, J. (1987). Effect of dose and food on the
bioavailability of cefuroxime axetil. Biopharmaceutics and Drug Disposition 8, 519-26.
Gerster, J. F., Rohlfing, S. R., Pecore, S. E., Winandy, R. M., Stern, R. M., Landmesser, J. E.
et al. (1987). Synthesis, absolute configuration and antibacterial activity of 6,7-dihydro-5,8-
dimethyl-9-fluoro-l-oxo-l//,5//-benzo[//]quinolizine-2-carboxylic acid. Journal of Medicinal
Chemistry 30, 839-43.
Gerster, J. F., Rohlfing, S. R., Rustad, N. J., Reiter, M. J., Pecore, S. E., Winandy, R. M. et al.
(1989). The synthesis and pharmacological profile of the stereoisomers of a tricyclic
quinolone antibacterial. In Quinolones, Proceedings of an International Telesvmposium
(Fernandes, P. B., Ed.), pp. 85-98. J. R. Prous Science, Barcelona.
Georgopapadakou, N. H., Dix, B. A., Angehrn, P., Wick, A. & Olson, G. L. (1987). Monocyclic
and tricyclic analogues of quinolones: mechanism of action. Antimicrobial Agents and
Chemotherapy 31, 614-6.
Graham, D. W., Ashton, W. T., Barash, L., Brown, J. E., Brown, R. D., Canning, L. F. et al.
(1987). Inhibition of the mammalian /Mactamase renal dipeptidase (dehydropeptidase-I)
by (Z)-2-(acylamino)-3-substituted-propenoic acids. Journal of Medicinal Chemistry 30,
1074-90.
Harding, S. M., Williams, P. E. O. & Ayrton, J. (1984). Pharmacology of cefuroxime as the
1-acetoxyethyl ester in volunteers. Antimicrobial Agents and Chemotherapy 25, 7882.
Hawkes, N. (1993). Lateral thinking. Times Magazine, 5th June, pp. 30-31.
Hayakawa, I., Atarashi, S., Kimura, Y., ICawakami, K., Saito, T., Yaiune, T. et al. (1991). Design
and structure-activity relationship of new Nrcis-2-fluorocyclopropyl quinolones. In Program
and Abstracts of the Thirty-First Interscience Conference on Antimicrobial Agents and
Chemotherapy, Chicago, 1991. Abstract 1504. American Society for Microbiology,
Washington, DC.
Hayakawa, I., Atarashi, S., Yokohama, S., Imamura, M., Sakano, K..-I. & Furukawa, M. (1986).
Synthesis and antibacterial activities of optically active ofloxacin. Antimicrobial Agents and
Chemotherapy 29, 163-4.
Hendel, J. & Brodthagen, H. (1984). Enterohepatic cycling of methotrexate estimated by use
of the D-isomer as a reference marker. European Journal of Clinical Pharmacology
26, 103-7.
Hoover, J. R. E. & Dunn, G. L. (1979). The /Mactam antibiotics. In Burger's Medicinal
Chemistry, 4th edn, part II (Wolff, M. E., Ed.), pp. 83-172. John Wiley, New York.
Hoshino, K., Hayakawa, I., Sato, K.., Sato, M., Kitamura, A. & Osada, Y. (1991a). Inhibitory
effects of DU-6859, a new fluorinated quinolone on type II topoisomerases. In Program and
Abstracts of the Thirtv-First Interscience Conference on Antimicrobial Agents and
Chemotherapy, Chicago, 1991. Abstract 1506. American Society for Microbiology,
Washington, DC.
Hoshino, K., Sato, K.., Akahane, K., Yoshida, A., Hayakawa, I., Sato, M. et al. (1991ft).
Significance of the methyl group on the oxazine ring of ofloxacin derivatives in the inhibition
of bacterial and mammalian type II topoisomerases. Antimicrobial Agents and Chemotherapy
35, 309-12.
Hutt, A. J. (1990). Enantiospecific analytical methodology: applications in drug metabolism
and pharmacokinetics. In Progress in Drug Metabolism, Vol. 12 (Gibson, G. G., Ed.),
pp. 257-361. Taylor and Francis, London.
Imamura, M., Shibamura, S., Hayakawa, I. & Osada, Y. (1987). Inhibition of DNA gyrase by
optically active ofloxacin. Antimicrobial Agents and Chemotherapy 31, 325-7.
Isert, D., Fischer, G., Klesel, N., Limbert, M., Markus, A., Riess, G. et al. (1992). Cefdaloxime
Pentexil Tosilate (HR 916 K): a diastereomerically pure novel oral cephalosporinester with
outstanding absorption characteristics. In Program and Abstracts of the Twenty-Second
Interscience Conference on Antimicrobial Agents and Chemotherapy, Anaheim, CA, 1992.
Abstract 188, p. 142. American Society for Microbiology, Washington, DC.
Kahan, F. M., K.ropp, H., Sundelof, J. G. & Birnbaum, J. (1983). Thienamycin: development
of iminpenem-cilastatin. Journal of Antimicrobial Chemotherapy 12, Suppl. D, 1-35.
Kim, H. S. & Campbell, B. J. (1982). /?-Lactamase activity of renal dipeptidase against
Chlrallty of antimicrobial agents 31
Parsons, W. H., Hajdu, R., Schoen, W. R., Combs, P. L., Sundelof, J. & Patchett, A. A. (1991).
A new class of potent, slowly reversible dehydropeptidase inhibitors. Biochemistry
International 23, 1107-15.
Pfaendler, H. R., Gosteli, J. & Woodward, R. B. (1979). The penems, a new class of 0-lactam
antibiotics. 4. Syntheses of racemic and enantiomeric penem carboxylic acids. Journal of the
American Chemical Society 101, 6306-10.
Ratcliffe, R. W., Wildonger, k . J., DiMichele, L., Douglas, A. W., Hajdu, R., Goegelman,
R. T. et al. (1989). Studies on the structures of imipenem, dehydropeptidase I hydrolyzed
imipenem and related analogues. Journal of Organic Chemistry 54, 65360.
Rauws, A. G. & Groen, K. (1994). Current regulatory (draft) guidance on chiral medicinal
products: Canada, EEC, Japan, United States. Chiralitv 6, 72-5.
Rebstock, M. C , Crooks, H. M., Controulis, J. & Bartz, Q. R. (1949). Chloramphenicol
(Chloromycetin) IV Chemical studies. Journal of the American Chemical Society 71, 2458-62.
Rosen, T., Chu, D. T. W., Lico, I. M., Fernandes, P. B., Shen, L., Borodkin, S. and Pernet,
A. G. (1988). Asymmetric synthesis and properties of the enantiomers of the antibacterial
agent 7-(3-aminopyrrolidin-l-yl)-1 -(2,4-difluorophenyl)-1,4-dihydro-6-fluoro-4-oxo-1,8-
naphthyndine-3-carboxylic acid hydrochloride. Journal of Medicinal Chemistry 31, 1586-90.
Sato, K., Hoshino, K. & Mitsuhashi, S. (1992a). Mode of action of new quinolones: the inhibitory
activity of DNA gyrase. In Progress in Drug Research, Vol.38, pp. 121-132. Birkhauser
Verlag, Basle.
Sato, K., Hoshino, K.., Tanaka, M., Hayakawa, I. & Osada, Y. (19926). Antimicrobial activity
of DU-6859, a new potent fluoroquinolone, against clinical isolates. Antimicrobial Agents and
Chemotherapy 36, 1491-8.
Sato, K., Hoshino, K., Tanaka, M., Hayakawa, I., Osada, Y., Nishino, T. et al. (1993).
Antibacterial activity of new quinolones and their modes of action. In Molecular Biology of
DNA Topoisomerases and its Application to Chemotherapy (Andoh, T., Ikeda, H. & Oguro,
M., Eds), pp. 167-76. CRC Press, Boca Raton.
Shen, L. L. (1993). Quinolone-DNA interaction. In Quinolone Antimicrobial Agents, 2ndedn
(Hooper, D. C. & Wolfson, J. S., Eds), pp. 77-95. American Society for Microbiology,
Washington, DC.
Shen, L. L. (1994). Molecular mechanisms of DNA Gyrase inhibition by quinolone antibacterials.
In DNA Topoisomerases: Biochemistry and Molecular Biology (Liu, L. F., Ed.) Advances in
Pharmacology, Vol. 29A, pp 285-304. Academic Press, London.
Shen, L. L., Bures, M. G., Chu, D. T. W. & Plattner, J. J. (1990). Quinolone-DNA interaction:
how a small drug molecule acquires high DNA binding affinity and specificity. In Molecular
Basis of Specificity in Nucleic AcidDrug Interactions (Pullman, B. & Jortner, J., Eds),
pp. 495-512. KJuwer Academic, Dordrecht.
Shen, L. L, Mitscher, L. A., Sharma, P N., O'Donnell, T. J., Chu, D. W. T., Cooper, C. S. et al.
(1989). Mechanism of inhibition of DNA gyrase by quinolone antibacterials: a cooperative
drug-DNA binding model. Biochemistry 28, 3886-94.
Tamai, I., Ling, H.-Y., Timbul, S. M., Nishikido, J. & Tsuji, A. (1988). Stereospecific absorption
and degradation of cephalexin. Journal of Pharmacy and Pharmacology 40, 320-4.
Testa, B., Carrupt, P.-A. & Gal, J. (1993). The so-called "interconversion" of stereoisomeric
drugs: an attempt at clarification. Chiralitv 5, 105-11.
Tucker, G. T. & Lennard, M. S. (1990). Enantiomer specific pharmacokinetics. Pharmacology and
Therapeutics 45, 309-29.
Une, T., Fujimoto, T., Sato, K. & Osada, Y. (1988). In vitro activity of DR-3355, an optically
active ofloxacin. Antimicrobial Agents and Chemotherapy 32, 1336-40.
Williams, K.. & Lee, E. (1985). Importance of drug enantiomers in clinical pharmacology. Drugs
30, 333-54.
Wise, R., Wills, P. J. & Bedford, K. A. (1981). Epimers of moxalactam: in vitro comparison of
activity and stability. Antimicrobial Agents and Chemotherapy 20, 30-2.
Yamada, H., Ichihashi, T., Hirano, K. & Kinoshita, H (1981). Plasma protein binding and
urinary excretion of R- and S-epimers of an arylmalonylamino-1-oxacephem I: in humans.
Journal of Pharmaceutical Sciences 70, 112-3.
Zak, O., Lang, M., Cozens, R., Konopka, E. A., Mett, H., Schneider, P. et al. (1988). Penems:
m vitro and in vivo experiments. Journal of Clinical Pharmacology 28, 128-35.
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: