SAR and species/stereo-selective metabolism of the sorbitol dehydrogenase inhibitor, CP-470,711.

Article abstract of DOI:10.1016/S0960-894X(02)00208-1

SAR studies on the stereoisomers of CP-470,711 suggested that in vivo epimerization was taking place in rats. Further metabolism studies revealed that no epimerization was occurring in dogs, and that no epimerization was expected in humans. A mechanism for the in vivo epimerization is proposed involving an oxidation-reduction pathway of the secondary benzylic alcohol, in contrast to an acid/base-promoted epimerization of the same center during chemical synthesis.

Full text of DOI:10.1016/S0960-894X(02)00208-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1477–1480
SAR and Species/Stereo-Selective Metabolism of the Sorbitol
Dehydrogenase Inhibitor, CP-470,711
Margaret Y. Chu-Moyer,* William E. Ballinger, David A. Beebe, James B. Coutcher,
Wesley W. Day, Jiancheng Li, Peter J. Oates and R. Matthew Weekly
Cardiovascular and Metabolic Diseases, Pfizer Global Research and Development, Groton Laboratories MS8220-3095,
Eastern Point Road, Groton, CT 06340, USA
Received 14 January 2002; accepted 13 March 2002
Abstract—SAR studies on the stereoisomers of CP-470,711 suggested that in vivo epimerization was taking place in rats. Further
metabolism studies revealed that no epimerization was occurring in dogs, and that no epimerization was expected in humans. A
mechanism for the in vivo epimerization is proposed involving an oxidation–reduction pathway of the secondary benzylic alcohol,
in contrast to an acid/base-promoted epimerization of the same center during chemical synthesis. # 2002 Elsevier Science Ltd. All
rights reserved.
Diabetes mellitus afflicts approximately 135 million
people worldwide, with an estimated diabetic popula-
tion of approximately 16 million in the United States
alone. With its complications, diabetes is the sixth-leading
cause of death by disease. These hyperglycemia-related
diabetic complications include neuropathy, nephro-
pathy, retinopathy and cardiovascular disease which can
lead to lower-limb amputations, end-stage renal failure,
loss of vision and myocardial infarction, respectively.1,2
CP-470,711 incorporates two chiral hydroxyethyl side
chains of the (R)-configuration on each of its pyrimidine
rings. Eariler SAR of molecules in this class showed that
pyrimidines bearing only one (R)-hydroxyethyl side
chain showed potent in vitro activity, while compounds
containing only one (S)-hydroxyethyl side chain pos-
sessed lower activity.5 Because of the fact that
CP-470,711 bears two hydroxyethyl side chains, its
(R,S)-, (S,R)-, and (S,S)-isomers were prepared to fur-
ther understand this SAR (Scheme 1).
Recently, we reported the synthesis and SAR leading up
to the identification of CP-470,711 (1), a potent inhi-
bitor of sorbitol dehydrogenase (SDH), the second
enzyme in the polyol pathway.3 The contribution to
nerve complications from the redox imbalance
(" NADH/NAD+ ratio) caused by increased flux
through this step in the pathway in the diabetic state
was shown earlier with a prototype inhibitor.4
Previously, we had shown that the condensation of
mono-benzyl protected piperazine 2 with reactive
pyrimidines, such as triflate 3, resulted in mixtures of
4R/4S.3 In the particular case when refluxing CH3CN
was employed, a 9:1 mixture was obtained. These enan-
tiomers are easily separated by chiral HPLC, and the
required (R)- or (S)-isomer can be carried on to sub-
sequent steps. In the event, removal of the benzyl pro-
tecting group from 4R or 4S under transfer
hydrogenolysis conditions provided 5R and 5S, respec-
tively. Subsequent condensation with the appropriate
chloropyrimidine,6 6R or 6S, and cleavage of the buty-
rate protecting groups, provided the target compounds,
1 and, 7–9, in >95% ee by chiral HPLC.7
In vitro evaluation3 of these analogues against both
human and rat SDH showed that those compounds
which contained at least one pyrimidine bearing the
(R)-ethanol side chain (1, 7 and 8) had similar activities.
*Corresponding author. Fax: +1-860-715-9628; e-mail: margaret_
y_chu-moyer@groton.pfizer.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00208-1

1478
M. Y. Chu-Moyer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1477–1480
Scheme 2. (a) MnO2, DCE, reflux, 7 h; (b) NaBH4, MeOH, rt, 2 h.
Scheme 1. (a) CH3CN, reflux, 15 h; (b) Chiracel AD, 90:10 hexane/
i-PrOH+1% diethylamine; (c) 10% Pd/C, HCO2NH4, HCl, MeOH,
.
reflux, 30 min; (d) Et3N, i-PrOH, reflux, 18 h; (e) LiOH H2O,
3:1 MeOH/H2O, rt, 2 h.
However, the compound bearing two (S)-ethanol side
chains (9) had ꢀ30- to 40-fold lower activity, consistent
with experience in our pyrimidine SDI series3,5 (Table
1). Surprisingly, when these compounds were assayed in
both the acute and chronic in vivo diabetic rat models,3
all the derivatives had similar activity, including the
(S,S)-isomer 9. This result suggested that in vivo con-
version of the isomers was occurring, possibly through
an oxidation/reduction mechanism. Therefore, several
new analogues were synthesized to evaluate this pro-
posed hypothesis.
Scheme 3. (a) MnO2, DCE, reflux, 18 h; (b) Tf2O, pyr, CH2Cl2, 0 ꢂC,
15 min; (c) Et3N, THF, rt, 45 min; (d) 12 N HCl/MeOH, rt, 18 h; (e)
.
LiOH H2O, 4:1 MeOH/H2O, rt, 1 h; (f) Et3N, i-PrOH, reflux, 15 h; (g)
MnO2, DCE, reflux, 19 h; (h) K2CO3, MeOH, rt, 4 h.
oxidized to ketone 13 and subsequently converted to
triflate 14 under standard conditions. Coupling with
piperazine 5R provided 15, which upon deprotection of
the butyrate group under acidic conditions gave mono-
ketone 16. The isomeric mono-ketone was prepared via
first removal of the butyrate ester from compound 5R to
provide free alcohol 17 followed by reaction with
chloropyrimidine 6R. The resulting differentially pro-
tected alcohol 18 was oxidized to butyrate mono-pro-
tected ketone 19. The butyrate protecting group was
then subsequently cleaved under basic conditions to give
the desired compound 20.
Diketone 10 was prepared via bis-oxidation of 1 under
the action of MnO2 in refluxing dichloroethane (Scheme
2). Bis-reduction of the diketone with NaBH4 then pro-
vided 11, the 1:1:1:1 mixture of all four isomers.
The two mono-ketone isomers were synthesized as fol-
lows (Scheme 3). (ꢁ)-Hydroxyethyl-pyrimidone 126 was
As expected in vitro, the diketone 10 is less potent by
ꢀ10- to 20-fold; the mixture of four isomers, compound
11, shows good activity which is consistent with 3/4 of
the material being active; and the mono-ketones 16 and
20, which each contain a pyrimidine bearing the
(R)-ethanol side chain, possess potent in vitro activity
against both human and rat SDH. However, once again,
all have similar activities in vivo suggesting that these
analogues are also possible intermediates in the biologi-
cal interconversion of 1 in the biological milieu of the
rat.
Table 1. Comparison of in vitro and in vivo activity for compounds
1, 7–9, 10, 11, 16 and 20
Compd
h-SDHa
IC50
(nM)
r-SDHb
IC50
(nM)
Acute assayc
ED90
(mg/kg/d)
Chronic assayd
% norm. @
1 mg/kg
1
7
8
9
10
11
16
20
10
33
19
831
212
28
19
6
17
29
27
640
227
8
0.2
0.4
0.8
0.7
0.9
0.6
0.2
0.5
90
81
81
84
92
92
94
77
18
5
In order to confirm the above pharmacological findings,
HPLC analysis of rat plasma samples was performed. An
achiral method8 was utilized to determine the compo-
sition of the mixture [e.g., presence or absence of di-
alcohol (11), mono-ketone (16 and 20) and di-ketone
(10)] and a chiral method9 was used to determine the
ratio of the di-alcohol stereoisomers (1, 7, 8 and 9).
aInhibition of human recombinant sorbitol dehydrogenase.
bInhibition of recombinant rat sorbitol dehydrogenase.
cPrevention of sciatic nerve fructose accumulation in streptozotocin
(STZ) diabetic rats (tid dosing, 4-h post STZ).
dReversal of sciatic nerve fructose accumulation in STZ rats (qd
dosing, 5-days post STZ).

M. Y. Chu-Moyer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1477–1480
1479
species that is non-selectively reduced to give a statis-
tical mixture of di-alcohols.
A comparable study in dogs revealed that no apparent
epimerization was taking place upon oral administra-
tion (30 mg/kg) of compound 1 (Fig. 2a). As well, there
was no evidence of formation of any mono-ketone or
di-ketone intermediates. There are two possible expla-
nations for apparent lack of epimerization of 1 in dog:
(1) the alcohols are not oxidized by the dog or (2) the
alcohols are indeed oxidized by the dog, but stereo-
specific reduction back to the (R)-alcohol takes place to
result in no net stereochemical change. In order to fur-
ther study the mechanism, di-ketone 10 was given orally
to dogs (10 mg/kg). Interestingly, the di-ketone was
indeed reduced to the di-alcohol, but chiral HPLC ana-
lysis revealed that it was completely of the (S,S )-con-
figuration (Fig. 2b). These data suggest that the lack of
epimerization seen in the dog is due to the inability of
the dog to oxidize the alcohol to the ketone, for if this
did occur, only compound with (S,S )-stereochemistry
would be detected.
Figure 1. Rat drug metabolism studies. Data were obtained at 1 h
post-dose for all experiments. (a) Achiral metabolism of compound 1;
(b) Achiral metabolism of compound 10; (c) distribution of the di-
alcohol stereoisomers 1, 7, 8 and 9 after dosing with compound 1.
As it became apparent that the oxido-reductive meta-
bolism of compound 1 was conspicuously different in
the rat versus the dog, we were interested in determining
which of these species would be a predictor for meta-
bolism of 1 in humans. In order to do so, a relevant in
vitro system was required. After extensive screening for
marker cells, only rat and dog hepatocytes were found
to mimic the metabolic pattern observed in the rat and
dog in vivo studies. That is, whereas metabolic turnover
of the chiral hydroxyethyl group was observed in the rat
hepatocytes,10 no turnover was seen with the dog hepato-
cytes11 (Fig. 3a). Human hepatocyte experiments revealed
that, like dog hepatocyte experiments, no metabolism of
the hydroxyethyl group was taking place (Fig. 3b). To the
extent that the metabolism of 1 in humans would be
entirely mediated by hepatocytes, it is expected that results
in humans would mirror the findings in dogs.
Figure 2. Dog drug metabolism studies. Data were obtained at 1 h
post-dose for all experiments. (a) Distribution of the di-alcohol
stereoisomers after dosing with compound 1; (b) distribution of the
di-alcohol stereoisomers after dosing with compound 10.
During the synthesis of the sorbitol dehydrogenase
inhibitor CP-470,711 (1), epimerization of the secondary
alcohol stereocenter was observed. Efforts to further
study this reaction as well as to characterize the isomers
arising from the two chiral centers present in 1 led to the
synthesis of 7, 8 and 9. Although in vitro activity was
consistent with known SAR, the in vivo results suggested
that these compounds were undergoing epimerization
through an oxidation–reduction mechanism. Various
redox isomers of 1 were also synthesized and pharmacolo-
gical data on these analogues supported the above finding.
Figure 3. (a) Comparison of in vivo rat/dog chiral metabolism of
compound 1 with in vitro chiral metabolism in rat/dog hepatocytes;
(b) in vitro chiral metabolism of compound 1 in human hepatocytes.
In the event, when rats were dosed with 1 (20 mg/kg,
po), achiral HPLC showed mainly the presence of di-
alcohol 11 with a very small amount of di-ketone 10 and
mono-ketones 16 and, 20, suggesting that oxidation and
reduction could be occurring to some extent (Fig. 1a).
When rats were dosed with di-ketone 10 (5 mg/kg, po),
di-alcohol 11 was again the major species in plasma
with a small amount of di-ketone 10 and mono-ketones
16 and 20 present (Fig. 1b). Analysis of the chiral com-
position of the di-alcohol from the former experiment
revealed that all four isomers, 1, 7, 8 and 9, were present
in nearly equal quantities (Fig. 1c). These results
strongly suggest the intermediacy of a transient ketone
Direct analysis of rat plasma revealed that in vivo chiral
metabolism of 1 was indeed taking place, very likely
through the intermediacy of ketone analogues 10, 16
and 20. However, in dogs, this metabolism was not
observed. Further mechanistic studies showed that dogs
did not oxidize the alcohol to the ketone and therefore,
‘isomerization’ of the alcohol stereocenter did not
occur. In vitro hepatocyte studies reflected the results in
rat and dog, thus suggesting that metabolism of 1 in
humans would be akin to that observed in dog.

1480
M. Y. Chu-Moyer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1477–1480
8. Serum samples (300 mL) were collected at selected time
References and Notes
points and combined with an internal standard. This mixture
was diluted with distilled H2O (1 mL), extracted with EtOAc
(5 mL), evaporated to dryness and reconstituted in mobile
phase (150 mL). These extracts (120 mL) were analyzed by
1. American Diabetes Association, Diabetes Facts and Fig-
ures: http://www.diabetes.org/ada/facts.asp.
2. (a) Nathan, D. M. N. Engl. J. Med. 1993, 328, 1676. (b)
Clark, C. M., Jr.; Lee, D. A. N. Engl. J. Med. 1995, 332, 1210.
3. Chu-Moyer, M. Y.; Ballinger, W. E.; Beebe, D. A.; Berger,
R.; Coutcher, J. B.; Day, W. W.; Li, J.; Mylari, B. L.; Oates,
P. J.; Weekly, R. M. J. Med. Chem. 2002, 45, 511.
4. Tilton, R. G.; Chang, K.; Nyengaard, J. R.; Van den
Enden, M.; Ido, Y.; Williamson, J. R. Diabetes 1995, 44, 232.
5. Mylari, B. L.; Oates, P. J.; Beebe, D. A.; Brackett, N. S.;
Coutcher, J. B.; Dina, M. S.; Zembrowski, W. J. J. Med.
Chem. 2001, 44, 2695.
reverse-phase HPLC using
a
Kromasil C4 column
(250ꢃ4.6 mm) and elution with a 1:1 MeOH/H2O mobile
phase containing 10 mLPic-B7/liter solution. The flow rate
was set at 1.0 mL/min and UV detection at 237 nM was used.
9. Serum samples (300 mL) were collected at selected time
points and combined with an internal standard. This mixture
was diluted with distilled H2O (1 mL), extracted with EtOAc
(5mL), evaporated to dryness and reconstituted in mobile phase
(150 mL). These extracts (120 mL) were analyzed by normal-
6. Chu-Moyer, M. Y.; Murry, J. A.; Mylari, B. L.; Zem-
browski, W. J. WO0059510, CAN 133:281794. The (S)-isomer
was synthesized from (S)-lactamide in a manner analogous to
the (R)-isomer.
phase chiral HPLC using
a
Chiralpak AS column
(250ꢃ4.6 mm) and elution with a 1:1 EtOH/hexane mobile
phase. The flow rate was set at 1.0 mL/min and UV detection
at 254 nM was used.
7. We gratefully acknowledge Stephen M. Chesnut and Ste-
phen M. Brown (PGRD Analytical Chemistry) for developing
the chiral HPLC assay and evaluating these compounds
according to the following conditions: Chiralpak AS
(250ꢃ4.6 mm), 85:15 hexane/EtOH, T=40 ꢂC, 1.0 mL/min,
UV detection at 254 nM.
10. Compound 1 (6.7 mM) was incubated with rat hepatocytes
for 2 h. Samples were prepared and assayed as described in ref
9.
11. Compound 1 (6.7 mM) was incubated with dog hepatocytes
for 4 h. Samples were prepared and assayed as described in
ref 9.