Sorbitol dehydrogenase inhibitors (SDIs): A new potent, enantiomeric SDI, 4-[2-1R-hydroxy-ethyl)-pyrimidin-4-yl]-piperazine-1-sulfonic acid dimethylamide

Article abstract of DOI:10.1021/jm0102001

We report here on our medicinal chemistry and pharmacology efforts to provide a potent sorbitol dehydrogenase inhibitor (SDI) as a tool to probe a recently disclosed hypothesis centered on the role of sorbitol dehydrogenase (SDH) in the second step of the polyol pathway, under conditions of high glucose flux. Starting from a weak literature lead, 2, and through newly developed structure-activity relationships, we have designed and executed an unambiguous synthesis of enantiomeric SDI, 6, which is at least 10 × more potent than 2. Also, 6 potently inhibits SDH in streptozotocin-diabetic rat sciatic nerve. We have described an expedient synthesis of a key building template, 33, for future research in the SDI area that may facilitate the discovery of even more potent SDIs with longer duration of action in vivo.

Full text of DOI:10.1021/jm0102001

J . Med. Chem. 2001, 44, 2695-2700
2695
Expedited Articles
Sor bitol Deh yd r ogen a se In h ibitor s (SDIs): A New P oten t, En a n tiom er ic SDI,
4-[2-1R-Hyd r oxy-eth yl)-p yr im id in -4-yl]-p ip er a zin e-1-su lfon ic Acid
Dim eth yla m id e
Banavara L. Mylari,* Peter J . Oates, David A. Beebe, Nathaniel S. Brackett, J ames B. Coutcher,
Michael S. Dina,† and William J . Zembrowski
Department of Cardiovascular and Metabolic Diseases, Pfizer Global Research & Development, Pfizer Inc.,
Groton, Connecticut 06340
Received May 4, 2001
We report here on our medicinal chemistry and pharmacology efforts to provide a potent sorbitol
dehydrogenase inhibitor (SDI) as a tool to probe a recently disclosed hypothesis centered on
the role of sorbitol dehydrogenase (SDH) in the second step of the polyol pathway, under
conditions of high glucose flux. Starting from a weak literature lead, 2, and through newly
developed structure-activity relationships, we have designed and executed an unambiguous
synthesis of enantiomeric SDI, 6, which is at least 10× more potent than 2. Also, 6 potently
inhibits SDH in streptozotocin-diabetic rat sciatic nerve. We have described an expedient
synthesis of a key building template, 33, for future research in the SDI area that may facilitate
the discovery of even more potent SDIs with longer duration of action in vivo.
Sustained elevation of blood glucose is the hallmark
of diabetes. Long-term diabetes results in maladies such
as peripheral neuropathy, nephropathy, and retino-
pathy. In clinical practice these disorders could eventu-
ally lead to one or more devastating consequences which
F igu r e 1.
include lower limb amputation, end-stage renal failure,
and loss of vision. In addition to human suffering, these
many of the early functional changes observed in
experimental diabetes, including alterations in vascular
albumin permeation, tissue blood flow, and function,
e.g., nerve conduction velocity (NCV). In turn, these
early functional changes may contribute to eventual
pathology in affected diabetic tissues.
complications impose a heavy burden both on individual
and national medical budgets. Two landmark studies,
the Diabetes Complications and Control Trial1 and the
United Kingdom Prospective Diabetes Study2 have
demonstrated both the benefits of lowering blood glucose
through intensive insulin therapy and the strong con-
nection between diabetic complications and chronic
excess circulating blood glucose. Over the last several
years, considerable effort has focused on developing an
oral therapy that would protect vulnerable diabetic
tissues from the detrimental effects of elevated blood
glucose. One of the most investigated approaches has
been to block excess glucose metabolism through the
first step of the polyol pathway (Figure 1) via aldose
reductase inhibitors3 (ARIs).
A significant aspect of the pseudohypoxia theory is
that it provides an exciting framework within which to
explore the potential consequences of selectively block-
ing the second step of the polyol pathway with a specific
sorbitol dehydrogenase inhibitor (SDI). Such an agent
would accentuate sorbitol concentrations while tending
to correct the tandem increases in NADH/NAD+ and
fructose levels in diabetic tissues. Obrosova et al.5
confirm the significant inhibition of fructose in diabetic
rats with SDI 157 (same as 1, a prodrug of 2), but did
not detect any changes in cytosolic NAD+/NADH. Seem-
ingly confounding results have been obtained in NCV
studies with a prototype SDI, 26-9 (Chart 1), and
genetically altered mice.10 However, discrepant results
with SDI 2 could be based on the use of different models
and experimental protocols (for example, prevention vs
reversal) and duration of dosing with 2, which is a weak
and short-lived SDI (vide infra). To provide a better tool
for future studies, our initial thrust has been to address
the SDI potency issue. Herein, we describe our efforts
leading to the discovery of enantiomeric SDI, 4-[2-(1R-
hydroxy-ethyl)-pyrimidin-4-yl]-piperazine-1-sulfonic acid
Recently, Williamson et al.4 have focused attention
on another aspect of the polyol pathway. Excess glucose
flux through the polyol pathway is proposed to create
an imbalance in the cytoplasmic redox status (increased
free NADH/NAD+ ratio) because of rapid consumption
of NAD+ during the oxidation of sorbitol to fructose by
sorbitol dehydrogenase (SDH) (Figure 1). They propose
that the reductive stress (“pseudohypoxia”) triggers a
cascade of biochemical changes which may underlie
* To whom correspondence should be addressed. Address: B. L.
Mylari, 6 Quinley Way, Waterford, CT 06385. Telephone: 860-443-
3085. Fax: 860-686-0620. E-mail address: lmylari@home.com.
† Current address: Genentech, San Francisco, CA.
10.1021/jm0102001 CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/17/2001

2696 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 17
Mylari et al.
Ch a r t 1
Sch em e 1a
Ch a r t 2
a
Reagents and conditions: (a) TEA/DMF; (b) AlBN, 80 °C; (c)
OsO4/NaIO4, dioxane/H2O, (d) NaBH4/EtOH.
dimethylamide, 6, which is significantly more potent
than 2, both in vitro and in vivo. We believe that the
recently disclosed rat pharmacokinetic and metabolism
results with 2,11,12 and our discovery of 6 coupled with
development of a practical preparation procedure for a
key building template, 33, have set the stage for future
clarification of the role of SDH in normal and diabetic
tissues. These could include both potential benefits6-9
and liabilities6,13 of SDIs and examination of the poten-
tial consequences of sequentially blocking both AR and
SDH, through ARI+SDI combinations.14
Geisen et al.6 report that 2 inhibits sheep liver SDH
and has a Ki of 92 µM for the oxidation of sorbitol to
fructose and causes a dose-dependent increase in the
sorbitol concentration in the sciatic nerve and in the
erythrocytes of normal and streptozotocin-diabetic rats.
We characterized the in vitro effects of 2 on both rat
and human SDHs and the in vivo effects of 2 in our
diabetic rat model. Both SDHs have been cloned,15 and
the recombinant enzymes have been made available to
us.16 We have confirmed that 2 is a moderately potent
inhibitor of the rat SDH and that it also inhibits human
SDH. Its IC50 values against the enzymes, respectively,
are 225 and 246 nM. It is exquisitely selective for SDH.
Against other dehydrogenases, including lactate, alco-
hol, fructose, and glyceraldehyde-3-phosphate, as well
as aldose reductase, it showed no effect even at 50 µM
(data not shown). The calculated oral ED50 from the
dose-response for inhibition of nerve fructose in strep-
tozotocin-diabetic rats is 14 mg/kg.
F igu r e 2.
SDH, like the well-studied liver alcohol dehydro-
genase (ADH), is a zinc-containing enzyme.17 It has been
speculated that the hydroxymethyl group of 2 serves as
a strong ligand to the zinc atom, mimicking the hy-
droxymethyl group of sorbitol.18 While we agree with
this speculation, we suggest that the proposed model is
incorrect in regard to the depicted N3 nitrogen ligand
to the zinc atom, because neither 1119 nor 12 (Scheme
1) show any SDH inhibition activity, even at the highest
dose tested, 10 µM. Also, it appears reasonable to expect
the less sterically hindered N-1 atom of 2 to be a better
co-ligand than the N-3 atom. The crucial role of the
hydroxymethyl group as a cooperative zinc ligand is
entirely congruent with the extremely poor SDI activi-
ties of 1 and 3 (IC50 > 10 µM).
On the basis of the above observations and the results
of the site-directed mutation on structure-function
carried out by Hoog et al.,20a,b we propose a revised
model shown in Figure 2. Definitive proof for this model
will have to await data from X-ray crystal structure
studies on SDH and complexes of SDH-SDI.
According to pharmacokinetic (PK) data11 obtained
with normal rats following oral administration of 2, it
has a very short t1/2 (<30 min) both in serum and sciatic
nerve. Rats fairly rapidly metabolize 2 to the corre-
sponding N-monodemethylated compound,12 10 (Chart
2), which is a very weak SDI (IC50 > 10 µM). The
excellent oral absorption of 2 as reflected in a high
serum concentration and its efficient distribution into
a highly protected nerve tissue are consistent with its
pKa (6.1), log P (0.55), high aqueous solubility (3 mg/
mL), and very low protein binding (<1%).
Because the specific conformation of the hydroxy-
methyl in 2 would be expected to have a significant
impact on its zinc-liganding strength, we examined the
effect of chirality at the hydroxymethyl carbon on SDH
inhibition potency. As a test case, we first prepared 4
(racemate) to make sure that it is at least as potent an
SDI as 2. It was prepared by oxidation of 2, under Swern
conditions, to the corresponding aldehyde 19 and then
allowing it to react with methyl Grignard (Scheme 2).
It was found that 19 was at least as potent as 2 with
an IC50 of 173 nM. Spurred by this finding, 19 was

Sorbitol Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 17 2697
Sch em e 2a
Ta ble 1. In Vitro and in Vivo Data for SDIs
a
entry
IC50 (h-SDH) (nM)
ED50 (mg/kg)
2
4
5
6
246 ( 35
173 ( 50
486 ( 24
27 ( 4
14.1 ( 1.6
2.4 ( 0.08
9.0 ( 1.4
1.6 ( 0.1
a
For inhibition of diabetic rat sciatic nerve fructose accumula-
tion.
a
Reagents and conditions: (a) (COCl)2/DMSO; (b) MeMgBr/
mixture of RR and RS diastereomeric esters, 29 and
29a . The structural assignment as well as enantiomeric
purity of these esters was determined from their NMR
spectra, by monitoring the signal for the methyl group
attached to the chiral carbon. All the esters showed a
pair of doublets at 1.61 and 1.68 ppm. The ratio of the
area under the signals for 4, 5, and 6, respectively, was
1:1 (as expected), 97:3, and 4:96. Results of in vitro
testing of the R and S enantiomers are shown in Table
1. The R enantiomer, 6, was found to be a significantly
more potent inhibitor (>10×) than the S enantiomer,
5, as well as 2 against h-SDH.
THF; (c) lipase P30/DME.
subjected to a lipase-mediated resolution of enanti-
omers, using lipase P30. Reaction monitoring showed
the formation of a less polar compound. Separation of
the reaction mixture gave 7, which showed optical
activity with a positive rotation. Base hydrolysis of 7
gave 6. A second component isolated from the reaction
mixture was 5 with an optical rotation equal in mag-
nitude, but of opposite sign relative to 6.
To ascertain the absolute configuration of the hy-
droxyethyl group of 6, we embarked on its total syn-
thesis starting from the methyl ester of R(+) lactic acid,
20 (Scheme 3). Through the first four steps, 20 was
converted to the known amide 21.21 Dehydration of 21
using P2O5 gave the corresponding nitrile, 22, which
served as a precursor to the desired amidine 15, via the
intermediate imidate, 23. A cyclo-condensation reaction
of 24 with 25 gave 26, which was converted to the
corresponding sulfonate ester 27. Displacement of the
sulfonate moiety by 14 gave 28. Demethylation of 28
gave the dextrorotatory 6, thus unambiguously estab-
lishing the R configuration for the chiral center in 6.
To be sure that the chiral center integrity was main-
tained through the synthetic steps, we reacted 4, 5
(Scheme 2), and 6 (Scheme 3) with R-R-methoxy-R-
trifluoromethylphenyl acetyl chloride and obtained a
At this juncture, we needed a more expedient syn-
thetic route to prepare larger supplies of 6 for testing
in our diabetic rat models. We therefore prepared
racemic 2-(1-hydroxyethyl)-3H-pyrimidin-4-one, 30, to
investigate the possibility of P30 lipase-mediated enan-
tiospecific acylation (Scheme 4A). On the basis of an
earlier result with 4, we expected that reaction of 30
with lipase P30 would, selectively or exclusively, give
an acylated alcohol of R configuration. We are pleased
to report the experimental design and execution of the
successful approach. Pyrimidone 30 is freely soluble in
water. Reaction of 30 with the lipase using acetic
anhydride as the acyl donor gave the (R)-acetate, 31,
which was also water soluble. So, we had to resort to
cumbersome separation procedures to isolate the (R)-
Sch em e 3a
a
Reagents and conditions: (a) Ag2O/MeI; (b) aq KOH; (c) SOCl2; (d) NH3; (e) P2O5; (f) HCl/EtOH; (g) NH3; (h) H2O; (i) ClSO2Me; (j)
THF/Et3N; (k) BBr3; (l) Py-CH3CN.

2698 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 17
Mylari et al.
Sch em e 4a
dehydrogenase and on our SAR data around 2 strongly
suggests that the pyrimidine N-1 atom and the oxygen
atom of the hydroxymethyl group of 2 function as key
ligands to the zinc atom in SDH. This rationale led to
the synthesis of the enantiospecific inhibitor, 6, which
is substantially more potent in vitro than 2 against
human SDH and also more potent than 2 in inhibiting
nerve fructose production in diabetic rats. This new and
more potent SDI could serve as a tool to further probe
the second step of the polyol pathway. To address the
rapid N-demethylation of the sulfamoyl moiety by rats,
we are now pursuing other modifications of 6, based on
the versatile pyrimidine template, 33. We have devel-
oped a highly practical route for the preparation of 33,
which facilitated in vivo evaluation of 6 as well as has
laid a strong foundation to pursue structure-activity
relationships for discovery of highly potent in vivo SDIs
with longer serum half-life.
Exp er im en ta l Section 22
4-(2-For m yl-p yr im idin -4-yl)-piper a zin e-1-su lfon ic Acid
Dim eth yla m id e (19). To a solution of 4-[2-hydroxy-methyl)-
pyrimidin-4-yl]-piperazine-1-sulfonic acid dimethylamide (12.12
g, 40.2 mmol) in methylene chloride (100 mL) was added a
solution of oxalyl chloride (3.9 mL, 44.2 mmol), and the
reaction mixture was cooled to -78 °C. To this was added a
solution of DMSO (6.27 mL, 88.4 mmol) in methylene chloride
(20 mL) dropwise so as to keep the reaction temperature from
rising above -70 °C. After 2 h, triethylamine (28.0 mL, 88.4
mmol) was added dropwise, and the reaction was allowed to
come to room temperature. The reaction mixture was diluted
with water (200 mL), and the collected organic layer was
washed with saturated sodium bicarbonate solution. The
washed organic layer was collected, dried over sodium sulfate,
and filtered, and the filtrate was evaporated to a solid. The
solid was triturated with ether (20 mL), and the mixture was
filtered to obtain the title compound (10.84 g, 94%): mp 124-
127 °C; 1H NMR (CDCl3, 300 MHz) δ 2.88 (s, 6H), 3.31 (m,
4H), 3.99 (m, 4H), 6.4 (d, J ) 8 Hz, 1H), 8.2 (d, J ) 8 Hz, 1H),
9.35 (s, 1H).
4-[2-(1RS-Hyd r oxy-eth yl)-p yr im id in -4-yl]-p ip er a zin e-
1-su lfon ic Acid Dim eth yla m id e (4). To a -5 °C cooled
solution of 4-(2-formyl-pyrimidin-4-yl)-piperazine-1-sulfonic
acid dimethylamide (419 g, 14 mmol) in dry tetrahydrofuran
(100 mL) was added an ether solution of methylmagnesium
bromide (7.6 mL, 23.1 mmol). After 20 min, the reaction was
allowed to warm to room temperature and then refluxed for
30 min. After cooling the reaction to room temperature, it
quenched with saturated ammonium chloride solution and
then extracted with ethyl acetate (2 × 50 mL). The organic
layer was collected, dried over sodium sulfate, and filtered,
and the filtrate was evaporated to a pale yellow solid (4.21 g,
95%): mp 115-116 °C; 1H NMR (CDCl3, 300 MHz) 1.48 (d, J
) 7 Hz, 6H), 2.83 (s, 3H), 3.31 (m, 4H), 3.74 (m, 4H), 4.70 (q,
J ) 8 Hz), 1H), 6.40 (d, J ) 8 Hz, 1H), 8.21 (d, J ) 8 Hz, 1H);
MS (CI) M+1316.
R-4-(4-Dim et h ylsu lfa m oyl-p ip er a zin -1-yl)-p yr im id in -
2-ylm eth yl Acetate (7) an d 4-[2-(1S-Hydr oxy-eth yl)-pyr im -
id in -4-yl]-p ip er zin e-1-su lfon ic Acid Dim eth yla m id e (5).
To a solution containing 4-[2-(1RS-hydroxy-ethyl)-pyrimidin-
4-yl]-piperazine-1-sulfonic acid dimethylamide (4, 0.9 g, 2.86
mmol), dimethoxyethane (5.7 mL), and vinyl acetate (10.5 mL)
was added Lipase P30 (90 mg, 10%), and the reaction mixture
was stirred at room temperature for 14 days. It was filtered,
and the filtrate was evaporated to brown oil, which was
chromatographed over silica gel. Elution with a 95:5 mixture
of ethyl acetate and methanol and evaporation of the eluent
gave the title compound (220 mg, 43%): mp 107-109 °C, [R]D
+41.9 (c ) 1, methanol); 1H NMR (CDCl3, 300 MHz) δ 1.55(d,
J ) 6H, 6H), 2.2(s, 3H), 2.9(s, 6H), 3.3 (m, 4H), 3.8(m, 4H),
6.35(q, J ) 8 Hz,1H), 6.4(d, 1H), 8.25(d, 1H). Subsequent
a
Reagents and conditions for part A: (a) H2O/rt; (b) CH2dCH-
O-(CdO)-R/dioxane, 50 °C. For part B: (a) ClSO2Me; (b) THF/
TEA; (c) HCl (conc).
acetate. We thought that the water solubility of pyrimi-
done 30 could be exploited to our advantage if we could
make the acyl derivative sufficiently lipophilic for facile
and complete extraction by organic solvents, which
would avoid the need for any separation procedures. Use
of propionic anhydride gave the propionate, 32, which
was still water soluble. Surprisingly, it turned out that
butyric anhydride was the optimum choice not only
because it led to the butyrate, 33, with the desired cri-
teria (easily and completely extracted by ethyl acetate)
but also because the rate of acylation by higher anhy-
drides including the next homologue, valeric anhydride,
was extremely slow. Now using the butyric anhydride
procedure, we have prepared 33 in kilogram quantities.
This easy access to 33 has provided an efficient route
to 6 (Scheme 4B), and has opened up opportunities for
further SAR development in this chemical family.
In our diabetic rat model, 6 was found to be more
potent than either 5 or 2 in inhibiting sciatic nerve
fructose accumulation. Thus we have provided a SDI
that is ∼10× more potent than the literature lead, 2,
both in vitro and in vivo. While 5 is less potent in vitro
than 2, it is nevertheless slightly more potent in vivo
than 2. This can be attributed to the higher lipophilicity
of 5, which could result in its more efficient nerve tissue
penetration. We have not yet determined the serum
half-life of 6. However, the finding12 that 2 is fairly
rapidly metabolized by the rat suggests that 6 would
undergo a similar metabolic fate. Compound 6 is
certainly a more potent tool for further understanding
the effects of SDIs. However, we are now focusing on
discovery of even more potent SDIs with a longer serum
half-life, and the progress we achieve in this endeavor
will be the subject of future publications.
Con clu sion
SDI 2 inhibits the accumulation of fructose in the
sciatic nerve of severely diabetic rats. Design of more
potent SDIs based on extant literature on alcohol

Sorbitol Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 17 2699
elution fractions were combined and evaporated to obtain 5
(45 mg, 9%): [R]D +15.9 (c ) 1, methanol); H NMR (CDCl3,
300 MHz) 1.48 (d, J ) 7 Hz, 6H), 2.83 (s, 3H), 3.31 (m, 4H),
3.74 (m, 4H), 4.70 (q, J ) 8 Hz), 1H), 6.40 (d, J ) 8 Hz, 1H),
8.21 (d, J ) 8 Hz, 1H); MS (CI) M+1 316.
+54.4° (c ) 1, methanol); 1H NMR (CDCl3, 300 MHz) δ 1.48(d,
3H), 2.89(s,6H), 3.31(m, 4H), 3.37(s, 3H), 3.78(m, 4H), 4.34(q,
J ) 8 Hz, 1H), 6.41(d, J ) 7.5 Hz, 1H), 8.30(d, J ) 7.5 Hz,
1H).
1
4-[2-(1R-Hyd r oxy-eth yl)-p yr im id in -4-yl]-p ip er a zin e-1-
su lfon ic Acid Dim eth yla m id e (6). Meth od 1. To an ice-
cold solution of 4-[2-(1-methoxy-ethyl)-pyrimidin-4-yl]-piper-
azine-1-sulfonic acid dimethylamide (1.75 g, 5.31 mmol) in
methylene chloride (53 mL) was added boron tribromide (10.6
mL, 10.6 mmol), and the reaction mixture was stirred for 1 h.
The mixture was allowed to warm to room temperature and
was quenched with a saturated aqueous solution of sodium
bicarbonate. The methylene chloride layer was washed with
water (20 mL), dried over sodium sulfate, and filtered, and
the filtrate was evaporated to a solid residue, which was
crystallized from a mixture of isopropyl ether and methylene
chloride to obtain the title compound as a white solid (642 mg,
38%): mp 103-105 oC; [R]D +16.1° (c ) 1, methanol); 1H NMR
(CDCl3, 300 MHz) 1.48 (d, 6H), 2.83 (s, 3H), 3.31 (m, 4H), 3.74
(m, 4H), 4.70 (q, 1H), 6.40 (d, 1H), 8.21 (d, 1H); MS (CI) M+1
316.
(R)-2-Meth oxy-p r op ion itr ile (22). A mixture of 2-meth-
oxy-propionamide, phosphorus pentoxide (4.13 g, 29.1 mmol),
and dry sand (3.0 g) was heated to 160 °C in a distillation
apparatus, and the distillate was collected (1.01 g, 44%): [R]D
+139.2 (c ) 1, methanol); 1H NMR (CDCl3, 300 MHz) δ 1.5 (d,
J ) 8 Hz, 3H), 3.45 (s, 3H), 4.15 (q, J ) 8 Hz, 1H).
(R)-2-Meth oxy-p r op ion im id ic Acid Eth yl Ester Hyd r o-
ch lor id e (23). Hydrogen chloride gas was passed into an ice-
cold solution of 2-methoxy-propionitrile (9.91 g, 116.5 mmol)
in anhydrous ethanol (100 mL) until the solution was satu-
rated with the gas. The reaction was stored in a refrigerator
overnight. The excess ethanol was removed under vacuum to
obtain the title compound as a deliquescent solid (100%, 19.5
g); [R]D +46.4 (c ) 1, methanol); 1H NMR (CDCl3, 300 MHz) δ
1.4 (d, J ) 8 Hz, 3H), 3.4 (s, 1H), 3.7 (q, J ) 8 Hz, 2H), 4.8 (q,
J ) 8 Hz, 1H), 6.2 (b, 1H).
Met h od 3. R-1-[4-(4-Dimethylsulfamoyl-piperazin-1-yl)-
pyrimidin-2-yl]-ethyl butyrate (230 mg, 0.60 mmol) was com-
bined with concentrated hydrochloric acid (5.0 mL) and stirred
at ambient temperature for 6 h and diluted with water, the
pH of the solution was adjusted to 9.0 with 6 N aqueous
sodium hydroxide, and it was extracted twice with ethyl
acetate. The extract was washed once with water, dried over
magnesium sulfate, and filtered, and the filtrate was concen-
trated to an oil which was purified by flash chromatography
(9:1 methylene chloride:methanol). Evaporation of the eluent
gave an oil which was crystallized from isopropyl ether to
obtain the title compound as a white solid identical to that
obtained by method 1.
P r ep a r a tion of 31. (R,S) 2-Hydroxyethyl 4-hydroxy pyrimi-
dine (2.1 g, 15.07 mol) was added to dioxane (63 mL) contain-
ing vinyl acetate (4.3 g, 50 mol), and the mixture was heated
to 50 °C. To the resulting solution was added lipase P30 (0.21
g), and the heating was continued for 24 h. The reaction
mixture was filtered, and the filtrate was evaporated to obtain
a thick syrupy liquid residue. The residue was chromato-
graphed over silica gel and eluted with a mixture of 95:5
methylene chloride and methanol. Evaporation of the collected
eluent gave the title compound as a colorless liquid (0.97 g,
92%): [R]D +39.9° (c ) 1, methanol); 1H NMR (CDCl3, 300
MHz) δ 1.61 (d, J ) 7 Hz, 3H), 2.2 (s, 3H), 5.65 (q, J ) 7 Hz,
1H), 6.35 (d, J ) 6 Hz, 1H), 7.97 (d, J ) 6 Hz, 1H), 11.94 (s,
1H).
(R)-2-Meth oxy-p r op ion a m id in e Hyd r och lor id e (24). A
solution of the 2-methoxy-propionimidic acid ethyl ester
hydrochloride in ethanol was cooled in an ice bath, gaseous
ammonia was passed into the solution until the reaction was
saturated with ammonia, and the reaction was stirred over-
night. Excess ethanol was removed to obtain a syrupy liquid,
which was triturated with diethyl ether (100 mL). The result-
ing solid was filtered to collect the title compound (12.2 g):
1
mp 60-75 °C; [R]D +45.4 (c ) 1, methanol); H NMR (CDCl3,
300 MHz) δ 1.4 (d, J ) 8 Hz, 3), 3.4 (s, 3H), 3.7 (q, J ) 8 Hz,
1H), 5.8 (b, 1H), 6.5 (b, 1H).
(R)-2-Meth oxym eth yl-p yr im id in e-3H-4-on e (26). A mix-
ture of (R)-2-methoxy-propionamidine hydrochloride (10.0 g,
72.15 mmol), sodium 2-ethoxycarbonyl ethenolate (25, 19.93
g, 144.3 mmol), and water (50 mL) was stirred at room
temperature for 24 h. The reaction was neutralized by addition
of sufficient concentrated HCl to adjust the pH to around 7.0
and extracted with methylene chloride. The extract was dried
over anhydrous sodium sulfate, evaporated to dryness, and the
residue was chromatographed over silica gel. Elution with a
mixture of 95:5 methylene chloride and methanol, and evapo-
ration of the eluent gave a thick viscous oil (1.7 g): [R]D +76.4
1
(c ) 1, methanol); H NMR (CDCl3 , 300 MHz) δ 1.51 (d, J )
8 Hz, 3H), 3.44 (s, 3H), 4.29 (q, J ) 8 Hz, 1H), 7.91 (d,1H),
11.00 (b, 1H).
(R)-2-(1-Me t h oxye t h yl)-p yr im id in -4-yl-m e t h a n e su l-
fon a te (27). To an ice-cold solution of (R)-2-methoxymethyl-
4-hydroxy-pyrimidine (1.69 g, 10.95 mmol) in methylene
chloride (10 mL) was first added triethylamine (1.7 mL, 12.05
mmol) and then mesyl chloride (1.38 g, 12.05 mmol). The
reaction was stirred at ice temperature for 20 min and later
allowed to come to room temperature. After 1 h, the reaction
was quenched with saturated aqueous sodium bicarbonate
solution, the organic layer was collected, and it was washed
with water (10 mL), dried over sodium sulfate, and filtered,
and the filtrate was evaporated to an oil (2.11 g, 83%): [R]D
+69.2 (c ) 1, methanol); 1H NMR (CDCl3, 300 MHz) δ 1.52 (d,
J ) 8 Hz, 3H), 3.38 (s, 3H), 3.62 (s,3H), 4.53 (q, J ) 8 Hz, 1H),
7.00 (d, 1H), 8.80 (d, 1H).
P r ep a r a tion of 33. (R,S) 2-Hydroxyethyl 4-hydroxy pyrimi-
dine (21.75 g, 155.2 mmol) was added to dioxane (650 mL)
containing vinyl butyrate (17.72 g, 310 mmol), and the mixture
was heated to 50 °C. To the resulting solution was added lipase
P30 (4.35 g), and the heating was continued for 24 h. The
reaction mixture was filtered, and the filtrate was evaporated
to obtain a thick syrupy liquid residue. The residue was
partitioned between methylene chloride (300 mL) and water
(600 mL), and the methylene chloride layer was collected, dried
over anhydrous sodium sulfate, and then filtered. The filtrate
was evaporated to obtain the title compound as a colorless
1
liquid (9.35 g, 86%): [R]D +29.5 (c ) 1, methanol); H NMR
(CDCl3, 300 MHz) δ 0.95 (t, J ) 8 Hz, 3H), 1.65 (m, 5H), 2.4
(m, 2H), 5.65 (q, J ) 8 Hz, 1H), 6.45 (d, J ) 8 Hz, 1H), 8.0 (d,
J ) 8 Hz, 1H).
(R)-4-[2-(1-Meth oxy-eth yl)-p yr im id in -4-yl]-p ip er a zin e-
1-su lfon ic Acid Dim eth yla m id e (28). To a solution of the
above sulfonate (2.11 g, 9.1 mmol) in tetrahydrofuran (20 mL)
was added N,N-dimethylsulfamoyl piperazine (14, 1.94 g, 10
mmol) followed by triethylamine (1.4 mL, 10 mmol). The
reaction was refluxed for 15 h and evaporated to an oily
residue. This was extracted with ethyl acetate (20 mL), and
the extract was washed first with a saturated aqueous solution
of sodium bicarbonate and then with water (10 mL). The ethyl
acetate extract was dried over sodium sulfate and filtered, and
the filtrate was evaporated to a crude product, which was
chromatographed over silica gel. Elution with a mixture of 9:1
ethyl acetate and methanol and evaporation of the solvents
gave the title compound (1.75 g, 59%): mp 65-70 °C; [R]D
R-1-[4-(4-Dim eth ylsu lfam oyl-piper azin -1-yl)-pyr im idin -
2-yl]-eth yl Bu tyr a te (36). To a solution of (R)-1-[4-piperazin-
1-yl)-pyrimidin-2-yl]-ethyl butyrate (1.49 g, 5.0 mmol), tri-
ethylamine (0.61 g, 6.0 mmol), and tetrahydrofuran (20.0 mL)
was added N,N-dimethylsulfamoyl chloride (0.86 g, 6.0 mmol)
at ambient temperature and stirred for 2 h. The mixture was
diluted with water and extracted twice with ethyl acetate. The
ethyl acetate extract was washed once with water (10 mL),
dried over magnesium sulfate, and filtered, and the filtrate
was concentrated to obtain the title compound as an oil (0.72
g, 87%): 1H NMR (CDCl3, 300 MHz) δ 0.95 (t, 3H), 1.55(d,

2700 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 17
Mylari et al.
(11) Ballinger, W. E.; Day, W. W.; Beebe, D. A.; Oates, P. J .;
Zembrowski, W. J .; Mylari. B. L. The Effect of Multiple Bolus
Dosing vs Chronic Water Dosing with CP-166,572, A Potent
Sorbitol Dehydrogenase Inhibitor, on Nerve Fructose Normaliza-
tion in Diabetic Rats. 7th North American ISSX Meeting,
California, October 20, 1996.
(12) Avery, M.; Ballinger, W. E.; Zembrowski, W. J .; Inskeep, P. B.;
Mylari, B. L.; Fouda, H. G. Microbore HPLC/MS of CP-166, 572
Metabolites in the Rat. American Society Mass Spectrometry
Conference, Atlanta, GA, May 21, 1995.
6H), 1.55 (d, 3H), 1.6-1.7 (m, 4H) 2.42 (t, J ) 8 Hz, 2H), 3.2
(m, 1H), 3.45 (m, 4H), 3.74 (m, 4H), 5.4 (q, J ) 8 Hz, 1H), 6.4
(d, J ) 8 Hz, 1H), 8.1 (d, J ) 8 Hz, 1H).
Su p p or tin g In for m a tion Ava ila ble: Supplementary in-
formation for the preparation of other new compounds de-
scribed in the text. This material is available free of charge
via the Internet at http://pubs.acs.org.
(13) Schmidt, R. E.; Dorsey, D. A.; Beaudet, L. N.; Plurad, S. B.;
Williamson, J . R.; Ido, Y. Effects of Sorbitol Dehydrogenase
Inhibition on Experimental Diabetic Autonomic Neuropathy. J .
Neuropathol. Exp. Neurol. 1998, 57, 1175-1189.
(14) US Patent 5,728,704; Chem. Abstr. 2000, 121, 35636.
(15) (a) Karlsson, C.; J ornvall, H.; Hoog, J .-O. Sorbitol Dehydro-
genase: cDNA Coding for the Rat Enzyme. Eur. J . Biochem.
1991, 198, 761-765. (b) Lee, F. K.; Cheung, M. C.; Chung, S.
The Human Sorbitol Dehydrogenase Gene: cDNA Cloning,
Sequence Determination, and Mapping by Fluorescence in situ
Hybridization. Genomics 1994, 21, 354-358.
(16) We thank Dr. J an-Olov Hoog of Karolinska Institutet for
providing us the rat and human sorbitol dehydrogenase cDNA
and Dr. A. Kamath and Mr. D. Cunningham in our Protein
Science Department for supplying us the rat and human
recombinant sorbitol dehydrogenases.
(17) J effery, J .; J ornvall, H. Sorbitol Dehydrogenase. Adv. Enzymol.
1988, 61, 47-106.
(18) Lindstad, R. I.; Koll, P.; McKinley-McKee, J . S. Substrate
Specificity of Sheep Liver Sorbitol Dehydrogenase. Biochem. J .
1998, 330, 479-487.
(19) After we completed our work on 11, a publication by Varlet and
co-workers disclosed both synthesis (same as ours) and SDH
inhibition activity of 11 relative to 2. For details, see Heterocycles
2000, 53, 797-804.
(20) (a) Hoog, J .-O.; Karlsson, C.; Eklund, H.; Shapiro, R.; J ornvall,
H. Enzymology and Molecular Biology of Carbonyl Metabolism
4; Weiner, H., Ed.; Plenum Press: New York, 1993; pp 439-
450. (b) Karlsson, C.; J ornvall, H.; Hoog, J .-O. Enzymology and
Molecular Biology of Carbonyl Metabolism 5; Weiner, H., Ed.;
Plenum Press: New York, 1995; pp 397-406.
Refer en ces
(1) The Diabetes Control and Complications Trial Research Group.
The Effect of Intensive Treatment of Diabetes on the Develop-
ment and Progression of Long-term Complications in Insulin-
dependent Diabetes Mellitus. N. Engl. J . Med. 1993, 329, 977-
986.
(2) U. K. Prospective Diabetes Study Group. Intensive Blood-glucose
Control with Sulfonylureas or Insulin Compared with Conven-
tional Treatment and Risk of Complications in Patients with
Type 2 Diabetes (UKPDS 33). Lancet 1998, 352, 837-853.
(3) Oates, P. J .; Mylari, B. L. Aldose Reductase Inhibitors: thera-
peutic Implications for Diabetic Complications. Exp. Opin. Invest.
Drugs 1999, 8, 2095-2119.
(4) Williamson, J . R.; Chang, K.; Frangos, M.; Hasan, K. S.; Ido,
Y.; Kawamura, T.; Nyengaard, J . R.; Enden Van Den.; Kilo, C.;
Tilton, R. G. Hypergylcemic Pseudohypoxia and Diabetic Com-
plications. Diabetes 1993, 42, 801-813.
(5) Obrosova, I. G.; Fathallah, L.; Lang, H. J .; Greene, D. A.
Evaluation of a Sorbitol Dehydrogenase Inhibitor on Diabetic
Peripheral Nerve Metabolism: a Prevention Study. Diabetologia
1999, 42, 1187-1194.
(6) Geisen, K.; Utz, R.; Grotsch, H.; Lang, J .; Nimmesgern. H.
Sorbitol-accumulating Pyrimidines: Arzneim. Forsch./ Drug Res.
1994, 44, 1032-1043.
(7) Tilton, R. G.; Chang, K.; Nyengaard, J . R.; Enden Van den.; Ido,
Y.; Williamson, J . R. Inhibition of Sorbitol dehydrogenase.
Effects on Vascular and Neural Dysfunction in Streptozotocin-
induced Diabetic rats. Diabetes 1995, 44, 234-242.
(8) Cameron, N. E.; Cotter, M. A.; Basso, M.; Hohman, T. C.
Comparison of the Effects of Inhibitors of Aldose Reductase and
Sorbitol Dehydrogenase on Neurovascular Function, Nerve
Conduction and Tissue Polyol Pathway Metabolites in Strepto-
zotocin-diabetic Rats. Diabetalogia 1997, 40, 271-281.
(9) Ido, Y.; Chang, K.; Oates, P. J .; Mylari, B. L.; Williamson, J . R.
Inhibitors of Sorbitol Dehdrogenase (SDI) and Aldose reductase
(ARI) Reverse Impaired Motor Nerve Conduction Velocity
(MNCV) in Diabetic Rats. Diabetes 1999, 48 (Suppl. 1), A150.
(10) Ng, T. F.; Lee, F. K.; Song, Z. T.; Calcutt, N. A.; Lee, A. Y.; Chung,
S. S.; Chung, S. K.; Ng, D. T.; Lee, L. W. Effects of Sorbitol
Dehydrogenase Deficiency on Nerve Conduction in Experimental
Diabetic Mice. Diabetes 1998, 47, 961-966.
(21) Levene, P. A.; Marker, R. E. Optical Rotations of Configuration-
ally Related Methoxypropionic and â- Methoxybutyric Nitriles.
J . Chem. Soc. 1933, 297-302.
(22) Satisfactory analytical and spectral data have been obtained for
all target compounds.
(23) Siegel, T. W.; Smith, S. R.; Ellery, C. A.; Williamson, J . R.; Oates,
P. J . An Enzymatic Fluorimetric Assay for Fructose. Anal.
Biochem. 2000, 280, 329-331.
J M0102001