Design and synthesis of a novel family of triazine-based inhibitors of sorbitol dehydrogenase with oral activity: 1-{4-[3R,5S-dimethyl-4-(4-methyl-[1,3,5]triazin-2-yl)-piperazin-1-yl]-[1,3,5] triazin-2-yl}-(R) ethanol

Article abstract of DOI:10.1016/S0968-0896(03)00490-5

Two new templates, (R) 2-hydroxyethyl-pyridine and (R) 2-hydroxyethyl-triazine, were used to design novel sorbitol dehydrogenase inhibitors (SDIs). The design concept included spawning of these templates to function as effective ligands to the catalytic z

Full text of DOI:10.1016/S0968-0896(03)00490-5

Bioorganic & Medicinal Chemistry 11 (2003) 4179–4188
Design and Synthesis of a Novel Family of Triazine-Based
Inhibitors of Sorbitol Dehydrogenase with Oral Activity:
1-{4-[3R,5S-Dimethyl-4-(4-methyl-[1,3,5]triazin-2-yl)-piperazin-1-
yl]-[1,3,5]triazin-2-yl}-(R) Ethanol
Banavara L. Mylari,* Gregory J. Withbroe, David A. Beebe, Nathaniel S. Brackett,
Edward L. Conn, James B. Coutcher, Peter J. Oates and William J. Zembrowski
Pfizer Global Research and Development, Groton Laboratories, Groton, CT 06340, USA
Received 3 April 2003; accepted 11 July 2003
Abstract—Two new templates, (R) 2-hydroxyethyl-pyridine and (R) 2-hydroxyethyl-triazine, were used to design novel sorbitol
dehydrogenase inhibitors (SDIs). The design concept included spawning of these templates to function as effective ligands to the
catalytic zinc within the enzyme through incorporation of optimally substituted piperazino-triazine side chains so as to accom-
modate the active site in the enzyme for efficient binding. This strategy resulted in orally active SDIs, which penetrate key tissues,
for example, sciatic nerve of chronically diabetic rats. The latter template led to the design of the title inhibitor, 33, which normal-
ized the elevated sciatic nerve fructose by 96% at an oral dose of 10 mg/kg.
# 2003 Elsevier Ltd. All rights reserved.
Introduction
The first reported in vivo active sorbitol dehydrogenase
inhibitor (SDI) is 4-(2-hydroxymethyl-pyrimidin-4-yl)-
piperazin-1-sulfonic acid dimethylamide, 1.3 It is a rela-
tively weak SDI, with a very short plasma half-life in
diabetic rats.8 The crucial role of 2-hydroxy-
methylpyrimidine of 1, in liganding to the catalytic zinc
atom in SDH and to manifest SDH inhibition activity is
supported by the observation that isomers of 1, 2 and
3,9 showed very little SDH inhibition activity and con-
firmed by the recently resolved single crystal X-ray
structure of the ternary complex of recombinant human
SDH with NADH and 1.10 The crystal structure shows
that 1 is positioned in the active site such that both the
oxygen of 2-hydroxymethyl and N-1 make contact with
the zinc in SDH while the piperazine ring with its sulfa-
moyl side chain extends out towards the surface of the
protein. The impact of replacing the hydroxymethyl
group by (R) 2-hydroxyethyl group, as in 4, in enhan-
cing enzyme inhibition potency has also been reported.9
Subsequent publications have disclosed the design and
synthesis of even more potent and longer acting inhibi-
tors, 5 and 6. Critical contributors to higher potency and
longer duration of action included the strategic place-
ment of small lipophilic methyl groups on the piperazine
linker and replacement of the metabolically vulnerable
Sorbitol dehydrogenase (SDH), the second enzyme in
the polyol pathway (Fig. 1), in concert with NAD+,
oxidizes sorbitol to fructose, and reduces NAD+ to
NADH. Excess flux through SDH, as would be pre-
vailing under diabetic conditions, creates an imbalance
in the cytoplasmic NAD+/NADH ratio. Williamson et
al.1 have coined the phrase ‘pseudohypoxia’ to describe
the altered cytoplasmic redox state. They also have
reported2 that inhibition of SDH can restore the altered
cytoplasmic red-ox state and can influence the early
functional changes observed in experimental diabetes,
including alterations in vascular albumin permeation,
tissue blood flow, and nerve conduction velocity. Also,
Geisen et al. have reported on the effects of inhibiting
SDH on renal hyperfiltration3 in diabetic rats. Because
these early functional changes may contribute to even-
tual pathology in affected diabetic tissues, this report
has spurred research targeted towards investigating the
role of SDH in the field of diabetic complications.4À7
*Corresponding author at: 6 Quinley Way, Waterford, CT 06385,
USA. Tel.: +1-860-443-3085; fax: +1-860-715-4483; e-mail: blmylari
@aol.com
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00490-5

4180
B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
Chemistry
The pyridine targets, 9, 11, and 13, were prepared
according to Scheme 1. Reaction of 4-chloro-2-hydroxy-
methylpyridine,13 7, with excess piperazine-1-sulfonic
acid dimethylamide,9 8, gave 9. Oxidation of 9 to 10
followed by reaction with methylmagnesium bromide,
gave the racemic alcohol, 11. Exposure of 11 to lipase
P309 in dioxane, separation of the resulting acetate, 12,
followed by base hydrolysis yielded 13. The assignment
of (R) configuration to the chiral center was not as rig-
orous as in the case of the corresponding pyrimidine
alcohol, 4.9 It was based on the assumption that the
lipase, as in the case of 4, would specifically acylate the
racemic alcohol substrate, 11, to give the (R) acetate, 12.
It also turned out that 13, like 4, showed positive optical
rotation (Scheme 1).
Figure 1. Polyol pathway.
dimethylsulfamoyl group by suitably substituted het-
erocycles.11 The most potent inhibitors to date feature
triazines substituted with methyl groups12 (cf. 6,
R1=R2=Me). Structures 1–6 are depicted in Chart 1.
Overall, our extensive SAR experience suggests that
highly potent SDIs feature an effective arrangement of
ligands to zinc (2-hydroxyethyl and pyrimidine N-1)
and a moderately polar linker with lipophilic methyl
groups (dimethylpiperazine) attached to planar nitrogen
heterocycles (e.g., pyrimidine, triazine) substituted with
lipophilic, small alkyl groups. The practice of seeking
multiple chemo-types, in projects directed towards
potential therapeutic target enzymes or receptors, is
fairly well established in medicinal chemistry (e.g.,
aldose reductase inhibitors, angiotensin converting
enzyme inhibitors and angiotensin II antagonists). We
desired SDIs of new chemo-types. At present, there is
no report of any in vivo active, non-pyrimidine SDI.
Herein, we report our investigations towards the goal
and disclose the first examples of new templates, (R)-2-
The triazine target, 20, was prepared by adapting an
intriguing literature procedure for the preparation of
certain substituted triazines.14a,b Compound 8 was
reacted with N,N-dimethylaminocarbamoyl chloride to
obtain 14. A mixture of 14 and phosphorous oxychlor-
ide was heated for 30 m to obtain the iminochloride, 15.
Exposure of methoxy-acetamidine, 16 to cyanogen bro-
mide resulted in N-cyano-acetamidine, 17. Thermal
cyclization of 15 with 17, gave the chlorotriazine, 18,
which was sequentially dechlorinated and demethylated
to yield the desired triazine, 20 (Scheme 2).
hydroxyethyl-s-triazine
(R)-2-hydroxyethyl-pyridine
and (R)-2-hydroxyethyl-s-triazine (R)-2-hydroxyethyl-s-
triazine, that we have used for the design of non-pyr-
imidine, orally active SDIs, effective in reducing accu-
mulation of fructose in the sciatic nerve of
streptozotocin diabetic rats.
Unlike in the pyridine example, we started with chiral
building blocks of established (R) configuration to syn-
thesize the desired triazine targets. (R) 2-Benzyloxy-
propionamide, 21,15 was reacted with chlorosulfonyl
isocyanate to obtain the urea, 22, which was once more
reacted with chlorosulfonyl isocyanate to obtain (R)-2-
benzyloxy-N-ureidocarbonyl-propionamide, 23. Base
promoted cyclization of 23 gave the triazine-dione, 24,
which was heated with phosphorous oxychloride to
obtain the dichloro-triazine, 25. Coupling of 25 with 8
gave 26, which upon catalytic reduction over Pd/C,
Biological Testing
Potential SDIs were screened for in vitro activity against
sheep liver and rat and human recombinant SDHs.
Some compounds prepared very early in the program
were tested only against sheep liver SDH. In general, we
found that there were only minor differences in the
responses of the inhibitors to different SDHs.9,11,12
Inhibitors with IC50s ꢀ250 nM were tested for oral in
vivo activity in two streptozotocin diabetic rat models,
acute and chronic, as measured by the ability of the
inhibitors to lower fructose that becomes elevated in
diabetic sciatic nerve (see Biological methods under
Experimental for detailed procedures).
Scheme 1. Reagents and conditions: (a) Et3N, reflux; (b) MnO2,
CH2Cl2, reflux; (c) MeMgBr, THF, reflux; (d) Lipase P30, dioxane,
45 ꢁC, 4d; (e) NaOH, MeOH, rt.
Scheme 2. Reagents and conditions: (a). ClCONMe2–Et3N–THF, rt;
(b) POCl3, 110 ꢁC; (c) CNBr, EtOH–Et3N; (d) MeCN, reflux; (e) Pd/C,
H2, EtOH; (f) BBr3, C HCl2.
Chart 1.
2

B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
4181
underwent both dechlorination and debenzylation to
give the desired triazine, 27. The enantiomeric purity of
27, as determined by chiral HPLC, was 90% (Scheme 3).
Results and Discussion
X-ray crystal structure of rat SDH shows that the triad,
Cys44, His69 and Glu155, forms the core of ligands to
the catalytic zinc in SDH.17 Based on SAR relative to
the different arrangements of pyrimidine nitrogen
atoms, as in 2 and 3, it was speculated that inhibitor 1
(pKa, 6.1) would bind to the catalytic zinc in SDH
through the N-1 of the pyrimidine and oxygen of the 2-
hydroxymethyl group.9 This speculation is now con-
firmed by Pauly et al.10 through X-ray of the ternary
complex of h-SDH with NADH and 1. We first pre-
pared hydroxymethylpyridine analogue, 9, and found it
to be a markedly weaker inhibitor (45% inhibition at
10 mM) than 1 (IC50=246 nM9). We expected the
hydroxymethylpyridine moiety to serve as well as the
hydroxymethylpyrimidine, to function as a zinc ligand.
However, the higher pKa of 9 (9.6) favors protonation,
rather than coordination to zinc, at physiological pH.
The racemate homologue, 11, showed only incremental
improvement in activity (IC50=5 mM), over 9. As
anticipated 13 showed a further increase in potency
(IC50=2.5 mM), however small it was. In the absence of
any change in pKa, the trend in SDI potency is in
agreement with SAR in the hydroxyethyl-pyrimidine
series. Because 13 was still ꢂ90Â less potent than 4 in
vitro, and because the higher pKa of the pyridine tem-
plate (99% protonated at blood pH) did not bode well
for penetration into diabetic complications tissues, for
example, peripheral nerve, we discontinued efforts
around the hydroxyethyl-pyridine template. SDH inhi-
bition data of these compounds is included in Table 1.
The triazino-triazines, 33 and 35, were prepared
according to Scheme 4. N,N-Dimethylcarbomoyl dime-
thylpiperazine, 28, prepared from 2,6-dimethylpiperazine,
was heated with (R)-N-cyano-methoxy-propionamidine,
29 to obtain piperazino-triazine, 30, (cf. preparation of
18). Reaction of 30 with 1,3 di-chloro-5-methyl (31a,
R=Me) and 1,3 di-chloro-5-phenyl (31b, R=Ph) tria-
zines16 gave, respectively, 32a and 32b. Dechlorination
of these compounds followed by their subsequent
demethylation by reaction with boron tribromide, gave
the targets, with positive optical rotation. As in the case
of 27, 33, and 35 had ee of 90%. Compound 33 was
obtained in higher enantiomeric purity (98% ee) by
preparative chiral HPLC. Sufficient sample of S enan-
tiomer (94% ee), 34, was obtained, by chiral separation.
Esterification of the hydroxy group in 33 with dimethy-
lamino acetyl chloride gave 36.
Scheme 5 outlines the synthesis of the cyclopropyl ana-
logue, 40. Demethylation of 30 with boron tribromide
gave 37, which was first coupled with 1,3-dichloro-5-
cyclopropyl-(1,3,5)-triazine, 38, to obtain 39, which was
then dechlorinated to yield the target.
In search of a template significantly less basic than pyr-
imidines (e.g., 4) and pyridines for designing essentially
non-basic SDIs (pKa ꢂ4), we decided to explore an s-
triazine with hydroxyalkyl side chain as a potential zinc
Scheme 3. Reagents and conditions: (a) ClSO2NCo, MeCN, rt;
(b) KOH, H2O, rt; (c) POCl3, PhNEt2, 70 ꢁC; (d) NaHCO3, DMF;
(e) Pd/C, HCOONH4, IPO, 90 ꢁC.
Scheme 5. Reagents and conditions: (a) BBr3, C HCl2; (b) NaHCO3,
2
DMF; (c) Pd/C, HCOONH4, HCl, IPO.
Table 1. In vitro SDH inhibition data of sulfamoyl-pyridines and
triazines
No.
X
Y
Z
IC50 (mM)
Sheepa
Rata
Humana
9
C
H
C
H
H>10
5
11
13
20
27
CH
CH
N
CH
CH
N
Me (RS)
Me (R)
H
2.5
5
Scheme 4. Reagents and conditions: (a) ClCONMe2,
C HCl2;
2
(b) POCl3, DMF, 110 ꢁC, (c) MeCN, reflux; (d) NaHCO3, DMF, rt;
N
N
Me (R)
0.20
0.22
(e) Pd/C, HCl, HCOONH4, IPO; (f) BBr3,
ClCOCH2NMe2, Et3N.
C 2HCl2; (g)
an=3, SE ꢀ20%.

4182
B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
ligand. Because s-triazine is non-basic and is strongly
electron withdrawing, it was expected that SDIs
according to the design strategy would have pKas close
to 4. This is supported by the stark contrast in pKas
between 2-aminopyrimidine (5.9), 2-aminotriazine (3.1).
As there is very little change in pKa from 2-aminopyr-
imidine to pyrimidine, 1 (6.1), it was expected that 20
would show a pKa not much different than that of ami-
notriazine. The measured pKa of 20 is 3.4. Because there
would be very little protonation of 20 at pH 7.1, we
hoped that 20 would compete more effectively for liga-
tion to zinc, rather than to proton, and thus would be at
least as potent an inhibitor as is 1. However, 20
(IC50=5 mM), like 9, was significantly less potent than
1. Because there is very little information in the litera-
ture on zinc coordinating ability of 2-hydroxymethyl-
triazine, we are left with the obvious speculation that
the potential liganding nitrogen atom of the s-triazine
backbone may not be an effective coordinator to the
catalytic zinc atom. However, as in the pyridine case,
the hydroxyethyl compound, 27, was more potent than
20, thus reiterating the importance of enantiomeric 1-
hydroxyethyl group. Because it was sufficiently potent
in vitro (IC50=200 nM), it was tested in vivo, in the
acute test, and was found to normalize the elevated
fructose level in the diabetic rat sciatic nerve, by 50%, at
an oral dose of 10 mg/kg. It was the first time that a
non-pyrimidine template had yielded an inhibitor,
which was active both in vitro and in vivo. However, 27
family of SDIs11 (cf. 6 with R1=R2=Me). Further-
more, in spite of the experience with 27, the discovery of
6 was the impetus to investigate whether 2-hydroxy-
methyl- and 2-hydroxyethyl-triazines with potency
enhancing piperazine-triazine side chains.
Substituents on the triazine not bearing the hydroxy-
methyl ligand included Me (33), c-Pr (40), and Ph (35).
All compounds showed SDH inhibition activity both in
vitro and in vivo (Table 2). SAR for in vitro activity in
this series is in agreement with the pyrimidino-triazine
series.12 The Me and c-Pr analogues are nearly equipo-
tent, while the Ph analogue is slightly less potent than
both Me and c-Pr analogues. The Me analogue, 33,
showed the best combination of in vitro and in vivo
activities. It was very potent orally giving near normal-
ization of sciatic nerve fructose in both acutely (at 5 mg/
kg) and chronically (at 10 mg/kg) diabetic rats. As
expected the S-enantiomer, 34, was less potent, both in
vitro and in vivo. The distinctly different fructose inhi-
bition responses with 33 and 35 are similar to those
obtained with 4 and its S-enantiomer,9 suggesting that
the chiral hydroxyethyl group is fairly stable to in vivo
oxido-reduction. The c-Pr analogue, 40, was only
slightly less potent than 33, while the more lipophilic Ph
analogue, 35, was significantly less potent than 33.
Unlike, SDIs with a pyrimidine moiety9,11 (e.g., 1), the
triazine-triazines were very poorly soluble in water, both
at physiological pH and at acidic pH, as low as 2. We
prepared 36, a potential prodrug of 33, with a basic
dimethylamino side-chain, which was freely soluble in
10% HCl and in 1 M citric acid. As expected, it showed
little SDI activity, in vitro. While it was hoped that the
vastly improved opportunity for solubility of 36 in gas-
tric fluids would manifest in greater vivo potency over
33, the prodrug was only as potent as the parent. In the
absence of pharmacokinetic data, it can be speculated
that administration of 36 to diabetic rats did not result
in more effective delivery of the parent drug to blood,
nerve tissue, or both.
was markedly less potent than
4 (IC50=27 nM,
ED50=1.6 mg/kg),9 both in vitro and in vivo.
According to Johansson et al.,17 the SDH substrate cleft
is significantly more polar than the alcohol dehy-
drogenase substrate cleft. However, sorbitol can stack
against the hydrophobic benzene ring of Phe118. Other
hydrophobic amino acid residues within the cleft
include Tyr200 and Tyr50. Our extensive SAR in the
pyrimidine-based SDIs is in agreement with this subtle
blend of polar and hydrophobic environment within the
SDH active site cleft (vide supra), and it recently, led to
the discovery of a highly potent pyrimidino-triazine
Conclusion
We examined the potential of two new templates, (R)-
hydroxyethyl-pyridine and (R)-hydroxyethyl-triazine
for building novel SDIs. Because there was no precedent
in the literature for either of these templates to serve as
strong ligands to zinc, we investigated the possibility of
spawning the new templates to construct novel SDIs,
using extant SAR developed in the design of potent
pyrimidine-based SDIs in our laboratory. Hydroxy-
ethyl-pyridine SDI, 13, was found to be a rather poor
inhibitor. It is surmised that this compound is so basic
that it is nearly fully protonated at physiological pH,
thus depleting its potential ability to effectively ligand to
the catalytic zinc in SDH. Progressing from 13, we syn-
thesized the first examples of non-basic non-pyrimidine
SDIs, 20 and 27. Inhibitor 27 showed good activity both
in vitro and in vivo. This provided the impetus to
append the earlier discovered potentiating substituted
triazine side chains as in 6. Several (R) hydroxyethyl-
triazino-triazines from this approach were found to be
Table 2. In vitro and in vivo SDH inhibition data of triazino-
triazines
No.
R
X
H
IC50 (nM)
% normalization of fructose
Rata Humana mg/kg
Acute
Chronic
96
33 Me
42
59
10
5
5
10
5
3
114
91
35
75
89
87
34b Me
35 Ph
36 Me COCH2NMe2
40 c-Pr
H
H
400
140
—
390
90
—
42
H
36
77c
an=3, SE for the assays, ꢀ15%.
b(S)-Enantiomer of 33.
cTested at 9 mg/kg.

B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
4183
quite potent inhibitors, both in vitro and in vivo. Inhi-
bitors, 33, 35 and 40 showed IC50s in the range of 40–
90 nM and all were orally active in reducing fructose
production in the sciatic nerve of streptozotocin diabetic
rats. The best inhibitor in this series was 33. It was quite
potent and normalized sciatic nerve fructose by 114 and
96% at an oral dose of 10 mg/kg, in the acute and
chronic tests, respectively. Inhibitor 40 was a close
second to 33, in vivo. Even the best triazine-triazine
inhibitor, 33, was still considerably less potent than
the best members of the hydroxyethyl-pyrimidine ser-
ies, for example, 6 (R1=R2=Me). It appears that
ligand strengths of the templates increase in the
order, hydroxyethyl-pyridine >hydroxyethyl-triazine
> >hydroxyethyl-pyrimidine.
4-[2-Hydroxymethyl)-pyridine-4-yl]-piperazine-1-sulfonic
acid dimethylamide (9). A mixture of 4-chloro-2-
hydroxymethyl-pyridine (21 mmol, 3.03 g), N,N-dime-
thylsulfamoyl-piperazine (37.1 mmol, 7.17 g) and Et3N
(42.2 mmol, 5.9 mL) was refluxed overnight. Excess
Et3N was removed and the residue was quenched with
water (25 mL) and the resulting solid was filtered and
the solid collected. This was crystallized from hot H2O,
(43%, 2.7 g): mp 150–151 ꢁC. H NMR (DMSO-d6) d
1
2.7 (s, 6H), 3.3 (m, 4H), 3.4 (m, 4H), 4.4 (s, 2H), 6.7 (dd,
J=5.7, 2.9 Hz, 1H), 6.9 (d, J=2.4 Hz, 1H), 8.1 (d,
J=5.7 Hz, 1H).
4-(2-Formyl-pyridin-4-yl)-piperazin-1-sulfonic acid dime-
thylamide (10). To a solution of 4-[2-hydroxymethyl)-
pyridine-4-yl]-piperazine-1-sulfonic acid dimethylamide,
8, (0.65 mmol, 195 mg) in CHCl3 (20 mL) was added
manganese-di-oxide (0.6 g) and refluxed for 1.5 h. The
reaction mixture was filtered and the filtrate was evap-
orated to obtain the crude product as an oil (68%,
Experimental
1
Melting points were determined on a Thomas-Hoover
capillary melting point apparatus, and are uncorrected.
1H NMR spectra were obtained on a Bruker AM-250
(Bruker Co., Billerica, MA, USA), a Bruker AM-300, a
Varian XL-300 (Varian Co., Palo Alto, CA, USA), or a
Varian Unity 400 at about 23 ꢁCat 250, 300, or
400 MHz for proton. The solvent was CDCl3, unless
otherwise indicated. Chemical shifts are reported in
parts per million (d) relative to residual chloroform
(7.26 ppm) or dimethylsulfoxide (2.49 ppm), as an inter-
nal reference. The peak shapes and descriptors for the
peak shapes are denoted as follows: s, singlet; d, doub-
let; t, triplet; q, quartet; m, multiplet; c, complex; br,
broad; app, apparent. Low-resolution mass spectra were
obtained under thermospray (TS) conditions on a
Fisons (now Micromass) Trio 1000 Mass Spectrometer
(Micromass Inc., Beverly, MA, USA), under chemical-
ionization (CI) conditions on a Hewlett Packard 5989A
Particle Beam Mass Spectrometer (Hewlett Packard
Co., Palo Alto, CA, USA), or under atmospheric pres-
sure chemical ionization (APCI) on a Fisons (now
Micromass) Platform II Spectrometer. Optical rotations
were obtained on a Perkin-Elmer 241 MCPolarimeter
(Perkin-Elmer, Norwalk, CT, USA) using a standard
path length of 1 dcm at about 23 ꢁCat the indicated
concentration in the indicated solvent.
133 mg). H NMR (DMSO-d6) d 2.7 (s, 6H), 3.3 (m,
4H), 3.4 (m, 4H), 6.7 (dd, J=5.7, 2.9 Hz, 1H), 6.9 (d,
J=2.4 Hz, 1H), 8.1 (d, J=5.7 Hz, 1H), 9.4 (s, 1H).
4-[R,S-2-Hydroxyethyl)-pyridin-4-yl]-piperazine-1-sulfonic
acid dimethylamide (11). To a solution of 4-(2-formyl-
pyridin-4-yl)-piperazin-1-sulfonic acid dimethylamide,
10, (4.4 mmol, 1.2 g), in dry THF, under a blanket of
dry nitrogen, was added methylmagnesium bromide
(3 M in Et2O, 2.64 mL) and the reaction was refluxed for
3 h. The reaction was quenched with NH4Cl and was
extracted with CH2Cl2 (20 mL). The CH2Cl2 extract was
washed with brine (10 mL) and the washed organic
portion was collected, dried, and the residue was puri-
fied by flash column chromatography (5% MeOH/
1
CH2Cl2 to yield the product (30%, 375 mg). H NMR
(DMSO-d6) d 1.5 (d, J=6.8 Hz, 3H), 2.7 (s, 6H), 3.3 (m,
4H), 3.4 (m, 4H), 4.6 (q, J=6.8 Hz, 1H), 6.7 (dd, J=5.7,
2.9 Hz 1H), 6.9 (d, J=2.4 Hz, 1H), 8.1 (d, J=5.7 Hz,
1H).
Acetic acid 1-[4-(4-dimethylsulfamoyl-piperazin-1-yl)-
pyridin-2-yl]-(R)-ethyl ester (12). A mixture of 4-[R,S-2-
hydroxyethyl)-pyridin-4-yl]-piperazine-1-sulfonic acid
dimethylamide, 11, (1.11 mmol, 350 mg), vinyl acetate
(111 mmol, 10.3 mL), dioxane (15 mL), and Lipase P30
(35 mg) was heated at 45 ꢁCfor 4 days. The reaction was
filtered and the filtrate was evaporated to dryness. The
residue was purified by flash column chromatography
(5% MeOH/CH2Cl2 to obtain the product (84%,
Liquid column chromatography was performed using
forced flow (flash chromatography) of the indicated
solvent on either Baker Flash column (40 mm, J. T.
Baker, Phillipsburg, NJ, USA) or Flash column60 (EM
Sciences, Gibbstown, NJ, USA) in glass columns or
using low nitrogen or air pressure in Flash 40TM or
Flash 12TM (Biotage, Charlottesville, VA, USA) car-
tridges. Analytical and preparative chiral HPLC
separations were done using chiralpak AD columns.
Unless otherwise indicated the optical purity of chiral
target analogues were at least 90% ee. Reactions
requiring the use of hydrogen gas at pressures greater
than 1 atmosphere were run using a Parr hydrogen
apparatus (Parr Instrument Co., Moline, IL, USA).
Unless otherwise specified, reagents were obtained from
commercial sources.
167 mg): [a]D +52.9ꢁ (c 1, MeOH). H NMR (DMSO-
1
d6) d 1.5 (d, J=6.8 Hz, 3H), 2.2 (s, 3H), 2.7 (s, 6H), 3.3
(m, 4H), 3.4 (m, 4H), 5.6 (q, J=6.8 Hz, 1H), 6.7 (dd,
J=5.7, 2.9 Hz 1H), 6.9 (d, J=2.4 Hz, 1H), 8.1 (d,
J=5.7, 1H).
4-[R-2-Hydroxyethyl)-pyridine-4-yl]-piperazine-1-sulfonic
acid dimethylamide (13). A solution of the acetate, 12
(0.39 mmol, 139 mg), in dioxane (3 mL), MeOH
(0.5 mL), and 1 N NaOH (0.25 mL) was stirred for 1 h at
room temperature. The reaction was evaporated to
dryness and the residue was slurried an a mixture of
EtOAc/H2O (1 mL/5 mL) and the resulting white solid

4184
B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
was collected (55%, 68 mg); mp 168–169 ꢁC ; a[]D
+24.0ꢁ (c 1, MeOH). 1H NMR (DMSO-d6) 1.5 (d,
J=6.8 Hz, 3H), 2.7 (s, 6H), 3.3 (m, 4H), 3.4 (m, 4H), 4.6
(q, J=6.8 Hz, 1H), 6.7 (dd, J=5.7, 2.9 Hz 1H), 6.9 (d,
J=2.4 Hz, 1H), 8.1 (d, J=5.7, 1H).
1H NMR d 2.8 (s, 6H), 3.3 (m, 4H), 3.5 (s, 3H), 3.9 (m,
4H), 4.4 (s, 2H), 8.6 (s, 1H).
4-(4-Hydroxymethyl-[1,3,5]triazin-2-yl)-piperazine-1-sul-
fonic acid dimethylamide (20). To an ice-cold solution
of 4-(4-chloro-6-methoxymethyl-[1,3,5]triazin-2-yl)-
piperazine-1-sulfonic acid dimethylamide, 19 (1.5 mmol,
474 mg) in CH2Cl2 (20 mL) was added dropwise a solu-
tion of boron tribromide (1 M in CH2Cl2, 3 mL). After
2 h, the reaction was quenched with H2O (5 mL) and
sufficient 10% KOH solution to raise the pH to 9. The
CH2Cl2 layer was collected, dried, filtered and the fil-
trate was evaporated to a solid residue. This was crys-
tallized from acetone to obtain the title compound
4-Dimethylsulfamoyl-piperazine-1-carboxylic acid dime-
thylamide (14). To a solution of N,N-dimethylsulfa-
moyl-piperazine, 8, (5.2 mmol, 1.0 g) in THF (10 mL)
and triethyl amine (0.75 mL) was added N,N-dimethyl-
aminocarbamoyl chloride (5.2 mmol, 0.5 mL) and the
reaction was stirred for 2 h at room temperature. The
precipitated triethyl amine hydrochloride was filtered
off and the filtrate was evaporated to obtain a white
solid, which was crystallized from a 1:1 mixture of
EtOAc and n-hexane to yield the title compound (88%).
1H NMR d 2.4 (s, 6H), 2.7 (s, 6H), 2.9 (m, 4H), 3.1 (m,
4H).
(68%, 275 mg): mp 157–159 ꢁC. H NMR d 2.8 (s, 6H),
1
3.3 (m, 4H), 3.9 (m, 4H), 4.4 (s, 2H), 5.2 (b. 1H), 8.6 (s,
1H).
2-Benzyloxy-N-ureidocarbonyl-(R)-propionamide
(23).
4-(Chloro-dimethylamino-methylene)-piperazine-1-sul-
fonic acid dimethylamide chloride (15). A mixture of 4-
dimethylsulfamoyl-piperazine-1-carboxylic acid dime-
thylamide, 14, (2 mmol, 530 mg) and phosphorus oxy-
chloride (2 mmol, 0.2 mL) was heated to 110 ꢁCfor
0.5 h. After cooling, the reaction solidified to yield the
title compound, which was immediately used in the next
step below.
To a suspension of 2-benzyloxy-propionamide, 21,
(26.6 mmol, 4.77 g) in CH3CN (100 mL) at room tem-
perature was added dropwise chlorosulfonyl isocyanate
(26.6 mmol, 4.1 mL) in CH3CN (20 mL). After 1 h the
reaction was concentrated, then carefully quenched with
H2O (20 mL) and allowed to stir at room temperature
for 1 h. The precipitated solid (22) was filtered, collected
and air-dried to obtain (2-benzyloxy-propionyl)-urea
(62%, 3.66 g). 1H NMR d 1.2 (d, J=6.5 Hz, 3H), 4.0 (q,
J=6.5 Hz, 1H), 4.4 (dd, J=11.7 Hz, 2H), 7.3 (m, 5H),
7.7 (s, 1H), 10.0 (S, 1H). This compound was re-sub-
jected to the above reaction conditions to convert (2-
benzyloxy-propionyl)-urea to 2-benzyloxy-N-ureido-
carbonyl-propionamide, using, (2-benzyloxy-propio-
nyl)-urea (16.5 mmol, 3.66 g), chlorosulfonyl isocyanate
(28.8 mmol, 2.5 mL), and acetonitrile (80 mL). The yield
of 2-benzyloxy-N-ureidocarbonyl-propionamide was
2-Methoxy-N-cyano-acetamidine. (17). To an ice-cold
solution of 2-methoxy-acetamidine hydrochloride, 16,
(0.1 mol, 12.5 g) in ethanol (100 mL) and Et3N (0.2 mol,
27.8 mL) was added dropwise a solution of cyanogen
bromide in acetonitrile. After 1 h, the solvents were
removed by evaporation and H2O (100 mL) was added
to the resulting residue. It was then extracted with
EtOAc (2 Â 100 mL). The EtOAc extract was collected,
dried, filtered and the filtrate was evaporated to obtain a
light yellow solid (72%, 8.2 g): mp 101–103 ꢁC. NMR d
3.4 (s, 3H), 3.7 (s, 2H), 5.8 (b, 1H), 6.5 (b, 1H).
1
59% (2.56 g). H NMR d 1.4 (d, J=6.5 Hz, 3H), 4.2 (q,
J=6.5 Hz, 1H), 4.6 (dd, J=12 Hz, 2H), 7.4 (m, 5H), 9.8
(s, 1H), 11.1 (s, 1H).
4-(4-Chloro-6-methoxymethyl-[1,3,5]triazin-2-yl)-pipera-
zine-1-sulfonic acid dimethylamide (18). 4-(chloro-
6-(1-Benzyloxy-(R)-ethyl)-1H-[1,3,5]triazine-2,4-dione (24).
To an ice-cold suspension of 2-benzyloxy-N-ureido-
carbonyl-propionamide, 23, (9.4 mmol, 2.5 g) in H2O
(15 mL) was added KOH (28 mmol, 1.6 g) in water
(10 mL). The reaction temperature was slowly raised to
room temperature, and the reaction was allowed to stir
for 1 h. Sufficient AcOH was added to adjust the pH of
the reaction to 5, and the resulting cloudy solution was
extracted with chloroform (3 Â 20 mL). The CHCl3
layer was collected, dried, filtered and the filtrate was
concentrated to obtain a residue, which was triturated
with IPE to obtain the title compound (74%, 1.72 g). 1H
NMR d 1.4 (d, J=6.5 Hz, 3H), 4.2 (q, J=6.5 Hz, 1H),
4.6 (dd, J=12 Hz, 2H), 7.4 (m, 5H), 11.2 (s, 1H), 12.1 (s,
1H).
dimethylamino-methylene)-piperazine-1-sulfonic
acid
dimethylamide chloride, 15, was dissolved in acetonitrile
(10 mL), to it was added 2-methoxy-N-cyano-acet-
amidine, 16 (2 mmol, 226 mg and refluxed for 2 h. Eva-
poration of the excess acetonitrile, gave a residue, which
was purified by flash column chromatography to yield
ꢁ
1
the title compound (58%, 410 mg): mp 143–144 C. H
NMR d 2.8 (s, 6H), 3.3 (m, 4H), 3.5 (s, 3H), 3.9 (b, 4H),
4.4 (s, 2H).
4-(4-Methoxymethyl-[1,3,5]triazin-2-yl)-piperazine-1-sul-
fonic acid dimethylamide (19). A mixture of 4-(4-chloro-
6-methoxymethyl-[1,3,5]triazin-2-yl)-piperazine-1-sul-
fonic acid dimethylamide, 18 (1.0 mmol, 350 mg), palla-
dium-carbon (100 mg), EtOH (10 mL), and NaOAc
(2.4 mmol, 196 mg) was hydrogenated in a Parr shaker
at 45 lb/sq inch for 1 h. The catalyst was filtered off and
the filtrate was concentrated. The resulting white pre-
cipitate was filtered and the residue was purified by flash
column chromatography (10% MeOH, CH2Cl2) to
obtain the title compound (70%, 224 mg): mp 78–81 ꢁC.
2-(1-Benzyloxy-(R)-ethyl)-4,6-dichloro-[1,3,5]triazine (25).
A mixture of 6-(1-benzyloxy-ethyl)-1H-[1,3,5]triazine-
2,4-dione, 24, (6.1 mmol, 1.5 g), phosphorus oxychloride
(18.2 mmol, 1.7 mL), and diethyl aniline (1 mL) was
heated at 70 ꢁCfor 1 h. Excess phosphorus oxychloride
was removed and the residual oil was extracted with
chloroform (2 Â 20 mL); the extract was washed with

B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
4185
1
water (3 Â 20 mL); the CHCl3 layer was collected, dried,
filtered and the filtrate was evaporated to obtain an oily
product. This oily product was purified by flash column
chromatography (90% hexane–EtOAc) to obtain the
compound (35%, 611 mg). 1H NMR d 1.6 (d, J=6.5 Hz,
3H), 4.0 (q, J=6.5 Hz, 1H), 4.6 (dd, J=12 Hz, 2H), 7.3
(m, 5H).
(85%, 410 mg). H NMR d 1.4 (d, J=6.5, 3H), 3.1 (q,
J=6.5 Hz, 1H), 3.4 (s, 3H).
2-Chloro-4-(3R,5S-dimethyl-piperazin-1-yl)-6-(R-1-meth-
oxy-ethyl)-[1,3,5]triazine (30). A mixture of 2,6-dime-
thyl-piperazine-1-carboxylic acid dimethylamide, 28
(51 mmol, 9.4 g) and phosphorus oxychloride (51 mmol,
4.8 mL) was heated at 110 ꢁCfor 30 min. After cooling
the reaction to room temperature, 2-methoxy-N-cyano-
propinamidine (51 mmol, 6.5 g) and acetonitrile was
added and then refluxed for 2 h. The reaction mixture
was evaporated to dryness and the residue was purified
by flash column chromatography (90% CH2Cl2/MeOH)
4-[4-(1-Benzyloxy-(R)-ethyl)-6-chloro-[1,3,5]triazin-2-yl]-
piperazine-1-sulfonic acid dimethylamide (26). A mixture
of 2-(1-benzyloxy-(R) ethyl)-4,6-dichloro-[1,3,5]triazine,
25, (0.75 mmol, 212 mg), N,N-dimethylsulfamoyl piper-
azine, 8, (0.75 mmol, 144 mg), NaHCO3 (1.5 mmol,
125 mg), and DMF (3 mL) was stirred overnight at
room temperature. EtOAc (15 mL) and H2O (20 mL)
were added and the EtOAc extract was collected and
washed with H2O (2 Â 10 mL). The EtOAc layer was
collected, dried and filtered, and the filtrate was evapo-
1
to obtain the compound (24%, 5.1 g). H NMR d 1.2
(dd, J=5.6 Hz, 6H), 1.4 (d, J=6.8 Hz, 3H), 2.8–3.1 (m,
4H), 3.4 (s, 3H), 4.1 (q, J=6.8 Hz, 1H), 4.7 (m, 1H), 4.8
(m, 1H). MS (AP+), 286.0.
1
rated to obtain the title compound (97%, 320 mg). H
2-Chloro-4-[2-(4-chloro-6-methyl-[1,3,5]triazine-2-yl)-
3R,5S-dimethyl-piperazin-1-yl]-6-(R-1-methoxy-ethyl)-
[1,3,5]triazine (32a). A mixture of 2-chloro-4-(3,5-dime-
thyl-piperazin-1-yl)-6-(1-methoxy-ethyl)-[1,3,5]triazine,
30, (1.12 mmol, 321 mg), 2,4-dichloro-6-methyl triazine,
31a (1.12 mmol, 184 mg), NaHCO3 (2.24 mmol, 189 mg)
and DMF (3 mL) was stirred at room temperature
overnight and was diluted with EtOAc (20 mL) and
water (30 mL). The EtOAc extract was collected, dried,
filtered and the filtrate was evaporated to a residue,
which was purified by flash column chromatography
(99% CH2Cl2/MeOH) to obtain the title compound
NMR 1.5 (d, J=6.6 Hz, 3H), 2.8 (s, 6H), 3.3 (m, 4H),
4.0 (m, 4H), 4.2 (q, J=6.6, 1H), 4.6 (dd, J=12 Hz, 2H),
7.3 (m, 5H); MS (AP+), 441.1
4-[4-(R)-1-Hydroxy-ethyl)-[1,3,5]triazin-2-yl]-piperazine-
1-sulfonic acid dimethylamide (27). A mixture of 4-[4-(1-
benzyloxy-R
ethyl)-6-chloro-[1,3,5]triazin-2-yl]-piper-
azine-1-sulfonic acid dimethylamide, 26, (0.73 mmol,
320 mg), (Pd/C910%, 640 mg), HCl (2 M in ether,
4.4 mmol, 2.2 mL), ammonium formate (15 mmol,
915 mg) and IPO (10 mL) was heated at 90 ꢁCfor 2 h.
The reaction was cooled and filtered, and to the filtrate
was added chloroform (20 mL) and aq saturated
NaHCO3 (20 mL). The CHCl3 layer was collected,
dried, filtered and the filtrate was evaporated to dryness.
The resulting residue was purified by flash column
chromatography (96% CHCl3/MeOH) to yield the title
1
(49%, 225 mg). H NMR d 1.2 (dd, J=5.6 Hz, 6H), 1.4
(d, J=6.8 Hz, 3H), 2.4 (s, 3H), 3.2 (m, 2H), 3.4 (s, 3H),
4.2 (q, J=6.8 Hz, 1H), 4.7 (m, 1H), 4.8 (m, 1H), 5.0 (m,
2H). MS (AP+), 412.9.
1-{4-[3R,5S-Dimethyl-4-(4-methyl-[1,3,5]triazin-2-yl)-pi-
perazin-1-yl]-[1,3,5]triazin-2-yl}-R-ethanol (33). A mix-
ture of 2-chloro-4-[2-(4-chloro-6-methyl-[1,3,5]triazine-
2-yl)-3R,5S-dimethyl-piperazin-1-yl]-6-(1-methoxy-ethyl)-
[1,3,5]triazine, 32a, (051 mmol, 211 mg), Pd/Ccatalyst
(10%, 84 mg), HCl (2 M in Et2O, 0.76 mmol, 0.38 mL),
ammonium formate (5.1 mmol, 322 mg) and IPO (8 mL)
was stirred at 90 ꢁCfor 2 h. After cooling the reaction it
was diluted with CH2Cl2 (20 mL) and was filtered. The
filtrate was evaporated to dryness and the residue was
partitioned between chloroform and aq saturated
sodium bicarbonate. The CHCl3 layer was collected,
dried, filtered and the filtrate was evaporated to a resi-
due, which was purified by flash column chromato-
graphy (99% CH2Cl2/MeOH) to obtain 2-[3R,5S-
dimethyl-2-(4-methyl-[1,3,5]triazine-4-yl)-piperazin-1-yl]-
4-(R 1-methoxy-ethyl)-[1,3,5]triazine (97%, 170 mg), [1H
NMR d 1.2 (dd, J=5.6 Hz, 6H), 1.4 (d, J=6.8 Hz, 3H),
2.4 (s, 3H), 3.2 (m, 2H), 3.4 (s, 3H), 4.2 (q, J=6.8 Hz,
1H), 4.7 (m, 1H), 4.8 (m, 1H), 5.0 (m, 2H), 8.5 (s, 1H),
8.6 (s, 1H). MS, AP+407.0] which was demethylated
according to the procedure for preparation of 5. This
material (ee, 90%) was further purified using pre-
compound (59%, 136 mg): mp 124–125 ꢁC. H NMR d
1
1.5 (d, J=6.6 Hz, 3H), 2.8 (s, 6H), 3.3 (m, 4H), 4.0 (m,
4H), 4.4 (s, 2H), 4.6 (q, J=6.6 Hz, 1H), 8.6 (s, 1H), 7.3
(m, 5H); MS (AP+), 317.1.
2,6-Dimethyl-piperazine-1-carboxylic acid dimethylamide
(28). To an ice-cold solution of 2,6-dimethyl piperazine
(179 mmol, 20.4 g), CH2Cl2 (200 mL), and (214 mmol,
29.9 mL) was added dropwise dimethylcarbamoyl
chloride (179 mmol, 16.4 mL). After 4 h the reaction was
quenched with a saturated NaHCO3 solution and the
CH2Cl2 layer was collected, dried, filtered and the fil-
trate was evaporated to dryness to obtain an orange oil
(70%, 23.1 g). NMR d 1.1 (d, J=4 Hz, 6H), 2.4 (m, 4H),
2.7 (s, 6H), 2.9 (m, 2H).
(R)-2-Methoxy-N-cyanopropinamidine (29). To an ice-
2-methoxy-propinamidine9
cold solution of
R
(3.8 mmol, 523 mg) in absolute EtOH (5 mL) and Et3N
(7.6 mmol, 1.1 mL) was added cyanogen bromide (2.9 M
in CH2Cl2, 1.3 mL). After the addition, the reaction
temperature was slowly raised to room temperature and
stirred for 3 h. Evaporation of all volatile liquids gave a
solid residue, which was extracted with EtOAc. The
EtOAc layer was washed with water and the EtOAc
layer was collected, dried, filtered and the filtrate was
evaporated to dryness to obtain the title compound
parative chiral HPLC[column, Chiralpak AD (5 cm
Â
50 cm; eluent, hexane–isopropanol, 85:15; flow rate,
75 mL/min; detect., 250 nM; tR, 60 min) to obtain the
title compound (76%; ee, 98%)]: mp 136–138 ꢁC ; d[]D
+14.4ꢁ (1.19 mg/mL, MeOH). H NMR d 1.2 (b, 6H),
1

4186
B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
1.5 (d, J=6.8 Hz, 3H), 2.4 (s, 3H), 3.2 (b, 2H), 4.6 (q,
J=6.8 Hz, 1H), 4.8 (m, 2H), 5.0 (m, 2H), 8.4 (s, 1H), 8.5
(s, 1H). MS (AP+)S, 331.0.
CH2Cl2, 11.04 mmol, 11 mL) and the reaction was stir-
red for 2 h. After the reaction was allowed to come to
room temperature, methylene chloride (10 mL) was
added to it followed by a small quantity of water (1 mL)
to quench the unreacted boron tribromide. Sufficient
saturated aq sodium bicarbonate was added to raise the
pH of the reaction solution to 8. The methylene chloride
layer was collected, dried, filtered and the filtrate was
evaporated to dryness to obtain a residue, which was
purified by flash column chromatography (98%
CH2Cl2/MeOH) to obtain the title compound (65%,
1-{4-[3R,5S-Dimethyl-4-(4-methyl-[1,3,5]triazin-2-yl)-pi-
perazin-1-yl]-[1,3,5]triazin-2-yl}-(S) ethanol (34). Pre-
parative chiral HPLCpurification of 1-{4-[3 R,5S-
dimethyl-4-(4-methyl-[1,3,5]triazin-2-yl)-piperazin-1-yl]-
[1,3,5]triazin-2-yl}-(R) ethanol of 90% ee (32) according
to the conditions described above gave the title com-
pound (20%; ee, tR, 62 min, 96%); mp 136–138 ꢁC ; a[ ]D
À13.8ꢁ (1.19 mg/mL, MeOH). H NMR d 1.2 (b, 6H),
392 mg): mp 126–128 ꢁC . a[]D +13.0ꢁ. H NMR d 1.1
1
1
1.5 (d, J=6.8 Hz, 3H), 2.4 (s, 3H), 3.2 (b, 2H), 4.6 (q,
J=6.8 Hz, 1H), 4.8 (m, 2H), 5.0 (m, 2H), 8.4 (s, 1H), 8.5
(s, 1H); MS (AP+), 331.0.
(m, 6H), 1.4 (d, J=6.8 Hz, 3H), 2.5 (m, 2H), 2.8 (m,
2H), 4.6 (m, 2H); MS (AP+), 272.0.
1-{4-Chloro-6-[4-(4-chloro-6-cyclopropyl-[1,3,5]triazin-2-
yl)-3R,5S-dimethyl-piperazin-1-yl]-[1,3,5]triazin-2-yl}-
ethanol (39). A mixture of 1-[4-chloro-6-(3R,5S-dime-
thyl-piperazin-1-yl)-[1,3,5]triazin-2-yl]-(R) ethanol, 37,
(0.74 mmol, 200 mg), 2,4-dichloro-6-cyclopropyl-tria-
zine, 38, (0.74 mmol, 140 mg), NaHCO3 (1.47 mmol,
124 mg) and DMF (4 mL) was stirred at room tempera-
ture overnight and was diluted with EtOAc (20 mL) and
water (20 mL). The EtOAc extract was collected, dried,
filtered and the filtrate was evaporated to a white solid
Dimethylamino-acetic acid 1-{4-[3R,5S-dimethyl-4-(4-
methyl-[1,3,5]triazin-2-yl)-piperazin-1-yl]-[1,3,5]triazin-2-
yl}-(R) ethyl ester (36). A mixture of 1-{4-[3R,5S-dime-
thyl-4-(4-methyl-[1,3,5]triazin-2-yl)-piperazin-1-yl]-[1,3,5]-
triazin-2-yl}-R-ethanol, 33, (0.45 mmol, 150 mg), N,N-
dimethylaminoacetyl chloride (2.7 mmol, 440 mg) and
Et3N (5.4 mmol, 0.75 mL) was refluxed overnight. The
reaction mixture was evaporated to a residue, which was
purified by flash column chromatography (94%
CH2Cl2/MeOH) to obtain the title compound as a vis-
1
(98%, 306 mg): H NMR d 1.0–1.2 (m, 4H), 1.3 (m,
1
cous oil (30%, 56 mg). H NMR d 1.2 (m, 6H), 1.6 (d,
6H), 1.5 (d, J=6.8 Hz, 3H), 2.0 (m, 1H), 3.2 (m, 2H),
4.6 (q, J=6.8 Hz, 1H), 4.8 (m, 2H), 5.0 (m, 2H); MS
(AP+), 424.9.
J=6.5 Hz, 3H), 2.4 (m, 9H), 3.2 (m, 2H), 3.3 (s, 2H), 4.7
(m, 2H), 5.0 (m, 2H), 5.6 (q, J=6.5 Hz, 1H), 8.5 (s, 1H),
8.5 (s, 1H).
1-{4-[4-(4-Cyclopropyl-[1,3,5]triazin-2-yl)-3R,5S-dime-
thyl-piperazin-1-yl]-[1,3,5]triazin-2-yl}-ethanol (40). A
2-Chloro-4-[2-(4-chloro-6-phenyl-[1,3,5]triazine-2-yl)-
3R,S5-dimethylpiperazin-1-yl]-6-(1-methoxy-R
ethyl)-
mixture of 1-{4-chloro-6-[4-(4-chloro-6-cyclopropyl-
[1,3,5]triazin-2-yl)-3,5-dimethyl-piperazin-1-yl]-[1,3,5]tri-
azin-2-yl}-ethanol, 39, (0.72 mmol, 306 mg), Pd/Ccata-
lyst (10%, 122 mg), HCl (2 M in Et2O, 1.08 mmol,
0.54 mL), ammonium formate (7.2 mmol, 454 mg) and
isopropanol (7 mL) was stirred at 90 ꢁCfor 2 h. After
cooling the reaction, it was diluted with methylene
chloride (20 mL) and was filtered. The filtrate was
evaporated to dryness and the residue was partitioned
between chloroform and aq saturated sodium bicarbo-
nate. The chloroform layer was collected, dried, filtered
and the filtrate was evaporated to a residue, which was
purified by flash column chromatography (99%
CH2Cl2/MeOH) to obtain a solid, which was triturated
with IPE to obtain the title compound (64%, 106 mg):
[1,3,5]triazine (32b). A mixture of 2-chloro-4-(3R,5S-
dimethyl-piperazin-1-yl)-6-(R-1-methoxy-ethyl)-[1,3,5]-
triazine, 30, (1.12 mmol, 321 mg), 6-phenyl-2,4-dichloro-
triazine, 31b (1.12 mmol, 184 mg), NaHCO3 (2.24 mmol,
189 mg) and DMF (4 mL) was stirred at room tempera-
ture overnight and was diluted with EtOAc (20 mL) and
water (20 mL). The EtOAc extract was collected, dried,
filtered and the filtrate was evaporated to a residue,
which was purified by flash chromatography (99%
CHCl3/MeOH), to give a white solid (48%, 225 mg). 1H
NMR d 1.3 (m, 6H), 1.5 (d, J=6.5 Hz, 3H), 3.2 (m, 2H),
3.4 (s, 3H), 4.6 (q, J=6.5 Hz, 1H), 4.8 (m, 2H), 5.1 (m,
2H), 7.2–7.4 (m, 3H), 8.3 (b, 2H); MS (AP+), 475.3.
1
1-{4-[3,5-Dimethyl-4-(4-phenyl-[1,3,5]triazin-2-yl)-pipera-
zin-1-yl]-[1,3,5]triazin-2-yl}-ethanol (35). 2-Chloro-4-[2-
(4-chloro-6-phenyl-[1,3,5]triazine-2-yl)-3R,S5-dimethyl-
mp 120–121 ꢁC. H NMR d 1.0–1.2 (m, 4H), 1.3 (m,
6H), 1.5 (d, J=6.8 Hz, 3H), 2.0 (m, 1H), 3.2 (m, 2H),
4.6 (q, J=6.8 Hz, 1H), 4.8 (m, 2H), 5.0 (m, 2H), 8.4 (s,
1H), 8.5 (s, 1H); MS, (AP+), 357.2.
piperazin-1-yl]-6-(1-methoxy-R
31b, was reacted with boron tribromide according to the
ethyl)-[1,3,5]triazine,
1
procedure for the preparation of 37 (54%). H NMR d
SDH inhibition activity
1.3 (m, 6H), 1.5 (d, J=6.5 Hz, 3H), 3.2 (m, 2H), 4.6 (q,
J=6.5 Hz, 1H), 4.8 (m, 2H), 5.1 (m, 2H), 7.2–7.4 (m,
3H), 8.3 (b, 2H), 8.5 (s, 1H), 8.6 (s, 1H); MS (AP+),
393.
Recombinant r-, h- or s-SDH was purchased from Pan
Vera Corp. (Madison, WI, USA) or Boehringer (Ger-
many). Experimental compounds were dissolved at
5 mM in 20% (v/v) aqueous DMSO, and 25-mL aliquots
were added to 0.2 mL of potassium phosphate buffer
(pH 7.0) containing NAD+, iodonitrotetrazolium violet
(INT), and h- or s-SDH. Following a 15-min incubation
at 25 ꢁC, the reaction was initiated by the addition of
20 mM sorbitol (25 mL). Final concentrations in the
1-[4-Chloro-6-(3R,5S-dimethyl-piperazin-1-yl)-[1,3,5]tria-
zin-2-yl]-(R) ethanol (37). To an ice-cold solution of 2-
chloro-4-(3R,5S-dimethyl-piperazin-1-yl)-6-(R-1-methoxy-
ethyl)-[1,3,5]triazine, 30, (2.2 mmol, 631 mg) in methyl-
ene chloride (7 mL) was added boron tribromide (1 M

B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
4187
assay were 90 mM potassium phosphate, 1 mM NAD+,
Appendix. Elemental analysis data
1 mM INT, 2 mM sorbitol, and 2 nM h-SDH or 8 nM s-
SDH. Enzyme activity was assayed with a SLT340
ATTCplate reader (model 16-925, SLT Lab-Instru-
ments, Austria), which measured the increase in the rate
of INT reduction at 495 nm at 25 ꢁCover 10 min. IC 50s
were calculated using a log-linear regression analysis.
Compd mol. formula
Calcd
Found
C
H
N
C
H
N
9
C12H20N4O3S 48.00 6.66 18.66 48.32 6.41 18.82
C13H22N4O3S 49.68 7.00 17.83 50.02 6.96 18.22
C13H22N4O3S 49.68 7.00 17.83 49.49 7.24 18.14
C10H18N6O3S 39.73 6.00 27.79 40.10 6.28 27.52
C11H20N6O3S 41.77 6.33 26.58 42.01 6.11 26.64
C15H22N8O 54.50 6.70 33.40 54.43 7.00 33.56
C15H22N8O 54.50 6.70 33.40 54.32 6.82 33.28
C20H24N8O 61.22 6.12 28.57 61.46 6.39 28.77
C24H31N8O2 62.20 6.69 24.19 61.98 6.86 24.32
C17H24N8O 57.30 6.74 31.46 57.58 6.55 31.72
11
13
20
27
33
34
35
36
40
Nerve fructose assay in diabetic rats: acute model
Male CD Sprague–Dawley rats (175–225 g) were made
diabetic by injection of STZ (17 mg/mL in 0.01 M citrate
buffer, pH 4.5, 85 mg/kg body wt) into the tail vein of
conscious rats. STZ was administered at approximately
9:00 a.m. on day 1. The animals were maintained with
free access to food and water. Test compound was
administered by oral gavage (volume of 5 mL/kg) at 4,
7, and 24 h after STZ administration. Four h after the
final dose (28 h after STZ administration), animals were
sacrificed by cervical dislocation. The left sciatic nerve
was excised, weighed, and placed in 1 mL of ice-cold 6%
perchloric acid and frozen for fructose analysis at a later
date. Weighed sciatic nerves in 6% (w/v) perchloric acid
(1 mL) were thawed and homogenized with a polytron
(Kinematica, Switzerland). After the mixture was cen-
trifuged, the decanted supernatant was neutralized by
the addition of 3.0 M potassium carbonate (100 mL) and
re-centrifuged. The fructose contents of the super-
natants were determined enzymatically. Briefly, fructose
was oxidized to 5-keto-fructose by fructose dehy-
drogenase (FDH) with concomitant reduction of resa-
zurin to the highly fluorescent resorufin. Final assay
concentrations were 0.2 M citric acid, pH 4.5, contain-
ing 13.2 M resazurin, 1.7 units/mL of FDH, and 0.068%
(v/v) Triton X-100. Reaction mixtures were incubated
for 60 min at room temperature in a closed drawer. The
sample fluorescence was determined at excitation )
(560 nm) and emission (580 nm), and slits of 5 mm each
(Perkin-Elmer model LS50B fluorescence spectro-
photometer). After the appropriate blanks from each
sample were subtracted, nanomole of fructose in each
sample was then determined by comparison with a lin-
ear regression of the fructose standards.
References and Notes
1. Williamson, J. R.; Chang, K.; Frangos, M.; Hasan, K. S.;
Ido, Y.; Kawamura, T.; Nyengaard, J. R.; Van Den Enden,
M.; Kilo, C.; Tilton, R. G. Diabetes 1993, 42, 801.
2. Tilton, R. G.; Chang, K.; Nyengaard, J. R.; Van den
Enden, M.; Ido, Y.; Williamson, J. R. Diabetes 1995, 44, 234.
3. (a) Geisen, K.; Utz, R.; Grotsch, H.; Lang, J.; Nimmes-
gern, H. Arzneim. Forsch./Drug Res. 1994, 44, 1032. (b) Also
known as SDI 158, CP-166572 and WAY135706.
4. Cameron, N. E.; Cotter, M. A.; Basso, M.; Hohman, T. C.
Diabetalogia 1997, 40, 271.
5. Ido, Y.; Chang, K.; Oates, P. J.; Mylari, B. L.; Williamson,
J. R. Diabetes 1999, 48, A150.
6. 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.
Diabetes 1998, 47, 961.
7. Oates, P. J. In The Neurobiology of Diabetic Neuropathy;
Tomlinson, D. R., Ed.; International Review of Neurobiol-
ogy, Vol 50; Academic: London, 2002; p 325.
8. 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 Nor-
malization in Diabetic Rats. Presented at the 7th North
American ISSX Meeting, San Diego, CA, Oct. 20, 1996.
9. Mylari, B. L.; Oates, P. J.; Beebe, D. A.; Brackett, N. S.;
Coutcher, J. B.; Michael, S.; Dina, M. S.; Zembrowski, W. J.
J. Med. Chem. 2001, 44, 2695.
10. Pauly, T. A.; Ekstrom, J. L.; Beebe, D. A.; Chrunyk, B.;
Cunningham, D.; Griffor, M.; Kamath, A.; Lee, M.; Madura,
R.; McGuire, D; Subashi, T.; Wasilko, D.; Watts, P.; Mylari,
B. L.; Oates, P. J.; Adams, P. D.; Rath. V. L. Structure 2003,
in press.
Nerve fructose assay in diabetic rats: chronic model
Male CD Sprague–Dawley rats (175–225 g) were made
diabetic by the injection of STZ (17 mg/mL in 0.01 M
citrate buffer, pH 4.5, 85 mg/kg body wt) into the tail
vein of conscious rats. STZ was administered at
approximately 9:00 a.m. on day 1. The animals were
maintained with free access to food and water. On day
8, at 9:00 a.m., test compound was administered by oral
gavage (volume of 5 mL/kg). Dosing continued once
daily at 9:00 a.m. for 5 days. Four hours after the final
dose (day 12), animals were sacrificed by cervical dis-
location. The left sciatic nerve was excised, weighed,
and placed in ice-cold 6% (w/v) perchloric acid (1 mL)
and frozen for fructose analysis at a later date, using the
methods described above.
11. Chu-Moyer, M. Y.; Ballinger, W. E.; Beebe, D. A.; Ber-
ger, R.; Coutcher, J. B.; Day, W. W.; Li, J.; Mylari, B. L.;
Oates, P. J.; Weekly, R. M. J. Med. Chem. 2002, 45, 511.
12. Mylari, B. L.; Oates, P. J.; Zembrowski, W. J.; Beebe,
D. A.; Conn, E. L.; Coutcher, J. B.; O’Gorman, M. T.; Lin-
hares, M. C.; Withbroe, G. J. J. Med. Chem. 2002, 45, 4398.
13. Hackley, B. E.; Daniher, F. A. J. Org. Chem. 1967, 32,
2624.
For elemental analysis data, see the Appendix.

4188
B. L. Mylari et al. / Bioorg. Med. Chem. 11 (2003) 4179–4188
14. (a) Harris, R. L. N. Synthesis 1980, 7, 841. (b) Harris,
R. L. N. Aust. J. Chem. 1981, 34, 623.
16. Wakabayashi, K.; Tsunoda, M.; Suzuki, Y. Bull.Chem.
Soc. Jap. 1969, 42, 2924.
15. Steiner, K.; Graf, U.; Hardegger, E. Helv. Chim. Acta
1971, 54, 845.
17. Johansson, H.; El-Ahmad, M.; Kaiser, C.; Jornvall, H.;
Hoog, J.; Ramaswamy, S. Chem. Biol. Interact. 2001, 30, 351.