Synthesis and biological activity of aminoguanidine and diaminoguanidine analogues of the antidiabetic/antiobesity agent 3-guanidinopropionic acid

Article abstract of DOI:10.1021/jm000094n

3-Guanidinopropionic acid (1) has been demonstrated both to improve insulin sensitivity and to promote weight loss selectively from adipose tissue in animal models of non-insulin-dependent diabetes mellitus (NIDDM). However, 1 has also been shown to be a substrate for both the creatine transporter and creatine kinase, leading to marked accumulation in muscle tissue as the corresponding N-phosphate. The corresponding aminoguanidine analogue 2 was recently discovered to retain the antidiabetic activity of 1 while being markedly less susceptible to creatine-like metabolism, suggesting that it should have less potential to accumulate in muscle. Further structural modification of 2 was undertaken to investigate whether the antidiabetic potency could be augmented while maintaining resistance to creatine-like metabolism. Modifications such as α-alkylation, homologation, and bioisosteric replacement of the aminoguanidine all were detrimental to antidiabetic activity. However, the simple regioisomeric aminoguanidinoacetic acid 9 and diaminoguanidinoacetic acid analogue 7 were found to be equipotent to 2, leading eventually to the discovery of the significantly more potent diaminoguanidinoacetic acid regioisomers 52 and 53. Further attempts to modify the more active template represented by 52 led only to reductions in antidiabetic activity. Each of the new active analogues displayed the same resistance to creatine-like metabolism as 2. Further testing of 7, 9, and 53 in obese diabetic ob / ob mice confirmed that weight loss is induced selectively from adipose tissue, similar to the lead 1. Administration of 53 to insulin-resistant rhesus monkeys led to reductions in both fasting and post-prandial plasma glucose levels with concomitant reductions in plasma insulin levels, suggesting that the compound improved the action of endogenous insulin. Compounds 7 and 53 were selected for further preclinical development.

Full text of DOI:10.1021/jm000094n

J . Med. Chem. 2001, 44, 1231-1248
1231
Syn th esis a n d Biologica l Activity of Am in ogu a n id in e a n d Dia m in ogu a n id in e
An a logu es of th e An tid ia betic/An tiobesity Agen t 3-Gu a n id in op r op ion ic Acid
Valerie A. Vaillancourt,*,† Scott D. Larsen,*,† Steven P. Tanis,† J effery E. Burr,§ Mark A. Connell,§
Michele M. Cudahy,† Bruce R. Evans,† Peter V. Fisher,† Paul D. May,† Martin D. Meglasson,§
Deborah D. Robinson,§ F. Craig Stevens,† J ohn A. Tucker,† Thomas J . Vidmar,‡ and J en H. Yu§
Departments of Medicinal Chemistry, Pharmacology, and Research Biostatistics, Pharmacia Corporation, 301 Henrietta Street,
Kalamazoo, Michigan 49007
Received March 1, 2000
3-Guanidinopropionic acid (1) has been demonstrated both to improve insulin sensitivity and
to promote weight loss selectively from adipose tissue in animal models of non-insulin-dependent
diabetes mellitus (NIDDM). However, 1 has also been shown to be a substrate for both the
creatine transporter and creatine kinase, leading to marked accumulation in muscle tissue as
the corresponding N-phosphate. The corresponding aminoguanidine analogue 2 was recently
discovered to retain the antidiabetic activity of 1 while being markedly less susceptible to
creatine-like metabolism, suggesting that it should have less potential to accumulate in muscle.
Further structural modification of 2 was undertaken to investigate whether the antidiabetic
potency could be augmented while maintaining resistance to creatine-like metabolism.
Modifications such as R-alkylation, homologation, and bioisosteric replacement of the ami-
noguanidine all were detrimental to antidiabetic activity. However, the simple regioisomeric
aminoguanidinoacetic acid 9 and diaminoguanidinoacetic acid analogue 7 were found to be
equipotent to 2, leading eventually to the discovery of the significantly more potent diamino-
guanidinoacetic acid regioisomers 52 and 53. Further attempts to modify the more active
template represented by 52 led only to reductions in antidiabetic activity. Each of the new
active analogues displayed the same resistance to creatine-like metabolism as 2. Further testing
of 7, 9, and 53 in obese diabetic ob/ ob mice confirmed that weight loss is induced selectively
from adipose tissue, similar to the lead 1. Administration of 53 to insulin-resistant rhesus
monkeys led to reductions in both fasting and post-prandial plasma glucose levels with
concomitant reductions in plasma insulin levels, suggesting that the compound improved the
action of endogenous insulin. Compounds 7 and 53 were selected for further preclinical
development.
In tr od u ction
ability to be incorporated into the creatine transport/
kinase cycle. Although 1 proved to be highly intolerant
of structural modification, the simple aminoguanidi-
noacetic acid analogue 2 was eventually discovered to
be an equipotent antidiabetic agent that was signifi-
cantly less susceptible to creatine-like metabolism,
suggesting that it should be less prone to accumulate
in muscle. A second phase of analogue work was thus
launched with the goal of discovering more potent
analogues of 2 which maintained this resistance to
creatine-like metabolism.
3-Guanidinopropionic acid (1) has been demonstrated
both to improve insulin sensitivity and to promote
weight loss selectively from adipose tissue in animal
models of non-insulin-dependent diabetes mellitus (NID-
DM).1,2 However, 1 has also been shown to be a
substrate for both the creatine transporter and creatine
kinase, leading to marked accumulation in muscle tissue
as the corresponding N-phosphate with concomitant
depletion of intracellular creatine phosphate.3 In the
Ch em istr y
Tables 1-8 depict the analogues that were prepared
and evaluated for this study. Compounds 8,4 9,5 and 186
were prepared according to literature procedures. Com-
pound 13 is an item of commerce.
Table 1 shows a variety of substituted analogues of
2. Esterification of 2 with methanol and HCl provides
3. R-Substituted analogues of 2 could be prepared from
the R-bromo acids 83 (eq 1). Treatment with hydrazine
produced the hydrazino acids7 which could be guany-
lated by reaction with 2-methyl-2-thiopseudourea sul-
fate to give the R-substituted aminoguanidinoacetic
acids 4-6. Preparation of the amino-substituted ana-
logue 7 could be accomplished by reaction of methyl
preceding paper, we described an analogue program
aimed at identifying novel, more potent compounds
retaining the antidiabetic activity of 1 but lacking its
* To whom correspondence should be addressed. V. A. Vaillan-
court: phone, (616)833-9552; fax, (616)833-2232; e-mail, valerie.a.
vaillancourt@am.pnu.com. S. D. Larsen: phone, (616)833-7713; fax,
(616)833-2516; e-mail, scott.d.larsen@am.pnu.com.
† Department of Medicinal Chemistry.
‡ Department of Research Biostatistics.
§ Department of Pharmacology.
10.1021/jm000094n CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/10/2001

1232 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
hydrazonothiocarbamate hydroiodide8 with hydrazino-
acetic acid (prepared in situ by saponification of ethyl-
Sch em e 2
hydrazinoacetate hydrochloride (84, X ) NH, R ) OEt)
(eq 2). Two regioisomeric analogues of 2 were also
(25-27) could be prepared by condensation of the
appropriately substituted aminoguanidine12,13 with gly-
oxylic acid (R3 ) H) (eq 3). Condensation of 2-hydrazino-
prepared. Alkylation of aminoguanidine with 2-bro-
moacetic acid gave 10. Saturation of the known triazi-
none 316 by catalytic hydrogenation in aqueous HCl
gave the cyclic analogue 11. A heteroaromatic replace-
ment of the aminoguanidine moiety was also investi-
gated. The heterocycle was prepared by first protecting
the terminal amino group of aminoguanidine as the
corresponding isopropylidene imine, followed by reaction
with 2-chloroacetaldehyde to give 86 (Scheme 1). The
heterocycle was then simultaneously deprotected and
alkylated with 2-chloroacetic acid to provide the desired
analogue 12.
4,5-dihydro-1H-imidazole with chloral hydrate provided
29. Condensation of diaminoguanidine with glyoxylic
acid in water at room temperature provided the dimer
30.
Biguanidine and oxamidine analogues related to 1
and 2 were prepared and are shown in Table 4. The
biguanide transfer reagent 90 (eq 4) was prepared by
Sch em e 1
alkylation of iminothiobiuret with methyl iodide.14 This
compound was then condensed with glycine, â-alanine,
3-aminopropionic acid, and taurine to give the desired
biguanide analogues 33-36. Esterification of 34 pro-
vided 37. The synthesis of the oxamidine analogues 38-
39 is shown in Scheme 3. The starting material 91, a
thiadiazole 1-oxide, was prepared according to a litera-
ture procedure.15 Addition of the requisite amino acid
(or ethyl ester) to 91 followed by direct ammonolysis of
Analogues of the regioisomeric aminoguanidine 9 are
shown in Table 2. Chain-extended analogues 14 and 15
were prepared by the procedure shown in eq 2 from the
amino acids 84 (X ) CH2, R ) H). Other analogues that
varied the terminal amino-substituent on the guanidine
of 9 could be prepared as shown in Scheme 2. The
thiourea 879 was alkylated with methyl iodide. Neu-
tralization of the resulting salt with aqueous sodium
hydroxide gave the desired isothiourea 88. Reactions
with 1,2-diaminoethane or hydroxylamine afforded the
aminoethyl (16) and hydroxy (17) analogues, respec-
tively.
Sch em e 3
Table 3 depicts a variety of imine analogues that were
prepared. Esterification of 18 with methanol provided
19. R-Substituted analogues 20-29 were also prepared
according to literature procedures (eq 3).10,11 Analogues
substituted on one of the nitrogens of the guanidine

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1233
the intermediate afforded 92. Hydrolysis of 92 with
concentrated aqueous HCl afforded the desired amidines
38-39. Esterification of 39 provided 40.
the diaminoisothiourea 97 (prepared by methylation of
thiocarbohydrazide, eq 6). The synthesis of the other
Analogues in which the guanidine is replaced by an
aminomethylamidine are shown in Table 5. Both pos-
sible isomers of these analogues were investigated.
Analogues with a terminal amino group were prepared
as shown in eq 5. The thioimidate 93 was prepared
regioisomeric diamoguanidinoacetic acid 53 is depicted
in Scheme 5. Starting from CBz-protected ethylhydrazi-
noacetate 98,21 saponification of the ester followed by
Sch em e 5
according to a literature procedure from glycinamide.16
Treatment of 93 with glycine or â-alanine followed by
deprotection with concentrated HCl gave the desired
analogues 41-42 as bis-hydrochloride salts. The syn-
thesis of the isomer in which the internal guanidine
nitrogen of 2 was replaced by an aminomethyl group is
summarized in Scheme 4. Glycine or â-alanine ethyl
treatment with trimethylsilyl isothiocyanate afforded
the thiourea 99. Alkylation with methyl iodide followed
by displacement of methanethiol with hydrazine and
deprotection of the benzyloxycarbonyl group afforded 53.
Analogues of the regioisomeric diaminoguanidinoace-
tic acid 52 are shown in Tables 7-8. R-Substituted
analogues 54-58 were synthesized from the corre-
sponding R-substituted amino acids 100 (Scheme 6).
Sch em e 4
Sch em e 6
Treatment with carbon disulfide followed by alkylation
with methyl iodide gave the dithiocarbamates 101.
Further alkylation with methyl triflate and neutraliza-
tion of the resulting salt led to the bis(methylthio)imines
102. Displacement of both methanethiol groups with
excess hydrazine gave the desired R-substituted ana-
logues 54-58. Analogues with variable chain lengths
and substituted analogues could be prepared as shown
in Scheme 7. Methylation of the known dithiocarbam-
ates 10322 with methyl iodide in acetone followed by
esters were formylated in the presence of sodium
cyanide to prepare the corresponding nitriles17 that were
Boc protected and treated with hydroxylamine to give
the hydroxyamidines 95. Saponification of the esters
provided the acids 96. At this point, deprotection with
HCl in dioxane furnished the R-aminohydroxyamidines
43-44. Alternatively, hydrogenolysis of the N-O bond18
and deprotection with HCl in dioxane gave the ami-
nomethylamidine analogues 45-46.
Additional diaminoguanidine analogues such as sub-
stituted analogues, a cyclic analogue, an analogue with
a heterocyclic replacement of the diaminoguanidine, and
regioisomeric analogues are depicted in Table 6. Con-
densation of the diaminoguanidine 7 with acetone or
benzaldehyde provided 47 and 48, respectively. Reduc-
tion of 47 with sodium cyanoborohydride provided 49.19
A cyclic analogue of 7 was prepared by hydrogenation
of the previously described cyclic imine 32.20 Alkylation
of 4H-1,3,4-triazole-3,4-diamine with 2-chloroacetic acid
afforded 51. The regioisomeric diamoguanidinoacetic
acid 52 was prepared by the reaction of glycine with
Sch em e 7

1234 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
Sch em e 10
neutralization of the resulting salts gave the bis-
(methylthio)imines 104. Exposure of 104 to excess
hydrazine gave the chain-extended analogues 59-60.
Alternatively, sequential displacement of the meth-
anethiol groups of the bis(methylthio)imine 104 (n )
1)23 with an amine and hydrazine provided the mono-
substituted analogues 61-64. Treatment of the bis-
(methylthio)imine 104 (n ) 1) with an excess of the
appropriate amine provided the disubstituted analogues
65-72. Deprotection of the bis-Boc-protected amino
analogue 67 with HCl in THF gave the bis(aminoethyl)
analogue 73. For heterocycle-substituted analogues 74-
75, addition of the nucleophile to 104 was too sluggish
to be useful. To facilitate these reactions, one of the
methanethiol groups of 105 was converted to a chloride
by treatment with sulfuryl chloride to give 106 (Scheme
8).24 The chloride could then be selectively displaced by
Sch em e 11
Sch em e 8
Sch em e 12
the less reactive heterocyclic amine. Saponification of
the ester followed by treatment with excess hydrazine
gave 74-75. The third approach to the synthesis of
substituted analogues is shown for the synthesis of 76
(Scheme 9). 2-Hydrazinopyridine was added to ethyl
Sch em e 9
Through a two-step condensation, tetrazino analogues
80 and 81 could also be prepared from the bis(thio-
methyl)imine 104a (Scheme 11). Reaction of 104a with
isothiocyanatoacetate. Saponification of the ester re-
sulted in the thiosemicarbazide 108. Alkylation with
methyl iodide followed by treatment with excess hydra-
zine gave 76. Condensation of 52 with excess 2,4-
pentanedione25 resulted in the formation of the bis-
(dimethylpyrazolyl) analogue 77. The final analogue
depicted in Table 8 is one which retains only the
terminal amino groups, replacing the remainder of the
diaminoguanidine moiety with carbon atoms. The syn-
thesis of this compound is outlined in Scheme 10. Raney
nickel reduction of the bis-nitrile 10926 afforded the
lactam 110 which was opened by heating in aqueous
HCl to afford the desired diamine 78.
N-acetylhydrazine or methyl hydrazinocarboxylate gave
the isothiosemicarbazide 111 (Scheme 11). Further
reaction with hydrazine gave the cyclic analogues 80
and 81. Finally, an additional cyclic analogue was
prepared as shown in Scheme 12. Ethyl isothiocyana-
toacetate was treated with methylhydrazine to give the
corresponding thiosemicarbazide.28 Saponification of the
ester afforded 113. Treatment with methyl iodide fol-
lowed by treatment with excess hydrazine gave the
cyclic analogue 82.
Conformationally restricted, cyclic diaminoguanidine
analogues are shown in Table 9. Compound 79 was
prepared as shown (eq 7) by the reaction of the bis-
(methylthio)imine 104a with ethylene bis(hydrazine).27

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1235
Ta ble 1. Substituted Aminoguanidine Analogues
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Ta ble 2. Isomeric Aminoguanidine Analogues
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Biology
An tih yp er glycem ic Activity in KKAy Mice. New
of 2 were detrimental to activity. The methyl ester 3
was inactive as were the R-substituted analogues 4-6
and the one-carbon homologue 8. Cyclization to the
corresponding lactam 11 also led to a loss of activity,
and incorporating the aminoguanidine into a cyclic
structure (12) was unfavorable as well. In this series
of substituted analogues, the only compounds that were
found to retain antidiabetic activity were the isomeric
aminoguanidine 9 and the diaminoguanidine 7. It was
intriguing to note that each of these analogues possessed
a terminal amino group. It was around these compounds
that we focused further SAR studies.
analogues were evaluated in obese hyperglycemic, hy-
perinsulinemic, insulin-resistant KKAy mice as previ-
ously described.29 Animals were grouped into treatment
and control groups (n ) 6) following pretest blood
glucose measurements with an Alpkem glucose au-
toanalyzer. Treatment groups had the selected com-
pound administered as a food mixture at 500 mg/kg for
4 days. The glucose level for the treated group (T) versus
the control group (C) was utilized to determine the
antihyperglycemic activity of the test compounds. The
results are summarized in Tables 1-9 as test/control
(T/C) values. For those compounds with T/C < 0.80, the
control values were statistically compared to the treat-
ment values using the nonparametric Wilcoxon rank
sum test. Those compounds determined to effect statis-
tically significant reductions in nonfasting blood glucose
are indicated with asterisks as defined in the tables.
Table 2 shows analogues of the isomeric aminoguani-
dine 9. Once again, most changes rendered the molecule
inactive. The one- and two-carbon homologues 14 and
15 were inactive as was the compound in which the
terminal amino group of 9 was removed by two carbons
(16). Replacement of the terminal amino group with
hydroxyl also resulted in an inactive analogue (17). For
comparison, guanidinoacetic acid 13 has been included
in the table, a compound we previously found to have
The substituted analogues in Table 1 show that, as
previously observed with 1, even simple modifications

1236 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Ta ble 3. Imine Analogues
Vaillancourt et al.
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Ta ble 4. Biguanides and Amidines
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control);. see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
reduced significance versus control due to poor repro-
ducibility. On the basis of the decreased activity of 13
and the inactivity of 16 and 17, we speculated that the
terminal amino group on the guanidine functionality
was a key component of the antidiabetic activity of 9.
The analogues in Table 3 represent a series of imine
analogues that were prepared. Although the unsubsti-
tuted analogue 18 and the heterocyclic analogue 28 did
display statistically significant activity in the assay,
these analogues were found to increase bilirubin levels
in a nondiabetic C57 mouse and were not investigated
further. Other modifications of the imine template,
including R-substitution, substitution on the guanidine
functionality and cyclization to the corresponding lac-
tams led to a loss of activity.
The structural similarity of the compounds reported
in this manuscript to Metformin was noted. With this
in mind, biguanide and amidine analogues were pre-
pared and are shown in Table 4. As can be seen from
the table, biguanide and amidine analogues were uni-
versally inactive within this series of compounds. This
may be the result of either the increased size of the
molecules or the change in the basicity of the amino-
guanidine functionality.

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1237
Ta ble 5. Carbon Insertions
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Ta ble 6. Diaminoguanidine Analogues
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Ta ble 7. Analogues of 52: Variations in the Carbon Chain
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
Analogues in which one of the amino groups of the
aminoguanidine moieties of 2 or 9 is extended out by
one methylene are shown in Table 5. These analogues
were targeted to try to avoid the potentially labile N-N
bonds of the leads. Unfortunately, these alterations were
uniformly detrimental to activity. The N-hydroxy ana-
logues 43 and 44 that were intermediates in the
syntheses of these analogues were also inactive.
Table 6 depicts analogues of the diaminoguanidi-
noacetic acid 7. As can be seen from the biological data,
simple substitutions, including imines (47 and 48) and
N-alkylation (49), resulted in a loss of activity. Cycliza-

1238 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Ta ble 8. Analogues of 52: Variation in the Aminoguanidine Region
Vaillancourt et al.
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
tion to the corresponding lactam (50) and incorporating
however, some of these compounds retain significant
activity. The one- and two-carbon homologues 59 and
60 were inactive.
the diaminoguanidine into a cyclic structure (51) also
led to inactive analogues. The isomeric diaminoguani-
dinoacetic acids 52 and 53, however, displayed improved
antidiabetic activity with T/C values of 0.23 and 0.43,
respectively. Dose-response testing of these compounds
indicated potencies 5-10 times that of 7 (Figures 2, 3).
Further SAR studies thus focused on these two diami-
noguanidine analogues.
The analogues shown in Table 7 are versions of 52 in
which variations have been made to the carbon chain
of the molecule. As in the other series that were previous
described, R-substitution (e.g. compounds 54-58) re-
sults in a decrease in activity. Unlike other series,
Table 8 depicts compounds in which the guanidine
moiety of 49 has been altered. Replacement of one or
both of the hydrazine fragments of 52 with various
amines and heterocycles is the basis of analogues 61-
77. All of these examples proved to be inactive. Replace-
ment of all of the nitrogens except for the terminal
amino group with carbon resulted in 78 which was also
inactive.
Conformationally restricted analogues of the diami-
noguanidines are shown in Table 9. Analogues 79-81
were attempts to restrict the diaminoguanidine func-

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1239
Ta ble 9. Cyclic Diaminoguanidine Analogues
a
Mouse insulin sensitizing screen nonfasting blood glucose levels (test/control); see Experimental Section: *p < 0.05, **p < 0.01, ***p
< 0.001.
F igu r e 1. Compound 9 given as admixture to food. Whole
blood glucose measured prior to treatment and day 3 using
the Alpkem 2700 Analyzer; *p < 0.05.
F igu r e 3. Compound 53 given as admixture to food. Whole
blood glucose measured prior to treatment and day 3 using
the Alpkem 2700 Analyzer; *p < 0.05.
multiple groups for dosing at different levels. Based on
the high degree of efficacy displayed by 52 and 53 at
500 mg/kg, dosing was undertaken at substantially
reduced levels relative to analogue 9. The results are
summarized graphically in Figures 1-3 (numerical
tabular data can be found in Supporting Information).
The least potent analogue 9 did not achieve a significant
reduction in NFBG under 400 mg/kg, indicating that
its potency is not markedly improved over that of 1 and
2. At the 400 and 500 mg/kg doses, statistically signifi-
cant reductions in NFBG were apparent without con-
comitant reductions in food intake. Both 52 and 53, on
the other hand, proved to be considerably more potent,
inducing significant reductions in NFBG at 60 and 50
mg/kg, respectively. Although food intake was clearly
suppressed at the highest doses administered, neither
compound induced statistically significant reductions in
food intake at the lowest effective doses for glucose
lowering. These results indicate that the onset of NFBG
lowering is not directly correlated to significant effects
on food intake and that the observed antidiabetic effects
are more likely to be caused by differences in energy
expenditure. These results are similar to those observed
in the ob/ ob mice studies (vide infra).
F igu r e 2. Compound 52 given as admixture to food. Whole
blood glucose measured prior to treatment and day 3 using
the Alpkem 2700 Analyzer; *p < 0.05.
tionality into a heterocycle. These analogues were also
inactive, as was the lactam 82.
Dose-Resp on se Stu d ies in KKAy Mice. Com-
pounds 52, 53, and 9 were evaluated at multiple doses
in KKAy mice to determine their potency relative to the
lead compounds 1 and 2. At the same time, food intake
was measured to determine the extent, if any, to which
nonfasting blood glucose (NFBG) lowering might be
correlated with reductions in food intake. Experiments
were performed under the identical protocol to that
described above, except that animals were divided into
Cr ea tin e Tr a n sp or t a n d Cr ea tin e Kin a se Activ-
ity. 3-Guanidinopropionic acid (1) was found to ac-
cumulate in muscle tissue. A likely mechanism for this
accumulation is that 1 is known to be a substrate for

1240 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Ta ble 10. Inhibition of Creatine Transport
Vaillancourt et al.
ment for 14 days, compounds 7 and 9 decreased the
blood glucose level of ob/ ob mice (p < 0.05). The effects
of these compounds were statistically different from
control but not significantly different from the initial
lead 1. By comparison, 53 also decreased the blood
glucose level of ob/ ob mice, but its effect was signifi-
cantly greater than that of the initial lead, 1 (p < 0.05).
test substance (1 mM)
inhib of [14C]creatine uptake (%)
creatine
1
2
7
92
97
36
26
88
52
The mean initial body weights for four of the experi-
mental groups were approximately the same (52-55 g;
Table 12). In one group (53) the mean initial body
weight was greater (59 g) due to a randomization error.
This represents a negative bias in the data for weight
loss. On the basis of comparison of the terminal body
weights in the mice, 1 and 9 produced a similar,
statistically significant loss of body weight by the mice
(p < 0.05). Compounds 7 and 53 produced a significant
loss of body weight compared to the controls (p < 0.05)
as well as compared to the mice treated with 1 or 9
(p < 0.05). The average daily food consumption was
significantly reduced by 1 and 53 (p < 0.05), but there
was no significant effect of 7 and 9 on this parameter.
This indicates that the relationship of reduced food
intake and weight loss is a weak one and that differ-
ences in energy expenditure by the experimental groups
are likely to have occurred.
Ta ble 11. Creatine Kinase Activity
substrate
reaction rate (relative to creatine)
100
2.4
creatine (1 mM)
1 (100 mM)
2 (50 mM)
7
52
53
0.24
0
0.1
0
the plasma membrane creatine transporter30 and to
reduce cellular creatine concentration by antagonizing
cellular creatine uptake.31 In addition, 1 is also known
to be a substrate for creatine kinase32 and to reduce
cellular ATP concentrations under conditions where the
forward flux rate (â-GPA + ATP f â-GPAP + ADP) is
high.33 The active analogues were therefore tested as
substrates for these two enzymes. Reduction in activity
with either of these enzymes should lead to decreased
accumulation in muscle tissue. To determine the rela-
tive antagonism of creatine transport by analogues of
1, the effect of these compounds on the uptake of [14C]-
creatine by cultured rat smooth muscle cells was
determined. The results are summarized in Table 10.
These data indicate that the creatine transporter dis-
criminates between the analogues of 1. Inhibition of A10
cell creatine uptake by 7 is low relative to that of 1.
To explore the relative substrate discrimination among
the analogues of 1 by creatine kinase, creatine kinase
activity was assayed in the forward direction by mea-
suring the conversion of ATP to ADP by enzyme
extracted from rabbit muscle (specific activity 350 U/mg;
Boehringer Mannheim, Indianapolis, IN). The results
are shown in Table 11. These data indicate that creatine
kinase discriminates between the analogues of 1. Phos-
phorylation of 7 and 53 was not detectable in this assay.
Compound 52 had a negligible rate of phosphorylation.
These results suggest that the likelihood of these
compounds to accumulate in muscle was very low.
P h a r m a cology in ob/ob Mice. Pharmacology stud-
ies with 3-guanidinopropionic acid (1) demonstrated a
reduction in hyperglycemia in obese, diabetic mice as
well as weight loss in the animals. To determine if these
activities were retained in the active compounds from
the current series, exemplary compounds from each of
the active series were tested with ob/ ob mice. These
mice are hyperglycemic and obese compared to normal,
wild-type mice due to a mutation in the leptin signal
sequence.34 The mice received test substances as an
admixture in the diet. Three of the test substances were
add to the diet at similar concentrations and produced
nearly similar levels of drug intake (Table 12). One of
the test substances, 53, was known to be more potent
based on earlier studies and was added to the diet at a
lower concentration. This resulted in an averaged daily
drug intake for 53 of approximately 6% of the dose of
the other test substances.
Body composition analysis demonstrated that the lean
body mass of mice that received the initial lead com-
pound, 1, was nearly similar to that in controls (Table
12). By comparison, the fat body mass of the mice was
significantly reduced by 1 compared to the control mice
(p < 0.05). Compounds 7, 9, and 53 produced small but
statistically significant reductions in lean body mass
compared to controls (p < 0.05). These compound also
produced statistically significant reductions in the fat
body mass of the mice (p < 0.05). For each of these
compounds, the reduction in fat body mass was 2 times
the lean tissue mass that was lost. This indicates that
the reduction of body weight in response to these
compounds may involve a relatively selective mobiliza-
tion of stored fat and subsequent clearance of this
material from the system.
P h a r m a cology in In su lin -Resista n t Mon k eys. A
previous study of the initial lead compound, 1, demon-
strated that the compound improved the disposal rate
for glucose in monkeys that were insulin-resistant. To
determine whether 53 retained this important property
of the initial lead, three insulin-resistant rhesus mon-
keys were given 53 or a placebo once daily for 4 days
every other week. The treatment sequence was placebo,
10 mg/kg/day 53, placebo, and 3 mg/kg/day 53. Plasma
glucose and insulin levels were determined after an
overnight fast and for 3 h after a liquid meal. The first
and second placebo phases produced similar results;
therefore, the placebo data were averaged for the
purpose of statistical analysis. The results are shown
in Table 13. Compound 53 produced a statistically
significant reduction in the fasting plasma glucose level
of rhesus monkeys (p ) 0.02 treatment effect in repeated
measures ANOVA). Post hoc analysis of the results
indicated that the effect was statistically significant at
a dose of 10 mg/kg/day. The fasting plasma insulin
concentration, post-prandial glucose response, and post-
prandial plasma insulin AUC all tended to be decreased
by drug treatment (p ) 0.10, 0.23, and 0.07, respectively,
The pretreatment blood glucose concentration was
similar in experimental groups (Table 12). After treat-

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1241
Ta ble 12. Effect of 2 Weeks of Oral Administration of Test Compounds on ob/ ob Mice
test substance
sample size (n)
control
1
7
9
53
15
15
8
8
8
average daily dose (mg/kg)
initial blood glucose (mg/dL)
terminal blood glucose (mg/dL)
initial body weight (g)
0
377 ( 20
229 ( 18
147 ( 18*
54 ( 1
48 ( 1*
75 ( 4*
24 ( 0.4
24 ( 1*
406 ( 12
226 ( 20
85 ( 13*
52 ( 1
436 ( 20
227 ( 20
97 ( 11*
52 ( 1
26 ( 1
228 ( 19
226 ( 29
55 ( 1
53 ( 1
90 ( 3
25 ( 1
28 ( 0.9
228 ( 34
57 ( 3*,‡
59 ( 2*
terminal body weight (g)
43 ( 1*,‡
81 ( 2
46 ( 1*
87 ( 5
44 ( 1*,‡
52 ( 2*,‡
22 ( 0.7*
22 ( 0.9*
average daily food consumption (g/kg)
terminal lean body weight (g)
terminal fat body weight (g)
22 ( 0.4*,‡
20 ( 0.7*,‡
23 ( 0.4*
23 ( 0.8*
*p < 0.05 vs control. ‡ p < 0.05 vs 1.
Ta ble 13. Effect of 53 in Insulin-Resistant Rhesus Monkeys
53 53
reductions in food consumption, suggesting that they
are altering the overall rate of metabolism. Finally,
when 53 was administered to insulin-resistant rhesus
monkeys, it induced reductions in fasting and post-
prandial serum glucose with concomitant reductions in
plasma insulin levels, suggesting that it is improving
the actions of endogenous insulin. Both 7 and 53 were
selected for further preclinical development.
placeboa (3 mg/kg/d) (10 mg/kg/d)
fasting plasma glucose
(mg/dL)
83 ( 11
81 ( 8
171 ( 12
41 ( 11
38 ( 24
64 ( 9*
146 ( 19
35 ( 7
post-prandial plasma glucose 185 ( 36
AUC (g‚min/mL)
fasting plasma insulin
(µU/mL)
post-prandial plasma insulin
AUC (mU‚min/dL)
111 ( 43
94 ( 44
30 ( 20
Exp er im en ta l Section
a
Mean of first and second placebo phases. *p < 0.05 vs control.
Ch em istr y. All melting points (mp) were obtained on a
capillary melting point apparatus and are uncorrected. Proton
magnetic resonance spectra were recorded on a Bruker AM-
300 spectrometer in the deuterated solvents indicated. Chemi-
cal shifts were recorded in parts per million (δ scale) and are
reported relative to internal tetramethylsilane or 3-(trimeth-
ylsilyl)-1-propanesulfonic acid, sodium salt. 13C NMR spectra
were recorded on a Bruker AM-300 spectrometer at 75.4 MHz.
Flash column chromatography separations were carried out
using EM Science silica (mesh 230-400). Electron impact (EI)
mass spectra were obtained with an ionization voltage of 70
eV. Alternatively, ionization was achieved by fast atom
bombardment (FAB). Reagents and solvents were purchased
from common suppliers and were used as received. All non-
aqueous reactions were run under a nitrogen atmosphere. All
starting materials were commercially available unless other-
wise noted. Compound 13 is available from commercial
sources. Compounds 8,4 9,5 18,6 28,35 31,7 and 3221 were
prepared according to literature procedures.
for the treatment effect in repeated measures ANOVA
analyses). The finding that the fasting and post-prandial
plasma glucose levels are reduced by 53 in the presence
of a reduced level of circulating insulin suggests that
the test compound improved the action of endogenous
insulin. These findings indicate that this compound
retains the activity of the initial lead in nonhuman
primates.
Con clu sion
3-Guanidinopropionic acid (1) was previously found
to possess antidiabetic activity and to induce weight loss
selectively from adipose tissue in obese animals. An
initial phase of analogue work led to the discovery of
an aminoguanidinoacetic acid analogue (2) that retained
equipotent antidiabetic activity but was less susceptible
to processing by the creatine transport/kinase system,
thereby possessing a reduced potential to accumulate
in muscle. A second phase of analogue work was
launched to try to identify compounds with greater
antidiabetic potency that retained the resistance to
creatine-like metabolism exhibited by 2. The SAR about
2 proved to be just as narrow as that established for 1,
with simple substituted, homologated, and bioisosteric
analogues losing all activity. Eventually it was discov-
ered that the simple regioisomeric aminoguanidinoacetic
acid 9 retained activity, as did three closely related
diaminoguandinoacetic acid analogues, one of which (7)
was equipotent to 2 and two of which (52 and 53) were
substantially more potent. After establishing that all
three diaminoguanidinoacetic acid regioisomers were
less susceptible to creatine-like metabolism than 1 and
therefore less likely to accumulation in muscle, testing
in additional animal models of obesity and diabetes was
undertaken. All three induced weight loss in obese
diabetic ob/ ob mice. The decrease in fat body mass was
twice that of lean body mass, suggesting that the weight
loss occurred selectively from adipose tissue in a manner
similar to 1. Furthermore, it was demonstrated that 52
and 53 induce reductions in NFBG in diabetic KKAy
mice that are independent of statistically significant
Gen er a l P r oced u r e for th e Syn th esis of 4-6. A solution
of the hydrazino acid (2.36 g, 20 mmol) and 2-methyl-2-
thiopsuedourea sulfate (2.78 g, 20 mmol) in 20 mL 1 N NaOH
was heated to reflux for 2 h. The solution was concentrated to
one-half volume and cooled. The resulting product was filtered
and washed with water and dried in vacuo.
2-[2-(Am in oim in om eth yl)h yd r a zin o]bu ta n oic Acid (4).
Compound 4 was prepared from 2-hydrazinobutyric acid7 [mp
187-189 °C] by the general method above: yield 63%; mp
1
229-232 °C dec; H NMR (D2O) δ 3.25 (t, J ) 6.5 Hz, 1 H),
1.64-1.54 (m, 2 H), 0.89 (t, J ) 7.5 Hz, 3 H); 13C NMR (DMSO)
δ 179.1, 160.4, 67.9, 25.3, 13.0; MS (FAB) m/z (relative
intensity) 161 (M + H, base), 115 (8), 74 (2), 60 (3). Anal.
(C5H12N4O2) C, H, N.
[2-(Hyd r a zin oim in om eth yl)h yd r a zin o]a cetic Acid (7).
Ethylhydrazinoacetate hydrochloride (9.28 g, 60 mmol) was
saponified by refluxing in 120 mL of 1 N NaOH for 2 h. To
the hot solution was then added N-amino-S-methylisothiourea
hydroiodide8 (13.98 g, 60 mmol) and the solution was refluxed
for an additional 2 h. The solvent was removed. The crude
product was dissolved in methanol and filtered to remove the
NaCl. The filtrate was condensed and dried by high vacuum.
The residue was then stirred with 150 mL MeOH overnight.
The resulting white solid was filtered. This solid was then
refluxed in 100 mL MeOH for 2 h to remove any impurities.
The mixture was then cooled and filtered. The resulting solid
was dried in vacuo to yield 2.14 g (24%) of 7 as an off-white
solid: mp 201-203 °C dec; 1H NMR (D2O) δ 3.39 (s, 2 H); 13
C
NMR (D2O) δ 178.6, 159.5, 54.0; MS (FAB) m/z (relative
intensity) 148 (M + H, base). Anal. (C3H9N5O2) C, H, N.

1242 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
2-{1-[Am in o(im in o)m eth yl]h yd r a zin o}a cetic Acid (10).
To a stirring suspension of aminoguanidine bicarbonate (100
g, 734 mmol) in water (200 mL) was added bromoacetic acid
(100 g, 720 mmol). After initial effervescence the homogeneous
solution was refluxed overnight, cooled to ambient tempera-
ture, and solvent evaporated to dryness. The residue was
suspended in EtOH (200 mL) and sonicated. The solid was
filtered to afford 13.6 g (9%) of the title compound as a white
solid: mp 163-165 °C; 1H NMR (D2O) δ 4.25 (s, 2 H); 13C NMR
(D2O) δ 60.6, 158.0, 175; MS (FAB+) m/z 133 (M + H). Anal.
(C3H8N4O2‚HBr) C, H, N, Br. To a stirring solution of LiOH‚
H2O (0.98 g, 23.4 mmol) in MeOH (20 mL) was added the salt
from above (5.0 g, 23.4 mmol) in MeOH (50 mL). After a solid
begins to form (∼15 min) stirring was continued an additional
30 min. The resulting solid was filtered and dried in vacuo at
25 °C to afford 2.08 g (66%) of the title compound as a white
solid: mp 118-119 °C; 1H NMR (D2O) δ 4.10 (s, 2 H); 13C NMR
(D2O) δ 56.19, 158.4, 171; MS (FAB+) m/z 133 (M + H). Anal.
(C3H8N4O2‚0.17H2O) C, H, N.
3-Am in o-1,6-d ih yd r o-1,2,4-tr ia zin -5(2H)-on e Mon oh y-
d r och lor id e (11). A solution of 316 (1.65 g, 14.7 mmol) in H2O
(40 mL) and concentrated HCl (8 mL) was hydrogenated over
10% Pd/C (165 mg) at 40 psi for 1.5 h. The reaction mixture
was filtered and the filtrate concentrated in vacuo. The residue
was recrystallized from MeOH to yield 11 (1.48 g, 68%) as a
white solid: mp 204-205 °C dec; 1H NMR (CD3OD) δ 3.60 (s,
2 H); 13C NMR (CD3OD) δ 175.2, 139.6, 51.7; MS (EI) m/z
(relative intensity) 114 (M+, 39), 85 (50), 42 (base). Anal.
(C3H6N4O‚HCl) C, H, N, Cl.
This material was dissolved in water (70 mL), and the solution
was chilled in ice before neutralization to pH 7 with 1 N NaOH
(required about 55 mL). Acetone was added to the resulting
solution until no more precipitate appeared (ca. 300 mL). The
mixture was stirred in the ice bath for 10 min before filtering,
giving 88 as a fine white solid (8.97 g, ca. 85% yield): mp
184 °C dec; 1H NMR (D2O) δ 3.98 (br s, 2 H), 2.63 (s, 3 H);
13C NMR (D2O) δ 174.0, 169.5, 46.9, 13.2; FAB MS m/z
(rel intensity) 149 (M + H, 100). Anal. (C4H8N2O2S) C, H,
N.
2-{[[(2-Am in oe t h y l)a m in o ](im in o)m e t h y l]a m in o }-
a cetic Acid (16). To a solution of isothiourea 88 (4.0 g, 27.0
mmol) in water (54 mL), cooled in a cold water bath, was added
in one portion ethylenediamine (9.02 mL, 135 mmol). The
solution was stirred at room temperature for 7 h. The solution
was then diluted with isopropyl alcohol to a volume of 250 mL.
Chilling at 0 °C overnight afforded 16 as a fine white solid
1
(2.83 g, 65%): mp 157-158 °C dec; H NMR (D2O) δ 3.81 (s,
2 H), 3.29 (t, J ) 6 Hz, 2 H), 2.82 (t, J ) 6 Hz, 2 H); 13C NMR
(D2O) δ (mult) 175.3 (s), 156.3 (s), 44.7 (t), 43.6 (t), 39.4 (t);
FAB MS m/z (rel intensity) 161 (M + H, 100). Anal. (C5H12N4O2‚
0.86H2O) C, H, N.
Gen er a l P r oced u r e for th e Syn th esis of 20-22. Con-
centrated (37%) hydrochloric acid (10 mL, 120 mmol) was
added slowly and cautiously to a magnetically stirred suspen-
sion of aminoguanidine bicarbonate (15.0 g, 110 mmol) in 20
mL of distilled water. Gas evolved and a clear solution of
aminoguanidine hydrochloride was formed. To this solution
was added 125 mmol of the keto acid all at once. The mixture
was diluted with 15 mL of distilled water to facilitate stirring,
and it was stirred 24 h at 25 °C. After an additional 24 h at 5
°C, it was filtered and the filtrant was washed with two
successive 20 mL portions of distilled water. The solid was
dried several hours at 25 °C in a stream of air, then at 20
Torr/45 °C/4 days.
2-[(Am in oim in om eth yl)h ydr azon o]pr opan oic Acid Hy-
d r och lor id e (20). Compound 20 was prepared from pyruvic
acid according to the general procedure above; yield 66% of
20 as a white powder: mp 238-239 °C dec (lit. 130°, 230 °C);
1H NMR (D2O) δ 1.97 (s, major (87%) stereoisomer, 3 H), 1.95
(s, minor (13%) stereoisomer, 3 H); 13C NMR (D2O) δ 166.9
(major stereoisomer), 155.9 (major stereoisomer), 145.8 (major
stereoisomer), 20.0 (minor stereoisomer), 12.1 (major stereoi-
somer); MS(FAB) m/z (relative intensity) 145 (M + H+, base).
Anal. (C4H8N4O2‚HCl) C, H, N, Cl.
Gen er a l P r oced u r e for th e Syn th esis of 33-36. To the
amino acid (4.33 g, 58.2 mmol) suspended in absolute ethanol
(0.3 L), in a 60 °C oil bath, was added in order, triethylamine
(11.67 g, 0.115 mol) and water (30 mL). The mixture was
stirred until the majority of the amino acid had dissolved (30
min), then 9014 (30 g, 0.115 mol) was added in 6 approximately
equal portions over 6 h. The mixture was allowed to stir for
12 h at 60 °C, then was cooled to room temperature. The solid
was isolated by filtration, washed with absolute ethanol and
dried in vacuo.
N-(2-Am in o-1H-im id a zol-1-yl)-N-(1-m eth yleth ylid en e)-
a m in e (86). To a mechanically stirring solution of amino-
guanidine carbonate (100 g, 734 mmol) in EtOH (1.2 L) and
concentrated aqueous HCl (1 mL) was added acetone (116.8
g, 2 mol) and the resulting solution heated at reflux overnight.
The resulting solution was cooled to ambient temperature and
solvent evaporated to ∼800 mL. To the resulting solution was
added chloroacetaldehyde (28.8 g, 367 mmol) and the mixture
heated to reflux for 4 h. The solvent was removed under
reduced pressure and the residue suspended between H2O and
CH2Cl2. The layers were shaken, the organics separated, dried
over Na2SO4 and the solvent removed in vacuo. The aqueous
layer was continuously extracted with CH2Cl2 overnight and
the organic layer evaporated. The extracted organic residues
were combined and purified via SiO2 flash column chomatog-
raphy (eluant 7% MeOH/CH2Cl2). The chromatographed resi-
due was recrystallized from EtOAc/hexane to afford 15.26 g
(15%) of the title compound as a highly crystalline white
1
solid: mp 123-125 °C; H NMR (CDCl3) δ 2.07 (s, 3 H), 2.19
(s, 3 H), 4.13 (br s, 2 H), 6.59 (d, J ) 2.0 Hz 1 H), 6.64 (d, J )
2.0 Hz, 1 H); MS (EI) m/z 138 (M+), 138, 82, 69, 56, 55, 53, 42,
41, 40, 25. Anal. (C6H10N4) C, H, N.
2-[(2-Am in o-1H-im id a zol-1-yl)a m in o]a cetic Acid (12).
To a stirring solution of 86 (5.0 g, 36.18 mmol) in H2O (40 mL)
in a pressure tube at ambient temperature was added chloro-
acetic acid (3.42 g, 36.18 mmol) and triethylamine (3.65 g,
36.18 mmol). The tube was closed and heated at 100 °C
overnight and cooled to ambient temperature. The aqueous
solution was washed with EtOAc (50 mL) and the aqueous
layer evaporated to dryness. The residue was taken up in
Im id oca r bon im id ic Dia m id e N-(Acetic a cid ) (33). Com-
pound 33 was prepared from glycine according to the general
procedure above; yield 28% of 33 as a fine, powdery, pale
1
yellow solid: mp 226-228 °C dec; H NMR (D2O) δ 3.79 (s, 2
H); 13C NMR (D2O) δ 172.5, 160.3, 45.3, 41.6; IR (Nujol) 3396,
3316, 3119, 1635, 1601, 1557, 1506, 1405, 1363, 1275, 733
cm-1; FAB/MS 319 (19.6), 160 (M+ + 1, base); HRMS m/z calcd
for C6H13N5O3 160.0834, found 160.0831.
EtOH (50 mL) and heated on
a steam bath to remove
byproducts and unreacted starting material. The solids were
filtered and dried in vacuo to afford 2.16 g (38%) of the title
1
compound as a tan solid: mp 265-267 °C; H NMR (D2O) δ
4.52 (s, 2 H), 6.73 (d, J ) 2.0 Hz, 1 H), 6.91 (d, J ) 2.0 Hz, 1
H); 13C NMR (D2O) δ 49.4, 114.6, 117.8, 146.9, 177.4; MS (FAB)
m/z 157 (M + H), 469, 314, 313, 189, 158, 157, 142, 141, 112,
99. Anal. (C5H8N4O2) H; C: calcd, 38.46; found, 36.39. N: calcd,
35.88; found, 33.98.
2-{[Im in o(m et h ylsu lfa n yl)m et h yl]a m in o}a cet ic Acid
(88). A solution of 87 (9.5 g, 71 mmol) and iodomethane (8.8
mL, 142 mmol) in methanol (140 mL) was stirred at room
temperature for 18 h. Attempted trituration of a small sample
yielded the hydroiodide salt of 88 as a gummy oil, so the entire
reaction mixture was concentrated in vacuo to a viscous oil.
Gen er a l P r oced u r e for th e Syn th esis of 92a ,b. To a
stirring solution of 9115 (10.0 g, 52.5 mmol) in EtOH (350 mL)
was added the amino acid (52.5 mmol) and triethylamine (5.35
g, 52.57 mmol). Water was added incrementally until a
solution was realized (∼50 mL) and the mixture stirred for
1.5 h. The solvent was evaporated to dryness to afford a
hygroscopic solid. Dry ammonia was slowly bubbled into a
stirring solution of this compound (45.8 mmol) in EtOH (125
mL). The resulting solution was stirred for 15 min in which
time a solid began to form. The suspension was stirred for 1
h, and the solid filtered.

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1243
92a was prepared from glycine according to the general
procedure above: yield 99% as an off-white solid; 1H NMR
(D2O) δ 4.03 (s, 2H); 13C NMR (D2O) δ 47.5, 159.0, 159.9, 175.3;
MS (FAB) m/z 191 (M + H). Anal. (C4H6N4O3S‚NH3‚0.4H2O)
H, N; C: calcd, 22.40; found, 22.83.
101-102 °C; 1H NMR (CDCl3) δ 5.04 (br d, 2 H, NH2), 4.19 (q,
J ) 7 Hz, 2 H), 3.97, 3.90 (two s, 4 H), 1.49, 1.44 (two s, 9 H),
1.28 (t, J ) 7 Hz, 3 H), mixture of rotamers and/or E, Z
isomers; 13C NMR (CDCl3) δ 170.3 (minor), 167.7 (major), 156.3
(major), 155.2 (minor), 151.7 (major), 151.3 (minor), 81.6
(minor), 81.3 (major), 61.4 (minor), 61.2 (major), 49.2 (major),
49.0 (minor), 48.4, 28.3 (minor), 28.2 (major), 14.2; MS m/z
276 (M + H), 260, 220 (base), 204, 176, 116, 57. Anal.
(C11H21N3O5) C, H, N.
2-[[2-Am in o-2-(h yd r oxyim in o)eth yl](ter t-bu toxyca r b-
on yl)a m in o]a cetic Acid (96a ). To a suspension of 95a (1.95
g, 7.1 mmol) in 95% EtOH (20 mL) was added 1.0 M NaOH
(28.5 mL). After the mixture was stirred at room temperature
for 2.5 h, 1.0 M HCl (28.5 mL) was added, and the mixture
was cooled at 0 °C overnight. The white solid formed was
filtered and washed with cold H2O to provide 1.48 g (85%) of
96a as a white solid: mp 219-220 °C; 1H NMR (DMSO) δ 9.33
(1 H), 5.81 (br s, 2 H), 3.87 (s, 2 H), 3.82, 3.79 (two s, 2 H),
1.38, 1.35 (two s, 9 H); 13C NMR (DMSO) δ 171.3, 154.5, 150.9,
79.6, 50.4, 50.0, 27.8; MS (FAB) m/z 249, 248 (M + H), 247,
232, 176, 192, 148, 57. Anal. (C9H17N3O5) C, H, N.
2-{[2-Am in o-2-(h yd r oxyim in o)eth yl]a m in o}a cetic Acid
(43). 96a (2.0 g, 8.1 mmol) was dissolved in a solution of
anhydrous HCl in dioxane (2.8 M, 15 mL). A solid precipitated
very rapidly, and the mixture was stirred for 5 h. The mixture
was concentrated under reduced pressure, and the solid which
was obtained was washed with Et2O to provide 1.6 g (89%) of
43 as a hygroscopic white solid: mp 78-80 °C; 1H NMR (D2O)
δ 3.97 (s, 2 H), 3.88 (s, 2 H); 13C NMR (D2O) δ 168.8, 153.3,
48.0, 43.8; MS (FAB) m/z 204, 148 (base), 40, 18; exact mass
calcd for C4H9N3O3 148.0722, found 148.0726. Anal. (C4H9N3O3‚
2HCl‚0.44H2O) C, H, N, Cl.
2-[(2-Am in o-2-im in oet h a n im id oyl)a m in o]a cet ic Acid
(38). 92a (8.6 g, 41.3 mmol) was dissolved in concentrated
aqueous HCl (25 mL) and allowed to stir for 2 h. The solution
was evaporated to dryness and dried under vacuo at 25 °C
overnight. The resulting solid (10.0 g, 37 mmol) was taken up
in minimal water and neutralized by dropwise addition of 1
M aqueous NaOH (74 mL, 74 mmol). The resulting solid was
filtered and washed with water (2 × 50 mL). The solid was
dried overnight in vacuo at 25 °C to afford 4.82 g (81%) of 38
1
as a tan solid: mp >300 °C; H NMR (D2O/DCl) δ 4.31 (s, 2
H); 13C NMR (D2O/DCl) δ 44.31, 154.0, 155.0, 168.9; MS
(FAB+) m/z 145 (M + H). Anal. (C4H8N4O2‚1.1H2O) C, H, N.
Gen er a l P r oced u r e for th e Syn th esis of 41-42. To a
solution of 9316 (20.0 g, 56.8 mmol) in CH3CN (250 mL) at 0
°C was added a solution of the neutralized amino acid ethyl
ester (56.8 mmol) in CH3CN (200 mL) over 10 min. After
warming to room temperature, the mixture was stirred for 2
h. The mixture was concentrated under reduced pressure to
provide the triflate salt as a waxy solid which was used directly
in the next reaction. The triflate salt was dissolved in 12 M
HCl (100 mL), and the mixture was heated at 95 °C for 1 h.
The reaction mixture was cooled to -15 °C in a freezer and
seeded. The mixture was allowed to stand in the freezer for
2.5 days. The solid was collected by filtration and washed with
THF.
2-[(2-Am in oeth a n im id oyl)a m in o]a cetic Acid Dih yd r o-
ch lor id e (41). Compound 41 was prepared from glycine ethyl
ester hydrochloride according to the general procedure above:
2-[(2-Am in o-2-im in oeth yl)a m in o]a cetic Acid Dih yd r o-
ch lor id e (45). To a suspension of 96a (3.0 g) in MeOH (200
mL) in a Parr reaction vessel was added Raney nickel (50%
slurry in H2O). The reaction vessel was pressurized with H2
(45 psi) and heated to 65 °C using a YSI Thermistemp
temperature controller for 2 h. The mixture was cooled to room
temperature and carefully filtered through Celite under N2.
The solvent was removed in vacuo to provide 2.5 g (89%) of a
1
yield 84% as a white solid; mp 191-192 °C; H NMR (D2O) δ
4.25 (s, 2 H), 4.18 (s, 2 H); MS (FAB) m/z 263, 133, 132 (base),
45, 30. Anal. (C4H9N3O2‚2HCl) C, H, N, Cl.
Gen er a l P r oced u r e for th e Syn th esis of 95. To a
solution of the amino acid ethyl ester hydrochloride (94) (36
mmol) in H2O (10 mL) at 0 °C was added 37% aqueous
formaldehyde (2.7 mL) and a solution of sodium cyanide (1.75
g) in H2O (3.6 mL). The resulting mixture was stirred at 0-5
°C for 4 h and extracted with Et2O (6 × 30 mL). The combined
organic phase was dried (Na2SO4) and filtered. Anhydrous HCl
(g) was bubbled through the ethereal extract until a milky
white mixture formed. After cooling to 0 °C, the white solid
which formed was collected by filtration to provide the desired
nitrile as an off-white solid. To a suspension of the nitrile (22.0
mmol) and triethylamine (3.1 mL, 22.0 mmol) in CH3CN (25
mL) was added di-tert-butyl dicarbonate (9.6 g, 44.0 mmol).
After stirring the mixture for several minutes, 4-(dimethy-
lamino)pyridine (0.67 g, 5.5 mmol) was added, and the mixture
was stirred at room temperature for 24 h. The reaction mixture
was concentrated to near dryness in vacuo and partitioned
between EtOAc (50 mL) and H2O. After the aqueous phase
was extracted with EtOAc (1 × 50 mL), the combined organic
phase was washed with 1 N HCl (1 × 25 mL) and dried
(MgSO4). The solvent was removed under reduced pressure,
and the residue was purified by chromatography (20% EtOAc:
hexane) to provide the Boc-protected aminonitrile as a colorless
oil. To a solution of NaHCO3 (3.5 g, 42.1 mmol) in H2O (30
mL) was added slowly a solution of hydroxylamine hydrochlo-
ride (2.9 g, 42.1 mmol) in H2O (5 mL). The aqueous solution
of hydroxylamine was added to a solution of the Boc-protected
aminonitrile (42.1 mmol) in 95% EtOH (400 mL). The mixture
was heated at reflux overnight. The EtOH was removed under
reduced pressure, and the residue was partitioned between
EtOAc (100 mL) and H2O (50 mL). The aqueous phase was
extracted with EtOAc (1 × 50 mL), and the combined organic
phase was dried (MgSO4). The solvent was removed in vacuo.
1
white solid: mp 250 °C dec; H NMR (DMSO) δ 11.0 (br s),
10.9 (br s), 10.2 (br s), 4.10 (s, 2 H), 3.67 (s, 2 H), 1.35 (s, 9 H);
13C NMR (DMSO) δ 175.5, 170.3, 169.7, 154.7, 153.8, 79.7,
79.6, 54.1, 53.3, 50.3, 50.0, 27.8, some signals doubled due to
rotamers; MS (FAB) m/z 232 (M + H, base), 176, 58, 57, 41.
Anal. (C9H17N3O4‚0.22H2O) C, H, N.
The solid from the above reaction (5.0 g, 21.6 mmol) was
suspended in a solution of anhydrous HCl in dioxane (2.8 M,
60 mL). The suspension was stirred at room temperature for
40 h. An 1H NMR of an aliquot showed that some starting
material remained. Anhydrous HCl in dioxane (3.3 M, 50 mL)
was added, and the suspension was stirred at room temper-
ature for another 24 h. The reaction was judged complete by
1H NMR. The suspension was filtered, and the filter cake was
washed with Et2O to obtain 3.97 g (90%) of 45 as a white
solid: mp 172-174 °C; 1H NMR (D2O) δ 4.19 (s, 2 H), 3.96 (s,
2 H); 13C NMR (D2O) δ 169.8, 162.5, 48.8, 46.4; MS (FAB) m/z
132 (base), 88, 58, 45. Anal. (C4H9N3O2‚2HCl) C, H, N, Cl.
Met h yl 1-H yd r a zin eca r b oh yd r a zon ot h ioa t e H yd r o-
iod id e (97). A mixture of thiocarbohydrazide (25.76 g, 242.7
mmol) and methyl iodide (18.2 mL, 291.9 mmol) in methanol
(240 mL) was stirred under nitrogen at reflux for 1 h 45 min.
The resulting pale yellow solution was cooled to room tem-
perature until crystallization commenced, and then it was
diluted with ether (200 mL) to facilitate complete crystalliza-
tion. After chilling in ice for 2 h, the mixture was filtered and
the collected solid was washed with ether. Drying in vacuo
gave pale yellow prisms (48.27 g, 80%). The solid discolors in
air, so it is stored in a dark bottle in the refrigerator: 1H NMR
(D2O) δ 2.49 (s); 13C NMR (DMSO-d6) δ 170.8, 12.7.
Compound 95a was prepared from glycine ethyl ester
hydrochloride according to the general procedure above; yield
87% of a solid that was approximately a 9:1 mixture of the
desired 95a and the cyclized lactam. The crude product could
be separated by chromatography (70% EtOAc:hexane): mp
2-[(Dih yd r a zin om eth ylen e)a m in o]Acetic Acid (52). A
solution of 97 (25.0 g, 101 mmol) and glycine (6.314 g, 83.98
mmol) in water (50 mL) and 12.5 N NaOH (8.89 mL, 111
mmol) was stirred under nitrogen at 75-80 °C for 3 h. The

1244 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
solution was chilled in ice while still under nitrogen before
the portionwise addition of absolute ethanol (550 mL in 50
mL portions), stirring between each addition until precipitation
was complete. Seeding with authentic product greatly facili-
tates this initial trituration. The mixture was then stirred for
15 min at 0 °C before filtering. The collected solid was washed
thoroughly with absolute ethanol. Drying gave a light pink
powder (8.04 g). The crude solid was dissolved in water (30
mL), filtered to remove some fine insoluble material, and then
diluted to a volume of 250 mL with absolute ethanol. Precipi-
tation began almost immediately and was accelerated by
sonication for a few seconds. After standing at room temper-
ature for 10 min, the mixture was filtered, giving a pale rose
powder (5.25 g, 42%): mp 200 °C dec. An analytical sample
could be prepared as follows: 2.04 g was dissolved in 6 mL
water, and the pink solution was stirred with activated
charcoal for 10 min at room temperature. The mixture was
filtered through a fine glass frit, and the nearly colorless
filtrate was diluted to 100 mL with absolute ethanol. After
standing 1 h at room temperature with occasional sonication,
1.61 g of white powder was obtained. This was dissolved again
in water, treated with charcoal again, filtered, and diluted to
100 mL with absolute ethanol. After 2 days at room temper-
ature, 1.35 g of white prisms (mp 202 °C, dec) were obtained.
The mp could not be raised any further by additional crystal-
lizations: 1H NMR (D2O) δ 3.78 (s); 13C NMR (D2O) δ 175.3,
158.1, 43.7; FAB MS m/z (rel intens.) 148 (M + H, 100), 133
(3), 116 (1). Anal. (C3H9N5O2) C, H, N.
not observable; 13C NMR (CD3OD) δ 54.8, 68.7, 129.1, 129.2,
129.4, 137.1, 171.8, 185.3, 198.6; MS (FAB) m/z (rel intensity)
284 (M + H, 99), 567 (9), 566 (8), 392 (8), 362 (7), 361 (10),
360 (57), 285 (14), 284 (99), 92 (6), 91 (72). Anal. (C11H13N3O4S)
C, H, N.
2-{1-[Hyd r a zin o(im in o)m eth yl]h yd r a zin o}a cetic Acid
(53). To a stirring solution of 99 (5.0 g, 17.64 mmol) in EtOH
(150 mL) at ambient temperature was added methyl iodide
(2.73 g, 19.41 mmol) and the resulting solution stirred
overnight. The solvent was removed in vacuo to afford 7.50 g
(quant) of the S-methyl isothiosemicarbazide as a yellow
foam: 1H NMR (CD3OD) δ 2.69 (brs, 0.6 H), 2.84 (brs, 0.4H),
4.40-4.70 (m, 2H), 5.31 (brs, 2H), 7.46 (m, 5H); MS (FAB) m/z
(rel intensity) 298 (M + H, 99), 452 (4), 374 (4), 327 (4), 326
(21), 300 (6), 299 (15), 298 (99), 208 (3), 164 (12), 91 (24). Anal.
(C12H15N3O4S) C, H, N.
To a vigorously stirring solution of this compound (25.5 g,
60 mmol) in H2O (100 mL) at ambient temperature was added
hydrazine hydrate (6.06 g, 120 mmol) slowly until 1/2 had been
added. H2O (10 mL) was added to the solid mass which had
formed and the solids broken up mechanically with a spatula.
The remaining hydrazine was then added and the solution
vigorously stirred for 1 h. The heterogeneous mixture was
sonicated and stirring continued until a thick mass had
formed. EtOH (50 mL) was added, the solid filtered, washed
with EtOH and dried in vacuo to afford 9.24 g (55%) of a white
1
solid: mp 168-170 °C; H NMR (D2O) δ 3.86 (brs, 1 H), 4.21
(brs, 1 H), 5.17 (s, 2 H), 7.39 (s, 5 H); 13C NMR (D2O) δ 54.4,
68.7, 128.1, 128.3, 128.9, 135.3, 162.0, 169.1, 173.1; MS (FAB)
m/z (rel intensity) 282 (M + H, 99), 358 (11), 283 (14), 282
(99), 281 (12), 264 (10), 238 (4), 192 (4), 148 (4), 91 (27), 87
(11). Anal. (C11H15N5O4) C, H, N.
Ben zyl 2-(2-Eth oxy-2-oxoeth yl)-1-h yd r a zin eca r boxy-
la te (98). To a stirring suspension of ethyl hydrazinoacetate
hydrochloride (5.0 g, 32.34 mmol) and N-methylmorpholine
(3.26 g, 32.34 mmol) at 0 °C was added solid N-(benzyloxy-
carbonyloxy)succinimide (8.06 g, 32.34 mmol). The mixture
was allowed to warm to ambient temperature overnight and
the solvent removed in vacuo. The residue was suspended
between EtOAc/H2O, the layers shaken, the organics separated
and dried over Na2SO4. The solvent was removed and the
residue chomatographed via SiO2 flash chromatography (elu-
ant 4:1 hexane/EtOAc) to afford 5.7 g (70%) of 98 as a white
To a solution of this solid (9.20 g, 32.71 mmol) in MeOH/
H2O (400 mL, ∼2:1 v/v) was added 10% Pd-C (1.0 g) and the
mixture hydrogenated at 30 psi for 4 h. The catalyst was
filtered through diatomaceous earth and 10% Pd-C (1.0 g) was
again added. The mixture was hydrogenated at 30 psi for 2.5
h and determined to be complete by TLC (eluant 85:14:1 CH2-
Cl2/MeOH/HCO2H). The mixture was filtered through diato-
maceous earth and solvent concentrated to ∼50 mL at which
time a solid precipitated. The solid was filtered, washed with
a minimum amount of H2O and dried in vacuo to afford 3.60
g (75%) of 53 as an off white solid: mp 196-198 °C. A second
crop was obtained by concentrating the filtrate until a solid
formed. Filtration afforded 0.90 g (19%, total yield: 94%)
additional material having identical melting point: 1H NMR
(D2O) δ 4.06 (s, 2 H); 13C NMR (D2O) δ 55.7, 160.1, 174.3; MS
(FAB) m/z (rel intensity) 148 (M + H, 99), 302 (8), 300 (3),
295 (8), 256 (3), 224 (16), 180 (3), 149 (5), 148 (99), 147 (8), 87
(5). Anal. (C3H9N5O2‚0.89H2O) C, H, N.
Gen er a l P r oced u r e for th e Syn th esis of 101. To a
suspension of the amino acid (10.0 g, 0.11 mol) and triethyl-
amine (33.5 mL, 0.24 mol) in EtOH (90 mL) and H2O (6 mL)
was added carbon disulfide (7.2 mL, 0.12 mol). After stirring
overnight, methyl iodide (7.5 mL, 0.12 mol) was added to the
yellow solution. The mixture was stirred for 1 h and concen-
trated to a slurry. The residue was dissolved in H2O (25 mL),
and concentrated HCl was added until acidic. The mixture was
extracted with Et2O (3 × 100 mL), and the organic phase was
dried (MgSO4) and concentrated to provide 101.
1
solid: mp 95-97 °C; H NMR (CDCl3) δ 1.27 (t, J ) 7 Hz, 3
H), 3.66 (s, 2 H), 4.19 (q, J ) 7 Hz, 2 H), 5.13 (s, 2 H), 6.77
(brs, 1 H), 7.33 (m, 5 H); MS (EI) m/z (rel intensity) 252 (M+,
5), 208 (16), 135 (21), 117 (33), 92 (36), 91 (99), 89 (11), 77
(13), 65 (32), 61 (15), 51 (10). Anal. (C12H16N2O4) C, H, N.
2-{1-(Am in oca r b ot h ioyl)-2-[(b e n zyloxy)ca r b on yl]-
h yd r a zin o}a cetic Acid (99). To a stirring suspension of 98
(3.0 g, 11.89 mmol) in EtOH (30 mL) at ambient temperature
was added aqueous NaOH (1 N, 11.89 mL). To the mixture
was added additional H2O (10 mL) and stirring was continued
for 1 h. (The mixture became a homogeneous solution and then
a solid precipitated.) Aqueous HCl (1 N, 11.89 mL) was then
added, the ethanol removed in vacuo and the aqueous ex-
tracted with EtOAc (2 × 100 mL). The organic layers were
combined, dried over Na2SO4, and the solvent removed to
afford 2.31 g (87%) of the CBz protected acid as white solid:
1
mp 131-133 °C; H NMR (CD3OD) δ 3.59 (s, 2 H), 5.15 (s, 2
H), 7.37 (m, 5 H); MS (FAB) m/z (rel intensity) 225 (M + H,
55), 449 (7), 359 (14), 247 (19), 237 (7), 226 (7), 225 (55), 224
(12), 181 (8), 92 (9), 91 (99). Anal. (C10H12N2O4‚0.12H2O) C,
H, N.
Compound 101 (R ) (S)-Me) was prepared from L-alanine
according to the general procedure above; yield 93% of 101 (R
) (S)-Me). A analytically pure sample was obtained by
To a stirring suspension of the acid (25.44 g, 112.7 mmol)
in EtOAc (500 mL) was added trimethylsilyl isothiocyanate
(14.79 g, 112.7 mmol) and the mixture was heated at gentle
reflux (80 °C) overnight. The resulting solution was cooled to
ambient temperature and washed with H2O (2 × 100 mL). The
organic layer was separated, dried over Na2SO4, and the
solvent evaporated to dryness. The oily residue was dissolved
in CH2Cl2 and allowed to stand at ambient temperature for 3
min in which time a solid formed. The solid was filtered,
washed with CH2Cl2 (100 mL) and dried in vacuo. The solid
was slurried in hot EtOAc (300 mL) to dissolve any sulfur
related byproducts and triturated with hexane (200 mL) to
afford 17.1 g of 99 (53%) as a white solid: mp 148-149 °C; 1H
NMR (CD3OD) δ 5.20 (s, 2 H), 7.30 (m, 5 H) remaining CH2
1
recrystallization from Et2O/hexane: mp 90-92 °C; H NMR
(D2O) δ 4.89 (q, J ) 7 Hz, 1 H), 2.59 (s, 3 H), 1.52 (d, J ) 7 Hz,
3 H); 13C NMR (D2O) δ 201.5, 176.0, 55.7, 17.5, 16.9; MS (EI)
m/z (rel intensity) 179 (M+, 93), 132 (26), 131 (14), 91 (30), 88
(16), 86 (99), 60 (20), 59 (10), 48 (10), 46 (13). Anal. (C5H9-
NO2S2) C, H, N, S.
Gen er a l P r oced u r e for th e Syn th esis of 102. To a
solution of the dithocarbamate 101 (28 mmol) in methylene
chloride (50 mL) at 0 °C was added methyl trifluoromethane-
sulfonate (3.5 mL, 31 mmol). The mixture was warmed to room
temperature and stirred for 20 h. The mixture was concen-

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1245
trated under reduced pressure to a colorless oil. The resulting
oil was dissolved in H2O (5 mL), and 1.0 M NaOH (28 mmol)
was added. The mixture was extracted with EtOAc (3 × 100
mL), and the organic phase was dried (MgSO4). After filtration,
the solvent was removed in vacuo to provide 102.
Compound 102 (R ) (S)-Me) was prepared from 101 (R )
(S)-Me) according to the general procedure above: 1H NMR
(MeOH) δ 4.46 (q, J ) 7.0 Hz, 1 H), 2.59 (s, 3 H), 2.42 (s, 3 H),
1.40 (d, J ) 7.0 Hz, 3 H).
Gen er a l P r oced u r e for th e Syn th esis of 54-58. 102 was
dissolved in absolute EtOH (25 mL), and anhydrous hydrazine
(0.14 mol) was added. The mixture was stirred for 1.5 h, and
the solid which formed was collected by filtration. The white
powder was further purified by crystallization from H2O/IPA
to give 54-58.
solution was diluted with 2-propanol (100 mL), seeded, and
left at 0 °C for several hours, affording a fine white solid (542
mg): mp 198 °C dec; MS (FAB) m/z (rel intensity) 145 (M +
H, 99), 299 (5), 290 (2), 289 (15), 221 (2), 146 (6), 145 (99), 143
(2), 100 (2), 99 (4), 87 (2); HRMS (FAB) calcd for C5H8N2O3 +
H 145.0613, found 145.0611. Anal. (C5H8N2O3) H, N; C: calcd,
41.67; found, 41.06.
A solution of the solid from above (3.00 g, 20.8 mmol) in
water (20 mL) was cooled to 0 °C before the addition of
hydrazine monohydrate (3.03 mL, 62.4 mmol). The solution
was stirred at 0 °C for 10 min and then at room temperature
for 5 h. The solution was diluted with 2-propanol to a volume
of 250 mL and seeded with authentic product. Precipitation
of product commenced immediately. After 30 min of occasional
stirring and sonication, the mixture was filtered, providing the
product as a white solid (2.87 g, 78%): mp 182 °C dec; 1H NMR
(2S)-2-[(Dih yd r a zin om eth ylen e)a m in o]p r op a n oic Acid
(54). Following the general procedure, 54 was prepared from
102 (R ) (S)-Me): yield 49% as a white powder; mp 174-176
(D2O) δ 3.82 (s, 2 H), 3.74 (t, J ) 5 Hz, 2 H), 3.39 (bs, 2 H); 13
C
NMR (D2O) δ 175.4, 157.2, 60.2, 44.3, 43.2; MS (FAB) m/z (rel
intensity) 177 (M + H, 99), 353 (10), 178 (7), 177 (99), 176 (4),
175 (2), 162 (2), 159 (3), 102 (2), 62 (2), 44 (2). Anal.
(C5H12N4O3) C, H, N.
1
°C dec; H NMR (D2O) δ 3.69 (q, J ) 7 Hz, 1 H), 1.20 (d, J )
7 Hz, 3 H); 13C NMR (D2O) δ 178.7, 157.6, 51.6, 18.3; MS (FAB)
m/z (rel intensity) 162 (M + H, 99), 323 (14), 296 (4), 163 (7),
162 (99), 161 (5), 147 (8), 144 (11), 116 (4), 44 (8), 18 (3); HRMS
(FAB) calcd for C4H11N5O2 + H 162.0991, found 162.0994.
Anal. (C4H11N5O2) C, H, N.
2-({Bis[2-(ter t-bu toxyca r bon yl)h yd r a zin o]m eth ylen e}-
a m in o)a cetic Acid (65). To a suspension of 104a (0.18 g, 1
mmol) in THF (1 mL) was added tert-butyl carbazate (0.66 g,
5 mmol). The resulting homogeneous solution was stirred at
room temperature overnight. The solid which formed was
collected by filtration and washed with cold THF to provide
N-[Bis(m eth ylsu lfa n yl)m eth ylen e]-â-a la n in e (104a ). A
solution of 103a 22 (12.12 g, 67.62 mmol) and methyl iodide (9.7
mL, 156 mmol) in acetone (100 mL) was stirred overnight at
room temperature. NMR analysis of the mixture indicated it
was only about 60% complete. The mixture was then stirred
at reflux for 5 h. After cooling to room temperature, the
mixture was filtered, giving 104a as a fine white solid (19.57
g, 90%) which was analytically pure: mp 138 °C dec; 1H NMR
(D2O) δ 3.99 (t, J ) 7 Hz, 2 H), 2.90 (t, J ) 7 Hz, 2 H), 2.84 (s,
3 H), 2.79 (s, 3 H); 13C NMR (D2O) δ 193.7, 174.9, 44.3, 31.8,
16.0, 15.7; MS (FAB) m/z 194 (M + H), 348, 224, 196, 195,
194, 180, 148, 146, 134, 73; HRMS (FAB) calcd for C6H11NO2S2
+ H 194.0309, found 194.0320. Anal. (C6H12INO2S2) C, H, N,
S.
3-[(Dih ydr azin om eth ylen e)am in o]pr opan oic Acid (59).
To a 0 °C solution of 104a (10.00 g, 31.15 mmol) in 1.00 M aq
NaOH (31 mL) was added hydrazine monohydrate (7.57 mL,
156 mmol). The solution was stirred 5 min before removing
the ice bath. Upon warming to room temperature, gas evolu-
tion became apparent. After stirring for 3.5 h at room tem-
perature, the solution was chilled in an ice bath and diluted
with 2-propanol in 50 mL portions (with seeding) to a total
volume of 600 mL. The mixture was sonicated and stirred 15
min at 0 °C before collecting the resulting light pink solid by
filtration (4.21 g). The crude solid was dissolved in water (50
mL), treated with Norite, and filtered through a fine glass frit,
giving an almost colorless solution. The solution was diluted
with 2-propanol to a volume of 1 L, seeded, and left at room
temperature overnight. The mixture was briefly sonicated to
losen crystals on the side, resulting in precipitation of ad-
ditional product. The mixture was chilled 4 h in a refrigerator
before filtering, affording 59 (3.16 g, 63%) as fine white
crystals: mp 192 °C dec; 1H NMR (D2O) δ 3.39 (t, J ) 7 Hz, 2
H), 2.48 (t, J ) 7 Hz, 2 H); 13C NMR (D2O) δ 180.1, 158.2,
37.8, 36.9; MS (FAB) m/z 162 (M + H), 323, 316, 296, 238,
163, 162, 161, 147, 118, 89; HRMS (FAB) calcd for C4H11N5O2
+ H 162.0991, found 162.0992. Anal. (C4H11N5O2‚0.43H2O) C,
H, N.
1
0.23 g (66%) of 65 as a white solid: mp 179-180 °C dec; H
NMR (D2O) δ 3.85 (s, 1 H), 1.46 (s, 1 H); 13C NMR (D2O) δ
174.0, 158.8, 157.5, 83.4, 44.5, 27.5; MS (FAB) m/z (rel
intensity) 348 (M + H, 99), 695 (11), 349 (17), 348 (99), 292
(17), 237 (7), 236 (81), 192 (6), 57 (23), 41 (6), 29 (6). Anal.
(C13H25N5O6) C, H; N: calcd, 20.16; found, 19.71.
Meth yl 2-{[Ch lor o(m eth ylsu lfan yl)m eth yliden e]am in o}-
a ceta te (106). To a stirring solution of bis(methylthio)-
methyleneglycine ethyl ester (105) (5.0 g, 25.9 mmol) in
CH2Cl2 (45 mL) at 0 °C was added SO2Cl2 (2.28 mL, 28.84
mmol) in CH2Cl2 (5 mL) dropwise over 20 min. After addition,
the mixture was stirred at 0 °C for 10 min then allowed to
warm to ambient temperature over 1h. The solvent was
removed in vacuo to afford 4.7 g (quant) of the desired product
as a slightly yellow liquid. This material was suitable for use
without further purification: 1H NMR (CDCl3) δ 2.5 (s, 3 H),
3.78 (s, 3 H), 4.37 (s, 2 H); 13C NMR (CDCl3) δ 16.4, 51.8, 54.0,
144.8, 168.8.
2-{[(E)-Hyd r a zin o(1H-p yr a zol-5-yla m in o)m eth ylid en e]-
a m in o}a cetic Acid (74). To a stirring solution of 3-amino-
pyrazole (4.30 g, 51.7 mmol) in CH2Cl2 (60 mL) at 0 °C was
added 106 (9.40 g, 51.7 mmol) in CH2Cl2 (20 mL) dropwise
over 15 min. The resulting green suspension was stirred at 0
°C for 30 min, then allowed to warm to ambient temperature
and stirred 1 h. H2O (100 mL) and CH2Cl2 (50 mL) were added,
the layers shaken, and the aqueous layer separated. The
aqueous layer was neutralized with solid NaHCO3 to pH 7 and
extracted with CH2Cl2 (3 × 75 mL). The organic layers were
combined, dried over Na2SO4 and solvent removed under
reduced pressure to afford 5.40 g (45%) of the desired com-
1
pound as a tan solid: mp 95-97 °C; H NMR (CDCl3) δ 3.74
(s, 3 H), 4.17 (s, 2 H), 6.06 (brs, 1 H), 7.41 (d, J ) 3 Hz, 1 H)
(NH's absent); MS (EI) m/z (rel intensity) 228 (M+, 64), 228
(64), 181 (49), 169 (12), 153 (39), 146 (16), 140 (27), 125 (19),
124 (40), 121 (99), 114 (38). To a stirring suspension of this
solid (5.0 g, 21.9 mmol) in H2O (5 mL) at 0 °C was added
aqueous NaOH (1 M, 21.9 mL) and EtOH (2 mL). The resulting
green solution was treated with hydrazine monohydrate (5.47
g, 109 mmol) and the resulting solution was warmed to room
temperature and stirred overnight. Aqueous HCl (1 M, 21.9
mL) was added and the resulting precipitate cooled to 0 °C
and filtered. The solid was dried in vacuo at 25 °C to afford
1.62 g of 74. Additional material which appeared in the filtrate
(2.0 g) was filtered and found to be identical to the first crop
2-({(Z)-Hydrazino[(2-hydroxyethyl)amino]methylidene}-
a m in o)a cetic Acid (61). A solution of 104a (5.00 g, 27.9
mmol) and ethanolamine (1.68 mL, 27.9 mmol) in water (13
mL) was stirred at room temperature for 24 h. The solution
was diluted with THF to a volume of 250 mL. An oil separated
which solidified with stirring. The mixture was stirred and
sonicated until the solid was finely divided. Filtration gave a
white solid (3.85 g), sufficiently pure to carry into the next
step. Structure analysis showed that both methanethiol groups
has been displaced by ethanolamine to give the cyclic oxazo-
lidine. An analytical sample was obtained by recrystallization
as follows: 0.85 g crude was dissolved in water (20 mL). The
1
(total yield 83%): mp 205-207 °C dec; H NMR (D2O) δ 3.97
(s, 2 H), 6.00 (d, J ) 3 Hz, 1 H), 7.49 (d, J ) 3 Hz, 1 H); MS
(FAB) m/z (rel intensity) 199 (M + H, 99), 397 (2), 353 (4),

1246 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
289 (2), 275 (3), 245 (2), 200 (9), 199 (99), 198 (7), 177 (2), 84
h
at which time the hydrogen pressure was reduced to
(2). Anal. (C6H10N6O2‚H2O) C, H, N.
atmospheric pressure. The reaction vessel was recharged with
H2 (40 psi) and shaken over a period of 17 h. The reaction
mixture was filtered over a plug of Celite (rinsing with wet
ethanol) and concentrated to give an amorphous solid. Puri-
fication via column chromatography (silica gel, 550 g, 15%
MeOH/CH2Cl2 doped with 3% concentrated ammonium hy-
droxide) afforded 5.1 g (61%) of 110 as an amorphous solid.
Analytical material could be obtained via recrystallization with
Et2O/CHCl3 (1:1, 2% methanol) at -20 °C: 1H NMR (300 MHz,
CDCl3) δ 7.53 (br s, 1 H), 3.46-3.37 (m, 1 H), 2.99 (dd, J ) 10,
10 Hz, 1 H), 2.71-2.65 (m, 2 H), 2.46-2.25 (m, 2 H), 1.98-
1.86 (m, 1 H), 1.86-1.74 (m, 3 H), 1.55-1.44 (m, 1 H); 13C NMR
(75 MHz, CDCl3) δ 172.6, 45.1, 44.5, 36.1, 30.1, 24.4; HRMS
calcd for C6H12N2O m/z 128.0905, found m/z 128.0960.
2-[{[2-(2-P yr id in yl)h yd r a zin o]ca r b ot h ioyl}a m in o]-
a cetic Acid (108). To a mixture of 2-hydrazinopyridine (4.6
g, 24.2 mmol) in absolute EtOH (25 mL) at 0 °C was added
ethyl isothiocyanatoacetate (3.0 mL, 23.0 mmol), and the
mixture was stirred at 0 °C for approximately 20 min. A thick
suspension had formed which made stirring difficult. The ice
bath was removed, and absolute EtOH was added until stirring
resumed. The mixture was then stirred at room temperature
for 30 min, cooled to 0 °C, and filtered. The solid was washed
with ice cold absolute EtOH and dried (house vacuum, 50 °C)
to yield 4.5 g (77%) the desired thiosemicarbazide as a white
1
solid: mp 149-151 °C dec; H NMR (DMSO) δ 9.63 (s, 1 H),
8.50 (s, 1 H), 8.40 (m, 1 H), 8.12 (d, J ) 4 Hz, 1 H), 7.62 (t, 1
H), 6.81 (t, 1 H), 6.57 (d, J ) 8 Hz, 1 H), 4.18 (d, J ) 6 Hz, 2
H), 4.08 (q, J ) 7 Hz, 2 H), 1.18 (t, J ) 7 Hz, 3 H); 13C NMR
(DMSO) δ 185, 172, 161, 148.5, 148, 140.0, 139, 117.7, 117.6,
109, 108.7, 62.3, 57, 46.4, 18, 14.5; MS (EI) m/z (rel intensity)
254 (M+, 26), 254 (26), 221 (99), 152 (20), 151 (93), 147 (29),
109 (36), 94 (50), 93 (17), 79 (41), 67 (21); HRMS (EI) calcd for
5-Am in o-4-(am in om eth yl)pen tan oic Acid Dih ydr och lo-
r id e (78). To a stirring solution of 110 (1.57 g, 12.3 mmol)
and water (24 mL) was added concentrated hydrochloric acid
(12 M, HCl, 4.0 mL, 50 mmol), and the solution was heated to
reflux over a period of 18 h and concentated to give an
amorphous solid. The crude solid was recrystallized by diss-
ovling in hot water/ethanol (1:4, 25 mL) and cooling to -20
°C over a period of 48 h. 78 was collected by filtration to give
C
10H14N4O2S 254.0837, found 254.0548. Anal. (C10H14N4O2S‚
0.21H2O) C, H, N, S.
1
1.43 g (53%) as a white solid: mp 183-184 °C; H NMR (300
A mixture of this solid (4.5 g, 17.7 mmol) in 1.0 M NaOH
(35 mL) was stirred at room temperature for 1.5 h. 1.0 M HCl
(35 mL) was added, and a solid immediately precipitated. The
solid was collected by filtration, and dried (house vacuum 50
°C, 18 h) to provide 3.9 g (98%) of 108 as a tan solid: mp 198-
199 °C dec; 1H NMR (DMSO) δ 9.55 (s, 1 H), 8.43 (s, 1 H),
8.10 (m, 1 H), 8.10 (d, 1 H), 7.59 (t, 1 H), 6.78 (t, 1 H), 6.56 (d,
J ) 6 Hz, 1 H), 4.11 (d, J ) 4 Hz, 2 H); 13C NMR (DMSO) δ
183.2, 171.4, 159.8, 148.1, 138.2, 116.1, 107.4, 45.5; MS (EI)
m/z (rel intensity) 226 (M+, 13), 208 (81), 193 (72), 175 (59),
151 (99), 136 (79), 109 (33), 94 (58), 93 (70), 79 (69), 78 (35);
HRMS (FAB) calcd for C8H10N4O2S + H 227.0603, found
227.0600. Anal. (C8H10N4O2S‚0.21H2O) C, H, N, S.
2-({(E)-Hyd r a zon o[2-(2-p yr id in yl)h yd r a zin o]m eth yl}-
a m in o)a cetic Acid (76). To a mixture of 108 (3.65 g, 16.1
mmol) and triethylamine (4.5 mL, 32.2 mmol) in MeOH (15
mL) at room temperature was added methyl iodide (3.0 mL,
48.3 mmol). After approximately 3 min, an exothermic reaction
was observed, and a yellow solid precipitated. The mixture was
stirred at room temperature for 3 h, and then cooled to 0 °C.
The solid was collected by filtration, washed with cold MeOH,
and dried (house vacuum, 50 °C, 18 h) to give 2.8 g (72%) of
the desired product (2:1 mixture of rotational isomers) as a
bright yellow solid: mp 149-151 °C; 1H NMR (DMSO + DCl)
δ 8.00 (m, 2 H), 7.25 (m, 1 H), 7.00 (m, 1 H), 4.28, 4.15 (s, 2
H), 2.72, 2.55 (s, 3 H); 13C NMR (DMSO + DCl) δ 170.9, 151.3,
144.4, 140.3, 136.8, 115.1, 112.9, 112.4, 45.6, 45.2, 14.5, 14.1;
MS (FAB) m/z (rel intensity) 241 (M + H, 99), 317 (3), 243 (5),
242 (13), 241 (99), 240 (7), 193 (8), 166 (6), 116 (4), 95 (6), 94
(4); HRMS (FAB) calcd for C9H12N4O2S + H 241.0759, found
241.0766.
To a suspension of this solid (2.6 g, 10.9 mmol) in absolute
EtOH (11 mL) at room temperature was added hydrazine
monohydrate (1.1 mL, 21.7 mmol). After stirring for ap-
proximately 2 h, a homogeneous solution formed. After stirring
for approximately an additional 1 h (3 h total), a solid
precipitated, and the mixture was cooled to 0 °C. The solid
was isolated by filtration, and washed with cold absolute EtOH
to yield 2.0 g (82%) of 76 as a white solid: mp 187-188 °C; 1H
NMR (D2O) δ 8.06 (d, 1 H), 7.70 (3, 1 H), 6.95 (3, 1 H), 6.85 (d,
1 H), 3.70 (br s, 2 H); MS (EI) m/z (rel intensity) 206 (39), 148
(12), 147 (99), 94 (44), 93 (10), 79 (22), 78 (38), 67 (19), 66 (9),
51 (10); MS (FAB) m/z (rel intensity) 225 (M + H, 99), 449 (7),
377 (3), 301 (15), 226 (11), 225 (99), 224 (27), 180 (4), 108 (4),
95 (11), 94 (6); HRMS (FAB) calcd for C8H12N6O2 + H 225.1100,
found 225.1098. Anal. (C8H12N6O2‚0.4H2O) C, H, N.
MHz, D2O) δ 3.19-2.99 (m, 4 H), 2.54 (t, J ) 7 Hz, 2 H), 2.21
(heptet, J ) 6 Hz, 1 H), 1.82 (q, J ) 6 Hz, 2 H); 13C NMR (75
MHz, D2O) δ 177.2, 40.1, 34.4, 29.9, 22.9; MS (FAB) m/z (rel
intensity) 147 (M + H, 99), 247 (3), 245 (7), 148 (8), 147 (99),
130 (3), 129 (6), 118 (5), 116 (3), 45 (3), 30 (5). Anal.
(C6H14N2O2‚2HCl) C, H, N.
2-[(4-Am in o-1,2,4-tr iazin an -3-yliden e)am in o]acetic Acid
Mon oh yd r a te (79). To a stirring solution of 104a (11.4 g, 37.0
mmol) in water (300 mL) was added a solution of ethylene bis-
(hydrazine)27 (3.3 g, 37.0 mmol) and aqueous sodium hydroxide
(1 M NaOH, 37 mL) via pipet. The reaction mixture was
allowed to stir at ambient temperature over a period of 22 h
and concentrated to give a tan solid. The crude solid was
dissolved in hot ethanol/water (120 mL, 10 mL) and filtered
to remove insoluble material and concentrated to give 3 g of a
viscous oil. Purification via column chromatography (silica gel,
500 g, 7:2:1 MeOH/CH2Cl2/NH4OH(conc)) gave 2.0 g (31%) of
79 ∼95% purity judged by 1H NMR. Analytically pure material
could be obtained by dissolving in hot ethanol and triturating
with diethyl ether at 0 °C: white solid; mp 130-132 °C
(sublimed); 1H NMR (300 MHz, D2O) δ 3.74 (s, 2 H), 3.57 (t, J
) 6 Hz, 2 H), 3.21 (t, J ) 6 Hz, 2 H); 13C NMR (75 MHz, D2O)
δ 175.7, 155.4, 50.6, 44.4, 42.4. Anal. (C5H11N5O‚H2O) C, H,
N.
2-{[(2-Acetylh yd r a zin o)(m eth ylsu lfa n yl)m eth ylid en e]-
a m in o}a cetic Acid (111 (R ) C(O)CH3). A solution of 104a
(1.00 g, 5.58 mmol) and acetic hydrazide (0.62 g, 8.4 mmol) in
2-propanol (5.6 mL) was stirred at room temperature for 1 h.
Cold water bath cooling was used for the first 5 min. The
precipitate was collected by filtration, affording 111 (R ) C(O)-
1
CH3) as a white solid (1.04 g, 91%): mp 118 °C dec; H NMR
(D2O) δ 4.00 (bs, 2 H), 2.68, 2.48 (2 bs, 3 H), 2.11 (bs, 3 H); 13
C
NMR (D2O) δ 173.9, 171.5, 47.3, 20.3 (broad), 13.3 (broad), one
peak missing; MS (FAB) m/z (rel intensity) 206 (M + H, 99),
296 (6), 260 (6), 216 (6), 207 (9), 206 (99), 205 (7), 149 (7), 132
(6), 109 (6), 45 (7); HRMS (FAB) calcd for C6H11N3O3S + H
206.0599, found 206.0605. Anal. (C6H11N3O3S) C, H, N.
2-{[6-Meth yl-1,4-dih ydr o-1,2,4,5-tetr aazin -3(2H)-yliden e]-
a m in o}a cetic Acid (80). To a mixture of 111 (R ) C(O)CH3)
(7.53 g, 36.7 mmol) in water (37 mL), cooled in a water bath,
was added hydrazine monohydrate (3.56 mL, 73.5 mmol). The
resulting solution was stirred at room temperature for 24 h.
The solution was diluted with 2-propanol to a volume of 500
mL. An oil separated which was induced to solidify by
sonication and seeding with authentic product. After standing
at 0 °C overnight, filtration afforded 80 as a white solid (1.74
5-(Am in om eth yl)-2-p ip er id in on e (110). In a hydrogena-
tion vessel was placed bisnitrile 10926 (11 g, 66 mmol) in
ethanol (absolute, 600 mL) and Raney nickel (50% aqueous
emulsion, 50 g). The heterogeneous solution was hydrogenated
at 40 psi H2 while being shaken vigorously over a period of 6
1
g, 28%): mp >280 °C; H NMR (D2O) δ 3.92 (s, 2 H), 2.39 (s,
3 H); 13C NMR (D2O) 175.7, 152.3, 151.1, 45.7, 9.1; MS (FAB)
m/z (rel intensity) 172 (M + H, 99), 344 (2), 343 (8), 326 (4),
280 (1), 248 (4), 173 (9), 172 (99), 127 (2), 126 (5), 111 (1);

(Di)aminoguanidine Analogues of Guanidinopropionic Acid
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8 1247
HRMS (FAB) calcd for C5H9N5O2 + H 172.0835, found
glucose are indicated with asterisks as defined in the tables.
For comparison, administration of troglitazone (Rezulin) in this
assay at a dose of 200 mg/kg gives a T/C of 0.78, and
administration of pioglitazone43 at a dose of 100 mg/kg gives
a T/C of 0.49. Metformin (Glucophage) is not active in the assay
when administered at 500 mg/kg.
172.0832. Anal. (C5H9N5O2) C, H, N.
2-{[(1-Meth ylh ydr azin o)car both ioyl]am in o}acetic Acid
(113). To a mixture of methyl hydrazine (0.86 mL, 16 mmol)
in absolute EtOH (15 mL) at 0 °C was added ethyl isothiocy-
anatoacetate (112) (2 mL, 16 mmol) dropwise over 5 min. The
mixture was stirred at 0 °C for an additional 10 min, warmed
to room temperature, and stirred for 0.5 h. After the mixture
was cooled back to 0 °C, the white solid was isolated by
filtration, washed with cold EtOH, and dried (house vacuum,
50 °C, 18 h) to yield 2.1 g (68%) of the desired thiosemicar-
bazide as a white solid: mp 121-123 °C; 1H NMR (CD3OD) δ
4.28 (s, 2 H), 4.18 (q, J ) 7 Hz, 2 H), 3.55 (s, 3 H), 1.27 (t, J
) 7 Hz, 3 H); 13C NMR (CD3OD) δ 183.2, 172.0, 62.0, 47.2,
43.3, 14.3; MS (FAB) m/z (rel intensity) 192 (M + H, 99), 383
(5), 300 (5), 268 (25), 194 (5), 193 (8), 192 (99), 191 (11), 160
(6), 89 (11), 46 (11); HRMS (FAB) calcd for C6H13N3O2S + H
192.0807, found 192.0805. Anal. (C6H13N3O2S) C, H, N.
Dose-Resp on se Stu d ies in KKAy Mice. Experiments
were run as described above. The difference between the pre-
and post-treatment blood glucose values was computed for
individual animals. These delta values were used to compare
treatment groups using a one-way analysis of variance.
Treatment groups were compared to the control group using
a t-test where the estimate of the standard error comes from
a pooled estimate using all the treatment groups. The food
difference data for each treatment group was compared in the
same fashion. All calculations were done using the SAS
computer package.36
Bod y Com p osition An a lysis. Diabetic, obese KKAy mice
of either sex, aged 12-16 weeks, were obtained from Clea
Nippon (Osaka, J apan37). Diabetic, obese female C57BL6J ob/
ob mice, aged 10-14 weeks, were obtained from J ackson
Laboratories (Bar Harbor, ME38). Mice were housed in the
Pharmacia Kalamazoo animal care facility at 21 ( 1 °C and
kept on a 12 h light-dark cycle. Fresh water and rodent chow
were available at all times. Diabetic mice received a diet of
Purina 5015 chow (26% of calories from fat; 4.35 kcal/g) and
normal mice and rats received Purina 5002 rodent chow (11.6%
of calories from fat; 4.06 kcal/g). Test compounds were
administered as an admixture in milled Purina chow for 14
days. Drug intake was determined by measuring the difference
in the weight of the food bowl before and after feeding. Blood
samples were obtained by puncture of the retro-orbital sinus.
Heparinized blood was stored on ice and promptly analyzed
or centrifuged for preparation of plasma. Plasma samples were
stored at -80 °C until assayed. Body composition was assessed
by injecting 3HOH intraperitoneally using the method of Pace
et al.39 The lean and fat body mass was determined based on
the difference in the 3HOH space of the two types of tissue.
The 3HOH space was converted to lean body mass according
to eq 1. Fat body mass was calculated according to eq 2:
To a portion of this solid (1.7 g, 8.8 mmol) was added 1.0 M
NaOH (18 mL) and the mixture was stirred for 1 h. The solid
dissolved after 10-15 min of stirring to produce a homoge-
neous solution. 1.0 M HCl (18 mL) was added, and the volume
was reduced in vacuo until a solid formed. The mixture was
cooled to 0 °C and filtered to yield 1.06 g (74%) of 113 as a
1
white solid: mp 151-153 °C; H NMR (CD3OD) δ 4.27 (s, 2
H), 3.55 (s, 3 H); 13C NMR (CD3OD) δ 183.2, 173.8, 47.2, 43.5;
MS (FAB) m/z (rel intensity) 164 (M + H, 99), 374 (8), 316 (6),
298 (21), 272 (11), 240 (39), 174 (6), 165 (6), 164 (99), 163 (18),
46 (6); HRMS (FAB) calcd for C4H9N3O2S + H 164.0494, found
164.0504. Anal. (C4H9N3O2S) C, H, N, S.
3-Hyd r a zin o-2-m eth yl-2,5-d ih yd r o-1,2,4-tr ia zin -6(1H)-
on e (82). To a suspension of 113 (0.40 g, 2.45 mmol) in MeOH
(2.5 mL) was added methyl iodide (0.46 mL, 7.40 mmol). The
mixture was heated at reflux for 1 h, then cooled to 0 °C. The
solid was isolated by filtration and washed with cold MeOH
to give 0.23 g of a white solid. To the filtrate was added Et2O
(30 mL), and the mixture was cooled to 0 °C. Filtration
provided a second crop 0.38 g of a white solid. The two crops
were combined to yield 0.61 g (86%) of a white solid: mp 188-
1
189 °C dec; H NMR (D2O) δ 3.89 (s, 2 H), 3.37 (s, 3 H), 2.47
(s, 3 H); MS (FAB) m/z (rel intensity) 160 (M + H, 99), 447
(3), 319 (7), 236 (2), 162 (7), 161 (7), 160 (99), 159 (4), 146 (1),
114 (1), 88 (4); HRMS (FAB) calcd for C5H9N3OS + H 160.0545,
found 160.0547. Anal. (C5H9N3OS‚HI) C, H, N, S.
LBM ) (3HOH space)/0.729
where LBM ) lean body mass and
(1)
To a suspension of the solid from above (7.42 g, 25.8 mmol)
in absolute EtOH (25 mL) was added 12.5 M NaOH (1.94 mL).
Hydrazine monohydrate (2.4 mL, 48.6 mmol) was added to the
mixture, and the mixture was stirred at room temperature for
1 h. A solid precipitated producing a thick suspension, and
absolute EtOH (30 mL) was added. The mixture was stirred
for an additional 1 h and filtered. The isolated solid was
washed with cold absolute EtOH and dried to give 3.1 g of a
white solid. A second crop (1.3 g) was realized by allowing the
filtrate to stand at room-temperature overnight. The two crops
were combined (4.4 g), and recrystallized (EtOH/H2O) to
provide 2.2 g (52%) of 82 (monohydrate) as a white solid: mp
FBM ) BW - LBM
(2)
where FBM ) fat body mass and BW ) body weight.
Stu d ies in Non h u m a n P r im a tes. Adult rhesus macaques
(Macaca mulatta) were maintained on a nutritionally adequate
diet of PMI Certified Primate bisquits (St. Louis, MO) and a
piece of fruit daily. The animals were all surgically ovarecto-
mized females and varied in age from 15-30 years. These
monkeys were selected because their fasting plasma insulin
and glucose concentrations were elevated relative to the mean
value for normal rhesus monkeys in our colony (20 µU/mL and
59 mg/dL, respectively; n ) 14). Oral dosing was performed
by dissolving the test substance in diet orange soda and
allowing the animals to voluntarily drink the solution from
the tip of a syringe. Dosing was performed once daily at
approximately 8:00 a.m. Meal tolerance tests were performed
in conscious monkeys that had been fasted overnight. A mixed
liquid meal was administered via an orogastric tube to start
the test. The composition of the meal was carbohydrate, 5.21
kcal/kg body weight; protein, 1.12 kcal/kg; fat, 2.34 kcal/kg.
Blood samples were obtained by percutaneous puncture of the
femoral vein at t ) 0, 30, 60, 90, and 120 min.
1
189-191 °C; H NMR (D2O) δ 3.52 (s, 2 H), 2.92 (s, 3 H); 13C
NMR (D2O) δ 165.3, 151.8, 40.8, 36.4; MS (EI) m/z (rel
intensity) 143 (M+, 99), 100 (24), 99 (20), 70 (26), 69 (20), 58
(21), 57 (33), 55 (38), 46 (62), 45 (30); HRMS (EI) calcd for
C4H9N5O 143.0807, found 143.0813. Anal. (C4H9N5O‚H2O) C,
H, N.
Biologica l P r oced u r es. In Vivo An tih yp er glycem ic
Activity in KKAy Mice. New analogues were evaluated in
obese hyperglycemic, hyperinsulinemic, insulin-resistant KKAy
mice as previously described.29 Animals were grouped into
treatment and control groups (n ) 6) following pretest blood
glucose measurements with an Alpkem glucose autoanalyzer.
Treatment groups had the selected compound administered
as a food mixture at 500 mg/kg for 4 days. For those
compounds with T/C < 0.80, the control values were statisti-
cally compared to the treatment values using the nonpara-
metric Wilcoxon rank sum test. Those compounds determined
to effect statistically significant reductions in nonfasting blood
Glucose was assayed using either a YSI Biochemical Ana-
lyzer or an Alpkem Autoanalyzer. Insulin concentration was
determined with a double-antibody radioimmunoassay tech-
nique using guinea pig antihuman insulin serum and 125I-
porcine insulin as the tracer (DuPont-NEN, Wilmington, DE).
For assays of monkey samples, human insulin was used as
the standard (Linco, St. Louis, MO).

1248 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 8
Vaillancourt et al.
(17) Kawashiro, K.; Morimoto, S.; Yoshida, H. The Ion-exchange
Chromatography of Imino Derivatives of Glycine. Bull. Chem.
Soc. J pn. 1983, 56, 792.
(18) Walker, G. N.; Alkalay, D.; Engle, A. R.; Kempton, R. J .
Synthesis of 5-Oxo-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-
10-carboxylic Acids, Corresponding Nitriles, and Related Bridged
Lactones, Hemiketals, Lactams, Amines, Amidoximines and
Amidines (5,10-Epoxymethano and 5, 10-Iminomethano Com-
pounds). J . Org. Chem. 1971, 36, 466.
(19) Borch, R. F.; Bernstein, M. D.; Durst, H. D. Cyanohydridoborate
Anion as as Selective Reducing Agent. J . Am. Chem. Soc. 1971,
93, 2897.
(20) Lee, J .; Paudler, W. W. Triazine Chemistry. VIII. 2,5-Dihydro-
5-oxo-1,2,4-triazines J . Heterocycl. Chem. 1972, 9, 995.
(21) Lecoq, A.; Marraud, M.; Aubry, A. Hydrazino and N-amino
Peptides. Chemical and Structural Aspects. Tetrahedron Lett.
1991, 32, 2765.
Statistical analysis was performed using SigmaStat Soft-
ware (J andel, San Rafael, CA). In studies that used mice, two-
way ANOVA was used. Post hoc comparison of sample means
was performed using the Student-Newman-Keuls procedure.
In the studies that used monkeys, the sample sizes were small
and animals were often heterogeneous with respect to the
measured parameters. To compensate for the high inter-
subject variability, the results were analyzed using a multiway
ANOVA design. Three-way ANOVA was used in order to
partition the sources of variation according to subject, treat-
ment, and time. With this approach, the variation due to inter-
subject differences was partitioned to allow concentration on
the effects attributable to the treatment and time. Post hoc
comparison of sample means was performed using the Tukey
Test.
(22) Berges, D. A.; Chan, G. W.; Polansky, T. J .; Taggart, J . J .; Dunn,
G. L. 4,5-Dihydro-5-thioxo-1H-tetrazole-1-alkanoic and Alkane-
sulfonec Acids and Their Amide Derivatives. J . Heterocycl.
Chem. 1978, 15, 981.
In Vitr o Eva lu a tion of Cr ea tin e Up ta k e/Kin a se. These
assays were run as described in the previous manuscript.
Results are reported in Tables 10 and 11.
(23) Hoppe, D.; Edgar, R. Studies in â-Lactam Synthesis. IV.
Synthesis of 3-[Bis(alkylthio)methyleneamino]-2-azetidinones by
Acid Chloride-Imine Reaction. Liebigs Ann. Chem. 1979, 2076.
(24) Saulnier, M. G.; Frennesson, D. B.; Deshpande, M. S.; Hansel,
S. B.; Vyas, D. M. An Efficient Method for the Synthesis of
Guanidino Prodrugs. Bioorg. Med. Chem. Lett. 1994, 4, 1985.
(25) Singh, S. P.; Savita; Kumar, D. Reaction of 2-Hydrazinobenz-
imidazole with â-Diketones: A Structural Reinvestigation.
Indian J . Chem. Sec. B 1993, 262.
Su p p or tin g In for m a tion Ava ila ble: Numerical tabular
data to support Figures 1-3 and experimental procedures for
analogues 3, 5, 6, 8, 14, 15, 17, 19, 21-27, 29, 30, 34-37, 39,
40, 42, 44, 46-51, 55-58, 60, 62-64, 66-73, 75, 77, and 81.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(26) Schoffstall, A. M. Synthesis of 5,6-Dihydropyrido[2,3-d]pyrimi-
dine Derivatives Directly From Acyclic Precursors. J . Org. Chem.
1971, 36, 2385.
Refer en ces
(1) Meglasson, M. D. Use of 3-Guanidinopropionic Acid in the
Treatment of Excess Adiposity. U.S. Patent 5,134,164, J uly 28,
1992.
(2) Meglasson, M. D.; Wilson, J . M.; Yu, J . H.; Robinson, D. D.; Wyse,
B. M.; deSouza, C. J . Antihyperglycemic Action of Guanidino-
alkanoic Acids: 3-Guanidinopropionic Acid Ameliorates Hyper-
glycemica in Diabetic KKAy and C57BL6J ob/ ob Mice and
Increases Glucose Disappearance in Rhesus Monkeys. J . Phar-
macol. Exp. Ther. 1993, 266, 1454-62.
(3) De Tata, V.; Cavallini, G.; Pollera, M.; Gori, Z.; Bergamini, E.
The Induction of Mitochondrial Mypopathy in the Rat by Feeding
â-Guanidinopropionic Acid and the Reversibility of the Induced
Mitochondrial Lesions: A Biochemical and Ultrastructural
Invesitigation. Int. J . Exp. Pathol. 1993, 74, 501-9 and refs cited
therein.
(4) Shirai, K.; Kumamoto, T.; Kobayashi, Y, Ri, T. 3-Guanidinoami-
nopropionic acid. J P Patent 54128523, Oct 5, 1979.
(5) Gante, J . Uber die Synthese Peptidanaloger Aminoguanidine.
Chem. Ber. 1968, 101, 1195.
(6) Gut, J .; Hesoun, D.; Novacek, A. Nucleic Acids Components and
Their Analogues. LXXXIII. Synthesis of 3-Amino-2,5-dihydro-
1,2,4-triazin-5-one (Azaisocytosine) and its N-Substituted De-
rivatives. Collect. Czech. Chem. Commun. 1966, 31, 2014.
(7) Carmi, A.; Pollak, G.; Yellin, H. R-Hydrazino-Acids. I. R-Hy-
drazinoaliphatic Acids and R-(1-Methylhydrazino)alaphatic Ac-
ids. J . Org. Chem. 1960, 25, 44.
(8) Freund, M.; Paradies, T. Zur Kenntniss des Tetrazols. Chem.
Ber. 1901, 34, 3110.
(9) Gaspert, B.; Skaric, V. Hydropyrimidines. I. Synthesis of 5,6-
Dihydro-2-thiouracil. Croat. Chem. Acta. 1963, 35, 171; Chem.
Abstr. 60:523h.
(27) (a) Oja, P. D. Preparation of 1,2-Dihydrazinoethane. U.S. Patent
3,317,608. (b) Deaniker, H. U.; Dreuy, J . Heilmittelchemische
Studien in der Heterocyclischen Reihe. Uber Polymethylen-bis-
sydnone und Polymethylen-bis-hydrazine. Helv. Chim. Acta
1957, 40, 918. (c) Evans, R. F.; J ones, J . I.; Kynaston, W.
Polymers Based on Hydrazine. Part III*. Reaction of Hydrazine
Hydrate with Ethylene Dihalides. J . Chem. Soc. 1963, 4023.
(28) (a) Gante, J .; Lautsch, W. Neuartige Thiosemicarbazino-(4)-
essigsaure-Derivate. Chem. Ber. 1964, 97, 989. (b) Klayman, D.
L.; Bartosevich, J . F.; Griffin, T. S.; Mason, C. J .; Scohill, J . P.
2-Acetylpyridine Thiosemicarbazones. 1. A New Class of Poten-
tial Antimalarial Agents. J . Med. Chem. 1979, 22, 855.
(29) Tanis, S. P.; Parker, T. T.; Colca, J . R.; Fisher, R. M.; Kletzien,
R. F. Synthesis and Biological Activity of Metabolites of the
Antidiabetic, Antihyperglycemic Agent Pioglitazone. J . Med.
Chem. 1996, 39, 5053. KKAy mice of either sex, aged 12-16
weeks, were obtained from Clea Nippon (Osaka, J apan).
(30) (a) Fitch, C. D.; Shields, R. P.; Payne, W. F.; Dacus, J . Creatine
Metabolism in Skeletal Muscle. III. Specificity of the Creatine
Entry Process. J . Biol. Chem. 1968, 243, 2024. (b) Moeller, A.;
Hamprecht, B. Creatine Transport in Cultured Cells of Rat and
Mouse Brain. J . Neurochem. 1989, 52, 544.
(31) Shoubridge, E. A.; Radda, G. K. A Phosphorus-31 Nuclear
Magnetic Resonance Study of Skeletal Muscle Metabolism in
Rats Depleted of Creatine with the Analogue â-Guanidinopro-
pionic Acid. Biochim. Biophys. Acta 1984, 805, 79.
(32) Chevli, R.; Fitch, C. D. â-Guanidinopropionate and Phosphory-
lated â-Guanidinopropionate as Substrates for Creatine Kinase.
Biochem. Med. 1979, 21, 162.
(33) Unitt, J . F.; Radda, G. K.; Seymour, A. M. The Acute Effects of
the Creatine Analogue, â-Guanidinopropionic Acid, on Cardiac
Energy Metabolism and Function. Biochim. Biophys. Acta 1993,
1143, 91.
(10) Miura, K.; Ikeda, M.; Kondo, t.; Setogawa, K. 3-Amino-as-triazine
Derivatives. Kanazawa Daigaku Yakugakuba Kenkyu Nempo
1961, 11, 25-9; Chem. Abstr. 1962, 56, 4767b.
(34) Zhang, Y.; Proenca, R.; Maffei, M.; Barone, M.; Leopold, L.;
Friedman, J . M. Positional Cloning of the Mouse Obese Gene
and Its Human Homologue. Nature 1994, 372, 425.
(35) Bradbury, A. F.; Smyth, D. G. Enzyme-catalyzed Peptide Ami-
dation. Isonation of a Stable Intermediate Formed by Reaction
of the Amidating Enzyme With an Amino Acid. Eur. J . Biochem.
1987, 169, 579.
(11) Pankaskie, M.; Abdel-Monem, M. M. Inhibitors of Polyamine
Biosynthesis. VII: Evaluation of Pyruvate Derivatives as Inhibi-
tors of S-Adenosyl-L-methionine Decarboxylase. J . Pharm. Sci.
1980, 69, 1000.
(12) Kirsten, G. W.; Smith, G. B. L. Reduction of Nitroguanidine. V.
The Synthesis of (a) R-Methyl-, (b) R-Ethyl-, (c) R-n-Butyl-γ-
aminoguanidine. J . Am. Chem. Soc. 1936, 58, 800.
(13) Phillips, R.; Williams, J . F. Nitro-aminoguanidine. J . Am. Chem.
Soc. 1928, 50, 2465.
(14) Ellingsfeld, H.; Scheuermann, H. Synthesis of 1,3,5-Triazines.
Chem. Ber. 1967, 100, 1874.
(15) Amato, J . S.; Karady, S.; Reamer, R. A.; Schlegel, H. B.; Springer,
J . P.; Weinstock, L. M. 1,2,5-Thiadiazole-1-oxides. An Experi-
mental and Theoretical Investigation of the Inversion Barrier.
J . Am. Chem. Soc. 1982, 104, 1375.
(36) SAS/ STAT User’s Guide, Version 6, 4th ed.; SAS Institute
Inc.: Cary, NC, 1989; pp 891-996.
(37) Chang, A. Y.; Wyse, B. M.; Copeland, E. J .; Peterson, T.;
Ledbetter, S. R. The Upjohn Colony of KKAy Mice: A Model for
Obese Type II Diabetes. In Diabetes 1985; Serrano-Rios, M.,
Lefebvre, P. J ., Eds.; Elsevier Publishing: New York, 1986.
(38) Coleman, D. L.; Diabetes-obesity Syndromes in Mice. Diabetes
1982, 31, 1.
(16) (a) Korte, A.; Hartke, K. Dithio and Thiono Esters. 53. Synthesis
of N-Protected R-Amino Dithio Esters from R-Amino Nitriles and
R-Amino Acids. Liebigs Ann. Chem. 1992, 595. (b) Bailly, C.;
Houssin, R.; Bernier, J .-L.; Henichart, J .-P. Depsipeptide Ana-
logues of the Anittumor Drug Distamycin Containing Thiazole
Amino Acid Residues. Tetrahedron 1988, 44, 5833.
(39) Pace, N.; Rahlmann, D. F.; Smith, A. H. Methodology for
estimation of total body composition in laboratory mammals.
NASA [Contract. Rep.] CR 1979, NASA-CR-162806, EPL-79-1,
44 pp.
J M000094N