Discovery of a series of phosphonic acid-containing thiazoles and orally bioavailable diamide prodrugs that lower glucose in diabetic animals through inhibition of fructose-1,6-bisphosphatase

Article abstract of DOI:10.1021/jm101035x

Oral delivery of previously disclosed purine and benzimidazole fructose-1,6-bisphosphatase (FBPase) inhibitors via prodrugs failed, which was likely due to their high molecular weight (>600). Therefore, a smaller scaffold was desired, and a series of phosphonic acid-containing thiazoles, which exhibited high potency against human liver FBPase (IC₅₀ of 10-30 nM) and high selectivity relative to other 5′-adenosinemonophosphate (AMP)-binding enzymes, were discovered using a structure-guided drug design approach. The initial lead compound (30j) produced profound glucose lowering in rodent models of type 2 diabetes mellitus (T2DM) after parenteral administration. Various phosphonate prodrugs were explored without success, until a novel phosphonic diamide prodrug approach was implemented, which delivered compound 30j with good oral bioavailability (OBAV) (22-47%). Extensive lead optimization of both the thiazole FBPase inhibitors and their prodrugs culminated in the discovery of compound 35n (MB06322) as the first oral FBPase inhibitor advancing to human clinical trials as a potential treatment for T2DM.

Full text of DOI:10.1021/jm101035x

J. Med. Chem. 2011, 54, 153–165 153
DOI: 10.1021/jm101035x
Discovery of a Series of Phosphonic Acid-Containing Thiazoles and Orally Bioavailable Diamide
Prodrugs That Lower Glucose in Diabetic Animals Through Inhibition of Fructose-1,6-Bisphosphatase
Qun Dang,*,† Yan Liu, Daniel K. Cashion,‡ Srinivas Rao Kasibhatla,§ Tao Jiang, Frank Taplin,
Jason D. Jacintho, Haiqing Li, Zhili Sun, Yi Fan, Jay DaRe,^ Feng Tian, Wenyu Li,# Tony Gibson,3
Robert Lemus, Paul D. van Poelje,O Scott C. Potter, and Mark D. Erion†
Departments of Medicinal Chemistry and Biochemistry, Metabasis Therapeutics, Inc., 11119 North Torrey Pines Road, La Jolla,
California 92037, United States. †Current address: Merck Research Laboratories, Rahway, New Jersey 07065.
‡ Current address: Siemens Medical Solutions, Culver City, California 90230. § Current address: Jubilant Biosys, Bangalore, India.
Current address: Genomics Institute of the Novartis Research Foundation, San Diego, California 92121. ^ Current address:
Neurocrine Biosciences, Inc., San Diego, California 92130. # Current address: ISIS Pharmaceuticals, Inc., Carlsbad, California 92008.
3 Current address: Takeda San Diego Inc., San Diego, California 92121. O Current address: Pfizer, Inc., Groton, Connecticut 06340.
Received August 9, 2010
Oral delivery of previously disclosed purine and benzimidazole fructose-1,6-bisphosphatase (FBPase)
inhibitors via prodrugs failed, which was likely due to their high molecular weight (>600). Therefore, a
smaller scaffold was desired, and a series of phosphonic acid-containing thiazoles, which exhibited high
potency against human liver FBPase (IC50 of 10-30 nM) and high selectivity relative to other 50-
adenosinemonophosphate (AMP)-binding enzymes, were discovered using a structure-guided drug
design approach. The initial lead compound (30j) produced profound glucose lowering in rodent models
of type 2 diabetes mellitus (T2DM) after parenteral administration. Various phosphonate prodrugs
were explored without success, until a novel phosphonic diamide prodrug approach was implemented,
which delivered compound 30j with good oral bioavailability (OBAV) (22-47%). Extensive lead opti-
mization of both the thiazole FBPase inhibitors and their prodrugs culminated in the discovery of com-
pound 35n (MB06322) as the first oral FBPase inhibitor advancing to human clinical trials as a potential
treatment for T2DM.
Introduction
negative charged phosphonic acid. To improve OBAV, lipo-
philic phosphonic acid prodrugs were prepared to mask the
negative charge, however, despite extensive efforts exploring
both known and novel prodrugs, we were unable to achieve
OBAV >10% in rodents or dogs. For example, benzimid-
azole 2 is a potent FBPase inhibitor with rapid in vivo glucose
lowering activity after intravenous administration, but var-
ious prodrugs including the diamide 3 were explored without
successfully achieving acceptable OBAV (>10%).
The factors limiting OBAV were attributed to oral absorp-
tion because both series appeared to be metabolically stable.
Most likely the high molecular weight (MW) of prodrugs for
both series (>600) was a key limitation,8 and as such we
eventually were forced to abandon those series and focus on
designing a new scaffold with a significantly lower MW while
retaining high FBPase potency and selectivity. Herein we report
the design, synthesis, and evaluation of thiazole phosphonic
acids as novel FBPase inhibitors. Replacement of the 5,6-fused
purine and benzimidazole ring systems with a smaller 5-mem-
bered thiazole of lower MW led to significantly improved FBPase
inhibitory potency. More importantly, the thiazole FBPase
inhibitors proved to be excellent candidates for oral delivery
via prodrugs, which led to the discovery of an oral FBPase
inhibitor, compound 35n (MB06322 or CS-917)3 as the first
oral FBPase inhibitor to advance to human clinical trials as a
potential treatment for T2DM.
Fructose-1,6-bisphosphatase (FBPasea) catalyzes a rate-
controlling step in the gluconeogenesis (GNG) pathway, thereby
representing a potential target for controlling the excessive
glucose production by the liver in patients with type 2 diabetes
mellitus (T2DM).1 While recognized as a potential target for
almost four decades, FBPase has proven to be an extraordi-
narily challenging target in which to find potent, selective, and
orally efficacious drug candidates.2 Using a structure-based
drug design approach, we ultimately were able to find a highly
promising series of phosphonicacids thatbind to the allosteric
AMP binding site of FBPase and act as an AMP mimetic
capable of inducing a protein conformational change that
results in FBPase inhibition, Figure 1.3-5 Our initial compound
series contained a purine (e.g., compound 1) and a benzimid-
azole phosphonic acid (e.g., compound 2), which demonstrated
potent inhibition of human FBPase and elicited significant
glucose lowering in both fasted normal ratsand rodent diabetic
animal models of T2DM after parenteral administration.6,7
However, not surprisingly, these compound series exhibited
very poor oral bioavailability (OBAV) due to the highly
*To whom correspondence should be addressed. Phone: 732-594-
2018. E-mail: qun_dang@merck.com. Address: 126 East Lincoln Avenue,
Rahway, New Jersey 07065.
a Abbreviations: FBPase, fructose-1,6-bisphosphatase; AMP, 50-
adenosinemonophosphate; T2DM, type 2 diabetes mellitus; OBAV, oral
bioavailability; GNG, gluconeogenesis; MS, molecular sieves; G-LOW,
glucose lowering.
Inhibitor Design. High MW was deemed as a potential
structural limitation of the purine and benzimidazole FBPase
r
2010 American Chemical Society
Published on Web 12/02/2010
pubs.acs.org/jmc

154 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
Scheme 1a
a Reagents and conditions: (i) R5-CH2CO2H, TFAA, BF3 OEt;
3
(ii) (CH2OH)2, p-TsOH, toluene; (iii) LDA, THF, 0 ꢀC, then ClPO-
(OEt)2; (iv) EtOH, HCl.
Figure 1. Design and evolution of novel AMP mimics.
Scheme 2a
a Conditions: (i) CuBr2, EtOAc-CHCl3; (ii) RSNa, EtOH.
Scheme 3a
Figure 2. AMP interactions with human FBPase.9
inhibitor series, which contributed to the inability of using
prodrug approaches to deliver these two series of FBPase
inhibitors orally. Therefore, we focused our redesign efforts
on reducing the number of atoms. The strategy to trim the
purine and benzimidazole scaffolds was guided by both
existing SAR6,7 and the X-ray crystal structures of FBPase-
inhibitor complexes (e.g., AMP-FBPase complex, Figure 2).9
Because the six-membered ring portion of the purine/benz-
imidazole forms few direct interactions with the binding site,
efforts focused on five-membered heterocycles that could be
substituted in a manner that retained the two hydrogen bond
interactions with 31Thr and 17Val gained by the amino sub-
stituent on the six-membered ring. Following extensive evalua-
tion of compounds containing various five-membered het-
erocycles and substituents by molecular modeling and in vitro
testing,5 the thiazole scaffold was selected for further inves-
tigation.
Synthesis of Thiazole Analogues. Thiazoles are most often
prepared from R-bromoketones and thioamides or thioureas
via cyclization reactions. Thus, various ketones containing
a diethyl phosphonate group were prepared as shown in
Schemes 1-7. The ketones required for thiazole analogues
with a 2,5-furanyl linker were readily prepared from either
furan or 2-acylfurans, as shown in Scheme 1. Friedel-Crafts
acylation of furan with a carboxylic acid (R5-CH2CO2H) and
trifluoroacetic anhydride (TFAA) in the presence of a cata-
lytic amount of BF3-OEt2 gave furan derivatives 4. Protec-
tion of 4 as a cyclic ketal followed by lithiation and addition
of diethyl chlorophosphate, and final ketal removal using
EtOH-HCl, gave ketones 5 (e.g., 5b and 5c were prepared
via this route, and 5a was prepared from 2-acetylfuran in a
similar manner).
a Conditions: (i) PhCH2COCl, AlCl3, (CH2Cl)2; (ii) NaOH, dioxane,
80 ꢀC; (iii) P(OEt)3, BrCCl3.
Scheme 4a
a Reagents and conditions: (i) t-BuOK, CH2Cl2, -78 ꢀC, AcCl,
-78 ꢀC; (ii) n-BuLi, THF, -78 ꢀC, ZnCl2, -78 to 0 ꢀC, CuI, 0 ꢀC,
PhCH2COCl.
Compound 5a was used to prepare other sulfide-containing
ketones following a previously reported procedure (Scheme 2).10
Thus, R-bromination of compound 5a with copper(II) bro-
mide followed by reaction with various sodium salts of mer-
captans (RSNa) in ethanol gave ketones 6.
The ketones required for analogues with a 1,3-pyrrolyl
linker were prepared in three steps, as shown in Scheme 3.
Friedel-Crafts acylation of N-toluenesulfonylpyrrole with
phenylacetyl chloride in the presence of aluminum trichlo-
ride gave exclusively the 3-phenylacetyl derivative in excellent
yield. Removal of the N-toluenesulfonyl protecting group using
sodium hydroxide and subsequent reaction with triethyl

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 155
Scheme 8a
Scheme 5
Scheme 6a
a Conditions: (i) CuBr2, EtOAc-CHCl3; (ii) thiourea, EtOH;
(iii) R2CSNH2, EtOH.
a Conditions: (i) (CH2OH)2, p-TsOH, PhMe; (ii) LDA, THF, 0 ꢀC,
then ClPO(OEt)2; (iii) EtOH, HCl.
Scheme 9a
Scheme 7a
a Conditions: (i) isoamyl nitrite, acetonitrile, CuCl2 (for 16) or CuBr2
(for 17) or TMSSMe (for 18).
a Conditions: (i) HP(O)(OEt)2, Pd(PPh3)4, TEA, toluene, 100 ꢀC;
(ii) R CH2MgBr, THF; (iii) PCC, 4 A MS, CH2Cl2.
5
˚
Scheme 10a
phosphite in BrCCl3 gave keto-pyrrole 7 in 69% yield over
the three steps.
Thiazole analogues containing a 2,4-oxazolyl linker were
synthesized from the corresponding keto-oxazoles; these in
turn were prepared using a two-step sequence involving pre-
paration of oxazole-phosphonate esters using a minor mod-
11
€
ification of Schollkopf’s procedure, and their subsequent
acylation using a minor modification of Anderson’s method;12
the synthesis of oxazole 9 is exemplified in Scheme 4. Treat-
ment of diethyl (isocyanomethyl)-phosphonate with potas-
sium tert-butoxide in dichloromethane followed by addition
of acetyl chloride gave oxazole 8 in 37% yield. Lithiation of
oxazole 8 using n-BuLi followed by transmetalation with
zinc(II) chloride produced the 2-ZnCl species of oxazole 8,
which was reacted with phenylacetyl chloride in the presence
of cuprous iodide to give the oxazole 9 in 33% yield.
The ester-linked keto-phosphonate 10 was prepared via
direct Fisher esterification of 2-ketohexanoic acid with diethyl
hydroxymethylphosphonate in the presence of ethanesulfo-
nic acid to give 10 in 30% yield (Scheme 5).
The 2,6-pyridyl linked keto-phosphonate 11 was prepared
using the three-step procedure described in Scheme 1 for the
synthesis of 5a. Protection of 2-acetylpyridine as the cyclic
ketal was accomplished by heating a toluene solution of
2-acetylpyridine and ethylene glycol with a catalytic amount
of p-TsOH; subsequent lithiation with LDA and reaction
with diethyl chlorophosphate, followed by a final cyclic ketal
deprotection, gave ketone 11 (Scheme 6).
a Reagents and conditions: (i) CuBr2, EtOAc-CHCl3; (ii) EtO2C-
CSNH2, EtOH; (iii) NH3, EtOH; (iv) TFAA, TEA, THF; (v) LiBH4,
EtOH.
introduced via Grignard addition to the aldehyde, followed
by oxidation of the resulting alcohol using pyridinium chloro-
˚
chromate (PCC) in the presence of 4 A molecular sieves
(MS), affording ketones 12a-d.
Phosphonate-containing ketones described in Schemes 1-7
were readily converted to 2-aminothiazoles via the R-bromo-
ketone, derived from bromination using copper(II) bromide.
Cyclization reactions with thiourea gave aminothiazoles 14,
whereas cyclization with thioamides gave C2-substituted thia-
zole analogues 15 (Scheme 8).
Other C2-thiazole analogues (16-18) were readily prepared
from compound 14a through Sandmeyer reactions13,14
(Scheme 9). Direct conversion of the 2-amino group to a
2-chloro group was accomplished by treatment of thiazole
14a with isoamyl nitrite in the presence of copper(II) chlo-
ride, producing thiazole 16. Alternatively, reactions of thia-
zole 14a with isoamyl nitrite and copper(II) bromide or
(methylthio)trimethylsilane (TMSSMe) gave thiazoles 17
and 18, respectively.
Phenylphosphonates 12 were prepared from their corre-
sponding aryl bromides in three steps as shown in Scheme 7.
Phosphonylation of 4-substituted 3-bromobenzaldehydes with
diethyl phosphite in the presence of Pd(PPh3)4 and triethyl-
amine (TEA) gave the corresponding 3-diethylphosphono-
benzaldehydes. The R5 groups (defined in Table 5) were

156 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
Scheme 11a
Table 1. SAR of C2-Thiazole Analogues of 30aa
R2
Me
HL IC50, μM
G-LOW, %
a Reagents and conditions: (i) DMF, NCS (for 22) or NBS (for 23) or
NIS (for 24).
30a
30b
30c
30d
30e
30f
30g
30h
30i
0.1
0.4
1.2
0.22
0.5
0.18
0.08
0.89
2
55
Et
NDb
ND
8
vinyl
CH2OH
H
Scheme 12a
ND
ND
3
Cl
Br
SMe
CN
ND
ND
65
30j
NH2
0.025
1
30k
30l
NHMe
NHAc
CONH2
CSNH2
Ph
ND
ND
ND
ND
ND
ND
ND
10
30m
30n
30o
30p
30q
2.75
0.5
13.5
8
2-thienyl
3-pyridyl
a Reagents and conditions: (i) Pd(PPh3)4, ArB(OH)2, K2CO3, DMF-
5
H2O, 80 ꢀC; (ii) Pd(PPh3)4, morpholine, EtOH, CH2Cl2.
a HL, human liver FBPase; IC50 is an average of 3 runs; G-LOW,
glucose lowering (Cmax) in fasted normal rats after iv administration at
doses of 10 mg/kg. b ND, not determined.
Scheme 13a
Table 2. SAR of C5-Thiazole Analogues with Various Alkyl Groupsa
R5
HL IC50, μM
G-LOW, %
30j
i-Bu
H
0.025
0.45
0.12
65
31a
31b
31c
31d
31e
31f
31g
31h
31i
NDb
7
Me
HOCH2
0.5
0.03
ND
64
n-Pr
i-Pr
a Reagents and conditions: (i) H2NCH2PO(OEt)2, EDCI, HOBt, DMF;
(ii) Br2, CH2Cl2; (iii) Pd(PPh3)4, 2-thienyl-SnBu3, DMF;(iv) TFA, CH2Cl2.
0.028
0.057
0.012
0.019
0.021
0.01
80
46
CF3CH2
neopentyl
80
24
Three other C2-analogues were prepared from ketone
5c as shown in Scheme 10. Bromination of ketone 5c with
copper(II) bromide followed by cyclization with ethyl thio-
oxamate gave a thiazole with an ethoxycarbonyl group at the
C2-position, which was treated with ammonia to give car-
boxamide 19 in 63% yield; alternatively, the C2-ester group
was reduced with lithium borohydride to give hydroxymethyl-
thiazole 21 in 87% yield. Dehydration of thiazole 19 using
TFAA produced 2-cyanothiazole 20 in 67% yield.
As illustrated in Schemes 1-8, R5 groups were introduced
as part of the ketone moiety prior to thiazole ring formation.
Alternatively, R5 groups could be introduced after the thiazole
was formed. For example, C5-halogenation of thiazole 14b
was readily achieved with NCS, NBS, and NIS, leading to
C5-chloro (22), C5-bromo (23), and C5-iodo (24) thiazoles
(Scheme 11), respectively.
cyclobutyl
cyclopentyl
cyclohexyl
cyclopropyl-CH2
cyclopentyl-CH2
cyclohexyl-CH2
PhCH2
68
78
31j
31k
31l
0.02
75
55
0.018
0.059
0.15
31m
31n
31o
ND
-5%
ND
morpholinyl-CH2
0.56
a HL, human liver FBPase; IC50 is an average of 3 runs; G-LOW,
glucose lowering (Cmax) in fasted normal rats after i.v. administration at
a dose of 10 mg/kg; b ND, not determined.
The amide-linked thiazole 28 was prepared in four steps as
shown in Scheme 13. Coupling of 2-(tert-butyloxycarbonyl-
amino)-4-thiazolecarboxylic acid with diethyl aminomethyl-
phosphonate using EDCI followed by bromination with
bromine gave thiazole 27. Stille coupling of thiazole 27 with
2-(tributylstannyl)thiophene followed by removal of the Boc
group gave thiazole 28.
Final thiazole phosphonic acids (Tables 1-5) were readily
obtained via TMSBr-mediated removal of the phosphonate
diethyl esters from thiazoles 29 (Scheme 14).
The C5-bromo group of thiazole 23 served as an excellent
handle to introduce other C5-groups via various coupling reac-
tions (Scheme 12). For instance, Suzuki coupling of thia-
zole 23 with boronic acids gave C5-aryl thiazoles 25, while
Buchwald coupling15 with morpholine produced 5-morpholinyl
thiazole 26.

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 157
Table 5. SAR of Linker Thiazole Analoguesa
Table 3. SAR of C5-Thiazole Analogues with Halo and Heteroatom-
Containing Groupsa
linkerb
31a 2,5-furanyl
R5
HL IC50, μM G-LOW, %
R5
HL IC50, μM
G-LOW, %
H
0.45
0.03
0.014
NDc
64
30j
i-Bu
Cl
Br
I
0.025
0.07
0.05
65
17
31d 2,5-furanyl
33a 2,5-furanyl
34a 2,5-thienyl
34b 1,3-pyrrolyl
n-Pr
Ph
32a
32b
32c
32d
32e
32f
32g
32h
32i
76
20
n-Pr
Ph
>10
ND
ND
ND
63
0.1
-2
75
NDb
82
>10
10
1-morpholinyl
EtS
n-PrS
0.016
0.033
0.016
0.024
0.024
0.3
34c 4,2-oxazolyl-(5-Me) Ph
34d -CH2OCO-
34e -CH2NHCO-
n-Pr
2-thienyl
H
0.05
0.95
2
ND
ND
ND
0
i-PrS
t-BuS
PhS
48
34f
2,6-pyridyl
ND
ND
ND
76
34g 1,3-phenyl
34h 1,3-phenyl
H
1.3
n-Pr
n-Pr
0.25
0.135
0.21
0.08
32j
CONMe2
CO2Et
CO2Bn
n-PrSO
1.7
34i
34j
1,3-phenyl-(6-Me)
0
32k
32l
0.014
0.015
0.858
1,3-phenyl-(6-OMe) i-Pr
Ph
9
42
ND
34k 1,3-phenyl-(6-F)
0
32m
a HL, human liver FBPase; IC50 is an average of 3 runs; G-LOW,
glucose lowering in fasted normal rats after iv administration at a dose
of 10 mg/kg. b The first term is connected to the phosphorus atom, e.g.
1,3-pyrrolyl: -PO(OH)2 is connected to the N1 of pyrrole. c ND, not
determined.
a HL, human liver FBPase; IC50 is an average of 3 runs; G-LOW,
glucose lowering in fasted normal rats after iv administration. b ND, not
determined.
Table 4. SAR of C5-Thiazole Analogues with Aryl Groupsa
Scheme 14
R5
HL IC50, μM
G-LOW, %
30j
i-Bu
Ph
0.025
0.014
0.043
0.021
0.022
0.021
0.088
0.014
0.016
0.013
0.032
0.041
0.034
0.012
0.04
65
76
66
0
33a
33b
33c
33d
33e
33f
33g
33h
33i
2-MeO-Ph
3-MeO-Ph
4-MeO-Ph
4-MeS-Ph
4-t-Bu-Ph
4-MeO2C-Ph
4-F-Ph
Scheme 15
63
72
6
78
60
40
43
78
36
20
21
24
4-Cl-Ph
4-Ac-Ph
33j
Evaluations of Thiazole Phosphonic Acids. Initially, we
expected that a thiazole ring with the amino substituent
could best be replaced with an amide group at the C2-position.
The first thiazole analogue prepared was compound 30a
because thioacetamide was available in our laboratory.
Interestingly, the C2-methyl thiazole 30a potently inhib-
ited human liver FBPase (IC50=100 nM). Moreover, after
intravenous (iv) administration to fasted normal rats at a
dose of 10 mg/kg, thiazole 30a lowered blood glucose
(Cmax) by 55% compared to vehicle-treated animals,6 indi-
cating that thiazole 30a was also able to inhibit GNG in vivo.
The initial encouraging activity exhibited by thiazole 30a
prompted us to explore the SAR of the C2 position further;
results are summarized in Table 1.
Larger C2-groups such as ethyl and vinyl groups resulted
in loss of potency (thiazoles 30b and 30c) compared to 30a, as
did adding a hydrogen bond donor HO group (thiazole 30d).
Eliminating the C2-Me group also led to a 5-fold weaker
compound in thiazole 30e, indicating that a suitable C2-
group is important for potency. Halo groups (Cl and Br) are
tolerated at C2 leading to analogues (30f and 30g) with po-
tency comparable to 30a, while analogues with C2-SMe and
cyano groups (30h and 30i) are significantly weaker (>8-fold).
33k
33l
4-MeSO2-Ph
4-Ph-Ph
33m
33n
33o
2-nathphyl
2-furanyl
2-thienyl
0.044
a HL, human liver FBPase; IC50 is an average of 3 runs; G-LOW,
glucose lowering in fasted normal rats after iv administration at a dose of
10 mg/kg.
Direct oxidation of analogue 32f10 using mCPBA gave the
C5-sulfoxide analogue 32m in 50% yield, Scheme 15.
To deliver thiazole phosphonate FBPase inhibitors orally,
various prodrugs were prepared to mask the double negative
charge. The classic acyloxyalkyl ester prodrugs (35a-35e) were
obtained via direct alkylation reactions, while other prodrug
esters (35f;35i) were prepared via the phosphonic dichlo-
ride coupling method as depicted in Scheme 16.
Saponification of the diphenyl ester 35g gave the mono-
phenyl ester of thiazole 30j as 35j. The monoester 35j was also
used to prepare mixed amide prodrugs 35k and 35l (Scheme 17).
Conversion of 35j to its corresponding monophosphonic chlo-
ride followed by reaction with either ammonia or L-alanine
ethyl ester gave prodrugs 35k and 35l, respectively.

158 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
Scheme 16a
alkyl groups and the thiazole core is tolerated, leading to com-
pounds (31k-31m) with good potency (IC50 18-59 nM),
while C5-benzyl and C5-morpholinylmethyl analogues (31n
and 31o) are much weaker than thiazole 30j. Thus, the SAR
of C5-alkyl thiazoles indicates that a proper size alkyl group
is preferred at the 5-position, and through this effort several
thiazole analogues (31d, 31e, 31g, 31i, 31j, and 31k) with
robust glucose lowering activity were discovered.
Exploring the SAR further, halo and heteroatom-contain-
ing groups were explored at the 5-position, and results are
summarized in Table 3.
Halogens are somewhat tolerated at the 5-position (32a-32c),
with the 5-bromo thiazole being the most potent (IC50 =50nM).
The 5-morpholinyl analogue 32d is a potent FBPase inhibi-
tor with an IC50 of 16 nM; however, it is not very stable
chemically at room temperature. The instability of 32d is
possibly due to the highly electron-rich thiazole ring that is
likely to be prone for oxidation. Alkylthio groups are well
tolerated at the 5-position (32e-32h), leading to potent
FBPase inhibitors (IC50 16-33 nM), while the phenylthio
analogue 32i is 12-fold weaker than thiazole 30j. The C5-
amide analogue 32j is 68-fold weaker than 30j, while C5-ester
analogues 32k and 32l are 2-fold more potent than 30j.
Conversion of the sulfide analogue 32f to its corresponding
sulfoxide 32m resulted in >34-fold loss in potency. The
diethyl ester of 5-bromothiazole 32b provided an excellent
handle to study aryl groups at the 5-position and results are
summarized in Table 4.
a Reagents and conditions: (i) R-I, Hunig’s base, DMF (for 35a-35e,
Table 6); (ii) SO2Cl2, then ROH, Hunig’s base, DMF (35f-35g, Table 7);
(iii) SO2Cl2, then HO(CH2)2CH(OH)X, TEA, THF.
Scheme 17a
The 5-phenylthiazole analogue 33a is 2-fold more potent
than 30j, while various substituted phenyl groups are also
tolerated, leading to potent FBPase inhibitors (thiazoles
33b-33l, IC50 = 13-88 nM). The 5-(2-naphthyl)thiazole
33m (IC50 = 12 nM) is 2-fold more potent than 30j, while two
C5-heteroaryl analogues (33n and 33o) are slightly weaker
than thiazole 30j. Consistent with modeling studies, the SAR
summarized in Tables 2-4 indicates that the C5 groups are
well tolerated, with alkyl, ester, and aryl substituents leading
to potent FBPase inhibitors (IC50 10-25 nM).
a Reagents and conditions: (i) NaOH, EtOH; (ii) SO2Cl2, then NH3
(for 35k) or L-alanine ethyl ester (for 35l), TEA, CH2Cl2.
Replacement of C2-methyl group with an amino group
gave analogue 30j, which is 4-fold more potent than 30a and
elicits potent glucose lowering in fasted normal rats (65%).
N-Methylation and acetylation of the 2-amino group of 30j
are detrimental for potency (thiazole 30k and 30l). The ini-
Although the furan linker was the best in the earlier purine
and benzimidazole series of FBPase inhibitors, various link-
ing groups were prepared prior to embarking on prodrug
efforts. The results are summarized in Table 5.
tially desired C2-amide 30m has only weak activity (IC50
=
2.75 μM), possibly due to desolvation energy costs associated
with the amide carbonyl group. The thioamide analogue 30n
is more potent than the corresponding amide 30m, which
supports this hypothesis. Finally, both aryl and hetero-
aryl groups were tested at the C2-position but all led to ana-
logues (30o-30q) with much weaker activity. Therefore,
the C2-SAR indicates the binding pocket around the
C2-position prefers a small hydrogen bond donating
group. Having identified that an amino group is optimal at
the C2-position, 2-aminothiazoles with various alkyl groups
at the C5-position were explored next and results are sum-
marized in Table 2.
Replacement of the furan with other five-membered het-
erocycles is detrimental to FBPase inhibition, leading to
inactive compounds (34a-34c). This reflects the ability of
the furan oxygen to receive a strong electrostatic interaction
from 30Leu based on the X-ray structure of human FBPase-
compound 30j complex (Figure 3).5 Interestingly, replace-
ment of the furan with ester- and amide linked thiazoles (34d
and 34e) led to compounds that retained FBPase inhibition,
with the ester linker analogue 34d showing potency compar-
able to the furan-linked thiazole 31d. Two six-membered
linker analogues (34f and 34g) were also explored, but both
lost potency (>4-fold for 34f and ca. 3-fold for 34g) com-
pared to their corresponding furan-linked analogue 31a. The
C5-thiazole SAR presented in Tables 2-4 indicates that an
alkyl group is required for high potency and because both 34f
and 34g do not have an equivalent C5-alkyl group, we
selected the more potent 1,3-phenyl-linked analogue 34g
for further optimization. Introduction of a propyl group
gave analogue 34h, which is 5-fold more potent than 34g but
still >8-fold weaker than its corresponding furan analogue
31d. Additional substitutions of the 1,3-phenyl linker and
changing the n-propyl group (analogues 34i-34k) led to
Removal of the C5-isobutyl group gave thiazole 31a, which
is 18-fold weaker than 30j. Other groups such as methyl and
hydroxymethyl at the C5-position (analogues 31b-31c) are
also weaker than thiazole 30j. On the other hand, larger C5-
alkyl analogues (31d-31f) exhibited potency comparable to
30j, and the bulky C5-neopentyl analogue 31g produced a
2-fold improvement in potency over thiazole 30j. C5-cyclic
alkyl thiazoles (31h-31j) are also potent FBPase inhibitors,
with thiazole 31j being the most potent inhibitor (IC50
=
10 nM). Insertion of a methylene group between the cyclic

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 159
Table 7. OBAV SAR of Thiazole 30j Prodrugsa
X
Y
MW OBAV, % G-LOW, %
30j
2
NDb
0c
35a t-BuCO2CH2O-
35b EtOCO2CH(Me)O-
35c PhOCO2CH2O-
35d EtOCO2CH2O-
35e i-PrOCO2CH2O-
35f 4-AcO-PhCH2O-
35g PhO-
X
X
X
X
X
X
X
530.6
534.5
602.6
506.5
534.5
598.6
454.5
452.9
419.4
378.4
377.4
11
ND
7
12
ND
40d
ND
ND
ND
53d
ND
ND
ND
46d
51
17
13
8
0.2
8
35h 3-Cl-Ph-cyclic estere
35i 4-pyridyl cyclic estere
35j PhO-
35k PhO-
35l L-alanine-OEt
8
OH
NH2
1
7
PhO- 477.5
12
26
22
47
Figure 3. X-ray structural analysis of 30j (in yellow) bound to FBPase.
Table 6. Pharmacoketic Profiles of Thiazole 30j and Benzimidazole 2a
35m glycine-OEt
35n L-alanine-OEt
35o EtO2CC(Me)2NH
X
X
X
472.5
500.6
528.6
59
45
a OBAV, determined by measuring urinary excretion of 30j following
oral administration of the prodrug vs iv administration of 30j; G-LOW,
glucose lowering in normal fasted rats after oral administration of a
10 mg/kg dose. b ND, not determined. c When 35a was dosed iv, blood
glucose was lowered by 53%. d Compound dosed at 30 mg/kg orally.
e Structure is shown in Scheme 17.
T1/2
h
,
Cl, Vd,
L/h/kg L/kg
protein
binding, %
OBAV,
%
compound
2
30j
0.75
0.65
0.26
0.24
0.25
0.18
99, rat; 99.9, human
90, rat; 93, human
2
4
a Fed male Sprague-Dawley rats were used; protein binding deter-
mined using heparinized rat and human plasma.
similar to the bisPOM prodrug 35a, it did not produce
significant glucose lowering. However, at a higher dose of
30 mg/kg, prodrug 35d lowered glucose by 40%. Four other
esters (35f-35i), a monophenyl ester (35j), and a monophe-
nyl amide (35k) were tested as prodrugs, but all gave <8%
OBAV. The 1-(3-chlorophenyl)-1,3-propanyl cyclic ester
(HepDirect)17 prodrug 35h was also tested at the higher dose
of 30 mg/kg, and it lowered blood glucose by 53%. An aryl
amidate (McGuigan) prodrug18 was prepared as compound
35l, which gave OBAV of 12% but only lowered glucose
orally at the higher dose of 30 mg/kg. Overall, although some
progress was made toward achieving OBAV of >15%, no
significant oral efficacy was obtained at the screening dose of
10 mg/kg. Therefore, the search was expanded to include novel
prodrug types; these efforts culminated in the discovery of
phosphonic diamides, a new prodrug class that enabled oral
delivery of these thiazole phosphonic acids.10 Optimization
of the phosphonic diamide prodrugs produced three pro-
drugs (35m-35o) that showed good OBAV (22-47%). More-
over, all three phosphonic diamide prodrugs elicited signifi-
cant glucose lowering in fasted rats after oral administration
of a 10 mg/kg dose. Extensive confirmatory studies led to the
selection of diamide 35n for further evaluation.
Efficacy Study in Rodent Models of T2DM. Compound
35n inhibits glucose production in both human and rat
primary hepatocytes in a concentration-dependent manner.3
The inhibitory potency of compound 35n was independent
of substrates used for the hepatocyte assay (i.e., lactate or
glycerol), which is consistent with the expected mechanism of
action, i.e., blocking of GNG via FBPase inhibition. More-
over, compound 35n demonstrated oral glucose-lowering
effects in several rodent models of T2DM (e.g., STZ, ZDF,
and GK rats, and db/db mice), establishing it as the first
FBPase inhibitor with reported oral activity in animal
models.19-21
further potency improvement over the initial lead 34g, with
analogue 34k being the most potent FBPase inhibitor of this
series (IC50 = 80 nM). However, analogue 34k is still ca.
6-fold weaker than its corresponding furan-linked analogue
33a and surprisingly did not lower glucose in the normal
fasted rat. The linker SAR indicated that the furan was still
optimal for the thiazole scaffold, thus thiazole 30j was selected
for further evaluation. A full pharmacokinetic (PK) study of
thiazole 30j was carried out, and the results are summarized
in Table 6.
Thiazole 30j showed similar PK profile compared to
benzimidazole 2, except plasma protein binding (Table 6);
the thiazole scaffold is significantly less protein bound com-
pared to the benzimidazole scaffold, which should translate
into more free drugs for the thiazole scaffold.
With the inhibitor optimization completed, attention turned
to preparation of phosphonate prodrugs in order to achieve
adequate OBAV.16 A wide variety of prodrugs was prepared
to attempt the orally delivery of thiazole 30j; the OBAV SAR
of these prodrugs in rats are summarized in Table 7.
The classic (pivaloyloxy)methyl phosphonate diester
(bisPOM) prodrug (35a) of thiazole 30j gave 11% OBAV in
our urinary screening assay,10 which represents a 5-fold im-
provement over the phosphonic acid 30j. However, in the
fasted normal rat assay, oral administration of prodrug 35a
did not lower glucose, while iv dosing did lower glucose by
53%. The glucose lowering after iv administration is likely
due to the rapid conversion of the bisPOM prodrug 35a to
the parent 30j by esterases in plasma such that iv adminis-
tration of 35a is essentially equal to iv dosing of the parent
30j. Four carbonate prodrugs (35b-35e) were also studied,
with the ethyl carbonate 35d showing the highest OBAV
among the carbonate prodrugs. Prodrug 35d was tested for
oral efficacy at the screening dose of 10 mg/kg; however,

160 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
Experimental Section
General Synthetic Methods. Compounds 30e, 30j, 31c, 32f, 32k,
35a, 35e, 35m, 35n, and 35o were prepared according to reported
procedures.3,5 All moisture sensitive reactions were performed
under a nitrogen atmosphere using flame-dried glassware and an-
hydrous solvents purchased from Aldrich. TLC was performed on
Analtech Uniplate silica gel GHLF (250 μm, 10 cm ꢀ 20 cm)
plates. Flash chromatography was performed on 230-400 mesh
EM Science silica gel 60. Melting points were recorded on a
Thomas-Hoover capillary melting point apparatus and uncor-
1
rected. H spectra were recorded on Varian Gemini or Mercury
spectrometers operating respectively at 200 or 300 MHz. Mass
spectra data were determined on a Perkin-Elmer Sciex API2000
LC-MS system. All compounds tested for biological activity have
g95% purity. Purity was determined via either analytical HPLC
(performed on a 4.6 mm ꢀ 250 mm YMC ODS-AQ 5 μm column
eluting with a 0.1% aqueous AcOH/MeOH gradient detected at
280 nm) or elemental (C, H, N) microanalyses (performed by
NuMega Resonance Laboratories, Inc., San Diego, CA, or by
Robertson Microlit Laboratories).
Figure 4. Effects of 35n prevention and intervention therapy in
male ZDF rats. Compound 35n was administered as a food ad-
mixture (0.4%, w/w) to 6-week-old (prevention) or 10-week-old
(intervention) rats. n = 8/group. *p < 0.05 compared to controls
(ANOVA).
2-(4-Methylpentanoyl)furan (4). A solution of furan (225 mL,
3.09 mol) and 4-methylpentanoic acid (276 g, 2.37 mol) in
anhydrous toluene (2.25 L) was treated with trifluoroacetic
anhydride (400 mL, 2.83 mol) under a nitrogen atmosphere
followed by boron trifluoride etherate (29 mL, 0.24 mol). After
heating the solution to 50-55 ꢀC for 2 h, the reaction was cooled
to 8-10 ꢀC (ice bath) and neutralized with a sodium carbonate
solution (21 wt %, 1.8 L, 3.5 mol). The layers were separated,
and the organic layer was filtered through a Celite pad. The filtrate
was dried (MgSO4) and concentrated in vacuo. The resulting dark
oil was purified by vacuum distillation (bp 66-68 ꢀC, 11 mmHg)
to give 2-(4-methylpentanoyl)furan as a colorless liquid (310 g,
A study was also carried out in six-week old male ZDF rats
in both prevention and intervention modes to demonstrate
further the potential of compound 35n as a potential treat-
ment for T2DM.19 Compound 35n was studied using three
cohorts of rats: one group of control animals treated with
vehicle, a second group in which drug treatment was initiated
at six weeks of age (prevention mode), and a third group in
which drug treatment was initiated at 10 weeks of age (inter-
vention mode). Compound 35n was administered as a food
admixture in these studies (0.4%). In the prevention mode,
drug treatment delayed the development of T2DM as in-
dicated by the slow progression of hyperglycemia compared
to vehicle-treated animals (Figure 4). In the intervention
mode, treatment with 35n resulted in a marked correction of
established hyperglycemia relative to the control group. This
study suggests that compound 35n might be useful to treat
both the early and late stages of T2DM.
1
78%). H NMR (CDCl3): δ 7.59 (m, 1H), 7.19 (m, 1H), 6.55
(m, 1H), 2.81 (t, 2H, J = 6.8 Hz), 1.61 (m, 3H), 0.98 (d, 6H, J =
6.8 Hz).
Diethyl (5-Acetylfuran-2-yl)phosphonate (5a). A solution of
2-acetylfuran (102 g, 926 mmol) and ethylene glycol (78 mL,
1399 mmol) in anhydrous benzene (500 mL) was treated with p-
toluenesulfonic acid (8 g, 42 mmol) and heated to reflux while
removing water with a Dean-Stark trap. After 3 h, the reac-
tion was cooled to room temperature, diluted with benzene
(150 mL), and washed with aqueous sodium hydroxide (0.5 M,
300 mL). The organic phase was dried (MgSO4), filtered, and
concentrated in vacuo to give the desired cyclic ketal as a dark
oil (134 g, 93%). A solution of the ketal in anhydrous THF (350 mL)
was treated with 1,4-diazabicyclo[2,2,2]octane (97.5 g, 869 mmol),
cooled to -42 ꢀC, and treated with a solution of n-butyl lithium
(2.5 M, 347 mL) in hexane via an addition funnel at a speed that
maintained the reaction temperature between -43 and -40 ꢀC.
After the addition was complete, the solution was stirred at
0-2 ꢀC for 1 h. In a separate flask, a solution of diethyl chloro-
phosphate (138 mL, 955 mmol) in anhydrous THF (350 mL)
was cooled to -45 ꢀC and treated with the above generated
lithiated 2-acetylfuran cyclic ketal solution (cooled to -8 ꢀC)
via a cannula at a rate that maintained the reaction temperature
between -44 and -40 ꢀC. After the addition was complete, the
reaction was stirred at room temperature for 2 h. The solvent
was removed in vacuo, and the residue was partitioned between
ethyl acetate (600 mL) and water (600 mL). The aqueous phase
was extracted again with ethyl acetate (500 mL), and the organic
extracts were combined, washed with brine (500 mL), dried
(MgSO4), filtered, and concentrated in vacuo to give a dark
oil (214 g). The crude 5-diethylphosphono-2-acetylfuran
cyclic ketal was dissolved in methanol (500 mL) and treated
with hydrochloric acid (1 N, 200 mL). After heating at 60 ꢀC
for 12 h, the methanol was removed in vacuo and the result-
ing dark oil was partitioned between ethyl acetate (600 mL)
and saturated sodium bicarbonate solution (300 mL). The
layers were separated, and the aqueous phase was extracted
Conclusion
Earlier prodrug efforts to achieve oral delivery of the pre-
viously reported purine and benzimidazole FBPase inhibitors
were unsuccessful, and high molecularweight(>600) wassus-
pected as a contributing factor. To discover a new scaffold
more amenable to oral delivery, a structure-guided drug de-
sign approach was used to trim the 5,6-fused heterocyclic sys-
tems of purine and benzimidazole to the five-membered hetero-
cycle thiazole. The thiazole series of FBPase inhibitors proved
to not only have improved potency against human liver FBPase
(many thiazoles with IC50 of 10-30 nM were discovered) but
also to be suitable for oral delivery via prodrugs, with the lead
compounds achieving oral bioavailability in the range of
25-50%. Compound 30j demonstrated potent inhibition of
glucose production in both rat and human hepatocytes3 and
elicited robust glucose lowering in rodent models of T2DM
after parental administration.19-21 Oral delivery of 30j was
achieved via its phosphonic diamide prodrug, a novel phos-
phonate prodrug approach designed and used to effectively
deliver phosphonate-containing FBPase inhibitors. Extensive
lead optimization of both the thiazole FBPase inhibitor series
and their prodrugs culminated in the discovery of compound
35n, the first oral FBPase inhibitor to advance to human
clinical trials in patients with T2DM.22

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 161
with EtOAc (400 mL). The combined organic extracts were
washed with brine (500 mL), dried (MgSO4), filtered, and
concentrated in vacuo to give a dark oil (167 g). The residue
was purified by vacuum distillation 126-145 ꢀC/0.10 Torr to
give compound 5a as a pale-orange oil (112 g, 61%). 1H NMR
(CDCl3): δ 7.18 (m, 2H), 4.19 (m, 4H), 2.49 (s, 3H), 1.28 (t, 6H,
J = 7.2 Hz).
ethyl acetate (2.5 L), and the organic phases were combined and
washed with a saturated solution of ammonium chloride, which
was subsequently re-extracted with dichloromethane (2 ꢀ 30 mL).
The organic extracts were combined, washed with brine (1.2 L),
dried (MgSO4), filtered, and concentrated in vacuo to give the
crude diethyl [5-(2-bromo-4-methyl-pentanoyl)-furan-2-yl]phos-
phonate as a greenish-yellow oil (587 g, 93%).
Diethyl (5-Pentanoylfuran-2-yl)phosphonate (5b). Prepared in
analogous fashion to compound 5a from 1-furan-2-yl-pentan-1-
one to give 5b as a brown oil. 1H NMR (DMSO-d6): δ 7.20 (m,
2H), 4.20 (m, 4H), 2.82 (t, 2H, J = 7.4 Hz), 1.76-0.90 (m, 13H).
Diethyl [5-(4-Methylpentanoyl)furan-2-yl]phosphonate (5c).
Prepared in analogous fashion to compound 5a from furan 4
to give 5c (purified via vacuum distillation, 126-145 ꢀC/0.10
The above bromoketone (7.16 g, 18.79 mmol) was then dis-
solved in anhydrous ethanol (90 mL), treated with ethyl thioox-
amate (5 g, 37.59 mmol), and heated at reflux. After 14 h, the
reaction mixture was cooled to room temperature, concentrated
under reduced pressure, neutralized with saturated sodium bicar-
bonate (50 mL), and extracted with ethyl acetate (3 ꢀ 50 mL).
The combined organic extracts were washed with brine (50 mL),
dried (MgSO4), filtered, and the filtrate concentrated under reduced
pressure to give a light-orange solid. The crude solid was puri-
fied by flash chromatography (SiO2, 5 cm ꢀ15 cm, 50% EtOAc-
hexane) to give 2-ethoxycarbonyl-4-[2-(5-diethylphosphono)-
furanyl]-5-isobutylthiazole as a yellow solid (4.5 g, 58%).
A solution of 2-ethoxycarbonyl-4-[2-(5-diethylphosphono)-
furanyl]-5-isobutylthiazole (250 mg, 0.60 mmol) in ammonia-
saturated methanol (5 mL) was stirred at room temperature for
18 h. The reaction solution was evaporated to give compound 19
as white solid (145 mg, 63%). 1H NMR (200 MHz, CDCl3): δ
7.23-7.26 (m, 1 H), 6.83-6.86 (m, 1 H), 4.11-4.22 (m, 4 H),
3.04 (d, J = 7 Hz, 2 H), 1.93-2.03 (m, 1 H), 1.30-1.37 (m, 6 H),
0.99 (d, J = 7 Hz, 6 H).
2-Cyano-4-[2-(5-diethylphosphono)furanyl]-5-isobutylthiazole
(20). A suspension of 19 (145 mg, 0.375 mmol) in THF (2 mL)
was treated with TEA (228 mg, 2.25 mmol), cooled to 0 ꢀC, and
then treated with TFAA (236 mg, 1.13 mmol). The reaction
mixture was stirred at room temperature for 18 h. The reaction
was quenched with saturated sodium bicarbonate (20 mL) and
extracted with EtOAc (3ꢀ 20 mL). The combined organic extracts
were dried (MgSO4) and evaporated, and the residue was puri-
fied by flash chromatography (SiO2, 2 cm ꢀ 15 cm, 60% EtOAc-
hexane) to give compound 20 (92 mg, 67%). 1H NMR (200 MHz,
CDCl3): δ 7.23-7.26 (m, 1 H), 6.90-6.98 (m, 1 H), 4.08-4.24
(m, 4 H), 3.11 (d, J = 7 Hz, 2 H), 1.93-2.03 (m, 1 H), 1.25-1.39
(m, 6 H), 1.11 (d, J = 7 Hz, 6 H).
2-Hydroxymethyl-4-[2-(5-diethylphosphono)furanyl]-5-isobutyl-
thiazole (21). A solution of 2-ethoxycarbonyl-4-[2-(5-diethyl-
phosphono)furanyl]-5-isobutylthiazole (see above procedure
for compound 19, 3.12 g, 7.50 mmol) in ethanol (60 mL) was
cooled to 0 ꢀC and treated with a solution of lithium borohy-
dride (2 M, 4.48 mL) in THF. The reaction solution was
stirred at room temperature for 1 h and then at 60 ꢀC for 1 h.
The cooled reaction solution was evaporated to dryness, and
the residue was partitioned between EtOAc and water. The
layers were separated, and the aqueous phase was extracted
with EtOAc (4 ꢀ 25 mL). The combined organic extracts were
dried (MgSO4) and evaporated, and the residue was purified
by flash chromatography (SiO2, 3 cm ꢀ 15 cm, 60% EtOAc-
hexane) to give compound 21 as colorless oil (2.4 g, 87%). 1H
NMR (200 MHz, CDCl3): δ 7.13-7.18 (m, 1 H), 6.69-6.74
(m, 1 H), 4.91 (s, 2 H), 3.97-4.17 (m, 4 H), 2.86 (d, J = 7 Hz,
2 H), 1.78-1.92 (m, 1 H), 1.21-1.32 (m, 6 H), 0.92 (d, J =
7 Hz, 6 H).
General Procedure for C5-Halogenation of Thiazole 14b. 2-
Amino-4-[2-(5-diethylphosphono)furanyl]-5-bromothiazole (23).
A solution of compound 14b (5.0 g, 17 mmol) in anhydrous
chloroform (70 mL) was cooled to 0 ꢀC and treated with N-
bromosuccinimide (3.63 g, 20 mmol) under nitrogen. The reac-
tion was allowed to warm to room temperature after 15 min and
then stirred for 2 h. After quenching with saturated Na2S2O3
(20 mL), the resulting mixture was diluted with water (20 mL)
and extracted with dichloromethane (2 ꢀ 30 mL). The combined
organics were dried (MgSO4) and evaporated to give a yellow
solid, which was purified by flash chromatography (SiO2, 5 cm ꢀ
10 cm, 100% EtOAc) to give compound 23 as a brown solid
1
Torr) as a pale-orange oil (112 g, 61%). H NMR (CDCl3): δ
7.19 (m, 2H), 4.22 (m, 4H), 2.87 (t, 2H, J = 6.6 Hz), 1.61 (m,
3H), 1.39 (m, 6H), 0.97 (d, 6H, J = 6.6 Hz).
General Procedure for Synthesis of Thiazoles 14 and 15 from
Ketones 13. 2-Amino-4-[2-(5-diethylphosphono)furanyl]thiazole
(14b). A solution of (5-acetyl-furan-2-yl)-phosphonic acid diethyl
ester (5a) (2 g, 8.12 mmol) in anhydrous ethyl acetate (20 mL) and
chloroform (20 mL) was treated with copper(II) bromide (5 g, 22.4
mmol) and stirred at room temperature under nitrogen for 24 h.
The reaction was quenched with saturated ammonium chloride
and extracted with dichloromethane (2 ꢀ 30 mL). The combined
organic extracts were washed with water (30 mL) and brine
(30 mL), dried (MgSO4), filtered, and concentrated in vacuo to
give a yellow oil. 1H NMR (CDCl3): δ 7.33 (m, 1H), 7.21 (s, 1H),
4.38 (s, 2H), 4.23 (m, 4H), 1.39 (m, 6H).
The crude 2-bromoacetyl-5-diethylphosphonofuran (137 mg,
0.423 mmol) was dissolved in anhydrous ethanol (2 mL) and
treated with thiourea (64 mg, 0.846 mmol). After heating at
reflux for 2 h, the reaction mixture was cooled, quenched with
saturated sodium bicarbonate, and extracted with ethyl acetate
(3 ꢀ 30 mL). The combined organic extracts were dried (MgSO4)
and filtered, and the filtrate was concentrated under reduced
pressure to give a yellow solid. The crude solid was purified by
flash chromatography (SiO2, 1 cm ꢀ 15 cm, 50, 65, 80, and 100%
EtOAc-hexane, gradient elution) to give compound 14b as a
yellow solid (90 mg, 70%). Melting point 144-148 ꢀC. 1H NMR
(CDCl3): δ 7.13 (m, 1H), 6.78 (s, 1H), 6.61 (m, 1H), 6.38 (bs, 2H),
4.35 (m, 4H), 1.35 (m, 6H). [MH]þ calcd for C11H15N2O4PS 303;
found:303. Anal. Calcd for C11H15N2O4PS þ 0.1H2O: C, 43.45;
H, 5.04; N, 9.21. Found: C, 43.14; H, 4.88; N, 9.14.
General Procedure for Syntheses of 2-Halothiazoles (16-18). 2-
Chloro-4-[2-(5-diethylphosphono)furanyl]-5-isobutylthiazole (16).
A solution of compound 14a (400 mg, 1.12 mmol) in acetonitrile
(10 mL) was treated with copper(II) chloride (180 mg, 1.34 mmol),
cooled to 0 ꢀC, and then added isoamyl nitrite (197 mg, 1.68 mmol).
The resulting reaction mixture was stirred at room temperature
for 1.5 h and then partitioned between EtOAc and water. The
layers were separated, and the aqueous phase was extracted with
EtOAc (3 ꢀ 20 mL). The combined organic extracts were dried
(MgSO4) and evaporated to dryness. The residue was purified
by flash chromatography (SiO2, 2 cm ꢀ 15 cm, 50% EtOAc-
hexane) to give compound 16 as a yellow solid (279 mg, 66%).
1H NMR (CDCl3): δ 7.21 (m, 1H), 6.82 (m, 1H), 4.17 (m, 4H),
2.99 (d, 2H, J = 7.0 Hz), 1.93 (m, 1H), 1.38 (t, 6H, J = 7.2 Hz),
0.99 (d, 6H, J = 7.0 Hz).
2-Carbamoyl-4-[2-(5-diethylphosphono)furanyl]-5-isobutyl-
thiazole (19). A suspension of 5c (500 g, 1.65 mol) and copper(II)
bromide (838 g, 3.76 mol) in anhydrous ethanol (5.5 L) was
heated to reflux under nitrogen for 1 h. The reaction was then
cooled to room temperature, filtered, and the solid washed with
ethanol (0.5 L). The combined filtrate and washing was con-
centrated under reduced pressure to an oil, which was dissolved
in ethyl acetate (4 L) and quenched with a saturated sodium
bicarbonate solution (2.8 L). The resulting mixture was filtered
through a Celite pad (washed with ethyl acetate, 0.5 L), and the
layers were separated. The aqueous phase was extracted with

162 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
(4.4 g, 68%). 1H NMR (200 MHz, CDCl3): δ 7.72-7.77
(m, 1 H), 7.08-7.13 (m, 1 H), 4.10-4.27 (m, 4 H), 1.32-1.40
(m, 6 H).
2-Bromo-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30g).
1H NMR (200 MHz, CD3OD) δ 1.00 (d, 6H, J = 7.0 Hz),
1.82-2.02 (m, 1H), 3.04 (d, 2H, J = 7.0 Hz), 6.82 (m, 1H), 7.17
(m, 1H). Anal. (C11H13NO4PSBr) C, H, N.
General Procedure for Suzuki Couplings of Thiazole 23. 2-
Amino-4-[2-(5-diethylphosphono)furanyl]-5-phenylthiazole (25).
Compound 23 (388 mg, 1 mmol), tetrakis(triphenylphosphine)
palladium (116 mg, 0.1 mmol), and phenyl boronic acid (247 mg,
2 mmol) were suspended in DME-MeOH (5:1, 5 mL) and
treated with a solution of saturated sodium carbonate (0.8 mL).
After heating at 80 ꢀC for 8 h, the reaction mixture was cooled,
diluted with dichloromethane (20 mL), dried (MgSO4), and fil-
tered through a Celite pad (washed with dichloromethane, 3 ꢀ
5 mL). The filtrate was evaporated to dryness, and the resulting
brown solid was crystallized in ethyl acetate and hexane to give
25 as a white solid (77 mg, 20%). White flakes (EtOAc-hexane).
Melting point 198-200 ꢀC. 1H NMR (CDCl3): δ 7.42-7.35
(m, 5H), 7.03 (m, 1H), 6.64 (bs, 2H), 6.39 (m, 1H), 4.15 (m, 4H),
1.25 (t, 6H, J = 7.2 Hz). [MH]þ calcd for C17H19N2O4PS, 379;
found, 379. Anal. Calcd for C17H19N2O4PS: C, 53.96; H, 5.06;
N, 7.40. Found: C, 54.06; H, 4.88; N, 7.33.
2-Amino-4-[2-(5-diethylphosphono)furanyl]-5-morpholinyl-
thiazole (26).Compound 23 (168 mg, 0.44 mmol), Pd2(dba)3 (20mg,
0.022 mmol), morpholine (77 mg, 0.88 mmol), 2-(di-tert-butylphos-
phino)biphenyl (3 mg, 0.01 mmol), and tBuOK (1M, 0.97 mL) in
DMF (2 mL) were heated at 80 ꢀC for 8 h under nitrogen. The
cooled reaction mixture was diluted with phosphate buffer (pH = 7)
and extracted with dichloromethane (3 ꢀ 30 mL). The combined
organic extracts were dried (MgSO4) and evaporated to dryness.
The residue was purified by preparative TLC (100% EtOAc) to give
26 as a yellow solid (25 mg, 15%). 1H NMR (200 MHz, CDCl3): δ
7.13-7.18 (m, 1H), 6.89-6.94 (m, 1H), 4.02-4.30 (m, 4H), 3.80-
3.88 (m, 4H), 2.81-2.89 (m, 4H), 1.29-1.39 (m, 6H).
General Procedures for TMSBr-Mediated Removal of Phos-
phonate Diesters. 2-Methyl-5-isobutyl-4-[2-(5-phosphono)furanyl]-
thiazole (30a). A solution of 2-methyl-5-isobutyl-4-[2-(5-diethylphos-
phono)furanyl]thiazole (1.31 g, 3.68 mmol) in anhydrous dichloro-
methane (18 mL) was treated with TMSBr (6.48 g, 36.8 mmol) at
room temperature. After stirring for 16 h, the reaction solution was
evaporated to dryness and the residue was azeotroped with methanol
(3 ꢀ 10 mL). The residue was suspended in water (20 mL). The
resulting solid was collected via filtration (washed with water, 3 ꢀ
10 mL; dichloromethane, 2 ꢀ 5 mL) and dried under vacuum to give
compound 30a as an off-white solid (1.02 g, 92%). 1H NMR
(200 MHz, CD3OD) δ0.98 (d, 6H, J= 6.8 Hz), 1.80-2.00(m,1H),
3.04 (d, 2H, J = 6.8 Hz), 6.78 (m, 1H), 7.08 (m, 1H). Anal.
2-Methylthio-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30h). 1H NMR (200 MHz, CD3OD) δ 0.99 (d, 6H, J = 7.0 Hz),
1.84-2.02 (m, 1H), 2.70 (s, 3H), 3.02 (d, 2H, J = 7.0 Hz), 6.80
(m, 1H), 7.08 (m, 1H). Anal. (C12H16NO4PS2) C, H, N.
2-Cyano-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30i). 1H NMR (200 MHz, CD3OD) δ 1.01 (d, 6H, J = 6.6 Hz),
1.94-2.04 (m, 1H), 3.18 (d, 2H, J = 6.6 Hz), 6.93 (m, 1H), 7.09
(m, 1H). Anal. (C12H13N2O4SP 0.09HBr) C, H, N.
3
2-Methylamino-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30k). Melting point 202-205 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 0.91 (d, 6H), 1.64-1.82 (m, 1H), 2.80 (m, 5H), 6.51 (m, 1H),
6.79 (brs, 1H). Anal. (C12H17N2O4PS 0.5H2O) C, H, N.
3
2-Acetamido-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30l). Melting point 179-181 ꢀC. 1H NMR (200 MHz, D2O þ
NaOD) δ 0.76 (d, 6H), 1.64-1.80 (m, 1H), 1.89 (s, 3H), 2.59 (d,
2H), 6.36 (brs, 1H), 6.50 (brs, 1H). Anal. (C13H17N2O5PS
0.25H2O) C, H, N.
3
2-Carbamoyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30m). As a solid. Melting point 185-186 ꢀC. 1H NMR (200 MHz,
D2O) δ 0.78 (d, 6H, J = 6.9 Hz), 1.67-1.86 (m, 1H), 3.78 (m, 2H),
6.54 (brs, 2H). Anal. (C12H15N2O5PS) C, H, N.
2-Thiocarbamoyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30n). 1H NMR (200 MHz, CD3OD) δ 1.03 (d, 6H), 1.94-2.08
(m, 1H), 3.08 (d, 2H), 6.82(m, 1H), 6.95 (m, 1H). Anal.
(C12H15N2O4PS2 0.1HBr 0.3EtOAc) C, H, N.
3
3
2-Phenyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30o).
1H NMR (200 MHz, CD3OD) δ 1.03 (d, J = 6.6 Hz, 6H),
1.94-214 (m, 1H), 3.11 (d, J = 7.0 Hz, 2H), 6.93 (m, 1H), 7.12
(m, 1H), 7.47 (m, 3H), 7.96 (m, 2H). Anal. (C17H18NO4PS
1HBr) C, H, N.
3
2-(2-Thienyl)-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30p). 1H NMR (200 MHz, D2O þ NaOD) δ 0.75 (d, J = 6.6 Hz,
6H), 1.64-1.88 (m, 1H), 2.71 (d, J = 7.2 Hz, 2H), 6.51 (m, 2H),
6.95 (brs, 1H). 7.39 (brs, 1H). Anal. (C15H16NO4PS2 0.75H2O)
3
C, H, N.
2-(3-Pyridyl)-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30q). 1H NMR (200 MHz, D2O þ NaOD) δ 0.72 (brs, 6H), 1.67
(brs, 1H), 2.64 (brs, 2H), 6.52 (m, 2H), 7.25 (brs, 1H), 7.94 (brs,
1H), 8.28 (brs, 1H), 8.65 (brs, 1H). Anal. (C16H17N2O4PS
3.75HBr) C, H, N.
2-Amino-5-methyl-4-[2-(5-phosphono)furanyl]thiazole (31a).
3
Melting point 200-220 ꢀC. 1H NMR (200 MHz, DMSO-d6) δ
(C12H16NO4PS 0.25CH2C12) C, H, N.
The following compounds were prepared in a similar manner
as compound 30a:
3
2.44 (s, 3H), 6.65 (m, 1H), 6.95 (m, 1H). Anal. (C8H9N2O4PS
0.65HBr) C, H, N.
2-Amino-5-hydroxymethyl-4-[2-(5-phosphono)furanyl]thiazole
3
2-Ethyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30b).
1H NMR (200 MHz, CD3OD) δ 0.96 (d, 6H, J = 6.6 Hz), 1.41
(t, 3H, J = 7.1 Hz), 1.84-2.02 (m, 1H), 3.04-3.20 (m, 4H), 6.87
1
(31b). Melting point 167 ꢀC (decomp). H NMR (200 MHz,
DMSO-d6) δ 4.72 (s, 2H), 5.71 (s, 1H), 6.49 (m, 1H), 6.81
(m, 1H), 7.02 (bs, 2H). Anal. (C8H9N2O5PS) C, H, N.
2-Amino-5-propyl-4-[2-(5-phosphono)furanyl]thiazole (31c).
(m, 1H), 7.15 (m, 1H). Anal. (C13H18NO4PS 1HBr) C, H, N.
3
2-Vinyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30c).
1H NMR (200 MHz, CD3OD) δ 0.9p (d, 6H, J = 6.8 Hz), 1.94
(m, 1H), 3.04 (d, 2H, J = 6.8 Hz), 3.30-3.80 (m, 2H), 5.65
1
Melting point 235-237 ꢀC. H NMR (200 MHz, D2O) δ 0.76
(t, 3H, J = 6.6 Hz), 1.38 (m, 2H), 2.72 (t, 2H, J = 6.6 Hz), 6.38
(brs, 1H), 6.55 (brs, 1H). Anal. (C10H13N2O4PS 0.3H2O) C, H, N.
3
(m, 1H), 6.82 (m, 1H), 7.11 (m, 1H). Anal. (C13H16NO4PS
3
2-Amino-5-isopropyl-4-[2-(5-phosphono)furanyl]thiazole (31d).
1H NMR (200 MHz, CD3OD) δ 1.37 (d, 6H), 3.58-3.78 (m,
1HBr 0.1H2O) C, H, N.
3
2-Hydroxymethyl-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole
(30d). Melting point 160-170 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 0.89 (d, 6H, J = 6.7 Hz), 1.72-1.80 (m, 1H), 2.93 (d, 2H,
J = 7.1 Hz), 4.64 (s, 2H), 6.87 (m, 1H), 6.91 (m, 2H). Anal.
1H), 6.84 (m, 1H), 7.10 (m, 1H). Anal. (C10H13N2O4PS 1HBr)
C, H, N.
2-Amino-5-(2,2,2-trifluoroethyl)-4-[2-(5-phosphono)furanyl]-
3
1
thiazole (31e). H NMR (200 MHz, DMSO-d6) δ 4.01 (q, 2H,
J = 7.6 Hz), 6.59 (m, 1H), 6.89 (m, 1H), 7.21 (bs, 2H). Anal.
(C9H8N2F3O4PS) C, H, N.
(C12H16NO5PS 0.75HBr) C, H, N.
3
5-Isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30e). Melting
point 164-166 ꢀC. 1H NMR (200 MHz, CDCl3): δ 8.66 (s, 1H),
6.50 (m, 2H), 2.78 (d, 2H, J = 6.6 Hz), 1.75 (m, 1H), 0.74 (d, 6H,
J = 6.6 Hz). Anal. (C11H14NO4PS) C, H, N.
2-Amino-5-neopentyl-4-[2-(5-phosphono)furanyl]thiazole (31f).
Melting point 240-241 ꢀC. 1H NMR (200 MHz, DMSO-d6) δ
0.90 (s, 9H), 2.81 (s, 2H), 6.50 (m, 1H), 6.82 (m, 1H), 6.95 (bs,
2-Chloro-5-isobutyl-4-[2-(5-phosphono)furanyl]thiazole (30f).
1H NMR (200 MHz, CD3OD) δ 0.98 (d, 6H, J = 7.1 Hz),
1.84-2.02 (m, 1H), 3.04 (d, 2H, J = 7.1 Hz), 6.81 (m, 1H), 7.18
(m, 1H). Anal. (C11H13NO4PSCl) C, H, N.
2H). Anal. (C12H12N2O4PS 0.2H2O) C, H, N.
2-Amino-5-cyclobutyl-4-[2-(5-phosphono)furanyl]thiazole (31g).
3
1H NMR (200 MHz, DMSO-d6) δ 1.72-2.42 (m, 6H), 4.01

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 163
(m, 1H), 6.48 (m, 1H), 6.85 (m, 1H), 7.05 (bs, 2H). Anal.
(C11H13N2O4PS 0.15HBr 0.15H2O) C, H, N.
2-Amino-5-cyclopentyl-4-[2-(5-phosphono)furanyl]thiazole (31h).
2-Amino-5-(2-methoxyphenyl)-4-[2-(5-phosphono)furanyl]-
thiazole (33b). Melting point 198-202 ꢀC. 1H NMR (200 MHz,
DMSO-d6) δ 3.65 (s, 3H), 6.01 (m, 1H), 6.74 (m, 1H), 6.90-7.40
3
3
1H NMR (200 MHz, DMSO-d6) δ 1.32-2.22 (m, 8H), 3.61 (m,
(m, 6H). Anal. (C14H13N2O5PS 0.24HBr) C, H, N.
3
1H), 6.47 (m, 1H), 6.95 (m, 3H). Anal. (C12H15N2O4PS
0.25HBr 0.2acetone) C, H, N.
2-Amino-5-cyclohexyl-4-[2-(5-phosphono)furanyl]thiazole (31i).
2-Amino-5-(3-methoxyphenyl)-4-[2-(5-phosphono)furanyl]-
thiazole (33c). Melting point 188-195 ꢀC. 1H NMR (200 MHz,
DMSO-d6) δ 3.71 (s, 3H), 6.31 (m, 1H), 6.80-7.30 (m, 7H).
3
3
1H NMR (200 MHz, D2O) δ 1.02-1.79 (m, 10H), 2.94 (m, 1H),
Anal. (C14H13N2O5PS 0.4HBr) C, H, N.
3
6.32 (m, 1H), 6.52 (m, 1H). Anal. (C13H17N2O4PS 0.2HBr)
C, H, N.
2-Amino-5-(4-methoxyphenyl)-4-[2-(5-phosphono)furanyl]-
thiazole (33d). Melting point 190-210 ꢀC. 1H NMR (200 MHz,
DMSO-d6) δ 3.78 (s, 3H), 6.21 (m, 1H), 6.79 (m, 1H), 6.84-7.31
3
2-Amino-5-cyclopropylmethyl-4-[2-(5-phosphono)furanyl]-
thiazole (31j). 1H NMR (200 MHz, DMSO-d6) δ 0.21 (m, 2H),
0.43 (m, 2H), 0.97 (m, 1H), 2.81 (d, 2H, J = 6.6 Hz), 6.51 (m,
(m, 6H). Anal. (C14H13N2O5PS 1.1H2O) C, H, N.
3
2-Amino-5-(4-methylthiophenyl)-4-[2-(5-phosphono)furanyl]-
thiazole (33e). Melting point 204-206 ꢀC. 1H NMR (200 MHz,
DMSO-d6) δ 2.45 (s, 3H), 6.31 (m, 1H), 6.81 (m, 1H), 7.20-7.32
1H), 6.91 (m, 3H). Anal. (C11H13N2O4PS 0.2EtOAc) C, H, N.
3
2-Amino-5-cyclopentylmethyl-4-[2-(5-phosphono)furanyl]-
thiazole (31k). 1H NMR (200 MHz, DMSO-d6) δ 1.13-2.13 (m,
9H), 2.80 (d, 2H, J = 7.0 Hz), 6.50 (m, 1H), 6.87 (m, 1H), 6.95
(bs, 2H). Anal. (C13H17N2O4PS 0.2HBr 0.2H2O) C, H, N.
(m, 6H). Anal. (C14H13N2O4PS2 1H2O) C, H, N.
3
2-Amino-5-(4-tert-butylphenyl)-4-[2-(5-phosphono)furanyl]-
thiazole (33f). Melting point 201-204 ꢀC. 1H NMR (200 MHz,
DMSO-d6) δ 1.25 (s, 9H), 6.26 (m, 1H), 6.81 (m, 1H), 7.20-7.40
3
3
2-Amino-5-cyclohexylmethyl-4-[2-(5-phosphono)furanyl]-
thiazole (31l). 1H NMR (200 MHz, DMSO-d6) δ 0.83-1.73 (m,
11H), 2.78 (d, 2H, J = 7.0 Hz), 6.49 (m, 1H), 6.87 (m, 1H), 6.94
(bs, 2H). Anal. (C14H19N2O4PS 0.2HBr 0.5H2O 0.1acetone)
(m, 6H). Anal. (C17H19N2O4PS 0.3HBr) C, H, N.
3
2-Amino-5-(4-methoxycarbonylphenyl)-4-[2-(5-phosphono)-
furanyl]thiazole (33g). Melting point 190-210 ꢀC. 1H NMR
(200 MHz, DMSO-d6) δ 3.82 (s, 3H), 6.46 (m, 1H), 6.84 (m, 1H),
3
3
3
C, H, N.
2-Amino-5-benzyl-4-[2-(5-phosphono)furanyl]thiazole (31m).
7.40-7.98 (m, 6H). Anal. (C15H13N2O6PS 0.2HBr) C, H, N.
3
1H NMR (200 MHz, DMSO-d6) δ 4.23 (s, 2H), 6.52 (m, 1H),
6.89 (m, 1H), 6.98 (bs, 2H), 7.26 (m, 5H). Anal. (C14H13N2O4PS
1H2O) C, H, N.
2-Amino-5-(4-fluorophenyl)-4-[2-(5-phosphono)furanyl]thiazole
(33h). Melting point 211-213 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 6.30 (m, 1H), 6.81 (m, 1H) 7.12-7.40 (m, 6H). Anal.
3
2-Amino-5-[(4-morpholinyl)methyl]-4-[2-(5-phosphono)furanyl]-
thiazole dihydrobromide (31n). 1H NMR (200 MHz, CD3OD) δ
3.37-4.04 (m, 8H), 4.82 (s, 2H), 7.18 (m, 2H). Anal. (C12H18Br2-
(C13H10N2O4PSF 0.5H2O) C, H, N.
3
2-Amino-5-(4-chlorophenyl)-4-[2-(5-phosphono)furanyl]thiazole
(33i). Melting point 204-208 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 6.39 (m, 1H), 6.81 (m, 1H), 7.22-7.40 (m, 6H). Anal.
N3O5PS 0.25HBr) C, H, N.
3
2-Amino-5-chloro-4-[2-(5-phosphono)furanyl]thiazole (32a).
Melting point >250 ꢀC. 1H NMR (200 MHz, DMSO-d6) δ 6.81
(m, 1H), 6.99 (m, 1H), 7,55 (bs, 2H). Anal. (C7H6N2O4PSCl)
C, H, N.
(C11H9N2O5PS 0.6H2O) C, H, N.
3
2-Amino-5-(4-acetylphenyl)-4-[2-(5-phosphono)furanyl]thiazole
(33j). Melting point 193-195 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 2.52 (s, 3H), 6.45 (m, 1H), 6.84 (m, 1H), 7.40-7.98 (m, 6H).
Anal. (C15H13N2O5PS) C, H, N.
2-Amino-5-bromo-4-[2-(5-phosphono)furanyl]thiazole (32b).
Melting point 224 ꢀC (decomp). 1H NMR (200 MHz, DMSO-
d6) δ 6.83 (m, 1H), 6.92 (m, 1H), 7.43 (bs, 2H). Anal. (C7H6N2-
O4PSBr) C, H, N.
2-Amino-5-(4-methanesulfonyl)-4-[2-(5-phosphono)furanyl]thiazole
(33k). Melting point 220-225 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 3.20 (s, 3H), 6.50 (m, 1H), 6.84 (m, 1H), 7.40- 7.86 (m, 6H).
2-Amino-5-iodo-4-[2-(5-phosphono)furanyl]thiazole (32c). 1H
NMR (200 MHz, DMSO-d6) δ 6.65 (m, 1H), 6.91(m, 3H). Anal.
(C7H6N2O4PSI) C, H, N.
Anal. (C14H13N2O6PS2 0.65H2O) C, H, N.
3
2-Amino-5-(4-phenylphenyl)-4-[2-(5-phosphono)furanyl]thiazole
(33l). Melting point 201-210 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 6.40 (m, 1H), 6.82 (m, 1H), 7.22 (m, 1H), 7.40-7.68
2-Amino-5-(N-morpholinyl)-4-[2-(5-phosphono)furanyl]thiazole
(32d). Melting point 215-220 ꢀC. 1H NMR (200 MHz, D2O) δ
2.78 (m, 4H), 3.74 (m, 4H), 6.52 (m, 1H), 6.62 (m, 1H). Anal.
(m, 11H). Anal. (C19H15N2O4PS 0.15HBr) C, H, N.
3
2-Amino-5-(2-naphthyl)-4-[2-(5-phosphono)furanyl]thiazole
(33m). Melting point 205-212 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 6.36 (m, 1H), 6.81 (m, 1H), 7.29 (bs, 2H), 7.40-7.98 (m,
7H). Anal. (C17H13N2O4PS 0.6HBr 0.7acetone) C, H, N.
(C11H14N3O5PS 0.45CH2Cl2) C, H, N.
3
2-Amino-5-ethylthio-4-[2-(5-phosphono)furanyl]thiazole (32e).
1H NMR (200 MHz, DMSO-d6) δ 1.15 (t, 3H, J = 7.4 Hz), 2.73
(q, 2H, J = 7.4 Hz), 6.91 (bs, 2H), 6.95 (m, 1H), 7.39 (m, 1H).
3
3
2-Amino-5-(2-furanyl)-4-[2-(5-phosphono)furanyl]thiazole
(33n). Melting point 190-210 ꢀC. 1H NMR (200 MHz, DMSO-
d6) δ 6.56 (m, 1H), 6.64 (m, 1H), 6.93 (m, 1H), 7.02 (m, 1H), 7.40
Anal. (C9H11N2O4PS2 0.45HBr) C, H, N.
3
2-Amino-5-isopropylthio-4-[2-(5-phosphono)furanyl]thiazole
hydrobromide (32g). 1H NMR (200 MHz, DMSO-d6) δ 1.18 (d,
6H, J = 6.6 Hz), 3.18 (m, 1H), 4.34 (bs, 2H), 6.92 (m, 1H), 7.02
(m, 1H). Anal. (C10H14N2O4PS2Br) C, H, N.
(brs, 2H), 7.68 (s, 1H). Anal. (C11H9N2O5PS 0.25HBr) C, H, N.
2-Amino-5-(2-thienyl)-4-[2-(5-phosphono)furanyl]thiazole
3
(33o). 1H NMR (200 MHz, D2O) δ 6.20 (m, 1H), 6.40 (m, 1H),
6.93 (m, 1H), 7.04 (m, 1H), 7.30 (d, 1H). Anal. (C11H9N2O4PS2
2-Amino-5-tert-butylthio-4-[2-(5-phosphono)furanyl]thiazole
3
(32h). 1H NMR (200 MHz, DMSO-d6) δ 1.25 (s, 9H), 6.85 (m,
1H), 7.12 (m, 1H), 7.41 (bs, 2H). Anal. (C11H15N2O4PS2
0.6CH2Cl2) C, H, N.
0.3EtOAc 0.11HBr) C, H, N.
3
2-Amino-5-phenyl-4-[3-(1-phosphono)pyrrolyl]thiazole (34b).
3
1
Melting point >200 ꢀC. H NMR (200 MHz, CDCl3) δ 8.0
(s, 1H), 7.4 (m, 5H), 6.79 (s, 1H), 5.60 (s, 1H). LC-MS m/z =
2-Amino-5-phenylthio-4-[2-(5-phosphono)furanyl]thiazole (32i).
Melting point 157 ꢀC (decomp). 1H NMR (200 MHz, DMSO-
d6) δ 6.86 (m, 1H), 7.18-7.39 (m, 6H), 7.65 (bs, 2H). Anal.
(C13H11N2O4PS2) C, H, N.
322.3 [C13H12N3O3PS þ H]þ. Anal. (C13H12N3O3PS 1.3H2O)
3
C, H, N.
2-Amino-5-phenyl-4-[2-(5-methyl-4-phosphono)oxazolyl]-
thiazole (34c). Melting point 190 ꢀC (decomp). 1H NMR (200 MHz,
2-Amino-5-[(N,N-dimethyl)carbamoyl]-4-[2-(5-phosphono)-
1
CD3OD) δ 2.39 (s, 3H), 7.39 (m, 5H). Anal. (C13H12N3O4PS
1H2O) C, H, N.
furanyl]thiazole (32j). H NMR (200 MHz, DMSO-d6) δ 2.06
(s, 6H), 5.81 (bs, 2H), 6.58 (m, 1H), 6.92 (m, 1H). Anal.
(C10H14N3O5PS 1.3HBr 1H2O 0.3acetone) C, H, N.
3
2-Amino-5-propyl-4-phosphonomethoxycarbonylthiazole (34d).
Melting point 134 ꢀC (decomp). 1H NMR (200 MHz, DMSO-
d6) δ 0.88 (t, 3H, J = 7.4 Hz), 1.54 (m, 2H), 2.98 (t, 2H, J =
7.4 Hz), 4.25 (d, 2H, J = 8.8 Hz), 7.02 (bs, 2H). Anal.
(C8H13N2O5PS) C, H, N.
3
3
3
2-Amino-5-benzyloxycarbonyl-4-[2-(5-phosphono)furanyl]-
thiazole (32l). Melting point 248 ꢀC (decomp). 1HNMR(200MHz,
DMSO-d6) δ 5.22 (s, 2H), 6.94 (m, 1H), 7.36 (m, 5H), 7.54 (m, 1H),
8.02 (bs, 2H). Anal. (C10H14N3O5PS 0.2H2O) C, H, N.
3

164 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Dang et al.
2-Amino-5-(2-thienyl)-4-[(N-phosphonomethyl)carbamoyl]-
thiazole (34e). Melting point 245 ꢀC(decomp). 1H NMR (200 MHz,
D2O) δ3.18 (d, 2H, J= 13 Hz), 6.90 (m, 1H), 7.09 (m, 1H), 7.31 (m,
1H). Anal. (C9H10N3O4PS2 1HBr 0.1EtOAc) C, H, N.
chloride (816 mg, 6.85 mmol) followed by anhydrous pyridine
(49 mg, 0.62 mmol). The resulting mixture was heated to reflux
for 2 h, cooled, and evaporated to dryness. The residue was
treated with a solution of 4-acetoxybenzyl alcohol (665 mg,
4 mmol) and N,N-diisopropylethylamine (1 mL, 6 mmol) in
anhydrous CH2Cl2 (5 mL) at room temperature. After stirring
at room temperature for 12 h, the reaction was quenched with
saturated NaHCO3 (20 mL) and extracted with CH2Cl2 (3 ꢀ
20 mL). The organics were combined, dried (MgSO4), and evap-
orated, and the resulting residue purified by flash chromatog-
raphy (SiO2, 1 cm ꢀ 15 cm, 65-100% EtOAc-hexane, gradient
elution) to give 35f as a brown foam (60 mg, 10%). 1H NMR
(CDCl3): δ 7.34 (d, 2H, J = 8.7 Hz), 7.14 (m, 1H), 7.05 (d, 2H,
J = 9.0 Hz), 6.58 (m, 1H), 5.70 (bs, 2H), 5.16-5.01 (m, 4H),
2.72 (d, 2H, J = 6.9 Hz), 2.29 (s, 6H), 1.81 (m, 1H), 0.91 (m,
6H). [MH]þ calcd for C29H31N2O8PS 599; found 599. Anal.
(C29H31N2O8PS) C, H, N.
2-Amino-4-[2-(5-diphenylphosphono)furanyl]-5-isobutylthiazole
(35g). This compound was prepared in a similar manner as com-
pound 35f from thiazole 30j (500 mg, 1.65 mmol) to give 35g as a
light-yellow solid (120 mg, 16%). Melting point 128-129 ꢀC. 1H
NMR (CDCl3): δ 7.38-7.18 (m, 11H), 6.63 (m, 1H), 5.51
(bs, 2H), 2.79 (d, 2H, J = 6.6 Hz), 1.83 (m, 1H), 0.97 (d, 6H,
J = 6.6 Hz). Anal. (C23H23N2O4PS) C, H, N.
3
3
2-Amino-4-[2-(6-phosphono)pyridyl]thiazole (34f). 1H NMR
(200 MHz, D2O) δ 7.09-8.40 (m, 4H). Anal. (C8H9N3O3PS2
1HBr) C, H, N.
3
2-Amino-4-[1-(3-phosphono)phenyl]thiazole (34g). Melting point
210-214 ꢀC. 1H NMR (200 MHz, DMSO-d6) δ 7.19-8.04
(m, 7H). Anal. (C9H9N2O3PS 0.8HBr 0.4CH2Cl2) C, H, N.
3
3
2-Amino-5-propyl-4-[1-(3-phosphono)phenyl]thiazole (34h). Melt-
ing point >220 ꢀC. 1H NMR (200 MHz, DMSO-d6) δ 0.92 (t, 3H,
J = 7.4 Hz), 1.57 (m, 2H, J = 7.4 Hz), 2.69 (t, 2H, J = 7.4 Hz),
6.92 (bs, 2H), 7.42-7.93 (m, 4H). Anal. (C12H15N2O3PS 0.09H2O)
C, H, N.
3
2-Amino-5-propyl- 4-[1-(4-methyl-3-phosphono)phenyl]thiazole
(34i). Melting point >220 ꢀC. 1H NMR (200 MHz, DMSO-d6)
δ 0.89 (t, 3H, J = 7.0 Hz), 1.5 (m, 2H), 2.49 (s, 3H), 2.65 (t, 2H,
J = 7.0 Hz), 6.96 (bs, 2H), 7.25-8.00 (m, 3H). Anal.
(C13H17N2O3PS 0.3H2O) C, H, N.
3
2-Amino-5-isopropyl- 4-[1-(4-methoxy-3-phosphono)phenyl]-
thiazole (34j). Melting point >220 ꢀC. 1H NMR (200 MHz, D2O)
δ 1.18 (d, 6H, J = 7.0 Hz), 3.18 (m, 1H), 3.81 (s, 3H), 7.02-7.85
(m, 3H). Anal. (C13H17N2O4PS) C, H, N.
2-Amino-5-phenyl- 4-[1-(4-fluoro-3-phosphono)phenyl]thiazole
(34k). Melting point >220 ꢀC. 1H NMR (200 MHz, DMSO-d6)
2-Amino-4-[2-(5-(1-(3-bromophenyl)-1,3-propyl)phosphono)-
furanyl]-5-isobutyl-thiazole (35h). A suspension 30j (420 mg,
1.5 mmol) in anhydrous dichloromethane (10 mL) was trea-
ted with thionyl chloride (10 mL, 137 mmol). The resulting
mixture was then heated to reflux. After 1.5 h, the reaction solu-
tion was cooled and evaporated to dryness and the residue
was treated with a solution of 1-(3-chlorophenyl)propane-1,3-diol
(252 mg, 1.35 mmol) and anhydrous pyridine (0.35 mL) in
anhydrous CH2Cl2 (10 mL). After stirring at room tempera-
ture for 12 h, the reaction was quenched with saturated NaHCO3
(20 mL) and extracted with CH2Cl2 (3 ꢀ 20 mL). The organics
were combined, dried (MgSO4), and evaporated, and the
resulting residue was purified by flash chromatography (SiO2,
1 cm ꢀ 15 cm, 1, 2, and 3% methanol-CH2Cl2, gradient elution)
to give 35h as a yellow gum (170 mg, 25%). 1H NMR (CDCl3): δ
7.41-7.20 (m, 5H), 6.71 (m, 1H), 5.78 (m, 1H), 5.13 (bs, 2H),
4.90-4.40 (m, 2H), 2.82 (d, 2H, J = 6.7 Hz), 2.60-1.79 (m, 3H),
δ 7.01-7.99 (m, 10H). Anal. (C15H12N2O3PSF 0.09HBr) C, H, N.
3
2-Amino-4-[2-(5-phosphono)furanyl]-5-(propane-1-sulfinyl)-
thiazole (32m). A solution of 32f10 (100 mg, 0.3 mmol) in MeOH
(10 mL) was treated with mCPBA (70 mg, 0.3 mmol), and the
reaction mixture was heated to 50 ꢀC. After stirring for 5 h, the
cooled reaction solution was evaporated in vauuo and the
residue was treated with water (5 mL). The resulting white solid
was removed by filtration, and the filtrate was concentrated to
give a yellow oil, which was purified by HPLC using a 10-70%
acetonitrile gradient containing 0.05% TFA. The pure fractions
were combined and lyophilized to give 32m (50 mg, 50%) as a
white solid. Melting point 120 ꢀC (decomp). 1H NMR (200 MHz,
CD3OD) δ 1.08 (t, 3H, J = 7.0 Hz), 1.79 (m, 2H), 3.15 (m, 2H),
6.89 (m, 1H), 7.08 (m, 1H). Anal. (C10H13N2O5PS2 0.2H2O
3
3
1TFA) C, H, N.
2-Amino-4-{2-[5-bis(1-(1-ethoxycarbonyloxy)ethyl)phosphono]-
0.99 (d, 6H, J = 6.7 Hz). Anal. (C20H22N2O4PSCl 0.25H2O)
3
C, H, N.
furanyl}-5-isobutyl-thiazole (35b). A suspension of 30j (284 mg,
0.94 mmol) in anhydrous DMPU (5 mL) was treated with TEA
(254 mg, 2.5 mmol) and 1-(1-ethoxycarbonyloxy)-iodoethane
(500 mg, 2.0 mmol) at room temperature under nitrogen. After
24 h, the reaction solution was evaporated to dryness and the
residue was purified by flash chromatography (SiO2, 3cm ꢀ 15 cm,
65-100% EtOAc-hexane, gradient elution) to give 35b as a
yellow solid (160 mg, 32%). Melting point 76-78 ꢀC. 1H NMR
(CDCl3): δ 7.25 (m, 1H), 6.62 (m, 1H), 6.45 (m, 2H), 5.19 (bs,
2H), 4.08 (m, 4H), 2.81 (d, 2H, J = 7.1 Hz), 1.85 (m, 1H), 1.63 (d,
6H, J = 7.0 Hz), 1.19 (m, 6H), 0.97 (d, 6H, J = 7.1 Hz). Anal.
(C21H31N2O10PS) C, H, N.
2-Amino-4-{2-[5-bis(phenoxycarbonyloxymethyl)phosphono]-
furanyl}-5-isobutyl-thiazole (35c). This compound was prepared
in a similar manner as compound 35b from thiazole 30j. 1H
NMR (CDCl3): δ 7.40-7.11 (m, 12H), 5.89 (d, 4H, J = 14 Hz),
5.10 (bs, 2H), 2.81 (d, 2H, J = 6.8 Hz), 1.81 (m, 1H), 0.93 (m,
6H). Anal. (C27H27N2O10PS) C, H, N.
2-Amino-4-{2-[5-(1-(4-pyridyl)-1,3-propyl)phosphono]furanyl}-
5-isobutylthiazole (35i). This compound was prepared in a sim-
ilar manner as compound 35h from thiazole 30j (500mg,1.65mmol)
to give 35i as a light-yellow solid (120 mg, 16%). Melting point
101-106 ꢀC. 1H NMR (CDCl3): δ 8.69 (m, 2H), 7.39 (m, 2H),
7.22 (m, 1H), 6.66 (m, 1H), 5.62 (m, 1H), 5.21 (bs, 2H), 4.65
(m, 2H), 2.81 (d, 2H, J = 6.6 Hz), 2.38 (m, 2H), 1.88 (m, 1H),
0.91 (d, 6H, J=6.6Hz).Anal. (C19H22N3O4PS 1.25H2O) C, H, N.
3
2-Amino-4-[2-(5-monophenylphosphono)furanyl]-5-isobutyl-
thiazole (35j). A solution of 35g (1.8 g, 3.96 mmol) in acetonitrile
(60 mL) and water (15 mL) was treated with lithium hydroxide
(1 N, 5.9 mL) at room temperature. After 4 h, the reaction solu-
tion was concentrated under vacuum and diluted with water (80 mL).
The pH of the resulting solution was adjusted to 4 by addition of
6 N HCl. The resulting white solid was collected through filtra-
tion (washed with water, 3 ꢀ 10 mL) and dried under vacuum to
give 35j as a white solid (1.4 g, 93%). Melting point 256 ꢀC
(decomp). 1H NMR (DMSO-d6): δ 7.31-7.01 (m, 6H), 6.59 (m,
1H), 5.80 (bs, 2H), 2.75 (d, 2H, J = 6.6 Hz), 1.73 (m, 1H), 0.82(d,
6H, J = 6.6 Hz). Anal. (C17H19N2O4PS) C, H, N.
2-Amino-4-{2-[5-bis(ethoxycarbonyloxymethyl)phosphono]-
furanyl}-5-isobutyl-thiazole (35d). This compound was prepared
in a similar manner as compound 35b from thiazole 30j. H
1
NMR (CDCl3): δ 7.21 (m, 1H), 6.61 (m, 1H), 5.75 (d, 4H, J =
13.2 Hz), 5.59 (bs, 2H), 4.20 (q, 4H, J = 7.4 Hz), 2.79 (d, 2H, J =
6.6 Hz), 1.89 (m, 1H), 1.275 (t, 6H, J = 7.4 Hz), 0.97 (d, 6H, J =
6.6 Hz). Anal. (C19H27N2O10PS) C, H, N.
2-Amino-4-[2-(5-bis(p-acetoxybenzyl)phosphono)furanyl]-5-
isobutylthiazole (35f). A suspension 30j (302 mg, 1 mmol) in
anhydrous 1,2-dichloroethane (5 mL) was treated with thionyl
2-Amino-4-[2-(5-(O-phenyl)-(N-((S)-1-ethoxycarbonyl)ethyl)-
phosphonamido)-furanyl]-5-isobutylthiazole (35l). A suspension
of 30j (302 mg, 1 mmol) in anhydrous 1,2-dichloroethane (5 mL)
was treated with thionyl chloride (815 mg, 6.85 mmol), followed
by anhydrous pyridine (49 mg, 0.62 mmol). After refluxing for
2 h, the reaction mixture was cooled and evaporated to dryness.
The resulting residue was dissolved in anhydrous CH2Cl2 (2 mL),

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 165
cooled to 0 ꢀC, and treated with a solution of phenol (1 mmol).
After stirring at room temperature for 1 h, the reaction was
treated with a solution of L-alanine ethyl ester hydrogen chloride
salt and N,N-diisopropylethylamine (0.5 mL) in anhydrous CH2Cl2
(2 mL) and stirred an additional 12 h at room temperature. The
reaction was then quenched with saturated NaHCO3 (20 mL)
and extracted with CH2Cl2 (3 ꢀ 20 mL). The combined organic
extracts were dried (MgSO4) and evaporated. The residue was puri-
fied by flash chromatography (SiO2, 1 cm ꢀ 15 cm, 65-100%
EtOAc-hexane, gradient elution) to give 35l as a yellow gum
(6) Dang, Q.; Brown, B. S.; Liu, Y.; Rydzewski, R. M.; Robinson,
E. D.; van Poelje, P. D.; Reddy, R. M.; Erion, M. D. Fructose-1,6-
bisphosphatase Inhibitors. 1. Purine Phosphonic Acids as Novel
AMP Mimics. J. Med. Chem. 2009, 52, 2880–2898.
(7) Dang, Q.; Kasibhatla, S. R.; Xiao, W.; Liu, Y.; DaRe, J.; Taplin,
F.; Reddy, K. R.; Scarlato, G. R.; Gibson, T.; van Poelje, P. D.;
Potter, S. C.; Erion, M. D. Fructose-1,6-bisphosphatase Inhibitors.
2. Design, Synthesis, and Structure-Activity Relationship of a
Series of Phosphonic Acid Containing Benzimidazoles that Func-
tion as 50-Adenosinemonophosphate (AMP) Mimics. J. Med.
Chem. 2010, 53, 441–451.
(8) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Experimental and computational approaches to estimate solubility
and permeability in drug discovery and development settings. Adv.
Drug Delivery Rev. 1997, 23, 3–26.
(9) Erion, M. D.; van Poelje, P. D.; Reddy, M. R. Computer-Assisted
Scanning of Ligand Interactions: Analysis of the Fructose 1,6-
Bisphosphatase-AMP Complex Using Free Energy Calculations.
J. Am. Chem. Soc. 2000, 122, 6114–6115.
(10) Dang, Q.; Kasibhatla, S. R.; Jiang, T.; Fan, K.; Liu, Y.; Taplin, F.;
Schulz, W.; Cashion, D. K.; Reddy, K. R.; van Poelje, P. D.;
Fujitaki, J. M.; Potter, S. C.; Erion, M. D. Discovery of phosphonic
diamide prodrugs and their use for the oral delivery of a series of
fructose 1,6-bisphosphatase inhibitors. J. Med. Chem. 2008, 51,
4331–4339.
(11) Schroder, R.; Schollkopf, U.; Blume, E.; Hoppe, I. In 2-stellung
unsubstituierte oxazole aus alpha-metallierten isocyaniden und
acylierungsreagenzien. Liebigs Ann. Chem. 1975, 533–546.
(12) Harn, N. K.; Gramer, C. J.; Anderson, B. A. Acylation of oxazoles
by the copper-mediated reaction of oxazol-2ylzinc chloride deriv-
atives. Tetrahedron Lett. 1995, 36, 9453–9456.
1
(125 mg, 26%). H NMR (CDCl3): δ 7.24-7.03 (m, 6H), 6.55
(m, 1H), 4.93 (bs, 2H), 4.15-4.01 (m, 3H), 3.77 (q, 1H,, J =
10.2 Hz), 2.77 (d, 2H, J = 6.9 Hz), 1.81 (m, 1H), 1.33-0.89
(m, 12H). [MH]þ calcd for C22H28N3O5PS 478; found 478.
Anal. (C22H28N3O5PS) C, H, N.
2-Amino-4-[2-(5-monophenyl-monoamino-phosphono)furanyl]-
5-isobutylthiazole (35k). This compound was prepared in a sim-
ilar manner as compound 35l from thiazole 30j. Melting point
205 ꢀC (decomp). 1H NMR (200 MHz, DMSO-d6): δ 7.39-6.98
(m, 6H), 6.53 (m, 1H), 5.41 (d, 2H, J = 7.6 Hz), 3.31 (bs, 2H),
2.78 (d, 2H, J = 6.6 Hz), 1.73 (m, 1H), 0.92 (d, 6H, J = 6.6 Hz).
Anal. (C17H20N3O3PS 0.3H2O 0.3HCl) C, H, N.
3
3
Biological Methods. FBPase enzyme assay, oral bioavailabil-
ity determination, and glucose lowering in Sprague-Dawley
rats and Zucker Diabetic rats assays were carried out as previ-
ously reported.3,6,10,19,20
(13) Sandmeyer, T. Ber. Dtsch. Chem. Ges. 1884, 17, 1633–1635.
(14) Galli, C. Substituent effects on the Sandmeyer reaction. Quantita-
tive evidence for rate-determining electron transfer. J. Chem. Soc.,
Perkin Trans. II 1984, 897–902.
(15) Yang, B. H.; Buchwald, S. L. Palladium-catalyzed amination of aryl
halides and sulfonates. J. Organomet. Chem. 1999, 576, 125–146.
(16) Hecker, S. J.; Erion, M. D. Prodrugs of Phosphates and Phospho-
nates. J. Med. Chem. 2008, 51, 2328–2345.
(17) Erion, M. D.; Reddy, K. R.; Boyer, S. H.; Matelich, M. C.; Gomez-
Galeno, J.; Lemus, R. H.; Ugarkar, B. G.; Colby, T. J.; Schanzer,
J.; Van Poelje, P. D. Design, synthesis, and characterization of a
series of cytochrome P(450) 3A-activated prodrugs (HepDirect
prodrugs) useful for targeting phosph(on)ate-based drugs to the
liver. J. Am. Chem. Soc. 2004, 126, 5154–5163.
(18) McGuigan, C.; Nickson, C.; O’Connor, T. J.; Kinchington, D.
Synthesis and anti-HIV activity of some novel lactyl and glycolyl
phosphate derivatives. Antiviral Res. 1992, 17, 197–212.
(19) van Poelje, P. D.; Potter, S. C.; Chandramouli, V. C.; Landau, B. R.;
Dang, Q.; Erion, M. D. Inhibition of fructose 1,6-bisphosphatase
reduces excessive endogenous glucose production and attenuates
hyperglycemia in zucker diabetic Fatty rats. Diabetes 2006, 55,
1747–1754.
(20) van Poelje, P. D.; Potter, S. C.; Topczewski, E.; Hou, J.; Linemeyer,
D. L.; Erion, M. D. Combination therapy with pioglitazone and a
fructose 1,6-bisphosphatase inhibitor (MB06322) improves glyce-
mic control and lactate homeostasis in male Zucker diabetic fatty
(ZDF) rats. Diabetologia 2006, 49 (Suppl 1), Abs 0852.
(21) Yoshida, T.; Okuno, A.; Izumi, M.; Takahashi, K.; Hagisawa, Y.;
Ohsumi, J.; Fujiwara, T. CS-917, a fructose 1,6-bisphosphatase
inhibitor, improves postprandial hyperglycemia after meal loading
in non-obese type 2 diabetic Goto-Kakizaki rats. Eur. J. Pharma-
col. 2008, 601, 192–197.
(22) Triscari, J.; Walker, J.; Feins, K.; Tao, B.; Bruce, S. R. In Multiple
Ascending Dose of CS-917, a Novel Fructose 1,6-bisphosphatase
(FBPase) Inhibitor, in Subjects with Type 2 Diabetes Treated for 14
Days; American Diabetes Association: Alexandria, VA, 2006.
Acknowledgment. We thank Drs. K. Raja Reddy and
M. Rami Reddy for their contributions to this project and
Dr. Scott Hecker for reviewing the manuscript. Metabasis
Therapeutics, Inc. was purchased by Ligand Pharmaceuticals,
Inc., La Jolla, California 92037.
Supporting Information Available: Experimental procedures
for the preparation of compounds 7-12, 27-28, and elemental
analysis data for all final compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Van Poelje, P. D.; Dang, Q.; Erion, M. D. Fructose 1,6-bisphos-
phatase as a therapeutic target for type 2 diabetes. Drug Discovery
Today: Ther. Strategies 2007, 4, 103–109.
(2) Van Poelje, P. D.; Dang, Q.; Erion, M. D. Discovery of fructose
1,6-bisphosphatase inhibitors for the treatment of type 2 diabetes.
Curr. Opin. Drug Discovery Dev. 2007, 10, 430–437.
(3) Erion, M. D.; van Poelje, P. D.; Dang, Q.; Kasibhatla, S. R.; Potter,
S. C.; Reddy, M. R.; Reddy, K. R.; Jiang, T.; Lipscomb, W. N.
MB06322 (CS-917): a potent and selective inhibitor of fructose 1,6-
bisphosphatase for controlling gluconeogenesis in type 2 diabetes.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 7970–7975.
(4) Erion, M. D.; Dang, Q.; Reddy, M. R.; Kasibhatla, S. R.; Huang,
J.; Lipscomb, W. N.; van Poelje, P. D. Structure-Guided Design of
AMP Mimics that Inhibit Fructose 1,6-Bisphosphatase with High
Affinity and Specificity. J. Am. Chem. Soc. 2007, 129, 15480–15490.
(5) Dang, Q.; Kasibhatla, S. R.; Reddy, K. R.; Jiang, T.; Reddy, M. R.;
Potter, S. C.; Fujitaki, J. M.; van Poelje, P. D.; Huang, J.;
Lipscomb, W. N.; Erion, M. D. Discovery of potent and specific
fructose-1,6-bisphosphatase inhibitors and a series of orally bio-
available phosphoramidase-sensitive prodrugs for the treatment of
type 2 diabetes. J. Am. Chem. Soc. 2007, 129, 15491–15502.