Discovery of [(3-bromo-7-cyano-2-naphthyl)(difluoro)methyl]phosphonic acid, a potent and orally active small molecule PTP1B inhibitor

Article abstract of DOI:10.1016/j.bmcl.2008.04.064

A series of quinoline/naphthalene-difluoromethylphosphonates were prepared and were found to be potent PTP1B inhibitors. Most of these compounds bearing polar functionalities or large lipophilic residues did not show appreciable oral bioavailability in rodents while small and less polar analogs displayed moderate to good oral bioavailability. The title compound was found to have the best overall potency and pharmacokinetic profile and was found to be efficacious in animal models of diabetes and cancer.

Full text of DOI:10.1016/j.bmcl.2008.04.064

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3200–3205
Discovery of [(3-bromo-7-cyano-2-naphthyl)(difluoro)-
methyl]phosphonic acid, a potent and orally active small
molecule PTP1B inhibitor
Yongxin Han,a,* Michel Belley,a Christopher I. Bayly,a John Colucci,a Claude Dufresne,a
Andre Giroux,a Cheuk K. Lau,a Yves Leblanc,a Daniel McKay,a Michel Therien,a
Marie-Claire Wilson,a Kathryn Skorey,b Chi-Chung Chan,c
Giovana Scapind and Brian P. Kennedyb
aDepartment of Medicinal Chemistry, Merck Frosst Centre for Therapeutic Research, Merck Frosst Canada Ltd, PO Box 1005,
Pointe-Claire-Dorval, Que., Canada H9R 4P8
bDepartment of Biochemistry & Molecular Biology, Merck Frosst Centre for Therapeutic Research, Merck Frosst Canada Ltd,
PO Box 1005, Pointe-Claire-Dorval, Que., Canada H9R 4P8
cDepartment of Pharmacology, Merck Frosst Centre for Therapeutic Research, Merck Frosst Canada Ltd, PO Box 1005,
Pointe-Claire-Dorval, Que., Canada H9R 4P8
dDepartment of Structural Biology, Merck & Co. Inc., 126 E. Lincoln Avenue, PO Box 2000, Rahway, NJ 07065-0900, USA
Received 5 March 2008; revised 22 April 2008; accepted 24 April 2008
Available online 29 April 2008
Abstract—A series of quinoline/naphthalene-difluoromethylphosphonates were prepared and were found to be potent PTP1B inhib-
itors. Most of these compounds bearing polar functionalities or large lipophilic residues did not show appreciable oral bioavailabil-
ity in rodents while small and less polar analogs displayed moderate to good oral bioavailability. The title compound was found to
have the best overall potency and pharmacokinetic profile and was found to be efficacious in animal models of diabetes and cancer.
ꢀ 2008 Elsevier Ltd. All rights reserved.
Type-2 diabetes and obesity are reaching epidemic pro-
portions and thus are becoming the leading causes of
health care burdens around the world. One of the hall-
marks of these diseases is insulin resistance, an attenuated
response of insulin to regulate glucose homeostasis. Pro-
tein tyrosine phosphatase 1B (PTP1B), a negative insulin
regulator, has been shown to play a role in developing
insulin resistance and obesity. Mice lacking the PTP1B
gene are resistant to diet-induced obesity and insulin resis-
tance.1 These findings are further corroborated by treat-
ing mice with PTP1B antisense oligonucleotides.2 In
addition, recent studies have shown that PTP1B also
plays a role in tumorigenesis.3 As a result, PTP1B inhibi-
tors represent attractive pharmaceutical agents for treat-
ing type-2 diabetes, obesity, and cancer. Several
approaches to find suitable PTP1B inhibitors have been
reviewed recently.4 A major drawback of most of these
inhibitors is the lack ofsufficient cell permeability andoral
bioavailability due to the presence of highly negative
charged polar residues in these inhibitors.5 Some non-
phosphonate inhibitors were recently shown to have off-
target activity which complicated efficacy data interpreta-
tion.6 Using oral bioavailability in rodents as a key
parameter, we optimized
a
series of dif-
luoromethylphosphonate PTP1B inhibitors with either
a naphthalene or a quinoline template and discovered
[(3-bromo-7-cyano-2-naphthyl)-(difluoro)methyl]phos-
phonic acid (3g) as a potent PTP1B inhibitor with a good
pharmacokinetic profile. We report herein the SAR that
led to the discovery of 3g, its synthesis, pharmacokinetic
profile in preclinical species, and its in vivo efficacies in
mouse models of diabetes and cancer.
Keywords: Quinoline/naphthalene-difluoromethylphosphonate; PTP1B
inhibitors; Diabetes; Obesity; Cancer.
Potent PTP1B inhibitors such as 1 (Fig. 1 and Table 1)
containing the difluoromethylphosphonate moiety have
*
Corresponding author. Tel.: +1 5144283301; fax: +1 5144284900;
e-mail: yongxin_han@merck.com
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.04.064

Y. Han et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3200–3205
3201
X
X
X
P
Br
CN
R
F
F
O
F
F
O
F
F
O
F
N
R
R
F
P
OH
OH
P
OH
OH
OH
OH
P
OH
OH
O
4
3
1
2
Figure 1. Structures of PTP1B inhibitors 1, 2, 3, and 4.
been reported previously.7 Most of these compounds,
however, exhibited poor cell permeability and poor
pharmacokinetic profile.5 Our current investigation
started with the small quinoline difluoromethylphospho-
nate 2a which did not display appreciable inhibitory
activity (IC50 > 50 lM) against PTP1B in the fluorescein
diphosphate (FDP) assay8 (Table 1). Incorporation of
an ortho-Br moiety gave 2b with improved potency
(IC50 = 14 lM). This observation was consistent with
previous findings as shown in Table 1, 1b versus 1c.7b
As a result, we focused our current effort on compounds
bearing an ortho-Br moiety as shown in Table 1, 2c–i.
ble oral absorption albeit with low plasma drug expo-
sure (Cmax = 0.3 lM at 10 mg/kg in C57BL/6 mice)
compared to compound 2b, again likely due to increased
polarity.
The synthesis of quinoline derivatives 2b–i is outlined in
Scheme 1. 4-Bromo-3-iodoaniline (5)11 was condensed
with crotonaldehyde in the presence of chloranil and
concentrated HCl in refluxing i-PrOH to give 6-bro-
mo-7-iodo-2-methylquinoline (6) after separating from
the region isomer by flash chromatography. The iodide
was reacted with [(diethoxyphosphoryl)(difluoro)-
methyl]zinc bromide using CuBr as the catalyst accord-
ing to literature procedures12 to yield compound 7. Sub-
sequent treatment of 7 with an excess of TMSBr12
followed by quenching with either water or a base gave
the corresponding phosphonate 2b. Compound 7 was
oxidized with SeO2 to the corresponding aldehyde 8.
Reaction of 8 with Wittig reagents ArCH2P+Ph3BrÀ
followed by deprotection of the diethyl phosphonate
moiety with TMSBr as described above gave phospho-
nates 2c–d. Alternatively, 8 was oxidized further to the
acid 9 by AgO in methanol. Reaction of acid 9 with
amines followed by deprotection of the diethyl phospho-
nate with TMSBr gave amides 2e–i.
Compound 2b displayed good pharmacokinetic profile
in rat with marginal oral bioavailability (F = 5%) but
good plasma drug exposure (Cmax = 43.3 lM at 20 mg/
kg) and long plasma elimination half life (t1/2 > 8 h).
This compound, however, lacked the desired inhibitory
activity against the PTP1B enzyme and thus was not a
suitable tool for studying PTP1B inhibition in vivo.
We then attempted to improve potency by exploiting
the secondary binding site.7a,9 From the high resolution
˚
(1.6 A) X-ray structure of the PTP1B-1c complex (Fig.
2, only active site shown),10 it was apparent that the sec-
ondary binding site, as characterized by the positively
charged Arg24 and Arg254 residues, was not exploited.
Docking the quinoline template into the active site sug-
gested that it should be possible to extend to the second-
ary binding site by modifying the methyl group in 2b. A
variety of groups were modeled and the carboxylate in
the 3-vinylbenzoate moiety (2d) was shown to have the
potential to make excellent hydrogen binding interac-
tions with Arg24 and Arg254 (Fig. 3). The styrene deriv-
ative 2c was also prepared as a comparison.
We next turned our attention to the naphthalene ana-
logs since these compounds generally had more favor-
able clogD values. For instance, using the commercial
software from ACD, the clogDs for 2b and 3b were cal-
culated to be À2.28 and À0.8, respectively. Phosphonate
3a, which was reported previously to have a Ki of
179 lM against PTP1B,13 was the starting point for fur-
ther SAR investigation. Incorporation of the ortho-Br
moiety and addition of a methyl group at the 7-position
resulted in a dramatic improvement in potency, giving
3b with an IC50 of 330 nM against PTP1B. The com-
pound was also significantly more potent than the corre-
sponding quinoline analog 2b. Attempts to improve
intrinsic activity by extending to the secondary binding
site, as was described above for the quinoline analogs,
also resulted in compounds with substantial improve-
ment in enzymatic potency but negligible oral absorp-
tion (results not shown). As a result, we focused on
small groups in hoping to improve potency while main-
taining oral bioavailability. Compounds 3b–f were syn-
thesized according to Scheme 2.
As illustrated in Table 1, incorporation of a styrene moi-
ety gave 2c with improved inhibitory activity over 2b in
the enzymatic assay (IC50 = 160 nM), suggesting that
hydrophobic interactions were beneficial for activity.
The incorporation of the carboxylate in 2d improved
PTP1B potency dramatically (IC50 = 7 nM), validating
the prediction by molecular modeling. We further dis-
covered that the double bond in 2c could be replaced
with an amide group, giving compounds 2e–g with good
intrinsic activities (IC50 = 60, 27 and 45 nM, respec-
tively) without the requirement of a charged carboxylate
residue. Unfortunately, all these compounds did not
show any appreciable oral bioavailability in rodents. It
was apparent that large lipophilic groups or polar resi-
dues were detrimental to oral bioavailability. The simple
amide derivative N-methyl amide (2h) was significantly
less active (IC50 = 1.26 lM) but the primary amide 2i
displayed good activity (IC50 = 250 nM) and apprecia-
Reaction of LiP(OEt)2 with aldehyde 10 gave alcohol
11, which was oxidized under the Swern conditions to
the corresponding ketone. The ketone was treated with
an excess of DAST to give the phosphonate 12, which
was processed to 3b as mentioned above. Compound

3202
Y. Han et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3200–3205
Table 1. Potency and selectivity of naphthalene- and quinoline-difluoromethylphosphonate PTP1B inhibitors
a
Compound
X
H
R
IC50 (lM) (FDP assay)
PTP1B
CD45
F
P
F
1a
0.016
50
N
N
N
O
HO
OH
O
1b
1c
H
2.0
>50
>50
O
Br
0.12
2a
2b
H
Br
Me
Me
>50
14.0
>50
>50
2c
2d
2e
2f
Br
Br
Br
Br
0.16
>50
37.8
>50
>50
CO2H
0.007
0.060
0.027
H
N
O
O
H
N
F
H
N
2g
Br
0.045
>50
O
2h
2i
Br
Br
H
C(O)NHMe
C(O)NH2
H
1.26
0.25
Ki = 179b
0.33
1.5
>50
>50
na
3a
3b
3c
3d
3e
3f
3g
4
Br
Br
Br
Br
Br
Br
—
Me
>50
>50
>50
>50
26.8
>50
>50
CH2NHC(O)Me
CH2S(O)2Me
CH2OMe
CH2CN
1.30
0.51
0.09
0.12
0.31
CN
—
a Values are averages from 2 to 8 titrations, all compounds show similar activities against T-cell phosphatase and the values are not included here.
b Literature value; na, not available.
12 was treated with N-bromosuccinamide (NBS) in the
presence of a catalytic amount of benzoyl peroxide to
give bromide 13, which was reacted with a variety of
nucleophiles and then processed to 3c–f as shown.
potency. Less polar groups such as CH2OMe (3e) and
CH2CN (3f) were tolerated. Compound 3f was potent
against PTP1B with an IC50 of 90 nM. It also showed
reasonable plasma exposure in mice after oral dosing
(Cmax = 6 lM at 2 h with 20 mg/kg dosing). The nitrile
analog 3g, however, had the best overall profile in terms
of potency and oral bioavailability. It had an IC50 of
As shown in Table 1, polar groups such as acetamide
(3c) and methylsulfone (3d) were detrimental to intrinsic

Y. Han et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3200–3205
3203
exhibited good pharmacokinetic profiles in several spe-
cies as shown in Table 2.
In diet-induced obese (DIO) mice, 3g exhibited good
oral bioavailability (F = 24%), slow clearance
(CL = 0.71 mL/kg/min), and good elimination half live
(t1/2 = 6 h). The oral bioavailability in higher species
such as rats (F = 4%) and squirrel monkeys (F = 2%)
were substantially lower but excellent exposures were
achieved with oral dosing as shown in Table 2. In addi-
tion, 3g had very low clearance rate in rats and squirrel
monkeys, resulting in very long elimination t1/2s (>17
and >30 h, respectively). The corresponding regio iso-
mer 4 also exhibited good pharmacokinetic profiles in
mice but was 3-fold less potent.
ARG 24
ILE 219
GLN 262
ARG 254
ARG 221
˚
Figure 2. High resolution (1.6 A) X-ray structure of the PTP1B-1c
complex. Active site residues important for binding are labeled in black
and the structure of compound 1c is in green.
Compound 3g and 4 were synthesized according to
Scheme 3. 6,7-Dibromo-2-naphthonitrile (14) was pre-
pared from commercially available 4,5-dibromo-o-xy-
lene in two steps according to literature procedures.15
Monolithiation with n-BuLi in the presence of TMSCl
followed by treating the resultant TMS derivatives with
iodine monochloride (ICl) gave a mixture of 6-bromo-7-
iodo- and 7-bromo-6-iodo-2-naphthonitrile (15 and 16).
The mixture was reacted with ((diethoxyphosphinyl)-
difluoromethyl)zinc bromide using CuBr as the catalyst
to yield a mixture of diethyl difluoromethyl-phospho-
nates 17 and 18 which was separated by normal phase
HPLC using a Zorbax silica column. Deprotection of
17 and 18 with TMSBr followed by treating the TMS-es-
ters with water gave the corresponding phosphonic acid
3g and 4. The free acids were easily converted to the
diammonium salts by treating with an excess of aqueous
ammonia in MeOH. The experimental procedures for
making 3g and 4 are described in note.16
ARG 24
ILE 219
GLN 262
ARG 254
ARG 221
Figure 3. Compound 2d modeled into the active site of PTP1B. The
ligand scaffold is in green. The difluoromethylphosphonate in both 1c
and 2d binds to the primary binding site similarly.
Compound 3g was studied in animal models of diabetes
and cancer. For example, in oral glucose tolerance tests
(oGTT) in DIO mice,17 3g exhibited dose dependent
inhibition (60%, 80% and 100% inhibition at 1, 3 and
10 mg/kg, respectively) of glucose excursion when given
orally 2 h before oral glucose challenge with an esti-
120 nM in the enzymatic assay and was the most active
compound in inhibiting yeast growth among a panel of
potent phosphonate inhibitors with comparable inhibi-
tory potencies against the PTP1B enzyme.14 It also
Br
Br
Br
a
b
PO(OEt)2
H2N
I
N
I
N
F
F
5
6
7
d
c
Br
Br
f
2b
PO(OEt)2
PO(OEt)2
HOOC
N
OHC
N
F
F
F
F
9
8
e, c
g, c
2e-i
2c-d
Scheme 1. Reagents and conditions: (a) chloranil, concd HCl in i-PrOH at reflux, then crotonaldehyde in i-PrOH via syringe pump; (b)
(EtO)2P(O)CF2ZnBr, CuBr, DMF, 60 ꢁC; (c) TMSBr, DCM, rt and then co-evaporate with water/ethanol (3·); (d) SeO2, ethanol, reflux; (e)
ArCH2P+Ph3BrÀ, LiHMDS, THF, À78 ꢁC—rt; (f) AgO, MeOH, rt; (g) EDCI, RNH2, TEA, trace DMAP, DCM, rt.

3204
Y. Han et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3200–3205
Br
Br
O
Br
a
b, c
PO(OEt)2
OH
H
PO(OEt)2
F
F
10
11
12
e
d
Br
f,g,h,i
3b
3c-f
PO(OEt)2
then d
F
F
Br
13
Scheme 2. Reagents and conditions: (a) (EtO)2PH, LiHMDS, THF, À78 ꢁC; (b) (COCl)2, DMSO TEA, DCM, À78 ꢁC—rt; (c) DAST, chloroform,
rt; (d) TMSBr, DCM, rt and then co-evaporate with water/ethanol; (e) NBS, cat. Benzoyl peroxide, CCl4, reflux; (f) 3c: aq NH3, then EDCI, AcOH;
(g) 3d: MeSNa in MeOH, then oxoneꢂ, acetone; (h) 3e: NaOMe, MeOH; (i) 3f: NaCN, MeOH.
Table 2. Pharmacokinetic profiles of 3g in mice, rats and monkeys
Species
Dosea (mg/kg) iv/po
F (%)
t1/2 (h)
CL (mL/min/kg)
Cmax/C24 h (lM)
DIO mice
Rat
Squirrel monkey
5/20b
5/20c
2/10c
24
4
6
>17
>30
0.71
0.03
0.02
17.4/9.0
75.6/45.3
24.0/19.0
2
a The diammonium salt was used for PK studies.
b Dosed as a solution in 0.9% NaCl for iv and as a 25% PEG200 solution for po.
c Dosed as a solution in 0.9% NaCl for iv and as a 0.5% methylcellulose solution for po.
Br
Br
d, e
3g
4
PO(OEt)2
NC
NC
NC
NC
I
Br
Br
c
F
F
F
a, b
17
18
15
16
I
NC
Br
d, e
14
PO(OEt)2
Br
F
Scheme 3. Reagents and conditions: (a) n-BuLi, TMSCl, THF, À78 ꢁC; (b) ICl, DCM, rt; (c) (EtO)2P(O)CF2ZnBr, CuBr, DMF, 45 ꢁC; (d) TMSBr,
DCM, rt and then co-evaporate with water/ethanol (3·); (e) NH4OH, MeOH, rt.
Himms-Hagen, J.; Chan, C.-C.; Ramachandran, C.;
Gresser, M. J.; Tremblay, M. L.; Kennedy, B. P. Science
1999, 283, 1544; (b) Klaman, L. D.; Boss, O.; Peroni, O.
D.; Kim, J. K.; Martino, J. L.; Zabolotny, J. M.; Moghal,
N.; Lubkin, M.; Kim, Y.-B.; Sharpe, A. H.; Stricker-
Krongrad, A.; Shulman, G. I.; Neel, B. G.; Kahn, B. B.
Mol. Cell. Biol. 2000, 20, 5479.
mated ED50 of 0.8 mg/kg. The anticancer effect of 3g
was studied in the NDL2 Ptpn1 transgenic mice.3 When
dosed orally at 30 mg/kg for 21 days, 3g caused a signif-
icant delay in the onset of tumor development in NDL2
Ptpn1+/+ mice when compared to vehicle treatment,
extending the median tumor free days (T50) from 28
days to 75 days. Interestingly, the blood glucose levels
of the drug treated animals in this study were lowered
to comparable levels seen in the NDL2 Ptpn1À/À ani-
mals, further demonstrating the antidiabetic potential
of this PTP1B inhibitor.
2. (a) Zinker, B. A.; Rondinone, C. M.; Trevillyan, J.
M.; Gum, R. J.; Clampit, J. E.; Waring, J. F.; Xie,
N.; Wilcox, D.; Jacobson, P.; Frost, L.; Kroeger, P.
E.; Reilly, R. M.; Koterski, S.; Opgenorth, T. J.;
Ulrich, R. G.; Crosby, S.; Butler, M.; Murray, S. F.;
McKay, R. A.; Bhanot, S.; Monica, B. P.; Jirousek,
M. R. PNAS U.S.A. 2002, 99, 11357; (b) Rondinone,
C. M.; Trevillyan, J. M.; Clampit, J. E.; Gum, R. J.;
Berg, C.; Kroeger, P.; Frost, L.; Zinker, B. A.; Reilly,
R.; Ulrich, R.; Butler, M.; Monia, B. P.; Jirousek, M.
R.; Waring, J. F. Diabetes 2002, 51, 2405; (c) Gum,
R. J.; Gaede, L. L.; Koterski, S.; Heindel, M.;
Clampit, J. E.; Zinker, B. A.; Trevillyan, J. M.;
Ulrich, R.; Jirousek, M. R.; Rondinone, C. M.
Diabetes 2003, 52, 21.
Acknowledgment
The authors thank Dr. Laird Trimble at Merck Frosst
for NMR support.
References and notes
3. Julien, S. G.; Dube, N.; Read, M.; Penney, J.; Paquet, M.;
Han, Y.; Kennedy, B. P.; Muller, W. J.; Tremblay, M. L.
Nat. Genet. 2007, 39, 338.
1. (a) Elchebly, M.; Pauette, P.; Michaliszyn, E.; Cromlish,
W.; Collins, S.; oy, A. L.; Normandin, D.; Cheng, A.;

Y. Han et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3200–3205
3205
4. Zhang, S.; Zhang, Z.-Y. DDT 2007, 12, 373.
5. Taylor, S. D.; Hill, B. Expert Opin. Investig. Drugs 2004,
13, 199.
6. Shrestha, S.; Bhattarai, B. R.; Cho, H.; Choi, J.-K.; Cho,
H. Bioorg. Med. Chem. Lett. 2007, 17, 2728.
7. (a) Lau, C. K.; Bayly, C. I.; Gauthier, J. Y.; Li, C. S.;
Therien, M.; Asante-Appiah, E.; Cromlish, W.; Skorey,
K.; Waddleton, D.; Payette, P.; Ramachandran, R.;
Kennedy, B. P.; Scapin, G. Bioorg. Med. Chem. Lett.
2004, 14, 1043; (b) Dufresne, C.; Roy, P.; Wang, Z.;
Asante-Appiah, E.; Cromlish, W.; Boie, Y.; Forghani, F.;
Desmarais, S.; Wang, Q.; Skorey, K.; Waddleton, D.;
Ramachandran, R.; Kennedy, B. P.; Xu, L.; Gordon, R.;
Chan, C. C.; Leblanc, Y. Bioorg. Med. Chem. Lett. 2004,
14, 1039.
8. Huang, Z.; Wang, Q.; Ly, H. D.; Gorvindarajan, A.;
Scheigetz, J.; Zamboni, R.; Desmarais, S.; Ramachan-
dran, C. Biomol. Screen. 1999, 4, 327.
9. Puius, Y. A.; Zhao, Y.; Sullivan, M.; Lawrence, D. S.;
Almo, S. C.; Zhang, Z. Y. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 13420.
10. Unpublished results. Coordinates and structure factors
have been deposited with the Protein Data Bank,
Accession No. 3CWE.
11. Johnsson, R.; Meijer, A.; Ellervik, U. Tetrahedron 2005,
61, 11657.
the above product in dichloromethane (250 mL) was
added excess ICl and the mixture was stirred at rt for
1 h. The solution was washed with 10% Na2S2O3 until all
ICl was consumed. The solution was then washed with
water, brine, dried over MgSO4, and filtered. The filtrate
was concentrated and the residue was recrystallized from
ether/hexanes to give the desired product. 1H NMR
(400 MHz, acetone-d6) (a mixture of two regioisomers): d
8.74 (s, 1H), 8.73 (s, 1H), 8.46–8.44 (m, 4H), 8.10–8.07 (m,
2H), 7.84–7.81 (m, 2H); (b) Compound 17 and 18: A flame
dried round-bottomed flask was charged with CuBr
(99.999%) and THF (10 mL), followed by ((diethoxyphos-
phinyl)difluoromethyl)zinc bromide (29 mL, 1.72 M in
THF) following the procedure of Shibuya et al.10 The
suspension was stirred under N2 for 15 min. A mixture of
15 and 16 (7.1 g) was then added as a solid and the mixture
was heated to 45 ꢁC overnight and cooled to rt. The
suspension was then quenched with half saturated NH4Cl
and extracted with 1:1 ether/ethyl acetate (3·). The
extracts were processed as usual to give the crude product
which was first purified by flash chromatograph (40%
ethyl acetate in hexanes). The two regioisomers were then
separated by normal phase HPLC using a Zorbax silica
column. Eluting with 50% ethyl acetate/hexanes first gave
the less polar isomer 17. 1H NMR (400 MHz, acetone-d6):
d8.72 (s, 1H), 8.54 (s, 1H), 8.46 (s, 1H), 8.19 (d, 1H), 7.95
(d, 1H), 4.26 (m, 4H), 1.33 (t, 6H). Continued elution gave
12. Yokomatsu, T.; Murano, T.; Suemune, K.; Shibuya, S.
Tetrahedron 1997, 53, 815.
1
the more polar isomer 20. H NMR (400 MHz, acetone-
13. Liu, D.-G.; Gao, Y.; Voigt, J. H.; Lee, K.; Nichlaus, M.
C.; Wu, L.; Zhang, Z. Y.; Burke, T. R., Jr. Bioorg. Med.
Chem. Lett. 2003, 13, 3005.
14. Montalibet, J.; Kennedy, B. P. Biochem. Pharmacol. 2004,
68, 1807.
d6): d 8.56 (s, 2H), 8.43 (s, 1H), 8.35 (d, 1H), 7.93 (d, 1H),
4.26 (m, 4H), 1.33 (t, 6H). The regiochemistry of the two
isomers was vigorously established by COSY, NOSEY,
HSQC, and HMBC experiments; (c) Compound 3g: A
solution of diethyl [(3-bromo-7-cyano-2-naphthyl)
(difluoro)methyl]phosphonate (2.2 g) in dichloromethane
(5 mL) and TMSBr (7 mL) was stirred at rt overnight and
concentrated. The residue was co-evaporated with dichlo-
romethane (2·), ethanol/water (2·) and then dissolved in
20 mL of methanol. Ammonia (30%) was then added with
vigorous stirring and the mixture was concentrated and
co-evaporated with methanol (3·). The solid residue was
washed with ether to give the desired product as a white
15. Hanack, M.; Grobhans, R. Chem. Ber. 1992, 125, 1243.
16. Experimental procedures for making compound 3g. (a) 7-
Bromo-6-iodo-2-naphthonitrile (15) and 6-bromo-7-iodo-2-
naphthonitrile (16): To a solution of 6,7-dibromo-2-
naphthonitrile (15 g) and TMSCl (6.73 mL) in THF
(250 mL) at À78 ꢁC was added n-BuLi (53 mL, 1.6 M in
hexanes, precooled to À20 ꢁC) rapidly with vigorous
stirring and the mixture was stirred for an additional
5 min and quenched with saturated aq NH4Cl. The
mixture was then extracted with ethyl acetate and the
organic layer was washed with water and brine, dried over
MgSO4, and filtered. The filtrate was concentrated and the
crude was purified by flash column chromatography to
give the desired product as a mixture of two regioisomers.
1H NMR (400 MHz, acetone-d6) (a mixture of two
regioisomers): d 8.53 (s, 1H), 8.42 (s, 1H), 8.33 (s, 1H),
8.30 (s, 1H), 8.23 (s, 1H), 8.20 (s, 1H), 8.18 (d, 1H), 8.07
(d, 1H), 7.82–7.77 (m, 2H), 0.50 (s, 18H). To a solution of
1
powder. H NMR (400 MHz, acetone-d6): d 8.65 (s, 1H),
8.58 (s, 1H), 8.54, (s, 1H), 8.16 (d, J = 5 Hz, 1H), 7.90 (d,
J = 5 Hz, 1H). MS (ÀESI): m/z 360.1 and 362.0 (MÀ1)À.
17. Lankas, G. R.; Leiting, B.; Roy, R. S.; Eiermann, G. J.;
Beconi, M. G.; Buftu, T.; Chan, C.-C.; Edmondson, S.;
Feeney, W. P.; He, H.; Ippolito, D. E.; Kim, D.; Lyons, K.
A.; Ok, H. O.; Patel, R. A.; Petrov, A. N.; Pryor, K. A.;
Qian, X.; Reigle, L.; Woods, A.; Wu, J. K.; Zaller, D.;
Zhang, X.; Zhu, L.; Weber, A. E.; Thornberry, N. A.
Diabetes 2005, 54, 2988.