5-Carboxamido-1,3,2-dioxaphosphorinanes, potent inhibitors of MTP

Article abstract of DOI:10.1016/j.bmcl.2004.07.069

5-Carboxamido-1,3,2-dioxaphosphorinanes have been identified as potent inhibitors of microsomal triglyceride-transfer protein. The 1,3,2- dioxaphosphorine functionality acted as a neutral and stable replacement for piperidine and piperidine N-oxide.

Full text of DOI:10.1016/j.bmcl.2004.07.069

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5067–5070
5-Carboxamido-1,3,2-dioxaphosphorinanes,
potent inhibitors of MTP
Richard Sulsky,a,* Jeffrey A. Robl,a Scott A. Biller,a Thomas W. Harrity,b John Wetterau,b
Fergal Connolly,b Kern Joliboisb and Lori Kunselmanb
aDepartment of Discovery Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
bDepartment of Metabolic Disease, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
Received 2 July 2004; revised 27 July 2004; accepted 29 July 2004
Available online 24 August 2004
Abstract—5-Carboxamido-1,3,2-dioxaphosphorinanes have been identified as potent inhibitors of microsomal triglyceride-transfer
protein. The 1,3,2-dioxaphosphorine functionality acted as a neutral and stable replacement for piperidine and piperidine N-oxide.
Ó 2004 Elsevier Ltd. All rights reserved.
Isoindolylpiperidine BMS-188063 (1) was identified by
high-throughput screening in our laboratories as an
inhibitor of microsomal triglyceride-transfer protein
(MTP).1 Inhibiting the function of MTP, a protein nec-
essary to the assembly and secretion of lipoproteins, has
been shown to be an effective method for reducing ser-
um cholesterol.1
contain the piperidine (or other basic amine) functional-
ity. Described here is the identification of one such chemo-
type, the cyclophosphonates of general structure 3.
In the course of exploring the SAR of 2, we synthesized
the N-oxide 4,3 a compound equipotent (in its ability to
inhibit the secretion of lipoproteins from HepG2 cells)
with 2 in vitro and slightly less potent in vivo (Table
1). In addition, we prepared simple phosphonates such
as 5, with somewhat diminished intrinsic potency and
in vivo activity.4 We hypothesized that the phospho-
nates and piperidines, having greatly different structures,
do not bind to the protein at precisely the same site.
O
N
N
1
On the other hand, N-oxides are isoelectronic with phos-
phonate esters, although the N–O bond is somewhat
shorter than the P–O bond.5 As a challenge to our ini-
tial hypothesis, we conjectured that a hybrid structure,
O
F
3
C
2
H
C
H
N
NH
N
O
C
2
CF3
H
O
N
CF3
O
P
O
Intense efforts, by both conventional and combinatorial
synthetic methods,2 led to BMS-201038 (2), which had
advanced as a clinical candidate in the treatment of
hypercholesterolemia. To expand the SAR of this series,
structurally diverse chemotypes were developed with the
aim of preparing potent MTP inhibitors that did not
O
O
N
H
R
3
CF3
H
N
H
O
CF3
O
N
CF3
O-
O
P
N+
O
OnBu
Keywords: MTP inhibitors; Cyclophosphonates.
*
OnBu
N
H
Corresponding author. Tel.: +1 609 8184730; fax: +1 609 8186810;
e-mail: sulskyr@bms.com
4
5
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.07.069

5068
R. Sulsky et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5067–5070
Table 1. MTP inhibition profile of selected compounds
F3
C
H2
C
H
R
N
O O
P
N
O
C
X
W or W' =
NH
O
Z =
NH
O
Y
O
Y
W, R = nil
W', R = O
Compound
X
Y
IC50 in vitro (nM)a
HepG2 apo B ED50 (nM)b
D% Serum cholesterol (mpk)c
2
W
W0
2-(40-Trifluorobiphenyl)
2-(40-Trifluorobiphenyl)
––
2-(40-Trifluorobiphenyl)
2-(40-Trifluorobiphenyl)
–OCH2C6H5
8
5
0.8
0.9
7
ꢀ92 (30)e
ꢀ77 (15)
ꢀ5 (30)
ꢀ73 (30)f
ꢀ4 (20)
ND
4
5
PO3Bu2
trans-Z
cis-Z
22
7
3a
3b
3c
3e
3f
3g
3h
3i
3j
0.65
18
40
44
14
11
9
trans-Zd
trans-Z
trans-Z
trans-Z
trans-Z
trans-Z
trans-Z
C6H5–
2-(N-Benzyl)piperidinyl
2-(20-Pyridyl)phenyl
2-(20-Benzothiazolyl)phenyl
2-(10-Morpholino)phenyl
2-(20-Benzoxazolyl)phenyl
23
ꢀ12 (15)
ꢀ22 (30)
ꢀ19 (15)
ꢀ76 (15)g
ꢀ3 (15)
ꢀ13 (30)
ND
79
<10
9
7.5
ND
ND
12
3
a In vitro values of triglyceride transfer of human MTP.
b Inhibition of secretion of apoB and apoA1 from HepG2 cells.
c Change in serum HDL cholesterol from control, after 3day study in male Golden Syrian hamsters dosed once daily.1
d For corresponding cis-isomer, I50 = 210nM.
e ED50 = 2.0mpk.
f ED50 = 3.3mpk.
g ED50 = 4.0mpk.
F
2C
H
C
H
F
3
2C
N
O
C
H
3
C
H
N
O
C
CO2H
b
a
PO3Et2
Br
7
6
F
2C
H
3
C
H
N
O
C
c
POCl2
8
Scheme 1. Reagents and conditions: (a) (i) 2n-BuLi/THF, (ii) Br(CH2)4Br, (iii) (COCl)2, DMF/CH2Cl2, (iv) CF3CH2NH2ÆHCl, Et3N/CH2Cl2, 71%;
(b) P(OEt)3, 110°C, 79%; (c) (i) TMSBr/CH2Cl2, (ii) (COCl)2, DMF/CH2Cl2, 88%.
CF3
NH2
a
OH
OH
b
O
HO
OH
N
H
CF3
CF3
CONHCH2CF3
O
CONHCH2CF3
O
O
O
+
P
O
P
O
O
O
N
H
N
H
cis-(3b)
trans-(3a)
Scheme 2. Reagents: (a) EDCI, HOBT, 2-(40-trifluoromethylphenyl)benzoic acid/THF, 72%; (b) 8, Et3N/CH2Cl2.
containing both phosphonate and piperidine N-oxide struc-
tural features, would have a synergistic effect on binding
potency. Thus we set about preparing compounds of
structural type 3.6 Preparation of the necessary phos-
phoryl chloride (8) (Scheme 1), proceeded in three steps
from 9-fluorenylcarboxylic acid. Carbodiimide coupling
of serinol with 2-(trifluoromethylphenyl)benzoic acid
provided 9a, which was then reacted with 8 to give the

R. Sulsky et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5067–5070
5069
NHTMS
OTMS
NHCO2CH2C6H5
OH
b
NH2
a
TMSO
HO
HO
c
OH
10
CONHCH2CF3
O
d
CONHCH2CF3
O
P
O
O
P
O
O
NHCO2CH2C6H5
3c
NH2
3d
CONHCH2CF3
O
e
O
P
O
NHCOR
3e-j
Scheme 3. Reagents and conditions: (a) (Me3Si)2NH, TMSBr, 180°C, 86%; (b) (i) CbzCl, Et3N/CH2Cl2, (ii) 4N HCl/dioxane, 92%; (c) (8), Et3N/
CH2Cl2, 23%; (d) H2, 10% Pd–C/EtOH, 95%; (e) EDCI, HOBt, Et3N, RCO2H/CH2Cl2.
trans-(3a) and cis-(3b) isomers, obtained in a 63:37 ratio
in 44% yield. Separation of the isomers was readily
accomplished by normal phase silica gel chromatography,
We have prepared a series of cyclophosphonate analogs
of the MTP inhibitor BMS-201038, a clinical candidate.
The compounds are comparatively rigid structural mim-
ics of the presumed bioactive chair form of the piper-
idine contained by BMS-201038. The cyclophosphonate
compounds are essentially equipotent in vitro with the
piperidine-containing MTP inhibitors. Two of the cyclo-
phosphonates are equipotent with the clinical candidate
in an in vivo model of cholesterol transport. Contrary to
our initial hypothesis, no synergistic effect of the phos-
phonate was uncovered. The cyclophosphonate func-
tionality neither helps nor interferes with the binding
affinity of molecules to MTP. However, the cyclophos-
phonate structure itself may serve as a useful, stable,
neutral surrogate for piperidines, their N-oxides, and
additional nonaromatic six-member rings in other drug
targets.
1
and structural assignments were made by H, 13C, and
31P NMR spectroscopy7 (Scheme 2).
The trans-isomers were consistently more active in vitro
than the cis-isomers (see Table 1). A more efficient tech-
nique for obtaining a variety of desired trans-analogs of
3a is described in Scheme 3. Serinol was exhaustively sil-
ylated and the resulting tris-silylated compound was
protected with CbzCl, followed by acidolysis (Scheme
3) to give 10. This diol was then reacted with phosphoryl
dichloride 8 to give carbamate 3c (isolated and sepa-
rated from the cis-analog in 23% yield). Hydrogenolysis
of 3c gave amine 3d, a common intermediate, which was
acylated to provide compounds 3e–j.
Biological data. The compounds were assayed in vitro,
in a cell-based assay and in a whole-animal model. Ini-
tial assay results showed that the trans-compounds were
5–6-fold more potent in vitro than the cis-isomers (Table
1).1 The acyl analogs of 3 revealed little difference in the
triglyceride transfer human MTP assay, with IC50 in the
range of 3–14nM for the trans-isomers and somewhat
less potent than piperidine 2.
References and notes
1. Wetterau, J. R.; Gregg, R. E.; Harrity, T. W.; Arbeeny, C.;
Cap, M.; Connolly, F.; Chu, C.-H.; George, R. J.; Gordon,
D. A.; Jamil, H.; Jolibois, K. G.; Kunselman, L. K.; Lan,
S.-J.; Maccagnan, T. J.; Ricci, B.; Yan, M.; Young, D.;
Chen, Y.; Fryszman, O. M.; Logan, J. V. H.; Musial, C. L.;
Poss, M. A.; Robl, J. A.; Simpkins, L. M.; Slusarchyk, W.
A.; Sulsky, R.; Taunk, P.; Magnin, D. R.; Tino, J. A.;
Lawrence, R. M.; Dickson, J. K., Jr.; Biller, S. A. Science
1998, 282, 751.
Apolipoprotein
B
(apoB)-containing lipoproteins
promote coronary artery atherosclerosis. We used a hu-
man liver-derived cell line, HepG2, which secrete apoB,
as a cell-based assay to further characterize these com-
pounds. Lipoprotein secretions of apo B in HepG2 cells
were inhibited by the cyclophosphonate analogs, the
most potent being 3a (ED50 = 0.65nM). This compares
favorably to clinical candidate BMS-201038, which also
possesses subnanomolar potency in the cell assay.
2. Lawrence, R. M.; Biller, S. A.; Fryszman, O. M.; Poss, M.
A. Synthesis 1997, 553.
3. (a) Biller, S. A.; Dickson, J. K.; Lawrence, R. M.; Magnin,
D. R.; Poss, M. A.; Sulsky, R. B.; Tino, J. A. U.S. Patent
5,739,135, 1998; (b) Biller, S. A.; Dickson, J. K.; Lawrence,
R. M.; Magnin, D. R.; Poss, M. A.; Sulsky, R. B.; Tino, J.
A. U.S. Patent 5,712,279, 1998.
4. (a) Biller, S. A.; Dickson, J. K.; Lawrence, R. M.; Magnin,
D. R.; Poss, M. A.; Robl, J. A.; Slusarchyk, W. A.; Sulsky,
R. B.; Tino, J. A. U.S. Patent 5,760,246, 1998. (b) Magnin,
D. R.; Biller, S. A.; Wetterau, J.; Robl, J. A.; Dickson, J.
K.; Taunk, P.; Harrity, T. W.; Lawrence, R. M.; Sun, C.-Q.;
Wang, T.; Logan, J.; Fryszman, O.; Connolly, F.; Jolibois,
K.; Kunselman, L. Bioorg. Med. Chem. Lett. 2003, 13(7),
1337–1340.
The compounds were tested in vivo in a three-day po
hamster study. Hamsters, unlike other rodents, trans-
port a substantial proportion of their cholesterol on
LDL, thus providing a useful animal model of the hu-
man system.1 Analogs 3a (ED50 = 3.3mpk) and 3g,
(ED50 = 4mpk) exhibited in vivo efficacy similar to that
of BMS-201038 (ED50 = 2.0mpk).
5. Jaskolski, M.; Olovsson, I.; Tellgren, R.; Mickiewicz-
Wichlacz, D. Acta Crystallogr., Sect. B 1982, B38(1), 291;

5070
R. Sulsky et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5067–5070
Yokomatsu, E.; Nakabayashi, N.; Matsumoto, K.; Shi-
CONHCH2CF3); 4.21 (dt, 1H, J = 4.1Hz, OCH2CHNH);
3.96 (dd, 4H, J = 4.1, 11.1Hz, OCH2CHNH); 3.68 (dq, 2H,
J = 6.9, 9.0Hz, NHCH2CF3); 2.35 (m, 2H, CH2(CH2)3PO);
1.4 (m, 4H, CH2(CH2)2CH2PO); 0.71 (m, 2H,
(CH2)3CH2PO). Compound 3b (cis): 1H 300MHz NMR
(CDCl3): d 7.2–7.8 (m, 16H, ArHs, NHCOAr); 6.27 (d, 1H,
J = 7.9Hz, NHCOAr); 5.44 (t, 1H, J = 6.3Hz, CON-
HCH2CF3); 4.54 (d, 2H, J = 11.4Hz, equatorial
OCH2CHNH); 4.07 (d, 1H, J = 6.1Hz, OCH2CHNH);
3.79 (dd, 2H, J = 6.1, 11.8Hz, axial OCH2CHNH); 3.65
(m, 2H, NHCH2CF3); 2.28 (m, 2H, CH2(CH2)3PO);
1.4 (m, 4H, CH2(CH2)2CH2PO); 0.80 (m, 2H, (CH2)3-
CH2PO).
buya, S. Tetrahedron: Asymmetry 1995, 6, 3055; Ruiz, M.;
Ojea, V.; Shapiro, G.; Weber, H.-P.; Pombo-Villar, E.
Tetrahedron Lett. 1994, 35, 4551.
6. Sulsky, R. U.S. Patent 5,962,440. Included are experimen-
tals for all cyclophosphonate compounds and intermediates
described in this paper.
7. Assignments of stereochemistry were made on the basis of
1H NMR spectroscopy and comparison to literature
examples of cyclophosphonates (see, Yee, K. C.; Bentrude,
W. G. Tetrahedron Lett. 1971, 12, 2775) NMR data:
1
Compound 3a (trans): H 300MHz NMR (CDCl3): d 7.2–
7.8 (m, 17H, ArHs, NHCOAr); 5.37 (t, 1H, J = 6.5Hz,