A new class of high affinity thyromimetics containing a phenyl-naphthylene core

Article abstract of DOI:10.1016/j.bmcl.2005.06.093

High affinity thyromimetics containing a novel phenyl-naphthylene core are reported. The functionalized core is readily accessible via a Suzuki coupling protocol. Examples of this new class of TR ligands have sub-nanomolar binding affinities for the TRβ r

Full text of DOI:10.1016/j.bmcl.2005.06.093

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4579–4584
A new class of high affinity thyromimetics containing
a phenyl-naphthylene core
Jon J. Hangeland,a,* Todd J. Friends,a Arthur M. Doweyko,a Karin Mellstro¨m,b
Johnny Sandberg,b Marlena Grynfarbb and Denis E. Ryonoa
aPharmaceutical Research Institute, Bristol-Myers Squibb, Princeton, NJ 08543-5400, USA
bKaro Bio AB, Novum, Huddinge S-141 57, Sweden
Received 13 May 2005; revised 28 June 2005; accepted 29 June 2005
Available online 15 August 2005
Abstract—High affinity thyromimetics containing a novel phenyl-naphthylene core are reported. The functionalized core is readily
accessible via a Suzuki coupling protocol. Examples of this new class of TR ligands have sub-nanomolar binding affinities for the
TRb receptor and low to modest selectivity for TRb. They also exhibit an SAR that diverges from other thyromimetics that are
based on the diaryl ether core found in 3,5,30-triiodothyronine.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Binding of the 3,5,30-triiodothyronine (T3; 1 in Fig. 1) to
the thyroid hormone receptor (TR) controls multiple
physiological endpoints (e.g., heart rate, bone and mus-
cle metabolism, cholesterol levels, and metabolic rate)
through transcriptional control of various gene targets.1
Levels of T3 in excess of normal physiologic amounts
cause cardiac hypertrophy, tachycardia, bone and mus-
cle wasting, and weight loss. Sub-physiological amounts
produce bradycardia and weight gain. The amount of
T3 needed to provide clinically relevant weight loss caus-
es unwanted cardiac side effects, preventing its use for
anti-obesity therapy.
to the receptor induces conformational changes to the
receptor surface that generate a co-factor binding site.5
Formation of the co-factor binding site is critical for
proper gene transcription control by the TR/ligand com-
plex. In addition, these structures show conclusively that
the bound conformation of the ligand has the 30-group
oriented away from the inner ring (i.e., in the distal po-
sition), confirming the outcome of earlier stereochemical
and biological activity studies using thyroxin analogs.6
While the association between the ligand and LBD is
very tight, side-chain and backbone movements within
the 30-binding pocket of the LBD9a,b permit groups sig-
nificantly larger than iodine to be appended onto the
core structure at the 30-position, leading to increased
selectivity for TRb while retaining full agonist activity.9
The thyroid hormone receptor is encoded by two sepa-
rate genes, TRa and TRb. Each produces multiple splice
variants, the predominant ones being TRa1 and TRb1.
TRa1 has been shown by tissue distribution data2 and
mouse knock-out studies3 to be the dominant isoform
in cardiac tissue, making it possible to separate unwant-
ed cardiac side effects (e.g., cardiac hypertrophy
and tachycardia) from beneficial effects by designing
thyromimetics selective for TRb1.4
The vast majority of reported thyromimetics retain two
features of the core structure of T3: a diphenyl ether and
a 3,5-disubstituted inner ring, the two acting in concert
to orient the phenyl rings in a skewed relationship.
The skewed conformation is labeled ÔidealÕ10 when the
inner phenyl ring is orthogonal to the C–O–C plane
(r = 90ꢁ; Fig. 2A) and the outer ring is coplanar to the
C–O–C plane (r0 = 0ꢁ; Fig. 2A). An analysis of in-house
generated X-ray crystal structures for a variety of thyr-
omimetics, including compound 4,8 complexed to either
the TRa or TRb LBD showed the average values for r
and r0 to be 79ꢁ and 47ꢁ, respectively (Fig. 2B). Thus,
the receptor bound conformation of the ligand deviates
from the ideal skewed conformation. Similar deviations
were observed in X-ray crystal structures of T3, and
X-ray crystal data of various thyromimetics bound to
the TR ligand binding domain (LBD) show that the li-
gand is buried within the core of the LBD.5 The ligand
and protein contacts are tight, and binding of the ligand
*
Corresponding author. Tel.: +1 609 818 49 58; fax: +1 609 818
3450; e-mail: jon.hangeland@bms.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.06.093

4580
J. J. Hangeland et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4579–4584
O
O
HO
I
4'
3'
1
5
HO
I
I
OH
OH
NH2
H3C
O
NH2
O
3
CH3
I
I
1
2
O
HO
Cl
OH
HO
O
OH
O
R
O
Cl
4: R = CH(CH3)2
5: R = H
6: R = Br
3
Figure 1. Structures of T3 (1), 20,30-substituted analogs of T3 from Jorgensen and Berteau (2),6 carbon-linked analogs from Scanlan and co-workers
(3),7 and phenyl acetic acid analogs from Malm and co-workers (4).8
reported to be a competitive antagonist in a transactiva-
tion assay, whereas compound 3 was reported to be a
low-potency agonist.7b
σ ave = 47˚
'
σ
'
= 0˚
HO
R
HO
The key carbon–carbon bond forming reaction used to
construct the phenyl-naphthylene core was carried out
using a Suzuki reaction (Scheme 1). The synthesis start-
ed from 6-methoxynaphth-1-ol (7),12 which was con-
verted to boronic ester 9 via triflate 8.13 The coupling
partners of boronic ester 9 were obtained by conver-
sion of methyl 2-(3,5-dichloro-4-hydroxy-phenyl)ace-
tate (10)8,9c and 6-dichloro-4-nitrophenol (11) to their
corresponding triflates 12 and 13. Suzuki coupling of
boronic ester 9 with triflate 12 required heating at
80 ꢁC for 4 h to give compound 14, albeit in low yield
(ca. 9%). In contrast, coupling of ester 9 with triflate 13
gave compound 15 in good yield (69%) in 30 min at
80 ꢁC. The higher yield obtained with triflate 13 was
a key improvement in the synthesis of 15, providing
sufficient material to permit more extensive SAR
studies than with 14.
R
4
1
σ
= 90˚
O
σ
ave = 79˚
A
B
Figure 2. Two views of the conformational relationship between the
phenyl rings of the diphenyl ether core of T3 and its analogs. (A)
depicts a side view of an ideal skewed conformation where the outer
ring is in the same plane as C–O–C and the inner ring is orthogonal to
the C–O–C plane; (B) depicts a view of the diphenyl ether core looking
down the C-1 ! C-4 axis showing the deviation from the ideal skewed.
Note that the outer ring has twisted such that, in this view, the 30 group
is positioned to the left and the outer ring has rotated
counterclockwise.
closely related analogs, not complexed to the receptor
LBD.10
The first two examples of this new class of thyromimetic,
compounds 16 and 17, were synthesized from compound
14 (Scheme 2) and were modeled after compound 4.8
For comparison, two diaryl ether analogs of 16 and 17
were synthesized14 (compounds 515 and 6, respectively,
in Fig. 1). Compound 16 had modest affinity for TRa
and TRb, and was slightly selective for TRb (Table 1).
Its diaryl ether analog, compound 5, had lower affinity
for TRa and TRb (2- and 4-fold, respectively), and
was unselective. Compound 17, which contained bro-
mine at the 50-position, had improved affinity for TRa
and TRb (15- and 42-fold, respectively) compared to
16, and was slightly more selective for TRb, reflecting
a greater improvement in affinity for TRb versus TRa.
The improved potency was consistent with the 50-substi-
tuent being located in the 30-binding pocket of the LBD.
In contrast to 5, compound 6 was equipotent to 17 and
had similar selectivity for TRb. In addition, compound
17 was a full agonist with EC50 values of 57 nM (80%
induction) for TRa and 40 nM (92% induction) for
TRb.16
To test if forcing the two phenyl rings to be orthogonal
would lead to high affinity thyromimetics or would con-
tribute to the selectivity for one TR isoform over the
other, a novel phenyl-naphthylene core was designed
(Fig. 3). The new core retained the essential relationship
between the two phenyl rings, locked into position the
distal conformational relationship of the 30-substitutent
to the inner phenyl ring, and enforced co-planarity of
the outer ring with the C–O–C plane (r0 = 0ꢁ). The 20-
position of the outer phenyl ring was tied to the bridging
carbon via a three carbon linker to form the second ring
of the naphthylene moiety. By doing this, the new struc-
ture combined substitution patterns reported in two sep-
arate series: (1) analogs of T3 containing small alkyl
substituents at the 20- and 30-positions6 (2 in Fig. 1),
and (2) analogs of GC-111 containing substitution on
the bridging methylene carbon linking the two phenyl
rings7 (3 in Fig. 1). An analog of compound 3 containing
a significantly larger group off the bridging carbon was

J. J. Hangeland et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4579–4584
4581
Figure 3. Schematic representation of the design of the phenyl-naphthylene core (B) as it relates to the traditional thyromimetic architecture (A).
Note that the numbering of the naphthyl ring system differs from that of the outer ring of the diaryl ether core (e.g., the 30 position of the diaryl ether
core is equivalent to the 50 position on the naphthylene ring).
Cl
O
Cl
X
O
RO
Cl
c
R
+
X
Cl
7: R = OH
8: R = Tf
10: R = H, X = CH2CO2CH3 14: X = CH2CO2CH3
a
b
11: R = H, X = NO2
15: X = NO2
a
12: R = Tf, X = CH2CO2CH3
13: R = Tf, X = NO2
O
B
9: R =
O
Scheme 1. Synthesis of phenyl-naphthyl core structures 14 and 15. Reagents and conditions: (a) Et3N, triflic anhydride, CH2Cl2, ꢁ40 ꢁC to ꢁ10 ꢁC
(99%); (b)13 bis(pinacolato)diborane, anhyd potassium acetate, PdCl2(dppf), 80 ꢁC, 4 h (58%); (c) compound 14: Pd(PPh3)4, K2CO3, toluene/acetone
(1:1), 80 ꢁC, 4 h (9%); compound 15: Pd(PPh3)4, Na2CO3, DME/H2O (3:1), 80 ꢁC, 30 min (69%).
a,b
O
HO
R
Cl
O
Cl
OH
or
c,a,b
O
O
Cl
Cl
14
16: X = H
17: X = Br
Scheme 2. Synthesis of compounds 16 and 17. Reagents and conditions: (a) BBr3, CH2Cl2, ꢁ78 ꢁC, stirred at 0 ꢁC for 90 min; (b) NaOH in CH3OH,
rt, 16 h (94% over two steps); (c)22 48% HBr, acetic acid, DMSO, rt, 16 h (97%).
Table 1. IC50 and selectivity data of phenyl-naphthylene thyromimet-
Further SAR studies of the phenyl-naphthylene core
ics and select analogs
were carried out starting from compound 15 owing to
Compound
IC50(nM)a,b
TR-b
0.26
1.1
2140
Selc
its greater synthetic flexibility and ready availability
(Scheme 3). Modifications at C-1 were selected to in-
crease affinity based, in part, on previously established
SAR.18 The rank ordering of C-1 side-chain affinities
for TRb was succinic acid (20e) < acetic acid (17) < 2-
aminoacetic acid (20a) ꢀ 3-aminopropanoic acid
(20b) < oxalic acid (20c) ꢀ malonic acid (20d), with the
two most potent compounds having IC50 = 0.3 nM for
TRb. No significant change in isoform selectivity was
observed for any of these analogs. Two analogs of com-
pound 20a were synthesized (21 and 22 in Scheme 3) to
assess the effect that changes to substituents at the 50-po-
sition would have on binding affinity and agonist activ-
ity. As expected, the 50-methyl analog 21 was slightly
less potent than 20a (2-fold) and was about equally
selective for TRb. This result is consistent with the 50-
position being located in a hydrophobic region of the
TR LBD. The 50-phenyl analog 22 was less potent
for both TR isoforms and less selective for TRb.
Interestingly, compound 22 exhibited partial agonist
Tr-a
0.24
1
0.9
14
1
4
25
3605
109
1538
101
19
5
6
9
545
13
7
16
17
20a
20b
20c
20d
20e
21
22
26
2
5
1.8
6
13
1.8
0.3
0.3
4
5.4
4.3
62
8
9
22
7.2
29
0.3
2
48
125
14
4
3
25
a Competitive binding affinities versus radioiodinated T3 using full-
length cloned hTRa1 and hTRb1.8,17
b IC50 values determined from a minimum of duplicate data. Values
have an average variability of 25%.
c Selectivity = (IC50 hTRa1)/(IC50 hTRb1 · 1.7).8,17

4582
J. J. Hangeland et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4579–4584
H
O
O
Cl
Cl
N
NO2
R
a
b,c
15
Br
Br
Cl
Cl
18
19a: R = CH2CO2Et
19b: R = (CH2)2CO2Me
19c: R = COCO2Et
f,d,e
19d: R = COCH2CO2Et
19e: R = CO(CH2)2CO2Et
O
H
d,e
HO
Cl
N
OH
H
HO
Br
Cl
N
R
R
Cl
21: R = CH3
22: R = Ph
Cl
20a: R = CH2CO2H
20b: R = (CH2)2CO2H
20c: R = COCO2H
20d: R = COCH2CO2H
20e: R = CO(CH2)2CO2H
Scheme 3. Synthesis of compounds 20a–e, 21, and 22. Reagents and conditions: (a) Br2, CH2Cl2, rt, 3 h (89%); (b) iron powder, acetic acid containing
10% H2O (v/v), rt, 16 h (89%); (c) compound 19a: ethyl bromoacetate, K2CO3, acetonitrile, 80 ꢁC, 16 h (85%); compound 19b: methyl acrylate, acetic
acid, 120 ꢁC, 16 h (99% as a 1:1 mixture of methyl ester and free acid); compounds 19c–e: C2H5CO2(CH2)nCOCl (n = 0–2), Et3N, CH2Cl2, 0 ꢁC to rt,
2 h; (d) BBr3, CH2Cl2, ꢁ78 ꢁC, stirred at 0 ꢁC for 90 min; (e) NaOH in CH3OH, rt, 16 h; (f) compound 21:19 CH3B(OH)2, PdCl2(dppf), K3PO4, 1,4-
dioxane, 110 ꢁC (sealed tube), 16 h (58%); compound 22: PhB(OH)2, Pd(PPh3)4, Na2CO3, DME/H2O (3:1), 80 ꢁC, 30 min (45%).
O
O
Cl
Cl
NO2
NHR
a
c,d
15
X
Cl
Cl
23 X = O
25 R = H
b
e-g
24 X = CH2
26 R = COCH2CO2H
Scheme 4. Synthesis of compound 26. Reagents and conditions: (a)20 acetic anhydride, (CH3)2SBF3, CH2Cl2, ꢁ78 ꢁC to rt, stirred at rt for 16 h
(77%); (b)21 Nysted reagent, TiCl4, THF/CH2Cl2 (1:1), ꢁ78 ꢁC to rt, stirred at rt for 16 h (27%); (c) iron powder, acetic acid containing 10% H2O
(v/v), rt, 16 h (99%); (d) PtO2, H2 (1 atm), EtOH, 2 h (97%); (e) C2H5CO2CH2COCl, Et3N, CH2Cl2, 0 ꢁC to rt, 2 h (91%); (f) BBr3, CH2Cl2, ꢁ78 ꢁC,
stirred at 0 ꢁC for 90 min; (g) NaOH in CH3OH, rt, 16 h (80% over two steps).
activity for both TR isoforms (19% induction for TRa
and 50% induction for TRb).16 Replacing the 50-bro-
mide of compound 20d with an isopropyl group gave
compound 26 (Scheme 4). This modification retained
high affinity for both receptor isoforms and produced
the most TRb selective compound discovered in this ser-
ies (25-fold for TRb).
naphthyl ring systems found in the bound model of 26
was 113ꢁ, while fully relaxed ring system preferred 91ꢁ
(in vacuo, SAM1D24). The energy cost to assume the
113ꢁ rotamer is estimated to be 2–3 kcal/mol, which
should be easily accommodated by minor reorganiza-
tion of the hydrophobic residue side chains surrounding
the naphthyl system (i.e., F272, I276, L341, and L346).
A proposed binding mode for 26 is shown in Figure 4 as
an overlay with the X-ray structure for compound 4
complexed to TRb.8 The energy-minimized structure
(Amber force field as implemented within Flo23) of 26
was found to adopt a favorable 0 H-bonding distance
The binding data indicate that the SAR of the phenyl-
naphthylene series diverges from thyromimetics contain-
ing a diphenyl ether core. A comparison of compounds
5 and 16 suggests that, in the absence of a lipophilic
group projecting into the 30-binding pocket of the TR
LBD, the phenyl-naphthylene core has a modestly high-
er affinity. Addition of a lipophilic bromide to C-30 and
C-50 (i.e., compounds 6 and 17, respectively) resulted in
a greater increase of affinity for the diphenyl ether series
(237-fold for TRb) than for the phenyl-naphthylene
˚
between H435 and the pendant 6 -OH of 2.29 A. The
model suggests a unique deep pocket interaction
between C-20, C-30, and C-40 of the naphthylene moiety
and residues L341 and F272 of the receptor LBD. In
addition, the dihedral angle between the phenyl and

J. J. Hangeland et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4579–4584
4583
2. (a) Lazar, M. A. Endocr. Rev. 1993, 14, 184; (b) Forrest,
D. Semin. Cancer Biol. 1994, 5, 167.
´
3. (a) Wilkstro¨m, L.; Johansson, C.; Salto, C.; Barlow, C.;
Barrow, A. G.; Baas, F.; Forrest, D.; Thoren, P.;
´
Vennstro¨m, B. EMBO J. 1998, 17, 455; (b) Johansson,
´
C.; Go¨the, S.; Forrest, D.; Vennstro¨m, B.; Thoren, P. Am.
J. Physiol. 1999, 276, H2006.
4. (a) Trost, S. U.; Swanson, E.; Gloss, B.; Wang-Iverson, D.
B.; Zhang, H.; Volodarsky, T.; Grover, G. J.; Baxter, J.
D.; Chiellini, G.; Scanlan, T. S.; Dillmann, W. H.;
Wolfgang, H. Endocrinology 2000, 141, 3057; (b) Grover,
G. J.; Mellstro¨m, K.; Ye, L.; Malm, J.; Li, Y.-L.; Bladh,
L.-G.; Sleph, P. G.; Smith, M. A.; George, R.; Vennstro¨m,
B.; Mookhtiar, K.; Horvath, R.; Speelman, J.; Egan, D.;
Baxter, J. D. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
10067.
5. (a) Wagner, R. L.; Apriletti, J. W.; McGrath, M. E.; West,
B. L.; Baxter, J. D.; Fletterick, R. J. Nature 1995, 378, 690;
(b) Ribeiro, R. C. J.; Apriletti, J. W.; Wagner, R. L.; Feng,
F.; Kiushner, P. J.; Nilsson, S.; Scanlan, T. S.; West, B. L.;
Fletterick, R. J.; Baxter, J. D. J. Steroid Biochem. Mol.
Biol. 1998, 65, 133.
Figure 4. Overlay of compounds 4 (white) and 26 (magenta) within the
ligand binding site of TRb. The structure of 26 was modeled after the
binding mode of 4 as found in the published X-ray structure8 (Protein
Data Bank entry 1NAX).
6. Jorgensen, E. C.; Berteau, P. E. J. Med. Chem. 1972, 14,
1199, and the previous papers in this series.
7. (a) Yoshihara, H. A. I.; Chiellini, G.; Mitchison, T. J.;
Scanlan, T. S. Bioorg. Med. Chem. 1998, 6; (b) Yoshihara,
H. A. I.; Apriletti, J. W.; Baxter, J. D.; Scanlan, T. S.
Bioorg. Med. Chem. Lett. 2001, 11, 2821.
series (42-fold for TRb). This suggests that the
conformationally flexible diphenyl ether core is able to
orient this group more optimally within the 30-binding
pocket of the LBD. This latter point is of consequence
when considering the phenyl-naphthylene core, or its
analogs, for further design of thyromimetics having
either isoform selectivity or agonist versus antagonist
properties. For example, analogs of compound 4, con-
taining a 30-phenyl or substituted phenyl group, were
selective for TRb (up to 38-fold) and were full agonists
for both isoforms.9c In contrast, compound 22, which
contains a 50-phenyl group, is the least selective analog
of the three closely related compounds (e.g., compounds
20a, 21, and 22). Indeed, the most selective compound
discovered (26; 25-fold for TRb) contains a 30-isopropyl
group, which is not commonly thought to contribute to
increased isoform selectivity.8,9 Furthermore, the partial
agonist activity exhibited by compound 22 suggests this
position to be more sensitive to changes in steric bulk,
both with respect to selectivity and agonist versus antag-
onist properties. Thus, this new class of thyromimetic
may provide additional opportunities to discover TR
ligands with improved separation of desired pharmaco-
logical outcome from unwanted cardiac side effects.
8. Ye, L.; Li, Y.-L.; Mellstro¨m, K.; Bladh, L.-G.; Koehler,
K.; Garg, N.; Garcia Collazo, A. M.; Litten, C.; Husman,
B.; Persson, K.; Ljunggren, J.; Grover, G.; Sleph, P. G.;
George, R.; Malm, J. J. Med. Chem. 2003, 46, 1580.
9. (a) Dow, R. L.; Schneider, S. R.; Paight, E. S.; Hank, R.
F.; Chaing, P.; Cornelius, P.; Lee, E.; Newsome, W. P.;
Swick, A. G.; Spitzer, J.; Hargrove, D. E.; Patterson, T.
A.; Pandit, J.; Chrunyk, B. A.; LeMotte, P. K.; Danley, D.
E.; Rosner, M. H.; Ammirati, M. J.; Simones, S. P.;
Schulte, G. K.; Tate, B. F.; DaSilva-Jardine, P. Bioorg.
Med. Chem. Lett. 2003, 13, 379; (b) Borngraeber, S.;
Budny, M.-J.; Chiellini, G.; Cunha-Lima, S. T.; Togashi,
M.; Webb, P.; Baxter, J. D.; Scanlan, T. S.; Fletterick, R.
J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100; (c) Hangeland,
J. J.; Doweyko, A. M.; Dejneka, T.; Friends, T. J.;
Devasthale, P.; Mellstro¨m, K.; Sandberg, J.; Grynfarb,
˚
M.; Sack, J. S.; Einspahr, H.; Fa¨rnegardh, M.; Husma, B.;
Ljunggren, J.; Koehler, K.; Sheppard, C.; Malm, J.;
Ryono, D. E. Bioorg. Med. Chem. Lett. 2004, 14, 3549.
10. The description of the conformational relationship
between the two phenyl rings has been adapted from:
Cody, V. J. Am. Chem. Soc. 1974, 96, 6720.
11. Chiellini, G.; Apriletti, J. W.; Yoshihara, H. A. I.; Baxter, J.
D.; Riberio, R. C.; Scanlan, T. S. Chem. Biol. 1998, 5, 299.
12. Kasturi, T. R.; Arunachalam, T. Can. J. Chem. 1968, 46,
3625.
13. Ishiyama, T.; Murata, M.; Miyaura, M. J. Org. Chem.
1995, 60, 7508.
14. Compounds 5 and 6 were synthesized using procedures
reported in Refs. 8 and 9c.
Acknowledgments
15. Dibbo, A.; Stephenson, L.; Walker, T.; Warburton, W. K.
J. Chem. Soc. Abstr. 1961, 2645.
The authors thank Dr. Minsheng Zhang, Dr. Johan
Malm, Ms. Yolanda Caringal, and Ms. Tamara Dejene-
ka for helpful suggestions related to this work.
16. Percent induction of reporter gene transcription relative to
T3, which is normalized to 100%. Percent gene induction is
measured using CHO cells transfected separately with
human TR-a1 or human TR-b1 gene constructs and a
reporter gene coding for soluble alkaline phosphatase
under control of a thyroid hormone receptor response
element.8,17
References and notes
1. Brent, G. In Endocrinology: Basic and Clinical Principles;
Conn, M. P., Melmed, S., Eds.; Humana Press: Totowa,
NJ, 1997, Chapter 19.
17. (a) Carlsson, B.; Singh, B. N.; Temciuc, M.; Nilsson, S.;
Li, Y.-L.; Mellin, C.; Malm, J. J. Med. Chem. 2002, 45,

4584
J. J. Hangeland et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4579–4584
623; (b) Barkhem, T.; Carlsson, B.; Simons, J.; Moeller,
˚
B.; Berkenstam, A.; Gustafsson, J.-A.; Nilsson, S. J.
Steroid Biochem. Mol. Biol. 1991, 38, 667.
21. (a) Matsubara, S.; Sugihara, M.; Utimoto, K. Synlett
1998, 313; (b) Imai, M.; Tanaka, M.; Suemune, H.
Tetrahedron 2001, 57, 1205.
18. Yokoyama, N.; Walker, G. N.; Main, A. J.; Stanton, J. L.;
Morrissey, M. M.; Boehm, C.; Engle, A.; Neubert, A. D.;
Wasvary, J. M.; Stephan, Z. F.; Steele, R. E. J. Med.
Chem. 1995, 38, 695.
22. Majetich, G.; Hicks, R.; Reister, S. J. Org. Chem. 1997,
62, 4321.
23. Flo/QXP molecular modeling program; Colin McMartin,
Thistlesoft, Colebrook, CT.
19. Harada, T.; Yoshida, T.; Inoue, A.; Takeuchi, M.; Oku,
A. Synlett 1995, 283.
20. Kiselyov, A. S.; Harvey, R. G. Tetrahedron Lett. 1995, 36,
4005.
24. Semi-empirical molecular orbital calculations carried
out using SAM1D within the AMPAC package as
implemented within Sybyl Version 7.0 (Tripos, Inc., St.
Louis, MO).