N2-Benzyl-N1-(1-(1-naphthyl)ethyl)-3-phenylpropane-1, 2-diamines and conformationally restrained indole analogues: Development of calindol as a new calcimimetic acting at the calcium sensing receptor

Article abstract of DOI:10.1016/j.bmcl.2004.03.056

The synthesis and calcimimetic activities of two new families of compounds are described. The most active derivatives of the first family, N ²-(2-chloro-(or 4-fluoro-)benzyl)-N¹-(1-(1-naphthyl)ethyl) -3-phenylpropane-1,2-diamine (4b

Full text of DOI:10.1016/j.bmcl.2004.03.056

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3345–3349
N 2-Benzyl-N1-(1-(1-naphthyl)ethyl)-3-phenylpropane-1,2-diamines
and conformationally restrained indole analogues: development
of calindol as a new calcimimetic acting at the calcium
sensing receptor
a
b
b
b
a,
*
ꢀ ꢁ
Albane Kessler, Helene Faure, Christophe Petrel, Martial Ruat, Philippe Dauban
and Robert H. Dodda,*
aInstitut de Chimie des Substances Naturelles, UPR 2301 Centre National de la Recherche Scientifique,
91198 Gif-sur-Yvette Cedex, France
bInstitut de Neurobiologie Alfred Fessard, IFR 2118 CNRS, Laboratoire de Neurobiologie Cellulaire et Molꢀeculaire,
UPR 9040 CNRS, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette, France
Received 12 November 2003; revised 12 February 2004; accepted 1 March 2004
Abstract—The synthesis and calcimimetic activities of two new families of compounds are described. The most active derivatives of
the first family, N2-(2-chloro-(or 4-fluoro-)benzyl)-N1-(1-(1-naphthyl)ethyl)-3-phenylpropane-1,2-diamine (4b and 4d, respectively,
tested at 10 lM) produced 98 6% and 95 4%, respectively, of the maximal stimulation of [3H]inositol phosphates production
obtained by 10 mM Ca2þ in CHO cells expressing the rat calcium sensing receptor (CaSR). The second family of calcimimetics was
obtained by conformationally restraining the compounds of type 4 to provide the 2-aminomethyl derivatives 5. One of these
compounds, (R)-2-[N-(1-(1-naphthyl)ethyl)aminomethyl]indole ((R)-5a, calindol), displayed improved calcimimetic activity com-
pared to 4b and 4d as well as stereoselectivity. In the presence of 2 mM Ca2þ, calindol stimulated [3H]inositol phosphates accu-
mulation with an EC50 of 1.0 0.1 or 0.31 0.05 lM in cells expressing the rat or the human CaSR, respectively. The calcimimetic
activities of these novel compounds were shown to be due to a specific interaction with the CaSR.
ꢀ 2004 Elsevier Ltd. All rights reserved.
The extracellular calciumsensing receptor (CaSR) is a G
protein-coupled receptor (GPCR), which senses extra-
acid B receptor as well as certain pheromone and taste
receptors.1;2 CaSR cDNA has been cloned fromsuch
tissues as the parathyroid, thyroid, kidney and brain.3–5
In the particular case of the parathyroid cell, the CaSR
at the surface detects and responds to small changes of
circulating [Ca2þ]e, thereby regulating parathyroid
2þ
cellular calcium[Ca
]
and regulates calciumion
e
homeostasis.1 The CaSR possesses a long amino termi-
nal tail typical of family 3 of GPCRs, which includes the
metabotropic glutamate receptors, the c-aminobutyric
H H
C
NH
N
R
H
O O
H
MeO
MeO
S
NH
4
5
N
N
OCH3
R
H
R
H
1 : R = Cl NPS 568
2 : R = H NPS 467
3
N
N
H
* Corresponding authors. Tel.: +33-1-69-82-45-94; fax: +33-1-69-07-72-47; e-mail: robert.dodd@icsn.cnrs-gif.fr
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.056

3346
A. Kessler et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3345–3349
hormone (PTH) secretion.1 Inactivating mutations of
the CaSR gene have been identified as the cause of
hypocalciuric hypercalcemia while activating mutations
lead to hypocalcemia.6
could have important repercussions on tissue distribu-
tion, metabolism, activity and selectivity. For the prep-
aration of these compounds, we chose as starting
material the N-(p-nitrophenylsulfonyl)aziridine 7 easily
synthesized, as shown in Scheme 1, by reaction of al-
lylbenzene 6 in acetonitrile with 1 equiv of [(N-p-nitro-
phenylsulfonyl)imino]phenyliodane in the presence
of copper(I) triflate as catalyst.22–24 The N-(p-nitro-
phenylsulfonyl) (N-nosyl) group presents the advantages
of providing highly reactive aziridines, of being easily
removed after aziridine ring opening25 and, in our par-
ticular case, of allowing selective alkylation of the N-
sulfonyl nitrogen. Thus, compound 7 reacted smoothly
with 1-(1-naphthyl)ethylamine in THF containing
0.1 equiv of triethylamine to give 60% of diamine 8. The
sulfonamide nitrogen could then be selectively alkylated
by 1.2–1.5 equiv of the appropriate benzyl bromide in
DMF using potassiumcarbonate as base, providing
compounds 9. Finally, removal of the N-(p-nitro-
phenylsulfonyl) group of each of the latter compounds
was effected by treatment with 3 equiv of thiophenol and
potassiumcarbonate at 50 ꢁC in acetonitrile containing
2% DMSO.
In addition to calcium, cloned CaSR is activated by
other divalent cations such as magnesium and by poly-
amines such as spermine, spermidine, aminoglycosides
(e.g., neomycin) and by b-amyloid peptides.7–11 In the
search for more selective CaSR ligands, the synthetic
phenylalkylamines, NPS 568 (1), and NPS 467 (2) have
been found to allosterically mimic or potentiate the ac-
8;12;13
tion of [Ca2þ
]
and to interact at the level of the
e
transmembrane domain.14 For instance, such calcimi-
metics are able to increase the concentration of cyto-
plasmic calcium ([Ca2þ]i) in bovine parathyroid cells and
inhibit PTH secretion.15;16
The therapeutic potential of NPS 568 was suggested by
experiments in rats in which this compound reduced or
eliminated osteitis fibrosa17 and led to an important
reduction of cell proliferation in the parathyroid gland
in a model reproducing chronic renal deficiency.18 PTH
levels were also decreased by NPS 568 in patients dis-
playing this same disorder19 as well as in menopaused
women exhibiting primary hyperparathyroidism.20 With
the aim of developing calcimimetics displaying improved
activities, tissue specificity and/or pharmacodynamic
properties, we recently described21 the synthesis of a new
family of compounds, the N2-arylsulfonyl-N1-(1-(1-
naphthyl)ethyl)-3-phenylpropane-1,2-diamines, exem-
plified by the 3,4-dimethoxy derivative 3 whose calci-
mimetic activities were comparable to that of NPS 568.
In this communication, we describe the use of CaSR
ligands of type 3 as a starting point for the development
of two further families of calcimimetics, their N-benzyl
analogues of general structure 4 and the 2-substituted
indole derivatives 5, which may be considered as rigid
analogues of 3 and of NPS 568 (1).
The resulting compounds 4a–f, in the formof their
hydrochloride salts, were then evaluated for their calci-
mimetic activity in Chinese hamster ovarian (CHO) cells
transfected [CHO(CaSR)] or not [CHO(WT*)] with an
expression vector containing cloned CaSR fromrat
brain7 previously shown to be useful for characterizing
calcimimetics.8;21 Activation of the CaSR by Ca2þ as
well as by NPS 568 has been shown to stimulate PLC
activity and accumulation of tritiated inositol phosphate
([3H]IP) in CHO(CaSR) cells prelabelled with myo-
[3H]inositol.7 The [3H]IP accumulation produced by
compounds 4a–f (10 lM) in the presence of 2 mM
[Ca2þ]e was measured and compared to the effect pro-
duced by NPS 568 as well as to the effect produced by
10 mM [Ca2þ]e (a concentration resulting in maximal
CaSR activation). As shown in Table 1 and Figure 1, all
the N-benzyl derivatives 4 prepared demonstrated calci-
mimetic activity, producing at 10 lM concentration
between approximately 70% and 100% of the maximal
response obtained by 10 mM Ca2þ. The effects of these
We first compared the effect on calcimimetic activity of
replacing the polar arylsulfonyl function of the more
active analogues of type 3 by a more lipophilic benzyl
group (i.e., ArSO2 fi ArCH2). Such a transformation
PhI=NSO2Ph-p-NO2
CuOTf, CH3CN
THF, Et3N
O O
S
NH
N
4A molecular sieves, 0˚C
N
O
O
H
S
O2N
6
7
NH2
8
R
DMF, K
2CO3
Br
NO2
PhSH, K2CO3
CH
3CN, DMSO, 50˚C
O O
S
NH
NH
N
N
H
R
O2N
4a - 4f
9
R
Scheme 1.

A. Kessler et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3345–3349
3347
Table 1. Accumulation of [3H]IP induced by the test compounds in CHO(CaSR) cells expressing rat cloned CaSR
R
H2
H
H
C
N
N
N
X
H
N
R
.HCl
.HCl
R'
4
5, 10, 13a, 14
Compound no. (10 lM)
R
R0
X
% [3H]IP accumulationa
1 (NPS 568)
4a
––
H
––
––
––
––
––
––
––
H
––
––
105 10
80 8 (78 7)b
98 6 (79 7)
78 6 (83 3)
95 4 (62 6)
80 6 (80 8)
74 17 (97 11)
4b
4c
2-Cl
3-Cl
4-F
––
––
4d
4e
––
––
4-OMe
3,4-Di-OMe
4f
––
(R,S)-5a
(R)-5a
(S)-5a
5b
H
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
C@O
CH2
103
117
32
87
60
55
34
54
2
6
5
H
H
H
5-F
H
5
7
1
5
1
5
H
5c
5d
5-Cl
5-OMe
5,6-Di-OMe
H
H
5e
10
H
SO2Ph(3,4-OMe)
H
H
H
13a
14
H
CH3
3
79 13
a CHO(CaSR) cells expressed rat Ca2þ sensing receptor cloned fromrat brain. 5 These cells were cultured in basal Ham’s F-12 medium (0.3 mM Ca2þ
,
0.6 mM Mg2þ) as previously described.7 Cells, in 24-well plates, were cultured overnight in their growth medium containing 0.5 lCi/mL myo-
[3H]inositol (Amersham Corp.). Drugs were dissolved at 10 mM in ethanol and then diluted in basal medium. Cells were incubated with drugs
(10 lM) at 37 ꢁC and activation of the phospholipase C enzyme was estimated after quantification of [3H]inositol phosphates ([3H]IP) accumulation
that was performed as described.7 Results are expressed as percentage of the maximal response obtained by 10 mM Ca2þ and are means SEM of
two to five independent experiments performed in triplicate. Results varied by less than 5% in any given experiment.
b Values in parentheses refer to the % [3H]IP accumulation for the equivalent sulfonamides at the same concentration; from Ref. 21.
700
600
500
400
300
200
100
0
unsubstituted (4a), 3-chloro (4c) and 4-methoxy deriv-
atives (4e) all stimulated [3H]IP accumulation by about
80%, while both the 2-chlorobenzyl (4b) and 4-fluoro-
benzyl (4d) derivatives were slightly more active (98 6%
and 95 4% stimulation, respectively). On the other
hand, the 3,4-dimethoxy substituents, shown to provide
the optimal calcimimetic activity in the N-sulfonyl
series21 produced a slight diminution of activity in the
N-benzyl series (74 17% stimulation for 4f vs 97 11%
for 3). The calcimimetic activities of the most active of
these N-benzyl derivatives (4b,d) were comparable,
though slightly inferior, to that of NPS 568 at the same
concentration of 10 lM (105 10% stimulation). When
compounds 4b–4d were tested in CHO(WT*) cells at the
same concentration (10 lM) (Fig. 1), no significant
increase of [3H]IP accumulation was measured thereby
indicating the specificity of the observed response in
CHO(CaSR) cells with this new series of calcimimetics
as was the case with 3.21
CHO(WT*)
CHO(CaSR)
10 mM NPS 4b
4c
4d
5a
5a
5a
(S)
5b
5c
5d
10
14
Ca2+
568
(R/S) (R)
Figure 1. Effects of test compounds of general structure 4 and 5 and of
NPS 568 in CHO(CaSR) cells expressing rat cloned CaSR or in
CHO(WT*) cells. CHO(CaSR) or CHO(WT*) cells were prelabeled
with myo-[3H]inositol in basal Ham’s F-12 medium as described in
legend to Table 1. [3H]IP accumulation was determined for each test
compound and for NPS 568 at 10 lM and in presence of 2 mM [Ca2þ]e.
The effect of 10 mM Ca2þ is presented for comparison. Results are
expressed as % of [3H]IP basal level and are means SEM of three to
five independent experiments performed in triplicate. In CHO(WT*)
cells, test compounds and NPS 568 have limited effects on [3H]IP basal
level (less than 40% of basal value) whereas they produced robust
accumulation of [3H]IP in CHO(CaSR) cells suggesting that these
compounds stimulate phospholipase C activity by interacting specifi-
cally with the CaSR.
While in vivo the N-benzyl derivatives may present
pharmacokinetic and/or pharmacodynamic advantages
over their N-sulfonyl analogues, it still seemed evident
that it would be advantageous to obtain calcimimetic-
type compounds demonstrating higher intrinsic activi-
ties with respect to the CaSR. In an effort to achieve this,
we turned our attention to the study of conformation-
ally restricted analogues of 3 and 4. Though there are
many ways in which a highly flexible molecule such as 3
can be made more rigid, an indole derivative of type 10,
compounds were, moreover, of the same order of mag-
nitude as the N-sulfonyl analogues 321 having the same
pattern of substitution on the phenyl ring. Thus, the

3348
A. Kessler et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3345–3349
H
N
N
MeO
MeO
X
NH
MeO
MeO
X
N
H
3 : X = SO
10: X = SO
2
2
4f :X = CH
2
Scheme 2.
in which the nitrogen atomof the N-sulfonyl group is
covalently bonded to the 3-phenyl ring, seemed partic-
ularly attractive since the resulting 2-substituted indole
derivative (Scheme 2) should be easily accessible, thus
allowing rapid validation of the working hypothesis.
The R and S enantiomers of the most active indole
derivative, racemic 5a, were then prepared separately
fromthe corresponding ( R)- and (S)-aminoethylnaph-
thalenes and their calcimimetic activities were compared
in CHO(CaSR) cells. As shown in Table 1, (R)-5a
(calindol) was shown to be considerably more active
than (S)-5a (117 5% vs 32 5% stimulation, respec-
tively) and as active as NPS 568 (105 10% stimulation,
Table 1 and Fig. 1). Increasing concentrations of (R,S)-
5a, (R)-5a or (S)-5a led to a dose-dependent increase of
[3H]IP accumulation in CHO(CaSR) cells (Fig. 2A),
confirming that (R)-5a is slightly more effective than the
racemic compound but at least 20-fold more-so than
(S)-5a. Both compounds gave a maximal stimulation
representing 130% of the 10 mM Ca2þ activation. We
then determined the potency of the more active com-
pound in eliciting the inositol phosphates response in
HEK293 cells transiently expressing the human CaSR.
Analysis of the dose response curve (Fig. 2B) gave an
EC50 for (R)-5a of 0.31 0.05 lM with a maximal
stimulation comparable to that obtained with 10 mM
Ca2þ. Moreover, when these indolic calcimimetics (5a–d,
10, 14) were tested on CHO(WT*) cells, [3H]IP accu-
mulation was very limited indicating that the CaSR was
responsible for the observed response in CHO(CaSR)
cells (Fig. 1). These data indicate that calindol is a po-
tent and stereoselective calcimimetic of both the rat and
the human CaSR.
Compound 10 was thus synthesized in straightforward
fashion as shown in Scheme 3. Treatment of indole-2-
carboxylic acid 11a with thionyl chloride in chloroform
afforded the acid chloride 12a. The latter was reacted
with 1-(1-naphthyl)ethylamine in chloroform in the
presence of triethylamine to give the carboxamide 13a,
which was reduced to the amine 5a with lithiumalu-
minum hydride/aluminum chloride in THF. Treatment
of this compound with 3,4-dimethoxybenzenesulfonyl
chloride then gave 10.
Evaluation of compound 10 for calcimimetic activity in
CHO(CaSR) cells showed, however, that it was con-
siderably less active than the unconstrained analogue 3
(54 5% stimulation of IP production for 10 vs
97 11% for 3 at 10 lM concentration) (Table 1).
Interestingly, investigation of the precursor of com-
pound 10 having the N-unsubstituted indole moiety
(compound 5a) showed that it was a considerably more
efficient stimulator of the IP response (103 6% at
10 lM). Other analogues of 5a having substituents on
the indole ring were subsequently prepared in the same
manner (compounds 5b–e), and tested for their calci-
mimetic properties. Substitution of the C-5 indole
position by halogens (F, Cl) or by an alkoxy group
(MeO) led to loss of activity relative to 5a as did
introduction of 5,6-dimethoxy groups (Table 1). Indole
N-methylation was also found to be detrimental to
activity (14: 79 13% stimulation) while the precursor
carboxamide 13a was inactive.
In conclusion, this study describes the discovery of two
new types of CaSR ligands demonstrating calcimimetic
activity. Both series of compounds were derived from
the previously described N-arylsulfonyl family of calci-
mimetics, the most active of which was the dimethoxy
analogue 3. In the first group of compounds, replace-
ment of the N-arylsulfonyl group by an N-benzyl group
R
R
R
NH2
SOCl2
CHCl3
H
N
CO H
COCl
N
H
N
H
N
H
2
CHCl3 , Et3N
O
11a : R = H
11b - 11e : R = H
12a : R = H
12b - 12e : R = H
13a : R = H
13b - 13e : R = H
LiAlH4, AlCl3, THF
R
R
R'X , base
H
N
H
N
N
N
H
'
R
O S
OCH3
OCH3
5a : R = H
13b - 13e : R = H
10: R = H; R' =
2
14: R = H; R' = CH3
Scheme 3.

A. Kessler et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3345–3349
3349
pending a thorough study of their structure–activity
relationships and their effects in vivo.
140
120
100
80
rat CaSR
NPS 568
(S)-5a
References and notes
(R)-5a
1. Brown, E. M.; MacLeod, R. J. Physiol. Rev. 2001, 81, 239.
ꢀ
2. Pin, J. P.; Galvez, T.; Prezeau, L. Pharmacol. Ther. 2003,
98, 325.
60
(R,S)-5a
40
3. Brown, E. M.; Gamba, G.; Riccardi, D.; Lombardi, M.;
Butters, R.; Kifor, O.; Sun, A.; Hediger, M.; Lytton, J.;
Hebert, S. C. Nature 1993, 366, 575.
4. Riccardi, D.; Park, J.; Lee, W. S.; Gamba, G.; Brown, E.
M.; Hebert, S. C. Proc. Natl. Acad. Sci. U.S.A. 1995, 92,
131.
20
0
-8
-7
-6
-5
-4
(A)
Log[drug], M
5. Ruat, M.; Molliver, M. E.; Snowman, A. M.; Snyder, S.
H. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 3161.
6. Hendy, G. N.; D’Souza-Li, L.; Yang, B.; Canaff, L.; Cole,
D. E. C. Human Mutat. 2000, 16, 281.
7. Ruat, M.; Snowman, A. M.; Hester, L. D.; Snyder, S. H.
J. Biol. Chem. 1996, 271, 5972.
8. Ferry, S.; Chatel, B.; Dodd, R. H.; Lair, C.; Gully, D.;
Maffrand, J.-P.; Ruat, M. Biochem. Biophys. Res. Com-
mun. 1997, 238, 866.
9. Brown, E. M.; Butters, R.; Katz, C.; Kifor, O. Endocri-
nology 1991, 128, 3047.
10. Ye, C.; Ho, C.; Kanazinska, M.; Quinn, S.; Rogers, K.;
Seidman, C. E.; Seidman, J. G.; Brown, E. M.; Vassilev, P.
M. J. Neurosci. Res. 1997, 47, 547.
11. McLarnon, S.; Holden, D.; Ward, D.; Jones, M.; Elliott,
A.; Riccardi, D. Biochem. Biophys. Res. Commun. 2002,
297, 71.
100
80
60
40
20
0
human CaSR
-8
-7
-6
-5
-4
(B)
Log [(R)-5a], M
12. Nemeth, E. F.; Steffey, M. E.; Hammerland, L. G.; Hung,
B. C. P.; Van Wagenen, B. C.; DelMar, E. G.; Balandrin,
M. F. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 4040.
13. Mailland, M.; Waelchli, R.; Ruat, M.; Boddeke, H. G.;
Seuwen, K. Endocrinology 1997, 138, 3601.
14. Hu, J.; Reyes-Cruz, G.; Chen, W.; Jackson, K. A.; Spiegel,
A. M. J. Biol. Chem. 2002, 277, 46622.
15. Nemeth, E. F.; Fox, J. Trends Endocrinol. Metab. 1999,
10, 66.
16. Goodman, W. G. Nephro. Dial. Transplant 2002, 17, 204.
17. Wada, M.; Ishii, H.; Furuya, Y.; Fox, J.; Nemeth, E. F.;
Nagano, N. Kidney Int. 1998, 53, 448.
Figure 2. Potency of 5a compounds and NPS 568 in stimulating
accumulation of [3H]IP in (A) CHO(CaSR) cells stably expressing rat
CaSR or (B) HEK293 cells transiently transfected26 with the human
CaSR. Cells were prelabeled with myo-[3H]inositol in basal Ham’s
F-12 medium as described in legend to Table 1. [3H]IP accumulation
induced by racemic (R,S)-5a or the (R)-5a and (S)-5a enantiomers was
determined in presence of 2 mM [Ca2þ]e as indicated. For NPS 568, the
determination was made in the presence of 2.3 mM Ca2þ. Results are
expressed as % of 10 mM Ca2þ response and are representative of three
to five independent experiments performed in triplicate. In CHO
(CaSR), the (R)-5a (calindol) and (R,S)-5a compounds display an EC50
of 1.0 0.1 and of 1.3 0.2 lM, respectively, and produced a similar
maximum stimulation representing 130% of the 10 mM Ca2þ activa-
tion. In HEK293 cells transfected with the human CaSR, (R)-5a
(calindol) displays an EC50 of 0.31 0.05 lM.
18. Wada, M.; Furuya, Y.; Sakiyama, J.; Kobayashi, N.;
Mitaya, S.; Ishii, H.; Nagano, S. J. Clin. Invest. 1997, 100,
2977.
19. Antonsen, J. E.; Sherrard, D. J.; Andress, D. L. Kidney
Int. 1998, 53, 223.
provided equally active calcimimetics (4b,d). An inter-
esting observation was that the pattern of substitution of
the N-2 aryl ring assuring maximal activity of the sul-
fonyl derivatives (e.g., the 3,4-dimethoxy derivative 3)
was not the same as for the N-benzyl derivatives (e.g.,
the 2-chloro derivative 4b). Moreover, in an effort to
further improve the efficacy of these compounds, rigid
analogues in the formof the 2-aminomethyl indole
derivatives 5 were prepared. One of these, calindol ((R)-
5a), evaluated in both rat and human CaSR, was found
to be one of the most active calcimimetic compounds
known to date. In view of the very restricted number of
such compounds so far described, both series of ligands
offer new perspectives for the development of clinically
useful calcimimetics acting specifically at the CaSR
20. Silverberg, S. J.; Bone, H. G.; Marriott, T. B.; Locker, F.
G.; Thys-Jacobs, S.; Dziem, G.; Kaatz, S.; Sanguinetti, E.
L.; Bilezikian, J. P. New Engl. J. Med. 1997, 337, 1506.
21. Dauban, P.; Ferry, S.; Faure, H.; Ruat, M.; Dodd, R. H.
Bioorg. Med. Chem. Lett. 2000, 10, 2001.
22. Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Am. Chem.
Soc. 1994, 116, 2742.
€
23. Sodergren, M. J.; Alonso, D. A.; Bedekar, A. V.;
Andersson, P. G. Tetrahedron Lett. 1997, 38, 6897.
24. Dauban, P.; Dodd, R. H. Tetrahedron Lett. 1998, 39,
5739.
25. Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett.
1995, 36, 6373.
26. Petrel, C.; Kessler, A.; Maslah, F.; Dauban, P.; Dodd, R.
H.; Rognan, D.; Ruat, M. J. Biol. Chem. 2003, 278,
49487–49494.