Replacement of the lactone moiety on podophyllotoxin and steganacin analogues with a 1,5-disubstituted 1,2,3-triazole via ruthenium-catalyzed click chemistry

Article abstract of DOI:10.1016/j.bmc.2007.08.020

Steganacin and podophyllotoxin are two naturally occurring lignans first isolated from plant sources, which share the capability to disrupt tubulin assembly. Although not strictly essential for its activity, the lactone ring on both structures represents

Full text of DOI:10.1016/j.bmc.2007.08.020

Bioorganic & Medicinal Chemistry 15 (2007) 6748–6757
Replacement of the lactone moiety on podophyllotoxin and
steganacin analogues with a 1,5-disubstituted 1,2,3-triazole
via ruthenium-catalyzed click chemistry
Daniela Imperio, Tracey Pirali, Ubaldina Galli, Francesca Pagliai, Laura Cafici,
Pier Luigi Canonico, Giovanni Sorba, Armando A. Genazzani and Gian Cesare Tron*
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche and Drug and Food Biotechnology Center,
Universit a` degli Studi del Piemonte Orientale ‘A. Avogadro’, Via Bovio 6, 28100 Novara, Italy
Received 2 April 2007; revised 30 July 2007; accepted 3 August 2007
Available online 22 August 2007
Abstract—Steganacin and podophyllotoxin are two naturally occurring lignans first isolated from plant sources, which share the
capability to disrupt tubulin assembly. Although not strictly essential for its activity, the lactone ring on both structures represents
Achilles’ heel, as it is a potential site of metabolic degradation and epimerization on its C2 carbon brings about a significant loss in
potency. In the present manuscript, we have used the ruthenium-catalyzed [3+2] azide–alkyne cycloaddition, a click-chemistry reac-
tion, to replace the lactone ring with a 1,5-disubstituted triazole in few synthetic steps. The compounds were cytotoxic, although to a
lesser degree compared to podophyllotoxin, while retaining antitubulin activity. The present structures might therefore represent
a good platform for the fast generation of metabolically stable compounds with few stereogenic centers that might be of value from
a medicinal chemistry point of view.
ꢀ
2007 Elsevier Ltd. All rights reserved.
1. Introduction
(
are two naturally occurring lignans first isolated from
À)-Steganacin (1) and (À)-podophyllotoxin (2) (Fig. 1)
1
plant sources. Both compounds inhibit the assembly
of tubulin into microtubules by interacting with the col-
chicine binding site and have been shown to possess
2
cytotoxic activity against several cancer cell lines. In
addition, podophyllotoxin and its congeners have also
been shown to possess other activities, for example anti-
3
viral and antirheumatic.
Figure 1. Steganacin and podophyllotoxin.
Over the years, a series of analogues of podophyllotoxin
and steganacin have been synthesized and a robust SAR
4
,5
is now available. The structure–activity relationships
of these two 5-ring compounds share a number of simi-
larities: (i) the dioxolane moiety is fundamental for the
toxin, the hydroxyl at the 4-position is not essential as 4-
deoxypodophyllotoxin maintains similar cytotoxic
8
6
properties in respect to its parent compound ; (iv) in ste-
ganacin, the acetate is not essential as stegane is equally
cytotoxic activity ; (ii) the lactone ring is not essential,
7
although it contributes to potency ; (iii) in podophyllo-
5
a
active ; and (v) the trimethoxyphenyl ring is essential in
both compounds for cytotoxic and antitubulic
4
,5
activities.
Keywords: Podophyllotoxin; Steganacin; Tubulin; Antiproliferative
activity; Click chemistry; 1,5-Disubstituted 1,2,3-triazole.
*
Corresponding author. Tel.: +39 0321 375857; fax: +39 0321
57821; e-mail: tron@pharm.unipmn.it
Although the lactone group is not strictly required on
podophyllotoxin, when this undergoes epimerization
3
0
968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.08.020

D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
6749
1
6
and freshly purified copper (I) bromide as catalyst
Scheme 3). The ethylnyltrimethylsilyl magnesium bro-
(
mide was obtained at 0 ꢁC by adding a solution of eth-
ylmagnesium bromide dropwise to ethynyltrimethyl-
silane in THF. After stirring at room temperature for
30 min, catalytic amounts of copper (I) bromide were
added and, after 30 min more, the corresponding bro-
mides 3 or 7 were added and the resulting solution
was heated at reflux for 16 h to obtain excellent yields
of the desired acetylenic derivatives (5 and 9). Reaction
times and the reagent amounts (ethynyltrimethylsilane/
ethylmagnesium bromide/bromide derivatives in ratio
Scheme 1. The copper or ruthenium 1,3-dipolar cycloaddition between
azides and alkynes.
4:4:1) were fundamental to obtain reproducible results
in high yields.
(
to yield the more thermodynamically stable C2
9
The two alkynes were then deprotected using the classi-
cal protocol (TBAF, THF, 0 ꢁC) but, to our surprise,
after the cleavage of the silyl group, an alkyne–allene
isomerization took place giving the corresponding allen-
ic derivatives. In our opinion, this might have been due
to the basicity of the fluoride ion, which might have been
epimer picropodophyllotoxin), the resulting compound
displays a significantly reduced cytotoxicity and antitub-
ulinic activity.
phenyl group in a quasi-axial position in respect to the
tetracyclic core is fundamental for potency and this is
only possible when the lactone ring is in a trans-junction.
Epimerization is rather common, as it occurs in slightly
basic medium and also under neutral conditions. Fur-
thermore, the lactone ring might be vulnerable to meta-
bolic cleavage and also for this reason it has been
4
,10
Indeed, the presence of the trimethoxy-
1
7
enough to catalyze this process. For this reason, we
1
8
carried out the deprotection using silver nitrate or
TBAF in acetic acid to give the desired alkynes 6 and
10, suppressing the isomerization (Scheme 3).
1
1
1
2
replaced by other functional groups.
With these intermediates in our hands, we performed the
click-chemistry reaction using Cp*Ru(PPh ) Cl as cata-
3
2
Among the most recent exciting developments in medic-
inal chemistry is the use of copper or ruthenium com-
pounds to regioselectively and efficiently catalyze the
lyst in refluxing benzene to give the 1,5-disubstitued triazo-
les 11 and 13 in good to excellent yields (Scheme 4). With
these substrates, we did not notice the formation of the
1,4-regioisomers (a parallel successful positive control
of the 1,4-regioisomer formation was carried out using
[
1
3+2] cycloaddition between azides and alkynes (Scheme
1
3–15
).
We therefore decided to exploit the ruthenium-
1
3b
catalyzed [3+2] cycloaddition to substitute the lactone
moiety in deoxypodophyllotoxin and stegane with a
copper (II) sulphate/sodium ascorbate).
1
,5-disubstituted 1,2,3-triazole. In theory, this heterocy-
In our stategy, we sought to generate azasteganacins and
azadeoxypodophyllotoxins by using different oxidative
clic ring would be metabolically stable and able to par-
ticipate as a hydrogen-bond acceptor, further allowing
the removal of two stereocenters.
1
9
coupling protocols from the same reactants (11 and
13). To our delight, using thallium (III) oxide/boron tri-
fluoride in trifluoroacetic acid, we were able to obtain
the racemic azasteganacin derivatives 12 and 14 in excel-
lent yields (Scheme 4).
In this manuscript, we report the synthesis and the bio-
logical evaluation of these new aza analogues of podo-
phyllotoxin and steganacin which can be readily
prepared in a few synthetic steps using the ruthenium
Unfortunately, using a number of protocols (e.g., Tl O
2
3
‘
click’ cycloaddition.
in CF COOH or Mn(OAc) in CF COOH) we were not
3 3 3
able to obtain the azadeoxypodophyllotoxin analogues
and for this reason a different strategy was used to get
these compounds. Indeed, a hydroxyl group at the ben-
zylic position might undergo an intramolecular Friedel–
Craft alkylation under acidic condition. For this reason,
intermediate 17 was prepared in the following manner:
3,4,5-trimethoxybenzaldehyde was reacted with ethynyl-
trimethylsilane in the presence of butyl lithium at
À78 ꢁC to give the corresponding secondary alcohol
15. This compound was deprotected using TBAF in
THF and the resulting terminal alkyne 16 was reacted
2
. Chemistry
The retrosynthetic strategy in Scheme 2 was planned to
synthesize the desired compounds. Commercially avail-
able piperonyl alcohol was brominated using phospho-
rous tribromide in ether to give piperonyl bromide (3)
which was subsequently turned into an azide (4) using
sodium azide in DMF/water. Trimethoxybenzyl azide
(
8) was synthesized using similar conditions starting
from 3,4,5-trimethoxybenzyl alcohol via bromide deriv-
ative (7). Preparation of the desired alkynes was carried
out starting from the corresponding bromide derivatives
of the piperonyl (3) and trimethoxybenzyl alcohol (7).
*
3 2
with 4 in the presence of Cp Ru(PPh ) Cl as catalyst.
Also in this case clean formation of the 1,5-disubstituted
triazole 17 was observed (Scheme 5).
Finally, acidic intramolecular Friedel–Craft reaction
using trifluoroacetic acid gave the desired azadeoxypodo-
phyllotoxin analogue 18 (Scheme 6). All attempts to
The acetylenic group was inserted using ethynyltrimeth-
ylsilane in the presence of ethylmagnesium bromide

6750
D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
Scheme 2. Retrosynthetic analysis for the synthesis of the target compounds.
generate the other azapodophyllotoxin analogue failed
approx. 7.0 ± 0.9 nM, a concentration comparable to
4
(see Supplementary materials for the synthetic
attempts).
those reported previously by others. The synthetic
intermediates where the triazole had not been closed
(
11, 13, 17), were devoid of any activity (data not shown
and Fig. 2), since no cytotoxicity was observable up to
100 lM. The steganacin analogue 12 displayed some
cytotoxic potential at concentrations over 10 lM, but
this was not further studied. On the contrary, 14 dis-
played cytotoxic activity which was comparable to 1 as
far as maximal effect, but with a significantly lower po-
tency (IC50 of approx. 1.1 ± 0.4 lM). Strikingly, the
cytotoxic potential of this compound is similar, if
not higher compared to that reported by others for
3
. Biological results and conclusion
To evaluate the biological potential of the synthesized
compounds, we first screened their cytotoxicity in com-
parison to podophyllotoxin (2). Compounds were
incubated for 48 h with a neuroblastoma cell line (SH-
SY5Y) and viability was evaluated via the MTT meth-
od. Under these conditions, 2 displayed an IC50 of

D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
6751
Scheme 3. Reagents and conditions: (a) PBr
0%; (d) CH
COOH, TBAF, 0 ꢁC, 90%; (e) PBr
reflux, 69%; (h) AgNO
3
, Et
2
O, 0 ꢁC, 70%; (b) NaN
3
, DMF, 40 ꢁC, 65%; (c) ethynyltrimethylsilane, EtMgBr, CuBr, THF, reflux,
, DMF, 40 ꢁC, 72%; (g) ethynyltrimethylsilane, EtMgBr, CuBr, THF,
9
3
3
, CH Cl
2
2
, 0 ꢁC, 81%; (f) NaN
3
3
, THF/H
2
O/2,3-lutidine, 0 ꢁC, 71%.
5
steganacin, its reference analogue. Last, the podophyl-
lotoxin analogue (18) was the most efficacious com-
pound tested, and displayed an IC50 of approx.
In conclusion, we report the synthesis of steganacin and
podophyllotoxin analogues that present a triazole moi-
ety in place of the lactone ring. The synthetic strategy
used is composed of few steps and makes use of a
click-chemistry reaction that capitalizes on the capacity
of ruthenium to catalyze [3+2] Huisgen’s cycloaddition
between azides and alkynes to yield the 1,5-disubstituted
triazole selectively. Indeed, the number of steps used for
the synthesis is considerably fewer than those utilized to
synthesize similar compounds, once again strengthen-
ing the claim that click chemistry is a valid aid to the
synthetic medicinal chemist. The synthesized com-
pounds appear to be active as cytotoxic agents and re-
tain antitubulin activity. Indeed, in our opinion they
could form the basis for novel analogues, as the trans
lactone presents low thermodynamic stability (com-
pared to the biologically inactive cis-form) and might
undergo metabolic cleavage. The aza analogues also
present a reduced number of stereogenic centers allow-
ing the possibility to generate a small library of com-
pounds in a relatively short period of time. Although
podophyllotoxin itself has not reached the market as
antitumoral agent, due to its low therapeutic index, it
was the molecular platform for the discovery of two
1
.5 ± 0.7 lM.
To evaluate whether the active compounds synthesized
still retained the antitubulin activity of their parent
compounds, we decided to investigate the effect of
18) in an in vivo tubulin assay, comparing its effects
to podophyllotoxin, combretastatin A-4, and paclitaxel
Fig. 3). In brief, cells were incubated with tubulin dis-
rupting agents for 16 h and then proteins were har-
vested in the presence of paclitaxel to freeze the
polymerized and unpolymerized forms of tubulin. Pro-
teins were then run on SDS–PAGE and an antitubulin
antibody was used to detect differences in polymeriza-
tion. In this assay, combretastatin A-4 and 2 both dis-
played, as expected, a depolymerizing effect, since
soluble protein was more than that found in the pellet.
As expected, paclitaxel displayed the opposite behav-
ior. As shown in Figure 3, compound 18 still retained
the capacity to disrupt tubulin polymerization, as it
displayed a pattern similar to 2 and combretastatin
A-4.
(
2
0
(

6752
D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
Scheme 4. Reagents and conditions: (i) Cp Ru(PPh
3
)
benzene, reflux, 94%; (l) Tl O , BF etherate, CF COOH, rt, 90%.
2
Cl, benzene, reflux, 85%; (j) Tl
*
2 3 3 3 3 2
O , BF etherate, CF COOH, rt, 62%; (k) Cp Ru(PPh ) Cl,
*
2 3 3 3
Scheme 5. Reagents and conditions: (m) ethynyltrimethylsilane, BuLi, THF, À78 ꢁC, 92%; (n) TBAF, THF, 0 ꢁC, 96%; (o) Cp*Ru(PPh
3 2
) Cl,
benzene, reflux, 56%.
3
Scheme 6. Reagents and conditions: (p) CF COOH, molecular sieves
˚
A, reflux, 22%.
4
blockbuster topoisomerase II inhibitors, such as etopo-
2
1
side and teniposide, and therefore such libraries could
be of use for multiple targets.
4. Experimental
Commercially available reagents and solvents were used
without further purification and were purchased from
Fluka–Aldrich or Lancaster. Dichloromethane was
Figure 2. Dose–response curve of cytotoxicity for the active com-
pounds synthesized. Compounds 11 and 13 were devoid of any activity
at the highest concentration tested (100 lM).

D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
6753
4.2. 1,3-Benzodioxol-5-ylmethyl azide (4)
To a solution of 3 (1.1 g; 5.1 mmol) in DMF (14 mL)
and water (3 mL), 500 mg of sodium azide (7.7 mmol;
1.5 equiv) were added. The resulting mixture was heated
at 40 ꢁC for 2 h. The reaction was worked up by dilution
with diethyl ether and washed with water (2·). The
aqueous layer was further washed with EtOAc (2·),
and the combined organic extracts were washed with
brine (1·). After drying over sodium sulphate, filtration,
and evaporation of the solvent, the crude was purified
by column chromatography using PE/EtOAc 95:5 as
eluant to give 600 mg of the corresponding azide 4 as
colorless oil (65%). IR (neat) 2094, 1489, 1444, 1243,
Figure 3. In vitro tubulin polymerization assay. S stands for soluble
non-polymerized) and P stands for pellet (polymerized). Tubulin was
visualized with a monoclonal mouse antibody.
(
dried by distillation from P O and stored on activated
5
2
˚
molecular sieves (4 A). Tetrahydrofuran (THF) and
diethyl ether were distilled immediately before use from
Na/benzophenone under a slight positive atmosphere of
À1
1
N . Dimethylformamide (DMF) was purified by distilla-
2
1037, 927 cm ; H NMR (300 MHz, CDCl ) d 6.79
3
13
(m, 3-H), 5.98 (s, 2-H), 4.23 (s, 2-H); C NMR
tion at reduced pressure collecting the fraction having
bp 76 ꢁC at 39 mmHg and stored on activated molecular
sieves (4 A). When needed, the reactions were performed
(75 MHz, CDCl ) d 148.2, 147.8, 129.3, 122.0, 108.8,
3
˚
108.4, 101.4, 54.7.
in flame- or oven-dried glassware under a positive pres-
sure of dry N2.
4.3. [3-(1,3-Benzodioxol-5-yl)-1-propynyl](trimethyl)silane
(5)
Melting points were determined in open glass capillary
with a Stuart scientific SMP3 apparatus and are uncor-
rected. All the compounds were checked by IR (FT-IR
In a three-necked round bottom flask in dry conditions
ethynyltrimethylsilyl magnesium bromide was prepared
in situ in the following manner: to a cooled (0 ꢁC) solu-
tion of ethynyltrimethylsilane (10.2 mL; 73.4 mmol; 4
equiv) in dry THF (60 mL), 73.4 mL of ethyl magnesium
bromide (1 M solution in THF; 73.4 mmol; 4 equiv) was
added dropwise. After the addition, the cooling bath
was removed and the resulting solution was stirred at
room temperature for 30 min. Then, 395 mg of copper
(I) bromide (2,75 mmol; 0.2 equiv) was added and after
30 min 3.9 g of 3 (18.35 mmol; 1 equiv) was added. The
resulting mixture was heated at reflux for 16 h. The reac-
1
13
THERMO-NICOLET AVATAR); H and C APT
JEOL ECP 300 MHz) and mass spectrometry (Ther-
(
mo Finningan LCQ-deca XP-plus) equipped with an
ESI source and an ion trap detector. Chemical shifts
are reported in parts per million (ppm). Column chro-
matography was performed on silica gel (Merck Kiesel-
gel 70–230 mesh ASTM) using the indicated eluants.
Thin layer chromatography (TLC) was carried out on
5
· 20 cm plates with a layer thickness of 0.25 mm
(Merck Silica gel 60F ). When necessary they were
developed with KMnO , Dragendorff reagent, and
2
54
tion was then cooled at 0 ꢁC and satd aq NH Cl was
4
.
4
Phosphomolibdic reagent. Elemental Analysis (C, H,
N) of the target compounds 11, 12, 13, 14, 17, and
added dropwise. The resulting solution was then diluted
with EtOAc and the organic layer was washed with
water (1·) and brine (1·). After drying over sodium sul-
phate, filtration, and evaporation of the solvent, the
crude was purified by column chromatography using
18 are within ±0.4% of the calculated values unless
otherwise noted.
Ruthenium catalyst Cp*Ru(PPh ) Cl was synthesized as
described previously.
PE/EtOAc 9:1 as eluant to give 3.85 g of 5 as pale yellow
oil (90%). IR (neat) 1502, 1488, 1245, 1039, 838 cm
3
2
2
2
À1
;
1
H NMR (300 MHz, CDCl ) d 6.76 (m, 3-H), 5.94 (s,
3
1
2-H), 3.56 (s, 2-H), 0.17 (s, 9-H); C NMR (75 MHz,
3
Compounds were evaluated via the MTT method and
2
3
via the in vitro tubulin assay as described previously.
.1. 5-(Bromomethyl)-1,3-benzodioxole (3)
CDCl ) d 147.8, 146.3, 130.2, 120.8, 108.6, 108.2,
3
104.5, 101.0, 86.9, 25.9, 0.18.
4
4.4. 5-(2-Propynyl)-1,3-benzodioxole (6)
To a cooled (0 ꢁC) solution of piperonyl alcohol (5 g;
2.86 mmol) in dry diethyl ether (50 mL), 3.1 mL of
3
To a solution of 5 (3.8 g; 16.37 mmol) in THF (38 mL),
1.1 mL of acetic acid was added (19.55 mmol; 1.2 equiv).
The resulting mixture was cooled at 0 ꢁC and 19.55 ml of
TBAF sol. 1 M in THF was added dropwise
(19.55 mmol; 1.2 equiv). After 4 h, the reaction was
worked up by evaporation of the solvent, and the resi-
due was dissolved in EtOAc and washed with water
(1·) and brine (1·). After drying over sodium sulphate,
filtration, and evaporation of the solvent, the crude was
purified by column chromatography using PE/EtOAc
9:1 as eluant to give 2.3 g of 6 as pale yellow oil
(90%). IR (neat) 3292, 1501, 1487,1243, 1037,
phosphorous tribromide (32.86 mmol; 1 equiv) dis-
solved in 40 mL of diethyl ether was added dropwise.
After 10 min, the reaction was worked up by dilution
with water. The organic layer was then washed with
brine (1·). After drying over sodium sulphate, filtra-
tion, and evaporation of the solvent, the crude was
recrystallized with petroleum ether to give 5 g of
the bromine 3 as white solid (70%). mp 97.0–
9
À1
8.0 ꢁC; IR (KBr) 2904, 1500, 1420, 1035, 776 cm
;
1
H NMR (300 MHz, CDCl ) d 6.86 (m, 2-H), 6.75
3
(
d, J = 8.2 Hz, 1-H), 5.97 (s, 2-H), 4.45 (s, 2-H);
C NMR (75 MHz, CDCl ) d 148.0, 147.9, 131.6,
1
3
À1
1
803 cm
J = 1.1 Hz, 1-H), 6.78 (m, 2-H), 5.93 (s, 2-H), 3.51 (d,
;
H NMR (300 MHz, CDCl ) d 6.85 (d,
3
3
1
22.9, 109.6, 108.4, 101.4, 34.4.

6754
D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
1
3
J = 2.7 Hz, 2-H), 2.18 (t, J = 2.7 Hz, 1-H); C NMR
75 MHz, CDCl ) d 147.8, 146.4, 129.8, 121.0, 108.4,
(69%). IR (neat) 1590, 1505, 1420, 1236, 1125,
À1
1
(
840 cm ; H NMR (300 MHz, CDCl ) d 6.56 (s, 2-
3
3
1
08.1, 101.0, 82.0, 70.5, 24.3.
H), 3.83 (s, 6-H), 3.80 (s, 3-H), 3.57 (s, 2-H), 0.17 (s,
1
3
9
132.0, 104.9, 104.3, 87.4, 60.9, 56.0, 26.4, 0.14. MS
-H); C NMR (75 MHz, CDCl ) d 153.2, 136.7,
3
4
.5. 5-(Bromomethyl)-1,2,3-trimethoxybenzene (7)
+
ESI) m/z 301 (M+Na) .
(
To a cooled (0 ꢁC) solution of trimethoxybenzyl alchool
3 g; 15.13 mmol) in dry dichloromethane (30 mL),
.4 mL of phosphorous tribromide, (15.13 mmol; 1
(
1
4.8. 1,2,3-Trimethoxy-5-(2-propynyl)benzene (10)
equiv) dissolved in 40 mL of diethyl ether, was added
dropwise. After 10 min, the reaction was worked up
by dilution with water. The organic layer was then
washed with brine (1·). After drying over sodium sul-
phate, filtration, and evaporation of the solvent, the
crude was recrystallized with petroleum ether to give
To a cooled (0 ꢁC) solution of 7 (700 mg; 2.69 mmol) in
THF (12 mL), water (12 mL), ethanol (12 mL), and 2,3-
lutidine (1.2 mL), 4.6 g of silver nitrate was added
(26.9 mmol; 10 equiv). After 30 min the reaction was
worked up by acidification with H SO 2 M. The result-
2
4
ing precipitate was filtered off and the solution was
diluted with EtOAc and washed with brine (1·). After
drying over sodium sulphate, filtration, and evaporation
of the solvent, the crude was purified by column chro-
matography using PE/EtOAc 9:1 as eluant to give
1.9 g of 10 as pale yellow oil (71%). IR (neat) 3283,
3
8
.2 g of the bromine 7 as white solid (81%). mp 86–
7 ꢁC; IR (KBr) 2942, 1589, 1465, 1245, 993 cm ; H
À1
1
NMR (300 MHz, CDCl ) d 6,59 (s, 2-H), 4.43 (s, 2-
3
1
H), 3.83 (s, 6-H), 3.81 (s,3-H); C NMR (75 MHz,
3
CDCl ) d 153.1, 138.2, 133.3, 106.2, 60.9, 56.2, 34.4;
3
+
MS (ESI) m/z 261–263 (M+H) .
À1
1
1590, 1504, 1420, 1234, 1121, 1004, 816 cm
;
H
NMR (300 MHz, CDCl ) d 6.57 (s, 2-H), 3.86
3
4
.6. 5-(Azidomethyl)-1,2,3-trimethoxybenzene (8)
(s, 6-H), 3.82 (s, 3-H), 3.56 (d, J = 2.7 Hz, 2-H), 2.20
1
3
(
t, J = 2.7 Hz, 1-H); C NMR (75 MHz, CDCl ) d
3
To a solution of 7 (3 g; 11.49 mmol) in DMF (36 mL)
and water (8 mL), 1.1 g of sodium azide (17.23 mmol;
153.4, 136.8, 131.8, 104.9, 81.9, 70.7, 60.9, 56.2, 25.1;
MS (ESI) m/z 207 (M+H) .
+
1
.5 equiv) was added. The resulting mixture was heated
at 40 ꢁC for 2 h. The reaction was worked up by dilution
with diethyl ether and washed with water (2·). The
aqueous layer was further washed with EtOAc (2·),
and the combined organic extracts were washed with
brine (1·). After drying over sodium sulphate, filtration,
and evaporation of the solvent, the crude was purified
by column chromatography using PE/EtOAc 9:1 and
PE/EtOAc 8:2 as eluants to give 1.8 g of the correspond-
ing azide 8 as colorless oil (72%). IR (neat) 2943, 2100,
4.9. 1-(1,3-Benzodioxol-5-ylmethyl)-5-(3,4,5-trimethoxy-
benzyl)-1H-1,2,3-triazole (11)
To a solution of the azide 4 (650 mg; 3.67 mmol; 1 equiv)
in benzene (40 mL), 1.3 g of the alkyne 10 (6.24 mmol;
1.7 equiv) and 52 mg of Cp*RuCl(PPh ) (0.073 mmol;
3
2
0.02 equiv) were added. The resulting mixture was heated
at reflux for 24 h and worked up by evaporation of the
solvent. The resulting crude was purified by column
chromatography using PE/EtOAc 7:3 and PE/EtOAc
4:6 as eluants to give 1.2 g of 11 as pale yellow solid
(85%). mp 101–102 ꢁC; IR (KBr) 2940, 1591, 1335,
À1
1
1
6
593, 1460, 1128 cm ; H NMR (300 MHz, CDCl ) d
3
.48 (s, 2-H), 4.22 (s, 2-H), 3.81 (s, 6-H), 3.78 (s, 3-H);
C NMR (75 MHz, CDCl ) d 153.5, 137.7, 131.1,
1
3
3
+
05.1, 60.1, 56.1, 55.1; MS (ESI) m/z 224 (M+H) .
À1
1
1
1127, 923 cm ; H NMR (300 MHz, CDCl ) d 7.55 (s,
3
1-H), 6.73 (d, J = 8.0 Hz, 1-H), 5.57 (br d, J = 8.0 Hz,
2-H), 6.18 (s, 2-H), 5.95 (s, 2-H), 5.34 (s, 2-H), 3.84 (s,
4.7. Trimethyl[3-(3,4,5-trimethoxyphenyl)-1-propynyl]
silane (9)
1
2-H), 3.82 (s, 3-H), 3.76 (s, 6-H); C NMR (75 MHz,
3
CDCl ) d 153.5, 148.3, 147.7, 137.1, 135.6, 134.4, 131.5,
3
In a three-necked round bottom flask in dry condition,
ethynyltrimethylsilyl magnesium bromide was prepared
in situ in the following manner: to a cooled (0 ꢁC) solu-
tion of ethynyltrimethylsilane (1.06 mL; 7.66 mmol; 4
equiv) in dry THF (4 mL), 7.66 mL of ethyl magnesium
bromide (1 M solution in THF; 7.66 mmol; 4 equiv) was
added dropwise. After the addition the cooling bath was
removed and the resulting solution was stirred at room
temperature for 30 min. Then, 41 mg of copper (I) bro-
mide (0.29 mmol; 0.2 equiv) was added and after 30 min
128.4, 120.9, 108.4, 107.9, 105.4, 101.4, 61.0, 56.2, 51.9,
29.6; MS (ESI) m/z 406 (M+Na) .
+
Anal. Calcd for C H N O : C, 62.65; H, 5.52; N,
3
2
0
21
5
10.95. Found: C, 62.64; H, 5.53; N, 10.96.
4.10. 6,7,8-Trimethoxy-4,14-dihydrobenzo[d][1,3]benzodi-
oxolo[5,6-f][1,2,3]triazolo[1,5-a]azocine (12)
To a solution of 11 (200 mg; 0.52 mmol; 1 equiv) in tri-
fluoroacetic acid (2 mL), 0.2 mL of boron trifluoride
diethyl ether (1.6 mmol; 3 equiv) was added. Then a sus-
pension containing 123 mg of thallium oxide
(0.27 mmol; 0.52 equiv) in trifluoroacetic acid (0.5 mL)
was added. After 15 min, the reaction was worked up
by dilution with EtOAc and the organic layer was
washed with NaOH 1 M (2·) and brine (1·). After dry-
ing over sodium sulphate, filtration, and evaporation of
the solvent, the crude was purified by column chroma-
500 mg of 7 (1.9 mmol; 1 equiv) was added. The result-
ing mixture was heated at reflux for 16 h. The reaction
was then cooled at 0 ꢁC and satd aq NH Cl was added
4
dropwise. The resulting solution was then diluted with
EtOAc and the organic layer was washed with water
(
1·) and brine (1·). After drying over sodium sulphate,
filtration, and evaporation of the solvent, the crude was
purified by column chromatography using PE/EtOAc
9
:1 as eluant to give 1.9 g of 9 as pale yellow oil

D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
6755
tography using PE/EtOAc 5:5 to give 124 mg of 12 as
white powder (62%). mp 203–204 ꢁC; IR (KBr) 2941,
Anal. Calcd for C H N O : C, 62.98; H, 5.02; N,
20 19 3 5
11.01. Found: C, 62.94; H, 5.23; N, 10.97.
À1
1
1
596, 1486, 1238, 1038 cm
;
H NMR (300 MHz,
CDCl ) d 7.47 (s, 1-H), 6.94 (s, 1-H), 6.83 (s, 1-H),
4.13. 1-(3,4,5-Trimethoxyphenyl)-3-(trimethylsilyl)-2-
propyn-1-ol (15)
3
6
.58 (s, 1-H), 6.03 (s, 1-H), 5.98 (s, 1-H), 5.53
(
(
(
(
1
1
d, J = 14.1 Hz; 1-H), 4.59 (d, J = 14.1 Hz, 1-H), 3.88
s, 3-H), 3.84 (s, 3-H), 3.83 (d, J = 15.3 Hz, 1-H), 3.70
To a cooled (À78 ꢁC) solution of ethynyltrimethylsilane
(3.9 mL, 28.03 mmol, 1.1 equiv) in dry THF, 12.2 mL of
butyl lithium 2.5 M solution in hexane (60.58 mmol; 1.2
equiv) was added dropwise. The resulting mixture was
stirred for 30 min, then 5 g of 3,4,5-trimethoxybenzalde-
hyde (25.5 mmol; 1 equiv) dissolved in 50 mL of dry
THF was added dropwise. After 1 h, the reaction was
1
3
s, 3-H), 3.30 (d, J = 15.3 Hz, 1-H);
C NMR
75 MHz, CDCl ) d 154.0, 150.9, 148.2, 148.1, 141.7,
3
34.2, 132.4, 129.8, 126.6, 124.4, 110.1, 110.0, 108.5,
01.7, 61.1, 61.0, 56.1, 52.2, 28.9; MS (ESI) m/z 382
+
(
M+H) .
Anal. Calcd for C H N O : C, 62.98; H, 5.02; N,
19
worked up by slow addition of satd aq NH Cl. When
4
20
3
5
1
1.01. Found: C, 62.73; H, 5.43; N, 11.30.
the resulting mixture reached room temperature, it was
diluted with EtOAc and washed with water (2·). The
aqueous layer was further washed with EtOAc (2·),
and the combined organic extracts were washed with
brine (1·). After drying over sodium sulphate, filtration,
and evaporation of the solvent, the crude as yellowish
oil (6.9 g, 92%) was enough pure to be used directly
for the following step. IR (neat) 3488, 1593, 1461,
4.11. 5-(1,3-Benzodioxol-5-ylmethyl)-1-(3,4,5-trimeth-
oxybenzyl)-1H-1,2,3-triazole (13)
To a solution of the azide 8 (1.5 g; 6.62 mmol; 1 equiv)
in benzene (62 mL), 1.8 g of the alkyne 6 (11.25 mmol;
1
0
.7 equiv) and 96 mg of Cp RuCl(PPh ) (0.13 mmol;
*
.02 equiv) was added. The resulting mixture was heated
3 2
À1
1
1124, 839 cm ; H NMR (300 MHz, CDCl ) d 6.81
3
at reflux for 24 h and worked up by evaporation of the
solvent. The resulting crude was purified by column
chromatography using PE/EtOAc 7:3 and PE/EtOAc
(s, 2-H), 5.39 (s, 1-H), 3.87 (s, 6-H), 3.84 (s, 3-H), 0.22
1
3
(s, 9-H); C NMR (75 MHz, CDCl ) d 152.9, 137.3,
3
136.5, 105.8, 103.7, 90.4, 64.4, 60.5, 55.8, À0.28; MS
+
(ESI) m/z 317 (M+Na) .
4
:6 as eluants to give 2.4 g of 13 as pale yellow solid
(
1
(
94%). mp 98–99 ꢁC; IR (KBr) 2941, 1592, 1461, 1242,
À1
1
037 cm
s, 1-H), 6.69 (d, J = 8.0 Hz, 1-H), 6.46 (m, 2-H), 6.26
;
H NMR (300 MHz, CDCl ) d 7.53
4.14. 1-(3,4,5-Trimethoxyphenyl)-2-propyn-1-ol (16)
3
(
s, 2-H), 5.93 (s, 2-H), 5.32 (s, 2-H), 3.84 (br s, 5-H),
To a cooled (0 ꢁC) solution of 15 (7.4 g, 25 mmol) in
THF (74 mL), 38 mL of TBAF sol. 1 M in THF
(38 mmol; 1.5 equiv) was added dropwise. After the
addition, the cooling bath was removed and the result-
ing solution was stirred at room temperature for 16 h.
The reaction was worked up by addition at 0 ꢁC of a
1
3
3
1
1
.81 (s, 3-H); C NMR (75 MHz, CDCl ) d 153.6,
3
48.6, 146.7, 137.8, 136.0, 134.1, 130.3, 129.5, 121.4,
08.7, 108.4, 104.3, 101.2, 60.8, 56.1, 52.0, 28.9; MS
+
(
ESI) m/z 384 (M+H) .
Anal. Calcd for C H N O : C, 62.65; H, 5.52; N,
5
satd aq NH Cl then it was diluted with EtOAc and
20
21
3
4
1
0.95. Found: C, 63.05; H, 5.65; N, 11.10.
washed with water (1·). The aqueous layer was further
washed with EtOAc (2·), and the combined organic ex-
tracts were washed with brine (1·). After drying over so-
dium sulphate, filtration, and evaporation of the solvent,
the crude was purified by column chromatography to
give 5.4 g of 16 as pale yellow solid (96%). IR (KBr)
4.12. 7,8,9-Trimethoxy-5,15-dihydro[1,3]benzodioxolo [5,6-
d]benzo[f][1,2,3]triazolo[1,5-a]azocine (14)
To a solution of 13 (200 mg; 0.52 mmol; 1 equiv) in tri-
fluoroacetic acid (2 mL), 0.2 mL of boron trifluoride
diethyl ether (1.6 mmol; 3 equiv) was added. Then a sus-
pension containing 123 mg of thallium oxide
À1
1
3242, 2942, 1593, 1460, 1235, 1124, 1006 cm
;
H
NMR (300 MHz, CDCl ) d 6.73 (s, 2-H), 5.34 (d,
3
J = 2.2 Hz, 1-H), 3.81 (s, 3-H), 3.25 (br s, –OH), 3.77
C NMR
1
3
(
0.27 mmol; 0.52 equiv) in trifluoroacetic acid (0.5 mL)
(s, 6-H), 2.62 (d, J = 2.2 Hz, 1-H);
(75 MHz, CDCl ) d 153.3, 137.8, 136.0, 83.7, 74.6,
was added. After 15 min the reaction was worked up
by dilution with EtOAc and the organic layer was
washed with NaOH 1 M (2·) and brine (1·). After dry-
ing over sodium sulphate, filtration and evaporation of
the solvent, the crude was purified by column chroma-
tography using PE/EtOAc 5:5 to give 183 mg of 14 as
white powder (90%). mp 199–200 ꢁC; IR (KBr) 2942,
3
+
64.3, 60.8, 60.5, 56.1; MS (ESI) m/z 245 (M+Na) .
4.15. [1-(1,3-Benzodioxol-5-ylmethyl)-1H-1,2,3-triazol-5-
yl](3,4,5-trimethoxyphenyl)methanol (17)
To a solution of the azide 4 (666 mg; 3.76 mmol; 1
equiv) in benzene (44 mL), 1.4 g of the alkyne 16
(6.39 mmol; 1.7 equiv) and 55 mg of Cp*RuCl(PPh3)2
(0.075 mmol; 0.02 equiv) were added. The resulting mix-
ture was heated at reflux for 24 h and worked up by
evaporation of the solvent. The resulting crude was puri-
fied by column chromatography using PE/EtOAc 7:3
and PE/EtOAc 4:6 as eluants to give 836 mg of 17as pale
yellow solid (56%). mp 124–125 ꢁC; IR (KBr) 3200,
À1
1
1
598, 1483, 1352, 1141, 995 cm
;
H
NMR
(
300 MHz, CDCl ) d 7.55 (s, 1-H), 6.81 (br s, 1-H),
3
6
4
3
.04 (s, 1-H), 5.99 (s, 1-H), 5.55 (d, J = 14.1 Hz, 1-H),
.59 (d, J = 14.1 Hz, 1-H), 3.92 (s, 3-H), 3.87 (s, 3-H),
.79 (d, J = 15.0 Hz, 1-H), 3.71 (s, 3-H), 3.36 (d,
1
3
J = 15.0 Hz, 1-H);
C NMR (75 MHz, CDCl ) d
3
1
1
2
53.9, 150.9, 148.1, 141.7, 134.0, 132.4, 129.9, 126.7,
24.4, 110.2, 110.0, 108.5, 101.7, 61.1, 61.0, 56.1, 52.1,
+
9.0; MS (ESI) m/z 382 (M+H) .
À1
1
1592, 1501, 1245, 797 cm
;
H NMR (300 MHz,

6756
D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
CDCl ) d 7.48 (s, 1-H), 6.72 (d, J = 8.0 Hz, 1-H), 6.63 (d,
J = 8.0 Hz, 1-H), 6.44 (br s, 3-H), 5.94 (s, 2-H), 5.69 (s,
Design 2000, 6, 1811; (c) Desb e` ne, S.; Giorgi-Renault, S.
Curr. Med. Chem. Anti-Cancer Agents 2002, 2, 71; (d)
Gordaliza, M.; Garc `ı a, P. A.; Miguel del Corral, J. M.;
Castro, M. A.; G o´ mez-Zurita, M. A. Toxicon 2004, 44,
3
1
H), 3.82 (s, 3-H), 3.79 (s, 6-H), 1.81 (br s, –OH) ;
-H), 5.51 (d, J = 14.7 Hz, 1-H), 5.41 (d, J = 14.7 Hz, 1-
C
1
3
4
41.
NMR (75 MHz, CDCl ) d 153.5, 148.0, 147.6, 138.8,
3
5
. For steganacin see: (a) Tomioka, K.; Ishiguro, T.;
Mizuguchi, H.; Komeshima, N.; Koga, K.; Tsuka-
goshi, S.; Tsuruo, T.; Tashiro, T.; Tanida, S.; Kishi,
T. J. Med. Chem. 1991, 34, 54; (b) Tomioka, K.;
Kawasaki, H.; Koga, K. Chem. Pharm. Bull. 1990,
38, 1898; (c) Tomioka, K.; Kubota, Y.; Kawasaki,
H.; Koga, K. Tetrahedron Lett. 1989, 30, 2949; (d)
Kubota, Y.; Kawasaki, H.; Tomioka, K.; Koga, K.
Tetrahedron 1993, 49, 3081; (e) Zavala, F.; Gue-
nard, D.; Robin, J. P.; Brown, E. J. Med. Chem.
1
1
37.6, 136.1, 135.5, 128.6, 121.4, 108.3, 108.2, 103.6,
01.3, 66.6, 60.8, 56.0, 52.1; MS (ESI) m/z 398 (MÀH) .
À
Anal. Calcd for C H N O : C, 60.14; H, 5.29; N,
3
2
0
21
6
1
0.52. Found: C, 60.36; H, 5.27; N, 10.74.
4
[
.16. 4-(3,4,5-Trimethoxyphenyl)-4,10-dihydro[1,3]dioxolo
4,5-g][1,2,3]triazolo[1,5-b]isoquinoline (18)
1
980, 23, 546.
Three hundred milligrams of 17 (0.75 mmol) was added
to a suspension of trifluoroacetic acid (3 mL) and molec-
6
. (a) Castro, A.; Miguel del Corral, J. M.; Gordaliza, M.;
Grande, C.; G o´ mez-Zurita, A.; Garc ´ı a- Gr a´ valos, D.; San
Feliciano, A. Eur. J. Med. Chem. 2003, 38, 65; The
benzodioxole moiety could also contribute to cytotoxicity
since it has been reported that it can be metabolized by
CYP to form metallo-carbene intermediates which could
be responsible for the antitumor activity of lignans, see:
˚
ular sieves (4 A). The resuling mixture was heated at re-
flux for 4 h. The reaction was then worked up by
dilution with EtOAc and washed with NaOH 1 M (2·)
and brine (1·). After drying over sodium sulphate, filtra-
tion, and evaporation of the solvent, the crude was puri-
fied by column chromatography using PE/EtOAc 7:3
and PE/EtOAc 4:6 as eluants to give 62 mg of 18 as pale
yellow solid (22%). mp 120–121 ꢁC; IR (KBr) 2931,
(
1
b) Landais, Y.; Robin, J. P.; Lebrun, A. Tetrahedron
991, 47, 3787, and references cited therein.
7
. (a) Gordaliza, M.; Angels-Castro, M.; Miguel del Corral,
J. M.; L o´ pez-V a´ zquez, M. L.; Garc ´ı a, P. A.; San Felic-
iano, A.; Garc ´ı a-Gr a´ valos, M. D.; Broughton, H. Tetra-
hedron 1997, 53, 15743; (b) Gordaliza, M .; Miguel del
Corral, J. M.; Angels-Castro, M.; L o´ pez-V a´ zquez, M. L.;
Garc ´ı a, P. A.; San Feliciano, A.; Garc ´ı a-Gr a´ valos, M. D.
Bioorg. Med. Chem. Lett. 1995, 5, 2465.
À1
1
1
590, 1507, 1249, 1127, 934 cm
;
H
NMR
(
(
300 MHz, CDCl ) d 7.43 (s, 1-H), 6.80 (s, 1-H), 6.56
s, 1-H), 6.30 (s, 2-H), 5.99 (s, 2-H), 5.54 (dd, J = 16.3/
3
1
(
.62 Hz; 1-H), 5.43 (dd, J = 16.3, 1.62 Hz, 1-H), 5.16
br s, 1-H), 3.84 (s, 3-H), 3.77 (s, 6-H); C NMR
1
3
(
1
6
75 MHz, CDCl ) d 153.8, 148.0, 147.6, 137.6, 136.8,
8. Gensler, W. J.; Murthy, C. D.; Trammell, M. H. J. Med.
Chem. 1977, 20, 635.
3
36.3, 131.2, 127.8, 122.3, 108.8, 106.1, 105.7, 101.8,
1.0, 56.3, 48.4, 42.9; MS (ESI) m/z 382 (M+H) .
+
9
. Gensler, W. J.; Gatsonis, C. D. J. Am. Chem. Soc. 1962,
4, 1748.
0. Brewer, C. F.; Loike, J. D.; Horwitz, S. B. J. Med. Chem.
979, 22, 215.
8
1
1
Anal. Calcd for C H N O : C, 62.98; H, 5.02; N,
2
1
0
19
3
5
1
1.01. Found: C, 63.10; H, 5.12; N, 10.85.
1. This could explain the reduced in vivo activity of
podophyllotoxin compared to its in vivo activity, see
Ref. 4b and references cited therein.
Acknowledgment
12. Cyclic ether, cyclic lactam, cyclic thioether, cyclopenta-
none, and carbocyclopentane ring have been prepared. See
Ref. 4a and references cited therein.
Financial support from Universit a` del Piemonte Orien-
tale and Regione Piemonte is gratefully acknowledged.
1
3. For copper catalyzed click chemistry, see: (a) Tornøe, C.
W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67,
3
057; (b) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 4,
596.
Supplementary data
2
1
4. For ruthenium-catalyzed click chemistry, see: Zhang, L.;
Chen, X.; Xue, P.; Sun, H. H. Y.; Williams, I. D.;
Sharpless, K. B.; Fokin, V. V.; Jia, G. J. Am. Chem. Soc.
2005, 127, 15998.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2007.08.020.
1
5. For medicinal chemistry application of click chemistry,
see: (a) Kolb, H. C.; Sharpless, K. B. Drug Discov. Today
2
003, 8, 1128; (b) Tron, G. C.; Pirali, T.; Billington, R. A.;
References and notes
Canonico, P. L.; Sorba, G.; Genazzani, A. A. Med. Res.
Rev. doi:10.1002/med.
1
2
. (a) Kupchan, S. M.; Britton, R. W.; Ziegler, M. F.;
Gilmore, C. J.; Restivo, R. J.; Bryan, R. F. J. Am. Chem.
Soc. 1973, 95, 1335; (b) Canel, C.; Moraes, R. M.; Dayan,
F. E.; Ferreira, D. Phytochemistry 2000, 54, 115.
. (a) Jordan, A.; Hadfield, J. A.; Lawrence, N. J.; McGown,
A. T. Med. Res. Rev. 1998, 18, 259; (b) Sackett, D. L.
Pharmacol. Ther. 1993, 59, 163.
16. (a) Rossi, R.; Carpita, A.; Lippolis, V.; Benetti, M. Gazz.
Chim. Ital. 1990, 120, 783; (b) Klunder, J. M.; Poshner, G.
H. In Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 3, pp
207–239.
17. Jung, M. E.; Hagenah, J. A. J. Org. Chem. 1987, 52,
1889.
3
4
. Bohlin, L.; Rosen, B. Drug Discovery Today 1996, 1, 343.
. For podophyllotoxin, see: (a) You, Y. Curr. Pharm.
Design 2005, 11, 1695; (b) Gordaliza, M.; Castro, M. A.;
Miguel del Corral, J. M.; San Feliciano, A. Curr. Pharm.
18. Carreira, E. M.; Du Bois, J. J. Am. Chem. Soc. 1994, 116,
10825.
19. (a) Ward, R. S.; Hughes, D. D. Tetrahedron 2001, 57,
2057; (b) Ref. 6b; (c) Ward, R. S.; Hughes, D. D.

D. Imperio et al. / Bioorg. Med. Chem. 15 (2007) 6748–6757
6757
Tetrahedron 2001, 57, 4015; (d) Planchenault, D.; Dhal,
R.; Robin, J. P. Tetrahedron 1993, 49, 5823; (e) Cambie,
R. C.; Craw, P. A.; Rutledge, P. S.; Woodgate, P. D. Aus.
J. Chem. 1988, 41, 897.
21. Baldwin, E. L.; Osheroff, N. Curr. Med. Chem. Anti-
Cancer Agents 2005, 5, 363.
22. Chinn, M. S.; Heinekey, D. M. J. Am. Chem. Soc. 1990,
112, 5166.
2
0. Beryozkina, T.; Appukkuttan, P.; Mont, N.; Van der
Eycken, E. Org. Lett. 2006, 8, 487.
23. Tron, G. C.; Pagliai, F.; Del Grosso, E.; Genazzani, A. A.;
Sorba, G. J. Med. Chem. 2005, 48, 3260.