Simplified dynemicin analogues: diastereoselective synthesis and evaluation of their activity against plasmid DNA.

Article abstract of DOI:10.1039/b307966j

The total synthesis of two diastereoisomeric simplified dynemicin analogues is reported. The key steps involved are: the regio- and diastereoselective functionalisation of an appropriate racemic quinoline precursor and the ring closure to give the 10-membered enediyne moiety through a Pd(0)-catalysed Stille reaction. After the successful conversion of one of these derivatives into a compound more readily activable under nearly physiological conditions, the activity against plasmid DNA was evaluated.

Full text of DOI:10.1039/b307966j

View Article Online / Journal Homepage / Table of Contents for this issue
Simplified dynemicin analogues: diastereoselective synthesis and
evaluation of their activity against plasmid DNA
Giuseppe Guanti* and Renata Riva*
Dipartimento di Chimica e Chimica Industriale, via Dodecaneso 31, 16146 Genova, Italy.
E-mail: guanti@chimica.unige.it (GG); riva@chimica.unige.it (RR);
Fax: ϩ390103536105 (GG); Fax: ϩ390103536118 (RR); Tel: ϩ390103536105 (GG);
Tel: ϩ390103536126 (RR)
Received 14th July 2003, Accepted 10th September 2003
First published as an Advance Article on the web 6th October 2003
The total synthesis of two diastereoisomeric simplified dynemicin analogues is reported. The key steps involved are:
the regio- and diastereoselective functionalisation of an appropriate racemic quinoline precursor and the ring closure
to give the 10-membered enediyne moiety through a Pd(0)-catalysed Stille reaction. After the successful conversion
of one of these derivatives into a compound more readily activable under nearly physiological conditions, the activity
against plasmid DNA was evaluated.
Introduction
Natural enediyne antibiotics (e.g. calicheamicin, esperamicin,
dynemicin 1) constitute a small family of cyclic derivatives,
all characterised by the presence of a 10-membered 3-ene-1,5-
diyne cyclic moiety. They are among the most powerful anti-
tumoral compounds known to date.1,2 All these compounds are
distinguished by a unique mode of action: they are actually
natural prodrugs, their biological properties being displayed
only after removal of an appropriate “safety lock”. This is
represented, in dynemicin and analogues, by an epoxide, whose
relative configuration is trans with respect to the 10-membered
unsaturated ring.1,3–7 The opening of the oxirane ring in
dynemicin, which is followed by Bergman cycloaromatisation
leading to a high reactive diradical responsible for DNA
damage, is a consequence of a series of cascade events, arising
from an initial bioreduction of the anthraquinone moiety.6
Since 1 is characterized by a rather complex structure, several
groups decided to synthesise simpler derivatives equipped
merely with the moieties essential to display the desired bio-
logical activity. However, when these simplified analogues lack
the anthraquinone portion, alternative triggering mechanisms
had to be exploited. The most widely used is based on the pro-
tection of the nitrogen (for example as carbamate). As soon as
this blocking group is removed, the restored nucleophilic power
Scheme 1 Retrosynthetic analysis.
of the nitrogen causes epoxide opening under physiological
conditions.
Within our project in the enediyne field, that has also wit-
1) the base-induced coupling of the bis(propargyl)bromide 312
or 2) the intramolecular pinacol coupling performed on the
bis(aldehyde) 4.9 On the other hand, following route B, only the
nessed various synthetic approaches to the family of artificial
acetylenic carbons are present on the acyclic derivative 5 and
enediynes called “lactenediynes”,8,9 we decided to synthesise
the remaining two sp2-carbon unit is joined directly during the
new simplified analogues of dynemicin 1, namely diastereo-
coupling reaction.3,13–15
The preparation of 3–5 requires the introduction of two
suitable acetylenic moieties, one at position 2 of the starting
isomeric derivatives 2a,b, and to explore their behaviour in
DNA cleavage experiments.10
quinoline nucleus, and the other in the side chain bonded at
C4. The selection of the most efficient order of introduction of
these two triple bonds was not trivial. Initial attempts to trans-
Results and discussion
Although derivatives with a structure resembling 2a,b have
already been prepared by Isobe and coworkers,5,11 we planned
a completely different strategy, depicted in Scheme 1. We
envisaged two different routes (A and B), having in common
the starting material [4-(1-hydroxyethyl)quinoline] and the
ring-closure of the enediyne ring at the double bond site. The
major difference between routes A and B is that, in the first
case, the required 6-carbon atom skeleton of the enediyne
moiety is already present on the acyclic precursor 3 or 4. The
ring closure may be realized through two different strategies:
form the formyl group of 2-(quinolin-4-yl)propanal into a
terminal triple bond, using several homologation procedures,
failed:16,17 only in one case16 did we isolate a modest amount of
a homologation product, but it was the allene derived from
isomerization of the triple bond.
From these preliminary experiments it was clear that the
triple bond directly attached at C2 of the quinoline nucleus
should be introduced first, thus transforming the side chain into
the second alkyne group at the level of a 1,2-dihydroquinoline
system. The regio- and diastereoselective addition of acetylides
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3967

View Article Online
Scheme 2 Reagents and conditions: a) (COCl)2, DMSO, iPr2EtN, Ϫ78 ЊC; b) CBr4, PPh3, Ϫ78
Ϫ50 ЊC; c) n-BuLi, toluene, Ϫ78 ЊC, argon; d)
r.t.; f ) m-CPBA, CH2Cl2, 0 ЊC.
NaHCO3, MeOH, 60 ЊC; e) MeCOCN2P(O)(OMe)2, K2CO3, MeOH, 0 ЊC
to quinolines 6 or 7 has been thoroughly studied by us and
recently reported.18 Applying this methodology, all derivatives
8–10a,b may be prepared with moderate to good stereo-
selectivity, depending upon the nature of the primary alcohol
protecting group.18
We chose to explore first route A, using 10a,b, in which a
protected propargylic hydroxymethyl group is already present,
as starting material. After trityl removal under acid conditions,
the inseparable diastereomeric mixture of 12a,b was oxidised to
the aldehydes 14a,b under slightly modified Swern conditions.19
However, treatment of the aldehydes with CBr4/PPh3, following
the Corey–Fuchs protocol,16 did not furnish the expected
dibromoolefins 16a,b, probably because of competitive electro-
philic reactions on the highly activated propargylic p-methoxy-
benzyl ether (Scheme 2).
into the corresponding alkynes 17a,b was more problematic,
representing probably the most crucial step of the whole
synthesis. Initial experiments using the standard conditions
(n-BuLi) under nitrogen for the dehydrobromination/debromi-
nation step furnished unsatisfactory (not higher than 40%!)
and poorly reproducible yields. The use of different bases
(t-BuLi, LDA) gave even worse results. We tried also a two step
sequence, passing through the formation of bromoalkyne
promoted under mild conditions by sodium bis(trimethyl-
silylamide) (NaHMDS):23 however this method led only to
partial desilylation of the preexisting alkyne as well as an
elimination reaction at positions 1–2 with the consequent
formation of quinoline derivatives. This latter experiment indi-
cated that the H in position 2 of dihydroquinoline was rather
prone to base-mediated deprotonation reactions.
Therefore we decided to introduce later the protected
hydroxymethyl group and carried out the same sequence on
trimethylsilyl substituted alkynes 11a or 11b.20 In this case
the two pure diastereoisomers could be readily separated by
chromatography, and were submitted independently to the
following transformations. The preparation of the aldehyde was
quite troublesome, due to the propensity of these derivatives to
undergo epimerisation. While alternative oxidants, like for
example Dess–Martin reagent,21 did not work well, a modified
Swern procedure allowed us to suppress epimerisation, pro-
vided that a rapid work-up under neutral or slightly acidic
conditions was performed and that aldehydes 13a or 13b were
used without chromatographic purification just after isolation.
Initially we tried the direct conversion of the aldehydes into
triple bonds, using Ohira’s protocol.17 Under these conditions
also the desired C-desilylation occurred. Although the overall
yield of 18a,b from 11a was satisfactory (59%), we observed an
extended epimerisation: starting from pure 11a22 we isolated
a 7:3 mixture of 18a and 18b. This is probably a consequence
of the required basic conditions and of the working tem-
perature, which must be higher than 0 ЊC in order to allow in
situ generation of the reagent.
Suspecting that this might also be the main reason for the
unsatisfactory yields of the n-BuLi mediated reaction, and
hoping that removal of the double bond could make this hydro-
gen less acidic, we tried to carry out the dehydrobromination
step after introduction of the epoxide. We first attempted to
prepare the epoxy alcohol by oxidation of 8a,24 followed by
TBDMS removal. However, deprotection24 gave the desired
epoxy alcohol only in moderate yield. Moreover this compound
turned out to be unstable, rapidly forming a by-product which
lacked both the epoxide and the hydroxymethyl groups and
which possessed an hexocyclic double bond. So we carried
out epoxidation of the dibromoalkene 15a (Scheme 3). This
reaction was found to be selective furnishing 20a in excellent
yield. However, treatment of 20a with n-BuLi gave again only
a moderate yield of 21a together with up to 5–10% of a by-
product identified as 22a. The formation of 22a is once again
due to the unsuspected high acidity of the propargylic proton
and this was confirmed also by treatment of 20a with NaH-
DMS. Actually, instead of the expected bromoalkyne corre-
sponding to 21a, we isolated only the analogue of 22a, having
the dibromoolefin moiety instead of the terminal triple bond, in
45% yield, thus demonstrating that proton abstraction from the
propargylic carbon is preferred to dehydroalogenation.
So we turned our attention to the Corey–Fuchs method-
ology. While the first step, that is the preparation of dibromo-
olefins 15a and 15b, worked well, the following transformation
Due to the difficulties encountered with these alternative
routes, we again turned our attention on the n-BuLi promoted
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3968

View Article Online
Scheme 3 Reagents and conditions: a) m-CPBA, CH2Cl2, 0 ЊC; b) n-
BuLi, toluene, Ϫ78 ЊC.
transformation of 15a,b into 17a,b searching for a way to
optimize it. We eventually found that careful control of tem-
perature (Ϫ78 ЊC), of the reaction time (15 min), and of the
amount of n-BuLi (2.2 eq) was beneficial. Most of all, perform-
ing the reaction under high purity argon instead of nitrogen,
the isolated yield was dramatically increased to 82–89%!
The following C-desilylation took place under unusually mild
conditions,25 thus affording 18a and 18b with an overall yield
of 72% and 76% from 11a and 11b respectively. Although this
route was longer than Ohira’s method, the yield was better and,
most importantly, no epimerisation was detected.
Scheme 4 Reagents and conditions: a) NIS, AgNO3, THF;. r.t., dark;
b) (Z)-Me3SnCH᎐CHSnMe3, Pd(PPh3)4, LiCl, DMF, 70 ЊC; c) i. p-TSA
᎐
(0.5 M sol. in THF), benzene, cyclohexa-1,4-diene, r.t., 30 min; ii. Et3N,
r.t. 24 h.
Diacetylenic derivatives 18a,b could be suitable for the
accomplishment of the synthesis following route A, by the intro-
duction of two hydroxymethyl groups at the alkyne termini.
For this purpose we tried several methods. Deprotonation of
alkynes with various bases (n-BuLi, LDA), followed by trap-
ping of the anions with different electrophiles [2-(trimethyl-
silyl)ethoxymethyl] chloride (SEM-Cl), (p-methoxy)benzyl-
oxymethyl chloride (PMBOM-Cl), ethyl chloroformate] did
not succeed. Moreover, when SEM-Cl was used, we isolated a
small amount of a new compound in which the electrophile was
bonded to the carbon α to nitrogen, thus demonstrating again
the acidity of proton in position 2. Also an attempt to treat 20a
with n-BuLi and trapping in situ the acetylide with PMBOM-Cl
failed.
For this reason we finally turned our attention to route B,
having in mind to build the cyclic enediyne through a Stille
double coupling reaction involving a bis(iodide) and (Z)-bis-
(trimethylstannyl)ethylene under Pd(0) catalysis. Our initial
reluctance toward this strategy was due to the fact that it has
been only seldom used for assembling these systems.3,13–15 More
importantly, in the reported examples the reaction was per-
formed only on systems conformationally more rigid than ours.
In order to check this strategy we first epoxidised the double
bond under standard conditions and, according to a previous
report by Isobe,5 isolated only one diastereoisomer in nearly
quantitative yield.
For the Stille coupling we had to transform 19a and 19b into
the bis(iodides) 23a and 23b and the best found conditions
employed the N-iodosuccinimide/AgNO3 system (Scheme 4),
for both stereoisomers.26 From molecular mechanics calcu-
lations27 we were able to predict 23b (derived from 8b or 9b) as
the best suited diastereoisomer for the coupling reaction, owing
to the proximity of the iodine–bearing carbons. Actually, for
both 23a,b we found about 8 favoured conformations charac-
terised by similar energy; however, the average distance between
the two sp hybridised carbons is 5.0–6.0 Å in the conformations
of 23a and only 4.1–4.4 Å in those of 23b. Moreover, the most
favourable conformations of 23b are those with the hydrogen
(bonded at the same carbon as the methyl) directed toward
the aromatic ring of the bicyclic system, as in the enediyne
derivative 2b, obtained from it.
Initially our results for the coupling reaction were dis-
appointing and we had to optimise a series of parameters in
order to succeed. In particular a correct choice of the Pd(0)
source was important. The in situ generation of Pd(0) by
reduction of more stable Pd(OAc)2 by means of diphenyl-
phosphinoethane (dppe), did not work. In our hands also
tris(dibenzylideneacetone)dipalladium chloroform complex,
successfully used for an enediyne ring formation,13 did not
give the coupling reaction. Finally, we found the best catalyst to
be tetrakis(triphenylphosphine)palladium but, in contrast with
literature data,28 good results were obtained only when working
under the strict exclusion of water and oxygen, with the catalyst
weighed in a glove box in an inert atmosphere. Moreover high
dilution conditions were important in order to minimise the
double attack of the tin reagent molecule to both iodine-bear-
ing carbons. Finally, the addition of LiCl also had a beneficial
effect, as previously reported by Nantz.14 The experimental
data confirmed the theoretical predictions. Stereoisomer 23b
gave an excellent 89% overall yield of 2b, under optimized con-
ditions, and 24b was never recovered. On the contrary, 2a was
obtained in 46–60% yield, variable amounts of 24a (16–23%)
being always isolated.
In the 1H NMR spectra of 2a and 2b some diagnostic signals
showed the same trend previously observed for the tricyclic lac-
tones derived from 9a and 9b, which allowed the demonstration
of their relative stereochemistry.18 Actually, the chemical shifts
of H-16 in 2a (3.04 ppm vs. 3.97 ppm in 2b) or of the methyl
group in 2b (1.25 ppm vs. 1.77 ppm in 2a) are strongly shifted
upfield for isomer b since this group is directed toward the
aromatic ring, thus providing further evidence for the relative
stereochemistry of the adducts 8–10a,b.
In order to demonstrate the propensity of enediynes 2a,b to
undergo Bergman cyclisation after epoxide opening, we first
simply opened the epoxide in the presence of p-toluenesulfonic
acid.5,6 This reaction was very fast and was always complete after
30 min; after neutralisation of the acid with triethylamine,29 the
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3969

View Article Online
solution was stirred at r.t. for 24 h in the presence of cyclohexa-
1,4-diene as hydrogen atom donor. Then we isolated two dif-
ferent products: monotosylate 25 from 2a and diol 26 from 2.30
We do not have a sure explanation for this divergent behaviour
of very similar compounds that differ only by the configuration
of one stereocentre, albeit near to the reaction site. However, we
feel that two different competing mechanisms, whose relative
rates are influenced by the different conformational equilibria
in the two epimers, may be operating.
Once sure that the epoxide opening will promote the cyclo-
aromatisation, we turned our attention on the transformation
of 2a,b into analogues that may be triggered under physio-
logical conditions. After a literature examination, we decided
to remove the phenyl carbamate, which is too stable,6,31 and to
substitute it with a 2-sulfonylethyl carbamate, previously
developed by Nicolaou32–35 and used also by Unno and Isobe.31
Such compounds are known to undergo a facile β-elimination
process at pH 8–8.5. The nucleophilic substitution promoted by
2-phenylthioethylate worked fine on 2b, giving 27b in 77% yield
(Scheme 5). The latter was then submitted to oxidation with
m-CPBA to afford target compound 28b in excellent yield.
Unexpectedly, the same sequence could not be reproduced on
2a. Actually, treatment of 2a with 2-phenylthioethylate fur-
nished not even traces of 27a, while the recovery of starting
material was low. Nor by switching to milder conditions (2-
phenylthioethanol, Cs2CO3, 18-crown-6)32 did we obtain 27a.
However, we isolated about 10% of 2b and a certain amount of
the symmetric carbonate derived from 2-phenylthioethanol.
Most likely, in this case, after expected nucleophilic attack of
the alkoxide, for some unclear reasons, the tetrahedric inter-
mediate prefers to eject the N-heterocycle instead of phenoxide,
with the formation of an unsymmetrical carbonate, which is
finally converted to the symmetrical one by reaction with
another molecule of alkoxide. Once deprotected, enediyne 29
may undergo extended decomposition through Bergman
cycloaromatization (in the absence of cyclohexa-1,4-diene,
the diradical may follow various degradation processes). On the
other hand, the basic reaction conditions can also promote
the epimerisation responsible for the formation of 2b. Another
possibility is removal of the phenyl carbamate by reduction
with LiAlH4,6 and to trap the resulting amine 29, without iso-
lation, with 2-phenylthioethyl chloroformate36 under classical
Schotten–Baumann conditions. In this case also, however, the
only identified product was the cycloaromatised derivative 30,
thus demonstrating the instability of free amine 29.
To avoid these undesired reactions we tried to protect the
triple bonds as cobalt complexes.6 Deep green solid 31 was
treated with 2-phenylthioethylate and compounds 31, 2a and
32 were isolated in 11, 25 and 43% yields, respectively. The
structure of 32 was demonstrated by analogy with literature
data and confirmed by treatment with Me3NO, under con-
ditions known to deprotect these complexes.6 Actually, 32 was
completely transformed into 2a.
Use of 2-phenylthioethyl chloroformate instead of phenyl
chloroformate at the beginning of the synthesis, during the
diastereoselective addition to 6,7, seems not to be promising on
the basis of our experience with this reaction.18,37 Moreover the
compatibility of this group with the following transformation
cannot be foreseen in advance. During the preparation of the
epoxide the oxidation to sulfone will take place as well, giving
a quite unstable product, which has to be submitted to a series
of other reactions. At the moment we have not yet solved the
problem of preparation of 27a or 28a.
Finally, we turned our attention on the activity evaluation of
compounds 2a,b and 28b. For this purpose we incubated com-
pounds 2a, 2b and 28b for 24 h at 37 ЊC and pH 8.48 with
supercoiled pBR322 plasmid (about 90% in form I) (Fig. 1).
As expected 2a and 2b did not show cleaving activity at all.
With 28b, in contrast, single-strand breaks were clearly evident
at concentrations of the enediyne as low as 1 µM. Moreover,
double-strand breaks, the most efficient in damaging DNA,
were still present in a 5 µM solution of 28b, and became much
more important when the incubation solution was more con-
centrated. The concentration that afforded equal quantities
of forms I and II was about 32 µM. The presence of DNA
scissions is most likely ascribed to carbamate removal through a
β-elimination reaction, followed by nitrogen-promoted epoxide
opening and final cycloaromatisation.
Compound 28b showed a very promising activity with
respect to DNA cleavage. Comparing its activity with other
dynemicin analogues this enediyne appears to be one of the
Scheme 5 Reagents and conditions: a) PhS(CH2)2OH, NaH, THF,
0 ЊC
r.t.; b) m-CPBA, CH2Cl2, 0 ЊC; c) i. LiAlH4, THF, 0 ЊC;
ii. PhS(CH2)2OCOCl, NaHCO3–H2O, r.t.; d) Co2(CO)8, CH2Cl2, r.t.;
e) Me3NO, CH2Cl2, r.t.
Fig. 1 Results of incubation of compounds 2a,b and 28b with
pBR322 plasmid (67.5 µM bpϪ1) for 24 h in pH 8.48 buffer at 37 ЊC.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3970

View Article Online
most active. Interestingly, by analysing structurally related
compounds prepared by Nicolaou,6,32,33,35,38 Isobe31,39 and
Wender7 it seems that the highest DNA cleavage ability is
showed by those compounds bearing a tertiary and not a
quaternary carbon in the propargylic position α to the benzylic
terminus of the epoxide.6,7,33,35 Moreover, a hydroxy (and an
alkoxy even more) group is usually responsible for a lower
activity, while the combination of a quaternary carbon with an
alkoxy group represents the worst situation.31,32,35,38,40
It’s noteworthy that often the dynemicin analogues syn-
thesized so far have a hydroxy or alkoxy group bonded to one
propargylic carbon as a consequence of the fact that enediyne
ring is formed through a Nozaki reaction. Our derivatives, by
contrast, lack this functionality since the cyclic enediyne is built
through a Stille coupling. Compound 28b, having a tertiary
carbon and lacking the oxygenated functionality, fits well for
the biological activity, as proved also by our experiments.
4-{(S*)-[(1-Methyl-2-oxo)ethyl]}-2-{(R*) [(trimethylsilyl)-
ethynyl]}-2H-quinoline-1-carboxylic acid phenyl ester 13a and
its epimer 13b
A solution of dry DMSO (1.06 ml, 15.04 mmol) in dry CH2Cl2
(12 ml) was cooled to Ϫ78 ЊC. A solution of oxalyl chloride
(2.22 M in CH2Cl2, 4.23 ml, 9.40 mmol) was added, followed,
after 10 min, by a solution of 11a or 11b (1.52 g, 3.76 mmol) in
dry CH2Cl2 (18 ml). After additional 10 min, ethyl diisopropyl-
amine (5.89 ml, 33.84 mmol) was added and the resulting slurry
was stirred overnight at Ϫ78 ЊC (about 15 h). Quenching with
5% aq (NH4)H2PO4 was followed by extractive work-up with
ether from pH 7. The combined organic layers were extracted
again with (NH4)H2PO4, and were finally washed with brine
with a small amount of 5% aq NaHCO3 solution added. After
drying over Na2SO4, the solution was concentrated. The
residue was dissolved in ether and dried again over Na2SO4.
After solvent removal crude aldehyde, as a pale yellow foam,
was used as such for the following step. Characterization of 13a:
Rf 0.33 (toluene/CH2Cl2/Et2O 70:29:1, A, B); GC-MS Rt 11.14;
m/z 403 (Mϩ, 10), 403 (10), 374 (6.8), 347 (6.1), 346 (20), 326
(14), 310 (9.2), 282 (8.9), 267 (5.3), 266 (16), 253 (6.0), 252 (16),
208 (8.2), 180 (7.8), 151 (5.3), 144 (10), 77 (28), 75 (9.5), 74 (8.3),
73 (100), 59 (5.3), 51 (6.0), 45 (9.4), 43 (6.6). Characterization
of 13b: Rf 0.42 (toluene/CH2Cl2/Et2O 70:29:1, A, B); GC-MS
Rt 11.17; m/z 403 (Mϩ, 24), 310 (13), 294 (8.5), 292 (6.6), 282
(18), 267 (12), 266 (28), 253 (6.3), 252 (24), 236 (6.2), 208 (9.6),
180 (6.0), 172 (6.7), 151 (5.2), 139 (9.1), 130 (8.6), 83 (7.9),
77 (20), 75 (14), 74 (8.5), 73 (100), 45 (7.9), 43 (6.1).
Conclusion
We have reported here a novel synthesis of dynemicin ana-
logues, characterised by an interesting DNA scission activity.
The presence of a consistent double-strand cleavage (about
70 : 30 form II : form III) makes 28b particularly promising.
A possible improvement of the activity is represented by the
synthesis of optically active analogues by a chemoenzymatic
procedure,41 having a functionalised group instead of the
methyl that will act as a handle for introducing suitable sub-
structures able to give additional interactions with DNA. This
strategy will enable us to prepare a series of similar compounds
with possibly enhanced activity after the enediyne formation.
Moreover, we are studying different “triggers” that can be
more easily introduced into the molecule, in order to promote
the epoxide opening through another mechanism.
4-{(R*)-[(3,3-Dibromo-1-methyl)allyl]}-2-{(R*) [(trimethyl-
silyl)ethynyl]}-2H-quinoline-1-carboxylic acid phenyl ester 15a
and its epimer 15b
A mixture of CBr4 (2.49 g, 7.52 mmol) and PPh3 (3.95 g,
15.04 mmol) was cooled to Ϫ78 ЊC and treated with dry CH2Cl2
(20 ml) for 10 min. The suspension was then allowed to stir
at Ϫ20 ЊC for 10 min and an orange solution was obtained.
After cooling again to Ϫ78 ЊC, a solution of crude aldehyde
(3.75 mmol) in dry CH2Cl2 (18 ml) was added and the mixture
was stirred at the same temperature for 20 min. Then the
temperature was allowed to raise to Ϫ50 ЊC. After 1.5–2 h the
reaction was usually complete and the brown solution was con-
centrated in vacuo and directly purified by chromatography with
Our results in this field will be reported in due course.
Experimental
NMR spectra were taken in CDCl3, at 200 MHz (1H), and
50 MHz (13C), using TMS as internal standard. Chemical shifts
are reported in ppm (δ scale), coupling constants are reported
1
in Hertz. Peak assignment in H NMR spectra was also made
with the aid of double resonance experiments. In AB systems,
the proton A is considered downfield and B upfield. Peak
assignment in 13C spectra was made with the aid of DEPT
experiments. GC-MS were carried out on a HP-5971A instru-
ment, using an HP-1 column (12 m long, 0.2 mm wide), electron
impact at 70 eV, and a mass temperature of about 170 ЊC.
Unless otherwise indicated analyses were performed with a
constant He flow of 0.9 ml minϪ1, initial temp. 100 ЊC, initial
time 2 min, rate 20 ЊC minϪ1, final temp. 260 ЊC, final time
4 min, initial temp. 250 ЊC, detector temp. 280 ЊC. Rt are in min.
IR spectra were measured with a Perkin-Elmer 881 instrument
as CHCl3 solutions. Melting points were measured on a Büchi
535 apparatus and are uncorrected. TLC analyses were carried
out on silica gel plates, which were developed by these detection
methods: A) UV; B) dipping into a solution of (NH4)4MoO4ؒ4
H2O (21 g) and Ce(SO4)2ؒ4 H2O (1 g) in H2SO4 (31 ml) and H2O
(469 ml) and warming. Rf were measured after an elution of
7–9 cm. Chromatograpy was carried out on 220–400 mesh silica
gel (if not otherwise specified) using the “flash” methodology.
Petroleum ether (40–60 ЊC) is abbreviated as PE. In extractive
work-up, aqueous solutions were always re-extracted thrice
with the appropriate organic solvent. Organic extracts were
washed with brine, dried over Na2SO4 and filtered, before
evaporation of the solvent under reduced pressure. All reac-
tions employing dry solvents were carried out under a nitrogen
atmosphere, if not otherwise specified. The purity of all com-
pounds was established by TLC, 1H and 13C-NMR, GC-MS.
PE/Et2O 95:5
8:2 to give 2.00 g of 15a or 1.96 g of 15b in
95% and 93% overall yield from 11a or 11b respectively as white
foams. Both compounds have been crystallised from Et2O/PE
mixtures to give white crystals. Characterization of 15a: Rf 0.41
(PE/Et2O 9:1, A, B); Mp 117.5–118.3 ЊC (PE/Et2O); found: C,
53.4; H, 4.55; N, 2.55. C25H25Br2NO2Si requires: C, 53.68;
H, 4.50; N, 2.50%; IR: νmax 2964, 2166, 1713, 1382, 1302, 1246,
1018, 842; GC-MS: GC-MS (usual method but with final temp.
290 ЊC) Rt 12.34; m/z 482 [Mϩ(79Br/81Br) Ϫ77, 9.5], 466 (6.9),
438 (6.7), 347 (57), 346 (62), 290 (6.2), 233 (7.6), 226 (12), 213
(6.2), 204 (9.1), 180 (10), 151 (7.0), 139 (11), 137 (8.6), 94 (5.9),
77 (34), 75 (8.4), 74 (9.6), 73 (100), 65 (5.3), 51 (7.6), 45 (7.7), 44
1
(9.3), 43 (5.6), 40 (12); H NMR: δ 0.046 [9 H, s, –Si(CH3)3],
1.37 [3 H, d, >CHCH3, J = 6.8], 3.82 [1 H, centre of m,
>CHCH3], 5.98 and 5.98 [2 H, AB system, H2 and H3, J = 7.2],
6.19 [1 H, d, –CH᎐CBr , J = 9.4], 7.17–7.44 [8 H, m, aromatics],
᎐
2
7.74 [1 H, broad d, H8, J = 7.0]; 13C NMR: δ Ϫ0.23 [3 C,
–Si(CH3)3], 17.03 [>CHCH3], 38.34 [>CHCH3], 45.17 [C2],
89.00 and 101.15 [2 C, –C᎐CTMS], 89.95 [᎐CBr ], 121.28,
123.60, 124.84, 124.98 and 127.96 [5 C, C3 and C5–8], 121.58
[2 C, C ortho of Ph], 125.68 [C para of Ph], 127.15 [C4a], 129.32
[2 C, C meta of Ph], 134.35 and 136.88 [2 C, C4 and C8a], 141.71
᎐
᎐
᎐
2
[–CH᎐CBr ], 151.06 [C ipso of Ph], 151.80 [>CO]. Characteriz-
᎐
2
ation of 15b: Rf 0.37 (PE/Et2O 9:1, A, B). Mp 152.8–153.2 ЊC
(PE/Et2O); found: C, 53.8; H, 4.6; N, 2.6. C25H25Br2NO2Si
requires: C, 53.68; H, 4.50; N, 2.50%; IR: νmax 2966, 2171, 1715,
1383, 1329, 1302, 1193; GC-MS (usual method but with final
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3971

View Article Online
temp. 290 ЊC) Rt 12.54; m/z 484 [Mϩ(79Br/81Br) Ϫ77, 5.5], 482
(11), 480 (5.5), 466 (7.3), 347 (17) 346 (62), 290 (5.5), 233 (7.5),
226 (12), 213 (5.9), 204 (8.6), 180 (9.9), 151 (6.8), 139 (13), 137
(11), 77 (38), 75 (6.1), 74 (7.9), 73 (100), 65 (5.6), 51 (6.2), 45
(72), 43 (5.2); 1H NMR: δ 0.099 [9 H, s, –Si(CH3)3], 1.33 [3 H, d,
>CHCH3, J = 7.0], 3.78 [1 H, centre of m, >CHCH3], 6.00 and
oxopropyl)phosphonate (45 mg, 234 µmol), prepared from
dimethyl 2-oxopropylphosphonate sodium anion and tosyl-
azide42 through a literature procedure,43 in dry methanol (2 ml)
was cooled to 0 ЊC and treated with anhydrous K2CO3 (22 mg,
156 µmol). After 30 min at 0 ЊC the mixture was stirred for
7.5 h at r.t. Quenching with saturated aq NH4Cl was followed
by usual extraction with Et2O. Crude product was purified by
chromatography with PE/Et2O 7:3, to give 18a,b (30 mg,
59% from 11a) as a 7:3 diastereomeric mixture, as determined
by 1H NMR.
Starting from 17a,b: a suspension of 17a or 17b (1.15 g,
2.88 mmol) was dissolved in dry methanol (22 ml), treated with
NaHCO3 (967 mg, 11.51 mmol) and stirred at 60 ЊC for 2 h. The
mixture was concentrated under vacuum, partitioned between
H2O/Et2O and extracted as usual. Chromatography with PE/
6.00 [2 H, AB system, H and H , J = 7.2], 6.55 [1 H, d, –CH᎐
᎐
2
3
CBr2, J = 8.8], 7.22–7.47 [8 H, m, aromatics], 7.79 [1 H, broad d,
H8, J = 7.0]; 13C NMR: δ Ϫ0.22 [3 C, –Si(CH3)3], 18.85
[>CHCH3], 38.92 [>CHCH3], 45.04 [C2], 89.20 and 100.99
᎐
[2 C, –C᎐CTMS], 89.93 [=CBr ], 121.12, 123.19, 124.97, 125.06
᎐
2
and 127.87 [5 C, C3 and C5–8], 121.57 [2 C, C ortho of Ph],
125.72 [C para of Ph], 127.37 [C4a], 129.34 [2 C, C meta of Ph],
134.26 and 137.08 [2 C, C and C ], 141.57 [–CH᎐CBr ], 150.94
᎐
4
8a
2
[C ipso of Ph], 151.81 [>CO].
Et2O 95:5
7:3 gave 18a or 18b (867 mg, 92% yield for both)
4-{(R*)-[(1-Methyl)prop-2-ynyl]}-2-{(R*) [(trimethylsilyl)-
ethynyl]}-2H-quinoline-1-carboxylic acid phenyl ester 17a and
its epimer 17b
as a pale yellow oil. Both compounds have been crystallised
from Et2O/PE mixtures to give white crystals. Characterization
of 18a: Rf 0.39 (PE/Et2O 8:2, A, B); Mp 138.2–138.7 ЊC (PE/
Et2O); found: C, 80.95; H, 5.25; N, 4.35. C22H17NO2 requires:
C, 80.71; H, 5.23; N, 4.28%; IR: νmax 3302, 2999, 2391, 1719,
1384, 1330, 1302, 1187; GC-MS: Rt 10.09; m/z 327 (Mϩ, 27),
326 (9.2), 283 (7.0), 282 (18), 274 (52), 250 (100), 234 (8.1), 230
(17), 219 (19), 216 (9.8), 206 (34), 205 (21), 204 (40), 192 (17),
191 (56), 190 (41), 189 (18), 180 (13), 179 (7.8), 178 (13), 168
(12), 166 (8.8), 165 (14), 164 (12), 163 (11), 154 (24), 153 (7.1),
152 (10), 139 (8.2), 128 (9.0), 127 (8.3), 77 (49), 65 (16), 63 (10),
To a solution of 15a or 15b (2.07 g, 3.70 mmol) in dry toluene
(20 ml) under argon and cooled to Ϫ78 ЊC a solution of n-BuLi
(5.09 ml, 1.6 M in hexanes, 8.14 mmol) was added dropwise.
After 15 min acetic acid (1.6 ml) was added, followed by water
(10 ml). After usual extraction with Et2O the solvent was
evaporated and crude product was purified by chromatography
with PE/Et2O 95:5
8:2 to give 17a (1.21 g, 82%) or 17b
(1.31 g, 89%) as a pale yellow foam. Both compounds have
been crystallised from Et2O/PE mixtures to give white crystals.
Characterization of 17a: Rf 0.31 (PE/Et2O 9:1, A, B); Mp
133.9–134.4 ЊC (PE/Et2O); found: C, 74.9; H, 6.4; N, 3.45.
C25H25NO2Si requires: C, 75.15; H, 6.31; N, 3.51%; IR: νmax
3304, 2962, 2170, 1717, 1382, 1328, 1245, 960, 842; GC-MS:
Rt 10.69; m/z 399 (Mϩ, 16), 347 (19), 346 (65), 323 (7.8), 322
(29), 307 (5.7), 306 (22), 291 (5.5), 290 (13), 278 (11), 276 (5.4),
262 (6.4), 232 (7.0), 226 (8.6), 217 (5.2), 204 (5.0), 180 (5.8), 151
(6.1), 97 (5.4), 77 (32), 73 (100), 59 (6.5), 53 (6.0), 51 (6.3), 45
53 (7.1), 51 (19.2), 39 (16); 1H NMR: δ 1.58 [3 H, d, >CHCH3,
᎐
J = 7.0], 2.19 and 2.24 [2 H, 2d, 2 –C᎐CH, J = 2.5 and 2.4], 3.79
᎐
[1 H, broad dq, >CHCH3, J = 2.0, 6.5], 6.02 [1 H, dd, H2, J =
2.3, 6.7], 6.18 [1 H, dd, H3, J = 1.1, 6.7], 7.17–7.44 [7 H, m,
aromatics], 7.55 [1 H, dd, H5, J = 1.7, 7.7], 7.78 [1 H, broad d,
H8, J = 7.7]; 13C NMR: δ 20.51 [>CHCH3], 27.33 [>CHCH3],
᎐
44.23 [C ], 70.26 and 71.99 [2 C, 2 –C᎐CH], 79.83 and 85.93
᎐
2
᎐
[2 C, 2 –C᎐CH], 121.07, 123.64, 124.78, 124.92 and 128.12 [5 C,
᎐
C3 and C5–8], 121.60 [2 C, C ortho of Ph], 125.80 [C para of Ph],
126.34 [C4a], 129.40 [2 C, C meta of Ph], 134.16 and 136.32
[2 C, C4 and C8a], 150.94 [C ipso of Ph], 151.87 [>CO]. Charac-
terization of 18b: Rf 0.39 (PE/Et2O 8:2, A, B). Mp 171.4–171.7
ЊC (PE/Et2O); found: C, 80.45; H, 5.35; N, 4.2. C22H17NO2
requires: C, 80.71; H, 5.23; N, 4.28%; IR: νmax 3303, 2984, 2116,
1722, 1382, 1328, 1301, 1262, 1189, 1024; GC-MS: Rt 10.36;
m/z 327 (Mϩ, 22), 282 (17), 275 (8.4), 274 (40), 251 (16), 250
(88), 230 (18), 219 (17), 216 (9.0), 206 (36), 205 (22), 204 (42),
192 (18), 191 (63), 190 (45), 180 (15), 179 (9.6), 178 (16), 168
(12), 165 (18), 164 (15), 163 (15), 154 (30), 153 (10), 152 (15),
151 (24), 139 (13), 128 (15), 127 (12), 126 (11), 115 (10), 89 (11),
78 (9.1), 77 (100), 76 (12), 75 (15), 65 (49), 63 (25), 53 (19),
1
(11), 43 (7.4), 39 (5.4); H NMR: δ 0.051 [9 H, s, –Si(CH3)3],
᎐
1.58 [3 H, d, >CHCH , J = 6.9], 2.18 [1 H, d, –C᎐CH, J = 2.5],
᎐
3
3.79 [1 H, broad dq, >CHCH3, J = 2.2, 6.8], 5.99 [1 H, d, H2,
J = 7.0], 6.16 [1 H, dd, H3, J = 0.9, 6.7], 7.16–7.44 [7 H, m,
aromatics], 7.52 [1 H, dd, H5, J = 1.6, 7.6], 7.76 [1 H, broad d,
H8, J = 7.7]; 13C NMR: δ Ϫ0.15 [3 C, –Si(CH3)3], 20.57
᎐
[>CHCH ], 27.32 [>CHCH ], 45.24 [C ], 70.14 [–C᎐CH], 86.11
᎐
3
3
2
᎐
᎐
[–C᎐CH], 89.04 and 101.16 [2 C, –C᎐CTMS], 121.64, 123.40,
᎐
᎐
124.68, 124.87 and 127.82 [5 C, C3 and C5–8], 121.64 [2 C,
C ortho of Ph], 125.69 [C para of Ph], 126.48 [C4a], 129.33 [2 C,
C meta of Ph], 134.28 and 135.74 [2 C, C4 and C8a], 150.99
[C ipso of Ph], 151.88 [>CO]. Characterization of 17b: Rf 0.34
(PE/Et2O 9:1, A, B). Mp 161.5–161.9 ЊC (PE/Et2O); found: C,
75.4; H, 6.45; N, 3.6. C25H25NO2Si requires: C, 75.15; H, 6.31;
N, 3.51%; IR: νmax 3305, 2962, 2169, 1714, 1381, 1328, 1303,
905; GC-MS: Rt 10.62; m/z 399 (Mϩ, 1.6), 346 (9.6), 306 (6.3),
323 (23), 77 (26), 75 (7.3), 74 (8.8), 73 (100), 59 (6.1), 53 (5.9), 51
1
51 (55), 50 (15), 39 (60); H NMR: δ 1.43 [3 H, d, >CHCH3,
᎐
J = 7.1], 2.24 and 2.39 [2 H, 2d, 2 –C᎐CH, J = 2.3 and 2.3], 3.89
᎐
[1 H, broad q, >CHCH3, J = 7.2], 6.01 [1 H, dd, H2, J = 2.3, 6.7],
6.46 [1 H, dd, H3, J = 1.3, 6.7], 7.18–7.44 [8 H, m, aromatics],
7.79 [1 H, broad d, H8, J = 7.8]; 13C NMR: δ 21.21 [>CHCH3],
᎐
27.47 [>CHCH ], 44.23 [C ], 72.02 and 72.29 [2 C, 2 –C᎐CH],
1
᎐
3
2
᎐
(6.0), 45 (11), 43 (6.1); H NMR: δ 0.084 [9 H, s, –Si(CH3)3],
79.64 and 85.07 [2 C, 2 –C᎐CH], 121.58, 123.13, 124.92, 125.04
᎐
᎐
1.45 [3 H, d, >CHCH , J = 7.2], 2.42 [1 H, d, –C᎐CH, J = 2.5],
᎐
3
and 128.06 [5 C, C3 and C5–8], 121.54 [2 C, C ortho of Ph],
125.79 [C para of Ph], 125.99 [C4a], 129.37 [2 C, C meta of Ph],
134.31 and 136.04 [2 C, C4 and C8a], 150.85 [C ipso of Ph],
151.77 [>CO].
3.91 [1 H, broad q, >CHCH3, J = 6.5], 6.01 [1 H, d, H2, J = 6.6],
6.46 [1 H, d, H3, J = 6.8], 7.19–7.46 [8 H, m, aromatics], 7.80
[1 H, broad d, H8, J = 7.0]; 13C NMR: δ Ϫ0.20 [3 C, –Si(CH3)3],
᎐
21.30 [>CHCH ], 27.56 [>CHCH ], 45.27 [C ], 72.14 [–C᎐CH],
᎐
3
3
2
᎐
᎐
85.29 [–C᎐CH], 88.92 and 100.99 [2 C, –C᎐CTMS], 122.20,
᎐
᎐
123.01, 124.85, 125.06 and 127.80 [5 C, C3 and C5–8], 121.62
[2 C, C ortho of Ph], 125.71 [C para of Ph], 126.29 [C4a], 129.35
[2 C, C meta of Ph], 134.57 and 135.61 [2 C, C4 and C8a], 151.03
[C ipso of Ph], 151.83 [>CO].
(2S*,3S*,4R*)-3,4-Epoxy-2-ethynyl-4-{(R*)-[(1-methyl)prop-2-
ynyl]}-2H-quinoline-1-carboxylic acid phenyl ester 19a and its
epimer 19b
A solution of 18a or 18b (778 mg, 2.38 mmol) in dry CH2Cl2
(15 ml) was cooled to 0 ЊC and treated with m-CPBA (≈ 80%,
2.05 g, 9.51 mmol). After 4–6 h the reaction is usually complete.
Excess m-CPBA was removed by treatment with Me2S (698 µl,
9.51 mmol) at 0 ЊC. Saturated aq NaHCO3 solution was added
(25 ml) and the biphasic system was vigorously stirred at r.t. for
2-[(R*)-Ethynyl]-4-{(R*)-[(1-methyl)prop-2-ynyl]}-2H-
quinoline-1-carboxylic acid phenyl ester 18a and its epimer 18b
Starting from 13a: a solution of crude aldehyde 13a (155 µmol),
prepared as described above from 11a, and dimethyl (1-diazo-2-
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3972

View Article Online
15 min. Usual work-up with Et2O, followed by chromatography
with PE/Et2O 9:1 1:1 gave 19a (751 mg, 92%) or 19b
(2S*,3S*,4R*)-3,4-Epoxy-4-{(R*)-[(1-methyl)prop-2-ynyl]}-2-
[(trimethylsilyl)ethynyl]}-2H-quinoline-1-carboxylic acid phenyl
ester 21a
(775 mg, 95%) as a pale yellow foam. Both compounds have
been crystallised from Et2O/PE mixtures to give white crystals.
Characterization of 19a: Rf 0.45 (toluene/CH2Cl2/Et2O 90:5:5,
A, B); Mp 138.2–138.7 ЊC (PE/Et2O); found: C, 76.5; H, 5.05;
N, 4.15. C22H17NO3 requires: C, 76.95; H, 4.99; N, 4.08%; IR:
νmax 3304, 3008, 2124, 1722, 1381, 1323, 1287, 1025; GC-MS:
Rt 10.46; m/z 343 (Mϩ, 64), 290 (12), 250 (33), 222 (18), 207 (11),
206 (18), 205 (27), 204 (61), 194 (38), 193 (16), 192 (23), 191
(69), 190 (20), 183 (11), 180 (16), 179 (18), 178 (33), 170 (12),
167 (16), 166 (12), 165 (17), 156 (11), 155 (15), 154 (69), 152
(15), 146 (15), 141 (17), 140 (24), 128 (22), 127 (18), 115 (23), 90
(20), 89 (11), 77 (100), 65 (26), 63 (18), 53 (39), 51 (52), 50 (11),
The same procedure described for the preparation of 17a,b, but
working under nitrogen, was followed, starting from 20a
(104 mg, 180 µmol) to give 21a in 40–45% yield as a yellow oil
after chromatography with PE/Et2O 9:1
85:15. Rf 0.45 (PE/
Et2O 7:3, A, B); found: C, 72.45; H, 6.10; N, 3.25. C25H25NO3Si
requires: C, 72.36; H, 6.06; N, 3.37%; IR: νmax 3306, 2960, 2178,
1720, 1379, 1323, 1244, 1015; GC-MS: Rt 11.64; m/z 415 (Mϩ,
11), 362 (8.3), 334 (5.3), 322 (19), 278 (5.7), 168 (5.4), 156 (9.8),
154 (19), 146 (6.2), 97 (12), 83 (6.8), 77 30), 75 (8.8), 71 (8.3), 73
(100), 65 (5.1), 59 (7.6), 53 (9.3), 51 (6.9), 45 (8.1), 43 (7.6), 39
(5.3); 1H NMR: δ 0.011 [9 H, s, –Si(CH3)3], 1.48 [3 H, d,
39 (36); 1H NMR: δ 1.48 [3 H, d, >CHCH3, J = 7.0], 2.30 [2 H,
᎐
>CHCH , J = 6.9], 2.22 [1 H, d, –C᎐CH, J = 2.4], 3.68 [1 H,
᎐
3
᎐
d, 2 –C᎐CH, J = 2.6], 3.65 [1 H, broad dq, >CHCH , J = 2.4,
᎐
3
broad dq, >CHCH3, J = 2.2, 6.9], 4.09 [1 H, d, H3, J = 3.0], 5.85
[1 H, d, H2, J = 3.0], 7.11–7.42 [7 H, m, aromatics], 7.57 [1 H,
broad d, H5 or H8, J = 8.0], 7.64 [1 H, dd, H5 or H8, J = 1.1, 8.0];
6.7], 4.10 [1 H, d, H3, J = 3.0], 5.92 [1 H, t, H2, J = 2.6], 7.11–7.43
[7 H, m, aromatics], 7.59 [1 H, broad d, H5 or H8, J = 7.8], 7.80
[1 H, dd, H5 or H8, J = 1.2, 8.0]; 13C NMR: δ 18.28 [>CHCH3],
28.05 [>CHCH3], 43.75 [C2], 56.91 [C4], 63.46 [C3], 71.67 and
13C NMR: δ Ϫ0.44 [3C, –Si(CH3)3], 18.47 [>CHCH3], 28.10
᎐
[>CHCH ], 44.52 [C ], 56.78 [C ], 63.52 [C ], 71.53 [–C᎐CH],
᎐
3
᎐
2
4
3
᎐
᎐
74.21 [2 C, 2 –C᎐CH], 77.35 and 82.70 [2 C, 2 –C᎐CH], 121.47,
᎐
᎐
᎐
82.62 [–C᎐CH], 91.86 and 98.78 [2 C, –C᎐CTMS], 121.49,
᎐
᎐
125.60, 125.69, 127.33, 128.63 and 129.27 [9 C, aromatic CH],44
125.50 [C4a], 135.11 [C8a], 150.87 [C ipso of Ph], 153.40 [>CO].45
Characterization of 19b: Rf 0.52 (toluene/CH2Cl2/Et2O 90:5:5,
A, B); Mp 101.5–101.8 ЊC (PE/Et2O); found: C, 76.55; H,
4.85; N, 4.2. C22H17NO3 requires: C, 76.95; H, 4.99; N, 4.08%;
IR: νmax 3304, 2995, 2423, 1722, 1381, 1324, 1122, 1024. GC-
MS: Rt 10.42; m/z 343 (Mϩ, 78), 343 (78), 250 (51), 222 (22), 207
(16), 206 (26), 205 (39), 204 (94), 194 (54), 193 (22), 192 (32),
191 (100), 183 (14), 180 (21), 179 (24), 178 (48), 167 (22), 166
(16), 165 (22), 155 (20), 154 (97), 152 (20), 146 (20), 141 (21),
140 (30), 128 (28), 127 (23), 115 (27), 90 (21), 77 (98), 65 (24),
125.39, 125.59, 127.05, 127.52 128.40 and 129.23 [10 C,
aromatic CH and C4a],44 135.43 [C8a], 150.97 [C ipso of Ph],
153.20 [>CO].45
(2S*,3S*,4R*)-3,4-Epoxy-2-iodoethynyl-4-{(R*)-[(3-iodo-1-
methyl)prop-2-ynyl]}-2H-quinoline-1-carboxylic acid phenyl
ester 23a and its epimer 23b
To a solution of 19a or 19b (823 mg, 2.40 mmol) in dry THF
(20 ml) silver nitrate (41 mg, 240 µmol) and N-iodosuccinimide
(1,32 g, 5.75 mmol) were added at r.t. and the suspension was
stirred in the dark for 1.5 h. After filtration the solution was
partitioned between water and ether and extracted as usual.
1
63 (16), 53 (33), 51 (40), 39 (24); H NMR: δ 1.46 [3 H, d,
᎐
>CHCH , J = 7.1], 2.20 and 2.26 [2 H, 2 d, 2 –C᎐CH, J = 2.2
᎐
3
and 2.5], 3.46 [1 H, centre of m, >CHCH3], 4.10 [1 H, d, H3, J =
2.6], 5.89 [1 H, t, H2, J = 2.6], 7.12–7.43 [7 H, m, aromatics],
7.57 [1 H, broad d, H5 or H8, J = 7.7], 7.84 [1 H, dd, H5 or H8,
J = 7.7]; 13C NMR: δ 16.40 [>CHCH3], 27.82 [>CHCH3],45
43.79 [C2], 57.54 [C4], 64.47 [C3], 72.09 and 74.22 [2 C, 2
Chromatography with PE/Et2O 100:0
0:100 gave desired 23a
(204 mg, 80%) or 23b (210 mg, 82%) as yellow foams. Both
compounds have been crystallised from iPr2O/PE/CH2Cl2
mixtures to give pale yellow crystals. Characterization of 23a: Rf
0.46 (toluene/CH2Cl2/Et2O 70:29:1, A, B); Mp 125.8–126.4 ЊC
(dec.) (iPr2O/PE/CH2Cl2); IR: νmax 2928, 2194, 1722, 1378,
1324, 1289, 907; GC-MS: unsuitable for this analysis; 1H NMR:
δ 1.46 [3 H, d, >CHCH3, J = 7.0], 3.80 [1 H, q, >CHCH3, J =
6.9], 4.06 [1 H, d, H3, J = 2.9], 6.04 [1 H, d, H2, J = 2.9], 7.10–
7.42 [7 H, m, aromatics], 7.58 [1 H, broad d, H5 or H8, J = 8.0],
᎐
᎐
–C᎐CH], 77.19 and 83.68 [2 C, 2 –C᎐CH], 121.48, 125.68,
᎐
᎐
125.71, 127.42, 127.77, 128.65 and 129.30 [9 C, aromatic CH],
125.99 [C4a], 135.07 [C8a], 150.93 [C ipso of Ph], 153.40 [>CO].45
(2S*,3S*,4R*)-4-{(R*)-[(3,3-Dibromo-1-methyl)allyl]}-3,4-
epoxy-2-[(trimethylsilyl)ethynyl]}-2H-quinoline-1-carboxylic
acid phenyl ester 20a
7.64 [1 H, dd, H5 or H8, J = 1.3, 7.7]; 13C NMR: δ 14.14 and
᎐
14.20 [2 C, 2 –C᎐CI], 18.30 [>CHCH ], 30.24 [>CHCH ], 45.53
᎐
3
3
᎐
[C ], 57.09 [C ], 63.52 [C ], 87.75 and 92.63 [2 C, 2 –C᎐CI],
᎐
2
4
3
The same procedure described for the preparation of 19a,b was
followed, starting from 15a (435 mg, 778 µmol) to give 20a
(403 mg, 90%) as an ivory foam after chromatography with PE/
121.53, 125.31, 125.68, 125.78, 127.31, 128.73, and 129.34
[10 C, aromatic CH and C4a],44 135.02 [C8a], 150.91 [C ipso of
Ph], 153.10 [>CO].45 Characterization of 23b: Rf 0.49 (toluene/
CH2Cl2/Et2O 70:29:1, A, B); Mp 96.6–97.0 ЊC (dec.) (iPr2O/PE/
CH2Cl2); IR: νmax 3007, 2394, 2192, 1721, 1379, 1324, 1187, 917;
Et2O 95:5
9:1. Crystallization from Et2O/PE gave 20a as a
white solid. Rf 0.31 (PE/Et2O 9:1, A, B); Mp 152.6–153.0 ЊC
(PE/Et2O); found: C, 52.45; H, 4.35; N, 2.5. C25H25Br2NO3Si
requires: C, 52.19; H, 4.38; N, 2.43%; IR: νmax 3005, 2180, 1721,
1375, 1322, 1286, 899; GC-MS (usual method but with final
temp. 290 ЊC) Rt 12.84; m/z 575 [Mϩ(79Br/81Br), 2.7], 482 (12),
374 (34), 363 (15), 362 (38), 334 (18), 315 (11), 263 (12), 226
(22), 242 (15), 234 (10), 215 (16), 213 (30), 211 (18), 196 (10),
183 (18), 180 (18), 168 (26), 167 (10), 154 (14), 146 (13), 139
(14), 133 (12), 131 (12), 97 (17), 83 (83), 77 (63), 75 (10), 73
1
GC-MS: unsuitable for this analysis; H NMR: δ 1.43 [3 H, d,
>CHCH3, J = 7.1], 3.60 [1 H, broad q, >CHCH3, J = 7.2], 4.30
[1 H, d, H3, J = 2.6], 6.01 [1 H, d, H2, J = 2.9], 7.11–7.43 [7 H, m,
aromatics], 7.57 [1 H, broad d, H5 or H8, J = 8.1], 7.78 [1 H,
broad d, H5 or H8, J = 7.3]; 13C NMR: δ 14.06 and 14.11 [2 C,
᎐
2 –C᎐CI], 16.32 [>CHCH ], 29.99 [>CHCH ], 45.56 [C ], 57.69
᎐
3
3
2
᎐
[C ], 64.35 [C ], 87.49 and 92.62 [2 C, 2 –C᎐CI], 121.46, 125.72,
᎐
4
3
127.32, 127.69, 128.69, and 129.30 [10 C, aromatic CH and
C4a],44 134.94 [C8a], 150.89 [C ipso of Ph], 153.18 [>CO].45
1
(100), 65 (11), 51 (13); H NMR: δ 0.029 [9 H, s, –Si(CH3)3],
1.22 [3 H, d, >CHCH3, J = 6.8], 3.66 [1 H, dq, >CHCH3, J =
6.6, 8.8], 3.96 [1 H, d, H3, J = 3.4], 5.84 [1 H, d, H2, J = 3.0], 6.50
(1R*,9S*,16R*,17S*)-1,17-Epoxy-16-methyl-8-phenoxycarb-
onyl-8-azatricyclo[7.7.1.02,7]heptadeca-2(3),4,6,12-tetraene-
10,14-diyne 2a and its epimer 2b
[1 H, d, –CH᎐CBr , J = 8.8], 7.13–7.43 [8 H, m, aromatics], 7.56
᎐
2
[1 H, dd, H5 or H8, J = 1.0, 7.6]; 13C NMR: δ Ϫ0.43 [3C,
–Si(CH3)3], 15.53 [>CHCH3], 37.74 [>CHCH3] 44.63 [C2],
᎐
57.98 [C ], 63.83 [C ], 90.99, 91.84, 98.61 [3C, –C᎐CTMS and
Bis(iodide) 23a or 23b (261 mg, 438 µmol) was poured into a
two necked flask equipped with an addition funnel and dis-
solved in dry DMF (34 ml). All the apparatus was carefully put
under an argon atmosphere and kept in dark. Then a solution
᎐
4
3
= CBr2], 121.51, 125.64, 125.84, 127.11, 127.56, 128.51, 129.28
[9 C, aromatic CH], 126.82 [C ], 135.33 [C ], 138.30 [–CH᎐
᎐
4a
8a
CBr2], 151.01 [C ipso of Ph], 153.39 [>CO].45
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3973

View Article Online
᎐
of (Z)-bis(trimethylstannyl)ethylene [91%, 171 µl, 658 µmol,
prepared bubbling acetylene into a dioxane solution of hexa-
methylditin in the presence of Pd(Ph3)4 46] in DMF (30 ml)
was transferred into the addition funnel. Bubbling of argon for
30 min into both solutions of reactants was performed using
two different needles. Then Pd(PPh3)4 (61 mg, 52.8 µmol),
previously weighed in a glove box under nitrogen, was added to
23a or 23b and the resulting slurry was stirred at r.t. for 10 min;
finally, dry LiCl (45 mg, 1.06 mmol) was added in the flask and
then it was warmed to 70 ЊC. A slow addition of the tin reagent
was then performed over a period of about 2 h.47 15 min after
the end of the addition the reaction was usually complete. The
mixture was poured into crushed ice and extracted as usual
with Et2O performing, if necessary, a filtration over a celite pad.
Crude products were usually purified twice by chromatography,
–C᎐C–], 121.57 [2 C, C ortho of Ph], 125.40, 126.61, 127.70 and
᎐
128.54 [4 C, C5–8], 125.56 [C4a], 125.68 [C para of Ph], 129.33
[2 C, C meta of Ph], 135.11 [C8a], 146.50 and 148.90 [4 C, –CH =
CH–], 150.06 [C ipso of Ph], 153.30 [>CO].45
(1R*,9S*,16R*,17S*)-17-Hydroxy-16-methyl-8-phenoxycarb-
onyl-1-(p-toluenesulfonyloxy)-8-azatetracyclo[7.7.1.02,7.010,15]-
heptadeca-2(3),4,6,10(11),12,14-esaene 25
A solution of 2b (13.5 mg, 36.7 µmol) in dry benzene (850 µl)
was treated, at r.t., with cyclohexa-1,4-diene (150 µl) and p-TSA
(0.5 M in THF, 77 µl, 38.6 µmol). After 30 min triethylamine
(54 µl, 38.6 µmol) was added and the pale pink suspension was
stirred again for 24 h. The mixture was partitioned between
Et2O and 5% aq NaHCO3 and extracted as usual. Crude
product was purified by preparative TLC, using PE/Et2O 3:7 as
eluent, to give 25 (4.5 mg, 23%) as an oil. Rf 0.37 (PE/Et2O 3:7,
A, B); GC-MS: Rt 13.29; m/z 370 (Mϩ Ϫ171, 14), 369 (56), 352
(5.1), 340 (14), 261 (11), 249 (17), 248 (100), 247 (53), 246 (67),
234 (5.9), 233 (40), 232 (16), 231 (5.3), 230 (10), 228 (6.2), 220
(16), 218 (14), 217 (12), 216 (5.3), 215 (5.3), 205 (13), 204 (35),
203 (9.8), 202 (7.4), 176 (5.7), 132 (7.9), 128 (5.7), 115 (7.2), 94
(7.7), 91 (6.8), 77 (30), 65 (9.0), 51 (7.8), 40 (5.6), 39 (8.0), 65
(7.6), 63 (6.8), 51 (11), 39 (7.3); 1H NMR: δ 1.51 [3 H, d,
>CHCH3, J = 7.2], 2.49 [3 H, s, CH3 of Ts], 4.01 [1 H, d (became
a s after exchange with D2O), –OH, J = 4.0], 4.33 [1 H, broad q,
H16, J = 6.7], 5.08 [1 H, t, H17, J = 4.2], 6.27 [1 H, d, H9, J = 5.2],
6.92–7.73 [15 H, m, aromatics], 7.94 [2 H, appar. d, H ortho to
S, J = 8.6]; 13C NMR: δ 13.73 [>CHCH3], 21.76 [CH3 of Ts],
43.99 [C16], 60.04 [C9], 67.12 [C17], 95.02 [C1], 121.78 [2 C,
C ortho of Ph], 123.13, 123.62, 129.01 and 129.70 [4 C, C3–6],
125.75 [C para of Ph], 126.14 [2 C, C12 and C13], 127.43 [2 C,
C ortho to S], 128.66 [2 C, C11 and C14], 129.45 [2 C, C meta of
Ph], 129.94 [2 C, C meta to S], 132.72 [2 C, C2 and –SO2C],
138.73 [C7], 145.32 [3 C, C10, C15 and CH3C of Ts], 151.22
[C ipso of Ph].
using PE/Et2O 100:0
1:1, to give 2a (97 mg, 60%) or 2b (144
mg, 89%) as very pale yellow foams. Compound 2a was then
crystallised from Et2O to give white crystals. Enediyne 2a was
obtained with a variable yield (46–60%), the other main isolated
product being compound 24a (16–23%) as a pale yellow oil.
Characterization of 2a: Rf 0.46 (toluene/CH2Cl2/Et2O 70:29:1,
A, B); Mp decomposes at 188 ЊC without melting (Et2O); found:
C, 78.35; H, 4.75; N, 3.75. C24H17NO3 requires: C, 78.46; H,
4.66; N, 3.81%; IR: νmax 3004, 2968, 2180, 1722, 1319, 1240,
1107, 922; GC-MS: Rt 13.46; m/z 367 (Mϩ, 23), 339 (23), 338
(11), 274 (42), 246 (29), 244 (12), 231 (14), 230 (34), 229 (39),
228 (77), 227 (11), 218 (57), 217 (78), 216 (55), 215 (66), 214
(21), 204 (25), 203 (47), 202 (78), 201 (20), 191 (16), 190 (21),
189 (36), 176 (12), 152 (11), 146 (33), 128 (12), 115 (11), 102
(12), 101 (11), 94 (11), 90 (34), 77 (100), 76 (13), 75 (10), 65 (26),
1
63 (23), 51 (38), 50 (11), 39 (24); H NMR: δ 1.77 [3 H, d,
>CHCH3, J = 7.4], 3.04 [1 H, broad q, H16, J = 7.2], 3.89 [1 H, d,
H17, J = 3.0], 5.65 [1 H, dt, H12, J = 1.5, 10.2], 5.81 [1 H, d, H13,
J = 10.0], 5.89 [1 H, dd, H9, J = 1.8, 3.0], 7.14–7.42 [7 H, m,
aromatics], 7.56 [1 H, broad d, H3 or H6, J = 7.8], 7.92 [1 H, dd,
H3 or H6, J = 1.8, 8.2]; 13C NMR: δ 15.19 [>CHCH3], 37.42
[C16], 45.98 [C9], 60.00 [C1], 70.07 and 70.16 [C17],48 87.80, 91.03,
(1R*,9S*,16S*,17S*)-1,17-Dihydroxy-16-methyl-8-phenoxy-
carbonyl-8-azatetracyclo[7.7.1.02,7.010,15]heptadeca-2(3),4,6,
10(11),12,14-esaene 26
᎐
92.04 and 102.68 [4 C, –C᎐C–], 121.40, 125.25, 127.05, 128.24,
᎐
129.01, and 129.16 [6 C, C3–6 and C12–13], 121.51 [2 C, C ortho of
Ph], 125.31 [C2], 125.74 [C para of Ph], 129.35 [2 C, C meta of
Ph], 135.79 [C7], 150.97 [C ipso of Ph], 153.10 [>CO].45 Charac-
terization of 2b: Rf 0.49 (toluene/CH2Cl2/Et2O 70:29:1, A, B);
found: C, 78.30; H, 4.60; N, 3.9. C24H17NO3 requires: C, 78.46;
H, 4.66; N, 3.81%; IR: νmax 3007, 2200, 1720, 1380, 1302, 1281,
1185; GC-MS: Rt 13.65; m/z 367 (Mϩ, 51), 339 (11), 338 (15),
274 (42), 246 (28), 231 (12), 230 (20), 229 (32), 228 (61), 227
(11), 219 (12), 218 (57), 217 (71), 216 (44), 215 (47), 214 (15),
204 (19), 203 (35), 202 (63), 201 (15), 191 (14), 190 (18), 189
(28), 176 (9.5), 146 (29), 115 (9.0), 90 (20), 78 (9.0), 77 (100), 75
(10), 65 (18), 63 (20), 51 (30), 39 (28); 1H NMR: δ 1.25 [3 H, d,
>CHCH3, J = 7.0], 3.97 [1 H, broad q, H16, J = 7.0], 4.17 [1 H, d,
H17, J = 2.8], 5.66 [1 H, dd, H12, J = 1.4, 10.0], 5.80 [1 H, dd, H13,
J = 1.5, 10.0], 5.96 [1 H, dd, H9, J = 1.7, 2.7], 7.12–7.41 [7 H, m,
aromatics], 7.57 [1 H, broad d, H3 or H6, J = 8.2], 7.70 [1 H, dd,
H3 or H6, J = 1.8, 8.0]; 13C NMR: δ 12.07 [>CHCH3], 26.90
The same procedure used for the preparation of 25 was fol-
lowed, starting from 7.9 mg of 2b (21.5 µmol). Crude product
was purified by preparative TLC, using PE/Et2O 4:6 as eluent,
to give 26 (4.3 mg, 52%) as an oil. Rf 0.31 (PE/Et2O 4:6, A, B);
GC-MS: Rt 9.59; m/z 293 (Mϩ Ϫ94, 0.85), 248 (9.5), 235 (16),
234 (100), 233 (8.8), 232 (4.7), 217 (16), 204 (5.1), 146 (7.3), 128
(6.0), 116 (14), 115 (9.4), 102 (5.2), 91 (4.8), 90 (5.8), 77 (11),
51 (5.2); 1H NMR: δ 1.43 [3 H, d, >CHCH3, J = 7.0], 2.60 [1 H,
broad d (became a s after exchange with D2O), C17–OH, J = 4.7]
3.02 [1 H, broad s, C1-OH ], 3.31 [1 H, q, H16, J = 7.2], 4.39 [1 H,
t, H17, J = 4.9], 6.06 [1 H, d, H9, J = 4.7], 6.98–7.71 [13 H,
m, aromatics]; 13C NMR: δ 18.65 [>CHCH3], 48.97 [C16],
59.44 [C9], 64.92 [C17], 72.40 [C1], 121.72 [2 C, C ortho of Ph],
122.93, 125.05, 126.98, 127.96, 128.79, 128.90, and 130.02 [8 C,
aromatic CH],44 125.81 [C para of Ph], 126.14 [2 C, C12 and
C13], 128.66 [2 C, C11 and C14], 129.48 [2 C, C meta of Ph],
132.41 [C2], 134.08 [C7], 141.10 [2 C, C10 and C15], 151.11 [C ipso
of Ph].
[C16], 45.47 [C9], 58.92 [C1], 62.63 and 62.74 [C17],48 85.88, 90.82,
᎐
92.43 and 103.36 [4 C, –C᎐C–], 121.51 [2 C, C ortho of Ph],
᎐
121.80, 125.66, 126.83, 127.01, 128.73 and 129.38 [9 C, C2, C3–6
,
C12–13 and C meta of Ph], 125.78 [C para of Ph], 135.07 [C7],
150.98 [C ipso of Ph], 153.21 [>CO].45 Characterization of 24a:
Rf 0.34 (PE/Et2O 9:1, A, B); found: C, 53.10; H, 5.25; N, 1.9.
C32H37NO3Sn2 requires: C, 53.30; H, 5.17; N, 1.94%; GC-MS:
unsuitable for this analysis; 1H NMR: δ 0.10 and 0.22 [18 H, 2 s,
2 –Sn(CH3)3], 1.44 [3 H, d, >CHCH3, J = 7.0], 3.75 [1 H, broad
q, >CHCH3, J = 7.0], 4.04 [1 H, d, H3, J = 2.8], 6.06 [1 H, broad
(1R*,9S*,16S*,17S*)-1,17-Epoxy-16-methyl-8-[(phenylthio)-
ethoxy]carbonyl-8-azatricyclo[7.7.1.02,7]heptadeca-2(3),4,6,12-
tetraene-10,14-diyne 27b
NaH (16 mg, 60% in mineral oil, 408 µmol) was suspended
in dry THF (2 ml). Then, after cooling to 0 ЊC, 2-(phenylthio)-
ethanol (58 µl, 430 µmol) was added. After 125 min a solution
of 2b (79 mg, 215 µmol) in THF (4 ml) was added and the
solution was stirred for 1.3 h at 0 ЊC and then for 10 min at
r.t. The solution was poured into chilled water and extracted
with Et2O. The combined organic layers were rapidly washed
t, H , J = 2.4], 6.27–6.61 [2 H, m, –CH᎐CH–] 7.12–7.40 [7 H,
᎐
2
m, aromatics], 7.57 [1 H, broad d, H5 or H8, J = 8.0], 7.74
[1 H, broad d, H5 or H8, J = 7.8]; 13C NMR: δ 9.18 and 8.99
[6 C, 2 –Sn(CH3)3], 17.78 [>CHCH3], 28.58 [>CHCH3], 44.60
[C2], 57.51 [C4], 63.66 [C3], 83.70, 84.43, 86.16 and 89.76 [4 C,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3974

View Article Online
with 1 N NaOH and then with 5% aq (NH4)H2PO4 and brine.
Chromatography with PE/Et2O 100:0 7:3 gave 27b as a white
Incubation with plasmid DNA and analysis by gel electrophoresis
Working buffer: this was prepared by dissolving TRIS (4.84 g),
EDTA (584 mg) and acetic acid (1.142 ml) in distilled water
(1 l).
Loading buffer: this was prepared from Ficoll (75 mg), the
working buffer (500 µl) and a solution containing 0.25% bromo-
phenol blue and 0.25% xylene cyanol in pH 8.3 TRIS–borate–
EDTA buffer (500 µl).
Compounds 2a, 2b and 28a were dissolved in DMSO
(7–10 mM sol.) and diluted with additional DMSO to the
desired concentrations just before incubation.
foam (71 mg, 77%), that was crystallised from CH2Cl2/Et2O to
give a white solid. Rf 0.44 (toluene/CH2Cl2/Et2O 70:29:1, A, B);
Mp 160.6–161.1 ЊC (CH2Cl2/Et2O); found: C, 72.75; H, 4.9;
N, 3.3. C26H21NO3S requires: C, 73.04; H, 4.95; N, 3.28%;
IR: νmax 2992, 2201, 1704, 1389, 1325, 1302, 815; GC-MS:
unsuitable for this analysis; 1H NMR: δ 1.21 [3 H, d, >CHCH3,
J = 7.0], 3.16 [2 H, centre of m, –CH2CH2S–], 3.92 [1 H, broad
q, H16, J = 7.0], 4.09 [1 H, d, H17, J = 3.8], 4.18–4.45 [2 H, m,
–CH2CH2S–], 5.64 [1 H, dd, H12, J = 1.4, 10.0], 5.77 [1 H, dd,
H13, J = 1.4, 9.8], ≈ 5.80 [1 H, overlapped with previous signal,
H9], 7.16–7.45 [7 H, m, aromatics], 7.47 [1 H, broad d, H3 or H6,
J = 8.0], 7.65 [1 H, dd, H3 or H6, J = 1.6, 8.2]; 13C NMR: δ 12.04
[>CHCH3], 26.83 [C16], 32.46 [–CH2CH2S–], 45.03 [C9], 58.83
Plasmid pBR322 (Fermentas, 90% form I, 500 µg mlϪ1) was
diluted (1 : 10) with a pH 8.48 TRIS (40 mM)/EDTA (4 mM)
buffer (prepared with molecular biology water) to provide a
concentration of 50 µg mlϪ1, 75 µM bpϪ1. This solution (18 µl)
was treated with the appropriately diluted solution of enediyne
(2 µl). The resulting mixtures were incubated at 37 ЊC for 24 h.
At the end of this period, the solutions (20 µl) were treated with
loading buffer (5 µl) and analysed on agarose gel [prepared
from agarose (300 mg), working buffer (32 ml) and ethidium
bromide (15 µg)], by the submarine methodology. The gel was
immersed in working buffer (325 ml) containing ethidium
bromide (165 µg) and eluted at 80 mV. After elution, the gel was
observed at 302 nm (transilluminator) and photographed.
[C1], 62.77 and 62.87 [C17],48 64.84 [–CH2CH2S–], 85.81, 90.55,
᎐
92.67 and 103.32 [4 C, –C᎐C–], 121.84, 125.32, 125.54, 126.86,
᎐
128.61, 129.11 and 130.03 [11 C, aromatic CH and C12–13],
127.06 [C2], 134.94 and 135.12 [C7 and C ipso of Ph], 154.32
[>CO].
(1R*,9S*,16S*,17S*)-1,17-Epoxy-16-methyl-8-[(phenylsul-
fonyl)ethoxy]carbonyl-8-azatricyclo[7.7.1.02,7]heptadeca-
2(3),4,6,12-tetraene-10,14-diyne 28b
The same procedure described for the preparation of 19a,b was
followed, starting from 27b (70 mg, 164 µmol) and using 106 mg
(491 µmol) of m-CPBA (≈80%). Quenching with Me2S (36 µl)
was directly followed by extraction with Et2O, avoiding stirring
of the solution in the presence of aq NaHCO3, due to lability of
28b under basic conditions. After extraction the combined
organic layers were very rapidly washed with 5% aq NaHCO3
and then with brine. Chromatography with PE/Et2O 4:6
0:100 gave sulfone 28b (69 mg, 91%) as a white foam. Rf 0.26
(PE/Et2O 6:4, A, B); found: C, 67.75; H, 4.65; N, 2.95.
C26H21NO5S requires: C, 67.96; H, 4.61; N, 3.05%; IR: νmax
3680, 3006, 1707, 1323, 1303, 1144; GC-MS: unsuitable for this
analysis; 1H NMR: δ 1.20 [3 H, d, >CHCH3, J = 7.0], 3.50 [2 H,
t, –CH2CH2SO2–, J = 6.0], 3.91 [1 H, broad q, H16, J = 6.9], 4.03
[1 H, d, H17, J = 3.0], 4.37–4.63 [2 H, m, –CH2CH2SO2–], 5.64
[1 H, dd, H12, J = 1.4, 9.8], ≈ 5.60 [1 H, overlapped with previous
signal, H9], 5.77 [1 H, dd, H13, J = 1.2, 9.8], 7.12–7.34 [3 H, m,
aromatics], 7.52–7.68 [4 H, m, aromatics], 7.92 [2 H, appar. d, H
ortho to SO2]; 13C NMR: δ 12.03 [>CHCH3], 26.82 [C16], 45.11
[C9], 55.24 and 59.60 [2 C, –CH2CH2S–], 58.81 [C1], 62.53 [C17],
Acknowledgements
We gratefully thank Dr Dina Cavallo and Dr Sara Perrozzi for
their contribution to this work, Professor Luca Banfi for helpful
suggestions for gel electrophoresis experiments and MIUR
(COFIN 98 and 00) for financial support.
References
1 K. C. Nicolaou and W.-M. Dai, Angew. Chem., Int. Ed. Engl., 1991,
30, 1387–1430.
2 K. C. Nicolaou and A. L. Smith, The Enediyne Antibiotics, in
Modern Acetylenic Chemistry, ed. P. J. Stang and F. Diederich, VCH,
Weinheim, 1995; p. 203–283.
3 M. D. Shair, T. Y. Yoon, K. K. Mosny, T. C. Chou and
S. J. Danishefsky, J. Am. Chem. Soc., 1996, 118, 9509–9525;
T. Takahashi, Y. Sakamoto, H. Yamada, S. Usui and Y. Fukazawa,
Angew. Chem., Int. Ed. Engl., 1995, 34, 1345–1348.
4 A. G. Myers, N. J. Tom, M. E. Fraley, S. B. Cohen and D. J. Madar,
J. Am. Chem. Soc., 1997, 119, 6072–6094; J. Taunton, J. L. Wood
and S. L. Schreiber, J. Am. Chem. Soc., 1993, 115, 10378–10379;
M. E. Maier, F. Boße and A. J. Niestroj, Eur. J. Org. Chem., 1999,
1–13.
᎐
85.85, 90.71, 92.32 and 103.26 [4 C, –C᎐C–], 121.77, 125.55,
᎐
5 T. Nishikawa, A. Ino and M. Isobe, Tetrahedron, 1994, 50,
1449–1468.
125.63, 126.67, 126.86, and 128.69 [6 C, C3–6 and C12–13], 127.17
[C2], 128.05 [2 C, C ortho of Ph], 129.48 [2 C, C meta of Ph],
134.04 [C para of Ph], 134.74 [C7], 139.05 [C ipso of Ph], 153.74
[>CO].
6 K. C. Nicolaou, A. L. Smith, S. W. Wendeborn and C.-H. Hwang,
J. Am. Chem. Soc., 1991, 113, 3106–3114 and references therein.
7 P. A. Wender, C. K. Zercher, S. Beckham and E.-M. Haubold,
J. Org. Chem., 1993, 58, 5867–5869; P. A. Wender, S. Beckham and
J. G. O’Leary, Synthesis, 1994, 1278–1282.
8 L. Banfi and G. Guanti, Angew. Chem., Int. Ed. Engl., 1995, 34,
2393–2395; L. Banfi, A. Basso and G. Guanti, Tetrahedron, 1997,
53, 3249–3268; L. Banfi and G. Guanti, Eur. J. Org. Chem., 1998,
1543–1548; L. Banfi and G. Guanti, Tetrahedron Lett., 2000, 41,
6523–6526; L. Banfi and G. Guanti, Tetrahedron Lett., 2002,
43, 7427–7429; L. Banfi and G. Guanti, Eur. J. Org. Chem., 2002,
3745–3755.
Compound 31
A solution of 2a (30 mg, 81.7 µmol) in dry CH2Cl2 (3 ml) was
treated with Co2(CO)8 (90%, 78 mg, 204 µmol) and stirred at
r.t. for 10 min. After solvent evaporation crude product was
directly purified by chromatography using PE/Et2O 100:0
9:1
to give 31 as a green solid (66 mg, 86%). Rf 0.28 (PE/Et2O 9:1,
9 L. Banfi, G. Guanti and A. Basso, Eur. J. Org. Chem., 2000, 939–946.
10 Part of this work was already reported; G. Guanti and R. Riva,
Chem. Commun., 2000, 1171–1172.
11 T. Nishikawa, M. Yoshikai, K. Obi, T. Kawai, R. Unno, T. Jomori
and M. Isobe, Tetrahedron, 1995, 51, 9339–9352.
12 R. S. Huber and G. B. Jones, Tetrahedron Lett., 1994, 35, 2655–2658.
13 H. Tanaka, H. Yamada, A. Matsuda and T. Takahashi, Synlett,
1998, 381–383.
A, B); IR: νmax 3012, 2083, 2061, 2028, 1722; GC-MS: unsuit-
able for this analysis; H NMR: δ 1.99 [3 H, d, >CHCH3, J =
7.4], 3.32 [1 H, q, H16, J = 7.3], 4.13 [1 H, d, H17, J = 3.0], 6.14
and 6.25 [2 H, AB system, H12–13, J = 12.8], 6.65 [1 H, d, H9,
J = 3.0], 7.08–7.42 [7 H, m, aromatics], 7.61 [1 H, broad d, H3
or H6, J = 8.4], 7.75 [1 H, dd, H3 or H6, J = 1.6, 7.8]; 13C NMR:
1
δ 18.32 [>CHCH3], 49.57 and 55.69 [C9 and C16], 59.02 [C1],
14 D. K. Moss, J. D. Spence and M. H. Nantz, J. Org. Chem., 1999, 64,
᎐
69.92 [C ], 80.60, 82.90, 87.74 and 95.42 [4 C, –C᎐C–], 121.52
᎐
17
4339–4343.
[2 C, C ortho of Ph], 123.58, 124.58, 125.00, 127.32, 129.28 and
129.91 [6 C, C3–6 and C12–13], 124.69 [C2], 125.88 [C para of Ph],
129.50 [2 C, C meta of Ph], 134.95 [C7], 150.96 [C ipso of Ph],
153.90 [>CO],45 198.06 and 199.06 [12 C, CO].45
15 D. L. J. Clive, Y. Tao, Y. Bo, Y.-Z. Hu, N. Selvakumar, S. Sun,
S. Daigneault and Y.-J. Wu, Chem. Commun., 2000, 1341–1350.
16 E. J. Corey and P. L. Fuchs, Tetrahedron Lett., 1972, 13, 3769–3772.
17 S. Ohira, Synth. Commun., 1989, 19, 561–564.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3975

View Article Online
18 G. Guanti, S. Perrozzi and R. Riva, Tetrahedron: Asymmetry, 1998,
9, 3923–3927; G. Guanti, S. Perrozzi and R. Riva, Tetrahedron:
Asymmetry, 2002, 13, 2703–2706.
19 G. Guanti, L. Banfi, R. Riva and M. T. Zannetti, Tetrahedron Lett.,
1993, 34, 5483–5486.
20 We started from 6 to get larger amounts of the minor stereoisomer
11b through a less stereoselective addition to the quinoline ring,
while we used 7 for the largest production of 11a.
21 D. B. Dess and J. C. Martin, J. Am. Chem. Soc., 1991, 113, 7277–
7287.
33 K. C. Nicolaou, Y.-P. Hong, Y. Torisawa, S.-C. Tsay and W.-M. Dai,
J. Am. Chem. Soc., 1991, 113, 9878–9880; K. C. Nicolaou,
Y.-P. Hong, W.-M. Dai, Z.-J. Zeng and W. Wrasidlo, J. Chem.
Soc., Chem. Commun., 1992, 1542–1544; K. C. Nicolaou,
W.-M. Dai, Y.-P. Hong, S.-C. Tsay, K. K. Baldridge and J. S. Siegel,
J. Am. Chem. Soc., 1993, 115, 7944–7953.
34 K. C. Nicolaou, P. Maligres, T. Suzuki, S. V. Wendeborn, W.-M. Dai
and R. K. Chadha, J. Am. Chem. Soc., 1992, 114, 8890–8907.
35 K. C. Nicolaou, W.-M. Dai, S. C. Tsay, V. A. Estevez and
W. Wrasidlo, Science, 1992, 256, 1172–1178.
22 We demonstrated that no epimerisation had occurred during Swern
oxidation.
36 This compound was prepared by reacting 2-phenylthioethanol with
a solution of phosgene in toluene and it was used as such, after
solvent removal, as a 0.86 M solution in dry THF.
37 Only with phenyl carbamate the reaction is fast even at Ϫ78 ЊC,
giving the addition products in high yield and moderate to good
d.r.
23 D. Grandjean, P. Pale and J. Chuche, Tetrahedron Lett., 1994, 35,
3529–3530.
24 While HF provoked epoxide opening, use of n-Bu4NF gave also the
expected C-desilylation.
25 This was beneficial, since the phenylcarbamate moiety is quite
reactive under basic conditions, due to the propensity to release
phenol.
26 T. Nishikawa, S. Shibuya, S. Hosokawa and M. Isobe, Synlett, 1994,
485–486. In our hands the direct transformation of TMS-alkyne
moiety into iodide described by the authors was sluggish and overall
yields lower.
27 Calculations were performed with the CS Chem3D program
(Cambridge Soft, Cambridge, MA, USA, version 4.0).
28 V. Farina, V. Krishnamurthy and W. J. Scott, Org. React., 1997, 50,
1–652.
29 Neutralization of p-TSA was important in order to minimize
complete decomposition that cannot be avoided otherwise.
30 The relative configuration of benzylic carbon of 25 and 26 has not
been demonstrated and was attributed only tentatively, by analogy
with reported data (refs. 5,6).
38 K. C. Nicolaou and W.-M. Dai, J. Am. Chem. Soc., 1992, 114,
8908–8921.
39 R. Unno, H. Michishita, H. Inagaki, Y. Suzuki, Y. Baba, T. Jomori,
M. Moku, T. Nishikawa and M. Isobe, Bioorg. Med. Chem., 1997, 5,
903–919.
40 R. Unno, H. Michishita, H. Inagaki, Y. Suzuki, Y. Baba,
T. Jomori, T. Nishikawa and M. Isobe, Bioorg. Med. Chem., 1997, 5,
883–901.
41 L. Banfi, G. Guanti, A. Mugnoli and R. Riva, Tetrahedron:
Asymmetry, 1998, 9, 2481–2492.
42 M. Regitz, J. Hocker and A. Liedhegener, Org. Synth., 1973, coll. vol.
5, 179–183.
43 P. Callant, T. D’Haenens and M. Vandewalle, Synth. Commun.,
1984, 14, 155–161.
44 Several signals are overlapped.
45 Very broad signal.
31 R. Unno, H. Michishita, H. Inagaki, Y. Baba, T. Jomori,
T. Nishikawa and M. Isobe, Chem. Pharm. Bull., 1997, 45, 125–133.
32 K. C. Nicolaou, W.-M. Dai, S. V. Wendeborn, A. L. Smith,
Y. Torisawa, P. Maligres and C.-K. Hwang, Angew. Chem., Int. Ed.
Engl., 1991, 30, 1032–1036.
46 T. N. Mitchell, A. Amamria, H. Killing and D. Rutschow,
J. Organomet. Chem., 1986, 304, 257–265.
47 High dilution conditions, realized employing a syringe pump, were
less satisfactory.
48 Both due to the same carbon.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 6 7 – 3 9 7 6
3976