Toward a total synthesis of stigmatellin; An unproductive C-1'-C-2' disconnection

Article abstract of DOI:10.1016/S0040-4039(00)00813-3

Although the lithio derivative of the dimethylchromone 6c could be condensed with reactive organic electrophilic species and iodine to give, respectively, homologated at C-1' (stigmatellin numeration) chromones and the iodide 6a, it proved impossible to convert efficiently either 6a or 6c into stigmatellin 1a by an anionic alkylation process. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00813-3

Tetrahedron Letters 41 (2000) 5495±5499
Toward a total synthesis of stigmatellin; an unproductive
C-10±C-20 disconnection
N. Adje, L. Domon, F. Vogeleisen-Mutterer and D. Uguen*
Á Â Á Â
Laboratoire de Synthese Organique, associe au CNRS, Ecole Europeenne de Chimie, Polymeres et Materiaux,
Â
Â
Universite Louis Pasteur, 25 rue Becquerel, 67087 Strasbourg France
Received 31 March 2000; accepted 12 May 2000
Abstract
Although the lithio derivative of the dimethylchromone 6c could be condensed with reactive organic
electrophilic species and iodine to give, respectively, homologated at C-10 (stigmatellin numeration) chromones
and the iodide 6a, it proved impossible to convert eciently either 6a or 6c into stigmatellin 1a by an anionic
alkylation process. # 2000 Elsevier Science Ltd. All rights reserved.
Isolated a few years ago by Ho¯e from a culture of the gliding bacteria Stigmatella auriantaca
as a mixture of the two stereomers 1a and 1b,1 stigmatellin 1 has proved to be a potent inhibitor
of the photosynthetic system. It is, accordingly, a useful tool for studying relevant electron-
transfer phenomena.2
Although most of the structural features of 1a and 1b were determined earlier,1 it was not until
recently that both the relative and absolute con®gurations of their four stereogenic centres have
been ascertained by identi®cation of a degradation product of 1a to an authentic, synthetic,
sample.3 To date, no total synthesis of these compounds has been claimed.
With the aim to synthesise stigmatellin 1a according to the indicated plan, we ®rst examined
the possibility to generate the fragment 2a from the aldehyde 3a, which is easily accessible from
the sul®de 4, via 3b, by means inter alia of a modi®ed Kiliani±Fischer methodology4 (Scheme 1).
Condensation of 3a with the dienylsulfone E-5 in the conditions of the Julia±Paris±Kocienski
(JPK) reaction and subsequent oxidation of the expected sul®de 2b would have provided the tri-
enyl sulfone 2a. Alkylation of 2a by the iodochromone 6a, followed by hydrogenolysis of the
sulfonyl group would have then delivered the target molecule.
Although, somewhat unexpectedly, this plan proved to be of little value at an advanced stage
of this synthesis, a few interesting points emerged from this exploratory study, especially those which
concern the elaboration of the chromone derivative 6c. These results are, accordingly, reported
herein, whereas a more valuable synthetic scheme is presented in the accompanying letter.5
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00813-3

5496
Scheme 1.
The required iodide 6a was prepared by starting from commercial 3,5-dimethoxyphenol, which
was ®rst converted into the chromone 6b as described.1 Due to the sensitivity of 6b to air, especially
in basic media, its free phenol functionality was immediately protected by treatment with
TBDMSCl and imidazole. The stable chromone derivative 6c thus obtained was then reacted
sequentially with LDA and NIS to give, after recrystallisation, the iodo derivative 6a as yellow
prisms (37% overall, from 6b) (Scheme 2).
Scheme 2. Reagents and conditions: (1) TBDMSCl (2.5 equiv.), imidazole (2.5 equiv.); DMF, rt, 15 h (73%); (2) LDA
(1 equiv.), THF; ^78ꢀC, 1 h, then NIS (1.5 equiv.), ^78ꢀC, 2 h (37% overall, from 6b)
Next, the preparation of the dienyl sulfone E-5 was examined. Condensation of the bis-allylic
alcohol 7 with p-toluenesul®nyl chloride a€orded a 2:3 mixture (NMR) of, respectively, the sulfone
E-5 and 8 in high yield (92%) (Scheme 3). By treatment with excess p-toluenesul®nic acid in
CHCl3,6 this mixture evoluted to give, after a few hours, an oily product which was mainly con-
stituted of the desired sulfone E-5, admixed, however, with its Z-isomer Z-5 (E:Z=85:15, by
NMR). Chromatography of this product on silica gel and recrystallisation of the purest fractions
then furnished the pure sulfone E-5a (43%).
Scheme 3. Reagents and conditions: (1) p-toluenesul®nyl chloride (1.5 equiv.), NEt3 (1.6 equiv.), CH2Cl2; ^78ꢀC, 5 min
(92%); (2) p-toluenesul®nic acid (2.5 equiv.), CHCl3; rt, 12 h; (3) column chromatography on silica gel (CH2Cl2/hex-
ane), then recrystallisation from ether/hexane (43%)

5497
The possibility to engage E-5 in a JPK coupling was tested by treating this sulfone with butyl-
lithium in THF at low temperature. The thus-formed lithio derivative was reacted, on the one
hand, with acetaldehyde, and on the other hand, with 2-phenylpropanal. In both cases, this con-
densation step was followed by in situ addition of Ac2O and DMAP, then reduction of the
resulting mixture of diastereomeric acetoxy-sulfones by sodium amalgam in MeOH, diluted with
a phosphate bu€er, at low temperature.
Surprisingly, whereas 2-phenylpropanal a€orded the triene 9 as expected, the dienylsulfone 107
was the only product isolated in the experiment with acetaldehyde (Scheme 4). Notwithstanding
this divergence, a similar JPK sequence was applied to the aldehyde 3a. In this case, the reduction
step proceeded very slowly, requiring repeated additions of both HgNa and the phosphate bu€er
to go to completion. Nevertheless, the pure (NMR, UV) trienyl sul®de 2b was isolated in a fairly
good yield (70% overall, from 3a).
Scheme 4. Reagents and conditions: (1) 1.2 M (in hexane) n-BuLi (1.05 equiv.), THF; ^78 to 35ꢀC, 15 min, then ^78ꢀC,
0.5 h, then acetaldehyde (1.3 equiv.); ^78 to ^20ꢀC, 15 min; (2) same conditions as in (1) with 2-phenylpropanal; (3)
same conditions as in (1) with the aldehyde 3a; (4) (i) Ac2O (2.1 equiv.), DMAP (0.11 equiv.), rt, 1 h; (ii) 6% HgNa
ꢀ
.
(400 mg/mmol, three times), HNa2PO4 12H2O (3Â4 equiv.), MeOH; ^20 C, 5±7 h
Attempted oxidation of this sul®de in mild conditions (n-Bu4NIO4) proved inecient. Due to
the sensitivity of the trienyl residue in 2b (or 2a), only traces of 2a could be characterised and
after several unsuccessful experiments with various oxidising reagents, it became clear that our
initial plan needed some adjustment.
The alkylation of the sulfone 3c, which formed by oxidation of 3b with oxone in a bu€er media
(80%), by the iodochromone 6a was then considered, our hope being that the aldehyde 11b,
which should form by hydrolysis of the expected acetal 11a, would su€er a JPK8 condensation
with the sulfone E-5 (Scheme 5). In the event, a treatment with HgNa as above would have then
provided the O-TBDMS derivative of 1a.
Scheme 5. Reagents and conditions: (1) oxone (5 equiv.), pH 8±9 phosphate bu€er, MeOH; 0ꢀC, 4 h (80%; (2) n-BuLi
(1.3 equiv.), HMPT (1.3 equiv.), THF; ^78ꢀC to rt
This plan proved no more ecient than the preceding one, however, although 3c could indeed
be anionised by treatment with butyllithium, as evidenced by quenching with either D2O or allyl
bromide, treatment of the thus-formed lithio derivative by the iodide 6a did not provide usefully

5498
the desired sulfone 11a: due presumably to exceeding acidity of the CH2 group at position 2 of the
chromone nucleus in 11a, extensive elimination of the sulfonyl group occurred and, at best, only
trace amounts of 11a could be detected at low conversion.
Another possibility was to condense the iodide 3d with the lithio derivative of 6c (or the related
organocopper and cerium species) to form 11c: hydrolysis of this acetal, followed by a JPK con-
densation with E-5a and desilylation would have then delivered 1a. The validity of this scheme
was tested by condensing anionised 6c with various reagents (Table 1).9
Table 1
Condensation of anionised 6c with organic electrophines
As can be seen, whereas the more electrophilic reagents eciently gave the expected product
(Entries 1±4), a g-oxygenated iodide modelling accurately the iodide 3d gave only traces of the
desired product 6h, the main product being the O-alkylation product 6i (Entry 5), a result not
improved by using the corresponding tri¯ate or by ®rst reacting the lithio derivative of 6c with
either CuI or CeCl3. Owing to these consecutive diculties, the initial plan and its various variations
were de®nitively thrown out.
In conclusion, a few positive points emerged from this study as, for instance, the possibility to
generate the triene residue of stigmatellin by using an ole®nation process. Additionally, clean
homologation at C-10 of a 2-methylchromone has been realised in several cases.10 However, all
attempts to elaborate the chromone 6c toward the stigmatellin molecule 1a proved to be not very
ecient and it now seems clear to us that any viable strategy for synthesizing stigmatellin 1a
should comprise an early appendage of an advanced fragment of the C-2 side chain to a suitable
precursor of the heterocyclic part of 1a, prior to the generation of the chromone ring system.
Such an approach is described in the accompanying Letter.5
References
1. Ho¯e, G.; Kunze, B.; Zorzin, C.; Reichenbach, H. Liebigs Ann. Chem. 1984, 12, 1883±1904.
2. Oettmeier, W.; Goolde, D.; Kunze, B.; Ho¯e, G. Biochem. Biophys. Acta 1985, 106, 216±223.
3. Enders, J.; Osborn, S. J. Chem. Soc., Chem. Commun. 1993, 424±426.
4. Adje, N.; Vogeleisen, F.; Uguen, D. Tetrahedron Lett. 1996, 37, 5893±5896.
5. Domon, L.; Uguen, D. Tetrahedron Lett. 2000, 41, 5501±5505.
6. (a) Julia, M.; Nel, A.; Righini, A.; Uguen, D. J. Organomet. Chem. 1983, 235, 113±120; (b) Julia, M.; Nel, M.;
Uguen, D. Bull. Soc. Chim. Fr. 1987, 487±492.
7. Presumably formed by basic elimination of an acetate ion in the transcient acetoxysulfone, followed by
hydrogenation of the thus-formed vinylsulfone residue.
8. Protocol for the JPK ole®nation process (all operations performed in an argon atmosphere): To a cooled (ca.
^78ꢀC) solution of the sulfone E-5 (25.1 mg, 0.1 mmol) in THF was added a 1.2 M solution of n-BuLi in hexane
(0.1 ml). The resulting orange±red mixture was warmed up to ^45ꢀC for 5 min, then cooled to ^78ꢀC. The

5499
aldehyde 3a (14.9 mg, 0.05 mmol), diluted with THF (0.6 ml), was then added, which induced a slight
decolouration. Stirring was pursued 1 hour at ^78ꢀC, then 5 min at ^20ꢀC, after that DMAP (0.06 mg) and
Ac2O (0.01 ml) were added sequentially. After 45 min stirring at rt, the resulting thick mixture was diluted with
ether (10 ml), and water (9 ml). Extraction with ether (4Â2 ml) of the aqueous phase was followed by washing of
the pooled organic layers with brine until neutral. After drying (MgSO4), the solvents were evaporated to give a
yellow oil (52.9 mg) which was immediately diluted with methanol (2 ml). To the resulting solution, HNa2PO4 (52
mg) and 6% NaHg (170 mg) were added sequentially, the addition of both reagent (same amount) being repeated
twice every 2 hours. Water (10 ml) and hexane (10 ml) were then added and the resulting mixture was stirred for
0.5 hour. The aqueous phase was further extracted with ether (3Â2 ml), the pooled organic extracts then being
washed with brine, and dried (MgSO4). Evaporation of the solvents, followed by ®ltration of the residue on
Florisil (pentane) a€orded the sul®de 2b (13.1 mg, 70%, from 3a) as a colourless oil. 13C NMR: 10.4, 12.1, 14.8,
15.5, 36.3, 38.1, 40.4, 55.8, 61.2, 83.5, 84, 125.1, 125.9, 128.2, 128.9, 130.7, 135.2, 138.2, 134.6; UV (CH2Cl2): 260.8
(0.85), 271.3 (0.9), 282.4 (0.69) (in bracket: ee values).
9. Typical experimental procedure: To a cooled (ca. ^78ꢀC) solution of the chromone 6c in THF (1.5 ml/mmol) was
added, under an argon atmosphere, a 1 M solution of LDA (1.5 equiv.) in THF. After 1 hour stirring, the
electrophilic reagent (see Table 1 in the text, 1.5 equiv.) and HMPA (2 equiv.) were added sequentially. This
mixture was stirred at rt for 15 min, then diluted with ether (10 ml/mmol), and 10% NaHCO3 (10 ml/mmol). After
a further 2 hours stirring, the aqueous layer was extracted with ether (2Â10 ml/mmol) and the pooled organic
extracts were washed with brine (5Â10 ml/mmol), and dried (K2CO3). Evaporation of the solvents was followed
by chromatography of the residue on 60 silica gel (AcOEt) to a€ord the pure chromone.
10. Selected data: compound E-5: m.p. 65±67ꢀC (ether/hexane, 1.49 (d, J=7.14 Hz, 3H), 1.81 (d, J=1 Hz, 3H), 2.42
(s, 3H), 3.65 (q, J=7.14 Hz, 1H), 5.01±5.13 (m, 2H), 5.69 (dq, J=10.5, 1 Hz, 1H), 6.48 (td, J=16.5, 10.5 Hz, 1H),
7.29 (m, 2H), 7.67 (m, 2H); 13C NMR: 12, 14, 21.7, 60.7, 111.4, 128.5, 129.7, 133.7, 135.8, 144, 144.6; compound
6a: m.p. 162ꢀC (diisopropyl ether); compound 6c: m.p. 108ꢀC (hexane); compound 6d: 1H NMR: 0.13 (s, 6H), 1.02
(s, 9H), 1.25 (t, J=7.5 Hz, 3H), 1.95 (s, 3H), 2.6 (q, J=7.5 Hz, 2H), 3.87 (s, 3H), 3.89 (s, 3H), 6.35 (s, 1H); 13C
NMR: ^4.4, 9.5, 11.8, 18.7, 25.4, 25.8, 55.7, 56.9, 92.4, 108.5, 116.2, 126, 150.1, 153.5, 154, 163.6, 177.9; compound
6e: 1H NMR: 0.14 (s, 6H), 1.02 (s, 9H), 1.96 (s, 3H), 2.45 (dd, J=7.9, 6.9 Hz, 2H), 2.71 (t, J=6.9 Hz, 2H), 3.88 (s,
3H), 3.91 (s, 3H), 5.08 (m, 2H), 5.78 (m, 1H), 6.35 (s, 1H); 13C NMR: ^4.4, 9.7, 18.7, 25.6, 31.5, 31.6, 55.7, 56.7,
1
92.3, 108.5, 116.1, 117.1, 126, 136.5, 149, 153.5, 154, 161.5, 185.9; compound 6f: H NMR: 0.19 (s, 3H), 1.09 (s,
9H), 1.85 (s, 3H), 2.87±3 (m, 4H), 3.92 (s, 3H), 3.94 (s, 3H), 6.4 (s, 1H), 7.15±7.32 (m, 5H); 13C NMR: ^4.3, 9.7,
18.8, 25.9, 33.7, 34.3, 55.7, 56.7, 92.3, 108.3, 117.4, 126.5, 127.3, 128.4, 140.4, 149.6, 153.6, 154.1, 161.1, 191.2;
1
compound 6g: H NMR: 0.08 (s, 3H), 0.15 (s, 3H), 1.03 (s, 15H), 1.8 (m, 1H), 1.98 (s, 3H), 2.69 (m, 2H), 3.4 (s,
1H, OH), 3.85 (s, 3H), 3.86 (m, 2H), 3.87 (s, 3H), 6.18 (s, 1H); 13C NMR: ^4.5, 10.2, 18, 18.7, 19.1, 25.94, 34.2,
1
37.7, 55.1, 56, 74.1, 91.1, 107.5, 118.8, 127, 149.8, 152.9, 153.7, 160.8, 177.4. H and 13C NMR in CDCl3 at 200
and 50 MHz, respectively.