Studies Directed to the Total Synthesis of Cepacin A: Preliminary Communication

Article abstract of DOI:10.1002/hlca.200490062

The synthesis of the bromoallenyl-substituted epoxide derivative 34 related to cepacin A (1) starting from D-arabinose as a chiral pool is reported.

Full text of DOI:10.1002/hlca.200490062

Helvetica Chimica Acta Vol. 87 (2004)
667
Studies Directed to the Total Synthesis of Cepacin A
Preliminary Communication
by Chao-Jun Tang and Yikang Wu*
State Key Laboratory of Bio-organic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, 354 Fenglin Road, Shanghai 200032, China
(e-mail: yikangwu@mail.sioc.ac.cn)
The synthesis of the bromoallenyl-substituted epoxide derivative 34 related to cepacin A (1) starting from
d-arabinose as a chiral pool is reported.
Cepacin A [1] (1; see the Figure) is a potent antibacterial substance produced by
Pseudomonas cepacia, SC 11,783 (a Gram-negative rod, motile by means of multitrious
glagella) isolated from a soil sample of West Windsor, New Jersey, USA. The structure
of cepacin A was established by spectroscopic means and chemical degradations, with
the allene configuration assigned on the basis of the optical rotation according to the
rules of Lowe and Brewster [2].
Figure. The structure of natural cepacin A (1)
The structure of cepacin A (1) is completely different from those of the existing
antibacterial drugs. Therefore, the investigation on this type of compounds might
eventually lead to effective agents against the bacteria resistant to the existing drugs.
The combination of the architectural complexity and the potential prospect as a new
antibacterial agent make cepacin A (1) an attractive target for synthetic endeavors.
Disclosed below are some of our efforts directed towards the total synthesis of 1. Our
synthesis tries to make full use of all chiral C-atoms of d-arabinose to construct the
C(6)ÀC(10) moiety of cepacin A (1)1). The lactone part C(1)ÀC(5) was planned to be
derived from levulinic acid (ˆ4-oxopentanoic acid), while the dialkynyl group
1
)
Arbitrary numbering; the systematic name of cepacin A (1) is 5-{3-[3-(hepta-1,2-diene-4,6-diynyl)oxir-
anyl]-3-hydroxyprop-1-enyl}-4,5-dihydrofuran-2(3H)-one.

668
Helvetica Chimica Acta Vol. 87 (2004)
(C(13)ÀC(16)) should be introduced at a later stage via a coupling reaction with a
bromoallene.
As shown in Scheme 1, d-arabinose was converted to dithioacetal 3 in 72% overall
yield by known transformations [3]. The dithioacetal protecting group was hydrolyzed
with HgCl2 and HgO [4] (80% yield). The intermediate aldehyde was directly treated
with Ph3PˆCHCO(CH2)2CO2Me to give 4, the reagent being prepared from levulinic
acid via bromination with Br2 in MeOH [5] and subsequent treatment of the resulting
methyl 5-bromolevulinate with Ph3P in refluxing benzene followed by deprotonation
with aq. Na2CO3 solution [6]. Reduction of the ketone carbonyl group of 4 with NaBH4
in MeOH at 08 afforded alcohol 5 (diastereoisomer mixture) in 80% yield.
Scheme 1
a) 1. EtSH, 6n HCl; 78%; 2. Me2CO, conc. H2SO4; 92%. b) 1. HgCl2, HgO (red), MeCN/H2O; 80%; 2.
Ph3PˆCHCO(CH2)2CO2Me, toluene, reflux; 91%. c) NaBH4, CeCl3 ¥ 7 H2O, MeOH; 80%.
Treatment of 5 with catalytic amounts of TsOH in anhydrous acetone under reflux
led to the formation of the lactone ring (Scheme 2). Partial hydrolysis of 6 under the
conditions of Xiao and Bai [7] gave the desired diol 7 (mixture of diatereoisomers with
respect to C(4)) in 40% yield. Later, we found that if MeOH was used as solvent, the
lactonization and selective hydrolysis of the terminal acetonide moiety could be
realized in a one-pot manner in ca. 64% yield. The diol moiety of 7 was protected as
(tert-butyl)dimethylsilyl (tBuMe2Si) ether with the hope that selective oxidation of the
primary silyl ether group of 8 with quinolinium fluorochromate (QFC) as reported by
Murugensan and Pandurangan [8] would afford the desired aldehyde 9. Unfortunately,
this reagent did not work in our case. A longer reaction sequence was then attempted.
First, the primary OH group of 7 was masked as pivaloyl (Piv) ester ( ! 10).
Subsequent protection of the secondary OH group as a silyl ether ( ! 11), however,
suffered seriously from low yields. These results incited us to search for better
alternatives.
The route shown in Scheme 3 appeared to work nicely all the way to aldehyde 9. The
treatment with CeCl3 ¥ 7 H2O and (CO2H)2 in MeCN [7] was much more effective on 3
than on 6, giving the corresponding diol in 92% yield. The primary and secondary OH
groups were readily protected as Piv ester ( ! 12) and silyl ether, respectively, yielding
the fully protected intermediate 13. The subsequent HgCl2/HgO-mediated hydrolysis
smoothly cleaved the dithioacetal to give the corresponding aldehyde. Further
treatment with the Wittig reagent used before afforded ketone 14 in 94% yield.
Reduction of 14 with NaBH4 in MeOH at 08 yielded alcohol 15 as a mixture of two
epimers, which (unlike 5) could be separated from each other after repeated
chromatography (SiO2, CH2Cl2/AcOEt). One of the epimers of 15 (of unknown
configuration at C(4)1)) was stirred in refluxing benzene in the presence of traces of

Helvetica Chimica Acta Vol. 87 (2004)
669
Scheme 2
a) TsOH, Me2CO, reflux; 70%. b) CeCl3 ¥ 7 H2O, (CO2H)2, MeCN; 40%. c) TsOH, MeOH, reflux; 64%. d)
tBuMe2SiCl, 1H-imidazole, N,N-dimethylpyridin-4-amine (DMAP), DMF; 54%. e) Quinolinium fluorochro-
mate (QFC). f) PivCl, Et3N, CH2Cl2; 58%; g) tBuMe2SiCl, 1H-imidazole, DMF; 52%.
pyridinum p-toluenesulfonate (PPTS) [9] leading to lactone 16 in quantitative yield.
With the lactone moiety satisfactorily established, we directed our attention again to
the other terminus of the C-chain. Cleavage of the Piv ester was rather difficult. Base-
catalyzed hydrolysis (NaOMe or tBuOK in the corresponding alcohol) failed to result
in a clean reaction. Finally, the Piv protecting group was removed by a diisobutyl-
aluminium hydride (DIBAL-H) reduction, with concurrent partial reduction of the
lactone to a hemiacetal. Oxidation of the hemiacetal back to the lactone and of the
primary OH group into an aldehyde was realized by pyridinium chlorochromate (PCC)
oxidation ( ! 9). The yield of this step was, however, only 50%. Subsequent addition of
the acetylide to aldehyde 9 to give 17 did not proceed as well either. It appeared to us
that the sluggish reaction was caused by the serious steric crowding associated with the
tBuMe2Si group.
To reduce the steric hindrance near the aldehyde group, we next chose to mask the
secondary OH group of 12 as a triethylsilyl (Et3Si) ether ( ! 18; Scheme 4). With the

670
Helvetica Chimica Acta Vol. 87 (2004)
Scheme 3
a) 1. CeCl3 ¥ 7 H2O, (CO2H)2, MeCN; 92%; 2. PivCl, Et3N, CH2Cl2; 77%. b) tBuMe2SiCl, 1H-imidazole, DMF;
95%. c) 1. HgCl2, HgO (yellow), MeCN/H2O; 2. Ph3PˆCHCO(CH2)2CO2Me, toluene, reflux; 94% from 13. d)
NaBH4, MeOH; 71%. e) PPTS, PhH, reflux; 100%. f) 1. DIBAL-H, CH2Cl2, À788; 95%; 2. PCC, NaOAc,
CH2Cl2; 50%. g) Me3SiCꢀCLi, THF, À788; 40%.
hope to avoid the undesired lactone reduction encountered with 16 during the Piv
removal, we decided to cleave the Piv protection group at an earlier stage. Thus,
reaction of 18 with DIBAL-H at À788 led to 19 in 91% yield. By using SO3 ¥ Py [10]
(which was much superior to other oxidants when substrates contained low-valent S-
atoms), the primary alcohol 19 was readily oxidized to give aldehyde 20. With the
tBuMe2Si group replaced by the smaller Et3Si group, the addition of the acetylide now
indeed proceeded smoothly, giving 21 in ca. 81% yield as a mixture of two epimers (21a/
21b 1.4 :1, separable by chromatography (SiO2)). One of the isomers (later shown to be
21b with (S)-configuration2) at the new stereogenic center) was converted into 22 with
the intention to test later whether acetate could also serve as a leaving group in the
bromoallene-formation step. The allene configuration was controlled by the pro-
pargylic OH group. Because much work had been done by the time when the
configuration of the isomer that we picked at random was established to be −wrong×,
and the allene isomer of cepacin was also one of the desired analogues for biological
testing, we went on with 21b.
Hydrolysis of the dithioacetal moiety of 22 was achieved most satisfactorily with I2
and NaHCO3 in acetone/H2O at 08 [11] ( ! 23). The lactone moiety was then
introduced via 24 and 25 by the same methodology as employed before. To simplify the
spectra of the intermediates, we tried to use the CBS reagent (ˆ(S)-B-methylox-
azaborolidine/borane ˆ (3aS)-tetrahydro-1-methyl-3,3-diphenyl-1H,3H-pyrrolo[1,2-c]-
[1,3,2]oxazaborole/borane; CBS ˆ CoreyÀBakshiÀShibata) [12] instead of NaBH4 to
stereoselectively reduce the ketone carbonyl group of 24, although this was not
2
)
The configuration was established by a NOESY experiment with a sample prepared from 21 by removal of
the Et3Si group and protecting the two OH groups as an acetonide.

Helvetica Chimica Acta Vol. 87 (2004)
671
Scheme 4
i
a) Et3SiCl, 1H-imidazole, DMF, r.t.; 90%. b) DIBAL-H, CH2Cl2, À788; 91%. c) SO3 ¥ Py, Pr2NEt, DMSO/
CH2Cl2 1 :1, 08 to r.t.; 82%. d) Me3SiCꢀCLi, THF, À788; 81%. e) Ac2O, Et3N, CH2Cl2, r.t.; 82%. f) I2, NaHCO3,
Me2CO/H2O 5 :1, 08; 79%. g) Ph3PˆCHCO(CH2)2CO2Me, toluene, reflux; 83%. h) BH3/(S)-B-methylox-
azaborolidine [12], 08; 61%. i) PPTS, PhH, 408; 59%.
necessary, the natural cepacin A (1) being a 1:1 epimer mixture with respect to this new
stereogenic center. A single isomer 25 of undetermined configuration was obtained in
61% yield (along with some unreduced 24). Unfortunately, the PPTS-catalyzed
lactonization of 25 to 26 occurred in only moderate yield (59%) (cf. 5 ! 6 and 15 !
16). But by that time, our synthesis along another route (Scheme 5) carried out in
parallel to that shown in Scheme 4 proceeded very well and, therefore, no further
attempts were made to carry on with acetate 26.
This other route started from 21b (Scheme 5). The dithioacetal moiety of 21b was
hydrolyzed with I2 and NaHCO3 as in the previous route, the formed aldehyde 27
immediately treated with Ph3PˆCHCO(CH2)2CO2Me, and the propargylic OH group
converted to the corresponding tosylate ( ! 28) to ensure higher reactivity in the
subsequent bromoallene-formation step. The Me3Si group at the acetylene terminus of
28 was removed with K2CO3 in MeOH/THF 2 :1 at 08 (92%). Further treatment of
propargyl tosylate 29 with anhydrous LiBr in the presence of CuBr in refluxing THF
under N2 according to the procedure reported by Mann et al. [13] afforded bromoallene
derivative 30 in 61% yield, along with 27% of unreacted 29. The TES protecting group
was then cleaved (97%) yield, the resulting OH group was transformed into the tosylate
(72%), and the ketone carbonyl group was reduced with borane/(S)-B-methyloxaza-
borolidine [12] to afford the hydroxy derivative 31 (of undetermined configuration;
77%). On treatment with PPTS, 31 was readily converted into lactone 32 (85%), whose

672
Helvetica Chimica Acta Vol. 87 (2004)
acetonide moiety was then removed by the reaction with propane-1,3-dithiol in CH2Cl2
with BF3 ¥ OEt2 as catalyst (69%). Finally, diol 33 was treated with K2CO3 in Et2O/
MeOH 50 :1 to generate the bromoallenyl-substituted epoxide derivative 343) in 77%
yield.
Scheme 5
a) I2, NaHCO3, Me2CO/H2O 5 :1, 08. b) Ph3PˆCHCO(CH2)2CO2Me, toluene, 70 808; 73% from 21b. c) TsCl,
Et3N, DMAP, CH2Cl2, 08 to r.t.; 74%. d) K2CO3, MeOH/THF 2 :1, 08; 92%. e) LiBr, CuBr¥ Me2S, THF, reflux;
61% (84% based on consumed 29). f) Bu4NF, THF; 97%. g) TsCl, Et3N, DMAP, CH2Cl2, 08 to r.t.; 72%. h)
BH3 ¥ SMe2/(S)-B-methyloxazaborolidine [12] 08; 77%. i) PPTS, PhH, 40 508; 85%. j) HS(CH2)3SH, BF3 ¥
OEt2, CH2Cl2; 69%. k) K2CO3, Et2O/H2O 50 :1; 77%.
Financial support from the NSF of China (No. 20025207, 20272071, 20321202, and 20372075), the CAS
(Knowledge Innovation Project, KGCX2-SW-209), the Major State Basic Research Development Program
(G2000077502), and Bayer AG is gratefully acknowledged. The authors also thank Dr. J¸rgen Scherkenbeck for
helpful discussions and some literature.
3
)
Data of 34: [a]2D0 ˆ À123.5( c ˆ 0.6, CHCl3). FT-IR (film): 3438, 1957, 1770, 1183 cmÀ1. 1H-NMR (300 MHz,
CDCl3; d in ppm, J in Hz)1): 6.25( d, J ˆ 5.7, HÀC(12)); 6.00 5.89 (m, HÀC(5), HÀC(6)); 5.16 (dd, J ˆ
7.9, 5.6, HÀC(10)); 5.03 (m, HÀC(4)); 4.26 (m, HÀC(7)); 3.55 (dd, J ˆ 8.1, 2.0, HÀC(9)); 3.07
(m, HÀC(8)); 2.62 2.54 (m, CH2(2)); 2.48 (m, 1 HÀC(3)); 2.10 (d, J ˆ 6.9, OH); 2.05( m, 1 HÀC(3)).
13C-NMR (75MHz, CDCl 3; d in ppm)1): 204.65(C(11)); 176.80 (C(1)); 130.97 (C(5) or C(6)); 129.95(C(6)
or C(5)); 98.11 (C(10)); 79.49 (C(4)); 74.86 (C(12)); 69.73 (C(7)); 61.58 (C(8)); 52.00 (C(9)); 28.40 (C(2));
28.24 (C(3)); assignments by COSY and HMQC experiments. EI-MS (m/z (%)): 161 (5.1, C(12) to C(8)
fragment), 159 (5, C(12) to C(8) fragment), 142 (6), 141 (7, C(7) to C(1) fragment), 123 (10), 113 (9), 95
‡
‡
(30), 81 (100). HR-ESI-MS: 322.9879 ([M ‡ Na] , C12H13O4BrNa ; calc. 322.9889).

Helvetica Chimica Acta Vol. 87 (2004)
673
REFERENCES
[1] W. L. Parker, M. L. Rathnum, V. Seiner, W. H. Trejo, P. A. Principe, R. B. Sykes, J. Antibiot. 1984, 37, 431.
[2] G. Lowe, J. Chem. Soc., Chem. Commun. 1965, 411; J. H. Brewster, in −Topics in Stereochemistry×, Vol. 2,
Eds. N. L. Allinger and E. L. Eliel, John Wiley & Sons, Inc., New York, 1967, p. 1.
[3] J. D. White, M. S. Jensen, J. Am. Chem. Soc. 1995, 117, 6224.
[4] A. Dondoni, A. Marra, P. Merino, J. Am. Chem. Soc. 1994, 116, 3324.
[5] S. F. MacDonald, Can. J. Chem. 1974, 52, 3257.
[6] R. C. Ronald, C. J. Wheeler, J. Org. Chem. 1983, 48, 139.
[7] X. Xiao, D. Bai, Synlett 2001, 535.
[8] V. Murugensan, A. Pandurangan, Ind. J. Chem., Sect. B 1992, 31, 377.
[9] M. Miyashita, A. Yoshikoshi, P. A. Grieco, J. Org. Chem. 1977, 42, 3772.
[10] L. F. Fieser, −Experiments in Organic Chemistry×, 3rd edn., D. C. Heath & Co., 1955, p. 337; H. H. Sisler,
L. F. Audrieth, Inorg. Synth. 1946, 2, 173.
[11] K. C. Nicolaou, M. E. Bunnage, D. G. McGarry, S. Shi, P. K. Somers, P. A. Wallace, X.-J. Chu, K. A. Agrios,
J. L. Gunzner, Z. Yang, Chem. Eur. J. 1999, 5, 599; G. A. Russell, L. A. Ochrymowy, J. Org. Chem. 1969, 34,
3618.
[12] E. J. Corey, R. K. Bakshi, S. Shibata, J. Am. Chem. Soc. 1987, 109, 551; V. K. Singh, Synthesis 1992, 605.
[13] F. D×Aniello, A. Mann, M. Taddei, C.-G. Wermuth, Tetrahedron Lett. 1994, 35, 7775; F. D×Aniello, A.
Schoenfelder, A. Mann, M. Taddei, J. Org. Chem. 1996, 61, 9631.
Received September 19, 2003