Total synthesis of (-)-ulapualide A: The danger of overdependence on NMR spectroscopy in assignment of stereochemistry

Article abstract of DOI:10.1002/anie.200700459

Lessons learnt: The asymmetric total synthesis of the macrolide (-)-ulapualide A has been accomplished. Interestingly, the ¹H NMR spectrum and chiroptical data of the macrolide and of a previously synthesized diastereoisomer with opposite stere

Full text of DOI:10.1002/anie.200700459

Angewandte
Chemie
DOI: 10.1002/anie.200700459
Macrolide Synthesis
Total Synthesis of (À)-Ulapualide A: The Danger of Overdependence
on NMR Spectroscopy in Assignment of Stereochemistry**
Gerald Pattenden,* Neil J. Ashweek, Charles A. G. Baker-Glenn, Gary M. Walker, and
James G. K. Yee
Ulapualide A (1) and its relatives, for example, mycalolide A
(2), are an intriguing family of trisoxazole-based macrolides
which have been isolated from marine nudibranchs (sea slugs)
and sponges.[1] The metabolites show antifungal activity and
ichthyotoxic properties, and they also show activity against
L1210 leukemia cells. Although the ulapualides have been
known since 1986, their complete stereochemistries were not
defined until quite recently following degradation and
synthetic studies,[2] and, in particular, X-ray studies of some
of their complexes with the protein actin.[3]
Earlier, we described a total synthesis of the structure 3
for ulapualide A, whose stereochemistry we assigned on the
basis of a molecular mechanics study of a hypothetical metal-
chelated complex.[1b] The synthetic ulapualide A showed
identical chromatographic behavior, as well as chiroptical
and 1H NMR spectroscopic data with those of naturally
derived material, but very small differences were observed in
the 13C NMR spectra associated with the C32–C34 portions of
the structures. The stereochemistry of natural ulapualide A
(1), determined from an X-ray analysis of its complex with
actin, differs from that of the diastereoisomer 3 at the
stereocenters C3, C28, C29, C30, and C32. This situation
makes it all the more bewildering, therefore, as to why the
1
synthetic diastereoisomer 3 gave H NMR spectroscopic and
other data which were superimposable on those of the natural
product. We have, therefore, undertaken a conceptually new,
second generation, total synthesis of ulapualide A[4] to resolve
any remaining ambiguities with regards to the stereochemis-
try of the free metabolite, which was isolated from the egg
masses of the nudibranch Hexabranchus sanguineus.[5]
Based on speculation of the biosynthetic origin of the
contiguous trisoxazole unit in the ulapualides 1 and 2,[6]
alongside those co-metabolites that contain only two oxazole
rings, for example, 4,[7] we designed a synthesis of 1 that
involved the macrolactamization of the intermediate 6,
leading to 5, followed by elaboration of the central oxazole
ring in 1 as a late step in the overall synthesis.[8] This approach
required a synthesis of the ester 6 from the acid 8 and the
alcohol 7, which together contain all of the ten stereocenters
in ulapualide A (1; Scheme 1).
[*] Prof. G. Pattenden, Dr. N. J. Ashweek, Dr. C. A. G. Baker-Glenn,
Dr. G. M. Walker, Dr. J. G. K. Yee
School of Chemistry
University of Nottingham
Nottingham, NG72RD (UK)
Fax: (+44)115-951-3535
E-mail: gp@nottingham.ac.uk
[**] We thank J. Kempson for his contributions to the synthesis of
intermediate 11. We also thank Astra Zeneca and Pfizer Ltd for
financial support towards this research program.
In our synthesis of the diastereoisomer 3 of ulapualide A
we applied a range of contemporary methods in asymmetric
Supporting information for thisarticle isavailable on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 4359 –4363
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4359

Communications
synthesis and examined a variety of hydroxy-protecting
groups to prepare the C26–C41(top-chain) portion in the
target.[1] We therefore followed this overall strategy in
preparing the functionalized C26–C41unit 13, with the
modified stereochemistry between C28 and C32, from the
aldehyde 9 and the phosphonate 10 en route to the alcohol 7.
Thus, the aldehyde 9 was made by using Evans aldol
À
chemistry to introduce the syn stereochemistry at C32 C33
À
and the anti stereochemistry at C29 C30, and we applied the
Brown allylboration chemistry to install the b-orientated
hydroxy group at C28.[9] The absolute stereochemistry of the
aldehyde 9 was established by X-ray analysis of the TBDPS
ether of the corresponding primary alcohol.[10]
A Wadsworth–Emmons reaction between the aldehyde 9
and the known phosphonate ester 10[1] using Ba(OH)2 as
base[11] gave the E alkene 11, which, in one step, was then
converted into the alcohol 12a following hydrogenation/
hydrogenolysis in the presence of the Pearlman catalyst[12]
(Scheme 2). The primary alcohol group in 12a was protected
as its TBDPS ether 12b and the dimethyl acetal group was
then unmasked selectively using Me2BBr in CH2Cl2 at
À788C[13] to give the aldehyde 13 in excellent yield. A
straightforward Wittig reaction between the aldehyde 13 and
the phosphorus ylide, derived in situ from the oxazolemethyl
bromide 14[14] and Bu3P in the presence of DBU, next gave the
E alkene 15a. Selective deprotection of the MOM ether
group in 15a using Me2BBr in Et2O at À788C then produced
the alcohol 15b (cf. 7) in preparation for coupling to the
carboxylic acid 8.
The protected form of the substituted carboxylic acid 8,
that is, 24b, which contains the two remaining stereocenters in
1, was conveniently synthesized using a cobalt-mediated
coupling reaction between the aldehyde 21 and the iodide 22
in the presence of vitamin B12.[15] Thus, a chelation-controlled
syn addition of lithium dimethylcuprate to the R a,b-unsatu-
rated ester 16[16] derived from the Garner aldehyde, first gave
the adduct 17 with 9:1 syn selectivity (Scheme 3). Successive
Scheme 1. Retrosynthesis of ulapualide A (1). TBS=tert-butyldimethyl-
silyl.
Scheme 2. Synthesis of the side chain 15b. Reagentsand conditions: a) Ba(OH) 2, THF, H2O, 68%; b) Pd(OH)2/C, H2, EtOAc, 96%; c) TBDPSCl,
imidazole, DMF, 93%; d) Me2BBr, CH2Cl2, À788C, 98%; e) PBu3, 14, DBU, DMF, 75%; f) Me2BBr, Et2O, À788C, 79%. Bn=benzyl, DBU=1,8-
diazabicyclo[5.4.0]undec-7-ene, DMF=N,N-dimethylformamide, MOM=methoxymethyl, TBDPS=tert-butyldiphenylsilyl.
4360 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4359 –4363

Angewandte
Chemie
functional group interconversions next led to the amino
alcohol 18 which was then reacted with the Garner acid,[17] to
give the corresponding amide 19. Oxidation of the primary
alcohol group in 19 followed by cyclization of the resulting
aldehyde[18] gave the substituted oxazole 20, which was then
elaborated to the new aldehyde 21 in two straightforward
steps.
then be converted into the methoxyoxazoline 26b by reaction
with NBS in MeOH followed by dehydrobromination in the
presence of Cs2CO3.[24] Finally, treatment of 26b with CSA in
benzene heated at reflux gave the same trisoxazole 27 as that
obtained earlier.
The synthesis of ulapualide A (1) from the substituted
macrolide intermediate 27 was completed in six relatively
straightforward steps, which were made possible by the
selective ease of removal of the silyl protecting groups in
27, that is, TBS > primary TBDPS > secondary TBDPS.[25]
Thus, deprotection of the TBS group in 27, using TMSOTf
at À788C followed by acetylation of the resulting secondary
alcohol 29a first gave the corresponding acetate 29b
(Scheme 5). Treatment of 29b with HF/pyridine in CH2Cl2
The aldehyde 21 was then coupled to the iodide 22[19] in
degassed DMF in the presence of vitamin B12 and CrCl2,[15] to
give a 1:1 mixture of epimers of the alcohol 23 in 88% yield.
Finally, oxidation of 23 to the corresponding ketone 24a
under Swern conditions followed by saponification of the
ester group in 24a using LiOH/MeOH gave the carboxylic
acid 24b.
À
The esterification of the carboxylic acid 24a with the
alcohol 15b proceeded smoothly under Yamaguchi condi-
tions[20] to give the coupled product 25 in an excellent 93%
yield (Scheme 4). The tert-butyl ester and the N-Boc protect-
ing groups in 25 were removed simultaneously using TMSOTf
in the presence of Et3N[21] to give the w-amino acid 6 (R =
TBDPS; Scheme 4). The amino acid 6 was not isolated but
instead was treated with HATU[22] at 08C which resulted in
smooth macrolactamization to 5 (R = TBDPS) in an agree-
able yield of 67% for the two steps. The conversion of the
macrolactam 5 into the corresponding trisoxazole 27 did not
turn out to be straightforward.[8] Although the intermediate 5
(R = TBDPS) was smoothly dehydrated to the oxazoline 26a
using DASTat À788C, the subsequent oxidation of 26a to the
corresponding trisoxazole 27 using NiO2 in benzene heated at
reflux[23] was unreliable and seldom gave yields greater than
25%. A more reliable, albeit more lengthy route to 27 from 5,
was via the enamide 28 that was produced after mesylation
and elimination of MeSO2H from 5. The enamide 28 could
for 1.5 h resulted in the selective deprotection of the
CH2OTBDPS group to give the primary alcohol 30a, which
was then oxidized to the corresponding aldehyde 30b using
Dess–Martin periodinane. When a solution of the aldehyde
30b in benzene was heated under reflux with N-methylform-
amide in the presence of pyridinium p-toluenesulfonate
(PPTS) for 10 h, workup and chromatography gave (E)-
alkenylformamide 31 which was isolated as a 3:2 mixture of
rotamers in 25–40% yield. Removal of the TBDPS protection
in 31 using HF/pyridine in THF at room temperature for 12 h,
finally gave (À)-ulapualide A (1), [a]2D6 = À52.8 (c = 0.11 in
MeOH) as a colorless, viscous oil, in 60% yield.
The synthetic ulapualide A displayed H and 13C NMR
1
spectroscopic data which were indistinguishable from those
obtained for naturally derived material.[26] Significantly,
whereas the synthetic diastereoisomer 3 displayed small
differences in the chemical shifts of the 13C NMR spectra
with the natural product for the C32 atom (d = 81.0 ppm; cf.
d = 81.8 ppm for natural), the C34 atom (d = 26.6 ppm; cf. d =
Scheme 3. Synthesis of the carboxylic acid 24b. Reagentsand conditions: a) Me 2CuLi, TMSCl, THF, 81%; b) DIBAL-H, THF, 100%; c) NaH, BnBr,
TBAI, THF, 78%; d) HCl, dioxane; e) Garner acid, EDC, 4-methylmorpholine, THF, HOBt, 79%; f) DMSO, (COCl)2, Et3N, CH2Cl2, À788C, 97%;
g) PPh3, 1,2-dibromotetrachloroethane, 2,6-di-tert-butylpyridine, CH2Cl2, DBU, MeCN, 72%; h) Pd/C, EtOAc, H2, 81%; i) DMSO, (COCl)2, Et3N,
CH2Cl2, À788C, 87%; j) Vitamin B12, CrCl2, DMF, 88%; k) DMSO, (COCl)2, Et3N, CH2Cl2, À788C, 85%; l) LiOH, MeOH, H2O, 88%. Boc=tert-
butoxycarbonyl, DIBAL-H=diisobutylaluminum hydride, DMSO=dimethyl sulfoxide, EDC=1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydro-
chloride, HOBt=1-hydroxy-lH-benzotriazole, TBAI=tetra-n-butylammonium iodide, TMS=trimethylsilyl.
Angew. Chem. Int. Ed. 2007, 46, 4359 –4363
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4361

Communications
Scheme 4. Synthesis of the trisoxazole 27. Reagentsand conditions: a) 2,4,6-trichlorobenzoyl chloride, Et 3N, toluene, DMAP, 08C, 93%;
b) TMSOTf, Et3N, CH2Cl2, 08C; c) HATU, Et3N, CH2Cl2, 08C, 67%; d) DAST, CH2Cl2, À788C, 89%; e) NiO2, benzene, reflux, 25%; f) MsCl, DIPEA,
CH2Cl2, DBU, 08C, 75%; g) NBS, MeOH, CH2Cl2, 92%; h) Cs2CO3, dioxane, 608C, 92%; i) CSA, benzene, 5-Š molecular sieves, reflux, 58%.
CSA=camphorsulfonic acid, DAST=(diethylamino)sulfur trifluoride, DIPEA=diisopropylethylamine, DMAP=4-dimethylaminopyridine,
HATU=N-[(dimethylamino)-1H-1,2,3-triazole[4,5-b]-pyridin-l-ylmethylene]-N-methylmethanaminium hexafluorophosphate, Ms=methanesulfonyl,
NBS=N-bromosuccinimide, Tf=trifluoromethanesulfonyl.
27.6 ppm for natural), and for the C33-methyl group (d =
14.2 ppm; cf. d = 15.5 ppm for natural), the synthetic ulapua-
lide A (1) showed matching chemical shift data for these
absorptions, that is, d = 81.8 (C32), 27.7 (C34), and 15.6 ppm
[2] a) S. Matsunaga, C. A. Celatka, N. Fusetani, P. Liu, J. S. Panek, J.
[1] For leading refereces, see a) S. K. Chattopadhyay, G. Pattenden,
Tetrahedron Lett. 1998, 39, 6095 – 6098; b) S. K. Chattopadhyay,
G. Pattenden, J. Chem. Soc. Perkin Trans. 1 2000, 2429 – 2454.
(C33-Me). This total synthesis of (À)-ulapualide A comple-
Am. Chem. Soc. 1999, 121, 5605 – 5606; b) P. Liu, J. S. Panek, J.
Am. Chem. Soc. 2000, 122, 1235 – 1236; c) P. Liu, J. S. Panek, J.
ments the X-ray studies made by Rayment and co-workers[3b]
Am. Chem. Soc. 2000, 122, 11090 – 11097.
on an actin–ulapualide A complex. It also emphasizes how
cautious synthetic chemists should be in relying on NMR
spectroscopy when comparing data for some complex natural
products and their synthetic counterparts. As this study has
shown, profound changes in the stereochemistry of a natural
product (compare 1 and 3) are not always reflected in a
significant way in their NMR spectroscopic data.
[3] a) J. S. Allingham, J. Tanaka, R. King V.A. Klenchin, I. Ray-
ment, G. Marriott, Nat. Struct. Biol. 2003, 10, 1058 – 1063; b) J. S.
Allingham, J. Tanaka, G. Marriott, I. Rayment, Org. Lett. 2004,
6, 597 – 599.
[4] For a synthesis of mycalolide A (2), see Ref. [2].
[5] For a review on some incorrectly assigned structures of natural
products, see: K. C. Nicolaou, S. A. Synder, Angew. Chem. 2005,
117, 1036 – 1069; Angew. Chem. Int. Ed. 2005, 44, 1012 – 1044;
diazonamide A is a particularly pertinent example of where X-
ray analysis has been misleading; see: J. Li, A. W. G. Burgett, L.
Esser, C. Amezcara, P. G. Harran, Angew. Chem. 2001, 113,
4905 – 4909; Angew. Chem. Int. Ed. 2001, 40, 4770 – 4773.
[6] G. M. Walker, PhD Thesis, University of Nottingham (UK),
2006.
Received: February 1, 2007
Published online: April 30, 2007
Keywords: macrocycles· natural products· oxazoles·
.
structure elucidation · total synthesis
4362 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4359 –4363

Angewandte
Chemie
Scheme 5. Completion of the synthesis of 1. Reagentsand conditions: a) TMSOTf, CH 2Cl2, À788C, 85%; b) acetic anhydride, DMAP, CH2Cl2,
pyridine, 84%; c) HF/pyridine, CH2Cl2, pyridine, 61%; d) Dess–Martin periodinane, CH2Cl2, 80%; e) N-methylformamide, PPTS, benzene, reflux,
40%; f) HF/pyridine, THF, 60%.
[7] a) M. R. Kernan, T. F. Molinski, D. J. Faulkner, J. Org. Chem.
1988, 53, 5014 – 5020; b) J. Kobayashi, M. Tsuda, H. Fuse, T.
Sasaki, Y. Mikami, J. Nat. Prod. 1997, 60, 150 – 154; c) J. Shin, H.-
S. Lee, J.-Y. Kim, H. J Shin, J.-W. Ahn, V. J. Paul, J. Nat. Prod.
2004, 67, 1889 – 1892.
[8] For a model study of this strategy, see: S. K. Chattopadhyay, J.
Kempson, A. McNeil, G. Pattenden, M. Reader, D. E. Rippon,
D. Waite, J. Chem. Soc. Perkin Trans. 1 2000, 2415 – 2428.
[9] For pertinent details see Ref [1].
[10] We thank Dr A. J. Blake for this X-ray analysis.
[11] a) C. Alvarez-Ibarra, S. Arias, G. Baæón, M. J. Fernµndez, M.
Rodriguez, V. Sinisterra, J. Chem. Soc. Chem. Commun. 1987,
1509 – 1511; b) I. Paterson, K. S. Yeung, J. B. Smaill, Synlett 1993,
774 – 776.
[12] W. M. Pearlman, Tetrahedron Lett. 1967, 8, 1663 – 1664.
[13] Y. Guindon, C. Yoakim, H. E. Morton, J. Org. Chem. 1984, 49,
3912 – 3920.
[14] The oxazole methyl bromide 14 was obtained from (S)-tert-butyl
2-(2-(benzyloxy)acetamido)hydroxypropanoate by cyclization
and oxidation using DAST, bromotrichloromethane, and DBU,
followed by manipulation of the functionality in the resulting
tert-butyl 2-(benzyloxy)methyl)oxazole-4-carboxylate.
[15] K. Takai, K. Nitta, O. Fujimura, K. Utimoto, J. Org. Chem. 1989,
54, 4732 – 4734.
[16] a) H. Yoda, T. Shirai, T. Katagiri, K. Takabe, K. Kimata, K.
Hosoya, Chem. Lett. 1990, 2037 – 2038; b) I. Jako, P. Uiber, A.
Mann, M. Taddei, C. Wermuth, Tetrahedron Lett. 1990, 31, 1011 –
1014; c) S. Hanessian, K. Sumi, Synthesis 1991, 1083 – 1089.
[17] a) P. Garner, J. M. Park, J. Org. Chem. 1987, 52, 2361– 2364;
b) A. McKillop, R. J. K. Taylor, R. J. Watson, N. Lewis, Synthesis
1994, 31– 33; c) A. D. Campbell, T. J. Raynham, R. J. K. Taylor,
Synthesis 1998, 1707 – 1709.
[18] a) R. Robinson, J. Chem. Soc. 1909, 95, 2167 – 2174; b) S.
Gabriel, Ber. Dtsch. Chem. Ges. 1910, 43, 1283 – 1287; c) P.
Wipf, S. Lim, J. Am. Chem. Soc. 1995, 117, 558 – 559.
[19] The S iodide 22 was obtained from the TBS ether of methyl 6-
hydroxy-3-oxohexanoate by asymmetric reduction using (S)-Ru-
binap (binap = 2,2’-bis(diphenylphosphanyl)-1,1’-binaphthyl);
see: a) M. Kitamura, T. Ohkuma, H. Takaya, R. Noyori, Org.
Synth. 1992, 71, 1– 13; b) H. Takaya, R. Noyori, Acc. Chem. Res.
1990, 23, 345 – 350, followed by manipulation of the functionality
in the resulting (S)-3-hydroxy hexanoate (> 98.5% ee).
[20] J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull.
Chem. Soc. Jpn. 1979, 52, 1989 – 1993.
[21] a) Y. Hamada, S. Kato, T. Shioiri, Tetrahedron Lett. 1985, 26,
3223 – 3226; b) M. Sakaitani, Y. Ohfune, Tetrahedron Lett. 1985,
26, 5543 – 5546; c) M. Sakaitani, Y. Ohfune, J. Org. Chem. 1990,
55, 870 – 876.
[22] L. A. Carpino, J. Am. Chem. Soc. 1993, 115, 4397 – 4398.
[23] A. J. Meyers, D. L. Evans, D. K. Minster, U. Jordis, S. M. Hecht,
A. L. Muzzu, J. Org. Chem. 1979, 44, 497 – 501.
[24] See Ref. [8] and also: N. Endoh, K. Tsuboi, R. Kim, Y.
Yonezawa, C. Shin, Heterocycles 2003, 60, 1567 – 1572.
[25] For the related steps in our “first-generation” synthesis of
structure 3 of ulapualide A, see Ref. [1].
[26] J. A. Roesener, P. J. Sheuer, J. Am. Chem. Soc. 1986, 108, 846 –
847.
Angew. Chem. Int. Ed. 2007, 46, 4359 –4363
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4363