Total synthesis and structural revision of laurefurenynes A and B

Article abstract of DOI:10.1002/chem.201302352

Structural reassignment: A total synthesis of the proposed structure of (-)-laurefurenynea A has been accomplished that relies on organocatalytic aldehyde α-chlorination and a flexible chlorohydrin-based strategy for stereocontrolled access to the bis-tetrahydrofuran core of the natural product. Analysis of incongruities between the ¹Ha NMR spectra of synthetic and natural material led to a configurational reassignment for the natural product, which was also confirmed by total synthesis. Copyright

Full text of DOI:10.1002/chem.201302352

COMMUNICATION
DOI: 10.1002/chem.201302352
Total Synthesis and Structural Revision of Laurefurenynes A and B
Michael T. Holmes and Robert Britton*[a]
The laurefurenynes A (1) and B (2)[1] (Figure 1) are mem-
bers of a family of 2,2’-bis-tetrahydrofuran-containing natu-
ral products isolated from various Laurencia sp. red algae.
This family of oxylipids also includes elatenyne (3),[2] the
lauroxolane 4,[3] laurendecumenyne B (5)[4] and the related
diastereomer notoryne[5] (not shown). Originally, elatenyne
(3) was structurally assigned as the pyranoACTHNUTRGNEUNG
[3,2-b]pyran 6[2]
based on spectral similarities with (Z)-dactomelyne (7).[6]
However, synthesis of 6[7] and analysis of calculated
13C NMR chemical shifts for a series of diastereomeric 2,2’-
bis-tetrahydrofurans[8] led to a revised structure 3, which
was recently confirmed by Burton and Kim through total
synthesis.[9] Biogenetically, the dihalo-bis-tetrahydrofurans
3–5 are proposed to derive from laurediol (8) through a
series of bromonium ion-induced cyclizations followed by
structural rearrangement (e.g., 10!3).[5,9,10] Considering
that the laurefurenynes A (1) and B (2) were isolated along
with the bromofucin (10) analogues 12 and 13, these com-
pounds may share a similar biosynthesis to elatenyne, raising
questions regarding their configurational assignment. Alter-
natively, the biogenesis of the laurefurenynes A and B could
involve a cascade of epoxide-mediated cyclizations,[1] or a
combination of these pathways. As fundamental understand-
ing of the biogenesis of the laurefurenynes will ultimately
rely on unequivocally establishing their stereochemical as-
signment, we endeavoured to extend our chlorohydrin-
based strategies for tetrahydrofuran synthesis[11,12] to the
preparation of laurefurenyne A (1). As described below, the
total synthesis of 1, as well as several epimers of the pro-
posed structure, led to the configurational revision for the
laurefurenynes A and B[13] and supports a biosynthetic path-
way for these molecules similar to that of elatenyne.
Figure 1. C15 acetogenins isolated from Laurencia sp. red algae and their
postulated biosynthesis from laurediol (8).
relation of the computed lowest energy conformations of
these diastereomers with the observed NOE data, from
which stereostructure 1 emerged as the best fit.[1] Consider-
ing, however, that both stereostructures were reasonable
candidates, we endeavored to develop a synthetic route to
laurefurenyne A that was sufficiently flexible to permit late
stage access to either stereostructure 1 or 16. Specifically, it
was envisioned that the stereochemically rich bis-tetrahydro-
furan core of the candidate stereostructures could be
formed selectively through the AgI-promoted cyclization of
a chlorodiol.[12a] The requisite chlorodiols 17 and 18 would
be derived from the common ketone 19 via a lithium aldol
reaction with either enantiomer of the a-chloroaldehyde
20,[14] followed by 1,3-anti-selective reduction of the carbon-
yl function.[12a] The common ketone 19 should also be ac-
cessed through a similar strategy (e.g., a-chloroaldehyde
The structural assignment of laurefurenyne A (1) and B
(2) was reported by Jaspars in 2010 and relied on a combina-
tion of 1D NOE and 2D NOESY studies to establish the rel-
ative stereochemistry of each of the tetrahydrofuran rings.[1]
Determination of the stereochemical relationship between
the tetrahydrofuran rings involved molecular modelling of
two potential stereostructures 1 and 16 (Scheme 1) and cor-
[a] M. T. Holmes, Dr. R. Britton
Department of Chemistry, Simon Fraser University
Burnaby, V5A 1S6, BC (Canada)
Fax : (+1)1-778-782-3765
E-mail: rbritton@sfu.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302352.
Chem. Eur. J. 2013, 19, 12649 – 12652
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12649

Scheme 2. a) LiCl, CuACHTNUGRTENUG(N CO2CF3)2, Na2S2O8, MeCN, H2O, 27 (20 mol%),
08C, 12 h; b) (3E)-3-methyl-3-penten-2-one, LDA, THF, ꢀ788C, 0.5 h;
then 22: 70% over 2 steps from 21; c) NaBH4, CeCl3·7H2O, MeOH, ꢀ78
8C, 40 min: 86% (d.r. 5:1); d) AgOTf, Ag2O, THF, 08C to RT, 36 h:
67%; e) TBSCl, imidazole, CH2Cl2, 08C to RT, 12 h: 92%; f) 1. O3,
CH2Cl2, ꢀ788C; 2.PPh3: 69%; g) DIBAL, THF, 608C, 5 d: 86%;
Scheme 1. Retrosynthesis of the candidate stereostructures for laurefure-
nyne A.
aldol, carbonyl reduction, cyclization) that builds on our
previous efforts in tetrahydrofuranol synthesis.[12]
h) DMSO, (COCl)2, ꢀ788C then Et3N, 1 h: 85%; i) LiCl, Cu
ACHTUNGTRENNUNG(CO2CF3)2,
Na2S2O8, MeCN, H2O, 27 (20 mol%), 08C, 12 h: 62%; j) LiCl, Cu-
AHCNUTRTGENN(GUN CO2CF3)2, Na2S2O8, MeCN, H2O, ent-27 (20 mol%), 08C, 12 h: 63%. Tf
= trifluoromethane-sulfonyl, LDA = lithium diisopropylamide, TBS =
tert-butyldimethylsilyl, DIBAL = diisobutylaluminum hydride.
As depicted in Scheme 2, synthesis of candidate stereo-
structures 1 and 16 initiated with the asymmetric a-chlorina-
tion of butanal (21) using the procedure described by Mac-
Millan.[14c] Due to difficulties associated with the purification
of the volatile a-chloroaldehyde 22, the crude product from
this reaction was treated directly with the lithium enolate
derived from 3-methyl-3-penten-2-one[15] to provide the anti-
ketochlorohydrin 23 along with minor amounts of the corre-
sponding syn-diastereomer (d.r. 6:1) in good yield over the
two steps. Luche reduction[16] of the ketochlorohydrin 23
then afforded the desired 1,3-anti-diol 24 as the major com-
ponent of an inseparable 5:1 mixture of diastereomers.
Treatment of this purified mixture with AgOTf/Ag2O[12a]
gave the tetrahydrofuranol 25 that was then protected as the
corresponding TBS ether. The relative stereochemistry of 25
was assigned following analysis of 1D NOESY spectra,
proved conditions for this transformation.[20] In addition,
both a Mukaiyama aldol reaction,[21] involving the TMS
enolsilyl ether derived from methyl ketone 26 and the a-
chloroaldehyde (R)-20, and a boron aldol reaction between
26 and (R)-20 provided very little (<5%) of the desired
product. Notwithstanding, a 1,3-anti-selective reduction of
the ketone function in 30 afforded the corresponding chloro-
diol (not shown) that cyclised using our AgOTf/Ag2O condi-
tions[12a] to yield the 2,2’-bis-tetrahydrofuran 31. Protection
of the alcohol function in 31 as a TBS ether, followed by
iodine-silicon exchange gave the (Z)-vinyl iodide 32 along
with small amounts of the corresponding E isomer. Finally,
a Sonogashira coupling[22] with TMS acetylene and global
deprotection completed the synthesis of the candidate ster-
eostructure 1. It is noteworthy that despite its linear nature,
the chlorohydrin-based strategy for tetrahydrofuran synthe-
sis[12] permits access to the correctly configured and stereo-
chemically rich 2,2’-bis-tetrahydrofuran 31 in only 9 steps
from butyraldehyde. Additionally, this asymmetric synthesis
relies solely on substrate-controlled stereoinduction con-
veyed by two readily available a-chloroaldehydes, each pre-
pared via organocatalysis.[14] As described in the Supporting
Information, following an identical sequence of reactions
that initiated with the lithium aldol coupling of methyl
ketone 26 and the enantiomeric a-chloroaldehyde (S)-20,
the candidate stereostructure 16 (Scheme 1) was also pre-
pared.
1
while comparison of H NMR spectra recorded on (R)- and
(S)-MTPA esters[17] of 25 and chiral GC analysis confirmed
both the absolute stereochemistry and high optical purity
(95% ee) of this material.[18] Oxidative cleavage of the
alkene function in 25 proceeded smoothly and completed
the synthesis of the methyl ketone 26. The aldehydes (S)-
and (R)-20 were obtained in a straightforward manner from
5-trimethylsilyl-4-pentynol (28) following DIBAL reduction
of the alkyne function and Swern oxidation.[19] Chlorination
of 29 using MacMillanꢁs SOMO a-chlorination procedure[14c]
provided the a-chloroaldehyde (S)-20 in good yield and op-
tical purity. Repetition of this reaction with catalyst ent-27
gave the antipode (R)-20.[18]
With the methyl ketone 26 and a-chloroaldehydes (R)-
and (S)-20 in hand, we next focused on combining these re-
agents in
a
lithium aldol reaction.[15] As detailed in
Having completed the total synthesis of both candidate
stereostructures 1 and 16, the spectral data derived from
these compounds was compared to that reported for the nat-
ural product laurefurenyne A. Surprisingly, the 1H and
13C NMR spectra of both 1 and 16 was clearly different from
that reported for laurefurenyne A. For example, compara-
Scheme 3, treatment of the lithium enolate derived from
ketone 26 with a-chloroaldehyde (S)-20 afforded the b-keto-
chlorohydrin 30, albeit in modest yield. Unfortunately, a
survey of bases (e.g., LiHMDS, KHMDS, NaHMDS), addi-
tives (e.g., LiCl, 12-C-4), and solvents failed to identify im-
12650
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12649 – 12652

Laurefurenynes A and B
COMMUNICATION
Fortunately, access to the revised candidate stereostruc-
tures 14 and 33 could be readily gained from intermediates
in the original synthesis of 1 and 16 (Scheme 3). As detailed
in Scheme 4, a Mitsunobu inversion[25] carried out on the
2,2’-bis-tetrahydrofurans 31 and 35 was followed by silicon–
iodine exchange to afford the bis-epimeric vinyl iodides 34
and 36, respectively. A series of reactions involving a Sono-
gashira coupling[22] with TMS acetylene, and subsequent re-
moval of the silyl and p-nitrobenzoate protecting groups
then provided the revised candidate stereostructures 33 and
14 in good overall yield. With these materials in hand, we
were delighted to find that the spectral data (1H NMR,
13C NMR, HRMS) derived from the 2,2’-bistetrahydrofuran
14 was in complete agreement with that reported for the
natural product laurefurenyne A[1] and differed significantly
from that of the diastereomer 33. In addition, the specific
rotation of 14 ([a]D =ꢀ6.2 cm3 gꢀ1 dm3, c=0.2 (CH3OH))
was consistent with that reported for the natural product
([a]D =ꢀ8.0 cm3 gꢀ1 dm3, c=0.1 (CH3OH)),[1] further con-
firming the absolute stereochemistry of laurefurenyne A to
be that depicted for compound 14 (Figure 2).
In summary, several diastereomers of the natural product
laurefurenyne A have been synthesized employing a stereo-
selective, chlorohydrin-based strategy to rapidly construct
their 2,2’-bis-tetrahydrofuran cores. Comparison of the spec-
tral data derived from the synthetic compounds to that re-
ported for the natural product led to a configurational revi-
sion for laurefurenyne A. Considering the similarities in
spectral data for laurefurenyne A and B (the natural E
isomer),[1] both the relative and absolute stereochemistry of
laurefurenyne B should also be revised accordingly. As lau-
refurenyne A (14) and elatenyne (3) (Figure 1) differ only in
Scheme 3. a) LiHMDS, THF, ꢀ40 to ꢀ788C, 0.5 h, then (S)-20, 0.5 h:
39%; b) Me4NBH
(OAc)3, MeCN, AcOH, ꢀ408C, 12 h: 73%; c) Ag2O,
AgOTf, THF, 08C to RT, 12 h: 52%; d) TBSOTf, 2,6-lutidine, CH2Cl2,
08C, 2 h: 82%; e) NIS, MeCN, 14 h: 70% (ratio Z/E isomers = 2:1);
f) [Pd
THF, 08C, 12 h: 75%. NIS = N-Iodosuccinimide, TBAF = tetra-n-butyl-
ACHTUNGTRENNUNGammonium fluoride.
ACHTUNGTRENNUNG(PPh3)4], CuI, Et3N, HCCSiMe3, 08C to RT, 14 h: 92%; g) TBAF,
tive 1H NMR data for compounds 1, 16, and the natural
product are presented in Figure 2 as the difference in chemi-
cal shift (ppm) between specific resonances reported for the
natural product[1] and the synthetic compounds 1 or 16.[18]
As indicated, resonances in the 1H NMR spectra of both
candidate stereostructures closely matched that reported for
the C10–C13 tetrahydrofuran ring and varied considerably
for the C6–C9 tetrahydrofuran ring. Based on these discrep-
ancies and the appreciated difficulty in the configurational
assignment of tetrahydrofurans owing to the multiple enve-
lope and twist conformations,[8,23] we re-evaluated the spec-
tral data reported for the natural product.[1] As depicted in
Figure 2, a key NOE correlation between H9 and H7 was
used to assign the relative stereochemistry at C7 in the natu-
ral product.[1] However, owing to partial overlap between
the resonances for H7 (d=
4.12 ppm) and H10 (d=
4.14 ppm), the observed NOE
correlation may well be be-
tween the adjacent protons H9
and H10.[24] Bearing this in
mind, and the consequent un-
certainty regarding the relative
stereochemistry at C7, we pro-
posed two revised candidate
stereostructures 14 and 33
(Figure 2) for the natural prod-
uct in which the C7 stereocen-
tres are inverted from those in
1 and 16. Furthermore, since
the spectral data derived from
the candidate stereostructure
16 was a closer match for H9
and H10 (bis-THF ring junc-
tion), the revised candidate
stereostructure 14, which main-
tains the same stereochemical
Figure 2. Difference plots of 1H NMR spectral data and revised candidate stereostructures 14 and 33. Bars in
graphs represent difference in chemical shift (ppm) between proton resonances in the candidate stereostruc-
tures 1 (top left) and 16 (top right) and those reported for the equivalent protons in the natural product laure-
furenyne A.
relationship between C9 and
C10, was deemed most likely
correct.
Chem. Eur. J. 2013, 19, 12649 – 12652
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12651

R. Britton and M. T. Holmes
[2] For the original isolation of elatenyne, see: J. G. Hall, J. A. Reiss,
Aust. J. Chem. 1986, 39, 1401–1409.
[3] For the re-isolation of elatenyne and its structural assignment as a
2,2’-bis-tetrahydrofuran, see: K. Kim, M. R. Brennan, K. L. Erick-
son, Tetrahedron Lett. 1989, 30, 1757–1760.
[4] a) N.-Y. Ji, X.-M. Li, K. Li, B.-G. Wang, J. Nat. Prod. 2007, 70,
1499–1502; b) N.-Y. Ji, X.-M. Li, K. Li, B.-G. Wang, J. Nat. Prod.
2010, 73, 1192.
[5] H. Kikuchi, T. Suzuki, E. Kurosawa, M. Suzuki, Bull. Chem. Soc.
Jpn. 1991, 64, 1763–1775.
[6] Y. Gopichand, F. J. Schmitz, J. Shelley, A. Rahman, D. Van der
Helm, J. Org. Chem. 1981, 46, 5192–5197.
Scheme 4. a) DIAD, PPh3, p-nitrobenzoic acid, THF, 08C: 87%; b) NIS,
MeCN, 14 h: 71% (ratio Z/E isomers = 2:1); c) [Pd(PPh3)4], CuI, Et3N,
AHCTUNGTRENNUNG
[7] H. M. Sheldrake, C. Jamieson, J. W. Burton, Angew. Chem. 2006,
118, 7357–7360; Angew. Chem. Int. Ed. 2006, 45, 7199–7202.
[8] S. G. Smith, R. S. Paton, J. W. Burton, J. M. Goodman, J. Org.
Chem. 2008, 73, 4053–4062.
HCCSiMe3, 08C to RT, 14 h: 86%; d) TBAF, THF, 08C, 12 h; e) NaOH,
MeOH, 1 h: 85% over two steps; f) DIAD, PPh3, p-nitrobenzoic acid,
THF, 08C: 87%; g) NIS, MeCN, 14 h, 61% (ratio Z/E isomers = 3:1);
h) [PdACHTUNGTRENNUNG(PPh3)4], CuI, Et3N, HCCSiMe3, 08C to RT, 14 h: 95%; i) TBAF,
[9] B. S. Dyson, J. W. Burton, T.-i. Sohn, B. Kim, H. Bae, D. Kim, J.
Am. Chem. Soc. 2012, 134, 11781.
THF, 08C, 12 h; j) NaOH, MeOH, 1 h: 87% over two steps. DIAD = dii-
sopropyl azodicarboxylate, PNB = p-nitrobenzoate.
[10] For the enzymatic cyclization of laurediols with bromoperoxidase,
see: a) A. Fukuzawa, M. Aye, Y. Takasugi, M. Nakamura, M.
Tamura, A. Murai, Chem. Lett. 1994, 2307–2310; b) A. Fukuzawa,
M. Aye, M. Nakamura, M. Tamura, A. Murai, Chem. Lett. 1990,
1287–1290.
[11] R. Britton, B. Kang, Nat. Prod. Rep. 2013, 30, 227–236.
[12] a) B. Kang, J. Mowat, T. Pinter, R. Britton, Org. Lett. 2009, 11,
1717–1720; b) B. Kang, S. Chang, S. Decker, R. Britton, Org. Lett.
2010, 12, 1716–1719; c) S. Chang, R. Britton, Org. Lett. 2012, 14,
5844–5847.
[13] For an alternative synthesis of laurefurenyne B that independently
confirms our findings, see the preceding Communication in this
issue: D. J. Shepherd, P. A. Broadwith, B. S. Dyson, R. S. Paton,
J. W. Burton, Chem. Eur. J. 2013, 19, 12644–12648. We thank Pro-
fessor Jonathan Burton for the exchange of information.
[14] For the asymmetric preparation of a-chloroaldehydes, see: a) M. P.
Brochu, S. P. Brown, D. W. C. MacMillan, J. Am. Chem. Soc. 2004,
126, 4108–4109; b) N. Halland, A. Braunton, S. Bachmann, M.
Marigo, K. A. Jørgensen, J. Am. Chem. Soc. 2004, 126, 4790–4791;
c) M. Amatore, T. D. Beeson, S. P. Brown, D. W. C. MacMillan,
Angew. Chem. 2009, 121, 5223–5226; Angew. Chem. Int. Ed. 2009,
48, 5121–5124; d) S. Dekeukeleire, M. D’hooghe, K. W. Tornroos, N.
De Kimpe, J. Org. Chem. 2010, 75, 5934–5940.
the relative configuration at C12, it is likely that these com-
pounds share a similar biosynthesis, with that of laurefure-
nyne A (14) involving attack of H2O (not Brꢀ) on the inter-
mediate oxonium 11 and an SN2 displacement of the C12
bromide by water (Figure 3). These results suggest that that
the Laurencia metabolites laurefurenynes C–F (e.g., 12 and
13, Figure 1),[1] which differ configurationally from laurefur-
enyne A at C6, may also require structural revision. The in-
terconversion of laurefurenyne A and C via a biomimetic
process[9] may well provide a means to access the latter com-
pounds and address this issue, work that is ongoing in our
laboratory.
[15] S. D. Halperin, B. Kang, R. Britton, Synthesis 2011, 1946–1953.
[16] J.-L. Luche, J. Am. Chem. Soc. 1978, 100, 2226–2227.
[17] J. A. Dale, D. L. Dull, H. S. Mosher, J. Org. Chem. 1969, 34, 2543–
2549.
[18] See Supporting Information for details.
Figure 3. Proposed biosynthesis of laurefurenyne A (14).
[19] K. Miura, T. Hondo, T. Nakagawa, T. Takahashi, A. Hosomi, Org.
Lett. 2000, 2, 385–388.
[20] The low isolated yield for 30 reflects poor conversion and/or decom-
position of the aldol product rather than difficulties in generating
the corresponding lithium enolate of 26. Treatment of the ketone 26
with LDA or LiHMDS (THF, ꢀ788C) followed by TMSCl afforded
the corresponding TMS enol silyl ether in good yield (>80%).
[21] T. Borg, J. Danielsson, P. Somfai, Chem. Commun. 2010, 1281–1283.
[22] K. Sonogashira, J. Organomet. Chem. 2002, 653, 46–49.
[23] A. Wu, D. Cremer, Int. J. Mol. Sci. 2003, 4, 158–192.
[24] In the Ph.D. thesis that describes the structural elucidation of laure-
furenyne A (“Investigations of Natural Products from Marine Or-
ganisms” Wael Mostafa Abdel-Mageed, 2009, Faculty of Chemistry,
University of Aberdeen, Scotland, UK), the NOE in question is as-
signed as H9 to H7 or H10. We thank Professor Marcel Jaspars for
providing us a copy of the appendix to this thesis.
Acknowledgements
This work was supported by an NSERC Discovery Grant and a Michael
Smith Foundation for Health Research Career Investigator Award to
R.B. The authors thank S. D. Halperin and B. Loosley (SFU) for prepara-
tion of some starting materials required for this work, and R. Gries
(SFU) for assistance with chiral GC analysis.
Keywords: chlorohydrin
· laurefurenyne A · laurencia
acetogenins · organocatalysis · total synthesis
[25] O. Mitsunobu, Synthesis 1981, 1–28.
[1] W. M. Abdel-Mageed, R. Ebel, F. A. Valeriote, M. Jaspars, Tetrahe-
dron 2010, 66, 2855–2862.
Received: June 19, 2013
Published online: August 19, 2013
12652
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12649 – 12652