Total synthesis of the antibiotic kendomycin: A macrocyclization using the Tsuji-Trost etherification

Article abstract of DOI:10.1002/anie.201400305

A highly stereocontrolled, convergent total synthesis of kendomycin [(-)-TAN2162], an ansa-macrocyclic antibiotic, is reported. The key of the strategy is an unprecedented Tsuji-Trost macrocyclic etherification, followed by a transannular Claisen rearrangement to construct the 18-membered carbocyclic framework. The oxa-six- and five-membered rings were also stereoselectively constructed respectively by a cascade oxidative cyclization at an unfunctionalized benzylic position and using a one-pot epoxidation/5-exo-tet epoxide opening. A new construct: The asymmetric total synthesis of the antibiotic kendomycin was accomplished by using a highly stereocontrolled convergent route. The key feature of the synthetic strategy is the construction of an 18-membered carbocycle based on an intramolecular Tsuji-Trost etherification/transannular Claisen rearrangement sequence. TBS=tert- butyldimethylsilyl.

Full text of DOI:10.1002/anie.201400305

Angewandte
Chemie
DOI: 10.1002/anie.201400305
Natural Products
Total Synthesis of the Antibiotic Kendomycin: A Macrocyclization
Using the Tsuji–Trost Etherification**
Tetsuya Sengoku, Shu Xu, Kenji Ogura, Yoshinori Emori, Kenji Kitada, Daisuke Uemura, and
Hirokazu Arimoto*
Abstract: A highly stereocontrolled, convergent total synthesis
of kendomycin [(À)-TAN2162], an ansa-macrocyclic anti-
biotic, is reported. The key of the strategy is an unprecedented
Tsuji–Trost macrocyclic etherification, followed by a trans-
annular Claisen rearrangement to construct the 18-membered
carbocyclic framework. The oxa-six- and five-membered rings
were also stereoselectively constructed respectively by a cascade
oxidative cyclization at an unfunctionalized benzylic position
and using a one-pot epoxidation/5-exo-tet epoxide opening.
albeit in modest yields. We have attempted an RCM strategy
[8]
À
for the macrocyclization at the C13 C14 double bond of 1.
Unfortunately, the unnatural Z diastereomer was the pre-
dominant product. However, in our studies the Claisen
rearrangement was found to work as a powerful tool for the
introduction of a carbon substituent at the highly hindered
C20a position.[8a] We finally found that the macrocyclic
etherification/Claisen rearrangement combination could be
used for the construction of kendomycinꢀs carbon framework.
The retrosynthetic analysis of 1 is outlined in Scheme 1. It
could be derived from 2 through aromatic oxidation and
oxidative removal of the terminal hydroxymethyl group. The
five-membered ring of 2 could possibly be constructed
selectively by a 5-exo cyclization of the phenol and the
neighboring olefin of 3. The macro-carbocycle of 3 was
envisioned to be derived from the transannular Claisen
rearrangement of the allyl aryl ether 4, which comes from
hydroquinone 5 using a selective macrocyclization at the less
hindered hydroxy group by means of Tsuji–Trost etherifica-
K
endomycin [(À)-TAN2162; 1; for structure see Scheme 1]
is an ansa-macrocyclic polyketide comprising a quinone
methide chromophore, and was originally isolated as an
antagonist for the endothelin receptor,[1] and as an antiosteo-
porotic agent.[2] Zeeck and Bode reported that 1 exhibited not
only cytotoxic effects against a number of human tumor cell
lines but also a strong antibacterial activity against both
Gram-positive and Gram-negative bacteria, notably methi-
cillin-resistant Staphylococcus aureus (MRSA) and vancomy-
cin intermediate S. aureus (VISA) Mu50 strain.[3] The
architectural structure and multiple biological activities of
1 have attracted significant interest in the past decade from
biological,[4] biosynthetic,[5] and chemically synthetic[6] per-
spectives.
The macrocyclic carbon framework of 1 represents
a major challenge for an efficient chemical synthesis.
Among all the achieved total syntheses and formal total
syntheses, several macrocyclization strategies, including C-
glycosydation,[7a] ring-closing metathesis (RCM),[7b,c,e,f,8] the
Barbier reaction,[7d] photo-Fries rearrangement,[7e,f] Dçtz
benzannulation,[7g] and Prins reaction[7h] have been reported,
[*] Dr. T. Sengoku,[$] [+] Dr. S. Xu,[+] K. Ogura, Y. Emori, K. Kitada,
Prof. Dr. H. Arimoto
Graduate School of Life Sciences, Tohoku University
2-1-1 Katahira, Sendai 980-8577 (Japan)
E-mail: arimoto@biochem.tohoku.ac.jp
Homepage: http://www.agri.tohoku.ac.jp/bunseki/index-j.html
Prof. Dr. D. Uemura
Department of Chemistry, Faculty of Science, Kanagawa University
Hiratsuka 259-1293 (Japan)
[$] Current address: Department of Materials Science, Faculty of
Engineering, Shizuoka University, Shizuoka 432-8561 (Japan)
[+] These authors contributed equally to this work.
[**] We are grateful for the financial support of Grants-in-Aid for
Scientific Research (Nos.18032010 and 21310136) from the JSPS.
T.S. acknowledges the JSPS for the Research Fellowships for Young
Scientists.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201400305.
Scheme 1. Retrosynthetic analysis. TBS=tert-butyldimethylsilyl.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
tion. The macrocyclization precursor 5 would be prepared
convergently from the aldehyde 6, alkylmagnesium 7, and
iodide 8. The tetrahydropyran ring could be cyclized diaste-
reoselectively at the unfunctionalized C5-position by a Ag2O-
triggered cascade reaction, as we have previously reported.[8a]
We have previously reported on the preparation of the
C1–C13 segment of 1,[8a] and herein, we present a modified
scalable synthetic route (Scheme 2). The modified route
compound 7[13] to 6 gave a 15:1 diastereomer mixture of the
secondary alcohol 13, favoring the expected Felkin–Anh-type
product. This selectivity is higher than our previous result
(5.5:1) which was obtained using a similar alkyllithium
nucleophile.[8a] After introduction of a phenolic hydroxy
group to the aromatic ring, 14 was stereoselectively cyclized at
the unfunctionalized benzylic position under phenol oxida-
tion conditions (Ag2O[14]) and quantitatively afforded the
tetrahydropyran 16 as a single diastereomer. The excellent
selectivity was thought to be due to the chair-form oxa-
Michael-addition transition-state 15 or 15’ with the bulky
quinone methide in an equatorial position. This result also
proposed another possibility for the unclear biosynthetic
mechanism of kendomycinꢀs pyran ring formation.[5] With this
modified route, 16 was obtained on a 3 gram scale.
The conversion of phenol 16 into 5 is shown in Scheme 3.
After a manipulation of the protecting groups, Swern
oxidation and Corey–Fuchs protocol furnished the alkyne
17. Hydrozirconation with the Schwartz reagent[15] and
subsequent iodination proceeded with high regioselectivity
(9:1) along with removal of a MOM ether. The regioisomers
18 and 18’ were not easily separable from each other, and
Scheme 2. Synthesis of C1–C13 segment 16: a) 11, NaHMDS, THF,
À608C; b) LiAlH4, THF, À788C to 08C, 82% (2 steps); c) (COCl)2,
DMSO, Et3N, CH2Cl2, À788C to 08C; d) (R,R)-[(E)-2-butenyl]diisopropyl
tartrate boronate, 4ꢀ M.S., PhMe, À788C, 77% (2 steps); e) TBSOTf,
2,6-lutidine, CH2Cl2, 08C, quant.; f) OsO4 (cat.), NMO, acetone/H2O
(4:1), RT; g) NaIO4, EtOH/H2O (1:1), RT, 93% (2 steps); h) 7, À788C,
THF, quant. (d.r. 15:1); i) H2, Pd(OH)2/C, EtOAc, RT; j) TBSCl, DMAP,
Et3N, CH2Cl2, RT, 85% (2 steps); k) HBr3·Py, K2CO3, Py, 08C to RT;
l) ethyl vinyl ether, PPTS, RT, quant. (2 steps); m) nBuLi, B(OMe)3,
THF, À788C to RT, then 30% H2O2 aq., Na2CO3, RT, 90%; n) PPTS,
nPrOH, RT, 99%. DMAP=4-(N,N-dimethylamino)pyridine, DMSO=
dimethylsulfoxide, HMDS=hexamethyldisilazide, M.S.=molecular
sieves, NMO=N-methylmorpholine N-oxide, PPTS=pyridinium
p-toluenesulfonate, Py=pyridine, Tf=trifluoromethanesulfonyl,
THF=tetrahydrofuran.
Scheme 3. Synthesis of the macrocyclization precursor 5: a) MOMCl,
NaH, THF, RT; b) PPTS, THF/MeOH (1:1), RT, 92% (2 steps);
c) (COCl)2·DMSO, Et3N, CH2Cl2, À788C; d) CBr4, PPh3, Et3N, CH2Cl2,
À788C, 72% (2 steps); e) nBuLi, MeI, THF, À788C to RT, 98%;
f) [HZrCp2Cl], PhH, RT, then I2; g) 8 (2 equiv), tBuLi, Et2O, À788C,
then MeO-9-BBN, THF, À788C to RT, then 18+18’, [Pd-
commenced with the commercially available benzoic acid 9,
which was easily converted into the aryl bromide 10 with
a two-step procedure.[9] Evans asymmetric alkylation[10] and
reduction furnished the S-alcohol 12 with more than 99% ee,
which was determined using the modified Mosherꢀs
method.[11] Swern oxidation and Roush crotylation[12]
afforded a single diastereomer containing the three desired
contiguous stereogenic centers, which was further trans-
formed to aldehyde 6. Nucleophilic addition of the Grignard
(dppf)2Cl2]·CH2Cl2 (20 mol%), 3mK3PO4 aq., DMF, RT, 87% (2 steps);
h) MOMCl, NaH, THF, RT, 95%; i) PPTS, THF/MeOH (1:1), RT, 73%;
j) ClCO2Et, Py, RT, 94%; k) (NH4)2Ce(NO3)6, MeCN/H2O (2:1), RT;
l) Na2S2O4, MeCN/H2O (1:1), 08C, quant. (2 steps). BBN=bora-
bicyclo[3.3.1]nonane, Cp=cyclopentadienyl, DMF=N,N-dimethylform-
amide, dppf=1,1’-bis(diphenylphosphino)ferrocene, MOM=methoxy-
methyl.
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
were thus used as a mixture for the Suzuki–Miyaura cross-
coupling reaction[16] with boranate 19. The boranate was
prepared by lithiation and transmetalation of 8.[17] The desired
coupling product 20 was isolated in 87% yield (two steps).
Protection of 20 with MOMCl, selective removal of the
terminal TBS group, carbonate formation, and exposure of
the two hydroxy groups of the p-hydroquinone yielded the
Tsuji–Trost cyclization precursor 5.
The examination of the key palladium-catalyzed Tsuji–
Trost cyclization is shown in Table 1. Although the Tsuji–
Trost reaction is a well-established synthetic protocol,[18] to
the best of our knowledge, the macroetherification with this
reaction has not been reported in total synthesis studies of
natural products. More challenging is that our substrate, 5, is
actually a hydroquinone which contains two hydroxy nucle-
ophiles, and is sensitive to oxidation, which may render the
reaction more complicated. However, to avoid the tedious
and poorly selective protecting strategy for the hindered C4-
OH group, we decided to perform the cyclization with both
free hydroxy groups. Eventually, it was found that the desired
macrocyclic ether 4 was obtained with [Pd2(dba)3] and dppb at
room temperature in THF,[19] though the regioisomer 4’ was
the major product (entry 1). Solvent effects were later found
to have great influence on the regioselectivity, with CH2Cl2
giving the best result (4/4’ = 10:1, entries 5 and 6). Notably,
when the concentration of 5 was higher than 10 mm,
intermolecular coupling became significant. Other palladium
catalysts such as [Pd(PPh3)4] or Pd(OAc)2 gave either no
reaction or decomposed byproducts.[20] Although the solvent
has been shown to play a key role in the ion-pairing in the
Tsuji–Trost reaction,[21] the specific reason for the selectivity
in this study is still not clear.
Scheme 4. Complete total synthesis of 1. mCPBA=m-chloroperoxyben-
zoic acid, TBAF=tetra-n-butylammonium fluoride.
tion with an acyclic substrate in refluxing Me2NPh.[8a] This
time, after TBS protection of the phenol 4 and heating at
2058C in Et2NPh (a more stable solvent), 21 was smoothly
converted into the desired carbocyclic 3 in 99% yield.[23] Next,
several strategies (iodoetherification, oxymercuration, etc.)
were attempted to cyclize the five-membered ring of kendo-
mycin. Eventually, epoxidation and subsequent intramolecu-
lar epoxide opening proved to be most efficient. Thus, after
alkene epoxidation with mCPBA in CH2Cl2, the 5-exo-tet
cyclization proceeded smoothly by just adding a protic
solvent, MeOH, and furnished predominately one diastereo-
mer of 22. Since the upper trisubstituted alkene was also
epoxidized during the process, a subsequent deoxygena-
tion[7b,c,24] afforded the compound 2 in 82% yield (two steps)
as a single diastereomer. Next a hypervalent-iodine-promoted
b-fragmentation of the alkoxyl radical removed the terminal
hydroxymethyl group, and presumably provided an unstable
intermediate tert-iodide, which was gradually eliminated to
give the benzofuran 23. This elimination process was accel-
erated by the addition of a small amount of water. The
compound 23 has the same skeleton as intermediates
obtained by the groups of Lee[7a] and Mulzer.[7e,f] Finally, the
selective removal of the phenolic TBS group, oxidation, and
acidic rearrangement of the resulting o-quinone, with the
simultaneous TBS removal, accomplished the total synthesis
of kendomycin.
After obtaining the 18-membered macrocyclic ether, we
next attempted a ring contraction through a Claisen rear-
rangement[22] (Scheme 4). We have reported this transforma-
Table 1: Regioselective macrocyclization using Tsuji–Trost etherifica-
tion.[a]
Entry
[Pd2(dba)3]
(equiv)
dppb
(equiv)
Solvent
T [8C]
4/4’[b]
Yield
4 [%][c]
1
2
3
4
5
6
0.15
0.5
1.1
0.2
0.4
0.4
2
20
1.6
10
0.86
THF
toluene
DMF
(ClCH2)2
CH2Cl2
CH2Cl2
RT
RT
60
70
50
70
1:3[d]
1:1
–
–
–
–
80
71
3:5
[e]
–
10:1
10:1
0.14
[a] The reaction was performed with 1–5 mm of 5 in the dark in a sealed
tube. [b] Determined by 1H NMR analysis of the crude reaction mixture.
[c] Yield of the isolated product. [d] Approximately 20% of the starting
material remained. [e] Only starting material recovered. dba=dibenzyl-
ideneacetone, dppb=1,4-bis(diphenylphosphino)butane.
In summary, we have achieved the total synthesis of the
antibiotic kendomycin (1) by using a highly stereocontrolled
convergent strategy. The key step of this synthetic route is the
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[10] D. A. Evans, M. D. Ennis, D. J. Mathre, J. Am. Chem. Soc. 1982,
104, 1737 – 1739.
[11] J. M. Seco, E. Quinoa, R. Riguera, Chem. Rev. 2004, 104, 17 –
118.
[12] W. R. Roush, A. D. Palkowitz, K. Ando, J. Am. Chem. Soc. 1990,
112, 6348 – 6359.
[13] The bromide precursor of the alkylmagnesium 7 was prepared
one step from the corresponding alcohol. For a three-step
synthesis of the alcohol from commercially available starting
materials, see: a) T. J. Donohoe, T. J. C. OꢀRiordan, M. Peifer,
C. R. Jones, T. J. Miles, Org. Lett. 2012, 14, 5460 – 5463; b) A. W.
van Zijl, F. Lꢂpez, A. J. Minnaard, B. L. Feringa, J. Org. Chem.
2007, 72, 2558 – 2563.
palladium-catalyzed regioselective Tsuji–Trost macroetheri-
fication/transannular Claisen rearrangement sequence to
construct the macrocyclic framework. Other synthetic fea-
tures include the tetrahydropyran installation by stereoselec-
tive nucleophilic addition to an aldehyde and a Ag2O-
triggered stereoselective cyclization and the establishment
of a five-membered ring by one-pot epoxidation/5-exo-tet
cyclization process.
Received: January 11, 2014
Published online: && &&, &&&&
[14] S. R. Angle, K. D. Turnbull, J. Am. Chem. Soc. 1989, 111, 1136 –
1138.
[15] D. W. Hart, T. F. Blackburn, J. Schwartz, J. Am. Chem. Soc. 1975,
97, 679 – 680.
[16] a) N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457 – 2483;
b) J. A. Marshall, B. A. Johns, J. Org. Chem. 1998, 63, 7885 –
7892.
[17] For the synthesis of 8, see the Supporting Information.
[18] B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103, 2921 – 2943.
[19] a) C. Damez, J.-R. Labrosse, P. Lhoste, D. Sinou, Synthesis 2001,
1456 – 1458; b) G. Wang, D. Niu, T.-L. Qiu, L. T. Phan, Z. Chen,
A. Polemeropoulos, Y. S. Or, Org. Lett. 2004, 6, 4455 – 4458; c) C.
Goux, P. Lhoste, D. Sinou, Synlett 1992, 725 – 727.
[20] Among the byproducts, the quinone A and methylenedioxy
arene B were structurally determined. A was speculated to
derive from oxidation of the hydroquinone 5 by high-valence
palladium intermediates. B might be generated from A by
a light-triggered transformation. For similar examples, see: C.
Thommen, C. K. Jana, M. Neuburger, K. Gademann, Org. Lett.
2003, 5, 1390 – 1393, and references therein.
Keywords: antibiotics · macrocycles · natural products ·
palladium · total synthesis
.
[1] a) Y. Funahashi, T. Ishimaru, N. Kawamura, JP Patent 08231551,
1996; b) Y. Funahashi, N. Kawamura, T. Ishimaru, JP Patent
08231552, 1996.
[2] M. H. Su, M. I. Hosken, B. J. Hotovec, T. L. Johnston, US Patent
5728727, 1998.
[3] H. B. Bode, A. Zeeck, J. Chem. Soc. Perkin Trans. 1 2000, 323 –
328.
[4] a) Y. A. Elnakady, M. Rohde, F. Sasse, C. Backes, A. Keller, H.-
P. Lenhof, K. J. Weissman, R. Mꢁller, ChemBioChem 2007, 8,
1261 – 1272; b) C. O. Janssen et al., Bioorg. Med. Chem. Lett.
2008, 18, 5771 – 5773 (see the Supporting Information for all
authors).
[5] a) H. B. Bode, A. Zeeck, J. Chem. Soc. Perkin Trans. 1 2000,
2665 – 2670; b) S. C. Wenzel, H. B. Bode, I. Kochems, R. Mꢁller,
ChemBioChem 2008, 9, 2711 – 2721.
[6] For reviews, see: a) H. J. Martin, T. Magauer, J. Mulzer, Angew.
Chem. 2010, 122, 5746 – 5758; Angew. Chem. Int. Ed. 2010, 49,
5614 – 5626; b) D. Bicchielli, Y. Borguet, L. Delaude, A. Demon-
ceau, I. Dragutan, V. Dragutan, M. Hans, C. Jossifov, F. Nicks, Q.
Willem, Curr. Org. Synth. 2012, 9, 397 – 405; c) M. Shan, L.
Wang, Q. Zhang, G. A. OꢀDooherty, Chemtracts 2009, 22, 1 – 17.
[7] For total syntheses, see: a) Y. Yu, H. Men, C. Lee, J. Am. Chem.
Soc. 2004, 126, 14720 – 14721; b) A. B. Smith III, E. F. Mesaros,
E. A. Meyer, J. Am. Chem. Soc. 2005, 127, 6948 – 6949; c) A. B.
Smith III, E. F. Mesaros, E. A. Meyer, J. Am. Chem. Soc. 2006,
128, 5292 – 5299; d) J. T. Lowe, J. S. Panek, Org. Lett. 2008, 10,
3813 – 3816; e) T. Magauer, H. J. Martin, J. Mulzer, Angew.
Chem. 2009, 121, 6148 – 6152; Angew. Chem. Int. Ed. 2009, 48,
6032 – 6036; f) T. Magauer, H. J. Martin, J. Mulzer, Chem. Eur. J.
2010, 16, 507 – 519; g) K. Tanaka, M. Watanabe, K. Ishibashi, H.
Matsuyama, Y. Saikawa, M. Nakata, Org. Lett. 2010, 12, 1700 –
1703; For formal total syntheses, see: h) K. B. Bahnck, S. D.
Rychnovsky, J. Am. Chem. Soc. 2008, 130, 13177 – 13181; i) K. B.
Bahnck, S. D. Rychnovsky, Chem. Commun. 2006, 2388 – 2390.
[8] a) T. Sengoku, H. Arimoto, D. Uemura, Chem. Commun. 2004,
1220 – 1221; b) T. Sengoku, D. Uemura, H. Arimoto, Chem. Lett.
2007, 36, 726 – 727.
[21] L. A. Evans, N. Fey, J. N. Harvey, D. Hose, G. C. Lloyd-Jones, P.
Murray, A. G. Orpen, R. Osborne, G. J. J. Owen-Smith, M.
Purdie, J. Am. Chem. Soc. 2008, 130, 14471 – 14473.
[22] E. A. Ilardi, C. E. Stivala, A. Zakarian, Chem. Soc. Rev. 2009, 38,
3133 – 3148.
[23] In a recent total synthesis of kendomycin (Ref. [7g]), Nakataꢀs
group applied a set of reaction conditions, similar to our
previously reported method, on a substrate which was more
acidic than 21. In their case, Ac2O had to be used for the in situ
trap of the Claisen product to avoid further decomposition. In
our substrate case, no decomposition was observed.
[24] K. B. Sharpless, M. A. Umbreit, M. T. Nieh, T. C. Flood, J. Am.
Chem. Soc. 1972, 94, 6538 – 6540.
[9] S. R. Kasibhatla, K. D. Hong, M. F. Boehm, M. A. Biamonte, L.
Zhang, US Patent 20050113340A1, 2005.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Natural Products
T. Sengoku, S. Xu, K. Ogura, Y. Emori,
K. Kitada, D. Uemura,
H. Arimoto*
&&&& — &&&&
Total Synthesis of the Antibiotic
Kendomycin: A Macrocyclization Using
the Tsuji–Trost Etherification
A new construct: The asymmetric total
synthesis of the antibiotic kendomycin
was accomplished by using a highly ste-
reocontrolled convergent route. The key
feature of the synthetic strategy is the
construction of an 18-membered carbo-
cycle based on an intramolecular Tsuji–
Trost etherification/transannular Claisen
rearrangement sequence. TBS=tert-
butyldimethylsilyl.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!