Total synthesis of (-)-quinocarcin by gold(I)-catalyzed regioselective hydroamination

Article abstract of DOI:10.1002/anie.201205106

In control: The novel and enantioselective total synthesis of (-)-quinocarcin includes the highly stereoselective preparation of the 2,5-cis-pyrrolidine by intramolecular amination, a selective substrate-controlled 6-endo-dig intramolecular alkyne hydroamination with a cationic Au^I catalyst, and Lewis-acid-mediated ring-opening/ halogenation sequence. Copyright

Full text of DOI:10.1002/anie.201205106

Angewandte
Chemie
DOI: 10.1002/anie.201205106
Natural Products
Total Synthesis of (ꢀ)-Quinocarcin by Gold(I)-Catalyzed
Regioselective Hydroamination**
Hiroaki Chiba, Shinya Oishi, Nobutaka Fujii,* and Hiroaki Ohno*
(ꢀ)-Quinocarcin is a pentacyclic tetrahydroisoquinoline alka-
loid (see Scheme 1 for structure). It has attracted consider-
able attention from both synthetic and biological chemists
interested in its intricate polycyclic architecture and potent
broad-spectrum antitumor activity.[1,2] After its isolation from
Streptomyces melanovinaceus in 1983 by Takahashi, Tomita
and co-workers,[3] several efficient syntheses of quinocarcin
have been reported by the groups of Fukuyama,[4] Garner,
Terashima, Myers, Zhu, and Stoltz.[5] The piperizinohydro-
isoquinoline motif, the core structure of quinocarcin, appears
in a variety of alkaloids such as tetrazomine, lemonomycin,
and ecteinascidins.[1] The development of a novel strategy that
facilitates convergent assembly of this tetracyclic core unit is
therefore desirable.
The tetrahydroisoquinoline ring system can be efficiently
constructed by a Pictet–Spengler condensation,[4,5d,e] but we
planned an alternative strategy using a gold(I)-catalyzed
intramolecular alkyne hydroamination, a simple and atom-
economical protocol for the construction of nitrogen-con-
Scheme 1. Retrosynthetic analysis of quinocarcin.
taining heterocycles from free amines, sulfonamides, and
carbamates.[6] Combined with a Sonogashira coupling, this
strategy provides a convergent approach to the quinocarcin
core structure.
intramolecular amination of the bromoallene 6 bearing an
amide group, a method originally developed by our group.[7]
Preparation of the bromoallene 6 began with the diaste-
reoselective propargylation of the g-butyrolactone
8
Our retrosynthetic analysis is shown in Scheme 1. We
expected that the known lactam 1[5e,f] could be synthesized
from the dihydroisoquinoline 2 by a diastereoselective hydro-
genation and intramolecular amide formation. The dihydroi-
soquinoline 2 could be obtained from the alkyne 3 by an
intramolecular hydroamination. Although control of the
regioselectivity of this hydroamination (6-endo versus 5-exo)
was a challenging issue in this strategy, we hoped that the
desired 6-endo-dig-selective reaction could be achieved by
appropriate tuning of the catalyst or substrate structure. The
alkyne 3 was retrosynthetically envisioned to arise from the
Sonogashira coupling of the phenylglycinol 4 and the 2,5-cis-
pyrrolidine 5. We thought that 5 could be stereoselectively
prepared by 2,5-cis-selective pyrrolidine formation through
(Scheme 2). Subsequent reduction of the lactone 9 with
LiBH4[8] afforded diol 10, the primary hydroxy group of which
was selectively acetylated to give the alcohol 11. Introduction
of the nitrogen functionality by a Mitsunobu reaction[9] of 11
with TsBocNH and subsequent selenium oxidation produced
the propargyl alcohol 13 as a mixture of diastereomers.
Treatment of 13 with MsCl and Et3N gave the corresponding
mesylate, which was then reacted with CuBr·Me2S/LiBr,[10]
followed by removal of the Boc group with TFA to afford the
bromoallene 6 as a mixture of diastereomers (d.r. = 55:45;
determined by 1H NMR spectroscopy). We attempted the 2,5-
cis-selective pyrrolidine formation by intramolecular amina-
tion of 6.[7] Treatment of 6 with NaH in DMF at room
temperature successfully produced the desired 2,5-cis-pyrrol-
idine 15 in 95% yield, with excellent diastereoselectivity (cis-
15/trans-15 = 96:4). In contrast, an intramolecular SN2 reac-
tion using the mesylate 18 under identical reaction conditions
afforded almost equal amounts of 2,5-cis- and trans-15 (cis-15/
trans-15 = 55:45, Scheme 3). The desired pyrrolidine 5 was
then prepared from cis-15 by removal of the TBDPS
[TBDPS = tert-butyl(diphenyl)silyl] and Ac groups, oxida-
tion, esterification, detosylation, and subsequent N-methyl-
ation using standard protocols (Scheme 2).
[*] H. Chiba, Dr. S. Oishi, Prof. Dr. N. Fujii, Prof. Dr. H. Ohno
Graduate School of Pharmaceutical Sciences, Kyoto University
Sakyo-ku, Kyoto 606-8501 (Japan)
E-mail: nfujii@pharm.kyoto-u.ac.jp
hohno@pharm.kyoto-u.ac.jp
[**] This work was supported by a Grant-in-Aid for the Encouragement
of Young Scientists (A) (H.O.) and a Platform for Drug Design,
Discovery, and Development from the MEXT (Japan). H.C. is
grateful for Research Fellowships from the Japan Society for the
Promotion of Science (JSPS) for Young Scientists.
With the key building block 5 secured, efforts were
directed toward a model study for construction of the
isoquinoline skeleton through gold(I)-catalyzed intramolecu-
lar hydroamination (Scheme 4). The model substrates 19a–d
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205106.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Scheme 2. Stereoselective synthesis of 5. Reagents and conditions:
a) TBDPSCl, imidazole, DMF, 208C; b) LHMDS, propargyl bromide,
THF, ꢀ808C; c) LiBH4, MeOH, Et2O, 208C; d) AcCl, 2,4,6-collidine,
CH2Cl2, ꢀ208C; e) TsBocNH, DIAD, PPh3, THF, 208C; f) SeO2, TBHP,
1,2-DCE, 608C; g) MsCl, Et3N, CH2Cl2, 208C; then CuBr·Me2S, LiBr,
THF, 508C; h) TFA, CH2Cl2, ꢀ208C; i) NaH, DMF, 208C; j) TBAF, THF,
208C; k) K2CO3, MeOH, 208C; l) DMP, CH2Cl2, 208C; m) NaClO2,
2-methylbut-2-ene, NaH2PO4, tBuOH/CH2Cl2/H2O (8:3:8), 08C;
n) HOBt, EDC·HCl, MeOH, 208C; o) Mg, MeOH, ꢀ208C; p) MeI,
Cs2CO3, acetone, 208C. Boc=tert-butoxycarbonyl, 1,2-DCE=1,2-
dichloroethane, DIAD=diisopropyl azodicarboxylate, DMF=N,N-
Scheme 4. Model study of the gold(I)-catalyzed intramolecular hydro-
amination (6-endo versus 5-exo). Tf=trifluoromethanesulfonyl.
=
dimethylformamide, DMP=Dess–Martin periodinane, EDC N-(3-
dimethylaminopropyl)-N’-ethylcarbodiimide, HOBt=1-hydroxybenzo-
triazole, LHMDS=lithium hexamethyldisilazide, Ms=methanesul-
fonyl, TBAF=tetra-n-butylammonium fluoride, TBDPS=tert-butyldi-
phenylsilyl, TBHP=tert-butylhydroperoxide, TFA=trifluoroacetic acid,
THF=tetrahydrofuran, Ts=4-toluenesulfonyl.
20b in 61% yield, along with the 5-exo-dig product 21b (32%
yield). However, the regioselectivity decreased when the
acetonide-type substrate 19c bearing a methyl ester was used.
The desired 6-endo-dig product 20d was exclusively obtained
in 73% yield using the dihydrobenzofuran-type substrate 19d
and the catalyst A, presumably as a result of the ring strain of
the 5-exo-dig product 21d. Changing the ligand showed that
the cationic gold catalyst B, generated in situ from IPr/AuCl
and AgNTf2, in 1,2-DCE afforded 20d in 96% yield.
Encouraged by this substrate-controlled switch in the regio-
selectivity, we proceeded to prepare an optically active
dihydrobenzofuran-type substrate for the total synthesis.
The preparation of the optically active 4d was initiated by
Wittig reaction of the aldehyde 22 (Scheme 5). Following
Sharplessꢀs protocol for the dihydroxylation,[14] the diol 24
was obtained from the resulting dihalostyrene 23 in good yield
with moderate enantioselectivity (83% yield, 81% ee). A
single recrystallization from chloroform provided the opti-
cally pure diol 24 (> 99% ee). Regioselective silylation of 24
followed by a Mitsunobu reaction with diphenylphosphoryl
azide (DPPA)[15] afforded the azide 26 in excellent yield.
After removing the TBS group with TBAF, an intramolecular
SNAr reaction of the resulting alcohol 27 using tBuOK in THF
at room temperature provided the dihydrobenzofuran 28 in
51% yield (60%, based on recovered starting material).[16]
Although the Staudinger reduction of the azide 28[17] did not
give satisfactory results, treatment of 28 with PhSH, SnCl2,
and Et3N in THF[18] successfully afforded the corresponding
Scheme 3. Intramolecular SN2 reaction of mesylate 18.
were prepared by copper-free Sonogashira coupling[11] using
the racemic aryl iodides 4.[12] After optimization of the
reaction conditions using 19a, a cationic gold(I) catalyst was
found to be the most efficient catalyst for promotion of the
hydroamination among the various transition-metal catalysts
tested (Cu, Rh, In, Pt, and Au).[12] However, the undesired 5-
exo-dig cyclization product 21a was obtained exclusively in
most cases (74% when using the cationic gold catalyst A).[13]
The observed regioselectivity can be rationalized by the
ꢀ
electronic properties of the alkyne: a new C N bond was
formed at the cationic carbon atom bearing an aryl substitu-
ent when activated by the gold catalyst. We thought that
substrate modification would affect the regioselectivity of the
hydroamination. Use of the seven-membered acetonide-type
substrate 19b overcame the inherent preference for 5-exo-dig
cyclization, thereby leading to the desired 6-endo-dig product
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Scheme 5. Synthesis of optically active dihydrobenzofuran 4d.
Reagents and conditions: a) MePPh3Br, LHMDS, THF, 08C; b) OsO4,
(DHQ)2PHAL, K2CO3, [K3Fe(CN)6], MeSO2NH2, tBuOH/H2O (1:1),
08C; c) TBSCl, DMAP, Et3N, CH2Cl2, 208C; d) DPPA, DEAD, PPh3,
THF, 08C; e) TBAF, THF, 08C; f) tBuOK, THF, 208C; g) PhSH, SnCl2,
Et3N, THF, 208C; h) Boc2O, Et3N, CH2Cl2, 208C. DEAD=diethyl
azodicarboxylate, (DHQ)2PHAL=hydroquinine 1,4-phthalazinediyl
diether, DMAP=4-(dimethylamino)pyridine, TBS=tert-butyldimethyl-
silyl.
amine, which was converted into the optically pure 4d in 83%
yield by treatment with Boc2O/Et3N.
With both the building blocks, 4d and 5, in hand, the stage
was set for their coupling and elaboration into quinocarcin
(Scheme 6). The treatment of equimolar amounts of 4d and 5
with [Pd(PPh3)4], CuSO4, and sodium ascorbate[19] in DMF
and Et3N at 808C provided the coupling product 3a in 92%
yield. It is noteworthy that the use of CuSO4/sodium
ascorbate was vital to prevent the generation of undesired
homocoupling products of the terminal alkyne 5. We then
began construction of the dihydroisoquinoline structure
through the established gold(I)-catalyzed hydroamination of
3a. Our initial attempt at the hydroamination of 3a resulted
in substantial decomposition of 3a under the reaction
conditions that were optimized for the model substrate 19d.
Further investigations using the catalyst B and increased
loadings of catalyst A were unsuccessful, thus resulting in
recovery of the starting material and a poor catalyst turnover
(46% yield using 40 mol% catalyst A), respectively. Consid-
ering that the serious steric repulsion between the methyl
ester and Boc groups would impair formation of the required
conformer for the hydroamination, we prepared the corre-
sponding amine 3b by cleavage of the Boc group. The desired
6-endo-dig cyclization proceeded efficiently upon treatment
of 3b with the catalyst A. Because the resulting enamine
product was unstable, we isolated the desired 6-endo-dig
product 30 in the tetrahydroisoquinoline form after stereo-
selective reduction with NaBH3(CN) (90% yield after
2 steps). Upon heating in the presence of AcOH at 808C,
the secondary amine selectively condensed with one of the
ester groups to form the diazabicyclo[3.2.1]octane core 31 in
96% yield. This is a related strategy to that of the late-stage
construction of the piperadine ring reported by Fukuyama
and Nunes[4] and Allan and Stoltz.[5f]
Scheme 6. Total synthesis of quinocarcin. Reagents and conditions:
a) [Pd(PPh3)4], CuSO4, sodium ascorbate, DMF/Et3N (3:2), 808C;
b) cat. A or B, 1,2-DCE; c) TFA, CH2Cl2, 08C; d) cat. A or B, 1,2-DCE;
then NaBH3(CN), MeOH, 1n HCl, 08C; e) AcOH, toluene, 808C;
f) BF3·Et2O, SiCl4, 1,2-DCE, 208C; then CsCl, MeCN, 608C; g) Me2SO4,
Cs2CO3, acetone, 208C; h) AgNO3, Et3N, acetone/H2O (3:1), 508C;
i) LiOH·H2O, THF/H2O (2:1), 208C; j) Li, NH3(l), THF, ꢀ788C!
ꢀ308C.
mediated ring-opening halogenation of benzofurans in the
presence of SiCl4 and BF3·AcOH,[20] we expected that the
neighboring lactam carbonyl group of 31 would assist the
benzofuran cleavage to provide the oxazolidinium intermedi-
ate 32, which could be converted into the phenylglycinol
derivative through hydrolysis. Our initial attempt revealed
that the exposure of 31 to BF3·Et2O and SiCl4 in 1,2-DCE
afforded a suspension, which possibly contained the expected
oxazolidinium intermediate 32.[21] However, aqueous workup
of this suspension only resulted in recovery of the starting
material (68%). This disappointing result can be attributed to
hydrolysis of the transient silyl ether in 32 prior to the
required cleavage of the oxazolidinium ring, thus promoting
the reverse reaction (benzofuran formation) to lactam 31.
After optimization of the workup procedure, use of CsCl gave
the desired result and produced the phenol 33 in 92% yield.
NOE experiments on this compound confirmed the relative
stereochemistry. Methylation of 33 with dimethyl sulfate
afforded the lactam 34 in 94% yield. After conversion of 34
into the alcohol 1a using acetone/H2O in the presence of
AgNO3 and Et3N, 1a was successfully converted into quino-
For the completion of the total synthesis, we had to
overcome the newly generated and challenging task of
cleaving the dihydrobenzofuran ring for construction of the
phenylglycinol moiety of quinocarcin. On the basis of the LiI-
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
carcin using procedure reported by Allan and Stoltz.[5f] The
spectroscopic data for our synthetic (ꢀ)-quinocarcin were
identical to those of the original isolation report and to the
reports on synthetic quinocarcin.[5]
In conclusion, we have achieved the asymmetric total
synthesis of quinocarcin. Notably, the regioselectivity of
intramolecular hydroamination of 19a–d could be completely
switched by substrate control. This novel and concise
construction of the tetrahydroisoquinoline core structure of
quinocarcin using the combination of a Sonogashira coupling
and an intramolecular hydroamination will enable the
syntheses of a variety of related tetrahydroisoquinoline
alkaloids.
[6] For reviews on gold-catalyzed hydroamination, see: a) A.
Fꢁrstner, P. W. Davies, Angew. Chem. 2007, 119, 3478; Angew.
Chem. Int. Ed. 2007, 46, 3410; b) D. J. Gorin, B. D. Sherry, F. D.
Toste, Chem. Rev. 2008, 108, 3351; c) S. F. Kirsch, Synthesis 2008,
3183; d) A. Das, S. M. A. Sohel, R.-S. Liu, Org. Biomol. Chem.
2010, 8, 960; e) N. Krause, C. Winter, Chem. Rev. 2011, 111, 1994.
For our selected publications on alkyne hydroamination, see:
f) H. Ohno, Y. Ohta, S. Oishi, N. Fujii, Angew. Chem. 2007, 119,
2345; Angew. Chem. Int. Ed. 2007, 46, 2295; g) Y. Ohta, H.
Chiba, S. Oishi, N. Fujii, H. Ohno, Org. Lett. 2008, 10, 3535; h) Y.
Ohta, H. Chiba, S. Oishi, N. Fujii, H. Ohno, J. Org. Chem. 2009,
74, 7052; i) Y. Ohta, Y. Kubota, T. Watabe, H. Chiba, S. Oishi, N.
Fujii, H. Ohno, J. Org. Chem. 2009, 74, 6299; j) K. Hirano, Y.
Inaba, N. Takahashi, M. Shimano, S. Oishi, N. Fujii, H. Ohno, J.
Org. Chem. 2011, 76, 1212.
[7] H. Ohno, K. Ando, H. Hamaguchi, Y. Takeoka, T. Tanaka, J.
Am. Chem. Soc. 2002, 124, 15255.
Received: June 29, 2012
[8] H. Driguez, B. Henrissat, Tetrahedron Lett. 1981, 22, 5061.
[9] O. Mitsunobu, M. Yamada, Bull. Chem. Soc. Jpn. 1967, 40, 935.
[10] a) M. Montury, J. Gorꢂ, Synth. Commun. 1980, 10, 873; b) C. J.
Elsevier, J. Meijer, G. Tadema, P. M. Stehouwer, H. J. T. Bos, P.
Vermeer, J. Org. Chem. 1982, 47, 2194.
[11] S. Urgaonkar, J. G. Verkade, J. Org. Chem. 2004, 69, 5752.
[12] See the Supporting Information.
[13] The Z configuration of 21a was determined by NOE analysis.
[14] S. G. Hentges, K. B. Sharpless, J. Am. Chem. Soc. 1980, 102, 4263.
[15] B. Lal, B. N. Pramanik, M. S. Manhas, A. K. Bose, Tetrahedron
Lett. 1977, 18, 1977.
[16] Shielding light was necessary to the success of the reaction.
[17] H. Staudinger, J. Meyer, Helv. Chim. Acta 1919, 2, 635.
[18] M. Bartra, V. Bou, J. Garcia, F. Urpi, J. Vilarrasa, Chem.
Commun. 1988, 270.
[19] S. S. Bag, R. Kundu, M. Das, J. Org. Chem. 2011, 76, 2332.
[20] D. Zewge, A. King, S. Weissman, D. Tschaen, Tetrahedron Lett.
2004, 45, 3729.
[21] The formation of the oxazolidinium intermediate 32 was
supported by isolation of the amidine derivative 35 after
treatment of the suspension with tert-butylamine. Unfortunately,
all our attempts to convert 35 into quinocarcin were unsuccess-
ful.
Published online: && &&, &&&&
Keywords: gold · heterocycles · hydroamination ·
.
natural products · total synthesis
[1] For reviews, see: a) J. D. Scott, R. M. Williams, Chem. Rev. 2002,
102, 1669; b) P. Siengalewicz, U. Rinner, J. Mulzer, Chem. Soc.
Rev. 2008, 37, 2676.
[2] a) K. Fujimoto, T. Oka, M. Morimoto, Cancer Res. 1987, 47,
1516; b) R. Kanamaru, Y. Konishi, C. Ishioka, H. Kakuta, T.
Sato, A. Ishikawa, M. Asamura, A. Wakui, Cancer Chemother.
Pharmacol. 1988, 22, 197; c) H. Saito, T. Hirata, M. Kasai, K.
Fujimoto, T. Ashizawa, M. Morimoto, A. Sato, J. Med. Chem.
1991, 34, 1959; d) A. W. Kahsai, S. Zhu, D. J. Wardrop, W. S.
Lane, G. Fenteany, Chem. Biol. 2006, 13, 973.
[3] a) F. Tomita, K. Takahashi, K. Shimizu, J. Antibiot. 1983, 36, 463;
b) K. Takahashi, F. Tomita, J. Antibiot. 1983, 36, 468; c) F.
Tomita, K. Takahashi, T. Tamaoki, J. Antibiot. 1984, 37, 1268.
[4] For the total synthesis of (ꢁ)-quinocarcin, see: T. Fukuyama, J. J.
Nunes, J. Am. Chem. Soc. 1988, 110, 5196.
[5] For asymmetric total syntheses of (ꢀ)-quinocarcin, see: a) P.
Garner, W. B. Ho, H. Shin, J. Am. Chem. Soc. 1992, 114, 2767;
b) P. Garner, W. B. Ho, H. Shin, J. Am. Chem. Soc. 1993, 115,
10742; c) T. Katoh, M. Kirihara, Y. Nagata, Y. Kobayashi, K.
Arai, J. Minami, S. Terashima, Tetrahedron 1994, 50, 6239; d) S.
Kwon, A. G. Myers, J. Am. Chem. Soc. 2005, 127, 16796; e) Y.-C.
Wu, M. Liron, J. Zhu, J. Am. Chem. Soc. 2008, 130, 7148; f) K. M.
Allan, B. M. Stoltz, J. Am. Chem. Soc. 2008, 130, 17270.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Natural Products
H. Chiba, S. Oishi, N. Fujii,*
H. Ohno*
&&&& — &&&&
Total Synthesis of (ꢀ)-Quinocarcin by
Gold(I)-Catalyzed Regioselective
Hydroamination
In control: The novel and enantioselective
total synthesis of (ꢀ)-quinocarcin
strate-controlled 6-endo-dig intramolecu-
lar alkyne hydroamination with a cationic
AuI catalyst, and Lewis-acid-mediated
ring-opening/halogenation sequence.
includes the highly stereoselective prep-
aration of the 2,5-cis-pyrrolidine by intra-
molecular amination, a selective sub-
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!