Enantioselective synthesis of 1-aryltetrahydroisoquinolines

Article abstract of DOI:10.1021/ol1004356

(Figure presented) 1-Aryltetrahydroisoquinolines (1-arylTHIQs) are important structural motifs in many alkaloids and biologically active compounds. Ligand 2a promotes the enantioselective addition of arylzinc reagents to 3,4-dihydroisoquinoline N-oxide to

Full text of DOI:10.1021/ol1004356

ORGANIC
LETTERS
2010
Vol. 12, No. 12
2690-2693
Enantioselective Synthesis of
1-Aryltetrahydroisoquinolines
Sa Wang, M. Burak Onaran, and Christopher T. Seto*
Department of Chemistry, Brown UniVersity, ProVidence, Rhode Island 02912
christopher_seto@brown.edu
Received February 21, 2010
ABSTRACT
1-Aryltetrahydroisoquinolines (1-arylTHIQs) are important structural motifs in many alkaloids and biologically active compounds. Ligand 2a
promotes the enantioselective addition of arylzinc reagents to 3,4-dihydroisoquinoline N-oxide to yield (S)-1-arylTHIQs in 97-99% ee. Pinacol
arylboronic esters are the optimal precursors for the arylzinc reagents. This method is applied to the enantioselective synthesis of Solifenacin.
1,2,3,4-Tetrahydroisoquinolines (THIQ) are an important
structural motif in many alkaloids, and they display signifi-
cant pharmacological properties.1 Since THIQ alkaloids often
incorporate a stereocenter at the 1-position with an attached
aryl substituent, this substructure has been the target for a
number of asymmetric syntheses.2 Several 1-arylTHIQs are
shown in Figure 1. Compounds 1a-e are antagonists of the
dopamine D1 (1c),3 NMDA (1a, b, d),4 and AMPA (1e)5
receptors that mediate dopamine and glutamate neurotrans-
(1) (a) Scott, J. D.; Williams, R. M. Chem. ReV. 2002, 102, 1669–1730.
(b) Charifson, P. S. Drug Future 1989, 14, 1179–1185.
(2) Chrzanowska, M.; Rozwadowska, M. D. Chem. ReV. 2004, 104,
3341–3370.
(3) (a) Charifson, P. S.; Wyrick, S. D.; Hoffman, A. J.; Simmons,
R. M. A.; Bowen, J. P.; Mcdougald, D. L.; Mailman, R. B. J. Med. Chem.
1988, 31, 1941–1946. (b) Minor, D. L.; Wyrick, S. D.; Charifson, P. S.;
Watts, V. J.; Nichols, D. E.; Mailman, R. B. J. Med. Chem. 1994, 37, 4317–
4328.
Figure 1. Bioactive 1-aryltetrahydroisoquinolines.
mission in the central nervous system. Overactivation of these
receptors can lead to psychiatric disorders, neuronal damage,
and cell death. Antagonism of the receptors modulates
excitatory synaptic neurotransmission and reduces neuro-
toxicity. The 1-arylTHIQ antagonists have been investigated
as anticonvulsants and neuroprotective agents for the treat-
ment of neurodegenerative diseases including ischemia,
epilepsy, Huntington’s, Alzheimer’s, and Parkinson’s dis-
eases. Other 1-arylTHIQs possess antibacterial6 and anti-
HIV activities.7 Compound 1f (Solifenacin, YM-905) is a
competitive antagonist that prevents binding of acetylcholine
(4) (a) Wong, E. H. F.; Kemp, J. A.; Priestley, T.; Knight, A. R.;
Woodruff, G. N.; Iversen, L. L. Proc. Natl. Acad. Sci. U.S.A. 1986, 83,
7104–7108. (b) Sherriffs, H. J.; Shirakawa, K.; Kelly, J. S.; Olverman, H. J.;
Kuno, A.; Okubo, M.; Butcher, S. P. Eur. J. Pharmacol. 1993, 247, 319–
324. (c) Ludwig, M.; Hoesl, C. E.; Ho¨fner, G.; Wanner, K. T. Eur. J. Med.
Chem. 2006, 41, 1003–1010.
(5) (a) Gitto, R.; Barreca, M. L.; De Luca, L.; De Sarro, G.; Ferreri, G.;
Quartarone, S.; Russo, E.; Constanti, A.; Chimirri, A. J. Med. Chem. 2003,
46, 197–200. (b) Gitto, R.; Caruso, R.; Pagano, B.; De Luca, L.; Citraro,
R.; Russo, E.; De Sarro, G.; Chimirri, A. J. Med. Chem. 2006, 49, 5618–
5622. (c) Gao, M.; Kong, D. Y.; Clearfield, A.; Zheng, Q. H. Bioorg. Med.
Chem. Lett. 2006, 16, 2229–2233. (d) Gitto, R.; Ficarra, R.; Stancanelli,
R.; Guardo, M.; De Luca, L.; Barreca, M. L.; Pagano, B.; Rotondo, A.;
Bruno, G.; Russo, E.; De Sarro, G.; Chimirri, A. Bioorg. Med. Chem. 2007,
15, 5417–5423.
10.1021/ol1004356  2010 American Chemical Society
Published on Web 05/18/2010

to the human muscarinic acetylcholine receptor and reduces
contractions of the bladder muscles.8 The succinate salt of
1f, marketed under the trade name Vesicare, is an FDA
approved treatment for overactive bladder.
Three main strategies have been developed for the enan-
tioselective syntheses of THIQs. The most common approach
involves a Bischler-Napieralski cyclization followed by
reduction of the resulting imine with a chiral hydride
reducing agent or catalytic hydrogenation in the presence of
a chiral catalyst.9 A second strategy employs asymmetric
Pictet-Spengler condensations.10 Finally, chiral THIQs have
been prepared by the introduction of nucleophiles or elec-
trophiles to the 1-position of isoquinoline derivatives.11
Herein, we report a direct and efficient method to access
chiral 1-arylTHIQs by the asymmetric addition of arylzinc
reagents to 3,4-dihydroisoquinoline N-oxide. In the optimized
procedure, we use pinacol esters of boronic acids as
precursors of the arylzinc reagents. Excellent enantioselec-
tivities (97%-99% ee) are achieved with a range of arylzinc
reagents using a chiral N-Boc ethylenediamine ligand. We
demonstrate the utility of this method by performing an
enantioselective synthesis of Solifenacin.
We selected arylboroxines as the precursors to arylzinc
reagents since they transmetallate smoothly with Et2Zn to
give predominantly the mixed arylethylzinc species (Figure
2, eq 2).12 In addition, a variety of boroxines can be prepared
by dehydrating commercially available boronic acids. The
modular, chiral ethylenediamine ligands (Figure 2, eq 1) are
derived from amino acids and incorporate tertiary amine,
amino acid side chain, and N-carbamoyl substituents that can
be tuned to optimize reactivity and stereoselectivity.11i,13
Figure 2. (1) Formation of the active catalyst L* by reaction of
ligands 2a-f with Et2Zn. (2) L* promoted enantioselective addition
of PhZnEt to 3,4-dihydroisoquinoline N-oxide.
Following deprotonation, the two nitrogen atoms provide a
bidentate coordination site for zinc. Initial studies indicated
that toluene was the best solvent for the arylation of 3,4-
dihydroisoquinoline N-oxide. We settled on a mixed solvent
of 2:1 toluene:CH2Cl2, which behaved similarly to pure
toluene but gave improved solubility of the boroxines.
Table 1 shows the effect of the ligand structure on the
yield and enantioselectivity of the reaction using phenyl-
boroxine as the arylzinc precursor. All of the reactions give
good yields of compound 3a. In terms of stereoselectivity,
ꢀ-branched amino acid side chains such as the sec-butyl
group of Ile or the isopropyl group of Val are preferred over
ligands with sterically demanding but non-ꢀ-branched side
chains such as Cys(Trt). In addition, ligands with cyclic
tertiary amines (entries 1 and 5) are preferred over noncyclic
analogues (entry 6). We selected ligand 2a for further study.
Ligand 2a promotes the addition of a variety of arylzinc
reagents, derived from the corresponding boroxines, to 3,4-
dihydroisoquinoline N-oxide to give the chiral hydroxy-
lamines 3a-h (Table 2). Yields for the reactions range from
83% to 97%. Enantioselectivities for most of the substrates
are good and range from 91% to 97% ee (entries 1-6).
However, for two of the substrates (entries 7 and 8) we
observed somewhat lower stereoselectivities. All of the other
(6) Tiwari, R. K.; Singh, D.; Singh, J.; Chhillar, A. K.; Chandra, R.;
Verma, A. K. Eur. J. Med. Chem. 2006, 41, 40–49.
(7) (a) Cheng, P.; Huang, N.; Jiang, Z. Y.; Zhang, Q.; Zheng, Y. T.;
Chen, J. J.; Zhang, X. M.; Ma, Y. B. Bioorg. Med. Chem. Lett. 2008, 18,
2475–2478. (b) Chen, K. X.; Xie, H. Y.; Li, Z. G.; Gao, J. R. Bioorg. Med.
Chem. Lett. 2008, 18, 5381–5386.
(8) (a) Naito, R.; Yonetoku, Y.; Okamoto, Y.; Toyoshima, A.; Ikeda,
K.; Takeuchi, M. J. Med. Chem. 2005, 48, 6597–6606. (b) Ohtake, A.;
Ukai, M.; Hatanaka, T.; Kobayashi, S.; Ikeda, K.; Sato, S.; Miyata, K.;
Sasamata, M. Eur. J. Pharmacol. 2004, 492, 243–250. (c) Cardozo, L.;
Lisec, M.; Millard, R.; Trip, O. V.; Kuzmin, I.; Drogendijk, T. E.; Huang,
M.; Ridder, A. M. J. Urology 2004, 172, 1919–1924. (d) Smulders, R. A.;
Krauwinkel, W. J.; Swart, P. J.; Huang, M. J. Clin. Pharmacol. 2004, 44,
1023–1033.
(9) (a) Hajipour, A. R.; Hantehzadeh, M. J. Org. Chem. 1999, 64, 8475–
8478. (b) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97–102.
(c) Lu, S. M.; Wang, Y. Q.; Han, X. W.; Zhou, Y. G. Angew. Chem., Int.
Ed. 2006, 45, 2260–2263.
(10) (a) Yamada, H.; Kawate, T.; Matsumizu, M.; Nishida, A.; Yamagu-
chi, K.; Nakagawa, M. J. Org. Chem. 1998, 63, 6348–6354. (b) Taylor,
M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126, 10558–10559. (c)
Seayad, J.; Seayad, A. M.; List, B. J. Am. Chem. Soc. 2006, 128, 1086–
1087. (d) Wanner, M. J.; van der Haas, R. N. S.; de Cuba, K. R.; van
Maarseveen, J. H.; Hiemstra, H. Angew. Chem., Int. Ed. 2007, 46, 7485–
7487.
(12) (a) Bolm, C.; Rudolph, J. J. Am. Chem. Soc. 2002, 124, 14850–
14851. (b) Liu, X. Y.; Wu, X. Y.; Chai, Z.; Wu, Y. Y.; Zhao, G.; Zhu,
S. Z. J. Org. Chem. 2005, 70, 7432–7435. (c) Wu, X. Y.; Liu, X. Y.; Zhao,
G. Tetrahedron: Asymmetry 2005, 16, 2299–2305. (d) Chai, Z.; Liu, X. Y.;
Wu, X. Y.; Zhao, G. Tetrahedron: Asymmetry 2006, 17, 2442–2447. (e)
Jimeno, C.; Sayalero, S.; Fjermestad, T.; Colet, G.; Maseras, F.; Perica`s,
M. A. Angew. Chem., Int. Ed. 2008, 47, 1098–1101. (f) Magnus, N. A.;
Anzeveno, P. B.; Coffey, D. S.; Hay, D. A.; Laurila, M. E.; Schkeryantz,
J. M.; Shaw, B. W.; Staszak, M. A. Org. Process Res. DeV. 2007, 11, 560–
567. (g) Fandrick, D. R.; Fandrick, K. R.; Reeves, J. T.; Tan, Z.; Johnson,
C. S.; Lee, H.; Song, J. J.; Yee, N. K.; Senanayake, C. H. Org. Lett. 2010,
12, 88–91. (h) Fujita, M.; Nagano, T.; Schneider, U.; Hamada, T.; Ogawa,
C.; Kobayashi, S. J. Am. Chem. Soc. 2008, 130, 2914–2915.
(11) (a) Rishel, M. J.; Amarasinghe, K. K. D.; Din, S. R.; Johnson, B. F.
J. Org. Chem. 2009, 74, 4001–4004. (b) Taylor, M. S.; Tokunaga, N. T.;
Jacobsen, E. N. Angew. Chem., Int. Ed. 2005, 44, 6700–6704. (c) Murahashi,
S.-I.; Imada, Y.; Kawakami, T.; Harada, K.; Yonemushi, Y.; Tomita, N.
J. Am. Chem. Soc. 2002, 124, 2888–2889. (d) Taylor, A. M.; Schreiber,
S. L. Org. Lett. 2006, 8, 143–146. (e) Itoh, T.; Miyazaki, M.; Fukuoka, H.;
Nagata, K.; Ohsawa, A. Org. Lett. 2006, 8, 1295–1297. (f) Ukaji, Y.;
Shimuzu, Y.; Kenmoku, Y.; Ahmed, A.; Inomata, K. Chem. Lett. 1997, 1,
59–60. (g) Ukaji, Y.; Shimizu, Y.; Kenmoku, Y.; Ahmed, A.; Inomata, K.
Bull. Chem. Soc. Jpn. 2000, 73, 447–452. (h) Ukaji, Y.; Yoshida, Y.;
Inomata, K. Tetrahedron: Asymmetry 2000, 11, 733–736. (i) Wang, S.; Seto,
C. T. Org. Lett. 2006, 8, 3979–3982.
(13) (a) Richmond, M. L.; Seto, C. T. J. Org. Chem. 2003, 68, 7505–
7508. (b) Richmond, M. L.; Sprout, C. M.; Seto, C. T. J. Org. Chem. 2005,
70, 8835–8840. (c) Sprout, C. M.; Richmond, M. L.; Seto, C. T. J. Org.
Chem. 2005, 70, 7408–7417.
Org. Lett., Vol. 12, No. 12, 2010
2691

derivatives are more soluble in nonpolar organic solvents
than boroxines but are not commonly used to generate
arylzinc species.12a,g,h Comparison of reactions using phe-
nylzinc reagents derived from phenylboroxine (Table 2, entry
1) and pinacolyl phenylboronic ester (Table 3, entry 1) shows
that they give the same yield of the product 3a, but the
reaction using the boronic ester gives a somewhat higher
enantioselectivity (92% vs 98% ee).
Entries 1-11 in Table 3 highlight the variety of N-hydroxy
1-arylTHIQs that can be prepared with excellent stereose-
lectivities (97-99% ee) starting from boronic esters. The
reaction works well with substrates that contain electron-
donating (entries 2, 5, 10, and 11) or electron-withdrawing
groups (entries 3 and 6). It also tolerates a larger naphthyl
group (entry 4) and heteroaromatic rings that contain oxygen,
sulfur, or nitrogen atoms (entries 7-9). Ether, ester, car-
bamate, and carbonate functional groups are all compatible
with the reaction conditions.
The yield for six of the eleven substrates that we examined
was 90% or better (entries 1, 3, 4, and 6-8). In contrast,
substrates with electron-rich aromatic rings typically gave
lower yields. For several of these reactions (entries 9-11)
we isolated a significant amount of N-hydroxy 1-ethylTHIQ
from the reaction mixture. This byproduct is formed by
transfer of an ethyl group from the ArZnEt reagent or excess
Et2Zn that is present in the reaction mixture. Two potential
explanations for the reduced yields we observe with electron-
rich substrates are: (1) incomplete boron/zinc transmetalation
caused by low reactivity of the boronic ester precursor with
Et2Zn or (2) slow transfer of the aryl group from the ArZnEt
reagent to the nitrone. In either case, transfer of an ethyl
group could become competitive with the desired aryl
transfer and result in formation of significant amounts of
N-hydroxy 1-ethylTHIQ.
Table 1. Effect of Ligand Structure on the Addition of PhZnEt
to 3,4-Dihydroisoquinoline N-Oxide to Yield 3aa
R2 ) side yield
chain of
ee
(%)c
entry ligand
R1
(%)b
1
2
3
4
5
6
2a
2b
2c
2d
2e
2f
-(CH2)2O(CH2)2-
-(CH2)2O(CH2)2-
-(CH2)2O(CH2)2-
-(CH2)2O(CH2)2-
-(CH2)5-
Ile
Val
92
93
94
95
90
80
92
90
86
84
89
73
Chgd
Cys(Trt)
Ile
Et
Ile
a Reaction conditions: (1) (ArBO)3 (0.47 molar equiv), Et2Zn (4.2 equiv),
60 °C, 12 h; (2) 2a-f (1.3 equiv), 2:1 toluene:CH2Cl2, -20 °C, 24 h.
Syntheses of these ligands have been reported in references 11i and 13.
b Yield of purified product. c Determined by chiral HPLC. d Cyclohexylglycine.
Table 2. Conversion of Boroxines to Arylzinc Reagents,
Followed by Enantioselective Addition to
3,4-Tetrahydroisoquinoline N-Oxide, Promoted by Ligand 2aa
entry
Ar
product
yield (%)b
ee (%)c
1
2
3
4
Ph
3a
3b
3c
3d
3e
3f
92
94
85
97
87
93
83
90
92
91
97
93
93
91
84
83
4-Me-C6H4
2-Me-C6H4
4-F-C6H4
4-Cl-C6H4
4-Br-C6H4
2-Br-C6H4
2-naphthyl
5
6
7d
3g
3h
8d
We also examined several of the substrates using 0.2 equiv
of the chiral ligand 2a. Reducing the ligand loading resulted
in modest to substantial decreases in enantioselectivity (Table
3, entries 12-16). Products 3a and 3b were still produced
with good stereoselectivity (91% and 92% ee, respectively).
However, the ee of the 4-chlorophenyl derivative 3e was
reduced to 67%. The rate of the uncatalyzed background
reaction between ArZnEt and 3,4-dihydroisoquinoline N-
oxide, which yields racemic products, can be competitive
with the rate of the ligand-promoted reaction. This back-
ground reaction erodes the enantioselectivity of the trans-
formation when the ligand is not present in stoichiometric
quantities.
The absolute configuration of 3a was determined to be
(S) by reducing it to 1-phenyl-1,2,3,4-tetrahydroisoquinoline
and comparing its specific optical rotation to a value from
the literature.15 By analogy, we believe that ligand 2a
controls addition of the other arylzinc reagents to the Si-
face of the nitrone to give products with the (S)-configuration.
Since the enantiomer of ligand 2a is readily prepared from
a Reaction conditions: (1) (ArBO)3 (0.47 molar equiv), Et2Zn (4.2 equiv),
60 °C, 12 h; (2) 2a (1.3 equiv), 2:1 toluene:CH2Cl2, -20 °C, 24 h. b Yield
of purified product. c Determined by chiral HPLC. d A small amount of
precipitate formed in these reactions.
substrates gave solutions that were homogeneous at both the
beginning and end of the reactions. In contrast, these two
reactions contained a precipitate in the product solutions. This
nonhomogeneous reaction mixture may be one factor leading
to the lower enantioselectivities (84% ee and 83% ee,
respectively) for entries 7 and 8.
Despite the encouraging results we obtained using aryl-
boroxines, we encountered several difficulties that limited
the scope of this procedure. First, it was difficult to prepare
pure boroxines by dehydration of arylboronic acids that
contained strong electron-withdrawing groups or very bulky
substituents.14 Second, a number of the arylboroxines had
low solubilities in toluene and CH2Cl2. This solubility
problem represented a significant limitation to the reaction.
These limitations prompted us to explore pinacol arylbo-
ronic esters as alternate precursors to arylzinc reagents. These
(15) Observed value for (S)-1-phenyl-1,2,3,4-tetrahydroisoquinoline:
[R]22D ) +36.1° (c ) 0.73, CH2Cl2). Literature value: [R]25D ) +12.3° (c
) 0.57, CH2Cl2). Ludwig, M.; Beer, H.; Lotter, H.; Wanner, K. T.
Tetrahedron: Asymmetry 1997, 8, 2693–2695.
(14) (a) Beckmann, J.; Dakternieks, D.; Duthie, A.; Lim, A. E. K.;
Tiekink, E. R. T. J. Organomet. Chem. 2001, 633, 149–156. (b) Tokunaga,
Y.; Ueno, H.; Shimomura, Y.; Seo, T. Heterocycles 2002, 57, 787–790.
2692
Org. Lett., Vol. 12, No. 12, 2010

Table 3. Conversion of Pinacol Boronic Esters to Arylzinc
Reagents, Followed by Enantioselective Addition to
3,4-Tetrahydroisoquinoline N-Oxide, Promoted by Ligand 2aa
Figure 3. Synthesis of Solifenacin.
require resolution of racemic 1-phenylTHIQ via its diaster-
eomeric salt with tartaric acid or separation of the diaster-
eomers of 1f by chromatography.16 Compound 3a (98% ee)
was reduced using Zn/Cu in AcOH/H2O to give (S)-1-
phenylTHIQ. Acylation of the amine with p-nitrophenyl
chloroformate gave carbamate 5 in 90% yield over two steps.
Finally, this carbamate was reacted with (R)-3-quinuclindinol,
NaH, and 18-crown-6 in refluxing toluene to give Solifenacin
in 74% yield.16
In summary, we have developed a procedure for the
enantioselective addition of arylzinc reagents to 3,4-dihy-
droisoquinoline N-oxide promoted by the chiral ligand 2a.
When boronic esters are used to generate the arylzinc
reagents, the reaction gives excellent enantioselectivities with
a broad range of aryl and heteroaryl substrates. This method
should be useful for preparing a variety of chiral 1-arylTH-
IQs.
a Reaction conditions: (1) Aryl pinacol boronic ester (1.4 equiv), Et2Zn
(4.2 equiv), 70 °C, 12 h; (2) 2a (1.3 or 0.2 equiv), 2:1 toluene:CH2Cl2,
-20 °C, 24 h. b Yield of purified product. c Determined by chiral HPLC.
d Reaction perfomed in toluene. e 1-Ethyltetrahydroisoquinoline was also
isolated from these reactions in the following yields: entry 9 (26%), entry
10 (37%), entry 11 (32%).
Supporting Information Available: Experimental pro-
cedures, spectroscopic data, and chiral HPLC traces. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL1004356
D-Ile, this method can be used to prepare either enantiomer
of 1-arylTHIQs.
(16) (a) Mealy, N.; Castan˜er, J. Drug Future 1999, 24, 871–874. (b)
Perlman, N.; Nidam, T. PCT Int. Appl. 2007, WO 2007076116 A2
20070705, 22pp.
We applied this method to an enantioselective synthesis
of Solifenacin 1f (Figure 3). Typical syntheses of this drug
Org. Lett., Vol. 12, No. 12, 2010
2693