Chemoenzymatic synthesis of the immunoglobulin domain of tim-3 carrying a complex-type N-glycan by using a one-pot ligation

Article abstract of DOI:10.1002/anie.201303073

Ligations: The title immunoglobin (Ig) domain was chemoenzymatically synthesized. The domain was divided into three segments and N- and S-protected peptide thioesters were synthesized by solid-phase peptide synthesis. The one-pot sequential ligation was achieved by using Ag⁺-free and Ag⁺-assisted activation of the thioester. After folding and enzymatic transfer of the synthetic glycan, the desired Ig domain carrying a nonasaccharide was successfully obtained. Copyright

Full text of DOI:10.1002/anie.201303073

Angewandte
Chemie
DOI: 10.1002/anie.201303073
Glycoprotein Synthesis
Chemoenzymatic Synthesis of the Immunoglobulin Domain of Tim-3
Carrying a Complex-Type N-Glycan by Using a One-pot Ligation**
Yuya Asahina, Shigehiro Kamitori, Toshifumi Takao, Nozomu Nishi, and Hironobu Hojo*
Ligation methods enable the efficient synthesis of (glyco)-
proteins with over 100 amino acid residues. At present, the
native chemical ligation (NCL) reaction[1] is used almost
exclusively owing to its high efficiency derived from the
chemoselective coupling between the thioester of one seg-
ment with the N-terminal cysteine residue of the other
segment. This realizes the protection-free ligation of the
segments, which facilitates the preparation of the peptide
segments by solid-phase peptide synthesis (SPPS). However,
owing to the low abundance of cysteine in proteins (ca.
1.5%), the requirement of the cysteine residue at the ligation
site often limits the application of the NCL method. To
overcome this problem, thiol auxiliary groups as well as
ligations that do not require cysteine have been developed,[2]
but every method has advantages and disadvantages.
Gly-Xaa, where epimerization-free ligation can be per-
formed.
However, further improvements are still required to
increase the general utility of our method. In the recent
syntheses of (glyco)polypeptides, we used an azido group for
the amino protection of the segments.[5] However, owing to its
hydrophobic nature, the azido-protected thioester sometimes
retains a low solubility in aqueous acetonitrile solution, which
makes its purification by reversed-phase (RP) HPLC difficult.
Thus, we attempted to use a more hydrophilic N-protecting
group and selected the isonicotinyloxycarbonyl (iNoc)[6]
group; it is hydrophilic and stable during the deprotection
by trifluoroacetic acid (TFA), purification and the ligation
steps, and it can be removed by a treatment with Zn in acetic
acid.
Since 1991, we have been developing a facile method for
(glyco)protein synthesis based on the direct aminolysis of the
N- and S-protected peptide thioester by the other segment.[3]
Various (glyco)proteins have been synthesized by using the
method.[4] Originally, the ligation reaction was performed by
activation of the alkyl thioester by silver ions. We recently
found that the reaction proceeds without the addition of silver
ions when more reactive aryl thioesters are used.[4c,d] These
Ag+-promoted and Ag+-free ligation methods have the
advantage that, in principle, the ligation can be performed
at any site, which enables a flexible synthetic plan. Never-
theless, as glycine residues exist abundantly in proteins
(average abundance of glycine is about 7%), we seldom
meet difficulty in finding appropriate ligation positions for
Another issue about our method is that the sequential
ligation requires the purification of the intermediates. In
previous syntheses, sequential ligation in the C- to N-terminal
direction was carried out.[4b,c] In this route, the 9-fluorenyl-
methoxycarbonyl (Fmoc) group, used for the N-terminal
protection of the intermediate, has to be removed by
piperidine after the ligation. To perform the second ligation
reaction, piperidine has to be eliminated by purification;
otherwise the remaining piperidine would consume the acid
component. In 2006, Bang et al. realized a sequential one-pot
native chemical ligation in the reverse N-to-C direction by
using the difference in the reactivity of aryl and alkyl
thioesters.[7] Chen et al. also demonstrated the sequential
ligation in organic solvents using ortho-mercaptophenol as an
aryl thioester precursor in the synthesis of small glycopep-
tides.[8] If the reactivity of the alkyl and aryl thioester can be
regulated by silver ions, the one-pot sequential ligation can
also be realized in our method, which further facilitates the
synthesis.
[*] Y. Asahina, Prof. Dr. H. Hojo
Department of Applied Biochemistry
Institute of Glycoscience
Tim-3 is a membrane protein located on the surface of
Th-1 cells and regulates the Th-1-mediated immune
response.[9] Recently, it has been shown that galectin-9, one
of the animal lectins, induces apoptosis of the Th-1 cells by
binding to the N-glycan of the Tim-3 immunoglobulin (Ig)
domain (Figure 1) and regulates the Th-1-cell-mediated
immunity. However, the precise mechanism of action of
galectin-9 is not yet known. To precisely study the interaction
between the Tim-3 Ig domain and galectin-9, the Ig domain
Tokai University
Kanagawa 259-1292 (Japan)
E-mail: hojo@keyaki.cc.u-tokai.ac.jp
Prof. Dr. S. Kamitori, Prof. Dr. N. Nishi
Life Science Research Center, Kagawa University
Kagawa, 761-0793 (Japan)
Prof. Dr. T. Takao
Institute for Protein Research, Osaka University
Osaka 565-0871 (Japan)
[**] We thank Prof. Dr. Yoon-Sik Lee of Seoul National University and Dr.
Sun-Jong Ryoo of BeadTech (Seoul, Korea) for providing core shell
resin. This work was supported by a grant-in-aid for scientific
research from the Ministry of Education, Sports, Science and
Technology of Japan (23380065). We thank Tokai University for their
support with a grant-in-aid for high technology research.
carrying
a distinct carbohydrate structure is required.
Recently, the transfer of a carbohydrate chain using an
endoglycosidase, such as endo-b-N-acetylglucosaminidase
from Mucor hiemalis (Endo-M),[10] was applied successfully
in the synthesis of homogeneous glycoproteins.[11] We also
succeeded in the chemoenzymatic synthesis of saposin C
carrying a complex-type carbohydrate chain by using the
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201303073.
Angew. Chem. Int. Ed. 2013, 52, 9733 –9737
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9733

.
Angewandte
Communications
reaction proceeded efficiently without significant side reac-
tions and was almost complete within 12 h (Figure 2). Thus,
perfect chemoselectivity between the alkyl and aryl thioesters
was achieved in the absence of silver ions. Without isolation of
the intermediate 5, the segment 4a was added to the reaction
mixture and the second ligation was carried out in a one-pot
manner using AgCl as an activator. After 6 h, the second
reaction also proceeded efficiently to give the protected Ig
domain 6a (Figure 2).
The deprotection and folding of the glycoprotein 6a were
carried out as shown in Scheme 2. The iNoc groups were
removed using zinc dust in an aqueous solution containing
mercaptopropionic acid (MPA) and 6m Gu·HCl. The MPA
was used as acid as well as the Zn2+ scavenger. Removal of the
Acm groups was then carried out by using silver ions. These
reactions proceeded efficiently to give the completely depro-
tected product 8a in good purity. Finally, the folding and the
disulfide bond formation was carried out in the redox buffer
Figure 1. Structure of the Tim-3 immunoglobin (Ig) domain 1. The
arrows indicate the sites of the ligation. CHO denotes the N-linked
carbohydrate.
Endo-M mutant (glycosynthase).[11 h] This method was now
applied to the synthesis of the Tim-3 Ig domain carrying
a homogeneous complex-type N-glycan 1.
The synthetic route of the target glycoprotein 1 is shown
in Scheme 1. The entire sequence of the Tim-3 Ig domain (1–
107) was divided into three segments, (1–31) 2, (32–68) 3, (69–
107) 4a (or 4b in the second route), and
the sequential ligation in the N-to-C-
terminal direction was performed using
aryl and alkyl thioesters. After removal of
the protecting groups and the disulfide
bond formation, enzymatic transfer of the
complex-type sugar was performed to
obtain the desired product 1.
The synthesis of each segment was
carried out by solid-phase synthesis. The
N-terminal peptide 2 was prepared as an
aryl thioester by Fmoc SPPS using the
described N-alkylcysteine-assisted thioes-
terification method.[12] In contrast, the
intermediate alkyl thioester 3 was synthe-
sized by the fast tert-butoxycarbonyl (Boc)
SPPS[13] using a stable thioester linker.[14]
To examine the usefulness of the iNoc
group for the protection of side chain
amino groups, all the Lys residues were
introduced using Boc-Lys(iNoc)-OH. The
preparation of the glycopeptide 4a was
Scheme 1. One-pot synthesis of the Ig domain of Tim-3 (1). HOOBt=3,4-dihydro-3-hydroxy-
4-oxo-1,2,3-benzotriazine, Acm=acetamidomethyl, DIEA=diisopropylethylamine.
also performed by using Boc chemistry.
Asn76 was introduced using the O-benzyl-
protected Boc-Asn(GlcNAc)-OH to real-
ize the enzymatic transfer of the complex-type sugar to this
GlcNAc after the chain assembly.
Segments 2, 3, and 4a were soluble in aqueous acetonitrile
when supplemented by guanidine HCl (Gu·HCl) and could be
purified by RP HPLC. In a control experiment, peptide 4c
(see the Supporting Information), the form of 4a in which the
Lys side chains are carrying azido groups, was synthesized.
However, the solubility of segment 4c under the RP-HPLC
purification conditions is extremely low and it was not eluted
from the HPLC column at all (data not shown). These data
demonstrate the suitability of the iNoc group, which is the
amino protecting group in the thioester method.
Figure 2. RP HPLC profile of the ligation reaction; a,b) reaction
mixture for the synthesis of peptide 5 at t=0 (a) and t=12 h (b);
c,d) reaction mixture for the synthesis of peptide 6a at t=0 (c) and
t=6 h (d). In each case, a linear gradient starting from 25%
acetonitrile at tret =0 min to 50% acetonitrile at tret =25 min was
applied.
For the ligation following the first route, the three
segments were condensed (Scheme 1). Segments 2 and 3
were condensed by using the Ag+-free ligation method,[4c,d]
keeping the alkyl thioester of the segment 3 inactive. The
9734
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 9733 –9737

Angewandte
Chemie
and cysteamine and kept at 48C. As shown in Figure SI3 in
the Supporting Information, the peptide 8b was converted to
peptide 8a with a half-life of about 30 min.
The transglycosylation reaction was carried out according
to the synthesis of saposin C (Scheme 3).[10 h] The Ig domain
Scheme 2. Synthetic route to glycopolypeptide 9 (first route). In (a–d)
a linear gradient starting from 25% acetonitrile at tret =0 min to 50%
acetonitrile at tret =25 min was applied.
Figure 3. RP HPLC profile of the folding of glycopolypeptide 8b at t=0
(a) and t=12 h (b). In (a,b) a linear gradient starting from 25%
acetonitrile at tret =0 min to 50% acetonitrile at tret =25 min was
applied.
overnight at 48C. However, the folding efficiency to 9 was
moderate (Scheme 2d), which seemed to be due to the low
solubility of the peptide 8a.
carrying GlcNAc 9 and the synthetic sugar oxazoline 10
(50 equiv to 9), which was prepared according to the
previously described procedure,[11h] were reacted in a phos-
phate buffer containing 2m urea (pH 7.0) in the presence of
glycosynthase. The solution was kept at room temperature for
6 h. RP HPLC analysis of the reaction mixture showed the
appearance of a peak of a slightly hydrophilic compound
(Figure SI4 in the Supporting Information). The matrix-
assisted laser desorption ionization time-of-flight (MALDI-
TOF) mass analysis of this peak showed a value that
corresponds well with the theoretical one of the desired
To improve the solubility of the glycopolypeptide 8a, we
examined the introduction of an O-acylisopeptide structure
(second route), which enables an increase in the solubility of
the peptide segments and can be easily converted to the
native peptide bond under neutral conditions.[15] We success-
fully applied this method previously to the chemoenzymatic
synthesis of the hydrophobic glycoprotein saposin C by using
NCL.[10h] In the present synthesis, the isopeptide structure was
introduced to the C-terminal segment 4a, which retained the
lowest solubility among 2, 3, and 4a during the RP HPLC
purification. In contrast to the synthesis of saposin C, where
the amino group of the isopeptide moiety was free and thus
the O-to-N acyl shift occurred during the ligation, in this
synthesis the iNoc protecting group was installed at this site to
keep the isopeptide structure even after the ligation. The
preparation of peptide 4b having an isopeptide structure at
Val77-Thr78 was carried out by introducing iNoc-Thr(Fmoc-
Val)-OH during the solid-phase synthesis. The solubility of
peptide 4b was significantly higher than that of glycopeptide
4a, thereby resulting in about a four times higher yield of
isolated peptide 4b than that of 4a.
The one-pot ligation was carried out as shown in Scheme 1
(second route). The reaction also proceeded efficiently and
the glycopolypeptide 6b was successfully obtained (Fig-
ure SI1 in the Supporting Information). After removal of
the iNoc and Acm groups, the completely deprotected
peptide 8b, which still has an isopeptide structure at Val77-
Thr78 was obtained. The folding and the disulfide bond
formation were then carried out. In contrast to glycopolypep-
tide 8a, 8b easily dissolved in 6m urea. The solution was then
diluted with the redox buffer and kept at 48C overnight. The
desired product 9 was obtained in a significantly higher purity
than obtained in the first route, thus demonstrating the
importance of complete solubilization prior to the disulfide
bond formation (Figure 3). To confirm that the isopeptide
bond was converted to the amide bond, the glycopolypeptide
8b was dissolved in the same folding buffer without cystamine
Scheme 3. Enzymatic transfer of a complex-type sugar.
Angew. Chem. Int. Ed. 2013, 52, 9733 –9737
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9735

.
Angewandte
Communications
product, thus indicating the success of the synthesis (see the
Supporting Information). The transfer efficiency was about
30%.
Received: April 12, 2013
Revised: May 29, 2013
Published online: July 19, 2013
To prove the correct folding, the disulfide pairings of
product 9 were determined (see the Supporting Information).
Owing to the lack of appropriate enzymatic cleavage sites in
the Ig domain, five sequential enzyme digestions were
performed. Among the two major fragments, one proved to
have a disulfide bond at Cys29-Cys40, as shown by mass
analysis. The disulfide bond pairs in the other fragment were
determined by comparing the retention time on HPLC with
that of synthetic samples as well as by MALDI MS/MS
analysis. The results proved the disulfide bonds at Cys15-Cys87
and Cys35-Cys86.
Keywords: glycoproteins · ligations · peptide thioester ·
synthesis design
.
[1] a) P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent,
Science 1994, 266, 776 – 779; b) P. E. Dawson, S. B. H. Kent,
Annu. Rev. Biochem. 2000, 69, 923 – 960.
[2] a) L. E. Canne, S. J. Bark, S. B. H. Kent, J. Am. Chem. Soc. 1996,
118, 5891 – 5896; b) J. Offer, P. E. Dawson, Org. Lett. 2000, 2, 23 –
26; c) D. W. Low, M. G. Hill, M. R. Carrasco, S. B. H. Kent, P.
Botti, Proc. Natl. Acad. Sci. USA 2001, 98, 6554 – 6559; d) T.
Kawakami, K. Akaji, S. Aimoto, Org. Lett. 2001, 3, 1403 – 1405;
e) P. Botti, M. R. Carrasco, S. B. H. Kent, Tetrahedron Lett. 2001,
42, 1831 – 1833; f) L. Z. Yan, P. E. Dawson, J. Am. Chem. Soc.
2001, 123, 526 – 533; g) J. Offer, C. N. C. Boddy, P. E. Dawson, J.
Am. Chem. Soc. 2002, 124, 4642 – 4646; h) T. Kawakami, S.
Aimoto, Tetrahedron Lett. 2003, 44, 6059 – 6061; i) D. Macmillan,
D. W. Anderson, Org. Lett. 2004, 6, 4659 – 4662; j) B. Wu, J.
Chen, J. D. Warren, G. Chen, Z. Hua, S. J. Danishefsky, Angew.
Chem. 2006, 118, 4222 – 4231; Angew. Chem. Int. Ed. 2006, 45,
4116 – 4125; k) A. Brik, S. Ficht, Y. Y. Yang, C. S. Bennett, C. H.
Wong, J. Am. Chem. Soc. 2006, 128, 5626 – 5627; l) A. Saporito,
D. Marasco, A. Chambery, P. Botti, S. M. Monti, C. Pedone, M.
Ruvo, Biopolymers 2006, 83, 508 – 518; m) B. L. Pentelute,
S. B. H. Kent, Org. Lett. 2007, 9, 687 – 690; n) Y. Y. Yang, S.
Ficht, A. Brik, C. H. Wong, J. Am. Chem. Soc. 2007, 129, 7690 –
7701; o) D. Crich, A. Banerjee, J. Am. Chem. Soc. 2007, 129,
10064 – 10065; p) Q. Wan, S. J. Danishefsky, Angew. Chem. 2007,
119, 9408 – 9412; Angew. Chem. Int. Ed. 2007, 46, 9248 – 9252;
q) C. Chatterjee, R. K. McGinty, J. P. Pellois, T. W. Muir, Angew.
Chem. 2007, 119, 2872 – 2876; Angew. Chem. Int. Ed. 2007, 46,
2814 – 2818; r) R. Okamoto, Y. Kajihara, Angew. Chem. 2008,
120, 5482 – 5486; Angew. Chem. Int. Ed. 2008, 47, 5402 – 5406;
s) C. Haase, H. Rohde, O. Seitz, Angew. Chem. 2008, 120, 6912 –
6915; Angew. Chem. Int. Ed. 2008, 47, 6807 – 6810; t) G. L.
Thomas, R. J. Payne, Chem. Commun. 2009, 4260 – 4262; u) H.
Hojo, C. Ozawa, H. Katayama, A. Ueki, Y. Nakahara, Y.
Nakahara, Angew. Chem. 2010, 122, 5446 – 5449; Angew. Chem.
Int. Ed. 2010, 49, 5318 – 5321.
The secondary structure of the Ig domain was next
characterized by CD spectroscopy. The products 1 and 9
showed typical spectra of the Ig-like b-sheet-rich structure,
indicating the correct folding of the products (Figure 4).
~
Figure 4. CD spectra of glycoproteins 1 (&) and 9 ( ).
The immunoreactivity of product 9 was analyzed by using
a sandwich ELISA assay with anti-Tim-3 antibodies (Fig-
ure SI7 in the Supporting Information). About 60% of the
product 9 was recognized by the antibodies, thus indicating
that it is folded to form correct epitope structures. The
decreased ratio of the recognition might be due to its reduced
size in peptide and/or carbohydrate chains compared to the
entire Tim-3 protein.
In conclusion, we successfully performed the one-pot
sequential ligation using the aryl and alkyl thioesters to obtain
the Tim-3 Ig domain carrying GlcNAc. The iNoc group
proved to be favorable for the amino group protection in this
ligation. The introduction of the O-acylisopeptide structure
markedly enhanced the solubility of the unfolded glycopro-
tein, which enabled the efficient folding. The obtained
glycoprotein was transformed into its complex-type homoge-
neous glycoform by using the enzymatic sugar chain transfer
by a glycosynthase. The interaction between the Tim-3 Ig
domain and galectin-9 is under investigation.
[3] a) H. Hojo, S. Aimoto, Bull. Chem. Soc. Jpn. 1991, 64, 111 – 117;
b) S. Aimoto, Biopolymers 1999, 51, 247 – 265.
[4] a) H. Hojo, E. Haginoya, Y. Matsumoto, Y. Nakahara, K.
Nabeshima, B. P. Toole, Y. Watanabe, Tetrahedron Lett. 2003, 44,
2961 – 2964; b) H. Hojo, Y. Matsumoto, Y. Nakahara, E. Ito, Y.
Suzuki, M. Suzuki, A. Suzuki, Y. Nakahara, J. Am. Chem. Soc.
2005, 127, 13720 – 13725; c) H. Hojo, Y. Murasawa, H.
Katayama, T. Ohira, Y. Nakahara, Y. Nakahara, Org. Biomol.
Chem. 2008, 6, 1808 – 1813; d) H. Katayama, H. Hojo, I. Shimizu,
Y. Nakahara, Y. Nakahara, Org. Biomol. Chem. 2010, 8, 1966 –
1972.
[5] H. Katayama, H. Hojo, T. Ohira, Y. Nakahara, Tetrahedron Lett.
2008, 49, 5492 – 5494.
[6] D. F. Veber, W. J. Paleveda, Jr., Y. C. Lee, R. Hirschmann, J.
Org. Chem. 1977, 42, 3286 – 3288.
[7] D. Bang, B. L. Pentelute, S. B. H. Kent, Angew. Chem. 2006, 118,
4089 – 4092; Angew. Chem. Int. Ed. 2006, 45, 3985 – 3988.
[8] G. Chen, Q. Wan, Z. Tan, C. Kan, Z. Hua, K. Ranganathan, S. J.
Danishefsky, Angew. Chem. 2007, 119, 7527 – 7531; Angew.
Chem. Int. Ed. 2007, 46, 7383 – 7387.
[9] E. Cao, X. Zang, U. A. Ramagopal, A. Mukhopadhaya, A.
Fedorov, E. Fedorov, W. D. Zenchek, J. W. Lary, J. L. Cole, H.
Deng, H. Xiao, T. P. DiLorenzo, J. P. Allison, S. G. Nathenson,
S. C. Almo, Immunity 2007, 26, 311 – 321.
9736 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 9733 –9737

Angewandte
Chemie
[10] a) K. Yamamoto, S. Kadowaki, M. Fujisaki, H. Kumagai, T.
Tochikura, Biosci. Biotechnol. Biochem. 1994, 58, 72 – 77; b) K.
Yamamoto, S. Kadowaki, J. Watanabe, H. Kumagai, Biochem.
Biophys. Res. Commun. 1994, 203, 244 – 252.
Tanaka, M. Noguchi, A. Kobayashi, S. Shoda, W. Huang, L.
Wang, H. Ashida, K. Yamamoto, Biochim. Biophys. Acta Gen.
Subj. 2010, 1800, 1203 – 1209; h) H. Hojo, H. Tanaka, M.
Hagiwara, Y. Asahina, A. Ueki, H. Katayama, Y. Nakahara,
A. Yoneshige, J. Matsuda, Y. Ito, Y. Nakahara, J. Org. Chem.
2012, 77, 9437 – 9446.
[11] a) M. Mizuno, K. Haneda, R. Iguchi, I. Muramoto, T. Kawakami,
S. Aimoto, K. Yamamoto, T. Inazu, J. Am. Chem. Soc. 1999, 121,
284 – 290; b) B. Li, Y. Zeng, S. Hauser, H. Song, L. X. Wang, J.
Am. Chem. Soc. 2005, 127, 9692 – 9693; c) H. Ochiai, W. Huang,
L. Wang, J. Am. Chem. Soc. 2008, 130, 13790 – 13803; d) Y. Wei,
C. Li, W. Huang, B. Li, S. Strome, L. X. Wang, Biochemistry
2008, 47, 10294 – 10304; e) M. Umekawa, W. Huang, B. Li, K.
Fujita, H. Ashida, L. X. Wang, K. Yamamoto, J. Biol. Chem.
2008, 283, 4469 – 4479; f) W. Huang, C. Li, B. Li, M. Umekawa,
K. Yamamoto, X. Zhang, L. X. Wang, J. Am. Chem. Soc. 2009,
131, 2214 – 2223; g) M. Umekawa, T. Higashiyama, Y. Koga, T.
[12] H. Hojo, Y. Onuma, Y. Akimoto, Y. Nakahara, Y. Nakahara,
Tetrahedron Lett. 2007, 48, 25 – 28.
[13] M. Schnçlzer, P. Alewood, A. Jones, D. Alewood, S. B. H. Kent,
Int. J. Pept. Protein Res. 1992, 40, 180 – 193.
[14] H. Hojo, Y. D. Kwon, Y. Kakuta, S. Tsuda, I. Tanaka, K. Hikichi,
S. Aimoto, Bull. Chem. Soc. Jpn. 1993, 66, 2700 – 2706.
[15] Y. Sohma, M. Sasaki, Y. Hayashi, T. Kimura, Y. Kiso, Chem.
Commun. 2004, 124 – 125.
Angew. Chem. Int. Ed. 2013, 52, 9733 –9737
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9737