Total Synthesis of the Posttranslationally Modified Polyazole Peptide Antibiotic Plantazolicin A

Article abstract of DOI:10.1002/anie.201507062

The power of rhodium-carbene methodology in chemistry is demonstrated by the synthesis of a structurally complex polyazole antibiotic. Plantazolicin A, a novel soil-bacterium metabolite, comprises a linear array of 10 five-membered rings in two pentacycli

Full text of DOI:10.1002/anie.201507062

Angewandte
Chemie
International Edition: DOI: 10.1002/anie.201507062
German Edition: DOI: 10.1002/ange.201507062
Natural Products Synthesis
Total Synthesis of the Posttranslationally Modified Polyazole Peptide
Antibiotic Plantazolicin A
Hiroki Wada, Huw E. L. Williams, and Christopher J. Moody*
Abstract: The power of rhodium–carbene methodology in
chemistry is demonstrated by the synthesis of a structurally
complex polyazole antibiotic. Plantazolicin A, a novel soil-
bacterium metabolite, comprises a linear array of 10 five-
membered rings in two pentacyclic regions that derive from
ribosomal peptide synthesis followed by extensive posttransla-
tional modification. The compound possesses potent anti-
microbial activity, and is selectively active against the anthrax-
causing organism. A conceptually different synthesis of
plantazolicin A is reported in which the key steps are the use
of rhodium(II)-catalyzed reactions of diazocarbonyl com-
Scheme 1. Synthesis of oxazoles from carboxamides and nitriles via
rhodium carbenes.
pounds to generate up to six of the seven oxazole rings of the
antibiotic. NMR spectroscopic studies and molecular modeling
reveal a likely dynamic hairpin conformation with a hinge
region around the two isoleucine residues. The compound has
modest activity against methicillin-resistant Staphylococcus
aureus (MRSA).
array of five-membered rings (azoles) that derive biosynthet-
ically from amino acids by ribosomal peptide synthesis
followed by wide-ranging posttranslational modification.[13–15]
However, it is the potent antimicrobial activity of plantazo-
licin A against Gram-positive organisms that has attracted
much attention. In particular, the compound is selectively
active against the anthrax-causing organism Bacillus anthracis
(Sterne).[11,13] Although the polyazole nature of plantazoli-
cin A is highly reminiscent of the thiopeptide antibiotics,[16–18]
a synthetic derivative of which has entered the clinic against
Clostridium difficile infections,[19,20] there are key differences.
Specifically, the linear nature of the antibiotic with its two
pentacyclic regions represents a challenge for chemical
synthesis that, in combination with the antimicrobial activity,
makes plantazolicin A highly attractive for further study. The
first total synthesis of plantazolicin A was reported by
Süssmuth and co-workers in 2013,[21] and very recently,
a second total synthesis was reported by Ley and co-work-
ers.[22] Both syntheses relied on classical peptide coupling. We
now report a conceptually different synthesis of plantazoli-
cin A on the basis of carbene chemistry.
I
n the one and a half centuries since August KekulØ and
Archibald Scott Couper independently proposed that
a carbon atom could form four bonds to other atoms
(including other carbon atoms),[1,2] the existence of divalent
carbon species with a six-electron valence shell has intrigued
chemists. Once regarded as mechanistic curiosities or fleeting
intermediates, such divalent species, now known as carbenes,
have moved center stage as a result of the isolation of the first
stable carbene in the late 1980s,[3] to be followed by the now
familiar stable N-heterocyclic carbenes,[4] which as ligands
have revolutionized transition-metal catalysis.[5] Also metallo-
carbene intermediates, derived from diazo compounds, par-
ticipate in a plethora of reactions that are useful in chemical
synthesis.[6] We now demonstrate the power of carbene
chemistry in the synthesis of the structurally unique, complex
polyazole antibiotic plantazolicin A, in which up to six of the
seven five-membered oxazole rings of the natural product are
formed from simple precursors, such as carboxamides or
Our strategy was to construct the azole rings by using
carbene intermediates, and hence our retrosynthetic analysis
divided the molecule into two fragments, 3 and 4, each
adorned with appropriate protecting groups, with the sensi-
tive oxazoline ring in the C-terminus right-hand fragment 4 to
nitriles,
as
facilitated
by
carbene
methodology
(Scheme 1).[7–10]
Plantazolicin A (1) and plantazolicin B (2) are novel
metabolites isolated from the soil bacterium Bacillus amylo-
liquefaciens FZB42.[11,12] The structures consist of a linear
be formed by
a late-stage cyclodehydration reaction
(Scheme 2). In contrast with other approaches, we elected
to introduce the guanidine moiety later in the synthesis. Thus,
the starting point for our synthesis was the known ornithine-
derived thiazole-4-ester 5,[23] which was readily converted into
the corresponding carboxamide 6 to set the scene for our first
step involving a carbene (Scheme 3). The key metallocarbene
[*] H. Wada, Dr. H. E. L. Williams, Prof. C. J. Moody
School of Chemistry, University of Nottingham
Nottingham NG7 2RD (UK)
E-mail: c.j.moody@nottingham.ac.uk
À
N H insertion was carried out by heating a mixture of methyl
Supporting information for this article, including experimental
procedures, characterization data for all compounds, and copies of
the 1H and 13C NMR spectra, is available on the WWW under http://
dx.doi.org/10.1002/anie.201507062.
2-diazo-3-oxobutanoate and amide 6 in the presence of
rhodium(II) acetate dimer (2.5 mol%) in dichloromethane
in a microwave reactor (200 W, 808C), to give the ketoamide
Angew. Chem. Int. Ed. 2015, 54, 15147 –15151
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
15147

.
Angewandte
Communications
Scheme 2. Retrosynthetic analysis of plantazolicin A.
7, which was immediately cyclized to oxazole 8 by cyclo-
dehydration with triphenylphosphine and iodine in the
presence of a base.[24] The difficulties associated with the
isolation of the product from triphenylphosphine oxide were
largely overcome by the use of polymer-supported triphenyl-
phosphine. The ester 8 was converted into the corresponding
carboxamide 9 and then into the thiocarboxamide 10 by
treatment with the Lawesson reagent. A Hantzsch reaction
formed the next ring and delivered the three-azole array 11
(Scheme 3). With two further oxazole rings to install, we
moiety onto amine 26. To this end, we developed a new
reagent, the pyrazole carboxamidine 27, which successfully
incorporated the protected guanidine to give penta-azole 28.
Hydrolysis of the ester finally gave the desired fragment 3
(Scheme 4) in 1.2% overall yield over 17 steps.
The presence of four C5-unsubstituted oxazoles in the C-
terminal fragment of the antibiotic, in contrast to the three 5-
methyloxazoles present in the N-terminal fragment 3, neces-
sitated a slight change in methodology. Thus, the order of
steps was reversed, and the carboxamide was initially
dehydrated to the nitrile, with subsequent direct conversion
into the oxazole in the rhodium-catalyzed carbene cyclo-
addition step (Scheme 1).[10,27] Hence, the synthesis of frag-
ment 4 started with the known isoleucine–isoleucine dipep-
tide 29,[28] which was converted into the carboxamide 30 and
then the required nitrile 31 (Scheme 5). These transforma-
tions set the stage for the direct installation of the first oxazole
32 by a dirhodium(II)-catalyzed reaction with the formyl
diazo compound ethyl 2-diazo-3-oxopropanoate, although
a change in catalyst to dirhodium tetrakis(heptafluorobutyr-
amide) was beneficial.[27] The sequence was iterated to deliver
the bisoxazole ester 35 by the conversion of oxazole ester 32
into nitrile 34 and thereafter a second rhodium–carbene step.
Meanwhile, nitrile 37 also underwent rhodium-catalyzed
carbene cycloaddition to give oxazole 38,[10] which was
deprotected to give amine 39 in preparation for union with
the bisoxazole fragment. Bisoxazole ester 35 was hydrolyzed
to acid 36, and subsequently coupled to amine 39 by use of the
HBTU protocol to give adduct 40. A DAST-mediated
cyclodehydration and aromatization completed the linear
array 41 of four oxazole rings, and this ester was subsequently
hydrolyzed to the corresponding carboxylic acid 42 for
further elaboration (Scheme 5). Separately, Boc-Phe-OH
was protected as its 2-trimethylsilylethyl (TMSE) ester 43,
the Boc group was cleaved, and the ensuing amine 44 was
united with N-Boc-protected allothreonine to give dipeptide
45. The dipeptide was deprotected in acid, and the resulting
amine 46 underwent coupling with the tetra-oxazole acid 42
to give, after removal of the Boc group, the complete C-
terminal fragment 4 in 1.7% overall yield over 13 steps from
the isoleucine dipeptide 29.
À
instigated an iterative carbene carboxamide N H insertion–
cyclodehydration sequence. Thus, ester 11 was converted into
carboxamide 12, which underwent the desired carbene
insertion to give ketoamide 13, the cyclodehydration of
which gave tetra-azole 14 in 52% yield over two steps. A
second iteration via carboxamide 15 and ketoamide 16
produced the desired penta-azole 17, although the yield of
this last ring-forming step was poor.
In view of the unsatisfactory formation of the fifth and
final ring in the penta-azole 17, we sought an alternative,
while still satisfying our desire to use carbene methodology to
construct the azole rings. Thus, the protected threoninamide
À
18 underwent rhodium(II)-catalyzed N H insertion followed
by cyclodehydration to give oxazole ester 20 in good yield, the
deprotection of which resulted in oxazole 21 with a threonine-
derived free amine (Scheme 4). This amine was then coupled
with the thiazolecarboxylate formed by simple hydrolysis of
the bis(thiazolyl)oxazole 11, thus resulting in the formation of
the linear tetra-azole 22. Subsequent ring closure of the
threonine fragment by the use of DAST methodology gave
oxazoline 23,[25] the dehydrogenative aromatization of which
with bromotrichloromethane in the presence of a base[26]
yielded the previously prepared penta-azole 17.
With an improved route available to the key penta-azole,
the synthesis rapidly progressed towards the complete left-
hand fragment 3 of the antibiotic. Acid-mediated removal of
the Boc group allowed for installation of the dimethylamino
group by a reductive amination reaction with formaldehyde
to give 25, which underwent cleavage of the remaining N-
protecting group in preparation for the introduction of the 2-
trimethylsilylethoxycarbonyl (Teoc)-protected guanidine
15148 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 15147 –15151

Angewandte
Chemie
Scheme 3. Synthesis of intermediate penta-azole 17: a) 35% aqueous NH3, EtOH, room temperature, 93%; b) methyl 2-diazo-3-oxobutanoate
(1.5 equiv), rhodium(II) acetate dimer (2.5 mol%), CH2Cl2, 808C, microwave (200 W); c) polymer–Ph3P (1.6 equiv), I2 (1.6 equiv), Et3N (3.2 equiv),
CH2Cl2, room temperature, 72% (2 steps); d) 35% aqueous ammonia, MeOH, THF, room temperature, 80%; e) Lawesson reagent (0.7 equiv),
CHCl3, room temperature, 55%; f) ethyl bromopyruvate (5.0 equiv), KHCO3 (10 equiv), DME, À108C; g) trifluoroacetic anhydride (5 equiv), 2,6-
lutidine (10 equiv), DME, À108C; h) K2CO3 (5.0 equiv), EtOH, H2O, room temperature, 71% (3 steps); i) 35% aqueous NH3, EtOH, room
temperature, 84%; j) methyl 2-diazo-3-oxobutanoate (1.4 equiv), rhodium(II) acetate dimer (2.5 mol%), CHCl3, 608C; k) polymer–Ph3P
(3.5 equiv), I2 (3.5 equiv), Et3N (7.0 equiv), CH2Cl2, room temperature, 52% (2 steps); l) LiOH, MeOH/THF/H2O (1:5:5), room temperature,
93%; m) EtO2CCl, Et3N, THF, 35% aqueous NH3, room temperature, 57%; n) methyl 2-diazo-3-oxobutanoate (1.5 equiv), rhodium(II) acetate
dimer (2.5 mol%), CHCl3, 608C; o) polymer–Ph3P (4.0 equiv), I2 (4.0 equiv), Et3N (8.0 equiv), CH2Cl2, room temperature, 11% (2 steps).
Boc=tert-butoxycarbonyl, Cbz=benzyloxycarbonyl, DME=1,2-dimethoxyethane.
Scheme 4. Synthesis of left-hand penta-azole fragment 3: a) methyl 2-diazo-3-oxobutanoate (1.4 equiv), rhodium(II) acetate dimer (2.5 mol%),
CHCl3, 708C, 16 h; b) Ph3P (2.0 equiv), I2 (2.0 equiv), Et3N (4.0 equiv), CH2Cl2, room temperature, 55% (2 steps); c) 4m HCl in 1,4-dioxane, room
temperature, 6 h, quantitative; d) LiOH, MeOH/THF/H2O (1:5:5), room temperature, 15 h, 82%; e) HBTU, Et3N, CH2Cl2, room temperature,
92%; f) DAST, K2CO3, CH2Cl2, À788C; g) BrCCl3, DBU, CH2Cl2, 08C!RT, 32%; h) 4m HCl in dioxane, room temperature, quantitative;
i) NaOAc·3H2O (56 equiv), formaldehyde (37% solution in H2O, 30 equiv), NaBH3CN (23 equiv), THF, room temperature, 63%; j) HBr (33%
solution in AcOH; 100 equiv), room temperature, quantitative; k) 27 (1.4 equiv), Et3N (2.8 equiv), CHCl3, 40%; l) Me3SnOH (10 equiv), DCE,
808C, quantitative. DBU=1,8-diazabicyclo[5.4.0]undec-7-ene, DCE=1,2-dichloroethane, DAST=dimethylamino sulfur trifluoride, HBTU=
(2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate.
Angew. Chem. Int. Ed. 2015, 54, 15147 –15151
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 15149

.
Angewandte
Communications
Scheme 5. Synthesis of right-hand tetra-azole fragment 4: a) 35% aqueous ammonia, MeOH, THF, room temperature, 76%; b) DBU (5.0 equiv),
ethyl dichlorophosphate (3.0 equiv), CH2Cl2, 08C, 80%; c) ethyl 2-diazo-3-oxopropanoate (3.0 equiv), rhodium(II) perfluorobutyramide dimer
(2.5 mol%), CHCl3, 608C, 53%; d) 35% aqueous ammonia, EtOH, THF, room temperature, 89%; e) DBU (5.0 equiv), ethyl dichlorophosphate
(3.0 equiv), CH2Cl2, 08C, 79%; f) ethyl 2-diazo-3-oxopropanoate (3.0 equiv), rhodium(II) perfluorobutyramide dimer (2.5 mol%), CHCl3, 608C,
59%; g) LiOH (58 equiv), EtOH/THF/H2O (1:5:5), room temperature, 65%; h) ethyl 2-diazo-3-oxopropanoate (3.0 equiv), rhodium(II) perfluoro-
butyramide dimer (2.5 mol%), CHCl3, 608C, 51%; i) 2 m HCl in ether, room temperature, quantitative; j) HBTU (1.5 equiv), 39 (1.5 equiv), Et3N
(2.0 equiv), CH2Cl2/DMF (1:1), room temperature, 76%; k) DAST (1.7 equiv), K2CO3 (5.0 equiv), CH2Cl2, À788C, quantitative; l) BrCCl3
(4.0 equiv), DBU (4.0 equiv), CH2Cl2, 08C, 59% (2 steps); m) LiOH (29 equiv), EtOH/THF/H2O (1:5:5), room temperature, 76%; n) 4m HCl in
dioxane, room temperature, 56%; o) Boc-protected allothreonine (1.0 equiv), HBTU (2.0 equiv), 44 (2.0 equiv), Et3N (4.0 equiv), CH2Cl2/DMF
(1:1), room temperature, 65%; p) 4m HCl in 1,4-dioxane, room temperature, 79%; q) HBTU (1.5 equiv), 46 (1.5 equiv), Et3N (3.0 equiv), CH2Cl2/
DMF (1:1), room temperature, 58%; r) 4m HCl in 1,4-dioxane, room temperature, quantitative; s) HBTU (1.8 equiv), 3, Et3N (3.7 equiv), CH2Cl2,
room temperature, 50%; t) DAST (30.0 equiv), CH2Cl2, À788C, 52%; u) TASF (30.0 equiv), DMSO, room temperature; v) HFIP, room temper-
ature, 31%. DMF=N,N-dimethylformamide, DMSO=dimethyl sulfoxide, TMSE=2-trimethylsilylethyl, TASF=tris(dimethylamino)sulfonium
difluorotrimethylsilicate, HFIP=hexafluoroisopropanol.
Having successfully obtained both the N- and C-terminal
halves of the antibiotic, we coupled polyazoles 3 and 4 in the
presence of HBTU and Et3N to give 48. The threonine moiety
of the intermediate 48 was cyclized in the presence of DAST
to afford the oxazoline 49, the 1H NMR and 13C NMR
spectroscopic data of which matched those of an intermediate
prepared in an earlier synthesis.[21] All that remained was to
remove the three silicon-based protecting groups, and in
common with previous workers, we found that the removal of
these groups was not completely straightforward, and
15150 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 15147 –15151

Angewandte
Chemie
required a two-stage strategy involving the sequential use of
two fluoride-based reagents (Scheme 5). Following purifica-
tion by HPLC, the NMR spectroscopic data of our synthetic
material matched those reported for the natural antibiotic
(see Tables S1 and S2 in the Supporting Information); the
compound also co-eluted on HPLC with authentic material
(see Figures S1–S3). Our synthetic material also exhibited
modest potency against methicillin-resistant Staphylococcus
aureus (MRSA) with a minimum inhibitory concentration
(MIC) of > 33 mgmLÀ1 (compare with the value of
> 128 mgmLÀ1 found for the natural material[13]).
Finally, we were interested in the conformation of
plantazolicin A. NMR spectroscopy NOESY and TOCSY
experiments were carried out (see the Supporting Informa-
tion) along with molecular modeling. The lack of long-range
NOEs suggests a moderately dynamic molecule with rigid
oxazole/thiazole arms that are not in close contact for any
appreciable time. However, the strong NOEs around the
central two isoleucine residues suggest that this portion of the
molecule could act as a hinge region leading to a dynamic
hairpin conformation. A structure consistent with these data
is shown in Figure 1.
Keywords: antibiotics · carbenes · diazo compounds ·
heterocycles · total synthesis
How to cite: Angew. Chem. Int. Ed. 2015, 54, 15147–15151
Angew. Chem. 2015, 127, 15362–15366
[1] A. KekulØ, Liebigs Ann. Chem. 1858, 106, 129 – 159.
[2] A. S. Couper, Ann. Chim. Phys. 1858, 53, 469 – 489.
[3] A. Igau, H. Grützmacher, A. Baceiredo, G. Bertrand, J. Am.
Chem. Soc. 1988, 110, 6463 – 6466.
[4] A. J. Arduengo, R. L. Harlow, M. Kline, J. Am. Chem. Soc. 1991,
113, 361 – 363.
[5] M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature
2014, 510, 485 – 496.
[6] M. P. Doyle, M. A. McKervey, T. Ye, Modern Catalytic Methods
for Organic Synthesis with Diazo Compounds, Wiley, New York,
1998.
[7] M. C. Bagley, K. E. Bashford, C. L. Hesketh, C. J. Moody, J. Am.
Chem. Soc. 2000, 122, 3301 – 3313.
[8] R. A. Hughes, S. P. Thompson, L. Alcaraz, C. J. Moody, J. Am.
Chem. Soc. 2005, 127, 15644 – 15651.
[9] J. Linder, C. J. Moody, Chem. Commun. 2007, 1508 – 1509.
[10] J. Linder, T. P. Garner, H. E. L. Williams, M. S. Searle, C. J.
Moody, J. Am. Chem. Soc. 2011, 133, 1044 – 1051.
[11] R. Scholz, K. J. Molohon, J. Nachtigall, J. Vater, A. L. Markley,
R. D. Süssmuth, D. A. Mitchell, R. Borriss, J. Bacteriol. 2011,
193, 215 – 224.
[12] B. Kalyon, S. E. Helaly, R. Scholz, J. Nachtigall, J. Vater, R.
Borriss, R. D. Süssmuth, Org. Lett. 2011, 13, 2996 – 2999.
[13] K. J. Molohon, J. O. Melby, J. Lee, B. S. Evans, K. L. Dunbar,
S. B. Bumpus, N. L. Kelleher, D. A. Mitchell, ACS Chem. Biol.
2011, 6, 1307 – 1313.
[14] C. D. Deane, J. O. Melby, K. J. Molohon, A. R. Susarrey, D. A.
Mitchell, ACS Chem. Biol. 2013, 8, 1998 – 2008.
[15] N. A. Piwowarska, S. Banala, H. S. Overkleeft, R. D. Suessmuth,
Chem. Commun. 2013, 49, 10703 – 10705.
[16] M. C. Bagley, J. W. Dale, E. A. Merritt, X. Xiong, Chem. Rev.
2005, 105, 685 – 714.
[17] R. A. Hughes, C. J. Moody, Angew. Chem. Int. Ed. 2007, 46,
7930 – 7954; Angew. Chem. 2007, 119, 8076 – 8101.
[18] X. Just-Baringo, F. Albericio, M. lvarez, Angew. Chem. Int. Ed.
2014, 53, 6602 – 6616; Angew. Chem. 2014, 126, 6720 – 6735.
[19] M. J. LaMarche, J. A. Leeds, A. Amaral, J. T. Brewer, S. M.
Bushell, G. Deng, J. M. Dewhurst, J. Ding, J. Dzink-Fox, G.
Gamber, A. Jain, K. Lee, L. Lee, T. Lister, D. McKenney, S.
Mullin, C. Osborne, D. Palestrant, M. A. Patane, E. M. Rann, M.
Sachdeva, J. Shao, S. Tiamfook, A. Trzasko, L. Whitehead, A.
Yifru, D. Yu, W. Yan, Q. Zhu, J. Med. Chem. 2012, 55, 2376 –
2387.
[20] J. A. Leeds, M. Sachdeva, S. Mullin, J. Dzink-Fox, M. J.
LaMarche, Antimicrob. Agents Chemother. 2012, 56, 4463 – 4465.
[21] S. Banala, P. Ensle, R. D. Süssmuth, Angew. Chem. Int. Ed. 2013,
52, 9518 – 9523; Angew. Chem. 2013, 125, 9696 – 9701.
[22] Z. E. Wilson, S. Fenner, S. V. Ley, Angew. Chem. Int. Ed. 2015,
54, 1284 – 1288; Angew. Chem. 2015, 127, 1300 – 1304.
[23] G. Pattenden, T. Thompson, Chem. Commun. 2001, 717 – 718.
[24] P. Wipf, C. P. Miller, J. Org. Chem. 1993, 58, 3604 – 3606.
[25] A. J. Phillips, Y. Uto, P. Wipf, M. J. Reno, D. R. Williams, Org.
Lett. 2000, 2, 1165 – 1168.
Figure 1. Most probable conformation of plantazolicin A on the basis
of NMR spectroscopy and molecular modeling.
The carbene-based synthesis described above is not only
a practical route that could be used to provide further
quantities of the fascinating antimicrobial agent plantazoli-
cin A, but it also makes available a wide range of fragment
structures and analogues that are not made available by
Natureꢀs biosynthetic machinery. The full biological evalua-
tion of these novel structures is under way.
Acknowledgements
[26] D. R. Williams, P. D. Lowder, Y. G. Gu, D. A. Brooks, Tetrahe-
dron Lett. 1997, 38, 331 – 334.
[27] K. J. Doyle, C. J. Moody, Synthesis 1994, 1021 – 1022.
[28] N. Umezawa, N. Matsumoto, S. Iwama, N. Kato, T. Higuchi,
Bioorg. Med. Chem. 2010, 18, 6340 – 6350.
We thank AstraZeneca for a Sten Gustafsson Scholarship to
H.W. and the University of Nottingham for financial support,
Dr. Ewan Murray and Professor Paul Williams (University of
Nottingham School of Life Sciences) for biological data, and
Dr. Zoe Wilson and Professor Steven Ley (University of
Cambridge) for a sample of plantazolicin A.
Received: July 31, 2015
Published online: October 16, 2015
Angew. Chem. Int. Ed. 2015, 54, 15147 –15151
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 15151