Chivosazole A - Elucidation of the absolute and relative configuration

Article abstract of DOI:10.1002/anie.200605198

The best of both worlds: The configurational assignment of a natural product - chivosazole A (1) isolated from myxobacteria - has been achieved for the first time through the combination of structure elucidation and genetic analysis. This was achieved by

Full text of DOI:10.1002/anie.200605198

Communications
DOI: 10.1002/anie.200605198
Structure Elucidation
Chivosazole A—Elucidation of the Absolute and Relative
Configuration
Dominic Janssen, Dieter Albert, Rolf Jansen, Rolf Müller, and Markus Kalesse*
Myxobacteria are a valuable source of structurally diverse
biologically active natural products. Among the various
strains isolated from Sorangium cellulosum, strain Soce12
produces several important antibiotic or cytotoxic compounds
such as soranigicin A,[1] sorangiolides,[2] disorazoles,[3] and
chivosazoles. The chivosazoles (Table 1) form a family of 31-
dissected into three polyene segments (C2–C9, C12–C15, and
C23–C28), hydroxylated ketide segments (C20–C23 and
C29–C35), and an oxazole moiety. Except for chivosazole F
(7), all the natural variants possess a 6-desoxyglucopyranose
unit (chinovose) at C11.
The structure of chivosazole A (1) was established by
mass spectrometry and NMR studies.[4] However, chemical
synthesis of this very potent natural product was obstructed
by the lack of stereochemical information on the ten
stereocenters. Nevertheless, its remarkable biological activity
and potency combined with its complex structure prompted
us to determine the configuration of chivosazole A (1) as a
prerequisite for its chemical synthesis.
Table 1:
Herein we describe how the absolute and relative config-
uration of chivosazole has been assigned through a combina-
tion of chemical degradation, partial synthesis, NMR spec-
troscopy, and genetic analysis.[5]
At the start of our investigations we analyzed the relative
configuration of C32 and C34. An acetonide was generated in
the side chain and examined with the aid of the method
described by Rychnovsky and Evans;[6] three new 13C NMR
signals were found at d = 24.8, 25.2, and 101.8 ppm, thereby
indicating a 1,3-anti-diol relationship between C32 and C34
(Scheme 1). To determine the configuration of the adjacent
Chivosazole
R1
R2
R3
R4
1
2
3
4
5
6
7
A
Me
Me
H
Me
H
2a
2a
2a
2a
2a
2a
2b
Me
Me
Me
H
H
H
Me
Me
Me
H
Me
H
A1 (6,7-E)
B
C
D
E
H
Me
F
–
–
membered macrolides which were isolated from S. cellulosum
Soce12 at the Helmholtz Centre for Infection Research (HZI,
formerly GBF) in 1997.[4] They are active against yeasts and
filamentous fungi and they are highly cytotoxic against
mammalian cell cultures (IC50 9 ngmLꢀ1 for L929 and
HeLa). Structurally, the 31-membered macrolactone can be
Scheme 1. Formation of acetonide 8. Ts=toluene-4-sulfonyl.
stereocenters, chivosazole A (1) was protected as a TBS ether
and then degraded by reductive ozonolysis to produce
fragment 9 (Scheme 2).
NOESY experiments (Figure 1) were used to assign the
relative configuration of the four contiguous stereocenters in
fragment 9. Since the relationship of the C32–C34 diol was
[*] D. Janssen, Dr. D. Albert, Prof. Dr. M. Kalesse
Zentrum für Biomolekulare Wirkstoffe (BMWZ)
Leibniz Universität Hannover
Schneiderberg 1B, 30167Hannover (Germany)
Fax : (+49)511-762-3011
E-mail: markus.kalesse@oci.uni-hannover.de
Dr. R. Jansen
Mikrobielle Wirkstoffe
Helmholtz Zentrum für Infektionsforschung
Inhoffenstrasse 7, 38124 Braunschweig (Germany)
Prof. Dr. R. Müller
Institut für Pharmazeutische Biotechnologie
Universität des Saarlandes
Postfach 151150, 66041 Saarbrücken (Germany)
Scheme 2. a) TBSOTf, 2,6-lutidine, CH2Cl2, 30% fourfold TBS-protected
1; b) O3, CH2Cl2, ꢀ788C; c) NaBH4, CH2Cl2, 46% over 2 steps.
TBS=tert-butyldimethylsilyl.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4898
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4898 –4901

Angewandte
Chemie
We used a combination of NMR data and molecular
modeling studies to determine the configuration of the
remaining five stereocenters of the macrocycle. The decisive
NMR signal in the conformational analysis was a transannular
NOE interaction between H7 and one of the diastereotopic
protons at C21.[7] On the basis of this NOE interaction, a
conformational search was initiated in which all the substitu-
ents were omitted. Only conformations that were consistent
with the observed transannular NOE interaction were
considered further.
Figure 1. NOE contacts in fragment 9. (relevant coupling constants:
The conformational analysis was performed by a Monte
Carlo search using the Macromodel program (version 8.0).[8]
The starting structure was a simplified molecule in which all
the side chains (except Me36) and the OH groups of the
macrocycle were replaced by hydrogen atoms. Ten torsion
angles (see Figure 2) were subjected to random changes
H29-H30: 9.7Hz, H30-H31: 1.0 Hz, H31-H32: 1.8 Hz).
already established as anti, the relative configuration of the
whole fragment could be determined.
To confirm our proposed configurations and to determine
the absolute configuration we synthesized segment 9 by an
independent experiment (Scheme 3). A syn-selective Evans
Figure 2. Macrocyclic backbone of chivosazole A subjected to a Monte
Carlo (MC) conformational search. The dotted line indicates the site of
the formal ring opening.
according to the MCMM[9] (Monte Carlo Multiple Minimum)
ꢀ
algorithm. Formal ring opening at bond C30 O1 (dotted
bond in Figure 2) was simulated, and a ring closing only
accepted if the C30···O1 distance was less than 4.5 Š. The
energies of the conformations found were minimized both in a
vacuum and in chloroform (GB/SA; generalized Born surface
area solvation model[10]) by using the MMFF force field and
the Polak-Ribi›re (PR) conjugate gradient method. The
calculated conformations were accepted if the configuration
of the double bonds was retained and if the distance between
H7 and one of the diastereotopic protons at C21 was less than
4 Š.
The identified structures were further grouped together
and ranked on the basis of the lowest energy representative of
each group. The omitted substituents were then added to
these representative conformations according to the observed
3J(H,H) coupling constants.[11] In every case only one config-
uration was in agreement with the coupling information.
After further structure optimization, the overall backbone
conformations remained unchanged compared to the initial
backbone conformation and were in accord with the exper-
imental NOE data. The key structural details are summarized
in Table 2. The structure obtained from the lowest energy
conformation corresponds to the structure derived from
NMR-derived data. The calculated structural parameters
which differ from the NMR data are shown in bold.
Scheme 3. a) PMBO(NH)CCl3, CSA, CH2Cl2, RT, 91%; b) DIBAl-H,
CH2Cl2, ꢀ788C, 76%; c) (COCl)2, DMSO, NEt3, CH2Cl2, ꢀ788C!08C,
95%; d) (R)-(+)-4-benzyl-3-propionyloxazolidin-2-one, TfOBBu2, NEt3,
CH2Cl2, 85%; e) DDQ, MS (4 Š), CH2Cl2, 08C!RT, 73%; f) LiBH4,
EtOH, THF, 08C!RT, 96%; g) (COCl)2, DMSO, NEt3, CH2Cl2,
ꢀ788C!08C, 99%; h) 2-(trimethylsilyloxy)propene, BF3·Et2O, toluene,
ꢀ788C, 86%, d.r. 97:3; i) Me4NHB(OAc)3, CH3CN/AcOH (1:1),
ꢀ408C, 57%; j) TBSOTf, 2,6-lutidine CH2Cl2, 72%; k) Pd(OH)2/C, H2,
iPrOH, 74%. PMB=para-methoxybenzyl, CSA=camphorsulfonic acid,
DIBAl-H=diisobutylaluminum hydride, Tf=trifluoromethanesulfonyl,
DDQ=2,3-dichloro-5,6-dicyan-1,4-benzoquinone, MS=molecular
sieves, PMP=para-methoxyphenyl, Bn=benzyl.
aldol reaction on the aldehyde derived from Roche ester (10),
followed by an anti-Felkin-selective Mukaiyama aldol reac-
tion afforded hydoxy ketone 13. Reduction of 13 with
Me4NHB(OAc)3 selectively generated anti-diol 14. Protec-
tion of the diol as TBS ethers and removal of the acetal
established segment 9. The spectroscopic data was identical to
those of the segment derived by degradation. Comparison of
the [a]D values enabled the absolute configuration to be
determined as 29S, 3 0S, 3 1S, 3 2S, and 34R.
Finally, the so-generated structure was resubmitted to a
Monte Carlo search to find out if the presence of the
substituents would result in a new energy minimum. This time
only the geometries of the double bonds and the configu-
Angew. Chem. Int. Ed. 2007, 46, 4898 –4901ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4899

Communications
Table 2: Key atom distances determined by Monte Carlo calculations; stereoisomers which do not satisfy the NOE data are indicated in bold.
Compound 10l-11l-19d-
10d-11d-19l-
20l-22d-29l-
30l
10l-11l-19l-
20l-22d-29l-
30l
10l-11l-19d-
20l-22d-29l-
30l
10l-11l-19d-
20l-22d-29l-
30l
10d-11d-19l-
20d-22l-29d-
30d
10l-11d-19d-
20l-22d-29l-
30l
10l-11l-19l-
20d-22l-29d-
30d
20d-22l-29d-
30d
Energy
[kJmolꢀ1 [a]
579.630
604.198
626.566
637.536
591.998
610.873
603.238
642.117
]
internuclear distances [Š]
H7-H9
3.4
3.8
2.8
2.6
4.1
3.1
3.5
3.73.1
3.4
3.2
2.7
2.5
2.9
3.1
3.3
3.6
2.4
5.8
4.2
3.5
7.6
2.3
2.8
3.2
3.6
2.5
4.0
4.1
3.3
2.8
3.2
2.5
3.9
2.3
2.7
3.5
3.5
2.6
4.0
4.1
3.3
4.1
3.0
2.6
3.6
3.1
3.4
3.2
2.5
2.5
2.2
2.8
3.3
3.8
2.6
2.6
2.5
3.1
3.4
4.1
2.5
2.5
2.1
2.7
3.5
3.8
2.6
2.72.6
3.5
3.1
2.7
3.3
2.7
3.2
3.3
3.3
H7-H21a
H7-H21b
H7-H36
H7-H38
H9-H10
3.0
2.2
2.7
H9-H11
H9-H14
H9-H372.5
H10-H11
H10-H36
H11-H36
H12-H372.5
H20-H21a
H20-H21b
H20-H38
H21a-H38
H21b-H38
H22-H23
H28-H29
H28-H30
H28-H39
H29-H26
H29-H40
2.5
3.6
2.4
2.2
2.6
2.5
2.1
3.0
2.72.5
2.1
4.1
4.0
4.0
2.6
2.4
2.6
2.5
2.6
2.8
2.6
3.1
2.6
3.8
2.3
2.8
2.5
3.1
2.4
2.9
2.2
2.2
3.1
2.6
3.8
2.4
2.8
2.5
3.1
2.5
2.8
2.1
2.2
3.1
2.4
3.9
2.3
3.4
2.6
3.1
2.4
2.9
2.2
2.2
3.1
2.5
2.5
3.2
2.3
2.5
3.1
2.5
2.8
2.1
2.2
3.1
2.6
2.4
3.5
2.3
2.5
3.1
2.5
2.9
2.2
2.2
3.1
2.5
2.5
3.5
2.3
2.5
3.1
2.5
2.72.9
2.1
2.2
3.1
2.5
2.5
3.3
2.2
2.5
3.1
2.5
3.1
2.5
2.5
3.1
2.3
2.5
3.1
2.6
3.1
2.2
2.2
2.2
2.4
[a] In chloroform (GB/SA) as solvent; force field: MMFF.
rations of the stereocenters were fixed during the calculations,
while no distance restraint between H7 and H21b was
applied. Gratifyingly, the global minimum found in this
Monte Carlo search had the same rigid carbon backbone as
found in the first conformational search (Figure 3). There is,
however, one small deviation (about 2 Hz) between the
calculated coupling constant 3JH10,H11 obtained by applying the
empirical Karplus equation proposed by Haasnoot et al.[11]
and the observed coupling constant. This difference indicates
that some conformational equilibrium is probable in this
region.
Analysis of the two conformations found directly above
the global minima indicate there is some flexibility in the C7–
C11 and the C19–C22 regions. In the two structures, a
hydrogen bond is present between HO22 and either N16 or
O20 (see the Supporting Information). A fast equilibrium
between the global minimum as the predominant conforma-
tion (about 90%) and the local minima found would explain
the above discrepancy of the coupling constant. By combining
this result with the absolute configurations of 29d and 30d
determined by degradation studies, the NMR experiments
and molecular modeling studies give an absolute configura-
tion of 10l-11l-19d-20d-22l-29d-30d.
Nevertheless, the limited availability and stability of
natural chivosazole A meant that we were not able to
obtain further structural assignment by chemical degradation.
To circumvent this drawback, we turned our attention to the
information available about chivosazole biosynthesis.[14] The
respective biosynthetic gene cluster encoding a complex
hybrid of polyketide synthases (PKS) and a nonribosomal
peptide synthetase was described recently by Müller and co-
workers.[12] By using the analysis of the central region of the
keto reductase (KR) described by Reid et al.[13] and Caffrey[14]
we have been able to determine the configuration at C30,
C32, and C34. Additionally, with the configuration of these
three secondary alcohols already established by degradation,
an internal standard was available which confirmed the
validity of the amino acid analysis for the assignment of the
Figure 3. Calculated structure of chivosazole A (1).
4900 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4898 –4901

Angewandte
Chemie
remaining secondary alcohols. The presence of an aspartate
residue in the keto reductase is indicative of the formation of
an alcohol with a d configuration. Correspondingly, the lack
of this aspartic acid leads to the l configuration. Application
of this method to chivosazole A (1) confirmed the known
configurations of C30, C32, and C34 in the side chain
(Table 3). Furthermore, the genetic analysis is in agreement
with the spectroscopically derived configurational assignment
for C11, C20, and C22.
Figure 4. Absolute configuration of chivosazole A (1).
Table 3: Keto reductase regions of the polyketide synthases (amino acids
146–155) involved in chivosazole biosynthesis.
[1] a) R. Jansen, H. Irschik, H. Reichenbach, D. Schomburg, V.
Wray, G. Hꢀfle, Liebigs Ann. Chem. 1989, 111 – 119; b) R.
Jansen, H. Irschik, H. Reichenbach, D. Schomburg, V. Wray, G.
Hꢀfle, Tetrahedron Lett. 1985, 26, 6031 – 6034; c) H. Irschik, R.
Jansen, K. Gerth, G. Hꢀfle, H. Reichenbach, J. Antibiot. 1987,
40, 7 – 13 .
[2] R. Jansen, H. Irschik, H. Meyer, H. Reichenbach, V. Wray, D.
Schomburg, G. Hꢀfle, Liebigs Ann. 1995, 867 – 872.
[3] a) R. Jansen, H. Irschik, H. Reichenbach, V. Wray, G. Hꢀfle,
Liebigs Ann. Chem. 1994, 759 – 773; b) H. Irschik , R. Jansen, K.
Gerth, G. Hꢀfle, H. Reichenbach, J. Antibiot. 1995, 48, 31 – 35.
[4] R. Jansen, H. Irschik, H. Reichenbach, G. Hꢀfle, Liebigs Ann.
1997, 1725 – 1732.
[5] For related approaches, see a) R. K. Kondru, P. Wipf, D. N.
Berantan, J. Am. Chem. Soc. 1998, 120, 2204 – 2205; b) S. Kiren,
L. J. Williams, Org. Lett. 2005, 7, 2905 – 2908; c) N. Matsumori,
D. Kaneno, M. Murata, H. Nakamura, K. Tachibana, J. Org.
Chem. 1999, 64, 866 – 876; d) J. Hassfeld, F. Christophe, H.
Steinmetz, T. Carlomagno, D. Menche, Org. Lett. 2006, 8, 4751 –
4754.
[6] a) S. D. Rychnovsky, B. N. Rogers, T. I. Richardson, Acc. Chem.
Res. 1998, 31, 9 – 17; b) D. A. Evans, D. L. Rieger, J. R. Gage,
Tetrahedron Lett. 1990, 31, 7099 – 7100.
KR
Central region[a]
KR product[b]
KR1
KR2
KR3
KR4
KR5
KR6
KR7
KR8
KR10
KR11
KR12
KR13
KR14
KR15
KR16
AGVLRDGLCL
ALSYQGAPLA
ALRLEDRTID
AGLAPSSNVA
AGVLRDGLAV
AIVMRDRSLV
AGGTDATRIG
AITLADGLLA
AGEMRTSTPA
AGLIRDALIP
AFLFASEPLA
AMVLADRTLM
AGLADHERPA
AGVLRDALIP
ALVLHQRSLA
C34 (d)
C32 (l)
C30 (d)
DB
DB
DB
C22 (l)
C20 (d)
DB
DB
C11 (l)
DB
DB
DB
DB
[a] Aspartate is highlighted in gray. [b] Absolute configuration of the KR
products 1,2,3,7,8,12 in parentheses. The other KR domains lead, after
elimination of water, to double bonds (DB).[12]
We have reported here for the first time a combination of
classical chemical methods and genetic analysis to assign the
configuration of a complex natural product. For parts of the
molecule three independent methods were used for the
analysis: chemical degradation and synthesis, analysis of
NMR data in combination with computational methods, and
amino acid assignments. All three methods gave the same
configurations for the individual centers. Only the combina-
tion of these various methods provides a structure with
reliable confidence for chivosazole A (1) (Figure 4).
[7] C. R. Landis, L. L. Luck, J. M. Wright, J. Magn. Reson. Ser. B
1995, 109, 44 – 59.
[8] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M.
Lipton, C. Caufield, G. Chang, T. Hendrickson, W. C. Still, J.
Comput. Chem. 1990, 11, 440 – 467.
[9] a) G. Chang, C. Wayne, W. C. Guida, W. C. Still, J. Am. Chem.
Soc. 1989, 111, 4379 – 4386; b) M. Saunders, K. N. Houk, Y. D.
Wu, W. C. Still, M. Lipton, G. Chang, W. C. Guida, J. Am. Chem.
Soc. 1990, 112, 1419 – 1427.
[10] W. C. Still, A. Tempzyk, R. Hawley, T. Hendrickson, J. Am.
Chem. Soc. 1990, 112, 6127 – 6129.
[11] C. A. G. Haasnoot, F. A. A. M. De Leeuw, C. Altona, Tetrahe-
dron 1980, 36, 2783– 2792.
[12] O. Perlova, K. Gerth, O. Kaiser, A. Hans, R. Müller, J.
Biotechnol. 2006, 121, 174 – 191.
Received: December 22, 2006
Published online: May 15, 2007
[13] R. Reid, M. Piagentini, E. Rodriguez, G. Ashley, N. Viswana-
than, J. Carney, D. V. Santi, C. R. Hutchinson, R. McDaniel,
Biochemistry 2003, 42, 72 – 79.
Keywords: configuration determination · keto reductase ·
molecular modeling · natural products · NMR spectroscopy
.
[14] P. Caffrey, ChemBioChem 2003, 4, 654 – 657.
Angew. Chem. Int. Ed. 2007, 46, 4898 –4901ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4901