Unnatural natural triterpenes produced by altering isoleucine into alanine at position 261 in hopene synthase and the importance of having the appropriate bulk size at this position for directing the stereochemical destiny during the polycyclization cascade

Article abstract of DOI:10.1039/b000711k

Incubation of squalene with the site-directed mutant of Ile261Ala of squalene-hopene cyclase from Alicyclobacillus acidocaldarius afforded multiple triterpenes consisting of tri-, tetra- and penta-cyclic skeletons, together with a final product of hopene, among which the previously unknown 6161616-fused tetracyclic skeletal compounds, denoted pro-hopene A and B, are included.

Full text of DOI:10.1039/b000711k

Unnatural natural triterpenes produced by altering isoleucine into alanine at
position 261 in hopene synthase and the importance of having the appropriate
bulk size at this position for directing the stereochemical destiny during the
polycyclization cascade†
Tsutomu Hoshino,* Takamasa Abe and Masanori Kouda
Department of Applied Biological Chemistry, Faculty of Agriculture and Graduate School of Science and
Technology, Niigata University, Ikarashi, Niigata 950-2181, Japan. E-mail: hoshitsu@agr.niigata-u.ac.jp
Received (in Cambridge, UK) 24th January 2000, Accepted 8th February 2000,
Published on the Web, 29th February 2000
Incubation of squalene with the site-directed mutant of
Ile261Ala of squalene-hopene cyclase from Alicyclobacillus
acidocaldarius afforded multiple triterpenes consisting of
tri-, tetra- and penta-cyclic skeletons, together with a final
product of hopene, among which the previously unknown
6/6/6/6-fused tetracyclic skeletal compounds, denoted pro-
hopene A and B, are included.
tetra-cyclic 8;2a,b,4b,c these discrete cationic intermediates were
produced by a single amino acid replacement of D377N
(D377C), F365A, Y420A, F601A, W169F, W169 H or W489F
in A. acidocaldarius SHC (Scheme 1). The abortive cyclization
products formed by each point mutation have revealed not only
the catalytic function of given amino acids but also the
cyclization mechanism for 1; the cyclization is triggered by the
DXDDTA motif;3 D377 stabilises monocyclic cation 4 proba-
bly via the carboxylate anion; and both F3654a and F6012b
stabilise the transient carbocations at each cyclization stage 5 ?
8 possibly via a cation–p interaction. The aromatic residues of
W169, W489 and Y420 are likely to play a significant role for
folding the substrate 1.2a,4c In a series of the site-directed
mutagenesis experiments, we have focused on Ile261 as a next
experiment, which is highly conserved among all the known
SHCs and oxidosqualene cyclases. The mutated SHCs of I261A
produced the unknown 6/6/6/6-fused tetracyclic 14 and 15
(‘unnatural’ natural products) together with the known 11, 12,
13 and 16.5 The stereochemistry at 13- and 17-positions of
premature products 11 and 12 was of a-configuration and in
contrast to those of true intermediates 6 and 8. We report here,
that the appropriate bulk size and shape at position 261 has a
critical role in directing the steric control for constructing the
hopane skeleton.
Squalene-hopene cyclase (SHC) catalyses the cyclization
reaction of squalene 1 into pentacyclic triterpenes, hopene 2 and
hopanol 3, under a fine regio- and stereo-chemical specificity.
The reaction has been believed to proceed through all pre-chair
conformation,1 but recent studies on the site-directed mutagene-
sis have indicated that the reaction proceeds via two expansion
steps from the 5-membered C- and D-ring intermediates, which
had been previously formed under Markovnikov closure, into
the corresponding 6-membered rings (anti-Markovnikov ad-
duct), as shown in Scheme 1.2a,b The ring-expansion processes
have been further supported by trapping experiments of the
6/6/5-fused tricyclic intermediate 6 and the 6/6/6/5-fused
tetracyclic intermediate 8 (Markovnikov adducts) using the
substrate analogues having a highly nucleophilic hydroxyl
group.2a,b The stereochemistry of 13H in 6 and 17H in 8 have
been established to be of b-configuration; no enzymic products
having the a-configurations have been obtained in the trapping
experiments. Recently, we and Poralla’s group have reported
the isolation of the enzymatic products derived from the
carbocation intermediates of mono- 4,3 bi- 5,4a,c tri- 6,2b,4c and
With the cell-free homogenates (150 ml), prepared from a 3
L culture of E. coli clone encoding I261A SHC, 100 mg of 1 was
incubated for 15 h at optimum catalytic conditions (pH 6 and
55 °C). The GC analysis (30 m DB1, capillary column) showed
six new enzymic products other than 2 in a hexane-extract of the
reaction mixture.
Column chromatography over SiO2 eluting with hexane
afforded five fractions in the following order: 16, 2 + 13, 12 +
14 + 15, 11 and the recovered 1. The complete separation of
each product was achieved by reverse-phase HPLC (C18) with
THF–H2O (6:4). The yields of each product, estimated by GC,
were as follows: 4.0, 19.0, 4.1, 4.1, 3.7, 3.2, 48.6 and 3.1 mg for
11, 12, 13, 14, 15, 16, 2 and 3, respectively, and 0.6 mg for the
recovered 1. No other products were observed in a detectable
amount.
Structures of all the isolated compounds were determined as
shown in Fig. 1 by EIMS and NMR spectra (H–H COSY 45,
HOHAHA, NOESY, DEPT, HMQC and HMBC).† The EIMS,
1H and 13C NMR spectra of 11 were identical to those of the
authentic (17E)-(13aH)-malabarica-14(27),17,21-triene,2b the
stereochemistry of H13 being further confirmed to be of a-
configuration by a strong NOE between H9 and H13. The
2b
epimer of 11, (13bH)-malabaricatriene 11A with a longer
retention time (rt) on the GC than 11, was never detected in the
reaction mixture. The fragmentation pattern of 12 in the EIMS
spectrum was superimposable to that of 17-epi-dammara-
20(21),24-diene 12A,2a but the rt of 12 on the GC is shorter than
that of 12A, and the NOE between 30-Me and H17 was found,
Scheme 1
thus proving the a-orientation of H17 in 12. The double bond of
† Electronic supplementary information (ESI) available: EIMS and NMR
assignments. See http://www.rsc.org/suppdata/cc/b0/b000711k/
13,17
13 was D
by HMBC correlation both from 30-Me to C13
DOI: 10.1039/b000711k
Chem. Commun., 2000, 441–442
This journal is © The Royal Society of Chemistry 2000
441

subsequent cyclizations to yield 2. Therefore, it is likely that the
(13bH)- and (17bH)-configurations in 6/6/5- and 6/6/6/5-ring
systems are required for the completion of two ring-expansion
processes to give 9. This hypothesis agrees well with the
previous results2a,b that only 6 and 8 were trapped when the
squalene analogues having a hydroxy group were incubated
with the native SHC, but that 6A and 8A were not. The secondary
cation 9 has been assumed as an intermediate in hopene
biosynthesis, but no experimental evidence has been given.
Formation of 14 and 15 gave definitive evidence for the
involvement of intermediate 9. The 1,2-shift of 28-Me to the
C17 cation could give 14 (path a in Scheme 2), whereas
rearrangement of the isoprenoid side chain to the cation could
afford 15 (path b). Triterpenes 14 and 15 have never been
reported before. We propose to denote the prohopane skeleton
for 9, prohopene A for 14 and prohopene B for 15. It is
noteworthy that 8A is the biosynthetic intermediate of some plant
triterpenes.1 A looser binding near the C/D-ring could also
perturb the correct positioning of the D21 double bond in 9, thus
leading to the accumulation of 14 and 15; the appropriate
arrangement of the double bond could allow a further
cyclization to give 10 through the nucleophilic attack of the
double bond toward cation 9. Product 16 could be formed
perhaps due to the erroneous positioning of the deprotonation
site for introducing the D22,29 double bond of 2, which may have
also occurred concomitantly with the local change near the C/D-
rings.
With respect to the mutated I261V SHC, no abortive
cyclization product was found and the kinetic data were almost
the same as that of the wild-type SHC,7 but the mutant I261A
gave a looser affinity and a slower velocity for hopene
biosynthesis.7 Site-directed mutagenesis not only helps to
understand the fundamental issue of the reaction mechanism
(molecular recognition and catalytic function), but also gen-
erates the previously unknown ‘unnatural’ natural products,8a
as represented by 14 and 15. The rational genetic engineering of
the active and/or recognition sites is a promising tool for the
creation of novel natural products.4a,8b
Fig. 1 Structures of the enzymatic products by the mutated I261A SHC.
and from 21-Me for C17. The proton chemical shift of 21-Me
was 0.919 ppm (J 6.7 Hz) in CDCl3, which allowed the
assignment of 20R-stereochemistry.6 The EIMS spectrum of 14
was essentially the same as that of 15. The 6/6/6/6-fused
tetracyclic skeletons of 14 and 15 were revealed by NMR
analyses. An HMBC correlation between 28-Me and C13 was
observed for 15, but not for 14. A strong NOE was observed
between H16 and H19 for 15, while no NOE was seen for 14.
Product 16 differed from 2 only in the double bond position,
which was determined by the HMBC cross peaks of 28-Me/
C17, 29-Me/C21 and 30-Me/C2l.
Tricyclic 11 could be produced via 6A and two tetracyclics 12
and 13, via 8A (Scheme 2). Proton elimination from 21-Me of 8A
could give 12, but a hydride shift to C18 of 8A followed by
deprotonation of the 13H could afford 13. Formation of tri- and
tetra-cyclic skeletons 11–13 may have occurred owing to the
local change, especially near the site(s) responsible for the C/D-
ring formation; the mutated SHC replaced by a smaller bulk size
of Ala could not perfectly fit with 1.7 Compounds 11–13 were
accumulated, while the corresponding epimers 11A and 12A from
6 and 8 were not detected. This suggests that 6A and 8A cannot
undergo further cyclizations, whereas the intermediates 6 and 8,
which had been produced by this mutated SHC, could undergo
This work was supported by a Grant-in-Aid to T. H. (No.
11660104) from the Ministry of Education, Science, Sports and
Culture, Japan.
Notes and references
1 I. Abe, M. Rohmer and G. D. Prestwich, Chem. Rev., 1993, 93, 2189.
2 (a) T. Sato, T. Abe and T. Hoshino, Chem. Commun., 1998, 2617; (b) T.
Hoshino, M. Kouda, T. Abe and S. Ohashi, Biosci. Biotechnol. Biochem.,
1999, 63, 2038.
3 T. Sato and T. Hoshino, Biosci. Biotechnol. Biochem., 1999, 63, 2189.
4 (a) T. Hoshino and T. Sato, Chem. Commun., 1999, 2205; (b) T.
Merkofer, C. Pale-Grosdemange, M. Rohmer and K. Poralla, Tetra-
hedron Lett., 1999, 40, 2121; (c) C. Pale-Grosdemange, T. Merkofer, M.
Rohmer and K. Poralla, Tetrahedron Lett., 1999, 40, 6009.
5 For 11; K. Masuda, K. Shiojima and H. Ageta, Chem. Pharm. Bull., 1989,
34, 1140; for 12; H. Yamashita, K. Masuda, T. Kobayashi, H. Ageta and
K. Shiojima, Phytochemistry, 1998, 49, 2461; for 13; Y. Arai, K. Masuda
and H. Ageta, Chem. Pharm. Bull., 1982, 30, 4219; for 16, H. Ageta, K.
Shiojima and Y. Arai, Chem. Pharm. Bull., 1987, 35, 2705.
6 I. Abe and M. Rohmer, J. Chem. Soc., Perkin Trans 1, 1994, 783.
7 Optimum catalysis was found at 55 °C for the mutants I261A, but at 60
°C for the wild-type and I261V. Kinetic values of Km and Vmax were
determined at 55 °C and pH 6.0 from Lineweaver–Burk plots as follows;
Km: 25.8, 34.7 and 178 mM; Vmax: 3.87, 3.57 and 0.43 nmol min21 mg21
,
respectively, for the wild-type, I261V, and I261A. The I261G mutant,
having the smallest size, also produced 11–13 in twice the quantity than
I261A, despite the cyclase activity being significantly decreased (the
amount of 2 was 17% of that of I261A), which further supports the
importance of a bulk size at 261 position for the stereoselective
cyclization reactions.
8 (a) Examples of ‘unnatural’ natural products by altering the active sites
of sesquiterpenes or polyketide synthases. D. E. Cane and Q. Xue, J. Am.
Chem. Soc., 1996, 118, 1563; R. McDaniel, A. Thamchaipenet, C.
Gustafsson, H. Fu, M. Betlach, M. Betlach and G. Ashley, Proc. Natl.
Acad. Sci., USA, 1999, 96, 1846; (b) C. A. Roessner and A. I. Scott,
Chem. Biol., 1996, 3, 325.
Scheme 2
Communication b000711k
442
Chem. Commun., 2000, 441–442