Importance of the methyl group at C(10) of squalene for hopene biosynthesis and novel carbocyclic skeletons with 6/5 + 5/5 + (6) ring system(s)

Article abstract of DOI:10.1021/ol026180u

Squalene-hopene cyclase (matrix presented) Incubation of (6E,10E,14E,18E)-2,6,10,19,23-pentamethyl-tetracosa-2,6,10,14,18,22-hexaene with Alicyclobacillus acidocaldarius hopene cyclase afforded four products having two types of carbocyclic skeletons, i.e., two hopane products and two products having an unprecedented carbocyclic skeleton of 6/5 + 5/5 +6 pentacyclic and 6/5 + 5/5 tetracyclic ring systems. The former two hopane skeletons were formed from the bioconversion of C(15)-desmethylsqualene and the latter two skeletons from that of C(10)-desmethylsqualene.

Full text of DOI:10.1021/ol026180u

ORGANIC
LETTERS
2002
Vol. 4, No. 15
2553-2556
Importance of the Methyl Group at C(10)
of Squalene for Hopene Biosynthesis
and Novel Carbocyclic Skeletons with
6/5 + 5/5 + (6) Ring System(s)
Tsutomu Hoshino* and Shumi Ohashi
Department of Applied Biological Chemistry, Faculty of Agriculture, and Graduate
School of Science and Technology, Niigata UniVersity, Ikarashi, Niigata 950-2181
hoshitsu@agr.niigata-u.ac.jp
Received May 14, 2002
ABSTRACT
Incubation of (6E,10E,14E,18E)-2,6,10,19,23-pentamethyl-tetracosa-2,6,10,14,18,22-hexaene with Alicyclobacillus acidocaldarius hopene cyclase
afforded four products having two types of carbocyclic skeletons, i.e., two hopane products and two products having an unprecedented
carbocyclic skeleton of 6/5 + 5/5 +6 pentacyclic and 6/5 + 5/5 tetracyclic ring systems. The former two hopane skeletons were formed from
the bioconversion of C(15)-desmethylsqualene and the latter two skeletons from that of C(10)-desmethylsqualene.
In the past several years, the catalytic mechanism of
squalene-hopene cyclase (SHC) from Alicyclobacillus aci-
docaldarius has been rapidly unraveled.1 This prokaryotic
cyclase converts a flexible squalene molecule 1 into penta-
cyclic hopene 2 and hopanol 3 (2:3 ) ca. 5:1) having the
6/6/6/6/5-fused ring system, in which all of the six-membered
rings have a chair structure (Scheme 1).1,2 The cyclization
mechanism is analogous to that of eukaryotic oxidosqualene
cyclases.2 We have established that the polycyclization
consists of eight reaction steps1 and that the ring expansion
reaction occurs twice for the formation of the six-membered
C- and D-rings of 2 on the basis of the successful trappings
of five-membered C- and D-ring intermediates (Markovnikov
adducts) by using site-directed mutants and substrate ana-
logues.3 Previously, we have shown by using C(23)-
desmethylsqualene that the isopropylidene moiety is indis-
pensable for the correct folding of 1 both to form the five-
membered E-ring and to initiate the polycyclization reaction;4
incubation of the norsqualene with the native SHC gave
products having a six-membered E-ring (i.e., a tetrahymanol
Scheme 1. Polycyclization Pathway of Squalene 1 into Hopene
2 and Hopanol 3 by Squalene-Hopene Cyclase
(1) (a) Review for squalene-hopene cyclase. Hoshino, T.; Sato, T. J.
Chem. Soc., Chem. Commun. 2002, 291. (b) Schmitz, S.; Fu¨ll, C.; Glaser,
T.; Albert, K.; Poralla, K. Tetrahedron Lett. 2001, 42, 883.
(2) Reviews for oxidosqualene and squalene cyclases. (a) Wendt, K. U.;
Schulz, G. E.; Corey, E. J.; Liu, D. R. Angew. Chem., Int. Ed. 2000, 39,
2812. (b) Abe, I.; Rohmer, M.; Prestwich, G. D. Chem. ReV. 1993, 93,
2189.
10.1021/ol026180u CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/27/2002

skeleton with a 6/6/6/6/6-fused A/B/C/D/E-ring system), but
afforded no enzymic product that is assumed to be formed
by starting the cyclization reaction from the methyl-deficient
side. However, the role of the methyl groups at positions5
of 6, 10, 15, and 19 for the cyclization cascade has remained
unknown. We planned to address the question of how the
methyl groups at the central part of 1 affect the polycycliza-
tion cascade. Here, we describe that when the Me at C(10)5
is missing, the polycyclization cascade is dramatically altered
to afford unprecedented cyclization products consisting of
the 6/5-fused (A/B) and the 5/5-fused (C/D) ring system, in
which the B- and C-rings are not fused. This finding strongly
indicates that the C(10)-Me5 is essential to the correct folding
of 1 for completion of the normal polycyclization reaction
leading to 2 and 3.
Norsqualene 4 lacking the methyl group at the central part
of 1, (6E,10E,14E,18E)-2,6,10,19,23-pentamethyl-tetracosa-
2,6,10,14,18,22-hexaene, was prepared according to the
synthetic scheme described in Supporting Information.
With the cell-free homogenates from 4 L of culture of the
recombinant E. coli encoding the SHC,6 50 mg of 4,
emulsified with Triton X-100, was incubated at catalytically
optimum conditions (pH 6.0, 60 °C) for 16 h. The enzymic
products were extracted with hexane. The GC and TLC
showed four enzymic products. Triton X-100 was removed
by passing the products through a short SiO2 column
(hexane). The repeated SiO2 column chromatography (5%
AgNO3, 100:0.01 hexane/EtOAc) gave 5 and 6 as an oil,
while 7 and 8 were obtained as a solid in a pure state, the
elution order being 8, 7, 6, and 5. The product amounts of
5-8 and the recovered 4 were estimated to be 16.8, 3.8,
10.5, 7.3, and 9.0 mg, respectively, from the GC analysis.
Figure 1. Structures of enzymic products 5-8.
NMR analyses. Product 7 has an isohopane skeleton,7 lacking
a methyl group at C(14).The eminent ion m/z 191 in the
EIMS spectra of 7 and 8 proved the 6/6-fused A/B-ring
(Scheme 2).8 However, the stereochemistry at C(21) of 7
was opposite to that of natural type 8; a strong NOE of H-21
with Me-27 was observed for 7, whereas no NOE was
observed between them for 8.
The detailed NMR analyses (COSY 45, HOHAHA,
NOESY, DEPT, HMQC, and HMBC) revealed the structures
of all the enzymic products (Figure 1). Products 5 and 6
have a novel skeleton having a 6/5 + 5/5 ring system, which
was unequivocally determined by HMBC and NOESY data
as shown in Figure 2. This ring system was further supported
by the fragment ion m/z 285 (∼100%) in the EIMS spectra
of 5 and 6. The strong NOEs of H-10 with Me-25 and Me-
26 verified the â-orientation for both H-10 and Me-26. A
clear NOE of H-5 with H-8 indicated that both protons were
in the R-orientation, but no NOE was observed between H-8
and H-10. H-18 of the E-ring in 6 must be arranged in an
axial disposition, because clear NOEs were observed for
H-18 with H-20 and Me-28. The stereochemistry at C(18)
of 5 and 6 has yet to be established. Products 7 and 8 had a
fused pentacyclic structure, which was established by the
Figure 2. Selected HMBC and NOE correlations of 6.
Previously, we have shown that an isopropylidene moiety
of 1 is required to start the polycyclization reaction. No
cyclization occurs when one of the two methyl groups is
missing.4 Analogue 4 has two isopropylidene moieties on
the left and right terminal sides of the squalene backbone.
Thus, 4 was cyclized not only as 10-desmethylsqualene 4a5
(3) (a) Sato, T.; Abe, T.; Hoshino, T. J. Chem. Soc., Chem. Commun.
1998, 2617. (b) Hoshino, T.; Kouda, M.; Abe, T.; Ohashi, S. Biosci.
Biotechnol. Biochem. 1999, 63, 2038.
(4) Hoshino, T.; Kondo, T. J. Chem. Soc., Chem. Commun. 1999, 731.
(5) The numbering of norsqualenes discussed in this paper is enumerated
from the direction of the cyclization of norsqualenes strats. Because analogue
4 was cyclized from the both terminal isopropylidene moieties (Scheme
2), this numbering system allows the discrimination of C(10)-4a from C(15)-
desmethylsqualenes 4b. However, both 4a and 4b should be named 15-
desmethylsqualene according to the nomenclature rule.
(7) Ageta, H.; Shiojima, K.; Suzuki, H.; Nakamura, S. Chem. Pharm.
Bull. 1993, 41, 1939.
(6) Sato, T.; Kanai, Y.; Hoshino, T. Biosci. Biotechnol. Biochem. 1998,
62, 404.
(8) Shiojima, K.; Arai, Y.; Masuda, K.; Takase, Y.; Ageta, T.; Ageta,
H. Chem. Pharm. Bull. 1992, 40, 1683.
2554
Org. Lett., Vol. 4, No. 15, 2002

Scheme 2. Polycyclization Pathway of 4 into Products 5-8
to give 5 and 6 via path a but also as 15-desmethylsqualene
produced via 11b according to the normal folding conforma-
tion, whereas 7 was produced via the incorrectly folded
structure 11a. The dramatically altered cyclization pathway
of 4a (path a), compared to that of 4b (path b), indicated
that C(10)-Me plays a more crucial role for construction of
the hopane skeleton than C(15)-Me.
4b5 to afford 7 and 8 via path b, as shown in Scheme 2. The
product ratio of 5, 6, 7, and 8 was 4.4:1:2.8:1.9, indicating
that both paths a and b had proceeded in a nearly equal ratio.
The Markovnikov cation 9 produced in path a could undergo
the 1,2-shift of hydride of H-17 to the C(18)-cation, followed
by deprotonation from H-13, resulting in the introduction
of the double bond between C(13) and C(17) (path c). H-13
and H-17 must be arranged in antiperiplanar geometry to
allow the double-bond formation. The â-orientation of Me-
26 indicates the R-arrangement of H-13, because all the
double bonds of 4 have an (E)-geometry; thus, H-17 of 9
must be in the â-orientation. The proton elimination from
Me-27 of 9 (path d) could give 10 having a methylidene
moiety, which could undergo the sequential reactions of
further cyclization, hydride shift, and deprotonation to give
6. Intermediate 10 was not trapped, indicating that this further
cyclization must be fast, which may have been catalyzed by
the acidic amino acid(s) located near the substrate entrance.9
The reaction mechanism shown in paths c and e may suggest
18S for 5 and 18R stereochemistry for 6.10 Product 8 was
Why were unprecedented carbocyclic 5 and 6 produced
from 4a? It is difficult to get a clear answer to this question
at the present time, but one possible explanation may be as
follows. The accurate cavity size or binding site to accept
C(10)-Me of 1 may be involved in the SHC; a strong
interdigitation between the C(10)-Me and its binding site may
enable 1 to adopt a six-membered chair structure for the
B-ring formation and to direct â-orientation for the C(10)-
Me during the polycyclization process (Scheme 1). However,
the less bulky hydrogen at C(10) of 4a could not precisely
interact with the binding site; in turn, the methyl group at
C(15) of 4a could be strongly captured by the binding site
intrinsic to C(10)-Me of 1 through the folding conformation
shown in path a. Thus, the C(15)-Me could have a â-orienta-
tion, in contrast to the R-orientation for the bioconversion
of 1f2 (Scheme 1). On the other hand, 4b having C(10)-
Me could adopt the normal folding structure through the
correct fitting of the C(10)-Me with the binding site inherent
to the C(10)-Me of 1 (path f), resulting in the production of
hopene homologue 8. However, substitution with a less bulky
hydrogen at C(15) could influence to some extent the normal
folding around the E-ring formation site to give 7 (path e).
(9) There may exist the possibility that DXDDTA motif, which is
responsible for the initiation of the polycyclization (Sato, T.; Hoshino, T.
Biosci. Biotechnol. Biochem. 1999, 63, 2189), could also participate in the
cyclization of intermediate 10. However, this possibility is unlikely, because
10 must again have access to the acidic motif to produce 6 after exiting the
cyclase cavity. In this case, 10 should be detected in the reaction mixture.
(10) To gain the opposite stereochemistry (18R for 5 and 18S for 6), a
large rotation of the side chain through the C(17)-C(18) axis of 9 and 10
must occur prior to the hydride shift of C(17) to C(18). This large motion
is unlikely inside the enzyme cavity. Through the biosynthetic studies of
lanosterol, Corey et al. have proposed that the side chain at C(17) of
protosteryl cation must have a â-orientation to get the 20R stereochemistry
of the lanosterol skeleton, in contrast to Cornforth’s hypothesis for the 17R-
orientation (Cornforth, J. W. Angew. Chem., Int. Ed. Engl. 1968, 7, 903).2b
Corey’s idea is based upon the successful trappings of the products having
the 17â-oriented side chain by using substrate analogues, and they explained
the reason for the 17â-orientation: 20R stereochemistry can be readily
obtained by the least motion (<60°) for the 17â-oriented side inside the
enzyme cavity, whereas the 17R-oriented one must have the large rotation
of 120° (Corey, E. J.; Virgil, S. C. J. Am. Chem. Soc. 1991, 113, 4025).2b
Thus, we assign 18S and 18R for 5 and 6, respectively, due to the least
motion of the side chain. To gain the opposite stereochemistry at C(10)
and C(14) in 5 and 6, a large rotation (180°) is also required around the
C(8)-C(10) axis during the polycyclization; thus, this possibility also is
unlikely. Furthermore, the lifetime of the secondary C(10)-cation formed
during the polycyclization is short, in contrast to stable tertiary C(18)-cation
in 9, because no abortive cyclization product having a bicyclic 6/5-fused
ring system, which is generated from C(10)-cation, was trapped. Thus, the
formation of the C-ring must be fast, and the epimerization at C(10) would
not occur. The stereochemistry at C(18) and C(10) in 5 and 6 must be further
ascertained by X-ray analysis to validate our assumption.
Org. Lett., Vol. 4, No. 15, 2002
2555

Corey et al. have shown that the C(10)-Me of (3S)-2,3-
oxidosqualene is important for the normal polycyclization
pathway in lanosterol biosynthesis. Lack of the two methyl
groups at both C(10) and C(15) led to the abnormal
cyclization product having a tricyclic 6/6/5-fused A/B/C-
ring system, in which the B-ring had a chair structure, despite
lanosterol synthase adopting a twist-boat form for the B-ring
formation.11 The absence of the C(15)-Me had no influence
on the polycyclization pathway leading to the formation of
a lanosterol homolog.12 It was suggested that the hydrophobic
group responsible for the interaction with the C(10)-Me is
involved in lanosterol cyclase.11 Introduction of a slightly
bulkier ethyl group at C(10) into 2,3-oxidosqualene also
significantly influenced the biosynthetic pathway of lanos-
terol;13 10-ethylated-2,3-oxidosqualene was converted into
the three abortive cyclization products consisting of two
monocycles and one 6/6/5-fused tricycle in addition to the
fully cyclized lanosterol homologue.13c Therefore, the bulk
size at C(10) must be accurate to correctly interact with the
hydrophobic group of the cyclase; a hydrogen atom is small,
but an ethyl group is large.
cycloartenol synthase allowed lanosterol production.15 The
flexible and linear molecules of 1 and oxidosqualene can
adopt numerous folding structures, but the steric bulk sizes
of active site residues involved in triterpene cyclases may
well direct a specific folding conformation to the substrates
for producing a given triterpene skeleton.
In conclusion, we have shown that the methyl group at
C(10) is essential to the correct folding of 1 for hopene
biosynthesis. To the best of our knowledge, the triterpene
skeletons of 5 and 6 possessing 6/5 + 5/5 + (6) ring systems
have not hitherto been reported. The norsqualene lacking a
methyl group at C(6) or C(19) gave a hopane skeleton via
the normal polycyclization pathway (unpublished results).1a
Thus, the following three methyl sites of 1 are crucial for
the correct folding in the hopene biosynthesis: (1) the site
for initiation,4 (2) the central site at C(10), and (3) the
termination site.4 To validate the hypothesis for involvement
of the binding sites in the SHC, mutagenesis experiments
are necessary in which the steric bulk sizes of the active
site residues (hydrophobic groups) are modified.
Acknowledgment. This work was supported by Grants-
in-Aid (Nos. 13660150 and 13306009) from the Ministry
of Education, Science, Sports and Culture, Japan.
The mutation experiments of squalene and oxidosqualene
cyclases also have highlighted the appropriate steric bulk
size at the active sites; modification of the bulk sizes for the
SHC (e.g., mutations of I261fA and L607fW) led to the
creation of unnatural natural products1,14 and that for the
Supporting Information Available: Synthetic scheme
of 4, incubation conditions, EIMS, NMR data, and specific
rotation of 5-8. This material is available free of charge
via the Internet at http://pubs.acs.org.
(11) Corey, E. J.; Virgil, S. C.; Liu, D. R.; Sarsha, S. J. Am. Chem. Soc.
1992, 114, 1524.
(12) van Tamlen, E. E. J. Am. Chem. Soc. 1968, 90, 3284.
(13) (a) Hoshino, T.; Ishibashi, E.; Kaneko, K. J. Chem. Soc., Chem.
Commun. 1995, 2401. (b) Hoshino, T.; Sakai, Y. J. Chem. Soc., Chem.
Commun. 1998, 1591. (c) Hoshino, T.; Sakai, Y. Tetrahedron Lett. 2001,
42, 7319.
(14) Hoshino, T.; Abe, T.; Kouda, M. J. Chem. Soc., Chem. Commun.
2000, 441.
OL026180U
(15) (a) Hart, E. A.; Hua, L.; Darr, L. B.; Wilson, W. K.; Pang, J.;
Matsuda, S. P. T. J. Am. Chem. Soc. 1999, 121, 9887. (b) Matsuda, S. P.
T.; Darr, L. B.; Hart, E. A.; Herrera, J. B. R.; McCann, K. E.; Meyer, M.
M.; Pang, J.; Schepmann, H. G. Org. Lett. 2000, 2, 2261.
2556
Org. Lett., Vol. 4, No. 15, 2002