The cyclization mechanism of squalene in hopene biosynthesis: The terminal methyl groups are critical to the correct folding of this substrate both for the formation of the five-membered E-ring and for the initiation of the polycyclization reaction

Article abstract of DOI:10.1039/a901351b

Incubations of C(23)-norsqualenes 5 and 6, lacking one of the two terminal methyl groups, with squalene-hopene cyclase gave unprecedented products 7 and 8 having a tetrahymanol skeleton together with a neohopane skeleton 12, strongly suggesting that the two geminal methyls of squalene 1 are critical to the formation of the five-membered E-ring in hopene biosynthesis and also are required to initiate the cyclization reactions of 1 into the pentacyclic triterpenes 2 and 3.

Full text of DOI:10.1039/a901351b

The cyclization mechanism of squalene in hopene biosynthesis: the terminal
methyl groups are critical to the correct folding of this substrate both for the
formation of the five-membered E-ring and for the initiation of the
polycyclization reaction†
Tsutomu Hoshino,*ab and Tomohiro Kondob
a
Department of Applied Biological Chemistry, Faculty of Agriculture, and b 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) 18th February 1999, Accepted 10th March 1999
Incubations of C(23)-norsqualenes 5 and 6, lacking one of
the two terminal methyl groups, with squalene-hopene
cyclase gave unprecedented products 7 and 8 having a
tetrahymanol skeleton together with a neohopane skeleton
also be responsible for tetrahymanol biosynthesis.2c Such a ring
expansion reaction was also demonstrated for the oxidosqua-
lene cyclization.3 Computational energy calculations of the
five- and six-membered intermediates have supported the
4
12, strongly suggesting that the two geminal methyls of
feasibility of such a ring expansion process. The structure of 2
squalene 1 are critical to the formation of the five-membered
E-ring in hopene biosynthesis and also are required to
initiate the cyclization reactions of 1 into the pentacyclic
triterpenes 2 and 3.
differs from that of 3 only in the terminal E-ring, i.e. five- and
six-membered rings for 2 and 3, respectively. To gain insight
into the different cyclization mechanism between SHC and
STC, the substrate analogues 5 and 6, lacking one of the two
terminal methyls in squalene backbone 1, were separately
incubated with SHC or STC, resulting in the formation of 7, 8
and 12 as SHC enzyme products, and 7 and 8 as STC products;
SHC, despite responsibility for the formation of the five-
membered E-ring, did produce a tetrahymanol skeleton having
a six-membered E-ring. We show herein that the terminal
methyl groups play a critical role for the determination of either
five- or six-membered E-ring formation.
The biocyclization of squalene 1 into hop-22(29)-ene 2
catalyzed by squalene-hopene cyclase [EC 5.4.99.-] (SHC) or
into tetrahymanol 3, by squalene-tetrahymanol cyclase [EC
5
.4.99.-] (STC), is one of the most intricate biochemical
1
reactions (Scheme 1). Compound 4, hopan-22-ol, is a minor
by-product. The cyclization proceeds via precise enzymatic
control to form the fused 6/6/6/6/5- or 6/6/6/6/6-ring system and
nine new stereocenters. The polyene cyclization reaction of 1 is
analogous to that of 2,3-oxidosqualene into lanosterol catalyzed
The analogues 5 and 6 were synthesized as follows:
(±)-2,3-oxidosqualene was treated with HIO
aldehyde, which was then subjected to a Wittig reaction with
4
to give the C27-
1
by lanosterol synthase. Recently, it has been proposed that a
n
ring expansion process from the five- to the six-membered ring
is involved in the D-ring formation of hopene biosynthesis,
EtPPh
norsqualenes 5 and 6 thus obtained were separated with a SiO
column (10% AgNO ) with hexane; 6 followed by 5. The 7
methyl signals of 5 in CDCl were as follows: d 1.66 [3H, s,
CN] and 1.58 [18H, (Z)-CH CN], while those of 6 were:
1.66 [3H, s, (E)-CH CN], 1.61[3H, brd, J 5.8, (E)-CH CN],
.58 [12H, s, (Z)-CH CN] and 1.57 [3H, s, (Z)-CH
Separate incubations of 5 and 6 with cell-free homogenates
3
Br in the presence of Bu Li in THF. The C(23)-
2
2a,b
prior to the further cyclization (Scheme 1).
Formation of the
3
five-membered intermediate D-ring is consistent with the
Markovnikov rule. The expansion process of the D-ring may
3
H
(E)-CH
H
3
3
d
3
3
5
1
3
3
CN].
H
H
D
C
Enz-AH+
from the cloned E. coli harboring SHC under catalytic optimal
A
B
6
conditions afforded highly polar products 7 and 8, respectively
(
1
Scheme 2). From the incubation mixtures of 5, another product
2 was found together with some minor products, both of which
had lower polarities than 7 on TLC. Separation of each product
was performed using a SiO column with hexane–EtOAc, but
the minor products, present in negligible amounts, were
inseparable even on AgNO –SiO TLC. Isolation yields of 7
1
STC
SHC
H
2
D
O
D
H
H
3
2
and 12 from 5 were 17 and 12%, respectively, while that of 8
from 6 was 16%; neither 12 nor the minor products were
detected from 6 (Scheme 3), using the same quantities of the
substrate and the cell-free extracts as for 5.
E
E
OH
Enz-AH+
Enz-AH+
R2
R1
3
23
2
R1
R2
E
5, 6
SHC and STC
5, 6
SHC and STC
OH
2
1
4
R1
OH
2OH
1
8
2
2R
Scheme 1
17
R1
R2
7
, 8
R1 = Me, R2 = H
R1 = H, R2 = Me
7
8
R1 = Me, R2 = H
R1 = H, R2 = Me
9
10
R
R
1
1
= Me, R = H
2
5
6
†
Details of orbital interactions and 2D NMR analyses for 7, 8 and 12 are
available from the RSC website, see: http://www.rsc.org/suppdata/cc/
999/731/
2
= H, R = Me
1
Scheme 2
Chem. Commun., 1999, 731–732
731

H
1
is more important than the (23E)-methyl for the formation of
the hopanyl cation 11 with the 5-membered E-ring. However,
the two terminal methyls would be necessary for the complete
building of the five-membered E-ring, since the six-membered
species 7 and 8 were produced in significant amounts when one
of the two methyls was absent. Why was the tetrahymanol
skeleton produced by the incubations of the norsqualene 5 and
H
H
E
H
2
1
12
2
2
H
1
1a
H
E
H
H
6
with SHC? One plausible answer may be that the two geminal
5
SHC
H
H
13
14
2
1
2
H
methyls strongly bind to SHC to acquire the desired conforma-
tion, shown in Scheme 1, during the formation of the five-
membered E-ring, and the substrate affinity would become
looser when one of the terminal methyls is absent, which would
have led to the formation of a tetrahymanol skeleton under
stereoelectronic control. The binding force of the Z-methyl to
the SHC would be stronger than the corresponding E-methyl,
because 12 was produced only from 5, and not from 6. The
absence of 13 from 5, although the hopanyl cation 11 has been
produced, suggests that the migration of the hydride (1, 2-shift)
must be fast compared with the two deprotonation reactions for
the formations of 13 and 14. This would be due to a greater
stability of the tertiary C21-carbocation intermediate 11b
compared to the secondary C22-cation 11a. The hopanol
analogue 15 was also not found. At the present time, we cannot
propose which terminal methyl of either the Z- or E-isomers of
the natural 1 is responsible for the proton elimination when the
double bond in 2 is introduced. Compounds 9, 10, 16, 17 and 18
were also not detected in any reaction mixture from either SHC
or STC, all of which are the presumed enzymic products based
on the idea that the polycyclization could be initiated from the
methyl-deficient part (Scheme 2 and 3). This finding strongly
suggested that the two geminal methyls are indispensable for
the initiation of polycyclizations by both STC and SHC, which
is in contrast to the report that 2,3-trans-1A-norsqualene
2,3-oxide, lacking one methyl on the epoxide ring, was cyclized
by the lanosterol synthase.1
OH
H
2
1
1b
1
5
H
H
OH
H
H
H
R2
R1
R2
R1
R2
R1
1
6
17
18
SHC
12, 13, 14, 15, 16, 17, 18
6
1
6, 17, 18: R1 = Me, R2 = H; or R1 = H, R2 = Me
Scheme 3
Independent incubations of 5 and 6 with cell-free extracts
7
from Tetrahymena pyriformis STC also gave 7 and 8 (1:2.5
ratio), respectively. No other products were detected. The
products with STC were indistinguishable from those with SHC
via GC-MS.
5
The structures of highly polar compounds 7 (C29
HRMS: m/z 414.3834; requires 414.3862) and 8 (C29
14.3854) were determined via NMR analysis. The signals at d
4.1 and 76.7 in 7 and 8, respectively, proved the involvement
of a hydroxy group. Detailed analyses using 2D NMR revealed
that 7 and 8 had a pentacyclic tetrahymanol skeleton possessing
a chair conformation of the E-ring. The hydroxy groups of both
and 8 were in the same equatorial orientation at the
1-position, but the arrangements of the methyl at the
2-position were different; axial for 7 by taking account of the
coupling constants of H-21 [d
Hz)] and equatorial for 8 owing to the ddd splitting of H-21 [d
H
H
50O, EI-
50O, m/z
5
4
7
C
,9
5
This work was partly supported by a Grant-in-Aid to T.H.
(No.0966011) from the Ministry of Education, Science and
Culture, Japan.We are also indebted to Dr M. Takahashi,
Tsukuba University, for the gift of Tetrahymena pyriformis
(GL strain).
7
2
2
H
3.69 (ddd, J = 11.2, 5.2, 5.2
H
5
3
.05 (J = 11.0, 11.0, 4.8 Hz)]. Product 12 (C29H48, EI-HRMS:
m/z 396.3722; requires 396.3756) did not have a hydroxy group.
One of the 7 methyl groups appeared as a triplet (J = 7.6 Hz)
Notes and references
1
I. Abe, M. Rohmer and G. D. Prestwich, Chem. Rev., 1993, 93, 2189.
1
2 (a) C. Pale-Grosdemange, C. Feil, M. Rohmer and K. Poralla, Angew.
Chem. Int. Ed., 1998, 37, 2237; (b) T. Sato, T. Abe and T. Hoshino,
Chem. Commun., 1998, 2617; (c) I. Abe and M. Rohmer, J. Chem. Soc.,
Chem. Commun., 1991, 902.
and the other 6 methyls as singlets in the H NMR spectrum,
showing the presence of one ethyl group in 12, and detailed
analysis revealed a neohopene skeleton,2a but with a carbon
skeleton of C29
.
3
E. J. Corey, S. C. Virgil, H. Cheng, C. H. Baker, S. P. T. Matsuda, V.
Singh and S. Sarshar, J. Am. Chem. Soc., 1995, 117, 11819; E. J. Corey
and H. Cheng, Tetrahedron Lett., 1996, 37, 2709; T. Hoshino and Y.
Sakai, Chem. Commun., 1998, 1591.
Formation of the tetrahymanol skeleton by SHC has never
been reported before, and also is quite interesting from the
evolutionary aspect of squalene cyclases. The E-ring formation
proceeded with complete stereoselectivity; the (23Z)-methyl
group of 5 was axial, while the corresponding E-methyl of 6 was
equatorial during the E-ring formation. The methyl orientations
of each product from SHC agreed with those from STC
4 C. Jenson and W. L. Jorgensen J. Am. Chem. Soc., 1997, 119, 10846.
1
1
5
Analyses of NMR data ( H- H COSY 45, HOHAHA, NOESY, DEPT,
HMQC and HMBC) unequivocally supported the structures of 5, 6, 7, 8
and 12.
6
7
T. Sato, Y. Kanai and T. Hoshino, Biosci. Biotechnol. Biochem., 1998,
(
Scheme 2), and were consistent with the previous report that
6
2, 407.
(a) P. Bouvier, Y. Berger, M. Rohmer and G. Ourisson, Eur. J. Biochem.,
980, 112, 549; (b) M. Renoux and M. Rohmer, Eur. J. Biochem., 1986,
155, 125.
the (23E)-methyl of natural 1 was arranged in an equatorial
orientation during the E-ring formation.7a The hydroxy groups
of both 7 and 8 were introduced in the same equatorial
disposition as a result of nucleophilic attack of a water molecule
in an equatorial direction on the C-21 cation. The cyclization of
the six-membered E-ring formation would be a concerted
reaction under stereoelectronic control and explained in terms
of HOMO–LUMO interactions.8
1
8 Given that a water molecule attacks in the axial direction, a more
hindered interaction would occur between the LUMO at C17–C18 and
the HOMO at C21–C22 due to constrained overlapping (carbon
numbering shown in Scheme 2).
9
R. B. Clayton, E. E. van Tamelene and R. G. Nadeau, J. Am. Chem. Soc.,
968, 90, 820.
1
It is noticeable that 12 was produced only from 5, and not
from 6 (Scheme 3). This fact implicates that the (23Z)-methyl of
Communication 9/01351B
732
Chem. Commun., 1999, 731–732