Steric bulk at position 454 in Saccharomyces cerevisiae lanosterol synthase influences B-ring formation but not deprotonation.

Article abstract of DOI:10.1021/ol9912940

[reaction: see text] Lanosterol synthase converts oxidosqualene to the tetracyclic sterol precursor lanosterol. The mutation experiments described here show that an active-site valine residue in lanosterol synthase contributes to cyclization control through steric effects. Mutating to smaller alanine or glycine residues allows formation of the monocyclic achilleol A, whereas the leucine and isoleucine mutants make exclusively lanosterol. The phenylalanine mutant is inactive.

Full text of DOI:10.1021/ol9912940

ORGANIC
LETTERS
2000
Vol. 2, No. 3
339-341
Steric Bulk at Position 454 in
Saccharomyces cerevisiae Lanosterol
Synthase Influences B-Ring Formation
but Not Deprotonation
Bridget M. Joubert, Ling Hua, and Seiichi P. T. Matsuda*
Department of Chemistry and Department of Biochemistry and Cell Biology,
Rice UniVersity, 6100 South Main Street, Houston, Texas 77005
matsuda@rice.edu
Received November 30, 1999
ABSTRACT
Lanosterol synthase converts oxidosqualene to the tetracyclic sterol precursor lanosterol. The mutation experiments described here show
that an active-site valine residue in lanosterol synthase contributes to cyclization control through steric effects. Mutating to smaller alanine
or glycine residues allows formation of the monocyclic achilleol A, whereas the leucine and isoleucine mutants make exclusively lanosterol.
The phenylalanine mutant is inactive.
Animals and fungi form the tetracyclic sterol ring system
from oxidosqualene in a single reaction catalyzed by lano-
sterol synthase (Scheme 1).1 Oxidosqualene is protonated
hindered efforts to determine how the enzyme directs
formation of the energetically unfavorable B-ring boat.
Nonenzymatic biomimetic cyclization studies have estab-
lished the structural and stereoelectronic requirements for
cation-olefin polycyclization,3 and a biomimetic cyclization
to construct the A/B-trans-9,10-syn system has recently been
achieved.4 In nonenzymatic systems, cyclization to the
desired product is achieved by optimizing intramolecular
interactions between the olefins, the electrophilic carbon, and
the leaving group. Enzymatic cyclizations are controlled by
intramolecular forces imposed by the active site. It has been
suggested that precisely localized point charges could control
cyclization.5 Aromatic residues that could contribute electron
density have been mutated in the related and structurally
characterized squalene hopene cyclase,6 and some of the
resultant mutants form different products.7 Recently, purely
Scheme 1 Reaction Catalyzed by Lanosterol Synthase
and cyclized to the chair-boat-chair protosteryl cation, and
a series of hydride and methyl shifts terminated by the
specific abstraction of the C9 proton produces lanosterol.2
The lack of a lanosterol synthase crystal structure has
(2) Corey, E. J.; Virgil, S. C. J. Am. Chem. Soc. 1991, 113, 4025-4026.
(3) Johnson, W. S.; Telfer, S. J.; Cheng, S.; Schubert, U. J. Am. Chem.
Soc. 1987, 109, 2517-2518.
(4) Corey, E. J.; Wood, H. B., Jr. J. Am. Chem. Soc. 1996, 118, 11982-
11983.
(1) (a) Matsuda, S. P. T. On the DiVersity of Oxidosqualene Cyclases;
Steinbu¨chel, A., Ed.; Wiley-VCH: Weinheim, 1998; pp 300-307. (b) Abe,
I.; Rohmer, M.; Prestwich, G. D. Chem. ReV. 1993, 93, 2189-2206.
(5) Johnson, W. S. Tetrahedron 1991, 47, xi-xxiii.
10.1021/ol9912940 CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/11/2000

steric effects have been shown to be important in controlling
deprotonation in the Arabidopsis thaliana cycloartenol syn-
thase (which also cyclizes to the protosteryl cation but
abstracts a different proton). The cycloartenol synthase
Ile481Val mutant converts oxidosqualene to a mixture of
cycloartenol, lanosterol, and parkeol (54:25:21).8 Ile481 is
strictly conserved in the known examples of cycloartenol
synthase,9 and valine is conserved at the corresponding
position in lanosterol synthase enzymes.1a These residues
correspond to active-site residue Asp374 in Alicyclobacillus
acidocaldarius squalene-hopene cyclase, and wild-type cy-
cloartenol synthase probably differs from the Ile481Val
mutant by active-site alterations rather than changes in global
architecture. We describe here a series of experiments in
which Saccharomyces cereVisiae10 lanosterol synthase Val454
(which corresponds to A. thaliana cycloartenol synthase
Ile481) was mutated to the hydrophobic residues phenyl-
alanine, leucine, isoleucine, alanine, and glycine to investi-
gate the role of sterics in catalysis.
mg/mL) were added to the resultant homogenates, which
were incubated at room temperature overnight. Triterpenes
were recovered from the reactions essentially as described
previously.14 Reactions were typically performed to provide
∼50 mg of total triterpene alcohol products.
The activities of the mutant enzymes were initially
analyzed by TLC (1:1 hexane:ether). This qualitative method
is useful for assaying total activity of an oxidosqualene
cyclase because diverse triterpene alcohols including lano-
sterol, cycloartenol, parkeol, lupeol, and â-amyrin have
sufficient structural similarity near the polar alcohol that they
are inseparable by TLC. The Val454Leu, Val454Ile,
Val454Ala, and Val454Gly mutants cyclized oxidosqualene
to material that comigrated with lanosterol. The Val454Ala
and Val454Gly mutants also produced a minor product less
polar than lanosterol. The Val454Phe mutant did not produce
detectable product (<1% the activity of wild-type, if any).
The products of the leucine and isoleucine mutants were
characterized further after acetylation with acetic anhydride
and pyridine. GC analysis of each reaction showed a single
signal that coeluted with lanosteryl acetate; no parkeyl acetate
or cycloartenyl acetate was observed (<0.5%, if any); 250
MHz 1H NMR analysis confirmed that each mutant produced
lanosterol. The acetylated triterpene produced by the iso-
leucine mutant showed methyl and acetate protons within
0.03 ppm of literature lanosteryl acetate signals:15 δ 0.688
(C18), 0.872 (C30), 0.884 (C28), 0.884 (C29), 0.912 (C21),
1.004 (C19), and 4.49 ppm (C3). The acetylated product of
the leucine mutant showed the same lanosteryl acetate
protons: δ 0.687 (C18), 0.871 (C30), 0.881 (C29), 0.881
(C28), 0.911 (C21), 1.003 (C19), and 4.50 ppm (C3).
The Val454Ala mutant was expressed and assayed simi-
larly. GC analysis of an acetylated aliquot showed two
compounds in a 91:9 ratio. The crude triterpene alcohols were
purified by silica gel chromatography (5% ether in hexane)
followed by HPLC (YMC-Pack-SIL, 15% tert-butyl methyl
ether:85% hexane). The 250 MHz 1H NMR spectrum of the
major compound matched that of literature lanosterol val-
ues,15 and that of the minor compound was consistent with
literature values for achilleol A.16
pSM61.21 (native S. cereVisiae lanosterol synthase in the
integrative galactose-inducible yeast expression vector
pRS305GAL)11 was subjected to single-stranded mutagen-
esis.12 The resultant lanosterol synthase Val454 derivatives
were used to transform the yeast lanosterol synthase mutant
SMY8.11 The strain expressing Val454Phe remained sterol-
dependent, but all other mutants produced colonies on sterol-
free media (1% yeast extract, 2% peptone, 2% galactose, 13
mg/L heme) that were indistinguishable from a control
expressing native lanosterol synthase in the same vector.
These genetic complementation experiments suggest that
when under galactose control Val454Leu, Val454Ile,
Val454Ala, and Val454Gly produce enough lanosterol to
meet the cell’s sterol requirements. If these mutants generate
byproducts, they are not sufficiently detrimental or abundant
to significantly inhibit yeast growth.
In vitro incubations with synthetic substrate generated
triterpene products for chromatographic and spectroscopic
analysis. Recombinant yeast was resuspended in 100 mM
potassium phosphate at pH 6.2 and lysed in a French Press.
Racemic oxidosqualene13 (1 mg/mL) and Triton X-100 (10
A control TLC assay suggested that this 91:9 quantitation
was inaccurate, because achilleol A (but not lanosterol) was
present in the homogenate before addition of substrate. The
yeast lanosterol synthase mutant SMY8 has a functional
biosynthetic pathway to oxidosqualene, which is converted
to lanosterol and achilleol A in vivo when the Val454Ala
mutant lanosterol synthase is expressed. Lanosterol is ap-
parently metabolized rapidly in vivo, but achilleol A is not.
Accumulated achilleol A produced in vivo would bias the
analysis to overestimate its production by the enzyme.
The Val454Ala and Val454Gly mutants were consequently
expressed in yeast strain LHY3,8 a lanosterol synthase mutant
(6) (a) Wendt, K. U.; Poralla, K.; Schulz, G. E. Science 1997, 277, 1811-
1815. (b) Wendt, K. U.; Lenhart, A.; Schulz, G. E. J. Mol. Biol. 1999, 286,
175-187.
(7) (a) Merkofer, T.; Pale-Grosdemange, C.; Wendt, K. U.; Rohmer, M.;
Poralla, K. Tetrahedron Lett. 1999, 40, 2121-2124. (b) Pale-Grosdemange,
C.; Merkofer, T.; Rohmer, M.; Poralla, K. Tetrahedron Lett. 1999, 40,
6009-6012. (c) Hoshino, T.; Sato, T. Chem. Commun. 1999, 2005-2006.
(8) Hart, E. A.; Hua, L.; Darr, L. A.; Wilson, W. K.; Pang, J.; Matsuda,
S. P. T. J. Am. Chem. Soc. 1999, 121, 9887-9888.
(9) (a) Corey, E. J.; Matsuda, S. P. T.; Bartel, B. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 11628-11632. (b) Morita, M.; Shibuya, M.; Lee, M.-S.;
Sankawa, U.; Ebizuka, Y. Biol. Pharm. Bull. 1997, 20, 770-775. (c)
Kushiro, T.; Shibuya, M.; Ebizuka, Y. Eur. J. Biochem. 1998, 256, 238-
244. (d) Darr, L. B.; Godzina, S. M.; Matsuda, S. P. T. Plant Physiol. 1999,
121, 686. (e) Godzina, S. M.; Lovato, M. A.; Meyer, M. M.; Foster, K. A.;
Wilson, W. K.; Gu, W.; de Hostos, E. L.; Matsuda, S. P. T. Lipids 2000,
in press.
(10) Corey, E. J.; Matsuda, S. P. T.; Bartel, B. Proc. Natl. Acad. Sci.
U.S.A. 1994, 91, 2211-2215.
(11) Corey, E. J.; Matsuda, S. P. T.; Baker, C. H.; Ting, A. Y.; Cheng,
H. Biochem. Biophys. Res. Commun. 1996, 219, 327-331.
(12) Ausubel, F. M.; Brent, R.; Kingston, R. E.; Moore, D. D.; Seidman,
J. G.; Smith, J. A.; Struhl, K. Current Protocols in Molecular Biology;
Wiley-Interscience: New York, 1999.
(13) Nadeau, R. G.; Hanzlik, R. P. Methods Enzymol. 1968, 15, 346-
351.
(14) Herrera, J. B. R.; Bartel, B.; Wilson, W. K.; Matsuda, S. P. T.
Phytochemistry 1998, 49, 1905-1911.
(15) Emmons, G. T.; Wilson, W. K.; Schroepfer, G. J., Jr. Magn. Res.
Chem. 1989, 27, 1012-1024.
(16) Barrero, A. F.; Alvarez-Manzaneda R.; E. J.; Alvarez-Manzaneda
R.; R. Tetrahedron Lett. 1989, 30, 3351-3352.
340
Org. Lett., Vol. 2, No. 3, 2000

that also has a squalene synthase deletion and therefore does
not accumulate oxidosqualene or achilleol A in vivo. GC
analysis of the acetylated products showed compounds with
relative retention times corresponding to those of lanosteryl
acetate and achilleyl A acetate in both the alanine (95:5)
under conditions that minimize lanosterol demethylation.
This oxidative reaction requires stoichiometric amounts of
the cofactor NADPH. Washed microsomes (a partially
purified enzyme preparation that should have very little
NADPH) of LHY3 expressing lanosterol synthase Val454Gly
were assayed to determine the product profile without
lanosterol metabolism. This assay gave essentially the same
results as crude homogenate (84:16), which therefore has
negligible lanosterol metabolism and is a suitable enzyme
source for product quantitation.
1
and glycine (83:17) mutants (Scheme 2). The H NMR
Scheme 2 S. cereVisiae Lanosterol Synthase Val454Ala and
Val454Gly Mutants Produce Significant Amounts of Achilleol A
The product profiles of the lanosterol synthase Val454
mutants are summarized in Table 1. The inactivity of the
Table 1. Product Composition of S. cereVisiae Lanosterol
Synthase Val454 Mutants
454 residue
lanosterol, %
achilleol A, %
phenylalanine
leucine
isoleucine
valine (wild-type)
alanine
0
100
100
100
95
0
0
0
0
5
glycine
83
17
Val454Phe mutant can be attributed to occluding the active
site with excessive bulk or prohibiting productive cyclization
by inappropriately positioned π-electron density. Alterna-
tively, the added steric bulk might disrupt protein folding.
Lanosterol synthase remains accurate when valine is mutated
to leucine or isoleucine. In contrast, cycloartenol synthase
produces a mixture of cycloartenol, lanosterol, and parkeol
when the corresponding conserved isoleucine residue is
mutated to valine.8 Even the more dramatic Val454Ala and
Val454Gly mutants do not alter deprotonation from the
protosteryl cation. Instead, these mutants acquire achilleol
A biosynthetic ability. Decreasing amino acid side chain
steric bulk from isopropyl to methyl or hydrogen allows
monocycle formation, suggesting that steric bulk contributed
by the valine side chain may participate in prefolding the
oxidosqualene ∆10 olefin to facilitate B-ring formation. It is
noteworthy that like deprotonation to form the tetracyclic
ring system, deprotonation to form the monocycle is accurate;
no other monocyclic compounds were observed.
spectra of the alanine and glycine mutant products showed
signals consistent with those of lanosteryl and achilleyl A
acetate literature values. The proton NMR spectrum of the
alanine mutant acetate mixture showed lanosteryl acetate
methyl and acetate protons within 0.02 ppm of literature
values as follows: δ 0.688 (C18), 1.004 (C19), 0.911 (C21),
0.882 (C28), 0.882 (C29), 0.872 (C30), and 4.51 ppm (C3).
Achilleyl A acetate signals were present within 0.02 ppm of
literature values16 as follows: methyl singlets at δ 0.793,
0.948, 1.603, and 1.685 ppm; side chain allylic methylene
protons at δ 2.015; exo-methylene olefinic singlets at δ 4.628
and 4.886 ppm; and C3 at δ 4.680. Proton NMR spectra of
the glycine mutant acetates supported the identification of
achilleyl A acetate with signals within 0.03 ppm of literature
values as follows: methyl protons at δ 0.793, 0.948, 1.603,
and 1.683 ppm; side chain allylic methylene protons at δ
2.015; exo-methylene protons at δ 4.626 and 4.882 ppm;
and C3 at δ 4.666. Lanosteryl acetate signals were present
in the glycine mutant acetate mixture within 0.02 ppm of
literature values: δ 0.687 (C18), 1.004 (C19), 0.911 (C21),
0.881 (C28), 0.881 (C29), 0.871 (C30), and 4.50 ppm (C3).
An additional incubation with the alanine mutant provided
crude triterpene alcohols, which were purified by silica gel
chromatography (5% ether in hexane) followed by HPLC
(YMC-Pack-SIL, 15% tert-butyl methyl ether:85% hexane).
Acknowledgment. The National Institutes of Health
(AI41598) and the Robert A. Welch Foundation (C-1323)
funded this research. We are grateful to William K. Wilson
for NMR analysis, Jihai Pang for mass spectral analysis, Lisa
B. Darr for helpful discussions and for constructing the
Val454Ile mutant, and Jennifer B. R. Herrera for helpful
discussions.
1
The 250 MHz H NMR spectra of the purified compounds
were then measured and found to be consistent with literature
values (see Supporting Information).
1
Supporting Information Available: A 250 MHz H
NMR spectrum of achilleol A produced by the Val454Ala
mutant and a GC trace of the acetylated Val454Ala products.
This material is available free of charge via the Internet at
http://pubs.acs.org.
A second potential complication with the quantitation is
that if yeast lanosterol demethylase were active under the in
vitro assay conditions, lanosterol would be consumed,
exaggerating the relative amount of achilleol A. To determine
whether this was a problem, a control assay was performed
OL9912940
Org. Lett., Vol. 2, No. 3, 2000
341