Enzyme redesign: Two mutations cooperate to convert cycloartenol synthase into an accurate lanosterol synthase

Article abstract of DOI:10.1021/ja053791j

Efforts to modify the catalytic specificity of enzymes consistently show that it is easier to broaden the substrate or product specificity of an accurate enzyme than to restrict the selectivity of one that is promiscuous. Described herein are experiments in which cycloartenol synthase was redesigned to become a highly accurate lanosterol synthase. Several single mutants have been described that modify the catalytic specificity of cycloartenol to form some lanosterol. Modeling studies were undertaken to identify combinations of mutations that cooperate to decrease the formation of products other than lanosterol. A double mutant was constructed and characterized and was shown to cyclize oxidosqualene accurately to lanosterol (99%). This catalytic change entailed both relocating polarity with a His477Asn mutation and modifying steric constraints with an Ile481Val mutation. Copyright

Full text of DOI:10.1021/ja053791j

Published on Web 09/24/2005
Enzyme Redesign: Two Mutations Cooperate to Convert Cycloartenol
Synthase into an Accurate Lanosterol Synthase
Silvia Lodeiro,† Tanja Schulz-Gasch,*,§ and Seiichi P. T. Matsuda*,†,|
Department of Chemistry and Department of Biochemistry and Cell Biology, Rice UniVersity, Houston,
Texas 77005, and F. Hoffmann-La Roche Ltd., Molecular Design PRBD-CS 92/2.10D, CH-4070 Basel, Switzerland
Received June 9, 2005; E-mail: matsuda@rice.edu; tanja.schulz-gasch@roche.com
Scheme 1. Oxidosqualene Cyclases Convert Oxidosqualene to
Diverse Cyclic Structures
Nature has used random mutagenesis and selection to generate
an enormous diversity of enzymes. Protein scientists have recently
begun to make progress in engineering enzymes with broadened
substrate, reaction, and product specificity.1 It has proven to be
dramatically more difficult to redesign enzymes to have stringent
specificity. Described herein are experiments in which homology
modeling was combined with site-directed mutagenesis to redesign
cycloartenol synthase to make lanosterol with 99% accuracy.
Oxidosqualene cyclases convert acyclic (S)-2,3-oxidosqualene
(1) to over 100 cyclic triterpenes (C30H50O and C30H52O2).2 The
reaction is initiated by epoxide protonation, followed by cation-
olefin cyclization and cation rearrangement, and terminated either
by deprotonation or by water addition. Cycloartenol synthase is a
member of this enzyme family that generates cycloartenol (2), the
sterol precursor in plants and some protists. Lanosterol synthase is
a paralog that forms lanosterol (3), the sterol precursor in animals
and fungi. These enzymes are only moderately related (∼40%
identical) but promote mechanistically similar reactions. Both
enzymes cyclize oxidosqualene to the protosteryl cation (4) and
promote hydride and methyl shifts to form the C-8 cation (5) and
then diverge by abstracting a different proton (Scheme 1). Cy-
cloartenol synthase promotes an additional hydride shift and
terminates the reaction with cyclopropyl ring formation and
deprotonation from C-19, whereas lanosterol synthase deprotonates
from C-9. We elucidated how the known catalytically important
residues impact the difference in deprotonation using a structural
model derived from the Alicyclobacillus acidocaldarius squalene-
hopene cyclase (AacSHC) crystal structure.3 The orientation of
active-site side chains was refined using the recently reported crystal
structure of the human lanosterol synthase (see Supporting Informa-
tion).4 We established previously that Tyr410, His477, and Ile481
are strictly conserved, catalytically important residues in Arabidopsis
thaliana cycloartenol synthase5 (AthCAS1). These residues syner-
gize to promote cycloartenol biosynthesis, and mutations at these
positions allow lanosterol formation. Ile481 is conserved in all
cycloartenol synthases, whereas Val is present in lanosterol
synthases. The Ile481 γ-methyl promotes accurate cycloartenol
formation by preventing rotation of the intermediate through steric
interactions with C-2 and the two axial methyl groups of the A-ring.
Removing the γ-methyl group with an Ile481Val mutation results
in 25% lanosterol in addition to cycloartenol and parkeol (6) (Table
1).6 Tyr410 hydrogen-bonds with His257 to form part of the ceiling
of the active site.3d Both Tyr410 and His257 are close to C-19,
and one of these could be the base that deprotonates to form
cycloartenol. Alternatively, their influence on deprotonation could
result from participation in the H-bonding network. Tyr410 is
present in all cycloartenol synthases, but animal and fungal
Table 1. % Yields of Cycloartenol (2), Lanosterol (3), Parkeol (6),
and 9â-∆7-Lanosterol (7) in Cycloartenol Synthase Mutants
mutant
2
3
6
7
native
99
0
0
54
0
0
0
0
0
0
0
1
0
0
5
H477N
H477Q
I481V
Y410T
88
22
25
65
78
78
78
99
94
12
73
21
2
<1
0
0
1
6
0
33
22
22
22
0
Y410T/I481V
Y410T/H477N/I481V
Y410T/H477Q/I481V
H477N/I481V
H477Q/I481V
0
lanosterol synthases maintain Thr at the corresponding position.
The AthCAS1 Tyr410Thr mutant forms 65% lanosterol along with
9â-lanosta-7,24-dien-3â-ol (7) and parkeol.7 Removing the aromatic
ring of Tyr410 decreases steric bulk above the intermediate. Because
the hydroxyl in Thr is closer to the R-carbon than in Tyr, the polar
groups of Tyr410Thr, Tyr532, and His257 are repositioned in the
Tyr410Thr mutant. This combination of steric and electronic
changes abolishes cycloartenol biosynthesis and allows deproto-
nation of the C-8/C-9 lanosteryl cation to form products 3, 6, and
7. His477 is not in the active site, but is a second-sphere residue
that affects the product profile through interactions with the side
chain of Tyr410.3d His477 is strictly conserved in the known
cycloartenol synthases, whereas lanosterol synthases maintain either
Gln or Cys. The AthCAS1 His477Gln mutant is a parkeol synthase
(73%) that also forms lanosterol (22%).8 Although known lanosterol
synthases lack Asn at this position, an AthCAS1 His477Asn mutant
biosynthesizes 88% lanosterol and 12% parkeol.8 The structural
model suggests that these His477 mutations alter steric and
electronic surfaces of the active site by offering a smaller interaction
partner to Tyr410, which is consequently withdrawn from the active
site. The Tyr410Thr mutation facilitates lanosterol production by
decreasing active-site bulk and relocating potential proton acceptors,
and the His477 mutants achieve a similar result by moving Tyr410
away from the intermediate cation.
† Department of Chemistry, Rice University.
§ F. Hoffmann-La Roche Ltd.
| Department of Biochemistry and Cell Biology, Rice University.
9
14132
J. AM. CHEM. SOC. 2005, 127, 14132-14133
10.1021/ja053791j CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
produces lanosterol more accurately (78%) than the parent mutants.3d,7
Neither mutation depends on direct interaction with the other or
preempts the effects of the other. In contrast, the His477Asn and
His477Gln mutations lack influence when both Ile481Val and
Tyr410Thr are present (Table 1).3d These three mutations are not
synergistic because, when Tyr410 is mutated to Thr, His477Asn/
Gln cannot interact with the smaller and more distant Thr side chain
and, therefore, cannot influence catalysis.
The AthCAS1 His477Asn Ile481Val double mutant (with native
Tyr410) is the most accurate example of an enzyme mutated to
biosynthesize lanosterol, and the AthCAS1 His477Gln Ile481Val
double mutant is nearly as accurate. Neither enzyme has the
Tyr410Thr Ile481Val maintained by animal and fungal lanosterol
synthases. Although no known native enzyme uses the motifs of
the more accurate AthCAS1 His477Asn Ile481Val double mutant,
the AthCAS1 His477Gln Ile481Val double mutant has the residues
of the phylogenetically distinct trypanosomal lanosterol synthases.12
The experiments described here show how mutagenesis coupled
with a high-quality model allows the design of an efficient and
highly accurate lanosterol synthase. An improved understanding
of structure and catalytic mechanisms should facilitate future
engineering of enzymes with tailored catalytic activities.13
Figure 1. Superposition of wild-type AthCAS1 (green) with (A) AthCAS1
H477Q I481V (yellow) and (B) AthCAS1 H477N I481V (blue). The
lanosteryl cation is shown with a transparent Connolly surface. The
hydrogen-bonding pattern is indicated by dotted lines.
Careful examination of the homology model suggested that the
His477Gln and His477Asn mutations should synergize with the
Ile481Val mutation to more accurately biosynthesize lanosterol. The
AthCAS1 His477Gln mutant has the polar functionality moved
toward C-11 and consequently biosynthesizes more parkeol than
lanosterol. AthCAS1 His477Asn forms lanosterol by positioning
the base near C-9/C-8, but close enough to C-11 to form some
parkeol. AthCAS1 Ile481Val allows some lanosterol biosynthesis
by introducing a smaller side chain, which enlarges the active-site
cavity, permitting rotation of the intermediate cation. Computer
models show that both the AthCAS1 His477Gln Ile481Val and the
AthCAS1 His477Asn Ile481Val double mutants relocate polarity
to a position more favorable for lanosterol formation. The decreased
sterics allow the intermediate to rotate, moving C-9/C-8 toward
the base. Because AthCAS1 His477Gln and AthCAS1 His477Asn
form 73 and 12% parkeol, respectively, the reduction in parkeol
biosynthesis in AthCAS1 His477Gln Ile481Val was expected to
be less than that in AthCAS1 His477Asn Ile481Val.
Acknowledgment. We thank Dr. W. K. Wilson and Dr. J.
Silberg for valuable advice. T.S.-G. thanks her colleagues from
Roche biostructure in Basel for stimulating discussions. This
research was financially supported by the National Science Founda-
tion (MCB-0209769), the Robert A. Welch Foundation (C-1323),
and the Herman Frasch Foundation.
Supporting Information Available: Details of experimental
procedures. This material is available free of charge via the Internet at
http://pubs.acs.org.
The AthCAS1 His477Asn Ile481Val and the AthCAS1 His477Gln
Ile481Val double mutants were expressed in the yeast lanosterol
synthase mutant SMY8.9 Both mutant enzymes genetically comple-
mented the mutation, whereas AthCAS1 did not. Expression in the
yeast squalene epoxidase/lanosterol synthase double mutant RXY610
provided cyclase free of in vivo products, and in vitro assay with
racemic oxidosqualene generated a triterpene alcohol fraction. After
purification by silica gel chromatography and derivatization as the
trimethylsilyl (TMS) ethers, GC-FID and GC-MS of the AthCAS1
His477Asn Ile481Val mutant showed a major peak (99%) with
retention time and MS identical to that of a TMS-lanosterol
standard. The GC-MS spectra also revealed the presence of TMS-
parkeol (1%). The structural assignments were confirmed by 500
MHz 1H NMR analysis. Similar analyses on the AthCAS1
His477Gln Ile481Val mutant products revealed the same products
in a ratio of 94 and 6%. If any additional byproducts were present,
they were at levels <1%. The GC-FID ratios are in close
agreement with the ratios obtained by NMR quantitation. These
rationally designed cycloartenol synthase derivatives are the best
known examples of terpene synthases modified to make different
products accurately, and they compare favorably with remodeled
proteins in other systems. The experimental results for both double
mutants confirmed our working hypothesis and further validate the
homology model. Activity assays indicated that the AthCAS1
His477Asn Ile481Val mutant has about half the activity of wild-
type AthCAS1.11
References
(1) (a) Cedrone, F.; Menez, A.; Quemeneur, E. Curr. Opin. Struct. Biol. 2000,
10, 405-410. (b) Penning, T. M.; Jez, J. M. Chem. ReV. 2001, 101, 3027-
3046. (c) Hult, K.; Berglund, P. Curr. Opin. Biotechnol. 2003, 14, 1-6.
(d) Bornscheuer, U. T.; Kazlauskas, R. J. Angew. Chem., Int. Ed. 2004,
43, 6032-6040.
(2) Xu, R.; Fazio, G. C.; Matsuda, S. P. T. Phytochemistry 2004, 65, 261-
290.
(3) (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. (c) Reinert, D. J.; Balliano, G.; Schulz, G. E. Chem. Biol.
2004, 11, 121-126. (d) Lodeiro, S.; Segura, M. J. R.; Stahl, M.; Schulz-
Gasch, T.; Matsuda, S. P. T. ChemBioChem 2004, 5, 1581-1585.
(4) Thoma, R.; Schulz-Gasch, T.; D’Arcy, B.; Benz, J.; Aebi, J.; Dehmlow,
H.; Hennig, M.; Stihle, M.; Ruf, A. Nature 2004, 432, 118-122.
(5) Corey, E. J.; Matsuda, S. P. T.; Bartel, B. Proc. Natl. Acad. Sci. U.S.A.
1993, 90, 11628-11632.
(6) Hart, E. A.; Hua, L.; Darr, L. B.; Wilson, W. K.; Pang, J.; Matsuda, S. P.
T. J. Am. Chem. Soc. 1999, 121, 9887-9888.
(7) Herrera, J. B. R.; Wilson, W. K.; Matsuda, S. P. T. J. Am. Chem. Soc.
2000, 122, 6765-6766.
(8) Segura, M. J. R.; Lodeiro, S.; Meyer, M. M.; Patel, A. J.; Matsuda, S. P.
T. Org. Lett. 2002, 4, 4459-4462.
(9) Corey, E. J.; Matsuda, S. P. T.; Baker, C. H.; Ting, A. Y.; Cheng, H.
Biochem. Biophys. Res. Commun. 1996, 219, 327-331.
(10) Fazio, G. C.; Xu, R.; Matsuda, S. P. T. J. Am. Chem. Soc. 2004, 126,
5678-5679.
(11) We did not undertake complete kinetic analysis because the hydrophobic
components require detergent for solubilization, and concentrations of
substrate and detergent in each phase of the aqueous/micellar system
cannot readily be determined. Experiments to measure the catalytic
competence of this mutant enzyme were restrained to concentrations well
above a plausible Km (see Supporting Information).
(12) (a) Buckner, F. S.; Ngyuen, L. N.; Joubert, B. M.; Matsuda, S. P. T. Mol.
Biochem. Parasitol. 2000, 110, 399-403. (b) Joubert, B. M.; Buckner,
F. S.; Matsuda, S. P. T. Org. Lett. 2001, 3, 1957-1960.
(13) Dwyer, M. A.; Looger, L. L.; Hellinga, H. W. Science 2004, 304,
1967-1971.
The synergistic behavior of the His477Asn/Gln and Ile481Val
mutations resembles that seen with the Tyr410Thr and Ile481Val
mutations; the AthCAS1 Tyr410Thr Ile481Val double mutant
JA053791J
9
J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005 14133