Effect of isotopically sensitive branching on product distribution for pentalenene synthase: Support for a mechanism predicted by quantum chemistry

Article abstract of DOI:10.1021/ja3043245

Mechanistic proposals for the carbocation cascade reaction leading to the tricyclic sesquiterpene pentalenene are assessed in light of the results of isotopically sensitive branching experiments with the H309A mutant of pentalenene synthase. These experimental results support a mechanism for pentalenene formation involving a 7-protoilludyl cation whose intermediacy was first predicted using quantum-chemical calculations.

Full text of DOI:10.1021/ja3043245

Communication
pubs.acs.org/JACS
Effect of Isotopically Sensitive Branching on Product Distribution for
Pentalenene Synthase: Support for a Mechanism Predicted by
Quantum Chemistry
Liansuo Zu,†,∥ Meimei Xu,‡,∥ Michael W. Lodewyk,†,∥ David E. Cane,*,§ Reuben J. Peters,*,‡
and Dean J. Tantillo*,†
†Department of Chemistry, University of California, Davis, California 95616, United States
‡Department of Biochemistry, Biophysics & Molecular Biology, Iowa State University, Ames, Iowa 50011, United States
§Department of Chemistry, Box H, Brown University, Providence, Rhode Island 02912-9108, United States
S
* Supporting Information
Scheme 1. Three Mechanisms Proposed for the Formation
of Pentalenene (1) and Δ6-Protoilludene (2) from (E,E)-
ABSTRACT: Mechanistic proposals for the carbocation
cascade reaction leading to the tricyclic sesquiterpene
pentalenene are assessed in light of the results of
isotopically sensitive branching experiments with the
H309A mutant of pentalenene synthase. These exper-
imental results support a mechanism for pentalenene
formation involving a 7-protoilludyl cation whose
intermediacy was first predicted using quantum-chemical
calculations.
Farnesyl Diphosphate
entalenene (1, Scheme 1) is a tricyclic sesquiterpene1,2 that
P
is produced in nature from farnesyl diphosphate (FPP)
through a cationic cascade reaction promoted by the enzyme
pentalenene synthase.3 The mechanism of this transformation
is one of the most highly studied among terpene-forming
reactions,3−5 in part because of the efficient generation of
complexity that accompanies conversion of FPPthe universal
acyclic, achiral precursor of all sesquiterpenesinto
pentalenene, a tricyclic, chiral, stereodense product.
At least two mechanisms have been suggested for the
formation of pentalenene from FPP. Path A in Scheme 1 (A →
B → D → E → F) represents the earliest and until recently the
most commonly accepted mechanistic proposal, involving
conversion of the humulenyl cations A and B to a secoillud-
6-en-3-yl cation (D) that then undergoes a 1,2-hydride shift
and subsequent cyclization to produce the penultimate
intermediate, the pentalenyl cation (F).3,4 The basic details of
this mechanism have been supported by a wide range of
experiments with stereospecifically labeled FPP and determi-
nation of the precise position and stereochemistry of isotopic
labeling in the enzymatically derived pentalenene product.3,4 In
2006, Gutta and Tantillo proposed an alternative cyclization
mechanism leading from B to F based on quantum-chemical
calculations [mPW1PW91/6-31+G(d,p)//B3LYP/6-31+G-
(d,p) in the absence of the enzyme active site].2d,5 In this
mechanism, the 7-protoilludyl cation (C), formed directly from
B, would be a mandatory intermediate along the pathway to
pentalenene (path B in Scheme 1; A → B → C → F).5
Although this mechanism invokes an unexpected intermediate
(C) followed by an unusual dyotropic rearrangement (C →
F),6,7 it is completely consistent with all of the reported
mechanistic and stereochemical results on the pentalenene
synthase reaction.3,4 Moreover, the predicted intermediacy of
protoilludyl cation C is also consistent with the previously
reported formation of the corresponding deprotonation
product, Δ6-protoilludene (2), as a minor (10−13%) coproduct
of pentalenene resulting from the cyclization of FPP by the four
pentalenene synthase active-site mutants H309A, H309C,
H309S, and H309F.4d It is also noteworthy that refluxing
Δ
7,13-protoilludene or either epimer of the 7-protoilludyl
Received: May 10, 2012
Published: June 27, 2012
© 2012 American Chemical Society
11369
dx.doi.org/10.1021/ja3043245 | J. Am. Chem. Soc. 2012, 134, 11369−11371

Journal of the American Chemical Society
Communication
alcohol in formic acid gives pentalenene in up to 28% yield,
consistent with the intermediacy of a species such as cation C.8
Although the enzymatic generation of cation C by the
pentalenene synthase mutants had previously been thought to
result from diversion (path C) of the natural cyclization path
A,4d,e the quantum-mechanical calculations would place the
protoilludyl cation C directly on the natural cyclization path B.
More recently, further quantum-mechanical calculations on
other possible conformations of intermediate C revealed that C
can be converted to F by an alternative stepwise rearrangement,
illustrated as path B′ in Scheme 1 (A → B → C → D′ → E′ →
F), in which D′ and E′ are geometric isomers of D and E,
having Z rather than E CC double bonds. In fact, path B′ is
predicted to have a barrier of only ∼6 kcal/mol for the
conversion of C to F from the lowest-energy conformer of C
(at the mPW1PW91/6-31+G(d,p)//B3LYP/6-31+G(d,p)
level),9 which is considerably lower than the barrier of nearly
20 kcal/mol for the direct dyotropic reaction.5,7
Figure 1. Effect of deuteration at C6 of FPP on the ratio of products
generated by H309A pentalenene synthase. GC−MS chromatograms
of reaction mixtures using (top) unlabeled FPP and (bottom)
[6-2H]FPP are shown. The peaks labeled * are due to β-elemene
from Cope rearrangement of germacrene A.
While these experimental observations and calculations are
consistent with a 7-protoilludyl cation intermediate, they do not
provide conclusive evidence as to whether such an intermediate
is on the direct pathway to pentalenene (paths B and B′) or
instead represents a diversion of the pentalenene pathway (path
C). The production of both 1 and 2 by H309A pentalenene
synthase does, however, provide us with an opportunity to
distinguish path A from paths B and B′ using the well-
established method of the isotopically sensitive branching
experiment.10,11 A key difference between these two mecha-
nistic scenarios is the point at which each pathway diverges
toward pentalenene and Δ6-protoilludene. In the path A
mechanism, the branch point for commitment to the formation
of either the natural product pentalenene or the diversion
product Δ6-protoilludene would be cation D. In contrast, for
both paths B and B′, the branch point is the protoilludyl cation
C itself. We therefore envisaged that substitution of the C6
proton of FPP by deuterium in [6-2H]FPP should have
essentially no effect on the ratio of 1 to 2 if the cyclization
mechanism proceeds through intermediate cation D via path A,
whereas this substitution would result in an increase in the 1:2
ratio if either path B or B′ is followed because of a primary
deuterium kinetic isotope effect (KIE) on the deprotonation of
C to give 2. The KIE suppressing the formation of Δ6-
protoilludene when [6-2H]FPP is used as the substrate if path B
or B′ is followed should result in an increased partition of the
common cation intermediate C toward pentalenene, resulting
in a net increase in the ratio of the final products 1 and 2.
Incubation of FPP with the purified recombinant
pentalenene synthase mutant H309A gave a 6.0:1 mixture of
1 (81%) and the coproduct 2 (13.4 0.3%), accompanied by
minor quantities (<6%) of germacrene A, detected as the
derived Cope rearrangement product β-elemene as described
previously, consistent with the results of previously reported
incubations with H309 mutants.4d,e The assays were carried out
in triplicate and analyzed by capillary GC−MS (Figure 1, top)
with the identity of each product confirmed by comparison of
both the electron-impact mass spectrum and retention index
with standards in the MassFinder 4.0 database.12 When
[6-2H]FPP was used as the substrate,13 the distribution of
sesquiterpene products was significantly shifted, with the
intensity of the protoilludene peak being reduced to only 7.5
0.4% of the total products while the intensity of the
pentalenene peak increased to 87% (Figure 1, bottom). This
nearly 2-fold increase in the 1:2 ratio (11.6 vs 6.0) as a result of
isotopically sensitive branching establishes that the protoilludyl
cation C is a common intermediate in the pathways for
formation of 1 and 2, as required by either path B or B′ but
inconsistent with formation of cation C as a diversion product of
path A to pentalenene (assuming that C and D do not rapidly
interconvert, i.e., for path A, conversion of D to C is effectively
irreversible).
The observed increase in the 1:2 ratio corresponds to a
primary KIE of kH/kD = 1.9 on the deprotonation of cation C
to yield 2, consistent with previously measured kH/kD values for
deprotonation of tertiary carbocations in terpene synthase-
promoted reactions (typically ranging from 2−6).11 Quantum-
chemical calculations using H2PO4− as a model base predicted a
kH/kD of 1.6−1.8.14,15 The conversion of C to F, whether by
path B or B′, would be expected to be subject to at most a small
normal secondary KIE as C6 changes from sp3 toward sp2
hybridization in the transition-state structures for the C → F5
and C → D′ reactions (these assumptions are supported by our
quantum-chemical calculations14). In contrast, the diversion of
cation D, formed by the previously postulated path A, to give
the protoilludyl cation C, would be expected to be subject to
only a minor secondary KIE (kH/kD slightly less than 1), since
C6 would be changing from sp2 toward sp3 hybridization.
Similarly, conversion of D to E along path A should have no
significant KIE since H6 is not directly involved in this step.
Path A for the cyclization of FPP to pentalenene by way of
cation D is therefore excluded by the observation of a decrease
in the proportion of 2 due to isotopically sensitive branching of
the common intermediate C, whether C is further converted to
1 by path B or path B′. The mechanisms proposed on the basis
of quantum-chemical calculations on the enzyme-free reaction
mechanism are therefore fully consistent with the experimental
results described herein. Further experimentation will be
necessary to distinguish between the downstream paths B
and B′.16
ASSOCIATED CONTENT
* Supporting Information
■
S
Additional details on computations, including the complete
Gaussian citation, and description of enzyme assays and
product identification. This material is available free of charge
via the Internet at http://pubs.acs.org.
11370
dx.doi.org/10.1021/ja3043245 | J. Am. Chem. Soc. 2012, 134, 11369−11371

Journal of the American Chemical Society
Communication
(10) Jones, J. P.; Korzekwa, K. R.; Rettie, A. E.; Trager, W. F. J. Am.
Chem. Soc. 1986, 108, 7074−7078.
AUTHOR INFORMATION
Corresponding Author
David_Cane@brown.edu; rjpeters@iastate.edu; djtantillo@
ucdavis.edu
■
(11) (a) Croteau, R. B.; Wheeler, C. J.; Cane, D. E.; Ebert, R.; Ha,
H.-J. Biochemistry 1987, 26, 5383−5389. (b) Wagschal, K. C.; Pyun,
H.-J.; Coates, R. M.; Croteau, R. Arch. Biochem. Biophys. 1994, 308,
477−487. (c) Nes, W. D.; McCourt, B. S.; Marshall, J. A.; Ma, J.;
Dennis, A. L.; Lopez, M.; Li, H.; He, L. J. Org. Chem. 1999, 64, 1535−
1542. (d) Schenk, D. J.; Starks, C. M.; Manna, K. R.; Chappell, J.;
Noel, J. P.; Coates, R. M. Arch. Biochem. Biophys. 2006, 448, 31−44.
(e) He, X.; Cane, D. E. J. Am. Chem. Soc. 2004, 126, 2678−2679.
(f) Wagschal, K.; Savage, T. J.; Croteau, R. Tetrahedron 1991, 47,
5933−5944. (g) Pyun, H. J.; Coates, R. M.; Wagschal, K. C.;
McGeady, P.; Croteau, R. B. J. Org. Chem. 1993, 58, 3998−4009.
(12) http://www.massfinder.com.
Author Contributions
∥L.Z., M.X., and M.W.L. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
D.J.T., L.Z., and M.W.L. gratefully acknowledge the National
Science Foundation (Grants CHE-0957416, CHE-0449845,
and CHE-030089 with the Pittsburgh Supercomputer Center)
for support. D.E.C. was supported by NIH Grant GM30301.
M.X. and R.J.P. were supported by NIH Grant GM076324. We
acknowledge Dr. Wayne K. W. Chou for performing the
analysis with the MassFinder 4.0 database, Ms. Taylor Chesnut
for assistance with the enzyme assays, and Drs. Pradeep Gutta
and Dan Willenbring for their work on preliminary quantum-
chemical calculations.
(13) (a) [6-2H]FPP was synthesized using a slightly modified
version of the reported procedure (see the Supporting Information for
details). (b) Cane, D. E.; Tandon, M. Tetrahedron Lett. 1994, 35,
5355−5358.
(14) (a) Predicted kH/kD values were computed using the Bigeleisen
and Mayer method, as implemented in the program Quiver (see:
Bigeleisen, J.; Mayer, M. G. J. Chem. Phys. 1947, 15, 261−267.
Saunders, M.; Laidig, K. E.; Wolfsberg, M. J. Am. Chem. Soc. 1989, 111,
8989−8994 ). A modified version of Quiver provided by Prof. Daniel
Singleton (Texas A&M) was utilized. The range for kH/kD reported in
the text reflects deprotonation from different conformers of C. See the
Supporting Information for additional details. (b) This is part 9 of our
series on sesquiterpene-related calculations. For part 8, see: Hong, Y.
J.; Tantillo, D. J. Chem. Commun. 2012, 48, 1571−1573.
REFERENCES
■
(1) (a) Seto, H.; Yonehara, H. J. Antibiot. 1980, 33, 92−93. (b) Cane,
D. E.; Sohng, J. K.; Williard, P. G. J. Org. Chem. 1992, 57, 844−852.
(2) (a) Cane, D. E. Compr. Nat. Prod. Chem. 1999, 2, 155−200.
(b) Cane, D. E. Chem. Rev. 1990, 90, 1089−1103. (c) Davis, E. M.;
Croteau, R. Top. Curr. Chem. 2000, 209, 53−95. (d) Tantillo, D. J.
Nat. Prod. Rep. 2011, 28, 1035−1053.
(3) (a) Cane, D. E.; Tillman, A. M. J. Am. Chem. Soc. 1983, 105,
122−124. (b) Cane, D. E.; Sohng, J.-K.; Lamberson, C. R.; Rudnicki,
S. M.; Wu, Z.; Lloyd, M. D.; Oliver, J. S.; Hubbard, B. R. Biochemistry
1994, 33, 5846−5857. (c) For a recent review on sesquiterpene
synthases, see: Miller, D. J.; Allemann, R. K. Nat. Prod. Rep. 2012, 29,
60−71.
(4) (a) Cane, D. E.; Abell, C.; Lattman, R.; Kane, C. T.; Hubbard, B.
R.; Harrison, P. H. M. J. Am. Chem. Soc. 1988, 110, 4081−4082.
(b) Cane, D. E.; Oliver, J. S.; Harrison, P. H. M.; Abell, C.; Hubbard,
B. R.; Kane, C. T.; Lattman, R. J. Am. Chem. Soc. 1990, 112, 4513−
4524. (c) Harrison, P. H. M.; Oliver, J. S.; Cane, D. E. J. Am. Chem.
Soc. 1988, 110, 5922−5923. (d) Seemann, M.; Zhai, G.; Umezawa, K.;
Cane, D. E. J. Am. Chem. Soc. 1999, 121, 591−592. (e) Seemann, M.;
Zhai, G.; de Kraker, J.-W.; Paschall, C. M.; Christianson, D. W.; Cane,
D. E. J. Am. Chem. Soc. 2002, 124, 7681−7689. (f) Lesburg, C. A.;
Zhai, G.; Cane, D. E.; Christianson, D. W. Science 1997, 277, 1820−
1824.
(15) The active-site histidine is not the base that performs the
deprotonation, since the H309A and related mutants all retain
significant pentalenene synthase activity.4 A plausible candidate for the
Brønsted base is the inorganic pyrophosphate originally released by
the pentalenene synthase-catalyzed ionization of FPP, thereby
−
prompting us to use H2PO4 as the model base in our calculations.
For reports in which pyrophosphate has been proposed to be the base
that carries out the final deprotonation to terminate terpene-forming
reactions, see ref 3c and: (a) Roy, A.; Roberts, F. G.; Wilderman, P.
R.; Zhou, K.; Peters, R. J.; Coates, R. M. J. Am. Chem. Soc. 2007, 129,
12453−12460. (b) Greenhagen, B. T.; O’Maille, P. E.; Noel, J. P.;
Chappell, J. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 9826−9831.
(c) Shishova, E. Y.; Di Costanzo, L.; Cane, D. E.; Christianson, D. W.
Biochemistry 2007, 46, 1941−1951. (d) Peters, R. J.; Croteau, R. B.
Arch. Biochem. Biophys. 2003, 417, 203−211. (e) Hong, Y. J.; Tantillo,
D. J. Org. Biomol. Chem. 2010, 8, 4589−4600. (f) Garms, S.; Chen, F.;
Boland, W.; Gershenzon, J.; Kollner, T. G. Phytochemistry 2012, 75, 6−
̈
13. (g) Zhou, K.; Peters, R. J. Chem. Commun. 2011, 47, 4074−4080.
(16) Note, however, that (6Z)-asterisca-3(15),6-diene, which would
arise from direct deprotonation of D′, is a known natural product. See:
(a) Mehta, G.; Umarye, J. D. Tetrahedron Lett. 2001, 42, 8101−8104.
(5) Gutta, P.; Tantillo, D. J. J. Am. Chem. Soc. 2006, 128, 6172−6179.
(6) Reetz, M. T. Angew. Chem., Int. Ed. Engl. 1972, 11, 129−130.
(b) Reetz, M. T. Angew. Chem., Int. Ed. Engl. 1972, 11, 130−131.
(c) Reetz, M. T. Tetrahedron 1973, 29, 2189−2194. (d) Reetz, M. T.
Adv. Organomet. Chem. 1977, 16, 33−65. (e) Hoffmann, R.; Williams,
J. E., Jr. Helv. Chim. Acta 1972, 55, 67−75. (f) Fernan
P.; Sierra, M. A. Chem. Rev. 2009, 109, 6687−6711.
(b) Fricke, C.; Hardt, I. H.; Konig, W. A.; Joulain, D.; Zygadlo, J. A.;
̈
Guzman, C. A. J. Nat. Prod. 1999, 62, 694−696.
́
dez, I.; Cossío, F.
(7) We have recently found that dyotropic rearrangement via a
different conformer of C than described in ref 5 can occur through a
transition-state structure that is ∼12 kcal/mol lower in energy than the
one described in ref 5, i.e., corresponding to a barrier for the dyotropic
rearrangement step of slightly less than 20 kcal/mol; a full account of
this work will be reported in due course.
(8) (a) Ohfune, Y.; Shirahama, H.; Matsumoto, T. Tetrahedron Lett.
1976, 17, 2869−2872. (b) Misumi, S.; Ohtsuka, T.; Ohfune, Y.; Sugita,
K.; Shirahama, H.; Matsumoto, T. Tetrahedron Lett. 1979, 20, 31−34.
(c) Pattenden, G.; Teague, S. J. Tetrahedron 1987, 43, 5637−5652.
(9) (a) A full account of the conformational potential energy surface
for C and D′ will be published in due course. (b) No minimum
corresponding to D has yet been located, likely because of the
proximity of the carbocation center to the π bond in such structures.
11371
dx.doi.org/10.1021/ja3043245 | J. Am. Chem. Soc. 2012, 134, 11369−11371