Stereochemical disposition of the geminal dimethyl groups in the enzymatic cyclization of geranyl diphosphate to (+)-bornyl diphosphate by recombinant (+)-bornyl diphosphate synthase from Salvia officinalis

Article abstract of DOI:10.1016/S0040-4020(01)00451-3

Regiospecifically deuterated geranyl diphosphate, in concert with NMR spectrometry, was employed to demonstrate that the trans-methyl group (C8) of geranyl diphosphate becomes the C9 carbon of (+)-bornyl diphosphate (geminal methyl syn to the diphosphate moiety) and that the cis-methyl group (C9) becomes the C8 (geminal methyl anti to the diphosphate). The syntheses of the relevant substrates and products, with accompanying spectrometric data are provided.

Full text of DOI:10.1016/S0040-4020(01)00451-3

TETRAHEDRON
Pergamon
Tetrahedron 57 -2001) 5327±5334
Stereochemical disposition of the geminal dimethyl groups in the
enzymatic cyclization of geranyl diphosphate to ꢀ1)-bornyl
diphosphate by recombinant ꢀ1)-bornyl diphosphate synthase
from Salvia of®cinalis
Mitchell L. Wise,a Hyung-Jung Pyun,b Greg Helms,c Bryce Assink,b Robert M. Coatesb
a,p
andRodney B. Croteau
aInstitute of Biological Chemistry, Washington State University, Pullman, WA 99164-6340, USA
bDepartment of Chemistry, University of Illinois, Urbana, IL 61801, USA
cDepartment of Chemistry, Washington State University, Pullman, WA 99164, USA
Received9 March 2001; accepted23 April 2001
AbstractÐRegiospeci®cally deuterated geranyl diphosphate, in concert with NMR spectrometry, was employed to demonstrate that the
trans-methyl group -C8) of geranyl diphosphate becomes the C9 carbon of -1)-bornyl diphosphate -geminal methyl syn to the diphosphate
moiety) andthat the cis-methyl group -C9) becomes the C8 -geminal methyl anti to the diphosphate). The syntheses of the relevant substrates
and products, with accompanying spectrometric data are provided. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
The biosynthesis of -1)- and- 2)-borneol in plants is unique
among monoterpenes in the incorporation of oxygen from
the diphosphate during cyclization of the acyclic precursor,
geranyl diphosphate -1, Scheme 1). Thus, -1)-camphor -5)
is biosynthesizedin culinary sage - Salvia of®cinalis) by
cyclization of 1 to -1)-bornyl diphosphate -2) followedby
sequential hydrolyses to -1)-bornyl monophosphate -3),
1±3
-1)-borneol -4), and®nally oxidation to the ketone.
Scheme 1. Biosynthesis of -1)-camphor -5) from geranyl diphosphate -1).
Investigations using -1)-bornyl diphosphate synthase
-BPPS) from this source ledto the recognition that 1 serves
tion is terminatedby a unique anti-Markovnikov capture of
the diphosphate anion to form 2. Oxygen-18 labeling
experiments have shown that both diphosphate ionizations
andrecombinations occur without detectable scrambling of
label from the C±O bonds into the other diphosphate
oxygens.6,7 Interestingly, while the sage BPPS produces
exclusively -1)--1R, 4R)-bornyl diphosphate -2), a comple-
mentary enzyme from tansy Tanacetum vulgare produces
the antipodal -2)--1S, 4S)-bornyl diphosphate by a mirror-
image cyclization reaction sequence.8
as the universal precursor of regular monoterpenes, and
subsequent studies with monoterpene syntheses from sage
provided evidence for the general mechanistic paradigm
4,5
illustratedin Scheme 2.
Thus, the substrate 1 undergoes
ionization to a geranyl1/OPP2 ion pair andinternal return of
the diphosphate anion generates an enzyme-bound linalyl
diphosphate -6A) intermediate. Rotation about the newly
formedC2±C3 single bondto the0 cisoidconformer - 6B)
followedby re-ionization and SN cyclization gives rise to
the central a-terpinyl carbocation -7) intermediate. Further
transformations of this highly reactive species occur by
cyclization, hydride shift, and Wagner±Meerwein
rearrangements. In the case of BPPS, the bridging cycliza-
A fundamental tenet of this general mechanistic model for
monoterpene biosynthesis4 is that the stereochemistry of the
a-terpinyl cation intermediate -and hence all subsequent
products) is dictated by the initial right- or left-handed
helical folding of the substrate that undergoes suprafacial
allylic rearrangement of the diphosphate group to generate
Keywords: biosynthesis; NMR; terpenes; labeling.
Corresponding author. Tel.: 1509-335-1790; fax: 1509-335-7643;
e-mail: croteau@mail.wsu.edu
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020-01)00451-3

5328
M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
Scheme 2. Enzymatic conversion of geranyl diphosphate to -1)-bornyl
diphosphate -2) through the intermediacy of -R)-linalyl diphosphate -6)
andthe - R)-a-terpinyl cation -7).
Scheme 3. Syntheses of regiospeci®cally deuterated -8,8,8-2H3) and
-9,9,9-2H3)geranyl diphosphates -14a, 14b). Reagents, conditions and
yields: -a) CH3C-vPPh3)CO2Et, 90%; -b) -CF3CH2O)2PvO-CH3)CH-
CO2Me, KHMDS, 18-C-6, THF, 2788C; 93%; -c) AlD3, ether, ,2108C;
91, 90%; -d) -EtO)2PvOCl, pyr, CH2Cl2: 88, 93%; -e) LiBEt3D, THF, 08C;
88, 93%; -f) Li, NH3, THF; 90, 100% -g) -i) MsCl, LiCl, collidine, DMF;
-ii) -Bu4N)3HOPP, CH3CN; 58, 61%.
the corresponding0 -S)- or -R)-linalyl diphosphate -6) and
then anti,endo-SN cyclization to the respective -S)- and
-R)-a-terpinyl ions.9±11 The stereochemical predictions of
this model were validated by labeling studies, kinetic
measurements, andstereochemical edterminations of
products.4,5,12±14 However, the factors that govern the
regio- andstereochemical outcome of the later steps remain
obscure. Previous studies established that the E methyl
group of 1 becomes the exo methyl of a- and b-pinene,15
andthat proton elimination in limonene biosynthesis occurs
either regioselectively at the Z methyl or almost randomly at
both positions, depending on the particular cyclization
enzyme.16±18 The principal objective of this collaboration
was to elucidate the geminal methyl correlation in the
cyclization catalyzedby BPPS.
2. Syntheses and NMR assignments
The requiredlabelededrivatives of
1 were preparedas
shown in Scheme 3 following previously publishedpro-
cedures in most cases.20,21 Reductive displacements of the
-Z)-allylic phosphates -11a and 11b) with LiBEt3D avoided
considerable Z±E isomerization observedin reudctions
conducted with the corresponding unstable mesylates.22,23
High purity -$98%) of the synthetic intermediates 9a,b±
13a,b with regardto the 6, 7 double bond was ensured by
careful chromatographic puri®cations together with GC
and/or NMR analyses at each stage.
1
Plans for direct H NMR analysis of deuterium labeled 2
requiredaccess to relatively large amounts of the cyclase. A
cDNA encoding the -1)-BPPS of S. of®cinalis was isolated
andthe enzyme functionally overexpressedin E. coli.19 The
One objective of this work was physical characterization of
-1)- and- 2)-2 andtheir unambiguous NMR assignments
which were crucial to localizing the deuterium label in the
enzymatic cyclization products. Conversion of -1)- and
-2)-4 to the diphosphates was carried out by the Cramer
procedure1,24,25 as modi®ed recently by Danilov.26 Conden-
sation of the alcohols with tetrabutylammonium hydrogen
phosphate -5 equiv.) andCCl 3CN -30 equiv.) in acetonitrile
-room temperature, 20 min) afforded a mixture of mono-,
di-, and polyphosphates that was separated on a Dowex ion
exchange resin by gradient elution with ammonium formate
in methanol in a manner similar to literature methods.27,28
²
pseudo-mature form of this recombinant enzyme produces
primarily -1)-2 -75% of product mix) along with a set of
stereochemically relatedole®ns -25% of proudct mix)
consisting of -1)-a-pinene -2.9% of total), -1)-camphene
-14.5%), limonene -1.9% -1) and2.4% - 2)), andthe
³
achiral compoundterpinolene -3.3%). The availability of
recombinant -1)-BPPS has allowed, for the ®rst time, the
assessment of additional stereochemical details of the
cyclization reaction. In this paper, we describe the synthesis
2
of -8-2H3) and-9- H3)geranyl diphosphate -14a and 13b),
1
The synthetic -1)- and- 2)-2 were characterizedby H and
31P NMR spectra that were identical to those of the
enzymatic cyclization products. HPLC retention time
comparisons between synthetic andenzymatic diphosphates
provided con®rming evidence.
andthe use of these labeledsubstrates, along with NMR
basedstructural evaluation, to ed®ne the stereochemical
dispositions of the terminal methyl groups in the course of
this remarkable enzymatic transformation.
The 1H spectrum of unlabeled bornyl diphosphate displayed
three distinct singlets at d 0.71, 0.72 and0.73 ppm, that
integratedfor a total of nine protons representing the three
methyl groups. The low®eldbroadtriplet at d 4.33 ppm was
assignedto the H2 proton. Selective irradiation at this
X
frequency produced an NOE to the signals at d 0.72 and
²
An N-terminally truncatedplasmidconstruct to remove the putative tran-
sit peptide region, thus improving soluble expression of the protein.
Stereochemical analyses of the ole®nic products produced by the full-
length protein -including transit peptide) previously reported were: -1)-
a-pinene [-16), 3.4%], camphene [9.5% -1)/0.5% -2)], limonene [3.9%
-1)/3.9% -2)], terpinolene -2.1%) andmyrcene -1.5%). 19
³

M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
5329
whereas the signals at d 1.57 and1.09 ed®ne protons
boundto the carbon at d 27.7. Selective irradiation at d
1.57 displayed an NOE to the methyl at d 0.71, thus
establishing this signal as H5X andthe singlet at d 0.71 as
that from the C8 methyl group.
3. Enzymatic cyclizations
Regiospeci®cally deuterated substrates -8-2H3)-14a and
-9-2H3)-14b, as well as unlabeledgeranyl idphosphate,
were separately incubatedwith puri®edbornyl diphosphate
synthase for 24±36 h under standard assay conditions19
resulting in the conversion of .95% of the substrate to
products. The ole®nic products -predominately a-pinene,
camphene andlimonene) were removedby extraction into
pentane. The ole®ns generatedin this reaction -approxi-
mately 25% of the total product) displayed some loss of
stereochemical ®delity, as previously reported.19 Speci®-
cally, the composition of the chiral ole®ns resulting from
reaction with unlabeledgeranyl diphosphate, as a percent
of total product, were camphene [14.5% -1)/trace -2)],
a-pinene [-16), 2.9% -1)/0% -2)], andlimonene [1.9%
-1)/2.4% -2)]. The relative proportions of the ole®n
products resulting from reaction with the two deuterium
labeled substrates did not vary signi®cantly from those indi-
catedabove. After removing the protein from the reaction
solutions by acetone precipitation andcentrifugation, the
remaining aqueous phase was ®rst vacuum dried then
Figure 1. Bornyl diphosphate with selected NOEs depicted by arrows.
0.73 ppm -Fig. 1), indicating that these are the protons on
C10 -the bridgehead methyl) and C9 -geminal methyl group
syn to the phosphate bearing carbon); additionally, an NOE
was observedto the signal at d 2.15 -H3X). Selective
irradiation of this signal -d 2.15) resultedin NOEs of the
singlet at d 0.72, a broad doublet of doublets centered at d
1.07, a broadtriplet at d 1.51, andthe broadtriplet at d 4.33,
thereby establishing the resonance at d 0.72 as the C9
protons, the signal at d 1.51 as the C4 methine proton,
and the doublet of doublets at d 1.07 as the geminal H3N
-the most intense NOE observed). Irradiation of the signal at
d 1.68 produced an NOE to the signal centered at d 1.13 and
a weak, but discernable, NOE at d 0.73 -H10), indicating
that this resonance represents H6N with through space
dipolar interaction to H6X andH5 at d1.09. An HMQC
N
experiment ®rmly establishedthat the proton signals at d
1.68 and1.13 originate from the same carbon at d 26.2,
Figure 2. High ®eld region of HMQC spectrum of from bornyl diphosphate derived from: -A) unlabeled geranyl diphosphate; -B) -9-2H3)-geranyl diphosphate
-14b) and-C) -8- 2H3)-geranyl diphosphate -14a).

5330
M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
Scheme 4. Mechanisms for pinene synthase andBPP synthase cyclizations of - R)-linalyl PP bearing tritium anddeuterium label in the trans terminal methyl
group.
lyophilizedbefore resuspending the resulting proudct in
D2O for NMR spectrometry. The enantiomeric purity of
the bornyl diphosphate -2) generatedby enzymatic reaction
was determined by phosphatase hydrolysis of an aliquot of
the unlabeled product to the corresponding alcohol
-con®rmedby GC/MS), andthen chiral phase column
capillary GC analysis. The enzymatically generated- 1)-
bornyl diphosphate was shown to be 100% enantiomerically
pure using synthetic -1)- and- 2)-bornyl diphosphate as
standards for this analytical protocol.
1 to form -R)-6 and SN0 cyclization to the -R)-a-terpinyl ion
7 shown in Scheme 2. Further labeling experiments proved
that the trans methyl group -C8ˆCH2T) of 1 becomes the
exo methyl of -1)-a-pinene and- 2)-b-pinene, consistent
with a least motion rotation -30 vs 1508) about the exocyclic
C6±C7 bondof 7 -Scheme 4).15
A similar radiochemical approach to elucidate the methyl±
methyl correlation with -1)-borneol was unattractive
because the corresponding hydrogen atoms of the bornane
ring are less accessible to unambiguous degradative
chemistry. However, the availability of the recombinant,
overexpressed- 1)-BPPS provided suf®cient biosynthetic
product to permit use of deuterium-labeled substrates
along with NMR-basedlocalization of the relevant hydro-
gen atoms to reveal the origin of the geminal methyl groups.
With the assignments of the methyl groups in bornyl
diphosphate ®rmly established, it was possible to interpret
an HMQC experiment with the enzymatic reaction products
derived from the speci®cally deuterated substrates. As
illustratedin Fig. 2, conversion of -8- 2H3)geranyl
diphosphate -14a) resultedin loss of the crosspeaks at d
18.2, F1 -13C) and d 0.72, F2 -1H); conversely, conversion
of -9-2H3)geranyl diphosphate -14b) resultedin loss of the
crosspeaks at d 19.3, F1 -13C) and d 0.71, F2 -1H). These
results were further substantiatedby NOE experiments, i.e.
irradiation of the signal at d 4.33 in the bornyl diphosphate
product generated from -9-2H3)geranyl diphosphate showed
a clear NOE to d 0.73 and0.72 -H10 andH9), andirradia-
tion at d 2.15 produced an NOE to d 0.72, d 1.51 andto d
1.07. The opposite results were seen in NOE experiments
with bornyl diphosphate enzymatically produced from 14a,
i.e. NOEs were not observedfrom irradiation of H2 X or H3X
to the H9 signal, whereas irradiation of the H5X proton
produced a clear NOE to the proton signal at d 0.71 -H8)
andto other protons consistent with those observedin the
unlabeled molecule. These data unequivocally demonstrate
that enzymatic cyclization of geranyl diphosphate to -1)-
bornyl diphosphate converts the substrate trans -C8) methyl
group to the syn -C9) methyl in bornyl diphosphate, and the
cis -C9) methyl of geranyl diphosphate to the anti -C8)
methyl group in bornyl diphosphate.
It is clear from the NMR data presented above that the trans
terminal methyl of the substrate is convertedto the syn
geminal methyl of 2. Thus, cyclization to the bornyl cation
occurs on the same face of the original 6,7-double bond to
0
which the SN cyclization took place to form the a-terpinyl
cation 7.
As illustratedin Scheme 4, the cyclizations leading to the
pinenes and- 1)-2 catalyzedby - 1)-pinene cyclase and
-1)-BPPS both take place on the Si face of a common
-R)-a-terpinyl intermediate. Whether in the latter case this
process occurs by direct anti-Markovnikov addition to the
cyclohexenyl double bond to generate the bornyl cation
-7!18), by indirect formation of the tertiary pinyl cation
through a Markovnikov-orientedring closure andrearrange-
ment -7!15!18), or by passage through a bridged, non-
classical carbocationic intermediate -see Scheme 2) is not
clear. Nevertheless, the lack of oxygen scrambling in the
tightly ion-paireddiphosphate anion throughout the reaction
coordinate6,7 implies more substantial participation in the
cyclization than that of a passive terminating nucleophile.
Thus, positioning of the diphosphate over C2 of the cyclo-
hexenyl intermediate 7 couldpolarize the p-double bond
suf®ciently to disfavor formation of the tertiary pinyl carbo-
cation 15, and both promote the anti-Markovnikov addition
andinternal return to the secondary position to generate 2.
Relevant chemical precedent for stereoselective endo
capture of bornyl ion pairs exists in the formation of bornyl
chloride by rearrangement of pinyl chloride in the hydro-
chlorination of a-pinene29 andin the methanolysis of pinyl
4. Discussion
Previous studies12±14 employing radiochemical methods and
native enzymes from S. of®cinalis establishedthat the
cyclizations of 1 to -1)-2, -1)-a-pinene -16), and- 2)-b-
pinene -17) take place with overall retention of con®gura-
tion at C1 of 1 -C3 of 2 andC7 of the pinenes), consistent
with suprafacial allylic isomerization of the diphosphate in

M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
5331
p-nitrobenzoates to small amounts of bornyl p-nitrobenzo-
ate andmethyl ether.
5.2.2. Methyl ꢀ2Z,6E)-8-benzyloxy-2,6-dimethyl-2,6-
octadienoate ꢀ9b). A suspension of 4.13 g -15.6 mmol) of
18-crown-6 and1.02 g -3.08 mmol) of bis-2,2,2-tri¯uoro-
ethyl) 1--methoxycarbonyl)ethylphosphonate21 in 78 mL
30
5. Experimental
of THF was stirredandcooledat
2788C as 623 mg
-3.1 mmol) of KN[Si-CH3)3]2 in 7 mL of THF was added.
After 20 min, 568 mg -2.60 mmol) of aldehyde 8 in 7 mL of
THF was added. After 30 min at 2788C, saturatedaq.
NH4Cl -50 mL) was added, and the THF was removed at
reduced pressure. The remaining aqueous mixture was
extractedwith hexane -25 mL £3) andthe combinedextracts
were washedwith H 2O -50 mL), dried -MgSO4), and
concentrated. Puri®cation by chromatography with ethyl
acetate±hexane -1:10) mixture as eluent gave 701 mg
-93%) of the product 9b -one isomer by GC) as a clear
5.1. Gas chromatographic analysis
GC±MS was performedon a Hewlett±Packard6890 GC
with a quadrupole mass selective detector and Hewlett±
PackardChemstation for adta analysis. Cool on-column
injection was employedusing a temperature program from
408C -1 min hold) to 2008C at 208/min, then to 3208C at
408/min on an HP-5MS column -Hewlett±Packardpart #
19091S-433, crosslinked5% phenyl methyl polysiloxane,
30 m£0.25 mm, 0.25 mm ®lm thickness) with He as the
carrier gas at 0.7 mL/min. Chiral phase separations were
performedon a Hewlett±Packard5890 chromatograph by
split injection -80:1) on a cyclodex-B capillary column -J
andW Scienti®c #112-2532, 30 m £0.25 mm, 0.25 mm ®lm
thickness) using H2 as the carrier gas at 0.6 mL/min with
¯ame ionization detection. For alcohols, temperature
programming was from 708C -1 min hold) to 1808C at
58/min, then to 2308C at 308C/min, andfor ole®ns was
from 708C -10 min hold) to 2308C at 108/min. The following
retention times were determined -min): a-pinene -16) 7.45
-2)/7.71 -1), camphene 8.79 -2)/9.07 -1), limonene 12.13
-2)/12.57 -1), borneol -4) 15.62 -2)/15.82 -1).
1
oil: H NMR -400 MHz, CDCl3) d 1.65 -s, 3H), 1.89 -d,
3H, Jˆ1.2 Hz), 2.14 -t, 2H, Jˆ7.6 Hz), 2.61 -q, 2H,
Jˆ7.5 Hz), 3.73 -s, 3H), 4.03 -d, 2H, Jˆ6.8 Hz), 4.50 -s,
2H), 5.43 -td, 1H, Jˆ6.7, 1.2 Hz), 5.93 -td, 1H, Jˆ7.2,
1.4 Hz), 7.26±7.35 -m, 5H); 13C NMR -100 MHz, CDCl3)
d 16.3, 20.6, 27.6, 39.0, 51.2, 66.5, 71.9, 121.4, 127.0,
127.5, 127.8, 128.3, 138.5, 139.5, 142.7, 168.3.
5.2.3. ꢀ2E,6E)-ꢀ1,1-2H2)-8-Benzyloxy-2,6-dimethyl-2,6-
octadien-1-ol ꢀ10a). A solution of 559 mg -13.3 mmol) of
LiAlD4 in 15 mL of ether was stirredandcooledin an ice-
salt bath as 593 mg -4.45 mmol) of AlCl3 was added.33,34
After 30 min, 907 mg -3.00 mmol) of 9a in 7 mL of ether
was added to the AlD3 reagent, andthe mixture was stirred
for 30 min. The reaction mixture was hydrolyzed at 08C by
adding 0.74 mL of H2O, 0.75 mL of 15% aq. NaOH, and
2.25 mL of H2O, successively. After 30 min, the solids were
®lteredoff andthe ®ltrate was concentrated. Puri®cation of
the residue by ¯ash chromatography with ethyl acetate±
hexane -1:2) mixture as eluent gave 717 mg -91%) of
dideutero alcohol 10a. 1H NMR -400 MHz, CDCl3) d
1.49 -br, 1H), 1.65 -s, 3H), 1.66 -d, 3H, Jˆ0.7 Hz), 2.07
-t, 2H, Jˆ7.6 Hz), 2.18 -q, 2H, Jˆ7.3 Hz), 4.02 -d, 2H,
Jˆ6.6 Hz), 4.51 -s, 2H), 5.39 -m, 2H), 7.27±7.35 -m, 5H);
13C NMR -100 MHz, CDCl3) d 13.6, 16.4, 25.7, 39.1, 66.5,
72.1, 121.1, 125.7, 127.5, 127.8, 128.3, 135.0, 138.4, 139.9.
1H NMR spectral data are in agreement with the literature
values35,36 except the peak for the CD2OH group -d 3.86)
was absent.
5.2. NMR analysis
NMR data for 2 in D2O were collectedat 125.7 MHz for 13C
andat 499.8 MHz for 1H on a Varian Inova spectrometer.
31P spectra were collectedat 121.5 MHz on a Varian
Mercury spectrometer. Phase sensitive HMQC data were
collectedas 2048 complex points in t2 and128 complex
points in t1 using the gradient selected echo-antiecho
method for F1 quadrature. The data were linear predicted
to 256 points in t1 then zero ®lledto 1K points andFourier
transformed, yielding a ®nal matrix of 1K£512 real points.
Sixty-four transients were collectedper t1 increment. One-
dimensional NOE data were collected using the Double
31
PulsedFieldGradient Spin Echo sequence of Stott et al.
with a mixing time of 700 ms for transient NOE buildup.
Scalar J couplings andexact chemical shifts were deter-
minedvia single frequency decouplings of each proton in
the proton spectrum followedby simulation of the resulting
spectrum using the Varian VNMR simulation program.32
Chemical shifts and J values were iteratively adjusted
until the simulatedspectrum matchedthe real spectrum as
judged by subtraction of the two spectra.
5.2.4. ꢀ2E,6E)-ꢀ1,1-2H2)-8-Benzyloxy-2,6-dimethyl-2,6-
octadien-1-yl diethyl phosphate ꢀ11a). A solution of
273 mg -1.04 mmol) of 10a and145 mL -1.79 mmol) of
pyridine in 3 mL of CH2Cl2 was stirredandcooledat 0 8C
as 225 mL -1.49 mmol) of diethyl chlorophosphate was
added.22 The suspension was stirredat 0 8C for 4 h, diluted
with ether -15 mL), andwashedwith H O -20 mL£2), satu-
2
ratedaq. NaHCO 3 -20 mL), andsaturatedaq. NaCl -20 mL).
After drying -MgSO4) andconcentration, the residue was
puri®edby ¯ash chromatography with ethyl acetate±hexane
-2:3) containing 1% Et3N as eluent to give 365 mg -88%) of
phosphate 11a. 1H NMR -400 MHz, CDCl3) d 1.29 -td, 6H,
Jˆ7.1, 0.7 Hz), 1.61 -s, 3H), 1.66 -d, 3H, Jˆ0.7 Hz), 2.04 -t,
2H, Jˆ7.4 Hz), 2.15 -q, 2H, Jˆ7.3 Hz), 3.99 -d, 2H,
Jˆ6.8 Hz), 4.08 -quintet, 4H, Jˆ7.3 Hz), 4.46 -s, 2H),
5.37 -td, 1H, Jˆ6.7, 1.0 Hz), 5.46 -td, 1H, Jˆ6.8, 1.2 Hz),
7.22±7.31 -m, 5H); 13C NMR -100 MHz, CDCl3) d 13.4,
16.0 -d, Jˆ6.9 Hz), 16.3, 25.8, 38.6, 63.4 -d, Jˆ5.4 Hz),
5.2.1. Ethyl ꢀ2E,6E)-8-benzyloxy-2,6-dimethyl-2,6-octa-
dienoate ꢀ9a). The E,E-dienoate was prepared according
to a literature procedure20 using 1.092 g -5.00 mmol) of
-4E)-6-benzyloxy-4-methyl-4-hexanal -8) and1.824 g
-5.03 mmol) of ethyl 2--triphenylphosphoranylidene)pro-
pionate. Puri®cation of the product -E/Z isomer
ratioˆ96/4 by GC) by ¯ash chromatography with ethyl
acetate±hexane -1:10) mixture as eluent afforded 1.365 g
-90%) of the product 9a -,99% E,E-isomer by GC) as a
1
clear oil. H NMR spectral data are in agreement with the
literature values.20

5332
M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
66.3, 71.9, 121.0, 127.4, 127.6, 128.2, 129.3, 130.4 -d,
Jˆ6.8 Hz), 138.3, 139.5; 31P NMR -161.9 MHz, CDCl3) d
2.10.
-400 MHz, CDCl3) d 1.65 -s, 3H), 1.79 -d, 3H, Jˆ
1.0 Hz), 2.05 -t, 2H, Jˆ7.3 Hz), 2.18 -q, 2H, Jˆ7.3 Hz),
4.00 -d, 2H, Jˆ6.8 Hz), 4.51 -s, 2H), 5.25 -td, 1H, Jˆ7.4,
1.0 Hz), 5.38 -t, 1H, Jˆ6.8 Hz), 7.27±7.36 -m, 5H); 13C
NMR -100 MHz, CDCl3) d 16.5, 21.1, 25.7, 39.3, 66.2,
72.0, 121.2, 127.1, 127.5, 127.8, 128.3, 134.9, 138.2, 140.0.
5.2.5. ꢀ2E,6E)-ꢀ8,8,8-2H3)-1-Benzyloxy-3,7-dimethyl-2,6-
octadiene ꢀ12a). A solution of 365 mg -0.92 mmol) of
11a in 5 mL of THF was stirredandcooledat 0
8C as
5.5 mL of 1.0 M LiBEt3D in THF was added. After 1 h,
15 mL of H2O was slowly added and the product was
extractedwith hexane -10 mL £3). The combinedextracts
5.2.9. ꢀ2Z,6E)-ꢀ1,1-2H2)-8-Benzyloxy-2,6-dimethyl-2,6-
octadien-1-yl diethyl phosphate ꢀ11b). Yield, 719 mg
-91%); 1H NMR -400 MHz, CDCl3) d 1.33 -td, 6H,
Jˆ7.1, 1.0 Hz), 1.64 -s, 3H), 1.78 -d, 3H, Jˆ1.5 Hz), 2.06
-t, 2H, Jˆ7.7 Hz), 2.20 -q, 2H, Jˆ7.6 Hz), 4.02 -d, 2H,
Jˆ6.6 Hz), 4.10 -quintet, 4H, Jˆ7.3 Hz), 4.50 -s, 2H),
5.39 -m, 2H), 7.27±7.35 -m, 5H); 31P NMR -161.9 MHz,
CDCl3) d 2.96.
were washedwith 1N HCl -30 mL), saturatedaq. NaHCO
3
-30 mL), andsaturatedaq. NaCl -30 mL); dried-MgSO 4);
and concentrated. Puri®cation of the residue by chromato-
graphy with ethyl acetate±hexane -1:30) mixture as eluent
provided 199 mg -88%) of 12a. 1H NMR -500 MHz,
CDCl3) d 1.61 -s, 3H), 1.65 -s, 3H), 2.06 -t, 2H,
Jˆ7.5 Hz), 2.12 -q, 2H, Jˆ7.6 Hz), 4.04 -d, 2H,
Jˆ6.5 Hz), 4.51 -s, 2H), 5.11 -td, 1H, Jˆ6.9, 1.3 Hz),
5.41 -td, 1H, Jˆ6.8, 1.3 Hz), 7.27±7.36 -m, 5H); 13C
NMR -125 MHz, CDCl3) d 16.4, 17.6, 26.4, 39.6, 66.6,
71.9, 120.9, 124.1, 127.5, 127.8, 128.3, 131.6, 138.7,
5.2.10.
ꢀ2E,6Z)-ꢀ8,8,8-2H3)-1-Benzyloxy-3,7-dimethyl-
2,6-octadiene ꢀ12b). Yield, 416 mg -93%). 1H NMR
-400 MHz, CDCl3) d 1.66 -s, 3H), 1.70 -d, 3H, Jˆ
1.2 Hz), 2.05±2.08 -m, 2H), 2.11 -m, 2H), 4.05 -d, 2H,
Jˆ6.8 Hz), 4.52 -s, 2H), 5.13 -td, 1H, Jˆ6.5, 1.0 Hz),
5.43 -td, 1H, Jˆ6.8, 1.2 Hz), 7.27±7.38 -m, 5H); 13C
NMR -100 MHz, CDCl3) d 16.4, 25.6, 26.3, 39.6, 66.5,
71.9, 120.7, 124.0, 127.5, 127.8, 128.3, 131.5, 138.5,
1
140.4. H NMR37±39 and 13C NMR38,39 spectral data are in
agreement with the literature values except the peaks for the
CD3 group -dH 1.68, dC 25.6) are absent.
1
140.4. H NMR37±39 and 13C NMR38,39 spectral data are in
5.2.6. ꢀ2E,6E)-ꢀ8,8,8-2H3)-3,7-Dimethyl-2,6-octadien-1-ol
ꢀ13a). Reduction of benzyl ether 12a -199 mg, 0.80 mmol)
with Li -31.3 mg, 4.51 mmol) in liquidNH -30 mL) and
agreement with the literature values except the peaks for the
CD3 group -dH 1.60, dC 17.6) are absent.
3
THF -2 mL) was carriedout as describedpreviously for
geranyl-d6 benzyl ether.16 Puri®cation by chromatography
with ethyl acetate±hexane -1:5) mixture as eluent followed
by Kugelrohr distillation afforded 114 mg -90%) of the
-8,8,8-2H3)geraniol 13a: 1H NMR -400 MHz, C6D6) d
0.58 -br, 1H), 1.45 -s, 3H), 1.52 -s, 3H), 1.97 -t, 2H, Jˆ
7.7 Hz), 2.09 -q, 2H, Jˆ7.4 Hz), 3.96 -d, 2H, Jˆ6.3 Hz),
5.16 -td, 1H, Jˆ7.0, 1.3 Hz), 5.38 -tq, 1H, Jˆ6.7, 1.3 Hz);
13C NMR -100 MHz, C6D6) d 16.1, 17.6, 26.8, 39.9, 59.3,
5.2.11. ꢀ2E,6Z)-ꢀ8,8,8-2H3)-3,7-Dimethyl-2,6-octadien-1-
ol ꢀ13b). Yield, 276 mg -100%); 1H NMR -400 MHz,
C6D6) d 0.69 -br, 1H), 1.46 -s, 3H), 1.65 -d, 3H,
Jˆ1.5 Hz), 1.97 -t, 2H, Jˆ7.7 Hz), 2.09 -qd, 2H, Jˆ7.2,
0.9 Hz), 3.96 -d, 2H, Jˆ6.6 Hz), 5.17 -td, 1H, Jˆ7.0,
1.2 Hz), 5.38 -tq, 1H, Jˆ6.7, 1.2 Hz); 13C NMR
-100 MHz, CDCl3) d 16.1, 25.7, 26.8, 39.8, 59.3, 124.6,
2
124.9, 131.3, 138.1; H NMR -77 MHz, C6H6) d 1.45;
Isotope ratio by FI MS, d3 93.20%, d2 3.16%, d1 2.96%.
2
124.6, 124.9, 131.3, 138.0; H NMR -77 MHz, C6H6) d
1.57; Isotope ratio by FI MS, d3 95.71%, d2 3.17%, d1
1.11%.
5.2.12. Diammonium hydrogen ꢀ2E,6Z)-ꢀ8,8,8-2H3)-3,7-
dimethyl-2,6-octadien-1-yl diphosphate ꢀ14b). Yield,
64 mg -61%); H NMR -400 MHz, D2O) d 1.47 -s, 3H),
1
5.2.7. Diammonium hydrogen ꢀ2E,6E)-ꢀ8,8,8-2H3)-3,7-
dimethyl-2,6-octadien-1-yl diphosphate ꢀ14a). -8,8,8-
2H3)Geraniol -13a, 40.3 mg, 0.256 mmol) was converted
to the chloride by Meyers' procedure40 using LiCl
-56.7 mg, 1.338 mmol), 2,4,6-collidine -275 mL, 2.08
mmol), andMsCl -120 mL, 1.55 mmol) in DMF. Conver-
sion of the chloride to -8,8,8-2H3)geranyl diphosphate was
carried out according to Poulter's procedure:41 yield, 52 mg
1.50 -s, 3H), 1.88 -t, 2H, Jˆ6.3 Hz), 1.94 -q, 2H,
Jˆ6.5 Hz), 4.25 -t, 2H, Jˆ6.7 Hz), 4.99 -t, 1H,
Jˆ6.4 Hz), 5.24 -t, 1H, Jˆ6.7 Hz); 13C NMR -100 MHz,
D2O) d 15.7, 24.8, 25.6, 38.9, 62.7 -d, Jˆ5.3 Hz), 119.8
-d, Jˆ9.1 Hz), 124.2, 133.7, 142.9; 31P NMR -161.9 MHz,
D2O) d 24.52 -d, Jˆ22.0 Hz), 27.00 -d, Jˆ20.7 Hz).
5.2.13. ꢀ1)- and ꢀ2)-Bornyl diphosphatesꢀꢀ1)-2 and ꢀ2)-
ent-2) by chemical synthesis. To a solution of either -1)-
or -2)-borneol -4) -191 mg, 1.04 mol) in CCl3CN -3.20 mL)
was added Bu4NH2PO4 -1.82 g) with constant stirring at
room temperature.26 After 20 min, 2% NH3 in methanol
-10 mL) was added and, after an additional 10 min, the
solvent was removedunder vacuum. The resulting yellow
oil was resuspended in 2% NH3 in methanol -40 mL) and
the mixture was centrifugedto remove precipitatedinor-
ganic salts. The supernatant was concentratedunedr
1
-58%); H NMR -400 MHz, D2O) d 1.43 -s, 3H), 1.52 -s,
3H), 1.91 -t, 2H, Jˆ6.8 Hz), 1.96 -q, 2H, Jˆ6.6 Hz), 4.28 -t,
2H, Jˆ6.8 Hz), 5.01 -t, 1H, Jˆ6.6 Hz), 5.25 -t, 1H,
Jˆ6.6 Hz); 13C NMR -100 MHz, D2O) d 15.7, 17.0, 25.6,
38.9, 62.9 -d, Jˆ5.3 Hz), 119.7 -d, Jˆ8.4 Hz), 124.2, 133.7,
143.0; 31P NMR -400 MHz, D2O) d 28.89 -d, Jˆ20.7 Hz),
29.93 -d, Jˆ20.7 Hz).
The Z,E -or E,Z) isomers of 10b±14b were preparedand
puri®edas describedabove for the E,E isomers 10a±14a.
Only the ®nal yields and characterization data are given.
w
vacuum, andthe resulting oil was loadedonto a Dowex
1£8±400 column -1.0£21 cm2, formate form) andeluted
under gravity with 1.0 L of a 0.05±0.5 M ammonium
formate gradient in methanol.27,28 Fractions containing
-1)- or -2)-bornyl monophosphate and- 1)- or -2)-2), as
5.2.8. ꢀ2Z,6E)-ꢀ1,1-2H2)-8-Benzyloxy-2,6-dimethyl-2,6-
octadien-1-ol ꢀ10b). Yield, 529 mg -90%); 1H NMR

M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
Table 1. H, 31P and 13C NMR spectral data for bornyl diphosphate in D2O
5333
1
Position
1H -ppm)a
31P -ppm)b
J -Hz)
13C -ppm)c
1
2
49.0
82.4
36.8
4.33
2.15
1.07
1.51
1.57
3J3Xˆ9.0, 4J6Xˆ3.0, 3J3Nˆ2.0
2J3Nˆ13.6, 3J2Xˆ9.0, 3J4ˆ4.5, 4J5Xˆ3.2
2J3Xˆ13.6, 3J2Xˆ2.0
3exo
3endo
4
3J3Xˆ4.5, 3J5Xˆ4.0, 4J6Xˆ1.0
3J6Xˆ12.4, 2J5Nˆ12.0, 3J6Nˆ4.0,
3J4ˆ4.0, 4J3Xˆ3.2
44.7
27.7
5exo
5endo
6exo
1.09
1.13
2J5Xˆ12.0, 3J6Nˆ9.0, 3J6Xˆ4.7
2J6Nˆ12.5, 3J5Xˆ12.4, 3J5Nˆ4.7,
4J2ˆ3.0, 4J4ˆ1.0
26.2
6endo
7
8
1.68
2J6Xˆ12.5, 3J5Nˆ9.0, 3J5Xˆ4.0
47.1
19.4
18.2
12.9
0.71
0.72
0.73
9
10
aP
bP
27.8
25.1
3JH2ˆ7.6, 3JC1ˆ5.8, 2JC2ˆ4.6, 3JC3ˆ1.0
JPPˆ16.4
a
Referencedto HOD at 4.65 ppm.
Referencedto H 3PO4 -85%) at 0 ppm.
Referencedto dioxane at 66.8 ppm.
b
c
determined by TLC -silica gel; 6:3:1 2-propanol, conc.
NH4OH, H2O) were pooledseparately andrepeateldy
dried by lyophilization -Scheme 2). NMR: -see Table 1).
The residue was then lyophilized -2±4 h) to reduce residual
NH4HCO3 salts before transfer of the bornyl diphosphate
product in D2O to an NMR tube.
5.2.14. Biosynthesis of ꢀ1)-bornyl diphosphate. The
cDNA encoding bornyl diphosphate synthase from
S. of®cinalis19 was subclonedas a truncatedpseudo-mature
Acknowledgements
42
form into the pSBET plasmidvector. Overexpression of
This work was supportedby National Institutes of Health
Grants GM31354 to R. B. C. andGM 13956 to R. M. C.
the recombinant protein in E. coli provided milligram quan-
tities of puri®edprotein, thereby allowing production of
suf®cient bornyl diphosphate -approximately 4 mg) for
NMR analysis. Details of the subcloning strategy andthe
expression andpuri®cation protocols will be published
elsewhere.
References
1. Croteau, R.; Karp, F. Arch. Biochem. Biophys. 1979, 198,
523±532.
To minimize contamination of the water soluble bio-
synthetic product with buffer components, enzymatic
reactions were conducted in 25 mM NH4HCO3 -pH 7.2)
2. Croteau, R.; Hooper, C. L.; Felton, M. Arch. Biochem.
Biophys. 1978, 188, 182±193.
3. Croteau, R.; Karp, F. Arch. Biochem. Biophys. 1979, 198,
512±522.
containing only the requiredcofactor 10 mM MgCl . Prior
2
to use, the puri®edenzyme solution was exchangedinto this
reaction buffer by repeatedconcentration anddilution using
an Amicon Centriprep 50 concentrator. Approximately
5 mg of geranyl diphosphate was added to 1.0 mg of the
puri®ed protein in a total volume of 10 mL buffer divided
into ®ve screwcap test tubes, each overlaidwith 1.0 mL
pentane to trap volatile ole®n products. Conversion of the
substrate to bornyl diphosphate was monitored by periodi-
cally removing 50 mL aliquots of the reaction mixture and
enzymatically hydrolyzing the diphosphate components
containedtherein using bovine intestinal alkaline phos-
phatase andpotato apyrase -in 100 mM Tris -pH 8.0) with
4. Croteau, R. Chem. Rev. 1987, 87, 929±954.
5. Wise, M. L.; Croteau, R. Monoterpene Biosynthesis. Compre-
hensive Natural Products Chemistry: Isoprenoids Including
Steroids and Carotenoids, Cane, D. E., Ed.; Elsevier: Oxford,
1999; Vol. 2, pp 97±153.
6. Croteau, R.; Shaskus, J. J.; Renstrùm, B.; Felton, N. M.; Cane,
D. E.; Saito, A.; Chang, C. Biochemistry 1985, 24, 7077±
7085.
7. Cane, D. E.; Saito, A.; Croteau, R.; Shaskus, J.; Felton, M.
J. Am. Chem. Soc. 1982, 104, 5831±5833.
8. Croteau, R.; Shaskus, J. Arch. Biochem. Biophys. 1985, 236,
535±543.
10 mM MgCl2 and0.1 mM ZnCl )43 andextracting the
9. Godtfredsen, S. V.; Obrecht, J. P.; Arigoni, D. Chimia 1977,
31, 62±63.
2
resulting alcohols -borneol andgeraniol) for GC/MS
analysis. The enzymatic reactions were allowedto proceed
at 318C until less than 5% of the substrate remained-24±
48 h). After removing the pentane overlay the protein in the
reaction mixture was precipitatedwith 20% -v/v) acetone.
The acetone was removedunder vacuum, andprecipitated
protein was separatedby ®ltration of the concentrated
aqueous phase through a glass wool plug into a separate
test tube which was brought to dryness under vacuum.
10. Godtfredsen S. V. PhD Dissertation, ETH, Zurich, 1978, No.
6243.
11. Poulter, C. D.; King, C.-H. R. J. Am. Chem. Soc. 1982, 104,
1420±1422.
12. Croteau, R.; Felton, M. Arch. Biochem. Biophys. 1981, 207,
460±464.
13. Croteau, R.; Felton, N. M.; Wheeler, C. J. J. Biol. Chem. 1985,
260, 5956±5962.

5334
M. L. Wise et al. / Tetrahedron 57 22001) 5327±5334
14. Croteau, R.; Satterwhite, D. M.; Wheeler, C. J.; Felton, N. M.
J. Biol. Chem. 1989, 264, 2075±2080.
28. Coates, R. M.; Kang, H.-J. J. Org. Chem. 1987, 52, 2065±
2074.
15. Coates, R. M.; Denissen, J. F.; Croteau, R. B.; Wheeler, C. J.
J. Am. Chem. Soc. 1987, 109, 4399±4401.
29. Meerwein, H.; Vorster, J. J. Prakt. Chem. 1936, 147, 83±92.
30. Salmon, J. R.; Whittaker, D. J. Chem. Soc. 1971, 1249±1254.
31. Stott, K.; Stonehouse, J.; Keeler, J.; Hwang, T.-L.; Shaka, A. J.
J. Am. Chem. Soc. 1995, 117, 4199±4200.
32. Bothnerby, A. A.; Castellano, S. J. J. Chem. Phys. 1964, 41,
3863±3869.
16. Pyun, H.-J.; Coates, R. M.; Wagschal, K. C.; McGeady, P.;
Croteau, R. J. Org. Chem. 1993, 58, 3998±4009.
17. Suga, T.; Hiraga, Y.; Mie, A.; Izumi, S. J. Chem. Soc. 1992,
1556±1558.
18. Coates, R. M.; Elmore, C. S.; Croteau, R. B.; Williams, D. C.;
Morimoto, H.; Williams, P. G. J. Chem. Soc. Chem. Commun.
1997 -1997), 2079±2080.
33. Bhalerao, U. T.; Plattner, J. J.; Rapoport, H. J. Am. Chem. Soc.
1970, 92, 3429±3433.
34. Dilling, W. L.; Plepys, R. A. J. Org. Chem. 1970, 35, 2971±
2976.
19. Wise, M. R.; Savage, T. J.; Katahira, E. J.; Croteau, R. J. Biol.
Chem. 1998, 273, 14,891±14,899.
35. Hofer, A.; Eugster, C. H. Helv. Chim. Acta 1982, 65, 365±370.
36. Altman, L. J.; Ash, L.; Marson, S. Synthesis 1974, 129±131.
37. Chen, K. M.; Semple, J. E.; Joullie, M. M. J. Org. Chem. 1985,
50, 3997±4005.
20. Coates, R. M.; Ley, D. A.; Cavender, P. L. J. Org. Chem.
1978, 43, 4915±4922.
21. Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405±
4408.
38. Hoemann, M. Z.; Agrios, K. A.; Aube, J. Tetrahedron 1997,
53, 11,087±11,098.
22. Ravn, M. M.; Jin, Q.; Coates, R. M. Eur. J. Org. Chem. 2000,
1401±1410.
39. Sajiki, H.; Hirota, K. Tetrahedron 1998, 54, 13,981±13,996.
40. Collington, E. W.; Meyers, A. I. J. Org. Chem. 1971, 36,
3044±3045.
23. Schenk, D. J. Ph.D. Dissertation, University of Illinois, 2000.
24. Cramer, F.; Rittersdorf, W. R. Tetrahedron 1967, 23, 3015±
3022.
41. Woodside, A. B.; Huang, Z.; Poulter, C. D. Org. Synth. 1988,
66, 211±219.
25. Cornforth, R. H.; Popjak, G. Methods Enzymol. 1969, 15,
385±388.
42. Schenk, P. M.; Baumann, S.; Mattes, R.; Steinbis, H.-H.
BioTechniques 1995, 19, 196±200.
26. Danilov, L. L.; Mal'tsev, S. D.; Shibaev, V. N. Soviet J. Biorg.
Chem. 1988, 14, 712±714.
27. Sofer, S. S.; Rilling, H. C. J. Lipid Res. 1971, 10, 183±187.
43. Croteau, R.; Cane, D. E. Methods Enzymol. 1985, 110, 383±
405.