Metabolism and mode of action of cis- and trans-3-pinanones (the active ingredients of hyssop oil)

Article abstract of DOI:10.1080/00498250110095745

1. Hyssop oil is an important food additive and herbal medicine and the principal active ingredients are (-)-cis- and (-)-trans-3-pinanones. No information is available on their metabolism or specific mode of action. 2. The metabolites of cis- and trans-3-pinanones were examined from mouse and human liver microsomes and human recombinant P4503A4 with NADPH and on administration to mouse by gas chromatography/chemical ionization mass spectrometry comparison with standards from synthesis. 3. The major metabolite of cis-3-pinanone in each P450 system and in brain of the i.p.-treated mouse in quantitative studies was 2-hydroxy-cis-3-pinanone, and two minor metabolites were hydroxypinanones other than 2-hydroxy-trans-3-pinanone and 4S-hydroxy-cfs-3-pinanone. The urine from oral cis-3-pinanone treatment examined on a qualitative basis contained conjugates of metabolites observed in the microsomal systems plus 2,10-dehydro-3-pinanone. 4. Trans-3-pinanone was metabolized more slowly than the cis-isomer in each system to give hydroxy derivatives different than those derived from cis-3-pinanone. 5. Cis- and trans-3-pinanones and hyssop oil act as γ-aminobutyric acid type A (GABAA) receptor antagonists based on inhibition of 4′-ethynyl-4-n-[2,3-³H₂] propylbicycloorthobenzoate ([³H]EBOB) binding in mouse brain membranes (IC₅₀ of 35-64μM) and supported by tonic/clonic convulsions in mouse (i.p. LD₅₀ 175 to >250 mg kg^-1) alleviated by diazepam. The cis-3-pinanone metabolites 2-hydroxy-cis-3-pinanone and 2,10-dehydro-3-pinanone exhibit reduced toxicity and potency for inhibition of [³H]EBOB binding.

Full text of DOI:10.1080/00498250110095745

xenobiotica, 2002, vol. 32, no. 4, 251±265
Metabolism and mode of action of cis- and
trans-3-pinanones (the active ingredients of
hyssop oil)
K. M. HOLD, N. S. SIRISOMA, S. E. SPARKS and
J. E. CASIDA*
Environmental Chemistry and Toxicology Laboratory, Department of Environmental
Science, Policy and Management, 114 Wellman Hall, University of California, Berkeley,
CA 94720-3112, USA.
Received 10 July 2001
1. Hyssop oil is an important food additive and herbal medicine and the principal
active ingredients are (¡)-cis- and (¡)-trans-3-pinanones. No information is available on
their metabolism or speci®c mode of action.
2. The metabolites of cis- and trans-3-pinanones were examined from mouse and
human liver microsomes and human recombinant P4503A4 with NADPH and on
administration to mouse by gas chromatography/chemical ionization mass spectrometry
comparison with standards from synthesis.
3. The major metabolite of cis-3-pinanone in each P450 system and in brain of the i.p.-
treated mouse in quantitative studies was 2-hydroxy-cis-3-pinanone, and two minor
metabolites were hydroxypinanones other than 2-hydroxy-trans-3-pinanone and 4S-
hydroxy-cis-3-pinanone. The urine from oral cis-3-pinanone treatment examined on a
qualitative basis contained conjugates of metabolites observed in the microsomal systems
plus 2,10-dehydro-3-pinanone.
4. Trans-3-pinanone was metabolized more slowly than the cis-isomer in each system
to give hydroxy derivatives di€erent than those derived from cis-3-pinanone.
5. Cis- and trans-3-pinanones and hyssop oil act as g-aminobutyric acid type A
(GABAA) receptor antagonists based on inhibition of 40-ethynyl-4-n-[2,3-3H2]propylbi-
cycloorthobenzoate ([3H]EBOB) binding in mouse brain membranes (IC50 of 35-64mM)
and supported by tonic/clonic convulsions in mouse (i.p. LD50 175 to >250 mg kg¡1
)
alleviated by diazepam. The cis-3-pinanone metabolites 2-hydroxy-cis-3-pinanone and
2,10-dehydro-3-pinanone exhibit reduced toxicity and potency for inhibition of
[3H]EBOB binding.
Introduction
Hyssop oil is produced by steam distillation of the ¯owers and stalks of
Hyssopus ocinalis L. of the mint family (Lamiaceae). This essential oil is used
as a ¯avouring for foods and liqueurs (Chartreuse and Benedictine) and a com-
ponent in some perfumes with a spicy odour. It has also been employed for
centuries (and possibly millennia) as an herbal medicine claimed to be e€ective for
fever, colds, and liver and gallbladder complaints among a wide variety of prob-
lems (Crowfoot and Baldensperger 1932, Blumenthal 1998, Gruenwald et al. 1998,
Alberta Agriculture 2001, Southall 2001).
Hyssop is `generally regarded as safe’ for food use by the US Food and Drug
Administration (Duke 2001) and `no health hazards are known in conjunction with
the proper administration of designated therapeutic doses’ (Gruenwald et al.
*Author for correspondence; e-mail: ectl@nature.berkeley.edu
#
Xenobiotica ISSN 0049±8254 print/ISSN 1366±5928 online
2002 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/00498250110095745

252
K. M. Hold et al.
1998). However, hyssop oil does pose a risk to the central nervous system. An oral
dose of 10±30 drops in the adult human or two to three drops in a child over several
days resulted in clonic and/or clonic-tonic spasms and 130 mgkg¡1 i.p. in rat
caused similar poisoning signs (Millet et al. 1981, Blumenthal 1998, Gruenwald et
al. 1998). These limited toxicological observations and studies are less than
adequate for an important food ¯avouring and herbal medicine.
The principal constituents of hyssop oil are (¡)-cis-3-pinanone (cP) and (¡)-
trans-3-pinanone (tP), also known as isopinocamphone and pinocamphone, re-
spectively (®gure 1) (Simonsen and Owen 1949, Phytochemical Database 2001).
Hyssop oil generally contains more tP than cP (Garg et al. 1999, Aqua-Oleum
2000). It is considered a convulsant essential oil with epileptogenic components
(Burkhard et al. 1999). Hyssop oil, tP and thujone induce the same type of
electrocortical seizures associated with myoclonic activities (Millet et al. 1981),
suggesting that hyssop oil and tP, as with thujone (Hold et al. 2000), are non-
competitive blockers of the g-aminobutyric acid (GABA)-gated chloride channel.
On this basis, the GABA type A (GABAA) receptor is a candidate target for the
action of hyssop oil and the 3-pinanones. The thujone diastereomers also serve as
models for the 3-pinanones in another respect, i.e. a- and b-thujones are rapidly
metabolized and detoxi®ed by cytochrome P450 (Hold et al. 2000, 2001).
The present study takes the ®rst steps in de®ning the metabolism and mode of
action of cP and tP. The data demonstrate that they are metabolized in vitro and in
vivo primarily by cytochrome P450 to a series of hydroxypinanones (quantitated in
brain but not in urine, where they appear as conjugates), that they act as blockers
of the GABA-gated chloride channel conveniently assayed as inhibition of 40-
ethynyl-4-n-[2,3-3H2Špropylbicycloorthobenzoate ([3H]EBOB) binding (Cole and
Figure 1. Partial metabolic pathways for cis-3-pinanone (cP) and trans-3-pinanone (tP). Metabolites
are designated by the position of hydroxylation (2OHcP or unspeci®ed for wOHcP, xOHcP,
yOHtP and zOHtP) or dehydrogenation (DHP).

3-Pinanone metabolism and mode of action
253
Casida 1992), and that metabolism is probably a detoxi®cation process relative to
their action at the GABAA receptor.
Materials and methods
Chemicals
Hyssop oil. The hyssop oil used was from steam-distilled ¯owering tops of H. ocinalis var. decumbens
grown in Switzerland and purchased from Lhasa Karnak Herb Co. (Berkeley, CA, USA).
General procedures. Preparation of cP, epi-cP, tP, and their hydroxy and dehydro derivatives involved
procedures shown in ®gure 2 and described below. The products were puri®ed by silica gel
chromatography and characterized by 1H-NMR (table 1), 13C NMR (table 2) and GC/MS (table 3),
which also established a purity in each case of 98±100%.
cP and epi-cP. cP was more easily obtained pure by oxidation of (‡)-isopinocampheol (Aldrich,
Milwaukee, WI, USA) than by chromatography of hyssop oil. Thus, pyridinium dichromate (1700 mg,
4.51 mmol) was added to (‡)-isopinocampheol (500 mg, 3.24 mmol) dissolved in methylene chloride
(5 ml) (Corey and Schmidt 1979). The reaction mixture was stirred overnight at room temperature,
diluted with diethyl ether (50 ml), ®ltered through silica gel and the ®ltrate washed with 5%
hydrochloric acid (50 ml), water (50 ml) and saturated sodium chloride (50 ml). The organic layer was
dried (anhydrous magnesium sulphate) and chromatographed on a silica gel column using 5% ethyl
acetate in hexane to obtain cP (482mg, 3.2 mmol, 98%). (¡)-Isopinocampheol (Aldrich) was converted
to epi-cP in an analogous manner. As expected (Simonsen and Owen 1949), cP from both hyssop oil and
the synthetic sample were observed to be levorotatory, whereas epi-cP was dextrorotatory.
tP. A cP/tP 1 : 2 diastereomeric mixture was used as the starting material rather than the more
complex hyssop oil. This mixture was obtained by epimerizing cP to tP (using sodium methoxide in
methanol) then conversion to the corresponding ketals (Satterwhite and Croteau 1987), leading to the
observation that the ketal with (2R; 3R)-2,3-butanediol forms much more rapidly with cP than tP. The
resulting reaction mixture greatly enriched in tP versus cP was chromatographed using 1±2% diethyl
ether in hexane to elute (in sequence) cP-ketal, tP and cP.
Enol and epoxy acetates. The hydroxypinanones were prepared via the enol and epoxy acetates.
To obtain the enol acetates, cP (1000mg, 6.57mmol) in isopropenyl acetate (6 ml) was treated with
p-toluenesulphonic acid (54 mg, 0.32 mmol) and re¯uxed for 3 days. The reaction mixture was
concentrated and chromatographed using 1±4% ethyl acetate in hexane to recover a non-resolved
mixture (860 mg, 67%) of 56% 2,3-enol acetate and 14% 3,4-enol acetate by GC/MS analysis. To obtain
the epoxy acetates, a solution of m-chloroperbenzoic acid (1.0 g, 75% pure, 4.4 mmol) in chloroform
(8 ml) was added to a solution of the above enol acetates (800mg, 4.11 mmol) in chloroform (8ml) at
08C. The mixture was stirred at 08C for 0.5 h then another portion of m-chloroperbenzoic acid (500 mg)
was added and warmed to room temperature with continued stirring for 3 h. The reaction mixture was
diluted with ethyl acetate (100ml) and washed in sequence with 5% sodium bisulphite (50 ml), 5%
sodium bicarbonate (50 ml) and saturated sodium chloride (50 ml). The organic layer was dried,
concentrated and chromatographed using 7% ethyl acetate in hexane to obtain the 2,3- and 3,4-epoxy
acetates (852 mg, 4.1 mmol, 98%) as a mixture.
2-Hydroxy-cis-3-pinanone (2OHcP) and (4S)-hydroxy-cis-3-pinanone (4SOHcP). The epoxy
acetates (210 mg, 1.0 mmol) were dissolved in absolute methanol (10 ml) and sodium methoxide
(0.1 mmol in methanol) was added at room temperature. The rearrangement was complete after 2 h.
A few drops of water were added then most of the solvent evaporated under vacuum and the residue
dissolved in ethyl acetate (25 ml) and washed with water (25ml) and saturated sodium chloride (25 ml).
The organic layer was dried and concentrated to an oily residue that was chromatographed using 3±5%
ethyl acetate in hexane to obtain 2OHcP (more polar product, 128 mg, 0.76mmol) (Suga et al. 1968,
Curci et al. 1993) and 4SOHcP (less polar product, 26 mg, 0.15 mmol).
2-Hydroxy-trans-3-pinanone (2OHtP). The mixture of epoxy acetates (120 mg, 0.57 mmol) in a 1 ml
round-bottomed ¯ask was heated at 1708C in an oil bath for 0.5 h. After cooling, the reaction mixture
was chromatographed using 4±6% ethyl acetate in hexane to obtain 2-acetoxy-tP (68 mg, 0.32mmol,

254
K. M. Hold et al.
Figure 2. Procedures for preparation of 3-pinanone isomers and their hydroxy and dehydro
derivatives. (a) Pyridinium dichromate, CH2Cl2, 258C, (b, i) NaOMe, MeOH, 258C, (b, ii)
(2R, 3R)-(¡)-2,3-butanediol, trimethylsilyl chloride, MeOH, 258C (tP does not form the ketal so
with a derivatized cP/tP mixture the unreacted tP is recovered by column chromatography),
(c) isopropenyl acetate, anhydrous p-toluenesulphonic acid, 1108C, (d) m-chloroperbenzoic acid,
CHCl3, 0±258C, (e) NaOMe, MeOH, 258C, (f) neat, 1408C, (g) tert-butyl hydroperoxide, SeO2,
CH2Cl2, 258C.
56%) as the major product. The acetate group was removed by treating with sodium methoxide in
methanol as above and the crude product was chromatographed using 15% ethyl acetate in hexane to
obtain 2OHtP (50 mg, 0.30 mmol, 94%).
2,10-Dehydro-3-pinanone (DHP). (1S)-b-pinene (Aldrich) (200 mg, 1.3 mmol) was dissolved in
methylene chloride (3ml) and selenium dioxide (6mg, 0.054 mmol) was added followed by tert-butyl
hydroperoxide (3ml, 90% pure). The reaction was stirred for 1 day at room temperature. Hexane
(50ml) was then added and the mixture washed with 10% potassium carbonate (30 ml) and saturated
sodium chloride (30 ml), dried over magnesium sulphate, ®ltered and concentrated. The oily residue
was chromatographed using 3% ethyl acetate in hexane to obtain the enone DHP (180 mg, 1.21mmol,
91%) (Mazzega et al. 1999).

3-Pinanone metabolism and mode of action
255

256
K. M. Hold et al.
Table 2. 13C-NMR (75 MHz) data (¯) in CDCl3.
Carbon
CH3
cPa
tPa
2OHcPb
2OHtPc
4SOHcP
DHPd
16.8
21.9
15.0
19.7
22.8
25.3
22.6
27.0
14.3
19.3
21.5
26.0
27.0
214.9
±
26.3
215.2
±
27.3
213.9
77.1
28.4
38.4
39.3
42.9
49.7
31.1
216.7
79.0
25.8
39.0
39.5
43.2
50.0
26.6
215.9
76.9
25.6
40.2
44.7
44.9
46.2
CO
199.7
-
CHOH
Other
34.3
39.0
39.2
44.7
45.1
51.3
29.1
38.4
39.4
44.3
44.5
46.4
32.4
38.6
40.7
42.4
48.3
117.2
149.1
a Assignment based on Badjah-Hadj-Ahmed et al. (1992).
b Assignment based on Solladie-Cavallo et al. (1994).
cAssignment based on Carlson and Pierce (1971).
d Assignment based on Mazzega et al. (1999).
GC/MS analysis
Using a previously described procedure (Hold et al. 2001) and brie¯y restated here, an HP5890 gas
chromatograph equipped with a DB-5 fused silica gel capillary column (30 m, 0.25 mm i.d., 0.25mm
®lm thickness) coupled to the HP5971 mass selective detector was used in the positive chemical
ionization mode. Helium was employed as the carrier gas and methane as the reagent gas at a ¯ow rate of
0.8 ml min¡1. The temperature programme was ®rst 80 to 2008C at 58C min¡1 and then 200 to 3008C at
208C min¡1, holding this ®nal temperature for 2 min. The injection port was at 2508C and the detector
at 2808C. The GC/MS was operated both in the full-scan and selected ion-monitoring (SIM) modes,
measuring m/z 135 for cP or 109 for tP (table 3) with a-thujone as the internal standard (m/z 135) and m/
z 151 for DHP and hydroxy-3-pinanones (table 3) with 7-hydroxy-a-thujone (m/z 151) (Sirisoma et al.
2001) as the internal standard. Trimethylsilyl (TMS) ethers were formed by adding N-(trimethylsi-
lyl)tri¯uoroacetamide (50 ml) to the extract residue and heating for 30 min at 808C (Hold et al. 2000,
2001). In each type of analysis, a sample aliquot of 1.0ml was injected splitless onto the column.
cP and epi-cP were analysed as their ketals formed with (2R,3R)-2,3-butanediol as described by
Satterwhite and Crotean (1987) for stereochemical assignment of the cis-pinanone fraction from hyssop
oil. Epi-cP was prepared for this purpose from (¡)-isopinocampheol (Aldrich) by the procedure above
for conversion of (‡)-isopinocampheol to cP.
Metabolism in cytochrome P450 (P450) systems
Mouse liver microsomes (10 000g supernatant, 100 000g pellet) prepared from a 20% (w/v) homo-
genate in phosphate bu€er (100 mm, pH 7.4) were washed once with phosphate bu€er and suspended in
0.25 M sucrose, 20 mM Tris, and 5.4 mM EDTA (pH 7.4). Other P450 sources were pooled human
microsomes and human P4503A4 Supersomes (Gentest, Woburn, MA, USA). The preparations were
stored frozen at ¡808C.
Liver microsomes (1 mg protein, 1.1 and 0.49 nmol P450 for mouse and human, respectively) or
P4503A4 (0.15 mg protein, 0.10 nmol P450) and NADPH (0 or 1 mm ®nal concentration) were
incubated with cP or tP (30mg added in ethanol, ®nal concentration 0.2 mm) in 100 mm phosphate,
pH 7.4, bu€er (1 ml) for 1h at 378C. Each reaction mixture was then cooled on ice, saturated with
sodium chloride and extracted with ethyl acetate (3ml). The ethyl acetate phase was evaporated to
200 ml for analysis of a 1.0-ml aliquot.

3-Pinanone metabolism and mode of action
257

258
K. M. Hold et al.
Animal treatment
Male albino Swiss-Webster mice (25±30 g) (Harlan, Indianapolis, IN, USA) were maintained under
standard conditions of controlled light (12 h light-dark cycle), temperature (20±228C) and relative
humidity (50±55%) with access to food and water ad libitum. The studies were carried out in accordance
with the Guiding Principles in the Use of Animals in Toxicology as adopted by the Society of Toxicology
in 1989 and were approved before commencement by the Animal Care and Use Committee of the
University of California at Berkeley.
Metabolites in the brain
Mice were treated i.p. with cP at 250 mg kg¡1 in propylene glycol (2 ml g¡1 mouse) and at appropriate
times thereafter the animals were killed and whole brains removed for analysis. The brains were rinsed
and homogenized in 100 mM phosphate, pH 7.4, bu€er (10ml). a-Thujone (0.5 mg) and 7-hydroxy-a-
thujone (2.0 mg) were added as internal standards. After centrifugation (1500g, 10min), the pellet was
resuspended in 2 ml phosphate bu€er, sonicated for 1 min and centrifuged, and the supernatant
fractions were combined. The samples were extracted with ethyl acetate (6 ml) and analysed as with
the extracts of microsomal incubations. Quantitation was based on standard curves for cP and 2OHcP
corrected for recoveries by the use of internal standards (a-thujone for cP and 7-hydroxy-a-thujone for
2OHcP). Mice administered tP at 250 mg kg¡1 were analysed the same way at 40 min after treatment.
Metabolites in the urine of mouse
Mice, as above, were treated orally with cP (100 mg kg¡1) in propylene glycol (2 ml g¡1 mouse) and
individually housed in all glass metabolism cages (Minor Metabowl, Jencons, Leighton Buzzard, UK).
Urine was collected for 24 h, an aliquot (1 ml) mixed with 200 mm phosphate, pH 7.0, bu€er (1 ml),
saturated with sodium chloride and extracted with ethyl acetate (4 ml). The unconjugated metabolites in
the ethyl acetate extract were analysed directly by GC/MS and the conjugate fraction in the aqueous
phase was treated with b-glucuronidase/arylsulphatase as described (Hold et al. 2001) then analysed
following deconjugation. Urine from the untreated mouse was used as a control but no interfering peaks
were observed.
Inhibition of [3H]EBOB binding
GABA-depleted mouse brain membranes (200 mg protein) (Cole et al. 1984) were incubated with the
test compound (added in dimethyl sulphoxide, ®nal concentration 1%) and [3H]EBOB (0.7 nM)
(38Ci mmol¡1; NEN Life Science Products, Boston, MA, USA) in 1.0 ml phosphate bu€er (10 mm,
pH 7.5) containing 200 mm sodium chloride for 70 min at 378C (Cole and Casida 1992). The mixtures
were then ®ltered through GF/C glass®bre ®lters using a cell harvester, followed by two rinses with 5 ml
ice-cold 0.9% sodium chloride. Speci®c binding was de®ned as the di€erence between total binding and
non-speci®c binding determined in the presence of 5 mm a-endosulphan.
Toxicity to mouse
Mice, as above, were treated i.p. with the test compound dissolved in dimethyl sulphoxide (2±
3 ml g¡1 mouse) for observation of poisoning signs shortly after treatment and mortality at 24h.
Alternatively, mice were treated with diazepam (0 or 1 mg kg¡1) in dimethyl sulphoxide (2 m g¡1
mouse) and 15min later with cP or hyssop oil at an approximate LD50 dose (200mg kg¡1) for
observation of poisoning signs and mortality.
Results
Analysis of hyssop oil and stereochemical assignment of 3-pinanones
The two principal constituents of the hyssop oil examined (66 and 22% of the
total peak area at tR 6.7 and 6.4 min, respectively) were identical in tR and MS
fragmentation patterns to cP prepared by oxidation of (‡)-isopinocampheol and tP
from epimerization of this synthetic cP sample, respectively (table 3). cP (both as a
synthetic standard and as a constituent of hyssop oil), on conversion to the ketal
with (2R 3S)-2,3-butanediol (®gure 2), gave t ˆ 11 3 min with no peak at
;
:
R
11.0 min (as would be the case for epi-cP), which con®rms the stereochemical
assignment and enantiomeric purity of cP in hyssop oil.

3-Pinanone metabolism and mode of action
259
GC/MS analysis of 3-pinanones and their metabolites and derivatives
Table 3 gives tR and protonated molecular ions and major fragments with their
relative abundances for each of the standards and metabolites. tP and cP are easily
di€erentiated bytR and relative abundances of their fragments at m/z 135, 109 and
93. DHP is baseline separated from tP and can be analysed in the presence of tP
and cP by [MH]‡ at m/z 151 versus 153. Seven hydroxy-3-pinanones were
analysed as standards or metabolites, four derived from cP and three from tP.
Each hydroxy-3-pinanone gives a small protonated molecular ion and fragments as
follows (m/z, relative abundance): [MH]‡ 169 (3±39), [MH-H2O]‡ 151 (47±100),
[MH-H2O-CO]‡ 123 (9±100) and [MH-H2O-C3H6]‡ 109 (9±89). The base peaks
for xOHcP and zOHtP appear at m/z 113. Corresponding values for the trimethyl-
silyl derivatives are [MH]‡ 241 (5±100), [MH-CH4Š‡ 225 (39±100), [MH-
TMSOH]‡ 151 (29±100), [MH-TMSOH-C3H6Š‡ 109 (17±81 except for
wOHcP and zOHtP) and other fragments at 169 (19±36 when present) and 123
(17±58).
Compounds identi®ed as metabolites (DHP and 2OHcP) and the trimethylsilyl
derivative of 2OHcP are identical to the standards in tR and give essentially the
same fragmentation patterns (table 3).
Metabolism of cis- and trans-3-pinanones in cytochrome P450 systems
cP and tP are stable on incubation with mouse or human liver microsomes or
human P4503A4 but are metabolized (except for tP with P4503A4) when NADPH
is added, with the extent of metabolism decreasing in the indicated order for P450
source (table 4). cP is much more extensively metabolized than tP in each system
(®gure 3, A and B). In every case there is a single major metabolite and one or two
minor metabolites, each at longer tR than the parent compound (®gure 3, A and B,
and table 4). The major metabolite of cP is identical in GC and MS features when
analysed directly and as the trimethylsilyl derivative to synthetic 2OHcP (table 3)
and the minor metabolites (wOHtP and xOHtP) di€er in chromatographic
properties and fragmentation patterns from synthetic standards of 4SOHcP and
Table 4. Metabolism of cis- and trans-3-pinanones in cytochrome P450 systems.
Relatively peak area (%)b
Substrate and
productsa
Mouse microsomes
Human microsomes
Human P4503A4
Cis-3-pinanone
cP
2OHcP
43 § 1
55 § 5
2 § 4
76 § 13
18 § 6
6 § 7
94 § 3
5 § 1
<1
wOHcP, xOHcPc
Trans-3-Pinanone
tP
yOHtP
zOHtP
86 § 14
3 § 1
92 § 7
1 § 1
7 § 6
100
<1
<1
11 § 14
a See ®gure 1 for structures and abbreviations.
b Area as total ion current of the indicated peak relative to that for the sum of all peaks
…mean § SD, n=3). All metabolism was dependent on NADPH.
c wOHcP and xOHcP appear as approximately equal peak areas.

260
K. M. Hold et al.
Figure 3. Representative GC/MS chromatograms for cis-3-pinanone (cP) and trans-3-pinanone (tP)
and their metabolites (OHcP, OHtP) extracted from the mouse liver microsome-NADPH
system and the brain of the treated mouse (250mg kg¡1, i.p.; 40 min after treatment). (A) cP
substrate, total ion current; (B) tP substrate, total ion current; (C) cP treatment, SIM at m/z 151
and 135, internal standard (is) a-thujone at 5.4min and 7-hydroxy-a-thujone at 8.7 min; (D) tP
treatment, SIM at m/z 151 and 109, internal standards as in (C). All metabolites fall within the
chromatographic region shown. See ®gure 1 for structures and abbreviations.
2OHtP. The hydroxypinanone metabolites of tP (yOHtP and zOHtP) were not
identi®ed.
Metabolites of cis- and trans-3-pinanones in the brain of mouse
GC/MS analysis of the brain of mouse treated with cP reveals the parent
compound, a single major metabolite and two minor metabolites (®gure 3, C).
The major metabolite is identi®ed as 2OHcP by tR and MS fragmentation pattern,
both directly and as the trimethylsilyl derivative. The two minor hydroxypina-
nones in brain, although not identi®ed, are the same ones (wOHcP and xOHcP)
observed in the cytochrome P450 systems (®gure 3, A and C). In a study relating
brain levels to the time after treatment, the parent compound drops to about half
its maximum amount in 10±20 min when 2OHcP reaches its maximum level
(®gure 4). The minor hydroxy metabolites (wOHcP and xOHcP) increase in
amount up to 40±80 min. The mouse at this 250 mg kg¡1 dose of cP with propylene
glycol as the treatment vehicle exhibit neurological poisoning signs throughout the
80-min experimental period.
The level of cP in brain for the ®rst 10 min after treatment greatly exceeds that
of the hydroxypinanone metabolites whereas with a-thujone the 7-hydroxy
metabolite during this period is at a much higher level than the parent compound
(®gure 4) (Hold et al. 2000).

3-Pinanone metabolism and mode of action
261
Figure 4. Brain levels of cis-3-pinanone (cP) and metabolites (2OHcP, wOHcP, xOHcP) as a function
of time following i.p. treatment with cP at 250 mg kg¡1. The curve for minor metabolites wOHcP
and xOHcP is based on the response factor for 2OHcP and because of their small amounts are
increased by a factor of 5 for plotting. The insert ®gure shows corresponding data for a-thujone
(aT) and its principal metabolite 7-hydroxy-a-thujone (7OHaT) after i.p. treatment with a-
thujone at 50 mg kg¡1 (Hold et al. 2000). Each datum point is the average of determinations on
two mice.
tP was also administered i.p. at 250 mg kg¡1 and the brain analysed 40 min after
treatment. The principal GC peaks were identical to those for the P450 metabo-
lites yOHtP and zOHtP (®gure 3, B and D).
Metabolites of cis-3-pinanone in the urine of mouse
The urine, 18 h after oral treatment of mouse with cP at 100 mg kg¡1, contains
DHP in the free but not the conjugate fraction with the same tR and almost
identical MS fragmentation pattern to the standard from synthesis (table 3). The
three hydroxypinanones (2OHcP, wOHcP, xOHcP) noted above for cytochrome
P450 systems in vitro and brain in vivo are also apparent in urine, but only in the
conjugate fraction and not free. An additional metabolite appears as a major
product in the free fraction of urine and the observed m/z 167 suggests that it may
be an aldehyde. It is very minor or absent in the brain or cytochrome P450 systems
examined.
Inhibition of [3H]EBOB binding
The IC50 of hyssop oil and 3-pinanone isomers and derivatives for inhibiting
[3H]EBOB binding in mouse brain membranes increase in the order of cP and tP
(35±36 mm), hyssop oil (64 mm), and DHP and 2OHcP (each 285 mm) (®gure 5 and

262
K. M. Hold et al.
Figure 5. Inhibition of [3H]EBOB binding in mouse brain membranes by hyssop oil and 3-pinanone
isomers, metabolites and analogues. See ®gure 1 for structures and abbreviations. Data shown
are derived from three determinations.
Table 5. Activities of hyssop oil and 3-pinanone isomers and derivatives as inhibitors of [3H]EBOB
binding and toxicants in mouse.
b
Compound
IC50 …mm)a
LD50 …mg kg¡1
†
Hyssop oil and constituents
Hyssop oil
cP
tP
64 § 14
36 § 9
35 § 2
235
175
>250
Dehydro and hydroxy derivatives
DHP
2OHcP
2OHtP
285c
285 § 48
1000d
>250
250
±
a Mean § SEM …n ˆ 4†. Hyssop oil is calculated as the molecular weight of 3-pinanone.
b I.p. treatment as 25% solution (w/v) in dimethyl sulphoxide.
c Extrapolation from 42% inhibition at the highest test concentration of 100mm (solubility
limitation) (®gure 5).
d Highest test concentration of 1000 mm gave a 50% inhibition.
table 5). Although not a metabolite, 2OHtP gives an IC50 ˆ 1000 mm. For
comparison, the IC50 of other monoterpenes is 13 mm for a-thujone and 29 mm
for b-thujone (Hold et al. 2001) and 300±1000 mm for d- and l-camphor, fenchone
and 1,8-cineole (also found in hyssop oil).
Toxicity in mouse
cP was more toxic than tP on i.p. administration to mouse with LD50 ˆ 175
and 250 mg kg¡1, respectively, consistent with cP being the principal acute
>

3-Pinanone metabolism and mode of action
263
toxicant in the hyssop oil examined (LD50 ˆ 235 mg kg¡1) (table 5). Mouse treated
i.p. with cP or hyssop oil at an LD50 dose exhibit neurological signs 1±2 min post-
injection, manifested as tail-raising and forelimb paddling, followed by inter-
mittent periods of dorsal and lateral trunk ¯exing and opisthotonos, progressing to
generalized and unremitting convulsions with running, hopping, chattering and
eventual death after 60±90 min or recovery from poisoning. DHP and 2OHcP were
only slightly less toxic (LD50 ¶ 250 mgkg¡1).
Mouse treated with diazepam (1 mg kg¡1) and 15min later with cP or hyssop oil
at the approximate LD50 dose (which caused neurological signs as above when
administered alone) exhibited depression only and appeared to have recovered
after 1 h.
Discussion
The current understanding of the metabolic fate of cP and tP is summarized in
®gure 1. 2OHcP is the principal metabolite of cP in P450 systems and mouse brain
and it also appears as a conjugate (glucuronide or sulphate) in mouse urine. Two
minor cP metabolites are hydroxypinanones (wOHcP and xOHcP with the
positions of hydroxylation not identi®ed) again appearing in P450 systems, brain
and as conjugates in urine. DHP is present in urine but is not detected in P450
systems as was also the case for 7,8-dehydro-a-thujone as an a-thujone metabolite
(Hold et al. 2001). The ®rst steps have also been taken in comparing the
metabolism of tP with that of cP. tP is converted to one principal hydroxypinanone
(yOHtP) and a minor hydroxypinanone (zOHtP) in P450 systems and brain, with
no common metabolite from cP and tP as currently understood. In each system
examined, cP is much more extensively metabolized than tP, and correspondingly
tP is more persistent. This may be due in part to the 2-position being more
sterically protected in the trans- than the cis-isomer. Studies on the fate of other
monoterpenes (Ishida et al. 1981, 1989, Scheline 1991, Hold et al. 2000, 2001) are
relevant to the site speci®city in oxidation and nature of the terminal metabolites of
the 3-pinanones.
Three lines of evidence indicate that the toxicity of cP, tP and hyssop oil results
from blocking the GABA-gated chloride channel. First, the similar potency of cP
and tP in the [3H]EBOB assay indicates their action at the non-competitive blocker
site of the GABAA receptor. Next, the poisoning signs are the same as those of
picrotoxinin, the classical chloride channel blocker, and of a-thujone and a variety
of other `small cage convulsants’ (Casida et al. 1976, Hold et al. 2000). Finally, as
with picrotoxinin and other GABAA receptor antagonists (Costa et al. 1975,
Klaassen 2001), the GABAA receptor modulator diazepam greatly ameliorates
the poisoning with life-sparing action.
The 2OHcP and DHP metabolites of cP are clearly detoxi®cation products
relative to cP based on reduced potency at the [3H]EBOB binding site and they are
also of lower i.p. toxicity to mouse. While not metabolites in systems examined
here, 2OHtP and 4SOHcP are also of low activity in the [3H]EBOB assay. This
parallels the ®ndings on a- and b-thujone metabolism where all of the hydroxy
metabolites and 7,8-dehydrothujone are less potent inhibitors of [3H]EBOB
binding than the parent compounds (Hold et al. 2000, 2001).
The main constituentof hyssop oil may beeither cP ortPdepending on the source
and puri®cation. The present study indicates that these isomeric 3-pinanones are

264
K. M. Hold et al.
of similar potency at the GABAA receptor with the trans- metabolized more
slowly. cP is more toxic than tP on i.p. administration to mouse and appears to be
the principal acute toxicant in the hyssop oil examined here
Acknowledgements
The work was supported by Grant R01 ES08419 from the National Institute of
Environmental Health Sciences, NIH, and its contents are solely the responsibility
of the authors and do not necessarily represent the ocial views of the NIEHS,
NIH. They thank Gary Quistad of this laboratory for helpful suggestions. The
®rst two authors contributed equally to the study.
References
Alberta Agriculture, 2001, Medicinal, culinary and aromatic crops. Food and rural development
website [http://www.agric.gov.ab.ca/crops/special/hyssop.html].
Aqua-Oleum, 2000, Hyssop oil [http://www.aqua-oleum.co.uk/essentials/long/hyssop.htm].
Badjah-Hadj-Ahmed, A. Y., Meklati, B. Y., Waton, H. and Pham, Q. T., 1992, Structural studies in
the bicyclo[3.1.1]heptane series by 1H and 13C NMR. Magnetic Resonance Chemistry, 30, 807±
816.
Blumenthal, M. (ed.), 1998, The Complete German Commission E Monographs: Therapeutic Guide to
Herbal Medicines (Austin: American Botanical Council), pp. 338±339.
Burkhard, P. R., Burkhardt, K., Haenggeli, C. A. and Landis, T., 1999, Plant-induced seizures:
reappearance of an old problem. Journal of Neurology, 246, 667±670.
Carlson, R. G. and Pierce, J. K., 1971, The synthesis and stereochemistry of the four isomeric pinane-
2,3-diols. Journal of Organic Chemistry, 36, 2319±2324.
Casida, J. E., Eto, M., Moscioni, A. D., Engel, J. L., Milbrath, D. S. and Verkade, J. G., 1976,
Structure-toxicity relationships of 2,6,7-trioxabicyclo[2.2.2]octanes and related compounds.
Toxicology and Applied Pharmacology, 36, 261±279.
Cole, L. M. and Casida, J. E., 1992, GABA-gated chloride channel: binding site for 40-ethynyl-4-n-
[2,3-3H2propylbicycloorthobenzoate ([3H]EBOB) in vertebrate brain and insect head. Pesticide
Biochemistry and Physiology, 44, 1±8.
Cole, L. M., Lawrence, L. J. and Casida, J. E., 1984, Similar properties of
[
35S]t-
butylbicyclophosphorothionate receptor and coupled components of the GABA receptor±
ionophore complex in brains of human, cow, rat, chicken and ®sh. Life Sciences, 35, 1755±1762.
Corey, E. J. and Schmidt, G., 1979, Useful procedures for the oxidation of alcohols involving
pyridinium dichromate in aprotic media. Tetrahedron Letters, 399±402.
Costa, E., Guidotti, A., Mao, C. C. and Suria, A., 1975, New concepts on the mechanism of action of
benzodiazepines. Life Sciences, 17, 167±186.
Crowfoot, G. M. and Baldensperger, L., 1932, From Cedar to Hyssop: A Study in the Folklore of
Plants in Palestine (London: Sheldon and New York: Macmillan), pp. 71±78.
Curci, R., Daccolti, L., Detomaso, A., Fusco, C., Takeuchi, K., Ohga, Y., Eaton, P. E. and Yip, Y.
C., 1993, Selective oxidation of tertiary-secondary vic-diols to a-hydroxy ketones by dioxiranes.
Tetrahedron Letters, 34, 4559±4562.
Duke, J. A., 2001, Syllabus of a mini-course in medical botany. Module 19 [http://www.ars-grin.gov/
duke/syllabus/gras.htm].
Garg, S. N., Naqvi, A. A., Singh, A., Ram, G. and Kumar, S., 1999, Composition of essential oil from
an annual crop of Hyssopus ocinalis grown in Indian plains. Flavour and Fragrance Journal, 14,
170±172.
Gruenwald, J., Brendler, T. and Jaenicke, C. (eds), 1998, PDR for Herbal Medicines (Montvale:
Medical Economics), pp. 908±909.
HOld, K. M., Sirisoma, N. S. and Casida, J. E., 2001, Detoxi®cation of a- and b-thujones (the active
ingredients of absinthe): Site speci®city and species di€erences in cytochrome P450 oxidation in
vitro and in vivo. Chemical Research in Toxicology, 14, 589±595.
HOld, K. M., Sirisoma, N. S., Ikeda, T., Narahashi, T. and Casida, J. E., 2000, a-Thujone (the active
component of absinthe): g-aminobutyric acid type A receptor modulation and metabolic
detoxi®cation. Proceedings of the National Academy of Sciences, USA, 97, 3826±3831.
Ishida, T., Asakawa, Y., Tekemoto, T. and Aratani, T., 1981, Terpenoids biotransformation in
mammals III: biotransformation of a-pinene, b-pinene, pinane, 3-carene, carane, myrcene, and
p-cymene in rabbits. Journal of Pharmaceutical Sciences, 70, 406±415.

3-Pinanone metabolism and mode of action
265
Ishida, T., Toyota, M. and Asakawa, Y., 1989, Terpenoid biotransformation in mammals. V.
Metabolism of (‡)-citronellal, (§)-7-hydroxycitronellal, citral, (¡)-perillaldehyde, (¡)-
myrtenal, cuminaldehyde, thujone, and (§)-carvone in rabbits. Xenobiotica, 19, 843±855.
Klaassen, C. D. (ed.), 2001, Casarett & Doull’s Toxicology, The Basic Science of Poisons, 6th edn (New
York: McGraw-Hill).
Mazzega, M., Fabris, F., Cossu, S., De Lucchi, O., Lucchini, V. and Valle, G., 1999, Synthesis,
atropoisomerism and chemistry of the dimers of R-(¡)-carvone, R-(¡)-hydrocarvone, 1R,5R-
(‡)-pinocarvone and other monoterpenic ketones. Tetrahedron, 55, 4427±4440.
Millet, Y., Jouglard, J., Steinmetz, M. D., Tognetti, P., Joanny, P. and Arditti, J., 1981, Toxicity
of some essential plant oils. Clinical and experimental study. Clinical Toxicology, 18, 1485±1498.
Phytochemical Database, 2001, Chemicals in Hyssop ocinalis L. (Lamiaceae)-Hyssop. In Dr. Duke’s
Phytochemical and Ethnobotanical Databases. USA-ARS-NGRL (Beltsville: Beltsville
Agricultural Research Center) [http://ars-grin.gov/duke].
Satterwhite, D. M. and Croteau, R. B., 1987, Resolution of monoterpene enantiomers by gas
chromatography. Journal of Chromatography, 407, 243±252.
Scheline, R. R., 1991, Handbook of Mammalian Metabolism of Plant Compounds (Boca Raton: CRC
Press), pp. 108±113.
Simonsen, J. and Owen, L. N., 1949, Pinocamphone. In The Terpenes, vol. 2: The Dicyclic Terpenes and
their Derivatives, 2nd edn (Cambridge: Cambridge University Press), pp. 226±232.
Sirisoma, N. S., HOld, K. M. and Casida, J. E., 2001, a- and b-Thujones (herbal medicines and food
additives): synthesis and analysis of hydroxy and dehydro metabolites. Journal of Agricultural
and Food Chemistry, 49, 1915±1921.
SolladiE-Cavallo, A., Schwarz, J. and Burger, V., 1994, A four-step, highly enantioselective
synthesis and enzymatic resolution of 3,4-dichloro-phenylalanine. Tetrahedron: Asymmetry, 5,
1621±1626.
Southall, W., 2001, in D. Bible-Larson, P. Young and D. Lyle (eds), Hyssop Superior Healing Power:
A 21st Century Herb, 2nd edn (Hollywood: Southall Research).
Suga, T., Shishibori, T., Hirata, T. and Matsuura, T., 1968, Stereochemical studies of monoterpene
compounds. VI. The stereochemistry of 2-hydroxypinocamphone and its reduction products.
Bulletin of the Chemical Society of Japan, 41, 1180±1184.
Youssefi, M. J., Boschung, A. F. and Thomas, A. F., 1994, Conformation dependence in the chemical
ionization mass spectra of 3-pinanones. Rapid Communications in Mass Spectrometry, 8, 313±316.