(E4,Z9)-Tetradecadienal, a sex pheromone for three North American moth species in the genus Saturnia

Article abstract of DOI:10.1023/A:1010310204552

The lepidopteran genus Saturnia has three representatives in North America, S. walterorum, S. mendocino, and S. albofasciata. (E4,Z9)-Tetradecadienal (E4,Z9-14:Aid) was identified as a sex pheromone component for all three species by combinations of coupled gas chromatography-electroantennogram detection (GC-EAD), GC-mass spectrometry (MS), and field trials. In field trials, all three species were strongly attracted to (E4,Z9-14:Ald) as a single component, Small amounts of (Z)-9-tetradecenal (Z9-14:Ald) also were found in extracts of all three species, but blends of this compound with E4,Z9-14:Aid were no more attractive to male moths than E4,Z9-14:Ald alone. Extracts of pheromone glands of female S. walterorum occasionally contained a third, trace compound eliciting responses from male antennae in GC-EAD experiments, but this compound was not identified. It is suggested that the three species can use the same, single component as a sex attractant because the flight period of S. albofasciata (fall) is different than that of the other two species (spring), whereas the geographic distributions of S. mendocino and S. walterorum overlap over only small portions of their ranges. Furthermore, the latter two species readily hybridize, so there may be minimal fitness cost to cross-attraction.

Full text of DOI:10.1023/A:1010310204552

Journal of Chemical Ecology, Vol. 27, No. 4, 2001
(
E4,Z9)-TETRADECADIENAL, A SEX PHEROMONE FOR
THREE NORTH AMERICAN MOTH SPECIES
IN THE GENUS Saturnia
1
1,
J. STEVEN MCELFRESH, JOCELYN G. MILLAR, *
and DANIEL RUBINOFF2
1Department of Entomology
University of California
Riverside California 92521
2Division of Insect Biology
Department of Environmental Science, Policy & Management
University of California, Berkeley California 94720
(Received July 10, 2000; accepted December 14, 2000)
Abstract—The lepidopteran genus Saturnia has three representatives in North
America, S. walterorum, S. mendocino, andS. albofasciata. (E4,Z9)-Tetradecadi-
enal (E4,Z9–14 : Ald) was identified as a sex pheromone component for all three
species by combinations of coupled gas chromatography–electroantennogram
detection (GC-EAD), GC–mass spectrometry (MS), and field trials. In field tri-
als, all three species were strongly attracted to (E4,Z9–14 : Ald) as a single
component. Small amounts of (Z)-9-tetradecenal (Z9–14 : Ald) also were found
in extracts of all three species, but blends of this compound with E4,Z9–14 : Ald
were no more attractive to male moths than E4,Z9–14 : Ald alone. Extracts of
pheromone glands of female S. walterorum occasionally contained a third, trace
compound eliciting responses from male antennae in GC-EAD experiments, but
this compound was not identified. It is suggested that the three species can use
the same, single component as a sex attractant because the flight period of S. al-
bofasciata (fall) is different than that of the other two species (spring), whereas
the geographic distributions of S. mendocino and S. walterorum overlap over
only small portions of their ranges. Furthermore, the latter two species readily
hybridize, so there may be minimal fitness cost to cross-attraction.
KeyWords—(4E,9Z)-tetradecadienal, (Z)-9-tetradecenal, Saturniaspp., pheromone.
*
To whom correspondence should be addressed. e-mail: jocelyn.millar@ucr.edu
7
91
0
098-0331/01/0400-0791$19.50/0 �C 2001 Plenum Publishing Corporation

7
92
MCELFRESH, MILLAR, AND RUBINOFF
INTRODUCTION
The genus Saturnia (Lepidoptera: Saturniidae) is represented by three species in
the New World, Saturnia mendocino Behrens, S. walterorum Hogue and Johnson,
and S. albofasciata (Johnson), all of whose distributions and biologies have been
described by Tuskes et al. (1996). All three species are univoltine. S. mendocino
and S. walterorum are sister species that are found in both montane and coastal
chaparral plant communities, whereas S. albofasciata is restricted to montane cha-
parral. S. mendocino ranges from central California to Oregon and is found in
both the coastal and Sierra Nevada mountain ranges, whereas S. walterorum is
found from southern California to northern Baja California, Mexico. The ranges
of the latter two species overlap in San Luis Obispo and Monterey counties of
California. The range of S. albofasciata spans that of the other two species, and it
is found at higher elevations from northern California to northern Baja California,
Mexico. Adults of S. mendocino and S. walterorum emerge from February through
May and are diurnally active. In contrast, S. albofasciata flies from mid-October
to mid-November, with males being active in the late afternoon until dusk, and
females ovipositing just after sunset. Eggs of all three species are laid singly or in
small clusters on the larval hosts, which include manzanita (Arctostaphylos spp.
Adanson) and madrone (Arbutus menziesii Pursh) for S. mendocino, manzanita
(Arctostaphylos spp.) and laurel sumac (Malosma laurina Nuttall) for S. wal-
terorum, and several Ceanothus spp. L. and Cercocarpos betuloides Nuttall for
S. albofasciata (Tuskes et al., 1996). Eggs of S. mendocino and S. walterorum
hatch in approximately three weeks, whereas those of S. albofasciata overwinter
and hatch the following spring. Larvae of all three species develop through the
spring, spin silk cocoons on or near the host, and subsequently emerge in the fall
of that year for S. albofasciata, or in the spring of the following year for the other
two species.
In conjunction with an ongoing study of sex attractant pheromones of sat-
urniid moth species in the western United States, we undertook a study of the sex
pheromone chemistry of these three Saturnia spp. We describe here the identifi-
cation and field testing of (E4,Z9)-tetradecadienal (E4,Z9–14 : Ald) as a female-
produced sex pheromone component for all three species.
METHODS AND MATERIALS
Insects
Pupae of S. mendocino were obtained from a culture maintained by one
of the authors (D.R.) at the Division of Insect Biology, University of Califor-
nia, Berkeley, California. This culture had been outcrossed frequently by mating

SEX PHEROMONE OF Saturnia SPP.
793
laboratory-reared females with wild males collected from several northern Cali-
fornia counties, so as to maintain the vigor of the laboratory colony. Specimens
of S. walterorum were obtained from C. Conlan and were from cultures that orig-
inated in San Diego County, California. Larvae of both species were reared on cut
bouquets of natural hosts in water or in nylon sleeves directly upon living plants.
Cocoons were kept outdoors in screen cages (approx. 30 cm square) in the shade
until adult emergence. Freshly emerged adults were separated by sex and were
either used immediately for preparation of pheromone gland extracts (females) or
�
for electroantennograms (males’ antennae), or were held at 4–10 C in glassine
envelopes for up to one week. Storing at cool temperatures was necessary because
the nonfeeding adult moths typically live less than one week.
A few pupae of S. albofasciata were obtained from C. Conlan, and these
produced a single female and several males. Further males used in GC-EAD studies
were collected from field sites, using E4,Z9–14 : Ald lures to attract them (see
below).
Extraction of Sex Pheromone Glands
Sex pheromone glands were removed from 1- to 4-day-old calling virgin fe-
males. The gland was extruded by applying gentle pressure to the abdomen to
evert the ovipositor, and the terminal abdominal segments were excised into a
0
.25-ml conical glass vial insert containing � 30 �l pentane. Care was taken to
avoid contamination of the samples with the dark-colored material that was abun-
dant in the abdomen, which may be used to glue eggs to the host plant. The glands
were extracted for 2–20 min, and the extract was transferred to a clean insert,
with two rinses with 1 drop of pentane. Extracts were concentrated by passive
evaporation as needed prior to analysis. If not used immediately, extracts were
�
stored at � 20 C. Approximately 20 extracts of female S. walterorum were ob-
tained, whereas fewer than 10 S. mendocino female extracts and only a single
S. albofasciata extract were available.
Gas Chromatography
GC analyses were conducted with Hewlett-Packard 5890A or 5890 Series II
gas chromatographs (H-P = Agilent, Palo Alto, California), with helium carrier
gas, in splitless mode. Columns and programs used to analyze female extracts
of S. walterorum included DB-225 (J&W Scientific, Folsom, California; 30 m �
� � � �
.26 mm ID, 0.25-�m film, 100 C/0 min, 15 C/min to 130 C/0 min, 3 C/min to
0
1
3
�
�
�
75 C/0 min, then 20 C/min to 220 C for 60 min), DB-WAX (J&W Scientific;
�
�
�
0 m � 0.32 mm ID, 0.25-�m film, 100 C for 1 min, 5 C/min to 240 C for 10 or
�
more min), DB-5 (J&W Scientific; 30 m � 0.25 mm ID, 0.25-�m film, 100 C for

7
94
MCELFRESH, MILLAR, AND RUBINOFF
�
�
1
min, 5 C/min to 275 C for 60 min). S. mendocino extracts were analyzed on the
DB-225 column as described above. Extracts also were analyzed by coupled GC-
EAD (see below). Tentative identifications were made by comparison of retention
times of compounds in extracts to those of synthetic standards.
Coupled Gas Chromatography–Electroantennography (GC-EAD)
Antennae from male S. mendocino (N = 3) were stimulated with standards
of the E and Z isomers of each of the monounsaturated tetraenals with the double
bond in positions 3–12. One-microliter aliquots containing approx. 50 ng of a given
standard in hexane were injected in random order onto a DB-5 column (30 m �
0
2
.32 mm, 0.25-�m film) in split mode with a 20 : 1 split ratio, run isothermally at
00 C. Antennal response peak heights were measured to the nearest 0.5 mm, then
�
analyzed using two-way analysis of variance of the log transformed data followed
by a Student-Neumann-Keuls means separation test. Because these standards were
the products of microscale oxidations and partial purifications, their concentrations
varied. Concentrations of individual samples were adjusted to keep the amount of
each compound injected within a twofold range, but no further compensation was
made for variability in amounts injected because antennal responses were relatively
insensitive to variations in the quantity injected. For example, halving the amount
of Z9–14 : Ald injected resulted in only a 27% reduction in antennal response.
Aliquots of pheromone gland extracts from females and synthetic standards
were analyzed by splitless coupled GC-EAD, using an H-P 5890 series II GC
�
equipped with a DB-5 column (30 m � 0.32 mm ID, 0.25-�m film, 100 C/1 min,
�
� � �
5
C/min to 150 C/5 or 10 min, then 25 C/min to 275 C/30 min for S. waltero-
rum and S. mendocino, and 100 C/1 min, then 10 C/min to 275 C/25 min for
S. albofasciata), using an injector temperature of 250 C. The column effluent was
�
�
�
�
split equally with a press-fit Y-connector (J&W Scientific) into two branches of
0.25-mm-ID, uncoated fused silica tubing, with one branch going to the GC flame
ionization detector (FID), and the other being directed through a heated trans-
�
fer line (275 C) to a 15-mm-diameter glass stimulus delivery tube. The capillary
effluent was diluted with humidified air (400 ml/min) and passed over the male
moth antennal preparation (see below). The moth antenna was suspended between
glass capillary electrodes filled with modified Locke’s saline (7.5 g NaCl, 0.21
CaCl2, 0.35 g KCl, 0.20 g NaHCO3 in 1 liter distilled water) (Humason, 1972)
or Kaissling’s saline (Kaissling and Thorson, 1980), with electrical contact to the
custom-built amplifier made with AgCl-coated silver wires. The amplifier and FID
outputs were simultaneously recorded on a matched pair of H-P 3394 recording
integrators.
Male antennae were removed using fine forceps to firmly hold the scape and
pull the antenna free of the head, without crushing the scape. The terminal rami and
tip of the antenna were removed so that the end of the antenna could be placed into

SEX PHEROMONE OF Saturnia SPP.
795
the saline-filled recording electrode. The antenna was positioned on the amplifier
mount at the end of the delivery tube such that the plane of the antennal branches
was perpendicular to the air flow from the stimulus delivery tube. A single antenna
preparation could be used for several hours with periodic additions of saline.
Field Trials
Pherocon 1C sticky traps (Tr e´ c e´ Inc., Salinas, California) were used in all
trials, with 11-mm red rubber septum lures (Wheaton Scientific, Millville, New
Jersey), loaded with heptane solutions (100 �l) of test compounds, plus 2 drops
of 10 mg/ml butylated hydroxytoluene (BHT) in heptane as an antioxidant. Lures
werekeptin20-mlglassvialsinafreezeroranice-chestwhennotinuse. Trapswere
placed in areas where host plants were common. Traps were spaced approximately
8
m apart and blocks were 600 m or more apart. Traps were suspended from shrubs
at heights of 0.5–1.5 m above ground. Trapped moths were counted at irregular
intervals of 3–11 days for S. mendocino and S. walterorum, with traps left in the
field for nearly one month in total.
S. mendocino field trials were conducted in the western Sierra Nevada moun-
tains near Hams Station, Amador County, California in April 1997 using 10 blocks
of treatments and controls. A second S. mendocino trial, comparing a blend of
E4,Z9–14 : Ald and Z9–14 : Ald (100 : 10 �g) to 100 �g E4,Z9–14 : Ald alone,
was conducted in Del Puerto Canyon, Stanislaus County, California in March
1998. S. walterorum field trials were conducted in late March to April 1997 with
five blocks of traps placed in the vicinity of Viejas grade near Descanso, and the
Laguna Mountains along Highway S1 between Monument peak and Kwaaymii
Point, San Diego County, California. S. albofasciata field trials were conducted
in November 1998 near Thomas Mountain, Riverside County, California. A sin-
gle block of traps was used, consisting of three traps baited with 100 �g E4,Z9–
14 : Ald and three traps baited with blank controls. Traps were left in place for one
day, after which time the treatment trap bottoms were completely covered with
male moths and so further testing was discontinued.
For the purposes of statistical analysis, a replicate consisted of a single block,
and counts for a particular treatment in a given block were pooled to eliminate
day effects. Trap catches were analyzed using two way analysis of variance of the
square root (x + 0.5) transformed counts.
Synthesis of Authentic Standards
Standards of tetradecenals with the double bonds at carbons 3–12 were pre-
pared by oxidation of samples of the corresponding alcohols, obtained from the
Research Institute for Plant Protection, Wageningen, The Netherlands, with excess
pyridinium dichromate in methylene chloride for 2 hr (Corey and Suggs, 1979).

7
96
MCELFRESH, MILLAR, AND RUBINOFF
The reaction mixtures were diluted with 10 volumes of pentane, then vortexed
for 30 sec. The precipitate was allowed to settle, and the supernatant solutions
were filtered through 2.5-cm plugs of Celite. The resulting solutions were used
immediately for electroantennogram studies as described above.
4,9-Tetradecadienals
Proton NMR spectra were taken on a General Electric QE-300 instrument
(300 MHz), as CDCl3 solutions. Electron impact mass spectra (70 eV) were taken
with a Hewlett-Packard 5970B mass selective detector interfaced to a 5890 gas
chromatographfittedwithaDB-5MScolumn(20m� 0.2mmID, J&WScientific).
Mass spectra are reported as m/z (abundance). All reactions were carried out in
oven-dried glassware. THF was purified by distillation from sodium benzophenone
ketyl under nitrogen. Unless otherwise specified, worked up reaction mixtures
were dried with anhydrous Na2SO4 and concentrated by rotary evaporation under
reduced pressure. Flash chromatography (Still et al., 1978) and vacuum liquid
chromatography (Coll and Bowden, 1986) were carried out with 230 to 400-mesh
silica gel (Aldrich Chemical Co.).
Synthesis of (E4,Z9)-Tetradecadienal (Scheme 1)
1-Bromonon-4-yne (3). Butyllithium (42.5 ml of 2.5 M solution in hexanes,
1
1
2
1
10 mmol) was added dropwise over � 20 min to an icebath-cooled solution of
-hexyne 1 (8.2 g, 100 mmol) in THF (100 ml) under Ar. The mixture was stirred
0 min, then warmed to room temperature, and 1,3-dibromopropane 2 (12.2 ml,
�
20 mmol) was added dropwise. The mixture was warmed to 55 C for 24 hr,
then cooled and quenched with water (250 ml). The mixture was extracted with
hexane, and the hexane extracts were washed with brine, dried, and concentrated,
yielding a mixture of unreacted starting material, the desired bromide 3, and the
alkadiyne from coupling at both ends of 1,3-dibromopropane (16 : 66 : 12 by
GC). The residue was slowly Kugelrohr distilled, yielding 20.11 g of a purified
�
monobromide fraction, oven temperature 55–60 C (0.4 mm Hg). NMR: � 3.54
SCHEME 1

SEX PHEROMONE OF Saturnia SPP.
797
(
2
t, 2H, J = 6.9 Hz), 2.34 (tt, 2H, J = 6.7, 2.1 Hz), 2.15 (tt, 2H, J = 6.7, 2.3 Hz),
.02 (quint, 2H, J = 6.6 Hz), 1.52–1.35 (m, 4H), 0.91 (t, 3H, J = 6.9 Hz). MS:
+
m/z 202/204 (1/1, M ), 174/176 (8/7), 132/134 (31/30), 123 (5), 107 (8), 95 (67),
81 (99), 79 (58), 77 (49), 67 (78), 53 (69), 41 (100).
(
Z)-1-Bromo-4-nonene (4). Cyclohexene (8.3 ml, 82 mmol) was added drop-
�
wise to BH3-dimethylsulfide complex (4.0 ml, 40 mmol) in 50 ml of THF at 0 C
under Ar. The mixture was warmed to room temperature for 2 hr, then cooled to
�
0
3
C again, followed by dropwise addition of 4.06 g (20 mmol) of bromoalkyne
to the white slurry. The mixture was warmed to room temperature, stirred 2 hr,
then cooled in an icebath while glacial acetic acid (16 ml) was added. The mixture
was stirred overnight, then cooled in an icebath during sequential slow addition of
5
M NaOH (60 ml) and 30% H2O2 (16 ml; Caution: Exothermic). The resulting
mixture was diluted with water, extracted twice with hexane, and the hexane ex-
tracts were washed with brine, dried, and concentrated. The more volatile fraction
of the residue (� 4 g) was removed by careful Kugelrohr distillation, oven temper-
�
ature 30–35 , � 0.4 mm Hg. The remaining liquid (� 6 g) was purified by vacuum
flash chromatography in a 150-ml sintered glass funnel, eluting with hexane. The
�
purified bromoalkene was Kugelrohr distilled (oven temperature � 50 C, 0.4 mm
Hg), yielding 2.22 g of 4 as a colorless oil. The yield was not optimized, and the
volatile fraction removed before chromatography contained approx. 25% of the
bromoalkene. NMR: � 5.49–5.62 (m, 2H), 3.42 (t, 2H, J = 6.5 Hz), 2.20 (br. quart,
2
H, J = 7.3 Hz), 2.1–2.0 (m, 2H), 1.92 (quint, 2H, J = 6.9 Hz), 1.4–1.28 (m, 4H).
MS: m/z 204/206 (4/4, M+), 162/164 (4/4), 148/150 (9/9), 135 (4), 109 (4), 95 (9),
3 (20), 81 (21), 69 (46), 55 (78), 41 (100).
Z)-9-Tetradecen-4-yn-1-ol (6). Butyllithium (6.4 ml of a 2.5 M solution
8
(
in hexanes; 16 mmol) was added to a solution of THP-protected pentynol 5
�
(
2
2.54 g, 15 mmol) in THF (25 ml) at � 10 C, under Ar. The solution was stirred
0 min, followed by sequential addition of 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-
pyrimidinone (DMPU, 8 ml) and bromoalkene 4 (2.05 g, 10 mmol). The reaction
was warmed to room temperature and stirred overnight, then poured into water
and extracted with hexane. The hexane extract was washed with dilute HCl, water,
and brine; dried; and concentrated. The residue was taken up in 25 ml MeOH, and
p-toluenesulfonic acid was added until the solution became acidic to moistened
litmus paper. The mixture was stirred at room temperature until the transacetaliza-
tion reaction was complete (� 2 hr), the reaction was quenched by addition of 1 g
NaHCO3, and the solvent was removed on a rotary evaporator. The residue was
partitioned between hexane and water, and the hexane extract was washed with
brine, dried, and pumped under vacuum (0.1 mm Hg) to remove the THP methyl
ether byproduct. The crude enynol 6 (96% pure by GC) was used without further
purification. NMR: � 5.46–5.28 (m, 2H), 3.77 (t, 2H, J = 6.1 Hz), 2.29 (tt, 2H,
J = 6.8, 2.3 Hz), 2.19–1.95 (m, 6H), 1.75 (quint, 2H, J = 6.4 Hz), 1.54 (quint, 2H,
J = 7.3 Hz), 1.39–1.25 (m, 4H), 0.91 (t, 3H, J = 6.9 Hz). MS: m/z 164 (2), 149

7
98
MCELFRESH, MILLAR, AND RUBINOFF
(
7
8), 137 (16), 121 (17), 119 (18), 108 (44), 107 (21), 105 (22), 93 (49), 91 (50),
9 (61), 67 (59), 55 (65), 41 (100).
E4,Z9) -Tetradecadienol (7). LiAlH4 (2 g, 53 mmol) was added in portions
(
to an ice-cooled mixture of dry diglyme (20 ml) and THF (2.5 ml) under Ar. When
the vigourous foaming had subsided, 2 g of the crude enynol 6 (� 9.6 mmol) in
5
ml diglyme was added to the thick slurry. When the foaming had subsided,
�
the mixture was heated to 110 C overnight. The next morning, the mixture was
cooled to 0 C, then hexane (150 ml) was added, followed sequentially by cautious
�
dropwise addition of 2.1 ml water (Foams!), 1.6 ml 20% NaOH, and 7.5 ml water.
The mixture was stirred for 15 min to allow the precipitate to granulate, then
the hexane layer was decanted, and the residue was rinsed several times with
aliquots of hexane. The combined hexane solutions were washed several times
with water and once with brine, then dried and concentrated. The residue was
purifed by flash chromatography (2.5 � 25 cm), eluting with a stepwise gradient of
�
1
5–25% EtOAc in hexane. The purifed product was Kugelrohr distilled (bp � 80 C
at 0.1 mm Hg), yielding 1.1 g of (E4,Z9)-tetradecadienol 7. NMR: � 5.48–5.41 (m,
2
2
H), 5.41–5.32 (m, 2H), 3.66 (t, 2H, J = 6.5 Hz), 2.12–1.94 (m, 8H), 1.64 (quint,
H, J = 6.9 Hz), 1.40 (quint, 2H, J = 7.5 Hz), 1.36–1.25 (m, 4H), 0.91 (t, 3H,
+
J = 6.9 Hz). MS: m/z 210 (trace, M ), 138 (5), 135 (4), 121 (6), 110 (9), 95 (22),
82 (25), 81 (50), 79 (59), 67 (61), 55 (62), 41 (100).
(
E4,Z9)-Tetradecadienal (8). Dienol 7 (0.42 g, 2 mmol) was added to a slurry
˚
of pyridinium dichromate (2.26 g, 6 mmol) and powdered 3A molecular sieve (2 g)
in CH2Cl2 (12 ml) (Herscovici et al., 1982). The mixture was stirred vigourously
for 30 min, then diluted with 50 ml hexane. After stirring for 5 min, the mixture
was filtered, decolorized with � 2 cc decolorizing charcoal, filtered again, and
concentrated. A few milliliters of toluene were added to the residue, which was
concentrated again to azeotrope out traces of pyridine. The clear liquid was then
�
Kugelrohr distilled (oven temperature 80–105 C, 0.1 mm Hg), yielding 325 mg
of aldehyde 8, >98% pure by GC. The aldehyde was diluted with hexane to a
concentration of 20 mg/ml, butylated hydroxytoluene was added (� 10 mg), and
�
the solution was stored at � 20 C under N2. NMR: � 9.77 (t, 1H, J = 1.2 Hz),
5
2
1
.54–5.28 (m, 4H), 2.50 (tt, 2H, J = 6.9,1.2 Hz), 2.34 (br. quart., 2H, J = 6.5 Hz),
+
.0 (m, 6H), 1.46–1.28 (m, 6H), 0.91 (t, 3H, J = 7.0 Hz). MS: m/z 208 (trace, M ),
64 (3), 151 (2), 135 (3), 121 (4), 110 (10), 95 (19), 82 (33), 81 (34), 67 (60), 55
(71), 41 (100).
Synthesis of (Z4,Z9)-Tetradecadienal (11) (Scheme 2)
,9-Tetradecadiynol (9). Butyllithium (10 ml of 2.5 M solution in hexanes,
5 mmol) was added dropwise to an icebath-cooled solution of 4-pentyn-1-ol THP
4
2
ether 5 (4.2 g, 25 mmol) in 25 ml THF under Ar. The mixture was stirred 10 min,
then bromoalkyne 3 (5.1 g, 25 mmol) was added. The mixture was warmed to

SEX PHEROMONE OF Saturnia SPP.
799
SCHEME 2
�
5
5 C and stirred overnight. GC analysis showed considerable quantities of starting
material remaining. However, further heating resulted in no change in the product–
starting material ratio so the mixture was worked up by quenching with water
and extracting with hexane. The hexane extracts were washed with brine, dried,
and concentrated. The residue was transacetalated by treatment with MeOH and
PTSA as described above. The crude diynol 9 (3 g) was purified by vacuum flash
chromatography in a 150-ml sintered glass funnel, eluting with 20% EtOAc in
hexanes. The purified alcohol was Kugelrohr distilled (oven temperature � 100–
�
1
2
1
20 C, 0.1 mm Hg), yielding 2.1 g of clear oil. NMR: � 3.76 (t, 2H, J = 6.2 Hz),
.33–2.22 (m, 6H), 2.14 (tt, 2H, J = 7.0, 2.1 Hz), 1.74 (quint, 2H, J = 6.3 Hz),
.67 (quint, 2H, J = 7.0 Hz), 1.52–1.33 (m, 4H), 0.91 (t, 3H, J = 7.1 Hz). MS:
m/z 177 (3), 163 (35), 150 (11), 135 (16), 131 (14), 119 (27), 117 (29), 105 (40),
91 (100), 79 (57), 67 (41), 55 (45), 41 (94).
(
Z4,Z9)-Tetradecadienol (10). Ni(OAc)2 � 4H2O (150 mg, 0.6 mmol) was
dissolved in 20 ml 95% EtOH (Brown and Ahuja, 1973). The solution was purged
with Ar, then 0.6 ml of a filtered 1 M solution of NaBH4 (prepared by dissolving 0.2
g NaBH4 in 4.75 ml EtOH and 0.25 ml 2 M aq. NaOH) was added by syringe. After
bubbling ceased, the reaction vessel was flushed with H2 and maintained under H2
for the duration of the reduction. Ethylene diamine (0.1 ml) was added, followed
5
min later by addition of diynol (1.04 g, 5 mmol). The mixture was stirred 2 hr,
monitoringbyGC, thenpouredinto300mlof0.2MHClandextractedwithhexane.
The hexane extracts were washed with water and brine, dried, concentrated and
�
Kugelrohr distilled (oven temperature � 110–120 C, 0.1 mm Hg), yielding 0.85 g
of dienol 10. NMR: � 5.48-5.33 (m, 4H), 3.67 (t, 2H, J = 6.5 Hz), 2.12 (br. quart,
2
2
H, J = 7.5 Hz), 2.10–1.98 (m, 6H), 1.64 (quint, 2H, J = 7.0 Hz), 1.42 (quint,
H, J = 7.6 Hz), 1.38–1.28 (m, 4H), 0.91 (t, 3H, J = 6.8 Hz). MS: m/z 210 (trace,
+
M ), 138 (5), 135 (4), 121 (8), 110 (8), 95 (27), 82 (27), 81 (49), 79 (58), 67 (65),
5
5 (70), 41 (100).
Z4,Z9)-Tetradecadienal (11). (Z4,Z9)-tetradecadienol 10 (20 mg) was oxi-
dized to the aldehyde as described above for the E4,Z9-isomer. The crude aldehyde
1 was purified by passage through a short column of silica gel in a disposable
(
1
pipet, eluting with ether. NMR: � 9.78 (t, 1H, J = 1.4 Hz), 5.49–5.29 (m, 4H),
2
1
.50 (tt, 2H, J = 7.3,1.3 Hz), 2.38 (br quart, 2H, J = 7.0 Hz), 2.11–1.99 (m, 6H),
.48–1.20 (m, 6H), 0.91 (t, 3H, J = 7.0 Hz).

8
00
MCELFRESH, MILLAR, AND RUBINOFF
Scrambling of Double Bonds of (Z4,Z9)-Tetradecadienol (Sonnet, 1974).
(Z4,Z9)-tetradecadienol 10 (0.21 g, 1 mmol) was mixed with 52 �l 2M aq. NaNO2
and 36 �l 6M HNO3 in a heavy-wall conical bottomed vial, and the mixture was
�
flushed with Ar, sealed, and heated 45 min at 75 C with stirring. The cooled mix-
ture was extracted with hexane, and the hexane solution was washed with dilute
NaHCO3, water, and brine; simultaneously dried and decolorized with a mixture
of charcoal and anhyd. Na2SO4; filtered; and concentrated. The crude mixture was
used in the next step.
4,9-Tetradecadienal Isomer Mixture. The mixture of 4,9-tetradecadienol iso-
mers was oxidized as described above for (E4,Z9)-tetradecadienal. In GC analyses
of the two pure isomers and the scrambled mixture, all four isomers were dis-
�
tinguishable on the DB-5 and DB-Wax columns using a 5 C/min temperature
program. The isomers were identified by comparison with the two authentic stan-
dards (Z4,Z9 and E4,Z9), and on the assumption that the most abundant isomer in
the scrambled mixture would be the most thermodynamically stable E4,E9-isomer.
By elimination, the fourth isomer was Z4,E9–14 : Ald.
RESULTS
Pheromone gland extracts of females contained only a single compound that
elicited large and consistent responses from male antennae in coupled GC-EAD
experiments, for all three species (Figure 1). The compound was the same for all
FIG. 1. Coupled gas chromatogram–electroantennograms of pheromone gland extracts
from single females and antennae from males of : (A) Saturnia walterorum, (B) S. mendo-
cino, and (C) S. albofasciata. Column: DB-5 (30 m � 0.32 mm, 0.25-�m film; program
�
�
�
for S. walterorum, 100 C for 0 min, 15 C/min to 275 C for 25 min; for S. mendocino,
� � � � �
1
00 C for 1 min, 15 C/min to 150 C for 5 min, then 25 C/min to 275 C for 40 min; for
�
�
�
S. albofasciata, 100 C for 1 min, 10 C/min to a final temperature of 275 C for 25 min. The
male S. albofasciata was field collected using a lure containing E4,Z9–14 : Ald at Thomas
Mountain, Riverside County, California.

SEX PHEROMONE OF Saturnia SPP.
801
three species, based on identical retention times on polar and nonpolar columns.
On a nonpolar DB-5 column, the retention time was slightly shorter than that of a
Z9–14:Aldstandard. Furthermore, duringpreliminarytrialstodeterminethereten-
tion characteristics of the antennal stimulatory compound on various GC columns,
it was discovered that the Z9–14 : Ald standard elicited strong responses from
male moth antennae, providing circumstantial evidence that the insect-produced
compound was probably a 14-carbon aldehyde, possibly with a double bond in
the 9 position. The fact that the retention time of the insect-produced compound
was slightly shorter than that of Z9–14 : Ald eliminated the possibility that the
compound might be an �, �-unsaturated aldehyde or a conjugated dienal, because
both of these structural units result in significant increases in retention times on
DB-5 columns versus a monoenal (Marques et al., 2000). The mass spectrum also
suggested a C14 dienal, with a small molecular ion at m/z 208, a small ion at m/z
+
1
8
90 from loss of water from the M ion, and abundant ions at m/z 67 (100%),
1 (70%), 95 (43%), and 110 (30%), characteristic of unsaturated hydrocarbon
fragment ions.
Further GC-EAD studies were carried out with the complete series of (Z)-
and (E)-tetradecenal isomers with double bonds in the C3 to C12 position, using
S. mendocino as the test species. These experiments provided compelling evidence
for a Z double bond at C9, with male moth antennae being strongly stimulated by
the Z9–14 : Ald standard (Figure 2). However, the slightly shorter retention time
of the insect-produced compound (10.32 min) versus the Z9–14 : Ald standard
(10.62 min) on the nonpolar DB-5 column, and the slightly longer retention time
(11.35 min) versus Z9–14 : Ald (11.19 min) on a polar DB-225 column suggested
the presence of a second, unconjugated double bond. The most likely position
and geometry appeared to be E4, based on the GC-EAD trials with tetradecenal
standards (Figure 2), but the antennal responses to E4-14 : Ald were much smaller
than those to the Z9–14 : Ald standard. On the assumption that alkenes with double
bonds in the 7, 8, 10, and 11 positions could be excluded because this would result
in conjugated dienes with significantly longer retention times than the monoenes;
a second analysis of treatments 3, 4, 5, 6, and 12 (Figure 2) showed that response
to E4–14 : Ald was significantly larger than the responses elicited by the other
possible isomers with double bonds at positions 3, 5, 6, and 12.
Comparison of the retention times of authentic E4,Z9–14 : Ald on polar and
nonpolar GC columns with those of the other three stereoisomers confirmed that
E4,Z9–14 : Ald was the isomer in the female extracts of all three species. On
�
DB-5 all four isomers were resolved using a 5 C/min ramp (ZZ = 15.47, ZE =
1
5.51, EZ = 15.59, and EE = 15.72 min), and an exact retention time match with
E4,Z9–14 : Ald was obtained upon coinjection of standards and an S. walterorum
female extract. On DB-Wax, all four isomers also were resolved (ZE = 10.89,
ZZ = 10.94, EZ = 11.02, and EE = 11.11 min), and the retention time of the
insect-produced compound exactly matched that of E4,Z9–14 : Ald. Final proof of

8
02
MCELFRESH, MILLAR, AND RUBINOFF
FIG. 2. Mean (� SE) electroantennographic responses from antennae of three male S. men-
docino to monounsaturated tetradecenals. Data were analyzed first by two-way ANOVA
of all data points (log10 x + 1 transformed) (bond position effect, F = 20.55, df = 19,38,
P < 0.0001; for male effect F = 9.91, df = 2,38, P = 0.0003) followed by a Student-
Neuman-Kuels means separation test (� = 0.05); bars surmounted by different lowercase
letters were significantly different. Second, the data were reanalyzed, assuming that one
double bond was at C-9, and that the compound was not conjugated or an allene, so that
only the C-3–C-6 and the C-12 positions were included in the two-way ANOVA (for bond
position effect F = 13.4, df = 8,26, P < 0.0001; for male effect F = 26, df = 2,26, P =
0
.0001) and Student-Neuman-Kuels means separation test (� = 0.05, bars surmounted by
different uppercase letters are significantly different).
the structure was obtained from match-up of the electron impact mass spectrum of
the compound with that of the standard and by the strong responses of male moths
to E4,Z9–14 : Ald in field trials.
In addition to the dienal, pheromone gland extracts of all three species con-
tained small amounts of Z9–14 : Ald (compound 2 in Figure 1A–C). The antennal
responses to this compound were not resolved from those of the slightly earlier
eluting E4,Z9–14 : Ald. The identification was confirmed by matching of reten-
tion times on three GC columns, and matching of the mass spectra with that of an
authentic standard. Although the molecular ion at m/z 210 was not seen, there was
a significant ion at m/z 192 from loss of water from the parent ion. Furthermore,
match-up of retention times on both DB-5 and DB-Wax columns excluded all
other monoene aldehyde positional and geometrical isomers from consideration
(Marques et al., 2000). The relative amount of Z9–14 : Ald as a percentage of
E4,Z9–14 : Ald for S. walterorum was 9.7 � 2.6 % (mean of eight single-female
extracts) and 10.9% for a four-female composite extract, whereas the amounts
found in single extracts of S. mendocino and S. albofasciata females was 9.3%
and 19.5%, respectively.

SEX PHEROMONE OF Saturnia SPP.
803
FIG. 3. ResultsofS. mendocinoandS. walterorumfieldtrialsusingblendsofE4,Z9–14:Ald
100 �g) and Z9–14 : Ald as trap lures. S. mendocino field trials were conducted in Amador
(
County, California from April 27 to May 10, 1997, with a total of 381 moths trapped. Two-
way ANOVA for Z9–14 : Ald dose effect, F = 2.733, df = 4,16, P = 0.0633, for block effect,
F = 0.909, df = 4,16; P = 0.4824. S. walterorum field trials conducted in San Diego County,
California from March 27 to April 24, 1997, total of 149 moths trapped. Two-way ANOVA
for Z9–14:Ald dose effect, F = 0.128, df = 4,16; P = 0.9702, for block effect, F = 4.203,
df = 4,16; P = 0.0163.
A second EAD response was seen sporadically to a trace component in
S. walterorum extracts. This compound was not identified, but comparisons of
its retention time on a DB-5 column with those of standards determined that it
was possibly E4,Z9–14 : OH or a geometrical isomer, a compound that might be
expected to be present as a precursor to E4,Z9–14 : Ald.
Field Trials. Sticky traps baited with synthetic E4,Z9–14 : Ald (100 �g) were
attractive to male S. mendocino and S. walterorum moths (Figure 3). For both
species, addition of Z9–14 : Ald did not significantly increase the attractiveness of
blends relative to the diene alone. These results were corroborated for S. mendocino
in a subsequent field trial in Del Puerto Canyon, in which there was no difference
in attractiveness between baits containing E4,Z9–14 : Ald alone and baits with
a 10 : 1 blend of E4,Z9–14 : Ald + Z9–14 : Ald (2-way ANOVA, for blend
effect: F = 0.0101, df = 1,9, P = 0.9221, for block effect F = 1.2018, df = 9,9,
P = 0.3944, based on totals of 100 moths in the treatment with Z9–14 : Ald, and
90mothsinthetreatmentwithoutZ9–14:Ald). Thus, theavailabledatasuggestthat
Z9–14 : Ald is not a pheromone component for these two species.
In a single field trial with S. albofasciata, three traps baited with E4,Z9–14 :
Ald (100 �g) caught a total of 151 moths in one night whereas the control traps
caught no moths. This number may underestimate the attractiveness of the lures
because the sticky trap surfaces were completely covered with moths. In addition,

8
04
MCELFRESH, MILLAR, AND RUBINOFF
males showed no hesitation in approaching lures, and, when in close proximity,
were observed to initiate copulatory attempts with lures, further suggesting that
the pheromone consists of a single component.
DISCUSSION
There have been few reports of 14-carbon 4,9-diene compounds as lepi-
dopteran pheromone components, and even in those few reports, details have
been sketchy. For example, Bestmann et al. (1987a) identified E4,Z9–14 : Ac
in pheromone gland extracts from the silk moth Antheraea pernyi, and located a
large-spiking antennal cell that was stimulated by this compound. However, the
effects of this compound on male moth behavior were not reported. A second study
by Bestmann et al. (1987b) examining the effects of differing functional groups on
the responses of single receptor cells used both E4,Z9–14 : Ac and E4,Z9–14 : Ald,
but in neither case were the syntheses of the compounds reported. Tomida et al.
(1987) prepared all four 4,9–14 : OH isomers and tested the formate derivatives in
laboratory bioassays as mimics of E6,Z11-16 : Ald, a sex pheromone component of
the eri-silk moth Philosamia cynthia ricini; no synthetic details were given. Other
syntheses of E4,Z9–14 : OH (Sharma et al., 1990; Du et al., 1994) and Z4,Z9–14
:
OH (Ducoux et al., 1992) have been published, but there have been no other
reports of biological activity associated with any C14 4,9-dienes as lepidopteran
pheromone components.
The numbers of moths of each of the three Saturnia species available for study
varied. Pheromone gland extracts were obtained from 21 female S. walterorum and
five female S. mendocino, providing sufficient material for GC and GC-EAD anal-
yses on several columns and GC-MS studies. In contrast, only a single extract from
a female S. albofasciata was obtained, due to the relative scarcity of this species
and difficulties in rearing it. This extract provided only sufficient material for
GC-EAD analysis on a DB-5 column. Nevertheless, given that the retention time
exactly matched that of E4,Z9–14 : Ald and that all four 4,9–14 : Ald stereoiso-
mers were separable under the conditions used, there can be little doubt that
E4,Z9–14 : Ald is indeed a sex pheromone component for this species.
All three species appear to use the same, single compound as their pheromone,
leading in theory to the possibility of interspecific interference in the pheromone
channels. However, the flight period of S. albofasciata is in the fall, whereas the
other two species fly in the spring, precluding any interference between
S. albofasciata and the other two species. Furthermore, experimental manipulation
of S. walterorum to encourage unnatural fall emergence of adults by exposing pu-
pae to an artificially lengthened cold period (November of one year until October
of the next) resulted in S. walterorum males that were attracted to and attempted to
mate with S. albofasciata females and vice versa (K. Wolfe and C. Conlan, personal

SEX PHEROMONE OF Saturnia SPP.
805
communication, 2000), despite gross differences in size. No copulations occurred
due to the physical incompatibility of the two species. Nevertheless, these results
suggest that separation in time provides an effective barrier to interspecific mating
between these two species, allowing them to share the same sex pheromone with
no fitness cost. For the other two species, only small parts of their ranges overlap,
and the two species are closely related and known to hybridize (Tuskes and Collins,
1981), so the costs of cross-attraction may be minimal. Overall, it is clear that the
pheromone chemistry is highly conserved within the North American members of
the genus.
Furthermore, there are obvious parallels with the pheromone chemistry of
other members of the Saturniinae subfamily. For example, the pheromone blends
of Antherea pernyi Gu e´ rin-M e´ neville (Bestmann et al., 1987a) and the congeneric
A. polyphemus Cramer (Kochansky et al., 1975) are reported to consist of
E6,Z11–16 : Al and the corresponding acetate, both of which contain the same
diene motif as E4,Z9–14 : Ald, but with two more carbons in the chain. E4,Z9–14 :
Ac also was reported from A. pernyi gland extracts. In another genus and tribe,
the pheromone blend of Samia cynthia ricini Donovan consists of E4,Z6,Z11–
16 : Ald and E4,Z6,Z11–16 : Ac, again with two of the three double bonds in the
same position relative to the nonfunctionalized end of the chain. Finally, Bestmann
et al. (1982) mentioned that E6,Z11–16 : Ac was attractive to Eudia (= Saturnia)
pavonia, but provided no further details.
In conclusion, we have presented spectroscopic, retention time, and field
trapping data that indicate that the three North American Saturnia species share
the same sex pheromone component, E4,Z9–14 : Ald. Possible negative effects of
cross-attraction between species are largely avoided by temporal and geographic
barriers that keep the species apart in space and time.
Acknowledgments—We thank Christopher Conlan and Kirby Wolfe for their generous donations
of insects and for sharing unpublished information used in this study.
REFERENCES
BESTMANN, H.-J., KOSCHATSKY, K. H., andVOSTROWSKY, O. 1982. StereoselektiveSynthesenvon1,5-
und1,6-Alkadienen als Sexualpheromone und Pheromonanaloge. Liebigs Ann. Chem. 1982:1478–
1
498.
BESTMANN, H.-J., ATTYGALLE, A. B., BROSCHE, T., ERLER, J., PLATZ, H., SCHWARZ, J., VOSTROWSKY,
O., WU, C.-H., KAISSLING, K. E., and CHEN, T.-M. 1987a. Identification of three sex pheromone
components of the female saturniid moth Antheraea pernyi (Lepidoptera: Saturniidae). Z. Natur-
forsch. 42c:631–636.
BESTMANN, H.-J., WU, C.-H., DOHLA, B., LI, K., and KAISSLING, K.-E. 1987b. Functional group
recognition of pheromone molecules by sensory cells of Antheraea polyphemus and Antheraea
pernyi (Lepidoptera: Saturniidae). Z. Naturforsch. 42c:435–441.

8
06
MCELFRESH, MILLAR, AND RUBINOFF
BROWN, C. A., and AHUJA, V. K. 1973. P-2 nickel catalyst with ethylenediamine, a novel system for
highly stereospecific reduction of alkynes to cis-olefins. J. Chem. Soc. Chem. Commun. 1973:553–
5
54.
COLL, J. C., and BOWDEN. B. F. 1986. The application of vacuum liquid chromatography to the
separation of terpene mixtures. J. Nat. Prod. 49:934–936.
COREY, E. J., and SUGGS, J. W. 1975. Pyridinium chlorochromate. An efficient reagent for oxidation
of primary and secondary alcohols to carbonyl compounds. Tetrahedron Lett. 1975:2647–2650.
DU, G., NG, C.-S., and PRESTWICH, G. D. 1994. Odorant binding by a pheromone binding protein:
Active site mapping by photoaffinity labelling. Biochemistry 33:4812–4819.
DUCOUX, J.-P., LE MENEZ, P., KUNESCH, N., KUNESCH, G., and WENKERT, E. 1992. An efficient
and stereoselective synthesis of insect pheromones by way of nickel-catalyzed Grignard reac-
tions. Syntheses of gossyplure and pheromones of Eudia pavonia and Drosophila melanogaster.
Tetrahedron 48:6403–6412.
HERSCOVICI, J., EGRON, M.-J., and ANTONAKIS, K. 1982. New oxidative systems for alcohols: molec-
ular sieves with chromium-(VI) reagents. J. Chem Soc. Perkin Trans. I 1982:1967–1973.
HUMASON, G. L. 1972. Animal Tissue Techniques, 3 ed. W. H. Freeman and Company. San Francisco.
KAISSLING, K.-E., and THORSON, J. 1980. Insect olfactory sensilla: Structural, chemical and electrical
aspects of the functional organization, pp. 261–282, D. B. Sattelle, L. M. Hall, and J. G. Hildebrand
(eds.), in Receptors for Neurotransmitters, Hormones, and Pheromones in Insects, Elsevier/North
Holland Biomedical Press, New York.
KOCHANSKY, J., TETTE, J., TASCHENBERG, E. F., CARDE´, R. T., KAISSLING, K.-E., and ROELOFS, W.
1
975. Sex pheromone of the moth Antheraea polyphemus. J. Insect. Physiol. 21:1977–1985.
MARQUES, F. A., MCELFRESH, J. S., and MILLAR, J. G. 2000. Kovats retention indexes of monoun-
saturated C12, C14, and C16 alcohols, acetates, and aldehydes commonly found in lepidopteran
pheromone blends. J. Braz. Chem. Soc. 11:592–599.
SHARMA, S., SABHARWAL, A., VOHRA, N., SHARMA, M. L., VIG, O. P., and KAD, G. L. 1990. Syntheses
of the sex pheromone components of wild silk moths, fruit fly, melon fly, and olive fly. J. Indian
Chem. Soc. 67:31–33.
SONNET, P. E. 1974. A practical synthesis of the sex pheromone of the pink bollworn. J. Org. Chem.
3
9:3793–3794.
STILL, W. C., KAHN, M., and MITRA, A. 1978. Rapid chromatographic technique for preparative
separations with moderate resolution. J. Org. Chem. 43:2923–2925.
TOMIDA, I., IWAKOSHI, K., KONO, S., TAKAHASHI, T., MATSUNAGA, T., and KAYAHARA, H. 1987.
Flutter-stimulating activities on male Eri-silk moths, Philosamia cynthia ricini, for four geometric
isomers of 4,9-tetradecadienyl formate. Agric. Biol. Chem. 51:939–940.
TUSKES, P. M., and COLLINS, M. M. 1981. Hybridization of Saturnia mendocino and S. walterorum,
and phylogenetic notes on Saturnia and Agapema (Saturniidae). J. Lepid. Soc. 35:1–21.
TUSKES, P. M., TUTTLE, J. P., and COLLINS, M. M. 1996. The Wild Silk Moths of North America:
A Natural History of the Saturniidae of the United States and Canada. Cornell University Press,
Ithaca, New York.