Content uploaded by Stephen L. Clement
Author content
All content in this area was uploaded by Stephen L. Clement on Dec 28, 2013
Content may be subject to copyright.
Bruchins: Insect-derived plant regulators that
stimulate neoplasm formation
Robert P. Doss*
†‡
, James E. Oliver
§
, William M. Proebsting
†
, Sandra W. Potter
¶
, SreyReath Kuy*, Stephen L. Clement
储
,
R. Thomas Williamson**, John R. Carney
††
, and E. David DeVilbiss
§
*Horticultural Crops Research Unit, United States Department of Agriculture, Agricultural Research Service, Corvallis, OR 97330; Departments of
†Horticulture and ¶Zoology, and **College of Pharmacy, Oregon State University, Corvallis, OR 97331; §Insect Chemical Ecology Laboratory, United States
Department of Agriculture, Agricultural Research Service, Beltsville, MD 20705; 储Regional Plant Introduction Station, United States Department of
Agriculture, Agricultural Research Service, Pullman, WA 99164; and ††KOSAN Biosciences, Hayward, CA 94545
Edited by Clarence A. Ryan, Jr., Washington State University, Pullman, WA, and approved March 29, 2000 (received for review February 8, 2000)
Pea weevil (Bruchus pisorum L.) oviposition on pods of specific
genetic lines of pea (Pisum sativum L.) stimulates cell division at the
sites of egg attachment. As a result, tumor-like growths of undif-
ferentiated cells (neoplasms) develop beneath the egg. These
neoplasms impede larval entry into the pod. This unique form of
induced resistance is conditioned by the Np allele and mediated by
a recently discovered class of natural products that we have
identified from both cowpea weevil (Callosobruchus maculatus F.)
and pea weevil. These compounds, which we refer to as
‘‘bruchins,’’ are long-chain
␣
,
-diols, esterified at one or both
oxygens with 3-hydroxypropanoic acid. Bruchins are potent plant
regulators, with application of as little as 1 fmol (0.5 pg) causing
neoplastic growth on pods of all of the pea lines tested. The
bruchins are, to our knowledge, the first natural products discov-
ered with the ability to induce neoplasm formation when applied
to intact plants.
Cell division in plants ordinarily occurs in meristems, and the
newly formed cells differentiate to form plant tissues and
organs (1, 2). In contrast, plant neoplasms, most commonly
represented by various galls (3–5), arise when cell division is
stimulated in nonmeristematic areas (1, 2). We study an inter-
esting phenomenon observed in lines of pea (Pisum sativum L.)
that exhibit the neoplastic pod phenotype, which is conferred by
the wild-type allele, Neoplastic pod (Np), often found in pea
germplasm (6, 7). This phenotype, first noted over 30 years ago,
is typified by the formation of large patches of callus tissue
(neoplasms) on the surface of pods grown under greenhouse
conditions. Greenhouse coverings filter out the UV wavelengths
from sunlight that ordinarily inhibit neoplasm formation (7).
Plants possessing the Np gene grown under unfiltered sunlight in
the field and plants homozygous for np do not form neoplasms.
The neoplastic pod trait remained a poorly understood botanical
curiosity until recently when it was reported that the Np gene
conferred responsiveness to oviposition by the pea weevil (Bru-
chus pisorum L.), an economically important insect pest of pea,
with neoplastic growth at the sites of egg attachment (8–10).
Herein, we show that this response is a previously unidentified
form of induced resistance to an insect, and we identify the
insect-derived chemical signal that stimulates cell division and
the resultant neoplastic growth.
Materials and Methods
Plant Material and Growing Conditions. Lines of pea (P. sativum)
homozygous for either the Np or np allele were used for all
studies. For the field study described below and most bioassays,
the lines used were selected from an F
4
heterozygote from a cross
between C887-332 (Np兾Np) and I
3
(np兾np) (10). Pods to be used
for bioassay were obtained from plants grown in a greenhouse
with set points of 16°C and 21°C. Natural sunlight was supple-
mented with light from high-pressure sodium lamps that were on
from 0600 until 2200 and a bank of eight cool white f luorescent
tubes (F96T12兾CW兾SS) that were on continuously. Such sup-
plemental lighting reduced the incidence of spontaneous neo-
plasm formation relative to that ordinarily seen on greenhouse-
grown plants (7). A field study to compare weevil infestation on
the Np兾Np line with that on the np兾np line was conducted in 1997
with five randomized blocks with 30 plants of each line per block.
Plants were grown in raised beds on the Oregon State University
Campus where they were exposed to a natural population of pea
weevil (B. pisorum), an insect common in this area (11).
Pea Pod Bioassays. Pods in the late flat pod stage (12) were split
along the suture, and the half pod was placed in a Petri dish on
moist filter paper with the outside surface exposed (10). Chro-
matographic fractions, appropriately diluted, and synthetic com-
pounds to be tested for neoplasm-inducing activity were applied
to the pod as 1-
l drops in 50% (vol兾vol) ethanol. The bioassays
were conducted in a growth chamber with continuous f luores-
cent light (30–40 microeinstein m
⫺2
䡠s
⫺1
) and a temperature of
23°C. The neoplastic tissue formed in 1 week was removed with
a scalpel and weighed.
Histology and Microscopy. Tissue to be used for light microscopy
was fixed in 2–2.5% (vol兾vol) glutaraldehyde in 0.1 M phosphate
buffer, dehydrated through an alcohol series, embedded in
paraffin, sectioned at 10
m, and mounted on slides. The
deparaffinized sections were stained with a Lillie–Mayer-type
hematoxylin-eosin procedure. Specimens prepared for scanning
electron microscopy were fixed as described above, critical point
dried, and coated with gold:palladium (60:40, wt兾wt).
Insect Collecting and Rearing. Adult pea weevils were collected
from experimental plantings of peas in eastern Washington,
frozen, and shipped on dry ice to Corvallis, OR. Before extrac-
tion, the male and female insects were separated (13). Sexually
mature female pea weevils were also obtained by collecting
adults as they emerged from infested seed, determining their sex,
and placing the females individually into Petri dishes containing
several detached pea flowers (to serve as a source of pea pollen).
As soon as egg deposition was noted (7–10 days), the insects were
frozen and stored at ⫺20°C.
A culture of cowpea weevils (Callosobruchus maculatus F.)
provided by W. E. Burkholder (Stored Products Insects Re-
search, U.S. Department of Agriculture, Agricultural Research
Service, Madison, WI) was increased and used to inoculate 24
20-liter containers, each holding 500 g of chickpeas (Cicer
arietinum L.). Cultures were held in the dark at room temper-
This paper was submitted directly (Track II) to the PNAS office.
Abbreviation: TMS, tetramethylsilane.
‡To whom reprint requests should be addressed. E-mail: dossr@bcc.orst.edu.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.
§1734 solely to indicate this fact.
Article published online before print: Proc. Natl. Acad. Sci. USA, 10.1073兾pnas.110054697.
Article and publication date are at www.pnas.org兾cgi兾doi兾10.1073兾pnas.110054697
6218–6223
兩
PNAS
兩
May 23, 2000
兩
vol. 97
兩
no. 11
ature. Adult insects were collected (by sieving) twice per week,
frozen, and stored at ⫺80°C. Chickpeas were replenished as
required to maintain the cultures.
The vetch weevil, Bruchus brachialis F., was collected from
wild vetch, Vicia villosa Roth., in eastern Washington (10). The
bruchids Stator limbatus Horn and Stator pruininus Horn were
provided in seeds of cat claw acacia, Acacia greggii A., by C. W.
Fox (University of Kentucky, Lexington, KY).
Extraction and Isolation of Neoplasm-Inducing Compounds. The first
successful isolation began with a total lipid extraction procedure
(14) that yielded9gofaviscousyellow oil from 100 g of adult
cowpea weevils (about 22,000 insects). Flash chromatography
was conducted with2gofthis material with a 5 ⫻15-cm column
of Florisil (Sigma F-9127) and with conditions as described for
separation of simple lipid classes (15). Fractions were obtained
by stepwise elution with hexane; 5, 15, and 30% (vol兾vol) diethyl
ether in hexane; diethyl ether; 2% (vol兾vol) acetic acid in diethyl
ether; and methanol.
An active fraction, 0.28 g of yellow oil, eluted with diethyl
ether and acetic acid in diethyl ether, was passed over a 1 ⫻
30-cm low-pressure, reversed-phase liquid chromatography col-
umn [C-18, J. T. Baker 7025-00, gradient elution starting with
85% (vol兾vol) methanol in water and ending with 100% (vol兾vol)
methanol]; 80 7.5-ml fractions were collected. Bioassay indicated
that activity was present in fractions 53– 67, which were pooled
(0.159 g of yellow oil).
Analysis of 10 mg of the pooled active fraction indicated that
it was comprised largely of free fatty acids that were themselves
inactive but were difficult to separate from the active materials.
Accordingly, the remaining sample was reacted with 2-bro-
moacetophenone (Sigma B3145) to form the phenacyl esters of
the fatty acids (16), thereby changing their chromatographic
behavior and facilitating their separation from the active
materials.
On repeating the low-pressure chromatographic separation
described above with the 2-bromoacetophenone-treated mate-
rial, two active samples were obtained, one eluting in fractions
56–60 (2.2 mg) and the other in fractions 64–67 (55 mg). The
2.2-mg fraction (light yellow oil) was subjected to TLC on a silver
nitrate-impregnated silica gel plate [Whatman K5F; layer thick-
ness ⫽0.25 mm; plate soaked in 12.5% (wt兾vol) AgNO
3
and
activated at 80°C). After development, the plate edges were
removed, sprayed with H
2
SO
4
:methanol (1:1, vol兾vol), and
charred; 11 fractions were taken based on appearance of the
plate edges. Fraction 7 (white solid; R
f
⫽0.38–0.42; 0.4 mg),
which was the most active on the basis of weight, was used for
structure determination.
A refined and scaled-up version of this procedure was used to
prepare additional active compounds starting with a 1,000-g
sample of cowpea weevils. A similar but much scaled-down
procedure was used to prepare active materials from a 4.96-g
sample of field-collected female pea weevils (⬇500 insects). In
this case, HPLC was used instead of low-pressure liquid chro-
matography. The final active fractions obtained f rom pea weevils
contained too little material for accurate weighing but could be
analyzed by GC-MS.
Instrumentation. NMR spectra were obtained on a Bruker DRX
600 spectrometer (Billerica, MA). Spectra were obtained in
deuterochloroform. GC-MS was carried out with a Finnigan-
MAT (San Jose, CA) Incos-50 GC-MS with a short (15-m ⫻
0.25-mm i.d.) fused silica capillary column (DB-5,J&W
Scientific, Folsom, CA). Both electron ionization-MS (70 V;
block source temperature ⫽150°C) and chemical ionization-MS
(ammonia or deuteroammonia as ionization gas; reagent gas
pressure ⫽0.5 torr, 1 torr ⫽133 Pa; block temperature ⫽60°C)
were used.
Derivatizations and Degradation. Hydrolyses were achieved with
8% (wt兾vol) NaOH in 85% (vol兾vol) methanol at 70°C for 30
min. After neutralization with 2 M HCl, solvent was evaporated,
and the residue was dissolved in ethyl acetate and passed through
a very small column of silica gel. The substances in the eluate
were analyzed by GC-MS or were derivatized for further anal-
ysis. Trimethylsilyl ethers were prepared by brief treatment with
N,O-bis[trimethylsilyl]trifluoroacetamide at 60°C. Exhaustive
hydrogenation兾hydrogenolysis of the hydrolysis product was
carried out over LiAlH
4
兾Pt兾Al
2
O
3
at 300°C (17). Ozonolyses
were conducted by collecting ozone in methylene chloride at
⫺78°C and then treating cold methylene chloride solutions of the
substrates with a few microliters of the ozone solution. The
reactions were held outside the cooling bath for a few minutes
and then quenched with dimethylsulfide.
Synthesis. Unsaturated
␣
,
-diols were synthesized by standard
routes involving acetylene alkylations and semihydrogenations
and兾or Wittig condensations. The (3-hydroxypropyl) esters were
initially prepared by oxidative desilylation of 3-(phenyldimeth-
ylsilyl)propanoates as described for 1,(Z)-9-docosene-1,22-diol,
1-(3-hydroxypropanoate)ester (bruchin A, see below); 9-decyn-
1-ol was deprotonated with butyllithium (two equivalents) in
tetrahydrofuran and alkylated with the tetrahydropyranyl ether
of 12-bromododecanol. The product was semihydrogenated
(Lindlar catalyst; cyclohexene as solvent), and the olefinic
alcohol was esterified with the acid chloride obtained by treating
3-(phenyldimethylsilyl) propanoic acid (18) with oxalyl chloride.
After removal of the tetrahydropyranyl group (methanol; tolu-
enesulfonic acid), the resulting monoester was treated with
fluoroboric acid etherate in dichloromethane (room tempera-
ture; 5 h). Flash chromatography on silica gel (increasing por-
tions of ethyl acetate in hexanes) was used to separate (Z)-9-
docosene-1,22-diol (from ester hydrolysis) from the desired
mono 3-(fluorodimethylsilyl)propanoate; the latter was then
stirred at room temperature in methanol-tetrahydrofuran solu-
tion containing sodium bicarbonate, potassium fluoride, and
30% (vol兾vol) hydrogen peroxide. After work-up and f lash
chromatography [40% then 50% (vol兾vol) ethyl acetate in
hexanes], 1(bruchin A) was obtained as a white solid (melting
point 47–48°C; shrink 45°C), after crystallization from heptane.
1
H NMR (CDCl
3
, 600 MHz) d 5.34 (2H, m, olefinic), 4.11 (2H,
t,J⫽4.4 Hz, CH
2
CO
2
R), 3.86 (2H, t, J ⫽3.9 Hz,
-O
2
CCH
2
CH
2
OH), 3.63 (2H, t, J ⫽4.6 Hz, alkyl-CH
2
OH), 2.70
(2H, t J ⫽3.8 Hz, -O
2
CCH
2
CH
2
OH), 2.09 (4H, m, allylic), 1.78
(br. s., OH). Electron ionization-MS, m兾z(in percentages): 412
[M]
⫹
(⬇0.4), 124 (14), 123 (16), 121 (14), 111 (11), 109 (28), 97
(25), 96 (59), 95 (56), 94 (21), 93 (12), 91 (58), 83 (34), 82 (74),
81 (69), 80 (36), 79 (18), 73 (83), 71 (11), 69 (51), 68 (36), 67 (67),
57 (19), 56 (14), 55 (100), 54 (31), 45 (13), 43 (42), 42 (14), 41
(58). 1-bis(Trimethylsilyl) ether, electron ionization-MS m兾z(in
percentages): 556 [M]
⫹
(⬇0.2), 541 (⬇0.2), 235 (10), 219 (22),
163 (31), 149 (10), 147 (100), 109 (13), 105 (28), 103 (79), 97 (12),
96 (14), 95 (20), 91 (12), 83 (17), 82 (14), 81 (23), 75 (58), 73 (57),
69 (24), 67 (23), 55 (41), 43 (15), 41 (16).
Results
Np
Provides Resistance to Pea Weevil. In a field trial with near-
isogenic pea lines exposed to a natural population of pea weevil
in Corvallis, OR, the rate of infestation of np兾np seed was 85.4%
versus 62.2% for Np兾Np seed (P⫽0.0017; blocks ⫽5). Neo-
plastic growth resulting from oviposition on Np兾Np pods was
clearly evident (Fig. 1 aand b).
Bruchids Contain Neoplasm-Inducing Activity. Application of crude
extracts, prepared with either freshly killed or frozen pea weevil
adults, to Np pea pods resulted in neoplastic growth (Fig. 1c; ref.
10). Extracts of either pea weevil eggs or accompanying fluid
Doss et al. PNAS
兩
May 23, 2000
兩
vol. 97
兩
no. 11
兩
6219
PLANT BIOLOGY
were also mitogenic (10). The cowpea weevil, a bruchid that was
reared more easily than the pea weevil, also yielded extracts that
induced formation of neoplasms. In contrast to the pea weevil,
where earlier work had demonstrated that sexually mature
female insects were much richer sources of neoplasm-inducing
activity than were males or immature females (10), newly
emerged female or male cowpea weevils were equally good
sources of activity. Consequently, this insect was used for initial
isolation of neoplasm-inducing compounds. The sensitive and
reliable Np pod assay (10) was used to guide fractionation of
weevil extracts.
Characterization of Neoplasm-Inducing Compounds in Cowpea Weevil.
The mass spectrum of the first compound isolated (compound 1;
Fig. 2) contained a barely detectable (0.4%) molecular ion. In
many respects, it resembled the spectrum of oleyl alcohol with
the addition of the prominent ions m兾z73 and 91. Chemical
ionization with ammonia produced an ion with m兾z430, indi-
cating a molecular weight of 412. Chemical ionization-MS with
deuteroammonia demonstrated the presence of two exchange-
able hydrogens, and formation of a bis-tetramethylsilane (bis-
TMS) derivative on treatment with N,O-bis[trimethylsilyl]trif lu-
oroacetamide was consistent with this conclusion.
On alkaline hydrolysis, 1gave a new compound with mo-
lecular weight 340 whose mass spectrum lacked the ions at m兾z
73 and 91. Exhaustive hydrogenation兾hydrogenolysis (17)
yielded n-docosane, n-heneicosane, and n-eicosane, indicating
that the hydrolysis product was a docosene-1,22-diol. Indeed,
catalytic hydrogenation (Pd兾C; 1 atm; 1 atm ⫽101.3 kPa)
provided docosane-1,22-diol. Ozonolysis gave rise to C
9
and
C
13
hydroxyaldehydes, indicating that the diol possessed a 9-10
double bond.
The fact that the intact natural product contained two active
hydrogens, one of which must have been present in the fragment
lost during hydrolysis, suggested that the original compound
could be a monoester of either lactic acid or the much less
common 3-hydroxypropanoic acid. Published spectra of lactic
acid esters did not contain the prominent m兾z73 and 91 ions, and
spectra of esters of 3-hydroxypropanoic acid were not available.
Accordingly, a small sample of the hydroxypropanoate of 1-de-
canol was prepared as a model compound. Its mass spectrum
contained prominent m兾z73 and 91 ions, suggesting that the
natural product was a mono 3-hydroxypropyl ester of the 9-do-
cosene-1,22-diol.
A 1-(3-hydroxypropanoate) ester with a (Z) double bond at
the 9 position was prepared as illustrated in Fig. 3; a key
element in the synthesis was the use of Fleming’s masked
hydroxy technology (19) for the construction of the 3-hy-
droxypropanoate group. We initially had no way of knowing
which of the two OHs was esterified, nor did we have any
information on the geometry of the double bond. We felt,
however, that the natural product was of lipid origin and that
the double-bond geometry would, accordingly, be (Z). (More
detailed synthetic details appear in J.E.O., R.P.D., R.T.W.,
J.R.C., and E.D.D., unpublished work.) The synthetic 1ex-
hibited the same mass spectrum as the natural product, and GC
retention times of the TMS derivatives were identical. Subse-
quent NMR experiments also confirmed the identity of the
natural and synthetic compounds and supported the (Z)
geometry assigned to the double bond. Synthetic 1was as
active in the pea pod bioassay as was the isolated compound.
Fig. 1. Stimulation of cell division on pods of an Np pea line (a derivative of
C887-332; ref. 10) by several treatments. (a) Scanning electron micrograph
showing the response of a pod from an Np pea line to oviposition by a pea
weevil. The pod was obtained from a field-grown plant several days after
oviposition. E, egg; N, neoplastic tissue formed in response to oviposition. (b)
Cross section through a pod showing neoplastic tissue formed in response to
pea weevil oviposition. Pod was harvested and fixed 8 days after oviposition.
(c) Neoplasms present 1 week after application of various amounts of 2
(bruchin B; Fig. 4a), a compound present in extracts from the cowpea and pea
weevil. Amounts applied as 1-
l drops in 50% (vol兾vol) ethanol were (from
left) 10, 5, 1, 0.5, and 0.0 pg. (Bars ⫽100
m.)
Fig. 2. Bruchin A, a monoester bruchin isolated from a lipid extract of
cowpea weevils.
Fig. 3. Synthesis scheme for (Z)-9-docosene-1,22-diol, 1-(3-hydroxypropano-
ate)ester (bruchin A), 1.
6220
兩
www.pnas.org Doss et al.
A second, somewhat refined, bioassay-guided fractionation of
a 1,000-g sample of insects yielded three major active compounds
(2–4, Fig. 4). All possessed the m兾z73 and 91 ions diagnostic for
3-(hydroxypropyl) esters. All formed bis-TMS ethers whose mass
spectra contained ions with m兾zof 103, 145, 147, and 163.
[Although the first three of these ions are not uncommon in
spectra of TMS ethers, particularly of polyfunctional compounds
(20), collectively the four proved useful for selected ion moni-
toring-based identifications of TMS derivatives of both mono-
and bis-3-hydroxypropyl esters.] Molecular weights from chem-
ical ionization-MS were 628, 656, and 654. Hydrolyses were
conducted as described above, and the molecular weight 628
compound (2) provided the same C
22
diol characterized earlier.
Hydrolysis of 3and 4gave monosaturated and diunsaturated C
24
diols, respectively. These data, collectively, suggested that the
active compounds were bis-3-(hydroxypropyl) diesters of
␣
,
-
diols. Ozonolyses of the C
24
diols demonstrated that the mono-
unsaturated compound, as was the case with the C
22
diol
analyzed earlier, possessed a double bond at C
9
, whereas the
diunsaturated diol had double bonds at C
9
and C
17.
The C
24
diols,
each with (Z) double-bond geometry, were also synthesized with
standard acetylene alkylations and兾or Wittig condensations.
Their NMR spectra confirmed the structural assignments, in-
cluding the (Z) configurations for all double bonds. These three
diesters were more abundant than the monoester characterized
from the first isolation, and we believe that the bis-3-
hydroxypropyl esters are the principal neoplasm-inducing
agents. A more detailed discussion of the characterization and
synthesis of 2–4(bruchins B, C, and D, respectively) is included
elsewhere (J.E.O., R.P.D., R.T.W., J.R.C., and E.D.D., unpub-
lished work).
Neoplasm-Inducing Compounds in Pea Weevil. GC-MS analysis of
fractionated whole-body extracts of adult female pea weevils
identified the same three bis-3-(hydroxypropyl) esters (2–4)
identified from the cowpea weevil. Monoester 1, if present, was
far less abundant.
Neoplasm Formation in Response to Synthetic Compounds. Synthetic
versions of all three diesters (J.E.O., R.P.D., R.T.W., J.R.C., and
E.D.D., unpublished work) were as active as the natural products
in the pea pod bioassays (Fig. 4). As little as 1 fmol, ⬇0.5 pg,
when applied to an Np pea pod, resulted in neoplasms large
enough to remove from the pods and weigh after 1 week.
Physiological Activity of Bruchins. We propose the name, ‘‘bruchins’’
to refer to these neoplasm-inducing long-chain diols esterified at
one or, more commonly, both oxygen atoms with 3-hydroxypro-
panoic acid. Bruchin application to Np pods causes browning of
tissue at the treatment site within 3– 6 h after initial exposure to
the chemical. Swelling, resulting from mitosis in cells underlying
the epidermis, is visible within 24–48 h (Fig. 5; ref. 10).
Neoplasms, weighing from several hundred micrograms to sev-
eral milligrams, are formed after 5–7 days (10). Bruchins do not
cause callus formation when applied to leaves or stems of pea.
Bruchin application also stimulates browning and swelling on
pods homozygous for the np allele; however, in this case, much
of the swelling results from cell enlargement rather than cell
division, and neoplasms are usually too small to remove and
weigh (Fig. 5). Pods of np兾np plants typically fail to respond to
pea weevil oviposition, although barely detectable swelling may
occur under some eggs.
Discussion
It has been suggested, without direct evidence, that Np is a
source of resistance to the pea weevil, a monophagous bruchid
(8, 9, 21) that is one of the most serious insect pests of peas
worldwide (22). Our use of isogenic lines confirmed such
resistance even with large weevil populations. Neoplastic
growth seems to provide resistance by reducing larval sur vival
(8, 9, 21). Ordinarily, pea weevil larvae burrow through the
ventral surface of the egg and directly into the pod to reach the
immature seed (11). In contrast, eggs on Np pods are displaced
from the pod surface by a mound of neoplastic tissue (Fig. 1
aand b). This mound causes the larvae to wander about before
Fig. 4. Structures and activities of the principal bruchins present in sexually
mature female pea weevils. Structures of bruchins B, C, and D are shown in a,
b, and c, respectively. Neoplasms (calli) resulting from application of the
indicated amounts of the bruchins were removed from pods with a scalpel and
weighed 1 week after treatment (10). Each bar represents the mean ⫾SEM for
six pods (blocks). Pea plants (derived from line C887-332; ref. 10) homozygous
for the Np gene were grown in a greenhouse with light provided by a
combination of high-pressure sodium lamps and fluorescent tubes. Bruchins
were applied as 1-
l drops in 50% (vol兾vol) ethanol to pods at the late flat pod
stage (12).
Doss et al. PNAS
兩
May 23, 2000
兩
vol. 97
兩
no. 11
兩
6221
PLANT BIOLOGY
attempting to burrow through the pod wall, thus exposing them
to environmental hazards including predators, parasites, and
desiccation (9, 21). Moreover, the neoplasms with attached
eggs are sometimes sloughed off of the pod before larval
emergence (8).
We originally attempted to isolate neoplasm-inducing com-
pounds from pea weevil but were unable to obtain sufficient
numbers of this insect to provide enough active material for
chemical characterization. Newly emerged female pea weevils
must ingest fresh pollen to become sexually mature (10, 13), and
only limited numbers of egg-bearing females could be collected
from the field. Fortunately, the easily reared cowpea weevil
provided a suitable alternative. Purified fractions were obtained
by using this insect, and the distinctive mass spectrum of
3-(hydroxypropanoic) esters provided the key to identifying the
bruchins.
Among the four bruchins described in this report, the mo-
noester (Fig. 2) was the first identified, not because it was more
abundant, but because it was the first active material obtained in
a sufficiently pure state to allow characterization. In fact,
subsequent isolations demonstrated that the diester bruchins
(Fig. 4) were more abundant than the monoester.
Thus far, bruchins have been identified in adult insects of
two bruchid genera. We have also detected strong neoplasm-
inducing activity in crude extracts of the three other bruchid
species that have been tested, namely: the vetch weevil, B.
brachialis (10); S. limbatus; and S. pruininus. Moreover, we
found that application of bruchin B (compound 2of Fig. 4) to
pods of Lathyrus tingitanus (Tangier peavine; PI 493288;
‘‘Raiano’’) results in neoplasm formation. L. tingitanus is one
of three legume species for which neoplastic growth in re-
sponse to bruchid oviposition has been reported (8 –10, 23).
Similarly, oviposition on pods of certain lines of common bean,
Phaseolus vulgaris L., by the bean-pod weevil, Apion godmani
Wagner, a nonbruchid Coleopteran, results in callus formation
that inhibits insect infestation (24). Hence, several legume
species possess similar resistance mechanisms. However,
bruchins have been identified only from pea weevil and
cowpea weevil, and it is possible that unrelated compounds
could induce the same response with peas as well as with other
legumes.
It is noteworthy that pods of all pea lines tested, both Np and
np, responded to bruchin application. The response of the np
lines was attenuated, however. Only slight swelling occurred,
even when large amounts of bruchin were applied (Fig. 5). We
assume that the failure of np pods to react strongly to pea
weevil oviposition is a result of this weak response and the low
bruchin dose delivered with the egg and accompanying f luid.
We believe the bruchins to be the first regulators isolated from
natural sources that stimulate neoplasm formation in intact
plants. In peas possessing the Np allele, bruchins mediate a
sensitive and efficient form of induced resistance wherein mi-
tosis is stimulated and neoplasms form only in tissue in contact
with the insect egg. Because bruchins are synthesized by the pea
weevil at the expense of increased lar val mortality, they must be
‘‘. . . the focus of intense selection pressure. . . ’’ (25). Their
presence in the insect despite this pressure suggests that they play
an important, but as yet undefined, role in the bruchid life cycle.
There have been several reports of host-marking pheromones
associated with the Bruchidae, including C. maculatus (26).
However, preliminary tests indicate that bruchins do not play
such a role in this insect.
The bruchins are relatively stable, low-melting, white solids,
sparingly soluble in common solvents. Structurally, these
long-chain
␣
,
-diols, monoesterified or diesterified with 3-hy-
droxypropanoic acid, represent a previously unknown class of
compounds. In contrast to 3-hydroxybutanoic acid (27), its
relatively important higher homolog, 3-hydroxypropanoic
acid, was previously unknown as an element of natural prod-
ucts. The free acid is a hygroscopic liquid that is normally
found only as a hydrate (28) and is mentioned only infre-
quently in the literature.
Preliminary results indicate that at least one 3-hydroxypro-
panoate functionality is required for activity. For example, the
mono- and bis-(3-hydroxypropanoate) esters, bruchins A and B
(1and 2) have approximately equivalent activity, but the
␣
,
-diol
itself is inactive. Although unsaturation in the diol chain is not
required and the bis (3-hydroxypropanoate) ester of docosane-
1,22-diol is fully active, the bis-propanoate, bis-lactate, and
bis-(3-hydroxybutanoate) esters of this diol are virtually inactive.
Further investigation will be necessary to identify the structural
requirements for activity.
The ability of insects to manipulate lipids chemically for their
own purposes is well established (29). It is noteworthy that the
four bruchins characterized thus far resemble the common
unsaturated fatty acids in possessing a (Z)-9 double bond.
Despite the widely separated double bonds in bruchin D (4) that
are in contrast to the usual 1,3 arrangement seen in polyunsat-
urated fatty acids, it seems likely that the bruchins are offshoots
of fatty acid synthesis or metabolism. If so, the bruchins would
join a slowly growing group of insect-produced difunctional fatty
acid derivatives with unprecedented structures such as the
defensive nitrogen-containing macrocycle from Epilachna
varivestis recently reported by Attygalle et al. (30) and the
remarkable volicitin from Pieris brassicae caterpillars. Volicitin,
like the bruchins, initiates a complex plant-signaling sequence
(31) that ultimately has a negative effect on the insect that
produced the chemical.
We thank J. K. Christian and H. Throop for technical support and John
Fellman for review of the manuscript. This work is technical paper 11642
of the Agricultural Experiment Station, Oregon State University.
1. Doonan, J. & Hunt, T. (1996) Nature (London) 380, 481– 482.
2. Hemerly, S. S., Ferreira, P. C. G., VanMont agu, M. & Inze´, D. (1999) BioEssays
21, 39– 47.
3. Meyer, J. (1987) Plant Galls and Gall Inducer s (Gebru¨der Borntraeger, Berlin).
4. Shorthouse, J. D. & Rohr fritsch, O. E., eds. (1992) Biology of Insect-Induced
Galls (Oxford Univ. Press, Oxford).
Fig. 5. Response of pea pods homozygous for either the Np or the np gene
96 h after application of 100 pg of 2, (bruchin B). The difference between np
(aand b) and Np (cand d) pods, reflected as variation in the amount of cell
division stimulated by the bruchin, is clearly visible. Arrows indicate margins
of bruchin application, and brackets (aand c) indicate areas enlarged in band
d. Note the difference between the two phenotypes in the number of small,
newly formed cells. Also note the influence of bruchin treatment on meso-
phyll cell size in treated and untreated regions of the np pods. The np pods
were from a line, I3, derived from the common cultivar ‘‘Alaska.’’ All of the
more than 20 np lines that we have tested show a similar response. The Np
pods were from a line derived from a cross between I3and line C887-332 (10).
(Bars ⫽1
m.)
6222
兩
www.pnas.org Doss et al.
5. Williams, M. J., ed. (1994) Plant Galls: Organisms, Interactions, Populations
(Clarendon, Oxford).
6. Dodds, K. S. & Matthews, P. (1966) J. Hered. 57, 83–85.
7. Snoad, B. & Matthews, P. (1969) in Chromosomes Today, eds. Parlington, C. D.
& Lewis, K. R. (Oliver Boyd, Edinburgh), pp. 126 –131.
8. Berdnikov, V. A., Trusov, Y. A., Bogdanova, V. S., Kosterin, O. E., Rozov,
S. M., Nedel’kina, S. V. & Nikulina, Y. N. (1992) Pisum Genet. 24, 37–39.
9. Hardie, D. (1993) Ph.D. thesis (Universit y of Adelaide, Adelaide, Australia).
10. Doss, R. P., Proebsting, W. M., Potter, S. W. & Clement, S. L. (1995) J. Chem.
Ecol. 21, 97–106.
11. L arson, A. O., Brindley, T. A. & Hinman, F. G. (1938) Technical Bulletin (U.S.
Dept. Agric., Washington, DC) Publ. No. 599.
12. Meicenheimer, R. D. & Muehlbauer, F. J. (1982) Exp. Agric. 18, 17–27.
13. Pesho, G. R. & VanHouten, R. J. (1982) Ann. Entomol. Soc. Am. 75, 439 – 443.
14. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37, 911–917.
15. Christie, W. W. (1973) Lipid Analysis (Pergamon Press, Oxford).
16. Wood, R. & Lee, T. (1983) J. Chromatogr. 254, 237–246.
17. Bierl-Leonhardt, B. A. & DeVilbiss, E. D. (1981) Anal. Chem. 53, 936–938.
18. Takeshita, K., Seik, Y., Kawamoto, K., Murai, S. & Sonoda, N. (1987) J. Or g.
Chem. 52, 4864– 4868.
19. Fleming, J. (1988) Pure Appl. Chem. 60, 71–78.
20. Halket, J. M. (1993) in Handbook of Der ivatives for Chromatography, eds. Lau,
K. & Halket, J. M. (Wiley, New York), pp. 297–325.
21. Clement, S. L., Cristofaro, M., Cowgill, S. E. & Weigand, S. (1999) in Global
Plant Genetic Resources for Insect Resistant Crops, eds. Clement, S. L. &
Quisenberry, S. S. (CRC, Boca Raton, FL), pp. 131–148.
22. Clement, S. L. (1992) Entomol. Exp. Appl. 63, 115–121.
23. Annis, B. & O’Keeffe, L. E. (1984) Entomol. Exp. Appl. 35, 83–87.
24. Garza, R., Cardona, C. & Singh, S. P. (1996) Theor. Appl . Genet. 92, 357–362.
25. Farmer, E. E. (1997) Science 276, 276–277.
26. Credmore, P. F. & Wright, A. W. (1990) Physiol. Entomol. 15, 285–298.
27. Mu¨ller, H.-M. & Seeback, D. (1992) Angew. Chem. Int. Ed. Engl. 32, 477–502.
28. Read, D. D. (1964) Or ganic Synthesis (Wiley, New York), Vol. I, pp. 321–324.
29. St anley-Samuelson, D. W. & Nelson, D. R., eds. (1994) Insect Lipids: Chemistry,
Biochemistry, and Biology (Un iversity of Nebraska Press, Lincoln, NE).
30. Attygalle, A. B., Blankespoor, C. L., Eisner, T. & Meinwald, J. (1994) Proc.
Natl. Acad. Sci. USA 91, 12790 –12793.
31. Pare´, P. W., Alborn, H. T. & Tumlinson, J. H. (1998) Proc. Natl. Acad. Sci. USA
95, 13971–13975.
Doss et al. PNAS
兩
May 23, 2000
兩
vol. 97
兩
no. 11
兩
6223
PLANT BIOLOGY
