Jasmonate-dependent plant defense restricts thrips performance and preference.

Abstract

The western flower thrips (Frankliniella occidentalis [Pergande]) is one of the most important insect herbivores of cultivated plants. However, no pesticide provides complete control of this species, and insecticide resistance has emerged around the world. We previously reported the important role of jasmonate (JA) in the plant's immediate response to thrips feeding by using an Arabidopsis leaf disc system. In this study, as the first step toward practical use of JA in thrips control, we analyzed the effect of JA-regulated Arabidopsis defense at the whole plant level on thrips behavior and life cycle at the population level over an extended period. We also studied the effectiveness of JA-regulated plant defense on thrips damage in Chinese cabbage (Brassica rapa subsp. pekinensis).
Thrips oviposited more on Arabidopsis JA-insensitive coi1-1 mutants than on WT plants, and the population density of the following thrips generation increased on coi1-1 mutants. Moreover, thrips preferred coi1-1 mutants more than WT plants. Application of JA to WT plants before thrips attack decreased the thrips population. To analyze these important functions of JA in a brassica crop plant, we analyzed the expression of marker genes for JA response in B. rapa. Thrips feeding induced expression of these marker genes and significantly increased the JA content in B. rapa. Application of JA to B. rapa enhanced plant resistance to thrips, restricted oviposition, and reduced the population density of the following generation.
Our results indicate that the JA-regulated plant defense restricts thrips performance and preference, and plays an important role in the resistance of Arabidopsis and B. rapa to thrips damage.

Full text

Page 1

BioMed Central
Page 1 of 12
(page number not for citation purposes)
BMC Plant Biology
Open Access
Research article
Jasmonate-dependent plant defense restricts thrips performance
and preference
Hiroshi Abe*1, Takeshi Shimoda2, Jun Ohnishi3, Soichi Kugimiya4,
Mari Narusaka5, Shigemi Seo6, Yoshihiro Narusaka5, Shinya Tsuda2 and
Masatomo Kobayashi1
Address: 1Experimental Plant Division, RIKEN BioResource Center, Tsukuba 305-0074, Japan, 2National Agricultural Research Center, Tsukuba
305-8666, Japan, 3National Institute of Vegetable and Tea Science, Tsu 514-2392, Japan, 4National Institute for Agro-Environmental Sciences,
Tsukuba 305-8604, Japan, 5Research Institute for Biological Sciences, Okayama 716-1241, Japan and 6National Institute of Agrobiological
Sciences, Tsukuba 305-8666, Japan
Email: Hiroshi Abe* - ahiroshi@rtc.riken.jp; Takeshi Shimoda - oligota@affrc.go.jp; Jun Ohnishi - jonishi@affrc.go.jp;
Soichi Kugimiya - kugimiya@affrc.go.jp; Mari Narusaka - ma_narusaka@bio-ribs.com; Shigemi Seo - sseo71@affrc.go.jp;
Yoshihiro Narusaka - yo_narusaka@bio-ribs.com; Shinya Tsuda - shinyat@affrc.go.jp; Masatomo Kobayashi - kobayasi@rtc.riken.jp
* Corresponding author
Abstract
Background: The western flower thrips (Frankliniella occidentalis [Pergande]) is one of the most
important insect herbivores of cultivated plants. However, no pesticide provides complete control
of this species, and insecticide resistance has emerged around the world. We previously reported
the important role of jasmonate (JA) in the plant's immediate response to thrips feeding by using
an Arabidopsis leaf disc system. In this study, as the first step toward practical use of JA in thrips
control, we analyzed the effect of JA-regulated Arabidopsis defense at the whole plant level on thrips
behavior and life cycle at the population level over an extended period. We also studied the
effectiveness of JA-regulated plant defense on thrips damage in Chinese cabbage (Brassica rapa
subsp. pekinensis).
Results: Thrips oviposited more on Arabidopsis JA-insensitive coi1-1 mutants than on WT plants,
and the population density of the following thrips generation increased on coi1-1 mutants.
Moreover, thrips preferred coi1-1 mutants more than WT plants. Application of JA to WT plants
before thrips attack decreased the thrips population. To analyze these important functions of JA in
a brassica crop plant, we analyzed the expression of marker genes for JA response in B. rapa. Thrips
feeding induced expression of these marker genes and significantly increased the JA content in B.
rapa. Application of JA to B. rapa enhanced plant resistance to thrips, restricted oviposition, and
reduced the population density of the following generation.
Conclusion: Our results indicate that the JA-regulated plant defense restricts thrips performance
and preference, and plays an important role in the resistance of Arabidopsis and B. rapa to thrips
damage.
Published: 27 July 2009
BMC Plant Biology 2009, 9:97doi:10.1186/1471-2229-9-97
Received: 23 January 2009
Accepted: 27 July 2009
This article is available from: http://www.biomedcentral.com/1471-2229/9/97
© 2009 Abe et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2

BMC Plant Biology 2009, 9:97http://www.biomedcentral.com/1471-2229/9/97
Page 2 of 12
(page number not for citation purposes)
Background
Insect attack is one of the most important factors retarding
plant growth, decreasing crop productivity, and causing
other agricultural problems. A constitutive and inducible
plant defense response confers immunity to herbivorous
insects [1-3]. Analyses at the molecular, metabolic, and
physiological levels [2,4] have focused on responses to
lepidopteran larvae (caterpillars) and aphids. Many anal-
yses of plant responses to feeding by caterpillars have
been conducted [e.g., [5-7]]. Caterpillars harm plants by
chewing-type feeding, the best understood of several feed-
ing modes. Although caterpillar feeding and mechanical
wounding are physically similar, plants show obvious
specific responses to caterpillar feeding [8]. Some of these
responses are induced by insect gut and oviposition
[9,10]. The sucking-type feeding by aphids and whiteflies
is also well understood. However, in contrast to caterpillar
feeding, sucking-type feeding rarely causes mechanical
damage to the host plant. Rossi et al. [11] reported that
the nematode resistance (R) gene Mi-1 of tomato is
involved in resistance to the potato aphid. Mi-1 also con-
fers resistance to whiteflies [12]. Other major classes of
insect feeding are also known. Leafminers feed within
leaves and stems, forming tunnels (mining-type feeding),
and thrips and spider mites feed by piercing and sucking
[13,14].
The western flower thrips (Frankliniella occidentalis [Per-
gande]) is one of the most important insect herbivores.
This tiny insect tends to occupy narrow crevices within or
between plant parts. The emergence worldwide of insecti-
cide resistance among western flower thrips makes them
difficult to control [15]. The thrips can also act as a vector
of tospoviruses such as tomato spotted wilt virus [16,17].
Damage by western flower thrips is increasing in many
countries; in particular, injury in greenhouse production
is serious [18-20]. Thus, the development of new methods
to control thrips damage by using the molecular mecha-
nisms of plant responses is needed.
Jasmonate (JA) has an important function in plant
responses to caterpillars and aphids [2]. Reymond et al.
[21] reported that the JA-insensitive coi1-1 mutant of Ara-
bidopsis is less resistant to cabbage butterfly (Pieris rapae).
Ellis et al. [22] reported that coi1-1 mutants are less resist-
ant to aphids, but the constitutive JA-signaling mutant
cev1 is more resistant. Our recent study focusing on Arabi-
dopsis response to thrips feeding also indicated the impor-
tant function of JA [23,24],
transcriptome analyses suggested a strong relationship
between JA treatment and thrips feeding [23]. Several
groups reported that JA-regulated gene expression is
induced by spider mites feeding [25,26], which have a
similar feeding mode to that of thrips. De Vos et al., using
Arabidopsis genome arrays [27], also reported the impor-
and comparative
tance of JA for feeding-inducible gene expression by thrips
and cabbage butterfly attack. Interestingly, they indicated
the existence of common genes in the response to both
feeding modes, and genes specific to each feeding mode.
Arabidopsis is a widely studied experimental plant for
which many useful genomic resources and much other
information are available. However, it is not suitable for
analyzing Arabidopsis responses to caterpillars, which can
quickly eat an entire plant. On the other hand, with the
tiny western flower thrips, it is possible to analyze Arabi-
dopsis responses to thrips attack over generations.
In this study, we focused on the effect of JA-regulated Ara-
bidopsis defense at the whole plant level on thrips behavior
and life cycle at the population level. We analyzed the
long-term effects of JA-regulated plant defense on thrips
oviposition, the population density of the following
thrips generation (larvae and pupae), and preference
between Arabidopsis WT and JA-insensitive coi1-1 mutant
host plants. The results show important effects of the JA-
dependent plant defense on both thrips performance and
preference. In addition, application of JA to Arabidopsis
WT plants before thrips attack decreased the thrips popu-
lation. Expression analyses of marker genes for JA
response in Chinese cabbage (Brassica rapa subsp. pekinen-
sis) suggested the occurrence of a JA-dependent defense
against thrips attack in this plant, too. The JA content of B.
rapa was significantly increased after thrips feeding, and
application of JA to plants enhanced their resistance to
thrips.
Results
Importance of jasmonate-regulated Arabidopsis defense
in resistance to thrips attack
We recently reported the role of JA in the short-term
response of Arabidopsis to thrips feeding on leaf discs over
1 or 2 days [23,24]. To analyze its role in long-term
defense at the whole plant level, we compared the feeding
damage between whole WT plants and JA-insensitive coi1-
1 mutants [28] inoculated with 20 thrips at 3 weeks. The
coi1-1 mutants had been completely devoured by 4 weeks
after inoculation, whereas WT plants were flowering and
producing siliques (Fig. 1A). These results suggest the
importance of JA-regulated defense in the resistance of
Arabidopsis to thrips attack.
To understand why coi1-1 mutants showed low resistance
to thrips attack, we first analyzed the number of thrips
eggs on the WT plants and coi1-1 mutants to compare the
asexual oviposition performance of thrips. Arabidopsis
rosette leaves were cut into leaf discs with 8-mm diameter.
One adult female thrips was put on each disc and allowed
to feed and oviposit. Because the females lay in the epider-
mal or mesophyll cells [29], we stained the eggs with

Page 3

BMC Plant Biology 2009, 9:97http://www.biomedcentral.com/1471-2229/9/97
Page 3 of 12
(page number not for citation purposes)
trypan blue to count them. As we reported previously
[23], the area of feeding scars on coi1-1 mutants was
greater than that on WT plants (data not shown). The
number of eggs on the coi1-1 discs was double that on the
WT discs (Fig. 2A–C). The decreased resistance of these
coi1-1 mutants could explain the increased oviposition
rate on these plants.
Effect of jasmonate-dependent Arabidopsis defense on
thrips population
Because the JA-regulated defense affected oviposition, we
analyzed its effect on the subsequent generation. We put
20 adult females on WT and coi1-1 plants and counted
adults, larvae, and pupae after 2 weeks. We covered the
soil with fine zirconia beads 0.4 mm in diameter to find
thrips easily. Thrips fed much more on coi1-1 mutants
than on WT plants (Fig. 3A, B). About 14 of the original
adult females remained on coi1-1 mutants, but only about
2 remained on WT plants (Fig. 3C). Similarly, while more
than 190 larvae lived on the coi1-1 mutants, only about 20
lived on the WT plants (Fig. 3D). We also found 5 times
as many pupae on coi1-1 mutants than on WT plants (Fig.
3E). These results demonstrate that the JA-regulated
defense can determine thrips population size.
Next, we analyzed the effect of JA-regulated plant defense
on host plant preference of thrips. We placed 100 adult
females halfway between WT and coi1-1 plants (Fig. 4A)
and counted the thrips on each plant after 2 days. The
coi1-1 mutants had many more thrips than the WT plants
(Fig. 4A): > 70% versus about 5% (χ2 test, χ2 = 175.879, df
= 1, p < 0.001; Fig. 4B); the remaining thrips roamed the
surroundings. These results indicate that the JA-regulated
plant defense influences the host plant preference of
thrips.
We next analyzed the effect of JA treatment on Arabidopsis
resistance to thrips attack. JA-treated plants had half as
many eggs as untreated plants (Fig. 5A). The numbers of
adults and larvae showed a similar contrast (Fig. 5B, C).
Together with the results from the coi1-1 mutants, these
results indicate that the JA-dependent defense response in
Arabidopsis plays an important role in resisting thrips.
Jasmonate-dependent plant resistance to thrips in B. rapa
To search for JA-dependent resistance to thrips in a
brassica crop, we analyzed the function of JA in B. rapa,
one of the most important brassica crops in the world,
especially in Asia. A search of the B. rapa EST database
(National Center for Biotechnology Information)
revealed putative counterparts of Arabidopsis JA-inducible
Function of JA in plant resistance to thrips feeding
Figure 1
Function of JA in plant resistance to thrips feeding.
Twenty adult females fed on 3-week-old WT plants (left) or
coi1-1 mutants (right). Typical plants after 4 weeks of feeding
are shown.
Effect of JA-dependent plant resistance on thrips oviposition on leaf discs
Figure 2
Effect of JA-dependent plant resistance on thrips ovi-
position on leaf discs. One adult female fed per leaf disc of
3-week-old WT plants (A) or coi1-1 mutants (B) for 4 days.
Eggs oviposited on leaf discs were stained with trypan blue.
Photos show typical leaf discs after staining; some eggs are
shown by red arrowheads. (C) Number of eggs per leaf disc
(mean ± SD) based on more than 10 independent determina-
tions. Asterisks indicate significant difference (Student's t-
test), ***p < 0.001.

Page 4

BMC Plant Biology 2009, 9:97http://www.biomedcentral.com/1471-2229/9/97
Page 4 of 12
(page number not for citation purposes)
marker genes. We analyzed the expression of marker genes
of the JA pathway corresponding to AtVSP2 and AtLOX2,
and genes corresponding to allene oxide synthase (AtAOS)
and allene oxide cyclase 2 (AtAOC2), both of which encode
enzymes that catalyze JA biosynthesis as shown by previ-
ous reports (Fig. 6E) [30,31]. Expression of the brassica
counterparts, BrVSP2, BrLOX2, BrAOS, and BrAOC2, was
induced by thrips feeding (Fig. 6A–D). In addition, the JA
content of B. rapa plants infested by thrips was signifi-
cantly higher than that of control plants (one-way
ANOVA, F = 13.938, df = 2, p < 0.01; Fig. 6F). These data
suggest the involvement of JA in the response to thrips
feeding in B. rapa also.
To confirm the functional role of JA in plant resistance to
thrips attack in B. rapa, we analyzed the effect of JA treat-
ment on thrips feeding. Injury from thrips attack was
lower in plants treated with JA than in untreated plants
(Fig. 7A–D), by a factor of about 15 (Fig. 7E). These results
indicate that the JA-dependent plant defense against
thrips is conserved in B. rapa.
We further analyzed JA's effect on thrips oviposition.
Rosette leaves of B. rapa were cut into leaf discs with 8-mm
diameter. One adult female thrips was put on each leaf
disc and allowed to feed and oviposit for 4 days. Applica-
tion of JA dose-dependently decreased the number of eggs
(one-way ANOVA, F = 10.367, df = 4, p < 0.001; Fig. 8A).
Finally, we analyzed the effect of JA on the next genera-
tion. JA treatment of plants restrained the thrips popula-
tion very effectively (Fig. 8B, C). These results clearly
indicate the important role of JA in resistance to thrips
attack in B. rapa also.
Discussion
The phytohormone JA regulates part of a plant's basal
defense system. Numerous studies have examined the
functions of JA in plant responses to pathogen attack,
mechanical wounding, UV irradiation, ozone exposure,
osmotic stress [32,33], and insect feeding [34,35]. The JAZ
(jasmonate ZIM-domain) family of repressors was identi-
fied in Arabidopsis as a negative regulator of JA signaling
[36-38]. JAZ interacts with COI1 protein, degrades, and so
Effect of the JA-dependent plant defense on thrips population
Figure 3
Effect of the JA-dependent plant defense on thrips population. (A, B) Twenty adult females fed on 3-week-old WT
plants (A) or coi1-1 mutants (B) for 2 weeks. (C-E) Number of adults (C), larvae (D), and pupae (E) after 2 weeks; mean ± SD
based on five independent determinations. Asterisks indicate significant differences (Student's t-test), **p < 0.005, ***p < 0.001.

Page 5

BMC Plant Biology 2009, 9:97http://www.biomedcentral.com/1471-2229/9/97
Page 5 of 12
(page number not for citation purposes)
induces JA-responsive gene expression. Overexpression of
a modified form of JAZ1 significantly decreased plant
resistance to the beet armyworm (Spodoptera exigua) [39].
Resistance by coi1-1 mutants to cabbage butterfly caterpil-
lar (Pieris rapae) was similarly decreased [21].
However, these analyses focused on plant responses to
lepidopteran larvae. Because caterpillars quickly devour
Arabidopsis plants and change to butterflies or moths,
which fly away, it is difficult to analyze the Arabidopsis
response and insect performance over generations on the
one Arabidopsis plant. For these reasons, we used thrips.
We found differences in symptoms between WT plants
and JA-insensitive coi1-1 mutants: thrips had demolished
coi1-1 mutants after 4 weeks, yet WT plants had flowers
and siliques (Fig. 1A). As it seemed unlikely that only 20
adult thrips could kill a plant in 4 weeks, we also studied
the effect of a JA-dependent Arabidopsis defense on ovipo-
sition. The number of eggs on coi1-1 was about double
that on the WT (Fig. 2A–C). As we described previously
[23], the area of feeding scars in coi1-1 was much greater
than that on WT plants (data not shown). The greater
number of eggs on coi1-1 might result from the better per-
formance of adult thrips. Alternatively, a difference in
plant metabolites between WT and coi1-1 might influence
oviposition. Annadana et al. [40] reported that cysteine
protease inhibitors restrict oviposition by western flower
thrips. Wounding and JA induce many genes encoding
cysteine protease inhibitors [41], including Arabidopsis
cystatin-1 (AtCYS1) [42]. Cysteine protease inhibitors
could explain the difference in thrips oviposition between
WT and coi1-1 plants.
Next, we analyzed the effect of JA-regulated plant defense
on the population density of the following generation of
thrips. Surprisingly, the population increased around 10-
fold after 2 weeks on the coi1-1 mutants, but changed little
on the WT plants (Fig. 3A–E). Most of the thrips on coi1-1
were larvae. We found some dead larvae on the WT plants
but none on coi1-1 (data not shown). These results indi-
cate that the JA-dependent plant defense in WT plants
reduces the survival of thrips larvae. We found about 7
times as many adult thrips on coi1-1 as on the WT, which
indicates that thrips can survive longer on coi1-1. We
attribute the much greater population of thrips on coi1-1
to this increased longevity and the greater egg production
on coi1-1 mutants, and the higher mortality of larvae on
the WT plants. Analysis of the hatching rate of eggs could
also help explain the increased population on coi1-1.
Barth et al. [43] reported that a double knock-out mutant
of Arabidopsis lacking two major genes for myrosinase
(tgg1, tgg2), which degrades glucosinolates to toxins such
as isothiocyanates, showed decreased resistance to the
cabbage looper (Trichoplusia ni) and tobacco hornworm
(Manduca sexta). Sasaki-Sekimoto et al. [33] reported that
JA regulates glucosinolate biosynthesis. Recently, Shroff et
al. [44] showed that the preferential allocation of glucosi-
nolates to the periphery of leaves may play a key role in
the defense of leaves by creating a barrier to chewing her-
bivores, which frequently approach leaves from the edge.
Several other compounds protect plants against insect
pests. Konno et al. [45] reported that cysteine proteases
such as papain, ficin, and bromelain showed toxicity to
two notorious pests, cabbage armyworm (Mamestra brassi-
cae) and cotton leafworm (Spodoptera litura). They later
reported that sugar-mimic alkaloids were toxic to cabbage
armyworm [46]. Further analyses will help to explain
which kinds of compounds, regulated by JA, reduce thrips
performance.
The choice test showed that coi1-1 mutants attracted 14
times as many thrips as did WT plants (Fig. 4A, B). As a
result, coi1-1 mutants suffered more damage. Aharoni et
Effect of the JA-dependent plant defense on host plant pref- erence of thrips
Figure 4
Effect of the JA-dependent plant defense on host
plant preference of thrips. (A) Three-week-old WT
plants (left) and coi1-1 mutants (right) were grown at each
end of a pot. One hundred adult females were collected in a
1-mL tube and laid between the plants. Photo shows plants
after 2 days. (B) Number of adult thrips on each plant after 2
days. Mean ± SD based on five independent determinations.
Asterisk indicates a significant difference between the two
plants, χ2 test, ***p < 0.001. The remaining thrips roamed the
surroundings were excluded from the statistical analysis.