Alkaloid metabolism in thrips-Papaveraceae interaction: Recognition and mutual response

Page 1

B

Aa

Ia

b

a

ARR1AA

KBCEFS

I

(Iybeotls((tifi

d

fT

0h

Journal of Plant Physiology 171 (2014) 119– 126

Contents lists available at ScienceDirect

Journal of Plant Physiology

j o ur na l ho me page: www.elsev ier .com/ locate / jp lph

iochemistry

lkaloid metabolism in thrips-Papaveraceae interaction: Recognitionnd mutual response

ngeborg Schütza,∗, Gerald B. Moritza, Werner Roosb

Martin-Luther-Universität, Institut für Biologie/Entwicklungsbiologie, Domplatz 4, 06108 Halle, GermanyMartin-Luther-Universität, Institut für Pharmazie/Molekulare Zellbiologie, Kurt-Mothes-Str. 3, 06120 Halle, Germany

r t i c l e i n f o

rticle history:eceived 1 August 2013eceived in revised form5 September 2013ccepted 21 October 2013vailable online 20 November 2013

eywords:

a b s t r a c t

Frankliniella occidentalis (Pergande), the Western Flower Thrips (WFT), is a polyphagous and highly adapt-able insect of the order Thysanoptera. It has a broad host range but is rarely found on Papaveraceae,which might be due to deterrent effects of alkaloids present in most species of this family. In order totest the adaptive potential of WFT, we investigated its interaction with two Papaveraceae offered as solefeeding source. We found that WFT are able to live and feed on leaves of Eschscholzia californica andChelidonium majus. Both plants respond to thrips feeding by the enhanced production of benzophenan-thridine alkaloids. Furthermore, cell cultures of E. californica react to water insoluble compounds prepared

enzophenanthridine alkaloidshelidonium majusschscholzia californicarankliniella occidentalisanguinarine reductase

from adult thrips with enhanced alkaloid production. During feeding, WFT take up benzophenanthridinealkaloids from either plant and from an artificial feeding medium and convert them to their less toxicdihydroderivatives. This was shown in detail with sanguinarine, the most cytotoxic benzophenanthri-dine. A similar conversion is used in plants to prevent self-intoxication by their own toxins. We concludethat WFT causes a phytoalexin-like response in Papaveraceae, but is able to adapt to such host plants bydetoxification of toxic alkaloids.

ntroduction

Western Flower thrips, Frankliniella occidentalis (Pergande)Thysanoptera, Thripidae), is an insect native to the Western USA.t spread since the 1980s and became a pest worldwide within 20ears (Kirk and Terry, 2003) affecting crop and ornamental plantsoth in greenhouses and in the field (Yudin et al., 1986). West-rn Flower thrips (WFT) is the most effective vector of a numberf Tospoviruses, of which Tomato Spotted Wilt Virus (TSWV) ishe most prominent (Whitfield et al., 2005). WFT lives on planteaves and fruits and inside inflorescences. More than 200 plantpecies from about 90 families can serve WFT as a feeding sourceBrødsgaard, 1989) which indicates its high adaptive potentialMoritz, 2006). It is noteworthy that larval and adult WFT occupyhe same ecological niches but individuals may change hosts dur-

ng their development (Moritz, 2006). Like many other thrips, WFTeeds by piercing epidermal cells and sucking up their contentncluding cytosol, chloroplasts, vacuoles and nuclei (Kirk, 1997).

Abbreviations: HPTLC, high performance thin layer chromatography; Rf, retar-ation factor; SD, standard deviation; WFT, Western Flower thrips.∗ Corresponding author at: Martin-Luther-Universität Halle-Wittenberg, Institut

ür Biologie/Entwicklungsbiologie, Domplatz 4, 06114 Halle, Germany.el.: +49 345 55 26473; fax: +49 345 55 27121.

E-mail address: [email protected] (I. Schütz).

176-1617/$ – see front matter © 2013 Elsevier GmbH. All rights reserved.ttp://dx.doi.org/10.1016/j.jplph.2013.10.009

© 2013 Elsevier GmbH. All rights reserved.

The damage resulting from thrips feeding thus differs from thatcaused by other herbivorous insects. It is characterized by induc-tion of local lesions, silvering around the punctures and darkeningof leaf due to desiccation (reviewed in Childers and Achor, 1995).Plant responses to thrips feeding have been reported in recentyears but the actual knowledge remains fragmentary. Arabidopsisthaliana leaf discs and plants reacted to thrips feeding by increasedbiosynthesis of jasmonate, an ubiquitous mediator of plant defence(Abe et al., 2008), while infection by a tospovirus increased the con-centration of salicylate, thus suppressing the plant’s anti-herbivoreresponse (Abe et al., 2012). The increase of jasmonate leads to areduced rate of oviposition and population growth by WFT on Ara-bidopsis thaliana, Brassica rapa and Triticum aestivum (Abe et al.,2009; El Wakeil et al., 2010). It is unclear, however, which kindof thrips-deterring activity was triggered via the jasmonate sys-tem. The same is true for ethylene, another plant hormone. Elevatedethylene emission has been observed in several plant–thrips inter-actions (reviewed in: Childers and Achor, 1995; Moritz, 2006) butno defence activities downstream of the ethylene peak are known.

An effective and widely distributed reaction of plants toherbivory is the enhanced production of toxic secondary com-pounds, as various alkaloids, terpenes, phenolic compounds, etc.

(cf. Dixon et al., 1995; Walling, 2000; Bruxelles and Roberts, 2001;Hammerschmidt and Kagan, 2001; Lambdon and Hassall, 2001).Similar responses to thrips feeding have not yet been reportedalthough such molecules most probably can deter thrips from

Page 2

1 t Phy

fsCspdsitMit

t(ta2nle

WacacihsslhamK1thtSamp2

bpewlA2Ajtefrcwiee

e

20 I. Schütz et al. / Journal of Plan

eeding: Maharijaya et al. (2012) have recently identified severalecondary products that are likely involved in the resistance ofapsicum (Solanaceae) species to WFT feeding, i.e. tocopherols, aesquiterpene, an unknown sterol and a number of alkanes. Plantsroducing cytotoxic pyrrolizidine alkaloids may benefit from theireterrent effects against WFT, as suggested from studies in twopecies of the genus Jacobaea (Asteraceae), showing higher feed-ng damage on those species that contain lower alkaloid levelshan others, and vice versa (Macel et al., 2005; Cheng et al., 2011;

acel, 2011). Up to now, neither enhanced production of alkaloidsn response to thrips invasion nor metabolic adaptation of the thripso plant defence compounds are known.

Metabolic adaptations of other herbivorous insects to the phy-ochemicals of food plants are known from insect-host systemsGlendinning, 2002). The most detailed data are available from arc-iids and similar insects that feed on Senecio (Asteraceae) speciesnd cope with their toxic pyrrolizidine alkaloids (Hartmann et al.,003, 2005; Langel and Ober, 2011). In this case, the metabolism isot only required to protect the insect but also to recruit the alka-

oids for the insect’s defence against animal predators (Narberhaust al., 2005).

In the present study we investigate the interaction betweenFT and two plant species that produce benzophenanthridine

lkaloids, members of the benzylisoquinoline family with a highytotoxic potential (Schmeller et al., 1997). The plants studied herere Chelidonium majus (L.) (Greater celandine) and Eschscholziaalifornica (Cham.) (California poppy), two species of the fam-ly Papaveraceae that are widespread both in the Americanemisphere and Europe (Tutin et al., 1993) and thus may be con-idered as potential host plants of WFT. They contain a broadpectrum of benzophenanthridines and other benzylisoquino-ines (cf. Preininger, 1986). Benzophenanthridine alkaloids areighly effective phytochemicals as they intercalate dsDNA, inter-ct with cytoskeletal proteins and various SH-enzymes and disturbembrane potential-dependent enzymatic processes (Wolff and

nipling, 1993; Faddeeva and Beliaeva, 1997; Schmeller et al.,997; Slaninova et al., 2001). Sanguinarine is regarded as the mostoxic among the benzophenanthridine alkaloids, due to its planar,ydrophobic and cationic molecular structure, making it one ofhe strongest antimicrobials produced in plants (Wink et al., 1998;laninova et al., 2001). The biochemical, enzymatic and molecularspects of benzophenanthridine biosynthesis have been studied inost detail in cell suspension cultures of E. californica and related

lants (Zenk, 1994; cf. Ziegler and Facchini, 2008; Klein and Roos,009 for reviews).

Contact with microbial elicitors causes enhanced expression ofiosynthetic and protective enzymes followed by the enhancedroduction of these benzophenanthridine alkaloids (Gundlacht al., 1992; Cline et al., 1993; Roos et al., 1998). Two signal path-ays operate between pathogen detection and gene expression:

ow elicitor concentrations signal via activation of phospholipase2 and subsequent shifts of cytoplasmic pH (Viehweger et al., 2002,006; Färber et al., 2003; Roos et al., 2006; Heinze et al., 2007;ngelova et al., 2010), high elicitor concentrations cause a peak of

asmonate, leading to enhanced expression of biosynthetic genesogether with an oxidative burst (Gundlach et al., 1992; Färbert al., 2003). Benzophenanthridine producing cells are protectedrom self-intoxication by several mechanisms, among them theeduction to less-toxic derivatives by sanguinarine reductase, aytoplasmic enzyme (Weiss et al., 2006; Vogel et al., 2010). Withhole plants, the highest alkaloid response to microbial elicitors

s found in young roots. Leaves originally contain low alkaloid lev-

ls, but are able to enhance alkaloid production upon contact withlicitor or wounding (Angelova et al., 2010).

The two plant species chosen for our study represent differ-nt modes of compartmentalization of benzophenanthridines: in

siology 171 (2014) 119– 126

C. majus, cells excrete alkaloids into laticifers which develop inleaves, stems and roots (Colombo and Bosisio, 1996), E. californicahas no laticifers but mainly accumulates the alkaloids in idioblastsof the root cortex (Angelova et al., 2010).

With this background in mind, we explored the following ques-tions:

1. Do WFT avoid feeding on benzophenanthridine producingPapaveraceae?

2. Can WFT adapt to food plants that produce benzophenanthridinealkaloids?

3. Can WFT detoxify these alkaloids?4. Are WFT recognized by alkaloid-producing Papaveraceae and do

the plants respond by enhanced alkaloid production?

Material and methods

Growth and maintenance of thrips population

The colony of Frankliniella occidentalis (WFT) was maintainedon common bean plants (Phaseolus vulgaris (L.)) with fully devel-oped leaves at a 16 h/8 h light-dark regime, 23 ◦C ± 2 ◦C and 75–80%relative humidity in a separate breeding room. Fresh bean plants(grown in greenhouse for about 10 days) were supplied every week.

For generation of first instar larvae, adult WFT were placed onbean leaf discs as described below and young larvae were collectedfrom these plates 0–4 h after hatching.

Plant material and cell culture

P. vulgaris and Eschscholzia californica (Cham.) were grown inpots of garden soil in the greenhouse at 21 ◦C. E. californica cellcultures were grown in a modified Linsmeyer-Skoog medium ongyratory shakers at 24 ◦C as described previously (Viehweger et al.,2002). Leaves of Chelidonium majus (L.) plants were taken from thewild (yard of the Zoological Institute in Halle).

WFT on leaf discs of P. vulgaris and C. majus

Each well of 12-well plates (Greiner Bio-One, Essen, Germany)was filled with 1.5 mL 0.4% (w/v) agar (Carl Roth, Karlsruhe,Germany). Leaf discs (15 mm in diameter) were punched out fromnewly emerged green leaves of P. vulgaris or from leaves of C. majus.Leaf discs were placed on the agar after it was cooled to room tem-perature. WFT (adults or larvae depending on the experiment) wereplaced on the leaf discs. The 12-well plates were closed with glasslids, sealed with Parafilm (Pechiney Plastic Packaging, Chicago) andkept in a climate chamber at 23 ◦C ± 1 ◦C, 16 h/8 h light-dark cyclesand relative humidity of 75%.

To study developmental time and survival rate, one freshlyemerged first instar larva (0–4 h after hatching) was placed on eachleaf disc and examined daily under a stereo microscope. To studyalkaloid uptake and induction of alkaloid production in the plant,15 first instar larvae were placed on each leaf disc. Larvae were col-lected and analysed for alkaloid uptake after a feeding period of 3days.

WFT on E. californica

To study alkaloid uptake from E. californica and alkaloid pro-duction in the plant, leaflets of 1.5–2 cm length (corresponding toapproximately 1.5–2.0 cm2 leaf surface, estimated from a digital

image) were cut from 20 to 25 d old E. californica plants from greenhouse and put in agar (1% (w/v) in distilled water) in 20 mL plasticjars. 25 first instar WFT larvae were placed on the leaflets in eachjar. The jars were sealed with Parafilm to avoid contaminations

Page 3

t Phy

adldl

C

gdoedw

2lTl(h

C

tmaSsftrret

Wa

jmaF3ptdf

A

va(p

mnEoa

I. Schütz et al. / Journal of Plan

nd drying and kept in the climate chamber under the conditionsescribed for leaf discs if P. vulgaris and C. majus. Larvae were col-

ected and analysed for alkaloid uptake after a feeding period of 3ays. To study survival rate of WFT, larvae were kept on E. californica

eaflets until pupation.

hoice of feeding plant – settlement preference test

Choice experiments were performed using leaf pieces of P. vul-aris and C. majus in 12-well plates. Leaf discs were punched out asescribed above and cut in half. One half of a P. vulgaris leaf disc andne half of a C. majus leaf disc were placed in each well adjacent toach other in order to allow movement of larvae between both halfiscs. On each half three first instar larvae were placed (6 larvae perell). Plates were closed and cultivated as described above.

Numbers of larvae on each half leaf disc were counted after4, 48 and 72 h. Dead thrips, developmental stages other than

arvae and thrips not on the leaf material were not included.his experiment was repeated three times. Data were ana-ysed and tested for significance (t-test) using WinSTAT softwareWinSTAT 2006.1; R. Fitch Software, Bad Krozingen, Germany;ttp://www.winstat.de/default.htm).

hoice between feeding solutions

40 adult WFT were placed in a glass Y-pipe which was closed athe lower end. Both upper openings were covered with a parafilm

embrane. 250 �L 3% sucrose solution (w/v in distilled water) werepplied on each side and covered with a second parafilm layer.ucrose solution on one side contained 20 �M, 100 �M or 200 �Manguinarine, respectively and no sanguinarine was added to theeeding solution on the other side. WFT are able to pierce throughhe parafilm membrane and ingest feeding solution by suckingesembling their feeding habit on natural sources. After 24 h WFTesiding on the parafilm membrane on either side were counted. Forach concentration of sanguinarine, the experiment was repeatedhree times. Data were analysed as described above.

FT feeding on sanguinarine solution – test of metabolization oflkaloids

20 adult WFT were placed in each of three empty 20 mL plasticars which were then sealed with parafilm. On top of the parafilm

embrane, 250 �L feeding solution were placed and covered with second parafilm layer as described in Kumm and Moritz (2008).eeding solution contained 3% sucrose (w/v in distilled water) and0 �M or 100 �M sanguinarine or no sanguinarine. WFT are able toierce through the parafilm membrane and ingest feeding solutionhus resembling their feeding habit on natural sources. After twoays, thrips were collected, frozen in liquid nitrogen and assayedor alkaloid content as described below.

ssay of benzophenanthridine alkaloids

Alkaloid content of WFT. After the indicated feeding period, lar-ae or adult WFT (cf. above) were collected in 50 �L ethanol p.a.nd shaken at room temperature for 2 min. After centrifugation10 min at 5000 × g), 5 �L of the supernatant were applied to higherformance thin layer chromatography (HPTLC, see below).

Alkaloid content of leaves. After three days of WFT feeding, C.ajus leaf discs (between 50 and 62 mg fresh weight) or E. califor-

ica leaflets (between 40 and 60 mg fresh weight) were collected inppendorf tubes containing 500 �L ethanol p.a. Tubes were thor-ughly shaken for 2 min at room temperature, centrifuged (5 mint 5000 × g) and the supernatants were diluted with water (1:2).

siology 171 (2014) 119– 126 121

Chlorophyll was removed by 3 subsequent extractions with hex-ane, each followed by another short centrifugation. The aqueousphase was evaporated under nitrogen and the dry residue was dis-solved in 50 �L ethanol p.a., 5 �L of this solution were examined byHPTLC analysis.

HPTLC analysis of benzophenanthridine alkaloids. 5 �L of analkaloid containing extract were spotted per lane onto a5 cm × 5 cm HPTLC plate (Merck, Germany) that was developed in aMicro chamber (Camag, Switzerland) with hexane/methanol/ethylacetate (20/4/4, vol/vol/vol) as the mobile phase. After dry-ing on air, alkaloid spots were visualized by their red-orangefluorescence (benzophenanthridines) or blue fluorescence (dihy-drobenzophenanthridines) as excited under UV (280–360 nm)illumination. Alkaloids were identified by their migration patterns(retardation factor, Rf) compared to that of authentic compoundsthat were commercially available (sanguinarine, chelerythrine) orisolated from E. californica cell cultures and identified by MS/MS(chelirubine).

Quantification was based upon digital fluorescence imagesobtained with an Olympus E10 camera. Each spot was analysedfor size and intensity by the Optimas 6.2 software and the productwas converted into amount by using a calibration graph that hadbeen obtained with aqueous sanguinarine or dihydrosanguinarinesolution of known concentration.

The average SD of the overall assay (sampling plus experimentalerror, excluding biological variations between different plants orthrips populations) was as estimated with 5 extracts of the sameleaf to about 12%.

Test for response of E. californica cell culture to WFT

An aqueous extract (homogenate) was prepared from adult WFTthat had been feeding on P. vulgaris. For that, 50 individuals werecollected in 100 �L distilled water, followed by 3 times freezing andthawing and homogenization with a plastic pestle in Eppendorfreaction tube. Part of the mixture was centrifuged at 100,000 × gfor 2 h at 4 ◦C. 10 �L of the supernatant after ultracentrifugation(containing the soluble molecules) or 10 �L of the homogenatebefore centrifugation (containing both soluble and solid compo-nents) were added to 1 mL of E. californica cell culture.

After 24 h, alkaloids produced in the cell culture were extractedfrom 500 �L samples by adding the same volume of KOH (0.3 Nin methanol) and 5 min of vigorous shaking at room temper-ature. After centrifugation (20 min at 13,000 rpm), 150 �L ofthe supernatant were pipetted into a micro plate well (blackmicro plates, Greiner, Germany) and mixed with 15 �L 3 N sul-phuric acid. Fluorescence was read at Ex 360 nm, Em 460 nm(dihydrobenzophenanthridines) and Ex 485 nm, Em 580 nm (ben-zophenanthridines) and converted into alkaloid concentration byusing standard solutions of known sanguinarine and dihydrosan-guinarine concentration.

Results

WFT in choice tests

Choice of feeding plant. In a first set of experiments WFT larvaewere allowed to choose between leaf pieces of P. vulgaris and C.majus for settlement and feeding. 24, 48 and 72 h after placing equalnumbers of larvae on the leaf pieces, significantly more larvae werefound on leaf pieces of P. vulgaris than on C. majus leaf pieces (Fig. 1).

WFT larvae clearly prefer P. vulgaris over C. majus, and more feedingscars were observed on P. vulgaris. It remained unclear whether thischoice is based upon the deterrent effect of the alkaloids present inC. majus.

Page 4

122 I. Schütz et al. / Journal of Plant Phy

Fig. 1. Settlement preference test. (a) WFT larvae feeding on leaf discs, upper right:P. vulgaris; lower left: C. majus. Arrows indicate feed scars. (b) WFT larvae per leafppP

ectogtt

Fascf(

iece were counted at the time points indicated. Data are means of 22 pairs of leafieces. Note that the numbers of larvae on C. majus are significantly lower than on. vulgaris (p < 0.01) at 24, 48 and 72 h.

Choice between artificial feeding solutions. To test for the influ-nce of alkaloids on the choice of feeding plant, we performed ahoice experiment with sanguinarine, a common benzophenan-hridine alkaloid present both in C. majus and E. californica. We

ffered adult WFT sucrose as feeding solution with or without san-uinarine in a Y-shaped glass pipe. The result is a clear avoidance ofhe sanguinarine-containing feeding medium at all concentrationsested (Fig. 2). This implies that WFT is able to sense sanguinarine

ig. 2. Effect of sanguinarine in a choice test in a Y-pipe. Feeding solution on onerm contained sanguinarine of the given concentration, on the other arm the sameolution without sanguinarine. After 24 h, WFT feeding at either solution wereounted. Data are means of three experiments, ±SD, n = 3. Significantly less WFTed on sanguinarine-containing medium at all tested sanguinarine concentrationsp < 0.05).

siology 171 (2014) 119– 126

and is deterred by this compound. This may explain the preferencefor bean over C. majus leaf pieces (cf. Fig. 1).

WFT on C. majus and E. californica

If C. majus leaves were provided as sole feeding source, a portionof larvae fed on these. 24% (57 out of 240) of the larvae suppliedwith C. majus leaves survived and developed into adults. At 23 ◦C,the developmental time spans were as follows: larval stages (firstand second instar) 5.6 ± 1 d, prepupal stage 1.2 ± 0.5 d and pupalstage of 2.4 ± 0.8 d. Life cycles were completed in 9.3 ± 1.2 d.

In qualitative experiments we observed that adult F. occidentalisbrought to C. majus leaf discs laid eggs there and that the larvaehatched from these eggs showed no obvious deformities. Thus, thealkaloids or other constituents of C. majus leaves did not influencethe developmental cycle of WFT feeding on this plant.

Leaflets of E. californica likely provide a less valuable feedingsource compared to C. majus, as they consist of small leaflets andthus have a lower cytoplasm-to-surface ratio. Nevertheless, 6% (13out of 210) of the larvae that were put on E. californica leaflets fedthere and developed into the pupal stage, showing that E. californicaprincipally can serve as sole feeding source as well. Feeding scarswere observed on E. californica leaflets.

Alkaloid uptake and metabolization

Alkaloid uptake from feeding plant. The finding that WFT avoidfeeding on the alkaloid containing leaves, but can use them assole feeding source, raised the question how WFT cope with thetoxic alkaloids. As seen from Fig. 3 and documented in detailin literature (e.g. Preininger, 1986), E. californica and C. majuscontain a variety of benzophenanthridines. In E. californica, san-guinarine, chelerythrine, chelirubine and macarpine are the maincomponents, whereas sanguinarine, chelerythrine, chelidonine andberberine dominate the alkaloid spectrum of C. majus.

We tested the alkaloid uptake of larvae feeding on C. majus andE. californica for three days by analysing body extracts of larvaethat had fed on these plants. Preliminary experiments with leafletsof E. californica had indicated that the alkaloid content in larvaeincreased with the time of larval feeding and saturated after 3–4days. As shown in Fig. 4, benzophenanthridine alkaloids accumu-late in larvae that had been feeding on C. majus and E. californica forthree days with sanguinarine and chelirubine as the main alkaloidspecies. Expectedly, larvae that fed on P. vulgaris contained no ben-zophenanthridine alkaloids or similarly fluorescing compounds.

Metabolization of alkaloids – detoxification of sanguinarine byWFT. The alkaloids that accumulated in WFT larvae during feedingdid not only comprise the dominating alkaloid species of the forageplant but in addition contained dihydroalkaloids, mainly dihy-drosanguinarine (Fig. 4). This compound is present in C. majus andE. californica as an intermediate of benzophenanthridine biosyn-thesis and also the product of detoxification of sanguinarine (cf.Weiss et al., 2006). The concentration ratio of dihydrosanguinarinevs. sanguinarine in the forage plant was lower than among the alka-loids accumulating in the insect (cf. Fig. 3). This was taken as a firsthint that dihydrosanguinarine was not (only) taken up from thefed plant, but rather metabolized from sanguinarine in the insect.In order to prove the suggested metabolizing activity, we providedadult WFT with an artificial feeding solution that contained san-guinarine as the only alkaloid. After a two-day feeding period,dihydrosanguinarine was clearly identified in the thrips (Fig. 5).As this alkaloid was not present in the feeding solution, it must

have been generated by WFT. Although the amount of sanguinar-ine ingested per insect is not known, our data show that the overallratio of sanguinarine to dihydrosanguinarine shifts towards the lat-ter if the insects fed at higher sanguinarine concentrations (Fig. 5,

Page 5

I. Schütz et al. / Journal of Plant Physiology 171 (2014) 119– 126 123

F ding. Lc d leavp

cci

P

afla(twr

Faob

ig. 3. Alkaloid content in leaves of E. californica and C. majus in response to WFT feeontent after 3 days of larval feeding (right lanes) and compared with unpopulatemol sanguinarine per mg fresh weight.

ompare lines 30 �M and 100 �M sanguinarine). This might indi-ate that the rate of reduction of sanguinarine overproportionallyncreases with the concentration of this alkaloid.

lant responses to WFT feeding

Thrips feeding induces enhanced alkaloid production in Papaver-ceae. The leaves of C. majus and E. californica responded to WFTeeding by an enhanced production of benzophenanthridine alka-oids. After a feeding period of 3 days, populated leaves of C. majusnd E. californica showed a pronounced increase in alkaloid content

Fig. 3). Interestingly, the increase almost exclusively involves theoxic sanguinarine (and chelerythrine in the case of E. californica)hich became the dominating alkaloids under these conditions,

ather than the less toxic dihydroalkaloids.

ig. 4. Alkaloids in WFT larvae after feeding on P. vulgaris, E. californica or C. majus. After

ssayed for their alkaloid content. Two experiments are combined: the left lanes show thf 4 experiments. Figures next to the bands of sanguinarine and dihydrosanguinarine givut cannot exactly be quantified.

eaflets of E. californica and leaf discs of C. majus were assayed for their total alkaloides (left lanes). Figures next to the sanguinarine bands give the alkaloid content in

WFT extracts as eliciting signal in E. californica cell cultures. Thealkaloid responses of the test plants to WFT feeding raised the ques-tion as to whether they are triggered by the mechanical injurycaused by thrips and/or a chemical agent derived from the insect.As the mechanical damage caused by thrips feeding can hardly bemimicked, we tested whether cell suspension cultures of E. cali-fornica responded to a chemical signal derived from the insects.These cultures are known to react very sensitive to elicitor-likeagents and have often been used as a test system for the expres-sion of biosynthetic and defense genes (e.g. Haider et al., 2000, cf.Introduction). For this purpose, water-soluble compounds or a totalhomogenate, prepared from adult WFT bodies, were added to cell

suspension cultures of E. californica. This allows an intense contactof plant cells with the insect material without the need for mechan-ical injuries. As seen from Fig. 6, water soluble molecules extractedfrom WFT bodies barely influenced alkaloid production, whereas

three days feeding on leaf pieces of the indicated plants, larvae were extracted ande lowest and the right lanes the highest alkaloid content per WFT found in a seriese the alkaloid content in pmol per larva. Contents below 0.5 pmol are detectable,

Page 6

124 I. Schütz et al. / Journal of Plant Physiology 171 (2014) 119– 126

F ing soa . Alkal

tpata

D

W

oildbei

FTcD

ig. 5. Alkaloids accumulated in adult WFT during feeding on sanguinarine-containnalysed. Left: sanguinarine solution as used in the experiment (control, no thrips)

he remaining homogenate containing the water-insoluble com-onents caused a strong increase in alkaloid production. Thus, itppears that water-insoluble compounds, and/or solid tissue struc-ures of F. occidentalis are recognized by E. californica cells and elicit

phytoalexin-like response.

iscussion

FT can feed and reproduce on Papaveraceae

Our data show that WFT avoid feeding on and populating leavesf Papaveraceae as long as an alternative plant such as P. vulgariss available. Although different criteria might have influenced theeaf choice experiments (as pre-adaptation of WFT to P. vulgaris,ifferent nutritional values or surface properties of the leaves),

enzophenanthridine alkaloids present in the Papaveraceae likelyxerted a deterrent effect. We identified sanguinarine to be at leastn part responsible for the negative choice of WFT for Papaveraceae.

ig. 6. Components of WFT bodies elicit alkaloid production in cells of E. californica.he total alkaloid content in a cell suspension culture of E. californica in untreatedontrol cells and 24 h after adding WFT homogenate or only soluble compounds.ata are means ± SD, n = 3.

lutions. After feeding for 2 days at the indicated solutions, WFT were extracted andoid content in pmol per insect.

When Papaveraceae are provided as sole feeding source, larvae canfeed and develop on them. WFT developmental time spans and sur-vival rates on C. majus are similar to those described by Kumm(2002) and Kumm and Moritz (2010) for WFT feeding on leavesof P. vulgaris. These were 5.6 ± 1.4 d for the larval stages (1st and2nd instar), 1.0 ± 0.2 d for the prepupal and 2.4 ± 0.4 d for the pupalstages, 9.1 ± 1.1 d in total. It appears that the development is notdeferred in WFT feeding on C. majus.

Alkaloids are taken up and detoxified by WFT

Despite the enhanced production of alkaloids in the populatedleaves, particularly the toxic sanguinarine (and chelerythrine), afraction of WFT larvae were able to feed on C. majus or E. cal-ifornica. WFT larvae cannot avoid contact with these toxins, asshown by the substantial ingestion of alkaloids during feeding(Fig. 4). Interestingly, the insects are able to reduce the ingestedbenzophenanthridines to dihydroalkaloids as seen by the accumu-lation of dihydrosanguinarine in larvae feeding on C. majus or E.californica, and in adult WFT after feeding on an artificial mediumcontaining pure sanguinarine. Dihydrosanguinarine is substantiallyless toxic than sanguinarine (Schmeller et al., 1997; Slaninova et al.,2001; Ulrichová et al., 2001).

While the reduction of sanguinarine to dihydrosanguinarine byWFT is shown here for the first time, it is well known in plants asa main activity in the management of toxic benzophenanthridines.In elicitor-treated cells of E. californica, sanguinarine first accumu-lates in the cell-wall region, followed by its re-absorption into thecells and reduction by the cytosolic enzyme sanguinarine reductase(Weiss et al., 2006). The formed dihydrosanguinarine is channelledinto the later reactions of benzophenanthridine biosynthesis. In thisway, the enzyme sanguinarine reductase drives a recycling processthat allows sanguinarine to be present at the cellular surface and atthe same time to keep this toxin below a critical cellular threshold(Weiss et al., 2006; Müller, 2013).

Sanguinarine reductase was recently cloned from E. califor-nica and sequenced (Vogel et al., 2010). Enzymes with a 3Dstructure similar to sanguinarine reductase are found among sev-eral human and even bacterial proteins and are considered

Page 7

t Phy

a3tmt(tanRetdWba

ibocppabih

tolrpTf

Pb

bsc2et(

mptMBs

maod(hiotae

I. Schütz et al. / Journal of Plan

s potential ancestors, e.g. biliverdine IX reductase and�-hydroxysteroid dehydrogenase (Vogel et al., 2010). The abilityo reduce benzophenanthridines was found even in cultured ani-

al cells (Zdarilova and Roos, unpublished) which correlates withhe broad toxicity of sanguinarine and other benzophenanthridinesVavreckova et al., 1996; Shemon et al., 2004). Our data suggesthat an enzyme with similar function is active in WFT. However,s complete WFT genome sequences are not yet available, it wasot surprising that our search in the only EST library (described byotenberg and Whitfield, 2010) was not successful. Another inter-sting hypothesis suggests that symbiont microorganisms providehe ability to reduce benzophenanthridine alkaloids. At least twoifferent bacterial strains have been described as symbionts ofFT (de Vries et al., 2001; Chanbusarakum and Ullman, 2008),

ut whether these are able to metabolize benzophenanthridinelkaloids remains to be clarified.

Despite the unexplored molecular background, the biolog-cal importance of the thrips’ ability to detect and detoxifyenzophenanthridine alkaloids in potential food plants appearsbvious: it allows the insect to thrive on Papaveraceae by over-oming one of their genuine defence mechanisms, i.e. the enhancedroduction of cytotoxic alkaloids. This extends the spectrum ofotential food plants and underlines the suggested, broad adapt-bility of these insects (Moritz, 2006). In our experiments, this isest demonstrated by the nearly unchanged life cycle of WFT feed-

ng on leaves of C. majus, similar to the benzophenanthridine-freeost P. vulgaris (Kumm and Moritz, 2010).

The differences in the performance of WFT larvae between thewo test plants are not yet understood: the low survival rates seenn E. californica compared to leaves of C. majus might be caused byower nutritional value of the leaflets of E. californica but could alsoeflect a higher sensitivity of thrips to chelerythrine, which wasresent in E. californica at a higher level than in C. majus (cf. Fig. 3).his is in good accordance with the data by Ripa et al. (2009) whoound WFT on flowers of E. californica only occasionally.

lants respond to WFT feeding by enhanced production ofenzophenanthridine alkaloids

Both C. majus and E. californica leaves respond to WFT feedingy the upregulation of benzophenanthridine alkaloid biosynthe-is. Similar responses are evoked by microbial elicitors in culturedells of E. californica (Schumacher et al., 1987; Viehweger et al.,006; Angelova et al., 2010). These similarities indicate that thenhanced production of alkaloids is part of the non-host resistanceo pathogens, i.e. a common evolutionarily heritage of land plantse.g. Nürnberger and Lipka, 2005).

While the molecular structures that are recognized in plant-icrobe interactions (MAMP, microbe associated molecular

atterns) are well investigated (e.g. Boller and Felix, 2009),he search for herbivore associated molecular patterns (HAMP,

ithöfer and Boland, 2008) or herbivore associated elicitors (HAE,onaventure et al., 2011) is only at its beginning, but might revealome overlapping components.

In our preliminary search for a thrips-derived stimulus (beyondechanical wounding) that might be recognized by the plant cells

nd trigger the alkaloid response, we focussed on body extractsf WFT. Although water-soluble components were inactive, thisoes not exclude the potential activity of salivary compoundsas discussed by Kirk, 1997) as these might be present at muchigher concentrations during feeding. On the other hand, water-

nsoluble WFT compounds are clearly recognized by cultured cells

f E. californica. These components need not necessarily be part ofhe insects’ body but might well include contaminating microbesnd/or symbionts found in the gut and other organs (de Vriest al., 2001; Chanbusarakum and Ullman, 2008). The involvement of

siology 171 (2014) 119– 126 125

tospoviruses in the recognition process can be excluded as the WFTstrain used here was free of this contaminant. While for the presentstudy it is not of central importance whether the insects themselvesor their specific symbionts constitute the elicitor-like compoundsto which the plants respond, this question opens several aspectsfor future work.

In summary, our results revealed for the first time mutualresponses by WFT and their hosts in plants producing alkaloids ofthe benzylisoquinoline family. These plants recognize when theyare being fed on by thrips, and respond with an ancestral defencemechanism of enhanced benzophenanthridine alkaloid produc-tion. The insects (and/or their symbionts) are obviously preparedfor these cytotoxic agents by the ability to detoxify them via reduc-tion to less dangerous compounds. This most probably reflectsevolutionary adaptation to food plants, which may include the pos-sibility of horizontal gene transfer, and thus must be considered asan essential element of the evolutionary success of WFT.

Although it is preliminary to speculate how plants and insectspecies develop in mutual response to the adaptations inventedby invader and responder, a WFT-Papaveraceae system might con-stitute an interesting model of research to learn more about theinitiation of an evolutionary arms race between plants and theirherbivores.

Acknowledgement

We thank Gabriele Danders, Renate Kranz, Beate Schöne andAngelika Steller for skilful technical assistance. We are grateful toDr. Laurence Mound, CSIRO, Black Mountain, Australia, for valuablediscussion and help with the English writing of the manuscript.

References

Abe H, Ohnishi J, Narusaka M, Seo S, Narusaka S, Tsuda S, et al. Function of jas-monate in response and tolerance of Arabidopsis to thrip feeding. Plant CellPhysiol 2008;49:68–80.

Abe H, Shimoda T, Ohnishi J, Kugimiya S, Narusaka M, Seo S, et al. Jasmonate-dependent plant defense restricts thrips performance and preference. BMC PlantBiol 2009;9:97.

Abe H, Tomitaka Y, Shimoda T, Seo S, Sakurai T, Kugimiya S, et al. Antagonistic plantdefense system regulated by phytohormones assists interactions among vectorinsect, thrips and a tospovirus. Plant Cell Physiol 2012;53:204–12.

Angelova S, Buchheim M, Frowitter D, Schierhorn A, Roos W. Overproduction of alka-loid phytoalexins in California poppy cells is associated with the co-expressionof biosynthetic and stress-protective enzymes. Mol Plant 2010;3:927–39.

Boller T, Felix G. A renaissance of elicitors: perception of microbe-associated molec-ular patterns and danger signals by pattern-recognition receptors. Annu RevPlant Biol 2009;60:379–406.

Bonaventure G, van Doorn A, Baldwin IT. Herbivore-associated elicitors: FACsignaling and metabolism. Trends Plant Sci 2011;16:294–9.

Bruxelles GL, de, Roberts MR. Signals regulating multiple responses to woundingand herbivores. Crit Rev Plant Sci 2001;20:487–521.

Brødsgaard HF. Frankliniella occidentalis (Thysanoptera; Thripidae): a new pest inDanish glasshouses: a review. Tidsskrift for Planteavl 1989;93:83–91.

Chanbusarakum L, Ullman D. Characterization of bacterial symbionts in Frankliniellaoccidentalis (Pergande), Western flower thrips. J Invert Pathol 2008;99:318–25.

Cheng D, Kirk H, Vrieling K, Mulder PPJ, Klinkhamer PGL. The relationship betweenstructurally different pyrrolizidine alkaloids and western flower thrips resis-tance in F2 hybrids of Jacobaea vulgaris and Jacobaea aquatica. J Chem Ecol2011;37:1071–80.

Childers CC, Achor DS. Thrips feeding and oviposition injuries to economic plants,subsequent damage and host responses to infestation. In: Parker BL, Skinner M,Lewis T, editors. Thrips Biology and Management. London: Plenum Press; 1995.p. 31–51.

Cline SD, McHale RJ, Coscia CJ. Differential enhancement of benzophenanthridinealkaloid content in cell-suspension cultures of Sanguinaria canadensis underconditions of combined hormonal deprivation and fungal elicitation. J Nat Prod1993;56:1219–28.

Colombo ML, Bosisio E. Pharmacological activities of Chelidonium majus L. (Papaver-

aceae). Pharmacol Res 1996;33:127–34.

Dixon RA, Harrison MJ, Paiva NL. The isoflavonoid phytoalexin pathway – fromenzymes to genes to transcription factors. Physiol Plant 1995;93:385–92.

El Wakeil NE, Volkmar C, Sallam AA. Jasmonic acid induces resistance to economi-cally important insect pests in winter wheat. Pest Manag Sci 2010;66:549–54.

Page 8

1 t Phy

F

F

G

G

H

H

H

H

H

K

K

K

K

K

K

L

L

M

M

M

M

M

M

N

N

26 I. Schütz et al. / Journal of Plan

addeeva MD, Beliaeva TN. Sanguinarine and ellipticine cytotoxic alkaloids iso-lated from well-known antitumor plants. Intracellular targets of their action.Tsitologiya 1997;39:181–208.

ärber K, Schumann B, Miersch O, Roos W. Selective desensitization of jasmonate-and pH-dependent signaling in the induction of benzophenanthridine biosyn-thesis in cells of Eschscholzia californica. Phytochemistry 2003;62:491–500.

lendinning JL. How do herbivorous insects cope with noxious secondary plantcompounds in their diet? Entomol Exp Appl 2002;104:15–25.

undlach H, Müller MJ, Kutchan TM, Zenk MH. Jasmonic acid is a signal trans-ducer in elicitor-induced plant cell cultures. Proc Natl Acad Sci USA 1992;89:2389–93.

aider G, von Schrader T, Füßlein M, Blechert S, Kutchan TM. Structure–activityrelationships of synthetic analogs of jasmonic acid and coronatine on inductionof benzo[c]phenanthridine alkaloid accumulation in Eschscholzia californica cellcultures. Biol Chem 2000;381:741–8.

ammerschmidt R, Kagan IA. Phytoalexins into the 21st century. Physiol Mol PlantPathol 2001;59:59–61.

artmann T, Theuring C, Witte L, Schulz S, Pasteels JM. Biochemical processing ofplant acquired pyrrolizidine alkaloids by the neotropical leaf-beetle Platyphoraboucardi. Insect Biochem Mol Biol 2003;33:515–23.

artmann T, Theuring C, Beuerle T, Klewer N, Schulz S, Singer MS, et al. Specificrecognition, detoxification and metabolism of pyrrolizidine alkaloids by thepolyphagous arctiid Estigmene acrea. Insect Biochem Mol 2005;35:391–411.

einze M, Steighardt J, Gesell A, Schwartze W, Roos W. Regulatory interaction ofthe G protein with phospholipase A2 in the plasma membrane of Eschscholziacalifornica. Plant J 2007;52:1041–51.

irk WDJ. Feeding. In: Lewis T, editor. Thrips as crop pests. New York: Cab Interna-tional Oxon; 1997. p. 119–74.

irk WDJ, Terry LI. The spread of the western flower thrips Frankliniella occidentalis(Pergande). Agr Forest Entomol 2003;5:301–10.

lein M, Roos W. Handling dangerous molecules: transport and compartmentationof plant secondary products. In: Osbourn A, Lanzetti C, editors. Plant naturalproducts. New York: Springer; 2009. p. 229–67.

umm S. Reproduction, progenesis, and embryogenesis of thrips (Thysanoptera.Insecta). Martin Luther University Halle & Wittenberg; 2002. p. 31–6, PhD theses.

umm S, Moritz G. First detection of Wolbachia in arrhenotokous populations ofthrips species (Thysanoptera: Thripidae and Phlaeothripidae) and its role inreproduction. Environ Entomol 2008;37:1422–8.

umm S, Moritz G. Life-cycle variation, including female production by vir-gin females in Frankliniella occidentalis (Thysanoptera: Thripidae). J Appl Ent2010;134:491–7.

ambdon PW, Hassall M. Do plant toxins impose constraints on herbivores? Aninvestigation using compartmental analysis. Oikos 2001;93:168–76.

angel D, Ober D. Evolutionary recruitment of a flavin-dependent monooxygenasefor stabilization of sequestered pyrrolizidine alkaloids in arctiids. Phytochem-istry 2011;72:1576–84.

acel M, Bruinsma M, Dijkstra SM, Ooijendijk T, Niemeyer MH, Klinkhamer PGL.Differences in effects of pyrrolizidine alkaloids on five generalist insect herbivorespecies. J Chem Ecol 2005;31:1493–508.

acel M. Attract and deter: a dual role for pyrrolizidine alkaloids in plant–insectinteractions. Phytochem Rev 2011;10:75–82.

aharijaya A, Vosman B, Verstappen F, Steenhuis-Broers G, Mumm R, Purwito A,et al. Resistance factors in pepper inhibit larval development of thrips (Franklin-iella occidentalis). Entomol Exp Appl 2012;145:62–71.

ithöfer A, Boland W. Recognition of herbivory-associated molecular patterns. PlantPhysiol 2008;146:825–31.

üller HU. Die Rolle der Sanguinarin-Reduktase in der Kontrolle desBenzophenanthridin-Metabolismus in Pflanzen. Martin Luther UniversityHalle & Wittenberg; 2013 (PhD theses).

oritz G. Die Thripse – Fransenflügler, Thysanoptera. Hohenwarsleben: WestarpWissenschaften; 2006.

arberhaus I, Zintgraf V, Dobler S. Pyrrolizidine alkaloids on three trophic levels –evidence for toxic and deterrent effects on phytophages and predators. Chemoe-cology 2005;15:121–5.

ürnberger T, Lipka V. Non-host resistance in plants. New insights into an old phe-nomenon. Mol Plant Pathol 2005;6:335–45.

siology 171 (2014) 119– 126

Preininger V. Chemotaxonomy of Papaveraceae and Fumariaceae. In: Brossi A, editor.The alkaloids: chemistry and pharmacology, vol. 29. Waltham: Academic Press;1986. p. 1–98.

Ripa R, Funderburk J, Rodriguez F, Espinoza F, Mound LA. Population abundanceof Frankliniella occidentalis (Thysanoptera: Thripidae) and natural enemies onplant hosts in Central Chile. Environ Entomol 2009;38:333–44.

Roos W, Evers S, Hieke M, Tschöpe M, Schumann B. Shifts of intracellular pHdistribution as a part of the signal mechanism leading to the elicitation of ben-zophenanthridine alkaloid biosynthesis. Plant Physiol 1998;118:349–64.

Roos W, Viehweger K, Dordschbal B, Schumann B, Evers S, Steighardt J, et al. Intracel-lular pH signals in the induction of secondary pathways – the case of Eschscholziacalifornica. J Plant Physiol 2006;163:369–81.

Rotenberg D, Whitfield AE. Analysis of expressed sequence tags for Frankliniellaoccidentalis, the western flower thrips. Insect Mol Biol 2010;19:537–51.

Schmeller T, Latz-Brüning B, Wink M. Biochemical activities of berberine, palma-tine and sanguinarine mediating chemical defence against microorganisms andherbivores. Phytochemistry 1997;44:257–66.

Schumacher HM, Gundlach H, Fiedler F, Zenk MH. Elicitations of benzophenan-thridine alkaloid synthesis in Eschscholtzia cell cultures. Plant Cell Rep1987;6:410–3.

Shemon AN, Sluyter R, Conigrave AD, Wiley JS. Chelerythrine and other ben-zophenanthridine alkaloids block the human P2X(7) receptor. Br J Pharmacol2004;142:1015–9.

Slaninova I, Taborska E, Bochorakova H, Slanina J. Interaction ofbenzo[c]phenanthridine and protoberberine alkaloids with animal andyeast cells. Cell Biol Toxicol 2001;17:51–63.

Tutin TG, Burges NA, Chater AO, Edmondson JR, Heywood VH, Moore DM, et al. FloraEuropaea, vol. 1, Psilotaceae to Platanaceae. Cambridge: Cambridge UniversityPress; 1993. p. 301–2.

Ulrichová J, Dvorák Z, Vicar J, Lata J, Smrzová J, Sedo A, et al. Cytotoxicity of nat-ural compounds in hepatocyte cell culture models. The case of quaternarybenzo[c]phenanthridine alkaloids. Toxicol Lett 2001;125:125–32.

Vavreckova C, Gawlik I, Muller K. Benzophenanthridine alkaloids of Chelidoniummajus. II. Potent inhibitory action against the growth of human keratinocytes.Planta Med 1996;62:491–4.

Viehweger K, Dordschbal B, Roos W. Elicitor-activated phospholipase A2 gen-erates lysophosphatidylcholines that mobilize the vacuolar H+ pool for phsignaling via the activation of Na+-dependent proton fluxes. Plant Cell Online2002;14:1509–25.

Viehweger K, Schwartze W, Schumann B, Lein W, Roos W, The G. Protein controlsa pH-dependent signal path to the induction of phytoalexin biosynthesis inEschscholzia californica. Plant Cell 2006;18:1510–23.

Vogel M, Lawson M, Conrad U, Sippl W, Roos W. Sanguinarine reductase – struc-ture and mechanism of a novel enzyme serving the protection from cytotoxicalkaloids. J Biol Chem 2010;285:18397–406.

de Vries EJ, Breeuwer JAJ, Jacobs G, Mollema C. The association of western flowerthrips, Frankliniella occidentalis, with a near Erwinia species gut bacteria: tran-sient or permanent? J Invert Pathol 2001;77:120–8.

Walling LL. The myriad plant responses to herbivores. J Plant Growth Regul2000;19:195–216.

Weiss D, Baumert A, Vogel M, Roos W. Sanguinarine reductase, a key enzyme ofbenzophenanthridine detoxification. Plant Cell Environ 2006;29:291–302.

Whitfield A, Ullman DE, German TL. Tospovirus–thrips interactions. Annu ReviewPhytopathol 2005;43:459–89.

Wink M, Schmeller T, Letz-Brünning B. Modes of action of allolochemical alkaloids:interaction with neuroreceptors, DNA, and other molecular targets. J Chem Ecol1998;24:1881–937.

Wolff J, Knipling L. Antimicrotubule properties of benzophenanthridine alkaloids.Biochemistry 1993;32:13334–9.

Yudin LS, Cho JJ, Mitchell WC. Host range of western flower thrips, Frankliniella occi-dentalis (Thysanoptera: Thripidae), with special reference to Leucaena glauca.

Environ Entomol 1986;15:1292–5.

Zenk MH. The formation of benzophenanthridine alkaloids. Pure Appl Chem1994;66:2023–8.

Ziegler J, Facchini P. Alkaloid biosynthesis: metabolism and trafficking. Annu RevPlant Biol 2008;59:735–69.