Validation of Serine/Threonine Protein Phosphatase As the Herbicide Target Site of Endothall ⇑ Joanna Bajsa, Zhiqiang Pan, Franck E

Pesticide Biochemistry and Physiology 102 (2012) 38–44

Contents lists available at SciVerse ScienceDirect

Pesticide Biochemistry and Physiology

journal homepage: www.elsevier.com/locate/pest

Validation of serine/threonine protein phosphatase as the herbicide target site of endothall ⇑ Joanna Bajsa, Zhiqiang Pan, Franck E. Dayan, Daniel K. Owens, Stephen O. Duke

USDA, ARS Natural Products Utilization Research Unit, University, MS 38677, USA article info abstract

Article history: Endothall, an older commercial herbicide, and cantharidin, a natural product from the blister beetle (Epic- Received 4 October 2011 auta spp.), are close chemical analogues. A comparison of the effect of endothall and cantharidin on plants Accepted 20 October 2011 revealed a similarity in their level of phytotoxicity on both Arabidopsis thaliana and Lemna paucicostata. Available online 29 October 2011 Cantharidin is a potent inhibitor of animal serine/threonine protein phosphatases. Protein phosphatases and kinases maintain a sensitive balance between phosphorylated and dephosphorylated forms of pro- Keywords: teins playing important roles in signal transduction pathways. In this study, we found endothall and can- Cantharidin tharidin to both completely inhibit plant serine/threonine protein phosphatases, and their relative Endothall inhibitory activities were similar to their relative phytotoxicities. Both compounds acted as slow, irre- Lemna paucicostata Serine/threonine protein phosphatase versible inactivators of the serine/threonine protein phosphatase activities. Transcription of several genes determined to be affected by the inhibition of these protein phosphatases by cantharidin in A. thaliana by transcriptome analyses were affected similarly by endothall, but in a more pronounced way. Therefore, the molecular target site of endothall in plants is similar to that of cantharidin in animals, namely, serine/threonine protein phosphatases responsible for regulating an array of biochemical processes. This mode of action is unlike any other commercial herbicide. Published by Elsevier Inc.

1. Introduction aphrodisiac associated with human poisonings [11,12].Itis currently being investigated as an anticancer drug lead [13]. In ani- Endothall (7-oxabicyclo[2.2.1]heptane-2,3-dicarboxylic acid, mals, cantharidin is a strong inhibitor of serine/threonine protein Fig. 1) was introduced in the 1950s, yet it is one of the few herbi- phosphatases (PPs) [14], a broad class of PPs associated with sig- cides still listed as having an unknown mode of action [1,2].Ina naling and control of numerous cellular processes in many organ- chemical class of its own, endothall induces symptoms similar to isms (e.g., [15]). The catalytic domain of all PP subfamilies is highly chilling injury on the foliage of treated plants and ultimately conserved in animals, plants and protozoans [16]. Inhibitors, such results in severe growth inhibition [2]. Investigations on the mode as cantharidin and endothall, bind to a hydrophobic pocket of the of action of endothall reported effects on mRNA [3] and protein [4] PP active site. According Bertini et al. [17], two metal ions (Mn synthesis. Additionally, endothall causes membrane dysfunction and/or Zn) and the amino acids Tyr272, Val429, Phe260, Gln242, [5], and the WSSA Herbicide Handbook suggests that its action Arg89 and Arg214 are involved in the binding of these compounds may initiate at the membrane level [2]. However, loss of mem- in the human member of subfamily PP5. Endothall has two binding brane integrity can occur in plants exposed to herbicides with conformations: one similar to cantharidin’s orientation and a modes of action involving a wide array of molecular target sites second rotated by ca. 120°. The structural similarity between [6,7]. The primary target site of endothall in plants has remained endothall and cantharidin has been apparent to animal scientists elusive for many years. and similar mechanisms of action on animal serine/threonine PPs Endothall is a close analog of cantharidin (Fig. 1) from the blis- has been conﬁrmed (e.g., [18]). ter beetle (Epicauta spp.) and the Spanish ﬂy (Lytta vesicatoria). Scientists often use herbicides as molecular probes to study Some insects accumulate up to 5% of their dry weight of this natu- plant biochemical processes [19]. Endothall has recently been used ral defense terpenoid. Cantharidin has had various uses as a phar- to manipulate serine/threonine PP activity and investigate a num- maceutical over the years (e.g., [8–10]), and is infamous as an illicit ber of plant processes including disease resistance [20–24], metal ion toxicity [25,26], apoptosis [27], and regulation of anthocyanin biosynthesis [28]. Endothall strongly inhibits alfalfa serine/ Abbreviation: PP, protein phosphatase. ⇑ Corresponding author. Fax: +1 662 915 1036. threonine PP2A (PP2A) activity, as well as serine/threonine PP1 E-mail address: [email protected] (S.O. Duke). (PP1) to a lesser degree, which is accompanied by a profound effect

0048-3575/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.pestbp.2011.10.007 J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44 39

100 lmol m2 s1 PAR. For RNA and protein preparations,

twelve-day-old seedlings were sprayed with the I30 concentration (concentration required to reduce chlorophyll content by 30% two days after treatment) of cantharidin (200 lM). Endothall was used at the same concentration. A. thaliana shoots were harvested 2, 10, and 24 h after spraying. Each experiment (time point) had three independent replicates of both treated and untreated (control) samples. Fig. 1. Structures of endothall, cantharidin and cantharidic acid. Duckweed (Lemna paucicostata (L.) Hegelm.) bioassays were conducted as previously described [35]. L. paucicostata stocks were grown on modified Hoagland media. The media contained: on plant cell coordination of chromosomal and microtubule events 1 1 1 1515 mg L KNO3, 680 mg L KH2PO4, 492 mg L MgSO4.7H2O, during mitosis [29]. Tresch et al. [30] recently connected the inhi- 1 1 1 20 mg L Na2CO3, 1180 mg L Ca(NO3)24H2O, 0.5 mg L H3BO3, bition of serine/threonine PPs by endothall to its pronounced ef- 1 1 1 0.05 mg L ZnSO4, 0.12 mg L Na2MoO4, 0.47 mg L MnCl2, fects on the plant cell cycle, which were similar to phenotypic 1 1 1 0.025 mg L CoCl2, 0.025 mg L CuSO4, 18.355 mg L Fe-EDTA. effects on plants with a mutant regulatory subunit (TON2) of a pro- The medium was adjusted to pH 5.5 with 1 M NaOH and then filter tein phosphatase (PP2A) in Arabidopsis thaliana. However, the phy- sterilized through a 0.2 lm sterile filter. All of the L. paucicostata totoxicity of endothall was not directly compared to its effect on used in these studies originated from a single L. paucicostata colony serine/threonine PP activity. We previously reported that canthar- (an aggregate of one mother and two daughter fronds) to assure idin was highly phytotoxic, apparently due to its effects on plant genetic uniformity. For the tests, plants were taken while still in serine/threonine PPs [31,32]. Herein, we validate the relationship exponential growth from a 4 to 5-day-old stock culture. between the herbicidal activity of endothall and inhibition of ser- Each concentration of a compound was tested in three repli- ine/threonine PPs supporting this data with an analysis of the com- cates. If required, the pH of the media was adjusted to 5.6 after pound’s effects on the expression of selected A. thaliana gene adding test chemicals. Both the initial screening and replicate ser- transcripts. ies tests were conducted in an incubator (Model #CU-36L, Percival Scientific, Boone, IA, USA) with white light (94 lEm2 s1) using 2. Materials and methods non-pyrogenic polystyrene sterile six-well plates (CoStar 3506, Corning Inc., Corning, NY, USA) with a lid. Each well contained 2.1. Chemicals 4950 lL of the Hoagland media plus 50 lL of water, or the solvent (as control), or the appropriate concentration of the compound dis- Technical grade endothall, cantharidin and all other chemicals solved in a suitable solvent. The final concentration of the solvent for the experiments were obtained from Sigma–Aldrich (St. Louis, was therefore 1% by volume. Each well was inoculated with two, MO, USA), with the exception of Tween 20 (Fisher Scientific, Pitts- three-frond colonies of approximately the same size. The selected burgh, PA, USA) and Murashige and Skoog (MS) basal salt mixture colonies had a ‘‘Mickey Mouse’’ appearance, where the developing (Phytotechnologies Laboratories, Shawnee Mission, KS, USA). daughter fronds were significantly smaller and still attached to the maternal frond. Total frond area per well was recorded by the im- 2.2. Culture and treatment of plants age-analysis-system Scanalyser (LemnaTec, Würselen, Germany) once per day from day 0 to day 7. Wild-type A. thaliana Columbia ecotype was grown on MS medium, five seedlings per well of a six-well plate at 24 °C and under constant light intensity of 100 lmol m2 s1 PAR. Twelve-day-old 2.3. Protein phosphatase assays seedlings were sprayed using chromatography sprayers (VWR) with a range of endothall or cantharidin concentrations in solution One gram of frozen plant material was ground in a mortar and of 0.1% acetone and 0.15% Tween 20 per container. Controls were pestle and suspended in 25 mM Tris-HCl pH 7.0%, 0.1% b-mercap- sprayed with a solution of 0.1% acetone and 0.15% Tween 20 solu- toethanol, 1 mM EDTA, 1 mM benzamidine, and 0.05% Tween 20 tion. The spray solution was applied to cover the leaf tissue and plant protease inhibitors cocktail (Sigma–Aldrich). The tissue entirely. Two days after spraying, shoots of the seedlings were har- was homogenized with a Polytron instrument (PT 3100, Kinemat- vested and chlorophyll content was determined by the method of ica AG, Littau-Lucerne, Switzerland) and centrifuged at 17,500g for Hiscox and Israelstam [33]. 30 min. The cell free extract was desalted over a Sephadex G25 For the root elongation assay, A. thaliana (Col. ecotype) seeds spin column to remove contaminating free phosphate. Protein con- were surface sterilized and cultured essentially as described in centration was determined by the Bradford method [36]. Baerson et al. [34] with the following minor modifications. Semi- Serine/threonine PP activity in the extracts was measured using solid media was composed of 0.5X Murashige and Skoog salts, 1X a serine/threonine PP assay system kit (Promega, Madison, WI, Gamborg’s B5 vitamins, and 1.0% (w/v) sucrose that was adjusted USA) in half-area 96-well plates (Corning Inc.). Dose-response to pH 5.7 with KOH prior to the addition of 1.0% (w/v) agar. Can- curves of the effect of cantharidin and endothall were performed tharidin and endothall were freshly prepared in acetone and added on desalted, cell-free extracts isolated from A. thaliana or L. pauci- to sterilized media which had been cooled to 50 °C such that final costata seedlings. Reactions were incubated with predetermined inhibitor concentrations of 0, 0.3, 1, 3, 10, 30, 100, 300, and concentrations of inhibitors at room temperature for 15 min, after 1000 lM were obtained in individual plates. Seeds were placed which the serine/threonine peptide substrate was added producing in a straight line approximately 3.0 cm. from the edge of an overall reaction volume of 50 lL. The reaction was carried out 9.0 9.0 1.5 cm square petri plates. Seeds were stratified hori- for 30 min, subsequently stopped by adding 50 lL of molybdate zontally at 4 °C for 3 days then transferred to a vertical orientation dye/additive mixture and incubated at room temperature for an in a growth chamber. additional 15 min for color development. The absorbance of a For transcript analysis, wild-type A. thaliana Columbia ecotype molybdate:malachite green phosphate complex was measured was grown on MS medium, 200 seedlings per 10 8 cm glass with a PowerWave XS plate reader (Bio-Tek, Winooski, VT, USA), container (PYREX) at 24 °C and under constant light intensity of with a 600 nm filter and software (KC Junior, version 1.41.a). 40 J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44

The mechanism of interaction between the inhibitors (endothall the genes of interest. Results were analyzed using the Bio-Rad CFX and cantharidin) with their molecular target sites was evaluated Manager software (version 2.0). PCR products were also assessed via two assays. First, a time-course study checked the serine/thre- by electrophoresis on 1.5% agarose gel. onine PP activity in L. paucicostata total soluble protein extracts upon pre-incubation with the inhibitors for 0, 5, 10, 20, 30, 40, 2.5. Statistical analysis and 50 min. The assay described above contained 2 lg total soluble protein per assay. A subsequent experiment was performed by Data from dose–response experiments were analyzed using the titrating the amount of total soluble protein per assay from 1 to dose–response curve (DRC) module [37] and R version 2.2.1 [38]. 6 lg per assay. Means and standard deviations were obtained using the raw data,

and I50 values were obtained from the parameters in the regression curves. The regression curves were imported into SigmaPlot ver- 2.4. Gene transcript analysis sion 10 (Systat Software Inc., San Jose, CA, USA). Pearson correlation coefficients were calculated using the statistical module of Total RNAs were isolated using the Trizol reagent (Invitrogen, SigmaPlot. Carlsbad, California, USA) according to manufacturer’s instructions. The RNA samples were treated with DNase I using an RNase-Free DNase kit as per the manufacturer’s instructions (Qiagen, Valencia, 3. Results CA, USA) to remove residual DNA contamination. The RNA recov- ered was then re-purified with the RNeasy Plant Mini Kit (Qiagen, 3.1. Symptoms and relative phytotoxicity of endothall and cantharidin Valencia, CA, USA). The integrity of purified total RNA was established by 1% agarose gel electrophoresis using ethidium bromide A. thaliana seedlings developed visible symptoms, such as re- staining. The concentration and purity of total RNA was deter- duced growth, loss of chlorophyll, and hypersensitive response- mined spectrophotometrically. like lesions within the first 24 h after treatment. The endothall First strand cDNAs were synthesized from total RNA using the and cantharidin concentrations that reduced chlorophyll in A. tha- iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA) as per the liana by 50% (I50) were 1.36 ± 0.22 and 0.73 ± 0.08 mM, respec- manufacturer’s instructions. Quantitative real time PCR was per- tively, after two days of treatment (Fig. 2A). A. thaliana seedlings formed in triplicate using the iQ SYBR Green Supermix (Bio-Rad) growing on media containing endothall or cantharidin were more for double-stranded DNA synthesis, and the fluorescence signal sensitive to these compounds, which reduced root growth by 50% was monitored on line using a Bio-Rad MiniOpticon real-time at 13.9 ± 0.73 and 73.2 ± 5.11 lM, respectively (Fig. 2B). Endothall PCR detection system (Bio-Rad) according to the manufacturer’s and cantharidin were even more potent against the growth of L. directions. Primers for PCR reaction were designed and analyzed paucicostata, with I50 values of 10.0 ± 1.9 and 0.24 ± 0.02 lM, using Primer Express Software (Table 1). Specificity and exclusivity respectively (Fig. 2C). of chosen primers was determined using BLAST program (http:// blast.ncbi.nlm.nih.gov/Blast.cgi) against the A. thaliana genome 3.2. Effects on serine/threonine protein phosphatase activity database. RNA levels were normalized using Actin 2 (At3g18780) as a reference gene. Efficiencies were calculated using a 10-fold se- Endothall and cantharidin inhibited serine/threonine PP activity rial dilution series of DNA. Reaction conditions were as follows: in a total soluble protein extract from A. thaliana seedlings. The initial denaturation 95 °C for 2 min, followed by forty cycles of apparent 50% inhibition (I50 app) values of endothall and canthar- 95 °C for 20 s, 60 °C for 10 s, 72 °C for 30 s, and a subsequent stan- idin in an in vitro serine/threonine PP assay were 2.0 ± 0.3 and dard dissociation protocol to check for the specificity of primers to 0.6 ± 0.07 lM, respectively (Fig. 3A). The dose/response curves

Table 1 Primers used for quantitative real time RT-PCR assays.

Symbol Locus Gene description Primer sequencea ACT2 At3g18780 Actin 2 F: 50 – TGGGCAAGTCATCACGATTGGTGC – 30 R: 50 – TCCATTCCCACAAACGAGGGCTGG – 30 WRKY6 At1g62300 Transcription factor for phosphate1 (Pho1) expression F: 50 – AATTTCACGGCGGCTCTTGCAGC – 30 R: 50 – TTGTTGTTTCCTTCGCCGTCGTGG – 30 ASA1 At5g05730 Alpha subunit of anthranilate synthase F: 50 – AACGGCGTGAGTGGGTGGCTTATC – 30 R: 50 – AGACCAGCGGCTTTGTTCTGGCAC – 30 DREB At5g25810 DREB subfamily A-4 of ERF/AP2 transcription factor family (TINY) F: 50 – TCTCTCGTGTCTTCCTCCGCGAC – 30 R: 50 – TCTGACCATTGCGGTGGATAAGGC – 30 ACS6 At4g11280 1-Aminocyclopropane-1-carboxylate (ACC) synthase F: 50 – TGACACGAGGCGGTTCGATGATGG – 30 R: 50 – TGTGCACGGACTAGCGGAGAAGGC – 30 T14L22.10 At1g51890 Leucine-rich repeat protein kinase family protein F: 50 – AGTGGCTAGCGTGTTTGCTGTGC – 30 R0 50 – CCGTGCCAGTAGTGAATGATCGGG – 30 PKS1 At2g02950 Soluble protein binding to PHYA or PHYB F: 50 – TGGGGTGTAAATGCGCATGCTCTG – 30 R: 50 – TCGGATCAGCGCTTCTCTTGACCG – 30 GASA1 At1g75750 GA-responsive GAST1 protein homolog F: 50 – TGTGTAGCACGGTGCAGGCTTTCG – 30 R: 50 – TTATGGGCACTTGCGGCGTCCAC – 30 C3CH4 At5g27420 CNI1 RING type ubiquitin ligase F: 50 – ACGAGCAGACCGGTGGGCTTTTC – 30 R: 50 – AACTCTCGGCGAAGGAACCGGC – 30 PBP1 At5g54490 PINOID (PID) – binding protein F: 50 – ACGCATCGACGGTTCTTGGACTCG – 30 R: 50 – GTCACAACACACCGTGACGCTTCC – 30 SAUR68 At1g29510 Auxin responsive SAUR protein F: 50 – GCAACAAAGAGCAGCCCTCCACAG – 30 R: 50 – CGCCGTGTAGACCACGAAACAACC – 30 RPM1 At3g07040 Disease resistance protein RPM1 F: 50 – AACGCATTCGCGGAGAAGGGAGTG – 30 R: 50 – AGCTTGCCTTGGCCGCCTAAGATG – 30

a F indicates forward and R indicates reverse. J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44 41

120 120 A A 100 100

80 80

60 60

40 40 Ser-Thre PPP Activity (%) Ser-Thre PPP Activity

Chlorophyll content (% control) Chlorophyll 20 20

0 0 0 10 100 1000 0 0.1 1 10 100 Inhibitor (µM) Inhibitor (µM) 40 B 100 B 30 80

20 60

Root length (mm) 40 10

Ser-Thr PPP Activity (%) Ser-Thr PPP Activity 20

0 0 0 1 10 100 1000 Inhibitor (µM) 00.010.11 10 Inhibitor (µM) 1000 C Fig. 3. Dose-response of endothall (N) and cantharidin (.) on serine/threonine protein phosphatase in total soluble protein extracts of (A) Arabidopsis thaliana and (B) Lemna paucicostata. Data represent means of 3 replications and the standard 800 deviation.

600 The inhibition exerted by these compounds on serine/threonine PP activity in L. paucicostata total soluble protein extracts increased 400 over time (Fig. 4A). The onset of inhibition was more rapid with cantharidin than with endothall (when tested at their respective I app concentrations). Titration of the amount of protein against 200 50 Lemna growth (% increase) Lemna growth a constant concentration of either endothall or cantharidin in the reaction assay (Fig. 4B) supports the time-dependent inactivation 0 of serine/threonine PP. However, the lines were not completely parallel. 00.1110100 Inhibitor (µM) 3.3. Effects on gene transcription Fig. 2. Dose-response of endothall (N) and cantharidin (.) on (A) chlorophyll content of Arabidopsis thaliana and (B) growth of Lemna paucicostata. Data represent Since the RT-PCR experiment required A. thaliana to be exposed means of 3 replications and the standard deviation. to endothall and cantharidin for relatively short period of times, growing the seedlings on plates as was done for the root length were very broad for both inhibitors, with complete inhibition of experiment (Fig. 2B) was not an option. Therefore, seedlings were serine/threonine PP at about 100 lM. This may be due to the crude sprayed with a 200 lM of each compound, a concentration that extract containing a number of different serine/threonine PPs with was near the beginning of phytotoxicity on the dose-response a range of sensitivities to the inhibitors. These compounds were curve for both compounds. After spraying, samples were taken at more active on serine/threonine PP activity extracted from 2, 10 and 24 h.

L. paucicostata, with I50 app values of 1.3 ± 0.05 and 0.068 ± The effects of endothall and cantharidin on A. thaliana gene 0.005 lM for endothall and cantharidin, respectively (Fig. 3B). transcription were compared by quantitative real-time RT-PCR 42 J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44

500 A 60 2 hour A 50 400 40 g/h

µ Pearson correlation 30 300 coefficient = 0.912 20 p = 0.0001 200 10

pmol phosphate/ 0 100 Normalized Fold Expression Normalized -10

0 -20 0 102030405060 4 1 6 6 1 1 B 1 1 8 G5 H BP KY M SA E SA KS 6 T1 P R ACSRP A DR P UR Incubation with inhibitor (min) A C3C W GA SA 80 10 hour 100 B B 60

80 40 -1 Pearson correlation

60 20 coefficient = 0.745 p = 0.0008

40 0 pmol phosphate min

Normalized Fold Expression Normalized -20 20

-40 P1 M1 EB S1 0 ASA1 ACS6 PB P R PK C3CH4 WRKY6AT1G5 R D GASA1 0246 SAUR68 20 Total soluble protein (µg) 24 hour C Fig. 4. (A) Time-dependence of serine/threonine PP inhibition in total soluble protein extract from L. paucicostata. The enzyme preparation and inhibitors were 10 incubated together as indicated on the x-axis prior to initiation of the enzymatic Pearson correlation reaction by the addition of the substrate. (B) Titration of serine/threonine PP coefficient = 0.864 activity alone (s) or in the presence of 1 or 0.3 lM endothall (N) and cantharidin (.), respectively. Data represent means of 3 replications and the standard p = 0.0006 deviation. 0 assays. Eleven genes whose expression responded to cantharidin in a previous microarray transcriptome analysis [31,32] were selected for this study (Table 1). Genes were chosen based on their -10 association with significant functions, and, additionally, some of Normalized Fold Expression Normalized these were among those most affected by cantharidin. These functions included various types of defense responses and hormonal, -20 and phytochrome signaling, such as hypersensitive response 1 6 6 4 1 1 1 B (RPM1 (AT3G07040)); flagellin signaling (WRKY6 (AT1G62300)); SA CS KY G5 BP M1 KS 68 SA E A A R AT1 P RP P AUR DR ethylene signaling (DREB (At5g25810), ACS6 (At4g11280) and W C3CH S GA ASA1 (AT5G05730)); abscisic acid responses (C3CH4 (At5g27420) Fig. 5. Relationship between the effect of endothall (h) and cantharidin (j) on the and GASA1 (AT1G75750)); auxin responses (PBP1 (At5g54490) normalized expression of selected Arabidopsis thaliana genes after (A) 2, (B) 10 and and SAUR68 (At1g29510); and phytochrome signaling (PBP1 (C) 24 h exposure. Data represent means of 3 replications and the standard (AT1G51890) and (PKS1) (At2g02950)). RNAs extracted from can- deviation. tharidin and endothall treated seedlings were analyzed. Statistical analysis revealed a strong correlation between the genes’ responses to cantharidin and endothall treatments, which is consis- 4. Discussion tent with the two compounds having similar effects on plants (Fig. 5). The Pearson correlation coefficient was highest after 2 h Endothall and cantharidin were more active on L. paucicostata of exposure to the compounds (Fig. 5A). The strength of the corre- than A. thaliana (Fig. 2A and B), which is in agreement with a lation decreased after 10 h (Fig. 5B), but appeared to be restored previous study showing that monocots are more sensitive than after 24 h of exposure to the herbicides (Fig. 5C). dicots [39]. The relative effects of cantharidin and endothall as J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44 43 phytotoxins on A. thaliana and L. paucicostata (Fig. 2) were similar several of these critical PPs by endothall causes signaling disrup- to their effects on serine/threonine PP enzyme activity (Fig. 3). tion that can be lethal at high enough doses of the herbicide. Cantharidin was more active than endothall as a phytotoxin on While the potency of endothall on serine/threonine PP activity chlorophyll content of A. thaliana and growth of L. paucicostata was lower (Fig. 3) and its binding appeared to develop more slowly and as an enzyme inhibitor, especially in L. paucicostata. This (Fig. 4) than that observed with cantharidin, it had a more pro- implies a connection between inhibition of serine/threonine PP nounced effect on the expression of genes than cantharidin activity (up to 19 forms of the enzyme in A. thaliana are presumed (Fig. 5). The general effects of these two serine/threonine PP inhib- to be inhibited by cantharidin [31]) and phytotoxicity. itors on transcription of the eleven genes selected for examination Endothall and cantharidin inhibited serine/threonine PP activity were similar. Both compounds caused an initial change in expres- in total soluble protein extracts from A. thaliana and L. paucicostata. sion, which resulted in a high Pearson correlation coefficient In all instances, cantharidin was more potent than endothall, and (Fig. 5A). However, the effect of cantharidin began to subside after inhibition was greater on L. paucicostata. This suggests that there 10 h, whereas the effect of endothall continued to be accentuated may be differences between serine/threonine PPs from dicotyle- (Fig. 5B). This suggests that the commercial herbicide remains bio- donous and monocotyledonous species (Fig. 3A and B). However, active for a longer period of time than the natural product in planta. little is known about the interactions of these compounds with The effect of endothall on transcription of these genes was still serine/threonine PP from different plant species. higher than cantharidin after 24 h exposure. Nevertheless, the phy- The anhydride-type serine/threonine PP inhibitors are more totoxicity of cantharidin to A. thaliana when sprayed onto plants active than their corresponding di-acids in vitro, which has led to was slightly higher than endothall (Fig. 2A). Examination of an- some debate about whether cantharidin or its hydrolyzed form, other set of the many genes affected by cantharidin [31,32] may cantharidic acid (Fig. 1), is the serine/threonine PP inhibitory form have given different relative results, but we expect that the general of the compound. However, analogues capable of undergoing rapid effects of the two compounds would be similar. ring opening of the anhydride moiety were more active than deriv- The complete inhibition of serine/threonine PP at the highest atives that could not undergo such a reaction [40], and a study concentrations of both inhibitors (Fig. 3) indicates that all serine/ finally demonstrated that the anhydride rings of both cantharidin threonine PPs in the plant are inhibited. This and the similar rela- and norcantharidin are hydrolyzed when bound to the catalytic tive phytotoxicity and enzyme inhibition (Figs. 2 and 3) provide domain of the human serine/threonine PP [17]. strong evidence that the herbicide mode of action of endothall is Cantharadic acid is a closer analog to the di-acid endothall than intimately related to inhibition of serine/threonine PP. Lastly, since cantharidin in that it varies by only two additional methyl groups there are at least 19 serine/threonine PPs in A. thaliana, one would (Fig. 1). However, the presence of these substituents contribute expect multiple forms of the endothall susceptible enzyme to be significantly to the overall activity of these inhibitors [41,42]. This found in all plant species. Thus, evolution of resistance to endothall is consistent with our data showing that cantharidin is a more at the target site is highly unlikely. potent inhibitor of plant serine/threonine PP (Fig. 3). The time-dependent interaction between the inhibitors and serine/threonine PP (Fig. 4A) suggests that they are slow inactivators Acknowledgment of these enzymes. This is in agreement with a previous report showing a marked time-dependent inhibition of serine/threonine We thank Robert Johnson and Susan Watson for their excellent PP activity by okadaic acid [43]. Typically, enzyme activity is line- technical assistance. arly correlated with enzyme concentration under conditions in which no other component of the reaction is limiting, as can be observed with the control plot in Fig. 4B. In such enzyme titration References assays, reversible inhibitors cause the lines to resolve toward the origin, whereas compounds that bind irreversibly to enzymes will [1] Herbicide Resistance Action Committee, http://www.hracglobal.com/ Publications/ClassificationofHerbicideModeofAction/tabid/222/Default.aspx, generate parallel lines, as can be observed with some HPPD inhib- Accessed Sept. 3, 2010. itors [44,45]. In our case, the regression curves obtained with [2] S.A. Senseman, Herbicide Handbook, 9th., Weed Science Society of America, endothall and cantharidin (Fig. 4B) support the time-dependent Lawrence, KS, 2007. pp. 458. [3] D. Penner, F.M. Ashton, Influence of dichlobenil, Endothall, and bromoxynil on inactivation of serine/threonine PP. However, the lines were not kinin control of prote?olytic activity, Weed Sci. 16 (1968) 323–326. completely parallel in comparison to the control, which may be [4] J.D. Mann, L.S. Jordan, B.E. Day, A survey of herbicides for their effect upon due to the fact that assays were performed in crude extracts. protein Synthesis, Plant Physiol. 40 (1965) 840–843. [5] A. Rikin, B. Rubin, Increase of cotton cotyledon resistance to the herbicide Indeed plants have a multitude of serine/threonine PP, and some endothall by abscisic acid, Physiol. Plant. 59 (1983) 161–164. that have the potential to cross react with the substrate used in [6] S.O. Duke, W.H. Kenyon, Peroxidizing activity determined by cellular leakage. our assays may not be sensitive to endothall or cantharidin. Target Assays for Modern Herbicides and Related Phytotoxic Compounds, in: P. Böger, G. Sandmann (Eds.), CRC Press, Boca Raton, FL, 1993, pp. 61–66. As no equilibrium can be established and inhibition of serine/ [7] F.E. Dayan, S. B. Watson. Plant cell membrane as a marker for light-dependent threonine PP activity increases over time, the dose–response curves and light-independent herbicide mechanisms of action. Pestic. Biochem. Physiol. (2011) doi:10.1016/j.pestbp.2011.09.004. can only yield I50 app, because this value will be time-dependent. [8] W.L. Epstein, A.M. Kligman, W.L. Epstein, A.M. Kligman, Treatment of warts Nonetheless, the dose–response curves were performed carefully with cantharidin, A. M. A. Arch. Dermatol. 77 (1958) 508–511. to ensure that the incubation time with the inhibitor and the subse- [9] L. Moed, T.A. Shwayder, M.W. Chang, Cantharidin revisited: a blistering quent enzymatic assays lasted precisely the same duration for each defense of an ancient medicine, Arch. Dermatol. 137 (2002) 1357–1360. experiment. Under these conditions, the I app values obtained [10] G.S. Wang, Medical uses of mylabris in ancient China and recent studies, J. 50 Ethnopharmacol. 26 (1989) 147–162. were highly reproducible and were used as a relative value to com- [11] A. Viehoever, I. Cohen, Mechanism of action of aphrodisiac, other irritant pare the potency of endothall and cantharidin. drugs. I. Physiological evaluation of yohimbine, cantharidin capsaicin and The inhibition of serine/threonine PPs by cantharidin at an IC piperine on Daphnia magna, Amer. J. Pharm. 110 (1938) 226–249. 30 [12] D.J. Karras, S.E. Farrell, R.A. Harrigan, F.M. Henretig, L. Gealt, Poisoning from dose based on chlorophyll content caused significant changes in ‘‘Spanish fly’’ (cantharidin), Am. J. Emer. Med. 14 (1996) 478–483. about 10% of the A. thaliana transcriptome [31]. Serine/threonine [13] D. Liu, Z. Chen, The effects of cantharidin, cantharidin derivates on tumour PPs control many signaling processes associated with hormones, cells. Anti-Cancer Agents, Med. Chem. 9 (2009) 392–396. [14] J. Bajsa, S.O. Duke, B.L. Tekwani, Plasmodium falciparum serine/threonine light, and response to pathogens and stress. Proper function of phosphoprotein phosphatases (PPP): from housekeeper to the ‘Holy Grail’, these processes is essential for plant survival. Thus, inhibition of Curr. Drug Targets 9 (2008) 997–1012. 44 J. Bajsa et al. / Pesticide Biochemistry and Physiology 102 (2012) 38–44

[15] X.H. Lin, J. Walter, K. Scheidmann, K. Ohst, J. Newport, G. Walter, Protein [31] J. Bajsa, Z. Pan, S.O. Duke, Transcriptional responses to cantharidin, a phosphatase 2A is required for the initiation of chromosomal DNA replication, protein phosphatase inhibitor, in Arabidopsis thaliana reveal the Proc. Natl. Acad. Sci. USA 95 (1998) 14693–14698. involvement of multiple signal transduction pathways, Physiol. Plant. [16] G.B. Moorhead, V. De Wever, G. Templeton, D. Kerk, Evolution of protein 143 (2011) 188–205. phosphatases in plants, Animals, Biochem. J. 417 (2009) 401–409. [32] J. Bajsa, Z. Pan, S.O. Duke. Serine/threonine protein phosphatases: Multi- [17] I. Bertini, V. Calderone, M. Fragai, C. Luchinat, E. Talluri, Structural basis of purpose enzymes in control of defense mechanisms. Plant Signal. Behav. (in serine/threonine phosphatase inhibition by the archetypal small molecules press). cantharidin and norcantharidin, J. Med. Chem. 52 (2009) 4838–4843. [33] J.D. Hiscox, G.F. Israelstam, A method for the extraction of chlorophyll from [18] Y.M. Li, C. Mackintosh, J.E. Casida, Protein phosphatase 2A and its leaf tissue without maceration, Can. J. Bot. 57 (1979) 1332–1334. [3H]cantharidin/[3H] endothall thioanhydride binding site. Inhibitor [34] S.R. Baerson, A. Sánchez-Moreiras, N. Pedrol-Bonjoch, M. Schulz, I.A. Kagan, specificity of cantharidin and ATP analogues, Biochem. Pharmacol. 46 (1993) A.K. Agarwal, M.J. Reigosa, S.O. Duke, Detoxification and transcriptome 1435–1443. response in Arabidopsis seedlings exposed to the allelochemical [19] F.E. Dayan, S.O. Duke, K. Grossmann, Herbicides as probes in plant biology, benzoxazolin-2(3H)-one, J. Biol. Chem. 280 (2005) 21867–21881. Weed Sci. 58 (2010) 340–350. [35] A. Michel, R.D. Johnson, S.O. Duke, B.E. Scheffler, Dose-response relationships [20] G.K. Agrawal, N.S. Jwa, R. Rakwal, A novel rice (Oryza sativa L.) acidic PR1 gene between herbicides with different modes of action and growth of Lemna highly responsive to cut, phytohormones, and protein phosphatase inhibitors, paucicostata – an improved ecotoxicological method, Environ. Toxicol. Chem. Biochem. Biophys. Res. Commun. 274 (2000) 157–165. 23 (2004) 1074–1079. [21] R. Rakwal, G.K. Agrawal, V.P. Agrawal, Jasmonate, salicylate, protein [36] M.M. Bradford, A rapid and sensitive method for the quantitation of phosphatase 2A inhibitors and kinetin up-regulate OsPR5 expression in cut- microgram quantities of protein utilizing the principle of protein-dye responsive rice (Oryza sativa), J. Plant Physiol. 158 (2001) 1357–1362. binding, Anal. Biochem. 72 (1976) 248–254. [22] R. Rakwal, K. Shii, G.K. Agrawal, M. Yonekura, Protein phosphatase inhibitors [37] R-Development-Core-Team. R: A language and environment for statistical activate defense responses in rice (Oryza sativa) leaves, Physiol. Plant. 111 computing. 2.9.0 edn. R Foundation for Statistical Computing. 2009. Vienna, (2001) 151–157. Austria. [23] R. Rakwal, G. Yang, S. Komatsu, Chitinase induced by jasmonic acid, methyl [38] C. Ritz, J. C. Streibig. Bioassay analysis using R. J. Stat. Softw. 12 (2005) 1–22. jasmonate, ethylene and protein phosphatase inhibitors in rice, Mol. Biol. Rep. [39] M. Matsuzawa, M.J. Graziano, J.E. Casida, Endothal and cantharidin analogues: 31 (2004) 113–119. Relation of structure to herbicidal activity and mammalian toxicity, J. Agric. [24] S. Hartney, J. Carson, L.A. Hadwiger, The use of chemical genomics to detect Food Chem. 35 (1987) 823–829. functional systems affecting the nonhost disease resistance of pea to Fusarium [40] A. McCluskey, M.A. Keane, L.-M. Mudgee, A.T.R. Sim, J. Sakoff, R.J. Quinn, solani f. sp. phaseoli, Plant Sci. 172 (2007) 45–56. Anhydride modified cantharidin analogues. Is ring opening important in the [25] T.-M. Tsai, H.-J. Huang, Effects of iron excess on cell viability and mitogen- inhibition of protein phosphatase 2A?, Eur J. Med. Chem. 35 (2000) 957–964. activated protein kinase activation in rice roots, Physiol. Plant. 127 (2006) [41] Y.-M. Li, C. Mackintosh, J.E. Casida, Protein phosphatase 2A and its 583–592. [3H]cantharidin/[3H]endothall thioanhydride binding site: Inhibitor [26] H.-B. Chang, C.-W. Lin, H.-J. Huang, Zinc-induced cell death in rice (Oryza sativa specificity of cantharidin and ATP analogues, Biochem. Pharmacol. 46 (1993) L.) roots, Plant Growth Regul. 46 (2005) 261–266. 1435–1443. [27] G.C. Vanlerberghe, C.A. Robson, J.Y. Yip, Induction of mitochondrial alternative [42] A. McCluskey, C. Taylor, R.J. Quinn, M. Suganuma, H. Fujiki, Inhibition of oxidase in response to a cell signal pathway down-regulating the cytochrome protein phosphatase 2A by cantharidin analogues, Bioorg. Med. Chem. Lett. 6 pathway prevents programmed cell death, Plant Physiol. 129 (2002) 1829– (1996) 1025–1028. 1842. [43] Y.M. Li, J.E. Casida, Cantharidin-binding protein – Identification as protein [28] X. Vitrac, F. Larronde, S. Krisa, A. Decendit, G. Deffieux, J.M. Merillon, Sugar phosphatase-2A, Proc. Nat. Acad. Sci. USA 89 (1992) 11867–11870. sensing and Ca2+-calmodulin requirement in Vitis vinifera cells producing [44] M.K. Ellis, A.C. Whitfield, L.A. Gowans, T.R. Auton, W.M. Provan, E.A. Lock, L.L. anthocyanins, Phytochemistry 53 (2000) 659–665. Smith, Inhibition of 4-hydroxyphenylpyruvate dioxygenase by 2-(2-nitro-4- [29] F. Ayaydin, E. Vissi, T. Mészáros, P. Miskolczi, I. Kovács, A. Fehér, V. Dombrádi, trifluoromethylbenzoyl)-cyclohexane-1,3-dione and 2-(2-chloro-4- F. Erdödi, P. Gergely, D. Dudits, Inhibition of serine/threonine-specific protein methanesulfonylbenzoyl)-cyclohexane-1,3-dione, Toxicol. Appl. Pharmacol. phosphatases causes premature activation of cdc2MsF kinase at G2/M 133 (1995) 12–19. transition and early mitotic microtubule organization in alfalfa, The Plant J. [45] G. Meazza, B.E. Scheffler, M.R. Tellez, A.M. Rimando, N.P.D. Nanayakkara, I.A. 23 (2000) 85–96. Khan, E.A. Abourashed, J.G. Romagni, S.O. Duke, F.E. Dayan, The inhibitory [30] S. Tresch, J. Schmotz, K. Grossmann, Probing mode of action in plant cell cycle activity of natural products on plant p-hydroxyphenylpyruvate dioxygenase, by the herbicide endothall, a protein phosphatase inhibitor, Pestic. Biochem. Phytochemistry 59 (2002) 281–288. Physiol. 99 (2011) 86–95.