Mode of action of fluopyram in plant-parasitic nematodes

Plant-parasitic nematodes (PPN) are responsible for severe yield losses in crop production. Management is challenging as effective and safe means are rare. Recently, it has been discovered that the succinate dehydrogenase (SDH) inhibitor fluopyram is highly effective against PPN while accompanying an excellent safety profile. Here we show that fluopyram is a potent inhibitor of SDH in nematodes but not in mammals, insects and earthworm, explaining the selectivity on molecular level. As a consequence of SDH inhibition, fluopyram impairs ATP generation and causes paralysis in PPN and Caenorhabditis elegans. Interestingly, efficacy differences of fluopyram amongst PPN species can be observed. Permanent exposure to micromolar to nanomolar amounts of fluopyram prevents Meloidogyne spp. and Heterodera schachtii infection and their development at the root. Preincubation of Meloidogyne incognita J2 with fluopyram followed by a recovery period effectively reduces gall formation. However, the same procedure does not inhibit H. schachtii infection and development. Sequence comparison of sites relevant for ligand binding identified amino acid differences in SDHC which likely mediate selectivity, coincidently revealing a unique amino acid difference within SDHC conserved among Heterodera spp. Docking and C. elegans mutant studies suggest that this minute difference mediates altered sensitivity of H. schachtii towards fluopyram.

www.nature.com/scientificreports/ both succininate:ubiquinone oxidoreductase (SQR) and succiniate dehydrogenase (SDH) activity. In literature, fluopyram is referred to as a succinate dehydrogenase inhibitor (SDHI) 12 . Thus, for simplicity the acronym SDH will be used to refer to the mitochondrial complex II.
Here we investigated the impact of fluopyram on the root-knot nematodes M. incognita and M. javanica as well as the cyst nematode H. schachtii in detail. Extreme differences in sensitivity of the two genera towards the compound were observed. In our study we aim at linking target features to the monitored differences in vivo. While fluopyram inhibition data on the molecular target, the complex II of the mitochondrial electron transport chain in fungi and C. elegans have been published 5,12 , detailed studies on the molecular (or biochemical) modeof-action in PPN and in non-target organisms are missing, limiting a detailed understanding of the molecular determinants of selectivity.

Results
Heterodera schachtii is far less sensitive to transient fluopyram exposure than Meloidogyne spp. We first investigated the direct effect of fluopyram on the root-knot nematode M. incognita and the cyst nematode H. schachtii. Exposing the second stage infective juveniles (J2) of M. incognita to different concentrations of fluopyram for 48 h revealed that 0.36 ppm (0.9 µM) fluopyram was sufficient to paralyze 50% of the nematodes. To obtain the same effect on H. schachtii J2, an about 14 times higher concentration of fluopyram (4.9 ppm/12.4 µM) was necessary. To determine whether the fluopyram-induced paralysis is reversible, we subsequently removed the compound and incubated the J2 in water for 6 days. 50% of the M. incognita J2 recovered from a fluopyram pre-treatment with 1.5 ppm (3.8 µM), while more than 90% were irreversibly affected by 5 ppm (12.6 µM) fluopyram. Interestingly, H. schachtii J2 fully recovered from fluopyram pre-treatments up to 20 ppm (50 µM). Despite 100% paralysis at 50 ppm (126 µM), only 21% remained immotile after the recovery phase ( Fig. 1).
To assess the virulence of the fluopyram-treated J2 and validate the obtained recovery data, we inoculated lettuce or A. thaliana with the 48 h incubated and 6 days recovered M. incognita or H. schachtii J2 (Fig. 2  Subsequently, nematodes were washed and incubated in water for further 6 days. The number of immobile nematodes (without (a) or with NaOH stimulus (b)) were counted at both of these time points. Values represented as mean ± SE of three independent biological replicates (n = 6-12 (a); n = 14-18 (b)). Asterisks indicate significant differences to control according to Dunn's method (b) (p < 0.05). Fluopyram is a highly selective inhibitor of the C. elegans SDH. To better understand the molecular determinants of activity differences we assessed the compound's effects on the putative molecular target SDH in more detail. Three known SDH inhibitors were tested to determine species-selectivity and respective efficacy using mitochondrial preparations from Musca domestica (house fly), the free-living nematode C. elegans, the earthworm species Eisenia fetida subsp. andrei, and Rattus sp. (rat) (Fig. 5).   www.nature.com/scientificreports/ As expected, the non-selective complex II inhibitor atpenin A5 showed strong SDH inhibition independent of the species. In contrast, the fungicide flutolanil was not active against the mammal SDH and only active against the SDH of C. elegans, E. fetida and house fly at highest concentrations. Interestingly, fluopyram was extremely potent for the C. elegans SDH (pIC 50 = 8.5) and two orders of magnitude more selective for C. elegans SDH than flutolanil. In addition, fluopyram did not inhibit the complex II of other test species within the applied concentration range.

SDHC amino acid substitution mediates nematode insensitivity towards fluopyram. A C.
elegans resistance screen highlighted that mutations of certain amino acids of the SDHB, SDHC and SDHD protein subunits of complex II mediate nematode insensitivity to wact-11 13 . The wact-11 family core structure and, in particular, wact-11 is closely related to that of fluopyram. Our sequence comparison of the relevant SDHB, SDHC and SDHD regions elucidated an amino acid exchange from G to A at position 90 in the H. schachtii SDHC subunit. Further analyses revealed that this amino acid difference is conserved within Heterodera spp. but could not be observed in any other nematode species analyzed, including Globodera spp. Besides Heterodera spp., B. xylophilus is the only other PPN we analyzed that harbors amino acid substitutions in two of the as relevant described sites (Fig. 6).
To validate that this amino acid difference observed in H. schachtii SDHC mediates increased tolerance towards fluopyram, we tested the susceptibility of a suitable corresponding C. elegans mutant (strain RP2699 with the G77D residue change in SDHC) to fluopyram. The mutant was significantly less sensitive to fluopyram compared to the C. elegans wild type. When incubated with 10 ppm (25 µM) fluopyram for 48 h more than 99% of the wild type nematodes were inactive while only about 19% of the juveniles harboring the mutation were affected (Fig. 7a). Although more than 99% of the C. elegans mutant nematodes were inactive after 48 h in 50 ppm (126 µM), they nearly completely recovered with less than 9% remaining immobile. In contrast, 100% of the C. elegans wild type nematodes were immobile after 48 h incubation in 50 ppm with this effect being nonreversible for about 38% (Fig. 7b). Permanent exposure of C. elegans L1 stage juveniles to 1 ppm (2.5 µM) fluopyram completely inhibited maturation of the wild type nematodes but not of the mutant strain. Compared to the untreated mutant strain control the percentage of mutant juveniles exposed to 1 ppm fluopyram reaching adulthood after 6 days was reduced by 10% only with no statistically significant difference compared to the control (Fig. 7c).
Protein homology modeling of nematode SDH sequences. Homology modeling studies based on the nematode SDHB, C and D sequences suggested the resistance mutation site to be placed about 9 Å apart from the assumed fluopyram binding site, in close proximity to one of the porphyrin rings of the catalytic heme. It is located in the middle of the first transmembrane helix of SDHC, with the sidechain pointing to the heme scaffold (Fig. 8a).
This SDHC helix plays a pivotal role in both substrate and inhibitor binding as well as aligning the heme moiety properly: Trp-sidechain is part of the hydrophobic fluopyram binding site, Ser-OH is involved in H-bonding,   www.nature.com/scientificreports/ and Arg-is required for the correct heme alignment by forming a salt bridge with the carboxylic groups of the porphyrin rings (Fig. 8b). A direct interaction between the fluopyram binding and the mapped resistance site G77D is unlikely owing to the relatively long distance of about 9 Å between the two regions. As pointed out, the putatively resistanceconferring position, in particular its sidechain is in close contact with one of the porphyrin bridges. Compared to the sterically undemanding Glycine introduction of an Alanine (introduction of an additional methyl group) and moreover Aspartate (insertion of a propionic acid with a negative charge) may alter the arrangement of the heme anchoring (Fig. 8b,c). A subsequent realignment of this pivotal helix may then affect fluopyram binding or access to the binding site.

Discussion
Fluopyram has recently been discovered to control PPN and commercial products for agricultural uses are available. Although there are several studies showing activity of fluopyram against nematodes, detailed mode-ofaction studies in PPN are missing. Here we show that fluopyram targets the complex II of the electron transport chain in nematodes and inhibits ATP generation. Fluopyram selectively inhibits SDH function in nematodes but not in non-target organisms such as rat, fly, and the earthworm species Eisenia fetida subsp. andrei. Exposure of nematodes to fluopyram causes reversible or irreversible paralysis in a concentration dependent manner revealing remarkable differences in sensitivity of Meloidogyne spp. and H. schachtii towards the compound. We identified a single amino acid difference within SDHC between Heterodera spp. and Meloidogyne spp. at a position crucial for interaction of fluopyram with its target. This genetic difference likely renders H. schachtii to be less sensitive towards the compound as validated by C. elegans mutant studies.
Exposure of nematodes to a test substance often causes inactivity of the organism and even inability to respond to external stimuli although the nematodes are not dead. To overcome this limitation and avoid misinterpretation of observations we incubated the J2 in compound-free medium or water (recovery phase) after a 2-day fluopyram treatment. Additionally, we investigated to what extent the infectivity of the treated PPN after the recovery phase was impaired. This experimental setup provides highly reliable data for making a statement on the antihelmintic properties of a compound. Secondly, it makes laborious microscopic phenotyping of compound-induced damages a b   8 . Reasons that possibly explain the differences between the reported LC 50 values and our observations and the conclusions are as follows: First, the authors do not specify how they decided whether a J2 is dead or not. We found that sodium hydroxide is a potent stimulating agent to trigger J2 movement which is especially helpful when evaluating the effect of a compound on H. schachtii J2 in in vitro assays. Secondly, the authors did not investigate whether the treated J2 are able to recover and able to infect a host plant. In general, it would be desirable to have a coordinated protocol to precisely determine the nematicidal or nematistatic activity of a compound in order to ascertain comparability of data generated in different institutions. To what extend nematistatic or nematicidal properties of a compound will affect its potential for nematode management in practice remains to be tested but will likely depend on its concentration and half-life in the soil.
In order to approach a mechanistic explanation for the observed differences between the nematode species we first determined the impact of fluopyram on ATP biosynthesis in C. elegans, M. incognita and H. schachtii. As ATP is essential for cellular processes including muscle contraction, fluopyram-induced ATP loss causes paralysis of nematodes. Prolonged ATP depletion to low levels leads to irreversible cell damages and finally to death of the organism. The AEC value is considered as a parameter that categorizes an organism to have normal metabolic activity at a value between 0.7 and 0.95. AEC values below 0.5 are associated with irreversible damage and thus cell death. At AEC values between 0.5 and 0.8 the cell/organism may recover from the stress situation as it is considered to stay viable 14 . We proved that fluopyram interferes with ATP biosynthesis in all three nematodes

Fluopyram [ppm]
wild type mutant Figure 7. SDHC G77D mutation reduces Caenorhabditis elegans sensitivity against fluopyram. Synchronized C. elegans L1 were exposed to different concentrations of fluopyram or DMSO control for 48 h. Subsequently, nematodes were washed and incubated in water for 48 h. The number of immobile nematodes was counted at both of these time points (a,b). The number of nematodes that developed from L1 into adults was determined after permanent exposure to fluopyram or DMSO control for 6 days (c). Bars display mean ± SE of at least six (a,b) or three (c) independent biological replicates (n = 23-36 (a,b); n = 11-12 (c)). Asterisks indicate significant differences between wild type and mutant at each condition according to Dunn's method (p < 0.05). www.nature.com/scientificreports/ nematodes and non-target organisms. Not having a functional assay to test PPN species we analyzed C. elegans as well as rat, fly and the earthworm E. fetida. We demonstrate that micromolar to nanomolar concentrations of fluopyram selectively inhibit complex II function of C. elegans which is not the case for rat, house fly and E. fetida. This is in accordance with Burns et al. 13 who reported that fluopyram impairs complex II activity of the C. elegans wild type (N2) 13 . Additionally, the structurally related wact-11 compound was shown to be active against C. elegans complex II but not against mouse complex II 13 . Moreover, our comparison of selected amino acid sequences of complex II sdh genes revealed that human, mouse and rat are very similar on genetic level, especially revealing amino acid differences compared to nematodes at three positions within SDHC known to be important for interaction of fluopyram with the target (Fig. 6, highlighted in red). Based on these data we are www.nature.com/scientificreports/ able to understand the molecular determinants of selectivity and provide target-based evidence that fluopyram does not possess activity against the SDH of mammals, insects and earthworm but is highly selective for nematodes and fungal pathogens. It is known that several SDH residues are important for binding of SDHIs to its target and are conserved among nematodes 13 . By aligning the respective amino acid sequence regions among several PPN we identified a single amino acid difference within Heterodera spp. at a crucial SDHC residue. We further showed that a C. elegans mutant strain with a point mutation at this position is much less sensitive towards fluopyram treatment compared to the wild type. After 48 h incubation in 10 ppm fluopyram nearly 100% of the wild type nematodes were immobile. In contrast, about 5 times less mutant nematodes were affected. Transient incubation in 50 ppm fluopyram immobilized all/nearly all wild type and mutant nematodes. However, approx. 90% of the mutant nematodes were able to recover-4.4 times more than for the wild type. Thus, our data indicate that this genetic difference between Heterodera spp. and Meloidogyne spp. is at least one major reason for the observed differences in sensitivity towards fluopyram. Further evidence is given by our in silico modeling studies showing that the amino acid difference is not located directly at the fluopyram binding pocket but could lead to steric reorganization of SDHC and thus indirectly interfere with ligand binding. Besides this, morphological characteristics like cuticle thickness, permeability and turnover rate of cyst and root-knot nematode J2 may play a role as well as discussed earlier to explain differences in sensitivity of these PPN towards fluazaindolizine 15 . Interestingly, we observed that C. elegans unlike M. incognita is able to recover from fluopyram treatment indicating that other so far not identified amino acids in the target might be important for ligand interaction and proper target function. In addition, other factors such as metabolic detoxification mechanisms and/or morphological differences could be involved as well. Another PPN genus that would be interesting to investigate is Bursaphelenchus spp. as we identified two amino acid differences within SDHC of B. xylophilius at positions important for ligand interaction. The fact that single amino acid differences can confer complete insensitivity of an organism towards a substance is established knowledge and thus supports our findings. In particular-utilizing a C. elegans mutant screening approach-it was demonstrated that amino acid differences within complex II proteins of C. elegans are responsible for variations in sensitivity of the nematode towards respective inhibitors 13 . As fluopyram was initially introduced as a fungicide in 2009, additional evidence is provided by studies reporting the emergence of resistant fungal strains of usually sensitive species due to individual amino acid substitutions 16 and investigations attesting that certain mutations within sdh genes confer selection advantages upon fluopyram application 17 .
Our study discloses variations in sensitivity of nematode genera towards fluopyram and provides evidence that this can be explained by a naturally occurring single amino acid difference in Heterodera spp. SDHC compared to other nematodes. Molecular understanding can moreover guide fluopyram use recommendations and the development of nematode management strategies in agricultural practices.

Methods
Fluopyram. Analytical grade fluopyram (Bayer AG, Monheim, Germany) was dissolved in dimethyl sulfoxide (DMSO) to obtain a stock solution and diluted to the appropriate working concentrations with water.

Propagation and collection of nematodes. Caenorhabditis elegans. Caenorhabditis elegans N2
wild type (originally obtained from Prof. Einhard Schierenberg, University of Cologne Germany) and mutant RP2699 (kindly provided by Prof. Peter Roy, Ph.D., University of Toronto, USA) were maintained according to standard methods released by the Caenorhabditis Genetics Center, University of Minnesota, Minneapolis, MN 55,455 USA 18 .
Heterodera schachtii. Heterodera schachtii was multiplied and harvested in vitro on mustard (Sinapsis alba cv. Albatros) roots growing on Knop medium supplemented with 2% sucrose 19 . Cysts were collected in a Baermann funnel 20 and hatching of larvae was stimulated by soaking cysts in 3 mM ZnCl 2 . The J2 were washed five times with water and the number of J2/10 µl was adapted to the respective experiment.
Meloidogyne incognita and M. javanica. Pepper seeds var. Feher (for M. incognita) and tomato seeds var. Rentita (for M. javanica) (Quedlingburger Saatgut GmbH, Aschersleben, Germany) were planted into 1 L plastic pots filled with 1,350 g of sandy loam soil (62.6% sand, 13% clay, 24.5% silt, 2.2% humus, pH 6.8). After emergence pepper or tomato plants were infested with a mixed population of 10,000 fully developed eggs and second stage juveniles (J2) of either M. incognita or M. javanica. Plants were kept for ten weeks in the greenhouse at 25 °C, 60-75% relative humidity, 60-80% field capacity and 14 h illumination under sodium vapor pressure lamps. The infested plants were then harvested and roots were washed-free of soil. Clean roots were incubated in an aeriated water bath to promote hatching of viable J2. Collection of J2 was carried out by filtration through multiple sieves (250 µm, 100 µm, 25 µm).

PPN viability and recovery assay. Heterodera schachtii.
To determine the sensitivity of H. schachtii to fluopyram, the nematodes were incubated for 48 h in a DMSO/water solution containing 0-50 ppm fluopyram. Control treatments received the same amount of DMSO as the 50 ppm variant. About 30 J2 were added to each well of a 96-well plate and incubated with the above described fluopyram concentrations or DMSO control for 48 h with gentle shaking of 30 rpm. For evaluation, 2.5% (v/v) final concentration of 1 M NaOH was added to each well and analyzed with a Leica DM4000 microscope equipped with an Olympus C-5050 digital camera. J2 remaining immobile upon NaOH stimulus were considered to be dead. In a parallel experiment, 10,000 J2 for each concentration were incubated for 48 h on a shaker applying a gentle speed of 30 rpm. After incubation, nematodes were washed in an 11 µm sieve with 500 ml sterile water and transferred to 96-well plates with flat www.nature.com/scientificreports/ bottom. J2 were then incubated for 6 days in water followed by evaluation with NaOH as described to obtain the percentage of recovered J2.
Meloidogyne incognita and M. javanica. J2 recovery assays were performed in vitro at room temperature. Fluopyram in DMSO was diluted with tap water to obtain 20 ml of the final test concentrations in 0.1% DMSO. A minimum of 10,000 J2 per treatment were added to each of the chemical solutions and incubated for 48 h on a plate shaker applying a gentle speed of 30 rpm. 48 h after incubation, J2 were transferred to a Büchner funnel to remove the chemical solution by vacuum filtration through 0.45 µm cellulose acetate filter disks. To ensure a complete removal of fluopyram, a volume of 250 ml tap water was added to each funnel and used to wash treated J2 by another step of vacuum filtration. Clean J2 were collected in fresh tap water. Two aliquots per treatment were transferred to 96-well plates and the number of mobile and immobile J2 was counted under a microscope. The activity of tested substances was calculated by comparing the number of immobilized J2 per treatment with that in the DMSO control. J2 were then incubated for additional 6 days followed by another assessment to record the percentage of recovered J2.
Infection assays. Meloidogyne incognita/M. javanica. The virulence of treated J2 was tested in 6-well plates. Each well was filled with 5 ml dry silica sand, 30 lettuce seeds var. Attractie (SPERLI GmbH, Everswinkel, Germany), 2.5 ml tap water and approximately 300 J2. Plates were kept in the greenhouse at growing conditions described above and irrigation was carried out daily. J2 virulence was evaluated after two weeks by counting the number of root galls per well. Then, nematodes were collected in 1 ml water and subsequently centrifuged at 12,300 rpm, 4 °C for 10 min. While the supernatant was discarded, the pellet was dissolved in 5% trichloracetic acid (TCA) and a 3 mm tungsten carbide ball (Qiagen, Cat. No. 69997) was added to the solution. Probes were inserted into a beat mill (MM300, Retsch, Haan, Germany) and shaken at 28 Hz for 1 min. Then, samples were incubated for 2 min on ice, before the shaking step was repeated (28 Hz, 1 min). Afterwards, the tungsten carbide ball was removed, and each probe was sonified for 30 s on ice. Then, 150 µL HPLC buffer was added, which contained: 5 mM tetrabutylammonium hydrogen sulfate, 110 mM NaH 2 PO 4 , 40 mM Na 2 HPO 4 , 6.6% MeOH (pH was adjusted with NaOH to 6.5). Samples were centrifuged at 13,200 rpm, 4 °C for 10 min, afterwards the supernatant was filtered through a 0.45 µm filter and filled into HPLC vials.
Calculation of the AEC value. After adenylate concentrations were quantitatively determined via an HPLC analysis, the AEC value was calculated with the following formula: Measuring succinate:ubiquinone reductase (complex II) activity. Succinate ubiquinone reductase activity was measured according to the method previously described by Barrientos et al. (2002) 21 .
Mitochondrial proteins were isolated from either housefly flight muscle, rat heart, earthworm or C. elegans by standard procedure using differential centrifugation 22 .
In brief, the assay was monitored at 600 nm using a Tecan M1000 Spectrophotometer (Tecan Group, Männedorf, Switzerland) in 384-well plates (Greiner Bio One, Kremsmünster, Austria) using 1 µg of mitochondrial protein per well in 50 mM KH 2 PO 4 (pH 7.2), 4% DMSO (f.c.), 100 µM 2,6-dichlorophenolindophenol (DCPIP), 20 µM decylubiquinone, 1 µM antimycin and 0.33% f.c. butanedioic acid (succinate substrate). Antimycin was added to prevent electron transfer to complex III of the respiratory chain. Enzyme activity was measured in the AEC = ATP + 0.5 ADP ATP + ADP + AMP Construction and refinement of the nematode SDH models was realized by Maestro's advanced homology modeling tool. From a series of complex II co-crystal structures of the parasitic worm Ascaris suum the PDB ID 4YSY was chosen because of the high chemical similarity of the co-crystallized inhibitor with fluopyram 27,28 . Fluopyram was fitted onto the respective positions of the amide group and the o-trifluoromethyl-phenyl moiety, while the chlorpyridyl was placed in a hydrophobic pocket formed by a nematode-specific tryptophan (sidechain).

Statistics.
Results are displayed as means + /− standard error (SE). At least three biological replicates of each experiment were performed including the number of technical replicates indicated in the results section. Significant differences were determined using one-way analysis of variance (ANOVA) and an appropriate post-hoc test (p < 0.05) (SigmaPlot 12.5, Systat Software, Inc., USA).

Ethics declaration.
For the experiments involving plants and vertebrates, all local, national or international guidelines and legislation were adhered to in the study.

Data availability
All data generated or analyzed during this study are included in this published article and the supplementary material. More details are available from the corresponding authors on reasonable request.