Virus effects on plant quality and vector behavior are species specific and do not depend on host physiological phenotype

Transcript

1 UC Riverside UC Riverside Previously Published Works Title Virus effects on plant quality and vector behavior are species specific and do not depend on host physiological phenotype Permalink https://escholarship.org/uc/item/2z46m64g Journal Journal of Pest Science, 92(2) ISSN 1612-4758 Authors Chesnais, Quentin Mauck, Kerry E Bogaert, Florent et al. Publication Date 2019-03-01 DOI 10.1007/s10340-019-01082-z Peer reviewed eScholarship.org Powered by the California Digital Library University of California

2 Journal of Pest Science https://doi.org/10.1007/s10340-019-01082-z ORIGINAL PAPER Virus effects on plant quality and vector behavior are species specific and do not depend on host physiological phenotype 1,2 1 1 2 1 · Florent Bogaert · Manuella Catterou · Kerry E. Mauck Quentin Chesnais · · Antoine Bamière 1 3 4 1 Fabien Spicher · Mark Tepfer · Arnaud Ameline · Véronique Brault Received: 9 August 2018 / Revised: 11 December 2018 / Accepted: 12 January 2019 © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract There is growing evidence that plant viruses manipulate host plants to increase transmission-conducive behaviors by vectors. Reports of this phenomenon frequently include only highly susceptible, domesticated annual plants as hosts, which constrains our ability to determine whether virus effects are a component of an adaptive strategy on the part of the pathogen or simply by-products of pathology. Here, we tested the hypothesis that transmission-conducive effects of a virus ( Turnip yellows virus [TuYV]) on host palatability and vector behavior ( ) are linked with host plant tolerance and physiological Myzus persicae phenotype. Our study system consisted of a cultivated crop, false flax ( Camelina sativa ) (Brassicales: Brassicaceae), a wild congener ( C. microcarpa ), and a viable F1 hybrid of these two species. We found that the most tolerant host ( C. microcarpa ) exhibited the most transmission-conducive changes in phenotype relative to mock-inoculated healthy plants: Aphids preferred to settle and feed on TuYV-infected C. microcarpa and did not experience fitness changes due to infection—both of which will increase viruliferous aphid numbers. In contrast, TuYV induced transmission-limiting phenotypes in the least tolerant host ( C. sativa ) and to a greater degree in the F1 hybrid, which exhibited intermediate tolerance to infection. Our results provide no evidence that virus effects track with infection tolerance or physiological phenotype. Instead, vector preferences and performance are driven by host-specific changes in carbohydrates under TuYV infection. These results provide evidence that induction of transmission-enhancing phenotypes by plant viruses is not simply a by-product of general pathology, as has been proposed as an explanation for putative instances of parasite manipulation by viruses and many other taxa. · Pathogen transmission · Physiological phenotypes · Plant domestication · Camelina genotypes · Myzus persicae Keywords Vector–host interactions · Vector manipulation Key message Communicated by C. Cutler. Electronic supplementary material The online version of this ) contains article ( https ://doi.org/10.1007/s1034 0-019-01082 -z • Most reports of viruses inducing transmission-conducive supplementary material, which is available to authorized users. vector behavior employ only host plants with very low Arnaud Ameline * tolerance to infection, which constrains our ability to [email protected] determine whether virus effects are part of an adaptive strategy or simply by-products of pathology. 1 UMR CNRS 7058 EDYSAN (Écologie et Dynamique des • On camelina plants with varying degrees of domesti- Systèmes Anthropisés), Université de Picardie Jules Verne, 33 rue St Leu, 80039 Amiens Cedex, France cation, we showed that virus-induced effects on vector 2 behavior do not track with infection severity, but rather Department of Entomology, University of California, Entomology Building, 900 University Ave, Riverside, appear driven by host-specific changes in carbohydrates. CA 92521, USA • Results of our study enrich our understanding of the rel- 3 UMR 1131, SVQV, INRA-UDS, 28, rue de Herrlisheim, evance of manipulative viruses in agricultural systems. 68021 Colmar Cedex, France 4 Institut Jean-Pierre Bourgin (IJPB), INRA, AgroParisTech, CNRS, Université Paris-Saclay, RD10, 78026 Versailles Cedex, France Vol.:(0123456789) 1 3

3 Journal of Pest Science to determine whether virus effects on host phenotypes are Introduction part of an adaptive strategy or simply by-products of infec- tion. For example, nearly all studies reporting putative Transmission of insect-vectored pathogens depends on virus manipulations included only a single host species or vectors engaging in specific sequences of behavior in ). But in nature most viruses cultivar (Mauck et al. 2018 response to host quality and palatability. Chemical, tactile, are capable of infecting multiple host species. Adapta- and visual host cues influence the frequencies and dura- tions that result in manipulation of one host should, at tions of vector probing and feeding behaviors mediating minimum, have neutral effects on the phenotypes of other parasite acquisition, retention, and inoculation (Lefèvre commonly infected hosts so as not to reduce the overall ; Mauck and Thomas 2008 ; Fereres and Moreno 2009 probability of transmission by vectors, but this has rarely 2012 , 2016 ). Given the importance of cue-driven et al. been explored. In addition to a lack of taxonomic diver - vector behavior for pathogen fitness, we may assume that sity in virus–host combinations, there is also a lack of vector-borne pathogens are frequently under selection to diversity in physiological phenotypes. Nearly all studies produce (or at least maintain) host phenotypes and inter - employ domesticated annual plants or laboratory models actions with vectors that are conducive to transmission ), 2018 ; Mauck et al. 2018 ; Eigenbrode et al. 2016 (Mauck ). (Lefèvre and Thomas 2008 ; Heil 2016 ; Mauck et al. 2016 most of which have been artificially selected for faster Consistent with this expectation, there are now over 100 growth and greater yield, and inadvertently for reduced published examples of plant viruses altering host pheno- ). In addition to being better 2015 defenses (Chen et al. types in ways that enhance virus transmission (reviewed in host for insects, domesticated annual plants are considered ; Mauck et al. 2012 , 2016 ; Eigen- 2009 Fereres and Moreno especially permissive for virus infections (Cronin et al. ) as well as theoretical work showing that brode et al. 2018 ). They often support higher titers of viruses relative 2010 virus effects on host–vector interactions can increase the to wild counterparts and display more severe symptoms rate and extent of pathogen spread (McElhany et al. 1995 ; 2015 (Nygren et al. ). Thus, we cannot rule out the pos- 2013 2017 Roosien et al. ; Shaw et al. ; Sisterson 2008 ). sibility that the host physiological phenotype is playing For a few of the most well-studied plant virus pathosys- a role in determining virus effects and that transmission- tems, functional genomics studies implicate specific viral conducive changes are only apparent in hosts that are eas- proteins as inducers of transmission-conducive changes ). 2016 ily exploited by the pathogen (Mauck to host phenotypes, lending support to the hypothesis that To explore the adaptive significance of putative virus phenotypic alterations are the result of virus adaptations manipulations and to understand how manipulative func- 2013 2014 ; Casteel et al. ; Bak et al. (Westwood et al. tions are maintained across agroecological boundaries, ). Comprehensive literature reviews further support 2017 studies are needed that quantify virus effects across taxo- this hypothesis: Virus-induced changes in host phenotypes nomic and physiological host diversity. To address this are not uniform, but exhibit convergence depending on the directly, we explored the effects of Turnip yellows virus specific frequency and duration of intracellular punctures , family Polerovirus (TuYV—genus ) on plant Luteoviridae and/or phloem ingestion required to transmit distinct types phenotypic traits mediating interactions with the primary of plant viruses (Bosque-Pérez and Eigenbrode 2011 ; Myzus persicae ) using a domesticated crop aphid vector ( Mauck et al. 2012 , 2018 ; Eigenbrode et al. 2018 ). Viruses micro- C. ), a wild congener host ( Camelina sativa host ( that are only acquired during sustained vector feeding in ), and an F1 hybrid of the two species. Preliminary carpa the phloem tend to increase host palatability and quality experiments on infected plants (infection status confirmed for vectors, which results in increased settling and uptake by DAS-ELISA) showed strong symptom expression (dis- of virions, while viruses that are only acquired during C. sativa , weak on C. microcarpa , and colorations) on short bouts of cellular content ingestion from non-vascular intermediate on the F1 hybrid plant. Based on these obser - tissues tend to decrease palatability, which enhances dis- vations, we predicted that virus pathogenicity would be persal immediately following virion acquisition (Mauck micro- C. reduced on the less virus-permissive wild host, ). 2018 ; Eigenbrode et al. , 2012 et al. 2016 carpa, relative to the more virus-permissive cultivated Convergence in the phenotypic effects of phylogeneti- sativa C. host, , with the hybrid having an intermediate cally distant viruses transmitted via the same sequences response to infection. We further predicted that patho- of vector behavior, combined with evidence from virus genicity would track with effects on vector behavior. In functional genomics studies, collectively suggests that this system, transmission-conducive effects are expected viruses can evolve to manipulate host phenotypes in ways to include enhanced plant palatability for aphids because , that enhance their own transmission (Mauck et al. 2012 TuYV is only acquired and inoculated during long-term ). However, there are several 2018 ; Eigenbrode et al. 2016 feeding in the phloem, and stimulation of aphid set- issues with the existing literature that constrain our ability tling and feeding will lead to greater virion acquisition 1 3

4 Journal of Pest Science across a genetic gradient that also matched host physiologi- (Bosque-Pérez and Eigenbrode ; Mauck et al. 2012 , 2011 cal phenotypes. ; Chesnais et al. 2016 2019 ). Based on this, we expected to observe the largest improvements in host palatability C. sativa, and quality for vectors due to TuYV infection in Cultivation of plants, insects, and TuYV neutral effects of infection on these traits in the F1 hybrid, and potentially transmission-limiting effects (reductions sativa C. Seeds of Bras- L. Crantz cv Céline (Brassicales: C. in palatability or quality) in microcarpa. To test our ) were provided by the CAVAC (Coopérative sicaceae predictions, we inoculated TuYV into each plant species agricole Vendéenne d’approvisionnement, de ventes de and the F1 hybrid and then measured physiological traits céréales et autres produits agricoles, La Roche-sur-Yon, as metrics of infection severity. We also assessed aphid France). Plants were transformed with a DsRed transgene responses to infected and healthy plants using bioassays 2014 as described (Julié-Galau et al. ) to facilitate rapid iden- that quantified settling and dispersal preferences, and by tification of hybrid plants versus selfed progeny after cross- - measuring aphid biomass and intrinsic rate of popula microcarpa ing with C. . The transformed line F was selfed, tion increase. To determine the mechanisms underlying and DsRed homozygous plants were selected for further aphid behavior and performance, we also quantified sug- C. crossing. Seeds of accession 03CF1063 of microcarpa ars, amino acids, and starch concentrations. These are key Brassicaceae (Andrz.) (Brassicales: ) originating from Guill- nutrients influencing host plant palatability and quality for estre (Hautes-Alpes, France) were provided by the Conserv - aphids and have been previously implicated as targets of atoire Botanique National du Bassin Parisien ( http://cbnbp manipulation by plant viruses (Casteel et al. 2014 ; Mauck C. ). For manual crossing, flowers of .mnhn.fr/cbnbp micro- ). et al. 2014 carpa were emasculated (i.e., anthers were removed) before anthesis to avoid selfing and were pollinated manually with sativa pollen of the homozygous, DsRed-expressing C. line F. Expression of the DsRed transgene in the progeny was Materials and methods confirmed as previously described (Julié-Galau et al. 2014 ). For experiments described below, seeds of each species Study system and (DsRed- C. microcarpa ) and the F1 hybrids C. sativa were sown in plastic pots (90 × 90 × 100 mm) containing TuYV is a globally distributed crop pathogen transmitted commercial sterilized potting soil and grown in a growth in a circulative, non-propagative manner by several aphid ± 5% relative humidity (RH), ± 1 °C, 60 chamber under 20 Brassicaceae species. It infects multiple genera within the , and 16L:8D photoperiod at 2500 lx. including crops, where it causes con- Camelina Brassica and (Sulzer) (Hemiptera: Aphididae) were Myzus persicae spicuous symptoms and considerable yield losses (Jay et al. established from one parthenogenetic female collected in is the most efficient vector of 1999 ). The aphid M. persicae 1999 in a potato field near Loos-en-Gohelle (France). Aphids TuYV and is a natural colonizer of crop and wild hosts in Brassica napus cv. “ were reared on oilseed rape ( ) ” Adriana the Brassicaceae as well as a globally distributed crop pest Brassicaceae ). Each pot (90 × 90 × 100 mm) (Brassicales: Camelina sativa is a re-emergent 2000 (Schliephake et al. ). containing 3–4 rapeseed plants was placed in a ventilated oilseed crop that is increasingly cultivated in western North × 360 mm) and maintained in a × 110 plastic cage (240 America and Europe, where it is being developed for pro- ± 1 °C, 60 growth chamber under 20 ± 5% relative humidity duction of lipids with multiple applications, including feed, (RH), and 16L:8D photoperiod at 2500 lx. green chemistry, and biodiesel (Faure and Tepfer 2016 ). (TuYV, Turnip yellows virus Polero- family, Luteoviridae In previous studies, we found that sativa was a highly C. genus) was provided by Véronique Ziegler-Graff at virus permissive host for TuYV infection and that TuYV induces IBMP-CNRS (Strasbourg, France) and maintained on Mon- physiological changes in this host which are conducive for ). Plants were tia perfoliata (Caryophyllales: Montiaceae persicae (Chesnais et al. 2019 M. transmission by ). Camel- inoculated with TuYV by allowing aphids to feed for 24 h is a wild plant that is endemic to Europe ina microcarpa and then by transferring five perfoliata on TuYV-infected M. and naturalized throughout North America. It is a com - aphids for 3 days on 7-day-old camelina plants. Adults and mon colonizer of field margins along agricultural produc- nymphs were then gently removed with a soft camel-hair tion areas, including those used for cultivation of sativa C. brush. Symptom development consisting of dwarfing, red- (Munoz et al. 2017 ). We recently found that C. sativa readily dening/yellowing of leaf margins, and interveinal discolora- 2013 hybridizes with C. microcarpa (Séguin-Swartz et al. ), tion was recorded 21 days post-inoculation (dpi) and virus and we took advantage of this natural hybridization to study infection was also confirmed using double-antibody sand- the effects of TuYV on plant performance (pathogenicity), wich enzyme-linked immunosorbent assay with polyclonal host chemistry, and transmission-conducive vector behavior TuYV antibodies (LOEWE, Germany) (Adams and Clark 1 3

5 Journal of Pest Science 1977 ). Sham-inoculated ( i.e. , non-infected) plants were Plant‑mediated effects of TuYV on aphid behavior treated similarly using non-viruliferous aphids. All bioassays described below were carried out on plants three weeks post- Plant palatability for herbivorous insects is a multidimen- ± 1 °C, 60 ± 5% inoculation under controlled conditions (20 sional plant trait. The relevant contributing components will RH, and 16L:8D photoperiod at 2500 lx). vary depending on the insect herbivore under study (e.g., phloem feeder vs. leaf chewer), but in general, palatability is a product of physical (e.g., trichome density, leaf tough- Plant susceptibility to infection by TuYV ness, physical spines) and chemical (e.g., primary and sec- ondary metabolites) aspects of tissues (Wardle et al. ; 1998 was persicae M. The transmission efficiency of TuYV by - Elger and Barrat-Segretain ). Plant palatability is meas 2004 ) on the two 1993 tested as described by Fereres et al. ( ured using behavioral assays that quantify the amount of Camelina species and F1 hybrid plants. Aphids (young time an insect spends investigating or feeding on a host. apterous adults) were placed in a Petri dish for a 1-h pre- Here, we used preference tests to assess both the relative acquisition starvation period. Then, for virus acquisition, attractiveness (pre-contact) and palatability (post-contact) M. starved aphids were deposited on an infected perfoliata of infected hosts relative to uninfected hosts of the same plant exhibiting visual symptoms of infection. After a 24-h genotype. The experimental setup used was adapted from acquisition access period, groups of five aphids were trans- Mauck et al. ( 2010 ). In these bioassays, we assessed the ferred to each 7-day-old camelina test plant for a 72-h inocu- propensity of 8-day-old apterous aphids to emigrate from lation access period before being manually removed. We plants over a 24-h period. infected or non-infected Camelina = 15 plants performed two independent experiments with n Ten aphids were released onto leaves (on the three basal per genotype within each experiment. The infec- Camelina leaves) of an infected or sham-inoculated plant (the “release” tion status of the inoculated plants was confirmed 21 days plant) placed adjacent to a second plant (the “choice” plant) post-infection (dpi) by symptom observation and double- which was of the opposite disease status, either infected or antibody sandwich enzyme-linked immunosorbent assay non-infected. The plants in the cage were linked by a bridge with polyclonal antibodies produced by LOEWE (Adams that provided an avenue for free movement between leaves ). and Clark 1977 of each treatment. The whole setup was placed in an aerated 360 × 240 × 110 mm plastic cage where the “release” and “choice” plants were randomly arranged in order to avoid Impact of TuYV on plant performance as a measure position effects. Aphids were then counted on each plant of infection severity 24 h after deposition. Each test was repeated 15 times for each Camelina species and the F1 hybrid. Plant biomass is commonly used as a means of assessing plant pathogen virulence and severity of infection (Sacris- Plant‑mediated effects of TuYV on aphid ), and we used biomass as the primary means tán et al. 2005 performance of quantifying the impact of TuYV on the performance of the three Camelina genotypes. Between 20 and 26 sham- Groups of synchronized first-instar nymphs (less than 24 h inoculated or TuYV-infected plants were harvested 21 days M. old) of were obtained from parthenogenetic persicae after inoculation, and their above-ground fresh biomass was adult females placed on leaves of B. napus (oilseed rape) measured using an electronic scale (Mettler Toledo ML204, set in 1.5% agar in Petri dishes (90 mm diameter). To quan- Max: 220 g, d = 0.1 mg). The plants were then placed in a tify effects of TuYV infection on aphid performance, groups freezer at − 80 °C to be used for metabolite profiling. of five first-instar nymphs were transferred onto sham- A decrease in leaf chlorophyll content is a typical plant C. sativa C. inoculated and infected , , and F1 microcarpa - response to stress imposed by biotic and abiotic attack = 15 per genotype x infection status). Nymphs n hybrids ( ers, including plant viruses (Carter and Knapp 2001 ). This were enclosed in clip cages on leaves at mid-height of each parameter is used extensively as a means of monitoring the plant, and survival was recorded daily until they reached degree of metabolic perturbation experienced by a plant adulthood. The time to reach adulthood, which corresponds 2001 ). To assess the impact under stress (Carter and Knapp , the pre-reproductive i.e. to the time of the first larviposition, of TuYV on overall plant metabolism, we measured the period, was recorded for each individual aphid. chlorophyll content index (CCI) on the third fully expanded Young adults were then randomly selected from the leaf of fifteen sham-inoculated or TuYV-infected plants pool of surviving individuals and transferred onto plants of 21 days after inoculation. CCI was measured with a chloro- and F1 hybrids to study adult per - C. sativa, C. microcarpa phyll content meter (CCM200, Opti-Sciences, Tyngsboro, formance. Each adult aphid was individually placed in a clip Massachusetts, USA) in growth chamber conditions. cage. Adult survival and the number of nymphs produced 1 3

6 Journal of Pest Science assess whether there was an effect of TuYV infection and were recorded daily for a duration equivalent to that of the genotype on plant physiological and biochemical Camelina pre-reproductive period. New nymphs were removed and persicae performance. We included parameters and on M. counted daily with a soft camel-hair brush to estimate the plant infection and genotype as main factors and also studied daily fecundity of each individual parent. For each combi- their interaction. Experimental data on plant physiological , nation of plant genotype ( C. or F1 C. microcarpa sativa and biochemical parameters and aphid weight were analyzed hybrid) per infection status (sham-inoculated or TuYV- using GLM that was based on a normal distribution and infected), 31–35 aphids were used. The daily fecundity and the function “Identity” was specified as the link function in intrinsic rate of natural increase ( ) were calculated using r m the model. Data on aphid performance were not normally the DEMP 1.5.4 software (Giordanengo 2014 ), which uses distributed (count data), accordingly we carried out a GLM a Jackknife resampling technique. The intrinsic rate of natu- ∑ − x r m m l = e 1 using a Poisson distribution, a quasi-likelihood function was ) was calculated as r ral increase ( , where x x m used to correct for over-dispersion, and Log was specified the mean l x the age-specific survival, and m is the age, x x as the link function in the model. The fit of all generalized number of female offspring produced in a unit of time by a linear models was checked by inspecting residuals and QQ female aged x (Birch 1948 ). To measure aphid body mass, plots. When a significant effect of one of the main factors = 30 per genotype and per infection n 8-day-old adult aphids ( was detected or when an interaction between factors was status) were randomly selected from the pool of surviving significant, a pairwise comparison using least-squares means individuals and weighed, one at a time, using a precision (package R: “lsmeans”) ( p value adjustment with Tukey electronic scale (Mettler M3, class 1, Max: 3 g, Low: 1 μg, method) at the 0.05 significance level was used to test for T = −3G [dd] = 1 μg). differences between treatments. Data on aphid retention and U attraction were analyzed using Mann–Whitney test. All Virus effects on host primary metabolites statistical analyses were performed using the R software Preserved plant material was ground to a fine powder in liq- 2016 (version 3.3.2) (R Core Team ). uid nitrogen using a ball mill (MM400, Retsch, Germany), Consistent with previous studies exploring virus effects and powders were kept in liquid nitrogen until metabo- on host traits relevant for host–vector interactions (Mauck lite extraction. Soluble sugars and total amino acids were 2018 et al. ), our explicit focus in statistical analyses was to determine whether, and to what extent, infection status quantified from a water–ethanol extract made from 100 mg relative to uninfected plants of the same geno- 2003 powder according to Harrison et al. ( ). Ethanolic and altered traits type . This is the most ecologically relevant comparison and aqueous fractions were combined, and an aliquot was con- centrated with an evaporator–concentrator and then stored that most often invoked in other studies because it addresses at − the following ecological questions: In a patch of any given 20 °C before analysis. Soluble sugars (glucose, fruc- - - tose, and sucrose) were assayed using the Boehringer enzy host plant, will aphids exhibit preferences for infected ver sus uninfected hosts? And, once a host of a given infection matic bioanalysis kit (R Biopharm, Mannheim, Germany, status is selected, will aphids perform differently on this host Bergmeyer 1974). The determination of the starch content relative to one on of the opposite infection status? Most of was carried out from the pellets resulting from the hydro- alcoholic extraction according to the protocol of Smith and our experiments also permitted an examination of whether Zeeman ( 2006 uninfected plants of the three genotypes differed in trait val- ). Total amino acids were determined from the un-concentrated water–ethanol extracts by the ninhy - ues, and whether relative trait differences were altered by 1957 ) and subtraction of drin colorimetric method (Rosen virus infection (the exception being choice tests, which were the ammonium contents quantified by the phenol hypochlo- pairwise). Comparisons by genotype are also relevant for our rite assay (Berthelot reaction). All metabolite assays were predictions because they confirm that our three host geno- performed in 96-well microtiter plates, with three technical types have inherent trait differences (uninfected condition) replicates per sample, and 14–15 biological replicates per and that responses to infection are not uniform across the treatment. gradient of physiological phenotypes (infected condition). Performing all pairwise comparisons would be relevant if Statistical analyses we were attempting to draw conclusions about outcomes for virus spread in a field scenario where vectors have the option Transmission efficiencies were analyzed by Pearson’s Chi- of choosing among all three hosts of both infection states. squared test. Virus titer (OD absorbance) was analyzed This was not the primary focus of our study, and we did not ), using a Kruskal–Wallis one-way analysis of variance ( H generate predictions about vector preference/performance - followed by a multiple comparison test using the R pack or trait values among all infection status x host genotype age “nparcomp” (type: Tukey). We used a generalized linear - combinations. However, we included an analysis of all pair model (GLM) with a likelihood ratio and Chi-square test to wise comparisons in the supplementary data as a first step 1 3

7 Journal of Pest Science toward exploring the broader implications of physiological phenotype x infection status differences in more complex ecological settings. Results Plant susceptibility to infection by TuYV In both virus susceptibility trials, the number of plants infected with TuYV by viruliferous M. persicae was not species and their significantly different for both Camelina ). Viral load of the hybrid was more similar 1 hybrid (Table to that of the cultivated species ( C. sativa ) and was signifi- C. microcarpa (Table 1 ). These results cantly higher than demonstrate that all three plant genotypes can be infected by TuYV and that persicae is capable of transmitting TuYV M. to both species and their F1 hybrid. Impact of TuYV on plant performance as a measure of infection severity C. sativa was significantly The biomass of TuYV-infected lower than that of the sham-inoculated plants, whereas virus infection had no effect on the biomass of C. microcarpa and Fig. 1 Physiological parameters measured on sham-inoculated hybrid plants (Fig. 1 a) (statistical results in Table S1a, Sup- (= control) or TuYV-infected Camelina plants 21 days after inocu- lation. Box plots show median (line), 25–75% percentiles (box), porting information). The chlorophyll content index (CCI) of 10–90% percentiles (whisker), and outliers (dots). a Above-ground sham-inoculated hybrid plants was significantly higher than chlorophyll content index (CCI). The asterisks indi- biomass and b of TuYV-infected plants for all genotypes, with the most cate a significant difference between TuYV-infected and sham- C. micro- drastic reductions occurring in TuYV-infected inoculated camelina for a plant host species (* P P < 0.01, < 0.05, ** NS < 0.001, P *** - not significant). Letters indicate significant differ 1 microcarpa carpa (Fig. b). The CCI of sham-inoculated C. ences between plant species associated with GLM followed by pair - was also significantly higher than that of sham-inoculated wise comparisons using least-squares means (lowercase letters for but lower than that of the hybrid (Fig. C. sativa, 1 b) (statisti- sham-inoculated plants, capital letters for TuYV-infected plants) cal analysis in Table S1a, Supporting information). microcarpa C. attracted 20% more aphids than the simi- Plant‑mediated effects of TuYV on aphid behavior U lar non-infected “choice” plants (Mann–Whitney tests, a, c). In < 0.001) (Fig. 2 U = 32.5, P = 34, U < 0.001, and P , TuYV-infected “release” For C. sativa and C. microcarpa contrast, for the camelina hybrid, TuYV infection of the M. persicae plants retained 20–25% more than non-infected “release” or the “choice” plants did not significantly influ- U “release” plants of similar genotype (Mann–Whitney > 0.05) tests, U ence aphid movement (Mann–Whitney P < 0.001). In the P = 49.5, U tests, U = 30, P < 0.001, and 2 (Fig. b). sativa and same trend, TuYV-infected “choice” plants of C. on the two Table 1 Transmission efficiency of TuYV by M. persicae Camelina species and their hybrid and virus titer (mean absorb- ± SEM) in the different plants ance C. sativa Significance C. microcarpa Hybrid 2 = 0.111 P = 4.399; 11/15 12/15 χ Repetition 1 15/15 2 P = 2.143; = 0.342 14/15 15/15 Repetition 2 13/15 χ b ab a f = 2; P 3.06 ± 0.13 = 0.008 H = 9.689; d 1.97 ± 0.26 Virus titer 2.36 ± 0.26 C. = 14 for N < 0.05) of virus titers between plant species associated with nparcomp type “Tukey” ( P Letters indicate significant differences ( = 15 for hybrid and sativa N = 13 for C. microcarpa ) , N 1 3

8 Journal of Pest Science “release plant” to an adjacent “choice plant” of the opposite viral Aphid behavioral responses to contact, volatile and visual cues Fig. 2 infectious status. Fifteen replicates were performed for each condi- , non-infected) and TuYV-infected plants after i.e. of sham-inoculated ( P < 0.001, not tion. Asterisks indicate significant differences (*** NS c Hybrid, and . Ten aphids Camelina sativa, a 24 h. microcarpa C. b U significant) associated with Mann–Whitney test were allowed to disperse from leaves of a non-infected or infected r information). A negative effect of TuYV infection on the Plant‑mediated effects of TuYV on aphid m and hybrid plants (− C. of aphids was observed on 7% sativa performance and − microcarpa . When C. 21%, respectively), but not on was sativa r , the C. aphids were reared on TuYV-infected The biomass of aphids reared on sham-inoculated plants var - m 12% higher compared to that of aphids on TuYV-infected ied depending on the plant genotype, with the highest aphid hybrid plants. sativa and the lowest on the hybrid (Fig. biomass on 3 C. a) (statistical results in Table S1b, Supporting information). Virus effects on plant chemical phenotype When reared on infected plants, the biomass of aphids decreased by 25% on C. and by 50% on the hybrid, sativa TuYV infection increased sucrose concentration in both but no change was observed for aphids fed on C. microcarpa and the hybrid leaves but had no effect on the C. sativa 3 a). (Fig. microcarpa (Fig. sucrose content of C. 4 a) (statistical r The intrinsic rate of population increase ( - ) was sig m results in Table S1c, Supporting information). The sucrose nificantly lower when aphids were reared on sham-inoc- C. sativa contents of both TuYV-infected and the hybrid sativa compared to C. or hybrid microcarpa C. ulated were more than ten times higher than of TuYV-infected plants (Fig. 3 b) (statistical results in Table S1b, Supporting 1 3

9 Journal of Pest Science TuYV-infected hybrid plants were significantly higher than and C. microcarpa those of TuYV-infected C. sativa. The total amino acid leaf content of sham-inoculated hybrid plants was two times higher than that of sham- microcarpa (Fig. inoculated C. a) (statistical 5 sativa and C. results in Table S1d, Supporting information). Compared to their respective sham-inoculated plants, TuYV-infected C. sativa and hybrid plants showed significantly higher total amino acid contents (+ 100% and + 25%, respectively), but . We also microcarpa no difference was observed for C. noted that the total amino acid contents of TuYV-infected hybrid plants were significantly higher than those of TuYV- infected and sativa C. For sativa. C. and microcarpa C. the F1 hybrid, the TuYV-infected plants had significantly higher sucrose/amino acid ratios compared to their respec- tive sham-inoculated plants but no difference was observed between sham-inoculated and TuYV-infected C. microcarpa plants. The sucrose/amino acid ratios of TuYV-infected C. sativa and hybrid plants were five to six times higher than C. that of TuYV-infected microcarpa (Fig. 5 b) (statistical results in Table S1d, Supporting information). No difference in the sucrose/amino acid ratio was observed for the sham- inoculated plants of the three species. Fig. 3 reared on three M. persicae Performance parameters of camelina genotypes sham-inoculated or TuYV-infected. Box plots show median (line), 25–75% percentiles (box), 10–90% percentiles a M. (whisker), and outliers (dots). Biomass of eight-day-old persi- Discussion cae aphids. b Intrinsic rate of population increase ( r ) of . persicae M. m The asterisk indicates a significant difference between TuYV-infected and sham-inoculated host plant species, and letters indicate signifi- Nearly all insect-transmitted plant viruses infect multiple cant differences between plant species associated with GLM followed hosts, but most empirical reports of viruses inducing trans- by pairwise comparisons using least-squares means (lowercase letters mission-conducive host phenotypes employ only a single for sham-inoculated plants, capital letters for TuYV-infected plants) host genotype, usually a domesticated annual plant with low NS < 0.001, P < 0. 01, *** P < 0.05, ** P (* not significant) tolerance to infection. To address this shortcoming, we tested the hypothesis that transmission-conducive virus effects on . TuYV infection increased glucose concen- microcarpa C. host phenotype and vector behavior track with host sus- Camelina genotypes compared to their tration in all the three ceptibility and infection tolerance. Thus, we predicted that respective sham-inoculated plants, although the effect was C. (domesticated) would be the least tolerant of TuYV sativa 4 b). We also C. microcarpa (Fig. of lesser magnitude for infection and exhibit the largest, and most transmission-con- sativa noted that the glucose content of TuYV-infected C. ducive, phenotypic changes in response to infection (Fig. 6 ). plants was significantly higher than those of TuYV-infected ) would be the We expected that the wild host ( C. microcarpa C. microcarpa and hybrid. Compared to their respective most tolerant and exhibit transmission-limiting trait changes Camelina sham-inoculated plants, TuYV-infected genotypes - when infected, with the F1 hybrid having intermediate toler showed significantly higher fructose contents (increased by ). Although we 6 ance and trait responses to infection (Fig. C. and by three times for both sativa C. ten times for micro- observed that TuYV infection caused stronger symptoms on C. sativa carpa c). TuYV-infected 4 and hybrid plants) (Fig. C. sativa , compared to the wild microcarpa the cultivated C. exhibited the highest fructose contents, followed by the - we also found that the F1 hybrid was actually the least toler C. hybrid plants, then . Starch content of hybrid microcarpa ance host (having the most severe symptoms) and supported plants was eight times higher than that of sham-inoculated the highest virus titers. These results suggest that there are sativa and six times higher than that of sham-inoculated C. physiological differences among the three plant genotypes in C. d). Compared to their respective sham- 4 microcarpa (Fig. responses to infection which do not track with domestication microcarpa C. inoculated plants, TuYV-infected hybrid and status and that virus accumulation is not strongly correlated plants showed significantly higher starch contents (+ 150% with pathological effects on host performance. We observed 200%, respectively), but no difference was observed and + similar disconnects in virus effects on host traits relevant for C. sativa . We also noted that the starch contents of for vector behavior across the three genotypes. TuYV infection 1 3

10 Journal of Pest Science Nutrient analysis of simple carbohydrates and starch in sham- Fig. 4 ▸ inoculated (= Camelina plants. Box plots control) or TuYV-infected show median (line), 25–75% percentiles (box), 10–90% percentiles (whisker), and outliers (dots). a Sucrose; b Glucose; c Fructose and d Starch. The asterisks indicate a significant difference between TuYV-infected and sham-inoculated camelina for a plant host spe- < 0.01, *** P < 0.05, ** P cies (* P < 0.001, NS not significant). Let- ters indicate significant differences between plant species associated with GLM followed by pairwise comparisons using least-squares means (lowercase letters for sham-inoculated plants, capital letters for TuYV-infected plants) increased palatability and attractiveness for vectors in both while for the F1 hybrid, TuYV C. C. sativa microcarpa, and had no effect on palatability (summarized in Fig. ). Indeed, 6 aphids were equally likely to settle on release plants regard- less of infection status and did not exhibit an emigration preference. Thus, while we do see variation in virus effects across the three genotypes, this variation does not follow our predictions and does not track with virus titer or plant tolerance of infection. Across the three genotypes, we also saw variation in virus effects on plant quality and aphid performance (body mass and intrinsic rate of increase) (summarized in Fig. 6 ). In some cases, TuYV effects on host quality for aphids were in conflict with effects on host palatability. TuYV infec- sativa tion increased palatability of C. but reduced quality. But there are no negative effects on host quality (relative to C. healthy hosts) when TuYV infects the wild host, micro- carpa, even though TuYV infection increased palatability of both and C. sativa. For the F1 hybrid, where C. microcarpa TuYV had no effects on palatability, we observed significant negative effects of TuYV infection on plant quality. In fact, infected F1 hybrids had the poorest quality for aphids of all the host genotype x infection status treatments. Collectively, - these results suggest that the wild C. microcarpa is a bet ter host for the virus to infect than the F1 hybrid, which we predicted might be less tolerant of infection and thus more easily manipulated. In C. microcarpa, virus accumulation was not significantly different from the domesticated con- gener, but plants did not suffer biomass losses as a result of infection. Given that plant size is also important for aphid visitation, the fact that infected C. microcarpa are the same size as healthy could equalize attraction of C. microcarpa aphids to infected hosts based on visual cues (Döring and Chittka 2007 ). Palatability and attractiveness of C. micro - carpa are also increased by virus infection, and although quality is unchanged, it is still roughly equivalent to the qual- C. sativa ity of infected . Similar disconnects between virus effects on palatabil- ity and quality are not common but have been documented in other systems. Notably, Wu et al. ( 2014 ) reported con- Pea enation [BLRV] and family ( Bean leafroll virus [PEMV]) alongside divergent virus effects on gruent aphid behavioral responses to infection of culti- mosaic virus vated peas by either of two members in the host quality. Aphids were attracted to plants infected with Luteoviridae 1 3

11 Journal of Pest Science ) suggest that host physiological phenotype 2013b et al. and associated fluctuations in metabolites may play a role. To explore underlying mechanisms in our system, we pro- filed several metabolites that are known to be key drivers of aphid preference (total free amino acids, sucrose, glucose, ; Singh et al. ). 2011 2010 fructose, and starch) (Hewer et al. Sucrose, glucose, and fructose are phagostimulatory, and ratios of sucrose to free amino acids between 4:1 and 8:1 1963 are particularly preferred by aphids (Auclair ; Abisgold et al. 1994 ). Thus, high levels of sucrose and free amino acids in non-vascular tissues tend to enhance aphid settling ; Hewer if ratios also remain favorable (Pescod et al. 2007 2010 ). However, in contrast to non-vascular tissue, et al. high sucrose and hexose levels in the phloem can be det- rimental to aphid performance because aphids are already under duress to extract diluted free amino acids from high osmolarity sugar-rich solutions (Douglas 2006 ). In our pathosystems, we found that TuYV infection more than doubled the levels of sucrose, glucose, and fructose in C. sativa leaf extract relative to healthy, sham-inoculated sativa C. controls. Free amino acids were also enhanced, and exhibited the most favorable ratio of sucrose to amino acids under TuYV infection (~ 2.5:1). This is consistent with aphid preferences. Although we did not quantify levels of these metabolites in the phloem, sucrose and hexose sugars Nutrient analysis of free amino acids and sugars in sham- Fig. 5 are likely also enhanced in this tissue (Shalitin and Wolf control) or TuYV-infected inoculated (= Camelina plants. Box plots 2000 ; Nadwodnik and Lohaus 2008 ). This could explain show median (line), 25–75% percentiles (box), 10–90% percentiles Total free amino acids in leaf tissue (whisker) and outliers (dots). a the relatively poor performance of aphids on TuYV-infected ratio of sucrose to amino acids. The asterisks indicate a signifi- and b C. because high phloem sucrose concentrations sativa cant difference between TuYV-infected and sham-inoculated camel- can increase osmotic stress and affect aphid performance ina for a plant host species (* < 0.001, P < 0.01, *** P < 0.05, ** P ; Douglas et al. ). In contrast to 1994 2006 (Abisgold et al. not significant). Letters indicate significant differences between NS plant species associated with GLM followed by pairwise compari- TuYV infection only increased C. microcarpa, C. sativa, in sons using least-squares means (lowercase letters for sham-inoculated glucose and fructose, and the relative increase was signifi- plants, capital letters for TuYV-infected plants) . In both cases, cantly lower than that observed in C. sativa these modifications are expected to enhance palatability for - aphids (Mittler et al. 1970 ; Chapman 2003 ), but the rela C. tively subtle changes in , and the absence of microcarpa both viruses and exhibited settling preferences for infected sucrose concentration modification, suggest that the phloem plants over sham-inoculated plants. But PEMV infection sap of this host is most likely not significantly altered in resulted in lower aphid survivorship and no benefits to sugar concentration. - fecundity, while BLRV enhanced almost all aphid perfor Our metabolite analysis also showed that the F1 hybrid mance metrics (Wu et al. 2014 ). Other studies with PEMV exhibits virus-induced changes in sugars and amino acids report variation in effects on host quality with disease pro- that are very similar to those of its domesticated parent 2010 gression. Hodge and Powell ( ) showed that PEMV ) with one notable difference. The infected F1 sativa C. ( - reduced quality of fava bean hosts immediately follow hybrid has starch concentrations that are drastically higher ing inoculation, but enhanced host quality 2 weeks after than all other infected hosts. And healthy F1 hybrids also inoculation. In both studies, it is unclear what mechanisms exhibit significantly elevated starch levels relative to both are driving variation in virus effects on aphid performance healthy parental genotypes. The reasons for extreme starch 2010 ). Virus titer and (Hodge and Powell ; Wu et al. 2014 accumulation in the hybrid are not clear but may be related within-plant defenses or metabolites were not quantified, to the mixing of two genetic backgrounds that differ in but previously documented changes in virus effects over carbon metabolism. For example, we found that healthy the course of disease progression (Werner et al. ; 2009 microcarpa despite having C. sativa is larger than healthy C. ) and across genotypes of crop Rajabaskar et al. 2013a a lower CCI. Healthy plants of the F1 hybrid have a higher plants with different levels of virus tolerance (Rajabaskar 1 3

12 Journal of Pest Science Fig. 6 A conceptual figure showing predictions and observations for TuYV effects on vector settling preference and performance across three Camelina genotypes: a culti- vated crop ( C. sativa ), a wild microcarpa congener ( C. ), and a viable F1 hybrid of these two species. On the left side, the width of each arrow is propor - tional to the predicted/observed probability of aphid settling preference on either infected or sham-inoculated plants. On the right side, the number of aphids above the plants illustrates the predicted/observed aphid performance on each of the sham S infected or sham plants. ( infected plant) I plant, equivalent (in terms of benefits for the virus) to changes mean CCI than both parents, but biomass is equivalent to and lower than C. sativa microcarpa induced in the less tolerant domesticated host ( C. ). C. sativa . These results sug- Far from being intermediate in its responses, the F1 hybrid gest that there are differences in carbon fixation efficiency was the least tolerant of infection and exhibited phenotypic and allocation strategies among the parent species that trans- - late into a novel physiological phenotype in the F1 hybrid changes that were also the least beneficial for virus transmis featuring high starch levels. Under TuYV infection, this sion. Our analysis of primary metabolites provides evidence that aphid behavior and performance are strongly influenced phenotype is exacerbated, possibly due to the influence of by both soluble sugar and starch levels. The nature of virus the virus infection on starch metabolism, which has been observed in many other virus pathosystems (Técsi et al. effects on these metabolites, and aphid preferences, poten- tially depend on genetically controlled variation in host car ). 1997 ; Balachandran et al. 1996 - bon fixation efficiency and allocation strategies. Although Regardless of the mechanisms underlying starch accu- mulation in the F1 hybrid, this phenotype likely explains not quantified here, glucosinolate compounds in Camelina spp. could also be targets for virus manipulation (Westwood why TuYV-induced changes in soluble sugars did not trans- ) as well as other general anti-herbivore defenses 2013 et al. late into increased palatability of this host for aphids and regulated by conserved phytohormone signaling pathways why aphids performed so poorly on infected hybrid plants. ). In future work, ) 1986 2014 2014 (Casteel et al. ; Mauck et al. Starch is deterrent to aphid feeding (Campbell et al. it would be informative to explore whether physiological and elevated starch levels are correlated with reduced aphid phenotype interacts with virus infection to augment defense ). In our system, starch accu- 2011 performance (Singh et al. mulation beyond a threshold level may override other pal- - responses and palatability for brassica specialist and gener atability cues that are modified by virus infection, such as alist TuYV vectors. It would also be beneficial to explore elevated simple soluble sugars resulting in no increase in variation on the virus side of these interactions. Nearly all palatability of infected F1 hybrids relative to healthy con- studies quantifying virus effects on host traits relevant for plant–vector interactions use viruses that were first identified trols and a decrease in quality. because they are pathogens in agriculture. And the majority Overall, our results fail to support our hypothesis that virus effects on host phenotype and vector behavior are of these studies have maintained viruses in culture for many years prior to performing experiments similar to those in our determined by host domestication status and tolerance for virus infection. We found that the host with the highest toler study (Mauck et al. ). While our work demonstrates the 2018 - microcarpa importance of considering host physiological phenotypes, C. ance for infection (wild ) exhibited phenotypic it is equally important to begin incorporating a broader changes in response to TuYV infection that were at least 1 3

13 Journal of Pest Science Bak A, Cheung AL, Yang C et al (2017) A viral protease relocalizes diversity of viruses, including those that are primarily found in the presence of the vector to promote vector performance. in non-crop hosts. Even though our study does employ a Nat Commun 8:14493. https ://doi.org/10.1038/ncomm s1449 3 typical crop-associated virus, our findings provide further Balachandran S, Hurry VM, Kelley SE et al (1997) Concepts of evidence that induction of transmission-enhancing pheno- plant biotic stress. Some insights into the stress physiology of virus-infected plants, from the perspective of photosyn- types by plant viruses is not strongly linked to pathology thesis. Physiol Plant 100:203–213. https ://doi.org/10.103 or host tolerance, as has been proposed as an explanation 4/j.1399-3054.1997.10002 01.x for putative instances of parasite manipulation by viruses Birch LC (1948) The intrinsic rate of natural increase of an insect and many other taxa (Poulin ). Rather, our 2016 ; Heil 2010 population. J Anim Ecol 17:15–26 Bosque-Pérez NA, Eigenbrode SD (2011) The influence of virus- results support the hypothesis that effects may be the prod- induced changes in plants on aphid vectors: insights from uct of viral proteins with specific functions, which are more luteovirus pathosystems. Virus Res 159:201–205. https ://doi. or less effective for induction of transmission-conducive org/10.1016/j.virus res.2011.04.020 changes across diverse host genotypes and species. Campbell BC, Jones KC, Dreyer DL (1986) Discriminative behavioral responses by aphids to various plant matrix https ://doi. polysaccharides. Entomol Exp Appl 41:17–24. org/10.1111/j.1570-7458.1986.tb021 66.x Carter GA, Knapp AK (2001) Leaf optical properties in higher Authors’ contributions plants: linking spectral characteristics to stress and chlorophyll concentration. Am J Bot 88:677–684 Casteel CL, Yang C, Nanduri AC et al (2014) The NIa-Pro protein QC, VB, MT, and AA conceived the ideas and designed Turnip mosaic virus of improves growth and reproduction of elements of the methodology; QC, FB, AB, FS, and MC Myzus persicae (green peach aphid). Plant J the aphid vector, performed the experiments; QC, FB, and AB analyzed the 77:653–663. https ://doi.org/10.1111/tpj.12417 Chapman RF (2003) Contact chemoreception in feeding by phy - data; QC and KEM led the writing of the manuscript. All tophagous insects. Ann Rev Entomol 48:455–484. https ://doi. the authors contributed critically to drafts and gave final org/10.1146/annur ev.ento.48.09180 1.11262 9 approval for publication. Chen YH, Gols R, Benrey B (2015) Crop domestication and its impact on naturally selected trophic interactions. Ann Rev Acknowledgements This work was performed, in partnership with the ://doi.org/10.1146/annur ev-ento-01081 Entomol 60:35–58. https SAS PIVERT, within the frame of the French Institute for the Energy 4-02060 1 Transition (Institut pour la Transition Energétique (ITE) P.I.V.E.R.T. Chesnais Q, Couty A, Uzest M et al (2019) Plant infection by t.com www.insti tut-piver ( ) selected as an Investment for the Future two different viruses induce contrasting changes of vec - (“Investissements d’Avenir”). This work was supported, as part of the https ://doi. tors fitness and behavior. Insect Sci. 26:86–96. Investments for the Future, by the French Government under the refer - org/10.1111/1744-7917.12508 ence ANR-001-01. This work was partially supported by a public grant Cronin JP, Welsh ME, Dekkers MG et al (2010) Host physiologi- overseen by the French National Research Agency (ANR) (Reference: cal phenotype explains pathogen reservoir potential. Ecol Lett VIRAPHIPLANT ANR-14-CE19-0010). Dr. Kerry Mauck is supported https ://doi.org/10.1111/j.1461-0248.2010.01513 13:1221–1232. by startup funds from the University of California, Riverside. .x Döring TF, Chittka L (2007) Visual ecology of aphids: a critical review on the role of colours in host finding. Arthropod-Plant Interact Compliance with ethical standards 1:3–16. https ://doi.org/10.1007/s1182 9-006-9000-1 Douglas AE (2006) Phloem-sap feeding by animals: problems and The authors declare that they have no conflict of Conflict of interest https ://doi.org/10.1093/jxb/ solutions. J Exp Bot 57:747–754. interest. erj06 7 Douglas AE, Price DRG, Minto LB et al (2006) Sweet problems: insect Ethical approval The article does not contain any studies with human traits defining the limits to dietary sugar utilisation by the pea participants or vertebrate animals. . J Exp Biol 209:1395–1403. Acyrthosiphon pisum aphid, https :// doi.org/10.1242/jeb.02148 Eigenbrode SD, Bosque-Pérez NA, Davis TS (2018) Insect-borne plant pathogens and their vectors: ecology, evolution, and com- plex Interactions. Ann Rev Entomol 63:169–191. https ://doi. References org/10.1146/annur ev-ento-02011 7-04311 9 Elger A, Barrat-Segretain MH (2004) Plant palatability can be inferred from a single-date feeding trial. Funct Ecol 18:483–488. https :// Abisgold JD, Simpson SJ, Douglas AE (1994) Nutrient regulation in doi.org/10.1111/j.0269-8463.2004.00846 .x : application of a novel geo- the pea aphid Acyrthosiphon pisum Faure J-D, Tepfer M (2016) Camelina, a swiss knife for plant lipid bio- metric framework to sugar and amino-acid consumption. Physiol technology. OCL 23:D503. https ://doi.org/10.1051/ocl/20160 23 Entomol 19:95–102. https ://doi.org/10.1111/j.1365-3032.1994. Fereres A, Moreno A (2009) Behavioural aspects influencing plant tb010 81.x virus transmission by homopteran insects. Virus Res 141:158– Adams AN, Clark MF (1977) Characteristics of the microplate https ://doi.org/10.1016/j.virus res.2008.10.020 168. method of enzyme-linked immunosorbent assay for the detec- Fereres A, Perez P, Gemeno C, Ponz F (1993) Transmission of spanish tion of plant viruses. J Gen Virol 34:475–483. https ://doi. pepper- and potato-PVY isolates by aphid (Homoptera: aphididae) org/10.1099/0022-1317-34-3-475 vectors: epidemiological implications. Environ Entomol 22:1260– Auclair JL (1963) Aphid feeding and nutrition. Ann Rev Entomol 1265. https ://doi.org/10.1093/ee/22.6.1260 ev.en.08.01016 3.00225 5 8:439–490. https ://doi.org/10.1146/annur 1 3

14 Journal of Pest Science . Planta 227:1079–1089. graveolens https ://doi.org/10.1007/ Giordanengo P (2014) EPG-Calc: a PHP-based script to calculate s0042 5-007-0682-0 electrical penetration graph (EPG) parameters. Arthropod Plant Nygren J, Shad N, Kvarnheden A, Westerbergh A (2015) Variation Interact 8:163–169. https ://doi.org/10.1007/s1182 9-014-9298-z in susceptibility to among wild and domes- Wheat dwarf virus Harrison J, Pou de Crescenzo MA, Sene O, Hirel B (2003) Does https ://doi.org/10.1371/ ticated wheat. PLoS One 10:e0121580. lowering glutamine synthetase activity in nodules modify nitro- journ al.pone.01215 80 ? Plant Physiol Lotus japonicus gen metabolism and growth of Pescod KV, Quick WP, Douglas AE (2007) Aphid responses to https ://doi.org/10.1104/pp.102.01676 6 133:253–262. - plants with genetically manipulated phloem nutrient lev Heil M (2016) Host manipulation by parasites: cases, patterns, and els. Physiol Entomol 32:253–258. https ://doi.org/10.111 https ://doi.org/10.3389/ remaining doubts. Front Ecol Evol 4:80. 1/j.1365-3032.2007.00577 .x fevo.2016.00080 Poulin R (2010) Parasite manipulation of host behavior: an update Hewer A, Will T, van Bel AJE (2010) Plant cues for aphid naviga- and frequently asked questions. Adv Stud Behav 41:151–186. tion in vascular tissues. J Exp Biol 213:4030–4042. https ://doi. https ://doi.org/10.1016/S0065 -3454(10)41005 -0 org/10.1242/jeb.04632 6 R Core Team (2016) R: a language and environment for statistical Hodge S, Powell G (2010) Conditional facilitation of an aphid vector, computing. R Found Stat Comput URL https ://www.R-proje Pea Enation Mosaic Acyrthosiphon pisum , by the plant pathogen, . Accessed 10 Dec 2018 ct.org/ Virus https ://doi.org/10.1673/031.010.14115 . J Insect Sci 10:1–14. Rajabaskar D, Wu Y, Bosque-Pérez NA, Eigenbrode SD (2013a) Beet western yellows Jay CN, Rossall S, Smith HG (1999) Effects of Dynamics of Myzus persicae arrestment by volatiles from Brassica napus on growth and yield of oilseed rape ( virus ). J Potato leafroll virus -infected potato plants during disease Agric Sci Camb 133:131–139 https ://doi. progression. Entomol Exp Appl 148:172–181. Julié-Galau S, Bellec Y, Faure J-D, Tepfer M (2014) Evaluation of the org/10.1111/eea.12087 potential for interspecific hybridization between Camelina sativa Rajabaskar D, Ding H, Wu Y, Eigenbrode SD (2013b) Different and related wild Brassicaceae in anticipation of field trials of GM Potato leafroll reactions of potato varieties to infection by https ://doi.org/10.1007/ camelina. Transgenic Res 23:67–74. Myzus persicae , and associated responses by its vector, virus s1124 8-013-9722-7 https ://doi.org/10.1007/ (Sulzer). J Chem Ecol 39:1027–1035. Lefèvre T, Thomas F (2008) Behind the scene, something else is s1088 6-013-0311-2 pulling the strings: emphasizing parasitic manipulation in vec - Roosien BK, Gomulkiewicz R, Ingwell LL, Bosque-Pérez NA, https ://doi. tor-borne diseases. Infect Genet Evol 8:504–519. Rajabaskar D, Eigenbrode SD (2013) Conditional vector pref- org/10.1016/j.meegi d.2007.05.008 erence aids the spread of plant pathogens: results from a model. Mauck KE (2016) Variation in virus effects on host plant phenotypes Environ Entomol 42:1299–1308. https ://doi.org/10.1111/ - and insect vector behavior: What can it teach us about virus evolu aab.12300 https ://doi.org/10.1016/j.covir tion? Curr Opin Virol 21:114–123. Rosen H (1957) A modified ninhydrin colorimetric analysis for o.2016.09.002 amino acids. Arch Biochem Biophys 67:10–15. https ://doi. Mauck KE, De Moraes CM, Mescher MC (2010) Deceptive chemi- org/10.1016/0003-9861(57)90241 -2 cal signals induced by a plant virus attract insect vectors to Sacristán S, Fraile A, Malpica JM, García-Arenal F (2005) An anal- ://doi. https inferior hosts. Proc Natl Acad Sci 107:3600–3605. ysis of host adaptation and its relationship with virulence in org/10.1073/pnas.09071 91107 Cucumber mosaic virus https :// . Phytopathology 95:827–833. Mauck KE, Bosque-Pérez NA, Eigenbrode SD et al (2012) Transmis- doi.org/10.1094/PHYTO -95-0827 sion mechanisms shape pathogen effects on host-vector interac- Schliephake E, Graichen K, Rabenstein F (2000) Investigations tions: evidence from plant viruses. Funct Ecol 26:1162–1175. on the vector transmission of the Beet mild yellowing virus https ://doi.org/10.1111/j.1365-2435.2012.02026 .x (BMYV) and the (TuYV). J Plant Dis Prot Turnip yellows virus Mauck KE, De Moraes CM, Mescher MC (2014) Biochemical and 107:81–87 physiological mechanisms underlying effects of Cucumber mosaic Séguin-Swartz G, Nettleton JA, Sauder C et al (2013) Hybridiza- on host-plant traits that mediate transmission by aphid vec- virus Camelina sativa (L.) Crantz (false flax) and North tion between tors. Plant Cell Environ 37:1427–1439. https ://doi.org/10.1111/ https :// American Camelina species. Plant Breed 132:390–396. pce.12249 doi.org/10.1111/pbr.12067 Mauck KE, De Moraes CM, Mescher MC (2016) Effects of pathogens Shalitin D, Wolf S (2000) infection affects Cucumber mosaic virus on sensory-mediated interactions between plants and insect vec- sugar transport in melon plants. Plant Physiol 123:597–604. https tors. Curr Opin Plant Biol 32:53–61. https ://doi.org/10.1016/j. ://doi.org/10.1104/pp.123.2.597 pbi.2016.06.012 Shaw AK, Peace A, Power AG, Bosque-Pérez NA (2017) Vector Mauck KE, Chesnais Q, Shapiro LR (2018) Evolutionary determinants population growth and condition-dependent movement drive the of host and vector manipulation by plant viruses. In: Malmstrom spread of plant pathogens. Ecology 98:2145–2157. https ://doi. CM (ed) Environmental virology and virus ecology, 1st edn. Else- org/10.1002/ecy.1907 vier Inc., New York, pp 189–250 Singh V, Louis J, Ayre BG et al (2011) TREHALOSE PHOSPHATE McElhany P, Real LA, Power AG (1995) Vector preference and dis- Arabi- SYNTHASE11-dependent trehalose metabolism promotes ease dynamics: a study of Barley yellow dwarf virus. Ecology Myzus dopsis thaliana defense against the phloem-feeding insect 76:444–457 persicae . Plant J 67:94–104. https ://doi.org/10.1111/j.1365- Mittler TE, Dadd RH, Daniels SC (1970) Utilization of different sugars 313X.2011.04583 .x by the aphid Myzus persicae . J Insect Physiol 16:1873–1890. https Sisterson MS (2008) Effects of insect-vector preference for healthy ://doi.org/10.1016/0022-1910(70)90234 -9 or infected plants on pathogen spread: insights from a model. Munoz F, Fried G, Armengot L, et al (2017) Database of weeds in J Econ Entomol 101:1–8. https ://doi.org/10.1603/0022- cultivation fields of France and UK, with ecological and biogeo- 0493(2008)101%5b1:EOIPF H%5d2.0.CO;2 Zenodo https ://doi.org/10.5281/ graphical information [Database]. Smith AM, Zeeman SC (2006) Quantification of starch in plant tissues. zenod o.11123 42 Nat Protoc 1:1342–1345. https ://doi.org/10.1038/nprot .2006.232 Nadwodnik J, Lohaus G (2008) Subcellular concentrations of sugar Técsi LI, Smith AM, Maule AJ, Leegood RC (1996) A spatial analysis alcohols and sugars in relation to phloem translocation in Plan- of physiological changes associated with infection of cotyledons Apium Prunus persica , Plantago maritima , tago major , and 1 3

15 Journal of Pest Science volatiles emitted during disease progression. Environ Entomol . Plant Physiol Cucumber Mosaic Virus of marrow plants with 38:1429–1438. https ://doi.org/10.1603/022.038.0511 https ://doi.org/10.1104/pp.111.4.975 111:975–985. Westwood JH, Groen SC, Du Z et al (2013) A trio of viral proteins Wardle DA, Barker GM, Bonner KI, Nicholson KS (1998) Can com- Arabidopsis thaliana tunes aphid-plant interactions in . PLoS One parative approaches based on plant ecophysiological traits pre- 8:e83066. https ://doi.org/10.1371/journ al.pone.00830 66 dict the nature of biotic interactions and individual plant species Wu Y, Davis TS, Eigenbrode SD (2014) Aphid behavioral responses https ://doi.org/10.104 effects in ecosystems? J Ecol 86:405–420. to virus-infected plants are similar despite divergent fitness 6/j.1365-2745.1998.00268 .x effects. Entomol Exp Appl 153:246–255. https ://doi.org/10.1111/ Werner BJ, Mowry TM, Bosque-Pérez NA et al (2009) Changes in eea.12246 –induced green peach aphid responses to Potato leafroll virus 1 3

Related documents