This is the submitted version of the article: Domingo Calap, María Luisa; Moreno, Ana Beatriz; Díaz Pendón, Juan Antonio; [et al.]. «Assessing the impact on virus transmission and insect vector behavior of a viral mixed infection in melon». Phytopathology, Vol. 110, issue 1 (Jan. 2020) p. 174-186. DOI 10.1094/PHYTO-04-19-0126-FI This version is avaible at https://ddd.uab.cat/record/221608 under the terms of the license 1 2 3 4 Assessing the impact on virus transmission and insect vector 5 behavior of a viral mixed infection in melon 6 7 8 Maria Luisa Domingo-Calap1, Ana Beatriz Moreno1, Juan Antonio Díaz Pendón2, 9 Aranzazu Moreno3, Alberto Fereres3 and Juan José López-Moya1* 10 1 11 Centre for Research in Agricultural Genomics CRAG, CSIC-IRTA-UAB-UB, 12 Cerdanyola del Vallès, Barcelona, Spain. 2 13 Institute for Mediterranean and Subtropical Horticulture "La Mayora" (IHSM-UMA- 14 CSIC), Algarrobo-Costa, Málaga, Spain. 3 15 Institute of Agricultural Sciences, ICA, CSIC, Madrid, Spain. 16 17 18 * Corresponding author 19 E-mail: juanjose.lopez@cragenomica.es (JJLM) 20 21 Running head: Virus transmission from mix-infected melon plants 22 Keywords: Mixed viral infection of plants; Homopteran vectors; Cucurbit crops; EPG 23 Word count: 6976 24 1 1 Summary 2 Mixed viral infections in plants are common, and can cause synergistic or 3 antagonistic interactions. Except in complex diseases with severe symptoms, mixed 4 infections frequently remain unnoticed, and their impact on insect vector transmission is 5 largely unknown. In this work, we considered mixed infections of two unrelated viruses 6 commonly found in melon plants, the crinivirus Cucurbit yellow stunting disorder virus 7 (CYSDV) and the potyvirus Watermelon mosaic virus (WMV), and evaluated their 8 vector-transmission by whiteflies and aphids, respectively. Their dynamics of 9 accumulation were analyzed until 60 days post-inoculation (dpi) in mix-infected plants, 10 finding reduced titers of WMV and much higher titers of CYSDV compared to single 11 infections. At 24 dpi, corresponding to the peak of CYSDV accumulation, similar 12 whitefly transmission rates were obtained when comparing either individual or mix- 13 infected plants as CYSDV sources, although its secondary dissemination was slightly 14 biased towards plants previously infected with WMV, regardless of the source plant. 15 However, at a later time point mix-infected plants partially recovered from the initially 16 severe symptoms, and CYSDV transmission became significantly higher. Interestingly, 17 aphid transmission rates both at early and late time points were unaltered when WMV 18 was acquired from mix-infected plants despite its reduced accumulation. This lack of 19 correlation between WMV accumulation and transmission could result from 20 compensatory effects observed in the analysis of the aphid feeding behavior by Electrical 21 Penetration Graphs. Thus, our results showed that mix-infected plants could provide 22 advantages for both viruses, directly favoring CYSDV dissemination while maintaining 23 WMV transmission. 24 2 1 Introduction 2 Transmission to susceptible hosts is a key step in the cycle of pathogens that relies 3 mainly on the activity of insect vectors, like many plant viruses (Whitfield, Falk, and 4 Rotenberg 2015). As in other parasitic relationships, plant pathogenic viruses could affect 5 the host-vector interaction in order to improve their transmissibility, inducing changes in 6 the host morphology or physiology (Culver and Padmanabhan 2007), or manipulating the 7 vector behavior (Lefèvre et al. 2009a; Mauck 2016a; Mauck, Chesnais, and Shapiro 8 2018). Here, we aimed to evaluate how mixed viral infections in plants could affect vector 9 transmission traits. 10 Plant viruses are distributed worldwide having a significant impact on plant survival and 11 production of crops (Alexander et al. 2014). Considering the important ecological roles 12 played by viruses, their relationships with hosts and organisms acting as vectors are being 13 explored (Eigenbrode et al. 2002; Mauck, De Moraes, and Mescher 2010; Mauck et al. 14 2015). Indeed, recent reports documented cases in which plant viruses can manipulate 15 their hosts, and consequently modify the attractiveness for vectors and modulate 16 transmission (Peñaflor et al. 2016; Mauck et al. 2012). Observed effects include 17 modification of host defense responses (Petek et al. 2014; Zhang et al. 2012), changes in 18 the nutritional quality of the plant (Alexander et al. 2014; Su et al. 2015), or alterations 19 of visual traits and production of volatiles conditioning attraction (Ingwell, Eigenbrode, 20 and Bosque-Pérez 2012; Liu et al. 2013). Since transmission to new hosts determines in 21 part the fitness of the pathogen, vector performance may have a direct impact on the 22 ecology and evolution of the viruses they transmit (Gutiérrez et al. 2013; Stafford, 23 Walker, and Ullman 2011). 24 Nevertheless, our understanding of how such interactions function in complex ecological 25 environments like natural and agricultural ecosystems is still limited. In these situations, 26 plants frequently interact with several pathogens simultaneously, and the composition 27 and dynamics of ecological communities can be significantly affected (Mauck, De 28 Moraes, and Mescher 2015; Mauck et al. 2015; Stout, Thaler, and Thomma 2006). In 29 particular, mixed infections of plant viruses are common, and current evidence suggests 30 that mixed infections are the rule and not the exception in nature (Mascia and Gallitelli 31 2016; Roossinck, Martin, and Roumagnac 2015; DaPalma et al. 2010). In addition, 32 several important viral diseases of plants are the outcome of interactions between several 33 viral agents (Syller 2012). 3 1 In nature, viral mixed infections can be generated by the transmission of different viruses 2 by different vectors, or by the same vector carrying the different viruses (Syller 2014). 3 The result of a mixed infection varies according to the type of interaction between the 4 viruses involved, from synergism to antagonism (Mascia and Gallitelli 2016; Syller 5 2012). In short, in a synergistic interaction, at least one of the viruses is benefited by the 6 presence of the other(s). The benefit is often quantifiable as an increase in the viral titer, 7 and consequently sometimes the resulting symptoms are more severe compared to a 8 single infection. On the contrary, during an antagonistic interaction, the presence of more 9 than one virus is detrimental for at least one of them, and this is often reflected in a 10 decrease of the viral titer. The order of infection might influence the outcome in virulence 11 and titers (Saldaña, Elena, and Solé 2003; Chávez-Calvillo et al. 2016). 12 As mentioned, few researchers have explored mixed viral infections with respect to their 13 effect on host-pathogen-vector relationships in agricultural ecosystems (Salvaudon, De 14 Moraes, and Mescher 2013; Syller 2012; Syller and Grupa 2016). In the present work we 15 focused our research in a mixed viral infection to shed some light in the possible 16 ecological outcomes regarding vector transmission. The work was carried out in melon 17 (Cucumis melo, L.), an economically relevant crop with global production close to 30 18 million tons in 2016 (www.fao.org). As in other crops, a significant part of the potential 19 production of melon could be lost due to diseases caused by different pathogens, 20 including viruses (Oerke 2006). Moreover, field surveys have reported high occurrence 21 of multiple viral infections in melon and other cucurbits (Juarez et al. 2013). To bring our 22 research closer to current problems, we worked with two viruses that frequently infect 23 melon: the crinivirus Cucurbit yellow stunting disorder virus (CYSDV), and the 24 potyvirus Watermelon mosaic virus (WMV), transmitted respectively by whiteflies and 25 aphids (Desbiez et al. 2011; Navas-Castillo, López-Moya, and Aranda 2014). Nowadays, 26 CYSDV can be found throughout many tropical and subtropical areas where cucurbits 27 are cultivated (Wintermantel et al. 2017). It has a bipartite single-strand plus-sense RNA 28 genome. As other criniviruses, CYSDV is restricted to the phloem of the susceptible 29 plants, being transmitted by whiteflies, such as Bemisia tabaci Gennadius, in a 30 semipersistent manner (Célix et al. 1996; Navas-Castillo, López-Moya, and Aranda 31 2014). On the other hand, WMV (formerly also known as WMV2) presents a worldwide 32 distribution, predominantly in temperate regions, and has a large host range, being able 33 to infect more than 170 plant species (Desbiez and Lecoq 2004). As other potyviruses, 34 WMV presents a positive-sense, single-stranded RNA genome that is translated into a 4 1 polyprotein subsequently processed by viral-encoded proteinases into functional products 2 (Valli, García, and López-Moya 2015). The infection occurs in most plant tissues, and its 3 transmission is mediated by different aphids, including Myzus persicae Sulzer, in a 4 nonpersistent manner (Díaz-Pendón et al. 2005). 5 The objective of our study was to evaluate mixed infections of CYSDV and WMV in 6 melon, in particular to know if the presence of both viruses could influence their vector 7 transmission. To answer this, we first compared the viral load of each virus either in single 8 or in mix-infected plants, following their accumulation along a period of 60 days, and 9 complemented the quantification measurements with observations of symptoms. Then, 10 transmission experiments were performed under laboratory conditions to evaluate the 11 influence of different plants (single- or mix-infected) as virus sources. Furthermore, the 12 crinivirus transmission was also analyzed under field conditions. Regarding the 13 transmission of WMV, the acquisition from single or mix-infected plants was compared 14 using Electrical Penetration Graph (EPG) to measure the number and duration of the 15 different phases and sub-phases. 16 Our results reveal that mixed infections could affect the transmissibility of these two 17 viruses. This knowledge could provide scientific-based advice to minimize virus damage, 18 thus improving melon production and quality. More generally, our findings may help to 19 recognize the importance and ecological impact of mixed viral infections on vector 20 transmission and virus dissemination. 21 22 Results 23 Mix-infected plants accumulated more CYSDV and less WMV compared to single 24 infections 25 Levels of CYSDV and WMV were compared between single and mix-infected 26 plants during 60 days after virus inoculation. To mimic natural conditions, infected plants 27 were obtained using viruliferous whiteflies for CYSDV and aphids for WMV. Samples 28 were collected every 12 days to quantify viral loads. In all time-points the titer of CYSDV 29 was higher in mix-infected compared to single infected plants, with the major peak of 30 accumulation occurring at 24 dpi during mixed infections (Fig. 1). On the other hand, 31 WMV titers were lower in mix-infected plants in comparison to single infections, for 32 most of the analyzed time points. Statistical comparisons were performed first for paired 33 values, using t-test for those cases where normality and homoscedasticity was confirmed, 5 1 and Mann-Whitney for the rest, finding significance at all time points except at 12 dpi 2 (not shown). The dynamics of the curves of virus accumulations were also analyzed by 3 ANOVA, finding dependency of viral accumulation both for single vs. mixed infections 4 in the two cases (CYSDV and WMV), and of time for CYSDV. 5 6 Mix-infected plants at late time points recovered from the severe symptoms caused 7 by WMV alone 8 Observation of symptoms were performed during 60 days to compare single and 9 mix-infected plants. In our growing conditions plants with individual infections showed 10 the expected symptoms described for each virus: severe mosaic and puckering with 11 distortion of all leaves in WMV infected plants; and mild leaf mottling, interveinal 12 chlorosis and yellowing, mainly visible in older leaves in CYSDV infected plants (Fig 13 2a). Interestingly, symptoms in mix-infected plants were very similar to those of a single 14 infection by WMV during approximately the first half of the experiment (Fig 2b), but 15 during the following thirty days, plants recovered and moderately regained growth and 16 vigor, with a final aspect that was more similar to CYSDV single infected plants (Fig 2b 17 and c). The same pattern of symptoms development was observed and reproduced in all 18 repetitions of the experiments. 19 20 CYSDV transmission rates were similar at 24 dpi when comparing single or mixed 21 infected source plants, and both under controlled or field conditions 22 To determine if the mixed infections had some impact on the dispersal of the 23 viruses by their natural vectors we performed transmission experiments. 24 First, we analyzed the transmission of CYSDV under controlled laboratory conditions, 25 comparing whiteflies that acquired the virus either in single or in mix-infected plants. 26 Plants at 24 dpi were chosen as virus sources due to the highest CYSDV accumulation in 27 mix-infected plants (Fig 1). Insects were allowed to acquire the virus during 48 h on 28 infected plants, and then the viruliferous whiteflies were transferred to uninfected 29 receptor plants for virus inoculation during 48 h. The transmission rates were determined 30 at 18 dpi, testing CYSDV presence by RT-PCR. No significant differences were found 31 (Chi-square) between transmission rates for each individual experiment when the 32 whiteflies acquired the virus in mixed infected compared to single infected plants (Fig 3), 33 although the pooled values reached a significative (P<0.05) higher transmission rate for 6 1 CYSDV, a result totally unexpected if we take into account the much lower viral load of 2 CYSDV alone compared to CYSDV+WMV at the chosen time point for the acquisition. 3 Transmission experiments were also performed in field conditions during two 4 consecutive years (see Experimental Procedures) in order to determine if the transmission 5 rates of CYSDV were influenced either by the source plants (single or mixed infected at 6 24 dpi) and/or by the receptor plants (uninfected or WMV-infected). In these field 7 conditions, the results (Figs 4 and 5, supplemental tables S2 and S3) showed larger 8 differences in transmission rates depending on the year, probably reflecting higher 9 mobility and activity of the vectors caused by diverse environmental clues (most likely 10 temperature), thus the comparisons were done separately for each year. Images of the 11 experimental setup, and the complete set of values are provided as supporting information 12 (Fig S1, Tables S2 and S3). 13 In the first experiment, transmission rates were almost identical when the viral source 14 plants were either single or mix-infected plants, suggesting that the acquisition of 15 CYSDV was not altered by the presence of WMV in mixed infections (Fig 4). Although 16 transmission was slightly higher to WMV-infected than to healthy receptor plants from 17 the two kinds of inoculum sources, the differences were not statistically significant 18 (t=1.06 and 1.62 for single or mixed infected source plants). 19 The experimental design in the second year was modified to allow comparison of 20 inoculation choices to either uninfected or WMV-infected plants alternated in every net- 21 structure. The results (Fig 5, sections a and b) confirmed absence of significant 22 differences (t=1.13 and 1.06 for single or mixed infected sources) between the 23 transmission rates to WMV-infected plants compared to uninfected plants, but the slight 24 preference of viruliferous whiteflies to transmit the crinivirus to plants previously 25 infected with the potyvirus was observed in two of the three repetitions performed, 26 regardless of the source of virus acquisition. In order to quantify this putative preference 27 to pre-infected plants, a model was tested for the dependance of percentage of infection 28 as linear function of the pre-infected state of the receptor plant, which resulted in 14% of 29 the infection rate variability explained by the model, although without statistical 30 significance (p=0.23). Therefore, further data will be required in order to confirm the 31 potential pre-infection effect. 32 33 Reduced WMV levels in mix-infected plants does not result in less transmission of 34 WMV by aphids at 24 dpi 7 1 To test if the mixed infection affected transmission of WMV, a different 2 methodology was adopted, taking into consideration the differences of the non-persistent 3 and semipersistent transmission processes. Apterous aphids were fed on a leaf of a single 4 or mixed infected melon plant to acquire the virus, and subsequently manually transferred 5 to test plants for virus inoculation. After insecticide treatment, the plants were observed 6 until appearence of WMV symptoms. Despite the lower viral titers in the mix-infected 7 plants at 24 dpi (Fig 1, and upper panel of Fig 6), the transmission rates when the insects 8 acquired the virus in a mixed infected plant compared to a single infected one were not 9 significantly different in two out of the three individual experiments, neither in the pooled 10 values (Fig 6). 11 12 Aphid feeding behavior was altered in mix-infected plants 13 The feeding behavior of aphids was monitored using EPG to determine if it was 14 influenced by single or mixed infections in the source plants. We fed wired aphids on i) 15 uninfected control plants; ii) single-infected plants with CYSDV; iii) single-infected 16 plants with WMV; and iv) mixed-infected plants with both CYSDV+WMV. The results 17 indicated that the presence of CYSDV alone did not have any major effect on the behavior 18 of aphid vectors compared to control uninfected plants, with the most relevant parameters 19 showing similar values or minor changes not associated with transmission-relevant 20 activities (Table S4). However, we observed that aphids spent more time without probing 21 on plants infected with WMV compared to double-infected plants. Also, longer duration 22 of the intracellular punctures (detectable as potential drops or Pds) was observed in plants 23 infected with CYSDV+WMV, with differences after grouping the data both by insect and 24 by event (Table 5). The statistical analysis of the duration of each of the sub-phases of 25 the Pds confirmed the existence of significant differences in the case of the sub-phase II- 26 3 (Fig 7 a), which was previously associated with acquisition of non-persistently 27 transmitted viruses (Collar, Avilla, and Fereres 1997). The duration of subphase II-3 of 28 the Pds (Fig 7 b, and complete data set in Table S5), and also the number of pulses (Fig 29 7 c) were significantly higher in plants infected with CYSDV+WMV. 30 31 At a late time point, mix-infected plants used as source produced similar rates of 32 aphid transmission of WMV, but higher rates of whitefly transmission of CYSDV. 33 The observed recovery of plants with mixed infections of WMV + CYSDV at late 34 time points (Fig 2) prompted us to perform additional measurements of vector 8 1 transmission rates. For these experiments, five groups of 10 test plants were considered 2 as biological repetitions for each one of the four cases: whitefly-transmission of CYSDV 3 and aphid-transmission of WMV, using as virus sources for acquisition either single 4 (CYSDV or WMV, respectively) or mixed (CYSDV+WMV) infected plants at 52 dpi. 5 Results are shown in Fig 8. 6 The transmission of CYSDV was much higher (98%) from mix-infected plants compared 7 to single infected source (62%), with significant differences (Chi-square, df=1, P < 8 0.0001), strongly contrasting with the results obtained at 24 dpi (Fig 3). Comparing the 9 source plants at these two time points, the recovered double-infected plants correlated 10 with a better performance as virus source for whitefly transmission. On the other hand, 11 WMV transmission rates were 74 and 67 % respectively from single and mixed infected 12 sources, not showing significant differences (Chi-square test, df=1, P value 0.4429), 13 reaching values in the same range as those found at 24 dpi (Fig 6), despite the clear 14 differences in WMV accumulation at late time points of infection between single and 15 mix-infected plants. 16 17 Discussion 18 Plants naturally infected by more than one virus has been reported often (Juarez 19 et al. 2013; Peñaflor et al. 2016; Roossinck et al. 2010; Tugume, Mukasa, and Valkonen 20 2016; Zinga et al. 2013). However, not many studies addressed the ecological impact of 21 such complex pathosystems (Elena, Fraile, and García-Arenal 2014). The interactions 22 between hosts, pathogens and vectors could result in an equilibrium with significant 23 effects on each of the three players (Mauck 2016b), as expected in any parasitic 24 interaction (Lefèvre et al. 2009b). Antecedents pointing to the importance of these 25 relationships during mixed infections have been described. For instance, potentially 26 relevant ecological repercussions of complex host-pathogens interactions have been 27 found between unrelated pathogens, like bacteria and viruses (Shapiro et al. 2012, 2013). 28 Concerning viruses, the recently reviewed ecological implications of host manipulation 29 (Mauck, Chesnais, and Shapiro 2018) derive mainly from reports related to single 30 infections (Eigenbrode et al. 2002; Ingwell, Eigenbrode, and Bosque-Pérez 2012; Mauck, 31 De Moraes, and Mescher 2010; Stafford, Walker, and Ullman 2011), although a recent 32 work considered double infections caused by two persistently transmitted viruses (Chen 33 et al. 2018). Repercussions on other combinations of viral infections, such as non- 9 1 persistently and semipersistently transmitted viruses (Bragard et al. 2013), are mostly 2 unknown. 3 In the case of cucurbits, available surveys revealed that mixed viral infections are quite 4 common (Gil-Salas et al. 2012; Sufrin-Ringwald and Lapidot 2011; Juarez et al. 2013). 5 We focused our work on melon plants infected by two virus that generate important 6 economic losses: WMV and CYSDV (Abrahamian and Abou-Jawdah 2014; Juarez et al. 7 2013). Mixed viral infections were established experimentally using aphids and 8 whiteflies, and the viral load for each individual virus was determined during 60 dpi. Our 9 results showed that CYSDV accumulated to higher titers in mixed compared to single 10 infected plants at all time-points with a peak at 24 dpi, while the viral load of WMV was 11 consistently lower in mix-infected plants than in single infections. These results 12 confirmed similar kinetics reported during mixed infections involving a potyvirus and 13 other plant viruses, including criniviruses: the most common pattern revealed that in 14 mixed infections the accumulation of the potyvirus tend to remain similar or lower 15 compared to single infections, while accumulation increases for the non-potyvirus partner 16 (Murphy and Bowen 2006; Wang et al. 2009). This usual outcome might be due, at least 17 partially, by attenuation of the host RNA silencing-related antiviral defense by the strong 18 activity of the RNA silencing suppressor helper-component protease HC-Pro present in 19 most potyviruses (Valli et al. 2018). This hypothesis is also indirectly supported by the 20 atypical case of the mixed infection between the potyvirus Sweet potato feathery mottle 21 virus (SPFMV) and the crinivirus Sweet potato chlorotic stunt virus (SPCSV) in sweet 22 potato plants, in which the potyvirus does not generate a synergism in the usual direction, 23 a fact that could be related with the almost complete lack of RSS activity in the SPFMV 24 HCPro (Mingot et al. 2016). In the case of a mixed infection of two potyviruses like 25 WMV and Zucchini yellow mosaic virus (ZYMV) in squash (Cucurbita pepo), authors 26 found that ZYMV replicated at similar rates in single and mixed infections, whereas 27 WMV accumulated to significantly lower levels in the presence of ZYMV, although it 28 was still readily transmitted from mix-infected plants (Salvaudon, De Moraes, and 29 Mescher 2013). 30 The virus-virus interaction during mixed infection is considered essential for the 31 combined disease. Outcomes can be classified as antagonistic or synergistic, with 32 synergisms being caused in many cases by combinations of unrelated viruses (Syller 33 2012). At the same time, the possible effects in the plant caused from mixed infections 34 could range from beneficial, to neutral or detrimental. CYSDV and WMV belong to very 10 1 distinct viral groups, crinivirus and potyvirus, being transmitted respectively by non- 2 related insects, semipersistently by whiteflies, and non-persistently by aphids. A 3 superficial analysis of these differences can lead to the assumption that they might not 4 share the same niches (such as host tissues being infected), and thus might not compete 5 for vector transmission or host resources, which could lead to synergism rather than 6 antagonism (DaPalma et al. 2010). In fact, this is roughly what we observed until around 7 the midpoint of the kinetics of virus accumulation. Unexpectedly, the situation appeared 8 to be somehow reversed after a prolonged double-infection period, revealing a more 9 complex interaction between the two viruses. The severity of symptoms caused by 10 CYSDV infection have been directly correlated to the viral load in the plant (Eid et al. 11 2006; Marco et al. 2003), and thus the higher titers of CYSDV in our mix-infected plants 12 could be one of the factors leading to stronger disease symptoms during the initial stages 13 of the double infection. However, we observed that mix-infected plants began to recover 14 gradually, starting around 36-48 dpi, and by 60 dpi the symptoms were considerably less 15 severe in the mix-infected plants compared to the ones infected only with WMV. 16 Symptom recovery has been described in some viral infections as related to RNA 17 silencing mechanisms (Lim et al. 2011), but further studies will be needed to determine 18 if this is also the case for CYSDV+WMV. 19 When we analyzed the CYSDV transmission rates to healthy plants in laboratory 20 conditions at 24 dpi, we found no significant differences when comparing single or mix- 21 infected plants as virus sources, despite the higher titers in the second case. In fact, a 22 report showed that accumulation of CYSDV associated with an increase of symptoms 23 might negatively influence transmission, at least in non-cucurbit hosts (Wintermantel et 24 al. 2016). However, the situation was totally reversed at 52 dpi, with mix-infected plants 25 becoming a better source for virus transmission. It is not easy to explain these differences 26 in transmission rates at 24 dpi and 52 dpi, especially considering that the viral load was 27 lower in the late time point (Fig 1). We can speculate that the severely deformed leaves 28 of the mix-infected plants at the early time point could cause steric difficulties for the 29 whiteflies to reach the phloem and acquire the virus, while the larger and less deformed 30 leaves of CYSDV single infected plants were more accessible for vectors. This 31 circumstance might disappear later after the recovery of the plants with WMV+CYSDV, 32 leading to a more comparable topology of leaves where the higher virus load would 33 become determinant for transmission success. Such interactions have been described in 34 certain conditions, for instance when the presence of a satellite modifies transmissibility 11 1 of the helper virus (Escriu, Perry, and García-Arenal 2000). However, we cannot rule out 2 other explanations. For instance, in other combinations involving phloem-restricted 3 viruses, co-infection with viruses not restricted to the vasculature can remove phloem 4 limitations and make the virus more accessible for vectors, as in the case of Potato leafrol 5 virus (PLRV) and the potyvirus Potato virus A (PVA) (Savenkov and Valkonen 2001). 6 Additional work will be needed to verify these or other hypothesis. 7 When CYSDV transmission was tested in field conditions, similar rates were obtained 8 when comparing single or mixed-infected plants, suggesting that the acquisition process 9 was not affected, regardless of the presence or absence of WMV in the source plants. 10 However, in these experiments we observed a slightly higher transmission towards WMV 11 pre-infected plants compared to uninfected test plants (Fig 4). This observation suggested 12 that the dissemination of CYSDV might be biased towards plants previously infected by 13 potyviruses. Indeed, a repetition of the field experiments allowing free-choice to the 14 viruliferous whiteflies during the inoculation period confirmed that the virus was again 15 preferentially inoculated to WMV pre-infected plants than to uninfected controls (Fig 5 16 a, b). In other words, the influence of the potyvirus presence in mixed infection could be 17 operating not only in the acquisition and transmission of the unrelated crinivirus at late 18 time points, but also in the inoculation process to favor mixed infections. Such effects of 19 vector attraction are more likely to be expected in semipersistently transmitted viruses, 20 while for non-persistently transmitted viruses such as WMV, effects on acquisition will 21 be more relevant for the transmission outcome. The explanation of this possible 22 preference for WMV pre-infected plants during inoculation of CYSDV by viruliferous 23 whiteflies will require further experimentation, but we can hypothesize that mechanisms 24 like a reduced expression of JA-responsive genes, shown in potyvirus-infected plants 25 (Westwood et al. 2014), could increase attraction to whiteflies. 26 The results of WMV transmission experiments showed mostly unaltered rates both at 27 early and late time points, regardless of the observed changes in viral load with reduced 28 virus accumulation in mix-infected plants. The transmission process was further analyzed 29 by EPG to determine the aphid feeding behavior. This technique has been used in 30 different studies (Garzo et al. 2016; Moreno et al. 2012; Munster et al. 2017), but to our 31 knowledge this is the first case in which plants single or mix-infected with viruses were 32 compared. The behavior of aphids was recorded when feeding on uninfected, CYSDV- 33 infected, WMV-infected, or mixed (CYSDV+WMV) infected plants, and significant 34 differences were found for transmission relevant parameters between WMV-single and 12 1 double-infected plants. Our results clearly showed that aphids, when feeding on mix- 2 infected plants, spent more time in the subphase correlated with acquisition of non- 3 persistent viruses (Collar, Avilla, and Fereres 1997), and also produced more pulses 4 during intracellular stylet punctures (Pds). Interestingly, no differences in number and 5 duration of Pd waveforms per insect were observed between uninfected plants and single 6 infected plants with either WMV or CYSDV alone, suggesting that the observed changes 7 in aphid behavior on mix-infected plants might be indeed attributed to the simultaneous 8 presence of both viruses. For the future, it remains to be determined if these changes 9 respond to differential attractiveness and/or nutritional conditions of plants, or alterations 10 in the defense mechanisms of plants during mixed infections, and how these specific 11 changes are caused by the simultaneous presence of the two viruses and not by each one 12 individually. It is relevant to recall that the transmission mechanisms are different for 13 criniviruses and potyviruses, which can have implications for the manipulative strategies 14 of vector-borne pathogens, as suggested recently (Mauck, Chesnais, and Shapiro 2018). 15 In fact, manipulative strategies by vectors have been reported only for a few cases of 16 mixed infections (Rochow 1972; Froissart, Michalakis, and Blanc 2002; Bourdin and 17 Lecoq 1991; Perry and Francki 1992), but the present work, addressing the 18 transmissibility from co-infections of a potyvirus and a crinivirus, suggests that the 19 possible effects of mixed infections on vector transmission might be more general than 20 previously considered. 21 To summarize, we observed that titers of CYSDV remained higher in doubly infected 22 plants, therefore increasing the chances of being transmitted, while the plants were able 23 also to support more whitefly vectors after recovering from the initial stronger symptoms. 24 Indeed, our experiment at 52 dpi showed a clear increase of the transmission rate for 25 CYSDV when acquired from mix-infected plants. Regarding the other virus partner, 26 transmission experiments indicate that WMV could also benefit from its presence in 27 mixed infections, even if the viral titers were lower compared to single infected plants in 28 most time points, because in this case the manipulation of aphids could somehow 29 compensate the lower titers by extending the feeding subphase related to virus 30 acquisition. In other words, during mixed infections CYSDV might enhance its 31 accumulation in plants less severely affected (and thus, prone to survive longer and in 32 better shape) and consequently its chances of dissemination, and at the same time the less 33 abundant WMV might be maintaining the same transmission rates thanks to alterations 34 in the feeding behavior of its vector. Thus, and from an ecological point of view, mix- 13 1 infected plants would confer adaptive advantages to both viruses for their respective 2 dissemination. The implications of multiple infections for the evolution of pathogens 3 (Tollenaere, Susi, and Laine 2016) are being explored in some virus pathosystems (Ali 4 and Roossinck 2017), and our results suggest that vectors might be key elements in the 5 outcome, along with some effects derived from the frequent presence of more than one 6 pathogen in mixed infections. Although additional work will be needed to elucidate all 7 the mechanisms operating during mixed infections, we believe that our results might be 8 useful to explore how the interactions and their dynamics can conditions the available 9 strategies to control viral diseases in crops. 10 11 Experimental Procedures 12 Plants, insects and viruses 13 Seeds of the virus-susceptible melon cultivar "Piel de sapo" PS T111, provided by 14 Semillas Fitó (Spain) were germinated after 5 min treatment with the fungicide Captan in 15 aqueous solution (3g/l). Plants were grown in chambers at 22-25ºC and 16/8 h light/dark 16 cycles. 17 A clonal population of the aphid Myzus persicae, named MP89 and originally from the 18 ICA-CSIC (Madrid, Spain), was maintained on tobacco plants (Nicotiana tabacum, cv 19 Xanthi). A colony of Bemisia tabaci biotype MEAM1 (formerly biotype B) was provided 20 by Dr. R. Gabarra (IRTA-Cabrils, Barcelona, Spain), and was raised on tomato plants 21 (Solanum lycopersicum, cv Micro-Tom). Plants supporting insects for laboratory 22 experiments were kept in separated growth chambers at 23-25ºC and 16/8 h light/dark 23 cycles. For field experiments with whiteflies, individuals from a colony of B. tabaci 24 biotype MED (former Q) were used (Rodríguez-López et al. 2012). 25 The Spanish isolate of the crinivirus CYSDV, named CYSDV-AlLM, was kindly 26 provided by Dr. M.L. Gómez-Guillamón (IHSM “La Mayora”, Málaga, Spain), and was 27 maintained through periodical whitefly transmission on susceptible host plants. 28 The initial infection of WMV was achieved using an infectious full-length clone kindly 29 provided by Dr. C. Desbiez (INRA–PACA, Avignon, France), following described 30 procedures (Desbiez et al. 2011), and the virus was later propagated through aphid 31 transmission. 32 33 CYSDV and WMV transmission experiments 14 1 Both CYSDV and WMV isolates were transmitted to melon plants using their 2 natural vectors. Experimental conditions for CYSDV transmission were set after testing 3 temperature and the number of viruliferous whiteflies, with acquisition and inoculation 4 access periods of 48h each, choosing a temperature of 25 ºC and a ratio of 20 whiteflies 5 per plant. A controlled-vacuum hand-trap and clamp-cages were used for handling insects 6 during transmission to individual plants. In experiments with a large number of test 7 plants, enough viruliferous whiteflies to reach the chosen rate were released inside insect- 8 proof cages containing the plants to be inoculated. 9 The transmission of WMV by aphids was performed essentially as described before for 10 other potyviruses (Ruiz-Ferrer et al. 2005). Briefly, aphids were fasted for 3h hours and 11 then allowed to acquire and inoculate the virus during periods of 10-30 min in each case, 12 transferring them manually from the source to the receptor plant with the help of a 13 paintbrush, until 10 viruliferous aphids were placed on each individual test plant. Vectors 14 were then killed with Confidor insecticide, and the plants were maintained to determine 15 transmission rates by presence of visual symptoms for WMV, or through molecular 16 diagnostic for CYSDV (see below). 17 The CYSDV field transmission experiments were performed inside net-structures (insect 18 proof walk-in 5 × 5 × 2 m) within a tunnel net house and under natural temperature and 19 light conditions (Fig S1) (Rodríguez-López et al. 2012). Tree plants infected with 20 CYSDV or co-infected with CYSDV+WMV were used as inoculum source, surrounded 21 by receptor plants (n=20), either uninfected or previously infected with WMV, distributed 22 in a circle of about 2 m diameter. About 800 non-viruliferous whiteflies were released in 23 the center of the source plants, leading to a vector density of 40 whiteflies/plant. After 4 24 days the receptor plants were treated with insecticide and transferred to a greenhouse for 25 up to 30 days, using tissue print to determine the presence of CYSDV. Field experiments 26 were performed in two consecutive years, using three independent net structures for each 27 combination of source (CYSDV or CYSDV+WMV) and receptor (uninfected or WMV) 28 plants (no choice, total 12 net structures). In free-choice experiments, 10 uninfected plants 29 and 10 WMV-infected plants were distributed in alternated positions (3 net structures for 30 each type of source plants). 31 32 Sampling, RNA extraction and detection of viruses 33 Sampling of representative tissues was standardized to account for growth 34 differences between uninfected, single, and double virus-infected plants. Two leaf discs 15 1 (1 cm diameter) were taken from the second to the latest fully-expanded leaf at each 2 sampling time point, regardless of size and position within the plant. Samples were 3 immediately frozen in liquid nitrogen and ground in Tissue lyser (QIAGEN). Total RNA 4 was extracted using TRIzol reagent (Invitrogen) according to provider instructions, 5 including an additional ethanol precipitation to improve purity. Quality and concentration 6 were estimated in NanoDropÒ spectrophotometer (ND-8000). 7 The High-Capacity cDNA Reverse Transcription kit (Applied Biosystems ™) was used 8 to generate cDNA from 1µg of total RNA, following the manufacturer protocol. The 9 cDNAs served as templates for amplification of viral sequences by polymerase chain 10 reaction (PCR) using TaKaRa ExTaq® commercial kit. The number of cycles and 11 conditions for amplification (temperature and extension time) were optimized for each 12 combination of primers. Virus specific oligos (Table S1) were used to amplify an 813 nt 13 fragment in the CP region of WMV, and a 563 nt fragment near the 3' end of the genomic 14 RNA2 of CYSDV (Marco et al. 2003). DNA fragments were analyzed by electrophoresis 15 on agarose gels. 16 17 Quantification of viral load by RT-qPCR 18 Virus quantification was performed with cDNAs generated from 1 µg of total 19 RNAs, that were previously treated with DNase (ThermoScientific), by RT-qPCR and 20 comparison to standard curves representing known dilutions of in vitro transcripts. 21 To generate virus transcripts, specific oligos were designed to amplify and clone 22 fragments of approximately 1.5 kb located near the 5' ends of viral genomic RNAs in 23 each case (Table S1). The plasmid pGEMT-Easy (Promega) was used to clone fragments 24 deriving from the 5' ends of RNA 1 and RNA 2 of CYSDV, generating constructs 25 RNA1ORF1CYSDV and RNA2Hsp70hCYSDV with the viral sequences under the 26 control of T7 promoter to allow in vitro transcription with genome orientation (+ sense). 27 For WMV, the complete P1 coding region was cloned into pENTR/D-TOPO, and 28 transferred by Gateway recombination to the vector pDEST 14 (Invitrogen) under the 29 control of T7 promoter in genome orientation (+ sense). Constructs were verified by 30 restriction analysis and sequencing. Transcripts were obtained with linearized plasmids 31 as templates using the MEGAscript T7 kit (Ambion), quantified and used as standards in 32 RT-qPCR. 33 Two pairs of specific oligos were designed (https://www.genscript.com/ssl- 34 bin/app/primer) to amplify fragments ranging from 80 to 150 base pairs (bp) for each 16 1 viral RNA (S1 Table), and those leading to better linear regression with the reference 2 were selected in each case. SYBRGreen (Roche) was used to detect PCR products in a 3 Ligth Cycler 480 (Roche) equipment, using 50 ng of cDNAs as template in the PCR 4 reactions 5 6 Screening of CYSDV infected plants through tissue print and molecular 7 hybridization 8 The presence of CYSDV before the onset of symptoms was determined by 9 tissue printing and hybridization with a virus-specific probe (Más and Pallás 1995). 10 Briefly, the plasmid pLM15/24 (Marco et al. 2003) kindly provided by Dr. M. Aranda 11 (CEBAS-CSIC, Murcia), containing a fragment of 5437 bp from the virus RNA1 under 12 the control of the SP6 promoter, was linearized with the restriction enzyme XbaI, and in 13 vitro transcription was carried out using the DIG NORTHERN kit (Roche), following 14 manufacturer's instructions, to obtain negative sense transcripts labeled with digoxigenin. 15 The transcripts were hydrolyzed to shorter size using a carbonate buffer treatment, 16 ethanol precipitated, and resuspended until use (Marco et al. 2003). 17 Cross sections of petioles were printed by pressing on positively charged Nylon 18 membranes (Roche), and the nucleic acids fixed by crosslinking with 120 mJ/cm2 19 (UVC500 crosslinker, Amersham Biosciences). The membrane was incubated with 20 prehybridization solution (20x SSC, 10% N-Lauryl sarcosine, 50% Formamide, 10% 21 SDS and 2% commercial blocking agent from Roche) at 65ºC for 3h, before adding the 22 denatured probe (heated for 10 minutes at 65 °C) and incubated for 14-16 h at 65°C. 23 Processing included: two 10 min rinses (2x SSC, 0.1% SDS); two 15 min rinses (0.1x 24 SSC, 0.1 % SDS); 5 min in washing solution (0.1 Maleic acid, 0.15M NaCl pH7.6, 0.3% 25 Tween 20); 1h blocking (0.1 Maleic acid, 0.15 M NaCl pH 7.6, 1% commercial blocking 26 agent of Roche); 1 h incubation with anti-DIG antibody (Roche) at the recommended 27 concentration; two 15 min rinses in washing solution. The signal was finally visualized 28 in a ChemiDoc chemiluminescence equipment (Bio-Rad), using CDP-Star substrate 29 (Roche) in detection buffer (0.1M TrisHCl pH 9.5, 0.1M NaCl). 30 31 Monitoring of the feeding behavior of aphids by EPG 32 Feeding behavior of M. persicae on melon plants was analyzed by EPG 33 (Tjallingii 1978). Activities up to 1-hour were recorded from individual insects feeding 34 on plants: i) not inoculated, ii) infected with CYSDV, iii) infected with WMV, or iv) 17 1 infected with CYSDV+WMV. Adult apterous aphids were connected to a gold wire (3 2 cm long, 20 µm of diameter) with a drop of silver conductive paint (Pelco® Colloidal 3 Silver no. 16034, Ted Pella Inc., Redding, CA, USA) on the dorsum. The wire was glued 4 to a copper nail and connected to an EPG probe following standard procedures (Carmo- 5 Sousa et al. 2014). Each wired insect was placed on the leaf of a plant, with the circuit 6 closed through the plant substrate. Four plants were tested in parallel using a 4-channel 7 device (Giga-4; EPG Systems, Wageningen, The Netherlands), and 20 to 30 repetitions 8 were recorded. EPG data acquisition was carried out using the Stylet+ computer program 9 (EPG-Systems, Wageningen, The Netherlands), to identify EPG waveforms described for 10 aphids (Tjallingii 1988) corresponding to non-probing (np), intercellular apoplastic stylet 11 pathway (C), and intracellular stylet puncture (Pd). No other EPG waveforms were 12 observed during the 1h recording time. Behavioural variables were processed using a 13 previously described MS Excel Workbook (Sarria et al. 2009). 14 EPG variables (mean ± SE) were calculated (Backus et al. 2007) to determine: proportion 15 of individuals that produced a specific waveform type (PPW); number of wave-form 16 events per insect (NWEI); total waveform duration (min) per insect (WDI); and total 17 waveform duration (min) per event (WDE). Data were analyzed after transformation 18 using ln(x+1). Because of the non-Gaussian distribution of the data, a Mann-Whitney U 19 test was conducted in GraphPad prism 6.0. 20 21 22 Acknowledgments 23 We thank Sonia Sánchez Campos and Gloria Lozano for their contribution during field 24 experiments, and Martí Bernardo, Victor González and members of CRAG "I love data 25 club" for assistance with statistical analysis. We acknowledge comments and suggestions 26 by Santiago Elena and Marilyn Roossinck. MLD and ABM were recipients of FPI 27 contract BES-2014-068970 from Mineco, and MSCA IF fellowship H2020-MSCA-IF- 28 2014 from EU, respectively. Work was financed by Mineco Grants AGL2013-42537-R 29 and AGL2016-75529-R to JJLM and JADP, and AGL2013-47603-C2-2-R to AF. We 30 also acknowledge support to CRAG from "Severo Ochoa programme" SEV-2015-0533, 31 and from Generalitat de Catalunya CERCA Institution. The authors declare no conflicts 32 of interest. 33 18 1 References 2 Abrahamian, P. E., and Abou-Jawdah, Y. 2014. Whitefly-transmitted criniviruses of 3 cucurbits: current status and future prospects. Virusdisease. 25:26–38. 4 Alexander, H. M., Mauck, K. E., Whitfield, A. E., Garrett, K. A., and Malmstrom, C. 5 M. 2014. Plant-virus interactions and the agro-ecological interface. Eur. J. Plant 6 Pathol. 138:529–547. 7 Ali, A., and Roossinck, M. J. 2017. Analysis of quasispecies variation in single and 8 mixed viral infection. Virus Evol. 3:1–7. 9 Backus, E. A., Cline, A. R., Ellerseick, M. R., and Serrano, M. S. 2007. Lygus hesperus 10 (Hemiptera: Miridae) Feeding on Cotton: New Methods and Parameters for Analysis 11 of Nonsequential Electrical Penetration Graph Data. Ann. Entomol. Soc. Am. 12 100:296–310. 13 Bourdin, D., and Lecoq, H. 1991. Evidence that Heteroencapsidation Between Two 14 Potyviruses Is Involved in Aphid Transmission of a Non-Aphid-Transmissible 15 Isolate from Mixed Infections. Phytopathology. 81:1459. 16 Bragard, C., Caciagli, P., Lemaire, O., Lopez-Moya, J. J., MacFarlane, S., Peters, D., et 17 al. 2013. Status and Prospects of Plant Virus Control Through Interference with 18 Vector Transmission. Annu. Rev. Phytopathol. 51:177–201. 19 Carmo-Sousa, M., Moreno, A., Garzo, E., and Fereres, A. 2014. A non-persistently 20 transmitted-virus induces a pull-push strategy in its aphid vector to optimize 21 transmission and spread. Virus Res. 186:38–46. 22 Célix, A., López-Sese, A., Almarza, N., Gómez-Guillamón, M. L., and Rodriguez- 23 Cerezo, E. 1996. Characterization of Cucurbit yellow stunting disorder virus, a 24 Bemisia tabaci-transmitted closterovirus. Phytopathology. 86:1370–1376. 25 Chávez-Calvillo, G., Contreras-Paredes, C. A., Mora-Macias, J., Noa-Carrazana, J. C., 26 Serrano-Rubio, A. A., Dinkova, T. D., et al. 2016. Antagonism or synergism between 27 papaya ringspot virus and papaya mosaic virus in Carica papaya is determined by 28 their order of infection. Virology. 489:179–191. 29 Chen, G., Su, Q., Shi, X., Pan, H., Jiao, X., and Zhang, Y. 2018. Persistently 30 Transmitted Viruses Restrict the Transmission of Other Viruses by Affecting Their 31 Vectors. Front. Physiol. 9:1261. 32 Collar, J. L., Avilla, C., and Fereres, A. 1997. New Correlations between Aphid Stylet 33 Paths and Nonpersistent Virus Transmission. Environ. Entomol. 26:537–544. 34 Culver, J. N., and Padmanabhan, M. S. 2007. Virus-Induced Disease: Altering Host 35 Physiology One Interaction at a Time. Annu. Rev. Phytopathol. 45:221–243. 36 DaPalma, T., Doonan, B. P., Trager, N. M., and Kasman, L. M. 2010. A systematic 37 approach to virus–virus interactions. Virus Res. 149:1–9. 38 Desbiez, C., Joannon, B., Wipf-Scheibel, C., Chandeysson, C., and Lecoq, H. 2011. 39 Recombination in natural populations of watermelon mosaic virus: New agronomic 40 threat or damp squib? J. Gen. Virol. 92:1939–1948. 41 Desbiez, C., and Lecoq, H. 2004. The nucleotide sequence of Watermelon mosaic virus 42 (WMV, Potyvirus) reveals interspecific recombination between two related 43 potyviruses in the 5’ part of the genome. Arch. Virol. 149:1619–1632. 44 Díaz-Pendón, J. A., Fernández-Muñoz, R., Gómez-Guillamón, M. L., and Moriones, E. 45 2005. Inheritance of resistance to Watermelon mosaic virus in Cucumis melo that 46 impairs virus accumulation, symptom expression, and aphid transmission. 47 Phytopathology. 95:840–6. 48 Eid, S., Abou-Jawdah, Y., El-Mohtar, C., Sobh, H., and Havey, M. 2006. Tolerance in 49 cucumber to Cucurbit yellow stunting disorder virus. Plant Dis. 90:645–649. 19 1 Eigenbrode, S. D., Ding, H., Shiel, P., and Berger, P. H. 2002. Volatiles from potato 2 plants infected with potato leafroll virus attract and arrest the virus vector, Myzus 3 persicae (Homoptera: Aphididae). Proc. R. Soc. B Biol. Sci. 269:455–460. 4 Elena, S. F., Fraile, A., and García-Arenal, F. 2014. Evolution and Emergence of Plant 5 Viruses. In Advances in Virus Research, , p. 161–191. 6 Escriu, F., Perry, K. L., and García-Arenal, F. 2000. Transmissibility of Cucumber 7 mosaic virus by Aphis gossypii Correlates with Viral Accumulation and Is Affected 8 by the Presence of Its Satellite RNA. Phytopathology. 90:1068–72. 9 Froissart, R., Michalakis, Y., and Blanc, S. 2002. Helper component- 10 transcomplementation in the vector transmission of plant viruse. Phytopathology. 11 92:576–579. 12 Garzo, E., Moreno, A., Hernando, S., Mariño, V., Torne, M., Santamaria, E., et al. 13 2016. Electrical penetration graph technique as a tool to monitor the early stages of 14 aphid resistance to insecticides. Pest Manag. Sci. 72:707–718. 15 Gil-Salas, F. M., Peters, J., Boonham, N., Cuadrado, I. M., and Janssen, D. 2012. Co- 16 infection with Cucumber vein yellowing virus and Cucurbit yellow stunting disorder 17 virus leading to synergism in cucumber. Plant Pathol. 61:468–478. 18 Gutiérrez, S., Michalakis, Y., Van Munster, M., and Blanc, S. 2013. Plant feeding by 19 insect vectors can affect life cycle, population genetics and evolution of plant viruses 20 ed. Arjen Biere. Funct. Ecol. 27:610–622. 21 Ingwell, L. L., Eigenbrode, S. D., and Bosque-Pérez, N. A. 2012. Plant viruses alter 22 insect behavior to enhance their spread. Sci. Rep. 2:578. 23 Juarez, M., Legua, P., Mengual, C. M., Kassem, M. A., Sempere, R. N., Gómez, P., et 24 al. 2013. Relative incidence, spatial distribution and genetic diversity of cucurbit 25 viruses in eastern Spain. Ann. Appl. Biol. 162:362–370. 26 Lefèvre, T., Lebarbenchon, C., Gauthier-Clerc, M., Missé, D., Poulin, R., and Thomas, 27 F. 2009a. The ecological significance of manipulative parasites. Trends Ecol. Evol. 28 24:41–48. 29 Lefèvre, T., Lebarbenchon, C., Gauthier-Clerc, M., Missé, D., Poulin, R., and Thomas, 30 F. 2009b. The ecological significance of manipulative parasites. Trends Ecol. Evol. 31 24:41–48. 32 Lim, H. S., Jang, C., Bae, H., Kim, J., Lee, C. H., Hong, J. S., et al. 2011. Soybean 33 mosaic virus infection and helper component-protease enhance accumulation of 34 Bean pod mottle virus-specific siRNAs. Plant Pathol. J. 27:315–323. 35 Liu, B., Preisser, E. L., Chu, D., Pan, H., Xie, W., Wang, S., et al. 2013. Multiple forms 36 of vector manipulation by a plant-infecting virus: Bemisia tabaci and Tomato yellow 37 leaf curl virus. J. Virol. 87:4929–4937. 38 Marco, C. F., Aguilar, J. M., Abad, J., Gómez-Guillamón, M. L., and Aranda, M. a. 39 2003. Melon Resistance to Cucurbit yellow stunting disorder virus Is Characterized 40 by Reduced Virus Accumulation. Phytopathology. 93:844–852. 41 Más, P., and Pallás, V. 1995. Non-isotopic tissue-printing hybridization: a new 42 technique to study long-distance plant virus movement. J. Virol. Methods. 52:317– 43 326. 44 Mascia, T., and Gallitelli, D. 2016. Synergies and antagonisms in virus interactions. 45 Plant Sci. 252:176–192. 46 Mauck, K., Bosque-Pérez, N. A., Eigenbrode, S. D., De Moraes, C. M., and Mescher, 47 M. C. 2012. Transmission mechanisms shape pathogen effects on host-vector 48 interactions: evidence from plant viruses ed. Charles Fox. Funct. Ecol. 26:1162– 49 1175. 50 Mauck, K. E. 2016a. Variation in virus effects on host plant phenotypes and insect 20 1 vector behavior: what can it teach us about virus evolution? Curr. Opin. Virol. 2 21:114–123. 3 Mauck, K. E. 2016b. Variation in virus effects on host plant phenotypes and insect 4 vector behavior: what can it teach us about virus evolution? Curr. Opin. Virol. 5 21:114–123. 6 Mauck, K. E., Chesnais, Q., and Shapiro, L. R. 2018. Evolutionary Determinants of 7 Host and Vector Manipulation by Plant Viruses. In Advances in Virus Research, , p. 8 189–250. 9 Mauck, K. E., De Moraes, C. M., and Mescher, M. C. 2010. Deceptive chemical signals 10 induced by a plant virus attract insect vectors to inferior hosts. Proc. Natl. Acad. Sci. 11 U. S. A. 107:3600–5. 12 Mauck, K. E., De Moraes, C. M., and Mescher, M. C. 2015. Infection of host plants by 13 Cucumber mosaic virus increases the susceptibility of Myzus persicae aphids to the 14 parasitoid Aphidius colemani. Sci. Rep. 5:10963. 15 Mauck, K. E., Smyers, E., De Moraes, C. M., and Mescher, M. C. 2015. Virus infection 16 influences host plant interactions with non-vector herbivores and predators ed. 17 Kwang Lee Pum. Funct. Ecol. 29:662–673. 18 Mingot, A., Valli, A., Rodamilans, B., San León, D., Baulcombe, D. C., García, J. A., et 19 al. 2016. The P1N-PISPO trans -Frame Gene of Sweet Potato Feathery Mottle 20 Potyvirus Is Produced during Virus Infection and Functions as an RNA Silencing 21 Suppressor ed. A. Simon. J. Virol. 90:3543–3557. 22 Moreno, A., Tjallingii, W. F., Fernandez-Mata, G., and Fereres, A. 2012. Differences in 23 the mechanism of inoculation between a semi-persistent and a non-persistent aphid- 24 transmitted plant virus. J. Gen. Virol. 93:662–667. 25 Munster, M. Van, Yvon, M., Vile, D., Dader, B., Fereres, A., and Blanc, S. 2017. Water 26 deficit enhances the transmission of plant viruses by insect vectors ed. Mikhail M. 27 Pooggin. PLoS One. 12:e0174398. 28 Murphy, J. F., and Bowen, K. L. 2006. Synergistic Disease in Pepper Caused by the 29 Mixed Infection of Cucumber mosaic virus and Pepper mottle virus. Phytopathology. 30 96:240–7. 31 Navas-Castillo, J., López-Moya, J. J., and Aranda, M. A. 2014. Whitefly-transmitted 32 RNA viruses that affect intensive vegetable production. Ann. Appl. Biol. 165:155– 33 171. 34 Oerke, E.-C. 2006. Crop losses to pests. J. Agric. Sci. 144:31–43. 35 Peñaflor, M. F. G. V., Mauck, K. E., Alves, K. J., De Moraes, C. M., and Mescher, M. 36 C. 2016. Effects of single and mixed infections of Bean pod mottle virus and 37 Soybean mosaic virus on host-plant chemistry and host–vector interactions ed. James 38 R. Bell. Funct. Ecol. 30:1648–1659. 39 Perry, K. L., and Francki, R. I. B. 1992. Insect-mediated transmission of mixed and 40 reassorted cucumovirus genomic RNAs. J. Gen. Virol. 73:2105–2114. 41 Petek, M., Rotter, A., Kogovšek, P., Baebler, S., Mithöfer, A., and Gruden, K. 2014. 42 Potato virus Y infection hinders potato defence response and renders plants more 43 vulnerable to Colorado potato beetle attack. Mol. Ecol. 23:5378–91. 44 Rochow, W. F. 1972. The Role of Mixed Infections in the Transmission of Plant 45 Viruses by Aphids. Annu. Rev. Phytopathol. 10:101–124. 46 Rodríguez-López, M. J., Garzo, E., Bonani, J. P., Fernández-Muñoz, R., Moriones, E., 47 and Fereres, A. 2012. Acylsucrose-producing tomato plants forces Bemisia tabaci to 48 shift its preferred settling and feeding site. ed. Gustavo Bonaventure. PLoS One. 49 7:e33064. 50 Roossinck, M. J., Martin, D. P., and Roumagnac, P. 2015. Plant Virus Metagenomics: 21 1 Advances in Virus Discovery. Phytopathology. 105:716–27. 2 Roossinck, M. J., Saha, P., Wiley, G. B., Quan, J., White, J. D., Lai, H., et al. 2010. 3 Ecogenomics: using massively parallel pyrosequencing to understand virus ecology. 4 Mol. Ecol. 19 Suppl 1:81–8. 5 Ruiz-Ferrer, V., Boskovic, J., Alfonso, C., Rivas, G., Llorca, O., Lopez-Abella, D., et 6 al. 2005. Structural Analysis of Tobacco Etch Potyvirus HC-Pro Oligomers Involved 7 in Aphid Transmission. J. Virol. 79:3758–3765. 8 Saldaña, J., Elena, S. F., and Solé, R. V. 2003. Coinfection and superinfection in RNA 9 virus populations: A selection-mutation model. Math. Biosci. 183:135–160. 10 Salvaudon, L., De Moraes, C. M., and Mescher, M. C. 2013. Outcomes of co-infection 11 by two potyviruses: implications for the evolution of manipulative strategies. Proc. 12 R. Soc. B Biol. Sci. 280:20122959–20122959. 13 Sarria, E., Cid, M., Garzo, E., and Fereres, A. 2009. Excel Workbook for automatic 14 parameter calculation of EPG data. Comput. Electron. Agric. 67:35–42. 15 Savenkov, E. I., and Valkonen, J. P. 2001. Potyviral helper-component proteinase 16 expressed in transgenic plants enhances titers of Potato leaf roll virus but does not 17 alleviate its phloem limitation. Virology. 283:285–93. 18 Shapiro, L., De Moraes, C. M., Stephenson, A. G., and Mescher, M. C. 2012. Pathogen 19 effects on vegetative and floral odours mediate vector attraction and host exposure in 20 a complex pathosystem ed. Wim van der Putten. Ecol. Lett. 15:1430–1438. 21 Shapiro, L. R., Salvaudon, L., Mauck, K. E., Pulido, H., De Moraes, C. M., Stephenson, 22 A. G., et al. 2013. Disease interactions in a shared host plant: effects of pre-existing 23 viral infection on cucurbit plant defense responses and resistance to bacterial wilt 24 disease. PLoS One. 8:e77393. 25 Stafford, C. A., Walker, G. P., and Ullman, D. E. 2011. Infection with a plant virus 26 modifies vector feeding behavior. Proc. Natl. Acad. Sci. 108:9350–9355. 27 Stout, M. J., Thaler, J. S., and Thomma, B. P. H. J. 2006. Plant-mediated interactions 28 between pathogenic microorganisms and herbivorous arthropods. Annu. Rev. 29 Entomol. 51:663–89. 30 Su, Q., Preisser, E. L., Zhou, X. M., Xie, W., Liu, B. M., Wang, S. L., et al. 2015. 31 Manipulation of host quality and defense by a plant virus improves performance of 32 whitefly vectors. J. Econ. Entomol. 108:11–19. 33 Sufrin-Ringwald, T., and Lapidot, M. 2011. Characterization of a synergistic interaction 34 between two cucurbit-infecting begomoviruses: Squash leaf curl virus and 35 Watermelon chlorotic stunt virus. Phytopathology. 101:281–289. 36 Syller, J. 2014. Biological and molecular events associated with simultaneous 37 transmission of plant viruses by invertebrate and fungal vectors. Mol. Plant Pathol. 38 15:417–426. 39 Syller, J. 2012. Facilitative and antagonistic interactions between plant viruses in mixed 40 infections. Mol. Plant Pathol. 13:204–216. 41 Syller, J., and Grupa, A. 2016. Antagonistic within-host interactions between plant 42 viruses: molecular basis and impact on viral and host fitness. Mol. Plant Pathol. 43 17:769–782. 44 Tjallingii, W. F. 1988. Electrical recording of stylet penetration activities. In Aphids: 45 Their Biology. Natural Enemies and Control, vol. 28, eds. A.K. Minks and P.J. 46 Harrewijn. Elsevier Science Publishers, p. 95–108. 47 Tjallingii, W. F. 1978. Electronic recording of penetration behaviour by aphids. 48 Entomol. Exp. Appl. 24:721–730. 49 Tollenaere, C., Susi, H., and Laine, A. L. 2016. Evolutionary and Epidemiological 50 Implications of Multiple Infection in Plants. Trends Plant Sci. 21:80–90. 22 1 Tugume, A. K., Mukasa, S. B., and Valkonen, J. P. T. 2016. Mixed infections of four 2 viruses, the incidence and phylogenetic relationships of Sweet potato chlorotic fleck 3 virus (Betaflexiviridae) isolates in wild species and sweetpotatoes in Uganda and 4 evidence of distinct isolates in East Africa ed. Ulrich Melcher. PLoS One. 5 11:e0167769. 6 Valli, A. A., Gallo, A., Rodamilans, B., López-Moya, J. J., and García, J. A. 2018. The 7 HCPro from the Potyviridae family: an enviable multitasking Helper Component that 8 every virus would like to have. Mol. Plant Pathol. 19:744–763. 9 Valli, A., García, J. a, and López-Moya, J. J. 2015. Potyviridae. eLS. :1–10. 10 Wang, J., Turina, M., Medina, V., and Falk, B. W. 2009. Synergistic interaction 11 between the Potyvirus, Turnip mosaic virus and the Crinivirus, Lettuce infectious 12 yellows virus in plants and protoplasts. Virus Res. 144:163–170. 13 Westwood, J. H., Lewsey, M. G., Murphy, A. M., Tungadi, T., Bates, A., Gilligan, C. 14 A., et al. 2014. Interference with jasmonic acid-regulated gene expression is a 15 general property of viral suppressors of RNA silencing but only partly explains 16 virus-induced changes in plant-aphid interactions. J. Gen. Virol. 95:733–739. 17 Whitfield, A. E., Falk, B. W., and Rotenberg, D. 2015. Insect vector-mediated 18 transmission of plant viruses. Virology. 479–480:278–289. 19 Wintermantel, W. M., Gilbertson, R. L., McCreight, J. D., and Natwick, E. T. 2016. 20 Host-Specific Relationship Between Virus Titer and Whitefly Transmission of 21 Cucurbit yellow stunting disorder virus. Plant Dis. 100:92–98. 22 Wintermantel, W. M., Gilbertson, R. L., Natwick, E. T., and McCreight, J. D. 2017. 23 Emergence and epidemiology of Cucurbit yellow stunting disorder virus in the 24 American Desert Southwest, and development of host plant resistance in melon. 25 Virus Res. 241:213–219. 26 Zhang, T., Luan, J.-B., Qi, J.-F., Huang, C.-J., Li, M., Zhou, X.-P., et al. 2012. 27 Begomovirus-whitefly mutualism is achieved through repression of plant defences 28 by a virus pathogenicity factor. Mol. Ecol. 21:1294–304. 29 Zinga, I., Chiroleu, F., Legg, J., Lefeuvre, P., Komba, E. K., Semballa, S., et al. 2013. 30 Epidemiological assessment of cassava mosaic disease in Central African Republic 31 reveals the importance of mixed viral infection and poor health of plant cuttings. 32 Crop Prot. 44:6–12. 33 34 23 1 Table Caption 2 3 Table 1. Selected EPG variables describing the feeding behavior of M. persicae on single 4 and mixed virus infected melon plant 5 6 7 Figure Legends 8 9 Figure 1. Kinetics of viral loads in single and mix-infected plants. Quantification of 10 CYSDV (left) and WMV (right), showing viral loads in single (blue) and mixed (red) 11 infected plants. Each point represents 2 independent biological replicates of individually 12 infected plants (dataset n=15), collected at 12, 24, 36, 48 and 60 d.p.i. The mean values 13 are indicated by the horizontal lines, connected through an adjusted curve, and surrounded 14 by gray areas representing 95% confidence intervals. 15 16 Figure 2. Symptomatology of infected plants. A) Details of leaves with symptoms 17 caused by single infection with WMV (upper panel) or CYSDV (lower panel). B) 18 Representative single and mix-infected plants photographed at the indicated time-points. 19 C) Side-by-side representative plants of the four categories (from left to right: non- 20 infected, CYSDV, WMV, CYSDV+WMV) photographed at 24 and 48 dpi. 21 corresponding with the development used as virus sources during vector-transmission 22 experiments 23 24 Figure 3. CYSDV transmission in controlled conditions at 24 dpi. Pooled values of 25 virus accumulation are plotted on the upper panel (compared by U Mann-Whitney test) 26 above the tabulated results of the three repetitions of transmission, with number of 27 infected plants/number of plants tested for each independent experiment, followed by the 28 values of two-tailed Chi-square tests (df=1). Transmission rates are showed graphically 29 using pink for CYSDV infected vs. green for uninfected test plants. 30 31 Figure 4. Transmission of CYSDV under field conditions. Average of transmission 32 rates in three individual net-structures, with 20 test plants each, either uninfected or 33 previously infected by WMV, and surrounding single infected CYSDV plants (left panel), 24 1 or mixed infected CYSDV+WMV plants (right panel) as sources of inoculum. No 2 significative differences were found (t-test). 3 4 Figure 5. Free-choice transmission of CYSDV under field conditions. A) Graphical 5 representation of comparative transmission of CYSDV to alternated plants, 10 uninfected 6 and 10 previously infected with WMV, using single infected CYSDV plants as source of 7 inoculum; results for each individual net-structure are represented individually. B) 8 Similar experiment as in a) with mixed infected CYSDV+WMV plants as inoculum 9 source. Analysis of data was performed with an association test. 10 11 Figure 6. WMV transmission in controlled conditions at 24 dpi. Pooled values of virus 12 accumulation are plotted on the upper panel (compared using t-test analysis) above the 13 tabulated results for the three repetitions of transmission, with number of infected 14 plants/number of plants tested for each independent experiment, followed by the values 15 of two-tailed Chi-square tests (df=1). Transmission rates are showed graphically using 16 blue for WMV infected vs. green for uninfected test plants. 17 18 Figure 7. EPG monitoring of aphid feeding behavior. A) Example of a representative 19 EPG record with a potential drop (Pd) in which the different sub-phases (II-1, II-2 and II- 20 3) can be identified. B) Comparison of the duration of the sub-phases corresponding to 21 Pds between aphids fed on single or mix-infected plants c) Comparison of number of 22 pulses of subphase II-3 between aphids fed on single or mix-infected plants. The error 23 bars represent ±1 SEM. ** P<0.005. 24 25 Figure 8. CYSDV and WMV transmission in controlled conditions at 52 dpi. The 26 transmission rate of CYSDV by whiteflies (upper panel) and WMV by aphids (lower 27 panel) is shown for five independent experiments using 10 plants each, with the number 28 of infected plants/number of plants tested for each independent experiment, and the 29 pooled total followed by the value of a two-tailed Chi-square test (df=1). Statistical 30 significant differences are indicated with P-value in bold. Transmission rates are depicted 31 graphically using the same color code as in previous figures: pink for CYSDV-infected, 32 blue for WMV-infected, and green for uninfected test plants. 33 34 25 1 Supporting Information Legends 2 3 Table S1. List of primer sequences and their uses 4 5 Table S2. CYSDV transmission in field conditions from single and mixed-infected plants 6 to uninfected or WMV-infected in different net-houses. 7 8 Table S3. CYSDV transmission in field conditions from single and mixed-infected plants 9 to uninfected or WMV-infected plants in the same net-structures, in alternate positions. 10 11 Table S4. Comparative analysis of EPG parameters in aphids feeding on uninfected, 12 CYSDV, WMV, and WMV+CYSDV infected plants 13 14 Table S5. Monitoring of intracellular subphases during potential drops (Pd) detected in 15 aphids feeding behavior by EPG 16 17 Figure S1. Secondary CYSDV transmission in field conditions. 18 19 26 Table 1. Selected EPG variables describing the feeding behavior of M. persicae on single and mixed virus infected melon plant Variables1 Treatment PPW NWEI P2 WDI P2 WDE P2 0.80±0.17 3.07±0.27 2.10±0.53 WMV 30/30 (0.48-1.04) (2.28-3.32) (0.65-3.18) Np 0.73 0.04 0.03 CYSDV + 0.82±0.23 2.96±0.28 2.00±0.46 28/28 WMV (0.30-1.18) (2.19-3.39) (0.84-2.98) 0.77±0.18 2.75±0.31 1.69±0.55 WMV 30/30 (0.30-1.00) (2.08-3.25) (0.94-3.05) C 0.48 0.06 0.74 CYSDV + 0.80±0.22 2.91±0.27 1.68±0.59 28/28 WMV (0.30-1.15) (2.36-3.26) (0.92-3.26) 1.14±0.26 1.75±0.25 0.70±0.08 WMV 30/30 (0.60-1.59) (1.11-2.13) (0.52-1.12) Pd 0.12 0.02 <0.0001 CYSDV + 1.24±0.23 1.90±0.22 0.74±0.11 28/28 WMV (0.70-1.62) (1.29-2.20) (0.54-1.18) 1 For each treatment, the following parameters were measured: Np, non-probe activity; C, intercellular stylet pathway; Pd, short intracellular punctures. PPW means proportion of individuals that produced the waveform type, with NWEI being the number of waveform events per insect, WDI the waveform duration per insect, and WDE the waveform duration per event. Durations are expressed in minutes for Np and C, and in seconds for Pd. Values are indicated as mean ± 1 SEM (range given in parentheses). 2 Statistical comparisons between the two treatments were performed for each parameter with non-parametric U-Mann Whitney test, considering non Gaussian distribution variables. Bold-type indicates significant differences (p<0.05) 1 27 1 2 3 Fig.1 4 28 1 2 Fig.2 3 29 1 2 Fig.3 3 30 1 2 Fig.4 3 31 1 2 Fig. 5 3 32 1 2 Fig. 6 3 33 1 2 Fig.7 3 34 1 2 Fig. 7 3 35 1 Supplemental Table 1. List of primer sequences and their uses Name Sequence Use RNA1 CYSDV Fw 5’ATGTCGTCGTCACTAGTTGC 3’ Template for in vitro RNA1 CYSDV Rv 5’GTCGAGACCACTCCAGTACA 3’ transcription 1 RNA2 CYSDV Fw 5’AGGTGGGTAGGTGTTGACAG 3’ Template for in vitro RNA2 CYSDV Rv 5’TGGGAAGTACCGTCTCCTTACC 3’ transcription 1 P1 WMV Fw 5’CACCATGGCAACAATCATGTTTGGAG 3’ Template for in vitro P1 WMV Rv 5’TCAATAATGTTGAATATCTTCTATCTCC 3’ transcription 1 qPCR RNA1 CYSDV Fw 5’GACTCAAACAAGCGCGATGA 3’ Analysis of viral load qPCR RNA1 CYSDV Rv 5’GCGTTGCCGTTACTGAGGAT 3’ (CYSDV RNA1) qPCR RNA2 CYSDV Fw 5’TTTGTCCGCTGACGTACTGG 3’ Analysis of viral load qPCR RNA2 CYSDV Rv 5’GGTTCAATGCAACGACTCAGA 3’ (CYSDV RNA2) qPCR P1 WMV Fw 5’TGCAACTATTGTGGCCTTCG 3’ Analysis of viral load qPCR P1 WMV Rv 5’CATGGCACCTCCACCTCATA 3’ (WMV) WMV Fw 5’CTTATGGTTGTCATTGCTATG 3’ Detection of WMV WMV Rv 5’ CCCACCAACTGTTGGAAG 3’ MA156 Fw 5’ GAAGAATTCCAGGCAAGG 3’ Detection of CYSDV 2 MA129 Rv 5’ TCACATCATCAATCCAAAAG 3’ 1 2 Primers were used for cloning partial virus sequences to generate in vitro transcripts as templates 3 for quantitative comparisons in RT-qPCR. Non-viral sequences used for cloning purposes are 4 highlighted in bold. 2 5 Primers described by Marco et al. 2003 6 7 36 1 Supplemental Table 2. CYSDV transmission in field conditions from single and mixed- 2 infected plants to uninfected or WMV-infected in different net-houses. Infected 1 Source Receptors Repetition plants2 Total Average3 CYSDV 1 19/20 Uninfected 2 18/20 53/60 88.33% 3 16/20 1 19/20 WMV 2 19/20 56/60 93.33% 3 18/20 CYSDV+WMV 1 15/20 Uninfected 2 20/20 50/58 86.2% 3 15/18 1 20/20 WMV 2 19/20 59/60 98.33% 3 20/20 1 3 Three independent net-structures, each with 20 receptor plants, all of the same type, are shown 2 4 Results obtained in each independent net-house are indicated for each treatment, and pooled 5 values and percentages are shown in the last columns. 3 6 Statistical analysis were performed by two-tail t-Test comparisons of every pair of treatments 7 (after verifying inequality of variances by F-Test). None of the paired comparisons indicated 8 significant differences between the samples means. 9 10 37 1 Supplemental Table 3. CYSDV transmission in field conditions from single and mixed- 2 infected plants to uninfected or WMV-infected plants in the same net-structures, in 3 alternate positions. Source Repetition1 Receptors Infected plants Average CYSDV 1 Uninfected 3/10 30% WMV 3/10 30% 2 Uninfected 4/10 40% WMV 8/10 80% 3 Uninfected 4/9 44,44% WMV 5/9 55,55% Total Uninfected 11/29 37,93% WMV 16/29 55,17% CYSDV+WMV 1 Uninfected 3/10 30% WMV 3/10 30% 2 Uninfected 5/10 50% WMV 6/10 60% 3 Uninfected 4/10 40% WMV 8/10 80% Total Uninfected 12/30 40% WMV 17/30 56,66% 1 4 The results of three independent net-structures (with 10 uninfected and 10 previously 5 WMV-infected receptor plants in alternated positions) are shown for each treatment, with 6 the pooled values for each category in the "Total" rows. 2 7 Statistical analysis were performed by two-tail t-Test comparisons of every pair of 8 treatments (after verifying inequality of variances by F-Test). None of the paired 9 comparisons indicated significant differences between the samples means. 10 11 38 1 39 1 Supplemental Table 4. Comparative analysis of EPG parameters in aphids feeding on 2 uninfected, CYSDV, WMV, and WMV+CYSDV infected plants Parameter1 Treatment PPW NWEI WDI WDE NP Uninfected 20/20 0.73±0.05 2.88±0.06 2.04±0.04 (0.30-1.04) a (2.24-3.16) a (0.84-2.87) a CYSDV 23/23 0.88±0.03 2.91±0.06 1.90±0.04 (0.60-1.20) a (2.34-3.33) a (0.86-2.95) b WMV 30/30 0.80±0.03 3.07±0.05 2.10±0.04 (0.48-1.04) a (2.28-3.32) b (0.65-3.18) ac CYSDV 28/28 0.82±0.04 2.96±0.05 2.00±0.03 +WMV (0.30-1.18) a (2.19-3.39) a (0.84-2.98) abd C Uninfected 20/20 0.71±0.05 2.96±0.04 1.91±0.06 (0.30-1.04) a (2.49-3.26) a (1.04-3.26) a CYSDV 23/23 0.87±0.03 2.90±0.07 1.75±0.05 (0.60-1.18) b (1.96-3.21) ab (0.84-3.14) b WMV 30/30 0.77±0.03 2.75±0.06 1.69±0.04 (0.30-1.00) ab (2.08-3.25) b (0.94-3.05) b CYSDV 28/28 0.80±0.04 2.91±0.05 1.68±0.05 +WMV (0.30-1.15) ab (2.36-3.26) ab (0.92-3.26) b Pd Uninfected 20/20 1.28±0.04 1.86±0.04 0.69±0.004 (0.85-1.53) a (1.38-2.09) ab (0.51-1.14) ab CYSDV 23/23 1.26±0.04 1.85±0.04 0.69±0.004 (0.85-1.59) a (1.45-2.21) ab (0.50-1.04) a WMV 30/30 1.14±0.05 1.75±0.05 0.70±0.005 (0.60-1.59) a (1.11-2.13) a (0.52-1.12) b CYSDV 28/28 1.24±0.04 1.90±0.04 0.74±0.005 +WMV (0.70-1.62) a (1.29-2.20) b (0.54-1.18) c 1 3 Parameters measured include: NP, non-probe activity; C, intercellular stylet pathway; Pd, short 4 intracellular punctures. PPW means proportion of individuals that produced the waveform type, 5 with NWEI being the number of waveform events per insect, WDI the waveform duration per 6 insect, and WDE the waveform duration per event. Durations are expressed in minutes for NP 7 and C, and in seconds for Pd. Values are indicated as mean ± SE with extremes given in 8 parentheses. The statistical comparisons were performed in pairs for each parameter with non- 9 parametric U-Mann Whitney tests, considering non Gaussian distribution variables. Only values 10 followed by different letters were found significantly different (p<0.05). 11 40 1 Supplemental Table 5. Monitoring of intracellular subphases during potential drops 2 (Pd) detected in aphids feeding behavior by EPG Sub-phase1 Treatment PPW WDI P2 II-1 WMV 30/30 1,36±0,05(0,75-1,82) 0,09 CYSDV+WMV 28/28 1,49±0,05(0,81-1,90) II-2 WMV 30/30 1,03±0,05(0,52-1,46) 0,08 CYSDV+WMV 28/28 1,15±0,05(0,53-1,53) II-3 WMV 30/30 1,34±0,04(0,77-1,69) 0,002 CYSDV+WMV 28/28 1,53±0,04(0,91-1,84) 1 3 The sub-phases of pd, associated with uptake (acquisition, II-3) and release (inoculation, II-1) 4 of non-persistently transmitted viruses are identified following Martin et al., 1997. PPW means 5 proportion of individuals that produced the waveform type, and WDI is the waveform duration 6 per insect in seconds. Values are indicated as mean ± SE with extremes given in parentheses. 2 7 Statistical comparisons between the two treatments was performed for each parameter with non- 8 parametric U-Mann Whitney test, considering non Gaussian distribution variables. Bold-type 9 indicates significant differences (p<0.05). 10 11 41 1 2 3 S1 Figure 42