SatRNA-mediated exacerbation of disease symptoms in plants reduces virus transmission potential by aphids

Abstract

Background

Plant viruses have evolved adaptations that enable them to alter host cues, thereby facilitating their replication and efficient transmission by insect vectors. Satellite RNAs (satRNAs), which accompany certain plant RNA viruses and are dependent on them for replication and transmission, can change the progression of pathogenesis and the expression of disease symptoms. This study aimed to analyse how the change in the course of infection by satRNA (exacerbation or mitigation of pathogenesis) impacts the subsequent stages of virus transmission. We hypothesised that satRNAs influence insect behaviour toward infected plants depending on their effect on pathogenesis progress; specifically, a significant disease exacerbation reduces plant attractiveness to aphids, whereas symptom mitigation promotes attraction.

Results

Using peanut stunt virus (PSV) and cucumber mosaic virus (CMV), and their satRNAs, which induce divergent infection symptoms on Nicotiana benthamiana or Solanum lycopersicum plants, olfactometry, electrical penetration graph (EPG) monitoring, virus acquisition and transmission by Myzus persicae were analysed. The results showed that satRNA, naturally associated with the helper virus, that alleviate disease symptoms caused the plants to be more attractive to the insect vectors. On the other hand, the presence of satRNAs leading to symptom exacerbation reduced plant attractiveness and discouraged phloem feeding. However, acquisition effects were host-dependent: while symptom-exacerbating satRNAs generally reduced acquisition, nc-satRNA markedly enhanced CMV acquisition from N. benthamiana. Moreover, virus transmission was significantly reduced only in S. lycopersicum.

Conclusions

These findings suggest that symptom-attenuating satRNAs have greater capacity to persist in the environment by limiting disease damage in host plants and maintaining plant attractiveness to aphid vectors, thereby facilitating virus acquisition and transmission. In contrast, symptom-exacerbating satRNA variants appear not only to diminish the number of cells available for viral replication due to severe symptoms but also to reduce plant palatability and to impair vector feeding and thus virus transmission, particularly in crop hosts such as S. lycopersicum, which may limit their environmental persistence. Together, these results underscore the epidemiological relevance of satRNA-mediated symptom modulation as a factor shaping virus spread in natural and agricultural settings.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09453-2.

Background

Plant viruses are intracellular parasites that alter the metabolism of infected plants. They have evolved adaptations not only to evade the host plant’s antiviral defences but also to ensure their successful replication and movement within the plant, as well as efficient spread between plants [1]. An infected plant undergoes such significant metabolic changes that it may become more attractive to insects acting as virus vectors, thereby facilitating the pathogen’s transmission to other plants [2–4]. These changes may affect the physiological and chemical properties of the host plant, which influence the attractiveness of the infected plants to vectors through olfactory, visual, and gustatory cues [5, 6].

Some virus strains can be accompanied by subviral RNAs – defective interfering RNAs or satRNAs – which, by influencing the course of pathogenesis and the biology of the host plant infected with the helper virus, may also affect virus transmission [7, 8]. SatRNAs may (as is the case with satRNA of many viruses) not encode any functional protein and are fully dependent on the helper virus for replication, encapsidation, movement, and transmission [9]. SatRNAs may influence the symptoms induced by the helper virus (either attenuating or exacerbating them), the accumulation level of the helper virus (usually decreasing it) [10–12], and the progression of pathogenesis [13, 14]. The outcome of the satRNA-helper virus-host plant interactions depends on the satRNA sequence, the strain of the helper virus, the host plant species, and external factors (e.g., temperature) [8]. SatRNAs are poorly studied in the context of their interactions with helper viruses and insect vectors. Recently, satRNA has been found to alter aphid behaviour. For example, cucumber mosaic virus (CMV) Y-sat, by inducing yellow leaf colouration in tobacco, was shown to enhance the attractiveness of Myzus persicae toward Y-sat-infected plants through visual cues [15]. However, since satRNAs usually lower CMV titre in the infected plants, they have been observed to decrease the efficiency of CMV transmission by Aphis gossypii in tomato and melon plants [16, 17]. Nevertheless, our full understanding of satRNA, viruses, and their interactions with viral vectors remains limited. The four-way interactions between virus, satRNA, host plant, and insect vector have not been studied so comprehensively so far, including aphid orientation, feeding behaviour, virus acquisition and transmission.

Here, we studied the effects of satRNAs on the three-way interaction between plants, viruses, and their insect vectors, using non-persistently aphid-transmitted viruses as an example: CMV and peanut stunt virus (PSV) (genus Cucumovirus, family Bromoviridae). CMV infects a wide range of hosts, encompassing more than 1,000 plant species, while the host range of PSV is mainly limited to plants belonging to the Fabaceae and Solanaceae families [18, 19]. Nevertheless, both viruses are distributed worldwide and cause serious diseases in agriculturally important plants, constituting a significant global problem. CMV and PSV have positive-sense single-stranded RNA (ss(+)RNA) genomes, consisting of three genomic and two subgenomic strands. RNA1 and RNA2 encode the 1a and 2a proteins, respectively, which form the viral replication complex. Additionally, RNA2 is a source of subgenomic RNA4A coding for the 2b protein. RNA3 codes for the movement protein (MP, 3a) and coat protein (CP); the latter is synthesised from subgenomic RNA4 [20, 21].

We hypothesised that satRNAs that attenuate symptoms promote helper virus transmission, whereas those that aggravate symptoms impede it. This hypothesis stems from the observation that severe pathogenic symptoms can sometimes lead to premature plant death, potentially reducing the time window during which the virus can be acquired by insect vectors and subsequently transmitted to neighbouring plants. Therefore, our goal was to investigate whether satRNAs that modify the host phenotypes induced by their respective helper viruses alter vector behaviour toward infected hosts (by olfactory and/or gustatory cues) and consequently affect transmission efficiency. To address this knowledge gap, we studied the transmission efficiency of the helper virus and the preceding behaviour of the green peach aphid (M. persicae), on the model plant Nicotiana benthamiana infected with CMV and PSV and the crop plant tomato (Solanum lycopersicum L.) infected with CMV, with or without their respective satRNAs. We conducted our analyses using two PSV strains: PSV-G, which is naturally devoid of satRNA but exhibits symptom exacerbation upon satRNA co-infection [11], and PSV-P, which naturally possesses satRNA and co-infected with this subviral particle shows slight symptom attenuation [13]. Additionally, we analysed the CMV Fny strain with non-necrogenic (non-nc-satRNA) and necrogenic satRNAs (nc-satRNAs). Since satRNAs can modify the symptoms of viral infections in plants, leading to their exacerbation or weakening, affect plant metabolism during viral infection, we assume that the presence of satRNA in virus-infected plants modifies their olfactory and gustatory attractiveness to the insect vectors and influences the aphid feeding behaviour. This, in turn, affects the efficiency of virus acquisition and transmission.

Materials and methods

Biological materials

N. benthamiana plants used to obtain viral particles from agroinfectious clones were grown under controlled conditions in greenhouse chambers with a 14 h light/10 h dark cycle at 24 °C day/20 °C night. Five-to-six-week-old N. benthamiana seedlings were infiltrated with agroinfectious clones of PSV-G (genomic RNAs sequences GenBank accession numbers JN135294, JN135293, JN135292) or PSV-P (genomic RNAs sequences GenBank accession numbers EU570236, EU570237, EU570238), either alone or in combination with satRNA-P (RNA sequence GenBank accession number EU570236) (PSV-G + satRNA-P, PSV-P + satRNA-P) inserted into Agrobacterium tumefaciens cells. Similarly, plants were infiltrated with agroinfectious clones of CMV-Fny (genomic RNAs sequences GenBank accession numbers NC_002034, NC_002035, NC_001440) alone or in combinations with non-necrogenic satRNA (sequence GenBank accession number Z75882) (CMV + non-nc-satRNA) or necrogenic satRNA (sequence GenBank accession number Z75883) (CMV + nc-satRNA). Plants infiltrated with the infiltrating buffer served as negative controls (mock). At 9 days post-infiltration (dpa), the fragments of systemically infected leaves were harvested for virus and satRNA presence verification (Additional File 1—Protocol S1).

At 14 dpa, systemically infected leaves from four plants (ca. 2.0 g) were collected and homogenised in 5 ml of 0.05 M phosphate buffer (pH 7.5) using a pestle and mortar. This sap was used for the mechanical inoculation of plants used in subsequent experiments. Two leaves of 4-week-old N. benthamiana (for PSV and CMV) and S. lycopersicum cv. Betalux (for CMV) plants were dusted with carborundum and inoculated with cucumovirus with or without satRNA combinations. Plants treated with the phosphate buffer only (mock-inoculated) were taken as the negative controls. Those plants were used in the subsequent experiments (Fig. 1). A separate set of plants was prepared for each experiment. All plants were grown in a greenhouse chamber with a 14 h light/10 h dark cycle at 21 °C day/18 °C night.

Fig. 1.

Experimental workflow used to assess the influence of cucumovirus-associated satellite RNAs (satRNAs) on aphid behaviour and virus transmission. Aphids (Myzus persicae) were exposed to plants of two host species – Nicotiana benthamiana (model plant) and Solanum lycopersicum cv. Betalux (crop plant) – inoculated with viruses with or without satRNAs inducing divergent symptoms. N. benthamiana was inoculated with peanut stunt virus (PSV) or cucumber mosaic virus (CMV), whereas S. lycopersicum was inoculated with CMV. Mock-inoculated plants served as negative controls. A separate set of plants was prepared for each of the four experimental stages: (1) Orientation behaviour – aphid host preference was assessed using a three-choice olfactometer under dark or light conditions; (2) Feeding behaviour – aphid stylet activity was recorded by Electrical Penetration Graph (EPG) monitoring to characterise probing and feeding parameters; (3) Virus acquisition – the efficiency of cucumovirus acquisition by aphids was quantified using droplet digital PCR (ddPCR); (4) Virus transmission – the efficiency of cucumovirus and satRNA co-transmission to healthy test plants was assessed by RT-qPCR. Treatment abbreviations: PSV-G, PSV strain G; PSV-G + satRNA-P, PSV strain G with satRNA-P (symptom exacerbation); PSV-P, PSV strain P; PSV-P + satRNA-P, PSV strain P with satRNA-P (symptom attenuation); CMV + non-nc-satRNA, CMV with non-necrogenic satRNA; CMV + nc-satRNA, CMV with necrogenic satRNA. Aphid image was obtained from TogoTV

Wingless individuals of M. persicae (Sulz.) (peach-potato or green peach aphid) were used in the experiments. The aphid stock colonies were maintained on N. tabacum or Brassica napus plants placed in an insect cage. The growth conditions were a 14 h light/10 h dark cycle at 18–21 °C day/18 °C night.

All greenhouse experiments were performed in the Research Centre of Quarantine, Invasive and Genetically Modified Organisms (IPP-NRI, Poznań, Poland).

Viral agroinfectious clones

These studies involved working with the following viruses: PSV-G, PSV-P, and CMV-Fny. Agroinfectious clones of PSV-P and satRNA-P, and CMV-Fny have been synthesised before [22, 23]. The agroinfectious clones of PSV-G and CMV non-nc- and nc-satRNAs were synthesised as described in Wrzesińska et al. [22]. Previously described infectious copies of PSV-G and cDNA clones of CMV non-nc- and nc-satRNAs cloned under T7 RNA polymerase promoter [11, 16] were used as templates for the RNA amplification through polymerase chain reaction (PCR) with the use of forward and reverse primers specified in Additional File 1—Table S2.

Analysis of insects’ orientation behaviour using an olfactometer

A four-arm olfactometer, configured as a three-choice assay by permanently blocking one arm with parafilm, was used to analyse insects’ orientation behaviour (Additional File 1—Figure S1). It comprised a central main arena (18 cm in diameter) and four tunnels (15 cm long) leading to side jars (18 cm in diameter). Each tunnel was terminated with a copper mesh to prevent physical contact between the plant and aphids. To conduct the experiment, plants infected with the virus, virus + satRNA, and mock-inoculated plants at 2–3 weeks post-inoculation (wpi), when the infection symptoms began to develop, were used. In each side jar, a plant from one treatment was placed and covered with a lid supplemented with a carbon filter to remove any odour from the air intake. Thirty aphids, after 2–3 h of starvation at 22 °C, were placed in the central arena and covered with a lid. An air pump connected to the central jar allowed air intake from the side jars. Under dark conditions, the olfactometer was covered with a black cloth and kept in a dark room to avoid visual cues. Under light conditions, the olfactometer was covered with a white permeable cloth and kept in the room under light. The choice made by the aphids was checked after 20 min. Those that entered the tunnel were considered to have made a choice. Three plants (biological replicates) from each treatment (mock, virus, virus + satRNA) were used at one measurement. After the test, the positions of the side jars with each biological replicate were swapped to avoid aphid directional cues (three technical replicates per one biological replicate). The number of aphids making a choice in each technical replicate for one plant was then summed. The experiment was conducted on 12 plants (biological replicates) from each treatment. Statistical analysis was performed in R (ver. 4.5.1). Aphid orientation data, expressed as counts of individuals making a choice per biological replicate, were analysed using a multinomial logistic regression model fitted with the ‘mblogit’ function from the ‘mclogit’ package (ver. 0.9.15). The model was fitted on individual-level data (one observation per choosing aphid), which maximises statistical power given the inherently low choice rate. Bioassays under light and dark conditions were conducted with independent sets of plants. Light condition was therefore treated as a between-group fixed-effect predictor. Since multinomial logistic regression estimates separate coefficients for each response category, the effect of light condition was modelled independently for each treatment contrast, which is equivalent to a full treatment × light condition interaction. The contribution of light condition to overall model fit was assessed by a likelihood ratio test comparing the full model against an intercept-only (null) model. Estimated choice probabilities with 95% confidence intervals for each treatment within each light condition were derived from the fitted model using the ‘emmeans’ package (ver. 1.11.2.8). Pairwise contrasts between treatments were evaluated separately within each light condition on the probability scale, and p-values were adjusted for multiple comparisons using the Benjamini–Hochberg method.

The analysis of the insects' feeding behaviour using electrical penetration graphs

The feeding of M. persicae on N. benthamiana and S. lycopersicum plants was monitored using the Electrical Penetration Graphs (EPG) technique [24]. The Giga-4 EPG system (Wageningen, Netherlands) was used. A microelectrode (gold wire with a diameter of 12.5 μm and a length of about 2–3 cm) was connected to the wingless aphid female’s body using a drop of silver paint. The aphid was placed on a leaf and monitored for 8 h. Its feeding behaviour was studied using the EPG Stylet + d computer program. Each recording included a record of the activity of one aphid’s stylets on one plant. Sixteen plants and sixteen aphids were used for each treatment (biological replicates) (one aphid per plant) before (7–8 days post-infection (dpi)) and after (27–28 dpi) fully developed infection symptoms (separate sets of plants were used for each infection stage). After the recording was finished, the leaves on which the aphids were feeding were collected for virus and satRNA presence verification (Additional File 1—Protocol S1). Raw EPG recordings were processed using the Stylet + software (EPG Systems, Wageningen, Netherlands). The following EPG waveforms were identified and analysed: non-probing (Np), penetration of peripheral tissues (ABC), potential drops (pd), salivation into sieve elements (E1), and ingestion of phloem sap (E2). For each waveform, the total duration and number were calculated as EPG-derived variables. Additionally, time to first probe was included as a pre-penetration behavioural parameter. Derailed stylet mechanism (F) and ingestion of xylem sap (G) waveforms were observed sporadically and occurred too infrequently to be subjected to reliable statistical analysis.

Statistical analysis was conducted with R (ver. 4.5.1). The effects of the infection variant (‘Variant’: mock, virus, virus + satRNA) and symptom development phase (‘Symptoms’: before and after symptom appearance) on EPG-derived variables (‘time’ and ‘number’) was assessed using Generalised Linear Models (GLM) including both main effects and their interaction term. Since separate sets of plants were used at each time point, a repeated-measures design was not applicable; the two time points were therefore treated as an independent between-subjects factor within the factorial model. For each response variable, multiple GLM distribution families were evaluated (Gaussian, Gamma, Inverse Gaussian, Poisson, Quasi-Poisson, negative binomial), and the optimal model was selected based on the Akaike Information Criterion (AIC) and simulation-based residual diagnostics implemented in the DHARMa package (ver. 0.4.7), including tests for uniformity and dispersion of residuals. The significance of the Variant × Symptoms interaction was assessed using analysis of deviance, with F-tests applied to Gaussian, Gamma, Inverse Gaussian and Quasi-Poisson models, and likelihood ratio tests (χ²) applied to Poisson and negative binomial models. Post-hoc pairwise comparisons between infection variants were performed separately within each time point using the ‘emmeans’ package (ver. 1.11.2.8) with p-values adjusted for multiple comparisons using the Tukey method, providing condition-specific p-values per time point. The proportion of aphids engaging in E1 and E2 waveforms was compared between treatments using a binomial Generalised Linear Model (GLM), with treatment as a fixed factor and CMV-infected plants as the reference group.

Viral RNAs and satRNAs acquisition assay

Apterous aphids were starved for 2–3 h, followed by 8 min of virus and satRNA acquisition access period (AAP) on a systemically infected leaf from N. benthamiana or S. lycopersicum plants infected with PSV-G, PSV-P, or CMV-Fny, with or without associated satRNAs 2 wpi. For each experimental condition, three leaves (one leaf from one plant each) were used with 7–8 aphids exposed on each leaf. Stylets from viruliferous and non-viruliferous aphids were carefully dissected using a scalpel under a stereomicroscope and transferred individually with a very thin brush into a single Eppendorf tube for molecular analyses. RNA extraction was performed using the Total RNA Mini Concentrator kit (A&A Biotechnology, Gdańsk, Poland). RNA was eluted in 12 µl of RNase-free water. The entire RNA eluate was used for first-strand cDNA synthesis with the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA), using random hexamer primers (EURx, Gdańsk, Poland).

Quantification analyses were performed using the BioRad QX200 System (Bio-Rad, Hercules, California, USA) with virus-specific primer pairs targeting individual viral RNA segments (RNA1, RNA2, and RNA3) and satRNAs (Additional File 1—Table S3). The detection and absolute quantification of viral RNA were carried out following a previous protocol by Budziszewska et al. [25]. Briefly, reactions were prepared in a 20 μl volume using QX200 ddPCR EvaGreen Supermix (Bio-Rad), with 0.1 μM of specific primer pairs (Additional File 1—Table S3), and 1 μl of a tenfold diluted cDNA derived from each sample (one stylet per replicate). Digital droplet PCR (ddPCR) was performed under the following thermal conditions: initial denaturation at 95 °C for 10 min, followed by 36 amplification cycles of 94 °C for 30 s, annealing at the primer-specific temperature for 60 s, ramp rate 2 °C/s, and a final step at 98 °C for 10 min. Droplet fluorescence was read using the Bio-Rad Droplet Reader (Bio-Rad) and viral RNA copy numbers were estimated using QuantaSoft v1.7 software.

The source leaves were subjected to virus and satRNA accumulation analysis by quantitative real-time PCR (RT-qPCR). Total RNA was extracted using 500 μL of TRI Reagent Solution (Thermo Fisher Scientific), followed by precipitation with 2-propanol and washing with 70% ethanol. The air-dried precipitate was suspended in nuclease-free water. The quantity and quality of the RNA were estimated in a DS-11 Series Spectrophotometer (DeNovix, Wilmington, DE, USA). The isolated RNA was subjected to DNA digestion with RNase-free DNase I (Thermo Fisher Scientific). One μg of RNA was reverse transcribed using RevertAid Reverse Transcriptase (Thermo Fisher Scientific) with random hexamer primers (EURx). RT-qPCR was performed in triplicate (technical replicates) for each sample using a LightCycler 480 Instrument (Roche Diagnostics, Basel, Switzerland). The reaction mixtures consisted of iTaq™ Universal SYBR Green Supermix (Bio-Rad), 0.5 μM forward and reverse primers (Additional File 1—Table S3), and 1 μL of cDNA. Reactions (qPCR) were performed as follows: 3 min of initial denaturation at 95 °C followed by 40 cycles of 20 s at 95 °C, 20 s annealing (at the temperatures listed in Additional File 1—Table S3), and 20 s at 72 °C. Dissociation curves were generated during temperature ramping from 65 °C to 95 °C. Absolute quantification was done based on the standard curves created using fivefold dilutions of cDNA synthesised from RNA isolated from the infected plants. The exact copy number of the analysed corresponding genomic RNAs was calculated using Avogadro’s number.

Statistical analysis of the virus and satRNA accumulation level in plants and aphids was performed with R (ver. 4.5.1). The data normality was checked, followed by the Mann–Whitney test (for PSV data) and the Kruskal–Wallis test (for CMV data) using the ‘ggstatsplot’ package (ver. 0.13.2) [26].

Virus and satRNA transmission assay

Apterous aphid individuals were starved for 2–3 h, followed by 8 min of virus and satRNA AAP on a leaf from N. benthamiana or S. lycopersicum plants infected with PSV-G, PSV-P, or CMV-Fny, with or without associated satRNAs 2 wpi. Ten aphids were collected and transferred to a healthy plant of a species corresponding to the virus and satRNA source plant. After 1 day (inoculation access period), the aphids were removed. After infection symptoms development, the plants’ leaves (systemically infected and apical ones) were harvested 2–3 weeks after transmission for virus and satRNA presence verification. Ten or twelve plants were used for each virus and satRNA treatment. The experiment was performed three times.

To assess the levels of viral RNAs and satRNAs in the harvested plants, total RNA was extracted, followed by reverse transcription of 500 ng of RNA as described above. RT-qPCR was performed in triplicate (technical replicates) for each sample using a LightCycler 480 Instrument (Roche Diagnostics) using the same qPCR reagents, qPCR conditions, and absolute quantification calculations were done as described above.

Statistical analysis was performed with R (ver. 4.5.1). The normality of the data was checked and the Mann–Whitney test for each experiment was performed to compare accumulation levels between the cucumovirus and cucumovirus + satRNA groups. To assess the overall effect, a meta-analysis of p-values was conducted using Fisher’s method.

Results

Nicotiana benthamiana and tomato plants inoculated with agroinfectious copies of the cucumovirus and satRNAs developed characteristic disease symptoms

The infectivity of the generated virus and virus + satRNA agroinfectious clones in the infiltrated N. benthamiana plants was confirmed by gel electrophoresis of RT-PCR products resulting from amplification of fragments corresponding to CP genes and the fragments of satRNAs sequences (Additional File 1—Figure S2). After confirmation of the virus and satRNA presence, plants infected with PSV or PSV + satRNA-P and CMV or CMV + satRNAs were used for passage to N. benthamiana (PSV and CMV) and S. lycopersicum (CMV). Those plants were used for further experiments. Plants displayed characteristic disease symptoms at 21–23 dpi (Fig. 2). The infection of N. benthamiana plants with PSV-G and PSV-P resulted in stunted growth, leaf malformations, and chlorosis (Fig. 2a). SatRNA presence in the PSV-G inoculum led to symptom aggravation, causing a more dwarfed phenotype, more severe systemically infected leaf distortion, and necroses, while satRNA addition to the PSV-P strain resulted in reduced symptoms (Fig. 2a). CMV infection in N. benthamiana resulted in the characteristic symptoms caused by the Fny strain: stunting and leaf malformation, whereas the addition of satRNAs resulted in the emergence of necroses, where nc-satRNA led to a higher number of necrotic spots compared to non-nc-satRNA (Fig. 2b). On the other hand, S. lycopersicum infected with CMV displayed slightly delayed growth, mild chlorosis and mosaic, systemically infected leaf distortion, and shoestring-like apical leaf blades (Fig. 2c). The presence of non-nc-satRNA visibly attenuated infection symptoms, as evidenced by reduced leaf distortion and improved plant stature compared to CMV-only infected plants The presence of nc-satRNA caused the appearance of necroses at the late stage of infection (Fig. 2c).

Fig. 2.

Disease symptoms induced by cucumoviruses with or without satellite RNAs (satRNAs) on two host plant species at 21–23 days post-inoculation (dpi). a Nicotiana benthamiana plants inoculated with peanut stunt virus strain G (PSV-G) alone show mild mosaic symptoms, whereas co-inoculation with satRNA-P (PSV-G + satRNA-P) causes symptom exacerbation, including severe leaf deformation and stunting. Plants inoculated with PSV strain P (PSV-P) alone show mosaic and mild distortion, while PSV-P + satRNA-P-infected plants display symptom attenuation relative to PSV-P alone. b N. benthamiana plants infected with cucumber mosaic virus (CMV) alone develop mosaic and mild leaf distortion. Co-infection with non-necrogenic satRNA (CMV + non-nc-satRNA) results in similar or slightly milder symptoms, whereas co-infection with necrogenic satRNA (CMV + nc-satRNA) causes visible chlorosis and early signs of tissue deterioration. c Solanum lycopersicum cv. Betalux plants infected with CMV alone show mosaic and leaf curling. CMV + non-nc-satRNA-infected plants display comparable symptoms, while CMV + nc-satRNA co-infection leads to severe systemic necrosis. The inset photograph (red arrow) shows an advanced stage of nc-satRNA-induced necrosis at 36 dpi, resulting in near-complete plant collapse. All plants were mechanically inoculated by leaf sap from N. benthamiana plants infiltrated with agroinfectious clones of each virus with or without satRNA. Mock-inoculated plants are shown as healthy controls in each panel

satRNA altered the orientation behaviour towards virus-infected plants

The effect of satRNA on plant attractiveness to aphids was examined in a three-choice olfactometer test, in which aphids could select between non-infected, cucumovirus-infected, and cucumovirus + satRNA-infected plants under dark or light conditions. Less than a quarter of the insect individuals moved toward any of the tested plants regardless of cucumovirus species, host plant species, or light/dark condition (Additional File 2—Table S4).

In the experiment with PSV G strain, aphids were least attracted to PSV-G + satRNA-infected N. benthamiana plants compared to both mock-inoculated (p-value 0.0016) and PSV-G -inoculated plants (p-value 0.0457) under light conditions (Fig. 3a). This demonstrates that light significantly reduces aphid preference for PSV-G + satRNA-infected plants (p-value 0.0157, Additional File 2—Table S6). Whereas, in the experiment with PSV P strain, PSV-P + satRNA-P-infected plants were the most attractive to aphids, significantly more so than mock-inoculated plants (p-value 0.0334) and there was a trend against PSV-P-infected plants (p-value 0.0886) under dark conditions (Fig. 3a).

Fig. 3.

Preference of Myzus persicae toward virus-infected plants with or without satellite RNAs (satRNAs) assessed by olfactometry under dark and light conditions. The number of aphids choosing each plant treatment in a three-choice olfactometer, in which aphids were simultaneously offered mock-inoculated (healthy), virus-infected, and virus + satRNA plants. Each data point represents the number of aphids that chose a given plant within one biological replicate; the red dot indicates the mean, and the boxplot range spans the minimum to maximum observed values. Left panels show results recorded under dark conditions (grey background), and right panels under light conditions (yellow background). Panels correspond to: a Nicotiana benthamiana plants infected with peanut stunt virus (PSV) strains G or P, with or without satRNA-P; b N. benthamiana plants infected with cucumber mosaic virus (CMV) with or without non-necrogenic (non-nc-satRNA) or necrogenic satRNA (nc-satRNA); c Solanum lycopersicum cv. Betalux plants infected with CMV with or without satRNAs. Statistical significance was assessed using a multinomial logistic regression model. Pairwise contrasts between treatments were evaluated separately within each light condition on the probability scale, and p-values were adjusted for multiple comparisons using the Benjamini–Hochberg method. Asterisks indicate statistically significant differences (* – p-value < 0.05) between treatments within a given light condition. Grey dots indicate borderline significance (0.05 ≤ p-value < 0.10). Estimated choice probabilities (%) and treatment × light condition interaction p-values are provided in Additional File 2 – Tables S5 and S6, respectively. Treatment abbreviations: PSV-G – PSV strain G, PSV-G + satRNA-P – PSV strain G with satellite RNA-P (satRNA-P), PSV-P – PSV strain P, PSV-P + satRNA-P – PSV strain P with satRNA-P, CMV + non-nc-satRNAs – CMV with non-necrogenic satRNA, CMV + nc-satRNA – CMV with necrogenic satRNA

The analysis of satRNA influence on aphid preferences toward N. benthamiana and S. lycopersicum plants mock-inoculated, infected with CMV, or infected with CMV + satRNAs showed that the most attractive plants are those infected with CMV compared to both mock-inoculated (p-value 0.0223 for N. benthamiana, p-value 0.0703 for S. lycopersicum) and CMV + nc-satRNA-infected plants (p-value 0.0111 for N. benthamiana, p-value 0.0347 for S. lycopersicum) under light conditions (Fig. 3b-c). Moreover, CMV + nc-satRNA-infected S. lycopersicum plants attracted the least number of aphids when compared with mock-, and CMV-inoculated ones (p-value 0.0001 and p-value 0.0008, respectively) under dark conditions (Fig. 3c). These results indicate that in this system (mock-inoculated, CMV-, and CMV + nc-satRNA-infected S. lycopersicum), the light condition has a significant influence on the preferences of aphids towards infected plants (p-value 0.002, Additional File 2—Table S6) and visual cues might play a considerable role in shaping aphid orientation behaviour. On the other hand, the presence of non-nc-satRNA has no effect on aphid preference between mock-inoculated, infected with CMV, or infected with CMV + satRNAs N. benthamiana and S. lycopersicum plants, irrespective of dark or light conditions (Fig. 3b-c, Additional File 2—Table S6).

Overall, viruses in co-infection with satRNAs that severely exacerbated symptoms were associated with reduced host plant attractiveness when compared to the virus-only-infected plants under light conditions. Plants infected with the virus alone (mostly applies to CMV) were often more attractive to aphids than non-infected ones. However, the effect of satRNAs that alleviate symptoms depends on the virus strain, in the case of PSV-P, it promoted attractiveness under dark conditions, while in the case of CMV, it has no significant effect on aphid orientation behaviour.

satRNA presence modified Myzus persicae probing and feeding activity

The analysis of EPG waveforms aimed to verify whether satRNA influences the probing and feeding behaviour of M. persicae on the cucumovirus-infected plants. For the acquisition and transmission of non-persistently transmitted viruses by aphids, the most important EPG events are pds, occurring during the ABC waveform phase associated with mesophyll tissue penetration, when the aphid stylet briefly punctures the cell membrane [27, 28]. Although the aphids were monitored for 8 h, their feeding activity lasted for approximately 2 h; therefore, the total duration and number of Np waveforms were calculated until the last E2 event, excluding prolonged final non-probing after feeding cessation.

The only event that changed significantly before disease symptom development was the pd event, the duration of which was significantly shortened on N. benthamiana plants infected with PSV-P + satRNA-P compared to mock- and PSV-P-inoculated plants (Additional File 2—Table S7). After symptom development, those feeding on PSV + satRNA-P-inoculated N. benthamiana were the least active (with statistically significant differences detected between insects placed on PSV-G + satRNA-P- or PSV-P + satRNA-P- and mock-inoculated plants) (Fig. 4a). The directions of these changes also concerned the number of Np waveforms (Fig. 4b). After symptom development, the analysis of the time to the 1 st probe showed a reduction during infection with PSV (without satRNA-P) and an elongation during infection with PSV + satRNA-P compared with healthy plants (Fig. 4a). The total duration of the ABC waveform (probing of non-phloem tissues) was significantly shorter on plants infected with PSV, and even shorter on those inoculated with PSV + satRNA-P- (regardless of the virus strain), compared to mock-inoculated N. benthamiana. A similar pattern was observed for the number and duration of pds, which were also significantly lower on PSV-infected plants and further reduced on those co-inoculated with PSV and satRNA-P (except for PSV-G + satRNA-P, on which pd duration was almost the same as on PSV-G). Additionally, the total duration of the E1 and E2 waveforms was significantly shortened in plants infected with PSV + satRNA-P compared to healthy plants. The Variant × Symptoms interaction was significant for pd duration and number in the experiments with both PSV strains (Additional File 2—Table S8), confirming that the effect of satRNA on pd-related feeding behaviour was contingent on symptom development. Similarly, a significant interaction was detected for Np number for both PSV strains, indicating that increased non-probing activity on satRNA-co-infected plants emerged only after symptom onset. These data indicate that the presence of satRNA-P in the PSV inoculum consistently exacerbates palatability loss in infected plants, irrespective of virus strain.

Fig. 4.

Probing and feeding behaviour of Myzus persicae on Nicotiana benthamiana plants infected with peanut stunt virus (PSV) strains (strain G or strain P) with or without satRNA-P, assessed by Electrical Penetration Graph (EPG) monitoring after symptom development (27–28 dpi). Heatmaps show the total duration (a, in minutes or seconds) and number of events (b) for each EPG waveform, representing distinct phases of aphid stylet activity within plant tissues. Sixteen plant–aphid pairs were analysed per treatment (one aphid per plant; n = 16). Values represent mean ± SEM (Standard Error of the Mean). The colour intensity reflects row-wise min–max scaled values of the mean duration or number of events for each EPG waveform across treatments: the deepest red indicates the highest value and the lightest shade the lowest value within each row, allowing visual comparison of relative differences across treatments. Statistically significant pairwise differences between treatments (p-value < 0.05), determined using Generalised Linear Models, are shown in bold and underlined. Detailed p-values for all pairwise comparisons are provided in Additional File 2—Table S9 (for ‘time’ parameter) and Table S10 (for ‘number’ parameter). Waveform definitions: pd – potential drops, Np – non-probing, ABC – penetration of peripheral tissues (mesophyll), E1 – accessing phloem elements and secretion of saliva, E2 – ingestion of phloem sap, min – minutes, sec – seconds. Treatment abbreviations: M – mock-inoculated (healthy control), G – PSV-G, Gs – PSV-G + satRNA-P, P – PSV-P, Ps – PSV-P + satRNA-P

The data relating to the feeding behaviour of aphids on N. benthamiana plants inoculated with CMV- or CMV + satRNA differ. There were no statistically significant differences in the total duration or the number of each EPG waveform between treatments before the development of infection symptoms (Additional File 2—Table S11). An exception was observed for pd duration, which was significantly longer on plants infected with CMV + nc-satRNA compared to those infected with CMV alone, whereas pd number was higher on plants infected with CMV + non-nc-satRNA than on those inoculated with the virus alone. However, after symptom development, the analysis of the time to the 1 st probe revealed that infection with CMV alone led to a shortening of this parameter, whereas the presence of nc-satRNA in the virus inoculum caused its extension (with statistically significant differences at p-value 0.0203) (Fig. 5a). Moreover, the aphids feeding on plants infected with CMV + non-nc-satRNA and CMV + nc-satRNA exhibited significantly longer total durations and higher numbers of Np waveforms compared to mock-inoculated plants (Fig. 5). Significant differences were observed for the duration and number of pds, which were reduced on N. benthamiana infected with CMV + satRNAs, with shorter pd duration on CMV + nc-satRNA-inoculated plants. The Variant × Symptoms interaction was significant for pd duration and number in both CMV + non-nc-satRNA and CMV + nc-satRNA comparisons (Additional File 2—Table S12), as well as for Np number and ABC number, confirming that these effects emerged after symptom development rather than being present throughout the experiment. Similarly, the total durations of the E1 waveform were significantly shorter on CMV + satRNA-inoculated plants when compared to mock-inoculated plants. Taken together, these data show that aphids reached the first probe more quickly on plants infected with the virus alone than on mock‑inoculated plants, whereas the presence of the highly symptom-exacerbating nc-satRNA in the inoculum significantly delayed the time to first probe compared with CMV alone. In addition, on CMV + satRNA‑infected plants, aphids spent more time in the non‑probing (Np) phase and showed reduced duration of phloem‑related E1 waveform compared with mock‑inoculated or CMV‑infected plants, consistent with reduced sustained feeding on these plants.

Fig. 5.

Probing and feeding behaviour of Myzus persicae on Nicotiana benthamiana plants infected with cucumber mosaic virus (CMV) with or without non-necrogenic satRNA (non-nc-satRNA) or necrogenic satRNA (nc-satRNA) assessed by Electrical Penetration Graph (EPG) monitoring after symptom development (27–28 dpi). Heatmaps show the total duration (a, in minutes or seconds) and number of events (b) for each EPG waveform, representing distinct phases of aphid stylet activity within plant tissues. Sixteen plant–aphid pairs were analysed per treatment (one aphid per plant; n = 16). Values represent mean ± SEM (Standard Error of the Mean). The colour intensity reflects row-wise min–max scaled values of the mean duration or number of events for each EPG waveform across treatments: the deepest red indicates the highest value and the lightest shade the lowest value within each row, allowing visual comparison of relative differences across treatments. Statistically significant pairwise differences between treatments (p-value < 0.05), determined using Generalised Linear Models, are shown in bold and underlined. Detailed p-values for all pairwise comparisons are provided in Additional File 2—Table S13 (for ‘time’ parameter) and Table S14 (for ‘number’ parameter). Waveform definitions: pd – potential drops, Np – non-probing, ABC – penetration of peripheral tissues (mesophyll), E1 – accessing phloem elements and secretion of saliva, E2 – ingestion of phloem sap, min – minutes, sec – seconds. Treatment abbreviations: M – mock-inoculated (healthy control), C – CMV, Cnonnc – CMV + non-nc-satRNA, Cnc – CMV + nc-satRNA

Regarding the feeding behaviour of M. persicae on S. lycopersicum inoculated with CMV or CMV + satRNAs, there were no statistically significant differences in the total duration or number of each EPG waveform between treatments before the development of virus infection symptoms (Additional File 2—Table S15). An exception was observed for duration and number of pds, which were significantly higher on plants infected with CMV + nc-satRNA compared to mock-inoculated and CMV-infected plants. After symptom development, the aphids were least active (the longest duration of Np waveform and its higher number) on plants infected with CMV + nc-satRNA, followed by CMV + non-nc-satRNA, compared to mock-inoculated plants (Fig. 6). The most pronounced treatment effects after symptom development were observed for pd duration and number. Relative to mock-inoculated plants, CMV infection increased both the mean duration and the number of pd events. Co‑infection with satRNAs further enhanced these parameters: aphids on CMV + non‑nc‑satRNA plants showed longer pd duration and more frequent pd events, and the strongest increase occurred on CMV + nc‑satRNA plants. The Variant × Symptoms interaction was significant for pd duration and number in both CMV + non-nc-satRNA and CMV + nc-satRNA comparisons (Additional File 2—Table S16), as well as for Np number, confirming that the effects of satRNA on pd-related feeding behaviour and non-probing activity were contingent on symptom development. Moreover, the greatest reduction in the total duration of the E2 waveform was observed for M. persicae feeding on S. lycopersicum infected with CMV + non-nc-satRNA and CMV + nc-satRNA, compared to healthy and CMV-infected plants. This reduction in the total duration of phloem waveforms (E1 and E2) during the monitoring period resulted from the absence of these phases in several individuals (Additional File 2—Table S19). All aphids engaged in the E1 waveform on CMV-infected plants, while one individual did not proceed to the E2 waveform. In contrast, co-infection with CMV and satRNAs reduced the engagement of M. persicae in the phloem phase, with a higher number of individuals failing to proceed to phloem sap ingestion on CMV + nc-satRNA-infected plants (Additional File 2—Table S19). This latter result indicates a statistically significant (p-value 0.0163) reduction in the number of M. persicae individuals proceeding to the E2 waveform. The obtained data suggest that satRNAs in co-infection with CMV lead to substantial deterioration of host plant palatability, which is enhanced by the presence of highly symptom-exacerbating nc-satRNA.

Fig. 6.

Probing and feeding behaviour of Myzus persicae on Solanum lycopersicum plants infected with cucumber mosaic virus (CMV) with or without non-necrogenic satRNA (non-nc-satRNA) or necrogenic satRNA (nc-satRNA) assessed by Electrical Penetration Graph (EPG) monitoring after symptom development (27–28 dpi). Heatmaps show the total duration (a, in minutes or seconds) and number of events (b) for each EPG waveform, representing distinct phases of aphid stylet activity within plant tissues. Sixteen plant–aphid pairs were analysed per treatment (one aphid per plant; n = 16). Values represent mean ± SEM (Standard Error of the Mean). The colour intensity reflects row-wise min–max scaled values of the mean duration or number of events for each EPG waveform across treatments: the deepest red indicates the highest value and the lightest shade the lowest value within each row, allowing visual comparison of relative differences across treatments. Statistically significant pairwise differences between treatments (p-value < 0.05), determined using Generalised Linear Models, are shown in bold and underlined. Detailed p-values for all pairwise comparisons are provided in Additional File 2—Table S17 (for ‘time’ parameter) and Table S18 (for ‘number’ parameter). Waveform definitions: pd – potential drops, Np – non-probing, ABC – penetration of peripheral tissues (mesophyll), E1 – accessing phloem elements and secretion of saliva, E2 – ingestion of phloem sap, min – minutes, sec – seconds. Treatment abbreviations: M – mock-inoculated (healthy control), C – CMV, Cnonnc – CMV + non-nc-satRNA, Cnc – CMV + nc-satRNA

SatRNA presence modified the efficiency of virus acquisition by M. persicae

To verify if the acquisition of virus by aphid stylets is altered by the presence of satRNA in cucumovirus inoculum, the virus accumulation analysis in aphid individuals and leaves on which the insects fed was performed. To compare virus accumulation between source leaves infected with virus or virus + satRNA and virus accumulation in aphids that foraged on those leaves, the ratio of viral genomic RNA copy number for virus alone versus virus in the presence of satRNA was calculated. The analysis of genomic RNAs of PSV in source leaves showed a trend where the accumulation of PSV was mainly higher in the virus-only-infected leaves of N. benthamiana compared to the PSV + satRNA-P-inoculated ones in the case of both PSV strains, except for the PSV-G RNA1 copy number, which was higher in leaves infected with PSV-G + satRNA-P (Fig. 7a). The mean copy number of PSV-G genomic strands in aphid stylets tended to be higher in aphids that fed on PSV-G-inoculated leaves than those feeding on PSV-G + satRNA-P infected N. benthamiana in the case of all three genomic strands (which was inferred from the calculation of ratios of viral genomic RNA copy numbers for virus alone versus virus in the presence of satRNA in plants and aphids). This showed higher efficiency of PSV-G acquisition than PSV-G + satRNA-P, although the differences in PSV-G accumulation in aphids were not statistically significant. Nevertheless, the accumulation of PSV-G in aphid stylets was 1.9-fold for RNA1, 1.7-fold for RNA2, and 1.8-fold for RNA3 higher than that of PSV-G + satRNA-P (when the viral copy number ratio in aphid stylets was compared to the ratio in source leaves). However, when the virus accumulation was measured in the aphid stylets probing on leaves inoculated with PSV-P and PSV-P + satRNA-P, it resulted in statistically significantly higher accumulation of RNA1 and RNA2 in individuals feeding on PSV-P + satRNA-P infected leaves. Here, the PSV-P + satRNA-P to PSV-P accumulation ratio in aphids was threefold higher than the corresponding ratio in the source leaves, suggesting more effective PSV-P acquisition in the presence of satRNA.

Fig. 7.

Cucumovirus genomic RNA strands and satRNA copy numbers in source leaves and aphids following virus acquisition, quantified by qPCR (source leaves) and ddPCR (aphids). Bar charts show mean copy numbers (log₁₀ scale) ± SEM for each viral genomic RNA segment (RNA1, RNA2, RNA3) and satRNA in source plant leaves (left panels) and in individual Myzus persicae aphids (right panels) after a defined acquisition access period. Data are presented for: aNicotiana benthamiana plants infected with peanut stunt virus (PSV) strain G (PSV-G) or PSV-G + satRNA-P (upper charts, blue), and PSV strain P (PSV-P) or PSV-P + satRNA-P (lower charts, red); bN. benthamiana plants infected with cucumber mosaic virus (CMV) alone, CMV + non-necrogenic satRNA (CMV + non-nc-satRNA), or CMV + necrogenic satRNA (CMV + nc-satRNA); cSolanum lycopersicum cv. Betalux plants infected with CMV alone, CMV + non-necrogenic satRNA (CMV + non-nc-satRNA), or CMV + necrogenic satRNA (CMV + nc-satRNA). Inset tables show the ratio of viral RNA copy numbers between virus-only and virus + satRNA treatments (e.g., G/Gs = PSV-G/PSV-G + satRNA-P), indicating the fold-difference in viral accumulation attributable to satRNA co-infection. Values above 1 indicate lower virus accumulation in the presence of satRNA, and values below 1 indicate higher virus accumulation. Statistical significance between groups was assessed using the Mann–Whitney test (for PSV data, the comparison of two groups: PSV and PSV + satRNA-P) and the Kruskal–Wallis test (for CMV data, the comparison of three groups: CMV, CMV + non-nc-satRNAs, and CMV + nc-satRNAs). Asterisks indicate statistically significant differences * – p-value < 0.05). Treatment abbreviations: PSV-G – PSV strain G (G), PSV-G + satRNA-P – PSV strain G with satellite RNA P (satRNA-P) (Gs), PSV-P – PSV strain P (P), PSV-P + satRNA-P – PSV strain P with satRNA P (Ps), CMV + non-nc-satRNAs – CMV (C) and non-necrogenic satRNA (Cnonnc), CMV + nc-satRNA – CMV and necrogenic satRNA (Cnc)

In the case of satRNAs' influence on CMV acquisition by M. persicae from N. benthamiana plants, viral genomic RNAs accumulation in the source leaves was significantly higher in plants inoculated with CMV than CMV + non-nc-satRNA or CMV + nc-satRNA (Fig. 7b). Additionally, the accumulation of non-nc-satRNA was significantly higher than that of nc-satRNA. However, the analysis of CMV RNAs accumulation ratios in M. persicae stylets showed that the CMV + non-nc-satRNA was acquired less efficiently (1.8-fold for RNA1 and 1.6-fold for RNA3) compared to the CMV RNAs accumulation ratio in source leaves. On the other hand, the analysis of CMV RNAs accumulation ratios in aphids feeding on CMV + nc-satRNA showed a 6.6-fold increase for RNA1, 12.6-fold for RNA2, and 7.2-fold for RNA3 in the acquisition of CMV by aphids, compared to their accumulation ratios in source leaves. More efficient acquisition of CMV in the presence of nc-satRNA than non-nc-satRNA was confirmed by statistically higher accumulation of nc-satRNA than non-nc-satRNA in aphid stylets (opposite to their relation in source leaves).

In S. lycopersicum plants infected with CMV, it was observed that the presence of both satRNAs reduced the accumulation of the virus in the source leaves and reduced the efficiency of virus acquisition by aphids (inferred from the comparison between CMV RNAs copy number ratios from leaves and aphid stylets) (Fig. 7c). In aphids that fed on CMV-infected plants, the accumulation ratio of CMV RNA1, RNA2, and RNA3 in stylets was lower by 2.4-fold, 1.3-fold, and 3.0-fold, respectively, in the presence of non-nc-satRNA, and 2.0-fold, 1.2-fold, 2.2-fold, respectively, in the presence of nc-satRNA.

In most cases, the presence of satRNA (regardless of the virus species, strain, and type of satRNA) resulted in a decrease in the level of genomic RNAs of the helper virus in infected plants. In turn, satRNA that exacerbated symptoms (in PSV-G-co-infected and CMV + non-nc-satRNA-infected N. benthamiana, or CMV-co-infected S. lycopersicum) reduced acquisition of the virus by aphids, while symptom-attenuating satRNA (in PSV-P-co-infected N. benthamiana) enhanced it. The exception was observed in N. benthamiana plants infected with CMV and highly symptom-exacerbating nc-satRNA, in which CMV acquisition efficiency was higher than in plants infected with CMV alone.

SatRNA significantly reduced virus transmission efficiency in S. lycopersicum plants

The aim of the next experiment was to verify whether satRNA has an influence on the transmission of cucumovirus to healthy plants by aphids. The transmission efficiency was determined by calculating the percentage of infected plants and the level of accumulation of cucumoviruses and satRNAs in these plants.

The analysis of the PSV transmission showed no statistically significant differences in the number of infected N. benthamiana plants after transmission between plants infected with PSV and PSV + satRNA-P, regardless of virus strain (Table 1). Moreover, there are no statistically significant differences in virus accumulation caused by satRNA-P in the plants after PSV-G transmission (Fig. 8a). However, in recipient plants after PSV-P and PSV-P + satRNA-P transmission, the accumulation level of viral RNA1 was higher in PSV-P + satRNA-P infected plants than in plants infected with the virus alone (Table 1, Fig. 8b). The analysis also showed that satRNA co-transmission with the helper virus was lowered compared to the transmission of the virus alone (Table 1).

Table 1.

Overall efficiency of virus transmission and satRNAs co-transmission from cucumovirus-infected source plants to healthy test plants by Myzus persicae across three independent experiments. Virus transmission efficiency (%) represents the proportion of test plants in which viral infection was confirmed by RT-qPCR. satRNA co-transmission efficiency (%) indicates the proportion of virus-positive test plants in which satRNA was additionally detected (conditional co-transmission rate). p-values refer to differences in virus transmission efficiency between virus-only and virus + satRNA treatments, assessed using Fisher's Exact Test; asterisks indicate statistical significance (* – p-value < 0.05). n/a = not applicable as the virus was not accompanied by satRNA. Transmission efficiencies from each independent experiment are provided in Additional File 2—Table S20

Host plant	Treatment	Efficiency [%]		p-value
		Virus transmission	satRNA co-transmission
N. benthamiana	PSV-G	88.2	n/a	-
	PSV-G + satRNA-P	84.8	76.5	0.7337
	PSV-P	58.8	n/a	-
	PSV-P + satRNA-P	58.8	52.9	1.000
	CMV	78.8	n/a	-
	CMV + non-nc-satRNA	76.5	47.1	1.000
	CMV + nc-satRNA	91.2	73.5	0.1863
S. lycopersicum	CMV	100.0	n/a	-
	CMV + non-nc-satRNA	82.3	47.1	0.02457 *
	CMV + nc-satRNA	50.0	35.3	1.038e−06 *

Fig. 8.

Cucumovirus genomic RNA strands (RNA1 and RNA3) and satRNA accumulation in test plants following virus transmission by Myzus persicae, quantified by RT-qPCR across three independent experiments. Boxplots show log₁₀-transformed copy numbers of viral genomic RNA segments (RNA1 and RNA3) and satRNA detected in test plants after aphid-mediated transmission. Each data point represents an individual test plant; the red dot indicates the mean and the horizontal line indicates the median. Data are shown for: a Nicotiana benthamiana plants infected with peanut stunt virus strain G (PSV-G) or PSV-G with satRNA-P (PSV-G + satRNA-P); b N. benthamiana plants infected with PSV strain P (PSV-P) or PSV-P with satRNA-P (PSV-P + satRNA-P); c N. benthamiana plants infected with cucumber mosaic virus (CMV) alone, CMV + non-necrogenic satRNA (CMV + non-nc-satRNA), or CMV + necrogenic satRNA (CMV + nc-satRNA); d Solanum lycopersicum cv. Betalux plants infected with CMV alone, CMV + non-necrogenic satRNA (CMV + non-nc-satRNA), or CMV + necrogenic satRNA (CMV + nc-satRNA). Statistical analysis was performed in two steps: within each experiment, accumulation levels between the virus-only and virus + satRNA groups were compared using the non-parametric Mann–Whitney test; the resulting p-values were then combined across the three experiments using Fisher’s method meta-analysis to assess the overall effect. The combined p-value < 0.05, indicating statistically significant differences across experiments, is marked with *. The accumulation values and p-values are provided in Additional File 2—Table S20

The analysis of transmission efficiency of CMV in the presence or absence of satRNAs to N. benthamiana plants showed that, despite a lack of significant differences in the number of infected plants (Table 1), the accumulation of RNA1 and RNA3 in those plants was significantly higher in CMV-infected plants than in those with CMV in co-infection with non-nc-satRNA or nc-satRNA (Fig. 8c). Moreover, the co-transmission of both satRNAs was not fully efficient (Table 1).

In the case of the analysis of CMV transmission to S. lycopersicum plants, the efficiency was significantly downregulated by the presence of both non-nc-satRNA and nc-satRNA (Table 1). Additionally, the accumulation level of CMV RNA1 and RNA3 was significantly lowered when CMV was in co-infection with non-nc-satRNA and nc-satRNA (Fig. 8d). Here, CMV transmission efficiency coincided with the symptom severity caused by satRNA presence – the more severe the infection symptoms, the more CMV and satRNA transmission was reduced. Moreover, the co-transmission of both satRNAs was lowered compared to the transmission of the virus alone (Table 1).

Discussion

The evolutionary success of plant viruses is largely due to their efficient transmission by vectors. Previous research indicated that virus-infected plants can emit signals that attract insects, which then settle on plants, acquire viral particles, and spread them to subsequent hosts [2]. Interestingly, while these signals are generally attractive to insects, in the case of aphids that transmit viruses in a non-persistent manner, the host plant palatability usually deteriorates [2, 6, 29]. After tasting the infected plant (thus acquiring the virus via the stylet in a very short time), the aphid moves to probe other plants, thereby transferring the virus [3, 29, 30]. The presence of satRNAs in co-infection with their helper viruses is a potential contributor to changes in the efficiency of virus transmission through insects. For instance, studies have shown that the presence of Y-sat during CMV infection, despite lower virus levels and a change in symptom expression on CMV-infected plants, does not affect olfactory attractiveness for aphids [15]. However, more aphids that fed on CMV + Y-satRNA-co-infected plants displayed winged morphs than those that fed on CMV-only infected plants. The presence of such satRNA, therefore, enables the spread of the virus in the environment over longer distances.

In this study, overall, the results from the experiments on aphid orientation behaviour were consistent with our hypothesis. The analyses showed that symptom-exacerbating satRNAs reduced the attractiveness of infected host plants to M. persicae compared to plants infected with the helper virus alone, under light conditions. This effect was observed in N. benthamiana (PSV-G + satRNA, CMV + nc-satRNA) and in S. lycopersicum (CMV + nc-satRNA). Thus, adding symptom-exacerbating satRNA to the viral inoculum appears to deter aphid infestation, may accelerate the death of the plant, and thus could shorten the window of aphid attraction and subsequent virus transmission to the healthy plants. Moreover, statistically significant differences and a trend in preference of aphids toward N. benthamiana and S. lycopersicum plants infected with CMV over mock-inoculated plants were detected under light conditions. This observation may suggest an involvement of visual cues in shaping aphid orientation behaviour as stated by Jayasinghe et al. [15], when it was shown that yellow leaf colour caused by the presence of Y-satRNA in CMV inoculum was more attractive to aphids. However, there were no statistically significant differences between the preferences of aphids for CMV- or mock-inoculated S. lycopersicum in darkness, similarly to other olfactometry assays done by Arinaitwe et al. [31].

In the case of symptom-attenuating satRNAs, the presence of satRNA in PSV-P inoculum naturally occurring in the environment attracted a higher number of aphids in darkness despite mitigating symptoms. In contrast, there was no difference in aphid preference towards CMV + non-nc-satRNA-infected N. benthamiana and S. lycopersicum plants under both dark and light conditions. Likewise, no satRNA-dependent changes in the aphid's orientation behaviour were reported for tobacco plants infected with CMV + Y-satRNA [15]. Together, these observations suggest that the effects of satRNA on vector behaviour are virus species-dependent. In some pathosystems, satRNA may contribute to helper virus maintenance by altering vector responses, whereas in others its effect appears neutral, with little impact on plant condition or insect behaviour. However, it is important to note that several pairwise contrasts showed borderline significance (0.05 ≤ p-value < 0.10), which may reflect limited statistical power arising from the low proportion of aphids making a host plant choice per replicate. Notably, the effects were consistently in the same direction across independent experiments. For example, nc-satRNA consistently reduced aphid preferences toward CMV + nc-satRNA-infected plants regardless of the host plant, while non-nc-satRNA had no effect in any experiment. This consistency provides stronger support for the biological conclusions than individual p-values alone. Future studies employing bioassay designs that generate higher choice rates or include larger numbers of biological replicates would increase power to detect subtle interaction effects.

Acquisition time of non-persistently transmitted viruses via stylet penetration of epidermal cells (probing) of the infected leaves requires a very short time. This strategy enables efficient uptake of viral particles [29]. Infection with non-persistently transmitted viruses can increase the attractiveness of host plants to aphids and encourage leaf probing (corresponding to potential drops occurring during ABC waveform) [1, 2]. However, the palatability of the infected plants is reported to be deteriorated and aphid feeding time shortens (corresponding to the E1 and E2 waveforms), causing insects to leave the plants, which in consequence may promote faster viral spread in the environment [1, 2, 6].

In the analyses involving N. benthamiana plants, following symptom development, both PSV and CMV infections caused elongation of the non-probing phase and shortening of the phloem phases. This tendency was more pronounced and statistically significant when plants were infected with virus and satRNAs. SatRNAs, in co-infections with both PSV (regardless of virus strain) and CMV, reduced the total duration and frequency of the ABC waveform (and pds appearing during this phase) and the duration of the phloem phase (E1 and E2 waveforms). Moreover, the prolonged time to the 1 st probe on virus and satRNA-inoculated plants suggests that these plants are less readily accepted for feeding by aphids. It can be concluded that satRNA co-infection with CMV and PSV shortens the duration of aphid feeding activity (total durations of ABC, E1, E2 waveforms) regardless of symptom severity (exacerbation in PSV-G + satRNA-P- and CMV + nc-satRNA-, and alleviation in PSV-P + satRNA-P- and CMV + non-nc-satRNA-infected plants). Moreover, we also show a trend of decreased total time of the E2 waveform, or no engagement in the phloem phase, in S. lycopersicum in the presence of CMV and satRNAs.. This phenomenon may promote earlier host departure and faster host switching. In the context of non-persistent transmission, such changes could favour more rapid transmission during brief feeding periods [3]. A change in aphid feeding behaviour associated with nc-satRNAs was previously observed in the case of A. gossypii during CMV acquisition on tomato plants, where aphids required more time before initiating probing on CMV + nc-satRNA plants than in CMV- and CMV + non-nc-satRNA-inoculated plants [16]. These data support a pattern similar to the’attract and deter’ manipulation described in orientation behaviour analyses in other studies [2, 29]. It may imply that PSV (and satRNA) attract aphids via olfactory cues (as in the case of PSV-P + satRNA) while repelling them by gustatory cues, thereby promoting spread to adjacent plants.

Our data showed that satRNA presence during CMV infection on tomato led to a higher number of pds, while on N. benthamiana lower. These differences may reflect satRNA-induced alterations in host plant palatability at the cellular level, detectable by M. persicae during intracellular stylet punctures. CMV infection is known to constitutively upregulate salicylic acid (SA) signalling while simultaneously suppressing jasmonate (JA)-dependent defences, a process mediated by the CMV 2b protein [32]. Since JA signalling is the primary pathway mediating aphid resistance in CMV-infected plants, its modulation by satRNA co-infection may underlie the differences in pd parameters observed between plant species. This interpretation is consistent with previous studies reporting differences in the expression of genes associated with defence responses to insects in virus and satRNA co-infected plants [11, 12, 33–35]. Ethylene may further contribute to these changes as a modulator of JA-mediated defences, potentially influencing cell composition and callose deposition detectable during pd [6]. It was also found that CMV infection in A. thaliana induced an aphid feeding-deterrent (4-methoxy-indol-3-yl-methylglucosinolate) [30].

Another important facet of virus survival in the environment is the acquisition of virus particles by aphids and their subsequent transmission to healthy plants. The efficiency of virus acquisition depends on whether the virus is present alone or in co-infection with satRNA, and this relation is shaped by the interplay between satRNA variant, helper virus strain, and host plant species. In the case of PSV, acquisition efficiency diverged between strains in a manner inconsistent with the virus accumulation. PSV-G showed a tendency toward less efficient acquisition in the presence of symptom-exacerbating satRNA despite higher virus accumulation, whereas PSV-P was more efficiently acquired when associated with symptom-attenuating satRNA despite lower virus titre. Importantly, the number of pds waveforms did not differ between the two PSV strains in the presence of satRNA-P, indicating that the divergence in acquisition cannot be attributed to differences in aphid probing behaviour. This dissociation between pd frequency, virus accumulation, and acquisition efficiency suggests that satRNA-P may modulate the accessibility of viral particles at epidermal puncture sites independently of total virus accumulation. Notably, PSV transmission efficiency was not changed by satRNA presence in the virus inoculum for both strains, indicating that the satRNA-induced changes in acquisition do not translate into proportional changes in transmission outcomes under the conditions tested. However, PSV accumulation in the test plants was higher in the presence of satRNA.

A similar but host-dependent pattern was observed for CMV. In N. benthamiana, highly efficient acquisition of CMV was observed in the presence of nc-satRNA relative to CMV alone, consistent with previously reported data on CMV and Y-sat, where the aphid stylet was proposed as a limiting element in acquiring viral particles [15]. In contrast, acquisition of CMV from N. benthamiana in the presence of non-nc-satRNA and from S. lycopersicum in the presence of both satRNA types was lower than from plants infected only with the virus. These results are consistent with previous data on CMV transmission in the presence of non-nc- and nc-satRNAs [17]. The effect of nc-satRNA on acquisition efficiency was host-dependent: nc-satRNA reduced acquisition from S. lycopersicum, whereas it enhanced it from N. benthamiana despite comparable virus accumulations in source plants. Notably, this divergence in acquisition was accompanied by opposing pd responses. Nc-satRNA presence increased pd frequency on S. lycopersicum but reduced it on N. benthamiana, indicating that neither probing frequency nor virus accumulation alone can account for the observed differences in acquisition efficiency. This host-dependent reversal of both pd behaviour and acquisition outcomes may reflect different cellular responses to nc-satRNA co-infection. In S. lycopersicum, nc-satRNA, which was shown to trigger programmed cell death and activate defence responses [36, 37], may compromise the integrity of epidermal cells at acquisition sites, reducing virus accessibility despite increased probing frequency. In N. benthamiana, in contrast, satRNA-induced changes in mesophyll cell composition may cause aphids to abandon individual cells more rapidly (fewer pds and shorter phloem phase) while preserving or enhancing virus accessibility at puncture sites, potentially due to the high susceptibility of this host to CMV infection. The mechanism underlying enhanced acquisition in N. benthamiana in the presence of nc-satRNA remains unclear. N. benthamiana differs from S. lycopersicum in its response to viral infection, including distinct sRNA profiles [38] and the absence of nc-satRNA-induced necrosis, which may collectively preserve tissue integrity at acquisition sites. However, the precise molecular basis of this host-specific facilitation of acquisition requires further investigation.

Acquisition efficiency did not predict transmission outcomes, indicating that additional factors modulate the CMV–satRNA–host–vector interaction. In S. lycopersicum, the reduced acquisition observed in the presence of both satRNA types was consistent with a significant decrease in CMV transmission efficiency. Moreover, the reduced accumulation of CMV RNA1 and RNA3 in recipient plants, together with the incomplete (partial) co-transmission of satRNAs, indicates that satRNAs may interfere not only with virus uptake but also with successful delivery and/or establishment in new hosts. Moreover, CMV transmission efficiency coincided with symptom severity caused by satRNA presence – the more severe the infection symptoms, the more CMV and satRNA transmission was reduced. This is further supported by data showing that necrogenic satRNAs can invade CMV populations only at high aphid densities, due to their decreased transmission rates relative to CMV alone or CMV associated with attenuative satRNA [39]. A similar final outcome was observed in N. benthamiana, where CMV acquisition was enhanced in the presence of nc-satRNA, but did not result in increased accumulation in plants after transmission, and no significant differences in the proportion of infected recipient plants were observed. Instead, CMV accumulation in recipient plants was reduced when co-transmitted with satRNAs, suggesting that post-acquisition processes, such as inoculation efficiency or early infection establishment, may be negatively affected by the presence of satRNA. Overall, these results indicate that satRNAs influence the spread of CMV primarily at the level of transmission success, not only acquisition, and that the overall epidemiological outcome is highly dependent on the host species and the severity of symptoms induced by the virus-satRNA co-infection.

Conclusions

Our data provide for the first time a collective analysis of the influence of symptom modulation direction by satRNA of cucumoviruses on the orientation, probing, and feeding behaviour of M. persicae, and cucumovirus acquisition and transmission efficiency. For feeding to occur, the aphid must be encouraged to move onto the infected plant, and this phase is crucial for the final transmission outcome. Here, we have shown that satRNAs that strongly exacerbate viral disease symptoms, by the example of PSV-G + satRNA-P and CMV + nc-satRNA, contributed to the reduced olfactory and feeding attractiveness of the host plants to M. persicae, as well as virus accumulation after transmission. There are differences in the insect feeding behaviour caused by the presence of symptom-modifying satRNAs, but they do not directly translate into altered virus acquisition and transmission efficiencies, although they may reflect strong cellular changes in the host plant. Moreover, despite the high CMV acquisition efficiency from N. benthamiana observed in the presence of nc-satRNA, viral accumulation in recipient plants after transmission was substantially reduced compared to CMV alone. On the other hand, the satRNA variants naturally associated with helper viruses, that alleviate disease symptoms and reduce helper virus accumulation level, as in the case of PSV-P + satRNA-P, are beneficial for both the virus and the host plant, as the plants become more attractive to the insect vectors. By preserving host plant in better conditions and maintaining the plant’s accessibility to aphid infestation, they ensure the continued helper virus replication, facilitating its acquisition and transmission to other plants. This conclusion is in accordance with the statement that satRNAs are considered a kind of buffering factor between the plant and the virus [8]. Under specific conditions, satRNAs are thought to stabilise the pathogenesis, which provides the survival of the virus and host plant in the ecosystem. By contrast, when satRNA leads to symptom exacerbation, it lowers the probability of virus transmission to the healthy plants. This may explain, to some extent, why such symptom-exacerbating satRNAs occur less frequently in the environment than those that attenuate symptoms. This effect might cause selection against satRNA variants that induce phenotypes that negatively impact vector-host interactions [40].

Authors’ contributions

BWK: Conceptualization, Methodology, Investigation, Data Curation, Formal Analysis, Data Analysis, Funding Acquisition, Project Administration, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing. PF: Investigation, Data Curation, Formal Analysis, Data Analysis, Writing – Review & Editing. MB: Investigation, Data Curation, Writing – Original Draft Preparation, Writing – Review & Editing. PS: Investigation, Data Curation, Writing – Original Draft Preparation, Writing – Review & Editing. AOS: Conceptualization, Methodology, Data Analysis, Writing – Review & Editing, Supervision. All authors read and approved the manuscript.

Funding

This work was supported by National Science Centre (Poland) Grant no. 2018/29/N/NZ9/02467.

Data availability

All data supporting the findings of this study are available within the paper and its Supplementary Information. The raw datasets and scripts used in this study are publicly available in the Zenodo repository at https://doi.org/10.5281/zenodo.20696898.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.