Effects of an Alphasatellite on the Life Cycle of the Nanovirus Faba Bean Necrotic Yellows Virus

ABSTRACT

Nanoviruses are plant viruses with a multipartite single-stranded DNA (ssDNA) genome. Alphasatellites are commonly associated with nanovirus infections, but their putative impact on their helper viruses is unknown. In this study, we investigated the role of subterranean clover stunt alphasatellite 1 (here named SCSA 1) on various important traits of Faba bean necrotic yellows virus (FBNYV) in its host plant Vicia faba and aphid vector Acyrthosiphon pisum, including disease symptoms, viral accumulation, and viral transmission. The results indicate that SCSA 1 does not affect the severity of symptoms nor overall FBNYV accumulation in V. faba, but it does change the relative amounts of its different genomic segments. Moreover, the association of SCSA 1 with FBNYV increases the rate of plant-to-plant transmission by a process seemingly unrelated to the simple increase of viral accumulation in the vector. These results represent the first study on the impact of an alphasatellite on the biology of its helper nanovirus. They suggest that SCSA 1 may benefit FBNYV, but the genericity of this conclusion is discussed and questioned.

IMPORTANCE Alphasatellites are circular single-stranded DNA molecules frequently found in association with natural isolates of nanoviruses and some geminiviruses, the two ssDNA plant-infecting virus families. While the implications of alphasatellite presence in geminivirus infections are relatively well documented, comparable studies on alphasatellites associated with nanoviruses are not available. Here, we confirm that subterranean clover stunt alphasatellite 1 affects different traits of its helper nanovirus, Faba bean necrotic yellows virus, both in the host plant and aphid vector. We show that the frequencies of the virus segments change in the presence of alphasatellite, in both the plant and the vector. We also confirm that although within-plant virus load and symptoms are not affected by alphasatellite, the presence of alphasatellite decreases within-aphid virus load but significantly increases virus transmission rate, and thus it may confer a possible evolutionary advantage for the helper virus.

INTRODUCTION

Faba bean necrotic yellows virus, a member species of the genus Nanovirus (family Nanoviridae), is a multipartite virus with eight circular single-stranded DNA segments, each about 1 kb in size and individually packaged in small (T = 1) isometric viral particles. Each segment contains a single open reading frame (ORF) and a noncoding sequence with two conserved regions known as the common region stem-loop (CR-SL) and the major common region (CR-M). The genome organization of nanoviruses corresponds to a one-segment/one-gene situation in which the different segments encode a cell cycle link protein (protein Clink; segment C), a movement protein (MP; M), a nuclear shuttle protein (NSP; N), a master-replication initiator protein (M-Rep; R), a capsid protein (CP; S) and three proteins with unknown functions (U1, U2, U4) (1, 2). Recently, it has been shown that the various segments of another nanovirus, Faba bean necrotic stunt virus (FBNSV), reproducibly accumulate in host plants at different relative frequencies, a phenomenon described as the genome formula. The genome formula has been shown to be host-specific and thus is assumed to be dependent on the environment in which the virus accumulates (3, 4), and similar observations have been reported for other unrelated multipartite viruses (5, 6).

In addition to the bona fide genome components, nanoviruses are known to be frequently associated with alphasatellites, which are self-replicating ssDNA molecules (about 1 kb) encapsidated in the nanoviral CP (7). Alphasatellite DNAs lack both CR-SL and CR-M and encode a replication initiator protein (alpha-Rep). Unlike the M-Rep proteins encoded by the nanoviral R segments, each alpha-Rep is unable to transreplicate nanovirus genome components or other alphasatellite DNAs (7). In fact, the replication-initiating activity of these alpha-Rep proteins is strictly restricted to their cognate satellite DNA component (8). Alphasatellites are dependent on their helper viruses not only for encapsidation but also for intra- and interhost movement (7).

Beyond nanoviruses, alphasatellites have been reported associated with geminiviruses, mostly with species of the genus Begomovirus (9–11), and with one species of the genus Mastrevirus (12). Recently, the alphasatellites of geminiviruses and nanoviruses have been classified as two subfamilies under the names Geminialphasatellitinae and Nanoalphasatellitinae, respectively, in the newly established family Alphasatellitidae (7). Despite a rapid increase in the number of alphasatellite species described, their actual role or impact on the different stages of their helper nanoviruses’ or geminiviruses’ cycles remains obscure. The fact that these extra components are encapsidated by the helper virus and transmitted together with it has raised many questions. The implications of the presence of alphasatellites with geminiviruses are relatively well documented, but available reports show a highly variable outcome. The Geminialphasatellitinae subfamily comprises genera related to both the Old World (OW) and New World (NW) geminiviruses (7). Based on the genus in which the alphasatellite is classified, its effect on the helper virus appears to differ; the OW geminialphasatellites classified in the genus Colecusatellite have no effect on their helper virus infection (13); however, the ones classified in the genus Ageyesisatellite seem to attenuate the symptoms (14). In contrast, NW geminalphasatellites in the genus Clecrusatellite increase symptoms but negatively impact transmission efficiency (15). Comparable studies on alphasatellites associated with nanoviruses are not available. Although not formally quantified, one study has reported the reduced efficiency of artificial co-inoculation of infectious FBNYV and “Rep11 DNA” satellite clones, as well as slightly attenuated symptoms induced on the corresponding Vicia faba host plants (16).

In the present paper, we further investigated the impact of the same Rep11 DNA alphasatellite, which has recently been classified as the species subterranean clover stunt alphasatellite 1 (SCSA 1) in the genus Subclovsatellite (7), on major biological traits of FBNYV in V. faba (broad bean). We chose the particular combination of FBNYV/alphasatellite, SCSA 1, for this study, because SCSA 1 is the most abundant satellite DNA found associated with FBNYV isolates in nature (8, 16). We first characterized the genome formula of FBNYV in its host plant V. faba and its aphid vector Acyrtosiphon pisum. We then analyzed the effect of SCSA 1 on FBNYV accumulation and genome formula in plants and insects, and finally also estimated its impact on symptom severity and transmission efficiency.

RESULTS

FBNYV genome formula in the presence/absence of SCSA 1 in faba bean plants.

Three weeks after aphid inoculation, the relative frequency of each FBNYV genomic segment was estimated by qPCR in 38 FBNYV-infected plants containing all eight segments but not SCSA 1. The median copy number of each segment was rounded to the nearest integer. The FBNYV genome formula was calculated as described for FBNSV by Sicard et al. (3) and was found to be 14^C 5^M 1^N 2^R 4^S 1^U1 4^U2 20^U4. Comparison between FBNYV and FBNSV (a series of FBNSV-infected plants analyzed as in [3]) showed that the genome formula of these nanoviruses were clearly distinct when infecting the same variety of V. faba (Fig. 1) under similar experimental conditions.

FIG 1.

Radar plot showing the genome formula of the FBNYV (red) and FBNSV (blue) in faba bean. The median frequency of each segment is reported on each axis of the radar plot. Colored dashes represent standard deviations.

Similarly, 3 weeks after aphid inoculation, the relative frequency of each FBNYV genomic segments was estimated by qPCR in 30 FBNYV infected plants containing all eight segments plus SCSA 1. In the presence of SCSA 1, the FBNYV formula shifted to 24^C 25^M 1^N 16^R 8^S 4^U1 11^U2 74^U4. Comparison of the FBNYV genome formula with and without the alphasatellite is shown in Fig. 2 (the formula without SCSA 1 represents the same data as in Fig. 1); here, the frequency of SCSA 1 is not used to calculate the relative frequency of the other segments to allow for direct comparison. However, it should be noted that, when present, the frequency of SCSA 1 over total viral DNA is rather high, ranging from 12.2% to 31.5% with a median value of 18.15%.

FIG 2.

FBNYV genome formula, in the presence (red) or absence (blue) of the SCSA 1 in faba bean plants, the median frequency of each segment is reported on each axis of the radar plot. Colored dashes represent standard deviations.

Our statistical analysis showed that the presence of SCSA 1 significantly modifies the FBNYV genome formula (interaction between segment: presence of SCSA 1, likelihood ratio test (LRT) = 38.221, P = 2.75e-06, model R² = 0.752).

Separate analyses were implemented to identify which segments were significantly impacted by the presence of the alphasatellite (Table 1). The results demonstrated that the relative abundances of the C, M, N, R, S, and U4 segments were significantly altered in the presence of SCSA 1, whereas no significant effect was detected on the relative abundances of U1 and U2.

TABLE 1.

Statistical tests on the effect of the presence/absence of SCSA 1 on relative frequencies of FBNYV genomic segments

Segment	F value	Pr(>F)a	Significancea
C	15.648	0.0005	***
M	6.9358	0.01682	*
N	14.07	0.00075	***
R	95.876	1.4E−13	***
S	26.623	9.8E−06	***
U1	0.9511	0.38057
U2	0.0739	0.7865
U4	4.7452	0.04395	*

^aThe Benjamini-Hochberg correction for multiple tests was applied on all P values. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Since we observed an effect of the presence of SCSA 1 on segment relative frequencies, we tried to go further and understand whether the frequency of SCSA 1 was correlated with that of FBNYV segments. To do so, we calculated a Spearman correlation matrix which determined that the frequency of SCSA 1 correlated positively with that of the C and S segments, and negatively with that of the M, R, and U2 segments (Fig. 3). No significance relationships were observed between the N, U1, and U4 frequencies and that of SCSA 1.

FIG 3.

Correlation matrix showing the correlation between the frequency of all FBNYV segments and SCSA 1. The intensity of the color as well as the shape of the ellipses depends on Spearman’s rank correlation coefficients, which are shown on the second half of the figure. P values (shown on top of ellipses) were calculated (and corrected with the Bonferroni-Holm multiple test correction) only for the correlations of interest, i.e., those involving SCSA 1.

Viral accumulation and symptoms in faba bean plants in the presence of SCSA 1.

To investigate the effect of the presence of SCSA 1 on the FBNYV accumulation level, the viral load in the 38 FBNYV-infected faba bean plants was compared to the viral load in the 30 plants coinfected with FBNYV and SCSA 1. The t test results showed that the presence of SCSA 1 had no significant effect on FBNYV accumulation in V. faba (t = −1.3142, degrees of freedom [df] = 30.477, P value = 0.1986). Further analyses indicated a strong correlation between the accumulation of SCSA 1 and that of the virus (Fig. 4. Spearman correlation test: ρ = 0.844, S = 702, P value = 7.674e-07), but no significant correlation between the relative frequency of SCSA 1 (relative to that of the other segments) and the accumulation level of FBNYV (Spearman correlation test: ρ = −0.048, S = 4712, P value = 0.8).

FIG 4.

Correlation between the accumulation of viral DNA and alphasatellite SCSA 1 in faba bean. The total accumulation of viral DNA is plotted against that of SCSA 1; qSCSA 1 = 0.134× qViral-DNA + 0.45, R² = 0.95.

In terms of symptoms, we used two indicators of the effect of viral infection on plant phenotype in our analysis: the height of individual plants, and the number of leaf levels they developed until the virus stopped their growth. Measurements were performed after 3 weeks, at which point the symptoms had fully developed in all infected plants, i.e., infected plants had stopped growing due to viral infection. Yellowing and necrosis symptoms in faba bean plants were very severe and indistinguishable between both treatments (FBNYV and FBNYV/SCSA 1), and they appeared after 4 weeks postinoculation in both cases. We noted no additional symptoms when SCSA 1 was present. Table 2 shows that the average height of faba bean plants infected with FBNYV + SCSA 1 (31.5 cm) was not significantly different from that of plants infected with FBNYV alone (33.7 cm). In fact, plant height could be more susceptible to viral accumulation (marginally significant) than to the presence of the alphasatellite.

TABLE 2.

Statistical analysis of host plant height^a

Variables	dfb	Sum Sq	Mean Sq	F value	Pr(>F)
SCSA 1 presence/absence	1	83.88	83.882	1.8965	0.17319
Viral accumulation	1	141.97	141.971	3.2099	0.07785
Residuals	65	2874.9	44.229

^aLinear model in which plant height was explained by viral load and the presence or absence of SCSA 1.

^bdf, degrees of freedom; Sum sq, sum of the squares; mean sq, mean of the squares; Pr(>F), F-test P value.

The average number of leaf levels that developed before the viral infection stopped the plant growth was not significantly different in the presence or absence of SCSA 1 (7.17 versus 7.63 leaf levels on average in the presence or absence of SCSA 1; linear model, F = 3.569, P = 0.063).

Frequencies of FBNYV segments in A. pisum in relation to the presence/absence of SCSA 1.

Given the importance of the aphid vector in the life cycle of FBNYV, we investigated the impact of the alphasatellite on FBNYV traits within its major vector species A. pisum (17). The FBNYV genome formula in aphid bodies was compared in the presence/absence of SCSA 1. For this, 23 and 24 pools of 5 aphids were collected from 2 sets of 3 source plants infected with or without the alphasatellite, respectively (see Materials and Methods), and analyzed by qPCR. First, we ran a global analysis considering both the effects of the presence/absence of SCSA 1 and those of the plants or aphids. This global analysis revealed that FBNYV segment frequencies varied when passing from plants to aphids, independent of the presence or absence of the alphasatellite (LTR = 65.187; P = 4.432e-11).

The next question was then whether these plant-to-aphid variations of the FBNYV genome formula are different when the alphasatellite is present. Statistical analysis of the two genome formulas observed in the presence (23^C 20^M 1^N 19^R 6^S 18^U1 12^U2 32^U4) and absence (37^C 24^M 1^N 7^R 13^S 15^U1 17^U2 24^U4) of SCSA 1 in the aphids showed that the effect of the alphasatellite on the FBNYV genome formula is significant (interaction between segment and alphasatellite presence, likelihood ratio LTR = 105.71, P value < 2.2e-16, model R² = 0.924) (Fig. 5). To identify the segments most significantly impacted by SCSA 1 during the acquisition by aphids, eight separate statistical analyses were performed (one per segment) and, in each analysis, the logit frequency of the segment was modeled as a function of the presence of the alphasatellite (Table 3). Comparison of these results with the segment frequencies determined for the faba bean source plants (Table 1) indicated that the plant-to-aphid frequency change of segments N, R and S was significantly altered by the presence of SCSA 1, with N and S being decreased and R being augmented.

FIG 5.

FBNYV genome formula, in the presence (red) or the absence (blue) of SCSA 1 in aphid vectors. The median frequency of each segment is reported on each axis of the radar plot. Colored dashes represent standard deviation.

TABLE 3.

Statistical test of the effect of SCSA 1 on the plant-to-aphid FBNYV formula changes

Segment	F value	Pr(>F)a	Significancea
C	3.8201	0.1137
M	2.8799	0.1546
N	18.332	0.0004	***
R	1.15E + 01	0.0039	**
S	29.319	1.83E−05	***
U1	1.85	0.2408
U2	0.8456	0.4145
U4	0.0184	0.8926

^aThe Benjamini-Hochberg correction for multiple tests was applied on all P values. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Viral accumulation in A. pisum in the presence of SCSA 1.

The viral DNA accumulation level was determined from the same qPCR data set as in the previous section. The amounts of the eight FBNYV DNAs were calculated and normalized by dividing them by the amount of DNA of the actin reference gene in each sample. Next, the mean values of viral DNAs found for the two treatments were compared. The results showed that the level of virus accumulation in the aphids was three times higher in the absence of SCSA 1 (0.0938 versus 0.0347), and that this effect was highly significant (LTR = 65.187, P value = 4.432e-11). Notably, this difference in viral accumulation within aphids cannot be explained by a different viral load in the source plants (the viral load of each source plant is visible in Fig. 6).

FIG 6.

Aphid transmission rate of FBNYV in the presence (red) or absence (blue) of SCSA 1. The x axis indicates the viral load in infected source plants, as estimated by qPCR. The y axis shows the transmission rate, measured as the proportion of infected plants. In each of these 6 independent experiments, 15 plants were inoculated, for a total of 90 plants. Error bars represent standard deviation.

Effect of SCSA 1 on the transmission rate of FBNYV by A. pisum.

In parallel, we also questioned whether the potential lower viral accumulation in aphid vectors in the presence of SCSA 1 would correlate to a distinct transmission rate. Using the same source plants (with and without SCSA 1) and the same aphid cohorts, we performed transmission tests as detailed in Materials and Methods. Surprisingly, the results of the Generalized Linear Model analysis showed that at comparable virus loads in the source plants, FBNYV transmission rate by aphids was higher in the presence of SCSA 1 (Fig. 6., resid. dev.= 10.634, P = 0.008273).

DISCUSSION

Gene Copy Number (GCN) variation has been identified in many organisms as a mechanism which affects gene expression (3). The concept of genome formula was first introduced for FBNSV by Sicard et al., (2013), and later confirmed for other unrelated multipartite viruses (5, 6). This phenomenon represents one of the potential benefits of the multipartite organization of viral genomes, as it would allow the regulation of the expression of different viral genes by adjusting their relative copy number (18, 19). Previous studies have shown that the FBNSV genome formula is host-dependent, likely adjusting/adapting the virus to distinct environments (3, 4). In the present study, the genome formula was calculated for a second nanovirus, FBNYV, and compared to that of FBNSV. It is remarkable that in the same host (V. faba, var. Sevilla), the genome formulae of these two nanoviruses are drastically different. Considering that FBNYV and FBNSV are closely related (20, 21), with nearly 75% overall identity in their genome sequences, and that they infect common plant hosts and use common aphid vectors (22, 23), such divergent genome formulae are intriguing and indicate that small sequence divergence might have considerable effects.

One hypothesis regarding the widespread presence of alphasatellites in the infections by many virus species in the families Geminiviridae and Nanoviridae (16, 24) is that alphasatellites confer a possible evolutionary advantage by assisting their helper viruses; for example, by raising the viral accumulation level, thereby changing the virus transmission rate. An opposing hypothesis is that alphasatellites compete with the helper nanovirus for replication, encapsidation, and transmission (16), illustrating a parasitic relationship. To argue which hypothesis better explains the alphasatellite-nanovirus relationship, we investigated the effect of SCSA 1 on various FBNYV traits. Statistical comparison of the FBNYV genome formula in faba bean showed that the genome formula is modified, not drastically but nevertheless significantly, when SCSA 1 is present. The process underlying this change is obscure thus far. One possibility is that the alphasatellite differentially competes with distinct viral segments for replication and/or encapsidation, so that the abundance of some segments would be altered and that of others would not. Alternatively, the presence of the alphasatellite could modify the environment, and the virus would adapt to the new conditions by accordingly adjusting its genome formula. One more possibility, although it contrasts with the current view on nanoviruses/alphasatellite replication, is that the alpha-Rep protein encoded by the alphasatellite can initiate the cross-replication of some viral DNAs, thereby changing their relative abundance in the host plant. We noted a significant negative correlation between the frequency of the R segment and that of SCSA 1, both of which encode Rep proteins. Given the similarity of the predicted functions of these two proteins (25), the negative correlation between their respective encoding segments is noteworthy. Although this result is interesting, further experiments are needed to definitively confirm or refute the capacity of SCSA1 expression protein to replicate at least some of the helper nanovirus segments.

Changes in the relative proportions of FBNSV genome segments between the sap of infected faba beans and aphids fed on these plants have been described previously (4). Replication of virus genome inside aphid bodies, differential decapsidation and degradation of some segments, and physicochemical differences of virus particles containing distinct segments have been discussed as potential explanations for these changes (4). Consistent with the prior report, we found that the relative frequencies of the FBNYV genome segments changed significantly within the aphids compared to their frequencies in the infected faba bean plants. Moreover, the presence of SCSA 1 affects the relative frequencies of viral genomic segments inside the aphid bodies. The various hypotheses described above also represent potential mechanisms underlying the effect of SCSA 1 on the relative frequencies of FBNYV segments inside aphid bodies, and thus the actual explanation remains undecided.

The specific effect of alphasatellites on the accumulation levels of geminiviruses or their associated betasatellites in different host plants has been extensively investigated (14, 26–28), contrasting with the situation in nanoviruses where this question is poorly documented. Our observation that the alphasatellite SCSA 1 has no effect on the helper nanovirus accumulation is similar to those of some earlier studies on geminiviruses (13, 14, 28, 29). It has been previously suggested that the genome formula may affect viral accumulation of FBNSV (3). Here, the presence of the satellite changes the FBNYV genome formula with no detectable effect on virus load. We have no definitive explanation for this observation. However, it must be considered that the results of the study on the FBNSV genome formula and its possible impact on virus load did not include any satellite, which may have interfered in some way and modified the relationship between formula and viral accumulation. Also, consistent with most studies on the geminivirus-alphasatellite association, the presence of SCSA 1 did not significantly modify the virus-induced symptoms in faba bean plants. The latter observation contradicts the results of a previous study, where the same SCSA 1 reduced the symptoms induced by the same FBNYV isolate (16). At this point, our only explanation is that our study and the earlier one used two distinct faba bean cultivars, “Sevilla” and “Condor,” respectively, suggesting that the effect of satellites on helper viruses is dependent on the environment (here, the host variety).

Vector transmission is one of the most important factors in the epidemiology of plant viruses, which depend upon vectors for survival and spread (30). The results of aphid transmission tested in the absence and presence of SCSA 1 revealed that, with comparable virus loads in the source plants, the rate of transmission is significantly higher when SCSA 1 is present. In this experiment, it is noticeable that the higher transmission rate was not due to increased virus load within aphids. The within-aphid virus load was even lower in the presence of SCSA 1, suggesting a putative negative effect of the alphasatellite on FBNYV accumulation within the body of its aphid vector; this effect appears to be compensated for by an elusive higher ability to transmit the virus or to cause successful infections after inoculation. Generally, vector transmission is one of the biological bottlenecks in the viral disease cycle that repeatedly and transiently reduces population size (31). In the case of FBNSV, the size of this bottleneck has been shown to be different for different segments of the genome, possibly related to the relative abundance of each segment in the insect body (32). Altered genome formulae in the presence of alphasatellite may change the size of transmission-related bottlenecks for each viral segment, but this more detailed analysis was beyond the scope of the present study.

In contrast to our findings, the presence of alphasatellites was linked to a reduction in the transmission rate of Euphorbia yellow mosaic virus (EuYMV) by Bemisia tabaci as reported previously (15). This could further indicate that, as is the case for symptom severity and helper virus accumulation, there is no general pattern, and that each type of alphasatellite-helper virus association exhibits unique features. Despite the possible existence of a phylogeny-related pattern of positive versus negative effects of alphasatellites on life traits of the helper virus (indicated in the introduction), we here insist on what we believe is an important consideration. The full understanding of satellite biology may never be reached at the laboratory scale, where we may merely reveal secondary (or collateral) effects not related to the actual “raison d’être” of satellites. The discrepancies found in the literature on distinct biological systems may stem from the fact that we only look at one small part of the picture. Complementary studies at the ecological level which include ecological networks of hosts, vectors, helper viruses, and satellites are now required, because they may represent the main piece of the puzzle of the biology of satellites.

MATERIALS AND METHODS

Plant, virus, and aphid.

Faba bean (Vicia faba var. “Sevilla;” Vilmorin) seeds were individually sown in pots and grown in a S2 restricted-access confinement facility with a 13/11-h day/night photoperiod, a temperature of 26/20°C day/night, and constant 70% humidity. The FBNYV infectious clone used in this study was kindly provided by Tania Timtchenko (CNRS, France). The FBNYV isolate [ES;Mu29] originated from Spain and the production of clones, each corresponding to one of the eight genome segments, is described as performed by Grigoras et. al (20). The infectious clone of alphasatellite SCSA 1 (AJ005968.1) was derived from an Egyptian isolate (8).

The Acyrthosiphon pisum colony was maintained on V. faba plants at a temperature of 23/21°C and a photoperiod of 13/11 h (day/night), ensuring reproduction through parthenogenesis. To obtain homogeneous cohorts of L3 to L4 instars, adult aphids were placed on faba bean plants and allowed to produce newborn larvae for 2 days. After removal of the adults, the progeny were allowed to grow for an additional 2 days before their use in transmission testing as described below.

Agroinoculation.

To determine the genome formula of FBNYV, faba bean seedlings were agroinoculated with the corresponding infectious clones, either with or without the presence of SCSA 1. The Rhizobium radiobacter (formerly Agrobacterium tumefaciens; COR 308 isolate) clones, each harboring a pBin19 plasmid containing a tandem repeat of one segment of FBNYV or SCSA 1 DNA (20), were grown separately for 18 h at 28°C and 150 rpm in 50 mL NZY medium (0.1% N-Z-Amine, 0.5% yeast extract, 0.5% NaCl, 12.5 mM MgCl₂, 12.5 mM MgSO4, and 0.4% glucose [pH 7.5]). These cultures were then centrifuged at 1,000 × g at 18°C for 30 min in 50-mL Falcon tubes. Each pellet was resuspended in 5 mL of MS buffer (33) containing 30 mM acetosyringone and 1 mM morpholinoethanesulfonic acid (pH 5.6) before being pooled. The mixture was kept at room temperature for 1 h, then 0.5 mL was needle-inoculated into the stem of each of the 10-day-old V. faba plants (26Gx 7/8-in needle, 1-mL syringe). Following the development of nanovirus-related symptoms (wrinkled uppermost leaves and downward rolling of the bottom leaves, appearing between 12 and 18 days postinoculation), the presence of the eight FBNYV genomic segments and eventually, of SCSA 1 DNA, was detected using specific quantitative PCR (qPCR). The FBNYV-infected plants containing all eight genomic segments with or without SCSA 1 were selected as the source plants for aphid transmission.

DNA extraction from plants.

DNA extraction from symptomatic plants was performed at 21 days postinoculation (dpi) following the protocol described earlier (34). Each plant sample consisted of four leaf disks (4-mm diameter) collected from the two upper leaf levels (two disks per leaf level). The four pooled leaf disks were ground in 400 μL extraction buffer (200 mM Tris-HCl [pH 7.5], 250 mM NaCl, 25 mM EDTA, 0.5% SDS 10%, 1% wt/vol polyvinylpyrrolidone 40 [PVP-40], 0.2% ascorbic acid) using metal beads and Fontainebleau sand, incubated at 65°C for 10 min, and then centrifuged at 10,000 × g for 10 min. An equal amount of isopropanol was added to the supernatant and centrifuged at 10,000 × g for 15 min. The pellet was finally washed with 1 mL of 70% ethanol and resuspended in 150 μL of distilled water.

Aphid transmission assays.

Because agroinoculation of the infectious clones is poorly efficient and often yields a low number of plants infected with all desired segments, we routinely use a few such plants as sources for aphid transmission; this is much more efficient and yields more fully infected plants.

To produce enough infected plants for experimental testing of the within-plants genome formula, viral accumulation, and symptom evaluation, aphid transmission of FBNYV was performed under two distinct conditions (with or without SCSA 1), using either FBNYV- or FBNYV + SCSA 1-infected V. faba source plants, in two independent experimental replicates. Three source plants per treatment were used in the first replicate and two source plants per treatment in the second. The infected source plants were obtained by agroinoculation of infectious clones, as described above, and the presence of all desired segments was confirmed using qPCR on systemically infected tissues where aphids were feeding. Approximately 100 L4 instars per source plant were allowed an acquisition access period (AAP) of 2 days. Following this AAP, for each source plant, groups of 3 aphids each were transferred to 24 healthy 9-day-old receiver V. faba plants for an inoculation access period (IAP) of two additional days. The receiver plants were finally treated with the insecticide Pirimor G (1 g/L in water; Syngenta), ensuring the elimination of all uncollected aphids.

To determine the effect of SCSA 1 on FBNYV transmission, we used three source plants per treatment (with and without SCSA 1) concurrently (to avoid a potential confounding date-effect). At the end of the AAP, aphids fed on source plants (all source plants were qPCR-tested to ascertain the presence of all inoculated segments, as indicated above) were then transferred to 15 healthy seedlings (three aphids per plant); the transmission rate for each treatment/replicate was calculated 21 days after inoculation as the number of infected plants over the total number of test plants. To estimate their viral load, aphids that fed on each of these source plants were also collected in parallel in pools of five, and were either processed immediately or stored until use at −20°C by adding 100 μL of 70% ethanol.

DNA extraction from aphids.

DNA from groups of five aphids was extracted using Pure Link Genomic DNA Minikit (Thermo Fisher Scientific, USA). Briefly, when aphids were not used immediately, ethanol was removed with a micropipette and evaporated at 55°C for 7 min. A mixture of 180 μL of genomic digestion buffer and 20 μL of proteinase K was added to each tube and the tubes were incubated at 55°C for 30 min. After grinding the aphids in the digestion buffer, the tubes were similarly incubated for an additional 30-min period followed by a centrifugation for 3 min at 10,000 × g. Then, 200 μL of genomic lysis/binding buffer and 200 μL of ethanol 90 to 100% was added to the supernatants. The mixture was then centrifuged for 1 min at 10,000 × g and washed twice using 500 μL of the washing buffer, followed by centrifugation at 10,000 × g. Finally, 60 μL of distilled water was used to elute DNA from the column. For each sample, 2 μL of the DNA solution was used in qPCR experiments.

qPCR conditions.

Quantitative PCR (qPCR) was used for the quantification of FBNYV and SCSA 1 segments. All qPCRs (40 cycles of 95°C for 10 s, 63°C for 12 s, and 72°C for 12 s) were carried out using the LightCycler FastStart DNA Master Plus SYBR Green I Kit (Roche) in a LightCycler 480 thermocycler (Roche, USA), following the manufacturer’s instructions. One exception was the reaction with the V. faba reference gene primers, which consisted of 40 cycles of 95°C for 15 s, 60°C for 30 s, and 72°C for 20 s. All primers used for the amplification of the genomic and alphasatellite segments were designed based on the sequence of FBNYV-ES;Mu29 and the Egyptian SCSA 1 isolates, using LightCycler Probe Design Software 2 (Roche, USA). The reference genes for V. faba and A. pisum were the cyp2 (Cytochrome P450) (35) and actin genes (36), respectively. All primers listed in Table 4 were used at a final concentration of 0.3 μM. Sample DNA (2 μL, without dilution) was added to the qPCR mix (5 μL of Roche 2× qPCR Mastermix, 2.7 μL of H₂O, 0.3 μL of primer mix, 8 μL total) after distribution in 384-well microtiter plates. The plasmid dilutions were initially used to establish standard curves; then, in each PCR plate, fluorescence data were normalized with the calibrated reference sample for each segment and analyzed using the LinRegPCR program. To determine overall viral accumulation, the sum of the estimated number of copies of each segment was divided by that of the host plant (cyp2) or aphid (actin) reference genes.

TABLE 4.

Identity and sequence of primers used in this study

Primer	Sequence 5′→3′	Target
FBNYV-C-137-F	TAAAGTCAGACCCATCTCCTTCAGAGCTGT	C
FBNYV-C-369-R	ATCATCGTCCTCTTCAGGCGGC
FBNYV-M-136-2F	ATGCTTGCTTGTTATGTTCCTGGGTTT	M
FBNYV-M-285-2R	AGCATCCCAATTACGTTCTCTTTCAGGAC
FBNYV-N-73-F	CCACAGCAAGAGATTATGTGCTGTGATAGT	N
FBNYV-N-240-R	ACAGACACCATCATCCTCAACCATGGATA
FBNYV-R-531-2F	GGTGTATGGCCCACAAGGTGGAGA	R
FBNYV-R-685-2R	AGCTTCGTGGAAAGTCGAAGAGCACTA
FBNYV-S-22-F	TCTGGTAAGAAAGGAAGAAGAACACCACGC	S
FBNYV-S-178-R	AACGAGCAACATCTCCCTCAGGC
FBNYV-U1-20-F	GCGATTCGTGGCTTGTTGATGAGGC	U1
FBNYV-U1-253-R	ACTGCGTGTAAACTTTCATACGACAACCC
FBNYV-U2-137-F	ACGGAAGGAGAGTACCAGCTTATCCC	U2
FBNYV-U2-292-R	CTCTATTTCTGACAGGCATACGGCTTCG
FBNYV-U4-152-2F	ATGAAGAAGAAGATCGTCCTGCTTCAGT	U4
FBNYV-U4-283-2R	GTTGCGTATCATTGATTGCCAGAACCGTTT
FBNYV-paraRep-201-F	TGCTCATTGGGAAATTGCGAAAGGAGAC	SCSA 1
FBNYV-paraRep-448-2R	GAAGCGATATCACAGCGAACGACGAACA
Cyp2-F	TGCCGATGTCACTCCCAGAA	V. faba reference gene (37)
Cyp2-R	CAGCGAACTTGGAACCGTAGA
Actin-F	CGTTACCAACTGGGACGATATG	A. pisum reference gene (36)
Actin-R	GGGTTCAATGGAGCTTCTGTTA

Statistical analysis.

All statistical tests were carried out with R software (version 3.4.2) (37). The nature and results of each statistical test are indicated in Results, in the figure legends, or in the titles of tables.

The FBNYV genome formula was calculated as described by Sicard et al. 2013 (3). Genome formulae were compared using the Linear Mixed Model (LMM) on the DNA concentrations of each segment (which were boxcox normalized). The factors “segment” and “presence/absence of SCSA” were used as explanatory variables, while a random effect was applied on the “plant replicates” factor.

To identify which segments were significantly impacted by the presence of SCSA 1, separate Linear Model analyses were implemented on each genomic segment, in which the logit frequency of the segment was modeled as a function of the presence of the alphasatellite. The same statistical test was used to identify which segments’ frequencies were altered in the aphid bodies in the presence of SCSA 1. The Benjamini-Hochberg Procedure was used in each case to limit the False Discovery Rates (FDR) due to multiple testing.

A t test was used to compare the accumulation levels of the virus in the presence and absence of SCSA 1, in both plants and the aphids’ bodies. It was also used to compare plant height and the number of leaf levels (two important symptoms considered to be proxies of disease severity) in the presence and absence of the alphasatellite.