Sipha maydis sensitivity to defences of Lolium multiflorum and its endophytic fungus Epichloë occultans

Plant and aphid stocks

We worked with the annual grass plants of Lolium multiflorum both symbiotic (E+) and non-symbiotic (E-) with its common fungal endophyte E. occultans (Moon et al., 2000). More than a decade ago, seeds of L. multiflorum with high percentages of endophyte infection were hand-collected from a naturalised population in the Pampean grassland (Argentina) (36°00′S, 61°5′W) (Gundel et al., 2009). Immediately after collection, a proportion of these seeds were treated with the systemic fungicide Triadimenol (150 g kg⁻¹; Baytan®) in order to obtain endophyte-free individuals. Since then, plants of these two biotypes (i.e., E+ and E-) have been cultivated annually in a common garden (recall that E. occultans is strictly vertically transmitted), multiplying fresh seeds for experimentation [IFEVA - CONICET, Universidad de Buenos Aires, Argentina (34°35′S, 58°28′W)]. Genetic segregation between plant biotypes has been prevented by allowing individual plants to freely exchange pollen during flowering (Gundel et al., 2012). Each late spring–early summer, ripe seeds produced by each plant biotype are harvested and evaluated for endophyte presence; after confirming the level of endophytes in each biotype, the seeds are stored in a 4 °C freezer. The endophyte detection is carried out by looking for fungal hypha in stained individual seeds following the “seed squash technique” (Bacon & White, 1994; Card et al., 2011). For this, 100 seeds from each seed lot (E+ and E-), are incubated in NaOH (5%), stained with rose bengal and examined under a light microscope at 40X power. The seeds produced in 2014 were examined for endophyte infection frequency (E+: 99%, and E-: 1%; n = 100 each) and stored in cold and dry conditions until use (2015).

In early-spring 2015, individual aphids S. maydis (Passerini) were collected from the local extant vegetation dominated by cereals and wild grasses. Starting with approximately 150 apterous adults, an aphid population was established within a growth chamber [21 °C (±1) constant, radiation 150 µmol m⁻² s⁻¹, and photoperiod L16:D8 h] on wheat plants (Var. Cronox; Don Mario). Wheat plants were replaced periodically to provide fresh material for the aphid population. After 6 weeks, the aphid population was large enough to provide the required number of individual adult aphids for the experiment (see next section).

Experimental description

In 2015 during the normal growing season for L. multiflorum (autumn-winter-spring), 50 E+ and 50 E- plants were grown in 1.5 L pots, filled with a mix of soil, sand, and peat in equal proportions. Plants were watered to field capacity, as needed, to avoid water deficits. In early-spring, 28 E+ and 28 E- healthy plants were selected and transferred to a growth chamber with the same environmental conditions as described earlier. At that time, the plants averaged 48 tillers (range: 25–73) and were starting the reproductive stage (spike appearance). After careful examination to ensure there were no invertebrates present on the L. multiflorum plants, each plant was individually enclosed within a white cotton voile fabric bag supported by a tubular plastic net. Before the application of the hormone treatment (see below), the plants were acclimated to the growth chamber conditions for one week.

The experiment comprised a 2 × 2 full factorial design, with endophyte (E+, E-) and salicylic acid (SA+, SA-) as the experimental treatments. Fourteen plants from each biotype were sprayed with 10 ml of SA solution (0.5 mM; Biopack®, Argentina). The same procedure was done on the other 14 E+ and 14 E- plants but sprayed with 10 ml of distilled water. Three days later, each plant was challenged with 5 apterous adult aphids. This number of aphids as starting population was chosen based on field observations, where the colonization of L. multiflorum plants by S. maydis aphids is usually carried out by only few individuals (Chaneton & Omacini, 2007). The 3-days period between the application of SA and the aphid challenge was previously identified as enough time for the plants to develop a defence response prior to contact with aphids (Bastías et al., 2018a).

The aphid populations on the L. multiflorum plants developed over the next 24 days. We counted the number of insects on each plant (aphid population size) at days 7 and 14 (corresponding to days 10 and 17 since the SA application, respectively). At day 24 (27th since SA application), a group of approximately 35 randomly chosen aphids were sampled from a subset of 5 individual plants per treatment to measure the mass-specific standard metabolic rate (SMR) by means of open-flow respirometry.

Two serial samples of plant tissues were taken to measure the physiological concentrations of plant defence hormones and fungal alkaloids. The tissues were sampled from a subset of 8 plants per treatment, selected at random. The first sample was taken on day 3, just prior to the introduction of the aphids. Two leaf-blades were excised from one tiller per plant just before the aphid challenge (day 3 post SA application) to evaluate the concentrations of SA and JA hormones. Since the recognized role of JA-signalling pathway responses in plant defences (Thaler, Humphrey & Whiteman, 2012), JA concentration levels were also measured in response to the endophyte presence and the SA treatment. The second harvest comprised the removal of one tiller per E+ plant 7 days after the aphid challenge (10 days post SA application). We used the pseudostem from this tiller to evaluate the concentration of loline alkaloids [note that the fungal endophyte E. occultans only produces this type of alkaloid (Bastias et al., 2017a; Bastías et al., 2017b)]. We selected tillers in visibly good conditions but with symptoms of aphids feeding activities. Even though E- plants are incapable of producing loline alkaloids, E- were subjected to the same sampling procedure as E+ plants to avoid any manipulation-dependent effects (Cahill Jr, Castelli & Casper, 2002). We estimated that the total tissue removed for hormone and alkaloid assessments represented around 1% of the aboveground plant biomass. At day 27 post SA application, the total aboveground biomass of the plants was harvested, and immediately dried in an oven (2 d at 60 °C) to evaluate the individual plant dry weight (Analytical scale, ± 0.01 g, Mettler Toledo).

Quantification of SA and JA hormones

Starting from freeze-dried and ground leaf material, subsamples of 50–100 mg were extracted with 100% Acetonitrile containing 100 ng of d6-SA and d5-JA as internal standards. The extracts were dried, derivatized with N-Methyl-N-(trimethylsilyl)trifluoroacetamide, and injected into an Agilent DB-5MS column (30 m, 0.25 mm inner diameter, 0.25 µm film thickness with a 10 m guard column). The column effluent was added into the ion source of a Scion TQ GC-MS/MS (Bruker Daltonics Inc.). The mass spectrometer was operated in positive ionization mode with multiple reactions monitoring (MRM) as previously described in Bastías et al. (2018a). SA and JA hormones were quantified relative to the peak area of their corresponding internal standards.

Quantification of loline alkaloids

Lolines were extracted from 50 mg of freeze-dried, ground plant samples using a solution of 40% methanol/5% ammonia and 1,2-dichloroethane containing 54.8 ng mL⁻¹ 4-phenylmorpholine as internal standard. Plant extracts were centrifugated, and supernatants transferred to glass GC vials via a 20 µm filter for analysis. The analysis was conducted using a GC flame ionization detector (GC2010Plus, Shimadzu Corporation, Japan) and separation was achieved on a ZB-5 capillary column (30 m × 0.32 mm × 0.25 µm film). More information can be found in Bastías et al. (2018a). The detection limit was 25 µg g⁻¹ DW.

Measurements of standard metabolic rate (SMR)

The SMR quantifies the energy budget required for insects to maintain homeostasis, thus this variable represents a measure of the general physiological status of insects (Nespolo, Roff & Fairbairn, 2008). Aphid maximum annual growth rates (r_m) are negatively correlated with SMR. Aphids with high metabolic rates (i.e., high levels of maintenance costs) might have less energy available for reproduction thus negatively impacting their population sizes (Castañeda, Figueroa & Nespolo, 2010). The SMR was measured by the production of CO₂ (VCO₂) of aphid groups placed within of an open-flow respirometry system (LI-6400; Li-Cor, Lincoln, USA). We estimated the mass-specific SMR that is, the amount of CO₂ produced per mass unit of aphid per hour (i.e., µL of VCO₂ per mg of aphid per hour). Aphids were obtained from a subset of 5 randomly chosen plants per treatment. From each of these plants, 35 adult and non-winged aphids were carefully removed and placed in Eppendorf tubes. We considered each group of aphids from one individual plant to be a replicate (i.e., 5 replicates per treatment). All the aphid groups were weighted (±0.01 mg, analytical balance, Mettler Toledo), and kept for one hour without food before the metabolic measurements. The VCO₂ of each aphid group was registered every second during a period of 10 min at 24 °C (±0.5). For this, the insects were placed in 10 mL-metabolic chamber that received CO₂-scrubbed air at a constant rate of 70 mL min⁻¹ and connected to a sensor of CO₂ (LI-6400; Li-Cor, Lincoln, USA). The mass-specific SMR was obtained averaging the 2-mins continuous and most stable VCO₂ values (from the 10-mins register) and dividing this averaged value by the insect group weight. Aphids were discarded after the SMR measurements.

Statistical analyses

The effects of the plant symbiotic status and SA application on the concentrations of SA and JA hormones, and on the plant above-ground biomass were analysed separately with linear effects models, using the function gls from the nlme package in R software (R Core Team, 2013), and assuming independent, identically distributed normal random errors (Pinheiro et al., 2009). The models included the plant’s symbiotic status (E+, E-) and SA treatment (SA+, SA-) as categorical factors. VarIdent variance structures were used on the SA treatments to accommodate deviations in the variance homogeneity in the SA and JA concentrations response variables (Zuur et al., 2009). After this procedure, all the ANOVA assumptions were met.

Similarly, the effects of the SA treatment on the concentrations of alkaloids (total lolines, NFL, and NANL) were analysed separately with lineal effect models, using the same software package mentioned earlier, and assuming independent, identically distributed normal random errors. The models included the SA treatment (SA+, SA-) as a categorical factor. When required, we used the function VarIdent on the SA treatment to accommodate deviations in the variance homogeneity (Zuur et al., 2009). ANOVA assumptions were then met.

The effects of plant symbiotic status and SA treatment on the population size of aphids (number of individuals) were analysed with linear mixed-effects models using the package glmmADMB in R software, and assuming that random errors were distributed independently and following a negative binomial distribution (Fournier et al., 2012). The model included plant symbiotic status (E+, E-), SA treatment (SA+, SA-), and time (7d and 14d since the aphid challenge) as categorical factors, and the random effect included the time nested in pot. Temporal autocorrelation between the repeated measurements was not observed.

The effects of the plant symbiotic status and the SA treatment on mass-specific SMR of aphids were analysed with linear effects models using the same R package and assuming the same random error distribution that for hormone concentration variables. The model included the plant symbiotic status (E+, E-) and SA treatment (SA+, SA-) as categorical factors. All the ANOVA assumptions were met. We performed post-hoc analyses between treatments when significant interactions were detected using the package lsmeans in R (Lenth, 2016). All the presented values in the result section are means ± standard errors (S.E.M). All data obtained in the present study are available in Table S1.

Effects of plant endophyte presence and SA on hormone levels and plant growth

The concentration of the hormones SA and JA within the plants responded differentially to the endophyte presence and the exogenous application of the SA. The plant SA concentration was independently affected by the plant endophyte status and the hormonal treatment (Table 1). The presence of the endophyte in the plant reduced the SA by ca. 10% [E-: 1114.00 ± 264.50, and E+: 1014.00 ± 248.00 (ng SA g⁻¹ DW)]. In addition, both plant biotypes (i.e., endophyte-symbiotic and non-symbiotic) showed increased SA concentrations of about 19-fold 3 days after the exogenous application of the hormone (Fig. 1A). Although the interaction effect between the treatments was not significant (Table 1), the reduction in SA concentration due to the endophyte presence was much more evident in plants not exposed to SA (the mean SA concentration difference between E+ and E- plants in the SA- treatment was around 22%) (Fig. 1A). The concentration of JA was not significantly modified by either the plant endophyte status or by the SA treatment (Fig. 1B) (Table 1).

The above-ground plant tissues at the end of the aphid challenge (day 27 from the application of SA) was 11% higher in endophyte-symbiotic than in non-symbiotic plants (E-: 4.42 ± 0.28 g, and E+: 4.99 ± 0.39 g). This effect was, however, not affected by the treatment with SA (Table 1).

Effects of SA on fungal loline concentrations

Ten days after plant exposure to the SA hormone, the concentration of loline alkaloids (total and the derivatives NFL and NANL) in endophyte-symbiotic plants did not vary among SA-treated and SA-untreated plants (Fig. 2) (Table 1).

Effects of plant endophyte presence and SA on S. maydis populations

The aphid population size on L. multiflorum plants increased over time but this increase was independent of both the endophyte presence or the plant exposure to salicylic acid (Table 2 and Fig. 3). On average, the aphid population size increased 2.60 fold in 7 days (from days 10 to 17 since plants exposure to SA, 5.96 ± 0.63 and 15.91 ± 1.74, respectively) (Fig. 3).

Effects of plant endophyte presence and SA on S. maydis metabolic rate

The aphid mass-specific SMR, evaluated at day 24 since the insects were placed on the plants, was unaffected by the endophytic fungus, the SA hormone, or the interaction between them (Table 2).