NbALD1 mediates resistance to turnip mosaic virus by regulating the accumulation of salicylic acid and the ethylene pathway in Nicotiana benthamiana.

AGD2-LIKE DEFENCE RESPONSE PROTEIN 1 (ALD1) triggers plant defence against bacterial and fungal pathogens by regulating the salicylic acid (SA) pathway and an unknown SA-independent pathway. We now show that Nicotiana benthamiana ALD1 is involved in defence against a virus and that the ethylene pathway also participates in ALD1-mediated resistance. NbALD1 was up-regulated in plants infected with turnip mosaic virus (TuMV). Silencing of NbALD1 facilitated TuMV infection, while overexpression of NbALD1 or exogenous application of pipecolic acid (Pip), the downstream product of ALD1, enhanced resistance to TuMV. The SA content was lower in NbALD1-silenced plants and higher where NbALD1 was overexpressed or following Pip treatments. SA mediated resistance to TuMV and was required for NbALD1-mediated resistance. However, on NahG plants (in which SA cannot accumulate), Pip treatment still alleviated susceptibility to TuMV, further demonstrating the presence of an SA-independent resistance pathway. The ethylene precursor, 1-aminocyclopropanecarboxylic acid (ACC), accumulated in NbALD1-silenced plants but was reduced in plants overexpressing NbALD1 or treated with Pip. Silencing of ACS1, a key gene in the ethylene pathway, alleviated the susceptibility of NbALD1-silenced plants to TuMV, while exogenous application of ACC compromised the resistance of Pip-treated or NbALD1 transgenic plants. The results indicate that NbALD1 mediates resistance to TuMV by positively regulating the resistant SA pathway and negatively regulating the susceptible ethylene pathway.

Introduction

Salicylic acid (SA) induces an important plant defence pathway against pathogens by producing pathogenesis‐related proteins (Durrant and Dong, 2004; Fu and Dong, 2013; Gao et al., 2015; Tsuda et al., 2009; Vlot et al., 2009). Research over the last decade has shown that AGD2‐LIKE DEFENCE RESPONSE PROTEIN 1 (ALD1) triggers the basal defence response and systemic acquired resistance (SAR) against bacterial infection in plants by regulating the SA pathway (Cecchini et al., 2015; Song et al., 2004a). Compared with wild‐type plants, ald1 mutants of Arabidopsis thaliana were more susceptible to Pseudomonas syringae and had reduced SA accumulation (Song et al., 2004a), while the overexpression of ALD1 conferred resistance to the pathogen by inducing the expression of PAD4 and ICS1, key components of the SA pathway that are both essential for the response (Cecchini et al., 2015; Song et al., 2004b). FMO1, another SAR regulatory gene, is also indispensable for systemic SA accumulation and SAR (Bartsch et al., 2006; Koch et al., 2006; Mishina and Zeier, 2006). ALD1 transcript induction depends on FMO1 during the establishment of SAR (Song et al., 2004b). The ALD1‐LIKE gene of Oryza sativa (OsALD1) was shown to be involved in the resistance response of rice to the rice blast fungus (Magnaporthe oryzae) (Jung et al., 2016). Interestingly, natural oviposition by Pieris brassicae or treatment with egg extract also induced SAR and inhibited growth of P. syringae in an ALD1‐ and FMO1‐dependent manner, implicating the ALD1‐mediated pathway in a response to insects (Hilfiker et al., 2014).

ALD1 encodes an aminotransferase with multiple substrates and products in vitro. l‐lysine is used as a substrate by ALD1 to produce enaminic 2,3‐dehydropipecolic acid (2,3‐DP), and then (2,3‐DP) is converted to pipecolic acid (Pip) by SARD4 (Ding et al., 2016; Hartmann et al., 2017; Navarova et al., 2012). Infection by P. syringae pv. maculicola (Psm) ES4326 induced a significant accumulation of Pip in the wild‐type, but not in ald1 mutants. Pretreatment with Pip led to increased pathogen resistance in wild‐type plants, and exogenous Pip complemented the resistance defect of ald1 mutants (Navarova et al., 2012). Recently, it was reported that FMO1 can convert Pip to N‐hydroxypipecolic acid (NHP), which then regulates the systemic acquired resistance to pathogen infection (Chen et al., 2018; Hartmann et al., 2018). Moreover, in the pathway it was found that Pip triggers activation of MPK3 and MPK6 that regulates the transcription factor WRKY33 to promote the expression of ALD1, indicating the positive regulatory loop in the ALD1‐mediated resistance pathway (Wang et al., 2018). Bernsdorff et al. investigated the relationships between SA and Pip, and showed that SA and Pip act independently from one another, but also synergistically in the basal immunity of A. thaliana to P. syringae (Bernsdorff et al., 2016). The results suggest that Pip acts as an indispensable switch for the activation of SAR, and that SA amplifies, and is required for, Pip‐triggered responses (Bernsdorff et al., 2016; Navarova et al., 2012).

Meanwhile, exogenous Pip strongly induced resistance in the SA‐deficient ics1 mutant, and both Pip and NHP pretreatment significantly increased the resistance of sid2 plants to Psm and Hpa, although in all cases resistance was markedly lower than in wild‐type plants, indicating that an SA‐independent defence pathway probably functions simultaneously (Hartmann et al., 2018; Navarova et al., 2012). However, no such SA‐independent defence pathway has yet been identified.

The defence role of ALD1 against bacterial pathogens has been well documented, and its function against fungal and oomycete pathogens has also been reported (Bernsdorff et al., 2016; Jung et al., 2016; Navarova et al., 2012). Recent data showed that tobacco mosaic virus (TMV) infection led to increased Pip accumulation, and that Pip‐treated tobacco had smaller TMV lesions (Adam et al. 2018). We now report that the expression of Nicotiana benthamiana ALD1 (NbALD1) is induced following infection by turnip mosaic virus (TuMV) and that NbALD1 participates in plant defence against TuMV by both SA‐dependent and SA‐independent pathways simultaneously. We also show that the ethylene pathway, regulated negatively by NbALD1 and mediating the susceptibility of N. benthamiana to TuMV, functions in NbALD1 or Pip‐mediated resistance in an SA‐independent manner.

Results

Expression of NbALD1 is induced in TuMV‐infected N. benthamiana

Arabidopsis thaliana ALD1 (AtALD1, Accession No. NM_126957.2) was used as query sequence in a basic local alignment search tool (BLAST) of the Nicotiana benthamiana Genome v1.0.1 predicted cDNA database (https://solgenomics.net/tools/blast/) to identify its homolog in N. benthamiana (Sequence ID: Niben101Scf04547g02001.1; named NbALD1). NbALD1 has 1350 nucleotides in an open reading frame encoding 450 amino acids. Sequence alignment showed that NbALD1 protein had 60.9% identity to AtALD1 (Accession No. NP_565359.1) and 54.2–68.7% identities to the ALD1 or AGD2 of other plants (Fig. S1A,B). All ALD1s contained the conserved residues for the pyridoxal‐5ʹ‐phosphate (PLP)‐binding site and the malate binding site (Fig. S1B). NbALD1 was expressed at a higher level in leaves than in flowers, stems or roots (Fig. 1A). At 5 days post inoculation (dpi) with TuMV, the expression of NbALD1 in leaves of N. benthamiana was remarkably induced (Fig. 1B).

Figure 1

NbALD1 was induced by TuMV in N. benthamiana and its silencing facilitated infection by TuMV. (A) Semi‐quantitative RT‐PCR showing that NbALD1 was expressed at a higher level in leaves than in flowers, stems and roots. (B) Semi‐quantitative RT‐PCR showing that the expression of NbALD1 in leaves was remarkably induced by TuMV infection at 5 dpi. Results from three independent biological replicates are shown; Ubiquitin C (UBC) was used as the internal reference gene. (C) Leaves of plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. On NbALD1‐silenced plants (TRV:ald1), there were more fluorescent spots (infection foci) on the inoculated leaves and a larger fluorescent area on systemically infected leaves than on the controls (TRV:00). (D) Numbers of infection foci on inoculated leaves of NbALD1‐silenced (TRV:ald1) and non‐silenced (TRV:00) plants at 4 dpi. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05). (E) Northern and western blots showing the increased accumulation of TuMV RNAs and CP protein in inoculated leaves and systemic leaves from NbALD1‐silenced plants (TRV:ald1) compared to the controls (TRV:00). (F) Percentage of plants with GFP fluorescence on newly emerged leaves (i.e. systemically infected) at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate).

Silencing of NbALD1 facilitates infection by TuMV

To investigate the potential roles of NbALD1 in TuMV infection, we used tobacco rattle virus (TRV)‐induced gene silencing (VIGS) to silence NbALD1 and then inoculated TuMV onto the plants (Liu et al., 2002). For this purpose, a sequence of 300 nt from NbALD1 was inserted into TRV‐RNA2 to produce TRV:ald1, at 8 dpi the expression of NbALD1 was only 18.4% of normal levels (Fig. S2). There were no obvious differences of phenotype between control (TRV:00) and NbALD1‐silenced plants (Fig. S2). Plants treated with TRV:ald1 or TRV:00 were then mechanically inoculated with green fluorescent protein (GFP)‐labelled TuMV (TuMV‐GFP) and subsequent viral infection was monitored by detecting GFP fluorescence. At 4 dpi, there were more fluorescent spots or infection foci on the inoculated leaves of NbALD1‐silenced plants (Fig. 1C,D). Both viral RNAs and coat protein (CP) accumulated much more in the inoculated leaves of NbALD1‐silenced plants than in the TRV:00‐treated controls (Fig. 1E). From 3 dpi, GFP fluorescence began to appear on newly emerged leaves, indicating systemic infection. Initially, the NbALD1‐silenced plants had a higher incidence of systemic infection and a larger area of GFP fluorescence than the controls, indicating a faster spread of systemic infection (Fig. 1C,F). Blotting analysis showed that viral RNAs and CP accumulated much more on the newly emerged leaves of silenced plants compared to the controls (Fig. 1E). Thus, silencing of NbALD1 facilitated infection by TuMV.

Overexpression of NbALD1 confers enhanced resistance to TuMV

To further determine the role(s) played by NbALD1 in defence against TuMV, we obtained transgenic N. benthamiana expressing HA‐tagged NbALD1 driven by the cauliflower mosaic virus (CaMV) 35S promotor (Fig. S3A). Transgenic plants had a similar phenotype to wild‐type at the early stage of development, but were slightly shorter at the flowering stage (Fig. S3B). Two lines of transgenic plants (OE4 and OE6) with increased levels of HA‐tagged NbALD1 (Fig. S3C) were then inoculated with TuMV‐GFP. At 4 dpi, the number of infection foci was fewer on the inoculated leaves of transgenic plants than on the controls, and the fluorescence associated with systemic infection was also less extensive and less intense (Fig. 2A,B). Viral RNAs and CP accumulated less in both the inoculated and newly emerged leaves of transgenic plants than in control wild‐type plants (Fig. 2C) and the systemic infection spread more slowly (Fig. 2D). Thus, overexpression of NbALD1 enhanced the resistance of N. benthamiana to TuMV.

Figure 2

Overexpression of NbALD1 enhanced N. benthamiana resistance to TuMV. (A) Leaves of plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. On plants overexpressing NbALD1 (lines OE4 and OE6), there were fewer fluorescent spots (infection foci) on the inoculated leaves and a smaller fluorescent area on systemically infected leaves than on the wild‐type (WT) controls. (B) Numbers of infection foci on inoculated leaves from WT, OE4 and OE6 plants at 4 dpi. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05). (C) Northern and western blots showing that TuMV RNAs and CP protein accumulated less in both the inoculated and the newly emerged leaves of transgenic plants compared to the control WT plants. (D) Percentage of wild‐type and transgenic plants with GFP fluorescence on newly emerged leaves (i.e. systemically infected) at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate).

Exogenous application of pipecolic acid boosts N. benthamiana resistance to TuMV

It has been reported that Pip, a downstream product of ALD1, is essential for ALD1‐mediated systemic acquired resistance and basal resistance (Navarova et al., 2012). Exogenous application of Pip strongly boosted basal and specific resistance of both wild‐type and ald1 mutant plants to compatible Psm and incompatible Psm avrRpm1 (Navarova et al., 2012). Exogenous application of Pip also enhanced resistance of tobacco to the adapted P. syringae pv. tabaci (Pstb) or non‐adapted Psm (Vogel‐Adghough et al., 2013). To confirm the defence role of NbALD1 against TuMV, we treated NbALD1‐silenced N. benthamiana with 100 μM Pip for 8 days. Treated plants were then inoculated with TuMV‐GFP. At 4 dpi, there were fewer infection foci on Pip‐treated plants than on the water‐treated controls (Fig. 3A,B) and there were corresponding decreases in the accumulation of viral RNAs and CP in both inoculated and systemic leaves (Fig. 3C). Systemic infection also spread more slowly on Pip‐treated plants (Fig. 3D). Thus, Pip treatment alleviated the resistance deficiency of NbALD1‐silenced plants. Pip treatment also decreased the number of infection foci and the accumulation of TuMV on non‐silenced plants (Fig. 3E–H). Taken together, the results further confirm that NbALD1 plays a role in defence against TuMV.

Figure 3

Exogenous application of pipecolic acid (Pip) enhanced N. benthamiana resistance to TuMV. (A) Leaves of plants inoculated with TuMV‐GFP on NbALD1‐silenced plants and examined under UV light at 4 dpi. (B) Numbers of infection foci on inoculated leaves from Pip‐ and H2O‐treated NbALD1‐silenced plants at 4 dpi. (C) Northern and western blots showing that TuMV RNAs and CP protein accumulated less in both the inoculated and the newly emerged leaves of Pip‐treated plants NbALD1‐silenced plants, compared to the H2O‐treated NbALD1‐silenced plants. (D) Percentage of plants systemically infected at different times after inoculation. (E) Leaves of plants inoculated with TuMV‐GFP on non‐silenced (TRV:00) plants and examined under UV light at 4 dpi. (F) Numbers of infection foci on inoculated leaves from Pip‐ and H2O‐treated non‐silenced (TRV:00) plants at 4 dpi. (G) Northern and western blots showing that TuMV RNAs and CP protein accumulated less in both the inoculated and the newly emerged leaves of Pip‐treated plants non‐silenced (TRV:00) plants, compared to the H2O‐treated non‐silenced (TRV:00) plants. (H) Percentage of plants systemically infected at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05).

SA accumulation is regulated positively by NbALD1 and plays a role in defence against TuMV

It has been reported that the ALD1 or Pip‐induced plant defence responses to P. syringae are accompanied by increased SA (Bernsdorff et al., 2016; Navarova et al., 2012). We therefore measured the SA content of our plants where NbALD1 was either silenced or overexpressed without virus infection. Compared to non‐silenced plants, the SA content of NbALD1‐silenced plants was decreased from 3.83 to 2.46 ng/g (Fig. 4A), but increased from 5.40 to 6.47 ng/g (OE4) or 7.55 ng/g (OE6) plants in those overexpressing NbALD1 without TuMV infection (Fig. 4B). SA content in the Pip‐treated plants was also enhanced significantly (Fig. 4C). The results indicate that NbALD1 positively regulates the SA content, similar to the previous reports on ALD1‐meditated resistance to bacterial pathogens.

Figure 4

The SA accumulation was regulated positively by NbALD1 and played a role in defense against TuMV. (A) SA content of NbALD1‐silenced (TRV:ald1) plants was significantly less than that in control plants (TRV:00). (B) SA content of NbALD1 transgenic lines OE4 and OE6 was significantly greater than that in wild‐type (WT) plants. (C) SA content of 100 μM Pip‐treated WT plants was significantly greater than that in H2O‐treated control plants for 8 dpi. (D) Leaves of WT and transgenic NahG plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (E) Numbers of infection foci on inoculated leaves. (F) Northern and western blots showing accumulation of TuMV RNAs and CP protein. (G) Percentage of plants systemically infected at different times after inoculation. (H) Leaves of H2O‐ and 10 μM SA‐treated WT plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (I) Numbers of infection foci on inoculated leaves. (J) Northern and western blots showing accumulation of TuMV RNAs and CP protein. (K) Percentage of plants systemically infected at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate). Different letters on histograms indicate significant differences (P < 0.05).

We next examined whether the SA plays a role in defence to TuMV by inoculating TuMV‐GFP to NahG (which encodes a salicylate hydroxylase converting SA to catechol) transgenic N. benthamiana in which SA cannot accumulate. The NahG plants were more susceptible to virus than the controls by all measures: there were more infection foci, TuMV accumulated to greater levels on inoculated leaves (Fig. 4D–F), systemic infection spread more rapidly (Fig. 4G) and there was a greater accumulation of viral RNAs and CP (Fig. 4D–F). These results imply that the SA is involved in resistance to TuMV. As further confirmation, wild‐type plants were treated with 10 μM SA for 48 h and then inoculated with TuMV‐GFP. Compared to the controls, fewer infection foci appeared, viral RNAs and CP accumulated less and systemic infection occurred more slowly (Fig. 4H–K).

The SA is an essential, but not the only, regulated pathway involved in NbALD1‐mediated resistance to TuMV

To examine the association of the SA with NbALD1‐mediated resistance to TuMV, NahG plants were pretreated with Pip and then inoculated with TuMV‐GFP. The number of infection foci on inoculated leaves of Pip‐treated NahG plants was significantly less than on leaves of H2O‐pretreated NahG plants (Fig. 5A,B). The numbers were greater than on the corresponding wild‐type plants (Fig. 5A,B). TuMV RNAs and CP accumulated less in Pip‐treated NahG leaves than in control NahG leaves (Fig. 5C) and systemic infection on Pip‐pretreated NahG plants progressed more slowly than that on H2O‐treated NahG plants, but still faster than that on Pip‐treated wild‐type plants (Fig. 5D). Consequently, TuMV RNAs and CP accumulated less in systemic leaves of Pip‐pretreated NahG plants than in systemic leaves of H2O‐pretreated NahG plants, but more than in systemic leaves of Pip‐pretreated wild‐type plants (Fig. 5C).

Figure 5

The SA was required but was not the only regulated pathway in NbALD1‐mediated resistance against TuMV. (A) Leaves of Pip‐treated wild type (WT) and NahG transgenic plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (B) Numbers of infection foci on inoculated leaves at 4 dpi. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05). (C) Northern and western blots showing the accumulation of TuMV RNAs and CP protein in the inoculated and systemic leaves. (D) Percentage of plants systemically infected at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate).

We next investigated the role of SA on NbALD1‐silenced plants. Non‐silenced and NbALD1‐silenced plants were pretreated with water or SA and then inoculated with TuMV‐GFP. The number of infection foci on inoculated leaves of SA‐treated NbALD1‐silenced plants was significantly less than on leaves of water‐treated NbALD1‐silenced plants (Fig. S4A,B). TuMV RNAs and CP accumulated less in SA‐treated NbALD1‐silenced leaves than in water‐treated NbALD1‐silenced leaves (Fig. S4C), but not to the level found in water‐treated non‐silenced leaves. This indicates that exogenous SA partially complements the resistance deficiency of NbALD1‐silenced plants.

Thus, the absence of SA compromised the Pip‐ or NbALD1‐mediated resistance, indicating that SA was required for NbALD1 to function effectively in resistance to TuMV. However, exogenous application of SA only partially attenuated the susceptibility caused by silencing of NbALD1 and Pip treatment still alleviated the susceptibility of NahG plants to TuMV, suggesting that there might be a SA‐independent pathway also participating in the NbALD1‐mediated resistance against TuMV.

The ethylene pathway is regulated negatively by NbALD1 and mediates N. benthamiana susceptibility to TuMV

Next we investigated whether a pathway other than SA might be involved in NbALD1‐mediated resistance to TuMV. An association between ALD1 and the ethylene pathway has been reported in Arabidopsis: mutations in ALD1 suppressed edr2‐mediated plant resistance to powdery mildew, programmed cell death and ethylene‐induced senescence, while an ald1 single mutant had delayed ethylene‐induced senescence (Nie et al., 2011). We therefore examined whether the ethylene pathway was involved in NbALD1‐mediated resistance to TuMV. The expression of ethylene‐responsive transcription factor 3 (ERF3), a key gene in the ethylene pathway, was significantly increased in NbALD1‐silenced plants (Fig. 6A) and there was increased accumulation of the precursor of ethylene, 1‐aminocyclopropanecarboxylic acid (ACC) (Fig. 6B). These results indicate the up‐regulation of the ethylene pathway in NbALD1‐silenced plants. Exogenous application of 100 μM Pip reduced the expression of ERF3 and ACC content (Fig. 6C,D). In OE4 and OE6 plants overexpressing NbALD1, ERF3 was remarkably less than in wild‐type plants, and ACC content was also less (48.3 ng/g in wild‐type, 26.5 ng/g in OE4 and 26.0 ng/g in OE6) (Fig. 6E,F). These results indicate that NbALD1 is a negative regulator of the ethylene pathway.

Figure 6

The Ethylene pathway was regulated negatively by NbALD1. (A) The increased expression of NbERF3 in NbALD1‐silenced (TRV:ald1) plants. Empty TRV (TRV:00)‐treated plants were used as controls. (B) The increased ACC content in NbALD1‐silenced (TRV:ald1) plants. (C) The decreased expression of NbERF3 in plants treated with 100 μM Pip. H2O‐treated plants were used as controls. (D) The decreased ACC content in plants treated with 100 μM Pip. (E) The decreased expression of NbERF3 in NbALD1 transgenic plants (OE4 and OE6). Wild type plants were used as controls. (F) The decreased ACC content in NbALD1 transgenic plants (OE4 and OE6). Error bars show the mean ± SD of three replicates (at least 20 plants per replicate). Different letters on histograms indicate significant differences (P < 0.05).

The role of the ethylene pathway in TuMV infection was next examined. Key genes in the ethylene pathway, aminocyclopropane‐1‐carboxylic acid synthases 1(ACS1), ACC oxidase 1 (ACO1) and ethylene insensitive 2 (EIN2), were silenced by TRV‐mediated VIGS. At 14 dpi, the expression level of each gene was about 10% of normal levels (Fig. S5). TuMV‐GFP was then inoculated onto silenced plants. Compared to control leaves, there were fewer infection foci on inoculated leaves of silenced plants at 4 dpi of TuMV, and there was less accumulation of TuMV RNAs and CP (Fig. 7A–C). Systemic infection was also delayed, with corresponding decreases in the accumulation of TuMV RNAs and CP (Fig. 7C,D). These results indicate that plants were more resistant to TuMV when ethylene was suppressed. As further confirmation, we used aminoethoxyvinylglycine (AVG), an inhibitor of ethylene synthesis, to inhibit the ethylene pathway and examined its influence on TuMV infection. After spraying 10 μM AVG onto the leaves of N. benthamiana, expression of ERF3 16 h later was 72% of that in control plants sprayed with H2O (Fig. S6). After inoculation of the treated leaves with TuMV‐GFP there were fewer infection foci at 4 dpi than on the controls (Fig. 7E,G), and less accumulation of TuMV RNAs and CP (Fig. 7F). Thus, suppression of the ethylene pathway enhanced N. benthamiana resistance to TuMV.

Figure 7

The ethylene pathway mediated susceptibility of N. benthamiana to TuMV. Silencing of three key genes in the ethylene pathway (ACS1, ACO1 and EIN2) (A–D) or AVG treatment (E–G) increased resistance to TuMV while treatment with ACC increased susceptibility (H–J). (A, E and H) Leaves of plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (B, G, and J) Numbers of infection foci on inoculated leaves at 4 dpi. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05). (C, F, and I) Northern and western blots showing accumulation of TuMV RNAs and CP protein. (D) Percentage of plants systemically infected at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate).

To determine the influence of an up‐regulated ethylene pathway on TuMV infection, we sprayed 100 μM ACC on leaves and confirmed that ERF3 was significantly up‐regulated 16 h later (Fig. S7). After inoculation of the ACC‐treated leaves with TuMV‐GFP, there were more infection foci at 4 dpi and much greater accumulation of viral RNAs and CP than on H2O‐treated control leaves (Fig. 7H–J), showing that up‐regulation of the ethylene pathway made N. benthamiana more susceptible to TuMV.

Taken together, the results suggest that the ethylene pathway was regulated negatively by NbALD1 and mediated N. benthamiana susceptibility to TuMV.

The negatively regulated ethylene pathway is involved in NbALD1‐mediated resistance against TuMV

The ethylene pathway was up‐regulated in plants where NbALD1 was silenced and down‐regulated where it was overexpressed or if plants were treated with Pip (Fig. 6). To clarify the relationship between the ethylene pathway and NbALD1‐mediated resistance, we silenced NbALD1 and ACS1 simultaneously by VIGS for 8 days (Fig. S8) and then inoculated TuMV‐GFP onto silenced plants. The number of infection foci on inoculated leaves of NbALD1/ACS1‐silenced plants was about half that on NbALD1‐silenced leaves at 4 dpi (Fig. 8A and B) and there was much less accumulation of TuMV RNAs and CP (Fig. 8C). Because silencing of NbACS1 lessened the susceptibility of NbALD1‐silenced plants to TuMV, it appears that the up‐regulated ethylene pathway may contribute to the susceptibility of NbALD1‐silenced plants to TuMV infection.

Figure 8

The negatively regulated ethylene pathway was involved in NbALD1‐mediated resistance to TuMV. (A) Leaves of plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (B). Numbers of infection foci on inoculated leaves. (C) Northern and western blots showing accumulation of TuMV RNAs and CP protein. (D) Leaves of plants inoculated with TuMV‐GFP and examined under UV light at 4 dpi. (E) Northern and western blots showing accumulation of TuMV RNAs and CP protein. (F) Percentage of plants systemically infected at different times after inoculation. Error bars show the mean ± SD of three replicates (at least 20 plants per replicate); different letters on histograms indicate significant differences (P < 0.05). Silencing of ACS1 enhanced the resistance of NbALD1‐silenced plants (A‐C) and ACC treatment compromised the resistance of Pip‐treated plants (D‐G).

Finally, Pip‐treated wild‐type plants in which the ethylene pathway was down‐regulated were sprayed with 20 μM ACC and 16 h later were inoculated with TuMV‐GFP. The ACC spray treatment increased the number of infection foci at 4 dpi and there was a greater accumulation of TuMV RNAs and CP, indicating that exogenous application of ACC counteracted the resistance of Pip‐treated plants to TuMV (Fig. 8D–G).

Discussion

Published evidence from Arabidopsis and rice suggests that the conserved plant ALD1 proteins may play a role in resistance to pathogens. AtALD1 was induced by P. syringae infection in Arabidopsis, an ald1‐deficient mutant was hyper‐susceptible to the bacterium (Song et al., 2004a) and overexpression of ALD1 inhibited the pathogen (Jong Tae et al., 2004; Jung et al., 2016). The expression level of OsALD1 increased when rice was infected by the fungus M. oryzae and overexpression of OsALD1 conferred a remarkable resistance to the fungus (Jung et al., 2016). Our finding that NbALD1 in N. benthamiana was induced by TuMV infection and mediated the resistance to the virus supports the view that ALD1 proteins have a conserved broad‐spectrum defence role in plants and provides the evidence of their activity against a viral pathogen.

The SA pathway plays a key role in ALD1‐mediated resistance to P. syringae in Arabidopsis. Arabidopsis overexpressing ALD1 had increased SA content, and ald1 mutants had decreased SA accumulation and were more susceptible to pathogens (Cecchini et al., 2015; Song et al., 2004a). Similar results were obtained from rice. Overexpression of OsALD1 significantly induced the expression of pathogenesis‐related protein1 (PR1), the resistance gene responding to SA. It has recently been shown that SA modulates the accumulation of NHP (a downstream product), without affecting NHP generation. Our results found that silencing of NbALD1 resulted in the reduction of SA content, while overexpression of NbALD1 led to induction of SA in N. benthamiana (Fig. 4). SA has been reported to function in plant defence against viral pathogens. Silencing of SA biosynthetic and signalling genes in N. benthamiana plants increased susceptibility to TMV (Chivasa et al., 1997). Exogenous application of SA or high levels of endogenous SA enhanced resistance to turnip crinkle virus (TCV) (Chandra‐Shekara et al., 2004). Citrus exocortis viroid (CEVd) or tomato spotted wilt tospovirus (TSWV) infection caused early and dramatic disease phenotypes in NahG tomato plants (Lopez‐Gresa et al., 2016). Interestingly, HCPro, the RNA silencing suppressor of TuMV, interacts with AtCA1 (the homologue of SABP3 (SA‐binding protein 3)) to compromise the SA pathway and facilitate viral infection (Poque et al., 2017). Recently, it was reported that high levels of guanosine pentaphosphate or tetraphosphate ((p)ppGpp) accumulation increased susceptibility to TuMV and reduced SA content, whereas plants with lower (p)ppGpp levels were more resistant to TuMV and had increased SA content, indicating a role for the SA pathway in defence against TuMV (Abdelkefi et al., 2018). Our finding that transgenic NahG N. benthamiana was susceptible to TuMV, while exogenous SA enhanced N. benthamiana resistance to TuMV, demonstrates that SA participates in defence against TuMV in N. benthamiana (Fig. 4). SA accumulation is regulated positively by NbALD1 and is essential for NbALD1‐mediated resistance to TuMV.

In examining the relationship between NbALD1 and SA, we found that Pip treatment compromised the susceptibility of NahG plants to TuMV and also that SA treatment did not restore the resistance of NbALD1‐silenced plants to that of wild‐type plants, indicating that there are one or more unknown pathways regulated by NbALD1 that participate in NbALD1‐mediated resistance to TuMV (Fig. 5). This is consistent with the results reported for the role of ALD1 in Arabidopsis against P. syringae. In those experiments, treatment with Pip and NHP increased the resistance of wild‐type plants but also of the sid2 mutant, suggesting that there is an SA‐independent role for ALD1 in induced resistance (Bernsdorff et al., 2016; Hartmann et al., 2018; Navarova et al., 2012; Vogel‐Adghough et al., 2013), although the identity of that SA‐independent pathway was not determined.

Our experiments suggest that ethylene may be an important factor in NbALD1‐mediated resistance, at least to TuMV. Ethylene is an important gaseous phytohormone that affects plant growth, development and biotic stress (Cao et al., 2007; Van Loon et al., 2006; Wang et al., 2013; Yang et al., 2017). Silencing of NbALD1 increased the content of ACC, the precursor of ethylene, and the expression level of ERF3, the ethylene response gene (Fig. 6). At the same time, Pip treatment or overexpression of NbALD1 inhibited ACC accumulation and the expression of ERF3 (Fig. 6). These results, therefore, indicate negative regulation of the ethylene pathway by NbALD1. However, the relationship between ALD1 and ethylene may be complex because it has been reported that the ald1 single mutant displayed an ethylene‐insensitive phenotype and delayed ethylene‐induced senescence (Nie et al., 2011). Here, preliminary results showed that ERF3 that was induced by TuMV infection on wild‐type plants was down‐regulated in TuMV‐infected plants overexpressing NbALD1, while in ACS‐silenced plants expression of NbALD1 was still up‐regulated by TuMV infection (Fig. S9). We suppose that there may be a complicated mechanism to manipulate the regulation of ALD1 and balance the levels of ethylene and salicylic acid, which is worth investigating further.

Increasing evidence shows that ethylene plays roles in viral infection. The symptoms of chlorosis and stunting shown by plants infected with CaMV or expressing CaMV‐P6 (the main symptom determinant protein) depended on interactions between P6 and ethylene‐associated components (Geri et al., 2004), while Arabidopsis mutants with defects in ethylene signalling showed reduced CaMV susceptibility (Love et al., 2007). Silencing ethylene biosynthetic and signalling genes strongly increased susceptibility to Chili veinal mottle virus (ChiVMV), indicating a key role for ethylene in plant systemic resistance against ChiVMV (Zhu et al., 2014). Ethylene‐inducible transcription factor RAV2 is required for suppression of RNA silencing by two unrelated plant viral proteins, potyvirus HC‐Pro and carmovirus P38 (Endres et al., 2010). While efficient infection by rice dwarf virus (RDV) depends upon a specific interaction between RDV‐Pns11 protein and OsSAMS, a key component of ethylene biosynthesis (Marco et al., 1976; Zhao et al., 2017), TuMV infection induce ethylene to benefit the virus (Casteel et al., 2015). These reports indicate a complicated role for ethylene during viral infection. In our experiments, silencing of genes in the ethylene pathway, ACS1, ACO1 and EIN2, increased plant resistance against TuMV (Fig. 7). Consistent with this, treatment with AVG, an inhibitor of ethylene, also enhanced plant resistance to TuMV (Fig. 7), while treatment with ACC made plants more susceptible (Fig. 7). The results thus demonstrate that the ethylene pathway is associated with plant susceptibility to TuMV and may be necessary for TuMV infection. Additionally, silencing of ACS1 in the ethylene pathway alleviated the susceptibility of NbALD1‐silenced plants to TuMV (Fig. 8), while exogenous application of ACC compromised the resistance of Pip‐treated plants and NbALD1 transgenic plants to TuMV (Fig. 8). These results demonstrate that the ethylene pathway is negatively regulated by NbALD1 and functions as an active defence against TuMV infection.

Taken together, the results reported here suggest a model for the role of NbALD1 in defence against TuMV that involves two regulated pathways: SA mediates resistance to TuMV and is positively regulated by NbALD1, while the ethylene pathway mediates susceptibility to the virus and is negatively regulated by NbALD1. It would be worth examining the possible relationship between these two pathways in NbALD1‐mediated resistance against TuMV.

Experimental procedures

Plants and virus inoculation

N. benthamiana plants and NahG transplants were grown in a glasshouse as previously described (Shi et al., 2016). Plant at 5 weeks old were chosen to be infiltrated with Agrobacterium tumefaciens carrying TuMV‐GFP vector. At 6 dpi, the systemically infected leaves were collected for 100 mg and ground in 10 mL PBS buffer (8 g/L NaCl, 0.2 g/L KCl, 1.44 g/L Na2HPO4, 0.24 g/L KH2PO4, pH 7.4), and suspended to use for mechanical virus inoculation. Then plants at 3 weeks old were used for mechanical virus inoculation. Plants were examined daily and the numbers with symptoms recorded. After the symptoms appeared on the upper leaves, the leaves were photographed and harvested. The number of infection foci was assessed on 20 plants for each treatment (Table S1).

Vector construction

For overexpressing NbALD1, the entire open reading frame (ORF) of NbALD1 was fused with HA‐tag and then introduced into the PCV vector. The inserted fragment was obtained by PCR using a pair of primers linked to BamHI and SacI sequences on either side. The PCV vector was digested with BamHI and SacI, then the digested vector and PCR product were treated with T4 DNA ligase. To suppress NbALD1, a segment of NbALD1 was obtained by PCR and then ligated into the TRV‐RNA2 vector as previously described (Liu et al., 2002). TRV‐RNA2 was digested with PstI, then the digested PCR and PCR products were treated with T4 DNA polymerase in the presence of dTTP and dATP. The resulting constructs were verified by sequencing and transformed to Agrobacterium tumefaciens strain GV3101. All primers used are listed in Table S2.

Exogenous application of Pip, SA, AVG and ACC

Ten millilitres of a 100 μM L‐Pip (P1404; TCI Europe) solution was pipetted onto the soil substrate of individually cultivated plants, 8 days prior to virus inoculation. SA was applied to leaves at a concentration of 10 μM and these leaves were inoculated with virus 48 h later. AVG and ACC were sprayed onto leaves at concentrations of 10 or 100 μM respectively, and leaves were inoculated with virus 16 h later. In all cases, plants were similarly treated with water as controls.

Quantification of SA and ACC from tobacco leaves

SA content was determined from 0.5 g samples of fresh leaves ground in liquid nitrogen using a mortar and pestle. After the addition of 5 mL exaction buffer (isopropanol/hydrochloric acid), the suspension was stirred for 30 min at 4 °C, 10 mL dichloromethane was added and the mixture stirred again for 30 min at 4 °C. After centrifugation at 13 000 g for 5 min at 4 °C, the supernatants were pooled, the organic phase was dried with nitrogen and then dissolved in 400 μL methanol containing 0.1% carboxylic acid. Samples were analysed by HPLC‐electrospray ionization/MS‐MS using a Waters ACQUITY UPLC coupled to an Xevo TQ. Chromatographic separation was carried out on a Waters ACQUITY UPLC R BEH C18 100 mm × 2.1 mm × 1.7 μm column at 40 °C. The solvent gradient used was 100% A (98/2 = H2O/CH3OH (V/V) +0.05% CHOOH + 5 mM CH3COONH4) to B (CH3CN) over 10 min. Solvent B was held at 100% for 5 min then the solvent returned to 100% A for 10 min equilibration prior to the next injection. The solvent flow rate was 0.3 mL/min. The MS was operated in the negative mode using electro‐spray ionization as the ion source.

ACC was extracted from the same leaf tissues which were ground in liquid nitrogen, then transferred into 50 mL microfuge tubes with 5 mL deionized water and treated in the ultrasonic cleaning instrument for 20 min. After centrifugation at 10 000 g for 5 min at 4°C, the supernatants were collected (pH = 4.0). Twenty millilitres of trichloromethane was added, the mixture was centrifuged at 10 000 g for 5 min and the supernatants were collected and infiltrated through a MCX polybase which had been pre‐activated by 3 mL CH3OH and 3 mL deionized water. The residue was eluted with 4 mL 1 M NH3·H2O. The ACC concentration in the supernatant was determined by HPLC‐electrospray ionization/MS‐MS.

SA and ACC assays used three replicates each of at least ten plants per treatment and genotype.

Northern blot, quantitative RT‐PCR and semi‐quantitative RT‐PCR

Total RNA was extracted from plants with TRIzol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. For northern blot, 10 μg of total RNA was separated on 1.5% formaldehyde agarose gels in 1 × MOPS buffer, and blotted onto Hybond‐N + membranes (GE Healthcare, Chicago, IL, USA). A probe of nearly 300 bp partially complementary to TuMV was labelled with Digoxigenin (DIG) (Roche, Basel, Switzerland). The sequences of the primers used are listed in Table S2.

For reverse transcription polymerase chain reaction (RT‐PCR), 1 μg of total RNA was reverse transcribed into cDNA by PrimeScript RT Enzyme (Takara). N. benthamiana Ubiquitin C gene (UBC) gene (Accession No. AB026056.1) was used as the internal reference gene for analysis. A Vazyme Ace qRT‐PCR system was used for the reaction and the results were analysed by the ∆∆CT method. The primer sequences used for this assay are listed in Table S2.

The UBC gene was also used as the internal reference gene in semiquantitative RT‐PCR analysis. Reactions were subjected to a varying number of cycles (28, 30, 32, 34, 36 and 38) of PCR amplification to determine the optimal cycle number within the linear range for PCR amplification. Products collected at various cycles were analysed by electrophoresis in 1.5% agarose‐GelRed gels, and those at the optimal cycle for each gene were used to compare relative expression levels. Primer sequences used in this assay are listed in Table S2.

Western blot

Total proteins were extracted from plant samples using lysis buffer (100 mM Tris‐HCl, pH 8.8, 60% SDS, 2% β‐mercaptoethanol) and separated in a 12% SDS‐PAGE gel as previously described (Jiang et al., 2014), then transferred onto nitrocellulose (Amersham, Uppsala, Sweden) by electroblotting, and detected with primary antibody to GFP and HA‐tag and secondary anti‐mouse and anti‐rabbit antibodies (Sigma‐Aldrich, St Louis, MO, USA). The antigen‐antibody (for secondary anti‐mouse) complexes were visualized using NBT/BCIP buffer (Sigma) at room temperature and the antigen‐antibody (for secondary anti‐rabbit) was visualized using ECL HRP Chemiluminescent Substrate (Invitrogen) under a chemiluminescence analyzer.

Quantification and statistical analysis

Three replicates, each of 20 plants per treatment and genotype, were used to assess the rate of systemic infection (Table S3) and three independent plants per treatment and genotype were sampled for qRT‐PCR analyses in all experiments. All results were confirmed by three independent experiments. ANOVA analyses with type II sum of squares were performed on log 10‐transformed values to assess statistical differences of datasets. Unpairwise comparisons of treatment versus control values were performed with a two‐tailed Student’s t‐test in Prism 6, and Holm–Sidak's multiple comparisons test was used for multiple comparison.

Fig. S1 Identities and alignment among ALD1 from different plants.

Click here for additional data file.

Fig. S2 Silencing of NbALD1 in N. benthamiana.

Click here for additional data file.

Fig. S3 Overexpression of NbALD1 in N. benthamiana.

Click here for additional data file.

Fig. S4 Exogenous SA partially complemented the resistance deficiency of NbALD1‐silenced plants.

Click here for additional data file.

Fig. S5 Silencing of ACS1, ACO1 and EIN2 in N. benthamiana.

Click here for additional data file.

Fig. S6 AVG treatment inhibited the expression of ERF3.

Click here for additional data file.

Fig. S7 ACC treatment induced the expression of ERF3.

Click here for additional data file.

Fig. S8 Silencing of NbALD1 and ACS1 in N. benthamiana simultaneously.

Click here for additional data file.

Fig. S9 Expression of ERF3 in TuMV‐infected wild‐type plants (A), NbALD1‐overexpressed plants (B) and expression of NbALD1 in TuMV‐infected ACS1‐silenced plants (C).

Click here for additional data file.

Table S1 Numbers of infection loci and statistical analysis.

Click here for additional data file.

Table S2 Primers used for analysis.

Click here for additional data file.

Table S3 Numbers of plants showing fluorescence under UV light in the systemic leaves at different times and statistical analysis of infection rates.

Click here for additional data file.