Developmentally regulated Arabidopsis thaliana susceptibility to tomato spotted wilt virus infection

Abstract Tomato spotted wilt virus (TSWV) is one of the most devastating plant viruses and often causes severe crop losses worldwide. Generally, mature plants become more resistant to pathogens, known as adult plant resistance. In this study, we demonstrated a new phenomenon involving developmentally regulated susceptibility of Arabidopsis thaliana to TSWV. We found that Arabidopsis plants become more susceptible to TSWV as plants mature. Most young 3‐week‐old Arabidopsis were not infected by TSWV. Infection of TSWV in 4‐, 5‐, and 6‐week‐old Arabidopsis increased from 9%, 21%, and 25%, respectively, to 100% in 7‐ to 8‐week‐old Arabidopsis plants. Different isolates of TSWV and different tospoviruses show a low rate of infection in young Arabidopsis but a high rate in mature plants. When Arabidopsis dcl2/3/4 or rdr1/2/6 mutant plants were inoculated with TSWV, similar results as observed for the wild‐type Arabidopsis plants were obtained. A cell‐to‐cell movement assay showed that the intercellular movement efficiency of TSWV NSm:GFP fusion was significantly higher in 8‐week‐old Arabidopsis leaves compared with 4‐week‐old Arabidopsis leaves. Moreover, the expression levels of pectin methylesterase and β‐1,3‐glucanase, which play critical roles in macromolecule cell‐to‐cell trafficking, were significantly up‐regulated in 8‐week‐old Arabidopsis leaves compared with 4‐week‐old Arabidopsis leaves during TSWV infection. To date, this mature plant susceptibility to pathogen infections has rarely been investigated. Thus, the findings presented here should advance our knowledge on the developmentally regulated mature host susceptibility to plant virus infection.

Tospovirus virions are spherical, enveloped, and 80-120 nm in diameter (Kormelink et al., 2011;Oliver & Whitfield, 2016). The virions contain three genomic RNA segments: large (L), medium (M), and small (S). The L RNA segment encodes an RNA-dependent RNA polymerase (RdRp) from the viral complementary strand van Knippenberg et al., 2002). The M RNA segment encodes a movement protein (NSm) from the viral sense strand and a glycoprotein precursor, which is further processed into a Gn and a Gc protein, from the viral complementary strand (Kormelink et al., 1992;Ribeiro et al., 2009). The S RNA segment encodes a nucleocapsid (N) protein from the viral sense strand and an RNA silencing suppressor (NSs) from the viral complementary strand (Takeda et al., 2002;Bucher et al., 2003;Schnettler et al., 2010).
Arabidopsis is an important model plant for studies on plant-virus interactions (Lellis et al., 2002;Garcia-Ruiz et al., 2010Wang et al., 2010Wang et al., , 2011bAregger et al., 2012;Uchiyama et al., 2014;Zhang et al., 2015;Gaguancela et al., 2016;Cheng et al., 2017;Hafren et al., 2017Hafren et al., , 2018Yuan et al., 2018). Arabidopsis has been shown to be susceptible to TSWV infection . In this study, we found that successful TSWV infection in Arabidopsis plants is linked to plant developmental stages. For example, at 3 weeks old, Arabidopsis plants are not susceptible to TSWV infection, but their susceptibility gradually increases as the plants mature. Inoculation of Arabidopsis dcl2/3/4 or rdr1/2/6 mutant plants with TSWV have shown that the resistance of young Arabidopsis plants to TSWV infection is not controlled by the genes involved in the RNA silencing pathway. Plasmids encoding the TSWV NSm:GFP fusion were delivered to leaves by particle bombardment and the fusion proteins moved efficiently between cells in 8-week-old Arabidopsis leaves compared with 4-week-old Arabidopsis leaves. Our results indicate that TSWV infection in Arabidopsis is controlled by the plant developmental stage, and that the age of the Arabidopsis plant should be considered when studying TSWV and possibly other viruses.

| Effect of Arabidopsis developmental stages on TSWV infection
In our initial experiments, we inoculated 5-week-old Arabidopsis Col-0 plants with TSWV lettuce isolate (TSWV-LE). By 30 days postinoculation (dpi), the inoculated plants did not show symptoms, contradicting the commonly accepted knowledge that younger plants are more susceptible to virus infections. To determine whether this TSWV isolate has the ability to infect Arabidopsis, we inoculated 9-week-old Arabidopsis plants with TSWV-LE and found that all the inoculated plants showed strong disease symptoms, including leaf curling and necrosis, and plant stunting followed by plant death at about 25-30 dpi ( Figure S1). To rule out the possibility that the lack of TSWV infection in the 5-week-old plants was caused by an inefficient inoculation method, we inoculated the same amount of TSWV-infected crude leaf sap (5 µl per leaf) to three leaves on each 5-or 9-week-old Arabidopsis plant. As expected, similar results as described above were obtained for the inoculated 5-and 9-week-old Arabidopsis plants, indicating that the susceptibility of Arabidopsis to TSWV infection is linked to the plant developmental stage.
Reverse transcription (RT)-PCR showed that by 30 dpi, only plants without visible TSWV symptoms lacked virus (Figure 1b,c). All plants inoculated at 7 or 8 weeks old showed TSWV symptoms (Table 1) To investigate whether this developmental stage-controlled susceptibility is specific to this isolate of TSWV, we inoculated Arabidopsis plants ranging from 3 to 8 weeks old with TSWV Yunnan tomato isolate (TSWV-YN). The 3-week-old Arabidopsis plants were not susceptible to TSWV-YN infection, while the 7-or 8-week-old plants were highly susceptible to TSWV-YN infection (Figure 1a,d,e and Table 1). To determine whether this developmental stage-controlled susceptibility is Arabidopsis ecotype-specific, we inoculated Arabidopsis  Figure   S2 and Table 1). Consequently, we conclude that the susceptibility of Arabidopsis to TSWV infection is linked to plant developmental stage.
To determine if Arabidopsis Col-0 susceptibility to other tospoviruses is also controlled by their developmental stages, we inoculated 3-to 8-week-old plants with tomato zonate spot virus (TZSV, a Euro/Asiantype tospovirus) or impatiens necrotic spot virus (INSV, an Americantype tospovirus). The results showed that none of the 3-week-old inoculated Arabidopsis plants developed virus-like symptoms and none accumulated the virus by 15 dpi as determined by RT-PCR (Table 1 and Figure 2a

| Arabidopsis susceptibility to CMV infection
To determine if this developmental stage-controlled susceptibility occurs for other plant viruses, we inoculated 3-to 8-week-old

| Effects of tomato, pepper, and N. benthamiana developmental stages on TSWV infection
We inoculated tomato, pepper, and N. benthamiana plants with TSWV-LE at the 2-, 3-, 4-, and 5-week-old stages. The results showed that the susceptibility of these three plants to TSWV infection was not influenced by plant developmental stage ( Figure 4 and Table 2).
Symptoms in the TSWV-infected plants were leaf chlorosis and plant stunting by 10 dpi. RT-PCR showed that by 10 dpi, TSWV had accumulated in these three assayed host plants (Figure 4b,c,e,f,h,i), indicating that this developmental stage-controlled plant susceptibility is not universal.

| Developmental stage-regulated Arabidopsis susceptibility to TSWV infection is not controlled by its RNA silencing pathway
To determine if the RNA silencing pathway is involved in this developmental stage-controlled Arabidopsis susceptibility to TSWV infection, we inoculated 3-to 8-week-old Arabidopsis dcl2/3/4 or rdr1/2/6 mutant plants with TSWV-LE as described above. The results showed that the TSWV-LE-inoculated dcl2/3/4 or rdr1/2/6 mutant Arabidopsis plants exhibited a similar developmental The systemic presence of virus in uninoculated leaves of the inoculated plants was determined by both ELISA and reverse transcription (RT)-PCR.
The plants inoculated at the 3-6-week-old stage were assayed at 30 days post-inoculation (dpi) and the plants inoculated at 7-8-week-old stage were assayed at 15 dpi. b The systemic presence of CMV in uninoculated leaves of the inoculated plants was determined by ELISA and RT-PCR at 40 dpi.

| Intercellular movement of TSWV NSm:GFP is regulated by Arabidopsis developmental stages
The TSWV-encoded NSm protein is critical for TSWV cell-tocell and long-distance movement in plants. We next investigated whether cell-to-cell movement of NSm:GFP fusion in Arabidopsis leaves is regulated by developmental stages through a particle bombardment assay. As a control, we bombarded leaves of the 4-or 8-week-old Arabidopsis plants with a control construct expressing green fluorescent protein (GFP):GFP fusion. The results showed that most loci expressing GFP:GFP fusion showed single cells with GFP green fluorescent signal (90.7% for the 8-week-old plant leaves and 85.3% for the 4-week-old plant leaves) (Table 3 and Figure 6a,c). In the 8-week-old plant leaves expressing NSm:GFP, a total of 296 loci were observed. Among these loci, 139 loci (47.0%) were single cell loci, 81 loci (27.4%) were clusters of 2-5 cells, 46 loci (15.5%) were clusters of 6-10 cells, and 30 loci (10.1%) were clusters of over 10 cells (Figure 6b and Table 3). In the 4-week-old plant leaves, a total of 244 loci showed NSm:GFP green fluorescent signal. Among these loci, 146 loci (59.8%) were single cell loci, 71 loci (29.1%) were clusters of 2-5 cells, 20 loci (8.2%) were clusters of 6-10 cells, and 7 loci (2.87%) were clusters of over 10 cells (Figure 6d and Table 3), indicating that intercellular movement of NSm:GFP is more efficient in the 8-week-old plant leaves than that in the 4-week-old plant leaves.

| Expression of PME and BGL genes was regulated by plant developmental stage and TSWV infection
Arabidopsis pectin methylesterase (PME) and β-1,3-glucanase (BGL) are known to regulate plasmodesmata functions during virus infection in plants (Dorokhov et al., 1999;Chen et al., 2000;Iglesias & Meins, 2000;Bucher et al., 2001). We analysed a previously published transcriptome data set obtained from TSWV-infected Arabidopsis plants (Xu et al., 2020) and found that the expression of AtPME (AT2G45220) and AtBGL (AT3G57260) was strongly up-regulated at 9, 12, and 15 days after TSWV inoculation (Figure 7a,b). To specifically examine expression of these genes, we analysed the expression of AtPME   Leisner and others reported that growth conditions important for Arabidopsis development can drastically affect cauliflower mosaic virus infection in Arabidopsis (Leisner et al., 1993). Tennant (Tennant et al., 2001). This developmental stage-controlled resistance has also been found in Arabidopsis plants to Pseudomonas syringae pv. tomato DC3000 or P. syringae pv. maculicola 4326 infection (Kus et al., 2002). In this study, after we transiently expressed TSWV NSm:GFP fusion in the leaves of the 4-or the 8-week-old Arabidopsis plants through particle bombardment, we found that the NSm:GFP fusion moved faster between cells in 8-week-old Arabidopsis leaves compared with 4-week-old Arabidopsis

leaves, suggesting a possible barrier between cells in the younger
Arabidopsis plant leaves. PME and Class I GBL of Nicotiana tabacum are known to regulate the size exclusion limit of plasmodesmata in cell walls and facilitate the cell-to-cell movement of many plant viruses (Dorokhov et al., 1999;Chen et al., 2000;Iglesias & Meins, 2000;Bucher et al., 2001). Our RT-qPCR results showed that the expression level of AtGBL in 8-week-old Arabidopsis plants is significantly higher than that in 4-week-old Arabidopsis plants. Although the average expression level of AtPME in 8-week-old Arabidopsis plants has no TA B L E 3 Cell-to-cell movement of TSWV NSm:GFP and GFP:GFP fusion in 4-or 8-week-old Arabidopsis plant leaves  of AtPME and AtGBL on TSWV infection. As expected, after the expression of AtPME or AtGBL was silenced through VIGS, the accumulation level of TSWV in the TSWV-inoculated Arabidopsis plants was significantly reduced, further supporting the above hypothesis that the expression of AtPME and AtGBL is crucial for the developmental stage-controlled susceptibility to TSWV infection in Arabidopsis plants. It is noteworthy that this developmental stage-controlled susceptibility was not observed in tomato, pepper, or N. benthamiana.
One possible explanation for this difference is that the basal levels of PME and GBL in young tomato, pepper, and N. benthamiana plants are sufficient for successful TSWV infection. The other possibility is that the expression of PME and GBL is readily induced in both young F I G U R E 7 Expression of AtPME (AT2G45220) and AtBGL (AT3G57260) in 4-or 8-week-old Arabidopsis plants inoculated with TSWV or phosphate-buffered saline (Mock). (a) and (b) Expression of AtPME and AtBGL in TSWV-inoculated or mock-inoculated Arabidopsis plants at 9, 12, and 15 days post-inoculation (dpi). These data were summarized from the recent published transcriptome profile of TSWV-infected or uninfected Arabidopsis plants. (c) and (d) The expression of AtPME and AtBGL was determined by quantitative reverse transcription PCR (RT-qPCR) using genespecific primers. Leaves of 4-or 8-week-old Arabidopsis plants were harvested at 7 dpi and analysed through RT-qPCR. The resulting data were calculated using the 2 −ΔΔCt method. *p < .05, determined using the unpaired two-tailed Student t test. (e) RT-qPCR analysis of AtPME and AtBGL expression in the ALSV-AtPME-, ALSV-AtGBL-, or ALSV-AtCH42-inoculated Arabidopsis plants. Error bars represent SD (n = 4); *p < .05. (f) A western blot result showing the accumulation level of TSWV in the AtPME-silenced + TSWV-inoculated, AtBGL-silenced + TSWV-inoculated or ALSV-empty (non-silenced) + TSWV-inoculated Arabidopsis plants. Five TSWV-inoculated leaves were harvested from each treatment at 5 dpi, pooled, and analysed through a western blot assay using a TSWV N-specific antibody. The Ponceau S stained gel is used to show the sample loadings and mature tomato, pepper, and N. benthamiana plants on TSWV infection.
It has been shown that a developmentally regulated plasma membrane protein of N. benthamiana (NbDREPP) can interact with two different movement proteins (P3N-PIPO and CI) of tobacco vein banding mosaic virus (TVBMV, a potyvirus) and facilitate the virus to move in infected leaves (Geng et al., 2015). Developmentally regulated intercellular trafficking has also been shown for CMV MP:GFP fusion movement in tobacco leaf cells (Itaya et al., 1998). In addition, these authors have shown that the walls of young tobacco leaf cells contain only primary plasmodesmata while the walls of mature leaf cells contain mostly secondary plasmodesmata that are more permeable to macromolecule trafficking. Iglesias and Meins have demonstrated that silencing of BGL expression in tobacco leaves can delay the cell-to-cell trafficking of CMV movement protein (Iglesias & Meins, 2000). In a different study, the authors found that the plasmodesmata between vascular bundle sheath cells and vein cells are structurally different from that between mesophyll cells, and have suggested that this difference may restrict brome mosaic virus infection predominantly in the mesophyll cells and the cells associated with vasculature in barley leaves at 24/20 °C (Ding et al., 1996(Ding et al., , 1999 (Li et al., 2014). Primers specific for Arabidopsis pectin methylesterase (AtPME, AT2G45220) and Arabidopsis β-1,3glucanase gene (AtBGL, AT3G57260) are listed in Table S1. The expression of Arabidopsis Actin 2 gene (AtACTIN 2, AT3G18780) was used as an internal control, and the relative expressions of AtPME and AtBGL were calculated using the 2 −∆∆Ct method (Livak & Schmittgen, 2001).

| Particle bombardment
Plasmid DNA pRTL2-NSm:GFP and pRTL2-GFP:GFP were made previously (Feng et al., 2016). Particle bombardment was conducted as described (Sanford et al., 1993) with specific modifications. Briefly, 60 mg of tungsten M-10 microcarrier (Bio-Rad) was mixed with 1 ml of freshly prepared 70% ethanol in a 1.5-ml Eppendorf tube, vortexed for 3 min, incubated at room temperature for 15 min, and then pelleted by 5 s centrifugation in a microfuge. The supernatant was removed from the microfuge tube and the pellet was rinsed three times with 70% ethanol. The pellet was resuspended in a 50% sterile glycerol solution to achieve a 60-mg tungsten particle/ml stock solution. Fifty microlitres of tungsten particle stock solution was mixed with 5 µg of pRTL2-NSm:GFP or pRTL2-GFP:GFP plasmid DNA, 50 µl of 2.5 M CaCl 2 , and 20 µl of 0.1 M spermidine. The mixed tungsten particle:plasmid DNA complexes were pelleted, resuspended in 140 µl of 70% ethanol, pelleted again, and then resuspended in 48 µl of 100% ethanol.
An aliquot (15 µl) of tungsten particle:plasmid DNA complex solution was loaded onto the centre of a macrocarrier (Bio-Rad), air-dried, and bombarded onto the abaxial side of leaves harvested from the 4-or the 8-week-old Arabidopsis Col-0 plants using the He/1000 particle delivery system (Bio-Rad). The bombarded leaves were incubated on wet filter papers inside Petri dishes for 21 hr at 22 °C in the dark.

| Confocal laser scanning microscopy
The bombarded Arabidopsis leaves were harvested and examined under an LSM 710 confocal laser scanning microscope equipped with a plan-apochromat 40×/0.95 Korr M27 lens (Carl Zeiss). The excitation wavelength was set at 488 nm and the emission wavelength was set at 528 nm as previously described (Wang et al., 2011a). The captured images were processed using the Zeiss 710 CLSM and Adobe Photoshop software.

| Virus-induced gene silencing in Arabidopsis
Silencing of AtPME or AtGBL expression in Arabidopsis was performed using ALSV-based vector as described previously (Igarashi et al., 2009). Full-length ALSV RNA1 (GenBank accession no. AB030940) and RNA2 (AB030941) sequences were synthesized by the GenScript (Nanjing, China) and cloned individually into the pCB301-2 × 35S-RZ-NOS vector (Feng et al., 2020) to produce pALSV1 and pALSV2, respectively. A partial sequence (300 bp) of AtPME or AtGBL was RT-PCR amplified from a total RNA sample using specific primers and cloned into the pALSV2 vector to produce pALSV2-AtPME or pALSV2-AtGBL. pALSV2 vector without an insert or with a 300 bp insert from AtCH42 (pALSV2-AtCH42) were used as control vectors. Agrobacterium tumefaciens cultures carrying pALSV-RNA1 and pALSV2, pALSV-RNA1 and pALSV2-AtPME, pALSV-RNA1 and pALSV2-AtGBL, or pALSV-RNA1 and pALSV2-AtCH42 were individually mixed with an equal volume of A. tumefaciens culture carrying a vector expressing an RNA silencing suppressor (tomato bushy stunt virus P19), and the mixed cultures were individually infiltrated into leaves of N. benthamiana plants. Eighteen days later, three newly emerged leaves were collected from each N. benthamiana plant and virus (i.e., ALSV, ALSV-AtPME, ALSV-AtGBL, or ALSV-AtCH42) in each leaf sample was isolated as described previously (Zhu et al., 2014). The isolated virus was inoculated to six leaves of each 6-weekold Arabidopsis plants. After 5 days, the three newly merged leaves on each ALSV-, ALSV-AtPME-, ALSV-AtGBL-, or ALSV-AtCH42inoculated Arabidopsis plant were inoculated with a crude sap prepared from TSWV-infected leaf tissues. Five days later, leaves of five plants representing a specific treatment were harvested, pooled, and analysed for TSWV accumulation using western blot assay.

ACK N OWLED G M ENTS
We thank Professor Baoping Li (Nanjing Agricultural University) for his helpful suggestions and comments on statistical analysis. This work was supported by the National Natural Science Foundation of China

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.