Cautionary notes on the use of Agrobacterium‐mediated transient gene expression upon SGT1 silencing in Nicotiana benthamiana

The suppressor of the G2 allele of skp1 (SGT1) is an essential component of the plant immune system, required for the induction of disease resistance mediated by many intracellular immune receptors containing nucleotide-binding domain and leucine-rich repeats (NLRs) (Azevedo et al., 2002; Kadota et al., 2010). Accordingly, virus-induced gene silencing (VIGS) of SGT1 in Nicotiana benthamiana is commonly used to test for immune- or SGT1-dependence of hypersensitive response (HR), or HR-like cell death induced by pathogen effectors, elicitors, or heterologous protein overexpression (Supporting Information Table S1). A reduction or loss of macroscopic cell death can be easily detected when the cellular reactions leading to such cell death require SGT1 activity, making this a powerful and straightforward assay. In the course of our experiments, aimed at determining the SGT1 dependence of immune-associated phenotypes, we performed VIGS in 2-wk-old N. benthamiana plants using tobacco rattle virus (TRV) following standard published procedures (Liu et al., 2002a). Three weeks after inoculation, silenced N. benthamiana plants showed the typical symptoms reported before for plants undergoing VIGS of NbSGT1, including shorter and more branched shoots, and thick, curly, and fragile leaves (Fig. S1). Accordingly, these tissues showed a dramatic reduction of NbSGT1 transcripts (Fig. 1a). To determine whether this causes a subsequent reduction of NbSGT1 protein accumulation, we raised an anti-AtSGT1 antibody (Methods S1; Fig. S2) and found that the accumulation of NbSGT1 is undetectable in SGT1 VIGS plants (Fig. 1b). Unexpectedly, when we tried to perform Agrobacterium-mediated transient expression of heterologous genes in SGT1 VIGS N. benthamiana plants, we noticed a very poor accumulation of the heterologous proteins in comparison with plants expressing the TRV empty vector (EV VIGS), including GFP (using both western blot and confocal imaging; Fig. 1c,d), GUS (Fig. S3), and several other proteins routinely used in transient expression assays in our laboratory (data not shown). It is noteworthy that we have not observed this phenomenon upon VIGS of other immune signaling components, such as NbNDR1 or NbEDS1, following the same VIGS protocol (Fig. S4). SGT1, in association with RAR1 and HSP90, acts as an important component of a chaperone complex for NLR proteins (Kadota et al., 2010) and is required for the appropriate accumulation of NLRs and, potentially, other proteins (Kadota et al., 2010; Ichimura et al., 2016). Therefore, we contemplated the possibility that SGT1 chaperone activity could be required for the appropriate accumulation of heterologous proteins. However, we found that the Agrobacterium-mediated expression of heterologous transcripts was also compromised in SGT1 VIGS plants in comparison with EV VIGS plants (Fig. 1e), suggesting an effect of SGT1 silencing upstream of the biosynthesis of heterologous proteins. In keeping with this notion, SGT1 silencing did not affect the accumulation of GFP in N. benthamiana 16C plants, which constitutively express the GFP gene under a 35S promoter (Fig. 1f). We thus considered that SGT1 silencing could have an impact on Agrobacterium survival in plant tissues. However, we found no significant difference between the numbers of Agrobacterium colony-forming units in SGT1 VIGS and EV VIGS plants 3 d after inoculation (Fig. 1g). Nevertheless, we observed that the expression of immune-associated genes, such as NbPR1 (commonly used to determine activation of salicylic acid (SA) signaling; Fig. S5), is up-regulated in SGT1 VIGS plants in comparison with EV VIGS plants, suggesting that SGT1 silencing may induce a weak activation of immune signaling not sufficient to reduce the survival of Agrobacterium in plant tissues. Our results indicate that, in our apparently standard experimental conditions, SGT1 silencing significantly impairs Agrobacterium-mediated transient expression and, subsequently, the accumulation of heterologous proteins. The same phenomenon has been shown in previous reports revealing a requirement of SGT1 and RAR1 for efficient Agrobacterium-mediated transformation (Anand et al., 2012; Anand & Mysore, 2013). This prompts for precaution when designing and interpreting experiments to determine whether an observed phenotype is SGT1-dependent. For instance, cell death triggered upon overexpression of a pathogen protein may not be detected in SGT1 VIGS tissues because the protein does not even accumulate, which may lead to the interpretation that the pathway required for such cell death is SGT1-dependent. Therefore, to avoid misinterpretations, it is essential to confirm (and show) the actual accumulation of the heterologous protein. Unfortunately, although several reports have unequivocally shown that heterologous proteins actually accumulate in SGT1 VIGS plants in their experimental conditions, this information is often missing in studies using approaches similar to the one exemplified here (Table S1). It is important to consider that even slight differences in the experimental conditions or materials used, including age of the plants used, TRV clone, environmental conditions, and time between the first and second Agrobacterium inoculation (to express TRV and the heterologous protein, respectively) may have a strong impact on protein accumulation in SGT1 VIGS plants. For example, in our experimental conditions, we realized that SGT1 silencing takes place as early as 1 wk after TRV inoculation, and, in these conditions, heterologous protein accumulation can be detected (Fig. 1h). Sustained SGT1 silencing led to the reported impairment in Agrobacterium-mediated transient expression (Fig. 1d–f,h). Since these conditions or materials may vary among different experiments, growth chambers, and laboratories, the confirmation of the accumulation of the heterologous protein should be performed in every experiment. This should anyway be a required practice in all the experimental approaches aimed at detecting the presence/absence of a response triggered by the expression of a heterologous protein upon genetic or chemical manipulation of plant tissues. Similarly, a recent report has shown that a bacterial effector, XopQ1, thought to suppress cell death triggered by co-expressed proteins, is, in fact, attenuating the Agrobacterium-mediated transient expression of other proteins (Adlung & Bonas, 2017). We observed that the expression of NbPR1 is up-regulated in SGT1 silenced tissues, suggesting that SA signaling is enhanced. This is consistent with the previous finding that silencing SGT1 in N. attenuata results in elevated accumulation of SA (Meldau et al., 2011). SA signaling has been shown to restrict Agrobacterium-mediated transient expression (Rosas-Díaz et al., 2017) without impacting significantly the survival of Agrobacterium cells in plant tissues. This suggests that the enhancement of SA signaling caused by sustained SGT1 silencing may be a factor in the reduced protein accumulation in SGT1 VIGS plants, although further detailed studies would be required to understand the nature of this phenomenon. In summary, our results suggest caution when interpreting the absence of a response caused by SGT1 silencing when the assays involve the overexpression of heterologous proteins, and recommend the confirmation of protein accumulation in plant tissues to avoid misinterpretations. There are two sets of different TRV-based VIGS systems mainly used by the research community (Ratcliff et al., 2001; Liu et al., 2002a). For this work, we used the system published by Liu et al. (2002a). Nicotiana benthamiana wild-type or 16C plants (Ruiz et al., 1998) were grown at 22°C in a walk-in chamber under 16 h : 8 h, light : dark cycle and a light intensity of 100 to 150 mE m−2 s−1. For VIGS assays, pTRV1 and pTRV2-PDS were transformed using Agrobacterium GV2260, while pTRV2 EV, pTRV2:NbSGT1 (Liu et al., 2002b), pTRV2:GFP (Senthil-Kumar & Mysore, 2014), pTRV2:NbEDS1, and pTRV2:NbNDR1 (Yang et al., 2017) were transformed using Agrobacterium GV3101 (pMP90). Agrobacterium cultures at OD600 = 1 containing pTRV1 or pTRV2 derivative plasmids were mixed in 1 : 1 ratio and infiltrated into two fully expanded leaves of 2- to 3-wk-old plants using a 1-ml needleless syringe (Liu et al., 2002a). All the following experiments were performed 3 wk after VIGS application, except for the time course-dependent Agrobacterium-mediated transient expression that was performed 1-, 2- and 3-wk post-VIGS application. Agrobacterium-mediated transient expression in N. benthamiana and protein detection were performed as previously described (Sang et al., 2018). Briefly, Agrobacterium GV3101 (pMP90) carrying pGWB505 or pGWB511:GUS was infiltrated into leaves of VIGS N. benthamiana for transient expression (OD600 = 0.5 in infiltration buffer). Before infiltration, Agrobacterium strains were incubated in the infiltration buffer (10 mM magnesium chloride (MgCl2), 10 mM 2-(N-morpholino)ethanesulfonic acid (MES) pH 5.8, and 150 μM acetosyringone) for 2 h. GFP fluorescence was observed using a Leica TCS SP8 (Leica, Mannheim, Germany) confocal microscope (excitation: 488 nm, emission: 500–550 nm, 20% strength). Gene expression was analyzed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) (Methods S1; Table S2). For protein accumulation, 10-d-old Arabidopsis seedlings (½ MS plate, 16 h : 8 h long-day growth chamber) and leaf discs (diameter = 18 mm) from N. benthamiana were ground with a Tissue Lyser (Qiagen, Hilden, Nordrhein-Westfalen, Germany) and samples were subsequently re-suspended in protein extraction buffer (100 mM Tris-HCl (pH 8), 150 mM sodium chloride (NaCl), 10% glycerol, 5 mM ethylenediaminetetraacetic acid (EDTA), 5 mM dithiothreitol (DTT), 1% protease inhibitor cocktail, 2 mM phenylmethylsulfonyl fluoride (PMSF), 1% NP40, 10 mM sodium molybdate, 10 mM sodium fluoride, 2 mM sodium orthovanadate), incubated at 70°C for 10 min, and centrifuged at 12 000 g for 5 min before loading in sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels. Western blots were performed using anti-GFP (Abiocode, Agoura Hills, CA, USA, M0802-3a), anti-FLAG (Abmart, Shanghai, China, M20008), anti-AtSGT1a (1 : 3000), anti-mouse IgG-peroxidase (Sigma A2554), and anti-rabbit IgG-peroxidase (Sigma A0545). The authors thank Yasuhiro Kadota and Ken Shirasu for helpful discussions and sharing biological materials, Suomeng Dong for sharing biological materials, Rosa Lozano-Duran for critical reading of this manuscript, and Xinyu Jian for technical and administrative assistance during this work. The authors thank the PSC Cell Biology core facility for assistance with confocal microscopy. The authors have no conflict of interest to declare. This work was supported by the Shanghai Center for Plant Stress Biology (Chinese Academy of Sciences) and the Chinese 1000 Talents Program. GY is partially supported by the China Postdoctoral Science Foundation (grant no. 2016M600339). GY and APM designed experiments. GY, LX and YS performed experiments and analyzed data. APM wrote the manuscript. Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Fig. S1 Representative images showing NbSGT1-silenced Nicotiana benthamiana at different time points. Fig. S2 Characterization of the custom anti-SGT1 antibody. Fig. S3 Silencing NbSGT1 reduces the Agrobacterium-mediated accumulation of GUS. Fig. S4 Silencing NbNDR1 or NbEDS1 does not affect the Agrobacterium-mediated accumulation of GUS. Fig. S5 Silencing SGT1 increases NbPR1 expression. Methods S1 Generation of anti-AtSGT1a antibody, Agrobacterium growth in NbSGT1-silenced Nicotiana benthamiana and qRT-PCR. Table S1 Published studies of cell death induction using Agrobacterium-mediated transient expression in pTRV2-based NbSGT1 VIGS Nicotiana benthamiana. Table S2 Primers used in this study. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.