1. IntroductionVarious plant effector molecules, which interact directly with viral components to inhibit infection, are induced during virus infection [1,2,3,4]. These include RNA-dependent RNA polymerases (RDRs) [1,5], pathogenesis-related proteins [2,6], the inhibitor of virus replication (IVR) [1,7] and Argonautes plus other RNases [3,8]. Most of these factors are induced by phytohormones [4,9,10,11]. RDR1 is induced by several phytohormones involved in defense responses, including salicylic acid (SA), jasmonic acid (JA), ethylene (ET) and abscisic acid [12]. Previously, it was shown that infection of tobacco (Nicotiana tabacum cv. Samsun NN; SNN) by potato virus Y (PVY) stimulated RDR1, as well as several other defense-related genes, namely, alternative oxidase 1 (AOX1), RDR6, IVR and the transcription factor (TF) SHE1 (previously called ERF5) [13]. On the other hand, silencing the expression of RDR1 in transgenic tobacco plants either reduced the level of expression or prevented the expression of these defense-related genes, suggesting that RDR1 had the ability to regulate the expression of such genes [13]. AOX1 was previously known to be upregulated in tobacco by SA, although in a pathway independent of RDR1 induction by SA [14], whereas SHE1 was shown not to be regulated by SA, JA or ET [15]; SA was considered to be unnecessary for induction of IVR (unpublished work given in [1]). More recently, gene expression of RDR1, SHE1 and AOX1 was shown to be upregulated during infection of tobacco by cucumber mosaic virus (CMV), potato virus X (PVX), PVY and tobacco mosaic virus (TMV), although at different amplitudes and to different extents via the different viruses [16,17].IVR and SHE1 are less well-known factors in the pathogen-defense signaling mechanism, although considerable research was conducted on the biology and biochemistry of IVR. IVR is a 21.6 kDa protein [18] that has been shown to be induced after transfection of (SNN) tobacco protoplasts with TMV. After IVR was purified from the culture medium of TMV-infected protoplasts and added to other NN tobacco protoplasts transfected with TMV, it inhibited virus replication [19]. The same inhibitory effect was demonstrated with several viruses (TMV, CMV, PVX and PVY) when applied to either tobacco leaf disks or whole plants recently infected [20,21]. IVR recovered from the intercellular fluid of TMV-infected NN tobacco leaves showed the same inhibitory effects [22]. Constitutive transgenic expression of cloned IVR, under the control of a cauliflower mosaic virus 35S RNA promoter, showed similar levels of resistance to infection by TMV in Samsun tobacco plants normally susceptible to infection (SNN) [23], relative to IVR applied to TMV-infected plants. The in vivo-expressed IVR protein showed a different electrophoretic mobility (c. 23 kDa) to the bacterially expressed IVR protein, suggesting that IVR was post-translationally modified [18]. IVR did not exhibit single-stranded RNase activity [20]. The 27 kDa SHE1 protein was first identified as an AP2/ERF class of TF, after isolation as a new TF able to bind (weakly) to GCC box cis-elements [15]. TMV infection induced SHE1, and transgenic over-expression of this TF was able to reduce the infection by TMV in SNN tobacco, both locally and systemically (the latter at a non-restrictive temperature, i.e., above 28 °C), apparently through a novel signal-transduction mechanism [15]. Later, SHE1 and IVR were analyzed for changes in transcription when nontransformed and RDR1 silenced tobacco plants were infected by either TMV (locally) or PVY (systemically) [13]. More recently, we showed that the expression of SHE1 was induced constitutively in transgenic tobacco plants expressing the CMV 1a protein, and that the CMV 1a protein interacted with SHE1, perhaps to prevent SHE1 from functioning in resistance to CMV [17]. To gain a better understanding of the role of SHE1 in plant defense against viruses, especially TMV, we examined the effect of silencing or over-expressing SHE1 in transgenic tobacco, whether SHE1 was involved in the same unknown pathway as IVR and/or whether SHE1 and IVR interacted with each other. 4. DiscussionThe induction of SHE1 by four viruses, the accumulation of which was not affected at 3 dpi in the infected leaf, indicates that induction of SHE1 does not require the N gene, although the activation of the N gene may greatly increase the level of induction (Figure 1). By contrast, the restriction of infection by TMV-GFP was determined to require the N gene, since it did not occur in either tobacco cv. Samsun NN or tobacco cv. SR1 rendered resistant to WT TMV by transgenically introducing the N gene [25]. However, this restriction response is mediated through effects on TMV-GFP movement [25], even at a temperature at which SHE1 is not expressed [15]. Thus, SHE1 is not a component of the novel resistance response in N gene tobacco plants, preventing the systemic infection of TMV-GFP.The silencing or over-expression of SHE1 had different effects on local infection by WT TMV, with over-expression of SHE1 showing a reduction in virus local movement, as well as the number of local lesions, whereas silencing of SHE1 had no notable effects on either phenotype (Figure 2). Curiously, there was also a change in the physical nature of the lesion on the OEx-SHE1 line, with the lesion being dark brown in appearance (Figure 2A). This is similar to the TMV-induced local lesions seen in leaf disks of both nontransformed and transgenic SNN tobacco plants silenced for the expression of both kinases WIPK and SIPK, while soaking in methyl jasmonate [34]. However, in that case, the lesions were larger with the phytohormone treatment (Figure 5B in [34]), whereas here, the lesions were smaller in the overexpressing plants than in the control. In the case of the effects on local and systemic movement of TMV-GFP at the non-restrictive temperature (for WT TMV), over-expression of SHE1 led to less intense fluorescence in the inoculated leaves than in the nontransformed tobacco, and no systemic infection, as also was the case for the control (Figure 3). However, silencing SHE1 had no effect on local movement of TMV-GFP, with no systemic movement in the first 2 wpi (Figure 3). At later times, WT TMV showed systemic movement at the restrictive temperature for plants silenced for SHE1 expression (Figure 4). Interestingly, TMV also was able to infect systemically the WIKP/SIPK doubly silenced transgenic plants, showing necrotic local lesions in distal leaves [34]. This indicates a perturbation in systemic virus resistance by altering the expression of other genes involved in the signaling defense response, as also was seen here (Figure 4). By contrast, over-expression of SHE1 reduced local accumulation and inhibited systemic infection, as also noted by Fischer and Dröge-Laser [15], although the latter effect could be a consequence of the former effect. The slow systemic movement that occurred when SHE1 expression was blocked by silencing during infection by WT TMV (Figure 4) may reflect non-vascular, systemic cell-to-cell movement, as seen in SNN tobacco plants by TMV mutants in the absence of functional capsid protein [35]. In that case, a slow cell-to-cell movement of the virus occurred to parenchyma cells in the petiole and up the parenchyma cells of the stem, exiting into parenchyma cells of an upper leaf petiole, and moving into mesophyll cells of a partially expanded upper leaf [35]. These effects are reminiscent of the delayed and incomplete systemic infection of transgenic SNN tobacco expressing the bacterial NahG gene, encoding salicylate dehydroxylase and inhibiting the accumulation of SA [36]. Given that there are multiple events in the resistance against TMV in SNN tobacco, involving components such as RDR1, AOX1, pathogenesis-related proteins and IVR [1], the loss of one pathway may partially compromise the overall resistance, but does not abolish it.Transiently expressed SHE1 fused to YFP was found to localize in the nucleus, but when interacting with the CMV 1a protein, SHE1 was found accumulating in the nucleus but also on the tonoplast membrane [17]. The ability of SHE1 to interact with IVR was shown using different approaches (Figure 6 and Figure 7), and the co-localization of the interacting SHE1-IVR complexes with the cytoplasm and tonoplast membrane (Figure 7B,C) suggests that IVR accumulation results in sequestering SHE1 from the nucleus, reducing further IVR gene expression. The co-regulation of the expression of SHE1 and IVR in SHE1 over-expressing and silenced plants (Figure 5; Figure S5) together suggest that IVR is in the same biosynthetic pathway downstream from SHE1. Thus, the binding of IVR to SHE1 to prevent its TF function would an example of end-product feedback inhibition. It is not known whether SHE1 itself binds to the uncharacterized promoter region of IVR, or whether it binds to another promoter for a gene encoding the actual TF for IVR mRNA synthesis.A further aspect of the relationship between IVR and SHE1 can be discerned by the inhibitory effect of high temperature on the expression of SHE1 [15] and IVR, as well as the function of IVR. IVR was found to be undetectable in either the culture fluid of SNN tobacco protoplasts infected with TMV or leaves of SNN tobacco infected with TMV, both maintained at 35 °C, using either serology or bioassay [37]. Interestingly, an amphidiploid interspecific N. glutinosa × N. debneyi hybrid line that showed a constitutive expression of IVR protein, giving rise to very few and smaller lesions when infected by TMV [38], and losing both IVR accumulation and resistance to TMV when maintained at 35 °C [37].The region of SHE1 that was delimited for interaction with IVR (Figure 6) constituted amino acids 86-185 of SHE1 (blue amino acids in Figure 8), although either these sequences or those from amino acids 1-86 could be involved with the proper folding required for the other region to interact with IVR. Nevertheless, this domain overlaps with the conserved DNA binding domain present in AP2/ERF TFs (blue underlined amino acids in Figure 8) [15]. A further deletion from amino acid 185 down to 135 weakened the binding ability of SHE1 for IVR. This region contains most of the amphipathic α–helix from the SHE1 protein (bold blue sequences in Figure 8). Thus, part of the C-terminal half of IVR may be interacting with the SHE1 DNA-binding domain and a key structural feature of this domain, preventing SHE1 from functioning as a TF. The SHE1 deletion construct missing ~20% of its C-terminal region was not as strong at binding IVR as the full-length SHE1 construct (Figure 6). This could be due to some structural alteration in part of the remaining protein, caused by the deletion, or it could be due to the sequences immediately adjacent to amino acid 185 also binding to the IVR. These adjacent SHE1 sequences (amino acids 186–203) contain a potential nuclear localization signal (NLS). Thus, if IVR binds to both the NLS and the DNA binding domain of SHE1, then the bound SHE1 would not be able to enter the nucleus.The published sequence of tobacco SNN IVR [18] has some errors in the C-terminal proximal coding region (Figure S7). These errors were noticed when sequencing the cDNA clone of IVR provided by [18]. In addition, the revised sequence is now nearly identical for this region with IVR sequences in the tobacco genome sequence, as well as IVR-like gene sequences identified in sequenced mRNAs from potato, tomato, pepper, Arabidopsis, rice and wheat [39]. This confirms that the original tobacco IVR cDNA sequence contained three errors: (1) an extra G shown as nucleotide 509; (2) a single rather than a double C present at nucleotide 542; and (3) the TC pair at nucleotides 575–576 (Figure 2 in [18]) should be CT (Figure S7). These errors led to a 12-amino-acid frameshift: 25-36 amino acids from the C-terminus of the 199-amino-acid IVR protein (Figure S7). Nevertheless, the statement by Akad et al. [18] that the C-terminal 120-199-amino-acid region is highly acidic (now containing 21 aspartate and glutamate residues), hydrophobic and containing a helical structure, is still the case. This also is the same region that contains sequences required for interaction with SHE1 (Figure 6). We presume that the level of IVR would be considerably higher than the level of a TF involved in IVR production, which would also facilitate feedback inhibition. However, after about a week, IVR was shown to accumulate in the intercellular spaces of infected SNN tobacco leaves [22]. In addition, this extracellular IVR, as well as the IVR that accumulated in the culture medium of TMV-transfected SNN tobacco protoplasts, between 24 and 96 hpi [19], were identical in electrophoretic movement, which was different from IVR expressed in E. coli [18]. This suggests that the in planta expressed IVR was post-translationally modified. That modification could also have some effect on IVR binding to SHE1.