|Home | About | Journals | Submit | Contact Us | Français|
Much of what we, as plant molecular biologists studying gene regulation, know comes from paradigms characterized or developed in mammalian systems. Although plants, animals and fungi have been diverging for a very long time, a great deal of the machineries and components discovered in yeast and mammals seem to have been maintained in plants. Nevertheless, despite this apparent conservation, evolutionary pressures on the mechanisms of gene regulation are likely to be different between these kingdoms, given their different environmental constraints. As such, it is imperative for plant molecular biologists to develop their own paradigms, even on seemingly conserved systems. It is with this intent that we compare and contrast the regulation of two pathogenesisrelated genes, the Arabidopsis PR-1 and potato PR-10a genes. The transcription factors regulating these genes present prime paradigms for the study of plant signal- and contextdependent dual-function transcription factors.
According to the conventional wisdom, transcription factors are typically classified as “activators” or “repressors”. Activators recruit coactivators, resulting in gene activation, while repressors recruit corepressors, leading to transcriptional repression. In this conventional realm, transcription factors serve only one master, that-is-to-say that their capacity to specifically recruit a coactivator or corepressor is a property intrinsic to the factor itself. In recent years however, it has become evident that transcription factors can be tempted and seduced by more than one master. This treasonous behavior is dictated by the cis-regulatory element to which the factor is bound, the structure of the surrounding chromatin, and the type of molecules available in the nuclear milieu. In essence, transcription factors are good politicians, formulating their function according to their environment. This dual-function role can be exerted in a “context-dependent” or “signal-dependent” fashion.
The context-dependent dual-function of transcription factors typically manifests itself as promoter-dependency, where a factor activates transcription in the context of one promoter but represses it in the context of another. This differential output is the direct result of the sequence of DNA at the cis-regulatory elements occupied by the factor.1–5 A classic example of context-dependent dual-function transcription factor is the glucocorticoid receptor. This Nuclear-Receptor transcription factor is recruited to two similar but different cis-regulatory elements termed the GRE (glucocorticoid response element) and nGRE (negative GRE). Upon binding of the cognate hormone to the glucocorticoid receptor, the complex is recruited to its target cis-elements which impose a certain stoichiometry upon the factor.1 The GRE serves as a scaffold to recruit and assemble a dimeric form of the glucocorticoid receptor, resulting in gene activation, while the nGRE will recruit and assemble trimers of the factor, which will eventually lead to transcriptional repression.
A good example of signal-dependent dual-function transcription factors can be demonstrated by the CCAAT/Enhancer Binding Protein β (C/EBPβ), a basic leucine-zipper (bZIP) transcription factor, and a part of the Ras signal transduction pathway. The repressing and activating functions of this factor are carried out at the same locus through the same cis-element in a promoter. Conversion of C/EBPβ from a transcriptional repressor to an activator is contingent upon phosphorylation.6 Another interesting signal-dependent dual-function transcription factor in which post-translational modification regulates the interplay between repression and activation is Sp3 (Specificity Protein 3). In this case, sumoylated Sp3 functions as a repressor, while acetylation is required for strong activation.7
Just as transcription factors can no longer be catalogued as “repressor” or “activator”, the boundaries between context-dependent and signal-dependent dual-function transcription factors will inevitably get blurry. In this current mini-review, we will highlight the recent realization of the existence of dual-function transcription factors in plants by contrasting the molecular mechanisms governing the regulation of the potato PR-10a gene with that of the Arabidopsis PR-1. PR-10a and PR-1 are, in their respective systems, molecular markers for the activation of inducible defense responses.
The PR-10a gene is activated in response to wounding, infection with the oomycete pathogen, Phytophthora infestans, or treatment with the pathogen-derived elicitor arachidonic acid (AA).8 The repression and activation of PR-10a expression are governed by different transcription factors operating through distinct cis-regulatory elements in the promoter. Repression of the PR-10a gene is mediated by the collective efforts of two transcription factors, the silencer element binding factor (SEBF) and the Pto interacting protein 4 (Pti4), which collectively form a repressosome.9 Pti4 and SEBF are mutually required for their recruitment to the cis-acting silencer element (SE), which spans nucleotides −52 and −27 of the promoter.10,11 In this repressosome, SEBF serves as the DNA-binding component, while Pti4 is essentially a corepressor. ChIP experiments have demonstrated that SEBF and Pti4 are only recruited to the promoter in the uninduced state, but following wounding or elicitor treatment, they are dismissed from the PR-10a regulatory region.9
The SEBF repressor is a 29 kDa protein which localizes to both the nucleus and the chloroplast.11 The negative regulatory properties of the SEBF at the PR-10 gene have been demonstrated using both RNA interference (RNAi) and transcription assay approaches.9,11 However, the repression function of the SEBF is not confined to the PR-10a locus as the factor has also been shown to negatively regulate reporter gene expression through a heterologous DNA-binding (DB) domain.9 The SEBF contains two consensus-RNA binding (cs-RBD) domains, termed cs-RBDI and cs-RBDII, which are responsible for Pti4-interaction and DNA-binding, respectively. Such domains are a defining feature of the heterogeneous nuclear ribonucleoproteins (hnRNPs) that, like the SEBF, are involved in transcriptional regulation and are found in an array of cellular compartments.12–15
Pti4 is an archetypical transcriptional activator of the inducible defense response in plants. The Pti4 factor was initially identified in a screen for interactors of the Pto kinase, which is responsible for conferring resistance to the bacterial spec disease in Solanum lycopersicum (tomato).16–18 The Pti4 is a member of the ethylene response factor (ERF) family of transcription factors found in a variety of plant species.18,19 ERFs are implicated in plant defense programs because they are induced in response to both biotic and abiotic stresses.19 A defining feature of these factors is the ERF DNA-binding domain. This domain specifically binds the GCC-box cis-element, which is a commonly occurring DNA motif in the regulatory region of many PR genes.16,17 Ectopic expression studies performed with the tomato Pti4 in the model system arabidopsis demonstrated that this factor activates the expression of an array of genes, the majority of which contain the GCC-box motif.16,17 Such results indicate that the Pti4 is a conserved transcriptional activator in plants. In contrast, the PR-10a promoter lacks a GCC-box, yet the Pti4 is recruited to this locus and is required for the repression of the gene. Given that the Pti4 behaves as both an activator and a repressor but demonstrates these opposing properties at distinct loci, this factor is a quintessential example of a context-dependent dual-function transcription factor in plants.
The Why1 (StWhy1) transcription activator induces the expression of the PR-10a through the elicitor response element (ERE), which is located between nucleotides −135 and −105.20,21 Under non-inducing conditions the Why1 is localized in the nucleus.21 However ChIP experiments have shown that under such conditions the nuclear localized Why1 is not associated with the PR-10a promoter.20 Prior to elicitation, the Why1 is likely to be distal from the PR-10a, sequestered in an inactive state by way of an uncharacterized inhibitor entity.21 These ChIP studies also demonstrated that the Why1 is only recruited to the promoter following wounding or elicitor treatment.20
Interestingly in the context of the EMSA, both the SEBF and Why1 behaved as strictly single-stranded DNA-binding proteins.11,21 The fact that both of the characterized transcriptional regulators of PR-10a are single-stranded binding factors would tend to indicate that the PR-10a promoter exists in a single-stranded or melted state. This uncommon conformation is likely to require unique regulatory mechanisms and presents a particularly interesting doorway22 for the RNA Polymerase II, which merits investigation.
The PR-1 gene is activated concomitant with the deployment of systemic acquired resistance (SAR), an inducible defense mechanism that is deployed in response to local pathogen attack, producing a long lasting heightened state of disease resistance throughout the plant.23,24 A necessary prerequisite for the establishment of SAR and PR-1 gene activation is the accumulation of the endogenous signaling molecule salicylic acid (SA).24,25 Exogenous application of SA or its chemical analogs, 2,6-dichloroisonicotinic acid (INA) and benzothiodiazole are also sufficient for PR-1 gene activation and the deployment of SAR, in a process referred to as chemical SAR.26 PR-1 gene expression is orchestrated through the concerted efforts of the transcriptional regulator, non-expressor of pathogenesis-related genes 1 (NPR1), the TGA2-containing clade of transcription factors, and cis-regulatory elements residing in the PR-1 promoter region.27–29
The TGA2-clade of transcription factors is required for the basal repression of PR-1.30 The first indication of this phenomenon was demonstrated by the elevated level of PR-1 expression in the tga2/5/6 triple knock-out plant under resting conditions.28 However, genetic and molecular approaches also suggested that TGA2 was a transcriptional activator, recruited to the promoter in an SA- and NPR1-dependent manner.28,31,32 The mechanism preventing the spurious expression of PR-1 remained elusive because it was difficult to comprehend how TGA2 could mediate PR-1 repression, if the factor is not present at the promoter under non-inducing conditions. The unquestionable role of the factor in PR-1 activation had also cast doubt on the repression function of TGA2.
We were able to definitively demonstrate that TGA2 was a constitutive transcriptional repressor through the use of an in planta transcription assay system.30 This system showed that TGA2 could repress an activated reporter gene through the heterologous GAL4 DB in resting or SA-stimulated leaves. This system was further used to demonstrate that the native TGA2 factor could also repress reporter gene expression in the context of the PR-1 promoter and that this repression function was independent of SA treatment.30
Independent ChIP-based investigations have established that, in planta, TGA2 is recruited to the PR-1 promoter in an SA- and NPR1-independent manner.30,33 The in planta transcription assay and ChIP experiments, supported by the PR-1 derepression observed in the tga2/5/6 mutant, demonstrated that the negative regulation of the PR-1 was indeed controlled by the TGA2-clade of transcription factors. Very recently, we have gone on to establish that TGA2-mediated repression is contingent upon the factor's ability to form oligomeric complexes on cognate DNA elements.34 It is unclear at present how the TGA2 oligomer actually represses transcription. However, a possible model could be that the TGA2-dependent oligomer simply functions to occlude the transcriptional machinery from the promoter and in doing so, prevents aberrant PR-1 expression. Such a model has been proposed for the oligomeric TEL transcriptional repressor.35 The TEL has been shown to effect reporter gene repression when operating from a DNA element up to 600 base pairs upstream of the transcriptional start site.36 The TGA2 cognate DNA elements in the PR-1 promoter are situated between 600–700 base pairs upstream of the transcriptional start site, further supporting the possibility of a TEL-type oligomer-mediated repression mechanism. Interestingly, replacement of the TEL homo-oligomerization domain with a bZIP motif (TGA2-type DNA-binding domain and oligomerization motif) did not affect the capacity of the factor to repress reporter gene activity.35
The two TGA-binding sites in the PR-1 promoter reside within cis-elements (LS5 and LS7) that possess contrasting functions. This arrangement presents an interesting regulatory situation particularly in light of the high molecular weight complex we now know occupies the promoter under resting conditions. The LS7 promoter element is required for SA-inducible PR-1 activation, whereas the LS5 negatively regulates PR-1 expression before and after SA-treatment.27 Unfortunately, with the limited resolution of the ChIP assay, we cannot determine which of these two sites is occupied by the TGA2/5/6-oligomer.30 Fluorescence anisotropy studies indicate that TGA2 has equal affinity for both the LS7 and LS5 sites in vitro.34 Given that these elements are separated by a mere 20 nucleotides,27 we cannot rule out the possibility that both positions are bound by the oligomeric complex. In vivo footprinting experiments performed on the PR-1 promoter found that under non-inducing condition the LS7 is protected.27 This protection may well be the result of the TGA2/5/6-oligomer. It could be reasoned that the TGA2/5/6-oligomer effects PR-1 repression by obstructing this positive cis-acting element.
There is some evidence that basal repression of PR-1 is also mediated in part by histone modifications. The SNI1 protein is proposed to function as a scaffolding protein recruiting various chromatin modifying activities.37 Plants bearing an sni1 mutation demonstrate increased basal PR-1 expression.37,38 The elevated basal expression of PR-1 in the sni1 mutant is attributed to modest increases in H3K4 methylation and general acetylation of H3 at the promoter.37 This explanation conflicts somewhat with the findings of Koornneef and colleagues (2008),39 which reported that PR-1 could be activated without significantly altering the general acetylation of H3 at the promoter. These results suggest that H3 acetylation does not control PR-1 expression. Such data coupled with the very mild changes in histone marks observed in the sni1 mutant do not make a convincing case that the factor negatively regulates PR-1 expression through chromatin modifications. However, it is not possible to discount that histone modifications contribute to the repression of the PR-1. It is widely acknowledged that the chromatin context is critical to basal transcriptional repression in eukaryotes and it undoubtedly plays a role in the negative regulation of PR-1, but the current studies have yet to identify the critical histone marks and chromatin modifying agents involved.
In unstimulated cells, the NPR1 protein is localized to both the nucleus and the cytoplasm.40,41 Furthermore, it has been determined that, not only is the NPR1 coactivator present in the regulatory region of the PR-1 gene under non-inducing conditions, but its recruitment to the promoter is independent of the TGA2-clade of transcription factors.30 NPR1, like most coactivators, does not possess a known DNA-binding domain and is therefore likely to be recruited to the repressed PR-1 locus by way of a bridging protein. At present there is no information as to what this NPR1-recruiting entity might be, nor is there any indication as to the role of NPR1 in this non-inducing situation. While the presence of coactivators at repressed promoters defies convention, a similar situation has been documented in Drosophila using the ChIP technique.42 It can be reasoned that the proximity of latent coactivators to cis-regulatory elements poises them to more readily induce gene expression in response to the appropriate stimulus. Despite the presence of NPR1 at the PR-1 promoter in resting cells, npr1 mutations do not affect the basal repression of the locus.43
Transcriptional reprogramming is critical to the deployment of plant inducible defenses. Both PR-1 and PR-10a are activated as components of defensive programs in arabidopsis and potato, respectively. Despite the fact that these PR genes are expressed under similar circumstances, mechanistic studies have demonstrated that the mode of induction is not conserved between the PR-1 and PR-10a.
In response to SA, the TGA2-clade of transcription factors retain their capacity for repression,30 but interaction with the NPR1, specifically involving the BTB/POZ domain, alters the conformation of the TGA2 at the PR-1 promoter.34 The repressive oligomeric TGA2 complex is cleared or redistributed such that only a low-order TGA2 structure, presumably a dimer, is present at the promoter in the context of an NPR1-TGA2 transactivating complex with a likely stoichiometry of 2 NPR1:2 TGA2.30,34 While it is tempting to believe that, in an activating situation, TGA2 is localized to the positive cis-acting PR-1 element, LS7, again it is impossible to specify the location of the factor on the promoter due to the resolution limits of the ChIP procedure.
Since TGA2 resides at the PR-1 promoter under both resting and inducing conditions and given that it is essential for both basal repression and activation of the gene, it could be reasoned that the factor's function is controlled by a simple signal-type mechanism mediated by post-translational modification, stimulated by SA treatment. However, as mentioned previously, in the absence of NPR1, TGA2 retains its capacity for repression following SA stimulation at both the PR-1 and in a reporter gene context.30 Casting further doubt on a simple signal-type mechanism dictating the duality of TGA2 is the finding that the mutation of residues known to be phosphorylated in the factor following SA treatment do not affect its capacity to induce PR-1 expression.44 At present, we understand that TGA2 must interact with the NPR1 protein in order to realize its activator function. However, while it is clear that this interaction occurs in response to SA stimulation, the mechanism and the nature of the signal regulating the interaction of these two constitutive PR-1 promoter residents remains uncharacterized. Nevertheless, given that both the repression and activation functions of TGA2 are manifested at the same locus (PR-1), this factor constitutes a paradigm for signal-dependent dual-function transcription factor in plants.
The activation of the PR-10a is a much different event from that of the PR-1 because, unlike the PR-1, the activating agents are not constitutive occupants of the PR-10a promoter. Under resting conditions the PR-10a promoter is bound by the SEBF-Pti4 repressosome.9 Upon wounding or elicitation, these negative regulators are cleared or dismissed, and the activator Why1 is recruited to the ERE promoter element at the PR-10a. Despite the unusual ssDNA-binding nature of the factors responsible for regulating expression of the PR-10a, this gene is controlled in a far more conventional fashion than is the PR-1. In the case of the PR-10a, repression and activation are effected through entirely different repressors and activators, which are recruited in a mutually exclusive manner to the promoter, where they occupy distinct negative and positive DNA elements, respectively. The situation at the PR-1 is much different because the TGA2 functions in both repression and activation and the NPR1 coactivator is always present at the promoter, even under non-inducing conditions.
Based on the current data, the PR-1 and PR-10a promoters seem to be much different environments under activating conditions. Unfortunately, very little is known about the actual DNA and chromatin architectures at either locus under these circumstances. Limited investigations into the histone marks associated with PR-1 expression have been conducted, but they are in many ways conflicting and, as a result, it remains unclear how such modifications contribute to the activation of this PR gene.24,37,39 This aspect of PR-10a regulation remains entirely unaddressed, but it would be very interesting to determine if this seemingly single-stranded locus is capable of supporting nucleosomes under any conditions.
The Why1 activator provides an interesting link between the mechanisms of activation for the PR-1 and PR-10. The Arabidopsis Why1 ortholog (AtWhy1) is required for the activation of PR-1 and the inducible defense response program SAR.20 EMSA analysis also demonstrated that the binding activity of the AtWhy1 for the 3′ portion of the ERE, termed the PBF-2 binding (PB) element, was stimulated in an SA-dependent, NPR1-independent manner. While recruitment of the Arabidopsis ortholog AtWhy1 to the PR-1 promoter has not been investigated, the recruitment pattern of the potato Why1 (StWhy1) is very conventional in that it is absent from its cognate site under repressing conditions, but is drafted to its regulatory element after defense-inducing cues.20
A common theme that emerged through the study of the regulation of PR-1 and PR-10a is the involvement of dual-function transcription factors. We demonstrated that the TGA2-clade of transcription factors is required for both the activation and repression of the PR-1 and we also showed that the known activator Pti4 is essential for the recruitment of the SEBF repressor to the PR-10a. It is rather surprising that one of the very few unifying features found in the regulation of these PR genes should be the involvement of uncommonly treasonous transcription factors. While such behaviors are deemed unconventional, there are a number of factors that are known to deviate from their designations as activators and repressors.1,3 While the PR-1 and PR-10a genes are induced by similar functional cues, that-is-to-say cues associated with disease resistance, their regulation is governed by very distinct molecular mechanisms, which are summarized in Figure 1.
The Arabidopsis PR-1 and potato PR-10a regulating factors present prime paradigms for the study of plant signal- and context-dependent dual-function transcription factors, respectively. The detailed models that have been evolved for the mechanisms regulating these loci now rank among the best descriptions of gene regulation in plants and will serve to reduce our reliance upon dual-function factor paradigms derived from other kingdoms. Given that plants present their very own idiosyncratic mechanisms to govern gene regulation, we would argue that greater emphasis should be placed on in-depth molecular studies of plant promoters and their associated trans-regulators.
We thank Ms. Jee Yan Chu for editorial assistance. Research in CD's lab is supported by the National Science and Engineering Council (NSERC) of Canada. P.B. was supported by an NSERC post-graduate scholarship and by the Ontario Graduate Scholarship program.
Previously published online: www.landesbioscience.com/journals/psb/article/11570