Finding a good balance: two distinct chromatin factors fine-tune stress response in Arabidopsis seedlings

Organisms maintain their integrity and achieve their goals of survival and reproduction by autonomously processing diverse inputs, such as environmental and intrinsic signals, into integrated responses; examples of this include initiating morphogenesis or changing the rate of growth. Plants, as sessile organisms, must continuously adapt to changing environmental conditions. Notable examples are water scarcity or salinity-induced stress faced by seeds or young seedlings, where the plants respond by delaying germination and by arresting postgermination development. These responses can be reversed upon the return of favorable conditions (Finch-Savage & Bassel, 2016; Brandizzi, 2020). While such plasticity offers advantages in nature, it is often undesirable in agricultural settings, as plants may also restrict their growth in response to moderate or temporary stress signals, which can lead to reductions in yield and harvest delays (Gupta et al., 2020; Zhang et al., 2020). Hence, unraveling the physiological and molecular basis of reaction to stress in Arabidopsis thaliana (Arabidopsis) and other plant species has been a major task in plant biology. It is now well established that stress responses are accompanied by changes in the activity of hundreds, or even thousands of genes, highlighting the remarkable ability of plants to adapt to conditions prevailing at a particular moment. These broad transcriptional responses are triggered and orchestrated by multiple signaling pathways comprising the action of hormones, transcription factors (TFs), and chromatin modifications (Takahashi et al., 2018; Waadt et al., 2022; Zhang et al., 2022). Yet, despite significant progress, the emerging picture is still far from complete. In an article published in this issue of New Phytologist, Perrella et al. (2023; 166–179) report that two distinct chromatin factors, Histone Deacetylase Complex 1 (HDC1) and linker (H1) histones, are involved in shaping the proper transcriptional response to salt stress in Arabidopsis seedlings. They propose a ‘dual brake’ mechanism where HDC1 and H1 impact two different epigenetic marks to suppress some key stress-induced genes, thereby preventing hypersensitive responses.

In addition to its purely structural functions, chromatin is considered to be a major regulatory system that coordinates various genetic networks, including those involved in stress perception and response (Badeaux & Shi, 2013). External and developmental cues can be transmitted through signaling cascades to chromatin-modifying enzymes, including histone acetyltransferases, methyltransferases, and many others (Bannister & Kouzarides, 2011). The introduced modifications can directly or indirectly modulate chromatin structure, affecting its accessibility to different factors (e.g. TFs and polymerases) and, ultimately, gene activity. Transcriptionally active chromatin regions are typically marked by histone acetylation and trimethylation of histone H3 on lysine 4 (H3K4me3), while transcriptionally repressed regions are deacetylated and distinguished by trimethylation of histone H3 on lysine 9 (H3K9me3) or 27 (H3K27me3). Notably, these chromatin modifications can be removed by other enzymes, such as histone deacetylases (HDACs; Millán-Zambrano et al., 2022). The main object of the study by Perrella et al. (2023), the HDC1 protein, has been shown to be a component of HDAC complexes (Perrella et al., 2013; Mehdi et al., 2016), and accordingly, the hdc1 knockout mutation was reported to cause global hyperacetylation of H3 lysines 9 and 14 (H3K9K14; Perrella et al., 2013).

Having previously established that hdc1 mutant seedlings show a hypersensitive response to salt (Perrella et al., 2013), the authors focused on the identification of gene targets regulated by HDC1 under these conditions. Surprisingly, despite the association of histone deacetylation with gene repression and increased global histone acetylation in the mutant, most of the differentially expressed genes (DEGs) found in salt-treated hdc1-1 were downregulated. The authors hypothesized that a small number of direct HDC1 targets tune the sensitivity of stress perception in the seedlings, while most of the observed transcriptional differences are a consequence of enhanced stress sensitivity in the mutant. Therefore, they selected several candidate genes exhibiting H3K9K14 hyperacetylation and upregulation of the transcript in salt-treated hdc1-1 as compared to control. Importantly, they provided genetic evidence in support of the key role of two of these genes by showing that knockout mutants of LATE EMBRYOGENESIS-ABUNDANT (LEA) or MADS AFFECTING FLOWERING 5 (MAF5) both displayed lower sensitivity to salt treatment than wild-type plants and suppressed the salt-hypersensitive phenotype of hdc1-1. Thus, LEA and MAF5 appear to mediate the stress-induced arrest of seedling development. It should be noted that while LEA encodes a member of a well-known protein family with roles in seed development (Candat et al., 2014), until the study by Perrella et al. (2023) the known functions of MAF5 were limited to flowering control (Ratcliffe et al., 2003). It has not been investigated whether two other genes, ABSCISIC ACID INSENSITIVE 3 (ABI3) and RAB GTPASE HOMOLOG B18 (RAB18), that show similar expression and H3K9/16Ac profiles to LEA and MAF5 are also critical for the hdc1 phenotype, and further studies await. Nonetheless, these findings highlight a recognized, but often neglected, possibility that the mutant phenotype can result from a small number of events (e.g. misregulation of few key genes), despite the wide changes in gene activity occurring in parallel.

Histone Deacetylase Complex 1 can interact with components of HDAC complexes as well as with several other proteins, including three variants of the linker histone H1 present in Arabidopsis (H1.1, H1.2, and H1.3; Rutowicz et al., 2015; Perrella et al., 2016). Perrella et al. (2023) revealed that the triple H1 mutation (3h1), similar to hdc1, confers sensitivity to salt. Furthermore, the transcript levels of the ABI3, LEA, MAF5, and RAB18 genes showed a stronger induction upon salt treatment in 3h1 than in wild-type seedlings, although the levels of H3K9/16Ac were not altered. This suggests that HDC1-mediated deacetylation under salt stress is independent of H1 functions, but somehow H1 contributes to the establishment or maintenance of the proper expression levels of key genes. Therefore, the authors investigated other epigenetic marks that were previously reported to be affected in h1 mutants (Zemach et al., 2013; Rutowicz et al., 2019). Remarkably, they found that in wild-type seedlings, salt treatment caused a strong increase in H3K27me3 levels in the studied genes, and both the hdc1 and h1 mutations prevented this effect. This indicates an additional role for HDC1 in facilitating stress-dependent deposition of H3K27me and suggests a functional link between HDC1 and H1. Linker histones have been shown to promote H3K27me3 enrichment at many genes in both animals and plants (Willcockson et al., 2021; Teano et al., 2023), and several other results support H1-HDC1 cooperation, including their direct interaction and nonadditive effects on transcript, as well as H3K9/16Ac and H3K27me3, levels under salt stress in the h1.2 h1.3 hdc1 triple mutant. Unfortunately, the attempts of the authors to generate a full quadruple 3h1 hdc1 knockout were unsuccessful, likely because of its lethality, thereby implying that HDC1 and H1s have overlapping functions in development under normal conditions. Obviously, further studies will be needed to dissect the H1-HDC1 relationship at the molecular level and to resolve some outstanding questions. For example, based on their experiments using a truncated version of HDC1 which is capable of binding H1, the authors have suggested that H3 deacetylation is a prerequisite for H3K27me3 deposition upon salt stress. While this is likely in general, the mechanistic details remain unclear. Other questions that require clarification include: what serves as the primary signal for recruiting PRC2 complexes responsible for H3K27me3 deposition? Is it histone deacetylation itself, the presence of the HDC1-containing HDAC complex, H1, or a combination of these factors? Additionally, it will be important to determine whether H1 binding changes under stress conditions and whether it depends on H3K9/14 (de)acetylation and the presence of HDC1. Intriguingly, in both animals and plants, H1s have been functionally linked with a different histone acetylation mark, H3K56ac (Bernier et al., 2015; Sheikh et al., 2023). Therefore, it would be interesting to test whether HDC1 mediates deacetylation of this and/or other sites, or even H1, under salt stress. In the recent report by Sheikh et al. (2023), changes in H3K56ac levels were correlated with the role of Arabidopsis H1 in the regulation of the defense gene expression after pathogen challenge. However, H3K27me3 levels were not significantly altered in the 3h1 mutant under these conditions. Thus, the function of H1 in promoting deposition of H3K27me3 can be specific to gene and stress type. Finally, given that H1.3 represents a stress-inducible variant of H1 conserved in all angiosperms, the question of whether and how it contributes to the different H1-dependent stress responses remains to be resolved.

From a broader perspective, the findings reported by Perrella et al. (2023) extend our understanding of how plants moderate their developmental responses under adverse conditions. In the model proposed by the authors, HDC1 within the HDAC complex mediates the H3K9/14 deacetylation of the key stress-responsive genes LEA and MAF5, counteracting salt-induced acetylation, which is likely to prevent their excessive upregulation. Additionally, H3K27me3 levels in these genes increase with salt stress (Fig. 1). As the authors propose, the seemingly counterintuitive deposition of a repressive mark on stress-induced genes can act as a ‘dual brake’ to prevent overly sensitive responses. While other explanations are also possible, they would need experimental verification. For example, H3K27 methylation, known for its relative stability (Gallusci et al., 2023), could potentially be deposited to restore efficient growth after the stress conditions disappear. It is also tempting to speculate that the increased H3K27me3 levels are important in limiting the activity of the selected genes following recurring stress. Otherwise, the genes could be primed for high expression, a process known to occur under different stress conditions (Bäurle, 2018). However, this may be undesirable in some circumstances (Crisp et al., 2016), as in the case of LEA and MAF5 which were shown by Perrella et al. (2023) to promote growth arrest. Regardless of possible interpretations, the results of Perrella et al. (2023) may have important implications in the future, as reinforcing the dual brake mechanism could serve as an effective strategy to improve the performance of crop species experiencing moderate stresses in the field.