INTRODUCTION
Plants employ the heat shock response (HSR) as a means to acclimate to heat stress caused by elevated temperatures1–4 in response to heat shock. Plant cells elevate expression of molecular chaperones to ensure protein stability, correct folding, triage, and degradation, thereby restoring protein homeostasis (proteostasis).5–8 The pivotal controllers of chaperones are heat shock transcription factors (HSFs).9–12 In the presence of high-temperature stimuli, HSFs amass in the nucleus and initiate the upregulation of chaperones and other genes associated with the HSR.13–15 Additionally, it has been observed that excessive light or high light intensity triggers the translocation of HSF gene family member HsfA1d into the nucleus, promoting an increase in heat shock protein (HSP) gene expression in Arabidopsis.16,17
On Earth, fluctuations in sunlight intensity correspond closely with shifts in temperature. Over time, plants have evolved mechanisms to interlink light and temperature sensing, given that exposure to light is recognized to augment resistance to heat.18 Genetic investigations have indeed linked plant survival at elevated temperatures to chloroplast signaling.19 Analyses of diurnal gene expression have unveiled the regulation of chaperone protein genes by light/dark cycles in Arabidopsis thaliana.19,20 Moreover, light has been shown to be involved in the amplification of the activity of HSRs and unfolded protein responses in the endoplasmic reticulum, both of which are integral to the HSR.16,21,22 It has been documented that the red light receptor phytochrome B (phyB) collaborates with chloroplasts as a temperature sensor, harmonizing light and temperature signals. As the temperature rises, phyB activity is repressed, leading to a reduction in phyB nuclear body size. Furthermore, genes regulated by phyB are modulated in response to temperature changes.23,24 Additionally, the blue-light receptor CRY1 interacts with phytochrome-interacting factor 4 (PIF4) to suppress auxin biosynthesis at elevated temperatures, thereby curtailing hypocotyl elongation in a process termed thermomorphogenesis.25 Collectively, these findings underscore the close intertwining of light and heat signaling, with light receptors serving as pivotal connectors between these two pathways.
However, it is noteworthy that the bulk of these studies have been conducted under conditions of moderate heat stress rather than in the face of severe or potentially lethal heat levels. Consequently, whether light plays a role in promoting adaptation to exceptionally high temperatures (i.e., thermotolerance) remains an enigma. Moreover, the precise light receptor governing plant thermotolerance and the mechanisms by which it exerts its influence remain elusive. In this study, we delve into the role of the blue-light receptor CRY1 in the induction of light-induced thermotolerance in plants. Through its capacity to interact with HsfA1d, CRY1 facilitates the nuclear import of HsfA1d, orchestrated by nuclear importin alpha 1 (IMPα1), thereby modulating the binding of HsfA1d to chromatin under light conditions. Our investigation dissects the intricate mechanisms that unite light signaling with the HSR.
RESULTS
Light-induced thermotolerance in plants
Prior research has elucidated that the thermotolerance of plants under light conditions is regulated by chloroplast signaling.19 To discern the specific influence of chloroplasts and ascertain whether the light-signaling pathway independently triggers heat tolerance, we conducted an experiment using three sets of Arabidopsis plants. These plants were initially grown in darkness at 22°C for a duration of 4 days. Subsequently, one set remained in darkness at 22°C, another was subjected to white light exposure (50 μmol m−2 S−1) for 1 h at 22°C, and the third set was exposed to 34°C for 30 min in the dark (Figure S1A). Following this, all plants were exposed to a lethal temperature of 45°C for 2 h, followed by a recovery phase at 22°C under white light for 7 days. Intriguingly, our observations revealed that pre-exposure to light facilitated thermotolerance akin to pre-treatment at 34°C. In stark contrast, plants that were continuously grown in darkness succumbed to heat stress (Figure 1A). This enhancement of thermotolerance subsequent to light treatment was also evident in wheat and canola (Figure S1). We termed this phenomenon light-induced thermotolerance (LIT). We further sought to investigate whether the LIT phenomenon remains valid for plants cultivated under light conditions. To address this, we cultivated two sets of Arabidopsis plants in low-intensity light (50 μmol m−2 S−1) for 4 days. Subsequently, the light intensity was augmented to 75 μmol m−2 S−1 for a duration of 1 h for one set, while the other set was maintained at the initial light intensity prior to subjecting them to daytime heat shock. Our findings demonstrated that elevating light intensity for 1 h heightens the heat resistance of Arabidopsis plants (Figure S1D).
Additionally, we sought to determine whether the LIT phenomenon is influenced by different photoperiods. We cultivated Arabidopsis seedlings under varying photoperiod conditions (day/night = 16 h/8 h, 12 h/12 h, or 8 h/16 h) for 4 days. Following this, the seedlings were pre-treated with 1 h of light exposure at midnight, followed by a heat shock at 45°C for 1 h. After a recovery period of 4 days, our results unveiled that longer daily light durations correlate with heightened thermotolerance in Arabidopsis seedlings. Notably, pretreatment with light exposure at midnight conferred a significant level of thermotolerance across all examined photoperiods (Figure S2A). Furthermore, extended daily light durations even augmented the thermotolerance of seedlings during the daytime (Figure S2B). These findings collectively suggest that a longer circadian rhythm day length contributes positively to the thermotolerance of Arabidopsis seedlings. Moreover, the LIT phenomenon demonstrated its applicability to seedlings cultivated under diverse photoperiod conditions.
Light-induced expression pattern resembles heat shock genes
In the realm of plants, the activation of the HSR in general and the specific transcriptional upregulation of HSP101 (encoding a disaggregate chaperone) have paramount significance for survival during instances of heat stress.26,27 Our investigation confirmed the increase in HSP101 mRNA expression following light exposure and heat treatment through semi-quantitative PCR (semi-qPCR) and qPCR analyses (Figure 1B).
To ascertain the extent of transcriptional response commonality between light treatment and heat stress, we conducted mRNA deep sequencing (mRNA-seq) on dark-grown etiolated Col-0 seedlings subjected to either 1 h of white light exposure or a temperature of 34°C for 30 min. The transcriptome-wide analysis revealed substantial concordance between gene expression in response to light and heat (Figures 1C–1E; Table S1). Remarkably, more than half of the genes induced by light exposure can also be induced by heat stress, culminating in a total of 582 genes commonly triggered by both stimuli (Figures 1C and 1D). The Gene Ontology (GO) terms were enriched, and the top 10 of each sector were listed (Figures 1E and S3). The results demonstrated that the 1,090 genes induced by light exposure were enriched in functions related to chloroplasts, photosystems, HSRs, light responses, and translation elongation factor activity (Figure S3A). The 2,518 genes induced by heat stress were primarily enriched in processes such as protein folding, chloroplast and photosystem structure and function, water transport, and translation elongation (Figure S3B). Notably, the 582 genes that responded to both light and heat exhibited enrichment in categories including protein folding, light response, translation elongation factor activity, hydrogen peroxide transmembrane transport, chlorophyll binding, and photosystem reaction (Figure 1E).
Conversely, treatments of 34°C heat and light pre-exposure resulted in the decreased expression of 2,081 and 909 genes, respectively. Interestingly, among these downregulated genes, 286 were shared between the two conditions (Table S1). GO analysis highlighted that these co-downregulated genes were implicated in light intensity response, lipid and hormone responses, transcription and gene expression, cellular components of lipid droplets, nuclei, and plastids (Figures S3C and S3D; Table S1). Notably, among the common genes induced by light and heat, more than 22 were annotated as HSPs or HSP-like genes, underscoring the shared capacity of light and heat to induce the HSR. These findings unveil a mutually reinforcing relationship between exposure to light and heat stress.
Blue light initiates HSR through CRY1
Plants exhibit distinct responses to different wavelengths within the white light spectrum. To explore the feasibility of inducing LIT through specific spectral ranges, we pre-conditioned Col-0 seedlings with light spanning far-red-, red-, or blue-spectrum light before exposing them to 45°C. Notably, seedlings subjected to pre-exposure with far-red or red light failed to manifest thermotolerance, whereas those pre-exposed to blue light exhibited enhanced survival rates akin to those pre-exposed to white light (Figure 2A). Significantly, only blue light pre-exposure led to the activation of HSP101 gene expression (Figure 2C).
Among light receptors, Cryptochrome 1 and 2 (CRY1 and CRY2) are attuned to blue light, while phytochrome A (phyA) and phyB are responsive to red and far-red light, respectively.28–31 In an endeavor to pinpoint the specific light receptor responsible for thermotolerance, we evaluated canonical heat-induced thermotolerance and LIT in null mutant lines (phyA-211, phyB-9, cry1–304, and cry2–1). The survival rates of all tested lines, including Col-0 and the mutant lines, were enhanced upon pre-heat treatment at 34°C before exposure to severe heat stress (Figure S4). Notably, LIT was evident in all mutant lines except for cry1–304, which exhibited a marked impairment in acquiring LIT, resembling the survival rate of dark-grown Col-0 (negative control), where less than 10% of seedlings survived (Figure 2B). While cry2–1 demonstrated a deficit in LIT, the extent of its sensitivity was notably lower compared with the cry1 mutant (Figure 2B). Corresponding with the phenotypic findings, cry1–304 displayed diminished induction of HSP101 in response to white light (Figure 2D).
Moreover, the impairment of LIT was further substantiated through the assessment of another cry1 null mutant, cry1–349, whose phenotype aligned with the lack of LIT induction upon blue light exposure. Intriguingly, in cry1–349, it was the pre-heat treatment at 34°C rather than light treatment that triggered seedling thermotolerance (Figure S5). These findings collectively underscore the pivotal role of the blue light receptor CRY1 in initiating LIT.
CRY1 interaction with HsfA1d
Employing the yeast two-hybrid system, we identified that HsfA1d-NT (1–220 amino acids [aa]) interacts with the full-length CRY1, its N-terminal segment (1–490 aa), and the C-terminal fragment (491–681 aa) of CRY1. This suggests that HsfA1d exhibits the capability to interact with at least two distinct regions of CRY1 (Figure 3A). We corroborated the interaction between full-length CRY1 and HsfA1d through a split-luciferase (split-LUC) assay within the Nicotiana benthamiana transient expression system (Figure 3B). Further substantiating this interaction, we introduced a 35S:: HsfA1d-FLAG expression cassette into Arabidopsis (HsfA1d-FLAG/Col-0) and observed a marked enhancement in Arabidopsis survival rates following 45°C treatment when HsfA1d was overexpressed (Figure S6A). Through immunoprecipitation (IP) utilizing an anti-FLAG antibody against the FLAG-tagged protein in the HsfA1d-FLAG/Col-0 line, we confirmed the interaction between CRY1 and HsfA1d (Figure 3C). Moreover, the interaction between CRY1 and HsfA1d in the nucleus was established through bimolecular fluorescence complementation (BiFC) (Figure 3D). Collectively, these findings underscore the stable interaction between CRY1 and HsfA1d in Arabidopsis. Notably, HsfA1d belongs to the HsfA1 subfamily, which serves as a crucial master transcriptional regulator of the HSR in Arabidopsis,1,32 making it a compelling candidate for mediating LIT.
HsfA1d mediates LIT
The HSR is evolutionarily conserved among eukaryotes, and HSFs stand as the principal transcriptional regulators within this response. Arabidopsis possesses 21 HSF homologs, while yeast only contains a solitary HSF gene.9,10 Among these, the HsfA1 subfamily encompasses HsfA1a, HsfA1b, HsfA1d, and HsfA1e and emerges as the primary regulator of the induced HSR in Arabidopsis. Intriguingly, our investigation illuminated that pre-treatment with light failed to elevate the survival rate of the HsfA1 quadruple mutant (hsfa1sqk) when subjected to 45°C exposure. Moreover, expression of the marker gene HSP101 exhibited poor induction in response to either light or heat exposure (Figures 4A and 4B). Remarkably, the reintroduction of HsfA1d into hsfa1sqk (hsfa1a/b/e mutant) effectively reinstated light- and heat-induced thermotolerance, concurrently re-establishing the expression of HSP101 (Figures 4A and 4B). Consequently, the HsfA1 subfamily is deemed essential for LIT, with HsfA1d emerging as the pivotal mediator sufficient to trigger this phenomenon.
Joint regulation of heat shock gene expression by HsfA1d and CRY1 in response to light
To comprehensively understand the combined influence of CRY1 and HsfA1d on the transcriptional responses triggered by light and heat, we conducted RNA-seq analysis on dark-grown Col-0, cry1–304, and hsfa1sqk seedlings following exposure to white light for 1 h or 34°C heat for 30 min (Table S1). In the hsfa1sqk mutant, among the 154 annotated light-responsive genes (Figure 4C, denoted in blue), 12 genes (8%) exhibited differential expression (fold change > 2 or < 0.5, adjusted p < 0.05) relative to Col-0 in response to light and heat shock. Strikingly, for the 186 annotated heat-responsive genes (Figure 4C, highlighted in red), 52 genes (28%) displayed differential expression in the hsfa1sqk mutant compared with Col-0 in response to light and heat shock. In the case of the cry1–304 mutant, only 3 heat-responsive genes demonstrated reduced expression (fold change ≤ 0.5) compared with Col-0 under heat stress, while 63 genes exhibited reduced expression in response to light. Similarly, under heat stress, only 2 light-responsive genes exhibited reduced expression in cry1–304 relative to Col-0, in contrast to 32 genes exhibiting reduced expression in response to light (Figure 4C). These observations suggest that HsfA1 transcription factors are instrumental in triggering the transcriptional response to heat shock under light and heat conditions. However, CRY1’s involvement is specific to the light-induced HSR.
Subsequently, to delineate the transcriptional signature associated with LIT, we identified the set of genes whose induction is dependent on the HsfA1 family and CRY1. In the hsfa1sqk strain, 74 genes exhibited a reduction in expression of >2-fold relative to Col-0 in response to heat, and 70 genes exhibited a similar reduction in response to light. Impressively, 54 of these genes showed reduced expression in response to light and heat stimuli (Figure 4D; Table S1). Among these 54 genes, 42 genes demonstrated a reduction of more than 2-fold in cry1–304 relative to Col-0 (Figures 4D and 4E). This final set of 42 genes predominantly encompassed HSP genes and HSF genes (Figure 4E). We designate these 42 genes as the set of genes co-regulated by CRY1 and the HsfA1 family in response to light, making them potential contributors to LIT.
Shared target genes of CRY1 and HsfA1d
Given the nuclear interaction between CRY1 and HsfA1d, coupled with the transcription factor role of HsfA1d, it is plausible that the CRY1-HsfA1d complex binds to chromatin, thereby influencing gene expression. To explore their potential co-occupancy on chromatin sites, we conducted chromatin IP sequencing (ChIP-seq) analysis on 5-day-old HsfA1d-FLAG/Col-0 and GFP-CRY133 plants, cultivated under a 16 h/8 h light/dark regimen (Table S2). The ChIP-seq data analysis unveiled that HsfA1d-FLAG and GFP-CRY1 were bound to 4,917 and 2,327 chromosomal regions, respectively, with a notable overlap of 1,483 sites (Figure 5A). Among the 42 genes exhibiting reduced expression under light and heat stimuli in the hsfa1sqk and cry1–304 mutants relative to Col-0, we identified 35 and 15 genes as targets of HsfA1d and CRY1, respectively, with a remarkable 12 genes being common to both (Table S1). This observation indicates that genes associated with the HSR are potentially regulated directly by HsfA1d and CRY1. Intriguingly, the co-occupying genes extend beyond core heat stress response genes like HSFB1 and HSFB2b. Among them RAV2, a gene linked to drought, salt, and oxidative stress, which are hallmarks of heat stress, also stands out. RAV2’s involvement in photo-protection against abiotic stress further suggests a possible connection between biotic stress and light signaling.34
Considering the divergence in binding maps between HsfA1d-FLAG and GFP-CRY1 lines as deduced from the ChIP-seq data, we postulate that the transcriptional co-occupancy of genes may be altered between the cry1–304 and hsfa1sqk mutants. Of note, we observed the presence of bromodomain-containing 4 (BRD4), a gene belonging to the bromodomain and extraterminal family, among the co-occupying genes of HsfA1d and CRY1. Notably, BRD4 plays a role in regulating transcription.35 Further analysis revealed changed binding patterns of HSFB1, HSFB2b, RAV2, and BRD4 by CRY1 and HsfA1d based on the ChIP-seq data (Figure 5C). This observation was substantiated by ChIP-qPCR experiments, which confirmed significantly altered associations of CRY1 and HsfA1d with these four target genes between the HsfA1d-FLAG and GFP-CRY1 lines (Figure 5B). Taken together, these findings underscore the physical interaction of CRY1 and HsfA1d with common gene sequences, including those pivotal for thermotolerance.
CRY1 facilitates nuclear localization of HsfA1d under blue light
Having observed the nuclear interaction between HsfA1d and CRY1 in A. thaliana16 (Figure 3D), we proceeded to investigate whether the nuclear localization of both proteins is influenced by light exposure. We initially examined the nuclear localization of CRY1 in response to light using 35S::GFP-CRY1 (GFP-CRY1) transgenic Arabidopsis seedlings33 (Figure S6B). Employing nuclear and cytoplasmic protein fractionation followed by western blotting, we observed increased nuclear accumulation of GFP-CRY1 after exposure to light (Figures 6A, left, and S7A). To delve further, we conducted a fluorescence-based analysis of nuclear accumulation of GFP-CRY1 through fluorescence microscopy and bleach recover after blue light treatment and dark (DK) conditions. The study revealed that blue light not only enhanced the nuclear accumulation of GFP-CRY1 in terms of quantity but also expedited the recovery rate after photobleaching (Figures S8A and S8B). Specifically, under DK conditions, the fluorescence intensity (FI) of nuclear GFP-CRY1 was only restored to approximately 66% of its initial value after bleaching, whereas with blue light exposure, FI restoration reached around 90%. Furthermore, confocal microscopy demonstrated a notable increase in global GFP FI in the hypocotyls of Arabidopsis seedlings subjected to light, underscoring the overall elevation in CRY1 accumulation (Figure S8C). In essence, blue light was found to augment the accumulation and nuclearization of CRY1 in Arabidopsis.
Conversely, we also noted that heat stress at 34°C triggered the nuclearization of HsfA1d (Figure S7D). To explore whether light has a similar effect on the nuclear localization of HsfA1d, we tracked localization of HsfA1d in HsfA1d-FLAG/Col-0 lines (Figure S6C). Employing fractionation and Western blot analysis, we demonstrated that light exposure also enhanced the nuclear accumulation of HsfA1d (Figures 6A, right, and S7B). Consequently, light exposure exhibited the ability to bolster nuclear localization of CRY1 and HsfA1d.
In the context of Arabidopsis, nuclear localization of the far-red receptor phyA is reliant on the interaction between far-red elongated hypocotyl 1 (FHY1) and IMPα1.36 Intriguingly, our investigations unveiled interactions between CRY1 and HsfA1d, respectively, with IMPα1 in the yeast two-hybrid system (Figure 6B), and these interactions were further corroborated in the split-LUC assay (Figure 6C). Notably, the presence of CRY1 was found to enhance the interaction between HsfA1d and IMPα1 (Figure 6D). Importantly, the effects of blue light on the interaction of CRY1-HsfA1d and CRY1-IMPα1 were examined through the split-LUC assay, revealing significant intensification of the interaction in response to blue light treatment (Figure S10). These findings led us to speculate that, in the presence of blue light, CRY1 may facilitate nuclear localization of HsfA1d through a physical interaction, potentially aided by IMPα1.
To test this hypothesis, we conducted a cross between HsfA1d-FLAG/Col-0 and the cry1–304 mutant, yielding HsfA1d-FLAG/cry1–304 (Figure S6D). Subsequent analysis involving fractionation of nuclear and cytoplasmic proteins, followed by western blotting, revealed that light-induced nuclear localization of HsfA1d was enhanced in HsfA1d-FLAG/Col-0 but remained unaltered in HsfA1d-FLAG/cry1–304 (Figures 6E and S7C). Hence, our results conclusively demonstrate that CRY1 actively promotes light-dependent nuclear localization of HsfA1d.
Decreased binding loci and affinity of HsfA1d in cry1 mutants
To delve deeper into the impact of CRY1 on chromatin binding of HsfA1d, we conducted ChIP-seq of HsfA1d in HsfA1d-FLAG/cry1–304 mutants under the same growth conditions as previously detailed (Table S2). The result unveiled a stark reduction in the number of HsfA1d binding sites in the cry1–304 background (721) compared with the Col-0 background (4,917), with 502 sites overlapping between the two (Figures 5A and 7A). Of significance, there were 190 overlapping binding sites between GFP-CRY1 and HsfA1d-FLAG in HsfA1d-FLAG/cry1–304 (GFP-CRY1 vs. HsfA1d-FLAG/cry1–304), exhibiting a notable decrease in contrast to the 1,483 overlapping sites in HsfA1d-FLAG/Col-0 (Figures 5A and 7B). These findings strongly suggest that, in cry1–304 mutants, about the majority (approximately 89%; Table S2) of shared binding loci between CRY1 and HsfA1d are lost.
Further assessment of CRY1’s impact on HsfA1d′s chromatin affinity involved counting binding peaks and generating a heatmap of HsfA1d binding to transcription start sites (TSSs) in HsfA1d-FLAG/Col-0 and HsfA1d-FLAG/cry1–304 mutant backgrounds. The heatmap clearly portrays diminished HsfA1d binding signals in the HsfA1d-FLAG/cry1–304 background, characterized by a dramatic reduction in binding sites at TSSs (Figures 7C and 7D). Moreover, IGV plots (Itegrative Genomics Viewer) illustrated that the binding peak of HsfA1d-FLAG targets, particularly BRD4, in the cry1–304 background were significantly diminished, as validated by ChIP-qPCR (Figures 7E and 7F). qRT-PCR results supported this observation, showing that, in terms of control and overexpression levels of HsfA1d, the transcription levels of HSFB1, HSFB2b, and BRD4 in the cry1–304 background were markedly decreased in comparison with the Col-0 background (Figures S9A and S9B). This indicates that the absence of CRY1 might exert an influence on the binding site and binding affinity.
The complex formed by CRY1 and HsfA1d governed the expression of a specific gene set, including heat response genes, which contribute to a plant’s survival under high temperature stress (Figures 4C–4E). In line with this, the overexpression of GFP-CRY1 in Col-0 led to an enhanced survival rate after exposure to harsh temperature treatment (Figure 7G), and the introduction of MYC-CRY1 complemented the heat tolerance phenotype observed in the cry1–304 line (Figure S9C).
DISCUSSION
This study uncovers a novel interplay between light signaling and the HSR, shedding light on the direct correlation between light exposure and the HSR. While previous studies have shown that high light/high irradiation triggers stress responses separately,16,17,37,38 transcriptome-wide analysis disclosed a significant overlap between light-induced genes and heat-induced genes, particularly in terms of functional categories such as the light-harvesting complex and HSPs (Figures 1C–1E and S3), indicating coordinated signaling that enables plants to pre-adapt to rising temperatures via LIT.
The investigation demonstrated that LIT is prompted by blue light, specifically through the action of the blue light receptor CRY1, rather than red or far-red light (Figure 2), unlike the PIF-dependent activities previously linked to CRY1, which revolve around photomorphogenesis at elevated temperatures, not thermotolerance.25,39 CRY1 physically interacts with HsfA1d in the nucleus, and plants that lose HsfA1s failed to trigger LIT (Figure 3D). It is speculated that the interaction between CRY1 and HsfA1d in the nucleus enables light-dependent nuclear localization of HsfA1d to initiate the light-dependent HSR and promotes LIT. Co-occupancy of these two factors at promoters implies their involvement in transcriptional regulation. CRY1 and HsfA1d share many common genomic binding sites, including heat stress genes such as HsfB1, HsfB2b, and HSPs (Figure 5). Importantly, CRY1 is responsible for the light-induced heat stress response but not high-temperature-induced heat stress. Nevertheless, HsfA1d is involved in light- and heat-induced heat-stress responses (Figure 4A). It indicates that CRY1 perceives light signals and transduces them to downstream genes through its interaction with HsfA1d. Thus, both CRY1and HsfA1d are required for LIT.
HsfA1s, including HsfA1a, HsfA1b, and HsfA1e, make up a major transcriptional family involved in acquired heat tolerance.40 This family triggers a transcriptional cascade composed of multiple transcription factors under high-temperature conditions, which is important for the activation of the HSR transcriptional regulatory network. While HsfA1d is pivotal for LIT, the roles of other members of the HsfA1 subfamily (HsfA1a, HsfA1b, and HsfA1e) remain to be explored because they often exhibit redundant functions. Indeed, we verified the interaction of CRY1 and HsfA1a/b/e (Figures S11A–S11C). As we know, there are two cryptochromes (CRY1 and CRY2) responsible for regulating various blue light responses in Arabidopsis.41 Moreover, considering the similarity in sequence and function between CRY1 and CRY2, it is plausible that CRY2 could also contribute to LIT, which was verified by a heat resistance test of the cry2–1 mutant line (Figure 2B), although it is less sensitive regarding heat tolerance than cry1–304, which may result from the variation of their roles in blue light signaling. Further, like CRY1, CRY2 also physically interacts with HsfA1d and IMPα1 (Figure S11D and S11E). The mechanism behind CRY2-mediated heat tolerance and the biological significance of the interaction between CRY2 and HsfA1d or IMPα1 deserves further exploration.
CRY1 and HsfA1d respond to light exposure by translocating into the nucleus, allowing them to transmit the light signal into the nucleus for the rewiring of the heat stress response (Figures 6A, S7A, and S7B). We found that IMPα1, which regulated far-red light receptor phyA nuclear translocation, also interacted with CRY1 and HsfA1d (Figures 6B and 6C). Moreover, the co-expressed CRY1 enhanced the interaction between HsfA1d and IMPα1 (Figure 6D). This may result in nuclear transport of HsfA1d by CRY1 when exposed to light. Consistent with this, HsfA1d failed to accumulate in the nucleus when exposed to light in cry1 mutations (Figures 6E and S7C). Notably, cry1 mutations reduced the number of occupied sites and the binding affinity of HsfA1d, and most of the lost sites previously bound by HsfA1d were co-occupied by CRY1 (Figures 7A–7D). Paradoxically, although the binding of HsfA1d to HsfB1 and HsfB2b remained unchanged, the expression of these two genes was decreased. This implies that CRY1 might also play a role in transcriptional activation in addition to affecting the nuclear localization of HsfA1d, possibly because of its transcription activation domain. Despite these complex interactions, overexpression of CRY1 enhanced the survival rate of DK-grown seedlings under high-temperature stress after light pre-treatment, while cry1 mutations reduced that (Figures 2B and 7G). The intricate interplay between CRY1, HsfA1d, and their complex potentially alters promoter binding and gene expression regulation to modulate plant thermotolerance. This study proposes a model in which blue-spectrum light triggers CRY1 to activate HSFs, inducing chaperone genes, and the increased chaperone expression promotes survival at elevated temperatures (Figures 7H and 7I).
The heat-shock response is activated with input from the CRY1 system when high-energy light is present. Similarly, the red/far-red light sensor phyB, which is also a thermosensor in plants,42,43 responds to lower-energy light and primes the HSR.44 Additionally, phyB is responsible for sensing the high proportion of far-red light in autumn to induce cold-responsive genes to prepare for winter frosts.45 Conversely, we propose that CRY1 is responsible for sensing the increasing blue light intensity during spring and summer to increase thermotolerance. On shorter timescales, these systems may operate on diurnal heating and cooling cycles to exert circadian control over proteostasis. From a practical standpoint, the crosstalk between CRY1 and HsfA1d endows plants with stress tolerance and could be a useful tool for modifying crop plants to adapt to high-temperature stress under global climate warming.
Limitations of the study
Our study identifies a molecular pathway underlying the phenomenon of blue LIT in plants. Light exposure enhances the translocation of CRY1 to the nucleus and facilitates nuclear accumulation of HsfA1d through physical interaction, which requires the cooperation of IMPα1, but it is not clear how the CRY1-HsfA1d interaction affects the HsfA1d-IMPα1 module; it is possible that additional factors are involved in CRY1-HsfA1d nuclear import. In the nucleus, CRY1 and HsfA1d co-occupy genomic binding sites and activate transcription of heat-related genes to confer LIT. However, HsfA1d’s binding on some genes, like HsfB1 and HsfB2b, could still be observed in cry1–304 while their transcription level was reduced. This suggests that CRY1 may regulate transcription of HsfB1 and HsfB2b independent of HsfA1d or that other transcription factors are involved in regulating the expression of these genes. Furthermore, we found that CRY1-HsfA1d physically interact with each other, but the mechanism of how light affects the association of CRY1-HsfA1d is lacking. More regulators associated with the CRY1-HsfA1d complex need to be identified to fully elucidate the underlying mechanisms. Finally, in this study we analyzed plant seedlings grown in controlled conditions with DK-to-light transfer, which is not a natural condition of plant growth. Additional regulation models need to be developed for natural day-light cycle growth conditions.
STAR★METHODS
RESOURCE AVAILABILITY
Lead contact
Further information and requests for reagents and source data should be directed to and will be fulfilled by the lead contact, Xu Zheng (zhengxu@123456henau.edu.cn).
Materials availability
All plant materials and plasmids generated in this paper will be shared by the lead contact upon request. This study did not generate new unique reagents.
Data and code availability
RNA-seq and ChIP-seq data generated in this study have been deposited in the Sequence Read Archive (SAR) of the National Center for Biotechnology Information (SAR: https://www.ncbi.nlm.nih.gov/sra/PRJNA842487), and are publicly available as of the date of publication.
This paper does not report original code.
All plant materials and plasmids generated in this paper will be shared by the lead contact upon request.
EXPERIMENTAL MODEL AND SUBJECT DETAILS
For cultivation of Triticun aestiun (Chinese spring), Brassica napus (Zhongshuang 11), and Arabidopsis seeds were surface sterilized for 15 min with 1% sodium hypochlorite. After 3 days germination, seedlings were transferred to soil and kept growing in the dark or 16/8 h light/dark regime (white light, WL) (50 μmol m−2S−1) under 22°C for 4 days, followed by treatment, and recovery at 22°C under WL for 4 days (Figure S1A).
All strains were in Col-0 background. phyA-211,46,47 phyB-9,47,48 cry1–304,49,50 cry1–349,51 and cry2–150,52 were null mutants to the corresponding genes. All light receptor mutant seeds were kindly provided by Dr Chentao Lin. HSFA1a/b/day/e quadruple mutation (hsfa1sqk) and HsfA1a/b/e were kindly provided by Dr Hsiang-chin Liu.1 Transgenic line over expressing GFP-CRY1 and MYC-CRY1 (MYC-CRY1/cry1–304) were kindly provided by Chentao Lin33 and Hongquan Yang,53 respectively.
The full length CDS of HsfA1d (AT1G32330, 1458bp) was amplified with gene specific primers (Table S3), and cDNAs reverse-transcribed from mRNA of Arabidopsis (Col-0) as template. The PCR product was cloned into the site XhoI-XbaI of binary vector pEarlyGate-100. The HsfA1d gene was under the control of the cauliflower mosaic virus 35S promoter. The construct was verified by sequencing and transformed into Agrobacterium tumefaciens strain GV3101 for following transformation of Arabidopsis. Arabidopsis plants were transformed with flower dip method described by Clough and Bent.54
METHOD DETAILS
RT-PCR
RNA was extracted from plants by TransZol regent (Cat No. ET101, TransGen Biotech, Beijing, China). For RT-PCR, the mRNA was converted to cDNA by PrimeScript RT reagent Kit with gDNA Eraser (Perfect Real Time) (Taraka, Code No: RR047Q). Quantitative RT-PCR analyses for expression of target genes were performed according to the manual of the SYBR supermix (BIO-RAD). Three replicates were performed for each sample, with ACTIN2 as the reference gene. The qPCR was performed on QuantStudio 6. The semi-quantitative RT-PCR was performed using the same primers for HSP101, using 18S rRNA as a reference. Primers were listed in Table S3.
Nuclear Fractionation
Nuclear fractionation was performed as described previously.55 Briefly, 1.5g fresh seedlings was homogenized in 3 mL Honda buffer (HB: 2.5% Ficoll 400, 5% dextran T40, 0.4 M Sucrose, 25 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 1 mM DTT, 1 mM PMSF, and 1×complete protease inhibitor mixture) and filtered through a 62 μm pore nylon mesh, then added Triton X-100 (with final concentration of 0.5%) followed by incubating on ice for 15 min and centrifuged (1500g for 5 min). The supernatant was centrifuged (12000 g, 10 min) and saved as nuclei free fraction and the pellet was gently resuspended in 1 mL HB (with 0.1% Triton X-100). Centrifuged the mixture (1500 g, 5 min) and resuspended the pellet in 1 mL HB and centrifuged once more (100 g for 1 min). Then, the supernatant was transferred to a new 1.5 mL eppendorf tube and centrifuged (2000 g, 5 min). The resulting pellet was resuspended in 300 μL buffer G (GB: 1.7 M Sucrose, 10mM Tris-HCl, pH 8.0, 0.15% Triton X-100, 2 mM MgCl2, 5 mM DTT, and 1 ×protease inhibitor mixture) and was added on the top of 300 μL GB in a new 1.5 mL eppendorf tube. Finally, the preparation of the sample was centrifuged (16,000 g, 60 min) and the resulting pellet was resuspended in 100 μL HB with 100 μL SDS loading buffer. The prepared samples were blotted with anti-Flag (Sigma, F1804) and Anti-CRY1 (kindly provided by Dr Chentao Lin) antibody for HsfA1d-Flag and GFP-CRY1 with Histone H3 and Actin (ACT) as loading control. Anti-Histone H3 (Phyto AB, PHY033S) and anti-ACTIN2 (Ambart, 26F7) antibodies were used to detect nuclear and cytosolic markers, respectively.
Yeast Two-Hybrid Analysis
The different combinations of bait plasmid BD-HsfA1d (1–220aa), BD-IMPα1, and prey plasmids AD-CRY1NT (1–490aa), AD-AD-CRY1CT (491–681aa) and AD-CRY1 were co-transformed into yeast strain Y2H. The empty GAL4 BD bait vector (pGBKT7) and GAL4 AD prey vector (pGADT7) combination was used as a negative control. For the construction of bait and prey plasmid, the coding sequences of HsfA1d, CRY1, CRY1 (1–490aa), CRY1 (491–681aa), IMPα1, were amplified from Arabidopsis cDNA. The deletion series of HsfA1d was generated from HsfA1d by inverse PCR. Yeast two-hybrid analysis was performed using the Matchmaker Y2H system (Clontech, Mountain View, CA, USA).
RNA sequencing
Total RNA was extracted using Trizol reagent following manufacturer’s instructions (Invitrogen, CA, USA). Sequencing libraries were generated from the purified mRNA using the VAHTS Universal V6 RNA-seq Library Kit for MGI (Vazyme, Nanjing, China) following the manufacturer’s recommendations with unique index codes. Sequencing was performed on MGI-SEQ 2000 platform by Frasergen Bioinformatics Co., Ltd. (Wuhan, China).
Low quality reads, such as reads with adaptor sequences, reads with >5% N, or >20% bases with quality <Q20 (percentage of sequences with sequencing error rates <1%), were removed using perl script. The clean reads were mapped to the Arabidopsis genome (Tair 10) using HISAT2.56
Differentially expression genes between sample groups was evaluated by DESeq2.57 The false discovery rate (FDR) was used to identify the threshold of P-value in multiple tests in order to compute the significance of the differences. Here, only genes with |log2(−FoldChange)| ≥1 and FDR significance score (padj) < 0.01 were used for subsequent analysis.
Chromatin immunoprecipitation and data analysis
ChIP was performed on 5-day old seedlings grown in the 16h/8h light/dark regime, then were transplanted or not transplanted to 34°C for 30min. Cross-linking was performed with 40 mL of 1% formaldehyde in a vacuum for 30 min followed by quenching with 125 mM glycine for 10 min. After rinsing seedlings with water, nuclei isolation buffer was added to grind the treated sample on ice. We filtered out the slag with a 200-purpose nylon mesh, and finally squeezed the filtrate into 5 mL EP tubes by hand (one glove for each sample) and place on ice for 15 min. After centrifugation at 2000 × g 4°C for 5 min, we discarded the supernatant. Tissues were ground in 500μL ChIP buffer (50 mM HEPES pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton x–100, 0.1% DOC) and sonicated on ice 10 times using a probe sonicator (75W 25% Duty Factor 12000s). 1 mL was transferred to a 1.5 mL tube and spun to remove tissues debris. Input was set aside, and a serial IP was performed. First, 25 μL of anti-FLAG magnetic beads (50% slurry, Sigma) were added the mixture was incubated for overnight at 4°C on a rotator. Beads were separated with a magnet and the supernatant was removed. Beads were washed twice with 1 mL ChIP buffer (5 min incubations at 4°C between each wash), washed again with LiCl buffer (1 mL), and finally wash once with TE buffer (1mL). Beads were separated with a magnet and eluate was transferred to a fresh tube. Bound material was eluted with 100 μL TE +1% SDS by incubating at 65°C for 2h. Beads were separated with a magnet and eluate was transferred to a fresh tube and Add 10 μl 5M NaCl solution to the liquid incubated 3–4 h at 65°C to reverse crosslinks. Protein was degraded by adding 5μL 20 mg/mL proteinase K in 243 μL TE (supplemented 2 μL GlycoBlue) and incubating at 37°C for 2h. DNA fragments were separated from protein by adding 500 μL phenol/chloroform/isoamyl alcohol (25:24:1), and the aqueous layer was added to a fresh tube. 55 μL of 4M LiCl was added along with 1 mL of 100% EtOH, and DNA was precipitated at −20°C overnight. DNA was pelleted by spinning for 30 min at 4°C and resuspended in 50 μL TE.
The PE library with an approximately 300 bp insert size was constructed according to the Illumina library preparation protocol (Il-lumina Inc., San Diego, CA, USA). The library quantification and size were measured using Qubit 3.0 Fluorometer (Life Technologies, Carlsbad, CA, USA) and Bioanalyzer 2100system (Agilent Technologies, CA, USA). Subsequently, sequencing was performed on the Illumina NovaSeq6000 platform.
In ChIP-Seq analysis pipeline. The interference information, containing adapter sequence, low-quality bases and undetected bases (indicated by N) were filtered using Trimmomatic58 (v0.38, with parameters: LEADING:3, TRAILING:3, SLIDINGWINDOW:4:15, MINLEN:8) and quality was assessed using FastQC59 in paired-end raw data. After obtaining clean data through quality control, the Bowtie2 were used to map clean reads to the reference genome60 (v2.3.5, with parameters: dovetail, no-unal, very sensitive local, no mixed, no discordant), and the low-quality mapping was further screen out. PCR redundancy and organelle alignment by Samtools61 (v1.9, with default parameters). The retained valid pairs were used for subsequent analysis.
MACS262 (v2.1.1.20160309) software was used to perform Peak Calling (with parameters: -f BAMPE -B –SPMR –keep-dup all). After that, the R scripts with Deeptools software63 were performed to illustrate the signal enrichment analysis of peak regions, gene body regions and the regions from transcription start site (TSS) with default parameters. ChIPseeker64 was used to obtain gene annotations on peak regions at the same time.
ChIP-qPCR assay
ChIP assays were performed as described above in chromatin immunoprecipitation part. Then, we performed real-time PCR to examine the relative enrichment of certain DNA fragments over the wild type (Col-0) by Roche LC480 with primers listed in Table S3 The 2−ΔΔCT values were calculated to analyze the relative gene expression.65
Co-Immunoprecipitation and western blotting
For in vivo Co-IP assays from plants, the samples were lysed in 600 μL plant IP buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 10 mM NaF, 25 mM b-glycerophosphate, 2 mM sodium orthovanadate, 10% glycerol, 0.1% Tween 20, 1 mM DTT) and 1×Complete Protease Inhibitor Cocktail (Roche). Then 500 μL of extract was incubated with anti-FLAG beads (SIGMA) for 4 h at 4°C. The beads were washed 4 times with 500 μL lysis buffer. The eluate was subjected to western blotting using anti-CRY1 and anti-FLAG antibodies.
The nuclear fraction, or the IP beads were boiled in SDS protein loading buffer at 90°C for 10 min 15 μL of each sample was loaded into SDS-PAGE gels. The gels were run at 30 mA for 1.5 h, and blotted to PVDF membrane at 225 mA for 1.5 h. After 1 h blocking in Li-Cor blocking buffer, the membrane was incubated with anti-FLAG primary antibody (SIGMA) for 1 h, anti-CRY1 (kindly provided as a gift by Chentao Lin). The fluorescent signal scanned with the Li-Cor/Odyssey system.
Luciferase reporter system
Tim-nLUC (nLUC) and Tim-cLUC (cLUC) vector system were used for the split-LUC assay. The coding sequences of CRY1, IMPα1, HsfA1a/b/e were amplified from Arabidopsis cDNA and inserted into the site Kpn I-Sal I of Tim-nLUC or Tim-nLUC. These vectors, including nLUC and cLUC, were introduced into Agrobacterium tumefaciens strain GV3101 separately. The different combinations of Agrobacterium tumefaciens strains with nLUC or cLUC were infiltrated into leaves of N. benthamiana. For the combination of IMPα1-nLUC and HsfA1d-cLUC, a GV3101 strain harboring pCAMBIA2300:CaMV35S::GFP (2300-GFP) or pCAMBIA2300:CaMV35-S::CRY1 (2300-CRY1) was co-infiltrated to elucidate effects of CRY1 to nuclear localization of HsfA1d.
Agroinfiltration of N. benthamiana was performed following the method of Shi et al.66 After 2 days of infiltration, the infiltrated leaves were sprayed with 1 mM D-luciferin (Sigma, 115144–35-9), and kept in the dark for 5 min. The luminescence signal was visualized and recorded using Night SHADE LB 985 system (Berthold, Germany) with a 30 s exposure time, 4 × 4 binning, slow readout, and high gain. Each combination of constructs was repeated for three times with highly similar results.
Bimolecular fluorescence complementation assay
For interaction analysis of CRY1 and HsfA1d, a bimolecular fluorescence complementation (BiFC) experiment was executed in leaf mesophyll protoplast of Arabidopsis. HsfA1d, CRY1 and HY5 genes were amplified from Arabidopsis cDNA and HsfA1d and CRY1 were inserted into the site Xho I-Hind III of PNYFP and PCYFP, respectively, to get PNYFP-CaMV35S::CRY1-nYFP (CRY1-nYFP) and PCYFP-CaMV35S::HsfA1d-cYFP (HsfA1d-cYFP). HY5 was cloned to the site Xba I-BamH I of pUC-CaMV35S:mCherry to get pUC-CaMV35S::HY5-mCherry (HY5-mCherry). HY5-mCherry was co-introduced into Arabidopsis protoplast with CRY1-nYFP and HsfA1d-cYFP as a nuclear maker. Preparation and transfection of Arabidopsis protoplast please refer to Wu. et al.67 Fluorescence signal were captured with Zeiss LSM710 confocal microscope. Wavelengths for excitation and emission were 488 nm and 525 nm for YFP, and 580 nm and 610 nm for mCherry.
QUANTIFICATION AND STATISTICAL ANALYSIS
GraphPad Prism 8 was used for all statistical analyses. Three independent parallel experiments were performed with similar results. Values were Means ± SD. p < 0.05 was considered to be statistically significant. Unless otherwise specified, the student t-test was used to determine the significance of the difference. All the Arabidopsis lines used in this study were Col-0 background, except as otherwise indicated.
