Dynamic regulation of proteostasis underlies thermopriming-mediated acquired thermotolerance in Arabidopsis

Abstract

Rapid climate change necessitates the development of heat-resilient plants. Elevated temperatures disrupt cellular protein homeostasis (proteostasis), and its timely restoration is crucial for plant survival following heat stress. Thermopriming, defined as pre-exposure to sublethal heat stress, has emerged as a promising strategy for enhancing heat tolerance. However, the impact of thermopriming on proteostasis remains unclear. Here, we demonstrate that priming-mediated acquired thermotolerance involves the dynamic regulation of protein maintenance and clearance pathways. Thermopriming facilitates activation of the heat shock response (HSR) via HSFA1 and the unfolded protein response (UPR). Concurrently, priming induces the protein clearance pathway, autophagy, potentially through dynamic modulation of autophagy-negative regulators. In contrast, unprimed seedlings fail to mount effective HSR and UPR responses, resulting in disrupted proteostasis, protein aggregate accumulation, and ultimately loss of viability under heat stress. While loss of UPR has minimal impact on priming-mediated thermotolerance, the HSR is essential, as its absence leads to lethality under heat stress. Notably, loss of HSR further enhances autophagy following stress. Our results highlight the critical role of protein maintenance mechanisms over clearance pathways in ensuring survival. Taken together, these findings demonstrate that thermopriming enhances heat stress resilience by temporally coordinating autophagy, HSR, and UPR to maintain proteostasis.

This study reports that thermopriming (pre-exposure to sublethal high temperature) restores proteostasis by temporally activating the HSFA1-dependent heat shock response, the unfolded protein response, and autophagy, thereby enhancing thermotolerance. In contrast, naive seedlings fail to survive severe heat stress due to impaired proteostasis restoration. These findings highlight the maintenance of protein homeostasis, rather than protein clearance alone, as the key determinant of thermotolerance.

Introduction

In nature, plants frequently encounter multiple abiotic and biotic stressors that disrupt cellular homeostasis. High temperature is a major abiotic stressor that is detrimental to plant growth, productivity, and even survival. Heat stress reduces protein, starch, and oil content (Fahad et al., 2017) and can lead to substantial economic losses in agriculture (Long and Ort, 2010; Schmitt et al., 2022). It has been estimated that every 2 °C rise in temperature could result in a 50% decline in crop productivity (Lobell et al., 2012). As sessile organisms, plants cannot escape adverse environmental conditions and must instead activate protective genes to survive. Recent studies have shown that pre-exposure to a stress stimulus, referred to as “priming,” can dramatically improve plant stress tolerance (Liu et al., 2022; Nair et al., 2022; Samantaray et al., 2023).

Priming refers to preconditioning with a non-lethal stress treatment that induces acquired stress tolerance. This priming treatment enables plants to respond more robustly to a subsequent severe stress, known as “triggering” (Conrath et al., 2007; Bäurle, 2016; Sedaghatmehr et al., 2016; Wilkinson et al., 2019). Various stimuli, including mild environmental stress factors, hormones, or chemical compounds, can act as priming agents (Nair et al., 2022). Priming-mediated acquired stress tolerance has been extensively studied in the model plant Arabidopsis thaliana (hereafter referred to as Arabidopsis), as well as in several crop species (Song et al., 2015; Balmer et al., 2018; Ye et al., 2019; Liu et al., 2022; Samantaray et al., 2023). Priming can remodel transcriptomic, epigenetic, and metabolic programs, thereby enhancing the robustness of plant responses to subsequent stress (Gamir et al., 2014; Espinas et al., 2016; Tugizimana et al., 2018; Serrano et al., 2019; Catoni et al., 2022). Although considerable effort has been devoted to understanding the transcriptional and epigenetic regulation underlying priming responses (Crespo-Salvador et al., 2018; Kotkar and Giri, 2020; Noh et al., 2021; Nair et al., 2022), much less is known about the contribution of post-translational processes—such as protein homeostasis, autophagy, and mitochondrial dynamics—which are critical for plant survival under stress (Sun et al., 2021; Raffeiner et al., 2023; Selinski et al., 2024).

Heat stress primarily disrupts proteostasis by destabilizing thermolabile proteins (Richter et al., 2010). In response, plants activate multiple proteostasis pathways—including molecular chaperones, autophagy, and the ubiquitin–proteasome system (UPS)—to restore proteostasis and promote stress tolerance. Upon heat stress, misfolded proteins accumulate, thereby activating the heat shock response (HSR) and the unfolded protein response (UPR) (Gao et al., 2023). The HSR involves the accumulation of heat shock proteins (HSPs), which facilitate the refolding and disaggregation of misfolded proteins (Richter et al., 2010; Rosenzweig et al., 2019). HSFA2 and HSFA3 are well-established regulators of the HSR that mediate transcriptional upregulation of chaperones during acclimatizing heat stress conditions (Friedrich et al., 2021). The UPR in plants is a complex endoplasmic reticulum (ER)-associated pathway that involves activation of multiple transducers, including members of the inositol-requiring enzyme 1 (IRE1), basic leucine zipper (bZIP), and NAC (NAM, ATAF1/2 and CUC2) families, to resolve proteotoxic stress (Nawkar et al., 2018; Pastor-Cantizano et al., 2020). In addition to these protective responses, proteotoxic stress also activates degradative pathways such as the UPS and autophagy. While the UPS mediates the clearance of unfolded or misfolded proteins, autophagy contributes by removing damaged organelles and promoting the recycling of dysfunctional proteins and protein aggregates during heat stress (Zhou et al., 2013; Su et al., 2020; Raffeiner et al., 2023; Sharma et al., 2024). Autophagy is tightly regulated by signaling pathways such as the TOR, SnRK, and IRE1 pathways. TOR (target of rapamycin) inhibits autophagy under nutrient-rich conditions by phosphorylating components of the autophagy initiation complex (Suttangkakul et al., 2011), whereas SnRK (sucrose non-fermenting 1 [SNF1]-related protein kinase) promotes autophagy during energy deprivation by inhibiting TOR (Soto-Burgos and Bassham, 2017). IRE1, an ER membrane protein, has also been proposed to positively regulate autophagy through its mRNA degradation activity (Bao et al., 2018). Additionally, HSFA1a has been shown to induce autophagy under heat and drought stress by upregulating ATG genes (Wang et al., 2015; Xie et al., 2022). Autophagy also contributes to resetting long-term heat stress memory (Sedaghatmehr et al., 2019) by degrading specific chaperones, including HSP21 and HSP90.1, through FtsH6- and NBR1-mediated mechanisms, respectively (Thirumalaikumar et al., 2020; Sedaghatmehr et al., 2021). However, how proteostasis mechanisms collectively contribute to priming-mediated responses remains unclear.

In this study, we examined the dynamics and regulation of proteostasis pathways to gain insight into the post-translational landscape during priming-mediated acquired thermotolerance. Our findings suggest that thermopriming enhances tolerance to severe heat stress by concurrently activating protein-protective responses through the accumulation of chaperones and folding catalysts via the HSR and UPR, while also facilitating the clearance of damaged proteins through dynamic modulation of autophagy. In contrast, when exposed to heat stress, unprimed seedlings fail to activate these protective responses, experience translational arrest, and accumulate protein aggregates, leading to disruption of proteostasis and ultimately failure to survive. Overall, our study provides novel insights into priming-mediated short-term acquired thermotolerance and highlights the critical role of proteostasis in maintaining cellular resilience under heat stress.

Results

Priming improves survival from severe heat stress

To understand the mechanisms driving short-term acquired thermotolerance, we exposed five-day-old wild-type (WT) Arabidopsis seedlings to 37 °C for 90 min and allowed them to briefly recover under optimal conditions for 90 min before exposing them to triggering heat stress (45 °C for 45 min) (Figure 1A). Hereafter, this condition is referred to as “primed.” The effect of thermopriming was analyzed by comparing the impact of lethal heat stress on primed vs. naive seedlings (unprimed: directly exposed to triggering heat stress). After heat stress, seedling growth was recorded over time under all tested conditions (control, primed, and unprimed) (Figure 1B). Following exposure to triggering heat stress, primed seedlings displayed a temporary halt in growth; however, growth resumed quickly and was significantly greater than that of unprimed seedlings, as quantified by leaf number, root length, and fresh weight (Figure 1C–1E). By five days post-stress (DPS), unprimed seedlings showed a visible and complete loss of chlorophyll and failed to survive (Figure 1B). In summary, heat stress affected growth under both primed and unprimed conditions; however, primed seedlings mounted a more robust response, facilitating rapid recovery and resumption of growth after heat stress.

Figure 1.

Priming improves survival under severe heat stress.

(A) Schematic representation of the experimental paradigm. Seedlings were exposed to 37°C as a priming stimulus, followed by recovery at 22°C (optimal growth temperature), and then subjected to a triggering heat stress at 45°C for the indicated durations. Control (untreated) and unprimed seedlings (exposed only to 45°C) were analyzed in parallel.

(B–E) (B) Following heat stress, seedlings were monitored over several days at 22°C to assess recovery. Representative images of control, primed, and unprimed seedlings are shown. Scale bars, 0.5 cm. Growth recovery was quantified based on (C) the number of leaves per plant (n = 3 biological replicates, 16 seedlings per replicate), (D) root length (n = 3 biological replicates, 6 seedlings per replicate), and (E) fresh weight (n = 3 biological replicates, with 16 seedlings pooled per replicate) at the indicated time points. Despite an initial growth arrest (as reflected in root length), primed seedlings resumed growth comparable to the control, whereas unprimed seedlings exhibited complete growth arrest and ultimately failed to survive.

Data are presented as mean ± SEM (for leaves per plant) and mean ± SD (for root length and fresh weight). Statistical significance was determined by two-way ANOVA. Asterisks indicate significant differences between primed and unprimed seedlings (∗∗P < 0.01; ∗∗∗∗P < 0.0001).

Quantification data are provided in Supplemental Data 2.

Thermopriming mitigates heat stress-induced protein damage

To determine the impact of heat stress (with or without priming) on protein homeostasis, we analyzed protein content in seedlings across all conditions (Figure 2A). The total protein content in control seedlings (grown under optimal conditions at 22 °C) and primed seedlings was similar across all time points analyzed after heat stress (Supplemental Figure 1A). In contrast, unprimed seedlings displayed a consistent decline in total protein levels from 1 to 3 DPS, indicating a reduction in total protein content (Supplemental Figure 1A).

Figure 2.

Priming increases chaperone abundance after heat stress via HSFA1.

(A) Schematic representation of the experimental paradigm. AP, BT, and 0–3 DPS indicate the time points at which seedlings were collected. PS denotes the priming stimulus, and TS denotes the triggering stimulus.

(B) Misfolded protein accumulation following heat stress was assessed by detecting the spliced form of bZIP60 (bZIP60s). Expression analyses indicate that primed seedlings initially accumulated misfolded proteins but recovered by 1 DPS, whereas unprimed seedlings exhibited sustained accumulation of misfolded proteins (n = 3 biological replicates). Statistical significance was determined by one-way ANOVA (∗∗∗∗ p < 0.0001; ∗∗∗ p < 0.001; ns, not significant).

(C) Schematic illustration of the processes involved in maintaining protein homeostasis.

(D–G) To assess the heat shock response (HSR), immunoblot analysis of HSP17.6, HSP21, HSP70, HSP90, and HSP101 proteins was performed in seedlings following heat stress. Priming increased the abundance of HSPs, and this elevation persisted after stress removal. In contrast, unprimed seedlings showed no increase in HSP abundance compared with the control (n = 3 biological replicates). α-Tubulin served as the loading control.

(H) Heat map showing the mean expression levels of HSP genes across different time points based on transcriptome data (n = 2 biological replicates). The data indicate that the priming stimulus increased HSP expression relative to the control. Although unprimed seedlings also exhibited increased HSP expression following stress, the levels remained lower than those in primed seedlings.

(I) Representative images of the hsfa1a/b/d (referred to as a/b/d) as well as WT Col-0 and Ws-0 following priming and triggering heat stress. The priming-induced HSR was absent in the a/b/d mutant compared with Col-0/Ws-0. Unprimed a/b/d seedlings failed to survive, similar to Col-0/Ws-0 after heat stress (n = 3 biological replicates).

(J) Immunoblot analysis of HSPs showed that primed WT seedlings displayed increased HSP abundance after priming, whereas primed a/b/d mutants failed to accumulate HSPs (n = 3 biological replicates).

Quantification data are provided in Supplemental Data 1, 3, and 4.

Heat stress causes protein denaturation, leading to the accumulation of misfolded or unfolded proteins, which, if unresolved, can lead to severe consequences. Accumulation of misfolded proteins triggers the splicing of bZIP60u (unspliced), generating the spliced form bZIP60s, which transcriptionally modulates genes involved in restoring cellular proteostasis under stress conditions (Deng et al., 2011; Humbert et al., 2012). To assess whether proteotoxic stress was induced under our experimental regime, we analyzed the transcript abundance of bZIP60s across all conditions at different time points following priming and triggering heat stress (Figure 2A). Interestingly, exposure to the priming temperature itself triggered the accumulation of misfolded proteins, as evidenced by the presence of bZIP60s at the AP (after priming) time point (Figure 2B and Supplemental Figure 1B and 1C). Following the triggering heat stress, both primed and unprimed seedlings accumulated misfolded proteins (elevated bZIP60s at AT/0 DPS); however, only primed seedlings recovered to control levels by 1 DPS (Figure 2B and Supplemental Figure 1B and 1C). In contrast, unprimed seedlings displayed sustained proteotoxic stress, as evidenced by elevated bZIP60 levels throughout the recovery phase (Figure 2B and Supplemental Figure 1B and 1C). These results suggest that priming-mediated responses may contribute to the efficient maintenance of the cellular protein pool. This observation prompted us to assess the status of major proteostasis maintenance pathways—(1) the HSR, (2) the UPR, and (3) autophagy—during our heat stress paradigm.

Thermopriming facilitates effective HSR mediated by HSFA1

We assessed the status of the HSR (Figure 2C) by analyzing chaperone levels (HSP70 and HSP90) after heat stress across all conditions. Exposure to the priming stimulus (AP) resulted in increased accumulation of HSP70 and HSP90, which persisted until 1 day after exposure to triggering heat stress in primed seedlings compared with the control (Figures 2D and 2E). However, unprimed seedlings failed to mount such a response, with HSP levels remaining lower than those in control seedlings (Figures 2D and 2E). In addition, we analyzed the abundance of other heat stress-inducible proteins, including HSP17.6, HSP21, and HSP101 (Bäurle, 2016; Friedrich et al., 2021), at the time points described above. Increased levels of all three HSPs were detected in primed seedlings but not in unprimed or control seedlings (Figure 2F and 2G). Immediately following exposure to the AP, increased abundance of HSP17.6, HSP21, and HSP101 was observed (Figure 2F), which persisted until 3 DPS (Figure 2G). Furthermore, to assess whether the observed HSP protein abundance paralleled transcript levels across all conditions, we profiled HSP expression using available RNA sequencing data (Supplemental Figure 2A–2C) and performed qPCR validation. In primed seedlings, HSP abundance correlated with elevated transcript levels compared with those in control seedlings (Figure 2H and Supplemental Figure 2D). In contrast, unprimed seedlings displayed a gradual increase in HSP transcript abundance after heat stress (Figure 2H and Supplemental Figure 2D), which did not correlate with the observed protein levels (Figure 2D–2G).

Furthermore, we examined the involvement of HSFA1, a core regulator of the plant HSR (Bakery et al., 2024). We profiled the expression of HSFA1a and its direct and indirect downstream targets, HSFA2 and HSFA3 (Ohama et al., 2017), following heat stress. HSFA1a levels did not change significantly across the tested time points, except at AT (after triggering)/1 DPS in unprimed seedlings (Supplemental Figure 2E). These observations are consistent with previous reports showing that HSFA1a levels remain largely unchanged (Löchli et al., 2023), likely because HSFA1 activity is predominantly regulated at the post-translational level under heat stress (Ohama et al., 2017; Mesihovic et al., 2022; Wang et al., 2023). In contrast, the downstream targets HSFA2 and HSFA3 were upregulated in both primed and unprimed seedlings following heat stress (Supplemental Figure 2E). Notably, HSFA2 induction persisted even at 2 DPS, whereas HSFA3 expression returned to basal levels by 1 DPS in primed seedlings (Supplemental Figure 2E).

HSFA1s are well-established regulators of the HSR that mediate the transcriptional upregulation of chaperones (Ohama et al., 2017). In our study, the downstream activity of HSFA1a (reflected by HSFA2 and HSFA3 expression levels) correlated with the abundance of HSPs across our experimental conditions. These observations prompted us to assess the impact of HSFA1 loss on the priming-mediated HSR. To overcome the functional redundancy of HSFA1s in Arabidopsis, we used a previously reported hsfa1a/b/d triple-mutant line (Yoshida et al., 2011) for our analysis. Under unprimed conditions, hsfa1a/b/d and Col-0/Ws-0 seedlings did not differ in their basal HSR (Figure 2I). However, even after priming, the hsfa1a/b/d mutant failed to survive following heat stress, unlike the Col-0/Ws-0 seedlings (Figure 2I). Furthermore, priming-mediated HSP induction was abolished in the hsfa1a/b/d mutant compared with the WT (Figure 2J). These findings underscore the essential role of HSFA1s in mediating priming-dependent chaperone induction and thermotolerance.

Unprimed seedlings experience translational arrest following heat stress

Notably, despite activation of HSFA1 and the concurrent increase in HSP transcript levels, unprimed seedlings did not show corresponding increases in HSP protein abundance. This prompted us to examine the translational status under these conditions. To this end, we analyzed the expression profiles of differentially expressed genes (DEGs) involved in protein translation using our transcriptome data. Exposure to the priming stimulus alone (the AP time point) had minimal impact on the expression levels of genes encoding translation-associated and ribosomal proteins. However, exposure to the triggering stimulus (the AT time point) led to strong downregulation of these genes, indicating repression of translation (Supplemental Figure 3). Interestingly, during the heat-stress recovery phase, the expression of most genes returned to near-control levels in primed seedlings, whereas unprimed seedlings exhibited sustained downregulation of these genes, suggesting a prolonged halt in translation (Supplemental Figure 3). These findings corroborate our earlier observation that although HSP transcripts accumulate in unprimed seedlings, HSP protein induction is impaired, likely due to impaired translation.

Priming activates the transcriptional UPR

Dysregulation of proteostasis is known to affect ER functional status, thereby triggering activation of the UPR (Humbert et al., 2012; Nawkar et al., 2018). The UPR restores ER proteostasis by upregulating genes involved in the ER quality control (ERQC) and ER-associated degradation (ERAD) pathways (Nawkar et al., 2018; Duan et al., 2023). To assess the status of the UPR, we examined the expression profiles of DEGs involved in the ERQC and ERAD pathways. The priming stimulus alone led to increased expression of several canonical UPR components, including members of the bZIP and NAC transcription factor families, as well as luminal binding proteins (BIPs), protein disulfide isomerases (PDILs), ER-resident DnaJ proteins (ERDJs/P58IPK), and other factors indicative of early ERQC activation (Figure 3A). Additionally, upregulation of various ubiquitin ligase genes implicated in the ERAD pathway was observed following priming (Figure 3A). Importantly, this transcriptional upregulation persisted in primed seedlings even at the post-triggering time point. In contrast, despite exposure to heat stress, unprimed seedlings failed to exhibit any increase in the abundance of these transcripts (Figure 3A). These results indicate that priming enables transcriptional activation of the UPR in response to heat stress.

Figure 3.

The priming stimulus activates transcriptional UPR, whereas loss of UPR delays seedling recovery after heat stress.

(A) Transcriptome data were used to assess activation of the UPR following heat stress. The heat map shows the mean expression levels of genes involved in endoplasmic reticulum (ER) quality control (ERQC) and ER-associated degradation (ERAD), representing two major branches of the UPR pathway. Primed seedlings exhibited enhanced induction of UPR-related genes compared with the control, whereas unprimed seedlings failed to mount a comparable response following heat stress.

(B) Representative images of UPR mutants (bzip60 and ire1a/b) following priming and triggering heat stress. Primed bzip60 and ire1a/b knockout mutants survived severe heat stress, whereas unprimed seedlings failed to survive (n = 3 biological replicates).

(C and D) Primary root length and fresh weight were quantified at 7 DPS across all genotypes. In WT, primed seedlings exhibited root lengths comparable to the control, whereas primed bzip60 and ire1a/b mutants showed a significant reduction in root length compared with their respective controls. Fresh weight was reduced in primed seedlings of all genotypes relative to their respective controls. On the x-axis, C denotes control, and P denotes primed seedlings. Data are presented as median ± SD (n = 3 biological replicates; 15–24 seedlings per genotype represented as individual gray circles). Statistical significance was assessed by one-way ANOVA followed by Tukey’s post hoc test. Compact letter display indicates statistically significant differences among groups (p < 0.05); groups not sharing a common letter are significantly different.

(E) To assess HSR activity in bzip60 and ire1a/b mutants, immunoblot analysis of HSP17.6, HSP21, HSP70, HSP90, and HSP101 was performed under all conditions following heat stress. Priming increased the abundance of HSPs in bzip60 and ire1a/b mutants, and this elevation persisted after stress removal, indicating an active HSR in these mutants. In contrast, unprimed bzip60 and ire1a/b seedlings showed no increase in HSP levels, similar to unprimed WT (n = 3 biological replicates). AP, after priming; AT, after triggering (day 0); DPS, days post-stress.

Quantification data are provided in Supplemental Data 1, 2, and 4.

To further evaluate whether loss of the UPR affects priming-mediated responses, we subjected homozygous bzip60 and ire1a ire1b (hereafter referred to as ire1a/b) mutants to our experimental regime. The IRE1–bZIP60 signaling axis represents one of the major branches of the plant UPR. Upon ER stress, IRE1 is activated, leading to the formation of bZIP60s, which transcriptionally modulate a subset of genes involved in the ERQC/ERAD pathways of the UPR (Nawkar et al., 2018; Pastor-Cantizano et al., 2020). The dynamics of bZIP60 accumulation during the experimental regime are shown in Figure 2B. Similar to WT (Col-0), both bzip60 and ire1a/b seedlings failed to survive when directly exposed to triggering heat stress (unprimed; Figure 3B). Upon priming, the UPR mutants were able to survive severe heat stress (Figure 3B); however, their growth recovery after stress was slower than that of the WT. At seven days post-stress, root length was significantly reduced in primed bzip60 and ire1a/b seedlings compared with their respective controls, whereas the root length of primed WT seedlings was comparable to that of the control (Figure 3C). In addition, primed seedlings of all genotypes exhibited lower fresh weights than their respective controls (Figure 3D).

As the UPR mutants can survive severe heat stress following priming treatment, we further evaluated whether loss of the UPR influences the HSR. To this end, we examined chaperone abundance in the bzip60 and ire1a/b mutants. Primed bzip60 and ire1a/b seedlings displayed elevated levels of all the HSPs tested (Figure 3E). However, slight differences were observed in the dynamics of HSP abundance across genotypes. For instance, HSP17.6 and HSP21 showed similar induction patterns in WT and ire1a/b, with both proteins reaching maximal induction by 1 DPS, whereas the bzip60 mutant displayed maximal HSP21 levels at 2 DPS (Supplemental Figure 4A). Nonetheless, primed WT seedlings showed slightly higher abundance of both HSP17.6 and HSP21 at 0 DPS compared with bzip60 and ire1a/b. Similarly, both mutants displayed differences in the induction patterns of HSP70 and HSP101 compared with WT. The induction of HSP101 was highest in WT at 0 DPS after heat stress and declined from 0 to 3 DPS. In contrast, both mutants showed the highest HSP101 abundance at 1 DPS, with bzip60 exhibiting a steady decline from 1 to 3 DPS (Supplemental Figure 4A). The induction level of HSP70 was minimal compared with the other HSPs in both the mutants and WT at all tested time points (Supplemental Figure 4A). Similar to unprimed WT, unprimed bzip60 and ire1a/b seedlings displayed no detectable HSP levels (Figure 3E). To determine whether the changes in protein levels correspond to changes at the transcript level, we analyzed the expression of HSP genes in both mutants. As expected, all tested HSPs showed elevated transcript levels in primed bzip60 and ire1a/b seedlings (Supplemental Figure 4B). Unprimed bzip60 and ire1a/b seedlings also displayed a delayed increase in HSP transcript levels (Supplemental Figure 4B), similar to WT (Supplemental Figure 2D). Together, these results suggest that unlike the HSR, UPR signaling is not essential for seedling survival following priming; however, it may be required for efficient growth recovery after heat stress.

Primed seedlings exhibit temporally regulated autophagy

After assessing the dynamics of protein maintenance pathways, we next examined the status of the protein clearance mechanism in response to priming and triggering heat stress. Autophagy, as a major clearance pathway, is regulated at both transcriptional and post-translational levels. At the transcript level, autophagy activation involves the upregulation of autophagy-related genes (ATGs) (Zhou et al., 2014; Wang et al., 2020), whereas at the protein level, autophagy induction is marked by the lipidation of ATG8 to ATG8–PE, a critical step for autophagosome formation (Huang et al., 2022). Upon exposure to the priming stimulus, several ATGs were upregulated in primed seedlings; however, this was not observed after triggering heat stress in either primed or unprimed conditions (Figure 4A).

Figure 4.

Primed seedlings display dynamic autophagy, whereas unprimed seedlings exhibit sustained autophagy.

(A) Heatmap showing the mean expression levels of differentially expressed autophagy-related genes (ATGs) across time points based on transcriptome analysis. The data indicate that the priming stimulus increases ATG expression compared with the control.

(B, C) Immunoblot analysis of ATG8 revealed that primed seedlings exhibited a slight induction of autophagy (as indicated by ATG8–PE levels) compared with the control at the AP time point, which further increased after the triggering stress (0–4 h). However, by 1 DPS, autophagy levels in primed seedlings were comparable to those of the control. In contrast, unprimed seedlings displayed a surge in ATG8–PE levels immediately after stress (0 h), which remained consistently higher than the control thereafter. Numbers below the blots indicate normalized intensity values relative to the corresponding control time point (n = 3 biological replicates).

(D) Immunoblot detection of GFP-ATG8 and free GFP (both detected using an anti-GFP antibody) at post-heat stress time points showed that autophagic flux in unprimed seedlings remained elevated during the stress recovery period, as indicated by high levels of free GFP. In contrast, autophagy in primed seedlings was comparable to that in the control. Numbers below represent the normalized mean intensity ratio of free GFP to total GFP (GFP-ATG8 + free GFP) relative to the corresponding control time point (n = 2 biological replicates).

(E) Expression of the autophagy inhibitors BGLU21 and PR-14 shown as fold change relative to the control. In primed seedlings, the expression of autophagy inhibitors was downregulated at AP, BT, and AT but returned to control levels by 1 DPS. In contrast, in unprimed seedlings, expression of autophagy inhibitors remained suppressed during the post-stress period. Data are presented as mean ± SD (n = 2–3 biological replicates). Statistical significance was determined by one-way ANOVA (∗∗∗∗ p < 0.0001, ∗∗∗ p < 0.001, ∗∗ p < 0.01; ns, not significant).

(F) Data-driven schematic illustrating the relationship between downregulation of the autophagy inhibitor BGLU21 and induction of ATG8 levels in primed and unprimed seedlings. AP, after priming; BT, before triggering; AT, after triggering (day 0); DPS, days post-stress.

Quantification data are provided in Supplemental Data 1 and 3.

We next quantified the protein-level autophagy response using ATG8 as a marker. Exposure to the priming temperature alone induced ATG8 levels (Figure 4B), consistent with the transcriptome data. However, immediately after exposure to triggering heat stress (0 h), autophagy (ATG8–PE) levels peaked in both primed and unprimed seedlings compared with the control (Figure 4B). To assess whether heat stress-induced autophagy is sustained, declines, or increases over time, we examined ATG8–PE levels from 1 to 3 DPS. In unprimed seedlings, ATG8-PE levels increased further by 1 DPS and remained elevated until 2 DPS (Figure 4C). Notably, the visible decline in ATG8 levels at 3 DPS correlated with reduced protein abundance in unprimed seedlings (Supplemental Figure 1A). In contrast, in primed seedlings, ATG8–PE levels returned to control levels by 1 DPS (Figure 4C). We also measured autophagy flux across our experimental paradigm using the GFP–ATG8 processing assay (Shin et al., 2014; Kang et al., 2018). Autophagy flux, as measured by the free GFP to total GFP ratio, was higher in unprimed seedlings, indicating more active autophagy compared with primed seedlings and controls (Figure 4D). The contrasting increase and decrease in autophagy levels in unprimed and primed seedlings, respectively, suggest dynamic regulation of autophagy under these conditions.

Regulation of autophagy under heat stress is independent of TOR and SnRK signaling

Previous studies have identified roles for TOR and SnRK signaling in regulating autophagy under stress (Pu et al., 2017; Soto-Burgos and Bassham, 2017). We therefore asked whether TOR/SnRK signaling is involved in the priming-mediated regulation of autophagy under heat stress. To test this, we analyzed phospho-S6K (pS6K) levels as a measure of TOR kinase activity (Upadhyaya et al., 2020) and active SnRK (pAMPK, detected using a phospho-AMPKα antibody) as an indicator of SnRK activity (Baena-González et al., 2007; Belda-Palazón et al., 2020). Interestingly, neither pS6K (Supplemental Figure 5A and 5B) nor pAMPK levels (Supplemental Figure 5C and 5D) were altered across the tested conditions, suggesting that heat stress-induced autophagy (as observed in Figure 4B and 4C) occurs independently of TOR- and SnRK-mediated regulation.

Induction of priming-mediated autophagy is associated with the attenuation of autophagy inhibitors

The UPR regulator IRE1 has also been linked to the modulation of autophagy by fine-tuning its RIDD (regulated IRE1-dependent decay of mRNA) activity (Maurel et al., 2014; Li and Howell, 2021). Notably, RIDD downregulates inhibitors of autophagy, such as BGLU21, ROSY1, and PR-14, by targeting their mRNAs for degradation (Bao et al., 2018). We surmised that the return to basal autophagy levels in primed seedlings, compared with the more sustained autophagy observed in unprimed seedlings, may involve RIDD-mediated regulation. Therefore, we examined the steady-state transcript levels of BGLU21 and PR-14 across all conditions. At the post-priming (AP), before-triggering (BT), and after-triggering (AT: 0 DPS) time points, primed seedlings exhibited a significant reduction in transcript levels of both genes compared with the control (Figure 4E). However, by 1–2 DPS, transcript abundance recovered (Figure 4E). In contrast, the transcript levels of BGLU21 and PR-14 remained significantly lower in unprimed seedlings at all time points compared with the control (Figure 4E). Importantly, these changes in transcript abundance correlated with ATG8/ATG8–PE levels in these seedlings. For instance, in primed seedlings, reduced transcript abundance of these autophagy inhibitors corresponded with increased autophagy at AP, BT, and AT. By 1–2 DPS, when transcript levels of these inhibitors returned to near-control levels, autophagy also reverted to basal levels in primed seedlings (Figure 4F). Conversely, sustained induction of autophagy in unprimed seedlings aligned with persistently reduced transcript abundance of the autophagy inhibitors following exposure to triggering heat stress (Figure 4F). Collectively, these observations suggest that priming modulates autophagy, potentially through RIDD-mediated attenuation of autophagy inhibitors.

Further, we assessed whether autophagy dynamics are affected by IRE1A/B and its downstream target bZIP60. Interestingly, autophagy remained induced in both bzip60 and ire1a/b mutants (Supplemental Figure 6A and 6B), and its dynamics were similar to those observed in WT (Figure 4B and 4C). Loss of IRE1A/B or bZIP60 did not appear to disrupt autophagy induction during heat stress. This was unexpected, as our results indicate that heat stress facilitated IRE1-dependent dynamic degradation of autophagy inhibitors, thereby modulating autophagy. However, autophagy was still induced in bzip60 and ire1a/b mutants. These findings suggest the existence of additional IRE1-independent mechanisms underlying autophagy regulation under heat stress.

Impairment of protein protective responses enhances protein clearance mechanisms

Given the potential involvement of alternative mechanisms regulating autophagy, we next investigated the role of HSFA1a, a transcription factor previously shown to induce autophagy under abiotic stress conditions (Wang et al., 2015; Xie et al., 2022). Notably, we observed induction of HSFA1 activity together with parallel upregulation of ATG genes following the priming stimulus (Figure 4A; Supplemental Figure 2E). We therefore assessed autophagy responses in the hsfa1a/b/d triple mutant and found that autophagy remained active in both primed and unprimed seedlings, as evidenced by sustained accumulation of ATG8–PE after stress (Figure 5A). These results suggest that autophagy may be regulated by a broader, multilayered regulatory system.

Figure 5.

Impairment of the heat shock response enhances autophagy during proteotoxic stress.

(A) Immunoblot analysis of ATG8 in the hsfa1a/b/d mutant shows that autophagy remains highly induced in both primed and unprimed seedlings following heat stress (n = 3 biological replicates). Numbers below the blots indicate normalized intensity values relative to the corresponding control time point.

(B) Data-driven schematic line plots comparing ATG8 and HSP profiles in primed WT, primed hsfa1a/b/d, and unprimed seedlings at the indicated time points. Bar plots above show the percentage of seedling survival at 7 DPS (n = 3 biological replicates per condition). In primed WT, chaperone (HSP) accumulation peaks early and is accompanied by transient ATG8 induction, suggesting efficient proteostasis and resolution of stress through timely attenuation of autophagy. In contrast, both unprimed WT and the primed hsfa1a/b/d mutant fail to induce HSPs, correlating with sustained ATG8 accumulation and complete post-stress mortality. Collectively, these data illustrate that thermotolerance is driven by coordinated HSP–autophagy interplay, whereas disruption of this balance leads to autophagy overactivation and seedling lethality. DPS, days post-stress.

Given that HSP induction was abolished in the hsfa1a/b/d mutant with or without priming (Figure 2J), we integrated all protein-level evidence and summarized it in a data-directed illustration (Figure 5B). Collectively, as evidenced in primed WT seedlings, activation of the protein protective response (chaperone accumulation) facilitates a dynamic protein clearance response (autophagy) that subsides during stress recovery (Figure 5B). However, as observed in unprimed seedlings (across all genotypes) and in primed hsfa1a/b/d seedlings, failure to accumulate chaperones upon heat stress may promote persistent activation of autophagy. Although this response likely aims to resolve stress, it ultimately leads to lethality (Figure 5B).

Unprimed seedlings accumulate protein aggregates following triggering heat stress

So far, our findings indicate that priming-mediated heat stress tolerance involves the accumulation of chaperones and temporary activation of autophagy. To test whether these changes contribute to the effective maintenance of protein homeostasis, we examined the proteostasis state by analyzing protein aggregates at 1 DPS, the time point at which we observed differential autophagy between the primed and unprimed conditions (Figure 4C). Unprimed seedlings displayed elevated levels of ubiquitin-conjugated aggregates compared with primed seedlings and controls (Figure 6A). Additionally, size distribution analysis of protein aggregates using dynamic light scattering (DLS) showed that unprimed seedlings accumulated a higher proportion of larger aggregates (13.03, 460.1, and 6662.6 nm), whereas primed seedlings accumulated comparatively smaller aggregates (18.02, 174.1, and 1977.0 nm), with distributions similar to those observed in the control (24.9, 136.5, and 1550.5 nm) (Figure 6B). We also compared the diffusion behavior of the aggregates using the DLS data. BSA was used as a standard monomeric protein control (8 nm) and exhibited the fastest (steepest) decay (Figure 6C). Aggregates derived from unprimed seedlings showed the slowest decay, reflecting larger population sizes, whereas those derived from primed seedlings showed comparatively faster decay, reflecting a relatively smaller population (Figure 6C). The control samples displayed a profile comparable to that of BSA. Taken together, these results indicate that thermopriming facilitates the timely activation of the HSR, UPR, and autophagy, thereby promoting effective aggregate clearance and enhancing heat stress tolerance.

Figure 6.

Priming facilitates efficient aggregate clearance following heat stress.

(A) Immunoblot analysis of ubiquitin-conjugated protein aggregates shows that unprimed seedlings accumulate higher levels of aggregates compared with primed and control seedlings by 1 DPS. The box plot represents the normalized mean intensity of ubiquitin in the insoluble fraction. The dotted line denotes the control value (n = 2 biological replicates).

(B) Dynamic light scattering (DLS) analysis showing the particle size distribution of protein aggregates isolated at 1 DPS from control, primed, and unprimed seedlings. Bovine serum albumin (BSA) was used as a monomeric protein standard. Unprimed seedlings accumulated a higher proportion of large aggregates (460.1 and 6662.6 nm), indicating extensive aggregation. In contrast, primed seedlings displayed smaller aggregate populations (174.1 and 1977.0 nm), comparable to the control, suggesting more efficient aggregate clearance.

(C) Autocorrelation function (g²) curves derived from DLS measurements reflecting particle size dynamics and diffusion behaviors. BSA exhibited the steepest (fastest) decay, consistent with uniform small particles. Aggregates derived from unprimed seedlings exhibited the slowest decay, indicative of larger and less diffusive aggregate populations. Primed samples exhibited faster decay than unprimed samples, indicating smaller and more dynamic aggregates. Data are representative of two biological replicates.

Discussion

In the current study, we demonstrate that efficient regulation of proteostasis underlies priming-mediated short-term acquired thermotolerance in Arabidopsis. Heat stress has a profound impact on thermolabile molecular complexes, resulting in protein dysfunction and aggregation (Mittler et al., 2012; McLoughlin et al., 2019; Rosenzweig et al., 2019). Maintaining proteostasis under stress requires a fine balance between the timely refolding of proteins through robust chaperone activity and the efficient clearance of irreversibly damaged proteins via the action of disaggregases, autophagy, and/or the UPS (Hartl et al., 2011; Mogk et al., 2018; McLoughlin et al., 2019). Disruption of this balance can lead to proteotoxic stress, ultimately compromising cellular viability. In this study, we observed that direct exposure to heat stress severely reduced protein content in unprimed seedlings, whereas this reduction was alleviated in primed seedlings. These observations are consistent with previous reports showing decreased protein content upon exposure to noxious heat stress (Gulen and Eris, 2004; Nagy-Réder et al., 2022). We found that thermopriming induces chaperone accumulation and activates the UPR, while also inducing a transient autophagy response, collectively contributing to the maintenance of proteostasis.

Unprimed seedlings, despite experiencing the same triggering heat stress, failed to accumulate HSPs even though HSP transcripts were significantly induced. Such failure to accumulate HSPs may result from sequestration of HSPs within protein aggregates or from impaired translation following stress. We observed that genes involved in translation remained downregulated in unprimed seedlings compared with primed seedlings after heat stress. Additionally, HSP101, a critical disaggregase that coordinates with other HSPs to resolve proteotoxic stress (McLoughlin et al., 2019), was detected only in primed seedlings (Supplemental Figure 3B). Collectively, these findings suggest impaired chaperone synthesis in unprimed seedlings. Moreover, unprimed seedlings displayed elevated levels of ubiquitinylated insoluble proteins, a phenotype also reported in HSP101-deficient plants under heat stress (McLoughlin et al., 2019). Similarly, Dannfald et al. (2025) showed that priming preserves global translational capacity under heat stress, whereas direct heat exposure leads to preferential mRNA decay.

Given the differences in autophagy responses between primed and unprimed seedlings, we examined the role of TOR/SnRK-dependent pathways, which are known to regulate autophagy (Soto-Burgos and Bassham, 2017; Signorelli et al., 2019; Yang et al., 2023) but detected no changes in TOR/SnRK activity under our experimental conditions. Indeed, TOR-independent mechanisms have been reported to modulate autophagy under oxidative and ER stress (Pu et al., 2017). Because thermopriming induced ER stress and subsequent UPR activation, we explored the potential involvement of IRE1. IRE1-mediated RIDD activity has previously been associated with autophagy induction under tunicamycin treatment (an ER stress–inducing agent), where RIDD-mediated degradation of autophagy inhibitors promotes autophagy (Bao et al., 2018). Our findings show that thermopriming facilitates the degradation of RIDD target transcripts, which correlates with autophagy induction in primed seedlings. Similarly, in unprimed seedlings, prolonged autophagy induction is associated with sustained transcriptional downregulation of autophagy inhibitors. Although these observations suggest that RIDD activity influences autophagy by modulating the abundance of autophagy inhibitors, the precise mechanism remains to be elucidated. Bao et al. (2018) suggested that BGLU21 (β-glucosidase 21) and PR-14 (pathogenesis-related protein 14) may interfere with the lipidation of ATG8—a crucial step in autophagosome formation—thereby affecting autophagy.

Notably, despite the loss of IRE1A/B or bZIP60, autophagy was still induced following heat stress. Our findings are consistent with previous reports showing that bZIP60 is dispensable for IRE1-dependent autophagy induction during ER stress (Bao et al., 2018). Moreover, although autophagy is induced in ire1b mutants following 42 °C heat treatment (Yang et al., 2016), it is not induced in ire1b or ire1a/b mutants under ER stress caused by DTT or tunicamycin treatment (Liu et al., 2012; Bao et al., 2018). Given that autophagy remains activated in ire1a/b mutants following thermopriming, we propose that alternative pathways may mediate heat-induced autophagy. We examined the involvement of HSFA1, a heat-inducible transcriptional activator of autophagy (Wang et al., 2015; Xie et al., 2022; Löchli et al., 2023), and found that autophagy was still induced in the hsfa1a/b/d mutant. Together, these observations suggest that the autophagy response is governed by multilayered regulation, a feature likely advantageous for sessile organisms in coping with environmental challenges.

Previous studies have also implicated NBR1 (next-to-BRCA1) and ATG8-interacting proteins 1–3 (ATI1, ATI2, and ATI3) in mediating selective autophagy (Zhou et al., 2018; Sjøgaard et al., 2019; Jung et al., 2020; Thirumalaikumar et al., 2020). Autophagy can occur either as selective autophagy, which involves recognition by specific receptors followed by clearance of damaged organelles or protein aggregates, or as bulk autophagy, which facilitates non-selective removal of cytoplasmic debris (Marshall and Vierstra, 2018). From our transcriptome data, we observed that selective autophagy receptors such as NBR1 and ATI1/2/3A were upregulated only in primed seedlings (Supplemental Figure 7), suggesting that the autophagy response in these seedlings may predominantly be selective in nature. In contrast, bulk autophagy may operate in unprimed seedlings, as we observed no changes in the expression of these receptors. Moreover, vacuolar cell death has previously been associated with high autophagic flux (Minina et al., 2013), a feature we observed in unprimed seedlings. Based on the present evidence, it is tempting to speculate that bulk autophagy may drive unprimed seedlings toward vacuolar cell death; however, this hypothesis requires further experimental validation.

Additionally, analysis of bzip60 and ire1a/b knockout mutants under primed conditions revealed reduced growth recovery after stress compared with WT. Previous studies have shown that bzip60 and ire1a/b mutants exhibit severe growth retardation under chemically induced ER stress (Deng et al., 2013; Pu et al., 2019). Although primed bzip60 and ire1a/b seedlings displayed increased HSP levels, their response was delayed compared with WT. One possible explanation is attenuated or delayed protein translation, a feature previously reported in ire1a/b mutants under ER stress (Yoo et al., 2024). Notably, Yoo et al. (2024) also demonstrated that supplementation with synthetic chaperones mitigates the effects of attenuated translation and growth retardation in ire1a/b mutants. Given that the ire1a/b mutant accumulates chaperones under the conditions tested, priming may facilitate faster and more robust recovery of these seedlings under stress. Taken together, the altered dynamics of HSP induction in UPR mutants suggest potential crosstalk between the HSR and UPR pathways during thermotolerance. In addition, the hsfa1a/b/d mutant failed to survive heat stress even after priming. This failure likely results from the inability of the hsfa1a/b/d mutant to produce HSPs, indicating that efficient protein maintenance mediated by chaperones is a critical determinant of stress response outcomes. An earlier report suggested that autophagy contributes to the degradation of HSPs, as autophagy mutants exhibit prolonged chaperone accumulation (Sedaghatmehr et al., 2019). However, our findings suggest that autophagy and the chaperone response may act cooperatively to modulate protein homeostasis, with impairment of one pathway potentially enhancing the activity of the other. For example, in hsfa1a/b/d, failure to induce HSPs is accompanied by sustained autophagy. These observations underscore the importance of maintaining a delicate balance between protein folding and clearance mechanisms in restoring cellular proteostasis following stress. In contrast, unprimed seedlings experience a collapse of this balance, as evidenced by failure to accumulate HSPs, lack of UPR activation, excessive aggregate accumulation, and sustained autophagy, ultimately compromising growth and survival (Figure 7).

Figure 7.

Priming-mediated acquired thermotolerance relies on the dynamic regulation of proteostasis.

The illustration compares the responses of primed (top) and unprimed (bottom) Arabidopsis seedlings following exposure to severe heat stress. In primed seedlings, HSFA1-mediated transcriptional activation drives a robust HSR, leading to the accumulation of HSPs. Accumulation of misfolded proteins in the ER activates the UPR, resulting in the induction of transcripts involved in ERQC and ERAD pathways, thereby alleviating ER stress. As a result of active HSR and UPR, protein aggregate formation remains minimal. Transient autophagy, potentially regulated by IRE1-dependent and -independent pathways, efficiently clears residual aggregates, thereby maintaining proteostasis and promoting seedling survival. In contrast, unprimed seedlings undergo translational repression and fail to efficiently activate HSR and UPR, resulting in prolonged ER stress and excessive protein aggregation. In response, autophagy remains persistently activated but eventually becomes maladaptive, coinciding with proteostasis collapse and ultimately leading to seedling death. Some graphical elements were created using BioRender.com.

Taken together, our study highlights that thermopriming facilitates efficient maintenance of proteostasis through the timely activation of the HSR, UPR, and autophagy, thereby promoting seedling survival under severe heat stress. Furthermore, we found that UPR activation is important for timely growth recovery after stress but may not be essential for the establishment of thermotolerance. In contrast, HSFA1-mediated HSR is indispensable for mounting a heat stress tolerance response. As we observed that chaperone accumulation is affected in UPR mutants, further studies are required to clarify the potential crosstalk between the UPR and HSR pathways during thermotolerance. Additionally, the mechanisms by which heat stress influences the induction of bulk vs. selective autophagy during acquired thermotolerance remain to be elucidated. Interestingly, recent studies have shown that HSRs can be heterogeneous and occur in a cell-type-specific manner (Jin et al., 2025; Wang et al., 2025). Single-cell-level analyses under primed conditions may therefore help resolve the spatial dynamics of proteostasis and priming-associated stress responses. Collectively, our findings reveal that priming-mediated acquired thermotolerance relies on the precise and coordinated regulation of protein folding and clearance pathways.

Methods

Plant materials and growth conditions

Arabidopsis (Col-0, referred to as WT, and Ws-0) seeds were surface-sterilized by washing with 80% ethanol for 30 seconds (s), followed by 1% bleach for 3 minutes (min), and then washed three times with sterile double-distilled water. Seeds were sown on ½ Murashige and Skoog medium supplemented with 1% sucrose and 0.8% agar. The seeds were stratified at 4°C for two days before being transferred to growth chambers under a 16-hour (h) light/8 h dark cycle at 65% humidity. Five-day-old Arabidopsis seedlings were primed by exposure to 37 °C for 90 min, followed by recovery at 22 °C for 90 min, and then subjected to triggering heat stress at 45 °C for 45 min. Unprimed seedlings were directly exposed to 45 °C for 45 min. Control seedlings were maintained under optimal growth conditions (22 °C, 16 h light/8 h dark cycle, 65% humidity) throughout the experiment. Seedlings were harvested at different stress-recovery time points for gene expression and protein analyses. The treatment conditions described above were also used for the analysis of GFP-ATG8a (CS39996) and bzip60 (CS813074), both obtained from the Arabidopsis Biological Resource Center (ABRC). The ire1a/b line was kindly provided by Chen et al. (2013) and Pu et al. (2019). The homozygous hsfa1a/b/d (psi00284) line was obtained from RIKEN-BRC.

Physiological measurements

To determine the effects of priming or triggering heat stress on growth, seedlings were analyzed for leaf number, root length, and fresh weight. Data on leaf number (48 seedlings from three independent experiments), primary root length (18 seedlings from three independent experiments), and fresh weight (three biological replicates) were collected and analyzed. For Figure 1E, 16 seedlings were pooled to form a single replicate, whereas for Figure 3D,fresh weight was measured for individual seedlings. For leaf number measurements, both true leaves and cotyledons were included. Root length was measured using ImageJ software (https://imagej.net/software/fiji/) (Schneider et al., 2012).

RNA extraction and gene expression analysis

For RNA extraction, approximately 40–50 mg of Arabidopsis seedlings were collected at several time points (post-priming, triggering, and during recovery) and frozen in liquid nitrogen. Samples were ground using a Retsch Mixer Mill MM400 at 25 Hz for 30 s (two cycles). RNA was then isolated using a combination of TRIzol/chloroform extraction and a column-based method (PureLink RNA Mini Kit, Thermo; 12183018A). The purified RNA was treated with DNase I (NEB; M0303) following the standard protocol. cDNA was synthesized from 1 μg of RNA according to the manufacturer’s instructions (1st Strand cDNA Synthesis Kit, Takara; 6110A). For gene expression analysis, reverse-transcription PCR (RT–PCR) was performed using SapphireAmp Fast PCR Master Mix (Takara; RR350A), whereas iTaq Universal SYBR Green Supermix (Bio-Rad; 1725121) was used for quantitative real-time PCR (qRT–PCR). The housekeeping gene ACT2 was used as an internal control to normalize target gene expression across samples. All primers used, together with their corresponding gene accession numbers, are listed in Supplemental Table 1.

Transcriptome analysis was performed with two biological replicates under the described conditions (Supplemental Figure 2A). Paired-end sequencing libraries were prepared using the SMART-seq Library Prep Kit (R40047) according to the manufacturer’s protocol, and sequencing was performed on the Illumina NovaSeq 6000 S4 platform. Adapter trimming and quality control were conducted using fastp (Chen et al., 2018) and FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), respectively.Kallisto (v0.50.1; Bray et al., 2016) was used to index the reference Arabidopsis transcriptome downloaded from Ensembl. Read mapping was performed using Kallisto, after which TxImport (Soneson et al., 2016) was used to import the Kallisto outputs into the R environment. Differential gene expression analysis between control, primed, and unprimed seedlings was performed using the edgeR package (Chen et al., 2025) in RStudio. Genes with an adjusted p value < 0.01 and a cutoff of log₂FC ≥ 1 were considered upregulated, whereas log₂FC ≤ −1 was used to define downregulated genes. The expression profiles of the DEGs used for plotting are provided in Supplemental Data 1.

Protein extraction and western blot analysis

Soluble proteins were extracted from Arabidopsis seedlings as previously described by Yamamoto et al. (2018) with minor modifications. Approximately 40–50 mg of seedlings were collected and frozen in liquid nitrogen. Samples were ground using a Retsch Mixer Mill MM400 at 25 Hz for 30 s (two cycles). After grinding, 200–300 μl of 1× RIPA lysis buffer (Merck; 20-188) supplemented with 10% glycerol, 1× phosphatase inhibitor cocktail (PhosSTOP, Roche; 4906837001), and 1× protease inhibitor cocktail (Roche; 4693132001) was added, and the samples were incubated at 4 °C for 1 h with shaking at 1500 rpm. The samples were then centrifuged at 14,000 rpm for 30 min at 4°C, after which the supernatant containing soluble proteins was collected and stored at −80 °C until further use. Plant protein aggregates were extracted using a standard protocol as described by Phukan et al. (2023).

For protein analysis, protein concentration was estimated using the Bicinchoninic acid (BCA) method according to the manufacturer’s instructions (cat. no. 23225). Proteins (20 μg per lane) were loaded onto an 8% SDS–PAGE gel, which was subsequently stained with Coomassie Brilliant Blue. Protein bands were visualized after destaining the gel following a standard protocol (Sreeramulu and Singh, 2007). For Western blot analysis, 20 μg of protein was loaded onto 6%, 8%, or 12% SDS–PAGE gels, after which the proteins were transferred onto a nitrocellulose membrane. The membrane was blocked with 5% BSA or 2% skim milk for 1 h at room temperature (RT), followed by incubation with the primary antibody overnight at 4°C. The blot was developed after incubation with the secondary antibody (1 h at RT) using Immobilon Forte HRP (horseradish peroxidase) substrate (cat. no. WBLUF0500). The following primary antibodies were used: α-Tub (CST 3873), ATG8 (Abcam, ab77003; Agrisera, AS142769), GFP (CST 2956), HSP70 (Abcam, ab2787), HSP90 (Sigma, CA1023; Agrisera, AS08346), HSP17.6 (Agrisera, AS07254), HSP21 (Agrisera, AS08285), HSP101 (Agrisera, AS08287), pS6K (CST 9234), S6K (CST 9202), pAMPK-α (CST 2535), ubiquitin (CST 3933), and actin (Agrisera, AS132640). Anti-rabbit (sc-2313; CST 7074) and anti-mouse (CST 7076) antibodies were used as secondary antibodies. Blots were quantified densitometrically using ImageJ.

Dynamic light scattering (DLS) measurements

To assess particle size distribution profiles, 20 μg of aggregate sample or standard protein (BSA) was appropriately diluted in ultrapure water and analyzed using a Litesizer 500 system equipped with a BM10 module (Anton Paar Instruments). Measurements were performed at 25°C using disposable cuvettes with a side-scatter detection angle in automatic mode, with a maximum of 60 runs per sample and 10 s per run.

Statistical analysis

Data were plotted using GraphPad Prism 10.0 (https://www.graphpad.com/). Unless otherwise stated, experiments were conducted with three biological replicates. Statistical significance was determined using one-way ANOVA (for qPCR and phenotypic analyses) or two-way ANOVA (for phenotypic analyses), as appropriate. p values < 0.05 were considered statistically significant.

Data and code availability

The transcriptome data generated in this study have been deposited in the Sequence Read Archive (SRA) of the National Center for Biotechnology Information (NCBI) under BioProject accession number PRJNA1312973.

Funding

This work was supported by the Department of Science and Technology (Sir J.C. Bose Fellowship to B.J.R.), the Science and Engineering Research Board (SPG/2021/004000) (A.D.A.), intramural funding from IISER Tirupati (A.D.A. and B.J.R.), and the University of Hyderabad (B.J.R.).

Acknowledgments

We sincerely thank Prof. Diane Bassham (Iowa State University) and Prof. Federica Brandizzi (Michigan State University) for providing the ire1a ire1b mutant seeds. M.B. acknowledges the Council of Scientific and Industrial Research (CSIR), Government of India, for fellowship support. The authors declare no competing interests.

Author contributions

Design of the experiments, conduct of the experiments, interpretation of the results, preparation of the figures, and writing of the manuscript, M.B.; conceptualization of the research question, supervision of the work, design of the experiments, interpretation of the results, writing and editing of the manuscript, and funding acquisition, A.D.A.; conceptualization of the research question, supervision of the work, design of the experiments, interpretation of the results, writing and editing of the manuscript, and acquisition of funding, B.J.R.

Published: February 26, 2026