Arabidopsis thaliana seeds were stratified at 4°C in the dark for 3 days on Murashige and Skoog (MS) agar medium supplemented with 3% sucrose and then transferred to a growth chamber at 19–20°C with ∼80 μmol m^–2 s^–1 of cool white fluorescent light or in compost soil with ∼120 μmol photons m^–2 s^–1 under long day conditions (LgD, 16 h light: 8 h dark).

To generate transgenic plant lines overexpressing the ZTL gene, the open reading frame (ORF) for the ZTL (At5g57360) was first ligated into pDONR221, and then cloned into pEarleyGate101 by the Gateway method (Earley et al., 2006). Primers used in vector construction are listed in Supplementary Table 1. The construct was verified by sequencing and introduced into the GV3101 strain of Agrobacterium tumefaciens. The construction was transformed by floral infiltration into WT (C24) plants. Transgenic plants were selected by basta resistance and confirmed by PCR. All positive lines showed a similar and reported phenotype: long-hypocotyl compared with wild-type (WT, C24) plants (Somers et al., 2004), suggesting these transgenic lines are ZTL-overexpressing plants, and the homozygous T3 seeds of the transgenic plants were used for further analysis. Mutant plants ztl-1 (C24; Somers et al., 2000) and ztl-3 (Col; Jarillo et al., 2001) were described in the corresponding published papers.

To assay germination, approximately 100 seeds were sterilized and planted on MS medium containing 3% sucrose and 0.8% agar (pH5.9), and was supplemented with or without different concentrations of (±) ABA. Germination was defined as an obvious emergence of the radicle through the seed coat. Seedling growth was also assessed by directly planting the seeds in ABA-free MS media and ABA-containing MS medium. For stomatal aperture assays, 3–4-week-old seedlings were used to study ABA-induced stomatal closure, leaves were detached at Zeitgeber Time 2 (ZT2) and floated in the buffer containing 50 mM KCl and 10 mM MES-KOH (pH 6.15) under a halogen cold-light source (Colo-Parmer) at 200 μmol m^–2 s^–1 for 3 h followed by addition of 20 μM (±) ABA. Stomatal apertures were recorded on epidermal strips before ABA treatment and 1 h or 3 h after ABA treatment. For water-loss assays, rosette leaves of comparable size from 3 to 4-week-old plants grown under LgD were detached, placed on a Petri dish and weighed at different times after detachment. For drought tolerance test, approximately 8-day-old seedlings were transplanted to the soil (each pot contained 4 plants) for another 2 weeks under standard growth conditions. Then, the plants were subjected to progressive drought by withholding water for 2 weeks and then rewatered for 2 days. The entire test was repeated five times.

Total RNA was isolated from 14-day-old seedlings with a Total RNA Rapid Extraction Kit (BioTeke, China), and treated with RNase-free DNase I (NEB) at 37°C for 1 h to degrade genomic DNA. Then RNA was purified by a RNA Purification Kit (BioTeke, China). Two micrograms of total RNA was subjected to first-strand cDNA synthesis using a Roche Transcriptor First Strand cDNA Synthesis Kit and an oligo (dT18) primer. qPCR was performed with SYBR Premix Ex Taq (Takara) using a BioRad Real-Time System CFX96TM C1000 Thermal Cycler (BioRad). Amplification of ACTIN2/8 genes was used as an internal control. The primers of CHLH and ACTIN2/8 for qPCR were described previously (Shang et al., 2010). For all quantitative real-time PCR analyses, the assays were repeated three times, and the means of four biological experiments were used to estimate gene expression.

Western-blot assays were essentially performed as previously described (Shen et al., 2006; Wu et al., 2009). Briefly, 2-week-old seedlings were ground in liquid nitrogen and proteins were extracted in extraction buffer [50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 0.1% (v/v) Triton X-100, 10% (v/v) glycerol and protease inhibitor cocktail (Roche)]. Protein concentration was calculated using the Bradford method (Bio-Rad), and 20–60 μg of total protein was loaded per lane. Proteins were transferred to nitrocellulose membranes (Amersham) and stained with red ponceae following standard protocols. Anti-CHLH (Wu et al., 2009), anti-ACTIN (Shang et al., 2010), anti-GFP (ZA009, Ythx Biotechnology), anti-ubiquitin (SC-166553, Santa Cruz Biotechnology) and anti-cLUC antibodies (Sigma, Cat No. L2164) were used to detect CHLH, ACTIN, ZTL-YFP, polyubiquitinated CHLH proteins and ZTL-cLUC, respectively. For hormone treatment, seedlings were sprayed with 7 ml of 0.005% Triton X-100 aqueous solution containing 100 μM (±) ABA, and samples were collected at the indicated time points.

The Matchmaker Gal4 two-hybrid system (Clontech, Cambridge, MA, United States) was used for screening the Arabidopsis cDNA library (ABRC, CD4-22). The ORFs of the N-terminus (amino acids 1–691), the middle fragment (amino acids 348–1,038) and the C-terminus (amino acids 692–1,381) of CHLH were fused to the GAL4 DNA binding domain in the plasmid pGBKT7 and were previously described (Shang et al., 2010; Zhang et al., 2014). The ORF of ZTL was cloned into pGADT7 plasmid by the NdeI (5’-end) and XhoI (3’-end) sites. Primers used in the vector construction are listed in Supplementary Table 1. Different combinations of plasmids were transformed into the yeast strain AH109. Transformants were plated on Leu-Trp-deficient and Leu-Trp-His-Ade-deficient medium separately and grew for 5–7 days at 30°C. pGBKT7-p53 and pGADT7-T were used as a positive control.

Firefly luciferase complementation imaging (LCI) assay was performed as previously described (Chen et al., 2008). The ORF of ZTL was inserted into cLUC at KpnI and SalI sites. Primers used in the vector construction are listed in Supplementary Table 1. The CHLH-nLUC construct was previously described (Shang et al., 2010). The constructs were transformed into A. tumefaciens strain GV3101. Bacterial suspensions were infiltrated into the leaves of 7-week-old Nicotiana benthamiana plants. After infiltration, plants were placed in dark for 44 h, followed by infiltration of an aqueous solution containing 100 μM (±) ABA and placed 4 h in dark again. LUC activity was recorded with a CCD imaging apparatus (Andor iXon).

The full-length cDNA of CHLH and ZTL was ligated into modified pMCSG7 protein expression vectors that contained GST or MBP tags at N-terminus, respectively. Primers used in the vector construction are listed in Supplementary Table 1. Protein purification, pull down and co-immunoprecipitation assays were conducted as previously described (Shang et al., 2010; Zhang et al., 2013).

For cell-free degradation assays, seedlings were sprayed with 7 ml of 0.005% Triton X-100 aqueous solution containing 100 μM (±) ABA for 2 h before sampling. Samples were ground in liquid nitrogen, resuspended in cell-free buffer (50 mM Tris–HCl pH 7.5, 100 mM NaCl, 5 mM MgCl₂, 5 mM DTT, 5 μM ATP) and clarified by centrifugation at 3,300 g. Equal amounts of extracts were transferred to individual tubes and incubated at 30°C with or without 50 μM MG132 as indicated. Reactions were stopped by adding protein gel-loading buffer. For in planta degradation assay, 2-week-old seedlings were transferred to MS liquid medium for 24 h, and then seedlings were incubated with 100 μM cycloheximide with or without 50 μM MG132 for 2 h in agitation, followed by the addition of (±) ABA to a final concentration of 100 μM. Then, samples were collected at the indicated time points and processed following the standard protocol.

For phosphorylation assays, proteins were extracted in AP buffer (10 mM Tris–HCl pH 8.0, 5 mM MgCl₂, 100 mM KCl, 0.02% Triton X-100) and separated by a Mn²⁺-Phos-tag (Wako) SDS-PAGE following manufacturer’s recommendations. Briefly, standard SDS-acrylamide gels were supplemented with or without 30 μM Phos-tag and 0.2 mM MnCl₂. Before transferring to nitrocellulose membranes, gels were washed two times with protein transfer buffer (25 mM Tris, 192 mM glycine, 20% (v/v) methanol) supplemented with 1 mM EDTA for 10 min, followed by an additional wash with transfer buffer without EDTA. Subsequent steps were performed according to standard western blot protocols. For alkaline phosphatase treatments, proteins were extracted in AP buffer and incubated with 10 U of FastAP alkaline phosphatase (Thermo Scientific) for 1 h at 30°C.

Statistical analyses were performed using the GraphPad Prism software (GraphPad Software, Inc. United States). For stomatal aperture and plant survival statistical analysis, two-way ANOVA tests followed by Dunnett’s multiple comparison tests were performed. Genotype and time after ABA addition or genotype and plant survival were considered as variables.

CHLH was reported to regulate a vital central ABA signaling network (Shang et al., 2010; Zhang et al., 2013, 2014; Mei et al., 2014), and the cytosolic C-terminus of CHLH plays a central role in ABA signaling (Wu et al., 2009). In a yeast two-hybrid screen using the C-terminus of CHLH (amino acids 692–1,381) as a bait, and from 3 × 10⁶ colonies screened, we found a CHLH interaction partner that is a clock-associated PAS protein, ZTL. Moreover, we conducted IP-MS assays to screen interaction partners of ZTL in vivo using an anti-GFP antibody in ZTL-overexpressing plants (with a YFP tag at the C-terminus of ZTL, named ZTL-ox). Our MS results confirmed previously reported results that TOC1, PRR5 and GIGANTEA (GI) interact with ZTL, and showed that CHLH was associated with ZTL in vivo (Supplementary Data 1, 2). Furthermore, we performed yeast two-hybrid, pull-down, firefly luciferase complementation imaging (LCI) and co-immunoprecipitation (Co-IP) assays to confirm the interaction in vitro and in vivo. In the yeast two-hybrid system, the C-terminus of CHLH_692–1,381 interacts with ZTL (Figure 1A), indicating that the CHLH-ZTL interaction takes place in the C-terminal cytosolic portion of CHLH. Then, we tested whether other truncated CHLHs interact with ZTL. Our results showed that the N-terminus of CHLH_1–691 (Shang et al., 2010) does not interact with ZTL (Supplementary Figure 2C), while the central region of CHLH_348–1,038 (Zhang et al., 2014) interacts with ZTL in yeast (Supplementary Figure 2A), suggesting that the interaction region of CHLH and ZTL may be from the C-terminal amino acids 692–1,038 of CHLH (Supplementary Figure 2B). Moreover, our pull-down assays indicated that ZTL directly interacts with CHLH in vitro (Figure 1C). Furthermore, we confirmed the CHLH-ZTL interaction in vivo with LCI and Co-IP assays (Figures 1B,D), and found that ABA enhances the physical interaction between CHLH and ZTL in tobacco (Figure 1B).

Previous studies have reported that TOC1 negatively participates in ABA-induced stomatal closure and plant responses to drought by inhibiting CHLH messenger RNA (mRNA) expression (Legnaioli et al., 2009). Then, we tested whether ZTL is involved in ABA signaling correlating with its known role in TOC1 degradation. To that end, we used two background Arabidopsis plants to conduct ABA phenotypic tests: the ztl-1 mutant (Somers et al., 2000) and ZTL-ox had a C24 background, and the ztl-3 mutant (Jarillo et al., 2001) had a Col background. In this work, we observed that ZTL mutants showed ABA hypersensitive phenotypes in ABA-inhibited early seedling growth and ABA-induced stomatal closure, while ZTL-overexpression plants displayed hyposensitive phenotypes (Figures 2A–D). However, ZTL miss-expressing seeds did not show significant differences from their corresponding WT seeds in ABA-induced inhibition of seed germination (Supplementary Figure 1).

Drought tolerance phenotypes are commonly associated with enhanced stomatal closure, which can minimize water transpiration from the leaves under drought conditions (Yamaguchi-Shinozaki and Shinozaki, 2006; Sirichandra et al., 2009). Therefore, we checked whether ZTL miss-expressing plants exhibited drought tolerance-related phenotypes. Consistent with the results of ABA-induced stomatal closure, ztl-1 and ztl-3 plants showed enhanced water-holding capacity and drought tolerance, whereas ZTL-ox plants exhibited reduced capacity to conserve water and drought tolerance (Figures 2E–H). Taken together, our results regarding ABA-induced stomatal closure, water loss and drought tolerance showed that the phenotype of ZTL mutants was similar to TOC1 mutants, and the phenotype of ZTL-overexpressing plants was similar to TOC1-overexpressing plants, implying that ZTL negatively regulates ABA signaling independent of its role in TOC1 regulation.

Given that ZTL modulates ABA signaling independent of its role in proteasome-mediated degradation of TOC1, we studied whether CHLH mRNA levels correlate with altered TOC1 protein levels in ZTL miss-expressing plants. To that end, we performed qPCR assays to determine CHLH mRNA levels during the diurnal cycle. Our results showed overall lower transcript levels of CHLH in both ztl-1 (Figure 3A) and ztl-3 (Figure 3B) plants, whereas a clear up-regulation of CHLH was observed in ZTL-overexpressing plants (Figure 3A). These results indicated that alteration of CHLH RNA levels in ZTL miss-expressing plants is most likely due to alterations in TOC1 protein levels, suggesting that ZTL regulates the expression of CHLH by TOC1. To obtain a deeper understanding of why altered transcript levels of CHLH in ZTL miss-expressing plants lead to unexpected phenotypes in ABA-induced stomatal closure and drought tolerance, we studied whether the mRNA and protein levels of CHLH correlate in ZTL miss-expressing plants. To that end, we performed western blot analysis to study CHLH protein abundance during the diurnal cycle. Our results showed that CHLH protein displays a clear correlation with mRNA levels in WT (C24 and Col) plants, exhibiting high levels during the first half of the day, gradually decreases during the second half and reaches a minimum before dawn (Figures 3C–F). However, CHLH mRNA levels do not correlate with protein levels in ZTL miss-expressing plants. ZTL mutants displayed overall lower levels of mRNA but higher levels of CHLH protein, and ZTL-ox plants exhibited overall higher levels of mRNA but lower levels of CHLH protein (Figure 3). Higher levels of CHLH protein in ZTL mutants and lower levels of CHLH protein in ZTL-overexpressing plants are consistent with observed stomatal closure and drought-related phenotypes in ZTL miss-expressing plants. Moreover, the diurnal oscillation of CHLH protein was severely disrupted in the ztl-1 and ztl-3 mutants (Figures 3C–F). Taken together, our qPCR and western blot results suggested that ZTL might regulate CHLH protein stability at the post-translational level.

Our qPCR and western blot results indicate a possible role of ZTL in the post-translational regulation of CHLH (Figure 3). Considering that ZTL is an E3 ligase of the SCF complex, we tested whether CHLH is degraded by the 26S proteasome and whether ABA has any effect on this regulation. Firstly, we examined CHLH protein stability using a previously described cell-free degradation assay (Más et al., 2003). Our results revealed that CHLH protein was unstable and exhibited a subtle reduction over time, and ABA treatment caused a clear reduction in CHLH abundance. However, protein extracts treated with MG132 markedly stabilized CHLH protein regardless of ABA treatment (Figures 4A,B), suggesting that CHLH is degraded by the 26S proteasome pathway, and ABA stimulates the proteasome-dependent degradation of CHLH. To confirm the cell-free results, we performed protein stability assays in vivo. De novo protein synthesis was first blocked by addition of the protein synthesis inhibitor cycloheximide (CHX). Our studies indicated a clear reduction in CHLH protein over the time course (Figures 4C,F), and this reduction was faster after ABA treatment (Figures 4D,F), while CHLH protein was stable in plants treated with both ABA and MG132 (Figures 4E,F), suggesting that ABA induces CHLH degradation by the 26S proteasome. Moreover, we found that CHLH degradation was repressed in the ztl-3 mutant, and ABA also did not induce CHLH degradation in the ztl-3 mutant (Supplementary Figure 3).

Meanwhile, we detected the polyubiquitination of CHLH in vivo. Total protein extract was immunoprecipitated with anti-CHLH antibody, and precipitated proteins were subjected to immunoblot analysis with anti-ubiquitin antibody (Figure 4G, upper panel). Our western blot results showed that ubiquitination of CHLH was significantly attenuated in the ztl-3 mutant at ZT19. Taken together, our results support the notion that ABA induces CHLH degradation via the ubiquitin-dependent 26S proteasome pathway, and ZTL is responsible for CHLH degradation in plants.

Our results indicate that external application of ABA enhances the interaction between ZTL and CHLH and ZTL-mediated CHLH proteasomal degradation (Figures 1B, 4). However, the molecular mechanism underlying how ABA changes the affinity of ZTL to CHLH needs to be explored.

Previous studies reported that phosphorylation of the E3 ligase smurf1 switches its substrate preference in support of axon development (Cheng et al., 2011). And protein phosphorylation plays crucial roles in circadian clock and ABA signaling (Sugano et al., 1998; Daniel et al., 2004; Fujiwara et al., 2008; Portolés and Más, 2010; Wang et al., 2010; Kulik et al., 2011; Kusakina and Dodd, 2012; Yang et al., 2017). Therefore, we tested whether ZTL is phosphorylated and whether ABA can impact this process. To that end, we used a compound called Phos-tag ligand, which provides phosphate affinity SDS-PAGE for mobility shift detection of phosphorylated proteins (Komis et al., 2014). Our results showed that conventional western blots only exhibited a unique band for the ZTL-YFP protein, while application of Phos-tag to the SDS gels allowed to resolve two ZTL-YFP bands (Figures 5A,B), indicating that ZTL may have at least two isoforms, one phosphorylated and one not. More importantly, our results showed that the external application of ABA progressively reduced the proportion of the lower band (non-phosphorylated ZTL), and the band completely disappeared after 4-h ABA treatment (Figures 5B,D), suggesting that ABA induces ZTL phosphorylation. To further confirm that the two ZTL-YFP bands observed in the Phos-tag blots correspond to phosphorylated and non-phosphorylated isoforms, we treated the plants with 100 μM (±) ABA for 4 h (ZTL was presumably all phosphorylated), and then we treated protein extracts with alkaline phosphatase to recover its un-phosphorylated state. Indeed, the alkaline phosphatase treated extracts displayed two specific bands (Figure 5C), indicating that alkaline phosphatase partially restored the un-phosphorylated state of the ZTL protein pull.

In summary, our results showed that ABA induces ZTL phosphorylation, ABA enhances the interaction between ZTL and CHLH, and induces ZTL-mediated CHLH proteasomal degradation (Figures 1B, 4, 5). These results suggest that the phosphorylated ZTL protein may exhibit higher affinity to CHLH, resulting in lower levels of CHLH protein after ABA treatment.

The CHLH mutant cch displayed strong insensitivity to ABA in stomatal movement and reduced drought tolerance (Shen et al., 2006; Wu et al., 2009). We introduced the cch mutant into the ztl-3 mutant, and the double mutant ztl-3 cch exhibited an extremely similar phenotype to cch in ABA-induced stomatal closure, water loss and drought tolerance (Figure 6). These solid genetic data showed that ZTL mediates ABA signaling upstream of CHLH, which is consistent with its role in proteasomal degradation of CHLH.

The key clock component TOC1 is targeted for proteasome-mediated degradation through specific interaction with ZTL (Más et al., 2003). Previous studies showed that overexpression of TOC1 led to enhanced abscisic acid sensitivity during seed germination (Castells et al., 2010). Moreover, TOC1 was reported to negatively mediate ABA signaling in ABA-induced stomatal closure and plant responses to drought by inhibition of CHLH expression (Legnaioli et al., 2009). Our phenotypic results showed that ZTL mutants display enhanced ABA-induced stomatal closure and drought tolerance, similar to TOC1 mutants, and ZTL-overexpressing plants exhibit reduced ABA-induced stomatal closure and drought tolerance, similar to TOC1-overexpressing plants (Figure 2). These results indicate that ZTL negatively regulates ABA signaling independent of its role in proteasomal degradation of TOC1. In addition, ZTL miss-expressing seeds showed a wild-type ABA response in ABA-inhibited seed germination (Supplementary Figure 1), implying that ZTL may have TOC1-independent roles during germination, which can compensate for its TOC1-dependent defects in ABA-inhibited seed germination. In agreement with this idea, PRR5, another known target of ZTL, was reported to negatively regulate the biosynthetic pathways of ABA (Kiba et al., 2007; Fukushima et al., 2009). Moreover, EARLY BIRD (EBI) was reported to associate with ZTL and regulate abiotic stress signaling and the circadian clock (Lisso et al., 2006; Johansson et al., 2011). A recent study reported that ZTL positively regulates ABA-induced stomatal closure with PRR5 and OST1 under LD (12 h light: 12 h dark) and CO₂-free aeration (Jurca et al., 2022). Our results showed that ZTL negatively regulates ABA signaling under long day (LgD, 16 h light: 8 h dark) and CO₂-normal conditions. Due to ZTL is a circadian clock protein, different growth conditions and inconsistent treatment conditions (CO₂-free aeration/CO₂-normal conditions) might lead to changes in ZTL interacting proteins at the same sampling time, which might account for the phenotypic inconsistency of ZTL. Taken together, these results suggested that ZTL might mediate complicated ABA signalings in plants. However, our genetic data showed that disruption of CHLH suppresses ABA hypersensitivity in the ztl-3 mutant (Figure 6), supporting the ZTL-CHLH module in ABA signaling. Additionally, our results indicated that ZTL plays an important role in ABA signaling independent of its regulation of TOC1 degradation, and this role is able to overcome its indirect effect on TOC1 protein turnover.

Our qPCR and western blot results showed that ZTL mutants displayed overall lower levels of CHLH mRNA but higher levels of CHLH protein, and ZTL-overexpression plants exhibited overall up-regulation of CHLH mRNA but down-regulation of CHLH protein (Figure 3). These results suggested that alteration of CHLH transcript levels in ZTL miss-expressing plants correlates with alterations in TOC1 protein levels, indicating that ZTL regulates CHLH mRNA expression through proteasomal degradation of TOC1. The inconsistent expression of CHLH mRNA and protein in ZTL miss-expressing plants suggests that ZTL may have a role in post-translational regulation of CHLH. Indeed, our genetic and biochemical results showed that CHLH is targeted by ZTL for proteasome-mediated degradation in modulating ABA signaling (Figures 3, 4, 6). CHLH protein exhibited diurnal oscillation in WT plants (Figures 3C–F). This is consistent with the circadian regulation of ZTL protein abundance by GIGANTEA (GI). The GI-ZTL interaction stabilizes the ZTL protein at the end of the day, and then allows higher accumulation of the ZTL protein during the evening (Kim et al., 2007), when CHLH protein abundance is lower (Figures 3C–F).

Turnover of internal chloroplast proteins is controlled by proteases inherited from the organelle’s prokaryotic ancestor, but the molecular mechanisms of ubiquitin-dependent degradation of chloroplast-associated proteins are still poorly understood. Although it is still controversial whether ZTL localizes in the nucleus, it has been well confirmed that ZTL localizes in the cytoplasm (Kim et al., 2007; Takase et al., 2011). CHLH localizes in the chloroplast outer membrane with its N and C termini exposed to the cytosol (Shang et al., 2010). Furthermore, our results in yeast two-hybrid assays indicate that ZTL interacts with the cytosolic C-terminus of CHLH (Figure 1A and Supplementary Figure 2). These results provide a plausible explanation for how CHLH and ZTL proteins can interact at a subcellular level. Previous studies reported that the chloroplast outer envelope membrane (OEM)-localized ubiquitin E3 ligase SP1 is responsible for the degradation of OEM-localized components of protein import translocases via the cytosolic 26S proteasome (Ling et al., 2012, 2019). Further studies found that SP2 and CDC48A cooperated to bring about retrotranslocation of ubiquitinated substrates from the OEM to the cytosol. Interestingly, our IP-MS results also found that CDC48A associates with ZTL in vivo (Supplementary Data 1, 2), implying CDC48A may also mediate CHLH retrotranslocation to the cytosol. Taken together, previous and our results together support the conclusion that chloroplast outer envelope membrane proteins are degraded through the cytoplasmic 26S proteasome. However, we first reported that chloroplast outer membrane-localized proteins are targeted by cytosolic ubiquitin E3 ligases for proteasome-mediated degradation, which deepens our understanding of ubiquitin-dependent chloroplast-associated protein degradation in plants.

Previous studies revealed that phosphorylation of some clock components affects their subcellular localization, turnover and affinity to their substrates, constituting an important mechanism for clock function in Arabidopsis, Neurospora, Drosophila and humans (Edery et al., 1994; Liu et al., 2000; Bao et al., 2001; Toh et al., 2001; Ko et al., 2002; Akten et al., 2003; Daniel et al., 2004; Merrow et al., 2006; Vanselow and Kramer, 2007; Portolés and Más, 2010; Robles et al., 2017). Our results also indicate that high levels of ABA induce ZTL phosphorylation (Figure 5). Furthermore, we found that ABA enhances ZTL and CHLH interaction (Figure 1B), and induces ZTL-mediated CHLH degradation (Figures 4A–F). These results suggest that phosphorylated ZTL may display a higher affinity for CHLH, resulting in the enhanced degradation of CHLH after ABA treatment. Previous studies showed that ABA-induced TOC1 expression enhances TOC1-mediated transcriptional repression of CHLH, while ABA-induced TOC1 expression also requires a functional CHLH, suggesting that the reciprocal regulation between TOC1 and CHLH might function as a fine-tuned switch that helps to modulate plant sensitivity to ABA, which in turn favors the temporal regulation of plant responses to drought stress (Legnaioli et al., 2009). Consistently, our results showed that ABA represses the expression of CHLH at both transcriptional and post-transcriptional levels. On the one hand, high levels of ABA induce the expression of TOC1 protein by ABA-induced transcriptional regulation of TOC1 (Legnaioli et al., 2009). On the other hand, ABA stimulates ZTL-mediated CHLH proteasomal degradation, and the “enhanced” repression of CHLH better modulates plant sensitivity to ABA under drought conditions (Figure 7).

In ABA signaling, the RCAR/PYR/PYL-PP2C-SnRK2 module is considered the main pathway, and SnRK2 proteins initiate a phosphorylation cascade in presence of ABA (Ma et al., 2009; Park et al., 2009; Miyakawa et al., 2013). Using the Arabidopsis Interaction Viewer,¹ we found at least two members of the SnRK2 family of kinases as putative ZTL interactors (Geisler-Lee et al., 2007). Additionally, previous studies reported that SKP1/ASK1, a conserved SCF (Skp1-cullin-F-box) ubiquitin ligase subunit, interacts with SnRK protein kinases to mediate proteasomal binding of a plant SCF ubiquitin ligase, suggesting that SnRK protein kinases are associated with the 26S proteasome in plants (Farrás et al., 2001). Moreover, our IP-MS results showed that SnRK2.1 and SnRK2.2 associate with ZTL in vivo (Supplementary Data 1). Taken together, SnRK kinases may be responsible for ZTL phosphorylation in response to ABA. Future studies are needed to confirm this regulation and will be crucial to provide the molecular link between ABA signaling and the circadian clock.