The plant material used was Arabidopsis (Arabidopsis thaliana). atann2 (AT5G65020: SALK_054223) and atann3 (AT2G38760: SALK_082344) were purchased from the Arabidopsis Biological Resource Center (ABRC). The corresponding T-DNA insertion mutants were analyzed by PCR using T-DNA primers and gene-specific primers. For identification at the RNA level by RT-PCR using gene-specific primers, AtACTIN2 (AT3G18780) was used as internal control. The primers are listed in Supplementary Table 1.

The native promoters and coding sequence (CDS) of AtANN2, AtANN3, and AtCRY2 (AT1G04400) were synthesized and cloned into pCAMBIA1300-nos by Sangon Biotech (Shanghai) Co., Ltd. Then, the vector was transformed into these null mutants to generate the complementary lines (AtANN2-COM and AtANN3-COM). The CDSs were cloned into pSuper1300-GFP, and the vector was also transformed into Col-0 to construct the overexpression lines (AtANN2-OE and AtANN3-OE).

Arabidopsis seeds were protected from light for 3 days at 4°C and then incubated under different photoperiods. The long-day conditions were as follows: 16 h light/8 h dark, light intensity 100 μmol m^–2 s^–1, temperature 22°C light/18°C dark, and relative humidity 70%. The short-day conditions were as follows: 8 h light/16 h dark, light intensity of 200 μmol m^–2 s^–1, temperature of 22°C light/18°C dark, and relative humidity of 70%.

The plants used in the growth phenotype assays in response to drought stress were grown in soil. The plants used in the gene expression, β-glucuronidase (GUS) staining, subcellular localization, and transmembrane Ca²⁺ flow assays were grown in 1/2 Murashige and Skoog (MS) medium and 1% (w/v) agar.

For the drought tolerance tests, long-day conditions included 9 days of drought treatment after 3 weeks of normal plant growth, and short-day conditions included 9 days of drought treatment after 4 weeks of normal growth. The mean number of senescing leaves per plant of different genotypes was counted (n = 50).

Quantitative real-time PCR analysis (qRT-PCR) was performed to analyze the gene expression levels. AtACTIN2 was used as internal control, and the RNA extraction was performed using the RNAprep Pure Plant Kit (Tiangen). cDNA was obtained by the reverse transcription of 2 μg of extracted RNA using iScript^TM gDNA Clear cDNA (Bio-Rad). qRT-PCR was performed using a PCR system according to the instructions of SYBR Premix EX Taq (Takara). The PCR program was applied by a StepOne Plus Real Time PCR system (QuantStudio 6 Flex, ABI). Three biological replicates were performed per sample.

The native promoters of AtANN2, AtANN3, and AtCRY2 were synthesized and cloned into pCAMBIA1391-GUS by Sangon Biotech (Shanghai) Co., Ltd. The AtANN2promoter:GUS and AtANN3promoter:GUS transgenic plants were obtained to observe GUS expression in the plants. The plant tissues were incubated in rinse solution [50 mM Na₂HPO₄, 50 mM NaH₂PO₄,0.5 mM K₃Fe(CN)₆, and 0.5 mM K₄Fe(CN)₆⋅3H₂O] for 5 min and then transferred to stain solution [50 mM Na₂HPO₄, 50 mM NaH₂PO₄,0.5 mM K₃Fe(CN)₆,0.5 mM K₄Fe(CN)₆⋅3H₂O, 10 mM EDTA Na₂, 1% Triton-100 and 2 mM X-Gluc] for incubation for 12 h at 37°C. The plants were eluted two- to three-fold using 95% ethanol at 65°C for 12 h. After decolorization, the specimens were observed under a stereomicroscope (Leica DVM6).

Transient expression of AtANN2-GFP and AtANN3-GFP was performed by Wuhan Edgene Bio-Tech. (Wuhan) Co., Ltd. using the leaf cell protoplasts. Arabidopsis leaves of 3-week-old plants were selected, soaked in enzyme solution [1.5% Cellulase R10,0.75% Macerozyme R10, 600 mM mannitol, 10 mM MES (pH 5.7), and 0.04% 2-hydroxy-1-ethanethiol] for 3 h, and evacuated at 23°C. The samples were filtered into Eppendorf tubes with 40-μm nylon gauze and centrifuged at 400 rpm for 5 min at 4°C. The supernatant was discarded, the samples were washed twice with pre-chilled W5 solution [154 mM NaCl, 125 mM CaCl₂, 2 mM KH₂PO₄, and 2 mM MES (pH 5.7)], W5 solution was added for suspension, and then, the samples were allowed to stand for 30 min at 4°C. Next, MMG solution [400 mM mannitol, 15 mM MgCl₂.6H₂O, and 4 mM MES (pH 5.7)] was added for suspension, and then microscopic examination was performed, with 20∼40 protoplasts per field of view obtained under 40 × magnification. The CDSs of AtANN2 and AtANN3 were synthesized by Sangon Biotech (Shanghai) Co., Ltd., cloned into the vector pEGOEP_35S-H-GFP, and transformed into the leaf cell protoplasts of Col-0 and AtCRY2-OE. Subsequently, 10 μl of target gene plasmid, 100 μl protoplasts, and 110 μl 40% PEG4000 solution were slowly mixed upside down, and then, the mixture was placed in a 22.5°C water bath for 15∼20 min. W5 solution was added to dilute the protoplasts, and the samples were mixed well to terminate the reaction. The samples were centrifuged at 400 rpm for 5 min at 4°C. The supernatant was discarded, and the samples were washed twice with pre-chilled W5 solution. The samples were incubated at 23°C under low light overnight and then observed and photographed under a laser confocal microscope (Olympus FV3000) at an excitation wavelength of 488 nm and an emission wavelength of 510∼530 nm.

The subcellular localization of the AtANN2-GFP and AtANN3-GFP fusion proteins was observed after 300 mM mannitol treatment using 10-day-old transgenic plants under different photoperiods, and GFP signals were observed under a laser confocal microscope (Olympus FV3000).

For the GST pull-down assay, the coding sequences of AtANN2, AtANN3, and AtCRY2 were synthesized by Sangon Biotech (Shanghai) Co., Ltd. The GST pull-down assay was performed by Wuhan Edgene Bio-Tech. (Wuhan) Co., Ltd. AtCRY2 was cloned into the vector pET28a(+), and AtANN2 or AtANN3 was cloned into the vector pGEX-6P-1. Proteins expressed in Escherichia coli were purified as previously described (Cui et al., 2018).

A bimolecular fluorescence complementation (BiFC) assay was performed by Wuhan Edgene Bio-Tech. (Wuhan) Co., Ltd. AtCRY2 was cloned into the vector pEGOEP_2X35S-cYFP, and AtANN2 or AtANN3 was cloned into the vector pEGOEP_2X35S-nYFP. The two plasmids were co-transformed into Arabidopsis leaf cell protoplasts. The samples were incubated for 16 h in the dark at 23°C. The fluorescence signals were observed using a microscope (Olympus FV3000) at an excitation wavelength of 488 nm and an emission wavelength of 510–530 nm. The parameters were consistent for all samples.

All plates were oriented vertically and maintained in the chamber described above for 5 days. After 5 days, seedlings germinated on 1/2 MS medium were transferred to the test solution [0.1 mM KCl, 0.1 mM CaCl₂, 0.1 mM MgCl₂, 0.5 mM NaCl, 0.3 mM MES, and 0.2 mM Na₂SO₄, pH 6.0] 30 min for balancing. The Ca²⁺ fluxes of the main root meristematic zone were measured non-invasively by using Non-invasive Micro-test Technology NMT System [NMT100 Series, YoungerUSA LLC, Amherst, MA 01002, United States; Xuyue (Beijing) Sci. & Tech. Co., Ltd., Beijing, China and imFluxes V2.0 (YoungerUSA LLC, Amherst, MA 01002, United States) Software]. The Ca²⁺-microsensor (2–4 μm aperture, XYPG120-2; Xuyue Sci. and Tech. Co., Ltd) was pulled and silanized using glass micropipettes after being filled with a backfill solution of 100 mM CaCl₂. The backfill solution was pushed into the tip and occupied a length of approximately 1 cm from the tip. The plant materials were collected after the seedlings were exposed to the test solution of 300 mM mannitol or the control system for 10 min. The data were collected every 6 s during the 3 min pretreatment period and 10 min measuring period. These data were obtained from approximately five roots per genotype. The positive values in the figures represent a cation efflux or anion influx.

All experiments were repeated at least three times. The statistical analyses were indicated in each figure legend and performed using the programs SigmaPlot 12.0 and SigmaStat 3.5. Different letters indicate significant differences with each treatment at the p < 0.05 level. Statistical analyses were performed by one-way or two-way ANOVA, followed by Tukey’s multiple comparisons test.

To investigate the functions of AtANN2 and AtANN3 in the drought stress response, their corresponding T-DNA insertion mutants were isolated (Figure 1A). In wild type Col-0, a specific band ranging from 1,000 to 1,100 bp in size was amplified using a gene-specific primer (LP and RP), while no band was detected using the T-DNA primer (LB and RP). In atann2 (salk_054223) and atann3 (salk_082344), the T-DNA primer amplification yielded bands with sizes of 890 and 500 bp, respectively, indicating the presence of T-DNA insertion into the target genomic regions. The qRT-PCR analysis showed that the expression levels of AtANN2 and AtANN3 genes were completely disrupted in the null mutants atann2 and atann3, and were able to be recovered in the complementary lines AtANN2-COM and AtANN3-COM to the same level as those in Col-0. In overexpression lines AtANN2-OE and AtANN3-OE, the expression levels of AtANN2 and AtANN3 genes were enhanced up to four- to five-fold (Figure 1B).

The plant growth phenotypes of the obtained mutants and transgenic lines in response to drought stress were determined under long and short-day conditions (Figures 1C,D). The number of senescing leaves of plants was significantly increased by drought treatment compared to that in the untreated plants. Under short-day condition, compared to that in wildtype Col-0 (5.78), a higher number of senescing leaves was determined in the atann2 mutant (7.48), while less numbers of senescing leaves in the AtANN2-OE (OE1 and OE2) lines (2.48–2.92), suggesting that AtANN2 plays a role in drought tolerance. As a control, the AtANN2-COM (COM1 and COM2, 6.06–6.08) lines showed no difference from that of Col-0. Under long-day condition, however, the contribution of AtANN2 to drought tolerance was not determined as all genotypes showed the similar numbers of senescing leaves under drought treatment. Unlike AtANN2, the AtANN3-related genotypes showed different responses to drought stress under long-day condition rather than short-day condition. Compared to that in wildtype Col-0 (5.52), a higher number of senescing leaves was determined in the atann3 mutant (8.30), while less numbers of senescing leaves in the AtANN3-OE (OE1 and OE2) lines (2.66∼3.24), suggesting that AtANN3 plays a role in drought tolerance under long-day condition. These results indicated the functions of AtANN2 and AtANN3 in drought tolerance dependent on the photoperiod.

By qRT-PCR analysis, the AtANN2 gene was found to be expressed mainly in hypocotyls, roots, vegetative rosettes, and siliques, and the AtANN3 gene was mainly expressed in hypocotyls, cotyledons, and roots (Figure 2A). Using GUS reporter gene transgenic plants, AtANN2 and AtANN3 promoter activities were mainly expressed in the roots and cotyledons of 3-day-old seedlings, slightly increased in the root tips and shoots of 7-day-old seedlings, and further decreased in the roots and shoots of 10-day-old seedlings (Figure 2B). These expression patterns of AtANN2 and AtANN3 did not differ between long-day and short-day conditions.

The relative expression levels of AtANN2 and AtANN3 were determined in plants treated with 300 mM mannitol (Col-0) for 0,0.5, 1, 3, 6, 12, and 24 h using qRT-PCR (Figure 3A). Compared to the control treatment (0 h), no difference was found in the relative expression levels of AtANN2 and AtANN3 in the shoots within 24 h of treatment, while AtANN2 expression in the roots was increased to 2.30- and 2.11-fold after 12 and 24 h of treatment in the seedlings grown under short-day conditions, respectively. In contrast, AtANN3 expression in the roots was increased to 4.68-fold after 24 h of treatment in seedlings grown under long-day conditions. AtANN2 promoter activity was elevated in roots after 24 h of treatment under short-day condition, while AtANN3 promoter activity was elevated in the roots of treated plants under long-day condition (Figure 3B). These results demonstrated that the photoperiod regulated the expression patterns of AtANN2 and AtANN3 in roots in response to mannitol treatment.

Leaf cell protoplasts of Col-0 plants grown under different photoperiodic conditions were used to investigate the subcellular localization of the AtANN2-GFP and AtANN3-GFP fusion proteins in response to mannitol treatment for 0, 5, 10, 30, and 60 min (Figure 4A and Supplementary Figure 2A). Under normal conditions, AtANN2- and AtANN3-dependent GFP signals were widely expressed within the cells, not only at the plasma membrane. In the protoplasts of the seedlings cultured under short-day, but not long-day conditions, the AtANN2-dependent GFP signals rapidly targeted to the plasma membrane after 5 min of mannitol treatment. However, regarding AtANN3-GFP, similar subcellular localization changes in response to mannitol treatment were only observed in protoplasts under long-day condition.

These regulatory patterns of the AtANN2 and AtANN3 proteins were further analyzed using transgenic lines expressing AtANN2-GFP and AtANN3-GFP under their own native promoters (Figure 4B and Supplementary Figure 2B). Again, the cytosolic localization of AtANN2-GFP and AtANN3-GFP rapidly targeted to the plasma membrane in the roots of plants treated with mannitol for 5 min grown under short day condition and long-day condition, respectively. Thus, in response to mannitol treatment, the AtANN2 and AtANN3 proteins could target to the plasma membrane, and this regulatory process was regulated by the photoperiod.

As AtANN1 mediates Ca²⁺ transport across the plasma membrane (Mohammad-Sidik et al., 2021), we examined the transmembrane Ca²⁺ flow in AtANN2- and AtANN3-related mutants and transgenic plants using an NMT system (Figure 5A). The main root meristematic zone (within approximately 80 μm from the root tip) of Arabidopsis seedlings was used as the test area (Figure 5B). To eliminate the error caused by the instability of Ca²⁺ flow within 1 min after mannitol treatment, the mean values were calculated for 3 min before (pre 3 min) and 1–5 min after the treatment (Figure 5D). No difference was detected in the mean values of Ca²⁺ flow for all tested genotypes before 3 min of treatment. Mannitol treatment caused Ca²⁺ efflux from the root tips. The function of AtANN2 was manifested under short-day condition: the activation of Ca²⁺ efflux from the root tip in atann2 was weakened after mannitol treatment relative to that in Col-0, and AtANN2-COM did not differ from Col-0. The Ca²⁺ efflux was activated to a greater extent in AtANN2-OE. In contrast, the treatment effects on the related mutants of AtANN2 were not observed under long-day condition (Figures 5C,D).

Unlike AtANN2, the function of AtANN3 manifested under long-day condition, but not short-day condition, as follows: the activation of Ca²⁺ efflux from the root tip in atann3 was weakened after mannitol treatment relative to that in Col-0 and AtANN3-COM, while Ca²⁺ efflux was activated to a greater extent in AtANN3-OE (Figures 5C,D). Therefore, these results suggested that AtANN2 and AtANN3 significantly affected Ca²⁺ flow across the plasma membrane, and these functions manifested only under specific photoperiodic conditions.

As the functions of AtANN2 and AtANN3 were regulated by photoperiod, we investigated whether the photoreceptor AtCRY2 could interact with the AtANN2 and AtANN3 proteins. By the pull-down approach to detect protein-protein interactions in vitro, AtANN2-GST was found to interact with AtCRY2-HIS, while AtANN3-GST interacted with AtCRY2-HIS (Figure 6A). The in vivo interaction was further revealed by bimolecular fluorescence complementation (BiFC) analysis (Figure 6B). AtCRY2-YCE could interact with AtANN2-YNE and AtANN3-YNE, which again demonstrated the interaction of AtCRY2 with AtANN2 and AtANN3 in Arabidopsis protoplasts. Based on these biochemical evidence, AtCRY2 could interact with AtANN2 and AtANN3 in plants and might play a role in the signaling pathway associated with the photoperiodic regulation of AtANN2 and AtANN3.

AtCRY2 has been found to play a negative regulatory role in drought stress, with the cry1cry2 double mutants showing better growth under drought stress. The cry2 single mutant showed a similar phenotype as WT, and AtCRY2-OE has more sensitive to blue light regulation of stomatal opening (Mao et al., 2005). To test whether the function of AtCRY2 in drought results from its regulation on AtANN2 and AtANN3, we generated transgenic lines overexpressing the AtANN2 and AtANN3 genes in the AtCRY2-OE background (AtANN2/CRY2-OE and AtANN3/CRY2-OE). The enhanced expression levels of AtANN2 and AtANN3 in the AtCRY2-OE background (3∼6-fold) did not differ from those in the Col-0 background (Figure 7A).

Overexpression of AtANN2 in Col-0 enhanced drought tolerance as revealed by less number of senescing leaves (2.48–2.92) under short-day condition (Figures 1C,D). However, the physiological contribution of the overexpressed AtANN2 protein was not observed in the AtCRY2-OE background with the number of senescing leaves (7.92–8.16) (Figures 7B,C). This finding indicated that overexpression of AtCRY2 could negatively regulate the function of AtANN2 in drought tolerance under short-day condition. The same negative regulation of AtANN3 under long-day condition was also shown in transgenic lines overexpressed AtANN3 in the AtCRY2-OE background with the number of senescing leaves (7.66–7.80) compared to those in the Col-0 background (2.66–3.24) (Figures 7B,C). Thus, these results indicated that the function of AtANN2 and AtANN3 in drought tolerance could be repressed by AtCRY2 overexpression under specific photoperiods.

To understand how the overexpression of AtCRY2 affected the functions of AtANN2 and AtANN3, we first investigated the gene expression patterns of AtANN2 and AtANN3 in the AtCRY2-OE lines by qRT-PCR. After mannitol treatment for 12 and 24 h, the expression of AtANN2 in roots was increased up to 2.22- to 2.50-fold under short-day condition, and the expression of AtANN3 was increased up to 2.06- to 3.28-fold under long-day condition (Figure 8A). These patterns were further confirmed by transgenic lines expressing the corresponding GUS reporter genes for AtANN2 and AtANN3 (Figure 8B). The expression levels of AtANN2 and AtANN3 in AtCRY2-OE were similar to that in Col-0 (Figure 3), indicating that AtCRY2 did not regulate AtANN2 and AtANN3 expression at the transcriptional level.

We then investigated changes in the subcellular localization of AtANN2 and AtANN3 by AtCRY2 overexpression. After mannitol treatment, unlike that in Col-0 (Figure 4A and Supplementary Figure 2A), neither AtANN2-GFP nor AtANN3-GFP was able to target to the plasma membrane within 60 min in the protoplast of AtCRY2-OE (Figure 9A and Supplementary Figure 3A) irrespective of long-day or short-day conditions. By analysis of transgenic lines expressing AtANN2-GFP and AtANN3-GFP in AtCRY2-OE background under control of their native promoters, it revealed that the subcellular localization of AtANN2-GFP and AtANN3-GFP appeared no longer at the plasma membrane after 5 min of mannitol treatment (Figure 9B and Supplementary Figure 3B) as they did in Col-0 background (Figure 4B and Supplementary Figure 2B). Thus, AtCRY2 was proposed to repress function of AtANN2 and AtANN3 by affecting their plasma membrane localization in response to mannitol treatment.

To test whether AtCRY2 affected the calcium channel activity of AtANN2 and AtANN3 in the drought signaling pathway, the transmembrane Ca²⁺ flow analysis was performed in transgenic lines overexpressing AtANN2 and AtANN3 either in Col-0 (AtANN2-OE and AtANN3-OE) or CRY2-OE background (AtANN2/CRY2-OE and AtANN3/CRY2-OE). Unlike that of AtANN2-OE and AtANN3-OE, mannitol activation of Ca²⁺ efflux was diminished in AtANN2/CRY2-OE under short-day condition and in AtANN3/CRY2-OE under long-day condition (Figure 10). This result suggested that AtCRY2 could negatively regulate the ion flow across the plasma membrane mediated by AtANN2 and AtANN3.

The functions of AtANN1 and AtANN4 depended on the photoperiod implied that the physiological functions of annexins might change under different environmental signals, such as different photoperiods. Konopka-Postupolska et al. (2009) found that atann1 was more sensitive to drought stress under short-day condition relative to Col-0. However, Huh et al. (2010) found that under long-day condition, AtANN1 and AtANN4 interacted, and atann1/ann4 was more resistant to drought stress than atann1 or atann4. The mechanism by which light signals regulated annexins’ function and thus affect plant stress tolerance response is currently unknown. In this study, the physiological functions of AtANN2 and AtANN3 were examined under the same experimental system, which were found to have a positive regulatory effect on drought stress tolerance, with AtANN2-related genotypes showed different responses to drought stress under short-day condition, while a role of AtANN3 revealed in drought tolerance under long-day condition. These results indicated that the functions of AtANN2 and AtANN3 in drought tolerance depended on the photoperiod (Figure 1).

Eight members of the Arabidopsis annexin family were highly expressed at the seed germination and seedling stages. In this study, we also found that AtANN2 and AtANN3 expressed mainly in roots and cotyledons of 3-day-old seedlings, slightly increased in root tips and shoots of 7-day-old seedlings (Figure 2B). Cantero et al. (2006) also found that AtANN2 was mainly expressed in root epidermal cells and lateral root primordium cells, and AtANN2 was assumed to play a role in root responses to external signals. In this study, we found that the photoperiod regulated the expression patterns of AtANN2 and AtANN3 in roots responded to mannitol treatment (Figure 3). Other studies have also found that light signals can regulate the expression of other annexin genes as follows (Baucher et al., 2011): AtANN1 and AtANN4 were regulated by AtHY5 (Lee et al., 2007). AtANN5 expression was increased in hypocotyls by red light irradiation, and far-red light reversed this response (Cantero et al., 2006). Annexin gene expression in tobacco (Nicotiana tabacum) was also regulated by light signals as follows: gene expression was evident in roots under light conditions, but no expression in roots was observed under dark conditions (Baucher et al., 2012).

Protein level regulation enabled a more rapid response to external signals and changed in function through regulation of protein activity and localization. The subcellular localization of proteins was closely related to the functions of annexins, and the same annexin could be localized in multiple regions in the cell (Laohavisit and Davies, 2009). This study found that under normal conditions AtANN2- and AtANN3-dependent GFP signals were widely expressed within the cells, not only at the plasma membrane. However, in response to mannitol treatment, the AtANN2 and AtANN3 proteins could be targeted to the plasma membrane, and this regulatory process was regulated by photoperiod (Figure 4). Other researchers have also found similar phenomena, such as AtANN1 translocation from the cytoplasm to the plasma membrane in root cells under salt stress (Lee et al., 2004). Wheat (Triticum aestivum) annexins under cold stress started to aggregate at the plasma membrane and functioned as calcium channels (Breton et al., 2000). Annexins are localized in other subcellular structures, e.g., the nucleus, vesicle membrane, and Golgi apparatus (Konopka-Postupolska and Clark, 2017). Different locations may be associated to different functions and regulatory proteins, and whether AtANN2- and AtANN3-regulated functions are related to other subcellular organelles, such as the nuclei, requires further investigation.

Plants receive drought signals, initiate regulation at the transcriptional level after a long period, and induce changes in subcellular localization within a short period. A multilayered pattern of regulation helps plants adapt to changing environments faster with less consumption.

In response to mannitol treatment, the AtANN2 and AtANN3 proteins could be targeted to the plasma membrane, and the altered subcellular localization was closely related to their protein functions. Certain annexins localized in the plasma membrane are currently known to be closely related to transmembrane ion transport, such as AtANN1, which constructs the root epidermal cell plasma membrane Ca²⁺ channel (Laohavisit et al., 2012; Richards et al., 2014; Zhao et al., 2019; Mohammad-Sidik et al., 2021). On the other hand, the ability of annexins to regulate ion transport may be related to their regulated resistance responses, for example, AtANN1 acted as a root plasma membrane calcium channel during salt, hyperosmotic, oxidative, and cold stress exposure by regulating the Ca²⁺ concentration in response to external changes (Laohavisit et al., 2012, 2013; Richards et al., 2014; Liu et al., 2021). AtANN4, which is transiently expressed in Xenopus (Xenopus laevis) oocytes, could also be embedded in the plasma membrane to caused calcium signaling in response to salt stress (Ma et al., 2019). Most previous studies have used membrane clamp techniques to investigate ion flow in transiently expressed Xenopus oocytes and plant protoplasts. In this study, the transmembrane Ca²⁺ flow in plant roots was examined using the NMT system (Figure 5A). This in vivo assay technique allows the detection of the rate and direction of ion flows in and out of living organisms without damaging roots. We found that AtANN2 and AtANN3 significantly affected Ca²⁺ flow across the plasma membrane, and these functions manifested only under specific photoperiodic conditions. Specifically, the function of AtANN2 was manifested under short-day condition, while the function of AtANN3 manifested under long-day condition (Figure 5). In addition to annexins, calcium signaling induced by stress signals involved multiple types of ion channels and signal transduction systems that together regulated local calcium signaling or long-distance calcium signaling. For example, AtCNGC19 could be rapidly activated as a Ca²⁺ channel in wounding stress, caused Ca²⁺ influx (Meena et al., 2019).

We found that the function of AtANN2 was manifested under short-day condition, while the function of AtANN3 manifested under long-day condition. How did annexins sense photoperiods? Photoreceptors sensed photoperiod and thus regulated plant growth, development, and stress responses (Yoshitake et al., 2015; Fantini and Facella, 2020), e.g., PHYs have been found to be involved in photoperiod sensing in short-day plant flowering. In long-day plants, both PHYs and CRYs were involved in photoperiod perception (Su et al., 2017; Fantini et al., 2019). Therefore, annexins were speculated to also sense photoperiods through photoreceptors. We found that AtCRY2 interacted with AtANN2 and AtANN3 (Figure 6).

On the other hand, CRYs were also involved in biotic and abiotic stresses, such as drought stress, salt stress, heat stress, low temperature stress, radiation, UV-B light, and pathogen attacks (Mao et al., 2005; Ahmad, 2016; Zhou et al., 2017, 2018). Most recent studies only focused on phenotypes, such as CRYs increased the sensitivity of plants to drought stress (D’Amico-Damiao and Carvalho, 2018; Aslam et al., 2020). Overexpression of CRYs in wheat and barley (Hordeum vulgare) resulted in worse growth status in Arabidopsis (Mao et al., 2005; Xu et al., 2009; Zhou et al., 2018). Knowledge regarding the mechanism is limited. We found that after AtCRY2 overexpression, the drought tolerance functions of AtANN2 and AtANN3 under a specific photoperiod were inhibited (Figure 7), their subcellular localization no longer appeared at the plasma membrane after 5 min of mannitol treatment (Figure 9), and mannitol activation of Ca²⁺ efflux was diminished (Figure 10). These results suggest that AtCRY2 is proposed to repress functions of AtANN2 and AtANN3 by affecting their plasma membrane localization to inhibited AtANN2 and AtANN3-dependent transmembrane Ca²⁺ flow in response to drought stress.

Taken together, these results demonstrated the existence of a “photoperiod - AtCRY2 -annexins (AtANN2 and AtANN3) - transmembrane Ca²⁺ flow - drought tolerance function” signaling pathway in Arabidopsis. This finding highlights a novel mechanism regulating the drought stress response by light signal transduction. Further research is needed to determine the role of the biological clock in regulating this signaling pathway and their downstream genes involved in drought tolerance.