ZEITLUPE Promotes ABA-Induced Stomatal Closure in Arabidopsis and Populus

Abstract

Plants balance water availability with gas exchange and photosynthesis by controlling stomatal aperture. This control is regulated in part by the circadian clock, but it remains unclear how signalling pathways of daily rhythms are integrated into stress responses. The serine/threonine protein kinase OPEN STOMATA 1 (OST1) contributes to the regulation of stomatal closure via activation of S-type anion channels. OST1 also mediates gene regulation in response to ABA/drought stress. We show that ZEITLUPE (ZTL), a blue light photoreceptor and clock component, also regulates ABA-induced stomatal closure in Arabidopsis thaliana, establishing a link between clock and ABA-signalling pathways. ZTL sustains expression of OST1 and ABA-signalling genes. Stomatal closure in response to ABA is reduced in ztl mutants, which maintain wider stomatal apertures and show higher rates of gas exchange and water loss than wild-type plants. Detached rosette leaf assays revealed a stronger water loss phenotype in ztl-3, ost1-3 double mutants, indicating that ZTL and OST1 contributed synergistically to the control of stomatal aperture. Experimental studies of Populus sp., revealed that ZTL regulated the circadian clock and stomata, indicating ZTL function was similar in these trees and Arabidopsis. PSEUDO-RESPONSE REGULATOR 5 (PRR5), a known target of ZTL, affects ABA-induced responses, including stomatal regulation. Like ZTL, PRR5 interacted physically with OST1 and contributed to the integration of ABA responses with circadian clock signalling. This suggests a novel mechanism whereby the PRR proteins-which are expressed from dawn to dusk-interact with OST1 to mediate ABA-dependent plant responses to reduce water loss in time of stress.

Keywords: OPEN STOMATA 1; PSEUDO-RESPONSE REGULATORs; ZEITLUPE; abiotic stress; abscisic acid; circadian clock; stomatal closure.

Figure 1

Stomatal opening and water loss phenotypes of detached Arabidopsis leaves. (A–C) Stomatal aperture in epidermal strips from leaves of 4-week-old seedlings treated with or without 10 μM ABA measured at ZT 8–9 under CO₂-free aeration. (A) ztl-1 and WT (C24) plants; (B) ztl-3 and WT (Col-0); and (C) ztl-21 and WT (Ws-2). (D) Stomatal conductance (gs) in leaves of 4-week-old Arabidopsis plants. Conductance was normalised against the WT means. (E,F) Rates of water loss in detached rosette leaves from 3-week-old Arabidopsis plants. Values are means ± SE of three biological replicates, each containing one leaf from six to eight plants of each genotype. (E) Rates of water loss in WT (Col-0); ztl-3 and ost1-3. Asterisks in (E) refer to both the comparison between WT and ztl-3 and between WT and ost1-3 (Student’s t-test); both mutants showed the same level of statistical difference from WT. (F) Rates of water loss in WT (Col-0); ztl-3 and ztl-3:p35S::HA::ZTL (oex1-3: independent lines of ztl-3 overexpressing ZTL). Asterisks in (F) represent the comparison between WT and ztl-3 plants (Student’s t-test); water loss from all other mutants did not differ statistically from that of WT plants. WT and ztl-3 plants (Student’s t-test); water loss from all other mutants did not differ statistically from that of WT plants. (A–F) ^*p < 0.05 and ^***p < 0.001 (Student’s t-test).

Figure 2

Delayed fluorescence (DF) in Populus leaves from WT (T89) trees and PttZTL 1,2 RNAi lines assayed under continuous light. (A) Best fitted Y traces of DF rhythms in leaves in constant conditions following entrainment in LD 18:6 cycles. (B) Values of circadian period and relative amplitude errors (RAE) of individual leaves included in the experiment shown in (A). RAE of rhythmic plants ≤0.6; period estimates are shown in Table 1.

Figure 3

Effect of ZTL expression on stomatal phenotype in Populus. (A) Relative expression levels of PttZTL at ZT 8 in two PttZTL 1,3 RNAi lines of Populus. Expression levels of PttZTL were normalised against expression of EF1a and the ratio was set at 1 in WT (T89) trees. Values are means ± SE of three or four biological replicates. (B) Effect of ABA treatment on stomatal ratios (width:length) in epidermal strips from wild-type and PttZTL 1,2 RNAi lines measured at ZT 8–9 under CO₂-free aeration. Data are means ± SE of three biological replicates, each containing 20 to 22 stomata. (C) Stomatal conductance (gs) in leaves from intact 6 to 9-week-old WT and PttZTL 1,2 RNAi trees. Values are means ± SE of three biological replicates, each containing six to eight plants with three leaves/genotype. Conductance was normalised against the WT means. (D) Rates of water loss from detached leaves from 13-week-old PttZTL RNAi and wild-type trees. Values are means ± SE of three leaves per biological replicates, from six plants per genotype. In (D), both RNAi lines differed from WT all time points: *** thus represents both the RNAi-5 vs. WT and RNAi-7 vs. WT comparisons at time points where RNAi lines differed from WT at the same level of significance. At 180 min, WT vs. RNAi line 5 differed from WT at p < 0.001 and RNAi line 7 differed from WT at p < 0.05; this is represented by ***/* on the Figure. (A–D) Differences between WT compared with both mutants are indicated where statistically significant at ^*p < 0.05; ^**p < 0.01; and ^***p < 0.001 (Student’s t-test).

Figure 4

Effects of genotype and ABA on expression of early and late ABA-signalling components and ABA-responsive genes in Arabidopsis. (A) ABA reception gene PYL5. (B–G) Early and progressing ABA-signalling genes. (B) ABI2. (C) HAB1. (D) OST1. (E) ABI5. (F) ABF3. (G) ABF4. (H,I) Late responsive ABA-signalling genes. (H) RD29A. (I) RAB18. Gene expression was measured in samples collected at ZT 8 from WT (Col-0) and ztl-3 plants treated with ABA or ethanol (control). Values are the means ± SE of pooled seedlings from three biological replicates, each containing two technical replicates. Expression levels were normalised against expression of EF1a and the ratio was set at 1 in untreated WT (Col-0). Results were analysed by two-way ANOVA to determine the effects of treatment (T), genotype (G) and the T × G interaction. Significance levels of the effects were determined using Sidak’s multiple comparisons post-hoc test; ^*p < 0.05; ^**p < 0.01; and ^***p < 0.001.

Figure 5

Coexpression of ZTL and OST1 in plant cells indicates a strong interaction between the proteins. (A) Representative co-immunoprecipitation assay showing in vivo interaction between ZTL and OST1. Tagged versions of ZTL and OST1 proteins were expressed in Arabidopsis protoplasts, singly or in combination. Proteins were immunoprecipitated using mouse anti-Myc antibody (IP: α-Myc) and subsequently analysed by Western blotting with an anti-HA-POD antibody (WB: α-HA) and anti-c-Myc chicken antibody (WB: α-Myc). The experiment was repeated three times in different Western blots and produced similar results. (B) Mean values (n = 3) ± SEM of the ratio of Co-IP signal to sample signal from the input reaction (40% of sample used in Co-IP) of the three different co-immunoprecipitation assays described in (A). The individual results from each experiment are shown on the plot: Filled circles (left-hand side): Myc-OST1 + HA-ZTL; filled squares (right-hand side): HA-ZTL.

Figure 6

ZTL and OST1 act synergistically to promote stomatal closure, and PRR5 partially blocks ABA signalling in the absence of both ZTL and OST1. (A, B) Water loss from detached rosette leaves of 3-week-old WT (Col-0) Arabidopsis and single, double and triple mutants carrying different combinations of alleles at the ZTL (ztl-3), PRR5 (prr5-1) and OST1 (ost1-3) loci. Values are means ± SE of two to three biological replicates, each containing one leaf from six to eight plants of each genotype. In (A), significant differences between the ztl-3 and ost1-3 double mutant and the ztl-3, prr5-1 and ost1-3 triple mutant are shown by cyan-coloured asterisks. As all the mutants differed from WT at p < 0.001 (Student’s t-test) at all time points, all the pairwise mutant vs. WT comparisons represented by three black asterisks for simplicity. In (B), pairwise comparisons found that ztl-3 differed significantly from WT but prr5-1 and the prr5-1 and ztl-3 double mutant did not. The result for ztl-3 vs WT (Student’s t-test) is represented by three black asterisks. Statistical levels in (A) and (B): ^*p < 0.05; ^**p < 0.01; and ^***p < 0.001.

Figure 7

The circadian clock protein PRR5 interacts with ZTL and OST1 to regulate stomatal responses in Arabidopsis. (A) ztl-3 and prr5-11 mutants show opposite stomatal aperture phenotypes in response to ABA. Time of ABA treatment was adjusted to an equivalent point in the circadian cycle to accommodate the difference in period (τ) between ztl-3 (τ = ~ 28 h), prr5-11 (τ = ~ 23 h) and wild-type (τ = 24 h) plants. Values are means ± SE of three biological replicates, each containing 20 stomata/genotype. (^***p < 0.001; Student’s t-test). (B) PRR5 and OST1 interact in vivo. Co-immunoprecipitation assay showing an interaction between PRR5 and OST1 in Arabidopsis protoplasts. Tagged versions of PRR5 and OST1 proteins were expressed in Arabidopsis protoplasts. Proteins were immunoprecipitated using mouse anti-Myc antibody (IP: α-Myc) and subsequently analysed by Western blotting with an anti-HA-POD antibody (WB: α-HA) and anti-c-Myc chicken antibody (WB: α-Myc).