Gain-of-function of the cytokinin response activator ARR1 increases heat shock tolerance in Arabidopsis thaliana

Abstract

In addition to its well-established role in plant development, the hormone cytokinin regulates plant responses to biotic and abiotic stresses. It was previously shown that cytokinin signaling acts negatively upon drought and osmotic stress tolerance and that gain-of-function of the cytokinin response regulator ARR1 causes osmotic stress hypersensitivity. Here we show that increased ARR1 action increases tolerance to heat shock and that this is correlated with increased accumulation of the heat shock proteins Hsp17.6 and Hsp70. These results show that the heat shock tolerance of plants can be elevated by increasing the expression of a cytokinin response activator.

Keywords: Arabidopsis; abiotic stress; cytokinin; heat shock; heat shock proteins.

Figure 1.

Heat-shock response of ARR1 gain- and loss-of-function lines. (a) Leaf discs excised from mature leaves of 30-day-old plants were floated in MES-KOH buffer (pH 6.8) for 2.5 h at 22°C (control) or 45°C (HS, heat shock) and then incubated in the same buffer at 22°C for 2 days before photography. Representative leaf discs of the Col-0 wild type (WT), arr1-1 single and double mutant, and two ARR1 gain-of-function lines are shown. Scale bar: 5 mm. (b) Chlorophyll levels in heat-shocked leaf discs. The chlorophyll content is presented as mean ± SD (n ≥ 3). *, P < .05; ****, P < .0001; ns, not significant (one-way ANOVA with Tukey’s multiple comparisons test). The insert shows the chlorophyll levels in the wild-type leaf discs compared to leaf discs of the arr1-1 and arr1-3 arr10-5 mutant. (c) Basal seed thermotolerance tests. Sterilized and stratified seeds were exposed to light for 4 h at 22°C (control) or at 45°C (heat shock) and then germinated and grown in darkness at 22°C. Etiolated seedlings were photographed after 3 days of growth, and hypocotyl lengths were measured using ImageJ. The data are presented as the percentage inhibition of hypocotyl length in response to heat shock (mean ± SD; n ≥ 20). **, P < .01; ****, P < .0001 (one-way ANOVA with Tukey’s multiple comparisons test).

Figure 2.

Cytokinin affects Hsp17.6 levels in ARR1 gain- and loss-of-function lines. (a) Representative immunoblot probed with anti-Hsp17.6 antibodies. Seven-day-old seedlings were kept for 2 h at 22°C or at 40°C. Ponceau S-stained membrane region encompassing the large subunit of RuBisCO (LSU) is shown as the loading control. (b) Quantification of the Hsp17.6 levels. Signal intensities of the control samples and the heat-shocked samples were quantified separately as they required quantification at images obtained after different exposure times. Data was normalized to the wild type and shown as mean± SD; n = 3 (**, P < .01; ****, P < .0001; one-way ANOVA with Tukey’s multiple comparisons test). (c) Relative expression levels of the Hsp20 gene family after 3-hour-long treatment with cytokinin trans-zeatin (t-z). Microarray data were extracted from publicly available series GSE5698 (AtGenExpress: Cytokinin treatment of seedlings,). Values are presented relative to the control. The significance of expression change between wild type and treatment and the intensity scale are shown on the right. Non-significant differences are not marked and significant changes are marked with an asterisks (*, P < .05; ** < 0.01; ****, P < .0001). Cytokinin-upregulated genes are circled in red.

Figure 3.

Cytokinin affects Hsp70 levels. (a) Expression levels of the Hsp70 gene family after 3-hour-long treatment of 21-day-old plants with cytokinin trans-zeatin (t-z). Values are presented relative to the control. Non-significant differences (ns) are marked. *, P < .05; ****, P < .0001. The AGI code of Hsp70-11 (BIP1) is At5g28540, Hsp70-12 (BIP2) is At5g42020, Hsp70-14 is At1g79930, and Hsp70-15 is At1g79920. (b) Representative immunoblot probed with anti-Hsp70 antibodies. The experimental conditions as the same as in Figure 2. Ponceau S-stained membrane region encompassing the large subunit of RuBisCO (LSU) is shown as the loading control. (c) Quantification of the Hsp70 levels. Data were normalized to the wild type and shown as mean± SD; n = 3 (* < 0.05; two-way ANOVA with Šídák’s multiple comparisons test).