Upstream kinases of plant SnRKs are involved in salt stress tolerance

Abstract

Sucrose non-fermenting 1-related protein kinases (SnRKs) are important for plant growth and stress responses. This family has three clades: SnRK1, SnRK2 and SnRK3. Although plant SnRKs are thought to be activated by upstream kinases, the overall mechanism remains obscure. Geminivirus Rep-Interacting Kinase (GRIK)1 and GRIK2 phosphorylate SnRK1s, which are involved in sugar/energy sensing, and the grik1-1 grik2-1 double mutant shows growth retardation under regular growth conditions. In this study, we established another Arabidopsis mutant line harbouring a different allele of gene GRIK1 (grik1-2 grik2-1) that grows similarly to the wild-type, enabling us to evaluate the function of GRIKs under stress conditions. In the grik1-2 grik2-1 double mutant, phosphorylation of SnRK1.1 was reduced, but not eliminated, suggesting that the grik1-2 mutation is a weak allele. In addition to high sensitivity to glucose, the grik1-2 grik2-1 mutant was sensitive to high salt, indicating that GRIKs are also involved in salinity signalling pathways. Salt Overly Sensitive (SOS)2, a member of the SnRK3 subfamily, is a critical mediator of the response to salinity. GRIK1 phosphorylated SOS2 in vitro, resulting in elevated kinase activity of SOS2. The salt tolerance of sos2 was restored to normal levels by wild-type SOS2, but not by a mutated form of SOS2 lacking the T168 residue phosphorylated by GRIK1. Activation of SOS2 by GRIK1 was also demonstrated in a reconstituted system in yeast. Our results indicate that GRIKs phosphorylate and activate SnRK1 and other members of the SnRK3 family, and that they play important roles in multiple signalling pathways in vivo.

Keywords: Arabidopsis thaliana; GRIKs; SOS2; SnRKs; phosphorylation; salinity; stress; sugar; upstream kinases.

Figure 1

Characterization of the grik1‐2 grik2‐1 mutant line. (a) Schematic diagram of grik1 T‐DNA insertion lines. (b) Reverse transcriptase‐polymerase chain reaction (RT‐PCR) of full‐length GRIK1, GRIK2 and tubulin8 in the wild‐type Col‐0, grik1‐2, grik2‐1 and grik1‐2 grik2‐1. (c) Phenotype of the wild‐type and grik1‐2 grik2‐1 under short‐day condition for 4 weeks. (d) Northern blotting with the middle region of GRIK1 as probe in the wild‐type, grik1‐1 and grik1‐2. (e) cDNA sequence of the 5′‐end region of grik1‐2 and potential ORFs. The longest ORF corresponded to an N‐terminal truncated GRIK1 protein starting at Met‐156 of the wild‐type protein. (f) Amount and phosphorylation status of SnRK1s in the wild‐type and grik1‐2 grik2‐1. Western blot with anti‐pAMPK antibody (upper panel) and anti‐SnRK1.1 antibody (middle panel). Coomassie staining was used to show protein amounts (lower panel). A greater amount of proteins from grik1‐2 grik2‐1 (150%) was also loaded to compare the phosphorylation ratio in equivalent amounts of SnRK1.1.

Figure 2

Salt‐sensitive phenotype of grik1‐2 grik2‐1 double mutant. (a) Seedlings of the wild‐type (Col‐0), grik1‐2, grik2‐1, grik1‐2 grik2‐1 and sos2‐2 mutants grown on Murashige and Skoog (MS) agar plate were transferred to fresh plates with or without 100 mm NaCl. Photographs were taken 12 days after transfer. The line graph represents primary root length of seedlings 12 days after transfer to MS agar plates with the indicated concentration of NaCl (mean ± SE, n = 20). Asterisks indicate significant differences from wild‐type (P < 0.05, in one‐way anova followed by Tukey's multiple comparison test). (b) Hydroponic culture of Col‐0, grik1‐2, grik2‐1, grik1‐2 grik2‐1 and sos2‐2 mutants in LAK medium supplemented with NaCl as indicated. Plants were grown for 4 weeks. The bar graph represents the dry weight of plants (mean ± SE, n = 7) at the end of the experiment. Asterisks indicate significant differences from wild‐type (P < 0.01, in one‐way anova followed by Tukey's multiple comparison test).

Figure 3

In vitro phosphorylation of SOS2 by GRIK1. (a) A kinase‐dead SOS2 fused to GST (SOS2‐KN, arrow) was incubated with GST‐fused GRIK1 or the kinase‐dead (K137R) of GRIK1 (GRIK1‐KR) also fused to GST (arrowhead). Proteins were separated in sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) followed by Coomassie staining (left) and autoradiograph (right). (b) GST‐fused mutant T168A of SOS2 was incubated with GST‐fused GRIK1. Proteins were separated in SDS–PAGE followed by Coomassie staining (left) and autoradiograph (right).

Figure 4

Expression of SOS2 mutants S159A (SA) and Y175A (YA), but not T168A (TA), partially rescued the sos2‐2 under salt stress condition. (a) Seedlings grown on Murashige and Skoog (MS) agar medium were transferred to fresh plates with or without 100 mm NaCl. Photographs were taken 12 days after transfer. The lines presenting the best resistance to the NaCl plates were selected for this test. (b) Northern blot for SOS2 transcript was performed with total RNA extracted from wild‐type (WT), sos2‐2 transgenic plants expressing the S159A (SA), T168A (TA) and Y175A (YA) mutant forms of SOS2, and non‐transformed sos2‐2. Total RNA was purified 12 h after transfer to MS agar plates with 100 mm NaCl. rRNA (ethidium bromide stained) was used as a loading control. (c) Survival rates of the seedlings on MS agar plates with or without 100 mm NaCl. The experiment was repeated three times, and mean values (± SE) are shown. (d) Primary root length of seedlings 12 days after transfer to MS agar plates with or without 100 mm NaCl (mean ± SE). Asterisks indicate significant differences from sos2‐2 transformed with the wild‐type SOS2 (P < 0.05, in one‐way anova followed by Tukey's multiple comparison test).

Figure 5

In vitro activation of SOS2‐∆308 by GRIK1. The GST‐fused SOS3‐independent form of SOS2 (SOS2‐∆308, arrowhead), or the combined T168A mutation in SOS2‐∆308 were incubated with GST‐fused GRIK1 or GST‐fused kinase‐dead (K137R) version of GRIK1 (GRIK1‐KR). The SOS1 C‐terminal fragment (SOS1‐CT, arrows) was added as a phosphorylation substrate of SOS2. Proteins were separated in sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) followed by Coomassie staining (left) and autoradiograph (right). Note that the small amount of GRIK1 used was enough to achieve the activation of SOS2, even though GRIK1 bands were not observed. The relative band intensity of SOS1‐CT bands was normalized to the signal from the SOS2‐∆308WT/GRIK1‐KR lane (mean ± SE, n = 6). Asterisks indicate significant differences (P < 0.05, binomial test).

Figure 6

SOS2‐GRIK1 interaction visualized by bimolecular fluorescence complementation (BiFC). Panels from (a) to (f) show several images of reconstituted BiFC by GRIK1‐SOS2 interaction (left: fluorescence images; right: overlay with transmitted light). Arrows indicate the presence of a transvacuolar strand in (c) and (d), and nuclei in (e) and (f). Panels (g) and (h) show a negative control with GRIK1 in pSPYCE(M) and the empty vector pSPYNE(R)173. Panels (i) and (j) show co‐transformation with SOS2 in pSPYNE(R)173 and the empty vector pSPYCE(M). Scale bar: 10 μm.

Figure 7

Activation of SOS2 by GRIK1 in yeast. (a) The constitutively active form of SOS2 lacking the C‐terminal autoinhibitory domain and with either the phosphomimic T168D mutation in the activation‐loop (SOS2‐T/D∆308) or with the three putative phosphorylation sites by GRIK kinases (S156/T168/Y175) converted to alanine residues (SOS2‐AAA∆308) were expressed in the yeast strain YP890 bearing mutations ena1‐4 nha1 nhx1 and expressing SOS1 from a chromosomal integration. The salt sensitivity of all transformants was analysed by spotting decimal dilutions of starting cultures in AP plates supplemented with the indicated amounts of NaCl. Growth in NaCl‐supplemented media reported the activation of SOS1 by SOS2. (b) The full‐length SOS2 protein and the triple mutant S156A/T168A/Y175A (SOS2‐AAA) were expressed in the yeast strain YP890 harbouring a chromosomal integration of SOS1. When indicated, SOS3 was also co‐expressed. The salt tolerance of the transformants was analysed in AP medium with increasing concentration of NaCl as described above. Failure to convey salt tolerance indicated that SOS2AAA is unable to form a productive complex with SOS3 to activate SOS1. (c) SOS1, SOS2 and SOS3 were expressed, in various combinations as indicated, in strain Δ3K4E, which lacks the three SNF1‐activating kinases SAK1, ELM1 and TOS3 in addition to the Na⁺ pumps ENA1‐4. When indicated, GRIK1 was also co‐expressed to test for the ability to complement the sak1 elm1 tos3 mutations with regard to SOS2 activation. Results indicated that only the full complement of GRIK1, SOS2, SOS3 and SOS1 was able to restore salt tolerance.

Figure 8

Model of the placement of GRIK kinases at the interface between sensing of the energy status and salinity stress signalling. Salt stress induces the activation of SOS2 through calcium/CBL (SOS3 and SCaBP8) binding. Sugar/energy status is transmitted through SOS2 phosphorylation by GRIK1/2, which also controls reciprocally the activity of SnRK1s.