The study started from a comparative analysis of salt stress-induced transcriptional responses in the salt-sensitive rice variety Oryza sativa (ssp. indica) line IR29 and in the salt tolerant grass F. rubra ssp. litoralis. Genes were identified that differentially respond to salinity in both species. F. rubra ssp. litoralis is characterized by substantial salt resistance, tolerates up to 500 mM NaCl and continues growth and development with 250 mM NaCl in hydroponic culture (not shown). In contrast, the rice line IR29 is severely damaged by exposure to salt concentrations of 150 mM NaCl 1617. The serine-threonine protein kinase SAPK4 was identified in a subtracted cDNA library from F. rubra ssp. litoralis enriched for salt-responsive genes. This experimental approach allowed to identify an EST-sequence from F. rubra ssp. litoralis that shared 89 and 93% identity on the nucleic acid and amino acid level with SAPK4 from rice (not shown). The present study aimed at a detailed analysis of the role of the kinase in plant salt acclimation. By semiquantitative RT-PCR (Fig. 1) and Northern-type RNA hybridizations (not shown), expression of SAPK4 was detected in non-stressed control plants of F. rubra and rice. In F. rubra, salt stress of 125 mM reduced and of 250 mM and 500 mM NaCl increased the transcript level of SAPK4. In the salt-sensitive rice line IR29, treatment with 125 mM NaCl for up to 48 h caused a decrease of SAPK4 transcript abundance. Due to lethality, 250 mM and 500 mM NaCl were not applied to rice.

For an analysis of the subcellular localization of the SAPK4 protein, constructs for the expression of the SAPK4 open reading frame cDNA fused to the green fluorescent protein (GFP) reporter gene driven by the 35S-CaMV promoter were generated. Onion epidermis cells were transformed with the translational fusion and fluorescence emission of GFP was monitored under a confocal laser scanning microscope (Fig. 2). In cells incubated for 24 hours in 0.5 × MS nutrient medium, strong GFP signals were detected in the nucleus (Fig. 2A, B). Cells bombarded with the empty vector as a negative control showed no fluorescence (Fig. 2D) As a positive control, onion epidermal cells were transformed with a translational construct of GFP. In this experiment, GFP showed localization throughout the cell with strongest signals in cytoplasm and nucleus (Fig. 2C). To extent the results obtained from the onion epidermal cells system, A. thaliana mesophyll protoplasts were isolated and were similarly transformed with the SAPK4-GFP transcriptional constructs. After incubation of transformed protoplasts for 24 hours, GFP-derived fluorescence emission was also detected in the cytoplasmic compartment (Fig. 2E).

To generate transgenic rice plants, the salt-sensitive variety IR29 was transformed with vectors containing the open reading frame rice SAPK4 cDNA for transcriptional over-expression under control of the 35S-CaMV promoter. Three independent transgenic rice lines designated S1, S4, and S5 were identified by kanamycin resistance and by the presence of the kanamycin resistance gene. Plants of the T2 generations were used for further investigations to examine the role of SAPK4 in plant salt tolerance. In non-stressed plants, a moderate increase in the transcript level of SAPK4 could be detected in the transgenic lines compared to wild-type rice indicating tight regulation of SAPK4 mRNA in rice (Fig. 3). No significant difference between wild-type and transgenic lines was seen in the phenotypes, growth rate, and development up to the age of 12 weeks (not shown). Transgenic plants exposed to elevated NaCl concentrations accumulated more SAPK4 transcript compared with wild-type rice (Fig. 3A) and the lines were analyzed for their salt stress responses. In two independent rice lines that were transformed with the empty plant expression vector the transcript amounts of SAPK4 were not changed in comparison to non-transformed wild-type rice (Fig. 3) and the phenotype of these plants was not changed as well (not shown). These data demonstrate that the observed effects of SAPK4 over-expression in transgenic rice are a result of SAPK4 and not of the empty plant expression vector.

Wild-type rice and the transgenic lines were grown in hydroponic culture to the age of 3 weeks under control conditions and subsequently were exposed to 150 mM NaCl. At 7 days of salt treatment the transgenic lines displayed improved salt tolerance (Fig. 3B, C). Salt-treated wild-type rice showed growth inhibition and developed chlorosis and necrosis (Fig. 3). In contrast, growth of the transgenic lines was rather unaffected and chlorosis was not apparent (Fig. 3B). To test developmental dependence of enhanced tolerance to increased NaCl concentrations by SAPK4, germination assays were performed with wild-type rice and T2 S1 and S4 seedlings. Control and transgenic plants were germinated on control nutrition medium and on medium supplemented with NaCl. Results showed that the germination decreased by approximately 20% in wild-type rice by salt treatment whereas germination was not significantly affected in the transgenic lines (Fig. 4A). In addition, leaf growth was inhibited by the salt stress in wild-type control plants whereas in the transgenic lines the leaf size of seedlings was not significantly changed compared with non-stressed control seedlings (Fig. 4B).

Accumulation of Cl^- was examined in wild-type rice and in the transgenic lines to test whether SAPK4 affects accumulation of Cl^- under salt stress. Rice plants were grown to the age of 3 weeks and treated with 150 mM NaCl for 48 hours. Under these stress conditions, the lines S1 and S4 contained only 60% of the Cl^- contents of wild-type rice (Fig. 5A). Cation homeostasis was addressed by comparing the contents of Na⁺, K⁺, and Ca²⁺ in wild-type rice and plants of the line S4 that were exposed to the same stress regime as described above. The line S4 accumulated 60% of Na⁺ and 80% K⁺ in comparison to wild-type plants whereas no differences in Ca²⁺ accumulation were observed (Fig. 5B). As a physiological reference chlorophyll a fluorescence kinetics were measured and photosynthetic yield calculated. Exposure to 150 mM NaCl for 48 hours resulted in a decreased photosynthetic activity in wild-type rice whereas no significant change occurred in the lines S1 and S4 in comparison with non-stressed control plants of the same lines (Fig. 5C).

The vacuolar ATPase, the vacuolar Na⁺/H⁺ antiporter NHX1, voltage-gated Cl^- channels, and catalase are well established targets of salt-dependent regulation (Fig. 6A) and it appeared interesting to study their transcription in wild-type and transgenic rice under salt stress to identify putative target genes regulated by SAPK4. The expression levels of the vacuolar ATPase, the vacuolar Na⁺/H⁺ antiporter OsNHX1, the Cl^- channel OsCLC1, and catalase isozyme A were assessed in wild-type and transgenic rice. The transcript amounts of the V-ATPase subunit B and of the catalase increased by treatment with 100 mM NaCl for 48 hours, whereas the transcript amounts of OsNHX1 and OsCLC1 decreased in response to NaCl stress (Fig. 6B). In addition, we were interested in analyzing transcription of the plasma membrane Na⁺/H⁺ antiporter SOS1 but were not able to detect its expression in both wild-type and transgenic rice.

Rice (Oryza sativa L. indica) var. IR29 and Festuca rubra ssp.litoralis were grown in a growth chamber with 14 h light (300 μE m^-2 sec^-1, 25°C) and 10 h dark (21°C) and 50% relative humidity. Seeds were germinated in vermiculite soaked with a modified half-strength Hoagland's nutrition solution 55. Seedlings were transferred to aerated hydroponic tanks 10 days after germination. For salt stress, the nutrition medium was supplemented with NaCl at a final concentration of 125 and 500 mM. Non-stressed control plants were grown in parallel and harvested at the same time. For transcript analyses, wild-type rice plants were grown to the age of 3 weeks. Experiments with F. rubra ssp.litoralis were performed at a comparable growth and developmental stage at the age of six weeks. Wild-type and transgenic rice lines were grown to the age of 8 weeks for growth and stress experiments. For germination analyses, seeds of wild-type and of transgenic rice lines were germinated in Petri dishes on sterile filter paper soaked with half-strength Hoagland's nutrition solution. The statistical significance of different germination rates was determined by Student's t-test (p < 0.05). Different NaCl concentrations were chosen for reasons of plant age. Rice plants of the age of 8 weeks were stressed with 125 and 150 mM NaCl that is a severe stress for rice 16. Germination and growth of seedlings were monitored at 50 mM NaCl. For studies of the transcription of genes regulated by SAPK4 the moderate salt stress of 100 mM NaCl was applied.

Total leaf RNA from rice and of F. rubra ssp. litoralis was extracted as described 55. Northern hybridizations were performed with 20 μg of total RNA per lane 55. cDNA was synthesized from 5 μg of total RNA with M-MLV RT II [H-] (Promega) and oligo-dT-priming in 20 μl reactions. cDNA probes for Northern detection were generated by PCR with cDNA synthesized from leaves of rice with gene-specific sense and antisense oligonucleotide primers and digoxigenin-dUTP (Roche, Germany) as a label. Transcript analyses were performed for the following genes: SAPK4 (AB125305, LOC_Os01g64970; primers for cloning the cDNA: 5'-CACCATGGAGAAGTACGAGGCG-3', 5'-TCATATGCGCAGTGAGCTCAT-3', primers for analyses of transcription: 5'-TGGCTACTCCAAGTCATC-3', 5'-TCGTACTCATCTTCCTCC-3'), OsNHX1 (LOC_Os07g47100; primer sequences: 5'-ATCTTCAATGCAGGCTTC-3' and 5'-TGCATCCATCCCAACATA-3'), OsVHA-B (LOC_Os06g37180; primer sequences: 5'-ATTGACAGGCAGCTGCAT-3' and 5'-GCAATGTCCATGCTAGGT-3'), OsCLC1 (LOC_Os01g65500; primer sequences: 5'-TGTACAAGCAGGACTGGA-3' and 5'-AGATAGGCCTTCACCTCA-3', and catalase isozyme A (LOC_Os02g02400; primer sequences: 5'-GGATGACACCAAGACATG-3', 5'-TCACGTTGAGCCTATTCG-3'). Actin was amplified as a loading control (primer sequences: 5'-GTGATCTCCTTGCTCATACG-3' and 5'-GGNACTGGAATGGTNAAGG-3'). Probes for array hybridization were prepared from each 25 μg of total RNA by incorporating digoxigenin-11-dUTP. Northern blot and cDNA-array membranes were washed with 0.5× SSC at 42°C for 30 minutes and hybridization signals were detected with anti-digoxigenin alkaline phosphatase conjugated Fab fragments and CSPD (Roche, Germany) as a substrate. RT-PCR analyses were performed in standard reactions as described 55. Actin was hybridized and amplified as a loading control. For densitometric analyses the Gelscan software (INTAS, Germany) was used.

mRNA was isolated from total RNA with the PolyATract kit (Promega, Mannheim, Germany). A subtraction cDNA-library of F. rubra ssp. litoralis was synthesized with the PCR-Select Kit (Clontech, Heidelberg, Germany) according to the manufacturer's protocol. Same amounts of mRNA from the salt stress treatments of F. rubra ssp. litoralis were pooled for the tester cDNA: 125 mM NaCl, 250 mM NaCl, and 500 mM NaCl for 6 h, 24 h, 48 h, and 7 days at the ages of 5 and 12 weeks, each leaf and root tissue. Same amounts of mRNA from control plants of the same developmental stage and harvested in parallel to the stressed plants were pooled for the driver cDNA. The subtracted cDNA was cloned into pCR-TOPO II (Invitrogen, Karlsruhe, Germany) and the inserts were amplified by PCR with the nested primers 1 and 2R (Clontech). The PCR products were analyzed on agarose gels and products that yielded single bands were selected for further procedures. BLAST analyses of sequenced PCR products (MWG Biotech, SeqLab, Germany) were performed in the rice TIGR database.

The PCR products from the subtraction cDNA-library were purified with QIAquick spin columns (Qiagen, Hilden, Germany) and dissolved in 50% (v/v) DMSO. cDNA-macroarrays with 129 functionally different ESTs were generated at the Center of Genome Research (ZfG) at the University of Bielefeld. The amplified PCR products were transferred in duplicates to nylon membranes (Hybond-N, Amersham Pharmacia Biotech, UK) in a 3 × 3 pattern leaving an empty area in the middle for local background subtraction. The membranes were fixed by baking and UV cross-linking. Labeled detection probes were prepared from each 25 μg RNA with incorporation of digoxigenin-11-dUTP (Roche, Mannheim, Germany) during first strand cDNA synthesis using SuperScript II (Invitrogen) and oligo(dT)-priming. The probes were purified using QIAquick spin columns (Qiagen) to remove unincorporated nucleotides. The probes were denatured at 100°C for 10 min prior to hybridization. The hybridization was performed with the same procedure as described above for Northern hybridizations.

The open reading frame cDNA of SAPK4 was amplified by RT-PCR from rice leaves (var. IR29) with sense and antisense oligonucleotide primers and cloned in pENTR/D-TOPO (Invitrogen, Netherlands). SAPK4 cDNA was subcloned in pK2GW7 for over-expression under the control of the constitutive 35S-CaMV promoter and in pK7FWG2 for over-expression with a C-terminal GFP-fusion under the control of the constitutive 35S-CaMV promoter 56. Agrobacterium tumefaciens strain LBA4404 was transformed with the constructs and used for transformation of rice. Mature dehusked rice seeds were sterilized in ethanol (94%, v/v) for 30 seconds, in formaldehyde (0.8%, v/v) for 40 min, and in sodium hypochlorite (1.8%, v/v). The seeds were incubated in Petri dishes containing two layers of sterile Whatman paper and 7 ml sterile deionized water at 42°C for 24 hours to allow pre-germination. The seeds were immersed for 10–15 min in liquid co-culture medium (CCL, modified Chu N6 medium with 2 mg/l naphthalene acetic acid, 1 mg/l kinetin, 55 mM glucose, 100 μM acetosyringone, pH 5,2) containing A. tumefaciens. Subsequently, the seeds were transferred into Petri dishes containing solid co-culture medium (CCL with 0.7% (w/v) agarose) and incubated for three days at 25°C in the dark. The seeds were transferred to CCS plates containing cefotaxime (400 mg/l) and were incubated at 25°C under 12/12 h (day/night) photoperiod for three days. To promote tiller and root development, the rice plants were incubated at a photoperiod of 12/12 h (day/night) in N6 medium without hormones plus kanamycin (50 mg/l). Obtained seeds were sterilized as described above and germinated under kanamycin (50 mg/l) treatment. Experiments were performed with transgenic T2 rice plants. Seeds from transformed rice were screened on 0.5 × MS agar medium containing 50 mg/l kanamycin and by RT-PCR of the kanamycin selection marker. Transgenic rice lines transformed with the empty expression vector served as a control.

Protoplasts were isolated from A. thaliana and were transformed with constructs for expression as described 57. Transformation of onion epidermal cells was performed according to 57. For subcellular localization of SAPK4-GFP fusion proteins a confocal laser scanning microscope (Leica DMRE) was used 57.

The photosynthetic quantum yield of photosystem II (ΦPSII) was calculated from chlorophyll a fluorescence data obtained from attached rice leaves using the Mini PAM (Walz, Germany). The actinic photon flux density was 300 μE m^-2 sec^-1 as during growth. Steady state maximum fluorescence emission was excited by application of a saturating light pulse (5000 μmol m^-2 s^-1). ΦPSII was calculated according to the manufacturer's instructions. The Cl^- content was measured in leaf extracts with a micro-chloro-counter (Marius-Utrecht, Netherlands).