A Universal Stress Protein Involved in Oxidative Stress Is a Phosphorylation Target for Protein Kinase CIPK6

Universal stress protein SlRd2 is a Cipk6 target and regulates Cipk6-mediated ROS

Abstract

Calcineurin B-like interacting protein kinases (CIPKs) decode calcium signals upon interaction with the calcium sensors calcineurin B like proteins into phosphorylation events that result into adaptation to environmental stresses. Few phosphorylation targets of CIPKs are known and therefore the molecular mechanisms underlying their downstream output responses are not fully understood. Tomato (Solanum lycopersicum) Cipk6 regulates immune and susceptible Programmed cell death in immunity transforming Ca²⁺ signals into reactive oxygen species (ROS) signaling. To investigate SlCipk6-induced molecular mechanisms and identify putative substrates, a yeast two-hybrid approach was carried on and a protein was identified that contained a Universal stress protein (Usp) domain present in bacteria, protozoa and plants, which we named “SlRd2”. SlRd2 was an ATP-binding protein that formed homodimers in planta. SlCipk6 and SlRd2 interacted using coimmunoprecipitation and bimolecular fluorescence complementation (BiFC) assays in Nicotiana benthamiana leaves and the complex localized in the cytosol. SlCipk6 phosphorylated SlRd2 in vitro, thus defining, to our knowledge, a novel target for CIPKs. Heterologous SlRd2 overexpression in yeast conferred resistance to highly toxic LiCl, whereas SlRd2 expression in Escherichia coli UspA mutant restored bacterial viability in response to H₂O₂ treatment. Finally, transient expression of SlCipk6 in transgenic N. benthamiana SlRd2 overexpressors resulted in reduced ROS accumulation as compared to wild-type plants. Taken together, our results establish that SlRd2, a tomato UspA, is, to our knowledge, a novel interactor and phosphorylation target of a member of the CIPK family, SlCipk6, and functionally regulates SlCipk6-mediated ROS generation.

Environmental factors, especially those imposing stress, stimulate endogenous cellular cues, which initiate protective responses in plants. Among the concurrent events during stress are changes in the intracellular Ca²⁺ concentration, which activate an overlapping set of downstream responses. Ca²⁺ changes are perceived and decoded by an array of Ca²⁺ sensors including calmodulins or calmodulin-related proteins, Ca²⁺-dependent protein kinases (CDPKs, CPKs), and calcineurin B-like proteins (CBLs; Dodd et al., 2010). Particularly, the CBL family has been shown to play a crucial role in different Ca²⁺-dependent processes in plants (Sanyal et al., 2015). CBL proteins present homology to the regulatory B-subunit of calcineurin and the neuronal calcium sensor proteins from animals and yeast (Luan, 2009). The overall structure of CBLs consists of four EF-hands. Spacing of EF-hands is invariable, while the C- and N-terminal extension of CBL proteins varies in length. Posttranslational modifications of CBLs, including protein phosphorylation and lipid modifications affect their subcellular localization and their stability to interact with other proteins (Sanyal et al., 2015; Nagae et al., 2003). Thus, phosphorylation of the conserved Ser residue in the C-terminal PFPF motif of the CBL proteins enhances the interaction with CBL-interacting protein kinases (CIPKs; Du et al., 2011; Hashimoto et al., 2012).

Upon Ca²⁺ binding, CBLs physically interact with CIPKs, Ser/Thr kinases that structurally belong to Suc nonfermenting 1-related kinases, group 3, also called protein kinases related to SOS2 (PKS; Gong et al., 2004; Yu et al., 2014). CIPKs are constituted of a C-terminal or regulatory domain and a conserved kinase catalytic domain at the N terminus. Within the divergent regulatory domain, CIPKs contain an autoinhibitory NAF/FISL motif and a type 2C protein phosphatase binding site called the “PPI motif”. It is well established that binding of CBLs to the NAF/FISL motif releases the C-terminal (autoinhibitory) domain from the kinase domain, thus leading the kinase into an active state (Guo et al., 2001; Chaves-Sanjuan et al., 2014). In Arabidopsis (Arabidopsis thaliana), there are 10 CBL and 26 CIPK homologs (Yu et al., 2014). By yeast two-hybrid (Y2H) and bimolecular fluorescence complementation (BiFC) assays it has been determined that CBLs show a level of specificity in targeting different CIPKs. On the other hand, a specific CIPK can also interact with different CBLs, thus allowing a single CIPK to access different cellular compartments and hence different substrates (Kim et al., 2000, 2007). It is believed that the specificity of the response to a given stimulus is achieved by decoding specific Ca²⁺ profiles by CBLs followed by the subsequent formation of different CBL/CIPK complexes in planta, and finally by phosphorylation of CIPK-specific substrates that contribute to the specific output response (Batistic et al., 2010).

At the moment, the most numerous and best characterized interactors or substrates for CBL/CIPK complexes are membrane proteins, which include salt overly sensitive 1 (SOS1; Quintero et al., 2002; Katiyar-Agarwal et al., 2006), H⁺-ATPase 2 (He et al., 2004), nitrate transporter (Ho et al., 2009), K⁺ transporter 1 (AKT1; Xu et al., 2006), high-affinity K⁺ transporter 5 (Ragel et al., 2015), and the respiratory burst oxidase homolog F (Drerup et al., 2013). Additionally, CIPKs have also been shown to interact with nonmembrane proteins; for example, SOS2-like protein kinase 5 interacts with the chaperone DnaJ (He et al., 2004; Yang et al., 2010); AtCIPK24 interacts with GIGANTEA (Kim et al., 2013), nucleoside diphosphate kinase 2, the catalases CAT2 and CAT3 (Verslues et al., 2007) and with ABA-insensitive 2 (ABI2), a type 2C Ser/Thr phosphatase (Guo et al., 2002; Ohta et al., 2003); and CIPK26 interacts with the RING-type E3 ligase “Keep on Going” and with ABI1, ABI2 (Lyzenga et al., 2013). Although it appears that CBL/CIPK complexes could interact with several proteins, at present only few CIPK phosphorylation targets have been identified.

Previously, our group demonstrated, to our knowledge, a novel role for tomato (Solanum lycopersicum) Cipk6 (SlCipk6) in plant innate immunity, thus functionally implicating for the first time the participation of a CBL/CIPK module in biotic stress signaling in plants (de la Torre et al., 2013). Other studies demonstrated the participation of Cipk6 orthologs from different plant species in diverse abiotic stress responses (Chen et al., 2012, 2013; Tsou et al., 2012; Tripathi et al., 2009). As a first step to investigate SlCipk6 downstream signaling molecular mechanisms, we set to identify SlCipk6-interacting proteins using a Y2H approach. We discovered that tomato Responsive to desiccation 2 (SlRd2), which contains a Universal stress protein (Usp) domain [Pfam (http://pfam.xfam.org/) accession no. PF00582], interacted with SlCipk6 and by means of a BiFC approach, we found that the complex SlCipk6/SlRd2 is localized in the cytoplasm. In addition, we demonstrated that SlRd2 is a phosphorylation substrate of SlCipk6, thus expanding the previously described substrates for the CIPK family. Interestingly, SlRd2 is an ATP-binding protein that forms homodimers, which is required for its biological role and for interacting with SlCipk6.

The universal stress protein A (UspA) superfamily was originally discovered in Escherichia coli, where its expression drastically increased in response to multiple stress conditions and to starvation (Vanbogelen et al., 1990). Importantly, UspA protein accumulation was necessary for bacterial survival at the stationary phase (Nyström and Neidhardt, 1994). It was found later that E. coli has six usp genes (uspA, uspC, uspD, uspF, uspF, and uspG); however, UspA set the nomenclature for the orthologous groups of proteins. UspA family members are classified into two major groups according to their ATP binding capability. The first group is constituted by ATP-binding proteins and is represented by Mj0577 from Methanococcus jannaschii (Zarembinski et al., 1998). Members of the second group have no ATP-binding capability and are represented by Haemophilus influenzae and E. coli UspAs (Sousa and McKay, 2001). Both Mj0577 and HiUspA form homodimers in vivo (Zarembinski et al., 1998). At present, more than 2000 UspA (or Usp containing domain) proteins have been identified from a wide range of organisms such as bacteria, archaea, fungi, protozoa and plants, constituting an ancient and conserved group of proteins (Aravind et al., 2002). In Arabidopsis, at least 44 proteins were found to contain an Usp domain, all of which resemble ATP-binding Mj0577 protein (Kerk et al., 2003) and several plant UspAs seemed to be involved in abiotic stress. In Arabidopsis, two UspA proteins, AtPHOS32 and AtPHOS34, were phosphorylated by AtMPK3 and AtMPK6 in response to bacterial elicitors in cell suspension cultures (Merkouropoulos et al., 2008). Other reports described several UspA members as effectors of low water potential (Merkouropoulos et al., 2008). Several UspA proteins have been characterized in rice (Sauter et al., 2002), tomato (Zegzouti et al., 1999; Loukehaich et al., 2012) legumes (Becker et al., 2001; Hohnjec et al., 2000), Salicornia (Udawat et al., 2016), and cotton (Zahur et al., 2009). Still, the precise structure, regulation, biochemical function, or mechanism of function of UspA proteins in planta, are largely unknown.

RESULTS

Identification of SlRd2 as a SlCipk6-Interacting Protein

To identify SlCipk6-interacting proteins (CIPs), we carried out a Y2H approach in two separate screens, using a tomato cDNA prey library previously developed (Zhou et al., 1995) and SlCipk6 and a mutant derivative, SlCipk6(T172D), as baits (Fig. 1A). SlCipk6(T172D) displayed enhanced kinase and autophosphorylation activity compared to SlCipk6 (de la Torre et al., 2013), and we hypothesized that using either SlCipk6 or SlCipk6(T172D) as a bait, might facilitate the identification of putative regulatory proteins or phosphorylation substrates respectively, because constitutive active kinase versions stabilize the interaction with their substrates (Uno et al., 2009). Approximately 1.1 × 10³ and 4.5 × 10³ yeast transformants were screened for SlCipk6- and SlCipk6(T172D)-interacting proteins respectively on selection plates (Supplemental Table S1). The inserts of 11 and 34 candidate prey clones from both screens were sequenced and compared with databases by BLAST in the Arabidopsis database (www.tair.org) for identification. The clone no. 29 (Cip29) was a partial open reading frame (ORF) that encoded a protein with a high similarity to Arabidopsis Response to desiccation 2 (AtRD2; At2g21620), a protein not yet characterized with no assigned function, containing a Universal stress protein domain (Usp). Clones containing different length fragments of Cip29 were identified once and two times in the screens performed with SlCipk6 and with SlCipk6(T172D), respectively, and they were among the strongest interactors. In light of these facts, its characterization was prioritized and the remaining CIPs will be published elsewhere. BLAST performed with Cip29 in the tomato database (www.solgenomics.net) identified unigene SGN-U567775 containing a full length ORF that corresponded to the tomato locus Solyc01g109710 and we will be referred to it as SlRd2 hereafter.

Figure 1.

SlRd2 and SlCipk6 Interact in a Y2H Assay. A, Schematic structure of SlCipk6 protein representing N-terminal catalytic, junction and C-terminal regulatory domains, PPi, NAF/FISL, and activation loop domains. Residues mutagenized for SlCipk6 characterization Lys-43 to Met (K43M), and Thr-172 to Asp (T172D) are marked. B, Y2H assay performed with SlCipk6 and SlCipk6 mutant derivatives in the bait vector pEG202 and SlRd2 in the prey vector pJG4-5. Wild-type, full-length SlCipk6; K43M and T172D, single amino acid substitution of Lys-43 to Met and Thr-172 to Asp, respectively; ΔNAF, amino acids 302–321 deletion; ΔCterm amino acids 302–432 deletion; and ΔNterm, amino acids 1–302 deletion. Clones were transformed into EGY48 strain, grown in liquid culture, and spotted at OD₆₀₀ = 0.1 and 0.01 on selective medium. Growth and blue patches indicates interaction. C, Interaction was quantitated by a β-galactosidase assay. Represented is the mean value of three independent experiments, each performed with three independent transformants. Error bars represent the sd of three different experiments.

We next obtained full-length SlRd2 ORF, cloned it in the prey vector, and confirmed its interaction with SlCipk6 by Y2H (Fig. 1B). Both SlRd2 and SlCipk6 did not show autoactivation activity. To characterize SlCipk6/SlRd2 interaction, different SlCipk6 mutant derivatives were tested (de la Torre et al., 2013). SlRd2 interacted with SlCipk6, SlCipk6(T172D), SlCipk6ΔNAF (a mutant version lacking the NAF/FISL domain necessary for CBL binding) or SlCipk6ΔCterm (a deletion mutant version that lacked the C-terminal regulatory domain), but did not interact with SlCipk6ΔNterm (a deletion mutant version lacking the kinase domain) or very weakly with SlCipk6 (K43M, a reduced kinase activity version where the catalytic Lys-43 has been mutated to Met; Fig. 1, A and B). The interaction was quantitated in a β-galactosidase activity assay (Fig. 1C) and protein expression in yeast was confirmed by immunoblot (Supplemental Fig. S1). SlRd2 did not interact with SlCipk11 (Solyc06g082440) or SlCipk14 (Solyc10g085450; Supplemental Fig. S2A) and SlCipk6 did not interact with SGN-U601569, the closest tomato sequence to SlRd2 (Supplemental Fig. S2B), indicating that SlCipk6 interaction with SlRd2 was specific. Altogether, we confirmed that SlCipk6 interacted with SlRd2 and concluded that SlCipk6 kinase domain and kinase activity seemed to be necessary and important, respectively, for SlCipk6/SlRd2 interaction.

SlCipk6 and SlRd2 Interact in Vivo

Next, we examined the in vivo interaction between SlCipk6 and SlRd2 by coimmunoprecipitation (co-IP). Although SlCipk6 interacted with SlRd2 in yeast at slightly higher levels, we used SlCipk6(T172D) for co-IP in planta because it has been described that constitutive active kinase versions stabilize the interaction with their substrates (Uno et al., 2009). For that purpose, both SlRd2 (fused to a GFP tag) and SlCipk6(T172D; fused to 2xIgG-BD and 9xMyc tags) were coexpressed via Agrobacterium-mediated transient transformation of leaves (agro-infiltration) in Nicotiana benthamiana. SlRd2 coimmunoprecipitated with SlCipk6(T172D), but did not with the GFP, used as a negative control in this experiment (Fig. 2A). Therefore, SlCipk6(T172D) and SlRd2 interacted in planta. Because kinases and their substrates need a coordinated expression and subcellular localization, we first assessed the likelihood that both SlCipk6 and SlRd2 localized at common cellular compartments. Both proteins were fused to GFP, agro-infiltrated individually into N. benthamiana leaves and the localizations were determined by confocal microscopy (Supplemental Fig. S3). Both SlCipk6-GFP (Supplemental Fig. S3, a and b) and GFP-SlRd2 (Supplemental Fig. S3, c and d) localized in the cytoplasm and in the nucleus, thus showing an overlapping subcellular localization.

Figure 2.

SlCipk6 and SlRd2 interact in planta and the complex localizes in the cytosol. A, Protein extracts from N. benthamiana leaves agro-infiltrated with GFP-SlRd2 and SlCipk6(T172D)-Myc, which is fused to two copies of the protein A IgG binding domains (2xIgG-BD), were incubated with IgG agarose beads, eluted and analyzed by immunoblotting using anti-Myc for SlCipk6 and anti-GFP for SlRd2. GFP protein was used as a negative control. No interaction was detected when SlCipk6(T172D) was agro-infiltrated with GFP (right upper panel). B, Fluorescence images of N. benthamiana leaf sections agroinfiltrated with SlRd2-YFP^C and SlCipk6-YFP^N. Colocalization analyses of SlCipk6/SlRd2 complexes (yellow, a and d) with the fluorescent membrane marker dye FM4-64 (red, b) and the cytoplasmic marker AtFKBP12-CFP (cyan, e; Faure et al., 1998). Merged images are shown in (c) and (f), respectively. White arrows show the cytoplasmic localization observed for the SlCipk6/SlRd2 complexes. Localization analysis of YFP^N (g) and YFP^C (h) empty vectors and SlCbl10/SlRd2 (i) were used as an experiment controls. Bars = 10 μm.

To gain insight where the complex localized intracellularly, we performed BiFC assays in N. benthamiana leaves (Walter et al., 2004). Fluorescence due to the reconstitution of yellow fluorescent protein (YFP) was observed at the confocal microscope when the combinations SlRd2-YFP^C and SlCipk6-YFP^N were agro-infiltrated in N. benthamiana leaves (Fig. 2B, a and d). The YFP signal showed a pattern similar to the one observed for AtFKBP12-CFP, a cytoplasmic marker fused to cyan fluorescent protein [CFP; (Faure et al., 1998)] (Fig. 2B, d–f), but did not overlap with the signal derived from the membrane-bound marker FM4-64 (Fig. 2B, a–c). No fluorescence was detected in the nucleus or when SlRd2-YFP^C or SlCipk6-YFP^N were coinfiltrated with empty vectors respectively, and neither when SlRd2-YFP^C was coexpressed with SlCbl10-YFP^C, a known interactor of SlCipk6 (Fig. 2B, g–i; de la Torre et al., 2013). These results indicate that SlCipk6 and SlRd2 interact in vivo and form a cytoplasmic complex in N. benthamiana leaves at the conditions assayed.

SlRd2 Transcript Accumulates in Response to Abiotic Stress

E. coli UspA (EcUspA) protein accumulates in response to a large and diverse number of stresses providing survival cues under adverse bacterial growth conditions (Freestone et al., 1997; Jung et al., 2015). In plants, members of the UspA family have been reported to accumulate in response to different abiotic stress conditions, including AtRD2 (Yamaguchi-Shinozaki et al., 1992; Jung et al., 2015). Both SlCipk6 and SlRd2 displayed an overlapping subcellular localization in N. benthamiana epidermal cells (Supplemental Fig. S3, a–d). Next, we inquired if SlRd2 and SlCipk6 expression were correlated in response to abiotic stress using quantitative (Q) RT-PCR to support synchronous activity in abiotic stress response.

SlRd2 mRNA was present in all the tissues examined, however it was substantially more abundant in roots (>25-fold compared to leaves), suggesting that it might play a prominent role in this tissue (Fig. 3A). AtRD2 accumulation increased in response to abiotic stress (Yamaguchi-Shinozaki et al., 1992). Therefore, to determine if SlRd2 and SlCipk6 transcript accumulation were responsive to salt and osmotic stress, we treated hydroponically grown tomato plants with 100 mm NaCl and 300 mm mannitol and analyzed SlRd2 accumulation. NaCl treatment rapidly increased SlRd2 transcript accumulation [approximately 12-fold, after 2 h (h)], which was maintained up to 12 h and decreased 24 h later (Fig. 3B). Osmotic pressure resulted in a similar accumulation of SlRd2 after 2 h and 4 h, but in contrast, SlRd2 accumulation increased up to 30-fold after 8 h to 12 h and decreased to 20-fold at 24 h (Fig. 3C). SlCipk6 orthologs from different plant species have been described to participate in different abiotic stress responses, and we tested if SlCipk6 transcript accumulated also in response to abiotic stress. For that purpose, we treated tomato plants with NaCl and mannitol, using the same conditions as described above. After NaCl treatment, accumulation of SlCipk6 mRNA increased 6-fold to 8-fold after 2 h and was maintained at the same levels up to 8 h, decreasing gradually thereafter, reaching a 2-fold increase at 24 h (Fig. 3D). Osmotic pressure also resulted in an increased SlCipk6 mRNA accumulation; however, followed a different pattern compared to SlRd2: 2 h after SlCipk6 transcripts increased 6-fold and steadily increased up to 15 fold 24 h after (Fig. 3E). Both SlCipk6 and SlRd2 transcript accumulation pattern showed similar kinetics, thus supporting the possibility of a coordinated role for SlCipk6 and SlRd2 in abiotic stress responses in tomato.

Figure 3.

SlRd2 and SlCipk6 transcripts highly accumulate after nacl and osmotic stress in tomato. A to E, y axis represents mean values of Q RT-PCR of three experiments, with three biological replicates in each. Error bars represent the se. The expression levels were normalized to SlActin2. Gene induction (fold increase) in infected or treated plants was compared with the expression level of control or mock inoculated plants at 0 h and is shown as relative expression, excepting in (A). A, SlRd2 transcript accumulates in tomato leaves, petioles, stems, flowers, and roots. Relative expression on y axis was compared with expression level in leaves. B and C, SlRd2; D and E, SlCipk6 mRNA accumulation in leaves during NaCl (B and D) and osmotic stress (C and E). Hydroponically grown tomato plants were treated with 100 mm NaCl (B and D) or 300 mm mannitol (C and E). OS, osmotic stress.

SlRd2 Belongs to the Universal Stress Protein A Superfamily

The translated sequence of SlRd2 ORF yields a 177 amino acid protein, with a M_r of 19.5 kD and a pI of 5.98. We aligned the Usp domain of SlRd2 with those of N. benthamiana and Arabidopsis orthologs NbRd2 and AtRD2 and other plant proteins containing the Usp domain, including tomato LeER6 (Zegzouti et al., 1999), rice (Oryza sativa) OsUSP1 (Sauter et al., 2002), and Vicia fava Enod18 (Becker et al., 2001) along with bacterial proteins belonging to the UspA family, including E. coli (Nyström and Neidhardt, 1992), H. influenzae (Sousa and McKay, 2001) and M. jannaschii (Zarembinski et al., 1998; Supplemental Fig. S4). Amino acid sequence alignment revealed that SlRd2 contained the residues involved in ATP binding including the Walker motif A or P-loop, (G-2X-G-9X-G(S/T)), also present in Mj0577 and UspA plant representatives but absent in E. coli paralogs (Sousa and McKay, 2001). This observation suggests that plant UspA proteins might also be functional ATP-binding proteins (Supplemental Figs. S4 and 5A). Overall, all UspA proteins (plant and bacterial) shared conservation within the dimerization subdomain (Supplemental Fig. S4), thus raising the possibility that plant UspAs could be present as dimers in the cell. In addition to the conserved Usp domain, SlRd2 has an N-terminal domain (amino acids 1 to 38) and a C-terminal extension (amino acids 168 to 177) with unknown function, which is shared with the Arabidopsis and N. benthamiana homologs (Figs. 4A and S5). Subsequently, a phylogenetic analysis was performed using E. coli UspG and the proteins included in the alignment shown in Supplemental Fig. S4. SlRd2 and NbRd2 were located in the same clade as AtRD2, thus supporting their orthology (Fig. 4B). Moreover, SlRd2, and all plant UspAs, are more related to Mj0577 subfamily and E. coli UspG than to the E. coli UspA (Fig. 4B).

Figure 5.

SlRd2 is an ATP-binding protein. A, Alignment of Usp domain of SlRd2 and AtRD2 and the full amino acid sequence of Mj0577. Conserved residues are marked in black boxes while similar residues are marked in gray boxes. Black bars underline residues putatively involved in ATP-binding. Walker motif or P loop motif G-2x-G-9x-G-(S/T) for ATP binding is marked with a blue line. Amino acid residues responsible for protein dimerization are marked as (D). B, For ATP-binding in vitro assay, 1 μg of purified His-SlRd2 and His-GFP proteins from E. coli were incubated with [α-³²P]ATP and corresponding buffer during 1 h at 30°C. Autoradiograph indicates ATP-binding (top panel); immunoblot analysis indicates protein level in the ATP binding assay (low panel). C, For ATP-binding in vivo assay, N. benthamiana leaf extracts from wild type and GFP-Rd2 overexpressing line 6 plants were incubated with 20 μg BHAcATP. Biotinylated proteins were purified using streptavidin beads, separated on protein gel, and detected using anti-GFP (top panel). Samples labeled with BHAcATP in presence of 10 mm of ATP were used as a control. Ponceau S staining was used as a loading control (bottom panel). Adenine, A; phosphate, P; ribose, R.

Figure 4.

Tomato SlRd2 protein structure and phylogenetic analysis. A, Schematic structure of SlRd2 containing N-terminal and C-terminal domains and a conserved Usp domain, which encompasses a dimerization motif. Special symbols indicate the conserved residues described to be involved in ATP-binding. B, Phylogenetic relationship of proteins containing Usp domains from plants and bacteria including A. thaliana, AtRD2; tomato, SlRd2 and LeER6; N. benthamiana, NbRd2; O. sativa, OsUSP1; V. faba, VfENOD18; E. coli, EcUspG and EcUspA; H. influenzae, HiUspA; and M. jannaschii, Mj0577. SlCbl10 was used as the outgroup. The numbers on the tree represent bootstrap scores.

SlRd2 Is an ATP-binding Protein and Forms Homodimers in Yeast, Bacteria, and Plants

In light of the high degree of conservation of SlRd2 amino acids putatively involved in nucleotide binding (see alignments in Figs. 5A and S4), we next inquired if SlRd2 also had the functional competence to bind ATP, as described for bacterial Mj0577. For that purpose, we purified SlRd2 protein and performed a nucleotide binding assay in vitro by incubating SlRd2 in the presence of [α-³²P]ATP. Samples were then analyzed by SDS-PAGE followed by autoradiography (Fig. 5B, upper panel). Indeed, SlRd2 was able to bind [α-³²P]ATP, whereas no [α-³²P]ATP binding was observed when the GFP protein was incubated in its presence. To further confirm that SlRd2 was also able to bind ATP in vivo, a synthetic analog of ATP, (+)-biotin-hex-acyl-ATP (BHAcATP), consisting of ATP, an acyl-P linker and a biotin tag, was used (Villamor et al., 2013). Incubation of GFP-SlRd2 expressing plant extracts with BHAcATP followed by streptavidin purification revealed that SlRd2 was able to bind BHAcATP in planta (Fig. 5C, upper panel). Moreover, ATP addition suppressed BHAcATP labeling by competing with BHAcATP and saturating the nucleotide-binding site of SlRd2 (Fig. 5C, upper panel). These findings indicate that SlRd2 binds ATP and therefore functionally belongs to the ATP-binding subgroup of the UspA family represented by Mj0577.

The dimerization motif, present in UspA family proteins, is also conserved in SlRd2 and is localized close to the C terminus (Figs. 5A and S4). Mj0577 and HiUspA were shown to exist as homodimers in vivo (Sousa and McKay, 2001; Zarembinski et al., 1998). Mj0577 crystallizes as a homodimer, and each monomer binds the other through antiparallel hydrogen bonds in the fifth beta sheet within each subunit. To test whether SlRd2 was also able to form homodimers, we performed a Y2H analysis in which the yeast strain expressed SlRd2 both in the prey (pJG4-5) and in the bait (pEG202) plasmids. Indeed, SlRd2 can form homodimers because growth was observed in restrictive media and blue color developed in the presence of X-gal (Fig. 6A). No interaction was detected when Y2H analysis was performed using a SlRd2 mutant version, SlRd2∆dim (amino acids 163 to 166 deletion), in which the putative dimerization domain VIIV was deleted (Fig. 6A). SlRd2 and SlRd2∆dim, were expressed in yeast (Fig. 6B). Therefore, SlRd2 forms dimers and the conserved VIIV domain is necessary for dimerization in vivo.

Figure 6.

SlRd2 homodimerizes in the cytosol and in the nucleus. A, Y2H assay with both SlRd2 and SlRd2Δdim cloned in the bait pEG202 (with a LexA fusion tag at the C terminus) and prey pJG45 (with a HA fusion tag at the C terminus) vectors in EGY48 strain, grown in liquid culture and spotted at OD₆₀₀ = 0.1 and 0.01 on selective medium. Growth and blue patches indicates interaction. B, Expression of both SlRd2 and SlRd2Δdim proteins in yeast were confirmed by immunoblotting using anti-LexA (pEG202) and anti-HA (pJG45) antibodies, respectively. C, The cross-linking agent DSG was added to a His-SlRd2 overexpressing E. coli culture for 30 min. Polymerization of His-SlRd2 protein was calculated according to its molecular weight: 1×, monomer; and 2×, dimer. D, Colocalization analyses of SlRd2/SlRd2 complexes (yellow, a, d, and g) with the membrane marker dye FM4-64 (red, b), cytoplasmic marker AtFKBP12 (cyan, e), or nuclear marker DAPI (blue, h). Merged images are shown in (c), (f), and (i). Arrow denotes localization in the nucleus. Bars = 10 μm.

It has been described that E. coli UspC is able to form tetramers in vivo (Nachin et al., 2008). To check if SlRd2 also formed homotetramers in vivo, an E. coli culture overexpressing SlRd2 tagged at the N terminus with the epitope His (His-SlRd2) was treated with a cross-linking agent, disuccinimidyl glutarate (DSG), for 30 min. Thereafter, protein extracts were obtained, analyzed by SDS-PAGE and His-SlRd2 detected by immunoblot. Two bands of approximately 42 and 24 kD were observed in the extracts treated with the crosslinker, which corresponded likely with SlRd2 dimer and monomer, whereas only the lower M_r band was observed in control conditions (Fig. 6C). Thus, we concluded that SlRd2 is able to form homodimers but not homotetramers in vivo.

To determine that SlRd2 formed homodimers in planta and their putative subcellular localization, BiFC assays were performed. SlRd2 cDNA was cloned in the BiFC vectors, transformed into Agrobacterium and agro-infiltrated in N. benthamiana leaves. Reconstitution of YFP fluorescence was observed 2 d after under the confocal microscope when SlRd2-YFP^C and SlRd2-YFP^N were coexpressed. The signal colocalized with that obtained by the expression of the cytoplasmic marker AtFKBP12-CFP (Fig. 6D, d–f) and with DAPI nuclear marker (Fig. 6D, g–i) but not with the cell membrane marker FM4-64 (Fig. 6D, a–c). These results demonstrate that SlRd2 can form homodimers in planta in the cytoplasm and in the nucleus.

SlRd2 Is a Phosphorylation Substrate of SlCipk6

Because SlCipk6 kinase domain and activity was necessary and important respectively for SlCipk6/SlRd2 interaction (Fig. 1), we inquired if SlRd2 was a phosphorylation target of SlCipk6. For that purpose, we set up an in vitro kinase assay in the presence of [γ-³²P]ATP using purified SlCipk6 (fused to the GST at the C terminus) and as substrates, SlRd2 (fused to the His epitope at the N terminus) and a mutant version, SlRd2(UspA; fused to GST at the N terminus), in which the specific putative regulatory N-terminal domain (amino acids 1 to 38) and C-terminal extension (amino acids 168 to 177) were deleted to leave only a core UspA domain. Myelin Basic Protein (MBP), a universal kinase substrate was included as a positive control for SlCipk6 kinase activity. SlCipk6 was expressed and purified from yeast, whereas SlRd2 and SlRd2(UspA) were purified from bacteria. Figure 7A shows that SlRd2 is phosphorylated by SlCipk6 but does not phosphorylate SlRd2(UspA), thus indicating that SlCipk6 phosphorylates SlRd2 specific regulatory N- or C-terminal regions. SlCipk6 also phosphorylates MBP. A faint signal can be still detected in the absence of SlCipk6, indicating that SlRd2 binds ATP (Fig. 7A, left). Protein loading levels were monitored by Coomassie staining (Fig. 7A). Quantification of SlRd2 [γ-³²P]ATP incorporation is presented in Figure 7B. Therefore, we conclude that SlRd2 is a phosphorylation substrate of SlCipk6 and represents, to our knowledge, a novel phosphorylation target of the CIPK family.

Figure 7.

SlCipk6 Phosphorylates SlRd2. A, SlCipk6, SlRd2, SlRd2(UspA) and MBP alone and with SlCipk6 were incubated in the presence of [γ-³²P] ATP and kinase buffer, electrophoresed on SDS polyacrylamide gel and autoradiographed. They were expressed and purified as fusion proteins as follows: SlCipk6-GST, His-SlRd2, GST-SlRd2(UspA). Autoradiographs indicate SlCipk6 autophosphorylation (upper panels) and SlRd2 phosphorylation (lower panels); Coomassie protein staining indicates SlRd2, SlRd2(UspA), MBP, and SlCipk6 protein loads. Protein amounts loaded: 0.3 μg for SlCipk6, 2.5 μg for SlRd2 and SlRd2(UspA), and 0.5 μg for MBP. B, Phosphorylation was quantified using Cyclone Phosphorimager Optiquant software (Packard Bioscience, Perkin Elmer). Signal from MBP phosphorylation was not normalized to protein content to better quantify SlRd2 phosphorylation and ATP binding. Similar results were obtained in two additional experiments. a.u., arbitrary units.

SlRd2 Overexpression in Yeast Confers Resistance to LiCl

In light of the strong transcriptional response of SlRd2 to NaCl and osmotic stress, we decided to functionally test SlRd2 participation in mediating different stress responses using yeasts as a model system. Yeast has been successfully used in functional studies to characterize the role of plant proteins in stress responses (Matsumoto et al., 2001; Quintero et al., 2002; Shitan et al., 2013; Bernard et al., 2012). SlRd2 was overexpressed in the yeast strain Saccharomyces cerevisiae BY4741 under control of the constitutive expression vector p426GPD (p426GPD-SlRd2). As a control, BY4741 strain transformed with empty vector (EV) p426GPD was used. Both yeasts lines were subjected to the following stress-causing treatments: hygromycin B, generates changes in membrane potential (Wang et al., 2009); MnCl_2, causes programmed cell death at high concentrations (Liang and Zhou, 2007); CaCl₂ and KCl, cause osmotic stress (Liang and Zhou, 2007); DTT and H₂O₂, generate oxidative stress (Babiychuk et al., 1995); bleomycin, causes DNA damage (Aouida et al., 2004); and finally, NaCl and LiCl, cause changes in ionic homeostasis (Ye et al., 2006). SlRd2 overexpressing line grew at a similar rate than EV strain in control conditions and in all the different treatments, except in 300 mm LiCl, where it grew significantly better than EV line (Fig. 8A). Next, we tested if SlRd2 dimerization was crucial for its biological role. For that purpose, we cloned SlRd2∆dim into p426GPD vector, transformed it into S. cerevisiae BY4741 strain and performed a LiCl resistance experiment along with p426GPD-SlRd2 and EV. Figure 8B shows that a functional SlRd2 dimer was required for conferring LiCl resistance, because SlRd2∆dim line grew at the same rate as EV line and loss its protection against LiCl. Both SlRd2 and SlRd2∆dim expression in yeast was confirmed by immunoblotting (Fig. 8C). We concluded that SlRd2 dimerization was necessary for its molecular function in conferring resistance to LiCl.

Figure 8.

SlRd2 expression confers LiCl resistance in yeast and its dimerization is required for its biological function. A, BY4741 yeast strain expressing SlRd2 (cloned into p426GPD) was subjected to different stress conditions. As a control, BY4741 was transformed with the EV. Both SlRd2 and EV lines were grown in liquid culture and spotted on specific medium at different dilutions. Overexpression of SlRd2 showed a higher growth than EV in response to LiCl (300 mM). B, Overexpression of SlRd2Δdim showed the same growth rate than EV line in response to LiCl. The experiment was performed as described in (A). C, Expression of SlRd2 and SlRd2Δdim proteins in yeast were confirmed by immunoblotting using anti-SlRd2 antibodies.

Next, we asked if SlRd2 dimerization was also necessary for its interaction with SlCipk6. For that purpose, we performed a Y2H assay with SlCipk6 and SlCipk6(T172D) as baits and SlRd2∆dim as a prey (Fig. 9A). Neither SlCipk6 nor SlCipk6(T172D) interacted with SlRd2∆dim by Y2H (Fig. 9A). To test their interaction in planta, we performed a BiFC experiment, and agro-infiltration SlRd2∆dim-YFP^C and SlCipk6-YFP^N in N. benthamiana leaves did not reconstitute YFP fluorescence, indicating that SlCipk6 was not able to interact with SlRd2∆dim (Fig. 9B). Hence, SlRd2 homodimerization is required for interaction with SlCipk6.

Figure 9.

SlRd2 dimerization is required for its interaction with SlCipk6. A, Y2H assay performed with SlCipk6 and SlCipk6(T1172D) in the bait vector pEG202 and SlRd2Δdim in the prey vector pJG4-5. Clones were transformed into EGY48 strain, grown in liquid culture, and spotted at OD₆₀₀ = 0.1 and 0.01 on selective medium. Growth and blue patches indicates interaction. A Y2H experiment performed with SlRd2 and SlCipk6 was used as a control. B, Fluorescence images of N. benthamiana epidermal cells expressing SlRd2Δdim-YFP^C and SlCipk6-YFP^N. YFP signal was not reconstituted in the BiFC experiment (a), confirming that SlRd2 dimerization is required for the interaction with SlCipk6 in planta. A BiFC experiment performed with SlRd2 and SlCipk6 was used as a control (b). Bars = 10 μm.

SlRd2 Protects Bacteria from Oxidative Stress and Negatively Regulates Reactive Oxygen Species in Plants

E. coli mutant strain TN3151 (a knock-out of the uspA gene) was susceptible to H₂O₂ treatments, which did not affect the wild-type strain, W3101 (Nyström and Neidhardt, 1994). Recent results demonstrated that a UspA protein from the pathogenic bacteria Mycobacterium tuberculosis provided protection for the parasite against host reactive oxygen species (ROS) generated by mammalian macrophages defense (Drumm et al., 2009). SlCipk6 was demonstrated to participate in ROS generation during plant responses to bacterial pathogen attack, which was dependent on the NADPH oxidase, RbohB (de la Torre et al., 2013). Given the striking conserved structural and functional features between SlRd2 and bacterial UspA proteins, we decided to test if SlRd2 could functionally complement TN3151 in protecting bacteria in response to H₂O₂ treatment. E. coli wild-type W3101 (wild type), uspa mutant TN3151 and TN3151 complemented with either SlRd2 (TN3151-SlRd2) or with the native UspA gene (TN3151-UspA) were grown in minimal media (morpholine propane sulphonic acid, MOPS) until OD₆₀₀ reached 0.7. Then, H₂O₂ (5 mM) was added and survival was measured at different time points after (Fig. 10A). A strong growth inhibition was observed for TN3151 strain after H₂O₂ treatment, as previously described (Nyström and Neidhardt, 1994); 40 min after H₂O₂ addition, only 45% of TN3151 survived whereas 100% TN3151-SlRd2 survived, showing a similar survival rate as either TN3151-EcUspA or wild-type W3101. SlRd2 and EcUspA expression in TN3151 was confirmed by immunoblotting (Fig. 10B) and lack of UspA gene expression in TN3151 was confirmed by PCR (Fig. 10C). In light of these results, we concluded that tomato SlRd2 functionally complements TN3151 mutant strain in protecting bacteria against oxidative stress damage and thus SlRd2 and E. coli UspA are functionally conserved in their ROS protection role in bacteria.

Figure 10.

Tomato SlRd2 restores E. coli TN3151 (Δuspa) survival and negatively regulates SlCipk6-mediated ROS. A, Wild type [W3101 (F-, IN(rrnD-rrnE)1, rph-1)], Uspa mutant TN3151 [F-, IN(rrnD-rrnE)1, UspA::kan, rph-1)], and the overexpressing TN3151 SlRd2 (OE-SlRd2) and TN3151 UspA (OE-UspA) E. coli lines were grown in MOPS medium and treated with H₂O₂ (5 mM) when the cultures reached OD₆₀₀ = 0.7. B, Immunoblot experiment confirmed SlRd2 (upper panel) and UspA (lower panel) expression in TN3151 E. coli lines. C, Transcript abundance of UspA cDNA in TN3151 and wild-type strains was determined by RT-PCR. 16S rRNA transcript abundance was used as an internal standard control. D, Measurement of ROS production in N. benthamiana leaf disks of GFP-SlRd2 overexpressing (OE-6 and OE-7) and wild-type plants, after agro-infiltration of SlCipk6-Myc (wild type), SlCipk6(T172D)-Myc (T172D), SlCipk6(K43M)-Myc (K43M), or EV. ROS accumulation was quantified as relative light units 8 h to 12 h after. Data are presented as means ± ses of four independent experiments. Means with different letters are significantly different at P < 0.05, Student’s t-test. Anti-GFP and anti-Myc were used as primary antibodies. RLU, relative light units; WT, wild type.

We have previously shown that kinase activity of SlCipk6 is associated with ROS production in N. benthamiana (de la Torre et al., 2013), so we next inquired if SlRd2 is required for SlCipk6-mediated ROS generation. To this end, c-Myc tagged versions of SlCipk6, SlCipk6(T172D) and SlCipk6(K43M; cloned into pTAPa-pYL436 vector) or EV were agro-infiltrated in N. benthamiana wild-type and overexpressing GFP-SlRd2 (OE-6, OE-7) leaves and production of ROS was quantified by a chemiluminescence assay in a luminometer (Fig. 10D). As expected, SlCipk6 and SlCipk6(T172D) expression in wild-type leaf discs resulted in ROS generation, which was significantly reduced in SlCipk6(K43M) expressing leaf discs (de la Torre et al., 2013). However, SlCipk6 and SlCipk6(T172D) agro-infiltration in OE-6 and OE-7 plants resulted in a significant reduction of ROS (Fig. 10D) whereas SlCipk6(K43M)-induced ROS levels remained as in wild-type discs. SlCipk6, SlCipk6(T172D), SlCipk6(K43M), and GFP-SlRd2 expression were confirmed by immunoblotting (Supplemental Fig. S6). This result clearly indicated that overexpression of SlRd2 negatively regulates SlCipk6-mediated ROS generation and that a functional link exists between both proteins.

DISCUSSION

SlRd2, a Member of the Universal Stress Protein Family, Interacts with SlCipk6

The identification and characterization of SlCipk6 targets and regulatory components is an important step for understanding how downstream SlCipk6 signaling specificity is achieved and to identify the pathways it regulates. As a first step, we set to identify SlCipk6-interacting proteins using a Y2H approach with either SlCipk6 or its enhanced kinase activity version SlCipk6(T172D) as baits, and a tomato library as a prey (Zhou et al., 1995). We expected to identify putative phosphorylation substrates or regulatory proteins. SlRd2 was identified in both screens and displayed one of the strongest interactions. SlCipk6/SlRd2 interaction was later confirmed and appeared to have some level of specificity. In addition, no assigned function was available for this protein. Hence, we prioritized its characterization. For SlCipk6/SlRd2 interaction, SlCipk6 kinase domain and kinase activity was required, thus suggesting that SlRd2 could be a phosphorylation substrate of SlCipk6. SlCipk6/SlRd2 interaction was further confirmed in planta, using coimmunoprecipitation and BiFC approaches. Thus, we compiled data using different approaches demonstrating that SlCipk6/SlRd2 interacted in N. benthamiana.

BiFC experiments indicated that SlCipk6/SlRd2 complex localized in the cytosol. Both GFP tagged SlCipk6 or SlRd2 localized mainly in the nucleus and in the cytoplasm of N. benthamiana epidermal cells, thus showing a coordinated subcellular localization pattern. However, SlCipk6/SlRd2 complex was only detected in the cytosol. We cannot rule out that the complex could localize to the nucleus or to different cellular compartments when complexed with CBLs or under stress conditions. In fact, it has been demonstrated that CIPKs can be targeted to different intracellular compartments via their interacting CBLs (Batistic et al., 2010). Accordingly, tomato Cipk6 was found to interact with SlCbl10 and with NbRbohB, a membrane based NADPH oxidase at the plasma membrane in N. benthamiana epidermal cells (de la Torre et al., 2013) whereas Brassica napus CIPK6 (also plasma membrane- and nucleus-localized), was found at the plasma membrane in complex with BnCBL1 (Chen et al., 2012). Here, we describe the localization of a CIPK in complex with its substrate, which poses additional unsolved questions, such as whether CBLs translocate CIPKs to meet their substrate at the final destination or if they could also transport the attached substrates as cargoes. CBLs and CIPKs interact with some, but not with all CIPK and CBL partners respectively, thus providing an enormous combinatorial signaling flexibility (Luan, 2009; Yu et al., 2014). At this point, we do not know which CBL(s) relay Ca²⁺ signals to SlCipk6 and SlRd2.

SlCipk6 Phosphorylates SlRd2, a Universal Stress Protein and a Novel Target of CIPK Family

We have demonstrated that SlCipk6 phosphorylates SlRd2 in vitro. In plants, Ca²⁺ intracellular increases occur after several intra- and extracellular stimuli, including biotic, abiotic, and developmental changes (Hashimoto and Kudla, 2011). It is noteworthy that tomato exposure to abiotic stress (osmotic stress and salinity) results in a rapid and strong accumulation of SlRd2 transcripts, which resembles SlCipk6 accumulation pattern. Similarly, in silico data indicates that under osmotic stress (mannitol 300 mM) in Arabidopsis AtCIPK6 and AtRD2 are transcriptionally upregulated (Arabidopsis eFP Browser). Based on our results, we propose that SlRd2 is a phosphorylation target of SlCipk6 and therefore, a Ca²⁺ signaling downstream effector or more likely, a downstream Ca²⁺ signaling component. Our results also indicate that CIPKs, in addition to membrane proteins, phosphorylate additional substrates, thus expanding their role in regulating different physiological processes. Previous work by others have shown that different UspA proteins from bacteria and plants are phosphorylated in response to stress, indicating that this posttranslational modification may be important for their role in the cell (Merkouropoulos et al., 2008; Lenman et al., 2008; Freestone et al., 1997). In addition, it was also found that the phosphosites were localized outside the Usp domain, as likely occurs in SlRd2 (Merkouropoulos et al., 2008). Likely the ancient and conserved function of the Usp domain adapts to different stress signals in different organisms acquiring specific signaling domains to modulate specific output responses. In the future, we will identify the exact SlRd2 residues phosphorylated by SlCipk6. Once identified, functional analysis in plants will be performed using transgenic plants with altered expression levels of SlRd2 or SlRd2 mutant versions in which the residues phosphorylated by SlCipk6 will be mutated to A (not a substrate) or to D (phosphomimics). These studies will help clarify SlRd2 role and SlCipk6 phosphorylation contribution in ROS generation and regulation in plants.

Using in vitro and in vivo assays, we demonstrate that SlRd2 is also an ATP-binding protein, like E. coli UspFG subfamily members and group representative Mj0577. Mj0577 and plant UspAs share the common motif present in Usps experimentally proven to bind ATP (Kvint et al., 2003). Obtaining SlRd2 nucleotide binding impaired mutant versions will help to understand the physiological relevance of this feature in plants under stress responses. The ATP binding capability in some members of the UspA family has led to the speculation that nucleotide binding Usps could function as molecular switches by sensing ATP levels during stress signaling detecting cellular energy or metabolic status (O’Toole et al., 2003; Persson et al., 2007; Drumm et al., 2009). In fact, autoadenylation is observed in bacteria in late stationary phase (Weber and Jung, 2006) and it has been shown to be a key factor in microbial survival under O₂ depletion, during growth arrest and in virulence. In this line, it has been described that the ability of UspA protein Rv2623 from M. tuberculosis to regulate its growth and latency in the host, is dependent on its ATP-binding activity (Drumm et al., 2009). Understanding the molecular mechanisms by which Usp-proteins act has broader implications in human health, because they contribute to human pathogen’s virulence and survival in the host (Seifart Gomes et al., 2011; Liu et al., 2007).

Although several reports described that UspA proteins contain a dimerization domain in its sequence, little is known about its functional implication. It has been described that EcUsp proteins have the capability to form homodimers and/or heterodimers in vivo leading to a higher adaptation to stress (Nachin et al., 2008; Heermann et al., 2009). In this work, we have found that SlRd2 forms homodimers in vivo at the cytoplasm and nucleus. However, we cannot discard that SlRd2 might form heterodimers with other UspA proteins in the plant cell. Interestingly we found that the dimerization of SlRd2 is required for both its biological stress-protection role in yeast and its interaction with SlCipk6. Similarly, Weber and Jung (2006) described that the dimerization of EcUspG was necessary for its cellular function. Unlike plants, the importance of UspA dimerization is well documented in E. coli. Thus, the Usp domain of KdpD (K⁺ transport system) functions as a binding surface for EcUspC and it is essential for its signaling role (Heermann et al., 2009). Recently it has been found that Arabidopsis AtUSP is able to switch from low M_r species to high M_r complexes, suggesting a chaperone function in stress tolerance to heat shock and oxidative stress (Jung et al., 2015). Notably, dimerization of Hypoxia Responsive Universal Stress Protein 1 is also important for ROS regulation and apparently for subcellular localization (Gonzali et al., 2015). An important question is whether SlRd2 dimerization is also required for ATP binding, which could in turn regulate the interaction with SlCipk6. These aspects deserve further analyses in the future.

SlRd2 Protects Bacteria against ROS and Regulates SlCipk6-mediated ROS in Plants

Despite UspA proteins are widely represented in plants [with 48 members in Arabidopsis (Kerk et al., 2003; Isokpehi et al., 2011)], very little is known about their function, regulation, molecular mechanisms or their participation in physiological responses. Previously, two Arabidopsis UspA proteins (At5g54430, At4g27320), were identified as differential phosphorylation substrates in response to pathogen derived elicitors (Lenman et al., 2008; Merkouropoulos et al., 2008). In vitro kinase assays identified AtPHOS32 (At5g54430) as a mitogen-activated protein kinases AtMPK3 and AtMPK6 substrate (Merkouropoulos et al., 2008). This observation along with SlRd2 being phosphorylated by SlCipk6 (a Ca²⁺-regulated kinase) indicates that plant UspA members might be regulated distinctly in response to a plethora of stimuli thus receiving and integrating signals from different pathways.

An interesting observation is that overexpression of SlRd2 in S. cerevisiae results in an increased tolerance to LiCl. In yeast, increased LiCl tolerance is promoted by a rise of activity of the K⁺ transporter Trk1/2, which is in turn controlled by phosphorylation and dephosphorylation modifications (Zaidi et al., 2012; Yenush et al., 2005). In plants CIPK/CBL complexes regulate the cellular K⁺ flux by interaction with the transporter AKT1 (Li et al., 2014). According to the yeast data, it is tempting to speculate that SlCipk6 and SlRd2 might act together in plants to regulate the activity of the plant K⁺ transporter AKT1. In fact, in Arabidopsis, AtCIPK6 interacts with different CBLs to regulate AKT1 and AKT2 (Lee et al., 2007; Lan et al., 2011; Held et al., 2011). However, further studies should be carried out to figure out this functional implication. Significantly, we have found that SlRd2 complements an E. coli Uspa mutant line restoring bacterial viability to wild-type levels to otherwise lethal doses of H₂O₂ for the mutant. Because EcUspA does not contain ATP-binding regions whereas SlRd2 can bind ATP, both proteins might play a common and conserved molecular role that does not imply ATP binding, at least in bacteria (Weber and Jung, 2006). The functional complementation of E. coli Uspa mutant by the expression of tomato Rd2 protein indicates that the molecular function of SlRd2, a plant UspA protein in protecting cells against the toxic effects of oxidative stress in bacteria have been conserved in evolution.

A rapid ROS burst have been implicated in different physiological responses. Recently, we reported that SlCipk6 contributed to ROS generation during biotic stress response in N. benthamiana, which largely depends on RbohB (de la Torre et al., 2013). Also, SlCipk6 overexpression resulted in ROS generation, which was dependent on SlCipk6 kinase activity but did not occur in RbohB silenced N. benthamiana plants (de la Torre et al., 2013). The fact that overexpression of SlRd2 results in reduced SlCipk6-mediated ROS in planta clearly indicates a functional relationship between both proteins in which SlRd2 negatively regulates SlCipk6-mediated ROS output. Similarly, overexpression of Salicornia brachiata Usp in tobacco plants resulted in reduced accumulation of ROS during stresses (Udawat et al., 2016). Because RBOHs require posttranslational modifications for their activation (Kobayashi et al., 2012), we propose that SlCipk6 phosphorylates the N-terminal regulatory domain of RbohB and SlRd2 directly or indirectly modulates this event, thus affecting ROS output. Because SlRd2 cancelled SlCipk6-dependent ROS, it might work in a negative feedback loop tempering the SlCipk6-RbohB signaling. If this was true, it will be important to determine whether SlRd2 inhibited SlCipk6 in a phosphorylation assay using MBP or RbohB as substrates. On the other hand, similar to UspA family members, the small GTP-binding proteins (Rac/Rop) have been postulated to act as molecular switches regulating a wide variety of important physiological functions in cells (Nibau et al., 2006; Xu et al., 2010). In this context, OsRac1 was required to activate OsRbohB in N. benthamiana cells (Wong et al., 2007). Interestingly, HRU1 has been found to interact with the GTPase ROP2 and RbohD, participating in the modulation of ROS levels under anoxia (Gonzali et al., 2015). Similarly, SlRd2 might act as a regulatory element of RbohB by affecting SlCipk6 activity. In the future, the molecular mechanism underlying SlRd2 regulation of SlCipk6-mediated ROS will be studied in higher detail.

MATERIALS AND METHODS

Bacteria, Yeast, and Plant Materials

Agrobacterium (Agrobacterium tumefaciens) strain C58C1 was grown at 30°C in Luria-Bertani (LB) medium with appropriate antibiotics. Yeast strains EGY48 (Matα trp1 his3 ura3 leu2::6LexAop-LEU2), BY4741 (Matα met15Δ0 his3Δ1 ura3Δ1 leu2Δ0) and GRF-167 (MATα his3Δ200 ura3167) were grown at 30°C in synthetic dropout medium with Glc as a carbon source. Escherichia coli strain W3101 [F-, galT22, λ-, IN (rrnD-rrnE)1, rph-1] and TN3151 [F-, λ-, IN (rrnD-rrnE)1, uspA1::kan, rph-1] were grown at 37°C in liquid morpholine propane sulphonic acid (MOPS) supplemented with Glc 0.4% (w/v). Nicotiana benthamiana) was grown in the greenhouse with 16 h of light and at 24°C (d) and 22°C (night). Tomatoes (Solanum lycopersicum) line Rio Grande-PtoR (Pto/Pto, Prf/Prf) was grown in hydroponic culture (Hoagland medium) in a growth chamber at the same growth conditions as N. benthamiana. The ORF of SlRd2 was cloned into the binary vector pGWB6 under the control of a cauliflower mosaic virus 35S promoter. Transgenic N. benthamiana plants were obtained according to the procedures of Rajput et al. (2014) and Park et al. (2013), and the homozygous transgenic lines 6 and 7 from T3 progeny were used (OE-6 and OE-7). Salt and osmotic shock treatments were performed adding NaCl or mannitol to final 100 mm or 300 mm concentration respectively to the media on 4-week-old tomato plants. For ROS detection and measurement, N. benthamiana leaf disks (0.28 cm²) were floated on 100 μL of distilled water in a 96-well white-bottom plate overnight at room temperature. Water was later replaced with 50 μL of distilled water and then incubated for 8 to 12 h at room temperature. For ROS detection, 50 μL of a 2× solution containing 100 μm luminol (Sigma-Aldrich) and 1 μg of horseradish peroxidase was quickly added to each well and ROS were measured in vivo as luminescence using a Varioskan Flash Multimode Reader.

SlRd2 Open Reading Frame cDNA Cloning and Deletion Mutant Generation

SlRd2 open reading frame was amplified from a tomato cDNA library prepared from leaf tissue by RT-PCR using primers OPS291 (5′-ATGGAAACGGTTATGGA-3′) / OPS292 (5′-TTAAATCACAGAGACTT-3′) and cloned into Gateway entry vector pDONR207 (Invitrogen). PCR-based site mutagenesis was performed to generate deletion of dimerization domain in SlRd2 using the QuikChange kit (Stratagene) using OPS405 (5′-CACAACTGTAAGATAGCACCGCCTGGAAAAGAAGCTGGGG-3′) and OPS406 (5′-CCCCAGCTTCTTTTCCAGGCGGTGCTATCTTACAGTTGTG-3′) primers.

Y2H Assay

Yeast strain EGY48 (containing pS18-34 vector) was used for Y2H assays. SlCipk6 cDNA and its mutant derivatives (de la Torre et al., 2013) were cloned into the bait vector pEG202, and SlRd2 cDNA was cloned into the prey vector pJG4-5. To generate ΔNterm mutant derivate in SlCipk6, a PCR reaction was performed using OPS501 (5′-ATGTTGAATGCTTTTCATATCATTTC-3′) and OPS165 (5′-CGGAATTCATGGGGACAGAAGAAAAATGTGC-3′) primers. Yeast transformation was performed by LiAc/PEG method as described in Yeast Protocols Handbook (Clontech). Growth and blue colonies on SD (X-Gal/Gal/-Ura,-His,-Trp,-Leu) plates indicated positive interaction. Finally, β-galactosidase assay was performed as described in Yeast Protocols Handbook. Expression of bait and prey fusion proteins was verified by immunoblotting using anti-LexA mouse monoclonal antibody (Santa Cruz Biotechnology) or anti-HA rat monoclonal antibody (Roche).

Quantitative Real-Time PCR

Total RNA was isolated from tomato leaves using TRIZOL reagent (Invitrogen) and subjected to DNAse treatment using TURBO DNA-free (Ambion). A quantity of 2 μg of total RNA was used to synthesize cDNA using random primers and Superscript II reverse transcriptase (Invitrogen) following the manufacturer’s protocol. Quantitative real-time (Q RT) PCR was performed with SsoFast EvaGreen Supermix (BIORAD) and a BIORAD real-time PCR system. The thermal cycle used was: 95°C for 10 min; 40 cycles of 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s. Quantitative real-time PCR reactions were carried out with the following oligonucleotides: OPS389 (5′-AGCAAACACGCTTTTGATTGGGC-3′)/OPS390 (5′-CACTGTCTTCACCATAGCAACC-3′) for SlRd2, OPS305 (5′-ATCCATGCACTTAATATCTTCC-3′) / OPS306 (5′-GCAATGATGGGTATCTGATAGCG-3′) for SlCipk6 and OPS281 (5′-AGCCACACAGTTCCCATCTAC-3′) / OPS282 (5′-AACTTCTCCTTCACTCCCTA-3′) for SlActin2 as an internal standard. Relative expression levels were determined as described previously in de la Torre et al. (2013).

Co-IP

For the co-IP experiment, SlRd2 and SlCipk6(T172D) coding sequences were cloned into pMDC43 (with a N-terminal GFP epitope) and pTAPa-pYL436 (with C-terminal 2xIgG-BD and 9xMyc tags) vectors, respectively, by gateway technology (Invitrogen; Rubio et al., 2005). Agrobacterium strains C58C1 carrying GFP-SlRd2 and SlCipk6(T172D)-Myc, respectively, were coinfiltrated into N. benthamiana leaves. Coinfiltration Agrobacterium C58C1 cultures carrying GFP and SlCipk6(T172D) were used as a negative control. Two d after, leaves were collected, frozen and grounded in liquid nitrogen and resuspended in three volumes of extraction buffer [50 mm Tris-HCl pH8, 0.1% NP-40, 1× Complete protease inhibitors (Roche)] and centrifuged at 14,000 g for 20 min at 4°C. Supernatant was filtered through two layers of Miracloth (Calbiochem). 2 mL was incubated with 50 μL IgG beads (Amersham Biosciences) for 2 h at 4°C with gentle rotation. Beads were washed five times with 2 mL of washing buffer (50 mm Tris-HCl pH 8, 0.1% NP-40). Elution from the IgG beads was performed by boiling the samples with 1× Laemmli buffer.

ATP-Binding Assay

For in vitro ATP binding assay, 1 μg of SlRd2 and GFP proteins purified from E. coli were incubated in binding buffer (20 mm Tris-HCl pH 7.5, 15 mm MgCl₂, 1 mm DTT), 50 μm ATP and 10 μCi of [α-³²P]ATP [Amersham; 3000 Ci/mmol (1 Ci = 37 GBq)] during 60 min at 30°C. Then, the reactions were stopped by adding 10 μL of 4× Laemmli loading buffer. The samples were denatured by boiling 5 min and later were run on 12% SDS-PAGE gel. ATP binding was visualized by autoradiography. Protein levels were verified by immunoblot using anti-His mouse monoclonal antibody (Roche). The in vivo ATP-binding assay was performed as described in Villamor et al. (2013). N. benthamiana leaf samples (1 g fresh weight) expressing GFP-SlRd2 were ground in liquid nitrogen and thawed in 2 volumes of extraction buffer (50 mm Tris pH 7.5). The lysate was later cleared via centrifugation and subjected to gel filtration using DG10 columns (Bio-Rad). Labeling was performed adding 10 mm MgCl₂ and 20 μm of BHAcATP (Thermo Scientific) to each sample and then incubated at room temperature for 1 h. For inhibition experiments, the lysate was incubated with 10 mm ATP for 30 min before labeling with BHAcATP. Biotinylated proteins were affinity purified by incubating the samples with Streptavidin beads (Thermo Scientific) for 1 h at room temperature. The beads were washed three times with 6 m urea. Finally, the purified proteins were boiled in 4× Laemmli buffer and analyzed by SDS-PAGE proteins gel and immunoblotted using anti-GFP mouse monoclonal antibody.

Cross-Linking Assay

The cross-linking assay was performed as described in Nachin et al. (2008). Fresh LB containing 50 μg/mL ampicillin was inoculated to a final OD₆₀₀ of 0.3 with an overnight culture of E. coli harboring pDEST17-SlRd2 construct and grown at 37°C. After 1 h, protein expression was induced by adding 1 mm isopropyl β-d-1-thiogalactopyranoside. After 4 h at 37°C, cells were harvested by centrifugation and resuspended in phosphate-buffered saline 1× to a final OD₆₀₀ of 0.7. Subsequently, 0.5 mm of cross-linking agent DSG was added to 200 μL of E. coli culture during 30 min at room temperature. The reaction was stopped by adding 40 μL of TS (200 mm Tris-HCl pH 8.8, 5 mm EDTA, 1 m Suc, and 0.05% (w/v) bromophenol blue)/TD (18% SDS and 0.3 m DTT; ratio 2:1). Finally, the samples were denatured by boiling 5 min and then were run on 12% SDS-PAGE gel. Protein M_r was determined by immunoblot using anti-His mouse monoclonal antibody (Roche).

In Vitro Phosphorylation Assays

For protein kinase assays, SlCipk6 full length was amplified by PCR using the primers OPS606 (5′-TCTAGACATGGGGACAGAAGAAAAATGT-3′) and OPS607 (5′-GTCGACCTCAAGCAATTGTTGGATTCTC-3′) and cloned as SalI/XbaI fragment in the yeast expression vector pEG(KT) (Mitchell et al., 1993), and then purified from yeast (strain GRF-167) using Glutathione Sepharose 4B affinity resin (GE Healthcare). SlRd2 cDNA was cloned into pET28a and purified from E. coli using a His column kit (GE Healthcare). SlRd2(UspA) was amplified by PCR using primers OPS675 (5′-AAAAAGCAGGCTCTATGGGCCGTGATATAGTGATC-3′) and OPS676 (5′-AGAAAGCTGGGTTTAAGGAACTATGATGACCGG-3′), cloned into pDEST15 vector (Invitrogen), and purified using Glutathione Sepharose 4B affinity beads. For kinase assays, 0.3 µg of SlCipk6 proteins and 2.5 µg of SlRd2, 2.5 µg of SlRd2(UspA) or 0.5 µg of MBP proteins were incubated in a final volume of 30 µL in kinase buffer (50 mm Tris-HCl pH 7.5, 2 mm MnCl₂, 2 mm DTT), 10 µM ATP and 10 μCi of [γ-³²P]ATP [Amersham; 3000 Ci/mmol (1 Ci = 37 GBq)] during 60 min at 30°C. The reactions were stopped by adding 10 μL of 4× Laemmli loading buffer. The samples were denatured by boiling 5 min and then were run on 12% SDS-PAGE gel. Kinase activity was visualized by autoradiography. Protein levels were verified by colloidal Coomassie Brilliant Blue G-250 staining.

BiFC Assay

For BiFC assays, SlRd2 and SlCipk6 coding sequences were cloned into pYFP^C (C-terminal YFP fragment) and pYFP^N (N-terminal YFP fragment) respectively by gateway technology (Invitrogen). Agrobacterium strains C58C1 carrying SlRd2-YFP^C and SlCipk6-YFP^N were coinfiltrated into N. benthamiana leaves. Agrobacterium cultures carrying YFP^C, YFP^N empty vectors and the mix SlRd2-YFP^C/SlCbl10-YFP^N were used as negative controls. Staining of N. benthamiana cells with FM4-64 was performed as described in Bolte et al. (2004). Fluorescence images were obtained 48 h after infiltration using a Leica TCS Sp2/DMRE confocal microscope with excitation wavelengths of 514 nm (YFP), 543 nm (FM4-64), and 440 nm (CFP). Transient expression of proteins in N. benthamiana leaves via agro-infiltration was performed as described in He et al. (2004).

Yeast Stress Tolerance Assays

For stress tolerance assays, yeast strain BY4741 harboring p426GPD EV and p426GPD-SlRd2 constructs were grown in liquid SD medium lacking Uracil (SD-Ura) containing 1% Glc (w/v) during 24 h at 30°C. Subsequently, they were diluted to the same concentrations (OD₆₀₀ = 10⁻¹, 10⁻², 10⁻³, and 10⁻⁴) and 10 μL of each dilution was spotted onto solid YPD medium supplemented with the different stress agents. Finally, yeast were grown at 30°C during 3 d and photographed.

Preparation and Purification of SlRd2 Antibody

SlRd2 open reading frame was cloned into pET-28a in frame with an N-terminal His-tag). The pET-28a-SlRd2 construct was transformed in BL21 (DE3) RIL (Stratagene) E. coli cells. Preparation of recombinant protein was performed as described in San-Miguel et al. (2013). Briefly, protein expression was induced at OD₆₀₀ = 0.5 by adding of 1 mm isopropyl β-d-1-thiogalactopyranoside to LB medium supplemented with 50 μg mL ampicillin. Cells were harvested by centrifugation at 3000 g for 10 min at room temperature and frozen overnight at −80°C. Preparation of His-tagged recombinant protein from pET-28a-SlRd2 was performed according to the manufacturer’s instructions (Qiagen). The antiserum was raised in rabbits using the full length SlRd2 as the antigen.

Oxidative Stress Complementation Assay in E. coli

The complementation assay was performed as described by Nyström and Neidhardt (1994). Fresh MOPS medium containing 50 μg/mL ampicillin was inoculated with overnight cultures of W3101 [F-, galT22, λ-, IN (rrnD-rrnE)1, rph-1] and TN3151 [F-, λ-, IN (rrnD-rrnE)1, uspA1::kan, rph-1] strains harboring pDEST17-SlRd2 and pDEST17-UspA constructs at an OD₆₀₀ of 0.1 and grown at 37°C until they reached an OD₆₀₀ = 0.7. A final concentration of 5 mm H₂O₂ was then added. 1 mL of each culture was harvested to perform the growth curve. Viability is expressed as the number of colony forming units at time divided by the number of colony forming units before the imposition of stress. Expression of SlRd2 and UspA proteins was determined by immunoblotting using anti-His mouse monoclonal antibody (Roche). mRNA UspA and 16S rRNA abundance were verified by RT-PCR using gene specific primers: OPS446 (5′-ATGGCTTATAAACACATTCTC-3′) and OPS448 (5′-TTATTCTTCTTCGTCGCGCAGC-3′) for UspA and OPS600 (5′-CTCCTACGGGAGGCAGCAG-3′) and OPS601 (5′-ATTACCGCGGCKGCTG-3′) for 16S rRNA.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: SlRd2 (KP843662), SlCipk6 (JF831200), SlCipk11 (JF831201), and SlCipk14 (JF831202).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. SlRd2, SlCipk6, and SlCipk6 mutant derivative proteins are expressed in yeast.

• Supplemental Figure S2. Specificity of SlCipk6/SlRd2 interaction.

• Supplemental Figure S3. Both SlCipk6 and SlRd2 are localized in the cytosol and in the nucleus.

• Supplemental Figure S4. SlRd2 protein sequence alignment.

• Supplemental Figure S5. SlRd2 Usp domain protein sequence alignment.

• Supplemental Figure S6. SlCipk6-Myc (wild type), SlCipk6(T172D)-Myc (T172D), and SlCipk6(K43M)-Myc (K43M) are expressed in N. benthamiana GFP-SlRd2 overexpressing (OE-6 and OE-7) and wild-type plants.

• Supplemental Table S1. Summary of the Y2H screen.