AtCPL5, a novel Ser-2-specific RNA polymerase II C-terminal domain phosphatase, positively regulates ABA and drought responses in Arabidopsis

Authors

  • Yong-Mei Jin,

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    Search for more papers by this author
  • Jinwook Jung,

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    Search for more papers by this author
  • Hyesung Jeon,

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    Search for more papers by this author
  • So Youn Won,

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    2. Rural Development Administration, Suwon, 441-707, Republic of Korea
    Search for more papers by this author
  • Yue Feng,

    1. Faculty of Molecular and Environmental Plant Science, Vegetable and Fruit Improvement Center, Department of Horticultural Sciences, Texas A&M University, College Station, TX 77843-2133, USA
    Search for more papers by this author
  • Jae-Sook Kang,

    1. Division of Applied Life Science (BK21 Program), Gyeongsang National University, Jinju, 660-701, Republic of Korea
    Search for more papers by this author
  • Sang Yeol Lee,

    1. Division of Applied Life Science (BK21 Program), Gyeongsang National University, Jinju, 660-701, Republic of Korea
    Search for more papers by this author
  • Jong-Joo Cheong,

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    Search for more papers by this author
  • Hisashi Koiwa,

    1. Faculty of Molecular and Environmental Plant Science, Vegetable and Fruit Improvement Center, Department of Horticultural Sciences, Texas A&M University, College Station, TX 77843-2133, USA
    Search for more papers by this author
  • Minkyun Kim

    1. Department of Agricultural Biotechnology, Center for Agricultural Biomaterials, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Republic of Korea
    Search for more papers by this author

Author for correspondence:
Minkyun Kim
Tel: 82 2 880 4641
Email: mkkim3@snu.ac.kr

Summary

  • Arabidopsis RNA polymerase II (RNAPII) C-terminal domain (CTD) phosphatases regulate stress-responsive gene expression and plant development via the dephosphorylation of serine (Ser) residues of the CTD. Some of these phosphatases (CTD phosphatase-like 1 (CPL1) to CPL3) negatively regulate ABA and stress responses. Here, we isolated AtCPL5, a cDNA encoding a protein containing two CTD phosphatase domains (CPDs).
  • To characterize AtCPL5, we analyzed the gene expression patterns and subcellular protein localization, investigated various phenotypes of AtCPL5-overexpressors and knockout mutants involved in ABA and drought responses, performed microarray and RNA hybridization analyses using AtCPL5-overexpressors, and assessed the CTD phosphatase activities of the purified AtCPL5 and each CPD of the protein.
  • Transcripts of the nucleus-localized AtCPL5 were induced by ABA and drought. AtCPL5-overexpressors exhibited ABA-hypersensitive phenotypes (increased inhibition of seed germination, seedling growth, and stomatal aperture), lower transpiration rates upon dehydration, and enhanced drought tolerance, while the knockout mutants showed weak ABA hyposensitivity. AtCPL5 overexpression changed the expression of numerous genes, including those involved in ABA-mediated responses. In contrast to Ser-5-specific phosphatase activity of the negative stress response regulators, purified AtCPL5 and each CPD of the protein specifically dephosphorylated Ser-2 in RNAPII CTD.
  • We conclude that AtCPL5 is a unique CPL family protein that positively regulates ABA-mediated development and drought responses in Arabidopsis.

Introduction

Environmental variables are critical factors controlling plant growth and development, and plants have evolved to acquire adaptive strategies to cope with adverse environmental conditions such as drought, cold and salt stresses (Knight et al., 1995; Fowler et al., 2005; Shi et al., 2005; Zhu et al., 2007; Miura & Hasegawa, 2008; Rensing et al., 2008; Nakashima et al., 2009). Abscisic acid (ABA) has been shown to be an important hormone regulator that coordinates plant developmental programs with environmental stress responses (Goode et al., 1992; Bopp & Werner, 1993). It has a wide range of important roles in plant growth and development, including embryogenesis, seed maturation and dormancy, and root and shoot growth, as well as tolerance to various environmental stresses (Finkelstein et al., 2002a; Zhang et al., 2006). For example, increased ABA biosynthesis upon drought stress in plants promotes stomatal closure, reducing transpirational water loss, and induces several drought- and ABA-responsive genes (Marion-Poll & Leung, 2006; Xiong et al., 2006; Yamaguchi-Shinozaki & Shinozaki, 2006; Wu, 2008).

Complex genetic regulatory networks for ABA or drought stress signaling in higher plants are interconnected with other signaling pathways controlled by light, sugars, and other plant hormones (Fedoroff, 2002; Finkelstein & Gibson, 2002b; Gazzarrini & McCourt, 2003; Zhang et al., 2004; Li et al., 2006; Feurtado & Kermode, 2007; Chen et al., 2008; Gong et al., 2008; Nakashima et al., 2009). A variety of proteins including protein kinases, phosphatases, and transcription factors are involved in ABA- and drought-responsive signaling pathways (Rodriguez, 1998; Merlot et al., 2001; Kwak et al., 2002; Yoshida et al., 2002; Osakabe et al., 2005; Hu et al., 2006; Kuhn et al., 2006; Chae et al., 2007). Expression of the ABA- or drought-responsive genes most often involves the interaction of cis-elements in their promoter regions, including ABRE, DRE, MYBR, MYCR, NACR and HDZFR, with their cognate transcription factors, thus forming a complex with RNA polymerase II (RNAPII) (for reviews, see Kim, 2006; Nakashima et al., 2009).

RNAPII is a core component of the transcription complex that catalyzes mRNA synthesis and is also involved in the regulation of various mRNA maturation processes such as capping, splicing, and polyadenylation (Hirose & Manley, 2000). Reversible phosphorylation of the carboxyl-terminal domain (CTD) of the largest RNA polymerase II subunit, which is composed of Y1S2P3T4S5P6S7 heptapeptide repeats, plays an important role in the regulation of gene expression. During the transcription process, CTD kinases and phosphatases serve as transcriptional activators or repressors (depending on the point in the transcriptional cycle) by altering the state of CTD phosphorylation at serine (Ser)-2 and Ser-5 (Koiwa et al., 2002, 2004; Lin et al., 2002; Xiong et al., 2002; Ueda et al., 2008). Recently, the RNAPII CTD phosphatase-like (CPL) genes CPL1/FRY2, CPL2, CPL3 and CPL4 were shown to be involved in stress responses and development in Arabidopsis (Koiwa et al., 2002, 2004; Xiong et al., 2002; Ueda et al., 2008; Matsuda et al., 2009). CPL1/FRY2, CPL2 and CPL3 are negative regulators of stress-responsive gene transcription as well as of osmotic stress, ABA, and wound responses (Xiong et al., 2002; Ueda et al., 2008; Matsuda et al., 2009). CPL2 is also known to be involved in auxin responses (Ueda et al., 2008), while CPL4 is necessary for normal plant growth (Bang et al., 2006).

The Arabidopsis RNAPII CTD phosphatase family is composed of > 20 members that are categorized into three groups based on protein-domain architecture similarity to other eukaryotic proteins (Koiwa et al., 2002; Koiwa, 2006). CPL1 and CPL2, belonging to the group I CPLs, which have an RNAPII CTD phosphatase domain (CPD) and double-stranded RNA-binding motifs (DRMs), specifically dephosphorylate Ser-5 of the CTD heptad of RNAPII in Arabidopsis (Koiwa et al., 2004; Koiwa, 2006). CPL3 and CPL4, belonging to the group II CPLs, which have a CPD and a breast cancer 1 (BRCA1) C-terminal (BRCT) domain, interact with AtRAP74 (Arabidopsis thaliana RNA polymerase II-associated protein), and this interaction is mediated by the BRCT domain in vitro (Yu et al., 2003; Koiwa et al., 2004; Bang et al., 2006; Koiwa, 2006). Recently, a family of small CTD phosphatases (SCPs) containing the Fcp1 (TFIIF-associating component of CTD phosphatase) homology (FCPH) domain, but not the BRCT domain, which exhibited Ser-5 phosphatase activity, was identified in human cells (Yeo et al., 2005). However, none of the plant proteins belonging to the group III CPLs, which only have CPD(s) without any other known functional domains, have been characterized.

The AtCPL5 gene, encoding a novel RNAPII CTD phosphatase that belongs to the group III CPL family, was cloned and characterized in this study. The encoded protein was unique in Arabidopsis, possessing two highly conserved CPDs. AtCPL5 was demonstrated to positively regulate ABA-mediated development and drought responses in Arabidopsis.

Materials and Methods

Plant material, growth conditions, and plant stress treatments

Arabidopsis (Arabidopsis thaliana (L.) Heynh) seeds of the wild type (WT) (Columbia-0) and the AtCPL5 T-DNA insertion mutant line GABI-Kat_532E08 (obtained from the Max Planck Institute for Plant Breeding Research, Cologne, Germany) were used in this study (Li et al., 2003). The homozygous GABI-Kat_532E08 mutant was screened by PCR-based genotyping using the primers shown in Supporting Information Table S1.

For plant growth, sterilized seeds were plated on half-strength MS medium (Murashige & Skoog, 1962) supplemented with 1% sucrose and 0.05% MES-KOH (pH 5.7). Plated seeds were stratified for 4 d in darkness at 4°C to break dormancy and transferred to a growth chamber at 23°C under long-day conditions (16 h : 8 h, light : dark cycle; light intensity 100 μE m−2 s−1).

To test AtCPL5 induction by stress treatments, 2-wk-old Arabidopsis seedlings grown on half-strength MS plates were treated with 0.1 mM ABA, 300 mM NaCl, cold (7°C under dim light), drought (dehydrated at 22°C; 60% humidity; dim light), 400 mM mannitol, and 1 mM salicylic acid, respectively.

AtCPL5 cDNA cloning, 35S:AtCPL5 construction, and Arabidopsis transformation

Total RNA was isolated from WT Arabidopsis plants grown under normal growth conditions using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). The full-length open reading frame of AtCPL5 cDNA was obtained by reverse transcription reaction using Superscript III reverse transcriptase (Invitrogen, Carlsbad, CA, USA) with the AtCPL5-specific reverse primer 5′-TCATTCTACAGATTCTACGTTGAGTTC-3′ (stop codon in italics). The PCR amplification of the full-length cDNA was performed using the forward primer 5′-TTTGTAGCCAAAAATCTTTCTCC-3′ and the reverse primer 5′-CGGCTAGCTTCTACAGATTCTACGTT-3′ (NheI site underlined; stop codon deleted) and cloned into the GUS-removed pCAMBIA1301 (MRC, Cambridge, UK). The resulting recombinant plasmid (p35S:AtCPL5) was transformed into Arabidopsis plants by the floral dip method (Clough & Bent, 1998) using Agrobacterium tumefaciens C58C1. The transformants were selected based on their hygromycin resistance segregation ratio and DNA and RNA hybridization analyses, as described by Jung et al. (2007).

Quantitative real-time RT-PCR

Total RNA was extracted from 2-wk-old Arabidopsis seedlings using the RNeasy Plant Mini Kit. Two micrograms of each RNA sample was reverse-transcribed to cDNA using the High Capacity RNA to cDNA kit (Applied Biosystems, Foster City, CA, USA). The resulting cDNA was pre-amplified with AtCPL5-specific primers for 14 cycles using the TaqMan PreAmp Master Mix Kit (Applied Biosystems). AtCPL5-specific primers were designed corresponding to the two exon regions flanking the internal intron (Table S2).

Quantitative real-time RT-PCR was performed using 5 μl of 5-fold diluted pre-amplified PCR products, TaqMan Gene Expression Master Mix (Applied Biosystems), and the AtCPL5-specific primers mentioned above, along with the AtCPL5-specific fluorescent probe annealing to the exon–exon boundaries located further inside the primer-annealing regions. 18S rRNA-specific primers and a fluorescent probe were also added for the internal control reaction. The probes and primers are shown in Table S2. All experiments were carried out in two biological replicates and three technical replicates for each sample. The relative quantification method (ΔΔCT) was used to evaluate the quantitative variation among replicates (Livak & Schmittgen, 2001).

Histochemical β-glucuronidase (GUS) assay

For the PAtCPL5::GUS construct, the 3955-bp DNA fragment located upstream from the initiation codon of the AtCPL5 was inserted into the pCAMBIA1391 binary vector (MRC). GUS staining was performed as previously described (Hemerly et al., 1993).

Seed germination, green cotyledon emergence, and root elongation in response to ABA

For seed germination assays, 30 seeds of each individual line were plated on half-strength MS medium containing different concentrations of ABA. After being stratified at 4°C for 4 d, the plates were transferred to long-day conditions (16 h : 8 h, light : dark) at 23°C. If seeds had a radicle length of ≥ 1 mm on the 5th day after the transfer, they were considered to have germinated. Experiments were performed in triplicate.

For measurements of green cotyledon emergence rates, 60 seeds of each individual line were plated on half-strength MS medium containing 0 or 1.0 μM ABA. Green cotyledons were scored on the 10th day after stratification for 4 d. Experiments were performed in duplicate.

For root elongation assays, seeds were germinated on an ABA-free medium for 6 d and well-grown seedlings were transferred to MS plates containing 0 or 4.0 μM ABA. Experiments were performed in quadruplicate.

Measurements of stomatal apertures and transpiration rates, and survival rates after drought stress treatment

To measure stomatal apertures, stomata were fully opened before ABA treatment. For this purpose, rosette leaves of 4-wk-old, soil-grown plants were detached and floated (abaxial side down) on assay buffer (10 mM MES-KOH (pH 6.1), 30 mM KCl, and 1 mM CaCl2), and then treated with different concentrations of ABA. Abaxial epidermal strips were peeled immediately, and stomatal apertures (pore width) were measured using an Axiophot (Carl Zeiss Jena GmbH, Jena, Germany) microscope coupled to a CCD camera. Sixty stomatal apertures per line were measured in duplicate.

For transpiration rate (water loss) measurements, aerial parts of 4-wk-old plants grown in soil were separated from the roots, placed abaxial side up on open weighing Petri dishes, and allowed to dry on the laboratory bench. The experiments were repeated three times.

For drought stress treatment, irrigation of 24-d-old Arabidopsis plants grown in soil was stopped for 12 d and then the plants were rewatered for 4 d. At least eight plants of each individual line were measured for survival rates in each experiment, each of which was performed four times.

Microarray analyses

Two independent biological replicates of microarray experiments were performed using 2-wk-old WT and AtCPL5-overexpressing plants without any treatment. Total RNA was isolated using the RNeasy Plant Mini Kit. Five micrograms of total RNA was used to generate cDNA using the GeneChip one-cycle cDNA synthesis kit (Affymetrix Inc., Santa Clara, CA, USA) and a T7 promoter-containing oligo(dT) primer. Synthesized cDNA was in vitro transcribed and biotinylated using the GeneChip IVT labeling kit (Affymetrix Inc.). Resulting biotin-labeled cRNA was fragmented and hybridized to Affymetrix ATH1 genome arrays using an Affymetrix Fluidics station 450 according to the manufacturer’s instructions. GeneChip hybridization data were generated by Microarray Suite 5.0 software (Affymetrix Inc.) and normalized using the quantile method (Bolstad et al., 2003). Genes exhibiting > 1.5-fold enhanced or reduced transcript levels in both biological replicates of 35S:AtCPL5 transgenic plants compared with WT were considered to show alterations in expression.

RT-PCR and RNA hybridization analyses

RT-PCR was carried out on 1 μg of RNA for 37 cycles using a one-step RT-PCR kit (Qiagen) and AtCPL5-specific primers (Table S3) spanning an intronic sequence in AtCPL5. For RNA hybridization analyses, 15–20 μg of total RNA was separated by electrophoresis on a 1% formamide/agarose gel and blotted onto Hybond-N+ nylon membranes (Amersham Biosciences, Buckinghamshire, UK). Hybridization was performed by adding various [α-32P] dCTP-labeled DNA probes for AtCPL5 and some stress-responsive genes from Arabidopsis. Primer sets for PCR products used as the DNA probes are shown in Table S3.

Recombinant protein expression and purification

For the expression of the full-length protein and individual CPDs of AtCPL5 recombinant proteins, cDNAs corresponding to the full-length protein (1806 bp), CPD1 (900 bp), and CPD2 (906 bp) were PCR-amplified using the appropriate primer sets (Table S4) and inserted into the pET-30a(+) vector (Novagen, Darmstadt, Germany). The recombinant proteins were expressed in Escherichia coli BL21 (DE3) Codon Plus-RIL (0.2 mM isopropyl-β-D-thiogalactoside; 18°C; 20 h). The purified recombinant protein was detected by immunoblotting using the His-tag antibody (Novagen).

For recombinant GST-AtCTD expression, cDNA encoding the CTD coding region of the largest subunit of Arabidopsis RNAPII (At4g35800) (Nawrath et al., 1990; Umeda et al., 1998) was PCR-amplified and introduced into the pGEX-5X-1 GST-fusion vector (Pharmacia, Uppsala, Sweden). The GST-AtCTD fusion protein was expressed in E. coli BL21 (DE3).

Assay for RNAPII CTD phosphatase activity

Purified recombinant GST-AtCTD was first phosphorylated by Cdc2 protein kinase (New England Biolabs, Ipswich, MA, USA). Dephosphorylation was then performed in a mixture (30 μl) containing 50 mM Tris-acetate buffer (pH 4.0), 3 μg of phosphorylated GST-AtCTD, 10 mM MgCl2, and 3 μg of purified recombinant protein (His6-Full, His6-CPD1 and His6-CPD2, respectively) at 37°C for 6 h (Koiwa et al., 2004; Zheng et al., 2005). After dephosphorylation, protein samples were subjected to 10% SDS-PAGE followed by blotting to Hybond-ECL nitrocellulose membranes (Amersham Biosciences). The membranes were blocked in 2% fatty acid-free bovine serum albumin (BSA; Sigma, St Louis, MO, USA), and then incubated with H5 and H14 monoclonal antibodies (Covance, Princeton, NJ, USA), followed by anti-mouse secondary antibody IgG-HRP. Finally, the membrane was visualized using the luminescent image analyzer LAS-3000 (Fuji Photo Film, Tokyo, Japan).

Statistical analysis

The t-tests (Figs 5, 6) were performed using the statistix for Windows 2.2 program (Analytical Software, Tallahassee, FL, USA).

Results

Cloning of AtCPL5 cDNA and analysis of the encoded protein

BLAST analysis of the four known CPLs, CPL1–CPL4, revealed that 24 CPD proteins may be encoded by the Arabidopsis genome. Further bioinformatics analyses using the Conserved Domain Architecture Retrieval Tool (CDART) showed that 20 of these proteins have CPD(s) located downstream of the conserved DXDX(T/V) acylphosphatase motif. This DXDX(T/V) motif is an intermediate phosphoryl acceptor endowed with phosphohydrolase activity, which is conserved in many acid phosphatases from bacteria and other eukaryotes (Collet et al., 1998; Thaller et al., 1998). Alternatively, the other four proteins have CPDs that are not associated with the DXDX(T/V) motif, indicating that they may not have phosphohydrolase activity.

Among the 20 proteins containing both CPD(s) and DXDX(T/V) motifs, 15 proteins have no other known functional domains, whereas five proteins (CPL1–CPL4, and the one encoded by At4g06599) have additional domains such as the DRM, BRCT (BRCA1 C-terminus), or ubiquitin processing protease (UBP). The protein encoded by the transcript NM_112850.1 is unique among these 15 proteins as it has two CPDs (CPD1, amino acids 113–243; CPD2, amino acids 410–538), which were previously designated small SCP1-like phosphatase 7 (SSP7) and SSP8, respectively (Koiwa, 2006).

The full-length cDNA for the transcript NM_112850.1 (1806 bp in size) was obtained by two-step RT-PCR using total RNA extracted from 2-wk-old seedlings. The same cDNA was also cloned independently by others (http://www.genscript.com/product_001/gene/code/GN115476/category/gene/Arabidopsis_thaliana_unknown_protein_AT3G19600_mRNA_complete_cds.html). The deduced amino acid sequence of the cloned cDNA (GenBank accession number FJ773993), designated Arabidopsis thaliana C-terminal domain phosphatase-like 5 (AtCPL5), was identical to that registered in the GenBank database (NP_188594.1; note that the originally annotated gene At3g19600 corresponding to AtCPL5 was computationally split into At3g19600.1 and At3g19595.1 in March 2008 by The Arabidopsis Information Resource). AtCPL5 consists of 601 amino acids with a predicted molecular mass of 69 kD and a pI of 7.0. In addition to the two CPDs, it has two DXDX(T/V) acylphosphatase motifs (motif 1, amino acids 94–98; motif 2, amino acids 391–395) in the vicinity of each CPD (Fig. 1a).

Figure 1.

 Analysis of C-terminal domain (CTD) phosphatase-like 5 (AtCPL5) protein architecture and its relationship to other CTD phosphatase domain (CPD)-containing proteins in Arabidopsis. (a) Schematic representation of the 601 amino acids of AtCPL5. Two RNA polymerase II (RNAPII) CPDs (CPD1 and CPD2) and two DXDX(T/V) motifs are indicated. (b) Phylogenetic analysis of CPD-containing proteins in Arabidopsis. A rooted phylogenic tree was drawn using CLUSTALW (1.81) in Biology WorkBench 3.2. AtCPL5 was divided into two polypeptides consisting of amino acid residues 1–300 and 301–601 for the analysis. CPD-containing proteins were classified into three groups (I–III) based on additional domains besides the CPD: groups I, II and III contain a double-stranded RNA-binding motif (DRM), a breast cancer 1 (BRCA1) C-terminal (BRCT) domain, and no other known functional domains in addition to CPD, respectively [correction added after online publication 19 January 2011: in Fig. 1(b), the proteins At3g55960 (SSP6) and At5g46410 (SSP4) were relabelled as Group III instead of Group II; additionally the protein At5g11860 (SSP5) was relabelled as Group III instead of Group I]. (c) Multiple alignment of CPD-containing proteins was performed using CLUSTALW (1.81) with amino acid sequences of AtCPL5 CPD1 (aa 1–300), AtCPL5 CPD2 (aa 301–601), SSP10, and SSP9. Identical amino acid residues are indicated in black, and conserved DXDX(T/V) motif sequences are shown in the box.

Phylogenetic analysis of two individual polypeptide regions containing each of the CPDs in AtCPL5 (amino acids 1–300 and 301–601, respectively) with 23 CPD-containing proteins was performed using CLUSTAL W (1.81) in Biology WorkBench 3.2 (http://workbench.sdsc.edu). The two regions showed the highest homology to SSP10 and SSP9, the group III CPL family proteins that only contain CPD without any other known functional domains (Fig. 1b). Amino acid sequence alignment revealed a 64% amino acid identity between the two polypeptide regions. In addition, they showed 58% and 71% identity with SSP10, and 57% and 67% identity with SSP9, respectively (Fig. 1c). Thus, AtCPL5 is a novel gene encoding a protein containing two tandem-repeated CPDs that belongs to the group III CPL family.

Expression of AtCPL5 in response to various treatments

To study the possible involvement of AtCPL5 in environmental stress responses, we first tested the regulation of AtCPL5 by environmental stresses. For this purpose, 2-wk-old Arabidopsis seedlings were subjected to treatments with phytohormone ABA or other stressors, including NaCl, drought, cold, mannitol, and salicylic acid. Quantitative real-time RT-PCR analyses showed that AtCPL5 was induced by ABA, NaCl, drought, and cold stress (Fig. 2). However, the induced accumulation of AtCPL5 transcripts, which showed different kinetic patterns among the treatments, appeared to be transient except in the case of the NaCl treatment. In other words, 0.1 mM ABA and drought treatments caused the transcript level to increase within 3 h, reach its peak (c. 5–6 fold) at 6 h, and then decline; in the case of cold treatment, the level increased c. 6-fold within 3 h and then rapidly declined; in the 300 mM NaCl treatment, however, it increased gradually up to c. 4-fold within 12 h. The expression of AtCPL5 did not alter when subjected to mannitol or salicylic acid (data not shown).

Figure 2.

 Quantitative real-time RT-PCR analyses of C-terminal domain phosphatase-like 5 (AtCPL5) expression in response to various stress treatments. 18S rRNA was used as an internal control. Data are the mean ± SE of two biological replicates.

Analysis of spatial expression of AtCPL5 with histochemical GUS staining

Spatial expression of AtCPL5 was examined in the promoter-GUS assay using PAtCPL5:GUS transgenic plants. The promoter region containing 3955 bp upstream of the start codon was used to drive the expression of GUS. In the vegetative stage, GUS activity was first detected in the seed coat 1 d after germination and in the hypocotyl of 3-d-old seedlings (Fig. 3a,b). GUS staining was also observed in roots, hypocotyl, rosette leaves, and cotyledons in 2-wk-old seedlings (Fig. 3c). GUS staining was more pronounced in primary than in secondary roots (Fig. 3d). In rosette leaves, GUS staining was observed in the vascular tissue of major veins, as well as guard cells and trichomes (Fig. 3e–h). During the reproductive stage, GUS staining was detected in flower buds, stems, stamens, carpels, and funiculi of siliques (Fig. 3i–k). Notably, a marked increase in GUS activity was observed in rosette leaves after ABA treatment (Fig. 3l,m), suggesting that AtCPL5 is involved in ABA-mediated plant responses in leaves.

Figure 3.

 Histochemical GUS staining of PAtCPL5:GUS transgenic plants. GUS expression was determined in vegetative and reproductive tissues: in the seed coat of a 1-d-old germinating seedling (a), the hypocotyl of a 3-d-old seedling (b), a 2-wk-old seedling (c), the root of a 2-wk-old seedling (d), a rosette leaf (e), guard cells (f), trichomes in rosette leaves (g, h), flower buds and stems (i), stamens (anthers and pollen tubes) and pistils (stigmas and styles) (j), funiculi in siliques (k), and a 7-d-old seedling treated with H2O (l) or 0.1 mM ABA (m) for 3 h. Scale bars in (a), (b), (d), (h) and (j) represent 100 μm; in (c), (i), (l) and (m), 1 mm; in (e), 0.5 mm; in (f), 20 μm; and in (g) and (k), 0.2 mm.

Targeting of AtCPL5 to the nucleus

To determine the subcellular localization of AtCPL5, the full-length cDNA of AtCPL5 was fused to the 5′-end of the soluble modified green fluorescent protein (smGFP) gene under the control of the cauliflower mosaic virus (CaMV) 35S promoter and introduced into onion epidermal cells by particle bombardment. As shown in Fig. 4, the chimeric protein, AtCPL5::smGFP, was localized predominantly to the nucleus of onion cells, whereas smGFP alone was distributed in both the cytosol and the nucleus.

Figure 4.

 Subcellular localization of the AtCPL5::smGFP fusion protein. A full-length cDNA of C-terminal domain phosphatase-like 5 (AtCPL5) was fused to the 5′-end of the smGFP reporter gene under the control of the 35S cauliflower mosaic virus (CaMV) promoter; the chimeric vector was introduced into onion epidermal cells by particle bombardment. After incubation for c. 20 h (23°C, in the dark), the bombarded specimen was treated with DAPI solution (3 μg ml−1) for at least 30 min. Photographs were taken using appropriate filters under a confocal laser scanning microscope.

AtCPL5-overexpressing transgenic plants differentially respond to exogenous ABA

The biological function of AtCPL5 was analyzed using a knockout mutant with a T-DNA insertion and transgenic Arabidopsis plants overexpressing AtCPL5. A T-DNA was known to be inserted 275 bp downstream from the start codon of the first AtCPL5 exon in the GABI-Kat 532E08 line (Fig. S1a). A homozygous mutant selected from the line by PCR-based genotyping was designated cpl5. RT-PCR and RNA hybridization analyses confirmed that AtCPL5 mRNA was absent in cpl5 (Fig. S1b,c). In addition, transgenic plants overexpressing AtCPL5 under the CaMV 35S promoter were also generated. Two independent T3 homozygous overexpression lines that contain a single copy of AtCPL5, designated OE#2 and OE#5, were obtained from 16 independent transformant lines based on the segregation ratio of hygromycin resistance and Southern hybridization analyses (Fig. S1d). RT-PCR and RNA hybridization results showed that AtCPL5 transcripts were highly expressed in OE#2 and OE#5 compared with WT plants (Fig. S1b,c). Neither the cpl5 mutant nor the overexpressor plants OE#2 and OE#5 were significantly different from the WT in their morphologies throughout the vegetative and reproductive growth stages (data not shown).

To examine the biological function of AtCPL5 in response to ABA in Arabidopsis, the ABA sensitivities of the cpl5 knockout and OE#2 and OE#5 overexpressing plants were investigated at different stages of development such as seed germination, green cotyledon emergence, and primary root elongation. In ABA-free medium, all of the lines tested showed almost the same seed germination rates (> 99%) and green cotyledon emergence rates (98–99%), and similar primary root elongation (74–76 mm) (Fig. 5). However, in ABA-containing medium (0.5–1.5 μM), the germination rates of OE#2 and OE#5 were significantly lower (< 0.05) than that of the WT; for example, at 1.0 μM ABA, OE#2 and OE#5 seeds showed 69% and 72% germination rates, respectively, while the WT had a germination rate of 87%. The cpl5 mutant showed similar germination rates compared with the WT in all ABA-containing media (Fig. 5a). A similar hypersensitivity of OE#2 and OE#5 and a marginal hyposensitivity of cpl5 to ABA were also observed during the early stages of seedling development. For example, on the 10th day after sowing seeds in 1.0 μM ABA-containing medium, OE#2 and OE#5 showed green cotyledon emergence rates of 32% and 29%, respectively; meanwhile, cpl5 and WT had rates of 68% and 61%, respectively (Fig. 5b). In addition, the inhibition of primary root elongation by ABA treatment was compared among the lines. For this purpose, seeds were germinated in ABA-free medium and grown in the absence or presence of 4 μM ABA for 10 d. OE#2 and OE#5 showed significant decreases (37% and 41%, respectively) in primary root elongation upon exposure to ABA, while WT and cpl5 seedlings showed decreases of only 19% and 18%, respectively (Fig. 5c). Taking these results together, overexpression of AtCPL5 appeared to confer ABA hypersensitivity in seed germination, green cotyledon emergence, and primary root elongation of Arabidopsis.

Figure 5.

 Effects of ABA treatment on the development and growth of C-terminal domain phosphatase-like 5 (cpl5), the wild type (WT), and AtCPL5-overexpressing transgenic plants (OE#2 and OE#5), respectively. (a) Germination rates. Seeds were germinated and grown on MS medium containing different concentrations of ABA. Data are the mean ± SD (= 30) of three independent experiments. (b) Green cotyledon emergence rates on ABA-free and 1.0 μM ABA media, respectively. Photographs were taken on the 10th day after stratification for 4 d. Data are the mean ± SD (= 60) of two independent experiments. (c) Primary root elongation. Seeds were germinated on ABA-free MS medium, and then transferred to medium containing 0 or 4.0 μM ABA. The newly grown primary root length was measured on the 10th day after transfer. Data are the mean ± SD (= 8) of four independent experiments.*, significantly different from the WT at P < 0.05. **, significantly different from the WT at P < 0.01.

Response of AtCPL5 knockout and overexpressing transgenic plants to drought stress

The ABA induction of AtCPL5 (Fig. 2) and its promoter activity in guard cells (Fig. 3) suggested that AtCPL5 may be involved in the drought tolerance response in Arabidopsis. To test this possibility, stomatal responses to ABA treatment, transpiration rates upon dehydration, and survival rates upon drought stress in loss-of-function and gain-of-function plants were analyzed. First, 10 μM ABA treatment of the AtCPL5-overexpressing plants, OE#2 and OE#5, induced dramatic decreases (69% and 60%, respectively) in stomatal apertures; WT plants showed a 40% decrease when given the same treatment. ABA-treated cpl5 showed a smaller decrease (33%) in stomatal aperture (Fig. 6a). Consistent with these results, the transpiration rates of the rosette leaves, measured based on the loss of fresh weight of the leaves after detachment from the main plant bodies, were significantly lower (< 0.05) in OE#2 and OE#5 than in the WT; for example, at c. 3.5 h after the start of dehydration, OE#2 and OE#5 lost less rosette leaf fresh weight (34% and 37%, respectively) compared with the WT (46%). The dehydration treatment resulted in similar losses in rosette leaf fresh weights in cpl5 and WT (Fig. 6b).

Figure 6.

 Stomatal apertures, transpiration rates, and drought stress responses of C-terminal domain phosphatase-like 5 (cpl5), the wild type (WT), and OE#2 and OE#5, respectively. (a) Stomatal apertures. Rosette leaves of 4-wk-old plants were floated on assay buffer and treated with 10 μM ABA for 2 h. Stomatal apertures (pore width) were measured under a microscope. Data are the mean ± SD (= 60) of two independent experiments. (b) Transpiration rates. Water loss was measured by placing the abaxial side of rosette leaves obtained from four individual plants upward on open weighing Petri dishes, and then allowing them to dry on the laboratory bench. Data are the mean ± SD of three independent experiments. (c) Drought stress responses. Arabidopsis plants grown in soil for 24 d were not watered for 12 d (Drought) and then rewatered for 4 d (Rewatering). Survival rates after drought stress treatment were calculated. Data represent the mean ± SD (= 8) of four independent experiments. *, significantly different from the WT at P < 0.05. **, significantly different from the WT at P < 0.01.

Plants were grown in well-irrigated soil for 24 d, and then irrigation was stopped for 12 d to compare their survival rates upon drought stress. The leaves of all plants tested began to wilt, showing an accumulation of anthocyanins, 7 d after withholding irrigation. On the 12th day, most of the WT plants were wilted. However, on the 4th day after rewatering, OE#2 and OE#5 overexpressors and cpl5 knockout plants showed significantly higher (P < 0.01) or lower (< 0.05) survival rates compared with WT plants: the survival rates of cpl5, WT, and OE#2 and OE#5 were 42, 50, 80 and 88%, respectively (Fig. 6c). All of these results indicate that AtCPL5 is likely to be involved in drought stress tolerance. Neither the AtCPL5 overexpressors nor the cpl5 mutant plants stressed with NaCl or cold, causing the induction of AtCPL5 (Fig. 2), showed phenotypic differences compared with WT plants treated with the same stresses (data not shown).

Activation of ABA- and drought stress-responsive genes by AtCPL5 overexpression in Arabidopsis

To test whether the change in AtCPL5 expression perturbs the accumulation of ABA- or drought stress-responsive genes, microarray experiments were performed to compare the transcriptome profiles of AtCPL5-overexpressing plants (line: OE#5) with those of the WT; cpl5 mutants were not used for this purpose because of their weak phenotypes in response to ABA and drought treatments. Two biological replicates of 2-wk-old plants were used for the experiments. Hybridization was conducted using Arabidopsis Affymetrix ATH1 genome arrays which contain > 22 500 probe sets, representing c. 24 000 genes. Microarray data were deposited in ArrayExpress (http://www.ebi.ac.uk/arrayexpress) under accession number E-MEXP-2552.

Thirty-three genes were found to be up-regulated by > 1.5-fold in AtCPL5-overexpressing plants compared with the WT (Table S5), whereas 101 genes were down-regulated by > 1.5-fold in AtCPL5-overexpressing plants (Table S6). As shown in Table 1, approximately one-third of the genes up-regulated by AtCPL5 were found to encode transcription factors, including the dehydration responsive element-binding protein (DREB)-type APETALLA 2 (AP2)/ethylene responsive factor (ERF) transcription factors, such as RAP2.4, RAP2 (related to AP2) and QRAP2 (glutamine-rich AP2), which are known to be responsive to ABA and drought stress (Seki et al., 2002; Wei et al., 2005; Lin et al., 2008). In addition, the genes encoding Aux/IAA transcriptional repressors for auxin signaling, such as IAA30, were also up-regulated in AtCPL5 overexpressors: IAA30 was previously shown to be up-regulated by ABA in Arabidopsis roots (Rodrigues et al., 2009). Furthermore, genes involved in various processes of cellular metabolism and cell proliferation regulation, along with those encoding protein kinases and phosphatases, were also up-regulated by AtCPL5. In particular, the up-regulation of CYP79C1 (CYTOCHROME-P450 MONOOXYGENASE 79C1) and At2g18320, which are involved in the biosynthesis of glucosinolates (Wittstock & Halkier, 2002; Schuler et al., 2006) and root-specific triterpenoids (Ehlting et al., 2008; Nelson et al., 2008), respectively, was noticeable. These secondary metabolites are known to be accumulated in response to various stresses including drought (Bouchereau et al., 1996; Champolivier & Merrien, 1996; Nelson et al., 2008). Another metabolic gene of interest was CCD7 (CARETENOID CLEAVAGE DIOXYGENASE 7), which is likely to control the ABA concentration in Arabidopsis through the competitive production of strigolactones from carotenoids, which also serve as precursors for ABA biosynthesis (Ohmiya, 2009). Genes down-regulated as a result of the overexpression of AtCPL5 included those involved in defense responses against pathogens, transcriptional regulation, the cell cycle, protein phosphorylation and dephosphorylation, RNA and protein modification, and other molecular functions (Tables 2,S6).

Table 1.   Genes showing > 1.5-fold enhanced transcript levels in C-terminal domain phosphatase-like 5 (AtCPL5)-overexpressing plants compared with wild-type (WT) plants
Probe IDAGI No.DescriptionFold changeaFold changeb
Exp. 1Exp. 2ABADrought
  1. aFold inductions of gene expression in AtCPL5-overexpressing plants compared with those of WT plants (OE#5/WT ≥ 1.5).

  2. bFold inductions of gene expression in WT plants treated with ABA or drought stress compared with those of untreated WT plants: collected from Genevestigator (http://www.genevestigator.ethz.ch). AGI, Arabidopsis Genome Initiative; AP2, APETALLA 2; ERF, ethylene responsive factor; MYB, myeloblast; IBR, in-between ring fingers.

Transcriptional regulation
262135_atAt1g78080AP2/ERF transcription factor (RAP2.4)2.61.71.42.7
255926_atAt1g22190AP2/ERF transcription factor (RAP2)1.51.91.75.4
253799_atAt4g28140AP2/ERF transcription factor (QRAP2)1.71.61.52.1
246401_atAt1g57560MYB transcription factor (AtMYB50)1.61.5  
251246_atAt3g62100Indole-3-acetic acid inducible transcription factor (IAA30)2.01.81.9 
253127_atAt4g36060Basic helix-loop-helix (bHLH) family protein1.72.0  
260431_atAt1g68190Zinc finger (B-box type) family protein1.61.6  
262917_atAt1g64800Transcription factor1.51.7 1.7
247767_atAt5g58890Agamous-like 82 (AGL82)2.81.72.6 
Cellular metabolism
264090_atAt1g79370Cytochrome P450 monooxygenase (CYP79C1)2.93.41.5 
266251_s_atAt2g18320Cytochrome P450, family 7082.11.71.81.5
265175_atAt1g23480Cellulose synthase-like 3 (AtCSLA03)1.71.6  
253934_atAt4g26830Hydrolase, hydrolyzing O-glycosyl compounds1.61.6 7.6
266129_atAt2g44990Carotenoid cleavage dioxygenase (CCD7)1.61.6  
265975_atAt2g11270Citrate synthase-related2.99.4 2.2
Protein kinase/phosphatase
255419_atAt4g03230Protein serine/threonine kinase1.91.81.92.5
260722_atAt1g48040Protein serine/threonine phosphatase1.91.71.92.5
254111_atAt4g24890Purple acid phosphatase 24 (PAP24)1.91.6 9.9
Others
247756_atAt5g59100Subtilase family protein1.91.8 10.2
257483_atAt1g49620Cyclin-dependent protein kinase inhibitor (ICK5)2.11.5  
247518_atAt5g61800Pentatricopeptide (PPR) repeat-containing protein1.71.92.94.8
251366_atAt3g61340F-box family protein1.61.61.5 
245590_atAt4g15060F-box protein-related3.11.9  
247344_atAt5g63750IBR domain-containing protein4.02.01.833.2
Genes with unknown function
257682_atAt3g13240Hypothetical protein1.61.61.812.0
254614_atAt4g19191Hypothetical protein1.51.51.53.4
256206_atAt1g50910Hypothetical protein1.52.5 2.4
250773_atAt5g05430Hypothetical protein1.62.1 2.2
249662_s_atAt5g36680Hypothetical protein3.01.8  
Table 2.   Some genes showing > 1.5-fold reduced transcript levels in C-terminal domain phosphatase-like 5 (AtCPL5)-overexpressing plants compared with wild-type (WT) plants
Probe IDAGI No.DescriptionsFold changeaFold changeb
Exp. 1Exp. 2ABADrought
  1. aFold reductions of gene expression in AtCPL5-overexpressing plants compared with those of WT plants (OE#5/WT ≤ 0.6).

  2. bFold reductions of gene expression in WT plants treated with ABA or drought stress compared with those of untreated plants: collected from Genevestigator (http://www.genevestigator.ethz.ch).

Defense
245454_atAt4g16920Disease resistance protein (TIR-NBS-LRR class)0.50.5  
265586_atAt2g19990Pathogenesis-related protein1-like (PR-1-like)0.50.50.40.5
259802_atAt1g72260Fungal defense thionin (THI2.1)0.60.50.40.6
264426_atAt1g61760Harpin-induced protein-related0.60.6  
Transcriptional regulation
245173_atAt2g47520AP2-domain-containing transcription factor0.20.6 0.3
246522_atAt5g15830Arabidopsis thaliana basic leucine-zipper 30.60.30.20.1
245685_atAt5g22220Transcription factor-1 (E2F1)0.60.6  
258204_atAt3g13960Growth-regulating factor 5 (AtGRF5)0.40.6  
Cell cycle
261780_atAt1g76310Cyclin-dependent protein kinase regulator (CYCB2;4)0.60.60.6 
Protein kinase/phosphatase
267289_atAt2g23770Protein kinase family protein0.60.30.4 
252964_atAt4g38830Protein kinase family protein0.50.30.6 
263111_s_atAt1g65190Protein kinase family protein0.40.40.20.3
246459_atAt5g16900Leucine-rich repeat protein kinase0.40.50.3 
247684_atAt5g59670Leucine-rich repeat protein kinase0.60.6 0.3
251160_atAt3g63240Phosphatase family protein0.40.50.50.5
267499_atAt2g45500Nucleoside-triphosphatase0.50.4  
RNA/Protein modification
255938_atAt1g12620Pentatricopeptide (PPR) repeat-containing protein0.60.50.60.1
267455_atAt2g33760Pentatricopeptide (PPR) repeat-containing protein0.50.5  
257765_atAt3g23020Pentatricopeptide (PPR) repeat-containing protein0.60.5 0.6
262509_atAt1g11290Chlororespiratory reduction 22 (CRR22)0.60.50.40.4
249405_atAt5g40280Farnesyltransferase beta subunit (ERA1)0.50.60.4 

Induction of the ABA- or drought stress-responsive genes by AtCPL5 overexpression was further investigated in RNA hybridization analyses, using WT and AtCPL5-overexpressing plants under either ABA-treated or drought-stressed conditions. The expression of RAP2.4, RAP2, and QRAP2 was found to be transiently induced by ABA or drought treatment in WT plants, as shown in Fig. 7. In particular, a dramatic induction of QRAP2 by drought stress was noted. Similar transient and dramatic inductions in response to drought stress were previously reported for RAP2 and QRAP2, respectively (Seki et al., 2002; Wei et al., 2005). In addition, consistent with the results obtained in the microarray experiments, all three transcripts showed greater accumulation in AtCPL5-overexpressing plants compared with the WT. Furthermore, when AtCPL5-overexpressing plants were treated with ABA or drought stress, further increases in the accumulation of the transcripts were observed, as was the case for AtCPL5 in ABA-treated AtCPL5-overexpressors. However, there were differences in the kinetics of the transcript accumulations: for example, the increased level of RAP2.4 transcripts following ABA or drought stress in AtCPL5-overexpressors was maintained, while RAP2 showed only transient increases in response to the same treatments.

Figure 7.

 Up-regulation of RAP2.4, RAP2 (RELATED TO AP2) and GLUTAMINE-RICH AP2 (QRAP) in C-terminal domain phosphatase-like 5 (AtCPL5)-overexpressing transgenic plants. (a) RNA hybridization analyses of 2-wk-old wild-type (WT), OE#5, and OE#2 plants treated with 0.1 mM ABA for 0, 1, 2 and 3 h. (b) RNA hybridization analyses of 2-wk-old WT, OE#5, and OE#2 plants subjected to drought stress for 0, 3, 6 and 9 h. All experiments were performed at least three times, and results from one representative experiment are shown.

In conclusion, these results indicate that AtCPL5 regulates the expression of certain ABA- or drought-responsive genes positively, including those encoding the DREB-type AP2/ERF transcription factors, such as RAP2.4, RAP2, and QRAP2, which are thought to contribute to ABA-mediated development and drought tolerance in Arabidopsis.

Ser-2-specific dephosphorylation of RNAPII CTD by AtCPL5 in vitro

His6-Full (1–601 aa), His6-CPD1 (1–300 aa), and His6-CPD2 (301–601 aa) were expressed in E. coli to test the phosphatase activities of the full-length protein and individual CPDs of the AtCPL5 protein toward the CTD of RNAPII. The bacterial transformants produced 77-, 42- and 42-kD polypeptides corresponding to His6-Full, His6-CPD1 and His6-CPD2, respectively, as detected by SDS-PAGE and anti-His-tag immunoblotting and further purified by nickel-affinity chromatography (Fig. 8a). As a potential substrate, the CTD of the largest subunit of Arabidopsis RNAPII was also expressed in the form of a GST-AtCTD fusion protein in E. coli. The purified recombinant protein was detected by SDS-PAGE and anti-GST-tag immunoblotting (Fig. 8b). The CTD phosphatase activities of the purified AtCPL5 and each CPD of the protein were determined using the GST-AtCTD protein phosphorylated by Cdc2 protein kinase (Zheng et al., 2005), and H5 and H14 monoclonal antibodies which are specific for CTD-Ser-2-PO4 and CTD-Ser-5-PO4, respectively (Bregman et al., 1995). In immunoblot analyses, phosphorylated GST-AtCTD, which has a lower electrophoretic mobility (Fig. 8c, lane 2) compared with unphosphorylated GST-AtCTD (Fig. 8c, lane 1), was recognized by both H5 and H14. Unexpectedly, H14 showed some cross-reactivity with unphosphorylated GST-AtCTD (Fig. 8c, panel H14, lane 1). After dephosphorylation with His6-Full, His6-CPD1 or His6-CPD2, phosphorylated GST-AtCTD was still strongly recognized by H14 (Fig. 8c, panel H14, lanes 3–5). However, it was rarely detected by H5 when dephosphorylated with His6-Full, and weakly detected by H5 when dephosphorylated with His6-CPD1 or His6-CPD2; the amount detected was slightly higher in the case of dephosphorylation with His6-CPD2 than with His6-CPD1 (Fig. 8c, panel H5, lanes 3–5). These results indicated that His6-Full, His6-CPD1, and His6-CPD2 removed phosphates from Ser-2, but not from Ser-5, of phosphorylated GST-AtCTD. Thus, AtCPL5 was concluded to have Ser-2-specific RNAPII CTD phosphatase activity in vitro, whose full activity is likely to result from the additive or synergistic effects of the partial activities of CPD1 and CPD2.

Figure 8.

 Arabidopsis RNA polymerase II (RNAPII) C-terminal domain (CTD) phosphatase activities of recombinant CTD phosphatase-like 5 (AtCPL5) proteins. (a) SDS-PAGE (left panel) and immunoblot analyses (right panel) of the soluble recombinant AtCPL5 proteins, His6-Full, His6-CPD1 and His6-CPD2, obtained from Escherichia coli transformants. The sizes of marker proteins (M) are indicated on the left in kD, and the arrows indicate the purified recombinant proteins. (b) SDS-PAGE (left panel) and immunoblot analyses (right panel) of the purified soluble recombinant GST-AtCTD. (c) Dephosphorylation of phosphorylated GST-AtCTD by purified His6-Full, His6-CPD1, and His6-CPD2. GST-AtCTD protein phosphorylated by Cdc2 kinase was used for dephosphorylation assays. Dephosphorylation status was detected by immunoblotting with either Ser-2-PO4-recognizing H5 or Ser-5-PO4-recognizing H14 antibody. The amount of GST-AtCTD applied to the gels was 0.3 μg for H14, 0.5 μg for H5 and 1 μg for Coomassie brilliant blue (CBB) staining.

Discussion

Eukaryotic transcriptional regulation and mRNA processing are both mediated by phosphorylation and dephosphorylation of the CTD of RNAPII. The Arabidopsis family of CPLs that dephosphorylate the CTD are thought to be involved in transcriptional regulation of various physiological processes, including hormone and stress responses, as well as plant development (Koiwa, 2006). In this study, we characterized an RNAPII CTD phosphatase-like gene, AtCPL5, which encodes the regulator of ABA-mediated responses in Arabidopsis.

One of the distinct features of AtCPL5 is its role in the positive regulation of ABA-mediated development and stress responses in Arabidopsis, while other well-characterized Arabidopsis CPLs, such as CPL1/FRY2, CPL2 and CPL3, act as negative regulators (Koiwa et al., 2002; Xiong et al., 2002; Ueda et al., 2008; Matsuda et al., 2009). The induction of AtCPL5 by ABA and drought stress was demonstrated by quantitative real-time RT-PCR analyses and histochemical GUS staining of PAtCPL5:GUS transgenic seedlings (Figs 2, 3l,m). In addition, perturbation of AtCPL5 transcript levels in WT Arabidopsis caused altered developmental responses to exogenous ABA treatments: for example, AtCPL5-overexpressing plants showed hypersensitivity in seed germination and seedling growth inhibition in response to ABA treatment (Fig. 5). Furthermore, involvement of AtCPL5 in the positive regulation of stress responses was supported by the altered stomatal apertures, transpiration rates, and survival rates of AtCPL5-overexpressing transgenic and cpl5 knockout plants in response to ABA and drought stress (Fig. 6). Considering AtCPL5 expression in guard cells and its ABA activation in rosette leaves (Fig. 3), it is tempting to speculate that the increased survival rates of the AtCPL5-overexpressing transgenic plants upon drought stress might have resulted in part from the regulation of stomatal closure by AtCPL5. The weak phenotypes of cpl5 mutants in response to ABA and drought treatments reflect the possible functional redundancy of AtCPL5 with other CPL protein(s) in Arabidopsis.

As mentioned in the Results, the phenotypic differences between AtCPL5-overexpressors or the cpl5 mutant and WT plants were only observed with ABA or drought treatment among the AtCPL5-inducing stressors. The induction of AtCPL5 by other stresses such as NaCl and cold might have occurred through the indirect or secondary effects of the stress treatments (e.g. the subsequent accumulation of ABA). This might be one of the possible reasons why AtCPL5 was only specific for ABA and drought responses. Similar differential phenotypic responses to the stressors that induce certain stress-responsive genes have been reported by other groups (Chen et al., 2007; Xiang et al., 2008).

A considerable number of genes involved in various molecular functions were found to be either up-regulated (33 genes) or down-regulated (101 genes) to moderate extents as a result of AtCPL5 overexpression in our microarray experiments (Tables 1,S5,S6). It was noted that the magnitudes of either increases or decreases in transcript levels in AtCPL5-overexpressing plants were pretty similar in general to those found in ABA- or drought-treated plants by other groups, when the microarray data obtained in this study were compared with those deposited in the Genevestigator database (see Tables 1, S5 for increases, and Tables 2, S6 for decreases). These results support the idea that AtCPL5 is likely to act as a transcriptional regulator in response to ABA or drought, and that most of the moderate changes in transcript levels in AtCPL5-overexpressing plants occurred mainly through the regular function of AtCPL5 with the minimal functional loss that might have resulted from the potential disturbance of normal protein–protein interactions as a consequence of AtCPL5 overexpression. The possible involvement of AtCPL5 in either transcriptional elongation or the termination stage might be the reason for the moderate changes in transcript levels in AtCPL5-overexpressors (see later discussion). However, some very large differences in fold increases or decreases in transcript levels were observed for certain genes between our microarray data and those obtained by other groups, especially for drought-stressed plants. We speculate that these differences were attributable to either the distraction effect of AtCPL5 overexpression or the different developmental stages of the plants used for microarray studies and differential expression of the stress-responsive genes depending on the severity of the encountered stress.

Interestingly, approximately one-third of the genes up-regulated as a result of AtCPL5 overexpression were found to encode transcription factors, whereas only a few down-regulated genes encoded transcription factors of unknown functions. These results indicate that AtCPL5 is likely to act as a transcriptional regulator required for the induction of transcription factors presumably involved in ABA-mediated plant development and drought tolerance. The predominant localization of the AtCPL5::smGFP fusion protein to the nucleus supports this idea (Fig. 4). Consistent with this notion, the up-regulated genes in AtCPL5-overexpressing plants included previously reported ABA- or drought-responsive genes that encode AP2/ERF transcription factors, such as RAP2.4, RAP2 and QRAP2 (Seki et al., 2002; Wei et al., 2005; Lin et al., 2008). In the present study, RAP2.4, RAP2 and QRAP2 were confirmed to be induced by ABA or drought stress (Fig. 7). These transcription factors are very closely related to one another phylogenetically, belonging to the A-6 subgroup of the DREB subfamily in the AP2/ERF transcription factor superfamily, which is known to be involved in the regulation of plant development and stress responses (Sakuma et al., 2002; Wei et al., 2005; Nakano et al., 2006): in particular, RAP2.4 is known to be involved in drought tolerance as well as the regulation of multiple developmental processes, including cotyledon expansion and root elongation (Lin et al., 2008). Consistent with the induction of RAP2.4, RAP2 and QRAP2 by ABA or drought (Table 1 and Fig. 7), the promoter regions of these closely related genes were found to contain one or two copies of the ABA-responsive cis element (ABRE) (data not shown).

Another up-regulated gene of interest involved in transcriptional regulation was IAA30, encoding the AUX/IAA transcriptional repressor for auxin signaling. It was previously shown to be up-regulated by ABA in Arabidopsis roots (Rodrigues et al., 2009). Therefore, up-regulation of the repressor gene for auxin signaling, along with the down-regulation of other root-abundant auxin-responsive genes (e.g. At5g47530 in Table S6; Zimmermann et al., 2004), in AtCPL5-overexpressing plants suggests that AtCPL5 might be involved in antagonism of ABA with auxin for the development of the root system (Deak & Malamy, 2005; De Smet et al., 2006). In addition, down-regulation of the pathogen defense genes in AtCPL5-overexpressing plants (Table S6), such as At4g16920 (TIR-NBS-LRR, Toll/interleukin receptor-nucleotide binding site-leucine rich repeat), At1g61760 (harpin-induced protein-related), fungal defense thionin (THI2.1) (Bohlmann et al., 1998), and pathogenesis-related protein 1-like (PR-1-like), reflects the interference of ABA with plant biotic stress responses (for a review, see Asselbergh et al., 2008). Therefore, it will be interesting to test the susceptibility of AtCPL5-overexpressing plants to pathogen attacks in the future.

Little is known about the downstream target genes regulated by RAP2.4, RAP2, QRAP2 and IAA30, which are up-regulated by AtCPL5. However, among the up- or down-regulated genes in AtCPL5-overexpressing plants, certain sets are likely to be regulated by these transcription factors under certain developmental or stress conditions, rather than directly regulated by AtCPL5, thereby contributing to ABA-mediated plant development and drought stress responses.

Irrespective of the possible regulation by the transcription factors mentioned above, the genes involved in cell proliferation regulation in Tables 1 and 2 (or Tables S5 and S6) are likely to attribute to the seedling growth inhibition by ABA, which was pronounced in ABA-treated AtCPL5-overexpressors (Fig. 5). Indeed, the negative regulator cyclin-dependent protein kinase inhibitor (ICK5; De Veylder et al., 2001) was up-regulated in AtCPL5-overexpressing plants, while the positive regulators, such as growth-regulating factor 5 (AtGRF5; Horiguchi et al., 2005) and B2-type cyclin (CYCB2; Lee et al., 2003), were down-regulated in the same transgenic plants. Also consistently, overexpression of ICK5 in Arabidopsis resulted in inhibition of plant growth and development (Chan, 2007), while overexpression of AtGRF5 and CYCB2 promoted the growth of the Arabidopsis leaf and the rice (Oryza sativa) root, respectively (Lee et al., 2003; Horiguchi et al., 2005).

Among the genes up-regulated by AtCPL5 overexpression, genes involved in ABA- or drought-induced biosynthesis of glucosinolate (CYP79C1), triterpenoid (At2g18320), and citrate (At2g11270) might contribute to the adaptive responses of plants subjected to drought stress, such as stomatal movements and osmotic adjustments (Venekamp, 1989; Lopez-Berenguer et al., 2008). For example, the potential increase in the cellular concentration of citrate, a competitive inhibitor of the vacuolar malate transporter, which is involved in stomatal movements (Rentsch & Martinoia, 1991; Emmerlich et al., 2003), might have been involved in the regulation of stomatal closure that was strongly manifested in ABA-treated AtCPL5-overexpressing plants (Fig. 6). Other genes that are thought to be involved in the regulation of stomatal closure via AtCPL5 in response to ABA or drought were also found among the list of down-regulated genes in AtCPL5-overexpressing plants: ENHANCED RESPONSE TO ABA1 (ERA1), encoding the β-subunit of protein farnesyltransferase (At5g40280), was an example, a loss-of-function mutation in which caused an enhanced response to ABA in stomatal closure as well as seed germination (Cutler et al., 1996; Pei et al., 1998).

Phosphorylation and dephosphorylation of different serine residues in RNAPII CTD predominate during different phases of transcription (Komarnitsky et al., 2000), and the phosphorylation status of the CTD in RNAPII was suggested as a focal regulatory point in plant stress response and development (Koiwa et al., 2002). Congruent with this notion, the previously reported repression of stress-responsive genes by CPL1, CPL2, or CPL3 (Koiwa et al., 2002; Xiong et al., 2002; Ueda et al., 2008; Matsuda et al., 2009) is thought to occur through the dephosphorylation of Ser-5 in the RNAPII CTD during either transcriptional preinitiation or the initiation step, which requires extensive phosphorylation of Ser-5 in the RNAPII CTD (Chesnut et al., 1992; Komarnitsky et al., 2000; Koiwa, 2006). CPL1 and CPL2 were indeed shown to be Ser-5-specific phosphatases (Koiwa et al., 2004; Ueda et al., 2008). In addition, various mutations in CPL1, CPL2 and CPL3 genes caused hyperactivation of stress-responsive genes by ABA, osmotic stress, and auxin, respectively (Koiwa et al., 2002; Xiong et al., 2002; Bang et al., 2006; Ueda et al., 2008). In contrast to this negative regulation of RNAPII activity by CPL1, CPL2 and CPL3, the activation of ABA- or drought-responsive genes, such as RAP2.4, RAP2 and QRAP2, in AtCPL5-overexpressing plants is thought to involve the dephosphorylation of Ser-2 in the RNAPII CTD by AtCPL5 (Fig. 8). Considering the fact that eukaryotic RNAPII CTD undergoes phosphorylation at Ser-2 residues during the transcriptional elongation step (Komarnitsky et al., 2000; Cho et al., 2001), dephosphorylation of Ser-2 in the RNAPII CTD by AtCPL5 might facilitate recycling of the hyperphosphorylated form of RNAPII, allowing RNAPII to enter another cycle of transcription. The possible involvement of AtCPL5 in transcriptional elongation or the termination stage might help to explain the moderate changes in the induction of its downstream target genes in AtCPL5-overexpressing plants (Table 1). A similar mechanism was previously proposed for yeast and human FCP1, which is a Ser-2 CTD phosphatase involved in regeneration of RNAPII at the end of the transcription cycle to facilitate another round of transcription (Cho et al., 1999).

It is possible that AtCPL5 may mediate ABA and drought responses based on transcriptional regulation through an interaction with certain transcription factors associated with the RNAPII complex, in light of the findings of interactions of another Ser-2 CTD phosphatase, FCP1, with various transcription factors involved in different stages of the transcription cycle in yeast and humans (e.g. RAP74 and TFIIF in the RNAPII complex, and the HIV transcriptional activator Tat) to regulate FCP1 phosphatase activity during the transcription cycle (Marshall et al., 1998; Kobor et al., 2000; Cho et al., 2001). A similar model has been proposed for CPL1 in Arabidopsis, suggesting that Arabidopsis thaliana myeloblast (AtMYB)3 function in abiotic stress signaling in concert with CPL1 (Bang et al., 2008). Actually, some transcription factors have been identified as candidates for CPL5-interacting proteins in our laboratory, and functional studies of them are underway. However, in light of the possible involvement of AtCPL5 in either transcriptional elongation or termination, the ABA and drought responses might also occur through an interaction of AtCPL5 with the chromatin-remodeling proteins that promote transcriptional elongation or termination (for a review, see Conaway et al., 2000). In any cases, elucidation of the mechanism by which AtCPL5 mediates ABA and drought responses will be an intriguing challenge for future studies: however, it should be kept in mind that a suboptimal stoichiometric interaction of AtCLP5 with the cognate binding proteins or an interaction of AtCPL5 with certain nonspecific proteins might possibly occur as a result of the overexpression of AtCPL5 in the transgenic plants.

Another novel feature of AtCPL5 is the unique structure of the protein. One of 24 CPD-containing CPLs in Arabidopsis (Fig. 1b), AtCPL5 consists of two separable polypeptide regions with high sequence homology to each other and contains highly conserved CPD and DXDX(T/V) motifs (Fig. 1a,c). The tandem DNA repeats corresponding to the two homologous regions in AtCPL5 are separated by a 1352-bp intron. Therefore, AtCPL5 is thought to have originated from the duplication of an ancestral gene followed by sequence-diverging and gene-fusing processes. Note that TBLASTN analyses of both halves of AtCPL5 using the DNA sequences available in GenBank revealed the presence of their homologs in lower plants whose genomes have recently been completely sequenced: that is, XM_001416694.1 (E-values: 2e-29 and 1e-30 for each half, respectively) in Ostreococcus lucimarinus, a unicellular green alga belonging to the Prasinophyceae which is one of the most ancient groups within the lineage giving rise to land plants (Palenik et al., 2007), as well as XM_001781932.1 (E-values: 6e-32 and 3e-33 for each half, respectively) and XM_001758975.1 (E-values: 2e-31 and 4e-31 for each half, respectively) in Physcomitrella patens, a moss belonging to the Bryophytes which is a remnant of a diverging land plant lineage (Rensing et al., 2008). All three proteins have the conserved DLDHT sequence upstream of their CPDs. On the basis of the presence of only one predominant homolog for AtCPL5 in the surface ocean-thriving green alga O. lucimarinus and two homologs with high amino acid sequence similarity to each other (63% identity: similar to the identity level (64%) between CPD1 and CPD2 in AtCPL5) in the land plant P. patens, the duplication of the AtCPL5 ancestral gene is speculated to have occurred by the early Silurian (430 million yr ago), when green plants expanded their habitat to land; this evolutionary movement involved an increase in gene family complexity, acquisition of genes for tolerating terrestrial stresses (e.g. water availability), and development of ABA signaling for dehydration responses (Rensing et al., 2008). The significance of the acquisition of double CPDs in a protein might be explained in a similar way. In other words, the additive or synergistic effect of the phosphatase activities of CPD1 and CPD2 in AtCPL5 (Fig. 8c), created by further evolutionary selective pressure after the gene duplication event, might have helped the development of stress-tolerance responses during the appearance or adaptation of higher plants to land.

On the basis of the high sequence homology of AtCPL5 to both SSP10 and SSP9, AtCPL5 is thought to be a paralog of the SSP10 and SSP9 genes, as is probably the case for the genes encoding CPL1 and CPL2, as well as CPL3 and CPL4 (Koiwa et al., 2004; Bang et al., 2006; Ueda et al., 2008). Therefore, complex multiplication and divergence processes for CPL genes are thought to have occurred during the evolution of higher plants. The accumulation of genome sequence information from diverse plant species along with biochemical and molecular genetic studies will help to elucidate the evolution of double CPD-containing proteins in higher plants.

In summary, AtCPL5 is proposed as a novel CPL family protein that positively regulates ABA-mediated responses in Arabidopsis through the regulation of a significant number of stress-responsive or developmental genes, at least some of which possibly involve the dephosphorylation of Ser-2 in RNAPII CTD. Further elucidation of the physiological significance of the presence of double CPDs, as well as the precise mechanism of AtCPL5 action in terms of how the Ser-2-specific RNAPII CTD phosphatase activity is regulated in the process of ABA signaling in planta, will help us better understand the evolution of CPLs, as well as the complex nature of the regulation of gene expression in the ABA-mediated development and stress responses of higher plants.

Acknowledgements

This study was supported by a research grant (CG1210 to M.K.) from the Crop Functional Genomics Center of the 21st Century Frontier Research Program funded by the Korean Ministry of Education, Science and Technology. H.K. received funding through USDA-CSREES grant 2006-34402-17121 ‘Designing Food for Health’.

Ancillary