Light regulation of CaS, a novel phosphoprotein in the thylakoid membrane of Arabidopsis thaliana


E.-M. Aro, Department of Biology, Plant Physiology and Molecular Biology, University of Turku, FI-20014 Turku, Finland
Fax: +358 2 333 5549
Tel: +358 2 333 5931


Exposure of Arabidopsis thaliana plants to high levels of light revealed specific phosphorylation of a 40 kDa protein in photosynthetic thylakoid membranes. The protein was identified by MS as extracellular calcium-sensing receptor (CaS), previously reported to be located in the plasma membrane. By confocal laser scanning microscopy and subcellular fractionation, it was demonstrated that CaS localizes to the chloroplasts and is enriched in stroma thylakoids. The phosphorylation level of CaS responded strongly to light intensity. The light-dependent thylakoid protein kinase STN8 is required for CaS phosphorylation. The phosphorylation site was mapped to the stroma-exposed Thr380, located in a motif for interaction with 14-3-3 proteins and proteins with forkhead-associated domains, which suggests the involvement of CaS in stress responses and signaling pathways. The knockout Arabidopsis lines revealed a significant role for CaS in plant growth and development.




calcium-sensing receptor




maximal fluorescence


plasma membrane-specific ferroxidase


variable fluorescence


green fluorescent protein


immobilized metal affinity chromatography


liquid chromatography


phosphorylated form of calcium-sensing receptor


photosystem I


photosystem II


yellow fluorescent protein

Protein phosphorylation is one of the key mechanisms used by all domains of life for regulation of cellular processes, from gene expression to metabolic control. In plants, protein phosphorylation plays crucial roles during acclimation of the photosynthetic apparatus to changing environmental cues [1]. Light- and redox-dependent protein phosphorylation is particularly important for regulation of photosynthetic protein complexes located in the thylakoid membranes of chloroplasts. Four major protein complexes are involved in photosynthetic light reactions: photosystem I (PSI), photosystem II (PSII), cytochrome b6f complex, and ATP synthase. The major phosphoproteins in the thylakoid membrane belong to PSII and its light-harvesting antenna II. The application of MS combined with affinity chromatography for phosphopeptide enrichment has allowed identification of the major phosphoproteins of PSII (D1, D2, CP43 and PsbH proteins) and light-harvesting antenna II [Lhcb1, Lhcb2 and the minor CP29 (Lhcb4) proteins] [2–4]. Phosphorylation of PSII core proteins is believed to play an important role in the repair cycle of the reaction center protein D1 and the assembly of PSII [5,6]. Reversible phosphorylation of light-harvesting antenna II proteins regulates state transitions, i.e. the mechanism that ensures a balanced excitation of PSI and PSII in changing environmental and metabolic conditions [7–11]. Two phosphorylated proteins have also been identified in PSI, but the biological significance of their phosphorylation still remains to be elucidated [4,12]. Likewise, two cytochrome b6f complex subunits undergo reversible phosphorylation: subunit IV, revealed by radioactive labeling [13], and Rieske Fe–S protein, which undergoes N-terminal phosphorylation, identified by MS [14]. Furthermore, a recent study has shown that a thylakoid membrane-associated protein, TSP9, is phosphorylated at multiple sites in response to increasing light intensity, and it is thought to play a role in plant stress acclimation and signal transduction [15].

A specific feature of environmentally induced thylakoid protein phosphorylation is an almost exclusive phosphorylation of Thr residues in the proteins of both plant and green algal photosynthetic membranes [1,16]. The use of reverse genetics has allowed identification of two light-dependent protein kinases involved in phosphorylation of thylakoid proteins. STN7 protein kinase is essential for phosphorylation of Lhcb1, Lhcb2 and Lhcb4 proteins [11,17] and, thus, for state transitions. The homologous STN8 protein kinase is involved in the phosphorylation of PSII core proteins and is absolutely essential for phosphorylation of PsbH protein of PSII at Thr4 [18,19].

Here we report the identification of a novel phosphoprotein, calcium-sensing receptor (CaS), from thylakoid membranes of Arabidopsis. The protein was previously named CaS and characterized as an extracellular calcium-sensing receptor localized in plasma membrane [20,21]. Both biochemical and immunolocalization studies, however, provide strong evidence that CaS is a chloroplast protein localized in the thylakoid membrane and not detectable in the plasma membrane. It is shown that the CaS protein level as well as its phosphorylation level increase in response to increasing light intensities. The phosphorylation site is mapped to Thr380, and is shown to be dependent on the STN8 protein kinase. Insertional mutagenesis of CaS resulted in reduced growth, indicating a significant role for CaS protein in plant growth and development.


Identification of CaS as a thylakoid 40 kDa phosphoprotein

In order to investigate the molecular mechanisms involved in acclimation of plant photosynthetic machinery to high light intensities, we isolated thylakoid membranes from the leaves of Arabidopsis and analyzed the light-induced changes in protein phosphorylation by immunoblotting with phosphothreonine-specific antibody (Fig. 1A). This analysis revealed the phosphorylation of a novel polypeptide with a molecular mass of about 40 kDa whose level of phosphorylation strongly increased with rising irradiance. To identify this 40 kDa phosphoprotein, thylakoids isolated from leaves exposed to high-light treatment were subjected to trypsin shaving [3,4]. The surface-exposed domains of membrane proteins were released and separated from the membranes by centrifugation. The resulting complex mixture of hydrophilic peptides was subjected to immobilized metal affinity chromatography (IMAC) [19] for phosphopeptide enrichment. The enriched phosphopeptides were analyzed by liquid chromatography (LC)-MS/MS.

Figure 1.

 Identification of CaS as a 40 kDa thylakoid phosphoprotein and its regulation by light in thylakoids. (A) Thylakoids were isolated from dark-adapted (D) leaves or leaves exposed for 3 h to low (30 μmol photon·m−2·s−1) (LL), growth (100 μmol photon·m−2·s−1) (GL) or high (600 μmol photon·m−2·s−1) (HL) light, and proteins were separated by SDS/PAGE and immunoblotted with phosphothreonine-specific antibody. Chlorophyll (0.75 μg) was loaded in each well. Well-known thylakoid phosphoproteins are marked, and the position of the 40 kDa phosphoprotein is indicated by an arrow. (B) The product ion spectrum of the doubly charged peptide ion with m/z 573.8 obtained by ESI and collision-induced fragmentation. The parent ion is labeled in the spectrum along with the fragment ion at m/z 524.8 produced after the characteristic neutral loss of phosphoric acid. The detected b-ions (N-terminal) and y-ions (C-terminal) are indicated in the spectrum as well as in the corresponding amino acid sequence. The ions marked with an asterisk indicate that the fragments underwent neutral loss of 98 Da (H3PO4). The lower-case ‘t’ indicates a phosphorylated Thr residue. (C) Immunoblot with CaS-specific antibody [for experimental settings, see (A)].

Besides several known phosphopeptides of the thylakoid membranes (supplementary Table S1), the analysis of data allowed the identification of a novel, previously uncharacterized phosphopeptide. The product ion spectrum of the corresponding doubly charged molecular ion with m/z 573.8 is presented in Fig. 1B. The series of b- and y-ions revealed the peptide sequence SGtKFLPSSD, with lowercase ‘t’ indicating phoshorylated Thr. A search in the Arabidopsis protein sequence database revealed that the amino acid sequence belongs to the C-terminus of the expressed protein At5g23060 with deduced molecular mass 41.3 kDa, previously described as an extracellular CaS [20].

In a parallel approach, the gel region corresponding to the 40 kDa phosphoprotein band in the gel (Fig. 1A) was cut out and subjected to in-gel digestion for protein identification by LC-MS/MS. CaS, together with 14 other proteins, was identified from this gel band (supplementary Table S2).

CaS-specific antibody was then used to determine whether the increased occurrence of phosphorylated CaS under high-light conditions (Fig. 1A) was related to an increase in the amount of CaS per se. As shown in Fig. 1C, the total amount of CaS protein was not drastically changed by increasing irradiance, but the phosphorylated form of CaS (P-CaS) clearly accumulated under high-light conditions as compared to darkness.

Chloroplast localization of CaS

Localization of the CaS phosphoprotein to the thylakoid membrane, as discussed above, is in good agreement with proteomics studies [22–24], but strongly contrasts with a previous report of the plasma membrane localization of CaS, using heterologous expression in onion epidermis cells, which unfortunately lack chloroplasts [20]. To address this apparent discrepancy, the subcellular localization of the endogenous CaS in Arabidopsis was investigated by exploiting purified membrane fractions and immunoblotting with purified CaS-specific antibody. CaS was not found in purified plasma membrane, whereas it was present in intact chloroplasts and in the thylakoid fraction but not in the stroma fraction (Fig. 2A). The purity of the membrane fractions was demonstrated by using plasma membrane-specific ferroxidase (FOX1) and thylakoid membrane-specific D1 antibodies as specific markers (Fig. 2A).

Figure 2.

 Localization of CaS to chloroplasts. (A) Plasma membrane (PM), intact chloroplasts (Chl), thylakoids (Th) and soluble stroma (S) were isolated from wild-type Arabidopsis, and proteins were separated by SDS/PAGE and immunoblotted with CaS-, D1- and FOX1-specific antibodies. Five micrograms (D1) or 10 μg (CaS and FOX1) of protein was loaded in each well. (B) The thylakoids (Th) isolated from leaves exposed to high light were fractionated to stroma-exposed (ST) and grana-exosed (GT) membranes. The fractions were separated by SDS/PAGE and immunoblotted with CaS-specific antibody. One microgram of chlorophyll was loaded in each well.

To further dissect the distribution of CaS in the thylakoid membrane, the thylakoids isolated from leaves exposed to high light were fractionated by digitonin [6]. Immunoblot analysis of thylakoid fractions revealed the presence of CaS both in grana and in stroma thylakoids, and its clear enrichment in the stroma-exposed membranes (Fig. 2B).

To further investigate the contradiction between our data and published reports showing the targeting of fluorescent-labeled CaS to the plasma membrane [20,21], we fused the yellow fluorescent protein (YFP) recombinantly to the C-terminus of CaS and transiently expressed this construct in Nicotiana benthamiana leaves. Observations by confocal laser scanning microscopy clearly demonstrated that the CaS–YFP fusion protein localized in chloroplasts (Fig. 3A–C). In stark contrast, the cytosolic YFP control accumulated YFP fluorescence signal in the cell periphery and nuclei (supplementary Fig. S1). These data clearly demonstrate that CaS predominantly resides in chloroplasts. Coexpression of CaS–YFP and GWD1tp–green fluorescent protein (GFP), a chloroplast-targeted protein used as a marker, showed perfect overlap of the YFP and GFP signals (Fig. 3D–F), and no signal was detected in the cell periphery. Coexpression of CAS–YFP and the cytosolic GFP further illustrated the exclusive localization of CAS–YFP in chloroplasts (supplementary Fig. S2).

Figure 3.

 Chloroplast localization of CaS–YFP in N. benthamiana. (A–C) A leaf section expressing CaS–YFP. (A) YFP fluorescence (excitation 514 nm; emission 545–600 nm). (B) Chloroplast autofluorescence (emission 650–707 nm). (C) Overlay image of (A) and (B). (D–F) A leaf section coexpressing CaS–YFP and GWD1tp–GFP. (D) YFP fluorescence (excitation 514 nm; emission 545–600 nm). (E) GFP fluorescence (excitation 488 nm; emission 495–510 nm). (F) Overlay image of (D) and (E).

Requirement of STN8 kinase for CaS phosphorylation

To address the question of whether one of the two light-regulated protein kinases, STN7 or STN8, is required for the light-dependent phosphorylation of CaS, we isolated thylakoids from the high-light-treated leaves of wild-type plants and two mutant lines lacking STN7 or STN8 (stn7 and stn8, respectively). Immunoblot analysis of isolated thylakoids with phosphothreonine-specific antibody revealed the absence of the 40 kDa CaS phosphorylation in the stn8 mutant (Fig. 4A). Analysis of the same fractions with CaS-specific antibody revealed similar levels of CaS in all samples. The migration of CaS in SDS/PAGE of thylakoid proteins isolated from the stn8 mutant was slightly faster than those of the wild-type and the stn7 mutant (Fig. 4B), which is typically observed when protein phosphorylation is altered (see also Fig. 1A). These data suggest that CaS is almost fully phosphorylated under high-light conditions, as the upper band corresponding to the phosphorylated form dominated under high-light conditions in the wild-type (Figs 1A and 4B) and the stn7 mutant (Fig. 4B).

Figure 4.

 CaS is a substrate for STN8 protein kinase. Thylakoids were isolated from leaves exposed to high light of wild-type (WT) and mutant plants lacking either STN7 (stn7) or STN8 (stn8). The proteins were separated by SDS/PAGE and immunoblotted with (A) phosphothreonine or (B) CaS-specific antibody. The positions of thylakoid phosphoproteins are indicated. (A) 0.75 μg Chlorophyll was loaded in each well. (B) one microgram of chlorophyll was loaded in each well.

The involvement of STN8 in the phosphorylation of CaS was further investigated by isolation of phosphopeptides from the wild-type and the stn7 and stn8 thylakoids, and analyzing them by LC-MS/MS. The mapping of phosphopeptides isolated from stn8 thylakoids in comparison to the wild-type and stn7 showed the specific absence of the CaS-originated phosphopeptide SGtKFLPSSD with m/z 573.82+ from the thylakoids of only the stn8 mutant. These results revealed that CaS in stn8 is not phosphorylated at Thr380, and suggest either that CaS is a direct target of the STN8 protein kinase or STN8 is a crucial component of the protein phosphorylation cascade involved in CaS phosphorylation.

Characterization of the CaS mutant lines

The mutant Arabidopsis lines with T-DNA insertion in the intron region of the CaS gene were obtained from GABI-Kat and SALK collections. Knockout plants were identified by immunoblot analysis of isolated thylakoids with CaS-specific antibody, and the D1-specific antibody was used as a control for equal protein loading (Fig. 5A). The specific absence of the 40 kDa phosphoprotein band in thylakoids isolated from knockout plants (Fig. 5B) provides definite evidence that this band represents CaS. To verify the lack of CaS transcripts in the mutant plants, RT-PCR analysis of mRNA from the mutant and wild-type plants was performed (Fig. 5C). The CaS knockout plants showed retarded growth even under normal unstressed conditions (Fig. 5D), indicating its important role in plant growth.

Figure 5.

 Phenotype revealed by the CaS knockout plants. Immunoblot analyses of thylakoids isolated from wild-type and CaS knockout plants using CaS-specific, D1-specific (A) or phosphothreonine-specific (B) antibody. (C) Ethidium bromide-stained gel with RT-PCR products showing no cas transcript in CaS knockout mutant lines and the presence of 18S rRNA in both mutant lines and the wild-type. (D) Retarded growth revealed by CaS knockout plants 3 weeks (upper panel) and 5 weeks (lower panel) after sowing the seeds.

To obtain further insights into the mechanisms responsible for the observed phenotype, we analyzed the photochemical efficiency of PSII by fluorescence measurements and the susceptibility of the CaS mutant to photoinhibition of PSII. However, no difference in the decrease of the variable fluorescence/maximal fluorescence (Fv/Fm) ratio during high light illumination (1500 μmol photon·m−2·s−1 for 3 h) or during subsequent recovery at low light (30 μmol photon·m−2·s−1 for another 3 h) was observed between the wild-type and the CaS mutant at any time point (supplementary Fig. S3). The whole chain electron transfer activities were also unaffected in the CaS mutant as compared to the wild-type (supplementary Table S3). As CaS is an intrinsic thylakoid protein, we then tested whether the absence of CaS exerts any effects on the composition of the thylakoid protein complexes. To this end, an immunoblot analysis was performed on the contents of representative proteins in different thylakoid protein complexes, including the PSI and PSII core complexes, ATP synthase, and the lumenal oxygen-evolving complex. This analysis revealed no significant changes in PSII, PSI and ATP synthase in the CaS mutant as compared to the wild-type (supplementary Fig. S4).

Sequence analysis and domain structure

The network-based tools targetp and chlorop ( strongly predict the CaS protein to be targeted to chloroplasts, with the transit peptide corresponding to residues 1–33 (Fig. 6A), which gives a molecular mass of 37.8 kDa for the mature protein. This calculated mass is in accordance with the MS identification of CaS in a gel region around 40 kDa, together with CYP38, FNR and several other known proteins (supplementary Table S2). The C-terminus contains two motifs: a noncatalytic rhodanese homology domain (amino acids 231–352), with the putative active residue Cys309 substituted by Asp, and a motif that is involved in interaction with 14-3-3 proteins and proteins with the ‘forkhead-associated’ (FHA) domain. These domains are found in a variety of signaling proteins, and can bind directly to the phosphothreonine residue [25]. The identified phosphorylation site, Thr380, of CaS lies within this motif (Fig. 6A).

Figure 6.

 Domain structure and homologous proteins of CaS. (A) Schematic representation of the domain structure of CaS. Polypeptide modules are indicated as follows: TP, chloroplast transit peptide; TM, transmembrane region; rhodanese-like, rhodanese homology domain; 14-3-3, motif for interaction with 14-3-3 proteins; FHA1, motif for interaction with forkhead-associated domain 1. The phosphorylated Thr380 is indicated by pThr. (B) Alignment of Arabidopsis CaS with the amino acid sequences of putative homologous proteins from higher plants and green algae. The lowercase ‘t’ above the sequence indicates phosphorylated Thr380. The predicted transmembrane domain is marked by a dashed line above the sequence.

CaS appears to be a plant-specific protein. It has homologs in Oryza sativa (gi:41352315) and Medicago truncatula (gi:92878521), as well as in the green algae Chlamydomonas reinhardtii (gi:46093489) and Ostreococcus tauri (gi:116059237) (Fig. 6B). No proteins with significant sequence similarity to CaS were found in cyanobacteria. According to hydropathy analysis (tmhmm at and sosui at, CaS in higher plants has one transmembrane helix (amino acids 188–210 in Arabidopsis), whereas the green algae proteins do not contain any transmembrane region. Alignment of protein sequences with clustalw (Fig. 6B) showed that phosphorylated Thr380 is conserved in homologous proteins of green algae.


CaS – a novel thylakoid phosphoprotein and a potential substrate of the STN8 protein kinase

The CaS protein (At5g23060) described here is a newly identified phosphoprotein in the thylakoid membrane of Arabidopsis, with its expression and phosphorylation level being strongly dependent on light intensity.

Studies of CaS (At5g23060) localization performed in onion epidermis using transient expression of a CaS–GFP fusion protein indicated the plasma membrane as the site of CaS localization [20]. However, the onion epidermis cells lack chloroplasts, and therefore the plasma membrane localization is inconclusive. Similarly, the use of human embryonic kidney cells for localization of CaS to the plasma membrane is questionable [21], as CaS is a plant-specific protein. To resolve the differences between those results and the present CaS localization to thylakoids, we performed immunoblot analysis of purified Arabidopsis plasma membrane with CaS-specific antibody, which clearly showed the absence of CaS in the plasma membrane (Fig. 2A). Neither was CaS found in the proteome study of Arabidopsis plasma membrane [26], whereas the respective studies with Arabidopsis thylakoids and mitochondria revealed the presence of CaS [22–24,27]. Moreover, we constructed the C-terminal YFP fusion of CaS and tested its subcellular localization in N. benthamiana. The overlap of CaS–YFP signal with chloroplast autofluorescence and the chloroplast-targeted control GWD1tp–GFP confirm chloroplast as the primary destination of CaS (Fig. 3).

Further subfractionation of thylakoids isolated from leaves exposed to high light and probing of these fractions with CaS-specific antibody showed that the majority of CaS protein is localized to the stromal thylakoids (Fig. 2B).

Evidence for CaS phosphorylation is provided by the mapping of the exact phoshorylation site, which corresponds to Thr380 in the C-terminus of the protein. Making use of two chloroplast protein kinase mutants of STN7 and STN8, it was possible to assign CaS as a likely substrate of the chloroplast-targeted STN8 protein kinase (Fig. 4A). As STN8 protein kinase phosphorylates stroma-exposed Thr residues of PSII core proteins [18,19], the C-terminus of CaS is most likely oriented to the stroma, where it can be involved in signal propagation from chloroplasts to other cellular compartments. STN8 kinase is selective for phosphorylation of easily accessible residues, such as N-terminal threonines of D1, D2, and CP43; this might be explained by long loops limiting access to the active site in the catalytic domain of STN8 [19]. The phosphorylation of CaS at the easily accessible C-terminus is in accordance with this selectivity of the STN8.

CaS is regulated at multiple levels according to environmental cues

The transcript level of CaS is significantly upregulated under normal growth irradiance as compared to darkness and low-light conditions [28]. Our results demonstrate that the high-light treatment increases the phosphorylation level of CaS, whereas the amount of the protein remains at the growth light level. Thus, CaS expression, and possibly its function, is tightly regulated by light at two levels: transcription, and post-translational modification by phosphorylation.

Physiological functions of CaS

CaS knockout mutants show clearly reduced growth as compared to the wild-type. As CaS is a thylakoid protein, it was first assumed that it possibly regulates the accumulation or stability of some thylakoid protein complexes. This, however, was not the case, as the contents of representative proteins in the four thylakoid protein complexes were not modified in CaS knockout mutants. Also, the light sensitivity of PSII, which is regulated by a number of thylakoid proteins [29], was unaffected in CaS knockout mutants. Therefore, the functional roles for CaS and its phosphorylation under stress conditions are more likely to be found in signaling cascades that coordinate the growth and responses of plants to environmental cues. The main location of CaS in stroma-exposed thylakoid regions is in line with its possible signaling function.

The stroma-exposed C-terminal part of CaS has a rhodanese-like protein domain (Fig. 6A). This domain, lacking the catalytic residues in some cases, is found in a wide variety of functionally distinct proteins in frequent association with other domain structures known to be involved in signal transduction [30], suggesting that CaS might play a role in sensing and signaling of environmental cues. It has been demonstrated that rhodanese domain proteins are associated with specific stress conditions, including the process of leaf senescence in Arabidopsis [31].

The C-terminus of CaS contains also a motif for interaction with 14-3-3 proteins and FHA domains, according to eukaryotic linear motif prediction at 14-3-3 proteins are known to function as adaptors that mediate protein–protein interactions and to be involved in signal transduction and stress responses and also in protein import into chloroplasts [32]. FHA domain proteins are directly involved in signal transduction, and the interaction between the FHA domain and target proteins is strictly dependent on phosphorylation of Thr residues of the target proteins [25,33]. The identified phosphorylation site of CaS at Thr380 is located within these predicted motifs, and its phosphorylation is intricately regulated by environmental cues. Although direct experimental evidence for such protein–protein interactions is still lacking, these structural features suggest a potential role of CaS protein in a signal transduction cascade sensing light or redox changes in chloroplasts and propagating the signal via direct protein–protein interactions.

Experimental procedures

Plant material and growth conditions

Arabidopsis ecotype Columbia (Col-0) was used for all other experiments except for the transient expression, which was carried out in tobacco. Plants were grown in a phytotron under the following conditions: 100 μmol photons·m−2·s−1 light intensity, 8 h photoperiod, 23 °C, and relative humidity 70%.

The T-DNA insertion lines of the stn7 gene (At1g68830) (SALK 073254) and the stn8 gene (At5g01920) (SALK 060869 and SALK 064913) in the Columbia background were obtained from the Salk Institute [34]. Plants homozygous for the T-DNA insertion were identified on the basis of PCR analysis [11,19].

The T-DNA insertion lines of the cas gene (At5g23060) (665G12 and SALK 070416) in the Columbia background were obtained from GABI-Kat [35] and Salk Institute collections [34]. CaS knockout plants were identified using purified CaS-specific antibody (see below).

Extraction of RNA and RT-PCR analysis

Total RNA of frozen leaf tissues was extracted with TRIzol (Invitrogen, Carlsbad, CA, USA). After RNase-free DNase treatment, 1 μg of total RNA was used to synthesize cDNA using SuperScript III reverse transcriptase (Invitrogen) in a 40 μL reaction volume. Four microliters (1/10) of RT product was used for PCR amplification with CaS-specific and 18S RNA control primers. The forward and reverse primers, respectively, for the 18S RNA were 5′-CTGCCAGTAGTCATATGCTTGTC-3′ and 5′-GTGTAGCGCGCGTGCGGCCC-3′. The forward and reverse primers, respectively, for CaS were 5′-AAATGGCAACGAAGTCTTCAC-3′ and 5′-CAGTCGGAGCTAGGAAGGAA-3′.

Isolation of plasma membrane, intact chloroplasts, stroma and thylakoids

The plasma membrane fraction of Arabidopsis was isolated as previously described [36]. Intact chloroplasts were isolated from mature Arabidopsis leaves using a two-step Percoll gradient [37]. The stroma fraction was obtained after chloroplast lysis in buffer and centrifugation at 15 000 g. Thylakoid membranes were isolated as described previously [38], including protease inhibitor cocktail (Complete; Roche, Mannheim, Germany). Thylakoids were subfractionated into grana, margin and stroma lamellae by using the digitonin method as previously described [6].

SDS/PAGE and immunoblotting

The proteins were separated by SDS/PAGE with 6 m urea and transferred to an Immobilon poly(vinylidene difluoride) membrane (Millipore, Bedford, MA, USA). The membranes were blocked with 5% (w/v) milk or BSA, and incubated with protein or phosphothreonine-specific antibody (polyclonal; New England Biolabs, Beverly, MA, USA). The amount of chloroplasts loaded in gels was tested for each antibody to give a linear response, and was varied between 0.5 and 5 μg of chloroplasts, depending on the antibody. The MicroLink Protein Coupling kit (Pierce, Rockford, IL, USA) was used for purification of CaS-specific antibody, raised against the full-length protein, kindly provided by Z. M. Pei (Duke University, Durham, NC, USA).

Phosphopeptide isolation

Isolated thylakoids were resuspended in 25 mm NH4HCO3 and 10 mm NaF to a final concentration of 3 mg of chloroplasts·mL−1 and incubated with MS-grade trypsin (Promega, Madison, WI, USA) (5 μg enzyme/mg chloroplasts) for 3 h at 22 °C. The digestion products were frozen, thawed, and centrifuged at 15 000 g. The supernatant was collected, and the membranes were resuspended in water and centrifuged again. The supernatants, both containing released thylakoid peptides, were pooled and centrifuged at 100 000 g for 20 min. The peptides were then lyophilized and methyl-esterified with 2 m methanolic HCl [39]. Phosphopeptides were enriched by IMAC as previously described [19], with modifications. The sample was first loaded on the IMAC column in 0.3% acetic acid in water; unbound peptides were lyophilized again, and loaded on the IMAC column in H2O/acetonitrile (ACN)/MeOH (1 : 1 : 1). Phosphopeptides were eluted with 4 × 10 μL of 20 mm Na2HPO4 with 20% ACN, and desalted using POROS R3 (PerSeptive Biosystems, Framingham, MA, USA).


In-gel trypsin digestion was performed as previously described [40]. Tandem MS was performed on an API QSTAR (Applied Biosystems, Foster City, CA, USA) equipped with a nanoelectrospray source (MDS Protana, Odense, Denmark) and connected in-line with the nano-HPLC system (LC Packings, Amsterdam, the Netherlands). Eluted and dried peptide samples were dissolved in 9 μL of 2% formic acid, centrifuged for 10 min at 12 000 g, and transferred to an autosampler vial. Aliquots (8 μL) of samples were loaded onto a C18 PepMap, 5 μm, 1 mm × 300 μm internal diameter nano-precolumn (LC Packing), desalted for 1.5 min, and subjected to reverse-phase chromatography on a C18 PepMap, 3 μm, 15 cm × 75 μm internal diameter nanoscale LC column (LC Packing). A gradient of 5–50% ACN in 0.1% formic acid was applied for 50 min with the flow rate of 0.2 μL·min−1. The acquisition of MS/MS data was performed on-line using the fully automated IDA feature of the analyst qs software (Applied Biosystems). The acquisition parameters were 1 s for TOF MS survey scans and 2–3 s for the product ion scans of two most intensive doubly or triply charged peptides. The major trypsin peptides were excluded from MS/MS acquisition. Analyses of MS/MS data were performed with the analyst qs software, and this was followed by protein identification by mascot with search parameters allowing for carbamidomethylation of Cys, one miscleavage of trypsin, oxidation of Met, and 200 p.p.m. mass accuracy. mascot search parameters in the case of phosphopeptide analysis allowed one miscleavage of trypsin, methylation of the C-terminus, Asp and Glut, and phosphorylation of Ser and Thr.

Fluorescence measurements at room temperature

PSII photochemical efficiency was determined as a ratio of Fv to Fm, measured from intact leaves with a Hansatech Plant Efficiency Analyser (Hansatech Instruments, King’s Lynn, UK) after a dark incubation for 30 min.

Construction of fluorescent protein fusions

The C-terminal YFP fusion of CaS was constructed by using a two-step USER cloning technique [41]. The CaS coding sequence (AY341888) was amplified by PCR using PfuTurbo CX Hotstart DNA polymerase (Strategene, La Jolla, CA, USA) and the uracil-containing primers nt114 (forward: GGCTTAAUATGGCTATGGCGGAAATGGCAACGA) and nt115 (reverse: GGTTTAAUTAAGGATCCTTAATTAAGCCTCAGCGGGTCGGAGCTAGGAAGGAACTT), where the underlined sequence was included for regeneration of a USER cloning cassette. The PCR product was mixed with the PacI/Nt.BbvCI-digested plasmid pCAMBIA330035Su and treated with USER enzyme mix (New England Biolabs) for 35 min at 37 °C and 25 min at 25 °C. The reaction mix was directly used to transform Escherichia coli DH10B chemically competent cells, the positive clone, pCAS, was obtained, and the correct insertion was verified by sequencing. A YFP fragment was amplified by PCR using the uracil-containing primers nt59 (forward primer: GGCTTAAUCTGGGTAGCGGTGGAATGGTGAGCAAGGGCGAGGAG) and nt34 (reverse primer: GGTTTAAUTTACTTGTACAGCTCGTCCAT). The product was mixed with the PacI/Nt.BbvCI-digested pCAS, treated with USER enzyme mix, and used to transform E. coli DH10B. The fusion construct, pCASYFP, was verified by sequencing and was subsequently introduced to Agrobacterium tumefaciens strain C58 pGV3850 for heterologous expression in tobacco. GWD1tp–GFP consisted of chloroplast transit peptide for glucan water dikinase 1 fused to GFP, and was used as a chloroplast marker.

Transient expression and subcellular localization in N. benthamiana

Overnight cultures of A. tumefaciens bearing appropriate plasmid constructs were harvested, resuspended in a buffer (100 μm acetosyringon, 10 mm MgCl2, 10 mm Mes, pH 5.6), and were incubated at room temperature for 2 h. The attenuance of each Agrobacterium strain was adjusted to 0.05 at 600 nm before infiltration. N. benthamiana was grown in a greenhouse for 4 weeks at 28 °C under 16 h of daylight and at 22 °C under 8 h of darkness. The Agrobacterium cell suspensions were infiltrated into leaves, and the plants were placed in a greenhouse. Observations of sections of the infiltrated leaves were carried out by 48 h after infiltration using a confocal scanning laser microscope (TCS SP2; Leica Microsystems, Wetzlar, Germany). Sequential scanning of GFP and YFP were carried out, with excitation at 488 nm and 514 nm, respectively, and emission at 495–510 nm and 545–600 nm, respectively. Chloroplast autofluorescence was detected at 650–707 nm. The scan speed was 800 Hz, and a line average of 8 was used.


The work was supported by the Academy of Finland, the Finnish Ministry of Agriculture and Forestry (the NKJ project), the Swedish Research Council for Environment, Agriculture and Space Planning (Formas), the Kone Foundation, and European Union FP6 contract 021313-Glytrans. We wish to thank Professor M. Sommarin for purified plasma membranes of Arabidopsis, Dr Z. M. Pei for CaS antibody, and Dr M. Glaring for the GDW1tp–GFP construct. We are grateful to the proteomics unit in the Turku Center of Biotechnology for maintenance of the MS unit.