Journal of Experimental Botany

JXB Advance Access originally published online on September 12, 2006
Journal of Experimental Botany 2006 57(14):3583-3594; doi:10.1093/jxb/erl104

This Article Abstract FREE Full Text (PDF) Supplementary Material All Versions of this Article: 57/14/3583    most recent erl104v1 E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (1) Request Permissions Disclaimer Google Scholar Articles by Yu, L. Articles by Moran, N. Search for Related Content PubMed PubMed Citation Articles by Yu, L. Articles by Moran, N. Agricola Articles by Yu, L. Articles by Moran, N. Social Bookmarking          What's this?

© The Author [2006]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org

RESEARCH PAPER

Phosphorylation of SPICK2, an AKT2 channel homologue from Samanea motor cells

Ling Yu^1 *, Dirk Becker^2 *, Hadas Levi¹, Menachem Moshelion¹, Rainer Hedrich², Ilana Lotan³, Arie Moran⁴, Uri Pick⁵, Leah Naveh¹, Yael Libal¹ and Nava Moran^1,

¹The Robert H. Smith Institute for Plant Sciences and Genetics in Agriculture, Faculty of Agricultural, Food and Environmental Quality Sciences, the Hebrew University of Jerusalem, Rehovot 76100, Israel
²Julius-von-Sachs-Insitute, Department of Botany I: Molecular Plant Physiology and Biophysics, University of Wuerzburg, Julius-von-Sachs-Platz 2, D-97082 Wuerzburg, Germany
³Department of Physiology and Pharmacology, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
⁴Department of Physiology, Faculty of Health Sciences, Ben-Gurion University, Beer Sheva, Israel
⁵Department of Biological Chemistry, Weizmann Instititute of Science, Rehovot 76100, Israel

To whom correspondence should be addressed. E-mail: nava.moran{at}huji.ac.il

Received 20 February 2006; Accepted 28 June 2006

	Abstract

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
SPICK2, a homologue of the weakly-inward-rectifying Shaker-like Arabidopsis K channel, AKT2, is a candidate K⁺-influx channel participating in light- and clock-regulated leaf movements of the legume, Samanea saman. Light and the biological clock regulate the in situ K⁺-influx channel activity differentially in extensor and flexor halves of the pulvinus (the S. saman leaf motor organ), and also—though differently—the transcript level of SPICK2 in the pulvinus. This disparity between the in situ channel activity versus its candidate transcript, along with the sequence analysis of SPICK2, suggest an in situ regulation of the activity of SPICK2, possibly by phosphorylation and/or by interaction with cAMP. Consistent with this (i) the activity of the voltage-dependent K⁺-selective fraction of the inward current in extensor and flexor cells was affected differentially in whole-cell patch–clamp assays promoting phosphorylation (using the protein phosphatase inhibitor okadaic acid); (ii) several proteins in isolated plasma membrane-enriched vesicles of the motor cells underwent phosphorylation without an added kinase in conditions similar to patch–clamp; and (iii) the SPICK2 protein was phosphorylated in vitro by the catalytic subunit of the broad-range cAMP-dependent protein kinase. All of these results are consistent with the notion that SPICK2 is the K⁺-influx channel, and is regulated in vivo directly by phosphorylation.

Key words: AKT2, gating, kinase, motor cells, nucleotides, phosphorylation, PKA, potassium channel, selectivity, Samanea

	Introduction

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
In the legume Samanea saman, the membrane permeability to K⁺ of motor cells in the leaf-moving organs, pulvini, is regulated by light and the biological clock (Lowen and Satter, 1989; Kim et al., 1992, 1993; Moran et al., 1996; Suh et al., 2000). K channels in the pulvinar cell membranes are the most likely conduits for the rhythmic transmembrane K⁺ fluxes constituting part of the osmotic motor (Moran et al., 1988; Yu et al., 2001). This, together with the rhythmic regulation of membrane permeability to water (Moshelion et al., 2002a), is very probably the basis for the diurnal regulation of the S. saman leaf movements (reviewed by Satter and Galston, 1981; Satter et al., 1988; Moshelion et al., 2002b). SPICK1 and SPICK2 (accession nos AF099095 and AF145272; 90% identical to one another) figure prominently among the ‘candidate channels’ for the K⁺-influx pathways in the Samanea motor cells because (i) they are 64% and 59%, respectively, identitical to AKT2, (Moshelion et al., 2002b); (ii) the in situ recorded inward K⁺ currents via K_H channels (described by Yu et al., 2001) resembled the already reported inward K⁺ currents mediated by members of the AKT2/3 subfamily expressed in heterologous systems, such as AKT2/3, ZMK2, and PTK2 (Marten et al., 1999; Philippar et al., 1999; Bauer et al., 2000; Langer et al., 2002; Ivashikina et al., 2005; see also the plant K channel classification by Pilot et al., 2003a; Very and Sentenac, 2003); (iii) the SPICK1 and SPICK2 genes, found in repeated scans of the cDNA library of the pulvinar tissues, were the only K⁺-selective influx channel-like genes (Moshelion et al., 2002b); (iv) the requirement for an efficient K⁺-efflux channel appears to be satisfied by the discovery of the pulvinar SPORK1—a Samanea homologue of SKOR and GORK (Moshelion et al., 2002b); and (v) the transcripts of SPICK1 and SPICK2 fluctuated in a diurnal rhythm in the pulvinar tissues, even in continuous darkness, suggesting their involvement in the rhythmic K⁺ fluxes in the pulvinus (Moshelion et al., 2002b).

The transcript level fluctuations of SPICK1 and SPICK2 suggested that K⁺ influx might be regulated through increased expression and abundance of their proteins. However, while the period of the rhythm of the fluctuations of the transcript level of SPICK2 in flexors (Moshelion et al., 2002b), responsible for pulvinus bending and leaf folding, corresponded satisfactorily to the rhythm in the activity of the K⁺-influx pathways (presumably the K_H channels; Kim et al., 1993), its phase did not: the mRNA level of SPICK2 peaked in the morning in the shrinking (i.e. K⁺-releasing) flexor tissue of the pulvinus, when and where K⁺-influx channels were expected to be inactive, and it was low in the evening, when and where K⁺-influx channels were expected to be active (Satter et al., 1988; Kim et al., 1993; Moran et al., 1996). In extensors, SPICK2 mRNA did not seem to fluctuate during continuous darkness. The ‘subjective morning’ swelling of extensor cells, responsible for pulvinus straightening and leaf unfolding, was preceded by a ‘subjective night’ increase of the mRNA level of SPICK1 in this tissue (Moshelion et al., 2002b). Thus, if the K⁺-influx channels (at least in flexors) are the SPICK2 protein product, either the high mRNA levels in flexors are translated into protein with a considerable delay or this protein is also regulated post-translationally, for example, by Ca²⁺, and/or by phosphorylation. To prove the latter, it is important to demonstrate (i) that the K_H channel activity is modulated in situ in response to specific light stimuli and/or to the clock phase; (ii) that phosphorylation (or dephosphorylation) mimics the effects of the original stimuli on these changes in activity; (iii) that the channel is capable of interacting directly with a kinase; and (iv) that it actually does so, i.e. becomes phosphorylated, in physiological conditions. Attaining these four objectives depends on the availability of means to detect the K_H channel protein in the plant tissue, or in tissue extracts, which, in turn, depends on the molecular identification of the K_H channel.

Following the achievement of the first objective by Kim et al. (1993), the second and the third objectives are addressed here: the sequence of SPICK2 was examined for potential modulation sites, and the resulting predictions on the in situ activity of the K_H channels were assayed using potential modulators—cAMP and phosphorylation—according to the second objective. Then, according to the third objective—the susceptibility to phosphorylation of the SPICK2 protein itself—the candidate channel was affirmed. The results contribute towards identifying the SPICK2 gene product with the K_H channel (see also the discussion by Yu et al., 2001).

	Materials and methods

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Plant material
Samanea saman Merr. trees [referred to also as Pithecellobium saman Benth. (Little and Wadsworth, 1964)] were grown in a greenhouse as described by Yu et al. (2001).

Protoplasts were isolated separately from extensor and flexor regions of the secondary terminal pulvinus of the second and third mature leaves (counting from the branch tip) during hours 2–6 of light (see the Samanea motor protoplast isolation procedure in the Supplementary data at JXB online).

Patch–clamp recording and data processing
Patch–clamp experiments, including filtering and sampling frequency, were performed as described earlier (Yu et al., 2001), using a Digidata 3122A interface and pClamp 8.2 program suite from Axon Instruments (Union City, CA, USA). Inward K⁺ currents were recorded in extensor and flexor protoplasts using patch–clamp in the whole-cell configuration, essentially as described by Yu et al. (2001). To aid in resolving the reversal potential of K⁺ from that of Cl^–, most of the internal Cl^– was substituted with gluconate (see Solutions, below). As previously, the whole-cell patch–clamp configuration was established with 5 mM KCl in the bath and then the bath solution was switched to the recording solution containing 150 mM KCl. To improve the chances of a good seal between the protoplast and the patch pipette, the bath contained ‘patching’ solution (see Solutions, below), and, after attaining a whole-cell configuration, the bath was flushed with 15 vols (5 ml) of ‘recording solution’. The ground-electrode bridge was filled with the ‘recording solution’. The pipette resistance in the ‘patching solution’ ranged between 5 and 10 M. The series resistance of the patch pipette ranged between 25 and 50 M (44 M on average). Since the recorded currents rarely exceeded 300 pA, the error in voltage clamping was usually below 15 mV. Moreover, in most cases, the mean currents compared in two different conditions differed by only a few tens of pA, which, multiplied by 50 M (worst case), yielded a possible difference in clamped membrane potential of <2 mV. When the differences in current were larger, the series resistance was either compensated online, or corrected during analysis by subtracting from each nominal E_M value the voltage drop due to the series resistance (i.e. the total membrane current at this E_M times the series resistance); G_K was then calculated using the corrected driving force (the reversal potential determination is usually much less prone to such errors because of the usually smaller currents) and replotted versus the corrected E_M. For averaging, the current values were interpolated, when necessary, to standard E_M values using the individual fitted Boltzmann curves. The Boltzmann equation, G_K = G_max/(1 + e^{–(EM–E1/2)*k}), was fitted to the G_K–E_M relationships of individual cells (k is ‘slope’, i.e. the steepness of the voltage dependence of G_K, and E_1/2 is the E_M at half-maximum G_K). The best-fit individual parameters values (G_max, E_1/2, and k) of all the cells of a treatment group were averaged.

The liquid junction potential resulting from the interphases between the bridge, bath, and pipette solutions before attaining the whole-cell configuration (see Solutions below) was measured, separately from the experiments, with 3 M KCl/agar bridges (Neher, 1992), verified by calculation, using the Clampex 8.2 program (Axon Instruments/Molecular Devices Co., Sunnyvale, CA, USA) and its value of 6 mV was added online (using the manual ‘holding potential’ dial capability of the patch–clamp amplifier) to the nominal value of the membrane potential.

Raising SPICK2 protein in Sf9 cells
Insect cell culture and inoculation:
Insect cells (Sf9, Spidoptera frugiperda cell line) were grown in 10 cm diameter Petri dishes (Nunc, cat. # 150350, Denmark) in a monolayer culture in GRACE's medium (cat. # 01-155-1A, Biological Industries, Beit-Haemek, Israel) supplemented with 10% fetal calf serum (cat. # 04-121-1A, Biological Industries), and 50 µg ml^–1 gentamycin (cat. # 03-035-1C, Biological Industries). Cells were incubated at 27–28 °C in a humidity chamber. SPICK2 (or KAT1 as a control) was cloned into the GFP-containing pFastBac Dual vector (Invitrogen, http://www.invitrogen.com/content.cfm?pageid=4015) under the control of the polyhedrin or p10 promoter, respectively. The recombinant virus DNA was obtained using the FastBac system according to the manufacturer's instruction. The recombinant virus was kept at 4 °C for 3 months, and for longer periods, at –70 °C. Upon reaching 70% confluence, the Sf9 cells were inoculated with the virus, according to an instruction manual (Gruenwald, 1996). The success of infection was indicated by the cytosolic expression of green fluorescent protein (GFP).

Crude membrane isolation and analysis of expressed SPICK2 protein:
Infected Sf9 cells were harvested 3–4 d post-infection by forcefully flushing the culture Petri dish with homogenization buffer (see Solutions below). Cells were pelletted by centrifugation at 1000 g for 10 min and the pellet from each dish of cells was resuspended in 2 ml of ice-cold Sf9 homogenization buffer. The resuspended cells were incubated on ice for 30 min and mechanically lysed with a polytron (Ystral Gmbh, Dottingen, Germany). The homogenate was centrifuged at 1000 g for 5 min at 4 °C to remove nuclei and non-lysed cells. The membranes were pelletted by centrifuging the supernatant at 100 000 g for 30 min. This membrane fraction was resuspended in 300 µl of homogenization buffer. The protein content was determined with Bio-Rad DC protein assay kit (cat. #500-0116, Bio-Rad) according to the manufacturer's instructions based on Lowry et al. (1951). The final suspension usually contained 3–4 µg µl^–1 protein.

The SF9 protein was resolved on a polyacrylamide gel (see below) and stained with Coomassie brilliant blue. An outstanding band (absent in a lane with protein from non-transfected SF9 cells) was excised from the gel and analysed at the Smoler Proteomics Center (Technion, Haifa, Israel), where it was subjected to trypsinization, and then underwent liquid chromatography–tandem mass spectrometry. The seven identified peptides (TWIGTTNPSFK, NNIEAATNFVSR, EILLSLAAK, DLVIDWSR, ILYQLCVQSDPHTAGDLLCK, DAMVTDEGGAEIDSIDLIR, and DNDKLFIVE) were compared with the available database including the predicted SPICK2 sequences, and were uniquely traced to their respective positions in the predicted SPICK2 sequence (249–259, 331–342, 404–412, 527–534, 624–643, 783–801, and 802–810), thus allowing the unique identification of the protein as SPICK2.

The predicted amino acid sequence of SPICK2 was examined using the program PROSITE (Release 17.49, of 1 June 2003, on http://www.expasy.ch/tools/scanprosite/).

SPICK2 phosphorylation
Phosphorylation consensus sites:
The ‘PhosphoBase’ database (version 2.0) was previously found at http://www.cbs.dtu.dk/databases/PhosphoBase/predict/predform.php, and is now hosted by EMBL at http://phospho.elm.eu.org/.

Antibody production:
Anti-SPICK2 antibody was produced by immunizing rabbits with a synthetic 22 amino acid peptide (amino acids 11–32 of the predicted SPICK2 protein sequence). The peptide was produced by Gramsch Laboratories (Schwabhausen, Germany), and certified as showing only a single peak using mass spectroscopy. Purified anti-Spick2 antibody was obtained by passing the ant-SPICK2 immune serum via an immunoaffinity column prepared by binding the immunizing peptide to SulfoLink Coupling Gel, according to the manufacturer's instructions (cat. #.20402, Pierce Biotechnology, Rockford, IL, USA).

Immunodetection of SPICK2:
To test the antibody, 50 µg of Sf9 membrane proteins, SPICK2, or KAT1, as a control, were dissolved in SDS sample buffer and heated to 65 °C for 3 min, then resolved by 10% SDS–PAGE according to the method of Laemmli (Benga, 2003). The proteins on the SDS–polyacrylamide gel were transferred to a Hybond-P PVDF membrane (Pack No. RPN303F, Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK), after activating the membrane according to the manufacturer's instructions in methanol for 20 s, and then washing it five times for 1 min with H₂O. After 1 h block with 5% of skimmed milk in TBS-T buffer, the blot was exposed overnight to the affinity-purified anti-SPICK2 antibody at a 1:1000 dilution, at 4 °C, followed by 5x 10 min washes with TBS-T, then, for 1 h, by an exposure to a second antibody [donkey anti-rabbit horseradish peroxidase-linked whole antibody (Amersham Life Science, UK)] at a 1:10 000 dilution. The immunoblot was then washed 5x 10 min with TBS-T, then reacted with ECL-plus (RPN 2132, Amersham Pharmacia Biotech), and exposed to Fuji medical X-ray film SUPER RX (Fuji Photo Film, Tokyo Japan).

SPICK2 immunoprecipitation:
An aliquot of 30–40 µl of crude Sf9 membrane final suspension, containing 100 µg of protein, was mixed with 10% SDS to a final concentration of 1% SDS and heated to 60 °C for 3 min. TBS/1% Triton X-100 (500 µl) was added and the mixture was shaken for 10 min at 4 °C. The sample was centrifuged for 10 min at 13 000 g in a microfuge, and the supernatant was collected and incubated overnight with 5 µl of anti-SPICK2 serum with gentle shaking at 4 °C. The antigen–antibody complex was then precipitated by mixing with 5 mg of protein A–Sepharose CL4B (cat. # 17-0780-01, Amersham Pharmacia Biotech AB), prewashed with H₂O and TBS/1% Triton X-100, followed by shaking at 4 °C for 1 h and centrifugation at 13 000 g for 1 min.

Phosphorylation of SPICK2 with the PKA catalytic subunit:
The protein A–SPICK2 pellet complex was washed three times with 1 ml of TBS/1% Triton X-100 and with 1 ml of cAMP-dependent protein kinase (PKA) phosphorylation buffer (see Solutions, below). Phosphorylation was performed by incubation of the pellet in the phosphorylation buffer and the addition of 2 U of the catalytic subunit of PKA [kept frozen in a stock of 2 mg ml^–1 in 0.5 mg ml^–1 of bovine serum albumin (BSA) and 5 mM dithiothreitol (DTT)] and 10 µCi of [-³²P]ATP (3000 Ci mmol^–1) for 30 min at 30 °C. The reaction was stopped by washing twice with 0.5 ml of TBS/1% Triton X-100. The proteins were dissociated by heating to 95 °C for 5 min in 40 µl of DTT sample buffer and resolved by 10% SDS–PAGE. The synthetic peptide fragments of the PKA inhibitor known as PKI (residues 6–22; kept frozen in 100 µM stock solutions in water) was mixed with PKA, at a 1:5 ratio (v/v) and incubated at 30 °C for 30 min; this mixture was then used at a 6:100 (v/v) ratio in some of the phosphorylation reactions.

Autoradiography and signal analysis:
The dried gel or membrane blot was exposed to a phosphorimage intensifier plate, usually for several hours (or overnight), sometimes for a few days, and the ³²P signals were then read by phosphorimager (model: FUJI FILM FLA-5000, Tokyo, Japan): and saved in an original format.

Isolation and ‘auto’phosphorylation of pulvinar cell membranes
Preparation of vesicles:
Vesicle preparation was conducted wholly on ice. Extensor and flexor tissue strips, excised from the terminal secondary pulvini of 140 second–eighth mature leaves of S. saman (1 g fresh weight each) were snap-frozen separately in liquid N₂, then finely crushed in a clay mortar with 1 ml of HM (homogenization medium; see Solutions, below), added gradually in small droplets, then transferred into a glass homogenizer (size C), using washes of HM, up to a total volume of 7 ml. The tissue was then homogenized with a teflon pestle for 5 min. The homogenate was centrifuged at 10 000 g for 10 min to remove various organelles. The supernatant was centrifuged at 80 000 g for 30 min. The pellet was resuspended in SB (solubilization buffer; a total volume of 1.3 ml), using a glass/teflon homogenizer (size A) for 1–2 min. The homogenate was layered on top of a two-phase sucrose gradient [34/40% (w/w), see Solutions, below] and centrifuged for 2 h. The plasma membrane-enriched fraction was collected at the interface (1 ml), resuspended in 10 ml of SB, centrifuged for 1 h, and the pellet was resuspended in 50 µl of SB to a final volume of 150 µl and kept frozen in liquid N₂ until use. The protein yield, determined by the modified Lowry method (Markwell et al., 1978), was 1.6–1.8 µg protein mg^–1 fresh tissue.

Phosphorylation of membrane vesicles:
Plasma membrane vesicles were thawed, resuspended in 10 ml of RS (‘reaction solution’; see below), centrifuged at 80 000 g for 30 min, resuspended in 150 µl of RS, and dispensed into Eppendorf vials in 50 µl aliquots. H7·HCl [1-(5-isoquinolinesulphonyl)-2-methylpiperazine hydrochloride; Research Biochemicals International, Natick, MA, USA], 1 µl out of a 5 mM stock in H₂O, was added to one of the tubes, to a final concentration of 100 µM. Okadaic acid (OA; free acid, cat. # 0-800, Alomone Labs Ltd, Jerusalem, Israel) was added to the second tube, after a two-stage dilution [immediately before use: 30-fold dilution into H₂O from a 0.3 mM stock in dimethylsulphoxide (DMSO), then 50-fold dilution into the reaction medium to a final concentration of 200 nM (the final DMSO concentration in the reaction medium was 0.07%)]. The third tube served as a control. The vesicles were incubated at room temperature (23 °C) for 10 min, then [-³²P]ATP was added to each tube (a total of 30 µCi in 3 µl in all the vials, specific activity 3000 Ci mmol^–1) and the incubation continued for an extra 30 min. The incubation was terminated by the addition of 300 µl of ‘stop solution’ (see below) and centrifugation at 100 000 g for 40 min. The pellet was washed again in 100 µl of stop solution, then resuspended in 50 µl of 4% SDS sample buffer, boiled at 100 °C for 5 min, and 40 µl of each protein suspension underwent an 8% PAGE together with standard molecular mass markers (205–29 kDa). The gel was stained with Coomassie blue, then dried, and autoradiographed (4–7 d exposure).

Solutions
cAMP (cat. # A9501, Sigma) was dissolved, just before use, in internal solution from a stock of 100 mM in H₂O, to a final concentration of 100 µM.

DTT sample buffer: 50 mM Na₂CO₃, 56 mM DTT, 1.75 M sucrose, 2% SDS, 0.04% bromophenol blue.

HM, homogenizing medium (in mM): 800 sucrose, 5 EDTA, 50 TRIS/MES, at pH 8, 5 DTT, 1 PMSF (phenylmethylsulphonyl fluoride), 5 ascorbic acid, and 0.6% PVP (polyvinylpyrolidone)-40, the last three added just before use.

Internal solution: 28 mM KCl, 60 mM K-gluconate, 1 mM MgCl₂, 20 mM HEPES, 2 mM ATP-K₂, and 2 mM BAPTA-K₄, 12 mM of NMG, pH 7.8, osmolarity adjusted to 750±10 mOsm with D-sorbitol.

OA free acid (cat. # 0-800, Alomone Labs, Jerusalem, Israel) was kept frozen lyophilized, or in stock (reusable, 5 µl aliquots) of 0.3 mM in DMSO.

OA sodium salt (cat. # 459620, Calbiochem) was dissolved in internal solution on the day of the experiment, just before use, from a stock of 5 mM in H₂O, to a final concentration of 5 or 300 nM.

Patching solution: 5 mM KCl, 0.3 mM CaCl₂, 10 mM HEPES, adjusted to pH 7.2 with 4 mM N-methylglucamine (NMG). The osmolarity of the solution was increased to 700±10 mOsm by D-sorbitol.

PKA phosphorylation buffer: 25 mM HEPES, 5 mM MgCl₂, 2 mM EGTA, 0.5 mM CaCl₂, and protease inhibitor, added just before use. pH=7.4.

RS, reaction solution included (in mM): 125 KCl, 20 HEPES, adjusted with 8.5 NMG to pH 7.2, 1 MgCl₂, 2 BAPTA-K₄, 1 CaCl₂ (free Ca²⁺ was 200 nM, based on the dissociation constant of BAPTA (of 10^–7 M, in the presence of 1 mM Mg²⁺ and 100 mM KCl) and 0.1 PMSF, the latter added just before use.

Recording solution: 150 mM KCl, 0.3 mM CaCl₂, 10 mM HEPES and 4 mM NMG, pH 7.2, osmolarity adjusted to 700±10 mOsm by D-sorbitol.

SB, solubilizing medium: (in mM): 300 sucrose, 5 EDTA, 1 TRIS/MES at pH 7, 1 DTT and 0.2 PMSF, the last two added just before use.

Sf9 homogenization buffer: 20 mM TRIS–HCl, 5 mM EDTA, 5 mM EGTA (pH 7.4), and complete^TM EDTA-free protease inhibitor cocktail (one tablet for 50 ml of solution) (cat. # 1873580, Roche, Basel, Switzerland).

Stop solution included (in mM): 125 KCl, 20 HEPES, adjusted with 8.5 NMG to pH 7.2, 50 NaF, 10 EDTA, 0.5 orthovanadate, and 1 PMSF, the latter added just before use.

TBS: 20 mM TRIS, 150 mM NaCl, pH 7.5.

4% SDS sample buffer (5x): glycerol 5 ml, SDS 1 g, ß-mercaptoethanol 2.56 ml, TRIS–HCl 0.5 M, pH 6.8, 2.13 ml, bromophenol blue, a trace; kept in aliquots of 1 ml at –20 °C.

	Results

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Phosphorylation affects K_H channels differentially in extensors and flexors
Square hyperpolarizing (negative) voltage pulses evoked inward (K_H) currents with two characteristic components: instantaneous currents, termed I_L, or ‘leak-like’ currents, and time-dependent currents, termed I_K currents (Yu et al., 2001). In the present work, the time-dependent currents, I_K, were 50% smaller in extensors and 30% smaller in flexors than those described by Yu et al. (2001), as might be expected based on a 25% smaller K⁺ concentration near the membrane in the present experiments and a lower pH in the bath (7.2 in the present experiments versus 7.8 previously (Fig. 2A, B, open circles). To decrease the contamination of I_L by non-specific leak current originating from insufficiently tight sealing between the patch pipette and the plasma membrane, only cells which had initial sealing resistance >5 G (conductance of <0.2 nS) before establishing the whole-cell configuration were included. Thus, while the instantaneous conductance reported in previous experiments ranged up to 2 nS in extensors and 1.5 nS in flexors (Yu et al., 2001), in the present experiments the respective conductance values did not exceed 0.2 nS (the slope of the control I–V in Fig. 2D, E).

View larger version (15K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 2 The effect of okadaic acid (OA) on KH channel activity. I–EM relationships of whole-cell currents (mean ±SE), without (Cntrl) or with 5 or 300 nM OA in the patch pipette solution, as indicated by the symbols; the number of cells at each treatment is indicated in the figure. (A, B) Time-dependent K+ currents, IK, from extensor and flexor protoplasts, at steady state, at the end of 4 s pulses. Note that the current scale for flexors is double that of extensors. (C) The effect of 5 nM OA on KH channel gating in flexors: chord conductance (GK)–EM relationships (mean ±SE) of control and 5 nM OA-treated flexor cells. EM values were corrected for series resistance error (see Materials and methods). Lines are Boltzmann functions drawn using the mean best-fit parameters resulting from fitting the GK–EM relationships of individual cells (see Materials and methods). Arrows mark the mean best-fit E1/2 values. Inset: best-fit values of Gmax of the data of (A) (mean ±SE). The asterisk indicates a significant difference (P <0.05). (D, E) IL: instantaneous currents (mean ±SE) from cells of (A) and (B), respectively.

 
To test the dependence of the K_H channel activity on phosphorylation, OA—a known inhibitor of serine-threonine phosphatases 1 and 2A (Ishihara et al., 1989)—was introduced into the patch pipette, and the time-dependent currents recorded between 25 and 35 min after establishing an open-cell contact (the lag period included changing the bath solution, and testing for the establishment of stable current amplitudes, see above) were examined. With 300 nM OA in the pipette, the time-dependent inward currents in extensors at potentials more negative than –95 mV were smaller by 50–80% relative to the control and, in flexors at potentials more negative than –125 mV they were smaller by 60–70% (Fig. 2A, B, filled circles). For example, in extensors with 300 nM OA at –170 mV, the mean steady-state I_K current was –45.5±11 pA (n=7) versus –95.5±19.4 pA (±SE, n=10) in control extensors, and in flexors with 300 nM OA at –170 mV it was –38.3±15.0 pA (±SE, n=7), as compared with –98.5±16.8 pA (n=8) in control flexors (Fig. 2).

The effect of 5 nM OA on K_H channels in extensors was not significant, but, surprisingly, in flexors, 5 nM OA significantly enhanced the inward time-dependent K⁺ currents at membrane potentials of –170 and –155 mV (P <0.05; Figs 1, 2A, B, triangles). For example, the mean steady-state I_K current at –170 mV in 5 nM OA-treated extensors was –72.2±18.2 pA (n=6; not different from control), while in flexors it was –180.3±32.4 pA (n=7), almost double that in the control cells: –98.5±16.8 pA (n=8).

View larger version (28K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 1 Representative inward membrane currents from flexor cells recorded using patch–clamp in a ‘whole-cell’ configuration. The superimposed traces were elicited by a series of pulses to the indicated membrane potentials, with okadaic acid (OA, concentrations in nM) in the patch pipette (Materials and methods). Downward pointing arrowheads mark instantaneous currents (IL) and upward pointing arrowheads mark steady-state currents (IK+IL). Note the increased currents at the low OA concentration and the inhibition of currents at the high OA concentration.

 
To resolve more specifically how 5 nM OA increased flexor K_H channel activity, the chord conductance (G_K) was calculated from the time-dependent currents of each cell at all membrane potentials tested, and each cell's own reversal potential was also calculated, and the individual G_K–E_M relationships of control cells and 5 nM OA-treated cells were fitted with the Boltzman equation (Materials and methods). This analysis revealed that in flexor cells, 5 nM OA more than doubled the mean maximum (asymptotic) conductance, G_max, of the K_H channels (from 0.80±0.21 nS, n=5, to 1.88±0.37 nS, n=5; Fig. 2C, inset). Notably, the individual E_1/2 values were within the G_K–E_M ranges, justifying the extrapolation to the asymptotic value of G_max. In contrast, the voltage dependence of the channel gating did not change. This is evident (i) from the similar values of membrane potential, E_1/2, at half-maximal conductance (Fig. 2C, arrows; these values were, respectively, –142.7±5.4 mV in the control cells and –153.3±2.7 mV in the OA-treated flexor cells); and (ii) from the sigmoid ‘slope’ values (which were 20.9±1.9 and 26.7±2.5, respectively). In other words, in flexors, 5 nM OA increased the total number of active I_K-type channels and/or their single channel conductance, while each membrane potential activated the same fraction of these channels as without OA.

This I_K-enhancing effect of OA could be due to a phosphorylation-induced conversion from ‘instantaneous’ I_L-type, to ‘time- and voltage-dependent’ I_K-type channels, as demonstrated to occur in the brain KCNK2 channels as a function of their phosphorylation (Bockenhauer et al., 2001) or dephosphorylation as suggested for AKT2 channels by Dreyer et al. (2001), and subsequently recently demonstrated by Michard et al. (2005a). To check whether the K⁺-influx channel of the pulvinar motor cells may indeed be interconverting between these two forms, the OA effect on I_K and I_L of both extensor and flexor cells was compared. In contrast to the mean steady-state I_K, however, the mean I_L from these cells, in both extensor and flexor, did not appear to be affected by OA (Fig. 2D, E).

Phosphorylation of membrane proteins can be cytosol independent
It is increasingly recognized that the interaction of channels with kinases occurs within multipartite complexes associated with membranes (as in Folco et al., 2004; Marble et al., 2005). Thus, if phosphorylation directly modulates the in situ K_H channel activity in the patch–clamp experiments, it would, most probably, interact with a membrane-associated kinase. Such membrane-associated kinases would be expected to impart an ‘auto’-phosphorylating potency even to cytosol-devoid isolated membranes. Indeed, membrane vesicles, enriched in plasma membrane vesicles on a discontinuous sucrose gradient, everted (i.e. induced to change their sidedness to inside-out) using Brij-58 detergent (Johansson et al., 1995) and incubated with [-³²P]ATP in a solution similar to the ‘internal solution’ used in patch–clamp experiments (compare ‘RS, reaction solution’ with ‘internal solution’, in Solutions, Materials and methods), allowed the phosphorylation of several proteins (Fig. 3). Notably, the pattern of phosphorylation was clearly different between extensor- and flexor-originated proteins, consistent with the functional asymmetry of the two tissue types. In both extensors and flexors, phosphorylation was enhanced in the presence of 300 nM OA during the reaction and was partially inhibited by H7, a general inhibitor of many kinases (Hidaka et al., 1984). At least some of these unidentified protein bands could, in principle, represent phosphorylated transport proteins, perhaps even channels. Thus, plasma membrane-enriched vesicles appear to possess an intimately associated kinase function, consistent with a requirement for direct phosphorylation of the channel protein and compatible with conditions of whole-cell patch–clamp recording.

View larger version (164K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 3 In vitro ‘auto’-phosphorylation of motor cell plasma membrane vesicles. Autoradiogram of an SDS-polyacrylamide gel with electrophoresed proteins solubilized from a plasma membrane-enriched fraction from S. saman pulvinar tissues. The vesicles were pre-incubated first, for several minutes, with okadaic acid (OA), a general kinase inhibitor (H7), or with a control buffer (C) and, for an additional 20 min, with added [-32P]ATP (see Materials and methods). The results are representative of three independent experiments.

 
SPICK2 protein can undergo direct phosphorylation
SPICK2 protein and anti-SPICK2 antibody:
Direct purification of K⁺ channels from plant tissue is highly challenging due to the very low abundance of these proteins in the membrane (Dreyer et al., 1999, and references therein). Therefore, to test the possibility that the K_H channel itself is phosphorylated directly, and hypothesizing that SPICK2 is the K_H channel, SPICK2 protein was expressed heterologously in Sf9 cells (Supplementary Fig. S-1 available at JXB online). KAT1 protein was also expressed similarly, for control experiments. The expression of a GFP protein, on the same plasmid but under a separate promoter (see Materials and methods), served to estimate the degree of transfection. Immune serum, raised against an immunizing peptide based on the N-terminus of the predicted SPICK2 sequence (Supplementary Fig. S-2 available at JXB online), precipitated the SPICK2 protein with great affinity out of the Sf9 microsomal proteins (Supplementary Fig. S-3A available at JXB online); the anti-SPICK2 antibody (purified on an affinity column using the immunizing peptide) detected the SPICK2 protein with high sensitivity and specificity in western blots, at 85 kDa (i.e. somewhat lower than 92.55 kDa, the expected molecular mass of the predicted protein sequence of 810 amino acids; Supplementary Fig. S-3B, C available at JXB online). Occasionally, an additional couple of bands could be observed in the immunoblots at >100 kDa (Supplemetary Fig. S-3C available at JXB online).

To prove the antibody specificity, the 85 kDa band was excised from the gel and its exact sequence was confirmed by mass spectrometry (see Materials and methods). In situ studies, using the purified anti-SPICK2 antibody, allowed the detection of the recombinant protein in the plasma membrane of SPICK2-expressing Sf9 cells, and/or in the vicinity of the plasma membrane (Supplementary Fig. S-4 available at JXB online). Therefore, for further studies, the Sf9 membrane fraction was used (see Materials and methods).

Kinases compatible with SPICK2 phosphorylation:
Out of 52 serine residues and 43 threonine residues in the predicted amino acid sequence of SPICK2 (Supplementary Fig. S-2 available at JXB online), 19 serine and three threonine residues were predicted by the ‘NetPhos’ program (version 2.0, Blom et al., 1999) to be potential phosphorylation sites. Of these, on the basis of the ‘PhosphoBase’ database (see Materials and methods), seven sites were predicted to be susceptible to phosphorylation by a cAMP-dependent kinase (S-61, RNSSPHH; S-154, RYLSTWF; S-212, RFSYFL; S-269, RYISAMY; T-483, RTKTLTQ; T-492, RLRTSDF; and S-739, RVSIFR; Supplementary Fig. S-2 available at JXB online), and one by a cGMP-dependent protein kinase (S-6, KKDSGSS). The general consensus sequences were R-X1-2-S/T-X or [R/K]2-3-X-S/T-X, respectively (Kennelly and Krebs, 1991); the classical, loose sequence R/L-R/L/X-X-S (Shabb, 2001; Gerhardstein et al., 1999) has not been included in this count. In addition, six possible calcium-dependent protein kinase (CDPK) phosphorylation sites (K/R/H_XX_S/T; A Harmon, personal communication) have been found, four of these overlapping with a PKA site (Fig. 4 and Supplementary Fig. SS-2 available at JXB online).

View larger version (46K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 4 Schematic representation of the predicted SPICK2 topology in the membrane and its functional motifs: six transmembrane domains (S1–S6), a pore, ankyrin domains (ANK), cyclic nucleotide-binding domain (CNB), PKA and CDPK recognition sequences. A series of plus signs represents the channel ‘voltage sensor’, S4 (see also Supplementary Fig. S-2 available at JXB online, and Materials and methods).

 
In vitro phosphorylation by PKA:
Because the predicted protein sequence of SPICK2 contains consensus motifs for phosphorylation by PKA, the susceptibility of the SPICK2 protein to phosphorylation by the commercially available catalytic subunit of PKA was assayed, using the method described by Ivanina et al. (1994). The protein was specifically precipitated by anti-SPICK2 immune serum IgGs captured by protein A beads (Im) and subsequently detected by the affinity-purified anti-SPICK2 antibody (see Materials and methods) in a western blot (Fig. 5A, lanes 1–3, arrow). Interestingly, in all of the experiments employing immunoprecipitation, the intense bands attributable to SPICK2 appeared on the immunoblots close to (and slightly above) the predicted molecular mass of SPICK2 (93 kDa; Fig. 5A), 10 kDa higher than in immunoblots blots performed without prior immunoprecipitation (85 kDa; e.g. Supplementary Fig. S-3 available at JXB online). Very probably, this difference reflected the different conformations of the SPICK2 protein resulting from the different procedures preceding the electrophoresis on the gel. The very faint bands seen occasionally at the same (93 kDa) position, where preimmune serum replaced the immune serum (lanes 4–5), or where KAT1 protein replaced the SPICK2 protein in the incubation with the sera (lanes 6–7), can be attributed to a pre-existing (pre-immune) IgG with an affinity for a protein of similar size among the Sf9 proteins (see also Supplementary Fig. S-5 available at JXB online). Notwithstanding this slightly imperfect differentiation between the immune and preimmune sera, SPICK2 phosphorylation by PKA is manifested, as expected, in the autoradiograph of the immunoblot (Fig. 5B), only in lane 2 (arrow), but not in the presence of the PKA-specific inhibitory peptide, PKI (Fig. 5B, lane 3), in the absence of PKA (lane 1), or in the absence of SPICK2 (lanes 6 and 7). Thus, SPICK2 is the first protein of the AKT2 subfamily demonstrated to be capable of undergoing direct phosphorylation (Li et al., 1998).

View larger version (48K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 5 In vitro phosphorylation of SPICK2 by PKA. (A) An immunoblot of SPICK2 and KAT1 proteins heterologously expressed in Sf9 and immunoprecipitated. PKA, the catalytic subunit of PKA; PKI, the PKA-inhibitory peptide; HC, heavy chains of the IgGs. Arrows indicate the location of the SPICK2 protein band. The molecular masses of protein standards are given in kDa. (B) A phosphorimager autoradiogram of the same membrane blot shown in (A) (see Supplementary Fig. S-5 available at JXB online, and Materials and methods for details). A representative of four independent experiments is shown.

 
cAMP action on K_H channels differs from the effects of OA-promoted phosphorylation
Lack of cAMP effect on I_K currents:
If the in situ K_H channel undergoes a direct phosphorylation by a plant kinase akin to PKA, similarly to the in vitro interaction of SPICK2 with PKA, cAMP (the PKA cofactor) would be expected to enhance the K_H channel phosphorylation, mimicking the effect of OA on the channel activity. The effect of cAMP on the K_H currents of the pulvinus was therefore tested (Fig. 6). cAMP (100 µM final concentration) was included in the ‘internal solution’ within the patch pipette and the currents were recorded between 20 and 30 min after attaining the whole-cell configuration (the time required for the exchange of bath solutions and for the stabilization of the inward currents, which normally increased gradually upon increasing the external [K⁺], Yu et al., 2001). At all membrane potentials tested, the mean steady-state amplitudes of the time-dependent (I_K) currents of cAMP-treated cells were not different from those of the control cells (Fig. 7A, B, open symbols). The mean amplitudes of the I_L currents were not affected either (Fig. 7C, D), even when, in order to decrease cell size-dependent variations in the current, the mean current densities were compared, rather than the means of the whole-cell currents (data not shown, but see Materials and methods). Finally, cAMP did not change the proportion (determined separately for each cell) between the instantaneous (I_L) currents and the time-dependent (I_K) currents (not shown).

View larger version (26K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 6 cAMP and Cs+ effects on inward currents. (A, B) Representative traces of inward currents recorded from two extensor cells, without (A) and with (B) 100 µM cAMP in the pipette (+cAMP). The currents were recorded between 10 and 20 min of bathing in the recording solution (i.e. 20–30 min after attaining the whole-cell configuration), and then a few minutes later, after Cs+ addition to the bath to a final concentration of 10 mM (+Cs+). CNTRL: an untreated cell. All the other details are as in Fig. 1.

 

View larger version (19K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 7 cAMP and Cs+ effects on inward currents. I–EM (current–membrane potential) relationships of whole-cell currents (mean ±SE), elicited during increasingly negative voltage pulses from a holding potential of –30 mV, without (Cntrl) or with cAMP applied intracellularly, before and after bath application of mM Cs+ (as indicated by the symbols). (A, B) I–EM relationships of mean (±SE) time-dependent currents (IK) from extensors [control, n=10 (n=9 at an EM of –170 mV); cAMP-treated cells, n=7] and from flexors [control, n=6 (n=5 at –170 mV); cAMP-treated cells, n=6]. (C, D) I–EM relationships of mean instantaneous current, IL (±SE), of all the cells in (A) and (B).

 
Yet, in spite of the apparent lack of change in the time-dependent currents, cAMP could have altered the proportions of the different channels conducting these inward K⁺ currents. For example, (i) cAMP could have activated cation-non-selective cyclic nucleotide-gated channels (CNGCs, such as described by Arazi et al., 1999; Leng et al., 1999; Sunkar et al., 2000; and as reported by Leng et al., 2002; Balague et al., 2003; Hua et al., 2003), with a concomitant repression of an equivalent fraction of K⁺-selective channels. Conversely, (ii) cAMP could have inhibited cation-non-selective channels (as in Maathuis and Sanders, 2001), while increasing the population of K⁺-selective channels. To distinguish between these possibilities, 10 mM Cs⁺ was added to the bath, as a probe for the channel K⁺ selectivity. Suggestion (i) would be supported by a larger fraction of the current inhibited by Cs⁺ in the absence of cAMP, than in the presence of cAMP, and suggestion (ii) by the inverse relationship.

The inward time-dependent (I_K) currents were blocked by Cs⁺ to a similar extent in both control and cAMP-treated cells, in both extensor and flexor cells (Figs 6, 7A, B, filled symbols). For example, the mean degree of block by Cs⁺ at –170 mV in seven control extensor cells was 81% and in five control flexor cells it was 71% (similar to previous experiments), and with cAMP, in five extensor cells 86% and in five flexor cells (three cells at –170 mV) 79%. Thus, inclusion of cAMP in the pipette did not seem to alter the proportion of the Cs⁺-sensitive (K⁺-selective) I_K currents relative to other time- and voltage-dependent currents.

cAMP and Cs do not affect I_L currents
I_L currents were not affected by cAMP (Fig. 7C, D, open symbols), and, in contrast to the time-dependent currents, they were not affected by Cs⁺ (Fig. 7C, D, closed symbols). In previous experiments (Yu et al., 2001), 10 mM Cs⁺ in the presence of 200 mM external K⁺ blocked roughly two-thirds of I_L. Different conditions between the present and the previous experiments may explain this difference in the observed I_L inhibition: (i) the choice, in the present experiments, to reject cells which did not form tight (5 G) seals with the patch pipette, may have favoured cells with Cs⁺-insensitive leak conduits; or (ii) the lower bath pH in the present experiments (7.2 versus pH 7.8 previously) could have diminished the interaction of Cs⁺ with the outer pore sites of the channel.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Phosphorylation and K_H channels
The novel contributions of this report are the demonstration that the SPICK2 channel protein can be phosphorylated, and also that phosphorylation affects the activity of the in situ S. saman K_H channels. Among plant K channels, only the inward-rectifying, Shaker-like, KAT1 has been shown so far to be amenable to direct phosphorylation. In one case, short peptides of KAT1 sequences were phosphorylated by an ABA-inducible Ca²⁺-insensitive kinase, ABRK (Mori et al., 2000). In another case, a soybean CDPK phosphorylated KAT1 was obtained in an in vitro translation system (Li et al., 1998). Thus, this is the first report, so far, on the phosphorylation of an AKT2 homologue protein.

This is also the first report on effects of phosphorylation on the in situ (in planta) AKT2 or an AKT2-type channel (such as PTK2 or ZMK2). Cherel et al. (2002) demonstrated a physical interaction between AKT2 and a phosphatase AtPP2CA. Upon the co-expression of AKT2 and the phosphatase in frog oocytes and in a mammalian cell line (COS), the instantaneous component, I_L, was more depressed by dephosphorylation than the time- and voltage-dependent component, I_K, thus, ‘increasing rectification’. Conversely, inhibition of the phosphatase by vanadate—hence, enhancement of phosphorylation—increased I_L more than it increased I_K (Cherel et al., 2002).

Recently, Michard et al. (2005a) pinpointed two PKA phosphorylation sites (S210 and S329)—presumably in the region of the AKT2 pore inner mouth (Michard et al., 2005a, b, and references therein) and demonstrated their importance in the switch between the two gating modes (voltage dependent and voltage independent; Dreyer et al., 2001) by electrophysiological assays of mutant channels in the same heterologous systems. The sensitivity of this switch to phosphorylation was intensified by the presence of a ‘voltage sensor’ lysine (K197) in the S4 region of AKT2 (Michard et al., 2005b).

Will phosphorylation affect the AKT2 channel similarly in situ? The present results on the in situ K⁺-influx channels in the S. saman motor cells are contrary to the above findings in the heterologous systems: (i) enhancing phosphorylation at a low OA concentration (5 nM) in flexors, increased the I_K currents without affecting the I_L currents, and (ii) enhancing phosphorylation at a high concentration of OA (300 nM), diminished I_K in both extensors and flexors but, again, without affecting I_L. Interestingly, while possessing the equivalent of a ‘voltage sensor’ lysine (here: K199), SPICK2 was the only one of the seven AKT2 homologues compared by Michard et al. (2005a) in which the site equivalent to S329 of AKT2 was permanently ‘mutated’ to N (asparagine), i.e. in a ‘permanently phosphorylated’-like state.

cAMP did not have any noticeable effect on the I_K or the I_L currents. This was contrary to expectations, based on (i) the predicted amino acid sequence of SPICK2 containing PKA phosphorylation sites; (ii) the reports about putative catalytic PKA subunits detected in planta (see, for example Ward, 2005, At2g20040 entry, #21517 in http://plantsp.genomics.purdue.edu/plantsp/family/class.html); and (iii) hints from earlier studies that a PKA-like activity might exist in plants (Assmann, 1995; Newton et al., 1999; Newton and Smith, 2004; for example, in vitro phosphorylation of guard cell proteins in Vicia faba by a PKA-like endogenous kinase, Friedrich et al., 1999). Indeed, the failure of cAMP to mimic the effect(s) of OA may be due to the lack of the ‘classical’ PKA regulatory subunit in plants (S Podel and M Gribskov, personal communication, 2004).

The lack of an effect of cAMP on the K_H channel activity is also somewhat surprising in view of the cyclic nucleotide-binding domain (CNB) between amino acids 413 and 504 of SPICK2 (Supplementary, Fig. S-2 available at JXB online), a domain common to all known plant K channels of the Shaker subfamily. Thus cAMP may have failed to bind to the CNB region, or its binding may have been ineffective.

Is the in vitro phosphorylation of SPICK2 relevant to the in situ regulation of K_H channels in Samanea?
The in vitro phosphorylation of SPICK2 is relevant—if SPICK2 is indeed the molecular entity underlying the K_H channel. However, a direct comparison between K_H channels and SPICK2 channels is yet to be achieved, since the functional expression of SPICK2 has not been successful so far. No SPICK2-specific inward current was detected in Xenopus oocytes or in CHO cells, or even in Sf9 cells, despite the large yields of SPICK2 protein in Sf9 cells (Supplementary Fig. S-1 available at JXB online) and despite its apparent localization to the plasma membrane in the latter (Supplementary Fig. S-4 available at JXB online). In all of the above systems, KAT1 constructs, within the same plasmids and at identical testing conditions, yielded prominent ‘classical’ inward currents (not shown). It remains to be seen whether this ‘silent’ behaviour in heterologous (Xenopus and mammalian) systems signifies failure of expression, or a lack of a partner subunit in the heterologous system [another , or ß, or another interacting protein, as has been already demonstrated in several other cases (Reintanz, 2002; Picco et al., 2004, and references therein)]. Baizabal-Aguirre et al. (1999) demonstrated an interaction of AKT2 with KAT1 by suppressing AKT2-mediated current in frog oocytes by a dominant negative point mutant of KAT1; AKT2 interactions with AKT1 and AtKC1 have been indicated in two-hybrid assays of their C-terminal polypeptides (Pilot et al., 2003b). Finally, by knocking out AKT2/3 in Arabidopsis guard cells and rendering the remaining inward K⁺ currents insensitive to Ca²⁺ block, Ivashikina et al. (2005) demonstrated AKT2 interactions with all or some of the channels KAT1, KAT2, AKT1, and AtKC1).

	Conclusions

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
The two objectives [(ii) and (iii)] set forth in the Introduction have been attained: K_H (I_K) channel activity was altered differentially by the phosphorylation-promoting OA, indicating the regulation of the I_K channel by phosphorylation. SPICK2 underwent in vitro phosphorylation by a broad range, not necessarily physiological, kinase. These phosphorylation events are consistent with a linkage between the in situ K_H channel and SPICK2. To establish this linkage firmly and to complete the physiological characterization of SPICK2, future work will focus on its homologous, in planta, overexpression, possibly co-expression with another potentially interacting channel, as well as silencing.

	Supplementary data

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Supplementary data can be found at JXB online.

	Note added in proof

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Two publications reported recently that a related Shaker-like K⁺ influx channel, the Arabidopsis AKT1 is subject to phosphorylation by CBL-interacting kinases is response to K starvation.

Li L, Kim B-G, Cheong YH, Pandey GK, Luah S. 2006. A Ca²⁺ signaling pathway regulates a K⁺ channel for low-K response in Arabidopsis. Proceedings of the National Academy of Sciences, USA 0605129103 [Epub ahead of print]

Xu J, Li HD, Chen LQ, Wang Y, Liu LL, He L, Wiu WH. 2006. A protein kinase, interacting with two calcineutin B-like proteins, regulates K⁺ transporter AKT1 in Arabidopsis. Cell 125, 1347–1360.

	Acknowledgements

 
We are grateful to Professor A Harmon for comments related to CDPKs, to Drs S Podell and M Gribskov for comments on the plant PKA, to Mr Juraj Sklenar for instruction on the preparation of membrane vesicles, to Professors Z Adam, M Cohen-Armon, H Fromm, D Hananshvili, Y Zik, and Y Lee, Drs Z Arazi, R Gurevitz, O Oesterzaetzer, and I. Sakler, and to Ms B Otto for friendly advice and helpful suggestions. We thank the Alomone Labs, Jerusalem, Israel for their professional advice on protein chemistry and a gift of okadaic acid. This research was supported by The Israel Science Foundation (Grant No. 550/01) to NM and, in part, by the Dead-Sea Works, Israel to NM, and an EMBO short-term fellowship to LY.

	Footnotes

 
^* These authors contributed equally to this work.

	Abbreviations

 
CDPK, calcium-dependent protein kinase; CNB, cyclic nucleotide-binding domain; DMSO, dimethylsulphoxide; DTT, dithiothreitol; GFP, green fluorescent protein; NMG, N-methylglucamine; OA, okadaic acid; PKA, cAMP-dependent protein kinase; PKI, PKA inhibitor; PMSF, phenylmethylsulphonyl fluoride.

	References

Top Abstract Introduction Materials and methods Results Discussion Conclusions Supplementary data Note added in proof References

 
Arazi T, Sunkar R, Kaplan B, Fromm H. (1999) A tobacco plasma membrane calmodulin-binding transporter confers Ni²⁺ tolerance and Pb²⁺ hypersensitivity in transgenic plants. The Plant Journal 20:171–182.[CrossRef][ISI][Medline]

Assmann SM. (1995) Cyclic AMP as a second messenger in higher plants. Plant Physiology 108:885–889.[ISI][Medline]

Baizabal-Aguirre VM, Clemens S, Uozumi N, Schroeder JI. (1999) Suppression of inward-rectifying K⁺ channels KAT1 and AKT2 by dominant negative point mutations in the KAT1 alpha-subunit. Journal of Membrane Biology 167:119–125.[CrossRef][ISI][Medline]

Balague C, Lin BQ, Alcon C, Flottes G, Malmstrom S, Kohler C, Neuhaus G, Pelletier G, Gaymard F, Roby D. (2003) HLM1, an essential signaling component in the hypersensitive response, is a member of the cyclic nucleotide-gated channel ion channel family. The Plant Cell 15:365–379.[Abstract/Free Full Text]

Bauer CS, Hoth S, Haga K, Philippar K, Aoki N, Hedrich R. (2000) Differential expression and regulation of K⁺ channels in the maize coleoptile: molecular and biophysical analysis of cells isolated from cortex and vasculature. The Plant Journal 24:139–145.[CrossRef][ISI][Medline]

Benga G. (2003) Birth of water channel proteins—the aquaporins. Cell Biology International 27:701–709.[CrossRef][ISI][Medline]

Blom N, Gammeltoft S, Brunak S. (1999) Sequence- and structure-based prediction of eukaryotic protein phosphorylation sites. Journal of Molecular Biology 294:1351–1362.[CrossRef][ISI][Medline]

Bockenhauer D, Zilberberg N, Goldstein SAN. (2001) KCNK2: reversible conversion of a hippocampal potassium leak into a voltage-dependent channel. Nature Neuroscience 4:486–491.[ISI][Medline]

Cherel I, Michard E, Platet N, Mouline K, Alcon C, Sentenac H, Thibaud JB. (2002) Physical and functional interaction of the Arabidopsis K⁺ channel AKT2 and phosphatase AtPP2CA. The Plant Cell 14:1133–1146.[Abstract/Free Full Text]

Dreyer I, Horeau C, Lemaillet G, Zimmermann S, Bush DR, Rodriguez-Navarro A, Schachtman DP, Spalding EP, Sentenac H, Gaber RF. (1999) Identification and characterization of plant transporters using heterologous expression systems. Journal of Experimental Botany 50:1073–1087.[Abstract]

Dreyer I, Michard E, Lacombe B, Thibaud JB. (2001) A plant Shaker-like K⁺ channel switches between two distinct gating modes resulting in either inward-rectifying or ‘leak’ current. FEBS Letters 505:233–239.[CrossRef][ISI][Medline]

Folco EJ, Liu GX, Koren G. (2004) Caveolin-3 and SAP97 form a scaffolding protein complex that regulates the voltage-gated potassium channel Kv1.5. American Journal of Physiology 287:H681–H690.

Friedrich P, Curvetto N, Giusto N. (1999) Cyclic AMP-dependent protein phosphorylation in guard cell protoplasts of Vicia faba L. Biocell 23:203–210.

Gerhardstein BL, Puri TS, Chien AJ, Hosey MM. (1999) Identification of the sites phosphorylated by cyclic AMP-dependent protein kinase on the beta(2) subunit of L-typ