Direct Phosphorylation and Activation of a Mitogen-Activated Protein Kinase by a Calcium-Dependent Protein Kinase in Rice

The mitogen-activated protein kinase (MAPK) is a pivotal point of convergence for many signaling pathways in eukaryotes. In the classical MAPK cascade, a signal is transmitted via sequential phosphorylation and activation of MAPK kinase kinase, MAPK kinase (MKK), and MAPK. The activation of MAPK is dependent on dual phosphorylation of a TXY motif by an MKK, which is considered the sole kinase to phosphorylate and activate MAPK. Here, we report a novel regulatory mechanism of MAPK phosphorylation and activation besides the canonical MAPK cascade. A rice (Oryza sativa) calcium-dependent protein kinase (CDPK), CPK18, was identified as an upstream kinase of MAPK (MPK5) in vitro and in vivo. Curiously, CPK18 was shown to phosphorylate and activate MPK5 without affecting the phosphorylation of its TXY motif. Instead, CPK18 was found to predominantly phosphorylate two Thr residues (Thr-14 and Thr-32) that are widely conserved in MAPKs from land plants. Further analyses reveal that the newly identified CPK18-MPK5 pathway represses defense gene expression and negatively regulates rice blast resistance. Our results suggest that land plants have evolved an MKK-independent phosphorylation pathway that directly connects calcium signaling to the MAPK machinery.     

Stress-Activated Protein Kinase/c-Jun N-Terminal Kinase Phosphorylates τ Protein

Abstract: A proportion of the neuronal microtubule-associated protein (MAP) τ is highly phosphorylated in foetal and adult brain, whereas the majority of τ in the neurofibrillary tangles of Alzheimer's patients is hyperphosphorylated; many of the phosphorylation sites are serines or threonines followed by prolines. Several kinases phosphorylate τ at such sites in vitro. We have now shown that purified recombinant stress-activated protein kinase/c-Jun N-terminal kinase, a proline-directed kinase of the MAP kinase extended family, phosphorylates recombinant τ in vitro on threonine and serine residues. Western blots using antibodies to phosphorylation-dependent τ epitopes demonstrated that phosphorylation occurs in both of the main phosphorylated regions of τ protein. Unlike glycogen synthase kinase-3, the c-Jun N-terminal kinase readily phosphorylates Thr²⁰⁵ and Ser⁴²², which are more highly phosphorylated in Alzheimer τ than in foetal or adult τ. Glycogen synthase kinase-3 may preferentially phosphorylate the sites found physiologically, in foetal and to a smaller extent in adult τ, whereas stress-activated/c-Jun N-terminal kinase and/or other members of the extended MAP kinase family may be responsible for pathological proline-directed phosphorylations. Inflammatory processes in Alzheimer brain might therefore contribute directly to the pathological formation of the hyperphosphorylated τ found in neurofibrillary tangles.     

Evidence for a Multi-ion Pore Behavior in the Plant Potassium Channel KAT1

KAT1 is a cloned voltage-gated K⁺ channel from the plant Arabidopsis thaliana L., which displays an inward rectification reminiscent of `anomalous' rectification of the i <sub>f</sub> pacemaker current recorded in animal cells. Macroscopic conductance of KAT1 expressed in Xenopus oocytes was 5-fold less in pure Rb<sup>+</sup> solution than in pure K<sup>+</sup> solution, and negligible in pure Na<sup>+</sup> solution. Experiments in different K<sup>+</sup>/Na<sup>+</sup> or K<sup>+</sup>/Rb<sup>+</sup> mixtures revealed deviations from the principle of independence and notably two anomalous effects of the K<sup>+</sup>/Rb<sup>+</sup> mole fraction (i.e., the ratio [K<sup>+</sup>]/([K<sup>+</sup>]+[Rb<sup>+</sup>])). First, the KAT1 deactivation time constant was both voltage- and mole fraction-dependent (a so-called `foot in the door' effect was thus observed in KAT1 channel). Second, when plotted against the K⁺/Rb⁺ mole fraction, KAT1 conductance values passed through a minimum. This minimum is more important for two pore mutants of KAT1 (T259S and T260S) that displayed an increase in P_Rb/P_K. These results are consistent with the idea that KAT1 conduction requires several ions to be present simultaneously within the pore. Therefore, this atypical `green' member of the Shaker superfamily of K⁺ channels further shows itself to be an interesting model as well for permeation as for gating mechanism studies. Received: 9 February 1998/Revised: 28 July 1998     

A Calcium-Dependent Protein Kinase Interactswith and Activates A Calcium Channel toRequlate Pollen Tube Growth

ABSTRACT Calcium, as a ubiquitous second messenger, plays essential roles in tip-growing cells, such as animal neu-rons, plant pollen tubes, and root hairs. However, little is known concerning the regulatory mechanisms that code anddecode Ca2＋ signals in plants. The evidence presented here indicates that a calcium-dependent protein kinase, CPK32,controls polar growth of pollen tubes. Overexpression of CPK32 disrupted the polar growth along with excessive Ca2＋accumulation in the tip. A search of downstream effector molecules for CPK32 led to identification of a cyclic nucleotide-gated channel, CNGC18, as an interacting partner for CPK32. Co-expression of CPK32 and CNGC18 resulted in activationof CNGC18 in Xenopus oocytes where expression of CNGC18 alone did not exhibit significant calcium channel activity.Overexpression of CNGC18 produced a growth arrest phenotype coupled with accumulation of calcium in the tip, simi-lar to that induced by CPK32 overexpression. Co-expression of CPK32 and CNGC18 had a synergistic effect leading tomore severe depolarization of pollen tube growth. These results provide a potential feed-forward mechanism in whichcalcium-activated CPK32 activates CNGC18, further promoting calcium entry during the elevation phase of Ca2＋ oscilla-tions in the polar growth of pollen tubes.     

Protein Kinase A Activation Phosphorylates the Rat ClC-2 Cl− Channel but Does Not Change Activity

Phosphorylation-dependent events have been shown to modulate the activity of several members of the mammalian CLC Cl⁻ channel gene family, including the inward rectifier ClC-2. In the present study we investigated the regulation of rat ClC-2 expressed in the TSA-201 cell line (a transformed HEK293 cell line that stably expresses the SV40 T-antigen) by protein kinases. Protein kinase A activation phosphorylated ClC-2 in vivo, whereas stimulation of protein kinase C with phorbol 12-myristate 13-acetate did not. In vitro labeling studies confirmed that protein kinase A could directly phosphorylate ClC-2, and that protein kinase C and Ca²⁺/calmodulin-dependent protein kinase II did not. Nevertheless, protein kinase A-dependent phosphorylation of CLC-2 failed to regulate either the magnitude or the kinetics of the hyperpolarization-activated Cl⁻ currents. Considered together, we demonstrate that protein kinase A activation results in the phosphorylation of rat ClC-2 in vivo, but this event is independent of Cl⁻ channel activity. Received: 20 November 2000/Revised: 28 March 2001     

Protein Kinase D Interacts with Neuronal Nitric Oxide Synthase and Phosphorylates the Activatory Residue Serine1412

Neuronal Nitric Oxide Synthase (nNOS) is the biosynthetic enzyme responsible for nitric oxide (·NO) production in muscles and in the nervous system. This constitutive enzyme, unlike its endothelial and inducible counterparts, presents an N-terminal PDZ domain known to display a preference for PDZ-binding motifs bearing acidic residues at -2 position. In a previous work, we discovered that the C-terminal end of two members of protein kinase D family (PKD1 and PKD2) constitutes a PDZ-ligand. PKD1 has been shown to regulate multiple cellular processes and, when activated, becomes autophosphorylated at Ser⁹¹⁶, a residue located at -2 position of its PDZ-binding motif. Since nNOS and PKD are spatially enriched in postsynaptic densities and dendrites, the main objective of our study was to determine whether PKD1 activation could result in a direct interaction with nNOS through their respective PDZ-ligand and PDZ domain, and to analyze the functional consequences of this interaction. Herein we demonstrate that PKD1 associates with nNOS in neurons and in transfected cells, and that kinase activation enhances PKD1-nNOS co-immunoprecipitation and subcellular colocalization. However, transfection of mammalian cells with PKD1 mutants and yeast two hybrid assays showed that the association of these two enzymes does not depend on PKD1 PDZ-ligand but its pleckstrin homology domain. Furthermore, this domain was able to pull-down nNOS from brain extracts and bind to purified nNOS, indicating that it mediates a direct PKD1-nNOS interaction. In addition, using mass spectrometry we demonstrate that PKD1 specifically phosphorylates nNOS in the activatory residue Ser¹⁴¹², and that this phosphorylation increases nNOS activity and ·NO production in living cells. In conclusion, these novel findings reveal a crucial role of PKD1 in the regulation of nNOS activation and synthesis of ·NO, a mediator involved in physiological neuronal signaling or neurotoxicity under pathological conditions such as ischemic stroke or neurodegeneration.     

Reactivating Kinase/p38 Phosphorylates τ Protein In Vitro

Abstract: Neurofibrillary tangles, one of the major pathological hallmarks of Alzheimer-diseased brains, consist primarily of aggregated paired helical filaments (PHFs) of hyperphosphorylated τ protein. τ from normal brain and especially from foetal brain is also phosphorylated on some of the sites phosphorylated in PHFs, mainly at serines or threonines followed by prolines. A number of protein kinases can phosphorylate τ in vitro; those that require or accept prolines include GSK3 and members of the mitogen-activated protein (MAP) kinase family, ERK1, ERK2, and SAP kinase-β/JNK. In this report, we show that another member of the MAP kinase family, the stress-activated kinase p38/RK, can phosphorylate τ in vitro. Western blots with phosphorylation-sensitive antibodies showed that p38, like ERK2 and SAP kinase-β/JNK, phosphorylated τ at sites found phosphorylated physiologically (Thr¹⁸¹, Ser²⁰², Thr²⁰⁵, and Ser³⁹⁶) and also at Ser⁴²², which is phosphorylated in neurofibrillary tangles but not in normal adult or foetal brain. These findings support the possibility that cellular stress might contribute to τ hyperphosphorylation during the formation of PHFs, and hence, to the development of τ pathology.     

A Casein Kinase-Like Kinase Phosphorylates β-Tubulin and May Be a Microtubule-Associated Protein

The hypothesis that casein kinase II (CKII) is a microtubule-associated protein kinase was investigated using a neuronal cell line and bovine brain. Heparin, an inhibitor of CKII, inhibited the phosphorylation of a PC12 cytosolic protein whose molecular weight was similar to that of beta-tubulin. Partially purified PC12 CKII was immunoreactive to an antibody directed against bovine CKII and was able to phosphorylate purified beta-tubulin in a heparin-inhibitable manner when the concentration of tubulin was less than 50 micrograms/ml. To better determine if CKII is a microtubule-associated protein kinase, bovine brain tubulin was chromatographed on FPLC Mono Q and phosphocellulose columns. Several tubulin casein kinase (TCK) activities were apparent. All TCK activities phosphorylated tubulin and casein, but none was able to phosphorylate the CKII-specific synthetic peptide RRREEETEEE. One of these TCK fractions was immunoreactive to the antibody directed against CKII, and this antibody labeled a 50-kDa molecular mass band that had a molecular mass distinctly different from those of the subunits of CKII. Thus, we suggest that a CKII-like protein, but not CKII, might be a microtubule-associated protein.     

Herpes Simplex Virus 1 Protein Kinase Us3 Phosphorylates Viral dUTPase and Regulates Its Catalytic Activity in Infected Cells

Us3 is a serine-threonine protein kinase encoded by herpes simplex virus 1 (HSV-1). In this study, a large-scale phosphoproteomic analysis of titanium dioxide affinity chromatography-enriched phosphopeptides from HSV-1-infected cells using high-accuracy mass spectrometry (MS) and subsequent analyses showed that Us3 phosphorylated HSV-1-encoded dUTPase (vdUTPase) at serine 187 (Ser-187) in HSV-1-infected cells. Thus, the following observations were made. (i) In in vitro kinase assays, Ser-187 in the vdUTPase domain was specifically phosphorylated by Us3. (ii) Phosphorylation of vdUTPase Ser-187 in HSV-1-infected cells was detected by phosphate-affinity polyacrylamide gel electrophoresis analyses and was dependent on the kinase activity of Us3. (iii) Replacement of Ser-187 with alanine (S187A) in vdUTPase and an amino acid substitution in Us3 that inactivated its kinase activity significantly downregulated the enzymatic activity of vdUTPase in HSV-1-infected cells, whereas a phosphomimetic substitution at vdUTPase Ser-187 restored the wild-type enzymatic activity of vdUTPase. (iv) The vdUTPase S187A mutation as well as the kinase-dead mutation in Us3 significantly reduced HSV-1 replication in human neuroblastoma SK-N-SH cells at a multiplicity of infection (MOI) of 5 but not at an MOI of 0.01, whereas the phosphomimetic substitution at vdUTPase Ser-187 restored the wild-type viral replication at an MOI of 5. In contrast, these mutations had no effect on HSV-1 replication in Vero and HEp-2 cells. Collectively, our results suggested that Us3 phosphorylation of vdUTPase Ser-187 promoted HSV-1 replication in a manner dependent on cell types and MOIs by regulating optimal enzymatic activity of vdUTPase.     

Characterization of a Nerve Growth Factor-Stimulated Protein Kinase in PC12 Cells Which Phosphorylates Microtubule-Associated Protein 2 and pp250

Treatment of PC12 cells with nerve growth factor (NGF) resulted in the rapid, but transient, activation of a protein kinase which specifically phosphorylated an endogenous 250-kDa cytoskeletal protein (pp250). We report that the microtubule-associated protein, MAP2, is an alternative substrate for the NGF-activated kinase. NGF treatment maximally activated the kinase within 5 min; however, the activity declined with longer exposure to NGF. The enzyme was localized predominantly in microsomal and soluble fractions and phosphorylated MAP2 on serine and threonine residues. The soluble enzyme was fractionated by DEAE chromatography and gel filtration and had an apparent Mr of 45,000. The enzyme was purified to near homogeneity by chromatofocussing and had a pI of 4.9. Kinetic analysis revealed that NGF treatment caused a sevenfold increase in Vmax for MAP2. The Km with respect to the MAP2 substrate was approximately 50 nM and was not altered by NGF treatment. A novel feature of the NGF-stimulated enzyme was its sharp dependence on Mn2+ concentration. The active enzyme is likely to be phosphorylated, because inclusion of phosphatase inhibitors was required for recovery of optimal activity and the activity was lost on treatment of the enzyme with alkaline phosphatase. Histones, tubulin, casein, bovine serum albumin, and the ribosomal subunit protein S-6 were not phosphorylated by this enzyme. The NGF-stimulated kinase was distinct from A kinase, C kinase, or other NGF-stimulated kinases. The rapid and transient activation of the protein kinase upon NGF treatment suggests that the enzyme may play a role in signal transduction in PC12 cells.     

A Novel Alpha Kinase EhAK1 Phosphorylates Actin and Regulates Phagocytosis in Entamoeba histolytica

Phagocytosis plays a key role in nutrient uptake and virulence of the protist parasite Entamoeba histolytica. Phagosomes have been characterized by proteomics, and their maturation in the cells has been studied. However, there is so far not much understanding about initiation of phagocytosis and formation of phagosomes at the molecular level. Our group has been studying initiation of phagocytosis and formation of phagosomes in E. histolytica, and have described some of the molecules that play key roles in the process. Here we show the involvement of EhAK1, an alpha kinase and a SH3 domain containing protein in the pathway that leads to formation of phagosomes using red blood cell as ligand particle. A number of approaches, such as proteomics, biochemical, confocal imaging using specific antibodies or GFP tagged molecules, expression down regulation by antisense RNA, over expression of wild type and mutant proteins, were used to understand the role of EhAK1 in phagocytosis. EhAK1 was found in the phagocytic cups during the progression of cups, until closure of phagosomes, but not in the phagosomes themselves. It is recruited to the phagosomes through interaction with the calcium binding protein EhCaBP1. A reduction in phagocytosis was observed when EhAK1 was down regulated by antisense RNA, or by over expression of the kinase dead mutant. G-actin was identified as one of the major substrates of EhAK1. Phosphorylated actin preferentially accumulated at the phagocytic cups and over expression of a phosphorylation defective actin led to defects in phagocytosis. In conclusion, we describe an important component of the pathway that is initiated on attachment of red blood cells to E. histolytica cells. The main function of EhAK1 is to couple signalling events initiated after accumulation of EhC2PK to actin dynamics.     

Staurosporine Activates a 60,000 Mr Protein Kinase in Bovine Chromaffin Cells That Phosphorylates Myelin Basic Protein In Vitro

Abstract: Bovine chromaffin cells contain a family of renaturable protein kinases. One of these, a 60,000 M_r kinase (PK60) that phosphorylated myelin basic protein in vitro, was activated fourfold when cells were treated with the protein kinase inhibitor Staurosporine. Because staurosporine inhibits protein kinase C, the role of this kinase in the regulation of PK60 activity was investigated. Fifty nanomolar Staurosporine produced half-maximal inhibition of protein kinase C activity in chromaffin cells, whereas ∼225 n M Staurosporine was required to induce half-maximal activation of PK60. Other protein kinase C inhibitors, H-7 and K-252a, did not mimic the effect of Staurosporine on PK60 activity. Chromaffin cells have three protein kinase C isoforms: α, ε, and ζ. Prolonged treatment with phorbol esters depleted the cells of protein kinase C α and ε, but not ζ. Neither activation nor depletion of protein kinase C affected the basal activity of PK60. Moreover, Staurosporine activated PK60 in cells depleted of protein kinase C α and e; thus, Staurosporine appeared to activate PK60 by a mechanism that does not require these protein kinase C isoforms. Incubation of cell extracts with Staurosporine in vitro did not activate PK60. Incubation of these extracts with adenosine 5'-O-(3-thiotriphosphate), however, caused a twofold activation of PK60. Although this suggests that PK60 activity is regulated by phosphorylation, the mechanism by which Staurosporine activates PK60 is not known. Staurosporine has been reported to promote neurite outgrowth from chromaffin cells. The role of PK60 in mediating the effects of Staurosporine on chromaffin cell function remains to be determined.     

An Abscisic Acid-Activated and Calcium-Independent Protein Kinase from Guard Cells of Fava Bean

Abscisic acid (ABA) regulation of stomatal aperture is known to involve both Ca2+-dependent and Ca2+-independent signal transduction pathways. Electrophysiological studies suggest that protein phosphorylation is involved in ABA action in guard cells. Using biochemical approaches, we identified an ABA-activated and Ca2+- independent protein kinase (AAPK) from guard cell protoplasts of fava bean. Autophosphorylation of AAPK was rapidly (~1 min) activated by ABA in a Ca2+- independent manner. ABA-activated autophosphorylation of AAPK occurred on serine but not on tyrosine residues and appeared to be guard cell specific. AAPK phosphorylated histone type III-S on serine and threonine residues, and its activity toward histone type III-S was markedly stimulated in ABA-treated guard cell protoplasts. Our results suggest that AAPK may play an important role in the Ca2+-independent ABA signaling pathways of guard cells.     

Functional Characterization and Determination of the Physiological Role of a Calcium-Dependent Potassium Channel from Cyanobacteria

Despite the important achievement of the high-resolution structures of several prokaryotic channels, current understanding of their physiological roles in bacteria themselves is still far from complete. We have identified a putative two transmembrane domain-containing channel, SynCaK, in the genome of the freshwater cyanobacterium Synechocystis sp. PCC 6803, a model photosynthetic organism. SynCaK displays significant sequence homology to MthK, a calcium-dependent potassium channel isolated from Methanobacterium thermoautotrophicum. Expression of SynCaK in fusion with enhanced GFP in mammalian Chinese hamster ovary cells’ plasma membrane gave rise to a calcium-activated, potassium-selective activity in patch clamp experiments. In cyanobacteria, Western blotting of isolated membrane fractions located SynCaK mainly to the plasma membrane. To understand its physiological function, a SynCaK-deficient mutant of Synechocystis sp. PCC 6803, ΔSynCaK, has been obtained. Although the potassium content in the mutant organisms was comparable to that observed in the wild type, ΔSynCaK was characterized by a depolarized resting membrane potential, as determined by a potential-sensitive fluorescent probe. Growth of the mutant under various conditions revealed that lack of SynCaK does not impair growth under osmotic or salt stress and that SynCaK is not involved in the regulation of photosynthesis. Instead, its lack conferred an increased resistance to the heavy metal zinc, an environmental pollutant. A similar result was obtained using barium, a general potassium channel inhibitor that also caused depolarization. Our findings thus indicate that SynCaK is a functional channel and identify the physiological consequences of its deletion in cyanobacteria.Detailed structural and mechanistic data now exist for many prokaryotic channels, but their physiological roles remain largely unclear (Martinac et al., 2008). This is especially true for potassium channels. K⁺ is the most abundant cation in organisms, and in general, it plays a crucial role in the survival and development of cells by regulating enzyme activity and tuning electrochemical membrane potential. Potassium channels in prokaryotes have been hypothesized to contribute to the setting of membrane potential rather than to high-affinity potassium uptake normally achieved thanks to specific ATP-dependent potassium transporters (Kuo et al., 2005). K⁺ channel genes are found in almost every prokaryotic genome that has been sequenced, but in most of the few studies where their deletion was obtained, no specific phenotype has been observed, suggesting either functional redundancy or that these channels are only required in case of rather specific environmental stresses. Gain-of-function potassium channel (Kch) mutants of Escherichia coli failed to grow in millimolar-added K⁺ but not Na⁺ (Kuo et al., 2003), and external H⁺ suppressed the gain-of-function phenotype, supporting the hypothesis that KCh might function to regulate membrane potential. However, a clear-cut role of prokaryotic potassium channels by genetic deletion was demonstrated only in a few cases.The model organism Synechocystis sp. PCC 6803 harbors an intracellular membrane system, the thylakoids, where both photosynthesis and respiration take place. In this work, we have identified in Synechocystis sp. PCC 6803 a so-far uncharacterized putative potassium channel, NP_440478, encoded by the open reading frame (ORF) sll0993, with sequence homology to MthK, a Ca²⁺-activated K⁺ channel of the archaeon Methanobacterium thermoautotrophicum. The structure of MthK in an open conformation has been determined (Jiang et al., 2002). The MthK subunit has two transmembrane segments and one pore region, followed by an extension of approximately 200 amino acid residues, which contains a region called the regulator of the conductance of K⁺ (RCK) domain. RCK of MthK binds divalent cations, such as Ca²⁺ or Cd²⁺ (Jiang et al., 2001, 2002). The physiological meaning of the activation of MthK with millimolar Ca²⁺ (Jiang et al., 2002) is, however, unclear, because Ca²⁺ as a second messenger operates at micromolar concentrations in eukaryotes, and the possible signaling roles of Ca²⁺ in prokaryotes are still unclear. Another RCK-containing 160-picoSiemens (pS) K⁺ channel from the archaeon Thermoplasma volcanium, TvoK, was also found to be activated by millimolar Ca²⁺ (Parfenova et al., 2007). Here, we report evidence that, similar to MthK, SynCaK can also be activated by calcium. Furthermore, we show localization of the protein in cyanobacteria and describe a phenotype associated with the lack of the channel in SynCaK-less mutant Synechocystis sp. PCC 6803 cells.     

Glycogen Synthase Kinase - 3 Phosphorylates and Regulates the Stability of p27kip1 Protein

p27 Kip1 is a critical regulator of the eukaryotic cell cycle. It acts as a check point proteinand regulates cell cycle progression at the G1 and G1/S phase as well as predominantlyblocks cell cycle progression in the absence of growth factors. Intracellular turnover of p27is tightly regulated at the level of translation as well as by posttranslational modification.The mechanism by which p27 protein is rapidly degraded during the G1 and G1/S phasetransition is well characterized. However, the process by which p27 remains extremelystable in the absence of growth factors remains unknown. Here, we report that GSK-3dependent phosphorylation of p27 protein is essential for its enhanced stability. p27 proteinharbours 2 functional GSK-3 phosphorylation sites at the C- terminus, which was found tobe effectively phosphorylated by the cognate enzyme both in vitro and in vivo. Combinedwith earlier observation which shows that it phosphorylates and triggers cyclin Ddegradation; GSK-3 now appears to be a central mediator of the cell-cycle regulatorynetwork, where it acts as a two-way switch, phosphorylating and targeting pro-proliferativefactors for degradation on one hand and simultaneously phosphorylating and stabilizing ananti-proliferative factor on the other hand. This dual mode of activity may doubly ensurethat cell cycle progression is aptly prohibited under conditions of limited growth factoravailability.     

Preferential KAT1-KAT2 Heteromerization Determines Inward K+ Current Properties in Arabidopsis Guard Cells

Guard cells adjust their volume by changing their ion content due to intense fluxes that, for K⁺, are believed to flow through inward or outward Shaker channels. Because Shaker channels can be homo- or heterotetramers and Arabidopsis guard cells express at least five genes encoding inward Shaker subunits, including the two major ones, KAT1 and KAT2, the molecular identity of inward Shaker channels operating therein is not yet completely elucidated. Here, we first addressed the properties of KAT1-KAT2 heteromers by expressing KAT1-KAT2 tandems in Xenopus oocytes. Then, computer analyses of the data suggested that coexpression of free KAT1 and KAT2 subunits resulted mainly in heteromeric channels made of two subunits of each type due to some preferential association of KAT1-KAT2 heterodimers at the first step of channel assembly. This was further supported by the analysis of KAT2 effect on KAT1 targeting in tobacco cells. Finally, patch-clamp recordings of native inward channels in wild-type and mutant genotypes strongly suggested that this preferential heteromerization occurs in planta and that Arabidopsis guard cell inward Shaker channels are mainly heteromers of KAT1 and KAT2 subunits.