S-acylation-dependent membrane microdomain localization of the regulatory Kvβ2.1 subunit

The voltage-dependent potassium (Kv) channel Kvβ family was the first identified group of modulators of Kv channels. Kvβ regulation of the α-subunits, in addition to their aldoketoreductase activity, has been under extensive study. However, scarce information about their specific α-subunit-independent biology is available. The expression of Kvβs is ubiquitous and, similar to Kv channels, is tightly regulated in leukocytes. Although Kvβ subunits exhibit cytosolic distribution, spatial localization, in close contact with plasma membrane Kv channels, is crucial for a proper immune response. Therefore, Kvβ2.1 is located near cell surface Kv1.3 channels within the immunological synapse during lymphocyte activation. The objective of this study was to analyze the structural elements that participate in the cellular distribution of Kvβs. It was demonstrated that Kvβ peptides, in addition to the cytoplasmic pattern, targeted the cell surface in the absence of Kv channels. Furthermore, Kvβ2.1, but not Kvβ1.1, targeted lipid raft microdomains in an S-acylation-dependent manner, which was concomitant with peptide localization within the immunological synapse. A pair of C-terminal cysteines (C301/C311) was mostly responsible for the specific palmitoylation of Kvβ2.1. Several insults altered Kvβ2.1 membrane localization. Therefore, growth factor-dependent proliferation enhanced surface targeting, whereas PKC activation impaired lipid raft expression. However, PSD95 stabilized Kvβ2.1 in these domains. This data shed light on the molecular mechanism by which Kvβ2.1 clusters into immunological synapses during leukocyte activation. Supplementary Information The online version contains supplementary material available at 10.1007/s00018-022-04269-3.


Introduction
Voltage-gated potassium channels (Kv) control action potentials in muscles and nerves. Furthermore, Kv plays fundamental functions during leukocyte activation, proliferation and apoptosis. In addition, the association with different accessory proteins further increases their functional diversity. These ancillary interactions affect not only the electrophysiological properties of the channels, but also their localization, trafficking and turnover [1,2].
The Kvβ family was the first group of proteins identified as Kv channel regulatory subunits. Three different genes (KCNAB1, KCNAB2 and KCNAB3) encode this family, including different splicing variants [1]. The crystal structure of the Kv1.2 channel in the presence of Kvβ2.1 confirmed the soluble nature of these Kvβ peptides [3]. These cytoplasmic proteins, via their conserved C-terminal domain, associate with α-subunits by interacting with an N-terminus preserved structure in Kv [4]. In addition to regulating the electrophysiological properties, some Kvβ members alter the trafficking of Kv channels. Therefore, an increase in Kv channels at the cell surface claims chaperone-like effects [5,6].
The regulatory voltage-gated Kvβ1.1 and Kvβ2.1 peptides are cytosolic proteins that clearly display intracellular phenotypes ( Fig. 1Aa-Ah). However, when the Kvβ pattern was analyzed further, pixel-by-pixel analysis revealed some partial colocalization with plasma membrane staining ( Fig. 1 Ad and Ah). This Kvβ cell surface expression was further confirmed in enriched plasma membrane preparations (Fig. 1B, C).

S-acetylation drives Kvß2.1 to cell plasma membrane lipid raft microdomains
Soluble proteins may attach to the eukaryotic plasma membrane by lipidic posttranslational modifications such as S-acetylation. Palmitoyl acyltransferases (PATs) add palmitate to cysteine residues, which can be reverted by thiosterases. Palmitoylation helps protein-protein interactions and confers protein stability but also improves trafficking and membrane association [29].

Human T lymphocytes concentrate palmitoylated Kvß2.1 at the immunological synapse
We showed that palmitoylated Kvβ2.1 targeted plasma membrane lipid raft microdomains. In addition, Kvβ2.1 interacts with ZIP1/2 and PKC, providing an effective phosphorylation cluster for signaling in the IS [20,22,25]. During T-cell activation, IS, enriched in lipid rafts, concentrates Kv1.3 channels, which in turn may associate with Kvβ2.1 to fine tune the immune response [32]. However, our data would indicate that Kvβ2.1 is located in IS via palmitoylation, even in the absence of Kv1.3. Therefore, the AKR activity of Kvβ2.1 in addition to further functions would be independent of channel expression. In this scenario, evidence claims Kvβ2.1 localization at IS [20]. Although palmitome studies identify Kvβ2 in brain [33,34], this protein is not found in some leukocyte analysis [35][36][37][38][39]. Human CD4 + lymphocytes and human Jurkat T cells expressed Kvβ2 (Fig. 5A, C). Biotin pulled down extracts (PD), in the presence of hydroxylamine, revealed that CD4 + lymphocytes (Fig. 5B) and Jurkat cells (Fig. 5D) expressed palmitoylated Kvβ2.1, which was located on the plasma membrane surface (Fig. 5E) and targeted lipid rafts (Fig. 5F) in Jurkat lymphocytes. Cell conjugates between SEE-activated human B lymphocytes (Raji) and human T lymphocytes (Jurkat) demonstrated that while Kvβ2.1 evenly stained the surface of T cells, in the absence of IS, Kvβ2.1 concentrated at the IS once the synapse was achieved (Fig. 5Ga-e). Non-SEE activated Raji cells generate no synapse with Jurkat T cells and Kvβ2 did not accumulate in cell-to-cell contact surfaces (Fig. 5Gf-i).

Distal C-terminal cysteines are responsible for the specific palmitoylation of Kvß2.1
We demonstrated that palmitoylated Kvβ2.1 reached plasma membrane lipid raft microdomains and polarized into IS during the immune response. Because 2-BP is promiscuous and targets PAT enzymes, transporters and many palmitoylated proteins [40], 2-BP's impairment of Kvβ2's targeting to lipid rafts (Fig. 4) must be taken with caution. Therefore, the molecular determinants involved were further analyzed. The five S-acylated putative Cys residues in Kvβ2.1 (51,212,248, 301 and 311) were mutated to Ser, maintaining structural similarity (Fig. 6A,  B). Surprisingly, while C(51-212-248-301)S preserved the regular intracellular pattern of Kvβ2.1, with limited cell surface localization, the introduction of the C311S mutation in combination with the rest of cysteine (C less ) triggered anomalous plasma membrane distributions (Fig.  S4). This altered behavior could be the consequence of the introduction of an artificial PKC phosphorylation site (Fig. S4B). Therefore, this substitution was discarded, and new Kvβ2.1 C301A, C311A and C less A mutants were generated, which conserved the WT phenotype (Fig. 6C). The sequential substitution of cysteines steadily triggered the loss of membrane localization, which was pronounced when the distal C-term Cys (301 and 311) were mutated (C51-212-248-301S; C less A) (Fig. 6D). Whether S-palmitoylation was concomitant with the cell surface expression of Kvβ2.1 was further investigated. The ABE data showed that only when Cys 301 and 311 were individually substituted (C301A, C311A) or combined with the rest of Cys (C51-212-248-301S; C less A) the Kvβ2.1 palmitoylation was clearly impaired (Fig. 7A, B; Fig. S5). Therefore, evidence demonstrated that the C301 and C311 C-terminal distal residues were the main targets for PATs in Kvβ2.1. Furthermore, palmitoylation-dependent lipid raft and plasma membrane targeting was further confirmed. Thus, unlike Kvβ2.1 WT, Kvβ2.1 C less A neither targeted to rafts ( Fig. 7C-E) nor to cell surface (Fig. S6).

Proliferation and activation differentially alter Kvβ2.1 localization
Lipid rafts are membrane platforms that initiate cell signaling. These microdomains reduce the spatial distance between signaling proteins and their targets [41]. Considering Kvβ oxidoreductase function and Kv modulation, the positioning of Kvβ2.1 in lipid raft microdomains within the IS was highly relevant. Because Kvβ subunits exhibit differential regulation upon proliferation or activation in macrophages [18], it was wondered whether targeting lipid rafts, which was dependent on S-palmitoylation, underwent regulation. Therefore, HEK cells were cultured in the absence of FBS and analyzed Kvβ2.1 spatial location. In the absence of growth factors (-FBS), the plasma membrane Kvβ2.1 localization slightly decreased ( Evidence demonstrates that PKC initiates signaling events within IS [21]. Previous works from the authors' laboratory report a PMA-dependent exit of Kv1.3 from raft domains prior to endocytosis of the channel [23]. Because Kvβ2.1, which is phosphorylated by PKC [42], was present at lipid rafts, it was analyzed whether PMA would also regulate the location of this auxiliary subunit. Therefore, lipid raft microdomains were isolated in the absence or presence of PMA, and Kvβ2.1 expression was analyzed (  lymphocyte proliferation. hDgl1 (synapse-associated protein 97, SAP97) is located at the IS, and hDgl4 (postsynaptic density protein 95, PSD95) is required for uropod formation in T-cells [43,44]. Furthermore, PSD95, interacting with Kv1.3 at the IS, stabilizes the channel in rafts upon PKC-dependent endocytosis. Palmitoylated PSD95 targets lipid rafts interacting with PKCα [45][46][47]. Taking all this in mind, the effect of PSD95 on Kvβ2.1 were analyzed upon PKC-dependent activation. The data indicated that PSD95 stabilized Kvβ2.1 into floating fractions, preventing the PKC-dependent lipid raft displacement observed with PMA ( Fig. 9C-E). However, unlike Kv1.3 [23], PSD95 did not interact with Kvβ2.1 (Fig. 9F). Therefore, these results indicated that PSD95 functions on Kvβ2.1 by an indirect mechanism stabilizing lipid rafts buffering PKC effects.
PKC activation triggers Kv1.3 endocytosis, increasing endosomal localization and lysosomal degradation of the channel, which is crucial for Kv1.3-related cell physiology [23]. Although mostly intracellular, Kvβ2.1 shares Kv1.3 spatial distribution and PKC-dependent regulation. Therefore, it was wondered whether PKC activation mediates a similar regulatory mechanism on Kvβ2.1. However, PMAdependent PKC activation neither increased the presence of Kvβ2.1 in early endosomes nor significantly augmented its ubiquitination (Fig. S7A-D). Furthermore, unlike Kv1.3 (Fig. S7E), the stability of Kvβ2.1 was not altered by PKCdependent activation. Thus, even after 24 h of PMA incubation neither lysosomal (BA, bafilomycin) nor proteasome MG (MG132) functions altered Kvβ2.1 abundance (Fig. S7  F,G). Our results indicated that Kvβ2.1, even in the absence of Kv1.3, targets the IS by a palmitoylation-dependent mechanism. S-acylation may also play a role in the functional coupling of Kvβ2 and Kv1.3, as has been for BK channel alpha and beta subunits [48], and such role will be discerned in future studies. The information points to the intrinsic importance of Kvβ2.1 function, independent of Kv1.3, during the immune response. In addition, the data show that palmitoylated Kvβ2.1 targets the lipid raft-enriched IS (Fig. 10). In fact, a signaling complex consisting of CD4, Kv1.3, Kvβ2, SAP97 (hDlg1) and ZIP clusters at the IS in human T cells [20,22,25]. In this scenario, PSD95 (also known as SAP90 from the synapseassociated protein-SAP-family), which stabilizes Kv1.3 at rafts, would share effects on Kvβ2.1. However, unlike Kv1.3, the mechanism would not involve direct association. Furthermore, while proliferation increased the amount of Kvβ2.1 in lipid rafts, PKC-dependent activation drove the peptide outside rafts, without compromising the peptide stability (Fig. 10).

Discussion
Voltage-dependent K + channels control the resting potential of mammalian cells. In addition, Kv participate in a myriad of physiological functions. To achieve that role, Kv associate with a number of ancillary proteins that further increase their function. The Kvβ regulatory subunit family is the most important group of peptides that, associating with many Kv channels, provide diversity [1]. We found that cytoplasmic Kvβ2.1 was located at the cell surface and, unlike Kvβ1.1, partially targeted lipid rafts in the absence of Kv subunits. This spatial localization, which situates Kvβ2.1 in the IS during the immune response, was mediated by specific S-acylation in two distal C-terminal cysteines (C301 and C311), which are far from the NADP + binding pocket. S-palmitoylation is related to lipid raft localization, and soluble proteins may route to the cell  [49]. Therefore, the results of this study are in line with those of palmitoylated SNAP-25, which routes to the plasma membrane, contributing to the formation of the SNARE complex [50]. Furthermore, the palmitoylation of the DLK kinase, critical for axon-tosoma retrograde signaling upon nerve injury, locates DLK in trafficking vesicles [51].
Kvβ1.1 and Kvβ2.1 are palmitoylated, but only the last targeted to lipid rafts. Although this location may be intriguing for a cytoplasmic protein, evidence situates this peptide in the IS complex in lymphocytes at the onset of the immunological response. Several palmitoyl acyltransferases, such as DHHC18 or DHHC13, are expressed in lymphocytes and may underlie the palmitoylation of Kvβs [35][36][37][38][39]. The data demonstrated that S-acylation controls the Kvβ2.1 location and concentrates this peptide in this location upon activation in cell conjugates. In addition, the differential lipid raft targeting of Kvβ2.1 by FBS-dependent proliferation or PMAinduced activation suggests that the abundance of Kvβ2.1 in these domains might serve to differentially regulate cell functions. Evidence demonstrates that Kvβ2.1 modulates the Kv1 and Kv4 families [52,53]. Upon specific combinations of Kvβ2.1/Kv channels, the half-voltage activation switches to more negative values [54]. Although Kvβ2.1 does not contain the ball-and-chain inactivation system, this protein enhances the surface expression of some channels. Therefore, a possible chaperone-like function for Kvβ2.1 was claimed [6]. However, little is known about the role of B Kvβ2.1+PMA A  [55]. These results suggest putative effects of Kvβ2.1-AKR activity on Kv currents. However, while the interaction with NADPH was required, most likely enhancing the proper folding of Kvβ2.1, the impairment of the enzymatic activity has no major changes on Kvβ2.1 function [56,57].
Considering that Kvβ2.1 underwent palmitoylation in the absence of α-subunits and that AKR activity is not necessarily coupled to control Kv channels, Kvβ2.1 could actindependent of Kv channels-as a sensor for the redox state of the cell. It was demonstrated that Kvβ1.1 does not target lipid rafts but is placed at the plasma membrane via actin interaction [7]. In this context, Kvβ1.3, closer to Kvβ1.1 than Kvβ2.1, exhibits a slower hydride transfer than Kvβ2.1 [58]. Therefore, the cell could place redox detectors with a specific sensing rate at the plasma membrane. Concomitantly, Kvβ2.1 can preferably reduce aldehydes and ketones of membrane-derived oxidized lipids, highlighting the importance of the Kvβ membrane location [13,14].
As mentioned above, Kvβ2.1 lipid raft localization was controlled by different insults. Upon proliferation, Kvβ2.1 increased its colocalization in lipid rafts. However, PMA impairs Kvβ2.1 presence in these subcellular domains and, similar to Kv1.3, this was partially counteracted by PSD95. The effect of PSD95 could involve an interaction with PKC rather than associating with Kvβ2.1. Therefore, Kvβ2.1 functions as a redox sensor at IS during the immune response, and MAGUK proteins stabilize the architecture of these multimeric domains. In these spots, several proteins interact with Kvβ2.1 [20,22]. For instance, ZIP1, ZIP2 and ZIP3 link this protein to PKCζ, which also supports the IS location. These partners may further interact with PKA, EB1, KIF3/kinesin II, TRPV1 and TREK, among others, triggering a complex scenario [11,21,[59][60][61]. Because Kvβ2.1 places in lipid rafts following palmitoylation, its partners may be translocated to those domains. This idea suggests that the Kvβ family, in the absence of Kv channels, could exert functions of signaling platforms. Kvβ2.1 interacts with TRPV1, augmenting channel trafficking to the plasma membrane [60]. This balance could be even more complex because other ion channel ancillary subunits, such as Navβ1, CaM, caveolin and KCNE4, modulate Kv1.3 [62][63][64]. In this context, the control of Kvβ2.1 distribution by FBS and PKC activation would fine-tune the Kv1.3 channelosome architecture. Further experiments need to be performed to decipher the redox control of these multicomplexes. Evidence highlights the idea that some functions of Kvβ2.1 have yet to be discovered. Kvβ2.1 null mice exhibit mild effects compared with severe consequences by annihilating Kv1.1 and Kv1.2 at the nervous system [16]. However, Kvβ2.1 association with Kv channels triggers sensitivity to hypoxia either by direct (Kv4.3) or indirect (Kv2.1) associations [65]. Furthermore, the fact that Kvβ2.1 AKR activity does not exert major changes in the control of Kv channels, suggesting that Kvβ2.1 could function as either a moonlighting protein or a scaffolding hub [56]. The NADP + binding ability of Kvβ subunits is required for cell surface trafficking of Kv1 channels in mammalian cells as well as axonal targeting [10,56]. The fact that palmitoylated Kvβ2 concentrates in IS during the immune response, also associating to many partners such as Kv1.3, ZIP1/2 and PKC in these microdomains, suggests an exciting yet deciphered physiological role.

Cell culture, transfections and pharmacological treatments
HEK-293 cells were cultured on DMEM (LONZA) containing 10% fetal bovine serum (FBS) supplemented with penicillin (10.000 U/ml), streptomycin (100 μg/ml) and 4 mM l-glutamine (Gibco). Human Jurkat T lymphocytes and Raji B cells were cultured in RPMI 1640 medium (Life Technologies) supplemented with 10% FBS and antibiotics. In some experiments, human CD4 + T cell subsets were used. Cells were isolated from peripheral whole blood using a negative selection Rosette Sep ™ kit from STEMCELL ™ Technologies. Human T lymphocytes were cultured at 37 °C and 5% CO 2 in RPMI 1640 medium (Life Technologies) supplemented with 10% FCS, 1% glutamine, 1% penicillin-streptomycin (Gibco), 1 × nonessential amino acid solution (Thermo Fisher Scientific), 10 mM HEPES (Life Technologies) and 50 U/ml IL-2 (Bionova). To generate T cell blasts, the Dynabeads Human T-Activator CD3/CD28 for T cell expansion and activation kit (Life Technologies) was used following the manufacturer's instructions. Human T cell blasts were used after 6-7 days of the expansion protocol. No IL-2 was supplemented in media the day before an experiment.
For confocal imaging and coimmunoprecipitation experiments, cells were seeded (70-80% confluence) in six-well dishes containing polylysine-coated coverslips or 100 mm dishes 24 h before transfection. Lipotransfectin ® (Attendbio Research) was used for transfection according to the supplier's instructions. The amount of transfected DNA was 4 µg for a 100 mm dish and 500 ng/well of a six-well dish (for coverslip use). Next, 4-6 h after transfection, the mixture was removed from the dishes and replaced with new fresh culture media. All experiments were performed 24 h after transfection.
In some experiments, 24 h transfected HEK cells were incubated with 1 µM PMA at 37 °C for indicated times. For 2-bromopalmitate treatment, HEK-transfected cells were starved for 3 h with DMEM supplemented with 1% dialyzed FBS, and 60 nM 2-bromopalmitate was added for 3 more hours at 37 °C. H 2 O 2 (400 µM) was added for 30 min at 37 °C.
Membrane-enriched preparations were purified from 24 h transfected HEK 293 cells. Cells were washed twice in cold PBS and scrapped in 1 mL of HB buffer (20 mM HEPES pH 7.4, 1 mM EDTA, 255 mM sucrose) supplemented with 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin and 1 mM phenylmethylsulfonyl fluoride as protease inhibitors. Samples were passed through a 25-gauge needle ten times, and lysates were centrifuged for 5 min at 3000 × g and 4 °C. The supernatant was diluted three times with HB buffer. To enrich the membrane fraction, lysates were centrifuged for 90 min at 150,000 × g. Finally, the pellet was resuspended in 30 μL 30 mM HEPES and analyzed by western blot.
For coimmunoprecipitation, 1 mg of protein was brought up to 500 μl with lysis buffer (NaCl 150 mM, HEPES 50 mM, Titron X-100 1%, pH 7.4) supplemented with protease inhibitors. Precleaning was performed with 40 μl of protein A Sepharose beads (GE Health care) for 1 h at 4 °C. Next, samples were incubated in a chromatography column (BioRad Micro spin Chromatography Columns), which contained 2.5 μg of anti-GFP antibody (Genescript) previously crosslinked to protein A Sepharose beads, for 2 h at room temperature (RT), with continuous mixing. Next, the columns were centrifuged for 30 s at 1000 × g. The supernatant (SN) was kept and stored at − 20 °C. Columns were washed four times with 500 μl of lysis buffer and centrifuged for 30 s at 1000 × g. Finally, elution was performed by incubating the columns with 100 μl of 0.2 M glycine pH 2.5 and spun for 30 s at 1000 × g. Eluted proteins (IPs) were prepared for western blotting by adding 20 μl of loading buffer (5×) and 5 μl of 1 M Tris-HCl pH 10.
Irreversible crosslinking of the antibody to the sepharose beads was performed by mixing the antibody with protein A sepharose beads for 1 h at RT. Next, the beads were incubated with 500 μl of dimethyl pimelimidate (DMP, Pierce) for 30 min at RT. Columns were washed four times with 500 μl of 1 × TBS, four times with 500 μl of 0.2 M glycine pH 2.5 and three more times with 1 × TBS. Finally, columns were incubated with protein lysates to perform immunoprecipitation, following the protocol described before.

Confocal microscopy and image analysis
For confocal analysis, cells were seeded on polylysinecoated coverslips 24 h prior to transfection. The next day, the cells were washed twice with PBS without K + (PBS-K + ), fixed with 4% paraformaldehyde for 10 min, and washed three times for 5 min with PBS-K + . Finally, coverslips were mounted on microscope slides (Acefesa) with house Mowiol mounting media. Coverslips were dried at RT for at least 1 day before imaging.
Staining with FITC-labeled cholera toxin β subunit (CTXβ) for lipid raft microdomains and wheat germ agglutinin (WGA)-Texas red (Invitrogen) for the plasma membrane was performed under non-permeabilized conditions [66]. Live cells (on ice) were quickly washed with PBS at 4 °C and stained with a dilution of WGA-Texas Red (1/1,500) in DMEM supplemented with 30 mM HEPES for 10 min at 4 °C. Subsequently, the cells were quickly washed twice and fixed with 4% paraformaldehyde in PBS for 10 min. Next, the cells were washed and mounted as described before. The EEA1 marker was used to stain early endosomes. Fixed cells were further permeabilized with 0.1% Triton X-100 for 10 min. After 60 min of incubation with blocking solution (10% goat serum, 5% nonfat powdered milk and 0.05% Triton X-100), primary mouse anti-EEA1 (1:500) antibody was added at 4 °C overnight (BD Transduction Laboratories) in 10% goat serum and 0.05% Triton X-100. Next, the cells were incubated with secondary goat anti-mouse antibody conjugated with Alexa Fluor 660 for 2 h at RT. Mounting as previously described [27].
All images were acquired with a Leica TCS SL laser scanning confocal spectral microscope (Leica Microsystems) equipped with argon and helium-neon lasers. All experiments were performed with a 63 × oil-immersion objective lens NA 1.32. Colocalization offline image analysis was performed using ImageJ software (https:// imagej. nih. gov/ ij/). A pixel-by-pixel colocalization study using JACoP (Just Another Colocalization Plugin) was used, and Mander's overlap coefficient (MOC) was calculated.

Immunological synapse formation
To analyze the Kvβ2 distribution during IS formation, human B cell lymphoma Raji cells were used as antigenpresenting cells. B cells were incubated with 10 µg/mL Staphylococcus enterotoxin E (SEE, Toxin Technologies) for 30 min. Next, Raji cells were mixed with Jurkat T cells (1:1) to form cell conjugates and spun at 200 × g for 1 min at 37 °C. Samples were plated on poly-L-lysine-coated coverslips and incubated for 15 min at 37 °C. The cells were washed once with PBS and fixed with 2% PFA in PBS for 10 min. Cells were rinsed three times with PBS between steps. Next, conjugates were labeled with anti-Kvβ2.1 (1/100, NeuroMab), anti-CD3 Alexa 647-conjugated antibody and anti-CD19 Alexa 488-conjugated antibody (1/100, BioLegend). Antibodies were diluted in PBS supplemented with 1% BSA and incubated for 2 h. Coverslips were rinsed and mounted in Mowiol.
Kvβ2 localization at the immune synapse was analyzed using ImageJ software. Briefly, region of interest (ROI) including the whole IS and equal surface areas in both Jurkat and Raji cell membranes outside the IS were drawn. The Kvβ2 signal intensity was measured in each membrane section and the intensity ratio calculated as follows: Therefore, ratios above 1 indicated that Kvβ2 intensities at the IS were higher than the sum of both Kvβ2 intensities in Jurkat and Raji membranes outside IS.

Cell unroofing preparations (CUP)
HEK-293 cells were seeded in poly-d-lysine-treated glass coverslips. Twenty-four hours after transfection, samples were cooled on ice for 5 min and washed twice in PBS-K + . Next, samples were incubated for 5 min in KHMgE buffer (70 mM KCl, 30 mM HEPES, 5 mM MgCl 2 , 3 mM EGTA, pH 7.5) diluted three times and gently washed with nondiluted KHMgE to induce hypotonic shock. Bursted cells were removed from the coverslip by intensive pipetting up and down. After two washes with KHMgE buffer, only membrane sheets remained attached. Preparations were fixed and mounted as previously described [67].

Lipid-raft isolation
Low-density Triton-insoluble complexes were isolated as previously described [68] from Jurkat T cells and HEK293 cells transiently transfected with Kvβ1.1CFP, Kvβ2.1CFP and the Kvβ2.1CFP ClessA mutant. Cells were homogenized in 1 ml of MBS (150 mM NaCl, 25 mM 2-morpholinoethanesulfonic acid 1-hydrate (MES), pH 6.5) 0.1% Triton X-100, and sucrose was added to a final concentration of 40%. A 5-30% linear sucrose gradient was layered on top and further centrifuged (39,000 rpm) for 20-22 h at 4 °C in a Beckman SW41-Ti rotor. Gradient fractions (1 ml) were sequentially collected from the top of the tube and analyzed by western blotting. While clathrin identified the nonbuoyant fractions, caveolin and flotillin labeled the floating fractions (lipid rafts) in HEK 293 cells and Jurkat T lymphocytes, respectively.

Half-life studies
Cells were preincubated with 100 µg/ml cycloheximide (CHX) for 3 h at 37 °C to halt protein synthesis. Next, cells were incubated in the presence or absence of 1 µM PMA at the indicated times. To inhibit lysosomal and proteasomal degradation, 60 nM bafilomycin A1 (BA) or 5 µM MG132 (MG), respectively, was added. Finally, the cells were washed twice with cold PBS, and protein extraction was performed as described above.

Acyl-biotin exchange (ABE) palmitoylation assay
Detection of protein palmitoylation by acyl-biotin exchange (ABE) was described previously [69]. Briefly, human CD4 + lymphocytes, Jurkat T cells and HEK 293 cells transfected with Kvβ2.1 wild type (WT) and Kvβ2.1 cysteine mutants were washed with cold PBS and harvested in buffer A (150 mM NaCl, 50 mM Tris HCl, 0.2% Triton X-100, 5 mM EDTA, pH 7.4) containing protease inhibitors and 10 mM NEM. Lysates were collected by scraping samples on ice, passed ten times through a 25-gauge needle and incubated for 1 h at 4 °C before centrifugation at 16,000 × g for 15 min. Supernatants were chloroform-methanol precipitated, and the pellet was allowed to air dry for 2-3 min. The pellet was resuspended in 300 µl of buffer B (4% SDS, 50 mM Tris HCl, 5 mM EDTA, pH 7.4) supplemented with 10 mM NEM and diluted fourfold in buffer A containing 1 mM NEM. Samples were incubated at 4 °C overnight with gentle agitation. NEM was removed by performing three sequential chloroform-methanol precipitations. Next, the pellet was resuspended in 500 µL of buffer B. The sample was split in two volumes, and 250 µL was diluted with 950 μL of buffer C (50 mM Tris-HCl, 1 mM HPDP-biotin, 0.2% Triton X-100, 100 µM PMSF) containing 0.7 M hydroxylamine (+ HA). The other 250 µL was diluted in buffer C without hydroxylamine (−HA). Samples were incubated at RT with gentle rocking for 1 h. Three rounds of chloroform-methanol precipitation were performed, and the final pellets were resuspended in 300 µL buffer B and diluted with 900 µL buffer D (150 mM NaCl, 50 mM Tris-HCl, 5 mM EDTA, 0.2 mM HPDP-biotin, 0.2% Triton X-100, pH 7.4). Samples were incubated at RT with gentle agitation for 1 h prior to 3 more sequential chloroform-methanol precipitations. The final pellet was resuspended in 120 µL of buffer E (2% SDS, 50 mM Tris HCl, 0,2% Triton X-100, 5 mM EDTA, pH 7.4) and diluted to 0.1% SDS with buffer A. Biotin-labeled proteins were captured with 50 µL of NeutrAvidin agarose beads, previously washed with 400 µL of buffer A and centrifuged for 1 min at 2000 × g. NeutrAvidine beads were pelleted by centrifugation at 2,000 × g for 30 s. After washing the beads four times with 400 µL of buffer A, captured proteins were boiled in 100 µL 1 × Laemmli buffer and 2% β-mercaptoethanol. Samples were then analyzed by SDS-PAGE western blot.