A Novel Mechanism for the Suppression of a Voltage-gated Potassium Channel by Glucose-dependent Insulinotropic Polypeptide PROTEIN KINASE A-DEPENDENT ENDOCYTOSIS*

The mechanisms involved in glucose regulation of insulin secretion by ATP-sensitive (K ATP ) and calcium-ac- tivated (K CA ) potassium channels have been extensively studied, but less is known about the role of voltage-gated (K V ) potassium channels in pancreatic (cid:1) -cells. The incretin hormone, glucose-dependent insulinotropic polypeptide (GIP) stimulates insulin secretion by poten-tiating events underlying membrane depolarization and exerting direct effects on exocytosis. In the present study, we identified a novel role for GIP in regulating K V 1.4 channel endocytosis. In GIP receptor-expressing HEK293 cells, GIP reduced A-type peak ionic current amplitude of K V 1.4 via activation of protein kinase A (PKA). Using mutant forms of K V 1.4 with Ala-Ser/Thr substitutions in a potential PKA phosphorylation site, C-terminal phosphorylation was shown to be linked to GIP-mediated current amplitude decreases. Proteinase K digestion and immunocytochemical studies on mutant K V 1.4 localization following GIP stimulation demon- strated phosphorylation-dependent rapid endocytosis of K V 1.4. Expression of K V 1.4 protein was also demon- strated in human (cid:1) -cells; GIP treatment resulting in similar decreases in A-type potassium current peak amplitude to those

Glucose-dependent insulinotropic polypeptide (GIP) 1 and glucagon-like peptide-1 are the two major intestinal hormones (incretins) involved in the stimulation of insulin secretion during a meal (1,2). glucose-stimulated insulin secretion is mediated via closure of ATP-sensitive K ϩ (K ATP ) channels resulting in membrane depolarization, activation of voltage-dependent Ca 2ϩ channels, and increases in intracellular Ca 2ϩ , followed by membrane repolarization by voltage-dependent K ϩ (K V ) and Ca 2ϩ -sensitive K (K CA ) channels (3)(4)(5). Incretins act by potentiating the events underlying membrane depolarization in addition to exerting direct effects on exocytosis. These events ultimately depend upon incretin interaction with its cognate seven-transmembrane G protein-coupled receptor and activation of proximal signal transduction pathways. In the case of GIP, these include stimulation of the adenylyl cyclase/cAMP/ protein kinase A (PKA) module, and activation of phospholipase A 2 (PLA 2 ) (6), protein kinase B (PKB) (7), and mitogenactivated protein kinases (8). There is little known regarding the effect of incretins on membrane repolarization of the ␤-cells (5).
Voltage-gated potassium channels (K V channels) belong to the six-transmembrane family of K ϩ channels consisting of K V 1 to K V 11 subfamilies (9) and are involved in repolarization of excitable cells (10). They are of interest as potential therapeutic targets in diabetes, because blockade of K V channels would be expected to prolong the pancreatic ␤-cell action potential, sustain the opening of voltage-dependent Ca 2ϩ channels, and thereby potentiate glucose-induced insulin release (5). Mammalian K V 1.Xs have been cloned and characterized in heterologous expression systems, and they generate at least two different types of outward potassium currents classified on the basis of their inactivation properties: delayed rectifier steady-state currents, which do not rapidly inactivate, and A-type transient currents, which inactivate rapidly by N-type inactivation (11,12). Delayed rectifier current is produced by K V 1.1, K V 1.2, K V 1.3, K V 1.5, and K V 1.6, and A-type current is produced by K V 1.4 (13,14). Although both types of currents are present in pancreatic ␤-cells, most of the studies have concentrated on the delayed-rectifier K V channels. The current study focused on a potential role for GIP in the regulation of K V 1.4, a prominent member of the K V 1 family and one of several channel proteins giving rise to potassium currents proposed to be involved in the regulation of insulin secretion. Using GIPRexpressing human embryonic kidney (HEK)-293 cells, we demonstrated that GIP decreases peak current amplitude of K V 1.4 via a process involving PKA activation. In parallel studies, K V 1.4 was shown to be expressed in human pancreatic ␤-cells, and GIP decreased A-type potassium current amplitude in these cells. Additionally, GIP was shown to induce phosphorylation and dynamin-dependent endocytosis of K V 1.4 and transient overexpression in INS-1 (832/13) ␤-cells of wild-type (WT) K V 1.4, or a T601A mutant form resistant to PKA phosphorylation, resulted in reduced glucose-stimulated insulin secretion; WT overexpression potentiated GIP-induced insulin secretion, whereas the loss of PKA-dependent phosphorylation (T601A cells) ablated this effect. These results therefore provide compelling evidence for a role for GIP-induced down-regulation of K V 1.4, via phosphorylation-dependent endocytosis of the channel protein, in the modulation of insulin secretion.

EXPERIMENTAL PROCEDURES
Generation of a GIPR-HEK293 Cell Line-HEK293 cells were grown in Dulbecco's modified Eagle's medium (Invitrogen), supplemented with 5% fetal bovine serum (Sigma-Aldrich), and penicillin-streptomycin (50 IU/ml, 50 g/ml, Invitrogen) and transfected with GIPR/pcDNA3 plasmid, expressing rat pancreatic islet GIP receptor cDNA under control of the cytomegalovirus promoter. Transfections were performed using Lipofectamine 2000 TM reagent (Invitrogen) for 4 h according to the manufacturer's instructions. Stably transfected cells were selected with G418 (Invitrogen), and GIPR-HEK293 cell clones were analyzed by quantitative real-time reverse transcription-PCR to check GIPR mRNA expression levels and by Western blotting to confirm GIPR protein expression, respectively.
cDNA Constructions of K V 1.4 Plasmids and Transient Transfections in GIPR-HEK293 Cells-K V 1.4 cDNA was cloned into the pEGFP-N2 vector (Clontech, Palo Alto, CA) and various constructs, as detailed under "Results," were prepared by PCR with HindIII and EcoRI insertions for directed cloning. Site-directed mutant constructs were prepared using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). All transfection plasmids were prepared using the Plasmid Midi kit (Qiagen, Valencia, CA). GIPR-HEK293 cells were plated at a density of 2 ϫ 10 5 cells/glass coverslip in 35-mm dishes. On the following day, transfection was performed with 1.5 g of the indicated K V 1.4 plasmids using Lipofectamine 2000 TM transfection reagent (Invitrogen), according to the manufacturer's instructions.
Enzyme Activity Assay of PKA-PKA activity was measured using a PKA kinase activity assay kit (Stressgen, Mississauga, Ontario) according to the manufacturer's protocol. The enzyme activity was normalized to protein concentration and shown as the relative activity to control. K V 1.4 protein Purification and in Vitro Phosphorylation-K V 1.4 cDNA was prepared by PCR and subcloned into the pGex4T3 vector (Amersham Biosciences). Glutathione S-transferase (GST)-K V 1.4 fusion protein was purified from BL21(DE3) Escherichia coli expressing pGex4T3/K V 1.4. GST fusion proteins were induced by treatment with 1 mM isopropyl-␤-D-thiogalactopyranoside (Sigma) for 4 h, and the bacteria were harvested by centrifugation, resuspended in PBST-100, pH 7.4 (phosphate-buffered saline, 1% Triton X-100, 1 mM EDTA) and sonicated (Branson Ultrasonic Corp, Brandury, CT) to release the proteins. GST-K V 1.4 fusion protein was affinity-purified using a glutathione-Sepharose 4B (Amersham Biosciences) column, and bound protein was eluted with 100 mM NaCl solution containing 25 mM reduced glutathione (Sigma). After affinity purification, K V 1.4 proteins were released from GST fusion protein through thrombin digestion. In vitro phosphorylation reactions were performed at 30°C in a volume of 40 l of MES buffer (50 mM MES, pH 6.9, 10 mM MgCl 2 , 0.5 mM EDTA, 1 mM dithiothreitol) and initiated with the addition of 5 Ci of [␥-32 P]ATP and unlabeled ATP to a final concentration of 100 M. For determination of the time course of phosphorylation of target protein, 50 g of K V 1.4 protein was incubated with 5 g/ml recombinant PKA catalytic subunit (active PKA, Sigma) or GIP-treated GIPR-HEK293 cellular extracts for 1 min to 2 h. Protein phosphorylation was determined by terminating the reactions with 2 ϫ SDS sample buffer, resolving the samples on 12.5% SDS-PAGE gels, and drying the gels for autoradiography.
Western Blot Analysis-Proteins (25 g of protein/well) from each sample were separated on a 12.5% SDS-PAGE gel and transferred onto nitrocellulose (Bio-Rad) membranes. Probing of the membranes was performed with antibodies against K V 1.4 and ␤-tubulin. Immunoreactive bands were visualized by enhanced chemiluminescence (Amersham Biosciences) using horseradish peroxidase-conjugated IgG secondary antibodies.
Islet Isolation and Cell Culture-Human islets were isolated from the pancreas of adult organ donors using collagenase duct perfusion, gentle dissociation, and density gradient purification at the Ike Barber Islet Transplantation Laboratory (Vancouver General Hospital, Vancouver, Canada) (15). The Research Ethics Board of the University of British Columbia provided ethics approval. Islets were dispersed to single cells and plated on laminin-coated glass coverslips in 35-mm dishes in CMRL Medium-1066 (Mediatech, Inc.) supplemented with 10% fetal bovine serum and penicillin-streptomycin (50 IU/ml-50 g/ml, Invitrogen).
Proteinase K Digestion Experiments-For proteinase K digestion (16,17), GIPR-HEK293 cells were transfected with a cDNA coding for K V 1.4 bearing a green fluorescent protein (GFP) tag at the C terminus (K V 1.4-EGFP). Cells incubated with 100 nM GIP for indicated periods of time were washed three times with ice-cold PBS and incubated with 10 mM HEPES, 150 mM NaCl, and 2 mM CaCl 2 (pH 7.4) with 200 g/ml proteinase K at 37°C for 30 min. The cells were then harvested, and proteinase K digestion was quenched by adding ice-cold PBS containing 6 mM phenylmethylsulfonyl fluoride and 25 mM EDTA. This was followed by SDS-PAGE and immunoblotting, and probing of the membranes was performed with antibodies against GFP and ␤-tubulin. Immunoreactive bands were visualized by enhanced chemiluminescence (Amersham Biosciences) using horseradish peroxidase-conjugated IgG secondary antibodies.
Confocal Microscopy-GIPR-HEK293 cells were transfected with WT or mutant forms of K V 1.4-EGFP plasmids and treated for 10 min with GIP (100 nM). Cells were then fixed and immunostained successively with ␣-GFP antibody and a secondary Alexa fluor® 488 dye-conjugated anti-rabbit antibody. Cell nuclei were counterstained with 4Ј,6-diamino-2-phenylindole, and transfected cells were imaged using a Zeiss laser scanning confocal microscope (Axioskop2).
Islet Embedding in Agar and Confocal Microscopy-Human islets were fixed in 4% paraformaldehyde in PBS for 30 min at room temperature. After spinning down, islets were resuspended in PBS, mixed with an equal volume of 2% agar solution, and added to a mold. Agaroseembedded sections were processed for a double immunostaining for K V 1.4 and insulin. The sections were incubated with K V 1.4 and insulin antibodies and visualized with Alexa fluor® 488-conjugated anti-mouse secondary antibody and Texas Red® dye-conjugated anti-rabbit antibody. Cell nuclei were counterstained with 4Ј,6-diamino-2-phenylindole and imaged using a Zeiss laser scanning confocal microscope (Axios-kop2). All imaging data were analyzed using the Northern Eclipse program (version 6).

GIP Decreases Peak Current Amplitude of K V 1.4 in GIPR-
HEK293 Cells-GIPR-expressing HEK293 cells (GIPR-HEK293) were used as an in vitro model system to characterize the modulation of K V 1.4 currents by GIP treatment. HEK293 cells were stably transfected with a rat islet GIPR cDNA construct under control of the cytomegalovirus promoter. The resulting GIPR-HEK293 cell clones transiently transfected with K V 1.4-EGFP cDNA (GIPR-HEK293-K V ) demonstrated voltageactivated, rapidly inactivating outward currents ( Fig. 1A) with properties similar to K V 1.4 currents previously reported in insulin-secreting cells (5). Administration of GIP-(1-42) to GIPR-HEK293-K V cells resulted in a decrease in peak current amplitude that was initiated within 1 min and decreased by 44.0 Ϯ 5.7% at 5 min (Fig. 1, A and B). The decrease in peak current amplitude was not observed in GIP-(1-42)-treated HEK293 cells, which do not express the GIPR (Control-I), or in GIPR-HEK293-K V cells treated with the non-insulinotropic truncated form of GIP (GIP- (19 -30); Control-II). The effect of GIPonthedecreaseofpeakcurrentamplitudewasconcentrationdependent, with 1 nM GIP producing a significant reduction (Fig. 1C). Other macroscopic properties of K V 1.4 were next studied to determine whether GIP affects the gating properties of K V 1.4. As shown in Fig. 1 (D-F), GIP did not alter the properties of activation (V 0.5 Control ϭ Ϫ27.6 Ϯ 2.5 mV versus V 0.5 GIP ϭ Ϫ25.1 Ϯ 1.1 mV), inactivation (V 0.5 Control ϭ Ϫ54.5 Ϯ 2.7 mV versus V 0.5 GIP ϭ Ϫ52.9 Ϯ 1.7 mV) or recovery from inactivation ( Control ϭ 4.6 Ϯ 0.6 s versus GIP ϭ 4.9 Ϯ 0.3 s), indicating that the effect of GIP is not mediated by changes in the macroscopic gating properties of K V 1.4.
GIP Activates PKA in GIPR-HEK293-K V Cells and GIP-stimulated PKA Activation Is Involved in the Decrease of K V 1.4 Peak Current Amplitude-Mechanisms involved in the decrease of K V 1.4 peak current amplitude by GIP treatment were next studied. The possibility that changes in K V 1.4 peak current amplitude were linked to activation of adenylyl cyclase and PKA was first examined, because this is a well established GIP signaling module involved in the regulation of insulin secretion. Treatment of GIPR-HEK293-K V cells with 100 nM GIP resulted in increased PKA activity ( Fig. 2A), apparent 10 min after initiation of GIP treatment, thereafter quickly decreasing with time. GIP stimulation was concentration-dependent, with an EC 50 value of 2.10 Ϯ 0.22 nM (Fig. 2B). To establish that GIP-induced activation of PKA was associated with decreases in K V 1.4 peak current amplitude, H-89 and (R p )-cAMP, selective inhibitors of PKA, were applied during electrophysiological recordings. The inhibitor H-89 significantly blocked GIP-stimulated PKA activation (Fig. 2, A and B) and the decreases of K V 1.4 peak current amplitude (Fig. 2C). Similarly, the cAMP antagonist (R p )-cAMP eliminated the effect of GIP on K V 1.4 peak current amplitude (Fig. 2C). Taken together, these results demonstrate that GIP-stimulated PKA activation is involved in the regulation of K V 1.4.  -II). C, concentration dependence of decreases in K V 1.4 peak current amplitude in response to GIP. Cells were prepared as described above and incubated with different concentration of GIP (in nM: 0, 1, 10, and 100). Currents were recorded at 5 min after GIP treatment. Peak current amplitudes at ϩ60 mV were normalized to cell capacitance and represented as percentage ratio to 0 nM GIP group (maximum ϭ 1). D-F, effect of GIP on macroscopic K V 1.4 current kinetics. Activation properties (D), inactivation properties (E), and recovery time from inactivation (F) of K V 1.4 were determined during treatment with GIP. All data represent the mean Ϯ S.E. from three to five independent experiments, and significance was tested using Student's t test or analysis of variance with a Newman-Keuls post hoc test, where the asterisk represents p Ͻ 0.05 versus basal.

GIP-stimulated PKA Activation Resulted in the Phosphorylation of K V 1.4 and Decreases in K V 1.4 Peak Current
Amplitude-To determine whether PKA could phosphorylate K V 1.4, purified recombinant K V 1.4 protein was incubated with the catalytic subunit of PKA and [ 32 P]ATP in vitro. PKA-induced phosphorylation of K V 1.4 protein was apparent 1 min after the initiation of PKA treatment, and it was sustained for 2 h (Fig.  3A). Similar rapid onset and sustained phosphorylation of K V 1.4 was observed with GIP-treated GIPR-HEK293-K V cellular extracts (Fig. 3B), indicating the involvement of GIP-stimulated PKA activation for the phosphorylation of K V 1.4. To identify the functional region of K V 1.4 involved in GIP-mediated phosphorylation, electrophysiological responses to GIP were recorded in K V 1.4 mutants with deletions at the N or C termini. As shown in Fig. 3C, ⌬N147 K V 1.4, with 147 amino acids deleted from the N terminus, showed delayed inactivation compared with WT K V 1.4. The delayed inactivation has been previously shown to result from deletion of an N-terminal inactivating "ball" domain, which acts as an intracellular tethered blocker of the open channel (12). GIP was found to retain the ability to reduce peak current amplitude expressed by ⌬N147 K V 1.4. On the other hand, the mutant ⌬C55 K V 1.4, with 55 amino acids deleted from C terminus, demonstrated no responsiveness to GIP (Fig. 3C). These results therefore indicate that the C terminus in K V 1.4 is responsible for GIPmediated effects. Inspection of the amino acid composition of the C terminus of K V 1.4 revealed several motifs that are potentially involved in regulating its subcellular localization and interaction with other proteins. The VXXSL motif is critical for glycosylation and surface membrane localization of the channel (18), and the sequence ETDV is the binding site for PDZ domain proteins, such as PSD-95 and SAP-97 (19,20). Additionally, we found a potential PKA phosphorylation site, SSTSSS, in the C terminus of K V 1.4. To determine whether phosphorylation of K V 1.4 is linked to GIP-mediated decreases in current amplitude, studies were performed on the effect of substituting alanine for each of the phosphorylatable serine and threonine residues: S599A, S600A, T601A, S602A, S603A, and S604A. When in vitro phosphorylation of the purified alanine-substitution mutants was examined, all proteins, except T601A, were phosphorylated by PKA (Fig. 3D). Electrophysiological recordings showed that GIP was able to reduce peak current amplitude with GIPR-HEK293 cells expressing S599A, S602A, S603A, and S604A, but not T601A (Fig. 3E). The mutant S600A was not functionally expressed in the plasma membrane (data not shown). Taken together, these results indicate that phosphorylation of K V 1.4 by GIP-stimulated PKA activation is linked to the decrease in peak current amplitude and phosphorylation of Thr-601 in the C terminus is a critical step in this process.
GIP-induced Decreases in K V 1.4 Peak Current Amplitude Result from Channel Endocytosis-Although phosphorylation of ion channels has been reported in several systems, in most cases it is associated with changes in inactivation rate and recovery from inactivation, rather than changes in peak current amplitude. Membrane protein phosphorylation directly affects interactions with intracellular proteins involved in many cellular pathways, and it was hypothesized that GIPmediated decreases in K V 1.4 peak current amplitude were mediated by phosphorylation-associated endocytosis. To examine this proposal, the cellular localization of K V 1.4 during treatment with GIP was studied using proteinase K digestion. The GFP tag on Kv1.4 is located intracellularly at the C terminus. Proteinase K is able to randomly digest extracellular regions of K V 1.4 channels in the plasma membrane, thus producing a digested form, whereas intracellular channel is protected. As shown in Fig. 4A, surface membrane expression levels of K V 1.4 (S) decreased with GIP treatment, and this was evident at 5 min following initiation of treatment. Confocal microscopy also revealed that K V 1.4 channels Peak current amplitudes at ϩ60 mV were normalized to cell capacitance and represented as percentage ratio to Control (maximum ϭ 1). All data represent the mean Ϯ S.E. from four to six independent experiments, and significance was tested using Student's t test or analysis of variance with a Newman-Keuls post hoc test, where the asterisk represents p Ͻ 0.05 versus basal.

FIG. 3. GIP-stimulated PKA activation resulted in the phosphorylation of K V 1.4 and decreases in ionic current. A and B, in vitro
phosphorylation of K V 1.4 by PKA. 50 g of purified K V 1.4 protein was incubated for the indicated periods of time at 30°C in kinase buffer containing 5 Ci of [␥-32 P]ATP in the presence of recombinant protein kinase A catalytic subunit (A) and GIP-stimulated GIPR-HEK293 cellular extracts (B). The reactions were terminated by boiling in SDS sample buffer, and proteins were separated by SDS-PAGE. The gels were Coomassie-stained to visualize protein, dried, and exposed to x-ray film to visualize incorporated 32 P. C, electrophysiological recordings. Schematic diagram of the K V 1.4 serial deletion constructs and their functional elements are presented. Peak current amplitudes are shown at the far right, for comparison between WT and N-terminally deleted (⌬N147) or C-terminally deleted mutants (⌬C55) of Kv1.4. Currents were recorded at zero time (black bar) and 5 min after GIP treatment (white bar), and peak current amplitudes at ϩ60 mV were normalized to cell capacitance and represented as percentage ratio (maximum ϭ 1). In present in the plasma membrane were greatly reduced by treatment with GIP (Fig. 4B). To determine whether GIPmediated decreases in K V 1.4 peak current amplitude were mediated by dynamin-dependent endocytosis, electrophysiological recordings were performed on GIPR-HEK293-Kv cells treated with GIP (100 nM) in the presence or absence of myristoylated dynamin inhibitory peptide (mry-DIP: myr-Gln-Val-Pro-Ser-Arg-Pro-Asn-Arg-Ala-Pro-NH 2 ) (21). Dynamin is a large GTPase implicated in the budding and scission of nascent vesicles from parent membranes (22). It has been extensively studied in the context of clathrin-coated vesicle budding from the plasma membrane, but it is also involved in the budding of clathrin-coated vesicles from other compartments and budding of caveoli, phagocytosis, and vesicle cycling at synapses (23,24). As shown in Fig. 4C, mry-DIP completely abolished the effect of GIP on K V 1.4. These results therefore strongly suggest that GIP induces the endocytosis of K V 1.4 and that these processes are responsible for the GIP-mediated decrease in K V 1.4 peak current amplitude.
Phosphorylation Is Involved in GIP-induced Retrograde Trafficking of K V 1.4 -Next, we investigated the potential relationships between phosphorylation and endocytosis of K V 1.4 in response to treatment with GIP. The mutant channels S599A, T601A, S602A, S603A, and S604A were transiently expressed in GIPR-HEK293 cells, and confocal microscopy was performed on fixed cells following treatment with GIP. Endocytosis was observed with S599A-, S602A-, S603A-, and S604Atransfected GIPR-HEK293 cells with GIP treatment, but not with T601A-transfected cells (Fig. 5). These results correlate well with the electrophysiological recordings, and the combined data demonstrate that phosphorylation and endocytosis are consecutive processes responsible for the effects of GIP on K V 1.4 channel distribution.
K v Channel Expression in Human Islets-K V channels have been shown to play an important role in the regulation of glucose-dependent insulin secretion in rodent islets (5). As shown in Fig. 6, K V 1.4 protein is also expressed in human islets , and cells with GIP treatment are shown in the right panels (ϩGIP), respectively. Immunocytochemical staining was performed using ␣-GFP antibody and imaged using a Zeiss laser scanning confocal microscope (Axioskop2). All imaging data were analyzed using the Northern Eclipse program (version 6), and the scale bar indicates 10 m. Immunocytochemical staining was performed using ␣-GFP antibody and imaged using a Zeiss laser scanning confocal microscope (Axioskop2). All imaging data were analyzed using the Northern Eclipse program (version 6), and the scale bar indicates 10 m. C, electrophysiological recordings. GIPR-expressing HEK293 cells were transfected with K V 1.4-EGFP and treated for the indicated periods of time with 100 nM GIP in the presence or absence of mry-DIP. Peak current magnitude at each time point was normalized to the peak current from zero time, and example traces from mry-DIP-treated cells are shown in the inset, and the scale bar indicates 10 nA. All data represent the mean Ϯ S.E. from four to six independent experiments, and significance was tested using Student's t test, where the asterisk represents p Ͻ 0.05 versus basal.
(A), mainly restricted to insulin-expressing pancreatic ␤-cells (B). Electrophysiological recordings from human islet cells revealed a typical A-type outward potassium current, and 100 nM GIP treatment resulted in a decrease in A-type peak current amplitude with a similar pattern to that observed in K V 1.4 currents in GIPR-HEK293-K V (Fig. 6, C and D). The effect of GIP on peak current amplitude was reversible by washing-out, and by 7 min following wash-out responses had almost returned to levels achieved prior to treatment with GIP. To determine whether A-type current in human islets is mediated by K V 1.4, 4-aminopyridine (4-AP; 1 mM), a conventional K V channel blocker, was applied (10). As shown in Fig. 6E, 4-AP treatment resulted in ϳ50% reduction in A-type potassium current. A-type potassium channels of different types exhibit variable sensitivity to 4-AP. For example, with cloned K V 1.4 channels 73% of peak current was found to be blocked by 1 mM 4-AP (25), whereas, in contrast, K V 4.2 was 7-fold less sensitive to 4-AP than K V 1.4 (26) and K V 3.4 exhibited much greater sensitivity to 4-AP (micromolar range) (27). The sensitivity to 4-AP exhibited by the A-type current in human islets suggests that it is mediated by K V 1.4, and that GIP is able to confer its effect on peak current amplitude in a similar manner to that observed with GIPR-HEK293-K V cells.
Phosphorylation-dependent Internalization of K V 1.4 Participates in the Effect of GIP on Insulin Secretion-To establish that phosphorylation-dependent internalization of K V 1.4 contributes to GIP stimulation of insulin secretion, the T601A mutant form, which is resistant to PKA phosphorylation, or WT K V 1.4 channel, were transiently expressed in insulin-secreting ␤-INS-1 cells and channel internalization determined by proteinase K treatment. The ␤-INS-1 cell line (clone 832/13) was chosen because it lacks functional A-type current (data not shown), thereby excluding the involvement of endogenous K V 1.4 current in responses to GIP. As shown in Fig. 7A, GIP treatment did not decrease surface membrane expression of mutant T601A under either low (2.5 mM) or high (25 mM) glucose conditions, whereas WT K V 1.4 was internalized by GIP treatment in the presence of high, but not low, glucose. Expression of either WT or T601A mutant K V 1.4 channels reduced glucose-stimulated insulin secretion, compared with vectortransfected cells (Fig. 7B), presumably because of a prolonged repolarization phase of the ␤-INS-1 cell action potential. As expected, GIP treatment did not significantly increase insulin secretion in the presence of low glucose in pEGFP-N2 vector-, T601A-, or WT K V 1.4-transfected groups. However, under high glucose conditions, GIP treatment resulted in increased insulin secretion in all groups. GIP responses of ␤-INS-1 cells overexpressing WT K V 1.4 channels were potentiated compared with pEGFP-N2-transfected cells, whereas the loss of PKA-dependent phosphorylation (T601A cells) ablated this effect. Taken together, these results strongly suggest that phosphorylationdependent internalization of K V 1.4 is an important component of GIP-potentiated insulin secretion.

DISCUSSION
The ability of GIP to directly enhance glucose-stimulated insulin secretion in pancreatic ␤-cells has been attributed to GIPR activation leading to enhanced depolarization and increases in the intracellular calcium concentration as well as direct effects on insulin exocytosis (3,4,28). Ion channels are the primary determinants of membrane excitability in most cells, and they are regulated to maintain membrane potentials within specific limits. Frequently this occurs through modulation of the functional responses of the ion channel to extracellular stimuli. In pancreatic ␤-cells, insulin secretion is modulated by the activity of several different ionic currents. Among FIG. 6. K V channel expression in human islets. A, determination of K V 1.4 expression in human islets by Western blot analysis. Total cellular extracts were prepared from human islets, and Western blot analyses were performed using antibodies against K V 1.4 and ␤-tubulin. Total cellular extracts from GIPR-HEK293 cells and K V 1.4-transfected GIPR-HEK293 cells (GIPR-HEK293 plus K V 1.4) were used as negative and positive controls for Western blotting, respectively. B, determination of K V 1.4 expression in human pancreatic ␤-cells by confocal microscopy. Human islets were embedded in agar, and immunohistochemical staining was performed using K V 1.4 and insulin antibody and imaging using a Zeiss laser scanning confocal microscope (Axioskop2). All imaging data were analyzed using the Northern Eclipse program (version 6), and the scale bar indicates 50 m. C-E, cells from human islets were briefly hyperpolarized to Ϫ100 mV followed by depolarization to ϩ60 mV for 900 ms from a holding potential of Ϫ80 mV. All whole cell recordings were performed at room temperature (20 -23°C). C, example of current traces measured in human islet in response to treatment with GIP. Scale bar indicates 2 nA. D, time course of decreases in A-type peak current amplitude in response to 100 nM GIP. Peak current magnitude at each time point was normalized to the peak current from zero time. E, effect of 1 mM 4-AP on A-type current in human islet. All data represent the mean Ϯ S.E. from three to six independent experiments, and significance was tested using Student's t test, where the asterisk represents p Ͻ 0.05 versus basal. The scale bar indicates 2 nA. these are the three main potassium currents: inward rectifying potassium currents, including the ATP-sensitive (K ATP ) channel and others, calcium-activated (K Ca ), and voltage-gated (K V ) currents (29 -32). The molecular mechanisms involved in the regulation of K ATP and K CA channels in pancreatic ␤-cells have been extensively studied, but considerably less is known about the K V channels. Because these channels are considered to be potential therapeutic targets for type 2 diabetes, it is important to establish their physiological role and mechanisms involved in regulating their activity. The current study was therefore initiated with the objective of identifying potential interactions between the incretin hormone GIP and K V channels. GIP transduces its biological effects on pancreatic ␤-cells by interacting with a seven-transmembrane receptor, GIPR, which is a member of the class II G protein-coupled receptor subfamily. The best characterized pathway by which GIP acts on insulin secretion in ␤-cells involves activation of the adenylyl cyclase/ cAMP/PKA pathway. Using HEK293 cells co-expressing the GIPR and Kv1.4 (GIPR-HEK293-K V cells), we have now shown that GIP reduces peak current amplitude of K V 1.4 channels via a pathway inhibited by the selective inhibitors of PKA, H-89, and the cAMP antagonist, (R p )-cAMP. In parallel experiments, it was shown that recombinant PKA catalytic subunits (Fig.  3A) or cell extracts from GIP-stimulated GIPR-HEK293-K V cells (Fig. 3B) increased phosphorylation of K V 1.4, and active PKA phosphorylated Thr-601 in the C terminus of K V 1.4 (Fig.  3D), thus substantiating the involvement of PKA signaling in GIP-induced effects on K V 1.4 current. This was confirmed by experiments showing that mutant T601A K V 1.4 channels could not be phosphorylated by PKA, and peak currents in this mutant were resistant to GIP (Fig. 3, D and E).
The macroscopic current in cells is regulated by the following two processes: 1) biophysical and biochemical modulation of surface membrane ion channel activity and 2) biosynthesis and trafficking of channel protein (33). Direct phosphorylation of channel proteins by serine/threonine and tyrosine kinases has been established as a mechanism by which ion channels are regulated. The delayed rectifier potassium channel K V 1.2 was the first example of a voltage-gated ion channel shown to be regulated by Ser/Thr phosphorylation (34,35) and a range of voltage-and ligand-gated channels have been found to be regulated by tyrosine kinases, including N-methyl-D-aspartic acid receptors, voltage-dependent Ca 2ϩ channels, and a variety of potassium channels (36,37). Previous studies have addressed the effects of Ser/Thr phosphorylation of the N-terminal domain of K V 1.4 on physiological responses. Calcium/calmodulindependent protein kinase has been shown to slow the inactivation of K V 1.4 currents by phosphorylating Ser-123 in the cytoplasmic N terminus (38). Treatment of K V 1.4-expressing Xenopus oocytes with phorbol 12-myristate 13-acetate, a protein kinase C activator, has been shown to lead to a biphasic change in the magnitude of peak current: an initial increase followed by a later reduction (39). Although in most cases the precise mechanisms underlying the effects of Ser/Thr phosphorylation on channel function are unclear, the most commonly suggested mechanism is that phosphorylation-induced changes in channel structure alter its biophysical properties (40). In the present study, GIP reduced K V 1.4 peak current amplitude, without affecting macroscopic gating properties of K V 1.4 (Fig.  1, A-F), and threonine phosphorylation of the C terminus by GIP-stimulated PKA activation also resulted in a decrease in K V 1.4 peak current amplitude (Fig. 3). These results imply that different mechanisms are involved, compared with previously reported phosphorylation of K V channels.
The trafficking of ion channels is one of the processes involved in the modulation of plasma membrane macroscopic currents (33). The regulation of expression of K V channels in the plasma membrane begins at the level of gene transcription and biosynthesis of the channel protein (41), with further control provided during insertion of the channel into the cell surface and by its regulated retrieval and degradation. Endocytosis was initially defined as a process by which substances are taken into the cell, but it is now recognized as an essential mechanism for the regulation of a variety of membrane proteins. Endocytosis is a first-order mechanism of internalization of membrane-bound proteins undergoing recycling or retrograde trafficking to be degraded. Endocytosis initiated by phosphorylation of K V channels results in decreased ionic current density (35). In the present study, it was demonstrated that direct phosphorylation of K V 1.4 by GIP-stimulated PKA activation is involved in endocytosis of the channel protein (Figs. 3 and 5). Retrograde trafficking of K V 1.4 resulting in decreased peak current amplitude was observed following treatment with GIP (Fig. 4, A and B), and dynamin-dependent endocytosis was involved in this process (Fig. 4C). In contrast, the nonphosphorylatable mutant T601A K V 1.4 was incapable of undergoing endocytosis, demonstrating the critical role played by phosphorylation in GIP-induced endocytosis of K V 1.4 (Fig. 5). The underlying molecular mechanism by which PKA-dependent phosphorylation is linked to endocytosis of K V 1.4 is not clear at the  (25 mM). Insulin release into the medium was determined by radioimmunoassay. All data represent the mean Ϯ S.E., and significance was tested using analysis of variance with a Newman-Keuls post hoc test, where the asterisk represents p Ͻ 0.05 versus high glucose (ϪGIP; WT K V 1.4); # represents p Ͻ 0.05 versus high glucose (ϩGIP; T601A); and & represents p Ͻ 0.05 versus high glucose (ϩGIP; pEGFP-N2). present time. Post-translational modifications of channel proteins by signaling molecules and resulting structural changes of channel proteins may affect protein-protein interactions between channel proteins and proteins involved in the endocytotic pathway.
The K V 1.4 channel was also demonstrated to be present in human pancreatic ␤-cells, and GIP treatment decreased A-type ionic current amplitude (Fig. 6). There have been controversial reports regarding the expression patterns of K V 1 family channels in human islets. It has been reported that K V 1.1, K V 1.2, and K V 1.4 are not found by reverse transcription-PCR in human islets, whereas K V 1.5 and K V 1.6 are present (5). On the other hand, Yan et al. (42) reported that only K V 1.3 and K V 1.6 were detected by reverse transcription-PCR, with a very weak indication for K V 1.7. In the current study, we have shown that K V 1.4 protein is expressed in human pancreatic ␤-cells (Fig.  6B). These discrepant results might arise from phenotypic differences in channel composition between races or individuals.
What is the likely effect of A-type potassium current downregulation on pancreatic ␤-cell function? A-type current has an important role in the early repolarization phase of action potentials (10,43). 4-AP potentiates insulin secretion from rat islets and insulinoma cells stimulated by sulfonylureas, even in the absence of glucose (5). Therefore, it is reasonable to predict that the down-regulation of A-type current mediated by GIPstimulated K V 1.4 endocytosis in pancreatic ␤-cells would enhance the duration and amplitude of action potentials, resulting in the prolongation of insulin secretion. This is supported by the demonstration that transient overexpression of WT K V 1.4 channels in INS-1 (832/13) ␤-cells resulted in potentiated insulin secretion in response to GIP, and that loss of a PKA phosphorylation site ablated this effect (Fig. 7B).
In summary, GIP-induced phosphorylation of K V 1.4 channel protein, resulting in endocytosis and decreases of ionic peak current amplitude, is likely to be an important pathway by which GIP acts as an insulinotropic hormone. This appears to be the first example of a physiological pathway directly linking hormone signaling to endocytosis of K V channels. The combined effects of GIP and glucagon-like peptide-1 account for ϳ50% of the total insulin response to a meal, and it is therefore clear that a deeper understanding of its mechanism of action is an important issue, with strong implications for the development of therapeutic agents for type 2 diabetes.