KCNMB1 regulates surface expression of a voltage and Ca2+-activated K+ channel via endocytic trafficking signals

Neuroscience 142 (2006) 661– 669

KCNMB1 REGULATES SURFACE EXPRESSION OF A VOLTAGE AND Ca2ⴙ-ACTIVATED Kⴙ CHANNEL VIA ENDOCYTIC TRAFFICKING SIGNALS Key words: voltage and Ca2ⴙ-activated Kⴙ channel, endoplasmic reticulum, human MaxiK ␣ subunit, KCNMB1, endocytic signals.

B. TORO,f N. COX,e R. J. WILSON,e E. GARRIDO-SANABRIA,e E. STEFANI,a,b,d L. TOROa,c,d AND M. M. ZAREIe* a

Department of Anesthesiology, UCLA, Los Angeles, CA 90095, USA

Voltage-dependent and Ca2⫹-activated K⫹ (MaxiK) channels are composed of the pore-forming ␣ and modulatory ␤ subunits. The ␣ subunit (Slo) has seven transmembrane domains (S0 –S6) (Meera et al., 1997) and its functional properties like voltage/Ca2⫹ sensitivities and activation/ deactivation kinetics are modulated by four ␤ subunits (␤1–␤4) (McManus et al., 1995; Meera et al., 1996, 2000; Wallner et al., 1999; Uebele et al., 2000; Xia et al., 1999; Brenner et al., 2000a). MaxiK channels are expressed in multiple tissues and play important roles in body functions like blood flow, uresis, and neurotransmission (review see (Lu et al., 2006)). Consistent with their role in neurotransmission, MaxiK channel ␣-subunit has been observed in axons and nerve terminals (Knaus et al., 1996), and its deficiency disturbs cerebellar function (Sausbier et al., 2004). On the other hand, the physiological role of ␤1 subunit has been highlighted by its reduction in models of hypertension, during aging and by gene silencing causing hypertension (Brenner et al., 2000b; Pluger et al., 2000; Amberg et al., 2003; Nishimaru et al., 2004). Thus, it is likely that expression levels of both subunits dictate physiological outcomes as is the case for reduced expression of Slo at the end of pregnancy perhaps supporting a timely delivery (Song et al., 1999); whereas in the brain, neuronal activity may be differentially regulated by Slo protein expression levels as predicted by its specific expression in particular regions of the brain (Knaus et al., 1996). Slo expression levels may be regulated by several mechanisms including regulation of transcript levels and regulation of intracellular trafficking. Previous studies have shown a direct correlation between the levels of Slo mRNA and its protein expression in smooth muscle (Song et al., 1999). On the other hand, few studies have addressed the mechanisms that govern Slo intracellular traffic; in particular, its forward traffic from the endoplasmic reticulum (ER) to the plasma membrane. Constitutive signals in Slo carboxyl terminus seem essential for its surface targeting (Bravo-Zehnder et al., 2000; Wang et al., 2003; Schmalhofer et al., 2005); while splice variation can also contribute to Slo surface regulation (Zarei et al., 2001). Specifically, Slo splice variant insert, SV1, diminishes Slo surface expression via a retention/retrieval signal sequence that prevents its forward traffic from the ER to the surface (Zarei et al., 2004). However, traffic mechanisms involved

b

Department of Physiology, UCLA, Los Angeles, CA 90095, USA

c

Department of Molecular and Medical Pharmacology, UCLA, Los Angeles, CA 90095, USA

d

Brain Research Institute, UCLA, Los Angeles, CA 90095, USA

e

Center for Biomedical Studies, UTB/TSC, Brownsville, TX 78520, USA

f

South Texas Arthritis Center, Brownsville, TX 78526, USA

Abstract—Voltage-dependent and calcium-activated Kⴙ (MaxiK, BK) channels are ubiquitously expressed and have various physiological roles including regulation of neurotransmitter release and smooth muscle tone. Coexpression of the pore-forming ␣ (hSlo) subunit of MaxiK channels with a regulatory ␤1 subunit (KCNMB1) produces noninactivating currents that are distinguished by high voltage/Ca2ⴙ sensitivities and altered pharmacology [McManus OB, Helms LM, Pallanck L, Ganetzky B, Swanson R, Leonard RJ (1995) Functional role of the beta subunit of high conductance calciumactivated potassium channels. Neuron 14:645– 650; Wallner M, Meera P, Ottolia M, Kaczorowski G, Latorre R, Garcia ML, Stefani E, Toro L (1995) Characterization of and modulation by a ␤-subunit of a human maxi KCa channel cloned from myometrium. Receptors Channels 3:185–199]. We now show that ␤1 can regulate hSlo traffic as well, resulting in decreased hSlo surface expression. ␤1 subunit expressed alone is able to reach the plasma membrane; in addition, it exhibits a distinct intracellular punctated pattern that colocalizes with an endosomal marker. Coexpressing ␤1 subunit with hSlo, switches hSlo’s rather diffuse intracellular expression to a punctate cytoplasmic localization that overlaps ␤1 expression. Furthermore, coexpressed ␤1 subunit reduces steady-state hSlo surface expression. Site-directed mutagenesis underscores a role of a putative endocytic signal at the ␤1 C-terminus in the control of hSlo surface expression. We propose that aside from its well-established role as regulator of hSlo electrical activity, ␤1 can regulate hSlo expression levels by means of an endocytic mechanism. This highlights a new ␤1 subunit feature that regulates hSlo channels by a trafficking mechanism. © 2006 IBRO. Published by Elsevier Ltd. All rights reserved. *Correspondence to: M. Zarei, UTB/TSC, Department of Biology, LHSB 2.804A, 80 Fort Brown, Brownsville, TX 78520, USA. Tel: ⫹1-956-882-5069; fax: ⫹1-956-882-5065. E-mail address: [email protected] (M. Zarei). Abbreviations: anti-c-Myc-FITC, anti-c-Myc monoclonal antibody conjugated to FITC; ER, endoplasmic reticulum; GFP, green fluorescent protein; hSlo, human voltage-dependent and Ca2⫹-activated K⫹ channel ␣ subunit; MaxiK, BK, voltage-dependent and Ca2⫹-activated K⫹.

0306-4522/06$30.00⫹0.00 © 2006 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2006.06.061

661

662

B. Toro et al. / Neuroscience 142 (2006) 661– 669

in Slo removal from the plasma membrane are yet unknown. Protein removal from the plasma membrane may be regulated by endocytic sorting signals. In most cases, these signals are located within the cytosolic domains of transmembrane proteins and can be categorized in two types: a) “tyrosine-based” (YXXØ) signals, where X is any amino acid, and Ø is a bulky hydrophobic amino acid residue such as leucine (Ktistakis et al., 1990); and b) “dileucine based” (LL) signals that can also have a leucine replaced with a hydrophobic residue such as isoleucine (Letourneur and Klausner, 1992; Bremnes et al., 1994; Bonifacino and Traub, 2003). Interestingly, these sequences can also target transmembrane proteins to lysosomes (Williams and Fukuda, 1990; Gough et al., 1999). A number of ion channels utilize endocytic signals to regulate their surface expression (Roche et al., 2001; Hu et al., 2001); however, the role of these type of signals in regulating surface expression of MaxiK channel remains unknown. Here we show that Slo surface expression levels can be regulated by ␤1 subunit coexpression via an endocytic mechanism that utilizes signals present in the ␤1 subunit.

EXPERIMENTAL PROCEDURES Molecular biology Two ␤1 mutant constructs were prepared by substituting tyrosine (Y), isoleucine (I) and/or leucine (L) to alanine (A) by site-directed mutagenesis: Mutant 1, I186A, L187A; mutant 2, Y183A, I186A. All constructs are in pcDNA3 vector (Invitrogen, Carlsbad, CA, USA). The human voltage-dependent and Ca2⫹-activated K⫹ channel pore-forming ␣ (hSlo) subunit was tagged with the c-Myc epitope at the N terminus (extracelluar). Tagged and untagged hSlo have similar properties (Meera et al., 1997) and for simplicity, we refer to the c-Myc-tagged construct as hSlo. GenBank accession numbers are U11058 for hSlo and U25138 for the human ␤1 subunit (KCNMB1). RhoB GTPase-GFP was from Clontech Laboratories, Inc (Palo Alto, CA, USA).

Transfection HEK293T cells were transiently transfected with the dextran method. In short, cells were grown to about 50% confluency and incubated for 2 h with the transfection mixture. The transfection mixture was made in warm culture media containing DEAE– dextran (Pharmacia, Piscataway, NJ, USA; 0.25 mg/ml), chloroquine (Sigma, St. Louis, MO, USA; 0.1 mM) and 2 ␮g endotoxin-free plasmid per 35⫻10 mm dish. Culture media contained Dulbecco’s Modified Eagle Medium⫹L-glutamine, 100 units/ml penicillin, 100 ␮g/ml streptomycin, and 10% heat-inactivated fetal bovine serum (Gibco BRL, Carlsbad, CA, USA). After transfection, cells were treated with 10% DMSO in PBS (in mM): 10 Na2HPO4, 2.3 NaH2PO4, 138 NaCl, and 2.7 KCl, pH 7.4 for 1 min. Cells were centrifuged, washed twice with culture media and plated. The day before experiments, cells were mechanically dissociated and transferred to chamber slides (Laboratory-Tek, Naperville, IL, USA) precoated with 0.1 mg/ml poly-D-lysine and 0.1 mg/ml collagen.

Immunocytochemistry hSlo endocytosis. As stated before, hSlo was tagged with the c-Myc epitope (AEEQKLISEEDL) at the N-terminus. This site was engineered in frame hSlo sequence to measure its extracellular

expression and endocytosis using anti-c-Myc antibodies. hSlo endocytosis was followed by labeling live cells with anti-c-Myc monoclonal antibody conjugated to FITC (anti-c-Myc-FITC). Cells transfected with hSlo with or without ␤1 subunit were cultured for 4 days, and incubated on ice for about 20 min prior to labeling. Labeling was performed on ice for 1 h with anti-c-Myc-FITC antibody. Cells were washed with cold media and endocytosis was stimulated by incubating cells at 37 °C for varying times. The best results were obtained with 0 and 120 min incubation times. After incubation at 37 °C, cells were immediately fixed with fixation solution, and permeabilized (see below) for double labeling with polyclonal antibody against ␤1 subunit. The fixation solution contained 4% paraformaldehyde (Fisher Scientific, Hampton, NH, USA), 14% picric acid (Sigma), 0.1 M phosphate buffer (Sigma) and was kept at 4 °C before use. hSlo surface labeling (non-permeabilized). Four days after transfection, live cells were incubated for 1 h at 4 °C with anti-cMyc monoclonal antibody (clone 9E10, Oncogene; 2 ␮g/ml). Cells were washed and fixed for 30 min with fixation solution. Cells were washed with PBS, permeabilized and double labeled with polyclonal antibodies (see below). Permeabilized labeling. Cells were permeabilized using solutions containing 0.2% Triton X-100 (Sigma). Nonspecific binding was blocked using 5% donkey serum in PBS for 30 min at room temperature (RT). Primary antibodies were added and incubated overnight at 4 °C. In double labeling cases, cells were washed and incubated for 1 h with Rhodamine RedTM-X-conjugated affinitypurified donkey anti-rabbit and FITC-conjugated affinity-purified donkey anti-mouse antibody (7.5 ␮g/ml) (Jackson ImmunoResearch Laboratories). Triple-labeling experiments used as secondary antibody (for anti-c-Myc primary antibody) Cy5-conjugated affinity-purified donkey anti-mouse antibody (6 ␮g/ml) (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) (see Fig. 2E) and RhoB GTPase-GFP fusion was directly detected by GFP fluorescence. Cells were washed and mounted using ProLong antifade (Molecular Probes, Eugene, OR, USA). Image analysis. Stacks of confocal images were acquired every 0.25 ␮m in the z axis with a confocal microscope (Olympus, Tokyo, Japan). “Total” protein expression was measured by outlining double-labeled cells (see Fig. 3) using the Image-Pro Plus (Media Cybernetics, Silver Spring, MD, USA) or ImageJ (NIH, Public Domain) under permeabilized conditions and measuring the mean intensity values (see Fig. 3, red bars). “Surface” expression was then evaluated by superimposing the same outlines used to measure the permeabilized cells on the same section labeled under non-permeabilized conditions and measuring the mean intensities (see Fig. 3, green bars). We quantified total (permeabilized) and surface (non-permeabilized) protein expression of each transfected cell in this manner and used the same approach to measure surface expression of hSlo and ␤1 or ␤1 mutants (see Fig. 4). All conditions including optical sectioning, number of sections and exposures were identical for hSlo and hSlo⫹␤1 or ␤1 mutants’ experiments. Quantified values are the mean⫾S.E. Colocalization images were drawn based on the strongest colocalized areas using the AutoVisualize module of AutoQuant (AutoQuant Imaging, Inc., Watervliet, NY, USA). Antibodies. Affinity-purified polyclonal anti-hSlo antibody was raised against a purified peptide corresponding to intracellular residues 883– 896 [VNDTNVQFLDQDDD] of hSlo and used at 1:500 dilution (Knaus et al., 1995; Song et al., 1999). The polyclonal antibody recognizing the human ␤1 subunit was raised against the extracellular residues 90 –103 [YHTEDTRDQNQQC] of KCNMB1 (Wanner et al., 1999) and used at 1/1000 dilution (Novus, Littleton, CO, USA). Anti-c-Myc antibody was a monoclonal antibody (clone 9E10, Oncogene) raised against the sequence AEEQKLISEEDLLRKRREQLKHKLEQLRNSCA and mouse antic-Myc-FITC conjugated antibody (Zymed Laboratory, South San

B. Toro et al. / Neuroscience 142 (2006) 661– 669 Francisco, CA, USA) was raised against the amino acid sequence EQKLISEEDL. The lowest concentration of antibody that gives the highest signal was used in these studies. Negative controls were performed by preadsorbing antibodies with the corresponding antigenic peptide and by mock transfections with no DNA.

RESULTS ␤1 Has a dominant punctated mode of expression Fig. 1 displays hSlo (c-Myc epitope, triangle) and ␤1 topologies, and HEK293T cells that were transfected with either hSlo (A), ␤1 (B, C) or both (D–I) subunits. Labeling was performed under permeabilized conditions using specific anti-c-Myc and anti-␤1 antibodies. When hSlo was expressed alone (A), cells displayed a fairly diffuse expression pattern that reaches the cell periphery (Zarei et al., 2001) (A, arrowhead). However, a clear punctated expression mode was evident when ␤1 subunit was expressed alone (B, C, arrows). Panels B and C display two typical expression patterns seen in ␤1 transfected cells. While both cells show a clear punctated expression pattern

663

(arrows), cell in panel C shows ␤1 expression that extends all the way to cell edges (arrowheads) confirming the ability of ␤1 subunit to reach the surface membrane (Meera et al., 2001). Strikingly, ␤1 co-transfection induced a large proportion of hSlo to appear clearly punctated (D, green) exhibiting a ␤1-like expression pattern (E, red) and a high degree of colocalization (F, yellow). Under closer inspection (box areas) the majority of hSlo puncta are colocalized with ␤1 (I, yellow). In addition, one can observe expression of the two subunits at the cell edges (D, E, arrows). The dominant effect of ␤1 punctated expression over hSlo led us to test the hypothesis that ␤1 could regulate hSlo surface expression by bringing hSlo to intracellular vesicles via an internalization mechanism. hSlo and ␤1 are co-targeted to an endosomal compartment hSlo endocytosis aided by coexpression with ␤1 subunit was directly visualized by live-labeling the extracellular

Fig. 1. ␤1 Punctated expression is dominant on hSlo pattern. HEK293T cells were transfected with hSlo (A), ␤1 (B, C) or hSlo⫹␤1 (D, E) subunits of MaxiK channel and labeled with monoclonal anti c-Myc (green) and polyclonal anti-␤1 antibodies (red) under permeabilized conditions. (A) hSlo has a diffuse expression pattern that extends to the cell periphery (arrow). (B) In contrast, ␤1 has a strong punctated expression pattern (arrow). In addition, some cells (C) appear to have both punctated (arrow) and cell periphery expression patterns (arrowhead). (D–E) When cells were transfected with hSlo⫹␤1, hSlo expression pattern (green) changed to resemble the ␤1 expression pattern (red). Under higher magnification (E), the punctated pattern appears like vesicles loaded with ␤1 subunits and almost all the hSlo containing vesicles also contain ␤1 subunit (yellow). Images are representative of 26 (A–C) or nine (D–E) different cells from three separate experiments. Images are overlaps of 14 (A–C) or 21 (D–E) confocal sections. Scale bars⫽10 ␮m.

664

B. Toro et al. / Neuroscience 142 (2006) 661– 669

c-Myc epitope engineered at hSlo amino terminus with anti-c-Myc-FITC in cells co-expressing both subunits. Labeling of live cells was performed at 0 °C to prevent potential endocytosis. After a period of 1 h on ice, cells were immediately fixed (Fig. 2A, 0 min) or the temperature was raised to 37 °C (Fig. 2B, 120 min) to favor endocytosis, which was stopped by fixation. Fixed cells were permeabilized and double labeled to detect ␤1 subunit as well. The results demonstrate that incubation at 0 °C (time zero) favored labeling of hSlo and ␤1 subunits at the cell periphery (Fig. 2A) with practically no detectable hSlo internalization. However, after 120 min at 37 °C, hSlo as well as ␤1 expression patterns were drastically changed from mostly surface labeling to intracellular puncta (Fig. 2B) with internalized hSlo largely colocalizing with ␤1 (Fig. 2B, right panel). hSlo alone was also internalized when incubated at 37 °C for 2 h though to a much lesser degree (Fig. 2C, arrowhead) with a considerable portion of hSlo remaining on the cell surface. Thus, ␤1 appears to be a more effective catalyst than temperature for hSlo internalization process. We next examined whether ␤1 and hSlo vesicles could colocalize within endosomal compartments. To this end, we utilized a well-characterized endosomal marker, RhoB GTPase. RhoB GTPase is targeted to a prelysosomal compartment that is involved in the sorting of internalized receptors for degradation (Ellis and Mellor, 2000) and is widely used as a specific endosomal marker (Mellor et al., 1998; Singh et al., 2004; Ellis and Mellor, 2000; Romsicki et al., 2004). We first cotransfected ␤1 with RhoB GTPase fused to green fluorescent protein (GFP) (Fig. 2D, RhoB⫹␤1) and found that RhoB showed a strong colocalization with ␤1 containing vesicles (colocalization panel) indicating that ␤1 is indeed targeted to endosomal compartments. We then expressed the three proteins, ␤1, RhoB and hSlo and examined their localization. Fig. 2E demonstrates that ␤1 is able to promote hSlo targeting to endosomes as a significant colocalization among, RhoB, ␤1 and hSlo was observed (colocalization panel, white).

␤1 Reduces hSlo surface expression Enhanced hSlo internalization by ␤1 expression in Fig. 2 already shows a downregulatory effect on hSlo surface expression by its ␤1 subunit. Fig. 3 shows a relative quantification of this phenomenon. Total and surface expression of hSlo were measured in cells expressing hSlo alone (Fig. 3A, B) or coexpressing hSlo⫹␤1 subunit (Fig. 3C, D). Panels A, B are confocal images at the middle of cells expressing only hSlo labeled either after (A, Total) or before (B, Surface) permeabilization. Surface (live) labeling was with anti-c-Myc antibody and total labeling was with an antibody raised against an intracellular epitope of hSlo. After permeabilization (A) cells show a strong intracellular labeling likely due to normal protein trafficking (Zarei et al., 2001). The same confocal sections but before permeabilization (B) clearly show hSlo labeling delineating the cell borders or cell surface. However, when cells coexpressing hSlo⫹␤1 were compared under identical conditions, there was a clear reduction in hSlo surface expression by ␤1 (D

vs. B) while total hSlo expression changes were difficult to discern by eye (C vs. A). Relative intensity quantification (E) normalized to hSlo values (see Experimental Procedures) demonstrated a slight non-significant reduction in the total hSlo protein in the presence of ␤1, probably due to competition of hSlo and ␤1 cDNA/mRNA’s for transcription/translation machineries (red bars). Yet, the latter explanation cannot account for the significant ␤1 subunitinduced decrease in hSlo surface expression (green bars). Modulatory trafficking sequences in ␤1 subunit control hSlo protein at the cell surface Sequence analysis of ␤1 subunit highlighted two overlapping putative endocytic signals (Fig. 4A, underlined sequences) located at its intracellular C-terminus: a tyrosinebased (YLSI) and a “dileucine”-like motif with one leucine replaced by isoleucine (IL). Thus, we went ahead and mutagenized these residues to test their potential role in downregulating hSlo surface expression. The sites selected for mutagenesis (Fig. 4A, asterisks) were replaced with alanine (A). This amino acid was chosen because it is hydrophobic like tyrosine (Y) and leucine (L), and because it has been effective in dissecting internalization signals present in other ion channels (Hu et al., 2001). Analysis was performed in HEK293T cells transfected with hSlo plus ␤1 as control ((Fig. 4B), or with hSlo and either of the two ␤1 mutants, I186A, L187A (Fig. 4C) or Y183A, I186A (Fig. 4D). In every experiment, cells were labeled under non-permeabilized conditions to measure surface expressions of hSlo, ␤1 and ␤1 mutants. Our results show that surface expression of hSlo was increased significantly when hSlo was coexpressed with I186A, L187A double mutant (C, green; E, green bar 2) suggesting a decrease in hSlo endocytosis. We observed a slight increase (⬇10%) with Y183A, I186A double mutant but overall this mutant was less effective in increasing surface expression of hSlo (D, green; E, green bar 3). In addition, surface expression of ␤1 and its mutants (E, red bars 1–3) follow the same expression trend as hSlo (E, green bars 1–3) with I186A, L187A mutant showing highest (red bar 2) and Y183A, I186A mutant less pronounced (red bar 3) increase in surface expressions. Taken together, the results indicate that disruption of the “dileucine”-like signal stabilizes ␤1 subunit on the plasma membrane that consequently has a stabilizing effect on hSlo surface expression levels. Mutagenesis does not interfere with normal trafficking of ␤1 mutants We next examined whether the above mutations affected the intracellular trafficking of ␤1 subunit. HEK293T cells were transfected with ␤1 or each of the two ␤1 mutants (I186A, L187A and Y183A, I186A) and their subcellular distribution patterns were assessed by labeling cells under permeabilized conditions. Fig. 5 shows examples of the total expression pattern of ␤1 and ␤1 mutants at low (A–C) and higher magnification (D–F, zoomed squares in A–C). While ␤1 expression appears to be more punctated (A), overall they have similar labeling patterns. However, a closer view (D–F) revealed that mutations reduced but did

B. Toro et al. / Neuroscience 142 (2006) 661– 669

665

Fig. 2. hSlo and ␤1 are internalized together and targeted to an endosomal compartment. HEK293T cells were co-transfected with hSlo and ␤1 subunits and the extracellular domain of hSlo was live-labeled with anti-c-Myc-FITC. ␤1 subunit was labeled after fixation and permeabilization (see Experimental Procedures). Cells were immediately fixed (A, 0 min) or incubated at 37 °C to stimulate endocytosis (B, 120 min). (A) At time zero, live-labeling of hSlo (left panel) reveals uniform surface expression around cells that colocalizes with ␤1 (middle and right panels) labeling at the cell periphery. (B) After 120 min incubation at 37 °C, hSlo label at the surface gets internalized showing clear puncta (left panel) with practically no visible surface expression. Strikingly, internalized hSlo colocalizes with ␤1 subunit as demonstrated in the “colocalization” panel (right panel). (C) Cells that were transfected only with hSlo show cell surface (arrow) and modest punctated expression (arrowhead). (D) RhoB GTPase-GFP (RhoB) and ␤1 coexpression shows a strong colocalization of ␤1 puncta with RhoB (endosomal marker). (E) RhoB, ␤1, and hSlo coexpression demonstrates “colocalization” of the three proteins. All images are single confocal sections acquired from the middle of the cells. Images are representative of nine (A), seven (B), 12 (C), 13 (D) and 16 (E) cells examined under high magnification. Scale bars⫽10 ␮m.

666

B. Toro et al. / Neuroscience 142 (2006) 661– 669

Fig. 3. ␤1 reduces hSlo on the cell surface. Total and surface expression of hSlo was studied in the absence (A, B) or presence of ␤1 subunit (C, D). (A, B) Cells transfected with hSlo and labeled after (A) or before (B) permeabilization. (C, D) Cells transfected with hSlo⫹␤1 and labeled after (C) or before (D) permeabilization. (E) Mean values of % pixel intensity after (red bars) or before (green bars) permeabilization. Surface expression was significantly reduced (green bars). A small reduction in total hSlo protein was observed in the presence of ␤1 subunit that can be attributed to ␤1 competition for resources (red bars). For a given set of experiments, images were acquired in the same day with identical exposures and normalized to hSlo values (see Experimental Procedures). Images are quantified from 74 (A, B) and 55 (C, D) cells from three separate experiments. Images are single confocal sections that were acquired at the middle of cells. Scale bar⫽10 ␮m.

not eliminate internalization. This observation poses the idea that ␤1 subunit may contain additional endocytic signals with unique sequences.

Fig. 4. ␤1 subunit contains functional endocytic trafficking signals. (A) Schematic representation of the ␤1 modulatory subunit of hSlo. The arrow marks two potential internalization signals in the C-terminus (underlined amino acids). The asterisks mark amino acids of consensus endocytic signals that were mutated to alanine. HEK293T cells were transfected with hSlo⫹␤1 (B) or with either of the two ␤1 double mutant constructs (C, D) and labeled under non-permeabilized condition. Thus, surface expression of both hSlo (green) and ␤1 (red) was measured and quantified. (E) Quantification of the % pixel intensity measurements from three separate experiments from three different transfections is shown. The results were normalized to the hSlo⫹␤1 intensity values (bar 1). Double mutants I186A, L187A significantly increased both the hSlo and ␤1 surface expressions (bar 2, green, red) while Y183A, I186A showed slight increase in the hSlo and ␤1 surface expressions (bar 3, green, red). Bars 1–3 correspond to panels B–D, respectively. For a given set of experiments (B–D), images were acquired in the same day with identical exposures and normalized to hSlo⫹␤1 values. Images are single confocal sections that were acquired at the middle of cells. Images (B–D) were quantified from 42 (B), 49 (C), and 60 (D) cells from three separate experiments. Scale bar⫽10 ␮m.

B. Toro et al. / Neuroscience 142 (2006) 661– 669

Fig. 5. ␤1 mutants have normal intracellular trafficking pattern. ␤1 or ␤1 mutant-transfected cells were labeled with anti-␤1 antibody after permeabilization. All mutants show normal protein trafficking pattern and I186A, L187A mutants tend to have less punctae compared with ␤1 or Y183A, I186A mutants. Images (A–F) are overlaps of 22 confocal sections and are representative of 35 cells examined under high magnifications. Scale bars⫽10 ␮m.

DISCUSSION Previously, we isolated a splice variant of MaxiK channel (SV1) that contains a retention/retrieval trafficking signal. Our findings demonstrated that SV1 trafficking signal was able to trap MaxiK ␣ subunit in the ER; thus, preventing forward movement of MaxiK channels from the ER to the plasma membrane (Zarei et al., 2001, 2004). In this study, we found that coexpression of a transmembrane ␤ subunit (␤1) with the pore-forming ␣ subunit (hSlo) of MaxiK channels can also regulate hSlo surface expression levels. This expression regulation appears to be originated from endocytic signals located proximal to the C terminus of the ␤1 subunit. In particular, site-directed mutagenesis within a dileucine-like signal significantly increased steady-state surface expression of hSlo (Fig. 4E, green bars 1, 2) that practically reversed the ␤1 effect on hSlo cell surface expression (Fig. 3, green bars). Furthermore, the two ␤1 mutants were able to reach the plasma membrane together with hSlo, and their intracellular distribution was quite similar to wild type except for an apparent decrease in ␤1-labeled vesicles indicating that mutations did not interfere with their normal forward trafficking. Taken together, retention/retrieval and endocytic signals collectively could fine tune the number of MaxiK channels expressed on the cell surface at anytime.

␤1 Subunit contains both tyrosine- and dileucinebased signals Most of tyrosine-based and dileucine-based sorting signals that are common in the large cytoplasmic domains of many proteins are inactive. These sorting signals are folded within the structure of the proteins; thus, are not accessible for interactions with components of the sorting machinery (review (Bonifacino and Traub, 2003)). In contrast, exposed sorting signals can interact with clathrin, adaptor protein (AP) and other accessory components of clathrin

667

coats (Pearse and Robinson, 1990). This interaction is significantly influenced by the relative position of sorting signals to transmembrane domains and the C-terminus. For instance, YXXØ signal behaves most often as purely endocytic sorting signal if it is situated 10 – 40 residues from the transmembrane domain. In contrast, the same sequence located ⬇six to nine residues from the transmembrane domain and in close proximity to the C-terminus can target protein to lysosomes (review (Bonifacino and Traub, 2003)). Strikingly, the ␤1 subunit of MaxiK channel contains two overlapping tyrosine- and dileucine-based sorting signals that are located about five to eight residues from the second transmembrane domain and in very close proximity to the C-terminal end. The C-terminus of the ␤1 subunit is quite short (⬇13 amino acids) that could provide sorting signals easy access to adaptor proteins. Thus, the ␤1 sorting signals are ideally located to act as endocytic as well as lysosomal-targeting signals. On the other hand, the ␣ subunit contains a potential endocytic signal that is 14 residues from the S6 transmembrane domain, and in the intracellular C-terminus that is ⬇790 amino acids long. These characteristics likely reduce the intrinsic internalization capability; thus, stabilizing the ␣ subunit on the cell surface. In this scheme, the ␤1 subunit would behave as a destabilizing factor directing the ␣ subunit to lysosomes and its subsequent catabolism. Although our data have been based on the ␤1 subunit, other MaxiK modulatory ␤ subunits may participate in regulating the ␣ subunit expression levels as well. For instance, sequence alignment shows that the two overlapping sorting signals are also conserved in the ␤2 subunit of MaxiK channel. ␤2 Subunit also shows a strong colocalization with RhoB GTPase (endosomal marker) that further supports our data (our unpublished observations). Another modulatory ␤ subunit of the MaxiK (␤3) (Brenner et al., 2000a) also contains two endocytic sorting signals at its C-terminus that unlike ␤1 and ␤2 are not overlapped. Finally, the ␤4 subunit of the MaxiK channel (Brenner et al., 2000a; Meera et al., 2000) has no endocytic signal at its C-terminus. These modulatory ␤ subunits are differentially distributed in the brain and other tissues (Weiger et al., 2000); thus, they could differentially regulate the amount of hSlo protein on the cell surface. It is interesting to note that in reduction of ␤1 in models of hypertension including that generated by ␤1 gene ablation there was no detectable increase in the number of active MaxiK channels in isolated cell membrane patches (Brenner et al., 2000b; Amberg et al., 2003; Amberg and Santana, 2003; Petkov et al., 2001) as would be predicted from our results. Although a precise mechanism to explain these intriguing findings is currently lacking, one could speculate that pathological conditions may trigger compensatory mechanisms that overcome the disrupted ␤1-mediated endocytosis or that cell-type specific mechanisms contribute to the role of ␤1 as an endocytic regulatory mechanism of ␣ subunit expression. Also, how ␤2, ␤3 and ␤4 modulatory subunits actually affect the ␣ subunit expression levels remains open to question.

668

B. Toro et al. / Neuroscience 142 (2006) 661– 669

Internalization as a mechanism to control surface expressions of ion channels Endocytic mechanisms are used by a number of ion channels as a means to regulate their surface expression. For instance, the NMDA receptor is a multimeric complex made up of homologous subunits (NR1, NR2A-D) (Hollmann and Heinemann, 1994). Of all of the NR1 and NR2 cytosolic domains, NR2B contains the strongest YXXØ and LL consensus internalization motifs (Ohno et al., 1998). These internalization motifs are located in close proximity to the C-terminal end of the NR2B subunit. Coexpression of the NR2B with pore forming NR1 subunit can significantly enhance receptor internalization (Roche et al., 2001) that resembles the ␤1 effect on the MaxiK ␣ subunit. CFTR chloride channel is another ion channel that regulates its surface expression by multiple endocytic sorting signals located in the C-terminal tail of the channel (Hu et al., 2001). In addition to endocytic signals, direct phosphorylation of ion channel by kinases has also been identified as a mechanism to control channel expression by endocytosis. The delayed rectifier potassium channel Kv1.2 is the first voltage-gated ion channel shown to be downregulated by tyrosine kinases (Huang et al., 1993). Although the precise mechanisms are not clear, it is believed that phosphorylation induces changes in channel structure that enhance the endocytosis of Kv1.2 from the cell surface (Nesti et al., 2004).

CONCLUSION In summary, we found an additional role for the modulatory ␤1 subunit of the MaxiK channel that is to enhance the internalization of the ␣ subunit. Thus, ␤1 may contribute to differential surface expression of MaxiK channels that is necessary for the proper function of tissues where it is expressed. Further studies are needed to evaluate the role that ␤1 may play to regulate MaxiK expression levels under physiological or pathological conditions. Acknowledgments—This work was supported by AHA grant 0230225N, NIH grants GM068855 (M.Z.) and MBRS RISE GM065925 (R.W.), HD046510 (E.S.) and HL54970 (L.T.).

REFERENCES Amberg GC, Bonev AD, Rossow CF, Nelson MT, Santana LF (2003) Modulation of the molecular composition of large conductance, Ca(2⫹) activated K(⫹) channels in vascular smooth muscle during hypertension. J Clin Invest 112:717–724. Amberg GC, Santana LF (2003) Downregulation of the BK channel beta1 subunit in genetic hypertension. Circ Res 93:965–971. Bonifacino JS, Traub LM (2003) Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu Rev Biochem 72:395– 447. Bravo-Zehnder M, Orio P, Norambuena A, Wallner M, Meera P, Toro L, Latorre R, Gonzalez A (2000) Apical sorting of a voltage- and Ca2⫹-activated K⫹ channel alpha-subunit in Madin-Darby canine kidney cells is independent of N-glycosylation. Proc Natl Acad Sci U S A 97:13114 –13119. Bremnes B, Madsen T, Gedde-Dahl M, Bakke O (1994) An LI and ML motif in the cytoplasmic tail of the MHC-associated invariant chain mediate rapid internalization. J Cell Sci 107 (Pt 7):2021–2032.

Brenner R, Jegla TJ, Wickenden A, Liu Y, Aldrich RW (2000a) Cloning and functional characterization of novel large conductance calcium-activated potassium channel beta subunits, hKCNMB3 and hKCNMB4. J Biol Chem 275:6453– 6461. Brenner R, Perez GJ, Bonev AD, Eckman DM, Kosek JC, Wiler SW, Patterson AJ, Nelson MT, Aldrich RW (2000b) Vasoregulation by the beta1 subunit of the calcium-activated potassium channel. Nature 407:870 – 876. Ellis S, Mellor H (2000) Regulation of endocytic traffic by rho family GTPases. Trends Cell Biol 10:85– 88. Gough NR, Zweifel ME, Martinez-Augustin O, Aguilar RC, Bonifacino JS, Fambrough DM (1999) Utilization of the indirect lysosome targeting pathway by lysosome-associated membrane proteins (LAMPs) is influenced largely by the C-terminal residue of their GYXXphi targeting signals. J Cell Sci 112 (Pt 23):4257– 4269. Hollmann M, Heinemann S (1994) Cloned glutamate receptors. Annu Rev Neurosci 17:31–108. Hu W, Howard M, Lukacs GL (2001) Multiple endocytic signals in the C-terminal tail of the cystic fibrosis transmembrane conductance regulator. Biochem J 354:561–572. Huang X-Y, Morielli AD, Peralta EG (1993) Tyrosine kinase-dependent suppression of a potassium channel by the G protein-coupled m1 muscarinic acetylcholine receptor. Cell 75:1145–1156. Knaus HG, Eberhart A, Koch ROA, Munujos P, Schmalhofer WA, Warmke JW, Kaczorowski GJ, Garcia ML (1995) Characterization of tissue-expressed alpha-subunits of the high-conductance Ca2⫹-activated K⫹ channel. J Biol Chem 270:22434 –22439. Knaus HG, Schwarzer C, Koch RO, Eberhart A, Kaczorowski GJ, Glossmann H, Wunder F, Pongs O, Garcia ML, Sperk G (1996) Distribution of high-conductance Ca2⫹-activated K⫹ channels in rat brain: targeting to axons and nerve terminals. J Neurosci 16:955–963. Ktistakis NT, Thomas D, Roth MG (1990) Characteristics of the tyrosine recognition signal for internalization of transmembrane surface glycoproteins. J Cell Biol 111:1393–1407. Letourneur F, Klausner RD (1992) A novel di-leucine motif and a tyrosine-based motif independently mediate lysosomal targeting and endocytosis of CD3 chains. Cell 69:1143–1157. Lu R, Alioua A, Kumar Y, Eghbali M, Stefani E, Toro L (2006) MaxiK channel partners: physiological impact. J Physiol 570:65–72. McManus OB, Helms LM, Pallanck L, Ganetzky B, Swanson R, Leonard RJ (1995) Functional role of the beta subunit of high conductance calcium-activated potassium channels. Neuron 14:645– 650. Meera P, Wallner M, Jiang Z, Toro L (1996) A calcium switch for the functional coupling between ␣ (hslo) and ␤ subunits (KV,Ca␤) of maxi K channels. FEBS Lett 382:84 – 88. Meera P, Wallner M, Song M, Toro L (1997) Large conductance voltage- and calcium-dependent K⫹ channel, a distinct member of voltage-dependent ion channels with seven N-terminal transmembrane segments (S0-S6), an extracellular N terminus, and an intracellular (S9-S10) C terminus. Proc Natl Acad Sci U S A 94:14066 –14071. Meera P, Wallner M, Toro L (2000) A neuronal beta subunit (KCNMB4) makes the large conductance, voltage- and Ca2⫹-activated K⫹ channel resistant to charybdotoxin and iberiotoxin. Proc Natl Acad Sci U S A 97:5562–5567. Meera P, Wallner M, Toro L (2001) Molecular biology of high-conductance, Ca2⫹-activated K⫹ channels. In: Potassium channels in cardiovascular biology (Archer SL, Rusch NJ, eds), pp 49 –70. New York: Plenum Press. Mellor H, Flynn P, Nobes CD, Hall A, Parker PJ (1998) PRK1 is targeted to endosomes by the small GTPase, RhoB. J Biol Chem 273:4811– 4814. Nesti E, Everill B, Morielli AD (2004) Endocytosis as a mechanism for tyrosine kinase-dependent suppression of a voltage-gated potassium channel. Mol Biol Cell 15:4073– 4088. Nishimaru K, Eghbali M, Lu R, Marijic J, Stefani E, Toro L (2004) Functional and molecular evidence of MaxiK channel beta1 sub-

B. Toro et al. / Neuroscience 142 (2006) 661– 669 unit decrease with coronary artery ageing in the rat. J Physiol 559:849 – 862. Ohno H, Aguilar RC, Yeh D, Taura D, Saito T, Bonifacino JS (1998) The medium subunits of adaptor complexes recognize distinct but overlapping sets of tyrosine-based sorting signals. J Biol Chem 273:25915–25921. Pearse BM, Robinson MS (1990) Clathrin, adaptors, and sorting. Annu Rev Cell Biol 6:151–171. Petkov GV, Bonev AD, Heppner TJ, Brenner R, Aldrich RW, Nelson MT (2001) Beta1-subunit of the Ca2⫹-activated K⫹ channel regulates contractile activity of mouse urinary bladder smooth muscle. J Physiol 537:443– 452. Pluger S, Faulhaber J, Furstenau M, Lohn M, Waldschutz R, Gollasch M, Haller H, Luft FC, Ehmke H, Pongs O (2000) Mice with disrupted BK channel beta1 subunit gene feature abnormal Ca(2⫹) spark/ STOC coupling and elevated blood pressure. Circ Res 87: E53–E60. Roche KW, Standley S, McCallum J, Dune LC, Ehlers MD, Wenthold RJ (2001) Molecular determinants of NMDA receptor internalization. Nat Neurosci 4:794 – 802. Romsicki Y, Reece M, Gauthier JY, Sante-Appiah E, Kennedy BP (2004) Protein tyrosine phosphatase-1B dephosphorylation of the insulin receptor occurs in a perinuclear endosome compartment in human embryonic kidney 293 cells. J Biol Chem 279:12868 – 12875. Sausbier M, Hu H, Arntz C, Feil S, Kamm S, Adelsberger H, Sausbier U, Sailer CA, Feil R, Hofmann F, Korth M, Shipston MJ, Knaus HG, Wolfer DP, Pedroarena CM, Storm JF, Ruth P (2004) Cerebellar ataxia and Purkinje cell dysfunction caused by Ca2⫹-activated K⫹ channel deficiency. Proc Natl Acad Sci U S A 101:9474 –9478. Schmalhofer WA, Sanchez M, Dai G, Dewan A, Secades L, Hanner M, Knaus HG, McManus OB, Kohler M, Kaczorowski GJ, Garcia ML (2005) Role of the C-terminus of the high-conductance calciumactivated potassium channel in channel structure and function. Biochemistry 44:10135–10144. Singh SP, McDonald D, Hope TJ, Prabhakar BS (2004) Upon thyrotropin binding the thyrotropin receptor is internalized and localized to endosome. Endocrinology 145:1003–1010. Song M, Zhu N, Olcese R, Barila B, Toro L, Stefani E (1999) Hormonal control of protein expression and mRNA levels of the MaxiK channel alpha subunit in myometrium. FEBS Lett 460:427– 432.

669

Uebele VN, Lagrutta A, Wade T, Figueroa DJ, Liu Y, McKenna E, Austin CP, Bennett PB, Swanson R (2000) Cloning and functional expression of two families of beta-subunits of the large conductance calcium-activated K⫹ channel. J Biol Chem 275:23211– 23218. Wallner M, Meera P, Ottolia M, Kaczorowski G, Latorre R, Garcia ML, Stefani E, Toro L (1995) Characterization of and modulation by a ␤-subunit of a human maxi KCa channel cloned from myometrium. Receptors Channels 3:185–199. Wallner M, Meera P, Toro L (1999) Molecular basis of fast inactivation in voltage and Ca2⫹-activated K⫹ channels: A transmembrane ␤-subunit homolog. Proc Natl Acad Sci U S A 96:4137– 4142. Wang SX, Ikeda M, Guggino WB (2003) The cytoplasmic tail of large conductance, voltage- and Ca2⫹-activated K⫹ (MaxiK) channel is necessary for its cell surface expression. J Biol Chem 278: 2713–2722. Wanner SG, Koch RO, Koschak A, Trieb M, Garcia ML, Kaczorowski GJ, Knaus HG (1999) High-conductance calcium-activated potassium channels in rat brain: pharmacology, distribution, and subunit composition. Biochemistry 38:5392–5400. Weiger TM, Holmqvist MH, Levitan IB, Clark FT, Sprague S, Huang WJ, Ge P, Wang C, Lawson D, Jurman ME, Glucksmann MA, Silos-Santiago I, DiStefano PS, Curtis R (2000) A novel nervous system beta subunit that downregulates human large conductance calcium-dependent potassium channels. J Neurosci 20:3563– 3570. Williams MA, Fukuda M (1990) Accumulation of membrane glycoproteins in lysosomes requires a tyrosine residue at a particular position in the cytoplasmic tail. J Cell Biol 111:955–966. Xia XM, Ding JP, Lingle CJ (1999) Molecular basis for the inactivation of Ca2⫹- and voltage-dependent BK channels in adrenal chromaffin cells and rat insulinoma tumor cells. J Neurosci 19:5255–5264. Zarei MM, Eghbali M, Alioua A, Song M, Knaus HG, Stefani E, Toro L (2004) An endoplasmic reticulum trafficking signal prevents surface expression of a voltage- and Ca2⫹-activated K⫹ channel splice variant. Proc Natl Acad Sci U S A 101:10072–10077. Zarei MM, Zhu N, Alioua A, Eghbali M, Stefani E, Toro L (2001) A novel MaxiK splice variant exhibits dominant-negative properties for surface expression. J Biol Chem 276:16232–16239.

(Accepted 28 June 2006) (Available online 14 August 2006)