Forward Transport of K2P3.1: Mediation by 14-3-3 and COPI, Modulation by P11
Total Page:16
File Type:pdf, Size:1020Kb
# 2007 The Authors Journal compilation # 2007 Blackwell Publishing Ltd Traffic 2008; 9: 72–78 Blackwell Munksgaard doi: 10.1111/j.1600-0854.2007.00663.x Forward Transport of K2P3.1: Mediation by 14-3-3 and COPI, Modulation by p11 Ita O’Kelly1,* and Steve A. N. Goldstein2,* endoplasmic reticulum (ER) retrieval signal (KRQN) that binds b coat protein (COP) to inhibit forward transport (12). 1The Faculty of Life Sciences, The University of We demonstrated that phosphorylation-dependent bind- Manchester, 46 Grafton Street, Manchester, ing of 14-3-3b, a cytoplasmic protein found in all cells (13), M13 9NT, UK suppressed b-COP binding and enabled forward transport 2 Department of Pediatrics and Institute for Molecular of the channel to the surface (12). Thus, two K 3.1 chan- Pediatric Sciences, Pritzker School of Medicine, 2P nel populations were observed in vivo; one with b-COP and University of Chicago, 5721 S. Maryland Avenue, Chicago, IL 60637, USA no 14-3-3 and another with 14-3-3b and no b-COP. This *Corresponding author: Ita O’Kelly, forward transport mechanism also appears to be exploited ita.o’[email protected] or Steve A. N. Goldstein, by other proteins (12,14–16). [email protected] While we observed b-COP to bind on the K2P3.1 N-terminus and 14-3-3b to bind on the C-terminus, Girard Surface expression of the K2P3.1 two-pore domain potas- et al. found basic residues in the C-terminus that sup- sium channel is regulated by phosphorylation-dependent pressed channel expression and observed C-terminal bind- binding of 14-3-3, leading to suppression of coatomer coat protein I (COPI)-mediated retention in endoplasmic reticu- ing of the annexin protein p11 (S100A10) in association lum (ER). Here, we investigate the nature of the macromo- with forward transport (22). p11 has also been reported to lecular regulatory complexes that mediate forward and enhance plasma membrane expression of Nav1.8 sodium retrograde transport. We demonstrate that (i) the channel channels (17), ASIC1a acid-sensing ion channels, transient employs two separate but interacting COPI binding sites receptor potential cation channels (TRPV5 and 6) and se- on the N- and C-termini; (ii) disrupting COPI binding to rotonin 1B receptor (18–20). In contrast, others found p11 either site interferes with the ER retention; (iii) p11 and 14- to bind to the K2P3.1 C-terminus, leading to ER retention 3-3 do not interact on their own; (iv) p11 binding to the C- through basic sites in p11 (21). terminal retention motif is dependent on 14-3-3; and (v) p11 is coexpressed in only a subset of tissues with K2P3.1, while 14-3-3 expression is ubiquitous. We conclude that To examine these discordant observations regarding roles K2P3.1 forward transport requires 14-3-3 suppression of of 14-3-3, COPI and p11 in regulation of K2P3.1 surface COPI binding, whereas p11 serves a modulatory role. expression, we sought to further delineate mechanistic bases for their action. Here, we report that basic motifs Key words: 14-3-3, annexin, background, COPI retention, on both the N- and C-termini of K2P3.1 must be intact to leak, p11, potassium channel, resting potential, TASK-1 achieve ER retention. Both motifs can bind b-COP, and the affinity of interaction is significantly reduced if either is Received 26 June 2007, revised and accepted for publica- tion 28 September 2007, uncorrected manuscript pub- altered by mutation or 14-3-3b is bound to the channel lished online 1 October 2007, published online 24 October C-terminus. We find p11 in only some tissues that express 2007 K2P3.1 (it is notably absent from heart) and show that p11 and 14-3-3 do not interact in the absence of the channel; rather, p11 binds to K2P3.1 only after phosphorylation- dependent binding of 14-3-3. K2P3.1 channels pass background potassium currents that stabilize cells at negative resting potentials and return cells to baseline after excitatory events. The channels show These observations argue that 14-3-3 is required to sup- widespread tissue distribution, have a variety of natural press COPI-dependent retention of K2P3.1 in ER and to and medicinal regulators and have putative roles in re- enable p11 modulation of channel forward transport in a sponses to hormones, neurotransmitters, acidosis, oxy- tissue-specific fashion. gen tension and local and volatile anesthetics (1–7). Quality control mechanisms that ensure its controlled expression Results on the plasma membrane are considered here. K2P3.1 has COPI binding motifs on both termini K2P channel subunits have cytoplasmic N- and C-termini, The N-terminal K2P3.1 sequence MKRQN (but not the four transmembrane domains and two pore-forming mutant sequence MNQQN) plays a critical role in reten- P loops, and two subunits appear both necessary and tion of the channel in the ER through COPI-mediated sufficient to form an ion conduction pore (8–11). In 2002, retrograde transport because of binding of the coatomer we observed that the K2P3.1 N-terminus has a dibasic subunit b-COP (12). Girard et al. found K2P3.1 to be 72 www.traffic.dk K2P3.1 Forward Transport: 14-3-3, p11 and COPI Retention released from ER retention on mutation of residues in the 14-3-3 disrupts COPI binding on the C- but not the C-terminus (e.g. K389RRSSV to K389A or R390A or R391A) N-terminus (22). Altering these residues might suppress COPI binding As 14-3-3 binding suppresses b-COP binding to intact to the C-terminus and/or has the potential to suppress K2P3.1 channels (12), we sought to study if 14-3-3 and phosphorylation of S393, which is required for 14-3-3 COPI competed for binding to the isolated N- or C-terminal binding and ablation of COPI-dependent ER retention motifs. Semi-quantitative analysis showed comparable through the N-terminus (12). We sought, therefore, to in- amounts of b-COP bound to N- and C-terminal peptides vestigate if b-COP interacts with the K2P3.1 C-terminus. As with a lower level isolated when the C-terminal peptide was observed previously, peptides analogous to the N-terminus phosphorylated (Figure 1C). We hypothesized that the phos- bound b-COP when the dibasic motif was intact, similarly. phopeptide bound 14-3-3 naturally present in the COS-7 Indeed, peptides identical to the C-terminal 16 residues of cells interfered with b-COP association. To test this hypoth- K2P3.1 (terminating -MKRRSSV) bound native b-COP from esis, the peptides were preincubated with recombinant COS-7 cells, whereas peptides with the basic residues 14-3-3b. Again, the N- and C-terminal peptides bound com- altered (-MNQQSSV) failed to bind (Figure 1A). These data parable amounts of b-COP, whereas binding to phosphor- suggest that either the N- or C-terminal site alone might ylated C-terminal peptide was suppressed to undetectable yield channel retention through the COPI pathway, how- levels (Figure 1B,C). This demonstrated a direct role for ever, this is not the case; both sites must be intact to 14-3-3 on the C-terminal motif not seen with the N-terminal achieve ER retention (12,22). sequence. Figure 1: b-COP interacts with both K2P3.1 N- and C-terminal motifs. A) Western blot analysis of native b-COP isolated from COS-7 cells on immobilized peptides analogous to the N- (MKRQNVA-) and C-termini (-MKRRSSV) of K2P3.1. Mutation of basic sites to MNQQNVA- and -MNQQSSV results in failure to bind b-COP. Wild-type C-terminal peptide with phosphate (-MKRRSSPV) binds b-COP less well because of binding of native 14-3-3 (12). B) Binding of recombinant 14-3-3b does not alter b-COP interaction with N-terminal or non- P phosphorylated C-terminal peptides but ablates binding to the phosphorylated K2P3.1 C-terminal peptide (-MKRRSS V). C) Bar chart shows semi-quantitative analysis of b-COP binding to the indicated peptides before and after incubation with recombinant 14-3-3b through computer-assisted densitometric scanning (IMAGE J software). Binding is presented as percentage of b-COP input based on total lysate lane. D) HA-tagged wild-type (WT) and mutant K2P3.1 channels expressed in COS-7 cells purified by immunoprecipitation with HA antibodies and analyzed by Western blot with antibodies to K2P3.1 (top) and b-COP (bottom) show WT channels affinity purified with b-COP, while channels lacking either N- or C-terminal basic motifs fail to copurify b-COP. Ab, antibody; Ha, hemagglutinin epitope tag. Traffic 2008; 9: 72–78 73 O’Kelly and Goldstein N- and C-termini dibasic motifs are required in and inhibition by external acidification (Figure 2B) as found 395 intact channels for NQ-K2P3.1DV channels before suggesting that traf- To confirm b-COP interaction with intact channels was ficking motifs do not alter biophysical function (12). mediated through both the N- and C-termini, hemaggluti- nin epitope-tagged K2P3.1 channels (HA-K2P3.1) were ex- p11 binds to the C-terminus of K2P3.1 through 14-3-3 pressed in COS-7 cells and their association with native Both p11 and 14-3-3 have been implicated in enabling b-COP evaluated. Figure 1D shows affinity purification forward transport of K2P3.1 through the same C-terminal through the tag with wild-type (WT) and mutant K2P3.1 residues (22). We sought to define the basis for action of channels. Wild-type channels were purified in association p11 by testing the hypothesis that it binds to the channel in with b-COP.