Search

PICK1 (protein interacting with C kinase 1) contains an N-terminal protein binding PDZ domain and a C-terminal lipid binding BAR domain. PICK1 plays a key role in several physiological processes, including synaptic plasticity. However, little is known about the cellular mechanisms governing the activity of PICK1 itself. Here we show that PICK1 is a substrate in vitro both for PKCα (protein kinase Cα), as previously shown, and for CaMKIIα (Ca(2+)-calmodulin-dependent protein kinase IIα). By mutation of predicted phosphorylation sites, we identify Ser77 in the PDZ domain as a major phosphorylation site for PKCα. Mutation of Ser77 reduced the level of PKCα-mediated phosphorylation ~50%, whereas no reduction was observed upon mutation of seven other predicted sites. Addition of lipid vesicles increased the level of phosphorylation of Ser77 10-fold, indicating that lipid binding is critical for optimal phosphorylation. Binding of PKCα to the PICK1 PDZ domain was not required for phosphorylation, but a PDZ domain peptide ligand reduced the overall level of phosphorylation ~30%. The phosphomimic S77D reduced the extent of cytosolic clustering of eYFP-PICK1 in COS7 cells and thereby conceivably its lipid binding and/or polymerization capacity. We propose that PICK1 is phosphorylated at Ser77 by PKCα preferentially when bound to membrane vesicles and that this phosphorylation in turn modulates its cellular distribution.

To investigate microdomain association of the dopamine transporter (DAT), we employed FCS (fluorescence correlation spectroscopy) and FRAP (fluorescence recovery after photobleaching). In non-neuronal cells (HEK293), FCS measurements revealed for the YFP-DAT (DAT tagged with yellow fluorescent protein) a diffusion coefficient (D) of approximately 3.6 x 10(-9) cm2/s, consistent with a relatively freely diffusible protein. In neuronally derived cells (N2a), we were unable to perform FCS measurements on plasma membrane-associated protein due to photobleaching, suggesting partial immobilization. This was supported by FRAP measurements that revealed a lower D and a mobile fraction of the YFP-DAT in N2a cells compared to HEK293 cells. Comparison with the EGFP-EGFR (epidermal growth factor receptor) and the EGFP-beta2AR (beta2 adrenergic receptor) demonstrated that this observation was DAT specific. Both the cytoskeleton-disrupting agent cytochalasin D and the cholesterol-depleting agent methyl-beta-cyclodextrin (mbetaCD) increased the lateral mobility of the YFP-DAT but not that of the EGFP-EGFR. The DAT associated in part with membrane raft markers both in the N2a cells and in rat striatal synaptosomes as assessed by sucrose density gradient centrifugation. Raft association was further confirmed in the N2a cells by cholera toxin B patching. It was, moreover, observed that cholesterol depletion, and thereby membrane raft disruption, decreased both the Vmax and KM values for [3H]dopamine uptake without altering DAT surface expression. In summary, we propose that association of the DAT with lipid microdomains in the plasma membrane and/or the cytoskeleton serves to regulate both the lateral mobility of the transporter and its transport capacity.

Treatment of the beta 2 adrenergic receptor with the reducing agent dithiothreitol (DTT) is known to abolish ligand binding to the receptor. Interestingly, the loss of binding can be prevented by preoccupation of the receptor with ligand. It is unclear, however, whether the ligand blocks access of DTT to the receptor, or the ligand stabilizes the receptor structure. In the present study, we have utilized circular dichroism (CD) and intrinsic tryptophan fluorescence to directly probe structural changes in the beta 2 adrenergic receptor in response to DTT treatment. Analysis of CD spectra of purified beta 2 receptor in the detergent micelle indicated that the receptor has an alpha-helix content of 60%, which is substantially more than what would be attributed to the seven transmembrane domains. The alpha-helix content was unchanged in the presence of DTT, suggesting that DTT treatment does not alter the secondary structure of the receptor. In contrast, the tryptophan fluorescence spectra demonstrated that DTT induces a reversible conformational change of the beta 2 receptor. Thus, DTT caused a red-shift in the maximum emission wavelength of the intrinsic tryptophan fluorescence. The change in emission spectrum correlated with a loss in the ability of the receptor to bind antagonist. Both changes in receptor binding and fluorescence emission were reversible, as removal of DTT allowed the receptor to restore 70% of ligand binding and return to the initial emission spectrum. Furthermore, we found adrenergic antagonists were able to slow the rate of the conformational change induced by DTT but not the rate of disulfide reduction, suggesting that the antagonists stabilize the structure of the reduced receptor.

To establish a purification procedure for the human serotonin transporter (hSERT) we expressed in Sf9 insect cells an epitope-tagged version of the transporter containing a FLAG epitope at the N-terminus and a polyhistidine tail at the C-terminus (FLAG-hSERT-12H). For purification, the transporter was solubilized in digitonin followed by nickel affinity and subsequent concanavalin A chromatography. Using this procedure we were able to obtain an overall purification of 700-fold and a yield of approximately 0.1 mg/L of cell culture. The purified transporter displayed pharmacological properties similar to those of hSERT expressed in native tissues and in transfected cell lines. Fluorescent labeling of the purified transporter with the thiol-reactive fluorophore nitrobenxoxadiazol-iodoacetamide (IANBD) and Texas Red bromoacetamide preserved the pharmacological profile of FLAG-hSERT-12H. Collisional quenching experiments revealed that the aqueous quencher iodide was able to cause marked quenching of the fluorescence of the IANBD labeled transporter with a K(SV) of 3.4 +/- 0.10 M(-)(1). In a mutant transporter with five cysteines mutated (5CysKO) we observed a significant reduction in this quenching (K(SV) = 2.1 +/- 0.16 M(-)(1), p < 0.01). This reduction was most likely due to labeling of (109)Cys since mutation of this cysteine alone resulted in a reduction in collisional quenching that was similar to that observed with 5CysKO (K(SV) = 2.2 +/- 0.15 M(-)(1)). These data suggest that labeling of (109)Cys contributes substantially to the overall fluorescence of IANBD labeled FLAG-hSERT-12H. On the basis of these data we infer that (109)Cys is embedded in a mixed hydrophobic/hydrophilic environment at the external ends of transmembrane segments 1 and 2. Further use of fluorescent techniques on purified hSERT should prove useful in future studies aimed at understanding the molecular structure and function of Na(+)/Cl(-)-dependent neurotransmitter transporters.

Previously, we have identified three Zn(2+) binding residues in an endogenous Zn(2+) binding site in the human dopamine transporter (hDAT): (193)His in extracellular loop 2 (ECL 2), (375)His at the external end of transmembrane segment (TM) 7, and (396)Glu at the external end of TM 8. Here we have generated a series of artificial Zn(2+) binding sites in a domain situated around the external ends of TMs 7 and 8 by taking advantage of the well-defined structural constraints for binding of the zinc(II) ion. Initially, we found that the Zn(2+)-coordinating (193)His in ECL 2 could be substituted with a histidine inserted at the i - 4 position relative to (375)His in TM 7. In this mutant (H193K/M371H), Zn(2+) potently inhibited [(3)H]dopamine uptake with an IC(50) value of 7 microM as compared to a value of 300 microM for the control (H193K). These data are consistent with the presence of an alpha-helical configuration of TM 7. This inference was further corroborated by the observation that no increase in the apparent Zn(2+) affinity was observed following introduction of histidines at the i - 2, i - 3, and i - 5 positions. In contrast, introduction of histidines at positions i + 2, i + 3, and i + 4 all resulted in potent inhibition of [(3)H]dopamine uptake by Zn(2+) (IC(50) = 3-32 microM). These observations are inconsistent with continuation of the helix beyond position 375 and indicate an approximate boundary between the end of the helix and the succeeding loop. In summary, the data presented here provide new insight into the structure of a functionally important domain in the hDAT and illustrate how engineering of Zn(2+) binding sites can be a useful approach for probing both secondary and tertiary structure relationships in membrane proteins of unknown structure.