点击显示 收起
【关键词】 N-Ethylmaleimide-Sensitive
Recycling of G protein-coupled receptors determines the functional resensitization of receptors and is implicated in switching β2 adrenoceptor (β2AR) G protein specificity in cardiomyocytes. The human β2AR carboxyl end binds to the N-ethylmaleimide-sensitive factor (NSF), an ATPase integral to membrane trafficking machinery. It is interesting that the human β2AR (hβ2AR) carboxyl end pulled down NSF from mouse heart lysates, whereas the murine one did not. Despite this difference, both β2ARs exhibited substantial agonist-induced internalization, recycling, and Gi coupling in cardiomyocytes. The hβ2AR, however, displayed faster rates of agonist-induced internalization and recycling compared with the murine β2AR (mβ2AR) and a more profound Gi component in its contraction response. Replacing the mβ2AR proline (-1) with a leucine generated a gain-of-function mutation, mβ2AR-P417L, with a rescued ability to bind NSF, faster internalization and recycling than the mβ2AR, and a significant enhancement in Gi signaling, which mimics the hβ2AR. Selective disruption of the mβ2AR-P417L binding to NSF inhibited the receptor coupling to Gi. Mean-while, inhibiting NSF with N-ethylmaleimide blocked the mβ2AR recycling after agonist-induced endocytosis. Expressing the NSF-E329Q mutant lacking ATPase activity inhibited the mβ2AR coupling to Gi in cardiomyocytes. Our results revealed a dual regulation on hβ2AR trafficking and signaling by NSF through direct binding to cargo receptor and its ATPase activity and uncovered an unprecedented role for the receptor binding to NSF in regulating G protein specificity that has diverged between mouse and human β2ARs.
β-Adrenoceptors play a pivotal role in regulating cardiomyocyte contraction through distinct signaling pathways. The β1AR couples to Gs protein(s), which increases cAMP/protein kinase A activity and the contraction rate, whereas the activated β2AR sequentially couples to both Gs and Gi in neonatal cardiomyocytes, creating a biphasic change in contraction. β2AR Gi coupling seems to be dependent on receptor trafficking, which includes both endocytosis and recycling. Inhibiting either process blocks receptor coupling to Gi in cardiomyocytes (Xiang et al., 2002; Xiang and Kobilka, 2003).
Many G protein-coupled receptors (GPCRs) undergo endocytosis in response to activation, yet their subsequent sorting in endosomes is variable, creating variable regulation of their activity during prolonged or repeated stimulation. Some receptors are targeted to lysosomes to down-regulate cellular responses mediated by the receptor, whereas many GPCRs possess the ability to efficiently return to the cell surface. This recycling of receptors underlies the resensitization of corresponding cellular responses (von Zastrow, 2003). Many GPCRs depend on sequences residing in their intracellular domains for recycling. A well-defined class of recycling sequences are PSD-95/Discs-large/ZO-1 (PDZ) domain binding motifs (also called PDZ ligands) that are usually located at the carboxyl-terminal end of different GPCR tails (Bockaert et al., 2004; Gage et al., 2005). The β2AR has a type I PDZ ligand at its carboxyl-terminal end that is necessary for recycling and sufficient to reroute the -opioid receptor from a degradative to a recycling pathway (Cao et al., 1999; Gage et al., 2001). In cultured neonatal mouse cardiomyocytes, this sequence is also required for the temporal switch from Gs to Gi-mediated signal transduction observed in the contraction-rate response to the agonist isoproterenol (Xiang and Kobilka, 2003). Several lines of evidence now indicate that membrane trafficking of this receptor dictates not only cellular resensitization but also signal transduction specificity. Despite progress in understanding the β2AR recycling process, numerous questions concerning the core mechanism and physiological variations remain.
Although the recycling sequence at the β2AR C terminus has been shown to bind PDZ domains in NHERF family proteins (NHERF-1/EBP50 and NHERF-2/E3KARP) (Hall et al., 1998; Cao et al., 1999), it also binds at least one protein with no identifiable PDZ domain: the N-ethylmaleimide sensitive factor (NSF) (Cong et al., 2001). NSF has been identified as an ATPase that binds SNAP receptor (SNARE) complexes in an ATP-dependent fashion to separate them during ATP hydrolysis; this and a wealth of other evidence has demonstrated its general role in vesicle fusion between various membrane compartments (Morgan and Burgoyne, 2004; Whiteheart and Matveeva, 2004). Moreover, NSF has been shown to bind to β-arrestin, an adaptor protein involved in GPCR desensitization and endocytosis upon agonist stimulation. β-Arrestin preferentially interacts with the ATP-bound form of NSF, and this NSF binding facilitates clathrin coat-mediated GPCR internalization (McDonald et al., 1999). In heterologous HEK293 cells, selective ablation of NSF binding to the β2AR was inferred to inhibit recycling of receptors, whereas imparting NSF binding on the -opioid receptor slightly enhanced its ability to recycle (Cong et al., 2001; Gage et al., 2005). Although there is also evidence to show that PDZ interactions promote receptor recycling (Cao et al., 1997) and are functionally important for Gi coupling in cardiomyocytes (Xiang and Kobilka, 2003), it is not clear how NSF may affect β2AR trafficking and signaling in these cells.
Here, we used neonatal mouse cardiomyocytes as a model system to address these questions. It is interesting that the NSF binding sites on the β2AR were not conserved among mammalian species, providing a naturally occurring divergence in NSF binding to exploit. The -1 position of the β2AR carboxyl terminus is proline in mβ2AR and leucine in hβ2AR. Because of this single amino acid difference, mβ2AR binding to NSF was not detectable. Nevertheless, despite the lack of detectable binding of the mβ2AR carboxyl terminus to NSF in biochemical assays, we found that inhibition of NSF activity with N-ethylmaleimide (NEM) inhibited murine β2AR (mβ2AR) recycling despite this poor affinity. In addition, both human and murine β2ARs sufficiently recycled after endocytosis and coupled to Gi pathways in cardiomyocytes. The different affinities for NSF seemed to have a minimum role on receptor trafficking and signaling. In contrast, inactivation of NSF ATPase activity with a point mutation was sufficient to block both human and murine β2AR recycling and coupling to Gi in cardiomyocytes, indicating that NSF is required for proper trafficking and signaling of β2ARs in cardiomyocytes independent of a high-affinity interaction with the receptor. This study strengthens the relationship between β2AR recycling and signaling specificity and demonstrates an unprecedented role for NSF in regulating physiologically relevant signal transduction.
cDNA Constructs and Mutagenesis. Constructs containing the cloned human and murine β2AR in pcDNA3 (Invitrogen, Carlsbad, CA) with a FLAG epitope attached at the N terminus were used for these studies and have been described before (Cao et al., 1999; Swaminath et al., 2004). Constructs encoding for GST-β2AR and GST-β2AR-alanine proteins (encompassing amino acids 328 to 413 of the human β2AR, and the latter with an additional alanine added to the C terminus) have also been reported (Cao et al., 1999). A comparable murine GST-β2AR construct was created by insertion of a polymerase chain reaction product of the region encoding amino acids 328 to 418 of the FLAG-mβ2AR construct using primers containing EcoRI and HindIII appendages and performing the appropriate digestion and ligation into pGEX-KG (Pfizer, New York, NY). The human NSF coding sequence was similarly ligated into pEGFP-N1 (Clontech, Mountain View, CA) after SacI digestion of the vector and a polymerase chain reaction product containing a 3'-SacI appendage and including the 5'-SacI restriction site from the source vector, NSF in pBluescriptR (American Type Culture Collection, Manassas, VA). The P417L mutation was introduced into the FLAG-mβ2AR and GST-mβ2AR constructs via the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA), as was the E329Q NSF mutation into the pEGFP-N1 construct. Plasmid amplification was done in DH5 Escherichia coli, and all sequences were verified by dideoxynucleotide sequencing (University of California San Francisco Biomolecular Resource Center, San Francisco, CA).
Cell Culture and Transfection. Spontaneously beating neonatal cardiomyocytes were prepared from hearts of 1-day-old β1/β2AR-KO mouse pups as before (Devic et al., 2001). The myocyte-enriched cells remaining in suspension after preplating were plated in 35-mm dishes for contraction-rate studies and in 12-well plates for immunological assays (with coverslips for immunofluorescent microscopy). Recombinant adenovirus encoding FLAG-mβ2AR has been described previously (Xiang et al., 2002), and the FLAG-mβ2AR/P417L, FLAG-hβ2AR, GFP-NSF, and GFP-NSF-E329Q adenoviral vectors were generated with the same pAdEasy system (Qbiogene Inc., Irvine, CA). Neonatal myocytes were infected with viruses at a multiplicity of infection of 100 after being cultured for 24 h. The receptor expression levels were determined by ligand binding assays as described previously (Xiang et al., 2002). They were expressed at equivalent levels in cardiac myocytes (FLAG-mβ2AR, 147.3 ± 22 fmol/mg; FLAG-mβ2AR/P417L, 171.6 ± 9.1 fmol/mg; and FLAG-hβ2AR, 160.3 ± 21.8 fmol/mg membrane).
GST Pulldown Assays. The various GST-β2AR fusion proteins were produced in BL21 E. coli and bound to glutathione-Sepharose agarose beads (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK). Beads containing 10 µg of the full-length fusion protein (assessed by densitometry of Coomassie-stained protein resolved by SDS-polyacrylamide gel electrophoresis) were incubated for 4 h at 4°C in 0.5 ml of clarified extracts from frozen mouse hearts with atria removed (Pel-Freez Biologicals, Rogers, AR), prepared to 10 mg/ml. Beads were washed four times in 1 ml of extract buffer [0.1% (v/v) Triton X-100, 150 mM NaCl, 25 mM KCl, and 10 mM Tris, pH 7.4, complete Roche protease inhibitor cocktail], and protein was eluted in lithium dodecyl sulfate sample buffer (Invitrogen) with dithiothreitol added to 20 mM. Samples were divided in two for SDS-polyacrylamide gel electrophoresis, transfer to nitrocellulose, and Western blotting using rabbit anti-EBP50 antibodies (courtesy of Dr. Anthony Bretscher, Cornell University, Ithaca, NY) or the mouse 2E5 anti-NSF antibody (courtesy of Dr. Sidney W. Whiteheart, University of Kentucky, Lexington, KY).
Immunofluorescence Microscopy. Myocyte images were obtained using a similar setup on a Zeiss Axioplan 2 microscope (Carl Zeiss Inc., Thornwood, NY). Fluorescent measurements of the myocyte receptor trafficking were made by a ratiometric normalization of fluorescent intensities measured using Metamorph software (Molecular Devices, Sunnyvale, CA). Epitope-tagged receptors were detected using M1 anti-FLAG antibody (Sigma, St. Louis, MO). Selective detection of surface relative to total pools of receptor and its use to estimate receptor recycling have been described previously (Tanowitz and von Zastrow, 2003). The recycling estimates were conducted without the EDTA strip. The primary antibody used in these experiments was M1 conjugated to Alexa Fluor 488 (Invitrogen) using standard procedures as described previously (Tanowitz and von Zastrow, 2003). Secondary staining was performed using a commercial goat antimouse IgG Alexa Fluor 594 conjugate (Invitrogen). Experiments were performed at least in triplicate, and representative results are shown.
Fig. 1. The binding of mβ2AR and hβ2AR to NSF and NHERF-1/EBP50. The binding of mβ2AR and hβ2AR to NSF and NHERF-1/EBP50 from mouse heart extracts. GST pull-downs were performed as described under Materials and Methods. Western detection of NHERF-1/EBP50 pulled down by the indicated GST-β2AR fusion protein from mouse heart extract, and the corresponding detection of NSF is shown in A, whereas the detection of NHERF-1/EBP50 from this extract is shown in B. Ponceau-stained GST fusion proteins are shown under each blot. Images are representative of three or more experiments.
Immunofluorescence Spectroscopy. Surface receptor levels were determined as before (Swaminath et al., 2004) in the indicated cell type expressing the indicated FLAG-β2AR. Media were refreshed 1 h before 10 µM isoproterenol (Sigma) stimulation for 10 or 30 min. Periods of agonist washout after 30-min isoproterenol stimulations were also performed for an additional 30 or 60 min as indicated.
Myocyte Contraction Rate Assay. Measurement of spontaneous contraction rates from myocytes expressing either the endogenous or the indicated FLAG-β2AR were carried out with and without the use of PTX as described previously (Devic et al., 2001). In some assays, NEM was applied 30 min before the addition of isoproterenol. Tat peptide, Tat-β2-DSAL consisting of Tat linked to GRQGFSSDSAL of β2AR, and Tat-β2-ASLL consisting of Tat linked to GRQGFSSASLL of β2AR through a cysteine bridge were synthesized in the Stanford Core facility and EZ-Biolab (Indianapolis, IN). Neonatal myocytes were preincubated at 37°C with 10 µM peptide for 25 min before isoproterenol (10 µM; Sigma) exposure.
Statistical Analysis. Curve-fitting and statistical analyses were performed using Prism (GraphPad Software, Inc., San Diego, CA).
NSF Had Higher Binding Affinity to Human β2AR than Murine β2AR. To understand the molecular mechanism of the NSF effect on β2AR signaling in cardiomyocytes, the interaction between β2AR and NSF from heart lysate was examined. NSF, a hexameric ATPase involved in membrane fusion, can bind to the carboxyl terminus of the hβ2AR. The protein-binding region on this receptor involves a four-residue stretch at the distal C terminus of the receptor (Cong et al., 2001). In addition, NHERF-1/EBP50, a cytoskeleton-associated protein, can also bind to the same stretch of residues on the carboxyl terminus of the hβ2AR. It is interesting that several rodent β2ARs, including the mβ2AR, are identical with the hβ2AR in the carboxyl-binding region except at one residue (leucine -1 of the hβ2AR, DSLL) that is required for binding to NSF but not to NHERF-1/EBP50 (Cong et al., 2001). Rather, the mβ2AR has a proline at the -1 position (DSPL). To determine whether both the human and murine β2ARs have a similar capacity to bind NSF, GST-fusion proteins, including various β2AR carboxyl-terminal tail sequences, were prepared. Protein binding was evaluated with a pull-down assay using the GST-fusion proteins coupled to glutathione-agarose beads.
GST fusion proteins were incubated with tissue lysate prepared from mouse hearts. NSF only bound to the cytoplasmic tail of the hβ2AR but not the mβ2AR (Fig. 1A). In contrast, NHERF-1/EBP50 bound to the cytoplasmic tail of both the hβ2AR and mβ2AR under these conditions (Fig. 1B). As a negative control for nonspecific binding, an addition of a single alanine residue to the hβ2AR carboxyl terminus (GSThβ2AR-Ala) was tested as well. As reported previously, this mutant failed to exhibit the PDZ domain-mediated and NSF protein binding (Fig. 1, A and B; Cao et al., 1999). In addition, we attempted to "rescue" an NSF interaction with the mβ2AR tail by substitution of the mβ2AR proline 417 with a leucine residue (mβ2AR-P417L). The mβ2AR-P417L pulled down similar amounts of NSF from lysates compared with the hβ2AR, indicating that the mβ2AR-P417L cytoplasmic tail fully rescued binding to NSF (Fig. 1A). Likewise, this mutant mβ2AR-P417L pulled down NHERF-1/EBP50 from mouse heart lysates (Fig. 1B).
The Binding of NSF and the NSF ATPase Activity Had Distinct Effects on β2AR Trafficking in Cardiomyocytes. To examine whether NSF plays any role in β2AR trafficking in cardiomyocytes, we analyzed the localization of flag-tagged β2ARs in cardiac myocytes. The mβ2AR, mβ2AR-P417L, and hβ2AR were transiently expressed in cardiac myocytes using recombinant adenovirus. Immunofluorescence studies showed that all three receptors had a cell-surface staining during a nonstimulated state (Fig. 2A). Upon isoproterenol stimulation, all three receptors had reduced cell-surface staining together with increased punctate intracellular staining, suggesting a significant internalization of the receptors in cardiac myocytes. These observations were confirmed quantitatively using an ELISA-based method for assaying surface receptor levels (Swaminath et al., 2004) in a large number of cells and a ratiometric method for analysis of fluorescence micrographs (Tanowitz and von Zastrow, 2003) (Fig. 2B; data not shown). It is interesting that when we measured the short-term decrease in cell-surface receptors after agonist stimulation in cardiac myocytes, we found that the mβ2AR-P417L had a faster rate (t = 2.63 ± 0.05 min) of cell surface-receptor decrease than the mβ2AR (t = 10.93 ± 0.01 min; Fig. 3). Because the receptor level change in the short time points after agonist stimulation is primarily determined by agonist-induced endocytosis, these data suggested a faster rate of endocytosis for the mβ2AR-P417L than for the mβ2AR in cardiac myocytes. The hβ2AR had a similar rate (t = 3.4 ± 0.03 min) of cell surface-receptor decrease compared with the mβ2AR-P417L (Fig. 3). However, after 30 min of agonist stimulation, we observed a similar amount of surface receptor decreases with the mβ2AR (30.54 ± 2.90%), the mβ2AR-P417L (28.43 ± 1.69%), and the hβ2AR (24.94 ± 2.74%). The observed decrease of receptor density at 30 min of stimulation should have been a composite of receptor endocytosis and recycling. The equivalent decreases of receptors at cell surface are usually due to much slower recycling process than endocytosis in cells.
Fig. 2. NSF binding enhances recycling of FLAG-β2ARs in neonatal cardiac myocytes from β1/β2AR-KO mice. A, human and murine β2ARs internalize and recycle in cardiac myocytes. Cardiac myocytes expressing a FLAG-tagged hβ2AR, mβ2AR, or mβ2AR-P417L were stained with M1 primary antibody conjugated to the Alexa-488 fluorophore to observe a starting "total" receptor population. After no treatment (0), 30-min 10 µM isoproterenol treatment (30), or 30-min isoproterenol treatment followed by a surface antibody strip and 60 min of agonist removal (30 + 60), cells were stained under nonpermeable conditions with a goat anti-mouse-IgG secondary antibody conjugated to the Alexa-594 fluorophore to observe the relative complement of "surface" receptor. Images are representative of three experiments. B, NSF binding β2ARs recycle faster. Surface levels of the three β2ARs were quantified by fluorescence spectroscopy measurements of M1-Alexa 488 associated with the cell surface receptors after the indicated periods of drug administration and removal (1, control; 2, 30 min of isoproterenol stimulation; 3, 30 min of isoproterenol followed by 30 min of drug removal; and 4, 30 min of isoproterenol followed by 60 min of drug removal). Surface levels are normalized as a percentage of untreated cell surface fluorescence, and error bars reflect standard deviations over three experiments. *, p < 0.05, significantly different between mβ2AR and hβ2AR or mβ2AR-P417L by t test.
Fig. 3. NSF binding enhances the internalization kinetics of FLAG-β2ARs expressed in neonatal cardiac myocytes from β1/β2AR KO mice. Surface levels of the three β2ARs were quantified by fluorescent measurement of M1-Alexa 488 associated with the cell surface receptors after the indicated periods of 10 µM isoproterenol administration. Data were normalized as a percentage decrease of untreated cell surface fluorescence, and error bars reflect standard deviations over three experiments. The data represent the mean ± S.E. of experiments from at least three different myocyte preparations.
When isoproterenol was removed, both the hβ2AR and mβ2AR-P417L recovered cell-surface staining almost completely after a 60-min incubation (Fig. 2). In contrast, the mβ2AR did not show a fully recovered cell-surface staining pattern, and some residual intracellular staining was observed in these cells (Fig. 2), even though the majority of the internalized receptors seemed to return to the surface within 60 min. When we examined the cell surface-receptor density at different time points with the fluorescent ELISA assay, the mβ2AR exhibited a lower recovery of cell surface-receptors after recycling for 60 min than the mβ2AR-P417L mutant and the hβ2AR (Fig. 2B; *, p < 0.05). A significant difference in surface recovery was also observed using ratiometric image measurements 60 min after agonist washout (data not shown). These data indicate that the mβ2AR, although capable of undergoing agonist-induced internalization and recycling in cardiomyocytes, differs in rates of recycling compared with the hβ2AR and mβ2AR-P417L.
It has been well-established that NSF ATPase activity plays an important role in membrane cargo trafficking. We then tested whether NSF activity was necessary for the endocytic recycling of the receptor by using NEM to inhibit NSF activity in myocytes. In the presence of NEM, endocytosis of the receptor was preserved; however, a return of the receptor to the cell surface after removal of agonist for 60 min was not (Fig. 4A). This observation was confirmed with measurements of surface receptor levels by a fluorescent ELISA assay. The cell surface receptor levels dropped after agonist addition and only recovered with agonist withdrawal in the absence of NEM (Fig. 4B). These data suggested that NEM treatment can block the receptor from recycling after endocytosis.
Fig. 4. Inhibiting NSF with NEM blocks the FLAG-β2AR recycling after agonist-induced endocytosis in cardiomyocytes. A, murine β2ARs internalize and recycle in cardiac myocytes. Cardiac myocytes expressing a FLAG-tagged mβ2AR were treated as described under Materials and Methods and Fig. 3. Images are representative of three experiments. Inhibiting NSF with NEM blocks the FLAG-β2AR recycling after agonist-induced endocytosis in cardiomyocytes. B, surface levels of the mβ2ARs were quantified by fluorescent measurement of M1-Alexa 488 associated with the cell surface receptors after the indicated periods of drug administration and removal (1, control; 2, 30 min of isoproterenol stimulation; 3, 30 min of isoproterenol followed by 30 min of drug removal; and 4, 30 min of isoproterenol followed by 60 min of drug removal). Surface levels were normalized as a percentage of untreated cell surface fluorescence, and error bars reflect standard deviations over three experiments. *, p < 0.05, significantly different between cells with and without NEM treatment by t test.
Dominant-Negative NSF Lacking ATPase Activity Inhibited Endogenous mβ2AR Coupling to Gi Pathway in Cardiomyocytes. Our finding that NEM inhibits β2AR recycling in cardiac myocytes suggested that NSF function is required for this process. To further probe whether NSF enzymatic activity can affect the receptor signaling independent from the direct NSF-receptor interaction, we examined the signaling mediated by the endogenous mβ2AR when overexpressing an inactivated NSF, the E329Q mutant (Whiteheart et al., 1994). This mutation abolishes ATPase activity and has been shown to block AMPA receptor trafficking (Whiteheart et al., 1994; Whiteheart and Matveeva, 2004). When the endogenous mβ2AR in the β1AR-KO myocyte was stimulated by isoproterenol, the activated receptor induced a biphasic contraction-rate response with an initial increase mediated by Gs coupling followed by a sustained Gi-dependent decrease to reduce the contraction rate below basal level (Fig. 5A; Xiang et al., 2002). When wild-type NSF was expressed in β1AR-KO cardiac myocytes, we did not observe any significant change in the endogenous mβ2AR-mediated contraction-rate response (Fig. 5A). In contrast, when the NSF-E329Q mutant was overexpressed in cardiomyocytes, the contraction rate mediated by the mβ2AR was significantly higher than the control and did not display a decrease lower than the basal level (Fig. 5B). This response profile was similar to that observed with an inhibition of Gi by PTX (Fig. 5C). Indeed, additional treatment of PTX did not generate any further increases in contraction rates (Fig. 5D). Therefore, the NSF-E329Q behaved as a dominant-negative to block the receptor coupling to Gi in cardiomyocytes. In addition, when myocytes are pretreated with NEM to inhibit the NSF ATPase activity, we also observed effects similar to those by NSF-E329Q mutant on mβ2AR signaling mediated contraction-rate response (data not shown).
Fig. 5. Dominant-negative NSF-E329Q mutant inhibits the mβ2AR coupling to Gi protein. Spontaneously beating cardiac myocytes from β1AR KO mice were transfected with a wild-type NSF (A and C) or NSF-E329Q (B and D) mutant adenovirus as indicated. The cells were administered 10 µM isoproterenol with inhibition of Gi by PTX. Overexpressing the NSF E329Q mutant enhanced the contraction-rate increase induced by isoproterenol stimulation. Additional PTX treatment did not further enhance the contraction-rate increase induced by the mβ2AR. The data represent the mean ± S.E. of experiments from at least three different myocyte preparations. *, p < 0.05, time course significantly different by two-way ANOVA.
The Divergent C Termini of the Human and Murine β2AR Had Different Effects on Contraction Rate Responses in Neonatal Cardiomyocytes. Our previous studies have shown that the localization and trafficking of the mβ2AR is important for the receptor's G protein signaling specificity and subsequent regulation of the myocyte contraction rate. In the course of this study, we found that the divergent PDZ ligand of the human and murine β2AR affected the receptor trafficking rates after agonist stimulation in cardiac myocytes. Thus, we wanted to examine whether differences in NSF binding and/or altered trafficking rates could modulate the receptor signaling in cardiac myocytes. When the mβ2AR was expressed in β1/β2AR-KO myocytes and stimulated by isoproterenol, the activated receptor induced a biphasic contraction-rate response with an initial increase followed by a sustained decrease to reduce the contraction rate lower than basal level (Fig. 6A; Xiang et al., 2002). This contraction-rate change is equivalent to that induced by the endogenous mβ2AR in β1AR-KO myocytes (Fig. 6A). The mβ2AR-P417L induced a similar contraction-rate response profile and initial increase compared with the mβ2AR in β1/β2AR-KO myocytes (Figs. 6C and 7D). However, the contraction rate decreased faster, and the contraction rate was lower than that induced by the mβ2AR during late stimulation in cardiac myocytes (Figs. 6C and 7E). In addition, when stimulating the hβ2AR expressed in β1/β2AR-KO myocytes with isoproterenol, the activated receptor also induced a biphasic, contraction-rate change with an initial increase followed by a sustained decrease (Fig. 6B). Although it is interesting that the initial contraction-rate increase was smaller than that induced by the activated mβ2AR and mβ2AR-P417L (Figs. 6B and 7D), it is more surprising that the late decrease in contraction rate induced by the exogenous hβ2AR was greater than that induced by the mβ2AR and mβ2AR-P417L (Figs. 6B and 7E).
Fig. 6. Differences in β2AR contraction-rate responses to isoproterenol in neonatal cardiac myocytes from β1/β2AR KO mice. The hβ2AR and mβ2AR-P417L exhibit different contraction rate profiles than the mβ2AR at comparable expression levels. Spontaneously beating, cardiac myocytes from β1/β2AR KO mice were infected with a FLAG-tagged mβ2AR (A), hβ2AR (B), or mβ2AR-P417L (C) recombinant adenovirus as indicated and infused with 10 µM isoproterenol. Contraction rates were measured and normalized as the change over baseline. The data represent the mean ± S.E. of experiments from at least three different myocyte preparations. *, p < 0.05, time course significantly different by two-way ANOVA.
Fig. 7. NSF binding enhances the Gi signaling components of β2ARs. Spontaneously beating cardiac myocytes from β1/β2AR KO mice were transfected with a FLAG-tagged mβ2AR (A), mβ2AR-P417L (B), or hβ2AR (C) adenovirus as indicated. The cells were administered 10 µM isoproterenol with inhibition of Gi with PTX. PTX treatment did not affect initial response usually mediated by receptor/Gs coupling (D) but significantly enhanced the contraction rate during the late stimulation induced by the Gi coupling to the activated hβ2AR, mβ2AR, or mβ2AR-P417L (E). The data represent the mean ± S.E. of experiments from at least three different myocyte preparations. *, p < 0.05, time course significantly different by two-way ANOVA. **, p < 0.05, unpaired t test significantly different on initial maximum contraction rate increases or late contraction-rate decreases mediated by different β2ARs. ***, p < 0.05, unpaired t test significantly different on late contraction rate decreases after PTX treatment.
The profound contraction-rate decrease induced by the mβ2AR-P417L and the hβ2AR suggests that these receptors may have enhanced coupling to Gi and/or reduced coupling to Gs compared with the mβ2AR. We therefore examined the Gi signaling induced by the activated receptors in cardiac myocytes. PTX was used to block Gi signaling in cardiac myocytes expressing the different β2ARs before isoproterenol stimulation. Upon inhibiting Gi with PTX, the isoproterenol-stimulated mβ2AR induced a slightly greater but not significant contraction-rate increase in myocytes compared with the control and prevented the late Gi-dependent contraction rate decrease (Fig. 7A; Xiang et al., 2002). PTX treatment also inhibited the contraction rate decrease mediated by the mβ2AR-P417L or the hβ2AR during the late phase of stimulation (Fig. 7, B and C). These data suggest that compared with the activated mβ2AR, the mβ2AR-P417L had an enhanced Gi coupling upon isoproterenol stimulation, and the activated hβ2AR coupled to Gi more efficiently in neonatal cardiac myocytes (Fig. 7E). This indicates that the divergent receptor C termini can induce different changes in contraction-rate responses that correlate with subtle changes in receptor transportation rates.
The Binding of NSF and PDZ Had Distinct Effects on β2AR Activation-Induced Contraction Rates in Cardiomyocyte. To further probe the effect of the β2AR binding to NSF and PDZ on receptor signaling in cardiomyocytes, we took advantage of the different binding affinities between receptor and proteins by using peptides to selectively disrupt the interactions. We expressed either mβ2AR or mβ2AR-P417L (the hβ2AR mimic) in β1/β2AR-KO cardiomyocyte for the contraction rate assay. Membrane-permeable peptides containing ASLL sequence and DSAL sequence were used to selectively disrupt NSF and NHERF/EBP50 binding, respectively. When mβ2AR-expressing myocytes were treated with NSF (ASLL) peptide, the activated receptor induced a slightly bigger but not significant initial increase than the cells without pretreatment (Fig. 8, A and C). The increase was sustained during stimulation and lacked a late decrease mediated by receptor/Gi coupling in control cells (Fig. 8, A and D). When mβ2AR-expressing myocytes were treated with PDZ (DSAL) peptide, the activated receptor induced a significantly greater initial increase than the control (Fig. 8, B and C), and the increase was sustained and lacked a late Gi-dependent decrease (Fig. 8, B and D).
Fig. 8. Selective disruptions of NSF and NHERF-1/EBP50 binding have distinct effects on β2AR signaling. Spontaneously beating cardiac myocytes from β1/β2AR KO mice were transfected with a FLAG-tagged mβ2AR (A-D) or mβ2AR-P417L (E-H) adenovirus as indicated. The cells were administered 10 µM isoproterenol with pretreatment of membrane-permeable NSF peptide ASLL and PDZ peptide DSAL to disrupt the receptor binding to NSF and PDZ protein, respectively. NSF peptide ASLL significantly affected the receptor-mediated contraction response during the late stimulation, which are usually mediated by receptor/Gi coupling (D and H). In contrast, PDZ peptide affected both initial contraction rate increase mediated by receptor/Gs coupling (C and G) and the late contraction-rate response mediated by receptor/Gi coupling (D and H). The data represent the mean ± S.E. of experiments from at least three different myocyte preparations. *, p < 0.05, time course significantly different by two-way ANOVA. **, p < 0.05, unpaired t test significantly different on initial maximum contraction-rate increases or late contraction-rate decreases after treatment with peptides.
In contrast, pretreatment with NSF (ASLL) peptide did not affect the activated mβ2AR-P417L-induced initial increase (Fig. 8, E and G). However, the increase was sustained and did not display a late Gi-induced decrease of contraction rate (Fig. 8, E and H). When myocytes expressing mβ2AR-P417L were treated with PDZ (DSAL) peptide, the activated receptor induced a significantly greater initial increase in contraction rate than the control (Fig. 8, F and G), and the increase was sustained and did not show a late Gi-induced decrease (Fig. 8, F and H). Together, these data showed that although disrupting the binding to PDZ protein (such as NHERF/EBP50) affects the receptor coupling to both Gs and Gi, disrupting the binding to NSF selectively affects the receptor coupling to Gi in cardiomyocytes.
In the present study, several approaches were used to test whether NSF regulates β2AR trafficking and physiological signaling. This idea was extended from the studies of β2AR-selective interactions with NSF and PDZ proteins. A distal portion of the cytoplasmic C terminus of the hβ2AR selectively binds to several PDZ domain-containing proteins, such as the cytoskeleton-associated protein NHERF/EBP50, which is implicated in receptor recycling (Cao et al., 1999). However, a subsequent study confirmed the importance of the PDZ ligand for receptor recycling to the cell surface but identified a distinct non-PDZ interaction of this sequence with NSF that was required for proper endocytic recycling (Cong et al., 2001). Although the reported difference could result from the difference between the derived HEK293 cell lines, we have tried to address the functional roles of these receptor-protein interactions in primary cultured cardiomyocytes—a native environment that may have a more precise regulation of receptor function. We have shown previously that the carboxyl-terminal sequence of mβ2AR was also required for efficient plasma membrane recycling and for receptor coupling to Gi in cardiomyocytes (Xiang and Kobilka, 2003). In this study, we showed that the binding to NSF enhanced both internalization and recycling rates of β2AR and increased the receptor coupling to Gi signaling in cardiomyocytes (Figs. 2, 3, and 8). We further distinguished the effects of NSF and PDZ binding on β2AR signaling in myocytes. Although the binding to NSF increases receptor/Gi coupling, the binding to PDZ proteins affects receptor coupling to both Gs and Gi proteins (Fig. 8).
It is interesting that, at the receptor's distal carboxyl terminus, the mβ2AR (DSPL) differs from the hβ2AR (DSLL) at the -1 position, at which the hβ2AR has a leucine critical for binding to NSF. Because the mβ2AR has a proline residue at the same relative position of the receptor cytoplasmic tail, we predicted that the receptor could not bind to NSF. Our experiments confirmed a very low affinity binding of NSF to the mβ2AR cytoplasmic tail (Fig. 1). We used a gain-of-function approach by replacing the proline with a leucine to generate a mutant mβ2AR-P417L. This mutant has a distal terminus identical with that of the hβ2AR (hβ2AR mimic) and displayed recovered binding to NSF (Fig. 1). The direct NSF-binding seemed to increase the rates of both agonist-induced endocytosis and recycling of the mβ2AR-P417L in cardiomyocytes (Fig. 2 and 3). We cannot exclude the possible contribution by the small increase in binding affinities of the mutant mβ2AR-P417L for PDZ proteins or new binding partners. However, our beating assay data supported that the increased trafficking rates are probably caused by the fact that the mutant mβ2AR-P417L gained binding to NSF (Fig. 8).
Consistent with the trafficking data, the mβ2AR-P417L and the hβ2AR also displayed a more profound coupling to Gi than the mβ2AR in cardiomyocytes (Fig. 7). We have established previously that activated mβ2AR undergo sequential coupling to Gs and Gi to modulate cardiomyocyte contraction rate, and the recycling of mβ2AR is necessary for coupling to Gi (Xiang and Kobilka, 2003). By using membrane-permeable peptides to selectively inhibit the receptor binding to NSF or PDZ proteins, we will be able to distinguish the subtle effects of a specific binding on receptor signaling. Although disruption of PDZ binding affects receptor coupling to both Gs and Gi, disruption of NSF binding selectively inhibits receptor coupling to Gi (Fig. 8). It is interesting that despite that the mβ2AR does not bind to NSF well, the NSF peptide ASLL affected the receptor signaling (Fig. 8A). This may result from a low basal interaction between the mβ2AR and NSF. On the other hand, NSF peptide ASLL is capable of binding to PDZ proteins (Cong et al., 2001); thus, it may compete against DSPL on the mβ2AR, which is not a perfect PDZ ligand because of the structure of proline. In contrast, the binding between DSLL of the mβ2AR-P417L and PDZ proteins is less likely to be affected by the NSF peptide (Fig. 8E). Thus, the effect on the mβ2AR-P417L signaling by the NSF peptides suggested that the binding to NSF affects the receptor/Gi coupling, which is consistent with its role in the modulation of receptor recycling.
NSF was identified as an ATPase, binding to SNARE complexes required for membrane fusion, thus playing critical roles in protein trafficking of many membrane receptors (Whiteheart and Matveeva, 2004). In agreement, we showed that NSF ATPase activity was essential for mβ2AR trafficking and signaling in cardiomyocytes (Fig. 4 and 5). It is interesting that NSF can bind to β-arrestin, an adaptor-like protein linking most GPCRs to clathrin-coated vesicles for endocytosis (McDonald et al., 1999). NSF binding to β-arrestin, like binding to classic SNARE substrates, is an ATP-dependent event (McDonald et al., 1999). Thus, NSF could play a role together with β-arrestin in recruiting the cargo receptors into clathrin-coated vesicles for budding. This process can be fine-tuned if NSF directly binds membrane cargo receptors, including hβ2AR (Heydorn et al., 2004). In addition, the binding of NSF to the hβ2AR is enhanced in the ATP-bound form (Gage et al., 2005), and the NSF ATPase activity dissociates the AMPA receptor from PDZ proteins allowing endocytosis (Osten et al., 1998; Hanley et al., 2002). Therefore, NSF can facilitate receptor recruitment into clathrin-coated vesicle by both direct binding to the cargo receptor and its ATPase activity. In the case of mβ2AR, the activated receptor recruits β-arrestin; this brings NSF to the receptor. NSF ATPase activity helps to dissociate the receptor from PDZ proteins to enter clathrin-coated vesicles, and later NSF regulates the vesicle fusion to endosome. In comparison, hβ2AR can directly bind to the NSF. When NSF is recruited to the receptor/arrestin complexes, it can compete against the receptor binding to PDZ proteins. This competition can lead to an increase in internalization rates (Fig. 3). During the receptor recycling, NSF binding can bridge the cargo receptor to SNARE complexes, which facilitate the docking of recycling vesicles to plasma membrane, hence enhancing the recycling rate of hβ2AR and mβ2AR-P417L but not mβ2AR. In this study, we only measured the cell surface receptor level during endocytosis and recycling. Any additional role of NSF in receptor trafficking among endosomal compartments remains to be addressed.
It is noteworthy that the subtle effects of NSF-hβ2AR binding on trafficking and signaling is not conserved throughout mammals; such an effect can be overlooked easily in an experimental procedure. Although NSF is a common factor involved in membrane receptor trafficking, the context of the NSF-receptor complex can further complicate the type and degree of receptor regulation. These regulations will probably include the binding of the receptor to PDZ-domain containing proteins and cytoskeleton-associated proteins and additional binding of NSF to other trafficking proteins such as SNARE complexes and arrestin. Further studies using NSF mutants with selective ablation of binding to the β2AR or other proteins such as arrestin will help to dissect any roles of individual protein-protein interactions on the β2AR trafficking and signaling in cardiomyocytes or physiological settings.
Indeed, an effect of NSF binding to the hβ2AR is likely to be complicated by competitive binding of PDZ proteins on the same sequence at the carboxyl-terminal end (Cong et al., 2001). A PDZ binding can have multiple effects on membrane receptor distribution and trafficking. One effect of PDZ binding is to stabilize and restrict the receptors at distinct subcellular domains. This is supported by the recent evidence that overexpressing NHERF-1/EBP50 reduced the agonist-induced internalization of two GPCRs, the parathyroid hormone receptor type-1 and thromboxane A(2)β receptor in HEK293 cells (Rochdi and Parent, 2003; Sneddon et al., 2003). By binding to cytoskeleton and/or scaffold proteins, the receptors can associate with signaling components and form complexes to either facilitate or restrict signal transduction. Consistent with the notion, our previous and current studies support that disrupting the PDZ binding to the β2ARs enhances the receptor coupling to Gs in cardiomyocytes (Xiang and Kobilka, 2003). In contrast, the PDZ protein GRIP/ABP binding has been shown to play a role in the stabilization of an intracellular pool of AMPA receptors that have been internalized with stimulation, thus inhibiting their recycling to the synaptic membrane (Braithwaite et al., 2002). Therefore, depending on the receptors and their binding partners, the PDZ domain-containing proteins can stabilize the receptor complexes at either the cell surface or intracellular compartments to fine-tune the receptor function in a given cell type. The third effect of PDZ binding seems to promote receptor trafficking to another subcellular location. Both PICK1 and NHERF-1/EBP50 have been shown to be critical for AMPA receptor and β2AR recycling back to the cell surface (Cao et al., 1999; Xiang and Kobilka, 2003; Lu and Ziff, 2005). In cardiomyocytes, selective disruption of PDZ binding with point mutations or with membrane-permeable peptide blocks receptor recycling and also inhibits receptor coupling to Gi ((Xiang and Kobilka, 2003) and Fig. 8). This PDZ-promoted trafficking may simply be a result of PDZ sequestration of receptors away from a competing trafficking fate, which could generalize PDZ interactions as hindrances to trafficking. The function of PDZ binding on receptor endocytosis and recycling could be further complicated by agonist-dependent phosphorylation of the receptor C-terminal end by G-protein receptor kinases and subsequent receptor dephosphorylation by pH-sensitive phosphatases (Sibley et al., 1986; Pitcher et al., 1995, 1998; Cao et al., 1999). The significance of this interplay in cardiomyocytes remains to be seen.
When the hβ2AR was expressed in murine cardiomyocytes, the receptor displayed sequential coupling to Gs and Gi to regulate the myocyte contraction rate (Fig. 6). This result reinforced the notion from our previous studies that the recycling of the β2AR is part of a mechanism necessary for the receptor to switch from Gs to Gi (Xiang et al., 2002, 2005; Xiang and Kobilka, 2003). Both the human and murine β2ARs displayed a dual coupling to both Gs and Gi proteins in cardiac myocytes. Our studies revealed a species-dependent difference between human and murine β2ARs. The hβ2AR seemed to have a lower efficiency in coupling to the Gs pathway and a significantly higher efficiency in coupling to Gi than the mβ2AR when regulating the myocyte contraction rate (Figs. 6 and 7). Our results suggest that the profound Gi coupling is in part due to the increased binding to NSF. The mechanism of the low Gs coupling efficiency is not clear, although the higher receptor endocytosis rate could be an indication of enhanced desensitization. Another clue lies in the differences between receptor species. Despite the fact that the mβ2AR-P417L had recovered the ability to bind NSF, much of the signaling properties of this mutant mβ2AR still resembled those of the mβ2AR rather than the hβ2AR (Fig. 6). The differences of other structural domains on the hβ2AR and mβ2AR must thus account for the differences observed between the mβ2AR-P417L and the hβ2AR in cardiomyocytes. The notable regions include both the third loop and the proximal region of the carboxyl tail, which can directly influence G protein coupling. Another species-dependent difference is that a unique sugar-modification site located on the second extracellular domain of the hβ2AR, but not rodent β2ARs, promotes receptor degradation upon long-term agonist stimulation (Mialet-Perez et al., 2004). When overexpressed in mice, the hβ2AR seems to enhance the cardiac contraction in animal hearts without developing heart failure (Milano et al., 1995). The β2AR/Gs signaling is proapoptotic (Zhu et al., 2001), whereas the β2AR/Gi signaling plays an antiapoptotic role in both mouse hearts and cultured mouse cardiac myocytes (Zhu et al., 2001; Patterson et al., 2004). Thus, the more preferential coupling of the hβ2AR to Gi over Gs observed in our experiments could explain the lack of pathologic changes observed with overexpression of the hβ2AR in the hearts of mice. Further studies characterizing the differences between the hβ2AR and mβ2AR are needed to advance our understanding of adrenergic physiology in vivo.
In conclusion, the present results indicate that NSF AT-Pase activity is necessary for agonist-dependent β2AR trafficking in cardiomyocytes, whereas NSF binding enhances the receptor transportation rates. Both the direct binding to NSF and its ATPase activity are important for the receptor coupling to Gi. Our data also showed different affinities of NSF binding to β2ARs from different species, and the direct binding to NSF contributes to the differences of receptor signaling in cardiomyocytes. Our data further revealed distinct effects of NSF and PDZ binding on β2AR signaling. In contrast to the selective effect on Gi coupling by the receptor binding to NSF, the receptor binding to PDZ proteins affects the receptor coupling to both Gs and Gi proteins. The present results add to the growing appreciation of diversified cellular factors as part of comprehensive mechanisms to fine-tune GPCR signaling and membrane trafficking in native mammalian cardiomyocytes.
Acknowledgements
We acknowledge Anthony Bretscher (Cornell) for anti-EBP50 and Sidney Whiteheart (University of Kentucky) for valuable discussion and for anti-NSF.
ABBREVIATIONS: β2AR, β2-adrenergic receptor; β1AR, β1-adrenergic receptor; GPCR, G protein-coupled receptor; KO, knockout; NSF, N-ethylmaleimide-sensitive factor; PDZ, PSD-95/Discs-large/ZO-1; NHERF/EBP50, Na+/H+ exchanger regulatory factor/ezrin/radixin/moesin-binding phosphoprotein of 50 kDa; GST, glutathione; PTX, pertussis toxin; NEM, N-ethylmaleimide; ELISA, enzyme-linked immunosorbent assay; HEK, human embryonic kidney; ANOVA, analysis of variance; AMPA, -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; SNARE, soluble N-ethylmaleimide-sensitive factor attachment protein receptor.
【参考文献】
Bockaert J, Dumuis A, Fagni L, and Marin P (2004) GPCR-GIP networks: a first step in the discovery of new therapeutic drugs? Curr Opin Drug Discov Dev 7: 649-657.
Braithwaite SP, Xia H, and Malenka RC (2002) Differential roles for NSF and GRIP/ABP in AMPA receptor cycling. Proc Natl Acad Sci U S A 99: 7096-7101.[Abstract/Free Full Text]
Cao TT, Deacon HW, Reczek D, Bretscher A, and von Zastrow M (1999) A kinase-regulated PDZ-domain interaction controls endocytic sorting of the beta2-adrenergic receptor. Nature 401: 286-290.
Cong M, Perry SJ, Hu LA, Hanson PI, Claing A, and Lefkowitz RJ (2001) Binding of the β2 adrenergic receptor to N-ethylmaleimide-sensitive factor regulates receptor recycling. J Biol Chem 276: 45145-45152.[Abstract/Free Full Text]
Devic E, Xiang Y, Gould D, and Kobilka B (2001) β-Adrenergic receptor subtype-specific signaling in cardiac myocytes from β1 and β2 adrenoceptor knockout mice. Mol Pharmacol 60: 577-583.[Abstract/Free Full Text]
Gage RM, Kim KA, Cao TT, and von Zastrow M (2001) A transplantable sorting signal that is sufficient to mediate rapid recycling of G protein-coupled receptors. J Biol Chem 276: 44712-44720.[Abstract/Free Full Text]
Gage RM, Matveeva EA, Whiteheart SW, and von Zastrow M (2005) Type I PDZ ligands are sufficient to promote rapid recycling of G protein-coupled receptors independent of binding to N-ethylmaleimide-sensitive factor. J Biol Chem 280: 3305-3313.[Abstract/Free Full Text]
Hall RA, Premont RT, Chow CW, Blitzer JT, Pitcher JA, Claing A, Stoffel RH, Barak LS, Shenolikar S, Weinman EJ, et al. (1998) The beta2-adrenergic receptor interacts with the Na+/H+-exchanger regulatory factor to control Na+/H+ exchange. Nature 392: 626-630.
Hanley JG, Khatri L, Hanson PI, and Ziff EB (2002) NSF ATPase and alpha-/beta-SNAPs disassemble the AMPA receptor-PICK1 complex. Neuron 34: 53-67.
Heydorn A, Sondergaard BP, Ersboll B, Holst B, Nielsen FC, Haft CR, Whistler J, and Schwartz TW (2004) A library of 7TM receptor C-terminal tails. Interactions with the proposed post-endocytic sorting proteins ERM-binding phosphoprotein 50 (EBP50), N-ethylmaleimide-sensitive factor (NSF), sorting nexin 1 (SNX1), and G protein-coupled receptor-associated sorting protein (GASP). J Biol Chem 279: 54291-54303.[Abstract/Free Full Text]
Lu W and Ziff EB (2005) PICK1 interacts with ABP/GRIP to regulate AMPA receptor trafficking. Neuron 47: 407-421.
McDonald PH, Cote NL, Lin FT, Premont RT, Pitcher JA, and Lefkowitz RJ (1999) Identification of NSF as a β-arrestin1-binding protein. Implications for β2-adrenergic receptor regulation. J Biol Chem 274: 10677-10680.[Abstract/Free Full Text]
Mialet-Perez J, Green SA, Miller WE, and Liggett SB (2004) A primate-dominant third glycosylation site of the β2-adrenergic receptor routes receptors to degradation during agonist regulation. J Biol Chem 279: 38603-38607.[Abstract/Free Full Text]
Milano CA, Allen LF, Dolber PC, Johnson TD, Rockman HA, Bond RA, and Lefkowitz RJ (1995) Marked enhancement in myocardial function resulting from overexpression of a human beta-adrenergic receptor gene. J Thorac Cardiovasc Surg 109: 236-241.[Abstract/Free Full Text]
Morgan A and Burgoyne RD (2004) Membrane traffic: controlling membrane fusion by modifying NSF. Curr Biol 14: R968-R970.
Osten P, Srivastava S, Inman GJ, Vilim FS, Khatri L, Lee LM, States BA, Einheber S, Milner TA, Hanson PI, et al. (1998) The AMPA receptor GluR2 C terminus can mediate a reversible, ATP-dependent interaction with NSF and alpha- and beta-SNAPs. Neuron 21: 99-110.
Patterson A, Zhu W, Chow A, Agrawal R, Kosek J, Xiao R, and Kobilka B (2004) Protecting the myocardium: A role for the beta2 adrenergic receptor in the heart. Crit Care Med 32: 1041-1048.
Pitcher JA, Freedman NJ, and Lefkowitz RJ (1998) G protein-coupled receptor kinases. Annu Rev Biochem 67: 653-692.
Pitcher JA, Payne ES, Csortos C, DePaoli-Roach AA, and Lefkowitz RJ (1995) The G-protein-coupled receptor phosphatase: a protein phosphatase type 2A with a distinct subcellular distribution and substrate specificity. Proc Natl Acad Sci U S A 92: 8343-8347.[Abstract/Free Full Text]
Rochdi MD and Parent JL (2003) Gq-coupled receptor internalization specifically induced by Gq signaling. Regulation by EBP50. J Biol Chem 278: 17827-17837.[Abstract/Free Full Text]
Sibley DR, Strasser RH, Benovic JL, Daniel K, and Lefkowitz RJ (1986) Phosphorylation/dephosphorylation of the β-adrenergic receptor regulates its functional coupling to adenylate cyclase and subcellular distribution. Proc Natl Acad Sci U S A 83: 9408-9412.[Abstract/Free Full Text]
Sneddon WB, Syme CA, Bisello A, Magyar CE, Rochdi MD, Parent JL, Weinman EJ, Abou-Samra AB, and Friedman PA (2003) Activation-independent parathyroid hormone receptor internalization is regulated by NHERF1 (EBP50). J Biol Chem 278: 43787-43796.[Abstract/Free Full Text]
Swaminath G, Xiang Y, Lee TW, Steenhuis J, Parnot C, and Kobilka BK (2004) Sequential binding of agonists to the β2 adrenoceptor. Kinetic evidence for intermediate conformational states. J Biol Chem 279: 686-691.[Abstract/Free Full Text]
Tanowitz M and von Zastrow M (2003) A novel endocytic recycling signal that distinguishes the membrane trafficking of naturally occurring opioid receptors. J Biol Chem 278: 45978-45986.[Abstract/Free Full Text]
von Zastrow M (2003) Mechanisms regulating membrane trafficking of G protein-coupled receptors in the endocytic pathway. Life Sci 74: 217-224.
Whiteheart SW and Matveeva EA (2004) Multiple binding proteins suggest diverse functions for the N-ethylmaleimide sensitive factor. J Struct Biol 146: 32-43.
Whiteheart SW, Rossnagel K, Buhrow SA, Brunner M, Jaenicke R, and Rothman JE (1994) N-ethylmaleimide-sensitive fusion protein: a trimeric ATPase whose hydrolysis of ATP is required for membrane fusion. J Cell Biol 126: 945-954.[Abstract/Free Full Text]
Xiang Y and Kobilka B (2003) The PDZ-binding motif of the β2-adrenoceptor is essential for physiologic signaling and trafficking in cardiac myocytes. Proc Natl Acad Sci U S A 100: 10776-10781.[Abstract/Free Full Text]
Xiang Y, Naro F, Zoudilova M, Jin SL, Conti M, and Kobilka B (2005) Phosphodiesterase 4D is required for β2 adrenoceptor subtype-specific signaling in cardiac myocytes. Proc Natl Acad Sci U S A 102: 909-914.[Abstract/Free Full Text]
Xiang Y, Rybin VO, Steinberg SF, and Kobilka B (2002) Caveolar localization dictates physiologic signaling of β2-adrenoceptors in neonatal cardiac myocytes. J Biol Chem 277: 34280-34286.[Abstract/Free Full Text]
Zhu WZ, Zheng M, Koch WJ, Lefkowitz RJ, Kobilka BK, and Xiao RP (2001) Dual modulation of cell survival and cell death by β2-adrenergic signaling in adult mouse cardiac myocytes. Proc Natl Acad Sci U S A 98: 1607-1612.[Abstract/Free Full Text]
作者单位：Department of Molecular and Integrative Physiology, University of Illinois at Urbana-Champaign, Urbana, Illinois (Y.W., Y.X.); Program in Pharmaceutical Sciences and Pharmacogenomics and Department of Pharmaceutical Chemistry (B.L.), Department of Psychiatry and Department of Cellular and Molecular