Visualization of Dynamic Trafficking of a Protein Kinase C βII/Green Fluorescent Protein Conjugate Reveals Differences in G Protein-coupled Receptor Activation and Desensitization*

Protein kinase C (PKC) links various extracellular signals to intracellular responses and is activated by diverse intracellular factors including diacylglycerol, Ca2+, and arachidonic acid. In this study, using a fully functional green fluorescent protein conjugated PKCβII (GFP-PKCβII), we demonstrate a novel approach to study the dynamic redistribution of PKC in live cells in response to G protein-coupled receptor activation. Agonist-induced PKC translocation was rapid, transient, and selectively mediated by the activation of Gqα- but not Gsα- or Giα-coupled receptors. Interestingly, although the stimuli were continuously present, only one brief peak of PKC membrane translocation was observed, consistent with rapid desensitization of the signaling pathway. Moreover, when GFP-PKCβII was used to examine cross-talk between two Gqα-coupled receptors, angiotensin II type 1A receptor (AT1AR) and endothelin A receptor (ETAR), activation of ETARs resulted in a subsequent loss of AT1AR responsiveness, whereas stimulation of AT1ARs did not cause desensitization of the ETAR signaling. The development of GFP-PKCβII has allowed not only the real time visualization of the dynamic PKC trafficking in live cells in response to physiological stimuli but has also provided a direct and sensitive means in the assessment of activation and desensitization of receptors implicated in the phospholipase C signaling pathway.

The protein kinase C (PKC) 1 family of phospholipid-dependent serine/threonine kinases plays key roles in the transduction and regulation of many cellular signaling processes by catalyzing specific substrate phosphorylation (1,2). Activation of protein kinase C can be triggered by stimulating a wide variety of plasma membrane hormone, neurotransmitter, and growth factor receptors, among which seven-transmembrane G protein-coupled receptors (GPCRs) relay extracellular signals to PKC by activating heterotrimeric guanine nucleotide-binding regulatory proteins (G proteins) (3). G proteins are composed of ␣and ␤␥-subunits, both of which serve as signal moieties and mediate cellular responses by modulating the activity of different effector systems and levels of various second messengers (4). G q ␣ is one of the major G protein ␣-subunits and is coupled to a large number of receptors including angiotensin receptors and endothelin receptors. The activation of G q ␣ by these receptors triggers the hydrolysis of membrane inositol phospholipids by phospholipase C␤ to form two important second messengers, inositol 1,4,5-triphosphate (IP 3 ) and diacylglycerol (DAG) (1,5,6). The binding of IP 3 to its intracellular receptor results in a rise in intracellular Ca 2ϩ (7). The increased membrane DAG and intracellular Ca 2ϩ lead to the mobilization of PKC to the plasma membrane and its subsequent activation (1). The activation of PKC by DAG, Ca 2ϩ , and many other lipid mediators such as arachidonic acid has been associated with many important biological functions including cell proliferation, differentiation, and gene expression (1,2,8).
In G q ␣-coupled receptor signal transduction, PKC activation not only serves to relay the signal from effector phospholipase C␤ to various PKC substrates for initiating downstream cellular responses, but it also exerts feedback effects on the system and is involved in the turning off of signaling from the receptors, an important regulatory process termed desensitization (9,10). The ability of PKC to phosphorylate and desensitize a receptor relies on the existence of consensus PKC phosphorylation site(s) at the intracellular domains of the receptors (11)(12)(13)(14). Since PKC does not discriminate between agonist-occupied and -unoccupied receptors (15), activation of PKC is believed to be associated with both homologous and heterologous desensitization of G protein-coupled receptors. Consequently, exposure of cells to a G q ␣-coupled receptor agonist that causes PKC activation potentially results not only in the loss of cellular response to that specific agonist but also the diminution of cellular responses to various other agonists.
The current methods of assessing G q ␣-coupled receptor signaling and desensitization are based mainly on the generation of IP 3 or the increase of intracellular Ca 2ϩ as a result of IP 3 generation (16 -18). However, neither approach assesses the equally important and independent branch of phospholipase C␤ signaling reflected by the production of DAG and the subsequent membrane translocation and activation of PKC. Although as important as IP 3 , DAG is rarely used as a measurement in assessing G q ␣ activation because of its quick turnover (1) and the lack of techniques to follow its real time cellular distribution or behavior. Moreover, biochemical studies have indicated that the activation of PKC by DAG is also transient and is mainly associated with many cellular events mediated by short term activation of PKC such as hormone secretion and muscle contraction (1,2). Phorbol esters mimic the action of DAG, but they are more potent, and their effects last longer in cells due to their persistence in the cell membrane (1). This has made feasible the study of PKC subcellular localizations before and after activation by phorbol esters by immunofluorescent microscopy in fixed cells (19,20). However, the same method is not applicable in following the more physiologically relevant redistribution of PKC in response to receptor activation in live cells. Furthermore, it is not clear whether the short term of PKC activation in response to receptor activation is related to the signaling desensitization at the level of receptors.
In the present study, we report the development of a green fluorescent protein conjugated PKC␤II (GFP-PKC␤II) to study PKC mobilization in response to agonist stimulation of G protein-coupled receptors and to assess the activation and desensitization of these receptors. GFP, originally identified in the jellyfish Aequorea victoria, displays an inherent green bioluminescence and has been used as a fluorescent reporter molecule in the localization of membrane receptors, cytoplasmic proteins, and secretory proteins (21)(22)(23)(24)(25)(26)(27)(28). When fused to the amino terminus of PKC␤II, the resulting fusion protein, GFP-PKC␤II, was found to be fully functional in terms of its phospholipiddependent kinase activity and its ability to translocate from cytoplasm to the plasma membrane in response to phorbol ester (PMA) stimulation, similar to that reported for a PKC␥/ GFP conjugate (28). Interestingly, while in PMA-treated cells GFP-PKC␤II remained on the plasma membrane, in cells stimulated with physiological signals that activate G q ␣-coupled receptors, the translocation of GFP-PKC␤II to the plasma membrane was found to be transient, reaching a peak and being reversed within minutes. This provided a real time visual demonstration of the cellular trafficking of a PKC isoenzyme in live cells, revealing a dynamic nature of the interaction of PKC␤II with lipid and/or protein molecules on the plasma membrane. Furthermore, since GFP-PKC␤II selectively responded to only signals activating G q ␣but not G s ␣and G i ␣coupled receptors and such responses were transitory, our results also demonstrate an important analytical role of GFP-PKC␤II as a reporter in the study of G q ␣-coupled receptor activation and desensitization.

EXPERIMENTAL PROCEDURES
Materials-[␥-32 P]ATP was purchased from NEN Life Science Products. Monoclonal antibody against GFP was from CLONTECH. Polyclonal rabbit antibody against PKC␤II was prepared and extensively characterized as described previously (29). Mammalian expression vector pBK-CMV and GFP plasmid pEGFP-N1 was from Stratagene and CLONTECH, respectively. Restriction enzymes were from Promega or New England Biolabs. Ampli-Taq DNA polymerase was obtained from Perkin-Elmer. Protein A-Sepharose CL-4B was from Amersham Pharmacia Biotech. Phosphatidylserine and sn-dioctanoyl-glycerol were purchased from Avanti Polar Lipids Inc. Eagle's minimum essential medium, phosphate-buffered saline (PBS), and 1 M HEPES buffer were from Life Technologies, Inc. Fetal bovine serum was obtained from Summit Biotechnology Inc. Fluorescein isothiocyanate-conjugated donkey anti-rabbit IgG was from Jackson ImmunoResearch Laboratories, Inc. All other chemicals were from Sigma.
Plasmid Constructs-The human PKC␤II cDNA was inserted between BamHI and KpnI sites of pBK-CMV. A unique BssHII restriction site was located three base pairs prior to the start codon (ATG) of PKC␤II cDNA. To construct GFP-PKC␤II, EGFP cDNA without the stop codon was first amplified from plasmid pEGFP-N1 by polymerase chain reaction using 5Ј-oligonucleotide primer 5Ј-GTGAACCGTCA-GATCCGCTAG-3Ј (based on the sequence of pEGFP-N1 from 575 to 595) and 3Ј-primer 5Ј-CCATCTTGCGCGCCTTGTACAGCTCGTC-CATGC-3Ј (with the native sequence of pEGFP-N1 from 1376 to 1396 underlined). The polymerase chain reaction fragment containing EGFP cassette was gel-purified, digested with BamHI and BssHII, and directly inserted between the BamHI and BssHII sites of plasmid pBK-CMV-PKC␤II prior to the 5Ј-end of the PKC␤II cDNA (Fig. 1). The sequence of the construct was confirmed by DNA sequencing. The cDNAs of AT 1A R, ET A R, ␤ 2 AR, and D 2 R were subcloned in pcDNA I or pcDNA I/amp mammalian expression vectors (Invitrogen).
Cell Culture and Transfection-HEK 293 cells from the American Type Culture Collection (ATCC) were maintained in Eagle's minimum essential medium supplemented with 10% (v/v) fetal bovine serum in a 5% CO 2 incubator at 37°C. Cells were seeded at a density of 2.0 ϫ 10 6 cells/100-mm dish and transfected using a modified calcium phosphate method with 1-10 g of plasmid (30).
Immunoprecipitation-HEK 293 cells were transfected with EGFP, PKC␤II, or GFP-PKC␤II. Immunoprecipitation was performed 48 h after transfection as follows. The cells were washed with cold PBS and solubilized in 0.5 ml of lysis buffer with protease inhibitors (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% Nonidet P-40, 10 mM NaF, 1 mM sodium orthovanadate, 1 mM dithiothreitol, 10 g/ml leupeptin, 10 g/ml aprotinin, 10 g/ml trypsin/chymotrypsin inhibitor, 5 g/ml pepstatin A, 1 mM phenylmethylsulfonyl fluoride) for 1 h, and the lysates of EGFP, PKC␤II, or GFP-PKC␤II transfected cells were immunoprecipitated with PKC␤II antibody. Protein A-Sepharose beads were used to absorb immunoprecipitates and were then washed four times with lysis buffer followed by one wash with kinase buffer (50 mM Tris-HCl, pH 7.4, 10 mM NaF, 1 mM Na 3 VO 4 , 0.5 mM EDTA, 0.5 mM EGTA, 2 mM MgCl 2 ). The kinase activity of immunoprecipitated proteins was analyzed by protein kinase C assay (see below).
Protein Kinase C Assay-PKC activity was measured using the vesicle assay for PKC as described previously (32). Standard assay conditions were as follows: 20 mM Tris-HCl, pH 7.4, 100 M MgCl 2 , 1 mM CaCl 2 , 10 M ATP, 10 -15 Ci/ml [␥-32 P]ATP, 40 g/ml phosphatidylserine/sn-dioctanoyl-glycerol vesicles, and 200 g/ml histone IIIS as substrate in a final volume of 250 l at 30°C for 10 min. 10 mM EGTA was used in determining basal kinase activity. The phosphorylated proteins were analyzed by SDS-polyacrylamide gel electrophoresis followed by autoradiography.
Indirect Immunofluorescence-Transfected cells were seeded on glass coverslips placed in six-well culture dishes at a density of 5 ϫ 10 5 cells/well. For experiments involving phorbol ester treatment, cells were treated with 100 nM PMA for 5 min. The cells were then rinsed briefly in PBS and fixed in 3.7% paraformaldehyde for 10 min. The fixed cells were permeabilized in PBS containing 0.2% Triton X-100 for 10 min and blocked in PBS containing 0.2% bovine serum albumin for 10 min. After 1 h of incubation with anti-PKC␤II polyclonal antibody (1:100 dilution), the cells were washed and incubated with fluorescein isothiocyanate-conjugated donkey anti-rabbit IgG (1:100 dilution) for 1 h. The coverslips were mounted onto the slides, and PKC␤II immunofluorescence was observed with a Zeiss LSM-410 laser scanning microscope at 488-nm excitation.
Confocal Microscopy-HEK 293 cells were transfected with GFP-PKC␤II and one or two of the G protein-coupled receptors as described in figure legends. 24 h after transfection, the cells were plated onto 35-mm glass-bottomed culture dishes (MatTek) at a density of 1 ϫ 10 5 and incubated for another 24 h for the cells to attach to glass. The cells expressing GFP-PKC␤II (25-40% of the total cell population) were observed under confocal microscopy. Confocal microscopy was performed on a Zeiss LSM-410 laser scanning microscope using a Zeiss 40 ϫ 1.3 NA oil immersion lens. The cells were kept warm during microscopy at 30°C in culture medium containing 20 mM HEPES on a heated microscope stage. GFP-PKC␤II fluorescent signals were collected sequentially using the Zeiss LSM software time series function with single line excitation (488 nm) with a time interval of 20 s between two scannings. Various drugs were applied to the cells during the scanning of GFP-PKC␤II transfected cells.

RESULTS
Comparison of Wild-type PKC␤II and GFP-PKC␤II-Due to its inherent fluorescence and unique compact structure, GFP has been reported in many studies to serve as a valuable reporter molecule in the localization of various proteins without interfering with their biological activity (21, 33). GFP-PKC␤II was constructed by fusing EGFP to the 5Ј-end of PKC␤II (Fig. 1). When examined using SDS-polyacrylamide gel electrophoresis followed by immunoblotting, GFP-PKC␤II was found to express to the same extent in HEK 293 cells as wildtype PKC␤II but migrated more slowly due to the added mass of GFP ( Fig. 2A). To compare the biological activity of GFP-PKC␤II with that of wild-type PKC␤II, both proteins were immunoprecipitated from HEK 293 cells transiently expressing GFP-PKC␤II or PKC␤II with an antibody against PKC␤II and analyzed for their relative kinase activity using a recently developed PKC assay with histone IIIS as the substrate (32). As shown in Fig. 2B, the cells transfected with EGFP contained very low kinase activity, but transfection of either GFP-PKC␤II or PKC␤II significantly increased the phosphorylation of histone IIIS. The ability of GFP-PKC␤II to phosphorylate histone IIIS in a Ca 2ϩ -and phospholipid-dependent manner was comparable with that of wild-type PKC␤II, indicating that the fusion of the GFP molecule to the N terminus of PKC␤II has no significant effect on PKC activity.
Phorbol esters such as PMA are well known to mimic the action of diacylglycerol in inducing PKC to translocate from cytosol to the plasma membrane (1,34). Thus, the ability of GFP-PKC␤II to respond to PMA was also examined (Fig. 2C, top). Under normal unstimulated conditions, GFP-PKC␤II was evenly distributed in the cytoplasm and excluded from the nuclei, a distribution exactly reflecting the cellular localization of wild-type PKC␤II detected with indirect immunofluorescence microscopy (Fig. 2C, middle). Stimulation with 1 M PMA triggered redistribution of GFP-PKC␤II to the plasma membrane. This membrane translocation was completed within 5 min, and about 90% of GFP-PKC␤II fluorescence was mobilized to the plasma membrane, similar to that observed with transfected wild-type PKC␤II in fixed HEK 293 cells (Fig. 2C, top and middle). The distribution of wild-type PKC␤II and GFP-PKC␤II observed in HEK 293 cells corresponded to the reported distribution of endogenous PKC␤II in U937, a cell line expressing endogenous PKC␤II (35). In contrast to GFP-PKC␤II, unconjugated GFP molecules were distributed throughout the cell body, and their distribution was not affected by PMA (Fig. 2C, bottom). These results indicate that GFP-PKC␤II fusion protein retains the biological activity of wild-type PKC␤II with respect to kinase activity as well as cellular localization.
Dynamic and Selective Trafficking of GFP-PKC␤II in Response to G Protein-coupled Receptor Activation-Although the redistribution of PKC upon phorbol ester stimulation has been extensively studied with indirect immunofluorescence microscopy (19,20), the kinetics of PKC membrane translocation in response to physiological signals have not been well characterized. One such physiological signal that activates PKC is through activation of G protein-coupled receptors. Therefore, initial studies using confocal microscopy examined the real time cellular distribution of GFP-PKC␤II in response to the activation of the G q ␣-coupled AT 1A R by its physiological ligand angiotensin II (Fig. 3A). This was done at 30°C in live HEK 293 cells, which had been transiently transfected to overexpress the AT 1A R and GFP-PKC␤II. In the absence of receptor activation, confocal microscopy revealed that GFP-PKC␤II was evenly distributed throughout the cytoplasm. However, upon agonist activation of the AT 1A R, a redistribution of GFP-PKC␤II to the plasma membrane and clearance of cytosolic fluorescence occurred and peaked within 40 s. This angiotensin II-induced GFP-PKC␤II membrane trafficking was rapid compared with the translocation caused by PMA stimulation, which did not peak until after 2-5 min. More interestingly, unlike PMAinduced translocation in which PKC remained persistently localized on the plasma membrane, the mobilization of GFP-PKC␤II to the plasma membrane following AT 1A R activation was transient, and the redistributed GFP-PKC␤II rapidly returned to the cytoplasm within 1 min after translocation. The time frame for GFP-PKC␤II recovery in the cytoplasm in the majority of the cells observed ranged from 20 s to 1 min after GFP-PKC␤II membrane translocation. Moreover, even in the continuous presence of angiotensin II for 30 min, only an initial brief peak of PKC membrane translocation was observed (data not shown). These visual results are consistent with biochemical studies indicating a dynamic nature of PKC␤II in interacting with molecules on the plasma membrane (1,34).
Among major groups of G protein-coupled receptors, those coupled to G q ␣ are directly associated with the activation of phospholipase C, resulting in the generation of diacylglycerol and a rise of intracellular Ca 2ϩ , whereas the effector for receptors coupled to G s ␣ and G i ␣ is mainly adenylyl cyclase associated with the production of cyclic AMP. To examine the specificity and selectivity of agonist-induced PKC trafficking, three additional G protein-coupled receptors were tested for their ability to trigger GFP-PKC␤II translocation, including ET A R (coupled to G q ␣), ␤ 2 AR (coupled to G s ␣), and D 2 R (coupled to G i ␣). In HEK 293 cells, transiently transfected with GFP-PKC␤II and ET A R, a rapid PKC membrane translocation was observed in response to the ET A R endogenous agonist endothelin (Fig. 3B). Similar to that caused by the activation of the AT 1A R, this ET A R-promoted redistribution of GFP-PKC␤II was transient and readily reversible. In contrast, when cells were transfected to coexpress GFP-PKC␤II with either ␤ 2 AR or D 2 R, the activation of neither receptor by its cognate ligand affected the distribution of GFP-PKC␤II fluorescence (Fig. 3, C and D). These results indicate that the mobilization of GFP-PKC␤II in live cells are selectively mediated by receptors activating the phospholipase C signaling pathway including GPCRs coupled to G q ␣ but not G s ␣ and G i ␣.
Desensitization of G q ␣-coupled Receptor Signaling Revealed by GFP-PKC␤II Trafficking-The observation that there was only one brief peak of PKC membrane translocation even in the continuous presence of angiotensin II or endothelin suggests that the transitory nature of PKC trafficking in response to these receptor agonists might be associated with the rapid desensitization of AT 1A R or ET A R signaling. To further investigate this phenomenon, HEK 293 cells transfected with GFP-PKC␤II and AT 1A R were sequentially exposed to two pulses of stimulation by angiotensin II separated by 10 min (Fig. 4A). Whereas the first exposure resulted in a transient GFP-PKC␤II trafficking between the cytoplasm and plasma membrane and GFP-PKC␤II returned to the cytoplasm within 1 min following its membrane translocation, there was no apparent PKC mobilization in response to the second angiotensin II stimulation, indicating the AT 1A R signaling was turned off as a result of initial receptor activation at a step prior to PKC translocation or at the level of PKC. The possibility that PKC itself might be desensitized and loses its ability to respond to further receptor activation was examined by exposing AT 1A R FIG. 1. Architecture of GFP-PKC␤II fusion protein in pBK-CMV vector. As described under "Experimental Procedures," the EGFP fragment without the stop codon was inserted at the 5Ј-end of human PKC␤II (hPKC␤II) cDNA with BamHI and BssHII sites. Also indicated are the numbers of amino acids and expected sizes of EGFP and human PKC␤II. and GFP-PKC␤II-containing cells to PMA following agonist activation of the AT 1A R. As shown in Fig. 4B, PMA triggered a second peak of stable GFP-PKC␤II redistribution to the plasma membrane in cells preexposed to angiotensin II. The time profile and extent of this PKC translocation were indistinguishable from PMA-induced PKC redistribution in cells untreated with receptor agonist. Similar results were obtained when cells were transfected with GFP-PKC␤II and ET A R and pretreated with endothelin (Fig. 4, C and D). These results strongly suggest that the transient nature of PKC trafficking in response to G q ␣-coupled receptor activation is a direct consequence of the rapid desensitization of receptor signaling at a step prior to PKC translocation.
Cross-talk of G q ␣-coupled Receptor Signaling Revealed by GFP-PKC␤II Trafficking-The AT 1A R and ET A R share the with 1 g of pBK-PKC␤II before (control) and after 5 min of treatment with 100 nM PMA (PMA) at 37°C. The cells were fixed and stained for wild-type PKC␤II as described under "Experimental Procedures." The experiments were performed independently on 5-10 cells on two or three different occasions. Bar, 10 m.

FIG. 2. Characterization of GFP-PKC␤II fusion protein expressed in HEK 293 cells by immunoblotting analysis, protein kinase C assay, and confocal microscopy. A, shown are immunoblots of homogenates from HEK 293 cells transfected with 1 g of pEGFP-N1 (lane 1), pBK-CMV-GFP-PKC␤II (lane 2), and pBK-CMV-PKC␤II (lane 3) probed with polyclonal antibody against PKC␤II (left)
and monoclonal antibody against GFP (right). 20 g of homogenate was loaded in each lane. As expected, the homogenate from cells transfected with pBK-CMV-GFP-PKC␤II showed strong immunoreactivity to both the PKC␤II antibody and GFP antibody. The band seen at approximate 50 kDa in each lane with anti-GFP represents nonspecific immunoreactivity. B, the above described HEK 293 cell homogenates (100 g) were immunoprecipitated with PKC␤II antibody, and kinase activity of the immunocomplexes was assayed. Protein kinase C assay was performed under standard conditions in the presence (ϩ) and absence (Ϫ) of PKC activator sn-dioctanoyl-glycerol (DiC8), and cofactors Ca 2ϩ and phosphatidylserine. Shown is a representative autoradiograph indicating GFP-PKC␤II and wild-type PKC␤II comparably phosphorylated substrate histone IIIS. C, shown are the confocal micrographs of HEK 293 cells transfected with 1 g of pBK-CMV-GFP-PKC␤II (top) or pEGFP-N1 (bottom) before (control) and after 5 min of PMA treatment (PMA). The transfected cells were stimulated with 1 M PMA at 30°C to activate PKC. The redistribution of GFP-PKC␤II started within 1 min, and almost 90% of fluorescence was located at the plasma membrane after 5 min of stimulation. In contrast, there was no change of the distribution of GFP itself in response to 1 M PMA stimulation. Shown in the middle are the confocal micrographs of HEK 293 cells transfected

FIG. 3. Confocal microscopy of the time-dependent GFP-PKC␤II redistribution in response to G protein-coupled receptor activation.
HEK 293 cells were transfected with 1 g of pBK-CMV-GFP-PKC␤II as well as 10 g of one of the following G protein-coupled receptor constructs: AT 1A R, ET A R, ␤ 2 AR, or D 2 R. All cells were visualized at 30°C with a confocal microscope. Before receptor activation (control), GFP-PKC␤II was evenly distributed in the cytoplasm. When cells transfected with AT 1A Rs or ET A Rs were stimulated with corresponding receptor agonists, 0.5 M angiotensin II (A) or 0.1 M endothelin (B), GFP-PKC␤II underwent a transient redistribution between the cytoplasm and plasma membrane. In contrast, no GFP-PKC␤II mobilization was observed in cells transfected with either ␤ 2 AR (C) stimulated with 10 M isoproterenol or D 2 R (D) stimulated with 10 M dopamine. Also indicated in the micrographs are the time points following agonist stimulation for each receptor. The experiments were performed independently on three different occasions, and each time 3-5 cells from independent stimulation by each agonist were recorded. All of the micrographs are representative of more than 70% of the cells observed. Bar, 10 m. same G q ␣-mediated signaling pathway in which G q ␣ activates phospholipase C␤ and results in the hydrolysis of phosphoinositol lipid to generate IP 3 and DAG, the latter serving as a second messenger for activation of PKC. To further study the rapid receptor desensitization as well as the relationship between signaling pathways of different receptors, AT 1A R and ET A R were cotransfected with GFP-PKC␤II into HEK 293 cells, and the effect of their potential cross-talk on the trafficking of GFP-PKC␤II was examined and visualized by confocal microscopy. In initial experiments, the cells were prestimulated with angiotensin II to activate the AT 1A R and induce PKC response. As described above, the resulting GFP-PKC␤II trafficking was transient and GFP-PKC␤II returned to the cytoplasm within 1 min following its membrane translocation. No second peak of PKC translocation was observed in the continuous presence of agonist for as long as 30 min. However, when endothelin was subsequently added to the angiotensin II-pretreated cells to activate the ET A R, a second peak of GFP-PKC␤II translocation was apparent within 1 min, the time profile and extent of which were indistinguishable from the first peak (Fig. 5A). As both AT 1A R and ET A R induce PKC response by activating G q ␣, which increases the activity of phospholipase C, the ability of endothelin to mobilize PKC in cells prestimulated with AT 1A R agonist further indicates that the activation of AT 1A R did not desensitize G q ␣-mediated signaling pathways, but instead the desensitization occurred at the level of the AT 1A R receptor itself. Moreover, these results also demonstrate that the activation of AT 1A Rs does not contribute to the desensitization of the ET A R.
To further investigate the effect of ET A R activation on AT 1A R signaling, we reversed the order of introducing agonists and sequentially added endothelin and then angiotensin II to the cells cotransfected with both receptors. Surprisingly, when the cells were preexposed to endothelin for 5 min, no additional peak of GFP-PKC␤II cellular movement was detected by confocal microscopy in response to a subsequent stimulation with angiotensin II (Fig. 5B). This result indicates that in contrast to the inability of AT 1A R to cause ET A R desensitization in HEK 293 cells, the activation of ET A Rs indeed led to the heterologous desensitization of the AT 1A R in the same cells, possibly mediated by the activation of PKC and subsequent phosphorylation of the receptor by PKC. This is consistent with the observation that inhibition of activation of PKC by 0.5 M staurosporine, a potent protein kinase inhibitor, resulted in a 42% lose of total agonist-stimulated phosphorylation of the AT 1A R but had no effect on that of the ET A R (16,36). The specificity of the heterologous desensitization was further assessed by examining the effect of activation of another second messenger-dependent kinase, cyclic AMP dependent protein kinase (PKA) on AT 1A R signaling in cells cotransfected with the ␤ 2 AR and AT 1A R (Fig. 5C). The activation of the ␤ 2 AR by isoproterenol is easily detectable by measuring the membrane translocation of a ␤-arrestin 2/GFP conjugate (26) and is known to stimulate G s ␣, which eventually leads to the accumulation of cyclic AMP and activation of PKA (3). As expected, stimulation of ␤ 2 AR by isoproterenol did not have a visible effect on the distribution of GFP-PKC␤II. However, this preexposure of the cells to isoproterenol did not prevent the rapid cellular redistribution of GFP-PKC␤II between the cytoplasm and plasma membrane triggered by subsequently added angiotensin II, indicating that activation of PKA did not result in the heterologous desensitization of the AT 1A R. DISCUSSION In the present work, the development of a fully functional green fluorescent protein conjugated PKC␤II has allowed visualization in live cells of the real time interaction of PKC with the plasma membrane in response to extracellular signals. Whereas stimulation by phorbol esters (e.g. PMA) causes a persistent localization of PKC to the plasma membrane, our results reveal a dynamic nature of PKC␤II trafficking between the cytoplasm and plasma membrane in response to physiological signals such as those activating G protein-coupled receptors. PKC␤II responds selectively only to specific signals that activate receptors coupled to G q ␣ but not G s ␣ and G i ␣ proteins. The membrane translocation of PKC␤II triggered by the activation of G q ␣-coupled receptors is rapid and transient and is followed immediately by the returning of PKC to the cytoplasm within minutes, indicating a desensitization of the signaling pathway. Moreover, when GFP-PKC␤II trafficking in response to the sequential activation of two distinct G q ␣-coupled receptors (i.e. AT 1A R and ET A R) was studied, the results indicated that the signaling desensitization occurs at the level of receptors and that there is cross-talk between the two receptors. Thus, GFP conjugated PKC fusion proteins serve as a novel useful tool not only for studying the dynamic localization of PKCs in signal transduction in live cells but also for detecting the activation and desensitization of receptors coupled to phospholipase C such as G q ␣-coupled receptors.
PKC cellular localization has been extensively studied in culture cells using antibody staining and immunofluorescent microscopy (19,20). However, many signal transduction events involving PKC are rapid, transient, and difficult to follow in fixed cells. Green fluorescent protein, because of its inherent bioluminescence and stoichiometric labeling, represents a sensitive optical reporter to follow the real time localization of many proteins in live cells and is thus well suited for the study of transitory and dynamic distribution of molecules in the process of signal transduction (26). Recently, the GFP technique has been used in the study of cellular distribution of PKC␥ (28). In our study, when EGFP was fused to the N terminus of PKC␤II, GFP-PKC␤II conjugate displayed Ca 2ϩ and phospholipid-dependent kinase activity comparable with that of wild-type PKC␤II and was localized mainly in the cytoplasm and excluded from the nuclei. Moreover, in cells stimulated with PMA, like wild-type PKC␤II, GFP-PKC␤II redistributed to the plasma membrane. The stable association of GFP-PKC␤II with the plasma membrane is a true reflection of the persistence of PMA in the plasma membrane as well as the stable interaction of PKC␤II to PMA. Therefore, although GFP is about one-third the size of PKC␤II (Fig. 1), as reported for many other proteins, our data demonstrate that GFP-PKC␤II conjugate retains the function of native PKC␤II in terms of biochemical behavior and cellular localization, although the possibility cannot be ruled out that the coupling of GFP to PKC␤II may alter the affinity of the enzyme to its interacting molecules.
The activation of PKC is triggered by a large number of extracellular signals including hormones, neurotransmitters, and growth factors that act through cell surface receptors. The activation of these receptors regulates the intracellular level of various PKC activators including DAG, Ca 2ϩ , and many other lipid mediators. G protein-coupled receptors were used in this work as an example to address PKC cellular trafficking and distribution in response to various physiological stimuli. Four receptors were examined for their ability to stimulate GFP-PKC␤II redistribution, including G s ␣-coupled ␤ 2 AR, G i ␣-coupled D 2 R, and G q ␣-coupled AT 1A R and ET A R. The physiological relevance of PKC activation mediated by these receptors is apparent since PKC responded selectively only to signals activating G q ␣-coupled receptors (i.e. AT 1A R and ET A R) but not G s ␣and G i ␣-coupled receptors, corresponding to the fact that among the three major G proteins only G q ␣ mediates the production of DAG at the plasma membrane and a rise of intracellular Ca 2ϩ . More importantly, unlike PMA-stimulated PKC translocation, physiological signals activating AT 1A Rs and ET A Rs induce a redistribution of GFP-PKC␤II to the plasma membrane, which appears to be transient, consistent with the FIG. 5. Cross-talk of G protein-coupled receptor signaling revealed by GFP-PKC␤II trafficking. A, effect of AT 1A R activation on the subsequent stimulation of ET A Rs by endothelin. HEK 293 cells were transfected to coexpress GFP-PKC␤II, AT 1A R, and ET A R. The cells were stimulated with 0.5 M angiotensin II (AgII), which induced GFP-PKC␤II redistribution between the cytoplasm and plasma membrane. This was followed by a second stimulation with 0.1 M endothelin (ET) 5 min after the first stimulation. A second peak of GFP-PKC␤II redistribution was observed in response to ET A R activation. B, effect of ET A R activation on the subsequent stimulation of AT 1A Rs by angiotensin II. The cells were transfected as described in A and stimulated with 0.1 M endothelin, which induced GFP-PKC␤II redistribution between the cytoplasm and plasma membrane. This was followed by a second stimulation with 0.5 M angiotensin II 5 min after the first stimulation. The preexposure of cells to endothelin blocked the subsequent activation of AT 1A Rs, as revealed by the lack of GFP-PKC␤II redistribution. C, effect of ␤ 2 AR activation on the subsequent stimulation of AT 1A Rs by angiotensin II. HEK 293 cells were transfected to coexpress GFP-PKC␤II, AT 1A R, and ␤ 2 AR. The cells were stimulated with 10 M isoproterenol to activate ␤ 2 ARs, and this was followed by a second stimulation with 0.5 M angiotensin II 5 min after the first stimulation. GFP-PKC␤II redistribution was observed following stimulation by angiotensin II, indicating activation of the AT 1A R. Also indicated in the micrographs are the time points following the first agonist stimulation. All cells were visualized at 30°C by confocal microscope. The experiments were performed independently on three or four different occasions, and each time 10 -15 cells from independent stimulation by different agonists were recorded. All of the confocal micrographs are representative of more than 60% of the cells observed. Bar, 10 m.
biochemical studies indicating that DAG is one of the plasma membrane stimuli for PKC and that the interaction between DAG and PKC is rapid and reversible (1,34).
However, it is somewhat unexpected that only one rapid cycle of PKC translocation from the cytoplasm to plasma membrane and back was observed although receptor agonists were present continuously. This lack of further PKC responsiveness is probably the result of a rapid desensitization of the agonistmediated signaling pathway upstream of PKC, since PKC itself still retains the ability to respond normally when subsequently exposed to PMA. In addition, sequential stimulation of the AT 1A R and then the ET A R with their corresponding agonists induced consecutive transient cycles of PKC translocation, one after each stimulation. As the AT 1A R and ET A Rs presumably share the same signaling components downstream of the receptors, this suggests that the desensitization of AT 1A R signaling occur at the level of the receptor itself. More interestingly, although activation of the AT 1A R does not affect the activity of the ET A R, activation of the ET A R not only shuts off its own signaling but also causes the desensitization of the AT 1A R, indicating differential regulation of the two receptors.
The rapid desensitization of G protein-coupled receptors is achieved mainly through phosphorylation of the receptors by two classes of serine/threonine protein kinases: the second messenger activated protein kinases, PKA and PKC; and the G protein-coupled receptor kinases that specifically phosphorylate agonist-activated receptors (15). Although the role of PKC in regulating AT 1A R desensitization in different tissues is still variable depending on experimental conditions, recently it was reported that the AT 1A R was phosphorylated by both G protein-coupled receptor kinases and PKC in response to short term angiotensin II stimulation in HEK 293, a cell line widely used in the study of receptor desensitization (16). In contrast, in the same cell line, the ET A R was mainly phosphorylated by G protein-coupled receptor kinases but not by activated PKC in response to agonist (36). These studies suggest that while both receptors serve as substrates for G protein-coupled receptor kinases, only the AT 1A R (but not ET A R) has the unique biochemical property to undergo agonist-dependent phosphorylation by PKC. The homologous desensitization of both receptors probably involves G protein-coupled receptor kinase or possibly PKC in the case of the AT 1A R. However, our results suggest that cross-talk (i.e. heterologous desensitization) between the two receptors is mainly mediated by PKC, and the distinct ability to be phosphorylated by PKC might underlie the observed differential desensitization properties of the two receptors. For instance, the inability of the ET A R to be phosphorylated by PKC might account for its lack of heterologous desensitization by activation of the AT 1A R. In addition, despite the finding that the AT 1A R was phosphorylated by PKA in intact aortic vascular smooth muscle cells (RASM) (37), we demonstrate that activation of PKA by stimulating the ␤ 2 AR does not lead to desensitization of the AT 1A R, consistent with the lack of PKA phosphorylation of the AT 1A R found in HEK 293 cells (16).
A large variety of signaling pathways are known to culminate in the activation of PKC, including those mediated by G q ␣ protein-coupled receptors (1,34). The numbers of such receptors are expected to expand rapidly with the progress of genomic sequencing and challenge the conventional biochemical measurements that assess individual receptor-specific properties for defining their corresponding ligands, detecting signaling activation and measuring change of second messenger levels. In this study, by combining the inherent fluorescence of GFP with the translocation property of PKC, we have developed a potential live cell biosensor that may provide sim-ple, sensitive, and rapid assessment of the involvement of PKC activation in the signaling pathways of these receptors. In addition, since GFP-PKC redistribution can serve as a sensitive indicator for receptor activation, it may provide a simple and universal tool for screening new ligands for receptors coupled to PKC as well as for associating newly discovered receptors with their cognate ligands and physiological functions. Moreover, by monitoring inhibition of PKC translocation, it may be potentially also applicable to the identification of the inhibitors of PKC itself as well as inhibitors that block the signaling components leading to activation of PKC. In the case of GPCRs, GFP-PKC translocation has been shown in this study to be a specific measure of G q ␣-coupled receptor activation. The transitory nature of the translocation makes GFP-PKC a useful tool for studying G q ␣-coupled receptor desensitization. Compared with detecting Ca 2ϩ signals, measuring PKC trafficking represents a more direct and accurate assessment of the properties of the plasma membrane receptors without concerns from the permeability, solubility, and compartmentalization of Ca 2ϩ indicators and interference from plasma membrane and intracellular Ca 2ϩ channels (e.g. the IP 3 receptor) (7,18,38). Furthermore, with the discovery of numerous novel G protein-coupled receptors by genomic sequencing, GFP-PKC should also be extremely useful in quickly identifying those receptors coupled to G q ␣ and their ligands and functions.
The visualization of GFP-PKC␤II dynamic translocation in this study provides a direct real time assessment of the distribution of a PKC in live cells in response to changes of intracellular PKC activators triggered by physiological stimuli. To date, 12 members are identified as belonging to the PKC superfamily, associated with a wide variety of cellular signaling events, such as mitogenesis and tumorigenesis (2,8). The use of GFP conjugates as optical reporters should provide valuable information concerning not only the specific cellular distribution of different PKC isoenzymes but also their dynamic trafficking in response to various physiological stimuli. Therefore, GFP-PKC conjugates may represent ideal optical tools in the study of specific functions and kinetics of each PKC isoenzyme in different signal transduction systems. Furthermore, when employed as a biosensor, GFP-PKC fusion proteins may also provide a unique and sensitive means for studying the kinetics and components of signal transduction pathways in which PKCs are involved.