Abstract
The regulation of the cellular distribution and intracellular signaling properties of the α1B- and α1D- adrenoceptor (α1-AR) subtypes was examined in stably transfected Rat 1 fibroblasts. In unstimulated cells, α1B-AR expression was noted primarily on the cell surface. Treatment with phenylephrine induced internalization of the α1B-AR and promoted association with arrestin 2. The internalized α1B-AR colocalized with the transferrin receptor, an endosomal marker. In unstimulated fibroblasts, the α1D-AR was detected in a perinuclear orientation and was colocalized with arrestin 2 in a compartment also containing the transferrin receptor. After treatment with prazosin, which exhibits inverse agonist properties, the α1D-AR was redistributed from intracellular sites to the cellular periphery and was no longer associated with the transferrin receptor or arrestin 2. α1D-AR-expressing cells exhibited a high degree of basal activity for both inositol phosphate formation and extracellular signal regulated kinase (ERK), which was reduced by treatment with prazosin. In these cells, phenylephrine induced a dose-dependent increase in inositol phosphate formation but had no effect on ERK activity. In α1B -AR-expressing cells, phenylephrine stimulated both inositol phosphate formation and ERK activity. These data show that: 1) there are differences in the cellular localization of the α1-AR subtypes; 2) the α1B-AR exhibits expected G protein-coupled receptor activity regarding cellular localization, agonist-mediated internalization, and coupling to second messengers; and 3) the α1D-AR is constitutively active and, as a result, is localized to intracellular compartments involved in receptor recycling.
Three subtypes of the α1-adrenoceptor (AR), the α1A, α1B, and α1D, have thus far been identified, cloned, and characterized (Cotecchia et al., 1988; Schwinn et al., 1990; Lomasney et al., 1991; Perez et al., 1991). Our understanding of the complexity of the α1-ARs and their interplay with other signaling systems has steadily increased. The α1-ARs are major effectors used by the sympathetic nervous system to regulate systemic arterial blood pressure, blood flow, and cell growth. Previous work by us and others has shown that the mRNA encoding each of these receptors is widely distributed in the peripheral vasculature (Ping and Faber, 1993;Piascik et al., 1994, 1995; Eckhart et al., 1996; Guarino et al., 1996) and that the α1-ARs are also expressed as protein in all of the peripheral arteries thus far examined (Piascik et al., 1997; Hrometz et al., 1999). Using both pharmacological approaches and antisense oligonucleotide technology, we have found evidence that the activation of smooth muscle contraction is caused by a single receptor in any given artery and that the α1-AR subtype that mediates contraction varies throughout the vasculature (Piascik et al., 1995, 1997; Hrometz et al., 1999).
These findings have lead us to explore possible mechanisms to account for the observation that expression in blood vessels is not sufficient to link an α1-AR to contraction. We have focused on the subcellular localization of the α1-ARs as a factor that modifies the ability of these receptors to regulate cellular function. In models of cellular signaling, the heptahelical G protein-coupled receptors (GPCRs) are postulated to reside in the cell membrane where they transduce a variety of intracellular signals initiated by binding of neurotransmitters or exogenously administered pharmacologic agents. Recent evidence has indicated that the α-ARs are not exclusively localized with the cell membrane. Using green fluorescent protein/α1-AR fusion proteins, Hirasawa et al. (1997) showed that the α1B-AR was expressed on the cell surface whereas the α1A-AR was expressed in intracellular compartments. In transfected fibroblasts, the α2C-AR was detected in intracellular compartments as well as on the cell surface, whereas the α2A-AR was found exclusively on the cell membrane (von Zastrow et al., 1993; Daunt et al., 1997).
Internalization of GPCR after activation by agonists is a well known phenomenon. Agonist-induced internalization of the GPCR regulates signal transduction by this class of transmembrane receptors. Like constitutively recycling receptors for transferrin and the low-density lipoprotein, GPCRs are internalized via clathrin-coated pits to endosomes (von Zastrow and Kobilka, 1992; reviewed by Krupnick and Benovic, 1998). In addition to activation of cellular signaling cascades, agonist occupancy of GPCR results in the recruitment of a GPCR kinase to the membrane, receptor phosphorylation, and promotion of arrestin binding to the receptor (Krupnick and Benovic, 1998). Arrestin binding is thought to uncouple the receptor from G proteins and initiate the process of internalization by directing the receptor to the clathrin-coated pits. This process has traditionally been thought of as a way of terminating cellular signaling. However, studies using transfected β2-ARs or endogenous opioid receptors indicate that internalization of certain GPCR also activates cellular signaling. Daaka et al. (1998) and Ignatova et al. (1999)showed that receptor internalization was necessary to activate extracellular signal-regulated kinase (ERK). In these studies, expression of dominant negative mutants of β-arrestin or dynamin prevented receptor internalization and receptor-stimulated ERK activity.
In this report we have examined the regulation of the cellular localization and signaling properties of the α1B- and α1D-ARs. We show that the α1B-AR is expressed primarily on the cell surface, is internalized by agonist occupation, and is coupled to inositol phosphate formation and ERK activity, consistent with traditional views of GPCR. The α1D-AR exhibits behavior atypical of a GPCR, and we postulate that this is due to the constitutive activity exhibited by this receptor.
Materials and Methods
Cell Culture Conditions.
Rat 1 fibroblasts stably transfected with cDNA for the human α1B-AR or α1D-AR (a gift from Glaxo Wellcome) were cultured in Dulbecco's modified Eagle's medium (DMEM; Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum (Life Technologies), a 1% antibiotic/antimycotic mixture (Life Technologies), and 500 μg/ml geneticin (Life Technologies). Cells were maintained in cell culture flasks at 37°C in 5% CO2, fed every 2 to 3 days with supplemented DMEM, and trypsinized every 5 to 7 days. After the fibroblasts reached confluency, they were plated out on sterile 20-mm × 20-mm glass coverslips and returned to the cell culture incubator for 72 h to allow attachment. Cells were deprived of serum overnight (18–24 h) before experimentation.
Immunocytochemistry and Laser Scanning Confocal Microscopy.
Cells were fixed with 3.7% formaldehyde in PBS for 10 min, washed with PBS containing 0.05% BSA, and permeabilized with 0.1% Triton in PBS for 5 min. Cells were incubated in blocking solution containing 10% fetal bovine serum in PBS for 1 h at room temperature, then washed and incubated with primary antibody for 2 h. The primary antibodies (against the α1-ARs, the transferrin receptor, and arrestin 2) are described in a section below. Fibroblasts were then incubated with either a fluorescein-conjugated (donkey anti-goat IgG, for α1-AR visualization) or a rhodamine-conjugated (donkey anti-rabbit or -mouse IgG, for arrestin or transferrin receptor visualization) secondary antibody. Dual-label immunocytochemistry was used to simultaneously assess the localization of the α1-ARs, arrestin, or transferrin. In certain experiments, the effect of phenylephrine or a series of antagonists on the cellular location of proteins of interest was assessed. We also examined α1B-AR immunoreactivity in COS 1 cells transiently transfected with the wild-type α1B-AR or a constitutively active mutant in which alanine 293 was mutated to lysine (A293K, Kjelsberg et al., 1992). Images were analyzed using a laser-scanning confocal microscope (Leica TCS, Exton, PA) with a 63× Water Immersion objective. Fluorescein isothiocyanate (FITC) fluorescence was excited by an argon laser at 488 nm and monitored at a wavelength no greater than 545 nm. Rhodamine fluorescence was excited using a 568-nm helium-neon laser and monitored at a wavelength no greater than 690 nm.
Quantitation of Intracellular Inositol Phosphates.
Rat 1 fibroblasts plated on 60-mm culture plates were grown in DMEM (Bio-Whittaker, Walkersville, MD) supplemented with fetal bovine serum. On reaching 90% confluency, 3 μCi of3H-myo-inositol (DuPont-NEN, Boston, MA) was added 24 h before use in experiments. Measurement of intracellular inositol phosphates was performed in serum-free DMEM after several changes of media. Assays were conducted in the presence of LiCl (10 mM) in a total assay volume of 3 ml. Cells were incubated at 37°C for 60 min in a 5% CO2 atmosphere in the presence or absence of increasing concentrations of phenylephrine. The effect of 1 μM prazosin, an inverse agonist, which was added 24 h before experimentation was also determined. Incubations were terminated by removal of the media and addition of 1 ml of a 0.4 M perchloric acid solution. The cell lysates were scraped, collected, and neutralized by the addition of 0.5 ml of a 0.72 N KOH/0.6 M KHCO3 solution. Soluble inositol phosphates in the cell lysates were isolated by passage through a Bio-Rad AG 1X-8 resin column that was buffered with a 0.1 M formic acid solution. After washing the column with 0.1 M formic acid, bound3H-IPs were displaced from the column by eluting with a 0.1 M formic acid solution containing 1 M ammonium formate and the radioactivity was detected by liquid scintillation counting. Data were analyzed using a one-tailed t test analysis (GraphPad Prism).
Assays for ERK Activity.
ERK activity was determined using in-gel kinase assays as we have described previously (Post et al., 1996). Briefly, fibroblasts were incubated with phenylephrine in the presence or absence of prazosin for the indicated times, washed with ice-cold PBS, and scraped into 1 ml of ice-cold buffered sucrose. The cell suspension was centrifuged (5 min, 1000g, 4°C), and the pellet was resuspended in 20 μl of lysis buffer (20 mM Tris-HCl, 250 mM NaCl, 2.5 mM EDTA, 3 mM EGTA, 20 mM β-glycerophosphate, 0.5% NP-40, 100 μM Na3VO4, 5 μM AEBSF (Calbiochem, La Jolla, CA), 1.5 nM aprotinin, 10 nM E-64, 10 nM leupeptin, pH 7.4). After incubation on ice for 30 min, the lysate was centrifuged (15 min, 15,000g, 4°C), the supernatant was collected, and protein content was determined. Protein was resolved on 10% SDS-polyacrylamide gels containing 0.5 mg/ml myelin basic protein (MBP). After electrophoresis, gels were first washed with 20% 2-propanol in 50 mM HEPES, pH 7.6 and then with 5 mM β-mercaptoethanol in HEPES buffer. Proteins were denatured by washing gels in 6 M urea and then renatured in HEPES buffer containing 0.05% (v/v) Tween-20 (renaturation buffer) at 4°C. After overnight incubation in renaturation buffer at 4°C, gels were preincubated in 25 ml of cold kinase buffer (20 mM HEPES, 20 mM MgCl2, 2 mM DTT, 5 mM β-glycerophosphate, 0.1 mM Na3VO4) for 30 min. Phosphorylation of MBP was performed in situ by incubating gels in kinase buffer containing 20 μM ATP and 150 to 160 μCi [γ32P]ATP for 90 to 120 min at 30°C. After extensive washing in 5% trichloroacetic acid/1% sodium pyrophosphate, gels were dried and exposed to film. 32P incorporation into MBP was quantified by densitometry. The ERK activity reported in Table 1 is in integrated optical density units, whereas those presented in Fig. 4 are normalized to a percentage of the control level of ERK activity obtained in each cell line. Data were analyzed by a one-way ANOVA.
Antibodies and Reagents.
The primary antibodies used were as follows: goat anti-human α1B IgG at 1:100 to 1:200, goat anti-rat α1D IgG at 1:25 to 1:50, (Santa Cruz Biotech, Santa Cruz, CA) and mouse anti-rat CD71, a marker for the transferrin receptor, at 1:100 (Research Diagnostics, Flanders, NJ). Arrestin antibodies were provided by Dr. Jeffrey Benovic (Thomas Jefferson University). All secondary antibodies were used at a dilution of 1:100, which included fluorescein isothiocyanate (FITC)-conjugated donkey anti-goat IgG, rhodamine-conjugated donkey anti-rabbit IgG, and rhodamine-conjugated donkey anti-mouse IgG (Jackson ImmunoResearch Labs, West Grove, PA). Prazosin and phenylephrine were purchased from Sigma (St. Louis, MO).
Results and Discussion
In this report we have examined the regulation of the cellular distribution of the α1B- and α1D-ARs in Rat 1 fibroblasts stably transfected with either subtype. We have previously used antibodies to examine the distribution of the α1-ARs in blood vessels, vascular smooth muscle cells, and stably transfected fibroblasts (Hrometz et al., 1999). In these studies we showed that the antibodies are specific and stain only those cells expressing the receptor against which the antibody was raised. Additionally, we show that preabsorption of the antibody with its immunizing peptide significantly reduces immunostaining.
In radioligand binding studies we determined that the α1B- and α1D-AR were expressed at similar levels in the Rat 1 fibroblast cell lines used in these studies (between 5.5 and 10 pmol/mg protein; data not shown). In α1B-AR-expressing fibroblasts, receptor immunoreactivity was detected on the margin of the cell, indicating that, as expected, this receptor was localized to the cell membrane (Fig. 1). In contrast, although there was α1D-AR immunoreactivity on the boundary of the cell, a significant degree of immunostaining was also detected in a perinuclear orientation (Fig. 1).
Figure 2 shows the immunostaining pattern obtained in fibroblasts expressing the α1B-AR after treatment with the α1-AR selective agonist phenylephrine. Agonist activation promotes a significant degree of α1B-AR internalization as evidenced by the enhanced FITC fluorescence detected in a perinuclear orientation (Fig.2, panel 1). Internalization can be detected within 5 min of receptor activation (see Fig. 2, panel 4 for time course) and is blocked by prior treatment of the cells with prazosin. Isoproterenol activation of the β2-AR promotes receptor internalization to endosomes (von Zastrow and Kobilka, 1992). We determined whether this is also true for the α1B-AR by performing dual-label immunofluorescence studies using an antibody against the transferrin receptor (CD71 in the figures), an endosomal marker. In these photomicrographs, the receptor is detected by a FITC-labeled secondary antibody, which emits a green color. The transferrin receptor is detected with a rhodamine-labeled secondary antibody that appears red (Fig. 2, panel 5). When these secondary antibodies are colocalized, the resulting fluorescent image appears yellow. Dual antibody labeling after activation of the α1B-AR with phenylephrine yields a perinuclear yellow fluorescence, indicating that the internalized α1B-AR is localized to endosomes (Fig. 2, panel 2). This colocalization and color change were prevented by pretreatment with prazosin.
Arrestin binding to phosphorylated GPCR is thought to uncouple the receptor from its cognate G protein(s) and participate in the process of internalization by directing the receptor to clathrin-coated pits. Dual-label immunocytochemistry shows that in unstimulated Rat 1 fibroblasts, there is little colocalization of the α1B-AR with arrestin 2 (detected with a rhodamine-labeled secondary antibody, see Fig. 2, panels 3 and 5). Agonist mediated α1B-AR internalization promotes the association of the receptor with arrestin 2 (see Fig. 2, panel 3). Receptor internalization and arrestin colocalization were blocked by prior treatment with prazosin. Detection of agonist-stimulated internalization of the α1B-AR into endosomes by confocal microscopy has been reported previously in stably transfected HEK 293 cells (Fonseca et al., 1995) and cells expressing an α1B-AR/green fluorescent protein fusion protein (Awaji et al., 1998). Association of constitutively active α1B-AR mutants with arrestin 2 has also been demonstrated (Mhaouty-Kodja et al., 1999).
The functional response to α1B-AR activation was assessed by measuring inositol phosphate formation and ERK activity in stably transfected fibroblasts. Phenylephrine stimulated the formation of inositol phosphates and ERK activity in cells expressing the α1B-AR (see Figs.3 and 4).
Immunocytochemical analysis and cell signaling studies in Rat 1 fibroblasts expressing the α1D-AR revealed a distinctly different cellular localization pattern and functional responses than those observed for the α1B-AR. Although some α1D-AR immunoreactivity was noted on the boundary of the cell, a significant degree of immunostaining was detected in a perinuclear orientation (see Fig. 1). The α1D-AR colocalizes with the transferrin receptor, indicating that in unstimulated fibroblasts, the α1D-AR is localized to endosomes (see Fig.5, panel 2). In unstimulated fibroblasts, the α1D-AR was also localized with arrestin 2 (see Fig. 5, panel 3). The control transferrin and arrestin immunoreactivities in α1D-AR-expressing fibroblasts were not different from those obtained in α1B-AR-expressing cells as shown in Fig. 2.
Fibroblasts expressing the α1D-AR exhibited a high degree of basal inositol phosphate formation compared with α1B-AR-expressing cells (see Table 1). Phenylephrine was capable of inducing an additional dose-dependent increase in inositol phosphate formation in α1D-AR-expressing fibroblasts (Fig. 3). Basal ERK activity in α1D-AR-expressing fibroblasts was also greater than that seen in α1B-AR-expressing cells. However, we did not detect phenylephrine-induced increases in ERK activity above basal levels in α1D-AR-expressing cells (Fig. 4). Interestingly, basal ERK activity was similar in magnitude to serum-stimulated levels in α1D-AR-expressing cells. It may be that the high basal levels of kinase activity we observed precluded any additional agonist-induced increases in the α1D-AR-expressing fibroblasts.
In unstimulated cells, we noted that the α1D-AR is localized with arrestin and endosomes, the same cellular locale as the α1B-AR after agonist activation. In addition, fibroblasts expressing the α1D-AR exhibit a high level of basal phospholipase C and ERK activity compared with the α1B-AR. These data argue that the α1D-AR is constitutively active in Rat 1 fibroblasts. To more fully substantiate that the α1D-AR is constitutively active, we performed immunocytochemistry, inositol phosphate, and ERK assays after culturing α1B- or α1D-AR-expressing fibroblasts in the presence of prazosin for 24 h. Prazosin has recently been shown to have inverse agonist properties (Scheer et al., 1997). Treatment of α1B-AR expressing fibroblasts with prazosin had no effect on the cellular localization of the α1B-AR (data not shown). Likewise, prazosin did not decrease either basal inositol phosphate formation or ERK activity (see Table 1) in α1B-AR expressing cells. Treatment of α1D-AR-expressing fibroblasts with prazosin caused a significant degree of redistribution of the α1D-AR from a perinuclear orientation to the cell periphery (Fig. 5, panel 1). After prazosin treatment, the α1D-AR was no longer colocalized with either the transferrin receptor or arrestin 2 (see Fig. 5, panels 2 and 3). Treatment of fibroblasts expressing the α1D-AR with prazosin also decreased basal levels of inositol phosphate formation and ERK activity (see Table 1). The cellular localization of a constitutively active mutant (A293K, Kjelsberg et al., 1992) of the α1B-AR was assessed after transient transfection in COS 1 cells. Like the α1D-AR, we detect a significant degree of α1B-AR internalization in cells expressing the constitutively active mutant (Fig.6). Therefore, receptor internalization in the absence of agonist is a property common to constitutively active α1-ARs.
To determine whether the effects of prazosin on α1D-AR immunolocalization were unique to this compound, we examined the effect of a series of antagonists on α1D -AR expression. Rat 1 fibroblasts expressing the α1D-AR were cultured for 24 h in the presence of phentolamine, BMY 7378, 5-methylurapidil, or WB 4101 (α1-AR antagonists), the α2-AR antagonist yohimbine, or the β-AR antagonist propranolol (see Fig. 7). Yohimbine and propranolol had no effect on the cellular localization of the α1D-AR. Phentolamine and BMY 7378, which have been reported to have inverse agonist properties (Scheer et al., 1997; Garcia-Sainz and Torres-Padilla, 1999), induced redistribution of the α1D-AR from internal compartments to the cell membrane. Phentolamine was as effective as prazosin in promoting α1D-AR redistribution, whereas BMY 7378, WB 4101, and 5-methylurapidil were less effective than either phentolamine or prazosin at promoting α1D-AR redistribution. Therefore, the ability to promote α1D-AR redistribution is not unique to prazosin but is also manifested by other ligands that interact with this receptor. These data also support the notion that inverse agonists, like traditional agonists, differ in their intrinsic activity.
Taken together, these data are strong evidence that the α1D-AR is constitutively active in fibroblasts. However, the data obtained in a stably transfected cell line may not accurately represent the localization and signaling characteristics of the α1D-AR in cells that natively express these receptors such as in vascular smooth muscle cells. Figure8 shows immunoreactivity of femoral artery smooth muscle cells immunostained with α1B- and α1D-AR antibodies. Similar to the immunoreactivity seen in fibroblasts, the α1D-AR is localized in a perinuclear fashion in vascular smooth muscle cells. This similarity in immunolocalization suggests that the α1D-AR may be constitutively active in cells that natively express all three α1-AR subtypes, an area of investigation that we are actively pursuing.
The α1D-AR has been somewhat of an enigma. It is not as well studied as the other α1-AR subtypes. Theroux et al. (1996) showed that this receptor was weakly coupled to second messenger pathways. Yang et al. (1997) were unable to detect a significant degree of expression of the α1D-AR in a variety of rat tissues. These findings are consistent with our observations that the receptor is predominantly expressed in intracellular compartments due to constitutive activity. As a result, the α1D-AR is not as responsive to agonist activation as the other subtypes of the α1-AR. We noted that phenylephrine stimulated inositol phosphate accumulation but had no effect on ERK activity in α1D-AR-expressing cells. This suggests that the effect of constitutive activity on functional responses differs depending on the second messenger/kinase being studied.
Studies with either transfected HEK 293 or SK-N-MC cells failed to demonstrate any constitutive function of the α1D-AR (Theroux et al., 1996). The constitutive activity of the α1D-AR we observe may be related to the level of α1D-AR expression. In our studies, we note differences in cellular localization and signaling of the α1B- and the α1D-AR, even though these receptors are expressed at similar levels in the Rat 1 fibroblasts (5–10 pmol/mg protein), suggesting that the level of receptor expression is not sufficient to account for our results. Previous studies also using the stably transfected α1D-AR subtype in Rat 1 fibroblasts showed constitutive activation of the calcium response (Garcia-Sainz and Torres-Padilla, 1999), a function linked to the production of IP3.
The α1D-AR also binds most agonists but not antagonists with higher affinity than the other two α1-AR subtypes. This phenotype is a hallmark of constitutive activity based on the revised ternary complex model (Samama et al., 1993). Constitutive activity is also associated with constitutive phosphorylation and, therefore, desensitization. Current models of GPCR signaling indicate that the receptor is dephosphorylated and recycled back to the cell surface. The internalized pool of α1D-ARs may represent the equilibrium component of the recycling process due to its constitutive nature.
In summary, we have shown that there are significant differences in the cellular expression of the α1-ARs. The α1B-AR exhibits characteristics of a typical GPCR as it is expressed on the cell surface, is subject to agonist-mediated internalization, and couples to second messenger/kinase activation. The results with the α1D-AR are not consistent with traditional models of GPCR. This receptor is expressed mainly in intracellular compartments. The unstimulated α1D-AR appears to be constitutively active and is subject to regulation by inverse agonists. In preliminary studies in vascular smooth muscle cells, we have observed that the α1D-AR is localized in a perinuclear orientation similar to that seen in fibroblasts. These data argue that the cellular localization we report in this paper is not an epiphenomenon of experiments in fibroblasts but rather is a characteristic of this receptor. We are currently investigating the possibility that the α1D-AR is constitutively active in cells natively expressing the receptor.
Acknowledgments
We thank Carol Swiderski for excellent technical assistance.
Footnotes
- Received July 28, 1999.
- Accepted December 20, 1999.
-
Send reprint requests to: Michael T. Piascik, Ph.D., Director, Vascular Biology Research Group, Dept. of Pharmacology, The University of Kentucky College of Medicine, 800 Rose St., UKMC MS 305, Lexington, KY 40536-0084. E-mail: mtp{at}pop.uky.edu
-
This work was supported by National Institutes of Health Grants HL-38120 (MTP), HL-52544 (DMP), and American Heart Association Established Investigator Award (DMP) and Scientist Development Award (GRP).
Abbreviations
- AR
- adrenoceptor
- DMEM
- Dulbecco's modified Eagle's medium
- ERK
- extracellular signal-regulated kinase
- FITC
- fluorescein isothiocyanate
- GPCR
- G protein-coupled receptor
- MBP
- myelin basic protein
- The American Society for Pharmacology and Experimental Therapeutics