Abstract
Human α1A-, α1B-, and α1D-adrenergic receptors were tagged at their amino termini with FLAG epitopes and stably expressed in human embryonic kidney (HEK)293 cells. Tagged receptors demonstrated a wild-type pharmacology and mobilization of intracellular Ca2+. After solubilization and immunoprecipitation, monomers, dimers, and trimers of each subtype were apparent on Western blots. Further denaturation with 6 M urea reduced most oligomers to monomers. Deglycosylation reduced the molecular size of α1A-, and to a lesser extent α1B- and α1D-adrenergic receptors. Radioligand binding site density was highest for α1A- and much lower for α1B- and α1D-adrenergic receptors, but did not correlate with protein expression. Commercial anti-α1-adrenergic receptor antibodies did not recognize the tagged receptors in Western blots of cell lysates, and substantial cross-reactivity was still observed after solubilization and immunoprecipitation. Surprisingly, only receptor monomers were apparent after photoaffinity labeling with 125I-arylazidoprazosin, and the intensity of photoaffinity-labeling correlated with the density of radioligand binding sites. We conclude that epitope-tagged α1-adrenergic receptors exist as both monomers and oligomers in HEK293 cells, but there is substantial discrepancy between protein and binding site expression. Because only monomers are detected by photoaffinity labeling, dimers and trimers observed on Western blots may be pharmacologically inactive.
The α1-adrenergic receptor (AR) family consists of three closely related gene products (α1A, α1B, and α1D) that mediate the actions of norepinephrine (NE) and epinephrine in sympathetically innervated tissues and brain (Ruffolo et al., 1994;Zhong and Minneman, 1999a). α1-ARs belong to the G protein-coupled receptor family and consist of single polypeptide chains predicted to have seven transmembrane spanning domains. α1-AR subtypes act through Gq proteins to activate phospholipase C, increase inositol 1,4,5-trisphosphate production, and increase intracellular Ca2+ (Esbenshade et al., 1993; Hieble and Ruffolo, 1994).
Characterization of the structural aspects of α1-AR subtypes has been hampered by the lack of good anti-receptor antibodies. Although the use of receptor-selective photoaffinity labels has identified proteins with molecular masses between 59 and 80 kDa in a variety of tissues (Siedman et al., 1984;Lomasney et al., 1986; Terman and Insel, 1986; Lattion et al., 1994), studies of the proteins themselves require subtype-specific antibodies. However, existing antibodies for α1-AR subtypes have generally been found to be inadequate for Western blots (A. Vicentic, M. Uberti, and K. Minneman, unpublished observations), suggesting that they may not be sufficiently specific and/or selective for this purpose. Therefore, we have epitope tagged the three human α1-AR subtypes to be able to isolate, purify, and detect the receptor proteins.
Herein, we describe the characteristics of α1-AR subtypes tagged at their amino termini with hexahistidine and FLAG epitopes to allow purification and immunological detection. We characterize the pharmacological and signaling properties of these receptors, and report on the detection and sizes of these proteins, their glycosylation and oligomerization, and the relationship of radioligand binding sites to receptor protein. In addition, we compare the specificity and selectivity of existing commercially available anti-α1-AR antibodies. These epitope-tagged constructs will facilitate further biochemical and functional investigations of closely related α1-adrenergic receptor subtypes until more selective anti-receptor antibodies are developed.
Experimental Procedures
Materials
Human α1A-AR cDNA (Hirasawa et al., 1993) was provided by Dr. Gozoh Tsujimoto (National Children's Hospital, Tokyo, Japan), human α1B-AR cDNA (Ramarao et al., 1992) was provided by Dr. Dianne Perez (Cleveland Clinic, Cleveland, OH), and the human α1D-AR cDNA was cloned in our laboratory (Esbenshade et al., 1995). Other materials were obtained from the following sources: HEK293 cells (American Type Culture Collection, Manassas, VA); fura-2/acetoxymethyl ester (Calbiochem, La Jolla, CA); (−)-NE, bitartrate, Dulbecco's modified Eagle's medium, penicillin, streptomycin, FLAG peptide, anti-FLAG M2 affinity resin, and HRP-conjugated anti-FLAG M2 antibody (Sigma-Aldrich, St. Louis, MO); prazosin (Pfizer, Groton, CT); and carrier-free Na125I and enhanced chemiluminescence reagent (Amersham Biosciences, Piscataway, NJ). Anti-α1-AR antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and Affinity Bioreagents (Golden, CO). Precast Tris-glycine gels for Western blot analysis were obtained from Novex (San Diego, CA).
Production of Epitope-Tagged α1-AR Subtypes
Receptor coding sequences were generated by polymerase chain reaction and subcloned into the mammalian expression plasmid pDT containing in frame N-terminal sequences encoding hexahistidine and FLAG (H/F) epitopes as described previously (Robeva et al., 1996). After sequencing the inserts, vectors containing one of the three human α1-AR subtypes were transfected into HEK293 cells by calcium phosphate precipitation (30 μg/15-cm plate). Cells were allowed to recover for 5 days before selection with 400 μg/ml geneticin.
Cell Culture
Transfected HEK293 cells were propagated on 15-cm cell culture dishes at 37°C in a humidified 5% CO2incubator in Dulbecco's modified Eagle's medium containing 4.5 g/l glucose, 1.4% glutamine, 20 mM HEPES, 100 mg/l streptomycin, 105 units/l penicillin, pyridoxine hydrochloride, and 10% fetal bovine serum (Esbenshade et al., 1995). The cells were detached by 0.25% trypsin/1 mM EDTA and gentle trituration and subcultured at a ratio of 1:3 when confluent. Cells were split 1 day before transfection to be about 50% confluent on the day of transfection. For studies involving Ca2+measurements, 10-cm dishes were seeded at a density of 6 × 106 cells/10 ml. Cells were grown to confluence before use.
Radioligand Binding and Ca2+ Measurements
For radioligand binding, confluent 15-cm plates were washed with phosphate-buffered saline (20 mM NaPO4 and 154 mM NaCl, pH 7.6) and harvested by scraping. Cells were collected by centrifugation, homogenized with a Polytron, membranes collected by centrifugation at 30,000gfor 20 min, and resuspended in 1× buffer (25 mM HEPES and 150 mM NaCl, pH 7.4) with a cocktail of protease inhibitors (1 mM benzamidine, 3 μM pepstatin, 3 μM phenylmethylsulfonyl fluoride, 3 μM aprotinin, 3 μM leupeptin, and 5 mM EDTA). Radioligand binding sites were measured by saturation analysis of specific binding of the α1-AR antagonist radioligand125I-BE (20–800 pM). Nonspecific binding was defined as binding in the presence of 10 μM phentolamine. The pharmacological specificity of radioligand binding sites was determined by displacement of 125I-BE (50–70 pM) by selected agonists and antagonists, and data were analyzed by nonlinear regression analysis (Theroux et al., 1996). Intracellular Ca2+ mobilization was measured using fura-2 as described previously (Theroux et al., 1996).
Solubilization and Immunoprecipitation
Anti-FLAG Affinity Chromatography.
Membranes (2–3 mg of protein) were prepared as described above and solubilized with 2%n-dodecyl-β-d-maltoside (DβM) in the presence of the same protease inhibitors. Membranes were rocked for 90 min at 4°C, centrifuged for 30 min at 16,000g, and the supernatant diluted 10-fold with 1× buffer containing protease inhibitors. Anti-FLAG M2 affinity column was pre-equilibrated with 5 volumes of 1× buffer with 0.4% DβM and protease inhibitors. Soluble fractions were loaded on the column, washed with 12 volumes of 1× buffer with 0.05% DβM, and eluted with 400 μg/ml FLAG peptide as 5 × 250-μl fractions in 1× buffer with 0.05% DβM. After column chromatography, a portion of each eluate was treated with 0.5 U of N-glycosidase F (PNGase F) for 2 h at room temperature. Samples were run on a 4 to 20% Tris-glycine polyacrylamide gel, transferred to nitrocellulose, and blotted with anti-FLAG M2-HRP-conjugated antibody using 1:600 dilution or with commercially available anti-α1-AR antibodies using 1:300 dilutions.
Ni2+-Nitrilotriacetic acid (NTA) Affinity Chromatography.
Eluted fractions from the anti-FLAG M2 affinity column were loaded by gravity onto columns containing 1 ml of 50% slurry Ni2+-NTA resin that had been pre-equilibrated with 1× buffer with 0.05% DβM. Columns were washed with 10 column volumes of 1× buffer with 0.05% DβM and 2 mM imidazole. Proteins were eluted with the same buffer containing 500 mM imidazole at pH 7.4.
Immunoprecipitation with M2 Anti-FLAG Affinity Resin.
Cells were solubilized as described above, diluted 10-fold, and incubated with anti-FLAG M2 affinity resin (100 μl of 50% slurry for 1 mg of protein) at 4°C overnight. Beads were centrifuged 5 min at 4000 rpm and washed three times with 1× buffer containing protease inhibitors. Proteins were eluted with 400 μg/ml FLAG peptide in 1× buffer and 0.05% DβM. In some cases when testing anti-α1-AR antibodies, a portion of immunoprecipitated receptors was concentrated 5-fold using Centricon YM-10 centrifugal filters (Amicon, Bedford, MA) with a 10,000-mol.wt. cutoff.
Photoaffinity Labeling
Photoaffinity labeling was performed on membranes from HEK293 cells selected for higher expression of each subtype, as described below. Membranes were prepared as described above for radioligand binding (0.5 μg of protein/μl), and treated in the dark for 1 h at room temperature with 6 nM125I-arylazidoprazosin. Nonspecific labeling was determined in the presence of 1 μM unlabeled prazosin. While still in the dark, open tubes were exposed to 6000 μJ/cm2 UV light for 3 min using a Stratalinker (Stratagene, La Jolla, CA). Membranes were then washed once with 1 ml of 1× buffer with protease inhibitors, centrifuged at 16,000g for 5 min, and the pellet homogenized in 0.5 ml of 2× buffer (50 mM HEPES and 300 mM NaCl, pH 7.4) containing protease inhibitors. Membranes were solubilized and immunoprecipitated as described above, and immunoprecipitates were run in parallel on SDS-PAGE. One gel was transferred to nitrocellulose for Western blotting, and the parallel gel was analyzed for radioactivity in a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Results
Expression of Epitope-Tagged α1-AR Subtypes in HEK293 Cells
Radioligand Binding.
Epitope-tagged α1-AR subtypes were stably expressed in HEK293 cells and the density (Bmax) of α1-AR binding sites determined by saturation analysis of the specific binding of the antagonist radioligand125I-BE. As shown in Fig.1, theBmax for α1A-ARs (3610 ± 30 fmol/mg) was about 20-fold higher than that for either α1B-ARs (151 ± 11 fmol/mg) or α1D-ARs (177 ± 27 fmol/mg). There were no significant differences in the affinity (KD) of the tagged α1-AR subtypes for the radioligand125I-BE (76 ± 31, 104 ± 23, and 117 ± 50 pM for α1A-, α1B-, and α1D-ARs, respectively). Competition curves for inhibition of specific125I-BE binding by NE and several selective antagonists were examined to further characterize the pharmacological properties of the tagged subtypes (Table1). KIvalues for the agonist NE, the nonsubtype-selective antagonists prazosin and yohimbine, the α1A-selective antagonist (+)-niguldipine, and the α1D-selective antagonist BMY7378 (Table 1) in competing for specific 125I-BE binding were indistinguishable from affinities for wild-type, nontagged receptors expressed in the same cell lines (data not shown).
Increases in Intracellular Ca2+.
The coupling of the epitope-tagged α1-ARs to mobilization of intracellular Ca2+ was also examined. Figure2 shows NE-stimulated increases in intracellular Ca2+ in HEK293 cell lines stably expressing each tagged subtype. Consistent with the much higher receptor density, activation of α1A-ARs caused a much larger increase in intracellular Ca2+ than did stimulation of α1B- or α1D-ARs. It should be noted that in other cell lines, the relative efficiencies for Ca2+mobilization by these three receptors differ substantially (α1A > α1B > α1D) (Theroux et al., 1996; Zhong and Minneman, 1999a,b).
Immunoprecipitation.
Membranes (2 mg of protein) from HEK293 cells stably expressing the tagged α1-AR subtypes were solubilized with 2% DβM, which has been used successfully to solubilize β2-ARs and other G protein-coupled receptors (Gether et al., 1995). Solubilized preparations were loaded on an anti-FLAG M2 affinity column and eluted with FLAG peptide. The eluates were subjected to treatment with or without PNGase F, run on SDS-PAGE, transferred to nitrocellulose, and blotted with anti-FLAG M2-HRP-conjugated antibody. After deglycosylation, all three receptor subtypes exhibited a major band (Fig. 3A) corresponding to their approximate predicted molecular masses (α1A, 466 aa, ∼50 kDa; α1B, 519 aa, ∼65 kDa; and α1D, 572 aa, ∼80 kDa). However, all three subtypes also exhibited bands of higher molecular mass (between 100 and 150 kDa), which based on their molecular weights seemed to run as SDS-resistant dimers and trimers. After purification on the anti-FLAG M2 column, the three fractions containing the largest amount of receptors were loaded onto a column with Ni2+-NTA. However, Fig. 3B shows that this second purification step yielded results of comparable purity as that obtained with the anti-FLAG M2 column alone and that substantial material was lost during the second purification. To determine whether similar results could be obtained by immunoprecipitation, soluble receptor fractions were incubated with anti-FLAG M2 affinity resin overnight, beads were washed, and receptors eluted with FLAG peptide and deglycosylated. When immunoprecipitates were subjected to SDS-PAGE and blotted with anti-FLAG M2-HRP-conjugated antibody, all three receptors were visible as monomers, dimers, and trimers similar to those observed after column purification (Fig. 4A). None of these bands were observed after solubilization and immunoprecipitation of membranes from untransfected HEK293 cells (data not shown). In contrast to the results obtained with immunoprecipitation, running crude cell lysates on SDS-PAGE and blotting with anti-FLAG M2-HRP-conjugated antibody produced no bands corresponding to any of the three α1-AR subtypes (Fig. 4B), probably because of limitations of total receptor number in the 20 μg of protein loaded in the lysates.
Effect of Urea and Dithiothreitol (DTT) on Higher Order Oligomers
To determine whether the higher order oligomers detected on Western blots of each tagged subtype were indeed SDS-resistant dimers and trimers, harsher denaturation conditions were used. As discussed above, monomers, dimers, and trimers were detected when α1-AR subtypes were solubilized, immunoprecipitated, deglycosylated, and eluted with the standard 4× loading buffer containing 2% SDS and 5% β-mercaptoethanol (Fig.5A). However, overnight treatment with 4× buffer containing 2% SDS, 6 M urea, and 100 mM DTT resulted in virtual elimination of the dimers and trimers and substantial increases in the density of the monomeric bands for all three subtypes. This suggests that the higher order bands observed on Western blots are indeed dimers and trimers.
Evaluation of Commercially Available Anti-α1-AR Antibodies
The ability to detect and purify the FLAG-tagged α1-AR subtypes with anti-FLAG M2 affinity resin provided us with an opportunity to evaluate the specificity of commercial antibodies for these receptors. Lysates of cell lines expressing each tagged subtype were run on SDS-PAGE and blotted with anti-α1A (C-19, called anti-α1C), anti-α1B(C-18), or anti-α1D-AR (R-20, called anti-α1A) antibodies from Santa Cruz Biotechnology. Similar to what was observed with anti-FLAG M2 antibody in cell lysates, after blotting with secondary antibodies and developing, no specific bands were observed corresponding to the tagged receptor subtypes (data not shown). The specificities of these antibodies were also examined after purification of epitope-tagged α1-AR subtypes. Receptors were solubilized, immunoprecipitated, and deglycosylated from 2 mg of membrane protein, and blotted with each of these antibodies. Little or no signal was observed after blotting (data not shown), so solubilized receptors were concentrated 5-fold using Centricon membranes to increase sensitivity. Detection of these concentrated, purified receptor samples with anti-FLAG antibody produced a strong signal corresponding to the predicted monomeric molecular weight for each α1-AR subtype (Fig.6A). Note that the high background observed on this blot is due to concentration of the samples.
In contrast, the subtype-specific antibodies detected their primary antigen but also seemed to produce substantial cross-reactivity. For example, anti-α1A-AR antibody recognized the α1A-AR but also the α1B-AR (Fig. 6B). The anti-α1B-AR antibody recognized both α1B- and α1D-ARs (Fig.6C), whereas anti-α1D-AR antibody did not produce the specific 80-kDa band detected by the anti-FLAG antibody, but picked up similar bands in each lysate (Fig. 6D). We also examined a polyclonal pan anti-α1-AR antibody (PA1-047; Affinity Bioreagents); however, this antibody produced no specific signals corresponding to any of the three subtypes, even after solubilization and immunoprecipitation of the receptors (data not shown).
Comparison of Radioligand Binding and Western Blots
It was interesting that the densities of α1-AR binding sites (Fig. 1) did not correlate with the relative protein expression observed with Western blots (Figs.3 and 4). Although the density of binding sites for each subtype varied by 20-fold, all three subtypes showed similar protein expression. To examine this further, additional HEK293 cell lines stably expressing each of the α1-AR subtypes were generated. In these cell lines, the Bmax for α1A-AR binding sites was similar to that obtained previously (3477 ± 594 fmol/mg), whereas theBmax for α1B-ARs was much higher (1627 ± 482 fmol/mg), and the Bmax for α1D-ARs was still relatively low (280 ± 146 fmol/mg). Again, KD values for125I-BE were similar for all three subtypes (107, 69, and 85 pM). However, despite the more than 10-fold differences in α1B-AR binding sites, these cell lines showed similar protein expression after immunoprecipitation and Western blotting (data not shown), suggesting that differences in protein expression are not due to differential solubilization and/or purification. This conclusion is also supported by the lack of binding sites remaining after membrane solubilization, and the observation that similar results were obtained with constructs with different epitope tags (data not shown).
Photoaffinity Labeling of α1-ARs
To further examine the biochemical and structural properties of the tagged α1-AR subtypes, we used the photoaffinity ligand 125I-arylazidoprazosin. To maximize the number of binding sites, we used membranes from the higher expressing stably transfected HEK293 cells discussed above. Expression levels of α1A-, α1B-, and α1D-AR binding sites in these cell lines were 3477, 1627, and 280 fmol/mg of protein, respectively. After the photoaffinity labeling procedure, receptors were solubilized, immunoprecipitated, and deglycosylated as described above, and parallel gels were run for Western blotting with anti-FLAG M2 antibody and detection of radioactivity by PhosphorImager. Figure7A shows a Western blot of α1-ARs detected with anti-FLAG antibody, whereas Fig. 7B shows a parallel autoradiograph of the identical samples labeled with 125I-arylazidoprazosin. Western blot analysis revealed multiple bands of different molecular weights, as shown in previous experiments. Surprisingly, photoaffinity labeling of α1-ARs resulted in labeling of a single major band, which overlaid exactly with the corresponding α1-AR monomers on the Western blots. The labeling of this band was completely blocked by the presence of 1 μM unlabeled prazosin, showing appropriate α1-AR specificity. Interestingly, the apparent labeling of receptor monomers by 125I-arylazidoprazosin seemed to roughly correlate with the density of α1-AR binding sites measured by radioligand binding assays (α1A > α1B > α1D). Due to the high photoaffinity labeling of α1A-ARs and poor labeling of α1D-ARs, the amount of protein loaded on the gels was altered, explaining why the relative expression of each receptor protein in Fig. 7A differs from previous figures. Similar photoaffinity labeling was obtained in other experiments where the amount of protein loaded and blotted was similar to those reported previously (data not shown).
Discussion
The objective of this work was to generate and characterize epitope-tagged α1-AR subtypes to study their biochemical properties. To facilitate purification and detection, we introduced hexahistidine and FLAG epitopes at their extracellular amino termini. This position was chosen to reduce the likelihood of alterations in binding or function, which depend primarily on determinants within the transmembrane and intracellular domains (Hwa et al., 1995; Graham et al., 1996). HEK293 cells have been used previously to study signaling by α1-ARs because they do not endogenously express these receptors (Theroux et al., 1996; Tao et al., 1997; Zhu et al., 1999) but do express Gαqand Gα11 and related proteins required for signaling. As shown for other receptors (Robeva et al., 1996), addition of N-terminal tags had no apparent effect on ligand binding or G protein coupling of α1-AR subtypes, and the potencies of agonists and antagonists in competing for radioligand binding sites were indistinguishable from wild-type receptors. NE also stimulated release of intracellular Ca2+ in cells stably expressing each of the tagged subtypes, and the rank order of coupling efficiency for release of Ca2+ by each subtype was similar to that observed with wild-type receptors (α1A > α1B ≥ α1D) (Theroux et al., 1996; Zhong and Minneman, 1999b).
The biochemical properties of tagged α1-ARs expressed in HEK293 cells were then determined after solubilization and purification. Western blots of lysates from cells stably expressing the different α1-AR subtypes resulted in a relatively high background, with no apparent bands corresponding to the expressed receptors. Use of an anti-FLAG M2 affinity column to purify the soluble receptors yielded more satisfactory results, with discrete bands corresponding to predicted sizes of receptor monomers, dimers, and trimers apparent. The use of a second Ni2+-NTA column resulted in comparable results but with a loss of material. Although Ni2+-NTA columns tolerate harsher wash conditions compared with antibody affinity columns (Hochuli et al., 1988; Hoffmann and Roeder, 1991), the use of a second purification step did not seem to be advantageous. In fact, simple immunoprecipitation using anti-FLAG M2 affinity resin yielded results similar to those from affinity chromatography, and the ease and simplicity of this approach led us to adopt it in further experiments.
We were surprised to find that the relative expression of α1-AR protein on Western blots did not correlate with the density of radioligand binding sites. Saturation analysis of specific 125I-BE binding showed that expression levels of the three tagged subtypes differed by more than 20-fold in the several stable cell lines that we developed (175–3500 fmol/mg). Despite these differences, comparable levels of protein expression were observed in Western blots from each of the stable cell lines. A similar discrepancy between protein expression and radioligand binding sites has been observed previously with somatostatin receptors, where increased protein levels were not associated with an equivalent increase in binding sites (Pfeiffer et al., 2001). To confirm these observations, we also performed Western blots on samples diluted to contain similar numbers of radioligand binding sites (data not shown), and we found a clear discrepancy between expression of protein and binding sites. This suggests that some proportion of α1-ARs may be synthesized to their full-length but improperly folded, preventing assembly of the receptor binding pocket. Improper folding and trafficking have been demonstrated for a number of other proteins, including ATP-sensitive K+ channels (Zerangue et al., 1999) and γ-aminobutyric acidB receptors (Margeta-Mitrovic et al., 2000).
The availability of epitope-tagged α1-AR subtypes also allowed us to determine the specificity and cross-reactivity of commercial antibodies. Antibodies from Santa Cruz Biotechnology were raised against unique peptide epitopes from the intracellular C-terminal tail of individual α1-AR subtypes and should not be affected by the N-terminal tags. However, these antibodies were not found to be useful for detecting α1-ARs on Western blots. When used on crude lysates of cells stably expressing the different subtypes, high nonspecific staining similar to that seen with the anti-FLAG M2 antibody was observed, and no bands corresponding to individual receptor subtypes were visible. Even after solubilization and purification by anti-FLAG M2 immunoprecipitation, Western blots with these antibodies showed cross-reactivity that was apparent after concentration of the purified receptor proteins. These subtype-specific antibodies have been used to examine receptor expression in fibroblasts and vascular smooth muscle cells using immunohistochemistry (Hrometz et al., 1999), particularly in cell lines with high heterologous expression (Gurdal et al., 1997). Because different characteristics are important in immunohistochemistry and Western blots, antibodies can often be useful for one technique but not another. However, the high background staining and cross-reactivity observed with these antibodies in Western blots suggest that their selectivity should be carefully evaluated each time they are used.
Glycosylation is important for structural maturation of cell-surface proteins (Kornfeld and Kornfeld, 1985), and the extent of glycosylation of α1-AR subtypes was determined by comparing receptor sizes before and after treatment with N-glycosidase F. A significant (∼15%) decrease in the molecular weight of α1A-ARs and intensification of the monomeric band were noted after treatment with N-glycosidase F, suggesting that this receptor is subject to N-linked glycosylation. This result is consistent with the existence of consensus sites for N-linked glycosylation on α1-ARs (Lomasney et al., 1991) that has been previously demonstrated for α1B-ARs (Graham et al., 1996). Little or no glycosylation was observed for α1B- or α1D-ARs in our studies, although minor decreases in molecular weight were observed after deglycosylation in some gels (data not shown), consistent with other reports (Lomasney et al., 1991; Perez et al., 1991; Graham et al., 1996).
Although Western blots of the three α1-AR subtypes showed the expected bands corresponding to the expected size for α1-AR monomers, clear bands corresponding to approximate molecular weights for dimers and trimers were also often apparent. A growing body of evidence suggests that G protein-coupled receptors, previously believed to function as monomers, are capable of forming functionally relevant homo- and/or heterodimers. Dimers corresponding to approximately twice the monomeric receptor molecular mass have been demonstrated for several G protein-coupled receptors, including β2-adrenergic (Hebert et al., 1996), δ-opioid (Cvejic and Devi, 1997), γ-aminobutyric acidB (Jones et al., 1998), mGluR5 (Romano et al., 1996), and m3 muscarinic receptors (Maggio et al., 1999). Our results suggest that α1-ARs form homo-oligomers, as evidenced by discrete bands on Western blots corresponding to 2 or 3 times the molecular size of individual monomeric receptors. The ability of high concentrations of urea and SDS to denature these dimers and higher order oligomers to monomers supports this conclusion.
We also examined the ability of ligand to bind to these receptors by photoaffinity labeling. Previous studies have used photoaffinity labeling to provide information on structural aspect of ARs, but conclusions were often difficult due to the coexistence of different subtypes of different molecular sizes (Siedman et al., 1984; Cornett and Norris, 1985; Terman et al., 1988). Surprisingly, we found that125I-arylazidoprazosin specifically labeled only a single band for each α1-AR subtype, corresponding exactly to the size of the α1-AR monomers seen on Western blots. Overlay of the Western blot and autoradiograph obtained from two gels with identical samples run in parallel showed a complete overlap of the autoradiographic band with the receptor monomers, but no radioactivity associated with the higher order oligomers seen on Western blots. Radioactivity incorporated into the bands corresponding to α1-AR monomers was completely eliminated by coincubation with 1 μM unlabeled prazosin, supporting the idea that the bands labeled are indeed receptor binding sites. Our data are generally consistent with previous results (Lomasney et al., 1986; Terman et al., 1988; Lattion et al., 1994;Vazquez-Prado et al., 2000) in which photoaffinity labeling of α1-ARs from mammalian smooth muscle cells was found to label only a single band with molecular weights similar to those expected for α1-AR monomers, providing little or no evidence for the existence of higher order oligomers. Similar results have been reported with photoaffinity labeling of D2-dopaminergic receptors with125I-azidophenethylspiperone. This photoaffinity probe was incorporated only by receptor monomers, whereas dimers and higher order oligomers were observed in Western blots with anti-D2 receptor antibody (Zawarynski et al., 1998). Although one previous study (Garcia-Sainz et al., 2001) demonstrated that 125I-arylazidoprazosin labeled several bands apparently corresponding to α1D-ARs in transfected cells, the major band that was sensitive to phentolamine was of 80 kDa, similar to our results with α1D-AR monomers. The photoaffinity labeling of only α1-AR monomers, but not dimers or trimers, raises the question of whether higher oligomeric structures are pharmacologically active.
In summary, we have described the use of epitope-tagged α1-AR subtypes to compare their biochemical and pharmacological properties. Introduction of the N-terminal tags did not affect the pharmacological specificities or signaling properties of these receptors, and Western blots after immunoprecipitation yielded monomers corresponding to the predicted receptor sizes. However, SDS-resistant dimers and trimers were also prominently observed on Western blots, although these could be reduced to monomers by harsher denaturing conditions. These tagged receptors were used to evaluate commercially available antibodies, and it was found that high nonspecific binding and significant cross-reactivity are likely to limit the utility of these reagents. Strikingly, the amount of receptor protein observed on Western blots did not correlate with the density of α1-AR binding sites, and photoaffinity labeling identified only the α1-AR monomers on Western blots. These observations suggest that a significant amount of full-length receptor protein may be inappropriately folded or processed such that it does not form functional binding sites, and raise questions about the pharmacological significance of the dimers or trimers. The availability of epitope-tagged receptor constructs allowing direct comparison of protein and binding site expression should be very useful in studying these receptor proteins, their potential for dimerization and formation of higher order oligomers, and the importance of this phenomenon in receptor-mediated signal transduction.
Acknowledgments
We thank Andre S. Pupo for helpful advice and discussions.
Footnotes
-
This study was supported by the National Institutes of Health.
- Abbreviations:
- AR
- adrenergic receptor
- NE
- norepinephrine
- HRP
- horseradish peroxidase
- H/F
- hexahistidine FLAG
- DβM
- 4% n-dodecyl-β-d-maltoside
- PNGase F
- N-glycosidase F
- Ni2+-NTA
- Ni2+-nitriloacetic acid resin
- PAGE
- polyacrylamide gel electrophoresis
- aa
- amino acid(s)
- DTT
- dithiothreitol
- BMY7378
- 8-[2-[4-(2-methoxyphenyl)-1-piperazinyl]ethyl]-8-azaspiro[4.5]decane-7,9-dione dihydrochloride
- 125I-BE
- 125I-BE 2254
- Received January 14, 2002.
- Accepted March 7, 2002.
- The American Society for Pharmacology and Experimental Therapeutics