Expression, purification , reconstitution and pharmacological characterization of a membrane tethered Gsα subunit from insect cells A model study using β2AR and tetGsα as a tool for ligand screening.

Venkata. R. P. Ratnala, Gayathri Swaminath, Tae Weon Lee, Brian K. Kobilka

1

INTRODUCTION G protein coupled receptors (GPCRs) constitute the largest family of membrane proteins in the human genome and are responsible for the majority of transmembrane signal transduction in response to hormones and neurotransmitters. GPCRs receive extracellular signals and interact with and activate G proteins. Binding of an agonist to the receptor induces formation of a complex comprising of a ligand, the receptor, and a G protein, followed by nucleotide exchange from GDP to GTP on the G protein α subunits. The complex then dissociates to the GTP-bound α-subunit (αGTP), βγ-complex, and receptor. The αGTP and βγ subunits both interact with downstream members of the signal cascades and initiate activation or inhibition of various effector enzymes and/or ion channels1-3. Heterotrimeric G proteins are essential components in transmission of signals from activated GPCRs to effector molecules. Gsα, the stimulatory G protein for adenylyl cyclase, is one of the first heterotrimeric G proteins to be identified, purified and cloned1, 4, 5

. G protein-mediated pathways are the most fundamental mechanisms of cell signaling.

In order to analyze these pathways, the availability of purified recombinant G proteins are critically important. More recently the three dimensional structure of this protein has been elucidated by X-ray crystallography. However, in spite of the remarkable success in characterizing the protein, it remains one of the most difficult G proteins to purify in a fully functional form. Expression and purification of this protein from E. coli6 was used for crystallography7, however, Gsα purified from E. coli is not palmitoylated, and therefore not capable of functionally interacting with GPCRs such as the β2 adrenoceptor (β2AR).

2

Gsα has been shown to be purified from Sf9 insect cells, yet this process is technically challenging, Using Sf9-Baculovirus expression system, a general and simplified method to purify tetGsα protein subunits is described here. This method can be useful for purification of most of G protein subunits. We have been able to purify sufficient quantities of tetGsα using this method to enable biophysical analysis or β2ARGsα coupling assays. We previously reported the generation of a modified, membrane tethered Gsα (tetGsα) that coupled more efficiently than wild type Gsα when expressed in insect cells.8,9. We now report that this membrane tether facilitates large scale production and purification of functional tetGsα from insect cells and also show that the tetGsα and reconstituting it with β2AR for its use as an efficient coupling system and in turn use it for screening efficacies of the diverse ligands. This would be further extended to other GPCR systems to examine if the receptor-tetGsα reconstitution systems can be used for high efficacy assay system and also could be used for screening of ligands for other classes of GPCRs. Development and improvement of novel subtype-specific ligands for known GPCRs is a key step in elucidating receptor functions and developing new drugs. Thus, screening many thousands of samples for each target GPCR is a regular practice to identify novel ligands or drugs in pharmacological research. Current conditions of ligand screening procedures promoted the development of a functional assay system with precise efficacies that could be extended for high-throughput screening. The highthroughput screening of GPCR ligands has generally been based on a cell-based assay system that monitors downstream events of signalling cascades. The representative downstream events for Gs and Gq coupled receptors are increase in intercellular cAMP

3

concentrations and Ca2+ concentration, respectively. Thus, measurements of cAMP and Ca2+ are now the most popular assay systems that are available for GPCRs. The cellbased assay systems were applied to several receptors and contributed to the finding of novel ligands like AC-90179 10 and ghrelin 11. However, these cell-based assays have the disadvantage that endogenous receptors on host cells respond to their ligands resulting in false positive signals and also since the signal being collected at the very distal end of the activation processes, which does not give the clear picture of the proximal events that take place near the receptor and seldom give false positives. Assay protocols that directly estimate G protein activation by a target receptor in vitro may constitute a simple screening system that avoids such false positive reactions in high-throughput screening. We planned to construct a new ligand screening system and apply it to β2AR. The reconstitution system between a GPCR and its partner G protein α subunit has been shown to be useful for studies of receptor-G protein interactions because of their efficient coupling8. Thus, the receptor-tetGsα reconstitution complex could be useful candidate for a new screening system. Using a β2AR receptor-tetGsα reconstitution system (β2AR-Gsα) as a model system, we have examined this system’s sensitivity and efficacies on various drugs. These results also suggest the potential usefulness and advantages of β2AR receptor-Gsα reconstitution system for screening of ligands, ligand efficacies for β2AR and which could be possibly extended to screen drugs for other GPCRs, which couple to the Gs subunit of the G proteins.

4

MATERIALS AND METHODS Materials Rat Gsα long isoform cDNA was kindly provided by Dr. R. R. Reed ( Johns Hopkins University., Baltimore, MD)12 Sf9 ESF921 culture medium is obtained from Expression Systems (Woodland, CA). [35S] GTPγS (1250 Ci/mmol) were obtained from PerkinElmer Life Sciences, Inc. (Boston, MA). Guanosine 5′−diphosphate sodium (GDP) was obtained from Sigma-Aldrich, Inc. (St. Louis, USA) Guanosine 5′-Othiotriphosphate (GTPγS) was from Boehringer Mannheim (Mannheim, Germany). NDodecyl-b-D-maltoside (nDM) was obtained from Anatrace (Maumee, OH, USA). Complete, EDTA-free Protease Inhibitor Cocktail Tablet was obtained from Roche Applied Science, Palo Alto, USA. 1,2-Dioleoyl-sn-Glycero-3-Phosphocholine (DOPC) Avanti Polar Lipids Inc., Alabaster, AL. “DC Protein Assay” was purchased from BioRad laboratories, (Hercules, CA) Sources of other materials have been described elsewhere8. Buffers Buffer A consisted of 50mM Tris-HCL pH7.4, 1mM EDTA, 3mM MgCl2, 100mM NaCl, 5mM NaF, 20uM AlCl3, 10uM GDP, 10mM β-mercaptoethanol and a mixture of protease inhibitors containing benzamidine and leupeptin at 1mM and 6µM final concentrations respectively. For the above 50 ml of lysis buffer one protease inhibitor cocktail tablet (roche) has been added. Buffer B consisting of containing 1% n-dodecyl maltoside, 50mM Tris-HCL pH7.4, 1mM EDTA, 3mM MgCl2, 100mM NaCl, 5mM NaF, 20uM AlCl3, 10µM GDP, 5mM CaCl2 1mM cysteine and protease inhibitors. Buffer C is consists of 0.1% n-dodecyl maltoside, 20mM Tris-HCl pH 7.4, 1mM MgCl2

5

100mM NaCl, 2mM CaCl2 and 10µM GDP. Buffer D is consisted of buffer C with 4mM EDTA and 200ug/ml flag peptide. Cell Culture and Viruses The recombinant baculovirus encoding the tetGsα was constructed as described earlier8. We generated recombinant baculovirus using the Bac to Bac baculovirus expression system (Invitrogen). Each virus strain went through 2 rounds of amplifications before using it for receptor expression as described9, 13. Virus stocks were characterized by infecting sf9 cells seeded at 3.0 x 106 cells/mL with dilutions ranging from 1:10, 1:100 1:500 and 1:1000 to determine the dilution that produced the highest expression of the recombinant protein. Sf9 insect cells were grown at 27° C in suspension cultures in shake flasks rotating at 120 rpm in ESF-921 medium (Expression Systems, CA) supplemented with 0.05 mg ml-1 genatamicin. Cells were routinely passaged every three days to 3-4 X 105 cells ml-1 The cell density was determined using a haemocytometer. Cells were infected using a 1:100 dilution of high titer virus stock, at a density of 3-4 million cells per milliliter. After infection, they were incubated for 48 hours, then harvested by centrifugation at 5000 g for 10 min. Cell pellets were stored at -80 °C until further use. Single step tetGsα purification Tethered Gsα was expressed in SF9 cells and lysed in lysis buffer (Buffer A). The lysate was dounced 20 times and centrifuged at 18,000 rpm for 10 min at 40C. The pellets were solubilized by douncing 20 times on ice in soubilization buffer (buffer B) The solubilized fraction was incubated for 1h at 40C with gentle stirring followed by centrifugation at 18,000 rpm for 20 min. To the supernatant 2mM CaCl2 was added to the solubilized fraction and passed onto M1 Flag column which is pre-equilibrated with

6

Buffer C with a flow rate of 40-50 ml per hour. The column was washed with 4 column volumes of buffer C and eluted with buffer D. The peak fractions of pure tetGsα subunit were pooled and the protein concentration is determined by using biorad protein DC assay. To the pooled fractions 10% glycerol by volume and MgCl2 to a final concentration of 5mM were added. The purified tetGsα subunits were aliquoted and stored at –80°C until further use. SDS gel electrophoresis and Western blotting Sf9 cells expressing the tetGsα were collected and centrifuged for 5 min at 2000 g. Cell pellets were taken up in an SDS/PAGE sample buffer (2% sodium dodecyl sulfate (SDS), 0.04 M dithioerythrol (DTE) and 0.015% bromophenol blue in 0.5 M Tris, final pH 6.8). Samples were run on a 12% SDS/PAGE gel at 100 V for the 5% acrylamide stacking gel and 200 V for the running gel. Protein staining was performed using Coomassie blue. For immunodetection proteins were blotted onto a nitrocellulose membrane (1 h at 100 V) in ice-cold blot buffer (25 mM Tris and 0.2 M glycine in 20% methanol) using a MiniProtean system (Bio-Rad, Melville, NY, USA). Blots were subsequently immunoassayed for the presence of flag-tagged tetGsα. Preparation of lipids Phospholipid 1, 2-Dioleoyl-sn-Glycero-3-Phosphocholine (DOPC) was purchased from Avanti Polar Lipids Inc. and their purity was ascertained by high performance liquid chromatography. Cholesterolhemisuccinate was purchased from steraloids Inc. A stock solution of 20 mg/ml of DOPC and 10 mg/ml of cholesterolhemisuccinate was made in chloroform. 3mg of DOPC and 0.3mg of cholesterolhemisuccinate was aliquoted into a glass vial from respective stock solutions. The chloroform was evaporated by passing

7

argon and then vacuum dried for 1h to remove any residual chloroform. The lipid mixture was then hydrated in a 20 mM Hepes buffer pH 7.5, 100 mM NaCl and 1% octylgluoside. Then the lipid mixture was vortexed vigorously and sonicated for over two hours with 10 min intervals in an ice/water bath. This mixture was stored at –80° C and used for reconstitution experiments. Preparation of reconstituted receptor and tetGsα 300 µl reconstitution mix of β2AR and tetGsα was made by mixing with the concentrations of 16µM receptor, 8µM tetGsα, 10mM GDP, 1 M MgCl2 in 37.5µL of lipid mixture (3mg/mlDOPC+0.3mg/mlCHS) and the volume was adjusted to 300µl with the reconstitution buffer (20mM Tris pH7.4, 3mM MgCl2, 100mM NaCl, GDP 10µM). The final concentrations would be of 1µM of receptor, 5µM of tetGsα 0.3mg/ml DOPC, 0.03mg/ml CHS, 33µM GDP and 3mM of MgCl2. The receptor-lipid mixture was mixed well and allowed to reconstitute on ice for 2 h. The vesicles formed were desalted on a 25x1.5 cm SephadexG-50 (fine) column to obtain uniform vesicles of about 100 nm. [35S] GTPγS binding assay Reconstituted receptor and tetGsα protein were suspended in 500 µl of binding buffer supplemented with 0.05% (w/v) BSA, 0.4 nM [35S] GTPγS, GDP (1 µM) with or without ICI and isoproterenol (10 µM). BSA was added in order to prevent absorption of membranes to the tube. Incubations were performed at 25 0C and shaking at 230 rpm for 30 min. Nonspecific binding was determined in the presence of 100 µM GTPγS and was less than 0.2% of total binding. Bound [35S] GTPγS was separated from free [35S] GTPγS

8

by filtration through GF/C filters, followed by three washes with 3 mL of cold binding buffer. Filter-bound radioactivity was determined by liquid scintillation counting. [3H] Dihydroalprenolol binding Reconstituted receptor tetGsα complex supplemented with 0.05% (w/v) BSA, was suspended in 500 µl of binding buffer supplemented with [3H] dihydroalprenolol (0.1210 nM). BSA was added in order to prevent absorption of membranes to the tube. Nonspecific binding was determined in the presence of 10µm alprenolol. Incubations were performed for 60 min at 25°C and shaking at 200 rpm. Competition binding experiments were carried out with reconstituted receptor tetGsα complexes with 1nM [3H] dihydroalprenolol in the presence of unlabeled ligand at various concentrations with or without GTPγ[S] (10 µM) as described elsewhere9, 13. Bound [3H] dihydroalprenolol was separated from free [3H] dihydroalprenolol by filtration through GF/C filters, followed by three washes with 2 ml binding buffer (4°C). Filter-bound radioactivity was determined by liquid scintillation counting.

9

RESULTS Purification of tetGsα Production conditions As the production level of a given recombinant protein is difficult to predict, in our hands, high production levels were obtained with protein-free medium (ESF921, Expression systems). In this medium Sf9 cells had a doubling time of about 20-24 h, and and could reach 7–8 X 106 cells/ML. Because of the potentially high cell densities and the good cellular production levels, resulting in optimal volumetric production levels, ESF921 protein free medium was used for all subsequent tetGsa production. Scaling up of our insect cell suspension cultures in ESF921 medium from 100 mL cultures to 1 litre and its multiples depending on the production needs was achieved in shaker flasks without significant loss in production level. Over a large number of experiments, we obtained production levels of functional tetGsα in Sf9 cells in the range of 3 mg per litre of cells. It has been previously reported that the MOI and the time-point of infection in the cellular growth cycle are important parameters in determining volumetric production levels and optimal time of harvesting. Final levels did not vary significantly when cells were infected in their early mid exponential growth phase with a good viral titre. Production levels levelled at 2–3 dpi, when cell viability had not yet suffered much. Therefore cells were routinely harvested at 2 dpi. We observed that the fragmentation of the tetGsα could be nearly completely suppressed by harvesting the cells at 2dpi with out the addition of any protease inhibitors in the sf9 cell cultures. The culture conditions yielding optimal volumetric production of recombinant tetGsα in 100 mL cultures, could

10

be directly scaled up to 1L flasks and could be easily maintained up to 10L, maintaining production yields of 3 mg of functional tetGsα per litre of culture. Functional soubilization and stabilization of tetGsα Much effort was put into finding optimal conditions to solubilize recombinant tetGsα from the insect cell membranes for subsequent binding to the flag-affinity matrix. Careful optimization of the detergent and the buffer composition was critical for obtaining maximum soubilization efficiency as well as optimal stability of functional tetGSa in micellar solution. First tests showed that detergent solubilized tetGsα was quite unstable and lost all activity within 2–3 days at 4°C. We could significantly stabilize the tetGsα receptor by addition of the 10mM GDP and 3mM MgCl2 in presence of 1% ndocyl maltoside (DDM). 70-90% soubilization efficiencies of functional tetGsα was obtained with DDM in presence of 100mM NaCl. While the pH had little effect on tetGsα stability in the range of 6.5–7.8. When a low density of cell membrane suspension was used in the solublization buffer, the extraction of functional tetGsα was nearly quantitative with sufficient stability to survive subsequent purification steps. Purification and reconstitution of the tetGsα A general and single step method for purification of G proteins from Sf9 cells is described here. The tetGsα protein subunit to be purified is expressed with an associated flag-tag. The expressed protein is adsorbed to a M1 liked flag containing resin and the desired untagged protein is eluted with EDTA and Flag peptide, which reversibly activates the α subunits of G proteins and causes dissociation of α from βγ. This method takes advantage of the high affinity and large capacity of flag resin for the flag tag, as

11

well as the extremely specific elution of the untagged subunit with EDTA and Flag peptide. We obtained 3 mg of pure tetGsα from 1 litre of insect cells. Figure 1a is a commasie stained PAGE of purified tetGsα. SDS PAGE of the purified preparation reveals a single protein band with an apparent molecular weight of 60kDa Fig. 1b shows the western blot. We obtained 3 mg of pure tetGsα from 1 litre of insect cells. Extending the tetGsα with a flag-tag aimed at rapid single-step affinity purification by immobilized m1 antibody. FLAG peptide at a concentration of 200 µg/ml was included in the elution buffer for efficient elution. Most of the low-affinity contamination could be removed by washing with buffer C, we routinely used 200 µg/ml of flag peptide along with 4mM EDTA (buffer D) for rapid and complete elution of the tetGSa. As estimated from SDS/PAGE analysis, the purity of the tetGsα after FLAG purification ranges between 75 and 95% (e.g. Fig. 1a, and 1b).The fractions eluted with 200 µg/ml along with 4mM EDTA were stored at -80°C and screened for functional tetGsα by SDS PAGE gels and western blotting. Those with a positive response were processed for reconstitution for GTPγS binding assays. A mixture of DOPC and CHS was added to the combined purified β2AR and tetGsα fractions in a molar lipid to protein ratio of about 100: 1. This is within the natural lipid to protein range of cellular membranes and with this ratio full functionality of recombinant β2AR which has been earlier demonstrated (Gayathri paper and ncb). The receptor-tetGsα-lipid mixture was mixed well and allowed to reconstitute on ice for 2 h. The vesicles formed were desalted on a 25x1.5 cm SephadexG-50 (fine) column to obtain uniform vesicles of 100 nm. Subsequently, the proteoliposomes containing reconstituted receptor and the tetGsα can be separated from non reconstituted receptor and from non reconstituted receptor and 12

tetGsα detergent complexes by selecting the right fractions from the eluate from the G-50 column. A proteoliposome fractions collected just after the void volume of the column would indicate proper reconstitution. The purity of the tetGsα is sufficient for functional and biophysical studies. Functional Characterization Characterization of β2AR-tetGsα reconstitution complex The efficiency of receptor-tetGsα protein coupling was assessed in several ways. GTP sensitive, high-affinity agonist binding reflects the formation of the ternary complex between agonist, receptor, and guanine nucleotide-free G protein. GTPγS binding assay measures the uptake of the non-hydrolysable GTP analogue GTPγS to the α-subunit and is therefore not a steady-state assay. However, this assay provides information about the number of G proteins accessible to receptors during a given period of time. Interaction of β2AR and tetGsα in the β2AR-tetGsα Purified and reconstituted β2AR-tetGsα complex derived from Sf9 cells were examined for binding with [3H] DHA or [35S] GTPγS. Figure 2 shows the displacement curve of [3H] DHA binding in the absence or presence of 0.1mM GTPγS for the purified and reconstituted β2AR-tetGsα complex. The displacement curve with a full agonist, isoproterenol, shifted to the right in presence of GTPγS for reconstituted β2AR-tetGsα complex. The displacement curve with a full agonist, isoproterenol, shifted to the left in absence of GTPγS for reconstituted β2AR-tetGsα complex. Approximately 50 % GTPγS sensitive high affinity agonist binding sites were observed for the reconstituted β2ARtetGsα complex in the molar ratio of 1:5. We then examined the effect of β1γ2 on the

13

above reconstitution complex. We reconstituted β2AR:tetGsα:β1γ2 in the ratio of 1:5:5 respectively and performed the above experiments. We observed a 70% GTPγS sensitive high affinity agonist binding sites were observed for the reconstitution complex. Effects of β2AR Ligands on [35S] GTPγS binding Figure 3 shows the time course of [35S] GTPγS binding to reconstitution complex in presence of 10 µM Isoproterenol, and 10 µM ICI. Figure 4a and 4b shows effect of isoproterenol and other ligands on the binding of [35S]GTPγS to the reconstitution complex in the presence of 1 µM GDP. The extent of [35S] GTPγS binding was approximately 13 times greater in the presence of isoproterenol than in its absence. ICI did three fold -ve stimulus the [35S] GTPγS binding. Effects of β2AR Ligands on the affinities of β2AR-tetGsα for Guanine Nucleotides We also examined effects of ligands on displacement of [35S] GTPγS binding by guanine nucleotides. The displacement curves by GDP shifted to right in the presence of agonist, indicating that the affinity for GDP of β2AR-tetGsα reconstitution system decreased by agonist binding (Fig. 5). The full agonist, partial agonist, and antagonist differed in their effects on the apparent affinity of the reconstitution system for GDP. These results indicated that the agonist bound β2AR-tetGsα reconstitution system had a lower affinity for GDP than the ligand-free or antagonist-bound β2AR-tetGsα reconstitution system, and that the affinity for GDP of salbutamol-bound β2AR-tetGsα reconstitution system was intermediate. A similar shift was also observed in the displacement curves by GTP in the presence of full and partial agonists, although the extent of the shift was smaller than in the displacement by GDP. No shifts were observed for displacement by GTPγS with and without ligands (data not shown). These results 14

indicate that the β2AR-tetGsα reconstitution system provides a useful means to characterize ligand–receptor–Gα-guanine nucleotide interactions.

15

DISCUSSION Purification of G proteins from natural tissue requires lengthy procedures with limited yields and it is also difficult to resolve closely related members of Gsα subunits. Reports show that expression of Gsα, Giα, and Goα in Escherichia coli (E. coli) yields large amounts of protein that can be myristoylated where appropriate (Giα and Goα)14, but the proteins are not palmitoylated and may be missing some other unknown modifications. Alpha subunits of Gq and G12 subfamilies and the βγ complex have not been successfully expressed in E. coli as active proteins yet. The Sf9-Baculovirus expression system has many advantages to overcome these problems. First, a variety of posttranslational modification mechanisms, especially lipid modifications, such as palmitoylation, myristoylation, and prenylation, are present in Sf9 cells. These lipid modifications are critically important for the interactions of G protein subunits with receptors, RGS proteins, or effectors. With these modifications present, the recombinant G protein subunits from Sf9 expression system are almost as active as native proteins15, 16

. For the optimal expression of the tetGsα it was better to store low passage viruses

as the progenitor stocks at –80°C. The operating virus stocks (usually 200–500 mL/batch) are amplified from these stocks and are stored at 4°C. These viral stocks are stable for at least several months. Early passage cells are frozen in medium containing 10% FBS and 10% DMSO and can also be used as cell stocks. Cells from one vial of frozen stock can be maintained for approx 4–6 mo. We change to a new frozen stock when we start to see cells with irregular shapes, a decrease of growth rate, or a reduction in the expression level of recombinant proteins. To increase the yield of tetGsα protein, the freshly

16

amplified recombinant viruses are recommended. It is likely that the myristoylation that is present in tetGsα from Sf9 cells, but not from E. coli. The final yield of recombinant TetGsα is approx 3mg per litre of sf9 cell culture. The same procedure can be applied for the purification of other G proteins. However, the yield of will be directly proportional to the level of expression in Sf9 cells and its inability to be activated by NaF17. The presence of glycerol prevents the aggregation of the protein during concentration and storing the tetGsα. The development of subtype-specific ligands for known GPCRs is also required as a means to determine the physiological function of each subtype and for use as drugs with low side effects. Thus, several protocols have already been developed to find ligands for GPCRs and successfully applied to identify ligands for GPCRs. One of the most popular methods for ligand screening is measurement of intracellular Ca2+ concentrations in cells expressing a target receptor. Cells that express greater amounts of receptors can respond to lower concentrations of ligands, because effectors may be fully activated when a smaller fraction of receptors is bound with agonist. In this paper, we have reported the sensitivity and specificity of β2AR-tetGsα reconstitution system in screening of ligands. Our results indicate that the β2AR-tetGsα Reconstitution complex requires only a minute fraction of the ligands to detect stimulation of [35S] GTPγS binding and is sensitive enough to be applicable for ligand efficacies. As this assay is at the very proximal to the receptor and to the ligand binding site, hence gives a much clearer picture of the events taking place after the ligand binds to the receptor, and reduces the number of false positives and increases the efficacy of the assay.

17

Since the GTPγS binding studies reflect the efficiency of G protein activation, we see much more efficient basal and agonist stimulated GTPγS binding activation in tetGsα membranes without the β1γ2 proteins compared to the presence of β1γ2 proteins. The most important advantage of [35S] GTPγS binding assay using the β2AR-tetGsα reconstitution system over cell-based assays is that false-positive reactions are negligible. The cell-based assay systems are affected by different circumstances causing falsepositive responses by endogenous receptors. These results indicate that the β2AR-TetGs reconstitution system is useful for ligand screening because of its high specificity and efficacy. Another benefit of the β2AR-tetGsα reconstitution system is the high signalnoise ratio of roughly ten fold to basal in the presence and absence of agonist (Fig. 4a). In addition, a large amount of membranes expressing a β2AR and tetGsα can be prepared easily by using the baculovirus-Sf9 system and could purify these proteins with ease. One 10 liter of cultured Sf9 cells of tetGsα is estimated to be enough for more than 50,000 assays using Millipore multiscreen HTS system. It is therefore not difficult to use the same batch of preparation with the same specific activity from start to end of a large-scale ligand screening for receptors. The [35S] GTPγS binding assay for the β2AR-tetGsα reconstitution complex is simple and amenable to high-throughput screening by using 96well microplates, as in cell-based ligand screening systems. The [35S] GTPγS assay using β2AR-tetGsα reconstitution system may, therefore, have wide application for GPCR ligand screening. β2AR-tetGsα reconstitution system is also useful to analyze the interaction of a receptor and Gsα. As shown in Fig. 5, β2AR-tetGsα reconstitution complex shows low, intermediate, and high affinity for GDP when it is bound with full agonist, partial agonist, 18

and antagonist, respectively. It is reasonable to assume that the agonist-bound receptor in the reconstitution system accelerates the dissociation of GDP from tetGsα, and that the partial agonist dose so to a lesser extent. We have previously shown that the receptor Gs fusion proteins, the agonist-bound receptors have been reported to effectively stimulate tetGsα and we have reported that the [35S] GTPγS binding activity of Sf9 membrane preparations expressing receptor-tetGsa receptor-Gsα fusion proteins could effectively stimulated by agonists9.The present results indicate that the interaction between β2AR and tetGsα in the reconstitution system is much more robust as compared with that in the fusion protein system. The receptor-tetGsα reconstitution system will be useful for ligand screening systems and will be amenable for detailed efficacy analysis more feasible, because agonists, partial agonists and antagonists can be discriminated by means of a simple binding assay, as shown here in our studies. In summary, we have shown that a membrane tether facilitates the expression and purification of functional tetGsα from Sf9 insect cells. The purified protein can be efficiently reconstituted with purified β2AR and functional coupling observed. This preparation should facilitate biophysical studies to characterize interactions between the β2AR and tetGsα. And also, we have shown that the receptor-tetGsα reconstitution system is a good model system for studies on the interaction of receptors and G proteins.

19

FIGURE LEGENDS: Figure 1. a) Purification of the TetGsa assayed by 12% PAGE. Molecular mass marker (Bio-rad, Hercules, CA, USA) are shown in lane 1 and lane 6. Purified TetGsα preparations are shown in lanes 2. b) Immunoblot analysis of purified tetGsα. Marker is shown in lane1. Lane 2 shows the flag-tagged TetGsa. The position of the intact flag tagged tetGsα is indicated by the arrow. Samples were subjected to SDS/PAGE (12% gel), followed by immunoblotting with anti- flag tag serum coupled to Alexa Fluor 568 maleimide as the primary antibody. Fluorescence of the gels was acquired using a FluorchemTM 8800 (FITC filter). Figure 2. Competition by isoproterenol of [3H] dihydroalperenol (DHA) binding in various reconstitution conditions of β2AR and G-protein Subunits with or without GTPγ[S]. [3H] Dihydroalpernol binding in reconstitution complex were performed as described in material and methods. Reconstitution complexes containing different G-protein subunits, [3H] dihydroalpernol and isoproterenol of the concentrations indicated on the abscissa. Reaction mixtures additionally contained binding buffer (control) (z) or GTPγS (10µM) (). Data points are expressed as percent of basal bound [3H] DHA. Data shown are the mean±SD of three independent experiments performed in triplicate. A) β2AR reconstituted with lipids alone. A) β2AR reconstituted with tetGSα in the molar ratio of 1:5 c) β2AR reconstituted with TetGsα and β1γ2 in the molar ratio of 1:5:5. D) β2AR reconstituted with β1γ2 complex alone in the molar ratio of 1:5.

20

Figure 3. Time course of [35S] GTPγS binding in reconstitution complex containing β2AR and TetGsα. The reconstitution complexes were incubated with 1nM [35S] GTPγS and 1µM GDP in the absence („) or in presence of 10µM isoproterenol (z) or in presence of 10µM ICI118551 () as described in the materials and methods section. Data shown are the mean±SD of three independent experiments performed in triplicates. Figure 4. Ligand-stimulated [35S] GTPγS binding to purified TetGsα reconstituted with purified B2AR. Reconstitution and [35S] GTPγS binding were performed as described under material and methods. The effects of agonist isoproterenol and the inverse agonist ICI118551 on basal [35S] GTPγS binding in the presence and absence of β2AR are shown in the histograms. The data for each panel represent the average ± S.E. of three determinations and representative of three independent experiments. Figure 5. Displacement of [35S] GTPγS binding by GDP in the presence of various β2AR ligands. The purified β2AR reconstituted with purified tetGSα in the molar ratio of 1:5 were incubated with 1mM Isoproterenol („),ICI118551(▲), salbutamol (♦) and in absence of ligand (▼) and with indicated concentrations of GDP at 30°C for 30 min in binding buffer with 100pM [35S] GTPγS, and 10mM MgCl2. The membranes were trapped on GF/B glass filters, which were washed three times with ice cold binding buffer. Radioactivity was then counted with liquid scintillation counter. Data points are expressed as percentage of basal bound [35S] GTPγS. Data shown are the mean ± S.E of three independent experiments performed in triplicate. 21

Figure 1a)

Figure 1b).

22

Figure 2.

23

Figure 3.

24

Figure 4a)

a) β 2 AR:Gsα :Gβ 1 γ 2 :: 1:5:0

[35S] GTPγS binding (c.p.m.)

70000 60000 50000 40000 30000 20000 10000

(+) receptor

Ba sa 10 l µM IC 10 µM I 10 I µM SO G TP γS

Ba sa 10 l µM IC 10 µM I 10 I µM SO G TP γS

0

(-) receptor

Figure 4b)

b) β 2 AR:Gsα :Gβ 1 γ 2 :: 1:5:5 40000 35000 30000 25000 20000 15000 10000 5000

(+) receptor

Ba sa 10 l µM IC 10 µM I 10 IS m O M G TP γS

0 Ba sa 10 l µM IC 10 µM I 10 I µM S O G TP γS

[35S] GTPγS binding (c.p.m.)

45000

(-) receptor

25

Figure 5)

26

REFERENCES: 1. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13.

14. 15.

16.

Gilman, A. G. G-Proteins - Transducers Of Receptor-Generated Signals. Annual Review Of Biochemistry 56, 615-649 (1987). Neer, E. J. Heterotrimeric G-Proteins - Organizers Of Transmembrane Signals. Cell 80, 249-257 (1995). Bourne, H. R. How receptors talk to trimeric G proteins. Current Opinion In Cell Biology 9, 134-142 (1997). Harris, B. A., Robishaw, J. D., Mumby, S. M. & Gilman, A. G. MolecularCloning Of Complementary-Dna For The Alpha-Subunit Of The G-Protein That Stimulates Adenylate-Cyclase. Science 229, 1274-1277 (1985). Neer, E. J. & Clapham, D. E. Roles Of G-Protein Subunits In Transmembrane Signaling. Nature 333, 129-134 (1988). Lee, E., Linder, M. E. & Gilman, A. G. in Heterotrimeric G Proteins 146-164 (1994). Sunahara, R. K., Tesmer, J. J. G., Gilman, A. G. & Sprang, S. R. Crystal structure of the adenylyl cyclase activator G(S alpha). Science 278, 1943-1947 (1997). Lee, T. W., Seifert, R., Guan, X. M. & Kobilka, B. K. Restricting the mobility of G(s)alpha: Impact on receptor and effector coupling. Biochemistry 38, 1380113809 (1999). Seifert, R., Lee, T. W., Lam, V. T. & Kobilka, B. K. Reconstitution of beta(2)adrenoceptor-GTP-binding-protein interaction in Sf9 cells - High coupling efficiency in a beta(2)-adrenoceptor-G(s alpha) fusion protein. European Journal Of Biochemistry 255, 369-382 (1998). Weiner, D. M. et al. 5-hydroxytryptamine(2A) receptor inverse agonists as antipsychotics. Journal Of Pharmacology And Experimental Therapeutics 299, 268-276 (2001). Kojima, M. et al. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 402, 656-660 (1999). Jones, D. T. & Reed, R. R. Molecular-Cloning Of 5 Gtp-Binding Protein Cdna Species From Rat Olfactory Neuroepithelium. Journal Of Biological Chemistry 262, 14241-14249 (1987). Seifert, R. et al. Different effects of G(s)alpha splice variants on beta(2)adrenoreceptor-mediated signaling - The beta(2)-adrenoreceptor coupled to the long splice variant of G(s)alpha has properties of a constitutively active receptor. Journal Of Biological Chemistry 273, 5109-5116 (1998). Mumby, S. M. & Linder, M. E. in Heterotrimeric G Proteins 254-268 (1994). Iniguezlluhi, J. A., Simon, M. I., Robishaw, J. D. & Gilman, A. G. G-Protein Beta-Gamma Subunits Synthesized In Sf9 Cells - Functional-Characterization And The Significance Of Prenylation Of Gamma. Journal Of Biological Chemistry 267, 23409-23417 (1992). Hepler, J. R. et al. Purification From Sf9 Cells And Characterization Of Recombinant Gq-Alpha And G11-Alpha - Activation Of Purified PhospholipaseC Isozymes By G-Alpha Subunits. Journal Of Biological Chemistry 268, 1436714375 (1993).

27

17.

Kozasa, T. et al. Purification And Characterization Of Recombinant-G(16-Alpha) From Sf9 Cells - Activation Of Purified Phospholipase-C Isozymes By G-Protein Alpha-Subunits. Proceedings Of The National Academy Of Sciences Of The United States Of America 90, 9176-9180 (1993).

28