INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ACTIVATION of a G protein-coupled receptor often triggers the sequestration of the receptor and internalization of the agonist. These processes contribute significantly to the regulation and maintenance of the responsiveness of a cell to external signals because receptor sequestration represents an important step in receptor desensitization and/or resensitization. Whereas the behavior of most wild-type receptors can be investigated in cells expressing them endogenously in significant amounts, only very few naturally expressed mutants are available for the study of structure-function relationships of receptor domains (11). The artificial expression of receptors and especially of their mutants in cultured cell lines either as transient or as stable clones has proved to be a very valuable substitute for this purpose.

This approach has also been used quite successfully with the kinin receptors (1, 13-16) to determine the role of their COOH-termini in ligand internalization and signal transduction. However, the various publications reported rather large differences in the internalization rates, even for wild-type B₂ receptors ranging from <30% ligand internalization after 10 min in COS-7 cells (14) to over 80% in Chinese hamster ovary (CHO) cells (1). This was considered to be at least partially due to the use of different cell lines and cell culture techniques, which hindered the comparison of these data (14). When we could not reproduce our own data (1) after changing our protocol to generate stable mutants, we were inspired to look for the molecular parameters determining the internalization and sequestration rates of the kinin receptors heterologously or homologously expressed in cultured cell lines and to define experimental conditions to get reproducible and comparable data with these cell lines.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Materials. CHO cells DUCXB1 were purchased from American Type Culture Collection, and Flp-In T-REx [human embryonic kidney (HEK)-293] cells were from Invitrogen (Groningen, The Netherlands). [2,3-Prolyl-3,4-³H]bradykinin (108 Ci/mmol) was from Perkin Elmer Life Sciences (Boston, MA). Peptides were bought from Bachem (Heidelberg, Germany). The primers were synthesized by Invitrogen and delivered desalted and lyophylized. Plaque-forming unit DNA polymerase was obtained from Stratagene Europe. Roche (Mannheim, Germany) delivered Fugene and Invitrogen Hygromycin B and Blasticidin. Polylysine, captopril, 1.10 phenanthroline, and bacitracin were purchased from Aldrich (Taufkirchen, Germany). Fetal calf serum, culture media, and penicillin-streptomycin were delivered by PAA Laboratories (Cölbe, Germany). All other reagents were of analytic grade and are commercially available.

Expression vectors. The sequence of the human B₂ kinin receptor, starting with the third encoded methionine, was cloned into the BamHI and the XhoI sites of the pcDNA5/FRT vector (where FRT represent Flp recombinant target) from Invitrogen. The receptor sequence was preceded at the NH₂ terminus by either a single hemagglutinin tag (MGYPYDVPDYAGSA) or a double-tag sequence (MGRSHHHHHH-GYPYDVPDYAGSA) cloned into the HindIII and BamHI site of the vector. To obtain lower expression levels of the B₂ kinin receptor (B₂KR) in HEK-293, the cytomegalovirus (CMV) promotor sequence was removed from the pcDNA5/FRT by cleavage with MfeI and HindIII and replaced by the simian virus 40 (SV40) promotor sequence derived from the pFRT/lacZeo vector (Invitrogen) using standard PCR technology. Alternatively, the tagged B₂KR coding sequences were cloned into a modified pIRESneo vector from Clontech. For B₂eGFP chimera, the sequence for enhanced green fluorescent protein (eGFP) taken from the pEGFP-C1 vector (Clontech) was added to the COOH terminus of the B₂KR with standard PCR methodology using a chimeric primer that excluded the stop codon of the B₂KR but included the first methionine of the eGFP-construct. For transient expression in COS-7 cells, the pCDNA3.1 (Invitrogen) was used because the other vectors do not contain the SV40 origin needed for plasmid amplification.

Cell culture. CHO cells were cultivated in MEM- with ribonucleosides and deoxyribonucleosides supplemented with 10% fetal calf serum and penicillin-streptomycin. A CHO host cell line harboring the FRT site was created with the pFRT/lacZeo-vector (Invitrogen) according to the instructions of the supplier. HEK-293 cells were grown in DMEM with 10% fetal calf serum and penicillin-streptomycin. Transfection of these cells was done using Fugene (Roche) by following the instructions of the producer [2 µg plasmid(s) plus 5 µl Fugene per 6-well dish]. Single, stably transfected clones were obtained after selection with either 300 µg/ml hygromycin B (pcDNA5/FRT-vectors) or after selection with G418 (800 µg/ml) for cells transfected with the pIRESneo-vector.

Cell transfection using electroporation. Four million cells (HEK-293 or COS-7) detached with trypsin were mixed together with 10 µg DNA in Bio-Rad gene pulser cuvettes (0.4 cm) and electroporated at 260 V,   , 960 µFD using a Bio-Rad gene pulser. The cells from up to six electroporations were pooled subsequently on a cell culture dish (10 cm), trypsinized after 24-h incubation in normal culture medium, and seeded on 12-well plates for binding experiments performed the next day.

[³H]BK binding studies. For the determination of dissociation constant K_d and receptor number B_max, confluent monolayers on 24-well plates were rinsed three times with PBS and equilibrated on ice with incubation buffer (40 mM PIPES, 109 mM NaCl, 5 mM KCl, 0.1% glucose, 0.05% BSA, 2 mM CaCl₂, and 1 mM MgCl₂, pH 7.4; degradation inhibitors: 2 mM bacitracin, 0.8 mM 1.10 phenanthroline, and 100 µM captopril) for 15 to 45 min. Subsequently, the plates were incubated on ice with 0.3 ml of ice-cold incubation buffer containing increasing concentrations of [³H]BK (10 concentrations ranging from 0.01 to 40 nM) for at least 90 min. The incubation was stopped by rinsing the cells four times with ice-cold PBS. The cell monolayer was lysed in 0.2 ml 0.3 M NaOH and transferred with another 0.2 ml buffer to a scintillation vial, and the radioactivity was measured in a betacounter after the addition of 2 ml scintillation fluid. Nonspecific binding was determined in the presence of 5 µM of unlabeled BK and subtracted from total binding being determined with [³H]BK alone to calculate specific binding.

Receptor sequestration assay. Monolayers of cells on 12-well plates were rinsed three times with PBS and incubated on ice with 5 µM unlabeled BK in 0.5 ml incubation buffer. After 60 min, the cells were placed in a water bath at 37°C to start receptor sequestration. At the indicated times, the trays were placed on ice, washed two times with ice-cold wash buffer, and treated with a solution of 0.05 M glycine, pH 3.0 (8) for 10 min on ice to remove all unlabeled extracellular ligand. The cells were again washed two times with ice-cold wash buffer, and remaining specific surface binding was determined with 1-2 nM [³H]BK at 0°C as described above.

Internalization of [³H]BK. Cells on 12-well plates were rinsed three times with PBS, preincubated with incubation buffer (15-45 min), and incubated with 0.3 ml of [³H]BK on ice for 90 min to reach equilibrium binding. To start [³H]BK internalization, the plates were placed in a water bath at 37°C. The internalization process was stopped after the indicated time by putting the plates on ice and washing the monolayers four times with ice-cold PBS. Surface-bound [³H]BK was then separated by treating the cell monolayer with 0.2 ml of an ice-cold dissociation solution (0.2 M acetic acid, 0.5 M NaCl, pH 2.7) for 10 min on ice. The supernatant with formerly surface-bound [³H]BK was quantitatively transferred to a scintillation vial by rinsing the cells with another 0.2 ml of PBS (surface-bound [³H]BK). The remaining cell monolayer containing the internalized [³H]BK was lysed in 0.2 ml 0.3 M NaOH and transferred with another 0.2 ml of water to a scintillation vial. The radioactivity of both samples was determined in a betacounter after 2 ml scintillation fluid were added. Nonreceptor-mediated [³H]BK internalization and surface binding were determined in the presence of 5 µM unlabeled BK and subtracted from total binding to give specific values.

Protein determination. Total protein was quantified with the Micro BCA Protein assay reagent kit from Pierce (Rockford, IL) using bovine serum albumin as the standard.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Sequestration of stably expressed B₂KRs in HEK-293 and CHO cells. Using the Flp-In system (Invitrogen), we obtained clones of CHO and HEK-293 cells that stably expressed the human wild-type B₂KRs or eGFP chimera thereof in high numbers exhibiting a B_max of ~4.5 pmol/mg protein for B₂KRs and B₂eGFP in CHO cells and 11 pmol/mg protein in HEK-293 cells and values of dissociation constant (K_d) of 1.2-2.9 nM for both cell types and both constructs. Surprisingly, however, the B₂KRs and the eGFP chimera of both the HEK-293 and the CHO cells responded to stimulation with 1 µM BK with only very poor sequestration. These data obviously deviated from those we had published earlier with respect to CHO cells (1) (Fig. 1).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   B2 kinin receptor (B2KR) sequestration in high-expressing cells. Human embryonic kidney (HEK)-293 and Chinese hamster ovary (CHO) cells, stably expressing high numbers of B2KRs, were preincubated with 1 µM bradykinin (BK) for the indicated times at 37°C. Remaining surface binding was then determined at 0°C with ~2 nM [3H]BK as described in MATERIALS AND METHODS. The results are given as means ± SD of triplicate determinations. Curves shown are representative of two [CHO; B2KR and B2 enhanced green fluorescent protein (B2eGFP)] and three [HEK-293; B2KR (twice) and B2eGFP] other experiments with similar results. CHO publ-labeled curve obtained with a B2KR construct is taken from Faussner et al. (1).

Internalization of [³H]BK by stably and highly overexpressed B₂KRs. Internalization of [³H]BK in these overexpressing HEK-293 cells (expressed as internalized radioactivity in percentage of specificly total bound radioactivity) is strongly dependent on the concentration of ligand applied. Fast internalization occurred with low amounts of ligand (<1 nM), which slowed down drastically with higher concentrations of [³H]BK, from more than 80% internalization at 0.4 nM [³H]BK after 5 min to as low as 10% at 10 nM [³H]BK (Fig. 2A).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Internalization rate in high-expressing cells strongly dependent on concentration of [3H]BK. HEK-293 cells, expressing high amount of B2KRs, were preincubated with the indicated concentrations of [3H]BK for 90 min on ice. The internalization was started by placing the plates in a water bath at 37°C. At the indicated times, the internalization process was stopped, and surface-bound and internalized ligand were determined as described in MATERIALS AND METHODS. A: internalization of [3H]BK in percentage of total bound [3H]BK; B: remaining surface binding; C: amount of internalized [3H]BK in counts per minute (cpm) per well. The results are given as means ± SD of triplicate determination. A representative experiment of two experiments is shown.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Internalization of [3H]BK in overexpressing CHO and HEK-293 cells after 10 min at 37°C. Cell monolayers on 12-well plates were incubated with the indicated amount of [3H]BK on ice for 90 min and thereafter placed in a water bath at 37°C for 10 min. The amount of surface-bound and internalized [3H]BK was determined as described in MATERIALS AND METHODS. Each experiment was done in triplicate. Results are given as means ± SD. The exact concentrations of [3H]BK used were the following: 0.4, 1.7, and 9.2 nM (HEK-293; B2eGFP); 0.4, 2, and 10 nM (HEK; B2); 0.5, 2.1, and 10.2 nM (CHO; B2 and B2eGFP).

[³H]BK internalization in low-expressing CHO and HEK-293 cells. To test the hypothesis that overexpression of the kinin receptors is responsible for the poor receptor sequestration and the concentration dependence of ligand internalization, we tried to generate clones with lower receptor expression. For this purpose, the pIRES expression system (Clontech) was employed. Usage of this expression vector, which inserts itself rather randomly into the cell genome, resulted in several stable CHO clones with receptor expression levels below 1 pmol/mg protein. Internalization of B₂KRs in these low-expressing CHO cells no longer showed a dependence on the concentration of [³H]BK (Fig. 4A) and led to internalization rates similar to those described for endogenously expressing cells (7).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Internalization of [3H]BK in low-expressing cells. CHO cells (A) and HEK-293 cells (B) expressing low amounts of B2KRs [<1 pmol/mg protein (CHO), 4.2 pmol/mg protein (HEK-293)] were treated as described in Fig. 2. All time points represent the means ± SD of triplicate determinations. For the representative experiments shown, five (CHO) and two (HEK-293) other experiments giving similar results were performed.

B₂KR sequestration in low-expressing CHO and HEK-293 cells. Both low-expressing CHO and HEK-293 cells responded to stimulation with 1 µM BK with rapid reduction of the receptor number on the cell surface, reaching a plateau of remaining surface binding of under 30% within 10 min, similar to the results published by us previously (Fig. 5).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Sequestration in low-expressing cells. CHO cells and HEK-293 cells expressing low amounts of B2KR [<1 pmol/mg protein (CHO) and 4.2 pmol/mg protein (HEK-293)] were treated as described in Fig. 1. All points represent the means ± SD of triplicate determinations. Six other experiments in CHO cells and one in HEK-293 gave similar results. CHO publ-labeled curve is taken from Faussner et al. (1). SV40, simian virus 40.

Internalization of transiently expressed B₂ kinin receptors. Transient transfection of cells with receptor genes is a very common approach to studying receptor-associated mechanisms because results can be obtained rather quickly. However, this method usually does not allow to discriminate between causes of the radioactive binding activity: whether this activity derives from a few cells expressing high amounts or from many cells expressing minor amounts of receptors. In [³H]BK-internalization studies with transiently B₂KR-transfected HEK-293 cells (Fig. 6A) and COS-7 (Fig. 6B), cells we observed a clear dependence of the internalization rate on the ligand concentration despite the fact that B_max with 1 pmol/mg protein (HEK-293) or 1.9 pmol/mg (COS-7) was apparently in the range of our low-expressing cells. This dependence indicates that most of the transiently transfected cells highly overexpressed the receptor. Yet, applying lower amounts of [³H]BK (0.6 nM) in both cell lines gave rise to internalization curves similar to those seen for endogenously expressing cells and for stable clones expressing lower amounts of B₂KRs.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Internalization in transiently transfected cells. HEK-293 cells (A) and COS-7 cells (B) were transfected transiently by electroporation as described in MATERIALS AND METHODS. After 2 days, the cells expressing ~1 pmol/mg protein (HEK) and 1.9 pmol/mg protein (COS-7) were treated, and the internalization was determined as described in Fig. 2. Each time point represents the means ± SD of triplicate determinations. One other experiment with HEK cells was performed giving similar results.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Transfection of various mammalian cell lines such as CHO, HEK-293, or COS cells, either transiently or stably with receptor genes and their variants, is a powerful and therefore widely applied tool for studying the function of the different domains of GPCRs. The same holds true for the (altered) behavior of these receptors when they are coexpressed with other proteins. The effects that mutations, deletions, and truncations of these receptors exert on their regulation often provide deeper insights into regulation mechanisms involving these receptors.

Stable and transient transfection often results in receptor numbers per cell that are far above those observed in endogenously expressing cells. The advantage of such an overexpression is the relatively low background. However, it has to be made certain that the overexpressed receptors do not differ too much in their behavior from naturally expressed ones. For this reason, experimental conditions have to be defined that enable the correlation of results obtained from artificially expressed wild-type receptors with those from naturally expressed ones, so that the probable different behavior of mutated receptors may be interpreted properly.

Using the Flp-In system, we obtained several clones expressing high amounts of B₂KR: 10-11 pmol/mg protein in HEK-293 cells and 4.5 pmol/mg protein in CHO cells. Similar high-expression levels in CHO cells have been reported by others (15). Our results clearly demonstrate that such an overexpression prevents the internalization machinery from rapidly translocating all occupied receptors. This can be deduced either from the poor sequestration of B₂KRs after stimulation with saturating amounts of unlabeled agonist or from the slow ligand internalization when concentrations are used that lead to occupation of a major part of the receptor compartment. This happens mostly in internalization assays in which (almost) saturating concentrations of tritiated ligands are used as opposed to assays in which iodinated ligands are used. In the latter, the ligand concentrations are usually far below the K_d due to the high specific labeling.

The data presented indicate that the term "overexpression" has to be interpreted in the context of the cell type used. Whereas a B₂KR expression level of 4.2 pmol/mg protein still enabled fast internalization rates in HEK-293 cells {even with almost saturating concentrations of [³H]BK (Fig. 4)}, a similar expression level of 4.5 pmol/mg protein in CHO cells led to a clear dependence of the internalization rate on the ligand concentration applied (Fig. 3). This difference may be caused by cell-type specific expression levels of proteins involved in receptor sequestration, as indicated by the different expression levels observed for -arrestin in CHO, COS-7, and HEK-293 cells (9). Although -arrestins are involved in the sequestration of many GPCRs via clathrin-coated pits (12), it is possible that they do not play a role in the sequestration of the human B₂KR, because coexpression of the B₂KR with a dominant-negative -arrestin mutant had no influence on its sequestration (8). Although phosphorylation of the B₂KR is required for its sequestration (14), it is rather unlikely that the phosphorylation by a specific receptor kinase is the rate-limiting step in the sequestration of overexpressed B₂KRs because this is a very rapid reaction in which one kinase molecule can phosphorylate many receptors in a very short time. Because there are several reports indicating that the B₂KR is sequestering via caveolae (3, 10), it is more probable that the longer-lasting interactions with adaptor proteins, either those involved in the translocation of the B₂KR into the caveolae or those in the following sequestration into intracellular compartments is the rate-limiting step. So far, however, little is known about the exact mechanism of GPCR sequestration via caveolae and therefore little about the kinds of protein-protein interactions that may be responsible for this rate limitation.

Our results indicate that published data on reduced internalization rates of receptor mutants have to be carefully interpreted in particular when these data were obtained from transiently transfected cells, COS-7 cells in particular, or from stable clones with very high expression. For example, it was reported that at an expression level of 8.59 pmol/mg protein in COS-7 the B₂ receptor double mutant C324A/C329A reveals a 30% reduction in internalization compared with the wild type (13). Yet, we found a perfectly normal internalization behavior for this mutant in HEK-293 cells with lower expression (not shown).

Cautious interpretation is also necessary when one wants to draw conclusions concerning the recycling of the B₂KRs from differences between the receptor sequestration and ligand internalization rates (8). Because determination of receptor sequestration, i.e., the change of the number of receptors on the cell surface after exposure to an agonist, always involves the full receptor complement, it has to be compared with ligand internalization rates under identical conditions, i.e., with (almost) saturating ligand concentrations. Our results show that under these conditions the relative reduction of receptors on the cell surface of high- and low-expressing cells (Figs. 1 and 5) corresponds well with the relative amount of ligand internalized (Figs. 2A and 4, respectively), indicating that there is no significant receptor recycling in the first 30 min after receptor activation.

Whereas the present report deals only with the loss of receptor sequestration in cells that overexpress B₂KRs, this loss might be a phenomenon that occurs in other GPCRs systems, too. For example, the vasopressin receptor V1a responds to 20-min exposure to an agonist with an 80% reduction of the number of receptors on the cell surface, when stably expressed in HEK-293 cells (5). However, only a 30% reduction under otherwise identical experimental conditions was observed when V1a is transiently expressed (4).

Another interesting aspect of the phenomenon described here concerns the possibility that these stably overexpressing cells may represent partially sequestration-deficient cell lines. Therefore, one could use these cell lines to identify proteins involved in receptor internalization by transiently overexpressing these proteins in these cells and trying to find out whether they are capable of rescuing receptor-sequestration.

In summary, our results demonstrate that highly expressed B₂KRs in HEK and CHO cells no longer show receptor-specific sequestration and internalization rates. To overcome this problem, usage of weaker or inducible expression promotors to obtain lower expression levels might be useful. Alternatively, a internalization assay could be used with a ligand concentration that enables the internalization rate independent of the ligand concentration; meaning that only as many receptors are occupied and activated as the cell can normally deal with. Because receptor wild types and mutants often have quite different expression levels, the highest ligand concentration that still gives rise to specific internalization rates might be considerably different for each, and therefore each has to be checked separately, particularly when a reduction in the internalization rate becomes obvious compared with the wild type.