Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The B₂ receptor (B₂R) subtype for bradykinin (BK) is preformed and widely distributed in mammalian tissues (endothelial and smooth muscle cells, afferent nerve endings, intestinal and renal epithelial cells, etc.; Marceau and Bachvarov, 1998). Accordingly, the physiological effects of kinin administration into normal tissues are both prominent and diversified: vasodilation and increased vascular permeability, hyperalgesia, natriuresis, etc. The B₂R type, for which BK is the minimal natural kinin sequence retaining high affinity, has been the target of most drug development efforts in this field.

As shown for several other G protein-coupled receptors, the B₂R has been observed to undergo agonist-induced redistribution, desensitization, and endocytosis in cultured cells. Early events following agonist stimulation in DDT1 MF-2 or A431 cells include redistribution of B₂R in caveolae and the formation of endocytic vesicles that are not clathrin-coated (de Weerd and Leeb-Lundberg, 1997; Haasemann et al., 1998). BK (100 nM) internalized about 60% of the surface receptors within 2.5 min in the HF-15 human fibroblast cell line, and a large fraction of the surface binding returned within 60 min (Blaukat et al., 1996). These authors mentioned the possibility of agonist-induced receptor down-regulation in that system. Parallel to endocytosis, functional desensitization and resensitization were observed (BK-induced intracellular calcium concentration increase); Ser/Thr phosphorylation/dephosphorylation events preceded the internalization and recycling to the surface of the B₂R, respectively (Blaukat et al., 1996). Further investigations progressively circumscribed a small Ser/Thr residue cluster located in the receptor C-terminal sequence that is the major structural determinant of agonist-induced B₂R phosphorylation and internalization (Faussner et al., 1998; Pizard et al., 1999). This sequence, SMGTLRTSIS, is highly conserved in the mammalian species for which the ortholog B₂Rs have been sequenced (Farmer et al., 1998). Mutation of all the Ser/Thr residues to Ala within this sequence abolishes BK-induced phosphorylation and massively inhibits internalization (Pizard et al., 1999).

There is evidence that receptor desensitization/resensitization cycle, paralleled by endocytosis, is mechanistically different from agonist-induced down-regulation for ₂-adrenoceptors (Jockers et al., 1999). As for the BK B₂R, the physiological correlates of desensitization/endocytosis may be found in biological systems where a specific time window (15 min) of desensitization follows the administration of a B₂R agonist, with substantial resensitization afterward; this is observed in systems pertaining to the vascular function in the rat (renal vasodilation, Praddaude et al., 1995; blood-brain barrier opening, Bartus et al., 1996). True B₂R down-regulation may rather be observed in some forms of intense, chronic inflammation (e.g., specific loss of BK functional effect in rat colonic mucosa submitted to persistent inflammation, Kachur et al., 1986; loss of B₂R immunoreactivity in renal grafts undergoing acute rejection in humans, Naidoo et al., 1996); the mechanism of this hypothetical down-regulation is unknown, as is the obligatory role of the autologous agonist in this phenomenon. Of potential interest for this line of investigation, the binding sites for [³H]BK corresponding to the B₂R are known to be destroyed by an acute trypsin treatment in bovine endothelial cells, and a corresponding loss of function was also noted (Sung et al., 1989). Similarly, the leukocyte receptor for anaphylatoxin C5a has been shown to be down-regulated by chymotrypsin or proteases derived from microbial pathogens (Jagels et al., 1996), suggesting agonist-independent pathways for pathological receptor down-regulation.

We previously reported a rabbit B₂R-green fluorescent protein (GFP) conjugate in a study focused on the analysis of noncompetitive antagonism (Houle et al., 2000). In the present study, the main objective pursued was to characterize agonist-induced changes in the receptor population in a heterologous system, using GFP-tagged B₂R constructs. The facilitation of imaging and of the immunological detection by GFP tagging allowed to probe the role of specific endocytic pathways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs. BK was purchased from Bachem (Torrence, CA), and the remaining drugs as well as sequencing grade trypsin were from Sigma (St. Louis, MO). Alexa Fluor 594 transferrin conjugate and low-density lipoprotein (LDL) DiI complex were purchased from Molecular Probes (Eugene, OR).

Construction of the Mutated B₂R-GFP S/T Variant. The construction of the rabbit B₂R-GFP eukaryotic expression vector based on pEGFP-N3 (Clontech, Palo Alto, CA) has been previously described (Houle et al., 2000). This construct was used as a template for the creation of B₂R-GFP S/T variant by site-directed mutagenesis using the Transformer Site-Directed Mutagenesis kit (Clontech) and a specific oligonucleotide primer. In the B₂R-GFP S/T construct, point mutations were created to replace six Ser/Thr residues in the carboxyl-terminal tail by Ala residues. The nucleotide sequence of the oligonucleotide primer was 5'-CGGAGAACGCCATGGGCGCGCTGCGGGCCGCCATCGCGGTGG-3'); the underlined nucleotides represent the nucleotide substitutions, leading to the simultaneous substitution of Ser³³⁸, Ser³³⁹, Thr³⁴², Thr³⁴⁵, Ser³⁴⁶, and Ser³⁴⁸ with Ala (residue numbering as in Bachvarov et al., 1995). All mutations were confirmed by sequencing, using the T7 Sequencing kit (Amersham Pharmacia Biotech, Piscataway, NJ). The last five of these Ser or Thr residues are necessary for efficient agonist-induced endocytosis of the human B₂R (Pizard et al., 1999); the Ser³³⁸ residue is specific to the rabbit sequence and has also been mutated to Ala, because it is very close to the postulated phosphorylation domain.

Cell Transfection. COS-1 cells, grown in 12-well plates until 70% confluent, or HEK 293 cells grown until 70% confluent in 35-mm Petri dishes were transiently transfected with the B₂R-GFP coding vector, its S/T variant, or the pEGFP-N3 control vector (coding for GFP) using the Ex-Gen 500 transfection reagent (MBI Fermentas Inc., Flamborough, ON, Canada), as directed by the manufacturer. Fluorescence microscopy proved that a large proportion of the cells expressed the fluorescent constructions 48 h after the transfection. The derivation of a HEK 293 cell line stably expressing B₂R-GFP is described elsewhere (Houle et al., 2000).

Binding Assays. The binding of [³H]BK (NEN Life Science Products, Boston, MA; 90 Ci/mmol) to adherent intact cells was evaluated in confluent 24-well plates as described (Bachvarov et al., 1995; Houle et al., 2000). The assay was applied to the evaluation of the affinity of B₂R conjugates in transiently transfected COS-1 cells (saturation curves constructed by establishing the radioligand binding at 0°C for 90 min with 1 µM cold BK in some wells to determine the nonspecific binding); the Scatchard plot parameters were evaluated using a computerized procedure (Tallarida and Murray, 1987). The COS-1 cells have been preferred for binding experiments due to the saturable, single-site properties of the B₂R constructs in these cells (Houle et al., 2000).

Effect of BK on the Subcellular Distribution of B₂R-GFP. These experiments were based on HEK 293 cells because the imaging of these cells is efficient and because a stable transfectant line for B₂R-GFP has been obtained, with very low intracellular fluorescence background (Houle et al., 2000). The agonist BK was added to the culture medium of the transfectant HEK 293 cells expressing B₂R-GFP or its S/T variant, and the subcellular fluorescence distribution observed without fixation or drug washout using a Bio-Rad 1024 laser beam confocal microscope as a function of treatment duration (up to 3 h; 60× objective with oil immersion; emission 488 nm, detection above 510 nm). The serum-containing culture medium was not removed or replaced to avoid any uncontrolled metabolic deprivation or stimulation; however, cycloheximide (71 µM) was always added 15 min before the agonist, drug treatment, or to control observations to avoid interference from newly synthesized receptors. Variations of this protocol involved addition of captopril (1 µM) to the culture medium, to influence BK half-life, or pretreatment with kinin receptor antagonists (15 min before agonist, simultaneously with cycloheximide).

Enzyme Immunoassay of BK in the Culture Medium. The culture medium of transfectant HEK 293 cells (containing 10% serum) is suspected to degrade the agonist as a function of time, thus influencing the cycling of receptor. To document this, BK (10 or 100 nM) was added to the culture medium (1 ml) of transfectant HEK 293 cells expressing B₂R-GFP (confluent 24-well plates incubated at 37°C). At various times (10-50 min), 200 µl of the medium was removed and transferred into 1 ml of ice-cold ethanol. The suspension was incubated in ice for 1 h, centrifuged to remove the precipitated proteins, and stored at 80°C until assayed. For that purpose, the ethanol extract was evaporated to dryness using a Speed vac system, and then processed precisely as described for the measurement of immunoreactive BK (Décarie et al., 1994). The results are expressed as the estimated BK concentration in the initial culture medium, and were analyzed for first order decay using a computer program (Tallarida and Murray, 1987).

Immunoblot. Monoclonal antibodies to GFP were purchased from Zymed (San Francisco, CA). For the analysis of B₂R-GFP, transfectant HEK 293 cells (confluent 75-cm² flasks) were put in boiling lysis buffer containing 10 mM Tris pH 7.4, 1.0 mM Na₃VO₄, and 1.0% SDS. The lysates were incubated for 5 min at 100°C and then centrifuged at 15,000g for 5 min. Total protein concentrations were then determined using the bicinchoninic acid protein assay (Pierce, Rockford, IL). Twenty-five micrograms of total proteins was run on a 9% SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane. The membranes were rinsed 20 s in methanol and dried for at least 1 h at room temperature. The blots were then incubated 1 h at room temperature in blocking buffer [washing buffer (10 mM Tris pH 7.5, 100 mM NaCl, 0.1% Tween 20) containing 5% skimmed milk]. The primary antibody (monoclonal anti-GFP, dilution 1:2000; Zymed) was added for 2 h at room temperature in fresh blocking buffer. The membranes were washed for 30 min in washing buffer at room temperature before adding the secondary antibody (horseradish peroxidase-conjugated, preadsorbed goat anti-mouse IgG; Santa Cruz Biotechnology, Santa Cruz, CA; dilution 1:16,000) for 1 h at room temperature in blocking buffer. The membranes were washed in washing buffer for another 30 min and then the antibodies were revealed using the Western Blot Chemoluminescence Reagent Plus (NEN Life Science Products), as directed. The immunoblot assay of B₂R-GFP has been applied to the analysis of receptor down-regulation initiated by agonist or extracellular protease pretreatments. In the case of the agonist stimulation, the cells were pretreated in the cell culture medium containing 10% serum (Houle et al., 2000). The trypsin treatment of intact cells and supporting control treatments were applied to cells washed with and incubated in serum-free -minimal essential medium for 30 min before extraction. The enzyme treatment (10 min; 1 µM) was not sufficient to detach cells, but rounded morphology was observed.

Analysis of Internalized Receptors and Ligand. We have previously shown the rapid agonist-induced translocation of B₂R-GFP into dense organelles that can be recovered in the 15,000g pellet of the cell extract (Houle et al., 2000). This cell fraction contains endosomes and lysosomes, among other organelles (Lodish et al., 1995). Homogenization of HEK 293 cells stably expressing B₂R-GFP in sucrose/tricine buffer, and sequential centrifugations were performed as described (Houle et al., 2000). To further study the significance of this form of cellular sequestration, this scheme was applied to cells treated with 10 nM cold BK or with 10 nM [³H]BK for various periods of time in the culture medium at 37°C. In the first case, the B₂R-GFP content of the 15,000g pellet from a whole 75-cm² flask was migrated and analyzed by immunoblot as described above; the 15,000g pellet of cells stimulated with the radioligand was counted by scintillation (untransfected HEK 293 cells were used as controls).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

[³H]BK Binding to Rabbit B₂R Fluorescent Conjugates. COS-1 cells transiently transfected with a GFP coding vector (sham transfection) failed to bind [³H]BK, whereas cells that expressed either of two rabbit B₂R fluorescent variants, B₂R-GFP and B₂R-GFP S/T, exhibited specific and saturable binding (Fig. 1A). The affinity estimates derived from Scatchard plot analysis (Fig. 1B) were close to each other (K_D = 3.1 and 1.28 nM, respectively). These values are similar to previously reported estimates in COS-1 cells for the wild-type receptor (K_D = 2.09 ± 0.39 nM) and the B₂R-GFP conjugate (Bachvarov et al., 1995; Houle et al., 2000). The B_max estimates understandably varied in the separate transient transfections shown in Fig. 1 (115 and 35 fmol/well for B₂R-GFP and B₂R-GFP S/T, respectively), but the mutant construction was consistently less expressed when the experiments were replicated, as reported with the human mutated B₂R (Pizard et al., 1999).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   A, specific binding of [3H]BK to COS-1 cells transiently expressing GFP (control), the B2R-GFP or B2R-GFP S/T fusion proteins. Values are mean of triplicate determinations in a representative experiment. B, Scatchard plot analysis of the binding data in A.

Subcellular Distribution of the B₂R-GFP Fusion Proteins as a Function of Treatment with Ligands. The stable transfectant HEK 293 cell line expressing the B₂R-GFP fusion protein has been reported previously (Houle et al., 2000). The cycling of this receptor induced by the agonist BK has been studied using confocal microscopy (Fig. 2; confocal plane halfway to cell thickness). These experiments were performed in the presence of cycloheximide (71 µM) to prevent the replacement of receptors by newly synthesized ones. The acute addition of BK (10 or 100 nM) was rapidly followed by translocation of the fluorescence into the cells (mostly vesicular structures, but also more diffuse labeling close to the plasma membrane; Fig. 2). These changes were reversible in 1 to 3 h, because most of the fluorescence label returned to membranes. Radioligand binding to these cells in the presence of cycloheximide evidenced a variable but small decline of surface receptors over 3 h (less than 10%; data not shown).

View larger version (81K): [in this window] [in a new window]   Fig. 2.   Subcellular localization of the B2R-GFP fusion protein in HEK 293 cells stably expressing this receptor conjugate, maintained in the complete culture medium and treated with cycloheximide (71 µM) and, optionally, with captopril (1 µM), and the indicated BK concentration for defined time periods (magnification 1800×). The selected confocal planes are halfway to the thickness of most cells.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Concentration of immunoreactive BK in the culture medium of HEK 293 cells stably expressing B2R-GFP exposed to exogenous BK (10 or 100 nM). Some cell wells had been pretreated with captopril (1 µM) for 30 min before adding BK. See text for data analysis.

View larger version (70K): [in this window] [in a new window]   Fig. 4.   Effect of drug treatments (applied 15 min before) on BK (10 nM)-induced internalization (assessed at 30 min) in HEK 293 cells stably expressing B2R-GFP treated with cycloheximide (71 µM). The treatments consisted of Ac-Lys-[Leu8]des-Arg9-BK (B1R antagonist; 1 µM) or LF 16.0687 (B2R antagonist; 1 µM). Presentation as in Fig. 2.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Immunoblot of HEK 293 cell extracts based on monoclonal anti-GFP antibodies. A, HEK 293 cells stably expressing B2R-GFP were extracted without treatment (lane 3), treated for 3 h with captopril 1 µM + cycloheximide 71 µM (lane 4) or captopril + cycloheximide + BK 100 nM (lane 5). B, effect of trypsin digestion on intact cells. HEK 293 cell stably expressing B2R-GFP were extracted following treatment for 30 min with cycloheximide (71 µM; lane 4) or cycloheximide + trypsin (1 µM) added for the last 10 min of incubation (lane 5). Other cells expressing B2R-GFP were pretreated 30 min before extraction with BK (100 nM) and captopril (1 µM; lane 6) optionally supplemented with trypsin (1 µM) for the last 10 min of incubation (lane 7). In both A and B, extracts from untransfected HEK 293 cells (lane 2) and HEK 293 cells transiently transfected with GFP (lane 1) were also immunoblotted as controls. Extract from untransfected HEK 293 cells digested with trypsin (B, lane 3) was also immunoblotted.

Agonist-Induced Translocation of a B₂R-GFP Conjugate into a Cellular Fraction. BK-induced translocation of B₂R-GFP into a 15,000g pellet of the cell homogenate has been previously demonstrated based on immunoblot involving anti-GFP antibodies (Houle et al., 2000). Further analysis of this system shows that the GFP-labeled receptors remain only temporarily in this compartment when cells are stimulated with BK (10 nM; Fig. 6). The time course is similar to that recorded in microscopy in the absence of captopril (Fig. 2). The same type of transfected cells internalized [³H]BK (10 nM) into the 15,000g pellet of the cellular extract with a comparable kinetics (progressive release of the radiolabeled peptide at 1 and 3 h; Fig. 6). By comparison, untransfected HEK 293 cells internalized very little [³H]BK into the 15,000g pellet of the cellular extract (Fig. 6).

View larger version (49K): [in this window] [in a new window]   Fig. 6.   Analysis of the 15,000g pellet of the cellular extract of HEK 293 cells stably expressing B2R-GFP treated with BK. Top, cells expressing B2R-GFP (shaded bars) or untransfected HEK 293 cells (black bars) were treated with 10 nM [3H]BK for various periods before harvesting the cells. The radioactivity was determined in the 15,000g pellet of the extract. Bottom, total extract from 75-cm2 confluent flasks of cells expressing B2R-GFP was migrated on a 11% SDS-polyacrylamide gel and immunoblotted using a monoclonal anti-GFP antibody. The cells had been treated with BK (10 nM) for various time periods.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Colocalization of B2R-GFP (green) and endosome markers (red fluorophore-labeled transferrin or LDL) in stably transfected HEK 293 cells stimulated or not with BK (100 nM; 15 min). See text for analysis. Presentation as in Fig. 2.

Effect of Mutations in the C-Terminal Receptor Tail on Agonist-Induced Endocytosis. The conjugate B₂R-GFP S/T is identical to B₂R-GFP, except for the mutation of C-terminal serine/threonine residues possibly involved in ligand-induced receptor phosphorylation. Experiments were performed in transiently transfected HEK 293 cells to compare the capacity of the two conjugates to be internalized by BK treatment (100 nM) (Fig. 8). Under these conditions, the background intracellular fluorescence is larger than that of stably transfected cells. The cells expressing B₂R-GFP S/T usually showed a higher intracellular fluorescence, under the form of clustered granular structures, and a weaker membrane labeling, relative to cells expressing B₂R-GFP. However, the mutant construction B₂R-GFP S/T failed to translocate into the cells under the form of multiple endocytic vesicles following treatment with BK, whereas B₂R-GFP did so in a reversible manner (Fig. 8).

View larger version (45K): [in this window] [in a new window]   Fig. 8.   Lack of BK-induced internalization of B2R-GFP S/T in transiently transfected HEK 293 cells maintained in the complete culture medium and treated with cycloheximide (71 µM). For comparison, other cells were transiently transfected with B2R-GFP. Presentation as in Fig. 2.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Receptor conjugates formed by the fusion of a G protein-coupled receptor with GFP have been used successfully for a number of applications (Milligan, 1999). As reported for other receptors, the rabbit B₂R-GFP conjugates showed an excellent conservation of function, because [³H]BK affinity remained very close to that of the wild-type B₂R. In addition, B₂R-GFP retained affinity for typical B₂R antagonists and mediated arachidonate release in response to BK (Houle et al., 2000). Another receptor-mediated effect, endocytosis, was observed with B₂R-GFP, and this was prevented by the nonpeptide B₂R antagonist LF 16.0687, but not by a B₁R antagonist (Fig. 4). The 101 kDa estimated molecular mass of B₂R-GFP (major band; Fig. 5) is consistent with the literature data. Indeed, this figure, minus the mass of GFP (27 kDa), gives an approximate molecular mass of 74 kDa for the rabbit B₂R; an estimate of 69 kDa has been previously given for the human form of this receptor (Blaukat et al., 1996). The large difference between the predicted peptide molecular mass (41.8 kDa for the rabbit sequence) and the found figures suggests that B₂Rs are heavily glycosylated, even under the GFP conjugate form.

The present results extend previous findings on agonist-induced B₂R endocytosis, and essentially show that all the internalized receptors are recycled to the cell surface, an outcome that was not definite so far (Blaukat et al., 1996). The present experiments also evidence the formation and disappearance of endocytic vesicles containing fluorescent receptors in the cells (confocal microscopy). Additionally, a dense cellular fraction containing the endosomes and lysosomes is enriched in immunoreactive B₂R-GFP following agonist stimulation (Fig. 6). In cells stimulated with tritiated BK, this fraction contains the radioactive agonist for a time period commensurate with the recycling of the receptor. [³H]BK was shown to be extensively hydrolyzed following receptor-mediated internalization in intact cells, and fragments are released as a function of time (Munoz and Leeb-Lundberg, 1992). The cell fraction transiently containing [³H]BK and B₂R-GFP in our experiments (Fig. 6) also contained the early endosome marker rab5 (Gorvel et al., 1991), observed in the immunoblots as a 25-kDa band that was not enriched by agonist stimulation (data not shown; primary monoclonal antibody from Synaptic Systems GmbH, Göttingen, Germany). However, the applied cell fractionation method is limited, because the 15,000g pellet contains other types of organelles (Lodish et al., 1995). Further analysis of the system using a colocalization approach supports the targeting of agonist-stimulated B₂R-GFP to the early recycling endosomes (labeled with transferrin), but not to the lysosome-bound late endosomes (labeled with LDL; Fig. 7), and in a time frame similar to the ones used in other systems (Pitts et al., 1999; Oksche et al., 2000).

ACE is absent from HEK 293 cells as assessed by a sensitive assay based on a tritiated substrate (data not shown; kit from Hycor Biomedical, Irvine, CA). ACE inhibition, under the form of captopril administration, has been used to increase the half-life of BK in the serum-containing medium, and ultimately to control the duration of B₂R-GFP internalization. A prolonged agonist stimulation (3 h) in the presence of captopril has been shown to induce persistent translocation of the receptor fluorescence into the cells. This is comparable to the persistent (6 h) loss of surface B₂R in IMR-90 fibroblasts exposed to BK in serum-free medium (Phagoo et al., 1999). It should be noted that the use of an ACE inhibitor in the present study has no conceptual relationship with the postulated cross talk between drug-occupied ACE and the B₂R in cells that coexpress both types of molecules (and, more specifically, in cells cotransfected with expression vectors for both molecules; Minshall et al., 1997; Benzing et al., 1999; Marcic et al., 1999). In these cases ACE inhibition rather limits agonist-induced B₂R desensitization and internalization by an unknown mechanism. Alternate serum kininase pathways probably explain the slow degradation of BK in the presence of captopril (Fig. 3). By using a separate enzyme immunoassay for des-Arg⁹-BK (Raymond et al., 1995), we have observed the accumulation of this metabolite produced by carboxypeptidase N in the presence of captopril (amounting to about 30% of the initial BK concentration over 50 min). However, the production of des-Arg⁹-BK was comparatively less abundant in the absence of captopril (data not shown).

We have conducted essentially all experiments on B₂R conjugate endocytosis in the presence of cycloheximide, to avoid interference from newly synthesized fluorescent receptors. This treatment was perhaps associated with a minor loss of B₂R-GFP density over 3 h; however, adding the agonist (even in the presence of captopril) was not associated with actual down-regulation (Fig. 5A). Thus, agonist-induced down-regulation, as opposed to desensitization, may not exist at the protein level for B₂Rs. This conclusion may be limited by the time frame of the experiments; however, a large down-regulation of ₂-adrenoceptors is observed at the protein level after 3 h of exposure to the agonist in the presence of cycloheximide (Jockers et al., 1999). It could be argued that GFP conjugation interferes with the membrane proteolytic mechanisms postulated to be recruited in ₂-receptor down-regulation. On the other hand, a fusion protein consisting of the endothelin B receptor conjugated with GFP is irreversibly sorted to the late endosomal/lysosomal pathway in transfected cells stimulated with the corresponding agonist, suggesting that GFP does not protect from this form of down-regulation (Oksche et al., 2000). Current evidence rather suggests that the effect of GFP conjugation is to slow down endocytic processes without qualitatively changing endocytic pathways for adrenoceptor subtypes (McLean and Milligan, 2000). However, it cannot not be excluded that BK B₂R down-regulation observed in pathology may result from transcriptional suppression of the gene, or other mechanisms different from agonist-induced endocytosis. A trypsin treatment produced a 28-kDa fragment detectable by the anti-GFP antibodies (Fig. 5B). This product is smaller than any of the C-terminal fragments predicted by the presence of Lys or Arg residues in the extracellular domains of the rabbit B₂R (all above 38 kDa). Cell treatment with the alternate serine protease endoproteinase Lys-C produced similar results (23-31-kDa immunoreactive products; data not shown). These preliminary results suggest that limited proteolysis initiated by extracellular enzymes may be swiftly completed by mechanisms endogenous to the cell. In this context, it is of interest that the B₂R may be a substrate for kallikreins (the kinin-forming serine proteases; Hecquet et al., 2000). Agonist-induced receptor endocytosis may protect receptors from degradation, as suggested by the reduction of trypsin effect in cells pretreated with BK and captopril (Fig. 5B).

Experiments involving the mutation of putative GRK-2 phosphorylation sites in the C-terminal tail of the B₂R seem to confirm their importance for agonist-induced endocytosis of B₂R-GFP (Fig. 8). No apparent receptor down-regulation independent from endocytosis was observed with the B₂R-GFP S/T variant. Consistent with the findings of Pizard et al. (1999) on human B₂ receptors, the mutated construction was less effectively expressed at the cell surface than the natural sequence in transiently transfected cells (fluorescence intensity); in addition, we show an increased intracellular content of receptor-associated fluorescence for B₂R-GFP S/T. This altered subcellular distribution could be explained by a defect in the synthesis and maturation of the mutated receptor, or by an accelerated degradation process. On the other hand, it has been claimed that the mutation of five Ser or Thr residues in the C-terminal tail of the human B₂R did not alter in a significant manner the agonist affinity, which we confirm, or receptor signaling (Pizard et al., 1999). B₂R-GFP S/T may be a limited model to assess the mechanism of endocytosis owing to its altered subcellular distribution. A further limitation is that the lower expression level of the mutant makes the capacity of the two-receptor conjugates to internalize difficult to compare.

In addition to the widely distributed B₂R, the kallikrein-kinin system includes a homolog G protein-coupled receptor, termed the B₁R (Marceau and Bachvarov, 1998). Several findings support B₁R importance in late inflammatory events: it is selectively stimulated by a class of abundant kinin metabolites, Lys-des-Arg⁹-BK or des-Arg⁹-BK, but not efficiently by the native kinins Lys-BK or BK. The B₁R is inducible following some types of tissue injury. The regulation of the two receptor subtypes differs at the protein level: the B₁R is not importantly internalized following agonist stimulation, relative to the B₂R (Faussner et al., 1998; Zhou et al., 2000). Accordingly, the B₁R fails to undergo ligand-induced phosphorylation, whereas the B₂R is phosphorylated in comparative experiments based on Sf9 cells (Blaukat et al., 1999). Finally, the B₁R is more resistant to functional desensitization than the B₂R in cell types that coexpress both receptor subtypes (for review, see Marceau and Bachvarov, 1998).

In summary, agonist stimulation of GFP conjugates of BK B₂R allowed to observe receptor endocytosis, a phenomenon followed by extensive recycling to the cell surface but no significant receptor down-regulation. Digestion initiated by extracellular proteases may be involved in pathological B₂R down-regulation, as suggested by the simulation involving trypsin.