Search

The mammalian bombesin receptor subfamily of G protein-coupled receptors currently consists of the gastrin-releasing peptide receptor (GRP-R), neuromedin B receptor, and bombesin receptor subtype 3. All three receptors contain a conserved aspartate residue (D98) at the extracellular boundary of transmembrane domain II and a conserved arginine residue (R309) near the extracellular boundary of transmembrane domain VII. To evaluate the functional role of these residues, site-directed GRP-R mutants were expressed in fibroblasts and assayed for their ability to both bind agonist and catalyze exchange of guanine nucleotides. Alanine substitution at GRP-R position 98 or 309 reduced agonist binding affinity by 24- and 56-fold, respectively, compared to wild-type GRP-R. Single swap GRP-R mutations either resulted in no receptor expression in the membrane (D98R) or the protein was not able to bind agonist (R309D). In contrast, the double swap mutation (D98R/R309D) had high-affinity agonist binding, reduced from wild-type GRP-R by only 6-fold. In situ reconstitution of urea-extracted membranes expressing either wild-type or mutant (D98A or R309A) GRP-R with G(q) indicated that alanine substitution greatly reduced G protein catalytic exchange compared to wild-type GRP-R. The D98R/R309D GRP-R had both a higher intrinsic basal activity and a higher overall catalytic exchange activity compared to wild-type; however, the wild-type GRP-R produced a larger agonist-stimulated response relative to the double swap mutant. Taken together, these data show that GRP-R residues D98 and R309 are critical for efficient coupling of GRP-R to G(q). Furthermore, our findings are consistent with a salt bridge interaction between these two polar and oppositely charged amino acids that maintains the proper receptor conformation necessary to interact with G proteins.

Recently, a fourth member of the bombesin (Bn) receptor family (fBB4-R) was isolated from a cDNA library from the brain of the frog, Bombina orientalis. Its pharmacology and cell biology are largely unknown, and no known natural cell lines or tissues possess sufficient numbers of fBB4-R's to allow either of these to be determined. To address these issues, we have used three different strategies. fBB4-R expression in cells widely used for other Bn receptor subtypes was unsuccessful as was expression in two frog cell lines. However, stable fBB4-R cell lines were obtained in CHO-K1 cells which were shown to faithfully demonstrate the correct pharmacology of the related Bn receptor, the GRP receptor, when expressed in these cells. [DPhe6,betaAla11,Phe13,Nle14]Bn(6-14) was found to have high affinity (Ki = 0.4 nM) for the fBB4 receptor and 125I-[DTyr6,betaala11,Phe13,Nle14]Bn(6-14) to be an excellent ligand for this receptor. The fBB4-R had a unique pharmacology for naturally occurring Bn-related agonists, with the presence of a penultimate phenylalanine being critical for high-affinity interaction. It also had a unique profile for six classes of Bn antagonists. The fBB4-R was coupled to phospholipase C with activation increasing [3H]inositol phosphates and mobilizing Ca2+ almost entirely from cellular sources. There was a close correlation between agonist the receptor occupation and the receptor activation. Three of the five classes of Bn receptor antagonists that interacted with higher affinity with the fBB4-R functioned as fBB4-R antagonists and two as partial agonists. fBB4-R activation stimulated increases in phospholipase D (PLD) over the same range of concentrations at which it activated phospholipase C. These results demonstrate that the fBB4 receptor has a unique pharmacology for agonists and antagonists and is coupled to phospholipase C and D. The availability of these cell lines, this novel ligand, and the identification of three classes of antagonists that can be used as lead compounds should facilitate the further investigation of the pharmacology and cell biology of the BB4 receptor.

Whereas baculovirus expression systems have been extensively used for high-level expression of steroid receptors and receptors coupled to adenylate cyclase, there are few studies on peptide receptors coupled to phospholipase C (PLC). In the present study we have expressed the murine gastrin-releasing peptide receptor (mGRP-R) in Sf9 cells using a recombinant baculovirus and characterized it structurally and functionally. mGRP-R was detectible 12 h post infection with recombinant baculovirus carrying mGRP-R cDNA and became maximal at 60 h post infection (Bmax = 6 pmol/mg protein), which is a 4-60-fold greater density than is found in native tissues. The mGRP-R in Sf9 cells assessed by affinity labeling or immunoblotting was smaller than that in native tissues (M(r) = 51 kD vs 82 kD), and the difference was due to the extent of glycosylation. In Sf9 cells the mGRP-R had at least two of the four potential extracellular glycosylation sites glycosylated, whereas in the native receptor all four were approximately equally glycosylated. In Sf9 cells the glycosylation was entirely biantennary complex, in contrast to the native mGRP-R, where it was entirely tri- and tetraantennary complex N-linked oligosaccharides. Affinity labeling studies revealed a band with an apparent molecular mass about 40 kDa higher than the 51-kDa mGRP-R band. The intensity of this band correlated with the extent of functional G protein coupling, suggesting that it may represent an mGRP-R-G protein complex. In binding studies the affinity of the mGRP-r in Sf9 cells for the agonists bombesin (Bn), GRP, and neuromedin B (NMB) varied differently with infection time: with Bn the affinity decreased 3-fold with longer infection times, with GRP it remained unchanged, and with NMB it decreased 10-fold. GPP(NH)p inhibited binding of either [125I]Tyr4Bn or [125I]GRP at 24 h post infection, but not at 96 h post infection. Agonists activated PLC, increasing both [3H]IP and [Ca2+]i; however, the efficacy of each agonist decreased with infection time. These results demonstrate that by the use of recombinant baculovirus infected Sf9 cells the PLC-linked receptor mGRP-R can be expressed in amounts significantly greater than those in native tissues. The mGRP-R expressed in these Sf9 cells is incompletely glycosylated and has less complex N-linked oligosaccharide chains, yet it is fully coupled to G proteins and activates phospholipase C, similar to the native receptor, if short infection times are used.(ABSTRACT TRUNCATED AT 400 WORDS)

In the present study, we investigated the nature and the importance of glycosylation of two mammalian bombesin receptors, the neuromedin B receptor (NMB-R) and the gastrin-releasing peptide receptor (GRP-R), using chemical cross-linking and enzymatic deglycosylation. [125I]-(D-Tyr0)NMB cross-linked to native NMB-R on rat C-6 glioblastoma cells or rat NMB-R transfected into BALB 3T3 cells revealed a single broad band, M(r) = 63,000, on both cell types that was not altered by DTT. NMB inhibited cross-linking specifically and saturably with an IC50 of 4.8 and 6.1 nM for C-6 and NMB-R transfected cells, respectively, and there was a close correlation between its ability to inhibit binding and its ability to inhibit cross-linking. A single broad band of M(r) = 82,000 was cross-linked with [125I]GRP on mouse GRP-R transfected BALB 3T3 cells. Peptide-N4-(N-acetyl-beta- glucosaminyl)asparagine amidase F (PNGase F) digestion increased the mobility of the original band in C-6, NMB-R, and GRP-R transfected cell membranes. Endoglycosidase H (Endo-H) and endoglycosidase F2 (Endo-F2) digestion had no effect on both transfected cells. Neuraminidase digestion slightly increased the mobility of the original band in NMB-R transfected cell membranes; however, it had no effect on GRP-R transfected cell membranes. Endo-alpha-N-acetylglucosaminidase (O-glycanase) digestion subsequent to neuraminidase treatment showed no additional effect on either receptor. Serial partial deglycosylation of cross-linked NMB-Rs with PNGase F treatment for different incubation periods revealed one band of partially glycosylated receptor (53 kDa) besides the fully glycosylated and fully deglycosylated ones, showing that NMB-R has two oligosaccharide chains. Similarly, three partially deglycosylated species (72, 62, and 52 kDa) are seen with the GRP-R, indicating that the GRP-R has four oligosaccharide chains. Treatment of unlabeled membranes with PNGase F followed by affinity labeling resulted in fully deglycosylated NMB-R or 75% deglycosylated GRP-R. Deglycosylation of the NMB-R did not alter its affinity for NMB or alter G-protein coupling; however, 75% deglycosylation of the GRP-R both decreased its affinity for GRP and altered its ability to couple to G-proteins. The present results demonstrate that NMB-R on native and transfected cells is an N-linked sialoglycoprotein with two triantenary and/or tetraantenary complex oligosaccharide chains. The apparent M(r) of this sialoglycoprotein is 63,000, and this protein does not contain disulfide-linked subunits or O-linked carbohydrates.(ABSTRACT TRUNCATED AT 400 WORDS)