657

NOTE

A bioactive peptide from the transmembrane 5 intracellular loop 3 region of the human 5HT1a receptor
Kenneth Hayataka, Mary-Frances OConnor, Nancy Kinzler, John T. Weber, and Keith K. Parker

Abstract: 15 amino acid peptide from the transmembrane 5 intracellular loop 3 region of the human 5HT1a receptor produced concentration-dependent decreases in agonist binding. This result is consistent with a competitive interaction between peptide, receptor, and G protein at the receptor G protein interface. Bombesin and a 13 amino acid peptide from the carboxyl terminus region of the receptor were inactive. Additionally, the peptide decreased forskolin-mediated cAMP elevation. Overall, these results suggest that amino acid residues from this region of the receptor are involved in receptor G protein coupling and that G protein is activated by the receptor. Key words: serotonin, 5HT1a receptor, G protein, cAMP, loop peptide. Rsum : Un peptide de 15 acides amins de la rgion de lhlice transmembranaire 5 boucle intracellulaire 3 du rcepteur 5HT1a humain entrane une diminution de la liaison dun agoniste, en fonction de sa concentration. Ce rsultat est en accord avec une comptition entre le peptide, le rcepteur et la protine G linterface rcepteur protine G. La bombsine et un peptide de 13 acides amins de la rgion C-terminale du rcepteur nont pas deffet. De plus, le peptide entrane une diminution de laugmentation de la concentration de lAMPc induite par la forskoline. Globalement, ces rsultats suggrent que des rsidus dacides amins de cette rgion du rcepteur interviennent dans le couplage du rcepteur et de la protine G permettant lactivation de la protine G par le rcepteur. Mots cls : srotonine, rcepteur 5HT1a, protine G, AMPc, peptide en boucle. [Traduit par la Rdaction] Note 660

A number of receptor regions have been implicated as contact sites between receptor and G protein in the seven transmembrane receptor (7TMR) superfamily (Savarese and Fraser 1992; Caron and Lefkowitz 1993; Strader et al. 1994; Baldwin 1994). Using a variety of techniques including sitedirected mutagenesis and receptor chimeras, investigators have particularly identified the transmembrane 5 (TM5)
Received February 11, 1998. Revised June 23, 1998. Accepted July 9, 1998. Abbreviations: IBMX, isobutylmethylxanthine; 7TMR, seven transmembrane receptor; TM5, transmembrane 5; 5HT, 5hydroxytryptamine. K. Hayataka, M.-F. OConnor, N. Kinzler, J.T. Weber,1 and K.K. Parker.2 Department of Pharmaceutical Sciences, School of Pharmacy, University of Montana, Missoula, MT 59812-1075, U.S.A.
1

intracellular loop 3 (i3) region of some 7TMR as a vital link to the G protein (Oksenberg et al. 1995; Burstein et al. 1996). To partially test this concept in the human serotonin (5-hydroxytryptamine; 5HT) 1a receptor (H5HT1aR) (Fargin et al. 1988), we have synthesized a 15 amino acid peptide (15MER) from the known primary structure of the receptor (Kobilka et al. 1987). We have used the 15MER in agonist binding inhibition assays (Peroutka et al. 1979; Maguire et al. 1976) and in determinations of cellular cyclic AMP to test for biological activity in this negatively coupled receptor system (Fargin et al. 1989). The results reported here implicate the TM5/i3 region of the H5HT1aR in G protein linkage while providing a tool in the 15MER for further exploration of the receptor G protein interface.

Peptide synthesis
Peptides were synthesized at the Murdock Molecular Biology Facility at the University of Montana using standard automated solid phase techniques with an ABI 431A peptide synthesizer. The 15MER has the following sequence (N-terminal to the left): IFRAARFRIRKTVKK. A 13MER from the C-terminus of the receptor near TM7 was also synthesized, with a sequence of
1998 NRC Canada

Present address: Virginia Medical College, Virginia Commonwealth University, Richmond, VA 23298-0613, U.S.A. 2 Author to whom all correspondence should be addressed (email: kparker@lewis.umt.edu).
Biochem. Cell Biol. 76: 657660 (1998)

658 Fig. 1. Concentration-dependent inhibition of [3H]8-OH-DPAT binding by 15MER in whole cells expressing the H5HT1aR (to the left of the control value). Concentrations are expressed as log M. The effect of bombesin at 105 and 108 M is shown on the right of the control value. Control is buffer alone. Results are given as the mean SEM for four experiments.

Biochem. Cell Biol. Vol. 76, 1998 Fig. 2. Concentration-dependent inhibition of [3H]8-OH-DPAT binding by 15MER in membranes expressing the H5HT1aR. Results are given as the mean SEM for five experiments.

KDFQNAFKKIIKC. To view these peptide regions in the context of the full sequence, readers are referred to Kobilka et al. (1987) and Savarese and Fraser. (1992), or in schematic form to Baldwin (1994), Strader et al. (1994), and Varrault et al. (1994). Following synthesis, peptides were subjected to chromatography on a Waters 625 HPLC system. Bombesin was purchased from either Calbiochem (LaJolla, Calif.) or Sigma Chemical Co. (St. Louis, Mo.). Peptides were stored desiccated at 20C.

Cell culture
CHO-K1 cultures transfected with the human 5HT1a receptor gene are the generous gift of John Raymond (Fargin et al. 1988, 1989). Cells were grown in monolayer cultures with Hams F-12 medium (Gibco) fortified with 10% fetal bovine serum and containing 200 g/mL geneticin. Culturing conditions were 37C, in a humidified atmosphere of 5% CO2, 95% air. Subculturing occurred about every 7 days, with 0.25% trypsin in physiological saline. When cells were used for assay, trypsinized cells were diluted in F-12 medium containing 10% fetal bovine serum and centrifuged for 10 min at 1000 rpm at 4C. The pellet was then rinsed in Earles Balanced Salt Solution and recentrifuged. In whole cell assays, this pellet was used directly by dilution in binding buffer (50 mM Tris, pH 7.4; 4 mM CaCl2; 10 M pargyline). In assays using prepared membranes, the pellet was homogenized on glass, followed by centrifugation at 450 000 gmin at 4C. The resulting pellet was then rehomogenized in binding buffer and used in the assay. In assays using solubilized receptor, the whole cell pellet was gently mixed for 1 h at 4C in buffer containing 0.6% sodium cholate, 20 mM Tris, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, and 100 mM NaCl (Mulheron et al. 1994). The mixture was centrifuged for 300 000 gmin at 4C. The resulting supernatant was diluted in binding buffer (1 to 30) and used in the assay. Receptor binding assays using the 5HT1a agonist [3H]8-OHDPAT ([ 3H]8-hydroxy-2-(di-n-propylamino)tetralin) (NEN) utilized conditions modified from a number of established protocols (Fargin et al. 1988; Pierce and Peroutka 1989; Arthur et al. 1993; Weber et al. 1997). Receptor preparation (700 L) was added to either 100 L of binding buffer (total counts) or 100 L of 5HT (105 M final concentration; for nonspecific binding), and either 100 L of binding buffer (control) or 100 L of binding buffer containing various concentrations of peptides, and 100 L of [3H]8-OH-DPAT (0.5 nM final concentration) in binding buffer. In

all cases, the total volume was 1.0 mL. Tubes were incubated at 30C for 30 min. All determinations were done in triplicate and were terminated by addition of 4 mL of ice-cold Tris buffer (50 mM, pH 7.4), followed by rapid vacuum filtration on GF/B (Whatman) filters (GF/F in the case of solubilized receptor). Filters were then rinsed with two successive 5-mL aliqouts of Tris buffer. Filters were counted in 5 mL of Ecoscint (National Diagnostics) in a Beckman LS 6500 liquid scintillation counter. In the case of experiments using solubilized receptor, parallel trials were conducted initially in which agonist bound receptor was separated from free agonist using gel filtration (Sephadex G- 50). Identical results were obtained.

cAMP determination
Cells were grown to confluency in 24-well plates. Medium was removed, and the cells were rinsed with two successive 1-mL volumes of serum-free F-12. Drugs were dissolved in water (5HT, 15MER) or in ethanol (forskolin, isobutylmethylxanthine (IBMX)) and then further diluted in serum-free F-12. Wells were treated with 0.5 mL of drug at the following final concentrations: 5HT (105 M), forskolin (3 105 M), IBMX (104 M), and 15MER (10 5 M). Incubation was for 20 min at 37C. All wells received IBMX. Reaction was terminated by addition of 0.3 mL 10 1 M HCl. Cells were scraped from the well bottoms and homogenates were centrifuged for 15 min at 4C at 2000 g. The resulting supernatant was extracted four times with five volumes of water-saturated ether. The aqueous remnant was then dried in a 60C oven. Residues were redissolved in 0.05 M acetate buffer, pH 5.8, and aliquots were assayed in duplicate or triplicate by radioimmunoassay (Amersham), using overall procedures parallel to those of Reppert et al. (1994).

Agonist inhibition studies

Agonist inhibition When whole cells containing the human 5HT1a receptor were assayed in the presence of 0.5 nM of the agonist [ 3 H]8-OH-DPAT, the 15MER produced concentrationdependent inhibition of binding (Fig. 1). On the other hand, parallel experiments in which the bioactive peptide bombesin (which is similar in length to the 15MER) was substituted for the 15MER produced no observable inhibition. Similarly, when agonist binding experiments were conducted with membranes containing the H5HT1aR (Fig. 2) or solubilized H5HT1aR (data shown in Fig. 3), concentration 1998 NRC Canada

Note Fig. 3. Inhibition of [3H]8-OH-DPAT binding by 105 M 15MER in various 5HT1aR preparations. Full concentration-dependency was examined in each experiment as shown in Fig. 1. Results in this figure are at the highest concentration used in each case, with sample sizes for each preparation of 4 (rabbit brain membrane), 4 (human whole cell), 5 (human membrane), and 4 (human solubilized).

659 Fig. 4. Inhibition of cAMP formation by 5HT and 15MER in whole cells expressing the H5HT1aR. Results are from three experiments with total sample sizes of 6 (basal: IBMX alone), 15 (15MER), 6 (5HT), 6 (5HT + 15MER), and 15 (forskolin alone).

dependent inhibition by the 15MER was observed. In each case, bombesin was unable to inhibit agonist binding. When identical experiments were conducted using membranes prepared from rabbit cerebral cortex, known to contain the 5HT1aR (Weber et al. 1997), the 15MER produced concentration-dependent inhibition. The comparative results of these experiments showing inhibition produced at the highest concentration of 15MER used (105 M) are shown in Fig. 3. The amount of inhibition produced in rabbit brain is different from that produced in solubilized human receptor (p < 0.05). All other preparations were not significantly different from each other. The pIC50 values were 5.0 for rabbit brain, 5.3 for human whole cell, 5.5 for human membrane, and 6.3 for human solubilized receptor. Use of solubilized receptor circumvented potential difficulties of delivering the 15MER across the cell membrane to the putative site of action at the G protein interface. The 15MER was nearly as active in preparations using membranes and whole cells as solubilized receptor. Thus, it appears that this amphipathic peptide has little trouble achieving intracellular access. The exact mechanism by which a peptide such as 15MER achieves intracellular access is not known; however, there is precedent for similar positively charged peptides, such as the venom melittin, for moving across membranes, perhaps by electrostatic attraction (Mousli et al. 1990). The 15MER is active against 5HT1aR from another species (rabbit) in which the receptor has not been cloned (Weber et al. 1997), creating the possibility that the function of this region has been conserved to some degree during evolution of the 5HT1aR. The rat and human 15MERs are identical except that an arginine is found in rat rather than a lysine in human at position two from the C-terminus of the peptide (Savarese and Fraser 1992). The specificity of 15MER activity in the context of 5HT1aR/G protein interactions has not been completely identified by these studies; it is possible that 15MER could interact, at least weakly, with closely related 5HTRs or even other receptors such as the beta adrenergic receptor. On the

other hand, as structural studies become more sophisticated, differences in loop sequence and conformation between receptors will undoubtedly be significant parameters in coupling and G protein activation specificity. A parallel series of experiments was conducted using the 13MER to concentrations as high as 104 M in all three human receptor preparations (whole cell, membrane, solubilized). Studies in other 7TMR suggested possible involvement of this C-terminal region in receptor G protein coupling (Savarese and Fraser 1992; Caron and Lefkowitz 1993; Strader et al. 1994; Baldwin 1994). In no case was the 13MER able to inhibit agonist binding in our studies. The agonist inhibition assays were modeled after the effects produced by GTP and its stable analogues (Peroutka et al. 1979; Maguire et al. 1976). When GTP binds to the G protein agonist bound receptor (high affinity) complex, the complex dissociates. Since high receptor affinity is dependent on receptor G protein coupling, the dissociated receptor is in the low-affinity state, observable by inhibition of agonist binding in a concentration (GTP)-dependent manner. We hypothesize that any agent (e.g., the 15MER) that has the potential to uncouple receptor and G protein has the capacity to drive receptors toward the lower affinity state. The results reported here for the 15MER are consistent with such an effect in which the 15MER competes with receptor for a binding surface on the G protein, effectively uncoupling the receptor G protein complex at higher peptide concentrations. Cellular cAMP levels The effects of 5HT and the 15MER were determined in whole-cell preparations in three separate experiments (Fig. 4). All wells contained IBMX; in addition, all other drug treatments contained forskolin. The 15MER alone produced a small but significant (p < 0.01) drop in forskolin-elevated cellular cAMP. 5HT alone produced a larger decrease in forskolin-elevated cellular cAMP (p < 0.01 compared with forskolin alone). The 15MER plus 5HT produced an even larger decrease in forskolin-elevated cellular cAMP (p < 0.01 compared with forskolin alone). The level for 5HT plus the 15MER is not statistically different from that for 5HT alone. Thus, at this time, it cannot be
1998 NRC Canada

660

Biochem. Cell Biol. Vol. 76, 1998 structure, function, and regulation. Recent Prog. Horm. Res. 48: 277290. Dratz, E.A., Furstenau, J.E., Lambert, C.G., Thireault, D.L., Rarick, H., Schepers, T., Pakhlevaniants, S., and Hamm, H.E. 1993. NMR structure of a receptor-bound G-protein peptide. Nature (London), 363: 276281. Fargin, A., Raymond, J.R., Lohse, M.J., Kobilka, B.K., Caron, M.G., and Lefkowitz, R.J. 1988. The genomic clone of G-21 which resembles a beta-adrenergic receptor sequence encodes the 5HT1a receptor. Nature (London), 335: 358360. Fargin, A., Raymond, J.R., Regan, J.W., Cotecchia, S., Lefkowitz, R.J., and Caron, M.G. 1989. Effector coupling mechanisms of the cloned 5HT1a receptor. J. Biol. Chem. 264: 14 848 14 852. Gettys, T.W., Fields, T.A., and Raymond, J.R. 1994. Selective activation of inhibitory G-protein alpha subunits by partial agonists of the human 5-HT1a receptor. Biochemistry, 33: 42834290. Herlitze, S., Hockerman, G.H., Scheuer, T., and Catterall, W.A. 1997. Molecular determinants of inactivation and G protein modulation in the intracellular loop connecting domains I and II of the calcium channel alpha1a subunit. Proc. Natl. Acad. Sci. U.S.A. 94: 15121516. Kobilka, B.K., Frielle, T., Collins, S., Yang-Feng, T., Kobilka, T.S., Francke, U., Lefkowitz, R.J., and Caron, M.G. 1987. An intronless gene encoding a potential member of the family of receptors coupled to guanine nucleotide regulatory proteins. Nature (London), 329: 7579. Maguire, M.E., Van Arsdale, P.M., and Gilman, A.G. 1976. An agonist specific effect of guanine nucleotides on binding of the beta adrenergic receptor. Mol. Pharmacol. 12: 335339. Mousli, M., Bueb, J.-L., Bronner, C., Rouot, B., and Landry, Y. 1990. G protein activation: a receptor-independent mode of action for cationic amphiphilic neuropeptides and venom peptides. Trends Pharmacol. Sci. 11: 358362. Mulheron, J.G., Casanas, S.J., Arthur, J.M., Garnovskaya, M.N., Gettys, T.W., and Raymond, J.R. 1994. Human 5HT1a receptor expressed in insect cells activates endogenous Go-like G protein(s). J. Biol. Chem. 269: 12 954 12 962. Oksenberg, D., Havlik, S., Peroutka, S.J., and Ashkenazi, A. 1995. The third intracellular loop of the 5-hydroxy-tryptamine2A receptor determines effector coupling specificity. J. Neurochem. 64: 14401447. Peroutka, S.J., Lebovitz, R.M., and Snyder, S.H. 1979. Serotonin receptor binding sites affected differentially by guanine nucleotides. Mol. Pharmacol. 16: 700708. Pierce, P.A., and Peroutka, S.J. 1989. Hallucinogenic drug interactions with neurotransmitter receptor binding sites in the human cortex. Psychopharmacology, 97: 118122. Reppert, S.M., Weaver, D.R., and Ebisawa, T. 1994. Cloning and characterization of a mammalian melatonin receptor that mediates reproduction and circadian responses. Neuron, 13: 1177 1185. Savarese, T.M., and Fraser, C.M. 1992. In vitro mutagenesis and the search for structurefunction relationships among G protein coupled receptors. Biochem. J. 283: 119. Strader, C.D., Fong, T.M., Tota, M.R., Underwood, D., and Dixon, R.A.F. 1994. Structure and function of G protein-coupled receptors. Annu. Rev. Biochem. 63: 101132. Varrault, A., Le Nguyen, D., McClue, S., Harris, B., Jouin, P., and Bockaert, J. 1994. 5-Hydroxytryptamine1a receptor synthetic peptides. J. Biol. Chem. 24: 16 720 16 725. Weber, J.T., Hayataka, K., OConnor, M.-F., and Parker, K.K. 1997. Rabbit cerebral cortex 5HT1a receptors. Comp. Biochem. Physiol. C, 117: 1924.
1998 NRC Canada

determined whether action of the 15MER and 5HT is additive or not. Both 5HT alone and 5HT plus 15MER give statistically lower values than 15MER alone (p < 0.01). 15MER out of the context of the entire i3 loop conformation may have a limited ability to stimulate G protein compared with the native loop or all receptor contact points may be required for full activation. The effect of the 15MER peptide alone on forskolinelevated c AMP levels suggests that the 15MER has some potential to mimic the action of agonist-stimulated receptor on G protein activation. Overall, the results reported here add to the growing body of information linking the i3 loop region of the receptor to coupling and activation of G protein. The studies of Varrault et al. (1994) demonstrated involvement of the TM6 portion of i3. The results presented here specifically implicate the TM5 end of i3 as reported in other 7TMRs (Oksenberg et al. 1995; Burstein et al. 1996). Further studies are required to demonstrate direct binding of the 15MER to purified preparations of G proteins (Gettys et al. 1994) and subsequent activation. Until direct binding of the 15MER to G protein is demonstrated, other sites of action cannot be eliminated. Additionally, structureactivity studies with the 15MER should be useful in determining the length and sequence of peptide optimally required for biological activity. Recently, synthetic peptides have been shown to be valuable as tools for elucidation of biologically active surfaces such as that between G protein and adenylyl cyclase of calcium channels (Herlitze et al. 1997). Thus, peptides such as the 15MER may be useful as probes of solution structures for biological surfaces such as that between receptor and G protein using multidimensional NMR (Dratz et al. 1993). The 15MER may be particularly interesting in that the apparent differential magnitude effects of uncoupling versus G protein activation may be exploitable in future structural experiments.

We are indebted to Dr. John Raymond for the CHO cell line expressing the human 5HT1a receptor. We gratefully appreciate the excellent work of Joan Strange of the University of Montanas Murdock Molecular Biology Facility for peptide synthesis. The work was greatly aided by the excellent technical assistance of Justin Iverson, Rene Potts, Shellene Rodriguez, and Joe Sure Chief. The work was supported by grants from National Science Foundation Experimental Program for Stimulation of Competitive Research (EPSCoR) and the National Institutes of Health (RR10169 and GM/OD54302-01).

Arthur, J.M., Casanas, S.J., and Raymond, J.R. 1993. Partial agonist properties of rauwolscine and yohimbine for the inhibition of adenylyl cyclase by recombinant human 5-HT1a receptors. Biochem. Pharmacol. 45: 23372341. Baldwin, J.M. 1994. Structure and function of receptors coupled to G proteins. Curr. Opin. Cell Biol. 6: 180190. Burstein, E.S., Spalding, T.A., and Brann, M.R. 1996. Amino acid side chains that define muscarinic receptor/G-protein coupling. J. Biol. Chem. 271: 28822885. Caron, M.G., and Lefkowitz, R.J. 1993. Catecholamine receptors: