The SK-N-MC cell line expresses an orexin binding site different from recombinant orexin 1-type receptor Heike A. Wieland 1, *, Richard M. So¨ll 2,3 , Henri N. Doods 1 , Dirk Stenkamp 1 , Rudolf Hurnaus 1 , Ba¨ rbel La¨ mmle 1 and Annette G. Beck-Sickinger 2 1 Division of Preclinical Research, Boehringer Ingelheim Pharma KG, Biberach, Germany; 2 Institute of Biochemistry, University of Leipzig, Germany; 3 Department of Applied Biosciences, Swiss Federal Institute of Technology, Zurich, Switzerland Orexin A and B (also known as hypocretins), two r ecently discovered neuropeptides, play an important role in food intake, sleep/wake cycle and neuroendocrine functions. Orexins are endogenous ligands of two G-protein-coupled receptors, termed OX 1 and OX 2 . This work presents the first short orexin A and B analogues, orexin A 23–33 and orexin B 18–28, with high affinity (119 ± 49 and 49 ± 23 n M )for OX 1 receptors expressed o n SK-N-MC cells and indicates the importance of the C-terminal part of the o rexin peptides for this ligand–receptor interaction. However, t hese C-ter- minal fragments of orexin did n ot displace the 125 I-labelled orexin B from the recombinant orexin 1 receptor stably expressed in Chinese hamster ovary cells. To examine the role of the shortened orexin A 23–33 in feeding, its e ffects in m imicking or a ntagonizing the effects of o rexin A were studied in rats after administration via the lateral hypothalamus. In contrast with orexin A, which potently induced feeding up to 4 h after administration, orexin A 23–33 neither induced feeding nor inhibited orexin A-induced feeding. Modafinil (VigilÒ), which was shown earlier to activate o rexin neurons, d isplayed binding ne ither to the orexin receptor e xpressed on SK-N-MC cells nor to the r ecombinant orexin 1 r eceptor, which indicates that modafinil d isplays i ts antinarcoleptic action via another yet unknown mechanism. PCR and subsequent sequencing revealed expression of the full-length o rexin 1 receptor mRNA in SK-N-MC and NT-2 cells. Interestingly, sequencing of several cDNA clones derived from RNA of both SK-N-MC and NT-2 cells differed from t he published nucleotide sequence at position 1375. Amino acid prediction of this AfiG change results in an isoleucinefivaline sub- stitution at the protein level, which may provide evidence for an ed iting process. Keywords: food intake; h ypocretin; ligand–receptor inter- action; obesity; orexin. Two novel neuropeptides, orexin A and B, were recently discovered independently by two groups and identified as potent stimulators of food intake after i ntracerebroventric- ular administration [1–4]. Further investigations revealed a broad involvement of these peptides in the regulation of many physiological and behavioural activities that are associated with f eeding behaviour [4–7], in the modulation of neuroendocrine function and the sleep/wake cycle [8–12]. Both peptide amides derive from prepro-orexin, a precursor p rotein produced in defined r egions of the lateral and perifornical hypothalamus, whose mRNA i s up-regu- lated upon fasting. O rexin immunoreactive n eurons are, however, distributed widely in the brain, including regions of the cerebral cortex, the medial groups of the thalamus, the circumventricular organs, the limbic s ystem and the brain s tem [13–15]. A key r ole for orexins in n arcolepsy has been described [10,16–18]. It was shown recently that the anti-narcoleptic drug Modafinil (VigilÒ), the mechanism of action of which is unknown, might act through the orexin pathway [10]. Orexin A consists of 33 amino acids, is C -terminally amidated and contains t wo intramolecular disulfide bonds, that connect cysteine residues from positions 6–12 and 7–14, respectively. Orexin B consists of 28 residues and sh ares 46% i dentity w ith o rexin A, mainly a t the C t ermin us. The three-dimensional solution structure of orexin B was recently determined b y two-dimensional NMR and shows two a helices, connected by a short linker sequence at position 20–23 [19]. The structure of orexin A is conserved among human, rat, mouse and cow, whereas r odent orexin B contains two amino acid substitutions compared with the human sequence: proline i nstead of serine in position two and asparagine instead of s erine in position 18. Xenopus laevis has orexins that differ slightly from the human sequence, but the C-terminal decapeptide of orexin A and B and the positions next to the disulfide bonds in orexin A remain conserved (Fig. 1), which suggests some importance in biological activity of these peptide regions [20]. Orexin A and B are endogenous ligands of two closely related (64% amino acid identity [1]) heptahelical G-protein-coupled receptors, termed OX 1 and OX 2 . They induce an intracellular i ncrease in free Ca 2+ concentration after activation of the receptors [1,21]. Orexin A shows Correspondence to A. G. Beck-Sickinger, Institute of Biochemistry, University of Le i pzig, Talstr. 33, 04103 L e ipzig, Germany. Fax: +49 341 9736 998, Tel.:+ 49 341 9735 901, E-mail: beck-sickinger@uni-leipzig.de Abbreviations:OX 1 receptor, orexin 1 receptor; OX 2 receptor, orexin 2 receptor;NPY,neuropeptideY;HOBt,N-h ydroxybenzotriazole; CHO, Chinese hamster ovary; IC 50 , 50% inhibitory concentration. *Present address: Aventis Phar ma Deutschland G mbH, DG Thrombotic Diseases/Degenerative J o int Diseases, H811, D-65926 Frankfurt, Germany. (Received 2 4 September 200 1, revised 7 December 2001, a ccepted 12 December 2 001) Eur. J. Biochem. 269, 1128–1135 (2002) Ó FEBS 2002 higher affinity to the OX 1 receptor, whereas the binding affinity of the two peptides to OX 2 receptor is in the same order of magnitude [10]. Up to now, little is known about the structure–activity relationship, except for the relevanc e of the C-terminal segment of o rexin A [22]. O nly r ecently, a subtype selective nonpeptide antagonist was described in vitro [23]. We describe here the shortest orexin A and B analogues that bind to OX-receptors. We have also determined that the orexin type 1 receptor is expressed by SK-N-MC cells, a human neuroblastoma cell line, although with a pharma- cological profile different from that of the recombinantly expressed OX 1 receptor. In addition, we describe an amino acid position that differs in clones derived from RNA that has been isolated from SK-N-MC cells. MATERIALS AND METHODS Materials N a -Fmoc-protected amino a cids w ere f rom A lexis (La ¨ ufelfingen, Switzerland). The side-chain protecting groups were tert-butyl for serine and threonine, and trityl for a sparagine and histidine. The 4-(2¢,4¢-dimethoxyphenyl- Fmoc-aminomethyl)-phenoxy (Rink Amide) resin was from Novabiochem (La ¨ ufelfingen). N-hydroxybenzotriazole (HOBt), trifluoroacetic a cid, thioanisole, p -thiocresol, tri- methylsilylbromide, 1,2-ethanedithiol, p iperidine, tert -buta- nol, 1,1,1-trifluoroethanol and dimethylformamide were from Fluka. N,N ¢-diisopropylcarbodiimide w as from Aldrich. Dimethylformamide (pure) and diethylether were from Scharlau (La Jota, Barcelona, Spain). Acetonitrile was from Romil (Cambridge, England). Dulbecco’s modified E agle’s medium was from B ioWhit- taker; OPTI-MEM and Lipofectamine were from G ibco BRL; fetal bovine serum was from BioWhittaker; Hepes was from Fluka; geneticin was from Gibco BRL; Pefabloc SC was from Merck; 125 I-labelled Tyr-human orexin B (specific activity 2130 CiÆmmol )1 ) was from Anawa (Zu ¨ rich, Switzerland); 125 I-labelled Tyr-human orexin A (spec ific activity 2130 CiÆmmol )1 was from NEN; orexin B was from Bachem (Heidelberg, Germany). Modafinil (Vigil Ò ) was from Laboratoire L. Lafon, Merckle, Blaubeuren (Germany), NT-2 cells were from Stratagene. Peptide synthesis The C-terminal undecapeptides of the o rexins, orexin A 23–33 and orexin B 18–28, and the analogues of orexin B and orexin A 23–33 were synthesized by automated multiple solid-phase peptide s ynthesis on a peptide synthe- sizer (Syro, MultiSynTech, Bochum, Germany) using Rink Amide resin (30 mg, resin loading 0.6 mmol Æg )1 ). Amino acids were attached by the Fmoc-strategy in a double coupling procedure, using a 10-fold excess of Fmoc-amino acid, H OBt and N,N ¢-diisopropylcarbodiimide in d imethyl- formamide a nd a reaction t ime o f 4 0 min per coupling. Fmoc-deprotection was accomplished w ith 40% piperidine in dimethylform amide f or 3 min, 2 0% piperidine for 7 min and finally 40% piperidine for a further 5 m in. The orexin A fragment was cleave d from the resin with a mixture o f trifluoroacetic a cid/thioanisole/p-thiocresol (90 : 5 : 5, v/v), precipitated from ice-cold diethylether, collected by ce ntrifugation and washe d four times with diethylether. The methionine-containing orexin B fragment was cleaved from the resin using a mixture of trifluoroacetic acid/thioanisol/ethanedithiol ( 90 : 7 : 3, v/v), precipitated and washed as described. Partial oxidation of the methio- nine residue was reduced by dissolving the p eptide (15 mg, 0.014 mmol) in 1 mL trifluoroacetic acid, followed by the addition of ethanedithiol (15.7 lL, 0.2 mol ÆL )1 )and trimethylsilylbromide (13 lL, 0.1 molÆL )1 ) [24]. The solu- tion was shaken for 40 min at room temperature and the peptide w as precipitated and washed as described. P urifica- tion of the peptide was achieved by preparative H PLC on a C18-column (Waters, 5 lm, 25 · 300 mm) with a linear gradient of 10–30% A in B; A ¼ 0.08% trifluoroacetic acid in acetonitrile, B ¼ 0.1% trifluoroacetic acid in water) and a flow rate of 15 mLÆmin )1 . The peptides were dissolved in tert-butanol/water (1 : 3) and lyophilized. Analytical characterization of the peptides w as achieved by electrospray ionization MS (SSQ 710, Finnigan MAT, Bremen, Germany) and by analytical reversed-phase HPLC on a LiChrospher RP18-column (5 lm, 3 · 125 mm, Merck, Darmstadt, Germany) using linear gradients of 5–50% over 30 min (I), 10–60% over 30 min (II), 10–40% over 30 min (III) or 20–40% over 30 min (IV). Analyt ical data were as expected [orexin A, 23–33: molecular mass (m), m expected 1036 Da; m found , 1036.1 ± 0.5 Da; HPLC reten- tion time (I), 17.3 min; [G23] orexin A 23–33: m expected 1022 Da; m found 1021.5 ± 0.6 Da; HPLC retention time (II), 11.4 min. Orexin B 18–28: m expected 1070 Da; m found 1069.9 ± 0.1 Da; HPLC retention t ime (III), 12.9 min. [L28] orexin B: m expected ,2881Da;m, 2881.3 ± 0.4 Da; HPLC retention time (IV) 16.6 min. cDNA subcloning and nucleotide sequence determination PCR w as used to amplify the full-length o rexin 1 receptor according t o a ccession number AF041243 [1]. Oligonucleo- tides f rom MWG Biotech (Ebersberg, Germany) were used as primers: OX 1 -f, 5¢-GTAGAGCCTAGGATGCCCCT- 3¢;OX 1 -r: 5¢-AGGAAGTGACTTATCCAGAGT-3¢. Total RNA from SK-N-MC cells and NT-2 cells were used as templates. Isolation of total RNA was performed with an RNeasy Total RNA Kit (Qiagen). RT-PCR was Fig. 1. Sequence o f (A) matu re orexin A peptides of human, bovine and rat origin and (B) mature orexin-B peptides. Deviations from the human s equences are underlined. U ¼ pyr oglutamic acid. Ó FEBS 2002 SK-N-MC cell line expresses different orexin binding sites (Eur. J. Biochem. 269) 1129 performed using the Superscript Preamp lification System (Gibco/BRL). After 3 min at 94 °C, the reactions were subjected t o 3 5 cycles of: denaturation, 1 min at 94 °C; annealing, 2 min at 60 °C; elongation 2 min at 72 °Cina primus plus cycler (MWG Biotech). PCR products of the expected size were cloned in pCR2.1TOPO using the TOPO TA Cloning Kit from Invitrogen. The sequence was confirmed using the BigDye Terminator Cycle Sequencing with an ABI 377 Sequencer using the M13 Forward ()20) and Reverse primers (Invitrogen). cDNA from orexin type 1 receptor was from Receptor Biology (Beltsville, MD, USA), sequenced and OX 1 R- cDNA was subcloned into pcDNA3.1/HisA vector from Invitrogen. Cell culture Transfection into Chinese hamster ovary (CHO) cells was performed using the lipofectamine PLUS method according to the m anufacturer’s protocol (Gibco/BRL) using expres- sion plasmids encoding the orexin 1 receptor. Binding assays with transfected cells CHO cells were grown in nutrient m ixture Ham’s F12 medium with 10% fetal bovine serum from BioWhittaker (Boehringer Ingelheim Bioproducts Partnership, Verviers, Belgium), nonessential amino acids, hygromycin B, 2 m M L -glutamine and 1% geneticin (Gibco/BRL) at 37 °Cand 5% CO 2 until they were confluent in a 24-well plate. The medium was aspirated. The cells were washed twice with 0.25 mL NaCl/P i . Incubation buffer [0.2 mL; 84.7 m M NaCl, 3 0 m M KCl, 1.2 m M MgSO 4 Æ7H 2 0, 11.2 m M NaH 2 PO 4 ,bufferedwithHepes,15 m M (4-(2-hydroxyethyl)- 1-piperazine e thanesulfonic acid, pH 7.5; from SERVA, Heidelberg, Germany)] and at the day of the experiment 5.5 m M glucose, 0.1% BSA, 0.05 mgÆmL )1 bacitracin was added. The total v olume (0.25 mL) contained 100 p M final concentration of 125 I-labelled Tyr-human (h) orexin B or 125 I-labelled Tyr-human orexin or 125 I-labelled h nueropep- tide Y (NPY)-Tyr36 (specific activity: 2000 CiÆmmol )1 ; Amersham) and increasing concentrations of the cold ligand orexin B or increasing concentrations of test compounds. After 120 min of gentle shaking a t room temperature the supernatant was removed followed by t wo washes with 0.25 mL NaCl/P i . Lysis buffer was a dded (NaCl/P i containing 2% Triton · 100) and after one wash with 0.5 mL radioactivity was counted. Membrane preparation and binding assay on SK-N-MC cells For m embrane p reparation of SK-N-MC ce lls, t he cells were grown in MEM (MEM with Earl’s salt, 10% fetal bovine serum, 1 m M sodium pyruvate, 1 % nonessential amino acids, 4 m M glutamine). Confluent cells were removed with 0.02% EDTA/ NaCl/P i and resu spended in 10 mL incubation buffer (MEM/25 m M Hepes containing 0.5% BSA, 50 l M phenylmethanesulfonyl fluoride, 0.1% bacitracin, 3 .75 m M CaCl 2 ), then washed twice with 10 mL NaCl/P i . After addition of 5 mL preparation buffer ( 5 m M Hepes, 0.32 M sucrose, 50 l M pefabloc, pH 7 .0) the cells were removed with a rubber policeman. After centrifugation at 4 °C, 10 min, 48 200 g the supernatant was decanted and centrifuged at 4 °C, 30 min, 48 200 g. The pellet was resuspended in 15 mL NaCl/P i . The sample was recentri- fuged at 4 °C, 50 min, 150 g and the pellet was resuspended in incubation buffer (84.7 m M NaCl, 3 0 m M KCl, 1.2 m M MgSO 4 Æ7H 2 O, 11.2 m M NaH 2 PO 4 bufferedwithHepes). After counting, the cells were diluted to a final concentration of 1.0 · 10 6 cellsÆmL )1 and homogenized using an Ultra- Thurrax. After the addition of 5.5 m M glucose, 0.1% BSA and 50 lg bacitracin, 200 ll of t his cell suspension was incubated for 2 h at room temperature with 100 p M 125 I-labelled orexin B and increasing concentrations of orexin, orexin a nalogues or NPY (Neosyste ` me, Strasbourg, France) in a total volume of 0.25 mL. Unbound radio- activity was separated by filtration through Whatman GF/ C filters presoaked in 0.5% po lyethylenimine. The filters were washed three times with ice-cold 0.9% NaCl. A ll tips and vials were siliconized. Competition binding experiments were analysed by a nonlinear least-squares fitting method with a one- or two- binding site model, respectively ( RS/1 software package, BBN Research Systems, Cambridge, MA, USA). The maximum specific r adioligand binding was s et to 100%. All data (n ¼ 3) are e xpressed a s mean ± SEM. Circular dichromism Conformational properties of the peptides were investigated by CD spectroscopy using a JASCO model J720 spectro- polarimeter ov er 190 –250 nm at 20 °CinaN 2 atmosphere. The peptides were dissolved in 20 m M NaCl/P i at neutral pH containing 0%, 30%, 50% or 70% trifluoroethanol and in pure trifluoroethan ol, in a concentration range of 200–300 l M . Each measurement was repeated three times using a thermostatable sample cell with a path of 0.02 cm and the following parameters: r esponse time, 2 s; scan speed, 2 0 nmÆmin )1 ; s ensitivity of 10 mdeg; s tep resolution, 0.2 nm; band width, 2 nm. The CD spectrum of the solvent was subtracted from the CD spectra of the peptide solutions to eliminate the interference from c ell, solvent and optical equipment. High f requency noise was reduced by means of a low-path Fourier-transform filter. The ellipticity was expressed as the mean-residue molar ellipticity [Q] R in degÆcm 2 Ædmol )1 . Rodent model of food intake Adult male Chbb:Thom rats weighing between 300 a nd 340 g were individually housed and maintained on a 12 light: 12 h dark cycle beginning at 06.00 hours. Tap water and standard laboratory chow were available throughout. After 1 week of habituation to their new housing conditions, t he animals w ere a naesthetized with sodium pentobarbital (60 mgÆkg )1 , intraperitoneally) for the placement of stainless steel guide cannulae. Cannulae (26 gauge) were placed 1 mm above the lateral hypothal- amus according to the stereotaxic coordinates: AP : 2.1, L : 2.0, V : 7.2 (+1 mm injection tip: 8.2). Guide cannulae were maintained in place on the skull w ith s mall metal screws and d ental a crylic cement. Cannulae were closed with a stainless steel stylet when not in use. Rats were allowed to recover for a t l east 1 week and were adapted t o the injection procedure. On the day of the experiments drugs 1130 H. A. Wieland et al. (Eur. J. Biochem. 269) Ó FEBS 2002 were injected between 01.00 and 02.00 p.m. Injection cannulae (33 gauge) were inserted 1 mm beyond the tips of the guide cannulae. The injection c annulae were attached by polyethylene t ubing to a Hamilton microsyringe mount- ed in an infusion pump. Injection volume was 0.4 lLgiven slowly over 40 s. Groups of six to eight rats received either saline ( c ontrol), 1.0 nmol Ærat )1 orexin A unilaterally, o r 1 nmolÆrat )1 orexin A and 3 nmol rat )1 orexin A 23–33 into the lateral hypothalamus and food intake was monitored for 4 h. In the s econd set o f experiments 1.0 nmol orexin A 23–33 was given with the injection of 1.0 nmol orexin A in order to antagonize the effects of o rexin A. RESULTS The peptides were synthesized by automated multiple peptide synthesis on a R ink Amide resin to directly obtain the peptide amides after cleavage of the peptides from the resin [25]. In addition to the native sequences h-orexin A and B, w e u sed t wo C-terminal segments h-orexin A 23–33 and h-orexin B 18–28. B ecause of the differ ent length o f the natural orexins, these two C-terminal segments are homo- logous and correspond to the C-terminal undecapeptide of orexin A and orexin B, respectively. Nine amino acids are identical, whereas orexin B contains a C-terminal methio- nine in contrast with leucine in orexin A (Fig. 1). This led to the orexin B analogue [L28] orexin B to make s ure that a ny identified differences are not owing to the different sequences. The second variable position is residue 23 (orexin A , alanine)/18 (orexin B, glycine). To investigate the role of this exchange we investigated [G23] orexin A 23–33 (Table 1) . The binding affinity of the peptides was tested on SK-N- MC cells. 125 I-labelled orexin B binding was inhibited in a dose-dependent fashion with a K i of 118 ± 57 n M (Table 1, Fig. 2) and to a similar order of magnitude on NT-2 cells, another human cell-line (data not shown). All curves displayed a monophasic shape with slopes close to unity. 125 I-labelled orexin B could be displaced by the orexin A and orexin B fragments in the range of human orexin B itself or with slightly improved affinity (Table 1). Sub stitution of orexin A at position 23 did not improve affinity significantly. Several atte mpts to detect specific 125 I-labelled orexin A binding was unsuccessful with SK-N-MC cells whereas recombinant CHO cells expressing the OX 1 receptor revealed a 5 0% inhibitory concentration (IC 50 )of 10 ± 6 n M for inhibition of 125 I-labelled orexin A binding by orexin A. The C-terminal fragments orexin A 23–33 and orexin B 18–28 do not displace 125 I-labelled orexin B from the recombinant receptor; neither does [G23] h-orexin A 23–33 displace 125 I-labelled orexin A. The first selective orexin 1 receptor antagonist (SB-334867-A) has been described recently, with a pK i value of 7.17 n M [23,26]. We tested a compound related to SB-334867, published earlier by G. Chan et al. [27], 1-(4- N,N-dimethylaminophenyl)-3-chinolin-4yl-urea), named EXBN8016BS. This compound displayed an IC 50 of 149 ± 3 n M for the inhibition of 125 I-labelled orexin A at the r ecombinant OX 1 receptor whereas it c annot inhibit the 125 I-labelled orexin B binding to both the recombinant OX 1 receptor or the orexin 1 receptor expressed on SK-N-MC cells. M odafinil was shown earlier to activate orexin-responsive n eurons. Therefore, w e e xamined whether Modafinil a cts indirectly via inhibitory orexin autoreceptors. Modafin il displayed no s ignificant a ffinity for the orexin B binding site of SK-N-MC cells or of recombinantly expressed OX 1 receptors. Sensitivity of orexin B binding to NPY has been observed (Table 1) with an IC 50 of % 450 n M . Table 1. Binding affinity of h-orexin A and B, C-terminal orexin A and B fragments and reported antagonists on SK-N-MC cells and CHO cells stably transfected w ith the huma n OX 1 receptor (1 00 p M radioligand). Ox1 receptor 125 I-labelled orexin B IC 50 [n M ] SK-N-MC cells 125 I-labelled orexin B K i [n M ] Ox1 receptor 125 I-labelled orexin A IC 50 [n M ] h-orexin A > 1000 (n ¼ 3) 882 ± 286 10 ± 6 h-orexin B 138 ± 28 118 ± 57 16 ± 15 [L28]h-orexin B 370 ± 200 – 180 ± 126 h-orexin A 23–33 > 10 000 (n ¼ 3) 119 ± 46 – [G23]h-orexin A 23–33 9400 ± 450 93 ± 80 > 10 000 (n ¼ 2) h-orexin B 18–28 > 10 000 (n ¼ 3) 49 ± 23 – Modafinil > 10 000 (n ¼ 3) > 10 000 (n ¼ 3) – EXBN8016BS a > 10 000 (n ¼ 3) > 10 000 (n ¼ 3) 149 ± 3 NPY 454 ± 241 (n ¼ 2) 450 ± 49 (n ¼ 2) – a 1-(4-N,N-dimethyl-aminophenyl)-3-chinolin-4yl-urea) [27]. Fig. 2. Receptor binding studies with 125 I-labelled orexin B and orexin B (m) using SK-N-MC cells. Ó FEBS 2002 SK-N-MC cell line expresses different orexin binding sites (Eur. J. Biochem. 269) 1131 Structure The structure of the peptides w as investigated by CD spectroscopy in aqueous solutions at neutral pH, containing increasing amounts of trifluoroethanol. All peptides adopted mainly random structure. Fi g. 3 s hows the CD spectra of orexin A 2 3–33 dissolved in water (A), 50% t rifluoroethanol in water ( B), 70% trifluoroethanol in water (C) and pure trifluoroethanol (D) in order to see any stabilizing effects of the solvent. All other peptides showed comparable CD spectra (data not shown). The negative band at 198 nm in aqueous solution, an indication of randomly structured pep tides, was shifted to 202 nm in all trifluoroethanol-containing samples. The negative CD value of the water solution at 190 n m was raised to positive values in trifluoroethanol- containing solutions. These shifts indicate partial forma- tion o f an a helix in trifluoroethanol-contain ing samples. Analysis of the spectra by a secondary structure estimation program ( JASCO , J-700 for Windows) based on the method of Yang et al. [ 28] revealed a slightly increasing amount of a helix with increasing amount of trifluoroethanol, although the maximum amount of helix was only % 11% (Fig. 3). Food intake Administration of 1 nmol orexin A into the third ventricle of rats significantly increased food intake after 2 and 4 h whereas a trend was seen after 6 and 8 h and no effec t was seen after 24 h. One nanomole of orexin A also significantly increased food intake after administration into the lateral hypothalamus (Fig. 4A). Administration of orexin A 23–33 together with orexin A in order to evaluate a potential antagonistic property of orexin A 23– 33 did not reveal any effect (Fig. 4A). Orexin A 23–33 in a dose range of 1 and 3 nmol per rat did not induce feeding (Fig. 4B). Sequencing To study orexin binding we searched for a suitable cellu lar system and screened neuronal cell lines of human origin (e.g. SK-N-MC and NT-2) for orexin receptor binding sites. Analysis of the cDNA derived from t otal RNA r evealed that these cell lines co ntain intronless o rexin 1 receptor transcripts that s eem t o b e partially edited in the c odon for the isoleucine/valine site at position 1375 (amino acid 408) beyond transmembrane region 7 (Fig. 5, Table 2). Fig. 5 shows that all other amino acids are 100% identical to the published sequence [1]. Analysis of the human genomic DNA revealed that adenosine is present at this position, which leads to an isoleucine at this position in the protein (personal commu- nication, Receptor Biology Inc). DISCUSSION The sequence o f the C-terminal decapeptide of o rexin A and B is conserved throughout all s pecies examined. Here we show that the C -terminal orexin fragments, orexin A 23–33 and orexin B 18–28, bind to the orexin receptor expressed on SK-N-MC cells with an affinity in the same range or with four- t o eightfold improved affinity c ompared Fig. 3. CD spectra and secondary structure according to the calculation of Yang et al. [28] of orexin A 23–33. Solvent: (A) water, pH 7.0; (B) 50% trifluoroethanol in water; (C) 70% trifluoroethanol in water; (D) pure t rifluoroethano l. Fig. 4. Food intake studies. Food intake after administration of orexin A via the lateral hypothalamus and after administration of orexin A 23–33 and orexin A (A). Food intake after administration of orexin A 23–33 (B). 1132 H. A. Wieland et al. (Eur. J. Biochem. 269) Ó FEBS 2002 with orexin B and orexin A, respective ly. On the contrary, initial data with N-terminal fragments, e.g. orexin B 1–10 or orexin B 1 –13, showed no significant affinity to th is orexin receptor (data not shown). This indicates an important sequence motif in the C-terminus of the peptides, that might play an essential role in binding of the peptides to t his receptor. This is in accordance with the recently published three-dimensional structure of orexin B, solved by two- dimentional NMR. I t s howed the peptide to consist of two a-helices, connec ted by a short linker [19]. The C-terminal helix of the mature human orexin B extended from r esidue 22 to residue 28. Although these seven residues constitute the major part of the ore xin A and B fragments, a helical structure, as postulated f or the matu re orexin B, could be found neither in orexin A 23–33 nor in orexin B 18–28. Even dissolving the peptides in solutions containing high amounts of trifluoroethanol, an a helix-inducing s olvent, resulted in peptides with maximal 11% a helix and still 89% random structure. This indicates that besides the confor- mational properties, t he amino acid side chains might also play an important role in binding of the orexin fragments to the receptors, in particular the trifunctional residues aspar- agine, histidine and threonine and additionally for orexin B 18–28 serine and the C-terminal methionine. The confor- mational influence of the C-terminal part of the o rexins remains unclear. The introduction of a helix inducers into C-terminal peptide fragments might increase binding affin- ity, because the native peptide contains a stable a helix at the C-terminus. Binding studies on SK-N-MC cells, a human n euroblas- toma cell line that is known to express NPY Y 1 receptors [29], revealed that this cell line also expresses receptors of the orexin family, which, after sequencing, turned out to correlate with the cloned OX 1 receptor. Both neuropeptides, NPY and orexins, are involved in the regulation of food intake. Sensitivity of orexin A binding to NPY has been described earlier by studying 125 I-labelled orexin A binding [30]. Interestingly some affinity of NPY could be identified fororexinreceptorsaswell.Assomecross-reactivityhas been reported [31], and orexin-induced food intake seems to involve the NPY pathways [22], this m ight b e one mode of regulation. Our results indicate the importance of the C-terminal part of th e orexins for binding interaction o f the ligand a nd its receptor expressed on SK-N-MC cells. The lack of affinity of the fragments for the recombinant OX 1 receptor suggests a different pharmacological profile a lthough a sequence almost i dentical t o the rec ombinant receptor is expressed in the neuroblastoma cell line. This discrepancy might be explained b y p ost-translational modifications, heterodimer- ization (re viewed in [33]) or different ac cessory proteins of the orexin 1 receptor e xpressed in SK-N-MC cells, similar to those found for the CGRP receptor [34] possibly resulting in different binding profiles. The binding to an orexin 2 receptor is possib le, but unlikely, because the dual e xpres- sion of both receptors with a different pharmacological profile should r esult in a biphasic inhibition curve. This w as not observed. The different IC 50 values of ligands, e.g. EXBN 8016BS or orexin A, to inhibit either 125 I-labelled ore xin A or Fig. 5. Sequence similarity of the o rexin 1 receptor cloned both from SK-N-MC and NT-2 cells. Posi tion X means G o r C at nucleotide 1375 coding for amino acid 408 which is either translated into isoleucine [1] or valine (Table 2). Table 2. Position 137 5 is nuc leotide 1375 within the O X 1 receptor cDNA which i s either t ranslated to isoleucine or valine. Cells/clone Pos 1375 Nucleotide at NT-2-cells 1 G Val 2G 3 A Ile 4A 5A 6A 7A 8A SK-N-MC 2G 3G Database (Acc.no AF 041243) [1] A Ile Ó FEBS 2002 SK-N-MC cell line expresses different orexin binding sites (Eur. J. Biochem. 269) 1133 125 I-labelled orexin B binding at the recombinan t OX 1 receptor might be due to different binding epitopes of the receptors as found for other agonist/antagonist systems of neuropeptides, such as NPY o r substance P [35,36]. It was hypothesized earlier [10] that modafinil may promote wakefulness t hrough orexin neurons. Sin ce, however, mod- afinil has low a ffinity fo r orexin 1 receptors, this activation might i nvolve other orexin receptors or could have another mechanism. The feeding effects with orexin A are a t variance w ith the strong and persistent feeding response observed by S akurai et al . [1], but similar to those d escribed by others in rats [3] and mice [37]. O rexin B showed no effects (data not shown) which is in agreement with earlier reports [3,4,37]. This is why we chose the shortened o rexin A fragment and not orexin B 18–28 for f eeding studies. The lack of feeding effects of orexin A 23–33 indicates that the orexin B binding site expressed on SK-N-MC cells is not represented in the lateral hypothalamus and therefore might not be involved in the regulation of food intake. However, this study has revealed a single nucleotide mismatch between corresponding cDNAs encoding orexin 1 receptors. Human genomic DNA analysis (personal com- munication, Receptor Biology) indic ated that alternative exons c ould be excluded as a potential source for this nucleotide exchange. Hence, editing of the RNA transcribed from these genes best explains our observation w hich is similartotheAfiG editing described in glutamate-gated channels [38,39] and with the G p rotein-c oupled seroton in- 2C receptor [40]. Single nucleotide polymorphism cannot be excluded at this point but the G was not found on a genomic level. Subtle kinds of regulation of G p rotein-coupled receptors coupling to G proteins have been described earlier [33]. For example, transcripts encoding the 5-HT 2C recep- tor, a phospholipase C-coupled receptor, un dergo R NA editing events in which the g enomically encoded adenosine residues are converted to i nosines by a double-stranded RNA adenosine deaminase(s). Seven major 5-HT 2C recep- tor isoforms are predicted, e ncoded by 11 distinct R NA species and differing in their second intracellular loops [40]. This post-transcriptional modification leads to a 10- to 15-fold reduction in efficacy of the coupling of 5-HT 2C to the G protein. We conclude that the s tructural s tudies o f C-terminal orexin A and B fragments and their binding affinity to the orexin receptors as presented in this w ork, provide a nother step in characterization o f the r eceptor binding mode of orexin A and B and provide important information for pharmacological a nd biochemical investigations. The observed A/G exchange in SK-N-MC and NT-2, two human cell lines, might indicate an editing process which will be further investigated. ACKNOWLEDGEMENTS The financial support of the Deutsche Forschungsgemeinschaft (Be 1264-3/1) is kindly acknowledged. We further acknowledge the skilled technical e xpertise of E. Liebhardt a nd S. Schacherl-Schmid. REFERENCES 1. Sakurai, T., Amemiya, A., Ishii, M., Matsuzaki, I., Chemelli, R.M., T anaka, H., Williams, S.C., Richardson, J.A., Kozlowski, G.P., Wilson, S., Arch., J.R.S., Buckingham, R.E., Haynes, A.C., Carr, S.A., Annan, R.S., McNulty, D.E., Liu, W S., Terrett, J.A., Elshourbagy, N.A., Bergsma, D.J. & Yanagisawa, M. ( 1998) Orexins and orexin receptors: a family of hypothalamic neuro- peptides and G protein-coupled receptors that regulate feeding behavior. Cel l 92, 573 –585. 2. De Lecea, L., Kilduff, T.S., Peyron, C., Gao, X B., Foye, P.E., Danielson, P.E., Fukuhara, C., Battenberg, E.L.F., Gautvik, V.T., Bartlett,F.S., I.I.Frankel, W.N., Van den Pol, A.N., Bloom, F.E., Gautvik, K.M. & Sut cliffe, J.G. (1998) The hypocretins: hypo- thalamus-specific peptides with neuro excitatory activity. Proc. Natl Acad. S ci. USA 95 , 322–327. 3. Haynes, A.C., Jackson, B., Overend, P., Buckingham, R.E., Wilson, S., T adayyon, M. & Arch., J.R.S. ( 1999) Effects of single and chronic intracerebroventricular administration o f the or exins on feeding in the rat. Peptides 20 , 1099–1105. 4. Ida, T., Nakahara, K., Katayama, T., Murakami, N. & Nakazato, M . (1999) Effect of lateral cerebroventricular injection of the appetite-stimulating neuropeptide, orexin and neuropeptide Y, on the various behavioral activities of rats. Brain Res. 821, 526–529. 5. Takahashi, N., Okumura, T., Yamad a, H. & Kohgo, Y. (1999) Stimulation of gastric acid secretion by centrally administered orexin-A in conscious r ats. Biochem. Biophys. Res. Commun. 254, 623–627. 6. Griffond, B., Risold, P.Y., Jacquemard, C ., Colard, C. & Fellmann, D. (1999) Insulin-induced hypoglycemia increases preprohypocretin (orexin) mRNA in the r at lateral hypothalamic area. Neurosci. L ett. 262, 77–80. 7. Dyer, C.J., Touchette, K.J., Carroll, J.A., Allee, G.L. & Matteri, R.L. (1999) Cloning of porcine prepro-orexin cDNA and effects of a n intramuscular injection of synthetic porcine orexin-B on food intake in young pigs. Domestic Anim Endocrinol. 16, 145–148. 8. Van den Pol, A.N., Gao, X.B., Obrietan, K., Kilduff, T.S. & Belousov, A.B. (1998) Presynaptic and postsynaptic actions and modulation of neuroendocrine neurons by a new hypothalamic peptide, hypocretin/orexin. J. Neurosci. 18, 7 962–7971. 9. Date, Y., Mondal, M.S., Matsukura, S., Ueta, Y., Yamashita, H., Kaiya, H., Kangawa, K . & Nakazato, M. ( 2000) Distribution of orexin/hypocretin in the rat median eminence and pituitary. Brain Res. Mol. Brain Res. 76, 1–6. 10. Chemelli, R.M., Willie, J.T., Sinton, C.M., Elmquist, J.K., Scammell, T., Lee, C., Richardson, J.A., Williams, S.C., Xiong, Y., Kisanuki, Y., Fitch, T.E., Nakazato, M., Hammer, R.E., Saper, C.B. & Yanagisawa, M . ( 1999) Narcolepsy in or exin knockout mice: molecular genetics of sleep regulation. Cell 98, 437–451. 11. Singer, C.M. & Lewy, A.J. (1999) Does our DNA determine when we sleep? Nat. Med. 5 , 983. 12. Hagan, J.J., Leslie, R.A., Patel, S., E vans, M.L., W attam, T.A., Holmes, S., Benham, C.D., Taylor, S.G., Routledge, C., Hemmati, P., Munton, R.P., Ashmeade, T.E., Shah, A.S., Hatcher, J.P., Hatcher, P.D., Jones, D.N.C., Smith, M.I., Piper, D.C., Hunter, A .J., Porter, R.A. & Upton, N. (1999) Orexin A activates locus coeruleus cell firing and increases arousal in the rat. Proc. N atl Acad. Sci. USA 96 , 10911–10916. 13. Nambu, T., Sakurai, T., Mizukami, K., Hosoya, Y., Yanagisawa, M. & Goto, K. (1999) Distribution of orexin neurons in the adult rat brain. Brain R es. 827, 2 43–260. 14. Dube, M.G., Kalra, S.P. & Kalra, P.S. (1999) Food intake elicited by central administration of orexins/hypocretins: identification of hypothalamic sites o f action. Brain Res. 842, 4 73–477. 15. Peyron, C., Tighe, D.K., van den Pol, A.N., de Lecea, L., H eller, H.C., Sutcliffe, J.G. & Kilduff, T.S. (1998) Neurons containing hypocretin (orexin) project to multiple neuronal systems. J. Ne urosci. 18, 9996–10015. 1134 H. A. Wieland et al. (Eur. J. Biochem. 269) Ó FEBS 2002 16. N ishino, S., Ripley, B., Overeem, S., Lammers, G.J. & Mignot, E. (2000) Hypocretin (orexin) deficiency in human narcolepsy. Lancet 35 5, 39–40. 17. Lin,L.,Faraco,J.,Li,R.,Kadotani,H.,Rogers,W.,Lin,X.,Qiu, X., de J on g, P.J., Nishino, S . & Mignot, E. ( 1999) the sleep d is- order canine narcolepsy is caused by a mutation in the hypocretin (orexin) receptor 2 gene. Cell 98, 365–376. 18. Siegel, J.M. (1999) Narcolepsy: a key role for hypocretins (orexins). Cell 98, 4 09–412. 19. Lee,J.H.,Bang,E.,Chae,K.J.,Kim,J.Y.,Lee,D.W.&Lee,W. (1999) Solution structure o f a n ew hypothalamic neuropeptide, human h ypocretin-2/orexin-B. Eu r. J. Bioc hem. 266 , 831–839. 20. S hibahara, M., Sakurai, T., N amb u, T., Takeno uchi, T., Iwaasa, H., Egashira, S.I., Ihara, M. & Goto, K. (1999) Structure, tissue distribution, a nd pharmacological characterization of Xenopus orexins. Peptides 20, 1 169–1176. 21. Smart, D ., Jerman, J.C., Brough, S.J., Rus hton, S.L., Murdoch, P.R., Jewitt, F., Elshourbagy, N.A., Ellis, C.E., Middlemiss, D.N. & B rown, F. (1999) Characterization of recombinant h uman orexin receptor pharmacology in a Chinese hamster ovary cell-line using F LIPR. Br. J. Pharmaco l. 128,1–3. 22. Darker, J.G., Porter, R.A., Eggleston, D.S., Smart, D., Brough, S.J., Sabido-David, C. & Jerman, J.C. (2001) Structure–activity analysis of truncated orexin-A analogues at the orexin-1 receptor. Bioorg. Med. Chem. L et t. 11, 737– 740. 23. S mart, D ., Sabido-David, C., Br ough, S.J., Jewitt, F., Johns, A., Porter, R.A. & Jerman, J.C. (2001) SB-334867-A: the first selective o rex in-1 receptor anta gonist. Br.J.Pharmacol.132, 1179–1182. 24. Beck,W.&Jung,G.(1994)ConvenientreductionofS-oxidesin synthetic peptides, lipopeptides and peptide libraries. Lett. Pept. Sci. 1, 3 1–37. 25. So ¨ ll, R. & Beck-Sickinger, A.G. (2000) On the synthesis of orexin A: a novel protocol to obtain peptides with two intramolecular disulfide bonds. J. Pept. Sci. 6,387. 26. Rodgers, R.J., Halford, J.C., Nunes de Souza, R.L., Canto de Souza, A.L., Piper, D.C., Arch., J.R., Upton, N., Porter, R.A., Johns, A. & B lundell, J .E. (2001) SB-334867, a selective orexin-1 receptor an tago nist, e nh ances behavioural satiety a n d b locks the hyperphagic e ffect of orexin-A in r ats. Eur. J. Neurosci. 13, 1444– 1452. 27. Chan,G.,Johns,A.,Jurewicz,A.,Porter,R.&Widdowson,K. Phenyl urea and phenyl thiourea derivatives as HFGAN72 antagonists. WO 99/09024. 28. Yang, J.T., Wu, C.S. & Martinez, H.M., (1986) Calculation of protein conformation from circular dichromism. Methods Enzy- mol. 130 ,208. 29. B eck-Sickinger, A.G., Wieland, H.A., Wittneben, H., Willim, K.D.,Rudolf,K.&Jung,G.(1994)Complete L -alanine scan of neuropeptide Y reveals ligands binding to Y1 and Y2 recep- tors with distinguished conformations. Eur. J. Biochem. 225 , 947–958. 30. Kane, J.K., Tanaka, H., Parker, S.L., Yanagisawa, M. & Li, M.D. (2000) Sensitivity of orexin-A binding to phospholipase C in hibi- tors, neuropeptide Y, and secretin. Biochem. Biophys. Res. Commun. 271, 959– 965. 31. Jaszberenyi, M ., Bujdoso, E. & T elegdy, G. (2001) The role of neuropeptide Y in orexin-induced hypothalamic-pituitary-adrenal activation. J. Neuroendocrinol. 13, 438–441. 32. Yamanaka, A., Kunii, K., Nambu, T., Tsujino, N., Sakai, A., Matsuzaki, I., Miwa, Y., Goto, K. & Sakurai, T. (2000) Orexin- induced food intake involves neuropeptide Y pathway. Brain Res. 859, 404–409. 33. Bockaert, J. & Pin, J .P. (1999) Molec ular tinkering of G protein- coupled receptors: an evolu tionary process. EMBO J. 18, 1723– 1729. 34. McLatchie, L., Fraser, N., Main, M., Wise, A., Brown, J., Thompson, N., S olari, R., Lee, M. & F o ord, S. ( 1998) RAMPs regulate the transport and ligand specificity of the calcitonin- receptor-like r eceptor. Nature 393, 333– 339. 35. S autel, M., Rudolf, K., Wittneben, H., Herzog, H., Martinez, R., Munoz, M., Eberlein, W., Engel, W., Walker, Ph & Beck-Sick- inger, A.G. (1996) Neuropeptide Y and the non peptide antagonist BIBP 3226 share a large overlapping binding site at the human Y 1 receptor. Mol . Pharmacol. 50 , 285–292. 36.Rosenkilde,M.M.,Cahir,M.,Gether,U.,Hjorth,S.A.& Schwartz, T .W. (1994) Mutations along transmembrane se gment II of the NK-1 r eceptor affect substance P competition with non-peptide antagonists but not substance P binding. J. Biol. Chem. 269, 2 8160–28164. 37. Lubkin, M. & Stricker-Krongrad, A. (1998) I ndependent feeding and metabolic a ctions of Orexins i n mice. Biochem. Biophys. Res. Commun. 253, 241– 245. 38. Sommer, B., Ko ¨ hler, M., Sprengel, R. & Seeburg, P.H. (1991) RNA editing in brain controls a d eterminant of io n flow in glutamate gated channels. Cell 67, 11– 19. 39. Higushi, M ., Single, F., Ko ¨ hler, M., Som mer, B., Sprengel, R. & Seeburg, P.H. (1993) RNA editing of AMPA receptor subunit GluR-B: a base-paired intron-exon structure determines position and efficiency. Cell 75, 1 361–1370. 40. B urns, C.M., Chu, H., R ueter, S.M., H utchinson, L.K., C an ton, H., Sanders- Bush, E . & Emeson, R.B. (1997) Regu lation of serotonin-2C receptor G-protein coupling by RNA editing. Nature 387, 303–308. Ó FEBS 2002 SK-N-MC cell line expresses different orexin binding sites (Eur. J. Biochem. 269) 1135 . The SK-N-MC cell line expresses an orexin binding site different from recombinant orexin 1-type receptor Heike A. Wieland 1, *, Richard. inhibitory orexin autoreceptors. Modafin il displayed no s ignificant a ffinity for the orexin B binding site of SK-N-MC cells or of recombinantly expressed OX 1 receptors.