View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Volume 296, number 3, 241-244 FEBS lOGO January 1992 0 1992 Federation of Europnn Biochemical Societies 00l45793/92&00

Stable expression of high affinity NIL1 () and NK, () receptors but low affinity NK3 () receptors in transfected CHO cells

Ulrik Gether, Taisser Marray, Thue W. Schwartz and Teit E. Johansen

Lnborarory ofMofectllar Ettdocrittology, Universiry Depurtmnnr of Ciinicuf Cftentisiry, Rfgsltospitafe[ 6Mi. Copenftagm, Dermrrrk

Received 21 November 1991

Stable CHO cell clones which selectively express all three rat tachykinin receptors were established by transfection. The binding of radiolabeled substance P and ncurokinin A (substance K) to CHO clones expressing the NK, and NK2 receptors, respectively. were snturatable and of high affinity (&=0.17 nM (NK,); 3.4 nM (NKJ). Scatchard analysis of the binding dam indicated for both receptors binding to a sin& population of binding sites, and competition binding studies showed that the binding specificities of the receptors corresponded to whoseof classical NK, and NKz receptors. In contrast, the binding of cledoisin to the NK, receptor expressed in the transfectcd CHO cells was of low affinity (ICI,=240 nM) compared to the high affinity of the receptor found when it was transiently expressed in COS-7 cells (I&=8 nM). Howcvcr, in both cases the receptor exhibited the specificity of a classical NK, receptor. The established cell clones may provide an important tool for further analysis ofthe molecular mechanisms involved in binding. activation, and coupling of receptors for tachykinin .

Tachykinin receptor; Transfection; CHO cell; COS-7 cell; G protein

1. INTRODUCTION ent receptor subtypes have bee2 cloned [12]. Hence, the tachykinin receptors constitute an optimal system for The mammalian tachykinins, substance P, neuroki- analyzing structure-function aspects of the binding and nin A (substance K) and ncurokinin B (neuromedin K) activation of receptors for . For example, are characterized by the common carboxy-terminal sc- it gives the opportunity to construct chimeric receptors quence Phe-X-Gly-Leu-Met-NH?. They are widely dis- between the receptor subtypes, an approach which has tributed neuropeptides in both the central and peri- proven very useful1 with respect to the adrenergic recep- pheral nervous system and involved in multiple biologi- tors [ 133. cal actions including modulation of sensory afferent In the present study we have established stable, trans- transmission, intestinal motility, stimulation of - fected clones of CHO cells which selectively express the tion, and activation of cells of the immune rat NK,, NK, and NK3 receptors. The binding proper- system (for reviews see refs. [l-4]). ties of the NKI and NK? receptors corresponded to The tachykinins act on at least three distinct receptors those of the physiological, classical receptors. In con- called NK,, NK? and NK3 which preferentially bind trast, the NKJ receptors expressed in CHO cells were of substance P, neurokinin A, and neurokinin J3, respec- low affmity as compared to NKJ receptors transiently tively [5-73. Recently Nakanishi and co-workers have expressed in COS-7 cells from the same plasmid. isolated all three cDNA’s encoding these receptors by use of a Xcnopus expression system [8-lo]. The cloning 2. EXPERlMENTAL of the NK, receptor was also recently reported by Her- shey and Krause 1990 [ll]. The predicted 2, I , Pepr ides Substance P. neurokinin A. ncurokinin B and elcdoisin were pur- sequences of the receptors revealed homologous struc- chased from Peninsula Laboratories. tures typical for G protein coupled receptors character- ized by the presence of seven putative transmembrane 2.2. Esprcssio~t wlors, rrunsfecrion artd cell cuhrrc domains [12]. Only a few neuropreptide receptors have The cDNA encoding the NK,, NK: and NK, receptors were gcner- ously provided by Dr. S. Nakanishi, Kyoto University, Dcpl. of Me- been cloned until now and ihe tachykinin family was the dicine, Kyoto, Japan. The NK, receptor cDNA was excised from the first family of neuropeptides for which a series of differ- original plasmid prTKR2 [9] as ti 2049 bp HindIII-EroRI fragment. the NKI receptor cDNA was excised from prTKRI-I [Id] as a 1,816 bp EcoRI-BmtWl fragment and the NK, receptor cDNA was excised Corresponrimce address: U. Gether, Laboratory of Moieculnr Endo- from prTKR3 [IO] as a 2.159 bp EcoRl-EcoKl fragment. These frag- crinology, University Dcpl. oT Clinical Chemistry, Rigshospitalct mcnts were cloned into the pTEJ-8 exprczion vector under control of 6321. Blegdamsvej 9, DK-2100 Copenhagen, Denmark. Fax: (45) the ubiquitin promotor and the SW0 polyadenylation signal. The (3135) 2995.

Published by ELwier Sciwtce Publishers B. V. 241 Volume 296, number 3 FEBS LETTERS January 1992 expression vector also comains the nromycin phosphotranskrase [15]. For stable expression fhe icsulting plasmids, pTE1&NKI, pTEJi3-NK2, and pTEJSmNK3 were iransfcclcd into CHOcclls hy the calcium phosphale precipilation method as described [l5,16]. For transient expression in COS-7 cells, IOhcells were seeded in a 90 mm culture plate and grown for 24 b before transfection by the calcium pbosphatc precipitalion method using 20 ,ug of DNA, After 24 h the cells were harvested, seeded in a 12~~ll culture plate and grown for another 24 h before analyzing. Cl-IO cells and COS-7 cells were grown as described [ 151.

2.3. Prepararion atuf purificarion oJrudioligatxis Monoiodinated ‘251.Bolion Hunter subsi?ncc P (‘““I-BH-SP), ‘2sI- Bohon Hunter neurokinin A (‘zJI-BH-NKA) and 1~51-Bol~onHunter (lzSI-BH-ELE) were prepared according to earlier described mclbods [6]. The reaction mixlure was subjected to revcrscephasc HPLC as described [17]. The fractions containing the labeled pcptides were pooled and stored at -2OOC in acid ethanol containing 0.2% mercaptoclbanol. Under thcsc condhions the radioligands were stable for 2-3 month:,. As the separation of labeled and unlnbcled was complete. :.he specific aclivity of ihe radioligands were assumed 10 be equal to the specific activily of the Bollon Hunier rcagcnt (4.400 cpm/fmol) after correctiorl for decay and counter eff&xcy. b -11 -10 -0 -u -7 -e 0 LO co ED no lo& ccnc.. ncutoklnln A Spcc~l~c bound 2.4. Bihcfing usqy (MI (fno1/1.Z6 I loo calls) Cells. Transfccted CHO-cells or CO!+7 cells were transferred 10 either IZ-well culture plales (0.3-I ,0x IO’ccells) or 6-well culrurc plates Fig. I. Upper left pcltrel: specific binding of “‘I-BH-SP to CHO-NK I- (l&l .5 x IO0cells) (Costar) 24 b before performing the binding assay. 5. The cells (CHO-NKI-5) were incubaled with 12’I-BH-SP (25 pM) Culls were washed twice in phosphate-buffered saline, pH 7.4, (PBS) and increasing concentralions of substance P as indicatr:d. Data arc followed by incubation for 3 h at 4’C with 25-50 pM radioligand and i:xpressed as the specific binding in cpm (0) and are from a rcpresenta- varying concentrations of cold peplides in 0.5 ml or I .O ml of 150 mM tive experiment of three. The binding of ‘s51-BH.SP to CHO cells N&I. 5 mM MnCI, in 50 mM HEPES, pH 7.4, supplemented with transrecied with the pTEZJ-8 expression vector alone is indicated (0). 0.1% bovine serum albumin (Sigma), 100 pgpl bacitracin (Sigma), 5 Scatchard analysis of the binding data are illuslrated in upper rig/u &+I leupeptin (Sims) and 10 p@l chymostatin (Sigma). 1.ne in- pmaf. toirrer fe_ffpurzef: the spcciflc binding of ‘2’l-BH-NKA lo CHO. cubation media were rcmovcd and the cells were washed twice in NK2-I. The cells (CHO-NK2-I), were incubated with “sl-BH-NKA ice-cold binding buffer. The cells were lysed in I ml lysis buffer (3 M (SO pM) and increasingconcrntrations or neurokininA as indicolcd. acetic acid, 8 M urea and 24 NonidetP-40) before counting. All Data arc expressed as specific binding in cpm (0) and are from a daermina.tions were performed in rriplicalc. The unspecific binding representative expcrimem of lhrec. The binding of lz51-BH-NKA to was determined as the binding of radioligand in the presence of I PM Cl-10 cells transfected with the pTEJ-8 expression vector alone is cold peptide. The spcciflc binding constituted more than 90% of toial indicated (0). In lower rig/u pnnel Scatchard analysis of the binding bound 011 the NK, receptor, more than 75% of total bound on the NK, data is shown. receptor and more ihan 50% of total bound on the NK, receptor (CHO cells). In selected experiments the stability of the WBolton Hunler labeled peptides during incubations was checked by analyzing radio&and (‘“‘I-BH-SP), the NE& receptor transfectant 100 ,ul aliquols of the incubation media on an HPLC system as de- bound 0.9&0.5% (mean t- S.E.M., n=6; range 0.05- scribed above. Aher 3 h of incubation more than 85% of the mdioac- 2.2%) (‘“‘I-BH-NKA) and the NK3 receptor trans- !ivity co-migraled with the radiolabeled peptides. fectants bound 1.31-0.18 (mean 2 S.E.M., ~25; range Mctttbrcrnes. The cells wcrc harvcstcd in 5 mM ice-cold EDTA in 0.3-2.8%)(~~SI-BH-ELE) when assaying 1.0 x IO” cells PBS, pelleled by centrifugntion at 200 x g and homogenized using an Ullralurrax homogenizer in 25 mM Tris-HCI, pH 7.4, containing 100 in the presence of 50 pM radioligand. No specific bind- &&I bacitracin. 4 ,u&l leupeptin, IO pg&l chymostatin and 1 mM ing could be determined in either untransfected CHO phenylmeihylsulfonyl fluorid (PMSFJ. The homogenate was cen- cells or in CHO cells transfected with the expression trifuged for I h at 25,000 x g and the pellct.ed membranes were re- vector alone. The NKI receptor clone with the highest suspended in the binding buffer. The membranes (2OOyg protein) were %), incubntcd for I.5 h ai 20°C with 50 pM ‘z51-BH-SP and various level of radioligand binding (8 1 CHO-NK 1- 11, and concentrations of cold pcptidc in 0.5 ml of a 25 mM HEPES buffer. a clone with an intermediate level of radioligand bind- pH 7.4, supplemented wilh 2 mM CaCI,, 1 mM M&la, 5 mM MnCI:, ing (24%), CHO-NKl-5, were selected for further char- I% BSA and rhe above menlioncd proiease inhibitors. The incubation acterization. To characterize the NI& receptor we was stopped by the addition of 0.5 ml ice-cold binding buffer before chose the clone, CHO-NK2-1, which bound 2.1% of the centrifugation nt 3,500 xg for IO min. The pellet was washed in buffer, repelleled and counted. All binding data were analyzed by compu- radioligand and to characterize the NI& receptor we terized non-linear regression analysis using GraphPAD, IS1 Software, chose CHO-NK3-7 which bound 2.8% of the radio- Philadelphia, PA. ligand. The specific binding of ‘zsI-BH-SP to the NK, re- ceptor reached equilibrium after 2.5 h (not shown). The 2. i?ESixTS displacement of ‘251-13H-SPwith increasing concentra- The stable NKl receptor transfectants bound 28-t-7% tions of ‘cold’ substance P showed that the specific bind- (mean f S.E.M., m=17; range 2.0-81%) of the added ing was of high affinity (Fig. 1, upper pacel). Scatchard

242 Volume 296, number 3 FEBS LETTERS January 1992

X of mar. bound teDI-Bli-8P' 100 - 'ehlH-ELE 100 CHO-?iK9-7

50 -

0 I -10 -e -0 -7 -6 0 -11 -1G -e -8 -7 -e Wb;\yx. log COAC., eledoisln I (Ml

Fig. 3. Specific binding of ‘?‘I-BH-ELE LOthe NK, receptor stably expressed in CHCJcells (O-O) and transiently cxpresscd in COS-7cells (c....Q The cells were incubated with SOpM ‘%DH-ELE and indi- cated concentrations of cledoisin. Data are cxprcsscd as percent of max. bound “‘I-BH-ELE (meant S.E.M., n=3).

determined both on intact cells and cell membranes, i CHO-NKP-1 Table I (not shown). 1 The binding of “SI-BH-NKA to the NK2 receptor 0 -11 -10 -8 -8 -7 -0 also reached equilibrium after 2.5 h (not shown). The log cont., peptide binding was saturatable, of high affinity and to a single 04 population of binding sites (approx. 20,000 per cell) Fig. 2. Binding specificity of the NK, and NK2 receptor cxprcssed in (Fig. 1, lower panel and Table I). CHO cells. Upper parwl: inhibition of ‘251-BH-SP binding to CHO- The specificity of the binding of ‘z51-13H-SPand tz51- NK l-5 by substance P (0) (n= 12) (K, = ! .8 x IO-” M). ncurokinin A BH-NKA to the NK, and NK, receptors expressed in (A) (/l=(i) (& = 9.5 x IO-* M) and neurokinin B (m) (n=6) (Ki = 4.0 x CHO-cells was analyzed by competition binding stud- IO-‘). Ltmw [~rrfrc(: inhibition of “‘1-BH-NKA binding to CHO-NK?- ies. As illustrated in Fig. 2, upper panel, substance P I by neurokinin A (A) (12~7)(& = 3.B x IO-yM), ncurokinin B (m) (rr=3) (Ki = I.3 x IO-’ M) and substance P (0) (11~3) (K; = 7.6 x IO-’ most potently displaced “51-BH-SP from the NK, re- M). All the data are expressed as percent of maximum bound radio- ceptor followed by neurokinin A and nerokinin B. On liyand (mm&D.). the NKz receptor, neurokinin A most potently displaced “‘I-BH-NKA followed by neurokinin B and substance analysis of the binding data showed that the CHO-NK I-5 P (Fig. 2, lower panel). clone expressed a single high-affinity binding site with In contrast to the results obtained on the NK, and a & of 0.19-tO.03 nM and a S,,, corresponding to NK2 receptors, we found that binding to the NK:, re- approx. 80,000 binding sites per cell (Fig. 1, upper panel ceptor transfectants was apparently of low affinity (Fig. and Table 1). The CHO-NKl-I 1 cell clone expressed, 3). Thus, using “‘I-BH-ELE as radioligand I&, for like CHO-NKl-5, a single high-affinity binding site as eledoisin was 240 nMkl20 nM (mean P S.E.M., n=3). However, when the NK3 receptor was transiently ex- pressed in COS-7 cells, we obtained high-affinity bind- Table I ing in the nanomolar range (ICso = 8-t-2 nM, mean 1 S.E.M., m=3) in accordance with previously published Binding properties of the iat NK, and NK, rccepiors stably expressed in transfected CHO cells results [IO]. Transient expression of the NK, and NKI receptor in COS-7 cells also resulted in high-affinity Cell clone &. -I,, n;e nm,l,brann. Binding sites binding (not shown). The specificity of the NKa re- (nM) (nM) (fr%lO5 per cell ceptor transiently expressed in COS-7 cells correspon- cells) ded likewise to previous results [IO] as the binding of CHO-NKI-5 0.19 t 0.03 ND. 13.7 t 0.2 80,000 ‘z51-BH-ELEwas inhibited most potently by neurokinin CHO-NKI-I 1 0.14 ?: 0.01 0.2 4 0,l 52 zk3 300,000 B (ICs, = 1.7 x lo-’ M) followed by eledoisin (I&, = 8 CHO-NK2-I 3.4 f 0.3 N.D. 3.6 f I.3 20,000 x lo-” M), neurokinin A (ICI0 = 3.2 x IO-’ M) and substance P (ICsO= 1.8 x lo-’ M) (data not shown). A The & and Em!,,, :‘nl~s arc cslcclated from Scatchard malysis of binding dota from ‘cold’ smration binding experiments as described similar rank order of potency was obtained for the NK; in Section 2. Data arc expressed as man I S.E.M. of three expcri- receptor stably expressed in the CHO cells although the merits. N.D. = not determined. I&-, values were approx. 30-SO-fold higher.

243 Volume 296, numbci 3 FEBS LETTERS January 1992

4. DISCUSSION An alternative explanation for the low affinity state of the NK3 receptor in CHO cells could possibly be The construction of ceil lines which selectively ex- found in the 71 amino acid-long, extracelhilar amino press a specific receptor subtype offers the advantage of terminus of the NK3 receptor which contains four gly- a controlled system for structural and functional char- cosylation sites compared to only two glycosylation acterization of the receptors. Of the tachykinin recep- sites in the 31-32 amino acid-long amino terminus of tors, the rat NK, and NK3 receptors have previously the NK, and NF_. receptors [9,10,14]. If CHO cells and been transiently expressed in COS-7 cells and the bovine COS-7 cells exhlblt different glycosylation patterns, the NK2 receptor has recently been stably expressed in glycosylated amino terminus of the NKJ receptor could mouse B82 fibroblasts [l&19]. In the present study we possibly interfere with ligand binding only in CHO cells. have established stable CHO cell clones which selec- This should also be elucidated in the future by analyzing tively express all three rat tachykinin receptors. The a NKJ receptor with a truncated amino terminus. NK, receptor and the NK, receptor demonstrated both A~~rto,vic~d~~,,ra,lrs:Nanni Din is thanked for helpful comments on the high affinity and the expected specificity [4-73. In con- manuscript and Tina Jakobsen and Margit Serenscn are thanked for trast, the NK3 receptor exhibited low-affinity binding superb technical assistance. The work was supported by grants from when expressed in CHO cells compared to the high the Danish Medical Research Council, the NOVO Foundation and affinity of the receptor when it was transiently expressed the Danish Biotechnology Center for Signal Peptides. in COS-7 cells. The high affinity of the NK3 receptor when transiently expressed in COS-7 cells has also been REFERENCES demonstrated in previous studies [lo]. It is a generally accepted hypothesis that high-affinity binding of PI Pernow, B. (1983) Pharmacol. Rev. 35, i35-141. Nakanishi, S. (1987) Physiol. Rev. 67, I 117-l 142. agonists to G protein-coupled receptors is dependent on ::; Maggie, J.E. (1988) Annu. Rev. Neurosci. I I, 13-28. the formation of a ternary complex between the agonist, PI Regoli, D., Drapcnu. G.. Dion, S. and Couture, R. (19SB)Trcnds the reqptor and at least the corresponding G protein Pharmucol. Sci. 9, 290-295. a-subunit [20]. Thus, an explanation for the expression 151Buck, S.H,, Burcher, E., Shults,C.W., Lovenbcrg, IV. and O’Do- nohuc. T.L. (19841 Science 226.987-989. of low affinity NKS receptors in CHO cells would be LB Cascicri. M.k.. Chicchini. G.d. iind Li:mg, T. (1985) J. Biol. that these cells do not express the correct G protein Chcm. 260, 1501-1507. a-subunit, or at least not in a functional form. If the 171 Lee. C.-M., Campbell. N.J., Williams, B.J. and Iverscn. L.L. optimal a-subunit is absent or non-functional, rhe re- (1986) Eur. 1. Pharmacol. 130,209-217. ceptor will interact with other a-subunits present, how- PI Masu, Y., Nakayama, K., Tanlaki, H., Harada, Y., Kuno, M. and Nakanishi. S. 11987) Nature 329. 836-838. ever, with less efficiency [211. This could explain why we 191Yokotu, Y., Sasai. ‘Y., Tinaka, K., Fihiwara, T., Tsuchida, K., at all observe binding to the NK3 receptor expressed in Shigemoto, R., Kakizyka, A., Ohkubo, H, and Nakanishi, S. the CM0 cells. In general, it is not expected to be pos- (1989) J. Biol. Chem. 30, 17649-17652. sible to detect binding of an agonist radioligand to a [lOI Shigcmoto, R.. Yokota, Y., Tsuchida, K. and Nakanishi, S. (1990) J, Biol. Chcm. 265, 623-628. receptor which has been uncoupled from its corre- [I I] Hcwhcy, A.D. and Krause. J.E. (1990) Science 247, 958-962. sponding G protein due to a too low agonist affinity 1121 Dohlman. H.G.. Thorner. J.. Caron. M.G. and Lefkowitz. R.J. . I POI. (1991) A&u. RI&. Biochcm.‘fiO. 653-688. It is known that the CHO cells lack at least the GzO [13] Kobilka, B.K.. Kobilka, T.S., Daniel, K.. Rcgan, J.W., Caron, a-subunit (P.H. Andenen, personal communication) M.G. and Lcfkowite, R.J. (198i3) Science 240, 1310-1316. [I41 Sasai, Y. and Nakanishi, S. (1989) Biochem. Biophya. Res. Com- but it is unclear to which G protein cz-subunits the ta- mun. 163, 695-702. chykinin receptors couple. The NK,, NK1, and NK, [I51 Johansen. T.E., Schollcr, MS., Tolstoy, S. and Schwartz, T.W. receptors are generally believed to couple to phospho- (1990) FEBS Lett. 267, 89-294, inosiride turnover via a pertussis toxin-insensitive mech- [IG] Grahnm. F.L. und Van der Eb, A.J. (1973) Virology 52,45&467. [I73 Gcthcr, U., Niclscn, H.V. and Schwartz, T.W. (1988) J. Chro- anism [3] a’nd it is likely that this involves ap a-subunit matogr. 447, 341-349. belonging to the so-called G,, class of a-subunits (for [IS] van Giersbergen, P.L.M., Shateer, %A., Henderson, A.K., Lai, review see ref. [22]). However, the NKJ receptor has also J., Nnkanishi, S., Yamamura, H-1. and Buck, S.H, (1991) Proc. been shown to inhibit CAMP accumulation through a Natl. Acad. Sci. USA 88. 1661-1665. pertussis toxin-sensitive mechanism which indicates [I91 Henderson, A.K.. Lai, J, Buck, S.H., Fujiwara, Y., Singh, G., Yamamura, MS., Nakanishi, S., Roeske, W.R. and Yamamnra, that the NK, receptor may couple to an &subunit H.I. (1990) Life Sci. 47, PLm7-PL-12. belonging to the Gi/G, class [233. Hence, the fact that [20] De Lean. A., Stadel, J.M. and Lefkowitz, R.J. (1980) J. Biol. the NKJ receptor is not expressed in its high-affmity Chem. 255, 7108-7117. form in CM0 cells may offer the opportunity to charac- [2l] Kurose. H., Regan. J.W., Caron, M.G. and Lelkowitz, R.J. terize the conversion of the receptor from its low-affin- (1991) Biochemistry 30, 3335-3341. [22] Simon, M.T., Strathmann, M.P. and Gautom, N. (1991) Science ity for.m to its high-affinity f~rrn by e.g. comp!emcnta- 252, 802-808. tion through co-expression of different G protein a- [23] Lnniyonu, A.. Sliwinski-lis. E. and Fleming, N. (1988) FEDS subunits in these cells. Lctt. 240, 186-190.

244