Proc. Nati. Acad. Sci. USA Vol. 90, pp. 1420-1424, February 1993 Neurobiology A f3-adrenergic receptor kinase-like enzyme is involved in olfactory signal termination (olfaction/slgnaling/kinase/subtypes/antibodies) SABINE SCHLEICHER*, INGRID BOEKHOFF*, JEFFREY ARRIZAtt, ROBERT J. LEFKOWITZt, AND HEINZ BREER*§ *University Stuttgart-Hohenheim, Institute of Zoophysiology, 7000 Stuttgart 70, Federal Republic of Germany; and tThe Howard Hughes Medical Institute, Departments of Medicine and Biochemistry, Duke University Medical Center, Durham, NC 27710 Contributed by Robert J. Leflcowitz, October 16, 1992

ABSTRACT We have previously shown that second- phorylation and cascade uncoupling may be mediated either messenger-dependent kinases (cAMP-dependent kinase, pro- by a cAMP-dependent protein kinase (PKA) or alternatively tein kinase C) in the olfactory system are essential in termi- by a cAMP-independent kinase (j3-adrenergic receptor ki- nating second-messenger signaling in response to odorants. We nase, DARK) (8-10). now document that subtype 2 of the ,B-adrenergic receptor In the present investigation we set out to explore whether kinase (fARK) is also involved in this process. By using a dual kinase pathway might also be active in olfactory subtype-specific antibodies to .ARK-1 and I3ARK-2, we show desensitization. The results indicate that besides second- that flARK-2 is preferentially expressed in the olfactory epi- messenger-controlled enzymes [PKA or PKC], a BARK-like thelium in contrast to 'mdings in most other tissues. Heparin, kinase, more specifically a JARK-2 subtype, is involved in an inhibitor of DARK, as well as anti-flARK-2 antibodies, (i) rapidly terminating olfactory signaling. completely prevents the rapid decline of second-messenger signals (desensitization) that follows odorant stimulation and (ii) strongly inhibits odorant-induced phosphorylation of ol- MATERIALS AND METHODS factory ciliary proteins. In contrast, (6ARK-1 antibodies are Materials. Sprague-Dawley rats were obtained from the without effect. Inhibitors ofprotein kinase A and protein kinase Zentralinstitut fur Versuchstierzucht, Hannover, F.R.G. Ci- C also block odorant-induced desensitization and phosphory- tralva (3,7-dimethyl-2,6-octadienenitrile) was obtained from lation. These data suggest that a sequential interplay ofsecond- International Flavors and Fragrances (Union Beach, NJ). messenger-dependent and receptor-specific kinases is function- Lilial [4-(1,1-dimethylethyl)-a-methylbenzenepropanol], eth- ally involved in olfactory desensitization. ylvanillin (3-ethoxy-4-hydroxybenzaldehyde), lyral [4-(4- hydroxy-4-methylpentyl)-3-cyclohexene-1-carboxyalde- The olfactory system responds precisely to iterative stimu- hydel, hedione (3-oxo-2-pentylcyclopentaneacetic acid lation (1) and thus can continuously monitor changes in the methyl ester) and eugenol [2-methoxy-4-(2-propenyl)phenol] odorous environment occurring between consecutive sniffs. were provided by W. Steiner, Baierbrunn, F.R.G. PKA This sensory behavior is due to characteristic phasic re- inhibitor (Walsh inhibitor) and heparin were obtained from sponses of receptor cells to short odor pulses (2), thereby Sigma. Calphostin C, a specific inhibitor of PKC, was pur- preventing a brief stimulus from being perceived as contin- chased from Calbiochem. [y-32P]ATP was supplied by Du- uous smell. The basis for this characteristic feature of olfac- Pont, and radioligand assay kits for cAMP and inositol tory neurons is a rapid termination of the odor-induced 1,4,5-triphosphate (InsP3) were purchased from Amersham. primary reaction-i.e., the second-messenger cascade initi- The purity grade of all chemicals used was >99%. ated by an odorant-occupied receptor is rapidly turned off (3, Antibodies. 8ARK-1 or f3ARK-2-specific polyclonal anti- 4). Recent studies have indicated that olfactory signaling is bodies were prepared as described in ref. 11 by injecting terminated by uncoupling the transduction cascade; this is rabbits with gluthathione-S-transferase fusion proteins con- accomplished via a negative-feedback reaction controlled by taining COOH-terminal sequences of either DARK isoform. the second messenger elicited upon odor stimulation. Sec- Sera were immunoaffinity-purified on antigen columns of ond-messenger-activated protein kinases [protein kinase A .BARK-1 or fARK-2 fusion proteins immobilized on Affi-Gel (PKA) or protein kinase C (PKC)] presumably phosphorylate 15. Antibodies that bound to one isotype column but passed elements ofthe transduction apparatus and thereby uncouple the other were that the signaling cascade (5). Subsequent experiments have through "subtype specific". Antibodies shown that stimulation with odorants indeed caused signifi- sequentially bound to and were eluted from both types of cant phosphorylation of olfactory ciliary proteins, and there column were subtype-nonspecific and recognized both en- is some circumstantial evidence indicating that receptor zyme forms. Specificity of antibodies was verified by immu- proteins for odorants may be modified via second-messenger- noblot analysis of COS cell extracts overexpressing f3ARK-1 controlled protein kinases (6). The emerging picture for signal or (3ARK-2 (11). termination in olfaction is, in many respects, reminiscent of Methods. Isolation of olfactory cilia. Partially purified desensitization phenomena in other systems-notably the preparations of chemosensory cilia from rat olfactory epi- visual and hormonal signaling pathways, where phosphory- thelia were produced according to the procedure described lation of receptors in an active state uncouples the reaction by Anholt et al. (12) and Chen et al. (13). Briefly, rat olfactory cascade by preventing further stimulation ofG proteins (7, 8). epithelia were dissected and collected in Ringer's solution For the most widely studied model, the p-adrenergic (120 mM NaCl/5 mM KCI/1.6 mM K2HPO4/1.2 mM system, two distinct pathways for phosphorylation and de- MgSO4/25 mM NaHCO3/7.5 mM glucose, pH 7.4). This sensitization have recently been discovered. Receptor phos- Abbreviations: PARK, P-adrenergic receptor kinase; PKA, protein kinase A; PKC, protein kinase C; InsP3, inositol 1,4,5-trisphosphate. The publication costs of this article were defrayed in part by page charge tPresent address: Oregon Health Sciences University, The Vollum payment. This article must therefore be hereby marked "advertisement" Institute, Portland, OR 97201. in accordance with 18 U.S.C. §1734 solely to indicate this fact. §To whom reprint requests should be addressed. 1420 Downloaded by guest on October 1, 2021 Neurobiology: Schleicher et al. Proc. Natl. Acad. Sci. USA 90 (1993) 1421 procedure and all subsequent operations were done at 40C. unaffected; however the "offset"-kinetic was significantly Cilia were detached from the epithelia by the calcium-shock reduced. The rapid termination of the signal is prevented by procedure, replacing the buffer by Ringer's solution/10 mM the PKA inhibitor as well as by heparin. A similar effect was CaCI2. After being agitated in an end-over-end shaker for 20 seen for the InsP3 response elicited by lyral (Fig. 1B); with min, the deciliated epithelia were removed by centrifugation either calphostin C, a specific blocker of PKC, or heparin at 7700 x g for 10 min. The supernatant was collected, and signal termination was prevented. These observations indicate the pellet was incubated again with Ringer's solution/10 mM that desensitization ofthe odorant-induced second-messenger CaC12 and then centrifuged. The combined supernatants cascade is blocked by inhibitors of second-messenger- containing the detached cilia were centrifuged at 27,000 x g dependent kinases but is also blocked by heparin. for 15 min. The pelleted cilia were resuspended in 10 mM On the basis ofthe results ofa previous study (6) it has been Tris HCl/3 mM MgCl2/2 mM EGTA, pH 8.0 and stored at hypothesized that one mechanism ofolfactory desensitization -700C. is odorant-induced phosphorylation of receptor proteins me- Phosphorylation procedure. The in vitro phosphorylation diated via second-messenger-activated kinases. To explore experiments were done by the procedure described by the mode of heparin action, its effect on odorant-induced Wilden and Kuhn (7). Odorants were dissolved in dimethyl incorporation of [32P]phosphate was analyzed. Fig. 1 C and D sulfoxide, giving a stock solution of 100 mM. To obtain the shows that stimulation ofolfactory cilia with either citralva or appropriate final odorant concentration different volumes of lyral in the presence of [y-32P]ATP significantly increased the the stock solution were added to the "stimulation buffer" incorporation of32p;. Labeling in the presence ofodorants was (200 mM NaCl/10 mM EGTA/50 mM Mops/2.5 mM about three times over basal level; -300 pmol of Pi per mg of MgCI2/1 mM dithiothreitol/0.05% sodium cholate/1 mM protein was incorporated after 1 sec. The odorant-induced ATP/1 ,M GTP/CaCl2 to give a free Ca2+ concentration of incorporation of 32p; was almost completely blocked by the 0.02 ,uM, pH 7.4). The odorant solutions were thoroughly Walsh inhibitor or calphostin C, respectively; in both cases mixed in an ultrasonic water bath and used immediately. heparin also reduced the 32p; incorporation level but only Controls were always performed by using the highest dimeth- reduced it by -60%. Both inhibitors together attenuated yl sulfoxide concentration, which never exceeded 0.05%. phosphorylation to the basal level (Fig. 1 C and D). Analysis The phosphorylation reaction was started by adding 20 ,ul of ciliary proteins incubated with [y-32P]ATP and 1 p.M lyral of the cilia preparation (20 ,ug of protein) to 80 ,ul of pre- for 1 sec indicated that a single polypeptide band (Mr 50,000) warmed (25°C) stimulation buffer containing odorants at the is labeled under these conditions (Fig. 1D, Inset), thus con- appropriate concentration and 1 ,uCi [-32P]ATP. After incu- firming recent observations (6); labeling of this protein was bation for the appropriate time intervals, usually 1 sec, the prevented by calphostin C. These observations suggest that, in reaction was stopped by adding 300 ,ul of an ice-cold solution addition to a second-messenger-controlled kinase, a heparin- of 25% trichloroacetic acid/5 mM phosphoric acid. The sensitive kinase is involved in olfactory signal termination and samples were placed on ice for 30 min followed by a thorough odorant-induced phosphorylation of olfactory proteins. mixing of the precipitated proteins. Aliquots (100 ,ul) of each To further characterize the involvement of a heparin- sample were spotted onto Whatman 3 MM filter paper (4 x sensitive kinase in olfactory signaling, similarity ofthe kinase 4 cm) and immersed in a solution of 10% trichloroacetic to the heparin-sensitive BARK was analyzed. Monospecific acid/5 mM phosphoric acid. After being thoroughly washed antiserum raised against ,ARK from rat (11) was used in the for 30 min, the filters were dried at room temperature for 2 hr. assay systems. The data in Fig. 2A show that the kinetics of The radioactivity adsorbed onto the filters was quantitated by citralva-induced cAMP signals are unchanged by a control scintillation counting. serum (1:1000); however, the anti-,BARK antiserum (1:1000) Measurement ofrapid changes in second-messenger con- completely prevents rapid desensitization of the cAMP re- centrations. A rapid-quench device with three syringes was sponse. The inhibitory effect of the ,ARK antibodies is also used to determine the subsecond kinetics of the odorant- observed in phosphorylation experiments (Fig. 2B); odorant- induced changes in second-messenger concentrations. Rapid induced incorporation of [32P]phosphate is significantly re- kinetic experiments were done as described (3). Briefly, duced by the kinase antibodies. These results suggest that a syringe I contained the olfactory cilia (0.5 ,ug/,ul) in 70 ,ul of ,lARK-like kinase is involved in olfactory signaling. buffer (10 mM Tris.HCI/3 mM MgCl2/2 mM EGTA); syringe Two closely related, yet distinct, BARK subtypes have II was filled with 360,41 of stimulation buffer containing the been identified (16, 17). Although some differences concern- odorants dissolved as described above. All solutions were ing activity and regional distribution in the brain have been adapted to 370C. Syringe III contained perchloric acid (7%) seen, the distinct physiological roles of the respective sub- cooled to 0°C. The reaction was started by mixing the types are still elusive. As a first step to evaluate whether a contents of syringe I and II and quenched with 300 Al of distinct kinase subtype was active in the olfactory system, perchloric acid ejected from syringe III. Activation of the different tissues were probed for immunoreactivity by using three syringes was controlled by an IBM AT computer. The subtype-specific antibodies (11). The results of ELISA as- quenched samples were collected and cooled on ice before says (Fig. 3A) show that j3ARK-i predominates in erythro- analysis for second-messenger concentrations. cytes as well as in lung and brain tissue; the samples are only weakly reactive with anti-pARK-2. In contrast, the olfactory RESULTS cilia show high levels of pARK-2 reactivity but show only Receptor-specific kinases, like rhodopsin kinase and BARK, basal activity for PARK-1. To scrutinize the specificity of are efficiently blocked by polyanions, such as heparin (14). In PARK subtypes in olfactory cilia, both subtype-specific a first approach to evaluate whether such second-messenger- antibodies were assayed over a broad range of dilutions. As independent kinases are involved in the rapid termination of depicted in Fig. 3B, at low antibody concentrations (high olfactory signaling, olfactory cilia pretreated with heparin dilution) only pARK-2 reactivity is detected; only at low were assayed for kinetics ofthe cAMP response elicited by the dilutions is some ,ARK-1 reactivity detectable. The specific fruity odorant citralva and the InsP3 response induced by lyral. and selective reaction of olfactory cilia with 8ARK-2 anti- As depicted in Fig. 1A, upon stimulation with citralva, control bodies was confirmed by immunocytochemical approaches samples show a very rapid and transient increase in cAMP (unpublished observation). level, thus confirming previous observations. In samples pre- The proposed specific role of a BARK-2-like kinase in treated with either Walsh inhibitor or heparin, the "onset"- olfactory cilia was confirmed in phosphorylation experi- kinetic of the odorant-induced cAMP-signal was virtually ments. Fig. 4 shows that the odorant-induced incorporation Downloaded by guest on October 1, 2021 1422 Neurobiology: Schleicher et al. Proc. Natl. Acad Sci. USA 90 (1993) A B 800-

600-

0e E 400-

200- t

-200-100 0 100 200 300 400 500 600 Time, msec Time, msec (. 1)

.'C il " 1(

^'! -.--

a_ J. -1 4I1) 1(

IJ I ( ) I( L I ( I()I II t ( ) OrIIr-l 11 ( )J1d r ( al)J -:LI t ( ) Jor41:I C1 ) T-;; ) dur1 ;1 - .'1K \ -I -l I lcr1 I1 IPIK \-I KTX( - I k r-'In *11 I K1\( I lv' ,i;lI !'

FIG. 1. Protein kinase inhibitors prevent desensitization of olfactory second-messenger signaling as well as odorant-induced protein phosphorylation. (A) Cilia preparations isolated from the olfactory epithelium ofrat were stimulated with citralva (1 /LM). In control samples a rapid and transient increase of cAMP concentration is seen, confirming reported results (3). In samples pretreated with either Walsh inhibitor (3.8 AM) or heparin (1 ,M) for 15 min at 40C, the "onset"-kinetic of the odorant-induced signal was unaffected; however, the "offset"-kinetic was significantly changed. Both inhibitors completely blocked desensitization. Data represent the means of three experiments; deviations are <10%. *, Odorant; A, Walsh inhibitor; *, heparin. (B) Stimulation with lyral (1 ,uM) induced a rapid and transient response of InsP3. Calphostin C (1 AM), the specific inhibitor of PKC or heparin, completely prevented rapid termination of the lyral-induced InsP3 cascade; the Walsh inhibitor had no effect. Data are the means of three experiments; deviation is

A B

1400 1000 FIG. 2. Effect of anti-pARK antise- 1200 rum on desensitization and phosphoryla- 800 tion of olfactory cilia. (A) Isolated olfac- tory cilia were preincubated either with 1()(.)(1 Er" control rabbit serum (dilution 1:1000) or with anti-PARK antiserum raised in rab- 600) 80(1 bit (dilution 1:1000). Pretreated cilia were stimulated with citralva (1 jAM), and the cAMP response was determined. Desen-

G 600) 400 sitization was completely prevented by anti-PARK antibodies. Data are the 0 means of three experiments. e, Odorant; 400 a, odorant plus anti-PARK. (B) Attenu- 200( ation of odorant-induced phosphoryla- tion. Pretreatment of olfactory cilia with 200 ---- P3ARK-specific antibodies (1:1000) signif- icantly reduced incorporation of 32p; in- control 0 Odorant Odorant duced by odorant stimulation; -20(1 + + DARK serum had no effect. Data are the means -10(1 100 200 500 control Time. miscc serum antiscrum of four experiments ± SD. DISCUSSION Some major differences also exist between the olfactory and fB-adrenergic systems. The odorant-induced second- suggest that two types ofkinases are The data presented here messenger response in olfactory cilia is turned off very off the olfactory transduction cascade. involved in turning rapidly-the cAMP or InsP3 signal decays after only 50 a kinase Inhibition of either second-messenger-controlled msec-whereas waning of the hormone response occurs (PKA or PKC) or a heparin-sensitive kinase prevents olfac- within minutes (20). The physiologically important rapidity of tory desensitization as well as odorant-induced phosphory- desensitization in olfaction probably requires specialized odorant receptors; Fig. lation of olfactory proteins (putative enzymes and a specific mechanism. In line with this notion is 1; ref. 6). Thus, the rapid termination of odorant-activated our discovery that in the olfactory neuroepithelium a char- transduction processes in olfactory cilia appears similar to acteristic subtype of the heparin-sensitive kinase, a pARK- the mechanisms mediating desensitization of the 8-adrener- 2-like kinase, is preferentially expressed, in contrast to brain gic system, which also includes two different kinases: the and peripheral tissues where the PARK-1 isoform predomi- cAMP-dependent protein kinase (PKA) and the heparin- nates (Fig. 4; ref. 11). Furthermore, it is a PARK-2-like sensitive 8-adrenergic kinase (PARK) phosphorylate the kinase that is specifically involved in the rapid desensitization f3-adrenergic receptor at specific sites and thereby disrupt ofolfactory signaling (Fig. 5). Thus, the 3ARK-2-type kinase further p-adrenergic receptor/G protein coupling (16, 18). may represent another example of particular isoforms of The two kinases represent alternative pathways of desensi- molecular entities that build the signal-transduction appara- tization active at different agonist concentrations and in- tus in olfactory neurons. Characteristic isoforms of a G volved in, respectively, heterologous or homologous desen- protein (Golf, ref. 21) and adenylate cyclase (type III, ref. 22) sitization (8). However, although agonist-occupied receptors have recently been described. have been conclusively shown to be phosphorylated by these Not only are different kinase subtypes active in desensi- kinases in the p-adrenergic receptor system (19), such proof tizing odorant and hormone receptors but there also may be has not been established in the olfactory system. Recent significant differences in the mechanisms. In the adrenergic findings (Fig. 1 Inset and ref. 6) are, however, consistent with system, blockade of PKA or pARK-1 partially inhibited this hypothesis. receptor phosphorylation as well as desensitization (14). In

A B

0.6 0.5 H3ARKS2

0.5 0.4 FIG. 3. Tissue-specific reactivity of an- tibodies discriminating between different

0.4 8ARK subtypes (o, 3ARK-1; u, pARK-2). 0).3 (A) Samples from different tissues of rat were probed for immunoreactivity by using I- =0.3) the ELISA technique and subtype-specific antibodies (dilution 1:1000). The 8ARK-1 0.2 subtype predominates in brain and lung tis- x 0.2- sue, as well as in erythrocytes (Erythr.); these samples display low reactivity with In cilia ().1 PARK- se anti-,3ARK-2. contrast, olfactory 0.1 show rather high levels of 8ARK-2 and only low activity for PARK-1. (B) Dose-response curve of the two subtype-specific antibodies

-u L-U- 0 on olfactory cilia preparation in ELISA as- Cilia Erx'thr. I.unit Cortex 100 1000 10.00() says. Data are the means of three separate [Dilution] experiments; deviation is <1%o. Downloaded by guest on October 1, 2021 1424 Neurobiology: Schleicher et al. Proc. Natl. Acad. Sci. USA 90 (1993)

1000 - sensitization. These observations raise the possibility that the two kinases act sequentially in a reaction cascade that is initiated by activating the second-messenger-dependent ki- ,1ARK-1 nase. Such an interrelationship between two different kinases 800- would represent a distinctive mechanism for receptor desen- sitization and might be necessary for rapid signal termination. The details of how these kinases interact, however, remain 600- elusive. Several schemes may be envisaged: (i) PKA could phosphorylate the odorant receptors and thereby trigger the exposure of specific sites for f3ARK-2-mediated phosphate incorporation. In this model, the action of 3ARK-2 would be 400- substrate-controlled, a mechanism characteristic for the structurally related rhodopsin kinase (23, 24). (ii) PKA might directly phosphorylate and thus activate /ARK-2; this type ofreaction would resemble a kinase cascade, as, for example, 200- described for regulation ofmetabolic pathways (25). (iii) PKA may phosphorylate and thereby inactivate endogenous in- hibitors of a f3ARK-2-like kinase. In summary, these data suggest a hypothesis different from 1^0 1000 10,000 signal termination in vision, where a single kinase (rhodopsin [Dilution] kinase) uncouples the rhodopsin/transducin cascade (26) and different from the 13-adrenergic system, where PKA and FIG. 4. Dose-response curve of the different subtype-specific j3ARK-1 work in parallel to phosphorylate and desensitize anti-.BARK antibodies on odorant-induced phosphorylation of olfac- the f3-adrenergic receptor under certain conditions (8). In tory cilia. Isolated olfactory cilia were pretreated with different olfaction, the rapid termination ofthe odorant-induced signal dilutions of subtype-specific antibodies and incubated with transduction cascade may be accomplished by a coordinated, [(-32P]ATP plus odorants for 1 sec. Note that the odorant-induced sequential action of a second-messenger-controlled kinase incorporation of [32P]phosphate was not affected by 8ARK-1 anti- and a PARK-2-like kinase. Although not yet proven, the bodies; however, labeling was significantly reduced with BARK-2 olfactory receptor proteins themselves are the most likely antibodies. Data are the means of three experiments ± SD. point of action of these kinases. contrast, in the olfactory system inhibitors of second- This work was supported by the Deutsche Forschungsgemein- messenger-dependent kinases completely blocked both phos- schaft Br 712/10-2 and Grant HL16037 from the National Institutes phorylation and desensitization, whereas inhibition of of Health to R.J.L. P3ARK-2 by either heparin or anti-PARK antibodies only partially blocked phosphorylation but completely prevented 1. Ottoson, D. (1956) Acta Physiol. Scand. 35, Suppl. 122, 1-83. 2. Getchell, T. V. & Shepherd, G. M. (1978) J. Physiol. (London) 282, desensitization (Fig. 1). Because PKA-mediated phosphory- 541-560. lation continues in the presence of anti-PARK antibodies or 3. Breer, H., Boekhoff, I. & Tareilus, E. (1990) Nature (London) 344, heparin, this phosphorylation is clearly insufficient for de- 65-68. 4. Boekhoff, I., Tareilus, E., Strotmann, J. & Breer, H. (1990) EMBO J. 9, 2453-2458. 1200- 5. Boekhoff, I. & Breer, H. (1992) Proc. Nati. Acad. Sci. USA 89,471-474. 6. Boekhoff, I., Schleicher, S., Strotmann, J. & Breer, H. (1992) Proc. Nati. Acad. Sci. USA 89, 11983-11987. anti-,3ARK-2 7. Wilden, U. & Kahn, H. (1982) Biochemistry 21, 3014-3022. 1000 - 8. Lefkowitz, R. J., Hausdorff, W. P. & Caron, M. G. (1990) Trends Pharmacol. Sci. 11, 190-194. 9. Hausdorff, W. P., Caron, M. G. & Lefkowitz, R. J. (1990) FASEBJ. 4, O 800- 2881-2889. 10. Dohlmann, H. G., Thorner, J., Caron, M. G. & Lefkowitz, R. J. (1991) Annu. Rev. Biochem. 60, 653-688. 0 11. Arriza, T. M., Dawson, J. L., Simerly, R. B., Martin, L. J., Caron, ~.600- M. G., Snyder, S. H. & Leflowitz, R. J. (1992) J. Neurosci., in press. 12. Anholt, R. R. H., Aebi, U. & Snyder, S. H. (1986) J. Neurosci. 6, 1962-1%9. ~~~~anti-p8ARK-1400 :'sse 13. Chen, Z., Pace, U., Heldman, J., Shapira, A. & Lancet, D. (1986) J. Neurosci. 6, 2146-2154. 14. Lohse, M. J., Benovic, J. L., Caron, M. G. & Lefkowitz, R. J. (1990) J. Control Biol. Chem. 265, 3202-3209. 200 15. Breer, H. & Boekhoff, I. (1991) Chem. Senses 16, 19-29. 16. Benovic, J. L., DeBlasi, A., Stone, W. C., Caron, M. G. & Lefkowitz, R. J. (1989) Science 246, 235-240. 17. J. 0' Benovic, L., Onorato, J. J., Arriza, J. L., Stone, W. C., Lohse, M., -200 -100 0 100 200 300 400 500 Jenkins, N. A., Gilbert, D. J., Copeland, N. G., Caron, M. G. & Lefkowitz, R. J. (1991) J. Biol. Chem. 266, 14939-14946. Time, msec 18. Hausdorff, W. P., Bouvier, M., O'Dowd, B. F., Irons, G. P., Caron, M. G. & Leflowitz, R. J. (1989) J. Biol. Chem. 264, 12657-12665. FIG. 5. Desensitization of odorant-induced second-messenger 19. Sibley, D. R., Benovic, J. L., Caron, M. G. & Lefkowitz, J. (1987) Cell signals is specifically prevented by 3ARK-2 antibodies. Isolated 48, 913-922. olfactory cilia pretreated with either control serum or subtype- 20. Shear, M., Insel, P. A., Melmon, K. L. & Coffino, P. (1976) J. Biol. specific antibodies were stimulated with citralva (1 1AM), and the Chem. 251, 7572-7576. induced second-messenger responses were determined. Note that 21. Jones, D. T. & Reed, R. R. (1989) J. Biol. Chem. 262, 14241-14249. 22. control serum (control) and PARK-1 antibodies (1:5000) had no effect Bakalyar, H. A. & Reed, R. R. (1990) Nature (London) 250, 1403-1406. 23. KOhn, H. & Dreyer, W. J. (1972) FEBS Lett. 20, 1-6. on the offset-kinetic of the odorant-induced second-messenger sig- 24. Bownds, M. D., Dawes, J., Miller, J. & Stahlman, M. (1972) Nature nal; however, with 83ARK-2 antibodies (dilution, 1:5000) rapid de- (London) 237, 125-127. sensitization was completely prevented. Data are the means of three 25. Cohen, P. (1985) Eur. J. Biochem. 151, 439-448. experiments; deviation was <10%0. 26. Stryer, L. (1986) Annu. Rev. Neurosci. 9, 87-119. Downloaded by guest on October 1, 2021