Proc. Nat. Acad. Sci. USAVol. 69, No. 6, pp. 1494-1498, June 1972

The Chemistry of Olfactory Reception: Stimulus-Specific Protection fromSulfhydryl Reagent Inhibition

(group-specific protein reagents/electro-olfactogram/frog/ethyl n-butyrate/N-ethyl maleimide)

MARILYN LEVISOHN GETCHELL* AND ROBERT C. GESTELAND

Department of Biological Sciences, Northwestern University, Evanston, Illinois 60201

Communicated by Irving M. Klotz, March 27, 1972

ABSTRACT The gtoup-specific protein reagent, N-ethylmaleimide, irreversibly blocks the electrical responseof the olfactory receptor organ of the frog to odorousstimuli. If the odorous substance, ethyl n-butyrate, inconcentrations high enough to saturate the receptor sys-tem, is present in the nasal cavity before and during a briefexposure to N-ethylmaleimide, the nose, after a wash anda recovery period, responds in nearly normal fashion tovapors of ethyl n-butyrate. Responses to other odoroussubstances, except those closely related to ethyl n-buty-rate, are abolished. We propose that we can use this pro-tection technique to identify the properties of the variousreceptor sites in the nose, and possibly to characterize thereceptor substances.

Specific receptor- sites on the dendritic membrane of theolfactory bipolar neuron have been postulated to account forthe ability to recognize and discriminate between diverseodorous compounds. This study provides direct evidencethat such specific receptor sites exist, and describes a methodthat can be used to determine the response properties ofspecific receptor sites.

Group-specific protein reagents disrupt functioning ofnerve cells. The compound action potential of frog sciaticnerve, a measure of summated spike activity of synchronouslyexcited axons, is abolished by HgCl2 (1), p-chloromercuri-benzoate (ClHgOBz) and N-ethylmaleimide (NEM) (2).These reagents, and the fluorinated dinitrobenzenes F(NO2)2-Bz and F2(NO2)2Bz (3), abolish action potentials in the squidgiant axon (4, 5). Synaptic transmission in frog and rat nerve-muscle preparations is blocked by ClHgOBz and o-iodosoben-zoic acid (6). HgCl2 and ClHgOBz reduce the response of theeel electroplax to synaptic activators, while NEM blocks re-polarization of the cell after synaptic activation (7). Inchemosensory systems, HgCl2, applied to the contact chemo-receptors of blowflies, eliminated behavioral responses tostimulation of these receptors (8). Nerve responses to stimula-tion of chemoreceptors in carp (9, 10) and salmon (11, 12)were reduced by HgCl2, NEM, and F(NO2)2Bz.We report that NEM, which reacts primarily with sulfhy-

dryl groups of proteins, irreversibly blocks olfactory

Abbreviations: ClHgOBz, chloromercuribenzoate; NEM, N-ethylmaleimide; F(NO2)2Bz, 1-fluoro-2,4-dinitrobenzene; F2r(NO2)sBz, 1,5-difluoro-2,4-dinitrobenzene; EOG, electro-olfacto-gram.* Present address. Monell Chemical Senses Center, University ofPennsylvania, Philadelphia, Pa. 19104.

receptor function in frog nose. The effect of NEM can beprevented by the presence in the nose of an odorous substancebefore and during the NEM exposure. After this treatment,the olfactory receptors respond to the odorous substance thatwas present during the exposure to NEM and to otherclosely related substances, but will not respond to mostsubstances that are normally effective stimuli.

Group-specific protein reagents have found wide applica-tion in the study of pure enzymes in solution. When anenzyme is exposed to a group-specific reagent, its enzymaticactivity may be reduced or abolished. In most cases, the lossof enzymatic activity can be correlated with reaction of thegroup-specific reagent with one or more amino-acid residues inthe active site of the protein. In these cases, if substrate ispresent when the enzyme is exposed to the group-specificreagent, the amino-acid residue in the active site is protectedagainst reaction with the inhibitor, and the protein retains allor a large percentage of its enzymatic activity; the substrateis bound to or near the amino acid in the active site of theenzyme. In our experiments, an electrical sign of nerve-cellresponse is used as an assay for the effect of the inhibitor.

Activity of olfactory receptor cells can be measured byrecording of summated activity at the surface of the olfactorybulb, of activity in the axons of the olfactory nerve, ormeasurement of the voltage between the surface of the.olfactory mucosa and an indifferent electrode. Substancesthat block the generation of action potentials, such as thelocal anaesthetics procaine and cocaine, abolish stimulus-induced changes in activity at the olfactory bulb and inolfactory nerve axons, but leave the stimulus-induced voltagerecorded from the olfactory mucosa unchanged (13). Thisresult agrees with studies on other receptors, where it hasbeen found that local anaesthetics block generation of actionpotentials in axons, but do not affect the events that occurwhen the stimulus interacts with the nonelectrically excitablereceptor membrane. The slow voltage change appearing atthe surface of the olfactory mucosa in response to a stimulusis called the electro-olfactogram (EOG), and is a measure ofreceptor membrane response (13, 14). We use the EOG as ameasure of response of olfactory receptors in these experi-ments to distinguish between effects of the group-specificreagent on the nerve-cell receptor membrane and its effectson generation and transmission of the action potential innerve axon. An EOG evoked by stimulation with ethyln-butyrate is shown in Fig. la; it is typical of the responseevoked by many odorous substances.

1494

Dow

nloa

ded

by g

uest

on

Apr

il 5,

202

1

Olfactory Receptor Chemistry 1495

Preliminary experiments showed that NEM irreversiblyabolished EOGs to all the stimuli used. Abolition of slowpotentials was accompanied by lack of response in theolfactory bulb. Lower concentrations of NEM or shortertimes of exposure to the olfactory epithelium caused reductionof the amplitude of the EOG and changes in its waveform(Fig. lb).

MATERIALS AND METHODSFrogs (Rana pipiens) were pithed anteriorly and posteriorlyalong the neuraxis. The nasal cavity was sealed by placingCenco Softseal Tackiwax over the buccal aperture of theinternal naris, and holding it in place with a piece of cottonin the frog's mouth. The olfactory epithelium was exposedby removal of the overlying skin, cartilage, and the dorsalmucosa. All recordings were made from the center of theolfactory eminence, the region of the frog nose where receptor-cell density is highest.A chlorided coil of silver wire wrapped in saline-moistened

cotton served as the ground electrode; it was placedthrough a slit in the skin into the submandibular lymphsac. The active electrode was chlorided silver wire bridgedto the olfactory mucosa by a fine-tipped glass pipette filledwith 3 M KCl. Signals were led through a FET input couplerand were displayed on an oscilloscope and recorded.

All solutions were made up in distilled water. Ringer'ssolution (15) had the following composition: 115 mM NaCl;2.5 mM KCl; 2.0 mM CaCl2; 1.1 mM Na2HPO4; 0.4 mMNaH2PO4. The pH of all solutions was 6.5.The stimuli used were: ethyl n-butyrate and methyl

n-butyrate (Eastman Kodak Co.), ethyl acetate and cis-1,2-dichloroethylene (Fisher Scientific Co.), and l-limonene(K and K Laboratories, Inc.). All were of reagent-gradepurity. Reagent-grade NEM was obtained from EastmanKodak Co.

Cofnpressed air was deodorized by passage throughactivated alumina, silica gel, and activated charcoal and wasmoisturized by passage through distilled water. The moistair stream flowed at a constant rate over the frog's olfactoryepithelium; odor vapors were diluted in this stream forstimulation. The stimulus strength was chosen so that anEOG of maximum amplitude for each stimulus was obtained.All stimulations were 1.5 sec in duration; an interval of5 min was allowed for recovery between stimulations. Glasssyringes with 27-gauge stainless steel needles were used tofill and drain the olfactory cavity; separate syringes for fillingand draining were used for each solution.

RESULTSTo demonstrate protection, the following experiments wereperformed:(1) Two or three EOGs to ethyl n-butyrate were recorded

from the untreated mucosa.(2) The nasal cavity was filled with Ringer's solution;

the solution was withdrawn and replaced immediatelyand after 5 min. After 10 min, the solution was withdrawn.

(3) Three EOGs to ethyl n-butyrate were recorded. Theaverage of the amplitudes of these three EOGs was thecontrol amplitude (100%).

(4) The epithelium was exposed to one of the followingsequences of treatments:(a) Ringer's solution for 5 min (the solution waschanged immediately and after 2.5 min); for 2 min (the

a

- -off rae

FIG. 1. Effects of NEM on the waveform of the ethyl n-bu-tyrate EOG. Curve a shows the typical EOG evoked by stimula-tion with ethyl n-butyrate vapor after exposure of the epitheliumto Ringer's solution (pH 6.5) for 10 min. An "off" response isevident. After exposure of the epithelium to Ringer's solution for 5min, 2 mM NEM in Ringer's solution for 2 min, and Ringer'ssolution for 3 min, followed by a 10-min wash with Ringer's solu-tion, the same stimulus evokes an EOG with an initial positivewave and a longer-lasting, smaller negative wave (curve b). Thebars above the EOGs represent the stimulus durations.

solution was changed once immediately); and for3 min (the solution was changed immediately, and after1.5 min). This sequence of solution changes was followedin b, c, and d as well.(b) 12.5 mM Ethyl n-butyrate in Ringer's solution for5 min; for 2 min; and for 3 min.(c) Ringer's solution for 5 min; 4 mM NEM in Ringer'ssolution for 2 min; and Ringer's solution for 3 min.(d) 12.5 mM Ethyl n-butyrate in Ringer's solution for5 min; 12.5 mM ethyl n-butyrate + 4 mM NEM inRinger's solution for 2 min; and 12.5 mM ethyl n-butyrate in Ringer's solution for 3 min.

(5) One EOG to ethyl n-butyrate was recorded.(6) The nasal cavity was filled with Ringer's solution for

10 min, with two initial solution changes.(7) EOGs to ethyl n-butyrate were elicited and recorded

every 5 min for the next 50 min.

There was no statistically significant change (Student'stwo-tailed t-test, P = 0.01) in the amplitudes of the EOGsmeasured in steps 1, 3, 5, and 7 when the above regimen wasdone with Ringer's solution alone (step 4a).When the epithelium was exposed to 12.5 mM ethyl

n-butyrate in Ringer's solution for 10 min (step 4b), theEOG amplitude was reduced by 66%. After the mucosa waswatied with Ringer's solution, the response amplitudeslowly recovered (circles, Fig. 2). This result agrees with theobservation that repeated, closely-spaced stimulations withvapors of a substance evoke successively smaller responses tothat substance and, to a lesser extent, diminish responses toother substances (13). Several minutes after stimulation arerequired for recovery to original sensitivity.When the mucosa was treated with Ringer's solution for

5 min, 4mM NEM in Ringer's solution for 2 min, and Ringer'ssolution again for 3 min (step 4c), the EOG amplitude wasreduced by 77%; washing with Ringer's solution failed tocause recovery (triangles, Fig. 2). However, when 12.5 mMethyl n-butyrate was present before, during, and afterexposure to NEM (step 4d), washing the mucosa withRinger's solution caused the EOG amplitude to recover to76% of its original value (squares, Fig. 2).

Interaction between the odorant and the protein reagent insolution was ruled out spectrophotometrically and physi-ologically. NEM absorbs strongly at 312 nm due to the carbon-

Proc. Nat. Acad. Sci. USA 69 (1972)

b

2"c,,

Dow

nloa

ded

by g

uest

on

Apr

il 5,

202

1

1496 Biochemistry: Getchell and Gesteland

Minutes

FIG. 2. Responses to ethyl n-butyrate. Relative EOG ampli-tudes evoked by vaporous stimulations with ethyl n-butyrateafter treatment with NEM and ethyl n-butyrate in solution (seeResults, Protection experiment). The amplitudes of these EOGs arerepresented by the points on the graph. Circles represent treat-ment 4a, triangles treatment 4b, and squares treatment 4c. Eachpoint represents the average of three experiments. Protection isdemonstrated. S--O, ethyl n-butyrate alone; a ethyln-butyrate + NEM; A-\A, NEM alone.

carbon double bond, the protein-reactive portion of the mole-cule. Addition to the double bond results in a decrease inabsorbance at this wavelength. A solution of 12.5 mM ethyln-butyrate and 4 mM NEM in Ringer's solution has an ab-sorbance equal to the sum of the absorbances of solutions ofthe two components. If a solution of 12.5 mM ethyl nbuty-rate + 4 mM NEM in Ringer's solution was placed on theolfactory epithelium for 2 min (without prior or subsequentexposure to the odorant in the solution), the EOG was ir-reversibly abolished. Thus, the effect called protection wasnot due to inactivation of NEM by ethyl n-butyrate in solu-tion.

Control experiments with the sciatic nerve of the frogdemonstrated that the ability of an odorant to protect itsEOG from inhibition by NEM was due to the presence in theolfactory epithelium of receptors that specifically interactwith the odorant, and was not due to nonspecific binding ofthe stimulus molecule to membrane components involved inbioelectrogenesis.The sciatic nerve was exposed and removed according to

standard dissecting procedures; the epineurium was peeledoff to expose 2.5- to 3.5-cm lengths of nerve to the reagents.The nerve was placed in a moist nerve chamber of Luciteplastic with 0.06 cm diameter platinum electrodes, and wasstimulated with biphasic square waves at a frequency of20 Hz. The electrical parameters measured were the stimulusrequired to elicit a minimal nerve response (the threshold)at stimulus durations ranging from 0.02 msec to 0.20 msec in0.01-msec increments, and the amplitude and shape of theaction potential to a stimulus of 0.80 V lasting for 0.20 msec.The electrodes were blotted between each set of measurementsto remove any excess solution that might have been carriedover with the nerve. At the end of each experiment, the ability

of the nerve to continue functioning after intense stimulationwas tested by stimulating the nerve with supramaximalstimuli for several minutes.Four series of experiments, analogous to the protection

experiments performed on the olfactory epithelium, were doneas follows:

(1) After 15 min in Ringer's solution in a glass perti dish, thedesheathed nerve was placed in the moist nerve chamber,and its response to electrical stimulation was measured.

(2) The nerve was returned to Ringer's solution for 10 min,lifted out of the solution once immediately after sub-mersion, and lifted once af ter 5 min to simulate the solu-tion changes in the nasal cavity and to agitate the solu-tion around the nerve.

(3) The response of the nerve to electrical stimulation wasmeasured.

(4) The nerve was exposed to one of the following series ofsolutions:(a) Ringer's solution for 5 min (the nerve was lifted outof the solution once immediately and once after 2.5 min);for 2 min (the nerve was lifted out of the solutiononce immediately); and for 3 min (the nerve was liftedout of the solution immediately and after 1.5 min). Theabove pattern of lifting the nerve out of the solutionswas followed for each of these treatments.(b) 12.5 mM Ethyl n-butyrate in Ringer's solution for5 min; for 2 min; and for 3 min.(c) Ringer's solution for 5 min; 4 mM NEM in Ringer'ssolution for 2 min; Ringer's solution for 3 min.(d) 12.5 mM Ethyl n-butyrate in Ringer's solution for5 min; 12.5 mM ethyl n-butyrate + 4 mM NEM inRinger's solution for 2 min; 12.5 mM ethyl n-butyrate inRinger's solution for 3 min.

(6) One measurement-the threshold of the action potentialto a stimulus of 0.20-msec duration-was taken.

(6) The nerve was returned to Ringer's solution for 10 minas in step 2 above.

(7) The response of the nerve to electrical stimulation wasmeasured.

The last two steps were repeated twice, to make the timeperiod allowed for recovery equivalent to that in the experi-ments on the olfactory epithelium.Treatment of the nerve with Ringer's solution (step 4a) had

no effect on consecutive threshold measurements. The am-plitude and shape of the compound action potential remainedconstant; a pronounced afterpotential was present. Thelatency of the action potential appeared to decrease slightlyafter repeated washings with Ringer's solution. Intense stimu-lation produced a temporary decrease in excitability thatlasted for less than 1 min; conduction block was never ob-served.Exposure of the nerve to 12.5 mM ethyl n-butyrate in

Ringer's solution (step 4b) produced a slight, irreversible in-crease in the nerve threshold. The amplitude and shape of theaction potential remained the same throughout the experi-ment. Intense stimulation did not produce a conduction block.Exposure of the nerve to 4 mM NEM in Ringer's solution

(step 4c) resulted in a complete conduction block after intensestimulation for less than 1 min. Three washings of the nervewith Ringer's solution failed to restore conduction. NEM, inthe absence of intense stimulation, abolished the after-po-

Proc. Nat. Acad. Sci. USA 69 (1972)

Dow

nloa

ded

by g

uest

on

Apr

il 5,

202

1

Olfactory Receptor Chemistry 1497

tential, increased the latency of the response noticeably, andcaused a decrease of more than 50% in the action potentialamplitude. Threshold measurements gave variable results.

In the series of experiments designed to test for a protectiveeffect of ethyl n-butyrate (step 4d), intense stimulation causeda conduction block that was not reversed by washing withRinger's solution. In the absence of intense stimulation, theafterpotential was abolished, the latency of the action poten-tial increased greatly, the action potential amplitude de-creased, and the threshold measurements gave variable re-sults, as with NEM alone. Thus, no protective effect of ethyln-butyrate on the nerve was observed.

In order to determine whether other odors interact withreceptor cells at the same site as ethyl n-butyrate, severalother odorous vapor-phase stimuli were used on noses thathad been exposed to 4 mM NEM while protected with12.5 mM ethyl n-butyrate in Ringer's solution. The results forcis-1,2-dichloroethylene, anisole, and l-limonene are clear-cut:ethyl n-butyrate is not able to protect the receptor sites forthese odorants from the effects of NEM. The EOG amplitudesevoked by these stimuli from a nose treated with NEM areidentical to the amplitudes from a nose protected by ethyln-butyrate and treated with NEM.

Results with ethyl acetate, an ester that does not have thetypical sweet ester odor, were complicated by the fact thatrepeated stimulations with this substance resulted in a seriesof EOGs with decreasing amplitudes. Since a significant differ-ence (P = 0.02) is seen between the effects of 4mM NEM and4 mM NEM + 12.5 mM ethyl n-butyrate on the amplitudeof the EOG to ethyl n-butyrate within the first 15 min afterwashing the epithelium with Ringer's solution, this timeperiod was used to compare the effects of 4 mM NEM withthe effects of 4 mM NEM + 12.5mM ethyl n-butyrate on theEOG elicited by puffs of ethyl acetate. There was no signifi-cant difference, at the same confidence level, between theamplitudes of EOGs to ethyl acetate from mucosae that hadbeen subjected to the two different treatments. Thus, thesetwo esters probably do not share the same receptor sites.When methyl n-butyrate, an ester structurally closely

related to ethyl n-butyrate and smelling more like it, was usedas the stimulus in the vapor phase, results similar to thoseseen with ethyl n-butyrate were obtained. The amplitude ofthe EOG to methyl n-butyrate elicited after treatment of theepithelium with ethyl n-butyrate-ethyl n-butyrate + NEM-ethyl n-butyrate (step 4d in the nose experiment), followed bya Ringer's solution wash, recovered to about 80% of the con-trol value. The result suggests that ethyl n-butyrate andmethyl n-butyrate interact with the same receptor sites.Treatment of the mucosa with ethyl n-butyrate in solution

reduced the amplitude of the EOGs to cis-1,2-dichloroeth-ylene, anisole, l-limonene, and ethyl acetate, even though theseodorants interact with receptor sites different than those forethyl n-butyrate. Studies of olfactory receptor cell responsesin the frog show that each cell is capable of responding to avery large number of chemically diverse stimuli (14). Thus,it is most likely, in view of the selectivity we find for a singlesite, that there are many different kinds of receptor sites on themembrane of a single cell. At low levels of stimulus intensity,we expect that each activated receptor site produces anincrement of current. Thus, at low stimulus intensities, theEOG, which is due to summated current flow, increases

stimulus, all of the receptor sites of one type will be active andthose cells with many of these sites will be strongly depolar-ized. Another stimulus acting at a different receptor site onthe same cell will evoke less current than it would if the cellwere not already strongly stimulated. Since infusion of asolution of ethyl n-butyrate into a nose represents maximalstimulation of all of the sites capable of reacting to it, reducedEOG amplitudes evoked by other stimuli delivered after theinfusion are not surprising. We, thus, have a reasonablemechanism to account for the selective, partial fatigue of theEOG observed previously (13).

DISCUSSION

The experiments reported here imply that the receptor forethyl n-butyrate is highly selective, interacting only withsubstances chemically and structurally closely related to thissubstance. It is possible, of course, that the ethyl n-butyratereceptors are normally not very specific, and that what we seein these experiments is an NEM-induced restructuring ofreceptor substance such that it intimately embraces ethyln-butyrate. However, with enzymes, cofactors unrelated insize and shape to the substrate can often protect active sitesfrom reaction with inhibitors. For example, MnSO4 protectsthe active site of isocitrate dehydrogenase (EC 1.1.1.42) fromreaction with NEM (16). In this case, protection results fromphysical blockage of the active site residues, rather than re-structuring of the active site around the protective substance.In a nose, modification of a receptor membrane around areceptor site might be expected to cause dramatic changes inits ability to generate an electrical response. Yet the EOGfrom the protected epithelium was nearly as large as that froman untreated epithelium.

Sulfhydryl groups are clearly implicated in the interactionsbetween the receptor and all of the odorous substances wehave used, as well as in the transduction process that couplesthe activated receptor to membrane permeability changes.Nearly all of the activity measurable in the nose is blocked bya short exposure to NEM or to HgCl2. However, the receptorprocess does not exclusively involve sulfhydryl binding. In an

experiment where we compared the amplitudes of EOGs tosix different stimuli after a 2-min exposure to 4mM NEM, thereduction of EOG amplitude varied directly with the dipolemoment of the stimulus (17). NEM most strongly inhibitedresponses to the most-polar stimuli. Presumably the less-polar stimuli are relatively more effective in an NEM-treatednose because they are reacting with groups that are less sus-ceptible to reaction with NEM.We have some evidence about the nature of the transduc-

tion process. The simplest hypothesis is that the stimulussubstance binds to a receptor molecule that is a structuralelement of the receptor cell membrane. The result of thisinteraction would be a local change in membrane structurethat would increase membrane permeability. Since single-cellstudies show that any particular substance can have bothexcitatory and inhibitory effects, we might suppose that thereare two molecular receptors for ethyl n-butyrate, one of whichwhen occupied increases membrane permeability to sodiumions and the other increases membrane permeability to K + orCl- ions. Alternatively, there could be one receptor substancethat is coupled to either of two membrance conductance"gate" substances, one gate producing excitatory effects, the

approximately with stimulus intensity (13). For a strong

Proc. Nat. Acad. Sci. USA 69 (1972)

other inhibitory effects. Exposure of a nose to 4 mM NEM for

Dow

nloa

ded

by g

uest

on

Apr

il 5,

202

1

1498 Biochemistry: Getchell and Gesteland

2 min changes the waveform of the EOG evoked by ethyln-butyrate. The time constant of the negative (excitatory)component is increased greatly during its falling phase and toa lesser degree during its rising phase. An initial positive wavewith a rapid time constant, which is not present normally, isseen after NEM treatment (Fig. lb). NEM affects excitatoryand inhibitory processes differentially and affects the rise andfall times of excitation differentially. This is also seen withother group-specific reagents. For example, F(NO2)2Bz leavesthe rise time of the negative component unaffected, whileincreasing the time constant of the falling phase; an initialpositive component, larger than that seen with NEM, appearsin the EOG to ethyl n-butyrate (17). These changes areobserved in the EOG whether or not a nose is protected by anodorant during the NEM exposure. Thus, we asssume thatNEM acts on two steps in the reception-excitation process.It binds strongly to the receptor molecules, preventing bindingof the stimulus. This action can be protected against. It alsomodifies the effectiveness of a stimulus in changing membraneexcitatory and inhibitory conductances. This modificationoccurs whether or not a protective stimulus is present duringNEM exposure. Therefore, the ion-gate molecules are prob-ably not identical with the receptor molecules, and NEM canact on the gates in the presence of a stimulus because thestimulus does not bind to the gates. It is clear that brief ex-posure to NEM disrupts gate functions much less than recep-tor functions, since a protected nose responds very well to theprotecting substance (about 70% normal) after exposure toNEM.

Lastly, we note that by treating a nose with labeled NEMafter a protection experiment, we may be able to isolate andcharacterize the substances that react with the protectingstimulus (18).

We thank Dr. Thomas V. Getchell and the Monell ChemicalSenses Center for providing the facilities in which part of thiswork was done, and Dr. Robert Cagan for the spectrophoto-metric measurements. This investigation was supported in partby a National Science Foundation Predoctoral Fellowship toM. G., and in part by grants from the National Science Founda-tion and the U.S. Air Force to R. G.

1. Del Castillo-Nicolau, J. & Hufschmidt, H. J. (1951). Nature167, 146-147.

2. Smith, H. M. (1958) J. Cell. Comp. Physiol. 51, 161-171.3. Cooke, I. M., Diamond, J. M., Grinnell, A. D., Hagiwara,

S. & Sakata, H. (1968) Proc. Nat. Acad. Sci. USA 60, 470-477.

4. Huneeus-Cox, F., Fernandez, H. L. & Smith, B. H. (1966)Biophys. J. 6, 675-689.

5. Hillman, G. R. & Mautner, H. F. (1968) Biol. Bull. 135,423.

6. Chang, C. C., Lu, S. E., Wang, P. N. & Chuang, S. T. (1970)Eur. J. Pharmacol. 11, 195-203.

7. Karlin, A. & Bartels, E. (1966) Biochim. Biophys. Acta 126,525-535.

8. Dethier, V. G. (1955) Quart. Rev. Biol. 30, 348-371.9. Hidaka, I. & Yokota, S. (1967) Jap. J. Physiol. 17, 652-666.

10. Hidaka, I. (1970) Jap. J. Physiol. 20, 599-609.11. Sutterlin, A. M. & Sutterlin, N. (1970) J. Fish. Res. Bd.

Can. 27, 1927-1942.12. Sutterlin, A. M. & Sutterlin, N. (1971) J. Fish. Res. Bd.

Can. 28, 565-572.13. Ottoson, D. (1956) Acta Physiol. Scand. 35, Suppl. 122, 1-83.14. Gesteland, R. C., Lettvin, J. Y. & Pitts, W. H. (1965) J.

Physiol. 181, 525-559.15. Takagi, S. F., Kitamura, H., Imai; K. & Takeuchi, H.

(1969) J. Gen. Physiol. 53, 115-130.16. Colman, R. F. & Chu, R. (1970) J. Biol. Chem. 245, 601-607.17. Getchell, M. L. (1971) Ph.D. thesis, Northwestern Uni-

versity, Evanston, Ill.18. Fox, C. F. & Kennedy, E. P. (1965) Proc. Nat. Acad. Sci.

USA 54, 891-899.

Proc. Nat. Acad. Sci. USA 69 (1972)

Dow

nloa

ded

by g

uest

on

Apr

il 5,

202

1