Brain Research, 97 (1975) 61-78
61
~ Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
SLOW P O T E N T I A L S OF T H E T U R T L E O L F A C T O R Y BULB IN RESPONSE TO O D O R S T I M U L A T I O N OF T H E N O S E
R. W. BEUERMAN*
Florida State University, Department of Biolo,~ieal Science, Tallahassee, Fla. 32306 (U.S.A.) (Accepted April 7th, 1975)
SU MMARY
Odor stimulation of the nose in the box turtle and the gopher tortoise produced a characteristic series of slow potentials in the olfactory bulb which were referred to as the odor evoked response. When recorded with direct coupling, the odor evoked response had 3 components: wave I, a short duration monophasic event; wave II, a long duration variation in the DC potential; and wave I11, an oscillatory potential superimposed on wave 1I. Waves I and II were negative at bulbar surfaces receiving olfactory input and positive deep within the bulb. This series of potentials could be evoked by 3 methods of odor stimulation: (1) large puffs delivered from odorant test bottles, (2) small puffs delivered from a syringe and (3) continuous flow with concentration and nasal flow rate parameters controlled by an olfactometer. When the odor evoked response was recorded at a bulbar locus, these potentials were seen in response to each stimulation and the amplitudes of each wave were reproducible with the same stimulus. The amplitudes of the 3 waves were compared in the gopher tortoise and differed with the 3 odorants tested - - high purity geranioi, technical grade geraniol and amyl acetate. Odorant concentration also directly affected the response amplitudes of all 3 wave components. The amplitudes of waves I and lli markedly decreased with closely spaced stimulations recovering to near the initial values when the interstimulus interval was increased severalfold. This series of sensory evoked potentials is considered to reflect the processing of odor information from the olfactory receptors by the olfactory bulb.
INTRODUCTION
A salient feature of physiological studies of the olfactory bulb in many verte* Present address: Department of Physiology and Biophysics, University of Washington SJ-40, Seattle, Wash. 98195, U.S.A.
62 brates, from fish to man. has been the appearance of electrical potential oscillat,,~m~s(~l large amplitude in response to olfactory stimuli in both acute ~,e,:',~. ,~.._,~.:~.;~7, ~:,.~ ~tnd chronic ~,'~,e:~,~4,~:~,'1:~,'~'~preparations. Oscillatory activity in the central nervous system often occurs in conjunction with sensory stimulation, it has been found as a lale component of the visually evoked potential ~,~ and as a superimposition on the b-wave of the electroretinogram ~. A problem with sensory evoked potentials has been to explain the relatio,~ship of the stimulus parameters to the activity of the cells giving rise to the response. Consequently, waveforms and amplitudes of visual, auditory, and somatic evoked potentials have been examined using many types and methods of stimulationeS,4L but: odor elicited slow potentials of the olfactory bulb have generally not come under such scrutiny. Ottoson :~.37, recording from the surface of the frog and rabbit oltactory bulbs, found that the oscillations were superimposed on a slow, DC potential that increased in amplitude with increasing concentrations of butanol. Leveteau ~md McLeod 26 have described an odor evoked slow potential in a single oll~actory glomerulus. Higashino ~z recorded oscillatory potentials differing in amplitude in the frog bulb in response to a homologous series of alcohols. In this study, olfactory bulb slow potentials were examined in two species of turtles and odor stimulation of the nose in preparations of either species was found to elicit a characteristic series of 3 slow potentials. These same slow potentials (referred to as the odor evoked response) were observed with 3 methods of odor delivery and several odorants of different chemical composition. The amplitudes of the component waves of the odor evoked response could be compared between preparations if the stimulation parameters of odorant concentration, odorant species and nasal flow rate were controlled. The interstimulus interval was found to be important fbr reproducibly eliciting the odor evoked response. Preliminary reports of some of these results have been presentedS.~L MATERIALS AND METHODS
Animal preparation Specimens of both sexes and differing ages of the box turtle (Terrapennecarolina) and the gopher tortoise (Gorpheruspolyphemus) were used. The box turtles ranged in weight from 0.4 to I. I kg and the gopher tortoises from 0.4 to 2.9 kg. Only animals in obvious good health, i.e., with clear eyes and no nasal discharge, responsive to handling were studied. The animal was anesthetized with ethylurethane (1.50 g/kg body weight, i.p.) and occasional movements of the jaw and neck muscles were blocked by an injection of D-tubocurarine chloride (1.2 mg/kg body weight) into the large, dorsal neck muscles. Additional curare was administered as needed during the experiment. The head was fixed in a holder and a tracheal cannula was inserted and connected to a respirator. The spinal cord was severed between the third and lburth vertebrae. The cranium was opened with a dental drill, the olfactory structures exposed, and the vomeronasal nerves identified and severed. In good preparations, the surface vascularization of the olfactory bulbs was highly visible due to copious blood flow. in
63 some experiments, the bulbs were acutely transected at the coronal sulcus ~8 to avert corticofuga113,14 and bulbofugal influences x3,14, which are considered to be inhibitory16, 25. Post mortem examination indicated that these bulb cuts were usually complete. The glass recording pipettes could not be advanced through the surface membrane of the bulb without severe dimpling. Small openings were made in the menix which allowed the electrodes to easily penetrate the bulb without observable dimpling as viewed under a dissecting microscope at × 16 magnification. Temperature was monitored periodically by a thermometer in the esophagus and was found usually to be between 24'~C and 20 °C.
Electrical recording methods Electrical activity from the bulbs was recorded with low-resistance (3-7 M~)~) micropipettes (1.0 m m O.D., 0.5 mm I.D.) pulled to have long shanks and filled with 3.0 M NaCI. The reference electrode was a coil of tungsten wire placed under the skin of the neck. Tungsten wire was inserted into the recording micropipette and connected to a coaxial cable which led to the input stage of a Grass P-16 amplifier with negative capacitance. Single-ended AC (band-pass 1-100 Hz) and DC (DC-100 Hz) outputs were viewed on an oscilloscope. The band-pass settings were selected on the basis of response frequencies obtained in preliminary studies. The highest dominant frequency component of the bulb oscillations was 25 Hz (gopher tortoise), whereas the more usual values were between 13 Hz and 17 Hz. Permanent records were made either on a Sanborn recorder or by an oscilloscope camera. Differential, AC-coupled activity from the primary olfactory nerve or vomeronasal nerve4S, 49 was recorded in conjunction with the bulbar activity in some experiments. The preamplifier output was conducted to a short time averaging circuit v and displayed on a Sanborn recorder. The use of time averaged multi-unit activity allows spike bursts to be treated as events and has been used in studies on the olfactory system 34,49, as well as the spinal cord. 42.
Stimulation procedures Odor stimulation was by 3 methods: (1) an air flow dilution olfactometer as described by Tucker 49, (2) manually operated syringes, and (3) polyethylene squeeze bottles. The olfactometer was employed to accurately specify the odor concentration and nasal flow rate and for long stimulus durations (I-30 sec). The gopher tortoise was primarily used for experiments with the olfactometer. The syringe delivery was used to provide brief ( 150-200 msec) and reproducible odor puffs. New, 10-ml glass syringes were specially cleaned and connected by Teflon tubing to a glass delivery tube placed directly in front of the naris to be stimulated. Various amounts of odorous air (1-4 ml) were drawn from the bottle head space over the liquid odorant and sent into the nasal cavity by depressing the plunger. Unintentional stimulation of the contralateral bulb was investigated and found not to be a problem. Technical grade geraniol in a 1:10 dilution with mineral oil was used with the syringe delivery system and will be called the standard stimulus. After an odorous stimulus, laboratory air filtered in a similar manner as by the olfactometer 49, was
64 passed through the delivery tube for 15--30 sec at a flow rate of lt) ml/sec. Another 30-40 sec was allowed betk)re presenting the next stimulus. Odor puffs were also delivered f'rom polyethylene squeeze bottles, containing a few milliliters of undiluted test odorant, were used to determine the responsiveness o(a preparation. A small amount of butyric acid was put on a pad in the wash bottle and allowed to evaporate to the point where it would not be overwhelming by subjective determination. With practice, reproducible odor pufl~ could be delivered as previously noted 49. Passage of the syringe delivered puff through the glass delivery tube was indicated by the output from a thermistor flowmeter. This flowmeter employed feedback circuitry which would supply current to the thermistor in the delivery tube in response to cooling by air-flow with a response time of less than 0.1 sec. Calibration of the electrical flowmeter was by a wet-test gasmeter and a stopwatch. The recorded output of the flowmeter indicated the shape, time course and flow rate of the stimulus puff (see Fig. 2). With the syringe delivery method, a 3-ml puff was about 150 msec in duration and had a maximal flow rate of 6 7.5 ml/sec. This flowmeter was also used in conjunction with the olfactometer.
Odorants Technical grade geraniol (Matheson, Coleman and Bell), used as the standard stimulus, was diluted with paraffin oil (Matheson, Coleman and Bell). In experiments with the olfactometer, technical grade geraniol, high purity geraniol (Givaudan Corp.) and amyl acetate (Matheson, Coleman and Bell) were compared. Subjectively both types of geraniol smelled similar; however, technical grade geraniol was somewhat stronger. Other test odorants were amyl acetate, isoamyl acetate and methyl salicylate (Matheson. Coleman and Bell) and butyric acid (Eastman Organic Chemicals).
Chromatographic analysis Instrumental analysis of the two geraniol samples was performed on a Varian Aerograph model 705 gas-liquid chromatograph with a 5-ft. stainless steel analytical column of 1/8 in. diameter. The column packing was 60/80 mesh Chromosorb W coated with 1090 Carbowax 20 M. Electrometer output was led to a Beckman pen recorder. Odorant samples in the gas phase were taken from the output of the olfactometer and deposited on the column by the method used by Pierson 39. Geraniot began to elute when the column reached 120 °C. The injector was at 155 ~C. Areas of odor peaks were measured with a planimeter: Areas obtained from olfactometer controls. wash air, were subtracted from the areas obtained from the odor samples. Areal analysis of the chromatograms showed that technical grade geraniol was only 50 % geraniol (the unidentified peaks may be citronellol or nerol 6) while the high purity geraniol was 97 ~,i pure (Fig. 1). Amyl acetate from this source (same lot no. and analyzed by these methods was previously shown not to have contaminants 3:~. The absence of odor peaks in the olfactometer control indicated the effectiveness of the procedures used to filter the wash air (Fig. 1).
65
I
GERANtOL, tech.
I
I
I
T
1
1
I ,,
GERANIOL,h.p,
m
] 9
m
i
CONTROL
] 15 MtNS.
Fig. 1. Chromatograms of 30-sec samples of technical grade geraniol (top trace), high purity geraniol (middle trace) and filtered air (bottom trace) were collected from the output of the olfactomete[ in the gas phase. The two types of geraniol were at saturation. Measurements of the areas under the peaks showed that technical grade geraniol was 50%0 pure whereas high purity geraniol was 9 7 ~ pure. The filtered air did not contain odor peaks. Gas chromatograph settings range 0.1, attenuator × 8.
66
I+
I
[
A
B
WAVE •
4 0 0 msec
~PLITOOE
400
Fig. 2. The odor evoked response was recorded using AC (upper trace) and DC (middle trace) coupling with the associated thermistor flowmeter analog record (lower trace). In this series of 4 consecutive responses (A and B) the recording micropipette was placed in the depths of the box turtle bulb. Each stimulation was a 3-ml puff of the standard stimulus which was indicated by the output of the thermistor flowmeter (C). The recorder was halted between stimulations while the wash air was applied. Waves I and II were measured as shown in B from an extension of the prestimulus base line, Wave llI was measured as in B and a factor for the envelope of wave Ill was the product of the dimensions shown in D. Amplitude indicators were 0.4 mV in each case. Also note the time base was expanded in B.
Data analysis The amplitudes of the wave components of the odor evoked response were measured in terms of potential as shown in Fig. 2. Amplitude of the early potential, wave 1, was measured at its maximum from an extension of the base line. Wave 1I amplitude was taken at the maximum of a curve drawn to fit the inferred time course with the prestimulus base line used as a reference level. The amplitudes of the oscillations, wave Ill, were referred to wave il, and the peak-to-peak values of complete cycles of the dominant frequency were combined to give an average value (Fig. 2B). Some of the records were retouched for photographic purposes. RESULTS
Components of the odor evoked response A micropipette driven into the deep layers of the olfactory bulb of either a box turtle or gopher tortoise recorded a characteristic series of slow potentials in response to a puff of odorous air directed into the nose. This series of slow potentials was referred to as the odor evoked response and when recorded from deep within the bulb appeared as in Fig, 2. There was an initial short duration positive event, wave I, with a latency of 150-200 msec, wave lI was also positive and of much longer duration on which was superimposed an oscillatory potential, wave lII (the induced waves of Adrian2). This series of potentials underwent predictable changes in amplitude and polarity which were dependent on the anatomical layering of the bulb s ] 0 . When the odor evoked response was recorded from surface regions of the bulb receiving olfac-
67 tory input, waves l and I 1 were negative; however, all 3 waves reached their greatest magnitudes in the deep layers of the bulb. The ongoing activity of the olfactory bulb in these two species of turtles was always of low amplitude, unsynchronized and was identical with the prestimulus baseline (Fig. 2).
Reproducibility of the slow waves Repeated applications of a single odor stimulus or stimulation with a number ot test odorants showed that the 3 wave components were elicited in a reproducible and consistent manner. Successive puffs of the standard stimulus showed that the wavetbrm, wave amplitudes and wave time courses were similar in each response as seen in Fig. 2A and B. The responses to 25 applications of the standard stimulus were recorded at a bulb location and gave the following values: wave I, X = 0.85 mV, S.E.M. -- 0.02 mV; wave II, X -= 0.62 mV, S.E.M. - 0.04 mV; envelope of wave Ill calculated as in Fig. 2D, X = 204.9 sq. ram, S.E.M. ~- 3.6 sq. mm. The wave components of the odor evoked response showed variations in amplitude and duration when test odorants of different chemical composition were used as stimuli (Fig. 3). Waves 1 in Fig. 3 were similar in duration, but varied in amplitude while waves II and III varied both in duration and amplitude. Butyric acid, which is least effective among these test odorants for eliciting the olfactory nerve response 4~,49 evoked a small response even with vigorous stimulation.
Duration of the odor evoked slow potentials Measurements of the half-amplitude duration of wave I (60 msec ± 5 msec) indicated no relationship to the method of stimulation, i.e., olfactometer, syringe, or squeeze bottle nor parameters of stimulation. The time course of wave l I depended on stimulation procedures. A short odor puff of less than 200 msec with the syringe delivery method produced a potential with a complete time course of 2-5 sec. Using the olfactometer, half-amplitude decay times of wave I1 were measured from the peak amplitude in the tortoise bulb for technical grade geraniol and amyl acetate, with stimulus duration at 5 sec and nasal flow rate at 10 ml/sec. This decay time for wave I I with geraniol increased with concentration; reaching 8 sec at 10 o.2,5 (of vapor saturation). These decay times for wave l! with amyl acetate were less than 5 sec at 10_3.0 (of vapor saturation) and more than 12 sec at 10 0.')5 (of vapor saturation). At a given recording locus, neither stimulus concentration nor the time of the stimulus application appeared to affect the duration of the oscillations, wave I Il. This observation was tested in detail with the olfactometer. The recording pipette was positioned in the depths of a box turtle bulb. Different combinations of stimulus concentration (amyl acetate at 10-1.°, 10 4.0 and 10 a.0 of saturation) and stimulus duration (1, 2, 4, 6 and 8 sec) were tested. The 58 responses were analyzed and the duration of the oscillations varied from 0.7 sec to 1.1 sec. Statistical tests for trends within groups were not significant nor was the two-way analysis of variance, P ,~- 0.1, Thus, the duration of wave II! under these conditions was not affected by stimulus concentration or duration. The duration of wave ill could be altered by modulation
68 STIMULATION
PUFF BOTTLE
herd :mff
frame Puff
t ~ - . ~ r
ACETATE
PC~
""""
f
i
~-
'r
T.YL SAL,CYL Vt
l
,
Fig. 3. The odor evoked response appeared consistently to test stimuli of different odorants. Odor puffs were delivered by squeeze bottles; puffs of various intensities were reproducible with practice. Wave II1 is seen more clearly in the AC (upper) trace of each pair which was recorded at higher gain. Records were from the depths of the box turtle bulb. of the nasal airflow rate alter the initial odor evoked response which produced a series of oscillations of long duration (3-5 sec), but of low frequencies (5-9 Hz).
Accessory olfactory bulb Although the vomeronasal nerve, the afferent input to the accessory bulb (AB), was routinely severed, it was desirable to determine whether the AB could contribute to the odor evoked response since the granule cell layer of the AB is contiguous with that of the olfactory bulb 9. By simultaneously recording from two micropipettes connected to different capacity coupled amplifiers, this possibility was tested. To insure activation of the AB, butyric acid was also used as a stimulus since the vomeronasal receptors are very responsive to this odorant 48.
69 With the olfactory nerve severed and the vomeronasal nerve left intact, the odor evoked response could not be induced in either the olfactory bulb or the AB with puffs of butyric acid. With the olfactory nerve intact, a response was evoked in both, but at the AB the response was considerably attenuated. In other experiments, the receptor spike discharge was recorded from a small twig of the vomeronasal nerve while AB activity was monitored with a recording pipette. Again, after severing the olfactory nerve no odor evoked response was found in the AB, even though the receptor axons conducted action potential responses. Thus, the odor evoked response of the olfactory bulb was considered not to be complicated by evoked potentials from the AB.
Comparison o[ the o!factory nerve response and the odor evoked response The odor evoked response of the bulb was recorded simultaneously with the averaged multi-unit activity of the olfactory receptor axons in 5 tortoises. The olfactometer was utilized to permit longer stimulus durations (2-30 sec) and so that the odorant concentration and nasal flowrate could be specified for comparisons to earlier studies48, 49. Unexpectedly, in the tortoise preparations in which olfactory nerve and bulb activity were recorded together, oscillations were constantly found as part of the olfactory nerve response at higher stimulus concentrations (Figs. 4 and 5). The use of a short time constant of integration allowed comparisons to be made although this maneuver decreased the phasic portion of the nerve response and typical records are seen in Figs. 4 and 5. Wave I of the odor evoked response coincided with the rising phase of the integrated olfactory nerve discharge while the bulb oscillations, wave Ill, occurred at the time of maximal amplitude in the olfactory nerve response as seen in Figs. 4 and 5. The olfactory nerve oscillations were observed only at higher stimulus concentrations which was in contrast to the bulb oscillations which occurred with the lowest stimulus concentrations used (Fig. 5). Also as demonstrated in Fig, 5, the olfactory nerve oscillations adapted rapidly to successive stimulations at a particular concentration.
I
I
2 seconds
Fig. 4. The odor evoked response of the bulb (A) was recorded simultaneously in the tortoise with the time averaged (0.02 sec) multiunit activity (B) from an olfactory nerve twig. The onset of the bulbar oscillations preceded the onset of the peripheral oscillations. Nasal flow rate was l0 ~.~ ml/ sec and amyl acetate concentration was 10 1.0 of saturation at 20 °C.
70
B
10-2.0 10-0.5
if!
B ,
10-2.5
Io-LO iO-3,O
N
-
J
~
'
~
"
-
~
5 Sec I,,
I
,' ',~,,~ ', ','~ ",~w '/, ,' w u k " ~ V A / ~ ~,~ t , ~ , w , ~ , v ~ , ~ , ~ , ¢ ~
10%5
'
CONTROL N
Fig. 5. The odor evoked response of the tortoise bulb (B) was recorded simultaneously with the olfactory nerve (N) discharge. The first stimulus of amyl acetate in each pair of records was 6 sec and succeeding stimulations were 2 sec, except at l0 0.5 (of saturation) where all stimulus durations were 2 sec in order to prevent damage to the olfactory receptors. The nose was not flushed with filtered air between stimulations of the same concentration, but was flushed for 60 sec between changes in concentrations. The olfactory nerve response (N) exhibited oscillations (arrows) at the first response for concentrations l0 °.L10 1.5. Amyl acetate concentration expressed in terms of saturation at 20 ~C and nasal flowrate was 10 ~.o ml/sec.
Successive stimuli delivered with short interstimulus intervals resulted in decreased amplitudes of the wave components of the odor evoked response. In contrast, these adaptive effects were not observed in the receptor axon discharge and repeated stimulations elicited similar amplitude responses (Fig. 5). As seen in Fig. 5, wave 1 adapted very rapidly and was usually present as part of only the first bulb response for successive stimulations at a particular concentration, even at lower stimulus concentrations. At higher stimulus concentrations, the amplitude of wave I11 showed a marked decrease in response to closely spaced stimulations (Fig. 5). After increasing the interstimulus interval several fold and applying a filtered air wash, waves I and Ill had large amplitudes for the first stimulus of the next concentration series. Adaptive effects were also tbund Ibr waves I and 111 in the box turtle when stimuli with short interstimulus intervals were delivered from the olfactometer or syringe delivery. Recovery depended, as in the tortoise, on increasing the interstimulus interval and the use of a filtered air wash. In these experiments wave I1 was not usually' recorded, but when observed it also exhibited adaptation effects. At high stimulus concentrations, the bulb oscillations reappeared as an 'off' response when short duration stimulations were used. When the stimulus duration was increased from 6 sec (as in Fig. 5) to as long as 30 sec, the receptor axon response maintained a tonic level of activity, but the odor evoked response was present only at
71 the beginning of the stimulation period. Geraniol was usually not effective in producing bulbar 'off' responses when the syringe or the olfactometer was employed for stimulus delivery.
Control stimulations Control stimulations with the olfactometer showed that the mechanical effects of nasal airflow or odor switching transients did not elicit the odor evoked response. Using the olfactometer filtered air was drawn through the nasal passageway in the same manner as for an odorous stimulus. At low nasal flow rates there was no detectable response in the olfactory bulb or nerve (Fig. 5, Control). With much higher nasal flowrates (10 ~.a ml/sec) low amplitudes of waves I and 11 were observed, but this was probably due to the high velocity wash air picking up adsorbed odorant from the odor delivery switches as has been previously found 49. Filtered air from the olfactometer was analyzed in this study and found not to contain odor peaks (Fig. I). Odor comparisons The amplitudes of the wave components of the odor evoked response were found to be a function of odorant species and concentration when amy[ acetate, and two types of geraniol, technical grade and high purity, were directly compared at several concentrations. Stimulation and recording procedures were standardized with tortoises whose olfactory bulbs were acutely transected at the coronal sulcus. The tip of the recording pipette was placed 300/tin below the level at which wave I became positive, as close to the midline of the bulb as possible and 1.5-2.0 mm posterior from the anterior pole. The odor was switched on for 5 sec, a 30-sec flush was conducted, another 3-5 rain were allowed to elapse, and the next stimulus-flush sequence was presented. Initially, a nasal flow rate of I ml/sec was used, but the response amplitudes to the two types of geraniol were almost unmeasurable at a concentration of 10 l.r, (ot" saturation). Increasing the nasal flow rate to 10 ml/sec overcame this difficulty. All 3 odorants were tested in each animal. The response amplitudes of the 3 wave components of the odor evoked response, plotted in Fig. 6, were shifted up the ordinate from the least effective stimulus, high purity geraniol, to technical grade geraniol and to amyl acetate the most effective stimulus. The two geraniols were particularly interesting since technical grade geraniol was found to be an odorant mixture by gas-liquid chromatography (Fig. 1). Although the initial amplitudes for the 3 waves were similar at 10 2.o (of saturation), the slopes were greater for the amplitudes of the 3 waves for technical grade geraniol. Apparently odorant differences are more effective than simply concentration difference (Fig. 6, compare A and B). Variations in amyl acetate concentration in these preparations produced a different set of curves for the wave amplitudes as shown in Fig. 6C. The amplitudes of waves I and II rose rapidly to maximums from relatively large values at 10 2.o (of saturation). In particular, wave Ii amplitude at 10- ~.0 (of saturation) for amyl acetate was approximately equal to the plateau for technical grade geraniol. Wave llI showed little amplitude variation for this concentration range of amyl acetate and had approximately the maximal value reached with technical grade geraniol (compare wave II1 in Fig. 6B and C).
72
GERANIOL, hKIh pur,fy
0 -20
-L5
40
-I0
-0
G~RANK~L, flc~P,k;ol
3O
20
i
~.0
-I.5
"I.0
"0.5
40
0
C
30
61
~ Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
SLOW P O T E N T I A L S OF T H E T U R T L E O L F A C T O R Y BULB IN RESPONSE TO O D O R S T I M U L A T I O N OF T H E N O S E
R. W. BEUERMAN*
Florida State University, Department of Biolo,~ieal Science, Tallahassee, Fla. 32306 (U.S.A.) (Accepted April 7th, 1975)
SU MMARY
Odor stimulation of the nose in the box turtle and the gopher tortoise produced a characteristic series of slow potentials in the olfactory bulb which were referred to as the odor evoked response. When recorded with direct coupling, the odor evoked response had 3 components: wave I, a short duration monophasic event; wave II, a long duration variation in the DC potential; and wave I11, an oscillatory potential superimposed on wave 1I. Waves I and II were negative at bulbar surfaces receiving olfactory input and positive deep within the bulb. This series of potentials could be evoked by 3 methods of odor stimulation: (1) large puffs delivered from odorant test bottles, (2) small puffs delivered from a syringe and (3) continuous flow with concentration and nasal flow rate parameters controlled by an olfactometer. When the odor evoked response was recorded at a bulbar locus, these potentials were seen in response to each stimulation and the amplitudes of each wave were reproducible with the same stimulus. The amplitudes of the 3 waves were compared in the gopher tortoise and differed with the 3 odorants tested - - high purity geranioi, technical grade geraniol and amyl acetate. Odorant concentration also directly affected the response amplitudes of all 3 wave components. The amplitudes of waves I and lli markedly decreased with closely spaced stimulations recovering to near the initial values when the interstimulus interval was increased severalfold. This series of sensory evoked potentials is considered to reflect the processing of odor information from the olfactory receptors by the olfactory bulb.
INTRODUCTION
A salient feature of physiological studies of the olfactory bulb in many verte* Present address: Department of Physiology and Biophysics, University of Washington SJ-40, Seattle, Wash. 98195, U.S.A.
62 brates, from fish to man. has been the appearance of electrical potential oscillat,,~m~s(~l large amplitude in response to olfactory stimuli in both acute ~,e,:',~. ,~.._,~.:~.;~7, ~:,.~ ~tnd chronic ~,'~,e:~,~4,~:~,'1:~,'~'~preparations. Oscillatory activity in the central nervous system often occurs in conjunction with sensory stimulation, it has been found as a lale component of the visually evoked potential ~,~ and as a superimposition on the b-wave of the electroretinogram ~. A problem with sensory evoked potentials has been to explain the relatio,~ship of the stimulus parameters to the activity of the cells giving rise to the response. Consequently, waveforms and amplitudes of visual, auditory, and somatic evoked potentials have been examined using many types and methods of stimulationeS,4L but: odor elicited slow potentials of the olfactory bulb have generally not come under such scrutiny. Ottoson :~.37, recording from the surface of the frog and rabbit oltactory bulbs, found that the oscillations were superimposed on a slow, DC potential that increased in amplitude with increasing concentrations of butanol. Leveteau ~md McLeod 26 have described an odor evoked slow potential in a single oll~actory glomerulus. Higashino ~z recorded oscillatory potentials differing in amplitude in the frog bulb in response to a homologous series of alcohols. In this study, olfactory bulb slow potentials were examined in two species of turtles and odor stimulation of the nose in preparations of either species was found to elicit a characteristic series of 3 slow potentials. These same slow potentials (referred to as the odor evoked response) were observed with 3 methods of odor delivery and several odorants of different chemical composition. The amplitudes of the component waves of the odor evoked response could be compared between preparations if the stimulation parameters of odorant concentration, odorant species and nasal flow rate were controlled. The interstimulus interval was found to be important fbr reproducibly eliciting the odor evoked response. Preliminary reports of some of these results have been presentedS.~L MATERIALS AND METHODS
Animal preparation Specimens of both sexes and differing ages of the box turtle (Terrapennecarolina) and the gopher tortoise (Gorpheruspolyphemus) were used. The box turtles ranged in weight from 0.4 to I. I kg and the gopher tortoises from 0.4 to 2.9 kg. Only animals in obvious good health, i.e., with clear eyes and no nasal discharge, responsive to handling were studied. The animal was anesthetized with ethylurethane (1.50 g/kg body weight, i.p.) and occasional movements of the jaw and neck muscles were blocked by an injection of D-tubocurarine chloride (1.2 mg/kg body weight) into the large, dorsal neck muscles. Additional curare was administered as needed during the experiment. The head was fixed in a holder and a tracheal cannula was inserted and connected to a respirator. The spinal cord was severed between the third and lburth vertebrae. The cranium was opened with a dental drill, the olfactory structures exposed, and the vomeronasal nerves identified and severed. In good preparations, the surface vascularization of the olfactory bulbs was highly visible due to copious blood flow. in
63 some experiments, the bulbs were acutely transected at the coronal sulcus ~8 to avert corticofuga113,14 and bulbofugal influences x3,14, which are considered to be inhibitory16, 25. Post mortem examination indicated that these bulb cuts were usually complete. The glass recording pipettes could not be advanced through the surface membrane of the bulb without severe dimpling. Small openings were made in the menix which allowed the electrodes to easily penetrate the bulb without observable dimpling as viewed under a dissecting microscope at × 16 magnification. Temperature was monitored periodically by a thermometer in the esophagus and was found usually to be between 24'~C and 20 °C.
Electrical recording methods Electrical activity from the bulbs was recorded with low-resistance (3-7 M~)~) micropipettes (1.0 m m O.D., 0.5 mm I.D.) pulled to have long shanks and filled with 3.0 M NaCI. The reference electrode was a coil of tungsten wire placed under the skin of the neck. Tungsten wire was inserted into the recording micropipette and connected to a coaxial cable which led to the input stage of a Grass P-16 amplifier with negative capacitance. Single-ended AC (band-pass 1-100 Hz) and DC (DC-100 Hz) outputs were viewed on an oscilloscope. The band-pass settings were selected on the basis of response frequencies obtained in preliminary studies. The highest dominant frequency component of the bulb oscillations was 25 Hz (gopher tortoise), whereas the more usual values were between 13 Hz and 17 Hz. Permanent records were made either on a Sanborn recorder or by an oscilloscope camera. Differential, AC-coupled activity from the primary olfactory nerve or vomeronasal nerve4S, 49 was recorded in conjunction with the bulbar activity in some experiments. The preamplifier output was conducted to a short time averaging circuit v and displayed on a Sanborn recorder. The use of time averaged multi-unit activity allows spike bursts to be treated as events and has been used in studies on the olfactory system 34,49, as well as the spinal cord. 42.
Stimulation procedures Odor stimulation was by 3 methods: (1) an air flow dilution olfactometer as described by Tucker 49, (2) manually operated syringes, and (3) polyethylene squeeze bottles. The olfactometer was employed to accurately specify the odor concentration and nasal flow rate and for long stimulus durations (I-30 sec). The gopher tortoise was primarily used for experiments with the olfactometer. The syringe delivery was used to provide brief ( 150-200 msec) and reproducible odor puffs. New, 10-ml glass syringes were specially cleaned and connected by Teflon tubing to a glass delivery tube placed directly in front of the naris to be stimulated. Various amounts of odorous air (1-4 ml) were drawn from the bottle head space over the liquid odorant and sent into the nasal cavity by depressing the plunger. Unintentional stimulation of the contralateral bulb was investigated and found not to be a problem. Technical grade geraniol in a 1:10 dilution with mineral oil was used with the syringe delivery system and will be called the standard stimulus. After an odorous stimulus, laboratory air filtered in a similar manner as by the olfactometer 49, was
64 passed through the delivery tube for 15--30 sec at a flow rate of lt) ml/sec. Another 30-40 sec was allowed betk)re presenting the next stimulus. Odor puffs were also delivered f'rom polyethylene squeeze bottles, containing a few milliliters of undiluted test odorant, were used to determine the responsiveness o(a preparation. A small amount of butyric acid was put on a pad in the wash bottle and allowed to evaporate to the point where it would not be overwhelming by subjective determination. With practice, reproducible odor pufl~ could be delivered as previously noted 49. Passage of the syringe delivered puff through the glass delivery tube was indicated by the output from a thermistor flowmeter. This flowmeter employed feedback circuitry which would supply current to the thermistor in the delivery tube in response to cooling by air-flow with a response time of less than 0.1 sec. Calibration of the electrical flowmeter was by a wet-test gasmeter and a stopwatch. The recorded output of the flowmeter indicated the shape, time course and flow rate of the stimulus puff (see Fig. 2). With the syringe delivery method, a 3-ml puff was about 150 msec in duration and had a maximal flow rate of 6 7.5 ml/sec. This flowmeter was also used in conjunction with the olfactometer.
Odorants Technical grade geraniol (Matheson, Coleman and Bell), used as the standard stimulus, was diluted with paraffin oil (Matheson, Coleman and Bell). In experiments with the olfactometer, technical grade geraniol, high purity geraniol (Givaudan Corp.) and amyl acetate (Matheson, Coleman and Bell) were compared. Subjectively both types of geraniol smelled similar; however, technical grade geraniol was somewhat stronger. Other test odorants were amyl acetate, isoamyl acetate and methyl salicylate (Matheson. Coleman and Bell) and butyric acid (Eastman Organic Chemicals).
Chromatographic analysis Instrumental analysis of the two geraniol samples was performed on a Varian Aerograph model 705 gas-liquid chromatograph with a 5-ft. stainless steel analytical column of 1/8 in. diameter. The column packing was 60/80 mesh Chromosorb W coated with 1090 Carbowax 20 M. Electrometer output was led to a Beckman pen recorder. Odorant samples in the gas phase were taken from the output of the olfactometer and deposited on the column by the method used by Pierson 39. Geraniot began to elute when the column reached 120 °C. The injector was at 155 ~C. Areas of odor peaks were measured with a planimeter: Areas obtained from olfactometer controls. wash air, were subtracted from the areas obtained from the odor samples. Areal analysis of the chromatograms showed that technical grade geraniol was only 50 % geraniol (the unidentified peaks may be citronellol or nerol 6) while the high purity geraniol was 97 ~,i pure (Fig. 1). Amyl acetate from this source (same lot no. and analyzed by these methods was previously shown not to have contaminants 3:~. The absence of odor peaks in the olfactometer control indicated the effectiveness of the procedures used to filter the wash air (Fig. 1).
65
I
GERANtOL, tech.
I
I
I
T
1
1
I ,,
GERANIOL,h.p,
m
] 9
m
i
CONTROL
] 15 MtNS.
Fig. 1. Chromatograms of 30-sec samples of technical grade geraniol (top trace), high purity geraniol (middle trace) and filtered air (bottom trace) were collected from the output of the olfactomete[ in the gas phase. The two types of geraniol were at saturation. Measurements of the areas under the peaks showed that technical grade geraniol was 50%0 pure whereas high purity geraniol was 9 7 ~ pure. The filtered air did not contain odor peaks. Gas chromatograph settings range 0.1, attenuator × 8.
66
I+
I
[
A
B
WAVE •
4 0 0 msec
~PLITOOE
400
Fig. 2. The odor evoked response was recorded using AC (upper trace) and DC (middle trace) coupling with the associated thermistor flowmeter analog record (lower trace). In this series of 4 consecutive responses (A and B) the recording micropipette was placed in the depths of the box turtle bulb. Each stimulation was a 3-ml puff of the standard stimulus which was indicated by the output of the thermistor flowmeter (C). The recorder was halted between stimulations while the wash air was applied. Waves I and II were measured as shown in B from an extension of the prestimulus base line, Wave llI was measured as in B and a factor for the envelope of wave Ill was the product of the dimensions shown in D. Amplitude indicators were 0.4 mV in each case. Also note the time base was expanded in B.
Data analysis The amplitudes of the wave components of the odor evoked response were measured in terms of potential as shown in Fig. 2. Amplitude of the early potential, wave 1, was measured at its maximum from an extension of the base line. Wave 1I amplitude was taken at the maximum of a curve drawn to fit the inferred time course with the prestimulus base line used as a reference level. The amplitudes of the oscillations, wave Ill, were referred to wave il, and the peak-to-peak values of complete cycles of the dominant frequency were combined to give an average value (Fig. 2B). Some of the records were retouched for photographic purposes. RESULTS
Components of the odor evoked response A micropipette driven into the deep layers of the olfactory bulb of either a box turtle or gopher tortoise recorded a characteristic series of slow potentials in response to a puff of odorous air directed into the nose. This series of slow potentials was referred to as the odor evoked response and when recorded from deep within the bulb appeared as in Fig, 2. There was an initial short duration positive event, wave I, with a latency of 150-200 msec, wave lI was also positive and of much longer duration on which was superimposed an oscillatory potential, wave lII (the induced waves of Adrian2). This series of potentials underwent predictable changes in amplitude and polarity which were dependent on the anatomical layering of the bulb s ] 0 . When the odor evoked response was recorded from surface regions of the bulb receiving olfac-
67 tory input, waves l and I 1 were negative; however, all 3 waves reached their greatest magnitudes in the deep layers of the bulb. The ongoing activity of the olfactory bulb in these two species of turtles was always of low amplitude, unsynchronized and was identical with the prestimulus baseline (Fig. 2).
Reproducibility of the slow waves Repeated applications of a single odor stimulus or stimulation with a number ot test odorants showed that the 3 wave components were elicited in a reproducible and consistent manner. Successive puffs of the standard stimulus showed that the wavetbrm, wave amplitudes and wave time courses were similar in each response as seen in Fig. 2A and B. The responses to 25 applications of the standard stimulus were recorded at a bulb location and gave the following values: wave I, X = 0.85 mV, S.E.M. -- 0.02 mV; wave II, X -= 0.62 mV, S.E.M. - 0.04 mV; envelope of wave Ill calculated as in Fig. 2D, X = 204.9 sq. ram, S.E.M. ~- 3.6 sq. mm. The wave components of the odor evoked response showed variations in amplitude and duration when test odorants of different chemical composition were used as stimuli (Fig. 3). Waves 1 in Fig. 3 were similar in duration, but varied in amplitude while waves II and III varied both in duration and amplitude. Butyric acid, which is least effective among these test odorants for eliciting the olfactory nerve response 4~,49 evoked a small response even with vigorous stimulation.
Duration of the odor evoked slow potentials Measurements of the half-amplitude duration of wave I (60 msec ± 5 msec) indicated no relationship to the method of stimulation, i.e., olfactometer, syringe, or squeeze bottle nor parameters of stimulation. The time course of wave l I depended on stimulation procedures. A short odor puff of less than 200 msec with the syringe delivery method produced a potential with a complete time course of 2-5 sec. Using the olfactometer, half-amplitude decay times of wave I1 were measured from the peak amplitude in the tortoise bulb for technical grade geraniol and amyl acetate, with stimulus duration at 5 sec and nasal flow rate at 10 ml/sec. This decay time for wave I I with geraniol increased with concentration; reaching 8 sec at 10 o.2,5 (of vapor saturation). These decay times for wave l! with amyl acetate were less than 5 sec at 10_3.0 (of vapor saturation) and more than 12 sec at 10 0.')5 (of vapor saturation). At a given recording locus, neither stimulus concentration nor the time of the stimulus application appeared to affect the duration of the oscillations, wave I Il. This observation was tested in detail with the olfactometer. The recording pipette was positioned in the depths of a box turtle bulb. Different combinations of stimulus concentration (amyl acetate at 10-1.°, 10 4.0 and 10 a.0 of saturation) and stimulus duration (1, 2, 4, 6 and 8 sec) were tested. The 58 responses were analyzed and the duration of the oscillations varied from 0.7 sec to 1.1 sec. Statistical tests for trends within groups were not significant nor was the two-way analysis of variance, P ,~- 0.1, Thus, the duration of wave II! under these conditions was not affected by stimulus concentration or duration. The duration of wave ill could be altered by modulation
68 STIMULATION
PUFF BOTTLE
herd :mff
frame Puff
t ~ - . ~ r
ACETATE
PC~
""""
f
i
~-
'r
T.YL SAL,CYL Vt
l
,
Fig. 3. The odor evoked response appeared consistently to test stimuli of different odorants. Odor puffs were delivered by squeeze bottles; puffs of various intensities were reproducible with practice. Wave II1 is seen more clearly in the AC (upper) trace of each pair which was recorded at higher gain. Records were from the depths of the box turtle bulb. of the nasal airflow rate alter the initial odor evoked response which produced a series of oscillations of long duration (3-5 sec), but of low frequencies (5-9 Hz).
Accessory olfactory bulb Although the vomeronasal nerve, the afferent input to the accessory bulb (AB), was routinely severed, it was desirable to determine whether the AB could contribute to the odor evoked response since the granule cell layer of the AB is contiguous with that of the olfactory bulb 9. By simultaneously recording from two micropipettes connected to different capacity coupled amplifiers, this possibility was tested. To insure activation of the AB, butyric acid was also used as a stimulus since the vomeronasal receptors are very responsive to this odorant 48.
69 With the olfactory nerve severed and the vomeronasal nerve left intact, the odor evoked response could not be induced in either the olfactory bulb or the AB with puffs of butyric acid. With the olfactory nerve intact, a response was evoked in both, but at the AB the response was considerably attenuated. In other experiments, the receptor spike discharge was recorded from a small twig of the vomeronasal nerve while AB activity was monitored with a recording pipette. Again, after severing the olfactory nerve no odor evoked response was found in the AB, even though the receptor axons conducted action potential responses. Thus, the odor evoked response of the olfactory bulb was considered not to be complicated by evoked potentials from the AB.
Comparison o[ the o!factory nerve response and the odor evoked response The odor evoked response of the bulb was recorded simultaneously with the averaged multi-unit activity of the olfactory receptor axons in 5 tortoises. The olfactometer was utilized to permit longer stimulus durations (2-30 sec) and so that the odorant concentration and nasal flowrate could be specified for comparisons to earlier studies48, 49. Unexpectedly, in the tortoise preparations in which olfactory nerve and bulb activity were recorded together, oscillations were constantly found as part of the olfactory nerve response at higher stimulus concentrations (Figs. 4 and 5). The use of a short time constant of integration allowed comparisons to be made although this maneuver decreased the phasic portion of the nerve response and typical records are seen in Figs. 4 and 5. Wave I of the odor evoked response coincided with the rising phase of the integrated olfactory nerve discharge while the bulb oscillations, wave Ill, occurred at the time of maximal amplitude in the olfactory nerve response as seen in Figs. 4 and 5. The olfactory nerve oscillations were observed only at higher stimulus concentrations which was in contrast to the bulb oscillations which occurred with the lowest stimulus concentrations used (Fig. 5). Also as demonstrated in Fig, 5, the olfactory nerve oscillations adapted rapidly to successive stimulations at a particular concentration.
I
I
2 seconds
Fig. 4. The odor evoked response of the bulb (A) was recorded simultaneously in the tortoise with the time averaged (0.02 sec) multiunit activity (B) from an olfactory nerve twig. The onset of the bulbar oscillations preceded the onset of the peripheral oscillations. Nasal flow rate was l0 ~.~ ml/ sec and amyl acetate concentration was 10 1.0 of saturation at 20 °C.
70
B
10-2.0 10-0.5
if!
B ,
10-2.5
Io-LO iO-3,O
N
-
J
~
'
~
"
-
~
5 Sec I,,
I
,' ',~,,~ ', ','~ ",~w '/, ,' w u k " ~ V A / ~ ~,~ t , ~ , w , ~ , v ~ , ~ , ~ , ¢ ~
10%5
'
CONTROL N
Fig. 5. The odor evoked response of the tortoise bulb (B) was recorded simultaneously with the olfactory nerve (N) discharge. The first stimulus of amyl acetate in each pair of records was 6 sec and succeeding stimulations were 2 sec, except at l0 0.5 (of saturation) where all stimulus durations were 2 sec in order to prevent damage to the olfactory receptors. The nose was not flushed with filtered air between stimulations of the same concentration, but was flushed for 60 sec between changes in concentrations. The olfactory nerve response (N) exhibited oscillations (arrows) at the first response for concentrations l0 °.L10 1.5. Amyl acetate concentration expressed in terms of saturation at 20 ~C and nasal flowrate was 10 ~.o ml/sec.
Successive stimuli delivered with short interstimulus intervals resulted in decreased amplitudes of the wave components of the odor evoked response. In contrast, these adaptive effects were not observed in the receptor axon discharge and repeated stimulations elicited similar amplitude responses (Fig. 5). As seen in Fig. 5, wave 1 adapted very rapidly and was usually present as part of only the first bulb response for successive stimulations at a particular concentration, even at lower stimulus concentrations. At higher stimulus concentrations, the amplitude of wave I11 showed a marked decrease in response to closely spaced stimulations (Fig. 5). After increasing the interstimulus interval several fold and applying a filtered air wash, waves I and Ill had large amplitudes for the first stimulus of the next concentration series. Adaptive effects were also tbund Ibr waves I and 111 in the box turtle when stimuli with short interstimulus intervals were delivered from the olfactometer or syringe delivery. Recovery depended, as in the tortoise, on increasing the interstimulus interval and the use of a filtered air wash. In these experiments wave I1 was not usually' recorded, but when observed it also exhibited adaptation effects. At high stimulus concentrations, the bulb oscillations reappeared as an 'off' response when short duration stimulations were used. When the stimulus duration was increased from 6 sec (as in Fig. 5) to as long as 30 sec, the receptor axon response maintained a tonic level of activity, but the odor evoked response was present only at
71 the beginning of the stimulation period. Geraniol was usually not effective in producing bulbar 'off' responses when the syringe or the olfactometer was employed for stimulus delivery.
Control stimulations Control stimulations with the olfactometer showed that the mechanical effects of nasal airflow or odor switching transients did not elicit the odor evoked response. Using the olfactometer filtered air was drawn through the nasal passageway in the same manner as for an odorous stimulus. At low nasal flow rates there was no detectable response in the olfactory bulb or nerve (Fig. 5, Control). With much higher nasal flowrates (10 ~.a ml/sec) low amplitudes of waves I and 11 were observed, but this was probably due to the high velocity wash air picking up adsorbed odorant from the odor delivery switches as has been previously found 49. Filtered air from the olfactometer was analyzed in this study and found not to contain odor peaks (Fig. I). Odor comparisons The amplitudes of the wave components of the odor evoked response were found to be a function of odorant species and concentration when amy[ acetate, and two types of geraniol, technical grade and high purity, were directly compared at several concentrations. Stimulation and recording procedures were standardized with tortoises whose olfactory bulbs were acutely transected at the coronal sulcus. The tip of the recording pipette was placed 300/tin below the level at which wave I became positive, as close to the midline of the bulb as possible and 1.5-2.0 mm posterior from the anterior pole. The odor was switched on for 5 sec, a 30-sec flush was conducted, another 3-5 rain were allowed to elapse, and the next stimulus-flush sequence was presented. Initially, a nasal flow rate of I ml/sec was used, but the response amplitudes to the two types of geraniol were almost unmeasurable at a concentration of 10 l.r, (ot" saturation). Increasing the nasal flow rate to 10 ml/sec overcame this difficulty. All 3 odorants were tested in each animal. The response amplitudes of the 3 wave components of the odor evoked response, plotted in Fig. 6, were shifted up the ordinate from the least effective stimulus, high purity geraniol, to technical grade geraniol and to amyl acetate the most effective stimulus. The two geraniols were particularly interesting since technical grade geraniol was found to be an odorant mixture by gas-liquid chromatography (Fig. 1). Although the initial amplitudes for the 3 waves were similar at 10 2.o (of saturation), the slopes were greater for the amplitudes of the 3 waves for technical grade geraniol. Apparently odorant differences are more effective than simply concentration difference (Fig. 6, compare A and B). Variations in amyl acetate concentration in these preparations produced a different set of curves for the wave amplitudes as shown in Fig. 6C. The amplitudes of waves I and II rose rapidly to maximums from relatively large values at 10 2.o (of saturation). In particular, wave Ii amplitude at 10- ~.0 (of saturation) for amyl acetate was approximately equal to the plateau for technical grade geraniol. Wave llI showed little amplitude variation for this concentration range of amyl acetate and had approximately the maximal value reached with technical grade geraniol (compare wave II1 in Fig. 6B and C).
72
GERANIOL, hKIh pur,fy
0 -20
-L5
40
-I0
-0
G~RANK~L, flc~P,k;ol
3O
20
i
~.0
-I.5
"I.0
"0.5
40
0
C
30
