- Email: [email protected]
Auditory response properties and directional sensitivity of cerebellar neurons of the echolocating bat, Eptesicus fuscus
Brain Res,'arch, 521~(1990) 123-129
123
Elsevier r BRES 24248
Short Communications
Auditory response properties and directional sensitivity of cerebellar neurons of the echolocating bat, Eptesicus fuscus Tsutomu Kamada* and Philip H.-S. Jen Division of Biological Sciences, University of Missouri-Columbia, Columbia, MO 65211 (U.S.A.) (Accepted 15 May 1990)
Key words: Cerebellar auditory neuron; Bat; Directional sensitivity
Auditory response properties and directional sensitivity of cerebellar neurons of Eptesicusfuscus were studied under free-field stimulation conditions. The best frequency (BF) and minimum threshold (MT) of a recorded neuron were first determined with a sound delivered in front of the bat. Discharge pattern and MT were studied with both BF stimuli and one-octave downward and upward sweep FM (frequency-modulated) stimuli. The directional sensitivity of cerebellar neurons was then studied by determining the variation of MT and response latency with BF and FM stimuli broadcast from each of 15 loudspeakers attached to a semicircular wooden track in front of the bat. All 85 cerebellar neurons recorded discharged phasically to acoustic stimuli. Only 20 were spontaneously active. Cerebellar neurons were generally more sensitive to FM stimuli than to pure tone pulses. Thus, they discharged more vigorously and had a lower MT to the former than the latter stimulus. Directional sensitivity of 47 neurons (BF = 23.4-81.1 kHz) was studied. All neurons varied their MTs with sound direction. Most neurons (n = 37, 79%) showed a lowest MT to a frontal sound. Directional sensitivity of cerebellar neurons appears to be sharper when determined with BF tone pulses than with FM stimuli. Thus the directional slope and the difference in MT between the best and worst angles of these neurons were larger when determined with the BF stimulus. Directional sensitivity of cerebellar neurons is not dependent upon stimulus frequency, unlike that of the inferior and cortical neurons of the same bat. Cerebellar neurons also varied their response latency with sound direction. Such a variation may provide the bat with another neural code for sound localization.
as a p a r t of the e x t r a p y r a m i d a l m o t o r system. Because of this, it has b e e n p r o p o s e d that the cerebellum m a y function as a type of c o m p u t e r that is responsible for the fine c o o r d i n a t i o n of muscle tone and m o v e m e n t 7'8'18'19. It is also assumed that the cerebellum integrates incoming i n f o r m a t i o n from various neural pathways and sends its o u t p u t to regulate m o t o r responses to different sensory stimuli. F o r e x a m p l e , m a n y studies have d e m o n s t r a t e d that the cerebellar n e u r o n s r e s p o n d to acoustic stimuli and m a y play an i m p o r t a n t role in modifying auditory responses 33'44'45 vocalization 29'3° and in orienting an animal t o w a r d a sound source 1-4'i7"46.
sensitivity of cerebellar n e u r o n s u n d e r the free field stimulation conditions by examining the relation of the response threshold and latency to a sound b r o a d c a s t from different azimuthal directions. W e r e p o r t here that cerebellar neurons are generally m o r e sensitive to oneoctave d o w n w a r d sweep F M ( f r e q u e n c y - m o d u l a t e d ) stimuli than to pure tone pulses. H o w e v e r , their freq u e n c y - i n d e p e n d e n t directional sensitivity a p p e a r s to b e sharper when d e t e r m i n e d with p u r e t o n e pulses than with F M stimuli. A total of 15 big b r o w n bats, Eptesicus fuscus ( b o d y weight 12.2-16.8 g) w e r e used for this study. T h e p r o c e d u r e s for surgery and r e c o r d i n g of neural responses were the same as in previous studies 25'26'43. Briefly, each
In o r d e r to u n d e r s t a n d how the cerebellum may utilize the a u d i t o r y directional information in orienting an animal t o w a r d the sound source, it is essential to study the directional sensitivity of cerebellar neurons. A l though two previous closed system studies in cats have shown that responses of cerebellar neurons are influenced by variation in binaural input 1'17, the auditory directional sensitivity of cerebellar neurons was only briefly e x a m i n e d in two free-field studies 2'46. Thus, using bats as a m o d e l system, we studied the directional
bat was anesthetized with N e m b u t a l (45-50 mg/kg b.w.) and the flat h e a d of a 1.8 cm nail was a t t a c h e d to t h e exposed dorsal surface of the skull with acrylic glue and dental cement. The animal was then tied onto a m e t a l plate and its h e a d fixed by locking the shank of the nail onto a metal rod with a set screw 42. M i c r o p i p e t t e glass electrodes (3 M KC1) w e r e used to r e c o r d neural activities extracellularly from the c e r e b e l l a r vermis and hemispheres inside a s o u n d - p r o o f r o o m ( t e m p e r a t u r e 35-38 °C). Acoustic stimuli (4 ms d u r a t i o n , 0.5 ms
T h e cerebellum is k n o w n to receive multisensory inputs 5'6'8'ii'18"i9"31'39 and at the same time is considered
* Present address: Department of Oral Physiology, Hokkaido University, Sapporo, Japan.
Correspondence: P.H.-S. Jen, Division of Biological Sciences, University of Missouri-Columbia, Columbia, MO 65211, U.S.A. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
124 of the neuron to both stimuli (generally obtained at 30-50 dB above the neuron's MT) was also determined for sounds delivered from each loudspeaker. A total of 85 cerebellar neurons (45 from cerebellar vermis, 40 from cerebellar hemispheres) responding to acoustic stimuli were isolated at depths between 20 and 1924 ~m. They all discharged either less than 3 impulses (i.e. phasic neurons) or a burst of impulses (i.e. phasic bursters) during the stimulus. Only 20 neurons were spontaneously active. The MT determined at the BF of each neuron ranged between 12 and 101 dB SPL. Similar to an earlier study 43, there was no appreciable correlation in the distribution of the BF, MT and recording depths of these neurons (Fig. 1A,B). Response latencies of these neurons determined with a BF stimulus delivered in front of the bat were between 1.4 and 21.4 ms, with most (71 neurons, 87%) being below 12.4 ms (Fig. 1C). A linear regression analysis showed a poor correlation between the latency and BF of these neurons (Fig. 1D). Frequency tuning curves measured for 31 neurons were all broad (Fig. 1E). Thus, most (25 neurons, 80%) Q10-dB
rise-decay times) were delivered at a rate of 2/s either a s pure tone pulses or one-octave downward or upward sweep FM stimuli with the center frequency sweeping across the best frequency (BF) of each recorded neuron. The bat's head was oriented toward an array of 15 loudspeakers (Polaroid) mounted on a semicircular wooden track. The distance between the bat and each loudspeaker was 55.7 cm creating a 1.6 ms acoustic delay. The 15 loudspeakers were carefully chosen so that all had a similar frequency response curve. They were calibrated with a Briiel and Kjaer 1/4 inch microphone placed at the bat's ear. Output was expressed in dB SPL referred to 0.0002 dyne/cm2 root mean square. For each cerebellar neuron encountered, the BF and minimum threshold (MT) for pure tone pulses and for both FM stimuli w e r e first determined with sounds delivered in front of the bat. The frequency tuning curve was plotted when necessary. The MT of the neuron to a BF stimulus and a one-octave downward sweep FM signal delivered from each loudspeaker was then sequentially determined. Similarly, the shortest response latency 5" 0.
~'~ 2000 N=83
•
c ~ . 1600 1200
A
•
• • °••o
"10
"~ Co) 800
00 •
o s 0
o4..,
r=0.0,
:.&:.,.., :-
•
B
.~
0 v
• •
120 N=84 mlOO
03
•
•
10 20 30 40 50 60 70 80 9 0 1 0 0
Best frequency(kHz)
O0
80 60 ~
oOooOo
e
~ r_=O.21
•
oe f D o . ~ ,
~
•
"E
0
C
40 50 60 70 80 90100
Best frequency(kHz) 3o N=82
D
°
" •
•
~ #
"S
lo
o
~,
.
.
0
.
.
.
2.5 5
.
.
.
V-I.
[-'].
0
.
•~ ¢. m
•
•
,
.
,
.
,
t" ool
.
,
.
,
.
,
.
o"•
,
.
,
.
,
.
Best frequency(kNz) 30 E
100
IN=31
=e)
25
80
~
20
60
> m
15
40
"13 6
O
d
F •
•
0
20
40
Frequency(kHz)
60
80100
r=0.44
•
lO ~
0
• • •
_c
10
*
10 20 30 40 50 80 70 80 9 0 1 0 0
Latency(ms)
rn "(3
•
lO ~ . " '• t ,
7.5 1 0 1 2 . 5 1 5 1 Z 5 2 0
120 " u3
O ~.~...._._. 000
10 2030
~
30 N=82
008
e
~
eeeoo ,i
,
i
,
i
.
i
• .
n
,
n
.
n
.
r
•
J •
10 20 30 40 50 60 70 80 9 0 1 0 0 Best frequency(kHz)
Fig. 1. The distribution of BFs of 83 cerebellar neurons of Eptesicus fuscus against their recording depth (gm) (A), minimum threshold (dB SPL), (B) latencies (ms) (D), and Qlo-dB values (F). The linear regression line and the correlation coefficient (r) for each distribution are respectively shown. The range of response latency and 4 representative tuning curves are shown in (C) and (E). The upper solid line in (E) is the frequency characteristic curve of the loudspeaker. Note: A total of 85 cerebellar neurons were sampled, recording depth of 2 neurons, MT of 1 neuron and latency of 3 neurons were carelessly neglected during experiment.
125 values (a value obtained by dividing the neuron's BF by the bandwidth at 10 dB above the neuron's MT) were below 15 (range: 1.6-24.0). There was a weak correlation between the Q~0-dB value and the BF of these neurons (Fig. 1F). CerebeUar auditory neurons were generally more sensitive to FM stimuli than to pure tone pulses. Fig. 2A,B show the post-stimulus time (PST) histograms of 2 representative cerebellar neurons to 3 different stimuli. Although the response patterns of these 2 neurons did not vary with the type of stimulus used, the timing of their discharged impulses always fluctuated from stimulus to stimulus so that the duration of their PST histograms
was often longer than 10 ms when obtained with stimulus at high intensity. They also responded more vigorously to FM stimuli than to pure tone pulses. A comparison of MT of cerebellar neurons to different stimuli shows that cerebellar neurons are most sensitive to 4 ms downward sweep FM stimuli (Fig. 2C,D). A 4 ms upward sweep FM stimulus is also more effective than a 4 ms pure tone (BF) pulse (Fig. 2E). However, strongly FM-selective or specialized neurons such as those found in the inferior colliculus 4°'41 were not recorded. The mean MT (dB SPL) of cerebellar neurons is 50.3 + 19.9 (n = 84) for BF stimuli; 46.8 + 18.5 (n = 63) for the upward sweep FM
lOO--
A
C dB SPL
FM down
106
:.:.:"
F M up
P u r e tone
.:~-i:-i.~....
BF kHz
•~
,,,.,,~.i.,
.,~~.~ .
•
i~ .
.
.
60 "o
~.'
, .,,"
80
39.30 96
N=65
/"
t~,
oJ
•
40 20
r-
I-
//" i
0
•
20
i
i
40
60
.
i
80
100
Threshold for FMI stimuli
6 56
~~s." ..
~
.:~.......
'---
._ 100 m
' 'lOms Samp 50 B.W. lOOps
E
~,~
80
LI.
60
D
N=63
•e"
•
•.•"
/// •
~ ,t D /
i
i
i
40
60
80
40-
B
FM
down
F M up
Pure tone if)
20-
e*
r--
101
........
44.78
i
,
20 91
.m
lOO
E
N=63
,K~"•
E
8 1 m
',~ 80
71
D. IXI 60 0 "D "~ 4O ¢-
..-~...a~-,.
..- • :-~.
61 ¸
51
:~....
100
Threshold for FMI stimuli
::....-..
_
.....
, ,lOms Samp 50 B.W. 1 0 0 ps
/
/ ,/
•~e- 2o l.-
/
/
0 2 0,
,
4 0~
6~0
8 0'
100
Threshold for FM! stimuli Fig. 2. A,B: post-stimulus time (PST) histograms showing response patterns of 2 cerebellar neurons of Eptesicus fuscus to 4 ms pure tone, one-octave downward and upward sweep FM stimuli. Stimulus intensity (dB SPL) is shown to the left of each histogram. Samp.: sample size; B.W.: bin width. C-E: Comparison of minimum threshold (dB SPL) of cerebellar auditory neurons of Eptesicus fuscus to different acoustic stimuli• The dashed lines represent equal values for both stimuli. Points above the line indicate that the neuron had a higher MT to the stimulus plotted in the ordinate.
126 1 O0 ~ A
N=25
12
E
N=12
10
BF FM
201 -
~
0~
,
"
.
,
,
BE=79.7 kHz
N=12 12
~"
60
/ l l
10
BF
O) 40
FM
m 2o ~" --
0~ 100]
.
.
~ .
BF=76.6 kllz ,
.
N-
iil 80 C
-3
_
%
E
......
lf~ 10
BF
,
•
~ ' o
,
14
~
FM BF
•
BF:46.1 kHz • , . , .
,
N=8
H
10
~
FM
402O 0
' 70
,
, 30
• "0
,
Ipsi
BF=40.4 kHz - , . , 30 70
21
kHz were studied. According to the variation of M T with sound direction, cerebellar n e u r o n s fall into the following 4 types. Type 1 cerebellar n e u r o n s (n = 25, 53%) showed their lowest M T at 0 ° (in f r o n t of the bat) when d e t e r m i n e d with both B F pulses and F M stimuli (Fig 3A). The M T increased as the sound was delivered from lateral angles. In general, the M T reached its highest value when the sound was delivered from the most lateral angle. F u r t h e r m o r e , the M T of all these n e u r o n s was always lower to an F M stimulus than to a B F pulse. Thus,
~/ N=7
B q ~ " ~;~. /
~ 14t m _.1
1 00 ] D
60 80 1
~BF F[ k,Hz,
2 100 t B
Directional response characteristics of 47 cerebellar auditory n e u r o n s with B F ranging b e t w e e n 23.4 and 81.1
70
Contra Azimuth(deg)
,
, 30
• 0
,
Ipsi
BF=61.4 kHz I , • , • 70
30
Contra
Fig. 3. A-D: Variation of minimum threshold (dB SPL) of 4 representative cerebellar auditory neurons determined with pure tone (BF) and one-octave downward sweep FM stimuli when the sound was delivered from 9 different azimuthal angles (see text for details). E-H: Variation of response latency (ms) of 4 representative cerebellar neurons determined with pure tone (BF) and one octave downward sweep FM stimuli when the sound was delivered from 9 different azimuthal angles (see text for details).
stimuli and 41.2 + 18.2 (n = 65) for the downward sweep F M stimuli.
the F M threshold curve is always lower than the B F threshold curve (Fig. 3A). N e u r o n s of type 2 (n = 12, 26%) showed a similar directional characteristic to that of type one n e u r o n s (Fig. 3B). However, these n e u r o n s were not always more sensitive to the F M stimulus than to the B F pulse when d e t e r m i n e d at different angles. Thus, the FM threshold curve intersects the B F threshold curve at certain azimuthal angles. Type 3 cerebellar n e u r o n s (n = 3, 6%) varied their MTs only when the sound was delivered from the azimuthal angles contralateral to the recording site (Fig. 3C). The M T of these 3 n e u r o n s r e m a i n e d u n c h a n g e d when the sound was delivered from the ipsilateral angles. Directional characteristics of type 4 cerebellar n e u r o n s (n = 7, 15%) showed a very m i n o r variation in their MTs to sound direction so that both B F and F M threshold curves are extremely flat (Fig. 3D). Nevertheless, the lowest M T was always obtained when the sound was delivered at 10° or 30 ° lateral. The difference in M T and directional slope between the angles of lowest M T (the best angle) and highest M T
TABLE I
The difference in minimum threshold, response latency and directional slope between the best and worst angles of cerebellar auditory neurons Minimum threshold Type
Latency
BF Difference (dB)
1
range mean
2
range mean
3
range mean
4
range mean
All
range mean
FM Slope (dB/deg)
n = 25 11-62 0.2-0.9 38.1___12.6 0.5+0.2 n=12 20-50 0.3-0.7 32.0+9.2 0.5+0.1 n=3 27-36 0.3-0.5 30.0_+5.2 0.4+0.1 n=7 12-42 0.1-0.5 22.7_+12.8 0.3_+0.2 n = 47 11-62 0.12-0.89 33.7_+12.5 0.50_+0.20
Difference (dB)
Type BF Slope (dB/deg)
n = 21 0.1-0.7 0.4+0.1 n=12 16-46 0.2-0.7 32.3_+1.0 0.4+0.1 n=3 9-40 0.1-0.6 23.7_+15.6 0.3+0.3 n=7 7-32 0.1-0.3 15.0_+8.6 0.2+0.1 n = 43 7-48 0.1-0.7 25.9_+11.1 0.4+0.2 9-48 26.0+9.3
Difference (ms) A
2-5 3.5+1.2
B
0-7 2.4_+1.8
C
2-5 4.0+1.4
D
0-1 0.4_+0.5
All
0-7 2.5_+1.8
FM Slope (ms/deg) n = 12 0.03-1.0 0.10+0.20 n=22 0.00-0.01 0.06+0.03 n=4 0.00-0.07 0.04_+0.03 n=8 0.00-0.05 0.02+0.04 n = 46 0.00-1.0 0.07+0.1
Difference (ms)
Slope (ms/deg)
n = 10 0.5-6 0.03-1.2 3.5_+2.0 0.20_+0.30 n =22 0.0-9 0.00-0.20 2.4___1.8 0.07+0.05 n=4 3.0-5 0.03-0.05 4.0_+0.8 0.04_+0.01 n=8 0.0-1 0.00-0.03 0.2_+0.4 0.01_+0.01 n = 44 0.0-9 0.00-1.16 2.4+2.0 0.08_+0.18
127 A
1"2]A"
"O
0.8
==1
CBL
.
N=47
o •
•
#0•
0.6
~
• •• •
}
• ••• •
•
r.0.25
0.4 0.2
•
•
,'." ; 0
20
....
40
60
80
0.8 0.6
•
O
0.4
,
~
0.2
100
"
r=0.2
6,
20 30 40 50 60 70 80 90 100 FM center frequency(kHz)
2.( IC
~"
•
"e
o.ol.','. ~.',. ,\ ......
Best frequency(kHz) °'s t O
0•0••
N=31 I
D
sc
N=58
1.E
"0 0.6 ~.~ 0.4
• e
_o 0.2
.
0.(
10
15
20
25
30
35
]E
1.2 "~1,01
CBL •
0.6
r=018
•
N=46
0.2
r=0.11
0,0 40
,
.
40
,
.
60
,
.
80
100
~.2~ F ~
0.8
~
0.4
"
CBL
N=43
0.6 •
20
.
Best f r e q u e n c y ( k H z )
•
~-0.4
,
20
40
Best frequency(kHz)
0
e• "l
O 0.~
60
80
Best f r e q u e n c y ( k H z )
100
•
o
0.2
•
o.o
~
•
'
r=0.07
~
20 30 40 50 60 70 80 90 100 FM center frequency(kHz)
Fig. 4. Six scatter plots showing the distribution of directional slope in MT variation (dB/deg) (A-D) and in latency variation (ms/deg) (E,F) of single neurons from 4 different brain centers with sound direction against the BF (A,C,D,E) or FM center frequency (B,F). The linear regression line and correlation coefficient for each distribution are respectively shown. CBL, cerebellum; IC, inferior colliculus; SC, superior colliculus.
(the worst angle) for each type of cerebellar neuron is shown in Table I. While the MT difference between the best and worst angles determined with BF and FM stimuli is comparable in type 2 cerebellar neurons, it is consistently larger for BF pulses than for FM stimulus for type 1, 3 and 4 cerebellar neurons. Thus, when the directional sensitivity of all cerebellar neurons is considered, it is clear that the MT difference and directional slope are larger when determined with BF pulses than with the FM stimuli (Table I, bottom row). The correlation between the directional sensitivity and the stimulus frequency was examined by plotting the directional slope (dB/deg) of cerebellar neurons against either the BF (Fig. 4A) or the center frequency of each FM stimulus (Fig. 4B). The fact that a linear regression analysis on both distributions reveals a low correlation coefficient (r = 0.25 and 0.2 respectively) suggests that the directional sensitivity of cerebellar neurons is relatively independent of the stimulus frequency. This finding is similar to data obtained from superior collicular neurons of the same bat (Fig. 4D). However, consistent with previous findings az-a4' 26,32,35,37 the directional slope of inferior collicular neurons is significantly dependent upon the stimulus frequency (Fig. 4C).
Cerebellar neurons also varied their response latencies when the sound was delivered from different azimuthal angles. They can also be generally classified into 4 types according to their latency variation. Type A (n = 12, 26%) neurons generally had a shorter latency when determined with FM signals than with BF pulses. Thus, the FM latency curve is always lower than the corresponding BF latency curve. Nevertheless, both curves show a dip at 0 ° and rise on both limbs when the sound is delivered at lateral angles (Fig. 3E). In other words, these neurons had their shortest latency when the sound was delivered from the front. Cerebellar neurons o f type B (n = 22, 48%) changed their latency unpredictably with sound direction when determined with both stimuli. Thus, the BF latency curve does not vary in parallel with the FM latency curve resulting in intersection of the two curves at certain azimuthal angles (Fig. 3F). Furthermore, while the latency of some neurons of this type varied with sound direction when determined with BF pulses, it remained unchanged or fluctuated only slightly with sound direction when determined with FM stimuli or vice versa. Cerebellar neurons of type C (n = 4, 9%) had their longest latency for both sound stimuli delivered at 70° ipsilateral. They systematically shortened their latencies when the sound was broadcast sequentially from ipsilateral azimuthal angles to contralateral ones (Fig. 3G). Response latency of type D cerebellar neurons (n = 8, 17%) either changed very little or remained the same with sound direction (Fig. 3H). Thus they have very poor directional characteristics in terms of latency variation. The difference in latency and directional slope between the angles of shortest (the best angle) and longest (the worst angle) latency for each type of cerebellar neuron is shown in Table I. It is clear that cerebellar neurons do not appear to vary the response latency to a greater extent for a particular acoustic stimulus, different from that of MT variation. However, similar to the frequency-independent MT variation with sound direction, a linear regression analysis also reveals no appreciable correlation between the sharpness of the directional sensitivity in latency variation and the stimulus frequency (Fig. 4B,D). Our present study of response properties of cerebellar neurons confirms an early observation on the same bat 28 that the phasically discharged cerebellar neurons are generally broadly tuned (Fig. 1E) and are more sensitive to FM stimuli than to pure tone pulses (Fig. 2). As discussed previously24"2s, such a finding suggests that the cerebellum is not involved in further fine frequency analysis but rather in integrating and utilizing processed auditory information for motor orientation. The majority of cerebellar neurons studied show a lowest MT to both stimuli delivered from the front (n = 37, 79%, Fig. 3A,B)
128 or from 10 ° to 30 ° lateral (n = 7, 15%, Fig. 3D). This finding is similar to previous studies in bats 43 and in cats 2 suggesting that the cerebellum is able to orient an animal's head toward a localized sound within its frontal gaze. We noted that cerebellar neurons have a sharper directional sensitivity to a pure tone pulse than to an FM stimulus. Thus the difference in MT and directional slope between the best and worst angles is larger when determined with the pure tone pulse (Table I). We believe that such a difference is perhaps due to the fact that the directional sensitivity of a neuron to a pure tone pulse is primarily determined by the directional characteristics of a single frequency but its directional sensitivity to an FM stimulus is the combination of directional characteristics of each contained frequency component. Thus, the directional sensitivity to an FM stimulus may become less sharp as a result of integration or averaging of directional characteristics of individual frequency components. Nevertheless, since the BF of each neuron was always set as the center frequency of the FM stimulus, the most sensitive angle of each neuron was always the same when determined with both stimuli (Fig. 3 A - D ) . A similar finding for inferior collicular neurons of the same bat has been reported previously 26. We observed that, excluding 8 neurons (e.g. Fig. 3H), all cerebellar neurons generally vary their response latencies significantly with sound direction (Fig. 3 E - G ) thus providing another possible neural code for sound localization. Such an observation has been reported in the auditory neurons of cats 9, bats 2°'21, insects 34 and frogs 1°. When a sound is broadcast from different directions, the interaural intensity changes as a result of directional characteristics of the animal's head and pinna. Such a change may be reflected by the variation of the number of impulses, MT and response latency of the recorded neuron with sound direction. Since all cerebellar neurons discharged phasically (no more than 7 impulses per stimulus), the variation of their number of impulses with sound direction per stimulus may be quite limited. Thus, auditory physiologists often study the variation of the number of impulses of a neuron with sound direction per 30 or 50 repeated stimuli. On the other hand, our observation on the variation of the response latency of these neurons with sound direction supports an earlier suggestion 9'1°'34 that the response
latency may also be a useful code for sound localization. A similar latency-shift mechanism in coding an object in the electric field by the electroreceptors of a weakly electric fish, mormyrids, has been proposed earlier 36. Previous studies of the inferior colliculus 12-14'26"32'35'37 (e.g. Fig. 4C) and the auditory cortex 25 have shown that the directional sensitivity of auditory neurons is frequency-dependent so that it sharpens with increasing stimulus frequency. We 35 have suggested that the frequency-dependent directional sensitivity is probably the result of directional filtering of sound by the pinna and head 22. However, our present study shows that the directional sensitivity of cerebeUar neurons is frequencyindependent (Fig. 4). Since the BF of our sampled neurons ranged from 23.4 to 81.1 kHz covering almost the entire range of the frequency spectrum of the bat's echolocation signals 23'3s, it is unlikely that our observation is simply due to a sampling bias of a particular population of cerebellar neurons. In one of our early unpublished observations, we noted that the directional sensitivity of superior collicular neurons of the same bat was also frequency-independent (Fig. 4D). Since both the superior colliculus and the cerebellum are involved ill auditory orientation 15'16'46, they apparently integrate and utilize the directional sensitivity of their presynaptic auditory neurons in order to orient an animal's head and pinnae toward a sound source. For sound localization, neurons with frequency-independent directional sensitivity are certainly more versatile and effective in providing information about the direction of the sound source. It appears that the frequency-dependent directional sensitivity is somehow transformed into the frequency-independent one after sensorimotor integration. The fact that both the superior colliculus and the cerebellum lack tonotopic organization and their neurons are more sensitive to FM stimuli 27'28 appears also to be the result of such integration process. Future studies of the directional sensitivity of the cerebellar neurons in terms of variation in the number of impulses with sound direction are necessary to reconfirm our present finding. This study was supported by a grant from the National Institute of Health (NS 20527) and a grant from the Research Council from the University of Missouri (Bio. Med 017). We thank Mr. D. Pinheiro for his fruitful discussion and an anonymous reviewer's comments on an earlier version of this paper and Ms. K. Cook for secretarial assistance.
1 Aitkin, L.M. and Boyd, J., Responses of single units in cerebellar vermis of the cat to monaural and binaural stimuli, J. Neurophysiol., 38 (1975) 418-429. 2 Aitkin, L.M. and Rawson, J.A., Frontal sound source location is represented in the cat cerebellum, Brain Research, 265 (1983) 317-321. 3 Altman, J.A., Bechterev, N.N., Radinova, E.A, Shimigidina,
G~M. and Syka, J., Electrical responses of the auditory area of the cerebellar cortex to acoustic stimulation, Exp. Brain Research., 26 (1976) 285-298. 4 Bechterev, N.N., Syka, J. and Altman, J.A., Responses of cerebellar units to stimuli simulating sound source movement and visual moving stimuli, Experientia, 31 (1975) 819-821. 5 Bloedel, J.R., Cerebellar afferent systems: a review, Prog.
129
Neurobiol., 2 (1975) 1-68. 6 Bloedei, J.R., Dichgans, J. and Precht, W., Cerebellar Functions, Springer, Berlin, 1985. 7 Eccles, J.C., The cerebellum as a computer: patterns in space and time, J. Physiol. (Lond.), 229 (1973) 1-32. 8 Eccles, J.C., Ito, M. and Szentagothai, J., The Cerebellum as a Neuronal Machine, Springer, New York, 1967. 9 Erulkar, S.D., The responses of single units of the inferior colliculus of the cat to acoustic stimulation, Proc. R. Soc. Lond., B, 150 (1959) 336-355. 10 Feng, A.S., Directional response characteristics of single neurons in the torus semicircularis of the leopard frog, (Rana pipiens), J. Comp. Physiol., 144 (1981) 419-428. 11 Fadiga, E. and Pupilli, G.C., Teleceptive components of the cerebellar function, Physiol. Rev., 44 (1964) 432-486. 12 Fuzessery, Z.M. and Pollak, G.D., Determinants of sound location selectivity in bat inferior colliculus: a combined dichotic and free-field stimulation study, J. Neurophysiol., 54 (1985) 757-781. 13 Grinnell, A.D., The neurophysiology of audition in bats: directional localization and binaural interaction, J. Physiol. (Lond.), 167 (1963) 97-113. 14 Grinnell, A.D.,and Grinnell, V.S., Neural correlates of vertical localization of echolocating bats, J. Physiol. (Lond.), 181 (1965) 830-851. 15 Harris, L.R., The superior colliculus and movements of the head and eyes in cats, J. Physiol. (Lond.), 300 (1980) 367-391. t6 Henkel, C.K. and Edwards, St. B., The superior colliculus control of pinna movements in the cat: possible anatomical connections J. Comp. Neurol., 182 (1978) 763-776. 17 Highstein, S. and Coleman, P.D., Responses of the cerebellar vermis to binaural auditory stimulation, Brain Research, 10 (1968) 470-473. 18 Ito, M., Is the cerebellum really a computer? Trends Neurosci., (1979) 122-126. 19 Ito, M., The Cerebellum and Neural Control, Raven, New York, 1984. 20 Jen, P.H.-S., Coding of Directional Information by Single Neurons in the Superior Olivary Complex of Echolocating Bats, Ph.D. thesis, Washington University-St. Louis, 1974. 21 Jen, P.H.-S., Coding of directional information by single neurons in the S-segment of the FM bat, Myotis lucifugus, J. Exp. Biol., 87 (1980) 203-216. 22 Jen, P.H.-S. and Chen, D.M., Directionality of sound pressure transformation at the pinna of echolocating bats, Hearing Res., 34 (1988) 101-118. 23 Jen, P.H.-S. and Kamada, T., Analysis of orientation signals emitted by the CF-FM bat Pteronotusp. parnellii and the FM bat Eptesicus fuscus during avoidance of moving and stationary obstacles, J. Comp. Physiol., 148 (1982) 389-398. 24 Jen, P.H.-S. and Schlegel, P.A., Neurons in the cerebellum of echolocating bats respond to acoustic signals, Brain Research, 196 (1980) 502-507. 25 Jen, P.H.-S., Sun, X.D. and Lin, P.J.J., Frequency and space representation in the primary auditory cortex of the frequency modulating bat, Eptesicus fuscus, J. Comp. Physiol., 165 (1989) 1-14. 26 Jen, P.H.-S., Sun, X.D., Chen, D.M. and Teng, H.B., Auditory space representation in the inferior cofliculus of the FM bat,
Eptesicus fuscus, Brain Research, 419 (1987) 7-18. 27 Jen, P.H.-S., Sun, X.D., Kamada, T., Zhang, S.Q. and Shimozawa, T., Auditory response properties and spatial response areas of superior coUicular neurons of the FM bat, Eptesicus fuscus, J. Comp. Physiol., 154 (1984) 407-413. 28 Jen, P.H.-S., Kamada, T., Sun, X.D., Schlegel, P., Vater, M., Harnischfeger, G. and Riibsamen, R., Responses of cerebellar neurons to acoustic stimuli in certain FM and CF-FM bats, Chinese J. Physiol., 27 (1984) 1-26. 29 Kirzinger, A., Cerebellar lesion effects on vocalization of the squirrel monkey, Behav. Brain Res., 16 (1985) 177-181. 30 Larson, C.R., Sutton, D. and Lindeman, R.C., Cerebellar regulation of phonation in rhesus monkey (Macaca mulatta), Exp. Brain Res., 33 (1978) 1-18. 31 MacKay, W.A. and Murphy, J.T., Cerebellar modulation of reflex gain, Prog. Neurobiol., 13 (1979) 361-417. 32 Makous, J.C. and O'Neill, W.E., Directional sensitivity of the auditory midbrain in the mustached bat to free-field tones, Hearing Res., 24 (1986) 73-88. 33 Mitra, J. and Snider, R.S., Cerebellar modification of unitary discharges in auditory system, Exp. Neurol., 23 (1969) 341-352. 34 M6rchen, A., Rheinlaender, J. and Schwarzkopf, J., Latency shift in insect auditory nerve fibers. A neural time cue of sound direction, Naturwissenschaften, 54 (1978) 656. 35 Poon, P.W.E, Sun, X.D., Kamada, T. and Jen, P.H.-S., Frequency and space representation in the inferior colliculus of the FM bat, Eptesicus fuscus, Exp. Brain Res., 79 (1990) 83-91. 36 Szabo, T, and Hagiwara, S., A latency-change mechanism involved in sensory coding of electric fish (Mormyrids), Physiol. Behav., 21 (1967) 331-335. 37 Shimozawa, T., Suga, N., Hendler, P. and Schuetze, S., Directional sensitivity of echolocation systems in bats producing frequency-modulated signals, J. Exp. Biol., 60 (1974) 53-59. 38 Simmons, J.A., Fenton, M.B. and O'Farrell, M.J., Echolocation and pursuit of prey by bats, Science, 203 (1979) 16-21. 39 Snider, R.S. and Stowell, A., Receiving areas of the tactile, auditory and visual systems in the cerebellum, J. Neurophysiol., 7 (1944) 331-358. 40 Suga, N., Analysis of frequency-modulated and complex sounds by single auditory neurons of bats, J. Physiol. (Lond.), 198 (1968) 51-80. 41 Suga, N., Classification of inferior collicular neurons of bats in terms of responses to pure tones, FM sounds, and noise bursts, J. Physiol. (Lond.), 200 (1969) 555-574. 42 Suga, N. and Schlegel, P.A., Neural attenuation of responses to emitted sounds in echolocating bats, Science, 177 (1972) 82-84. 43 Sun, X.D., Jen, P.H.-S. and Zhang, W.P., Auditory spatial response areas of single neurons and space representation in the cerebellum of echolocating bats, Brain Research, 414 (1987) 314-322. 44 Teramoto, S. and Snider, R.S., Modification of auditory responses by cerebellar stimulation, Exp. Neurol., 16 (1966) 191-200. 45 Velluti, R . and Crispino, L., Cerebellar actions on cochlear microphonics and on auditory nerve action potential, Brain Res. Bull., 4 (1979) 621-624. 46 Wolfe, J.W., Responses of the cerebellar auditory area to pure tone stimuli, Exp. Neurol., 563 (1972) 295-309.
123
Elsevier r BRES 24248
Short Communications
Auditory response properties and directional sensitivity of cerebellar neurons of the echolocating bat, Eptesicus fuscus Tsutomu Kamada* and Philip H.-S. Jen Division of Biological Sciences, University of Missouri-Columbia, Columbia, MO 65211 (U.S.A.) (Accepted 15 May 1990)
Key words: Cerebellar auditory neuron; Bat; Directional sensitivity
Auditory response properties and directional sensitivity of cerebellar neurons of Eptesicusfuscus were studied under free-field stimulation conditions. The best frequency (BF) and minimum threshold (MT) of a recorded neuron were first determined with a sound delivered in front of the bat. Discharge pattern and MT were studied with both BF stimuli and one-octave downward and upward sweep FM (frequency-modulated) stimuli. The directional sensitivity of cerebellar neurons was then studied by determining the variation of MT and response latency with BF and FM stimuli broadcast from each of 15 loudspeakers attached to a semicircular wooden track in front of the bat. All 85 cerebellar neurons recorded discharged phasically to acoustic stimuli. Only 20 were spontaneously active. Cerebellar neurons were generally more sensitive to FM stimuli than to pure tone pulses. Thus, they discharged more vigorously and had a lower MT to the former than the latter stimulus. Directional sensitivity of 47 neurons (BF = 23.4-81.1 kHz) was studied. All neurons varied their MTs with sound direction. Most neurons (n = 37, 79%) showed a lowest MT to a frontal sound. Directional sensitivity of cerebellar neurons appears to be sharper when determined with BF tone pulses than with FM stimuli. Thus the directional slope and the difference in MT between the best and worst angles of these neurons were larger when determined with the BF stimulus. Directional sensitivity of cerebellar neurons is not dependent upon stimulus frequency, unlike that of the inferior and cortical neurons of the same bat. Cerebellar neurons also varied their response latency with sound direction. Such a variation may provide the bat with another neural code for sound localization.
as a p a r t of the e x t r a p y r a m i d a l m o t o r system. Because of this, it has b e e n p r o p o s e d that the cerebellum m a y function as a type of c o m p u t e r that is responsible for the fine c o o r d i n a t i o n of muscle tone and m o v e m e n t 7'8'18'19. It is also assumed that the cerebellum integrates incoming i n f o r m a t i o n from various neural pathways and sends its o u t p u t to regulate m o t o r responses to different sensory stimuli. F o r e x a m p l e , m a n y studies have d e m o n s t r a t e d that the cerebellar n e u r o n s r e s p o n d to acoustic stimuli and m a y play an i m p o r t a n t role in modifying auditory responses 33'44'45 vocalization 29'3° and in orienting an animal t o w a r d a sound source 1-4'i7"46.
sensitivity of cerebellar n e u r o n s u n d e r the free field stimulation conditions by examining the relation of the response threshold and latency to a sound b r o a d c a s t from different azimuthal directions. W e r e p o r t here that cerebellar neurons are generally m o r e sensitive to oneoctave d o w n w a r d sweep F M ( f r e q u e n c y - m o d u l a t e d ) stimuli than to pure tone pulses. H o w e v e r , their freq u e n c y - i n d e p e n d e n t directional sensitivity a p p e a r s to b e sharper when d e t e r m i n e d with p u r e t o n e pulses than with F M stimuli. A total of 15 big b r o w n bats, Eptesicus fuscus ( b o d y weight 12.2-16.8 g) w e r e used for this study. T h e p r o c e d u r e s for surgery and r e c o r d i n g of neural responses were the same as in previous studies 25'26'43. Briefly, each
In o r d e r to u n d e r s t a n d how the cerebellum may utilize the a u d i t o r y directional information in orienting an animal t o w a r d the sound source, it is essential to study the directional sensitivity of cerebellar neurons. A l though two previous closed system studies in cats have shown that responses of cerebellar neurons are influenced by variation in binaural input 1'17, the auditory directional sensitivity of cerebellar neurons was only briefly e x a m i n e d in two free-field studies 2'46. Thus, using bats as a m o d e l system, we studied the directional
bat was anesthetized with N e m b u t a l (45-50 mg/kg b.w.) and the flat h e a d of a 1.8 cm nail was a t t a c h e d to t h e exposed dorsal surface of the skull with acrylic glue and dental cement. The animal was then tied onto a m e t a l plate and its h e a d fixed by locking the shank of the nail onto a metal rod with a set screw 42. M i c r o p i p e t t e glass electrodes (3 M KC1) w e r e used to r e c o r d neural activities extracellularly from the c e r e b e l l a r vermis and hemispheres inside a s o u n d - p r o o f r o o m ( t e m p e r a t u r e 35-38 °C). Acoustic stimuli (4 ms d u r a t i o n , 0.5 ms
T h e cerebellum is k n o w n to receive multisensory inputs 5'6'8'ii'18"i9"31'39 and at the same time is considered
* Present address: Department of Oral Physiology, Hokkaido University, Sapporo, Japan.
Correspondence: P.H.-S. Jen, Division of Biological Sciences, University of Missouri-Columbia, Columbia, MO 65211, U.S.A. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
124 of the neuron to both stimuli (generally obtained at 30-50 dB above the neuron's MT) was also determined for sounds delivered from each loudspeaker. A total of 85 cerebellar neurons (45 from cerebellar vermis, 40 from cerebellar hemispheres) responding to acoustic stimuli were isolated at depths between 20 and 1924 ~m. They all discharged either less than 3 impulses (i.e. phasic neurons) or a burst of impulses (i.e. phasic bursters) during the stimulus. Only 20 neurons were spontaneously active. The MT determined at the BF of each neuron ranged between 12 and 101 dB SPL. Similar to an earlier study 43, there was no appreciable correlation in the distribution of the BF, MT and recording depths of these neurons (Fig. 1A,B). Response latencies of these neurons determined with a BF stimulus delivered in front of the bat were between 1.4 and 21.4 ms, with most (71 neurons, 87%) being below 12.4 ms (Fig. 1C). A linear regression analysis showed a poor correlation between the latency and BF of these neurons (Fig. 1D). Frequency tuning curves measured for 31 neurons were all broad (Fig. 1E). Thus, most (25 neurons, 80%) Q10-dB
rise-decay times) were delivered at a rate of 2/s either a s pure tone pulses or one-octave downward or upward sweep FM stimuli with the center frequency sweeping across the best frequency (BF) of each recorded neuron. The bat's head was oriented toward an array of 15 loudspeakers (Polaroid) mounted on a semicircular wooden track. The distance between the bat and each loudspeaker was 55.7 cm creating a 1.6 ms acoustic delay. The 15 loudspeakers were carefully chosen so that all had a similar frequency response curve. They were calibrated with a Briiel and Kjaer 1/4 inch microphone placed at the bat's ear. Output was expressed in dB SPL referred to 0.0002 dyne/cm2 root mean square. For each cerebellar neuron encountered, the BF and minimum threshold (MT) for pure tone pulses and for both FM stimuli w e r e first determined with sounds delivered in front of the bat. The frequency tuning curve was plotted when necessary. The MT of the neuron to a BF stimulus and a one-octave downward sweep FM signal delivered from each loudspeaker was then sequentially determined. Similarly, the shortest response latency 5" 0.
~'~ 2000 N=83
•
c ~ . 1600 1200
A
•
• • °••o
"10
"~ Co) 800
00 •
o s 0
o4..,
r=0.0,
:.&:.,.., :-
•
B
.~
0 v
• •
120 N=84 mlOO
03
•
•
10 20 30 40 50 60 70 80 9 0 1 0 0
Best frequency(kHz)
O0
80 60 ~
oOooOo
e
~ r_=O.21
•
oe f D o . ~ ,
~
•
"E
0
C
40 50 60 70 80 90100
Best frequency(kHz) 3o N=82
D
°
" •
•
~ #
"S
lo
o
~,
.
.
0
.
.
.
2.5 5
.
.
.
V-I.
[-'].
0
.
•~ ¢. m
•
•
,
.
,
.
,
t" ool
.
,
.
,
.
,
.
o"•
,
.
,
.
,
.
Best frequency(kNz) 30 E
100
IN=31
=e)
25
80
~
20
60
> m
15
40
"13 6
O
d
F •
•
0
20
40
Frequency(kHz)
60
80100
r=0.44
•
lO ~
0
• • •
_c
10
*
10 20 30 40 50 80 70 80 9 0 1 0 0
Latency(ms)
rn "(3
•
lO ~ . " '• t ,
7.5 1 0 1 2 . 5 1 5 1 Z 5 2 0
120 " u3
O ~.~...._._. 000
10 2030
~
30 N=82
008
e
~
eeeoo ,i
,
i
,
i
.
i
• .
n
,
n
.
n
.
r
•
J •
10 20 30 40 50 60 70 80 9 0 1 0 0 Best frequency(kHz)
Fig. 1. The distribution of BFs of 83 cerebellar neurons of Eptesicus fuscus against their recording depth (gm) (A), minimum threshold (dB SPL), (B) latencies (ms) (D), and Qlo-dB values (F). The linear regression line and the correlation coefficient (r) for each distribution are respectively shown. The range of response latency and 4 representative tuning curves are shown in (C) and (E). The upper solid line in (E) is the frequency characteristic curve of the loudspeaker. Note: A total of 85 cerebellar neurons were sampled, recording depth of 2 neurons, MT of 1 neuron and latency of 3 neurons were carelessly neglected during experiment.
125 values (a value obtained by dividing the neuron's BF by the bandwidth at 10 dB above the neuron's MT) were below 15 (range: 1.6-24.0). There was a weak correlation between the Q~0-dB value and the BF of these neurons (Fig. 1F). CerebeUar auditory neurons were generally more sensitive to FM stimuli than to pure tone pulses. Fig. 2A,B show the post-stimulus time (PST) histograms of 2 representative cerebellar neurons to 3 different stimuli. Although the response patterns of these 2 neurons did not vary with the type of stimulus used, the timing of their discharged impulses always fluctuated from stimulus to stimulus so that the duration of their PST histograms
was often longer than 10 ms when obtained with stimulus at high intensity. They also responded more vigorously to FM stimuli than to pure tone pulses. A comparison of MT of cerebellar neurons to different stimuli shows that cerebellar neurons are most sensitive to 4 ms downward sweep FM stimuli (Fig. 2C,D). A 4 ms upward sweep FM stimulus is also more effective than a 4 ms pure tone (BF) pulse (Fig. 2E). However, strongly FM-selective or specialized neurons such as those found in the inferior colliculus 4°'41 were not recorded. The mean MT (dB SPL) of cerebellar neurons is 50.3 + 19.9 (n = 84) for BF stimuli; 46.8 + 18.5 (n = 63) for the upward sweep FM
lOO--
A
C dB SPL
FM down
106
:.:.:"
F M up
P u r e tone
.:~-i:-i.~....
BF kHz
•~
,,,.,,~.i.,
.,~~.~ .
•
i~ .
.
.
60 "o
~.'
, .,,"
80
39.30 96
N=65
/"
t~,
oJ
•
40 20
r-
I-
//" i
0
•
20
i
i
40
60
.
i
80
100
Threshold for FMI stimuli
6 56
~~s." ..
~
.:~.......
'---
._ 100 m
' 'lOms Samp 50 B.W. lOOps
E
~,~
80
LI.
60
D
N=63
•e"
•
•.•"
/// •
~ ,t D /
i
i
i
40
60
80
40-
B
FM
down
F M up
Pure tone if)
20-
e*
r--
101
........
44.78
i
,
20 91
.m
lOO
E
N=63
,K~"•
E
8 1 m
',~ 80
71
D. IXI 60 0 "D "~ 4O ¢-
..-~...a~-,.
..- • :-~.
61 ¸
51
:~....
100
Threshold for FMI stimuli
::....-..
_
.....
, ,lOms Samp 50 B.W. 1 0 0 ps
/
/ ,/
•~e- 2o l.-
/
/
0 2 0,
,
4 0~
6~0
8 0'
100
Threshold for FM! stimuli Fig. 2. A,B: post-stimulus time (PST) histograms showing response patterns of 2 cerebellar neurons of Eptesicus fuscus to 4 ms pure tone, one-octave downward and upward sweep FM stimuli. Stimulus intensity (dB SPL) is shown to the left of each histogram. Samp.: sample size; B.W.: bin width. C-E: Comparison of minimum threshold (dB SPL) of cerebellar auditory neurons of Eptesicus fuscus to different acoustic stimuli• The dashed lines represent equal values for both stimuli. Points above the line indicate that the neuron had a higher MT to the stimulus plotted in the ordinate.
126 1 O0 ~ A
N=25
12
E
N=12
10
BF FM
201 -
~
0~
,
"
.
,
,
BE=79.7 kHz
N=12 12
~"
60
/ l l
10
BF
O) 40
FM
m 2o ~" --
0~ 100]
.
.
~ .
BF=76.6 kllz ,
.
N-
iil 80 C
-3
_
%
E
......
lf~ 10
BF
,
•
~ ' o
,
14
~
FM BF
•
BF:46.1 kHz • , . , .
,
N=8
H
10
~
FM
402O 0
' 70
,
, 30
• "0
,
Ipsi
BF=40.4 kHz - , . , 30 70
21
kHz were studied. According to the variation of M T with sound direction, cerebellar n e u r o n s fall into the following 4 types. Type 1 cerebellar n e u r o n s (n = 25, 53%) showed their lowest M T at 0 ° (in f r o n t of the bat) when d e t e r m i n e d with both B F pulses and F M stimuli (Fig 3A). The M T increased as the sound was delivered from lateral angles. In general, the M T reached its highest value when the sound was delivered from the most lateral angle. F u r t h e r m o r e , the M T of all these n e u r o n s was always lower to an F M stimulus than to a B F pulse. Thus,
~/ N=7
B q ~ " ~;~. /
~ 14t m _.1
1 00 ] D
60 80 1
~BF F[ k,Hz,
2 100 t B
Directional response characteristics of 47 cerebellar auditory n e u r o n s with B F ranging b e t w e e n 23.4 and 81.1
70
Contra Azimuth(deg)
,
, 30
• 0
,
Ipsi
BF=61.4 kHz I , • , • 70
30
Contra
Fig. 3. A-D: Variation of minimum threshold (dB SPL) of 4 representative cerebellar auditory neurons determined with pure tone (BF) and one-octave downward sweep FM stimuli when the sound was delivered from 9 different azimuthal angles (see text for details). E-H: Variation of response latency (ms) of 4 representative cerebellar neurons determined with pure tone (BF) and one octave downward sweep FM stimuli when the sound was delivered from 9 different azimuthal angles (see text for details).
stimuli and 41.2 + 18.2 (n = 65) for the downward sweep F M stimuli.
the F M threshold curve is always lower than the B F threshold curve (Fig. 3A). N e u r o n s of type 2 (n = 12, 26%) showed a similar directional characteristic to that of type one n e u r o n s (Fig. 3B). However, these n e u r o n s were not always more sensitive to the F M stimulus than to the B F pulse when d e t e r m i n e d at different angles. Thus, the FM threshold curve intersects the B F threshold curve at certain azimuthal angles. Type 3 cerebellar n e u r o n s (n = 3, 6%) varied their MTs only when the sound was delivered from the azimuthal angles contralateral to the recording site (Fig. 3C). The M T of these 3 n e u r o n s r e m a i n e d u n c h a n g e d when the sound was delivered from the ipsilateral angles. Directional characteristics of type 4 cerebellar n e u r o n s (n = 7, 15%) showed a very m i n o r variation in their MTs to sound direction so that both B F and F M threshold curves are extremely flat (Fig. 3D). Nevertheless, the lowest M T was always obtained when the sound was delivered at 10° or 30 ° lateral. The difference in M T and directional slope between the angles of lowest M T (the best angle) and highest M T
TABLE I
The difference in minimum threshold, response latency and directional slope between the best and worst angles of cerebellar auditory neurons Minimum threshold Type
Latency
BF Difference (dB)
1
range mean
2
range mean
3
range mean
4
range mean
All
range mean
FM Slope (dB/deg)
n = 25 11-62 0.2-0.9 38.1___12.6 0.5+0.2 n=12 20-50 0.3-0.7 32.0+9.2 0.5+0.1 n=3 27-36 0.3-0.5 30.0_+5.2 0.4+0.1 n=7 12-42 0.1-0.5 22.7_+12.8 0.3_+0.2 n = 47 11-62 0.12-0.89 33.7_+12.5 0.50_+0.20
Difference (dB)
Type BF Slope (dB/deg)
n = 21 0.1-0.7 0.4+0.1 n=12 16-46 0.2-0.7 32.3_+1.0 0.4+0.1 n=3 9-40 0.1-0.6 23.7_+15.6 0.3+0.3 n=7 7-32 0.1-0.3 15.0_+8.6 0.2+0.1 n = 43 7-48 0.1-0.7 25.9_+11.1 0.4+0.2 9-48 26.0+9.3
Difference (ms) A
2-5 3.5+1.2
B
0-7 2.4_+1.8
C
2-5 4.0+1.4
D
0-1 0.4_+0.5
All
0-7 2.5_+1.8
FM Slope (ms/deg) n = 12 0.03-1.0 0.10+0.20 n=22 0.00-0.01 0.06+0.03 n=4 0.00-0.07 0.04_+0.03 n=8 0.00-0.05 0.02+0.04 n = 46 0.00-1.0 0.07+0.1
Difference (ms)
Slope (ms/deg)
n = 10 0.5-6 0.03-1.2 3.5_+2.0 0.20_+0.30 n =22 0.0-9 0.00-0.20 2.4___1.8 0.07+0.05 n=4 3.0-5 0.03-0.05 4.0_+0.8 0.04_+0.01 n=8 0.0-1 0.00-0.03 0.2_+0.4 0.01_+0.01 n = 44 0.0-9 0.00-1.16 2.4+2.0 0.08_+0.18
127 A
1"2]A"
"O
0.8
==1
CBL
.
N=47
o •
•
#0•
0.6
~
• •• •
}
• ••• •
•
r.0.25
0.4 0.2
•
•
,'." ; 0
20
....
40
60
80
0.8 0.6
•
O
0.4
,
~
0.2
100
"
r=0.2
6,
20 30 40 50 60 70 80 90 100 FM center frequency(kHz)
2.( IC
~"
•
"e
o.ol.','. ~.',. ,\ ......
Best frequency(kHz) °'s t O
0•0••
N=31 I
D
sc
N=58
1.E
"0 0.6 ~.~ 0.4
• e
_o 0.2
.
0.(
10
15
20
25
30
35
]E
1.2 "~1,01
CBL •
0.6
r=018
•
N=46
0.2
r=0.11
0,0 40
,
.
40
,
.
60
,
.
80
100
~.2~ F ~
0.8
~
0.4
"
CBL
N=43
0.6 •
20
.
Best f r e q u e n c y ( k H z )
•
~-0.4
,
20
40
Best frequency(kHz)
0
e• "l
O 0.~
60
80
Best f r e q u e n c y ( k H z )
100
•
o
0.2
•
o.o
~
•
'
r=0.07
~
20 30 40 50 60 70 80 90 100 FM center frequency(kHz)
Fig. 4. Six scatter plots showing the distribution of directional slope in MT variation (dB/deg) (A-D) and in latency variation (ms/deg) (E,F) of single neurons from 4 different brain centers with sound direction against the BF (A,C,D,E) or FM center frequency (B,F). The linear regression line and correlation coefficient for each distribution are respectively shown. CBL, cerebellum; IC, inferior colliculus; SC, superior colliculus.
(the worst angle) for each type of cerebellar neuron is shown in Table I. While the MT difference between the best and worst angles determined with BF and FM stimuli is comparable in type 2 cerebellar neurons, it is consistently larger for BF pulses than for FM stimulus for type 1, 3 and 4 cerebellar neurons. Thus, when the directional sensitivity of all cerebellar neurons is considered, it is clear that the MT difference and directional slope are larger when determined with BF pulses than with the FM stimuli (Table I, bottom row). The correlation between the directional sensitivity and the stimulus frequency was examined by plotting the directional slope (dB/deg) of cerebellar neurons against either the BF (Fig. 4A) or the center frequency of each FM stimulus (Fig. 4B). The fact that a linear regression analysis on both distributions reveals a low correlation coefficient (r = 0.25 and 0.2 respectively) suggests that the directional sensitivity of cerebellar neurons is relatively independent of the stimulus frequency. This finding is similar to data obtained from superior collicular neurons of the same bat (Fig. 4D). However, consistent with previous findings az-a4' 26,32,35,37 the directional slope of inferior collicular neurons is significantly dependent upon the stimulus frequency (Fig. 4C).
Cerebellar neurons also varied their response latencies when the sound was delivered from different azimuthal angles. They can also be generally classified into 4 types according to their latency variation. Type A (n = 12, 26%) neurons generally had a shorter latency when determined with FM signals than with BF pulses. Thus, the FM latency curve is always lower than the corresponding BF latency curve. Nevertheless, both curves show a dip at 0 ° and rise on both limbs when the sound is delivered at lateral angles (Fig. 3E). In other words, these neurons had their shortest latency when the sound was delivered from the front. Cerebellar neurons o f type B (n = 22, 48%) changed their latency unpredictably with sound direction when determined with both stimuli. Thus, the BF latency curve does not vary in parallel with the FM latency curve resulting in intersection of the two curves at certain azimuthal angles (Fig. 3F). Furthermore, while the latency of some neurons of this type varied with sound direction when determined with BF pulses, it remained unchanged or fluctuated only slightly with sound direction when determined with FM stimuli or vice versa. Cerebellar neurons of type C (n = 4, 9%) had their longest latency for both sound stimuli delivered at 70° ipsilateral. They systematically shortened their latencies when the sound was broadcast sequentially from ipsilateral azimuthal angles to contralateral ones (Fig. 3G). Response latency of type D cerebellar neurons (n = 8, 17%) either changed very little or remained the same with sound direction (Fig. 3H). Thus they have very poor directional characteristics in terms of latency variation. The difference in latency and directional slope between the angles of shortest (the best angle) and longest (the worst angle) latency for each type of cerebellar neuron is shown in Table I. It is clear that cerebellar neurons do not appear to vary the response latency to a greater extent for a particular acoustic stimulus, different from that of MT variation. However, similar to the frequency-independent MT variation with sound direction, a linear regression analysis also reveals no appreciable correlation between the sharpness of the directional sensitivity in latency variation and the stimulus frequency (Fig. 4B,D). Our present study of response properties of cerebellar neurons confirms an early observation on the same bat 28 that the phasically discharged cerebellar neurons are generally broadly tuned (Fig. 1E) and are more sensitive to FM stimuli than to pure tone pulses (Fig. 2). As discussed previously24"2s, such a finding suggests that the cerebellum is not involved in further fine frequency analysis but rather in integrating and utilizing processed auditory information for motor orientation. The majority of cerebellar neurons studied show a lowest MT to both stimuli delivered from the front (n = 37, 79%, Fig. 3A,B)
128 or from 10 ° to 30 ° lateral (n = 7, 15%, Fig. 3D). This finding is similar to previous studies in bats 43 and in cats 2 suggesting that the cerebellum is able to orient an animal's head toward a localized sound within its frontal gaze. We noted that cerebellar neurons have a sharper directional sensitivity to a pure tone pulse than to an FM stimulus. Thus the difference in MT and directional slope between the best and worst angles is larger when determined with the pure tone pulse (Table I). We believe that such a difference is perhaps due to the fact that the directional sensitivity of a neuron to a pure tone pulse is primarily determined by the directional characteristics of a single frequency but its directional sensitivity to an FM stimulus is the combination of directional characteristics of each contained frequency component. Thus, the directional sensitivity to an FM stimulus may become less sharp as a result of integration or averaging of directional characteristics of individual frequency components. Nevertheless, since the BF of each neuron was always set as the center frequency of the FM stimulus, the most sensitive angle of each neuron was always the same when determined with both stimuli (Fig. 3 A - D ) . A similar finding for inferior collicular neurons of the same bat has been reported previously 26. We observed that, excluding 8 neurons (e.g. Fig. 3H), all cerebellar neurons generally vary their response latencies significantly with sound direction (Fig. 3 E - G ) thus providing another possible neural code for sound localization. Such an observation has been reported in the auditory neurons of cats 9, bats 2°'21, insects 34 and frogs 1°. When a sound is broadcast from different directions, the interaural intensity changes as a result of directional characteristics of the animal's head and pinna. Such a change may be reflected by the variation of the number of impulses, MT and response latency of the recorded neuron with sound direction. Since all cerebellar neurons discharged phasically (no more than 7 impulses per stimulus), the variation of their number of impulses with sound direction per stimulus may be quite limited. Thus, auditory physiologists often study the variation of the number of impulses of a neuron with sound direction per 30 or 50 repeated stimuli. On the other hand, our observation on the variation of the response latency of these neurons with sound direction supports an earlier suggestion 9'1°'34 that the response
latency may also be a useful code for sound localization. A similar latency-shift mechanism in coding an object in the electric field by the electroreceptors of a weakly electric fish, mormyrids, has been proposed earlier 36. Previous studies of the inferior colliculus 12-14'26"32'35'37 (e.g. Fig. 4C) and the auditory cortex 25 have shown that the directional sensitivity of auditory neurons is frequency-dependent so that it sharpens with increasing stimulus frequency. We 35 have suggested that the frequency-dependent directional sensitivity is probably the result of directional filtering of sound by the pinna and head 22. However, our present study shows that the directional sensitivity of cerebeUar neurons is frequencyindependent (Fig. 4). Since the BF of our sampled neurons ranged from 23.4 to 81.1 kHz covering almost the entire range of the frequency spectrum of the bat's echolocation signals 23'3s, it is unlikely that our observation is simply due to a sampling bias of a particular population of cerebellar neurons. In one of our early unpublished observations, we noted that the directional sensitivity of superior collicular neurons of the same bat was also frequency-independent (Fig. 4D). Since both the superior colliculus and the cerebellum are involved ill auditory orientation 15'16'46, they apparently integrate and utilize the directional sensitivity of their presynaptic auditory neurons in order to orient an animal's head and pinnae toward a sound source. For sound localization, neurons with frequency-independent directional sensitivity are certainly more versatile and effective in providing information about the direction of the sound source. It appears that the frequency-dependent directional sensitivity is somehow transformed into the frequency-independent one after sensorimotor integration. The fact that both the superior colliculus and the cerebellum lack tonotopic organization and their neurons are more sensitive to FM stimuli 27'28 appears also to be the result of such integration process. Future studies of the directional sensitivity of the cerebellar neurons in terms of variation in the number of impulses with sound direction are necessary to reconfirm our present finding. This study was supported by a grant from the National Institute of Health (NS 20527) and a grant from the Research Council from the University of Missouri (Bio. Med 017). We thank Mr. D. Pinheiro for his fruitful discussion and an anonymous reviewer's comments on an earlier version of this paper and Ms. K. Cook for secretarial assistance.
1 Aitkin, L.M. and Boyd, J., Responses of single units in cerebellar vermis of the cat to monaural and binaural stimuli, J. Neurophysiol., 38 (1975) 418-429. 2 Aitkin, L.M. and Rawson, J.A., Frontal sound source location is represented in the cat cerebellum, Brain Research, 265 (1983) 317-321. 3 Altman, J.A., Bechterev, N.N., Radinova, E.A, Shimigidina,
G~M. and Syka, J., Electrical responses of the auditory area of the cerebellar cortex to acoustic stimulation, Exp. Brain Research., 26 (1976) 285-298. 4 Bechterev, N.N., Syka, J. and Altman, J.A., Responses of cerebellar units to stimuli simulating sound source movement and visual moving stimuli, Experientia, 31 (1975) 819-821. 5 Bloedel, J.R., Cerebellar afferent systems: a review, Prog.
129
Neurobiol., 2 (1975) 1-68. 6 Bloedei, J.R., Dichgans, J. and Precht, W., Cerebellar Functions, Springer, Berlin, 1985. 7 Eccles, J.C., The cerebellum as a computer: patterns in space and time, J. Physiol. (Lond.), 229 (1973) 1-32. 8 Eccles, J.C., Ito, M. and Szentagothai, J., The Cerebellum as a Neuronal Machine, Springer, New York, 1967. 9 Erulkar, S.D., The responses of single units of the inferior colliculus of the cat to acoustic stimulation, Proc. R. Soc. Lond., B, 150 (1959) 336-355. 10 Feng, A.S., Directional response characteristics of single neurons in the torus semicircularis of the leopard frog, (Rana pipiens), J. Comp. Physiol., 144 (1981) 419-428. 11 Fadiga, E. and Pupilli, G.C., Teleceptive components of the cerebellar function, Physiol. Rev., 44 (1964) 432-486. 12 Fuzessery, Z.M. and Pollak, G.D., Determinants of sound location selectivity in bat inferior colliculus: a combined dichotic and free-field stimulation study, J. Neurophysiol., 54 (1985) 757-781. 13 Grinnell, A.D., The neurophysiology of audition in bats: directional localization and binaural interaction, J. Physiol. (Lond.), 167 (1963) 97-113. 14 Grinnell, A.D.,and Grinnell, V.S., Neural correlates of vertical localization of echolocating bats, J. Physiol. (Lond.), 181 (1965) 830-851. 15 Harris, L.R., The superior colliculus and movements of the head and eyes in cats, J. Physiol. (Lond.), 300 (1980) 367-391. t6 Henkel, C.K. and Edwards, St. B., The superior colliculus control of pinna movements in the cat: possible anatomical connections J. Comp. Neurol., 182 (1978) 763-776. 17 Highstein, S. and Coleman, P.D., Responses of the cerebellar vermis to binaural auditory stimulation, Brain Research, 10 (1968) 470-473. 18 Ito, M., Is the cerebellum really a computer? Trends Neurosci., (1979) 122-126. 19 Ito, M., The Cerebellum and Neural Control, Raven, New York, 1984. 20 Jen, P.H.-S., Coding of Directional Information by Single Neurons in the Superior Olivary Complex of Echolocating Bats, Ph.D. thesis, Washington University-St. Louis, 1974. 21 Jen, P.H.-S., Coding of directional information by single neurons in the S-segment of the FM bat, Myotis lucifugus, J. Exp. Biol., 87 (1980) 203-216. 22 Jen, P.H.-S. and Chen, D.M., Directionality of sound pressure transformation at the pinna of echolocating bats, Hearing Res., 34 (1988) 101-118. 23 Jen, P.H.-S. and Kamada, T., Analysis of orientation signals emitted by the CF-FM bat Pteronotusp. parnellii and the FM bat Eptesicus fuscus during avoidance of moving and stationary obstacles, J. Comp. Physiol., 148 (1982) 389-398. 24 Jen, P.H.-S. and Schlegel, P.A., Neurons in the cerebellum of echolocating bats respond to acoustic signals, Brain Research, 196 (1980) 502-507. 25 Jen, P.H.-S., Sun, X.D. and Lin, P.J.J., Frequency and space representation in the primary auditory cortex of the frequency modulating bat, Eptesicus fuscus, J. Comp. Physiol., 165 (1989) 1-14. 26 Jen, P.H.-S., Sun, X.D., Chen, D.M. and Teng, H.B., Auditory space representation in the inferior cofliculus of the FM bat,
Eptesicus fuscus, Brain Research, 419 (1987) 7-18. 27 Jen, P.H.-S., Sun, X.D., Kamada, T., Zhang, S.Q. and Shimozawa, T., Auditory response properties and spatial response areas of superior coUicular neurons of the FM bat, Eptesicus fuscus, J. Comp. Physiol., 154 (1984) 407-413. 28 Jen, P.H.-S., Kamada, T., Sun, X.D., Schlegel, P., Vater, M., Harnischfeger, G. and Riibsamen, R., Responses of cerebellar neurons to acoustic stimuli in certain FM and CF-FM bats, Chinese J. Physiol., 27 (1984) 1-26. 29 Kirzinger, A., Cerebellar lesion effects on vocalization of the squirrel monkey, Behav. Brain Res., 16 (1985) 177-181. 30 Larson, C.R., Sutton, D. and Lindeman, R.C., Cerebellar regulation of phonation in rhesus monkey (Macaca mulatta), Exp. Brain Res., 33 (1978) 1-18. 31 MacKay, W.A. and Murphy, J.T., Cerebellar modulation of reflex gain, Prog. Neurobiol., 13 (1979) 361-417. 32 Makous, J.C. and O'Neill, W.E., Directional sensitivity of the auditory midbrain in the mustached bat to free-field tones, Hearing Res., 24 (1986) 73-88. 33 Mitra, J. and Snider, R.S., Cerebellar modification of unitary discharges in auditory system, Exp. Neurol., 23 (1969) 341-352. 34 M6rchen, A., Rheinlaender, J. and Schwarzkopf, J., Latency shift in insect auditory nerve fibers. A neural time cue of sound direction, Naturwissenschaften, 54 (1978) 656. 35 Poon, P.W.E, Sun, X.D., Kamada, T. and Jen, P.H.-S., Frequency and space representation in the inferior colliculus of the FM bat, Eptesicus fuscus, Exp. Brain Res., 79 (1990) 83-91. 36 Szabo, T, and Hagiwara, S., A latency-change mechanism involved in sensory coding of electric fish (Mormyrids), Physiol. Behav., 21 (1967) 331-335. 37 Shimozawa, T., Suga, N., Hendler, P. and Schuetze, S., Directional sensitivity of echolocation systems in bats producing frequency-modulated signals, J. Exp. Biol., 60 (1974) 53-59. 38 Simmons, J.A., Fenton, M.B. and O'Farrell, M.J., Echolocation and pursuit of prey by bats, Science, 203 (1979) 16-21. 39 Snider, R.S. and Stowell, A., Receiving areas of the tactile, auditory and visual systems in the cerebellum, J. Neurophysiol., 7 (1944) 331-358. 40 Suga, N., Analysis of frequency-modulated and complex sounds by single auditory neurons of bats, J. Physiol. (Lond.), 198 (1968) 51-80. 41 Suga, N., Classification of inferior collicular neurons of bats in terms of responses to pure tones, FM sounds, and noise bursts, J. Physiol. (Lond.), 200 (1969) 555-574. 42 Suga, N. and Schlegel, P.A., Neural attenuation of responses to emitted sounds in echolocating bats, Science, 177 (1972) 82-84. 43 Sun, X.D., Jen, P.H.-S. and Zhang, W.P., Auditory spatial response areas of single neurons and space representation in the cerebellum of echolocating bats, Brain Research, 414 (1987) 314-322. 44 Teramoto, S. and Snider, R.S., Modification of auditory responses by cerebellar stimulation, Exp. Neurol., 16 (1966) 191-200. 45 Velluti, R . and Crispino, L., Cerebellar actions on cochlear microphonics and on auditory nerve action potential, Brain Res. Bull., 4 (1979) 621-624. 46 Wolfe, J.W., Responses of the cerebellar auditory area to pure tone stimuli, Exp. Neurol., 563 (1972) 295-309.