Inferior

Colliculus.

Response

Properties

Pericentral, L.

M.

AITKIN,

I. Comparison of Neurons

and External W.

R. WEBSTER,

Neuropsychology Laboratory, Monash University, Clayton,

J.

L.

Nuclei VEALE,

Departments of Psychology Victoria 3168, Australia

INFERIOR COLLICULUS of the cat is, in Nissl-stained sections, seen to have three main divisions (5). The central nucleus (ICC) is the largest component and is characterized in Golgi material by the presence of laminae composed of cell bodies and dendrites (23, 28). It is surrounded by an irregular annulus of tissue cytoarchitecturally divisible into two further components: the pericentral nucleus (ICP), dorsal and posterior to ICC, and the external nucleus (ICX), lateral and rostra1 to ICC. Most physiological studies, where histological divisions are specified, have concentrated their attention on the properties of neurons of the central nucleus (2 1, 3 1, 32), and only brief comments have appeared in the literature relating to the external nucleus or the pericentral nucleus (21, 31, 37) of the cat. More substantial material is available describing ICX of the rabbit (2). In view of the fact that ICX and ICP together form a large proportion of the inferior colliculus, and the fact that cell types and connections are often very different from those ircl ICC (9, 23, 24, 28, 29, 30, 35), we studied responses to tonal stimuli of neurons in these three regions of the anesthetized cat. Particular attention has been given to their tuning characteristics and binaural responses since it is probable that these response parameters are less likely to be modified by the presence of an anesthetic agent than are the temporal discharge patterns of neurons (6). Some of the results preTHE

Received 1196

for

publication

December

16, 1974.

AND and

of in Central, of Adult D.

C.

Cat

CROSBY

Physiolog-y,

sented here are simil.ar to those reported in the recent study of Merzenich and Reid (21) . METHODS

Twenty-five adult cats with clean outer and middle ears were used in this study. Twentytwo animals were anesthetized (40 mg/kg) and maintained with intraperitoneal injections of sodium pentobarbital, while the remaining three cats were anesthetized with 1% chloralose -10% urethan in saline (5 ml/kg intraperitoneal, initial dose). After tracheal cannulation the pinnae were removed bilaterally to allow the -insertion of sound-path assemblies. The inferior colliculus was exposed by aspiration of the occipital cortex and the brain stem was stabilized with a 3% agar gel. The bullae were opened and silver ball electrodes placed on the round windows to monitor the cochlear responses to clicks. Animals were supported in a m odified TrentWells stereotaxic apparatu .S located in a sound-attenuated room, and were maintained at near 38.5OC rectal temperature with a feedback blanket unit. Recordings of the activity of single units were made with glass-coated tungsten microelectrodes or stainless steel electrodes coated with epoxylite, connected to a high-gain highimpedance preamplifier. Most of the observations made in this study were based on oscilloscope display and aural monitoring, but in certain cases action potentials we re used to trigger s a discriminator generating standard pulses, which were stored on magnetic tape f&r later analysis. Methods of analysis have been documented in previous reports from this laboratory (3). Tone pips, routinely of 300 ms duration and 10 ms rise and fall time, were generated as described previously (4) and passed to paired Beyer DT48S earphones in metal housings,

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS

OF

INFERIOR

which were connected by lo-cm-long soft polyethylene tubing to sound-path assemblies. In the 13 later animals, calibrated probe tubes concentrically located with their tips 2-3 mm from the eardrum, were connected to Briiel and Kjaer 0.5-inch microphones for accurate measurement of sound intensity. Such measurements in the 12 earlier animals were achieved by coupling the sound-path assembly directly to a Briiel and Kjaer 0.5-inch microphone in a sealed system to derive a standard set of calibrations used in all the early animals. The results for the two methods of calibration were harmonious, although some irregularities in tuning curves, using the direct-coupling methods, were probably attributable to resonances not revealed by this method of measurement. Estimates of acoustic crossover were obtained from 18 animals, and interaural attenuation was never less than 47 dB over the range of frequencies employed. Successful electrode penetrations were terminated by making a lesion at the terminal point. In many cases a second lesion was produced at the presumed physiological border between ICP and ICC. Postmortem fixation of the brain in 10% formal-saline was succeeded by section of the midbrain in a sagittal plane at 50 pm. Relevant sections were stained either with cresyl fast violet or Thionine. All units reported in this study could be assigned a location in one of the three subnuclei of the inferior colliculus. RESULTS

Recordings were made from 292 units, of which 52 were studied in cats anesthetized with chloralose-urethan. The results to be described are based on the properties of 150 of these units for which adequate studies were made of tuning characteristics or binaural behavior, in addition to accurate establishment of their histological location. There appeared to be no major differences in the thresholds, binaural properties, or tuning characteristics of the units isolated under the two anesthetic conditions. All penetrations were made from dorsorostral to ventrocaudal through the inferior colliculus, although the point of entry of the microelectrode was varied in order to sample the different subdivisions of this nucleus. Electrodes first traversed either ICP (19 penetrations), in which dendarkly staining cells were sely packed, associated with neurons having long den-

COLLICULUS

1197

drites prominent even in Nissl stains (Fig. 2) or, in more rostra1 approaches penetrated the external nucleus (5 penetrations), characterized by large cells with dendrites loosely arranged among glial elements (Fig. 4). Subsequently, the majority of unit recordings were made in the central nucleus where closely packed neurons of diameters midway between those of ICP and ICX were encountered (Figs. 2, 4). Tuning curves were measured for 141 located units (Table 1). It was usually possible to derive such curves for units in ICC, but considerable difficulty was often encountered in tuning units in ICP. These units frequently responded to tonal stimuli irregularly and exhibited what has been referred to as “habituation” (1, 2, 37). This phenomenon, whose occurrence may be regarded as a distinguishing feature of cells in ICP compared with ICC, aPPea red as a tendency for a unit to fire to the fiS r t of a train of stimuli, and thereafter to discharge erratically (Fig ‘. 1) or not at all . Frequently, an alteration in tone frequency wo&ld lead to a resumption of stimulus-locked firing, but this again would wane as stimuli were presented at the new tone frequency. For some units, habituation occurred at frequencies on the low- and high-frequency boundaries of tuning curves for elements in ICP and ICX (e.g., unit 11, Fig. 5). Habituation was a function of stimulus repetition rate and occurred markedly for 4-l

l

4 5 IMULUS

H

WJ 67-2

6

7 8 9 SEQUENCE

FIG. 1. Habituation of resDonses with stimulus repetition for two onset units. Fbr WJ67-2, responses to successive blocks of four stimuli are summed, while for WJ67-3, successive blocks of two stimuli are summed. Repetition rate for both units, I/2 s, stimuli 1,000 Hz at 80 dB SPL for unit WJ67-2, 1,000 Hz at 75 dB SPL for unit WJ67-3.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

1198

AITKIN,

WEBSTER,

VEALE,

AND

CROSBY

TABLE 1. Summary of tuning curve data for units in three divisions of inferior colliculus

No. of units (N) No. of tuning curves Too difficult to tune

ICC

%N

ICP

%N

105 105 0

100 0

25 17 8

68 32

some units at presentation rates as low as l/2 s (Fig. 1). As a result, 32% of the units encountered in ICP could not be tuned at all to. tonal stimuli (Table 1). Similar habituation was observed for units in ICX, although in this division it was usually possible to select a low repetition rate at which a regular discharge could be

ICX

%

20 19 1

95 5

elicited; thus, only one ICX unit could not be tuned. Tuning The (WJ64) nated quence

characterzstics electrode track depicted in Fig. 2 passed through ICP and termiventrally in ICC. The short seof observations in ICP was charac-

WJ-64 E/O

-30.0. SWISH

E/O E/g (1) E/O (F) E/E E/O (F)

EjOF) E/E E/E g (1) E/O (1) i/E E/O E/O

FIG. 2. Electrode penetration in experiment WJ-64, traversing both pericentral and central nucleus. H.U. = habituating unit; swish = background response to tones with no discriminable spikes; E/O = contralateral response only; LLD = dorsal nucleus of the lateral lemniscus. Other nomenclature explained in text. Each number is the best frequency for an isolated unit. Only units for which tuning curves were measured have been included in the central nucleus sequence.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018. Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS

OF

INFERIOR

terized by the presence of habituating units which were broadly tuned (e.g., WJ64-1, Fig. 6). Some suggestion of diminishing best frequency was discerned in ICP, although the concept of a best frequency is difficult to apply to such broadly tuned units. The major part of the penetration sampled units from the central nucleus, and a selection of tuning curves from this region is shown in Fig. 3. These curves characteristically occupied a restricted part of the frequency spectrum and were distinctly different from those of ICP (Figs. 9 and 10). The tonotopic organization of the central nucleus, originally demonstrated by Rose et al. (31) and precisely documented by the recent study of Merzenich and Reid (21), was also routinely observed in our study. The more rostra1 and lateral penetration of Fig. 4 (BL-IO) sampled from units in both ICX and ICC. A clear tendency

2d

COLLICULUS

,

,

03

OS

,

1199

,,I,

I

‘I”“1

I

50

10 FREO&CY

10 0

200

(kHz)

FIG. 3. Tuning curves for seven units encountered in experiment WJ-64. Unit number, related to depth at which the unit was isolated, appended to each curve.

for best frequencies to decrease initially in the track correlated with the passage of the electrode through the outer rind of the nucleus, in this case identifiable as ICX. The first tuning curves were broad and irregular (Fig. 5, units ‘7 and 8) with

BL-10

L.F.

oL.:-o*3 0:7' E/E 0.8 E/O F 0.7 E/O IIF 1.6 I/I 0.5-2.0 2.1 E/O ; .; $F) IFIG. 4. Electrode penetration in experiment BLIP, traversing spontaneously firing unit inhibited by both ipsilateral and contralateral swish. Other nomenclature as for Fig. 2.

the external and stimuli. L=lesion,

central nuclei. I/I L.F.=low-frequency

=

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

AITKIN,

1200

WEBSTER,

VEALE,

100

BL-10 X

I

005

01

5.

02

1.0 20 FRECNJENCY (kk)

50

10 0

200

300

for six units isolated in the The letter H on the lower and upper frequency limits of unit 11 signifies that habituation was so marked at upper and lower frequencies that the tuning curve could not be completed. FIG.

external

Tuning nucleus

05

curves in BL-IO.

lowest thresholds between 12 and 15 kHz. Thereafter, lower frequencies were encountered, culminating in the isolation of unit 14 which had a best frequency of 100-200 Hz, at the border of ICC. The subsequent observations were from the central nucleus, as determined by histological considerations, but these did not show any tonotopic ordering. Since the penetration was rostral, lateral, and oblique to the vertical plane, it is possible that only the lateralmost laminae of ICC were here penetrated. The extremely broad tuning of some units in ICP and TCX is exemplified by the tuning curves of Figs. 6 and 9. Units BL7-9 and 3-8 (Fig. 6) were isolated in ICX, and thresholds for these units could be obtained over a range of 6.5 octaves. Unit W”64-1 (Fig. 6) from ICP exhibited a

AND

CROSBY

region of best frequency, but nonetheless could be .fired by a similar range of tone P frequencies. Most tuning curves of units within ICC were V shaped with a rapid roll-off at frequencies above the best frequency, but the extent of the low-frequency arms of the curves was variable. Some units with high best frequencies could be excited by relatively intense stimuli at 4 octaves below best frequency, while other units at equivalent intensities fired only to frequencies within half an octave of best frequency. The extremes of this gradation are illustrated by selected tuning curves from ICC in Fig. 7. The low-frequency arms of some high-frequency tuning curves closely resembled those of cochlear nerve fibers (18) (Fig. 7) in that they appeared as “tails” to the more sensitive part of the curve, but a subset of the total

NRk

y7-4

63-2OJ

1

looloo80-l $5

80-

60-

m 0

60-

. 40-

20-

B O-I,

I

o-2

I

1

Y

I”“1

0.5

1-o

, 2.0

t

FREQUENCY FIG.

central

BL7-9).

,I,,,,

5.0

Y

10.0

n

I

uI 20.0

I 30.0

r

3

T

5

7

1.

3I

FREQUENCY

(kHz)

6. Tuning curves for units nucleus (I&“641) and external

2

1.

isolated nucleus

in peri(BW-8,

FIG.

curves

’

5I

*l”l’“‘l

7

10

15 201

301

10 13

7. Tailed tuning (B) for units located

(ktir)

curves (A) and sharp tuning in ICC. NR = no response

at any intensity tested.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS

OF

INFERIOR

population were extremely sharply tuned and their curves occupied a very restricted range of frequencies, even at intensities near 100 dB SPL (Fig. 7B). Unit DC5-I I (Fig. ‘7B) was one of the most sharply tuned units in our sample-at 60 dB it discharged only at frequencies varying from 19,liOO to 22,300 Hz (0.2 octave range). Ninety-two tuning curves from ICC (88%) were sufficiently complete and regular to enable measurement of the bandwidth at 10 dB above best-frequency threshold, as a function of best frequency (Q). Such a measure of sharpness of tuning is important since it removes the bias that a logarithmic plot of a tuning curve necessarily introduces. Only half of the tuning curves from ICP and ICX were amenable to bandwidth measurements because they were often very irregular in shape (e.g., Figs. 6, lOA). Q factors obtained for neurons in the three divisions are plotted in Fig. 8. The distribution of Q is a nonmonotonic function of frequency, greatest values of Q (i.e., sharpest tuning) occurring between 10 and 17 kHz and diminishing on either side of this frequency range. This relationship is generally similar to that observed in the ventral division of the medical geniculate body (MGB) (4). Q values for units from ICP

l

ICC

N=92

o x

ICP ICX

N=8 N=lO l l

COLLICULUS

1201

and ICX were intermingled with those from ICC, but tended to have low values. While clear differences existed between the shapes of tuning curves of units in ICC and its neighboring nuclei, no differences could be detected in the thresholds of units at best frequency. Thresholds from both regions were distributed over a range of 60 dB, depending on the sound frequency. Broadness

of tuning

at higher

Binaural responses of units three divisions

FIG.

8.

(bandwidth threshold)) 1 10 units

Value of Q ((best frequency in Hz)/ in Hz at 10 dB above best-frequency plotted as a function of best frequency for in inferior colliculus.

intensities

Figures 9 and 10 relate the tuning curves of four units in ICP to their response ranges at higher intensities. With the exception of unit lV\54-4 (Fig. 9A) . which discharged in a sustained fashion throughout the tone, all discharges were at the onset of the stimulus and only these are included in the isointensity response areas. For three units (Figs. 9A, and lOA, B), the tuning curves gave an accurate indication of the response ranges at high intensities. Unit DC&5 (Fig. 1OB) was one of the more discretely tuned units of ICP, although at 80 dB responses were evoked over a range of 3 octaves. This, however, contrasted with the remarkable broadness demonstrated by the sustained discharges of unit wJ54-4 (Fig. 9A). It is obvious that firing could be elicited by tones separated by at least 6 octaves and could have extended well above 30 kHz. Unit DUO-6 (Fig. 9B) demonstrated a clear best frequency near 3,000 Hz for its threshold tuning curve, but the best frequency and shape of this curve appeared in no way related to the isointensity spike-count functions at higher intensity levels. This unit, which habituated markedly, demonstrates the characteristically irregular, broad firing to tones observed for most units in ICP. in

The contralateral, ipsilateral, and binaural responses to tones were routinely and carefully assessed for units in this study. Reports of detailed binaural interactions in the inferior colliculus revealed in some of these data have been published elsewhere (38, 39) and will not be further pursued here. It was sufficient

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

AITKIN,

1202

WEBSTER,

100 60

w

40

E

20

AND

CROSBY

A ““1 70-

v,,, 2

VEALE,

i

--

A -100 80I

,,,,,, o-5

1.0

2.0

5.0

10.0

800

20.0

30.0

0:426

100

A

600 I5 8400

3'4

510 -8OdB

1:O

2:o

30

510

1.0

2.0

1.0

2.0 3.0

0

ol

, , ,“, 0.5 1.0

7.

J 1 2.0

’

’ I’“‘1 50

1

1

so40--

,;,

m 302014-

1.0

3.0

10.0

20.0

20.0

Ok

0:6

1

L/--_--z ,,, , , 10.0

R

B 6o

u

2:o

520 6.2

;

1:O

580 2 u 60 !$40

30.0

30.0

-1

70 10.2

60

I

‘I

I

I’lI-1

3.0

5.0

o10.0

5.0

10.0

m80dBSPL x-x60dBSPL

2 50 840

I

0.4 0.6

I

w 30

+ -0 80dB -4 60dB x---x 40 dB

i

20 10

1

A

d

0' 0.2

I

I’I

0.4

0.6

1

11’1’

FREQUENCY(kHz) f

0.5

’

“‘I

I

10

I

“““I

2.0 3.0 FREQUENCY(kHz)

I

10.0

20.0

I

30.0

FIG. 10. Threshold tuning curves (upper) and isointensity response areas (lower) for units DC204 (A) and DCb-5 (B). Labeling as for Fig. 9.

FIG. 9. Threshold tuning curves (upper) and isointensity response areas (lower) for units WJ54-4 (A) and DUO-6 (B). All p oints on response area curves were obtained by summing the responses to 20 repetitions of a 300 ms duration tone pip at the stated frequencies and intensities, repeated at l/2 s. The time over which spikes were counted for unit wJ54-4 corresponded to the total tone-on period. That for unit DC20-6 and the further two units of Fig. 10 was a period beginning at 5 and finishing at 20 ms after the beginning of the tone (onset response).

binaural stimuli as did units in ICC. The 22 units in ICP were, with 6 exceptions, driven only by contralateral stimulation. Those neurons which were also influenced by the ipsilateral ear tended to have narrower tuning curves than the broad monaural curves of the 64% majority. However, one exception was unit WJ54-4 (Fig. 9A), which was strongly for our purposes to define the nature of binaural yet fired over an extremely the input from each ear and the result of broad range. simultaneous binaural stimulation at best Table 3 gives a breakdown of the nature frequency. of the binaural input for the units in this Table 2 indicates that, for 87 units of sample. At best frequency, three principal ICC, 70 (80%) were influenced by stimu- types of binaural response were recognized lation of each ear and 17 (20%) were only with binaural stimuli of equal magnitude Excitatory influenced by stimulation of one ear. An presented simultaneously. identical proportion was observed for the influences could be manifested bv excitaslender sample of units in ICX and, gen- tory responses to simulation of either ear erally, units in this subdivision behaved alone (EE) or by an excitatory response to similarly and equally sensitively toward stimulation of one ear, no response to

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS

TABLE

-

2.

No. of units Monaural Binaural Monaural: stimulated

Binaural

influence

(N)

OF

on units

INFERIOR

ICC

%N

ICP

%N

ICX

%N

87 17 70

20 80

22 14 8

64 36

16 3 13

19 81

response to one ear stimulus (EO). Binaural: some influence

only and no modifications observed on stimulation

DISCUSSION

A number of important findings have emerged from this study. First, a population of neurons in ICP exhibited very broad tuning compared with the narrow tuning of TABLE 3. Binaural response tee with units in three divisions

observed

EE

EO (F)

EO (1)

25 6 4

11 (2) 2 (1) 2

34 (1) 5 2

Numbers in parentheses which ipsilateral and not produced excitation.

1203

in three divisions

stimulation of the other ear alone, and a facilitation of the response when both ears were stimulated together (EO (F)). Inhibitory responses appeared in the form of an inhibition of a monaural response when both ears were stimulated together (EO (I)). In the latter two categories of responses, the contralateral ear was dominant in producing excitation and only four units were observed for which ipsilateral stimulation excited while the contralateral ear did not. For some-units, considerable variation in the magnitude of the binaural response occurred with alterations in the interaural time delay between the ipsilateral and contralateral stimuli, as has been reported previously (32). However, with the classification illustrated in Table 3, each delaysensitive neuron could be assigned to one of the three categories of EE, EO (F), and EO (I), when simultaneous binaural stimulation was used. All types of binaural response were observed in the three nuclei of the inferior colliculus, and the breakdown of binaural response type is very similar to that observed in the ventral division of MGB (4).

ICC ICX ICP

COLLICULUS

are number contralateral

of units for stimulation

of this response of each ear.

when

the other

ear was

the majority of units in ICC. Broadness and irregularity were also common features of units in ICX and, in both regions, were associated with the phenomenon of habituation. Second, the majority of units in ICP were driven only by the contralateral ear, while at least 80% of neurons in ICC and ICX were binaurally influenced. Third, further evidence was found to support the concept of reverse tonotopic sequences in both ICP and ICX. Broad tuning has previously been observed in the outer rind of tissue surrounding the central nucleus of the inferior colliculus (2, 2 1, 37), and it is also a common feature of the responses in both the magnocellular division of MGB (1) and in area AI1 of the auditory cortex (20) It’ is unlikely that broadness of tuning would be attributable to the presence of an anesthetic agent since anesthesia would be likely to restrict, rather than to spread, the range of effective frequencies by suppressing information transmission at the edges of the response area (3, 6). Furthermore, thresholds at best frequency were similarly distributed for units in all divisions, again suggesting that no selective effects due to anesthetic had occurred. This fact, and the existence of broadly tuned units in other areas of the brain, also argues against the presence of local damage to ICP due to operative or surface trauma as being the cause of broad tuning. Relationship between cell morphology

unit

response and

Three recent reports (9, 29, 30) have presented structural details for cells in the inferior colliculus. The central nucleus contains bitufted (29) or disc-shaped (9)

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

1204

AITKIN,

WEBSTER,

cells arranged in fibrodendritic laminae. In addition, stellate (9) or multipolar (29) neurons have dendrites which lie across one or several laminae. A physiological correlate for these cell types could be that sharply tuned units of ICC may be derived from bitufted neurons while more broadly tuned elements may relate to the stellate forms, since bitufted neurons are likely to receive a more restricted fiber spectrum than are stellate neurons. Descriptions of cell types in ICP are hampered by disagreement between the two principal studies (9, 30) as to what is included in this nucleus. It is perhaps more useful to consider only the studies of Rockel and Jones (29, 30) since they were confined to the cat. These authors described four cell types-small and large spiny cells, multipolar, and fusiform cells. The locally ramifying axons of the small spiny cells indicate that they may be interneurons (30). The other three cell types were much larger in dendritic spread and received contacts from most of the axons traversing ICP. It seems likely that the larger neurons of ICP with their numerous synaptic contacts were responsible for the broadly tuned discharges of the present study. Similar neurons of ICX (28) have equally large dendritic fields, and our results suggest that some units in this region are also broadly tuned. Both unit behavior and cell morphology are strikingly similar to observations made in the medial division of MGB 0 The source of afferent input to the three regions of inferior colliculus may explain differences in their binaural properties. The high proportion of binaural units in both ICC and ICX is compatible with the observations that neurons in ICC receive a strong projection from lower binaural structures (35) and that these neurons may, in turn, project to ICX (29). The preponderance of monaural responses in ICP may be indicative of a nonlemniscal projection and correlates with difficulties observed in anatomical studies in identifying the afferent auditory projection to this region (30). It is likely that some auditory input

VEALE,

AND

CROSBY

to ICP descends from the auditory cortex, since a projection exists in this direction (30) . Signijicance of “sharp” “broad” tuning

and

It has long been considered that auditory neurons respond to more restricted ranges of sound frequency at central than at peripheral levels of the auditory pathway. Katsuki and his colleagues ( 16, 17) demonstrated that tuning curves measured for units in the cochlear nerve were broader than those in the inferior colliculus and medial geniculate body, and comparable observations were made by Erulkar (8) for inferior colliculus. While there is some debate as to whether the medial geniculate body represents the locus of greatest tuning sensitivity for units in the auditory pathway (4, 17, 36), there is little doubt, in view of the recent observations of Kiang and Moxon (18) on cochlear nerve fibers that considerable refinement and sharpening occurs in the tuning of many neurons in both the inferior colliculus and medial geniculate body. Comparisons with cochlear nerve measurements suggest that sharp tuning curves may be the manifestation of a loss of the tail of the tuning curve which, for high best-frequency cochlear nerve fibers, extends to very low frequencies ( 18). The existence of both tailed and sharp tuning curves in ICC implies a duality of function in which sharp tuning curves may reflect information relevant to pitch discrimination, while tailed neurons may have other auditory functions. This dichotomy is similar to that proposed by Erulkar (8), but it is not clear whether his broad group A tuning curves were located in the central nucleus or in its surrounding regions. The presence of both types of neuron in a tonotopic array is reasonable if it be assumed that the connections of such cells to the peripheral auditory nuclei differ. For example, fibers reach the central nucleus from both the contralateral, dorsal, and ventral cochlear nuclei as well as from the superior olivary nuclei (25, 35), and it is known that neurons in different parts of the cochlear nuclear

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS

OF

INFERIOR

complex differ in their reception and integration of the discharges of cochlear nerve fibers (10-12, 33). With respect to tuning curves, it is of great interest that units in the dorsal cochlear nucleus may have very broad curves compared with the restricted tuning of units in the ventral cochlear nucleus ( 11). It is likely that the broad tuning curves, qualitatively different from those in ICC and exhibited by most units in ICP and ICX, result from the convergence of many fibers on cells in these regions. Such a convergence may be of fibers individually sensitive to a restricted tonal frequency range, as would be the case for collaterals of lateral lemniscal afferents to ICC, or of axons of neurons which already demonstrate a tendency to respond to a broad frequency range, as do some units in ICC and the dorsal cochlear nucleus (11). A similar suggestion has been made for the source of broad tuning in the medial division of MGB (1). Speculations on functional differences between divisions of inferior colliculus

Our results suggest that different regions of the inferior colliculus may be suited to different functions in audition. The central nucleus is a tightly organized structure in relation to the cochlea, containing elements tuned sharply to different sound frequencies. Most cells are sensitive to binaural stimulation and manv of them may encode, within the spatiotemporal patterns of their discharges, information about sound localization (32). The majority of the axons of the central nucleus project to the ventral nucleus of MGB (22), and from there to cortical area AI (34). These three regions contain neurons with similar properties and organization (4, 20) and should together be regarded as the central core of the upper auditory pathway. The pericentral nucleus, although demonstrating some evidence of tonotopic organization, contains units with such broad tuning that a concept of best frequency is difficult to apply to them. Many units are verv labile and habituate to repeated identical stimuli, but respond read-

COLLICULUS

1205

ily to changes in tonal frequency. Axons of these neurons probably terminate in areas around the ventral division of MGB (1, 24) which, in turn, project on nonprimary auditory cortex (34). Some axons of neurons in ICP also terminate in ICC (30). The overall behavior of units in ICP and their monaural input suggest that cells in this region may subserve a role in auditory attention, especially in view of the observations of Jane, Masterton, and Diamond (14) that damage to the dorsal parts of the inferior colliculus in cats leads to modification of attention to auditory stimuli. Finally, although it has been shown in this and a previous study (31) that units in ICX may respond sensitively to auditory stimuli, the number of penetrations in the present study yielding responses in ICX was small and they were restricted to the rostrolateral section of this subdivision. In view of the existence of connections to ICX from somatosensory regions (13, 15, 19) and to the existence in the former region of cells unresponsive to acoustic stimuli (21), it is probable that the external nucleus of the inferior colliculus is not a primarily auditory structure in the sense that the central nucleus is, but may be related to acousticomotor reflexes (7). The afferent projection of this area to such structures as the pontine nuclei and cerebellum support this contention (26, 27). SUMMARY

The responses of 150 units in the central (ICC), pericentral (ICP), and external nuclei (ICX) of the inferior colliculus of the anesthetized cat were studied in relation to their tuning characteristics and binaural responses to tonal stimuli. Units in ICC were characterized by sharp tuning and binaural responses, while those in ICP and ICX were frequently very broadly tuned with a poorly defined best frequency. Nonetheless, in the latter nuclei a tendency existed for tonotopic organization to occur with high frequencies located externally and low frequencies at the margins of the central nucleus. Tuning measurements were hampered by the common occurrence of

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

1206

AITKIN,

WEBSTER,

habituation in the discharges of single units in ICP and, to a lesser extent, ICX. The majority of units in ICP could be differentiated from those in ICX by their monaural input. Speculations were advanced linking anatomical cell types to physiological responses in the three nuclei and into the possible functional significance of the different behavior of units to tonal stimuli.

VEALE,

AND

CROSBY

ACKNOWLEDGMENTS The authors thank Debbi Spain, Di Harrison, Moyra Farrington, Jill Maplesden, and Louise Yeo for secretarial and technical assistance. This study was supported by grants from the Australian Research Grants Committee.

Present

Physiology Adelaide,

address

and

Adelaide,

of J. L. Veale:

Dept. of Human University of Alstralia.

Pharmacology S.A.

5000,

REFERENCES 1. AITKIN, L. M. Medial geniculate body of the cat: responses to tonal stimuli of neurons in the medial division. J. Neurofhysiol. 36: 275283, 1973. L. M., FRYMAN, S., BLAKE, D. W., AND 2. AITKIN, WEBSTER, W. R. Responses of neurons in the rabbit inferior colliculus. I. Frequency specificity and topographic arrangement. Brain Res. 47: 77-90, 1972. 3. AITKIN, L. M. AND PRAIN, S. M. Medial geniculate body: unit responses in the awake cat. J. Neurophysiol. 37: 5 12-52 1, I974. 4. AITKIN, L. M. AND WEBSTER, W. R. Medial geniculate body of the cat: organization and responses to tonal stimuli of neurons in the ventral division. J. Neurophysiol. 35: 365-380, 1972. 5. BERMAN, A. L. The Brain Stem of the Cat. Madison: Univ. of Wisconsin Press, 1968. 6. BOCK, G. R., WEBSTER, W. R., AND AITKIN, L. M. Discharge patterns of single units in inferior colliculus of alert cat. J. Neurophysiol. 35: 265-277, 1972. 7. BUSER, P., ST. LAURENT, J., AND MENINI, C. H. Intervention du colliculus infirieur dans l’klaboration et le controle cortical spircifique des dkcharges clonique au son chez le chat sur chloralose. Exptl. Brain Res. 1: 102-126, 1966. 8. ERULKAR, S. D. The responses of single units in the inferior colliculus of the cat to acoustic stimulation. Proc. Roy. Sot., London, Ser. B 150: 336355, 1959. 9. GENIEC, P. AND MOREST, D. K. The neuronal architecture of the human posterior colliculus. Acta Oto-Laryngol. Suppl. 295, 197 1. 10. GODFREY, D. Localization of Single Units in the Cochlear Nucleus of the Cat: An Attempt to Correlate Neuronal Structure and Function (doctoral dissertation). Cambridge: Harvard University, 1971. 11. GOLDBERG, J. M. AND BROWNELL, W. E. Discharge characteristics of neurons in anteroventral and dorsal cochlear nuclei of cat. Brain Res. 64: 35-54, 1973. D. D. AND MARUYAMA, N. Excitatory 12. GREENWOOD, and inhibitory response areas of auditory neurons in the cochlear nucleus. J. Neurophysiol. 28: 863-892, 1965. 13. HAND, P. AND LIU, C. N. Efferent projections of the nucleus gracilis. Anat. Record 154: 353354, 1966.

J. A., MASTERTON, R. B., AND DIAMOND, I. T. The function of the tectum for attention to auditory stimuli in the cat. J. Corn@ Neurol. 125: 165-192, 1965. 15. JANE, J. A. AND SCHROEDER, D. M. A comparision of dorsal column nuclei and spinal afferents in the European hedgehog (Erinuceus [email protected]). Exptl. Neurol. 30: I-17, 1971. 16. KATSUKI, Y., SUMI, T., UCHIYAMA, H., AND WATANABE, T. Electric responses of auditory neurons in cat to sound stimulation. J. Neurophysiol. 2 1: 569-588, 1958. Y., WATANABE, T., AND MARUYAMA, N. 17. KATSUKI, Activity of auditory neurons in the upper levels of brain of cat. J. Neurophysiol. 22: 343-359, 1959. E. C. Tails of tuning 18. KI~NG, N. Y.-S. AND MOXON, curves of auditory-nerve fibers. J. Acoust. Sot. Am. 55: 620-630, 1974. 19. MEHLER, W. R., FEFFERMAN, M. E., AND NAUTA, W. J. H. Ascending axon degeneration following anterolateral chordotomy in the monkey. Anat. Record. 124: 332-333, 1956. 20. MERZENICH, M. M., KNIGHT, P. L., AND ROTH, G. L. Representation of cochlea within primary auditory cortex in the cat. J. Neurophysiol. 38: 231249, 1975. 21. MERZENICH, M. M. AND REID, M. D. Representation of the cochlea within the inferior colliculus of the cat. Brain Res. 77: 397-415, 1974. R. Y. AND GOLDBERG, J. M. Ascending 22. MOORE, projections of the inferior colliculus in the cat. J. Camp. Neurol. 121: 109-136, 1963. 23. MOREST, D. K. The laminar structure of the inferior colliculus of the cat. Anat. Record 148: 314, 1964. 24. MOREST, D. K. The cortical structure of the inferior quadrigeminal lamina of the cat. Anat. Record 154: 389-390, 1966. 25. OSEN, K. K. Projection of the cochlear nuclei on the inferior colliculus in the cat. J. Camp. Neurol. 144: 355-372, 1972. 26. POWELL, E. W. AND HATTON, J. B. Projections of the inferior colliculus in cat. J. Camp. Neurol. 136: 183-192, 1969. G. L. Anatomic relationships of the 27. RASMUSSEN, ascending and descending auditory systems. In: Neurological Aspects of Auditory and Vestibular Disorders, edited by W. S. Fields and B. R. Alford. Springfield: Thomas, 1964, p. 1-19. 14.

JANE,

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.

SUBDIVISIONS 28.

29.

30.

31.

32.

33.

OF

INFERIOR

ROCKEL, A. J. Observations on the inferior colliculus of the adult cat, stained by the Golgi technique. Brain Res. 30: 407-410, 1971. ROCKEL, A. J. AND JONES, E. G. The neuronal organization of the inferior colliculus of the adult cat. I. The central nuc1eus.J. Camp. Neural. 147: 1 l-60, 1973. ROCKEL, A. J. AND JONES, E. G. The neuronal organization of the inferior colliculus of the adult cat. II. The pericentral nuc1eus.J. Camp. Neurol. 149: 301-334, 1973. ROSE, J. E., GREENWOOD, D. D., GOLDBERG, J. M., AND HIND, J. E. Some discharge characteristics of single neurons in the inferior colliculus of the cat. I. Tonotopical organization, relation of spike counts to tone intensity, and firing patterns of single elements. J. Neurophysiol. 26: 294-320, 1963. ROSE, J. E., GROSS, N. B., GEISLER, C. D., AND HIND, J. E. Some neural mechanisms in the inferior colliculus of the cat which may be relevant to localization of a sound s0urce.J. Neurophysiol. 29: 288-314, 1966. ROSE, J. E., KITZES, L. M., GIBSON, M. M., AND HIND, J. E. Observations on phase-sensitive

34.

35.

36.

37.

38.

39.

COLLICULUS

1207

neurons of anteroventral cochlear nucleus of the cat: nonlinearity of cochlear output. J. Neurophysiol. 37: 2 18-253, 1974. SOUSA-PINTO, A. Cortical projections of the medial geniculate body in the cat. Advan. Anat. Embryol. Cell Biol. 48: l-42, 1973. VAN NOORT, J. The Structure and Connections of the Inferior Colliculus. An Investigation of the Lower Auditory System. Leiden: van Gorcum, 1969, p. l-l 18. WATANABE, T. AND KATSUKI, Y. Response patterns of single auditory neurons of the cat to species-specific vocalization, Japan. J. Physiol. 24: 135-155, 1974. WEBSTER, W. R. AND AITKIN, L. M. Central auditory’ processing. In: Handbook of Psychobiology, edited by M. Gazzaniga and C. Blakemore. New York: Academic, 1975, p. 325-364. WEBSTER, W. R. AND VEALE, J. L. Binaural response patterns in the inferior colliculus. Proc. Australian WEBSTER,

Physiol.

Pharmacol.

Sot.

1: 60-6

1, 1970.

W. R. AND VEALE, J. L. Patterns of discharge of cat inferior colliculus

binaural units. Proc. Australian 84-85, 1971.

Physiol.

Pharmacol.

Sot.

2:

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (138.026.031.003) on August 21, 2018 Copyright © 1975 American Physiological Society. All rights reserved.