efferent terminals in the cochlea of the mustached bat: quantitative data

Hearing Research, 66 (1993) 81-90

% 1993 Elsevier Science Publishers B.V. All rights reserved 0378-S95S/93/$n6.00

Xl

HEARES 0188 1

Efferent terminals in the cochlea of the mustached bat: Quantitative data

D.H. Xie ‘, M.M. Henson b, A.L. Bishop ’ and O.W. Henson, Jr. ’ ‘I Department of Cell Biology and Anatomy and ’ Department of Surgery, Dic~ision of Otolaryngolo~y / Head and Neck Surgery,

The ~~il,t~rsi~ of North Carolina at Chapel Hill? Chapel HiK sorrel Carolina, USA and “ Respirat#~ Clinical Re.~earch~ Glaxo, inc., Research Triangle Park, North Carolina, USA

(Received 29 June 1992; Revision received 29 October 1992; Accepted 13 November 1992)

Efferent terminals in the cochlea of the mustached bat were stained for acetylcholinesterase (AChE) and quantitative data were obtained for

the number and size of the endings on the outer hair cells (OHCsI in each row, from base to apex. From TEM micrographs and AChE-stained,

surface preparations it was determined that every OHC had a single, large terminal. The mean size of the terminals was signifi~ntly different in

each row, with the largest occurring in the first row (7.1 pm*); the mean size in the second and third rows was 5.7 and 5.0 pm’ respectively. In

specific frequency processing regions, the largest mean size (8.4 pm*) for first row OHCs was consistently found in the distal densely innervated

(DDI) area. This region has afferent neurons that are sharply tuned to the second harmonic, constant frequency component of the bat’s biosonar

signals. Sudden changes in the size of the terminals were observed exactly at the boundaries of the DDI with adjacent sparsely innervated

regions. Similar, but less striking, size changes also occurred in and adjacent to the proximal densely innervated (PDI) region, a harmonically

related, sharply tuned region, which processes the bat’s 91.5 kHz, third harmonic, constant frequency signals. The region of the cochlea with the

smallest first row terminals (mean 5.3 pm*) was the large, sparsely innervated region of the basal turn, a region that does not appear to process

biosonar signals. Although the significance of differences in efferent terminal size is not known, the data suggest a possible correlation between

OHC stimulation and sharp tuning. The potentially greater influence of the efferent fibers on the first row of OHCs. compared to other rows, is

consistent with observations made on other mammals; in the latter, however, the greater influence has been suggested more by number than size.

Unlike other mammals, the OHC efferents in the mustached bat have no clear base-to-apex gradient in the number or size of the efferent

terminals. It is suggested that this might reflect the high frequency nature of the ear 16-120 kHz) and absence of low frequency hearing.

Cochlea; Efferent; Olivocochlear; Hair cell; Bat

Introduction

The cochlea of the mustached bat (P~er~n~~u~ parnellii) shows a number of unusual anatomical and physiological specializations for the detection and analysis of constant frequency (CF) biosonar signals (Henson and Henson, 1991; Henson et al., 1985; Kiissl and Vater, 1985; 1990; Pollak and Bodenhamer, 1981; Pol- lak et al., 1972; Suga et al., 1975; Suga and Jen, 1977; Zook and Leake, 1989). Sharp tuning is especially evident in two narrow bands centered around 60 and 90 kHz. These bands correspond to the prominent second and third harmonics of the bat’s CF echolocation calls. As shown in Fig. 1, the cochlea shows striking differences in afferent innervation density in different parts of the basal turn. Near the basal end is a proximal densely innervated (PDI) region and at the

Correspondence to: O.W. Henson, Jr., Department of Cell Biology

and Anatomy, 108 Taylor Hall, CB No. 7090, University of North

Carolina at Chapel Hill, Chapel Hill, NC 27599, USA. Fax: (919) 966-1856.

other end of the basal turn is a distal densely innervated (DDI) region. Between these two densely innervated regions is a sparsely innervated W-1) area that terminates apically as a straight region (SR) (Henson, 1973; Henson and Henson, 1991). Hair cells and neurons in the PDI are responsible for detecting the +91.5 kHz, third harmonic, CF and those in the DDI, the f61 kHz second harmonic, CF. The intervening SI-1 region processes sound in the 75-64 kHz band, and it does not appear to be related to any of the animal’s biosonar signal components. In the second turn, there is a region just apical to the SI-2 region which may show an increase in density of afferent fibers and where other specializations are sometimes found (Henson and Henson, 1991; K&l and Vater, 1985; Zook and Leake, 1989). There is a peak in neural density in the second turn which appears to represent an FM region of the second harmonic and another peak in neural density occurs in the apex. The apical region appears to contain neurons that respond to both the CF and FM components of the first harmonic of the bat’s biosonar signais. Physiological studies of afferent neurons associated with the densefy innervated

regions of the cochlea have shown that the C‘F units are very sharply tuned, with (_I,,, dB values commonly above 60 and sometimes on the order of 200-400; by contrast, neurons in other regions are not as sharply tuned, as has been well demonstrated by a number of

BASALTURN SECONDTURN

B -- PDI I l

DDI

0 20 40 60 00 100

BASE COCHLEAR LENGTH (X) APEX

Fig. 1. The regions of the cochlea of the mustached bat. Scanned images of sudan black B-stained, surface preparations of the basal and second turns are shown in (A); a graphic display of the nerve fiber density as a function of basilar membrane length is shown in (B) and frequency map data are shown in (0. In A, the large basal turn is characterized by the PDI, SI-1, SR and DDI regions. The second turn shows additional, less pronounced, sparsely and densely innervated regions; here the sparsely innervated regions are desig- nated SI-2 and SI-3. The graph showing neural density has peaks which correspond to the nerve densities shown in (A). From the frequency map data in (C), it is evident that the PDI and DDI regions are associated with the reception of second and third harmonic, CF, biosonar signal components (H2-CF; H3-CF). Note that the SI-1 region appears to be associated with frequencies between about 75 and 63 kHz. Harmonic FM regions are also shown (HS-FM and H2-FM. (A and C, after Henson and Henson, 1991; B, redrawn from Zook and Leake, 1989). The frequency map data represented by solid symbols are from Zook and Leake (1989) and the open

symbols are from K&s1 and Vater (1985).

studies (Suga et al., 1975; Suga and Jcn. 1077; Kiissl and Vater, 1985; Huffman and Hcnson, 1902). Exam- ples of representative, broadly tuned and sharply tuned curves for single units are shown in Fig. 2.

The mustached bat’s olivocochlear system is like that of most other mammals (Bishop and Henson. 1987) in that there are medial and lateral subsystems (see Warr, 1992). Previous studies on the olivocochlear system in the mustached bat have suggested that each OHC has one large efferent terminal (Bishop and Henson, 19881, but complete base to apex analyses were not made.

The purpose of this study was to examine and quan- tify the efferent endings on the OHCs throughout the cochlea. Special attention was directed to gradients in size among the rows of OHCs, and to characteristics of efferent terminals in sharply tuned vs broadly tuned regions.

Methods

The animals used in this study were mustached bats, Pteronotus p, parnellii, from Jamaica, WI. For tissue fixation, animals were deeply anesthetized with methoxyflurane (Metofane, Pitman-Moore, Inc.) and killed by decapitation; the heads were cut in the mid- sagittal plane, the cochleae rapidly removed and washed in 0.9% NaCl. The stapes and membrane cov- ering the round window were removed and the tissue was placed in cold (4°C) fixative for 8-12 h. The latter consisted of 4% paraformaldehyde, 0.5% glutaralde- hyde, and 0.2% picric acid in 0.1 M phosphate buffer, pH 7.4. After fixation, cochleae were decalcified in 0.1 M EDTA in 0.1 M phosphate buffer, pH 7.4 for 2-4 days in the cold. The EDTA solution was changed each day. The cochleae were then cut into four segments (hook, basal turn, second turn, apex), stained for acetylcholinesterase and viewed as surface preparations.

All steps in the staining procedure were carried out at room temperature and followed the method of Tago et al. (1986). Tissues were first washed in 0.1 M phosphate buffer three times, 5 min each. Next, they were treated with 0.1% H,O, for 30 min to destroy endoge- nous peroxidase and reduce background staining. This was followed by washing in 0.1 M phosphate buffer, three times, 5 min each, and then treating with 1% Triton X 100 for 1 h. After another wash in phosphate buffer (three times, 5 min each) the tissue was incubated for 30 min in a solution containing 35 /IM acetylthiocholine iodide, 5 ,uM K,Fe(CN),, 30 WM CuSO,, and 50 PM sodium citrate in 0.1 M maleate buffer, pH 6.0. Following another series of washes, with five changes of 50 mM Tris-HCl, pH 8.2, 5 min each, the tissue was incubated for 5 min in a solution

FKEQUENCY (KHZ)

Fig. 2. Examples af tuning curves OF broadly and sharply tuned l-ochlear lrucleus units in the mustached bat. Sharply tuned (high Q,,, dB) units have best frequencies at or near hl and Yl.5 kIIz, and they process the second and third harmonic, CF cnmponcnts of the bat’s biosunar signals. The high-Q, 61 kIJz unirs are typical& tuned in a narrolv band near the rescmancc frequency of the cochlea (CRFX Figure courtesy of Rusrrll

Huffman (unpublished data ).

containing O&3% 3’3’ diaminobenzidine and 0.3% with 0.003% H,O, for lfl-12 min. Finafly, the tissue nickel ammonium sulfate in 50 mM Tris-HCI, pH 8.2. was washed with three changes of 5 mM Tris-HCI, pH The tissue was then incubated in the same solution X.2, 5 min each and then dehydrated with a graded

Fig. 3. Photomicrographs (A and B) and drawings (C and 0) ot AthE stained surface preparations. (A) shows numernlls, efferent termlnals in the ST. 1 region. iB1 is a higher magnification of terminals In the DDI region. (C? and (D) are filled contours of images rraccd with the aid of a drawing t&e. Note the differences in the size and spacing af the terminals &I the S-1 CC) and DDI Of regions. The bar in CA) = 20 pm: bar in

fB)= 10 pm. Arruws in iA) and (Bj point to the first row of OHCs. The bar in CD1 also applies to tC>.

x4

ethanol series, cleared with xylene, and coverslipped with DPX. The following points were especially impor- tant for good results: 1) rapid removal of cochleae and immediate immersion in cold fixative; 2) maintaining a high pH (8.2) with Tris-HCI; and 3) treatment with 1% Triton X 100 prior to histochemical staining. Consis- tent reactions were obtained when the solutions containing the tissue were constantly, vigorously agitated.

To establish that AChE staining was concentrated in the large efferent terminals, the crossed olivocochlear bundle was sectioned in the floor of the fourth ventri- cle; the animals were allowed to survive for 12 to 37 days. Subsequent examination of the cochleae revealed complete degeneration of about 60% of the large OHC terminals. Small, approx. 1.0 pm, AChE positive terminals were evident on OHCs where the large terminals had degenerated, but these could clearly be identified as type II afferent terminals. Details concerning the lesioning technique and the distribution of crossed vs uncrossed terminals and type II afferent fibers will be the subject of companion papers (in preparation). Transmission electron microscopy was also used to

Fig. 4. Transmission electron micrograph showing the single, large, efferent terminal associated with the OHCs of the mustached bat. The large efferent terminal (E) typically occupies the center of the neural pole of the OHC and the small, afferent terminals (* ) form a

ring around the efferent ending. Bar = 2.0 pm.

PDI SI DDI 2ND APEX

5: B - 121 ---- cn Pfil 243 1

P lo ; 6 6

g 4 2

5 0 PDI SI DDI 2ND APEX

C 4 12

P62 la9T ___ 64

PDI SI DDI 2ND APEX

REGION OF COCHLEA

Fig. 5. Area of efferent terminals on outer hair cells as a function of distance along the basilar membrane. Each data point represents the average size of the terminals over a distance of 100 pm. Note: 1) no clear base to apex gradient; 2) the mean large size of the terminals in the first row compared to the second and third rows; and 3) the large size of the first row terminals in the DDI region, centered about

50-60% of the length along the cochlear duct.

verify the number and size of the efferent terminals on OHCs. Techniques used for electron microscopy have been discussed in previous reports from our laboratory (Henson et al., 1982; 1983; 1984).

Drawings of efferent endings in surface preparations were made with a drawing tube attached to a microscope. Terminals were viewed under high magnification (X 1000); an outline of each image was drawn on paper and then retraced onto a data tablet. The area of each ending was determined with a morphome- tric analysis system (Eutectics Electronics VDP3 with DEC PDP-11 interface).

ReSUltS

Charactetitics of OHC efferent terminals When the cochleae were stained for AChE, a single,

large, circular-shaped mass of reaction product ap- peared at the base of each OHC (Fig. 3A,B). This was consistent from row to row and base to apex. The large size, position and vessiculated nature of these terminals was confirmed by transmission electron microscopy (Fig. 4); micrographs of sections near the central part of the OHC base were obtained from the

TABLE I

Areas of efferent terminals by region (+m*)

PDI SI DDI 2nd turn Apex Total

1st row 6.87 f 0.98 (418) 6.21 f 0.62 (873) 8.44* 1.18 (877) 7.09 f 0.28 (238) 6.68+ 1.28 (152) 7.06 + 0.86 (2558)

2nd row 5.53 + 0.60 (423) 5.06 + 0.48 (858) 6.50 f 0.87 (926) 6.25 k 0.65 (233) 5.34+ 1.07 (152) 5.74 f 0.73 (2592)

3rd row 4.84+0.60 (411) 4.70 kO.55 (813) 5.55 * 0.72 (880) 5.2OkO.32 (197) 4.67 k 0.93 (147) 5.00 i 0.62 (2448)

Total 5.75 + 0.72 (1252) 5.32 k 0.55 (2544) 6.83 f 0.92 (2683) 6.18 k 0.42 (668) 5.56+ 1.09 (451) 5.93 f 0.74 (7598)

Values are means? SD for (N); Number of endings measured are shown in parentheses; P < 0.005 for 2nd row vs 1st row in the same region;

P < 0.01 for 3rd row vs 2nd row in the same region; P < 0.025 for each region vs DDI in the same row.

hook, PDI, Sl-1, DDI and apex. A minimum of 15 part of the neural pole of each OHC. Afferent termi- OHCs were examined with TEM in each area. In all nals, by contrast, were small (approx. 1.0 pm diameter) cases, large efferent terminals occupied the central and occupied a peripheral position (Fig. 4).

PDI SI

3 4 5 6 7 8 9 1011121314 3 4 5 6 7 6 9 1011121314

DDI

AREA OF ENDING (SQqM) AREA OF ENDING (SQ.pM)

APEX 60

50 ROW,

40

3 4 5 6 7 8 9 1011121314 3 4 5 5 7 6 9 1011121314

ROW3

AREA OF ENDING (SQqM) AREA OF ENDING (SQ+M)

Fig. 6. Bar graph showing terminal size in three preparations (P52, P61 and P62) where measurements were made in the five major regions of the

cochlea. Each bar on the left (black) shows area measurements for the first row terminals; the next two bars show the data for the second and

third rows, respectively. The error bars represent standard deviations and the numbers above each bar show the number of measurements made.

Note the progressive change in size from row one to three and the large size of the terminals in the first row of the DDI region.

Terminal size The areas of 7,598 single efferent terminals were

measured in the cochleae of five animals. When the data on terminal size were pooled and examined as a function of cochlear length (Fig. 51, or organized according to specialized cochlear regions (Table I; Fig. 61, it was clear that: 1) the area of the majority of the terminals was between 4.7 and 7.1 pm* (mean 5.9 pm’): 2) there was a clear row-to-row size gradient; 3) there was no base-to-apex size gradient; and 41 many large terminals were found in the DDI and many small terminals occurred in the SI-1 region.

Data in Table I and in Fig. 6 show that the mean size of terminals associated with the first row of OHCs is consistently larger than that for the second, and the size of the second row terminals is larger than in the third row. The standard deviations shown in Table I emphasize that there is a significant range in size for the endings in each row. Also, as indicated in Fig. 3C and 3D and the histograms (Fig. 6), there is a broadly overlapping distribution of the size of the terminals in any hair cell row, i.e. at any given point in the cochlea, a terminal in the second or third row may be as large or larger than one in the first row. Data for the mean values, however, clearly show a row-to-row size gradient.

The marked difference in size between the terminals in the SI-1 and DDI regions was not only apparent from the tabulated and graphic data, but from visual inspection of the AChE-stained terminals as well (Fig. 3). In these figures, it is evident that the terminals are not only larger in the DDI than in the SI-1 region, but the entire OHC region is very different in width, the latter being markedly wider in the DDI region.

For different frequency processing regions, the most striking finding was the large size of terminals on the first row of OHCs in the DDI (Table I, mean = 8.4 pm*) compared with first row terminals in other areas (6.2-7.1 pm2). The terminals in the first row average about 30% larger in the DDI than in the adjacent $9-1 region; the terminals in the second row are about 23% larger and in the third row only about 16% larger than in the adjacent sparsely innervated region.

Fig. 6 shows data for three animals where terminal sizes in each area were measured in all segments of the cochlea. These figures again emphasize the consistent occurrence of large mean values for terminals in the first row and the large value for first row terminals in the DDI. They also show that the row-to-row gradient occurs in each area. Here, as in Fig. 5, it is obvious that there is no distinct base-to-apex gradient in terminal size. Analyses (t-tests) were carried out to determine if there were statistical differences between the size of the terminals in the different rows. For each region, the size of the terminals in rows 1, 2 and 3 were statistically different from each other. The large first

__ REGION OF COCHLEA

lo TXKCTI~N JUNCTION

4

REGION OF COCHLEA

Fig. 7. Area of efferent terminals on the first row of outer hair cells in, and adjacent to, the DDI region (upper graph) and PDI region (lower graph). Data are for a single preparation. Each data point and the standard deviation bars in the upper graph represents the mean for 10 consecutive measurements; in the lower graph each point is for 20 consecutive measurements. The arrows represent the position of the junctions with adjacent regions. Note how size changes corre-

spond to the position of the junctions.

row terminals in the DDI were also statistically different from those in all other regions (Table I>.

As previously noted, there are several areas in the cochlea where the exact extent of specialized regions can be seen in surface preparations and where boundaries between sparsely and densely innervated regions are sharp Gee Fig. 1). When the size of the terminals in the first row was measured within, and immediately adjacent to, these regions, it was clear that abrupt changes in the mean size occurred exactly at the junctions between regions, and large terminals were pri- marily confined to the densely innervated regions (Fig. 7). This was especially well marked in relation to the DDI region. For example, one transition zone is represented by the SR and contains only 80 first row OHCs. The plot of data for the PDI region and its junction with the SI-1 re,gion sometimes showed a similar trend. (Fig. 7). Boundaries between sparsely and densely innervated regions in the second and apical turns were not always clear and size changes were more difficult to analyze than in the regions in the large basal turn. There were, however, indications that large terminals were also more characteristic of densely innervated

x7

than sparsely innervated regions in these turns. In one case, the terminals near the beginning of the SI-2 region of the second turn had a mean size of about 6.5 wrnZ while the mean size for the first row endings in the middle part of the second turn in the same animal was about 8 grn’.

Discussion

In this study, efferent terminal size was assessed from surface preparations. There would appear to be considerable potential for variation in the profiles determined by this technique compared to the actual volume of the endings. This has been recognized as a possible problem in previous studies on cats (Liberman et al., 1990) and guinea pigs (Brown, 1987); however, spot checks in different regions with TEM imaging compared to the size evaluated from antisynapto- physin-treated surface mounts revealed a close similar- ity (Liberman et al., 1990). Therefore it seems likely that the methods used in this study provide a reason- able technique for studying the size of terminals.

The main points we wish to emphasize are that the cochlea of the mustached bat has: 31 a single efferent ending on each OHC in all three rows from base to apex; 2) a row by row gradient in the size of terminals with a notable concentration of the largest terminals on the OHCs of the first row; and 3) a significantly greater population of large terminals in the densely innervated regions. As previously noted (see Fig. I), the latter are areas of the cochlea known to be associated with sharp neural tuning, fine frequency resolution and the perception of the constant frequency components of the animal’s biosonar signals. The smallest terminals occur in the SI-1 region which does not process frequencies within the animal’s biosonar signals.

The presence of a single efferent ending on each OHC from base to apex has not been reported for the cochleae of other species. In some species, such as the chinchilla (Iurato et al., 1978) and mouse (Wilson et al., 1991), a single terminal on one OHC is often encountered, but this is not consistent and there is a longitudinal gradient in the innervation pattern. The basic mammalian plan, as represented by cats (Liber- man and Brown, 1986; Liberman et al., 1990; Ischii and Balogh, 1968; Fex and Altschuler, 1981; Ginsburg and Morest. 19841, guinea pigs (Hashimoto and Kimura, 1987; Takasaka and Shinkawa, 1987; Altschuler et al., 1985; Warr and Guinan, 1979; Brown, 1987; Smith, 197S), rats (Dannhof et al., 1991; Fex and Altschuler, 1984; Godfrey and Ross, 1985; Roth et al., 19911, hamsters (Simmons et al., 1990) and primates (Bodian, 1983; Bodian and Gucer. 1980; Ishii et al., 1967a, b; Nakai and Igarashi, 1974; Smith. 19731 is that there are

many terminals on the outer hair cells in the basal and middle turns of the cochlea and few, if any, in the apical part. In general, the more sensitive regions of the ear have more efferent fibers and terminals than the less sensitive regions. The number of terminals per OHC has been reported to range from 2-S in the basal end of the guinea pig cochlea, and from 6-13 in the third turn (Robertson, 1984; Robertson and Gummer, 1985; Hashimoto and Kimura, 1987). In cats, Liberman et al. (1990) found a maximum of nine terminals per OHC and Spoendlin and Gacek (19631 reported 6-10 in the basal turn. In cases where an OHC has many efferent terminals, it appears that some. perhaps most, are derived from different fibers, although one fiber may have more than one ending on the same OHC. The total size of the terminals associated with a given

OHC may be more than 50 ym’ in the cat (Liberman et al., 1990) and thus the area of the terminals is much more extensive than that seen in the mustached bat.

The mustached bat, like the old world horseshoe bats (Rhinolophidae) has been of special interest bc- cause of the complex nature of their sonar signals. The sonar signals emitted by representatives of these two groups of bats are characterized by a combination of CF and FM components: in both groups the CF component and associated sharp neural tuning allow the animals to detect small frequency differences and to change the frequency of their pulses in such a way that the echoes of the CF component are maintained in a narrow frequency band. This process is well known as Doppler-shift compensation (Schnitzler, 197Oa,bl. The cochleae in all of these bats show marked changes in afferent innervation density, yet the effcrcnt distribution to the OHCs is remarkably different in the two groups and is clearly very different from that in other mammals. The rhinolophids are unique in having no OHC efferents in any part of the cochlea (Bishop and Henson. 1988; Bruns and Schmieszek, 1980; Firbas and Welleschick, 1970); the mustached bat is unique in having only one efferent terminal on each OHC throughout the cochlea.

It is of interest that the absence of efferent terminals in rhinolophids, and their apparent reduction in the mustached bat, occurs throughout the cochlea, and not simply in those areas (e.g. PDI and DDI) that have been identified as ‘acoustic foveae’. With such sharply tuned systems, one can speculate that if efferent nerves were retained in some areas, and lost in others, changes in cochlear micromechanics might be far reaching and not limited to the areas with or without efferent con- trol. Detailed studies of OHC efferent terminals and gradients in other species of bats are wanting and a comparison of the terminals in the more common FM bats with CF-FM bats is necessary before an adequate comparison of the system can be fully discussed. Of special interest would be the size, number and arrange-

ment of terminals in the frog-eating bat, Truchops cirrhosus, which unlike other Microchiroptera, has a specialized apical region of the cochlea for low frequency (frog-call) perception and a basal, ultrasonic segment for biosonar signal detection (Bruns et al., 1989). The representation of the OHC efferents in the apex of the cochlea has long been of interest and it is generally held that the signal processing at low frequencies may be different from that at high frequencies. According to Liberman et al. (1990) this change occurs at about the 3 kHz region in the cat cochlea. In the part of the cochlea that responds to lower frequencies, the efferent nerve population is greatly reduced, the terminals may not occupy the usual basal position, and neural transmitters appear to be different (see Liberman et al., 1990; Warr, 1992). Since hearing in the mustached bat covers frequencies extending from 6 to 10 kHz up to 120 kHz, it may be that the absence of typical mammalian changes in the apex reflects an absence of low frequency hearing.

In the common mammalian plan there is not only a base to apex gradient, but also a row by row gradient in the number of efferent terminals on OHCs; this has been noted by many investigators (Churchill and Schuknecht, 1959; Fex and Altschuler, 1981; Hashimoto and Kimura, 1987; Ishii et al., 1967a,b; Ginsberg and Morest, 1984; Guinan et al., 1984; Warr and Guinan, 1979; Simmons et al., 1990; Takasaka and Shinkawa, 1987; Fex et al., 1982; Altschuler et al., 1985). The typical pattern is that the first row of OHCs has a larger number of terminals and/or larger terminals than the second row. In the same manner, the second row is more heavily innervated than the third row. This pattern is evident in the basal and middle turns of the cochlea but not in the apex, where the innervation becomes irregular or disappears. Even in the apex, it is the first row that persists the furthest. The fact that the cochlea of the mustached bat shows a terminal size gradient rather than a population gradient suggests that the gradient seen in most mammals is not simply one that may have been determined by developmental factors, i.e. that the first row is the one to be invaded first by the growth cones (Simmons et al., 1990) of the developing efferent fibers and thus the one that becomes the most heavily innervated. It seems rather to suggest different levels of synaptic activity between the first row compared to rows two and three. This idea has been expressed in some of the early observations on efferent terminal gradients (Smith and Sjostrand, 1961). The large size of the terminals of the first row suggests a larger reservoir of synaptic vesicles and perhaps the possibility of more stable responses. The greater degree of innervation of first row OHCs has been of special interest in view of evidence that the first row contributes more to the establishment of neural tuning curve tips of the primary afferent fibers

than do the OHCs of other rows (Liberman and Dodds, 1984). In this respect it is also of interest to mention the cochlear amplifier and related theories that suggest a resonance segment of the cochlea exists just basal to a region of sharp tuning (Davis, 1983; Neely and Kim, 1983, 1986). At least in the mustached bat, it seems clear that the changes in size of the efferent terminals correspond in position to the boundaries of the sharply tuned DDI region; there is no obvious basal displace- ment (Fig. 7).

The association of a greater influence of efferent innervation with regions of low thresholds has been noted by Weiderhold (19671 and Liberman et al. (19901 for the cat. Coupled with audiometric data, it appears that the efferents have a maximum effect in the most sensitive region of hearing. In the guinea pig it has also been noted that the maximum effect of efferent stimulation is in the most sensitive, 7-10 kHz range (Teas et al., 1972). There is, however, no evidence that efferent activity affects thresholds. There is also no evidence in the mustached bat that the lowest thresholds for sounds occur in the PDI or DDI; in fact, single unit data suggest that units associated with the broadly tuned FM areas of the cochlea have just as low thresholds as units in the CF processing areas (Suga et al., 19751. Judging from efferent terminal size, the data for P. pamellii suggest a relationship of efferent innervation to sharp tuning and frequency resolution. We cannot, however, discount the possibility that the large size of the terminals in the DDI, and the slight tendency for larger terminals in other CF areas, simply reflects a greater degree of afferent activity and reflexive efferent activity, or stimulation of the cochlea by long CF rather than short FM signals.

Acknowledgements

This work was supported by NIH grant DC 00114 from the National Institute on Deafness and Other Communicative Disorders. Critical reading of this manuscript by Art Keating, Russell Huffman and Win- ston Lancaster is greatly appreciated.

References

Altschuler, R.A., Kachar, B., Rubio, J.A., Parakkal, M.H. and Fex, J. (1985) Immunochemical localization of choline acetyltransferase- like immunoreactivity in the guinea pig cochlea. Brain Res. 338, l-11.

Bishop, A.L. and Henson, O.W., Jr. (1987) The efferent cochlear projections of the superior olivary complex in the mustached bat. Hear. Res. 31, 175-182.

Bishop, A.L. and Henson, O.W., Jr. (1988) The efferent auditory system in Doppler-shift compensating bats. In: P.E. Nachtigall

89

and W.B. Moore (Eds.), Animal Sonar, Plenum Publishing Corp.,

pp. 307-310.

Bodian, D. (1983) Electron microscopic atlas of the Simian cochlea.

Hear. Res. 9, 201-246.

Bodian, D. and Gucer, C. (1980) Denervation study of organ of Corti

of old world monkeys. _I. Comp. Neural. 192, 785-796.

Brown, M.C. (19X7) Morphology of labeled efferent fibers in the

guinea pig cochlea. J. Comp. Neurol. 260, 605-610.

Bruns, V. and Schmieszek, E. (1980) Cochlear innervation in the

greater horseshoe bat: Demonstration of an acoustic fovea. Hear.

Res. 3, 27-43.

Bruns, V., Burda, H. and Ryan, M.J. (1989) Ear morphology of the

frog-eating bat, (Truchops cirrhosus, Family Phyllostomidae): Ap-

parent specializations for low-frequency hearing. J. Morph. 199,

103-l 18.

Churchill, J.A. and Schuknecht, H.F. (1959) The relationship of

acetylcholinterase in the cochlea to the olivocochlear bundle.

Henry Ford Hospital Med. Bull. 7, 202-205.

Dannhof, B.J., Roth, B. and Bruns, V. (1991) Anatomical mapping of

choline acetyltransferase (ChATI-like and glutamate decarboxyl-

ase (Gad&like immunoreactivity in outer hair cell efferents in

adult rats. Cell Tissue Res. 266, 89-95.

Davis, H. (1983) An active process in cochlear mechanics. Hear. Res.

9, 79-90.

Fex. J. and Altschuler, R.A. (1981) Enkephalin-like immunoreactivity

of olivocochlear nerve fibers in cochlea of guinea pig and cat.

Proc. Nat. Acad. Sci. 78, 1255-1259.

Fex, J. and Altschuler, R.A. (1984) Glutamic acid decarboxylase

immunoreactivity of olivocochlear neurons in the organ or Corti

of guinea pig and rat. Hear. Res. 15, 123-131.

Fex, J.. Altschuler. R.A.. Wenthold, R.J. and Parakkal, M.H. (1982).

Aspartate aminotransfersae immunoreactivity in cochlea of guinea

pig. Hear. Res. 7, 149-160.

Firbas, W. and Welleschick, B. (1970) Uber die verteilung der

acetylcholinesterase aktivitat im Corti;schen Organ vonn Fled-

ermiiusen. Acta Otolaryngol. 70, 329-335.

Ginzberg, R.D. and Morest, D.K. (1984) Fine structure of cochlear

innervation in the cat. Hear. Res. 14. 109-127.

Godfrey, D. and Ross, C. (1985) Enzymes of acetylcholine metabolism

in the rat cochlea. Ann. Otol. Rhinol. Laryngol. 94, 409-414.

Guinan J.J., Jr., Warr. W.B. and Norris, B.E. (1984) Topographic

organization of the olivocochlear projections from the lateral and

medial zones of the superior olivary complex. J. Comp. Neurol.

226, 21-27.

Hashimoto, S. and Kimura, R.S. (1987) Computer-aided three-di-

mension reconstruction and morphometry of the outer hair cells

of the guinea pig cochlea. Acta Otolaryngol. (Stockh.) 103, 64-74.

Henson. M.M. (1973) Unusual nerve-fiber distribution in the cochlea

of the bat, Pteronotus p. parnellii (Gray). J. Acoust Sot. Am. 53.

1739% 1740.

Henson. M.M. and Henson. O.W., Jr. (1991) Specializations for

sharp tuning in the mustached bat: the tectorial membrane and

spiral limbus. Hear. Res. 56, 122-132.

Henson, M.M., Henson, O.W., Jr. and Jenkins, D.B. (1984) The

attachment of the spiral ligament to the cochlear wail: anchoring

cells and the creation of tension. Hear. Res. 16, 231-242.

Henson, M.M.. Jenkins, D.B. and Henson, O.W., Jr. (1982) The cells

of Boettcher in the bat, Pteronotus p. parnellii. Hear. Res. 7,

91-103.

Henson, M.M., Jenkins, D.B. and Henson, O.W.,Jr. (1983) Susten- tacular cells of the organ of Corti- the tectal cells of the outer

tunnel. Hear. Res. 10, 153-166.

Henson, O.W., Jr., Schuller, G. and Vater, M. (1985) A comparative

study of the physiological properties of the inner ear in Doppler

shift compensating bats (Rhinolophus roui and Pteronotus parnellii). J. Comp. Physiol. A 157, 587-597.

Huffman. R.F. and Henson, O.W., Jr. (1992) Labile Cochlear Tuning

in the Mustached Bat. II. Concomitant Shifts in Neural Tuning.

J. Comp. Physiol. A. (in press).

Ishii, J. and Balogh, K. (1968) Distribution of efferent nerve endings

in the organ of Corti. Acta Otolaryngol. 66, 282-288.

Ishii, T., Murakami, Y. and Balogh, K.F. (1967a) Acetyl-

cholinesterase activity in the efferent nerve fibers of the human

inner ear. Ann. Otol. Rhinol. Laryngol. 76. 69-82.

Ishii, T., Murakami, Y. and Gacek, R.R. (1967b3 Histochemical study

of the acetylcholinesterase activity in the inner ear of the squirrel

monkey. Acta Oto-Laryngol. 64, 267-279.

Iurato, S., Smith, C.A., Eldredge, D.H., Henderson, D., Carr, C.,

Ueno, Y., Cameron, S. and Richter, R. (1978) Distribution of the

crossed olivocochlear bundle in the chinchilla’s cochlea. J. Comp.

Neurol. 182, 57-76.

Kiissl, M. and Vater, M. (1985) The cochlear frequency map of the

mustached bat, Pteronotus parnellii. J. Comp. Physiol. A 157,

687-697.

Kiissl, M. and Vater, M. (1990) Resonance phenomena in the cochlea

of the mustached bat and their contribution to neuronal response

characteristics in the cochlear nucleus. J. Comp. Physiol. A. 166.

71 l-720.

Liberman. M.C. and Dodds, L.W. (1984) Single neuron labeling and

chronic cochlear pathology: III stereocilia damage and alterations

of threshold tuning curves. Hear. Res. 16, 55-74.

Liberman, M.C., Dodds. I.W. and Pierce. S. (1990) Afferent and

efferent innervation of the cat cochlea: Quantitative analysis with

light and electron microscopy. J. Comp. Neural. 301, 443-460.

Liberman, M.C. and Brown, M.C. (1986) Physiology and anatomy of

single olivocochlear neurons in the cat. Hear. Res. 24, 17-36.

Nakai. Y. and Igarashi, M. (1974) Distribution of the crossed olivo-

cochlear bundle terminals in the squirrel monkey cochlea. Acta

Otolaryngol. (Stockh.) 77. 393-404.

Neely, S.T. and Kim, D.O. (1983) An active cochlear model showing

sharp tuning and high sensitivity. Hear. Res. 9, 123-130.

Neely, S. and Kim, D.O. (1986) A model for active elements in

cochlear biomechanics. J. Acoust. Sot. Am. 79. 1472-1480.

Pollak, G.D. and Bodenhamer, R.D. (1981) Specialized characteris-

tics of single units in the inferior colliculus of mustache bats.

Frequency representation, tuning. and discharge patterns, J. Neu-

rophysiol. 46, 605-620.

Pollak, G.D., Henson, O.W., Jr. and Novick. A. (1972) Cochlear

microphonic audiograms in the ‘pure tone’ hat, Chilonycteris parndii. Science 176, 66-68.

Robertson, D. (1984) Horseradish peroxidase injection of physiologi-

cally characterized afferent and efferent neurones in the guinea

pig spiral ganglion. Hear. Res. 15, 113-122.

Robertson, D. and Gummer, M. (1985) Physiological and morpholog-

ical characterization of efferent neurons in the guinea pig cochlea.

Hear. Res. 20, 63-77.

Roth. B., Dannhof, B. and Bruns. V. (1991) ChAT-like immunoactiv-

ity of olivocochlear fibres on rat outer hair cells during postnatal

development. Anat. Embryol. 183, 483-489.

Schnitzler. H.-U. (1970a) Echoortung bie der fledermaus. Chilonyc- trris ruhiginosa. Z. Vergl. Physiol. 68, 25-39.

Schnitzler H.-U. (1970b) Comparison of echolocation behavior in

Rhinolophus ferrumequinum and Chilonycteris ruhiginosa. Bijdr.

Dierk. 40, 77-80.

Simmons, D.D., Manson-Gieseke, L.. Hendrix, T.W. and McCarter,

S. (1990) Reconstructions of efferent fibers in the postnatal

hamster cochlea. Hear. Res. 49, 127-39. Smith, C.A. (1973) The efferent neural supply to the vertebrate ear.

Adv. Oto. Rhino. Laryngol. 20, 296-310.

Smith, CA. (1975) Innervation of the cochlea of the guinea pig by

use of the Golgi stain. Ann. Otol. 84. 443-458.

Smith, C.A. and Sjostrand, S. (1961) Structure of the nerve endings

on the external hair cells of the guinea pig cochlea as studied by

serial sections. Ultrastruct. Res. 5, 523-556.

90

Spoendlin. H. and Gacek, R.R. (1963) Electron microscopic study of

the efferent and afferent innervation of the organ of Corti in the

cat. Ann. Otol. Rhinol. Laryngol. 72, 660-686.

Suga, N., Simmons, J.A. and Jen, P.H-S. (1975) Peripheral special-

izations for fine frequency analysis of Doppler-shifted echoes in

the auditory system of the “CF-FM” bat, Preronot~s pamellii. J.

Exp. Biol. 63, 161-192.

Suga. N. and Jen, P.H-S. (1977) Further studies on the peripheral

auditory system of the CF-FM bats specialized for fine frequency

analysis of Doppler-shifted echoes. J. Exp. Biol. 69, 207-232.

Takasaka, T. and Shinkawa, H. (1987) Serial section reconstruction

of the guinea pig outer hair cells as studied with high voltage

electron microscope and a computer graphic display. Acta Oto-

laryng. (Stockh.) Suppl. 435, 7-20.

Tago, H., Kimura, H. and Maeda, T. (1986) Visualization of detailed

acetylcholinesterase fiber and neuron staining in rat brain by a

sensitive histochemical procedure. J. Histochem-Cytochem. 34,

1431-1438.

Teas, D.C., Konishi, T. and Nielsen, D.W. (1972) Electrophysiologi-

cal studies on the spatial distribution of the crossed olivocochlcar

bundle along the guinea pig cochlea. J. Acoust. Sot. Am. 51.

1256-1264.

Warr, W.B. (1992) Organization of ohvocochlear efferent systems in

mammals. In: D.B. Webster, A.N. Popper and R.R. Fay (Eds.).

The Mammalian Auditory Pathway: Neuroanatomy, Springer-

Verlag, pp. 410-448.

Warr, W.B. and Guinan. J.J. (1979) Efferent innervation of the organ

of Corti: two separate systems. Brain Res 173. 152-155.

Wiederhold, M.L. (1967) A study of efferent inhibition of auditory

nerve activity. Doctoral dissertation, MIT. (Cited by Teas et al.,

1972).

Wilson, J.L., Henson, M.M. and Henson, O.W., Jr. (1991) Course

and distribution of efferent fibers in the cochlea of the mouse.

Hear. Res. 55, 98-108.

Zook, J.M. and Leake, P.A. (1989) Connections and frequency

representation in the auditory brainstem of the mustache hat,

Prrronoftrs pamellii. J. Comp. Neural. 290, 243-261.

efferent terminals in the cochlea of the mustached bat: quantitative data

Documents