Developmental changes of frequency representation in the rat cochlea

Hearing Research, 56 (1991) l-7 0 1991 Elsevier Science Publishers

HEARES

B.V. All rights reserved

037%5955/91/$03.50

01621

Developmental

changes of frequency representation

in the rat cochlea

Marcus Miiller Zoologisches Institut der J. W. Goethe Universitiit, Frankfurt am Main, F.R. G. (Received

12 September

1990; accepted

10 April 1991)

The place-frequency map of the developing rat cochlea was measured by iontophoretic HRP-injections into the ventral cochlear nucleus at electrophysiologically characterized positions. Distribution of retrograde HRP transport in cochlear spiral ganglion cells was analysed by means of a three dimensional reconstruction of the cochlea. Cochlear place-frequency maps were derived in rats of two ages groups: 13 to 22 days, and 36/37 day old animals. These maps were compared with the place-frequency map of adult rats (Miiller, 1991). No systematic difference in the place-frequency map between 36/37 day old and adult rats was observed. In animals of the younger age group the place-frequency map (for frequencies above 4 kHz) was shifted towards lower frequencies for a given place along the basilar membrane. The morphological and physiological basis for this frequency shift during development is discussed. Rat; Cochlea;

Development;

Place-frequency

map; HRP

Introduction

It is generally accepted, and described consistently, that in birds and mammals the development of hearing begins with low to mid-frequency ranges (Rubel, 1978). Furthermore it has been reported, that the cochlea matures morphologically from the cochlear base to the cochlear apex (for review see Romand, 1983; Rubel, 1978). These two findings lead to a paradox because in the adult animal high frequencies are represented at the cochlear base and low frequencies at the cochlear apex. This contradiction was solved by a hypothesis formulated by Rubel et al. (1983). They hypothesized that during development the cochlear place-frequency code is not stable, in that the place where a given frequency is represented shifts from the cochlear base towards the cochlear apex. The hypothesis was based on observations of cochlear hair cell loss produced by overstimulation with different frequencies in an age graded series of chickens. They found that the location of maximum hair cell loss shifted progressively apically as a function of age. The hypothesis was further supported by evidence from electrophysiological mapping in auditory brainstem nuclei of chicken (Lippe and Rubel, 1983). As in the cochlea, at a given position within the nuclei investigated progressively higher frequencies were represented while development proceeded.

Correspondence foe: Marcus Miiller, Klinikum der J.W. Goethe-Universitit, Zentrum der Physiologie, Theordor-Stern-Kai 7, 6000 Frankfurt am Main 70, F.R.G.

Recently the hypothesis of Rubel and coworkers generated contradictory discussion. Manley et al. (1987) investigated the development of frequency representation in Leghorn chickens of different posthatching ages by labelling physiologically characterized auditory nerve fibres. They found that, after hatching, the cochlear place-frequency map remained stable. Their findings were further supported by results from Cotanche et al. (1987) and Tilney et al. (1987) who demonstrated that the location of maximum hair cell damage is a function of stimulus intensity. They argued that, since hearing sensitivity increases during the age stages Rubel et al. (1983) investigated, animals were stimulated with different intensities above threshold. However, the results of Manley et al. (1987), Cotanche et al. (1987) and Tilney et al. (1987) do not explain the frequency shift found in the brainstem nuclei (Lippe and Rubel, 1983; Lippe, 1987). In mammals experimental evidence concerning the hypothesis of Rubel et al. (1983) seems to be somewhat clearer. On the basis of local differential cochlear microphonic recordings, Harris and Dallas (1984) reported for the first time a developmental frequency shift in the basal turn of the cochlea in the gerbil Meriones unguiculats. These results were confirmed with the same method by Yancey and Dallos (19851, and further modified by Arjmand et al. (1988). In the latter study two locations in the cochlea were compared. They found no significant shift in characteristic frequency in the second cochlea turn, while a frequency shift of 1.5 octave occurred in the mid-basal cochlea turn. Further positive evidence for Rubels theory comes from very recent results of Echteler et al.

2

(1989) in the gerbil Meriones unguiculatus. During postnatal development they recorded from spiral ganglion cells at a basal cochlear position and found a progressive change to higher characteristic frequencies. Also investigations in different nuclei in the central auditory system of the mongolian gerbil (Sanes et al., 1989; Ryan and Woolf, 1988) house mice (Romand and Ehret, 1990) and of a hipposiderid bat (Rubsamen et al., 1989) demonstrated a shift in characteristic frequency during development. In the present paper the development of tonotopicity in the cochlea of the rat after onset of hearing is reported using a different technique. In contrast to the other studies in mammals it was attempted to investigate tonotopy over a large frequency range at different postnatal stages.

Material and Methods

In order to determine the characteristic position of a specific frequency along the basilar membrane (BM) iontophoretic HRP-injections were made into the cochlear nucleus at sites electrophysiologically characterized with respect to their characteristic frequency. The pattern of retrograde transport was evaluated in cochlear spiral ganglion cells. Cochlear tonotopy was investigated in animals in two age groups: The first group consisted of animals in age stages between onset of hearing until adult-like hearing sensitivity is reached (13-22 days after birth, DAB), the second group consisted of animals in an age of two weeks after adult-like sensitivity is reached (36/37 DAB). Successful experiments were conducted in 13 animals (Rattus noruegicus, Wistar SPF, supplied by Hoechst AG; 13 DAB N=1,14DABN=2,15DABN=1,16DABN=3, 21 DAB N= 1, 22 DAB N= 1, 36 DAB N=2, 37 DAB N = 2). The age refers to the day when the HRP injections and recordings were made, with day of birth being 0 DAB. Methods were already described in detail previously (Miiller, 1990, 1991), so only a brief description is given here. During surgery and electrophysiological recordings animals were anaesthetized with Ethrane (2-3% in oxygen/nitrousoxide 4: 6, flow 1 l/min). The cochlear nucleus was approached from dorsal through a hole in the skull passing the electrode through the cerebellum. As the cochlear nucleus was reached, the characteristic frequency (CF) of single units or multiunit clusters was measured in 50 mm steps. Stimuli consisted of 40 ms pure tone pips (incl. 5 ms rise- and fall-time). For recording and HRP-application micropipettes filled with 10% HRP in 0.5 M NaCl were used. Only when no sudden changes in CF were de-

tected during a penetration, the precise excitatory threshold tuning curve from a single-or multi-unit was determined, and HRP was applied iontophoretically (5 min, 1 PA, 800 ms on, 200 ms off). In all animals it was attempted to deposit HRP at two sites which had a difference in CF of at least one octave. The tonotopicity in the cochlear nuclei was similar as described in adult rats (Miller, 1991). While lowering the electrode, in the anteroventral cochlear nucleus no or little change in CF was observed, however in the posteroventral cochlear nucleus a gradient from low to high frequencies was found. Consequently, HRP-applications were made in the anteroventral cochlear nucleus whenever possible. During the survival time of 20-24 h animals of the younger age group were placed in a cage with a heating pad (30°C) and fed several times with warm milk. The animals kept or even gained weight and were in good condition during this time. After the survival time animals were killed with an overdose of Nembutal, and perfused with saline followed by fixative. The cochleae were decalcified, sectioned in a cryostat (48 km), and reacted with tetramethylbenzidine. Alternate sections (48 pm) of the brain were reacted with diaminobenzidine and tetramethylbenzidine. The position of labelled spiral ganglion cells and their processes to the inner hair cells along the cochlear duct was determined by means of a computer program which allows a three-dimensional reconstruction of the cochlea coil (method modified after Kraus et al. 1982). The inner and outer anchoring points of the BM (defined by the lower end of the inner pillar cell and spiral lamina, respectively) in all sections were drawn using a microscope equipped with a drawing tube (Zeiss, 2.5/0.08, 37 X ). In addition the outline of the cochlea, Scala tympani and Scala vestibuli were drawn as landmarks. The drawings were fitted and marked with reference points. The coordinates of the inner and outer anchoring points of the BM were determined with a digitizer (Morphomat 10, Zeiss) and analysed with a computer programme which allows the reconstruction of the cochlea coil. Data points were sorted with the programme in sequential order, depending on their location in the cochlea spiral. Length position for every BM-anchoring point was calculated by summing the three-dimensional distance between neighbouring points. BM-length was normalized to lOO%, 0% refers to the cochlea base, 100% to the cochlea apex. The position of the hair cells which were innervated by the labelled spiral ganglion cells was determined on the horizontal projection of the three dimensional BM-reconstruction. For each injection, the position where the maximum number of distal spiral ganglion cell processes to the inner hair cells was found, was defined as the characteristic position of that frequency on the BM.

3

100

Results Physiology Since the sample of recorded neurons was quite small in the different ages only a brief description of the physiological data can be given (Figs. 1 and 2). Tone evoked single or multi-unit responses could first be obtained in 12 day old animals. However, at this age tuning curves with a clearly defined CF (i.e. a difference in tip to tail threshold of at least 20 dB) were not recordable. The thresholds for these responses were about 100 dB SPL. At 13 DAB mean thresholds fell to about 70 dB SPL and the first signs of tuned neurons were visible. At 14 DAB mean thresholds further decreased (53 dB SPL) and neurons became tuned to a clearly defined CF. Mean thresholds further decreased to 42 dB SPL at 16 DAB, and 20 dB SPL at 21/22 DAB. This value is comparable to the mean thresholds found in the 36/37 DAB animals. Although the recordings made in cochlear nucleus are clearly not a representative sample, there was a general trend of an expansion of the hearing range with increasing age: 13,‘14 DAB 6-18 kHz; 15/16 DAB 4.25-27.25 kHz; 21/22 DAB 4-32 kHz; adult 1.25-56 kHz’ (Miiller, 19911. Cochiear to~oto~y Following iontophoretic HRP applications located within the CN, labelled axons of cochlear neurons could be traced within the cochlea and all subdivisions of the cochlear nucleus. Injection diameters in the cochlear nucleus ranged from 80 to 250 pm (measured in DAB reacted sections). Injections sites were located in either anteroventral or posteroventral cochlear nucleus. The labelling patterns within cochlear nucleus resembled those found in adult rats (Miiller 1991) and gerbils (Miiller, 1990). Within a restricted area of the cochlea, the somata of spiral ganglion cells were la-

100

3 % 8 E iii 5

sr

I

80

60

40

H A-a a-m H 0-m M

!z 20

I 10.0

0, 1.0

13MR 14DA6 16DAB 16M6 22 DA6 36 DA8

I 100.0

FREQUENCY [kHz] Fig. 1. Individual

tuning curves from animals

bols represent

in diffrent

animals in different

ages.

ages. Sym-

13DAB l 14MB d 16DAB .$9 : 0 21Ma r22OAB O+ o 36/‘37LMEl

-12DAB

0

80 T %

E

.

-13

DAB

_ -14

DAB

.

60-

A

“9.

d

d EI

0 l

.

0

d

40-

0

z

_

-165AR

-

-21/22

43* 00

-t

20-

0

m

o 0

A

I

“$oI

’

DA0

0

mean threshold

I

0 1

10

100

FREQUENCY [ kHz] Fig. 2. Thresholds

at characteristic

ual tuning curves in different different

frequency

revealed

ages. Symbols represent

from individanimals in the

age classes. To the right of the graph mean thresholds the different

at

ages stages are given.

belled. In the region where the maximum number of labelled cells was located their distal afferent processes up to the inner hair cells were also stained with HRP reaction-product. The number of labelled cells ranged from a few up to 250 labelled cells. These labelling patterns resembled that of the adult rats and gerbils obtained with the same methods (Miiller, 1990, 1991). The mean cochlea length was 8.36 mm (SD = 0.35, N = 13), a value similar to that found in adult rats (Miiiler, 1991). No significant changes in BM-length during development were observed (Table I>. In animals up to 22 DAB, frequency locations could be determined from 16 HRP deposits in nine animals, in the 36/37 DAB 7 frequency locations in four animals were traced. The location of HRP labelled spiral ganglion cells in the cochlea varied systematically with the CF recorded at the injection site (Table I, Fig. 31. HRP applications at CN locations with high CFs resulted in labelling spiral ganglion cells at the cochlear base, with decreasing CF at the injection site the location of labelled spiral ganglion cells was found more apically. Compared to adult rats (Miiller, 1991) and the 36/37 DAB animals, in all rats younger than 22 DAB the frequency positions were systematically shifted towards the base of the cochlea (Fig. 3). The frequency positions obtained for the 36/37 DAB rats did not differ systematically from those found in adult rats. Cochlear frequency representation pattern in the adult rat was described in mathematical terms with an exponential function (Miiller, 1991). Since there was no obvious difference in the frequency positions within the 13-22 days ofd rats (and not enough data is available at individual ages) the data from all these animals were pooled and used to calculate a function regression line with the exponential function. The following

4

TABLE

apex

I

PHYSIOLOGICAL AND ANATOMICAL DIVIDUAL EXPERIMENTS

DATA

FOR

THE

1 oo-

IN80-

Age DAB

CF kHL

Characteristic max

basal

apical

13

IO.0

so

3s

Sh

7.75

%I

I4 I4 I4 I4

h.00 10.00 14.50 1X.00

63 49 39 2x

sx

75 53 4x 3x

8.02 X.27 x.02 x.27

is

place (c/c BM-length)

4s 21 21

cochlea length (mm)

E

l -__ . -. _ .---___-__ --___ -._-._ -.

60-

!2 40-

zo-

base

o

0

13-21

DAB

0

36-37

DAB

l

Od:lt

,

'x0 :;t:*.

‘:)

, , , ,,,

h

I 100

IS IS

12.60 27.7s

46 21

40 17

54 27

X.hS X.65

Ih I6 I6 Ih Ih

4.2.5 X.30 15.50 24.00 24.50

75 5s 3x 20 23

65 51 35 IO 20

7X 59 45 26 31

8.40 7.7X x.40 X.23 7.7x

21 21

IO.10 IX.10

54 32

52 27

56 47

8.3X x.4x

function for the rat pup cochlear place-frequency map was computed (coefficient of correlation = 0.987):

22 22

IO.40 32.00

52 I2

47 6

61 25

X.80 8.X0

x(f) = 98.34* exp( -O.O5063*f[kHz])

37 37 37 36 37 37 36

12.50 16.50 24.00 33.7s 44.00 55.00 56.00

52 43 37 IX I8 I 4

47 30 29 IO 8 0 3

55 52 47 22 26 3

8.4X 8.20 9.30 x57 X.2 x.43 X.SJ

Age, frequency recorded at the injection site (CF), corresponding characteristic place (maxl, positions where the basal and apical limit of labelling was seen, and cochlear length determined from the three dimensional reconstruction.

apex

100

---___ -*. I --__

80

1

10

FREQUENCY

[kHz]

Fig. 4. Cochlear tonotopic map in rat pups and adult rats (dashed lines). The regression lines and confidence belts for P = 0.005 are shown.

- 7.969

where x(f) is the characteristic place expressed in % BM-length from the base, and f is frequency in kHz. In Fig. 4 the function regression lines of the cochlear place-frequency map in rat pups and adult rats are shown. Since the confidence belts (P = O.OOS>of the two functions do not intersect at frequencies above 5 kHz it can be concluded that the function regression lines are different for frequencies above 5 kHz. Since no data could be obtained below 4 kHz, no conclusions can be drawn about the frequency representation in the apical 25 or 30% of the BM. The frequency shift between rat pups and adult rats at a given position along the BM spans from 0.4 octaves at 20% BM-length to 0.67 octaves at 70% BM-length. This corresponds to a shift for a given frequency of 9-10% BM-length ( = 0.X mm).

Discussion Methodological considerations

base

FREQUENCY

[kHz]

Fig. 3. Comparison of cochlear place-frequency maps in rat pups and adult rats (dashed line; Miiller, 1991) determined by iontophoretic HRP applications into the cochlear nucleus and following retrograde HRP transport in auditory nerve fibers. Symbols represent the location where the maximum number of labelled processes of spiral ganglion cells were found in the different age classes. The lines show the functions calculated with the non-linear regression analysis.

One important issue to be addressed is whether the method applied allows differentiation of such a small change of only about half of an octave (10% BM-length) between the rat pup and adult cochlear tonotopic map. It is evident from the data that all frequency locations are shifted to lower frequencies in the rat pups as there is no random scatter around the adult cochlear tonotopic map. Furthermore, frequency locations in 36/37 DAB animals did not differ from those in the adult rats (Miiller, 1991). The confidence belts of the function regression lines (Fig. 4) suggest that the resolution capacity of the mapping method applied here is

5

about + 4% BM-length (at P = 0.005). Therefore there seems to be little doubt that the method used is adequate to determine a shift in the place-frequency map of about 10%. However, small frequency shifts in the 13 to 22 day old animals are not unambiguously detectable with the methods applied. It is clear that only a more accurate mapping method (i.e. single fibre labelling) can detect these very small changes in the place-frequency map. Despite the methodological problems the conclusion to be drawn from the data presented here, is that the place-frequency map in the developing rat cochlea seems to be quite stable during at least the first week after onset of hearing, whilst an adult-like placefrequency map is present at 36/37 DAB. Since the 36/37 DAB animals were investigated within the same series of experiments as the younger rat pups, any methodological bias or any minor changes in dataevaluation can be excluded as a source of error in determining characteristic places in the cochlea. Physiology

The developmental changes in the tuning curves reported here are in accordance with data from previous studies in different species (reviews: Rubel 1978; Brugge, 1983; Romand, 1983; Sanes and Rubel, 1988). The expansion of the hearing range during ontogenesis, particularly in the high frequency range, has been reported from cochlear microphonic responses in rat by Crowly and Hepp-Reymond (19661, and, also in rat, with auditory brainstem responses by Blatchley et al. (1987). Puel and Uziel (1987) found that adult-like sensitivity (CAP threshold) in rat is reached at 16 DAB for 16 kHz, at 17 DAB for 4 kHz, and at 19 DAB for 1 kHz, thus also in the low-frequency range a later maturity in sensitivity is present. Despite the very small sample of neurons recorded in the present study, this trend for hearing to start in a mid-frequency range and subsequently expand to lower and higher frequencies was observed. The high variability in CF-thresholds found in the present study might be explained by the fact that the data stem from different animals with some variability in maturity. Since it was shown in the present paper that a place-frequency code is already established at 13/14 DAB, and these frequencies are located in the middle of the cochlea (between 28 and 63% BM-length), one must conclude that cochlear function begins in the middle of the cochlea rather than at the cochlear base. On the other hand, looking qualitatively at the middle ear of a rat pup at this age, it is hard to imagine how high frequencies are transduced by this immature middle ear conductive apparatus. At this age the middle ear is still partly filled with fluid and mesenchyme and middle ear ossicles do not seem to be fully ossified. In a recent investigation in the mongolian gerbil, Woolf

and Ryan (1988) indeed showed that high-frequency cochlear microphonic responses were elicited two days earlier by direct stapes stimulation compared to stimulation of the intact middle and outer ear. Thus the observation that hearing starts in low to mid-frequency ranges rather than high frequencies, might be due more to middle ear transmission capabilities, rather than to cochlear maturity. However, this conclusion is still to be verified in rat. Shift in place-frequency

code

To the authors knowledge the present report is the most complete cochlear place-frequency map in a developing mammal. The findings in the present report support the hypothesis of Rubel et al. (1983) that frequencies shift further apically during ontogenesis. At least in the frequency range between 4 and 30 kHz (corresponding to positions of = 80-10% BM-length), a frequency shift of about half of an octave was found. In the basal turn (adult CF = 17 kHz) of the cochlea in the gerbil a frequency shift of about 1.5 octaves was observed (Echteler et al., 1989; Yancey and Dallos, 1985), whereas for a BM-position further apically (adult CF = 2.6 kHz), no shift in frequency was revealed (Arjmand et al., 1988). Since the apical 25% of the BM was not mapped in the present study, either because the BM is not responsive, or the tonotopic fields in CN were not recorded from (which seems more likely, see also Miiller, 19911, it cannot be ruled out that there is no shift in the cochlear place-frequency map in the low-frequency region of the rat cochlea. A shift in frequency representation has been found in the central auditory system in other mammals. Sanes et al. (19891, investigating the development of frequency representation in the lateral superior olive of gerbils, found that the tonotopic map changed with age such that the characteristic frequencies at a given anatomical position became progressively higher during development. A similar observation was made by Riibsamen et al. (19891 in the inferior colliculus of a bat species. The hypothesis of the development of the place principle was based on the observation that the cochlea matures anatomically from the cochlear base to the cochlear apex (in rat: Wada, 1923). It is known however, that there are cochlear structures which do not follow this baso-apical developmental gradient, For example Burda (1985) could show in two rat strains that the width of the triad of the outer hair cells reaches its maturity first at 30% BM-length. Romand (19831 discussed that in the late stage of development onset of final cochlea maturation occurs at 20% BMlength (myelination of spiral ganglion cells), thus spatial development proceeds in two directions. Passive mechanical cochlear elements

In order to explain a frequency-shift during development an exact knowledge of which anatomical struc-

6

tures in the cochlea (or organ of Corn) contribute to frequency representation is necessary. For example these structures could be the width and thickness of the BM. It has recently been shown in echolocating bats that the overrepresentation of a certain frequency range is correlated with a constant width and thickness of the BM (Vater et al., 1985, Kiissl and Vater, 1985). Unfortunately, except for the early study of Wada (1923), no data about the anatomical development of the rat BM-width and thickness are available in literature. The data of Wada (1923) show some indication that, while the BM-width remains constant, the BMthickness might increase during development. This has been confirmed in a current study in our laboratory, but it is premature to draw conclusions from these data. However, an increasing thickness, or an increasing number of filaments within the BM, would result in an increased stiffness of the membrane and thus might lead to a shift in frequency representation towards higher frequencies. There are a number of structural changes, occurring after the onset of hearing, which might contribute to a change in the vibration pattern of the basilar membrane. For example, Kraus and Aulbach-Kraus (1981) found a reduction in the mass of the tympanic cover layer. This reduction in mass might not only favour a less restricted vibration of the BM (gain in sensitivity), but also account for a shift in frequency representation towards higher frequencies. Changes in the angles between the pillar cells, and the filamentous structure of pillar cells (Kraus and Aulbach-Kraus, 19811, might also contribute to an alteration of the vibration patterns of the BM. The disappearance of marginal pillar cells (a connection between reticular and tectorial membrane, Kraus and Aulbach-~aus, 1981; Lenoir et al., 19871, might result in a reduction of the stiffness (stability) of the connection between the tectorial membrane and the outer hair cells, thus causing a shift in frequency towards lower frequencies. Active mechanical cochlear elements

The current models of cochlear physiology include an active process which is made responsible for the high sensitivity and sharp tuning in the cochlea of adult mammals (Neely and Kim, 1986; Patuzzi and Robertson, 1988). The anatomical basis for this active process is thought to include the outer hair cells and the tectorial membrane. The active process is highly nonlinear, and it has been shown that the outer hair cells are the major source of nonlinearity affecting the motion of the basilar membrane. A number of studies show that the disruption of the active process in the cochlea, due to a number of different manipulations, lead to a loss in sensitivity of about 50-60 dB, furthermore a shift in frequency of about half of an octave to lower frequencies (Robertson et al., 1980; Sellick et al.,

1982; Liberman, 1984; Liberman and Dodds, 19844). The remaining tuning and sensitivity might be regarded as passive cochlear tuning. Thus, it would be very intriguing and elegant to argue that the shift in frequency observed during development is the result of the development of an active biomechanical process in the cochlea. On the basis of measurements of acoustic distortion products (2f,--f2) during development in gerbils, this hypothesis has recently been promoted by Norton et al. (1991). If this is the case one should expect that a gradual increase in sensitivity would be a~ompanied by a gradual shift in place-frequency code. This line of arguing could explain the findings of Echteier et al. (1989) in the gerbil. Their results suggest that developmental changes in place-frequency map are intimately linked with an increase in sensitivity. At a more apical position (adult CF = 2.4 kHz1 in the gerbil cochlea, no shift in characteristic place was observed (Arjmand et al. 19881. This might argue against the developing active process being linked to a shift in frequency, however the active process might not play such an important role in the low frequency region (Dallas, 19%; Johnstone et al., 1986; Patuzzi and Robertson, 1988). The hypothesis that the development of the frequency place code can be explained by the development of the active process, obviously seems not valid for the rat cochlea: in 21/22 DAB animals adult-like sensitivity is reached, but the place-frequency map appears still not to be adult-like. This suggests that the difference in frequency representation observed in the neonate may not be caused by outer hair cell immaturity. Since it appears that there are so many factors contributing to the development frequency place code, some of which might have the opposite effects on the frequency representation pattern, clearly, more quantitative anatomical and physiological studies are needed to clarify the exact origin of frequency representation patterns and it’s development. For example it might be very important to investigate whether manipulations of the active process have similar effects in adult and neonate animals. However, it seems that there is now overwhelming evidence that Rubels hypothesis of the place-frequency shift during ontogenesis is confirmed with different approaches and different mammalian species.

Acknowledgements I thank Prof V. Bruns, Dr. W Plassmann and Dr. D. Robertson for critical reading and discussions on the manuscript. I also thank Dr. D. Caird for correcting the English of an earlier version of the manuscript.

This work was supported by the Deutsche Forschungsgemeinschaft (SFB 45, B21).

References Arjmand, E., Harris, D. and Dallas, P. (1988) Developmental changes in frequency mapping of the gerbil cochlea: Comparison of two cochlear locations. Hear. Res. 32, 93-96. Blatchley, B.J., Cooper, W.A. and Coleman, J.R. (1987) Development of auditory brainstem response to tone pip stimuli in the rat. Dev. Brain Res. 32, 75-84. Brugge, J.F. (1983) Development of lower auditory nuclei. In: R. Romand (Ed.), Development of Auditory and Vestibular Systems, Academic Press, pp. 89-119. Burda, H.(1985) Qualitative asessment of postnatal maturation of the organ of corti in ttwo rat strains. Hear. Res. 17, 201-208. Cotanche, D.A., Saunders, J.C. and Tilney, L.G. (1987) Hair cell damage produced by acoustic trauma in the chick cochlea. Hear. Res. 25, 267-286. Crowly, D.E. and Hepp-Reymond, MC. (1966) Development of cochlear function in the ear of the infant rat. J. Comp. Physiol. Psychol. 62, 427-432. Dallas, P. (1986) Neurobiology of cochlear inner and outer hair cells: intracellular recordings. Hear. Res. 22, 185-198. Echteler, SM., Arjmand, E. and Dallas, P. (1989) Developmental alterations in the frequency map of the mammalian cochlea. Nature 341, 147-149. Harris, D.M. and Dallas, P. (1984) Ontogenetic changes in frequency mapping of mammalian ear. Science 225, 741-743. Johnstone, B.M., Patuzzi, R. and Yates G.K. (1986) Basilar membrane measurements and the travelling wave. Hear. Res. 22, 147-153. Kiissl, M. and Vater, M. (1985) The cochlea frequency map of the mustache bat Pteronotus purnellii. J. Comp. Physiol. A 157, 687687. Kraus, H.J. and Aulbach-Kraus, K. (1981) Morphological changes in the cochlea of the mouse after the onset of hearing. Hear. Res. 4, 89-102. Kraus, H.J., Fiedler, J., Ziiller, H. and Bruns, V. (1982) Three-dimensional reconstruction and quantitative analysis of the mammalian cochlea. Verh. Dtsch. Zool. Ges. 279. Lenoir, M., Puel, J.-C. and Pujol, R. (1987) Stereocilia and tectorial membrane development in the rat cochlea. A SEM study. Anat. Embryol. 175, 477-487. Liberman, M.C. (1984) Single neuron labeling and chronic cochlear pathology. I. Threshold shift and characteristic frequency shift. Hear. Res. 16, 33-41. Liberman, M.C. and Dodds, L.W. (1984) Single neuron labeling and chronic cochlear pathology. III. Stereocilia damage and alterations to threshold tuning curves. Hear. Res. 16, 55-74. Lippe, W.R. (1987) Shift of tonotopic organization in brain stem auditory nuclei of the chicken during late embryonic development. Hear. Res. 25, 205-208. Lippe, W.R. and Rubel, E.W. (1983) Development of the place principle: tonotopic organization. Science 219, 514-516. Manley, G.A., Brix, J. and Kaiser, A. (1987) Developmental stability of the tonotopic organization of the chick’s basilar papilla. Science 237, 655-656. Miiller, M. (1990) Quantitative comparison of frequency representation in the auditory brainstem nuclei of the gerbil Pachyuromys duprusi. Exp. Brain Res. 81,140-149. Miiller, M. (1991) Frequency representation in the rat cochlea. Hear Res. (in press)

Neely, S.T. and Kim, D.O. (1986) A model for active elements in cochlear biomechanics. J. Acoust. Sot. Am. 79, 1472-1480. Norton S.J., Bargones J.Y. and Rubel E.W. (1991) Development of otoacoustic emissions in gerbil: Evidence for micromechanical changes underlying development of the place code. Hear. Res. 51, 73-92. Patuzzi, R. and Robertson D (1988) Tuning in the mammalian cochlea. Physiological Rev. 68, 1005-1083. Patuzzi, R.B., Yates G.K., and Johnstone B.M. (1989) Outer hair cell receptor current and sensorineural hearing loss. Hear Res. 42, 47-72. Puel, J-L. and Uziel, A. (1987) Correlative development of cochlear action potential sensitivity, latency, and frequency selectivity. Dev. Brain Res. 37, 179-188. Robertson, D., Cody, A.R., Bredberg, G. and Johnstone, B.M. (19801 Response properties of spiral ganglion neurons in cochlea damaged by direct mechanical trauma. J. Acoust. Sot. Am. 67, 1295-1303. Romand, R. (1983) Development of the cochlea. In: R. Romand (Ed.), Development of Auditory and Vestibular Systems, Academic Press, pp. 47-88. Romand, R. and Ehret, G. (1990) Development of tonotopy in the inferior colliculus. I. Electrophysiological mapping in house mice. Dev. Brain Res. 54, 221-234. Rubel, E.W. (1978) Ontogeny of structure and function in vertebrate auditory systems. In: M. Jacobson (Ed.), Handbook of Sensory Physiology, Vol IX, Development of Sensory Systems, Springer, New York, pp. 135-237. Rubel. E.W., Lippe, W.R. and Ryals, B.M. (1983) Development of the place principle. Ann. Otol. Rhinol. Laryngol. 93, 609-615. Riibsamen, R., Neuweiler, G. and Marimuthu, G. (1989) Ontogenesis of tonotopy in inferior colliculus of a hipposiderid bat reveals postnatal shift in frequency-place code. J. Comp. Physiol. A 165, 755-769. Ryan, A.F. and Woolf, N.K. (1988) Development of tonotopic representation in the Mongolian gerbil: A 2-deoxyglucose study. Dev. Brain Res. 41, 61-70. Saner,, D.H. and Rubel, E.W. (1988) The development of stimulus coding in the auditory system. In: A.F. Jahn and J. Santos-Sacchi (Eds.1, Physiology of the Ear, Raven Press, New York, pp. 431456. Sanes, D.H., Mernickel, M. and Rubel, E.W. (1989) Evidence for an alteration of the tonotopic map in the cochlea during development. J. Comp. Neurol. 279, 436-444. Sellick, P.M., Patuzzi, R. and Johnstone B.M. (1982) Measurement of basilar membrane motion in the guinea pig using the Miissbauer technique. J. Acoust. Sot. Am. 72, 131-141. Tilney, M.S., Tilney, L.G. and DeRosier, D.J. (1987) The distribution of hair cell bundle lengths and orientations suggests an unexpected pattern of hair cell stimulation in chick cochlea. Hear. Res. 25, 141-151. Vater, M., Feng, A.S. and Betz, M. (1985) An HRP-study on the frequency place map of the horseshoe bat cochlea: Morphological correlates of the sharp tuning to a narrow frequency band. J. Comp. Physiol. A 157, 671-686. Wada (1923) Anatomical and physiological studies on the growth of the inner ear in the albino rat. Am. Anat. Mem. 10, l-74. Woolf, N.K. and Ryan, A.F. (1988) Contributions of the middle ear to the development of function in the cochlea. Hear. Res. 35, 131-142. Yancy. C. and Dallos, P. (1985) Ontogenetic changes in cochlear characteristic frequency at a basal turn location as reflected in the summating potential. Hear. Res. 18, 189-195.