6 mouse model of presbycusis

IQIlUI IIl .aKIt

ELSEVIER

Hearing Research 89 (1995) 109-120

Synaptic loss in the central nucleus of the inferior colliculus correlates with sensorineural hearing loss in the C57BL/6 mouse model of presbycusis Ann M. Kazee

a, *

Li Ying Han

a

Vlasta P. Spongr d Joseph P. Walton c, Richard J. Salvi a, Dorothy G. Flood b

a Department of Pathology and Laboratory Medicine, University of Rochester, 601 ElmwoodAvenue, Box 626, Rochester, NY 14642, USA b Department of Neurology, University of Rochester, Rochester, NY 14642, USA c Department of Surgery (Otolaryngology Division), University of Rochester, Rochester, NY 14642, USA d Hearing Research Laboratory, University of Buffalo, Buffalo, N~, USA Received 27 December 1994; revised 3 May 1995; accepted 7 June 1995

Abstract Between 3 and 25 months of age, light and electron microscopic features of principal neurons in the central nucleus of the inferior colliculus of the C57BL/6 mouse were quantitated. This mouse strain has a genetic defect producing progressive sensorineural hearing loss which starts during young adulthood (2 months of age) with high-frequency sounds. During the second year of life, hearing is severely impaired, progressively involving all frequencies. The hearing loss was documented in the present study by auditory brainstem recordings of the mice at various ages. The cochleas from many of the same animals showed massive loss of both inner and outer hair cells beginning at the base (high-frequency region) and progressing with age along the entire length to the apex (low-frequency region). In the inferior colliculi, there was a significant decrease in the size of principal neurons in the central nucleus. There was a dramatic decrease in the number of synapses of all morphologic types on principal neuronal somas. The percentage of somatic membrane covered by synapses decreased by 67%. A ventral (high frequency) to dorsal (low frequency) gradient of synaptic loss could not be identified within the central nucleus. These synaptic changes may be related to the equally dramatic physiologic changes which have been noted in the central nucleus of the inferior colliculus, in which response properties of neurons normally sensitive to high-frequency sounds become more sensitive to low-frequency sounds. The synaptic loss noted in this study may be due to more than the loss of primary afferent pathways. It may represent alterations of the complex synaptic circuitry related to the central deficits of presbycusis. Keywords: Presbycusis; C57 mouse; Inferior colliculus; Synaptic morphometry; Cytocochleogram; Auditory brainstem response

1. Introduction The auditory system of the C 5 7 B L / 6 (C57) mouse is mature by about 1 month of age (Willott and Shnerson, 1978; Shnerson and Willott, 1979; Shnerson et al., 1982; Shnerson and Pujol, 1982). Starting soon after sexual maturity, however, the cochlea in the C57 mouse undergoes progressive degeneration starting with the basal (high frequency) region and progressing toward the apical (low frequency) region (Mikaelian, 1979; Henry and Chole, 1980; Willott, 1986; Li and Borg, 1991). By the end of the

* Corresponding author: Tel.: (716) 275-3202; Fax: (716) 273-1027. 0378-5955/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 3 7 8 - 5 9 5 5 ( 9 5 ) 0 0 1 2 8 - X

second year of life, the hearing of the C57 mouse has diminished to near total deafness. The C57 mouse shows a pattern of hearing loss similar to human presbycusis in which the hearing loss progresses from high to low frequencies. The inferior colliculus (IC) is one of the major nuclei of the auditory brainstem. The central nucleus of the inferior colliculus (ICCN) receives ascending auditory input while the cortex of the IC receives descending input from auditory neocortex. Most of the previous anatomical studies of the ICCN have been done in the cat or rat (Rockel, 1971; Rockel and Jones, 1972a, b; Morest and Oliver, 1984; Oliver and Morest, 1984; Faye-Lund and Osen, 1985). Based on cytoarchitecture, the ICCN is divided into central, medial, lateral and ventral parts. Principal or disc-

110

A.M. Kazee et al. / Hearing Research 89 (1995) 109-120

shaped cells (60-80%) and multipolar or stellate cells have been identified in the ICCN. In Golgi preparations, the dendrites of the principal neurons are bipolar and are arranged parallel to the afferent fibers of the lateral lemniscus into laminae. Although the orientation of the laminae varies slightly among the divisions of the ICCN, generally they run from ventro-lateral to dorso-medial in transverse sections at 20-50 ° angles. Principal neurons have been classified according to size as small, medium, mediumlarge, or large while multipolar neurons have been classified as simple, complex, or small. Although the nomenclature is slightly different in the mouse, the basic organization of the ICCN is similar (Meininger et al., 1986). In mice, too, the ICCN has a well-defined tonotopic organization, with high frequencies roughly being represented ventrally and low frequencies dorsally, although the tonotopic map does not exactly correspond to anatomic subdivisions (Semple and Aitkin, 1979; Syka et al., 1981; Stiebler and Ehret, 1985). Willott et al. (1988) found dramatic changes with aging in the response properties of ICCN neurons to contralateral auditory stimuli. There were increased thresholds in the > 25 kHz range as early as 6 months of age; and by 15 months the thresholds increased at all frequencies, most profoundly in the high-frequency range (ventro-medial IC). By the end of the second year of life, all thresholds in the ICCN were greatly elevated and most best frequencies fell within a narrow mid-frequency range (8-16 kHz) and respond only at very high intensities. There were also increased numbers of 'sluggish' neurons (auditory, but poorly driven by sound) and an increase in the number of spontaneously active neurons, suggesting a decrease in inhibitory input. Age-related alterations in temporal processing has also been observed in the C57 mouse. Walton et al. (1995) have shown that C57 mice having hearing loss on the order of 25-45 dB show significant increases in recovery from short-term adaption at the level of the midbrain, as measured by the P5 component of the auditory brainstem response (ABR). Moreover, the magnitude of the shift was comparable to that reported behaviorally for human hearing impaired listeners. One of Willott's most interesting observations though was that neurons in the ventral portion of the IC, which normally respond to high-frequency sounds, become more sensitive to low-frequency sounds with aging (Willott, 1984). Willott postulated that this change in auditory sensitivities may be due to changes in either the auditory periphery if the basal cochlea becomes more responsive to low frequencies, or to changes in central auditory pathways. There is no evidence to support the first hypothesis that the basal region of the cochlea responds to low frequency sounds at low sound levels. The later hypothesis could conceivable be due to changes in anatomic pathways such as collateral growth of remaining axons sensitive to low frequencies. There is, however, no age-related change in the projections of the anteroventral cochlear nucleus to

the ICCN based on retrograde transport of horseradish peroxidase tracer (Willott et al., 1985). Alternatively, the high-frequency neurons could become more sensitive to low-frequency sounds if low-frequency inputs to the ICCN are unmasked by denervation supersensitivity or diminished inhibition. The ICCN receives strong input from both the ventral and the dorsal cochlear nuclei (VCN and DCN). VCN neurons do not become sensitized to low frequencies like ICCN neurons do, while highfrequency DCN neurons become sensitive to low frequencies when the base of the cochlea is damaged (Willott et al., 1991). Unmasking DCN input to the ICCN might be the cause of the change in ICCN sensitivity by a mechanism of synaptic reorganization or sprouting. This change in sensitivity is carried through to the primary auditory cortex where middle frequency representation is increased with aging in the C57 mouse with concomitant loss of units responding to high frequencies (Willott et al., 1993). In an effort to determine what neuroanatomical changes may underlie these central auditory processing defects, we present a quantitative ultrastructural analysis of the principal cells of the central nucleus of the inferior colliculus (ICCN) across the lifespan of the C57 mouse.

2. Materials and methods A total of 15 C57 mice, aged 3-25 months, were used in the present study. The animals were obtained from the National Institute of Aging colony maintained by Charles River Breeding Labs, except for the oldest group of animals which were from Jackson Labs and aged at the University of Rochester. In most, but not all cases, the same animal was used for the neurophysiology and neuroanatomy parts of this study. The external and middle ears of all the mice were examined. Only those animals which had no evidence of obstruction/inflammation were used in the present study.

3. Neurophysiology The auditory brainstem response (ABR) was used to measure peripheral and brainstem auditory sensitivity prior to killing the animals. The animals were mildly tranquilized (chlorprothixene 7 /zl/mg), restrained using a fullbody form-fitting sleeve, and their body temperature was maintained at 38°C with a hot water heating pad. The ABR was recorded via subcutaneous platinum needle electrodes inserted at the vertex (non-inverted input), right mastoid prominence (inverted input), and tail (indifferent site). The EEG activity was differentially amplified ( × 50000) and filtered (100-3000 Hz) before averaging (Biologic Navigator). Artifact rejection was used to limit contamination by muscle activity by automatically rejecting data epochs where the EEG amplitude exceeded 7/xV. Each averaged

111

A.M. Kazee et al. / H e a r i n g Research 89 (1995) 109-120

response was summed for 500 stimulus presentations over a time period of 10 ms. Peak identification was based on the initial positive peak, referred to as P1, with subsequent positive peaks labeled P2-P5. Stimuli were presented 13 cm from the ear canal opening via an ultra-high-frequency leaf-tweeter (Panasonic AA102) having an effective bandwidth from 1 to 100 kHz. Stimulus intensity was calibrated using a Bruel and Kjaer 1 / 4 " microphone placed at the metal opening of the pinna. An ABR audiogram was obtained by presenting tone bursts at l l / s having the following characteristics: 5 ms durations, 1 ms cosineshaped rise-fall times (10-90%), center frequencies of 4, 8, 10, 12, 16, 24, 36, and 50 kHz. Thresholds were determined by lowering stimulus intensity in 10 dB steps to below thresholds, then bracketing thresholds to 5 dB steps. ABR testing was not performed in animals older than 14 months due to the severity of the hearing loss.

4. L i g h t m i c r o s c o p y :

lOO

\../

90

Q-, rl (73 0:3

8o

g

x~

70

•

60 5o

o o3 (-

4o 3o 20 lO o

•

I mon.

x

8

•

6 mon.

•

14

. . . . . . .

2.0

4.0

I

.

.

.

mon. .

mon. .

.

l

I

.

1010 2 0 . 0

100.0

Frequency (kHz) Fig. l. Mean thresholds for ABRs in 1 (O) n = l l , 6 (11) n = 10, 8 ( X ) n = 10 and 14 (A) n = 4, month old C57 mice. Standard deviations ranged from 4 to 9 dB.

cochleas

Following ABR recording, the animals were killed for cochlear and brainstem anatomy. They were deeply anesthetized with 0.5 ml of sodium pentobarbital (6.5 m g / m l , i.p.) and perfused with 0.1% heparin sulfate and 0.5% sodium nitrite in 0.85% sodium chloride for 10 min, followed by 1% glutaraldehyde and 4% paraformaldehyde in phosphate buffer (pH 7.3) for 25 min at a rate of 3 m l / m i n . The cochleas were removed from the temporal bones and the round and oval window membranes were opened. The cochleas were perfused by direct injection of buffered glutaraldehyde fixative through the round and oval windows. They were then immersed in fixative and stored at 4°C for further processing. The cochleas were subsequently post-fixed in 1% buffered osmium tetroxide, decalcified in ethylene diamine tetracetic acid, and then gradually dehydrated up to 70% ethyl alcohol. The organ of Corti and the stria vascularis were dissected out as half-turns and evaluated. Hair cell counts were obtained under an oil immersion lens at 0.1-0.18 mm intervals over the entire length of the cochlea. The percentage of missing hair cells was computed on the basis of hair cell density data obtained from 10 normal hearing, 1-month-old animals. A cytocochleogram of hair cell loss was obtained by plotting the percentage of missing inner hair cells (IHC) and outer hair cells (OHC) along the length of the cochlea. Percent distance from the apex of the cochlea was related to frequency using a generalized frequency place map (Greenwood, 1990).

5. L i g h t a n d e l e c t r o n m i c r o s c o p y :

inferior colliculi

After removal from the skull, the brains were fixed overnight at 4°C. The following day, the inferior colliculi were identified and each was cut in the sagittal and

horizontal planes. These cuts divided each inferior colliculi into 4 quadrants: dorso-lateral, dorso-medial, ventro-lateral and ventro-medial. The blocks were rinsed in phosphatebuffered saline, postfixed in 1% buffered osmium tetroxide for 1.5 h, dehydrated in graded alcohols, and embedded in Spurr epoxy resin. Sections (1.5 /zm thick) were stained with toluidine blue and the central nucleus identified. The central nucleus was identified by its location relative to the central aqueduct and the external cortex. Axons of the lateral lemniscus formed a dense network of parallel fiber laminae within the ICCN. Principal neurons had ovalshaped somas and prominent bipolar dendrites. These dendrites were oriented along the fiber laminae of the lateral lemniscus (Fig. 3). Multipolar neurons could be distinguished from principal neurons by their multiple dendrites which were not oriented parallel to the fiber laminae. Some dendrites of multipolar neurons were perpendicular to the laminae. Quantitative data were obtained at the light microscopic level using a microcomputer (as described by Curcio and Sloan, 1986). For the morphometry, the animals were divided into 3 age groups (young, 3 months; middle-aged, 6 - 1 8 months; old, 24 months) consisting of 4 - 6 animals in each group. The age groups were compared with respect to neuronal, nuclear, and nucleolar area, and major and minor axis diameter. Only those principal neurons which contained visible nucleoli were used for quantitative light microscopy. A total of 646 neurons were measured at the light microscopic level at a total magnification of × 4 550. For the electron microscopy, the blocks were trimmed to include only the ICCN, and silver thin sections were cut and stained with uranyl acetate and lead citrate, and examined on a Hitachi H7100 electron microscope. Principal neurons were identified for ultrastructural examination on the basis of their size, location and orientation within the

112

A.M. Kazee et aL / Hearing Research 89 (1995) 109-120 A

Frequency (kHz) 1 iiiif~iq

i

f

10 I

i 1rill

f

100 J J irll I

i

100

-

~

so

.--.

40

~

il

-i~ 2o 0

I

I

I

0

[

I

I

I

20

I

I

I

I

40

[

I

I

I

60

I

I

I

80

100

% Distance from apex B

Frequency (kHz) 1 i~lJJl,l 100

10

I

,

- -

o

i

0

i

IHC

i

i

i

i

i

J

i

i

i

1Ill

I

y

,,

i

100

1~1111 -

I

i

i

i

i

i

,

t

i

20 40 60 80 % Distance from apex

,

~

~

I

100

Fig. 2. Representative cytocochleograms from C57 mice, aged 6 months old (A) and 18 months old (B), showing the typical pattern of hair cell loss for IHCs and the 3 rows of OHCs as a function of their distance from the cochlear apex.

were quantitated by tracing them into the microcomputer using the graphic tablet. Synaptic profiles were identified by the presence of a presynaptic terminal containing synaptic vesicles, a synaptic cleft and p r e - a n d posts y n a p t i c p a r a m e m b r a n o u s thickenings. A s y m m e t r i c synapses were sometimes curved and had a postsynaptic membrane density thicker than the presynaptic density; symmetric synapses were characterized by pre- and postsynaptic membrane densities of equivalent thickness. Vesicle shape was also used to describe the synaptic type. The data as a function of age for the ICCN as a whole were examined statistically using parametric analyses of variance (ANOVA). Newman-Keuls post hoc analyses were also performed between pairs of age groups. Split-plot factorial design A N O V A s were used to test for significant differences among age groups with respect to neuronal location within the ICCN according to the dorso-ventral and medio-lateral quadrants of the ICCN. Tests yielding P values of less than 0.05 were considered statistically significant. P values of less than 0.01 are noted when found. This design properly adjusts for the repeated sampling of the subdivisions of the ICCN. The care and use of the animals reported on in this study were approved by the National Institutes of Health and the University of Rochester Committee on Animal Resources regulations.

6. R e s u l t s

ICCN. Only principal cells which contained nuclei were examined quantitatively at the ultrastructural level. The entire neuron was photographed at low magnification ( × 9 0 0 0 ) for measurement of the somatic membrane perimeter. Then, the same neuron was rephotographed at a higher magnification ( x 36 000), and the number and length of synaptic appositions as well as presynaptic terminal area

6.1. Auditory physiology

Mean A B R tone-burst thresholds plotted as a function of probe frequency for 4 age groups of C57 mice are displayed in Fig. 1 and show the progression of sensorineural hearing loss over the first 14 months of life. Threshold shifts of 2 5 - 3 0 dB are initially observed by 6

Fig. 3. Light microscopic field from a toluidine blue-stained 1.5 /xm thick plastic section showing principal neurons (white arrows) oriented parallel to afferent fiber laminae in the ICCN of a 3-month-old C57 mouse. Calibration bar = 40 k~m.

A.M. Kazee et al. / Hearing Research 89 (1995) I09-120

113

Fig. 4. A small principal neuron is shown in this low magnification electron micrograph from a 3-month-old C57 mouse. The neuron has a large nucleus, scant cytoplasm and axosomatic synapse (arrow and inset). Original magnification: × 9000; calibration bar = 2 /zm.

Table 1 Quantitative light microscopic data from young, middle-age and old C57 mice

Neuronal area ( / z m 2) ( ± SEM) Nuclear area ( b~m2) ( ± SEM) Nucleolar area ( / ~ m 2 ) ( + SEM) Neuronal major axis diameter ( ~ m ) ( ± SEM) Neuronal minor axis diameter ( / ~ m ) ( ± SEM)

Y o u n g (3 months) (n = 5, 284 neurons)

Middle-age ( 6 - 1 8 months) (n = 6, 219 neurons)

Old (24 months) (n = 4, 143 neurons)

211 + 7 66+2

161 _+ 10 * * 57+ 3

145 ± 13 * * 50± 3 *

9 ± 0.5

8 ± 0.4

6 ± 0.9 *

22 + 0.5

19 + 0.6 *

18 5- 0.6 * *

13 ± 0.3

11 ± 0.4 * *

11 ± 0.5 * *

Newman-Keuls post hoc tests. SEM = standard error of the mean. * Significant difference from young at P < 0.05. * * Significant difference from young at P < 0.01.

114

A.M. Kazee et al. / Hearing Research 89 (1995) 109-120

months of age in the high frequencies. The loss progresses rapidly and by 8 months of age, at least a 20 dB threshold shift extends across the entire auditory spectrum. By 14 months the animals are functionally deaf. These mean thresholds are in agreement with those obtained in several other laboratories for C57 mice (Henry and Lepkowski, 1978; Li and Borg, 1991). In the 1- and 6-month-old age groups, standard deviations ranged from 4 to 9 dB. 6.2. C o c h l e a r p a t h o l o g y

The cochleas from the C57 mice show a progressive loss of both OHC and IHC as a function of age. The typical pattern of hair cell loss from a 6-month-old mouse

is illustrated in Fig. 2A. At this age, there is nearly complete loss of OHC from the base to 75% of the distance from the apex of the cochlea; this region corresponds to frequencies above 40 kHz. The loss then gradually decreases to almost zero approximately 65% of the distance from the apex. There was also a complete loss of IHC from 90 to 100% of the distance from the apex. This corresponds to frequencies above 70 kHz. By 18 months of age, most C57 mice show a significant loss of hair cells over most of the cochlea. Fig. 2B shows the typical pattern of hair cell loss at 18 months of age. All of the OHC and IHC were missing from 55% to 100% of the distance from the apex of the cochlea (frequencies > 12 kHz). Fifty to 80% of the OHC were missing

Fig. 5. Low magnification electron micrograph of a large principal neuron in the central nucleus of the inferior colliculus from a 25-month-old C57 mouse showing the oval-shaped soma with prominent granular endoplasmic reticulum (white arrows) and numerous lipofuscin granules (arrowheads). Original magnification: × 9000; calibration bar = 2 /xm.

A.M. Kazee et al. // Hearing Research 89 (1995) 109-120

throughout the remaining cochlea (regions encoding frequencies < 12 kHz). In addition, 10-50% of the IHC were missing at frequencies below 12 kHz.

6.3. Inferior collicular morphology Principal neurons from young C57 mice (3 months old) had mostly oval-shaped somas, but occasional neurons

115

with more rounded or fusiform contours were seen. The long axis of their somas were oriented parallel to the fibers of the lateral lemniscus. They had two large dendrites extending from opposite poles of the somas also oriented along the fiber laminae. They were more common than multipolar neurons which were also found within the ICCN. They could be distinguished from multipolar neurons by the more numerous, randomly oriented dendrites

Fig. 6. Electron micrographs showing various types of axosomatic synaptic contacts seen in the ICCN of the C57 mouse. A: from a 3-month-old C57 mouse, the terminal contains synaptic vesicles which are small, oval and clear with occasional dense core vesicles (white arrows). These terminals usually make symmetric synaptic contacts. B: from a 3-month-old C57 mouse, shows terminals making symmetric axosomatic synapses which contain clearly rounded synaptic vesicles (white arrows) or oval synaptic vesicles (black arrow). Occasionally, as in this example, there were more than one synaptic contact per terminal and the synapses were slightly concave. C: an asymmetric synapse (arrow) from an 18-month-old C57 mouse. The postynaptic membrane density is thick; the synaptic vesicles are rounded to pleomorphic. The adjacent terminals show a more symmetric synaptic complex (arrowheads) with oval to flatten synaptic vesicles. Original magnification: × 36 000; calibration bar = 1 /zm.

116

A.M. Kazee et aL / Hearing Research 89 (1995) 109-120

Table 2 Quantitative electron microscopic data from young, middle-age and old C57 mice Young (3 months) (n = 5, 70 neurons) Synapses/neuron (n) (+ SEM) Synaptic length (/.tm) (+ SEM) Somatic membrane perimeter (/xm) ( + SEM) Somatic membrane covered by synapses (%) ( + SEM) Synaptic terminal area (/.tm 2) ( _+SEM)

Middle-age (6-18 months) (n = 6, 66 neurons)

Old (24 months) (n = 4, 72 neurons)

3.77 + 0.46

1.37 + 0.32 * *

1.37 _+0.43 * *

0.26 + 0.01

0.28 + 0.01

0.22 + 0.01 *

42.00 + 1.53

41.63 + 1.54

34.59 _+ 1.51 *

2.36 _+0.24

0.84 + 0.16 * *

0.78 _+0.19 * *

0.70 + 0.13

0.50 + 0.07

0.34 + 0.04

Newman-Keuls post hoc tests. SEM = standard error of the means. * Significant difference from young at P < 0.05. * ~ Significant difference from young at P < 0.01.

of the multipolar neurons. The size of principal neurons varied considerably. Large and small principal neurons could be identified scattered throughout each of the 4 quadrants of the ICCN. The majority of principal neurons were oval-to-fusiform in shape, with a nuclear to cytoplasmic ratio of 1:1.5-1:2. In young mice, the neuronal nuclei were relatively large and had several deep infoldings (Fig. 4). The cytoplasmic volume was variable with only a thin rim of cytoplasm in small principal neurons, while the larger neurons had abundant cytoplasm, filled with organelles. In smaller neurons, the rough and smooth endoplasmic reticulum were sparsely scattered throughout the cytoplasm, while in the larger neurons there were regular linear arrays of rough endoplasmic reticulum (Nissl substance) as well as abundant free ribosomes and polyribosomal clusters. Mitochondria were more numerous in larger neurons and in older mice. A well-developed Golgi apparatus was usually found in a perinuclear location. There was some condensation of the nuclear chromatin near the nuclear envelop and near the single large nucleolus, especially in the older age groups. There is extensive accumulation of lipofuscin in the older mice (Fig. 5). A x o s o m a t i c synapses o f the symmetric type were frequently identified on principal neurons throughout the ICCN (Fig. 6A, B). They were characterized by pre- and postsynaptic membrane densities of equivalent thickness. Synaptic terminals making symmetric synapses contained

synaptic vesicles which were mostly oval in shape, but some terminals had vesicles which were clearly small and round; in others, the vesicles were more pleomorphic and irregular. The symmetric axosomatic synapses were usually flat (non-concave). Terminals making symmetric synapses sometimes had 2 - 3 synaptic contacts per terminal (Fig. 6B). Axosomatic synapses with asymmetric pre- and postsynaptic densities could be found throughout the ICCN, but they were less c o m m o n than symmetric axosomatic synapses. The presynaptic terminals which made asymmetric synapses usually contained synaptic vesicles which were rounded in shape and often had a few dense core vesicles as well. The synaptic vesicles were usually densely packed within the terminal. The postsynaptic membrane on the principal neuron soma was thicker than the presynaptic membrane density (Fig. 6C). These synapses were often concave in configuration. Occasionally, they could be found on somatic spicules. There was a striking loss of axosomatic synapses of both the symmetric and asymmetric type in the middle-age group mice ( 6 - 1 8 month) compared with the young (3 month) mice. This loss continued in the oldest age group examined (24 month). There appeared to be a slightly greater loss of the symmetric type synapses, but this could not be quantified due to the sample size and inability to clearly classify each individual synapse. It was evident, though, that synapses of both types were being lost.

Table 3 Percentage of principal cell somatic membrane covered by synapses ( + SEM) according to tonotopic location within ICCN in C57 mice Dorsal Ventral Lateral Medial (low frequency) (high frequency) (low frequency) (high frequency) Young (3 months) Middle-aged (6-18 months) Old (24 months)

2.92 + 0.47 0.54 + 0.33 0.66 + 0.17

1.36 ± 0.30 1.00 _ 0.06 1.05 + 0.42

2.56 + 0.54 1.06 + 0.22 0.86 + 0.26

1.85 _+0.55 0.56 _+0.19 1.03 + 0.39

A.M. Kazeeet al./ HearingResearch89 (1995)109-120 6.4. Inferior colliculus morphometry The neuronal area, nuclear area, and nucleolar area of principal neurons are presented in Table 1. Only principal neurons in the ICCN which contained a visible nucleolus were used to make the measurements. Blocks from all 4 quadrants were examined in each animal. On average, 43 neurons per animal (215 neurons/age group) were measured. The mean and standard deviation from each animal were calculated and the mean and standard error of the mean (SEM) for each age group were determined. The neuronal, nuclear, and nucleolar areas decreased 30% ( F = 10.584; df = 2,12; P < 0.01), 24% ( F = 6.066; dr= 2,12; P < 0.05) and 33% ( F = 4.331; df= 2,12; P < 0.05) respectively from the young (3 months) to the old (24 month) mice. The major axis (long diameter) and minor axis (short diameter) of principal neuron cell bodies also decreased significantly, 18% ( F = 11.047; df= 2,12; P < 0.01) and 15% ( F = 9.518; df= 2,12; P < 0.01) respectively. Results of the post hoc pairwise comparisons (Table 1) showed that: (1) the young group was significantly different from the old group for all measures, (2) the young group was significantly different from the middleaged group for some of the measures, and (3) the middleaged and old groups did not vary significantly from one another. Table 2 shows various ultrastructural parameters which were assessed quantitatively. A total of 216 complete principal neuron somas were examined, averaging 14 neurons per animal and 72 neurons per age group. The somatic membrane perimeter was examined at low magnification ( X 9 0 0 0 ) and showed an 18% decrease ( F = 6.327; dr= 2,12; P < 0.05) in old mice compared with young mice. The number of synapses contacting the somatic membrane decreased from 3 . 7 7 / n e u r o n to 1.37/neuron, a 63% decrease ( F = 12.187; dr= 2,12; P < 0.01). The length of synaptic apposition also decreased slightly (15% decrease; F = 8.114; dr= 2,12; P < 0.01), resulting in a decrease in the percentage of somatic membrane being contacted by synapses from 2.36% to 0.78% (67% decrease; F = 20.408; dr= 2,12; P < 0.01). The area of the presynaptic terminal did not change significantly with age ( F = 3.601, dr= 2,12, P = NS). Post hoc pairwise comparisons (Table 2) showed the middle-age and old animals had significantly fewer synapses than young animals (by synapse number and percentage of

117

somatic membrane covered by synapses). Old animals differed significantly from young and middle-aged animals only in synapse length and somatic membrane perimeter. There was no continued loss of synapses as the animal aged. When the synaptic loss was examined by location according to the tonotopic organization of the ICCN (dorsolateral, low frequency; ventro-medial, high frequency), it is clear that synapses are being lost at all frequencies along the tonotopic map (Table 3). When the synaptic loss was analyzed by split-plot factorial design by location and age group, we see that age is the most significant variable, but there is also a consistent difference, unassociated with age, along the dorso-ventral gradient (Table 4).

7. Discussion

7.1. Anatomical organization of the ICCN There are no detailed ultrastructural studies of the ICCN of the C57 mouse currently available. Rockel and Jones (1972a, b) studied the ultrastructure of the ICCN in the cat. They described the soma and proximal dendrites of principal cells as being covered by axon terminals. Many of these were terminals of the lateral lemniscus which contained large spherical synaptic vesicles and made asymmetric synapses. Paloff et al. also studied the fine structure of the ICCN in the cat (Paloff et al., 1989; Paloff and Usunoff, 1992). They found that large neurons ( > 22 /xm) had numerous (up to 25/ce11) axosomatic synapses of both the asymmetric and symmetric type. Smaller neurons ( < 22 /xm) had much fewer axosomatic synapses. They described 9 different types of synapses, based on the size and shape of the terminals, the size, shape and distribution of the vesicles, the characteristics of the synaptic zone, and the postsynaptic target. In the ICCN of the rat, Ribak and Roberts (1986) described findings similar to those in the cat. The largest neurons ( > 25 /xm), whether they were disc-shaped (principal cells) or stellate (multipolar cells) had many axosomatic synapses, mostly of the symmetric type, while neurons 15-25 /xm in diameter had sparse somatic synapses. They described the largest neurons with synapses covering the vast majority of the somatic surface, 30-40/ce11. These largest neurons projected to the medial geniculate

Table 4 Split-plot factorial analysis of synaptic loss by age and location within ICCN by F test (with levels of significance) Age(F-A) Location(F-B) Synapses/neuron (n) 11.28 (0.01) 2.47 (NS) medio-lateral Synapses/perimeter(%) 9.22 (0.01) 1.71 (NS) medio-lateral Synapses/neuron (n) 13.72 (0.01) 9.00 (0.05)dorso-ventral Synapses/perimeter(%) 13.49 (0.01) 8.04 (0.05)dorso-ventral

Interaction(F-AB) 1.00 (NS) 0.63 (NS) 2.07 (NS) 1.55 (NS)

118

A.M. Kazee et al. / Hearing Research 89 (199.5) 109-120

body and were excitatory. Most of the small neurons ( < 15 /zm) were interneurons and used gamma-aminobutyric acid (GABA) as their neurotransmitter (Roberts and Ribak, 1987). GABA is the major inhibitory transmitter of the IC. GABA terminals contain pleomorphic or flattened vesicles and make symmetric synapses. Terminals with small round vesicles were described by Oliver as arising from fusiform cells of the dorsal cochlear nucleus (Oliver, 1985). We describe principal neurons with many fewer axosomatic synapses than in the cat or rat with only 1-2% of the somatic surface being covered by synapses. Axosomatic synapses of both the asymmetric as well as symmetric type were identified, although the symmetric type appeared to be more common. Both types of axosomatic synapses diminished with aging. Although the number of axosomatic synapses per principal cell soma was small in a 2-dimensional micrograph, the impact of a 63% decrease in synaptic number and a 67% decrease in synaptic coverage over the entire surface area of the somatic membrane may be quite significant functionally. Willott et al. (1994) has previously studied aging in the ICCN at the light microscopic level in the C57 mouse. They found little effect of aging with no change in the total volume of the IC nucleus or in the volume of the IC neuropil. They also saw no age-related neuronal loss or change in the size of IC neurons. This last finding is not confirmed in the present study, where we noted a significant decrease in all neuronal parameters measured (neuronal area, nuclear area, nucleolar area, major axis diameter, minor axis diameter, and somatic membrane perimeter). The discrepancy with regard to the decrease in neuronal size observed with aging between the present study and that of Willott may be due to methodological differences between the two studies. The methods of fixation (formalin versus paraformaldehyde/glutaraldehyde), tissue preparation (paraffin versus Spurr epoxy resin) and staining (cresyl violet versus toluidine blue) were all different. Furthermore, the sections were 10/xm thick in Willott's study but only 1.5 /xm thick in the present study. We measured only principal neurons while Willott measured all inferior colliculi neurons. This difference in which neurons were being measured suggests another explanation for the discrepancy. Namely, that the decrease in principal cell size was being masked by multipolar cells which did not change and that only principal cells were affected (personal communication, Willott). Our major finding is the extensive synaptic loss which correlates well with the chronic progressive hearing loss in the C57 mouse. This massive loss of synapses on the principal cell somas of the ICCN begins in middle age when the C57 mouse is already beginning to experience profound sensorineural hearing loss in the high-frequency range. The dramatic loss of hair cells seen throughout much of the cochlea could be a major contributor to the effects seen in the ICCN. But, this synaptic loss was noted

at all locations in the tonotopic organization of the ICCN. The loss could not be temporally related to the progression of hearing loss (high to low frequency) noted in both the cytocochleograms and auditory brainstem recordings. No specific type of synaptic terminal appeared to be preferentially lost in this study, a finding which may be due to the limited sampling which is feasible on ultrastructural morphometry and the emphasis axosomatic synapses. Auditory nuclei, other than the ICCN, have been examined anatomically in various species with regard to aging. The changes that have been identified morphologically in the cochlear nucleus can be summarized as follows. (1) DCN layer III, which receives primary cochlear afferents (Martin, 1981; Browner and Baruch, 1982; Ryugo and Willard, 1985) shows a decrease in neuronal number and size, with some decrease also observed in DCN volume (Willott et al., 1992). Layers I and 1I show little age-related change prior to the terminal stage suggesting that the age-related changes in layer III are due to the progressive peripheral hearing loss and degeneration of spiral ganglion cells. (2) The octopus cell area (OCA) in the posterior VCN (PVCN) shows age-related changes only in extreme old age despite the heavy primary afferent input to the area (Harrison and Irving, 1966; Kane, 1973). There is a decrease in OCA volume due to a decrease in both the number and size of neurons and an increase in glial density (Willott and Bross, 1990). (3) The anterior VCN (AVCN) in the C57 mouse shows no change in volume but there were decreases in neuron number and density between 1 and 7 months (Willott et al., 1987). These parameters remained stable thereafter, despite the chronic severe hearing loss. Ultrastructurally, in the dorsal AVCN (high-frequency region), there was a loss in the percentage of somatic surface apposed by terminals (Briner and Willott, 1989). Degenerating neurons were not observed and many ultrastructural features were robust despite aging and chronic deafferentation. Webster (1983a Webster (1983b) noted that if the hearing loss occurred in an immature animal, there was cell loss in the AVCN, but not if the hearing loss occurred in a mature animal. Elsewhere in the central auditory system, Casey and Feldman (1988) found that the synaptic input of primary calycine endings on the soma of principal cells in the medial nucleus of the trapezoid body decreased 18% in aged Sprague-Dawley rats. Non-primary input did not change with aging. The length of the calycine and noncalycine endings did not change, nor the size of the neuronal soma, confirming that the result was not due to age-related shrinkage. Feldman (1982) also showed synaptic loss in the rostral cochlear nucleus with aging. In 3-month-old Sprague-Dawley rats, he estimated 60-80% of somatic surfaces were contacted by axon terminals (about 25% of these contacts were by primary calyceal end bulbs of Held). After 27 months of age, the total coverage

A.M. Kazee et aL / Hearing Research 89 (1995) 109-120

was reduced to 40%, with preferential loss of the end bulbs of Held. There was no evidence of reinnervation by nonprimary synapses. The significance of the extensive synaptic loss with aging which the C57 mouse exhibits is likely due to the chronic progressive hearing loss which is genetically predetermined in this mouse species. The loss occurs during middle age when the primary denervation is occurring in the peripheral auditory system. This synaptic loss continues to be seen in very old mice but is not progressively worse, therefore, is probably n o t just a phenomenon of age. The synaptic loss may be related to the dramatic physiologic changes which are occurring throughout the ICCN in the C57 mouse, in which high-frequency units become more sensitive to low-frequency sounds. Identification of the cells of origin of the different types of synaptic terminals would be helpful in determining the significance of the synaptic loss.

Acknowledgements This work was supported in part by funding from Grant P01 AG09524 from the National Institutes of Health, and the International Center for Hearing and Speech Research (RICHS), Rochester, New York. We gratefully acknowledge the critical reading of the manuscript and the suggestions of Dr. Robert D. Frisina.

References Briner, W. and Willott, J.F. (1989) Ultrastructural features of neurons in the C 5 7 B L / 6 J mouse anteroventral cochlear nucleus: young mice versus old mice with chronic presbycusis. Neurobiol. Aging 10, 295 -303. Browner, R.H. and Baruch, A. (1982) The cytoarchitecture of the dorsal cochlear nucleus in the 3 - m o n t h - a n d 26-month-old C57BL/6J mouse: a Goigi study, J. Comp. Neurol. 211, 115-138. Casey, M.A. and Feldman, M.L. (1988) Age-related loss of synaptic terminals in the rat medial nucleus of the trapezoid body. Neuroscience 24, 189-194. Curcio, C.A. and Sloan, K.R., Jr. (1986). Computer-assisted morphometry using video-mixed microscopic images and computer graphics. Anat. Rec. 214, 329-337. Faye-Lund, H. and Osen, K.K. (1985) Anatomy of the inferior colliculus in rat. Anat. Embryol. 171, 1-20. Feldman, M.L. (1982) Synaptic loss with advanced age in the cochlear nucleus. Soc. Neurosci. Abstr. 7, $67. Greenwood, D.D. (1990) A cochlear frequency-position function for several species - - 29 years later. J. Acoust. Soc. Am. 87, 2592-2605. Harrison, J.M. and Irving, R. (1966) The organization of the posterior ventral cochlear nucleus in the rat. J. Comp. Neurol. 162, 269-284. Henry, K.R. and Chole, R.A. (1980) Genotypic differences in behavioral physiological and anatomical expressions of age related hearing loss in the laboratory mouse. Audiology 19, 369-383. Henry, K.R. and Lepkowski, C.M. (1978) Evoked potential correlates of genetic progressive hearing loss. Age-related changes from the ear to the inferior colliculus of C 5 7 B L / 6 and C B A / J mice. Acta Otolaryngol. 86, 366-374.

119

Kane, E.C. (1973) Octopus cells in the cochlear nucleus of the cat: heterotypic synapses on homeotypic neurons. Int. J. Neurosci. 5, 251-279. Li, H.-S. and Borg, E. (1991) Age-related loss of auditory sensitivity in two mouse genotypes. Acta Otolaryngol. 111, 827-834. Martin, M. (1981) Morphology of the cochlear nucleus of the normal and reeler mutant mouse. J. Comp. Neurol. 197, 141-152. Meininger, V., Pol, D. and Derer, P. (1986) The inferior colliculus of the mouse: a Nissl and Golgi study. Neuroscience 17, 1159-1179. Mikaelian, D.O. (1979) Development and degeneration of hearing in the C57B16 mouse: Relation of electrophysiologic responses from the round window and cochlear nucleus to AVCN anatomy and behavioral responses. Laryngoscope 89, 1-15. Morest, D.K. and Oliver, D . L (1984) The neuronal architecture of the inferior colliculus in the cat: Defining the functional anatomy of the auditory midbrain. J. Comp. Neurol. 222, 209-236. Oliver, D.L. (1985) Quantitative analyses of axonal endings in the central nucleus of the inferior colliculus and distribution of 3H-labeling after injections in the dorsal cochlear nucleus. J. Comp. Neurol. 237, 343-359. Oliver, D.L. and Morest, D.K. (1984) The central nucleus of the inferior colliculus of the cat. J. Comp. Neurol. 222, 237-264. Paloff, A.M. and Usunoff, K.G. (1992) The fine structure of the inferior colliculus in the cat. II. Synaptic organization. J. Hirnforsch. 32, 77-106. Paloff, A.M., Usunoff, K.G., Hinova-Palova, D.V. and Ivanov, D.P. (1989) The fine structure of the inferior colliculus in the cat. I. Neuronal perikarya in the central nucleus. J. Hirnforsch. 30, 69-90. Ribak, C.E. and Roberts, R.C. (1986) The ultrastructure of the central nucleus of the inferior colliculus of the Sprague-Dawley rat. J. Neurocytol. 15, 421-438. Roberts, R.C. and Ribak, C.E. (1987) An electron microscopic study of the GABAergic neurons and terminals in the central nucleus of the inferior colliculus of the rat. J. Neurocytol. 16, 333-345. Rockel, A.J. (1971) Observations on the inferior colliculus of the adult cat stained by the Golgi technique. Brain Res. 30, 407-410. Rockel, A.J. and Jones, E.G. (1972a) The neuronal organization of the inferior colliculus of the adult cat. I. The central nucleus. J. Comp. Neurol. 147, 11-60. Rockel, A.J. and Jones, E.G. (1972b) Observations on the fine structure of the central nucleus of the inferior colliculus of the cat. J. Comp. Neurol. 147, 61-92. Ryugo, D.K. and Willard, F.A. (1985) The dorsal cochlear nucleus of the mouse: a light microscopic analysis of neurons that project to the inferior colliculus. J. Comp. Neurol. 242, 381-396. Semple, M.N. and Aitkin, L. (1979) Representation of sound frequency and laterality by units in central nucleus of cat inferior colliculus. J. Neurophysiol. 42, 1626-1639. Shnerson, A., Devigne, C. and Pujol, R. (1982) Age-related changes in the C57BL/6J mouse cochlea. II. Ultrastructural findings. Dev. Brain Res. 2, 77-80. Shnerson, A. and Pujol, R. (1982) Age-related changes in the C57BL/6J mouse cochlea. I. Physiological findings. 2, 65-75. Shnerson, A. and Willott, J.F. (1979) Development of inferior colliculus response properties in C 5 7 B L / 6 J mouse pups. Exp. Brain Res. 37, 373-385. Stiebler, I. and Ehret, G. (1985) Inferior colliculus of the house mouse. I. A quantitative study of tonotopic organization, frequency representation, and tone-threshold distribution. J. Comp. Neurol. 238, 65-76. Syka, J., Druga, R., Popelar, J. and Kalinova, B. (1981) Functional organization of the inferior colliculus. In: J. Syka and L. Aitkin (Eds.), Neuronal Mechanism of Hearing, Plenum, New York, pp. 137-153. Walton, J., Frisina, R. and Meierhans, L. (1995) Sensorineural hearing loss effects recovery from short term adaptation in the CBA and C 5 7 B L / 6 models of presbycusis. Hearing Res., 1995.

120

A.M. Kazee et aL / Hearing Research 89 (1995) 109-120

Webster, D.B. (1983a) Auditory neuronal sizes after a unilateral conductive hearing loss. Exp. Neurol. 79, 130-140. Webster, D.B. (1983b) Late onset of auditory deprivation does not affect brainstem auditory neuron soma size. Hearing Res. 12, 145-147. Willott, J.F. (1984) Changes in frequency representation in the auditory system of mice with age-related hearing impairment. Brain Res. 209, 159-162. Willott, J.F. (1986) Effects of aging, hearing loss and anatomical location on thresholds of inferior colliculus neurons in C57BL/6 and CBA mice. J. Neurophys. 56, 391-408. Willott, J.F. and Bross, L.S. (1990) Morphology of the octopus cell area of the cochlear nucleus in young and aging C57BL/6J and C B A / J mice. J. Comp. Neurol. 300, 61-81. Willott, J.F. and Shnerson, A. (1978) Rapid development of tuning characteristics of inferior coUiculus neurons in mouse pups. Brain Res. 148, 230-233. Willott, J.F., Pankow, D., Hunter, K. and Kordyban, M. (1985) Projections from the anterior ventral cochlear nucleus to the central nucleus of the inferior colliculus in young and aging C57BL/6 mice. J. Comp. Neurol. 237, 545-551.

Willott, J.F., Jackson, L.M. and Hunter, K.P. (1987) Morphometric study of the anteroventral cochlear nucleus of two mouse models of presbycusis. J. Comp. Neurol. 260, 472-480. Willott, J.F., Parham, K. and Hunter, K.P. (1988) Response properties of inferior colliculus neurons in middle-aged C57BL/6J mice with presbycusis. Hearing Res. 37, 15-28. Willott, J.F., Parham, K., Hunter, K.P. (1991) Comparison of the auditory sensitivity of neurons in the cochlear nucleus and inferior colliculus of young and aging C57BL/6J and C B A / J mice. Hearing Res. 53, 78-94. Willott, J.F., Bross, L.S. and McFadden, S.L. (1992) Morphology of the dorsal cochlear nucleus in C57BL/6J and C B A / J mice across the life span. J. Comp. Neurol. 321, 666-678. Willott, J.F., Aitkin, L.M. and McFadden, S.L. (1993) Plasticity of auditory cortex associated with sensorineural hearing loss in adult C57BL/6J mice. J. Comp. Neurol. 329, 402-411. Willott, J.F., Bross, L.S., McFadden, S.L. (1994) Morphology of the inferior colliculus in C57BL/6J and C B A / J mice across the lifespan. Neurobiol. Aging 15, 175-183.