- Email: [email protected]
Refinement of dendritic arbors along the tonotopic axis of the gerbil lateral superior olive
Brain Research, 67 (1992) 47-55 I~) 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-89931921505.00
47
BRESD 51440
Refinement of dendritic arbors along the tonotopic axis of the gerbil lateral superior olive Dan H. Sanes, John Song and James Tyson Dept~rtments of Otolaryngology, Physiology and Biophysics, New York University School of Medicine, New York, NY 10016 (USA)
(Accepted 31 December 1991) Key words: Auditory pathway; Development; Frequency tuning; Lateral superior olive; Morphometry; Dendrite
We have investigated the development of dendritic arbors in a central auditory nucleus in the Mongolian gerb.il, the lateral superior olive (LSO). The morphology of these arbors has been shown to vary with tonotopic position i:~ adults, with high frequency neurons having a more restricted field. In the prcscnt study, qualitative observations were made on horseradish peroxidase-filled neurons from animals 1-11 days postnatal, and quantitative results were obtained from Golgi-impregnated material from animals 10 days postnatal and older. The tonetopic position of each cell was computed as a percent of the total distance along the LSO. The dendritic arbors of high frequency neurons became spatially constrained along the frequency axis during the 3rd postnatal week, while those in the low frequency region retained a broader arborization into adulthood. This refinement was correlated with a decrease in total dendritic length and the number of branch points per neuron, particularly in the high frequency projection region. The distribution of octave bandwidths t,a which single LSO neurons responded in 13-16 day animals showed a similar course of maturation across the tonotopic axis: high frequency neurons responded to a larger number of octaves, and with greater variability, than those in adults. These data suggest that a specific alteration in dendrite morphology, which occurs after the onset of response to airborne sound, may contribute to adult frequenc) selectivity. INTRODUCTION The modification of dendritic processes during maturation has been a common observation in a wide range of neuronal systems. The proximal cause for this phenomenon is thought to be related to the finite availability of afferent innervation a2. However, it is not clear, from the systems thus far examined, whether these reductions in dendrite structure are functionally meaningful to the neuron. In the present study we focus on a discrete alteration of central auditory neuron dendrites that largely occurs following the onset of response to airborne sound. While these alterations reflect only a fraction of the intra- and intercellular forces acting during dendrogenesis, they offer an expedient assay system for experimental manipulations. The morphology of neurons within the lateral superior olive (LSO) have been well described in several species 12'21'29'34'43'46. The major neuronal class, the principal cells, compose approximately 80% of the neuronal population in adult cats 2~, and have modest dendritic trees that are oriented perpendicular to the tonotopic axis 42'43'46'53. The electrophysiological properties of LSO neurons also appear to be relatively uniform, being excited by ipsilateral sound stimuli and inhibited by contralateral sound stimuli 7's'9A5'19.39.
We have previously shown that the dendritic arb6dzations of principal neurons in the gerbil LSO are mote constrained along the tonotopic axis at successively higher frequency projection regions, and that this morphological variation correlates with the octave frequency range of single LSO neurons 4a. In addition, the results of a functional study indicate that frequency selectivity within the LSO improves during the period when airborne sound first evokes a neural response, and that this improved selectivity cannot be solely attributed to cochlear maturation 39. An analysis of the inhibitory receptor molecule, the glycine receptor 41, and single inhibitory afferents from the medial nucleus of the trapezoid body in the LSO 45 suggest that a pre- and postsynaptic modification could contribute to the functional maturation. In the present study, we present further evidence that the refinement of postsynaptic morphology in the LSO may play a role in functional maturation. MATERIALS AND METHODS Golgi impregnations Tissue was prepared according to the rapid Golgi technique of Adams (1979). Mongolian gerbils (Meriones unguiculatus) aged 10-21 days of either sex were given a lethal dose of anesthetic (sodium pentobarbital; 80 mg/kg; i.p.), and perfused transcardially with 0.9% NaCI, 10% buffered formalin, and potassium dichromate mordant. The brains were stored in mordant for 2-4 days, and
Correspondence: D.H. Sanes, Center for Neural Science, New York University, 4 Washington Place, New York, NY 10003, USA. Fax: (1) (212) 995-4011.
48 transferred to a silver nitrate solution for 2-4 days. Transverse vibratome sections at 150 pm were cleared in methyl salicylate, and mounted on gelatin-coated slides.
Horseradish peroxidase iontophoresis Brain slices of 400 pm from gerbils aged 2-11 days postnatal were prepared as described previously44. The tissue was placed in a chamber, and superfused with artificial cerebrospinal fluid (ACSF (in raM): NaC! 127.4, KCi 5, KH2PO 4 1.2, MgSO4.7H20 1.3, NaHCO 3 26, glucose 15, CaCI2.2H20 2.4, pH = 7.4). Electrodes filled with 6% HRP in 0.05 M "Iris and 0.5 M KCI (pH = 7.6) were beveled to a final resistance of 200-300 MQ, and positioned over the LSO. The parameters for HRP iontophoresis were 200 ms cathodal pulses at 0.8-1.5 nA delivered at 3 Hz for 1-3 rain. The slices were left to transport the HRP for 2.5 h after the last neuron was injected. The brain slices were placed in fixative (1% paraformaldehyde, 2.5% glutaraldehyde, 0.12 M phosphate buffered saline) for 2 h, rinsed in phosphate-buffered saline (PBS) at 4°C for 12-48 h, and washed in PBS with 0.5% Triton X-100 for 2 h. Each slice was then incubated in a filtered solution of 0.05% diaminobenzidine (DAB) with 0.02 % CoC!2 in 0.1 M Tris buffer2 for 30 min, and transferred to a fresh filtered DAB solution containing H202 for 5-15 min. Slices were placed onto gelatin-subbed slides, and dehydrated, cleared in xylenes, and mounted for viewing.
from 23 HRP-filled n e u r o n s f r o m 10 animals using the brain slice p r e p a r a t i o n . For the purposes of this study, the p a r a m e t r i c measures were p o o l e d for n e u r o n s with a position < 5 0 % from the apical m o s t projection r e g i o n or --_50% f r o m the apical most p r o j e c t i o n region. This corr e s p o n d s to the area of the L S O that responds to app r o x i m a t e l y 4 kHz 42. T h e s e 2 groups will h e r e a f t e r be r e f e r r e d to as the low f r e q u e n c y n e u r o n s a n d high freq u e n c y n e u r o n s , respectively. T h e data from Golgi-filled tissue was g r o u p by age at 10-12 days (n - 12 low a n d 32 high), 13-14 days (n = 24 low a n d 33 high), 15-16 days (n = 26 low and 22 high), 17-19 days (n = 27 low and 23 high), and 21-22 days (n = 29 low and 31 high), and the s a m p l e d n e u r o n s were e v e n l y distributed a l o n g
Image analysis and 3-D reconstruction Oolgi impregnated neurons in the LSO were selected for quantitative measurement as previously described43. Briefly, principal neurons with distinguishable processes, and no significant cut ends were observed with a light microscope (Zeiss Standard 16; Planapo 40 oil), and displayed on a color monitor (MTI-Dage Series 68 Neuvicon Video Camera). This selection procedure was meant to exclude multiplaner and marginal neurons from the analysis. Data acquisition and analysis was accomplished by using custom designed software for computer aided quantification and 3-D reconstruction (Cellmate/Treemate, R&M Biometrics) ~. The number of primary dendrites, the number of branch points, the total length of all dendritic segments, and the transverse soma area were all computed by the Treemate software. Measures of arbor spread along the tonotopic axis were also obtained from calibrated plots of the neurons, The tonotopic position of each neuron was computed as percent distance from the a.~ical most projection region (i,e, ventrolateral; see Sanes et al, 4", Fig, [B), and parametric measures were plotted against each neuron's position,
I
'%.
•4 . ,
Auditory neurophysiology The dehvery' of acoustic stimuli and recording. ~°f single neuron response properties have been described in detad ~ . Gerbils, aged 13-16 days, were studied under ketamine and chloral hydrate anesthesia such that nociceptive reflexes were eliminated. Sound stimuli were presently bilaterally through a calibrated, closed delivery system within a sound attenuated chamber. Single neurons were isolated with glass micropipettes, and responses to tonal stimuli of 50 ms duration with a 4 ms rise/fall time were led into a window discrimator for on-line data analysis. Response areas derived from monaural ipsilateral stimuli were generated for neurons that were determined to be ipsilaterally excited and contralaterally inhibited. Neuron location was determined histologically after iontophoresis of Fast green. The lowest and highest frequency to evoke a reliable increase in discharge rate at 30 dB above threshold was then determined, and this frequency range was converted to an octave band about the characteristic frequency. RESULTS T h e quantitative results o f this study are d r a w n from 259 n e u r o n s taken from 93 M o n g o l i a n gerbil brains sectioned in the transverse plane, and the qualitative results
Fig. tion Bar ron.
1. A neuron in the LSO of a 1 day animal. Top: a reconstrucof the neuron showing its position within the LSO boundary. - 100 pro. Bottom: a photomontage of the HRP-filled neuBar - 50 gm.
49 the transverse axis of the LSO. In addition, we compared values previously obtained from adult animals 43.
of growth through the first 2 postnatal weeks.
Development of arbor spread and frequency selectivity The morphology of a low and high frequency neuron
Initial postnatal arborization Figs. 1-4 show examples of HRP-fiUed LSO neurons at several postnatal ages. At 1-3 days, the HRP-filled neurons displayed a profusion of processes emanating from the soma, and a considerable number of branches (Figs. 1 and 2). Even at this early stage some neurons displayed a preferential orientation ~!ong the presumptive isofrequency laminae of the LSO (Fig. 1). At 10-11 days, LSO neurons displayed a high degree of branching, yet occupied a variable fraction of the LSO nucleus (Figs. 3 and 4). Although the numbers of HRP-filled neurons were not sufficient for quantitative analyses, their morphologies were consistent with a rapid period
/
I
\
\ .~
are illustrated along with reconst~'uctions rotated about the transverse axis for 14-15 day animals (Figs. 5 and
....... ~
o ,
i
"" ,..
:'
~,,
'/ f
i J
/
'~. i
/
.(
/
t/
)
\, /
/
~'%.. '~,, "L..
.j
%
0
\,
°
.
,///' .x ,.4
. . ~
~
....
t
i' ol /
J
I
•
Q ,\. m Fig. 2. A neuron in the LSO of a 3 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary. Bar = 100/am. Bottom: a photomontage of the HRP-filled neuron. Bar = 50/~m.
I[
Fig. 3. A neuron in the LSO of a 10 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary. Bar = 100/~m. Bottom: a photomontage of the HRP-fiiled neuron. Bar = 50 ~tm.
50 6), and 21 d~'~yanimals (Figs. 7 and 8). In Golgi-stained material, it was found that the distance that LSO dendritic arborizations spread across the tonotopic axis became more refined during development, but only for high frequency neurons (Fig. 9). This distance decreased significantly between 13-14 and 21-22 days from 71 + 3.7/~m (aS + S.E.M.) to 49 + 3.5 (t = 4.24; P < 0.0001; df = 62). There was no further change in animals aged _> 90 days. The spread of low frequency dendritic arbors did not vary significantly from the earliest age examined to the adult (Fig. 9). The octave bandwidths of neurons at 13-16 days were
calculated at 30 dB above threshold, and a comparison was made to the maturation of dendrite form along the frequency axis. As illustrated in Fig. 10, the octave bands of high frequency neurons were greater, and much more variable, in 13-16 day animals compared to adults. However, it must be noted that the octave bandwidths calculated for neurons in adult animals were computed at 50-60 dB above threshold. The octave bands of low frequency neurons had the same distribution in 13-16 day and adult animals. The change in distribution of octave bandwidths from 13-16 day to adult animals bears a striking resemblance to the change in distribution of dendritic arborizations across the frequency axis (Fig. 10).
Total dendritic length and number of branch points t' J .....~-~-.,
I
%
..
'l .,
i. t
\
/
t
i •,./'
The total dendritic length of high frequency LSO neurons first increased from 885 _ 60/~m at 10-12 days to 1082 + 70/tm at 13-14 days postnatal (t = -2.13; P < 0.05; df -- 60). As shown in Fig. 9, there was a significant reduction in total dendritic length to a value of 865 +_ 47/~m by 21-22 days (t -- 2.12; P < 0.05; df = 51), and a further reduction thereafter. There was no significant variation in the total dendritic length of low frequency LSO neurons across age. The mean number of branch points per neuron decreased significantly between 13-14 days and 90+ days postnatal, but was more apparent for high frequency neurons (Fig. 9). The values changed from 11.7 + 1 (~ ± S.E.M.) to 4.5 ± 0.3 branches for high frequency neurons (t - 8.04; P < 0.0001; df = 76), and from 7.6 ± 0.6 to 4.7 ± 0.4 branches for low frequency neurons (t - 4.06; P < 0.0005; df - 61). The number of branch points was consistently greater for high frequency neurons at all ages through 21-22 days, but was equivalent in adult animals. Among the parameters that did not vary significantly between 10-12-day- and adult animals were the number of primary dendrites (i.e. arising from the cell body), the cross-sectional area of the soma, and the distance that dendrites extended perpendicular to the tonotopic axis (data not shown). DISCUSSION
!! or
Fig. 4. A neuron in the LSO of an 11 day animal. Top: a reconstructior of the neuron showing its position within the LSO boundary. Bar -- 100 l~m. Bottom: a photomontage of the HRP-filled neuron. Bar -- 50 urn.
We have demonstrated a refinement of principal cell dendritic arborizations that appears largely restricted to one area of the LSO, the high frequency projection region. This morphological transformation is correlated with an improvement in the frequency selectivity of single LSO neurons over the same age range. The structural refinement of high frequency dendrites resulted from regressive events including the loss of branch points, and a decrease in total dendritic length.
51
/
\ I !
",,
] ;
•,-, i"
/
l' "-t.
a, -¸~. l~ .it,...
Fig. 5. A neuron in the LSO of a 14 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar = 100/~m. Bottom: a photomontage of the Golgi-stained neuron. Bar = 50/~m.
Similar modifications of dendrite form have been described in a wide variety of systems. These regressive events include a decrease in the number of spines 2¢'33, a decrease in the total process length 23'5s, a decrease in the number of primary dendrites 5°, and a decrease in the number of branch points 26'36. The physical dimensions occupied by a dendritic field may also change during development 35. It has been shown that dendrogenesis is partially related to afferent density or synaptic activity5'2s'3°'31'54'55, the availability of postsynaptic targets 57, the interaction with a specific substrate 3't3, and intrinsic cues 4'24'51'52. The concomitant change in dendrite morphology and frequency selectivity during development could result from an activity-mediated process. However, since the observed changes in dendrite form were largely restricted to the high frequency projection region of the LSO (Fig. 10), one would have to postulate a different level or pattern of synaptic activity in the low and high frequency
Fig. 6. A neuron in the LSO of a 15 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar 100/~m. Bottom: a photomontage of the Golgi-stained neuron. Bar = 50/~m.
regions. There is a small amount of evidence for a difference in spontaneous and evoked activity along the frequency axis. Measures of auditory nerve spontaneous activity in the adult gerbil demonstrate that a larger percentage of fibers with characteristic frequencies above 5 kHz have very low discharge rates 47. There is no reported difference in sound evoked activity or metabolism along the tonotopic axes of the gerbil central auditory nuclei 37'39' 49.~o. However, sound evoked 2-deoxyglucose uptake is conspicuously absent from the low frequency projection region at the onset of response to airborne sound 38. Whether or not these physiological factors play a role in the pattern of dendrogenesis seen in the gerbil LSO awaits experimental analysis. There is some evidence that the local environment differs between the high and low frequency projection
52
".'.\ •.
' ~ .
~
7
~
I[
,..
',.
,
"".
,i
[
~ -° . . . . . . . . -
.,
I
Fig. 8. A high frequency neuron in the LSO of a 21 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 300 intervals about the sagittal plane. Bar = 100 ~m. Bottom: a photomontage of the Golgistained neuron showing a highly restricted arborization along the tonotopic axis. Bar = 50 # m .
I
Ill Ill
Fig. 7. A low frequency neuron in the LSO of a 21 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar = 100 ~m. Bottom: a photomontage of the Oolgistained neuron illustrating the relatively large spread along the tonotopic axis compared to the ~euron shown in Fig. 8. Bar = 50
regions. The packing density of neuronal somas is nearly double in the high frequency region, and the density of glycine receptors is approximately fourfold greater 'm. In contrast, there is a greater density of astrocytes in the low frequency region Is, and a greater number of CAT301 positive neurons as well4~, Therefore, it is possible that dendritic branching patterns reflect 2 distinct extracellular environments that influence process outgrowth and stability.
Does dendrite form influence frequency tuning? The maturation of stimulus coding within the gerbil auditory system has been described for both the cochlea and cochlear nucleus s9'6°. These authors found an electric response to sound, as evidenced by a cochlear microphonic or N~ compound action potential, as early as postnatal day 12. In addition, Woolf and Ryan ~° demonstrated that high frequency neurons in the cochlear nucleus have adult-like Qlo values from the onset of hearing, whereas the tuning of low frequency neurons continued to improve. The contrast with the present results may be explained in 2 ways. First, measures of tuning taken near the neurons absolute threshold (e.g. Qto) may not be as sensitive to the convergence of afferents as would be measures taken at 30-60 dB above threshold (e.g. octave bandwidths used in the present study). Second, in young animals the appearance of
53 Dendrite Width (A) o
I ...... L...... !
.k.,
I
-,:,=-
60"
40
i
~+
'|
/ ~ ~o* *° * •
Adult
:t. !
oo
13-16
8O
E
tt
15or
100
1001 -~
/:
,
~
o
~o
/
o o.
~p o
"~,,,,o
~6" o .°.,~i#~
10-12 13"14 15"~16 1¥-19 21"22 90.
Age (days postnatal) 0
Branches
/
I
I
I
100 % Distance from Apical End
I
12~
I
0
25
50
75
10; -----
,ID E
-
-
-
High 7
Low o
13-16 Adult
6
s
xF= 13:14 ,m',le it'-Ill e~e= Q~I. '"-'
o
o go OOOOO o ~ ~ % R~b°~7 ~°° o
o
Age (days postnatal)
3 Total Dendritic Length
I
1100"
o
I
liO0'
E
....
.o°'°'"
......
o,
O ,
::L 700'
600
~ ~.0o O~
o~ ~
~o ~ o ~
i
i
1
i . .....
2
~
10;1218'o14lli'-lil l#'-lli li-'illl lib. Age (days postnatal)
Fig. 9. Parametric changes in dendrit'c arbors from the high frequency projection region (solid line) and low frequency projection region (dashed line) of the LSO. Top: the dendritic width (A) along the frequency axis remains unchanged for low frequency neurons, but decreases significantly between 13 and 19 days postnatal. Middle: the number of branch points per neuron decreases for both high and low frequency neurons between 15 and 904- days. Bottom: the total dendritic length remains unchanged for low frequency neurons, but decreases significantly between 15 and 904days postnatal for high frequency neurons. Adult values taken from Sanes et al. 'u. adult-like functional properties are commoilly found in conjunction with immature responses, and the process of maturation consists of probabilistic refinements within the whole population. In this regard, it is interesting to note the distribution of octave bands in the high frequency region of 14-16 day animals. Neurons appeared to exhibit two extremes in tuning. On the other hand, the distribution of dendrite widths appeared as an uniform shift in the distribution towards larger values. It should be noted that the afferent arborizations are becoming refined during the same time period 4s. Therefore, it is possible that the physiological result is magnified by the immaturity of
0
~
oo ~o ~ i
i
100 % Distance from Apical End 25
50
75
Fig. 10. A comparison of LSO dendritic width along the frequency axis (Top) and frequency selectivity (bottom) at 13-16 days postnatal and in the adult. The frequency axis has been converted to % distance from the apical most projection region to correct for changes in the cochlear tonotopic map during development2°'42. Top: the dendritic width is virtually identical from 0 to 50% of the distance from the apical most projection region. However, above 50% the dendritic width is generally greater for 13-16 day animals. Bottom: the octave bands largely overlap from 0 to 50% of the distance from the apical most projection region, but are much greater for 13-16 day animals above 50%. Adult values taken from Sanes et al.43.
other elements in the circuit, including the cochlea. Although the frequency response of central auditory neurons is primarily dependent upon cochlear processing, there is evidence to indicate that the convergence of afferents is also a factor. For example, neurons in the cat cochlear nuclei exhibit both excitatory and inhibitory frequency response areas from separate afferent populations 17'25's6. If the dendritic field dimensions regulate the number of afferents that will form synaptic contacts, then they may also contribute to the frequency selectivity. There is a precedent for dendritic morphology influencing a specific visual coding property. In the rabbit retina, orientation-sensitive ganglion and amacrine cells can have markedly asymmetric dendritic arborizations that correlate with the preferred angle of a bar stimu-
54 iUS6.t°. In addition, there have been 2 reports of central auditory neurons showing a preferential orientation along the frequency axis, but there was no direct correlation with frequency tuning 16'27. It should be stressed that the emergence of 2 different dendritic arborization patterns may be related less to frequency tuning than to the binaural processing strategies of the LSO. For example, recent studies have demonstrated a strong temporal dependency of LSO neurons, particularly for low frequency cells 14'22. The broader spread of both dendrites and inhibitory arborizations 45 across the frequency axis in the low frequency region may subserve the phase-sensitivity of these cells. Low frequency LSO neurons may sample from a maximum number of phas,.Mocking afferents irrespective of their characteristic frequency.
REFERENCES 1 Adams, J.C., A fast, reliable silver-chromate Golgi method for pedusion-fixed tissue. Stain Technoi., 54 (1979) 225-226. 2 Adams, J.C., Heavy metal intensification of DAB-based reaction product, J. Histochem. Cytochem., 29 (1981) 775. 3 Adler, R. and Varon, S., Neuritic guidance by nonneuronal cells of ganglionic origin, Dev. Biol., 86 (1981) 69-80. 4 Banker, G.A. and Cowan, W.M., Further observations on hippocampal neurons in dispersed cell culture, J. Comp Neurol., 187 (1979) 469-494. 5 Berry, M., McConnell, E and Sievers, J., Dendritic growth and the control of neuronal form, Current Topics Dev. Biol., 15 (1980) 67-102. 6 Bloomfield, S.A., Two types of orientation-sensitive responses of amacrine cells in the mammalian retina, Nature, 350 (1991) 347-350. 7 Boudreau, J.C. and Tsuchitani, C., Binaural interaction in the cat superior olive s-segment, J. Neurophysiol., 3t (1968) 442-451. 8 Boudreau, J.C. and Tsuchitani, C., Cat superior olive s-segment cell discharge to tonal stimulation, in W.D. Neff (Ed.), Contri. butions to Sensory Physiology, Vol. 4, Academic Press, New York, 1970, pp, 143-213. 9 Caird, D. and Klinke, A.R., Processing of binaural stimuli by cat superior olivary complex neurons, Exp. Brain Res., 52 (1983) 358-399. 10 Caldwell, J.H. and Daw, N.W., New properties of rabbit retinal ganglion cells, J. Physiol., 276 (1978) 257-276. 11 Canaday, M., Hillman, D.E. and Mahoney, G., Neuronal arbor recording and analysis. In A. Toga (Ed.), Three.Dimen. sional Neurolmaging, Raven, New York, 1990, pp. 87-100. 12 Cant, N.B., The fine structure of the lateral superior olivary nucleus of the cat, J. Comp. Neurol., 227 (1984) 63-77. 13 Denis-Donini, S., Giowinski, J. and Prochiantz, A., Glial heterogeneity may define the three-dimensional shape of mouse mesencephalic dopaminergic neurones, Nature, 307 (1984) 641643. 14 Finlayson, P.G. and Caspary, D.M., Phase sensitive binaural inhibition in lateral superior olivary neurons, Soc. Neurosci. Abstr., 16 (1990) 875. 15 Galambos, R., Schwartzkopf, J. and Rupert, A., Microelectrode study of the superior olivary nuclei, Am. J. Physiol., 197 (1959) 527-536. 16 Glaser, E.M., Van der Loos, H. and Gissler, M., Tangential orientation and spatial order in dendrites of cat auditory cortex: a computer microscope study of golgi-impregnated material, Exp. Brain Res., 36 (1979) 411-431.
In conjunction with our previous report of a relationship between LSO dendrite morphology and frequency selectivity in adults 43, the present correlation in young animals adds further support, and suggests that the modification of neuronal morphology can effect the appearance of mature neuronal response properties.
Acknowledgements. We thank Dr. Philip Smith for a critical reading of the manuscript, Dr. Dean E. Hillman and Mr. M. Canaday for help with the computer graphic reconstruction, Ms. V. Siverls for histological assistance, and Mr. Adam Start for photographic assistance. The physiological characterizations of LSO neurons were collected in Dr. E.W. Rubel's laboratory, then at the University of Virginia Medical Center. This work was supported by NIH DC00540-02 and the Mendik Research Fund. D.H.S. is a Sloan Foundation Fellow.
17 Greenwood, D.D. and Maruyama, N., Excitatory and inhibitory response areas of auditory neurons in the cochlear nucleus, J. Neurophysiol., 28 (1965) 863-892. 18 Hafidi, A., Sanes, D.H. and Cammer, W., Distribution of astrocytes along the frequency axis of the gerbil LSO, Soc. Neurosci. Abstr., 17 (1991). 19 Harnischfeger, G., Neuweiler, G. and Schlegel, P., Interaural time and intensity coding in superior olivary complex and inferior colliculus of the echo locating bat Molossus ater, J. Neurophysiol., 53 (1985) 89-109. 20 Harris, D. and Dallos, P., Ontogenetic changes in frequency mapping in a mammalian cochlea, Science, 225 (1984) 741-743. 21 Helfort, R.H. and Schwartz, I.R., Morphological evidence for the existence of multiple neuronal classes in the cat lateral superior olivary nucleus, J. Comp. Neurol., 244 (1986) 533-$49. 22 Joris, P.X. and Yia, T.C.T., Time sensitivity of cells in the lateral superior olive (LSO) to monaural and binaural amplitudemodulated complexes, Assoc. Res. Otolarnygol. Abstr., 13 (1990) 267. 23 Leuba, G. and Garey, LJ., Development of dendritic patterns in the lateral geniculate nucleus of monkey: a quantitative golgi study, Dev. Brain Res., 16 (1984) 285-299. 24 Levinthal, F., Macagno, E. and Levinthal, C., Anatomy and development of identified cells in isogonic organisms, Cold Spring Harbor Syrup. Quant. Biol., 40 (1975) 321-333. 25 Martin, M.R. and Dickson, J.W., Lateral inhibition in the anteroventral cochlear nucleus of the cat: a microiontophoretic study, Hearing Res., 9 (1983) 35-41. 26 McMullen, N.T., Goldberger, B. and Glaser, E.M., Postnatal development of lamina Ill/IV nonpyramidal neurons in rabbit auditory cortex: quantitative and spatial analyses of golgi-impregnated material, J. Comp. Neurol., 278 (1988) 139-155. 27 Meininger, V. and Baudrimont, M., Postnatal modifications of the dendritic tree of cells in the inferior ¢olliculus of the cat. A quantitative golgi analysis, J. Comp. Neurol., 200 (1981) 339355. 28 Morest, D.K., The differentiation of cerebral dendrites: a study of the post-migratory neuroblast in the medial nucleus of the trapezoid body, Z. Anat. Entwickl.-Gesch., 128 (1969) 271-289. 29 OUo, C. and Schwartz, I.R., The superior olivary complex in C57BL/6 mice, Am. J. Anat., 155 (1979) 349-374. 30 Pinto Lord, M.C. and Caviness, V.S., Determinant of cell shape and orientation: a comparative golgi analysis of cell-axon interrelationships in the developing neocortex of normal and reeler mice, J. Comp. Neurol., 187 (1979) 49-70. 31 Rakic, P., Extrinsic cytological determinants of basket and stellate cell dendritic pattern in the cerebellar molecular layer, J.
55
Comp. Neurol., 146 (1972) 335-354. 32 Rakic, E, Genetic and epigenetic determinants of local circuits in the mammalian central nervous system. In F.O. Schmidt and EG. Worden (Eds.), The Fourth Neurosciences Study Program, MIT Press, Cambridge, 1979, pp. 109-127. 33 Ramoa, A.S., Campbell, G. and Schatz, C.J., Dendritic growth and remodeling of cat retinal ganglion cells during fetal and postnatal development, J. Neurosci., 8 (1988) 4239-4261. 34 Ramon y Cajal, S., Histoiogie du Systeme de l'Homme et des Vertebres, Maloine, Paris, 1909, pp. 774-838. 35 Rihn, L.L. and Ciaiborne, B.J., Dendritic growth and regression in rat dentate granule cells during late postnatal development, Dev. Brain Res., 54 (1990) 115-124. 36 Rogowski, B.A. and Feng, A., Normal postnatal development of medial superior olivary neurons in the albino rat: a golgi and nissl study, J. Comp. Neurol., 196 (1981) 85-97. 37 Ryan, A.E, Woolf, N.K. and Sharp, ER., Tonotopic organization in the central auditor.v pathway of the mongolian gerbil: a 2-deoxyglucose study, J. Comp. Neurol., 207 (1982) 369-380. 38 Ryan, A.F. and Woolf, N.K., Development of tonotopic representation in the mongolian gerbil: a 2-deoxyglucose study, Dev Brain Res., 41 (1988) 61-70. 39 Sanes, D.H. and Rubel, E.W., The excitation in the gerbil lateral superior olive, J. Neurosci., 8 (1988) 682-700. 40 Sanes, D.H., Geary, W.A., Wooten, G.F. and Rubel, E.W., Quantitative distribution of the glycine receptor in the auditory brain stem of the gerbil, J. Neurosci., 7 (1987) 3793-3802. 41 Sanes, D.H. and Wooten, G.F., Development of glycine receptor distribution in the lateral superior olive of the gerbil, J. Neurosci., 7 (1987) 3803-3811. 42 Sanes, D.H., Merickel, M. and Rubel, E.W., Evidence for an alteration of the tonotopic map in the gerbil cochlea during development, J. Comp. Neurol., 279 (1989) 436-444. 43 Sanes, D.H., Goldstein, N.A., Ostad, M. and Hillman, D.E., Dendritic morphology of central auditory neurons correlates with their tonotopie position, J. Comp. Neurol., 294 (1990) 443454. 44 Sanes, D.H., An in vitro analysis of sound localization mechanisms in the gerbil lateral superior olive, J. Neurosci., 10 (1990) 3494-3506. 45 Sanes, D.H. and Siveris, V., The development and specificity of inhibitory axonal arborizations in the lateral superior olive, J. Neurobiol., 22 (1991) 837-854.
46 Scheibel, M.E. and Scheibel, A.B., Neuropii organization in the superior olive of the cat, Exp. Neurol., 43 (1974) 339-348. 47 Schmeidt, R.A., Spontaneous rates, thresholds and tuning of auditory-nerve fibers in the gerbil: comparisons to cat data, Hearing Res., 42 (1989) 23-36. 48 Schwartz, I.R. and Hockfieid, S., Localization of CAT-301 immunoreactivity on neurons in the brainstem auditory nuclei, Soc. Neurosci. Abstr., 15 (1989) 110. 49 Semple, M.N. and Kitzes, L.M., Single-unit ~esponses in the inferior colliculus: different consequences of contralateral and ipsilateral stimulation, J. Neurophysiol., 53 (1985) 1467-1482. 50 Smith, Z.D.J., Organization and development of brain stem auditory nuclei of the chicken: dendritic development in n. laminaris, J. Comp. Neurol., 203 (1981) 309-333. 51 Solomon, E, Detailed neurite morphologies of sister neuroblastoma cells are related, Cell, 16 (1979) 165-169. 52 Solomon, E, Neuroblastoma cells recapitulate their detailed neurite morphologies after reversible microtubule disassembly, Cell, 21 (1980) 333-338. 53 Tsuchitani, C. and Boudreau, J.C., Single unit analysis of cat superior olive S segment with tonal stimuli, J. Neurophysiol., 29 (1966) 684-697. 54 Valverde, E, Structural changes in the area striata of the mouse after enucleation, Exp. Brain Res., 5 (1968) 274-292. 55 Vaughn, J.E., Barber, R.E and Sims, T.J., Dendritic development and preferential growth into synaptogenic fields: a quantitative study of Golgi-impregnated spinal motor neurons, Synapse, 2 (1988) 69-78. 56 Voigt, H.E and Young, E.D., Evidence of inhibitory interactions between neurons in the dorsal cochlear nucleus, Z Neurophysiol., 44 (1980) 76-96. 57 Voyvodic, J.T., Development and regulation of dendrites in the rat superior cervical ganglion, J. Neurosci., 7 (1987) 904-912. 58 Weiss, G.M. and Pysh, J.J., Evidence for loss of purkinje cell dendrites during late development: a morphometric golgi analysis in the mouse, Brain Res., 154 (1978) 219-230. 59 Woolf, N.K. and Ryan, A.E, The development of auditory function in the cochlea of the mongolian gerbil, Hearing Res., 13 (1984) 277-283. 60 Woolf, N.K. and Ryan, A.F., Ontogeny of neural discharge patterns in the ventral cochlear nucleus of the mongolian gerbil, Dev. Brain Res., 17 (1985) 131-147.
47
BRESD 51440
Refinement of dendritic arbors along the tonotopic axis of the gerbil lateral superior olive Dan H. Sanes, John Song and James Tyson Dept~rtments of Otolaryngology, Physiology and Biophysics, New York University School of Medicine, New York, NY 10016 (USA)
(Accepted 31 December 1991) Key words: Auditory pathway; Development; Frequency tuning; Lateral superior olive; Morphometry; Dendrite
We have investigated the development of dendritic arbors in a central auditory nucleus in the Mongolian gerb.il, the lateral superior olive (LSO). The morphology of these arbors has been shown to vary with tonotopic position i:~ adults, with high frequency neurons having a more restricted field. In the prcscnt study, qualitative observations were made on horseradish peroxidase-filled neurons from animals 1-11 days postnatal, and quantitative results were obtained from Golgi-impregnated material from animals 10 days postnatal and older. The tonetopic position of each cell was computed as a percent of the total distance along the LSO. The dendritic arbors of high frequency neurons became spatially constrained along the frequency axis during the 3rd postnatal week, while those in the low frequency region retained a broader arborization into adulthood. This refinement was correlated with a decrease in total dendritic length and the number of branch points per neuron, particularly in the high frequency projection region. The distribution of octave bandwidths t,a which single LSO neurons responded in 13-16 day animals showed a similar course of maturation across the tonotopic axis: high frequency neurons responded to a larger number of octaves, and with greater variability, than those in adults. These data suggest that a specific alteration in dendrite morphology, which occurs after the onset of response to airborne sound, may contribute to adult frequenc) selectivity. INTRODUCTION The modification of dendritic processes during maturation has been a common observation in a wide range of neuronal systems. The proximal cause for this phenomenon is thought to be related to the finite availability of afferent innervation a2. However, it is not clear, from the systems thus far examined, whether these reductions in dendrite structure are functionally meaningful to the neuron. In the present study we focus on a discrete alteration of central auditory neuron dendrites that largely occurs following the onset of response to airborne sound. While these alterations reflect only a fraction of the intra- and intercellular forces acting during dendrogenesis, they offer an expedient assay system for experimental manipulations. The morphology of neurons within the lateral superior olive (LSO) have been well described in several species 12'21'29'34'43'46. The major neuronal class, the principal cells, compose approximately 80% of the neuronal population in adult cats 2~, and have modest dendritic trees that are oriented perpendicular to the tonotopic axis 42'43'46'53. The electrophysiological properties of LSO neurons also appear to be relatively uniform, being excited by ipsilateral sound stimuli and inhibited by contralateral sound stimuli 7's'9A5'19.39.
We have previously shown that the dendritic arb6dzations of principal neurons in the gerbil LSO are mote constrained along the tonotopic axis at successively higher frequency projection regions, and that this morphological variation correlates with the octave frequency range of single LSO neurons 4a. In addition, the results of a functional study indicate that frequency selectivity within the LSO improves during the period when airborne sound first evokes a neural response, and that this improved selectivity cannot be solely attributed to cochlear maturation 39. An analysis of the inhibitory receptor molecule, the glycine receptor 41, and single inhibitory afferents from the medial nucleus of the trapezoid body in the LSO 45 suggest that a pre- and postsynaptic modification could contribute to the functional maturation. In the present study, we present further evidence that the refinement of postsynaptic morphology in the LSO may play a role in functional maturation. MATERIALS AND METHODS Golgi impregnations Tissue was prepared according to the rapid Golgi technique of Adams (1979). Mongolian gerbils (Meriones unguiculatus) aged 10-21 days of either sex were given a lethal dose of anesthetic (sodium pentobarbital; 80 mg/kg; i.p.), and perfused transcardially with 0.9% NaCI, 10% buffered formalin, and potassium dichromate mordant. The brains were stored in mordant for 2-4 days, and
Correspondence: D.H. Sanes, Center for Neural Science, New York University, 4 Washington Place, New York, NY 10003, USA. Fax: (1) (212) 995-4011.
48 transferred to a silver nitrate solution for 2-4 days. Transverse vibratome sections at 150 pm were cleared in methyl salicylate, and mounted on gelatin-coated slides.
Horseradish peroxidase iontophoresis Brain slices of 400 pm from gerbils aged 2-11 days postnatal were prepared as described previously44. The tissue was placed in a chamber, and superfused with artificial cerebrospinal fluid (ACSF (in raM): NaC! 127.4, KCi 5, KH2PO 4 1.2, MgSO4.7H20 1.3, NaHCO 3 26, glucose 15, CaCI2.2H20 2.4, pH = 7.4). Electrodes filled with 6% HRP in 0.05 M "Iris and 0.5 M KCI (pH = 7.6) were beveled to a final resistance of 200-300 MQ, and positioned over the LSO. The parameters for HRP iontophoresis were 200 ms cathodal pulses at 0.8-1.5 nA delivered at 3 Hz for 1-3 rain. The slices were left to transport the HRP for 2.5 h after the last neuron was injected. The brain slices were placed in fixative (1% paraformaldehyde, 2.5% glutaraldehyde, 0.12 M phosphate buffered saline) for 2 h, rinsed in phosphate-buffered saline (PBS) at 4°C for 12-48 h, and washed in PBS with 0.5% Triton X-100 for 2 h. Each slice was then incubated in a filtered solution of 0.05% diaminobenzidine (DAB) with 0.02 % CoC!2 in 0.1 M Tris buffer2 for 30 min, and transferred to a fresh filtered DAB solution containing H202 for 5-15 min. Slices were placed onto gelatin-subbed slides, and dehydrated, cleared in xylenes, and mounted for viewing.
from 23 HRP-filled n e u r o n s f r o m 10 animals using the brain slice p r e p a r a t i o n . For the purposes of this study, the p a r a m e t r i c measures were p o o l e d for n e u r o n s with a position < 5 0 % from the apical m o s t projection r e g i o n or --_50% f r o m the apical most p r o j e c t i o n region. This corr e s p o n d s to the area of the L S O that responds to app r o x i m a t e l y 4 kHz 42. T h e s e 2 groups will h e r e a f t e r be r e f e r r e d to as the low f r e q u e n c y n e u r o n s a n d high freq u e n c y n e u r o n s , respectively. T h e data from Golgi-filled tissue was g r o u p by age at 10-12 days (n - 12 low a n d 32 high), 13-14 days (n = 24 low a n d 33 high), 15-16 days (n = 26 low and 22 high), 17-19 days (n = 27 low and 23 high), and 21-22 days (n = 29 low and 31 high), and the s a m p l e d n e u r o n s were e v e n l y distributed a l o n g
Image analysis and 3-D reconstruction Oolgi impregnated neurons in the LSO were selected for quantitative measurement as previously described43. Briefly, principal neurons with distinguishable processes, and no significant cut ends were observed with a light microscope (Zeiss Standard 16; Planapo 40 oil), and displayed on a color monitor (MTI-Dage Series 68 Neuvicon Video Camera). This selection procedure was meant to exclude multiplaner and marginal neurons from the analysis. Data acquisition and analysis was accomplished by using custom designed software for computer aided quantification and 3-D reconstruction (Cellmate/Treemate, R&M Biometrics) ~. The number of primary dendrites, the number of branch points, the total length of all dendritic segments, and the transverse soma area were all computed by the Treemate software. Measures of arbor spread along the tonotopic axis were also obtained from calibrated plots of the neurons, The tonotopic position of each neuron was computed as percent distance from the a.~ical most projection region (i,e, ventrolateral; see Sanes et al, 4", Fig, [B), and parametric measures were plotted against each neuron's position,
I
'%.
•4 . ,
Auditory neurophysiology The dehvery' of acoustic stimuli and recording. ~°f single neuron response properties have been described in detad ~ . Gerbils, aged 13-16 days, were studied under ketamine and chloral hydrate anesthesia such that nociceptive reflexes were eliminated. Sound stimuli were presently bilaterally through a calibrated, closed delivery system within a sound attenuated chamber. Single neurons were isolated with glass micropipettes, and responses to tonal stimuli of 50 ms duration with a 4 ms rise/fall time were led into a window discrimator for on-line data analysis. Response areas derived from monaural ipsilateral stimuli were generated for neurons that were determined to be ipsilaterally excited and contralaterally inhibited. Neuron location was determined histologically after iontophoresis of Fast green. The lowest and highest frequency to evoke a reliable increase in discharge rate at 30 dB above threshold was then determined, and this frequency range was converted to an octave band about the characteristic frequency. RESULTS T h e quantitative results o f this study are d r a w n from 259 n e u r o n s taken from 93 M o n g o l i a n gerbil brains sectioned in the transverse plane, and the qualitative results
Fig. tion Bar ron.
1. A neuron in the LSO of a 1 day animal. Top: a reconstrucof the neuron showing its position within the LSO boundary. - 100 pro. Bottom: a photomontage of the HRP-filled neuBar - 50 gm.
49 the transverse axis of the LSO. In addition, we compared values previously obtained from adult animals 43.
of growth through the first 2 postnatal weeks.
Development of arbor spread and frequency selectivity The morphology of a low and high frequency neuron
Initial postnatal arborization Figs. 1-4 show examples of HRP-fiUed LSO neurons at several postnatal ages. At 1-3 days, the HRP-filled neurons displayed a profusion of processes emanating from the soma, and a considerable number of branches (Figs. 1 and 2). Even at this early stage some neurons displayed a preferential orientation ~!ong the presumptive isofrequency laminae of the LSO (Fig. 1). At 10-11 days, LSO neurons displayed a high degree of branching, yet occupied a variable fraction of the LSO nucleus (Figs. 3 and 4). Although the numbers of HRP-filled neurons were not sufficient for quantitative analyses, their morphologies were consistent with a rapid period
/
I
\
\ .~
are illustrated along with reconst~'uctions rotated about the transverse axis for 14-15 day animals (Figs. 5 and
....... ~
o ,
i
"" ,..
:'
~,,
'/ f
i J
/
'~. i
/
.(
/
t/
)
\, /
/
~'%.. '~,, "L..
.j
%
0
\,
°
.
,///' .x ,.4
. . ~
~
....
t
i' ol /
J
I
•
Q ,\. m Fig. 2. A neuron in the LSO of a 3 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary. Bar = 100/am. Bottom: a photomontage of the HRP-filled neuron. Bar = 50/~m.
I[
Fig. 3. A neuron in the LSO of a 10 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary. Bar = 100/~m. Bottom: a photomontage of the HRP-fiiled neuron. Bar = 50 ~tm.
50 6), and 21 d~'~yanimals (Figs. 7 and 8). In Golgi-stained material, it was found that the distance that LSO dendritic arborizations spread across the tonotopic axis became more refined during development, but only for high frequency neurons (Fig. 9). This distance decreased significantly between 13-14 and 21-22 days from 71 + 3.7/~m (aS + S.E.M.) to 49 + 3.5 (t = 4.24; P < 0.0001; df = 62). There was no further change in animals aged _> 90 days. The spread of low frequency dendritic arbors did not vary significantly from the earliest age examined to the adult (Fig. 9). The octave bandwidths of neurons at 13-16 days were
calculated at 30 dB above threshold, and a comparison was made to the maturation of dendrite form along the frequency axis. As illustrated in Fig. 10, the octave bands of high frequency neurons were greater, and much more variable, in 13-16 day animals compared to adults. However, it must be noted that the octave bandwidths calculated for neurons in adult animals were computed at 50-60 dB above threshold. The octave bands of low frequency neurons had the same distribution in 13-16 day and adult animals. The change in distribution of octave bandwidths from 13-16 day to adult animals bears a striking resemblance to the change in distribution of dendritic arborizations across the frequency axis (Fig. 10).
Total dendritic length and number of branch points t' J .....~-~-.,
I
%
..
'l .,
i. t
\
/
t
i •,./'
The total dendritic length of high frequency LSO neurons first increased from 885 _ 60/~m at 10-12 days to 1082 + 70/tm at 13-14 days postnatal (t = -2.13; P < 0.05; df -- 60). As shown in Fig. 9, there was a significant reduction in total dendritic length to a value of 865 +_ 47/~m by 21-22 days (t -- 2.12; P < 0.05; df = 51), and a further reduction thereafter. There was no significant variation in the total dendritic length of low frequency LSO neurons across age. The mean number of branch points per neuron decreased significantly between 13-14 days and 90+ days postnatal, but was more apparent for high frequency neurons (Fig. 9). The values changed from 11.7 + 1 (~ ± S.E.M.) to 4.5 ± 0.3 branches for high frequency neurons (t - 8.04; P < 0.0001; df = 76), and from 7.6 ± 0.6 to 4.7 ± 0.4 branches for low frequency neurons (t - 4.06; P < 0.0005; df - 61). The number of branch points was consistently greater for high frequency neurons at all ages through 21-22 days, but was equivalent in adult animals. Among the parameters that did not vary significantly between 10-12-day- and adult animals were the number of primary dendrites (i.e. arising from the cell body), the cross-sectional area of the soma, and the distance that dendrites extended perpendicular to the tonotopic axis (data not shown). DISCUSSION
!! or
Fig. 4. A neuron in the LSO of an 11 day animal. Top: a reconstructior of the neuron showing its position within the LSO boundary. Bar -- 100 l~m. Bottom: a photomontage of the HRP-filled neuron. Bar -- 50 urn.
We have demonstrated a refinement of principal cell dendritic arborizations that appears largely restricted to one area of the LSO, the high frequency projection region. This morphological transformation is correlated with an improvement in the frequency selectivity of single LSO neurons over the same age range. The structural refinement of high frequency dendrites resulted from regressive events including the loss of branch points, and a decrease in total dendritic length.
51
/
\ I !
",,
] ;
•,-, i"
/
l' "-t.
a, -¸~. l~ .it,...
Fig. 5. A neuron in the LSO of a 14 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar = 100/~m. Bottom: a photomontage of the Golgi-stained neuron. Bar = 50/~m.
Similar modifications of dendrite form have been described in a wide variety of systems. These regressive events include a decrease in the number of spines 2¢'33, a decrease in the total process length 23'5s, a decrease in the number of primary dendrites 5°, and a decrease in the number of branch points 26'36. The physical dimensions occupied by a dendritic field may also change during development 35. It has been shown that dendrogenesis is partially related to afferent density or synaptic activity5'2s'3°'31'54'55, the availability of postsynaptic targets 57, the interaction with a specific substrate 3't3, and intrinsic cues 4'24'51'52. The concomitant change in dendrite morphology and frequency selectivity during development could result from an activity-mediated process. However, since the observed changes in dendrite form were largely restricted to the high frequency projection region of the LSO (Fig. 10), one would have to postulate a different level or pattern of synaptic activity in the low and high frequency
Fig. 6. A neuron in the LSO of a 15 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar 100/~m. Bottom: a photomontage of the Golgi-stained neuron. Bar = 50/~m.
regions. There is a small amount of evidence for a difference in spontaneous and evoked activity along the frequency axis. Measures of auditory nerve spontaneous activity in the adult gerbil demonstrate that a larger percentage of fibers with characteristic frequencies above 5 kHz have very low discharge rates 47. There is no reported difference in sound evoked activity or metabolism along the tonotopic axes of the gerbil central auditory nuclei 37'39' 49.~o. However, sound evoked 2-deoxyglucose uptake is conspicuously absent from the low frequency projection region at the onset of response to airborne sound 38. Whether or not these physiological factors play a role in the pattern of dendrogenesis seen in the gerbil LSO awaits experimental analysis. There is some evidence that the local environment differs between the high and low frequency projection
52
".'.\ •.
' ~ .
~
7
~
I[
,..
',.
,
"".
,i
[
~ -° . . . . . . . . -
.,
I
Fig. 8. A high frequency neuron in the LSO of a 21 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 300 intervals about the sagittal plane. Bar = 100 ~m. Bottom: a photomontage of the Golgistained neuron showing a highly restricted arborization along the tonotopic axis. Bar = 50 # m .
I
Ill Ill
Fig. 7. A low frequency neuron in the LSO of a 21 day animal. Top: a reconstruction of the neuron showing its position within the LSO boundary, and rotated at 30° intervals about the sagittal plane. Bar = 100 ~m. Bottom: a photomontage of the Oolgistained neuron illustrating the relatively large spread along the tonotopic axis compared to the ~euron shown in Fig. 8. Bar = 50
regions. The packing density of neuronal somas is nearly double in the high frequency region, and the density of glycine receptors is approximately fourfold greater 'm. In contrast, there is a greater density of astrocytes in the low frequency region Is, and a greater number of CAT301 positive neurons as well4~, Therefore, it is possible that dendritic branching patterns reflect 2 distinct extracellular environments that influence process outgrowth and stability.
Does dendrite form influence frequency tuning? The maturation of stimulus coding within the gerbil auditory system has been described for both the cochlea and cochlear nucleus s9'6°. These authors found an electric response to sound, as evidenced by a cochlear microphonic or N~ compound action potential, as early as postnatal day 12. In addition, Woolf and Ryan ~° demonstrated that high frequency neurons in the cochlear nucleus have adult-like Qlo values from the onset of hearing, whereas the tuning of low frequency neurons continued to improve. The contrast with the present results may be explained in 2 ways. First, measures of tuning taken near the neurons absolute threshold (e.g. Qto) may not be as sensitive to the convergence of afferents as would be measures taken at 30-60 dB above threshold (e.g. octave bandwidths used in the present study). Second, in young animals the appearance of
53 Dendrite Width (A) o
I ...... L...... !
.k.,
I
-,:,=-
60"
40
i
~+
'|
/ ~ ~o* *° * •
Adult
:t. !
oo
13-16
8O
E
tt
15or
100
1001 -~
/:
,
~
o
~o
/
o o.
~p o
"~,,,,o
~6" o .°.,~i#~
10-12 13"14 15"~16 1¥-19 21"22 90.
Age (days postnatal) 0
Branches
/
I
I
I
100 % Distance from Apical End
I
12~
I
0
25
50
75
10; -----
,ID E
-
-
-
High 7
Low o
13-16 Adult
6
s
xF= 13:14 ,m',le it'-Ill e~e= Q~I. '"-'
o
o go OOOOO o ~ ~ % R~b°~7 ~°° o
o
Age (days postnatal)
3 Total Dendritic Length
I
1100"
o
I
liO0'
E
....
.o°'°'"
......
o,
O ,
::L 700'
600
~ ~.0o O~
o~ ~
~o ~ o ~
i
i
1
i . .....
2
~
10;1218'o14lli'-lil l#'-lli li-'illl lib. Age (days postnatal)
Fig. 9. Parametric changes in dendrit'c arbors from the high frequency projection region (solid line) and low frequency projection region (dashed line) of the LSO. Top: the dendritic width (A) along the frequency axis remains unchanged for low frequency neurons, but decreases significantly between 13 and 19 days postnatal. Middle: the number of branch points per neuron decreases for both high and low frequency neurons between 15 and 904- days. Bottom: the total dendritic length remains unchanged for low frequency neurons, but decreases significantly between 15 and 904days postnatal for high frequency neurons. Adult values taken from Sanes et al. 'u. adult-like functional properties are commoilly found in conjunction with immature responses, and the process of maturation consists of probabilistic refinements within the whole population. In this regard, it is interesting to note the distribution of octave bands in the high frequency region of 14-16 day animals. Neurons appeared to exhibit two extremes in tuning. On the other hand, the distribution of dendrite widths appeared as an uniform shift in the distribution towards larger values. It should be noted that the afferent arborizations are becoming refined during the same time period 4s. Therefore, it is possible that the physiological result is magnified by the immaturity of
0
~
oo ~o ~ i
i
100 % Distance from Apical End 25
50
75
Fig. 10. A comparison of LSO dendritic width along the frequency axis (Top) and frequency selectivity (bottom) at 13-16 days postnatal and in the adult. The frequency axis has been converted to % distance from the apical most projection region to correct for changes in the cochlear tonotopic map during development2°'42. Top: the dendritic width is virtually identical from 0 to 50% of the distance from the apical most projection region. However, above 50% the dendritic width is generally greater for 13-16 day animals. Bottom: the octave bands largely overlap from 0 to 50% of the distance from the apical most projection region, but are much greater for 13-16 day animals above 50%. Adult values taken from Sanes et al.43.
other elements in the circuit, including the cochlea. Although the frequency response of central auditory neurons is primarily dependent upon cochlear processing, there is evidence to indicate that the convergence of afferents is also a factor. For example, neurons in the cat cochlear nuclei exhibit both excitatory and inhibitory frequency response areas from separate afferent populations 17'25's6. If the dendritic field dimensions regulate the number of afferents that will form synaptic contacts, then they may also contribute to the frequency selectivity. There is a precedent for dendritic morphology influencing a specific visual coding property. In the rabbit retina, orientation-sensitive ganglion and amacrine cells can have markedly asymmetric dendritic arborizations that correlate with the preferred angle of a bar stimu-
54 iUS6.t°. In addition, there have been 2 reports of central auditory neurons showing a preferential orientation along the frequency axis, but there was no direct correlation with frequency tuning 16'27. It should be stressed that the emergence of 2 different dendritic arborization patterns may be related less to frequency tuning than to the binaural processing strategies of the LSO. For example, recent studies have demonstrated a strong temporal dependency of LSO neurons, particularly for low frequency cells 14'22. The broader spread of both dendrites and inhibitory arborizations 45 across the frequency axis in the low frequency region may subserve the phase-sensitivity of these cells. Low frequency LSO neurons may sample from a maximum number of phas,.Mocking afferents irrespective of their characteristic frequency.
REFERENCES 1 Adams, J.C., A fast, reliable silver-chromate Golgi method for pedusion-fixed tissue. Stain Technoi., 54 (1979) 225-226. 2 Adams, J.C., Heavy metal intensification of DAB-based reaction product, J. Histochem. Cytochem., 29 (1981) 775. 3 Adler, R. and Varon, S., Neuritic guidance by nonneuronal cells of ganglionic origin, Dev. Biol., 86 (1981) 69-80. 4 Banker, G.A. and Cowan, W.M., Further observations on hippocampal neurons in dispersed cell culture, J. Comp Neurol., 187 (1979) 469-494. 5 Berry, M., McConnell, E and Sievers, J., Dendritic growth and the control of neuronal form, Current Topics Dev. Biol., 15 (1980) 67-102. 6 Bloomfield, S.A., Two types of orientation-sensitive responses of amacrine cells in the mammalian retina, Nature, 350 (1991) 347-350. 7 Boudreau, J.C. and Tsuchitani, C., Binaural interaction in the cat superior olive s-segment, J. Neurophysiol., 3t (1968) 442-451. 8 Boudreau, J.C. and Tsuchitani, C., Cat superior olive s-segment cell discharge to tonal stimulation, in W.D. Neff (Ed.), Contri. butions to Sensory Physiology, Vol. 4, Academic Press, New York, 1970, pp, 143-213. 9 Caird, D. and Klinke, A.R., Processing of binaural stimuli by cat superior olivary complex neurons, Exp. Brain Res., 52 (1983) 358-399. 10 Caldwell, J.H. and Daw, N.W., New properties of rabbit retinal ganglion cells, J. Physiol., 276 (1978) 257-276. 11 Canaday, M., Hillman, D.E. and Mahoney, G., Neuronal arbor recording and analysis. In A. Toga (Ed.), Three.Dimen. sional Neurolmaging, Raven, New York, 1990, pp. 87-100. 12 Cant, N.B., The fine structure of the lateral superior olivary nucleus of the cat, J. Comp. Neurol., 227 (1984) 63-77. 13 Denis-Donini, S., Giowinski, J. and Prochiantz, A., Glial heterogeneity may define the three-dimensional shape of mouse mesencephalic dopaminergic neurones, Nature, 307 (1984) 641643. 14 Finlayson, P.G. and Caspary, D.M., Phase sensitive binaural inhibition in lateral superior olivary neurons, Soc. Neurosci. Abstr., 16 (1990) 875. 15 Galambos, R., Schwartzkopf, J. and Rupert, A., Microelectrode study of the superior olivary nuclei, Am. J. Physiol., 197 (1959) 527-536. 16 Glaser, E.M., Van der Loos, H. and Gissler, M., Tangential orientation and spatial order in dendrites of cat auditory cortex: a computer microscope study of golgi-impregnated material, Exp. Brain Res., 36 (1979) 411-431.
In conjunction with our previous report of a relationship between LSO dendrite morphology and frequency selectivity in adults 43, the present correlation in young animals adds further support, and suggests that the modification of neuronal morphology can effect the appearance of mature neuronal response properties.
Acknowledgements. We thank Dr. Philip Smith for a critical reading of the manuscript, Dr. Dean E. Hillman and Mr. M. Canaday for help with the computer graphic reconstruction, Ms. V. Siverls for histological assistance, and Mr. Adam Start for photographic assistance. The physiological characterizations of LSO neurons were collected in Dr. E.W. Rubel's laboratory, then at the University of Virginia Medical Center. This work was supported by NIH DC00540-02 and the Mendik Research Fund. D.H.S. is a Sloan Foundation Fellow.
17 Greenwood, D.D. and Maruyama, N., Excitatory and inhibitory response areas of auditory neurons in the cochlear nucleus, J. Neurophysiol., 28 (1965) 863-892. 18 Hafidi, A., Sanes, D.H. and Cammer, W., Distribution of astrocytes along the frequency axis of the gerbil LSO, Soc. Neurosci. Abstr., 17 (1991). 19 Harnischfeger, G., Neuweiler, G. and Schlegel, P., Interaural time and intensity coding in superior olivary complex and inferior colliculus of the echo locating bat Molossus ater, J. Neurophysiol., 53 (1985) 89-109. 20 Harris, D. and Dallos, P., Ontogenetic changes in frequency mapping in a mammalian cochlea, Science, 225 (1984) 741-743. 21 Helfort, R.H. and Schwartz, I.R., Morphological evidence for the existence of multiple neuronal classes in the cat lateral superior olivary nucleus, J. Comp. Neurol., 244 (1986) 533-$49. 22 Joris, P.X. and Yia, T.C.T., Time sensitivity of cells in the lateral superior olive (LSO) to monaural and binaural amplitudemodulated complexes, Assoc. Res. Otolarnygol. Abstr., 13 (1990) 267. 23 Leuba, G. and Garey, LJ., Development of dendritic patterns in the lateral geniculate nucleus of monkey: a quantitative golgi study, Dev. Brain Res., 16 (1984) 285-299. 24 Levinthal, F., Macagno, E. and Levinthal, C., Anatomy and development of identified cells in isogonic organisms, Cold Spring Harbor Syrup. Quant. Biol., 40 (1975) 321-333. 25 Martin, M.R. and Dickson, J.W., Lateral inhibition in the anteroventral cochlear nucleus of the cat: a microiontophoretic study, Hearing Res., 9 (1983) 35-41. 26 McMullen, N.T., Goldberger, B. and Glaser, E.M., Postnatal development of lamina Ill/IV nonpyramidal neurons in rabbit auditory cortex: quantitative and spatial analyses of golgi-impregnated material, J. Comp. Neurol., 278 (1988) 139-155. 27 Meininger, V. and Baudrimont, M., Postnatal modifications of the dendritic tree of cells in the inferior ¢olliculus of the cat. A quantitative golgi analysis, J. Comp. Neurol., 200 (1981) 339355. 28 Morest, D.K., The differentiation of cerebral dendrites: a study of the post-migratory neuroblast in the medial nucleus of the trapezoid body, Z. Anat. Entwickl.-Gesch., 128 (1969) 271-289. 29 OUo, C. and Schwartz, I.R., The superior olivary complex in C57BL/6 mice, Am. J. Anat., 155 (1979) 349-374. 30 Pinto Lord, M.C. and Caviness, V.S., Determinant of cell shape and orientation: a comparative golgi analysis of cell-axon interrelationships in the developing neocortex of normal and reeler mice, J. Comp. Neurol., 187 (1979) 49-70. 31 Rakic, P., Extrinsic cytological determinants of basket and stellate cell dendritic pattern in the cerebellar molecular layer, J.
55
Comp. Neurol., 146 (1972) 335-354. 32 Rakic, E, Genetic and epigenetic determinants of local circuits in the mammalian central nervous system. In F.O. Schmidt and EG. Worden (Eds.), The Fourth Neurosciences Study Program, MIT Press, Cambridge, 1979, pp. 109-127. 33 Ramoa, A.S., Campbell, G. and Schatz, C.J., Dendritic growth and remodeling of cat retinal ganglion cells during fetal and postnatal development, J. Neurosci., 8 (1988) 4239-4261. 34 Ramon y Cajal, S., Histoiogie du Systeme de l'Homme et des Vertebres, Maloine, Paris, 1909, pp. 774-838. 35 Rihn, L.L. and Ciaiborne, B.J., Dendritic growth and regression in rat dentate granule cells during late postnatal development, Dev. Brain Res., 54 (1990) 115-124. 36 Rogowski, B.A. and Feng, A., Normal postnatal development of medial superior olivary neurons in the albino rat: a golgi and nissl study, J. Comp. Neurol., 196 (1981) 85-97. 37 Ryan, A.E, Woolf, N.K. and Sharp, ER., Tonotopic organization in the central auditor.v pathway of the mongolian gerbil: a 2-deoxyglucose study, J. Comp. Neurol., 207 (1982) 369-380. 38 Ryan, A.F. and Woolf, N.K., Development of tonotopic representation in the mongolian gerbil: a 2-deoxyglucose study, Dev Brain Res., 41 (1988) 61-70. 39 Sanes, D.H. and Rubel, E.W., The excitation in the gerbil lateral superior olive, J. Neurosci., 8 (1988) 682-700. 40 Sanes, D.H., Geary, W.A., Wooten, G.F. and Rubel, E.W., Quantitative distribution of the glycine receptor in the auditory brain stem of the gerbil, J. Neurosci., 7 (1987) 3793-3802. 41 Sanes, D.H. and Wooten, G.F., Development of glycine receptor distribution in the lateral superior olive of the gerbil, J. Neurosci., 7 (1987) 3803-3811. 42 Sanes, D.H., Merickel, M. and Rubel, E.W., Evidence for an alteration of the tonotopic map in the gerbil cochlea during development, J. Comp. Neurol., 279 (1989) 436-444. 43 Sanes, D.H., Goldstein, N.A., Ostad, M. and Hillman, D.E., Dendritic morphology of central auditory neurons correlates with their tonotopie position, J. Comp. Neurol., 294 (1990) 443454. 44 Sanes, D.H., An in vitro analysis of sound localization mechanisms in the gerbil lateral superior olive, J. Neurosci., 10 (1990) 3494-3506. 45 Sanes, D.H. and Siveris, V., The development and specificity of inhibitory axonal arborizations in the lateral superior olive, J. Neurobiol., 22 (1991) 837-854.
46 Scheibel, M.E. and Scheibel, A.B., Neuropii organization in the superior olive of the cat, Exp. Neurol., 43 (1974) 339-348. 47 Schmeidt, R.A., Spontaneous rates, thresholds and tuning of auditory-nerve fibers in the gerbil: comparisons to cat data, Hearing Res., 42 (1989) 23-36. 48 Schwartz, I.R. and Hockfieid, S., Localization of CAT-301 immunoreactivity on neurons in the brainstem auditory nuclei, Soc. Neurosci. Abstr., 15 (1989) 110. 49 Semple, M.N. and Kitzes, L.M., Single-unit ~esponses in the inferior colliculus: different consequences of contralateral and ipsilateral stimulation, J. Neurophysiol., 53 (1985) 1467-1482. 50 Smith, Z.D.J., Organization and development of brain stem auditory nuclei of the chicken: dendritic development in n. laminaris, J. Comp. Neurol., 203 (1981) 309-333. 51 Solomon, E, Detailed neurite morphologies of sister neuroblastoma cells are related, Cell, 16 (1979) 165-169. 52 Solomon, E, Neuroblastoma cells recapitulate their detailed neurite morphologies after reversible microtubule disassembly, Cell, 21 (1980) 333-338. 53 Tsuchitani, C. and Boudreau, J.C., Single unit analysis of cat superior olive S segment with tonal stimuli, J. Neurophysiol., 29 (1966) 684-697. 54 Valverde, E, Structural changes in the area striata of the mouse after enucleation, Exp. Brain Res., 5 (1968) 274-292. 55 Vaughn, J.E., Barber, R.E and Sims, T.J., Dendritic development and preferential growth into synaptogenic fields: a quantitative study of Golgi-impregnated spinal motor neurons, Synapse, 2 (1988) 69-78. 56 Voigt, H.E and Young, E.D., Evidence of inhibitory interactions between neurons in the dorsal cochlear nucleus, Z Neurophysiol., 44 (1980) 76-96. 57 Voyvodic, J.T., Development and regulation of dendrites in the rat superior cervical ganglion, J. Neurosci., 7 (1987) 904-912. 58 Weiss, G.M. and Pysh, J.J., Evidence for loss of purkinje cell dendrites during late development: a morphometric golgi analysis in the mouse, Brain Res., 154 (1978) 219-230. 59 Woolf, N.K. and Ryan, A.E, The development of auditory function in the cochlea of the mongolian gerbil, Hearing Res., 13 (1984) 277-283. 60 Woolf, N.K. and Ryan, A.F., Ontogeny of neural discharge patterns in the ventral cochlear nucleus of the mongolian gerbil, Dev. Brain Res., 17 (1985) 131-147.