- Email: [email protected]
Localization of optic tectal input to the ventral dendrite of the goldfish Mauthner cell
113
Brain Re'search, 401 (1987) 113-121 Elsevier BRE 12270
Localization of optic tectal input to the ventral dendrite of the goldfish Mauthner cell Steven J. Zottoli 1, A n d r e w R. H o r d e s 2 and D o n a l d S. F a b e r 2 1Department of Biology, Williams College, Williamstown, MA 01267 (U.S.A.) and 2Division of Neurobiology, Department of Physiology, State University of New York. Buffalo. NY14214 (U.S.A.) (Accepted 3 June 1986)
Key words: Mauthner cell, Segregated synaptic input; Visual input; Startle response; Mauthner cell ventral dendrite
Although visually evoked Mauthner cell (M-cell) startle responses occur in the goldfish, the afferent projections underlying these reactions have not been previously studied. We have recorded from the M-cell while stimulating the left optic nerve and/or right optic tectum and have traced projections of the optic nerve and restricted areas of the optic rectum using HRP histoehemistry and autoradiography. Tectal stimulation elicits similar postsynaptic potentials (PSPs) in both M-cells. The responses recorded in the right (ipsilateral) cell were localized to its ventral dendrite. The existence of uncrossed tectal projections to the ventral dendrite was confirmed morphologically following application of horseradish peroxidase (HRP) to the optic tectum. The PSPs contained both inhibitory and excitatory components, but with adequate stimulus strength, excitation of either M-cell dominated. Thus, this pathway is probably sufficient to trigger visually evoked startle responses mediated by the M-cell. Stimulation of the left optic nerve also evoked PSPs capable of bringing both M-cells to threshold. The blockage of this response by conditioning stimulation of the right tectum suggests that the visual information is relayed to the M-cells through this structure. In support of these findings, no label was found near any portion of the M-cell after either intraocular injection of tritiated proline or application of HRP to the cut end of the optic nerve. In summary, visual input to the M-cell is mediated via projections from the tectum, is segregated onto the ventral dendrite, and is capable of bringing this neuron to threshold. This pathway presumably accounts for the demonstrated behavioral efficacy of visual stimuli in evoking a startle response. INTRODUCTION The M a u t h n e r cell (M-cell) of goldfish mediates startle responses evoked by sound9'3°and by visual stimuli924. The afferent pathway subserving the auditory response is quite direct, primarily involving the large saccular afferents which end as myelinated clubs on the distal lateral dendrite 1'17'21'23. In contrast, the afferent projections underlying the visually evoked response remain u n k n o w n , although Bartelmez 1 traced fibers from the optic tectum to the inferior and superior ventral dendrites of the M-cell in Amieurus. To better u n d e r s t a n d the input circuitry involved in M-cell-initiated startle behaviors resulting from visual stimuli, we recorded from this n e u r o n while stimulating the optic nerve and/or the optic tecturn and traced projections of the optic nerve and restricted areas of the optic tectum using H R P histochemistry and autoradiography. We found that post-
synaptic potentials (PSPs) due to tectal stimulation are localized onto the M a u t h n e r cell ventral dendrite and that this input is capable of bringing the cell to threshold. The existence of tectal projections to the ventral dendrite was confirmed morphologically, thereby providing another example of segregated synaptic input onto the M-cell. In addition, the results indicate that this dendrite is accessible for electrophysiological investigations and is electrically inexcitable, like the well-studied lateral process. MATERIALS AND METHODS C o m m o n goldfish, 12-18 cm in total length were used for electrophysiological and histological experiments. The fish were exposed to a 12-12 h light/dark cycle and kept between 15 and 16 °C. Electrophysiological experiments were conducted using standard surgical and recording techniques.
Correspondence: S. Zottoli, Department of Biology, Williams College, Williamstown, MA 01267, U.S.A.
114 Fish were anesthetized with MS-222 (tricaine methanesulfonate) and intracellular recordings were obtained from either M-cell with glass microelectrodes filled with 2.7 M KC1 (R 1 and R2 of Fig. 1). Bipolar stimulating electrodes were placed on the left optic nerve (S1) and right optic tectum ($2; the majority of retinal afferents to the tectum are known to terminate in the contralateral tectum 22) for activation of orthodromic inputs to the Mauthner cell and on the vertebrae overlying the spinal cord near the caudal edge of the dorsal fin ($3) for antidromic activation (Fig. 1). The optic nerve was exposed by cutting the conjuctiva around the rim of the orbit, severing the extraocular muscles and rotating the eyeball. The caudolateral edge of the optic tectum was chosen as the area to stimulate because of its accessibility and the fact that stimulation resulted in PSPs capable of firing the Mauthner cell. The M-cell was first located by its large antidromically activated extracellular spike recorded in the axon cap 12A3, and subsequent intracellular penetrations were made from the M-cell soma, from the lateral dendrite as far as 300/,tm caudolateral to the axon cap, and from the ventral den-
$2
7---~ Sa Fig. 1. Schematic diagram of the stimulating and recording arrangement. The left optic nerve (S 0 and the right optic rectum ($2) were stimulated for orthodromic activation of the Mauthner cells (shown caudal to the tecta). The cells were antidromically activated with an electrode ($3) placed on the vertebrae over the spinal cord. Intracellular recordings were obtained from the left (R 0 and right (R2) Mauthner cells.
drite as far as 100/~m rostral and ventral to the axon cap. In previous experiments with the M-cell 31, fish were routinely positioned within the holding chamber so that the surface of the medulla oblongata was approximately horizontal. The recording electrode was in a sagittal plane with an orientation of about 70 ° to the horizontal and was inserted into the brain in a rostroventral direction. Since the ventral dendrite courses in the same general direction, such an approach would have reduced the accessibility of this neurite and increased the likelihood of damage. Certainly, it would have been difficult to determine the exact position of the electrode tip relative to the soma and the length of the dendrite. Therefore, in the present experimental series, fish were positioned in the holding chamber so that the head tilted upwards (about 20°). With this orientation the initial portion of the ventral dendrite was approximately perpendicular to the recording electrode and distal segments were more accessible.
Morphological studies Autoradiography. Unilateral injections of tritiated proline (20 ~Ci; New England Nuclear) were made into a single eye of each of 6 goldfish under anesthesia (MS-222). The fish were held at 15-16 °C and sacrificed, 14 days postoperatively, by perfusion with 0.7% saline followed with fixative (90 m180% E T O H , 5 ml formalin, 5 ml glacial acetic acid). The brains were removed, dehydrated (ethanol, t-butyl mixture), embedded in paraffin, transversely sectioned at 15 ~m and mounted on gelatin-coated slides. The slides were dipped in the dark into Kodak NT2B emulsion and left in the refrigerator for 31 days. They were then developed in Dektol and stained with Cresyl violet acetate 4. Horseradish peroxidase. A bead of recrystallized horseradish peroxidase (HRP; about 250 ktg) was applied 3~ to either the completely transected optic nerve (one fish) or to a superficial cut of the optic tecturn (5 fish) at sites equivalent to those used for stimulation in the electrophysiological studies. After HRP application to the optic nerve the space behind the eyeball was filled with a Vasoline-paraffin oil mixture to reduce spread of the enzyme and the eye was placed back in its socket. In the experiments utilizing H R P application to the tectum the brain cavity
115 was filled with a Vaseline-paraffin oil mixture covered by thin plastic and c a p p e d with an epoxy plug anchored to the skull 3°'32. In one instance the M a u t h n e r cell soma was iontophoretically filled with H R P (4% H R P , Sigma VI, in 0.05 M Tris with 0.2 M KC1, p H 8.62; initial resistance of 2 0 - 5 0 Mr2) and H R P was also applied to the tectum. A f t e r intervals of 5 - 7 days (15.6 °C holding t e m p e r a t u r e ) , the fish were perfused with 20 ml of 2.5% glutaraldehyde in phosphate buffer (pH 7.2), and their brains were removed and placed successively in fresh fixative for 3 h, in 20%. sucrose for 5 h and in a gelatin albumin mixture overnight. The brains were subsequently sectioned at either 40 or 5 0 - 6 0 # m and stained (freefloating at 40 t~m and after drying to microscope slides at between 50 and 60 Bm) with either the chromogen benzidine dihydrochloride ls.25 or H a n k e r Yates reagent >. Brains stained with H a n k e r Yates reagent were counterstained with Neutral red or Cresyl violet acetate. RESULTS M-cell response to optic tectal stimulation
Stimulation of the caudolateral edge of the right rectum resulted in postsynaptic potentials (PSPs) recorded in both M-cell s o m a t a (50 # m lateral to the
A1 AD ~Left M-cell
A2
lOV
B1
_
axon cap). This bilateral response is shown in Fig. 2, which is from an e x p e r i m e n t where recordings were first obtained from the left M-cell ( A l, A2) and then from the right one (B 1, B2). In both, the antidromic spike amplitudes were at least 30 mV ( A 1, B1). Stimulation of the right optic tectum, at low stimulus strengths, e v o k e d c o m p a r a b l e PSPs in the two cells, while when the strength was d o u b l e d , the response in the right M-cell was greater. In 8 experiments the latency of the response in the ipsilateral M-cell was 0.63 _+ 0.09 ms (mean _+ S.D.; n = 12), with a m e a n peak latency of 0.96 ms, while the PSP r e c o r d e d in the cell contralateral to the stimulus site had a latency of 0.60 _+ 0.08 ms (n = 5) and a mean p e a k latency of 1.03 ms. That is, there was no difference in the timing of the ipsi- and contralateral responses to stimulation of the right optic rectum. A s shown in Fig. 2, when the stimulation voltage was increased a second p e a k with a latency of about 1.9 ms was discernible in both cells. In general, the bilateral PSPs were similar in nature, and the following description focuses on the ipsilateral M-cell responses to right optic tectum stimulation. Typically, if the tectal stimulus strength was increased sufficiently, both M-cells could be excited b e y o n d threshold. This o r t h o d r o m i c excitation is shown for the ipsilateral M-cell in Fig. 3, where it can be c o m p a r e d with antidromic activation of the cell.
Right M-cell
.L__: B2
A1
Extracell.
B1
Intracell.,
,=@~
AD
J m
w
20V
'~11~ .,. ~
1msec
Fig. 2. Bilateral Mauthner cell inputs from the right optic tecturn. A t, A 2 and BI, B2: recordings from, respectively, the left and right M-cells of responses to stimulation of the surface of the right optic tectum at its caudolateral border. The electrodes were located in the soma, as indicated by the amplitudes of the antidromic spikes (Al, B1). In A 2 and B 2, the optic tectum was stimulated at 10 (upper traces) and 20 V (lower traces) and compound PSPs were evoked. Note that at 10 V one peak dominates while at 20 V a second later peak is comparable in amplitude to the first.
J
2-~-~-c
lmsec
Fig. 3. Extra- and intracellular recordings near the M-cell soma. AI: antidromic activation of the right M-cell resulted in an extracellular spike recorded within the axon cap and followed by an EHP (arrow). At: right tectal stimulation also evoked a positivity within the axon cap (arrow). In these superimposed traces, the M-cell fired in one case. BI, B2: intracellular recordings obtained in the same experiment from the soma 50 l~m lateral to the axon cap. BI: the antidromic spike is approximately 40 mV in amplitude. B2: tectal stimulation resulted in an EPSP which in this example was capable of bringing the cell to threshold.
116 Both extracellularly and intracellularly r e c o r d e d spikes are illustrated, and in B 2 a clear EPSP precedes the orthodromic action potential. Even though the PSP e v o k e d in the ipsilateral M-cell was often sufficiently excitatory to evoke an o r t h o d r o m i c action potential it contained an inhibitory c o m p o n e n t , as shown by two methods. First, antidromic stimulation activates a pool of interneurons (PHP neurons 12) inhibitory to the M-cell. Their impulses produce an extracellular positivity ( E H P ) in the axon cap, which is the first phase of a t w o - c o m p o nent (electrical and chemical) inhibition. A similar potential was r e c o r d e d in the axon cap after tectal stimulation (Figs. 3 A 2 and 4A1). In fact, as shown in
A1
OpT
A2
AD-FOpT
,I
B1
.I
opt
lmsec
Fig. 4. Inhibitory component of the tectal PSP. A1, A2: extracellular recordings from the right M-cell's axon cap. AI: stimulation of the right tectum (OPT) produced an extracellular positivity. A2: when a conditioning antidromic stimulus, which evoked a negative M-cell spike and a subsequent extrinsic hyperpolarizing potential (EHP) was followed by the tectal stimulation, the amplitude of the tectal response was reduced (the bar above the test response in A 2 indicates the control amplitude in AI). This reduction suggests that some of the same inhibitory interneurons which produce the EHP after antidromic stimulation also contributed to the positivity recorded after tectal stimulation (superimposed traces with and without tectal stimulation). B 1, B2: intracellular recordings in the same experiment, of M-cell responses evoked by the stimulus paradigms used for A 1 and A 2. BI: composite PSP recorded in the M-cell soma after right optic tectum stimulation. B2: a conditioning tectal PSP produces a conductance increase, as indicated by the decrease in amplitude of test antidromic action potentials superimposed at different intervals. The first spike is equivalent in amplitude to a control and thus acts as a reference, Note that the conductance change was maximal approximately 4 ms after the first peak of the tectal PSP. Two or more superimposed sweeps in all records.
Fig. 4A2, when a conditioning antidromically e v o k e d E H P was followed by a test tectally e v o k e d positivity, the latter was reduced in amplitude. This finding suggests that tectal neurons may activate some of the same P H P neurons fired by antidromic stimulation of the M-cell's recurrent inhibitory n e t w o r k (alternatively, the interneurons involved m a y have been inhibited by the spinal stimulation). Second, a m o r e direct means of unmasking the inhibitory c o m p o n e n t of tectal PSPs involved superimposing antidromic action potentials on a conditioning PSP (Fig. 4B1) e v o k e d by optic tectum stimulation. A s is a p p a r e n t in Fig. 4B2, this protocol decreases the antidromic spike height, with the maximal reduction occurring about 4 ms following the first p e a k of the tectal PSP. The explanation for this result is that the antidromic spike, which is conducted passively to the recording site in the soma, is shunted by an inhibitory conductance change 14. Thus, stimulation of the caudolateral tecturn produces mixed excitation and inhibition of the M-cell and at high-stimulation voltages and short latencies the excitatory c o m p o n e n t may be dominant. Since the lateral dendrite does not normally support antidromic impulses, spike amplitude decreases as recordings are m a d e m o r e laterally (the last active zone for this potential is the initial segment and axon hillock) 13. The same is a p p a r e n t l y true for the ventral dendrite as d e m o n s t r a t e d in Fig. 5A. In this figure, the reference point is the axon hillock, at the center of the axon cap, and the maximal spike amplitude was r e c o r d e d 50 ~ m lateral, in the soma. Recordings from two sites on the ventral dendrite, 50 and 100 ktm rostral, are shown, and antidromic spike height decreased progressively in the rostral direction. The PSPs e v o k e d by tectal stimulation and r e c o r d e d at the same positions are shown in Fig. 5B, and the resultant spatial distribution of the response amplitudes, from 100 ktm rostral to 300/~m lateral, are plotted in Fig. 6. Two i m p o r t a n t p r o p e r t i e s of the tectally e v o k e d PSPs should be noted. First, their amplitudes were progressively smaller at the m o r e lateral recording sites. Second, the maximal amplitudes were in the ventral dendrite. Since maximal voltages should be r e c o r d e d closest to the source of current, the results in this e x p e r i m e n t thus suggest that tectal input to the ipsilateral M-cell is localized to the ventral dendrite rather than to the soma or lateral dendrite. Similar d a t a were o b t a i n e d in 5 additional ex-
117 rapid decay in antidromic spike height in the ventral dendrite compared to that of the lateral one (Fig. 6) may reflect inaccuracies in the m e a s u r e m e n t of distances along the former. A detailed comparison of
A: Antidromic
B: Optic
Tectum
tmsec
the two dendrites electrotonic properties suggests they are comparable (Chang et al., unpublished observations). M - c e l l r e s p o n s e s to
lmsee
Fig. 5. Distribution of tectal input on the Mauthner cell. A: antidromic action potentials are shown recorded rostrally (R) and laterally (L) from the axon hillock of the right M-cell, at the indicated distances (in j~m). The rostral recording sites were consistently more ventral than the M-cell axon hillock and were therefore in the ventral dendrite while the lateral recording sites were in the lateral dendrite. B1, B2: PSPs evoked by right tectal stimulation at two different stimulus strengths (10 and 20 V) and recorded at the corresponding sites. Note that the PSPs are maximal in amplitude at the most distal recording locus on the ventral dendrite (100#m rostral). periments using one or more recordings from the ventral dendrites.
optic n e r v e
stimulation
To determine whether the M-cell responses to optic rectum stimulation could also be evoked by visual input, we tested effects of optic nerve activation. Stimulation of the left optic nerve produced comparable responses in both M a u t h n e r cells, as demonstrated by the PSPs recorded in the left cell (Fig. 7A, B) and in the right one (Fig. 7C1). There appear to be at least two major components to the evoked PSPs, with peak latencies of about 3.5 and 7.0 ms, respectively. This was a c o m m o n finding for the responses recorded in both M-cells. Specifically, in all experiments, the mean latency for the early PSP was 3.75 + 0.54 ms in the left M-cell (n = 8) and 3.8 +_ 0.47 ms
The distribution of the PSPs evoked in the left
(n = 5) in the right one. The latency of the second
M-cell by right optic tectum stimulation was not explored systematically in the ventral dendrite, but it was clear that these responses also decreased progressively along the lateral process. That is, the input to the contralateral M-cell is most likely localized to
peak was less easily discerned. As with tectal stimulation, when the stimulus strength to the optic nerve was increased, the PSPs were large enough to fire
the soma or ventral dendrite. Also, the apparent
A
B
L.OpN~L.M-cell
IO0-
X
80-
E i
o~ 60
i|
• 2msec
"0 "~
40 •
E
<
20.
L OpN~Rt.M-cell
OpT 20V
C 2 ,'~ OpT+O~N
• OpT lOV
0 100
50
Rostral
0
50
Wm
100
/,~// , 200
300
Lateral
Fig. 6. Normalized amplitudes of the antidromic spikes and tectal PSPs from Fig, 5 plotted with reference to recording distance from the axon hillock (reference recording site; 0). While the largest antidromic spike was 50ktm lateral to the hillock, the largest PSPs were found 100 ~tm rostral to the cap, in the ventral dendrite.
2msec
Fig. 7. PSPs recorded in the M-cell in response to left optic nerve stimulation. A: stimulation of the left optic nerve (L. OpN) produces a PSP in the left M-cell soma. B: with an increase in stimulus strength the PSP evoked an orthodromic impulse. C1, C2: responses recorded in the right M-cell. When the optic nerve-evoked PSP shown in C1 followed a conditioning tectal PSP, the former was blocked (C2). Upper and lower traces in A and B are at high and low gain, respectively.
118 either M-cell. The records in Fig. 7B illustrate this effect for the left M-cell. In addition, the spatial distributions of PSP amplitudes r e c o r d e d in the left M-celt following both left nerve and right tectal stimulations were the same (not shown). Finally, to d e t e r m i n e whether the optic nerve input projects to the M-cell through connections in the tectum, a test stimulus to the left optic nerve was p a i r e d with a conditioning stimulus to the right tectum. A s shown in Fig. 7C2, the PSPs e v o k e d by optic nerve stimulation were blocked, suggesting that tectal activation reduces the excitability of the same neurons which relay optic nerve input to the M-cell.
drite of the M-cell. The c a m e r a lucida drawing in Fig. 8 demonstrates this relationship. It should be e m p h a sized that this is a 2-dimensional drawing of a 3-dimensional cell and that the soma of the M-cell is 500 um caudal to the level of the ventral d e n d r i t e - t e c t o bulbar interactions and the tip of the lateral dendrite is another 380 ~ m m o r e caudal to the s o m a in this preparation. A l t h o u g h fibers from the t e c t o b u l b a r tract project into the reticular formation throughout their caudal extent, no fibers were found in the vicinity of the M-cell soma, proximal ventral dendrite or its lateral dendrite. The contralateral t e c t o b u l b a r fibers were fewer in n u m b e r than the ipsilateral group, and few projections into the reticular formation were seen. F u r t h e r -
Histology Optic nerve projections. Projections of the optic nerve were traced to the diencephalon, p r e t e c t u m and contralateral optic tectum after t r i a t e d proline injections. In no case could label be found n e a r any portion of the M-cell (level of V I I I t h nerve entry). The results were similar after H R P application to the optic nerve, with the exception that this enzyme spread to some of the ocular muscles. A s a result, m o t o r neurons of the III, IV, V I r and VIc nuclei and a few fibers of the descending sensory branch of the Vth nerve were labeled. H o w e v e r , in no instance did any fiber a p p r o a c h the vicinity of the M-cell or its dendrites.
Fibers labeled after application of HRP to a portion of the caudolateral tectum. O b s e r v a t i o n of the wound sites indicated that H R P was confined to the tectum. W h e n the H R P was applied to the right caudolateral tectum, some fibers could be traced ventrally from the tectum along the lateral edge of the tegmentum. These fibers eventually f o r m e d an ipsilateral tectobulbar tract which was located at the ventrolateral edge of the brain. O t h e r fibers from the caudolateral rectum crossed the midline at the rostral m i d b r a i n levels and p r o j e c t e d v e n t r o m e d i a l l y in the medulla. Both the ipsi- and contralateral t e c t o b u l b a r tracts extended to the level of the M-cell, and their extent caudal to this area was not investigated. As the ipsilateral fibers pass ventrocaudally from the tectum, some extend into the reticular formation and a p p e a r to end on neurons in the vicinity. The fibers reach their ventrolateral position before they come into close apposition to the distal ventral den-
# /,/iY
"
Fig. 8. Projections of tectal fibers to the vicinity of the right Mauthner cell. This camera lucida drawing is a composite of 11 sections, each 80 ,um in thickness. Thus what appears to be in two dimensions exists over 880~m, the lateral dendrite extending caudally out of the paper and the ventral dendrite rostrally, into the paper. After the right M-cell was filled, a small wound was made in the right dorsal tectum near its caudal border and HRP was applied. The fibers that were filled projected from the tectum down the lateral edge of the brain to an ipsilateral position in the ventrolateral portion of the brain. Note that there are no filled processes in association with the lateral dendrite and soma while the ventral dendrite, particularly its distal region, is in close proximity to the fibers.
119 more, the relationship of these fibers to the contralateral M-cell is currently not known. DISCUSSION Bartelmez 1 was the first to trace crossed and uncrossed tectobulbar fibers to the vicinity of the Mauthner cell ventral dendrites of Amieurus. Specifically, he noted that the ventral dendrites which project ventrolaterally end amongst tectobulbar fibers near the periphery of the oblongata. This study confirms and expands his work and presents baseline information on the physiological properties of the ventral dendrite.
Physiological pathways from the retina to the Mauthner cell Stimulation of one optic nerve resulted in PSPs which were capable of bringing both M-cells to threshold. The blockage of this response by conditioning stimulation of the tectum reveals that visual information is not passed directly to either M-cell, but is relayed through the corresponding tectum. That is, in the context of the visual pathway studied here, the flow of information was from the left optic nerve to the right tectum and then to both M-cells. Thus the primary focus of the etectrophysiological studies was on tectal inputs to the M-cell. Stimulation of a limited portion of the tectum (the same portion injured and exposed to HRP) evoked composite PSPs in both M-cells. Although these PSPs contained inhibitory and excitatory components, excitation of both cells dominated with the greater stimulus strengths used. Thus, we speculate that this pathway could trigger visually evoked startle responses mediated by the M-cell. The tectally evoked inhibitory input seems to involve, in part, the PHP neurons, since an antidromic conditioning stimulus which elicits an EHP (caused by impulses in these interneurons) 12 can reduce the amplitude of a similar extracellular positivity resulting from tectal stimulation. PHP neurons consist of at least two classes. The first have their cell somata lateral and caudal to the M-cell lateral dendrite 26'31 and are not part of the collateral inhibitory network while collateral interneurons have cell bodies ventral to the M-cell soma and lateral to the ventral dendrite 12. This second class of PHP neurons lies in closer prox-
imity to the tectal fibers than do the former, and therefore we speculate that they may mediate at least a portion of the M-cell inhibition elicited by tectal stimulation. However, no attempt was made to localize the inhibitory synaptic input to the M-cell. The excitatory synaptic input to the M-cell was tentatively localized to the proximal ventral dendrite. However, the precise locus of these synapses was not determined because the most distal recording site on the ventral dendrite was only 100 pm from the soma. Since that point consistently exhibited the largest amplitude PSPs, and since the majority of stained uncrossed tectobulbar fibers were in close apposition to the distal ventral dendrite, we ex'pect that more distal penetrations along that dendrite will reveal even larger PSPs. We were not able to discern a clear morphological relationship between crossed tectobulbar fibers and the contralateral M-cell. Thus, the source of PSPs recorded in the left M-cell following right tectal stimulation remains ambiguous. The similar latencies of the tectally evoked PSPs recorded in the two M-cells suggests that the crossed pathway does not involve additional interneurons but that rather, terminal processes of this projection have not been adequately resolved. Furthermore, it should be stressed that the crossed tract is relatively direct and is not significantly longer than the uncrossed one (see also ref. 15). Nevertheless, we considered the possibility that the similar latencies were due to current spread from the stimulus electrodes. In one experiment, the stimulating electrode was positioned on the caudolateral surface of the right tectum, at a site where maximal amplitude PSPs were recorded from the right M-cell soma. When the stimulating electrode was moved along the surface, 0.5-1.0 mm away from this site, PSP amplitude rapidly decreased. More significantly, when the stimulus electrode was advanced through the tectum to a depth of approximately 0.5 ram, the PSPs were lost. Thus, current spread to deeper structures is unlikely and the responses studied here can reasonably be attributed to localized stimulation of the tectum. In confirmation, the latencies of the PSPs recorded in the two M-cells following optic nerve stimulation are comparable, a finding which again cannot be explained by current spread. The lateral dendritic membrane does not support spike electrogenesis and there consequently is a pro-
120 gressive reduction in antidromic spike amplitude when the recording electrode is moved to sites increasingly more lateral to the initial segment-axon hillock region. Since the antidromic spike height apparently decrements in a similar fashion along the ventral dendrite, we suggest that the two processes have comparable properties and that the localization of synaptic input to the ventral dendrite on the basis of PSP amplitudes is not contaminated by contributions from active dendritic membrane.
Morphological pathways from Mauthner cell
the retina to the
Direct projections from the retina to the Mauthner cell were not found utilizing HRP histochemistry or autoradiography, in confirmation of the electrophysiological experiments (see above). Therefore visual input to the M-cell must come via other neurons, most likely those in the diencephalon, pretectum or contralateral optic rectum (areas to which retinal fibers project) 3. Since the majority of visual input projects to the contralateral tectum in fish 22, w e concentrated our efforts on a retinotectal-M-cell pathway, although other pathways cannot be excluded. Projections of the right tectum into the medulla oblongata of fish consist of ipsi- and contralateral tectobulbar tracts which do not continue into the spinal cord ua5'22. Our studies confirm the presence and position of these tracts, which seem to be the same as those described by Bartelmez 1. The ipsilateral tract is larger than the contralateral one and both tracts extend caudally to the level of the M-cells but their zone of final termination is not established. The tectum is thought to receive afferent projections from every major brainstem division 22. Since our HRP studies do not allow the distinction between afferent and efferent tectal projections, fibers in the vicinity of the distal ventral M-cell dendrite may not be part of the tectobulbar tract. However given that degeneration studies have demonstrated a similar tectobulbar projection 15, we feel that the majority of fibers observed in this study originated from cell bodies in the tectum. The ipsilateral tract is distributed and appears to intermingle with the end of the
ventral dendrite while the contralateral input is more compact and found more medially. Indeed, it is not clear whether this latter tract comes in the vicinity of the ventral dendrite and currently there is no evidence that these fibers actually synapse on the M-cell. All uncrossed fibers were restricted to the distal portion of the ventral dendrite. That is, no tectal fibers were found in the vicinity of the lateral dendrite, soma or proximal ventral dendrite of either cell. Since the labeling was restricted to a small portion of the tectum, it is possible that only a fraction of the tectal input to the M-cell was labeled and that the overall projection is more widespread. However, when larger injection sites were used in Eugerres and Holocentrus 11 (the dorsal and dorsolateral portions of the tectum were sectioned) the size and extent of the crossed and uncrossed tectobulbar tracts were similar to those observed in the present study.
Behavioral correlates Startle responses of teleost fish are elicited by abrupt visual stimulation 6'24, acoustic or vibratory stimuli 2"19"2°'2z'z8, and electric fields 27'28. Furthermore the Mauthner cell initiated C-bend (the first stage of the startle response believed to be involved in predator avoidance) 1° of the goldfish can be elicited by similar stimuli (visual, auditory or vibratory) 7-9'3°. Although most predatory attacks would result in a combination of these stimuli to the fish, visual stimuli may provide the initial and most important input in many instances 5'6. For example fish respond to visual stimulation of predatory birds with rapid C-bends in the wild 29. In this context, it would be particularly important to know what visual cues are relayed from the tectum to the M-cell. ACKNOWLEDGEMENTS We thank M. Agostini for histological assistance, J. Lakatos, L. Marek and E. Williams for help in preparing figures. This work was supported in part by NINCDS Grant NS-15335 to D.S.F. and NSF Grant BNS 8216138 and a Faculty Research Grant to S.J.Z.
121 REFERENCES 1 Bartelmez, G.W., Mauthner's cell and the nucleus motorius tegmenti, J. Comp. Neurol., 25 (1915) 87-128. 2 Bigelow, H.B., The sense of hearing in the goldfish Carassius auratus L., Am. Naturalist, 38 (1904) 275-284. 3 Braford, M.R. Jr. and Northcutt, R.G., Organization of the diencephalon and pretectum of the ray-finned fishes. In R, Davis and R.G. Northcutt (Eds.), Fish Neurobiology, Vol. 2, Univ. of Michigan Press, Ann Arbor, 1983, pp. 117-163. 4 Cowan, W.M., Gottlieb, D.I., Hendrickson, A.E., Price, J.L. and Woolsey, T.A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 317 (19721 21-51. 5 Diamond, J., The Mauthner cell. In W.S. Hoar and D.J. Randall (Eds.), Fish Physiology, Vol. 5, Academic Press, New York, 1971, pp. 265-346. 6 Dill, L.M., The escape response of the Zebra danio (Brachydanio rerio). I. The stimulus for escape, Anim. Behav., 22 (1974) 711-722. 7 Eaton, R.C. and Farley, R.D., Mauthner neuron field potential in newly hatched larvae of the zebrafish, J. Neurophysiol., 38 (1975) 502-512. 8 Eaton, R.C., Bombardieri, R.A. and Meyer, D., The Mauthner initiated startle response in teleost fish, J. Exp. Biol., 66 (1977) 65-81. 9 Eaton, R.C., Lavender, W,A. and Wieland, C.M., Identification of Mauthner-initiated response patterns in goldfish: evidence from simultaneous cinematography and eleetrophysiology, J. Comp. Physiol., 144 (1981) 521-531. 10 Eaton, R.C. and Hackett, J.T., The role of the Mauthner cell in fast-starts involving escape in teleost fishes. In R.C. Eaton (Ed.), Neural Mechanisms of Startle Behavior, Plenum Press, New York, 1984, pp. 213-266. 11 Ebbesson, S.O.E. and Vanegas, H., Projections of the optic tectum in two teleost species, J. Comp. Neurol,, 165 (1976) 161-180. 12 Faber, D,S. and Korn, H., Electrophysiology of the Mauthner cell: basic properties, synaptic mechanisms and associated networks. In D,S. Faber and H. Korn (Eds.), Neurobiology of the Mauthner Cell, Raven Press, New York, 1978, pp. 47-131. 13 Furshpan, E.J. and Furukawa, T., Intracellular and extracellular responses of the several regions of the Mauthner cell of the goldfish, J, Neurophysiol., 25 (1962) 732-771. 14 Furukawa, T. and Furshpan, E.J., Two inhibitory mechanisms in the Mauthner neurons of goldfish, J. Neurophysiol., 26 (1963) 140-176. 15 Grover, B.G. and Sharma, S.C., Tectal projections in the goldfish (Carassius auratus): a degeneration study, J. Comp. Neurol., 184 (1979) 435-454. 16 Hanker, J.S., Yates, P.E., Metz, C.B. and Rustioni, A., A new specific sensitive and non-carcinogenic reagent for the
17
18
19
20
21
22
23
24
25
26
27
28 29 30
31
32
demonstration of horseradish peroxidase, J. Histochem., 9 (1977) 789-792, Lin, J.-W., Faber, D.S. and Wood, M,R., Organized projection of the goldfish saccular nerve onto the Mauthner cell lateral dendrite, Brain Research, 274 (1983) 319-324. Mesu[am, M.-M., The blue reaction product in horseradish peroxidase neurohistochemistry. Incubation parameters and visibility, J. Histochem. Cytochem., 24 (1976) 1273-1280. Moulton, J.M. and Dixon, R.H., Directional hearing in fishes. In W.N. Tavolga (Ed.), Marine Bio-acousties, Vol. 2, Pergamon, New York, 1967, pp. 187-232. Muller, T.J., Goldfish Respond to Sound Direction in the Mauthner Initiated Startle Behavior, Ph.D. Dissertation, University of Southern California, 1981. Nakijima, Y., Fine structure of the synaptic endings on the Mauthner cell of the goldfish, J. Comp. Neurol., 156 (1974) 375-402. Northeutt, R.G., Evolution of the optic tectum in rayfinned fishes. In R.E. Davis and R.G. Northcutt (Eds.), Fish Neurobiology, Vol. 2, Univ. of Michigan Press, Ann Arbor, 1983, pp. 1-42. Robertson, J.D., Bodenheimer, T.S. and Stage, D.E., The ultrastructure of Mauthner cell synapses and nodes in goldfish brains, J. Cell Biol., 19 {1963) 159-199. Rodgers, W.L., Melzack, R. and Segal, J.R., 'Tail-flip' re* sponse in goldfish, J. Comp. Physiol. Psychol., 56 (1963) 917-923. Rosene, D.L. and Mesulam, M.-M., Fixation variables in horseradish peroxidase neurohistochemistry. I. The effects of fixation time and perfusion procedures upon enzyme activity, J. Histochem. Cytochem., 26 (1978) 28-39. Triller, A. and Korn, H., Morphologically distinct classes of inhibitory synapses arise from the same neurons: ultrastructural identification from crossed vestibular interneurons intracellularly stained with HRP, J. Comp. Neurol., 203 (1982) 131-155. Webb, P., Acceleration performance of rainbow trout, Salmo gairdneri, and green sunfish, Lepom& cyanellus, J. Exp. BioL, 63 (1975) 451-465. Wilson, D.M., Function of giant Mauthner's neurons in the lungfish, Science, 129 (1959) 841-842, Zottoli, S.J., Fishing behavior of common grackles, Auk, 93 (1976) 640-642. Zottoli, S.J., Correlation of the startle reflex and Mauthner cell auditory responses in unrestrained goldfish, J. Exp. Biol., 66 (1977) 243-254. Zottoli, S,J, and Faber, D.S., An identifiable class of statoacoustic interneurons with bilateral projections in the goldfish medulla, Neuroscience, 5 (1980) 1287-1302. Zottoli, S.J., Hangen, D.H. and Faber, D.S., The axon reaction of the goldfish Mauthner cell and factors that influence its morphological variability, J. Comp. Neurol,, 230 (1984) 497-516.
Brain Re'search, 401 (1987) 113-121 Elsevier BRE 12270
Localization of optic tectal input to the ventral dendrite of the goldfish Mauthner cell Steven J. Zottoli 1, A n d r e w R. H o r d e s 2 and D o n a l d S. F a b e r 2 1Department of Biology, Williams College, Williamstown, MA 01267 (U.S.A.) and 2Division of Neurobiology, Department of Physiology, State University of New York. Buffalo. NY14214 (U.S.A.) (Accepted 3 June 1986)
Key words: Mauthner cell, Segregated synaptic input; Visual input; Startle response; Mauthner cell ventral dendrite
Although visually evoked Mauthner cell (M-cell) startle responses occur in the goldfish, the afferent projections underlying these reactions have not been previously studied. We have recorded from the M-cell while stimulating the left optic nerve and/or right optic tectum and have traced projections of the optic nerve and restricted areas of the optic rectum using HRP histoehemistry and autoradiography. Tectal stimulation elicits similar postsynaptic potentials (PSPs) in both M-cells. The responses recorded in the right (ipsilateral) cell were localized to its ventral dendrite. The existence of uncrossed tectal projections to the ventral dendrite was confirmed morphologically following application of horseradish peroxidase (HRP) to the optic tectum. The PSPs contained both inhibitory and excitatory components, but with adequate stimulus strength, excitation of either M-cell dominated. Thus, this pathway is probably sufficient to trigger visually evoked startle responses mediated by the M-cell. Stimulation of the left optic nerve also evoked PSPs capable of bringing both M-cells to threshold. The blockage of this response by conditioning stimulation of the right tectum suggests that the visual information is relayed to the M-cells through this structure. In support of these findings, no label was found near any portion of the M-cell after either intraocular injection of tritiated proline or application of HRP to the cut end of the optic nerve. In summary, visual input to the M-cell is mediated via projections from the tectum, is segregated onto the ventral dendrite, and is capable of bringing this neuron to threshold. This pathway presumably accounts for the demonstrated behavioral efficacy of visual stimuli in evoking a startle response. INTRODUCTION The M a u t h n e r cell (M-cell) of goldfish mediates startle responses evoked by sound9'3°and by visual stimuli924. The afferent pathway subserving the auditory response is quite direct, primarily involving the large saccular afferents which end as myelinated clubs on the distal lateral dendrite 1'17'21'23. In contrast, the afferent projections underlying the visually evoked response remain u n k n o w n , although Bartelmez 1 traced fibers from the optic tectum to the inferior and superior ventral dendrites of the M-cell in Amieurus. To better u n d e r s t a n d the input circuitry involved in M-cell-initiated startle behaviors resulting from visual stimuli, we recorded from this n e u r o n while stimulating the optic nerve and/or the optic tecturn and traced projections of the optic nerve and restricted areas of the optic tectum using H R P histochemistry and autoradiography. We found that post-
synaptic potentials (PSPs) due to tectal stimulation are localized onto the M a u t h n e r cell ventral dendrite and that this input is capable of bringing the cell to threshold. The existence of tectal projections to the ventral dendrite was confirmed morphologically, thereby providing another example of segregated synaptic input onto the M-cell. In addition, the results indicate that this dendrite is accessible for electrophysiological investigations and is electrically inexcitable, like the well-studied lateral process. MATERIALS AND METHODS C o m m o n goldfish, 12-18 cm in total length were used for electrophysiological and histological experiments. The fish were exposed to a 12-12 h light/dark cycle and kept between 15 and 16 °C. Electrophysiological experiments were conducted using standard surgical and recording techniques.
Correspondence: S. Zottoli, Department of Biology, Williams College, Williamstown, MA 01267, U.S.A.
114 Fish were anesthetized with MS-222 (tricaine methanesulfonate) and intracellular recordings were obtained from either M-cell with glass microelectrodes filled with 2.7 M KC1 (R 1 and R2 of Fig. 1). Bipolar stimulating electrodes were placed on the left optic nerve (S1) and right optic tectum ($2; the majority of retinal afferents to the tectum are known to terminate in the contralateral tectum 22) for activation of orthodromic inputs to the Mauthner cell and on the vertebrae overlying the spinal cord near the caudal edge of the dorsal fin ($3) for antidromic activation (Fig. 1). The optic nerve was exposed by cutting the conjuctiva around the rim of the orbit, severing the extraocular muscles and rotating the eyeball. The caudolateral edge of the optic tectum was chosen as the area to stimulate because of its accessibility and the fact that stimulation resulted in PSPs capable of firing the Mauthner cell. The M-cell was first located by its large antidromically activated extracellular spike recorded in the axon cap 12A3, and subsequent intracellular penetrations were made from the M-cell soma, from the lateral dendrite as far as 300/,tm caudolateral to the axon cap, and from the ventral den-
$2
7---~ Sa Fig. 1. Schematic diagram of the stimulating and recording arrangement. The left optic nerve (S 0 and the right optic rectum ($2) were stimulated for orthodromic activation of the Mauthner cells (shown caudal to the tecta). The cells were antidromically activated with an electrode ($3) placed on the vertebrae over the spinal cord. Intracellular recordings were obtained from the left (R 0 and right (R2) Mauthner cells.
drite as far as 100/~m rostral and ventral to the axon cap. In previous experiments with the M-cell 31, fish were routinely positioned within the holding chamber so that the surface of the medulla oblongata was approximately horizontal. The recording electrode was in a sagittal plane with an orientation of about 70 ° to the horizontal and was inserted into the brain in a rostroventral direction. Since the ventral dendrite courses in the same general direction, such an approach would have reduced the accessibility of this neurite and increased the likelihood of damage. Certainly, it would have been difficult to determine the exact position of the electrode tip relative to the soma and the length of the dendrite. Therefore, in the present experimental series, fish were positioned in the holding chamber so that the head tilted upwards (about 20°). With this orientation the initial portion of the ventral dendrite was approximately perpendicular to the recording electrode and distal segments were more accessible.
Morphological studies Autoradiography. Unilateral injections of tritiated proline (20 ~Ci; New England Nuclear) were made into a single eye of each of 6 goldfish under anesthesia (MS-222). The fish were held at 15-16 °C and sacrificed, 14 days postoperatively, by perfusion with 0.7% saline followed with fixative (90 m180% E T O H , 5 ml formalin, 5 ml glacial acetic acid). The brains were removed, dehydrated (ethanol, t-butyl mixture), embedded in paraffin, transversely sectioned at 15 ~m and mounted on gelatin-coated slides. The slides were dipped in the dark into Kodak NT2B emulsion and left in the refrigerator for 31 days. They were then developed in Dektol and stained with Cresyl violet acetate 4. Horseradish peroxidase. A bead of recrystallized horseradish peroxidase (HRP; about 250 ktg) was applied 3~ to either the completely transected optic nerve (one fish) or to a superficial cut of the optic tecturn (5 fish) at sites equivalent to those used for stimulation in the electrophysiological studies. After HRP application to the optic nerve the space behind the eyeball was filled with a Vasoline-paraffin oil mixture to reduce spread of the enzyme and the eye was placed back in its socket. In the experiments utilizing H R P application to the tectum the brain cavity
115 was filled with a Vaseline-paraffin oil mixture covered by thin plastic and c a p p e d with an epoxy plug anchored to the skull 3°'32. In one instance the M a u t h n e r cell soma was iontophoretically filled with H R P (4% H R P , Sigma VI, in 0.05 M Tris with 0.2 M KC1, p H 8.62; initial resistance of 2 0 - 5 0 Mr2) and H R P was also applied to the tectum. A f t e r intervals of 5 - 7 days (15.6 °C holding t e m p e r a t u r e ) , the fish were perfused with 20 ml of 2.5% glutaraldehyde in phosphate buffer (pH 7.2), and their brains were removed and placed successively in fresh fixative for 3 h, in 20%. sucrose for 5 h and in a gelatin albumin mixture overnight. The brains were subsequently sectioned at either 40 or 5 0 - 6 0 # m and stained (freefloating at 40 t~m and after drying to microscope slides at between 50 and 60 Bm) with either the chromogen benzidine dihydrochloride ls.25 or H a n k e r Yates reagent >. Brains stained with H a n k e r Yates reagent were counterstained with Neutral red or Cresyl violet acetate. RESULTS M-cell response to optic tectal stimulation
Stimulation of the caudolateral edge of the right rectum resulted in postsynaptic potentials (PSPs) recorded in both M-cell s o m a t a (50 # m lateral to the
A1 AD ~Left M-cell
A2
lOV
B1
_
axon cap). This bilateral response is shown in Fig. 2, which is from an e x p e r i m e n t where recordings were first obtained from the left M-cell ( A l, A2) and then from the right one (B 1, B2). In both, the antidromic spike amplitudes were at least 30 mV ( A 1, B1). Stimulation of the right optic tectum, at low stimulus strengths, e v o k e d c o m p a r a b l e PSPs in the two cells, while when the strength was d o u b l e d , the response in the right M-cell was greater. In 8 experiments the latency of the response in the ipsilateral M-cell was 0.63 _+ 0.09 ms (mean _+ S.D.; n = 12), with a m e a n peak latency of 0.96 ms, while the PSP r e c o r d e d in the cell contralateral to the stimulus site had a latency of 0.60 _+ 0.08 ms (n = 5) and a mean p e a k latency of 1.03 ms. That is, there was no difference in the timing of the ipsi- and contralateral responses to stimulation of the right optic rectum. A s shown in Fig. 2, when the stimulation voltage was increased a second p e a k with a latency of about 1.9 ms was discernible in both cells. In general, the bilateral PSPs were similar in nature, and the following description focuses on the ipsilateral M-cell responses to right optic tectum stimulation. Typically, if the tectal stimulus strength was increased sufficiently, both M-cells could be excited b e y o n d threshold. This o r t h o d r o m i c excitation is shown for the ipsilateral M-cell in Fig. 3, where it can be c o m p a r e d with antidromic activation of the cell.
Right M-cell
.L__: B2
A1
Extracell.
B1
Intracell.,
,=@~
AD
J m
w
20V
'~11~ .,. ~
1msec
Fig. 2. Bilateral Mauthner cell inputs from the right optic tecturn. A t, A 2 and BI, B2: recordings from, respectively, the left and right M-cells of responses to stimulation of the surface of the right optic tectum at its caudolateral border. The electrodes were located in the soma, as indicated by the amplitudes of the antidromic spikes (Al, B1). In A 2 and B 2, the optic tectum was stimulated at 10 (upper traces) and 20 V (lower traces) and compound PSPs were evoked. Note that at 10 V one peak dominates while at 20 V a second later peak is comparable in amplitude to the first.
J
2-~-~-c
lmsec
Fig. 3. Extra- and intracellular recordings near the M-cell soma. AI: antidromic activation of the right M-cell resulted in an extracellular spike recorded within the axon cap and followed by an EHP (arrow). At: right tectal stimulation also evoked a positivity within the axon cap (arrow). In these superimposed traces, the M-cell fired in one case. BI, B2: intracellular recordings obtained in the same experiment from the soma 50 l~m lateral to the axon cap. BI: the antidromic spike is approximately 40 mV in amplitude. B2: tectal stimulation resulted in an EPSP which in this example was capable of bringing the cell to threshold.
116 Both extracellularly and intracellularly r e c o r d e d spikes are illustrated, and in B 2 a clear EPSP precedes the orthodromic action potential. Even though the PSP e v o k e d in the ipsilateral M-cell was often sufficiently excitatory to evoke an o r t h o d r o m i c action potential it contained an inhibitory c o m p o n e n t , as shown by two methods. First, antidromic stimulation activates a pool of interneurons (PHP neurons 12) inhibitory to the M-cell. Their impulses produce an extracellular positivity ( E H P ) in the axon cap, which is the first phase of a t w o - c o m p o nent (electrical and chemical) inhibition. A similar potential was r e c o r d e d in the axon cap after tectal stimulation (Figs. 3 A 2 and 4A1). In fact, as shown in
A1
OpT
A2
AD-FOpT
,I
B1
.I
opt
lmsec
Fig. 4. Inhibitory component of the tectal PSP. A1, A2: extracellular recordings from the right M-cell's axon cap. AI: stimulation of the right tectum (OPT) produced an extracellular positivity. A2: when a conditioning antidromic stimulus, which evoked a negative M-cell spike and a subsequent extrinsic hyperpolarizing potential (EHP) was followed by the tectal stimulation, the amplitude of the tectal response was reduced (the bar above the test response in A 2 indicates the control amplitude in AI). This reduction suggests that some of the same inhibitory interneurons which produce the EHP after antidromic stimulation also contributed to the positivity recorded after tectal stimulation (superimposed traces with and without tectal stimulation). B 1, B2: intracellular recordings in the same experiment, of M-cell responses evoked by the stimulus paradigms used for A 1 and A 2. BI: composite PSP recorded in the M-cell soma after right optic tectum stimulation. B2: a conditioning tectal PSP produces a conductance increase, as indicated by the decrease in amplitude of test antidromic action potentials superimposed at different intervals. The first spike is equivalent in amplitude to a control and thus acts as a reference, Note that the conductance change was maximal approximately 4 ms after the first peak of the tectal PSP. Two or more superimposed sweeps in all records.
Fig. 4A2, when a conditioning antidromically e v o k e d E H P was followed by a test tectally e v o k e d positivity, the latter was reduced in amplitude. This finding suggests that tectal neurons may activate some of the same P H P neurons fired by antidromic stimulation of the M-cell's recurrent inhibitory n e t w o r k (alternatively, the interneurons involved m a y have been inhibited by the spinal stimulation). Second, a m o r e direct means of unmasking the inhibitory c o m p o n e n t of tectal PSPs involved superimposing antidromic action potentials on a conditioning PSP (Fig. 4B1) e v o k e d by optic tectum stimulation. A s is a p p a r e n t in Fig. 4B2, this protocol decreases the antidromic spike height, with the maximal reduction occurring about 4 ms following the first p e a k of the tectal PSP. The explanation for this result is that the antidromic spike, which is conducted passively to the recording site in the soma, is shunted by an inhibitory conductance change 14. Thus, stimulation of the caudolateral tecturn produces mixed excitation and inhibition of the M-cell and at high-stimulation voltages and short latencies the excitatory c o m p o n e n t may be dominant. Since the lateral dendrite does not normally support antidromic impulses, spike amplitude decreases as recordings are m a d e m o r e laterally (the last active zone for this potential is the initial segment and axon hillock) 13. The same is a p p a r e n t l y true for the ventral dendrite as d e m o n s t r a t e d in Fig. 5A. In this figure, the reference point is the axon hillock, at the center of the axon cap, and the maximal spike amplitude was r e c o r d e d 50 ~ m lateral, in the soma. Recordings from two sites on the ventral dendrite, 50 and 100 ktm rostral, are shown, and antidromic spike height decreased progressively in the rostral direction. The PSPs e v o k e d by tectal stimulation and r e c o r d e d at the same positions are shown in Fig. 5B, and the resultant spatial distribution of the response amplitudes, from 100 ktm rostral to 300/~m lateral, are plotted in Fig. 6. Two i m p o r t a n t p r o p e r t i e s of the tectally e v o k e d PSPs should be noted. First, their amplitudes were progressively smaller at the m o r e lateral recording sites. Second, the maximal amplitudes were in the ventral dendrite. Since maximal voltages should be r e c o r d e d closest to the source of current, the results in this e x p e r i m e n t thus suggest that tectal input to the ipsilateral M-cell is localized to the ventral dendrite rather than to the soma or lateral dendrite. Similar d a t a were o b t a i n e d in 5 additional ex-
117 rapid decay in antidromic spike height in the ventral dendrite compared to that of the lateral one (Fig. 6) may reflect inaccuracies in the m e a s u r e m e n t of distances along the former. A detailed comparison of
A: Antidromic
B: Optic
Tectum
tmsec
the two dendrites electrotonic properties suggests they are comparable (Chang et al., unpublished observations). M - c e l l r e s p o n s e s to
lmsee
Fig. 5. Distribution of tectal input on the Mauthner cell. A: antidromic action potentials are shown recorded rostrally (R) and laterally (L) from the axon hillock of the right M-cell, at the indicated distances (in j~m). The rostral recording sites were consistently more ventral than the M-cell axon hillock and were therefore in the ventral dendrite while the lateral recording sites were in the lateral dendrite. B1, B2: PSPs evoked by right tectal stimulation at two different stimulus strengths (10 and 20 V) and recorded at the corresponding sites. Note that the PSPs are maximal in amplitude at the most distal recording locus on the ventral dendrite (100#m rostral). periments using one or more recordings from the ventral dendrites.
optic n e r v e
stimulation
To determine whether the M-cell responses to optic rectum stimulation could also be evoked by visual input, we tested effects of optic nerve activation. Stimulation of the left optic nerve produced comparable responses in both M a u t h n e r cells, as demonstrated by the PSPs recorded in the left cell (Fig. 7A, B) and in the right one (Fig. 7C1). There appear to be at least two major components to the evoked PSPs, with peak latencies of about 3.5 and 7.0 ms, respectively. This was a c o m m o n finding for the responses recorded in both M-cells. Specifically, in all experiments, the mean latency for the early PSP was 3.75 + 0.54 ms in the left M-cell (n = 8) and 3.8 +_ 0.47 ms
The distribution of the PSPs evoked in the left
(n = 5) in the right one. The latency of the second
M-cell by right optic tectum stimulation was not explored systematically in the ventral dendrite, but it was clear that these responses also decreased progressively along the lateral process. That is, the input to the contralateral M-cell is most likely localized to
peak was less easily discerned. As with tectal stimulation, when the stimulus strength to the optic nerve was increased, the PSPs were large enough to fire
the soma or ventral dendrite. Also, the apparent
A
B
L.OpN~L.M-cell
IO0-
X
80-
E i
o~ 60
i|
• 2msec
"0 "~
40 •
E
<
20.
L OpN~Rt.M-cell
OpT 20V
C 2 ,'~ OpT+O~N
• OpT lOV
0 100
50
Rostral
0
50
Wm
100
/,~// , 200
300
Lateral
Fig. 6. Normalized amplitudes of the antidromic spikes and tectal PSPs from Fig, 5 plotted with reference to recording distance from the axon hillock (reference recording site; 0). While the largest antidromic spike was 50ktm lateral to the hillock, the largest PSPs were found 100 ~tm rostral to the cap, in the ventral dendrite.
2msec
Fig. 7. PSPs recorded in the M-cell in response to left optic nerve stimulation. A: stimulation of the left optic nerve (L. OpN) produces a PSP in the left M-cell soma. B: with an increase in stimulus strength the PSP evoked an orthodromic impulse. C1, C2: responses recorded in the right M-cell. When the optic nerve-evoked PSP shown in C1 followed a conditioning tectal PSP, the former was blocked (C2). Upper and lower traces in A and B are at high and low gain, respectively.
118 either M-cell. The records in Fig. 7B illustrate this effect for the left M-cell. In addition, the spatial distributions of PSP amplitudes r e c o r d e d in the left M-celt following both left nerve and right tectal stimulations were the same (not shown). Finally, to d e t e r m i n e whether the optic nerve input projects to the M-cell through connections in the tectum, a test stimulus to the left optic nerve was p a i r e d with a conditioning stimulus to the right tectum. A s shown in Fig. 7C2, the PSPs e v o k e d by optic nerve stimulation were blocked, suggesting that tectal activation reduces the excitability of the same neurons which relay optic nerve input to the M-cell.
drite of the M-cell. The c a m e r a lucida drawing in Fig. 8 demonstrates this relationship. It should be e m p h a sized that this is a 2-dimensional drawing of a 3-dimensional cell and that the soma of the M-cell is 500 um caudal to the level of the ventral d e n d r i t e - t e c t o bulbar interactions and the tip of the lateral dendrite is another 380 ~ m m o r e caudal to the s o m a in this preparation. A l t h o u g h fibers from the t e c t o b u l b a r tract project into the reticular formation throughout their caudal extent, no fibers were found in the vicinity of the M-cell soma, proximal ventral dendrite or its lateral dendrite. The contralateral t e c t o b u l b a r fibers were fewer in n u m b e r than the ipsilateral group, and few projections into the reticular formation were seen. F u r t h e r -
Histology Optic nerve projections. Projections of the optic nerve were traced to the diencephalon, p r e t e c t u m and contralateral optic tectum after t r i a t e d proline injections. In no case could label be found n e a r any portion of the M-cell (level of V I I I t h nerve entry). The results were similar after H R P application to the optic nerve, with the exception that this enzyme spread to some of the ocular muscles. A s a result, m o t o r neurons of the III, IV, V I r and VIc nuclei and a few fibers of the descending sensory branch of the Vth nerve were labeled. H o w e v e r , in no instance did any fiber a p p r o a c h the vicinity of the M-cell or its dendrites.
Fibers labeled after application of HRP to a portion of the caudolateral tectum. O b s e r v a t i o n of the wound sites indicated that H R P was confined to the tectum. W h e n the H R P was applied to the right caudolateral tectum, some fibers could be traced ventrally from the tectum along the lateral edge of the tegmentum. These fibers eventually f o r m e d an ipsilateral tectobulbar tract which was located at the ventrolateral edge of the brain. O t h e r fibers from the caudolateral rectum crossed the midline at the rostral m i d b r a i n levels and p r o j e c t e d v e n t r o m e d i a l l y in the medulla. Both the ipsi- and contralateral t e c t o b u l b a r tracts extended to the level of the M-cell, and their extent caudal to this area was not investigated. As the ipsilateral fibers pass ventrocaudally from the tectum, some extend into the reticular formation and a p p e a r to end on neurons in the vicinity. The fibers reach their ventrolateral position before they come into close apposition to the distal ventral den-
# /,/iY
"
Fig. 8. Projections of tectal fibers to the vicinity of the right Mauthner cell. This camera lucida drawing is a composite of 11 sections, each 80 ,um in thickness. Thus what appears to be in two dimensions exists over 880~m, the lateral dendrite extending caudally out of the paper and the ventral dendrite rostrally, into the paper. After the right M-cell was filled, a small wound was made in the right dorsal tectum near its caudal border and HRP was applied. The fibers that were filled projected from the tectum down the lateral edge of the brain to an ipsilateral position in the ventrolateral portion of the brain. Note that there are no filled processes in association with the lateral dendrite and soma while the ventral dendrite, particularly its distal region, is in close proximity to the fibers.
119 more, the relationship of these fibers to the contralateral M-cell is currently not known. DISCUSSION Bartelmez 1 was the first to trace crossed and uncrossed tectobulbar fibers to the vicinity of the Mauthner cell ventral dendrites of Amieurus. Specifically, he noted that the ventral dendrites which project ventrolaterally end amongst tectobulbar fibers near the periphery of the oblongata. This study confirms and expands his work and presents baseline information on the physiological properties of the ventral dendrite.
Physiological pathways from the retina to the Mauthner cell Stimulation of one optic nerve resulted in PSPs which were capable of bringing both M-cells to threshold. The blockage of this response by conditioning stimulation of the tectum reveals that visual information is not passed directly to either M-cell, but is relayed through the corresponding tectum. That is, in the context of the visual pathway studied here, the flow of information was from the left optic nerve to the right tectum and then to both M-cells. Thus the primary focus of the etectrophysiological studies was on tectal inputs to the M-cell. Stimulation of a limited portion of the tectum (the same portion injured and exposed to HRP) evoked composite PSPs in both M-cells. Although these PSPs contained inhibitory and excitatory components, excitation of both cells dominated with the greater stimulus strengths used. Thus, we speculate that this pathway could trigger visually evoked startle responses mediated by the M-cell. The tectally evoked inhibitory input seems to involve, in part, the PHP neurons, since an antidromic conditioning stimulus which elicits an EHP (caused by impulses in these interneurons) 12 can reduce the amplitude of a similar extracellular positivity resulting from tectal stimulation. PHP neurons consist of at least two classes. The first have their cell somata lateral and caudal to the M-cell lateral dendrite 26'31 and are not part of the collateral inhibitory network while collateral interneurons have cell bodies ventral to the M-cell soma and lateral to the ventral dendrite 12. This second class of PHP neurons lies in closer prox-
imity to the tectal fibers than do the former, and therefore we speculate that they may mediate at least a portion of the M-cell inhibition elicited by tectal stimulation. However, no attempt was made to localize the inhibitory synaptic input to the M-cell. The excitatory synaptic input to the M-cell was tentatively localized to the proximal ventral dendrite. However, the precise locus of these synapses was not determined because the most distal recording site on the ventral dendrite was only 100 pm from the soma. Since that point consistently exhibited the largest amplitude PSPs, and since the majority of stained uncrossed tectobulbar fibers were in close apposition to the distal ventral dendrite, we ex'pect that more distal penetrations along that dendrite will reveal even larger PSPs. We were not able to discern a clear morphological relationship between crossed tectobulbar fibers and the contralateral M-cell. Thus, the source of PSPs recorded in the left M-cell following right tectal stimulation remains ambiguous. The similar latencies of the tectally evoked PSPs recorded in the two M-cells suggests that the crossed pathway does not involve additional interneurons but that rather, terminal processes of this projection have not been adequately resolved. Furthermore, it should be stressed that the crossed tract is relatively direct and is not significantly longer than the uncrossed one (see also ref. 15). Nevertheless, we considered the possibility that the similar latencies were due to current spread from the stimulus electrodes. In one experiment, the stimulating electrode was positioned on the caudolateral surface of the right tectum, at a site where maximal amplitude PSPs were recorded from the right M-cell soma. When the stimulating electrode was moved along the surface, 0.5-1.0 mm away from this site, PSP amplitude rapidly decreased. More significantly, when the stimulus electrode was advanced through the tectum to a depth of approximately 0.5 ram, the PSPs were lost. Thus, current spread to deeper structures is unlikely and the responses studied here can reasonably be attributed to localized stimulation of the tectum. In confirmation, the latencies of the PSPs recorded in the two M-cells following optic nerve stimulation are comparable, a finding which again cannot be explained by current spread. The lateral dendritic membrane does not support spike electrogenesis and there consequently is a pro-
120 gressive reduction in antidromic spike amplitude when the recording electrode is moved to sites increasingly more lateral to the initial segment-axon hillock region. Since the antidromic spike height apparently decrements in a similar fashion along the ventral dendrite, we suggest that the two processes have comparable properties and that the localization of synaptic input to the ventral dendrite on the basis of PSP amplitudes is not contaminated by contributions from active dendritic membrane.
Morphological pathways from Mauthner cell
the retina to the
Direct projections from the retina to the Mauthner cell were not found utilizing HRP histochemistry or autoradiography, in confirmation of the electrophysiological experiments (see above). Therefore visual input to the M-cell must come via other neurons, most likely those in the diencephalon, pretectum or contralateral optic rectum (areas to which retinal fibers project) 3. Since the majority of visual input projects to the contralateral tectum in fish 22, w e concentrated our efforts on a retinotectal-M-cell pathway, although other pathways cannot be excluded. Projections of the right tectum into the medulla oblongata of fish consist of ipsi- and contralateral tectobulbar tracts which do not continue into the spinal cord ua5'22. Our studies confirm the presence and position of these tracts, which seem to be the same as those described by Bartelmez 1. The ipsilateral tract is larger than the contralateral one and both tracts extend caudally to the level of the M-cells but their zone of final termination is not established. The tectum is thought to receive afferent projections from every major brainstem division 22. Since our HRP studies do not allow the distinction between afferent and efferent tectal projections, fibers in the vicinity of the distal ventral M-cell dendrite may not be part of the tectobulbar tract. However given that degeneration studies have demonstrated a similar tectobulbar projection 15, we feel that the majority of fibers observed in this study originated from cell bodies in the tectum. The ipsilateral tract is distributed and appears to intermingle with the end of the
ventral dendrite while the contralateral input is more compact and found more medially. Indeed, it is not clear whether this latter tract comes in the vicinity of the ventral dendrite and currently there is no evidence that these fibers actually synapse on the M-cell. All uncrossed fibers were restricted to the distal portion of the ventral dendrite. That is, no tectal fibers were found in the vicinity of the lateral dendrite, soma or proximal ventral dendrite of either cell. Since the labeling was restricted to a small portion of the tectum, it is possible that only a fraction of the tectal input to the M-cell was labeled and that the overall projection is more widespread. However, when larger injection sites were used in Eugerres and Holocentrus 11 (the dorsal and dorsolateral portions of the tectum were sectioned) the size and extent of the crossed and uncrossed tectobulbar tracts were similar to those observed in the present study.
Behavioral correlates Startle responses of teleost fish are elicited by abrupt visual stimulation 6'24, acoustic or vibratory stimuli 2"19"2°'2z'z8, and electric fields 27'28. Furthermore the Mauthner cell initiated C-bend (the first stage of the startle response believed to be involved in predator avoidance) 1° of the goldfish can be elicited by similar stimuli (visual, auditory or vibratory) 7-9'3°. Although most predatory attacks would result in a combination of these stimuli to the fish, visual stimuli may provide the initial and most important input in many instances 5'6. For example fish respond to visual stimulation of predatory birds with rapid C-bends in the wild 29. In this context, it would be particularly important to know what visual cues are relayed from the tectum to the M-cell. ACKNOWLEDGEMENTS We thank M. Agostini for histological assistance, J. Lakatos, L. Marek and E. Williams for help in preparing figures. This work was supported in part by NINCDS Grant NS-15335 to D.S.F. and NSF Grant BNS 8216138 and a Faculty Research Grant to S.J.Z.
121 REFERENCES 1 Bartelmez, G.W., Mauthner's cell and the nucleus motorius tegmenti, J. Comp. Neurol., 25 (1915) 87-128. 2 Bigelow, H.B., The sense of hearing in the goldfish Carassius auratus L., Am. Naturalist, 38 (1904) 275-284. 3 Braford, M.R. Jr. and Northcutt, R.G., Organization of the diencephalon and pretectum of the ray-finned fishes. In R, Davis and R.G. Northcutt (Eds.), Fish Neurobiology, Vol. 2, Univ. of Michigan Press, Ann Arbor, 1983, pp. 117-163. 4 Cowan, W.M., Gottlieb, D.I., Hendrickson, A.E., Price, J.L. and Woolsey, T.A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 317 (19721 21-51. 5 Diamond, J., The Mauthner cell. In W.S. Hoar and D.J. Randall (Eds.), Fish Physiology, Vol. 5, Academic Press, New York, 1971, pp. 265-346. 6 Dill, L.M., The escape response of the Zebra danio (Brachydanio rerio). I. The stimulus for escape, Anim. Behav., 22 (1974) 711-722. 7 Eaton, R.C. and Farley, R.D., Mauthner neuron field potential in newly hatched larvae of the zebrafish, J. Neurophysiol., 38 (1975) 502-512. 8 Eaton, R.C., Bombardieri, R.A. and Meyer, D., The Mauthner initiated startle response in teleost fish, J. Exp. Biol., 66 (1977) 65-81. 9 Eaton, R.C., Lavender, W,A. and Wieland, C.M., Identification of Mauthner-initiated response patterns in goldfish: evidence from simultaneous cinematography and eleetrophysiology, J. Comp. Physiol., 144 (1981) 521-531. 10 Eaton, R.C. and Hackett, J.T., The role of the Mauthner cell in fast-starts involving escape in teleost fishes. In R.C. Eaton (Ed.), Neural Mechanisms of Startle Behavior, Plenum Press, New York, 1984, pp. 213-266. 11 Ebbesson, S.O.E. and Vanegas, H., Projections of the optic tectum in two teleost species, J. Comp. Neurol,, 165 (1976) 161-180. 12 Faber, D,S. and Korn, H., Electrophysiology of the Mauthner cell: basic properties, synaptic mechanisms and associated networks. In D,S. Faber and H. Korn (Eds.), Neurobiology of the Mauthner Cell, Raven Press, New York, 1978, pp. 47-131. 13 Furshpan, E.J. and Furukawa, T., Intracellular and extracellular responses of the several regions of the Mauthner cell of the goldfish, J, Neurophysiol., 25 (1962) 732-771. 14 Furukawa, T. and Furshpan, E.J., Two inhibitory mechanisms in the Mauthner neurons of goldfish, J. Neurophysiol., 26 (1963) 140-176. 15 Grover, B.G. and Sharma, S.C., Tectal projections in the goldfish (Carassius auratus): a degeneration study, J. Comp. Neurol., 184 (1979) 435-454. 16 Hanker, J.S., Yates, P.E., Metz, C.B. and Rustioni, A., A new specific sensitive and non-carcinogenic reagent for the
17
18
19
20
21
22
23
24
25
26
27
28 29 30
31
32
demonstration of horseradish peroxidase, J. Histochem., 9 (1977) 789-792, Lin, J.-W., Faber, D.S. and Wood, M,R., Organized projection of the goldfish saccular nerve onto the Mauthner cell lateral dendrite, Brain Research, 274 (1983) 319-324. Mesu[am, M.-M., The blue reaction product in horseradish peroxidase neurohistochemistry. Incubation parameters and visibility, J. Histochem. Cytochem., 24 (1976) 1273-1280. Moulton, J.M. and Dixon, R.H., Directional hearing in fishes. In W.N. Tavolga (Ed.), Marine Bio-acousties, Vol. 2, Pergamon, New York, 1967, pp. 187-232. Muller, T.J., Goldfish Respond to Sound Direction in the Mauthner Initiated Startle Behavior, Ph.D. Dissertation, University of Southern California, 1981. Nakijima, Y., Fine structure of the synaptic endings on the Mauthner cell of the goldfish, J. Comp. Neurol., 156 (1974) 375-402. Northeutt, R.G., Evolution of the optic tectum in rayfinned fishes. In R.E. Davis and R.G. Northcutt (Eds.), Fish Neurobiology, Vol. 2, Univ. of Michigan Press, Ann Arbor, 1983, pp. 1-42. Robertson, J.D., Bodenheimer, T.S. and Stage, D.E., The ultrastructure of Mauthner cell synapses and nodes in goldfish brains, J. Cell Biol., 19 {1963) 159-199. Rodgers, W.L., Melzack, R. and Segal, J.R., 'Tail-flip' re* sponse in goldfish, J. Comp. Physiol. Psychol., 56 (1963) 917-923. Rosene, D.L. and Mesulam, M.-M., Fixation variables in horseradish peroxidase neurohistochemistry. I. The effects of fixation time and perfusion procedures upon enzyme activity, J. Histochem. Cytochem., 26 (1978) 28-39. Triller, A. and Korn, H., Morphologically distinct classes of inhibitory synapses arise from the same neurons: ultrastructural identification from crossed vestibular interneurons intracellularly stained with HRP, J. Comp. Neurol., 203 (1982) 131-155. Webb, P., Acceleration performance of rainbow trout, Salmo gairdneri, and green sunfish, Lepom& cyanellus, J. Exp. BioL, 63 (1975) 451-465. Wilson, D.M., Function of giant Mauthner's neurons in the lungfish, Science, 129 (1959) 841-842, Zottoli, S.J., Fishing behavior of common grackles, Auk, 93 (1976) 640-642. Zottoli, S.J., Correlation of the startle reflex and Mauthner cell auditory responses in unrestrained goldfish, J. Exp. Biol., 66 (1977) 243-254. Zottoli, S,J, and Faber, D.S., An identifiable class of statoacoustic interneurons with bilateral projections in the goldfish medulla, Neuroscience, 5 (1980) 1287-1302. Zottoli, S.J., Hangen, D.H. and Faber, D.S., The axon reaction of the goldfish Mauthner cell and factors that influence its morphological variability, J. Comp. Neurol,, 230 (1984) 497-516.