0306-4522/90 $3.00 + 0.00 Pcrpmon Press plc 0 1990IBRO

Neuroscience Vol. 35, No. 2, pp. 221-248, 1990 Printed in Great Britain

RESPONSES OF LOCUS COERULEUS AND SUBCOERULEUS NEURONS TO SINUSOIDAL STIMULATION OF LABYRINTH RECEPTORS 0. POMPEXANO,*~D. MANZONI,* C. D. BARNES,: G. STAMPACCH~A*and P. D’ASCANIO+ *Dipartimento di Fisiologia e Biochimica, Universid di &a, Via S. Zeno 31, 56100 Fisa, Italy IDepartment of VCAPP, Washington State University, Pullman, Washington 99164-6520, U.S.A. Abstract-In precollicular decerebrate cats the electrical activity of 141 individual neurons located in the locus coeruleus-complex, i.e. in the dorsal (n = 41) and ventral parts (n = 67) as well as in the locus subcoeruleus (n = 33), was recorded during sinusoidal tilt about the longitudinal axis of the whole animal, leading to stimulation of labyrinth receptors. Some of these neurons showed physiological characteristics attributed to the norepinephrine-containing locus coeruleus neurons, namely, (i) a slow and regular resting discharge, and (ii) a typical biphasic response to fore- and hindpaw compression consisting of short impulse bursts followed by a silent period, which has been attributed to recurrent and/or lateral i~bition of the norepineph~ne~on~ining neurons. Furthermore, 16 out of the 141 neurons were activated antidromically by stimulation of the spinal cord at T,, and L, , thus being considered coeruleospinal or subcoeruleospinal neurons. A large number of tested neurons (80 out of 141, i.e. 56.7%) responded to animal rotation at the standard frequency of 0.15 Hz and at the peak amplitude of 10”. However, the proportion of responsive neurons was higher in the locus subcoeruleus (72.7%) and the dorsal locus coeruleus (61.0%) than in the ventral locus coeruleus (46.3%). A periodic modulation of firing rate of the units was observed during the sinusoidal stimulus. In particular, 45 out of the 80 units (i.e. 56.2%) were excited during side-up and depressed during side-down tilt (B-responses), whereas 20 of 80 units (i.e. 25.0%) showed the opposite behavior (a-responses). In both instances, the response peak occurred with an average phase lead of about + 18”, with respect to the extreme side-up or side-down position of the animal; however, the response gain (imp& per deg) was, on average, more than two-fold higher in the former than in the latter group. The remaining 15 units (i.e. 18.7%) showed a prominent phase shift of this response peak with respect to animal position. Similar results were obtained from the subpopulation of locus coeruleus-complex neurons which fired at a low rate (O.OS).

Following the criteria used in the present study (see Experimental Procedures) we were able to identify unit responses with gain values of 0.05 imp+ per deg or less. These low values were particularly obtained when: (i) the units fired at a frequency lower than 5 imp&; (ii) their discharge was so regular that a peak variation of the discharge frequency of about 0.5 imp./s or less for 10” of animal tilt could easily be detected in the averaged records; (iii) these responses represented a real modulation of the neuronal discharge rather than a chance variation in impulse density, since they were obtained by averaging a large number of sweeps (usually 18) covering two cycles of table movement; on the other hand, no responses were observed by averaging the neuronal discharge

LCa 67 (4 anti.) 31 (46.3%) (I anti.) 36 (53.7%) (3 anti.) 7.4 + IO.0 6.5 f 8.3 8.1 & 11.3 0.09 * 0.13 2.0 + 2.0, n = 30* 7 (22.6%) 14(45.1%) 10 (32.2%)

LCd

41 (2 anti.) 25 (61.0%) (1 anti.j 16 (39.0%) (1 anti.) 11.7+ 17.1 13.2 rt: 20.2 9.3 + 10.9 0.19 + 0.23 2.4 k I .9, n = 25

7 (28.0%) 17(68.0%)

I (4.0%) 4 (16.7%)

6 (25.0%) 14 (58.3%)

SC __-. 33 (10 anti.) 24 (72.7%) (9 anti.) 9(27.3%) (1 anti.) 11.4& 13.4 11.8 + 15.1 10.5 + 8.0 0.32 + 0.36 4.0 + 3.6, n = 23*

11(19.6%)

14 (25.0%) 31 (55.4%)

LCd + LCa ._._ ._. 108 (6 anti.) 56(51.9%) (2 anti.) 52 (48.1%) (4 anti.) 9.0 & 13.2 9.5 + 15.1 8.5 & 11.1 0.14*0.19 2.2 & 1.9, n = 55*

1.5(18.7%)

20 (25.0%) 45 (56.2%)

Total ...II_ 141 (16 anti.) 80 (56.7%) (I 1 anti.) 61 (43.3%) (5 anti.) 9.6 + 13.3 10.2 + 15.0 8.8 + 10.7 0.19 & 0.26 2.7 L 2.7 n = 78*

neurons to sinusoidal tilt around the longitudinal axis of the whole animal

Values are means + S.D. of responses recorded during roll tilt of the animal at standard parameters of 0.15 Hz, 2 10”. Figures in parentheses are percentages. Base frequency, mean tiring rate in imp-/s evaluated during animal tilt (in two units showing cut-off responses the base frequency was substituted by the resting discharge); gain, in imp./s per deg; sensitivity, in %/deg; phase angle, in degrees of phase lead (positive values) or phase lag (negative values) with respect to the side-down tilt. Number of units marked by asterisks are slightly lower than the corresponding number of responsive units due to two units showing cut-off responses. whose sensitivity could not be evaluated. Some of the tested units were activated antidromically from T,,-L, segments of the spinal cord (anti.).

. _._~. -.-No. of units Responsive (R) units Unresponsive (R) units Base frequency of R + R Base frequency of R-units Base frequency of R-units Gain of R-units Sensitivity of R-units Phase angle of R-units from +75” lead to -45” lag from $135”lead to - 105” lag from +75” to + 135” and from -45” to -105”

Table 1. Average response characteristics of locus coeruleus-complex

233

Responses of LC neurons to labyrinth input

The phase angle of the first harmonic of responses of all the LC-complex units to standard parameters of tilt was also evaluated. As shown in Fig. 3, two main groups of units were observed. The first group of units (45/80, i.e. 56.2%) which were excited during side-up tilt of the animal, showed a phase angle of responses that ranged from a lead of + 135” to a lag of - 105”, with an average phase lag of - 162.3 + 27.4, SD., deg, corresponding to an average lead of f17.7” with respect to the extreme side-up position @-responses). On the other hand, the second group of units (20/80, i.e. 25.0%), which were excited during side-down tilt of the animal, showed a phase angle of responses that varied from a lead of +75” to a lag of -45”, with an average phase lead of + 18.5 f 32.7, SD., deg (cc-responses). In addition to these two populations of units which showed an average phase lead of + 17.9 + 28.9, SD. deg with respect to the side-up or the side-down animal displacement (n = 65), there were 15 units (i.e. 18.7%) whose phase angle of responses was intermediate between the values reported above. A detailed analysis of the results indicated that the LCd and the SC units (49units), showed c(- or /3-responses (Fig. 3, dotted and striped columns, respectively), while the LCcr units (31 units) did not show any preferential dist~bution of their phase angle (Fig. 3, white columns). Figure 4 represents, in a polar diagram, both the gain and the phase relation of the responses. It appears that most of the LCd neurons (a) and the SC neurons (A) responded more or less in phase with

during an equal number of sweeps in the absence of stimuli; finalty, (iv) in contrast to the latter activity, the first harmonic responses obtained during animal tilt were highly coherent and the corresponding coefficient remained unmodified during successive tests. Within the whole population of responsive units, the gain of the first harmonic of responses of the LC-complex neurons to the labyrinth input, evaluated at the standard parameters, varied from 0.01-0.05 to 1.60 imp+ per deg, with a mean value of 0.19 + 0.26, SD., for the 80 responsive units. The corresponding mean value was higher for the SC and the LCd neurons than for the LCLYneurons, the difference between either LCD and LCd units, or LCct and SC units being statistically significant (t-test between the means, P < 0.05 and P < 0.01, respectively) (Table 1). The ~nsitivity of the first harmonic of responses of the same neurons varied from 0.29 to lOSO%/deg, with an average of 2.7 f 2.7, SD., %/deg for 78 out of 80 responsive units. Even in this case the mean response sensitivity was higher for the SC and the LCd neurons than for the LCcr neurons (Table 1). Finally, the average gain and sensitivity of responses of the SC neurons was higher than those obtained from the whole population of LC neurons (LCd f LCcr units) (t-test between the means, P < 0.01for differences in gain values of responses) (Table 1). Histograms of both gain and sensitivity display a unimodal long-tailed distribution (not shown).

NUMBER

OF UNITS LABYRINTH

LC complex

0 z

LCo( ;“,”

INPUT N=31 N=25 N=24

10

PHASE

LAG

(DEG)

PHASE LEAD

Fig. 3. Distribution of the phase angle of the first harmonic of responses of LC-complex neurons tested during roll tilt of the animal at 0.15 Hz, & IO”. Positive numbers in the abscissa indicate, in degrees, the phase lead, whereas negative numbers indicate the phase lag of responses with respect to the extreme side-down position of the animal, as indicated by 0”. Responses of LC-complex neurons to tilt, underlined by horizontal bars, have been used to evaluate the average phase angle of units excited during side-down (O”) or side-up displacement of the animal (180”). Most of the units were excited during side-up tilt (45/80 neurons) while a smaller proportion of units were excited during side-down tilt of the animal (20/80 neurons). The two populations of LCd-SC neurons greatly contributed to these opposite response patterns, while the responses recorded from the LCa did not show any preferential distribution of their phase angle.

234

0. POMPEIANO ef al LABYRINTHINPUT .LCd NEURONS N=25 oLCo( NEURONS N=31 rSC NE;ILJr\

\

\

.

/

-‘..-._,y 90' Pig. 4. Polar diagram showing the gain and the phase angle of the first harmonic of responses of LC-complex neurons to sinusoidal tilt at 0.15 Hz, + IO”. The 25 LCd neurons are indicated by filled circles, the 31 LCc( neurons by open circles, while the 24 SC neurons are indicated by closed triangles. The response gain of each unit is indicated by the distance of the corresponding symbol from the center of the diagram (see the scale along the vertical meridian); eight units had a gain higher than 0.5 imp+ per deg. The relative position of the symbol with respect to the 0” meridian indicates in degrees the phase lead (positive values) or the phase lag (negative values) of responses with respect to the extreme side-down position of the animal. The dashed lines outline the standard deviation of the phase angle of response of the two main populations of LC-complex units that showed a positional sensitivity, as indicated in Fig. 3. However, the two populations of LCd-SC units which were excited during side-down or side-up tilt of the animal showed, on average, a phase lead of f27.0 + 27.6, S.D., deg (n = 13) or a phase lag of - 163.9 If: 25.2, SD., deg (n = 31) with respect to the extreme side-down position of the animal, respectively; on the other hand, the LCa units did not show any preferential distribution of their phase angle of responses. (i.e. near 0 and 1800) and showed a larger gain than the LCa neurons (0) which did not display any preferential distribution of their phase angle. Moreover, the LCd and SC neurons, excited by side-up tilt showed, on average, a higher gain and sensitivity (0.28 & 0.26, S.D., imp./s per deg and 4.0 f 3.3, S.D., %/deg) than the LCd and SC units excited by side-down tilt (0.12 + 0.17, S.D., imp./s per deg and 1.5 i: 0.8, S.D., %fdeg). Table 2 indicates the relative patterns of responses of the LCd, the LCcl and the SC neurons as well as the average gain and sensitivity of the corresponding populations of units displaying different response patterns. In particular, the units excited by side-up tilt of the animal showed a larger than two-fold gain and sensitivity with respect to the units excited by side-down tilt (t-test between the means, P < 0.05 for differences in gain values of responses). No significant difference, however, was found between position

the average base frequencies of these two populations

of units. Gain and sensitivity values of the LCd neurons or the SC neurons showing /I- or cr-responses to tilt were, on average, higher than gain and sensitivity values of the LCc( neurons showing the corresponding patterns of response to animal tilt; however, the difference was statistically significant only between the SC and the LCa neurons showing /?-responses (t-test between the means, P < 0.01 for differences in gain values of responses). Similarly, the average gain and sensitivity of responses of the SC neurons showing p-responses were, on average, higher than those obtained from the whoie population of LC neurons (LCd + LCa units) showing the same response pattern (t-test between the means, P < 0.05 for differences in gain vahtes of responses). The relation between base frequency and the response gain or sensitivity of the LC-complex neurons,

1 62.1 0.25 0.4, n = 1 + 122.3

Other units No. of units Base frequency Gain Sensitivity Phase angle

17

SC

10 10.5 * 13.0 0.12+0.18 1.6 f 1.2, n = 10 - 22.9 f 74.2’*

7 (1 anti.) 5.9 & 5.7 0.05 & 0.03 1.5 + 1.1, n = 7 i2.7 & 37.7 4(1 anti.) 15.2 f 20.6 0.48 + 0.75 2.5 k 1.4, n = 4 + 2.3 + 92.7*+

6(1 anti.) 18.5 If: 24.7 0.f7 rfrO.18 1.3 + 0.7, ii = 6 +32.7 + 28.2

14 14 (7 anti.) 4.0 f 2.8 7.9 f 5.6 0.09 + 0.13 0.34 + 0.26 2.6 f 2.1, n = 13* 5.6 f 4.1, n = 13’ - 158.8 + 32.5 - 152.3 & 19.2

LCU

11 15.9 + 19.9 0.13 rf 0.18 IS& 1.2, n = 11

14 (I anti.) 6.0 & 6.5 0.06 f 0.05 1.6& 1.0, n = 14 -I-12.4 f 37.7

31 (1 anti.) 9.0 zfr15.7 0.17 rt. 0.22 2.1 + 2.3, n = 30+ - 166.9 & 29.6

LCd + LCu

15 (1 15.2 f 0.22 f 1.8 f - 18.5 f

anti.) 19.3 0.41 1.3, n = 15 75.0**

20 (2 anti.) 9.7 + 15.0 0.09 f 0.11 1.5 i: 0.9, n = 20 t 18.5 & 32.7

45 (8 anti.) 8.7 & 13.4 0.22 & 0.24 3.6 & 3.2, n = 43* - 162.3 + 27.4

Total

For the definition of units excited during side-up or side-down tilt of the animal, see the text. Number of units marked by one asterisk are slightly lower than the corresponding number of responsive units due to two units showing cut-off responses, whose sensitivity could not be evaluated. The phase angle of the unit responses marked by two asterisks was evaluated with respect to the extreme side-up or side-down animal displacement; otherwise, the phase angle was evaluated with respect to the extreme side-down animal tilt. The mean phase angle of the positional responses evaluated with respect to the extreme animal displacements, corresponded to + 11.1 f 27.0, SD., de8 for the 24 LCd units, $15.0 rf: 34.5, SD., deg for the 21 LCa units and +29.2 + 21.6, SD., deg for the 20 SC units (f 17.7 + 28.9, S.D., deg for the total ~p~ation of 65 LC-complex units).

7 6.1 + 7.7 0.07 f 0.06 1.6kO.9, n =7 + 22.1 * 28.2

Units excited by side-down tilt No. of units Base frequency Gain Sensitivity Phase angle

=

17 (1 anti.) 13.2 & 20.4 0.23 f 0.26 2.8 & 2.1, n - 173.5 * 26.0

LCd

Units excited by side-up tilt No. of units Base frequency Gain Sensitivity Phase angle

-

Table 2. Average response characteristics of locus coeruleus-complex neurons showing different patterns of response to sinusoidal tilt around the longitudinal axis of the whole animal

4

P

R ti 0

“0

236

0.

POMPEIANO et at

recorded during roll tilt of the animal at 0.15 Hz, + lo”, was also evaluated. If we consider the whole population of 80 units, the response gain increased by increasing the log of the base frequency (coefficient of correlation, r = 0.48, P < O.OOl), while the response sensitivity decreased (coefficient of correlation, r = 0.23, P < 0.05). Moreover, while the a-responsive units (n = 20) behave as an homogeneous population of neurons, characterized by a slight increase in gain

Al.0

LC complex LAEYRINTH

for increasing base frequency (Fig. .5A), the p-responsive units (n = 45) included apparently two populations of units, one (n = 30) behaving similarly to the a-units, the other (n = 15) being characterized by a more prominent increase in response gain for increasing levels of base frequency (Fig. 5B). These latter units were particularly located in the SC (9/14 units, i.e. 64.3%) than in the LCd and LCa (6/31. i.e. 19.3%).

INPUT

N=20 y=0.141xt0004 GO75

/

05

/ 5

lb

I

10

50

100

50

f 100

log BASE FREQUENCY lImp,/secl

B to.

l

N=45 y&2J2xtO.O56

r=O.41

*

3 =:

*

t

4

$ 0.5” E =: z 3

_

. O_

#

03

0.5

1.0

5

10

log BASE FREQUENCY (Imp./secl

Fig. 5. Relations between mean base frequency and gain values of the first harmonic response of LC-complex neurons which were excited during side-down (A) or side-up tilt of the animal (B). (A) Slight positive correlation between the log of the base frequency and the gain of 20 LC-complex neurons (seven LCd, seven LCa, six SC units) which were excited during ipsilateral side-down tilt of the animal at 0.15 Hz, f IO”, thus showing u-responses (r =0.75, P 0.05, not significant).

serotonergic,43*84acetylcholine neurons,34 as well as peptidergic neurons,” the neurochemical identity of the recorded neurons can be assessed only by combining intracellular recording experiments with histoimmunochemistry. It is of interest that some of the LC-complex units recorded in the present experiments were found to be antidromically activated from the ipsilateral cord at T,2-L,. This is in agreement with the anatomical finding that in the cat the axons of the LC-complex neurons projecting to the lumbar segments descend mainly ipsilaterally. 39 It should be noted, however, that in the present study the conduction velocity of

the antidromically activated CS neurons was, on average, higher than that reported in the rat.30 This faster conduction velocity, which fits with the results of previous experiments performed in cats,” should be compared with those obtained in the same animal species by Nakazato,” who reported conduction velocities from less than 1 to 33m/s for CS neurons antidromically activated by stimulating the high cervical cord at C,-C,. Although in this study the distance between stimulating and recording electrodes was very short (20-24 mm) and the repetition rate of the stimulus used for antidromic activation of the CS neurons was so high (2.5/s) to induce slowing

241

Responses of LC neurons to labyrinth input

LC complex

0.01

005

1.0

0.1

log GAIN OF VESTWJLAR

B

RESPONSES

(0.026 Hz)

NUMBER OF UNITS

N=19

Fig. 9. Relatjonships between response characteristic of 19 LC-complex units elicited by tow and high frequencies of animal tilt. Among the recorded units, seven of which (six SC and one LCd) antidromically identified as CS neurons, three were located in the LCd (0). three in the LCa (O), 13 in the SC (A) as shown in the diagram (A). (A) For each unit the gain of the first ha~onic of response elicited at the tilting frequency of 0.15 Hz (ordinate) is compared to that of the response elicited at the frequency of 0.026 Hz (abscissa). The line of slope I would apply if both response gains were equal. The inset illustrates the dist~bution of units according to the ratio between gain at 0.15 Hz and gain at 0.026Hz (mean corresponding to 1.49 f 0.78, SD.; paired r-test, P < 0.002). (B) Distribution of the LC-complex neurons according to the difference in the phase angle of the first harmonic of the response elicited at the tilting frequency of 0.15 Hz with respect to that elicited at the frequency of 0.026 Hz (A$). Coincidence in the phase angle of responses at the two different frequencies is indicated by 0”. Positive and negative numbers in the abscissa indicate phase lead and phase lag of the response at 0.15 Hz, with respect to that elicited at 0.026 Hz. A unimodal dist~bution of units can be observed, with a A$ co~s~nding on average to -7.7 + 51.9”, SD. (paired t-test, P > 0.05, not significant). Similar results were also obtained from the eight units tested during the whole frequency series (Fig. 8), where the ratio between gains, as defined above, corresponded on average to 1.32 + 0.39, SD., while the phase angie of the unit responses at 0.15 Hz showed on average a slight phase lead of +7.2 + 15.9”, S.D. deg with respect to the value obtained at 0.026 Hz.

242

0. POMPEIANO et al.

in the LC axon conduction velocity in the rat,s fewer than 10% of these neurons conducted impulses of 1 m/s or less. The assumption put forward by Nakazato,” i.e. that only unmyelinated fibers with conduction velocities of less than l/m were noradrenergic, would exclude the noradrenergic nature of most of our recorded units. This conclusion, however, is not supported by the fact that about 8.5% of all CS neurons in the cat contain NE.” On the other hand the hypothesis of this author that the fast conducting fibers are choline@ is contradicted because none of the cholinergic cells in the LC-complex of cat project to the lumbar cord.36 The demonstration that most of the CS neurons originated from the ventral part of the LC and the SC 8’,*2where medium-sized and large multipolar cells are’located,47*48explains why, in our experiments, the conduction velocity of the spinally projecting axons was higher than originally anticipated. Direct experiments, however, are required to find out whether CS neurons with myelinated axons are noradrenergic in origin, as the CS units provided with unmyelinated axons. It is worth mentioning that some of the LCcomplex units, which were not antidromically activated by spinal cord stimulation at T,,-L,, could still reach the spinal cord, if we assume that the corresponding descending fibers projected only to the cervical and/or thoracic segments. Other units, however, might project to supraspinal structures. If we consider now the physiological characteristics attributed by previous authors*,2*9.‘8.30 (cf. Ref. 23) to NE-containing LC-complex neurons, it appears that in addition to a typical positive-negative extracellular spike of long duration (1.5-2 ms), these units: (i) showed a slow and regular resting discharge, and (ii) responded to a noxious stimulus with a transient excitatory response followed by a reduced discharge, a finding which has been attributed to recurrent and/or lateral inhibition of the noradrenergic neurons (see the introduction). In the present experiments on cats, the mean firing rate of the LC-complex units (9.6 imp+) was higher than that reported in the same animal species by previous authors@ (cf. Ref. 231, a finding which can be attributed in part at least to different preparations used (decerebrate vs intact animals). Decerebration may in fact interrupt a descending pathway exerting a tonic inhibitory influence on the LC neurons. That the resulting increase in resting discharge of LC-complex neurons may actually contribute to the postural rigidity in the decerebrate cat@ (cf. Ref. 61), is shown by the fact that just the opposite effect, i.e. a postural atonia appeared in the same preparation after anatomical or functional inactivation of the LC neurons.‘4.63 In our experiments, interestingly, the results obtained from the whole population of LC-complex neurons were also found for the selected population of neurons (68 out

of 141) which fired at a frequency lower than 5 imp+. As to the typical pattern of response of the LCcomplex units to compression of the paws, we found that even after decerebration a large proportion of units responded with a burst of excitation followed by a period of quiescence, as reported previously in rats.9 The units which were simply inhibited by the stimulus probably escaped the direct excitatory input from the periphery, being submitted to recurrent and/or lateral inhibition of neighboring LC-complex neurons driven by the pinch stimulus. In conclusion, not all the LC-complex neurons recorded in our decerebrate cats displayed properties attributed to NE-containing LC neurons in rats; yet, some of the recorded neurons behave as if they were noradrenergic in nature. Response characteristics of locus coeruleus-complex neurons to labyrinth stimulation

The most striking finding of our experiments was that a large proportion of LC-complex neurons (X7%), particularly located in the caudal part of the LCd and the rostraf part of the SC, exhibited a periodic modulation of their firing rate during sinusoidal tilt of the animal at the standard parameters of 0.15 Hz, + 10”. The majority of these neurons, some of which were identified as CS neurons, responded preferentially to the extreme animal displacements, being either excited during side-up (56.2%) or during side-down (25.0%) animal tilt. Moreover, these populations of units showed an average phase lead of + 17.9” with respect to the side-up or the side-down animal displa~ment, a finding which may depend on stimulation of otolith receptors (see next section). The LC-complex units affected by tilt showed a positive correlation between response gain and base frequency. Since the base frequency closely corresponded to the mean firing rate of the units recorded in the animal at rest, it appears that the more prominent the self-inhibitory mechanism which controls the resting discharge of the NE-containing LC-complex neurons, the lower was the response gain of the same units to animal tilt. It is of interest that the predominant population of units, which showed ~-responds to tilt, was organized in two different groups, one displa~ng a slight positive correlation between gain and base frequency, as documented also for the a-responsive units, the other showing a more prominent positive correlation. This finding indicates that vestibular afferents activated during side-up tilt are not equally effective on the LC-complex neurons, due to differences either in connectivity or in intrinsic excitability of this neuronal population. In all the units tested, the response gain remained unmodified by increasing the peak amplitude of displacement from 5” to 20” (at the frequency of 0.15 HZ), an indication that the system was linear with respect to the amplitude used. Moreover, the activity of eight LC-complex neurons, some of which were

Responses of LC neurons to labyrinth input antidromically identified as projecting to the lumbosacral segments of the spinal cord, was recorded to increasing frequency of stimulation from 0.008 to 0.32 Hz at the peak amplitude of IO”, which raised the maximum angular acceleration from 0.025 to 41.7”/s2. Although a few units &owed a stability in their response gain other units displayed a slight, but progressive, increase in gain by increasing the frequency of tilt. Among all these neurons, four units showed a stability in the phase lead of the responses. This finding, which occurred in spite of the increase in peak angular acceleration above threshold for canal-induced responses of vestibular nuclei neurons, can be attributed to stimulation of otolith receptors. Experiments performed in cats’ and monkeys22 have in fact shown that units recorded from macular (utricular) afferents responded to sinusoidal tilt with small phase lead which remained almost constant upon increasing frequency of rotation, as described in the present units. The finding that the response gain of the four LC neurons either did not change or showed only a slight increase by increasing frequency of stimulation, suggested that both static and dynamic components of the response were due to the static and dynamic sensitivity of otolith receptors. In addition to these units, two LC units exhibited an increase in phase lead of their response that tended to be related to the velocity signal during increase in head angular acceleration; this increase in phase lead was also associated with an increase in response gain, as expected if both macular and canal inputs converged on the same unit. As to the last two units, which exhibited a decrease in phase lead as a function of frequency, their responses could still be attributed to stimulation of otolith receptors if we assume that some interaction had occurred between the afferent volleys originating from both labyrinths. It is also likely that neuronal processes (i.e. recurrent and/or lateral inhibition of the same or neighboring NE-containing neurons) became so effective at increasing frequency of tilt as to contribute to the reduction in phase lead of the responses. We cannot conclude this section without referring to the results of previous experiments showing that a large proportion of LC-complex neurons received a neck input.6 In particular, it appeared that 73 out of 99 (i.e. 73.7%) units, some of which antidromically identified as CS neurons, responded to neck rotation at 0.15 Hz, + 10”. The majority of these neurons were excited during side-down neck rotation (_54.8%), while a smaller proportion were excited during sideup neck rotation (24.7%). If we also consider the whole population of LC-complex neurons tested during sinusoidal stimulation of labyrinth and neck receptors, it appeared that 52 out of 90 units (57.8%) received a convergent input from both types of receptors.

243

Comparison between responses of coeruleospinal neur ons and lateral vestibulospinal neurons to labyrinth stimulation

The responses of LC-complex neurons, including the CS neurons, to animal tilt should be compared with those recorded in decerebrate cats from lateral vestibular nucleus (LVN) neurons.s If we consider in particular the activity of VS neurons antidromically activated by summation of the spinal cord at Tn-L, , thus projecting to the lumbosacral segments,” it appears that the proportion of these responsive neurons to roll tilt of the animal (58.9%) was almost comparable to that found in the present study for the LC-complex neurons (56.7%). Moreover, the responses of antidromically identified VS neurons to animal tilt showed, on average, a phase lead of +21.0” with respect to the extreme animal displacement. This lead, which closely corresponded to that of LVN neurons receiving the otolith input in canalplugged preparations,72 was comparable to the value reported in the previous section for the LC-complex neurons. A final comment concerns the predominant pattern of response of the VS neurons to animal tilt. The majority of the VS neurons activated by the extreme animal displacements (77.3%) were excited by sidedown tilt (ct-responses), a finding consistent with the pattern of utricular input.3~21~46~78 Moreover, most of these unit responses showed a larger gain than that of the units displaying the opposite response pattern.” In contrast to these findings, the majority of the LC-complex neurons activated by the extreme animal displacements (69.2%) were excited during side-up tilt of the animal (B-responses). Moreover, these units showed more than a two-fold larger gain with respect to the units excited by side-down tilt. In particular, these response characteristics involved the SC and the LCd neurons rather than those of the LCa. The responses of the LC and the SC neurons to stimulation of labyrinth receptors can be attributed to direct afferent projections of vestibular nuclear neurons to the ipsilateral LC~omplexi~~1*~z7~s8 (cf,, however, Ref. 4), provided that these vestibular neurons show a B-response to tilt. An alternative possibility, however, is that the macular input of one side is transmitted not only to the ipsilateral LVN, but also, via a crossed pathway, to the ventral aspect of the contralateral medullary reticular formation,s3 whose neurons project, at least in part, to the LCcomplex (cf. Ref. 4), thus exerting an excitatory influence on the corresponding neurons.ig~” One of the main channels transmitting the macular input of one side to the contralateral medullary reticular structures is represented by the lateral VS tract acting on neurons of the crossed spinoreticular pathway.53 In this way the laby~nthine volleys originating from macular receptors of one side during side-down tilt would be responsible not only for the a-responses of

244

0. POMPElAN el ai

VS neurons originating from the ipsilateral LVN, but also for the /?-responses of CS neurons originating from the contralateral LC-complex (Fig. 10). It is of interest that the spinoreticular pathway transmitting the macular input of one side to the contralateral medullary reticular formation may also involve reticulospinals3 as well as reticu~o~rebellar neurons.* A final comment concerns the comparison between &he c~ruleospina1 and ~estibu~ospina~ neuron responses to increasing frequency of sinusoidal tilting. Three types of LVN neurons were observed in decerebrate animals with the intact IabyrinthP’ (cf. also Manzoni et al., unpublished observations). In

LEFT SIDE

particular, some units exhibited the response characteristics which would occur if only macular afferents were assumed to converge onto the neurons, i.e. static gain as well as phase angle in spite of the increase in angular acceleration. Other units, however, were characterized by an increase in response gain associated either with an increase in phase lead or with an increase in phase lag of the response for increasing frequencies of tilt. The former effect depended on aRerent volleys originating from semicircular canals, since for high frequencies of tilt the responses became related to the velocity of animaf rotation. On the other hand, the latter effect was still observed after

RIGHT SIDE

Lf D----

IN

Fig. 10. Anatomical pathways which might contribute to the control of posture during the VS reflex. cSR, crossed spinoreticular pathway transmitting the macular input of one side to the contralateral mRF; LC, locus coeruleus from which the ~eruleospinal (CS) pathway originates; LVN, lateral vestibular nucleus of Deiters from which the excitatory lateral vestibuIospina1 (VS) pathway originates; mRF, ventral aspect of the medullary reticuiar fo~ation whose neurons project to the LC neurons; M, a-motoneurons innervating limb extensors; R, Renshaw cells coupled with the group of extensor motoneurons. Open and closed triangles represent excitatory and inhibitory synapses, respectively. IN indicates a sinusoidal labyrinth input resulting from + IO”, roll tilt of the animal. Each cell group has represented its temporal response to IN during a complete cycle of tilt started towards the left side, as indicated by the curved arrow. OUT indicates the temporal electromyogram discharge pattern of the triceps bra&ii to the same cycle of tilt. Upward deflections in the sine waves indicate excitation during left side-down animal di~la~ment. IN utilizes the VS rerlex arc comprised of primary vestibular afferents originating from labyrinth receptors, LVN neurons and M, which generate the electromyogram output (left side). This reflex arc is supplemented by the reflex arc comprised of: contralateral primary vestibular afferents (whose discharge, IN, would be 180” out of phase), cSR cells whose axons cross the midline to reach the mRF and the related LC (thus determining the 180” out of phase response of the LC neurons to tilt). Since the CS pathway inhibits the Renshaw cells, the 180” out of phase signal of the LC neurons would be reversed to in phase signal at the level of Renshaw cells, thus reducing the gain of motoneuronal response M, genemting the e?~tromyo~~ output (OUT). The CS projection ~~~jbut~ not only to inhibitory synapses on Renshaw cells but also to excitatory synapses on M; these opposite influences may depend upon different types of adrenoceptors (i.e. both a- and ~-receptors) in the cord (Ref. 37). Further details appear in the text.

Responses of LC neurons to labyrinth input

245

increased discharge of excitatory VS neurons (a-responses**~~‘*),but also on the reduced discharge of CS neurons (@-responses) projecting to the corresponding segments of the spinal cord (Fig. 9). In order to understand how the descending volleys originating from the LVN and LC interact at the spinal cord level, experiments were performed in To understand the role of the LC and the SC decerebrate cats, in which there is a continuous discharge of LC neurons (cf. Ref. 64). In these neurons in the postural adjustments occurring during the VS reflexes, we should consider that the lateral VS preparations, the Renshaw-cells anatomically linked with the GS motoneurons either fired at a low rate or neurons send axons to the ventral horn region of the spinal cord, where they terminate particularly in were silent at rest,” probably due to the tonic inlaminae VII-IX.73 These neurons actually exert a hibitory influence exerted by the CS neurons on Renshaw-cells activity. Interestingly, the same Rendirect excitatory influence on ipsilateral limb extensor shaw-cells either did not respond to labyrinth stimumotoneurons~,4g,84 and, through their recurrent collation or showed only a small amplitude a-response.@’ laterals, on the related Renshaw-cells. On the other The increased discharge of these Renshaw-cells durhand, the CS neurons send terminals to the whole ing side-down tilt could be attributed not only to an gray matter of the spinal cord, where they terminate increased discharge of VS neurons (e-responses), but also within the laminae VII and IX; these afferents also to a reduced activity of CS neurons elicited by are considered to be aminergic, as shown either the same direction of head rotation (/I-responses). by histofluorescent techniques28s42*57 or by immunoThis would lead to disinhibition of the Renshawcytochemical staining for dopamine-fi-hydroxylcells anatomically linked with these extensor asesi-83 or by anterograde autoradiographic tracing motoneurons,26 thus enhancing the functional coufollowing injections of labeled amino acids into the pling of the Renshaw-cells with their own extensor LC and the SC.33,55,83 Whatever the neurochemical motoneurons (Fig. 10). The opposite would occur nature of the projecting neurons may be, this pathduring side-up tilt. These findings explain why the way exerts an excitatory influence on ipsilateral limb extensor (and flexor) motoneurons,“.25 an effect amplitude of the electromyogram modulation, and thus the response gain of muscle extensors was quite which can be attributed, at least in part, to inhibition small in amplitude in the forelimbs or even absent in of spinal cord inhibitory interneurons. In particular, the hindlimbs.51,52 experiments performed in precollicular decerebrate This interpretation of the experimental findings is cats have shown that stimulation of the LC induced supported by the results of experiments performed in a removal of recurrent inhibition acting on both decerebrate cats, showing that anatomicali or funcextensor and flexor motoneurons.26 This finding was tional inactivation of the LC and SC neurons produe to an inhibitory LC control on Renshaw-cell duced by local injection of minute doses of the activity, as shown by experiments of unit recording.26 a,-adrenergic agonist clonidine,63 which acts on the The descending influences on recurrent inhibition and somatodendritic a2-adrenoceptors by enhancing reRenshaw-cells described above may originate from current and/or lateral inhibition of the corresponding noradrenergic LC and SC neurons, since iontophoretic application of NE also exerts an inhibitory NE-containing neurons (cf. Ref. 23), decreased the postural activity, but greatly increased the gain of the action on Renshaw-cells.7*‘7+80 VS reflexes. The LC-complex neurons may thus play The demonstration that the LC inhibits Renshawcells linked with both extensor and flexor motoa prominent role in the control of posture as well as neurons suggested that this structure could, by in the gain regulation of the VS reflexes. Since the LC increasing or decreasing their neuronal discharge, and SC neurons undergo spontaneous fluctuations in modify the functional coupling between Renshawtheir firing rate, leading to changes in postural activcells and the corresponding motoneurons, thus alterity during the sleep-waking cycle (cf. Ref. 32), they ing the input~utput relation of ~-motoneurons to a may intervene in order to adapt, to the animal state, given excitatory input (cf. Ref. 62). the response gain of limb extensors to labyrinth As pointed out in the previous section, a large stimulation. proportion of LC-complex neurons, including CS It is noteworthy that the LC could act in more than neurons projecting to the lumbosacral segments of one way to modify the response gain of extensor the spinal cord, responded to natural stimulation of muscles to labyrinth stimulation. In addition to the labyrinth receptors; moreover, the majority of these reported LC-induced inhibition of Renshaw-cell acneurons showed a response pattern which was just tivity, another means involves the direct LC enhanceopposite to that eiicited by the VS neurons projecting ment of extensor motoneuron firing efficacy’i~zsthus to the same segments of the spinal cord. Since during modifying the excitatory vestibulospinal drive to the VS reflexes there is a contraction of ipsilateral extensors. Another possibility depends on the LC limb extensors during side-down tilt of the animal,51~7’ inhibition of pontine reticular neurons and the reit is likely that this response depends not only on the lated medullary inhibitory RS neurons.‘4%63The latter

canal plugging, 72thus being attributed to stimulation of macular receptors. Unit responses similar to those described above were also found among the LC-complex neurons tested over the whole frequency range, as discussed in the previous section.

NSC 35,2--B

246

0. POMPEIANO rt uf

neurons may actually inhibit the extensor motoneurons by exciting the corresponding Renshaw-cells (cf. Ref. 67). Through this interplay among VS, CS and RS systems, it becomes possible for the animals to execute balanced postural adjustments in response to labyrinthine signals.

CONCLUSION

It should be mentioned that not all LC-complex neUt’OnS affected by tilt were activated ant~dromically as projecting to the lumbosacral segments of the

spinal cord. Since these unidentified least in part, project to supraspinal might be involved in the labyrinthine functions than posture.

units could, at structures, they control of other

Acknawle~ge~enrs-This work was supported by National Institute of Neurological and Communicative Disorders and Stroke Research Grant NS 07685-20, by grants from the Ministero della Pubblica Istruzione, and the Consiglio Nazionale delle Ricerche, Roma, Italy to 0. Pompeiano and by NSF grant INT 8516441 to C. D. Barnes. We thank E. Biagetti, Il. Corti and Mm M. Yaglini for their valuable technical assistance and Mrs C. pUcci for editinn and tvning .I Y the manuscript.

REFERENCES

I. Aghajanian G. K., Cedarbaum J. M. and Wang R. Y. (1977) Evidence for norepinephrine-mediated collateral inhibition of locus coeruleus neurons. Bruin Res. 136, 570-577. 2 Aghajanian G. K. and Van Der Maelen C. P. (1982) a~-Adrenoceptor-m~iated hype~olarization of locus coeruieus neurons: intracellular studies in uivo. Science 215, 1394-1396. 3. Anderson J. H., Blanks R. H. I. and Precht W. (1978) Response characteristics of semicircular canal and otolith systems in cat. I. Dynamic response of primary vestibular fibers, Expl Brain Res. 32, 491-507. 4. Aston-Jones G., Ennis M., Pieribone V. A., Nickelf W. T. and Shipley M. T. (1986) The brain nucleus locus coeruleus: restricted afferent control of a broad efferent network. Science 234, 7346737. 5. Aston-Jones G., Legal M. and Bloom F. E. (1980) Brain aminergic axons exhibit marked variability in conduction velocity. Brain Res. 195, 215-222. 6. Barnes C. D., Manzoni D., Pompeiano O., Stampacchia G. and d’Ascanio P. (1989) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal neck rotation in decerebrate cat. Neuroscience 31, 371-392. I. Biscoe T. J. and Curtis D. R. (1966) Noradrenaline and inhibition of Renshaw cells. Science 151, 1230-1231. 8. Boyle R. and Pompeiano 0. (1980) Reciprocal responses to sinusoidal tilt of neurons in Deiters’ nucleus and their dynamic characteristics, Arch. ital. Biol. 118, l-32. 9. Cedarbaum J. M. and Aghajanian G. K. (1978) Activation of locus coeruleus neurons by peripheral stimuli: modulation by a collateral inhibitory mechanism. L:if Sci. 23, 1383-1392. 10. Cedarbaum J. M. and Aghajanian G. K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. 1. camp. Neural. 178, 1-16. II. Chan J. Y. H., Fun8 S. J., Chan S. H. H. and Barnes C. D. (1986) Facilitation of lumbar monosynaptic reflexes by locus coeruleus in the rat. Brain Res. 369, 103-109. 12. Clavier R. M. (1979) Afferent projections to the self-stimulation regions of the dorsal pons, including the locus coeruleus, in the rat as demonstrated by the horseradish peroxidase technique. Brain Res. Bull. 4, 497-504. 13. Dahlstriim A. and Fuxe K. (1964) Evidence for the existence of monoamine containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acra physiol. scund. 62, Sup@. 232, l-55. 14. d’Ascanio P., Bettini E. and Pompeiano 0. (1985) Tonic inhibitory influences of locus coeruieus on the response gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. ita/. Eiol. 123, 69-100. IS. Denoth F., Magherini P. C., Pompeiano 0. and StanojeviC M. (1979) Responses of Purkinje cells of the cerebellar vermis to neck and macular vestibular inputs. Ppiigrs Arch. 381, 87-98. 16. Egan T. M. and North R. A. (1985) Acetylcholine acts on mr-muscarinic receptors to excite rat locus coeruleus neurons. Br. 1. Fhar~ac.

85, 733-735.

17. Engberg I. and Ryall R. W. (1966) The inhibitor

action of noradrenaline and other monoamines on spinal neurones.

J. Physiol., Land. 185, 298-322.

18. Ennis M. and Aston-Jones G. (1986) Evidence for self- and neighbor-mediated postactivation inhibition of locus coeruleus neurons. Brain Res. 374, 299--305. 19. Ennis M. and Aston-Jones G. (1986) A potent excitatory input to nucleus locus coeruleus from the ventrolateral medulla. Neurosci. Lett. 71, 299-305. 20. Ennis M. and Aston-Jones G. (1987) Two physiologically distinct populations of neurons in the ventrolateral medulla innervate the locus coeruleus. Brain Res. 425, 275-282. 21. Fern&ndez C. and Goldberg J. M. (1976) Physiology of peripheral neurons innervating otolith organs of the squirrel monkey. I. Response to static tilts and to long-duration centrifugal force. J. Neurophysiot. 39, 970-984. 22. Fernandez C. and Goldberg J. M. (1976) Physiology of peripheral neurons innervating otolith organs of the squirrel monkey. III. Response dynamics. J. Neurophysiol. 39, 996-1008. 23. Foote S. L., Bloom F. E. and Aston-Jones G. (1983) Nucleus locus coeruleus: new evidence of anatomical and physiological specificity. Physiol. Rev. 93, 844-914. Brain 24. Fung S. J. and Barnes C. D. (1981) Evidence of facilitatory coeruleospinal action in lumbar motoneurons ofcats. Res. 216, 299-3 11.

2s. Fmg S. J. and Barnes C. D. (1987) Membrane excitability changes in hindlimb motoneurons induced by stimulation of the locus coeruleus in cats. Bruin Res. 402, 230-242. 26. Fung S. J., Pompeiano 0. and Barnes C. D. (1987) Suppression of the recurrent inhibitory pathway in lumbar cord segments during locus coeruleus stimulation in cats. Brain Res. 402, 351-354. 27. Fung S. J., Reddy V. K., Bowker R. M. and Barnes C, D, (1987) Differential labeling of the vestibular complex following unilateral injections of horseradish peroxidase into the cat and rat locus coeruleus. Brain Res. 401, W-352.

Responses of LC neurons to labyrinth input

247

28. Fuxe K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. IV. Distribution of monoamine terminals in the central nervous system. ACM phyiol. stand. 64, Suppl. 247, 37-85. 29. Grillner S.. Honsro T. and Lund S. (1970) The vestib~ospinal tract: effects on alpha motoneurones in the lumbosacral spinal cord in the cat. Expl Brain Res. i0, 94-120. _ 30. Guyenet P. G. (1980) The coeruleospinal noradrenergic neurons: anatomical and electrophysiological studies in the rat. Brain Res. 189, 121-133. 31. Hancock M. B. and Fougerousse C. I. (1976) Spinal projections from the nucleus locus coeruleus and nudeus s&oeruleus in the cat and monkey as demonstrated by the retrograde transport of ho~radish peroxidase. Brain Res. BUN. 1, 229-234.

32.

33. 34. 35. 36. 37. 38.

39. 40. 41.

42.

Steriade M. (1986) Neuronal basis of behavioral state control. In Znfrinsic Regulatory System of the The Nervous System (ed. Bloom F. E.), Vol. IV, pp. 701-823, American Physiological Society, Bethesda, Maryland. Holstege G., Kuypers H. G. J. M. and Boer R. C. (1979) Anatomical evidence for direct brainstem projections to the somatic motoneuronal cell groups and anatomical preganglionic cell groups in cat spinal cord. Brain Res. 171, 329-333. Jones B. E. and Beaudet A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brainstem: a choline acetyltransferase and tyrosine hydroxylase immunohistochemical study. J. camp. Neural. 261, 15-32. Jones B. E. and Friedman L. (1983) Atlas of catechol~~ne perikarya varicosities and pathways in the brainstem of the cat. J. camp. Neural. 215, 382-396. Jones B. E., Care M. and Beaudet A. (1986) Retrograde labeling of neurons in the brain stem following injections of [‘HIcholine into the rat spinal cord. Neuroscience 18, 907-916. Jones D. J. and McKenna L. F. (1980) Norepinephrine-stimulated cyclic AMP formation in rat spinal cord. J. ~euroc~em. 34, 467-469. Jones G., Segal M., Foote S. L. and Bloom F. E. (1979) Locus coeruleus neurons in freely moving rats exhibit pronounced alterations of discharge rate during sensory stimulation and stages of the sleepwake cycle. In Catecholumines: Basic and Clinic Frontiers (eds Usdin E., Kopin I. J. and Barchas J.), pp. 643-645. Pergamon Press, New York. Krausz M. (1986) Distribution of neurons in the lateral pontine te~entum projecting to thoracic, lumbar and sacral spinal segments in the cat. J. Hirnfbrsch. 27, 485-493. Kubin L., Magherini P. C., Manzoni D. and Pompeiano 0. (1980) Responses of lateral reticular neurons to sinusoidal stimulation of labyrinth receptors in decerebrate cat. J. Neurophysiol. 44, 922-936. Kuypers H. G. J. M. and Huisman A. M. (1982) The new anatomy of the descending brain pathways. In Brain Stem Control of Spinal ~eehanjsms (eds Sjiilund B. and Bj6rklund A.), pp. 29-54, El~vier/North-Holland, Amsterdam. Lackner K. J. (1980) Mapping of monoamine neurones and fibers in the cat lower brainstem and spinal cord. Anar. H&son J. A. and Brain, Section I,

Embryol. 161, 169-195.

43. Lai Y.-Y. and Barnes C. D. (1985) A spinal projection of serotonergic neurons to the locus coeruleus in the cat. Ne~rosc~. Left. 58, 159-164. 44. Leger L., Charney Y., Chayvialle J. A., Berod A., Dray F., Pujol J. F., Jouvet M. and Dubois P. M. (1983) Localization of substance P- and enkephalin-like immunoreactivity in relation to catecholamine-containing cell bodies in the cat dorsolateral pontine tegmentum: an immunofluorescence study. Neuroscience 8, 525-546. 45. Leger L. and Hernandez-Nicaise M. L. (1980) The cat locus coeruleus, light and electron microscopic study of the neuronal somata. Anut. Embryol. 159, 181-189. 46. Loe P. R., Tomko D. L. and Werner G. (1973) The neural signal of angular head position in primary afferent vestibular nerve axons. J. Physiol., Lond. 230, 29-50. 47. Loughlin S. E., Foote S. L. and Bloom F. E. (1986) Efferent projections of nucleus locus coeruleus: topographic organization of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience 18, 291-306. 48. Loughlin S. E., Foote S. L. and Grzanna R. (1986) Efferent projections of nucleus locus coeruleus: mo~holo~c subpopulations have different efferent targets. Neuroscience 18, 307-319. 49. Lund S. and Pompeiano 0. (1968) Monosynaptic excitation of alpha-motoneurons from supraspinal structures in the cat. Acta physiol. stand. 73, I-21. 50. Manzoni D., Marchand A. R., Pompeiano 0. and Stampacchia G. (1987) Coding of velocity signal by lumbar projecting vestibulospinal neurons in decerebrate cats. Neuroscience 22, Suppi S739. 51. Manzoni D., Pompeiano O., Srivastava U. C. and Stampacchia G. (1983) Responses of forelimb extensors to sinusoidal stimulation of macular labyrinth and neck receptors. Archs ital. Bioi. 121, 205-214. 52. Manzoni D., Pompeiano O., Srivastava U. C. and Stampacchia G. (1984) Gain regulation of vestibular reflexes in foreand hindlimb muscles evoked by roll tilt. Boll. Sot. ital. Biol. sper. 60, Suppl. 3, 9-10. 53. Manzoni D., Pompeiano O., Stampacchia G. and Srivastava U. C. (1983) Responses of medullary reticulospinal neurons to sinusoidal stimulation of labyrinth receptors in decerebrate cat. J. Ne~rophysio~. 50, 1059-1079. 54. Marchand A. R., Manzoni D., Pompeiano 0. and Stampacchia G. (1987) Effects of stimulation of vestibular and neck receptors on Deiters neurons projecting to the lumbosacral cord. Pfugers Arch. 409, 13-23. 55. McBride R. L. and Sutin J. (1976) Projections of the locus coeruleus and adjacent pontine tegmentum in the cat. J. camp. Neurol. 165, 265-284.

56. Miachon S., Berod A., Leger L., Chat M., Hartman B. and Pujol J. F. (1984) Identification of catecholamine cell bodies in the eons and pons-mesencephalon junction of the cat brain, using tyrosine hydroxylase and dopamine-/l-hydroxylase immunohistochemistry. Brain Res. 305, 369-374. 57. Mizukawa K. (1980) The segmental detailed topographical distribution of monoaminergic terminals and their pathways in the spinal cord of the cat. Anat. Anz. 147, 125-144. 58 Morgane P. J. and Jacobs M. S. (1979) Raphe projections to the locus coeruleus in the rat, Brain Rc.s. Bull. 4,519-534. 59. Nakazato T. (1987) LOCUS coeruleus neurons projecting to the forebrain and the spinal cord in the cat. Neuroscience 23, 529-538.

60. P&as D. and Parent A. (1978) Atlas of the distribution of monoamine containing nerve cell bodies in the brain stem of the cat. J. camp. Neurol. X79, 699-718.

248

0. POMPEIANO et al.

61. Pompeiano 0. (l980) Choline& activation of reticular and vestibular mechanisms controlling posture and eye movements. In IBRO Monograph Series, Vol. 6, The Reticular Formation Revisited (eds Hobson J. A. and Brazier M. A. B.), pp. 473-512. Raven Press, New York. 62. Pompeiano 0. (1984) Recurrent inhibition. In Handbook of the Spinal Cord (ed. Davidoff R. A.), Vols 2 and 3, pp. 461-557. Marcel1 Dekker, New York. 63. Pompeiano O., d’Ascanio P., Horn E. and Stampacchia G. (1987) Effects of local injection of the a,-adrenergic agonist clonidine in the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Archs ital. Biol. 125, 225-269. 64. Pompeiano 0. and Hoshino K. (1976) Tonic inhibition of dorsal pontine neurons during the postural atonia produced by an anticholinester~ in the decerebrate cat. Archs ital. Biol. 114, 310-340. 65. Pompeiano O., Manzoni D., Barnes C. D., Stampacchia G. and d’Ascanio P. (1988) ~byrinthine influences on locus coeruleus neurons. Acta Otolaryngol., Stockh. 105, 576-58 1. 66. Pompeiano O., Wand P. and Srivastava U. C. (1985) Responses of Renshaw cells coupled with hindlimb extensor motoneurons to sinusoidal stimulation of labyrinth receptors in the decerebrate cat. Pftigers Arch. 493, 245-257. 67. Pompeiano O., Wand P. and Srivastava U. C. (1985) influence of Renshaw celis on the gain of hindlimb extensor muscles to sinusoidal labyrinth stimulation. P’iigers Arch. 494, 107-l 18. 68. Reddy V. K., Fung S. J., Zhuo H. and Barnes C. D. (1989) Spinally projecting noradrenergic neurons of the dorsotateral pontine tegmentum: a combined immunocytochemical and retrograde labeling study. Brain. Res. 491, 144-149. 69. Sakai K. (1980) Some anatomical and physiological properties of pontomesencepbalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep. In IBRO Monograph Series, Vol. 6, The Reiicular Formation R~~sited (eds Hobson J. A. and Brazier M. A. B.), pp. 427-447. Raven Press, New York. 70. Sakai K., Sastre J.-P., Salvert D., Touret M., Tohyama M. and Jouvet M. (1979) Tegmentoreticular projections with special reference to the muscular atonia during paradoxical sleep in the cat: an HRP study. Brain Res. 176, 233-254. 71, Schor R. H. and Miller A. D. (1981) Vestibular reflexes in neck and forelimb muscles evoked by roll tilt. J. Neurophysiol. 46, 167-178. 72. Schor R. H. and Miller A. D. (1982) Relationship of cat vestibular neurons to otolith-spinal retlexes. Expf Brain Res.

47, 137-144. 73. Shinoda Y., Ohgaki T. and Futami T. (1986) The morphology of single lateral vestibulospinal tract axons in the lower cervical cord ofthe cat. .I. camp. Neurol. 249, 226-241. _74. Stamoacchia G.. Barnes C. D.. d’Ascanio P. and Pomneiano 0. (1987) Effects of microiniection of a cholinergic agonist into ;he locus cberuleus on the gain of vestibulospi~al reflexes-in decerebrate cats. Archs itaf. Bioi. 125, 1%7-l-38. 75. Svensson T. H., Bunney B. S. and Aghajanian G. K. (1975) Inhibition of both noradrenergic and serotonergic neurons in brain by the a-adrenergic agonist clonidine. Brain Res. 92, 291-306. 76. Tohyama M., Sakai K., Touret M., Salvert D. and Jouvet M. (1979) Spinal projection from the lower brainstem in the cat as demonstrated by the horseradish peroxidase technique. II. Projection from the dorsolateral pontine tegmentum and raphe nuclei. Brain Res. 176, 215-231. 77. Van Dongen P. A. M. (1981) The central noradrenergic transmission and the locus coeruleus: a review of the data, and their implications for neurotransmission and neuromodulation. Prog. Neurobiol. 16, 117-143. 78. Vidal J., Jeannerod M., Lifschitz W., Levitan H., Rosenberg J. and Segundo J. P. (1971) Static and dynamic properties of gravity-sensitive receptors in the cat vestibular system. Kybernetik 9, 205215. 79. Watabe K. and Satoh T. (1979) M~hanism underlying prolonged inhi~tion of rat locus coeruleus neurons following anti- and orthodromic activation. Bruin Res. 165, 343--347. 80. Weight F. F. and Salmoiraghi G. C. (1966) Adrenergic responses of Renshaw cells. J. Pharmac. exp. Ther. H&391-397. 81. Westlund K. N., Bowker R. M., Ziegler M. G. and Coulter J. D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res. 263, 15-31. 82. Westlund K. N., Bowker R. M., Ziegler M. G. and Coutter J. D. (1984) Origins and terminations of descending noradrenergic projections to the spinal cord of monkey. Brain Res. 292, 1-16. 83. Westlund K. N. and Coulter J. D. (1980) Descending projections of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: axonal transport studies and dopamine-fl-hydroxylase immunocytochemistry. Brain Res. Rev. 2, 235.--264.

84. Wikiund L., Leger L. and Persson M. (1981) Monoamine cell dist~bution in the cat brain stem. A fluorescence histochemical study with quantification of indolaminergic and locus coeruleus cell groups. J. camp. Neurol. 203, 613-647. 85. Wilson V. J., Kato M., Thomas R. C. and Peterson B. W. (1966) Excitation of lateral vestibular neurons by peripheral arerent input. J. Neurophysiol. 29, 5088529. 86. Wilson V. J, and Yoshida M. (1969) Comparison of effects of simulation of Deiters’ nucleus and medial longitudinal fasciculus on neck, forelimb and hindlimb motoneurons. J. N~rophysio~. 32, 743-758. 87. Wolstencroft J. H. (1984) Reticulospinal neurons. J. Physioi., Lond. 174, 91-108. (Accepted 15 November 1989)