ACTA NEUROBIOL. EXP. 1915. 35: 821-840

Memorial Paper i n Hanor of Jerzy Konorski

THE BIOLOGICAL SIGNIFICANCE OF PGO SPIKES IN THE SLEEPING CAT

Adrian R. MORRISON and Robert M. BOWKER

Laboratories of Anatomy, School of Veterinary Medicine and Institute of Neurological Sciences, University of Pennsylvania

Philadelphia, USA

Abstract. Large-amplitude waves recorded in the pontine tegmen- tum, lateral geniculate body and visual cortex herald the onset and con- tinue throughout paradoxical sleep. The role of these ponto-geniculo-oc- cipital (PGO) waves, or spikes, has puzzled researchers since their disco- very. This paper reports experiments in cats which have demonstrated that PGO spikes are essentially an epiphenomenon, an electrical sign of the activation of a "startle network" by the neural turmoil of paradoxi- cal sleep. Internal stimulation provided by the bursts of neural activity which characterize paradoxical sleep produces PGO spikes in the lateral geniculate body which are identical in appearance to those elicited dur- ing synchronized and paradoxical sleep by 1,55 Hz tone bursts or taps on the cage in normal cats. Cerebellar lesions result in behavioral responses to the intrinsic startles during synchronized sleep in the form of extensor or flexor jerks of the forelimbs. The jerks occur in con- junction with each PGO spike. Identical movements can be induced in the same cats in wakefulness by such startling stimuli as dropping the cat or hissing with an aerosol can. Lesions involving the auditory-visual area permit cats to be stimulated by sound in synchronized sleep with- out arousal, but anterior lobe lesions produce an easily aroused animal. We postulate that the phenomena observed following cerebellar lesions are the result of alteration in the control of serotonergic neurons of the pontine raphe nuclei.

822 A. R. MORRISON AND R. M. BOWKER

Since the discovery of paradoxical sleep, during which the electro- encephalogram resembles that of wakefulness and bursts of rapid eye movements occur (2), a variety of new and interesting phenomena have been discovered to be a part of what is now commonly called rapid-eye-. movement, or paradoxical, sleep. Workers soon revealed that although an animal (a majo'rity of studies having been performed on the cat) lay paralyzed with atonic muscles due to postsynaptic inhibition of the spinal motor neurons (24, 40), the central nervous system itself churned with activity. Neurons in a variety of areas, such as the lateral geniculate body (4), vestibular nuclei (6), motor cortex (20), and pontine tegmentum (29), fire in unusually high-frequency bursts, much as if they were in- duced by "electrical stimulation or strychninization" (20). The only out- ward manifestations of this turbulence are the twitchings of limb, facial and extraocular muscles which periodically interrupt the otherwise still, paralyzed organism. Among the phenomena associated with paradoxical

EMG 0.2

2 s e c

Fig. 1. Suppressions (arrows) of dorsal cervical muscle tone in the transition period before paradoxical sleep in a normal cat. (Reprinted from Fig. 4 of Morrison and

Bowker (39) by permission of Gustav Fischer Verlag.)

PGO SPIKES IN THE SLEEPING CAT 823

sleep, one of the most-studied has been the large-amplitude waves oc- curring 30 sec or so prior to the onset and during paradoxical sleep, in the pontine reticular formation, lateral geniculate nucleus (LGN) and visual cortex of the cat (Fig. 1) (5, 11, 31). Because of the recording sites these waves have been termed ponto-geniculo-occipital waves or spikes (PGO spikes). A series of PGO spikes appears during synchroni- zed sleep and heralds the onset of each episode of paradoxical sleep, and then spikes continue throughout the episode. They appear in clusters whenever rapid eye movements occur.

Both the pathways and neurotransmitters involved in the produc- tion of PGO spikes have been intensively investigated (31). Abundant work (5, 34, 44) indicates that their site of origin is the pontine reticular formation and that activity in this region then leads to spikes in LGN and the visual cortex. Their occurrence in these areas is independent of retinal input (10, 41). The function of PGO spikes during sleep, how- ever, remains a mystery. In this paper we shall discuss experiments

Fig. 2. An unfolded dorsal view of the cerebellum illustrates the longitudinal cor- ticonuclear zones of the medial vermal zone (stars) and more lateral paravermal zone (circles). Anterior lobe is subdivided into lobules I-V. FL, fore limb area;

A-V, auditory-visual area; Prim. Fiss., primary fissure.

824 A. R. MORRISON AND R. M. BOWKER

which have led us to the conclusion that PGO spikes are a sign of the neural turmoil of paradoxical sleep, nothing more. We shall argue that they are a neural sign of the operation of a startle response to the inter- nal stimulation provided by altered neuronal activity which appears prior to and during paradoxical sleep. Our conclusions concerning the biologi- cal significance of PGO spikes were reached as a result of experiments designed to investigate the role of the cerebellum in the control of motor activity during sleep. In addition to showing that the cerebellum parti- cipates in the modulation of motor quiescence during synchronized sleep and the smooth transition into the atonia of paradoxical sleep (39), our results now suggest that the cerebellum probably participates in the regulation of serotonergic mechanisms involved in the occurrence of PGO spikes. We shall proceed by first describing our earlier observa- tions made on cerebellar-lesioned cats and then focusing on later fin- dings which support our hypothesis.

Lesions have been placed by aspiration in the vermal or paraver- ma1 zones of the cerebellar cortex (Fig. 2) or by electrocoagulation with- in the fastigial nuclei to which the vermis and parts of the paravermis project (19, 50). Twenty-six cats were implanted with standard stainless steel screw electrodes in the skull for recording the electrocephalogram (EEG) and in the frontal sinus to monitor the electro-oculogram; with stainless steel wires in the dorsal cervical muscles and forelimb muscu- lature for electromyographic (EMG) recording; and with bipolar elec- trodes in the lateral geniculate body in order to record PGO spikes. The latter were aimed for the point of the greatest oscilloscopic response to a penlight f lashd in the eye, usually a t coordinates anterior 6, lateral 11, and vertical 2.5-3.0 (42). The cats were housed and recorded from while in a Lehigh Valley behavioral chamber. They could be observed through a one-way mirror or photographed and videotaped through plain glass. All cats were recorded from for a t least 5 hours during the middle of the day pre- and post-operatively when sleep is a t a maximum for laboratory cats (47). In addition, neurological examinations were re- gularly performed before and after the operations. These consisted of tests of general locomotion and posture, tactile placing, hopping reflexes, muscle tone, vestibular placing, responsiveness to visual and auditory stimuli, vestibular responses to rotation, alterations in muscle tone and deep tendon reflexes (46). Lesions and lateral geniculate electrode place- ments were examined histologically in sections prepared from saline and formalin-perfused brains embedded in celloidin. Cresyl violet and Weil stains were employed.

In our first experiments (39) we observed that bilateral lesions in- volving the vermal and paravermal cortices (Fig. 3) to some degree re-

PGO SPIKES IN THE SLEEPING CAT

Fig. 3. RB10: Cerebellar cortical lesion involving both vermal and paravermal zones in a cat showing ipsilateral flexor jerks of the right forelimb in synchroniz- ed sleep. RB13: Cerebellar cortical lesion with destruction of the majority of both fastigial and right interpositus nuclei. This cat exhibited signs of severe cerebellar .damage in wakefulness and strong extensor jerks of the right forelimb in syn-

chronized sleep.

sulted, during synchronized sleep, in repetitive extensor or flexor jerks 'of primarily the forelimb;, the elbow being the joint most prominently involved in the movements (Fig. 4). Bilateral electrolytic lesions placed in the fastigial nuclei, to which the vermis projects (19, 50), resulted in extensor thrusts of the forelimbs. These movements were most vigorous and occurred with the highest frequency (more than 30Isec) during the first few postoperative days. In cases with large lesions jerks continued with a gradually reducing frequency to 1 or 2 per minute in the ensuing weeks. In the case of large bilateral vermal or fastigial lesions the ex- tensor thrusts were strong enough to raise the cat o~ne or two inches from the floor. In fact, when first noticed after a fastigial lesion made as a control for underlying pontine tegmental lesions (28), our impres- sion was that the animal was being shocked electrically. The jerks did not occur in isolation from other phenomena associated with synchro- nized sleep. Curiously, the jerks were always associated with PGO spikes on a 1 : 1 basis (Fig. 4) and less obviously with large spindles (Fig. 5). If the lesions were large, the jerks continued into the early phase of paradoxical sleep in conjunction with a prolonged presence of cervical muscle tone.

A. R. MORRISON AND R. M. BOWKER

A EOG ----- <-?------- - -- 7------

EEG - A

w w w I I

EEG-M ~,W'NI~W~**

NECK

R B I 3 - 2 N D DAY POST-OP Fig. 4. A continuous record, showing the transition period to paradoxical sleep, in a cat with a large cerebellar lesion. (See RBI3 in Fig. 3). The frequency of the jerks can be seen as artifacts in the neck muscle EMG tracing. They were accom- panied by extensor thrusts of the right forelimb primarily. The prolonged transi- tion period was sometimes observed in the first few postoperative days. EEG-A and EEG-M: Electroencephalograms recorded from the anterior suprasylvian and precruciate gyri. EOG, electro-oculogram; LGN, lateral geniculate nucleus; Neck, EMG from dorsal cervical muscles. (Reprinted from Fig. 3 of Morrison and Bowker

(39) by permission of Gustav Fischer Verlag.)

LGN OJO I

I 5 sec

R B I 3 - 2ND DAY POST-OP Fig. 5. Demonstration of extensor jerks associated with spindle bursts on the EEG

seen as artifacts in the dorsal cervical muscle EMG (See RB13 in Fig. 3.).

PGO SPIKES IN THE SLEEPING CAT 82 7

The significance of these movements remained a puzzle until we observed that they paralleled the forelimb movements elicited in the vestibular placing test in character, time of onset, and gradual recovery. The vestibular placing test consists of quickly dropping a cat several centimeters while cradling the head and body in ones arms and allowing the forelimbs to hang freely. When we apply the test, the normal cat exhibits a varying degree of toe fanning, only the lateral togs moving with the slightest drop. The behavioral response occurs due to stimula- tion of the labyrinthine and visual receptors (16, and personal observa- tions) and, quite possibly, other proprio- and exteroceptcrrs. It should be remembered for the purpose of the general argument that dropping a cat suddenly is startling and that the response is exaggerated when the cat is excited. Following cerebellar lesions the response is altered in one of two ways. With damage to the vermis or underlying fastigial nuclei, one observes an explosive extensor thrust of the forelimbs and extreme fanning of the toes. In cases where the lesion involves paraver- ma1 cortex to spme degree, the ipsilateral forelimb will flex at the el- bow while the toes fan (13, 53). In both sleep and wakefulness the jerks were less obvious on the first postoperative day but became more vigo- rous and exaggerated during the second postoperative day. Recovery began during the latter part of the week.

Thus, it seemed evident that a common mechanism must underly the release of jerks in sleep and wakefulness, stimulation of the vestibular apparatus. In wakefulness during the vestibular placing test a sudden downward acceleration elicits the forelimb responses. What event might simulate such an acceleration in synchronized sleep? Thus far no chan- ges in transmission through the vestibular system have been detected at any time other than during rapid-eye-movement bursts (35). However, Pivik and Dement (43) reported that in some cats they could detect brief suppressions of activity recorded from the dorsal cervical muscles in conjunction with PGO spikes. We later confirmed this observation (39). In Fig. 1 one can observe complete suppressions of EMG activity in association with PGO spike appearance. We thought it reasonable to assume that such abrupt cessations of antigravity muscle tone would be sufficient to produce a minimal head-drop, which would excite the vestibular receptors in an animal primed for an exaggerated response as a result of a cerebellar lesion.

Our explanation of the jerks seen in connection with large sleep spindles followed a similar argument. Gassel and Pompeiano (26) have reported that during sleep spindles gamma motor neuron activity is brief- ly depressed, resulting in a slight suppression of spinal myotatic reflexes.

,828 A. R. MORRISON AND R. M. BOWKER

This could also simulate a slight postural collapse which would excite vestibular receptors.

Additional experiments, however, indicate that the jerks appearing in synchronized sleep were more than a response to purely vestibular stimulation (7, 8). If the lesion included lobules VI and VII of the ver- mis, the auditory-visual area of the cerebellum (45), spontaneous jerks associated with PGO spikes and sleep spindles would still occur in sleep provided that the more rostra1 forelimb areas of lobule V of the vermis or paravermis were also involved (Fig. 2 and 6). However, jerks and

Fig. 6. Cortical lesion involving bilateral forelimb areas and auditory and visual areas. The cat exhibited decreased responsiveness to auditory and visual stimula- tion with bilateral forelimb hyperflexion in wakefulness. Similar flexion jerks

appeared spontaneously in synchronized sleep.

PGO spikes could also be induced in such cats by the slightest sound of a pencil tapping on the side of the chamber or by 1,500 Hz pure tone bursts of 100 msec duration (The output of the tone generator was display- ed on an oscilloscope so that relative measurements of sound intensity, i.e., amplitue of the sine wave, could be made.) The jerks were associat-

PGO SPIKES IN THE SLEEPING CAT 829

led with the PGO spikes on a 1 : 1 basis (Fig. 7), but the cats were not awa- kened by these stimuli. Cats varied in how rapidly the PGO spikes ha- bituated, from after a few stimuli to the maximum number of stimuli administered, 14, (Fig. 7). During wakefulness cats with such lesions showed a decreased responsiveness to auditory stimuli, such as handclaps,

LGN 0.04 . . . . . . . . . . . . . .

EMG 0.04

Fig. 7. Illustration of the effectiveness of repetitive auditory stimuli in inducing forelimb jerks and PGO spikes. Following cerebellar damage (see Fig. 6), 14 con- secutive light taps on the cage side during synchronized sleep induced forelimb jerks and PGO spikes 1 for 1. Note that after 2-3 taps cortical desynchronization occurred, but the cat showed no behavioral awakening and continued in synchro-

nized sleep upon cessation of the stimuli. Time calibration, 1 sec.

and tended to stumble into objects jn their path, which confirms earlier observations of Chambers and Sprague (13). Cats with more rostra1 ce- rebellar lesions, on the other hand, were exceedingly easily aroused, much more so than were normal cats, and exhibited no auditory or visual deficits.

The contrasting behaviors of these two groups of cats with cerebel- lar lesions provided us with the first clue as to the real significance of PGO spikes. We suggested (7) that the elevated thresholds of arousal in cats with auditory-visual area lesions permitted us more easily to intro-

830 A. R . MORRISON AND R. M. BOWKER

duce stimuli which would elicit a startle reflex in a sleeping animal without awakening it. Because of the poor motor control resulting from the lesion in the forelimb area tho most obvious response then became a jerk of affected limbs. The startle reflex is, of course, a primitive re- flex, not requiring a high level of awareness. It is easily produced by auditory stimuli in neonatal animals (23) and adult cats with transec- tions of the brainstem at the mesencephalic level (3).

Having elicited PGO spikes and limb jerks in cats with cerebellar lesions, we attempted to dissociate the PGO spikes from the limb jerks by placing lesions in the cerebellar auditory-visual area alone. In a cat with such a lesion (Fig. 8) PGO spikes followed the sound stimuli 1 : 1,

Fig. 8. Cortical lesion involving auditory-visual area in lobules VI, VII A and B, and VIII A and B with no damage to forelimb areas.

but jerks were not evident (Fig. 9) (7). If sound intensity were low enough. no change from the EEG pattern of synchronized sleep occurred; how- ever, an eye blink could occur. This observation indicated that jerks d e p end upon cerebellar damage and are peripheral to the mechanism generat- ing PGO spikes.

PGO SPIKES IN THE SLEEPLNG CAT

A B A U D

1 . . . - -

EOG 1.Omv

LGN 0.1

EEG . 0.1

Fig. 9. A: Spontaneous PGO spikes at the transition from synchronized to para- doxical sleep in a cat with cerebellar damage to the auditory-visual area similar

to that shown in Fig. 8. B: Sound-induced PGO spikes in synchronized sleep. Note similar configurations of PGO spikes in A and B although paper speed differed. Ilatency to spike appe-

arance 80-120 msec. AUD, Auditory stimulus marker; Time calibrations, 1 sec.

But is cerebellar damage necessary for the production of PGO spikes by auditory stimulation? We noted that PGO spikes had been induced in intact, reserpinized cats in the pons, LGN, and the visual cortex by auditory stimuli (12, 30), but no reports of their induction in normal animals had appeared. That cerebellar damage or reserpinization is not necessary is illustrated in Fig. 10, which demonstrates PGO spikes elicited by low-intensity, pure tone bursts in a normal cat in synchroniz- ed sleep (8). These spikes were identical in appearance to those observed spontaneously at the transition period and during paradoxical sleep in the same cat (Fig. 11). An important behavioral observation was that a generalized body twitch (startle reaction) accompanied the auditory sti- mulus and the PGO spike, if the sound were loud enough. On the other hand, by decreasing the intensity of the stimulus it was possible to eli- cit a PGO spike with no observable body movement or EEG change. Fi- gure 11 illustrates that PGO spikes could even be elicited during para- doxical sleep. Thus, these results demonstrate that the same system which spontaneously produces PGO spikes in LGN during sleep can be excited by natural external stimuli during sleep. It is of interest that we have observed that natural stimuli will cause PGO spikes to appear in LGN

A. R. MORiRISON AND R. M. BOWKER

AUD

EOG 1 .O mv

LGN I 0.1

Fig. 10. Normal cat in synchronized sleep exhibiting a spontaneous and sound- induced PGO spike. A slight body twitch, or startle, occurred following the sound as seen in the EMG record. A very subtle EEG change also occurred, but the cat then continued in synchronized sleep. Latency of induced PGO spike, 80-120 msec.

Time calibration. 1 sec.

in synchronized sleep but that others have found that pontine stimulation will not (12, 14).

But in the normal situation in which PGO spikes appear spontaneous- ly before EEG signs of paradoxical sleep and then continue throughout the episode, what is the source of the startle stimulus inducing PGO spi- kes recorded in the lateral geniculate body? We suggest that an internal stimulus for the startle is provided by the increased neuronal activity which begins prior to the onset of the typical low-voltage, high-frequency EEG pattern and complete muscular atonia (29). Firing patterns alter i n the direction of high frequency bursts of activity and have been likened to bursts produced by strychninization or electrical stimulation (20). The pontine reticular formation appears to be the region which responds t o such spontaneous neuronal activity or startling auditory stimuli in sleep and then excites neurons in the other regions in which PGO spikes a r e recorded, LGN and the visual cortex (5).

PGO SPIKES IN THE SLEEPING CAT

A B AUD

m,_ EOG 1.Omv

f - . ------I E E G - Ymw:4

R B 23 N O R M A L Fig. 11. A: Normal cat in paradoxical sleep illustrates spontaneous PGO spikes recorded from LGN. Also, note some individual variation in the wave form of

each spike. B: During the same paradoxical sleep episode a spontaneous PGO spike preceeds a sound-induced PGO spike. EOG activity appears in conjuction with or after the corresponding PGO spike. Latency of induced PGO spike, 80-120 msec. Time cal-

ibration, 1 sec.

Viewed in this context, PGO spikes are merely an epiphenomenon, a consequence of the neural turmoil characterizing paradoxical sleep which seems to center in the lower brainstem and to stimulate a system which can just as well be excited by external stimuli. This conception bestows far les importance on PGO spikes than is common in the sleep literature. Thinking of PGO spikes in this way, one is not surprised that spikes can not be recorded in LGN of the albino rat (48), even though they can be recorded in the brainstem (27). The basic mechanism of in- trinsic excitation of brainstem neurons obviously occurs during transi- tion to paradoxical sleep in this species, but excitation of LGN in such a visually inept animal would not be necessary in a startle situation or when the "startle network" is self-excited during transition to paradox- ical sleep.

Our hypothesis gains in plausibility if one returns to our original observation (39) that PGO spikes and limb jerks in cerebellar-lesioned

834 4 . R. MORRISON AND R. M. BOWKFR

cats are associated on a 1 : 1 basis. Although we originally suggested that the jerks result from a simulation of the vestibular placing reflex induced by brief suppressions of dorsal cervical muscle activity, stimuli other than those exciting vestibular receptors are actually capable of eliciting the same response in wakefulness. Since coming to the conclu- sion that PGO spikes, as well as jerks observed in cerebellar-lesioned cats, are part of a startle mechanism, we have examined the effect of a sudden noise on a cat exhibiting unilateral hyperflexions of the fore- limb in the vestibular placing reflex. A hiss from an aerosol can, quite startling to the cat, would evoke hyperflexion of the involved limb while the cat was standing, so that the animal would stumble to the affected side. Such stimuli, presented repetitively, lost their effectiveness in eli- citing the exaggerated behavior. Yu (53) has shown that painful stimuli, as well as noise, will produce this reaction. After vermal or fastigial le- sions startling stimikli, such as noise or sudden handling, will cause fore- limb extension with toe fanning as well as opisthotonus (personal ob- servations). The link between the similarity of the jerks which occur

w Fig. 12. Cortical lesion shows left paravermal and auditory-visual area destruc- tion. Ipsilateral forelimb flexor movements were observed during walking in wake-

fulness and spontaneously during synchronized sleep.

PGO SPIKES IN THE SLEEPLNG CAT 835

spontaneously or following auditory stimuli in sleep and those produced by the vestibular placing test, then, is not that the response depends solely upon labyrinthine stimulation, but that dropping a cat suddenly is as startling as a hissing aerosol can or an explosive brainstem discharge at the transition to paradoxical sleep. In the normal cat the common res- ponse, postural adjustment preparatory to some reaction, involves the same output mechanism. This response is inappropriately regulated as a result of a cerebellar lesion. This assertion is further bolstered by a recent observation in one oat with bilateral vestibular nerve transection and a paravermal lesion (Fig. 12). Sudden cutaneous stimulation or shin- ing of a light in an eye could elicit a unilateral forelimb hyperflexion in quiet wakefulness similar to the jerks still observed spontaneously in sleep. Forelimb hyperflexion in the vestibular placing test was absent, of course, if the animal were blindfolded.

The jerks observed in conjunction with large sleep spindles (Fig. 5) must be evidence of the "startling" effects of synchronous discharges originating in the cerebral cortex (1. 36), red nucleus (25), or other struc- tures. The phasic cessations of dorsal cervical muscle activity (Fig. I), which sometimes accompany PGO spikes in normal cats (43), are likely to be parallel events rather than instigators of the jerks via stimulation of the labyrinthine receptors as we originally thought (39). They may be caused by sudden excitations, by brainstem discharges, of the medial and descending vestibular nuclei, which Wilson and Yoshida (51) have shown to be inhibitors of cervical motor neurons.

The results Dement and his collaborators (18, 21) have obtained with cats administered para-chlorophenylalanine (PCPA), which depletes the brain of serotonin, also support the hypothesis that PGO spikes are a sign of an intrinsic excitation of a startle mechanism. They have ob- served that after PCPA injections PGO spikes gradually begin to make an appearance during wakefulness or, at least, while the cat exhibits electrographic signs of wakefulness. These results suggest that serotonin normally inhibits the appearance of PGO spikes. This conclusion is sup- ported by McGinty et al. 's (37) demonstration that serotonergic neurons of the dorsal raphe nucleus cease firing prior to each PGO spike. Behav- iorally, cats administered PCPA appear to hallucinate. They exhibit hy- peractive "episodes of perceptual behavior" and "overreact to slight noises". This behavior coilsists of "rapid darting eye movements", "orien- ting movements of the ears" and "visual searching" (21). Most impor- tantly for the present argument, a serotonin-depleted cat shows this orientation reaction (or startle?) each time a PGO spike occurs, much as if the cat were responding to some form of internal stimulation (18).

We have demonstrated that even in the normal sleeping animal the

28 - Acta Neurobiologiae Experimentalis

836 A. R. MORRISON AND R. M. BOWKER

startle network operates and is associated with a PGO spike (Fig. 10 and 11). In paradoxical sleep the normal cat is incapable of responding behaviorally to the internal stimuli generating PGO spikes due to inhi- bition of the spinal motor neurons (24, 40) and is apparently depressed sufficiently to resist the alerting effects of these stimuli during the tran- sition to paradoxical sleep. Cats exhibiting episodes of paradoxical sleep without atonia following pontine tegmental lesions (28, 32), however, do respond to such internal stimuli; for they have hallucinatory-like beha- vior when PGO spikes are occurring. Thus, in this situation the bursts of neuronal activity during paradoxical sleep are capable of eliciting be- havioral responses because the muscles are no longer atonic; however, they are incapable of arousing cats with pontine lesions from the parado- xical sleep state because such animals can move grossly for several mi- nutes without awakening.

In the first days following cerebellar damage caused by the place- ment of pontine tegmental lesions (28, 32) by a transcerebellar approach or lesions of the anterior lobe of the cerebellum alone (Fig. 3) (39), though, cats are often abruptly awakened from synchronized sleep with a jerk after varying lengths of times, much as if they were startled. Cats with cerebellar damage continue to be easily awakened from all stages of sleep by the slightest noise as well as to be hyperirritable dur- ing wakefulness throughout postoperative survival periods lasting se- veral months. The latter was noted earlier by Chambers and Sprague (13). Thus, there is something about anterior lobe cerebellar lesions which affects a cat's ability to maintain itself in synchronized sleep in the early postoperative period in particular. Lesions of the auditory-visual area of the cerebellum, on t h ~ . other hand, raise the threshold for arousal from synchronized sleep (Fig. 7) during the first postoperative week, which permits one to elicit PGO spikes with auditory stimuli without EEG arousal (7). In wakefulness cats with such lesions are inattentive to both auditory and visual stimuli.

We now suggest a plausible explanation for these effects of cerebel- lar lesions: Serotonin depletion by inhibition of synthesis with PCPA suppresses sleep (17, 33, 52), enhances pain perception (49), increases arousal and emotionality (22), and prolongs habituation to auditory sti- muli (15). Serotonin is largely produced within the raphe nuclei of t he brainstem (38), which receive some projections from the fastigial nuclei (9). These nuclei, in turn, receive projections from the cerebellar cortical areas involved by our lesions (19, 50). The effects of anterior lobe o r auditory-visual area lesions, with regard to ease or difficulty of arousal and long-term change in affect of the anterior lobe-lesioned cats, could be due to an alteration in the modulation of the activity of the seroto-

PGO SPIKES IN THE SLEEPING CAT 83 7

nergic raphe neurons by the cerebellum, making cats more or less suscep- tible to the startling effects of external stimuli feeding into a brainstem neural system and to those originating within that same system as the organism's central nervous system begins the transition into paradoxical sleep. In the normal cat the nervous system has reached a state in which the increased activity of neurons in the pontine or brainstem "startle network", when released gradually from the inhibitory effects of sero- tonin, is insufficient to arouse the animal. Only when serotonin is re- moved by raphe lesions (31) or, as we hypothesize, its release is altered by cerebellar lesions are the startling effects of the brainstem discharges either frequent or intense enough to produce insomnia or difficulties in cycling through various stages of sleep. I t is obvious, however, recogni- zing the complexity of the nervous system, that other fiber systems (38) also play a role in sleep onset and maintenance. I t is probably their ac- tivity which allows the normal organism to remain unresponsive to the startle stimuli provided by the neural turmoil generating PGO spikes and permits a cat to recover sleep during chronic administration of PCPA (18).

This investigation was supported by N.I.H. Grant NS 08377. We thank Ms. G. Mann for her excellent technical assistance.

REFERENCES

1. ARDUINI, A., BERLUCCHI, G. and STRATA, P. 1963. Pyramidal activity dur- ing sleep and wakefulness. Arch. Ital. Biol. 101: 530-544.

2. ASERINSKY, E. and KLEITMAN, N. 1953. Regularly occurring periods of eye motility, and concomitant phenomena during sleep. Science 118: 273-274.

3. BARD, P. and MACHT, M. B. 1958. The behavior of chronically decerebrate cats. In G. E. W. Wolstenholme and C. M. O'Connor Symposium. Churchill, London, p. 55-71.

4. BIZZI, E. 1966. Discharge patterns of single geniculate neurons during the rapid eye movements of sleep. J. Neurophysiol. 29: 1087-1096.

5. BIZZI, E. and BROOKS, D. C.* 1963. Functional connections between pontine reticular formation and lateral geniculate nucleus during deep sleep. 101: 666-680.

6. BIZZI, E., POMPEIANO, 0 . and SOMOGYI, I. 1964. Spontaneous activity of single vestibular neurons of unrestrained cats during sleep and wakefulness. Arch. Ital. Biol. 102: 308-330.

7. BOWKER, R. M. and MORRISON. A. R. 1974. A role for the cerebellar auditory area in sleep arousal mechanisms. Presented a t Association for Psychophy-

siological Study of Sleep a t Jackson Hole, Wyoming. 8. BOWKER, R. M. and MORRISON, A. R. 1975. The startle reflex and PGO

spikes. Brain Res. (in press).

838 A. R. MO~RISON AND R. M. BOWKER

9. BRODAL, A., WALBERG, F. and TABER, E. 1960. The raphe nuclei of the brain stem in the cat. 111. Afferent connections. J. Comp. Neurol. 114: 261-281.

10. BROOKS, D. C. 1967. Effect of bilateral optic nerve section on visual monopha- sic wave activity in the cat. Electroenceph. Clin. Neurophysiol. 23: 134-141.

11. BROOKS, D. C. and BIZZI, E. 1963 Brainstem electrical activity during sleep. Arch. Ital. Biol. 101: 648-665.

12. BROOKS, D. C. and GERSHON, M. D. 1971. Eye movement potentials in oculo- motor and visual systems of the cat: A comparison of reserpine-induced waves with those present during wakefulness and rapid eye movement sleep. Brain Res. 27: 223-239.

13. CHAMBERS, W. W. and SPRAGUE, J. M. 1955. Functional localization in the cerebellum. 11. Somatotopic organization in cortex and nuclei. Arch. Neu- rol. Psychiat. 4: 653-680.

14. COHEN, B. and FELDMAN, M. 1968. Relationship to electrical activity in pon- tine reticular formation and lateral geniculate body to rapid eye move- ments. J. Neurophysiol. 31: 806-817.

15. CONNER, R. L., STOLK, J. M., BARCHAS, J. D. and LEVINE, S. 1970. Para- chlorophenylalanine and habituation to repetitive auditory startle stimuli in rats. Physiol. Behav. 5: 1215-'1219.

16. De KLEIJN, A. 1924. Experimental physiology of the labyrinth. Proc. Royal Soc. Med., Sect. Otol. 17: 6-23.

17. DELORME, F., FROMENT, J. L. and JOUVET, M. 1966. Suppression du som- meil par la P, chlormethamphetamine et de la P. chlorophenylalanine. C. R. Soc. Biol. 160: 2347-2351.

18. DEMENT, W. C. 1969. The biological role of REM sleep. In A. Kales (ed.), Sleep, physiology and pathology. J. B. Lippincott Co., Phila., p. 245-265.

19. EAGER, R. P. 1963. Efferent cortico-nuclear pathways in cerebellum of the cat. J. Comp. Neurol. 120: 81-104.

20. EVARTS, E. V. 1967. Activity of individual cerebral neurons during sleep and arousal. In S. S. Kety, E. V. Evarts, and H. L. Williams, (ed.). Sleep and altered states of consciousness. Res. Publ. Assn. Nerv. Ment. Dis., Williams and Wilkins, Baltimore, p. 319-336.

21. FERGUSON, J., HENRIKSEN, S., COHEN, H., MITCHELL, G., BARCHAS, J. and DEMENT, W. 1970. "Hypersexuality" and behavioral changes in cats caused by administration of p. chloraphenylalanine. Science 168: 499-501.

22. FIBIGER, H. C. and CAMPBELL, B. A. 1971. The effect of para-chlorophenyla- lanine on spontaneous locomotor activity in the rat. Neuropharmacol. 10: 25-32.

23. FOX, M. W. 1963. Conditioned reflexes andrinnate behavior of neonate dog. J. Small Anim. Pract. 4: 85-99.

24. GASSEL, M. M., MARCHIAFAVA, P. L. and POMPEIANO, 0. 1965. An analysis of supraspinal influences acting on motoneurons during sleep in the unre- strained cat. Modification of the recurrent discharge of the alpha motoneu- rons during sleep. Arch. Ital. Biol. 103: 25-44.

25. GASSEL, M. M., MARCHIAFAVA, P. L. and POMPEIANO, 0. 1965. Activity of red nucleus during deep, desynchronized sleep in unrestrained cats. Arch. Ital. Biol. 103: 369-396.

26. GASSEL, M. M. and POMPEIANO, 0. 1965. Fusimotor functions during sleep in unrestrained cats. Arch. Ital. Biol. 103: 347-368.

PGO SPIKES IN THE SLEEPING CAT 839

27. GOTTESMAN, C. 1967. Recherche sur la psychophysiologie du sommeil chez le rat. Ph. D. Thesis. Palais Royal. Paris. 156 p.

28. HENLEY, K. and MORRISON, A. R. 1974. A re-evaluation of the effects of lesions of the pontine tegmentum and locus coeruleus on phenomena of pa- radoxical sleep in the cat. Acta. Neurobiol. Exp. 34: 215-232.

29. HOBSON, J. A., McCARLEY, R. W., PIVIK, R. T. and FREEDMAN, R. 1964. Selective firing by cat pontine brainstem neurons in desynchronized sleep. J . Neurophysiol. 37: 497411.

30. JEANNEROD, M. and KIYONO, S 1969 Effets de la rberpine sur la response reticulaire aux stimulations sensorielles Brain Res. 12: 129-137.

31. JOUVET, M. 1972. The role of monoamines and acetylcholine containing neurons in the regulation of the sleep-waking cycle. In Ergebnisse der Physiologie: Neurophysiology and neurochemistry of sleep and wakefulness. Springer Verlag, New York, 64: 166-307.

32. JOUVET, M. and DELORME, I?. 1965. Locus coeruleus et sommeil paradoxal. C. R. Soc. Biol. 159: 895-899.

33. KOELLA, W. P. and CZICMAN, J. 1966. Mechanism of the EEG synchronizing action of serotonin. Amer. J. Physiol. 211: 926-935.

34. LAURENT, J., CESPUGLIO, R. and JOUVET, M. 1974. =limitation des voies ascendantes de l'activite ponto-gkniculo-occipitale chez le chat. Brain Res. 65: 29-52.

35. LENZI, G. L., POMPEIANO, 0. and SATOH, T. 1969. Transmission of labyrin- thine volleys through the vestibular nuclei during sleep. Arch. Ital. Biol. 107: 19-42.

36. MARCHIAFAVA, P. L. and POMPEIANO, 0. 1964. Pyramidal influences on the spinal cord during desynchronized sleep. Arch. Ital. Biol. 102: 500429.

37. McGINTY, D. J., HARPER, R. M. and FAIRBANKS, M. K. 1973. 5-HT-contain- ing neurons: unit activity in behaving cats. In J . Barchas and E. Usdin (ed.), Serotonin and behavior. Acad. Press, Inc., New York, p. 267-279.

38. MORGANE, P. J. and STERN., W. C. Chemical anatomy of brain circuits in relation to sleep and wakefulness. In E. Weitzman (ed.), Advances in sleep research. Vol. 1 Spectrum Publications, Inc., New York, p. 1-131.

39. MORRISON, A. R. and BOWKER, R. M. 1973. Cerebellar and spinal contribu- tions to the regulation of muscle tone and movement during sleep. In U. J. JovanoviC (ed.), The nature of sleep. Gustav Fischer, Stuttgart, p. 270-277.

40. MORRISON, A. R. and POMPEIANO, 0. 1965. An analysis of supraspinal in- fluences acting on motoneurons during sleep in the unrestrained cat. Res- ponses of the alpha motoneurons to direct electrical stimulation during sleep. Arch. Ital. Biol. 103: 497-516.

41. MUNSON, J. B. 1974. Eye movement potentials following visual deafferentation. Brain Res. 66: 435-442.

42. MUNSON, J. B. and GRAHAM, R. B. 1973. Localization of lateral geniculate spikes in alert, sleeping and reserpine-treated cats. Electroenceph. Clin. Neurophysiol. 35: 323-326.

43. PIVIK, T. and DEMENT, W. C., 1970. Phasic changes in muscular and reflex activity during non-REM sleep. Exp. Neurol. 27: 115-124.

44. SIMON, R. P., GERSHON, M. D. and BROOKS, D. C. 1973. The role of raphe nuclei in the regulation of ponto-geniculo-occipital wave activity. Brain Res: 58: 313-330.

810 A. R. MORRISON AND R. M. BOWKER

45. SNIDER, R. S. and STOWELL, A. 1944. Receiving areas of the tactile, auditory, and visual systems in the cerebellum. J. Neurophysiol. 7: 331-357.

46. SPRAGUE, J. M., LEVITT, M., ROBSON, K., LIU, C. N., STELLAR, E. and CHAMBERS, W. W. 1963. A neuroanatomical and behavioral analysis of the syndromes resulting from midbrain lemniscal and reticular lesions in the cat. Arch. Ital. Biol. 101: 225-295.

47. STERMAN, M. B., KNAUSS, T., LEHMANN, D. and CLEMENTE, C. D. 1965. Circadian sleep and waking patterns in the laboratory cat. Electroenceph. Clin. Neurophysiol. 19: 509-517.

48. STERN, W. C., FORBES, W. B. and MORGANE, P. G. 1974. Absence of ponto- geniculo-occipital (PGO) spikes in rats. Physiol. Behav. 12: 293-295.

49. TENEN, S. S. 1967. The effects of p-chorophenylalanine, a serotonin depletor, on avoidance acquisition, pain sensitivity and related behavior in the rat. Psychopharmacologia 10: 204-220.

50. WALBERG, F. and JANSEN, J. 1964. Cerebellar corticonuclear projection stu- died experimentally with silver impregnation methods. J. Hirnforsch. 6 : 338-354.

51. WILSON, V. J. and YOSHIDA, M. 1969. Bilateral connections between labyrinths and neck rnotorneurons. Brain Res. 13: 603-607.

52. WEITZMAN, E. D., RAPPORT, M. M. McGREGOR, P. and JACOBY, J. 1968. Sleep patterns of the monkey and brain serotonin concentration: Effect of p-chlorophenylalanine. Science 160: 1361-1363.

53. YU, J. 1972. The pathway mediating ipsilateral limb hyperflexion after cerebel- lar paravermal cortical ablation or cooling in cats. Exp. Neurol. 36: 549-562.

Received 13 December 1974

Adrian R. MORRISON, Laboratories of Anatomy, School of Veterinary Medicine, and Institute of Neurological Sciences, University of Pennsylvania, Philadelphia, Pennsylvania 19174, USA. Robert M. BOWKER, Laboratories of Anatomy and Institute of Neurological Sciences, Univer- sity of Pennsylvania, Philadelphia, Pennsylvania 19174, USA.