(D-4Ota) and Other Central Muscle Relaxants on Spinal and Supraspinal Reflexes in Cats

Japan. J. Pharmacol, 23, 83-96 (1973) 83

EFFECTS OF NEW s-TRIAZOLOBENZODIAZEPINE (D-4OTA) AND OTHER CENTRAL MUSCLE RELAXANTS ON SPINAL AND SUPRASPINAL REFLEXES IN CATS

Sukehiro CHIBA and Yuji NAGAWA Biological Research Laboratories, Central Research Division, Takeda Chemical Industries, Ltd., Higashiyodogawa-ku, Osaka, Japan

Accepted 29 August 1972

Abstract-The inhibitory effect of benzodiazepine-type agents such as 8-chloro-6- phenyl-4H-s-triazolo [4, 3a] [1, 4] benzodiazepine (D-40TA), diazepam and nitrazepam on the spinal and supraspinal polysynaptic reflexes was compared with that of mephenesin, methocarbamol, chlorzoxazone and chlormezanone in cats. Benzodiazepines did not depress the spinal and supraspinal polysynaptic reflexes or reflex potentials in the spinal and the gallamine-immobilized cats. In these respects, chlormezanone resembled benzodiazepines. On the other hand, mephenesin, methocarbamol and chlorzoxazone blocked these reflexes in all kinds of preparations such as the anesthetized, the spinal, the decerebrate and the gallamine-immobilized preparations. D-40TA depressed gamma rigidity at the dose below that necessary to depress alpha rigidity. Moreover, it inhibited more profoundly the tonic stretch reflex than the phasic one. Spontaneous and evoked discharges of the muscle spindle in the decerebrate cat were significantly depressed by D-40TA, while those of spinal cat were unaffected. These results suggest that skeletal muscle relaxation by benzodiazepines including D-40TA is attributed to the primary depression of the brain stem reticular system and in turn the ascending inhibitory action via the gamma system on the spinal and supraspinal polysynaptic neurons. It should be stressed that the integrity of the connection between the brain stem and gamma system in the spinal cord and its related muscles is also required for eliciting the depression of supraspinal polysynaptic reflex by these agents. Mephenesin-type agents presumably inhibit directly the interneurons at the spinal and supraspinal levels.

Mephenesin, first introduced by Berger and Bradley (1) as the centrally acting muscle relaxant, has been postulated to have a selective action on spinal interneuron as polysynaptic reflex responses were abolished without effect on monosynaptic reflex responses (2). Later, the benzodiazepine agents have been found to possess tranquilizing and muscle relaxant properties. The depression of polysynaptic spinal reflexes by diazepam, one of these benzodiazepines, has been proposed to be due to the action principally on the brain stem reticular formation, on the basis of the finding that spinal polysynaptic depression in decerebrate cats is nullified by spinal transection (3). Our current study on the benzodiazepines led to the discovery of a new type of com- pound, 8-chloro-6-phenyl-4H-s-triazolo [4, 3a] [1, 4]-benzodiazepine (D-40TA) which was highly active in tranquilizing, sedative-hypnotic and muscle relaxant effects in experi- mental animals (4). In our neuropharmacological study on this agent, the depression of 84 S. CHIBA & Y. NAGA WA not only spinal polysynaptic reflex but also supraspinal polysynaptic reflex observed in decerebrate cat was not demonstrated in spinal cat or in gallamine-immobilizedcat. The purpose of the present experiments was to clarify whether or not the mode of action of D-40TA is different from that of diazepam, nitrazepam, mephenesin and other central muscle relaxants. The effect of D-40TA on the gamma system was also studied.

METHODS Seventy-nine cats of both sexes weighing from 2.4 to 4.9 kg were used and the following four kinds of preparations were made : Immobilized cat with gallamine triethiodide, anesthetized cat with combination of a-chloralose (40 mg/kg, i.v.) and urethane (400 mg/ kg, i.p.), spinal cat with transection at Cl level and decerebrated cat at intercollicular level. Immobilization with gallamine was preceded by ether anesthesia for surgical operation. Transection of either the spinal cord or the brain stem was accomplished with a blunt spatula under ether anesthesia. The gallamine-immobilizedand the spinal cats were maintained on artificial respiration. After surgical operation, the ether was withdrawn and at least 3 hr elapsed before experimentation began. In gallamine-immobilizedcats, the pressure points were anesthetized with procaine. Experiments with mechanical responses of spinal and supraspinal reflexes were per- formed in three kinds of preparations, the anesthetized, the spinal and the decerebrate cats. After the head and bilateral femurs of animal were fixed on a stereotaxic instruments, the patellar reflexes were elicited by tapping the right patellar tendon with an automatically operated magnetic hammer once every 5-10 sec. On the left hind limb, either flexor reflex response of the anterior tibialis muscle to supramaximal stimulation of the peroneal nerve or ipsilateral extensor reflex response of the quadriceps femoris muscle to supramaximal stimulation of the sciatic nerve were obtained. In some preparations, the left peroneal nerve was stimulated to provoke both flexor and contralateral extensor reflex responses. The central cut end of either the peroneal or the sciatic nerve was placed on bipolar platinum stimulating electrodes. The parameters of stimuli, which were delivered from a Nihon Koden electronic stimulator (MSE-3), were rectangular, 0.3 Hz in frequency and 5 msec in pulse duration. In addition to these spinal reflexes, supraspinal linguomandibular reflex was also produced once per 3.3-5 sec by supramaximal stimulation of the ton- gue through needle electrodes. These reflex responses were recorded by means of force- displacement transducers on a Nihon Koden polygraph (RM-150). At the same time, the electromyogram (EMG) of extensor lateral digitorum muscle in forelimb and the bra- chial arterial blood pressure were routinely recorded. Experiments with electrical responses of spinal and supraspinal reflexes were per- formed in the gallamine-immobilizedcats and in the spinal or the decerebrate cats. After the head, spine and ankle of animal were rigidly mounted on a stereotaxic instrument, the animal was subjected to exposure of the lumbosacral cord and the popliteal fossa under ether anesthesia. Following the surgical procedure, the exposed regions were covered by liquid paraffin warmed at normal body temperature. All the ventral roots from L4 D-40TA ON SPINAL REFLEX 85 to Si were bilaterally severed in order to eliminate the feedback control through the gamma loop. The central cut end of the tibial nerve was placed on the bipolar platinum stimulating electrodes. The stimulating pulses were rectangular, 0.3 Hz in frequency, 0.5 msec in pulse duration and 4 times threshold for the spinal monosynaptic reflex potentials in intensity. The spinal mono-and polysynaptic reflex potentials and dorsal root reflex potentials were recorded by means of bipolar platinum recording electrodes attached to the central cut ends of L7 ventral root filaments and L6 dorsal root filaments, respectively. For recording of supraspinal reflex potentials, the nerve innervating the digastric muscle was exposed at the temporal fossa, and the central cut end of the lingual nerve was supra- maximally stimulated. These reflex potentials were displayed on a Nihon Koden dual beam oscilloscope (VC-7A) and photographed. In some decerebrate cats, only the ventral and dorsal root filaments for recording and the tibial nerve for stimulation were severed in order to observe the indirect effect on spinal reflex potentials through the gamma loop. On the other hand, some other decerebrate cats were pretreated with gallamine to produce functional deafferentation of muscle spindles. Effects on phasic and tonic reflex responses were investigated, using decerebrate cats according to the same method as that devised by Smith et al. (5). The Achilles tendon was sectioned and attached to a force-displacement transducer mounted on an automatic traction apparatus. The phasic stretch reflex was evoked by displacing the transducer 10 mm distally at a velocity of 100 mm/sec. Subsequently, by maintaining the displaced state for 10 sec the tonic stretch reflex was evoked. The muscle tension caused by stretch reflexes was determined by subtracting "passive" tension obtained after complete paralysis induced by adequate dose of gallamine and nerve sectioning. For the test of gamma rigidity of the decerebrated cats at intercollicular level, the animals exhibiting marked rigidity were selected about 2 hr after the termination of ether anesthesia. The resistance of shoulder, elbow, carpal, hip, knee and tarsal joints in four limbs to forced flexor movement was scored from 0 for complete loss of resistance to 3 for severe rigidity before and after administration of the agent. Effects on alpha rigidity of the decerebrated cats by ligating both the carotid and basilar arteries were also investigated by recording of the EMGs of the biceps brachii muscle, triceps brachii muscle and trapezius muscle. Effects on muscle spindle activity were observed in both the spinal and the decerebrate cats. A single afferent fiber (G1a) originating from muscle spindle in the gastrocnemius muscle was identified in the dorsal root filaments of L7 segment according to the criteria of Hunt (6). The spontaneous and evoked spindle discharges caused by stretching the muscle were recorded on the oscilloscope and photographed. For simultaneous recording of the electrical activities of the alpha as well as gamma fibers, the decerebrate cats were used. Filaments which showed sustained rhythmical discharges of relatively low amplitude or impulses of comparatively high amplitude were selected from the severed ventral root of L7 segment (7). With the exception of these ventral filaments for recording and tibial nerve for stimulation, all the peripheral nerves of the hind limb tested and 86 S. CHIBA & Y. NAGA WA the ventral and dorsal roots were left intact. The spontaneous gamma and reflexivelyevoked alpha and gamma efferent discharges by the tactile stimulation or the afferent stimulation of the tibial nerve were displayed on dual beam oscilloscope and photographed. D-40TA, nitrazepam and diazepam synthesized in this chemical laboratory were dis- solved in 20 % glycofurol solution and mephenesin (MyoserolR, Sankyo) in 10 % pro- pyleneglycol solution for injection into the radial vein. For intraperitoneal injection, chlorzoxazone (SolaxinR, Eisai), methocarbamol (RobaxinrR, Takeda) and chlormezanone (TrancopalrR,Daiichi) were suspended in 5% gummi arabicum solution.

RESULTS 1) Effects on the patellar, flexor, ipsi-and contralateral extensor, linguomandibularre- flexes and the EMG activity D-40TA showed a marked depressive effect on the flexor, ipsi- and contralateral extensor and linguomandibular reflexes in all of 5 decerebrate cats, as indicated in Fig. 1. These effects reached a peak within 5 min after administration and lasted for 20 to 30 min at the intravenous dose of 0.1 mg/kg, which was the minimum effective dose for inhibition of the linguomandibular reflexes. The higher dose, 0.2 mg/kg, produced more than

FIG. 1. Effect of D-40TA on patellar reflex (PR), linguomandibular reflex (LMR) and ipsilateral extensor reflex (ER) in decerebrate cat. BP.: blood pressure. D-4OTA ON SPINAL REFLEX 87

80% depression in the flexor, ipsi-and contralateral extensor and linguomandibular reflexes for more than 2 hr. Increase in dosage levels of D-40TA caused longer depression in duration. The EMG activity was also reduced for a long time, synchronized with the depression of the spinal and linguomandibular reflexes. On the other hand, the change in blood pressure induced by D-40TA was only a transient hypotension ranging from 20 to 40 mmHg, indicating no direct relationship between the fall of blood pressure and the

TABLE 1. Effects of various agents on linguomandibular (LMR), extensor (ER) and patellar (PR) reflexes in decerebrate cat.

―: Nochange ↓: Depress, ↑: Augment, .

FIG. 2. Effect of D-40TA on patellar reflex (PR), linguomandibular reflex (LMR) and ipsilateral extensor reflex (ER) in spinal cat. BP.: blood pressure. 88 S. CHIBA & Y. NAGA WA observed effects on these reflexes. The patellar reflexeswere never depressed at the dosage levels ranging from 0.2 to 2.0 mg/kg in the decerebrate preparations. Conversely, in some animals, the seeminglydiminished amplitude of the patellar reflexescaused by the increased tonus of the quadriceps femoris muscle was enhanced to its normal amplitude by decreasing the muscle tonus as the result of relieving decerebrate rigidity, the patellar reflexes being normalized. Comparative results obtained with D-40TA, nitrazepam, diazepam, methocarbamol, chlormezanone, chlorzoxazone and mephenesin in the dosage levels capable of producing more than 80 % depression in the contralateral extensor and linguomandibular reflexes of 3-5 decerebrate preparations are presented in Table 1. On the other hand, the patellar, flexor, ipsi- and contralateral extensor and linguomandibular reflexes in the spinal cats were not depressed by the intravenous injection of 0.2 to 2.0 mg/kg of D-40TA but rather augmented in 3 of 5 animals, as shown in Fig. 2. Increase of dosage levels up to 10 mg/kg caused no modification. As shown in Table 2, the data obtained using 3-5 spinal cats for each agent indicated that D-40TA resembled nitrazepam, diazepam and chlormezanone, but differed from mephenesin, chlorzoxazone and methocarbamol. All the latter three agents depressed the flexor, ispi- and contralateral extensor and linguomandibular reflexes even in the spinal preparations.

TABLE 2. Effects of various agents on linguomandibular, extensor and patellar reflexes in spinal cat.

―: Nochange ↑ : Depress, ↓: Augment,

In 4 ƒ¿-chloralose-urethanized cats, the intravenous dose of 0.125 mg/kg of D-40TA depressed almost completely the flexor, ipsi-and contralateral extensor and linguomandibular reflexes without affecting the patellar reflexes, for more or less than 4 hr. In contrast to the transient hypotension observed in the decerebrate and the spinal preparations, a fall in blood pressure of the anesthetized preparations persisted during the depression of the reflexes. The flexor and linguomandibular reflexes depressed by D-40TA were restored to 70-80 % of control level after the spinal cord transection at Cl level. There- after the restored reflexes were unchanged even after the additional dose of 2 to 10 mg/ kg of D-40TA, nitrazepam and diazepam, respectively, but clearly inhibited by the dose of 25 mg/kg of mephenesin (Fig. 3). D-4OTA ON SPINAL REFLEX 89

A) Anesthetized cat

B) 15 min after spinal section

FIG. 3 A) Effect of D-40TA on linguomandibular reflex (LMR), flexor reflex (FR) and blood pressure (BP) in anesthetized cat. B) Effect of D-49TA and mephenesin on both reflexes after spinal transection.

2) Effects on the spinal and supraspinal reflex potentials The spinal polysynaptic reflex potentials recorded from L7 ventral root of the decerebrate cats, which maintained the gamma efferent and spindle afferent nerves intact except the nerves for stimulation and recording, were reduced by more than 50 % following the intravenous dose of 0.5 mg/kg of benzodiazepine-type agents such as D-40TA, nitrazepam and diazepam, but the monosynaptic reflex potentials were unchanged or only slightly augmented. At the same time, the dorsal root reflex potentials from L6 dorsal root were augmented by about 200 %. (Fig. 4A). These effects persisted for more than 2 hr. On the other hand, polysynaptic reflex potentials in the preparations such as the gallamine-immobilizedand the spinal cats were never depressed even at the accumula- tive dose of 2 mg/kg injected intravenously at 30 min intervals. The same results were obtained in the decerebrate cats pretreated with gallamine. The dorsal root reflex potentials, however, were maximally enhanced already at the intravenous dose of 0.2-0.5 mg/kg of these agents in all preparations. Fig. 5 shows comparative results of D-40TA on the dorsal root reflex and mono-and polysynaptic reflex potentials in the gallamine- immobilized, the spinal and the decerebrate cats. Intravenous injection of 25 mg/kg of mephenesin, which is known to block the interneurons at the spinal and reticular formation levels (8), considerably depressed the poly- 90 S. CHIBA & Y. NAGA WA A) Decerebrate cat CONTROL

B) Gallamine-immobilized cat

FIG. 4 A) Effect of D-40TA on dorsal root reflex potentials (DRR) and spinal mono- and polysynaptic reflex potentials (SR) in decerebrate cat. B) Comparison of effects of D-40TA and mephenesin on DRR and spinal polysynaptic reflex potentials in gallamine-immobilized cat.

FIG. 5. Comparative results of D-40TA on dorsal root reflex potentials (DRR), spinal monosynaptic reflex potentials (DRR), spinal monosynaptic reflex potentials (MSR) and polysynaptic reflex potentials (PSR) in three kinds of preparations, the gallamine-immobilized, the spinal and the decerebrate cats. Each point represents mean % change obtained from 5 animals. D-40TA ON SPINAL REFLEX 91

A) Decerebrate cat

CONTROL

B) Galiamine-immobilized cat

FIG. 6. Effect of D-40TA on reflex potentials of the digastric nerve in response to afferent stimulation of the lingual nerve in decerebrate cat (A) and in gallamine- immobilized cat (B). Note failure of D-40TA to depress the reflex of gallamine- immobilized cat.

synaptic reflex potentials without augmenting the dorsal root reflex potentials in all preparations such as the decerebrate, the spinal and the gallamine-immobilized cats. (Fig.

4B). In these respects, mephenesin resembled methocarbamol and chlorzoxazone.

Intravenous dose of 0.2 mg/kg of D-40TA also depressed the supraspinal polysynaptic reflex potentials of the digastric nerve in response to afferent stimulation of the lingual nerve in the decerebrate cats, as indicated in Fig. 6. Such a depressive effect of D-40TA, however, was not observed in both the spinal and the gallamine-immobilized cats and in the decerebrate cats pretreated with gallamine.

3) Effects on gamma and alpha rigidities and stretch reflexes

The inhibitory effect of D-40TA on the gamma driven rigidity was compared with that of nitrazepam and diazepam, using the decerebrate cats exhibiting marked rigidity.

These agents at the intravenous dose of 0.5 mg/kg were applied at the hourly intervals.

The mean•}SEM. (mg/kg) of the dose which reduced the total score to 50 % of value obtained before administration was 0.84•}0.09 (4 animals) for D-40TA, 2.27•}0.10 (4 animals) for nitrazepam and 1.92•}0.55 (5 animals) for diazepam, respectively. D-40TA proved to be twice as potent as nitrazepam and diazepam. The alpha rigidity induced by anemic decerebration was unaffected by D-40TA at the intravenous dose of 1 mg/kg, but a significant depression was observed following an intravenous dose above 3 mg/kg of D-40TA in all of 3 preparations. These results indicate that the principal relaxant effect is on the gamma rather than on the alpha motor system.

The effect of D-40TA on the stretch reflex caused by the afferent activity from muscle spindle was investigated using 4 decerebrate cats. The abrupt stretching of the gastrocnemius muscle produced a rapidly developing reflexive contraction of the muscle (phasic

stretch reflex) only at the early period and the successively persistent reflexive contraction 92 S. CHIBA & Y. NAGA WA

(tonic stretch reflex) during maintained stretch. Maximum % depression (mean•}SEM.) obtained after the intravenous injection of 0.5 mg/kg of D-40TA was 10.3•}5.9 for phasic stretch reflex and 42.6•}4.4 for tonic stretch reflex. The depressing effect by this dose lasted for about 4 hr.

4) Effects on gamma motor system

It was of interest to study the site of action of D-40TA in the gamma system. For this purpose, the effects of D-40TA on the spontaneous and evoked spindle discharges by stretching of the gastrocnemius muscle were observed in the decerebrate and the spinal cats. As shown in Fig. 7, frequency of the spontaneous spindle discharges of Gla fiber

(conduction velocity, about 100 m/s) was 40 to 70 Hz and that of the discharges in response to the muscle tension by stretching was 200 to 400 Hz. Intravenous injection of 0.5 mg/ kg of D-40TA depressed the spontaneous and evoked discharges of the muscle spindle for more than 4 hr in all of 4 decerebrate cats, whereas, the spontaneous and evoked spindle discharges of the same muscle in 3 spinal cats were not affected by D-40TA even at a high dose of 2 mg/kg.

Further investigations were made by recording of the gamma efferent activity from

L7 ventral root of 3 decerebrate cats. Discharges of low amplitude occurring during a slight stretching of the gastrocnemius muscle were regarded as originating from the efferent activity of the gamma motoneurons. In addition to these gamma efferent discharges, the spikes of high amplitude which were regarded as originating from the efferent activity of the alpha motoneurons, were also evoked by tactile stimulation. As indicated in Fig.

8, D-40TA at an intravenous dose of 0.5 mg/kg depressed almost simultaneously both the evoked discharges of high and low amplitudes. The reflexive alpha and gamma moto- neur onal activities recorded from L7 ventral root in response to the afferent stimulation

CONTROL

10 min after D-40TA 0.5 mg/kg iv

FIG. 7. Effect of D-40TA on spontaneous and evoked spindle discharges recorded from Gla fiber of L7 dorsal root in decerebrate cat. D-40TA ON SPINAL REFLEX 93

a) spont.

b) tactile stim.

c) N. tibial stim.

FIG. 8. Effect of D-40TA on spontaneous gamma motoneuronal discharges (a), and evoked alpha (high spike) and gamma (low spike) motoneuronal discharges (b, c) recorded from alpha and gamma efferent fibers of L7 ventral root in decerebrate cat. of the tibial nerve were also depressed by the same treatment of D-40TA for more or less than 4 hr.

DISCUSSION The skeletal muscle relaxation by mephenesin-type agents such as methocarbamol and chlorzoxazone is reported to be derived from the block of interneurons at the spinal and supraspinal levels (9-12). In the present experiments, these agents blocked the spinal polysynaptic reflexes, extensor and flexor reflexes, and the supraspinal polysynaptic reflexes, linguomandibular reflex in three kinds of preparations such as the anesthetized, the spinal and the decerebrate cats. Furthermore, the spinal polysynaptic reflex potentials in the gallamine-immobilized, the spinal and the decerebrate cats were also reduced. These results strongly suggest that mephenesin-type agents act on the interneurons at the spinal and supraspinal levels. On the other hand, benzodiazepine-type agents did not depress these reflexes and reflex potentials of the spinal and the gallamine-immobilized cats. Ghelarducci et al. also reported no significant effect of nitrazepam on the spinal polysynaptic reflex potentials in the gallamine-immobilized, decerebrated cats (13). These reflex responses and reflex potentials in the anesthetized and the decerebrate cats, which are regarded as maintaining feedback control mechanism proposed by Hunt and Paintal (14) and Eldred et al. (15), were however significantly inhibited by benzodiazepine-type agents, In these respects, the action of chlormezanone was similar to benzodiazepine- 94 S. CHIBA & Y. NAGA WA type agents. It has been suggested that diazepam and nitrazepam augment presynaptic inhibition in the spinal cord on the basis of the observation that the dorsal root reflex potentials were remarkably augmented by these agents (16, 17). Eccles et al. (18) have postulated that this increase in presynaptic inhibition gives rise to depression of the spinal polysynaptic reflex potentials. Benzodiazepine-type agents, however, did not affect the spinal polysynaptic reflex potentials in the gallamine-immobilizedand the spinal cats and in the decerebrate cats pretreated with gallamine, although these agents augmented the reflex potentials of the dorsal root. This augmenting effect, therefore, seems unlikely to be responsible for the depression of the spinal polysynaptic reflex potentials. Ngai et al. (3) found that benzodiazepine-type agents are more active on the decerebrate than on the spinal cats concerning the skeletal muscle relaxing effect. It was also reported that doses of benzodiazepine which did not affect the alpha motor system depressed gamma rigidity (19). Many investigators have postulated that the mechanism of nitrazepam or diazepam producing skeletal muscle relaxation is due to an inhibitory action on the brain stem reticular formation, most likely on the reticular facilitatory system (13, 20, 21). In the present experiments, marked inhibition of gamma rigidity was demonstrated after the injection of nitrazepam, diazepam and D-40TA. Moreover, D-40TA depressed gamma rigidity at the dose below that necessary to depress alpha rigidity. Also, D-40TA inhibited the tonic stretch reflexes of the gastrocnemius muscle, which is presumed to be more dependent on the state of the gamma system than the phasic components (22). Such high sensitivity of the gamma system to D-40TA in the decerebrate preparation implies that the depression of the facilitatory control of the reticular formation is an important factor for the manifestation of reduction of the polysynaptic reflexes by D-40TA. This suggestion is supported by the reports (23-25) that the reticular formation has been demonstrated to have a functionally close relation with the gamma system. However, the possibility that the inhibitory effect on the reticular facilitatory system is a prerequisite for the depression of the polysynaptic reflexes is indicated by the following results obtained in the present experiments. The linguomandibular reflex responses of the spinal preparation were not depressed by benzodiazepine-type agents but augmented in some instances. The depressed linguomandibular reflexes by D-40TA in the anesthetized cat were restored to 70-80 % of control level after spinal transection at Cl level, and thereafter the restored reflexes were unaffected even by the large doses of benzodiazepine-type agents. Furthermore, the reflex potentials from the digastric nerve in response to the afferent stimulation of the lingual nerve in the functionally deafferented preparations such as the gallamine-immobilized cat or the decerebrate cat pretreated with gallamine were also not modified by D-40TA. These results suggest that the depression of the reticular facilitatory system by D-40TA affects in turn an ascending influence via feedback control mechanism from the spinal cord and its related muscle on supraspinal reflex pathway and results in the depression of the supra- D-40TA ON SPINAL REFLEX 95 spinal polysynaptic reflexes. Such an interpretation is supported by the findings that the digastric muscle seems to be devoid of muscle spindle (26), and that the controlling and activating function of the gamma system on the alpha motoneuron is under inhibitory and facilitatory control of the reticular formation (27). For explanation of the inhibition of the spinal polysynaptic reflexes, this interpretation is also favorable, since benzodiazepine-type agents did not affect the reflexes in the preparation which lacked the feedback control mechanism by immobilizing the skeletal muscle with gallamine or the supraspinal control by transecting the spinal cord at Cl level. In the decerebrate cat, D-40TA caused a depression of the spontaneous and evoked discharges of the G1a fiber derived from the muscle spindle, in contrast to no effect on these discharges of the same fiber in the spinal preparation. Moreover, block of the reflexive alpha motoneuronal activity appeared synchronously with that of the gamma mononeuron. These results indicate that D-40TA depresses primarily the reticular formation, affects in turn the gamma activity and thereby reduces the activity level of the polysynaptic neurons. In conclusion, it is suggested that mephenesin-type agents inhibit interneurons at the spinal and supraspinal levels to produce skeletal muscle relaxation. On the other hand, the effect of benzodiazepine-type agents is attributed to an ascending inhibitory action through the gamma system caused by the primary effect on the reticular formation. Information gained from the present experiments suggest the clinical efficacy of D- 40TA for treatment of rigidity. Acknowledgements: We are very grateful to Dr. K. Shimamoto for assistance with the manuscript and to Mr. M. Funado for technical assistance.

REFERENCES 1) BERGER, F.M. AND BRADLEY,W.: Br. J. Pharmacol. Chemother. 1, 265 (1946) 2) BERGER, F.M.: Pharmacol. Rev. 1, 243 (1949) 3) NGAI, S.H., TSENG, D.T.C. AND WANG, S.C.: .1. Pharmacol. exp. Ther. 153, 344 (1966) 4) NAKAJIMA, R., TAKE, Y., MORIYA, R., SAJI, Y., Yui, T. AND NAGAWA, Y.: Japan. J. Pharmacol. 21, 497 (1971) 5) SMITH, C.M., BUDRIS, A.V. AND PAUL, J.W.: J. Pharmacol. exp. Ther. 136, 267 (1962) 6) HUNT, C.C.: J. gen. Physiol. 38, 117 (1954) 7) HUNT, C.C.: J. Physiol. 115, 456 (1951) 8) HENNEMAN,E., KAPLAN, A. AND UNNA, K.: J. Pharmacol. exp. Ther. 97, 331 (1949) 9) KAADA, B.: J. Neurophysiol. 13, 89 (1950) 10) RoszKowsKI, A.P.:J. Pharmacol. exp. Ther. 129, 75 (1960) 11) TRUITT, E.B. AND LITTLE, J.M.: J. Pharmacol. exp. Ther. 122, 239 (1958) 12) WITKIN, L.B. AND SPITALLETTA,P.: Toxicol. Appl. Pharmacol. 2, 151 (1960) 13) GHELARDUCCI,B., LENZI, G. AND POMPEIANO,O.: Arch. int. Pharmacodyn. Thew. 163, 403 (1966) 14) HUNT, C.C. AND PAINTAL,A.S.: J. Physiol. 143, 195 (1958) 15) ELDRED, E., GRANIT, R. AND MERTON, P.A.: J. Physiol. 122, 498 (1953) 16) YOSHINO, T.: Folia pharmacol. japon. 65, 315 (1969) 17) STRATTEN,W.P. AND BARNES,C.D.: Fed. Proc. 27, 571 (1968) 18) ECCLES, J.C., SCHMIDT,R.F. AND WILLIS, W.D.: J. Physiol. 168, 500 (1963) 19) ZBINDEN, G. AND RANDALL,L.O.: Advance Pharmacol. 5, 213 (1967) 96 S. CHIBA & Y. NAGA WA

20) PRZYBYLA,A.C. AND WANG, S.C. J. Pharmacol. exp. Ther. 163, 439 (1968) 21) JIMENEZ-PABON,E. AND NELSON, E.A.: Neurology 15, 1120 (1965) 22) CHIN, J.H. AND SMITH, C.M.: Pharmacol. exp. Ther. 136, 276 (1962) 23) ELDRED, E., GRANIT, R. AND MERTON, P.A.: Acta physiol. scand. 29, 83 (1953) 24) MERTON, P.A.: The Spinal Cord, p. 247, Ciba Foundation Symposium, Churchill, London (1955) 25) TOKIZANE,T.: Kagaku, Tokyo 25, 229 (1955) 26) SZENTAGOTHAI,J.: J. comp. Neurol. 90, 111 (1949) 27) GRANIT, R. AND KAADA, B.: Acta physiol. scand. 27, 130 (1952)