Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
Forearm oscillation during cooling of the dentate nucleus in the monkey1 J. B. COOKE AND J. S. THOMAS') L)epartment ofPhysiology,I'11ivc.rsity of Western Onturio, London, Onr., Canada N6A SCI Received July 24, 1995
COOKE, J. D., and TIIOMAS, J. S. 1996. Forearm oscillation during cooling of the dentate nucleus in the monkey. Can. J. Physiol. Pharmacol. 54, 430-436. A study was madc of oscillations in arm acceleration in monkeys performing a selfpaced manual stcp-tracking task. Power spectral density analyses of segments of arm acceleration data from normal monkeys containing both flexion and extension movements and intermovement holding periods showed three major peaks at 1-1.5 Hz, 3-5 Hz and 5-9 Hz. Cooling of the dentate nucleus produced a marked increase in the rclativc magnitude of the 3-5 Hz spectral band. The spcctral peak in this frequency range was larger than the other two which were also present during cooling. Autocorrelation functions from long segments of data containing flexion and extension movements and intermovement holding periods showed rcgular periodic variations in both normal and cooled animals. This suggests that the ongoing oscillations were not changed in phase by the occurrence of the self-initiated arnl movements. J. S. 1976. Forearm oscillation during cooling of the dentate COOKE,J. D. et THSM.~S, nucleus in the monkey. Can. J. Physiol. Pharmacol. 54,430-436. UIIC @tudedes oscillations dans lqacc21eration du bras est faite chez lc singc au cours d'un travail manuel accompli par Ctapes au rythmc de ['animal. Les analyscs de densite spectrale de puissance faites sur les rGsultats obtenils par accelkration du bras chez les singes normaux au cours des mouvernents de flexion et d'extension et ail cours des periodes comprises entre les mouvements montrent trois pics majeurs h 1-1.5 Hz. 3-5 Hz et 5.7 Hz. Ida rCfrigCration du noyau dentel6 augmente de faqon marquee i'importance relative de la bande spectrale 3-5 Hz. Le pic spectral dans ces limites dc frCquence, est plus grand que les deux autres, q i ~ isont aussi retrouves au cows du rcfroidisscment. Les fonctiorls d'autocorrClation tirCes de grandes series de rCsultats portant sur les mouvements dc flexion et d'exrension et lcs periodes erltre les niouvcments montrent dcs variations pkriodiques rCgulihres h la fois chez les animaux normailx et trait&. Cela suggkre quc la survenue des mouvements spontanes du bras n'entraine aucun changement de phase des oscillations continues. [Traduit par 1e journal]
Introduction In a study on primates performing self-initiated step-tracking arm movements, oscillations in arm acceleration at 3-5 Hz and 5-7 Hz were observed during normal movements (Brooks ct cal. 19730). In a later study (Conrad et (d. 1975) it was shown that arm oscillations at similar frequencies could be induced during and at the end of movements by application of a f i x e pulse to ABBREVIATIONS: PSD. power spectral density: AC'F, autocorrelation function. 'Supported in part by grants from the hledical Kesearch Council of Canada (MT-4465) and the National Institutes of Health (NS-103 1 1 ) . 'Present address: 1,aboratory of Neural Control, NINDS, National Institutes of Health, Bethesda, hlD. 20014.
the arm. Conrad ct ul. (1975) suggested that these oscillations might arise from loop delays in a supraspinal stretch reillex traversing the motor cortex. That physiological tremor may be influenced peripherally was indicated in studies by Lippold (1966, 1970) who showed that a suddeil mechanical displacen~entof the extended finger reset the phase of ongoing tremor. In the present study a comparison has been madc of 'physiological' tremor and 'cerebellar' tremor as produced by reversible local cooling of the dentate nucleus (Brooks et al. 19736). We have asked two questions: ( I ) what tremor fiequencies are present in the normal and the dentate-cooled monkey; and (2) is ongoing trcmor 'reset' in phasc by self-initiated arm movements. To quantitate the oscillatio~lspres-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
COOKE AND THOMAS
43 1
ent during performance of forearm movements in the monkey, power spectral density functions of arm acceleration records were determined. Spectral peaks in both the 3-5 Hz and the 5-7 Hz bands were observed in normal animals. Reversible dentate dysfunction increased the relative power in the 3-5 Hz band. As indicated by the autocorrelation function, both the 'physiological' tremor and the experimental 'cerebellar' tremor appear independent of the self-initiated movements in that they showed coherence across several movements.
Methods Experimental Arrangenzents The data was obtained from two monkeys (Cebus alhifrons, h420 and M12) which had been trained to perform a manual self-paced step-tracking task. Following initial training in the task, the monkeys were implanted with a head-holder on the skull and a cooling probe sheath placed lateral to the dentate nucleus as described previously (Brooks ct ul. 19736). Following completion of the experimental series the monkeys were killed and their brains were fixed in 10% formalin for later histological verification of the position of the cooling probes. Histological reconstruction of the probe location and temperature distribution have been given elsewhere for one animal (M12, MeyerLohmann ct al. 1945) and are shown for the other animal (A420) in Fig. I . In Fig. 1 it is seen that the probe was positioned dorsal to the caudal portion of the dentate nucleus, passed through the lateral part of the nucleus and was ventral to the rostra1 part of the nucleus. From the estimated isotherms (dashed lines, of 30% and a Fig. 1 ) it inay be seen that a minlim~~m maximum of 100% of the dentate nucleus was cooled to less than 20 " C in the sections show11 when the probe refcrcnce temperature was 10 " C . Only in one section did the 20 " C isotherm include the interpositus nucleus. Task Requirernents The details of the motor task have been described in detail elsewhcrc (Brooks et al. 1973b; MeyerLohmann et rrl. 1945). The n~onkeyswere trained to make alternate flexions and extensions about the elbow, moving a maniprilandum handle into electrotonically determined target zones. The target zones were 10" of arc in widths, their centres were separated by 50" of arc, and they were not bounded by mechanical stops at their outer edges. After holding in one target zone for approximately 1 s, the monkey was required to move to the other target. The movements were 'self-paced'; the animal could move at any time following completion of the 'hold'. Light and sound cues were provided to the monkey when he entered the appropriate target zone and when he had performed a siaccessful movement. Juice reinforcement was given upon successful completion of each movement.
FIG.1. Estimated isotherms for 20 " C in cooled regions of left cerebellum when probe reference temperature was 10 "C. Frontal sections drawn for M20L at levels of 20, 40. 60 and 80% rostral-caudal extent of the dentate nucle~ls.Midline is at zero on the scale below the sections. D, dentate nucleus; IP, interpositus nucleus; P, cooling probc. Data Acquisition Angular position of the handle was obtained from a precision potentiometer mounted in the manipulandum. A torque motor mounted in the manipulandum was used as a tachometer to provide an analog voltage proportional to angular velocity (Schmidt 1973). Thc angular position and velocity of the handle were recorded on a penwriter for visual inspection and also on magnetic tape for later analysis (Honeywell 4600). Drrta A 11~1y,si~ Parts of the experimental record selected for analysis were digitized with a 12-bit A i D converter at an effective sampling rate of 100 Hz (PDP-12, Digital Equipment Corp.. Maynard, MA.). When available the torque motor velocity signal was digitized: otherwise the position signal was used. Acceleration was obtained by digital sequential fifth point differentiation of the torqiie motor velocity signal when it was available. When only the position signal was available it was differentiated twice. Analyses were performed on either a PDP-12 or a PDP-10 computer. Power spectral density analyses and autocorrelation functions were obtained from 20.48-s
CAN. J. PHYSIOL. PHARMACOL. VOL. 54, 1976
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
M 2 0 CONTROL
MI2
CONTROL
95 20 25 8 5 10 15 20 25 (Hz) FREQ ( H z ) FIG.2. Upper traces-Representative control records at normal brain temperature of position. velocity and acceleration, respectively, obtained from two monkeys (left hand traces, h120-22573; right hand traces, M 12-22573). Vertical calibrations represent 50" of position, 200°/ s of velocity and 2000°/s%f acceleration. Horizontal calibration represents 5 s. Lower traces-Representative nor-nlal PSD plots from the same two monkeys obtained from a randomly selected 20.48-s segment of data in the same experimental control periods as the sample movements shown in the upper traces.
0
5
10 FREQ
segments of acceleration data (2048 words at 100 Hz sampling rate). Acceleration was used as it accentuates any tremor or oscillation in position. Sarnple PSD and ACF plots were obtained from several different experimental sessions with each monkey. Each experimental session usually consisted of three cooling and four control periods. PSD and ACF analyses were perfornled on at least three 20.48-s segments of data in each control and cooling period. Each data segment contained both flexion and extension movements as well as inter-movement holding periods. The data presented here are derived from one experimental session with M12 (M12, 22/5/73. control No. 1, cool No. 2 ) and two with M20 (M20,22,/5/73. control No. 2, cool No. 3; 29/6/73, cool No. 2 ) .
Results
Representative records of movements made by
the monkeys are shown in Fig. 2 (upper traces). The monkeys performed regular sequences of alternating flexions and extensions separated by holding periods in the target zones. Regular oscillations are apparent in the acceleration records. In some cases these oscillations produced a series of zero acceleration crossings (e.g. movement No. 2 of M12 in Fig. 2); in others they appeared as low amplitude modulations superi~nposed on a slower, larger swing in acceleration (e.g. movement No. 5 of MI2 in Fig. 2). A representative PSD function obtained from the same control periods as the movement traces is shown for each animal in the lower curves of Fig. 2. The major peak in the PSD at 1-1.5 Hz presumably reflects the slow acceleration pattern associated with movements of the handle from
COOKE A N D THOMAS
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
M 2 0 DENTATE 10 'C
MI2 DENTATE 10
FREQ ( H Z ) FREQ (HZ) FIG.3 . Upper traces-Representative records of position, velocity and acceleration during dentate cooling (10 " C reference temperature) obtained from two monkeys (left hand traces, M20-22573; right hand traces, M12-22573). Calibration of position, velocity, acceleration and time as in Fig. 2 except for acceleration for M12, which represents 2500°/s? Lower tracesPlots of the power spectral density functions obtained from the same two monkeys. The plots were derived from a randomly selected 20.48 s segment of data during the s a n e experimental cooling periods as shown in the movement records in the upper traces.
one target to the other. This is in agreement with the range of mean movement durations (approximately 0.8-1.1 s). Two higher frequency peaks were also usually visible. In M20 for example, there is a secondary peak at about 3.5 Hz and another at about 5 Hz. In MI2 a clear secondary peak, in this case of about the same magnitude as the low frequency peak, is centered at about 5 Hz.
Dentate Cooling Representative records of movements made by the same two aniinals during dentate cooling (Fig. 3) show an increased proportion of movoments with oscillations occurring at 3-5 Hz as described previously (Brooks st al. 1973b). The increase in oscillation magnitude is apparent in the acceleration traces and can even be seen in
the position records. The power spectra changed correspondingly: there is a marked concentration of power in the 3-5 Hz band. The power in this band increased by one to two orders of magnitude during cooling. Both the 1-1.5 Hz and the 5-7 Hz frequency bands were still present. Arc Oscillations 'Reset' bg) SelJrinitiated Moven~e~zts? Oscillations of the finger or the forearm may be evoked by a sudden mechanical displacement (Lippold 1970; Conrad et ul. 1975). Indeed Lippold (1970) has shown that such a sudden displacement produces an oscillation which is phase-locked to the mechanical displacement. Is tremor 'reset' by a self-initiated movement? It is apparent from Fig. 3 that during dentate cooling oscillations at 3-5 Hz increased not only
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
CAW. J. PHYSIOL.
5
0
10
FREB Ma0
0
j 0
i
15
20
PHARMACOL. VOL.
25
0
2
(Hz)
DENTATE 10
54, 1976
4 SECONDS
6
8
m~
/ \
\/I,
i 5
FWEO 10 (1 H5Z ) 2 0
25
4 0
2
SECONDS 4
6
8
FIG.4. Power spectral densities (left hand graphs) and related autocorrelation functions (right hand graphs) obtained from one monkey (M20) under control conditions (upper graphs: hI2022573) and during dentate cooling to a reference temperature of 10 (lower graphs: M2029673). The data were derived from ramdsm%y selected 20.48-s segments of data within the control and the cooling periods and do not cori-cspond lo data segments in prcvioa~sfigures. OC
during the movements but also in the intermovement holding periods, indicating the oscillations during movements and during holding may be related. Indeed, in some cases it appeared as if the oscillations during ~novcment were merely a continuation of the oscillations during holding rather than an oscillatlolm 'triggered' by the nlovement. Support for this observation is provided by the ACF of long data segments. Tf tlac self-initiated arm movements do not affect the phase of the ongoing oscillation, the ACF should show a maintained cyclic variation with a period equal to the inverse of the oscillation frequency. 111
Fig. 4 (upper traces) are shown PSD and ACF plots from a segment of control data from M20. The ACF shows one mqor periodic vririation having a period of approximately 1 s which is maintained over periods of at least 8 s. Since, during the period from which this data was obtained, the interval between movements was about I s it is most probable that in the seg~nent from which the ACF was derived several movements must have been made. This pattcrn in the ACF presumably reflects the regularity with which lnovements were made and the base (about 1 Hz) frequency of thc ixovements themselves; the similarity of the base movement fre-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
COOKE AND THOMAS
quency and movement repetition rate leading t o an A C F pattern of the observed type. A relatively small and regular variation in the ACF is superimposed on the larger variations described above. This small, regular variation has a period of about 175-200 ms and is apparent throughout the record. The power spectrum, :is cxpccted, shows a peak, at around 6 HL.This frequency cornpoilent, if it were maintained unchanged in phasc throughout most of the data segment, would give rise t o such an oscillation in the ACF. Even in those parts of the 4 C F where therc are large changes in its magnitude, variations in the A C F can be observed with a timing suggesting they are a continuation of the small 6 Ilz oscillations. Since these larger swings in the A C F can be attributed t o periods at which movements are in phase with one another, it appears that this 6-Piz resting oscillation continues, unchanged in phase, over time periods in which several movements were made. This suggests that the ongoing tremor in a normal monkey is not 'reset' by the occurrence of the self-initiated movement. The lower traces in Fig. 4 show PSD and A C F plots from data obtained during dentate cooling in M20. The ACF shows cyclic variation with a period of about 225 ms corresponding to the PSD peak at 4-5 Ilr. The cyclic A C F variations are of greater magnitude than those seen during tlae control period, presumably reflecting the relatively greater (approximately 10 times) power in the 3-5 Hz band during cooling as compared with that in the 5-7 Piz band during the control period. The cyclic variation in the '4CF during dentatc cooling is relatively constant for time shifts up t o at least 8 s. Again, it is highly probable that several movements were made within this data segment and thus that the periodicity in the .4CF is not altered by the presence of movements. The amplitude modulation of the A C F may be related to the presence of movements in the data segment.
Discussion The present study has confirmed the existence of oscillation about thc primate elbow joint in two different frequency bands as was reported by Brooks el ul. (1973~).In that study, frequencies of 3-5 Hz were determined for what were called 'steps9 in 'discontinuous' movenlents and
435
of 5-7 Hz for deviations in 'continuous9 movenaents. In the present study, power spectral density analyses of long segments of data. containing both flexion and extension movements and inter-movement holding p-eriods show three major frequency bands in whrch power is contained in the normal monkey: 1-1.5 Hz, 3-5 Hz and 5-7 Hz. The magnitude of the spectral peaks varied: either or both of the higher frequency bands could be present in any given sample. Differelaces in the relativc amplitudes of the various spectral peaks may be due t o variation in the relative numbers of these naovcments of different types in tlac samples analysed. One other possible reason for the presence of the 3-5 Hz and 5-7 H L spectral peaks may be that they arise froin oscillations occurring about the elbow and the wrist. Variation in the frequency of oscillation depending upon the joint about which the oscillation is taking place is known to occur. the frequency of oscillation being greater at distal than at proximal joints (Halliday and Redfearn 1956; Randall and Stiles 1964; Walsh 1960; Fox and Randall 1970). Due t o our experimental arrangement in which the rlinin-mal's forearm was unsupported except at the elbow and hand, it is possible that tremor occurring about both the elbow and wrist could produce movement of the n-manipulanduna handle. What effect do peripherally occurring events have on tremor? Oscillations in the monkey forearm naay be evoked by a sudden mechanical displacement during arm movement (Conrad cr!. 1975; Meyer-kolamann c.t (11. 1975). 14s noted earlier, Lippold (I 969, 1970) demonstrated that finger os~illation in the human may be altered by a sudden displacement of the finger, the oscillations being reset in phase by the displacement. It has also been shown that tremor frequency and magnitude can be changed by imposing loads on the tremoring member (Stiles and Randall 1967; Fox and Randall 1970; Joyce and Rack 1974). It is thus clcar that peripheral events nnay influence or modify limb oscillations. The present experiments suggest however that the ongoing oscillation is not reset in phase by volitional self-initiated moven-ments of the limb. Rather it would appear that the oscillations during such movements are amplitude-modulated continuations of the holding tremor. The arm displacements in the present study are qualitatively different in timc scale from the displace-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
436
CAN. J. PI1YSlOL. PMARMACOL. VOL. 54, 1976
ments produced by the finger taps of kippold (1969, 1970) or the force pulses of Conrad et al. (1975). It may thus be that a rather sudden, synchronous peripheral input is required to reset the phase of ongoing tremor. Dentate dysfunction produced an increase in the magnitude of the 3-5 Hz oscillations. This is in agreement with Brooks et ul. (197%) who found an increased proportion of movements containing 3-5 H a oscillations during dentate cooling compared with those containing a 5-7 Hz oscillation. Since in the present study, both frequency bands were present in the power spectra of normal animals and during dentate cooling, the effect of dentate dysfunction may be to cause a shift in the dominant oscillatioil frequency rather than to bring about a qualitatively different oscillation. It should be noted in addition that the oscillations in arin acceleration and in the modulation of firing frequency of units in the motor cortex as produced by a sudden arm displacement are at the same frequency as the normal tremor in the present work, i.e. 5-7 Hz. In addition, cooling of the dentate nucleus has the same effect on these oscillations as it did on the tremor in the present work. Thus, the oscillations produced by a sudden arm displacement and those occurring naturally may arise f r o n ~ the same mechanisms. as was suggested by kippold (1970) for finger trcmor. MeyerLohmann ct al. (1975) suggested that the oscillation following arm displacement may be caused by titne delays in a long-loop stretch reflex involving the motor cortex. It may be through such a mechanism that the dentate nucleus is involved in determining the dominant oscillatioi~ frequency by, for example, changing loop p i n by altering fusimotor activity (Matthews 1972; Gilman 1973). Acknowledgments The authors wish to thank Professor Vernon Brooks, in whose laboratories this work was done, for the many useful discussions and for criticizing the manuscript. We also thank Dr. Tutis Vilis and Dr. Jon Ilore for the comments on the manuscript.
BROOKS, V. B., COOKE,3. B., and THOMAS. J. S. 1 9 7 3 ~ . The continuity of movements. I n Control of posture and locomotion. Edited by R. 13. Stein, K. G. Pearson, R. S. Smith and J. B. Redford. Plen~rnlPublishing Corp., New York. pp. 257272. BROOKS,V. B., KOZLOVSKAYA, I. B., ATKIN,A., HORVATII,F. E., and UNO,M. 1973b. Effects of cooling dentate nucleus on tracking-task performance in monkeys. J. Ne~rrophysiol.36, 974-995. B., MEYER-LOHMANN, J., MATSUNAMI, J., and CONRAD, BROOKS,V. B. 1975. Precentral unit activity following torque pulse injections into elbow movements. Brain Res. 94, 219-236. Fox, J. R., and RANDALL, J. E. 1970. Relation between forearm tremor and the biceps electromyogram. J. Appl. Physiol. 29, 103-108. GILMAN,S. 1973. A cerebello-thalamo-cortical pathway controlling fusimotor activity. I n Control of posture and locomotion. Edited b y K. B. Stein, K. C . Pearson, R. S. Smith and J. B. Redford. Plenum Publishing Corp., New York. pp. 309330. J. W. T. 1956. An HALLIDAY, A. M., and REDFEARN, analysis of the frequencies of finger tremor in healthy subjects. J. Physiol. (London), 134, 600611. JOYCE,G. C., and RACK,P. M. H. 1974. The effects of load and force on tremor at the normal human elbow joint. J. Physiol. (London), 240, 375-396. LIPPOLD,0. C. J. 1969. Tremor and osciBlation in the stretch reflex arc. J. Physiol. (London), 202, 55P-57P. 1970. Oscillation in the stretch reflex arc and the origin of the rhythmical 8-12 c/s component of physiological tremor. 3. Physiol. (London), 206, 359-382. MATTHEWS, P. B. C. 1972. Mammalian muscle receptors and their central connections. Edward Arnold (Publishers) Ltd., London. MEYER-LOHMANN, J., CONRAD,R., MATSUNAMI. K., and BROOKS,V. B. 1975. Effects of dentate cooling on precentral unit activity following torque pulse injections into elbow nmovements. Brain Res. 94,237-25 1. RANDALL, J. E., and STILES.R. N o 1964. Power spectral analysis of finger acceleration tremor. J. Appl. Physiol. 19, 357-360. SCHMIDT, E. M. 1973. Electrotonically controlled load for monkey manipuIandum. Electroencephalogr. Clin. Neurophysiol. 35, 95-97. STILES,R. N., and RANDALL, 3. E. 1967. Mechanical factors in human tremor frequency. J. Appl. Physiol. 23, 324-336). VV~ALSH, E. G. 1969. Interference with the tremor of F,ukinsonism by the application of a rhythmic force. J. Physiol. (London), 202, 109P-110P.
Forearm oscillation during cooling of the dentate nucleus in the monkey1 J. B. COOKE AND J. S. THOMAS') L)epartment ofPhysiology,I'11ivc.rsity of Western Onturio, London, Onr., Canada N6A SCI Received July 24, 1995
COOKE, J. D., and TIIOMAS, J. S. 1996. Forearm oscillation during cooling of the dentate nucleus in the monkey. Can. J. Physiol. Pharmacol. 54, 430-436. A study was madc of oscillations in arm acceleration in monkeys performing a selfpaced manual stcp-tracking task. Power spectral density analyses of segments of arm acceleration data from normal monkeys containing both flexion and extension movements and intermovement holding periods showed three major peaks at 1-1.5 Hz, 3-5 Hz and 5-9 Hz. Cooling of the dentate nucleus produced a marked increase in the rclativc magnitude of the 3-5 Hz spectral band. The spcctral peak in this frequency range was larger than the other two which were also present during cooling. Autocorrelation functions from long segments of data containing flexion and extension movements and intermovement holding periods showed rcgular periodic variations in both normal and cooled animals. This suggests that the ongoing oscillations were not changed in phase by the occurrence of the self-initiated arnl movements. J. S. 1976. Forearm oscillation during cooling of the dentate COOKE,J. D. et THSM.~S, nucleus in the monkey. Can. J. Physiol. Pharmacol. 54,430-436. UIIC @tudedes oscillations dans lqacc21eration du bras est faite chez lc singc au cours d'un travail manuel accompli par Ctapes au rythmc de ['animal. Les analyscs de densite spectrale de puissance faites sur les rGsultats obtenils par accelkration du bras chez les singes normaux au cours des mouvernents de flexion et d'extension et ail cours des periodes comprises entre les mouvements montrent trois pics majeurs h 1-1.5 Hz. 3-5 Hz et 5.7 Hz. Ida rCfrigCration du noyau dentel6 augmente de faqon marquee i'importance relative de la bande spectrale 3-5 Hz. Le pic spectral dans ces limites dc frCquence, est plus grand que les deux autres, q i ~ isont aussi retrouves au cows du rcfroidisscment. Les fonctiorls d'autocorrClation tirCes de grandes series de rCsultats portant sur les mouvements dc flexion et d'exrension et lcs periodes erltre les niouvcments montrent dcs variations pkriodiques rCgulihres h la fois chez les animaux normailx et trait&. Cela suggkre quc la survenue des mouvements spontanes du bras n'entraine aucun changement de phase des oscillations continues. [Traduit par 1e journal]
Introduction In a study on primates performing self-initiated step-tracking arm movements, oscillations in arm acceleration at 3-5 Hz and 5-7 Hz were observed during normal movements (Brooks ct cal. 19730). In a later study (Conrad et (d. 1975) it was shown that arm oscillations at similar frequencies could be induced during and at the end of movements by application of a f i x e pulse to ABBREVIATIONS: PSD. power spectral density: AC'F, autocorrelation function. 'Supported in part by grants from the hledical Kesearch Council of Canada (MT-4465) and the National Institutes of Health (NS-103 1 1 ) . 'Present address: 1,aboratory of Neural Control, NINDS, National Institutes of Health, Bethesda, hlD. 20014.
the arm. Conrad ct ul. (1975) suggested that these oscillations might arise from loop delays in a supraspinal stretch reillex traversing the motor cortex. That physiological tremor may be influenced peripherally was indicated in studies by Lippold (1966, 1970) who showed that a suddeil mechanical displacen~entof the extended finger reset the phase of ongoing tremor. In the present study a comparison has been madc of 'physiological' tremor and 'cerebellar' tremor as produced by reversible local cooling of the dentate nucleus (Brooks et al. 19736). We have asked two questions: ( I ) what tremor fiequencies are present in the normal and the dentate-cooled monkey; and (2) is ongoing trcmor 'reset' in phasc by self-initiated arm movements. To quantitate the oscillatio~lspres-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
COOKE AND THOMAS
43 1
ent during performance of forearm movements in the monkey, power spectral density functions of arm acceleration records were determined. Spectral peaks in both the 3-5 Hz and the 5-7 Hz bands were observed in normal animals. Reversible dentate dysfunction increased the relative power in the 3-5 Hz band. As indicated by the autocorrelation function, both the 'physiological' tremor and the experimental 'cerebellar' tremor appear independent of the self-initiated movements in that they showed coherence across several movements.
Methods Experimental Arrangenzents The data was obtained from two monkeys (Cebus alhifrons, h420 and M12) which had been trained to perform a manual self-paced step-tracking task. Following initial training in the task, the monkeys were implanted with a head-holder on the skull and a cooling probe sheath placed lateral to the dentate nucleus as described previously (Brooks ct ul. 19736). Following completion of the experimental series the monkeys were killed and their brains were fixed in 10% formalin for later histological verification of the position of the cooling probes. Histological reconstruction of the probe location and temperature distribution have been given elsewhere for one animal (M12, MeyerLohmann ct al. 1945) and are shown for the other animal (A420) in Fig. I . In Fig. 1 it is seen that the probe was positioned dorsal to the caudal portion of the dentate nucleus, passed through the lateral part of the nucleus and was ventral to the rostra1 part of the nucleus. From the estimated isotherms (dashed lines, of 30% and a Fig. 1 ) it inay be seen that a minlim~~m maximum of 100% of the dentate nucleus was cooled to less than 20 " C in the sections show11 when the probe refcrcnce temperature was 10 " C . Only in one section did the 20 " C isotherm include the interpositus nucleus. Task Requirernents The details of the motor task have been described in detail elsewhcrc (Brooks et al. 1973b; MeyerLohmann et rrl. 1945). The n~onkeyswere trained to make alternate flexions and extensions about the elbow, moving a maniprilandum handle into electrotonically determined target zones. The target zones were 10" of arc in widths, their centres were separated by 50" of arc, and they were not bounded by mechanical stops at their outer edges. After holding in one target zone for approximately 1 s, the monkey was required to move to the other target. The movements were 'self-paced'; the animal could move at any time following completion of the 'hold'. Light and sound cues were provided to the monkey when he entered the appropriate target zone and when he had performed a siaccessful movement. Juice reinforcement was given upon successful completion of each movement.
FIG.1. Estimated isotherms for 20 " C in cooled regions of left cerebellum when probe reference temperature was 10 "C. Frontal sections drawn for M20L at levels of 20, 40. 60 and 80% rostral-caudal extent of the dentate nucle~ls.Midline is at zero on the scale below the sections. D, dentate nucleus; IP, interpositus nucleus; P, cooling probc. Data Acquisition Angular position of the handle was obtained from a precision potentiometer mounted in the manipulandum. A torque motor mounted in the manipulandum was used as a tachometer to provide an analog voltage proportional to angular velocity (Schmidt 1973). Thc angular position and velocity of the handle were recorded on a penwriter for visual inspection and also on magnetic tape for later analysis (Honeywell 4600). Drrta A 11~1y,si~ Parts of the experimental record selected for analysis were digitized with a 12-bit A i D converter at an effective sampling rate of 100 Hz (PDP-12, Digital Equipment Corp.. Maynard, MA.). When available the torque motor velocity signal was digitized: otherwise the position signal was used. Acceleration was obtained by digital sequential fifth point differentiation of the torqiie motor velocity signal when it was available. When only the position signal was available it was differentiated twice. Analyses were performed on either a PDP-12 or a PDP-10 computer. Power spectral density analyses and autocorrelation functions were obtained from 20.48-s
CAN. J. PHYSIOL. PHARMACOL. VOL. 54, 1976
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
M 2 0 CONTROL
MI2
CONTROL
95 20 25 8 5 10 15 20 25 (Hz) FREQ ( H z ) FIG.2. Upper traces-Representative control records at normal brain temperature of position. velocity and acceleration, respectively, obtained from two monkeys (left hand traces, h120-22573; right hand traces, M 12-22573). Vertical calibrations represent 50" of position, 200°/ s of velocity and 2000°/s%f acceleration. Horizontal calibration represents 5 s. Lower traces-Representative nor-nlal PSD plots from the same two monkeys obtained from a randomly selected 20.48-s segment of data in the same experimental control periods as the sample movements shown in the upper traces.
0
5
10 FREQ
segments of acceleration data (2048 words at 100 Hz sampling rate). Acceleration was used as it accentuates any tremor or oscillation in position. Sarnple PSD and ACF plots were obtained from several different experimental sessions with each monkey. Each experimental session usually consisted of three cooling and four control periods. PSD and ACF analyses were perfornled on at least three 20.48-s segments of data in each control and cooling period. Each data segment contained both flexion and extension movements as well as inter-movement holding periods. The data presented here are derived from one experimental session with M12 (M12, 22/5/73. control No. 1, cool No. 2 ) and two with M20 (M20,22,/5/73. control No. 2, cool No. 3; 29/6/73, cool No. 2 ) .
Results
Representative records of movements made by
the monkeys are shown in Fig. 2 (upper traces). The monkeys performed regular sequences of alternating flexions and extensions separated by holding periods in the target zones. Regular oscillations are apparent in the acceleration records. In some cases these oscillations produced a series of zero acceleration crossings (e.g. movement No. 2 of M12 in Fig. 2); in others they appeared as low amplitude modulations superi~nposed on a slower, larger swing in acceleration (e.g. movement No. 5 of MI2 in Fig. 2). A representative PSD function obtained from the same control periods as the movement traces is shown for each animal in the lower curves of Fig. 2. The major peak in the PSD at 1-1.5 Hz presumably reflects the slow acceleration pattern associated with movements of the handle from
COOKE A N D THOMAS
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
M 2 0 DENTATE 10 'C
MI2 DENTATE 10
FREQ ( H Z ) FREQ (HZ) FIG.3 . Upper traces-Representative records of position, velocity and acceleration during dentate cooling (10 " C reference temperature) obtained from two monkeys (left hand traces, M20-22573; right hand traces, M12-22573). Calibration of position, velocity, acceleration and time as in Fig. 2 except for acceleration for M12, which represents 2500°/s? Lower tracesPlots of the power spectral density functions obtained from the same two monkeys. The plots were derived from a randomly selected 20.48 s segment of data during the s a n e experimental cooling periods as shown in the movement records in the upper traces.
one target to the other. This is in agreement with the range of mean movement durations (approximately 0.8-1.1 s). Two higher frequency peaks were also usually visible. In M20 for example, there is a secondary peak at about 3.5 Hz and another at about 5 Hz. In MI2 a clear secondary peak, in this case of about the same magnitude as the low frequency peak, is centered at about 5 Hz.
Dentate Cooling Representative records of movements made by the same two aniinals during dentate cooling (Fig. 3) show an increased proportion of movoments with oscillations occurring at 3-5 Hz as described previously (Brooks st al. 1973b). The increase in oscillation magnitude is apparent in the acceleration traces and can even be seen in
the position records. The power spectra changed correspondingly: there is a marked concentration of power in the 3-5 Hz band. The power in this band increased by one to two orders of magnitude during cooling. Both the 1-1.5 Hz and the 5-7 Hz frequency bands were still present. Arc Oscillations 'Reset' bg) SelJrinitiated Moven~e~zts? Oscillations of the finger or the forearm may be evoked by a sudden mechanical displacement (Lippold 1970; Conrad et ul. 1975). Indeed Lippold (1970) has shown that such a sudden displacement produces an oscillation which is phase-locked to the mechanical displacement. Is tremor 'reset' by a self-initiated movement? It is apparent from Fig. 3 that during dentate cooling oscillations at 3-5 Hz increased not only
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
CAW. J. PHYSIOL.
5
0
10
FREB Ma0
0
j 0
i
15
20
PHARMACOL. VOL.
25
0
2
(Hz)
DENTATE 10
54, 1976
4 SECONDS
6
8
m~
/ \
\/I,
i 5
FWEO 10 (1 H5Z ) 2 0
25
4 0
2
SECONDS 4
6
8
FIG.4. Power spectral densities (left hand graphs) and related autocorrelation functions (right hand graphs) obtained from one monkey (M20) under control conditions (upper graphs: hI2022573) and during dentate cooling to a reference temperature of 10 (lower graphs: M2029673). The data were derived from ramdsm%y selected 20.48-s segments of data within the control and the cooling periods and do not cori-cspond lo data segments in prcvioa~sfigures. OC
during the movements but also in the intermovement holding periods, indicating the oscillations during movements and during holding may be related. Indeed, in some cases it appeared as if the oscillations during ~novcment were merely a continuation of the oscillations during holding rather than an oscillatlolm 'triggered' by the nlovement. Support for this observation is provided by the ACF of long data segments. Tf tlac self-initiated arm movements do not affect the phase of the ongoing oscillation, the ACF should show a maintained cyclic variation with a period equal to the inverse of the oscillation frequency. 111
Fig. 4 (upper traces) are shown PSD and ACF plots from a segment of control data from M20. The ACF shows one mqor periodic vririation having a period of approximately 1 s which is maintained over periods of at least 8 s. Since, during the period from which this data was obtained, the interval between movements was about I s it is most probable that in the seg~nent from which the ACF was derived several movements must have been made. This pattcrn in the ACF presumably reflects the regularity with which lnovements were made and the base (about 1 Hz) frequency of thc ixovements themselves; the similarity of the base movement fre-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
COOKE AND THOMAS
quency and movement repetition rate leading t o an A C F pattern of the observed type. A relatively small and regular variation in the ACF is superimposed on the larger variations described above. This small, regular variation has a period of about 175-200 ms and is apparent throughout the record. The power spectrum, :is cxpccted, shows a peak, at around 6 HL.This frequency cornpoilent, if it were maintained unchanged in phasc throughout most of the data segment, would give rise t o such an oscillation in the ACF. Even in those parts of the 4 C F where therc are large changes in its magnitude, variations in the A C F can be observed with a timing suggesting they are a continuation of the small 6 Ilz oscillations. Since these larger swings in the A C F can be attributed t o periods at which movements are in phase with one another, it appears that this 6-Piz resting oscillation continues, unchanged in phase, over time periods in which several movements were made. This suggests that the ongoing tremor in a normal monkey is not 'reset' by the occurrence of the self-initiated movement. The lower traces in Fig. 4 show PSD and A C F plots from data obtained during dentate cooling in M20. The ACF shows cyclic variation with a period of about 225 ms corresponding to the PSD peak at 4-5 Ilr. The cyclic A C F variations are of greater magnitude than those seen during tlae control period, presumably reflecting the relatively greater (approximately 10 times) power in the 3-5 Hz band during cooling as compared with that in the 5-7 Piz band during the control period. The cyclic variation in the '4CF during dentatc cooling is relatively constant for time shifts up t o at least 8 s. Again, it is highly probable that several movements were made within this data segment and thus that the periodicity in the .4CF is not altered by the presence of movements. The amplitude modulation of the A C F may be related to the presence of movements in the data segment.
Discussion The present study has confirmed the existence of oscillation about thc primate elbow joint in two different frequency bands as was reported by Brooks el ul. (1973~).In that study, frequencies of 3-5 Hz were determined for what were called 'steps9 in 'discontinuous' movenlents and
435
of 5-7 Hz for deviations in 'continuous9 movenaents. In the present study, power spectral density analyses of long segments of data. containing both flexion and extension movements and inter-movement holding p-eriods show three major frequency bands in whrch power is contained in the normal monkey: 1-1.5 Hz, 3-5 Hz and 5-7 Hz. The magnitude of the spectral peaks varied: either or both of the higher frequency bands could be present in any given sample. Differelaces in the relativc amplitudes of the various spectral peaks may be due t o variation in the relative numbers of these naovcments of different types in tlac samples analysed. One other possible reason for the presence of the 3-5 Hz and 5-7 H L spectral peaks may be that they arise froin oscillations occurring about the elbow and the wrist. Variation in the frequency of oscillation depending upon the joint about which the oscillation is taking place is known to occur. the frequency of oscillation being greater at distal than at proximal joints (Halliday and Redfearn 1956; Randall and Stiles 1964; Walsh 1960; Fox and Randall 1970). Due t o our experimental arrangement in which the rlinin-mal's forearm was unsupported except at the elbow and hand, it is possible that tremor occurring about both the elbow and wrist could produce movement of the n-manipulanduna handle. What effect do peripherally occurring events have on tremor? Oscillations in the monkey forearm naay be evoked by a sudden mechanical displacement during arm movement (Conrad cr!. 1975; Meyer-kolamann c.t (11. 1975). 14s noted earlier, Lippold (I 969, 1970) demonstrated that finger os~illation in the human may be altered by a sudden displacement of the finger, the oscillations being reset in phase by the displacement. It has also been shown that tremor frequency and magnitude can be changed by imposing loads on the tremoring member (Stiles and Randall 1967; Fox and Randall 1970; Joyce and Rack 1974). It is thus clcar that peripheral events nnay influence or modify limb oscillations. The present experiments suggest however that the ongoing oscillation is not reset in phase by volitional self-initiated moven-ments of the limb. Rather it would appear that the oscillations during such movements are amplitude-modulated continuations of the holding tremor. The arm displacements in the present study are qualitatively different in timc scale from the displace-
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Santa Cruz (UCSC) on 11/20/14 For personal use only.
436
CAN. J. PI1YSlOL. PMARMACOL. VOL. 54, 1976
ments produced by the finger taps of kippold (1969, 1970) or the force pulses of Conrad et al. (1975). It may thus be that a rather sudden, synchronous peripheral input is required to reset the phase of ongoing tremor. Dentate dysfunction produced an increase in the magnitude of the 3-5 Hz oscillations. This is in agreement with Brooks et ul. (197%) who found an increased proportion of movements containing 3-5 H a oscillations during dentate cooling compared with those containing a 5-7 Hz oscillation. Since in the present study, both frequency bands were present in the power spectra of normal animals and during dentate cooling, the effect of dentate dysfunction may be to cause a shift in the dominant oscillatioil frequency rather than to bring about a qualitatively different oscillation. It should be noted in addition that the oscillations in arin acceleration and in the modulation of firing frequency of units in the motor cortex as produced by a sudden arm displacement are at the same frequency as the normal tremor in the present work, i.e. 5-7 Hz. In addition, cooling of the dentate nucleus has the same effect on these oscillations as it did on the tremor in the present work. Thus, the oscillations produced by a sudden arm displacement and those occurring naturally may arise f r o n ~ the same mechanisms. as was suggested by kippold (1970) for finger trcmor. MeyerLohmann ct al. (1975) suggested that the oscillation following arm displacement may be caused by titne delays in a long-loop stretch reflex involving the motor cortex. It may be through such a mechanism that the dentate nucleus is involved in determining the dominant oscillatioi~ frequency by, for example, changing loop p i n by altering fusimotor activity (Matthews 1972; Gilman 1973). Acknowledgments The authors wish to thank Professor Vernon Brooks, in whose laboratories this work was done, for the many useful discussions and for criticizing the manuscript. We also thank Dr. Tutis Vilis and Dr. Jon Ilore for the comments on the manuscript.
BROOKS, V. B., COOKE,3. B., and THOMAS. J. S. 1 9 7 3 ~ . The continuity of movements. I n Control of posture and locomotion. Edited by R. 13. Stein, K. G. Pearson, R. S. Smith and J. B. Redford. Plen~rnlPublishing Corp., New York. pp. 257272. BROOKS,V. B., KOZLOVSKAYA, I. B., ATKIN,A., HORVATII,F. E., and UNO,M. 1973b. Effects of cooling dentate nucleus on tracking-task performance in monkeys. J. Ne~rrophysiol.36, 974-995. B., MEYER-LOHMANN, J., MATSUNAMI, J., and CONRAD, BROOKS,V. B. 1975. Precentral unit activity following torque pulse injections into elbow movements. Brain Res. 94, 219-236. Fox, J. R., and RANDALL, J. E. 1970. Relation between forearm tremor and the biceps electromyogram. J. Appl. Physiol. 29, 103-108. GILMAN,S. 1973. A cerebello-thalamo-cortical pathway controlling fusimotor activity. I n Control of posture and locomotion. Edited b y K. B. Stein, K. C . Pearson, R. S. Smith and J. B. Redford. Plenum Publishing Corp., New York. pp. 309330. J. W. T. 1956. An HALLIDAY, A. M., and REDFEARN, analysis of the frequencies of finger tremor in healthy subjects. J. Physiol. (London), 134, 600611. JOYCE,G. C., and RACK,P. M. H. 1974. The effects of load and force on tremor at the normal human elbow joint. J. Physiol. (London), 240, 375-396. LIPPOLD,0. C. J. 1969. Tremor and osciBlation in the stretch reflex arc. J. Physiol. (London), 202, 55P-57P. 1970. Oscillation in the stretch reflex arc and the origin of the rhythmical 8-12 c/s component of physiological tremor. 3. Physiol. (London), 206, 359-382. MATTHEWS, P. B. C. 1972. Mammalian muscle receptors and their central connections. Edward Arnold (Publishers) Ltd., London. MEYER-LOHMANN, J., CONRAD,R., MATSUNAMI. K., and BROOKS,V. B. 1975. Effects of dentate cooling on precentral unit activity following torque pulse injections into elbow nmovements. Brain Res. 94,237-25 1. RANDALL, J. E., and STILES.R. N o 1964. Power spectral analysis of finger acceleration tremor. J. Appl. Physiol. 19, 357-360. SCHMIDT, E. M. 1973. Electrotonically controlled load for monkey manipuIandum. Electroencephalogr. Clin. Neurophysiol. 35, 95-97. STILES,R. N., and RANDALL, 3. E. 1967. Mechanical factors in human tremor frequency. J. Appl. Physiol. 23, 324-336). VV~ALSH, E. G. 1969. Interference with the tremor of F,ukinsonism by the application of a rhythmic force. J. Physiol. (London), 202, 109P-110P.
