The effects of muscle stretch and vibration on fusimotor activity in the lightly anaesthetised cat

The effects of muscle stretch and vibration on fusimotor activity in the lightly anaesthetised cat

BRAIN RESEARCH 55 T H E EFFECTS OF MUSCLE S T R E T C H A N D VIBRATION ON F U S I M O T O R ACTIVITY IN T H E L I G H T L Y A N A E S T H E T I S E...

868KB Sizes 0 Downloads 15 Views

BRAIN RESEARCH

55

T H E EFFECTS OF MUSCLE S T R E T C H A N D VIBRATION ON F U S I M O T O R ACTIVITY IN T H E L I G H T L Y A N A E S T H E T I S E D CAT

U. PROSKE* ANDD. M. LEWIS Department of Physiology, The Medical School, University Walk, Bristol (Great Britain)

(Accepted May 18th, 1972)

INTRODUCTION An indirect but convenient method of measuring the activity of fusimotor neurones is to record from afferents of muscle spindles 7. In a recent paper, Lewis (in preparation) using this technique found, in lightly anaesthetised animals, that the size of the reflex response of fusimotor neurones to cutaneous stimulation depended on muscle length. He observed in several preparations a reduction in the peak of the reflex activity recorded by a spindle if the muscle containing the spindle was stretched to near its maximum body length. Bessou et al. 3 demonstrated that the peak rate of firing of a spindle in response to fusimotor stimulation, with the ventral roots cut, increased with increase in muscle length over the whole range of normal body lengths. Lewis concluded that the reduction in the reflex amplitude he had observed at long muscle lengths might be attributable to the action of stretch sensitive afferents within the muscle. Observations concerning the effect of muscle stretch on the activity of fusimotor neurones have remained controversialS, 9. The experiments reported here examine whether the reduction of the fusimotor reflex activity at long muscle lengths can be explained in terms of autogenetic mechanisms. METHODS The preparation

The experiments were carried out on cats weighing between 1.5 and 2.8 kg (mean 2.2 kg). Anaesthesia was induced by an initial intraperitoneal injection of pentobarbitone sodium in a dose of 40 mg/kg, and maintained by further intravenous injections of 0.25-1 ml of a solution containing 6 mg/ml. During the recording proce* Wellcome Research Fellow. Present address: Department of Physiology, Monash University, Clayton, Victoria 3168, Australia. Brain Research, 46 (1972) 55-69

56

U. PROSKE A N D D. M. LI!WIS

dure sufficient additional anaesthetic was given to preserve a brisk response to pinching a forepaw without producing general arousal. No recordings were made immediately after an injection of anaesthetic. The medial gastrocnemius muscle and its tendon were dissected free, which involved section of muscle fibres of lateral gastrocnemius where the two muscles were fused. Care was taken to preserve a vascular connection between the muscles at this point. In experiments involving lateral gastrocnemius-soleus the tendon and nerves of plantaris were cut but the muscle itself was not separated. Before cutting the tendon, markers were placed to allow identification of the lengths in the body corresponding to maximum flexion and extension of the ankle joint with the knee joint held at 90 ° . The nerves to medial gastrocnemius and when necessary lateral gastrocnemius-soleus were freed from the sciatic for a length of 20 m m and other limb nerves were cut. The upper and lower ends of the tibia and fibula were rigidly fixed to a metal table. The skin flaps overlying the dissected muscles were used to retain a pool of mineral oil warmed to 37 °C, which covered exposed tissues. The muscle tendons were fixed firmly to an electromagnetic length servo via a tension recorder. Details of these methods have been described by Lewis and ProskelL Dorsal roots L5 to $2 were exposed by laminectomy and covered by warmed mineral oil. A subdivision of the L6 or L5 dorsal root was cut and the central portion stimulated by 0.1 msec pulses of l V amplitude. A second thin strand of dorsal root fibres was cut in the L7-SI region and the distal end split into fine filaments for the isolation of single primary or secondary afferent fibres from medial gastrocnemius.

Stimulating procedures Two different ways of tonically activating muscle afferents were employed: stretch and vibration. Since vibration activated spindles within the muscle being vibrated, this procedure was used with a muscle showing a close functional synergism to the test muscle. Thus in the first series of experiments, the effects of stretch and vibration of lateral gastrocnemius-soleus were examined on the background and reflex evoked activity of fusimotor neurones supplying a spindle in the medial gastrocnemius muscle. The second experimental approach used a measure of fusimotor activity in medial gastrocnemius during stretch of that muscle. (a) Stretch. The term stretch as used in these experiments is meant to imply static stretch. The muscle was stretched to a number of lengths covering a range within the m a x i m u m and minimum body lengths for the muscle. At each of these lengths the muscle was held constant for 1 min before measurements of background firing and reflex responses from the test spindle were commenced. Throughout the period of stretch of the synergist muscle, the test muscle was held at a length midway between m a x i m u m and minimum body length. (b) Vibration. The experiments of Brown et al. 4 have demonstrated that vibration of a muscle at an appropriate amplitude and frequency is a powerful stimulus for the primary endings of the muscle spindles. If the primary endings of spindles are involved in modifying the reflex activity of fusimotor neurones then vibration of

Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATION ON GAMMA ACTIVITY

57

spindles might be a sufficiently powerful stimulus to produce measurable changes in the reflex discharges of fusimotor neurones. To test this, lateral gastrocnemius-soleus was vibrated for varying periods at 300 Hz using peak to peak amplitudes of 10-100 #m, while recording from a spindle in medial gastrocnemius. In decerebrate cats, vibrating a muscle produces powerful reflex contraction of extrafusal fibres in the muscle itself and in its synergists. Since the animals used in our experiments were under light Nembutal anaesthesia, this reflex was depressed. The depression was sufficiently profound for vibration to produce only a weak nonsustained contraction in the vibrated muscle and no contraction in its synergist. The same held true for the reflex activation of fusimotor neurones during cutaneous nerve or dorsal root stimulation. Thus any extrafusal effects on the responses of spindles were avoided (cf. Wuerker and Henneman14). In spite of the depression of firing of alpha-motoneurones, fusimotor reflex activity could be readily provoked in the test spindle by stimulating a cutaneous nerve or dorsal root filament. The high firing rates achieved during the reflex response of the spindle would require powerful and synchronous activation of several fusimotor neurones. It was concluded that the level of anaesthesia employed did not significantly depress reflex activity of fusimotor neurones.

Criteria for Jusimotor activity The reflex activity of fusimotor neurones, as represented by changes in firing of the spindle, was measured either in the absence of any external stimulus (background activity) or following an electric shock applied to a skin nerve or dorsal root filament. (a) Background activity. The high frequency bursts of firing of the spindle in the absence of electrical stimulation (Fig. 1) were considered to be related to the fusimotor activity and a measure of the frequency distribution of action potentials from the spindle represented the amount of such activity. (b) Reflex response to stimulation. A single shock applied to a skin nerve evokes in an animal under light anaesthesia a high frequency burst of action potentials from spindles of flexor and extensor muscles without producing any extrafusal contraction. The magnitude and duration of the spindle discharge suggests an almost synchronous reflex activation of several fusimotor fibres (Lewis, in preparation). While the effects of stretch and vibration of the synergist muscle were examined on the reflex activation of a spindle in the test muscle following a shock to the sural nerve, this procedure was not used when measuring the effect of stretch of the homonymous muscle. Here, rather than producing fusimotor reflex activity by stimulating a skin nerve, shocks were applied to filaments of a dorsal root, cut distally. The filaments selected came from the L5-L6 region since here no afferents from medial gastrocnemius were ever encountered. When the amplitude of the reflex change in spindle discharge had been measured at a number of different muscle lengths, the dorsal roots containing the afferents from medial gastrocnemius were cut (L6, L7 and S1) and the reflex versus length measure-

Brain Research, 46 (1972) 55-69

58

U. PROSKE AND D. M. LEWIS

ments repeated. This second set of measurements gave the size of the fusimotor reflex in the absence of inflow from afferents in the stretched muscle.

Data presentation Spikes recorded from afferent nerves were used to trigger an instantaneous frequency meter 11 the output of which was displayed on a storage oscilloscope. By the superposition of a number of responses to single stimuli, 'frequencygrams' were constructed 3. In addition, 'average frequencygrams' were measured on-line using a general purpose digital computer. This was programmed to measure the interval between action potentials and calculate the corresponding instantaneous frequency. A number of responses, typically 20, could be averaged by dividing the time after the stimulus into a number of collecting periods (usually 500 each of 1 msec). When a spike occurred in a time bin the corresponding instantaneous frequency was summed with the previous contents. At the end of a series the averaged response was presented for measurement of amplitude and time course by means of a cursor which could be made to scan the display. Thus the peak of such an averaged response represents the peak instantaneous frequency averaged over 20 sweeps. The horizontal and vertical scales and bin size could be altered for convenient measurement. In a period before the stimulus various parameters of the background discharge were measured. In most experiments this period was set at 3 sec. The mean and root mean square deviation of instantaneous frequencies were calculated during each cycle. These were averaged over the period of measurement prior to stimulation and were shown at the beginning of each display by 3 rows of dots, representing the mean plus and minus the root mean square (see Fig. 3). A distribution of instantaneous frequencies was also assembled by grouping the frequencies in 39 bins of 10 Hz and a further bin representing all frequencies above 390 Hz. This distribution of frequencies was printed out and later displayed graphically. RESULTS

In initial experiments activity was measured only for the primary endings of spindles. A total of 20 primary endings with a range of conduction velocities from 79 to 107 m/sec were isolated in dorsal root filaments. The effect of muscle stretch was tested on both primary and secondary endings. Five secondary endings with a range of conduction velocities between 35 and 51 m/sec were isolated.

Stretch and vibration of the synergbt muscle (1) Effects on the background activity An example of the background activity from a spindle in the medial gastrocnemius muscle is shown in Fig. 1. The irregularities in the discharge rate of the spindle are related to fusimotor activity (see Matthews and Steinla). The frequency of firing reached during the larger bursts of activity might imply that several fusimotor fibres Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATIONON GAMMAACTIVITY

59

0.5s 300Hz

Fig. 1. The discharge of a single primary ending from a medial gastrocnemius spindle, showing the background activity in the absence of stimulation. Each action potential is represented by a dot whose vertical displacement is proportional to the reciprocal of the time interval between it and the previous action potential, i.e. the instantaneous frequency. The dots in the row at the bottom are at 500 msec intervals. The vertical bar (right) represents the frequencies 0-300 Hz, the bottom of the bar being 0 Hz.

are firing synchronously while their brief time course suggests the fibres to be static (Lewis, in preparation). (a) Stretch. Stretch is not as specific a form of stimulation as vibration. The stretch would be expected to activate secondary spindle afferents and tendon organs as well as primary spindle afferents so that any effect of stretch could not be attributed to spindle primaries alone. When the lateral gastrocnemius-soleus muscle was stretched to a number of lengths, no effect could be detected from the background activity of the primary endings of muscle spindles in the medial gastrocnemius. At the longest lengths of the lateral muscles a small increase in the mean level of activity of the medial gastrocnemius spindle was recorded, but since this effect persisted after section of the nerve to the lateral muscles it was considered to be the result of transmission of tension from the stretched muscle. (b) Vibration. Fig. 2B represents the response of a spindle in medial gastrocnemius during vibration of the lateral muscle for 5 sec at 300 Hz and 15 # m amplitude. The record represents 10 superimposed traces. The vibration clearly reduces the bursts of firing that are evident prior to vibration, almost down to the base line level. It may be noted that the effect has a slow onset and persists for at least 1 sec after the end of the vibration. As a control, the nerve to the lateral muscle was sectioned and the vibration experiment repeated. The result is shown in Fig. 2A and reveals the vibration now to be ineffective. While such a clear cut effect was seen in several preparations it could not always be demonstrated. Even in a single animal, under certain conditions vibration was ineffective in suppressing the bursts. Such a condition was a sudden increase in the background fusimotor activity associated with coughing. It was concluded that the central nervous input to fusimotor neurones could under certain conditions completely mask any reflex effects of the vibration. No sign of spread of vibration to the test muscle could ever be detected in these experiments. It required a high resting tension on the vibrated muscle before such spread would occur, and this was prevented by holding the muscle at a length corresponding to its o p t i m u m body length.

Brain Research, 46 (1972) 55-69

60

U.

PROSKE

AND

D.

M.

LEWIS

A

• ....'.,' r~.

~

:;~i~.;~~~":':"".~"""'

. , ~ .:" . . ~ . . . . ,. •.... :'.... ~ . ,

" ....

....



.

~;'" " :'"~.~':"°" ¢ "~"

B

......................

~1 . . . . . . . .

r

0.5s 300Hz Fig. 2. The effect of vibration on the background activity of a single primary afferent from a spindle in medial gastrocnemius. As in Fig. 1, instantaneous frequency of the spindle discharge is displayed but 10 successive oscilloscope sweeps have been superimposed in each record. The time calibration below each record represents 500 msec intervals while the vertical bar on the right of the figure represents the frequencies 0-300 Hz. In record B lateral gastrocnemius-soleus was vibrated at 300 Hz with an amplitude of 15/~m for the period indicated by the horizontal bar. Record A shows the response after section of the nerve to lateral gastrocnemius-soleus, vibration being applied during the same portion of the trace as in the lower record. (2)

E f f e c t s on t h e r e f l e x r e s p o n s e to c u t a n e o u s s t i m u l a t i o n

Quite weak volleys applied to the sural nerve produced a measurable change in the spindle discharge. Larger volleys were necessary for maximal and consistent reflex activation of spindles. An example of the response of a spindle in the medial gastrocnemius muscle to stimulation of the sural nerve is shown in Fig. 3A. The size of the stimulus intensity was just above threshold for the delta group of fibres. The height of each dot in the record represents the reciprocal of the mean interval between action potentials from 20 superimposed responses. The time during the response has been subdivided into 1 msec bins and the mean intervals between the action potentials in each bin have computed. The mean and root mean square deviation of the firing rate for a period of 3 sec prior to stimulation are represented by the 3 rows of dots on the left of each record A and B. At the bottom of the figure is a calibration marker at 100 Hz. Below the calibration signal and below the records A and B is a line which represents zero frequency. Thus the height of the 10 msec pulses above the line represents 100 Hz. ( a ) S t r e t c h . Stretch of the synergist muscle had no effect on the reflex activity as measured by the test spindle. Since stretch had been ineffective in modifying the background firing, this was not a surprising result. ( b ) V i b r a t i o n . The response illustrated in Fig. 3B was measured during vibration of the lateral gastrocnemius-soleus muscles at a frequency of 300 Hz and 15/~m amplitude. The period of the vibration was 5 sec, the reflex response being measured 4 sec after the onset of the vibration. Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATIONON GAMMAACTIVITY

61

A

.. -.'..

.......~'..'

,-

.



B

...... ..

.

,'~.

C a t 10 m s

Fig. 3. The response of a primary ending from a spindle in medial gastrocnemius to stimulation of the sural nerve. Each of the 3 records A, B and Cal is the output of a digital computer and represents the average of 20 responses. The line below each record is at zero frequency. The 3 rows of dots on the left of records A and B represent the mean :~ the root mean square deviation of the background activity averaged for 3 sec before each stimulus. The time after the stimulus has been divided into 1 msec periods. The mean instantaneous frequency was calculated for all action potentials that occurred in each 1 msec period and was displayed as a spot whose vertical displacement was proportional to the mean. (If no potentials occurred no spot was displayed.) The bottom record (Cal) shows the output to calibration pulses at 100 Hz. Record A represents the response to nerve stimulation in the absence of vibration. In record B the response was measured 4 sec after the onset of a 5 sec period of vibration of lateral gastrocnemius-soleus at 300 Hz and 15 #m.

The reflex response d u r i n g v i b r a t i o n is little different from the control. Its peak is slightly lower a n d the r o o t m e a n square deviation a b o u t the m e a n has become smaller. This is the expected result from a stimulus that reduces the b a c k g r o u n d firing of f u s i m o t o r neurones. The reduction of the peak a m p l i t u d e was n o t a consistent finding a n d it was concluded that v i b r a t i o n of a synergist did n o t significantly modify the f u s i m o t o r reflex in the test muscle. Thus it was concluded that while stretch of a synergist was ineffective in modifying either the b a c k g r o u n d firing or the f u s i m o t o r reflex in the test muscle, vibration of the synergist could reduce the b a c k g r o u n d firing b u t left the reflex unaltered.

Stretch of the homonymous muscle Since stretch of a synergist had been w i t h o u t effect of the fusimotor reflex activity it was decided to explore whether stretch of the test muscle itself modified the activity.

(1) Effects of stretch on the background activity The b a c k g r o u n d activity was m e a s u r e d over a range of muscle lengths corre-

Brain Research, 46 (1972) 55-69

62

U. PROSKE AND

I). M. L E W I S

14 12 10 o x 8

4 2

20

40

60

80

100

120

140

FREQUENCY Hz

Fig. 4. Distribution of frequencies in the background activity of a primary afferent from medial gastrocnemius held at 6 mm below its maximum body length. The frequencies, in classes of 10 Hz are represented on the horizontal scale and the number of spikes counted during a 60 sec period on the vertical scale. The unshaded histogram was obtained with dorsal roots intact, while the activity observed after section of dorsal roots is represented by the shaded areas.

sponding to Lmax - - 14 m m to Lmax, Lmax being m a x i m u m body length. The distribution of frequencies was measured over a 3 sec period at each length. Then the dorsal roots containing the afferent inflow from the muscle were cut and the series of measurements repeated. The 'instantaneous frequencies' were grouped into classes l0 Hz wide up to 390 Hz with a final class for all frequencies above this; the highest frequency accepted for counting was 800 Hz. At each length the mean frequency and root mean square were computed. A sample of the distribution of frequencies for a spindle from medial gastrocnemius with the muscle held at Lmax - - 6 m m is given in Fig. 4. This length corresponded approximately to the muscle's optimum length in the body. The shaded histogram represents the distribution of frequencies after section of L6, L7 and S1 dorsal roots. While before section of dorsal roots the distribution has its peak in the 50-60 Hz range and extends to 150 Hz, after section it has a narrower range and a peak in the next class below. Both distributions appear skewed towards the high frequency end, probably reflecting the occurrence of bursts of action potentials. But it would seem that there are fewer bursts after dorsal root section. In an attempt to compare the number of bursts before and after dorsal root section, the mean frequency corresponding to the 90 ~ level of the distribution was computed and compared at different muscle lengths. This is represented in Fig. 5 for the same spindle. The filled triangles are for values before dorsal root section, the open triangles after section. While there is a drop in the high frequencies at each length after section of the dorsal roots, for most of the range the drop is approximately constant. Thus it has been concluded that muscle length has little or no effect on the distribution of high frequencies. Since the high frequencies are considered to be related to fusimotor activity it is concluded that muscle stretch under these conditions does not significantly modify the background fusimotor firing. The drop in frequency that does occur on dorsal root section is probably related to a reduction of the number of tonically Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATION ON GAMMA ACTIVITY

63

/

140 120 .~

100

z

80 6o

40 20 I

I

I

I

I

I

I

I

14

12

10

8

6

4

2

0

MUSCLE LENGTH

turn below Lma x

Fig. 5. The effect of muscle length on the proportion of high frequencies contained in the background activity of the same spindle as in Fig. 4. Responses shown before (A), and after (A) section of the dorsal roots. The abscissa represents muscle length in mm below maximum length in the body. The ordinate represents the frequency at the 90 ~ level of the distribution of frequencies, i.e. the frequency below which contained 90 % of the distribution of frequencies.

discharging sensory fibres exerting their excitatory effects on the fusimotor neurones but not to stretch sensitive afferents in the test muscle.

(2) Effects of stretch on the fusimotor reflex The amplitude of the fusimotor reflex in response to stimulation of a filament of dorsal root was measured with the muscle held at a series of lengths just as for the measurement of background activity. Since it was known that after ventral root section the response of a muscle spindle to fusimotor stimulation increased at all muscle lengths in the soleus muscle 12, a similar increase in the reflex response with length was predicted here provided that stretch sensitive afferents in the muscle did not modify it. I f any effects were being exerted by such afferents, it should be revealed by section of the dorsal roots containing the afferents. In Fig. 6 are represented curves for the peak and the mean frequencies of firing of the spindle in response to a single shock applied to the central end of an L6 dorsal root filament. The mean (dotted line) was measured over a 3 sec period prior to stimulation. The filled circles represent values before section of the dorsal roots containing afferents from the muscle, the open circles after section. Because of the small differences encountered for the means different symbols were not used to distinguish between them. For lengths Lmax - - 14 to Lmax - - 6 m m the amplitude of the reflex after dorsal root section was reduced at each length, the amount of reduction remaining approximately constant. As for the background activity, this effect can not be accounted for by afferents within the muscle itself but is more likely to reflect a change in excitability of the fusimotor neurones on sectioning the dorsal roots. Brain Research, 46 (1972) 55-69

64

tJ. PROSKE AND D. M. LEWIS

The same explanation cannot be applied to the values of the peak frequency of the reflex when the muscle was stretched to the lengths Lmax - - 4 m m to [-max. At Lmax - - 4 m m and Lmax - - 2 m m the amplitude of the reflex did not change as a result of dorsal root section, while at Lrnax it was actually larger than the pre-section value. Note also that the mean frequency increased at all these lengths. Such an effect of length on the peak of the reflex can not be simply explained in terms of changes in overall excitability, but must invoke the action of stretch sensitive afferents within the muscle itself. Similar evidence for the action of stretch sensitive afferents on reflex amplitude was observed with 5 other spindles. The remainder showed a reduction in the reflex at all muscle lengths after dorsal root section, but no change in the slope of the curve. For two spindles a drop in the peak of the reflex was observed both before and after dorsal root section at the longest lengths. Even the spindle stimulated synchronously by several fusimotor fibres after ventral root section showed a drop in the size of the response at these lengths. It was concluded that such behaviour involved mechanisms at the receptor and could not be attributed to reflex action.

The effect of stretch on the responses from secondary endings of spindles While most of the attention had been paid to primary endings of spindles some 350 O

300

250

£ 200 U Z

150

lOO

50

o~ ....

I 14

---" ....

I 12

MUSCLE

.'--

.-

I 10 LENGTH

.'--

--

I 8

o

I 6 mm

I 4 below

I 2

l

0

Lrnax

Fig. 6. Effect of muscle length on the reflexly evoked increase in firing of a spindle (the same as in Figs. 4 and 5) in response to a single shock applied to the central end of a filament of dorsal root, and on the mean rate of background firing measured during the 3 sec period prior to stimulation. Muscle length in mm below maximum body length is represented on the abscissa. The peak frequency, i.e. the peak instantaneous frequency averaged over 20 sweeps, is plotted as circles (see Methods). The filled circles represent the response before and the open circles after section of dorsal roots containing the afferents from the muscle. The mean rate of background firing, i.e. the mean of instantaneous frequencies measured during the 3 sec and averaged over 20 sweeps (cf. Fig. 3) is repiesented by the small dots. The drop in the mean background firing after dorsal root section was so small that separate symbols have not been used and only a single line drawn to join the points. Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATIONON GAMMAACTIVITY

65

A

..

.'

,

.

...'

."

0-5s 150Hz 13

-..'. •

• • ....



.'.. .. ..........\

.'... ..

lOres.

Fig. 7. Background activity, A, and reflex response to stimulation of the central end of a filament of dorsal root, B, of secondary endings from medial gastrocnemius spindles. Trace A represents the instantaneous frequency of the background activity from one secondary ending. Time calibrations are at 500 msec intervals and the vertical frequency bar represents 0-150 Hz. Trace B represents the reflex response of another secondary ending to stimulating a filament of L5 dorsal root (0.1 msec duration 1.0 V strength). Below it is a calibration signal at 100 Hz. The lines under trace B and under the calibration marks represent zero frequency.

preliminary observations were made on secondary endings. A total of 5 afferents from secondary endings from the medial gastrocnemius muscle were isolated and stretch of the muscle was found to have an effect of 3 on them. Effects of stretch or vibration of a synergist muscle were not examined. Secondary endings of spindles showed irregularities and bursts of firing similar to those from primary endings. Since it is known that secondary endings are only activated by static fusimotor neurones z, these bursts, in the absence of extrafusal contraction, can be attributed to static fusimotor activity. An example of the background fusimotor activity for a secondary ending is shown in Fig. 7A. Note the much more regular firing than for primaries and the large bursts. The distribution of impulse frequencies for secondary endings did not differ from that for primary endings and increased in the same regular manner with muscle length. After dorsal root section the mean rate of firing of secondary endings was reduced as well as the proportion of high frequency bursts. In Fig. 7B is represented the increase in firing of a secondary ending to stimulation of the central end of a filament of L6 dorsal root. Similar reflex responses could be produced by stimulating cutaneous nerves. As for the background activity this response must represent the activity of only static fusimotor neurones. The latency of the response was 25 msec with a time to peak of 8 msec which is somewhat longer than for Brain Research, 46 (1972) 55-69

66

tJ. PROSKE A N D D. M. LEWIS

I

o/-°

I

1

J

I

1

I

10

8

6

4

2

0

MUSCLE

LENGTH m m below Lma x

Fig. 8. Effect of muscle length on the reflexlyevoked increase in firing of a secondary spindle afferent (same as in Fig. 7 record B) in response to a single shock applied to the central end of a filament of dorsal root, and on the mean rate of background firing measured during the 3 sec period prior to stimulation. Muscle length, in mm below maximum body length is represented on the abscissa. The ordinate represents spindle firing rates. The peak frequency (circles) and mean background firing rates (triangles) were measured as for Fig. 6 (see also Methods). Filled symbols before and open symbols after section of dorsal roots containing the afferents from the muscle. the responses from primary endings (see Lewis, in preparation). The peak frequency of firing reached was 170 Hz at this muscle length, Lmax - - 4 mm. A plot of the mean frequency of the spontaneous activity (triangles) and peak of the reflex (circles) before (filled symbols) and after (open symbols) dorsal root section is represented in Fig. 8. While the mean frequency after section is a little lower, the peak from Lnaax - - 7 m m onwards is much larger. The difference in slope between the curves before and after dorsal root section is interpreted as an effect mediated via stretch sensitive afferents within the muscle. While the effect demonstrated in Fig. 8 was the largest observed, two further secondaries showed a similar if less dramatic behaviour. The other two endings examined showed no change in slope after dorsal root section. DISCUSSION The reflex control of movement requires activity in both alpha and fusimotor neurones. While many reflexes involving activity of alpha-motoneurones have been identified, the part played by fusimotor neurones has remained uncertain. In 1951 Hunt s demonstrated in decerebrate cats that stretch of a muscle reduced the fusimotor activity in the muscle. Hunt and Paintal 9 were unable to observe this result in spinal cats, although characterising several other reflexes involving fusimotor neurones. The results of Hunt and Paintal's experiments led to the conclusion that the pattern of Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATION ON GAMMA ACTIVITY

67

reflex activation of fusimotor neurones did not simply mimic that of alpha-motoneurones. Since it has been suggested that much of the fusimotor reflex activity in the preparations used here comes from static fusimotor neurones, the lightly anaesthetised cat might be compared with a decerebrate preparation 1°. This proposition is further supported by the finding that values for the latency and duration of the reflex firing evoked in the gastrocnemius spindles by cutaneous stimulation were found to be similar to values from spindles in a flexor muscle, tenuissimus (Lewis and Proske, unpublished observations). A recent contribution to the question of autogenetic activation of fusimotor neurones has been made by Brown et al. 5. They observed in ventral root filaments a reduction in the background firing of some fusimotor neurones on vibrating the triceps surae muscle. Such inhibition, while possibly the result of direct action of spindle afferents on fusimotor neurones, could also represent recurrent inhibition mediated via alpha-motoneurones6. In the experiments reported here, the inhibition of the background firing of fusimotor neurones in medial gastrocnemius produced by vibrating lateral gastrocnemius-soleus was sometimes sufficiently powerful to abolish all high frequency bursts. The long latency of this depression suggests that it is not brought about by a simple, segmental pathway but may involve supraspinal levels. The possibility might be considered that firing of tendon organs in the vibrated synergist might contribute to the observed depression. Indeed some reflex tension developed in the synergist during the initial period of vibration, but rapidly declined; after the first half second following the onset of vibration no active tension could be detected. This agrees with the findings of Westbury (personal communication) who, while recording under similar conditions of anaesthesia, showed that most motoneurones discharge only briefly in response to muscle vibration. In contrast to the effect of vibration of the synergist muscle, stretch of the synergist had no effect. The explanation for this discrepancy probably lies in the fact that vibration is a much more powerful stimulus of spindle group I terminals than is muscle stretch. Dorsal root section reduced the background activity in most fusimotor neurones as demonstrated by the smaller number of bursts of firing and a reduced reflex response (cf. HuntS). This depression probably arises from a reduction in the tonic cutaneous input to fusimotor neurones after dorsal root section, since the limb denervation did not entirely eliminate all afferents entering the spinal cord via these roots. The reduction in background firing and in the reflex after dorsal root section was constant at all muscle lengths for two-thirds of the spindles tested. If autogenetic effects were being exerted by stretch receptors within the muscle, the depression would be expected to change with length. For the remaining 6 spindles the depression was less marked at long muscle lengths; it seems reasonable to attribute this to the operation of an inhibitory process. Muscle stretch, unlike vibration, activates all of the stretch receptors in the muscle. Thus any explanation of the effect of muscle stretch on fusimotor activity must include a consideration of spindle secondary endings and tendon organs. While the action of these can not be distinguished, it seems more Brain Research, 46 (1972) 55-69

68

U. PROSKE A N D D. M. L E W I S

probable that tendon organs are responsible for this type of autogenetic effect, because of their high threshold to stretch, so that their reflex effectiveness might not be adequate except at the longest muscle lengths. While these comments also apply to reflexly evoked changes in the firing of secondary endings, the same cannot be said for the reflex illustrated in Fig. 8 as here the depression was registered over a wider range of lengths. Since only a few secondary endings were examined, it remains uncertain whether they respond differently from primary endings to muscle stretch. It has been pointed out that the inhibitory action produced by vibrating the synergist muscle or stretching the muscle itself could not be detected in all experiments. This is not surprising in the face of the observed sensitivity of the fusimotor discharge to small changes in the state of the preparation or level of anaesthesia. Furthermore it was repeatedly observed that any change in the proportion of fusimotor activity contributed by descending facilitation (see also Alnaes et al.1), cutaneous and muscle input seemed able to modify the magnitude of the observed reflex. It is concluded that while autogenetic effects on fusimotor reflex activity are detectable, under the experimental conditions employed they are weak and variable. SUMMARY

Responses from primary and secondary endings of muscle spindles were recorded in the medial gastrocnemius muscle of the cat under light pentobarbitone anaesthesia and intact ventral roots. The mean, root mean square deviation and distribution of frequencies of action potentials were measured for each spindle. High frequency bursts of action potentials, observed in the background firing of primary endings of spindles could be inhibited by vibration of a synergist muscle. The reflexly evoked increase in firing of spindles in response to a single stimulus applied to skin nerve or filament of dorsal root was measured over a range of muscle lengths. The reflex response to electrical stimulation as well as the background firing of a spindle were reduced by section of dorsal roots containing the afferent of the test muscle. For 6 of the primary endings of spindles tested the evoked reflex increased in amplitude after section of dorsal roots when the muscle was stretched near to its maximum length in the body. Three of the 5 secondary endings showed an increase in the reflex following dorsal root section over a wider range of muscle lengths. It is concluded that autogenetic effects on the reflex activity of fusimotor neurones can be detected but under the experimental conditions employed they are weak and variable.

Brain Research, 46 (1972) 55-69

STRETCH AND VIBRATIONON GAMMAACTIVITY

69

ACKNOWLEDGEMENTS This work was supported by a Wellcome Research Fellowship. We are also grateful to the Medical Research C o u n c i l of G r e a t Britain for p r o v i d i n g the M o d u l a r I c o m p u t e r used in the experiments. Mrs. Lyn Dowsett has provided c o n t i n u o u s help.

REFERENCES 1 ALNAES,E., JANSEN,J. K. S., AND RUDJORD,T., Fusimotor activity in the spinal cat, Actaphysiol. scand., 63 (1965) 197-212. 2 APPELBERG,B., BESSOU,P., AND LAPORTE,Y., Action of static and dynamic fusimotor fibres on secondary endings of cat's spindle, J. Physiol. (Lond.), 185 (1966) 160-171. 3 BEssOU,P., LAPORTE,Y., AND PAGES,B., Frequencygrams of spindle primary endings elicited by stimulation of static and dynamic fusimotor fibres, J. PhysioL (Lond.), 196 (1968) 47-64. 4 BROWN, M. C., ENGBERG,I., AND MATTHEWS,P. B. C., The relative sensitivity to vibration of muscle receptors of the cat, J. PhysioL (Lond.), 192 (1967) 773-800. 5 BROWN,M. C., LAWRENCE,D. G., AND MATrrmws,P. B. C., Reflex inhibition by Ia afferent input of spontaneously discharging motoneurones in the decerebrate cat, J. Physiol. (Lond.), 198 (1968) 5-7P. 6 ELLAWAY,P. H., Recurrent inhibition of fusimotor neurones exhibiting background discharges in the decerebrate and the spinal cat, J. Physiol. (Lond.), 216 (1971) 419-439. 7 GRA~T, R., AND KAADA,B. P., Influence of stimulation of central nervous structures on muscle spindles in cat, Acta physiol, scand., 27 (1953) 130-160. 8 HuNt, C. C., The reflex activity of mammalian small nerve fibres, J. Physiol. (Lond.), 115 (1951) 456--469. 9 HuNt, C. C., AND PAINrAL, A. A., Spinal reflex regulation of fusimotor neurones, J. Physiol. (Lond.), 143 (1958) 195-212. 10 JANSEN,J. K. S., AND MATrnEWS,P. B. C., The effects of fusimotor activity on the static responsiveness of primary and secondary endings of muscle spindles in the decerebrate cat, Acta physiol. scand., 55 (1962) 376-386. 11 KAY,R. H., A reciprocal time-interval display using transistor circuits, Electron. Engng., 37 (1965) 543-545. 12 LEWlS,D. M., ANDPROSKE,U., The effect of muscle length and rate of fusimotor stimulation on the frequency of discharge in primary endings from muscle spindles in the cat, J. Physiol. (Lond.), 222 (1972) 511-535. 13 MATrrIEWS,P. B. C., ANOSTEIN,R. B., The sensitivity of muscle spindle afferents to small sinusoidal changes of length, J. Physiol. (Lond.), 200 (1969) 723-743. 14 WUERKER,R. B., AND HENNEMAN,E., Reflex regulation of primary stretch receptors via gamma motoneurones in the cat, 3". Neurophysiol., 26 (1963) 539-550.

Brain Research, 46 (1972) 55-69