feedback of isotonic muscle contraction on neuromuscular impulse transmission in the frog

Pfliigers Arch. 355, 291--306 (1975) �9 by Springer-Verlag 1975

Feedback of Isotonic Muscle Contraction on Neuromuscular Impulse Transmission in the Frog*

D. L. Ypey

Department of Physiology, University of Leiden, Leiden, The Netherlands

Received September 30, 1974

Summary. The effect of isotonic contraction on neuromuscular impulse transmission was studied in the in vivo M. gastrocnemins preparation of the frog. The N. ischiadicus was stimulated regularly with single pulses at different frequencies in the range of 1/8 to 8 Hz. When steady state conditions were reached, each half- minute a second stimulus was added to one of these pulses. The interval between the pulses of the pair was varied within the contraction cycle of the first stimulus. Compound extraeellular or single intracellular action potentials were recorded from the muscle.

At frequencies of 1/2--2 Hz a depression of the amplitude of the second compound muscle action potential of up to 50% of the first response was found when the muscle contracted isotenieally. However, the second response was facilitated during an isometric contraction. The time course of the depression was equal to that of muscle shortening during the twitch, while the time course of facilitation corresponded roughly to that of facilitation of transmitter release. At lower frequencies the "isotonic depression" or "isometric facilitation" was not or only slightly present. However, at 1/8 Hz the depression could be evoked or increased by curarization. At frequencies higher than 2 Hz facilitation dominated over depression under isotonic conditions.

With flexible intracellular micro-electrodes it was shown that the depression of the amplitude of the compound muscle action potential observed during the isotonic twitch was due to a reduction in neuromuscular impulse transmission.

I t is concluded that the isotonic depression is a negative feedback effect of the change of length of the contracting muscle on s3maptie impulse transmission, probably due to an effect of length on transmitter release.

Key words: Neuromuscular Transmiss ion- Isotonic Cont rac t ion- Muscle Length -- Feedback.

A slight increase of muscle length in the physiological range m a y cause a drast ic increase of neuromuscula r impulse t ransmiss ion in in vivo frog muscle (u et al., 1974). I t has been argued t h a t this phenomenon is due to an effect of a change of length of the presynapt ic nerve endings on t r ansmi t t e r release (Hur te r and Trautwein , 1956), resul t ing in a change in the n u m b e r of endpla te potent ia ls which reach the muscle

* This study was partially supported by the Netherlands Organization for Pure Research (Z.W.O.).

292 D.L. Ypey

fibre threshold potential . The observat ions presented by these authors are suggestive for the existence of a feedback mechanism for the control of muscle contract ion, which is located in the nerve muscle junct ions. This (local) mechanism was expected to func t ion dur ing an isotonic, b u t no t dur ing an isometric contract ion (Ypey et al., 1974).

I n this paper the existence of such a feedback mechanism is shown: 1Neuromuscular impulse t ransmiss ion is depressed dur ing isotonic muscle

contract ion, b u t no t dur ing isometric contract ion.

Methods The in vivo ischiadicus-gastroenemius preparations of deeerebrate frogs (Rana

temporaria) were used at room temperature as described before (Ypey et al., 1974)- the proximal head of the free dissected muscle remained attached to the mechani- cally immobilized leg, while the blood supply to the muscle was kept intact. For recording under isotonic conditions the Achilles' tendon was connected to a hanging weight of about 10 grams. This load was chosen in such a way that the length of the muscle was about equal to the maximal length in situ. The nerve was stimulated with supramaximal pulses at frequencies in the range of 1/8--8 Hz. In the experiments in which the influence of stimulus frequency was investigated, each frequency was first applied for 10 rain to reach steady state conditions (Ypey e~ al., 1974). Then, in the subsequent 10--20 rain, the response to pulse pairs was measured. These pairs were generated by adding once per half-minute an additional stimulus to one of the stimuli of the regular pulse train at that frequency. The interval between the pulses of the pair was varied within the contraction cycle of the first stimulus. This procedure was repeated in one preparation for up to seven increasing frequencies. Isotonic contractions were recorded with a Sanborn model 7 DCDT-050 displacement transducer. During stimulation isotonic conditions could be changed to isometric conditions by clamping the tendon.

Extraeellular Recording. With a metal needle electrode in the Achilles' tendon a biphasie positive-negative compound muscle action potential was recorded from the gastrocnemius, which was bathed in paraffin oil. The amplitude of the positive peak of this signal was used as a measure for the number of muscle fibre action potentials triggered at the neuromuscular junctions by the endplate potentials. Details of this recording method have been described previously (Ypey et aL, i974). Errors in the measurement of the compound action potential amplitude were in general less than 60/0.

Intracellular Recording. Flexible micro-electrodes were used to record intracellular action potentials from muscle fibres of a contracting M. gastrocnemius, bathed in frog l~inger. To prepare floating microelectrodes Woodbury and Brady's method (1956) was slightly modified, in part as described by Coraboeuf (1969): Flexible silver wires of 1.5 cm length and 25 ~m diaraetor were soldered on copper wires of 5 cm length and 1 mm diameter. The terminal 0.5 cm part of the silver wire was coated with AgC1 and inserted into an electrode tip, which had been broken off from a 2.5 M KC1 filled micro-pipette. A firm connection with the silver wire was made by the application of glue (cyanolith) on the opening of the pipette. The copper wire, which could be bent to an angle for easy penetration was fixed to a mieromanipulator and made contact with a Grass micro-electrode amplifier (P 16). The tip resistance of the electrode was 5--10 M~. Muscle fibres were penetrated close to the proximal attachment of the muscle to the knee joint where connective

Feedback of Isotonic Contraction on Neuromuscular Transmission 293

tissue is thin and movement of the muscle, due to contraction or to applied changes of length, is minimal. I t was often possible ~o record stable resting membrane potentials and muscle fibre action potentials for periods of up to 10 rain of repetitive stimulation and contraction.

Results

Depression o/ the Compound Muscle Action Potential during Isotonic Contraction

Fig. 1 A shows the phenomenon which has been discovered: a depression of the compound muscle action potential amplitude during isotonic contraction. During 1 t tz stimulation with single pulses under steady state conditions both the compound muscle action potential and the subsequent isotonic twitch cycle were recorded. The duration of the (isotonic) contraction cycle was about 90 msec; its maximal shortening occurred at about 50 msec. Every half-minute one stimulus of the regular 1 t tz pulse train was followed by a second stimulus with a variable interval (see methods).

As is clear from Fig. 1 A, a depression of the compound action potential amplitude was observed for the second response. I t s t ime course is equal to tha t of muscle length during the twitch cycle of the first response. The maximal amplitude reduction of the compound action potential is about 500/0 of the first response and it occurs a t the t ime of maximal shortening (50 msee). I t is known (Ypey et al., 1974) tha t an applied decrease of muscle length of about 1 m m decreases synaptic impulse transmission, as measured from the compound action potential amplitude, ra ther strongly. We expected, therefore, tha t the observed depression is a feedback effect of muscle shortening during contraction on neuromuscular impulse transmission. This implies tha t contraction under ideal isometric conditions does not influence the compound muscle action potential, which was found to be the case (Fig. 1 B): When the muscle was clamped at a fixed length no decrease (depression) of the amplitude was observed during the twitch, but an interval dependent increase (facilitation). Sometimes a small depression was present in the facilitation curve (cf. Fig.dA). I t is probably due to some remaining internal shortening in the muscle, because ideal isometric contractions do not occur.

I t can also be observed in Fig. 1 A tha t the second twitch has a larger peak shortening a t an interval of 40 or 60 msec than the first one, although the amplitude of the corresponding compound action potential is reduced to about 500/0 . Therefore, a facilitation mechanism is likely to be involved in the contraction process, which overcompensates for the isotonic twitch dependent depression of impulse transmission at the junctions.

294 D. L. Ypey

2o my L

r

VV

cle shortening

A

~_ i i t i

0 50 100 msec

mV comp. muscle AP amplitude

,0

0,6

1.2

1.8

2,4

mm

B

40 isometric

2O

~ , . ~ / i s o t o n i c

Fig. i . (A) The depression of the amplitude of the compound muscle action potential evoked at different times during the isotonic twitch cycle under 1 Hz steady state conditions. Both the action potentials and the single or double twitches were drawn from superimposed sweep photographs of electrical and mechanical responses to pulse pairs with different intervals. (B) Depression of the compound muscle action potential during the isotonic twitch and facilitation during the isometric twitch. The amplitude of the second response of the pulse pair was plotted against the pulse interval. The dotted line is the average first response amplitude. Both figures are

of the same preparation and have the same time base

A feedback effect of i sotonic muscle con t rac t ion on neuromuscu la r t r ansmiss ion as descr ibed above has nor y e t been r e p o r t e d in t h e l i tera- ture . Such an effect m a y p l ay an i m p o r t a n t role in t he cont ro l of muscle cont rac t ion . Therefore, th is phenomenon , here cal led the " isotonic depress ion" , has been ana lysed fur ther .


~ comp, muscle AP ampl i tude

1co

,~.

I/2

8O

6O

40 -

2O

4 0 I I I I I I I

0 20 40 60 80 100 120 140

st imulus interval (msec)

Fig. 2. The effect of stimulus frequency on the isotonic twitch dependent depression of the compound muscle action potential (AP), as measured with double pulses with a variable interval. Symbols of values at t ---- 0 msee represent the average percentual amplitude values of the first response of the pairs at a certain frequency. At each frequency the first response was constan~ within a range of • 8 % during the variation of the intervals (steady state situation). The 4 and 8 Hz curve are of a different preparation. The average amplitude of the first action potential at

1/8 Hz is taken as 100~

The Effect el Stimulus Frequency on the Isotoniv Depression I t has been shown before (u et al., 1974) that the effect of muscle

length on frog neuromuscular transmission is stimulus frequency dependent. I t is to be expected, therefore, that a feedback effect of isotonic muscle shortening should depend also on the frequency of presynaptie stimulation.

To check this we did the experiment of Fig. 1 at different frequencies within the range of 1/8-- 8 Hz. Fig. ~ shows the result of a representative experiment [eft Ypey and Meyer (1972) for a preliminary communication]. Both at the lowest frequencies used (1/8, 1/4 and 1/2 Hz) ~nd at the highest frequencies (4 and 8 Hz) no or a sm~tl depression-like effect of twitch-shortening is visible, while at the intermediate frequencies of 1 and 2 I-Iz a clear depression can be observed.

296 D.L. Ypey

~ comp, muscle AP ampli tude

100 ~_ . . . . . .

I- ~ �9 �9 1

60

4Q _._ i.

20

0 I I r I ~ I I

0 20 40 60 80 100 120 140

stimulus interval (msec)

Fig. 3. The effec~ of stimulus frequency on the compound action potential response to pairs of stimuli with a variable interval at isometric condition, l~uscle length was kept at the initial length for isotonic contraction (same preparations as in Fig.2). Symbols at t ~ 0 represent the average response amplitudes of the first stimulus of all the pairs at a certain frequency. The steady state amplitude of the first response was constant within a range of =k 10% during the interval experiment

at each frequency

Peak shortening of the twitch decreased in this experiment from 1.80 mm at 1/8 Hz to 0.85 mm at 2 ttz, while the duration of the twitch increased from 80 to 100 mseo. At 4 Hz the twitch peak was only 90/o of the 1/8 Hz peak while the twitch duration was about 200 msec, corresponding to the smaller and longer lasting depression of the action potential at this frequency.

The depression is :not present in a similar set of curves under isometric conditions (Fig. 3 for the same preparations). At the lowest frequencies there is no clear difference between isotonic and isometric contraction. But at the intermediate and higher frequencies facilitation of the second response can be observed. The time course of this facilitation roughly corresponds to that of facilitation of transmitter release (Mallart and Martin, 1967). The isometric facilitation curves of Fig. 3 reflect, therefore, this synaptic property.


These curves for the control situation (isometric instead of isotonic) make it clear tha t the isotonic depression is in fact a depression in a facilitation curve. Sometimes the isometric facilitation curves seemed to have a relative depression in their t ime course, as has already been observed by Bobbert (1969). This depression is minimal in stretched muscles and has the t ime course of the contraction. I t is therefore probably a depression due to internal shortening of a not ideally isometrically contracting muscle.

I t has been shown tha t a certain change of length affects transmission most at the intermediate frequencies (u et al., 1974). I t is also known tha t an increase in stimulus frequency within the range used causes a decrease of t ransmit ter release per pulse in vitro (Capek, 1971) while muscle lengthening causes an increase in the endplate potential amplitude (I-Iutter and Trautwein, 1956). These data can be used to explain the effect of stimulus frequency on the isotonic twitch dependent depression: a reduction of the average endplate potential amplitude by isotonic muscle shortening only results in a reduction of impulse transmission when the mean of the distribution of endplate potentials lies not too far above or below the average threshold. This is probably the case at the intermediate stimulus frequencies (around 1--2 I-Iz). Similarly, it can be explained tha t facilitation of t ransmit ter release in pulse pairs and under isometric conditions cannot increase impulse transmission at the lowest frequencies used because all or most of the endplate potentials are already supra-threshold at those frequencies.

A factor which was expected to play a role in the disappearance of the isotonic depression at higher stimulus frequencies, is a decrease of the twitch amplitude, due to an increased failing of neuromuscular impulse transmission. We observed in all experiments a decrease of twitch amplitude. At about 2 Hz the twitch peak was half maximal. This decrease was often accompanied by an increase of the twitch duration, which was correlated with the prolonged duration of the isotonic depression. At 8 Hz stimulation an isotonic depression could also be evoked by increasing the strength of the twitch with an extra condition- ing stimulus. Such a stimulus reestablishes a strong contraction, due to the process of synaptic facilitation of impulse transmission.

To show the reproducibility of the isotonic depression for different preparations and to demonstrate the dependence of this depression on the mean ratio of impulse transmission in the junctions, all the experiments on stimulus frequency are replotted in Fig.4A. The percentual difference between second and first action potential amplitude for pulse pairs with an interval of 50 msec, has been plotted against the amplitude of the first action potential. The first response amplitude is a measure of the ratio of impulse transmission (u et al., 1974). Both variables

20 Pfl(igers Arch., Vol, 355

O/o

2 0

0 J

I

2oi

_ 4 0

298 D.L. Ypey

AP2- APt 100

APt max

�9 I

I �9

A *

~ 41 �9 �9

2 5 O '

D � 9

0

q30

A

|

5Q 75 m"~ 10O ~ P1 , " .,i. ~;

0 b O0 �9

D

[] D

B

, j . 2 , o _ ~ _ . _ 7 ~ �9 ,~ ~ooo/oa B ' , < , - . - .

o O

�9 O o

~176 c o t , t P % o 0

o % 0 o o _20 ~ % 0 0 o

o o o 0 %o %

%

0 0 _40

0

l~'ig.4A and B. Isotonic depression and isometric facilitation of the response to a second stimulus of a pair as a function of the pereentuaI compound amplitude of the response to the first stimulus of the pair. geplotted from experiments on stimulus frequency [(A) rectangular and triangle symbols] and curarization [(B) rounded symbols]. Open symbols: isotonic conditions. Closed symbols: isometric conditions. Different symbols represent different preparations. AP1 is the amplitude of the first compound muscle action potential of the pair, AP2 is the second one while A P i max

is the maximal measured first response amplitude at 1/8 Hz stimulation

are expressed as a percentage of the m a x i m a l compound ac t ion po ten t i a l ampl i tude , t h a t could be measu red a t 1/8 Hz. I somet r i c condi t ions are compared wi th isotonic ones in the plot .

There is a sca t t e r of depress ion and /o r fac i l i t a t ion values for different p repara t ions , p r o b a b l y due to differences in t he th ickness and in i t ia l l ength of the muscle or in the condi t ion of the frogs. The graph shows, nevertheless , t h a t the m a x i m a l shor tening depress ion occurs a t a r o u n d


75% of the maximal amplitude, while the maximal facilitation under isometric conditions occurs at around 300/0 . The actual depression at 50 msec can be read off in this figure as the difference between isotonic and isometric values at the same first response amplitude percentage.

The observed effect of the amplitude and duration of the isotonic twitch on the degree and duration of the depression further supports the interpretation of the depression phenomenon as a feedback effect of isotonic shortening on neuromuscular impulse transmission. Because stimulus frequency influences the ratio of neuromuscular impulse transmission, the experiments suggest, tha t this feedback effect of isotonic muscle contraction on impulse transmission, depends on the safety factor of impulse transmission. The same dependence probably holds for the effect of facilitation of t ransmit ter release on action potential transmission, as occurs in pulse pairs under isometric conditions.

The E//ect o/Curarization on the Isotonic Depression

Another experimental way to demonstrate the dependence of the isotonic shortening depression on the safety factor of impulse transmission, is by curarization of the endptates. In the experiment of Fig.5 a dose of 0.15 mg flaxedil/g bodyweight was injected intraperitone- ally into a frog at t = 0 min during 1/7.5 Hz stimulation with pairs of stimuli at 50 msec intervals. After every 1 or 2 pulse pairs muscle contraction was changed from isotonic to isometric or vice versa.

Before the injection, the second compound muscle action potential under isotonic conditions is often about equal to, or as in this ease already smaller than the first one. After the injection, the amplitude of both responses declines and gradually the depression of the second response becomes larger. The isotonic twitch amplitude also declines (not shown in the figure), but in general more slowly than the amplitude of the first action potential. However, the depression is absent when the muscle is clamped at its initial length. The isotonic depression first grows to a maximum of 30 ~ of the uncurarized amplitude of the first response at t = 2.5 rain. Then it declines again. For several such curarization experiments the depression was plotted against the percentual amplitude of the first response in Fig.4B. A similar relation as obtained in the experiments on stimulus frequency (ef. Fig. 4A) was found for the depression during curarization. However, during isometric contractions at the lower first response amplitude values, facilitation is absent or is not so strong as in the frequency experiments.

Because the eurarizing agent flaxedil lowers impulse transmission by reducing the endplate potential (Bovet et al., 1959), it also follows from these experiments tha t the isotonic depression, which was inter-

20*

300 D. L. Ypey

m V c o m p . m u s c l e A P a m p l i t u d e

8 0 E [] u a

~ 1 7 6 ii

6 0 i ~176 o ' ' a �9

[ ~176

4 0 ~ o

2 O

0 i i i i

0 2 4

_1C x x x

X _ 2 0

_ 3 0 Xx x X x x

A P~-AP I

i s o t o n i c i s � 9 �9 �9 A P 1

o u A P 2

~e

~e o I

o ~o ~ o

o ~~ o

~ 1 7 6 o ~ e o

~ o ~eilq o o

o ~' '~. ~ . . o

o ~ o o o � 9

X X X X

X

6 8 10

X X X X X X

X X X X

_ _ J miD

1 0 0 ~

AP~ r n o x

Fig.5. The effect of curarization on the isotonic twitch dependent depression. The closed symbols give the responses (APi) to the first stimulus, the open symbols those (APe) to the second stimulus. The interval between the stimuli in the pairs, which are applied at a frequency of 1/7.5 ttz, is 50 msec. Circles: isotonic condition. Squares: isometric condition. Crosses: isotonic depression, defined as (AP2--APi) /

API max. Flaxedil injection at t = 0 rain

preted as the feedback effect of muscle shortening on impulse transmission, depends on the safety factor of impulse transmission, i.e. on the difference between the average endplate potential amplitude and the average threshold potential of the population of muscle fibres. Apart from this fator, the shape of the endplate potential peak-amplitude distribution is expected to play an important role (Ypey et al., 1974).

Intracellular Demonstration o/ the Feedback E//ect o] Isotonic Muscle Contraction on Impulse Transmission

Intracellular measurements were used to show that it is neuromuscular impulse transmission which is depressed during the isotonic twitch. Some introductory experiments were performed to determine the effect of stimulus frequency and muscle length on action potential transmission. Muscle fibres of a frog gastrocnemius contracting isometrically or


isotonically were penetrated with flexible intracellular micro-electrodes. Action potentials were recorded during supramaximal nerve stimulation with single pulses in the frequency range of 1/8--4 I-Iz.

For most fibres it was found that impulse transmission under steady state conditions failed occasionally at frequencies of 1--3 Itz. At lower frequencies transmission was in general 100~ Fig. 6 shows, for example, a fibre in which transmission had not yet failed at 1 tIz. That muscle length is an important factor in impulse transmission (tIutter and Traut- wein, 1956; Ypey et al., 1974) was now also verified: When a fibre was failing and responded only occasionally with an all or none action potential, it was possible to increase the ratio of transmission--often to 100~ stretching the muscle by 1--2 mm, without clearly changing the resting membrane potential or the peak amplitude of the action potential (Edman, 1971). Some fibres were impaled not too far from the junction to observe endplate potentials in the ease of failing spike genera- tion. This shows that changes of firing upon changes of stimulus frequency or muscle length are due to changes of the probability of threshold crossing by the endplate potentials. This preliminary result is in good agreement with our conclusions from previous extraecllular measurements (u et al., 1974).

In the experiment of Fig. 6 the preparation was stimulated with single pulses under isotonic conditions, first at a frequency of 1/2 tIz, then at 1 Hz. At 1/2 Hz transmission never failed, at 1 tIz only occasionally. At both frequencies single pulse stimulation was changed to stimulation with pulse pairs at the same frequency. The interval between the pulses of the pair was 60 msec. After every five pairs the isotonic condition was changed to isometric and vice versa. At 1/2 tIz during isometric contraction both action potentials of the pulse pair never failed, while during isotonic contraction the second impulse sometimes failed. At 1 tIz during isometric contraction both impulses again almost never failed, but during isotonic contraction the second impulse nearly always failed (Fig. 6). In several other fibres this isotonic depression of impulse transmission was found at critical frequencies of 1/2--4 Hz, while under isometric conditions facilitation of impulse transmission was observed for the second impulse of the 60 msec pair. At 1/8 IIz depression and facilitation were, in general, absent. Finally, it was possible to observe in a fibre the interval dependence of the isotonic reduction of impulse transmission, which was observed before (Fig. 1) for the compound muscle action potential.

These intracellular experiments are, therefore, a direct proof that the stimulus frequency dependent isotonic depression and the isometric facilitation of the amplitude of the compound muscle action potential respectively reflect a depression and a facilitation of neuromuscular impulse transmission. I t follows from these direct observations that

302 D. L. Ypey

1/2 Hz stimulation

I I

6 0 m s e c

2 sec

1 Hz

i s o t o n i c

i s o m e t r i c

i s o t o n i c

~ 1 6 m V

Fig. 6. The effect of shortening during the single isotonic twitch (upper and lower records at each figure) on neuromuscular impulse transmission at 2 frequencies of stimulation. Te recorded intracellular muscle fibre action potentials are responses to pulse pairs. The interval between the stimuli in the pair is 60 msec, so that the second response occurs during the twitch of the first. Every 5 pairs, the isotonic condition is changed to isometric and vice versa. Isometric condition is indicated

by white parallel lines (middle record at each frequency)

muscle shortening dur ing isotonic contract ion has a feedback on neuromuscular impulse t ransmiss ion and tha t this effect depends on the safety factor of impulse t ransmiss ion across the junct ions.

Feedback of isotonic Contraction on Neuromuscular Transmission 303

Discussion

The results on single muscle fibres demonstrate tha t the depression of the amplitude of a compound muscle action potential which is evoked during an isotonic twitch, reflects a depression in neuromuscular impulse transmission in the population of synapses. A plausible explanation of this depression of impulse transmission is in the effect of muscle length on t ransmit ter release (Hutter and Trautwein, 1956). Recent intracellular in vitro experiments by us confirm this effect and also exclude the possi- bility tha t muscle length interferes at the frequencies employed with the occurrence of an all or none type of presynaptic block of action potential conduction in the nerve fibre: At supramaximal stimulation with these frequencies failures of endplate potentials never occur.

Whether the change of length during contraction also affects the absolute threshold potential of the muscle fibres or not cannot be concluded from the described experiments. I t is only known tha t the threshold of nerve fibres is independent of stretch of the fibres (Goldman, 1964). Therefore, this point needs further investigation.

I t is possible tha t other factors than a reduction of impulse transmission also play a role in the depression of the compound action potential during isotonic contraction.

One such factor might be a change of impedance of the muscle tissue during contraction (Dubnisson, 1937), causing a change in the extra- cellular action potential amplitude. With a sine wave input for the measurement of the absolute magnitude of the impedance of the muscle between the recording electrodes at the frequency content of the compound muscle action potential (ca. 1 ]~Hz), we found tha t this value is indeed decreased during an isotonic contraction, but not during an isometric contraction. For example in one muscle stimulated at a rate of 1 I tz this depression was 7.5~ for an interval of 50 msec while the depression of the compound potential amplitude at the same interval in the same muscle was 31.5 ~ Because the relation between voltage and applied current was linear within the voltage range of the recorded action potentials, this change in impedance during isotonic contraction is not likely to be the main cause of the compound amplitude depression, as already seemed clear from the intraeellular measurements. Another factor which could cause an amplitude depression during shortening is a desynchronization of the single muscle fibre action potentials, due to a reduction of the peak amplitude of the endplate potentials at the shorter muscle length (Hurter and Trautwein, 1956). In at least two experiments we did not observe a prolonged duration of the depressed compound action potential. I t is however difficult to estimate the relative contribution to the depression of desynchronization and increased

304 D.L. Ypey

failing of action potential transmission without a more complete knowl- edge of the behavior of the distributions of endplate peak amplitudes and of single action potential latencics upon changes of length (u et al., 1974). The only thing, which can be stated is that failing and desynchronization occur at the same time when supra-threshold endplate potentials change to below-threshold. This makes the compound amplitude extra sensitive to changes of the probability of impulse transmission.

One result of interest in this investigation is the conclusion that facilitation of transmitter release, as occurs in the responses to pulse pairs (Mallart and Martin, 1967) can increase the probability of in vivo impulse transmission strongly if a muscle is stimulated under isometric conditions at frequencies higher than about 1 Hz. This means that facilitation of release may be an important factor in the control of impulse transmission, especially in fatigued muscles. Because neuromuscular impulse transmission is often considered to be a process with a high safety factor (Katz, 1966), hardly any attention has been paid in the literature to the functional importance of the facilitation process for the control of neuromuscular impulse transmission. Bobbert (1968) has already shown in a preliminary communication that this safety factor is not always high and that facilitation of neuromuscular impulse transmission occurs in pairs of stimuli 25 msec apart. The time course of this facilitation, as measured in the present paper with double pulses with a variable stimulus interval and for different stimulus frequencies shows that this facilitation must be that of transmitter release (Mallart and Martin, 1967): Facilitation of transmission decays with approximately the same time constant as that of release (ca. 30 msec).

The main finding of the present paper is that nerve impulses, which arrive at the neuromuscular junction at the time that the muscle is shortened due to a contraction, have a lower probability of triggering postsynaptically a muscle fibre impulse than the preceding impulse which caused the contraction. This phenomenon shows that the neuromuscular junctions combine stretch receptor properties with effector properties, as is known for example for smooth muscle fibres of the gut (Burnstock et al., 1963).

Whether the negative feedback effect of contraction on synaptic transmission plays an essential role in the control of frog muscle contraction, depends--as this paper shows--on what kinds of natural activation patterns and mechanical behaviour occur in a specific muscle and on how the safety factor of impulse transmission is influenced by these factors. At around 50 ~ impulse transmission, changes of muscle length are most effective (Ypey et al., 1974). Such ratios of impulse transmission occur in the frog gastrocnemius during fatigue, caused by repetitive use of the


muscle at frequencies of 1--3 Hz. Negative feedback of contraction on transmission can only occur in the free living animal when the muscle shortens while nerve impulses are still arriving at the junctions. How- ever, depression during shortening may be compensated by cumulative facilitation of t ransmit ter release, if a t rain of pulses arrives in the same junction. In addition, facilitatory compensating mechanisms can be located in the contraction process itself (see Fig. 1 A). The muscle spindle system and the actin-myosin system with its force-length relation (Gordon et al., 1966) are two other well known feedback systems functioning in the control of muscle length. Evidence has been obtained tha t muscle length also affects the effectiveness of excitation-contraction coupling (Taylor and Riidel, 1970; Allen et al., 1974). Therefore, several processes within the chain of events leading to contraction, now seem to be sensitive to changes of length.

Negative feedback of changes of length of isometrically and tetani- cMly contracting muscle, functioning for example in the control of posture, could be another important consequence of the stretch receptor properties of the junctions (Kiessling, 1971). Such a feedback system may have a similar function as the muscle spindle system, although the mechanisms of both systems are quite different. I t is striking tha t it was shown more than twenty years ago tha t the muscle spindle receptor discharge is depressed during isotonic contraction (Hunt and Kuffler, 195l) in a similar way as it has been found now for neuromuscular impulse transmission. The role of the muscle spindle system in the control of motor act ivi ty has been investigated extensively. The functional importance of the synaptic feedback system described in this paper is however an interesting problem still open for investigation.

Acknowledgements. Thanks is due to Dr. A. C. Bobbert for the introduction into the problem of this investigation and to Professor Dr. A. A. Verveen for encouraging discussions. Deanna Anderson en Arie de Vos provided good experimental and technical assistance respectively.

References

Allen, D. G., Jewell, B. R., Murray, J. W. : The contribution of activation processes to the length-tension relation of cardiac muscles. Nature (Lend.) 248, 606--607 (1974)

Bobbert, A. C. : Impulse transmission at frog myoneural junctions. Proc. I.U.P.S. VII, 49 (1968)

Bobbert, A. C. : The transfer function of frog myoneural junctions. I. The influences of fatique, facilitation, and depression upon impulse transmission in vitro. Acta ph:isiol, pharmacol, neerl. 15, 289--328 (1969)

Borer, I)., Bovet-Nitti, F., Marini-Bettolo, G. B.: Curare and curare-like agents. Amsterdam: Elsevier Pub. Comp. 1959

306 D.L. Ypey

Burnstock, G., Holman, M. E., Prosser, C. L. : Elcctrophysiology of smooth muscle. Physiol. Rev. 48, 482--527 (1963)

Capek, R., Esplin, D. W., Salehmoghaddam, S. : Rates of transmitter turnover at the frog neuromuscular junction estimated by electrophysiological techniques. J. Neurophysiol. 84, 831--841 (1971)

Coraboeuf, E.: Resistance measurements by means of microelectrodcs in cardiac tissues, chapter 12. In: Glass microelectrodes. Edited by M. Lavall@e, 0. F. Schanne and N. C. Hebert. New York-London-Sydney-Toronto: John Wiley & Sons, Inc. 1969

Dubuisson, M.: Impedance changes in muscle during contraction and their possible relation to chemical processes. J. Physiol. (Lond.) 89, 132--152 (1937)

Edman, K. A. P., Kiessling, A. : The time course of the active state in relation to sarcomere length and movement studied in single skeletal muscle fibres of the frog. Acta physiol, scand. 81, 182--196 (1971)

Goldman, L.: The effects of stretch on cable and spike parameter single nerve fibres; some implications for the theory of impulse propagation. J. Physiol. (Lond.) 175, 425--444 (1964)

Gordon, A. M., Huxley, A. F., Julian, F. J. : The variation in isometric tension with sarcomere length in vertebrate muscle fibres, J. Physiol. (Lond.) 184, 170--182 (1966)

Hunt, C. C., Kuffler, S. W. : Stretch receptor discharges during muscle contraction. J. Physiol. (Lond.) 113, 298--315 (1951)

Hutter, O. F., Trautwein, W. : Neuromuscular facilitation by stretch of motor nerve endings. J. Physiol. (Lond.) 183, 610--625 (1956)

Katz, B. : Nerve, muscle and synapse. New York: Mac Graw Hill 1966 Kiessling, A. : Transfer characteristics of subsystems responsible for muscular con-

traction as controlled by muscle length. Biokybernetik III, 145--149 (1971) Mallart, A., Martin, A. R. : An analysis of facilitation of transmitter release at the

neuromuscular junction of the frog. J. Physiol. (Lond.) 198, 679--694 (1967) Taylor, S. R., Rfidel, R.: Striated muscle fibres: Inactivation of contraction in-

duced by shortening. Science 167, 882 (1970) Woodbury, J. W., Brady, A. J.: Intracellular recording from moving tissues with

a flexibly mounted ultra-microelectrode. Science 128, 100 (1956) Ypey, D. L., Kerkhof, P. L. M., Bobbert, A. C. : Muscle length and neuromuscular

transmission in the frog. Pfliigers Arch. 847, 309--322 (1974) Ypey, D. L., Meyer, A. W. A. : Muscle length and neuromuscular transmission of

rana temporaria. Abstracts of contributed papers 3, p. 271. IUPAB Congress IV, Moscow 1972

D. L. Ypey Dept. of Physiology, University of Leiden, Wassenaarseweg 62, Leiden, The Netherlands

feedback of isotonic muscle contraction on neuromuscular impulse transmission in the frog

Documents