Changes in primary muscle spindle ending excitability induced by a ramp-and-hold stretch

BRAIN RESEARCH ELSEVIER

Brain Research 705 (1995) 216-224

Research report

Changes in primary muscle spindle ending excitability induced by a ramp-and-hold stretch S.S. Sch~ifer * Abteilung Neurophysiologie 4230, Medizinische Hochschule Hanno~:er, Konstan~-Gutschow-Str. 8, 30625 Hannover, Germany

Accepted 5 September 1995

Abstract 36 primary (la) muscle spindle afferents from the tibial anterior muscle of the cat were subjected to a ramp-and-hold stretch (stretch rate 10 mm/s, stretch amplitude 5 or 8.5 mm) of the muscle, on which a sinusoidal stretch (50 Hz) of four different amplitudes (25, 50, 250 and 500 /zm) was superimposed. In 54 discharge patterns a Ia afferent subjected to a ramp-and-hold stretch with a sinusoidal stretch superimposed responded only to the superimposed sinusoidal stretch. In 25 of the cases the Ia afferent responded with an one-to-one driven action potential (AP) and in 29 of the cases with two APs per sinusoidal stretch. For these 54 discharge patterns the phase of the sinusoidal cycle was determined at which each AP occurred. Where the Ia afferent responded with one AP per cycle an accelerating phase advance was observed during the ramp stage of the underlying ramp-and-hold stretch and a decelerating phase advance during the plateau. This phase shift means that the excitability of the site generating the AP increased during the ramp stage and decreased during the plateau. If the Ia afferent responded with two APs per superimposed cycle, the second AP per cycle evinced a decelerating phase advance during the ramp and an accelerating phase advance during the plateau. The phase of the second AP per cycle showed a second, contrary change in excitability at the AP generating site. The excitability decreased during the ramp and increased during the plateau. The first kind of excitability change is interpreted as a consequence of an inward current at the AP generating site. The second, contrary type of excitability points to an interplay between an inward and an outward current. An increasing outward current lowers the excitability for the second AP per cycle during the ramp. A decreasing outward current raises the excitability during the plateau. Keywords: Primary muscle spindle afferent; Ramp stretch response; Sinusoidal superimposition; Ia ending excitability; Receptor current

1. Introduction The excitability of a Ia muscle spindle afferent is related to the threshold of the site initiating the action potential (AP) [5,19]. If the threshold is low excitability is high and consequently so is the discharge frequency of the Ia afferent, and vice versa, Under a ramp-and-hold stretch the excitability of the A P generating site is high during the ramp and lower during the plateau. This change in the excitability is evinced by the discharge frequency, which is high during the ramp stage and lower during the plateau. The reason for a different degree of excitability is assumed to be the effect of a different magnitude of receptor potential depolarization at the A P initiating site. Receptor

* Corresponding author. Fax: (49) (511) 532-2776. 0006-8993/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0006-8993(95)01 183-8

potential depolarization is high during the ramp and lower during the plateau [9,10,12]. Therefore the changes in the excitability of the AP initiating site are an approximate reproduction of the changes in receptor potential. Beyond this, Ottoson et al. [20] look at sensitivities of the receptor potential which are displayed under a rampand-hold stretch of the isolated frog muscle spindle. To investigate these the authors superimpose small test stretches on a ramp-and-hold stretch and record the receptor potential. The deviations in receptor potential with which the spindles respond to the test stretches are high at the beginning of the ramp. However, the magnitude of the deviations declines continuously until the end of the ramp. Therefore, the sensitivity of the receptor potential falls with the increasing duration of the ramp. As for the plateau stage, the receptor potential deviations are small at beginning. However, the magnitude of the deviations increases continuously with the duration of the plateau (see also

s.s. Schiller/Brain Research 705 (1995) 216-224 Hunt and Ottoson [11], Fig. 2). The receptor potential has low sensitivity at the end of the ramp and at the beginning of the plateau. Concomitantly with the decrease in the sensitivity of the receptor potential the threshold of the AP generating site increases [19]. The threshold is raised at the end of the ramp and at the beginning of the plateau. We repeated the experiments of Ottoson et al. [20] in a modified way. We superimposed a sinusoidal stretch of small amplitude on a ramp-and-hold stretch of high amplitude. The sinusoidal stretch is a test stretch, testing the ability of the Ia afferent to respond under the ramp-and-hold stretch. In this way we determine the excitability, i.e. the threshold of the AP generating site. We observe two contrary excitabilities. In the first place the excitability of the AP initiating site changes, the change being dependent on the ramp-and-hold stretch, i.e. increasing during the ramp and decreasing during the plateau. In addition, however, a second less noticeable excitability change, in the opposite direction, overlays the first, obvious, excitability change. It leads to a decrease in the excitability of the AP initiating site during the ramp and to an increase in its excitability during the plateau. Both contrary excitability changes are interpreted by the currents acting at the impulse initiating site.

2. Materials and methods 2.1. Surgery The experiments were performed on 10 cats ( 2 - 4 kg in weight) anaesthetized with sodium pentobarbitone (initial dose, 40 m g / k g i.v.; continuation of anaesthesia, 5 m g / h i.v.). The subsequent surgical procedure has been described previously [31]. The left hind limb was denervated with the exception of the tibial anterior muscle. The ventral and dorsal roots of L 7 and L 6 w e r e cut. 36 Ia afferents (conduction velocity between 80 and 120 m / s ) were investigated. 2.2. Stretching and obtaining of the discharge patterns The muscle was stretched in a ramp fashion at a rate of 10 m m / s and with an amplitude of 5 or 8.5 mm from the medium prestretch of the muscle of 6 mm. A prestretch of 0 mm corresponds to the length of the muscle under maximum plantar flexion. The plateau of the ramp-and-hold stretch was held for a period of 2.5 s for 18 Ia afferents and of 5 s for the remaining 18 Ia afferents. The release of the stretch was carried out in ramp fashion at a rate of 10 m m / s . A sinusoidal stretch of 50 Hz and with a variety of different amplitudes, 0, 25, 50, and 250 /zm, was superimposed on the ramp-and-hold stretch. In addition, a sinusoidal stretch with an amplitude of 500 /xm was used with 18 of the Ia afferents. With 18 Ia afferents the sinusoidal stretch was started 1 s before the ramp-and-hold stretch

217

and continued for 1 s after the end of the release. With the remaining 18 Ia afferents the sinusoidal stretch started simultaneously with the ramp stretch and stopped with the end of the release. Each stretch with the same set of parameters was repeated 5 times with a waiting time of 7 s between each stretch. The responses of each Ia afferent in the form of its discharge frequency from the second to the fifth stretch were superimposed to obtain a discharge pattern [30]. Stretching was effected using a linear-controlled electromagnetic stretching device (own construction). The controller has a tolerance of 2 /xm as between the target value and the actual length of stretch effected: the length measuring device has a measurement uncertainty of 1 /xm. The target value was the length of a ramp-and-hold stretch with sinusoidal superimposition. The actual length change of the stretching device was recorded and monitored in all experiments. The amplitude of the superimposed sine stretch was constant before the ramp, during the ramp, and during the plateau of the ramp-and-hold stretch. In addition, the tension of the muscle was recorded. The sinusoidally modulated tension change was small at the initial length of the muscle, i.e. before the ramp; it increased during the ramp, and remained constant and high during the plateau. This increasing magnitude of the changes in oscillatory tension as the ramp-and-hold stretch progresses is due to the increasing stiffness of the muscle. We posed the question as to whether there was a correlation between the increasing magnitude of the changes in oscillatory tension and the extent to which the superimposed sine stretch was transferred to the spindle. First of all it should be said that the muscle was not slack: we chose a medium prestretch of 6 ram. There was no slackness in the spindle either: where no sinusoidal stretch was superimposed, the 36 Ia afferents responded to the ramp-and-hold stretch with an initial peak at the beginning of the ramp. Therefore, if we proceed on the assumption that the muscle is homogeneous, the constant amplitude of the superimposed sine stretch will be transferred uniformly to all parts of the muscle irrespective of its stiffness. The spindle, lying parallel to the muscle fibres, will receive a constant fraction of the superimposed sine stretch. However, the homogeneity of the muscle is only approximate. We therefore extended the scope of the experiment in respect of 18 Ia afferents. A sinusoidally superimposed ramp stretch of constant magnitude was carried out under five degrees of prestretch of the muscle (0, 3, 6, 9, 12 mm) [32]. Eight of the 18 Ia afferents responded with one action potential per superimposed sine stretch (as in Fig. la) at each of the five degrees of prestretch. Phase graphs (as Fig. lb) were obtained for each discharge pattern of the 8 Ia afferents. In respect of each individual Ia afferent phase graphs were obtained for each of the five degrees of prestretch; these were almost identical. Thus the increasing stiffness of the muscle as a function of the length of stretch does not influence the results reported in this work.

218

S.S. Schiifer / Brain Research 705 (1995) 216-224

3. R e s u l t s The experimental method produces 126 discharge patterns under ramp stretching with sinusoidal superimposition for the 36 Ia afferents investigated. In every discharge pattern the Ia afferent showed a response to the superimposed sinusoidal stretch. At the same time it was possible to observe that the ramp of the ramp-and-hold stretch basically had an excitatory effect on the Ia fibre: thus an Ia afferent was in general able to generate one A P per cycle during the ramp, whereas during the plateau it responded with an A P only to every second cycle. However, it also occurred that a Ia afferent generated one AP per cycle throughout the duration of both the ramp and the plateau; or, it responded with two APs per cycle during the ramp and with only one A P per cycle during the plateau; or else it generated two A P s per cycle throughout the duration of both the ramp and the plateau. For this work, only those discharge patterns were selected in which the Ia afferent generated one or two A P s per cycle throughout both the ramp and the plateau stages. Accordingly there were 25 discharge patterns available with one A P per cycle and 29 with two APs per cycle. Even though the remaining discharge patterns show basically the same changes in excitability, this investigation refers only to these 54 discharge patterns which display an unequivocal response by the Ia afferent. Fig. l a shows the discharge pattern of a Ia afferent generating one A P per superimposed sinusoidal stretch before the ramp, during the ramp and during the plateau. This is shown by the constant discharge frequency of 50 i m p / s . The underlying ramp-and-hold stretch can be recognized from two discharge frequency dots which lie higher at the beginning of the ramp and one that lies slightly lower at the end of the ramp. Fig. l b gives the phase of the sinusoidal stretch at which each A P occurred. In order to determine the sinusoidal stretch without the ramp-and-hold stretch the acceleration of the stretch was recorded during the experiment. The ' n o i s e ' of the acceleration was smoothed out by the moving average procedure. The phase of each A P in relation to the sinusoidally-modulated acceleration was calculated with a program developed for an IBM A T compatible computer. The choice of the sinusoidally modulated acceleration as the reference value for determining the phase of an AP was made entirely for methodological reasons: we do not believe that a spindle subjected to a sine stretch generates its APs only in response to the acceleration of the sine stretch [18]. A cycle has a rising and a falling stage. The transition from the falling to the rising stage was defined as the 0 ° phase. In Fig. l b the phase is shown on the y-axis. Phases from 0 ° to 180 ° correspond to the rising stage of the sinusoidally-modulated acceleration and those from 180 ° to 360 ° to the falling stage. Each cross describes the phase at which one A P occurs in relation to the acceleration. As we do not know the phase that exists at the spindle site at any

r-f --~:-: - . ~--%---g . ,

~ .: :

. ....

slnusoldal stretch

gO.

L.

<,

o

2'70" 360-

!

Is

!

xinuloldal w~t'e~h

|.3U

Fig. 1. a: the response of a la afferent to a ramp-and-hold stretch with sinusoidal superimposition. Velocity of the ramp = 10 ram/s; sinusoidal stretch, frequency = 50 Hz, amplitude = 50 /~m. The la afferent responds with a one-to-one driven action potential to the sinusoidal stretch, b: phases of the action potentials of a in relation to the sinnsoidally-modulated acceleration. Each cross represents the phase of an action potential in response to one sinusoidal stretch. The time-dependent changes in the phase reveal the effect of the underlying ramp-and-hold stretch on the phase of the one-to-one driven action potential. DP, dynamic peak; MSt, maximum static value; FSt, final static value, c: a representative discharge pattern of a Ia afferent in response to a ramp-and-hold stretch. Ramp rate = 10 mm/s. The time-dependent similarity between the phase (b) and the discharge frequency is clearly visible.

given moment the absolute phase values are not meaningful: only the changes of the phase which can be identified in the phase plot shown in Fig. l b are useful data for us. Because the Ia afferent responds to each sinusoidal stretch with one A P (Fig. la), each succeeding cross occurring along the time axis (x-axis) corresponds to one sinusoidal stretch (Fig. lb). The phase of the A P shows a little scatter but is otherwise constant during the first second shown in Fig. l b before the beginning of the ramp. The A P jumps to a lower

S.S. Schiifer /Brain Research 705 (1995) 216-224 300" o

.

Fig. 2. The phase at the time of the dynamicpeak (closed circles) and of the final static value (open circles) plotted against the phase at the time of the maximum static value of a one-to-one driven action potential in response to a ramp-and-hold stretch with sinusoidal superimposition. Bisector: phase of the maximumstatic value. The ramp of the underlying ramp-and-hold stretch leads to an accelerating phase advance of the one-to-one driven action potential, the plateau to a decelerating phase advance comparedto the phase of the maximumstatic value.

phase with the beginning of the ramp. Subsequently, the phase forms an initial peak. Thereafter the AP remains at a lower phase during the remaining part of the ramp. With the beginning of the plateau the phase increases, at first steeply and then gradually, until the end of the plateau. If the Ia afferent generates one AP per cycle, then the phase, being a response to the underlying macroscopic ramp-andhold stretch, shows all the characteristic features that we know from the discharge pattern of a Ia afferent. Fig. lc shows such a discharge pattern for comparison. In both panels the excitability of the AP generating site is elevated during the ramp. This is because the Ia afferent is able to generate its AP at an earlier point of the sinusoidal stretch during the ramp stage (Fig. lb), even though the sinusoidal stretch remains unchanged, just as it is able to diminish its interspike intervals during the ramp stage (Fig. lc). During the plateau the excitability of the AP generating site decreases continuously. This is because in Fig. lb the phase increases during the plateau, in the same way as in Fig. lc the interspike intervals increase. Fig. 2 generalizes our observations of the phase graphs obtained from the 25 discharge patterns in which the Ia afferents respond with one AP per cycle during both the ramp and the plateau. (The amplitude of the superimposed sinusoidal stretch was 25 ~ m for 11 discharge patterns, 50 /xm for 10 discharge patterns, 250 /xm for 3 discharge patterns and 500 p,m for one discharge pattern. The differing amplitudes do not lead to any systematic phase changes in the APs being evident in the phase graphs.) Three phase values were read from each phase graph. The phase during the last 50 ms at the end of the ramp represents the phase of the dynamic peak. The phase during the first 20 ms of the plateau represents the phase of the maximum static value. The phase during the second quarter of the second second of the plateau represents the phase of the final

219

static value. In Fig. 2 the phases of the dynamic peak (open circles) and the final static value (closed circles) are plotted against the phase of the maximum static value. In this way one is not dependent upon the absolute value of the phase, which is not meaningful in our method. However, the relative changes of the phase are clearly identifiable in the diagram. The 45 bisector shows the phase of the maximum static value. The maximum static value separates the Ia afferent's response to the ramp from its response to the plateau of the ramp-and-hold stretch. Fig. 2 shows that the APs occur at a lower phase at the time of the dynamic peak than at the time of the maximum static value in every phase pattern. The phase of the AP at the time of the final static value is higher in every phase pattern than at the time of the maximum static value. Therefore the excitability of the Ia afferent increases during the ramp and decreases during the plateau in each of the 25 phase graphs. Fig. 3a shows a discharge pattern in which the Ia afferent responds to the superimposed sinusoidal stretch with two APs per cycle during the ramp and during the plateau of the ramp-and-hold stretch. In this discharge pattern the sinusoidal stretch starts simultaneously with the beginning of the ramp. The two APs per cycle lead to the occurrence of two discharge frequency bands. The lower discharge band jumps from the initial activity present before the ramp to a discharge frequency of 58 i m p / s with the beginning of the ramp-and-hold stretch. The upper discharge band has a discharge frequency of 420 i m p / s at the beginning of the ramp. However, the discharge frequency of the upper band declines after a second peak continuously during the course of the ramp to reach such a low level at its end (the time of the dynamic peak) that the discharge frequency differs only negligibly from the discharge frequency evident at the beginning of the plateau (the time of the maximum static value). By contrast, the discharge frequency increases again during the plateau until its end (the time of the final static value). During the release stage the Ia afferent responds only with one AP per cycle. The phase graph of Fig. 3b explains the course of the discharge frequency of the upper discharge band in terms of the interval between the phases of the sinusoidal stretch at which the two APs per cycle occur. The APs show no regularity in their phases before the ramp, because the sinusoidal stretch has not yet started at that time. With the beginning of the ramp the Ia afferent responds with two APs to each superimposed sinusoidal stretch. The two APs show the characteristic phase changes described in Fig. lb as a response to the underlying ramp-and-hold stretch. The phase advance accelerates during the ramp, then decelerates slightly during the plateau and steeply during the release. However, if the phase interval between the two APs per cycle during the ramp is examined, it is found to be smaller at the beginning of the ramp than at its end. This means that while the ramp of the ramp-and-hold

S.S. Schiifer / Brain Research 705 (1995) 216-224

220 a

DV

I

sinuoidld s~-o~h

0

b 4.

e~gO

* 4" ÷

4"

180. sinmotd81 sd.ro~h

Fig. 3. The response of a Ia afferent to a ramp-and-hold stretch with sinusoidal superimposition. Ramp rate = 10 m m / s ; sinusoidal stretch, frequency = 50 Hz, amplitude = 500 p.m. DV, dynamic value; DP, dynamic peak; MSt, maximum static value; FSt, final static value. The la afferent responds to the stimulus with two action potentials per cycle. The response of the Ia afferent is shown (a) by its discharge frequency, and (b) by the phases of the action potentials in relation to the sinusoidallymodulated acceleration. The phase change of the second action potential per cycle (2nd AP) is smaller both during the ramp and during the plateau than that of the first action potential per cycle (lst AP).

stretch elevates ~ e excitability of the AP initiating site for both APs, for the second AP per cycle it does so to a lesser degree. As a result the interval between the two APs increases from the beginning of the ramp until its end. The reciprocal value of the interval between the first and the second AP per cycle is the discharge frequency of the upper discharge band. Thus the discharge frequency of the upper discharge band decreases during the ramp. During the plateau the phase interval between the two APs per cycle is larger at its beginning (at the time of the maximum static value) than at its end (at the time of the final static value). This diminution of the interval between the first and the second AP per cycle leads to the continuous increase in the discharge frequency of the upper discharge band during the plateau that can be seen in Fig. 3a. The cause for the decrease in the phase interval is the fact that the excitability of the AP generating site de-

creases more slowly for the second than for the first AP per cycle during the plateau. The excitability in respect of the second AP per cycle is enhanced during the plateau and diminishes during the ramp. During the release stage the excitability of the AP generating site decreases at first in a small step. Thereafter, the phase and thus also the excitability of the AP generating site remains nearly constant over some cycles before it declines steeply. To summarize, one can say that a ramp-and-hold stretch induces two contrary changes in excitability at the AP generating site. On the one hand excitability increases during the ramp, declines continuously during the plateau and falls steeply during the release. This kind of excitability change shows the obvious course of the phase of both APs. On the other hand, a second less noticeable change in excitability can be evinced from the course of the phase of the second AP. The excitability of the AP generating site diminishes during the course of the ramp, and is enhanced again during the plateau. The two contrary changes in excitability lead in five cases to unexpected discharge patterns, one of which is shown in Fig. 4. The Ia afferent generates two APs per cycle from the beginning of the ramp-and-hold stretch until the mid-point of the release. Additionally, a third AP is generated during the last third of the ramp stage. The reciprocal value of the interval between the first and the second AP per cycle describes the discharge frequency of the upper discharge band in Fig. 4a. The upper discharge band declines smoothly at first and then steeply during the ramp to reach the maximum static value. Correspondingly, the distance between the first and the second AP per cycle increases (Fig. 4b). Nevertheless, the Ia afferent generates a third AP per cycle during the last third of the ramp. It is the effect of the ramp in increasing excitability that leads to the third AP per cycle. Simultaneously the ramp exerts the contrary effect of diminishing excitability, with the result that the interval between the second and third APs per cycle is so large that it prevents any higher discharge frequency from being observed in Fig. 4a. The third AP leads in Fig. 4a to the cloud of points at the end of the ramp which lie far below the upper discharge band. During the plateau the upper discharge band in Fig. 4a shows the second excitability-increasing effect on the threshold of the second AP, in the same way as in Fig. 3a. During the release the upper discharge band shows first an excitability-increasing effect on the threshold of the second AP, followed by a steep fall in the excitability of the second AP. Thereafter the Ia afferent responds with only one AP per cycle. An excitability increase for the second AP occurring at the beginning of the release is generally observed in the cases of those Ia afferents that generate two APs per cycle during the release. The change in the interval between the first and the second AP during the ramp and the plateau is small measured in terms of the absolute phase values. In seeking

S.S. Schiifer /Brain Research 705 (1995) 216-224

to generalize the change in the excitability of the AP generating site for the second AP we did not examine the change in the absolute phase values. Instead we determined the changes in the discharge frequency of the upper discharge band. If the interval between two APs is small (the interval between the first and the second AP lies between 2.5 and 5 ms in the 29 discharge patterns available) small changes in an already small interspike interval exert a large effect on the reciprocal value. The highest discharge frequency of the upper discharge band during the ramp was determined for each of the 29 discharge patterns available: this value is referred to as the dynamic value. In Fig. 3a and Fig. 4a, the dynamic value lies at the beginning of the ramp. Often, however, it was found later in the course of the ramp. We also took readings of the dynamic peak discharge, which is the discharge frequency of the upper discharge band at the end of the ramp; of the maximum static value, which is the discharge frequency of the upper discharge band at the beginning of the plateau; and finally of the final static value, the discharge frequency of the upper discharge band

221

o 0 ~

50-

45O-

b

L -',

-

...j

its'° Fig. 5. a: dynamic value plotted against dynamic peak of the upper discharge band of 29 discharge patterns with two action potentials per cycle in response to a ramp-and-hold stretch with sinusoidal superimposition. Bisector: discharge frequency of the dynamic peak. The discharge frequency of the upper discharge band declines during the ramp phase, b: final static value plotted against maximum static value of the upper discharge band from the same 29 discharge patterns as a. Bisector: discharge frequency of the maximum static value. The discharge frequency of the upper discharge band increases during the plateau stage in 16 of the 29 discharge patterns.

l_._._

sinusoidoJ

O-

stretnh

b

90 m~80

270-

~w_l_~'~-~ i , ÷

3ntAp

"-5~" "±. r L ~ m ~

'~

I~AP

2~tAp

stnuaoklal s ~ ' e t e h Fig. 4. The response of a Ia afferent to a ramp-and-hold stretch with sinusoidal superimposition. Ramp rate = 10 m m / s ; sinusoidal stretch, frequency = 50 Hz, amplitude = 500/xm. The Ia afferent responds to the stimulus with two action potentials per cycle, and during the last third of the ramp with three action potentials per cycle. The response of the Ia afferent is shown (a) by its discharge frequency, and (b) by the phases of the action potentials in relation to the sinusoidally-modulated acceleration. 1st AP, first action potential per cycle; 2nd AP, second action potential per cycle; 3rd AP, third action potential per cycle.

after 2.5 s of the plateau. In Fig. 5a the dynamic x(alue is plotted against the dynamic peak of a discharg~ ~attern. (The discharge patterns used for Fig. 5 were obtained with the following amplitudes of the superimposed sinusoidal stretch: two discharge patterns at 50 /zm, 9 discharge patterns at 250 /zm and 18 discharge patterns at 500 /zm.) It can be seen that the discharge frequency declines during the ramp from the dynamic value to the dynamic peak, because with one exception the dynamic value lies above the 45 bisector. (In Fig. 5a the 45 bisector represents the magnitude of the dynamic peak.) In 28 of the 29 discharge patterns, therefore, the excitability of the AP generating site decreases for the second AP during the ramp. In Fig. 5b, the final static value is plotted against the maximum static value read from the same 29 discharge patterns. It can be seen that the final static value is higher than the maximum static value in 16 of the discharge patterns. (In Fig. 5b the 45 bisector represents the magnitude of the maximum static value.) We thus observe that in 55% of

222

S.S. Schiifer / Brain Research 705 (1995) 216-224

the discharge patterns the excitability of the AP generating site increases for the second AP during the plateau.

4. Discussion Two contrary excitability changes are observed at the AP generating site under a ramp-and-hold stretch. Where the Ia afferent responds to the sinusoidally superimposed ramp-and-hold stretch with one AP per cycle the phase of the one-to-one driven AP indicates an increasing excitability during the ramp and a decreasing excitability during the plateau. A second contrary excitability change is superimposed on the first one. This second excitability change is observable where the Ia afferent generates two APs per cycle. In respect of the second AP per cycle excitability decreases during the ramp and increases during the plateau. These excitability changes are due to the fact that there is a change in the threshold at the AP generating site. The threshold is reached when the inward current is just higher than the outward current at the AP generating site. We shall consider possible causes of changes in the currents at the AP generating site. In the first place the currents acting at the AP generating site are generated in the sensory terminals, so that the receptor currents and their changes need to be considered. Secondly, the APs themselves may initiate feedback effects on the AP generating site which need to be examined as well. Sensory terminals experience elongation when subjected to a stretch. This causes the stretch-activated (SA) channels lying in the membrane of the sensory terminals to open in parallel with the deformation of the sensory endings [4,16]. As a result the membrane becomes unspecifically permeable for cations [29]. The activation of the SA channels generates an inward current carried by Na + and Ca 2+ [12] and a K + outward current. The inward current, however, predominates over the outward current [29]. The net inward current is enhanced under a ramp-and-hold stretch as long as the ramp lasts, because the SA channels are subjected to steadily increasing activation. However, the general view at present is that the inward current appears to activate a secondary K + outward current. On the one hand there is discussion of a Ca 2+ activated K + outward current [4,15,21,30,33]. Swerup et al. [21,33] explain in relation to the crayfish that a stretch builds up a C a 2+ activated K + current at the stretch receptor, which causes the initial steep repolarization of the receptor potential after the end of the ramp. Kruse and Poppele [15] show that the dynamic property of the isolated cat muscle spindle decreases if C a 2+ activated K + channels o r C a 2+ channels are blocked. The authors postulate a Ca 2+ activated K + outward current which stabilizes the dynamic properties of the Ia afferent through a negative feedback. On the other hand, a voltage-dependent K + current cannot be excluded. Such a voltage-dependent K + outward current has been demonstrated for the crayfish stretch receptor

[21,22,27,28,33]. It has the properties of a delayed rectifier. It is activated by a depolarization higher than - 7 0 mV. The K + current has a fast activation constant (r = 1 - 4 ms) and a slow inactivation constant (r = 500 ms). The stronger the inward current via the SA channels is, the stronger both K + outward currents become. The free intracellular Ca 2+ concentration increases with increasing inward current, so that the Ca 2+ activated K + outward current is enhanced. Simultaneously the receptor potential depolarizes so that the voltage-dependent K + outward current is also enhanced. However, the secondary K + outward currents counteract the inward current, with the result that net inward current increases more slowly during the dynamic phase of stretching. To the extent that the net inward current increases, the excitability of the AP generating site is enhanced during the ramp. The increasing excitability can be evinced either from the increasing discharge frequency (Fig. lc), or else, as described in this paper, from the increasing phase advance in a one-to-one driven AP under a sinusoidally superimposed ramp stretch (Fig. lb). Such a synchronism between the phase change and the discharge frequency change is predicted by the model of Awiszus [1] if the inward current increases at the AP generating site. Awiszus chooses as an AP generating site the Hodgkin-Huxley model neuron of the squid giant axon [7]. Therefore only two kinds of channel are taken into account in this model namely, an Na + channel and a delayed rectifier K + channel. In a mammalian model neuron, by contrast, a much larger diversity of ionic channels would need to be taken into consideration [6]. Thus the particularities of the model simulations of Awiszus cannot be transferred to the muscle spindle. Nevertheless the underlying results would appear to apply to a mammalian neuron. Poppele and Chen [23] simulate the AP generating site of the muscle spindle through a leaky integrator model. Just as in the case of the Hodgkin-Huxley model neuron under a sine stretch, the authors describe an increasing phase advance of a one-toone driven AP if the inward current increases at the AP generating site. We therefore believe that the increasing phase advance of a one-to-one driven AP under a ramp stretch with sinusoidal superimposition is a consequence of the increasing net inward current. If the Ia fibre generates two APs per superimposed sine stretch, the phase advance of the second AP increases more slowly during the ramp than that of the first AP (Fig. 3b). We can only observe the lesser phase advance increase for the second AP because the phase of the second AP has a reference point in the phase of the first AP. It may therefore well be that a lesser phase advance increase exists for both the first and the second AP: in which case the following considerations although here confined to the phase of the second AP per sine stretch, might also be valid for the first AP. If the second AP per sine stretch evinces a lesser phase advance increase, this must mean that the excitability of

S.S. Schiifer / Brain Research 705 (1995) 216-224

the AP generating site also increases more slowly, so that the threshold is reached later and later during the ramp. The threshold is reached when the inward current is just stronger than the outward current at the AP generating site. We proceed on the assumption that an increasing inward current through the SA channels produced by the ramp initiates a K + outward current which gains in strength the stronger the inward current is. Thus the K ÷ outward current and the inward current increase concurrently during the ramp. To the same degree as the K ÷ outward current increases, the threshold of the AP generating site is raised. It is our belief that this effect of the K ÷ outward current can be observed in the slower increase in the excitability of the AP generating site, i.e. in the more slowly increasing phase advance of the second AP per superimposed sine stretch. During the plateau of the ramp-and-hold stretch the sensory spirals of the nuclear bag 1 fibres close again (display 'creep', [2,24]). In the case of the nuclear bag 2 and chain fibres such a 'creep' of the sensory terminals, if it occurs at all, is only very slight. In line with this deformation of the sensory terminals, however, the receptor current running through the SA channels will tend to weaken during the plateau. A secondary K + outward current appears to be superimposed on this inward current. Swerup and co-workers [21,22,27,28,33], with regard to the crayfish stretch receptor, postulate a voltage-dependent K ÷ outward current which inactivates during the plateau with a time constant of 500 ms. This means that the net inward current will be constantly diminished by the effect of such a K ÷ outward current. Correspondingly, the excitability of the AP generating site decreases. This decreasing excitability is manifested by a continuous decrease in the discharge frequency during the plateau (Fig. lc) and also in the phase advance of a one-to-one driven AP (Fig. lb). If two APs per superimposed sinusoidal stretch are generated during the plateau, the phase advance of the second AP diminishes to a lesser degree during the plateau than that of the first AP. This phase change of the second AP is the manifestation of an enhancement in the excitability of the AP generating site. The excitability is enhanced if the secondary K + outward current decreases continuously during the course of the plateau. These considerations so far explain the phase changes of a one-to-one driven AP under a sinusoidally superimposed ramp-and-hold stretch in terms of a net inward current. If the Ia afferent generates two APs per superimposed sine stretch the phase changes of the second AP, determined relatively to those of the first AP, are explained in terms of a secondary K + outward current which is activated by the inward current. Inward and outward currents are thought to arise in the terminal endings as a consequence of their deformation. However, we cannot exclude the possibility that the secondary K ÷ outward current or components of it are generated directly at the

223

AP generating site. A K ÷ outward current activated by the APs themselves has been demonstrated in respect of various cells of the CNS, which leads to afterhyperpolarization increasing from AP to AP [17] (IAHP). The Ca 2÷ conductance increases during the APs. The free intracellular Ca 2÷ concentration activates K + channels which remain open over several hundred milliseconds [26]. A K ÷ outward current of this kind has not so far been discussed in connection with the AP generating sites of stretch receptors (muscle spindle [14], Golgi tendon organ [25], lobster stretch receptor [3]), and therefore appears to be a less attractive hypothesis. Positive charges may reach the extracellular space not only by means of K ÷, but also by means of Na ÷, through the agency of an electrogenic pump. Na ions diffuse into the intracellular space during the APs. The free intracellular Na + concentration activates the electrogenic pump which counteracts the inward current [8,13]. Leak currents too will counteract the inward current. These outward currents arising at the AP generating site will impair the excitability of the AP generating site during the ramp because they increase during the ramp; they will enhance its excitability during the plateau because they decrease during the plateau. We take these further possibilities into consideration because our experiments do not allow us to determine where in the muscle spindle the inward and outward currents are generated which act on the AP initiating site. We have recorded AP sequences from the Ia afferent under a sinusoidally superimposed ramp-and-hold stretch. We have determined the phase changes of the APs by reference to the superimposed sine stretch. The AP sequences are generated at the AP generating site of the Ia afferent. Therefore the phase changes reveal excitability changes, i.e. changes in the threshold of the AP generating site. The changes in the threshold are interpreted in terms of an interplay of inward and outward currents acting on the AP initiating site. We believe that these inward and outward currents are mainly generated at the sensory terminals. The results of Hunt and Ottoson [11] furnish evidence to support our assumption. These authors record the receptor potential of isolated cat muscle spindles under a step-and-hold stretch on which a sine stretch is superimposed. The amplitude of the receptor potential sinusoidal modulations increases continuously from the beginning to the end of the plateau. In the same way, in our experiments the excitability of the AP generating site increases continuously during the plateau in respect of the second AP per sine stretch.

References [1] Awiszus, F., Continuousfunctionsfor the analysis of sensory transduction, Biol. Cybern., 61 (1989) 153-161. [2] Boyd, I.A., Gladden, M.H. and Ward, J., The contribution of

224

[3]

[4]

[5]

[6] [7]

[8] [9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

S.S. Schiifer / Brain Research 705 (1995) 216-224 mechanical events in the dynamic bag 1 intrafusal fibre in isolated cat muscle spindles in the form of the Ia afferent axon discharge, J. Physiol., 317 (1981) 80-81 P. Edman, A., Gestrelius, S. and Gramp, W., Analysis of gated membrane currents and mechanisms of firing control in the rapidly adapting lobster stretch receptor, J. Physiol., 384 (1987) 649-669. Erxleben, Ch.F.J., Calcium influx through stretch-activated cation channels mediates adaptation by potassium current activation, NeuroReport, 4 (1993) 616-618. Gregory, J.E., Harvey, R.J. and Proske, U., A 'late supernormal period' in the recovery of excitability following an action potential in muscle spindle and tendon organ receptors, J. Physiol., 271 (1977) 449-472. Hille, B., Ionic Channels of Excitable Membranes, Sinaur, Sunderland, Massachusetts, 1984. Hodgkin, A.L. and Huxley, A.F., A quantitative description of membrane current and its application to conduction and excitation in nerve, J. Physiol., 117 (1952) 500-544. Holloway, S.F. and Poppele, R.E., Evidence for electrogenic pumping in cat muscle spindle, Brain Res., 154 (1978) 144-147. Hunt, C.C. and Ottoson, D., Impulse activity and receptor potential of primary and secondary endings of isolated mammalian muscle spindles, J. Physiol., 252 (1975) 259-281. Hunt, C.C. and Ottoson, D., Initial burst of primary endings of isolated mammalian muscle spindles, .L Neurophysiol., 39 (1976) 324-330. Hunt, C.C. and Ottoson, D., Responses of primary and secondary endings of isolated mammalian muscle spindles to sinusoidal length changes, ./. Neurophysiol., 40 (1977) 1113-1121. Hunt, C.C., Wilkinson, R.S. and Fukami, Y., Ionic basis of the receptor potential in primary endings of mammalian muscle spindles, J. Gen. Physiol., 71 (1978) 683-698. Kaila, K., Rydquist, B., Swerup, C. and Voipio, J., Stimulation induced changes in the intracellular sodium activity of the crayfish stretch receptor, Neurosci. Lett., 74 (1987) 53-57. Kr611er, J., Griisser, O.-J., Weis, L.-R., A study of encoder properties of muscle spindle primary afferent fibres by a random noise disturbance of the steady stretch response, BioL Cybern., 63 (1990) 91-97. Kruse, M.N. and Poppele, R.E., Electrical 'tuning' in muscle spindle receptors, Brain Res., 496 (1989) 303-306. Kruse, M.N. and Poppele, R.E., Components of the dynamic response of mammalian muscle spindles that originate in the sensory terminals, Exp. Brain Res., 86 (1991) 359-366. Lancaster, B. and Adams, P.R., Calcium-dependent current generating the afterhyperpolarization of hippocampal neurons, J. Neurophysiol., 55 (1986) 1268-1282.

[18] Matthews, P.B.C. and Stein, R.B., The sensitivity of muscle spindle afferents to small sinusoidal changes of length, J. Physiol., 200 (1969) 723-743. [19] Ottoson, D. and Shepherd, G.M., Relation of afferent nerve excitability to impulse generation in the frog spindle, Acta Physiol. Scand., 75 (1969) 49-63. [20] Ottoson, D., McReynolds, J.S. and Shepherd, G.M., Sensitivity of isolated frog muscle spindle during and after stretching, J. NeurophysioL, 32 (1969) 24-34. [21] Ottoson, D. and Swerup, C., Ionic dependence of early adaptation in the crustacean stretch receptor, Brain Res., 336 (1985) 1-8. [22] Ottoson, D. and Swerup, C., Effects of intracellular TEA injection an early adaptation of crustacean stretch receptor, Brain Res., 336 (1985) 9-17. [23] Poppele, R.E. and Chen, W.J., Repetitive firing behavior of mammalian muscle spindle, J. Neurophysiol., 35 (1972) 357-364. [24] Poppele, R.E. and Quick, D.C., Stretch induced contraction of intrafusal muscle in cat muscle spindle, J. Neurosci., 1 (1981) 1069-1074. [25] Proske, U., The case of multiple sites of impulse initiation in muscle spindles and tendon organs. In W. Hamann and A. Iggo (Eds.), Sensory Receptor Mechanisms, 1984, pp. 133-143. [26] Rudy, B., Diversity and ubiquity of the K channels, Neuroscience, 25 (1988) 729-749. [27] Rydquist, B. and Zou, J.-H., Potential dependent potassium currents in the slowly adapting stretch receptor neuron of the crayfish, Acta Physiol. Scand., 137 (1989) 409-419. [28] Rydquist, B. and Swerup, C., Stimulus-response properties of the slowly adapting stretch receptor neuron of the crayfish, Acta Physiol. Scand., 143 (1991) 11-19. [29] Sachs, F., Mechanical transduction in biological systems, CRC Crit. ReL,. Biomed. Eng., 16 (1988) 141-'169. [30] Sch~ifer, S.S., Oscillations in the discharge frequency of cat primary muscle spindle afferents during the dynamic phase of a ramp-andhold stretch, Brain Res., 686 (1995) 61-69. [31] Sch~ifer, S.S., Simulation of dynamic fusimotor cffects in the discharge frequency of Ia afferents by prestretching the muscle spindle, Exp. Brain Res., (1995) in press. [32] Stoev, G. and Goll, M., Erregbarkeitsiinderungen der Muskelspindel in Abhiingigkeit t~on ihrer Liinge, Dissertation an der Medizinischen Hochschule Hannover, Abteilung Neurophysiologie. 1989. [33] Swerup, C., Rydquist, B. and Ottoson, D., Time characteristics and potential dependence of early and late adaptation in the crustacean stretch receptor, Acta Physiol. Scand., 119 (1983) 91-99.