Effects of chlorobutanol on primary and secondary endings of isolated cat muscle spindles

Effects of chlorobutanol on primary and secondary endings of isolated cat muscle spindles

Brain Research 854 Ž2000. 106–121 www.elsevier.comrlocaterbres Research report Effects of chlorobutanol on primary and secondary endings of isolated...

538KB Sizes 0 Downloads 0 Views

Recommend Documents

No documents
Brain Research 854 Ž2000. 106–121 www.elsevier.comrlocaterbres

Research report

Effects of chlorobutanol on primary and secondary endings of isolated cat muscle spindles M. Fischer

)

Department of Neurophysiology (Unit 4230), HannoÕer Medical School, Carl-Neuberg-Str. 1, D-30625 HannoÕer, Germany Accepted 2 November 1999

Abstract The effects of the preservative chlorobutanol on primary and secondary endings of muscle spindles isolated from the tenuissimus muscle of the cat were investigated in this study. Chlorobutanol was applied to the bathing solution in final concentrations of between 10 and 100 mgrml. It induced a reversible and dose dependent decrease in the discharge frequency of both types of ending without any visible length change in the sensory region of the receptor. The initial activity, the peak dynamic discharge, the maximum static discharge value and the final static discharge value were evaluated from an ending’s discharge pattern obtained during ramp-and-hold stretches. These four basic discharge frequencies decreased in parallel with increasing concentrations of chlorobutanol. Their sensitivities to chlorobutanol were similar Žmean values: y0.11 to y0.29 imprs per mgrml chlorobutanol. and were independent of the amplitude of stretch. The dynamic response and the static response of both primary and secondary endings remained unchanged, indicating that the sensitivity of the spindle to stretch was not influenced by chlorobutanol. Chlorobutanol also reduced the discharge activity of the muscle spindle afferents during sinusoidal stretches. The amplitude of the receptor potential ŽAC component. remained unchanged under chlorobutanol. With the available recording technique it was not possible to measure slow shifts of the membrane potential. However, a hyperpolarization of the ending’s membrane might explain why the afferent discharge frequency is reduced by chlorobutanol. The calcium dynamics of the spindle do not appear to be altered by CB, as the effect exerted on the afferent discharge by a change in the extracellular calcium concentration and a blockage of calcium channels was different from the CB effect. As the inhibitory effect of CB was reduced by ouabain, it is possible that CB activates the electrogenic NarK pump or affects a mechanism that is closely related to the activity of the pump. The properties of the axonal membrane appear not to be altered, as chlorobutanol did not change the shape of action potentials. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Isolated muscle spindle; Ramp-and-hold stretch; Chlorobutanol; Calcium; Electrogenic NarK pump

1. Introduction The preservative chlorobutanol ŽCB s 1,1,1-tri-chloro2-methyl-2-propanol. is added to various pharmaceutical preparations because of its antibacterial and antifungal properties. However, this substance induces diverse inhibitory side effects on smooth muscle cells w2,3x, myocardial cells w1,10,24x and several other types of tissue w8,26,28–30x. The mechanisms underlying, the inhibitory effects of CB are not fully understood. Hermsmeyer and Aprigliano w10x demonstrated a prolongation of action potentials and a dysrhythmogenic effect of CB on amphib-

)

Fax: q49-511-532-2776

ian myocardial cells. Arimura et al. w1x showed that CB reduced the amplitude of slow action potentials and the contractile tension of the partially depolarized papillary muscle in guinea pigs. This effect was reversed by an elevation of the extracellular calcium concentration. The authors assumed that CB induced a reduction of an inward calcium current. Barrigon et al. w3x observed a decrease in the 45 Ca influx and an increase in the 45 Ca efflux in rat aortic strips that were dependent on the presence of CB. Contractile responses induced by noradrenaline and potassium chloride were reduced by CB. Habara and Kanno w8x demonstrated that the intracellular calcium dynamics and the secretory response of isolated pancreatic acini of the rat were altered by CB. The authors discussed an influence of CB on receptor proteins, ionic channels or carriers and an effect on intracellular signal processing.

0006-8993r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 2 3 2 5 - 2

M. Fischerr Brain Research 854 (2000) 106–121

The aim of this study was to investigate the effect of CB on the isolated cat muscle spindle. We recorded the receptor’s afferent discharge when CB was added to the bathing solution in varying final concentrations. Even at very low doses of CB Ž10–100 mgrml. the afferent discharge was clearly reduced. Furthermore, the influence of CB on the spindle’s sensitivity to stretch was studied by stimulating the receptor with ramp-and-hold and sinusoidal stretches. Additionally, we investigated whether CB affects the calcium dynamics or the electrogenic NarK pump of the muscle spindle.

107

tude, because the signal size was heavily dependent on the recording conditions Žsignal amplitude of the receptor potential: F 10 mV; signal amplitude of action potentials: F 500 mV.. In particular, the signal amplitude was affected by the amount of extracellular fluid that covered the nerve in the oil chamber. The best recording conditions were thus achieved when the nerve was completely freed from connective tissue, since it was that tissue that mainly increased the aqueous layer surrounding the nerve. Experiments were carried out 1 h after the completion of preparation. 2.2. Stretching the isolated spindle

2. Materials and methods 2.1. Spindle preparation and recording technique The preparation of isolated muscle spindles was similar to that previously described by Fischer and Schafer ¨ w4x. The tenuissimus muscle and its nerve supply were excised from cats anesthetized with sodium pentobarbital Ž45 mgrkg i.v... The excised tissue was transferred into a modified Ringer’s solution with the following composition: NaCl 118.6 mM; KCl 4.75 mM; CaCl 2 1.80 mM; NaHCO 3 23.2 mM; KH 2 PO4 1.19 mM; MgSO4 0.84 mM; glutamine 2.40 mM; glycine 3.20 mM; histidine 0.97 mM; glutamic acid 1.02 mM; glucose 1 grl w20x. The solution was continuously aerated with 95% O 2 and 5% CO 2 . The pH was adjusted to 7.4. The temperature was kept constant at 358C. The isolation of a muscle spindle and its nerve supply was effected by removing the attached extrafusal fibers and connective tissue. The spindle was not isolated over its full length since a number of extrafusal fibers were usually left around the poles of the intrafusal muscle fibers. The isolated spindle was transferred into an experimental chamber that was continuously perfused with Ringer’s solution. Then the poles of the spindle were fixed to the holding rods of a stretching device, using a histoacryl glue ŽB. Braun-Melsungen.. The afferent nerve of the spindle was drawn into an oil-filled chamber and placed on an Ag electrode for the extracellular recording of discharges. The reference electrode was placed in the Ringer’s solution close to the spindle. Action potentials of the muscle spindle endings were amplified ŽGrass P511 AC pre-amplifier. and recorded on tape. Using the same technique it was possible to record the AC component of the receptor potential of muscle spindle endings while action potentials were blocked by lidocaine. It should be mentioned that frequencies lower than 0.1 Hz could not be reproduced by the AC pre-amplifier, so that slow components of the receptor potential could not be recorded. The graphs accompanying this study that show the receptor potential andror action potentials do not include a vertical scale for the signal ampli-

Isolated muscle spindles were stretched by the movements of two holding rods fixed to the membranes of two loudspeakers. The movements of these membranes were driven by DrA converted signals from a personal computer and controlled with the aid of photocells which converted the intensity of a beam of light reflected from mirrors attached to the membranes into a DC signal. The initial length of a spindle corresponded to its length in situ when the tenuissimus muscle was slightly prestretched Žangle of femurrtibia joint: 1358.. L0 was defined as the part of the spindle that lay between the holding rods. Values of L 0 were 3–5 mm. The ramp-andhold stretches had amplitudes of 2.5–10% of L 0 . The velocity of the ramp phase was 40% of L0 per second. The plateau phase was held for a period of 3 s. There was a pause of 10–12 s between individual stretches. Sinusoidal stretch was applied in a few experiments, with frequencies ranging from 10 to 400 Hz and amplitudes between 5 and 200 mm. 2.3. Treatment with bioactiÕe agents CB ŽSigma. was added to the Ringer’s solution to give final concentrations of 10, 25, 50, 75 and 100 mgrml. The solutions containing CB were applied to the isolated muscle spindle by switching the source of influx into the experimental chamber. With the chamber having a volume of 0.5 ml and a perfusion rate of 3–5 mlrmin, the bathing fluid was exchanged in 10 s at the most. CB induced effects on the discharge frequency of muscle spindle afferents were measured 3 min after the exchange of the solutions. CB was always completely washed out before a new concentration was tested. In a few experiments a bathing fluid containing 0.2% lidocaine hydrochloride ŽSigma. was used to block the generation of action potentials, in order to investigate effects of CB on the receptor potential of the muscle spindle afferents. The concentration of extracellular calcium ions was varied in some experiments by substituting CaCl 2 for an isoosmolar concentration of NaCl and vice versa. In further tests the calcium channel blockers nifedipine, verapamil or D600 ŽSigma.

108

M. Fischerr Brain Research 854 (2000) 106–121

were added to the Ringer’s solution. Ouabain ŽSigma. was used to inhibit the NarK pump. 2.4. EÕaluation of afferent discharges Primary and secondary muscle spindle endings were distinguished by criteria based on their different physiological properties w4x. Primary but not secondary endings discharged with each cycle of sinusoidal stretch when the stretch amplitude was 5 mm and the frequency was increased from 10 to 100 Hz. During ramp-and-hold stretches primary endings showed an initial burst and a postdynamic minimum, and often fell silent during the release of the stretch; with secondary endings these characteristic features did not occur. The extracellularly recorded action potentials of primary endings were often larger than the action potentials of secondary endings, as a consequence of their differing axon diameters. The extracellularly recorded action potentials of the muscle spindle endings were evaluated as follows. Discharge patterns as shown in Fig. 2c Žfilled circles. were built up by determining the instantaneous discharge frequencies of one ending and superimposing that ending’s responses to five successive ramp-and-hold stretches. Four basic discharge frequencies were taken from such discharge patterns. The initial activity ŽIA. was the median value of the instantaneous discharge frequencies during the last 500 ms before the start of the stretch. The peak dynamic discharge ŽPD. was the median value of the discharge frequencies during the last 25 ms of the ramp phase. The maximum static value ŽMST. was the highest discharge frequency during the plateau of the stretch. MST was evaluated as the median frequency over a period of 50 ms, beginning 20 ms after the start of the plateau phase. Sometimes the highest discharge frequencies were found up to 70 ms later in the plateau phase; in such cases the period of evaluation was shifted accordingly. MST represents a static value that characterizes the spindle’s impulse activity before the beginning of a slow adaptive process which continuously diminishes the firing rate during the plateau phase of the stimulus. Thus MST divides the full adaptive decay of the discharge frequency into a fast component and a slow component w25x: The fast adaptive decay is represented by the decrease in the discharge frequency from PD to MST and depends mainly on the velocity of the ramp. The slow decay is represented by the decrease from MST to the final static value ŽFST. and depends mainly on the amplitude of the stretch. FST was evaluated as the median value of the discharge frequencies during the last 250 ms of the plateau phase. The responses were quantified using median values rather than mean values because the median values provided a very good depiction of the response characteristics even when single impulses occasionally caused large deviations of the instantaneous discharge frequency from the firing rate usually observed.

3. Results 3.1. Effects of CB on muscle spindle endings at rest and during ramp-and-hold stretches Fig. 1 shows the effect of 25, 50, 75 and 100 mgrml CB on the resting discharge of a secondary ending of an isolated muscle spindle. The four curves that are superimposed in this diagram represent the responses of the ending to the four concentrations of CB. Each curve displays the resting discharge frequencies measured at intervals of about 13 s. In each experiment CB was added over a period of 4 min Žhorizontal bar at the top of the diagram.. With CB the afferent discharge frequency decreased without any visible length change in the sensory region of the receptor. The discharge frequency was only slightly reduced when CB was added to the bathing fluid at a concentration of 25 mgrml, but it declined from 35 to 22 imprs when a final concentration of 100 mgrml CB was tested. The reduction of the discharge frequency was totally reversible irrespective of the dose of CB: when the bathing solution was replaced by normal Ringer’s solution recovery took only a few minutes and the previous discharge frequency of about 35 imprs was restored. The resting discharge of most muscle spindle endings is dependent on the pre-stretch of the receptor, and the reduction of the firing rate under CB might reflect a change in the ending’s stretch sensitivity. On the other hand, the resting discharge of at least some muscle spindle endings does not depend on the muscle pre-stretch w7x, and may depend on a stretch independent conductance of the sensory ending membrane. We therefore investigated the effect of CB on the stretch sensitivity of muscle spindle

Fig. 1. Effect of chlorobutanol ŽCB. on the resting discharge of a secondary ending of an isolated muscle spindle. The four curves show the resting discharge frequencies that were measured at intervals of about 13 s when 25, 50, 75 and 100 mgrml CB were added to the bathing fluid Žsee bar at the top of the diagram.. Symbols for different concentrations of CB are depicted below the diagram.

M. Fischerr Brain Research 854 (2000) 106–121

endings under ramp-and-hold stretches of different amplitudes. The influence of CB on the resting discharge was compared with the influence of CB on the discharge frequency during a stretch. Fig. 2 represents the effect of 100 mgrml CB on a primary ending Ža; c. and on a secondary ending Žb; d. under ramp-and-hold stretches. The diagrams in a and b show the instantaneous discharge frequencies of the endings when the isolated spindle was repeatedly stimulated by stretches with an amplitude of 5% of the spindle’s initial length L0 . The stretches are depicted in the bottom line under each diagram. The horizontal bar at the top of each diagram represents the period of exposure to CB. The replacement of the bathing fluid with a solution containing CB resulted in a decrease in the afferent discharge fre-

109

quency of both the primary and the secondary ending within the first 2–3 min. Then a new and stable level of activity was reached. Before CB was added, the primary ending fired with a resting discharge frequency of nearly 50 imprs in each pause between individual stretches ŽFig. 2a.. With CB the resting activity declined continuously to about 30 imprs and then dropped to 0 imprs. However, firing did not stop during the stretches. In Fig. 2c the responses of the primary ending to the first five ramp-and-hold stretches of Fig. 2a were superimposed using filled circles, and additionally the responses to the last five stretches of Fig. 2a were superimposed in the same diagram using open circles. It may be estimated from this diagram that CB induced a decrease in the discharge frequency of about 20

Fig. 2. Effect of CB on a primary Ža; c. and on a secondary Žb; d. ending of one isolated muscle spindle. Ža and b. Instantaneous discharge frequencies showing the decrease of the ending’s discharge frequency when 100 mgrml CB was added to the bathing solution Žsee bar at the top of each diagram.. The spindle was repeatedly stimulated by ramp-and-hold stretches as depicted beneath each diagram Žamplitude of stretch: 5% of the spindle’s initial length L0 .. Žc and d. Discharge patterns obtained from the data of Ža and b., superimposing the responses to five successive stretches in normal Ringer’s solution Žfilled circles. and in Ringer’s solution containing 100 mgrml CB Žopen circles.. IA: initial activity; PD: peak dynamic discharge; MST: maximum static value; FST: final static value.

110

M. Fischerr Brain Research 854 (2000) 106–121

imprs during a stretch. The course of activity changes during the ramp and during the plateau phase of the stretch remained unchanged under CB. However, the absence of the resting activity before and after the stretch was a striking effect of CB on this primary ending. The resting activity of the secondary ending did not cease when the muscle spindle was exposed to 100 mgrml CB ŽFig. 2b and d.. It declined from about 60 imprs to about 45 imprs. The instantaneous discharge frequency during the stretches declined by nearly the same degree. Therefore the course of the discharge frequency in normal Ringer’s solution ŽFig. 2d, filled circles. and in a solution containing 100 mgrml CB Žopen circles. changed roughly in parallel. The effect of CB was tested on 16 primary endings and 21 secondary endings under ramp-and-hold stretches. In general, CB induced a reversible reduction of the afferent discharge frequency and the previous discharge pattern was completely restored when the solution containing CB was replaced by normal Ringer’s solution. Under CB the basic discharge frequencies IA, PD, MST and FST Žsee Fig. 2c. either showed new activity levels, or else dropped to 0 imprs if the firing rate decreased below a frequency of about 10–20 imprs. This critical discharge frequency was different for each individual ending and showed no significant difference between primary and secondary endings. The abrupt cessation of firing at a critical discharge frequency is a well known feature of the impulse activity of muscle spindle afferents and other mechanoreceptors w6,16x. It is not an effect of CB and can be observed in normal Ringer’s solution as well. However, this behavior emphasized the inhibitory effect of CB where the overall activity of an ending was low. The reduction of each basic discharge frequency was dependent on the concentration of CB. The dose–response relationship was studied in eight primary and 10 secondary endings under four different amplitudes of ramp-and-hold stretch. The amplitudes of the stretches were 2.5% L0 , 5% L0 , 7.5% L0 and 10% L0 . The diagrams in Fig. 3 show the dependence of the impulse activity of primary endings Ža. and secondary endings Žb. on the concentration of CB when the spindles were repeatedly stretched by 2.5% L0 . Mean values of the four basic discharge frequencies PD, MST, FST and IA were plotted against the concentration of CB. The standard deviations included in this figure mainly depended on the different specific levels of activity of each individual ending. Therefore they do not display a valid estimation of the variance of CB effects. For the sake of greater clarity the symbols for MST and FST are plotted with a slight horizontal displacement. The curves in Fig. 3 show that on average the four basic discharge frequencies of primary Ža. and secondary Žb. endings decreased dose dependently in an almost linear manner. The slopes of these curves represent the sensitivity of each basic discharge frequency to CB. For purposes of comparison the gradients of the linear regression lines

Fig. 3. Dependence of the mean values of the four basic discharge frequencies IA, PD, MST and FST on the concentration of CB. Ža. Primary endings Ž ns8.; Žb. secondary endings Ž ns10.. The basic discharge frequencies of both types of ending decrease in a dose dependent manner. The symbols of MST and FST were plotted with a slight horizontal displacement. ŽAbbreviations: see Fig. 2..

were calculated from the data of each curve. The four curves of Fig. 3a had negative gradients of between y0.19 and y0.21 imprs per mgrml CB. Thus in primary endings the decrease in each of the four basic discharge frequencies was about 0.2 imprs when the CB concentration was increased by one mgrml. In secondary endings ŽFig. 3b. the impulse activity fell by nearly the same amount ŽPD: y0.19 imprs per mgrml CB; MST: y0.18 imprs per mgrml CB; FST: y0.18 imprs per mgrml CB; IA: y0.21 imprs per mgrml CB.. What the mean values plotted in Fig. 3 do not show is that the basic discharge frequencies of a number of individual endings dropped to 0 imprs when CB was added: Three out of eight primary endings Ža. did not show an IA even in normal Ringer’s solution. With 50 mgrml CB the IA was 0 imprs in a further two primary endings, and it fell to 0 imprs in all the primary endings when the spindles were exposed to 100 mgrml CB. The other basic discharge frequencies that were obtained during a stretch also dropped to 0 imprs if the firing rate decreased below the critical discharge frequency Ž10–20 imprs. of the

M. Fischerr Brain Research 854 (2000) 106–121

ending concerned. Thus five out of eight primary endings did not show an FST in 100 mgrml CB. The PD of two endings and the MST of three endings were also 0 imprs at this concentration of CB. With secondary endings Žb. a similar response to CB was observed: the higher the concentration of CB, the more endings showed a fall of their IA to 0 imprs. At a concentration of 100 mgrml IA fell below the critical discharge frequency in seven out of 10 secondary endings. PD was 0 imprs in two cases. MST and FST dropped to 0 imprs in three endings. With ramp-and-hold stretches of larger amplitude Ž5– 10% L0 ., firing occurred during the stretches in every primary and secondary ending. At least the PD exceeded the critical discharge frequency, even when a CB concentration of 100 mgrml was used. However, in these cases the gradients of the dose–response curves were smaller Žy0.11 to y0.15 imprs per mgrml CB.. This indicates that the slopes of the curves in Fig. 3 might have been rendered steeper by the sudden drop in the activity of several endings from their critical discharge frequency to 0 imprs. Because of this supposition the gradients of the dose–response curves were reinvestigated and all readings of 0 imprs were excluded from the calculation. The results are shown in Fig. 4a for primary endings and in Fig. 4b for secondary endings. The mean values and standard deviations of the gradients of the dose–response curves, i.e., the sensitivity of the basic discharge frequencies to CB, were plotted against the amplitudes of the ramp-and-hold stretches. The white bars in Fig. 4a represent the sensitivity of IA of primary endings to CB. The values were in the range y0.25–y 0.29 imprs per mgrml CB and were higher than the sensitivities of PD, MST and FST. The latter are illustrated using filled bars. The values for PD, MST and FST were in the range y0.11–y 0.17 imprs per mgrml CB. There was one exception: The sensitivity of FST was y0.21 imprs per mgrml CB when the amplitude of the stretch was 7.5% L0 . An analysis of variance ŽScheffe´ test. indicated that the sensitivity of IA was not significantly different from the other sensitivities when the data of each stretch amplitude was separately analysed. The sensitivities to CB being mainly independent of the amplitude of stretch, it is reasonable to combine the values obtained under different amplitudes of stretch in order to enlarge the number of samples per group. In this case an analysis of variance showed that CB affected IA significantly more strongly than the other basic discharge frequencies ŽScheffe´ test: P - 0.05.. With secondary endings ŽFig. 4b., no significant difference was observed between the sensitivity of IA and that of PD, MST and FST Žrange: y0.13 to y0.22 imprs per mgrml CB.. But once again, the sensitivities of the basic discharge frequencies to CB were independent of the amplitudes of the ramp-and-hold stretches, even though the values varied, within a narrow range, between one amplitude and another.

111

Fig. 4. Mean values and standard deviations of the endings’ sensitivity to CB evaluated for the basic discharge frequencies IA Žwhite bars., PD Ždotted bars., MST Žhatched bars. and FST Žstriped bars. and plotted against the amplitude of stretch; Ža. includes the data of eight primary endings; Žb. includes the data of 10 secondary endings. In primary endings only, the sensitivity of IA was significantly different from the sensitivities of the other basic discharge frequencies Žsee text for further explanation.. In both types of ending, the sensitivities of all four basic discharge frequencies were independent of the amplitude of stretch. ŽAbbreviations: see Fig. 2..

This finding demonstrates indirectly that the sensitivity of the spindle to stretch was not influenced by CB. In order to confirm this result, the dynamic response ŽDR. and the static response ŽSR. of the muscle spindle endings were evaluated ŽFig. 5.. DR is the difference between PD and FST and represents the dynamic properties of an ending. SR describes the static properties of an ending and is calculated as the difference between FST and IA. Fig. 5 shows the mean values and standard deviations of DR and SR of primary Ža. and secondary Žb. endings, plotted against the concentration of CB. The different filling patterns of the bars correspond to different amplitudes of ramp-and-hold stretches. The figure displays the data from eight primary and 10 secondary endings. Readings of 0 imprs have been eliminated. Consequently, at a concentration of 100 mgrml CB no SR values appear in Fig. 5a, as IA dropped to 0 imprs in all primary endings. The two

112

M. Fischerr Brain Research 854 (2000) 106–121

velocity and a slow adaptive decay from MST to FST that is affected by the stretch amplitude w25x. These two components of adaptation were also separately investigated. However, neither the fast decay nor the slow decay were influenced by CB in either type of ending Žnot shown.. Fig. 5 additionally demonstrates that the SR of both primary and secondary endings remained unaffected when the concentration of CB was varied between 0 mgrml and 100 mgrml. The mean values of SR increased with increasing amplitudes of stretch in both types of ending, but for each amplitude of stretch SR remained constant under different concentrations of CB. These results confirm that the static sensitivity of the muscle spindle endings, i.e., the change in the discharge frequency ŽSR. per unit of change in the spindle length Žamplitude of stretch., was not influenced by CB. If the static sensitivity to stretch were affected by CB, one would expect a change in SR when the concentration of CB was varied and the amplitude of stretch was kept constant. Similarly the dynamic sensitivity to stretch Žimprs per unit of the stretch velocity. remained unchanged under CB, because DR and in particular the fast adaptive decay did not change when CB was varied and the ramp velocity was kept constant. 3.2. Effects of CB on muscle spindle endings during sinusoidal stretch

Fig. 5. Mean values and standard deviations of the dynamic response ŽDR. and static response ŽSR. of eight primary Ža. and 10 secondary Žb. endings plotted against the concentration of CB. DR and SR were evaluated under four different amplitudes of stretch: 2.5% L0 Žwhite bars.; 5% L0 Ždotted bars.; 7.5% L0 Žhatched bars.; 10% L0 Žstriped bars.. SR was not calculated where IA was 0 imprs. DR was not calculated where FST was 0 imprs. Therefore, mean values of SR are missing in Ža. where IA was 0 imprs for each of the eight primary endings. DR and SR of primary and secondary endings were not affected by CB.

bars representing SR at 75 mgrml CB were calculated from a single primary ending, the only one that was active during the pauses between individual stretches at this concentration of CB. The mean DR value of primary endings increased very slightly with an increasing concentration of CB, especially when the amplitude of stretch was 2.5% of L0 Žwhite bars in Fig. 5a.. However, the DR values at 0 mgrml CB and at 100 mgrml CB were not significantly different ŽStudent’s t-test: P ) 0.3 for each amplitude of stretch.. With secondary endings too ŽFig. 5b., DR remained unaffected when the concentration of CB was varied. DR breaks down into a fast adaptive decay of the discharge frequency from PD to MST that is mainly affected by the stretch

In a few experiments, the effect of CB on the discharge frequency of muscle spindle afferents during sinusoidal stretches was investigated. The results taken from these studies confirmed the finding that CB reduces the impulse activity of muscle spindle endings. The primary ending of a spindle that was stretched sinusoidally with an amplitude of 10 mm and a constant frequency of 60 Hz fired in normal Ringer’s solution with one action potential per cycle of stretch: occasionally one stretch failed to elicit an impulse. When the bathing fluid was replaced by a solution containing 100 mgrml CB, the impulse activity decreased to one discharge per 10 cycles of sinusoidal stretch. As in normal Ringer’s solution, the action potentials remained phase locked in respect of the sinusoidal stretch. Further experiments were done with another primary ending, using a larger amplitude of sinusoidal stretch Ž50 mm. and increasing the frequency of mechanical stimulation from 10 to 400 Hz in steps of 10 Hz and with a step duration of 2 s ŽFig. 6.. The amplitude of stretch was kept constant during the whole experiment. Fig. 6a shows the instantaneous discharge frequency of the ending plotted against time when the spindle was immersed in normal Ringer’s solution. A second scale has been added beneath the diagram to represent the frequency of stimulation. As long as the stimulation frequency was below 250 Hz, one action potential occurred during each cycle of stretch. With stimulation frequencies of between 290 and 350 Hz, there was only one action potential to every two cycles of

M. Fischerr Brain Research 854 (2000) 106–121

113

Fig. 6. Comparison of the impulse activity of a primary ending in a normal Ringer’s solution Ža;b. and in a Ringer’s solution containing 100 mgrml CB Žc;d. when the spindle is sinusoidally stretched using a stimulation frequency increasing stepwise from 10 Hz to 400 Hz Žstep size: 10 Hz; step duration: 2 s.. Ža and c. Instantaneous discharge frequencies plotted against time and against the stimulation frequency. Žb and d. Sequence of action potentials obtained at a stimulation frequency of 400 Hz marked by arrows in Ža and c.. CB reduced the discharge activity in the high frequency range.

stretch. When the stimulation frequency was within the ranges of 260–280 Hz and 360–400 Hz, the ending discharged either in every cycle of stretch or else in every other cycle. Fig. 6b shows a sequence of action potentials that were evoked by sinusoidal stimulation at a frequency of 400 Hz Žsee arrow in Fig. 6a.. Some stimuli failed to elicit an action potential, so that the ratio of impulses to stimuli varied between 1:1 and 1:2. When CB was added to the bathing solution at a final concentration of 100 mgrml, the activity of the ending was clearly reduced ŽFig. 6c and d.. At low stimulation frequencies, the strength of the sinusoidal stretch was enough to elicit one action potential per cycle, as in normal Ringer’s solution. However, at stimulation frequencies above 250 Hz the impulse frequency progressively declined. At a stimulation frequency of 400 Hz Žarrow in Fig. 6c;d. only an average of one action potential to every nine cycles of stretch occurred. The reduction of the excitability of the ending was reversible when CB was washed out of the bathing fluid.

With sinusoidal stretches of large amplitudes it was also possible to investigate the receptor potential of some muscle spindle endings. Unfortunately, slow DC shifts of the receptor potential could not be studied with the available pre-amplifier Žsee under Methods.. However, it was possible to record the AC component of the receptor potential at a very low high-pass-filter frequency of 0.1 Hz. Results from a primary ending are shown in Fig. 7. The isolated muscle spindle was sinusoidally stretched by 200 mm, i.e., 5% of its initial length L0 . The frequency of stimulation was 10 Hz. The stimulus is depicted in the lowest curve of the figure. With this amplitude of sinusoidal stretch the ending discharged at a rate of three impulses per cycle of stretch, as shown in the top section of Fig. 7. The action potentials have been curtailed in order to focus on the fluctuations of the receptor potential when a high amplification factor is used. The second curve of Fig. 7 shows the receptor potential while the action potentials were blocked by 0.2% lidocaine. In the third section of the figure the

114

M. Fischerr Brain Research 854 (2000) 106–121

ent action potentials in a graduated manner before firing totally stopped. By blocking sodium channels lidocaine inhibits the generation and active propagation of action potentials. However, lidocaine could not directly reach the axon membrane at the recording site in the oil-filled chamber of our experimental apparatus. We suppose that the signals recorded were due to a passive electrotonic spread of action potentials that were elicited at a node of Ranvier immersed in the solution of the experimental chamber. In view of these observations it should be possible to study the influence of CB on the axon membrane as well, even if the recording site is not directly accessible to CB. We compared the shape of the action potentials while the isolated receptor was bathed in normal Ringer’s solution on the one hand and in Ringer’s solution containing 100 mgrml CB on the other, but neither the amplitude nor the shape of the action potentials changed when CB was added to the bath. 3.4. Effects of CB on the calcium dynamics of muscle spindle endings Fig. 7. Effect of CB on the receptor potential ŽAC component. of a primary ending during sinusoidal stretch. Top line: Oscillogram representing the ending’s response to the stretch stimuli in normal Ringer’s solution. Action potentials have been curtailed. Second line: Receptor potential recorded after the generation of action potentials has been blocked by adding 0.2% lidocaine to normal Ringer’s solution. Third line: Receptor potential recorded with the spindle immersed in Ringer’s solution containing 0.2% lidocaine and 100 mgrml CB. Each record is an average of responses to 25 sinusoidal stretches. CB did not change the amplitude or the phase advance Žarrows. of the receptor potential. Bottom line: Sinusoidal stretch Ž200 mm, 10 Hz..

receptor potential is shown while the spindle was immersed in a solution containing 0.2% lidocaine and 100 mgrml CB. Each record is an average of the responses to 25 sinusoidal stretches. The amplification factor is identical in each case and recording conditions did not change. The fluctuations of the receptor potential were synchronous with the sinusoidal stimulation. It is to be noted that the receptor potential showed a phase advance of 50–608 relative to the sinusoidal stretch Žarrows.; this phase advance was not affected by CB. However, the main result of this experiment was that the amplitude of the receptor potential remained unchanged when CB was added to the bathing fluid. But it should be emphasized again that only the AC component of the receptor potential was tested and that the existence of slow shifts of the membrane potential cannot be ruled out with the recording technique available Žsee Section 4.. 3.3. Effect of CB on the shape of action potentials During the experiments on the receptor potential we observed that lidocaine diminished the amplitude of affer-

Various investigations using non-neural tissue have shown that the effect of CB is most probably related to changes in the transmembrane calcium dynamics w1,3,8x. As far as the role of calcium in isolated muscle spindles is concerned, CB might influence the discharge frequency of the sensory endings in at least two different ways. Ž1. CB might affect the exchange of calcium across the spindle capsule, so that the ionic milieu of the capsular space is changed and the excitability of the sensory endings might be influenced. Ž2. The conductance or open probability of calcium channels in the sensory ending membrane itself might be changed by CB. Both of these hypotheses were tested. First the calcium concentration of the bathing solution was changed from 1.8 mM in normal Ringer’s solution to 0.9 or 2.7 mM. However, a change in the calcium concentration hardly affected the afferent discharge as long as the spindle capsule was intact. Opening the capsule by microdissection drastically increased the sensitivity of the sensory endings to changes in the extracellular calcium concentration. Thus the spindle capsule represents an effective barrier for calcium ions. Increasing the calcium concentration reduced the afferent discharge frequency and decreasing the calcium concentration increased the firing rate. Fig. 8 shows discharge patterns of a secondary ending obtained in Ringer’s solution containing 1.8 mM calcium Ža, filled circles. and 0.9 mM calcium Žb, filled circles.. The responses to five successive ramp-and-hold stretches were superimposed. Additionally, the responses to a further five ramp-and-hold stretches were included in each diagram showing the effect of CB Ž50 mgrml. on the same ending when CB was added to the solutions with the different calcium concentrations Ža and b, open circles..

M. Fischerr Brain Research 854 (2000) 106–121

115

Fig. 8. Discharge patterns of a secondary ending obtained under different extracellular calcium concentrations Ža: 1.8 mM; b: 0.9 mM.. Filled circles represent the superimposed responses of the ending to five successive ramp-and-hold stretches in solutions not containing CB. Open circles represent five responses when 50 mgrml CB was added to the bath solutions. The stretches are depicted beneath each diagram. A reduction in the extracellular calcium concentration increased the ending’s discharge rate and its sensitivity to stretch. CB reduced the discharge frequency by a constant degree during each phase of the stretch irrespective of the concentration of calcium.

The ramp-and-hold stretches are depicted in the bottom line under each diagram. The discharge patterns show that a reduction of the extracellular calcium concentration increased not only the spindle’s discharge frequency but also the ending’s sensitivity to stretch. It can be estimated from these discharge patterns that DR increased from 25 imprs to nearly 50 imprs and SR from about 17 imprs to nearly 25 imprs when the concentration of calcium was reduced, even though there was no change to the amplitude or the velocity of stretch. The figure additionally shows that CB remained effective when the spindle capsule was opened. CB did not counteract the change in the sensitivity to stretch that was induced by a reduction of the calcium concentration, but it did reduce the discharge frequency by a constant degree of about 10 imprs during each phase of the ramp-and-hold stretch irrespective of the concentration of calcium. Elevating the extracellular calcium concentration reduced the firing of muscle spindle endings and diminished their sensitivity to stretch Žnot shown.. SR and DR declined. Thus the extracellular calcium concentration influences the sensory activity in a different way than CB does. This finding was observed for both primary and secondary endings. In a further series of experiments the effect of CB on muscle spindle endings was compared with the effect of calcium channel blockers that were added to the Ringer’s solution. Fig. 9 demonstrates the effect of 10 mM nifedipine on the activity of a secondary ending. The ending was repeatedly stimulated by ramp-and-hold stretches and the basic discharge frequencies PD, MST and FST were plotted against time. The ending did not discharge in the pause between individual stretches. Therefore IA was 0 imprs during the whole experiment and has not been depicted.

The spindle was exposed to nifedipine during a period that is marked by the horizontal bar at the top of the diagram. The basic discharge frequencies MST and FST that represent the static properties of the muscle spindle ending declined to 0 imprs within 2 min after exposure to nifedipine, and recovered with several minutes delay when nifedipine was washed out. PD, however, which represents the dynamic properties of the ending, remained nearly unchanged during the experiment. In further experiments using increasing concentrations of nifedipine PD was also diminished. However, in comparison with the other basic discharge frequencies the reduction of PD was smaller, or its decline to 0 imprs took place over a different course of time. The calcium channel blockers D600 and verapamil

Fig. 9. Effect of 10 mM nifedipine on the basic discharge frequencies PD, MST and FST obtained from the responses of a secondary ending when the muscle spindle was repeatedly stimulated by ramp-and-hold stretches Žamplitude: 5% of L0 .. MST and FST declined to 0 imprs when nifedipine Žnife, see horizontal bar at the top of the diagram. was added to the bathing solution, but PD remained almost unchanged.

116

M. Fischerr Brain Research 854 (2000) 106–121

induced similar effects to those of nifedipine. To sum up: for both types of ending the effects of calcium channel blockers were clearly different from the inhibitory effect of CB, which generally reduced the four basic discharge frequencies in a parallel manner and by nearly the same degree. 3.5. Effects of CB on the electrogenic Na r K pump of muscle spindle endings The effect of CB might be due to a hyperpolarization of the sensory ending membrane that is independent of the stretch applied to the receptor. The electrogenic NarK pump is a good candidate to bear responsibility for the stretch independent decline of the afferent discharge fre-

quency for the following reasons. Ž1. A powerful influence of the NarK pump on the membrane potential has been shown for the crayfish stretch receptor w17,27x, for the isolated frog muscle spindle w21,22x and for the cat muscle spindle w11x. The electrogenic pump causes strong posttetanic hyperpolarization and a post-tetanic depression of the impulse activity following a high frequency discharge of the sensory ending. Ž2. Yamamoto et al. w31x localized the NarK pump in rat muscle spindles immunocyto-chemically, and found the highest concentration of this protein in the sensory membrane of the afferent nerves. Ž3. The NarK pump acts independently of the stretch applied to a muscle spindle. We tested the hypothesis by comparing the effect of CB in Ringer’s solution with the effect of CB in a Ringer’s

Fig. 10. Reduction of the CB effect when the electrogenic NarK pump was inhibited by ouabain. Ža. Effect of 25 mgrml CB and 1 mM ouabain on the basic discharge frequencies IA, PD, MST and FST obtained from the responses of a secondary ending when the muscle spindle was repeatedly stimulated by ramp-and-hold stretches Žamplitude: 5% of L 0 .. The periods of exposure to the different drugs are represented by horizontal bars at the top of the diagram. The first application of CB exerted a stronger inhibitory effect on the four basic discharge frequencies than the second application of CB under ouabain. Žb–d. Discharge patterns obtained from the ending’s responses to five successive ramp-and-hold stretches in a solution not containing CB Žfilled circles. and in a solution containing 25 mgrml CB Žopen circles.. The discharge patterns were obtained at the five different points in time during the experiment that are marked by arrows Ž1–5. in Ža.. Žb. Discharge patterns showing the effect of CB before ouabain was added Žarrows 1 and 2.. Žc and d. Discharge patterns showing the effect of CB after ouabain had been added Žarrows 3 and 4 in Žc.; arrows 4 and 5 in Žd... See text for further explanations.

M. Fischerr Brain Research 854 (2000) 106–121

solution containing ouabain that selectively inhibits the NarK pump. If CB did indeed activate the pump, the effect of CB should be reduced under ouabain. Unfortunately, ouabain itself irreversibly affected the spindle’s afferent discharge. A few minutes after the Ringer’s solution had been replaced by a Ringer’s solution containing 1 mM ouabain the discharge frequency of primary and secondary endings started to increase steadily. After about 10–15 min, the firing rate suddenly decreased, and then the firing ceased irreversibly. However, ramp-and-hold stretches occasionally evoked a single action potential at the beginning of the ramp phase. Very similar effects have been described by Holloway and Poppele who used strophanthidin to inhibit the NarK pump in cat muscle spindles w11x. Fig. 10 shows the results of a representative experiment before the activity declined. The four basic discharge frequencies IA, PD, MST and FST were obtained from responses of a secondary ending to repeated ramp-and-hold stretches and were plotted against time in Fig. 10a. We administered 25 mgrml CB before 1 mM ouabain was applied to the bath and compared its effect with the CB effect under ouabain before firing stopped. Horizontal bars at the top of the diagram represent the periods of exposure to ouabain and CB. From a qualitative point of view, Fig. 10a shows that the first application of CB affected the four basic discharge frequencies more strongly than the second application. For example, IA declined to 0 imprs during the first exposure to CB but not during the second application. Ouabain increased the four basic discharge frequencies after about 5 min of exposure. The response of the ending to the second application of CB appears to interfere with the response to ouabain. For a quantitative analysis of the experiment, discharge patterns of the ending were obtained at the five different times that are marked by arrows Ž1–5. in Fig. 10a. These discharge patterns are displayed in Fig. 10b–d. Filled circles represent superimposed responses to five successive ramp-and-hold stretches in a solution that does not include CB. Open circles represent five superimposed responses in a solution containing 25 mgrml CB. Fig. 10b shows the effect of CB before ouabain was applied to the bath Žarrows 1 and 2 in Fig. 10a.. PD, MST and FST decreased by 7–8 imprs when CB was added. IA declined from 30 to 18 imprs in one out of five responses. In the remaining four responses IA dropped to 0 imprs. Fig. 10c shows the effect of CB after 1 mM ouabain had been applied to the bathing fluid Žarrows 3 and 4 in Fig. 10a.. In this case CB reduced PD, MST and FST only slightly, by about 2 imprs. IA declined from 27 to 21 imprs. Thus the effect of CB was drastically diminished by ouabain. However, this immense reduction in the CB effect might be misleading because the decrease in the discharge frequency that was induced by the second application of CB appears to interfere with the beginning of an increase in the discharge frequency induced by ouabain Žarrow 4 in Fig. 10a.. This

117

increase in the discharge frequency might influence the result by reducing the difference between the discharge frequencies obtained at the arrows 3 and 4, making the CB effect as calculated from these discharge patterns erroneously small. For this reason the responses under CB and ouabain Žarrow 4. were additionally compared with responses that were obtained under ouabain after CB had been washed out Žarrow 5 in Fig. 10a.. The discharge patterns are shown in Fig. 10d. In this case the calculated CB effect is erroneously enlarged, because the discharge frequencies of the control responses Žfilled circles. are certainly increased by the influence of ouabain and the difference between the discharge frequencies obtained at the arrows 4 and 5 is not diminished but raised by the ouabain effect. Nevertheless, the CB effect calculated from these discharge patterns is still smaller than the CB effect calculated from the discharge patterns in Fig. 10b that were obtained before ouabain was added to the bathing solution. PD and MST decreased by about 5 imprs. FST declined by 7 imprs as in Fig. 10b and IA declined from 30 to 21 imprs. Similar results were obtained from two further secondary endings and two primary endings. A reduction of the CB effect by ouabain was obvious in each experiment but a quantitative evaluation of the finding was difficult. Using a lower dose of ouabain Ž0.1 mM. delayed both effects of the inhibitor, i.e., the increase in the firing rate and the reduction of the CB effect.

4. Discussion The results show that CB reduces the afferent discharge frequency of primary and secondary endings of isolated cat muscle spindles in a dose dependent manner ŽFig. 3.. At the same time, the receptor’s response to stretches of a constant amplitude and velocity remain unchanged, i.e., the sensitivity to stretch is not affected by CB. The experiments showed that in respect of ramp-and-hold stretches SR and DR remain constant under varying concentrations of CB ŽFig. 5., and that in respect of sinusoidal stretches the amplitude of the AC component of the receptor potential is not altered ŽFig. 7.. The results of Fig. 6 might be thought to indicate that it is the dynamic properties of the muscle spindle ending that are mainly affected by CB, as in this figure firing is preferentially reduced in the high frequency range of sinusoidal stretch. However, it is known from frog muscle spindle endings that the AC component of the receptor potential decreases and the DC component increases with an increasing frequency of a sinusoidal stimulation, resulting in a decrease in the peak depolarization when the stimulation frequency is raised and the muscle spindle is not pre-stretched w23x. During the control experiment in Fig. 6a the receptor potentials appeared to remain superthreshold even if the peak depolarization was reduced in the high frequency range. With CB we suppose

118

M. Fischerr Brain Research 854 (2000) 106–121

an inhibitory effect on the DC component of the receptor potential that hyperpolarizes the membrane by a constant degree at each stimulation frequency. If the AC component remains constant under CB and the DC component decreases, then it is to be expected that when a high stimulation frequency leads to reduced overall activity and to a diminution of the peak depolarization of the receptor potential Žsee Fig. 6b., firing will first and foremost be affected in the high frequency range where the receptor potential first becomes subthreshold. 4.1. The site of action of CB The inhibitory effect of CB may be due to its action on various sites of the receptor. If it is the intrafusal fibers that are affected, CB might indirectly influence the ending’s membrane potential by altering the spindle’s inherent stretch of its sensory region. On the other hand, the membrane potential might be modified directly if the open probability or the conductance of channels in the sensory membrane or the encoder membrane is altered. Similar effects are to be expected if an electrogenic transport of ions is influenced or if the ionic milieu of the capsular space is altered. Finally, the axon of the afferent nerve fibers might be affected. 4.1.1. Intrafusal fibers The effects of CB on the contraction of smooth muscle cells and myocardial cells w1–3,10,24x favor the possibility that CB acts on the intrafusal muscle fibers of the receptor. However, when the isolated muscle spindle was exposed to CB, no visible length change occurred at the polar ends or in the equatorial region of these muscle fibers while the spindle was being observed under a microscope with 600fold magnification during the experiments. Thus it is very unlikely that CB itself elicited or released a contraction of the intrafusal fibers. By contrast, it is easy to recognize length changes of bag fibers induced by adding succinylcholine to the bathing fluid. However, it might be assumed that CB alters the viscoelastic properties of the intrafusal fibers. The finding that the spindle’s sensitivity to stretch was not affected by CB contradicts this assumption. If the viscoelastic properties had changed, the increase in SR with increasing amplitudes of stretch would have been expected to change. Fig. 5 shows that this was not the case. Therefore the intrafusal fibers do not appear to be the site of action of CB. 4.1.2. Spindle capsule The response properties of an isolated muscle spindle remains nearly unchanged when its capsule is removed ŽRef. w14x; own observations.. On the other hand, Fukami w5x demonstrated that a transcapsular potential of 10–20 mV exists which is negative inside the capsular space. He additionally showed that the ionic milieu of the capsular

space is different from that outside. The impulse activity of primary and secondary endings was notably reduced when the capsule was penetrated and hyaluronate removed enzymatically from the capsular space. However, the impulse activity was restored when the concentration of potassium was increased or that of calcium reduced. If CB were able to raise the calcium concentration in the intracapsular space, it would be possible to explain the decrease in impulse frequency. However, our results show that the basic discharge frequencies of a spindle during a rampand-hold stretch did not change in parallel when the calcium concentration was altered ŽFig. 8.. Similar results were obtained by Fukami w5x and also by Ottoson w18x who studied the effect of calcium on the isolated frog muscle spindle. An excess of calcium produced a gradual decrease in the spontaneous activity and diminished the sensitivity to stretch. Therefore the hypothesis that CB changes the ionic milieu of the capsular space should be rejected, because CB induced a parallel decrease in the four basic discharge frequencies without a change in the sensitivity to stretch. 4.1.3. Terminal sensory membrane and encoder membrane Our investigations do not allow us to distinguish between the effects of CB on the sensory membrane and those on the encoder membrane of muscle spindle afferents. However, the following discussion deals mainly with effects on the receptor membrane, because very little is known about the properties of the encoder membrane and its pharmacology. The sensory membrane of muscle spindle endings contains different ion channels that might be directly or indirectly influenced by CB. These channels can be divided into depolarizing and hyperpolarizing channels. A further subdivision into channels which are opened by stretch and channels which open independently of stretch might be helpful. Mechanogated ion channels that depolarize the membrane are stretch activated ŽSA. channels. Their existence in sensory endings of muscle spindles is very likely, but has not yet been directly proven. Most such channels investigated in various other tissues are selective for cations w9,12x. Our results show that CB affects the isolated muscle spindles by reducing the afferent discharge frequency without changing the spindle’s sensitivity to stretch. Therefore the ionic currents passing through SA channels appear to remain unchanged when CB is added. This view is supported by the finding that the AC component of the receptor potential obtained during sinusoidal stretches was not affected by CB ŽFig. 7.. Non-SA channels that depolarize the membrane are voltage dependent sodium or calcium channels. A decreased open probability of such channels hyperpolarizes the membrane and reduces the impulse activity. However, the contribution of voltage dependent sodium channels to the receptor potential seems to be negligible. TTX and lidocaine are well known to block these sodium channels,

M. Fischerr Brain Research 854 (2000) 106–121

but the receptor potential is little affected by these drugs Žsee Fig. 7.. When using these drugs, a slow shift of the membrane potential cannot be excluded. However, this shift appears to be very small. Otherwise the reduction of the AP amplitude should be accompanied by a reduction of the impulse frequency, and this has not been observed. Therefore a blockage of voltage dependent sodium channels can hardly explain the CB induced decline in the firing rate of muscle spindle afferents. As far as calcium channels are concerned, the blocking agents nifedipine, D600 and verapamil were tested. All these blockers induced a decrease in the spindle’s discharge frequency. However, during ramp-and-hold stretches PD did not decline by the same degree as the other basic discharge frequencies ŽFig. 9., indicating that the static properties of the spindle endings were more strongly affected than the dynamic properties. Such a difference was not observed when the muscle spindle was treated with CB. Therefore CB probably does not influence the gating of calcium channels in the muscle spindle. Very little is known about hyperpolarizing currents that contribute to the receptor potential of mammalian muscle spindles. The existence of a calcium dependent potassium channel has been shown by Kruse and Poppele w15x using the blocking agent apamin. Hunt et al. w14x sometimes observed a small depolarizing shift in the membrane potential of muscle spindle endings when they blocked potassium channels with TEA. Unfortunately it was not possible to measure slow shifts in the membrane potential in our experiments. A hyperpolarizing shift in the membrane potential might explain the inhibitory effect of CB on the discharge activity. Such a hyperpolarization could be driven by an increase in an outward potassium current or by an activation of the electrogenic NarK pump. The latter was tested by studying the influence of ouabain on the CB induced decrease in the discharge frequency. Ouabain itself raised the discharge frequency of the endings and finally caused their activity to cease. These ouabain effects appear to indicate the ongoing inhibition of the electrogenic NarK pump. When CB was added to the bathing solution while ouabain was effective, the CB induced reduction of the discharge frequency was diminished. Even though a quantitative evaluation was difficult to obtain, ouabain obviously reduced the inhibitory effect of CB ŽFig. 10.. Thus the site of action of CB might be the electrogenic NarK pump, which hyperpolarizes the membrane when its activity is increased. However, our results do not allow us to conclude that the pump is directly activated by CB because we did not measure the turn over rate of the pump. Additionally, we cannot exclude the possibility that CB affects a cellular mechanism that is closely related to the activity of the NarK pump, e.g., a secondary transport mechanism such as a sodium dependent exchange of ions across the membrane. Further experiments are needed to determine the causes of the observed effects more specifically.

119

4.1.4. Axon There are some findings that contradict the possibility that CB interacts with channels of the nodes of Ranvier: Fig. 2 indirectly shows that the propagation of action potentials along the nerve fibers is not blocked by CB. Even if the background activity of some endings ceased, firing could always be evoked by stretching the spindle. Besides, the blocking of propagation may be expected to result in a sudden stop of activity. Our investigations show a graduated decrease in the discharge frequency before firing stopped at a critical frequency ŽFig. 2a.. Additionally, the shape of the action potentials was not altered by CB. It is therefore very unlikely that CB exerts an effect on the axonal membrane of muscle spindle afferents. In summary, most probably CB reduces the afferent discharge frequency of primary and secondary muscle spindle endings by hyperpolarizing their membrane potential. The site of action of CB appears to be the NarK pump or a cellular process that is closely dependent on the activity of the pump. 4.2. Comparison between the effect of CB on primary and secondary endings In general, primary and secondary endings were affected by CB in a very similar way ŽFigs. 2–5., indicating that the mechanism of inhibition is the same in both types of endings. However, there were some slight differences between the results obtained from primary and secondary endings. Ž1. With primary endings the sensitivity of IA to CB was higher than the sensitivities of the other basic discharge frequencies ŽFig. 4a.. With secondary endings the sensitivities of the four basic discharge frequencies to CB were not significantly different ŽFig. 4b.. Ž2. The DR of primary endings rose slightly with increasing concentrations of CB where the amplitude of stretch was very low Ž2.5% of L0 ; Fig. 5a.. Where the amplitude of stretch was larger, DR was not affected by CB. The DR of secondary endings remained unchanged for each amplitude of stretch when the concentration of CB was increased ŽFig. 5b.. Both phenomena indicate that for primary endings the effect of CB is intensified in the low discharge frequency range. This can be explained by properties of the ending’s encoder. Hunt and Ottoson w13x showed that the relation between the static discharge frequency and the static receptor potential was nearly linear in primary and secondary endings. However, it is not known if this linearity holds true for very low discharge frequencies that correlate to membrane potentials just above the firing threshold. A stronger effect of CB in the low frequency range may be explained if small shifts in the receptor potential in the threshold range can be related to enhanced alterations of the firing rate. Consequently IA, which is induced by only moderate depolarization and represents firing in the low discharge frequency range, should be more influenced by

120

M. Fischerr Brain Research 854 (2000) 106–121

the depressive effect of CB than the other basic discharge frequencies which relate to stronger depolarization during a stretch. Additionally, if the amplitude of stretch is very low Ž2.5% of L0 ., then IA and FST are both in the low discharge frequency range and might both be affected by the depressive effect of CB to an increased extent. Thus under a small amplitude stretch FST decreases more than PD with increasing concentrations of CB, since PD is related to a depolarization of the membrane that is far away from its threshold. Therefore DR increases with increasing concentrations of CB. However, as long as the amplitude of stretch was large, DR was little affected by CB; this indicates that under a large amplitude both PD and FST are related to a depolarization of the membrane that is far away from its threshold, so that the relationship between the discharge frequency and the receptor potential becomes nearly linear as has been shown by Hunt and Ottoson w13x. With secondary endings DR did not rise with increasing concentrations of CB, and the sensitivity of IA to CB did not differ from the sensitivities of the other basic discharge frequencies. These results show that the effect of CB on secondary endings is not dependent on their firing rate. The difference as compared with primary endings might be due to different properties of the encoder in the low discharge frequency range. This point of view accords with the results of Poppele and Bowman w19x, who described the behavior of muscle spindle afferents quantitatively and derived different transfer functions for primary and secondary endings which reflect different encoder properties of the endings. Thus the slightly different effects of CB on primary and secondary endings are very probably not related to different inhibitory mechanisms.

Acknowledgements The author wishes to thank Prof. S.S. Schafer for ¨ reading the manuscript and for helpful suggestions. The technical assistance of B. Begemann is also gratefully acknowledged.

References w1x H. Arimura, Y. Ikemoto, T. Ito, J. Yoshitake, Lack of effects of D-tubocurarine and pancuronium on the slow action potential of the guinea pig papillary muscle, Can. Anaesth. Soc. J. 32 Ž1985. 484–490. w2x A.B. Ayers, B.N. Davies, P.G. Withrington, Responses of the isolated, perfused human spleen to sympathetic nerve stimulation, catecholamines and polypeptides, Br. J. Pharmacol. 44 Ž1972. 17–30. w3x S. Barrigon, T. Tejerina, C. Delgado, J. Tamargo, Effects of chlorbutol on 45 Ca movements and contractile responses of rat aorta and its relevance to the actions of Syntocinon, J. Pharm. Pharmacol. 36 Ž1984. 521–526.

w4x M. Fischer, S.S. Schafer, Temperature effects on the discharge ¨ frequency of primary and secondary endings of isolated cat muscle spindles recorded under a ramp-and-hold stretch, Brain Res. 840 Ž1999. 1–15. w5x Y. Fukami, Studies of capsule and capsular space of cat muscle spindles, J. Physiol. Lond. 376 Ž1986. 281–297. w6x J.E. Gregory, R.J. Harvey, U. Proske, A ‘‘late supernormal period’’ in the recovery of excitability following an action potential in muscle spindle and tendon organ receptors, J. Physiol. Lond. 271 Ž1977. 449–472. w7x J.E. Gregory, D.L. Morgan, U. Proske, Two kinds of resting discharge in cat muscle spindles, J. Neurophysiol. 66 Ž1991. 602–612. w8x Y. Habara, T. Kanno, Dual effects of chlorobutanol on secretory response and intracellular Ca2q dynamics in isolated pancreatic acini of the rat, Br. J. Pharmacol. 109 Ž1993. 685–692. w9x O.P. Hamill, D.W. McBride Jr., The pharmacology of mechanogated membrane ion channels, Pharmacol. Rev. 48 Ž1996. 231–252. w10x K. Hermsmeyer, O. Aprigliano, Effects of chlorobutanol and bradykinin on myocardial excitation, Am. J. Physiol. 230 Ž1976. 306–310. w11x S.F. Holloway, R.E. Poppele, Evidence for electrogenic pumping in cat muscle spindle, Brain. Res. 154 Ž1978. 144–147. w12x H. Hu, F. Sachs, Stretch-activated ion channels in the heart, J. Mol. Cell. Cardiol. 29 Ž1997. 1511–1523. w13x C.C. Hunt, D. Ottoson, Impulse activity and receptor potential of primary and secondary endings of isolated mammalian muscle spindles, J. Physiol. Lond. 252 Ž1975. 259–281. w14x C.C. Hunt, R.S. Wilkinson, Y. Fukami, Ionic basis of the receptor potential in primary endings of mammalian muscle spindles, J. Gen. Physiol. 71 Ž1978. 683–698. w15x M.N. Kruse, R.E. Poppele, Components of the dynamic response of mammalian muscle spindles that originate in the sensory terminals, Exp. Brain. Res. 86 Ž1991. 359–366. w16x O.C.J. Lippold, J.G. Nicholls, J.W.T. Redfearn, Electrical and mechanical factors in the adaptation of a mammalian muscle spindle, J. Physiol. 153 Ž1960. 209–217. w17x S. Nakajima, K. Takahashi, Post-tetanic hyperpolarization and electrogenic Na pump in stretch receptor neurone of crayfish, J. Physiol. Lond. 187 Ž1966. 105–127. w18x D. Ottoson, The action of calcium on the frog’s isolated muscle spindle, J. Physiol. 178 Ž1965. 68–79. w19x R.E. Poppele, R.J. Bowman, Quantitative description of linear behavior of mammalian muscle spindles, J. Neurophysiol. 33 Ž1970. 59–72. w20x R.E. Poppele, W.R. Kennedy, D.C. Quick, A determination of static mechanical properties of intrafusal muscle in isolated cat muscle spindles, Neuroscience 4 Ž1979. 401–411. w21x H. Querfurth, Action potential patterns of isolated frog muscle spindle in response to sinusoidal stimulation, J. Neurophysiol. 53 Ž1985. 76–88. w22x H. Querfurth, Processing vibratory stimuli in isolated frog muscle spindle, Exp. Brain. Res. 61 Ž1985. 11–20. w23x H. Querfurth, Receptor potentials of isolated frog muscle spindle evoked by sinusoidal stimulation, J. Neurophysiol. 53 Ž1985. 60–75. w24x O.P. Rosaeg, N.J. Cicutti, R.S. Labow, The effect of oxytocin on the contractile force of human atrial trabeculae, Anesth. Analg. 86 Ž1998. 40–44. w25x S.S. Schafer, ¨ The fast and slow components of receptor adaptation in the discharge frequency of the primary muscle spindles in the cat, Z. EEG-EMG 23 Ž1992. 12–19. w26x S.J. Sequeira, T.J. McKenna, Chlorbutol, a new inhibitor of aldosterone biosynthesis identified during examination of heparin effect on aldosterone production, J. Clin. Endocrinol. Metab. 63 Ž1986. 780–784. w27x P.G. Sokolove, I.M. Cooke, Inhibition of impulse activity in a sensory neuron by an electrogenic pump, J. Gen. Physiol. 57 Ž1971. 125–163.

M. Fischerr Brain Research 854 (2000) 106–121 w28x V.J. Wulff, J.L. Fahy, The latent period of the Limulus lateral eye receptor potential: action of chlorobutanol, Brain. Res. Bull. 2 Ž1977. 197–202. w29x V.J. Wulff, J.L. Fahy, Effect of chlorobutanol on Limulus ventral photoreceptor latency: a model of a two component latent period, Brain. Res. Bull. 4 Ž1979. 803–807.

121

w30x V.J. Wulff, J.L. Fahy, Interaction of light and chlorobutanol on Limulus ventral photoreceptor potential latency, Brain. Res. Bull. 5 Ž1980. 469–472. w31x A. Yamamoto, S. Ozono, M. Onozuka, K. Omori, Y. Tashiro, Y. Fukami, Immunocytochemical localization of Na, K-ATPase in rat muscle spindles, Cell. Struct. Funct. 19 Ž1994. 179–187.