JOURNALOF NEUROPHYSIOLOGY 1978. Printed Vol. 41, No. 6. November
in U.S.A.
Receptor Potential and Spike Initiation Varieties of Snake Muscle Spindles YASUSHI
SUMMARY
FUKAMI
AND
CONCLUSIONS
/. Receptor potentials, in response to ramp-and-hold stretch, have been recorded from two varieties of snake muscle spindles. 2. The two types of spindles have a similar sensitivity of impulse discharge to amplitude of receptor potential during the static phase of stretch. 3. Receptor potentials from short-capsule Spindles
show
a high dynamic
sensitivity
to
velocity of stretch. Amplitude of dynamic receptor potentials is well related to frequency of dynamic discharge except beyond a certain velocity of stretch where the frequency deviates progressively more than expected from linearity. 4. Receptor potentials from long-capsule spindles show a low dynamic sensitivity to velocity of stretch and amplitude of dynamic receptor potentials is well correlated with dynamic firing frequency. 5. The threshold level of receptor potential for initiating spike discharge varies with the velocity of stretch, the relation being similar for the two types of spindles. 6. It is concluded that the basis for functional differentiation of snake spindles may lie in the mechanism by which deformation of sensory endings is transformed into receptor potential. 7. Late adaptation of impulse discharge, a characteristic feature of the response of the short-capsule spindle to maintained stretch, has been related to length changes of the sensory region measured directly with Nomarski optics. The linear relation found between the slow adaptive fall of imReceived 1546
in Two
for publication
March
16, 1978.
0022-3077/78/0000-OOOO$O
pulse discharge and the simultaneous shortening of the sensory region strongly suggests a mechanical basis for the late adap. tat1on’ r NT Ro D u CT1 o N Two morphologically and functionally distinct varieties of muscle spindle occur in snakes and lizards (1, 8, 11, 18, 26, 27, 29, 3 1). Several
features
of their
responses
to ramp-and-hold stretch distinguish these two types of spindles. One type, the shortcapsule spindle, responds to both velocity and amplitude of stretch, whereas the other type, the long-capsule spindle, is mainly sensitive to the amplitude of stretch. During maintained stretch the rate of afferent discharge from the short-capsule spindle falls progressively (late adaptation). In contrast, the response of the long-capsule spindle, except during the early adaptive fall, remains at a steady firing level Sensory transduction in muscle spindles involves the following steps: transmission of mechanical stimulus (stretch) through intrafusal fibers, deformation of sensory endings followed by generation of receptor potential in the sensory terminal, and finally, impulse initiation in the sensory nerve. Using high-speed cinematography and Nomarski optics, Fukami and Hunt (12) measured 1ength changes between various structural elements in the sensory region of snake muscle spindles. They demonstrated that th e 1ength change is essentially similar in the two types of spindles. The purpose of this study is to examine the relationship between receptor potential and spike discharge in both types of snake
I .25 Copyright
0 1978 The
American
Physiological
Society
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.
SNAKE
MUSCLE
muscle spindles. There has been, to date, no systematic study of the receptor potential of reptilian muscle spindles. A preliminary report of this study has been published (IO). METHODS
Muscle spindles in segmental ventral costocutaneous muscles of the garter snake (Thtrmno/~/zi.s) were used. The technique for dissection of the muscle and the ionic composition of the bathing solution have been described elsewhere (17). After visual identification of the type of spindle under the dissecting microscope with dark-field illumination, a single spindle together with the nerve was dissected out with some extrafusal fibers that served as a support. Each end of the muscle was tied to a nylon rod which was connected to one of a pair of Brush pen motors with position and velocity feedback control. The muscle spindle was stretched at both ends and the drive to each motor was adjusted so that the innervated portion of the capsular region remained at rest during stretch. to reduce motion artifact. The position and velocity feedback signal was used to record displacement. The Brush pen motors were driven at a rate of once every 5 s by a computer (LINC) programmed to give ramp-and-hold stretches. The isolated spindle was brought near the interface between snake Ringer solution (17) and liquid paraffin and the nerve was lifted into liquid paraffin on one of a pair of glass pipettes filled with agar-Ringer solution. The other pipette was placed in the bathing solution. Each pipette was connected to a calomel half-cell. Signals from the nerve were amplified differentially by a Princeton Applied Research mode1 113 amplifier and, together with displacement signals, recorded on analog magnetic tape (HewlettPackard 3906 instrumentation recorder). Frequency of spike discharge was determined using an interspike-frequency converter or by manual counting from records reproduced from tapes by using a chart recorder. Tetrodotoxin (TTX) ( 10s7- 10MGw/v) was used to block impulse activity while recording receptor potentials. Most of the experiments were done at the muscle length at which slack was just taken up. At the conclusion of each experiment the nerve was crushed close to the capsule to check for movement artifact. In all cases examined no noticeable movement artifact was observed. For the study of length change of the intrafusal fiber during late adaptation, Nomarski optics and either a high-speed camera (Locam, Red Lake Laboratories) or a 35mm camera were used. The experimental procedure was essen-
SPINDLES
1547
tially the same as one already described (12). After the end of the dynamic stretch the sensory region was photographed at various intervals to measure length change between various structural elements visible under Nomarski optics. The afferent discharge was recorded on tape, shortly before, during, and after each photograph, to determine the average impulse frequency during this period. All experiments were performed at room temperature of about 25OC. RESULTS
The short-capsule spindle characteristically shows a high-frequency discharge during ramp stretch, followed by a much lower discharge frequency during hold stretch (Fig. 1A). The frequency of discharge is greatest early in the ramp (initial burst). When the ramp continues long enough, this is followed by a late dynamic discharge at a slightly lower frequency (Fig. lA:J. Following the dynamic response to ramp stretch, a pause in impulse activity may occur prior to the static discharge. The staticdischarge frequency varies with stretch amplitude. The receptor potential, recorded after block of impulse activity by TTX, has a configuration very similar to the changes in impulse frequency to an identical stretch. The response of the long-capsule spindle is quite different. Relatively little dynamic sensitivity is seen in either the frequency response or receptor potential. The greater static sensitivity is seen in both impulse discharge, which is more regular than that of the short-capsule spindle, and in the receptor potential. To examine the above features of receptor potentials, characteristic of each type of spindle, several components of receptor potential were related to impulse discharge obtained from the same spindle before application of TTX. Relation oj’ rcwptor potuztiul to jkqwncy oj’ impulse disc-hurgc STATIC COMPONENT. In the experiments on graded amplitude of stretch (cf. Fig. 1) the average amplitude of receptor potential during the last one-half or one-third of the hold phase of stretch was plotted against the average imPulse freauencv during the same
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
IS48
t-IG.
Y. FUKAMI
1.
amplitude of records. in the left of stretch
Kesponse of a short-capsule (A) and a long-capsule (B) spindle to ramp-and-hold stretch of graded before (left column) and after (right column) application of TTX. Stretch is identical for each pair Note in A that the stretch signals in the right column are recorded at twice amplification of those shown column. Amplitude of stretch from 1 to 3 in both A and B (in pm): 100, 500, and 700. Velocity (5 mm-s-‘) is identical for all records. Voltage calibration, 500 @. Time calibration, 0.5 s.
period. The results obtained from one longcapsule and two short-capsule spindles are shown in Fig. 2. Each group of data may be related by a straight line with roughly the same slope. This suggests that the impulse-initiating site, irrespective of the type of spindle, has a similar sensitivity to steady depolarization of the ending and that the receptor potential may account largely for the difference in impulse response to graded
amplitude of stretch between the two varieties of snake spindles. It has been shown (13) that the spike-initiation sites of the two types of snake spindles have almost identical sensitivities to externally applied current pulses: the average rate of impulse discharge increases linearly with the increase in current intensity with an almost identical slope. In Fig. 2 the intercept of linear extrapolation of a single group of data points with the x
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
SPINDLES
axis may be taken as a threshold depolarization for initiating static discharge. This threshold value differs among differe nt s pindles and may be ascribed mainly to the resting length of the spindle. The response of the short-capsule spindle is highly sensitive to stretch velocity not only in the spike frequency, but also in the amplitude of receptor potential (Fig. 3A). In contrast, the long-capsule spindle shows much less dynamic sensitivity in both impulse discharge and receptor potential (Fig. 3B). In order to examine the relation between dynamic components of the receptor potential and impulse discharge, the threshold depolarization of the receptor potential for initiating a train of impulses was determined by extrapolating the approximately linear relation between mean static frequency and receptor potential 0.5 s after the ramp phase of stretch of varying amplitude (cf. Fig. 2). The scaling of frequency on the ordinate was then adjusted so that the mean static frequency and the receptor potential 0.5 s after the ramp coincide with each other. The results thus obtained are shown in Fig. 3 as filled triangles superimposed on each receptor potential. In both types of spindle when the velocity of stretch is slow, the frequency of impulse discharge follows relatively well the dynamic component of receptor potential. As the velocity of stretch increases, the discrepancy between the two parameters increases progressively, particularly in short-capsule spindles. This progressive discrepancy between discharge frequencv and the receptor potential with the increase in velocity of stretch may also be seen in Fig. 4 where the peak frequency of dynamic discharge obtained from two short-capsule (filled circles and squares) and one long-capsule spindles (open circles) is plotted against the peak amplitude of dynamic receptor potential expressed in percent of its maximum value. The relation of late dynamic discharge to the late dynamic component of receptor potential is also shown as open squares. In addition to the higher peak frequency of discharge of the short-capsule spindle at any stretch velocities examined, the relation differs in the following way between the two types of spindles: in th e long-ca.psule spindle the DYNAMIC
COMPONENT.
1549
20
% 10
3 g -
n
100 Receptor Potential
200 ()N)
FIG. 2. Relation between average amplitude of receptor potential measured during the last one-half or one-third of the hold phase of stretch and average impulse frequency during the same period. Triangles, data from a long-capsule spindle. Filled and open circles, data obtained from two short-capsule spindles. Regression lines to fit each group of data are: y (impulses/s) = 0.077,~ - 1.2 14 (correlation coefficient, I = 0.988) for triangles, y = 0.084x - 4.950 (r = 0.949) for open circles, and ,\’ = 0.058 - 2.363 (1. = 0.977) for filled circles.
increase of the frequency may be related linearly with the increase in amplitude of the receptor potential, whereas in the shortcapsule spindle the peak frequency, beyond a certain velocity of stretch (about 10 mm/s, which corresponds roughly to 80% in Fig. 4) increases progressively more than expected from the linear relationship. Since the dynamic receptor potential of the short-capsule spindle is highly velocity sensitive, the progressive deviation from linearity may be due to the spike-generating mechanism, which is more sensitive to the rate of rise than to the amplitude of receptor potential. The possibility that TTX might affect preferentially the dynamic receptor potential seems to be less likely because neither the dynamic nor static components of the subthreshold receptor potentials were affected by TTX. Chcrngc in firing thrushold with wlocit~ of stretch In both types of spindles the threshold level of receptor potential for initiating spike discharge decreasesas the velocity of stretch is increased (arrows in Fig. 3 and upper records in Fig. 5). For comparison of this phenomenon between the two types of spindles, the th reshold le vel of re ceptor poten tial for spike initiation wa s ploitted against the la-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1550
Y. FUKAMI
tency of the first spike measured from the onset of the receptor potential. The results thus obtained from the two types of spindles are almost identical, as may be seen in Fig. 5. The change in firing threshold with velocity of stretch may be due either to accommodation, that is, the threshold change occurring in a single spike-initiation site, or to shift
FIG.
velocity records. 500 pm Triangles, impulses.
3.
of the spike-initiation site from one to another terminal branch. For the latter possibility one has to assume several spike-initiation sites with different firing threshold and accommodative properties. With a rapid rise of receptor potential, the one with lowest threshold for firing and rapid accommodation would fire first. As the rate of rise of
Response of a short-capsule (A) and a long-capsule (B) spindle to ramp-and-hold stretch of graded before (left column) and after (right column) application of TTX. Stretch is identical for each pair of Velocity of stretch from 1 to 3 in both A and B (in mms-l): 1 .O, 5.0, and 10.0. Amplitude of stretch, for A and 350 pm for B. Arrows indicate threshold level of receptor potential for the first spike. frequency of impulse discharge before application of TTX (explanation of scaling in text). Ordinate, s- l. Voltage calibration, 250 PV for A and 200 IV for B. Time calibration, 0.5 s.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
receptor potential decreased, the first site, because of the rapid accommodation, would not fire, and the second site, with higher firing threshold and slower accommodation, would fire in turn. In this way the shift would continue, with the decrease in the rate of rise of receptor potential, until the site with highest threshold and slowest accommodation would fire first by slow rate of stretch. Because of the lack of information on the spike-initiation site in snake muscle spindles, the above argument is highly speculative. As discussed later (see DISCUSSION), however, circumstantial evidence seems to support the possibility of a single accommodative mechanism for this phenomenon. Late adaptation After the ramp phase of stretch the rate of impulse discharge rapidly falls to a lower level, followed by a further slow decrease during the hold phase. These two adaptive processes (initial fast and late slow adaptation) are both approximately exponential in time course and may occur in both types of spindles.’ In long-capsule spindles, however, the late adaptation is less marked and a steady frequency level is attained usually within a few seconds. In contrast, in shortcapsule spindles the late adaptation continues until spike discharge ceases. This remarkable difference in the time course of the late adaptation has been repeatedly demonstrated in snake muscle spindles (9, 13). Because of this late adaptation, short-capsule spindles, within the range of muscle length in situ, usually lack background discharge. In this respect the sensory ending of short-capsule spindles differs from the primary ending of mammalian muscle spindles, which usually shows background discharge and lacks this degree of adaptation. Shown in the upper graph of Fig. 6 are responses to an identical stretch of a shortcapsule (open circles) and a long-capsule spindle (filled circles) in a muscle plotted on a semilog scale. The difference of time course of late adaptation between these l Houk and Henneman (16) have shown that the adaptation of the response of a tendon organ of the cat can be described by an equation containing two exponential terms, whereas the adaptation of the response to a passive stretch requires no fewer than three exponential components to explain the entire time course of adaptation.
1551
SPINDLES
impulseshec 200 r 150' t loo50-
"
40
60
80
100 %
FIG. 4. Relation between dynamic spike frequency and amplitude of dynamic receptor potential expressed in percent of the maximum value. Filled circles and squares, plot of frequency of initial burst against the amplitude of initial dynamic component of receptor potential, obtained from two short-capsule spindles. Open squares, frequency of late dynamic discharge near the end of ramp stretch plotted against the amplitude of late dynamic component of receptor potential measured just before the end of dynamic stretch. Filled circles and open squares were obtained from the same spindle and records are partly shown in Fig. 3A. Open circles, plot of peak dynamic frequency against peak amplitude of dynamic receptor potential of a long-capsule spindle. Records are partly shown in Fig. 323. In all cases the velocity of stretch was increased by 1 mm. s-l -step from 1.0 to 15.0 mms? 100% corresponds to 1,100 PV for filled circles, 2,250 PV for filled squares, 725 PV for open squares, and 180 PV for open circles. Amplitude of stretch, 350 pm for both filled circles and squares and 500 pm for open circles. Two straight lines are drawn, one through filled circles and squares below 80%, and the other through open circles.
spindles is evident. Two straight lines, one through the later part of the discharge from the short-capsule spindle and the other through triangles obtained by a plot of the difference between the straight line connecting the later part of discharge and the early part of discharge, demonstrate the two phases of adaptation menti oned above. The rate of late adaptatio n, as in frog spindles ( 31, appeared to be dependent on the stretch amplitude: the larger the stretch the slower became the rate. The lower half of Fig. 6 shows another example of the response of a short-capsule spindle with a much slower rate of adaptation. The measurements of the rate of adaptation have not been carried out on a sufficient number of snake spindles to specify the range of variation. The possible basis for the late adaptation might be either the mechanical properties of the intrafusal fiber (shortening) or the recep-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1552
Y. FUKAMI
tor mechanism (receptor adaptation) in the sensory ending. The spike-initiating mechanism (spike adaptation) may also contribute to the late adaptation. To examine the above possibilities an attempt was made to measure changes in distance between structural elements visible under Nomarski optics during maintained stretch. The sensory region was photographed at various intervals during stretch while recording afferent discharge. The result obtained from a shortcapsule spindle is shown in Fig. 7 where the average impulse frequency is plotted as a function of percent decrease of distance between two pairs of structural elements, indicated by arrow heads in the photomicro-
100
a
graph. Except for the first point lying on the ordinate, all the data points may be correlated with a straight line (see legend). This finding strongly suggeststhat the shortening of the sensory region may account largely for the late adaptation. The deviation of the first point from linearity may be taken to be due to the stretch velocity-dependent dynamic discharge, which lacks direct relationship with the overall length change of the sensory region (12). The same experiment was repeated on long-capsule spindles. As expected, during steady discharge under maintained stretch, the distance between any pairs of structural components in the sensory region remained unchanged. The
0
0
0
0
0
80
0 0
z 60 s ii a 40
Q, 0
Cc0 d 0
20
0
0
0
10
20
30 40 50 Latency (msec)
60
FIG. 5. Change in firing threshold with velocity of stretch. Threshold level of receptor potential (% of the maximum value) for initiating a first spike by stretch is plotted against the latency measured as illustrated in an upper record. Filled circles represent data from a short-capsule spindle and the maximal threshold level (100%) corresponds to 150 pV. Open circles represent data from a long-capsule spindle and the maximal threshold level (100%) corresponds to 140 pV. Inset shows sample records from the short-capsule spindle to show decrease in the threshold level (arrows) as the slope of receptor potential or the stretch velocity increases from records 1 to 3. Stretch velocity (mms-I) from records 1 to 3: 1.0, 5.0, and 15.0. Voltage calibration, 100 pV. Time calibration, 20 ms.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
corollary of the above conclusion is that the receptor potential of short-capsule spindles should decay, during a maintained stretch, in parallel with the late adaptation of spike discharge.
i mpulsedsec 60 40 20 10
DISCUSSION
The uniqueness of snake muscle spindles among those of vertebrates lies not only in their simple structure, but also in the occurrence of two varieties of spindles, which can be distinguished on the basis of their morphological features as well as their distinct response patterns to stretch. By passing electrical pulses through the sensory nerve it was shown that the relation between relative intensity of applied current and average rate of impulse discharge was almost identical for spike-initiation sites of the two types of snake spindles (13). The present results (Fig. 2) agree with the above finding in that the sensitivity of the spikeinitiation site to the static component of receptor potential is similar in the two types of spindles. Taken together, the evidence indicates that the spike-initiation site in the sensory ending of snake muscle spindles does not play an important role for determining the stretch-velocity sensitivity of the two types of spindles. The amplitude of the dynamic component of receptor potential recorded from the shortcapsule spindle has been shown in the present study to be highly sensitive to the velocity of stretch. It has been demonstrated (12) that the high stretch-velocity sensitivity of the short-capsule spindle is, like in frog spindles (24), not directly related to length changes of the overall sensory region during stretch. In view of this evidence the factor responsible for the stretch-velocity sensitivity must lie somewhere between overall length changes of the sensory region and generation of receptor potentials. The question of whether the high stretch-velocity sensitivity is due to some localized mechanical event or to the ionic mechanisms of the receptor membrane itself remains to be answered. For the progressive deviation from linearity in the relation between the peak frequency of dynamic discharge and the peak amplitude of dynamic receptor potential of the short-capsule spindle (filled symbols in Fig.
1553
SPINDLES
5
1
0
1
I 2
I 3
0
5
10
15
I 4
J
5 set
nn-
I
-
20 min
FIG. 6. Time course of adaptation of impulse discharge from a short-capsule spindle during a maintained stretch. Frequency of discharge is plotted against time after dynamic stretch on a semilog scale. Upper graph, response to an identical stretch of a short- (open circles) and a long-capsule spindle (filled circles) situated nearby in a muscle, and thus favorable for comparing responses of the two types of spindle to stretch. Note the lack of late adaptation of discharge from the longcapsule spindle. Adaptation of discharge from the shortcapsule spindle is related by two straight lines, one through later part of discharge (late slow adaptation) and the other through triangles obtained by a plot of difference between the straight line connecting later part of discharge and the early part of discharge. Lower graph, another example of adaptation of afferent discharge from a short-capsule spindle. About 2-mm stretch was applied by using a micromanipulator. A straight line is drawn through points of late slow adaptation. Note much slower rate of adaptation in this case.
4), an additional factor to the spike-generating mechanism (see RESULTS) would be the distributed membrane capacity along the sensory ending. The receptor potential generated at the receptor site, transmitted electrotoni tally , and recorded at some distance along the nerve, will be distorted in its highfrequency component leading to the observed discrepancy. In contrast to snake muscle spindles, the functional difference between the fast and the slowly adapting stretch ret eptor neurons of crayfish has been ascribed to the spikegenerating mechanism (spike adaptation)
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1554
Y. FUKAMI
imp/set
0
161
77jGO%
0 45q.G
Om
I
2
I
4
I
6 %decrease
I
s of length
0%
I
1
IO
12
FIG. 7. Relation between late slow adaptation of impulse discharge and length changes of sensory region of a short-capsule spindle. The rate of impulse discharge is plotted against decrease in distance (expressed in percent of the initial value) between two pairs of structural elements visible under Nomarski optics. Time course of adaptation of impulse discharge is shown in the lower graph of Fig. 6. Open circles, distance between structures indicated by a pair of white arrowheads in the photomicrograph. Filled circles, distance between structures indicated by black arrowheads. Initial distance is 77 (open circles) and 45.5 Frn (filled circles), respectively, which corresponds to 0% on abscissa. Regression line to fit data (except the point on y axis) is y (impulses/s) = -1.02x + 15.49. Correlation coefficient, -0.97. The black bar in the micrograph represents 50 pm.
rather than to the mechanisms producing the receptor potential (23). Intracellularly applied constant currents induce discharges in the slowly adapting neurons as long as
the depolarizing stimulus is applied, whereas in the rapidly adapting cell current pulses never evoke long-lasting trains of impulses. Furthermore, intracellularly recorded gen-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.
SNAKE
MUSCLE
erator potentials induced by various degrees of stretch are similar in both types of neurons. Resetting of activity in spike-initiation sites (or pacemakers) by antidromic invasion of impulses originated from other sites or probabilistic mixing of impulse activities in different pacemaker sites have been suggested for varieties of vertebrate tonic mechanoreceptors (3-7, 14, 15, 19, 20, 22, 30). In cat muscle spindles (4), based on findings on the effect of fusimotor stimulation, the possibility has been discussed that two spikeinitiation sites, one situated in the terminal branches of Ia axon innervating nuclear bag fibers and the other situated in those innervating nuclear chain fibers might shift from one to the other during stretch, thus transmitting high-frequency dynamic and lower frequency static discharge in the parent axon. The short-capsule spindle of snake, unlike those mechanoreceptors where multipacemakers are suggested, lacks terminal branching of the sensory axon (1, 18): the axon terminates abruptly without branching before the myelin is lost and unmyelinated bulbous terminals distribute for 100 ,um or less along the single intrafusal fiber. In the long-capsule spindle the sensory axon may bifurcate before losing the myelin. In addition, we have never observed in snake spindles the “abortive” spikes that have been observed in frog spindles. These small spikes of various amplitude and frequency are taken to be due to action potentials initiated in the first nodes of sensory nerve branches and failed to propagate into the parent axon (19-21). The above features of snake spindles, particularly of the short-capsule type, seem to point toward the possibility of a single rather than multipacemaker sites and support the assumption of accommodation as the basis for the change in firing threshold level of receptor potential with the velocity of stretch (Fig. 5). The previous finding (13) that a regular discharge elicited from the sensory ending of snake spindles by extracellular application of a constant current simply adds to the stretch-evoked response, may easily be explained in terms of a single pacemaker. The similarity between the two types of snake spindles in length changes during ramp-and-hold stretch (12) raises the question of the functional significance of the remarkably distinct morphological features
SPINDLES
1555
of the sensory region characteristic of each type. The intrafusal fiber in the polar region of the two types of snake spindles also differs’ in both the ultrastructure and histochemical profile (25, 28): the short-capsule intrafusal fiber is similar to the tonic extrafusal muscle fibers, whereas the intrafusal fiber of long-capsule spindles is similar to the twitch extrafusal fibers. The overall structural and histochemical features of the short-capsule spindle might provide a basis for the progressive slow mechanical adaptation of the intrafusal fiber. To sum up, the sensory ending of the shortcapsule spindle senses not only the magnitude of the length change, but also, through unknown mechanisms (see above), its velocity, thus producing various components of the receptor potential. Due to the mechanical adaptation of the intrafusal fiber, the short-capsule spindle usually lacks background sustained discharge and may be regarded as being specialized for registering rate of change in muscle length. In contrast, the ending of the long-capsule spindle is rather sensitive, in both receptor potential and impulse discharge, to the degree of length change, and may be regarded as being specialized to monitor muscle length itself. It is tempting to speculate on how the behavior of primary and secondary endings of mammalian muscle spindles might be derived from two types of sensory endings such as occur in snake spindles. If a single sensory axon innervated the two types of snake spindles, the response to stretch of this ending would reveal both dynamic and static sensitivities, similar to the mammalian primary ending, whereas if an axon innervated only long-capsule spindles, the ending would have major sensitivity to muscle length, similar to the secondary ending. It is interesting in this respect that shortening of the equatorial region of nuclear bag fibers has been demonstrated for cat spindles by Boyd (2) and for rat spindles by Smith (32) during a hold stretch, which may correspond to the shortening of the sensory region of the short-capsule spindle revealed in the present study. ACKNOWLEDGMENTS
The author is grateful to Drs. C. C. Hunt Rovainen for their valuable suggestions. This work was supported by Public Health Grant NS 07907.
and C. Service
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1556
Y. FUKAMI
REFERENCES 1. BARKER, D. The morphology of muscle receptors. In: Handbook of Sensory Physiology. Vol. 11 l/2. Muscle Receptors, edited by C. C. Hunt. Berlin: Springer, 1974, p. l- 190. 2. BOYD, I. A. The mammalian muscle spindle-an advanced study (film). J. Physiol. London 214: IP, 1971. G. AND WESTBURY, D. R. Adapta3. BROKENSHA, tion of the discharge of frog muscle spindles following a stretch. J. Physiol. London 242: 383403, 1974. A. AND MATTHEWS, P. B. C. The effects 4. CROWE, of stimulation of static and dynamic fusimotor fibres on the response to stretching of the primary endings of muscle spindles. J. Physiol. London 174: 109-131, 1964. 5. DUCLAUX, R. AND KENSHALO, D. R. Cutaneous receptive fields of primate cold fibres. Brain Res. 55: 437-442, 1973. 6. EMONET-DENAND, F., HULLIGER, M., MATTHEWS, P. B. C., AND PETET, J. Factors affecting modulation in post-stimulus histograms on static fusimotor stimulation. Brain Res. 134: 180- 184, 1977. 7. FLOYD, K. AND MORRISON, J. F. B. Interactions between afferent impulses within a peripheral receptive field. J. Physiol. London 238: 62-63P, 1974. 8. FUKAMI, Y. Tonic and phasic muscle spindles in snake. J. Neurophysiol. 33: 28-35, 1970. 9. FUKAMI, Y. Electrical and mechanical factors in the adaptation of reinnervated muscle spindles in the snake. In: Research in Muscle Development and the Muscle Spindle. Excerpta Med. Intern. Congr. Ser. 240, 1972, p. 379-399. 10. FUKAMI, Y. Receptor potential and impulse initiation in two varieties of reptilian muscle spindle. Nature 257: 240-241, 1975. 11. FUKAMI, Y. AND HUNT, C. C. Structure of snake muscle spindles. J. Neurophysiol. 33: 9-27, 1970. 12. FUKAMI, Y. AND HUNT, C. C. Structures in sensory region of snake spindles and their displacement during stretch. J. Neurophysiol. 40: 112 l1131, 1977. 13. FUKAMI, Y., ICHIKI, M., AND KONISHI, A. Responses of snake muscle spindles to mechanical and electrical stimulation. Brain Res. 103: 477486, 1976. 14. GOTTSCHALDT, K. M., IGGO, A., AND YOUNG, D. W. Functional characteristics of mechanoreceptors in sinus hair follicles of the cat. J. Physiol. London 235: 287-315, 1973. 15. HORCH, K. W., WHITEHORN, D., AND BURGESS, P. R. Impulse generation in type I cutaneous mechanoreceptors. J. Neurophysiol. 37: 267 -28 1, 1974. 16. HOUK, J. AND HENNEMAN, E. Response of Golgi tendon organs to active contractions of the soleus
17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
muscle of the cat. J. Neurophysiol. 30: 14661481, 1967. HUNT; C. C. AND WYLIE, R. M. Responses of snake muscle spindles to stretches and intrafusal fiber contraction. J. Neurophysiol. 33: l-8, 1970. ICHIKI, M., NAKAGAKI, I., KONISHI, A., AND FUKAMI, Y. The innervation of muscle spindles in the snake, Elaphe yuadrivirgata. J. Anat. 122: 141- 167, 1976. ITO, F. AND ITO, Y. Role of abortive spikes on encoding mechanism in frog muscle spindle. In: Understanding the stretch reflex, edited by S. Homma. Progr. Brain Res. 44: 141-154, 1976. ITO, F., KANAMORI, N., AND KURODA, H. Structural and functional asymmetries of myelinated branches in the frog muscle spindle. J. Physiol. London 241: 389-405, 1974. KATZ, B. Action potentials from a sensory nerve ending. J. Physiol. London 111: 248-260, 1950. MATTHEWS, B. Responses of intradental nerves to electrical and thermal stimulation of teeth in dogs. J. Physiol. London 264: 641-664, 1977. NAKAJIMA, S. AND ONODERA, K. Membrane properties of the stretch receptor neurones of crayfish with particular reference to mechanisms of sensory adaptation. J. Physiol. London 200: 161- 185, 1969. OTTOSON, D. AND SHEPHERD, G. M. Length changes within isolated frog muscle spindle during and after stretching. J. Physiol. London 207: 747-759, 1970. PALLOT, D. J. The structure and motor innervation of intrafusal fibers in snakes of Natrix sp. J. Anat. 118: 281-293, 1974. PALLOT, D. J. AND RIDGE, R. M. A. P. The fine structure of the long-capsule muscle spindles of the snake Natrix sp. J. Anat. 113: 61-74, 1973. PALLOT, D. J. AND RIDGE, R. M. A. P. The fine structure of the short capsule muscle spindles in snakes of Natrix sp. J. Anat. 114: 13-24, 1973. PALLOT, D. J. AND TABERNER, J. Histochemical studies on snake muscle spindles. J. Histochem. Cytochem. 22: 881-886, 1974. PROSKE, U. An electrophysiological analysis of responses from lizard muscle spindles. J. Phpsiol. London 205: 289-307, 1969. PROSKE, U. AND GREGORY, J. E. Multiple sites of impulse initiation in a tendon organ. Exptl. Neural. 50: 5 15-520, 1976. PROSKE, U. AND RIDGE, R. M. A. P. Extrafusal muscle and muscle spindles in reptiles. Progr. Neurohiol. 3: 3-29, 1974. SMITH, R. S. Properties of intrafusal muscle fibres. In: Muscle Affertwt and Motor Control, edited by R. Granit. Uppsala, Sweden: Almqvist & Wiksell, 1966, p. 69-80.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
in U.S.A.
Receptor Potential and Spike Initiation Varieties of Snake Muscle Spindles YASUSHI
SUMMARY
FUKAMI
AND
CONCLUSIONS
/. Receptor potentials, in response to ramp-and-hold stretch, have been recorded from two varieties of snake muscle spindles. 2. The two types of spindles have a similar sensitivity of impulse discharge to amplitude of receptor potential during the static phase of stretch. 3. Receptor potentials from short-capsule Spindles
show
a high dynamic
sensitivity
to
velocity of stretch. Amplitude of dynamic receptor potentials is well related to frequency of dynamic discharge except beyond a certain velocity of stretch where the frequency deviates progressively more than expected from linearity. 4. Receptor potentials from long-capsule spindles show a low dynamic sensitivity to velocity of stretch and amplitude of dynamic receptor potentials is well correlated with dynamic firing frequency. 5. The threshold level of receptor potential for initiating spike discharge varies with the velocity of stretch, the relation being similar for the two types of spindles. 6. It is concluded that the basis for functional differentiation of snake spindles may lie in the mechanism by which deformation of sensory endings is transformed into receptor potential. 7. Late adaptation of impulse discharge, a characteristic feature of the response of the short-capsule spindle to maintained stretch, has been related to length changes of the sensory region measured directly with Nomarski optics. The linear relation found between the slow adaptive fall of imReceived 1546
in Two
for publication
March
16, 1978.
0022-3077/78/0000-OOOO$O
pulse discharge and the simultaneous shortening of the sensory region strongly suggests a mechanical basis for the late adap. tat1on’ r NT Ro D u CT1 o N Two morphologically and functionally distinct varieties of muscle spindle occur in snakes and lizards (1, 8, 11, 18, 26, 27, 29, 3 1). Several
features
of their
responses
to ramp-and-hold stretch distinguish these two types of spindles. One type, the shortcapsule spindle, responds to both velocity and amplitude of stretch, whereas the other type, the long-capsule spindle, is mainly sensitive to the amplitude of stretch. During maintained stretch the rate of afferent discharge from the short-capsule spindle falls progressively (late adaptation). In contrast, the response of the long-capsule spindle, except during the early adaptive fall, remains at a steady firing level Sensory transduction in muscle spindles involves the following steps: transmission of mechanical stimulus (stretch) through intrafusal fibers, deformation of sensory endings followed by generation of receptor potential in the sensory terminal, and finally, impulse initiation in the sensory nerve. Using high-speed cinematography and Nomarski optics, Fukami and Hunt (12) measured 1ength changes between various structural elements in the sensory region of snake muscle spindles. They demonstrated that th e 1ength change is essentially similar in the two types of spindles. The purpose of this study is to examine the relationship between receptor potential and spike discharge in both types of snake
I .25 Copyright
0 1978 The
American
Physiological
Society
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.
SNAKE
MUSCLE
muscle spindles. There has been, to date, no systematic study of the receptor potential of reptilian muscle spindles. A preliminary report of this study has been published (IO). METHODS
Muscle spindles in segmental ventral costocutaneous muscles of the garter snake (Thtrmno/~/zi.s) were used. The technique for dissection of the muscle and the ionic composition of the bathing solution have been described elsewhere (17). After visual identification of the type of spindle under the dissecting microscope with dark-field illumination, a single spindle together with the nerve was dissected out with some extrafusal fibers that served as a support. Each end of the muscle was tied to a nylon rod which was connected to one of a pair of Brush pen motors with position and velocity feedback control. The muscle spindle was stretched at both ends and the drive to each motor was adjusted so that the innervated portion of the capsular region remained at rest during stretch. to reduce motion artifact. The position and velocity feedback signal was used to record displacement. The Brush pen motors were driven at a rate of once every 5 s by a computer (LINC) programmed to give ramp-and-hold stretches. The isolated spindle was brought near the interface between snake Ringer solution (17) and liquid paraffin and the nerve was lifted into liquid paraffin on one of a pair of glass pipettes filled with agar-Ringer solution. The other pipette was placed in the bathing solution. Each pipette was connected to a calomel half-cell. Signals from the nerve were amplified differentially by a Princeton Applied Research mode1 113 amplifier and, together with displacement signals, recorded on analog magnetic tape (HewlettPackard 3906 instrumentation recorder). Frequency of spike discharge was determined using an interspike-frequency converter or by manual counting from records reproduced from tapes by using a chart recorder. Tetrodotoxin (TTX) ( 10s7- 10MGw/v) was used to block impulse activity while recording receptor potentials. Most of the experiments were done at the muscle length at which slack was just taken up. At the conclusion of each experiment the nerve was crushed close to the capsule to check for movement artifact. In all cases examined no noticeable movement artifact was observed. For the study of length change of the intrafusal fiber during late adaptation, Nomarski optics and either a high-speed camera (Locam, Red Lake Laboratories) or a 35mm camera were used. The experimental procedure was essen-
SPINDLES
1547
tially the same as one already described (12). After the end of the dynamic stretch the sensory region was photographed at various intervals to measure length change between various structural elements visible under Nomarski optics. The afferent discharge was recorded on tape, shortly before, during, and after each photograph, to determine the average impulse frequency during this period. All experiments were performed at room temperature of about 25OC. RESULTS
The short-capsule spindle characteristically shows a high-frequency discharge during ramp stretch, followed by a much lower discharge frequency during hold stretch (Fig. 1A). The frequency of discharge is greatest early in the ramp (initial burst). When the ramp continues long enough, this is followed by a late dynamic discharge at a slightly lower frequency (Fig. lA:J. Following the dynamic response to ramp stretch, a pause in impulse activity may occur prior to the static discharge. The staticdischarge frequency varies with stretch amplitude. The receptor potential, recorded after block of impulse activity by TTX, has a configuration very similar to the changes in impulse frequency to an identical stretch. The response of the long-capsule spindle is quite different. Relatively little dynamic sensitivity is seen in either the frequency response or receptor potential. The greater static sensitivity is seen in both impulse discharge, which is more regular than that of the short-capsule spindle, and in the receptor potential. To examine the above features of receptor potentials, characteristic of each type of spindle, several components of receptor potential were related to impulse discharge obtained from the same spindle before application of TTX. Relation oj’ rcwptor potuztiul to jkqwncy oj’ impulse disc-hurgc STATIC COMPONENT. In the experiments on graded amplitude of stretch (cf. Fig. 1) the average amplitude of receptor potential during the last one-half or one-third of the hold phase of stretch was plotted against the average imPulse freauencv during the same
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
IS48
t-IG.
Y. FUKAMI
1.
amplitude of records. in the left of stretch
Kesponse of a short-capsule (A) and a long-capsule (B) spindle to ramp-and-hold stretch of graded before (left column) and after (right column) application of TTX. Stretch is identical for each pair Note in A that the stretch signals in the right column are recorded at twice amplification of those shown column. Amplitude of stretch from 1 to 3 in both A and B (in pm): 100, 500, and 700. Velocity (5 mm-s-‘) is identical for all records. Voltage calibration, 500 @. Time calibration, 0.5 s.
period. The results obtained from one longcapsule and two short-capsule spindles are shown in Fig. 2. Each group of data may be related by a straight line with roughly the same slope. This suggests that the impulse-initiating site, irrespective of the type of spindle, has a similar sensitivity to steady depolarization of the ending and that the receptor potential may account largely for the difference in impulse response to graded
amplitude of stretch between the two varieties of snake spindles. It has been shown (13) that the spike-initiation sites of the two types of snake spindles have almost identical sensitivities to externally applied current pulses: the average rate of impulse discharge increases linearly with the increase in current intensity with an almost identical slope. In Fig. 2 the intercept of linear extrapolation of a single group of data points with the x
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
SPINDLES
axis may be taken as a threshold depolarization for initiating static discharge. This threshold value differs among differe nt s pindles and may be ascribed mainly to the resting length of the spindle. The response of the short-capsule spindle is highly sensitive to stretch velocity not only in the spike frequency, but also in the amplitude of receptor potential (Fig. 3A). In contrast, the long-capsule spindle shows much less dynamic sensitivity in both impulse discharge and receptor potential (Fig. 3B). In order to examine the relation between dynamic components of the receptor potential and impulse discharge, the threshold depolarization of the receptor potential for initiating a train of impulses was determined by extrapolating the approximately linear relation between mean static frequency and receptor potential 0.5 s after the ramp phase of stretch of varying amplitude (cf. Fig. 2). The scaling of frequency on the ordinate was then adjusted so that the mean static frequency and the receptor potential 0.5 s after the ramp coincide with each other. The results thus obtained are shown in Fig. 3 as filled triangles superimposed on each receptor potential. In both types of spindle when the velocity of stretch is slow, the frequency of impulse discharge follows relatively well the dynamic component of receptor potential. As the velocity of stretch increases, the discrepancy between the two parameters increases progressively, particularly in short-capsule spindles. This progressive discrepancy between discharge frequencv and the receptor potential with the increase in velocity of stretch may also be seen in Fig. 4 where the peak frequency of dynamic discharge obtained from two short-capsule (filled circles and squares) and one long-capsule spindles (open circles) is plotted against the peak amplitude of dynamic receptor potential expressed in percent of its maximum value. The relation of late dynamic discharge to the late dynamic component of receptor potential is also shown as open squares. In addition to the higher peak frequency of discharge of the short-capsule spindle at any stretch velocities examined, the relation differs in the following way between the two types of spindles: in th e long-ca.psule spindle the DYNAMIC
COMPONENT.
1549
20
% 10
3 g -
n
100 Receptor Potential
200 ()N)
FIG. 2. Relation between average amplitude of receptor potential measured during the last one-half or one-third of the hold phase of stretch and average impulse frequency during the same period. Triangles, data from a long-capsule spindle. Filled and open circles, data obtained from two short-capsule spindles. Regression lines to fit each group of data are: y (impulses/s) = 0.077,~ - 1.2 14 (correlation coefficient, I = 0.988) for triangles, y = 0.084x - 4.950 (r = 0.949) for open circles, and ,\’ = 0.058 - 2.363 (1. = 0.977) for filled circles.
increase of the frequency may be related linearly with the increase in amplitude of the receptor potential, whereas in the shortcapsule spindle the peak frequency, beyond a certain velocity of stretch (about 10 mm/s, which corresponds roughly to 80% in Fig. 4) increases progressively more than expected from the linear relationship. Since the dynamic receptor potential of the short-capsule spindle is highly velocity sensitive, the progressive deviation from linearity may be due to the spike-generating mechanism, which is more sensitive to the rate of rise than to the amplitude of receptor potential. The possibility that TTX might affect preferentially the dynamic receptor potential seems to be less likely because neither the dynamic nor static components of the subthreshold receptor potentials were affected by TTX. Chcrngc in firing thrushold with wlocit~ of stretch In both types of spindles the threshold level of receptor potential for initiating spike discharge decreasesas the velocity of stretch is increased (arrows in Fig. 3 and upper records in Fig. 5). For comparison of this phenomenon between the two types of spindles, the th reshold le vel of re ceptor poten tial for spike initiation wa s ploitted against the la-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1550
Y. FUKAMI
tency of the first spike measured from the onset of the receptor potential. The results thus obtained from the two types of spindles are almost identical, as may be seen in Fig. 5. The change in firing threshold with velocity of stretch may be due either to accommodation, that is, the threshold change occurring in a single spike-initiation site, or to shift
FIG.
velocity records. 500 pm Triangles, impulses.
3.
of the spike-initiation site from one to another terminal branch. For the latter possibility one has to assume several spike-initiation sites with different firing threshold and accommodative properties. With a rapid rise of receptor potential, the one with lowest threshold for firing and rapid accommodation would fire first. As the rate of rise of
Response of a short-capsule (A) and a long-capsule (B) spindle to ramp-and-hold stretch of graded before (left column) and after (right column) application of TTX. Stretch is identical for each pair of Velocity of stretch from 1 to 3 in both A and B (in mms-l): 1 .O, 5.0, and 10.0. Amplitude of stretch, for A and 350 pm for B. Arrows indicate threshold level of receptor potential for the first spike. frequency of impulse discharge before application of TTX (explanation of scaling in text). Ordinate, s- l. Voltage calibration, 250 PV for A and 200 IV for B. Time calibration, 0.5 s.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
receptor potential decreased, the first site, because of the rapid accommodation, would not fire, and the second site, with higher firing threshold and slower accommodation, would fire in turn. In this way the shift would continue, with the decrease in the rate of rise of receptor potential, until the site with highest threshold and slowest accommodation would fire first by slow rate of stretch. Because of the lack of information on the spike-initiation site in snake muscle spindles, the above argument is highly speculative. As discussed later (see DISCUSSION), however, circumstantial evidence seems to support the possibility of a single accommodative mechanism for this phenomenon. Late adaptation After the ramp phase of stretch the rate of impulse discharge rapidly falls to a lower level, followed by a further slow decrease during the hold phase. These two adaptive processes (initial fast and late slow adaptation) are both approximately exponential in time course and may occur in both types of spindles.’ In long-capsule spindles, however, the late adaptation is less marked and a steady frequency level is attained usually within a few seconds. In contrast, in shortcapsule spindles the late adaptation continues until spike discharge ceases. This remarkable difference in the time course of the late adaptation has been repeatedly demonstrated in snake muscle spindles (9, 13). Because of this late adaptation, short-capsule spindles, within the range of muscle length in situ, usually lack background discharge. In this respect the sensory ending of short-capsule spindles differs from the primary ending of mammalian muscle spindles, which usually shows background discharge and lacks this degree of adaptation. Shown in the upper graph of Fig. 6 are responses to an identical stretch of a shortcapsule (open circles) and a long-capsule spindle (filled circles) in a muscle plotted on a semilog scale. The difference of time course of late adaptation between these l Houk and Henneman (16) have shown that the adaptation of the response of a tendon organ of the cat can be described by an equation containing two exponential terms, whereas the adaptation of the response to a passive stretch requires no fewer than three exponential components to explain the entire time course of adaptation.
1551
SPINDLES
impulseshec 200 r 150' t loo50-
"
40
60
80
100 %
FIG. 4. Relation between dynamic spike frequency and amplitude of dynamic receptor potential expressed in percent of the maximum value. Filled circles and squares, plot of frequency of initial burst against the amplitude of initial dynamic component of receptor potential, obtained from two short-capsule spindles. Open squares, frequency of late dynamic discharge near the end of ramp stretch plotted against the amplitude of late dynamic component of receptor potential measured just before the end of dynamic stretch. Filled circles and open squares were obtained from the same spindle and records are partly shown in Fig. 3A. Open circles, plot of peak dynamic frequency against peak amplitude of dynamic receptor potential of a long-capsule spindle. Records are partly shown in Fig. 323. In all cases the velocity of stretch was increased by 1 mm. s-l -step from 1.0 to 15.0 mms? 100% corresponds to 1,100 PV for filled circles, 2,250 PV for filled squares, 725 PV for open squares, and 180 PV for open circles. Amplitude of stretch, 350 pm for both filled circles and squares and 500 pm for open circles. Two straight lines are drawn, one through filled circles and squares below 80%, and the other through open circles.
spindles is evident. Two straight lines, one through the later part of the discharge from the short-capsule spindle and the other through triangles obtained by a plot of the difference between the straight line connecting the later part of discharge and the early part of discharge, demonstrate the two phases of adaptation menti oned above. The rate of late adaptatio n, as in frog spindles ( 31, appeared to be dependent on the stretch amplitude: the larger the stretch the slower became the rate. The lower half of Fig. 6 shows another example of the response of a short-capsule spindle with a much slower rate of adaptation. The measurements of the rate of adaptation have not been carried out on a sufficient number of snake spindles to specify the range of variation. The possible basis for the late adaptation might be either the mechanical properties of the intrafusal fiber (shortening) or the recep-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1552
Y. FUKAMI
tor mechanism (receptor adaptation) in the sensory ending. The spike-initiating mechanism (spike adaptation) may also contribute to the late adaptation. To examine the above possibilities an attempt was made to measure changes in distance between structural elements visible under Nomarski optics during maintained stretch. The sensory region was photographed at various intervals during stretch while recording afferent discharge. The result obtained from a shortcapsule spindle is shown in Fig. 7 where the average impulse frequency is plotted as a function of percent decrease of distance between two pairs of structural elements, indicated by arrow heads in the photomicro-
100
a
graph. Except for the first point lying on the ordinate, all the data points may be correlated with a straight line (see legend). This finding strongly suggeststhat the shortening of the sensory region may account largely for the late adaptation. The deviation of the first point from linearity may be taken to be due to the stretch velocity-dependent dynamic discharge, which lacks direct relationship with the overall length change of the sensory region (12). The same experiment was repeated on long-capsule spindles. As expected, during steady discharge under maintained stretch, the distance between any pairs of structural components in the sensory region remained unchanged. The
0
0
0
0
0
80
0 0
z 60 s ii a 40
Q, 0
Cc0 d 0
20
0
0
0
10
20
30 40 50 Latency (msec)
60
FIG. 5. Change in firing threshold with velocity of stretch. Threshold level of receptor potential (% of the maximum value) for initiating a first spike by stretch is plotted against the latency measured as illustrated in an upper record. Filled circles represent data from a short-capsule spindle and the maximal threshold level (100%) corresponds to 150 pV. Open circles represent data from a long-capsule spindle and the maximal threshold level (100%) corresponds to 140 pV. Inset shows sample records from the short-capsule spindle to show decrease in the threshold level (arrows) as the slope of receptor potential or the stretch velocity increases from records 1 to 3. Stretch velocity (mms-I) from records 1 to 3: 1.0, 5.0, and 15.0. Voltage calibration, 100 pV. Time calibration, 20 ms.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
SNAKE
MUSCLE
corollary of the above conclusion is that the receptor potential of short-capsule spindles should decay, during a maintained stretch, in parallel with the late adaptation of spike discharge.
i mpulsedsec 60 40 20 10
DISCUSSION
The uniqueness of snake muscle spindles among those of vertebrates lies not only in their simple structure, but also in the occurrence of two varieties of spindles, which can be distinguished on the basis of their morphological features as well as their distinct response patterns to stretch. By passing electrical pulses through the sensory nerve it was shown that the relation between relative intensity of applied current and average rate of impulse discharge was almost identical for spike-initiation sites of the two types of snake spindles (13). The present results (Fig. 2) agree with the above finding in that the sensitivity of the spikeinitiation site to the static component of receptor potential is similar in the two types of spindles. Taken together, the evidence indicates that the spike-initiation site in the sensory ending of snake muscle spindles does not play an important role for determining the stretch-velocity sensitivity of the two types of spindles. The amplitude of the dynamic component of receptor potential recorded from the shortcapsule spindle has been shown in the present study to be highly sensitive to the velocity of stretch. It has been demonstrated (12) that the high stretch-velocity sensitivity of the short-capsule spindle is, like in frog spindles (24), not directly related to length changes of the overall sensory region during stretch. In view of this evidence the factor responsible for the stretch-velocity sensitivity must lie somewhere between overall length changes of the sensory region and generation of receptor potentials. The question of whether the high stretch-velocity sensitivity is due to some localized mechanical event or to the ionic mechanisms of the receptor membrane itself remains to be answered. For the progressive deviation from linearity in the relation between the peak frequency of dynamic discharge and the peak amplitude of dynamic receptor potential of the short-capsule spindle (filled symbols in Fig.
1553
SPINDLES
5
1
0
1
I 2
I 3
0
5
10
15
I 4
J
5 set
nn-
I
-
20 min
FIG. 6. Time course of adaptation of impulse discharge from a short-capsule spindle during a maintained stretch. Frequency of discharge is plotted against time after dynamic stretch on a semilog scale. Upper graph, response to an identical stretch of a short- (open circles) and a long-capsule spindle (filled circles) situated nearby in a muscle, and thus favorable for comparing responses of the two types of spindle to stretch. Note the lack of late adaptation of discharge from the longcapsule spindle. Adaptation of discharge from the shortcapsule spindle is related by two straight lines, one through later part of discharge (late slow adaptation) and the other through triangles obtained by a plot of difference between the straight line connecting later part of discharge and the early part of discharge. Lower graph, another example of adaptation of afferent discharge from a short-capsule spindle. About 2-mm stretch was applied by using a micromanipulator. A straight line is drawn through points of late slow adaptation. Note much slower rate of adaptation in this case.
4), an additional factor to the spike-generating mechanism (see RESULTS) would be the distributed membrane capacity along the sensory ending. The receptor potential generated at the receptor site, transmitted electrotoni tally , and recorded at some distance along the nerve, will be distorted in its highfrequency component leading to the observed discrepancy. In contrast to snake muscle spindles, the functional difference between the fast and the slowly adapting stretch ret eptor neurons of crayfish has been ascribed to the spikegenerating mechanism (spike adaptation)
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1554
Y. FUKAMI
imp/set
0
161
77jGO%
0 45q.G
Om
I
2
I
4
I
6 %decrease
I
s of length
0%
I
1
IO
12
FIG. 7. Relation between late slow adaptation of impulse discharge and length changes of sensory region of a short-capsule spindle. The rate of impulse discharge is plotted against decrease in distance (expressed in percent of the initial value) between two pairs of structural elements visible under Nomarski optics. Time course of adaptation of impulse discharge is shown in the lower graph of Fig. 6. Open circles, distance between structures indicated by a pair of white arrowheads in the photomicrograph. Filled circles, distance between structures indicated by black arrowheads. Initial distance is 77 (open circles) and 45.5 Frn (filled circles), respectively, which corresponds to 0% on abscissa. Regression line to fit data (except the point on y axis) is y (impulses/s) = -1.02x + 15.49. Correlation coefficient, -0.97. The black bar in the micrograph represents 50 pm.
rather than to the mechanisms producing the receptor potential (23). Intracellularly applied constant currents induce discharges in the slowly adapting neurons as long as
the depolarizing stimulus is applied, whereas in the rapidly adapting cell current pulses never evoke long-lasting trains of impulses. Furthermore, intracellularly recorded gen-
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.
SNAKE
MUSCLE
erator potentials induced by various degrees of stretch are similar in both types of neurons. Resetting of activity in spike-initiation sites (or pacemakers) by antidromic invasion of impulses originated from other sites or probabilistic mixing of impulse activities in different pacemaker sites have been suggested for varieties of vertebrate tonic mechanoreceptors (3-7, 14, 15, 19, 20, 22, 30). In cat muscle spindles (4), based on findings on the effect of fusimotor stimulation, the possibility has been discussed that two spikeinitiation sites, one situated in the terminal branches of Ia axon innervating nuclear bag fibers and the other situated in those innervating nuclear chain fibers might shift from one to the other during stretch, thus transmitting high-frequency dynamic and lower frequency static discharge in the parent axon. The short-capsule spindle of snake, unlike those mechanoreceptors where multipacemakers are suggested, lacks terminal branching of the sensory axon (1, 18): the axon terminates abruptly without branching before the myelin is lost and unmyelinated bulbous terminals distribute for 100 ,um or less along the single intrafusal fiber. In the long-capsule spindle the sensory axon may bifurcate before losing the myelin. In addition, we have never observed in snake spindles the “abortive” spikes that have been observed in frog spindles. These small spikes of various amplitude and frequency are taken to be due to action potentials initiated in the first nodes of sensory nerve branches and failed to propagate into the parent axon (19-21). The above features of snake spindles, particularly of the short-capsule type, seem to point toward the possibility of a single rather than multipacemaker sites and support the assumption of accommodation as the basis for the change in firing threshold level of receptor potential with the velocity of stretch (Fig. 5). The previous finding (13) that a regular discharge elicited from the sensory ending of snake spindles by extracellular application of a constant current simply adds to the stretch-evoked response, may easily be explained in terms of a single pacemaker. The similarity between the two types of snake spindles in length changes during ramp-and-hold stretch (12) raises the question of the functional significance of the remarkably distinct morphological features
SPINDLES
1555
of the sensory region characteristic of each type. The intrafusal fiber in the polar region of the two types of snake spindles also differs’ in both the ultrastructure and histochemical profile (25, 28): the short-capsule intrafusal fiber is similar to the tonic extrafusal muscle fibers, whereas the intrafusal fiber of long-capsule spindles is similar to the twitch extrafusal fibers. The overall structural and histochemical features of the short-capsule spindle might provide a basis for the progressive slow mechanical adaptation of the intrafusal fiber. To sum up, the sensory ending of the shortcapsule spindle senses not only the magnitude of the length change, but also, through unknown mechanisms (see above), its velocity, thus producing various components of the receptor potential. Due to the mechanical adaptation of the intrafusal fiber, the short-capsule spindle usually lacks background sustained discharge and may be regarded as being specialized for registering rate of change in muscle length. In contrast, the ending of the long-capsule spindle is rather sensitive, in both receptor potential and impulse discharge, to the degree of length change, and may be regarded as being specialized to monitor muscle length itself. It is tempting to speculate on how the behavior of primary and secondary endings of mammalian muscle spindles might be derived from two types of sensory endings such as occur in snake spindles. If a single sensory axon innervated the two types of snake spindles, the response to stretch of this ending would reveal both dynamic and static sensitivities, similar to the mammalian primary ending, whereas if an axon innervated only long-capsule spindles, the ending would have major sensitivity to muscle length, similar to the secondary ending. It is interesting in this respect that shortening of the equatorial region of nuclear bag fibers has been demonstrated for cat spindles by Boyd (2) and for rat spindles by Smith (32) during a hold stretch, which may correspond to the shortening of the sensory region of the short-capsule spindle revealed in the present study. ACKNOWLEDGMENTS
The author is grateful to Drs. C. C. Hunt Rovainen for their valuable suggestions. This work was supported by Public Health Grant NS 07907.
and C. Service
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
1556
Y. FUKAMI
REFERENCES 1. BARKER, D. The morphology of muscle receptors. In: Handbook of Sensory Physiology. Vol. 11 l/2. Muscle Receptors, edited by C. C. Hunt. Berlin: Springer, 1974, p. l- 190. 2. BOYD, I. A. The mammalian muscle spindle-an advanced study (film). J. Physiol. London 214: IP, 1971. G. AND WESTBURY, D. R. Adapta3. BROKENSHA, tion of the discharge of frog muscle spindles following a stretch. J. Physiol. London 242: 383403, 1974. A. AND MATTHEWS, P. B. C. The effects 4. CROWE, of stimulation of static and dynamic fusimotor fibres on the response to stretching of the primary endings of muscle spindles. J. Physiol. London 174: 109-131, 1964. 5. DUCLAUX, R. AND KENSHALO, D. R. Cutaneous receptive fields of primate cold fibres. Brain Res. 55: 437-442, 1973. 6. EMONET-DENAND, F., HULLIGER, M., MATTHEWS, P. B. C., AND PETET, J. Factors affecting modulation in post-stimulus histograms on static fusimotor stimulation. Brain Res. 134: 180- 184, 1977. 7. FLOYD, K. AND MORRISON, J. F. B. Interactions between afferent impulses within a peripheral receptive field. J. Physiol. London 238: 62-63P, 1974. 8. FUKAMI, Y. Tonic and phasic muscle spindles in snake. J. Neurophysiol. 33: 28-35, 1970. 9. FUKAMI, Y. Electrical and mechanical factors in the adaptation of reinnervated muscle spindles in the snake. In: Research in Muscle Development and the Muscle Spindle. Excerpta Med. Intern. Congr. Ser. 240, 1972, p. 379-399. 10. FUKAMI, Y. Receptor potential and impulse initiation in two varieties of reptilian muscle spindle. Nature 257: 240-241, 1975. 11. FUKAMI, Y. AND HUNT, C. C. Structure of snake muscle spindles. J. Neurophysiol. 33: 9-27, 1970. 12. FUKAMI, Y. AND HUNT, C. C. Structures in sensory region of snake spindles and their displacement during stretch. J. Neurophysiol. 40: 112 l1131, 1977. 13. FUKAMI, Y., ICHIKI, M., AND KONISHI, A. Responses of snake muscle spindles to mechanical and electrical stimulation. Brain Res. 103: 477486, 1976. 14. GOTTSCHALDT, K. M., IGGO, A., AND YOUNG, D. W. Functional characteristics of mechanoreceptors in sinus hair follicles of the cat. J. Physiol. London 235: 287-315, 1973. 15. HORCH, K. W., WHITEHORN, D., AND BURGESS, P. R. Impulse generation in type I cutaneous mechanoreceptors. J. Neurophysiol. 37: 267 -28 1, 1974. 16. HOUK, J. AND HENNEMAN, E. Response of Golgi tendon organs to active contractions of the soleus
17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
muscle of the cat. J. Neurophysiol. 30: 14661481, 1967. HUNT; C. C. AND WYLIE, R. M. Responses of snake muscle spindles to stretches and intrafusal fiber contraction. J. Neurophysiol. 33: l-8, 1970. ICHIKI, M., NAKAGAKI, I., KONISHI, A., AND FUKAMI, Y. The innervation of muscle spindles in the snake, Elaphe yuadrivirgata. J. Anat. 122: 141- 167, 1976. ITO, F. AND ITO, Y. Role of abortive spikes on encoding mechanism in frog muscle spindle. In: Understanding the stretch reflex, edited by S. Homma. Progr. Brain Res. 44: 141-154, 1976. ITO, F., KANAMORI, N., AND KURODA, H. Structural and functional asymmetries of myelinated branches in the frog muscle spindle. J. Physiol. London 241: 389-405, 1974. KATZ, B. Action potentials from a sensory nerve ending. J. Physiol. London 111: 248-260, 1950. MATTHEWS, B. Responses of intradental nerves to electrical and thermal stimulation of teeth in dogs. J. Physiol. London 264: 641-664, 1977. NAKAJIMA, S. AND ONODERA, K. Membrane properties of the stretch receptor neurones of crayfish with particular reference to mechanisms of sensory adaptation. J. Physiol. London 200: 161- 185, 1969. OTTOSON, D. AND SHEPHERD, G. M. Length changes within isolated frog muscle spindle during and after stretching. J. Physiol. London 207: 747-759, 1970. PALLOT, D. J. The structure and motor innervation of intrafusal fibers in snakes of Natrix sp. J. Anat. 118: 281-293, 1974. PALLOT, D. J. AND RIDGE, R. M. A. P. The fine structure of the long-capsule muscle spindles of the snake Natrix sp. J. Anat. 113: 61-74, 1973. PALLOT, D. J. AND RIDGE, R. M. A. P. The fine structure of the short capsule muscle spindles in snakes of Natrix sp. J. Anat. 114: 13-24, 1973. PALLOT, D. J. AND TABERNER, J. Histochemical studies on snake muscle spindles. J. Histochem. Cytochem. 22: 881-886, 1974. PROSKE, U. An electrophysiological analysis of responses from lizard muscle spindles. J. Phpsiol. London 205: 289-307, 1969. PROSKE, U. AND GREGORY, J. E. Multiple sites of impulse initiation in a tendon organ. Exptl. Neural. 50: 5 15-520, 1976. PROSKE, U. AND RIDGE, R. M. A. P. Extrafusal muscle and muscle spindles in reptiles. Progr. Neurohiol. 3: 3-29, 1974. SMITH, R. S. Properties of intrafusal muscle fibres. In: Muscle Affertwt and Motor Control, edited by R. Granit. Uppsala, Sweden: Almqvist & Wiksell, 1966, p. 69-80.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019
