Length changes of the cat soleus muscle under frequency-modulated distributed stimulation of efferents in isotony

The length changes of the cat soleus muscle have been examined in isotony within closed cycles of the stimulation rate change. Successive stimuli were applied in a cycle to five filaments of the preliminary dissected L7-S1 ventral roots (method of distributed stimulation), maximal rate did not exceed 80-90/s (16-18/s per single filament). At the beginning of a slow linear increase in the rate a muscle began shortening rather quickly, the rate changes without muscle reaction consisted of 5.91 +/- 0.28/s (mean +/- S.E.M.). A substantially linear movement was observed during increase of the input rate up to 40-50/s (i.e. 8-10/s per filament), a further rate increment could evoke a somewhat slowing of the shortening velocity with a distinctive infection of the rate-length curve. During a significant part of the rate decrease phase no movement was seen, the rate range with the absence of muscle lengthening was 29.93 +/- 1.57/s. After such a pronounced period of length fixation, a muscle began to elongate, the steady-state velocity at this part of the movement trajectory was invariably higher as compared with shortening velocity at the leading edge of the cycle. The striking feature of a powerful length clamping seen in active muscle at the phase of rate decrease preceded by previous rate increment allows us to suppose that this strongly non-linear behaviour of muscle might be a main reason for the existence of powerful dynamic components in efferent activity every time when muscle should shorten against external load in a ramp-and-hold fashion. We used the following experimental paradigm to check the assumption. Stimulation began with regular rate of 15/s; then, after transition from isometry to isotony and cessation of movement transients, the rate was raised linearly; after reaching a peak value of 50-90/s, it decreased linearly or exponentially, being afterwards fixed at several different levels between the maximal and initial values of the rate. It was demonstrated that such pattern of stimulation could be effective for a linear transition between two equilibrium lengths, provided that corresponding parameters in modulation signal were chosen in an optimal way. Duration and amplitude of the leading edge of the dynamic component seemed to completely define amplitude and velocity of the ramp phase of movement, parameters of decay to the final steady rate were, on the other hand, important for efficacy of the length clamping at hold phase. It was concluded that hysteresis effects of muscle contraction seemed to be an extremely important nonlinear property of muscle dynamics in producing any transition movements between two steady-states. Thus, oversimplified muscle models, like the spring model based on the isometric length-tension dependencies, seem to be incorrect. Possible mechanisms for the muscle hysteresis and its role in motor control are discussed.