Joint angle signaling by muscle spindle receptors

Abstract

Nerve impulses were recorded from sensory fibers supplying the tibialis anterior and soleus muscles of anesthetized cats as the ankle joint was moved from one end of the flexion-extension axis to the other and back again in steps of 6–7 °. The rate of movement from one position to the next was 40 deg/s and each position was held for 16–18 s. Plots were made of receptor discharge frequency as a function of ankle joint angle during joint movement (dynamic input-output (I-O) functions) as well as 2 and 15 s after movement terminated (2 and 15 s static I-O functions). Only receptors with a sustained (5 s) static response within the physiological range were studied. A total of 229 tibialis anterior receptors met this criterion, of which 11 were identified as tendon organs. One hundred and five soleus receptors were studied, of which 6 were tendon organs. Thus tendon organ activity accounted for only a small part of the muscle afferent signal under passive conditions. The spindle receptors in soleus and tibialis anterior divided the ankle flexion-extension range about equally between them, those in soleus signaling over the flexion half of the range and those in tibialis anterior over the extension half. At angles where the receptors in a particular muscle did not signal joint angle, the tendon of the muscle was observed to be slack. Thus the total muscle afferent discharged in a relaxed animal is high at one end of the range, declines progressively as the ankle is displaced to an intermediate position, and then increases again as the joint moves toward the opposite end of the range. The spindle receptors within an individual muscle were recruited rather early as the muscle came under tension so that over most of a muscle's signaling range joint angle could have been coded by changes in receptor discharge frequency but not by which spindle receptors were active. To evaluate the information signaled by individual muscle spindle receptors, the following measurements were made from plots of impulse frequency vs joint angle: (1) dynamic response, defined as the frequency difference between the dynamic and 2 s static I-O functions during muscle lengthening; (2) adaptation, defined as the frequency difference between the 2 and 15 s static I-O functions during muscle lengthening; (3) linear directionality of the 2 s static response, defined as the angular distance between the 2 s static I-O functions for muscle lengthening and muscle shortening where they crossed a common frequency line about half way through the coded region; (4) resting discharge; and (5) the overall frequency increment, (6) slope and (7) threshold of the 2 s static I-O function. Dynamic response, adaptation and linear directionality displayed a strong positive correlation in both tibialis anterior and soleus and the receptors were grouped into several subpopulations by a computer programmed to cluster the receptors that were closest in a 3 dimensional space specified by these parameters. The subpopulations at the phasic end of the range had large dynamic responses, much linear directionality, and adapted at a relatively rapid rate. These features decreased markedly as the subpopulations became more tonic. The 2 s static I-O functions of the subpopulations did not differ systematically in overall frequency increment or slope but by 15 s both tended to be greater for the more tonic receptors. Also, the more tonic neurons tended to have higher thresholds and lower conduction velocities. Since the discharge frequency of spindle receptors is likely to be important in specifying joint angle, an effort was made to evaluate the signaling ‘errors’ caused by rate sensitivity, adaptation and linear directionality. Linear directionality gives a direct error measurement and was determined not only for the 2 s static response but also during movement. The error due to rate sensitivity was considered to be the angular distance between the dynamic and 2 s static I-O functions where they crossed a common frequency line about half way through the coded region during muscle lengthening. The adaptation error was the angular distance between the 2 and 15 s static I-O functions where they crossed a common frequency line that was similarly located. The rate errors of the more phasic receptors were in excess of 50° during the staircase stimulus sequence used in this study and these receptors were silenced during the movements that shortened the muscle. The static discharge of these receptors showed adaptation errors of 5–20° and linear directionality errors of 30–60° 2 s after movement terminated. The more tonic receptors had rate errors of 5–20°, adaptation errors of 0–5°, and linear directionality errors of 20–50° during movement, and 6–25° 2 s after movement terminated. Unless ‘corrected’ in some way, such errors would significantly compromise an animal's information about limb position under passive conditions, especially during and just after an imposed movement.