1. The role of proprioceptive pathways linking the direct antagonists soleus (S) and tibialis anterior (TA) muscles in governing the mechanical properties of the ankle joint were studied in the decerebrate cat. Actions of these heterogenic pathways were compared with those between S and extensor digitorum longus (EDL), a muscle that also acts at the metatarsophalangeal joint. These neurally mediated interactions between S and either TA or EDL were studied by applying controlled length changes to the isolated tendons of pairs of these muscles and recording the resulting changes in force. The muscles were activated with the use of electrically evoked crossed-extension reflexes, flexion reflexes, and brain stem stimulation. 2. Heterogenic inhibition from TA or EDL onto S was well developed whether S was initially quiescent or activated by a crossed-extension reflex. The inhibition persisted for the duration of the stretch of TA or EDL. During a crossed-extension reflex, TA did not generate background force, but brief stretch reflexes could be obtained. During flexion reflexes, stretch reflexes in S were usually abolished, and heterogenic inhibition from S to TA was weak or absent. 3. The strength of the heterogenic inhibition onto S was dependent on the initial length and activation level of TA and EDL. Changes in flexor length or activation level per se did not alter the background force or strength of the stretch reflex in S. Even taking into account the variation of strength of inhibition with the initial state of the muscle of origin, the strength of the inhibition was stronger from TA to S than the other way around. 4. The contributions of heterogenic inhibition from TA and EDL to S were independent in the sense that these components summed linearly with each other and with the autogenic reflex in S. In addition, the magnitude of the inhibition from TA to S was proportional to the amplitude of stretch for low to intermediate levels of initial force in S. The inhibition appeared to affect the mechanical responses of S essentially as rapidly as the stretch reflex in this muscle. 5. The heterogenic inhibition from TA to S was reduced or abolished by intravenous injections of strychnine but unaffected by injections of picrotoxin or bicuculline. These results, together with the observation that the inhibition sums linearly with the stretch reflex, suggest that the mechanism of this heterogenic inhibition is glycinergic and postsynaptic and, therefore, may include la-disynaptic reciprocal inhibition. 6. The heterogenic inhibition from TA and EDL to S was found to increase the in situ stiffness of S for both directions of ankle movement when physiological mechanical coupling between S, TA, and EDL was restored. This enhancement of stiffness is expected from the approximately linear summation of autogenic and heterogenic inputs. The magnitude of this action depends on the length and activation level of TA and EDL. 7. These results indicate that reciprocal inhibition to an actively contracting muscle constitutes an important channel of proprioceptive input when autogenic reflexes in the antagonist are active. One action of this proprioceptive pathway is to enhance the stiffness of the recipient muscle and, as a consequence, the stiffness of the spanned joint. The results also suggest that the strength of reciprocal inhibition may increase during coactivation. This heterogenic reflex operates simultaneously with the autogenic pathways that compensate for nonlinear properties of muscle.
