Restoring balance to the reflex actions of the muscle spindle: the secondary endings also matter



The muscle spindle has two quite different kinds of sensory endings. Classically, attention was concentrated upon the primary endings and their large group Ia afferents while the secondary endings and their smaller group II afferents were neglected, even though the latter are just as numerous. The potential importance of secondary endings was recognized on finding that they better signal the absolute length of muscle, and undergo characteristic fusimotor regulation. So an ‘ordinary’ stretch receptor possesses two separate output channels that send linked but distinct signals to the CNS. Understanding why this occurs and how the CNS uses these differing signals remains a challenge.

The Ia afferents underlie the reflex response to untoward mechanical events, as in stumbling. Their actions have been studied exhaustively. They are easy to excite in isolation, both by small sharp tendon taps (because of their high dynamic sensitivity) and by weak electrical stimuli (being large, they have a low electrical threshold). In contrast, the spindle secondary afferents require larger stimuli, whether electrical or mechanical, and can only be activated along with other afferents. The reflex discharge then consists of serial responses. When its latency is suitably short the earliest component can be ascribed to the ‘knee jerk’ monosynaptic action of the Ia afferents; interestingly, human hand muscles with their refined skills have largely lost this simple uncontrollable response.

The origin of later reflex components remains problematic, with no shortage of candidates. The spindle group II afferents come first, with extra delay introduced by their slower peripheral conduction. However, based on animal experimentation the prevailing orthodoxy became that the spindle group II afferents joined other smaller fibres to comprise the flexor reflex afferents (FRA), producing non-specific flexor responses irrespective of their originating muscle. In functional terms this was nonsense, entailing the belief that the refined signals from spindle secondaries were wasted. They then slipped into limbo, while the ability of the Ia afferents to produce delayed responses began to emerge. Two technical considerations were clarified: mechanical stimuli can oscillate, triggering successive Ia bursts eliciting multiple monosynaptic responses, and motoneurones' natural rhythmicity can segment responses. More significantly, Ia afferents were found to produce delayed excitation via spinal interneurones, over and above acting monosynaptically. Further, evidence accumulated that they could also excite motonerones via ‘long loop’ transcortical pathways that were particularly potent for human finger muscles. Crucially, unlike the monosynaptic reflex, the putative ‘transcortical response’ occurred bilaterally in patients with mirror movements; due to aberrations in neural wiring, each side of their cerebral cortex activates MNs for both hands. Moreover, cooling the arm to slow nervous conduction produces only minor slowing of ‘transcortical’ responses, appropriate for Ia action (large axons lose less time than small axons). Thus, the importance of the muscle spindle primaries in reflex control became ever more fully established, along with recognition that monosynaptic excitation was only part of the story.

Human investigation now predominates, searching for group II action. The logic is simple. Different muscles have different functions and are subject to differing levels of control by the cerebral cortex. Thus the reflex control of different muscles probably varies, with some aspects sharpened and others damped down. Demonstration of powerful cortically mediated Ia action for highly encephalized muscles leaves the group II question open. Recently, evidence for a meaningful reflex role for spindle secondary afferents has built up for the lower limb. The paper by Lourenço et al. (2006) in the current issue of The Journal of Physiology importantly extends this to the upper limb, for which they have analysed the response of arm muscles to electrical stimulation of a mixed nerve, supplying functionally related hand muscles.

Every available manoeuvre was employed to circumvent the difficulties, and detailed perusal of their straightforward findings presents difficulties for the uninitiated. Ia afferents provide a tripartite contribution. Initial monosynaptic activation is followed successively by variable spinal interneuronal and transcortical responses. With strong stimuli Group II afferents elicit a delayed spinal response, falling in between the two late Ia responses. This is not from cutaneous afferents, being absent on cutaneous stimulation. It is not from Ia afferents, because it has a high threshold and there is excessive delay on peripheral cooling. It is not transcortical because it occurred unilaterally in a patient with mirror movements.

We have moved on. The question ‘Which one of the various potential afferent inputs is responsible for the delayed responses?’ has been replaced by ‘What physiological contribution does each of several pathways provide to human reflex motor control in each particular situation?’ The power of electrical stimulation as an initial tool is shown by its success in dealing with the first question; here it demonstrated the existence and functioning of particular pathways. Its weakness is brought out by its inability to address the emerging second question. The elegant analysis needed to detect small responses necessarily vitiated quantitative comparisons. Individual motor units were studied with the stimulus timing locked to their ongoing firing. The initial Ia monosynaptic volley reached the motoneurone early in its recovery cycle when it was still largely refractory, and hence the motoneurone with its omnipresent synaptic noise was only rarely excited to discharge. The polysynaptic Ia and group II inputs arrived later, as the motoneurone recovered, and so were able to elicit a significant discharge, detectable with averaging. Conventional random timing of the stimulus would have elicited more monosynaptic discharge from the motoneurone, leaving it refractory, unready to respond to later inputs. The existence of delayed pathways was demonstrated unequivocally for synchronous activation of afferents from functionally related muscles, but their synaptic potency remains uncertain. Quantitative measures of autogenetic reflexes are required for full understanding of physiological functioning.

So, both primary and secondary spindle afferents take part in reflexly regulating human voluntary contraction. This seems a teleological truism, for why else should both exist? However, experiment is all, the essential arbiter of untested beliefs.