Tactile suppression commonly refers to the reduction in tactile perception that occurs during movement, or what is also called movement-related gating. The function of tactile suppression is most likely to suppress redundant movement-related feedback that can be predicted from the motor command so that the perception of unexpected or novel inputs is enhanced. The central motor command plays a key role in generating tactile suppression. Peripheral feedback from the moving limb also contributes because tactile suppression is seen during passive movement, i.e. in the absence of a motor command.
Contents |
The senses provide continuous information about the extra- and intra-personal environment that we inhabit and, together with the motor systems, allow us to both interpret and interact with the environment. The somatosensory system occupies a particularly privileged position in this schema since primary somatosensory cortex, S1, is the only primary sensory receiving area to have direct, and reciprocal, connections with primary motor cortex, M1, and so direct access to the motor system. One challenge for the Central Nervous System (CNS) is, however, to process the vast amount of somatosensory input to the CNS at any moment of time since the receptor sheet for discriminative touch covers the entire surface of the body (∼ 2 m2). Not too surprisingly, the brain has developed mechanisms to control the flow of sensory inputs to the central processing regions, with tactile suppression during movement (also referred to as movement-related gating) being an example of one such control mechanism. While the following text refers mainly to results obtained in primates, tactile suppression is not limited to higher mammals but has also been intensively studied in rodents specifically in relation to whisking movements of the vibrissae (for a review, see Kleinfeld, Ahissar, & Diamond, 2006).
During movement, near threshold tactile stimuli (intensities that can be detected at rest) are not detected, i.e. they are totally suppressed (Chapman, Bushnell, Miron, Duncan & Lund, 1987; Post, Zompa, & Chapman, 1994). Similar results are obtained with mechanical (e.g. vibration) and electrical stimulation. The decrease in perceptual sensitivity begins well before the onset of movement (Figure 1B(a)). Tactile suppression is maximal (no stimuli detected) at about the same time as the limb begins to move. When the timing of the decrease is examined relative to the onset of electrical activity in the muscles involved in the movement (EMG, electromyographic activity), i.e. relative to the time that the motor command arrives at the periphery (Figure 1B(b)), it can be seen that detection concomitantly begins to decrease prior to EMG onset (Williams, Shenasa, & Chapman, 1998). Such observations suggest that the motor command contributes to generating tactile suppression.
As the intensity of the test stimulus is increased (not illustrated) the degree of suppression declines so that stronger tactile stimuli are all perceived, but their subjective intensity is diminished (Williams & Chapman, 2000). If the test stimulus is sufficiently strong, then there may be no drop in subjective intensity during movement. Finally, the ability to discriminate differences in the intensity of pairs of tactile stimuli (just noticeable difference) is unaffected during movement, indicating that the relative differences between tactile inputs are preserved during movement even though their subjective intensity may be decreased (Chapman et al., 1987; Post et al., 1994).
The degree of tactile suppression with movement is modulated by several factors. The speed of the movement is important (Cybulska-Klosowicz, Meftah, Raby, Lemieux, & Chapman, 2011): there is no attenuation of tactile detection for very slow movements, < 200 mm/s or 25°/s, corresponding to speeds often used in tactile exploration (Smith, Chapman, Deslandes, Langlais, & Thibodeau, 2002). Beyond this, faster movements result in greater suppression. The physical relation between the movement and the stimulus is also critical, with greater and earlier suppression occurring with closer physical proximity between the active muscles and the tactile stimulus (Williams, Shenasa & Chapman, 1998). Moreover, there is an anatomical limit to the suppressive effects with the effects being largely, if not exclusively, restricted to the moving limb. The critical factor for tactile suppression is motor activity and not movement per se, since similar sensory attenuation is seen with isometric contractions, when no overt movement is produced (Williams & Chapman, 2002). Tactile suppression can be elicited in the absence of descending motor commands when the limb is passively moved, indicating that reafferent feedback also has a role in modulating sensory perception (Williams & Chapman, 2002; Chapman & Beauchamp, 2006).
Sensory inputs that are self-induced also show evidence of a modest reduction in their subjective intensity during the movement that generates the sensation (e.g. Bays, Wolpert, & Flanagan, 2005; Bays, Flanagan, & Wolpert, 2006). Under certain conditions, this modulation appears to be entirely central in origin, dependent on the predicted sensory consequences of the movement. This mechanism differs from the modulation seen for externally applied stimuli, however, since the effects are not limited to the moving limb but extend to include the contralateral limb. This predictive mechanism, triggered by the motor command, helps to explain why we cannot tickle ourselves (Weiskrantz, Elliott, & Darlington, 1971).
To summarize, the signals responsible for suppressing or modulating tactile perception arise from at least two different sources – the motor command (efference copy or corollary discharge) and the sensory feedback associated with the movement or motor activity (reafference). Finally, peripheral feedback alone can suppress the detection of threshold-level stimuli prior to the onset of movement (Williams & Chapman, 2002; Chapman & Beauchamp, 2006), an effect attributed to backward masking whereby the strong signal from movement reafference suppresses the perception of an earlier, but weak, test stimulus. Interestingly, similar effects are also reported in the visual system (reviewed in Wurtz, 2008), with both the motor command and visual feedback contributing to visual suppression. Furthermore, there is evidence for a suppression of proprioception ("muscular sense") as well (Collins, Cameron, Gillard & Prochazka, 1998).
Tactile inputs from cutaneous mechanoreceptors of, for example, the hand are relayed through the dorsal column-medial lemniscal (ML) pathway to sensory thalamus (ventral posterior lateral nucleus, VPL) and thence to S1. Inputs to the dorsal column nuclei are both direct (collaterals from primary afferents) and indirect (after first synapsing in the spinal dorsal horn). The central neural mechanisms underlying tactile suppression have been studied by recording the amplitude of short-latency somatosensory evoked responses (SEPs) to tactile stimulation in the alert monkey, with the test stimulus being applied at various times prior to and during movement. Figure 2 shows representative recordings from S1 taken with the animal at rest (left, inter trial interval) and while performing a simple reaction-time task (right). In this example, the tactile stimulus, an air puff directed to the centre of the receptive field on the forearm, was applied shortly after the onset of movement (elbow flexion), i.e. after the onset of electromyographic (EMG) activity in the biceps, the main agonist. There was a pronounced decrease in the amplitude of the S1 cortical SEP during movement (Chapman, Jiang, & Lamarre, 1988), consistent with a reduction in the transmission of tactile inputs to S1. Short-latency single unit responses to natural tactile stimuli (air puff) in S1 are likewise decreased prior to and during voluntary movement, consistent with a profound decrease in S1 responsiveness to tactile inputs before and during movement (Jiang, Chapman, & Lamarre, 1991).
The time-course for movement-related SEP modulation in S1 is shown in Figure 3A. The SEP begins to decrease at about the onset of EMG activity (first vertical line), with the maximal decrease occurring after the onset of movement (second vertical line). Similar effects are seen with (isotonic) and without (isometric) movement about the joint (Jiang, Lamarre, & Chapman, 1990). The effects are, moreover, independent of the direction of the movement (flexion versus extension of the elbow).The degree of modulation covaries with the speed of movement – greater gating with faster movements (not shown). As in the perceptual studies, the gating effects seen with active movement are also observed with passive movements (Figure 3B), confirming the importance of peripheral reafference to tactile gating. The motor command clearly also contribute: the onset of the decrease in the S1 SEP during passive movement is delayed by about 80 ms compared to active movement, beginning at the onset of passive movement. Thus, the signals contributing to tactile suppression are both peripheral and central in origin.
Tactile suppression is present at the earliest possible relay stations, the spinal cord (Seki, Perlmutter, & Fetz, 2003) and the dorsal column nuclei (Ghez & Lenzi, 1971), and precedes the onset of movement. Pioneering studies by Ghez & Lenzi showed that the underlying synaptic mechanism involves presynaptic inhibition (primary afferent depolarization, PAD) of primary cutaneous afferents at the level of the dorsal column nuclei, thus reducing transmission at the first main relay from the periphery to S1 cortex. More recently Fetz and colleagues (Seki, Perlmutter, & Fetz, 2003) have extended these results to show that PAD is also present at the level of the dorsal horn of the spinal cord. Their recordings, made in awake, behaving primates, showed that the inhibition of cutaneous afferent input to spinal cord interneurones begins well before the onset of EMG activity and voluntary movement.
The suppressive effects seen at the earliest relay stations are, however, quite modest in amplitude (Figure 3A,B – ML). The maximal degree of suppression during movement gradually increases on ascending to higher levels of the pathway (VPL, S1), with the time-course for this modulation being almost identical at all levels (Chapman et al., 1988). Most importantly, the time-course for the decrease in SEP amplitude is almost identical to the time-course for tactile suppression of detection in humans (Figure 1). Peripheral feedback from the moving limb (passive movements) decreases the amplitude of SEPs at all levels of the pathway (Figure 3B), with greater decreases occurring at higher levels (S1 versus ML). In contrast, the motor command appears to particularly act at lower levels of the pathway (spinal cord, dorsal column nuclei) since S1 cortical responses to stimulation of either the medial lemniscus or VPL thalamus are only modulated after the onset of movement, active or passive (Figure 3C-D).
While tactile suppression is present at all levels of transmission of tactile inputs, there is little or no evidence for visual suppression during saccades in, for example, primary visual cortex, V1 (Wurtz, 1968) although it is found in the extrastriate areas (see Wurtz 2008). This suggests that the neuronal mechanisms underlying visual suppression are different from those underlying tactile suppression.
Evidence that the motor commands contribute to generating tactile suppression has been obtained in studies carried out in awake monkeys. Jiang, Chapman, & Lamarre (1990) showed that intracortical microstimulation applied to the arm representation of M1, essentially generating a weak artificial motor command, leads to a decrease in the amplitude of SEPs in S1 elicited by air puff stimuli applied to the arm. Similar results are obtained even when the stimulation intensity is reduced below the threshold for eliciting EMG activity. Moreover, the effects show a spatial gradient, limited to gating inputs from skin areas overlying, or distal to, the motor output. Thus there is direct evidence that the motor command can diminish the transmission of cutaneous inputs to S1 cortex. Similar observations have since been made in humans using transcranial magnetic stimulation to activate M1 (Voss, Ingram, Haggard, & Wolpert, 2005).
Finally, there is evidence that these suppressive effects extend beyond the pathway relaying tactile inputs to S1. Seki & Fetz (2012) have shown that cutaneous inputs to M1 and premotor cortex are also diminished during active, as well as passive, movement.
The function of tactile suppression, or movement-related gating of tactile stimuli, is most likely to suppress inputs that can be predicted from the motor command, so as to enhance the detection of other novel stimuli (Coulter, 1974; Chapman, Jiang & Lamarre, 1988). These controls act to limit the quantity of afferent feedback that is processed at higher levels of the neuraxis.