Following inflammation or nerve injury, stimuli that are normally perceived as innocuous can evoke persistent pain. A population of neurons that contributes to this syndrome has now been identified.
For many people with persistent pain syndromes, even the touch of something as innocuous as a shirt on their skin can be agonizing. Determining the underlying cause of this debilitating pain is essential for developing effective therapies. However, the precise identity of the neural circuits involved has so far remained elusive. In this issue (page 651), Seal et al.1 pinpoint a previously enigmatic population of small sensory neurons that respond to light touch, and show that they have a crucial role in the painful sensitivity to touch or pressure that follows injury or inflammation.
The sensations of pain and touch, like other bodily sensations such as temperature and itch, begin with the activation of distinct subsets of primary sensory neurons by physical stimuli affecting the body. Once activated, these neurons excite circuits in the dorsal horn of the spinal cord. It is there that the sensory input is processed and then relayed to centres in the brain, where it is perceived and interpreted.
The dorsal horn has a laminated, or layered, structure, with each lamina receiving input from multiple but distinct classes of sensory neurons that extend from the body's periphery and convey information about various sensations (Fig. 1). These neurons release the excitatory neurotransmitter glutamate to activate neurons in the dorsal horn. Glutamate is packaged into synaptic vesicles in neurons by proteins called vesicular glutamate transporters (VGLUTs), of which there are three types. Previously, only the two most common types, VGLUT1 and VGLUT2, had been characterized in sensory neurons. VGLUT1 is expressed in touch-sensing fibres (heavily myelinated Aβ fibres with low activation thresholds) that terminate in lamina III/IV and the inner part of lamina II (IIi) in the dorsal horn2,3. By contrast, sensory neurons that innervate the more superficial laminae, I and II, are generally smaller and mainly associated with sensing painful (including strong mechanical) stimuli, temperature and itch. These neurons have thin, either lightly myelinated Aδ fibres or unmyelinated C fibres, most of which express VGLUT2 (ref. 3). (Myelination increases the speed at which electrical impulses pass along a neuron.)
Seal et al.1 show that the little-studied VGLUT3 is expressed in a specific subpopulation of small sensory neurons that have unmyelinated C fibres. These fibres terminate in lamina I and also in lamina IIi, where they overlap with the most dorsal innervation of conventional, touch-sensing Aβ fibres (Fig. 1). It has been known for some time that there is a small subset of sensory neurons in the C-fibre population that terminates primarily in lamina II and that responds to innocuous mechanical stimuli, such as light touch, rather than to noxious stimuli4. By performing electrical recordings on a preparation of sensory neurons with their fibres still attached to skin, Seal et al. demonstrate that the VGLUT3-expressing sensory neurons are activated by brush and light touch, and so represent these previously ill-defined, low-threshold C-fibre mechanoreceptors.
The authors then assessed the function of low-threshold C-fibre mechanoreceptors by studying mice in which the vGlut3 gene had been knocked out5. The absence of VGLUT3 functionally disconnects this subset of peripheral sensory neurons from their dorsal-horn targets by preventing the normal release of glutamate. Surprisingly, in the absence of injury or inflammation, these knockout mice had small defects in their responses to noxious mechanical stimuli, but responded normally to low-intensity mechanical stimuli and also to hot and cold stimuli. This somewhat paradoxical result suggests that, despite the low-threshold mechanical sensitivity of the VGLUT3-expressing neurons, they can contribute to the detection of noxious mechanical pain.
Seal and colleagues' most striking findings1, however, came in the next round of tests, in which the authors used three manipulations in mice as models of persistent pain states. These comprised: injection of an irritant substance into the footpad, causing a massive inflammatory response; an incision in the paw that is considered a model of post-surgical pain; and a lesion of the peripheral nerve to model neuropathic pain. In the last case, injury to peripheral nervous tissue causes a persistent pain state, best defined6 as a pathological response of the pain system to nerve damage. In all of these conditions, animals whose paws are exposed to mechanical and thermal stimulation will usually tolerate a far lower stimulation intensity than normal mice before withdrawing their paws. In all three models, the VGLUT3-knockout mice and normal control mice showed similar hypersensitive responses to heat. However, knockout animals showed much less hypersensitivity to mechanical stimulation than did control animals. In fact, in the model of inflammation, the VGLUT3-knockout mice did not develop mechanical hypersensitivity. Thus, preventing synaptic input from these low-threshold C-fibre mechanoreceptors seems to remove a crucial mechanical input that is either partially or completely required for responses to mechanical touch in these hypersensitive states.
One of many questions highlighted by this work is why two very different types of low-threshold mechanoreceptor — a subset of Aβ fibres and a subset of VGLUT3-expressing C fibres — both terminate in the same lamina IIi region of the dorsal horn. This region has been implicated7,8 as a key processor of low-threshold inputs that can become painful when there is inflammation or nerve injury. Understanding the physiological changes occurring in this region in such conditions may be crucial for understanding how normally innocuous inputs can activate pain pathways.
Why are there two distinct low-threshold inputs? If we consider that the Aβ-fibre input is important for identifying the location of touch, perhaps the C-fibre input instead works to amplify the Aβ-mediated responses under injury conditions and so to modify the accompanying protective responses and perception of pain. The authors note that, under normal conditions in humans, activation of low-threshold mechanoreceptor C fibres by brushing correlates with the sensation of pleasurable touch9. Although it remains to be confirmed that the VGLUT3 neurons are responsible for sensing pleasurable touch in humans, a sensory subsystem that can change from eliciting feelings of pleasure to evoking pain following inflammation or injury would indicate a hitherto unexpected level of plasticity in an already remarkably flexible system.
Seal, R. P. et al. Nature 462, 651–655 (2009).
Neumann, S., Braz, J. M., Skinner, K., Llewellyn-Smith, I. J. & Basbaum, A. I. J. Neurosci. 28, 7936–7944 (2008).
Todd, A. J. et al. Eur. J. Neurosci. 17, 13–27 (2003).
Sugiura, Y., Lee, C. L. & Perl, E. R. Science 234, 358–361 (1986).
Seal, R. P. et al. Neuron 57, 263–275 (2008).
Costigan, M., Scholz, J. & Woolf, C. J. Annu. Rev. Neurosci. 32, 1–32 (2009).
Malmberg, A. B., Chen, C., Tonegawa, S. & Basbaum, A. I. Science 278, 279–283 (1997).
Miraucourt, L. S., Dallel, R. & Voisin, D. L. PLoS ONE 2, e1116 (2007).
Löken, L. S., Wessberg, J., Morrison, I., McGlone, F. & Olausson, H. Nature Neurosci. 12, 547–548 (2009).
About this article
The effects of preferential A- and C-fibre blocks and T-type calcium channel antagonist on detection of low-force monofilaments in healthy human participants
BMC Neuroscience (2015)
Experimental Physiology (2013)
Journal of Cognitive Neuroscience (2012)
Neuroscience Bulletin (2012)