Allodynia is a hallmark of neuropathic pain in which normally non-painful stimuli, such as a light touch, evoke pain. Researchers have recently taken strides toward understanding the neuronal circuitry in the dorsal horn of the spinal cord that might cause non-noxious touch to be interpreted as painful. Those advances include recent findings from Reza Sharif-Naeini, McGill University, Montreal, Canada, who presented his work during Dorsal Horn Circuits Underlying Touch-Evoked Pain After Nerve Injury, a webinar hosted by the Pain Research Forum on June 15, 2016. Sharif-Naeini was joined by panelists Rebecca Seal, University of Pittsburgh, US, Daniel Voisin, Université de Bordeaux, France, and moderator Andrew Todd, University of Glasgow, UK.
When touch causes pain
Researchers have long postulated that allodynia arises from increased excitability in spinal cord signaling circuits. That excitation could arise from augmented neurotransmission to lamina I neurons projecting from the spinal cord to pain centers of the brain. A large body of evidence now supports the idea that allodynia is the result of decreased inhibitory signaling—disinhibition—in the spinal cord dorsal horn after nerve injury, and that nerve pain leads to an “unmasking of polysynaptic pathways [in the spinal cord] that link non-noxious stimuli to the pain pathway,” Sharif-Naeini said. Multiple mechanisms probably contribute to the reduction in inhibitory tone in the dorsal horn.
Building on previous work from many researchers, Sharif-Naeini presented a model of a polysynaptic circuit in which Aβ fibers, carrying innocuous touch signals, enter the spinal cord at the deep laminae and make connections with excitatory interneurons containing the gamma isoform of protein kinase C (PKCγ), and with inhibitory interneurons containing parvalbumin (PV), which Sharif-Naeini focused on (see PRF related news story). PV interneurons receive inputs from and make synaptic contacts with central terminals of large, myelinated sensory fibers. The PV neurons contain both the inhibitory neurotransmitters GABA and glycine. Sharif-Naeini wanted to determine whether PV neurons could serve as the “gate” that normally keeps innocuous touch stimuli from activating the pain pathway, and how they might change after nerve injury.
To do that, he created transgenic reporter mice expressing a fluorescent protein specifically in the inhibitory PV-containing neurons. He first asked whether the loss of inhibition leading to mechanical allodynia was due to apoptotic death of the PV neurons, but there was no evidence of cell loss in the spared-nerve injury (SNI) model of neuropathic pain.
Sharif-Naeini and colleagues next used a DREADD (designer receptors exclusively activated by designer drugs) strategy to activate the PV neurons. The researchers unilaterally injected an adeno-associated virus carrying a Cre-dependent DREADD gene to the spinal cord dorsal horn of knock-in mice expressing Cre recombinase under control of the PV promoter. When they delivered a synthetic ligand to activate the DREADD, it activated the PV neurons for about an hour. During that time, naïve (uninjured) mice became hyposensitive to mechanical (but not thermal) stimuli on the side where the virus was injected, but not on the contralateral side, indicating that PV neuron activity dampened mechanical sensitivity. Next, the team took the inverse approach and silenced the PV neurons using an inhibitory DREADD in naïve mice. Those mice, in contrast, displayed significant mechanical allodynia in response to testing with von Frey hairs, even in the absence of nerve injury. Together, the results suggested that “PV neurons appear to be modality-specific filters of incoming information whereby they modulate mechanical inputs but do not influence thermal inputs,” Sharif-Naeini said.
Sharif-Naeini then asked, “If we can turn the PV neurons on in mice that have mechanical allodynia, can we attenuate that?” To that end, they determined the mechanical stimulus intensity necessary to evoke a response in each animal 60 percent of the time. Compared to naïve mice, those with SNI required a much lower intensity, but in mice transfected with the excitatory DREADD and treated with the ligand to activate the PV neurons, the responses shifted back toward those of uninjured animals. “PV neuron activation significantly attenuated the allodynia in these animals,” Sharif-Naeini said.
Using a different model in which capsaicin is injected into the paw to produce mechanical and thermal hypersensitivity, the researchers discovered that activation of the PV neurons again attenuated mechanical allodynia without affecting thermal hypersensitivity. That supports the idea that “these neurons are really specialized in controlling mechanical inputs to the spinal cord,” Sharif-Naeini said.
What’s the target?
Next, Sharif-Naeini and his fellow investigators wanted to identify the molecular targets of the PV interneurons. They started with the PKCγ interneurons, because previous reports had suggested an interaction between the two cell types. Using confocal microscopy and three-dimensional reconstruction, the researchers detected synapses between PV neurons and the somata of PKCγ neurons. “They may also have other targets, but one of the targets of the PV neurons are these PKCγ neurons,” he said.
PKCγ is more than a marker for the interneurons that express it; the enzyme likely makes essential contributions to their function. In other types of neurons, PKCγ enhances glutamatergic signaling. “If we inhibit that enzyme, we can remove the contribution of these neurons to the polysynaptic pathway,” Sharif-Naeini said. After intrathecal injection of a PKCγ inhibitor, the researchers saw no change in acute tactile sensitivity, but it attenuated allodynia in nerve-injured mice. The team also looked at Fos expression—an indication of neuronal activation—in nerve-injured animals. “If we simply brush the paw surface after injury, that induces Fos in neurons in the superficial layers of the dorsal horn, but if we pretreat with a PKCγ antagonist, that significantly decreased activation of the lamina I neurons,” Sharif-Naeini said.
The researchers also observed that eight weeks after nerve injury, appositions onto PKCγ soma had drastically decreased, due to some withdrawal or pruning process, Sharif-Naeini hypothesized. That loss of inhibitory inputs “increases the probability that Aβ inputs onto PKCγ neurons will activate them and essentially carry the excitation to the lamina I neurons” in the pain pathway, Sharif-Naeini said. To artificially recreate that pruning process, Sharif-Naeini and colleagues genetically manipulated mice to selectively express in PV neurons the neurotoxic peptide saporin, which kills the cells. Confocal microscopy showed a decrease in appositions onto PKCγ neurons that corresponded with the development of mechanical allodynia in uninjured mice. Injection of the PKCγ inhibitor attenuated allodynia, indicating that the PV neurons normally inhibit the PKCγ neurons.
Sharif-Naeini proposed an addition to the model of known dorsal horn circuitry: touch-sensitive Aβ fibers contact PKCγ neurons, but the Aβ fibers may have a collateral branch that recruits PV neurons, which in turn feed back onto PKCγ neurons to prevent their excitation. After nerve injury, decreased PV neuronal synapses onto PKCγ neurons could allow Aβ fibers to engage the polysynaptic pathway leading to lamina I neurons and cause mechanical allodynia. “If we can increase output from the PV neurons, we can rescue or even prevent allodynia,” Sharif-Naeini said. “We think that they could be one of the gates that normally prevents Aβ fibers from activating nociceptive circuits.”
Panel discussion and audience questions
Following his presentation, Sharif-Naeini took questions from the panel and audience members. Rebecca Seal asked about the role of GABA in the PV neurons, considering that they contain both glycine and GABA, and because other work suggests that glycine drives the neurons’ inhibitory signaling. Sharif-Naeini said that, in addition to the synapses onto PKCγ neurons, PV interneurons also make axo-axonic synapses onto the central terminals of Aβ fibers, where GABA may be more important. One possibility is that the PV neurons release both neurotransmitters indiscriminately, but the response is determined by the receptors present postsynaptically: by GABA receptors on Aβ fibers and by glycine receptors on PKCγ neurons. Daniel Voisin added that the two neurotransmitters may play different roles in mediating static (pressure) versus dynamic (brush) mechanical allodynia.
Andrew Todd wondered about the relative functional importance of the PV neuronal synapses onto Aβ fibers versus onto PKCγ neurons; Sharif-Naeini said that both types of synapses could be contributing to maintaining inhibition under normal circumstances, and that further studies will be needed to tease apart the contributions of each synapse type. In ongoing studies, he is examining the time course and spatial distribution of the loss of PV-PKCγ synapses. He noted three ways that changes could lead to disinhibition and allodynia: decreased excitatory drive from Aβ fibers onto PV neurons; alterations in the intrinsic membrane properties of the PV neurons; and decreased inhibitory inputs from PV neurons to PKCγ and/or Aβ neurons. For now, he added, his group is focused on the last two possibilities.
Seal then asked about the possibility of multiple neuronal types that could act as inhibitory “gates” between touch and pain sensations, considering recent findings of inhibition by dynorphin-containing interneurons in lamina II, and her own recent studies of excitatory calretinin-positive neurons, also in lamina II (see PRF related news). Seal’s work suggests that calretinin and PV neurons may be differentially disinhibited to produce allodynia depending on the type of injury. Sharif-Naeini said it makes sense that more than one inhibitory neuron type is important. “You would want multiple gates to prevent innocuous information from reaching pain neurons.” Supporting this idea, he added, anterograde tracing studies point to additional downstream targets of the PV neurons.
Several audience members asked about the potential for translating the findings into human therapeutics. While any clinical application is certainly years away, Sharif-Naeini did see a path forward. “The data show that if we turn on PV neurons in mice, we can attenuate allodynia. So we are now trying to see what makes PV neurons unique. If we can find a receptor that is specifically expressed on PV neurons in the dorsal horn and then screen for agonists, [such an agent] might provide analgesia.” Seal added, “Overall, this is an exciting time to think about the possibility of new drugs, because these circuits have been mysterious for so long. If we can start to get a handle on how they’re wired and the mechanisms underlying the transition from acute to chronic pain, I think we’re going to find many new targets for therapeutics, so I feel very optimistic.”
Voisin agreed that it is an exciting time and that researchers are making important advances, including some changes to Melzack and Wall’s initial model, which proposed that wide dynamic range (WDR) neurons rather than Aβ neurons were responsible for mechanical allodynia, as more recent work has shown. Todd agreed, saying that after 40 years in the field, he has seen an extraordinary acceleration of new findings in recent years thanks to new technology and tools. “We know a lot more than we did 10 years ago, but there is still an enormous amount to do,” he concluded.
Stephani Sutherland, PhD, is a neuroscientist, yogi, and freelance writer in Southern California.
Image credit: maxximmm/123RF Stock Photo