Spinal Interneurons: Much More Than Middlemen
Nociceptive afferents, which carry pain signals to the spinal cord, and the projection neurons that relay those signals to the brain, are understandably the focus of most research on the neural basis of pain. But recent papers suggest that spinal interneurons can be just as important for pain perception. In one of the latest such studies, Allan Basbaum’s group at the University of California, San Francisco (UCSF), US, found that knockout mice with a selective deletion of excitatory dorsal horn interneurons exhibited profound reductions in brain-mediated behavioral responses to a variety of painful and itch-inducing stimuli. The work appears in the April 24 issue of Neuron.
“I think it’s a great paper,” said Sarah Ross of the University of Pittsburgh, Pennsylvania, US, who has recently reported on an elevated itch phenotype in mice lacking inhibitory spinal interneurons (Ross et al., 2010). “These new findings really underscore the key role of the spinal cord in the modulation of sensory input. This has been underappreciated, and I think understanding the wiring of the dorsal horn is going to yield a great deal of insight into how sensory input is integrated and processed.”
Adopting an orphan
The new study had its genesis several years ago at a conference, where Basbaum heard that mice in which the testicular orphan nuclear receptor (TR4) gene was globally knocked out showed virtually no response to painful heat in the hot plate test, which measures supraspinal responses to thermal stimuli. The description was sufficiently intriguing to Basbaum to begin what became a two-year quest to obtain a mouse model with a conditional knockout of TR4 restricted to the central nervous system (CNS).
TR4 carries an “orphan” designation because the endogenous ligand for this nuclear receptor is unknown, and Basbaum said that little is understood about what the TR4 gene ordinarily does. However, when first author Xidao Wang and coworkers knocked out TR4 only in the CNS, the effects were remarkable. The mice exhibited normal reflexive withdrawal responses to heat in the Hargreaves and tail immersion tests, but in the hot plate test, the team writes, “Many of the animals went to cutoff, which is the maximal response permitted to avoid injury.” The TR4 knockouts were also significantly less responsive to acute chemically induced pain, and in the von Frey test of mechanically induced pain, the knockout mice showed considerably higher withdrawal thresholds than wild-type mice.
As for itch, which the researchers attempted to induce with three different agents, the behavior of the mice was “dramatic,” Basbaum said, unlike anything he has seen during his career in pain research. “No itch at all—nothing. It was really pretty amazing.”
In normal mice, standard models of inflammatory pain and nerve injury reliably generate long-term hypersensitivity to both thermal and mechanical stimuli, which is thought to result from “central sensitization” in the dorsal horn (reviewed in Basbaum et al., 2009). In a model of persistent inflammatory pain (injection of complete Freund’s adjuvant to the paw), the TR4 knockout mice developed thermal and mechanical hyperalgesia at levels comparable to wild-type mice.
However, in two models of nerve injury-induced neuropathic pain—the chronic-constriction injury (CCI) and spared nerve injury (SNI) models—the TR4 knockout mice revealed intriguing responses, especially in light of the reduced baseline sensitivity to mechanical pain in these mice noted above. In the CCI model, knockout mice were comparable to wild-type mice in thermal hyperalgesia, but exhibited no significant hypersensitivity to mechanical stimuli. Since thermal sensitivity is not manifest in the SNI model, the researchers tested only mechanical sensitivity, and again the TR4 knockout mice did not develop hyperalgesia.
Presented with this multifaceted phenotype, Basbaum said, the group began the “very long, drawn-out affair” of determining its anatomical and physiological foundations. The most obvious possibility, that primary nociceptive afferents were somehow damaged by the gene deletion, was quickly ruled out, as these neurons were shown to be structurally and functionally sound in the TR4 knockout mice. Moreover, a selective deletion of TR4 from afferents had no effect on pain behaviors.
The next hypothesis, that spinal projection neurons were faulty or reduced in number in the TR4 knockouts, also didn’t pan out. Retrograde labeling of dorsal horn projection neurons from supraspinal targets showed no difference in the number of cells in TR4 knockouts compared to wild-type animals. However, intracellular recordings of presumptive projection neurons in the dorsal horn in response to a variety of noxious stimuli revealed an altered pattern of activity: The firing patterns accurately reflected the presence and intensity of the stimuli, but the strength and duration of the neural response was reduced overall. This finding was complemented by measures of reduced Fos expression in a supraspinal target region.
This led the group to the notion that the TR4 knockout was somehow affecting interneurons, so as to weaken the afferent signal before it was passed on to projection neurons for supraspinal processing.
After what Basbaum described as considerable “grunt work” labeling and characterizing dorsal horn cells, the team eventually established that knocking out TR4 had indeed caused about a 40 percent reduction in the overall number of interneurons in the superficial dorsal horn. Drilling down further with a series of markers, the group observed no change in presumptive inhibitory interneurons, which led to the conclusion that the TR4 knockout predominantly affects excitatory cells.
Previous work by others (Sun and Chen, 2007) had shown that a subpopulation of spinal cord cells expressing the gastrin-releasing peptide receptor (GRPR) is necessary for itch. Given the striking itch phenotype in the TR4 knockout mice, it is notable that in these experiments the Basbaum group observed a nearly 80 percent loss of a subset of cells positive for gastrin-releasing peptide and the GRPR.
Next steps
Basbaum interprets these combined findings as showing that “in order for the brain to know what’s going on, it’s not sufficient to just activate the projection neurons, and it’s not a matter of interneurons being mere regulators of the output of the projection neurons—they’re absolutely essential” for the full expression of pain- and itch-related behaviors.
In the 2007 study of GRPR cells and itch by Sun and Chen referenced above, “It wasn’t established whether those cells were interneurons or projection neurons,” Basbaum said. “I think our data would argue that that’s a population of interneurons.” And because pain and itch were both profoundly affected by the loss of excitatory interneurons in the TR4 mice, “It raises the question of how pain and itch send their message to the brain,” Basbaum said. “Do they do it through the same projection neuron? Are there separate pathways? We haven’t answered that one yet, but we’re working on it.”
Basbaum speculates that the precise anatomical segregation of dorsal horn interneurons may allow subpopulations to regulate different responses within a single sensory modality—as in the TR4 knockout mice, which display marked mechanical hypersensitivity after tissue inflammation, but not after nerve injury.
Both Basbaum and Ross said this emerging picture is both scientifically exciting and clinically promising, and each is conducting follow-up studies on the wiring and function of these cells.
Ross has also studied bipolar cells (Kim et al., 2008), retinal interneurons that parse important features of the visual field, such as the edges of objects or their direction of movement, for further processing by amacrine and ganglion cells. She believes spinal interneurons may be playing a similar role, detecting and discriminating various properties of painful or itchy stimuli. “Just like in the retina, where bipolar cells and amacrine cells are involved in feature extraction for visual input,” Ross said, “I think it’s likely that spinal interneurons are going to be found to be involved in feature extraction for somatosensory input.”
Basbaum said his group also plans to begin screening excitatory spinal interneurons in search of unique or enriched G protein-coupled receptors in these cells that could be useful drug targets. “If we figure out a way to selectively regulate key populations of interneurons, we could probably regulate itch and pain independent of shutting down the projection neuron,” Basbaum said, allowing projection neurons to deliver informative, but sharply diminished, signals to the brain. “There will likely be some awareness of the stimulus by the brain, but you could selectively reduce the perception sufficiently to be clinically relevant without eliminating the response completely. That’s kind of interesting, because you don’t want to eliminate pain—you want to reduce the magnitude, the duration, the emotional significance of the stimulus, and the best way to do that may be via the interneurons.”
Pete Farley is a science writer based in Connecticut, US.
Article Contributors:
