Neuropathic Pain Symptoms Arise From Different Mechanisms
Neuropathic pain typically features three hallmark symptoms: spontaneous pain, cold allodynia, and tactile allodynia. New research examining gene expression changes in mice now shows that the last two symptoms develop independently, at different times, via two distinct mechanisms: cold allodynia depends entirely on nociceptive sensory neurons, whereas tactile allodynia requires immune cells but not thermally sensitive nociceptors.
“There is a clear temporal separation of neuronal and immune involvement,” said Simon Beggs, University College London, UK, who was not involved in the study. “With this very detailed gene expression profile, they have shown what people have been thinking lately: that immune cell activity is probably important for maintenance of chronic pain states, but not for the earliest times,” when neuronal dysfunction dominates.
The work, published January 30 in Cell Reports, was led by Michael Costigan and Clifford Woolf, both of Boston Children’s Hospital and Harvard Medical School, Boston, US.
A different time course
In the clinic, cold and tactile sensitivity have been observed to arise separately from one another, and basic researchers have recently focused on determining the mechanisms underlying each phenomenon (Jensen and Finnerup, 2014).
“Our approach was that if we looked very carefully at the two pain mechanisms, we might see differences in how the symptoms develop,” Costigan told PRF. Using the spared nerve injury (SNI) model of neuropathic pain, the group found that cold allodynia developed very early—beginning the day after injury, peaking about four to five days and continuing for at least 15 days. Tactile allodynia, in contrast, came on more slowly, becoming significant only after five days and peaking about seven days.
“Not everyone would agree with the idea that tactile allodynia takes seven days to develop,” Beggs said. “Some people see behaviors peak at one day after injury,” and the original paper describing the SNI model reported that tactile allodynia was significant the first day after injury (Decosterd and Woolf, 2000). In neuropathic pain models, Beggs said, behavioral responses can be extremely variable from one lab to another, depending on the specific methods used.
Costigan agreed that timing can indeed vary significantly from one study to another, and that allodynia develops sooner in rats than in mice, contributing to variability. But despite that, he said, “it is almost universally accepted that full neuropathic pain sensitivity takes at least a week to peak, and cold allodynia usually precedes tactile allodynia.”
In any case, Beggs said, the timing discrepancy does not take away from the study’s mechanistic findings.
Gene expression changes point to distinct mechanisms
The researchers next wanted to examine gene expression changes in temporal detail following injury, performing gene expression microarrays eight and 16 hours after injury, and then daily for 10 days. “The idea was that if there were differences in when the phenotypes developed, we might be able to correlate those with what’s happening at the molecular level and see what’s going on mechanistically,” Costigan said.
“They have achieved a high degree of granularity with their data—they got a very high-resolution look at gene changes by doing the arrays as often as they did,” Beggs said, which he called the strength of the paper.
First and third authors Enrique Cobos and Fuying Gao analyzed the arrays with a technique called weighted gene co-expression network analysis (WGCNA), which grouped genes into “clusters” that behaved similarly—that is, their expression level changed at the same time, whether they were up- or downregulated. From the 1,704 transcripts deemed significantly regulated across the array gene set, the WGCNA produced eight such clusters. The very first genes to show changes in expression were associated with chemotaxis in immediate response to nerve damage. Next were three distinct waves of genes related to neuronal function, followed by a group of mixed immune, neuronal, and other genes, and finally a late immune wave of genes regulated about a week after injury.
In general, the time course of expression changes in neuronal genes matched the time course of the development of cold allodynia, and the late immune-associated gene changes matched the timing of the tactile allodynia, the investigators observed.
Next the authors formally correlated the onset curves of cold and tactile allodynia in the mice to the full regulated gene set to produce a set of transcripts temporally related to the development of each pain modality. They determined the cellular mechanisms associated with each wave of gene expression based on multiple pathway analysis tools. “Overwhelmingly, the cold group was neuronal, and the tactile group was immune,” Costigan said. “We showed that in three different ways.”
The first two methods used computer programs to determine the cells associated with the gene expression changes. One was a commercial program called Ingenuity Pathway Analysis (IPA), and the other the Gene Ontology (GO) program developed at the National Institutes of Health (NIH).
“The data were very strong, so we wanted to hammer home that we weren’t pushing it in any way, shape, or form” by manipulating the computer programs’ parameters, Costigan said. So next the team compared the genes closely related to cold and tactile allodynia with genomewide expression studies of isolated sensory neurons or activated immune cells (macrophages and T cells) grown in culture. The association held up: Genes that were regulated at the onset of cold allodynia were expressed in neurons, whereas genes associated with tactile allodynia were expressed in immune cells.
“You can see fundamental differences in the genes regulated in neurons compared to immune cells,” Costigan told PRF. “Damaged neurons respond relatively quickly, but the immune genes were regulated later; they’re moving from the blood to the DRG [dorsal root ganglia], so it’s a slower process.”
Neuropathic pain is contributed by two distinct processes: one that occurs in neurons, leading to cold allodynia, and another in immune cells, leading to tactile allodynia. Image and caption from Cobos et al., 2018.
Testing the model in animals
The researchers next wanted to test if the findings from the genetic analyses also applied to behaving animals. If cold allodynia was, in fact, rooted in sensory neurons, the symptom should not develop in animals lacking those neurons, the investigators reasoned. To support this idea, they engineered mice such that neurons expressing the transient receptor potential vanilloid type 1 (TRPV1) channel also produced a toxin called DTA under control of the TRPV1 promoter, which essentially kills the neurons before the animals are born. As a result, the TRPV1-DTA mice contained no TRPV1-lineage nociceptive neurons, but mechanically sensitive neurons remained intact.
TRPV1-lineage neurons also contain TRPM8, the primary sensor of noxious cold. Without those receptors, uninjured TRPV1-DTA mice didn’t sense noxious cold. In fact, “they don’t feel any temperature, really,” Costigan said. Wild-type mice given access to two thermal plates—one at ambient temperature and another a noxious cold plate—strongly prefer the ambient plate. In contrast, uninjured TRPV1-DTA mice split time randomly between the plates, indicating their lack of cold sensation.
The TRPV1-DTA mice also displayed tactile allodynia similar to wild-type mice at seven days after SNI, and this lasted at least 21 days. But cold allodynia was predictably absent in TRPV1-DTA mice at seven or 14 days after injury. The animals did develop very slight cold allodynia 21 days after SNI, which the authors speculate might rely on compensatory changes in remaining sensory neurons, perhaps involving TRPA1 receptors, which have been implicated in cold sensitivity after nerve injury.
Leukocyte collaboration
In order to perform the complementary experiment testing the role of immune cells in tactile allodynia, the researchers needed to remove the animals’ macrophages, a cell type found in the DRG after nerve injury and part of the innate immune system. “This system developed early in evolution—it’s the one we absolutely need,” Costigan said.
Mice can’t survive without macrophages altogether, so the group used a trick to temporarily kill off the immune cells. Immediately before nerve injury, mice received an intravenous injection of clodronate, a toxic drug that was encapsulated in tiny liposome packages. “In the blood, the macrophage comes along and eats the liposomes—because that’s what they do—and that releases the drug, killing macrophages specifically,” Costigan said.
This cell population recovers over the next two weeks, but that window allowed the researchers to test the animals while macrophages were absent. “We did the nerve injury then, and the tactile allodynia was really knocked down almost completely” at seven and 10 days, compared to control mice treated with empty liposomes. Strikingly, as the macrophages recovered, the animals developed tactile allodynia like the control mice. “That really suggests that macrophages are very important,” Costigan said.
Finally, the researchers examined the role of T and B immune cells, which they also observed in the DRG after nerve injury. Mice lacking the Rag1 gene lack T and B cells, and these animals developed cold allodynia similar to wild-type animals but displayed significantly reduced tactile allodynia. The tactile allodynia phenotype was rescued by introducing CD4+/CD8+ T cells, suggesting that B cells are not required. Cold allodynia remained intact similar to control mice with the loss of macrophages or T and B cells, suggesting that process develops independent of immune cells.
“It’s a nice use of the models,” Beggs said of the clodronate depletion and the use of the Rag1 knockout mice. “It is interesting that these two types of leukocytes were required, so there’s something about macrophages and T cells perhaps working together,” he said.
“There is a question still hovering,” said Beggs: “Is there a sex difference?” He said it was a missed opportunity of the study, which comprised only male mice. “To go back and collect these data in females would be quite daunting, but it leaves a bit of a hole where we’d be interested to know what’s going on in females peripherally.” Beggs and other groups have observed sex differences in immune cell activity in the central nervous system in neuropathic pain models. A recent publication from another group suggests that peripheral immune reactions in the DRG are sexually dimorphic (Lopes et al., 2017).
Costigan said the new findings “show that two of the main neuropathic pain symptoms that patients suffer from the most have different mechanisms.” All the pain symptoms presumably require sensory neuron inputs to the spinal cord, “but they use different neurons. The tactile allodynia needs the component of the immune system likely acting on large mechanosensitive A-beta fibers, whereas cold allodynia can work on its own in nociceptive neurons.” That finding represents a step forward in the fundamental understanding of neuropathic pain, and also in the potential to identify target molecules for new pain drugs.
Stephani Sutherland, PhD, is a neuroscientist, yogi, and freelance writer in Southern California.
Image credit: Cobos et al., 2018
Article Contributors:
