Among the hallmarks of neuropathic pain is the phenomenon of mechanical allodynia, in which normally non-painful stimuli—such as clothing rubbing against the skin, or even a person simply walking—produce pain. New research led by Ru-Rong Ji, Duke University Medical Center, Durham, US, now presents the immune system receptor Toll-like receptor 5 (TLR5) as a new marker to distinguish among A-fiber subclasses. The group also shows that TLR5 activation provides a portal to allow sodium channel-blocking drugs into sensory neurons, which might allow for a “guided missile” approach to stop mechanical allodynia. Moreover, the work confirms that mechanical allodynia is mediated by large, myelinated A-beta fibers in mouse models of neuropathic pain.
“What’s really exciting is that, for the first time, they have identified a selective marker for A-fibers—through which you can actually inhibit or silence neurons,” said Cheryl Stucky, a pain researcher at the Medical College of Wisconsin, Milwaukee, US, who was not involved in the work. “We really have had no way to distinguish cell bodies of A-beta fibers from A-deltas—no way at all. And now, TLR5 seems to be in a subset of A-fibers that are A-beta and not A-delta or A-alpha. That is an important advance that should provide a valuable tool for the field.”
The work was published online October 19 in Nature Medicine.
An A-beta fiber-specific immune receptor?
Co-first authors Zhen-Zhong Xu, Yong Ho Kim, Sangsu Bang, and Yi Zhang began by investigating dorsal root ganglia (DRG) neuronal expression of Toll-like receptors (TLRs), which are widely expressed in immune and glial cells. “We were interested in TLR expression in neurons following studies that have documented roles for TLR3 and TLR7 in neurons for pain and itch sensation,” Ji told PRF (Liu et al., 2012). Quantitative polymerase chain reaction (PCR) showed that most of the dozen known TLRs were expressed in DRG neurons, but in situ hybridization and immunohistochemistry revealed that TLR5 was mostly confined to medium- to large-diameter neurons that also expressed a marker for myelinated A-fibers. Pain-sensing neurons are classically described as small, unmyelinated C-fibers, but larger, myelinated mechanosensitive neurons are increasingly appreciated for their role in chronic pain. A small proportion of the DRG neurons containing TLR5—about 7 percent—appeared to be peptidergic C-fibers, and some A-deltas were also labeled. “We would not say that TLR5 is found exclusively in A-beta fibers, but it is primarily in those neurons,” Ji said.
At peripheral nerve endings, TLR5 staining was concentrated at Meissner corpuscles—structural end-organs in glabrous skin important for tactile function. Some TLR5-positive neurons were also seen at the base of hair follicles, but not at Merkel cells, another type of end-organ found in hairy and glabrous skin specialized to sense texture. “Because A-beta fibers are known to innervate both Merkel cells and Meissner corpuscles, I would have liked to see the TLR5-positive subpopulation defined in more detail,” Stucky said, which she noted could be achieved in future work with electrophysiological experiments using a skin-nerve preparation. In the spinal cord, TLR5 labeling was restricted to laminae III-V in the deep dorsal horn, where A-beta fibers terminate. Together, the results demonstrate that the TLR5 receptor is concentrated in a subpopulation of A-beta fibers, providing a marker that will aid researchers in investigating the functions of the broad class of A-fiber sensory neurons.
A new doorway for drugs
Previous work showed that activation of transient receptor potential vanilloid type 1 (TRPV1) in C-fibers provides an entry portal for QX-314, a membrane-impermeable sodium-channel blocker derived from lidocaine; co-application of the TRPV1 ligand capsaicin and QX-314 blocks C-fibers (Binshtok et al., 2007). In the current work, the researchers hypothesized that TLR5 activation might somehow allow for QX-314 entry, in part because associations have been shown between TLRs and ion channels.
Consequently, Ji and colleagues looked to the TLR5 ligand flagellin, found in bacterial flagella, which activates the receptor on immune cells to trigger an adaptive immune response; the role of TLRs in neurons is not yet understood, nor is it known with which proteins TLR5 might associate. Co-application of flagellin and QX-314 blocked evoked sodium currents and action potentials from dissociated mouse large-diameter A-fiber but not small C-fiber DRG neurons, indicating that TLR5 provided entry for the sodium channel blocker. Neither flagellin nor QX-314 alone affected sodium currents or action potentials, nor were these blocked in transgenic Tlr5-deficient mice.
To further confirm that the drugs targeted A-beta fibers selectively, the authors made electrophysiological recordings from the sciatic nerve while electrically stimulating the hindpaw of anesthetized mice. A-beta conduction was significantly slowed by co-application of flagellin and QX-314, but A-alpha, A-delta, and C-fiber conduction were unaffected. The results suggest that TLR5 activation opens the pore of some as-yet unidentified protein to allow QX-314 entry, Ji said.
The researchers also studied human DRG neurons isolated postmortem from four non-diseased men and women ranging in age from 29 to 67. Ji’s team procured the cells from the nonprofit National Disease Research Interchange (see PRF related news). Similar to the results using mouse neurons, co-application of flagellin and QX-314 blunted evoked sodium currents in large-diameter A-fiber DRG neurons but not small-diameter C-fibers.
Effects on pain
Would co-application of flagellin plus QX-314 have behavioral consequences in animals? “In three models of neuropathic pain, the drug combination to block A-beta fibers suppressed allodynia without affecting heat hyperalgesia,” Ji said. In the paclitaxel (PAX) model of chemotherapy-induced pain, systemic injection of PAX produces lasting mechanical allodynia. A week after PAX injection, the researchers co-applied flagellin and QX-314 to the hindpaws of mice. The treatment normalized withdrawal thresholds to mechanical stimulation for several hours in wild-type but not Tlr5-deficient mice, indicating that flagellin, acting via TLR5, and QX-314 had blocked allodynia. The treatment alleviated mechanical allodynia continuously when the researchers treated mice daily for five days, which produced no adverse effects in the mice. In the streptozotocin (STZ) model of diabetic neuropathy and the chronic constriction injury (CCI) model of nerve injury, treatment with flagellin and QX-314 also temporarily reversed long-lasting mechanical allodynia, whereas blockade of C-fibers with capsaicin plus QX-314 had little or no effect on mechanical allodynia but did reverse heat hyperalgesia.
In order to assess animals’ ongoing pain, the researchers employed the conditioned place preference (CPP) test. PAX-treated animals preferred the chamber associated with flagellin/QX-314 treatment, whereas blockade of C-fibers with capsaicin/QX-314 had no effect on place preference. The findings are in line with the idea that ongoing pain arises from spontaneous firing of A-fibers, and A-beta fibers in particular. “It’s really important to look at ongoing pain in addition to stimulus-evoked pain, because that’s primarily the problem that affects patients with chronic pain,” Stucky said. “A lot of what patients experience is ongoing or movement-evoked pain,” she added, which could also be measured in future studies with non-reflexive behaviors such as wheel running.
The new study provides a way to identify A-beta fibers and indicates that those fibers are responsible for mechanical allodynia—which previous evidence had suggested but could not confirm without a specific marker. Together with other recent work, the new report adds to researchers’ understanding of the neurons that are responsible for mechanical allodynia, including neurons in the spinal cord (see PRF related news). For example, Stucky’s group recently used calcitonin gene-related peptide (CGRP) to differentiate among subpopulations of large, myelinated sensory neurons (Weyer et al., 2015). CGRP is traditionally associated with small nociceptive C-fibers, but has recently been found in up to half of all large myelinated sensory neurons (McCoy et al., 2013). Stucky found that subpopulations of A-fiber neurons that innervate muscle and skin differentially upregulated mechanically sensitive ion channels in the complete Freund’s adjuvant (CFA) model of inflammatory pain, contributing to mechanical allodynia. “These papers clearly demonstrate there are molecular and electrophysiological changes in A-fiber neurons that can drive mechanical allodynia,” Ji told PRF.
But perhaps most importantly, the new discovery could provide a way to target and silence the cells that give rise to allodynia. “This combination therapy could work like a guided missile,” Ji said. “TLR5 gives us a way to open the door, to get another drug inside, whether it’s QX-314 or something else” that might be used to block neuronal activity, he said. Because TLR5 is present at peripheral nerve endings, he could envision a medication that acts in the periphery—such as a cream to penetrate skin, much like commercially available capsaicin creams. However, Ji said, “the concentration of flagellin needed to open TLR5 is much lower—1,000 times lower—than the concentration of capsaicin required to open TRPV1, so this may get QX-314 into neurons with less risk of side effects.”
Stephani Sutherland, PhD, is a neuroscientist, yogi, and freelance writer in Southern California.
Image and caption provided by Xu et al.