Pinning down a perpetrator in postsurgical eye pain

Neuropathic pain is a pathological condition arising from a diseased or damaged somatosensory nervous system. Therapeutics for this condition remain insufficiently effective and, a better understanding of the mechanisms underlying neuropathic pain is therefore greatly needed. Neuropathic pain can, in some cases, result from iatrogenic trauma. In two types of refractive eye surgery, laser-assisted in situ keratomileusis (LASIK) and photorefractive keratectomy (PRK), a laser is used to reshape the corneal surface and thereby readjust the focal point of incoming light relative to the retina (1). The cornea has an exceptionally high density of nociceptive neuron terminals arising from the trigeminal nerve (Fig. 1A), however, and the surgical reshaping process unavoidably damages some of these terminals. This damage is thought to be a contributor to the sensory dysfunction experienced by a subset of patients after refractive surgery. Multiple types of corneal discomfort have been reported in these patients. One ensemble of symptoms is referred to as dry eye and may include irritation, a sense of dryness, ocular fatigue, and red eye (2). Dry eye occurs in almost half of patients after refractive surgeries but typically resolves within 3 to 6 mo. More impactfully, a smaller subset of patients experience corneal neuralgia (3), characterized by corneal pain that can last for months after surgery, even without continued evidence of corneal tissue injury (4). These patients often show sensory symptoms out of proportion to objective clinical signs and do not respond well to dry eye treatments. While this view dichotomizes dry eye and corneal neuralgia, nerve terminal damage and local inflammation could contribute to each. In fact, dry eye is sometimes treated with autologous serum or recombinant neurotrophins, to promote corneal reinnervation (5). The heterogeneous incidence, nature, and persistence of sensory symptoms in patients after refractive surgery suggest the existence of genetic or other factors that govern predisposition to these conditions. Recent reports have identified rare ion channel variants in corneal neuralgia patients (6, 7). In this issue of PNAS, Gualdani et al. examined the biophysical properties of a rare transient receptor potential cation channel subfamily V member 1 (TRPV1) variant, V527M, found in a patient who developed corneal neuralgia after LASIK and subsequent PRK enhancement (8). Their findings suggest a potential pathophysiological model (Fig. 1B) for the pain experienced by this patient.

Fig. 1.

(A) Schematic diagram of the corneal nociception pathway. The dense network of corneal epithelial nerve terminals is shown at Top. In the example terminal expanded at Right, TRPV1 channels are activated by stimuli such as heat, bioactive lipids, protons, and capsaicin, resulting in neuronal depolarization and action potential generation. The signal propagates from the periphery to the trigeminal ganglion and then to the brainstem, subsequently triggering pain sensation. (B) Proposed mechanism of how TRPV1 variant V527M, in concert with other factors, could predispose one to enhanced corneal pain after refractive eye surgeries.

Using patch-clamp and calcium imaging experiments in human embryonic kidney (HEK) 293 cells transfected with TRPV1 V527M, Gualdani et al. showed that this variant channel displayed enhanced amplitudes of responses to acidic pH and heat, but not to capsaicin, as compared to wild-type (WT) TRPV1, indicative of a modality-specific gain-of-function. Protons both directly activate and potentiate TRPV1 (9). Gualdani et al. found that TRPV1 V527M exhibited a larger leftward shift in its voltage activation curve under acidic conditions than did WT TRPV1, providing further evidence for altered proton gating. They also showed that the V527M variant exhibited slower deactivation during continuous capsaicin stimulation and less attenuated response amplitude during consecutive capsaicin challenges, compared to WT, indicating defective activity-dependent desensitization. Heterogeneous impacts of perturbations on different TRPV1 stimulus mechanisms have precedents. Mutations in distinct TRPV1 domains can differentially suppress its activation by capsaicin, protons, or heat (10), and some small molecule TRPV1 antagonists suppress TRPV1 activity less efficaciously at low pH (11). Future modeling of the impact of the V to M mutation on TRPV1 structure and gating in the presence of different stimuli might reveal how its specific gain-of-function pattern arises and how it might be most effectively inhibited.

In PNAS, Gualdani et al. examined the biophysical properties of a rare transient receptor potential cation channel subfamily V member 1 (TRPV1) variant, V527M, found in a patient who developed corneal neuralgia after LASIK and subsequent PRK enhancement.

To determine whether TRPV1 V527M could alter corneal sensory neuron excitability, Gualdani et al. performed patch-clamp recordings on cultured rat primary trigeminal ganglion (TG) neurons transiently transfected with this variant. They showed that neurons overexpressing TRPV1 V527M exhibited a higher resting membrane potential, a lower current threshold to generate action potentials, more firing in response to a given current stimulus, and less capsaicin-induced desensitization, compared to neurons overexpressing WT TRPV1. They also found that neurons expressing TRPV1 V527M exhibited increased calcium rises in response to low pH, when compared with neurons expressing WT TRPV1, all findings consistent with those in HEK 293 cells.

Gualdani et al. also examined potential upstream modulators of TRPV1 V527M that might contribute to corneal neuralgia. Many proinflammatory mediators, such as bradykinin, are released locally after corneal injury (12). Using calcium imaging, Gualdani et al. found that TG neurons expressing TRPV1 V527M exhibited increased sensitization to capsaicin by bradykinin, compared to neurons expressing WT TRPV1. Twelve-hydroxyeicosatetraenoic acid (12-HETE), an endogenous TRPV1 activating lipid, is released after corneal injury and inflammation (13). Gualdani et al. observed that TG neurons expressing TRPV1 V527M also exhibited increased responsiveness to 12-HETE and increased 12-HETE potentiation of capsaicin responses, compared to neurons transfected with WT TRPV1. Based on mechanistic understanding of bradykinin-induced nociceptor sensitization, they speculated that 12-HETE produced downstream of bradykinin might have contributed to TRPV1 V527M-dependent TG nociceptor sensitization, and thus corneal neuralgia, in the patient bearing that variant. They also acknowledged, however, a potential role for bradykinin-induced TRPV1 phosphorylation by protein kinase C (PKC). Further experiments to pharmacologically block 12-HETE generation or PKC activity might help explicitly distinguish the relative contributions of these two mechanisms. Moreover, distinct PKC phosphorylation sites on TRPV1 differentially impact its sensitivity to different stimuli (14). The basal phosphorylation state of TRPV1 V527M might thus be distinct from that of WT TRPV1, with potential implications for channel activity.

One limitation of the study by Gualdani et al. is that the human TRPV1 variant was overexpressed in rat TG neurons, some but not all of which expressed endogenous WT rat TRPV1. Overexpression of ion channels in any cell can strain cell health, though this was controlled for by the parallel overexpression of WT TRPV1. Perhaps more importantly, it is likely that the exogenous human TRPV1 V527M heteromultimerized with endogenous rat WT TRPV1 subunits in those neurons already expressing the latter, but formed homomultimers in cells not doing so. Such heterogeneity could complicate the interpretation of results. The sensory neurons of the patient in whom the TRPV1 V527M variant was identified are likely to have uniformly coexpressed WT TRPV1 with this variant. The rat TG neurons assayed by Gualdani et al. were limited to small diameter neurons, partially constraining their heterogeneity. Yet, the cornea is innervated by heterogeneous Aδ and C free nerve endings, including peptidergic and nonpeptidergic polymodal nociceptors variably expressing TRPV1, TRPA1, and/or ASIC channels, cold receptors expressing TRPM8, and mechanoreceptors expressing PIEZO2 (15). Generation and study of gene edited animals in which the TRPV1 V527M mutation is knocked into the endogenous TRPV1 locus might allow more consistent allelic coexpression with endogenous WT TRPV1, in addition to eliminating confounds of transient transfection and overexpression. With the application of retrograde tracing and neuronal subtype labeling strategies, the identities and properties of the specific cornea-innervating neuronal subpopulation(s) expressing the TRPV1 variant could then be studied in isolation. Such an animal model would also permit direct in vivo electrophysiological and calcium imaging experiments on the relevant corneal nociceptors and corresponding pain behavior studies. A complementary approach, based on the work of the laboratory that published the Gualdani et al. study (16), would be to generate and study nociceptive sensory neurons derived from induced pluripotent stem cells from the patient with the TRPV1 V527M variant. This approach might also permit the identification of any modifying genes that contributed to that patient’s pain condition.

Another issue relevant to the study of Gualdani et al. is whether corneal neuralgia is a purely neuropathic pain state. Abnormal reinnervation of the cornea is often observed after refractive eye surgeries (2). However, these surgeries also damage the corneal epithelial cells and subepithelial stromal cells. Thus, it is possible that local tissue inflammation or ongoing secretion of pronociceptive factors without obvious inflammation also contributes to corneal pain after LASIK or PRK. Characterization and comparison of corneal pain in animal models produced by direct nerve injury or by, for instance, lacrimal gland resection, would potentially be useful for distinguishing these possibilities (17). TRP channel mutations have also been associated with trigeminal neuralgia (18). However, the presumed pathology in classical trigeminal neuralgia is at or central to the TG, which is distinct from distal terminal damage experienced by patients undergoing refractive surgeries. The extent to which the pathophysiological mechanisms triggered by nerve injury at these distant loci overlap remains unclear. Of note, TRPV1 activation can also result in degeneration of distal nociceptor terminals (19). It is therefore possible that hyperactivity of the TRPV1 V527M might itself contribute to corneal nerve terminal damage.

Finally, it is intriguing that only after the PRK enhancement but not after the initial LASIK surgery did the patient with the TRPV1 V527M variant experience corneal neuralgia. Were both injuries necessary for the advent of corneal neuralgia or would PRK alone have been sufficient? Did other factors allow the TRPV1 variant to manifest as corneal neuralgia in this patient? Insights into these questions might emerge from more extensive confocal microscopic analysis of corneal innervation in patients after LASIK and again following any PRK enhancement, and from correlation of those findings with corneal pain incidence and genotyping at TRPV1 and other nociception-relevant loci. Animal models of repetitive refractive surgeries might also prove useful to address this issue.