Parkin Inhibits Static Mechanical Pain by Suppressing Membrane Trafficking of Mechano-transducing Ion Channel TACAN

Dear Editor,

Clinical practice shows that harmless mechanical stimulation can cause pain in humans after inflammation or nerve injury, leading to a sensation called mechanical allodynia [1]. Mechanical allodynia comes in many forms: dynamic allodynia caused by something as light as a brushstroke, static allodynia caused by stimulation by weak von Frey filaments [2], which seriously disturbs the quality of life and working ability of patients. Parkin, an E3 ubiquitin ligase, is linked to neurodegenerative diseases such as Parkinson’s disease (PD) and Alzheimer’s disease (AD) [3, 4]. It has been reported that the neuropathic pain induced by diabetes reduces the expression of Parkin in sensory neurons [5]. The role of Parkin down-regulation in pathological pain and the underlying molecular mechanism remain unclear.

TACAN (TMEM120A) was screened and identified from a transmembrane protein that regulates the stretching activation current of vascular smooth muscle cells [6], and is an ion channel involved in mechano-sensing [7]. TACAN is expressed in cytomembrane of the sensory neurons and participates in the regulation of mechanical pain sensation [7]. However, the mechanism of TACAN membrane trafficking in sensory neurons has not yet been clarified.

Parkin is an important regulator of ion channels, which is involved in regulating the structure and function of various ion channels. In the rat brain, Parkin, as an interacting partner in the Cav2.2 channel complex, reduces the expression and current density of Cav2.2 [8]. Loss of Parkin in hippocampal neurons significantly increases acid-sensing ion channel currents [9]. Parkin regulates epidermal growth factor receptor trafficking and is responsible for the occurrence and progression of D [10]. However, it is still unclear whether Parkin interacts with the TACAN ion channel to regulate the membrane trafficking of TACAN.

First, we assessed the expression of genes in dorsal root ganglia (DRG) using RNA-seq analysis after complete Freund's adjuvant treatment (CFA, an inflammatory pain model). We identified 49 genes down-regulated more than two-fold in the DRGs of mice after CFA treatment (Fig. S1). Among these reduced genes, Prkn (the gene for Parkin) attracted our attention, because previous study had shown that Parkin is down-regulated in sensory neurons under the state of pathological pain induced by diabetes [5]. However, whether Parkin down-regulation is involved in pathologic pain and the possible molecular mechanism remains unclear. Quantitative polymerase chain reaction (qPCR) analysis was further applied to assess changes in the Prkn gene. The data showed that Parkin mRNA was significantly decreased in DRGs after CFA treatment (Fig. 1A). Immunoblotting showed that CFA also decreased the expression level of Parkin protein in the DRGs of mice (Fig. 1B). Double immunostaining showed that Parkin was co-localized with IB4, CGRP, and NF200-positive DRG neurons (Fig. 1C). Colocalization studies showed that 35.7%, 11.1%, and 40.6% of Parkin-positive cells expressed IB4, CGRP, and NF200, respectively, in the DRGs (Fig. S2A). Meanwhile, 92.1% of IB4-, 73.0% of CGRP- and 82.6% of NF200-positive neurons expressed Parkin in the DRGs (Fig. S2B). Together, these results demonstrate that Parkin is expressed in primary sensory neurons and its expression in DRGs is decreased under inflammatory pain.

Fig. 1.

Parkin selectively regulates static mechanical sensitivity. A The mRNA level of Parkin in DRGs of mice on day 5 after CFA treatment (n = 4, two mice mixed as a sample). B The immunoblotting showed the protein level of Parkin in DRG of mice on day 5 after CFA treatment (n = 3, two mice mixed as a sample). C Double immunostaining of Parkin with IB4, CGRP and NF200 in DRG sections of mice. Scale bar, 200 µm. D Behaviors of mice injected with Parkin-siRNA and nontargeting (NT)-siRNA evaluated by the von Frey test (n = 6). E Behaviors of Parkin-overexpressing and control mice evaluated by the von Frey test (n = 5). F–H Effects of overexpression of Parkin on CFA-, SNL- and VCR-induced static mechanical allodynia at different time points (n = 5). *P < 0.05, **P < 0.01, ***P < 0.001. n.s. means not significant. A, B, D, E, independent Student’s t test. F–H, two-way ANOVA followed by Bonferroni’s multiple comparisons test. Error bars indicate SEM.

To determine the role of Parkin in pain regulation, we knocked down its expression in DRGs by intrathecal injection (i.t) of siRNA (Fig. S2C-E). We firstly compared the mechanosensitivity of nontargeting (NT)-siRNA to that of Parkin-siRNA mice. Compared with NT-siRNA mice, Parkin-siRNA mice had a lower withdrawal threshold in von Frey filament-induced static mechanical stimulation (Fig. 1D), but they showed no change in brush-induced dynamic stimulation (Fig. S2F). Parkin-siRNA had no significant effect on the sensitivity to intense mechanical stimulation in pinprick tests (Fig. S2G). Furthermore, the sensitivity of Parkin-siRNA mice to noxious heat stimulation in Hargreaves test (Fig. S2H) and in tail-flick test (Fig. S2I) were not altered. No difference was found in responses to the evaporative cooling test between the NT-siRNA and Parkin-siRNA mice (Supplementary Fig. 2J). To further confirm the role of Parkin in static mechanical sensitivity, we intrathecally injected adeno-associated virus encoding Prkn (rAAV-Syn-Prkn-2A-EGFP) in mice. Twenty-one days after virus injection, marked green fluorescence in DRG neurons suggested a high efficiency of transfection (Fig. S3A). The immunostaining and immunoblotting results indicated that the expression of Parkin in the DRGs were significantly increased on day 21 after AAV injection (Fig. S3B–D). As expected, overexpression of Parkin (day 21 after AAV injection) increased the withdrawal threshold to von Frey filament-evoked static stimulation (Fig. 1E), but did not alter their sensitivity to stimuli induced by dynamic, pinprick, Hargreaves, tail-flick or evaporative cooling test (Fig. S3E–I). Then we examined the regulation of Parkin in pathological pain models. Behavioral data showed that overexpression of Parkin (day 21 after AAV injection) relieved the static mechanical allodynia induced by CFA, spinal nerve ligation (SNL, a neuropathic pain model) and vincristine (VCR, a chemotherapeutic pain model) (Fig. 1F–H). Compared with control mice, mice overexpressing Parkin (day 21 after AAV injection) showed normal dynamic allodynia and thermal hyperalgesia induced by CFA, SNL and VCR (Fig. S4A–F). Thus, these data demonstrate that Parkin selectively regulates static mechanical sensitivity.

TACAN (TMEM120A) and Piezo2 are mechano-transducing ion channels, which are highly expressed in sensory neurons and have recently been proved to be helpful in detecting mechanical stimulation [7, 11]. To determine the potential mechanism by which Parkin regulates static mechanical sensitivity, we investigated the interaction of Parkin with TACAN and Piezo2 channels in DRGs using co-immunoprecipitation. The results showed that Parkin potentially interacted with TACAN (Fig. 2A), but not with Piezo2 (Fig. 2B) in mouse DRG protein extracts. Furthermore, we captured high-resolution images by structured illumination microscopy (SIM) and found that 11.5% of Parkin was colocalized with TACAN and that 10.6% of TACAN was colocalized with Parkin in DRG neurons of mice (Fig. 2C and Fig. S5A). SIM images also showed co-localization of Parkin and TACAN in human sciatic nerve (Fig. 2D). The specificity of the TACAN antibody was validated by a blocking peptide (Fig. S5B). To determine the cellular type specificity of Parkin/TACAN interactions, we performed triple immunostaining experiments and found that Parkin and TACAN were co-expressed in IB4-, CGRP- and NF200-positive DRG neurons, but not in macrophages (CD68) (Fig. S5C). This suggests that the cell type of Parkin/TACAN interactions are IB4-, CGRP- and NF200-positive DRG neurons, not macrophages. In addition, we tested the effect of Parkin knockdown or overexpression on membrane, cytoplasmic, and total expression of TACAN in DRGs. Compared with NT-siRNA, Parkin-siRNA increased the membrane abundance of TACAN in DRG neurons (Fig. 2E). Parkin knockdown decreased the accumulation of TACAN in the cytoplasmic of DRG neurons (Fig. 2F). No difference in total expression of TACAN in DRG neurons was found between NT-siRNA and Parkin-siRNA mice (Fig. 2G). Notably, the proportion of TACAN membrane trafficking significantly increased in the DRG neurons after Parkin-siRNA treatment (Fig. 2H). Consistently, overexpression of Parkin significantly decreased the membrane level (Fig. 2I), increased the cytoplasmic level (Fig. 2J), and did not alter the total (Fig. 2K) abundance of TACAN in sensory neurons. The proportion of TACAN membrane trafficking in sensory neurons markedly decreased in mice overexpressing Parkin (Fig. 2L). Thus, these results suggest that Parkin interacts with TACAN and inhibits membrane trafficking of TACAN in sensory neurons. Further, we used TACAN-siRNA to determine whether Parkin modulated static mechanical pain by TACAN. The results showed that knockdown of TACAN significantly alleviated static mechanical allodynia induced by Parkin-siRNA (Fig. S5D). This suggests that TACAN may involve in Parkin regulating the static mechanical pain.

Fig. 2.

Parkin interacts with TACAN and regulates the membrane trafficking of TACAN in sensory neurons. A, B Co-immunoprecipitation showing the interaction between Parkin and TACAN, or Piezo2 in DRGs. DRG lysate was immunoprecipitated with Parkin antibody and immunoblotted with Parkin, TACAN, and Piezo2 antibody as indicated. This experiment was repeated three times. C, D High-resolution images by structured illumination microscopy showing co-localization between Parkin and TACAN in mouse DRG neurons (C) and human sciatic nerve (D). Scale bars, 5 μm. E–G The membranous (E), cytoplasmic (F), and total (G) TACAN expression in DRG neurons of Parkin-siRNA and nontargeting (NT)-siRNA mice (n = 3, two mice mixed as a sample). H The TACAN membrane trafficking ratio in DRG neurons of Parkin-siRNA and nontargeting (NT)-siRNA mice (n = 3, two mice mixed as a sample). I–K The membranous (I), cytoplasmic (J), and total (K) TACAN expression in DRG neurons of Parkin-overexpressing and control mice (n = 3, two mice mixed as a sample). L The TACAN membrane trafficking ratio in DRG neurons of Parkin-overexpressing and control mice (n = 3, two mice mixed as a sample). Independent Student’s t test. *P < 0.05, **P < 0.01. n.s. means not significant. Error bars indicate SEM.

Here, our data showed that Parkin is expressed in primary sensory neurons and its expression in DRGs is decreased after inflammatory pain. This is similar to previous findings that neuropathic pain caused by diabetes reduces Parkin expression in sensory neurons [5]. Studies have shown that inhibiting the ATF4 (activating transcription factor 4) signaling pathway significantly reduces Parkin expression in the nervous system [12]. Our previous work has revealed that ATF4 is also involved in the regulation of pathological pain [13]. However, the cellular and molecular mechanisms (role of ATF4) of Parkin down-regulation in sensory neurons induced by pathological pain remain unclear, and further research is needed. Mechanical allodynia is a kind of sensitized mechanical pain state, which is mainly induced by peripheral sensory neuron sensitization. Mechanical allodynia mainly contains two forms, including static allodynia and dynamic allodynia [2]. Parkin, an E3 ubiquitin ligase, is highly expressed in the nervous system and participates in the regulation of various functions, such as PD or AD [3, 4]. Our results showed that overexpression of Parkin alleviates static mechanical allodynia induced by inflammation, nerve injury or chemotherapeutic drugs. Clinical research has shown that Parkin mutation carriers have a high frequency of loss of mechanical or vibratory detection on the hands and feet [14]. Therefore, the regulation of endogenous Parkin expression may be an important way to treat AD, PD and mechanical allodynia. Innocuous mechanical stimuli are detected by a subpopulation of sensory neurons called low-threshold mechanoreceptors (NF200-positive neurons) and noxious mechanical pain are detected by a subpopulation of sensory neurons called nociceptors (IB4-positive neurons, CGRP-positive neurons). In humans, inflammation or nerve injury can cause abnormal sensory processing, leading to a painful response to innocuous mechanical stimuli, called mechanical allodynia [1]. Static mechanical allodynia, induced by normally innocuous low-threshold static mechanical stimulation, represents a cardinal feature of neuropathic pain. Our data showed that Parkin is involved not only in pain responses to noxious static mechanical stimuli but also in allodynia induced by innocuous static mechanical stimuli. At the same time, Parkin is also widely expressed in low-threshold mechanoreceptors (NF200-positive neurons) and nociceptors (IB4-positive neurons, CGRP-positive neurons) of DRG neurons. This may suggest that Parkin expressed in nociceptors is involved in the regulation of static mechanical pain, while Parkin expressed in low-threshold mechanoreceptors is involved in the regulation of static mechanical allodynia. This needs to be further studied. Parkin, an important regulator of ion channels, is involved in the regulation of various ion channel functions. For example, loss of Parkin in hippocampal neurons significantly increases the activity of acid-sensing ion channel [9]. Our data showed that Parkin interacts with TACAN channels in sensory neurons. According to the present study, the TACAN channel only senses static mechanical stimulation [7]. This may be the theoretical basis by which Parkin selectively regulates static mechanical sensitivity. However, it is not clear whether Parkin binds with TACAN directly or indirectly, as well as the specific binding domain. And the effects of Parkin on TACAN mechanical current density also needs further research in the future. Our data showed that knockdown of TACAN significantly alleviated static mechanical allodynia induced by Parkin-siRNA. Considering that knockdown of TACAN did not completely restore the mechanical pain threshold of Parkin-siRNA mice to the normal level, other targets except for TACAN may be involved. Further research is needed. Previous research has shown that Parkin modulating EGFR trafficking is responsible for the occurrence and progression of PD [10]. Our results showed that the overexpression of Parkin decreased the membrane expression of TACAN, and increased its cytoplasmic expression without affecting its total expression in sensory neurons. In other words, Parkin may regulate the negative trafficking of TACAN in sensory neurons. Cytoplasmic Dynein have been shown to be involved in negative trafficking of ion channels and receptors in cells [15]. Whether cytoplasmic Dynein contributes to the negative trafficking of TACAN by Parkin in sensory neuron remains to be further studied. In conclusion, our study indicated that Parkin interacts with TACAN and decreases its membrane trafficking in sensory neurons, which contributes to the reduction of the static mechanical sensitivity. Therefore, our research provides potential therapeutic targets and strategies for patients with static mechanical hypersensitivity.

Supplementary Information

Below is the link to the electronic supplementary material.

Conflict of interest

All authors claim that there are no conflicts of interest.