Action potential–independent spontaneous microdomain Ca²⁺ transients–mediated continuous neurotransmission regulates hyperalgesia

Significance

The neurotransmission between dorsal root ganglion (DRG) and spinal cord neurons is essential for pain sensation. Mounting evidence has established the crucial roles of action potential (AP)-evoked neurotransmission from DRG neurons in chronic pain. However, the function and mechanism of AP-independent neurotransmission are not fully understood. Here, we observed frequent spontaneous microdomain Ca²⁺ (smCa) activities across DRG neurons’ somata and axons independent of APs. These smCa activities, mediated by spontaneous activation of TRPA1, triggered continuous neurotransmission from DRG to spinal cord, which was substantially elevated in inflammatory pain conditions. Importantly, inhibiting smCa or the associated continuous neurotransmission alleviated hyperalgesia. Our study may have identified a unique mechanism underlying nociceptive sensitization in chronic pain.

Abstract

Neurotransmitters and neuromodulators can be released via either action potential (AP)–evoked transient or AP-independent continuous neurotransmission. The elevated AP-evoked neurotransmission in the primary sensory neurons plays crucial roles in hyperalgesia. However, whether and how the AP-independent continuous neurotransmission contributes to hyperalgesia remains largely unknown. Here, we show that primary sensory dorsal root ganglion (DRG) neurons exhibit frequent spontaneous microdomain Ca²⁺ (smCa) activities independent of APs across the cell bodies and axons, which are mediated by the spontaneous opening of TRPA1 channels and trigger continuous neurotransmission via the cyclic adenosine monophosphate-protein kinase A signaling pathway. More importantly, the frequency of smCa activity and its triggered continuous neurotransmission in DRG neurons increased dramatically in mice experiencing inflammatory pain, inhibition of which alleviates hyperalgesia. Collectively, this work revealed the AP-independent continuous neurotransmission triggered by smCa activities in DRG neurons, which may serve as a unique mechanism underlying the nociceptive sensitization in hyperalgesia and offer a potential target for the treatment of chronic pain.

Chronic pain is a multifaceted and persistent disease that affects more than 30% of people all over the world and costs billions of dollars every year (1, 2). Pain signals start from primary sensory dorsal root ganglion (DRG) neurons (3–5), which function as environmental detectors (heat, cold, mechanical force, etc.) and transmit sensory information from the periphery to the central nervous system. The communication between DRG neurons and central spinal cord neurons through neurotransmission is essential for pain processing. In addition to pain sensation in physiological conditions, DRG neurons also play a key role in the progression and maintenance of chronic pain (6, 7). However, the molecular and cellular mechanisms remain not fully understood.

A large population of neurotransmitters (glutamate, ATP, etc.) and neuromodulators [Substance P, CGRP, neuropeptide Y (NPY), etc.] are secreted from DRG neurons and involved in the development and maintenance of chronic pain (8, 9). The concentrations of these neurotransmitters and neuromodulators in DRG and spinal cord are greatly elevated in individuals experiencing chronic pain (10–15). Neurotransmitters and neuromodulators can be released via either action potential (AP)–evoked or AP-independent neurotransmission. Mounting evidence has established the crucial roles of AP-coupled neurotransmission from DRG neurons in chronic pain. However, whether and how AP-independent continuous neurotransmission contributes to chronic pain remains largely unknown.

In the current study, we observed frequent spontaneous microdomain Ca²⁺ (smCa) activities independent of APs across the cell bodies and axons of primary sensory DRG neurons using high-resolution real-time total internal reflection fluorescence (TIRF) imaging. The smCa activities trigger continuous neurotransmission in a protein kinase A (PKA)-dependent manner. Importantly, the smCa-induced continuous neurotransmission in DRG neurons increased dramatically in chronic pain, inhibition of which alleviates hyperalgesia. Collectively, this study defines the AP-independent continuous neurotransmission triggered by smCa activities as a possible unique mechanism underlying the nociceptive sensitization of primary sensory neurons, providing a potential therapeutic target for the treatment of chronic pain.

Results

smCa Activities in Sensory DRG Neurons.

The real-time TIRF imaging (16, 17) of a membrane-localized Ca²⁺-indicator, Lck-GCaMP5g (18), was used to simultaneously monitor local Ca²⁺ signals beneath the plasma membrane in both the soma and axons of individual primary sensory DRG neurons. We strikingly observed frequent spontaneous Ca²⁺ activities across the somatic and axonal regions of DRG neurons in normal Ca²⁺-containing bath solution (Fig. 1 A–E, SI Appendix, Fig. S1 A and B, and Movies S1–S4). These spontaneous Ca²⁺ activities were confined to a restrained area without spreading out. In addition, the number of active regions of interest (ROIs) and firing rate of the spontaneous Ca²⁺ signals in each DRG neuron were significantly decreased after incubation with N,N’-[1,2-ethanediylbis(oxy-2,1-phenylene)]bis[N-[2-[(acetyloxy)methoxy]-2-oxoethyl]-1,1’-bis[(acetyloxy)methyl] ester-glycine (BAPTA-AM), a fast and potent Ca²⁺-chelator (3, 19), but not affected by the slow Ca²⁺-chelator 3,12-bis[2-[(acetyloxy)methoxy]-2-oxoethyl]-6,9-dioxa-3,12-diazatetradecanedioic acid, 1,14-bis[(acetyloxy)methyl] ester (EGTA-AM) (Fig. 1 F–M), indicating that the spontaneous Ca²⁺ activities in DRG neurons are microdomain Ca²⁺ (smCa) signals. To determine the origin of the smCa signals, we removed the extracellular Ca²⁺ or depleted the intracellular Ca²⁺ store with 2-Aminoethyl diphenylborinate (2-APB). The smCa signals in the somata and axons of DRG neurons disappeared in Ca²⁺-free bath solution and then fully recovered after switching back to normal Ca²⁺-containing solution (Fig. 2 A–C and SI Appendix, Fig. S1 C–E), suggesting that the smCa signals originate from Ca²⁺-influx via channels on the plasma membrane of DRG neurons, instead of the intracellular Ca²⁺ store because 2-APB did not affect smCa signals (SI Appendix, Fig. S2 A and B).

Fig. 1.

smCa activities in DRG neurons. (A, Upper) A TIRF image of cultured DRG neuron transfected with Lck-GCaMP5g; (Lower) A cartoon illustrating that the Lck-GCaMP5g (black) binds with Ca²⁺ and emits green fluorescence. (Scale bar, 10 μm.) (B and C) The surface plot of the Ca²⁺ signal from a typical ROI (defined in Materials and Methods) in A at inactive and active states. (D) The fluorescence intensity of an active ROI in 1 min with the hot image shown below. The smCa events (defined in methods) are indicated by black bars at the Top. (E) The whole-cell surface plot of the Ca²⁺ signal from the DRG neuron in A. (F) The hot image of the Ca²⁺ signals before and after the treatment of EGTA-AM (10 μM) from 12 active ROIs in the DRG neuron shown in A. (G) Quantification for the total active ROIs per cell per minute before and after the treatment of EGTA-AM. (H) The smCa events from 11 cells before and after the treatment of EGTA-AM. Each row shows the total smCa events in 1 min from one DRG neuron. (I) Quantification for the total smCa events per cell per minute before and after the treatment of EGTA-AM. (J–M) The same as F–I, but replace EGTA-AM with BAPTA-AM (10 μM). Data were shown as a violin plot with all data points included. Paired student’s t test. ns, not significant; *P < 0.05; **P < 0.01.

Fig. 2.

SmCa is dependent on extracellular Ca²⁺ and TRPA1 channels. (A, Left) A TIRF image of cultured DRG neuron transfected with Lck-GCaMP5g; (Right) The whole-cell surface plot of the Ca²⁺ signal from the DRG neuron in the Left panel of A. (Scale bar, 10 μm.) (B) The hot image of the Ca²⁺ signals from 12 active smCa ROIs in the DRG neuron shown in A, in 2 mM, 0 mM, and 2 mM Ca²⁺-containing baths, respectively. (C) Quantification for the total active ROIs per cell per minute in 2 mM, 0 mM, and 2 mM Ca²⁺-containing baths, respectively. (D) Quantification for the total active ROIs per cell per minute before and after whole-cell voltage clamping to −70 mV. (E, Left) Quantification for the total active ROIs per cell per minute before and after the treatment of cocktail of inhibitors [1 μM TTX (antagonist of voltage-gated sodium channel), 1 μM GVIA (antagonist of voltage-gated N-type Ca²⁺ channel), 5 μM Nifedipine (antagonist of voltage-gated L-type Ca²⁺ channel), and 200 μM CdCl₂ (the pore blocker of all VGCCs)]. (Right) Quantification for the total active ROIs per cell per min before and after the treatment of 1 μM TTA-A2 (antagonist of voltage-gated T-type Ca²⁺ channel). (F) Quantification for the total active ROIs per cell per minute before and after the treatment of broad-spectrum antagonist of TRP channels (50 μM RR, Ruthenium Red). (G) Quantification for the total active ROIs per cell per minute before and after the treatment of TRPA1-specific antagonist, 1 μM A967079 (Left) or 20 μM HC030031 (Right), respectively. (H) Quantification for the total active ROIs per cell per minute before and after the treatment of TRPV1-specific antagonist, 10 μM AMG9810. (I) Quantification showing the total active ROIs per cell per minute (Left) and total smCa events per cell per minute (Right) in WT, TRPA1^−/− and TRPV1^−/− DRG neurons, respectively. (J) Quantification showing the total active ROIs per cell per minute (Left) and total smCa events per cell per minute (Right) in control and TRPA1 overexpression DRG neurons, respectively. (K) A cartoon illustrating that smCa activity is mediated by the Ca²⁺ influx through TRPA1 channel. Data were shown as a violin plot with all data points included. Paired student’s t test for D–H. Student’s t test for J. One-way ANOVA with a post hoc test enabling multiple comparisons for C and I. ns, not significant; **P < 0.01; ***P < 0.001.

TRPA1 Mediates smCa Activities in DRG Neurons.

To determine the Ca²⁺ channel responsible for smCa signals, we first clamped the membrane potential of DRG neurons at −70 mV to block the activity of voltage-gated Ca²⁺ channels (VGCCs) and found that it did not affect the smCa signals (Fig. 2D and SI Appendix, Fig. S2 C and D). In addition, the cocktail of inhibitors (including 1 μM TTX for voltage-gated Na⁺ channel, 1 μM ω-conotoxin GVIA for voltage-gated N-type Ca²⁺ channel, 5 μM nifedipine for voltage-gated L-type Ca²⁺ channel and 200 μM CdCl₂ for all VGCCs) and 1 μM TTA-A2 (antagonist of voltage-gated T-type Ca²⁺ channel) (3, 19) showed no effect on the smCa as well (Fig. 2E and SI Appendix, Fig. S2 E–H), excluding the contribution of VGCCs and supporting the AP-independency of these smCa signals. Next, we tried ruthenium red (RR, 50 μM), a nonspecific broad-spectrum antagonist for all transient receptor potential (TRP) channels. Surprisingly, RR almost eliminated the smCa signals (Fig. 2F and SI Appendix, Fig. S2 I and J), suggesting the involvement of TRP channels in smCa activities. Furthermore, two specific inhibitors for TRPA1 (A967079, 1 μM or HC030031, 20 μM) blocked the smCa signals significantly (Fig. 2G and SI Appendix, Fig. S3), but the antagonist of TRPV1 (AMG9810, 10 μM) showed no effect (Fig. 2H), suggesting an essential role of TRPA1 in smCa signals. In addition, the spontaneous inward membrane currents of DRG neurons at resting membrane potential were also inhibited by TRPA1 antagonist (SI Appendix, Fig. S1 F and G). Consistently, the number of active regions and firing rate of smCa signals in TRPA1^−/−, but not TRPV1^−/−, DRG neurons were reduced significantly compared with that of wild-type (WT) DRG neurons (Fig. 2I and SI Appendix, Fig. S4). Contrarily, overexpression of TRPA1 in DRG neurons increased the number of active regions and firing rate of smCa signals dramatically (Fig. 2J and SI Appendix, Fig. S5). Taken together, these findings demonstrate that the smCa signals in DRG neurons are mediated by the spontaneous activation of TRPA1 channels and independent of APs (Fig. 2K).

SmCa Triggers Continuous Neurotransmission in Both the Somata and Terminals of DRG Neurons.

As an important second messenger, Ca²⁺ has diverse functions in different cellular procedures, especially its crucial roles in neurotransmission (20, 21). To investigate the function of smCa activities in the neurotransmission of DRG neurons, we labeled the large dense-core vesicles and small vesicles of DRG neurons with NPY-pHluorin and synaptophysin (Spy)-pHluorin, respectively (3, 18, 22) (Fig. 3 A and F). Consistent with the smCa signals, we observed abundant spontaneous release events of both NPY- and Spy-pHluorin labeled vesicles from DRG neurons in Ca²⁺-containing bath solution by using high-resolution TIRF live-imaging, indicated by abrupt fluorescence increase followed by the decrease to baseline (Fig. 3 B and G and Movies S5 and S6), which were diminished when switching to Ca²⁺-free bath solution (Fig. 3 C and H). In addition, the spontaneous release of DRG neurons was largely inhibited by BAPTA-AM, instead of EGTA-AM (SI Appendix, Fig. S6 A–D). In contrast, clamping the membrane potential at −70 mV did not affect the spontaneous release, indicating that smCa, but not spontaneous AP, is the trigger for the spontaneous vesicle release in DRG neurons (SI Appendix, Fig. S6E). Furthermore, the nonspecific TRP channel blocker (RR) and specific antagonist of TRPA1 (A967079), shown to block smCa in DRG neurons (Fig. 2 F and G), largely inhibited the spontaneous release of both NPY- and Spy-pHluorin labeled vesicles in DRG neurons (Fig. 3 D and I). Moreover, the spontaneous release events were significantly fewer in TRPA1^−/− DRG neurons compared with that in WT neurons (Fig. 3E), further confirming the essential role of TRPA1-mediated smCa in the spontaneous vesicle release of DRG neurons (Fig. 3J).

Fig. 3.

SmCa regulates spontaneous quantal vesicle release in the somata of DRG neurons. (A, Upper) the TIRF image of a 24 h cultured DRG neuron transfected with NPY-pHluorin; (Lower) A cartoon illustrating that NPY-pHluorin emits green fluorescence when vesicle lumen’s pH increases from 5.5 to 7.4 during fusion. (Scale bar, 10 μm.) (B) The fluorescence intensity versus time and the TIRF image montage showing a single quantal vesicle release process from three events marked in A, which are aligned by the onset point of each release. Each release event happens spontaneously and randomly at different time in 1 min recording. (C) Quantification for the total spontaneous quantal release events per cell per minute in 2 mM Ca²⁺ and 0 mM Ca²⁺-containing baths, respectively. (D) Quantification for the total spontaneous quantal release events per cell per minute, before and after the treatment of RR (Left) or A967079 (Right). (E) Quantification for the total spontaneous quantal release events per cell per minute in WT and TRPA1^−/− DRG neurons. (F, Upper) the TIRF image of a 24 h cultured DRG neuron transfected with Spy-pHluorin; (Lower) A cartoon illustrating that Spy-pHluorin emits green fluorescence when vesicle lumen’s pH increases from 5.5 to 7.4 during fusion. (Scale bar, 10 μm.) (G) The fluorescence intensity versus time and the TIRF image montage showing a single quantal vesicle release process from three events marked in F. (H) Quantification for the total spontaneous quantal release events per cell per minute in 2 mM Ca²⁺ and 0 mM Ca²⁺-containing baths, respectively. (I) Quantification for the total spontaneous quantal release events per cell per minute, before and after the treatment of RR (Left) or A967079 (Right). (J) A cartoon illustrating that spontaneous Ca²⁺ influx through the TRPA1 channel (smCa) regulates spontaneous quantal vesicle release. Data were shown as a violin plot with all points included. Paired student’s t test for C/D/H/I. Student’s t test for E. *P < 0.05; **P < 0.01; ***P < 0.001.

Since smCa activities exist not only in the somata but also in the axons of DRG neurons (SI Appendix, Fig. S1 A–E), we set out to study the role of smCa activities in synaptic transmission of DRG neurons cocultured with spinal dorsal horn neurons (Fig. 4 A and B). Consistent with our TIRF imaging data, the frequency of miniature excitatory postsynaptic current (mEPSC) recorded from the dorsal horn neurons was reduced remarkably after switching from the normal Ca²⁺-containing solution to a Ca²⁺-free solution (Fig. 4 C and D). Besides, the mEPSC events were also largely inhibited by the nonspecific TRP channel blocker (RR) and the specific TRPA1 channel antagonist (A967079) (Fig. 4 E–H). Furthermore, the frequency of mEPSC events from WT dorsal horn neurons cocultured with TRPA1^−/− DRG neurons was significantly lower than that cocultured with WT DRG neurons (Fig. 4 I and J). In addition, we recorded the mEPSCs of dorsal horn neurons from adult fresh spinal cord slices with intact DRG central terminals and found that the mEPSC events were also inhibited by TRPA1 antagonist (SI Appendix, Fig. S6 F−H). Taken together, smCa triggers continuous neurotransmission in both the somata and terminals of DRG neurons.

Fig. 4.

SmCa regulates continuous miniature synaptic transmission from DRG to spinal dorsal horn neurons. (A) A light-field image showing patch-clamp recording on postsynaptic spinal dorsal horn (DH) neurons cocultured with DRG neurons. (Scale bar, 10 μm.) (B) A cartoon illustrates the recording for mEPSC. (C) The typical mEPSC traces from cocultured DRG-dorsal horn synapses in 2 mM Ca²⁺ and 0 mM Ca²⁺-containing baths, respectively. (D) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right) in 2 mM Ca²⁺ and 0 mM Ca²⁺-containing baths shown in C. (E) The typical mEPSC traces from cocultured DRG-dorsal horn synapses before and after RR treatment. (F) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right), before and after RR treatment shown in E. (G) The typical mEPSC traces from cocultured DRG-dorsal horn synapses before and after A967079 treatment. (H) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right), before and after A967079 treatment shown in G. (I) The typical mEPSC traces from WT spinal dorsal horn neurons cocultured with WT and TRPA1^−/− DRG neurons. (J) Quantification for the mEPSC event number per minute (Left) and the amplitude (pA, Right) shown in I. Data were shown as a violin plot with all points included. Paired student’s t test for D/F/H. Student’s t test for J. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001.

PKA Plays an Essential Role in smCa-Triggered Continuous Neurotransmission.

PKA and protein kinase C (PKC) are important kinases known to be activated by Ca²⁺ and function in spontaneous synaptic transmission (23, 24). To identify the key player mediating smCa-triggered continuous neurotransmission in DRG neurons, we tested the specific inhibitors of PKA (H89, 2 μM) and PKC (Bis, 1 μM) and found that blocking PKA showed no effect on the smCa activities (SI Appendix, Fig. S7 A and B), while blocking PKC reduced smCa activities in DRG neurons (SI Appendix, Fig. S7 C and D), implying that PKA may be downstream of smCa activities. Intriguingly, the spontaneous release of NPY-pHluorin labeled vesicles was reduced significantly by H89, the PKA antagonist (Fig. 5 A–C), suggesting the functional involvement of PKA in smCa-triggered spontaneous neurotransmission in DRG neurons. Consistently, H89 also blocked the mEPSC events recorded from dorsal horn neurons cocultured with DRG neurons (Fig. 5 D and E). To further confirm that PKA is the downstream protein kinase of smCa activities, we first inhibited the smCa signals by using the TRPA1 antagonist (A967079) and then applied H89 to inhibit PKA activity. The frequency of mEPSC events recorded from spinal dorsal horn neurons cocultured with DRG was reduced by A967079 as shown before (Fig. 4 G and H), but did not further decrease with the addition of H89 (Fig. 5 F and G). Similarly, A967079 did not further lower the mEPSC event frequency after H89 application (Fig. 5 H and I), suggesting that smCa and PKA are in the same signaling pathway. Furthermore, after inhibiting the smCa-triggered mEPSC with A967079, the PKA agonist (25 μM sp-cAMP) rescued the frequency of mEPSC events recorded from spinal dorsal horn neurons cocultured with DRG (Fig. 5 J and K), indicating that PKA is downstream of smCa signals. Altogether, PKA plays an essential role in smCa-triggered continuous neurotransmission in DRG neurons.

Fig. 5.

SmCa regulates spontaneous vesicle release via PKA. (A) A TIRF image of a 24 h cultured DRG neuron transfected with NPY-pHluorin. (Scale bar, 10 μm.) (B) A cartoon illustrating that NPY-pHluorin emits green fluorescence when the vesicle lumen’s pH increases from 5.5 to 7.4 during fusion. (C) Quantification for total spontaneous quantal release events per cell per minute by TIRF imaging, before and after treatment of the PKA antagonist (2 μM H89). (D) The typical mEPSC traces from cocultured DRG and dorsal horn neurons before and after H89 treatment. (E) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right) before and after H89 treatment. (F) Typical mEPSC traces recorded from cocultured DRG and dorsal horn neurons, first in normal Ca²⁺-containing bath solution (Top), then TRPA1 antagonist (A967079) was applied (Middle), and last PKA inhibitor (H89) was added on top of TRPA1 antagonist (Bottom). (G) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right) in F. (H) Typical mEPSC traces recorded from cocultured DRG and dorsal horn neurons, first in normal Ca²⁺-containing bath solution (Top), then PKA inhibitor (H89) was applied (Middle), and last TRPA1 antagonist (A967079) was added on top of PKA inhibitor (Bottom). (I) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right) in H. (J) Typical mEPSC traces recorded from cocultured DRG and dorsal horn neurons, first in normal Ca²⁺-containing bath solution (Top), then TRPA1 antagonist (A967079) was applied (Middle), and last PKA agonist (25 μM sp-cAMP) was added on top of TRPA1 antagonist (Bottom). (K) Quantification for the normalized mEPSC event number per minute (Left) and the amplitude (pA, Right) in J. Data were shown as a violin plot with all points included. Paired student’s t test for C and E. One-way ANOVA with a post hoc test enabling multiple comparisons for G, I, and K. ns, not significant. *P < 0.05; **P < 0.01; ***P < 0.001.

SmCa-Mediated Continuous Neurotransmission Regulates Chronic Pain Behaviors.

To study the physiological relevance of smCa-mediated spontaneous vesicle release in DRG neurons, we took advantage of the Complete Freund’s adjuvant (CFA)-induced chronic inflammatory pain model (Fig. 6 A and B) (25). Intriguingly, the firing rate and the number of active regions of smCa activities in DRG neurons from CFA-inflamed WT mice were significantly higher than those from saline-control mice (Fig. 6 C and D and Movies S7 and S8). However, the effect of CFA was diminished in TRPA1^−/− mice (Fig. 6D and Movies S9 and S10). Consistent with smCa activities, the spontaneous release of NPY-pHluorin-labeled vesicles was increased remarkably in DRG neurons from CFA-injected WT mice than that receiving saline injection (Fig. 6 E and F and Movies S11 and S12). Again, the spontaneous release of TRPA1^−/− DRG neurons was insensitive to CFA inflammation (Fig. 6F and Movies S13 and S14). These data indicate that smCa-mediated continuous neurotransmission is elevated in CFA-induced chronic pain. Next, we performed hot-plate and mechanical Von Frey behavior tests for CFA- and saline-injected WT or TRPA1^−/− mice. As expected, compared with the saline control group, CFA-inflamed WT mice exhibited significant pain algesia with shorter latencies of paw licking on the hot plate and lower mechanical pain threshold (Fig. 6 G and I). Importantly, the CFA-induced pain algesia was alleviated in TRPA1^−/− mice (Fig. 6 G and I), or in WT mice preinjected with different doses of PKA inhibitor H89 (Fig. 6 H and J). Consistent with the blockade effects on smCa activities (Fig. 1 F–M), local application of BAPTA-AM, but not EGTA-AM, effectively alleviated the CFA-induced hyperalgesia in WT mice (SI Appendix, Fig. S8 A and B). Furthermore, BAPTA-AM, but not EGTA-AM, increased the threshold of mechanical pain in CFA-treated WT mice to a similar level as that of the CFA-inflamed TRPA1^−/− mice (SI Appendix, Fig. S8 C–E). Thus, smCa-mediated continuous neurotransmission may play an essential role in chronic pain.

Fig. 6.

Spontaneous smCa-mediated continuous neurotransmission regulates chronic pain behavior. (A) A photograph showing the hind paw’s thickness between the ipsilateral and contralateral sides after saline or CFA injection, respectively. (B) Quantification of the hind paw’s thickness between the ipsilateral and contralateral side after saline or CFA injection, respectively. (C) The whole-cell surface plot of the spontaneous Ca²⁺ signals in DRG neurons from WT C57 injected with saline or CFA. (D) Quantification for the total smCa ROIs number per cell per minute in DRG neurons from WT C57 or TRPA1^−/− mice injected with saline and CFA, respectively. (E) The TIRF images of spontaneous vesicle release in DRG neurons from WT C57 mice injected with saline or CFA group. (Scale bar, 10 μm.) (F) Quantification for the spontaneous quantal vesicle release events per cell per minute in DRG neurons from WT C57 or TRPA1^−/− mice injected with saline and CFA, respectively. (G) Statistics of the latency of paw licking in hot plate behavior test for WT or TRPA1^−/− mice injected with saline or CFA. (H) Statistics of the latency of paw licking in hot plate behavior test for CFA-inflamed WT mice 1-h after the acute hind paw injection of saline, 20 μM H89, or 200 μM H89. (I) Statistics of the mechanical pain threshold in Von Frey behavior test for WT or TRPA1^−/− mice injected with saline or CFA. (J) Statistics of the mechanical pain threshold in Von Frey behavior test for CFA-inflamed WT mice 1-h after the acute hind paw injection of saline, 20 μM H89, or 200 μM H89. Data were shown as a violin plot with all points included. One-way ANOVA-test for H and J. Two-way ANOVA-test for G and I. Paired student’s t test for B. Student’s t test for D and F. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001.

Discussion

Chronic pain is a multifaceted and unpleasant condition that affects more than 30% of individuals worldwide (1, 2). In the current study, we found that the primary sensory DRG neurons exhibit frequent AP-independent smCa activities, which trigger continuous neurotransmission via the cAMP-PKA signaling pathway. Importantly, the smCa-mediated continuous neurotransmission increased dramatically in chronic pain and may play an important role in hyperalgesia (Fig. 7).

Fig. 7.

Cartoon illustration showing that the AP-independent continuous neurotransmission regulates hyperalgesia behavior. Spontaneous Ca²⁺ influx through the TRPA1 channel forms a local microdomain Ca²⁺ signal (smCa) independent of spontaneous APs, which continuously trigger the spontaneous vesicle release via PKA. The smCa-mediated continuous neurotransmission increased remarkably in chronic pain and may serve to regulate hyperalgesia behaviors in vivo. The image of spinal cord structure was modified from D. Michael McKeough’s work

One of the major findings of the current work is the hotspots of AP-independent smCa activities in the somatic and axonal regions of primary sensory DRG neurons, supported by the following evidence: i) By using TIRF live-imaging, we observed frequent local Ca²⁺ transients in both the somata and axons of DRG neurons without excitatory stimuli (Fig. 1 A–E, SI Appendix, Fig. S1 A and B, and Movies S1–S4). ii) The spontaneous local Ca²⁺ transients were largely blocked by the fast Ca²⁺-chelator BAPTA-AM, but not the slow Ca²⁺-chelator EGTA-AM (Fig. 1 F–M), indicating the local Ca²⁺ transients to be microdomain Ca²⁺ signals. iii) The microdomain Ca²⁺ signal is independent of membrane potential or the firing of APs (Fig. 2 D and E, Left and SI Appendix, Fig. S2 C–F). iv) The smCa signals are critically dependent on extracellular Ca²⁺ influx (Fig. 2 A–C and SI Appendix, Fig. S1 A–E). v) The nonselective TRP channel blocker, RR almost eliminated the smCa signals in DRG neurons (Fig. 2F and SI Appendix, Fig. S2 I and J). vi) The TRPA1 channel antagonists, A967079 and HC030031, reduced the smCa signals in both the somata and axons of DRG neurons significantly (Fig. 2G and SI Appendix, Fig. S3). vii) The smCa events were diminished in TRPA1^−/− DRG neurons (Fig. 2I and SI Appendix, Fig. S4) and increased in the DRG neurons with TRPA1 overexpression (Fig. 2J and SI Appendix, Fig. S5). Taken together, the spontaneous opening of TRPA1 channels mediates smCa activities in DRG neurons.

TRPA1 is one of the TRP channels expressed in the nociceptive DRG neurons and involved in multiple sensations, including cold, mechanical, and itchy sensations (26–28). In addition, TRPA1 is activated by a broad range of reactive chemicals and inflammatory agents (protons, fatty acid derivatives, cytokines, chemokines, etc.), and thus plays an important role in analgesia (29, 30). Our current study demonstrates that the spontaneous activation of TRPA1 is the primary source of the frequent smCa activities in DRG neurons. Interestingly, TRPA1 may not be the only channel responsible for smCa signals because we observed smCa activities in nearly all types of DRG neurons, not just the small-sized DRG neurons where the TRPA1 channel is commonly found (31). Our present study focused on small-sized DRG neurons because of their crucial roles in pain sensation. However, there was still a small number of smCa activities remaining in the small-sized DRG neurons incubated with TRPA1 antagonist or from TRPA1^−/− mice (Fig. 2 G and I and SI Appendix, Figs. S3 and S4). Since the nonselective TRP channel antagonist, RR almost eliminated the smCa events (Fig. 2F), the remaining TRPA1-independent smCa activities might be contributed by some other TRP channels. The origin and function of the remaining smCa signals in DRG neurons need further investigation. In addition, our current work is mainly performed in cultured neurons, and we shall continue to explore the smCa activities in more physiological conditions in the future.

As an important second messenger, local Ca²⁺ signals are involved in multiple cellular procedures, including excitation–contraction coupling (32), gene expression (33), cell migration (34), and neurotransmission (21, 35). Our data demonstrate that TRPA1-mediated smCa activities trigger spontaneous neurotransmission in DRG neurons, including somatic and axonal release of neurotransmitters or neuromodulators observed with real-time TIRF imaging of the vesicles labeled with Spy-pHluorin or NPY-pHluorin (Fig. 3 and Movies S5 and S6), and glutamate release from the synapses formed between DRG and dorsal horn neurons detected by mEPSC recordings (Fig. 4). The crucial roles of Ca²⁺ in AP-triggered neurotransmission have been well established that VGCCs-mediated Ca²⁺ influx triggers vesicle release via the Ca²⁺ sensor of synaptotagmin (36, 37). On the other hand, Ca²⁺ also contributes essentially to spontaneous neurotransmission, but the Ca²⁺-sensor for spontaneous neurotransmission is still controversial (38, 39). PKA has been shown to enhance spontaneous neurotransmitter release via the phosphorylation of complexin (24). PKA also plays an important role in regulating the releasable vesicle pool (40), which in another way facilitates spontaneous neurotransmitter release. Importantly, two adenylyl cyclases (AC1 and AC8) are known to be activated by Ca²⁺/calmodulin (41), rendering PKA a potent candidate involved in Ca²⁺-triggered spontaneous neurotransmission. Indeed, our data showed that the specific antagonist of PKA (H89) substantially reduced the spontaneous release of NPY-pHluorin labeled vesicles (Fig. 5 A–C) and mEPSC events from dorsal horn neurons cocultured with DRG neurons (Fig. 5 D and E). In addition, sp-cAMP, an activator of PKA restored the mEPSC events after inhibiting the smCa signals with TRPA1 antagonist (Fig. 5 J and K), indicating that PKA is downstream of smCa. However, we could not fully exclude the potential long-term effect of PKA on smCa, which might have a systematic impact on the plasticity changes at molecular, structural, and functional levels. Taken together, PKA plays an essential role in smCa-mediated continuous neurotransmission from DRG neurons to the central spinal cord.

As the primary sensory neurons, DRG neurons play a pivotal role in the development and maintenance of chronic pain. In dorsal root ganglia and dorsal horn where pain signal is transmitted from DRG neurons to the central nervous system, the extracellular concentration of different neurotransmitters (glutamate, ATP, etc.) and neuromodulators (Substance P, CGRP, NPY, etc.) are accumulated in chronic pain, which contributes essentially to pain behaviors (10–15). The elevated concentration of these neurotransmitters and neuromodulators is thought to be caused by the increased firing rate of spontaneous APs of DRG neurons in chronic pain (42). Our data, however, offers direct evidence that spontaneous vesicle release from both the somata and the axons of DRG neurons is independent of APs (SI Appendix, Fig. S6E). Instead, TRPA1-mediated smCa activities and the downstream PKA signaling pathway are essential for the AP-independent spontaneous neurotransmission in DRG neurons (Figs. 3–5). Importantly, the smCa activity and its triggered spontaneous neurotransmission were increased significantly in animals experiencing inflammatory pain (Fig. 6 C–F and Movies S7–S14), and the CFA-induced pain hyperalgesia was alleviated by TRPA1^−/− or PKA antagonist, shown to inhibit smCa-mediated spontaneous neurotransmission (Fig. 6 G–J). It is not surprising that TRPA1 and PKA are involved in CFA-induced hyperalgesia because they are well studied in previous literature (43, 44). TRPA1 is known to be expressed in nociceptive DRG neurons and involved in both physiological noxious-stimuli sensations and chronic pain hyperalgesia (26–28). For PKA, it has been reported that continuous activation of the cAMP-PKA pathway maintains the hyperexcitability of sensory neurons and behavioral hyperalgesia (45, 46), and the antagonists targeting the cAMP-PKA pathway greatly suppress thermal and mechanical hyperalgesia (47–49). Additionally, we found that BAPTA-AM, but not EGTA-AM, significantly alleviated the CFA-induced hyperalgesia in vivo (SI Appendix, Fig. S8), suggesting the specific role of TRPA1-mediated smCa in chronic pain. Thus, our study reveals a potential role of AP-independent continuous neurotransmission from DRG neurons in hyperalgesia, which may provide a unique target for the treatment of chronic pain in the future.

Materials and Methods

The use and care of animals were approved and directed by the Institutional Animal Care and Use Committee of Xi’an Jiaotong University and Peking University and the Association for Assessment and Accreditation of Laboratory Animal Care. High-resolution live-cell TIRF imaging was adopted to observe smCa transients and quantal vesicle release in cultured sensory DRG neurons from juvenile or adult rodents. Patch-Clamp was adopted to record mEPSCs in cocultured DRG-spinal dorsal horn synapses or fresh spinal cord slices with DRG afferent fibers from adult male rats. Hot Plate and Von Frey behavior tests were adopted to record the pain hyperalgesia in the CFA-inflamed mice model. For more details about the materials and methods, please see our SI Appendix file.

Supplementary Material

Appendix 01 (PDF)

Movie S1.

The TIRF image of smCa in the soma of a DRG neuron.

Movie S2.

The surface plot of smCa in the soma of a DRG neuron.

Movie S3.

The TIRF image of smCa in the axons of a DRG neuron.

Movie S4.

The surface plot of smCa in the axons of a DRG neuron.

Movie S5.

The spontaneous release of NPY-pHluorin labeled vesicles in the soma of a DRG neuron.

Movie S6.

The spontaneous release of Spy-pHluorin labeled vesicles in the soma of a DRG neuron.

Movie S7.

The TIRF image of smCa in the soma of a DRG neuron from the saline-treated WT mice.

Movie S8.

The TIRF image of smCa in the soma of a DRG neuron from the CFA-treated WT mice.

Movie S9.

The TIRF image of smCa in the soma of a DRG neuron from the saline-treated TRPA1^-/- mice.

Movie S10.

The TIRF image of smCa in the soma of a DRG neuron from the CFA-treated TRPA1^-/- mice.

Movie S11.

The spontaneous release of NPY-pHluorin labeled vesicles in the soma of a DRG neuron from the saline-treated WT mice.

Movie S12.

The spontaneous release of NPY-pHluorin labeled vesicles in the soma of a DRG neuron from the CFA-treated WT mice.

Movie S13.

The spontaneous release of NPY-pHluorin labeled vesicles in the soma of a DRG neuron from the saline-treated TRPA1^-/- mice.

Movie S14.

The spontaneous release of NPY-pHluorin labeled vesicles in the soma of a DRG neuron from the CFA-treated TRPA1^-/- mice.

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.