Shang et al. (2016. J. Cell Biol.https://doi.org/10.1083/jcb.201603081) reported that activation of lysosomal TRPA1 channels led to intracellular calcium transients and CGRP release from DRG neurons. We argue that both findings are more likely due to influx of insufficiently buffered extracellular calcium rather than lysosomal release.
A recent article (Shang et al., 2016) proposed functional expression of the chemoreceptive ion channel TRPA1 (transient receptor potential channel A1) in calcium-storing lysosomes of nociceptive mouse dorsal root ganglion (DRG) neurons: The agonist allyl isothiocyanate (AITC) increased cytosolic calcium in the nominal absence of extracellular calcium ions that would otherwise flow in through the activated TRPA1 channels in the plasma membrane. 40% of the overall stimulated calcium rise was attributed to lysosomal release and reported as sufficient to evoke vesicular exocytosis of calcitonin gene–related peptide (CGRP) from the neuronal somata. Cultured rodent DRG nerve cell bodies have shown merit as surrogates for their inaccessible peripheral terminals, but they remain what they are: a model, with differences in geometry, epigenetics, protein infrastructure, environment, and function compared with their peripheral nerve endings. In those terminals, in the trachea for example, we do not find detectable CGRP release upon TRPA1 activation by AITC and formaldehyde, if extracellular calcium is safely removed (Fig. 1). This physiological result is in accord with CGRP measurements of spinal sensory terminals stimulated with various electrophilic TRPA1 agonists and of DRG neurons stimulated with bradykinin, which also acts through TRPA1 activation (Vasko et al., 1994; Bandell et al., 2004; Quallo, 2015; Benemei et al., 2017; Li Puma et al., 2019).
To examine this discrepancy in DRG neurons, we replicated experiments from Shang et al. (2016). However, there were two purpose-designed changes in our approach. We employed complete bath exchange using 10 mM in comparison to 1 mM EGTA as extracellular calcium buffer; Shang et al. (2016) puffed the extracellular solution onto the cells and used 1 mM EGTA. We consider 1 mM insufficient to detach and chelate all loosely attached calcium ions binding to the anionic extracellular moieties of membrane proteins. We repeated the key experiment of Shang et al. (2016) (their Fig. 1 C), stimulating hundreds of mouse DRG neurons (cultured for 4–6 h) with 100 µM AITC in the presence of 1 mM EGTA and the absence of calcium (Fig. 2, A and D). Indeed, 14% of the cells responded with small calcium transients, apparently indicating some intracellular calcium release. However, we did not observe any calcium transients in the presence of 10 mM EGTA, while 35% of DRG neurons responded with a large calcium influx in the presence of extracellular calcium. Using DRG neurons that had been rested overnight in culture, 4% responded to AITC with calcium transients in the presence of 1 mM EGTA, and no calcium transients were observed in the presence of 10 mM EGTA (Fig. 2, B, C, and E). Hsu and Lee (2015) used 5 mM EGTA in calcium-free external solution and found no intracellular calcium increase in response to 100 µM AITC applied to sensory neurons of the jugular ganglion. Given the limited selectivity of many TRPA1 agonists, including AITC (Gees et al., 2013), it is important that other electrophilic TRPA1 agonists also did not increase cytosolic calcium in DRG neurons in the safe absence of external calcium (Eberhardt et al., 2012; Fischer et al., 2015).
Our procedure to achieve calcium-free conditions involves a 10-min washout of extracellular calcium with EGTA buffer, and this might deplete the intracellular stores. However, using a novel custom-made miniature dish accommodating the freshly dissociated DRG neurons, we achieved nominally calcium-free conditions within <1 s (Fig. 3, A and B). Maintaining these conditions, AITC was applied after 1 min and slightly activated 6 of 202 neurons in the presence of 1 mM EGTA; in the presence of 10 mM EGTA, 0 of 211 neurons exhibited calcium transients in response to AITC (Fig. 3, C and D).
Residual calcium ions in nominally calcium-free physiological buffers result from inevitable contaminations of the commercial American Chemical Society–grade constituents, which can amount to 12 µM calcium in a final solution (Patton et al., 2004). Exactly this extracellular concentration is sufficient to cause a significant potentiation of agonist-evoked rTRPA1 currents. This potentiation depends on the influx of calcium ions through the membrane channel; intracellular calcium is an agonist of TRPA1, and half-maximal activation of the channel is achieved by 225 nM calcium (Wang et al., 2008). The problem of contamination can be minimized by raising the EGTA concentration, for example from 1 to 10 mM, which reduces the relative error from 12% to 1% in case of an assumed 10 µM calcium contamination (Patton et al., 2004).
Shang et al. (2016) used glycylphenylalanine 2-naphtylamide (GPN) to osmotically disrupt the lysosomes, which caused a small cytosolic calcium transient (Fig. 3 C in Shang et al. ; and our Fig. S1). We used Fura-2 wide-field microfluorimetry and confirmed GPN-induced intracellular release of calcium ions that is quenched by pretreatment with the membrane permeable BAPTA-AM (Fig. 3 I). We also confirmed that GPN disrupts lysosomes in DRG neurons using LysoTracker, a fluorescent dye for tracking acidic organelles (Fig. 3, E–G). In addition, we found that the membrane-permeable AITC had no effect on LysoTracker fluorescence (Fig. 3 H), although TRPA1 should pass protons when activated (Ye et al., 2018). As shown in Shang et al. (2016), we observed that prolonged GPN pretreatment of the DRG neurons reduced the large AITC response in the presence of external calcium as much as the removal of extracellular calcium (Fig. 3, G and H, in Shang et al. ; and inserts in our Fig. 2 A). However, in our hands, the yield of responsive neurons was also greatly reduced (mean n = 318 vs. 63 per field of view), as was the KCl control response of the cells that underwent the GPN pretreatment (Fig. S2). These findings may suggest off-target effects of GPN, compromising the ability of the DRG neurons to respond to AITC and to nonspecific depolarization.
In conclusion, we propose that in DRG neurons cytosolic calcium increase evoked by AITC in the absence of external calcium, as described by Shang et al. (2016), is actually due to insufficient buffering of extracellular calcium ions and influx through activated TRPA1 channels in the plasma membrane. Whether calcium-induced calcium release plays an amplifying role remains to be elucidated (Ouyang et al., 2005). In addition, we found that TRPA1-evoked CGRP release from isolated peripheral nerves depends entirely on the availability of extracellular calcium.
Materials and methods
Animals and trachea preparation
The experiments were performed using C57BL/6 mice (in-house breeding colony; originally from Jackson Laboratory) and mice carrying targeted deletions of Trpv1 or Trpa1; the initial heterozygous breeding pairs were generous gifts of GlaxoSmithKline (Dr. John Davis, formerly Harlow, UK) or Drs. Kelvin Kwan and David Corey (Harvard University, Cambridge, MA), respectively, and the animals had been continuously back–crossed on C57BL/6 since 2001 or 2006. Mice were housed in group cages in a temperature-controlled environment on a 12-h light–dark cycle and were supplied with water and food ad libitum. Mice of either sex, body weight 15–25 g, were euthanized by exposure to a rising CO2 atmosphere. The trachea was excised from the thyroid cartilage to the carina and hemisected along the midline. One half of the preparation was then used as control, and the other half for experimental variations. The preparations were washed for 30 min at 37°C in synthetic interstitial fluid (SIF; in mM): NaCl 107.8, KCl 3.5, CaCl2 1.53, MgSO4 0.69, NaHCO3 26.2, NaH2PO4 1.67, sodium gluconate 9.64, glucose 5.55, and sucrose 7.6. The fluid had a temperature of 37°C and was gassed with carbogen (95% O2, 5% CO2) to obtain pH 7.4. The husbandry and usage of the animals were approved by the District Government of Mittelfranken (Ansbach, Germany), and experiments were performed in accordance with the guidelines of the International Association for the Study of Pain.
Stimulation and chemicals
After the initial washing period, the isolated trachea was consecutively incubated for 5 min in each of four tubes mounted in a temperature-controlled shaking bath at 37°C. The first two incubations were performed to determine basal CGRP release. The third tube contained the stimulating chemicals, 100 µM AITC (Sigma-Aldrich) or formaldehyde (Sigma-Aldrich), as TRPA1 agonists. Formaldehyde was bought as a saturated aqueous solution (formalin, 37–37.5 mass %, stabilized by methanol 8–12%; Carl Roth). For calcium-free conditions, CaCl2 was omitted from the SIF, and either 1 or 10 mM EGTA (Sigma-Aldrich) was added; this solution was then used to apply the stimulants and as pretreatment in case of subsequent AITC application.
CGRP enzyme immunoassay (EIA)
The CGRP content of the incubation fluid was measured using commercial EIA kits with a detection threshold of 5 pg/ml (Bertin Pharma). For this purpose 100 µl sample fluid was stored on ice and, immediately after the tissue exposure period, mixed with 25 µl of fivefold-concentrated commercial CGRP-EIA buffer that contained a proprietary cocktail of peptidase inhibitors. The antibody reactions took place overnight. The EIA plates were read photometrically using a microplate reader (Dynatech). All results were processed as measured by the EIA in picograms of CGRP per milliliter of SIF. To reduce interindividual variability and day-to-day baseline variability, the data refer to the second individual baseline value (before stimulation). This value was subtracted from all four data points of a typical experiment so that only the absolute change in CGRP release (Δ picograms per milliliter) is displayed in the figures.
Statistical comparisons were performed using Statistica 7 (Statsoft). All time series of experimental values were first analyzed for the effect of chemical stimulation compared with baseline CGRP secretion using nonparametric Wilcoxon matched pairs test. The baseline-corrected (i.e., Δ pg/ml) CGRP values were entered into a one-way ANOVA followed by Tukey’s honestly significant difference test, focusing on the peak values of stimulated CGRP release to compare the chemically stimulated responses in WT, TRPA1−/− knockout mice and double-knockout mice (TRPA1−/− and transient receptor potential channel V1 [TRPV1−/−]). P < 0.05 was considered statistically significant. Data points represent means ± SEM of the given number (n) of experiments on different animals. Triple asterisk symbols in the figures (***) indicate significant differences at P < 0.001.
DRGs were taken from the spinal columns of C57BL/6 mice. After removing DRGs from all spinal levels, the nerves and roots were cut off. The cleaned DRGs were incubated in 0.5% streptomyces protease and 1% clostridium collagenase at 37°C for 30 min (Sigma-Aldrich). The cells were first washed and then mechanically dissociated in serum-free TNB 100 medium supplemented with TNB 100 protein–lipid complex (Biochrom), penicillin, and streptomycin (100 U/ml each; Life Technologies) and murine nerve growth factor (mouse NGF 2.5S, 100 ng/ml; Alomone Labs). They were incubated on glass coverslips coated with poly-d-lysine (200 µg/ml; Sigma-Aldrich) in the above medium at 37°C in a 5% CO2 atmosphere for ∼24 h before the Ca-imaging measurements. Alternatively, the neurons were imaged for 4–6 h after incubation.
For Ca imaging, the fluorescent calcium indicator dye Fura-2-AM was used. It was mixed with 0.02% pluronic F-127 (Thermo Fisher Scientific or Biotrend) and diluted to 3 µM in extracellular solution containing (in mM): 1.25 calcium, 145 NaCl, 5 KCl, 10 glucose, 10 Hepes, and 1 MgCl2. DRG cells were loaded for ∼30 min at 37°C with 5% CO2 and then washed for 10 min at room temperature in either the Ca-containing extracellular solution for measurements with calcium or nominally Ca-free solutions containing 2 mM BAPTA, 1 mM EGTA, or 10 mM EGTA. In additional experiments, 3D-printed miniature superfusion chambers were used to accommodate freshly dissociated DRGs neurons; these allowed for rapid and complete bath exchange in <60 s before the stimulation with 100 µM AITC (Fig. 3). In some experiments, BAPTA-AM (10 µM; Merck) was applied to the cells during Fura-2 staining for 30 min and washed out before imaging.
The Ca imaging measurements were performed using an inverted microscope and excitation at 358- and 391-nm wavelengths with a Polychrome V monochromator (Till Photonics) at 1 Hz. A Peltier-cooled slow-scan charge-coupled device camera recorded the fluorescence emission >440 nm, controlled by TillVision software. The software was also used to calculate the fluorescence ratio for all selected neurons with background correction.
During the experiments, a gravity-driven and software-controlled common outlet perfusion system superfused the cells continuously with a flow rate of ∼1 ml/min. The protocols included a final application of 60 mM KCl in calcium-containing solution (even for the Ca-free measurements) to detect all responsive cells.
Lysotracker staining and imaging
DRG cells were loaded after 4–6 h in culture with Lysotracker 50 nM (Lysotracker Red DND-99; Thermo Fisher Scientific) for 1 h at 37°C with 5% CO2. After washing for 10 min at room temperature, Lysotracker fluorescence was monitored at 518 nm every 2 s. Control experiments allowed to correct for a slow but steady signal loss due to bleaching without intervention.
The measurements on the confocal microscope (Zeiss LSM 780) were performed to gain a higher resolution for the detection of lysosomes. The DRG cells were loaded with Lysotracker as described above and washed for 10 min at room temperature. The fluorescence was acquired before and after GPN 200 µM application, as indicated (in Fig. 3).
Online supplemental material
Fig. S1: In the absence of external calcium, we do not observe calcium transients in DRG neurons stimulated by the lysosome disruptor GPN. Fig. S2: Following GPN pretreatment, calcium imaging shows a massive reduction of DRG neurons responding to AITC and unspecific depolarisation by KCl.
We thank S. Haux-Oertel, A. Kuhn, I. Izidorczyk, and J. Schramm for excellent technical assistance and Birgit Vogler for assistance with confocal microscopy. Dr. Cosmin Ciotu is gratefully acknowledged for constructing the miniature dishes described in Fig. 3.
Funded by Deutsche Forschungsgemeinschaft (Interdisziplinäre Zentrum für Klinische Forschung Erlangen) Project E 27 to M.J.M. Fischer and DFG 704/3-1 to P.W. Reeh.
The authors declare no competing financial interests.
Author contributions: T.I. Kichko, M.J.M. Fischer, and P.W. Reeh: conception and design of research; T.I. Kichko and L. Gebhardt performed experiments; T.I. Kichko, M.J.M. Fischer, L. Gebhardt, and P.W. Reeh analyzed data and interpreted results of experiments; T.I. Kichko and L. Gebhardt prepared the figures; TIK and P.W. Reeh drafted the manuscript; T.I. Kichko, M.J.M. Fischer, L. Gebhardt, and P.W. Reeh approved the final version of the manuscript.
L. Gebhardt and T.I. Kichko contributed equally to this paper.