Reactive oxygen species mediate TNFR1 increase after TRPV1 activation in mouse DRG neurons

Background Transient receptor potential vanilloid subtype 1 (TRPV1) is activated by low pH/protons and is well known to be involved in hyperalgesia during inflammation. Tumor necrosis factor α (TNF-α), a proinflammatory cytokine, is involved in nociceptive responses causing hyperalgesia through TNF receptor type 1 (TNFR1) activation. Reactive oxygen species (ROS) production is also prominently increased in inflamed tissue. The present study investigated TNFR1 receptors in primary cultured mouse dorsal root ganglion (DRG) neurons after TRPV1 activation and the involvement of ROS. C57BL/6 mice, both TRPV1 knockout and wild type, were used for immunofluorescent and live cell imaging. The L4 and L5 DRGs were dissected bilaterally and cultured overnight. TRPV1 was stimulated with capsaicin or its potent analog, resiniferatoxin. ROS production was measured with live cell imaging and TNFR1 was detected with immunofluorescence in DRG primary cultures. The TRPV1 knockout mice, TRPV1 antagonist, capsazepine, and ROS scavenger, N-tert-Butyl-α-phenylnitrone (PBN), were employed to explore the functional relationship among TRPV1, ROS and TNFR1 in these studies. Results The results demonstrate that TRPV1 activation increases TNFR1 receptors and ROS generation in primary cultures of mouse DRG neurons. Activated increases in TNFR1 receptors and ROS production are absent in TRPV1 deficient mice. The PBN blocks increases in TNFR1 and ROS production induced by capsaicin/resiniferatoxin. Conclusion TRPV1 activation increases TNFR1 in cultured mouse DRG neurons through a ROS signaling pathway, a novel sensitization mechanism in DRG neurons.


Background
Inflammation comes with sensitization of specialized sensory neurons that comprise the nociceptive pathway, leading to enhanced pain sensation in response to both noxious and non-noxious stimuli [1]. The transient receptor potential vanilloid 1 (TRPV1) participates in the development of chemical and thermal hyperalgesia in inflammation, through the action of low pH, heat and inflammatory mediators [2]. The most striking feature of TRPV1 knockout mice is the virtual absence of thermal hypersensitivity in the setting of inflammation [3][4][5][6]. Inflammation causes the release of inflammatory mediators which lower the threshold of heat stimulation, a process known as heat hyperalgesia depending on TRPV1 activation [7].
While reactive oxygen species (ROS) are primarily important for regulating many metabolic cellular activities, excess ROS have been implicated in neuropathic and inflammatory pain [20][21][22][23]. The transient but dramatic analgesic effect of freeradical scavengers in pain models suggests that ROS are critically involved in the generation of pain. Some research on the signaling pathways has been directed towards documenting a role for ROS in cell death and the induction of cell proliferation [24][25][26]. The effects of ROS may be mediated by reversible effects on intracellular proteins which lead to alterations in intracellular signaling pathways. Effects of ROS on aberrant signaling pathways in the nervous system remain to be determined.
Although some reports show that TNF-α sensitizes TRPV1, we are interested in the role that TRPV1 activation might have on TNFR1 as the drive for maintaining nociceptor sensitization. The present study investigated changes in TNFR1 in response to TRPV1 activation in primary DRG neuronal cultures from both TRPV1 knockout and wild type mice while monitoring involvement of ROS as a potential drive in this signaling pathway. Primary DRG neurons were dissected from wild type and TRPV1 knockout C57BL/6 mice. After overnight incubation, DRG cultures were challenged with either capsaicin or resiniferatoxin.

Capsaicin induces TNFR1 staining increases in DRG neurons from wild type mice
In untreated control primary cultures (24 hr) of DRG neurons from wild type mice, 22.82 ± 1.68% of the neurons stained for TRPV1, and 9.24 ± 0.27% stained for TNFR1 (Table 1). Photomicrographs shown in Figure 1 illustrate the green immunofluorescent staining for TRPV1 and red immunofluorescence identifying TNFR1. In order to explore expression of TNFR1 in mouse DRG neurons after capsaicin stimulation, we first examined the co-localization of TRPV1 with TNFR1. Color combination (yellow) microscopy indicated co-localization of TRPV1 and TNFR1 in some mouse DRG neurons. Dual staining for TRPV1 and TNFR1 was observed in 41.04 ± 3.94% of TRPV1 immunostaining positive neurons ( Table 1, Fig. 1 Merge) (n = 592). Staining for TRPV1 and TNFR1 were both observed primarily in small size neurons. Both capsaicin (1 μM, 20 min) and resiniferatoxin (200 nM, 20 min) treated groups had significantly increased TNFR1 staining (capsaicin treated group: 14.62 ± 0.72%, n = 431, P < 0.05; resiniferatoxin treated group: 16.5 ± 1.37%, n = 622, P < 0.01), compared with the untreated group ( Fig. 2A, B). There was an increase in the percentage of neurons showing dual staining for TNFR1 and TRPV1 after treatment with capsaicin or resiniferatoxin. In the capsaicin treated group, TNFR1 was identified in 52.89 ± 2.89% of neurons with TRPV1 immunoreactivity (P < 0.05 compared with the untreated group), and TNFR1 was found in 73.63 ± 3.54% of TRPV1 neurons in the resiniferatoxin treated group (P < 0.01 compared with the untreated group; P < 0.05 compared with the capsaicin treated group). This indicates that primary mouse DRG neurons are more responsive to potent resiniferatoxin as expected (Table 1). However, the TRPV1 staining was not changed after capsaicin or resiniferatoxin stimulation (capsaicin treated group: 27.91 ± 2.58%; resiniferatoxin treated group: 23.19 ± 2.15% (P > 0.05 for both compared with the untreated group 22.82 ± 1.68%)) ( Table  1).

PBN blocks TRPV1 activated increases in ROS production in DRG neurons of wild type mice
Experiments were performed to analyze ROS production after capsaicin treatment in cultured DRG neurons from wild type mice. N-tert-butyl-α-phenylnitrone (PBN) is an electron spin trap with high avidity for free radical species and functions as an antioxidant in many biological systems. PBN was used to block ROS production after TRPV1 activation with capsaicin. The color intensity increases in the microscope image indicate that oxidized derivative DCF is accumulating in cells (Fig. 4A, B). The DCF accumulation indicates that ROS production is increasing.
Although there was a slow, steady increase of ROS production in all neurons in normal physiological experimental solution due to spontaneous oxidation of the dye over time, ROS production was significantly increased in cultured DRG neurons after bath application of capsaicin (

Capsaicin increases ROS production in DRG neurons of wild type mice
Experiments were performed in cultured DRG neurons from wild type and TRPV1 knockout mice to determine ROS production changes after capsaicin or resiniferatoxin stimulation. In the experiment with wild type mouse DRG neurons, capsazepine (10 μM) added 10 min prior to start of recording blocked both capsaicin (1 μM (Fig. 5C).

Discussion
The present study demonstrates that capsaicin increased both the TNFR1 staining and ROS production in cultured DRG neurons from wild type mice through TRPV1 activation. These phenomena did not occur in TRPV1 deficient mice. ROS scavenger PBN and TRPV1 antagonist, capsazepine, blocked both the TNFR1 and ROS increases. We conclude that TRPV1 activation increases TNFR1 through a signaling pathway driven by ROS generation.

Regulation of TNFR1 by capsaicin
Two subtypes of TNF-α receptors, TNFR1 and TNFR2, are expressed in peripheral nerves [16]. TNFR1 is expressed in dorsal root ganglion neurons and plays a potential role in nociception [13]. In present experiment, TNFR1 is posi-tively immunostained in 10% of mouse DRG neurons and almost all of which were small size neurons which suggests a potential role of TNFR1 in transmission of nociception. TNFR1 in mouse DRG neurons is detected after sciatic nerve injury but not in control mice [27]. The process of dissociating and culturing DRG neurons apparently increases expression in the cells somewhat since TNFR1 is not found in normal mouse DRG tissue sections. In rat inflammation or nerve injury models, increases in TNFR1 expression are reported by several other researchers [15,17,[28][29][30]. Although there are reports that TNF-α increases TRPV1 expression through TNF-α receptors and TRPV1 induced activity is enhanced by TNF-α administration [31-33], it would be interesting to know if they can reciprocally regulate one another. The interactive relationship of these two types of receptors will be the target of further investigation since the present study suggests TRPV1 activation is inducing TNFR1 increases and/or activation. In the present study using mouse DRG neuronal cultures, TNFR1 was significantly increased after stimulation with capsaicin/resiniferatoxin. The responses to capsaicin and resiniferatoxin were different since resiniferatoxin is a more robust agonist for TRPV1. This result is supported by calcium influx experiments where resiniferatoxin produces a stronger calcium increase [3].
To confirm that TNFR1 increases induced by capsaicin are mediated through TRPV1, TRPV1 knockout mice were employed and additional experiments using capsazepine were also performed. The results of both these experiments showed that capsaicin/resiniferatoxin induced increases in TNFR1 in small size DRG neurons co-expressing TRPV1, and thus are likely to be nociceptors. This novel finding of capsaicin induced TNFR1 increases in cultured mouse DRG neurons has not been reported by other investigators. The result is in agreement with those in a rat inflammation model [13]. In CFA arthritis models in vivo, upregulation of TNFR1 receptors is detected in peripheral joint tissue, DRG neurons, and small fiber bundles within the root entry zone and fibers bordering the trigeminal and spinal cord dorsal horn [28,34,35].
During inflammation, high levels of H + are released from injured tissue and are important mediators activating TRPV1. In fact, there are several other mediators activating TRPV1 such as anandamide and high temperature (43°C) that are present in inflamed tissue targeted by sensory nerve terminals. Thermal hyperalgesia and mechanical allodynia are triggered by TRPV1 over-activation and plastic changes occur in neurons with influxes of ions such as calcium and sodium. One of the changes we find is increasing TNFR1 in TRPV1 stained neurons which would allow neurons to receive information presented in inflammatory conditions of which TNF-α plays an important role. The present study demonstrates the mechanistic linkage between TNFR1 and TRPV1 activation. TNF-α is pro-duced by satellite cells in DRG, Schwann cells in the peripheral nervous system and local macrophage cells which increase TNF-α release from peripheral tissue attracting considerable attention in nociceptive sensation and inflammatory cell invasion during inflammation. TNF-α mRNA increases are observed 24 hours after periganglionic injections of complete Freund's adjuvant [36]. Anti-TNF therapy used in treatment of rheumatoid arthritis produces profound analgesia [8,11], indicating that TNF-α and its receptors may be good therapeutic targets in other chronic inflammatory pain conditions. The two distinct subtypes of TNF-α receptors have different functions in inflammation. TNFR1 functions mainly in the process that leads to hyperalgesia and most of TNFR2's effects are involved in immunological responses. While depressing the immunological responses is also a way to reduce inflammation induced hyperalgesia, we have chosen to study neuronal TNFR1 effects in this investigation. Significant increases in TNFR1 are relevant based on behavioral evidence showing that antibody against TNFR1 applied to the epineuria abolishes hyperalgesia caused by chronic constriction nerve injury [14]. During inflammation, more TNF-α is released, and TNFR1 expression in neurons is increasing in proportion to the changes in TNF-α release from peripheral tissue and DRG satellite cells [8,29]. A highly active condition in which ROS, TNF-α and TNFR1 are increasing would cause hypersensitivity of peripheral nerves based on our studies presented here for TRPV1 activation and others in literature.

PBN blocks capsaicin induced TNFR1 increases
ROS are not only endpoint products of TRPV1 activation but also signaling pathway elements. Capsaicin or resiniferatoxin activation of TRPV1 increases TNFR1 staining in cultured mouse DRG neurons (C) There is no change of TNFR1 in DRG neurons from TRPV1 knockout mice after capsaicin or resiniferatoxin stimulation.

TRPV1 activation induces increases in ROS production
To determine further if ROS is involved in regulation of TNFR1 increases after TRPV1 activation, we also studied ROS production after TRPV1 activation by capsaicin in the mouse DRG neuronal cultures. We assessed cytosolic oxidant activity with use of CM-H2DCFDA, a redox-sensitive fluorochrome probe to monitor ROS production. CM-H2DCFDA oxidation to DCF is a nonspecific reaction used to monitor generation of biological oxidants that include ROS. It is important to recognize that intracellular conversion of H2DCFDA to DCF occurs in competition with other oxidation reactions. The overall rate of H2DCFDA conversion to DCF therefore reflects a complex, highly dynamic balance between simultaneous rates of oxidant production and oxidant buffering by multiple pathways. We interpret this integrated signal as net oxidant activity under a given condition. The data indicate that ROS production is increased by challenging neurons with capsaicin/resiniferatoxin. The results from TRPV1 deficient mice show no increase in ROS production after capsaicin stimulation. The data for the capsazepine pretreatment experiment also confirmed this finding.  [3,42,43]. However, the mechanism of generating calcium influx in response to ROS production increase is still unclear. Significant increases in ROS production do not occur in the present study until 30 minutes since ROS levels in cells are tightly regulated. ROS generation during inflammation has attracted increasing attention in the literature. It is also known that the ROS scavenger used to block ROS production also reverses hypersensitivity [23,44]. While much work has been done in vivo in rat, our work in cultured mouse DRG neurons challenging TRPV1 with capsaicin to increase ROS production is similar to in vivo microenvironment conditions present during inflammation. The finding that PBN abolished ROS production induced by capsaicin in the study further confirms that elevated ROS production is triggered by TRPV1 activation. Thus, ROS are implicated as important functional and potentially therapeutic targets influencing neurotransmitter responses, inflammatory mediator receptors and neurogenic vasodilatation relevant to treatment of inflammation and pain.

Conclusion
The present study reveals that TRPV1 activation increases ROS which act as signaling molecules to increase TNFR1 in mouse DRG neurons. Thus, in conditions of injured or inflamed tissues, TRPV1 activation increases of ROS and inflammatory mediator TNF-α receptor content could enhance peripheral nerve sensitization, inflammatory pain and potentially other pathological events.

Experimental methods
This study was performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health and was approved by the The ROS scavenger PBN blocks TNFR1 increases in cultured mouse DRG neurons induced by capsaicin or resiniferatoxin Resiniferatoxin significantly increases the TNFR1 increase in DRG neurons (**P < 0.01). PBN blocks the TNFR1 increase induced by resiniferatoxin ( ## P < 0.01 compared to resiniferatoxin). Results indicate that capsaicin and resiniferatoxin increase TNFR1 in DRG neurons through a ROS signaling pathway.

(page number not for citation purposes)
Capsaicin or resiniferatoxin induces ROS production in cultured mouse DRG neurons through TRPV1 activation Figure 5 Capsaicin or resiniferatoxin induces ROS production in cultured mouse DRG neurons through TRPV1 activation. (A) Capsaicin significantly increases ROS production and capsazepine blocks capsaicin induced ROS production (30 minutes: ## P < 0.01 compared to capsaicin). (B) Resiniferatoxin significantly increases ROS production and resiniferatoxin blocks capsaicin induced ROS production (30 minutes: ### P < 0.001 compared to resiniferatoxin). (C) There is no change in ROS production in DRG neurons from TRPV1 deficient mice after capsaicin or resiniferatoxin stimulation.

University of Kentucky Institutional Animal Care and Use
Committee.

Primary DRG neuronal cultures
Male and female TRPV1 knockout or wild type C57BL/6 mice (19 -20 g, Jackson Lab, as a tissue sharing gift from Dr. Lu-Yuan Lee) were used in these experiments [42]. Mice were anesthetized and the vertebral columns were exposed, removed, and put into the ice-cold dissection solution (NaCl 135, KCl 5, KH 2 PO 4

Determination of ROS production by H2DCFDA fluorescent probe
All experiments were done at room temperature. Cells were kept at room temperature for 30 min before replacing culture medium with DPBS, pH 7.4. On the day of the experiment, the ROS-sensitive fluorescent dye, 5-(and-6)chloromethyl-2', 7'-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA, Invitrogen) was freshly reconstituted. Cells were pre-loaded with CM-H2DCFDA in DPBS at a final concentration of 4 μM for 30 min. The loading buffer was removed and replaced with DPBS allowing 20 min to 2 h recovery time for cells to exhibit low levels of esterase activity. ROS production was monitored for 30 minutes in the live cell imaging experiment. Cells were stimulated with bath application of capsaicin (1 μM) or resiniferatoxin (200 nM) and in some cases neurons were pretreated 10 minutes with TRPV1 antagonist capsazepine (10 μM) before beginning the recording and capsaicin stimulation. ROS scavenger N-tert-Butyl-αphenylnitrone (PBN, 2 mM) was administered for 90 minutes before capsaicin stimulation.
The change in fluorescent intensity over time is proportional to DCF accumulation, an index of net oxidant activity in the cytoplasm. Fluorescent imaging was performed using an Eclipse 2000-S microscope (Nikon) equipped with a high sensitivity digital monochrome Cool SNAP ES camera (Photometrics, Tucson, AZ) controlled by Meta-Morph software (Molecular Devices, Sunnyvale, CA). DCF was excited at 480 nm and emitted light was collected using a 520 nm long pass filter. The photos were captured under 10 ms exposure at 1 min intervals for 30 min. Emissions from DCFH-free buffer are not detectable under these conditions, and background correction is not required. Photooxidation artifact was controlled by conducting experiments in a darkened laboratory and by standardizing excitation parameters to minimize the cumulative delivered energy. The baseline fluorescent intensity was determined prior to exposing the cells to experimental inducements. MetaMorph offline software (Molecular Devices, Sunnyvale, CA) was used for image data analysis.

Statistics
All data were expressed as mean ± SEM. Comparisons among multiple means were analyzed with the one-way ANOVA. Post hoc test was analyzed with the Newman-Keuls Multiple Comparison test. A P < 0.05 was considered significant.