Meclizine and Metabotropic Glutamate Receptor Agonists Attenuate Severe Pain and Ca²⁺ Activity of Primary Sensory Neurons in Chemotherapy-Induced Peripheral Neuropathy

Abstract

Chemotherapy-induced peripheral neuropathy (CIPN) affects ∼68% of patients undergoing chemotherapy, causing debilitating neuropathic pain and reducing quality of life. Cisplatin is a commonly used platinum-based chemotherapeutic drug known to cause CIPN, possibly by causing oxidative stress damage to primary sensory neurons. Metabotropic glutamate receptors (mGluRs) are widely hypothesized to be involved in pain processing and pain mitigation. Meclizine is an H1 histamine receptor antagonist known to have neuroprotective effects, including an anti-oxidative effect. Here, we used a mouse model of cisplatin-induced CIPN using male and female mice to test agonists of mGluR8 and Group II mGluR as well as meclizine as interventions to reduce cisplatin-induced pain. We performed behavioral pain tests, and we imaged Ca²⁺ activity of the large population of dorsal root ganglia (DRG) neurons in vivo. For the latter, we used a genetically-encoded Ca²⁺ indicator, Pirt-GCaMP3, which enabled us to monitor different drug interventions at the level of the intact DRG neuronal ensemble. We found that CIPN increased spontaneous Ca²⁺ activity in DRG neurons, increased number of Ca²⁺ transients, and increased hyper-responses to mechanical, thermal, and chemical stimuli. We found that mechanical and thermal pain caused by CIPN was significantly attenuated by the mGluR8 agonist, (S)−3,4-DCPG, the Group II mGluR agonist, LY379268, and the H1 histamine receptor antagonist, meclizine. DRG neuronal Ca²⁺ activity elevated by CIPN was attenuated by LY379268 and meclizine, but not by (S)−3,4-DCPG. Furthermore, meclizine and LY379268 attenuated cisplatin-induced weight loss. These results suggest that Group II mGluR agonist, mGluR8 agonist, and meclizine are promising candidates as new treatment options for CIPN, and studies of their mechanisms are warranted.

SIGNIFICANCE STATEMENT Chemotherapy-induced peripheral neuropathy (CIPN) is a painful condition that affects most chemotherapy patients and persists several months or longer after treatment ends. Research on CIPN mechanism is extensive but has produced only few clinically useful treatments. Using in vivo GCaMP Ca²⁺ imaging in live animals over 1800 neurons/dorsal root ganglia (DRG) at once, we have characterized the effects of the chemotherapeutic drug, cisplatin and three treatments that decrease CIPN pain. Cisplatin increases sensory neuronal Ca²⁺ activity and develops various sensitization. Metabotropic glutamate receptor (mGluR) agonist, LY379268 or the H1 histamine receptor antagonist, meclizine decreases cisplatin's effects on neuronal Ca²⁺ activity and reduces pain hypersensitivity. Our results and experiments provide insights into cellular effects of cisplatin and drugs preventing CIPN pain.

Introduction

Chemotherapy-induced peripheral neuropathy (CIPN) is a side effect of chemotherapy, which affects ∼68% of patients, causing severe neuropathic pain that may persist for six months or longer in ∼30% of patients who have undergone chemotherapy (Quasthoff and Hartung, 2002; Seretny et al., 2014; Addington and Freimer, 2016; Cioroiu and Weimer, 2017; Flatters et al., 2017). Cisplatin is a platinum-based chemotherapeutic agent that inhibits tumor growth by crosslinking DNA nucleotides (Greystoke et al., 2017; Armstrong-Gordon et al., 2018), and which is widely used to treat cancer in various locations including lung, stomach, and head (Seretny et al., 2014; Calls et al., 2020). There have been extensive studies on strategies to reduce neuropathic pain and to understand its mechanisms at the level of individual cells, using in vitro dorsal root ganglia (DRG) explants and histologic analysis in animal models (Quasthoff and Hartung, 2002; Jaggi and Singh, 2012; Addington and Freimer, 2016; Cioroiu and Weimer, 2017). However, because of a lack of suitable tools and techniques, CIPN has not been studied in vivo at the level of an intact DRG neuronal ensemble, leaving an important gap in the understanding of CIPN mechanisms necessary to develop better therapeutics.

There are numerous nonmutually exclusive proposed mechanisms for CIPN caused by platinum-based chemotherapeutic drugs (Jaggi and Singh, 2012). These potential mechanisms including mitochondrial malfunction and cytoplasmic calcium unbalance (Melli et al., 2008), alteration of potassium channel subtype expression (Descoeur et al., 2011), upregulation of transient receptor potential vanilloid receptors (Anand et al., 2010; Ta et al., 2010), endoplasmic reticulum (ER) stress (Nawrocki et al., 2005; Yi et al., 2020), primary sensory neuron senescence (Calls et al., 2021), and oxidative stress via mitochondrial disruption generating reactive oxygen species (ROS; Joseph et al., 2008; Gorgun et al., 2017; Khasabova et al., 2019). ROS produced by mitochondria are normally broken down by the superoxide dismutase pathway (McCord and Fridovich, 1988). An SOD mimetic drug, calmangafodipir, reduced CIPN symptoms caused by oxaliplatin, a platinum-based drug (Glimelius et al., 2018). Of particular relevance to our work here, physiological changes associated with cisplatin-induced CIPN overlap extensively with known mechanisms of inflammatory pain (Hu et al., 2018; Quintão et al., 2019; Y. Zhu et al., 2019).

Concomitant administration of anti-inflammatory drugs with cisplatin has reduced CIPN-induced neuropathic pain both in animal models and in clinical trials (Glimelius et al., 2018; Khasabova et al., 2019). In vivo DRG imaging may be a particularly useful strategy for studying how concomitant drug administration modifies the effect of cisplatin on primary sensory neurons. All mGluRs except mGluR6 are found on central and/or peripheral neurons known to be involved in nociception, and are known to affect pain (Yang and Gereau, 2002, 2003; Marabese et al., 2007; Chiechio et al., 2010; Jaggi and Singh, 2012; Mazzitelli et al., 2018; Pereira and Goudet, 2018). In particular, inflammatory pain is attenuated by Group II mGluRs (mGluR2 and mGluR3) and Group III mGluRs (mGlurR4, mGluR7, and mGluR8; Zammataro et al., 2014; Mazzitelli et al., 2018; Pereira and Goudet, 2018), and is enhanced or modulated by Group I mGluRs (mGluR1 and mGluR5; Bhave et al., 2001; Karim et al., 2001; Zammataro et al., 2014; Radwani et al., 2017; Mazzitelli et al., 2018; Pereira and Goudet, 2018; Masuoka et al., 2020). Thus, metabotropic glutamate receptors (mGluRs), especially Groups II and III, are promising therapeutic targets for CIPN-induced neuropathic pain (Mazzitelli et al., 2018; Pereira and Goudet, 2018).

Meclizine is also an attractive candidate for reducing CIPN caused by platinum-based drugs. Meclizine is an H1 receptor antagonist and a pregnane X receptor agonist in mice shown to be neuroprotective in animal models of conditions as diverse as Huntington's (Gohil et al., 2011), Parkinson's (Hong et al., 2016), and hypoxia (Zhuo et al., 2016). Meclizine protects cultured DRG primary sensory neurons from cisplatin-induced damage by enhancing pentose phosphate pathway activity, enhancing NADPH production, and improving clearance of DNA damage (Gorgun et al., 2017). Furthermore, meclizine shifts retinal ganglion cells toward the glycolysis metabolism pathway and away from the mitochondrial respiratory pathway (J. Zhu et al., 2020), suggesting that meclizine could protect against cisplatin-induced oxidative stress and mitochondrial toxicity.

In this study, we used mechanical and thermal pain behavioral assays and confocal microscopic imaging with the genetically-encoded fluorescent calcium indicator, Pirt-GCaMP3, to investigate cisplatin-based CIPN-induced neuropathic pain in vivo at the level of an intact ensemble comprising the large DRG primary sensory neuronal population. We found that cisplatin induced mechanical and thermal hypersensitivity in pain behavioral assays, increased levels of spontaneous Ca²⁺ activity in DRG primary sensory neurons, and increased Ca²⁺ responses to mechanical, thermal, and chemical stimuli. mGluR8 agonist (S)−3,4-DCPG, Group II mGluRs (mGluR2 and mGluR3) agonist LY379268, and histamine H1 receptor antagonist, meclizine, each attenuated cisplatin-induced pain in behavioral assays. Furthermore, LY379268 and meclizine attenuated spontaneous calcium activity in DRG neurons in vivo. Finally, meclizine and LY379268 reduced cisplatin-induced weight loss.

Materials and Methods

Animals

All experiments were performed in compliance with policies and procedures of the Institutional Animal Care and Use Committee at University of Texas Health Science Center at San Antonio (UTHSA) and in accordance with National Institutes of Health and American Association for Accreditation of Laboratory Animal Care. Pirt-GCaMP3 mice (Kim et al., 2014, 2016) on a C57BL/6J background or C57BL/6J mice (The Jackson Laboratory) were used in all experiments. Pirt-GCaMP3 mice used in experiments were all heterozygous. Both males and females were used. All animals were at least eight weeks old, were kept in a 14/10 h light/dark cycle, and were provided ad libitum access to food and water.

Drugs and drug treatment

Cisplatin, LY379268, and (S)−3,4-DCPG were purchased from Abcam. Meclizine, HC 030031, AMG 333, NF 110, and AMG 9810 were purchased from Tocris. Cisplatin stocks were dissolved in 1-methyl-2-pyrrolidinone at 8 mg/ml (Sigma-Aldrich). LY379268, (S)−3,4-DCPG, and NF 110 stocks were dissolved in water. Meclizine stocks were dissolved in DMSO. HC 030031, AMG 333, and AMG 9810 were suspended at 50 µm in water employing repeated vortexing and freeze-thaw cycles as necessary until no pellets could be observed after centrifuging at 18,000 rcf. Aliquots of all drugs except cisplatin were stored at −80°C until use. Cisplatin stock was kept at −20°C until use. All drugs were added to 0.9% saline immediately before injection, intraperitoneally. Meclizine was added to water immediately before injection, intraperitoneally. On days that both behavioral testing and injection occurred, animals were weighed before behavioral testing. Mouse weights were used to calculate drug dosage. Injections were performed after behavioral testing. For drugs used in imaging experiments, aliquots from 50 µm stock were vortexed vigorously and diluted to 10 µm final concentration in PBS. Isoflurane was purchased from the Piramal Group (Mumbai, India). Pentobarbital was purchased from Diamondback Drugs. Cisplatin and saline vehicle injections were performed on days 1, 3, 5, and 7. In experiments using multiple rounds of cisplatin injection, cisplatin was also injected on days 12, 14, 16, and 18. When (S)−3,4-DCPG or LY379268 were concomitantly administered to animals receiving cisplatin, injections were given 30 min before cisplatin. When meclizine was concomitantly administered with cisplatin, injections were given 3 h before cisplatin. In imaging experiments where drugs were applied directly to the DRG, the order of drug administration was HC 030031, AMG 333, NF 110, and AMG 9810.

Behavioral tests

von Frey mechanical test

Mice were placed in a 4.5 × 5 × 10 cm transparent container on a metal mesh and allowed to habituate for at least 30 min before testing. Each mouse was tested at a specific force eight times to determine the lowest force required to elicit a paw withdrawal response more than 50% of the time.

Hot plate thermal test

Mice were placed in a 4.5 × 5 × 10 cm transparent container on a temperature controlled hot plate set to 45°C. Latency to acute nocifensive behavior was determined by onset of hindpaw lifts, licking, jumping, or flinching.

DRG exposure surgery

L5 DRG exposure surgery and imaging was performed as previously described (Kim et al., 2016). Mice were anesthetized with pentobarbital (40–50 mg/kg). After deep anesthesia was achieved, the animal's back was shaved, and the shaved skin was aseptically prepared by cleaning with alcohol and iodine pad, and ophthalmic ointment was applied to keep the eyes moist (Lacrilube, Allergan Pharmaceuticals). Mice were kept on a heating pad and monitored with a rectal thermometer to maintain body temperature at 37 ± 0.5°C.

Dorsal laminectomy was performed in the L4–L6 area. A 2-cm midline incision was made in the lower back around the lumbar enlargement area. Paravertebral muscles were dissected away to expose the lower lumbar enlargement area, and the bones were cleaned. Small rongeurs were used to remove the surface aspect of the L5 DRG transverse process bone near the vertebra to expose the DRG without disrupting the neurons, axons, and other cells in the DRG.

In vivo DRG GCaMP Ca²⁺ imaging

In vivo GCaMP Ca²⁺ imaging of the large DRG was performed over 1–6 h following exposure surgery on day 8 after the last of four drug injections on days 1, 3, 5, and 7. Mice were kept on a heating pad and monitored with a rectal thermometer to maintain body temperature at 37 ± 0.5°C. Mice were laid abdomen down on a custom-designed imaging stage. Movement from breathing, heart beats, etc., was minimized by holding the head in a custom designed holder with an anesthesia/gas mask and custom designed vertebral clamps. Continuous anesthesia was maintained with 1−2% isoflurane in pure oxygen.

For imaging and analysis, the stage was fixed under a single photon confocal microscope (Carl Zeiss AG). Raw image stacks (512 × 512–1024 × 1024 pixels in the x-y plane and 20- to 30-µm voxel depth) were converted into time lapse movies (∼6.5–7 s were required to produce a single frame) and analyzed using Zeiss Zen 3.1 Blue Edition software (Carl Zeiss AG).

Putative responding cells were identified by visual observation of raw image time lapse movies. Ca²⁺ transient intensities were calculated by ΔF/F₀ = (F_t – F₀)/F₀, where F_t is the pixel intensity in a region of interest (Cioroiu and Weimer) at the time point of interest, and F₀ is the baseline intensity determined by averaging the intensities of the first two to six frames of the ROI in the experiment. For calcium responses to stimuli that produced multiple transients, only the first peak was analyzed. Cells showing calcium transients before the stimulus were assumed to be spontaneously active and were not included in analysis of stimuli-induced transients. Each peak of spontaneous Ca²⁺ transients was analyzed individually. Suitable transients (not too much movement of images, clear baseline fluorescence, no nearby cells with Ca²⁺ transients) were randomly sampled for analysis. An effort was made to sample the same number of cells from each DRG for calculating average and SEM for area under the curve (AUC) and ΔF/F₀ intensity; no more than 40% of cells came from a single DRG.

Stimuli were applied carefully so as to not cause movement during imaging. Brush stimuli were applied by repeated gentle brushing from heel to toes approximately once per second using small (5 mm) and large (40 mm) brush bristles. Press (SMALGO Algometer, Bioseb Instruments) and von Frey filaments were applied directly to the hindpaw ipsilateral to the DRG being imaged for 15–20 s. Thermal stimuli were applied by immersing the hindpaw in water at a specific temperature (0°C or 45°C) for 15–20 s. Stimuli were applied after 35–40 s of baseline imaging. HC 030031, AMG 333, NF 110, and AMG 9810 were applied during imaging by topically pipetting onto DRG neurons and allowed to sit for at least 5 min. Between drug applications, neurons were rinsed three to four times in PBS, and the final wash was allowed to sit on the DRG for at least 5 min.

Statistical analysis

Statistics were performed on GraphPad Prism 9.0.1. Ca²⁺ transient intensities and numbers of activated cells showing spontaneous Ca²⁺ activity or Ca²⁺ activity following stimuli were analyzed by Student's t test or one-way or two-way ANOVA followed by post hoc Tukey's test or post hoc Dunnett's test, as appropriate. Specific statistical tests performed are indicated in the text of the results and figure and table legends.

Results

Cisplatin induces mechanical and thermal hyperalgesia, increases spontaneous Ca²⁺ activity in the DRG neurons, and increases the sensitization of DRG neurons to mechanical and thermal stimuli

To determine the effects of cisplatin on DRG Ca²⁺ signaling, animals were injected with cisplatin (3.5 mg/kg, i.p.) every other day for four total injections (Deng et al., 2012). Compared with saline-treated controls, cisplatin-injected animals developed mechanical hyperalgesia after the second injection of cisplatin (Fig. 1A, Movie 1; drug effect F_(1,48) = 475.7, p < 0.001; time effect F_(3,48) = 34.4, p < 0.001; interaction F_(3,48) = 40.43, p < 0.001; n = 7 per group; Tukey's df = 48, day 4 q = 17.52, p < 0.001; day 8 q = 20.88, p < 0.001; day 11 q = 21.12, p < 0.001) and thermal hyperalgesia after the fourth cisplatin injection (Fig. 1B; drug effect F_(1,48) = 35.32, p < 0.001; time effect F_(3,48) = 15.69, p < 0.001; interaction F_(3,48) = 5.378, p = 0.0028; n = 7 per group; Tukey's df = 48, day 8 q = 6.722, p = 0.0005; day 11 q = 6.974, p = 0.0003). Ca²⁺ imaging experiments were performed as previously described (Kim et al., 2016). Animals were anesthetized and DRGs were surgically exposed and imaged. In cisplatin-injected mice, we found significant increases in the numbers of cells with spontaneous Ca²⁺ oscillation (t₍₁₈₎ = 5.150, p = 0.000067, n = 11 saline and 9 cisplatin) and in total spontaneous Ca²⁺ activity (total of spontaneous Ca²⁺ oscillation and steady-state high Ca²⁺ levels; t₍₁₈₎ = 4.380, p = 0.000361, n = 11 saline and 9 cisplatin) compared with saline-injected mice, but no significant increase in the number of cells exhibiting steady-state high Ca²⁺ activity compared with saline-injected mice (t₍₁₈₎ = 1.800, p = 0.088631, n = 11 saline and 9 cisplatin; Fig. 1C, Movie 2), indicating that the increase was primarily because of spontaneous Ca²⁺ oscillation. The numbers of DRG neurons responding to press (100, 300, and 600 g), 10-g von Frey, and 45°C stimuli were increased compared with saline controls (Figs. 1D, 2D, 2G, Movie 1; 100 g t₍₁₅₎ = 3.457, p = 0.003523, n = 8 saline and 9 cisplatin; 300 g t₍₆₎ = 3.179, p = 0.01910, n = 4 each; 600 g t₍₆₎ = 2.945, p = 0.02,577, n = 4 each; 10 g t₍₁₈₎ = 2.386, p = 0.02,825, n = 11 saline and 9 cisplatin; 45°C t₍₁₅₎ = 4.625, p = 0.000330, n = 11 saline and 9 cisplatin).

Figure 1.

Cisplatin treatment causes increased sensitivity to pain, increases the number of DRG neurons showing Ca²⁺ activity spontaneously or in response to press stimuli and alters Ca²⁺ transient fluorescence intensities and areas under the curve to press stimuli. A, Mechanical pain sensitivity tests using von Frey filaments were performed on days 0, 4, 8, and 11. Red arrows indicate saline or cisplatin injection (days 1, 3, 5, and 7). Mechanical sensitivity is plotted as the 50% withdrawal threshold in grams. B, Thermal pain sensitivity tests using the hot plate assay were performed on days 0, 4, 8, and 11. Arrows indicate day of cisplatin or vehicle (saline) injection (days 1, 3, 5, and 7). Thermal sensitivity is plotted as paw withdrawal latency in seconds. C, DRGs were imaged for spontaneous Ca²⁺ activity following injection of cisplatin or saline in the absence of stimuli. Graphs show number of cells with Ca²⁺ oscillation, steady-state high Ca²⁺, and total number of cells with Ca²⁺ activity. D, DRGs were imaged during paw press at forces of 100, 300, and 600 g. Graphs show cells producing Ca²⁺ transients in response to press stimuli. Ea, Representative maximum intensity Z-projection images of spontaneous DRG neuronal Ca²⁺ activity in saline-injected and cisplatin-injected animals. Imaged DRGs are shown on left. Area outlined by red boxes are magnified on the right. Yellow circles in Ea, Fa, Ga, Ha indicate Ca²⁺ activity of neurons in response to different stimuli. Eb, Graphs of spontaneous DRG neuronal Ca²⁺ transients shown with maximum fluorescence intensity set at 0 s. Y-axis is expressed as ΔF/F₀ fluorescence intensity. Saline n = 5 DRGs, 28 cells, 85 transient peaks. Cisplatin n = 5 DRGs, 90 cells, 189 transient peaks. Ec, Mean ΔF/F₀ fluorescence intensities of spontaneous Ca²⁺ transients shown as a function of time. Left y-axis, AUC; arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. Fa, Ga, Ha, Representative maximum intensity Z-projection images of Ca²⁺ activity in response to 100-g (Fa), 300-g (Ga), or 600-g (Ha) press stimulus among DRG neurons in saline-injected and cisplatin-injected animals. All panels (and subpanels) E–H are sequentially derived from a single DRG neuronal ensemble from either a saline-injected animal or a cisplatin-injected animal. Areas outlined by red boxes are magnified on the right. Fb, Gb, Hb, Graphs of Ca²⁺ transients in response to press stimuli. Arrow indicates start of press stimulus. Y-axis, ΔF/F₀ fluorescence intensity. Saline, n = 3, 3, and 3 DRGs; n = 88, 105, and 48 cells for 100, 300, and 600 g, respectively. Cisplatin n = 3, 3, and 3 DRGs; n = 51, 189, and 186 cells for 100, 300, and 600 g, respectively. Fc, Gc, Hc, Mean fluorescence intensities of Ca²⁺ transients in response to press stimuli. Left y-axis, AUC; arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity for each frame. All graphs with error bars show mean ± SEM. Transient plots show all individual analyzed cells in gray lines and mean ΔF/F₀ in red. For behavior, animal numbers are saline, n = 14 (8 males, 6 females), cisplatin, n = 13 (8 males, 5 females). Numbers of DRGs analyzed for spontaneous activity, 100-, 300-, and 600-g press were saline, n = 11, 8, 4, and 4, respectively, cisplatin, n = 9, 9, 4, and 4, respectively. Please see further detail in Tables 1–3 and in Figure 4C. Comparisons of numbers of cells showing Ca²⁺ transients were performed using Student's t test. Comparisons of fluorescence intensities of Ca²⁺ transients were performed with two-way ANOVA followed by Tukey's post hoc tests; saline versus cisplatin *p < 0.05, **p < 0. 01, ***p < 0.001.

Figure 2.

Cisplatin treatment increases the number of Ca²⁺-activated DRG neurons responding to thermal and strong von Frey stimuli and alters Ca²⁺ transient fluorescence intensities and AUC in response to thermal stimuli. A, D, G, DRGs were imaged during stimulation using small brush and large brush (A), paw immersion in 0°C and 45°C water (D), and 0.07- and 10-g von Frey filaments (G). Graphs show cells producing Ca²⁺ transients in response to different stimuli. Ba, Ca, Ea, Fa, Ha, Ia, Representative images (maximum intensity Z-projections from saline-injected and cisplatin-injected animals) of DRG neuronal Ca²⁺ activity in response to stimuli: small brush (Ba), large brush (Ca), 0°C cold (Ea), 45°C hot (Fa), 0.07-g von Frey (Ha), and 10-g von Frey (Ia). All panels (and subpanels) B, C, E, F, H, I are sequentially derived from a single DRG neuronal ensemble from either a saline-injected animal or a cisplatin-injected animal. Areas outlined by red boxes are magnified on the right. Yellow circles track Ca²⁺ activity of two sets of DRG neurons (in saline-injected mice vs cisplatin-injected mice) through different stimuli. Bb, Cb, Eb, Fb, Hb, Ib, Graphs of Ca²⁺ transients in response to various stimuli. Arrow indicates start of stimulus. Y-axis, ΔF/F₀ fluorescence intensity. Bc, Cc, Ec, Fc, Hc, Ic, Mean fluorescence intensities of Ca²⁺ transients in response to various stimuli. Left y-axis, AUC; arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity for each frame. Numbers of DRGs analyzed for cell counts for small brush, large brush, 0°C water, 45°C water, 0.07-g von Frey, and 10-g von Frey were: saline, n = 7, 6, 5, 9, 6, and 11, respectively; cisplatin, n = 4, 5, 4, 8, 5, and 9, respectively. Numbers of cells analyzed for Ca²⁺ transients for small brush, large brush, 0°C water, 45°C water, 0.07-g von Frey, and 10-g von Frey were: saline, n = 10, 11, 27, 61, 7, and 45, respectively; cisplatin, n = 10, 16, 40, 51, 8, and 55, respectively. Please see further detail in Tables 2 and 3 and in Figure 4C. All graphs with error bars show mean ± SEM. Transient plots show all individual analyzed cells in gray lines and mean ΔF/F₀ in red. Comparisons of numbers of cells showing Ca²⁺ transients were performed using Student's t test. Comparisons of fluorescence intensities of Ca²⁺ transients were performed with two-way ANOVA followed by Tukey's post hoc tests; saline versus cisplatin *p < 0.05, **p < 0.01, ***p < 0.001.

Movie 1.

Cisplatin dramatically increases the number of DRG neurons responding to mechanical stimulus. Representative movies generated from microscope images compare DRGs during 100-g press between saline-treated and cisplatin-treated animals. In response to a non-noxious press (100 g), an average of 42.38 cells produced Ca²⁺ transients in saline-treated animals (A), and an average of 88.44 cells did so in cisplatin-treated animals (B). Arrows show examples of activating neurons.

Movie 1B.

Cisplatin dramatically increases the number of DRG neurons responding to mechanical stimulus. Representative movies generated from microscope images compare DRGs during 100-g press between saline-treated and cisplatin-treated animals. In response to a non-noxious press (100 g), an average of 42.38 cells produced Ca²⁺ transients in saline-treated animals (A), and an average of 88.44 cells did so in cisplatin-treated animals (B). Arrows show examples of activating neurons.

Movie 2.

Cisplatin increases the number of cells showing spontaneous Ca²⁺ transients. Representative movies generated from microscope images compare DRGs in the absence of stimuli. An average of 2.39 cells spontaneously produced Ca²⁺ transients in saline-treated animals (A) and an average of 17.87 cells produced Ca²⁺ transients in cisplatin-treated animals (B). Arrows show examples of activating neurons.

Movie 2B.

Cisplatin increases the number of cells showing spontaneous Ca²⁺ transients. Representative movies generated from microscope images compare DRGs in the absence of stimuli. An average of 2.39 cells spontaneously produced Ca²⁺ transients in saline-treated animals (A) and an average of 17.87 cells produced Ca²⁺ transients in cisplatin-treated animals (B). Arrows show examples of activating neurons.

Stimulus-induced Ca²⁺ transients were monitored in saline-treated and cisplatin-treated animals. The numbers of cells showing spontaneous Ca²⁺ activity and Ca²⁺ transients in response to stimuli were higher in DRGs of cisplatin-treated animals than in DRGs of saline-treated animals (F_(1,47) = 49.21, p < 0.001 by two-way ANOVA). Average AUC of Ca²⁺ transients of cisplatin group was decreased by a large number of low-amplitude Ca²⁺ transients (many neurons are activated; Fig. 4C). Analysis of spontaneous and press-induced Ca²⁺ transients showed that spontaneous Ca²⁺ transients in the DRG of saline-treated controls took longer to return to baseline than in cisplatin-treated animals (Fig. 1Eb,Ec; drug effect F_(1,684) = 17.38, p < 0.001; time effect F_(2,684) = 147.6, p < 0.001; interaction F_(2,684) = 1.207, p = 0.2997; n = 43 saline, 189 cisplatin; Tukey's df = 684, 0 s q = 4.256, p = 0.0323; 7 s q = 1.621, p = 0.8617; 14 s q = 6.263, p = 0.0002). At 100- and 300-g press, cisplatin-treated animals had greater AUC and were slower to return to baseline than saline-treated animals (Fig. 1Fb,Fc,Gb,Gc; 100 g, AUC t₍₂₃₃₎ = 3.533, p = 0.000499; drug effect F_(1,1068) = 16.88, p < 0.001; time effect F_(5,1068) = 13.33, p < 0.001; interaction F_(5,1068) = 2.788, p = 0.0165; n = 88 saline, 137 cisplatin; Tukey's df = 1068, 7 s q = 2.276, p = 0.9057; 14 s q = 6.274, p = 0.0006; 21 s q = 7.668, p < 0.0001; 28 s q = 0.5069, p > 0.9999; 35 s q = 6.544, p = 0.0003; 300 g, AUC t₍₂₉₂₎ = 3.878, p = 0.000130; drug effect F_(1,1554) = 15.04, p = 0.001; time effect F_(5,1554) = 48.06, p < 0.001; interaction F_(5,1554) = 1.779, p = 0.1140; n = 105 saline, 189 cisplatin; Tukey's df = 1554, 7 s q = 3.724, p = 0.2619; 14 s q = 5.281, p = 0.0105; 21 s q = 4.777, p = 0.0360; 28 s q = 6.345, p 0.0005). Ca²⁺ transients at 600-g press were not significantly different between saline-treated and cisplatin-treated animals (Fig. 1H; drug effect F_(1,1073) = 1.277, p = 0.2587; time effect F_(5,1073) = 29.62, p < 0.0001; interaction F_(5,1073) = 1.066, p = 0.3774). Compared with Ca²⁺ transients in DRG neurons of saline-treated mice, cisplatin treatment increased AUC of Ca²⁺ transients in response to small brush stimulus and decreased AUC of Ca²⁺ transients in response to paw immersion in 45°C water (Fig. 2Bb,Bc,Fb,Fc; small brush AUC t₍₁₈₎ = 6.069, p = 0.00001; drug effect F_(1,104) = 23.54, p < 0.0001; time effect F_(5,104) = 498.9, p < 0.0001; interaction F_(5,104) = 123.8, p < 0.0001; n = 10 saline, 10 cisplatin; Tukey's df = 104, 7 s q = 34.87, p < 0.0001; 14 s q = 5.798, p = 0.0045; 45°C drug effect F_(1,623) = 5.675, p = 0.0175; time effect F_(5,623) = 62.38, p < 0.0001; interaction F_(5,623) = 2.559, p = 0.0264; n = 61 saline, 51 cisplatin; Tukey's df = 623, 7 s q = 0.511, p > 0.9999; 14 s q = 5.185, p = 0.0141). Compared with saline-treated controls, more Ca²⁺ transients were delayed in cisplatin-treated animals in response to large brush stimulus (Fig. 2Cc; drug effect F_(1,136) = 0.5561, p = 0.4571; time effect F_(5,136) = 37.74, p < 0.0001; interaction F_(5,136) = 10.31, p < 0.0001; n = 11 saline, 16 cisplatin; Tukey's df = 136, 7 s q = 2.479, p = 0.8400; 14 s q = 2.726, p = 0.7398; 21 s q = 7.500, p < 0.0001; 28 s q = 5.980, p = 0.0024; 35 s q = 3.181, p = 0.5175). Amplitude of Ca²⁺ transients in response to paw immersion in 45°C water was significantly higher in saline-treated animals than in cisplatin-treated animals (Fig. 2Fc; AUC t₍₁₁₀₎ = 2.568, p = 0.01156; drug effect F_(1,623) = 5.675, p = 0.0175; time effect F_(5,623) = 62.38, p < 0.0001; interaction F_(5,623) = 2.559, p = 0.0264; n = 61 saline, 51 cisplatin; Tukey's df = 623, 7 s q = 0.5110, p > 0.9999; 14 s q = 5.815, p = 0.0141; 21 s q = 3.223, p = 0.4928; 28 s q = 0.4958, p > 0.9999; 35 s q = 0.9730, p > 0.9999). Paw immersion in 0°C water, or paw stimulus with 0.07- and 10-g von Frey, did not show significant differences between Ca²⁺ transients of two groups (Fig. 2Eb,Ec,Hb,Hc,Ib,Ic; drug effect 0°C F_(1,364) = 0.9285, p = 0.3359; 0.07 g F_(1,73) = 0.3066, p = 0.5815; 10 g F_(1,525) = 1.199, p = 0.2740).

Figure 4.

mGluRs 2/3 agonist, LY379268, reduces mechanical and thermal pain sensitivity in cisplatin-treated animals and reduces DRG neuronal Ca²⁺ activity in response to thermal stimulus. A, B, Mechanical pain sensitivity tests using von Frey filaments and thermal pain sensitivity tests using hot plate were performed before injection and on days 8, 12, and 19 following the start of injections. Red arrows indicate injection of cisplatin (days 3, 5, 14, 16). Blue arrows indicate injection of LY379268 30 min before cisplatin injection. Mechanical sensitivity is plotted as 50% withdrawal threshold (grams). Thermal sensitivity is plotted as paw withdrawal latency (seconds). Numbers of animals were: saline, n = 12 (6 males, 6 females); cisplatin, n = 12 (6 males, 6 females); LY379268, n = 7 (3 males, 4 females). C, AUC of Ca²⁺ transients of individual spontaneous oscillating neurons. Groups are saline controls, cisplatin, cisplatin plus DCPG, cisplatin plus LY379268, and cisplatin plus meclizine. AUCs are shown. Except panel C, graphs show mean ± SEM. D, DRGs were imaged for spontaneous Ca²⁺ activity in the absence of stimuli and during stimulation of the hindpaw with 45°C, 10-g von Frey, and 100-g press. Graphs show numbers of cells producing Ca²⁺ transients spontaneously, steady-state high cytosol Ca²⁺, total spontaneous Ca²⁺ activity, and activity in response to stimuli. Ea, Representative images (maximum intensity Z-projections) of spontaneous DRG neuronal Ca²⁺ activity in cisplatin-injected, or cisplatin plus LY379268-injected animals. Imaged DRGs are shown on left. Areas outlined by red boxes are magnified on the right. Yellow circles indicate spontaneous Ca²⁺ activity and Ca²⁺ activity following various stimuli in later panels. Eb, Graphs of spontaneous DRG neuronal Ca²⁺ transients are shown with the first frame of fluorescence intensity increase set at 0 s. Y-axis, ΔF/F₀ fluorescence intensity. Cisplatin, n = 5 DRGs, 90 cells, 189 transient peaks. LY379268, n = 3 DRGs, 21 cells, 51 transient peaks. Ec, Mean fluorescence intensities of spontaneous DRG neuronal Ca²⁺ transients as a function of time frame. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. Fa, Ga, Ha, Representative images (maximum intensity Z-projections) of Ca²⁺ transients in response to different stimuli: 45°C hot (Fb), 10-g von Frey (Gb), and 100-g press (Hb) of cisplatin-injected and cisplatin plus LY379268-injected animals. Areas outlined by red boxes are magnified on the right. Yellow circles indicate DRG neurons activated by different stimuli. Fb, Gb, Hb, Graphs of DRG neuronal Ca²⁺ transients in response to different stimuli: 45°C hot (Fb), 10-g von Frey (Gb), and 100-g press (Hb). Red arrow indicates start of each stimulus. Y-axis, ΔF/F₀ fluorescence intensity. Fc, Gc, Hc, Mean fluorescence intensities of DRG neuronal Ca²⁺ transients in response to various stimuli. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. All graphs with error bars show mean ± SEM. Transient plots show all individually analyzed cells in gray lines and mean ΔF/F₀ in red. Numbers of DRGs analyzed for 45°C water, 10-g von Frey, and 100-g press were, respectively: saline, n = 9, 11, and 8: cisplatin, n = 8, 9, and 9; cisplatin plus LY379268, n = 3, 3, and 3. Numbers of cells analyzed for 45°C water, 10-g von Frey, and 100-g press were, respectively: cisplatin, n = 51, 55, and 137 and LY379268, n = 66, 51, and 255. Please see further detail in Tables 2 and 3. Behavioral, AUC, and Ca²⁺ transient intensity comparisons of saline-injected and LY379268 plus cisplatin-injected animals versus cisplatin-injected animals are two-way ANOVA followed by Dunnett's post hoc test. Cell count comparisons of saline-injected and LY379268 plus cisplatin-injected versus cisplatin-injected animals were performed by one-way ANOVA followed by Dunnett's post hoc test. For behavioral tests: saline versus cisplatin *p < 0.05, **p < 0.01, ***p < 0.001; cisplatin versus cisplatin plus LY379268 #p < 0.05, ##p < 0.01, ###p < 0.001, for cell counts and Ca²⁺ transients: cisplatin versus cisplatin plus LY379268 *p < 0.05, **p < 0.01, ***p < 0.001.

The relative proportions of small, medium, and large diameter neurons that exhibited spontaneous Ca²⁺ oscillation and response to stimuli were indistinguishable between cisplatin-treated and saline-treated mice. Nonetheless, following cell activation, spontaneous Ca²⁺ transients and Ca²⁺ transients following most stimuli predominantly occurred in small diameter neurons (<20 µm), followed by medium diameter neurons (20–25 µm), and finally, relatively few large diameter neurons (>25 µm). However, more large diameter neurons were activated by stronger mechanical stimuli (10-g von Frey, 300- and 600-g press) than by other stimuli, albeit still a lower proportion than small and medium diameter neurons (Tables 1, 2).

Table 1.

Comparisons of cell diameter of DRG neurons responding to press stimuli in saline-treated controls and cisplatin-treated animals

Press responses
Diameter	Saline	Cisplatin
100-g pinch: number of activated cells (proportion of activated cells)
Small (<20 µm)	22.00 ± 4.51, n = 3 (55.9 ± 8.4%)	46.33 ± 7.75, n = 3 (67.6 ± 3.8%)
Medium (20–25 µm)	12.67 ± 3.53, n = 3 (33.6 ± 8.7%)	17.67 ± 2.40, n = 3 (25.2 ± 0.2%)
Large (>25 µm)	4.00 ± 0.58, n = 3 (10.5 ± 2.0%)	4.33 ± 2.19, n = 3 (7.2 ± 3.7%)
300-g pinch: number of activated cells (proportion of activated cells)
Small	40.33 ± 5.90, n = 3 (53.4 ± 6.7%)	52.00 ± 12.12, n = 3 (51.9 ± 12.7%)
Medium	23.67 ± 2.03, n = 3 (31.6 ± 3.4%)	29.67 ± 6.12, n = 3 (29.4 ± 18.7%)
Large	11.33 ± 2.33, n = 3 (15.0 ± 3.6%)	19.00 ± 8.62, n = 3 (18.7 ± 7.9%)
600-g pinch: number of activated cells (proportion of activated cells)
Small	49.00 ± 15.72, n = 3 (51.2 ± 7.5%)	64.00 ± 7.77, n = 3 (49.5 ± 5.6%)
Medium	26.33 ± 2.96, n = 3 (31.4 ± 5.6%)	41.67 ± 2.03, n = 3 (32.0 ± 2.1%)
Large	15.00 ± 2.52, n = 3 (17.4 ± 3.1%)	23.67 ± 5.36, n = 3 (18.4 ± 4.3%)

Most activated neurons were small (<20 µm) or medium (20–25 µm) diameter. Strong mechanical stimuli (300- and 600-g press) activated a higher proportion of large diameter (>25 µm) neurons. Proportions of small versus medium versus large diameter neurons were not statistically distinguishable between groups. N, numbers of DRGs analyzed for cell diameter.

Table 2.

Comparisons of cell diameter of DRG neurons responding to different stimuli in saline-treated controls and cisplatin-treated, cisplatin plus (S)−3,4-DCPG-treated, and cisplatin plus meclizine-treated animals.

Spontaneous and stimulus responses following pharmacological treatment
Diameter	Saline	Cisplatin	Cisplatin + DCPG	Cisplatin + meclizine
Spontaneous Ca2+ activation
Small	3.80 ± 1.20, n =s 5	13.17 ± 2.77, n = 6	15.67 ± 13.67, n = 3	5.40 ± 1.60, n = 5
Medium	1.80 ± 0.58, n = 5	2.67 ± 0.67, n = 6	1.33 ± 0.88, n = 3	2.20 ± 1.32, n = 5
Large	0.00 ± 0.00, n = 5	1.67 ± 0.71, n = 6	0.00 ± 0.00, n = 3	0.80 ± 0.58, n = 5
0°C cold
Small	12.67 ± 2.73, n = 3	17.33 ± 1.20, n = 3	29.33 ± 1.20, n = 3	10.33 ± 4.84, n = 3
Medium	1.33 ± 0.33, n = 3	1.33 ± 1.33, n = 3	3.67 ± 1.45, n = 3	2.33 ± 0.88, n = 3
Large	0 ± 0.00, n = 3	0 ± 0.00, n = 3	1.00 ± 0.58, n = 3	0.00 ± 0.00, n = 3
45°C hot
Small	18.67 ± 2.03, n = 3	34.33 ± 10.33, n = 3	29.00 ± 7.23, n = 3	17.33 ± 3.84, n = 3
Medium	6.33 ± 0.88, n = 3	10.33 ± 0.33, n = 3	5.33 ± 1.20, n = 3	3.67 ± 1.86, n = 3
Large	0 ± 0.00, n = 3	0 ± 0.00, n = 3	0.33 ± 0.33, n = 3	0.33 ± 0.33, n = 3
0.07-g von Frey
Small	1 ± 0.00, n = 3	3.33 ± 0.33, n = 3	1.8 ± 0.58, n = 5	3.00 ± 2.52, n = 3
Medium	1.33 ± 0.67, n = 3	0.33 ± 0.33, n = 3	0.2 ± 0.20, n = 5	0.33 ± 0.33, n = 3
Large	0 ± 0.00, n = 3	0 ± 0.00, n = 3	0 ± 0.00, n = 5	0.33 ± 0.33, n = 3
10-g von Frey
Small	3.67 ± 2.33, n = 3	11 ± 3.06, n = 3	22.67 ± 7.31, n = 3	11.33 ± 4.41, n = 3
Medium	8.33 ± 4.37, n = 3	7.67 ± 1.33, n = 3	6.00 ± 1.53, n = 3	6.33 ± 2.19, n = 3
Large	4.00 ± 2.08, n = 3	2.67 ± 1.76, n = 3	5.00 ± 1.15, n = 3	3.67 ± 1.45, n = 3

Most activated neurons were small (<20 µm) or medium (20–25 µm) diameter. Strong mechanical stimulus (10-g von Frey) tended to activate a higher proportion of large diameter (>25 µm) neurons. N, number of DRGs analyzed for cell diameter.

Cisplatin-induced hyperalgesia is attenuated by concomitant administration of mGluR8 agonist, (S)−3,4-DCPG

Since mGluR8 agonists are known to attenuate some kinds of inflammatory pain (Marabese et al., 2007; Palazzo et al., 2011), we tested mGluR8 agonist, (S)−3,4-DCPG, in the cisplatin-induced pain model by injecting (S)−3,4-DCPG (30 mg/kg, i.p.) 30 min before each cisplatin injection. (S)−3,4-DCPG attenuated cisplatin-induced mechanical and thermal hyperalgesia (Fig. 3A,B; mechanical drug effect F_(2,221) = 463.7, p < 0.0001; time effect F_(12,221) = 6.148, p < 0.0001; interaction F_(24,221) = 3.563, p < 0.001; n = 7 saline, 7 cisplatin, 6 cisplatin plus (S)−3,4-DCPG; Tukey's df = 221 cisplatin vs cisplatin plus (S)−3,4-DCPG df = 221, day 0 q = 1.120, p = 0.7085; day 4 q = 1.120, p < 0.0001; cisplatin vs saline day 0 q = 1.311, p = 0.6238, day 4 q = 10.63, p < 0.0001; thermal drug effect F_(2,234) = 40.91, p < 0.0001; time effect F_(12,234) = 8.010, p < 0.001; interaction F_(24,234) = 2.424, p = 0.0004; n = 6 saline, 6 cisplatin, 9 cisplatin + (S)−3,4-DCPG; Tukey's cisplatin vs cisplatin plus (S)−3,4-DCPG df = 234, day 0 q = 1.007, p = 0.7566; day 8 q = 4.964, p = 0.0016). However, neither the number of activated neurons showing spontaneous Ca²⁺ activity nor the number of activated neurons responding to 45°C, 0.07-g von Frey, 10-g von Frey, or 100-g press stimuli were different between cisplatin-treated and cisplatin plus (S)−3,4-DCPG-treated animals (Fig. 3C; 45°C t₍₁₆₎ = 1.448, p = 0.1696, n = 8 each; 0.07 g t₍₈₎ = 1.177, p = 0.2731, n = 5 each; 10 g t₍₁₅₎ = 1.305, p = 0.2114, n = 9 cisplatin, 8 cisplatin plus (S)−3,4-DCPG; 100 g t₍₁₁₎ = 0.4550, p = 0.6580, n = 9 cisplatin, 4 cisplatin + (S)−3,4-DCPG). The number of activated neurons following 0°C cold stimulus was increased in cisplatin plus (S)−3,4-DCPG animals compared with cisplatin-only controls (Fig. 3C; t₍₅₎ = 4.845, p = 0.04903, n = 4 cisplatin, 3 cisplatin plus (S)−3,4-DCPG). (S)−3,4-DCPG slowed spontaneous decay time of Ca²⁺ transients (return to baseline; Fig. 3Db,Dc; drug effect F_(1,655) = 34.58, p < 0.001; time effect F_(2,684) = 454.8, p < 0.001; interaction F_(2,684) = 1.987, p = 0.1379; n = 189 cisplatin, 33 (S)−3,4-DCPG + cisplatin; Tukey's df = 655, 0 s q = 3.889, p = 0.0671; 7 s q = 7.092, p < 0.001; 14 s q = 3.425, p = 0.1502). (S)−3,4-DCPG also increased the neuronal number of delayed Ca²⁺ transients in response to 0°C cold (Fig. 3Eb,Ec; drug effect F_(1,366) = 6.630, p = 0.0104; time effect F_(5,366) = 20.26, p < 0.001; interaction F_(5,366) = 6.794, p < 0.001; n = 40 cisplatin, 25 cisplatin plus (S)−3,4-DCPG; Tukey's df = 366, 7 s q = 3.887, p = 0.2072; 14 s q = 0.2487, p > 0.9999; 21 s q = 6.720, p = 0.0002; 28 s q = 4.831, p = 0.0338) but decreased AUC of Ca²⁺ transients in response to 10-g von Frey stimulus (Fig. 3Hb,Hc; t₍₇₃₎ = 3.469, p = 0.000879, n = 55 cisplatin, 20 cisplatin plus (S)−3,4-DCPG).

Figure 3.

mGluR 8 agonist, (S)−3,4-DCPG reduces mechanical and thermal pain sensitivity in cisplatin-treated animals and affects Ca²⁺ activity in response to some stimuli. A, Mechanical pain sensitivity tests using von Frey filaments were performed every 3–15 d up to 80 d. Red arrows indicate injection of saline or cisplatin (days 1, 3, 5, 7, 12, 14, 16, 18); orange arrows indicate injection of (S)−3,4-DCPG or saline vehicle 30 min before cisplatin injection. Mechanical sensitivity is plotted as 50% withdrawal threshold (grams). Saline n = 7 (5 males, 2 females), cisplatin n = 7 (5 males, 2 females), cisplatin + (S)−3,4-DCPG n = 6 (3 males, 3 females). B, Thermal pain sensitivity tests using hot plate assay were performed every 3–15 d up to 80 d. Red arrows indicate injection of saline or cisplatin (days 1, 3, 5, 7, 12, 14, 16, 18); orange arrows indicate injection of (S)−3,4-DCPG or saline vehicle 30 min before cisplatin injection. Thermal sensitivity is plotted as paw withdrawal latency in seconds. Animal numbers are the same as in panel A. C, DRGs were imaged for spontaneous Ca²⁺ activity in the absence of stimuli and during stimulation of the hindpaw with 0°C, 45°C, 0.07-g von Frey, 10-g von Frey, and 100-g press. Graphs show numbers of cells producing Ca²⁺ transients spontaneously, steady-state high cytosol Ca²⁺, total spontaneous activity, and in response to stimuli. Da, Representative images (maximum intensity Z-projections) of spontaneous DRG neuronal Ca²⁺ activity in cisplatin-injected or cisplatin plus (S)−3,4-DCPG-injected animals. Imaged DRGs are shown on left. Areas outlined by red boxes are magnified on the right. Yellow circles track Ca²⁺ activity of two sets of DRG neurons (in cisplatin-injected mice versus cisplatin plus (S)−3,4-DCPG-injected mice) through different stimuli (in later panels). Db, Graphs of spontaneous DRG neuronal Ca²⁺ transients are shown with maximum fluorescence intensity set at 0 s. Y-axis, ΔF/F₀ fluorescence intensity. Cisplatin, n = 5 DRGs, 90 cells, 189 transient peaks. (S)−3,4-DCPG, n = 4 DRGs, 37 cells, 83 transient peaks. Dc, Mean fluorescence intensities of spontaneous DRG neuronal Ca²⁺ transients as a function of time. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. Ea, Fa, Ga, Ha, Ia, Representative images (maximum intensity Z-projections) of Ca²⁺ transients in response to different stimuli: 0°C cold (Ea), 45°C hot (Fa), 0.07-g von Frey (Ga), 10-g von Frey (Ha), and 100-g press (Ia) of cisplatin-injected and cisplatin plus (S)−3,4-DCPG-injected animals. Areas outlined by red boxes are magnified on the right. Yellow circles permit comparison of DRG neurons activated by different stimuli. Eb, Fb, Gb, Hb, Ib, Graphs of DRG neuronal Ca²⁺ transients in response to different stimuli: 0°C cold (Eb), 45°C hot (Fb), 0.07-g von Frey (Gb),10-g von Frey, and 100-g press (Ib). Red arrow indicates start of each stimulus. Y-axis, ΔF/F₀ fluorescence intensity. Ec, Fc, Gc, Hc, Ic, Mean fluorescence intensities of DRG neuronal Ca²⁺ transients in response to various stimuli. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. All panels (and subpanels) D, E, F, G, H (except I) are sequentially derived from a single DRG neuronal ensemble from either a cisplatin -injected animal or a cisplatin plus (S)−3,4-DCPG-injected animal. Transient plots show all individually analyzed cells in gray lines and mean ΔF/F₀ in red. Numbers of DRGs analyzed for 0°C water, 45°C water, 0.07-g von Frey, 10-g von Frey, and 100-g press were, respectively: saline, n = 5, 9, 6, 11, and 8; cisplatin, n = 4, 8, 5, 9 and 9; cisplatin plus (S)−3,4-DCPG, n = 5, 4, 3, 4, and 4. Numbers of cells analyzed for transients for 0°C water, 45°C water, 0.07-g von Frey, 10-g von Frey, and 100-g press were, respectively: cisplatin, n = 40, 51, 8, 55, and 137 and (S)−3,4-DCPG, n = 25, 40, 4, 20, and 243. Please see further detail in Tables 2 and 3 and in Figure 4C. All graphs with error bars show mean ± SEM. Behavioral, AUC, and transient intensity comparisons were performed using two-way ANOVA followed by Tukey's post hoc tests. Cell count comparisons of saline-injected and (S)−3,4-DCPG plus cisplatin-injected versus cisplatin-injected animals were performed by one-way ANOVA followed by Dunnett's post hoc test; for behavioral comparisons: saline versus cisplatin and cisplatin versus cisplatin + (S)−3,4-DCPG *p < 0.05, **p < 0.01, ***p < 0.001; saline versus cisplatin #p < 0.05, ##p < 0.01, ###p < 0.001, for cell counts and Ca²⁺ transients *p < 0.05, **p < 0.01, ***p < 0.001.

Cisplatin-induced hyperalgesia and the numbers of spontaneous Ca²⁺-activated cells are attenuated by Group II mGluR agonist, LY379268

Group II mGluRs are strong candidate analgesic targets for inflammatory pain (Mazzitelli et al., 2018). We tested effects of the mGlu2/3 receptor agonist, LY379268, on cisplatin-induced pain in the cisplatin-induced animal model. We found that injection of LY379268 (10 mg/kg, i.p.) 30 min before each cisplatin injection reduced cisplatin-induced mechanical and thermal hyperalgesia by day 19 (Fig. 4A,B; mechanical drug effect F_(2,109) = 60.27, p < 0.0001; time effect F_(3,109) = 38.62, p < 0.0001; interaction F_(6,109) = 8.617, p < 0.0001; n = 12 saline, 12 cisplatin, 7 cisplatin plus LY379268; Dunnett's df = 109, cisplatin vs cisplatin plus LY379268, day 0 q = 0.5559, p = 0.8075; day 8 q = 1.149, p = 0.4264; day 12 q = 1.380, p = 0.3016; day 19 q = 3.404, p = 0.0019; cisplatin vs saline, df = 109 d 0 q = 0.08549, p = 0.9949; day 8 q = 8.339, p < 0.0001; day 12 q = 6.727, p < 0.0001; day 19 q = 6.847, p < 0.0001; thermal drug effect F_(2,67) = 17.68, p < 0.0001; time effect F_(3,67) = 11.57, p < 0.0001; interaction F_(6,67) = 1.889, p = 0.0955; n = 7 per group; Dunnett's cisplatin vs cisplatin plus LY379268, df = 67, day 0 q = 0.6970, p = 0.7072; day 8 q = 0.5504, p = 0.8049; day 12 q = 2.213, p = 0.0550; day 19 q = 2.551, p = 0.0244; cisplatin vs saline, df = 67, day 0 q = 0.4930, p = 0.8378; day 8 q = 3.180, p = 0.0043; day 12 q = 3.507, p = 0.0016; day 19 q = 4.819, p < 0.0001). The numbers of activated neurons following 45°C hot stimulus was decreased in cisplatin plus LY379268-treated animals compared with cisplatin-treated controls (Fig. 4D; t₍₉₎ = 2.279, p = 0.04868, n = 8 cisplatin, 3 cisplatin plus LY379268). LY379268 increased AUC and slowed decay time of spontaneous Ca²⁺ transients (Fig. 4Ec; AUC t₍₂₆₄₎ = 3.641, p = 0.000327, n = 189 cisplatin, 77 cisplatin plus LY379268 drug effect F_(1,783) = 40.79, p < 0.0001; time effect F_(2,783) = 609.2, p < 0.0001; interaction F_(2,783) = 0.2899, p = 0.7484; Tukey's df = 783 0 s q = 3.767, p = 0.0005; 7 s q = 4.234, p < 0.0001; 14 s q = 2.684, p = 0.0221) Spontaneous amplitude was not statistically distinguishable from saline (Dunnett's saline vs cisplatin plus LY379268 df = 340 q = 0.8542 p > 0.999). LY379268 slowed decay time of Ca²⁺ transients following 45°C (Fig. 4Fc; drug effect F_(1,543) = 20.27, p < 0.0001; time effect F_(5,543) = 14.28, p < 0.0001; interaction F_(5,543) = 19.43, p < 0.0001; n = 51 cisplatin, 66 cisplatin plus LY379268; Tukey's df = 543 7 s q = 5.766, p = 0.0030; 14 s q = 4.781, p = 0.0369; 21 s q = 2.2800, p = 0.9042; 28 s q = 8.299, p < 0.0001; 35 s q = 6.328, p = 0.0006; 42 s q = 6.198, p = 0.0009).

Cisplatin-induced mechanical hyperalgesia, spontaneous DRG neuronal Ca²⁺ activity, and Ca²⁺ response to 45°C hot stimulus are attenuated by histamine receptor 1 antagonist, meclizine

Since meclizine has a protective role against DNA damage by cisplatin (Gorgun et al., 2017), we tested meclizine in cisplatin-treated animals for its effects on behavior and Ca²⁺ activity. Meclizine was injected (16 mg/kg, i.p.) 3 h before cisplatin injection (3.5 mg/kg, i.p.) four times over 8 d. Meclizine reduced cisplatin-induced mechanical hypersensitivity (Fig. 5A; drug effect F_(2,129) = 45.55, p < 0.0001; time effect F_(3,109) = 27.65, p < 0.0001; interaction F_(6,109) = 5.729, p < 0.0001; n = 12 saline, 15 cisplatin, 16 cisplatin plus meclizine; Dunnett's cisplatin vs cisplatin plus meclizine, df = 129 d 0 q = 0.004792, p > 0.9999; day 8 q = 3.078, p = 0.0050; day 12 q = 3.052, p = 0.0054; day 19 q = 2.958, p = 0.0073; cisplatin vs saline, df = 129 d 0 q = 0.07429, p = 0.9960; day 8 q = 7.246, p < 0.0001; day 12 q = 5.845, p < 0.0001; day 19 q = 5.949, p < 0.0001) and reduced cisplatin-induced thermal hypersensitivity (Fig. 5B; drug effect F_(2,63) = 19.04, p < 0.0001; time effect F_(3,63) = 7.403, p = 0.0003; interaction F_(6,63) = 2.958, p = 0.0131; n = 7 saline, 7 cisplatin, 6 cisplatin plus meclizine; Dunnett's cisplatin vs cisplatin plus meclizine, df = 63 d 0 q = 0.004792, p > 0.9999; day 8 q = 3.078, p = 0.0050; day 12 q = 3.052, p = 0.0054; day 19 q = 2.958, p = 0.0073; cisplatin vs saline, df = 129 d 0 q = 0.07429, p = 0.9960; day 8 q = 7.246, p < 0.0001; day 12 q = 5.845, p < 0.0001; day 19 q = 5.949, p < 0.0001). Meclizine also attenuated spontaneous Ca²⁺ activity in cisplatin-treated animals, reducing the numbers of neurons exhibiting spontaneous Ca²⁺ activity (Ca²⁺ oscillation plus steady-state high Ca²⁺) down to numbers comparable to numbers seen in saline-treated animals (Fig. 5C,D; oscillation t₍₁₅₎ = 2.872, p = 0.01163, n = 9 cisplatin, 8 cisplatin plus meclizine; steady high t₍₁₄₎ = 2.134, p = 0.05098, n = 9 cisplatin, 7 cisplatin plus meclizine; total spontaneous activity t₍₁₄₎ = 4.112, p = 0.01056, n = 9 cisplatin, 7 cisplatin plus meclizine), and reduced the numbers of Ca^2+-activated neurons in response to 100-g press (mechanical stimulus) and 45°C water (thermal stimulus; Fig. 5C,Da, Movie 3; 100 g t₍₁₄₎ = 2.197, p = 0.04532, n = 9 cisplatin, 7 cisplatin plus meclizine; 45°C t₍₁₅₎ = 3.500, p = 0.00322, n = 9 cisplatin, 8 cisplatin plus meclizine). In addition, meclizine treatment significantly increased the AUC of Ca²⁺ transients and decay time (return to baseline) in response to 0°C water (cold stimulus; Fig. 5Eb,Ec; 0°C AUC t₍₈₇₎ = 2.187, p = 0.03146, drug effect F_(1,505) = 9.078, p = 0.0027; time effect F_(5,505) = 35.90, p < 0.0001; interaction F_(5,505) = 7.338, p < 0.0001; n = 40 cisplatin, 49 cisplatin plus meclizine; Tukey's df = 505 7 s q = 4.679, p = 0.0463; 14 s q = 1.973, p = 0.9642; 21 s q = 6.654, p = 0.0002; 28 s q = 4.284, p = 0.1032; 35 s q = 1.589, p = 0.9935; 42 s q = 0.5821, p > 0.9999). In contrast, meclizine reduced AUC of Ca²⁺ transients in response to 45°C water (Fig. 5Fb,Fc; AUC t₍₅₉₎ = 2.519, p = 0.01451, drug effect F_(1,501) = 0.3214, p = 0.5710; time effect F_(5,501) = 45.65, p < 0.0001; interaction F_(5,501) = 0.1126, p = 0.1126; n = 40 cisplatin, 21 cisplatin plus meclizine; Tukey's df = 501 7 s q = 5.132, p = 0.0163; 14 s q = 1.513, p = 0.9958; 21 s q = 0.7778, p > 0.9999; 28 s q = 0.07723, p > 0.9999; 35 s q = 2.394, p = 0.8709; 42 s q = 0.0000, p > 0.9999), 10-g von Frey (Fig. 5Hb,Hc; 10 g AUC t₍₈₀₎ = 3.580, p = 0.000587, drug effect F_(1,424) = 12.13, p = 0.0005; time effect F_(5,424) = 8.542, p < 0.0001; interaction F_(5,424) = 2.385, p = 0.0376; n = 55 cisplatin, 27 cisplatin plus meclizine; Tukey's df = 424 7 s q = 5.415, p = 0.0081; 14 s q = 1.739, p = 0.9864, 21 s q = 4.117, p = 0.1406, 28 s q = 2.583, p = 0.8026, 35 s q = 1.523, p = 0.9955, 42 s q = 1.019, p = 0.9999), and 100-g press (Fig. 5Ib,Ic; AUC t₍₂₅₇₎ = 4.063, p = 0.000065, drug effect F_(1,1089) = 33.17, p < 0.0001; time effect F_(5,1089) = 9.116, p < 0.0001; interaction F_(5,1089) = 4.919, p = 0.0001; n = 137 cisplatin, 122 cisplatin plus meclizine; Tukey's df = 1089 7 s q = 8.499, p < 0.0001; 14 s q = 5.082, p = 0.0176, 21 s q = 4.6700, p = 0.0460, 28 s q = 0.5574, p > 0.9999, 35 s q = 5.8300, p = 0.0024, 42 s q = 0.1629, p > 0.9999). Meclizine caused no detectable change on Ca²⁺ transients in response to 0.07-g von Frey (p > 0.5; Fig. 5Gb,Gc; drug effect F_(1,85) = 0.08211, p = 0.7752; time effect F_(5,85) = 6.552, p < 0.0001; interaction F_(5,85) = 0.9157, p = 0.4749; n = 8 cisplatin, 10 cisplatin plus meclizine).

Figure 5.

Meclizine attenuates cisplatin-induced mechanical and thermal pain sensitivity and reduces spontaneous DRG neuronal Ca²⁺ activity and Ca²⁺ activity in response to thermal and mechanical stimuli. A, B, Graphs show mechanical and thermal pain sensitivity tests using von Frey filaments or 45°C hot plate before injection and on days 8, 12, and 19 following the start of injections. Red arrows indicate saline or cisplatin injection, and green arrows indicate meclizine or vehicle injection 3 h before cisplatin injection (days 3, 5, 14, 16). Mechanical sensitivity is plotted as 50% withdrawal threshold in grams. Thermal sensitivity is plotted as paw withdrawal latency (seconds). Saline, n = 14 (8 males, 6 females); cisplatin, n = 13 (8 males, 5 females); meclizine, n = 12 (6 males, 6 females), days 1, 8, 12, and n = 6 (3 males, 3 females) day 19. C, DRGs were imaged for spontaneous Ca²⁺ activity in the absence of stimuli and during stimulation of the hindpaw with 0°C, 45°C, 0.07g von Frey, 10g von Frey, and 100g press. Graphs show numbers of cells producing Ca²⁺ transients spontaneously, steady-state high cytosol Ca²⁺, total spontaneous Ca²⁺ activity, and activity in response to stimuli. Da, Representative images of spontaneous DRG neuronal Ca²⁺ activity in cisplatin-injected, or cisplatin plus meclizine-injected animals. Imaged DRGs are shown on left. Areas outlined by red boxes are magnified on the right. Yellow circles indicate spontaneously activating neurons. Db, Graphs of spontaneous DRG neuronal Ca²⁺ transients are shown, maximum fluorescence intensity set at 0 s. Y-axis, ΔF/F₀ fluorescence intensity. Cisplatin, n = 5 DRGs, 90 cells, 189 transient peaks. Meclizine n = 4 DRGs, 24 cells, 30 transient peaks. Dc, Mean fluorescence intensities of spontaneous Ca²⁺ transients as a function of time frame. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. Ea, Fa, Ga, Ha, Ia, Representative images of DRG neuronal Ca²⁺ activity in response to stimuli: 0°C cold (Ea), 45°C hot (Fa), 0.07-g von Frey (Ga), 10-g von Frey (Ha), and 100-g press (Ia) of cisplatin-injected and cisplatin plus meclizine-injected animals. Areas outlined by red boxes are magnified on the right. Yellow circles indicate Ca²⁺ activating neurons with different stimuli. Eb, Fb, Gb, Hb, Ib, Graphs of DRG neuronal Ca²⁺ transients in response to various stimuli. Red arrow indicates start of each stimulus. Y-axis, ΔF/F₀ fluorescence intensity. Ec, Fc, Gc, Hc, Ic, Mean fluorescence intensities of DRG neuronal Ca²⁺ transients in response to various stimuli. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. All graphs with error bars show mean ± SEM. Transient plots show all individually analyzed cells in gray lines and mean ΔF/F₀ in red. Numbers of DRGs analyzed for 0°C water, 45°C water, 0.07-g von Frey, 10-g von Frey, and 100-g press were, respectively: saline, n = 5, 9, 6, 11, and 8; cisplatin, n = 4, 8, 5, 9, and 9; cisplatin plus meclizine, n = 5, 4, 9, 7, and 4. Numbers of cells analyzed for transients for 0°C water, 45°C water, 0.07-g von Frey, 10-g von Frey, and 100-g press were, respectively: cisplatin, n = 40, 51, 8, 55, and 137 and meclizine, n = 49, 40, 10, 27, and 122. Please see further detail in Tables 2 and 3 and Figure 4C. Behavioral comparisons of saline-injected and meclizine plus cisplatin-injected versus cisplatin-injected animals are two-way ANOVA followed by Dunnett's post hoc test. AUC and transient intensity comparisons are two-way ANOVA followed by Tukey's test. Cell count comparisons of saline-injected and meclizine plus cisplatin-injected versus cisplatin-injected animals were performed by one-way ANOVA followed by Dunnett's post hoc test. For behavioral tests: saline versus cisplatin, *p < 0.05, **p < 0.01, ***p < 0.001; cisplatin versus cisplatin plus meclizine, #p < 0.05, ##p < 0.01, ###p < 0.001. For cell counts and Ca²⁺ transients: cisplatin versus cisplatin plus meclizine *p < 0.05, **p < 0.01, ***p < 0.001.

Movie 3.

Meclizine dramatically decreases the number of DRG neurons responding to paw immersion in 45°C water in cisplatin-treated animals. In animals treated with cisplatin, an average of 56.88 neurons showed Ca²⁺ transients in response to 45°C hot stimulus (A). Treatment with meclizine 3 h prior to cisplatin injection decreased the average number of activated DRG neurons to 26.44 (B). Arrows show examples of activating neurons.

Movie 3B.

Meclizine dramatically decreases the number of DRG neurons responding to paw immersion in 45°C water in cisplatin-treated animals. In animals treated with cisplatin, an average of 56.88 neurons showed Ca²⁺ transients in response to 45°C hot stimulus (A). Treatment with meclizine 3 h prior to cisplatin injection decreased the average number of activated DRG neurons to 26.44 (B). Arrows show examples of activating neurons.

Cisplatin-induced weight loss is attenuated by meclizine and LY379268 but unaffected by (S)−3,4-DCPG

Cachexia, a condition associated with cancer and characterized by weight loss, muscle loss, and adipose loss, increases cancer mortality (Fearon et al., 2013), and is exacerbated by nearly every form of chemotherapy (Langer et al., 2002; Kazemi-Bajestani et al., 2016). Here, we monitored cisplatin-associated weight loss over the time course of experiments. Cisplatin treatment was associated with rapid, persistent weight loss (Fig. 6A–C). Concomitant (S)−3,4-DCPG treatment had no attenuating effect on weight loss (Fig. 6A). Cisplatin-associated body weight loss among (S)−3,4-DCPG-treated animals returned to baseline levels 32 d after the last cisplatin injection (Fig. 6A; Dunnett's df = 425 cisplatin vs cisplatin plus (S)−3,4-DCPG day 58 q = 1.020, p = 0.4924). Animals that received concomitant LY379268 experienced less weight loss over the course of injections than cisplatin-treated controls (Fig. 6B; drug effect F_(2,191) = 124.9, p < 0.0001; time effect F_(10,191) = 25.30, p < 0.0001; interaction F_(20,191) = 3.375, p < 0.0001; n = 7 each, Dunnett's df = 191 cisplatin vs cisplatin plus LY379268 day 18 q = 2.294, p = 0.421). Animals that received concomitant meclizine administration experienced moderate weight loss (drug effect F_(2,154) = 50.01, p < 0.0001; time effect F_(10,154) = 3.595, p = 0.0003; interaction F_(20,154) = 1.753, p = 0.0306; n = 7 saline, 7 cisplatin, 6 cisplatin plus meclizine) and returned to baseline within 10 d after the fourth injection. We also observed a late effect, as body weight of meclizine-treated animals began decreasing 50 d after the last injection (Fig. 6C; Dunnett's df = 191 cisplatin vs cisplatin plus meclizine day 18 q = 1.385, p = 0.2922).

Figure 6.

Cisplatin treatment results in significant body weight loss in mice which is attenuated by meclizine and LY379268 treatment but not by (S)−3,4-DCPG treatment. A, Changes in body weight over the course of mechanical and thermal sensitivity experiments using animals injected with saline, cisplatin, or cisplatin plus (S)−3,4-DCPG (from Fig. 3). Y-axis, Percent of starting body weight. Red arrow indicates cisplatin or saline vehicle injection. Orange arrow indicates (S)−3,4-DCPG or saline vehicle injection 30 min before cisplatin injection. Saline, n = 7 (5 males, 2 females); cisplatin, n = 7 (5 males, 2 females); cisplatin plus (S)−3,4-DCPG, n = 6 (3 males, 3 females). B, Changes in body weight over the course of mechanical and thermal sensitivity experiments using animals injected with saline, cisplatin, or cisplatin plus LY379268 (from Fig. 4). Red arrow indicates cisplatin or saline vehicle injection. Blue arrow indicates LY379268 or saline vehicle injection 30 min before cisplatin injection. Saline, n = 14 (8 males, 6 females); cisplatin, n = 13 (8 males, 5 females); cisplatin plus LY379268, n = 6 (3 males, 3 females). C, Changes in body weight over the course of mechanical and thermal sensitivity experiments using animals injected with saline, cisplatin, or cisplatin plus meclizine (from Fig. 5). Red arrow indicates injection of cisplatin or saline vehicle. Green arrow indicates injection of meclizine or saline vehicle 3 h before cisplatin injection. Saline, n = 6 (3 males, 3 females); cisplatin, n = 8 (5 males, 3 females); cisplatin plus meclizine, n = 6 (3 males, 3 females). All graphs show mean ± SEM. Comparisons were performed using two-way ANOVA followed by Dunnett's post hoc test. Saline versus cisplatin, *p < 0.05, **p < 0.01, ***p < 0.001; cisplatin versus cisplatin plus drug #p < 0.05, ##p < 0.01, ###p < 0.001.

TRP channel inhibitors administered directly onto DRGs during imaging experiments reduces the number of cells producing spontaneous Ca²⁺ transients and NF 110 and AMG 9810 sensitize saline-treated animals to 45°C stimulus

In order to study somatosensory ion channels or receptors involved in cisplatin-induced drug hypersensitivity, we applied four TRP channel inhibitors to DRG to study their effects on spontaneous activity and four stimuli: large brush, 10-g von Frey, 100-g press, and hindpaw immersion in 45°C water. The TRP channel inhibitors used and their targets were HC 030031 (TRPA1; McNamara et al., 2007), AMG 333 (TRPM8; Horne et al., 2018), NF 110 (P2X3 and P2X2/3 heteromultimers; Hausmann et al., 2006), and AMG 9810 (TRPV1; Doherty et al., 2005).

The number of cells with Ca²⁺ activity in response to stimuli was largely unaffected by inhibitor application. Treatment with inhibitors reduced spontaneous Ca²⁺ transients (drug effect F_(4,82) = 4.843, p = 0.0015) but this effect was indistinguishable between treatment groups (drug x treatment group interaction F_(16,82) = 0.4225, p = 0.9725). The treatment group of the animals was more likely to have an effect on the number of cells with Ca²⁺ activity (for a complete list of F and p values, see Table 3), but the treatment group was not associated with a significant effect of any TRP channel inhibitor.

Table 3.

F values from two-way ANOVA on the number of cells producing Ca²⁺ transients spontaneously in the absence of stimuli, spontaneous steady-state high cytosol Ca²⁺ activity, total spontaneous Ca²⁺ activity, and response to large brush, 10-g von Frey, 100-g press, and paw immersion in 45°C water before and after application of TRP channel inhibitor

Stimulus	Main baseline and TRP channel inhibitor effect	Main treatment group effect	Interaction
Spontaneous oscillation	F(4,82) = 4.843, p = 0.0015**	F(4,82) = 2.667, p = 0.0380*	F(16,82) = 0.4225, p = 0.9725
Spontaneous steady-state high	F(4,77) = 0.8213, p = 0.5156	F(4,77) = 10.60, p = 0.001**	F(16,77) = 0.1395, p > 0.999
Spontaneous activity, total	F(4,79) = 1.491, p = 0.2129	F(4,79) = 9.285, p < 0.001***	F(16,79) = 0.6552, p = 0.8281
Large brush	F(4,69) = 1.870, p = 0.1255	F(4,69) = 0.8954, p = 0.4715	F(16,69) = 1.114, p = 0.3600
10-g von Frey	F(4,76) = 1.396, p = 0.2436	F(4,76) = 2.372, p = 0.0597	F(16,76) = 0.8977, p = 0.5743
100-g pinch	F(4,67) = 1.169, p = 0.3323	F(4,67) = 0.3033, p = 0.8748	F(16,67) = 0.8516, p = 0.6246
45°C hot	F(4,78) = 1.590, p = 0.1853	F(4,78) = 7.194, p < 0.001***	F(16,78) = 1.395, p = 0.1663

Inhibitors and their targets were HC 030031 (TRPA1), AMG 333 (TRPM8), NF 110 (P2X3 and P2X2/3 heteromultimers), and AMG 9810 (TRPV1). Treatment groups were saline, cisplatin, cisplatin plus LY379268, cisplatin plus (S)−3,4-DCPG, and cisplatin plus meclizine.

*p < 0.05,

**p < 0.01,

***p < 0.001.

Application of NF 110 to DRG had a striking effect on Ca²⁺ transients in response to 45°C (Fig. 7Ia,Ib,Ic; saline controls AUC t₍₁₀₂₎ = 3.242, p = 0.001607; inhibitor effect F_(1,387) = 16.17, p < 0.001; time effect F_(5,387) = 4.493, p = 0.0005; interaction F_(5,387) = 2.204, p = 0.0532; n = 44 baseline, 60 NF 110; Tukey's df = 387, 7 s q = 0.635, p > 0.9999; 14 s q = 7.050, p < 0.0001; Cisplatin AUC t₍₂₁₁₎ = 4.407, p = 0.000017; inhibitor effect F_(1,874) = 12.62, p = 0.0004; time effect F_(5,874) = 9.165, p < 0.0001; interaction F_(5,874) = 3.070, p = 0.0094; n = 126 baseline, 79 NF 110; Tukey's df = 874, 7 s q = 1.079, p = 0.9998; 14 s q = 6.297, p = 0.0006; 21 s q = 5.469, p = 0.0066). No other TRP channel inhibitor had a significant effect on Ca²⁺ transient fluorescence intensity in response to 45°C (data not shown). NF 110 had no detectable effect on Ca²⁺ transient fluorescence intensity in response to mechanical stimuli (data not shown). In addition, in saline-treated animals, NF 110 and AMG 9810 increased the number of cells producing Ca²⁺ transients in response to 45°C, but there were no effects on any other treatment group (Fig. 7H; saline F_(4,19) = 4.506, p = 0.0099; n = baseline 9, HC 030031 4, AMG 333 4, NF 110 3, AMG 9810 3; Dunnett's df = 18, HC 030031 q = 0.2484, p = 0.9979; AMG 333 q = 0.9985, p = 0.7607; NF 110 q = 3.060, p = 0.0238; AMG 9810 q = 3.329, p = 0.0132).

Figure 7.

TRP channel inhibitors reduce spontaneous Ca²⁺ oscillation and P2X3 inhibitor, NF 110, increases amplitude of Ca²⁺ transients in response to 45°C stimulus. DRGs of animals-treated with saline vehicle, cisplatin, cisplatin plus LY379268, cisplatin plus (S)−3,4-DCPG, and cisplatin plus meclizine were imaged for spontaneous Ca²⁺ activity in the absence of stimuli, and cells with Ca²⁺ oscillation and steady-state high Ca²⁺ were counted. TRP channel inhibitors were topically applied to the DRGs, and the imaging process was repeated with three to four PBS washes between drug applications. Inhibitors and their targets were HC 030031 (TRPA1), AMG 333 (TRPM8), NF 110 (P2X3 and P2X2/3 heteromultimers), and AMG 9810 (TRPV1). A–C, Graphs show numbers of cells producing Ca²⁺ transients spontaneously in the absence of stimuli, steady-state high cytosol Ca²⁺, and total spontaneously Ca²⁺ active cells. Da, Graphs of DRG neurons of animals-treated with saline or cisplatin before and after NF 110 application. Spontaneous Ca²⁺ transients are shown with maximum fluorescence intensity set at 0 s. Y-axis, ΔF/F₀ fluorescence intensity. Db, Dc, Mean fluorescence intensities of spontaneous Ca²⁺ transients are shown as a function of time frame in saline-treated, and cisplatin-treated animals. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. E–H, Mice from panels A–C were subjected to hindpaw stimuli before and after application of each inhibitor to DRG. Graphs show numbers of cells producing Ca²⁺ transients in response to large brush, 10-g von Frey, 100-g press, and paw immersion in 45°C water. For baseline, HC 030031, AMG 333, NF 110, and AMG 9810 cell numbers were, respectively: saline, n = 10, 10, 13, 5, and 6 cells and 15, 28, 28, 13, and 10 transients; and cisplatin n = 18, 19, 7, 13, and 16 cells and 56, 28, 15, 21, and 14 transients. Ia, Graphs of DRG neuronal Ca²⁺ transits of animals treated with saline or cisplatin before and after NF 110 application. Ca²⁺ transients in response to paw immersion in 45°C water are shown, maximum fluorescence intensity is set at 0 s. Y-axis, ΔF/F₀ fluorescence intensity. For baseline, HC 030031, AMG 333, NF 110, and AMG 9810, numbers of cells were, respectively: saline, n = 44, 61, 86, 60, and 60 cells and cisplatin, n = 135, 92, 120, 86, and 80 cells. Ib, Ic, Mean fluorescence intensities of Ca²⁺ transients in response to paw immersion in 45°C water as a function of time frame in saline vehicle-treated (Ib) and cisplatin-treated animals (Ic) before and after NF 110 treatment. Left y-axis, AUC, arbitrary units. Right y-axis, ΔF/F₀ fluorescence intensity. Numbers of DRG analyzed for cell counts are shown in Table 4. All graphs show mean ± SEM. Comparisons of cell counts before and after inhibitor application (within treatment groups vs baseline) and for each inhibitor (between treatment groups vs cisplatin) were performed using one-way ANOVA followed by Dunnett's post hoc test. AUC and Ca²⁺ transient ΔF/F₀ fluorescence intensity comparisons versus no inhibitor baseline were performed using two-way ANOVA followed by post hoc Dunnett's test. *p < 0.05, **p < 0.01, ***p < 0.001. Cell counts compared with baseline before inhibitor treatments; #p <0.05.

Table 4.

Numbers of DRGs analyzed for cell counts during TRP channel inhibitor treatment experiments

Stimulus	Saline	Cisplatin	Cisplatin + LY379268	Cisplatin + (S)−3,4-DCPG	Cisplatin + meclizine
Baseline
Spontaneous activity	11	9	3	9	8
Large brush	10	13	3	4	4
10-g von Frey	11	9	3	8	7
100-g pinch	8	9	3	4	4
45°C hot	9	8	3	8	9
After HC 030031
Spontaneous activity	4	3	3	3	4
Large brush	3	3	3	3	3
10-g von Frey	4	4	3	3	3
100-g pinch	4	3	3	3	3
45°C hot	4	3	3	3	4
After AMG 333
Spontaneous activity	4	3	3	3	4
Large brush	3	3	3	3	3
10-g von Frey	4	3	3	3	3
100-g pinch	5	3	3	3	3
45°C hot	5	3	3	3	4
After NF 110
Spontaneous activity	3	3	3	3	4
Large brush	3	3	3	3	3
10-g von Frey	3	3	3	3	3
100-g pinch	3	4	3	3	3
45°C hot	3	3	3	3	4
After AMG 9810
Spontaneous activity	3	3	3	3	4
Large brush	3	3	3	3	3
10-g von Frey	3	3	3	3	3
100-g pinch	3	3	3	3	3
45°C hot	3	3	3	3	4

Discussion

CIPN is a painful, debilitating, and complex phenomenon that occurs commonly in patients treated with chemotherapeutic drugs. Various chemotherapeutic drugs act through different molecular and cellular mechanisms, suggesting that causes of CIPN are similarly diverse (Quasthoff and Hartung, 2002; Seretny et al., 2014; Addington and Freimer, 2016; Cioroiu and Weimer, 2017; Flatters et al., 2017). This suggests that the treatment of CIPN should be guided by the chemotherapeutic drug used. The DRG is a site of peripheral sensory neuropathy. Therefore, we sought to study the effects of specific chemotherapeutic agents and specific therapies on peripheral neurons and to understand how specific therapies affect responses to different mechanical, thermal, and chemical stimuli. To accomplish this, we used in vivo GCaMP Ca²⁺ imaging of the large population of DRG neurons (>1800/DRG), which enabled us to analyze neuronal firing and activity at a population level as an ensemble. Such in vivo imaging allows for the detection and study of physiologically important phenomena that would be a challenge to detect using more conventional methods such as electrophysiological recordings (Kim et al., 2016).

Prior research has shown that cisplatin treatment induces spontaneous activity and hyperexcitability in dissociated primary sensory neurons in vitro (Laumet et al., 2020). Here, using in vivo ensemble analysis of the large array of DRG neurons in an animal model of CIPN, we found that in response to several mechanical and thermal stimuli Ca²⁺ activity and Ca²⁺ transients were increased by cisplatin treatment. The increased numbers of DRG neurons showing Ca²⁺ activity in response to mechanical and thermal stimuli are consistent with mechanical and thermal hypersensitivity observed in behavioral tests. As expected, strong mechanical stimuli caused larger diameter (>25 µm) neurons to activate more often, while thermal, cold, and weak mechanical stimuli (0.07-g von Frey) activated few if any large diameter neurons, consistent with other reports (Kim et al., 2016; Chisholm et al., 2018; Ishida et al., 2021). Previous work has shown activation of increased numbers of DRG neurons in response to stimuli in a variety of pain models (Kim et al., 2016; Chisholm et al., 2018; Kucharczyk et al., 2020; Ishida et al., 2021). Taken together, these results show that the numbers of DRG neurons activated in response to stimuli measured by Pirt-GCaMP3 fluorescence is a valid indicator of peripheral pain hypersensitivity. Our results show that concomitant administration of (S)−3,4-DCPG, LY379268, or meclizine attenuated CIPN mechanical and thermal pain hypersensitivity. Our results also show that two of three drugs (meclizine, LY379268) which attenuated CIPN pain hypersensitivity also attenuated the number of neurons exhibiting spontaneous Ca²⁺ activation. Furthermore, concomitant meclizine or LY379268 administration reduced cisplatin-induced weight loss.

Cisplatin increased the fluorescence intensity of Ca²⁺ transients for several mechanical stimuli (100- or 300-g press or small brush). Interpretation of AUC and plots of fluorescence intensity must be done carefully. Increases in average AUC and amplitude of Ca²⁺ transients should indicate a stronger response to stimuli (e.g., neurons firing more action potentials, upregulation of voltage-gated and other Ca²⁺ channels, lowered thresholds for Ca²⁺ channel activation, increased efflux of Ca²⁺ from ER or mitochondria, etc.). However, these same factors that cause increased numbers of neurons to produce Ca²⁺ transients with higher amplitudes can increase baseline cytosolic Ca²⁺ concentrations and, therefore, increased baseline fluorescence. It is important to note that plots of fluorescence depend on the change in fluorescent intensity relative to baseline. Cisplatin-treated animals generally exhibit higher overall baseline fluorescence intensity (increased F₀). This reduces the ΔF/F₀ ratio, reducing the calculated AUC. Despite the apparently elevated baseline intensities, the calculated AUC in cisplatin-treated animals was nearly always at least as high as in saline vehicle-treated control animals and often higher. An alternative, nonmutually exclusive possibility, is that the increased number of cells producing Ca²⁺ transients in response to cisplatin treatment produce low amplitude transients, which would reduce the average amplitude (Fig. 4C).

The effect of (S)−3,4-DCPG on cisplatin-induced CIPN pain does not appear to be mediated by DRG neurons. mGluR8 is present in most DRG (Carlton and Hargett, 2007; Govea et al., 2012) and trigeminal ganglia (Boye Larsen et al., 2014) neurons. However, the role of mGluR8 in the DRG in CIPN has not been determined. Peripheral administration of mGluR8 agonists (such as (S)−3,4-DCPG) can attenuate thermal and mechanical hyperalgesia caused by carrageenan, formalin, and capsaicin (Marabese et al., 2007; Govea et al., 2012; Pereira and Goudet, 2018). (S)−3,4-DCPG can modulate pain centrally, as well. Infusion of (S)−3,4-DCPG into the periaqueductal gray area was sufficient to reduce nociceptive responses and thermal and mechanical hypersensitivity caused by formalin and carrageenan injection. In addition, infusion of an antagonist of Group III mGluRs to the periaqueductal gray area reduced the analgesic effect of peripheral (S)−3,4-DCPG, showing that mGluR8 can act fully or partially through a central anti-nociceptive pathway (Marabese et al., 2007). Administration of mGluR8 agonist (S)−3,4-DCPG 30 min before cisplatin produced a striking reduction in cisplatin-induced mechanical hypersensitivity and reduced thermal hypersensitivity; yet, (S)−3,4-DCPG-treated animals were not distinguishable in terms of numbers of cells with spontaneous Ca²⁺ activity or numbers of activated cells responding to any stimulus analyzed in this study except 0°C, where (S)−3,4-DCPG increased the number of responding neurons. While there were some subtle differences in Ca²⁺ transients, (S)−3,4-DCPG failed to produce the striking effects on DRG neuronal Ca²⁺ activity seen with LY379268 and meclizine. Since (S)−3,4-DCPG administration was systemic, it was not possible to distinguish central and peripheral effects. The lack of striking decreases in numbers of spontaneously Ca²⁺-activated neurons or activation from stimuli combined with known effects of (S)−3,4-DCPG in the periaqueductal gray area suggests that attenuation of cisplatin-induced mechanical and thermal hypersensitivity by (S)−3,4-DCPG is central rather than peripheral or that (S)−3,4-DCPG works peripherally through a mechanism that does not affect DRG neuronal Ca²⁺ activity. mGluR8 may be a target for CIPN pain although it does not seem to act peripherally.

Group II mGluR agonists attenuate hyperalgesia and allodynia from a broad range of causes (Carlton et al., 2009; Davidson et al., 2016; Johnson et al., 2017; Mazzitelli et al., 2018; Sheahan et al., 2018), and antagonists of Group II receptors may aggravate hyperalgesia and allodynia, block analgesia, and prolong recovery time from these conditions (Yang and Gereau, 2002, 2003; Zhuo et al., 2016), and may increase the number of action potentials in primary sensory neurons in response to capsaicin or heat (Carlton et al., 2011). Group II mGluRs (mGluR2/3) are broadly expressed in primary sensory neurons (Carlton and Hargett, 2007; Boye Larsen et al., 2014; Sheahan et al., 2018). Our data show that in this cisplatin-induced CIPN model, mGluR2/3 agonist LY379268 reduces mechanical and thermal hyperalgesia and decreases the number of DRG neurons exhibiting spontaneous Ca²⁺ activity as well as the number of DRG neurons producing Ca²⁺ transients in response to 45°C. Mechanical and thermal sensitivity assays were based on systemic administration, so it is not possible to determine the extent to which the effect was peripheral or central. However, reduced numbers of DRG Ca²⁺-activated neurons indicates that analgesia via primary sensory neuron activity is one mechanism. Other studies have shown peripheral mGluR2 or mGluR2/3 agonists can block inflammatory hyperalgesia (Yang and Gereau, 2002, 2003; Yamamoto et al., 2007). Side effects remain a major concern for Group II mGluR agonists, as convulsions have been observed in preclinical animal models (Dunayevich et al., 2008). Limiting a mGluR2/3 drug to the periphery may reduce side effects (Chishty et al., 2001; Gao et al., 2017; Fan et al., 2018). Our results combined with earlier research suggest that a mGluR2/3 agonist with poor blood brain barrier penetration could treat CIPN pain without disrupting mGluR2/3 signaling in the central nervous system.

Concomitant meclizine treatment reduces mechanical and thermal hyperalgesia and decreases the number of DRG neurons exhibiting spontaneous Ca²⁺ activity (both Ca²⁺ oscillation and total cells with steady-state high Ca²⁺ activity) and reduces the number of DRG neurons producing Ca²⁺ transients in response to 45°C. Meclizine also reduced the intensity of Ca²⁺ transients arising from two different mechanical stimuli: 10-g von Frey and 100-g press, suggesting that meclizine may act on mechanical pain by reducing the activity of neurons responding to stimuli. Because meclizine treatment was systemic, it is impossible to distinguish a central from peripheral effect. However, there are good reasons to believe that meclizine acts via peripheral mitochondrial repair and/or by reducing oxidative stress. Cisplatin binds mitochondrial DNA in DRG neurons, inhibiting mitochondrial transcription and replication, and causing mitochondrial DNA degradation (Podratz et al., 2011). Cisplatin directly inhibits mitochondrial respiration system in vivo and reduces transcription and increases degradation of mitochondrial RNA (Garrido et al., 2008). Meclizine improves clearance of damaged mitochondrial DNA (Gorgun et al., 2017) and shifts mitochondrial respiration toward glycolytic metabolism (Hong et al., 2016; Zhuo et al., 2016). Meclizine also maintains an elevated ratio of reduced:oxidized glutathione and NADPH (Zhuo et al., 2016; Gorgun et al., 2017). Thus, meclizine employs multiple mechanisms to protect neurons from oxidative damage. ROS and oxidative stress are known to activate and/or potentiate a variety of transient receptor potential channels (Mori et al., 2016; Carrasco et al., 2018), which may partially depolarize neurons, thus rendering them more excitable and resulting in increased spontaneous Ca²⁺ activity and decreased neuron firing threshold in response to stimuli. ROS also affect potassium and sodium channels (Annunziato et al., 2002; Sahoo et al., 2014) and can damage a wide range of other molecules in the cell (Juan et al., 2021). By improving the redox state, enhancing DNA repair, and shifting primary sensory neuron metabolism away from mitochondrial respiration and toward glycolysis, meclizine may attenuate hyperexcitability through multiple known mechanisms, and thereby reduce hypersensitivity to painful stimuli. Meclizine is also an approved drug for use in humans, making it an excellent candidate for possible use in treating CIPN.

One of the more common side effects of platinum-based chemotherapeutic drugs is weight loss (Langer et al., 2002; Fearon et al., 2013; Kazemi-Bajestani et al., 2016). Numerous mechanisms by which cisplatin and other platinum-based chemotherapeutic drugs cause weight loss have been proposed (Yakabi et al., 2010; Hiura et al., 2012; Garcia et al., 2013; Matsumura et al., 2013; Yoshimura et al., 2013; Woo et al., 2016; Guo et al., 2018; Wong et al., 2020). Meclizine reduced both cisplatin-induced hypersensitivity to mechanical pain and cisplatin-induced weight loss. There were no detectable effects of (S)−3,4-DCPG on body weight of cisplatin-treated animals. Of relevance, meclizine is commonly used for treatment of nausea and vomiting during pregnancy (Heitmann et al., 2016) and motion sickness (Paule et al., 2004), and cisplatin induces nausea in humans (Dilruba and Kalayda, 2016). Meclizine is known to be a strong antagonist of H₁ histamine receptor, a weak antagonist of muscarinic acetylcholine receptor (Kubo et al., 1987), and, in mice but not in humans, a weak antagonist of constitutive androstane receptor (Huang et al., 2004). Histamine suppresses food intake in the central nervous system (Clineschmidt and Lotti, 1973; Provensi et al., 2016), and histaminergic neurons play a large role in food intake, metabolism, weight gain or loss, and foraging-related locomotor activities (Provensi et al., 2016). Mice lacking the muscarinic three acetylcholine receptor eat less and weigh less than wild-type mice (Yamada et al., 2001; Gautam et al., 2006). Stimulation of hypothalamic muscarinic acetylcholine receptors stimulates food intake (Jeong et al., 2017). These observations suggest that meclizine may attenuate weight loss through antagonism of anorexigenic H₁ histamine receptors or muscarinic acetylcholine receptors.

We investigated specific receptor roles in our imaging system by applying specific TRP channel inhibitors to DRG cell bodies. Unsurprisingly, the inhibitors decreased the number of spontaneously activating cells. The P2X3 and P2X2/3 heteromultimer inhibitor, NF 110, appeared to sensitize DRGs to 45°C stimulus, increasing the number of cells responding in saline injected control animals and dramatically increasing the fluorescence intensity of Ca²⁺ transients in both cisplatin-injected and saline-injected animals. No TRP channel inhibitors decreased cisplatin-induced sensitivity to stimuli. TRPV1 inhibitor, AMG 9810, also increased the number of Ca²⁺-activated cells responding to 45°C stimulus. These 45°C stimulus results seem paradoxical, but the inhibitor screen was for indirect effects on stimuli. Inhibitors were applied to the cell bodies rather than into the hindpaw where the receptors that respond directly to the stimuli are found. Importantly, these effects were detectable in saline-treated control animals, but not in cisplatin-treated or cisplatin-plus-TRP channel inhibitor-treated animals. These results suggest that P2X3-injected and TRPV1-induced sensitization is either blocked by, downstream of, or redundant with the effects of cisplatin.

We have used a cisplatin-based CIPN model to study DRG neurons in vivo at the level of an intact neuronal ensemble. We found that two (LY379268 and meclizine) of three drug treatments tested attenuated CIPN-induced pain hypersensitivity, reduced spontaneous DRG neuron Ca²⁺ activity, and reduced the number of cells responding to 45°C stimulus. The third drug, (S)−3,4-DCPG, did not reduce spontaneous Ca²⁺ activity, but (S)−3,4-DCPG is known to reduce pain hypersensitivity centrally in some inflammatory pain models. Finally, our results demonstrate that in vivo Ca²⁺ imaging of DRG neurons can be used to study mechanisms of CIPN and peripherally-acting candidates for therapeutic intervention.