Share this article on:

Orally active Epac inhibitor reverses mechanical allodynia and loss of intraepidermal nerve fibers in a mouse model of chemotherapy-induced peripheral neuropathy

Singhmar, Poojaa; Huo, XiaoJiaoa; Li, Yanb; Dougherty, Patrick, M.b; Mei, Fangc; Cheng, Xiaodongc; Heijnen, Cobi, J.a; Kavelaars, Annemiekea,*

doi: 10.1097/j.pain.0000000000001160
Research Paper

Chemotherapy-induced peripheral neuropathy (CIPN) is a major side effect of cancer treatment that significantly compromises quality of life of cancer patients and survivors. Identification of targets for pharmacological intervention to prevent or reverse CIPN is needed. We investigated exchange protein regulated by cAMP (Epac) as a potential target. Epacs are cAMP-binding proteins known to play a pivotal role in mechanical allodynia induced by nerve injury and inflammation. We demonstrate that global Epac1-knockout (Epac1−/−) male and female mice are protected against paclitaxel-induced mechanical allodynia. In addition, spinal cord astrocyte activation and intraepidermal nerve fiber (IENF) loss are significantly reduced in Epac1−/− mice as compared to wild-type mice. Moreover, Epac1−/− mice do not develop the paclitaxel-induced deficits in mitochondrial bioenergetics in the sciatic nerve that are a hallmark of CIPN. Notably, mice with cell-specific deletion of Epac1 in Nav1.8-positive neurons (N-Epac1−/−) also show reduced paclitaxel-induced mechanical allodynia, astrocyte activation, and IENF loss, indicating that CIPN develops downstream of Epac1 activation in nociceptors. The Epac-inhibitor ESI-09 reversed established paclitaxel-induced mechanical allodynia in wild-type mice even when dosing started 10 days after completion of paclitaxel treatment. In addition, oral administration of ESI-09 suppressed spinal cord astrocyte activation in the spinal cord and protected against IENF loss. Ex vivo, ESI-09 blocked paclitaxel-induced abnormal spontaneous discharges in dorsal root ganglion neurons. Collectively, these findings implicate Epac1 in nociceptors as a novel target for treatment of CIPN. This is clinically relevant because ESI-09 has the potential to reverse a debilitating and long-lasting side effect of cancer treatment.

Epac-inhibitor ESI-09 reverses paclitaxel-induced mechanical allodynia and associated hallmarks of chemotherapy-induced peripheral neuropathy in rodents including spinal cord astrocyte activation, loss of intraepidermal nerve fibers, and spontaneous neuronal discharges.

aLaboratory of Neuroimmunology, Department of Symptom Research, The University of Texas MD Anderson Cancer Center, Houston, TX, USA

bDepartments of Anesthesia and Pain Medicine Research, The University of Texas MD Anderson Cancer Center, Houston, TX, USA

cDepartment of Integrative Biology and Pharmacology, The University of Texas Health Science Center, Houston, TX, USA

Corresponding author. Address: Laboratory of Neuroimmunology, Department of Symptom Research, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Unit 384, Houston, TX 77030, USA. Tel.: (713) 794-5297. E-mail address: (A. Kavelaars).

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

P. Singhmar and X. Huo contributed equally to this study.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (

Received May 22, 2017

Received in revised form December 18, 2017

Accepted January 08, 2018

Back to Top | Article Outline

1. Introduction

Chemotherapy-induced peripheral neuropathy (CIPN) is a major dose-limiting side effect of standard antineoplastic treatments.48,50,63 Approximately 25% to 75% of cancer patients treated with paclitaxel experience CIPN symptoms5,10,11 including tingling, pain, hypersensitivity to pain, and numbness. These symptoms often persist long after completion of treatment and can be irreversible.12,29,38 There is no FDA-approved treatment for CIPN. Identification of preventive or curative interventions and underlying mechanisms contributing to CIPN would be a significant step forward in improving cancer treatment adherence and the quality of life of cancer survivors.

Paclitaxel is widely used for the treatment of solid tumors including breast, ovarian, lung, and head and neck cancer.50 Paclitaxel acts by hyperstabilizing microtubules to cause mitotic arrest and cell death. Pain-sensing neurons (nociceptors) are vulnerable to chemotherapy drugs because of the presence of their ganglia outside the blood-brain barrier and lack of a restrictive vascular barrier in the periphery.63

Cellular and molecular mechanisms associated with CIPN include mitochondrial dysfunction, oxidative stress, production of cytokines and chemokines in dorsal root ganglion (DRG) and spinal cord, and altered expression of ion channels in DRG neurons.10,65,66,68 Chemotherapy-induced peripheral neuropathy is associated with astrocyte activation in the spinal cord and loss of intraepidermal nerve fibers (IENFs).20,37,69 Mitochondrial structure and functional abnormalities in the peripheral nerve axons is a key mechanism underlying CIPN.25,37 It is hypothesized that abnormal mitochondrial function creates energy deficits leading to an increase in spontaneous discharges by myelinated A-fibers (Aβ and Aδ) and unmyelinated C-fibers, and IENF loss.8,68

cAMP signaling has long been recognized as a major pathway in the sensitization of nociceptors.22,24,54,55 The cAMP effectors protein kinase A (PKA), exchange protein directly activated by cAMP (Epac, also known as RAP-GEF), and cyclic nucleotide-gated ion channels are involved in pain signaling.16,27,35,58 The cAMP/PKA pathway has been intensely studied for its role in inflammation-induced peripheral sensitization.19,22,31,54,55 Research by our group and others has identified an important contribution of a cAMP-EPAC (exchange proteins activated by cAMP) pathway to mechanical allodynia and thermal hyperalgesia in rodent models of inflammatory pain and in surgical models of neuropathic pain.23,28,31,52,59,60 For example, Epac1 knockout animals (Epac1−/−) are protected against mechanical allodynia in the complete Freund adjuvant model of chronic inflammatory pain and in the spinal nerve transection model of neuropathic pain.21,52,60 Inhibitors of Epac have been developed recently and the competitive inhibitor ESI-09 has emerged as an attractive therapeutic option for blocking Epac activity in vivo.1,4,26,52

Here, we tested the hypothesis that Epac1 is a target for prevention or reversal of paclitaxel-induced peripheral neuropathy using pharmacological and genetic interventions. We identified ESI-09 as a novel inhibitor of chemotherapy-induced pain. Administration of ESI-09 in mice reversed established paclitaxel-induced mechanical allodynia and associated hallmarks of CIPN, that is, astrocyte activation, IENF loss, and spontaneous neuronal discharges.

Back to Top | Article Outline

2. Methods

2.1. Animals

Male and female C57BL/6J mice aged 10 to 12 weeks and male and female mice homozygous for global53 deletion of the Epac1 gene (Epac1−/− mice) were used in this study. Nav1.8 Cre mice were used to drive cell-specific deletion of Epac1 in nociceptors (N-Epac1−/−). Mice were bred and maintained in the animal facility of the Institute of Biosciences and Technology of Texas A&M University and of The University of Texas MD Anderson Cancer Center, Houston, Texas. All experiments were performed in accordance with the National Institute of Health Guidelines for the Care and Use of Laboratory Animals and the Ethical Issues of the International Association for the Study of Pain and were approved by the Institutional Animal Care and Use Committees. Spared nerve injury (SNI) was performed in male C57BL/6J mice as previously described.17 In brief, under 1.5% isoflurane anesthesia, the 3 terminal branches of the sciatic nerve were exposed carefully with minimal damage. Leaving the sural nerve intact, the tibial and the common peroneal nerves were tightly ligated with 6-0 silk surgical sutures and cut 1 to 2 mm distal to the ligation. Sham controls underwent anesthesia, incision, and exposure of the nerve only.

Back to Top | Article Outline

2.2. Drug

Paclitaxel (6 mg/mL) in 50% El Kolipher (Sigma Aldrich, St. Louis, MO) and 50% ethanol (Sigma Aldrich) or vehicle alone was diluted in sterile saline to final concentration of 1 mg/mL and administered intraperitoneally (i.p.) at a dose of 10 mg/kg every other day for 2 weeks. This dosing schedule of paclitaxel was used in our previous studies to induce CIPN in mice.37,44 The Epac inhibitor ESI-093,67 was formulated with corn oil (Sigma Aldrich) and administrated daily by oral gavage at a dose of 20 mg/kg. CE3F4 (Tocris) was dissolved in dimethyl sulfoxide and used at a concentration of 2.5 μM.

Back to Top | Article Outline

2.3. Measurement of mechanical allodynia

Mechanical hyperalgesia was measured using von Frey hairs as previously described.44 Briefly, mice were placed in a plastic cage (10 × 10 × 13 cm3) with a mesh floor for 30 minutes before testing. Animals were placed on a wire grid floor, through which the von Frey hairs (North Coast Medical Inc, CA) were applied (bending force range from 0.02 to 1.4g, starting with 0.16g) by an investigator blinded to treatment. The hair force was increased or decreased according to the response. The 50% paw withdrawal threshold was calculated using the up-and-down method.13 Clear paw withdrawal, shaking, or licking was considered as nociceptive-like responses. Ambulation was considered an ambiguous response, and in such cases, the stimulus was repeated.

Back to Top | Article Outline

2.4. Immunohistochemistry

For IENFs, biopsies from the plantar surface of the hind paws were collected and processed as described.9,44,51 Biopsies were immediately saved in Zamboni's fixative for 24 hours, cryoprotected in 20% sucrose, frozen in optimal cutting temperature compound, and sliced into 25 μm sections. Sections were blocked in phosphate-buffered saline (PBS) containing 5% normal donkey serum/0.1% saponin and stained with antibodies against the panneuronal marker PGP9.5 (Rabbit; AbD Serotec, Oxford, United Kingdom) and collagen IV (Goat; Southern Biotech, Birmingham, AL) followed by Alexa-594 donkey anti-rabbit (Life Technologies, Grand Island, NY) and Alexa-488 donkey anti-goat (Invitrogen, Grand Island, NY). For negative control sections, primary antibody was omitted. Sections were visualized using a Leica fluorescence microscope and 3 to 4 randomly chosen sections from each paw were quantified. The number of nerve fibers crossing the collagen-stained dermal/epidermal junction into the epidermis was counted using the 40X objective in a blinded setup. The length of the epidermis within each field was measured using Leica software. Intraepidermal nerve fiber density is expressed as the total number of fibers/length of the epidermis (IENFs/mm). For glial fibrillary acidic protein (GFAP) staining, mice were perfused intracardially with PBS followed by 4% paraformaldehyde in PBS, and the spinal cord was collected. Cryosections (6 μm) of lumbar segments L2 to L5 were incubated with 1:200 rabbit anti-GFAP (Acris Antibodies GmbH, Hiddenhausen, Germany) followed by Alexa-488 goat anti-Rabbit (Invitrogen). Glial fibrillary acidic protein expression in the dorsal horn was quantified in 6 to 8 sections per group (3-4 mice per group). The mean intensity of fluorescence and the percent area positive were calculated using ImageJ software.

Back to Top | Article Outline

2.5. Oxygen consumption rate

Oxygen consumption rate (OCR) was measured using the Seahorse XFe 24 (Seahorse Bioscience, North Billerica, MA) as previously described.36,42 Tibial nerves were dissected starting at their start point from sciatic nerve till the ankle of the foot and placed into islet capture XF24 microplate in XF media supplemented with 5.5 mM glucose, 0.5 mM sodium pyruvate, and 1 mM glutamine without desheathing. Oxygen consumption rate was measured before (baseline respiration) and after addition of mitochondrial inhibitors. The following inhibitors were used during the assay: 12 μM oligomycin, 20 μM carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and 20 μM of rotenone and 20 μM of antimycin A. An assay cycle of 3-minute mix, 3-minute wait, and 4-minute measure combination was repeated 4 times for baseline rates and after each drug injection. Adenosine triphosphate (ATP)-coupled respiration is the OCR used for mitochondrial ATP synthesis and is reduction in OCR because of addition of oligomycin. Maximal mitochondrial respiration (maximal respiratory capacity) is the reduction in OCR because of uncoupler FCCP. Oxygen consumption rate values were normalized to the total protein content of each well.

Back to Top | Article Outline

2.6. Acute dissociation of dorsal root ganglion neurons for whole-cell patch clamp recordings

Rats were given paclitaxel as described previously.41 Animals were deeply anesthetized with isoflurane, and the bilateral L4 and L5 ganglia were surgically exposed and removed. The ganglia were placed in a flask containing trypsin (0.0625 mg/mL, Hyclone) and type IA collagenase (1 mg/mL, Sigma-Aldrich) in Dulbecco's Minimal Essential Medium and shaken for 50 minutes in a heated (37°C) bath. After the cells were mechanically dispersed, they were plated on poly-L-lysine-coated glass sheets and held in culture dishes with Dulbecco's Minimal Essential Medium (10% fetal bovine serum) until use; the cells were used for patch clamp within 6 hours after plating.39 Whole-cell patch recording was performed as previously described and the cells were transferred to a recording chamber and perfused with oxygenated artificial cerebrospinal fluid (2 mL/min) at room temperature. Only neurons with a stable resting membrane potential of at least −40 mV and evoked spikes that overshot 0 mV were used. Series resistance was compensated to above 70%.

Back to Top | Article Outline

2.7. Statistical analysis

Data are expressed as mean ± SEM. Statistical analysis was performed using 1-way or 2-way analysis of variance followed by Turkey analysis. P < 0.05 was considered statistically significant.

Back to Top | Article Outline

3. Results

3.1. Effect of the Epac inhibitor ESI-09 on paclitaxel-induced mechanical allodynia

Male mice were treated with paclitaxel every other day for 2 weeks (10 mg/kg, 3 injections per week with 2 days' rest in between) to induce peripheral neuropathy.37 Paclitaxel-treated mice developed mechanical allodynia as tested using von Frey hairs (Fig. 1A, open circles). To test whether Epac inhibition reverses paclitaxel-induced peripheral neuropathy, we administered the Epac-inhibitor ESI-09 (20 mg/kg daily for 6 days) starting at day 1 after completion of paclitaxel treatment. Treatment with ESI-09 strongly inhibited mechanical allodynia (Fig. 1A, filled circles). ESI-09 was still effective in inhibiting mechanical allodynia even when we delayed administration of ESI-09 until 10 days after the last dose of paclitaxel (Fig. 1B). ESI-09 also inhibits paclitaxel-induced mechanical allodynia in female mice (Supplementary Figure 1A, available online at In mice treated with ESI-09 alone, mechanical sensitivity was not changed.52

Figure 1

Figure 1

The beneficial effect of ESI-09 is not limited to chemotherapy-induced neuropathic pain; the results in Figures 1C and D demonstrate that ESI-09 also reverses mechanical allodynia induced by SNI when the treatment is started early (3 days) or late (8 days) after injury. ESI-09 did not affect mechanical sensitivity in the contralateral paw (Supplementary Figures 1B and 1C, available online at

Back to Top | Article Outline

3.2. Effect of Epac1 deletion on paclitaxel-induced mechanical allodynia

In vitro, ESI-09 inhibits activation of both Epac1 and Epac2.56,72 To investigate the contribution of Epac1, we tested mice lacking the Epac1 gene. The results in Figure 2 demonstrate that Epac1−/− mice are protected from paclitaxel-induced mechanical allodynia, indicating that only deleting the Epac1 gene is enough to protect against taxol-induced mechanical allodynia. We did not detect differences between male and female Epac1−/− mice and therefore pooled data for the 2 sexes. In line with previous studies,21,52 we did not observe any effect of genetic deletion of Epac1 on mechanical sensitivity at baseline (Fig. 2; bar graph). Using in situ hybridization analysis, we showed previously that Epac1 mRNA is expressed in all cells in the DRG.60 To determine whether nociceptor Epac1 is required for the development of mechanical allodynia in response to paclitaxel treatment, we used male and female mice with cell-specific deletion of Epac1 in Nav1.8+ neurons (N-Epac1−/− mice). N-Epac1−/− male and female mice were protected against paclitaxel-induced mechanical allodynia (Fig. 2). Consistent with what we observed in mice with global deletion of Epac1 (Epac1−/− mice), cell-specific deletion of Epac1 from nociceptors did not affect baseline mechanical sensitivity (Fig. 2).

Figure 2

Figure 2

Collectively, these findings indicate that the beneficial effects of ESI-09 in the paclitaxel model of neuropathy are mediated through inhibition of nociceptor Epac1, although we cannot fully exclude a partial contribution of Epac2 or other potential cAMP targets to the observed effects of the pharmacological intervention.

Back to Top | Article Outline

3.3. Impact of Epac inhibition or Epac1 deletion on paclitaxel-induced astrocyte activation

Paclitaxel-induced peripheral neuropathy is associated with activation of astrocytes in the spinal cord.47,69 Therefore, we determined whether the inhibition of Epac signaling by oral administration of ESI-09 or genetic deletion of Epac1 affected astrocyte activation. Consistent with earlier studies,47,69 paclitaxel increased GFAP staining indicating astrocyte activation in the dorsal horn of the spinal cord of wild-type (WT) mice (Fig. 3A). Treatment with ESI-09 starting at 10 days after completion of paclitaxel normalized the expression of GFAP (Figs. 3A and B). Moreover, in Epac1−/− mice, paclitaxel did not increase spinal cord GFAP expression (Fig. 3C). Cell-specific deletion of Epac1 from NaV1.8+ neurons also prevented astrocyte activation in response to paclitaxel (Figs. 3C and D). These findings indicate that paclitaxel-induced astrocyte activation occurs downstream of Epac1 signaling in sensitized nociceptors. In line with previous studies,43,49,69 we did not observe changes in microglia activation as assessed by staining for Iba-1 in the spinal cord of paclitaxel–treated mice (not shown).

Figure 3

Figure 3

Back to Top | Article Outline

3.4. Impact of Epac inhibition or deletion on paclitaxel-induced intraepidermal nerve fiber loss

Chemotherapy-induced peripheral neuropathy is known to be associated with a reduction in density of IENFs in the plantar surface of the hind paws, and this phenomenon is believed to be related to the impaired mitochondrial function in the peripheral nerve.33 The data in Figure 4 show that also in our hands, paclitaxel treatment reduces IENF density as assessed at day 18 after completion of paclitaxel treatment. ESI-09 was administered daily for 6 days, starting 10 days after completion of paclitaxel treatment. The results show that this schedule of ESI-09 administration significantly reduced the loss in IENF density (Figs. 4A and B). In addition, mice with either global- or nociceptor-specific deletion of Epac1 are protected from the paclitaxel-induced reduction in IENF density (Figs. 4C and D).

Figure 4

Figure 4

Back to Top | Article Outline

3.5. Contribution of Epac1 to paclitaxel-induced mitochondrial dysfunction

Multiple studies have reported mitochondrial abnormalities in the peripheral neurons in rodent models of CIPN.37,66 We examined whether the protection against paclitaxel-induced mechanical allodynia in Epac1-deficient mice was associated with preservation of mitochondrial function in the peripheral nerves. Ex vivo mitochondrial respiration of tibial nerve samples from WT and Epac1−/− mice was compared using Seahorse extracellular flux analyzer. In WT mice, paclitaxel treatment resulted in reduced basal, ATP-coupled, and maximal (maximal respiratory capacity) mitochondrial respiration. Proton leak was not affected by paclitaxel treatment. Epac1−/− mice were protected against all the paclitaxel-induced defects in mitochondrial bioenergetics (Figs. 5A and B).

Figure 5

Figure 5

Back to Top | Article Outline

3.6. Impact of Epac inhibition on paclitaxel-induced spontaneous discharges by dorsal root ganglion neurons

Next, we investigated the effect of ESI-09 on the paclitaxel-induced spontaneous discharges in DRG neurons. Similar to previous studies, paclitaxel treatment led to high incidence of abnormal spontaneous activity in DRG neurons. The data in Figure 6A show that treatment with ESI-09 reversed these spontaneous discharges in DRG neurons, returning it to normal activity. We also used Epac1-selective inhibitor, CE3F4, and found it inhibited spontaneous discharges as well. These results indicate that Epac1 is involved in the development of abnormal spontaneous activity in the DRG neurons.

Figure 6

Figure 6

Back to Top | Article Outline

4. Discussion

We demonstrate for the first time that pharmacological inhibition of Epac using the orally active compound ESI-09 inhibits already established mechanical allodynia in mice treated with paclitaxel (Figs. 1A and B). ESI-09 inhibits the activation of astrocytes in the spinal cord and loss of pain sensory nerve endings in the periphery, 2 well-established hallmarks of CIPN (Figs. 3A and 4A). Moreover, ESI-09 significantly reduced the incidence of abnormal spontaneous discharges in DRG neurons from paclitaxel-treated rats (Fig. 6). We show that genetic deletion of Epac1 recapitulates the effect of the inhibitor ESI-09. Epac1 knockout animals show reduced paclitaxel-induced mechanical allodynia, spinal cord astrocyte activation, and loss of IENF density compared with WT mice. Moreover, Epac1−/− knockout animals do not develop the paclitaxel-induced energy deficits in mitochondria in the peripheral nerve. We further show that Epac1 in nociceptors is critical for development of CIPN symptoms. Mice with a selective Epac1 deficiency in Nav1.8-positive primary sensory neurons do not show spinal cord astrocyte activation and IENF loss in response to paclitaxel treatment (Figs. 3C and 4C). The protective effect of Epac1 inhibition is not limited to CIPN induced by paclitaxel but extends to mechanical allodynia in response to SNI-induced neuropathy (Fig. 1C) and cisplatin (unpublished). The finding that ESI-09 and genetic deletion of Epac1 both inhibit chemotherapy-induced mechanical allodynia and identifies Epac1 as a novel potential therapeutic target for the treatment and prevention of CIPN. This is important because CIPN is a major adverse side effect of chemotherapy treatment and there are currently no FDA-approved drugs available to prevent or alleviate CIPN.

There have been many studies which investigated the effects of opioids, antiepileptic's, and antidepressants on CIPN, but clinical results have been disappointing.46,64 As far as preclinical studies aimed at identifying targets to treat CIPN is concerned, most drugs have been shown to only prevent the CIPN. Our present data clearly show that ESI-09 is effective even when treatment is initiated up to 10 days after completion of paclitaxel treatment in both males and females (Fig. 1). Not all patients treated for cancer develop CIPN and therefore, the potential to target Epac signaling to alleviate already existing CIPN is promising. Before clinical translation can be considered, further studies are needed to examine the effect of Epac inhibition on pain and IENF density in additional rodent studies in which paclitaxel is given i.v. in different dosing schedules. It should be noted, however, that the cumulative dose of paclitaxel used here is in the range of what is used to treat tumors in mice.14,32,34 In addition to general analysis of potential toxic side effects, the potential interference of ESI-09 treatment with the anticancer effect of treatment needs to be examined. It is encouraging that ESI-09 inhibits invasion and metastasis of pancreatic cancer cells.2,4 These findings indicate that Epac1 inhibition during or after chemotherapeutic treatment may serve as a 2 pronged approach, on the one hand reducing CIPN and on the other hand reducing tumor cell metastasis.

ESI-09 acts as a competitive inhibitor of both Epac isoforms (Epac1 and Epac2) but has especially emerged as an attractive therapeutic option for blocking Epac1 cellular functioning.1,4,52 ESI-09 inhibits cAMP-mediated Epac1 GEF activity and inhibits GTP transfer to Rap1 in response to the specific Epac1 agonist 007-AM.1,4 In vivo treatment of ESI-09 recapitulates the phenotype of Epac1 knockout animals in multiple model systems. For example, ESI-09 administration and genetic deletion of Epac1 both inhibit pancreatic cancer metastasis,1 bacterial invasion in a model of rickettsiosis,26 complete Freund adjuvant–induced pain,52 and neointima formation in a mouse carotid artery ligation model.61 In this study, we demonstrate that genetic deletion of Epac1 recapitulates the effect of ESI-09 at different levels of CIPN, that is, at the level of inhibition of paclitaxel-induced mechanical allodynia, spinal cord glia activation, and IENF loss. This indicates Epac1 rather than Epac2 or other potential targets of the cAMP analogue ESI-09 is critical for CIPN. In addition, ex vivo, the competitive Epac inhibitor ESI-09 and the noncompetitive Epac1 inhibitor CE3F4 both suppress the paclitaxel-induced spontaneous discharges of DRG neurons.

Our findings showing that oral administration of the Epac inhibitor ESI-09 attenuates mechanical allodynia in paclitaxel-treated mice indicate that the persistent sensitization is dependent on persistent activation of a cAMP-to-Epac1 signaling pathway. In this respect, it is of interest that there is evidence for persistent activation of nociceptor membrane adenylyl cyclase7 in a model of neuropathic pain induced by spinal cord injury, and it is possible that the same is happening in response to paclitaxel. Moreover, in both the spinal cord injury model and the model of CIPN we use here, abnormal spontaneous activity is detected in the DRG and we show here that this is blocked by ESI-09. In addition, it is known that cAMP can be produced locally in mitochondria by soluble adenylyl cyclase which is activated by bicarbonate ions and regulated by cellular ATP and calcium levels.18,57,62 It would be interesting to explore whether the mitochondrial abnormalities that develop in response to paclitaxel treatment activate this cAMP-inducing pathway.

Epac1 signaling can also be activated independently of cAMP and we recently showed that cAMP-independent activation of Epac1 regulates neurite outgrowth in vitro.6 Specifically, we showed that a decrease in levels of the protein importin-β1 promotes translocation of Epac1 to the plasma membrane to inhibit neurite outgrowth. It would be interesting to investigate whether paclitaxel treatment affects the level of importin-β1 in IENFs with potential consequences for Epac1 regulation and IENF density.

It is known that cAMP-mediated activation of Epac1 leads to translocation of Epac1 to the plasma membrane and activation of Rap1 and other downstream effectors including PKCε.15,31 Activated PKCε can translocate to different subcellular compartments, including to the mitochondria where it can reduce complex-I respiration and Na+-K+-ATPase activity,45 potentially leading to mitochondrial insufficiency. Levine and group originally proposed a role for PKCε in paclitaxel-induced neuropathy and showed that paclitaxel-induced hyperalgesia was attenuated by an antagonist of PKCε.19 More recently, Wang et al. reported that paclitaxel induces long-term activation of PKCε, PKCβII, and PKCδ as monitored by their plasma membrane translocation in DRG neurons of paclitaxel-treated mice. Intrathecal administration of a PKCε inhibitor attenuated paclitaxel-induced mechanical allodynia. As a mechanism of action, these authors proposed that PKCε activates TRP and TRPV channels.30

Mitochondrial structural and functional abnormalities in peripheral sensory neurons have been documented in various animal models of CIPN.25,36,37,71 In paclitaxel- and oxaliplatin-treated rats, sciatic nerves showed decreased complex I- and complex II-mediated respiration and ATP production.70 We show here that in WT mice, paclitaxel treatment results in an overall decrease in mitochondrial OCR; basal, ATP coupled, and maximal respiration in the tibial nerve were all reduced indicating an overall reduction in mitochondrial content or mitochondrial health in the distal axons. Notably, we recently showed that prevention of mitochondrial damage using an inhibitor of mitochondrial p53 accumulation also prevented CIPN induced by either paclitaxel37 or cisplatin.42 These earlier findings indicate a causal relation between chemotherapy-induced mitochondrial deficits in sensory neurons and CIPN. Our present findings show that protection against CIPN by genetic deletion of Epac1 was associated with prevention of the paclitaxel-induced deficits in mitochondrial bioenergetics in tibial nerves in mice (Fig. 5), supporting the causal relation between mitochondrial damage and signs of CIPN. It remains to be determined how Epac1 activation contributes to mitochondrial deficiencies in models of CIPN and if this is a cAMP-dependent mechanism.

Interestingly, the protective effect of genetic deletion of Epac1 from all cells or from nociceptors only on mechanical allodynia was associated with prevention of the paclitaxel-induced loss of IENFs. Moreover, inhibition of Epac1 by oral administration of ESI-09 restored IENF density (Fig. 4). It is not known how the loss of IENFs in models of CIPN is regulated. The prevailing idea is that the reduction of IENF density is not due to neuronal death, but rather that reduced mitochondrial function leading to energy deficiency promotes retraction of distal endings. In line with this model, we showed earlier in the paclitaxel as well as the cisplatin model of CIPN that prevention of mitochondrial damage by the small compound PFT-µ not only protected against mitochondrial deficits but also against the reduction in IENF density.37,42 Our present findings support a model in which mitochondrial deficits in the distal axons are related to IENF loss.

Chemotherapy-induced peripheral neuropathy in rodent models is known to be associated with spontaneous firing by DRG neurons.40,65,68 Limited availability of ATP in the sensory axons is believed to generate spontaneous and irregular discharges.8,65,71 Here, we show that ex vivo ESI-09 treatment inhibits the abnormal spontaneous discharges in DRG neurons from paclitaxel-treated animals (Fig. 6). Similar results were obtained with a different Epac1 inhibitor, indicating that the effects are indeed mediated by inhibition of Epac1. We propose that at least part of the beneficial effect of ESI-09 on mechanical allodynia is mediated through the suppression of spontaneous activity in nociceptive neurons. Notably, the suppressive effect of ESI-09 on spontaneous discharges is rapid and reversible, indicating that persistent Epac activation is required for these spontaneous discharges. This observation is consistent with our observation that mechanical allodynia is inhibited but does not resolve in response to ESI-09 treatment. It would be interesting to know whether and how mitochondrial dysfunction in nociceptive neurons contributes to the spontaneous discharges.

In summary, we demonstrate that Epac1 is required for paclitaxel-induced mechanical allodynia and ESI-09 can be a potentially effective treatment to prevent and treat already existing paclitaxel-induced neuropathy. We propose that Epac1 inhibition could develop into an effective strategy to improve quality of life of patients with cancer.

Back to Top | Article Outline

Conflict of interest statement

The authors have no conflict of interest to declare.

Supported by grants NIH RO1 NS073939, NIH RO1 NS074999, NIH R21 CA183736, a STARS grant of the University of Texas System, and MD Anderson's Cancer Center Support Grant P30CA016672.

Back to Top | Article Outline

Supplemental digital content

Supplemental digital content associated with this article can be found online at

Back to Top | Article Outline


[1]. Almahariq M, Chao C, Mei FC, Hellmich MR, Patrikeev I, Motamedi M, Cheng X. Pharmacological inhibition and genetic knockdown of exchange protein directly activated by cAMP 1 reduce pancreatic cancer metastasis in vivo. Mol Pharmacol 2015;87:142–9.
[2]. Almahariq M, Mei FC, Cheng X. The pleiotropic role of exchange protein directly activated by cAMP 1 (EPAC1) in cancer: implications for therapeutic intervention. Acta Biochim Biophys Sinica 2016;48:75–81.
[3]. Almahariq M, Mei FC, Wang H, Cao AT, Yao S, Soong L, Sun J, Cong Y, Chen J, Cheng X. Exchange protein directly activated by cAMP modulates regulatory T-cell-mediated immunosuppression. Biochem J 2015;465:295–303.
[4]. Almahariq M, Tsalkova T, Mei FC, Chen H, Zhou J, Sastry SK, Schwede F, Cheng X. A novel EPAC-specific inhibitor suppresses pancreatic cancer cell migration and invasion. Mol Pharmacol 2013;83:122–8.
[5]. Argyriou AA, Polychronopoulos P, Koutras A, Xiros N, Petsas T, Argyriou K, Kalofonos HP, Chroni E. Clinical and electrophysiological features of peripheral neuropathy induced by administration of cisplatin plus paclitaxel-based chemotherapy. Eur J Cancer Care 2007;16:231–7.
[6]. Baameur F, Singhmar P, Zhou Y, Hancock JF, Cheng X, Heijnen CJ, Kavelaars A. Epac1 interacts with importin beta1 and controls neurite outgrowth independently of cAMP and Rap1. Sci Rep 2016;6:36370.
[7]. Bavencoffe A, Li Y, Wu Z, Yang Q, Herrera J, Kennedy EJ, Walters ET, Dessauer CW. Persistent electrical activity in primary nociceptors after spinal cord injury is maintained by scaffolded adenylyl cyclase and protein kinase a and is associated with altered adenylyl cyclase regulation. J Neurosci 2016;36:1660–8.
[8]. Bennett GJ, Doyle T, Salvemini D. Mitotoxicity in distal symmetrical sensory peripheral neuropathies. Nat Rev Neurol 2014;10:326–36.
[9]. Boyette-Davis J, Xin W, Zhang H, Dougherty PM. Intraepidermal nerve fiber loss corresponds to the development of taxol-induced hyperalgesia and can be prevented by treatment with minocycline. PAIN 2011;152:308–13.
[10]. Boyette-Davis JA, Walters ET, Dougherty PM. Mechanisms involved in the development of chemotherapy-induced neuropathy. PAIN 2015;5:285–96.
[11]. Brown MR, Ramirez JD, Farquhar-Smith P. Pain in cancer survivors. Br J pain 2014;8:139–53.
[12]. Cavaletti G, Marmiroli P. Chemotherapy-induced peripheral neurotoxicity. Curr Opin Neurol 2015;28:500–7.
[13]. Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994;53:55–63.
[14]. Choi CH, Ryu JY, Cho YJ, Jeon HK, Choi JJ, Ylaya K, Lee YY, Kim TJ, Chung JY, Hewitt SM, Kim BG, Bae DS, Lee JW. The anti-cancer effects of itraconazole in epithelial ovarian cancer. Scientific Rep 2017;7:6552.
[15]. Consonni SV, Gloerich M, Spanjaard E, Bos JL. cAMP regulates DEP domain-mediated binding of the guanine nucleotide exchange factor Epac1 to phosphatidic acid at the plasma membrane. Proceedings of the National Academy of Sciences of the United States of America, 2012;109:3814–19.
[16]. de Rooij J, Zwartkruis FJ, Verheijen MH, Cool RH, Nijman SM, Wittinghofer A, Bos JL. Epac is a Rap1 guanine-nucleotide-exchange factor directly activated by cyclic AMP. Nature 1998;396:474–7.
[17]. Decosterd I, Woolf CJ. Spared nerve injury: an animal model of persistent peripheral neuropathic pain. PAIN 2000;87:149–58.
[18]. Di Benedetto G, Scalzotto E, Mongillo M, Pozzan T. Mitochondrial Ca(2)(+) uptake induces cyclic AMP generation in the matrix and modulates organelle ATP levels. Cell Metab 2013;17:965–75.
[19]. Dina O, Chen X, Reichling D, Levine J. Role of protein kinase Cε and protein kinase A in a model of paclitaxel-induced painful peripheral neuropathy in the rat. Neuroscience 2001;108:507–15.
[20]. Dougherty PM, Cata JP, Cordella JV, Burton A, Weng HR. Taxol-induced sensory disturbance is characterized by preferential impairment of myelinated fiber function in cancer patients. PAIN 2004;109:132–42.
[21]. Eijkelkamp N, Linley J, Torres J, Bee L, Dickenson A, Gringhuis M, Minett M, Hong G, Lee E, Oh U. A role for Piezo2 in EPAC1-dependent mechanical allodynia. Nat Commun 2013;4:1682.
[22]. Eijkelkamp N, Singhmar P, Heijnen CJ, Kavelaars A. Sensory neuron cAMP signaling in chronic pain. Cyclic Nucleotide Signaling 2015;13:113.
[23]. Eijkelkamp N, Wang H, Garza-Carbajal A, Willemen HL, Zwartkruis FJ, Wood JN, Dantzer R, Kelley KW, Heijnen CJ, Kavelaars A. Low nociceptor GRK2 prolongs prostaglandin E2 hyperalgesia via biased cAMP signaling to Epac/Rap1, protein kinase Cε, and MEK/ERK. J Neurosci 2010;30:12806–15.
[24]. Ferreira SH, Nakamura M. I—prostaglandin hyperalgesia, a cAMP/Ca2+ dependent process. Prostaglandins 1979;18:179–90.
[25]. Flatters SJ, Bennett GJ. Studies of peripheral sensory nerves in paclitaxel-induced painful peripheral neuropathy: evidence for mitochondrial dysfunction. PAIN 2006;122:245–57.
[26]. Gong B, Shelite T, Mei FC, Ha T, Hu Y, Xu G, Chang Q, Wakamiya M, Ksiazek TG, Boor PJ, Bouyer DH, Popov VL, Chen J, Walker DH, Cheng X. Exchange protein directly activated by cAMP plays a critical role in bacterial invasion during fatal rickettsioses. Proceedings of the National Academy of Sciences of the United States of America, 2013;110:19615–20.
[27]. Grandoch M, Roscioni SS, Schmidt M. The role of Epac proteins, novel cAMP mediators, in the regulation of immune, lung and neuronal function. Br J Pharmacol 2010;159:265–84.
[28]. Gu Y, Wang C, Li G, Huang LM. EXPRESS: F-actin links Epac-PKC signaling to purinergic P2X3 receptors sensitization in dorsal root ganglia following inflammation. Mol pain 2016;12.
[29]. Hagiwara H, Sunada Y. Mechanism of taxane neurotoxicity. Breast Cancer (Tokyo, Japan) 2004;11:82–5.
[30]. He Y, Wang ZJ. Nociceptor beta II, delta, and epsilon isoforms of PKC differentially mediate paclitaxel-induced spontaneous and evoked pain. J Neurosci 2015;35:4614–25.
[31]. Hucho TB, Dina OA, Levine JD. Epac mediates a cAMP-to-PKC signaling in inflammatory pain: an isolectin B4 (+) neuron-specific mechanism. J Neurosci 2005;25:6119–26.
[32]. Ji W, Wang B, Fan Q, Xu C, He Y, Chen Y. Chemosensitizing indomethacin-conjugated dextran-based micelles for effective delivery of paclitaxel in resistant breast cancer therapy. PLoS One 2017;12:e0180037.
[33]. Jin HW, Flatters SJ, Xiao WH, Mulhern HL, Bennett GJ. Prevention of paclitaxel-evoked painful peripheral neuropathy by acetyl-L-carnitine: effects on axonal mitochondria, sensory nerve fiber terminal arbors, and cutaneous Langerhans cells. Exp Neurol 2008;210:229–37.
[34]. Jing C, Gao Z, Wang R, Yang Z, Shi B, Hou P. Lenvatinib enhances the antitumor effects of paclitaxel in anaplastic thyroid cancer. Am J Cancer Res 2017;7:903–12.
[35]. Kawasaki H, Springett GM, Mochizuki N, Toki S, Nakaya M, Matsuda M, Housman DE, Graybiel AM. A family of cAMP-binding proteins that directly activate Rap1. Science (New York, NY) 1998;282:2275–9.
[36]. Krukowski K, Ma J, Golonzhka O, Laumet GO, Gutti T, van Duzer JH, Mazitschek R, Jarpe MB, Heijnen CJ, Kavelaars A. HDAC6 inhibition effectively reverses chemotherapy-induced peripheral neuropathy. PAIN 2017;158:1126–37.
[37]. Krukowski K, Nijboer CH, Huo X, Kavelaars A, Heijnen CJ. Prevention of chemotherapy-induced peripheral neuropathy by the small-molecule inhibitor pifithrin-mu. PAIN 2015;156:2184–92.
[38]. Lee JJ, Swain SM. Peripheral neuropathy induced by microtubule-stabilizing agents. J Clin Oncol 2006;24:1633–42.
[39]. Li Y, Adamek P, Zhang H, Tatsui CE, Rhines LD, Mrozkova P, Li Q, Kosturakis AK, Cassidy RM, Harrison DS, Cata JP, Sapire K, Zhang H, Kennamer-Chapman RM, Jawad AB, Ghetti A, Yan J, Palecek J, Dougherty PM. The cancer chemotherapeutic paclitaxel increases human and rodent sensory neuron responses to TRPV1 by activation of TLR4. J Neurosci 2015;35:13487–500.
[40]. Li Y, Tatsui CE, Rhines LD, North RY, Harrison DS, Cassidy RM, Johansson CA, Kosturakis AK, Edwards DD, Zhang H, Dougherty PM. Dorsal root ganglion neurons become hyperexcitable and increase expression of voltage-gated T-type calcium channels (Cav3.2) in paclitaxel-induced peripheral neuropathy. PAIN 2017;158:417–29.
[41]. Li Y, Zhang H, Zhang H, Kosturakis AK, Jawad AB, Dougherty PM. Toll-like receptor 4 signaling contributes to paclitaxel-induced peripheral neuropathy. J Pain 2014;15:712–25.
[42]. Maj M, Ma J, Krukowski K, Kavelaars A, Heijnen C. Inhibition of mitochondrial p53 accumulation by PFT-μ prevents cisplatin-induced peripheral neuropathy. Front Mol Neurosci 2017;10:108.
[43]. Makker PG, Duffy SS, Lees JG, Perera CJ, Tonkin RS, Butovsky O, Park SB, Goldstein D, Moalem-Taylor G. Characterisation of immune and neuroinflammatory changes associated with chemotherapy-induced peripheral neuropathy. PLoS One 2017;12:e0170814.
[44]. Mao-Ying QL, Kavelaars A, Krukowski K, Huo XJ, Zhou W, Price TJ, Cleeland C, Heijnen CJ. The anti-diabetic drug metformin protects against chemotherapy-induced peripheral neuropathy in a mouse model. PLoS One 2014;9:e100701.
[45]. Nowak G, Bakajsova D, Clifton GL. Protein kinase C-epsilon modulates mitochondrial function and active Na+ transport after oxidant injury in renal cells. Am J Physiol Ren Physiol 2004;286:F307–316.
[46]. Pascual D, Goicoechea C, Burgos E, Martin MI. Antinociceptive effect of three common analgesic drugs on peripheral neuropathy induced by paclitaxel in rats. Pharmacol Biochem Behav 2010;95:331–7.
[47]. Peters CM, Jimenez-Andrade JM, Kuskowski MA, Ghilardi JR, Mantyh PW. An evolving cellular pathology occurs in dorsal root ganglia, peripheral nerve and spinal cord following intravenous administration of paclitaxel in the rat. Brain Res 2007;1168:46–59.
[48]. Polomano RC, Bennett GJ. Chemotherapy-evoked painful peripheral neuropathy. Pain Med (Malden, Mass) 2001;2:8–14.
[49]. Robinson CR, Zhang H, Dougherty PM. Astrocytes, but not microglia, are activated in oxaliplatin and bortezomib-induced peripheral neuropathy in the rat. Neuroscience 2014;274:308–17.
[50]. Rowinsky EK, Eisenhauer EA, Chaudhry V, Arbuck SG, Donehower RC. Clinical toxicities encountered with paclitaxel (Taxol). Semin Oncol 1993;20(4 suppl 3):1–15.
[51]. Siau C, Xiao W, Bennett GJ. Paclitaxel- and vincristine-evoked painful peripheral neuropathies: loss of epidermal innervation and activation of Langerhans cells. Exp Neurol 2006;201:507–14.
[52]. Singhmar P, Huo X, Eijkelkamp N, Berciano SR, Baameur F, Mei FC, Zhu Y, Cheng X, Hawke D, Mayor F Jr, Murga C, Heijnen CJ, Kavelaars A. Critical role for Epac1 in inflammatory pain controlled by GRK2-mediated phosphorylation of Epac1. Proceedings of the National Academy of Sciences of the United States of America, 2016;113:3036–41.
[53]. Suzuki S, Yokoyama U, Abe T, Kiyonari H, Yamashita N, Kato Y, Kurotani R, Sato M, Okumura S, Ishikawa Y. Differential roles of Epac in regulating cell death in neuronal and myocardial cells. J Biol Chem 2010;285:24248–59.
[54]. Taiwo Y, Heller P, Levine J. Mediation of serotonin hyperalgesia by the cAMP second messenger system. Neuroscience 1992;48:479–83.
[55]. Taiwo YO, Bjerknes LK, Goetzl EJ, Levine JD. Mediation of primary afferent peripheral hyperalgesia by the cAMP second messenger system. Neuroscience 1989;32:577–80.
[56]. Tsalkova T, Mei FC, Li S, Chepurny OG, Leech CA, Liu T, Holz GG, Woods VL Jr, Cheng X. Isoform-specific antagonists of exchange proteins directly activated by cAMP. Proceedings of the National Academy of Sciences of the United States of America, 2012;109:18613–18.
[57]. Valsecchi F, Konrad C, Manfredi G. Role of soluble adenylyl cyclase in mitochondria. Biochim Biophys Acta 2014;1842:2555–60.
[58]. Walsh D, Perkins JP, Krebs EG. An adenosine 3', 5'-monophosphate-dependant protein kinase from rabbit skeletal muscle. J Biol Chem 1968;243:3763–5.
[59]. Wang C, Gu Y, Li GW, Huang LY. A critical role of the cAMP sensor Epac in switching protein kinase signalling in prostaglandin E2-induced potentiation of P2X3 receptor currents in inflamed rats. J Physiol 2007;584:191–203.
[60]. Wang H, Heijnen CJ, van Velthoven CT, Willemen HL, Ishikawa Y, Zhang X, Sood AK, Vroon A, Eijkelkamp N, Kavelaars A. Balancing GRK2 and EPAC1 levels prevents and relieves chronic pain. J Clin Invest 2013;123:5023–34.
[61]. Wang H, Robichaux WG, Wang Z, Mei FC, Cai M, Du G, Chen J, Cheng X. Inhibition of Epac1 suppresses mitochondrial fission and reduces neointima formation induced by vascular injury. Sci Rep 2016;6:36552.
[62]. Wang Z, Liu D, Varin A, Nicolas V, Courilleau D, Mateo P, Caubere C, Rouet P, Gomez AM, Vandecasteele G, Fischmeister R, Brenner C. A cardiac mitochondrial cAMP signaling pathway regulates calcium accumulation, permeability transition and cell death. Cell Death Dis 2016;7:e2198.
[63]. Windebank AJ, Grisold W. Chemotherapy-induced neuropathy. J Peripher Nervous Syst 2008;13:27–46.
[64]. Xiao W, Boroujerdi A, Bennett GJ, Luo ZD. Chemotherapy-evoked painful peripheral neuropathy: analgesic effects of gabapentin and effects on expression of the alpha-2-delta type-1 calcium channel subunit. Neuroscience 2007;144:714–20.
[65]. Xiao WH, Bennett GJ. Chemotherapy-evoked neuropathic pain: abnormal spontaneous discharge in A-fiber and C-fiber primary afferent neurons and its suppression by acetyl-L-carnitine. PAIN 2008;135:262–70.
[66]. Xiao WH, Bennett GJ. Effects of mitochondrial poisons on the neuropathic pain produced by the chemotherapeutic agents, paclitaxel and oxaliplatin. PAIN 2012;153:704–9.
[67]. Yan J, Mei FC, Cheng H, Lao DH, Hu Y, Wei J, Patrikeev I, Hao D, Stutz SJ, Dineley KT, Motamedi M, Hommel JD, Cunningham KA, Chen J, Cheng X. Enhanced leptin sensitivity, reduced adiposity, and improved glucose homeostasis in mice lacking exchange protein directly activated by cyclic AMP isoform 1. Mol Cell Biol 2013;33:918–26.
[68]. Zhang H, Dougherty PM. Enhanced excitability of primary sensory neurons and altered gene expression of neuronal ion channels in dorsal root ganglion in paclitaxel-induced peripheral neuropathy. Anesthesiology 2014;120:1463–75.
[69]. Zhang H, Yoon SY, Zhang H, Dougherty PM. Evidence that spinal astrocytes but not microglia contribute to the pathogenesis of Paclitaxel-induced painful neuropathy. J Pain 2012;13:293–303.
[70]. Zheng H, Xiao WH, Bennett GJ. Functional deficits in peripheral nerve mitochondria in rats with paclitaxel- and oxaliplatin-evoked painful peripheral neuropathy. Exp Neurol 2011;232:154–61.
[71]. Zheng H, Xiao WH, Bennett GJ. Mitotoxicity and bortezomib-induced chronic painful peripheral neuropathy. Exp Neurol 2012;238:225–34.
[72]. Zhu Y, Chen H, Boulton S, Mei F, Ye N, Melacini G, Zhou J, Cheng X. Biochemical and pharmacological characterizations of ESI-09 based EPAC inhibitors: defining the ESI-09 “therapeutic window”. Scientific Rep 2015;5:9344.

Epac; CIPN; Neuropathy; ESI-09

Supplemental Digital Content

Back to Top | Article Outline
© 2018 International Association for the Study of Pain