The online version of this article (https://doi.org/10.1186/s12974-018-1116-6) contains supplementary material, which is available to authorized users.
C-C motif chemokine ligand 2
C-C chemokine receptor type 2
Chemotherapy-induced painful neuropathy
C-X3-C motif chemokine ligand 1
C-X3-C motif chemokine receptor 1
P38 mitogen-activated protein kinase
Phospho-p44/42 MAPK (Erk1/2)
Short interfering ribonucleic acid
Tamm-Horsfall protein 1
Tumour necrosis factor alpha
Transient receptor potential cation channel subfamily V member 1
Chemotherapy-induced painful neuropathy (CIPN) is a dose-limiting side effect of chemotherapeutic drugs, which is poorly managed by analgesics at present . Current treatments for CIPN such as gabapentin and opioids show limited efficacy and engender a host of undesirable side effects such as dizziness and nausea [2, 3]. Indeed, it has been reported that in as many as 40% of patients , CIPN results in the premature cessation of chemotherapy and thus jeopardises cancer treatment success. It is therefore essential that novel, more efficacious analgesics are developed and tailored according to different chemotherapeutic agents, which have a broad range of actions. This warrants a deeper understanding of the underlying drug-specific mechanisms of CIPN.
Most chemotherapeutic agents do not cross the blood-brain barrier but are able to penetrate the blood-nerve barrier, where they accumulate in dorsal root ganglia (DRG) and peripheral nerves, causing toxicity that is intensified by the absence of a lymphatic system in the endoneurial compartment [5–7]. Treatments that are currently used for CIPN are predominantly ‘neurocentric’ therapies, which target neuronal responses to injury inflicted by chemotherapeutic agents. Compelling evidence from preclinical models of CIPN in recent years however has uncovered the importance of immune cells, such as monocytes/macrophages, and their communication at the endothelial-neuronal interface peripherally [8, 9].
One chemotherapeutic agent associated with neuropathic pain is the vinca alkaloid, vincristine (VCR), used most commonly in the treatment of lymphomas and leukaemias . In a preclinical model of VCR-induced pain, VCR-mediated activation of the endothelium promotes the infiltration of CX3CR1-expressing monocytes into peripheral nerves, where they differentiate into macrophages and activate TRPA1 channels on sensory neurons via the release of reactive oxygen species . An increase in monocyte infiltration into the sciatic nerve is causative of mechanical hypersensitivity (allodynia), as opposed to being an epiphenomenon or downstream event, on the basis that transient depletion of macrophages delays the onset of allodynia until such depletion is ceased and macrophages can populate the sciatic nerve . Communication between monocytes/macrophages and sensory neurons in the periphery has been found to be a key underlying mechanism in preclinical CIPN associated with other chemotherapeutic agents such as paclitaxel. Indeed, depletion of macrophages in the DRG prevents paclitaxel-induced allodynia [11, 12]. Furthermore, inhibition of CX3CR1-mediated activation of macrophages specifically, as well as inhibition of p38 MAPK, a downstream target of CX3CR1 activation, also blocks the development of paclitaxel-induced allodynia .
Despite the promise of targeting CX3CR1 receptors in monocytes/macrophages as a prophylaxis for VCR pain, and, indeed, pain associated with other chemotherapeutic agents, a critical caveat remains: CX3CR1-deficient mice still develop VCR-induced allodynia, albeit delayed . It therefore appears that the role of CX3CR1 in monocytes is dynamic and alternative mechanisms emerge with increased exposure to VCR. In this study, we attempt to elucidate such a mechanism that could account for the eventual onset of VCR-induced allodynia in CX3CR1-deficient mice.
Circulating monocytes can, very broadly speaking, exist as two phenotypes—‘patrolling’ or ‘inflammatory’, each of which expresses distinct markers . Patrolling monocytes, characterised by low expression of Ly6C (Ly6Clow) and high levels of CX3CR1 (CX3CR1high), are predominantly associated with homeostatic functions . Inflammatory or ‘classical’ monocytes however express high levels of Ly6C (Ly6Chigh) and relatively low levels of CX3CR1 (CX3CR1low). Unlike patrolling monocytes, inflammatory monocytes express the chemokine receptor CCR2 (CCR2+). In the sciatic nerve, CCR2 receptor is activated by CCL2, which is expressed by endothelial cells, monocytes/macrophages and sensory neurons [15–17]. Inflammatory monocytes are recruited to sites of injury and inflammation in a CCL2/CCR2-regulated fashion, releasing a multitude of factors, many of which exert pronociceptive functions . It is well-established that neuronal injury, such as that caused by chemotherapy, is accompanied by an increase in macrophages with the inflammatory CCR2+ phenotype [19, 20].
In this study, we therefore investigate the role of CCL2/CCR2 signalling in inflammatory monocytes/macrophages in VCR-induced allodynia—specifically at later stages of treatment, when the role of CX3CR1 is less pertinent. Given that the CCR2 receptor is also expressed in sensory neurons , we also consider the role of CCR2-mediated neuronal activation in order to differentiate between the roles of CCR2 signalling in sensory neurons and monocytes. As well as attempting to elucidate the role of CCR2+ monocytes/macrophages in VCR-induced allodynia in CX3CR1-deficient mice, we also investigate a potential interaction between CX3CR1 and CCR2 expression in immortalised human monocytes (THP-1 cells). Furthermore, we assess the regulation of pronociceptive cytokines from THP-1 cell downstream of the CX3CR1-CCR2 interaction, which could constitute a potential monocyte-derived signal that mediates monocyte communication with nociceptive neurons, which are known to express cytokine receptors .
Experiments were performed in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986 and local animal care and use guidelines. Male and female mice (25–30 g, 12–17 weeks of age) were used. They did not differ with respect to thresholds. Mice were randomly assigned to groups. Each group contained approximately equal numbers of age-matched mice of both sexes.
For Cx3cr1-gfp (green fluorescent protein) mice, an original breeding stock was a kind gift from Steffen Jung (Weizmann Institute of Science, Israel). Cx3cr1-gfp heterozygous mice were generated and genotyped as described . For Ccr2-rfp (red fluorescent protein) mice, an original breeding stock was purchased from Jackson Laboratory. Ccr2 disruption/RFP expression was confirmed by PCR using published primers . All mice used were generated from a C57Bl/6 background.
Static mechanical withdrawal thresholds were assessed as described previously . Briefly, unrestrained animals were acclimatised for 45 min prior to testing. Calibrated von Frey filaments (0.008–1.4 g) were applied to the plantar surface of the hind paw until they bent and were held for 3 s or until paw withdrawal. A 50% paw withdrawal threshold (in grams) was calculated using the ‘up-down’ method starting with the 0.6-g filament. Three baseline (B) measurements were made prior to treatment, the mean of which is presented. Bilateral measurements were made with no difference in threshold being recorded between left and right (data not shown). Data presented are means of the left and right hind paw thresholds. All tests were conducted blind.
Vincristine sulphate (VCR; Sigma-Aldrich) was dissolved in sterile saline. 0.5 mg/kg/day VCR was injected intraperitoneally (i.p.) using a 25-g needle for two 5-day cycles (days 0–4 and 7–11) with a 2-day break as previously described ). RS-102895 hydrochloride (Sigma), which is small molecule antagonist of CCR2 receptor belonging to the spiro-piperidine class, was dissolved in dimethyl sulphoxide (DMSO) at 10 mg/ml and diluted in sterile saline; 20 mg/kg/day was injected i.p. for 5 days  during the second VCR cycle.
Mice, under pentobarbital anaesthesia, were transcardially perfused with saline, followed by 4% paraformaldehyde (PFA). The lumbar enlargement, bilateral L3–5 DRG, and sciatic nerves were excised. Tissue was post-fixed for 4 h and then dehydrated in 30% sucrose in PBS before being embedded in optimum cutting temperature-embedding medium (VWR) and frozen on dry ice. Fifteen-micrometer cryosections were mounted onto Superfrost Plus slides (VWR). Sections were permeabilised for 15 min at room temperature in 0.1% PBS-Triton X-100 (Sigma-Aldrich). Slides were blocked in 0.1% PBS-Triton X-100 + 3% BSA (Sigma-Aldrich) for 1 h at room temperature before being incubated in primary antibody at 4 °C for 16 h. Slides were washed thrice in 0.1% PBS-Triton X-100, before being incubated with the appropriate fluorescently tagged secondary antibody for 1.5 h at room temperature. Slides were washed again and then mounted using FluorSave™ (Merck, UK) and visualised under a Zeiss LSM710 confocal microscope (Zeiss). Five sections per mouse were randomly selected and analysed blind as described previously .
To visualise microglia, anti-Iba1 (1:100, WAKO) and anti-rabbit Alexa Fluor 488 (1:1000, Invitrogen) were used. For F4/80 expression, anti-F4/80 (1:400, Abcam) was used followed by anti-rat Alexa 488 or 568 (both 1:1000, Invitrogen). For CCL2 in the sciatic nerve, anti-CCL2 (1:200, Invitrogen) and anti-rabbit Alexa 568 (1:1000, Invitrogen) were used. P-ERK visualisation in DRG was carried out using anti-p-ERK (1:300, Cell Signaling Technology) followed by anti-rabbit Alexa 568 (1 in 1000 Invitrogen). GFP did not require amplification. In all cases, control staining was performed in which no primary antibody was applied.
Sciatic nerves were fresh-dissected, snap-frozen and stored at − 80 °C. Tissue was homogenised in ice-cold RIPA buffer (20 mM tris(hydroxymethyl)aminomethane; 10 mM NaF; 150 mM NaCl; 1% nonyl-phenoxylpolyethoxylethanol; 1 mM phenylmethanesulfonyl fluoride; 1 mM Na3VO4 (all Sigma-Aldrich); and 10 mg/ml proteinase inhibitor (Roche)) before being incubated at 4 °C with agitation for 2 h. Samples were centrifuged at 4 °C and 13,000 rpm for 20 min. Supernatant was used for Western blot. For THP-1 cultures, medium was spun at 2000g for 5 min at 4 °C. The pellet was washed, re-suspended in radioimmunoprecipitation assay (RIPA) buffer, vortexed and kept on ice for 30 min, and then spun at 13,000 rpm at 4 °C for 10 min. The supernatant was collected. Protein concentration was determined using the bicinchoninic acid (BCA) protein assay (Pierce). Twenty micrograms of protein per sample was separated on a 15% SDS-PAGE gel and transferred onto a nitrocellulose membrane which was performed according to the manufacturer’s instructions (Bio-Rad, UK). Blots were blocked for 1 h at room temperature in 0.1% TBS-Tween-20 (TBST) + 5% skimmed milk before being incubated in primary antibody (F4/80), CCL2 (anti-mouse or human), CCR2 (anti-mouse or human) (all 1:1000, Abcam) and p-ERK (1:1000 Cell Signaling Technology) at 4 °C overnight along with a loading control (α-tubulin, β-actin both 1:1000, Abcam). After washing in 0.1% TBST, blots were incubated for 1 h at room temperature with HRP-conjugated secondary antibody (1:2000, DAKO). Blots were washed before being developed with SuperSignal™ West Femto substrate (Thermo Fisher Scientific). Bands were visualised using BioSpectrum© Imaging and quantified using Quantity One (Bio-Rad, UK).
Mice were sacrificed by a rising concentration of CO2. Peritoneal cells were then elicited by injecting saline into the peritoneal cavity. Lavages were collected and centrifuged at 2400 rpm for 10 min. Pellets were suspended in red phenol-free DMEM (Gibco) supplemented with 10% heat-inactivated foetal bovine serum (HI-FBS), 1% pen/strep and 1% sodium pyruvate. Cells were incubated with conjugated antibodies (CD45.1-Pacific Blue (BioLegend); F4/80-PE (eBioscience); Alexa Fluor® 647 anti-mouse CD192 (CCR2) Antibody (BioLegend); Ly-6C APC (eBioscience)) on ice for 30 min and then washed twice with PBS containing 1.5% BSA before being re-suspended prior to analysis. Cells were gated and analysed with LSRFortessa™ (BD bioscience) and FlowJo software (Tree Start).
THP-1 cells were a kind gift from Mauro Perretti (Queen Mary, University of London). Cells were cultured in RPMI 1640 with 10% FBS and 2 mM glutamine (Sigma-Aldrich) and maintained in upright T25 tissue culture flasks. For all experiments, cells were cultured in 24-well plates. Cells were transfected using Lipofectamine RNAiMAX (Thermo Fisher Scientific) with CX3CR1 siRNA (Insight Biotechnology) or control SignalSilence®siRNA (New England Biolabs) 24 h after seeding, and assays were performed 48 h post-transfection. Efficiency of CX3CR1 knockdown was determined using Western blot. In some experiments, cells were pre-treated, 1 h before transfection with siRNA, with SB203580 or PD98059 (both 25 μM, Cambridge Bioscience). For CCL2 stimulation, cells were treated for 3 h with human recombinant CCL2 (Cambridge Bioscience) at 10, 50 and 100 ng/ml in HEPES buffer.
CCR2 expression was measured using human chemokine C-C-motif receptor 2 (CCR2) ELISA Kit (2B Scientific) according to the manufacturer’s instructions. For cytokine expression, a multi-analyte ELISA for proinflammatory cytokines (Qiagen) was performed on culture medium to assess which cytokines were upregulated. ELISA was then performed specifically for TNFα and IL1β (Abcam) according to the manufacturer’s instructions.
All data were analysed using SPSS (IBM Analytics). Behavioural data were analysed by repeated measures (RM) two-way ANOVA, followed by Tukey’s test. All other in vivo data were analysed by one-way ANOVA, followed by Tukey’s test. All in vitro data was analysed using Student’s paired t test. All data are shown as mean ± SEM, and data were considered significant when p < 0.05.
In order to establish whether CCR2 receptor signalling plays a role in VCR-induced allodynia, CCR2-deficient mice (CCR2RFP/RFP) were treated with two 5-day cycles of VCR [9, 24]. In the first cycle, they were found to develop significant VCR-induced allodynia to the same severity and within the same time frame as heterozygous (CCR2+/RFP), littermate controls (Fig. 1a). Specifically, within 24 h of the first VCR dose and throughout the first treatment cycle, mechanical withdrawal thresholds of both CCR2-deficient and heterozygous mice were significantly reduced relative to saline-treated mice. Indeed, there was no difference in threshold between VCR-treated CCR2-deficient and heterozygous mice throughout the first VCR cycle (days 0–4). During the second cycle of VCR however, CCR2-deficient mice had significantly higher thresholds, thus reduced allodynia, relative to heterozygous controls (Fig. 1a).
We initially confirmed that VCR administration did not cause a measurable microglial response (Additional file 1: Figure S1A, B), before focussing our attention on the sciatic nerve, where VCR accumulates and mechanisms underlying preclinical VCR pain are most pertinent . In CCR2 heterozygous mice, the elevation of macrophages in the sciatic nerve occurred to the same degree during both the first and second cycles of VCR treatment as demonstrated by immunohistochemical (Fig. 1b, c) and Western blot detection (Fig. 1d, e) of the macrophage marker F4/80. In CCR2-deficient mice however, when allodynia was significantly reduced during the second cycle relative to the first cycle, we observed that F4/80 expression was also significantly reduced in the sciatic nerve relative to the first VCR cycle (Fig. 1b–e).
In light of the apparent involvement of the CCR2 receptor in VCR allodynia during the second VCR cycle, we wanted to establish if treatment with a CCR2 antagonist on VCR-induced allodynia mirrors genetic deficiency, as well as identify whether CCR2 does indeed play a role in VCR-induced allodynia in CX3CR1-deficient mice during the second cycle. To address the first of these questions, we began by administering the CCR2 antagonist RS-102895 during the second VCR cycle to CX3CR1 heterozygous mice, which possess functional CX3CR1. Our rationale for therapeutic (VCR cycle 2) as opposed to prophylactic (VCR cycle 1) dosing was derived not only from the apparent effect of CCR2 deficiency on VCR-induced allodynia during the second VCR cycle specifically but also from previous observations that whilst the development of VCR-induced allodynia is orchestrated by CX3CR1 signalling in monocytes, other mechanisms appear to regulate allodynia during the second VCR cycle . Indeed, prophylactic treatment with RS-102895 during the first VCR cycle had no effect on withdrawal thresholds (Additional file 1: Figure S2). However, we found that when CX3CR1 heterozygous mice were given RS-102895 during the second VCR cycle, there was still no change in VCR-induced allodynia (Additional file 1: Figure S3A). Concurrently, VCR-induced monocyte/macrophage infiltration into the sciatic nerve was also unaltered by administration of RS-102895 during the second VCR cycle (Additional file 1: Figure S3B, C).
Although our data so far suggests that CCR2 antagonism during the second VCR cycle is not necessarily therapeutic in mice possessing functional CX3CR1 receptor, this does not rule out a role for CCR2-mediated signalling in VCR-induced allodynia that has been observed in CX3CR1-deficient mice at later stages of treatment . We therefore treated CX3CR1-deficient (CX3CR1GFP/GFP) mice with two cycles of VCR in the presence of RS-102895 (or vehicle) during the second VCR cycle, when allodynia manifests in CX3CR1-deficient mice . As observed previously, CX3CR1-deficient mice displayed a delayed onset of allodynia in response to VCR treatment relative to heterozygous littermates. Whilst CX3CR1 heterozygous mice developed significant allodynia within 24 h of the first VCR dose (Additional file 1: Figure S3A), mechanical thresholds only dropped significantly in CX3CR1-deficient mice relative to saline-treated controls at the end of the first VCR cycle (Fig. 2a). In addition, dissimilar to the results obtained in CX3CR1 heterozygous mice, systemic treatment of CX3CR1-deficient mice with RS-102895 during the second VCR cycle significantly reduced VCR-induced allodynia (Fig. 2a). Although withdrawal thresholds in CX3CR1-deficient mice treated with RS-102895 alongside VCR were still significantly lower than baseline levels, a significant increase relative to CX3CR1-deficient mice treated with VCR alone was observed within 24 h of the first RS-102895 dose and persisted throughout the second VCR cycle. Crucially, administration of RS-102895 in all saline-treated mice did not alter mechanical thresholds (Fig. 2a, Additional file 1: Figure S3). This data exposes a potential role for CCL2/R2 signalling in VCR-induced allodynia in CX3CR1-deficient mice during the second cycle specifically.
Given that RS-102895 significantly reduced VCR-induced allodynia in CX3CR1-deficient mice during the second VCR cycle, we assessed whether it also resulted in a simultaneous reduction in VCR-induced monocyte infiltration into the sciatic nerve. In CX3CR1 heterozygous mice, RS-102985 administration during the second VCR cycle did not significantly reduce VCR-induced monocyte infiltration into the sciatic nerve, as demonstrated by immunohistochemistry (Fig. 2b, c) and Western blot analysis (Fig. 2d, e) of F4/80. However, the administration of RS-102895 to VCR-treated CX3CR1-deficient mice did indeed result in a significant reduction in macrophage number in the sciatic nerve (Fig. 2b–e). This suggests that the mechanism by which RS-102895 significantly reduces VCR-induced allodynia in CX3CR1-deficient mice is, at least in part, regulated by CCR2 signalling in, or CCR2-mediated recruitment of, monocytes/macrophages in peripheral nerves. Furthermore, the reduction in VCR-induced allodynia and monocyte infiltration that accompanied the administration of RS-102895 to CX3CR1-deficient mice specifically is suggestive of changes in CCR2 expression and/or activity as a result of loss of CX3CR1 in monocytes/macrophages in the periphery.
Unlike CX3CR1, which is predominantly expressed by monocytes/macrophages, CCR2 is additionally expressed on sensory neurons [16, 17]. To investigate the potential contribution of CCR2 signalling in sensory neurons to the reduction in VCR-induced allodynia observed in RS-102895-treated CX3CR1-deficient mice, we considered the effect of RS-102895 on CCR2-mediated neuronal activation. In both CX3CR1-deficient and heterozygous mice, two cycles of VCR administration resulted in the upregulation of phospho-p44/42 MAPK2 (p-ERK2), a marker for neuronal activation and downstream target of CCR2 signalling [25, 26] as demonstrated by both immunohistochemistry (Fig. 3a, b) and Western blot analysis in lumbar DRG (Fig. 3c, d). Administration of RS-102895 for the duration of the second VCR cycle however resulted in a significant reduction of p-ERK2 in lumbar DRG from both CX3CR1-deficient and heterozygous mice (Fig. 3a–d) suggesting that the effect of RS-102895 on VCR-induced elevation of p-ERK2 did not differ between genotypes. Crucially, at day 14 (3 days after treatment cessation), VCR-treated CX3CR1-deficient mice that had received RS-102895 still had significantly reduced allodynia relative to CX3CR1 heterozygous mice (Fig. 3e). However, p-ERK2 activation was observed in lumbar DRG at this time point regardless of genotype, as demonstrated by immunohistochemistry (Fig. 3f, g) and Western blot analysis (Fig. 3h). This could suggest that the reduction in VCR-associated allodynia in RS-102895-treated CX3CR1-deficient mice is not entirely dependent on the diminution of sensory neuronal activation via CCR2 receptor antagonism and could therefore also be regulated by CCR2 signalling elsewhere, such as that present in monocytes/macrophages.
This data also indicates that it is less likely that the absence of effect of RS-102895 in CX3CR1 heterozygous mice was due to lack of bioavailability. Activation of p-ERK is a well-established downstream event of CCR2 receptor activation; thus, the fact that VCR-induced p-ERK activation was only reduced when RS-102895 was administered and not at day 14, at which point treatment was terminated (Fig. 3e–g), provides us with a measureable outcome of RS-102895 bioavailability and target engagement.
Given the effect of RS-102895 treatment on CX3CR1-deficient mice specifically and that CX3CR1 deficiency is associated with CCL2 upregulation in phagocytic cells [27, 28], we went on to examine CCL2 expression in the sciatic nerve of CX3CR1-deficient mice. Western blot analysis of CCL2 expression in sciatic nerve homogenates from CX3CR1-deficient and heterozygous mice that had been treated with saline only demonstrated that CCL2 elevation in the sciatic nerve accompanied CX3CR1 deficiency (Fig. 4a, b). In addition, we found that two VCR cycles increased CCL2 expression in the sciatic nerve of both CX3CR1-deficient and heterozygous mice (Fig. 4a, b). Immunohistochemical analysis of GFP+ (macrophages) and CCL2+ profiles in the sciatic nerve suggested that increases in CCL2 in the sciatic nerve in CX3CR1-deficient mice and in both genotypes in response to VCR treatment were, at least in part, a result of an increase of CCL2 in macrophages (Fig. 4c, d). In order to confirm that the results obtained were not a misrepresentation afforded by potential antibody artifacts, we confirmed that CCL2 could not be detected in sciatic nerve tissue from CCL2-deficient mice (data not shown).
Our data to this point demonstrates that (i) treatment with a CCR2 antagonist (RS-102895) reduces VCR-induced allodynia and monocyte/macrophage infiltration in sciatic nerves in CX3CR1-deficient mice during cycle two specifically and (ii) VCR induces CCL2 elevation in monocytes/macrophages. However, whether VCR allodynia is associated with peripheral nerve infiltration of CCR2-expressing monocytes, which provide a cellular target for RS-102895, has yet to be determined. We therefore considered evaluating monocytes/macrophages in peripheral nerve tissue specifically; however, the yield obtained during isolation was suboptimal for reliable characterisation. We therefore used peritoneal lavages as a surrogate source of monocytes . Importantly, we confirmed that CCR2 receptor is indeed detectable in both control and lipopolysaccharide (LPS)-stimulated peritoneal cells using both Western blot and FACS analysis of Ly6C+ CCR2+ cells (Additional file 1: Figure S4).
In light of our evidence that CCR2 receptor is detectable, we measured peritoneal CCR2+ monocytes (F4/80− cells) in CX3CR1-deficient and heterozygous mice following two cycles of VCR using FACS analysis. We observed that in both genotypes, two cycles of VCR treatment significantly increased F4/80−Ly6C+CCR2+ cells relative to saline-treated controls (Fig. 4e, f). Furthermore, the number of F4/80−Ly6C+CCR2+ cells in VCR-treated CX3CR1-deficient mice was significantly greater than that measured in heterozygous controls, indicating that CX3CR1 deficiency was associated with an increase of CCR2+ monocyte numbers in vivo. Critically, treatment with RS-102895 during the second VCR cycle significantly reduced the number of F4/80−Ly6C+CCR2+ cells to a greater extent in CX3CR1-deficient mice (Fig. 4e, f). Taken together, this data is suggestive of an interaction between CX3CR1 and CCR2 receptors in monocytes.
In order to investigate whether CX3CR1 regulates CCR2 expression in monocytes, we knocked down CX3CR1 expression using siRNA in the human monocyte THP-1 cell line, which resulted in an approximately 80% reduction in CX3CR1 expression (Additional file 1: Figure S5A). As shown previously , we confirmed that downregulation of the CX3CR1 receptor in monocytes is accompanied by an increase in CCL2 expression (Fig. 5a). In addition, however, a significant increase in CCR2 expression was also observed, with CX3CR1 siRNA-transfected THP-1 cells expressing significantly higher levels of CCR2 relative to those transfected with control siRNA (Fig. 5b, Additional file 1: Figure S4B). Furthermore, when cells were pre-treated with the p38 MAP kinase inhibitor, SB203580, downregulation of CX3CR1 (which is unaffected by SB203580 administration (Additional file 1: Figure S5C)) did not result in an increase in CCR2 expression (Fig. 5c). When cells were pre-treated with the MAPK/ERK kinase inhibitor, PD98059, however, CX3CR1 downregulation resulted in an increase in CCR2 expression (Fig. 5c). Taken together, these data suggest that downregulation of CX3CR1 in THP-1 cells in vitro results in upregulation of CCR2 receptor via p38 MAP Kinase signalling.
In order to begin to identify a mechanism by which an interaction between CX3CR1 and CCR2 in monocytes could potentially result in mechanical allodynia, we then went on to establish whether this interaction regulates the release of pronociceptive mediators. Indeed, downregulation of CX3CR1 resulted in a significant increase in the basal release of the proinflammatory cytokines TNF-α and IL1β to the same extent as stimulation of non-transfected THP-1 cells with 10 ng/ml CCL2 (Fig. 5d, e). Furthermore, release of TNF-α and IL1β, following stimulation with higher doses of CCL2 (50 and 100 ng/ml), was not significantly higher than those observed following CX3CR1 receptor downregulation alone (Fig. 5d, e).
In short, our THP-1 cell data demonstrates the presence of a novel interaction between CX3CR1 and CCR2 receptor expression, which is mediated by p38 MAP kinase signalling. In addition, it demonstrates that the CX3CR1-CCR2 interaction in vitro regulates the release of pronociceptive cytokines, thus providing a potential means by which monocytes could communicate with sensory neurons in the mediation of allodynia in CX3CR1-deficient mice during the second VCR cycle.
Chemotherapy-induced painful neuropathy (CIPN) is a dose-limiting side effect that jeopardises the success of cancer therapy . The need for novel, more effective pain therapies, which necessitates a deeper understanding of the underlying mechanisms, is therefore profound. Different chemotherapeutic agents are likely to induce CIPN via a range of mechanisms, and thus, future therapies are likely to be most effective if they are tailored for specific cancer treatments.
In this study, we elucidate a mechanism underlying pain associated with vincristine (VCR) treatment and present evidence to suggest that CCR2 receptor signalling in monocytes infiltrating the sciatic nerve plays a role in VCR-induced allodynia in CX3CR1-deficient mice during the second cycle specifically. We begin to elucidate a novel interaction between CX3CR1 and CCR2 in monocytes and demonstrate in vitro that it regulates the release of proinflammatory cytokines that have been previously shown to modulate neuronal mechanisms underlying chronic pain [23, 29].
We initially show that in CCR2-deficient mice, VCR-induced allodynia and concurrent monocyte infiltration into the sciatic nerve develop to the same severity and within the same time frame as CCR2 heterozygous littermates but are both significantly reduced during the second VCR cycle. Our findings from CCR2-deficient mice however is not enough to confirm a role for CCR2 signalling in monocytes per se in VCR-induced pain. CCR2 mediates the egression of monocytes from the bone marrow into the bloodstream where they survey the environment surrounding tissue [30, 31]. It is therefore plausible that the reduction in allodynia that we observe is a result of prevention of monocyte egression as opposed to CCR2 signalling in macrophages in the sciatic nerve specifically. Indeed, pharmacological inhibition of CCR2 affords a very different scenario biologically to CCR2 deficiency and as such is accompanied by different behavioural and biological manifestations. Specifically, when we administered the CCR2 antagonist RS-102895 to VCR-treated mice during the second cycle, we observed that neither VCR-induced allodynia nor monocyte infiltration into the sciatic nerve was reduced.
Our pivotal observation however was that RS-102895 administration during the second VCR cycle did significantly reduce established VCR-induced allodynia and monocyte infiltration into the sciatic nerve of CX3CR1-deficient mice. Indeed, in keeping with this, CCR2 receptor inhibition in CX3CR1-deficient mice specifically has also been found to be protective in other pathological conditions such as macular degeneration .
In addition we found that the (i) VCR-induced increase in CCR2+ monocytes in the peritoneal cavity was exacerbated in CX3CR1-deficient mice and (ii) RS-102895 treatment significantly reduced the VCR-induced increase in CCR2+ monocytes in CX3CR1-deficient mice specifically. Taken together, these data are suggestive of an interaction between CX3CL1/CX3CR1 and CCL2/CCR2 chemokine/receptor systems in monocytes that could underlie the onset of VCR-induced allodynia in CX3CR1-deficient mice.
Interactions between CX3CR1 receptor and CCL2 have been reported previously in other contexts. Indeed, in addition to our observation that CCL2 is elevated in the sciatic nerve of CX3CR1-deficient mice, the expression of CCL2 has also been shown to be elevated in mononuclear phagocytes in the retina in CX3CR1-deficient mice . Here, we present novel evidence that CCR2 receptor expression is also altered downstream of CX3CR1 downregulation in vitro. Specifically, when we downregulated CX3CR1 expression in immortalised human monocyte THP-1 cells, we observed a significant upregulation in both CCL2 and CCR2. Pharmacological inhibition of p38 MAPK but not MEK prevented this effect, and thus, we suggest that downregulation of CX3CR1 results in upregulation of the CCR2 receptor via p38 MAPK signalling.
As well as CX3CR1-mediated regulation or CCL2 and CCR2 expression in monocytes, the converse interaction has also been observed previously. Specifically, CCL2 stimulation of monocytes has been found to increase CX3CR1 receptor expression at monocyte membranes, most likely due to the stimulation of receptor translocation from intracellular stores as opposed to the stimulation of de novo expression . This regulation has been shown to occur via CCR2-mediated activation of p38 MAPK, with pharmacological inhibition of p38 MAPK, attenuating the effect . This data, in conjunction with the data presented in our study, suggests that the interaction between CX3CR1 and CCR2 receptors in monocytes is reciprocal and that expression of CX3CR1 and CCR2 is mutually regulated.
As well as CX3CR1 deficiency being associated with an increase in CCL2-CCR2 expression, VCR treatment was also found to increase CCL2. Elevated CCL2/CCR2 signalling in CX3CR1-deificient mice under basal conditions could explain the efficacy of RS-102895. The basal elevation of CCL2/CCR2 signalling alone in CX3CR1-deficient mice is not likely to be sufficient to cause allodynia as withdrawal thresholds did not vary between saline-treated CX3CR1-deficient and heterozygous mice. However, the potentially additive effect of CX3CR1 deficiency and two cycles of VCR on CCL2/CCR2 expression could indeed result in susceptibility to hypersensitivity. This could account for the onset of allodynia in CX3CR1-deficient mice at later stages of VCR treatment and the therapeutic benefit of CCR2 inhibition, which could also be attributed to increases in CCR2+ monocytes found in CX3CR1-deficient mice. Indeed, CCL2/R2 signalling in sensory neurons is a well-established mediator of pain. The elevation of CCR2+ monocytes and CCL2 expression in sciatic nerves that we observe in CX3CR1-deficient mice, in addition to VCR-induced increases in CCL2, is likely to cumulatively increase activity of the CCL2/CCR2 signalling axis in sensory neurons, thus activating pain pathways [32–34].
Whereas CX3CR1 is mainly found in monocytes/macrophages, the CCR2 receptor is expressed by other cell types. Elucidating the role of CCR2 signalling in monocytes/macrophages specifically is therefore a more intricate process than unravelling the role of CX3CR1 signalling in monocytes. For instance, the CCR2 receptor is also expressed in sensory neurons . To attribute at least part of our observed effect of RS-102895 in CX3CR1-deficient mice to inhibition of CCR2 in monocytes specifically, we needed to consider a role for CCR2 inhibition in sensory neurons. A well-established downstream target of CCR2 signalling in neurons is p-ERK activation, which is associated with noxious activation of sensory neurons . At the end of two cycles of VCR treatment, we found that p-ERK2 is elevated in DRG neurons in both CX3CR1-deficient and heterozygous mice. However, RS-102895 reduced VCR-associated p-ERK2 expression to the same extent in both genotypes despite the fact that allodynia is significantly reduced in CX3CR1-deficient mice. Although this does not rule out a role for CCR2-mediated neuronal activation in VCR allodynia, since thresholds observed in CX3CR1-deficient mice are still considered to be allodynic, it could suggest that other CCR2-mediated mechanisms are just as pivotal in regulating the severity of allodynia. Another key observation was that after the treatment was terminated, p-ERK2 activation in VCR-treated mice that also received RS-102895 was no longer reduced, regardless of the genotype. Crucially, however, allodynia was still significantly reduced in CX3CR1-deficient mice, despite measureable p-ERK2 activation. This could suggest one of two things: either the reduction in VCR allodynia in RS-102895-treated CX3CR1-deficient mice does not depend on the effect on neuronal activation or the dominant underlying mechanisms regulating VCR allodynia change as VCR pain evolves into ongoing after VCR treatment is terminated. Regardless of the explanation, however, our data suggests that there is scope for CCR2-mediated mechanisms that do not only require activation of CCR2 in sensory neurons to regulate VCR-associated allodynia in CX3CR1-deficient mice. CCR2 signalling in monocytes could be a candidate for such a mechanism.
Indeed, our final, crucial observation was that downregulation of CX3CR1 in human monocytes in vitro regulated basal release of the proinflammatory cytokines TNFα and IL1β. It is well-established that both TNFα and IL1β are pronociceptive [29, 35], and thus, this provides us with a crucial link between the CX3CR1-CCR2 interaction in monocytes and noxious signalling. Indeed, we confirmed that CCL2 stimulation of THP-1 cells also results in TNFα and IL1β release and observed that both CCL2 and CCR2 are elevated in THP-1 cells when CX3CR1 is downregulated thus suggesting that downregulation of CX3CR1 results in the release of proinflammatory cytokines through heightened CCL2/CCR2 signalling. Our data suggests the presence of a novel interaction between CX3CR1 and CCR2 receptors, which regulates the release of monocyte-derived signals that mediate communication between monocytes and nociceptive neurons (Fig. 6). Critically, in vivo, we observe that under basal conditions, CX3CR1 deficiency is associated with an increase in CCL2 in macrophages in the sciatic nerve and that VCR treatment increases CCL2 expression further, which would cumulatively result in the release of pronociceptive cytokines and thus mediate pain signalling.
Our data provides exciting evidence for a novel role of CCR2 receptor signalling in monocytes in VCR-associated allodynia in CX3CR1-deficient mice, which is likely to manifest as a result of an interaction between CX3CR1 and CCR2 expression in monocytes in the periphery. Identification of this interaction could thus uncover a new therapeutic target for longer-term treatment of vincristine pain.
We thank Thomas Pitcher for the technical assistance.
KM is supported by the Medical Research Council, UK (MR/M023893/1).
The datasets used and/or analysed during the current study are available from the corresponding author on a reasonable request.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Additional file 1: Table S1. Antibodies used for immunohistochemistry. Table S2. Antibodies used for Western blot analysis. Figure S1. No microglial response is detectable in VCR-treated CCR2 heterozygous or knockout mice. Figure S2. Prophylactic treatment of CX3CR1 heterozygous mice with RS-102895 does not affect the onset or severity of allodynia in cycle 1. Figure S3. Treatment with RS-102895 does not reduce VCR-associated allodynia in CX3CR1+/GFP mice. Figure S4. Peritoneal monocytes/macrophages express CCR2 under basal conditions. Figure S5. Transfection of THP-1 cells with CX3CR1 siRNA downregulates CX3CR1 expression and upregulates CCR2 expression via p39 MAP kinase. (DOCX 960 kb)
Seretny M, Currie GL, Sena ES, Ramnarine S, Grant R, MacLeod MR, Colvin LA, Fallon M. Incidence, prevalence, and predictors of chemotherapy-induced peripheral neuropathy: a systematic review and meta-analysis. Pain. 2014;155(12):2461–70. https://doi.org/10.1016/j.pain.2014.09.020. CrossRefPubMed
Dougherty PM, Cata JP, Burton AW, Vu K, Weng HR. Dysfunction in multiple primary afferent fiber subtypes revealed by quantitative sensory testing in patients with chronic vincristine-induced pain. J Pain Symptom Manag. 2007;33(2):166–79. CrossRef
Bloodworth D. Issues in opioid management. Am J Phys Med Rehabil. 2005;84(Suppl 3):S42–55. PubMed
Langhert KA, Von Zee CL, Stubbs Jr EB. Tumour necrosis factor α enhances CCL2 and ICAM-1 expression in peripheral nerve microvascular endoneurial endothelial cells. ASN Neuro. 2013;5(1):e00104.
Zhang H, Boyette-Davis JA, Kosturakis AK, Li Y, Yoon SY, Walters ET, Dougherty PM. Induction of monocyte chemoattractant protein-1 (MCP-1) and its receptor CCR2 in primary sensory neurons contributes to paclitaxel-induced peripheral neuropathy. J Pain. 2013;14(10):1031–44. CrossRefPubMedPubMedCentral
Miller RJ, Jung H, Bhangoo SK, White FA. Cytokine and chemokine regulation of sensory neuron function. Handb Exp Pharmacol. 2009;194:417–49. CrossRef
Saederup N, Cardona AE, Croft K, Mizutani M, Cotleur AC, Tsou CL, Ransohoff RM, Charo IF. Selective chemokine receptor usage by central nervous system myeloid cells in CCR2-red fluorescent protein knock-in mice. PLoS One. 2010;27:5(10).
Yuan F, Yosef N, Lakshmana Reddy C, Huang A, Chiang SC, Tithi HR, Ubogu EE. CCR2 gene deletion and pharmacologic blockade ameliorate a severe murine experimental autoimmune neuritis model of Guillain-Barré syndrome. PLoS One. 2014;14:9(3).
- A novel interaction between CX3CR1 and CCR2 signalling in monocytes constitutes an underlying mechanism for persistent vincristine-induced pain
- BioMed Central