If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
University Hospital Würzburg, Department of Anesthesiology, Intensive Care, Emergency Care and Pain Management, Center for Interdisciplinary Pain Medicine, Würzburg, Germany
University Hospital Würzburg, Department of Anesthesiology, Intensive Care, Emergency Care and Pain Management, Center for Interdisciplinary Pain Medicine, Würzburg, GermanyUniversity Hospital Würzburg, Department of Neurosurgery, Tumorbiology Laboratory, Würzburg, Germany
University Hospital Würzburg, Department of Anesthesiology, Intensive Care, Emergency Care and Pain Management, Center for Interdisciplinary Pain Medicine, Würzburg, Germany
Address reprint requests to Heike L. Rittner, MD, University Hospital Würzburg, Center for Interdisciplinary Pain Medicine, Department of Anesthesiology, Intensive Care, Emergency Care and Pain Management, Oberdürrbacher Street 6, 97080 Würzburg, Germany.
University Hospital Würzburg, Department of Anesthesiology, Intensive Care, Emergency Care and Pain Management, Center for Interdisciplinary Pain Medicine, Würzburg, Germany
miR-183 is reduced in chronic constriction injury, local substitution reverses hyperalgesia.
•
Circulating exosomes from CRPS patients contain less miR-183.
•
Sera of both CRPS patients and neuropathic mice impair the microvascular barrier.
•
miR-183 stabilizes the microvascular barrier via FOXO1-mediated claudin-5 increase.
•
miR-183, FoxO1 and claudin-5 form a functional pathway in neuropathy and CRPS.
Abstract
Blood nerve barrier disruption and edema are common in neuropathic pain as well as in complex regional pain syndrome (CRPS). MicroRNAs (miRNA) are epigenetic multitarget switches controlling neuronal and non-neuronal cells in pain. The miR-183 complex attenuates hyperexcitability in nociceptors, but additional non-neuronal effects via transcription factors could contribute as well. This study explored exosomal miR-183 in CRPS and murine neuropathy, its effect on the microvascular barrier via transcription factor FoxO1 and tight junction protein claudin-5, and its antihyperalgesic potential. Sciatic miR-183 decreased after CCI. Substitution with perineural miR-183 mimic attenuated mechanical hypersensitivity and restored blood nerve barrier function. In vitro, serum from CCI mice und CRPS patients weakened the microvascular barrier of murine cerebellar endothelial cells, increased active FoxO1 and reduced claudin-5, concomitant with a lack of exosomal miR-183 in CRPS patients. Cellular stress also compromised the microvascular barrier which was rescued either by miR-183 mimic via FoxO1 repression or by prior silencing of Foxo1.
Perspective
Low miR-183 leading to barrier impairment via FoxO1 and subsequent claudin-5 suppression is a new aspect in the pathophysiology of CRPS and neuropathic pain. This pathway might help untangle the wide symptomatic range of CRPS and nurture further research into miRNA mimics or FoxO1 inhibitors.
Although widely prevalent, neuropathic and chronic pain remains a challenge to treat due to its diverse underlying etiologies, varying clinical presentations and a still incomplete molecular understanding. Complex regional pain syndrome (CRPS) is a rare primarily chronic pain syndrome – occurring after fracture or surgery of an extremity. In addition to pain, common features of CRPS include allodynia, edema, sweating, temperature, skin color, and trophic changes as well as motor disturbances at the affected extremity.
While an entity of its own, CRPS exhibits features comparable to painful traumatic peripheral neuropathies including pain characteristics, gain in response to painful stimuli, and loss in response of nonpainful stimuli in quantitative sensory testing.
Substance P signaling contributes to the vascular and nociceptive abnormalities observed in a tibial fracture rat model of complex regional pain syndrome type I.
Chronic post-ischemia pain (CPIP): A novel animal model of complex regional pain syndrome-type I (CRPS-I; reflex sympathetic dystrophy) produced by prolonged hindpaw ischemia and reperfusion in the rat.
– are available for CRPS each covering certain aspects of the disease. The neuropathic and neurogenic phenotype of CRPS can be modeled by chronic constriction injury (CCI).
It protects the nerve from external, potentially noxious stimuli. Main components are the perineurium and the endoneurial endothelium with sealed tight junctions: transmembrane tight junction proteins like claudins, occludin, and tricellulin form complexes with cytosolic anchor proteins like Tjp1 (ZO-1).
claudin-5 expression is also reduced in traumatic neuropathy like CCI, as soon as 3 hour post-surgery, with a nadir at day 7, associated with a leaky BNB.
Early alterations of Hedgehog signaling pathway in vascular endothelial cells after peripheral nerve injury elicit blood-nerve barrier disruption, nerve inflammation, and neuropathic pain development.
Early alterations of Hedgehog signaling pathway in vascular endothelial cells after peripheral nerve injury elicit blood-nerve barrier disruption, nerve inflammation, and neuropathic pain development.
Endoneurial edema elevates local hydrostatic pressure, compromises the microcirculation, promotes ischemia, harms Schwann cells and axons, and facilitates neuropathic pain. One modifier of microvascular barriers are microRNAs (miRNA).
Hypoxia-Induced MicroRNA-212/132 alter blood-brain barrier integrity through inhibition of tight junction-associated proteins in human and mouse brain microvascular endothelial cells.
MiR-183 is part of the miR-182/-183/-96 cluster: A sensory-neuron specific conditional knockout mouse of this whole cluster develops mechanic hypersensitivity,
MicroRNA expression profile in bovine granulosa cells of preovulatory dominant and subordinate follicles during the late follicular phase of the estrous cycle.
Differential regulation of endogenous glucose-6-phosphatase and phosphoenolpyruvate carboxykinase gene expression by the forkhead transcription factor FKHR in H4IIE-hepatoma cells.
: Active, non-phosphorylated FoxO1 accumulates in the cell nucleus, while inactive phosphorylated FoxO1 relocates to the cytoplasm. FoxO1 also regulates claudin-5: Active FoxO1 – together with β-catenin and transcription factor TCF – binds to the claudin-5 promotor region and inhibits transcription.
Expression of the ALS-causing variant hSOD1(G93A) leads to an impaired integrity and altered regulation of claudin-5 expression in an in vitro blood-spinal cord barrier model.
While both miR-183 – FoxO1 binding and FoxO1 – claudin-5 interaction are established, the complete pathway has never been explored. Moreover, research is sparse on miR-183 in barriers. In this study, we assessed antihyperalgesic and barrier-protective effects of miR-183 in vivo and in vitro, and tested barrier breakdown with sera from CCI mice and CRPS patients. We hypothesized that miR-183 suppresses its target FoxO1, thereby allows for Cldn5 transcription, restores endoneurial barrier integrity, and ultimately relieves mechanical hyperalgesia (Fig 1).
Figure 1Graphical abstract: MiR-183 protects the barrier via FoxO1 suppression. A In neuropathic pain, miR-183 is decreased. This leads to overexpression of FoxO1 and consecutive inhibition of Cldn5 transcription. Ultimately, the BNB is impaired. B. Through perineural application of miR-183 mimic, Foxo1 is repressed. Claudin-5 expression is increased, causing BNB recovery and re-sealing. Created with biorender.
Male C57BL/6 mice aged 8 to 12 weeks (Janvier) were held in sawdust coated cages in groups of four, with environmental enrichment, at a 12 hours light-dark cycle with food and water ad libitum. Littermates were randomized: Two animals of each cage were allocated to the study group, their littermates to the control group. Throughout the duration of the experiments, animals’ welfare was assessed through daily score sheets. Sample sizes were calculated by a biostatistician, based on literature research and previous experience. All animal experiments were approved by the regional government and are in accordance with the guidelines of the International Association for the Study of Pain.
CCI Surgery
CCI was performed under deep isoflurane anesthesia.
The left dorsal proximal and medial thigh was shaved and disinfected using a propanol-based spray (Kodan, Schülke, Norderstedt). For CCI, the sciatic nerve at the left midthigh was exposed, and three loose silk ligatures (Perma Silk 6.0, Ethicon Inc., Somerville, NJ) with about 1-mm spacings in between were loosely tightened around the sciatic nerve.
The wound was closed with sutures (Prolene 5.0, Ethicon Inc., Somerville, NJ). Sham-operated mice served as control. Here, the sciatic nerve was briefly exposed without performing ligatures around the nerve. The wound was closed with sutures.
Perineural Injection of miR-183 mimic
Customized lyophilized in vivo mimics (mmu-miR-183, control cel-miR-67) were reconstituted in PBS at pH 7,4 and aliquoted at 20 pmol/µl at -20°C to avoid multiple freeze-thaw cycles. In vivo mimics were designed as follows (Dharmacon, Cambridge, UK): mmu-miR-183-5p: Active: 5′ 5′-P.U.A.U.G.G.C.A.C.U.G.G.U.A.G.A.A.U.U.C.A.C.U.3′-Fl 3′, Passenger: 5′ U.G.A.A.U.U.C.U.A.C.C.A.G.U.G.C.C.A.U.A.U.U.3′-Chl 3′. cel-miR-67: Active: 5′ 5′-P.U.C.A.C.A.A.C.C.U.C.C.U.A.G.A.A.A.G.A.G.U.A.G.A.3′-Fl 3′, Passenger: 5′ U.A.C.U.C.U.U.U.C.U.A.G.G.A.G.G.U.U.G.U.G.A.U.U.3′-Chl 3′.
Starting 7 days after CCI, perineural injections were performed daily for 4 consecutive mornings: Animals were put under deep isoflurane anesthesia. The left buttock and proximal dorsal thigh were disinfected using a propanol-based spray (Kodan, Schülke, Norderstedt). 50 µl (corresponding to 1 nmol) of thawed mimics were injected into the proximal left thigh using a 1 ml insulin syringe and a 26 G cannula. Twitching of the ipsilateral hind paw indicated proximity to the sciatic nerve. This injection technique without prior skin incision is advantageous for repetitive injections and well-established.
Macrophage inflammatory protein-1alpha mediates the development of neuropathic pain following peripheral nerve injury through interleukin-1beta up-regulation.
Behavioral testing in mice focused on mechanical sensitivity, as mechanoceptive pathologies are more robust in CRPS patients than thermoceptive pathologies.
Prior to the first injection and 6 hours after each mimic injection, the paw withdrawal threshold as indicator of the mechanical sensitivity was measured using von Frey filaments. Thresholds were determined using Dixon's Up-and-Down method. Before starting the experiment, animals were allowed to acclimatize to the setting. All behavioural experiments took place in the home cage room.
Harvesting of Sciatic Nerve
Following decapitation under deep isoflurane anesthesia, the thighs were disinfected, incised and the sciatic nerves harvested. Fresh tissue was snap-frozen in liquid nitrogen and stored at -80°C; histology samples were initially embedded in TissueTek (Sakura, Alphen aan den Rijn, Netherlands).
Permeability Assay
To assess capillary permeability, anesthetized mice were laid down in a supine position on a pad. The thoracic cavity was opened through sternotomy and the heart exposed. 50 µl of 70 kDa blue dextran solution (TdB, Uppsala, Sweden) was injected into the left ventricle using a 1 ml-insulin syringe. The dye-injected mice were sacrificed by decapitation after 2 minutes. The sciatic nerves were dissected and embedded into Tissue-Tek. Frozen samples were cut into 12 µm-thick sections on a cryostat at -20°C. Without any fixation, microscope glass slides containing tissue sections were mounted and imaged by fluorescence microscopy (Olympus. XB 51). Unchanged RGB-8-bit images were processed with Fiji/ImageJ (versions 1.51d-j and 1.52e, Open Source), as described before.
For semiquantification, endoneurium was manually defined in captured brightfield images, saved as region of interest and transferred to the immunofluorescence images. Immunofluorescence intensity was measured in images with same settings and magnification. Blue dextran intensity was determined with the integrated measure function in Fiji/ImageJ [determined intensity (IntD) = “selected area” ∗ “mean color intensity”]. To correct the intensity for noise (corrected immunofluorescence intensity, IntC), the background signal was registered in three areas and the given mean subtracted from the determined intensity [IntC = IntD – (“selected area” ∗ “mean of the color intensity of 3 background areas”)]. Referring to the initially selected area, a normalized immunofluorescence intensity (NIFI) was determined (NIFI = IntC / selected area). The mean NIFI from up to three fascicles of each animal was used to quantify and compare stainings. NIFI means of the two groups (CCI + mmu-miR-183 mimic, CCI + cel-miR-67 mimic) were compared.
Patients
Patients participated in the ncRNAPAIN project (EU FP 7, grant agreement 602133) following the ncRNApain study protocol (German registry for clinical studies (https://www.germanctr.de/, Registration Number DRKS00008964). They underwent extensive clinical and neurological examination and blood sampling. Clinical features are based on observed signs according to the CRPS Severity Score.
; controls comprised healthy subjects and subjects after trauma who did not develop CRPS. Pain symptoms in controls included prolonged healing pain after trauma and back pain. For serum incubation, three patients reflecting typical CRPS features, including a strong phenotype, and healthy controls matched for age and gender were selected from the cohort. All patients had given written informed consent; the study was approved by the local ethics committee.
Serum Preparation and Exosome Isolation
Blood samples (Sarstedt S-monovette with beads, 7.5 ml, Sarstedt, Nümbrecht, Germany) were taken from patients in the morning, after overnight fasting. They rested for 30 minutes (room temperature) before centrifugation at 1300 g for 10 minutes (room temperature). Aliquots were stored at -80°C. Total exosomes were extracted from 1 mL of patient plasma using miRCURY exosome isolation kit serum and plasma (Exiqon, Woburn, MA; catalog# 300101). The kit was used according to the manufacturer's instructions. After incubating the plasma with 10 µL thrombin for 5 minutes, an initial spin at 10,000 g was performed to remove cells and debris. The supernatant was removed, and the corresponding amount of precipitation buffer according to the manufacturer's instruction was added. The mixtures were incubated at 4°C for an hour and spun at 500 g for 5 minutes (room temperature) to precipitate the pellet. The buffer was completely removed, and the pellet was resuspended in 140 µL of manufacturer-supplied resuspension buffer. Electron microscopy confirmed the isolation of exosomes as described before.
Differential susceptibility of cerebral and cerebellar murine brain microvascular endothelial cells to loss of barrier properties in response to inflammatory stimuli.
and cultured in Dulbecco´s Modified Eagles Medium (DMEM) (Sigma-Aldrich, Burlington, MA) supplemented with 50 U/ml penicillin/streptomycin and 10% fetal calf serum in a 37°C incubator with 5% CO2/95% air atmosphere (Steri-Cult 200, Forma Scientific, ThermoFischer Scientific, Waltham, MA) until confluent. The medium was replenished thrice per week and the cells were split once a week in a 1:3 ratio by dissociation with 0.25% trypsin-EDTA (Sigma-Aldrich). Meanwhile, rat glioma astrocytes C6 (ATCC) were cultured in the same medium and culture conditions but sub-cultivated in a 1:20 ratio. Both cell lines were grown in 0.5% gelatin (Serva)-coated cell culture tissue flasks.
For C6/OGD treatment, cells were differentiated in DMEM medium supplemented with 1% fetal calf serum for 24 hours. Next, medium was changed into glucose and pyruvate-free DMEM prior to placing the cells in the oxygen/glucose deprivation (OGD) chamber (Heracell 150i, Thermo Fischer Scientific) for 4 hours with the following conditions: 37°C, 5% CO2and 1% O2.
To induce inflammation-based effects, medium of stretched C6 astrocytes was collected and centrifuged. Supernatant was used to replace the medium of cerebEND cells. Prior to collecting medium from the C6 cells, they were grown in BioFlex 6-well culture plates with collagen-coated silastic membranes (Flexcell International, Burlington, NC) and subjected to stretch injury using the Cell Injury Controller II system (Virginia Commonwealth University) with 50 ms pulse duration, as previously described.
Before serum incubation, sera from CCI mice as well as CRPS patients were diluted to an end concentration of 2% in cell culture medium, as described before.
Glucocorticoid effects on endothelial barrier function in the murine brain endothelial cell line cEND incubated with sera from patients with multiple sclerosis.
The sera were individually added to cerebEND cells and incubated for 24 hours before further analyses were conducted.
For incubation with miRNA mimics, hsa-miR-183-5p and negative control cel-miR-39 mimic (miRCURY LNA microRNA Mimic, Exiqon, Vedbaek, Denmark) were re-suspended in diethyl pyrocarbonate (DEPC)-water, to a final concentration of 5 pmol/μl. 1 μl of this was added to 4 μl of i-Fect siRNA transfection reagent (Neuromics, Minneapolis, MN) and 195 μl of cell culture medium. The mixture was incubated for 5 minutes at room temperature before adding dropwise to the cells. The cells were then allowed to incubate at 37°C for 24 hours. As the hsa-miR-183-5p mimic is also designed to target mmu-miR-183-5p, we will use the abbreviation “mmu-miR-183″ to avoid confusion. To transfect cerebEND cells with FoxO1 siRNA, FoxO1 Stealth siRNA (1299003, Thermo Fisher Scientific) was diluted in Opti-MEM I medium (Sigma Aldrich) to a final concentration of 6 pmol. After gentle mixing, 100 µl was placed onto 24-wells. Next, 1 µl Lipofectamine RNAiMAX (Thermo Fisher Scientific) was added into the wells and allowed to incubate for 20 minutes at room temperature. Afterwards, 500 µl of cerebEND cells with a density of 4.5 x 105 was added. The plate was mixed gently and allowed to incubate up to 72 hours at 37°C. Medium change was done every two days. The cells were expanded by seeding into 12 wells, then 6 wells, then into a T25 flask.
Transendothelial Electric Resistance (TEER) Measurements
Cells were grown on transwell-inserts (Corning, Corning, NY) with a pore diameter of 0.4 µm. Following treatment, TEER was measured atop a warm plate set to 37°C, using the volt-ohm meter device EVOM (World Precision Instruments, Sarasota, FL). Blank filters were used as internal control.
Immunohistochemistry of cerebEND Cells
After removing medium from transwell inserts and washing with PBS, cells were fixed in cold 500 µl methanol for 20 minutes at -20°C. After repeated washing and rehydration in PBS, cells were blocked with 400 µl PBS containing 5% porcine serum for 1 hours at room temperature. Next, 100 µl of the antibody (1:500 in PBS with 5% donkey serum) were added for 1 hours at 37°C. Following repetitive washing steps in PBS, cells were mounted on cover glasses using a drop of Vecta Shield mounting medium (Vector Laboratories, Burlingame, CA). Antibody used: Claudin-5 Alexa Fluor 488, Thermo Fisher Scientific (catalog no 352588).
qPCR
CerebEND cells were washed twice with PBS before lysing with QIAzol lysis reagent (Qiagen, Venlo, Netherlands). The cells were then scraped from the wells with a cell scraper. Next, the cells were further dissociated by passing through an insulin syringe (Omnican, U40, B. Braun, Melsungen, Germany). Nerve tissue was resuspended in QIAzol lysis agent and dissociated using the Tissue Lyser (Qiagen) for to 4 minutes at 20/s. RNA was isolated from the homogenates with the Qiagen miRNeasy kit (Qiagen), following manufacturer´s instructions.
RNA concentrations were measured with a Nanodrop ND 2000 spectrophotometer (Thermo Fisher Scientific) . For mRNA experiments, 1 μg total RNA per sample were reverse-transcribed to cDNA by means of the high-capacity cDNA kit (with random primer and RNAse inhibitor, Applied Biosystems, Waltham, MA) according to the manufacturer's instruction. qPCR analyses were performed using FAM-labeled probes for all investigated targets using TaqMan technique according to the manufacturer's protocol (Taqman, Applied Biosystems). Gapdh served as reference gene. For miRNA analysis, TaqMan technique was applied using the TaqMan MicroRNA Reverse Transcription Kit (input 1 ng total RNA per sample) and TaqMan Universal PCR Master Mix (Applied Biosystems) according to the manufacturer's protocol. For miRNA, snU6 served as reference gene. The following Taqman probes were used: Cldn5: mm_00727012_s1; Foxo1: mm_00490671_m1; Gapdh: mm_99999915_g1; miR-183-5p: 002269; U6 snRNA: 001973. Samples were analyzed as triplicates. Relative mRNA expression was analyzed using the ΔCt method.
Western Blot
Protein was isolated from supernatant derived from RNA extraction. The supernatant was pelleted by centrifugation and re-suspended in 0.3 M guanidine hydrochloride. Samples were again centrifuged and washed with ethanol. The ethanol wash was removed, and the pellet was allowed to dry for 5 to 10 minutes at 37 °C. The pellet was dissolved in 50 μl of protein buffer (8 M urea, 3M thiourea, 4% Chaps, 1% DTT, all from Sigma Aldrich). Samples were sonicated for 10 seconds at a pulse of 0.0005/0.0003 ps. Protein concentration was determined using the Pierce BCA Protein Assay kit (Thermo Fisher Scientific). 4x Laemmli buffer (8% SDS, 40% glycerol, 0.004% bromphenolic blue, 0.25 M Tris-HCl supplemented with 6% β-mercaptoethanol shortly before usage) was added to the samples. Afterwards, 20 μg of protein was loaded onto 12% SDS-PAGE gels (1.5 mm thick). Following electrophoresis, proteins were transferred onto polyvinylidene difluoride membranes (Biorad) using a tank blotter with the following conditions: 40 mA per gel at 4°C overnight. Blocking of membranes were done using 5% milk powder. Incubations with primary and secondary antibodies were carried out as previously described.
The following antibodies were used: FoxO1: ab179450 (abcam, Cambridge, UK) 1:100; FoxO1-P (S256): ab131339, (abcam, Cambridge, UK) 1:100; claudin-5: 35-2500 (Invitrogen, Carlsbad, CA) 1:500; β-actin: A-3854 (Sigma Aldrich) 1:20,000. Membranes were incubated in ECL solution for 2 minutes for visualization using the FluorChem FC2 Multiimager II (Alpha Innotech). Densitometric analyses of the protein bands were carried out using the ImageJ software (public domain, developed by NIH).
Statistics
For statistical analysis, GraphPad Prism (Version 9, GraphPad Software Inc., San Diego, CA) and Microsoft Excel 2016 (Version 16.11, Microsoft, Redmond, WA) was used. Tests performed included Student's t-test (incl. Welch's correction for unequal variance) for comparison between two groups, and one- or two-way ANOVA + post-hoc test for multiple groups (see figures). Statistical significance was determined as P < .05.
Results
MiR-183 is Downregulated in CCI
While miR-183 decrease in neuropathy is known in the DRG, this is not the case for the peripheral nerve. To assess CCI-induced changes of miR-183 over time, we measured sciatic miR-183 expression for 14 days after surgery. Levels decreased continuously, by 82 % at 7 days, when a full neuropathic phenotype had developed (Fig 2A). In parallel, we observed an 85 % reduction in Cldn5 after 7 days (Fig 2B).
Figure 2CCI reduces miR-183 and Cldn5 in mice. A MiR-183 expression in the sciatic nerve 6 h, 1, 3, 7 and 14 d after CCI. Expression values were normalized on naïve animals (n = 4, ddCt, one-way ANOVA and Bonferroni multiple comparison test). B Cldn5 levels in the sciatic nerve 7 d after CCI (n = 8, dCt, t-test). All data are shown as mean ± SEM, *P < .05, **P < .01, ***P < .001.
Local miR-183 Injection Attenuates Mechanical Hypersensitivity
To study the effect of miR-183 on pain behavior and barrier function, we injected mmu-miR-183 mimic perineurally at the sciatic nerve, once a day, starting 7 days after CCI surgery (Fig 3A). Mechanical allodynia rapidly subsided (Fig 3B): After two injections, paw withdrawal thresholds rose progressively as compared to control cel-miR-39 injections. After 4 consecutive injections, miR-183 had accumulated in the nerve (Fig 3C). In parallel, miR-183 mimic restored the BNB: Clnd5 expression in the sciatic nerve increased (Fig 3D) and extravasation of 70 kDa blue dextran into the endoneurium vanished (Fig 3E). This suggests a reconstituted endoneurial barrier function. At the same time, expression of active FoxO1, a miR-183 target and claudin-5 transcriptional repressor, decreased by 50% (Fig 3F) – whereas phosphorylated inactive FoxO1 changed only insignificantly (Fig 3G). Thus, decreased active FoxO1 links miR-183 treatment to barrier resealing.
Figure 3MiR-183 mimic in vivo injection alleviates pain, restores the barrier, and reduces FoxO1. A Experimental timetable: Starting 7 d after CCI, mimics were injected daily, followed by measurement of mechanical sensitivity 6 h after each injection. Control group: cel-miR-67. B Mechanical sensitivity after perineural mimic injection, measured by von Frey filaments (n = 8, two-way ANOVA with Sidak‘s multiple comparison test). Contralateral values shown as reference. C Sciatic miR-183 expression after perineurial mimic application (n = 8, t-test). D Claudin-5 protein expression after mimic injection (n = 5, t-test). E Endoneurial permeability measured as normalized immunofluorescence intensity of endoneurial 70kDa blue dextran after intracardial injection (n = 4, t-test). F Non-phosphorylated FoxO1 protein expression after mimic injection (n = 6, t-test). G Phosphorylated FoxO1 protein expression after mimic injection (n = 6, t-test). All data are shown as mean ± SEM, *P < .05, **P < .01. Abbreviations: ns, not significant; IDV, integrated density value.
Its barrier function is dependent on claudin-5. Sera from CCI mice taken at day 7 transiently impaired the transendothelial resistance and therefore barrier function after an incubation time of 24 hours (Fig 4A). In parallel, active FoxO1 expression doubled (Fig 4B), while phosphorylated FoxO1 levels remained unchanged compared to sera from sham controls (Fig 4C).
Figure 4CCI serum mice impairs microvascular barrier function and increases FoxO1. A Change in transendothelial resistance (TEER) after incubation of microvascular cerebEND cells with 2% mouse serum (CCI vs sham-operated animals) for 24 h, normalized to pre-incubation values. Cell medium: Incubation with cell culture medium only. (n = 3, one-way ANOVA with Newman-Keuls multiple comparison test). B Non-phosphorylated FoxO1 protein expression after serum incubation (n = 4, t-test). C Phosphorylated FoxO1 protein expression after serum incubation (n = 4, t-test). All data are shown as mean ± SEM, *P < .05. Abbreviations: ns, not significant; IDV, integrated density value.
Sera of CRPS Patients Lacking Exosomal miR-183 Impair the Microvascular Barrier in vitro
We next quantified exosomal miR-183 in plasma from CRPS patients and matched controls without CRPS from the ncRNAPain Study cohort (Table 1): Exosomal miR-183 was markedly decreased in CRPS patients (Fig 5A). To further explore the effect of CRPS serum, we selected three patients with typical CRPS features from the cohort: These patients were middle-aged females suffering from acute (<1 year), warm CRPS type I of the upper extremity and from moderate to high pain with neuropathic features. We compared these to healthy controls, matched for age and gender (Tab 2). Also in this small cohort, we observed a trend towards reduced exosomal miR-183, although the effect was not significant (P = .08) (Fig 5B). Further analysis of CRPS patients showed a moderate negative correlation of miR-183 levels with age (r = -0.45). Body mass index, mean pain, and CSS were not correlated (Tab 3). Moreover, we did not observe any impact of sex, CRPS type (I vs II, warm vs cold), or observed symptoms (Suppl Tab). Incubation of microvascular cells with CRPS serum had a similar effect on the microvascular barrier as CCI serum: It compromised the barrier integrity transiently after 24 hours (Fig 5C) and led to an increase of active FOXO1; phosphorylated, inactive FOXO1 remained stable (Fig 5D, E).
Table 1Epidemiological and Clinical Characteristics of the CRPS Cohort.
Data from entire cohort compared to controls (n = 30–40, t-test). Presumed etiology of pain in control subjects: prolonged healing of fractures or nonspecific back pain.
Figure 5Circulating exosomal miR-183 is reduced in CRPS patients, and incubation with serum impairs the microvascular barrier through upregulation of non-phosphorylated FoxO1. A, B Relative exosomal miR-183 expression of the entire cohort (A, t-test with Welch‘s correction, n = 30–40) and in subjects of the serum study (B, t-test with Welch‘s correction, n = 3). C Change in transendothelial resistance (TEER) after incubation with 2 % patient serum (CRPS vs Controls), normalized on pre-incubation values. Cell Medium: Incubation with cell culture medium only (n = 3, one-way ANOVA with Holm-Sidak‘s multiple comparison test). D Non-phosphorylated FoxO1 protein expression after incubation (n = 3, t-test). E Phosphorylated FoxO1 protein expression after incubation (n = 3, t-test). All data are shown as mean ± SEM, *P < .05, **P < .01. Abbreviations: ns, not significant; IDV, integrated density value.
MiR-183 Restores the Microvascular Barrier in vitro via FoxO1 Inhibition
In a next step, we investigated the putative pathway of barrier resealing by using the C6/OGD protocol on microvascular cells. This established model of cellular stress - due to hypoxia and an inflammatory environment - impairs barrier function.
Indeed, C6/OGD strongly reduced TEER (Fig 6A) and Cldn5 (Fig 6B) in these cells. Incubation with miR-183 mimic restored the barrier: TEER values nearly doubled, consistent with an improvement by 36 % compared to the control mimic (Fig 6C). Immunohistochemistry shows a recovered claudin-5 intensity and distribution pattern (Fig 6D). In line, Cldn5 was significantly increased (Fig 6E), while FoxO1 mRNA and protein expression declined (Fig 6F, G). This confirms miR-183 mimic treatment as a rescue to stressed microvascular cells. Finally, we analyzed the direct influence of miR-183 target FoxO1 to confirm the first part of our model (Fig 1): We silenced FoxO1 by transfecting cells with Foxo1 siRNA (transfection efficacy 50%, Fig 7A) before C6/OGD treatment. Such prior inhibition of FoxO1 attenuated the barrier breakdown (Fig 7B). Thus, FoxO1 regulation and barrier recovery after miR-183 treatment are not merely correlational but causal: FoxO1 expression directly affects the barrier.
Figure 6MiR-183 reseals the microvascular barrier and represses FoxO1. A, B. Microvascular cells treated with C6/OGD, compared to normoxia: Transendothelial resistance TEER (A, n = 12, t-test) and Cldn5 (B, n = 5, t-test with Welch‘s correction). C-G: C6/OGD treatment and consecutive mimic incubation (control: cel-miR-39): Transendothelial resistance (C, n = 6, two-way RM ANOVA with Bonferroni‘s multiple comparison test), Cldn5 immunoreactivity – representative example (D), relative Cldn5 RNA levels (E, n = 3, t-test), FoxO1 protein expression (F, n = 6, t-test) and Foxo1 (G, n = 6, t-test). All data are shown as mean ± SEM, *P < .05, **P < .01, ***P < .001, ****P < .0001. Abbreviations: Bar, 100 µm; ns, not significant; IDV, integrated density value.
Figure 7Foxo1 silencing protects the microvascular barrier. A Transfection efficacy of Foxo1 siRNA measured as Foxo1 (n = 6, t-test). Control: only vehicle, without siRNA. B Transendothelial resistance (TEER) after Foxo1 siRNA transfection (n = 6, t-test). All data are shown as mean ± SEM, *P < .05, **P < .01.
In this study, we showed not only an antihyperalgesic but also barrier-protective function of local miR-183 in peripheral neuropathy. MiR-183 inhibits FoxO1 and thus allows for claudin-5 recovery in endoneurial cells, functional barrier resealing and, ultimately, antinociception (Fig 1). This effect is due to quantitative changes of FoxO1 expression and not a shift in FoxO1 activity due to phosphorylation – a direct effect of miR-183 targeting Foxo1 posttranscriptionally. Sera of both CCI mice and CRPS patients impaired the microvascular barrier. Thus, both diseases entail a common circulating factor that initiates breakdown. This observation not only closes a gap between animal models and human disease, but also strengthens the idea of barrier impairment as a common underlying principle. This is endorsed by miR-183 deficiency in both CCI and CRPS.
It is still unclear what causes such a miR-183 deficiency. In general, miRNAs can be regulated both on a transcriptional level (eg, via transcription factors or promoter methylation) and post-transcriptionally (eg, through SNPs).
Interestingly, one upstream regulator of miR-183, in natural killer cells, is transforming growth factor beta (TGFβ). This cytokine is decreased in blood of CRPS patients.
Thus, such reduction in a promoter might explain the low miR-183 levels we observed. Yet, this relation has not been described in pain research and requires further research.
MiR-183 expressed in sensory neurons is downregulated in neuropathy or osteoarthritis.
But also non-neuronal pathways have been implicated to explain the antihyperalgesic effect of miR-183: MAP kinase MAP3K4 is a direct target, downregulating proinflammatory cytokines IL-6, IL-1ß, and cyclooxygenase 2 in the DRG.
In our study, we extended miR-183 effects to non-neuronal barrier-forming cells at the site of the injury itself – documented by in vivo and in vitro evidence.
We showed that miR-183 mimic incubation re-establishes the electric resistance of a previously leaky barrier and reduces its known target FoxO1. The miR-183-FoxO1 axis regulates cell cycles: FoxO1 is pro-apoptotic, and inhibition by miR-183 leads to cancer progression, eg, in mesothelioma.
A long non-coding RNA SNHG15 acts as a “sponge”/competing endogenous RNA (ceRNA) of miR-183; via FoxO1, it leads to higher apoptosis and infarct size in an OGD + reperfusion model of cerebral ischemia.
Similar to our results, the authors observed a miR-183-dependent regulation of FoxO1 after OGD. As in our study, FoxO1 siRNA-transfected cells maintained a more stable microvascular barrier after OGD treatment. This supports the role of FoxO1 targeting claudin-5, the main constituent of the endothelial barrier: FoxO1 also regulates claudin-5 in an in vitro model of the blood-spinal cord barrier.
Expression of the ALS-causing variant hSOD1(G93A) leads to an impaired integrity and altered regulation of claudin-5 expression in an in vitro blood-spinal cord barrier model.
Another pathway describes FoxO1 activation with consequent downregulation of Cldn5 in the BBB via IL-1ß, dependent on non-muscle myosin light chain kinase.
Non-muscle Mlck is required for beta-catenin- and FoxO1-dependent downregulation of Cldn5 in IL-1beta-mediated barrier dysfunction in brain endothelial cells.
In conclusion, our results support a complete pathway of miR-183-Foxo1-Cldn5 in neuropathic pain, the single components of which are described for other barriers.
We found a decrease in circulating, exosomal miR-183 in CRPS patients. Exosomes are produced and secreted from a variety of cell types and facilitate intercellular communication –either in the immediate environment or over larger distances, eg, via blood stream. MiR-183-enriched exosomes play a role in cancer: deriving from tumor cells, they regulate FoxO1 and cytokine expression.
Exosomal miRNA can also regulate pain states. In neuropathic pain, sensory neurons secrete miR-21-loaded exosomes that are taken up by macrophages and induce the pro-inflammatory M1 phenotype.
Increased endothelial permeability is described after uptake of exosomal miR-155: Induced in exosomes from vascular smooth muscle cells, miR-155 inhibits tight junction protein expression in recipient endothelial cells.
Our findings combine these observations that a) miR-183 is a relevant exosomal miRNA and b) exosomal miRNA signature regulates chronic and neuropathic pain.
In this study, we demonstrate the antihyperalgesic effect of local, perineural miR-183 application. This is particularly interesting given the pleiotropic effect of miR-183: As with many miRNA, possible off-site targets could be a concern.
Therefore, a local application seems advisable, especially in regional pain syndromes. While no miR-183 mimic has been approved yet for therapeutic purposes, the field of miRNA therapeutics is advancing fast.
In addition, FoxO1 might be another treatment target: FoxO1 inhibitors were developed as treatment in obesity and diabetes. AS1842856 showed promising effects
Also here, local administration might be an interesting option in regional pain syndromes.
Interestingly, two studies found an opposite role of miR-183 involved in barrier opening: Firstly, in Dengue virus-infected HUVEC cells, miR-183 is increased; miR-183 overexpression suppressed Cldn5.
ERG-Associated lncRNA (ERGAL) promotes the stability and integrity of vascular endothelial barrier during dengue viral infection via interaction with miR-183-5p.
Secondly, our group has earlier observed an increase of sciatic miR-183 after perineurial barrier opening via local recombinant tissue plasminogen application, and BNB opening after local miR-183 application at the intact nerve.
These studies differ in several aspects. (i) The barriers targeted are different from ours: HUVECs are macrovascular cells and perineurial cells have an epithelial origin. Thus, regulations might be different in different barriers. (ii) The specific inflammatory microenvironment might be important: MiR-183 is upregulated in several viral infections
Thus, the observed Cldn5 downregulation might be a by-product of this context rather than a direct effect of miR-183 on endothelial cells. Furthermore, miR-183 might have a different function in an intact, non-injured environment. In summary, our data support the idea of a microvascular-specific barrier protection via miR-183 in nerve injury.
There are limitations to this study. While perineural injection of miRNA mimics proved highly efficient, we did not systemically study possible side effects. Mice behavior was unaltered except for the analgesic effect, and we did not observe increased local inflammation or necrosis, macroscopically or microscopically. Furthermore, CCI and CRPS present with several differences: The CCI model represents a posttraumatic peripheral mononeuropathy that develops reliably after a standardized procedure and resolves after 2 to 3 months, whereas CRPS is a complex painful condition with or without previous nerve damage that occurs in about 1% of patients with upper extremity trauma
Yet, in both conditions, serum incubation led to impairment of the microvascular barrier and FoxO1 increase. Thus, they do share common circulatory features as well as clinical symptoms like allodynia, mechanical hypersensitivity and neurogenic inflammation. While we show clear evidence of a barrier-protective function of miR-183, we do not suggest it to be the only pathway that explains the analgesic effect. Indeed, we argue that miR-183 has a pleiotrophic effect, setting off different pathways and signaling cascades that concertedly attenuate the hypersensitivity.
This study highlights the importance of barrier impairment and its control by serum factors in neuropathy. Humoral factors may contribute to the etiology of CRPS
Both studies were able to transfer these primary chronic pain diseases into animal models using patients’ sera. While the authors – as well as other groups before
– focused on autoantibodies, other factors like miRNAs are also plausible: Several circulating miRNA are regulated in CRPS and might serve as potential biomarkers.
This is the first study to provide clear mechanistic insight in vitro and in vivo into the impact of a humoral miRNA in CPRS.
Further research should study the mechanisms of miR-183 suppression in neuropathy to provide further insight into upstream pathways, eg, TGFβ as an upstream regulator. Moreover, the origin of exosomal miR-183 in pain is still unclear. Further characterization of the cellular origin and uptake might elucidate the background of exosomal miR-183 in pain.
The association of neuropathic pain with barrier alterations makes FoxO1 an interesting candidate to investigate in other neuropathies, such as diabetic polyneuropathy (dPNP): As mentioned above, FoxO1 is a crucial regulator in gluconeogenesis.
Differential regulation of endogenous glucose-6-phosphatase and phosphoenolpyruvate carboxykinase gene expression by the forkhead transcription factor FKHR in H4IIE-hepatoma cells.
Hence, an investigation of the miR-183-FoxO1-claudin-5 axis in dPNP is promising.
Conclusion
This is the first description of miR-183, FoxO1 and claudin-5 as a functional pathway. Moreover, we have identified miR-183 as a stabilizer of microvascular cells of the BNB and as an important exosomal serum factor possibly protecting against CPRS development after trauma. Together with the evidence that local application of miR-183 has an antihyperalgesic effect, miR-183 is a putative novel target for treatment of peripheral neuropathies or CPRS.
Acknowledgments
The authors thank Andrea Prappacher for mouse husbandry and Bernhard Schwab and Solange R. Sauer for technical support.
Differential regulation of endogenous glucose-6-phosphatase and phosphoenolpyruvate carboxykinase gene expression by the forkhead transcription factor FKHR in H4IIE-hepatoma cells.
Non-muscle Mlck is required for beta-catenin- and FoxO1-dependent downregulation of Cldn5 in IL-1beta-mediated barrier dysfunction in brain endothelial cells.
Glucocorticoid effects on endothelial barrier function in the murine brain endothelial cell line cEND incubated with sera from patients with multiple sclerosis.
Hypoxia-Induced MicroRNA-212/132 alter blood-brain barrier integrity through inhibition of tight junction-associated proteins in human and mouse brain microvascular endothelial cells.
Chronic post-ischemia pain (CPIP): A novel animal model of complex regional pain syndrome-type I (CRPS-I; reflex sympathetic dystrophy) produced by prolonged hindpaw ischemia and reperfusion in the rat.
MicroRNA expression profile in bovine granulosa cells of preovulatory dominant and subordinate follicles during the late follicular phase of the estrous cycle.
Substance P signaling contributes to the vascular and nociceptive abnormalities observed in a tibial fracture rat model of complex regional pain syndrome type I.
Macrophage inflammatory protein-1alpha mediates the development of neuropathic pain following peripheral nerve injury through interleukin-1beta up-regulation.
Expression of the ALS-causing variant hSOD1(G93A) leads to an impaired integrity and altered regulation of claudin-5 expression in an in vitro blood-spinal cord barrier model.
Early alterations of Hedgehog signaling pathway in vascular endothelial cells after peripheral nerve injury elicit blood-nerve barrier disruption, nerve inflammation, and neuropathic pain development.
Differential susceptibility of cerebral and cerebellar murine brain microvascular endothelial cells to loss of barrier properties in response to inflammatory stimuli.
ERG-Associated lncRNA (ERGAL) promotes the stability and integrity of vascular endothelial barrier during dengue viral infection via interaction with miR-183-5p.
This work was supported by the IZKF Würzburg (Z-2/61-P), the European Union FP7 grant “ncRNAPain” (grant agreement 602133) and the German Research Foundation (Clinical Research Group KFO5001).
P < .001.
00389-8&id=gr1.jpg){kind=link}
00389-8&id=gr2.jpg){kind=link}
00389-8&id=gr3.jpg){kind=link}
00389-8&id=gr4.jpg){kind=link}
00389-8&id=gr5.jpg){kind=link}
00389-8&id=gr6.jpg){kind=link}
00389-8&id=gr7.jpg){kind=link}