IMT504 blocks allodynia in rats with spared nerve injury by promoting the migration of mesenchymal stem cells and by favoring an anti-inflammatory milieu at the injured nerve

1. Introduction

Despite our growing understanding of the mechanisms of neuropathic pain, its management remains remarkably difficult due to lack of effective and/or safe treatments.^10,63 An underlying problem relates to difficulties in therapeutically addressing a multitude of mechanisms involved in the pathophysiology of neuropathic pain; a challenge that seems difficult to overcome by using mono-target drugs.⁹ Approaches targeting multiple systems, such as the regulation of neuroimmune interactions, are emerging as alternatives.⁴⁶

Exogenous administration of mesenchymal stem cells (MSCs)¹ is also being explored as an alternative therapeutic strategy for neuropathic pain.^6,27,64 We previously demonstrated that intraganglionic or systemic administration of MSCs in rats with single ligature nerve constriction or sciatic nerve crush prevents or resolves mechanical and cold allodynia.^11,41 Since then, the anti-hyperalgesic and anti-allodynic effects of systemic^23,53 or intrathecal⁶ administration of MSCs in rats with chronic orofacial pain²³ and mice with sciatic nerve ligation or spared nerve injury (SNI)^6,53 have been confirmed. The therapeutic potential of MSCs is currently addressed in several clinical trials, and a variety of biobanks have emerged in preparation for improvements in the protocols that regulate their use as medicinal products.¹ However, while their push into clinical practice remains strong,⁶⁰ MSCs represent a small fraction of cells derived from the bone marrow (or, in fact, from any other source),¹ making their availability a major limitation. An additional major pitfall is the need of lengthy and strict aseptic processes involving sample acquisition, culturing, expansion, and maintenance to obtain enough cells to be administered.^1,59 Therefore, strategies promoting endogenous MSCs mobilization seem more desirable from a therapeutic stand-point.³⁹

Two classes of synthetic immunomodulatory oligodeoxynucleotides (ODNs) have been identified according to their CpG content: CpG ODNs and non-CpG ODNs.^12,28 IMT504, belonging to the latter class, is characterized by presence of at least one active site bearing the sequence PyNTTTTGT, in which Py is a C or T, and N is A, T, C or G (Py: Pyrimidine; A: Adenine; C: Cytosin; T: Thymidine; G: Guanine).¹² We previously showed that IMT504 target B-cells, plasmacytoid dendritic cells, and MSCs.^13,25,45 We also demonstrated that multiple subcutaneous doses of IMT504 result in accelerated recovery from unilateral sciatic nerve crush-induced mechanical and cold allodynia in rats, and that the effect compares to a single intravenous administration of MSCs.¹¹ Interestingly, in vitro and in vivo analyses in naïve rats revealed that IMT504 induces expansion of MSCs in bone marrow and peripheral blood.²⁵

Here we test the hypothesis that the pro-mobilization effects of IMT504 on endogenous MSCs result in their homing within injured nerves, and that this causes microenvironmental changes that may explain the pain-like behavior reducing effects of the ODN. We do so by addressing, in vivo, the effects of a single subcutaneous dose of IMT504 in rats undergoing SNI on pain-like behavior, migration and homing of MSCs in injured nerves, and changes in pro- and anti-inflammatory cytokines at the injured site. We also analyze in vitro, potential IMT504-dependent changes in migration behavior and cytokine releasing properties of MSCs, and on the inflammatory milieu present in injured nerves.

2. Methods

2.1. Animals

A total of 171 adult Sprague–Dawley male rats (200–300 g; obtained from BioFucal, Buenos Aires, Argentina) were used for all experiments in this study. Rats were housed in a light- and temperature-controlled room with a 12-hour-light/dark cycle and ad libitum access to food and water. Animals were allowed 1 week for housing habituation before starting any experiment. All experiments performed were approved by the Institutional Animal Care and Use Committee (IACUC; 21–04) of the Instituto de Investigaciones en Medicina Traslacional, and were carried out according to the Guide for the Care and Use of Laboratory Animals (NIH Publication 86–23).

2.2. Spared nerve injury (SNI)

In 122 rats anaesthetized using Isoflurane (5% induction, 2.5% maintenance, 0.8 l/min O₂ flow rate; Piramal Healthcare, UK) the right hindlimb was shaved and the surgical site cleaned using 70% alcohol and iodine solutions. After incision of the skin and the biceps femoris muscle at the mid-thigh level, the sciatic nerve and its three terminal branches were exposed (sural, common peroneal and tibial). The tibial and common peroneal branches were tightly ligated using a non-absorbable 4.0 silk suture (Ethicon, NJ, USA), followed by removal of 2–4 mm of the distal nerve stump, being very careful to leave intact the sural branch. Muscle and skin were sutured in layers using vicryl 4.0 (Ethicon) and mononylon ethilon 4.0 (Ethicon) sutures, respectively. Postsurgical pain control was achieved by a single subcutaneous dose of dexketoprofen trometamol (5 mg/kg; Lab Argentina, Buenos Aires, Argentina) administrated before surgery and additional topical application of a 2% lidocaine hydrochloride gel (AstraZeneca, Buenos Aires, Argentina) on the sutured area.

2.3. Experimental drug

In all experiments, the oligodeoxynucleotide (ODN) IMT504, with sequence 5′-TCATCATTTTGTCATTTTGTCATT-3′ was used. The HPLC-grade phosphorothioate ODN (Biosynthesis Inc., Lewisville, Texas, USA) was suspended in sterile saline solution (0.9% NaCl; 20 mg/ml; storage concentration), and assayed for LPS contamination. Working concentration doses (6 mg/kg) were obtained by dissolving in sterile saline solution (vehicle; VEH). The ODN was administered subcutaneously, at a final volume of 200–300μl, depending on animal weight. Some experimental groups were injected using a 24-mer PolyC ODN (6 mg/kg) with no immune stimulating capacity, as an additional control.³

2.4. Experimental design

Rats (n= 40) received a single subcutaneous dose of IMT504, 7 days after injury (SNI+IMT504). Control groups included injured rats receiving a single subcutaneous injection of vehicle (SNI+VEH) (n= 40) or PolyC ODN (SNI+PolyC) (n= 6). Finally, naïve (uninjured and untreated; n= 49) and untreated 7 day-SNI rats (n= 36) were also included in the study.

A single injection, as opposed to multiple injections used in some of our previous studies,^11,31 was chosen based on our most recent publication where we identified 6 mg/kg as the optimal single effective dose in rats that had the most acceptable dose-response effect.³² A systemic delivery for IMT504 was chosen based on our previously published work, where we established the anti-allodynic efficacy of the ODN upon systemic administration.^11,31,32 Both choices were also based on our interest to provide translational value to the approach, as lower number of drug administration are better tolerated by patients.

2.5. Behavioral testing

Behavioral studies were performed in a quiet room during daytime (9.00–18.00). Animals were placed in transparent plastic domes (20×7×10 cm) on a metal mesh floor with 3×3 mm holes. Thirty minutes after habituation, the lateral area of both, ipsilateral and contralateral hindpaws was tested for sensitivity to non-noxious mechanical and cold stimuli (Fig. 1A). In all treatment protocols, animals were tested for pain-like behavior 1 day before SNI (basal responses) and at different time-points during the following weeks after injury. For experiments specifically determining the anti-allodynic effects of IMT504, an n= 6 per group was used. In addition, mechanical allodynia was also tested in all other animals before euthanasia and collection of tissues/cells to be used for other in vivo and in vitro experiments. This was done to confirm that any cellular-molecular observation adequately matched behavioral data at the time of analysis.

Fig. 1. Experimental design to study the anti-allodynic effects and the mechanisms of action of IMT504 in rats with spared nerve injury.

(A) Characterization of the time course of mechanical and cold allodynia and acute thermal nociception effects in naïve, SNI+PolyC and SNI+IMT504-treated rats (n= 6 rats per group, per time-point). (B) Quantification of MSCs (CD45-CD90+CD29+ cells) in sciatic nerve, peripheral blood and bone marrow were done by flow cytometry (n= 5 rats per group, per time-point). (C) BMMCs were collected from femur and tibia of naïve, SNI+VEH and SNI+IMT504 rats and cultured during 14 days to quantify the number of CFU-Fs (n= 6 rats per group, per time-point). (D) Rat and human MSCs in vitro, pre-treated or not with IMT504 (7μg/ml), were co-cultured with sciatic nerve conditioned medium (from naïve rats or ipsilateral and contralateral from 7-day SNI rats) for the assessment of migration behavior (n= 4). (E) For analysis of IMT504 internalization, MSCs were cultured with IMT504-Alexa Fluor 488 or Texas Red. (F) ELISA analysis was performed to address changes in pro- (TNF-α and IL-1β) and anti-inflammatory (IL-10 and TGF-β1) cytokines in injured nerves (n= 3 rats per group, per time-point). (G) Analysis of cxcr4 and cxcl12 expression levels in ipsilateral sciatic nerve from naïve, SNI+VEH and SNI+IMT504 rats and analysis of pro- and anti-inflammatory cytokines expression levels in MSCs and ex vivo 7 day-SNI nerves obtained from co-cultures, by RT-qPCR. Acronyms: SNI, spared nerve injury; VEH, vehicle; BMMCs, bone marrow mononuclear cells; PBMCs, peripheral blood mononuclear cells; Ipsi, ipsilateral; CFU-F, colony forming units-fibroblast; MSC, mesenchymal stem cell; CM, sciatic nerve conditioned medium; ELISA, enzyme linked immunosorbent assay; RT-qPCR, quantitative real time reverse-transcription-PCR. Figure was created using BioRender.com.

2.5.1. Mechanical allodynia assessment

Mechanical withdrawal thresholds were assessed by using von Frey nylon monofilaments of different bending forces (1.4, 2, 6, 8, 10, 15 and 26 g; Stoelting Inc., Wooddale, USA) and the modified up-down method of Dixon, as described by Chaplan et al.,⁵ to establish the 50% withdrawal threshold. Results were obtained using the formula: 50% g threshold = (10 Xf+kδ)/10.000, where Xf = value (in log units) of the final von Frey hair used; k = tabular value for the pattern of positive/negative responses; and δ = mean difference (in log units) between stimuli.

2.5.2. Cold allodynia assessment

Cold withdrawal frequency was evaluated following the Choi method and quantified by foot withdrawal frequency and allodynia score.^8,15 Briefly, a drop of acetone (50 μL) was applied to the lateral plantar surface of the hindpaw using the tip of a 1 mL syringe, 5 times to each paw at an interval of 5 min. Cold withdrawal frequency was expressed as the percentage of positive responses (foot withdrawal scored as positive (20%), and lack for withdrawal as negative (0%); 5 positive responses = 100%). For cold allodynia scoring, responses to acetone were graded using a 4-point scale: (0), no response; (1), quick withdrawal, flick or stamp of the paw; (2), prolonged withdrawal or repeated flicking (≥ 2) of the paw; (3), repeated flicking of the paw with licking directed at the ventral side of the paw. Cumulative scores were generated, with the minimum score being 0 (no response to any of the 5 iterations) and the maximum score being 15 (repeated flicking and licking of paws on each of the 5 iterations).

2.5.3. Acute thermal nociception

Acute thermal nociception was assessed 1 day before injury (baseline) and 3 and 21 days after treatment in naïve, SNI+VEH, SNI+PolyC and SNI+IMT504 rats as previously described.⁶⁵ Briefly, after two days of habituation to handling (by gentle restraining inside a clean cloth), the latency of tail-flick reflex to swift immersion of the last 3 cm of the tip of the tail on a water bath kept at a constant temperature of 52 ± 0.58 °C was measured using a stopwatch (resolution of 0.01 s) (Fig. 1A). This was repeated 3 times, with 10 s intervals, and an average response obtained.

2.6. Flow cytometry

Peripheral blood, bone marrow and sciatic nerves were harvested from naïve, SNI+VEH and SNI+IMT504 rats to obtain a single cell suspension (Fig. 1B). Briefly, rats were deeply anesthetized 3 and 21 days after treatment (n= 5 per group), using Isoflurane (as above). For analysis of peripheral blood, 5 ml samples were quickly obtained by cardiac puncture, collected in a tube containing heparin and processed through a gradient density (Histopaque-1077; Sigma, St. Louis, USA) for peripheral blood mononuclear cells (PBMCs) isolation, followed by suspension in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Grand Island, NY, USA). For analysis of bone marrow (BM), whole BM plugs were flushed out from femoral and tibial bones using a 10 ml syringe with DMEM supplemented with penicillin (100 U/ml; Sigma) and streptomycin sulfate (100 μg/ml; Sigma). Bone marrow mononuclear cells (BMMCs) were isolated by gradient density (Histopaque-1077) and suspended in fresh medium. Finally, for sciatic nerve analysis, a cardiac perfusion was first performed using 1x HBSS (Hanks’ Balanced Salt solution, Sigma, St. Louis) to completely remove blood from tissues. This was followed by careful removal of the suture in the ipsilateral nerves and the excision of a 1-cm–long segment proximal to the injury (or right-side nerve in the case of N rats). These samples were disaggregated using a Papain Solution (Papain (15 U/ml) /DNase (10 μg/ml)), (Papain, Roche; DNase, Sigma) at 37°C for 30 min, followed by filtration using a cell strainer cap and centrifugation at 1200 rpm for 10 s. Contralateral nerve samples were also harvested 21 days after treatment.

For all samples, cells were suspended in 1 ml of FACS buffer (1 % BSA and 0.05 % Sodium Azide (NaN₃) in 1x PBS (phosphate-buffered saline). Cells were harvested and stained with monoclonal antibodies (mAb) for 30 min in FACS buffer. The following mAb were used for phenotypic characterization of MSCs: Alexa Fluor 488-labeled anti-CD45 (clone OX-1; BioLegend, San Diego, CA, USA), PE-labeled anti-CD90 (clone KW322; BioLegend) and Alexa Fluor 647-labeled anti-CD29 (clone HMβ1–1; BioLegend). As a negative control, cells were incubated with the same species isotype controls as the primary antibodies (Alexa Fluor 488-labeled or PE-labeled mouse anit-IgG1, κ and Alexa Fluor 647-labeled Armenian hamster anti-IgG; BioLegend). All samples were first gated on a forward scatter (FS)/side scatter (SS) plot and selected accordingly to exclude non-viable cells and debris.

Data were collected using a FACS ARIA II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) and analyzed by FlowJo software 9.9.3 (LLC, Ashland, OR, USA).

2.7. Colony Forming Unit-Fibroblast (CFU-F) assay

Naïve, SNI+VEH and SNI+IMT504 rats were decapitated (n= 6 per group) and BMMCs were obtained as described above. Following previously published protocols,⁵² 2×10⁶ BMMCs were plated into 6-well tissue culture plate (Greiner Bio-One, Kremsmünster, Austria) in 3 ml DMEM supplemented with glutamine (2 mM; Gibco), penicillin (100 U/ml), streptomycin sulfate (100 μg/ml), amphotericin B (2.5 g/ml; Sigma) and 10% fetal bovine serum (FBS; Natocor, Córdoba, Argentina) and incubated at 37°C in an atmosphere of 5% CO₂ air. Unattached cells were removed by washing adherent cells twice with PBS. On day 14, cultures were fixed and stained with Giemsa (Biopack, Buenos Aires, Argentina) (Fig. 1C).

Triplicate analysis per rat was performed, followed by quantification of the number of CFU-F (colonies containing 50 or more cells) in each plate, using a light microscopy (Eclipse E-800, Nikon, Tokyo, Japan) at 10X magnification. Data was expressed as mean ± SEM CFU-F from 2×10⁶ plated BMMCs.

2.8. Isolation of rat and human MSCs

Naïve rat BMMCs were obtained as mentioned in Section 2.7 (8 rats were used for collection) and plated in p100 Petri dish (Greiner Bio-One) with DMEM supplemented with glutamine (2 mM), penicillin (100 U/ml), streptomycin sulfate (100 μg/ml), amphotericin B (2.5 g/ml) and 10% FBS and incubated at 37°C in an atmosphere of 5% CO₂ air. Non-adherent cells were discarded and cultures continued until confluence, renewing the medium every 3–4 days. At this time, cells were harvested with trypsin-EDTA (0.05%–0.02%; Sigma) and cultured at 10⁴ cells per cm² (passage 1). This process was repeated until passage number 4, for enrichment of MSCs.

Human MSCs were obtained from healthy donors (Hospital Naval Pedro Mallo, Buenos Aires, Argentina) and were characterized according to the International Society for Cellular Therapy (ISCT) guidelines as described previously.¹⁹

2.9. In vitro migration assays of IMT504-pre-treated rat and human MSCs

In vitro migration assays were performed to determine the effect of in vitro IMT504 pre-treatment on the ability of MSCs to migrate towards conditioned medium obtained from naïve or injured rat sciatic nerves (Fig. 1D).

2.9.1. Sciatic Nerve Conditioned Medium

To obtain sciatic nerve conditioned medium (CM), naïve rats or rats sacrificed 7 days after SNI (n= 4 per group) were deeply sedated using a CO₂ chamber followed by quick decapitation. After dissection of the right/ipsilateral and left/contralateral sciatic nerve segments from naïve and injured rats (approximately 1 cm proximal from its trifurcation/ligation), were immediately placed in DMEM supplemented with penicillin (100 U/m) and streptomycin sulfate (100 μg/ml), and left for 30 min in ice before further processing. Subsequently, sciatic nerves belonging to each experimental group and side were transferred in pairs to a 6-well tissue culture plate loaded with 2 ml of DMEM supplemented without FBS and incubated at 37°C in an atmosphere of 5% CO₂ air. Conditioned medium was harvested 16 h later and stored separately at −80°C until use.

2.9.2. In vitro migration assay

In vitro migration was performed using a 48-Well Micro Chemotaxis Chamber (Neuroprobe Inc., Maryland, USA) as previously described.¹⁹ Rat and human MSCs, in vitro pre-treated or not with IMT504 (7μg /ml IMT504 in supplemented DMEM without FBS; 16 h), were placed at a density of 1.2×10³ cells/well in the upper wells, which were separated from the lower wells by an 8 μm pore polycarbonate filter (Nucleopore membrane, Neuroprobe Inc.). DMEM or CM were applied to the lower wells. Chambers were incubated for 5 h at 37°C in a 5% CO₂ humidified atmosphere. After that, the membrane was carefully removed and cells attached to the lower side were fixed in 2% paraformaldehyde and washed 3 times with PBS 1x. Cells on the upper side of the membrane were scraped off with a blade. Attached cells were stained with DAPI (Sigma) and counted using fluorescent-field microscopy (Eclipse Ni, H-600L, Intensilight C-HGFIE, Nikon, Tokyo, Japan) at 10x.

Three independent experiments were run, and within each experiment, quadruplicate analysis for each condition was tested. Quantification was performed on 3 representative visual fields per iteration, per experiment, per group, using the Cell Profiler software. Data was expressed as mean of migrating cells/field ± SEM.

2.10. In vitro analysis of IMT504 internalization in rat and human MSCs

For analysis of IMT504 internalization, MSCs (obtained as described above) were cultured in coverslips loaded with 7μg/ml IMT504 conjugated either with Alexa Fluor 488 or Texas Red (Biosynthesis, Lewisville, TX, USA) (Fig. 1E). Sixteen h after incubation, cultures were washed with PBS, fixed with 4% paraformaldehyde in PBS for 30 min and washed 3 times (5 min each) with PBS and permeabilized with Triton-X100 (0,2% v/v, Sigma) for 5 min. This was followed by 1 h incubation at room temperature in BSA (5% w/v in PBS, blocking solution (Sigma)) for non-specific binding blockade. Cells were incubated overnight at 4°C in anti-vimentin antiserum (1:1000, ab45939; Abcam, Cambridge, UK) diluted in blocking solution. After three 5 min washes in PBS, cultures were incubated for 1 h at room temperature either with an FITC-conjugated goat anti-rabbit IgG antiserum (1:200; Jackson ImmunoResearch, West Grove, USA) or an rhodamine-conjugated goat anti-rabbit IgG antiserum (1:200; Jackson ImmunoResearch), diluted in blocking solution. Cell nuclei were visualized by DAPI staining (4’,6-diamidino2-phenylindole dihydrochloride; 1 ug/ml; Sigma, diluted in PBS) after a 10 min incubation with the nuclear tracer. Coverslips were mounted with 2.5 % DABCO (1,4-diazobicyclo-[2.2.2]-octane; Sigma) in glycerol and stored at −20°C until analysis. Photomicrographs of cultured cells incubated with IMT504-Alexa Fluor 488 were taken using a fluorescence microscope (Eclipse Ni, H-600L, Intensilight C-HGFIE, Nikon, Tokyo, Japan) and captured with a digital camera (Nikon, DS-Qi2, Tokyo, Japan) using 20x (Plan Apo λ 0.75, OFN25, DIC N2) and 40x (Plan Apo λ 0.95, 0.11–0.23, WD0.25 − 0.17) objectives. All images were taken using the same light intensity and exposure parameters. In addition, cultured cells incubated with Texas Red were viewed under an Olympus Fluoview FV1000 laser scanning confocal 60X (Olympus, Tokyo, Japan). Images were processed and analyzed using the FV10-ASW Fluoview software (Olympus) and ImageJ (NIH, Bethesda, MD).

2.11. ELISA quantification of TNF-α, IL-1β, IL-10 and TGF-β1 levels in rat sciatic nerve

For the measurement of cytokines levels, the ipsilateral sciatic nerve (or right-side nerve in the case of naïve rats) was collected from deeply anesthetized (as above) naïve, SNI+VEH and SNI+IMT504 rats, 3 and 21 days after treatment (n= 3 per group) (Fig. 1F). Approximately 1 cm of each ipsilateral nerve was removed proximal to the position of the injury and rapidly frozen and stored at −80°C. Tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), interleukin-10 (IL-10) and transforming growth factor-β1 (TGF-β1) levels were measured using commercially available immunoassay ELISA kits for rat (Abcam, Cambridge, UK) according to the manufacturer’s instructions. The results are expressed as picograms of cytokine per milligram of tissue protein (pg/mg).

2.12. Co-culture of MSC (pre-treated with IMT504 or not) with ipsilateral SNI nerves

MSCs were seeded in 6-well plates and incubated in DMEM supplemented with glutamine (2 mM), penicillin (100 U/ml), streptomycin sulphate (100 μg/ml), amphotericin B (2.5 g/ml) and 10% FBS at 37°C in an atmosphere of 5% CO₂ air. Once cultured MSCs reached 80% confluence, cells were pre-treated or not with 7μg /ml IMT504 or 7μg/ml PolyC (in supplemented DMEM without FBS; 16 h) overnight. Cells were then washed with PBS and a polycarbonate cell culture insert (pore size of 0.4 μm, Millipore, Billerica, MA) was placed into each well, on top of the cells. This was followed by placement of two ipsilateral 7 day-SNI nerves (approximately 1 cm from the ligation) dissected out from injured rats, on top of the insert (n= 32). Co-cultures were maintained overnight in supplemented DMEM without FBS at 37°C and 5% CO₂ in a humidified atmosphere (Fig. 1G). Additional wells were prepared with MSCs or ipsilateral 7 day-SNI alone, and incubated or not with IMT504 (7μg /ml) or PolyC (7μg /ml), in supplemented DMEM without FBS at 37°C and 5% CO₂ in a humidified atmosphere during 16 h. Four independent experiments were run for each condition being tested.

2.13. Quantitative real time reverse-transcription-PCR (RT-qPCR) analysis

Total RNA from in vivo ipsilateral sciatic nerve (from naïve, SNI+VEH and SNI+IMT504 rats, 3 and 21 days after treatment; n= 6 per group), in vitro MSCs (n= 4 per group) and ex vivo ipsilateral 7 day-SNI nerves (n= 4 per group) was extracted and purified using an RNA clean and concentrator kit (Zymo Research, Irvine, USA) according to manufacturer’s protocol followed by cDNA synthesis using Moloney Murine Leukemia Virus Reverse Transcriptase (M-MLV RT) (Promega, Fitchburg, USA). cDNAs were subjected to real-time polymerase chain reaction (AriaMax, Agilent Technologies, California, USA). The mRNA levels were quantified by SYBR® Green (Bio-rad, California, USA): Cxcl12, Cxcr4, Tnf-α, Il-1β, Il-10 and Tgf-β1, according to the experiment. The primers are detailed in the Supplementary Table 1. Values were normalized to levels of β- actin (used as housekeeping) transcript. Data were processed by the ΔΔCt method and relative to the amount of the PCR product amplified from in vivo naïve nerves, in vitro MSCs cultured with DMEM or ex vivo ipsilateral 7 day-SNI nerves culture with DMEM. A non-template control was run in every assay, and all determinations were performed in triplicates (Fig. 1G).

2.14. Statistical Analysis

In all experiments, data is expressed as mean ± S.E.M and evaluated using GraphPad Prism 7.0. Behavioral data was statistically analyzed using two-way ANOVA followed by Tukey’s post-hoc test. Flow cytometry, CFU-F assay, migration assay, ELISA and RT-qPCR data was statistically analyzed using one-way ANOVA followed by Dunnett’s post-hoc test for comparisons with the control group (in vivo naïve nerves, in vitro MSCs culture with DMEM or ex vivo ipsilateral 7 day-SNI nerves culture with DMEM) and for multiple comparisons between groups one-way ANOVA followed by Sidak’s post-hoc test. Alpha value was set at P<0.05 (*P< 0.05, **P< 0.01, ***P< 0.001, ****P< 0.0001). Exact P-values were presented when it corresponded.

3. Results

3.1. IMT504 blocks SNI-induced mechanical and cold allodynia during prolonged periods of time, without affecting regular nocifensive responses

We began our analysis by determining if administration of IMT504 in rats with ongoing neuropathic pain induced by SNI was effective in reducing or blocking mechanical and cold allodynia. All injured rats, previous to IMT504, VEH or PolyC administration, exhibited clear signs of ipsilateral mechanical allodynia, as well as an increase in cold withdrawal frequency and score, starting 3 days after injury and onwards. Moreover, and as expected, SNI+VEH and SNI+PolyC rats remained allodynic for at least 42 days after injury (Fig. 2).

Fig. 2. In vivo administration of IMT504 blocks mechanical and cold allodynia.

(A-B) Mechanical withdrawal thresholds (A) and cold withdrawal frequencies (B upper) or cold withdrawal scores (B lower), expressed as mean ± SEM, from the right hindpaws of naïve rats (black squares, black line) and ipsilateral hindpaws of SNI+VEH (black circles, black line), SNI+PolyC (black rhombus, brown line) and SNI+IMT504 (black triangle, blue line) rats. Data represents an n= 6 per group. Statistical significance was established by means of two-way ANOVA followed by Tukey’s post-hoc test: P< 0.0001 among groups; *P< 0.05, **P< 0.01 and ****P< 0.0001 ipsilateral SNI+VEH or ipsilateral SNI+PolyC vs ipsilateral SNI+IMT504; ^#P< 0.05, ^##P< 0.01, ^###P< 0.001 and ^####P< 0.0001 ipsilateral SNI+VEH, ipsilateral SNI+PolyC or ipsilateral SNI+IMT504 vs naïve.

Administration of a single dose of IMT504 in SNI rats, 7 days after injury, resulted in a quick recovery towards basal withdrawal thresholds, as compared to SNI+VEH or SNI+PolyC rats (Fig. 2A). The anti-allodynic effect was evident already 1 day after treatment, becoming comparable to naïve basal withdrawal levels 2 weeks after treatment and onwards (Fig. 2A). In addition, IMT504 significantly decreased the frequency and score of allodynic responses elicited by paw stimulation with acetone from day 1 after treatment, and approaching basal values, comparable to those observed in naïve rats, for several weeks after (Fig. 2B). The contralateral hindpaws of all injured rats did not reveal any significant change in basal withdrawal thresholds or frequencies (data not shown).

Analysis of responses to acute hot thermal stimuli, at baseline for all animals, or 3 or 21 days after treatment in SNI+VEH, SNI+PolyC and SNI+IMT504 rats, displayed lack of statistically significant differences in tail-flick latencies (Supplementary Fig. 1). This supports the notion that the anti-allodynic effects of IMT504 detected thus far do not interfere with regular nocifensive responses to hot thermal stimulation.

3.2. IMT504 potentiates the mobilization and homing of MSCs into injured peripheral nerves, with participation of the CXCL12-CXCR4 axis

Having established that a single dose of IMT504 efficiently modulates pain-like behavior in rats undergoing persistent neuropathic pain, we evaluated whether the effect of the ODN is associated with the mobilization and homing of endogenous MSCs in the injured nerves. We chose 3 and 21 days post-treatment (10 or 28 days after injury, respectively), as sampling time-points to evaluate the presence of MSCs by flow cytometry in sciatic nerve, peripheral blood, and bone marrow in naïve, SNI+VEH and SNI+IMT504 rats, and potential time-dependent effects (see Supplementary Fig. 2 for flow cytometry gating strategies and panels). This analysis was complemented with an in vitro quantification of the number of CFU-F generated by bone marrow collected from all three groups, and an in vivo evaluation of the transcript expression levels of cxcl12 and cxcr4 at the injured nerves, at both time-points (Fig. 3).

Fig. 3. In vivo administration of IMT504 promotes the mobilization and homing of MSCs into injured sciatic nerves, and modulates the CXCL12-CXCR4 axis.

(A-C and E-G) Quantification of MSCs (CD45^-CD90⁺CD29⁺ cells) in ipsilateral sciatic nerve (A and E), peripheral blood (B and F) and bone marrow (C and G), and analysis of CFU-F units in bone marrow (D and H) from naïve (gray columns), SNI+VEH (black columns) or SNI+IMT504 (blue columns) rats, 3 (A-D) or 21 (E-H) days after treatment. Data represents an n= 5 (A-C and E-G) or n= 6 (D and H) rats per group, per time-point and are expressed as mean ± SEM. (I-L) Cxcl12 (I and K) and Cxcr4 (J and L) mRNA levels were measured in naïve (gray columns), SNI+VEH (black columns) or SNI+IMT504 (blue columns) rats, 3 (I and K) or 21 (J and L) days after treatment. Data represents an n= 6 rats per group, per time-point and are expressed as mean ± SEM. Statistical significance was established by means of one-way ANOVA followed by the Dunnettś multiple comparison test: P <0.0001 among groups; *P<0.05, **P< 0.01, ***P< 0.001 and ****P< 0.0001 naïve vs SNI+VEH or SNI+IMT504, and followed by the Sidakś multiple comparison test: P <0.0001 among groups; ^#P<0.05, ^##P< 0.01, ^###P< 0.001 and ^####P< 0.0001 SNI+VEH vs SNI+IMT504.

Analysis 3 days after treatment in SNI+VEH rats showed considerable increases in the percentage of endogenous MSCs present in ipsilateral sciatic nerve (Fig. 3A) and peripheral blood (Fig. 3B), as compared to naïve rats. This was paralleled by a reduction in the percentage of MSCs in the major reservoir of MSCs, the bone marrow (Fig. 3C), a result that was confirmed through in vitro CFU-F analysis, where the number of colonies in SNI+VEH animals was lower compared to naïve rats (Fig. 3D). Treatment with IMT504 potentiated the effects mentioned above. Thus, SNI+IMT504 rats exhibited a higher percentage of MSCs detected in ipsilateral sciatic nerve (Fig. 3A) and peripheral blood (Fig. 3B) than observed in SNI+VEH, and also resulted in a significant reduction of MSCs percentage and CFU-F counts in the bone marrow (Fig. 3C, D).

Evaluation 21 days after treatment revealed that MSCs percentage (Fig. 3E–G) and CFU-F counts (Fig. 3H) in SNI+VEH animals returned to values comparable to those observed in naïve rats in all analyzed tissues (Fig. 3E–H). In contrast, a higher percentage of MSCs was observed in ipsilateral sciatic nerves and peripheral blood in SNI+IMT504 than in SNI+VEH rats (Fig. 3E, F). Furthermore, a modest increase in the percentage of MSCs was observed in the bone marrow of SNI+IMT504 rats, although still remaining lower than those observed in SNI+VEH or naïve rats (Fig. 3G); CFU-F counting confirmed this observation (Fig. 3H).

Analysis in contralateral nerves of SNI+VEH and SNI+IMT504 rats, 21 days after treatment, revealed no significant changes in the percentage of MSCs, remaining comparable to those observed in naïve rats (compare Fig. 3E and Supplementary Fig. 2H), and suggesting that the pro-migratory effects of IMT504 were localized to the site of the nerve injury.

Finally, evaluation of cxcl12 and cxcr4 transcript levels in injured nerves, 3 days after treatment, showed upregulated expression of the chemokine and its associated receptor both, in VEH- and IMT504-treated rats; however, ODN-treated rats exhibited the strongest upregulations (Fig. 3I, J). In contrast, analysis performed 21 days after treatment revealed that while VEH-treated rats retained upregulated levels of cxcl12 transcript (although much lesser than observed 3 days after vehicle treatment), SNI+IMT504 rats exhibited a return to naïve-comparable levels in injured nerves. Cxcr4 transcript levels, on the other hand, remained upregulated in both experimental groups 21 days after treatment, although VEH-treated rats exhibited a higher upregulation than IMT504-treated rats (Fig. 3I, J).

3.3. Exposure to IMT504 upregulates the expression of cxcr4 in MSCs and stimulates their migration

We next evaluated, through in vitro and ex vivo experimentation, whether the expression of cxcr4 transcript is modulated by in vitro IMT504 pre-treatment in rat MSCs, and the influence of exposure of these cells to injured nerves. This was followed by analysis of the effect of in vitro IMT504 pre-treatment on rat and human MSCs migration abilities towards conditioned media derived from rat injured nerves (CM) obtained under different conditions.

While the expression levels of cxcr4 in rat MSCs was regularly low, either with or without PolyC pre-treatment, previous treatment with IMT504 resulted in a very significant induction of the receptor (Fig. 4A). A comparable induction was obtained when untreated or PolyC-pre-treated MSCs were exposed to injured nerves. Interestingly, exposure of IMT504-pre-treated MSCs to injured nerves resulted in the highest upregulation of cxcr4 levels (Fig. 4A). Collectively, these results are consistent with the upregulated expression of cxcr4 observed in nerve samples collected in vivo, 3 days after injury (Fig. 3J).

Fig. 4. In vitro IMT504 exposure induces the expression of Cxcr4 transcript in rat MSCs and promotes rat or human MSCs chemotaxis.

(A) Cxcr4 mRNA levels were measured in untreated (gray column), PolyC-pre-treated (light-brown column) and IMT504-pre-treated (light-blue column) rat MSCs cultured in DMEM, as well as in untreated (black column), PolyC-pre-treated (brown column) or IMT504-pre-treated (blue column) rat MSCs co-cultured with 7 day-SNI ipsilateral nerves (rat MSCs cultured in DMEM served as control). Data represents an n= 4 and are expressed as mean ± SEM. Statistical significance was established by means of one-way ANOVA followed by the Dunnetś multiple comparison test: P< 0.0001 among groups; ****P< 0.0001 between control (untreated MSCs cultured in DMEM) and other groups, and followed by the Sidakś multiple comparison test: P <0.0001; ^##P< 0.01, ^###P< 0.001 and ^####P< 0.0001 between all groups tested. (B-C) In vitro chemotaxis assay of untreated (black columns) or IMT504-treated rat (B) or human (C) MSCs (blue columns) towards injured nerve conditioned media (CM) derived from rat naïve, 7 day-SNI ipsilateral or contralateral nerves. Results are expressed as the mean of migrating cells/field ± SEM. Statistical significance was established by means of one-way ANOVA followed by the Dunnettś multiple comparison test: P< 0.0001 among groups; **P< 0.01, ***P< 0.001 and ****P< 0.0001 MSCs vs MSCs+IMT504. Additional analysis comparing the effect on MSCs or MSCs+IMT504 groups to different CM exposures was analyzed by one-way ANOVA followed by Sidak’s multiple comparison test: P< 0.0001 among groups; ^#P< 0.05, ^##P< 0.01 and ^####P< 0.0001.

Analysis of migratory behavior revealed that only a small number of either rat (Fig. 4B and Supplementary Fig. 3a, e) or human (Fig. 4C and Supplementary Fig. 3i, m) MSCs, incubated or not with IMT504, migrated towards the polycarbonate filter in the presence of DMEM. In contrast, rat (Fig. 4B) or human (Fig. 4C) MSCs exhibited a natural tendency to migrate towards CM from naïve (Supplementary Fig. 3b, j), ipsilateral (Supplementary Fig. 3c, k) and contralateral (Supplementary Fig. 3d, l) nerves (of note, the CM from naïve and contralateral nerves in this experimental setting could represent a 1 day injury – nerve collection, which is the time these nerves spent in culture to obtain the corresponding CM. This likely explains their pro-migratory effect upon MSCs). IMT504 pre-treatment enhanced the migration of rat (Fig. 4B) and human (Fig. 4C) MSCs towards the CM from most nerve conditions (Fig. 4B, 4C and Supplementary Fig. 3f–h, 3n–p). However, and importantly, the strongest pro-migratory effect of IMT504 was observed when rat (Fig. 4B and Supplementary Fig. 3g) or human (Fig. 4C and Supplementary Fig. 3o) MSCs were exposed to CM from 7 day-SNI ipsilateral sciatic nerves.

Finally, analysis of rat (Fig. 5a–h) and human (Fig. 5i–p) MSCs incubated overnight with 7 μg/mL of an IMT504-Alexa Fluor 488 conjugate revealed that most cells have the ability to internalize naked ODN, appearing as positive fluorescent granules distributed throughout the cytoplasm. Confocal analysis of MSCs incubated overnight with 7 μg/mL of an IMT504-Texas Red conjugate confirmed the cytoplasmic distribution of the ODN, without apparent presence within the nucleus (Supplementary Fig. 4a, b).

Fig. 5. In vitro exposure of rat or human MSCs to IMT504 reveals cellular internalization of the ODN.

(a-p) Photomicrographs of rat (a-h) and human (i-p) MSCs. In green, IMT504-Alexa Fluor 488 fluorescent signal (a, e, i, m); in red, vimentin immunofluorescent signal (red; b, f, j, o), and in blue DAPI fluorescent signal (c, g, k, o) (d, h, l, p show merged images). Arrowheads point to MSCS exhibiting presence of IMT504-Alexa Fluor 488-positive intracellular granules. Scale bars: 100 μm (d=a-c, i-l); 50 μm (h=e-g, m-p).

3.4. IMT504 modulates the inflammatory milieu in the injured nerve of SNI rats

To address the possibility that the anti-allodynic and MSCs-pro-migratory effects of IMT504 is associated with the modulation of inflammatory processes taking place during SNI, we evaluated the protein levels of pro- and anti-inflammatory cytokines in the injured nerves.

Cytokine protein levels were measured 3 and 21 days after treatment in the ipsilateral sciatic nerves of SNI+VEH and SNI+IMT504 rats, and also in naïve rats (Fig. 6). Analysis in SNI+VEH rats showed a marked increase in the expression levels of the pro-inflammatory cytokines, TNF-α (Fig. 6A) and IL-1β (Fig. 6B) at both time-points evaluated, relative to naïve rats. Analysis of anti-inflammatory cytokines in these animals showed a non-statistically significant trend to increase in IL-10, 21 days after injury (Fig. 6C). Increased levels of TGF-β1 were only observed 3 days after treatment, with values returning to near basal levels 21 days after treatment (Fig. 6D).

Fig. 6. In vivo administration of IMT504 modulates the protein levels of pro- and anti-inflammatory cytokines in the injured nerve of SNI rats.

(A) TNF-α, (B) IL-1β, (C) IL-10 and (D) TGF-β1 were measured in naïve (gray columns), SNI+VEH (black columns) or SNI+IMT504 (blue columns) rats, 3 and 21 days after treatment. Data represents an n= 3 rats per group, per time-point and are expressed as mean ± SEM. Statistical significance was established by means of one-way ANOVA followed by the Dunnetś multiple comparison test: P< 0.0001 among groups; *P< 0.05, **P< 0.01, ***P< 0.001 and ****P< 0.0001 naïve vs SNI+VEH or SNI+IMT504, and followed by the Sidakś multiple comparison test: P< 0.0001 among groups; ^#P< 0.05 and ^##P< 0.01 SNI+VEH vs SNI+IMT504.

In contrast to SNI+VEH rats, SNI+IMT504 rats did not exhibit upregulated levels of TNF-α and IL-1β (Fig. 6A, B), as demonstrated 3 days after treatment. In fact, the values detected in ODN-treated rats were comparable to those observed in naïve rats at this time-point. Twenty-one days after treatment, the expression levels of TNF-α in SNI+IMT504 rats exhibited values comparable to those detected in SNI+VEH rats (Fig. 6A); however, the expression levels of IL-1β remained significantly downregulated in IMT504-treated rats (Fig. 6B). As for the expression levels of anti-inflammatory cytokines, IL-10 levels were found strongly enhanced 3 and 21 days after treatment in rats receiving the ODN, although no significant differences were detected between SNI+VEH and SNI+IMT504 rats 21 days after treatment. In contrast, TGF-β1 was found significantly upregulated in IMT504-treated rats at both time-points evaluated (Fig. 6C, D).

3.5. IMT504 potentiates an anti-inflammatory profile in rat MSCs and promotes changes in the inflammatory milieu of injured nerves in vitro

We next determined whether IMT504 modulates, ex vivo, the transcript expression levels of the main immunoregulatory cytokines produced by rat MSCs, Tgf-β1 and Il-10, and the impact these changes may have on the inflammatory environment present in the injured sciatic nerve collected from SNI rats. For this purpose, we carried out single or co-culture experiments of MSCs and 7 day-SNI ipsilateral sciatic nerves, and the effect of IMT504 in different scenarios (Fig. 1).

First, we evaluated whether the incubation with IMT504 changed the expression levels of anti-inflammatory cytokine genes in MSCs. We found that IMT504 significantly increased the expression of Il10 and Tgf-β1 mRNA in MSCs when compared with cells exposed to DMEM or PolyC (Fig. 7A, B). Next, we evaluated whether exposure of untreated or PolyC-pre-treated MSCs to injured nerves obtained from SNI rats alters the changes mentioned above in Il10 and Tgf-β1 mRNA expression. Both conditions resulted in increases in Tgf-β1 and Il-10 mRNA expressions, virtually identical to what was observed for IMT504-pre-treated MSCs cultured in DMEM (Fig. 7A, B). Finally, IMT504-pre-treated MSCs exposed to injured sciatic nerves showed the most pronounced increase in Il-10, while a decrease in of Tgf-β1 mRNA expression was observed (Fig. 7B,C).

Fig. 7. IMT504 acts on MSCs and the injured nerve to promote an anti-inflammatory milieu.

(A-B) Il-10 (A) and Tgf-β1 (B) mRNA levels were measured in untreated (gray columns), PolyC-pre-treated (light-brown columns) and IMT504-pre-treated (light-blue columns) MSCs cultured in DMEM, and in untreated (black columns), PolyC-pre-treated (brown columns) or IMT504-pre-treated (blue columns) MSCs co-cultured with 7 day-SNI ipsilateral nerves (MSCs cultured in DMEM served as control). (C-F) Tnf-α (C), Il-1β (D), Il-10 (E) and TGF-β1 (F) mRNA levels were measured in untreated (gray columns) PolyC-pre-treated (light-brown columns) or IMT504 treated (light-blue columns) 7 day-SNI ipsilateral nerves, or in untreated (black columns), PolyC-pre-treated (brown columns) or IMT504 pre-treated (light-blue columns) MSCs co-cultured with 7 day-SNI ipsilateral nerves (nerves incubated in DMEM served as control). Data are expressed as mean ± SEM. Statistical significance was established by means of one-way ANOVA followed by the Dunnetś multiple comparison test: P< 0.0001 among groups; *P< 0.05, ** P< 0.01, ***P< 0.001 and ****P< 0.0001 between control (untreated MSCs cultured in DMEM or 7 day-SNI ipsilateral nerves cultured in DMEM, as appropriate) and other groups, and followed by the Sidakś multiple comparison test: P <0.0001; ^#P< 0.05, ^##P< 0.01, ^###P< 0.001 and ^####P< 0.0001 between all groups tested.

Taking advantage of the two-way crosstalk nature of the transwell co-culture system, we also evaluated effects on the injured nerves. Ipsilateral sciatic nerves cultured only with DMEM, PolyC or IMT504 were included in the analysis. Compared to nerves cultured in DMEM or PolyC, their incubation with IMT504 alone resulted in decreased Tnf-α (Fig. 7C) and Il-1β mRNA levels (Fig. 7D), but no changes in the expression of Il-10 (Fig. 7E) or Tgf-β1 mRNA (Fig. 7F). Co-culture of injured nerves with untreated or PolyC-pre-treated MSCs also resulted in the downregulation of Tnf-α and Il-1β1 mRNAs (Fig. 7C, D) and a significant upregulation of Tgf-β1 mRNA; Il-10 mRNA levels remained unchanged (Fig. 7E–F). Finally, co-culturing injured nerves with IMT504-pre-treated MSCs resulted in significantly downregulated mRNA levels of Tnf-α (Fig. 7C) and Il-1β (Fig. 7D), and considerable upregulation in Il-10 (Fig. 7E) and Tgf-β1 (Fig. 7F) mRNA expressions. Moreover, compared to the other experimental groups, the latter condition resulted in the most pronounced downregulation of Tnf-α and upregulation of Il-10 and Tgf-β1 mRNA (Fig. 7C–F).

4. Discussion

Our present study closes the conjectural circle that originally supported the association of the anti-allodynic effects of IMT504 with the ODN’s pro-mobilization effects on endogenous MSCs,¹¹ by showing: 1) That the rapid onset and long-lasting anti-allodynic effects of IMT504 in rats with SNI directly correlates with the rapid and long-lasting mobilization and homing of MSCs within the injured peripheral nerves. 2) That IMT504 evokes an anti-inflammatory phenotype in MSCs, which likely engage in complex interactions with local and infiltrating immune cells within injured nerves, and promote the resolution of the local inflammatory process. And 3) that IMT504 retains the pro-migratory effects in human MSCs in vitro, strongly supporting the translational potential of this ODN to treat neuropathic pain conditions with inflammatory components. In addition, our results also support the hypothesis that IMT504 may directly modulate local and/or infiltrating immune cells at the site of injury.

The migratory capacity of MSCs towards injured nerves, or conditioned media obtained from injured nerves, in the absence of IMT504 treatment implies their readily response to local chemical cues.⁴⁸ In fact, peripheral nerve injury results in increased local expression of a variety of molecules involved in neuroinflammation,⁴⁸ including some acting as chemoattractants.^37,55,62 Factors known to influence MSCs migration and homing include pro-inflammatory cytokines such as IL-1β and TNF-α, both of which were found upregulated in injured nerves as shown here in vivo and in vitro. The chemokine CXCL12 and its counterpart receptor CXCR4,^35,50 synthesized by rodent nociceptive DRG neurons⁴³ and locally upregulated after nerve injury⁴⁹ were also found to be key players in the migration of intrathecally administered MSCs in mice with chronic constriction injury (CCI).⁶ Here we validated the influence of the CXCL12-CXCR4 axis over endogenous MSCs under nerve injury conditions. Remarkably, IMT504 potentiated this axis in vivo and in vitro, which provides a mechanistic explanation to the pro-migratory role of the ODN. However, it should be noted that the in vivo effect of IMT504 on the CXCL12-CXCR4 axis appeared to weaken over time, suggesting the participation of additional mechanisms contributing to the pro-migratory actions of the ODN over MSCs.

The migration of MSCs to the injured nerve could be understood as a mechanism to prevent an excessive inflammatory reaction after injury. However, it seems that in SNI-rats this is not sufficient to reduce pain-like behaviors, nor the upregulated expression of pro-inflammatory cytokines in injured nerves. This may suggest that the anti-allodynic and anti-inflammatory actions of IMT504 not only depend on potentiating the migratory capacity of MSCs towards injured nerves, but also their anti-inflammatory functionality.

Notably, here we show that IMT504 induces the transcript expression of two potent anti-inflammatory and anti-nociceptive cytokines in cultured MSCs: Il10 and Tgf-β1. IL-10, secreted not only by MSCs, but also by activated T and B cells, macrophages and mast cells, is a powerful anti-inflammatory cytokine that inhibits the release of IL-1β, IL-6, and TNF-α.^2,58,61 Similarly, TGF-β1, a multifunctional cytokine belonging to a superfamily of transforming growth factors, has been shown to induce anti-inflammatory and anti-nociceptive effects.^30,56 In addition, our ex vivo analysis also revealed the existence of complex cross-talk interactions between MSCs and the inflammatory milieu, where exposure of untreated MSCs to injured nerves induced the expression levels of Il-10 and TGF-β1, supporting the protective role of these cells and the potentiation of this function by the ODN. Interestingly, co-incubation of IMT504-pre-treated MSCs with injured nerves further induced the expression of Il-10 transcript in these cells, but resulted in a considerable reduction in their expression levels of TGF-β1. Nevertheless, the net effect of these interactions in injured nerves, in the presence of IMT504, remained to be anti-inflammatory.

The mechanistic hypothesis emerging from our in vitro and ex vivo experiments is supported by the molecular findings of our in vivo studies. Thus, IMT504-treated rats display an early, long-lasting (>21 days) and robust increase in IL-10 and TGF-β1 protein levels associated with marked reductions of TNF-α and IL-1β proteins at the injured nerve site. The latter are two potent pro-inflammatory cytokines, strongly increased in injured peripheral nerves^2,44 at the expense of resident macrophages, mast and Schwann cells, as well as of infiltrating neutrophils and macrophages^33,42,47, and promoting the development and maintenance of neuropathic pain.^2,21 Altogether, our present data supports that the mechanism of anti-nociceptive action of infiltrating MSCs in IMT504-treated rats is the modulation of neuroimmune interactions at the local milieu and the restoration of a balance between pro- and anti-inflammatory cytokines, where IL-10 and TGF-β1 acquire key roles. Such an homeostatic mode of action agrees well with the concept that an efficient treatment against neuropathic pain should modulate the balance towards an anti-inflammatory scenario, acting on more than one cytokine.¹⁷

Our results are also in line with increasing evidence on the therapeutic actions of MSCs, which appear to depend on the paracrine release of a variety of mediators^4,26,51 capable of influencing the transformation of macrophages from an M1 (pro-inflammatory) into an M2 (anti-inflammatory) phenotype;¹⁸ inhibiting T-cell proliferation;⁷ and suppressing activated CD4+ and CD8+ T cells and B-cells.²⁰ Similarly, systemic administration of human MSCs in mice undergoing a 4 day-long SNI, reduces pain-like behaviors, in association with their homing in the spinal cord and pre-frontal cortex (not the sciatic nerve), plus decreased levels of IL-1β and IL-17, and increased levels of IL-10 in the spinal cord.⁵³ Intrathecal MSCs in mice with CCI also block pain, likely through release of TGF-β1, but not IL-10.⁶ And in rats with spinal nerve ligation, intrathecal IL-1β-pre-treated MSCs alleviate neuropathic pain more potently than MSCs alone, through an IL-10 and TGF-β1-mediated modulation of microglia and astrocytes.³⁴ More recently, systemic¹⁷ or intrathecal⁵¹ administration of MSCs by-products, such as MSC-derived conditioned media¹⁷ or -exosomes⁵¹, were shown to efficiently reduced pain-like behavior in mice with partial sciatic nerve ligation¹⁷ or rats with SNL.⁵¹ The effects were also associated with reductions in sciatic nerve and spinal cord protein levels of TNF-α and IL-1β,¹⁷ and reduced expression of TNF-α and IL-1β plus upregulation of IL-10, brain-derived- and glial cell line-derived neurotrophic factors in L4–5 DRGs.⁵¹ All these data are remarkably similar to our IMT504 findings, further supporting our mechanistic hypotheses.

The observation of reduced transcript expression of pro-inflammatory cytokines in nerves only incubated with IMT504 also suggests direct effects on the microenvironment of the injured peripheral nerve. Accordingly, we recently showed that systemic IMT504 downregulates the expression of a large number of pro-inflammatory molecules, with parallel decreases in the number of macrophages and B-lymphocytes, in the hindpaw skin of rats with persistent inflammation.³² An early direct action onto immune cells at the site of injury could be possible, since IMT504 is detectable in the blood stream during the first 4 hours after subcutaneous administration.¹⁶ Curiously, direct ex vivo application of IMT504 onto injured nerves did not induce the expression of Il-10 or TGF-β1 as was the case when nerves were exposed to IMT504-treated MSCs. Therefore, it could be speculated that while direct, early immunomodulation by IMT504 is plausible, the substantial ex vivo and in vivo effects on anti-inflammatory cytokines, plus the long-lasting anti-allodynic actions of IMT504 depended on the modulation of MSCs, particularly as time after treatment progresses.

One limitation of our study is that it does not allow us to determine whether IMT504 is equally effective, or modulates pain signaling through similar mechanisms in male and female rats. Recent studies have brought to attention the influence of sex^22,36,40,54 or developmental stage⁵⁷ in the physiopathology of chronic pain. In particular, differences in the participation of spinal microglia^38,40,54 and peripheral immune cells^36,54 have been exposed. Future studies on the influence of sex on the effects and mechanisms of action of IMT504 are warranted.

Finally, from a therapeutic standpoint, IMT504 emerges as a promising option against neuropathic pain. On the one hand, the replication of pro-migratory effects in human MSCs further enhances the clinical potential of IMT504. Previously failed clinical trials with promising drugs at preclinical stages, such as propentofyline, have shown the importance of further validation using human tissues or cells.²⁹ On the other hand, IMT504 appears as a remarkably superior approach than the gold standard treatment for neuropathic pain models including SNI, and also the clinical standard of care, gabapentin. A single dose of gabapentin (the same paradigm we used for IMT504) exerts acute and effective, although transitory (∼3 h), anti-allodynic actions in rats with a 7 day-long SNI.¹⁴ However, the effects of IMT504 already established SNI remain to be investigated, as it is possible that its efficacy could be less robust due to potential central mechanisms occurring at later stages after nerve injury.²⁴ Nevertheless, our current study sets the foundation to plan clinical safety studies of IMT504 in human subjects, and to test its potential as a novel therapeutic for neuropathic pain with a unique mechanism of action.