Epidural Spinal Cord Stimulation Promotes Motor Functional Recovery by Enhancing Oligodendrocyte Survival and Differentiation and by Protecting Myelin after Spinal Cord Injury in Rats

Abstract

Epidural spinal cord stimulation (ESCS) markedly improves motor and sensory function after spinal cord injury (SCI), but the underlying mechanisms are unclear. Here, we investigated whether ESCS affects oligodendrocyte differentiation and its cellular and molecular mechanisms in rats with SCI. ESCS improved hindlimb motor function at 7 days, 14 days, 21 days, and 28 days after SCI. ESCS also significantly increased the myelinated area at 28 days, and reduced the number of apoptotic cells in the spinal white matter at 7 days. SCI decreased the expression of 2′,3′-cyclic-nucleotide 3′-phosphodiesterase (CNPase, an oligodendrocyte marker) at 7 days and that of myelin basic protein at 28 days. ESCS significantly upregulated these markers and increased the percentage of Sox2/CNPase/DAPI-positive cells (newly differentiated oligodendrocytes) at 7 days. Recombinant human bone morphogenetic protein 4 (rhBMP4) markedly downregulated these factors after ESCS. Furthermore, ESCS significantly decreased BMP4 and p-Smad1/5/9 expression after SCI, and rhBMP4 reduced this effect of ESCS. These findings indicate that ESCS enhances the survival and differentiation of oligodendrocytes, protects myelin, and promotes motor functional recovery by inhibiting the BMP4-Smad1/5/9 signaling pathway after SCI.

Introduction

Spinal cord injury (SCI) leads to motor, sensory, and autonomic dysfunction below the site of injury [1]. In the USA, the recent estimated number of persons with SCI was ~291,000, with ~17,730 new cases each year (www.uab.edu/nscisc). Although there have been numerous studies on SCI, effective treatment is still lacking [2]. Accumulating evidence indicates that epidural spinal cord stimulation (ESCS) not only effectively relieves neuropathic pain [3, 4], but also significantly improves motor function [5–9], and even restores cough after SCI [10]. As a result, ESCS is becoming a leading candidate for the treatment of SCI [11]. Apart from activating residual neuronal pathways [12, 13], ESCS also promotes anatomical plasticity of the nervous system after SCI [7, 12–14]. However, it is unclear whether ESCS promotes oligodendrocyte differentiation or reduces myelin loss after SCI.

Oligodendrocytes wrap axons with myelin sheaths, provide trophic support, and protect neurons and their axons [15]. Oligodendrocyte loss and axonal demyelination are major pathological events hindering functional recovery after SCI [16]. Increased numbers of mature oligodendrocytes can significantly improve motor functional recovery [17]. Although inhibition of oligodendrocyte apoptosis after SCI reduces axonal demyelination and promotes neurological recovery, stimulating oligodendrocyte differentiation is also a promising approach for replacing damaged oligodendrocytes and enhancing remyelination [18–23]. Studies have shown that spontaneous oligodendrocyte replacement and remyelination occur naturally from both endogenous neural stem cells (eNSCs) and oligodendrocyte progenitor cells (eOPCs) in several animal models of SCI [16, 18, 19, 24–28]. However, this self-repair response is not sufficient to compensate for the oligodendrocyte loss and demyelination caused by SCI, ultimately leading to spinal cord dysfunction [16, 29]. Therefore, exploring novel therapeutic strategies for promoting oligodendrocyte differentiation and remyelination is key to enhancing functional recovery after SCI [30, 31].

ESCS is an important neuromodulatory approach, whereby an electrode is implanted into the spinal epidural space and directly stimulates the spinal cord tissue with a suitable current to exert a therapeutic effect [32]. Many studies have suggested that other electrical stimulation methods also enhance neurological functional recovery by promoting eNSC proliferation and their differentiation into oligodendrocytes, thereby stimulating axonal remyelination after SCI [33–36]. For example, Becker et al. [36] reported that electrical stimulation promotes the proliferation of endogenous neural progenitor cells and their differentiation into oligodendrocytes after SCI in adult rats. Geng et al. [34] reported that electroacupuncture promotes the proliferation of eNSCs and their differentiation into oligodendrocytes by inhibiting the activation of the Notch signaling pathway after SCI in rats. During CNS development, bone morphogenetic proteins (BMPs) repress oligodendrogenesis while enhancing the development of astrocytes [37]. BMP4 suppresses the differentiation of adult eOPCs into oligodendrocytes, and inhibition of BMP4 signaling promotes remyelination and functional recovery from CNS demyelinating disease [38]. Our previous studies showed that ESCS promotes eNSC proliferation (unpublished), but the effects of ESCS on oligodendrocyte differentiation and myelination after SCI are still unclear. Therefore, in the present study, we investigated whether ESCS affects oligodendrocyte differentiation in the spinal cord white matter after SCI, and explored the underlying mechanisms.

Materials and Methods

Animals, Chemicals, Electrodes, and External Electrical Stimulator

All the adult female Sprague-Dawley rats (250 g–300 g) were provided by the Experimental Animals Center of Jinzhou Medical University. All the animal experiments were conducted strictly following the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care Ethics Commission of Shanghai Tenth People’s Hospital. Recombinant human bone morphogenetic protein 4 (rhBMP4, HY-P7007, MedChemexpress, Monmouth Junction, NJ) was dissolved in dimethyl sulfoxide (DMSO). The stimulating electrodes were independently developed by the Control Science and Engineering Department of Huazhong University of Science and Technology (Fig. 1A). The external electrical stimulator (Isostim A320) was from World Precision Instruments Inc. (Sarasota, FL).

Fig. 1.

Electrode and direct-current square-wave pulses. A Electrode circuit diagram (A1) and front (A2) and back (A3) photographs. The electrode is made by a flexible circuit board technique and implanted into the dorsal epidural space. The electrode material is polyimine, which has good biocompatibility. Three circular gold contacts 1 mm in diameter are separated by 7 mm gaps, while eight small holes at the center and two at the proximal end are used to fix the electrode to the paravertebral ligament and muscles [47]. We set the middle gold contact as the anode and the proximal gold contact as the cathode. B Traces of direct-current square-wave pulses in a simulated rat environment. C Parameters of ESCS (frequency 50 Hz, pulse width 200 µS, amplitude 90% MT: 0.045 mA). MT, motor threshold.

Experimental Groups and Drug Administration

Sprague-Dawley rats were randomly divided into four groups: Sham (T10 and T13 laminectomy only), SCI (T10 contusive injury and T13 laminectomy), SCI+ESCS (abbreviated to ESCS), and SCI+ESCS+rhBMP4 (ESCS+BM; 2.5 μL, 1 μg/μL rhBMP4 [39]) groups. Rats in the sham and SCI groups were not given electrical stimulation. Rats in the ESCS and ESCS+rhBMP4 groups were given monophasic direct current pulse stimulation immediately after the incision was sutured [14], and this treatment lasted for 30 min per day over the course of 1 week [7, 40]. The rats in the ESCS+rhBMP4 group were given rhBMP4 immediately after SCI, while the rats in the Sham, SCI, and ESCS groups received an equivalent volume of DMSO. A microsyringe (F519159, Sangon Biotech, Shanghai, China) was used to inject 2.5 μL rhBMP4 at 0.5 μL/min into the spinal cord [41]. Each group was randomly divided into three subgroups for subsequent experiments: (A) assessment of hindlimb motor function for 7 days and 28 days [42, 43] (n = 6); (B) semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) and western blot (n = 6 each); (C) immunofluorescent labeling, fluorescence TUNEL staining, and Luxol fast blue (LFB) (n = 6 each) histological examination when the hindlimb motor function test was finished at 28 days after SCI. Myelin basic protein (MBP) and LFB staining were assessed at 28 days, while other indicators were measured at 7 days after SCI.

SCI Model Establishment, Epidural Electrode Implantation, and Stimulation Parameters

Adult rats were anesthetized with pentobarbital sodium (40 mg/kg, i.p.). The incomplete contusive SCI model was established using a self-made Allen’s impact device and postoperative care was as described previously [44]. In brief, after anesthesia and sterilization, the skin and muscle overlying the spinal column were incised and a laminectomy was performed at T10 and T13, leaving the dura intact. A firing pin (20 g in weight and 2 mm in diameter) was then dropped onto the exposed T10 region through a glass casing (2.2 mm inner diameter) from a height of 25.0 mm. After SCI, the stimulating electrode was implanted into the epidural space from the T13 segment with the distal end of the electrode oriented rostrally and the anode center facing the epicenter of the lesion of T10, and the positioning holes on the electrode were fixed to the paravertebral ligament and muscles. The cathode and anode wires were tunneled to the animal’s neck, where plugs were attached firmly to the skin. As the motor threshold (MT) could not be measured after SCI due to the severity of injury and possible spinal shock, we measured it in rats after sham operation. These rats were given direct-current square-wave pulse stimulation (frequency 2 Hz, pulse width 200 µs) when they were awake; the current intensity was gradually increased until the first occurrence of symmetrical contractions of the lower trunk and/or hindlimbs was seen or palpated; the current intensity at that point was the MT [45]. The MT was 0.05 ± 0.009 mA (n = 6), so the stimulus parameters were set at a frequency of 50 Hz, pulse width 200 µS, and intensity 0.045 mA [7, 45, 46] (Fig. 1B, C).

Basso, Beattie, and Bresnahan (BBB) Score Assessment

The BBB score has 22 grades ranging from 0 to 21, and is used to assess the recovery of hindlimb motor function after SCI in rats [48]. A score of 0 indicates complete paralysis, and the score of 21 indicates that motor function is completely normal. The BBB scores were collected and analyzed by two evaluators who were blinded to the experimental groups at 1 day, 7 days, 14 days, 21 days, and 28 days after SCI.

Semi-quantitative RT-PCR

At 7 days and 28 days after SCI, 6 mm of spinal cord (3 mm at the rostral and caudal ends of the epicenter) was removed and equally divided into two parts centered the epicenter. One part was used to extract total RNA with TRIzol reagent (Invitrogen, Waltham, MA), and the other part was used for extracting total protein. The purity of the RNA was evaluated by the A260/A280 ratio (normal range 1.7–2.1) using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA). RT-PCR was performed following the instructions in the PrimeScript™ One Step RT-PCR Kit Ver.2 (Takara, Beijing, China). A 50-µL reaction system containing 0.5 μg RNA was placed in an S1000 PCR instrument (Bio-Rad, Hercules, CA) with the reaction conditions described in the operating instructions. The expression of 2′,3′-cyclic-nucleotide 3′-phosphodiesterase (CNPase) and MBP was assessed using the following primers sets (Beijing Dingguochangsheng Biotechnology Co., Ltd, Beijing, China): CNPase; F–GAG ACA TAG TGC CCG CAA AG, R–ATC TTG GTG CCG TTG TGG TA (287 bp); MBP; F–GAC GAG CTT CAG ACC ATC CA, R–CCA TAG TTC CTC TAC GCC TCG (263 bp); β-actin F–CTC TGT GTG GAT TGG TGG CT, R–AGC TCA GTA ACA GTC CGC CT (136 bp). Products of RT-PCR and a 50-bp DNA ladder marker (MD1001, Simgen, China) were separated by 2% agarose gel electrophoresis and the imaging results were recorded with ethidium bromide staining. NIH ImageJ software was used to determine the relative band intensity.

Western Blot Analysis

The protein extract was transferred to a 0.22- or 0.45- μm polyvinylidene fluoride (PVDF) membrane after separation on 12% SDS-polyacrylamide gel. The PVDF membranes were blocked with 0.1% bovine serum albumin at room temperature for 2 h, and then incubated with the primary antibody on a shaking table at 4°C overnight. The primary antibodies were as follows: anti-CNPase (marker of oligodendrocytes, ab6319, 1:250; Abcam, Cambridge, UK), anti-MBP (78896S, 1:1000; CST, Danvers, MA), anti-BMP4 (ab39973, 1:1000; Abcam, Cambridge, UK), anti-phospho-Smad1/5/9 (#13820, 1:1000; CST, Danvers, MA), anti-Smad1/5/9 (ab66737, 1:1500; Abcam, Cambridge, UK), and β-actin (sc-47778, 1:1000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). After fully cleaning the PVDF membrane, the corresponding secondary antibodies [goat anti-rabbit and goat anti-mouse IgG-HRP (sc-2004 or sc-2005; all 1:2000; Santa Cruz Biotechnology, Inc.)] were incubated for 2 h at room temperature. The PVDF membranes were illuminated with electrochemiluminescence and photographed with a gel imaging system. ImageJ was used to measure the gray values of the target proteins and β-actin bands, and target proteins were normalized to β-actin to reflect their relative expression levels.

Immunofluorescent Single and Double Labeling

The rats were sacrificed and the injured spinal cords were fixed in 4% paraformaldehyde at 7 days and 28 days after SCI. Then 6 mm of spinal cord (3 mm at the rostral and caudal ends of the epicenter) were removed and used to cut coronal and sagittal frozen sections. The coronal sections were cut after trimming 1 mm from the epicenter of the lesion. Sagittal sections with gray and white matter on both sides were cut parallel to the posterior median sulcus. Serial 10-μm frozen sections were used for histological examination. Sections were permeabilized with 0.25% Triton X-100 for 20 min, and then incubated with 5% normal goat serum (abs933, Absin, Shanghai, China) for 1 h at room temperature to block non-specific binding sites. The sections were incubated with the primary antibodies (diluted in 5% normal goat serum) anti-CNPase (ab6319, 1:50; Abcam, Cambridge, UK), anti-Sox2 (marker of NSCs, ab97959, 1:400; Abcam, Cambridge, UK), anti-MBP (78896S, 1:200; CST, Danvers, MA), anti-BMP4 (ab39973, 1:200; Abcam, Cambridge, UK), and anti-phospho-Smad1/5/9 (#13820, 1:250; CST, Danvers, MA) in a humidified chamber at 4°C overnight. After fully cleaning the unbound primary antibody from the sections, they were incubated with the secondary antibodies Alexa Fluor 594 goat anti-mouse IgG and/or Alexa Fluor 488 goat anti-rabbit IgG (A-11005 and A-11034, all 1:250; Thermo Fisher Scientific, Waltham, MA) for 2 h at room temperature in the dark. After immunostaining, the sections were incubated with 4′,6-diamidino-2-phenylindole (DAPI, 1 µg/mL) for 2 min in the dark to mark the nuclei. The primary antibody was replaced with PBS buffer in the negative control. A fluorescence microscope (Leica DMI4000B, Wetzlar, Germany) was used to visualize all indicators. Two sections separated by >100 μm for each indicator in one rat were used for immunofluorescent labeling. The counting function in Photoshop CS3 was used to count the numbers of sox2/CNPase double-positive and phospho-Smad1/5/9-positive cells. and ImageJ was used to quantify the mean fluorescence intensity of CNPase, MBP, and BMP4. Mean fluorescence intensity = integrated density / area; percentage immunopositive cells = 100 × number of immunopositive cells / total number of cells (DAPI-stained cells) [44].

Fluorescence TUNEL Staining

Serial 10-μm transverse frozen sections (two sections separated by 100 μm per rat) were prepared for TUNEL labeling which was performed according to the In Situ Cell Death Detection Kit, TMR red (Roche, Basel, Switzerland). After TUNEL staining, all the sections were incubated with 1 µg/mL DAPI for 2 min to counterstain nuclei. Apoptotic cells (red) in the spinal white matter were visualized under a fluorescence microscope and counted using the counting function of Photoshop CS3 [49]. Percentage apoptotic cells = 100 × (red cells / blue cells).

LFB Staining

Serial 10-μm transverse frozen sections (two sections separated by 100 μm per rat) were incubated in 1×PBS for 20 min and then stained according to the instructions with the LFB kit (G3245, Beijing Solarbio Science and Technology Co., Ltd, Beijing, China). Finally, these sections were observed under a light microscope. The proportion of LFB-positive area in the spinal white matter at 28 days after SCI was analyzed with ImageJ [50].

Statistical Analysis

All data are expressed as the mean ± SD, and were analyzed using one-way analysis of variance (ANOVA) with the LSD-t test. SPSS version 19.0 was used for all analyses. P < 0.05 was considered statistically significant.

Results

ESCS Improves Hindlimb Motor Function at 7 Days, 14 Days, 21 Days, and 28 Days after SCI in Rats

BBB scores were lower in the SCI group than in the sham group at 7 days, 14 days, 21 days, and 28 days (P < 0.05). In contrast, compared with the SCI group, the BBB scores were markedly higher in the ESCS group on these days (P < 0.05). In addition, the BBB scores were lower in the ESCS+BM group than in the ESCS group at 7 days, 14 days, and 28 days (P < 0.05). These results indicate that ESCS improves hindlimb motor function at 7 days, 14 days, 21 days, and 28 days following SCI in rats (Fig. 2).

Fig. 2.

Effects of ESCS on hindlimb motor dysfunction at 1 day, 7 days, 14 days, 21 days, and 28 days after SCI in rats (n = 6). ^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS.

ESCS Reduces Myelin Loss in Spinal White Matter 28 Days after SCI

The LFB stain binds myelin to reveal the myelin structure of neural tissue. The relative LFB-stained area can therefore be used to compare myelination among groups [50]. Compared with the Sham group, the relative (proportional) LFB-stained area was reduced in the SCI group at 28 days (P < 0.05). Notably, this area was markedly higher in the ESCS group than in the SCI group (P < 0.05). Moreover, in the ESCS+BM group, the area was lower than in the ESCS group (P < 0.05). These results demonstrate that ESCS reduces myelin loss in the spinal white matter by inhibiting the BMP4-Smad1/5/9 signaling pathway after SCI (Fig. 3).

Fig. 3.

Effects of ESCS on the proportion of LFB-positive area in the spinal white matter at 28 days after SCI (n = 6). A Representative images of LFB staining (scale bar, 100 μm). B Quantification of the proportion of LFB-positive area (^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS).

ESCS Inhibits Apoptosis in Spinal White Matter 7 Days after SCI

TUNEL staining showed that, compared with the sham group, the percentage of apoptotic cells in the white matter was higher in the SCI group (P < 0.05). Furthermore, compared with the SCI group, this percentage was lower in the ESCS group (P < 0.05). There was no significant difference between the ESCS+BM and ESCS groups (P > 0.05) (Fig. 4).

Fig. 4.

ESCS decreased apoptosis in the spinal white matter at 7 days after SCI (n = 6). A Representative images of fluorescent TUNEL staining in the spinal white matter (arrows, apoptotic cells; scale bar, 100 μm). B Quantitative analysis of apoptosis (^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS).

ESCS Up-regulates CNPase Expression 7 Days after SCI in Rats

CNPase is an enzyme expressed by immature and mature oligodendrocytes, and is therefore used as an oligodendrocyte marker [51, 52]. We used RT-PCR and western blotting to evaluate CNPase expression to assess the effects of ESCS on oligodendrocyte survival. The mRNA and protein levels of CNPase were lower in the SCI group than in the sham group (P < 0.05; Fig. 5), indicating oligodendrocyte loss after SCI. Notably, compared with the SCI group, CNPase expression was markedly higher in the ESCS group (P < 0.05), suggesting that ESCS reduces oligodendrocyte loss after SCI. In the ESCS+BM group, CNPase expression was lower than in the ESCS group (P < 0.05), suggesting that ESCS reduces oligodendrocyte loss after SCI by inhibiting the BMP4-Smad1/5/9 signaling pathway.

Fig. 5.

Effects of ESCS on the CNPase expression 7 days after SCI. A Representative images of RT-PCR. B Quantitative analysis of CNPase (n = 6). C Representative images of western blots and quantitative analysis of CNPase (n = 6). ^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS.

ESCS Promotes Oligodendrocyte Differentiation after SCI

Sox2 is a universal marker of eNSCs and endogenous neural progenitor cells [50, 53], and is also expressed by eOPCs [54]. We performed immunofluorescence double labeling for Sox2/CNPase/DAPI (newly-differentiated oligodendrocytes) [51, 53–55] to investigate the effects of ESCS on oligodendrocyte differentiation in the spinal white matter 7 days after SCI. As shown in Fig. 6, the mean fluorescence intensity of CNPase (Fig. 5) was consistent with the RT-PCR and western blot results. Moreover, the percentage of newly-differentiated oligodendrocytes increased after SCI compared with the sham group (P < 0.05). Interestingly, the percentage of newly-differentiated oligodendrocytes increased after ESCS compared with the SCI group (P < 0.05), suggesting that ESCS promotes oligodendrocyte differentiation. rhBMP4 decreased the percentage of newly-differentiated oligodendrocytes compared with the ESCS group (P < 0.05; Fig. 6). These results suggest that ESCS reduces oligodendrocyte loss and promotes oligodendrocyte differentiation by inhibiting the BMP4-Smad1/5/9 signaling pathway after SCI.

Fig. 6.

Effects of ESCS on oligodendrocyte differentiation at 7 days after SCI. A–D Representative images of Sox2/CNPase/DAPI immunofluorescent double labeling (coronal sections in A; sagittal sections in B; arrows, eOPCs; boxes mark the enlarged area; scale bar, 100 μm) and quantitative analysis (mean fluorescence in C and double-labeling in D; n = 6; ^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS).

ESCS Increases MBP Expression in Spinal White Matter 28 Days after SCI

MBP is a structural protein in the plasma membrane of oligodendrocytes that plays a critical role in myelin compaction and thickening in the CNS [56]. Mature oligodendrocytes express myelin proteins, among which MBP is one of the most abundant [51]. We used RT-PCR, western blots, and immunofluorescence staining to evaluate MBP expression after SCI. The MBP mRNA and protein expression levels were markedly lower after SCI compared with the sham group (P < 0.05). In contrast, compared with the SCI group, MBP expression was higher in the ESCS group (P < 0.05). Moreover, in the ESCS+BM group, MBP expression was decreased compared with the ESCS group (P < 0.05), indicating that ESCS protects the myelin in the spinal white matter after SCI by inhibiting the BMP4-Smad1/5/9 signaling pathway (Fig. 7).

Fig. 7.

Effects of ESCS on the expression of MBP in the spinal white matter 28 days after SCI. A, B Representative images of RT-PCR (A) and quantitative analysis of MBP (B, n = 6). C Representative western blots and quantitative analysis of MBP (n = 6). D, E Representative images of immunofluorescent labeling in the spinal white matter (coronal sections in D; sagittal sections in E; scale bars, 100 μm). F Quantification of MBP (n = 6; ^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS).

ESCS Inhibits the BMP4-Smad1/5/9 Signaling Pathway 7 Days after SCI in Rats

The BMP4-Smad1/5/9 signaling pathway regulates the proliferation and differentiation of OPCs and NSCs [38, 57, 58]. To further clarify the mechanisms by which ESCS affects oligodendrocyte differentiation and myelination, we examined the expression of BMP4 and p-Smad1/5/9 after SCI by western blot and immunofluorescence staining. As shown in Fig. 8A–C, compared with the sham group, the expression levels of BMP4 and p-Smad1/5/9 were higher in the SCI group (P < 0.05), suggesting that the BMP4-Smad1/5/9 signaling pathway is activated after SCI. These levels were lower in the ESCS group than in the SCI group (P < 0.05). Moreover, compared with the ESCS group, p-Smad1/5/9 expression was markedly higher in the ESCS+BM group (P < 0.05).

Fig. 8.

Effects of ESCS on the BMP4 and Smad1/5/9 expression and their phosphorylation levels 7 days after SCI. A–C Representative western blots (A) and quantitative analysis of BMP4 (B), and Smad1/5/9 and p-Smad1/5/9 (C) (n = 6). D–G Representative images of immunofluorescent labeling (D, F; scale bars, 100 μm) and quantification of BMP4 (E) and p-Smad1/5/9 (G) (n = 6). ^aP < 0.05 vs Sham, ^bP < 0.05 vs SCI, ^cP < 0.05 vs ESCS.

Immunofluorescence staining for BMP4 and p-Smad1/5/9 (Fig. 8D–G) confirmed the western blot results (Fig. 8A–C). Compared with the sham group, the BMP4 mean fluorescence intensity and percentage of p-Smad1/5/9-positive cells were higher in the SCI group (P < 0.05). Compared with the SCI group, these proteins were downregulated in the ESCS group (P < 0.05). Furthermore, these proteins were upregulated in the ESCS+BM group compared with the ESCS group (P < 0.05; Fig. 8D–G). These findings indicate that ESCS reduces oligodendrocyte loss, promotes oligodendrocyte differentiation, and protects myelin after SCI by inhibiting the BMP4-Smad1/5/9 signaling pathway.

Discussion

SCI is common and devastating; it results in motor, sensory and autonomic dysfunction. ESCS is becoming a promising treatment for SCI [11]. Studies indicate that, apart from the activation of residual neuronal pathways, ESCS also promotes anatomical plasticity of the nervous system after SCI [7, 12–14]. However, the effects of ESCS on oligodendrocyte differentiation and myelination after SCI remain unclear. In this study, we found that ESCS treatment reduced oligodendrocyte and myelin loss, and enhanced oligodendrocyte differentiation in the spinal white matter after SCI by inhibiting the BMP4-Smad1/5/9 signaling pathway.

Oligodendrocyte loss and widespread demyelination [59] contribute to the loss of motor functions after SCI [22, 60, 61]. Electrical stimulation has been shown to promote the proliferation and differentiation of eNSCs and eOPCs, and improve neurological function after CNS injury [34, 36, 55, 62–66]. For example, electroacupuncture promotes functional recovery in the brain by enhancing the proliferation of eNSCs and by stimulating their differentiation into astrocytes, oligodendrocytes, and neurons after cerebral ischemic injury in rodents [65, 66]. Following SCI, electroacupuncture significantly enhances the proliferation of eNSCs and eOPCs, promotes eOPC differentiation, and inhibits oligodendrocyte death. Together, these effects promote remyelination and motor functional recovery in rats [35, 63].

In the present study, ESCS significantly alleviated locomotor dysfunction at 7 days, 14 days, 21 days, and 28 days after SCI, supporting previous research findings [5–9]. ESCS also reduced apoptosis in the spinal white matter 7 days after SCI. Furthermore, the protein and mRNA expression levels of CNPase were significantly decreased, while the percentage of newly-differentiated oligodendrocytes was increased after SCI. Notably, ESCS significantly upregulated CNPase and the percentage of newly-differentiated oligodendrocytes 7 days after SCI. These findings indicate that ESCS reduces oligodendrocyte loss by inhibiting apoptosis and promoting oligodendrocyte differentiation in the spinal white matter after SCI in rats, and support previous studies [22, 34, 36, 55, 62–64]. Although in some studies Sox2 has been used as a marker of eNSCs [50, 53], it is also expressed in eOPCs [54]. Therefore, it is unclear whether the newly-differentiated oligodendrocytes in this study were derived from eOPCs, eNSCs, or both.

Myelin integrity is essential for the control of motor function and for integrating sensory information, and remyelination plays a critical role in the reconnection of neuronal pathways after SCI [67]. Mature oligodendrocytes are the main cells expressing myelin proteins such as MBP, and are derived from eOPCs [51, 68]. Therefore, promoting the differentiation of eOPCs or eNSCs into mature oligodendrocytes may contribute to remyelination [69] and motor functional recovery after SCI [70]. Oscillating field stimulation has been shown to improve motor function by promoting the differentiation of eOPCs and remyelination after SCI in rats [71]. In the current study, MBP mRNA and protein levels in the white matter of the spinal cord were significantly decreased at 28 days after SCI, and ESCS markedly upregulated their levels. ESCS also significantly increased the myelinated area 28 days after SCI, indicating that it significantly reduces the loss of myelin from neuronal axons in the spinal white matter, likely by inhibiting oligodendrocyte loss and promoting their differentiation.

Following SCI, eOPCs that differentiate into oligodendrocytes are few, and some differentiate into astrocytes [72]. Moreover, Uemura et al. [73] reported that upregulation of BMP4 stimulates the differentiation of eOPCs into astrocytes and simultaneously inhibits their proliferation and differentiation into oligodendrocytes, thereby aggravating white matter damage after chronic cerebral hypoperfusion. Notably, studies have shown that inhibiting the activation of the BMP signaling pathway is an effective strategy for promoting oligodendrocyte differentiation and remyelination and for improving motor function after SCI [74, 75]. In the present study, rhBMP4 markedly reduced CNPase expression and the percentage of newly-differentiated oligodendrocytes at 7 days, as well as MBP expression and the myelinated area 28 days after ESCS treatment. Moreover, ESCS significantly inhibited the upregulation of BMP4 and p-Smad1/5/9 after SCI, while rhBMP4 increased the levels of BMP4 and p-Smad1/5/9 after ESCS treatment. These results indicate that ESCS promotes oligodendrocyte differentiation and reduces myelin loss by inhibiting the BMP4-Smad1/5/9 signaling pathway after SCI in rats. However, perplexingly, if inhibition of the BMP4-Smad1/5/9 signaling pathway promotes oligodendrocyte differentiation and reduces myelin loss, why then do activation of the BMP4-Smad1/5/9 signaling pathway and induction of oligodendrocyte differentiation occur simultaneously following SCI? The reason is likely to be complex, but we speculate that multiple factors [76] and signaling pathways [34, 77] are involved in spontaneous oligodendrocyte differentiation after SCI.

In conclusion, ESCS treatment reduces oligodendrocyte and myelin loss and enhances oligodendrocyte differentiation in the spinal white matter by inhibiting the BMP4-Smad1/5/9 signaling pathway after SCI. However, further study is needed to clarify the anti-apoptotic mechanisms of ESCS as well as its effect on neurons. Nevertheless, our study provides important insight into the therapeutic mechanism of action of ESCS and suggests that ESCS in combination with other treatments which promote oligodendrocyte differentiation and enhance remyelination may have better therapeutic efficacy after SCI.

Conflict of interest

The authors declare that they have no conflicts of interest, financial or otherwise.