Brain-Derived Neurotrophic Factor Administration Mediated Oligodendrocyte Differentiation and Myelin Formation in Subcortical Ischemic Stroke
Background and Purpose—Translational research is beginning to reveal the importance of trophic factors as a therapy for cellular brain repair. The purpose of this study was to analyze whether brain-derived neurotrophic factor (BDNF) administration could mediate oligodendrogenesis and remyelination after white matter injury in subcortical stroke.
Methods—Ischemia was induced in rats by injection of endothelin-1. At 24 hours, 0.4 μg/kg of BDNF or saline was intravenously administered to the treatment and control groups, respectively. Functional evaluation, MRI, and fiber tract integrity on tractography images were analyzed. Proliferation (KI-67) and white matter repair markers (A2B5, 2',3'-cyclic-nucleotide 3'-phosphodiesterase [CNPase], adenomatous polyposis coli [APC], platelet-derived growth factor receptor alpha [PDGFR-α], oligodendrocyte marker O4 [O4], oligodendrocyte transcription factor [Olig-2], and myelin basic protein [MBP]) were analyzed at 7 and 28 days.
Results—The BDNF-treated animals showed less functional deficit at 28 days after treatment than the controls (P<0.05). Although T2-MRI did not show differences in lesion size at 7 and 28 days between groups, diffusion tensor imaging tractography analysis revealed significantly better tract connectivity at 28 days in the BDNF group than in the controls (P<0.05). Increased proliferation of oligodendrocyte progenitors was observed in treated animals at 7 days (P<0.05). Finally, the levels of white matter repair markers (A2B5, CNPase, and O4 at 7 days; Olig-2 and MBP at 28 days) were higher in the BDNF group than in the controls (P<0.05).
Conclusions—BDNF administration exerted better functional outcome, oligodendrogenesis, remyelination, and fiber connectivity than controls in rats subjected to subcortical damage in ischemic stroke.
After decades of research focused on the search for a treatment for cortical infarcts in experimental models in which the gray matter is most affected, a few translational studies are beginning to highlight the importance of considering the white matter component after stroke.1 Not only are ≤25% of ischemic strokes in humans subcortical or lacunar and confined to white matter regions2 but also cortical infarcts produce white matter injury. The high frequency of this damage motivates the search for an effective therapy to enhance the mechanisms underlying the repair of damaged white matter (axons and myelin) after a stroke.2
Trophic factors are emerging as a viable repair therapy in stroke, and they can strongly promote a favorable environment for cellular repair after brain injury.3,4 One of the prominent trophic molecules is brain-derived neurotrophic factor (BDNF), which is secreted in an activity-dependent manner and crucially promotes synaptic regulation and axonal plasticity associated with learning, memory, and sensorimotor recovery.4,5 Furthermore, in vitro and BDNF knockout studies have demonstrated that this trophic factor has direct effects on oligodendroglia, promoting the proliferation and differentiation of oligodendrocyte precursor cells (OPC) and myelination.6–8
All of the above indicate that BDNF could exert a possible effect on oligodendrogenesis and remyelination after a stroke. Thus, the present study explored the possible effect of BDNF administration on white matter remodeling via oligodendrogenesis and myelinogenesis and whether this effect might correlate with functional recovery in an animal model of subcortical stroke.
Materials and Methods
This translational study followed all stroke therapy academic industry roundtable and RIGOR guidelines in terms of randomization, blinding and statistical power.9 In addition, the experiments were designed to minimize animal suffering in compliance with our medical school’s Ethical Committee for the Care and Use of Animals in Research (EU directives 86/609/CEE and 2003/65/CE). A total of 74 adult Sprague–Dawley rats (200–250 g) were randomly distributed into 3 groups: sham: surgery without endothelin-1 injection+intravenous saline administration (n=24); control: subcortical stroke+intravenous saline administration (n=24); and BDNF animals: subcortical stroke+intravenous BDNF treatment (n=24). In the BDNF group, recombinant human BDNF (Peprotech, United Kingdom) was diluted in saline to a final volume of 1 mL and administered through the tail vein at a final dose of 0.4 μg/kg at 24 hours after surgery. A saline solution of the same volume was delivered to the control and sham animals. The rats were then randomly divided into 3 subgroups that were euthanized at 4 hours (n=4 per each group), at 7 days (n=10 per each group) or at 28 days (n=10 per each group). Two rats were excluded from the study because one died after surgery and the other died during the magnetic resonance procedure.
Endothelin-1 Subcortical Stroke Model
In all the animals, physiological parameters and body temperature were continuously monitored during surgery (online-only Data Supplement). To provoke white matter injury, a subcortical stroke was induced to preanesthetized rats after a small craniotomy by the injection of a potent vasoconstrictor, endothelin-1, using a stereotactic apparatus using stereotactic references (+0.4 mm anteroposterior, +3.5 mm lateral, +6 mm dorsoventral from the bregma). One microliter of endothelin-1 (0.25 μg/μL) was delivered at a final speed of 0.2 μL/m. Immediately after surgery, analgesia was provided to all groups by an intraperitoneal injection of meloxicam at 2 mg/kg.
BDNF Quantification After Treatment
BDNF was quantified by the human BDNF ELISA kit (ABCAM) and Western blot in brain tissue and serum in both the control and the treated animals at 4 hours, 7 days, and 28 days (online-only Data Supplement). Immunofluorescence (using a primary antibody anti-BDNF [1:1000, Millipore] and a secondary antibody antirabbit Alexa Fluor 594 [1:750, Invitrogen]; n=4 per group).
Functional Evaluation Scales
Functional evaluation was scored in all animals by a blinded observer before surgery and at 1, 3, 7, and 28 days after treatment. Motor performance evaluation was measured by beam walking test,10 the rotarod test,11 and the Modified Rogers Scale12 (online-only Data Supplement).
In Vivo Analysis by MRI and Diffusion Tensor Imaging Tractography
To analyze whether the BDNF effect could be examined with in vivo imaging techniques, we studied the ipsilateral hemisphere at the site of the endothelin-1–induced lesion using T2-MRI, diffusion on apparent diffusion coefficient (ADC) maps, and tract connectivity using diffusion tensor imaging (DTI) tractography (n=6 of each group) at 7 and 28 days after treatment (online-only Data Supplement).
Cell Proliferation Analysis
KI-67 staining was analyzed on days 7 and 28, using 10 sections selected from each animal (n=6 animals per group; online-only Data Supplement).
To evaluate differentiation markers in proliferating cells, we double-stained KI-67–positive cells with 2',3'-cyclic-nucleotide 3'-phosphodiesterase (CNPase), A2B5, adenomatous polyposis coli (APC), and platelet-derived growth factor receptor alpha (PDGFR-α) by immunofluorescence. All the images were acquired as a maximum confocal projection.
Immunohistochemistry, Immunofluorescence, and Western Blot
CryoMyelin staining kit (Hito Biotech) that allows sensitive localization and visualization of the myelin fibers was performed on frozen sections. The mean region of interest intensity in the CryoMyelin staining was quantified using a Nikon Eclipse-Ti inverted microscope and NIS-elements software. The lesion zone was studied in more detail using immunofluorescence and Western blot as previously described (n=4 animals of each group).11 CNPase, A2B5, oligodendrocyte marker O4 (O4), Nogo-A, myelin basic protein (MBP), and oligodendrocyte transcription factor (Olig-2) markers were studied at 7 and 28 days (online-only Data Supplement).
The data are presented as mean±SEM. The Kruskal–Wallis test followed by the Mann–Whitney U test was used to compare data. Values of P<0.05 were considered significant at a 95% confidence interval; the data were calculated using statistical software programs SPSS 16 and GraphPad.
BDNF Levels Were Increased After Treatment
ELISA analysis of serum BDNF showed that the levels were augmented at 4 hours after its intravenous administration (4.33±0.32 and 5.61±0.62; P<0.05; Figure 1A). BDNF was analyzed in the BDNF-treated group and in the control animals throughout the brain. At 4 hours, significant higher levels of BDNF were found in the area of injury in BDNF animals (6.67±0.43) compared with the controls (0.79±0.24; P<0.05). Also Western blot analysis detected higher levels of BDNF in the ischemic lesion in the treated animals (0.97±0.08) compared with the control group (0.79±0.24) at 4 hours after treatment (P<0.05). At 7 and 28 days after administration, the BDNF levels were not significantly different in the treated animals (0.28±0.03 and 0.20±0.08) compared with the control group (0.27±0.063 and 0.204±0.049), respectively (Figure 1B and 1C). Immunofluorescence for BDNF showed both intracellular and extracellular staining (Figure 1C).
BDNF Improved Functional Recovery in the Subcortical White Matter Injury Model
No significant differences were found in the functional outcome of the treated and control animals at 1, 3, and 7 days after treatment. However, compared with the control rats, 28 days after treatment the BDNF-treated rats showed significantly better rotarod test performances (67.25±25.10 and 103.00±16.10, respectively; P=0.049), beam walking (3.5±0.70 and 1.4±0.54, respectively; P=0.042), and Modified Rogers Test (2.5±0.57 and 1.2±0.44, respectively; P=0.018; Figure 2A).
BDNF Effects on White Matter Were Negligible by In Vivo MRI but Perceptible by Tractography and Myelin Staining
The lesion size viewed on MRI in the BDNF-treated animals was indistinguishable from the control group at 7 days (7.30±2.56 and 14.75±1.72, respectively) and 28 days (7.49±1.79 and 8.77±3.97, respectively) after treatment (P>0.05; Figure 2C).
DTI tractography data showed similar results in axial diffusivity (124.297±10.72 and 133.73±14.497; P>0.05) and diffusion in ADC maps (93.58±10.033 and 123.307±23.839; P>0.05) in both the control and the treated groups, respectively, at 7 days after stroke. However, compared with the control rats, 28 days after treatment the BDNF-treated rats showed significantly improved axial diffusivity (126.39±7.82 and 159.715±14.761, respectively; P<0.05) and diffusion in ADC maps (129.743±11.934 and 157.926±12.562, respectively; P<0.05; Figure 2C). This result suggests that there was a significant improvement in white matter thickness (width, breadth, and depth) and restoration of tract connectivity in the BDNF-treated animals compared with controls at 28 days.
These results are in agreement with the morphological study of myelin fibers by CryoMyelin staining. The mean region of interest intensity was calculated at the lesion site in the control and BDNF-treated animals, in which white intensity indicated absence of myelin and black intensity the presence of myelinated axons. The results showed higher intensity (absence of axons) in controls (198.61±26.30) compared with BDNF-treated animals (172.18±17.40; Figure 2B).
BDNF Administration Enhances OPC Proliferation After White Matter Injury
Numerous KI-67–positive cells were observed in the ischemic lesion in the control animals and in the BDNF-treated group. The number of KI-67–positive cells (170±17.32) was significantly higher in the ischemic lesion compared with control group (34.25±18.28; P=0.032) 7 days after BDNF treatment (Figure 3A).
Double staining of KI-67–positive cells with the OPC markers was observed to be higher in the subventricular zone in treated animals than in controls for A2B5 (9.12±1.87% and 5.45±2.10%), CNPase (31.00±7.32% and 22.12±5.34%), APC (26.00±5.12% and 17.32±4.21%) and PDGFR-α (36.00±6.23% and 28.00±7.34%), suggesting that OPC proliferation is enhanced after BDNF injection (Figure 3B).
BDNF Injection Increases OPC Markers 7 Days After Axonal Disruption
The levels of OPC markers were analyzed by Western blot in the lesion area at 7 and 28 days after treatment (Figure 4B). There was a significant increase in CNPase marker levels in the BDNF-treated animals (217.00±14.10) compared with the controls at 7 days (183.20±16.32; P<0.05). A2B5 was also higher in the BDNF-treated animals (141.20±41.20) than in the controls (75.39±13.20) at 7 days after treatment (P<0.05). We also found higher levels O4 marker at 7 days, in the BDNF-treated group (190.88±20.19) than in the control animals (82.10±18.01; P<0.05). At 28 days, the levels of OPC markers were lower in both control and BDNF animals. Also, when comparing treated and control groups at 28 days, no significant differences were found (A2B5 [91.12±41.10 versus 112.10±13.81], CNPase [89.23±21.10 versus 110.10±21.30], and O4 [75.20±19.10 versus 69.60±16.13], respectively). Immunofluorescence analysis confirmed these results (Figure 4A).
BDNF Administration Enhances Oligodendrocyte Maturation and Axonal Growth-Associated Markers 28 Days After Injury
Western blot analysis showed significantly higher levels of the oligodendrocyte marker Olig-2 in the BDNF-treated animals than in the control group at 28 days (242.85±15.50 and 94.21±10.20, respectively; P<0.05) and a significant increase in MBP in the treated animals compared with the controls (192.18±29.99 and 106.55±26.74, respectively; P<0.0.05). We also found a significant decrease in Nogo-A levels in the BDNF-treated animals compared with the controls at 7 days (160.11±17.23 and 213.21±12.41, respectively) and at 28 days (163.01±18.88 and 200.10±9.87, respectively; P=0.014). At 7 days, the levels of Olig-2 (150.25±19.43 and 70.91±21.50) and MBP (51.21±23.10 and 25.01±10.80) were too low compared with 28 days in both the treated group and the control group because all these markers are related to later developmental steps in white matter differentiation.
In the current search for new therapeutic strategies to improve functional and cognitive deficits after stroke, it is worth remembering that myelination failure prevention is necessary for brain repair processes. Thus, the present study used an intravenous infusion (0.4 μg/kg) of BDNF as a therapeutic strategy to prevent myelination failure. The group treated with BDNF injection showed improved functional recovery and a significant increase in the number of proliferating cells, including OPC, after white matter injury. After using BDNF, large numbers of cells expressed OPC markers, such as CNPase, A2B5, and O4 at 7 days. At 28 days after treatment, the cells began to acquire specific markers of oligodendrocyte differentiation, such as Olig-2 and MBP, suggesting that repair of white matter fiber tracts was induced by the BDNF injection. The results support a role for BDNF in improving the repair of white matter and in OPC proliferation and differentiation in this experimental subcortical stroke model.
This treatment was chosen because systemic administration is already known to allow BDNF to cross the blood brain barrier.4,13 In this sense, the BDNF levels in our study were augmented in brain tissue and peripheral serum at 4 hours after injection. This increase of BDNF levels observed in the brain might enhance recovery mechanisms after stroke.
BDNF as therapy to induce brain protection and brain repair is becoming more common in translational research. After a cortical stroke, BDNF has been shown to control neuronal circuits, increase the number of newborn neurons in several brain areas,3,4 reduce astrogliosis, enhance axonal growth in the ischemic border zone,5 and stimulate the plasticity of dendritic branching and synaptic transmission.14 However, the effects of BDNF administration on nerve fiber repair, myelin formation, and remodeling after white matter injury are still unknown.
In a translational study, it is important to analyze whether BDNF injections act on the motor dysfunctions that are characteristic to subcortical stroke. Previous authors found that the beam walking test,10 the rotarod test,11 and the Modified Rogers Test12 to be effective in assessing the motor deficit associated primarily with subcortical stroke. In this study, BDNF treatment induced a significant improvement in functional recovery that was particularly notable at 28 days after treatment when compared with the controls. These results clearly suggest a true recovery-enhancing effect for BDNF. Although there are no previous studies that have administered BDNF in a subcortical stroke model, our results are consistent with previous data showing a better functional outcome in animals treated intravenously with BDNF after cortical ischemia.4
Focal injection of the vasoconstrictor endothelin-1 into the subcortical white matter produces a visible infarct on MRI, as previously shown in subcortical stroke studies.15,16 To analyze whether the BDNF effect could be visualized with in vivo imaging techniques, we performed T2-MRI, ADC maps, and DTI tractography. No significant differences in lesion size were observed in the BDNF-treated animals compared with the control group on the T2-MRI images; however, analyzing fiber tract integrity by DTI tractography showed that tract thickness was recovered at 28 days after BDNF administration. This result suggests that functional recovery at 28 days could be related to the process by which restructured axons, which had previously been compromised and demyelinated, recover not only their proper structure but also tract connectivity. The lack of a relationship between the T2-MRI and DTI tractography images, however, remains unclear. This is why we intensified the histological analysis of the injured tissue in this study.
After a stroke, myelin injury activates the OPC distributed throughout the white matter, which proliferate and migrate to the site of damage where they subsequently mature into myelinating oligodendrocytes that ensheath axons to form the myelin membrane.1,17 Our study using colocalization of the KI-67 antigen and the CNPase, A2B5, APC, and PDGFR-α markers showed BDNF to be a potent factor for enhancing the OPC proliferation response after white matter injury. Our findings agree with other nonstroke studies in vitro and in BDNF knockout mice, showing that BDNF exerts direct effects on oligodendroglia, promoting OPC proliferation and differentiation, as well as myelination.6–8 To elucidate whether BDNF administration also acts on oligodendrocytes and shattered white matter fibers and increases some repair mechanisms, including nerve fiber remodeling, axonal sprouting, oligodendrogenesis and myelinogenesis, we measured white matter repair-associated markers in both the treated and the control animals. Various white matter repair markers were studied at 2 different time points, 7 and 28 days. Related to the first steps of the genesis and migration of white matter progenitor cells, markers such as A2B5 (a characteristic OPC marker), O4, and CNPase (markers related to immature oligodendrocytes) were studied in the lesion area. Markers related to white matter differentiation and myelin fiber maturation such as MBP (myelin marker), Olig-2 (mature oligodendrocytes), and the myelin inhibitor Nogo-A were also analyzed.
The best-characterized oligodendroglial progenitor marker is A2B5.18,19 This molecule is a cell surface ganglioside expressed on developing oligodendroglial progenitors. In our study, the levels of the A2B5 marker were higher in the BDNF-treated animals at 7 days than in the control group. O4 is another marker expressed in pro-oligodendrocytes during oligodendrocyte differentiation.20 Our study found an increase in the amount of O4 protein in the animals subjected to BDNF administration compared with their controls. This observation agrees with a study describing the levels of CNPase.19 The presence of CNPase seems to be one of the earliest events in oligodendrocyte differentiation, and BDNF-treated animals showed higher levels of this marker than the control group. The levels of A2B5, CNPase, and O4 were too low at 28 days because all these markers are related to early events in oligodendrocyte differentiation. These results are in line with previous nonstroke in vitro studies in which the effects of BDNF are examined in oligodendrocyte progenitors, finding enhanced A2B5 and O4 expression after neurotrophin administration.21
About the oligodendrocyte maturation-associated proteins, the levels of these markers were found to be too low at 7 days. However, Olig-2 levels were elevated in the treated animals compared with the control animals at 28 days, an observation that could be explained by BDNF having enhanced oligodendroglial cell differentiation. However, both the presence of mature oligodendrocytes and coating the myelin sheath are important for white matter repair. There was a significant increase of MBP reactivity in the white matter tracts at 28 days in the treated animals when compared with the control group. Although there are no previous studies that relate BDNF to remyelination after stroke, these increases in the levels of MBP and Olig-2 concur well with previous nonstroke studies, indicating that BDNF influences differentiating oligodendrocytes by increasing both the number of MBP cells and the expression of the MBP protein.22 What is unknown is whether the increased myelination was because more axons were being myelinated or whether there was an overall increase in myelin thickness.
Among the myelin-associated proteins, Nogo-A has been shown to be particularly powerful in preventing axonal growth and plasticity.23 Nogo-A not only inhibits axonal growth but also prevents neurotrophins such as BDNF from binding to this receptor inhibiting axonal growth.24 Some studies have demonstrated that inhibition mediated by Nogo-A is blocked if neurons are exposed to BDNF before encountering the inhibitor.25 The present study concurs with these observations because at 28 days after treatment, there was a significant decrease in the levels of Nogo-A in BDNF animals. These results indicate that BDNF injection could enhance axonal growth and plasticity by decreasing Nogo-A levels. The observations found at 28 days suggest that functional recovery might be related to the axonal sprouting subsequent to BDNF administration and could be indicative of the process by which growing and as-yet demyelinated axons are recovered with new myelin sheaths. Furthermore, this interpretation of these observations agrees with the tractography images that suggest that tract connectivity was being restored in the infarcted area at the same time point.
Our study helped us to identify a clear role for BDNF in improving functional outcome by mediating axonal growth, OPC proliferation, oligodendrocyte differentiation, remyelination, and fiber tract connectivity restoration in an experimental animal model of white matter injury.
We greatly appreciate the support of Inés Barahona and Esperanza Medina and we thank ServingMed.com for linguistic assistance.
Sources of Funding
This study was supported by research grants PS12/01754 (P.I.: Dr Díez-Tejedor), INVICTUS Spanish Neurovascular Network RD12/0014/0006 (Dr Fuentes and J. Ramos-Cejudo) and Sara Borrell postdoctoral fellowship CD12/00706 (Dr Otero-Ortega) from the Research Institute Carlos III, Ministry of Science and Innovation of Spain.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.006692/-/DC1.
- Received July 7, 2014.
- Revision received September 25, 2014.
- Accepted October 16, 2014.
- © 2014 American Heart Association, Inc.
- Matute C,
- Domercq M,
- Pérez-Samartín A,
- Ransom BR
- Schäbitz WR,
- Steigleder T,
- Cooper-Kuhn CM,
- Schwab S,
- Sommer C,
- Schneider A,
- et al
- Schäbitz WR,
- Berger C,
- Kollmar R,
- Seitz M,
- Tanay E,
- Kiessling M,
- et al
- Fisher M,
- Feuerstein G,
- Howells DW,
- Hurn PD,
- Kent TA,
- Savitz SI,
- et al
- Britton GL,
- Kim H,
- Kee PH,
- Aronowski J,
- Holland CK,
- McPherson DD,
- et al
- Gutiérrez-Fernández M,
- Rodríguez-Frutos B,
- Ramos-Cejudo J,
- Teresa Vallejo-Cremades M,
- Fuentes B,
- Cerdán S,
- et al
- Rogers DC,
- Campbell CA,
- Stretton JL,
- Mackay KB
- Goldberg MP,
- Ransom BR
- Blasi F,
- Wei Y,
- Balkaya M,
- Tikka S,
- Mandeville JB,
- Waeber C,
- et al
- Baumann N,
- Pham-Dinh D
- Back SA,
- Riddle A,
- McClure MM
- Cheatwood JL,
- Emerick AJ,
- Schwab ME,
- Kartje GL