Background
Spontaneous remyelination enables regeneration of white matter after lesions. This process is possible thanks to the presence, in the adult central nervous system (CNS), of oligodendrocyte precursor cells (OPCs) which have the capacity to replace dead oligodendrocytes and to form new myelin sheaths at the sites of demyelination [
1]. Spontaneous remyelination is linked to the ability of OPCs to proliferate, migrate and differentiate into mature oligodendrocytes in physiological conditions (during development and in adulthood) and in response to injuries.
This process is important in various pathological situations in the CNS, including multiple sclerosis (MS). In relapsing-remitting MS (RR-MS), the most frequent form of the disease, patients experience successive phases of relapses (worsening of symptoms) and remissions (neurological improvement), which are considered to mostly follow a pattern of successive demyelination and remyelination, giving rise to the so-called ‘shadow plaques’. However, the extent of remyelination is very variable among patients, and, in most cases, incomplete [
2,
3]. A second form of MS exists, called primary progressive MS (PP-MS), in which the disease progresses gradually without remission phases. Finally, in a significant subset of RR-MS patients, the course of the disease eventually shifts to a progressive course, termed secondary progressive MS (SP-MS). In these progressive forms of MS (PP-MS and SP-MS), although partial remyelination has been observed [
4], its efficiency is not sufficient for counterbalance the progression of disease. Therefore, there is great interest in understanding the endogenous factors which govern spontaneous remyelination in order to explain why this process fails in the different forms of MS. This better knowledge could help designing new therapeutic strategies aiming at boosting the capacity of remyelination in MS and other demyelinating diseases [
5,
6].
Most remyelinating oligodendrocytes are derived from adult OPCs, which have a widespread distribution throughout the CNS [
7]. Effective remyelination depends in great part on the ability of these OPCs to reach the demyelinated areas. Migration of OPCs depends on the action of chemotactic factors able to attract or repulse OPCs, thus controlling the direction of migration, and chemokinetic factors able to accelerate or slowdown OPCs, thus controlling the speed of migration [
8]. Although embryonic OPCs, responsible for myelination during development, and adult OPCs, responsible for remyelination after lesion, form two distinct populations, developmental myelination and post-lesion remyelination share mechanistic aspects [
9,
10]. For this reason, lessons learned from the description of myelination during development can be useful to better understand the chemotactic and chemokinetic factors involved in remyelination.
Tissue-type plasminogen activator (tPA) is a serine protease long described for its role in fibrinolysis in the circulation. Beyond this property, tPA has been reported to be expressed by many cell types in the brain, including oligodendrocytes [
11,
12] and to exert multiple functions in the healthy and diseased CNS [
13]. Among these functions, tPA has been shown to facilitate neuron migration during development [
14]. tPA has also been shown to influence neuron and OPC survival, in particular by exerting anti-apoptotic effects on these cells via the action of a structural domain homologous to epidermal growth factor (EGF) [
11,
15], a growth factor previously described to increase OPC migration [
16]. However, whether tPA could influence OPC migration was never investigated so far.
In the present study, we hypothesized that tPA could facilitate OPC migration based on the following facts described above: (i) tPA facilitates neuronal migration [
14], (ii) tPA can act on OPCs through its EGF-like domain [
11] and (iii) EGF increases OPC migration [
16]. To study this, we first explored the migration of OPCs in the developing spinal cord and telencephalon (giving raise to corpus callosum, the vastest white matter area in murine brain) by comparing wild type (WT) and tPA knock-out (tPA
−/−) animals. Because we observed a delay in OPC migration during myelination, we studied whether a similar phenotype could appear in adult tPA
−/− mice after focal white matter lesion. Indeed, remyelination was also delayed in these animals, indicating that endogenous tPA could facilitate OPC migration. Finally, we show here that tPA exerts a chemokinetic effect on OPCs in vitro, through its domain of homology with EGF.
Methods
Ethical statement
Experiments were performed in our laboratory (agreement number D14118001) in accordance with French ethical laws (act no. 87–848;
Ministère de l’Agriculture et de la Forêt), European Union Council Directives of November 24, 1986 (2010/63/EEC) and guidelines for the care and use of laboratory animals. Experiments have been approved by the ethics committee n°52 on animal experiments (CENOMEXA) and by the French Ministry of Research under the project license number 02653.2 (30/06/2016). None of the experimental procedures induced animal mortality. All experiments were performed following the ARRIVE guidelines (
www.nc3rs.org.uk), including randomization of treatment as well as analysis blind to the treatment.
Animals and surgery
Wild type and tPA
−/− [
17] C57BL6-J mice (aged 2 months, CURB, Caen, France) were housed in a temperature-controlled room on a 12 h light/12 h dark cycle with access to food and water ad libitum. Demyelination was induced by stereotaxic injection of 1% LPC (L1381, Sigma-Aldrich) in 0.1 M phosphate-buffered saline (Sigma-Aldrich). Mice were deeply anaesthetized with isoflurane (induction 5%, maintenance 2.5% in 70/30% NO
2/O
2). Rectal temperature was maintained at 37 °C using a feedback-regulated heating pad. Mice were placed in a stereotaxic frame. The demyelinating agent was injected unilaterally into the corpus callosum with a microinjection pipette (Hecht Assistent), using the following stereotaxic coordinates: 1.1 mm anterior to the bregma, 2.7 mm lateral to the bregma and 1.3 mm deep from the cortex surface, using an angle of 30°. Scalp incisions were closed with Vetbond thread.
Magnetic resonance imaging (MRI) analysis
MRI analyses were performed on 1, 3, 7, 14 and 21 days post-injection. Experiments were carried out on a Pharmascan 7 T MRI (Bruker, Germany). T2-weighted images were acquired using a multislice/multiecho sequence: TE/TR 12.7 ms/2500 ms and a flip angle of 180° with spatial resolution of 70 μm*70 μm and slice thickness 350 μm interpolated to an isotropic resolution of 70 μm (acquisition time = 8 min). Lesion sizes were quantified on these images by an experimenter blinded to the genotype using ImageJ (NIH software v1.49e, National Institute of Health, Bethesda, MD, USA).
Immunohistochemistry
1, 3, 7, 14 and 21 days post-injection mice were deeply anesthetized and perfused transcardially with 20 ml of cold heparinized NaCl 0.9%, followed by 2% paraformaldehyde and 0.2% picric acid in 150 ml of 0.1 M sodium phosphate buffer, pH 7.4. Brains were removed, washed in veronal buffer containing 20% sucrose, and frozen in Tissue-Tek (Miles Scientific). Embryos were collected at noon of embryonic days. Heads and trunks were fixed in 2% paraformaldehyde and 0.2% picric acid for 24 h, cryoprotected in veronal buffer containing 20% sucrose, and frozen in Tissue-Tek. Cryomicrotome-cut sections (10 μm) were collected on poly-lysine slides and stored at −80 °C before processing. Sections were incubated overnight at room temperature with a primary antibody or a cocktail of primary antibodies (Additional file
1: Table S1) diluted in veronal buffer containing 0.25% Triton ×100. Three rinses were performed in veronal buffer at room temperature. All secondary antibodies (Additional file
1: Table S1) were diluted (1:600) in veronal buffer containing 0.25% Triton × 100. Incubation was performed at room temperature for 1h30min followed by three washes. Washed sections were coverslipped with antifade medium containing DAPI. For each set of immunostaining, the following controls were systematically performed on adjacent sections: Omission of primary or secondary antibodies in single immunolabelling experiments resulted in no specific labelling. Additionally, the absence of cross-reactivity between the antibodies in multi-immunolabelling experiments was demonstrated by omission of one of the primary antibodies and consecutively the absence of relevant signal detection with the cocktail of secondary antibodies.
Images were digitally captured using a Leica DM6000B microscope-coupled coolsnap camera (ORCA Flash 4-LT; Hamamatsu), visualized with Metamorph 5.0 software (Molecular Devices) and further processed using ImageJ1.5 h software (NIH).
3D reconstruction
Images were collected using a Leica SP5 confocal microscope with a 100× oil-immersive objective (Leica Microsystems). Confocal images were taken at a 1024 × 1024 pixel resolution with a z-step of 0.45 μm. The 3D structure was reconstructed from confocal images using Imaris software (version 5.5, Bitplane, Zurich, Switzerland). Volume and surface functions were used.
Cell counting and quantification of MBP immunofluoresence
The total number of positive cells was counted in the lesion area. In the contralateral corpus callosum, an equivalent area for each mice was selected and the number of positive cells was counted in the selected area. During development, the area of the whole spinal cord or telencephalon section was measured and the total number of positive cells was counted in the section. For MBP immunofluorescence quantification, the mean gray value was evaluated in the lesion area or the equivalent contralateral area. All quantitative analyses were done in three randomly selected sections per mouse and the individual values for the number of cells/mm2 or for fluorescence intensity for each mouse was estimated by averaging the values of all counted sections for the same mouse.
Oligodendrocyte precursor cell (OPC) cultures
Primary mixed glial cultures were prepared according to the modified technique of McCarthy and de Vellis [
18]. Briefly, forebrains of P
0-P
1 newborn Wistar rats were dissociated mechanically and resuspended in DMEM (D5671, Sigma-Aldrich) containing 10% fetal bovine serum, 10% horse serum, 2 mM glutamine (25030024, Thermo Fisher Scientific), 0.5% penicillin streptomycin (15140122, Thermo Fisher Scientific) and plated on poly-D-lysine-coated (0.1 mg/mL) (P6407, Sigma-Aldrich) 75 cm
2 flasks (Nunc, Wiesbaden, Germany). After 10 days in culture the flasks were shaken at 210 rpm at 37 °C for 3 h to remove loosely adherent microglia. The remaining OPCs present on the top of the confluent monolayer of astrocytes were dislodged by shaking overnight at 270 rpm. The cell suspension was filtered through a 40 μm nylon mesh and then pre-plated on bacterial grade Petri dishes for 1 h. The nonadherent OPCs that remained in suspension were recovered, filtered through a 40 μm nylon mesh, and plated again on bacterial grade Petri dishes for 30 min. The resulting enriched OPCs cell suspension was counted and seeded in accordance to the assay performed.
Cell proliferation assay
OPC proliferation was estimated by examination of the cultures under bright-field microscopy and quantified with WST-1 assay. Briefly, oligodendrocyte precursor cell were plated on poly-D-lysine-coated (0.1 mg/mL), 24-well (2 cm2/well) tissue culture dish at a density of 6 × 104 cells/cm2 and cultured in DMEM (D5671, Sigma-Aldrich) supplemented with recombinant PDGF-AA (PHG0035, Thermo Fisher Scientific, 5 ng/ml) and bFGF (15140–122, Invitrogen, 5 ng/ml), N2 supplement (17502–048, Thermo Fisher Scientific), 2 mM glutamine (25030024, Thermo Fisher Scientific), 0.5% penicillin streptomycin (15140122, Thermo Fisher Scientific). Cells were used for proliferation assay at 2 days in vitro. To assess cell proliferation, WST-1 reagent (Roche Applied Science, Indianapolis, IN), a tetrazolium salt, was added to the medium and incubated for 1 h at 37 °C in 5% CO2. After WST-1 incubation with cells, bathing media from 24-well plates was transferred into 96-plates and cell proliferation was determined by measuring the absorbance at 460 nm (reference wavelength 600 nm) for cleavage of the tetrazolium salt to formazan. A portion of the wells were used to evaluate the number of cells at the beginning of the experiment (t0 value), by adding WST-1 directly to the medium of 3 wells. Other wells were treated with tPA, vehicle or control medium by renewing medium. After 24 h of treatment, cells were incubated with WST-1 to estimate the number of cells after treatment (t1 value). Percentage of proliferation during 24 h (with or without treatment) was calculated using the following formula: [% of proliferation = ((t1 value/t0 value) × 100)-100].
Chemokinesis assay and immunocytochemistry
OPC migration was assessed in chemotaxis chambers with polycarbonate membranes (pore size 8 μm; Corning Costar). The membranes were coated with poly-
L-lysine (0.1 mg/mL) and laminin (0.1 mg/mL) (23017015, Thermo Fisher Scientific) as described previously [
19]. OPCs from rat were seeded (40000 cells/transwell) in the upper chamber while in the lower compartment the DMEM culture medium containing N2 supplement, 2 mM glutamine, penicillin streptomycin was supplemented for the different experimental groups as follows: Control; FGF2 (0.2 μg/mL; RD Systems 233-FB); tPA 0.1, 1, 10 μg/mL (Actilyse; BoehringerIngelheim); tPA buffer (vehicle); tPA 10 μg/mL + inhibitor of EGFR kinase, AG1478 (5 μM; 1276, Tocris); AG1478 (5 μM); tPA GGACK 10 μg/mL. Concerning experimental groups, tPA 10 μg/mL + AG1478 5 μM and AG1478 5 μM, the cells were also treated one hour before and during the experiment with the EGFR blocker AG1478 5 μM and the rest of the cultures were exposed to an equal volume of the vehicle DMSO (Sigma-Aldrich-Aldrich) during the course of the experiment which was carried out at 37 °C, 5% CO
2, and at 95% relative humidity. After 24 h, cells were fixed with 4% paraformaldehyde (PFA; for 10 min at RT), washed 3 times with phosphate buffer saline (PBS, pH 7.4) and the non-migratory cells on the upper membrane surface were removed with a cotton swab. The presence of transmigrated OPCs in the lower chamber was evaluated by immunostaining with Olig2 antibody (1:200, AB9610 Millipore) and its corresponding fluorescent secondary antibody. After immunostaining, the Boyden filters were examined and images were digitally captured using a Leica DM6000 microscope-coupled coolsnap camera and visualized with Metamorph 5.0 software (Molecular Devices). To quantify chemokinesis, 16 fields per well (×20 objective) taken randomly were photographed and the number of transmigrated OPCs per field was counted using ImageJ 1.49e software (NIH). The data were expressed as number of migrating OPCs per mm
2 ± standard error to the mean (SEM).
Production of proteolytically inactive tPA (GGACK tPA)
GGACK (1,5-dansyl-L-glutamyl-L-glycyl-L-arginine chloromethylketone; EMD) was added to Actilyse (BoehringerIngelheim) in a fourfold molar excess. The solution was allowed to react for 24 h at room temperature and dialysed for 48 h at 4 °C with PBS to remove all unbound GGACK. The actilyse buffer was reconstituted with arginine monohydrochloride (Sigma-Aldrich-Aldrich) added to GGACK-tPA. Finally, the lack of proteolytic activity of GGACK-tPA was confirmed with a spectrozyme assay (American Diagnostica).
Immunoblotting
After dissociation with ice-cold TNT buffer (50 mM Tris–HCl pH 7.4; 150 mM NaCl; 0.5% Triton X-100), cells were centrifuged (12,000 g, 4 °C, 15 min) and protein content assessed by the BCA method (Pierce, France). Proteins (20 μg) were separated by 10% SDS-PAGE and transferred onto a PVDF membrane. Membranes were blocked with TBS (10 mM Tris; 200 mM NaCl; pH 7.4) containing 0.05% Tween-20, 5% BSA, and incubated overnight at 4 °C with primary antibodies against Erk 1/2 (42 kDA) and phosphorylated Erk 1/2 (42 kDA) (Cell signaling). After incubation with the anti-rabbit peroxydase-conjugated secondary antibodies (1:50000), proteins were visualized with an enhanced chemiluminescence ECL-Plus detection system (Perkin ElmerNEN, France).
Statistical analysis
All results are expressed as mean ± SEM. For in vitro experiments, the n value corresponds to n different well pools derived from independent dissections. For group comparison, Kruskal-Wallis tests were used followed by Mann–Whitney U-tests as post hoc tests.
Discussion
The present study reports the effects of tPA on the migration of OPCs during embryonic development and during remyelination after white matter lesion. We show that, in both situations, OPC migration is delayed in tPA−/− animals. This was illustrated by a lower number of OPCs leaving the pMN zone (spinal cord development) or reaching the remyelinating corpus callosum (post-lesion recovery) in tPA−/− mice. During development and regeneration, we observed that tPA was expressed in vessels. Interestingly, in both cases, OPC migration occurred along vessels and the proportion of OPCs surrounding these vessels was reduced in tPA−/− animals. Finally, we highlighted the role of EGFR signaling in these processes by showing that EGFR was expressed on migrating OPCs and that, in vitro, tPA exerted a chemokinetic effect on OPCs which was mediated by the activation EGFR.
The involvement of tPA in the migration of neurons during development of the CNS has been first shown more than 15 years ago [
14] and has been confirmed later [
24]. The present study brings new advances by showing that oligodendrocyte migration is also facilitated by tPA. In fact, tPA increases the migration of several other cell types, within or outside the central and peripheral nervous system, such as macrophages [
25,
26] or cancer cells [
27]. However, in the CNS, while most studies report the influence of tPA on migration during development, the issue of how the pro-migratory effects of tPA can influence pathological situations has been scarcely addressed [
28]. Our study brings new elements by highlighting the potential role of tPA on oligodendrocyte migration during white matter damage. This suggests implication for several pathologies in which white matter is damaged, such as multiple sclerosis, leukodystrophies, periventricular leukomalacia, white matter stroke, head trauma or spinal cord injury.
Interestingly, in our study, we show that tPA invalidation leads to a delay in oligodendrocyte migration during development and after white matter demyelination. This suggests that, despite the fact that the oligodendrocyte populations involved in these two situations are different [
29], their chemokinetic response to tPA would be equivalent. Pro-migratory effect of tPA was observed in the corpus callosum and in the spinal cord, which suggests equivalent responses in these two regions of the CNS. Thus, tPA appears to rise homogeneous responses in oligodendrocytes migration regardless of the age (embryo, adult), region (telencephalon, spinal cord) or pathological state.
Although the mechanism of action of tPA on migration was suggested before to involve proteolytic degradation of the extracellular matrix [
24], our data show that tPA can also exert protease independent chemokinetic effects on OPCs by activating the EGFR. Several previous in vitro studies reported protease-independent effects of tPA on neurons [
30‐
32] and oligodendrocytes precursors [
11] via a mechanism termed as cytokine- or growth factor–like. In OPCs, EGF receptor activation by tPA was shown to mediate antiapoptotic effects, thus sustaining protection of the white matter after experimental ischemic stroke [
11]. The present study indicates that tPA can also exert pro-migratory effect on OPCs by activating EGFR. Several differences in the experimental approach explain why pro-migratory, rather than anti-apoptotic effects are unveiled in the present work. First, the lesion model used here is based on the detergent properties of lysolecithin which toxic effects largely supplants apoptosis. In effect, the lesion size at 1dpi is mostly similar in wild type and tPA
−/− animals, and the effect of tPA invalidation appears only from 3dpi, when OPC migration and remyelination starts. Second, our study addresses the effect of endogenous tPA, while in previous works [
11] tPA was injected to the animals. The impact of tPA injection largely differs from that of endogenous tPA in terms of dose, timing and accessibility to injured tissues. Finally, the timing in which the study was conducted is also different: while in the Correa et al. paper, the impact of tPA injection was assessed in the subacute phase (1 day post ischemia), in the present study, histological analysis were performed in subacute and chronic phases (up to 21dpi), which enables observing tissue regeneration, in particular OPC migration.
The activation of EGFR pathway has been previously described as an activator of OPC migration. Migratory NG2
+ cells express more EGFR than non-migratory cells [
16] and overexpression in non-migratory cells prompts them to migrate [
16]. Accordingly, EGFR-expressing retinal progenitor cells show increased chemokinesis in the presence of EGF [
33]. In the context of white matter lesion, overexpression or invalidation of EGFR signaling in oligodendrocytes respectively accelerate [
34] or reduces [
35] remyelination after focal demyelination of corpus callosum. In light of these findings, the effect of intranasal EGF treatment was tested in a mouse model of preterm diffuse white matter injury [
36]. This treatment led to enhanced generation of OPCs, leading to functional recovery [
36]. Chemokinetic effects were proposed to sustain these effects of EGFR signaling activation, as EGF induces neural precursor cells to differentiate to glial cells and to acquire a motile phenotype [
37]. In our study, we suggest that this EGFR-dependent pathway could be activated by tPA to exert pro-migratory effects on OPCs in pathological conditions where white matter is damaged, without inducing hyperplasia of OPCs as reported when EGFR is constitutively expressed [
38].
The increase in the number of OPCs during remyelination upon EGFR activation has been suggested in the past to occur via increased proliferation, enhanced differentiation from neural progenitors, migration or combinations of these different mechanisms [
39‐
41]. Here we show that tPA does not influence OPC proliferation in vitro, but rather enhances their migration via a chemokinetic dependent on the activation of EGFR. One possibility to explain this difference is that the activation of EGFR by tPA, by its natural ligand (EGF) or by genetic strategies can induce different response in target cells, in particular because the degree of activation of downstream targets in these different situations may vary considerably.
Our in vitro data suggest that tPA mediates its promigratory effects, at least in part, by activation of EGFR and independently of its proteolytic activity. Nevertheless, proteolytic effects have also been suggested: tPA was shown to proteolytically activate proBDNF into mature BDNF, which increases proliferation of OPCs in vitro [
42] andpromotes remyelination when injected intra veinously after WM stroke [
43]. In our hands, tPA invalidation did not modify proliferation of OPC, which claims for a different mechanism. Activation of growth factor pathways by tPA can also occur by a regulation of expression, rather than maturation: tPA upregulates vascular endothelial growth factor (VEGF) by endothelial cells [
44] which others have reported to promote the migration of oligodendrocyte precursor cells in vitro [
45,
46]. Notheworthy, whether the VEGF pathway targets OPC proliferation [
47] or not [
46] is still a matter of debate. In any case, we report here that tPA deficiency did not influence OPC proliferation. Further studies may help decipher possible links between tPA and the VEGF pathway in OPC migration in vivo, during development and/or remyelination.
The previous point highlights the possible cooperative role between endothelial cells and OPCs during (re)myelination. This has led to the concept of « oligovascular niche » [
48]: vascular cells would secrete soluble factors (including VEGF) which promotes migration and survival of oligodendrocytes [
49]. Because tPA is mainly produced by endothelial cells, it could be one of these oligotrophic factors. In our hands, the common cellular source of tPA in development and adult remyelination was endothelial cells, which corroborates a possible role of endothelial-derived tPA in the observed effects on OPC migration. This is particularly relevant to the increasing literature describing that migration of progenitor cells, including OPCs, is guided by vessels. Blood vessels form a scaffold for migration of neuroblasts to the adult olfactory bulb [
50], and more recently, a similar mechanism has been described specifically for OPCs migration [
21]. A cooperation between vessels and progenitors exists, in which neural progenitors (including OPCs) induce local angiogenesis while migrating, which in turns facilitates progenitor emigration from the niche [
51]. This « vasophilic » migration [
50] was described during development in previous studies, but the present work is to our knowledge the first to describe it during remyelination after white matter lesion. This is particularly interesting in regard to the description of perivascular tPA deposits in acute MS lesions [
52], a type of demyelinated plaque where remyelination may succeed spontaneously.
Pericytes have also been suggested to influence OPC migration [
53], which raises the question of the expression of tPA in pericytes. During development, we did not detect pericytes in the developing spinal cord at E13, when the phenotypic differences between wt and tPA
−/− first occurred (data not shown), which is likely due to the immaturity of the vessels of the spinal cord at this stage, in which pericytes are usually not reported [
54‐
56]. After white matter lesion, tPA was not detected in pericytes surrounding blood vessels at 3 dpi (Additional file
4: Figure S3), when the delay in OPC migration was observed in tPA
−/− as compared to wt mice. This is in accordance with our previous report of the absence of tPA in adult pericytes [
12]. Nevertheless, we not cannot exclude with certitude that tPA is not expressed in pericytes, and this should be the purpose of future studies.
This link between angiogenesis and OPC migration is relevant to white matter pathology, in particular multiple sclerosis, in which angiogenesis has been described [
57]. tPA could be involved in these processes in regard to a previous study in experimental autoimmune encephalomyelitis, a model of multiple sclerosis, in which recovery is reduced in tPA
−/− animals [
58]. Noteworthy, while wild type animals showed remyelination in late phases of the disease, this remyelination was severely reduced in tPA
−/− animals [
58]. The present study brings new elements in a model where demyelination is the primary insult:we here show that the delay in OPC migration to the damaged area results in reduced remyelination in tPA
−/− mice. This strengthens the idea that the failure in recovery in tPA
−/− in EAE could be the result of a reduced capacity of remyelination. Overall, the present data suggest that endogenous tPA could be important for remyelination in multiple sclerosis. In accordance with this, tPA activity has been shown to be decreased in post-mortem tissues from MS patients [
59], alleging that a drop of tPA may participate in reducing the opportunity of remyelination in MS patients.