Oligodendrocytes (OL) are the myelin-forming cells of the central nervous system (CNS). These cells originate from progenitors generated in specific regions of ventral forebrain and migrate extensively throughout the CNS during development (reviewed by Baumann and Pham-Dinh,2001; Bradl and Lassmann,2010). When an OL progenitor cell (OPC) arrives at its appropriate location, it begins to differentiate into a mature OL, extending multiple processes, each of which contacts and ensheathes an axonal segment. However, the molecular cues that coordinate the migration and timing of myelin formation in vivo are not well understood at the present time.
Numerous in vitro studies have demonstrated that OPC differentiation can be influenced by extrinsic and intrinsic factors, including growth factors, extracellular matrix molecules, and transcription factors (reviewed by Pfeiffer et al.,1993; Jakovcevski et al.,2009; Li et al.,2009). To further understanding of the molecular events that regulate OPC differentiation, we conducted a detailed microarray screening of OPCs in vitro using cells at the progenitor, premyelinating, and mature stages of differentiation. One unexpected finding was the expression of the mRNA for reelin. Reelin is a large glycoprotein that is secreted by Cajal–Retzius cells in the marginal zone, and functions as a critical regulator of neuronal migration (reviewed by D'Arcangelo,2006). The results of this study reveal that OPCs and OLs not only synthesize and secrete reelin but also express a receptor and essential components of the intracellular signaling pathway for reelin. These data suggest that oligodendroglial (ODG) cells are a source of reelin in the CNS and that reelin may potentially modulate OL behavior during development.
MATERIALS AND METHODS
All procedures involving animals were in compliance with guidelines established by the institutional Committee for the Humane Use of Animals and Department of Laboratory Animal Resources following recommendations from the Association for Assessment and Accreditation of Laboratory Animal Care.
Primary Cell Culture
Primary mixed glial cell cultures were isolated from neonatal (P2) Sprague-Dawley rats (Taconic Farms, Germantown, NY) as described previously (Osterhout et al.,1997). In brief, the cortices of neonatal brains were disassociated, and the resulting cell suspension was plated in poly-D-lysine polylysine coated tissue culture flasks. When the monolayer of cells was confluent, OPCs were isolated by shaking overnight at 37°C and further purified by differential plating.
Purified OPCs were plated onto culture dishes precoated with either laminin (Invitrogen, Carlsbad, CA) or poly-D-lysine (Sigma, St. Louis, MO). Cells were cultured for the first 24 hr in defined serum free media (adapted from Osterhout et al.,1999, containing insulin, selenium, transferrin, and triiodothyronine) supplemented with B104 neuroblastoma conditioned media (CM; B104-CM, 30% v/v) to prevent OPC differentiation. For the Western blot analysis of reelin levels, OPCs were maintained in defined serum free media containing 10 ng/mL PDGF-AA (Peprotech, Rocky Hill, NJ) and 20 ng/mL FGF2 (Peprotech, Rocky Hill, NJ) to prevent OPC differentiation. B104-CM was not included in the media for these experiments as it contains reelin (data not shown), and as such, we could not detect reelin secretion by OPCs in the presence of B104-CM. To initiate OPC differentiation, the culture media was changed to the defined serum free media without growth factors.
Six dishes of OPCs were harvested at 0, 4, or 6 days postplating for each substrate. Culture dishes were placed on ice and rinsed three times in ice-cold 1x PBS. Cells were lysed by adding 300-μL RLT lysis buffer (Qiagen, Valencia, CA) with 1% β-mercaptoethanol (Sigma Aldrich, St. Louis, MO). RNA samples were snap frozen on dry ice and stored at −80°C until further processing. RNA was purified and amplified using the Ovation™ Biotin RNA amplification and Labeling System (NuGen, San Carlos, CA). An enriched antisense cDNA probe was produced from the RNA using the basic 3′ Ribo-SPIA™ (NuGen, San Carlos, CA).
Amplification and labeling of cDNA probes was performed by the SUNY Upstate Medical University Microarray Core Facility. cDNA probes were hybridized to the Rat genome 230 2.0 gene array (Affymetrix, Santa Clara, CA). All hybridization, washing, labeling, and detecting steps were performed according to the manufacturers established protocols.
Expression values were obtained using RMA normalization of the CEL files and data formatted for import into Microarray Experiment Viewer 4.2 (Saeed et al.,2003). Significant changes in gene expression were detected using a two-way ANOVA to compare the effects of plating substrate and differentiation time across the three time points.
Coverslips were fixed every day over the time course of 1 week, using 4% paraformaldehyde. The fixed cultures were permeabilized in 0.25% Triton-X 100 in 1x PBS for 5 min and blocked in an antibody blocking serum (3% FBS and 3% BSA in HBSS) for 1 hr at room temperature. All primary antibodies were diluted in antibody blocking serum and left on the cultures overnight at 4°C. Primary antibodies included three for reelin (CR50, 1:200, a gift from Dr. Huaiyu Hu; clone 142, 1:200, Millipore, Bellerica, MA, and G10, 1:200, Abcam, Cambridge, MA), Dab1 polyclonal antibody (1:200; a gift from Dr. Brian Howell) and very low density lipid receptor (VLDLR, 1:1000; gift from Dr. Brian Howell). The coverslips were rinsed and incubated with secondary antibodies, a goat-anti-mouse or goat-anti-rabbit Alexafluor 488 or 555 (1:500, Invitrogen, Carlsbad, CA), for 1 hr at room temperature. Coverslips were rinsed and mounted on slides using Prolong Gold mounting media with DAPI (Invitrogen, Carlsbad, CA).
CM and cell lysate collection.
OPC CM was collected every 2 days and concentrated using the Amicon Ultra-4 Centrifugal Filter Unit (Millipore, Bellerica, MA). OPC cell harvest was done on ice. Cells were washed three times in ice-cold 1x PBS and detached by scraping. The resulting cell suspension was transferred to a microfuge tube and pelleted by centrifugation. The resulting cell pellet was resuspended in ice-cold lysis buffer with protease inhibitors added fresh. All harvested CM and lysates were kept at –80°C.
Protein lysates were harvested in the same manner as for the Western blots. Lysates were incubated overnight at 4°C with a polyclonal antibody to Dab1 (1:200, Biodesign, Saco, ME) and precipitated out with Protein-A agarose beads (Santa Cruz Biotech, Santa Cruz, CA).
Samples were run on a precast gradient, range 3%–8%, Tris-Acetate Nupage gel (Invitrogen), followed by transfer to a PVDF membrane. Membranes were blocked overnight in a solution of 5% instant milk in TBS containing 0.1% Tween-20 and then probed for reelin (G10, 1:200, Abcam, Cambridge, MA), Dab1 (1:1000, from Dr. Howell), phosphorylated tyrosine (4G10; 1 μg/mL), and VLDLR (1:1000; from Dr. Howell). Primary antibodies were diluted in the blocking solution and incubated overnight at 4°C. Membranes were rinsed, and the bound antibodies were detected using an ECF kit (GE Health Care) per the manufacturer protocol. Labeled membranes were imaged on a STORM 840.
ODG Cells Express the mRNA for Reelin, Dab1, and VLDLR
To better understand the changes in gene expression that occur as the OPC matures into an OL, a detailed microarray analysis was performed on ODG cells at several stages of differentiation: the progenitor, premyelinating progenitors, and mature myelinating OL. Genes that demonstrated a significant change in expression were determined by a two-factor ANOVA analysis, (P ≤ 0.01) level based on 1,000 permutations, comparing the effect of the substrate, the OPCs were plated on against the stage of differentiation. The results of this analysis revealed two specific genes, which unexpectedly showed a significant change in expression as OPCs underwent differentiation.
The heat map in Fig. 1A summarizes the expression profile for reelin and reelin-related genes. Reelin (Reln) and disabled homolog 1 (Dab1) are both expressed at high levels in the progenitors. However, at the premyelinating phase of differentiation, there is a significant decrease in the expression levels of Reln (P < 0.0001) and Dab1 (P < 0.001). As cell differentiation continues, Reln and Dab1 maintain a constant but low level of expression. The expression of Reln and Dab1 are independent of the substrate, with similar expression patterns on laminin and polylysine. Interestingly, the gene for a reelin surface receptor, Vldlr (P = 0.564) was moderately expressed during all time points and did not change over the course of differentiation.
In addition to the Reln, Dab1, and Vldlr genes, two specific myelin genes with known expression patterns were also identified: myelin-associated glycoprotein, (Mag), and myelin basic protein, (Mbp). These genes served as internal controls confirming the differentiation of OPCs. The discovery that Reln and Dab1 genes were present in ODG was unexpected and indicates that ODG cells may express reelin and associated proteins.
ODG Cells Express Reelin Protein
Western blotting analysis of ODG whole cell lysates demonstrated that reelin protein was expressed at all stages, from the progenitor phase (lane 1; Fig. 1B) and all subsequent days after the initiation of differentiation (2, 3, 4, and 5 days; Fig. 1B). Reelin expression appears to be maximal in progenitors and decreases as the cells start to differentiate. A lower level of reelin expression remains constant through the premyelinating progenitor phase (3 and 4 days in vitro) and mature OLs (6 days in vitro). Immunocytochemical studies showed the cellular localization of reelin, which is observed throughout the cell body and proximal cell processes of progenitor cells (day 1; Fig. 1C). However, as the cell differentiates into the premyelinating stage, reelin localizes to the perinuclear area and expression in the proximal processes is barely visible. In the mature OL, reelin is highly perinuclear and completely absent in any of the cellular processes (day 6; Fig. 1C).
Collectively, the western blotting and immunocytochemistry results confirm the genetic expression data, revealing that ODG cells express reelin at every stage of maturation and suggesting that ODG could be a source of reelin in the CNS.
ODG Cells Express Components of the Reelin Signaling Pathway
As with reelin, both Western blots and immunocytochemical data confirmed the protein expression of Dab1, the reelin surface receptor, and VLDLR. VLDLR is present in all ODG cells, from progenitors to mature cells (5 days postplating; Fig. 1B). VLDLR levels steadily increase as OPCs differentiate into a mature OL, with maximal expression being found at 5 days postplating (4 days differentiation time). Within the cell, VLDLR is observed at all stages of differentiation throughout the cell body and processes (Fig. 1C). In mature OLs, VLDLR expression can be seen not only in the cell body and cytoplasmic processes but also in patchy areas of the myelin membrane sheet proximal to the branching points of processes (Fig. 1C).
The expression of the cytoplasmic adaptor protein Dab1 appears to mirror the expression pattern of reelin. Dab1 protein is present in both progenitors and differentiating cultures (Fig. 1B). Dab1 levels are highest in the progenitor cells, and as OPCs differentiate, Dab1 levels decrease. A low, but steady expression level of Dab1 is maintained at all stages of differentiation. Dab1 appears in the cytoplasmic processes and cell body in both the progenitor and premyelinating progenitor phases (1 and 3 days postplating; Fig. 1C). In mature OLs, Dab1 remains in both the cell body and cytoplasmic processes but appears to be excluded from the developing myelin membrane sheet (Fig. 1C).
Collectively, these data show that ODG cells express reelin, as well as key components of the reelin signaling pathway, the receptor for reelin (VLDLR) and the cytoplasmic signaling molecule for reelin (Dab1). This suggests that ODG cells not only produce reelin but they may also be responsive to the reelin in an autocrine manner.
ODG Cells Secrete the Reelin Protein and Display an Intracellular Dab1-Fyn Interaction
With evidence of reelin protein expression in the cell lysates, it was next determined if ODG actually secrete reelin. Conditioned OPC culture media (24 hr incubation) was analyzed by Western blotting to see if reelin was detectable. When the CM is compared with the progenitor cell lysate (Lys), the amount of reelin found in the media was dramatically higher than the protein levels found in the cell lysate (Fig. 1D). Culture medias were conditioned by ODG cells at various time of differentiation (1, 3, 5, and 7 days) and probed for reelin, which was detected at all time points (Fig. 1E). This finding suggests that most of the reelin being produced by these cells is being secreted.
Given that ODG express all three components of the reelin signaling pathway, it might be expected that reelin can bind to ODG and activate intracellular signaling pathways associated with reelin. Reelin–Dab1 interactions are critical for a cell to respond to reelin (Howell et al.,1997). Dab1 can also interact with the Src family kinase Fyn, which is a critical regulator of ODG differentiation (Osterhout et al.,1999). If these pathways are involved, Dab1 would be phosphorylated in the presence of reelin. To examine this, we immunoprecipitated Dab1 from the cell lysate and probed it for phosphorylated tyrosine, which would indicate protein activation. There were consistent low levels of tyrosine phosphorylation at all time points, but it is unclear if this is truly in response to reelin in the culture media (data not shown). We also probed for a Dab1–Fyn interaction by immunoprecipitating with Dab1 and probing for Fyn. We observed Fyn–Dab1 associations at all time points along a differentiation time course (Fig. 1F). Interestingly, the interaction looks strongest in the ODG progenitors, a stage before Fyn activation and the beginning of OPC differentiation (Osterhout et al.,1999).
These findings reveal that ODG cells, regardless of their differentiation status, will secrete reelin into their environment. ODG cells are also unique in that they express all of the essential components of the reelin signaling pathway: reelin, a reelin receptor, and the cytoplasmic adaptor protein. Taken together, this suggests that ODG cells may be a major source of reelin in the CNS, and moreover, reelin may play a role in the biology of ODG cells.
Reelin is a critical modulator of neuronal positioning and migration during brain development (reviewed by Rice and Curran,2001; Tissir and Goffinet,2003). It is expressed by specific populations of migrating neurons, whereas the reelin receptors are expressed by different target cells. Reelin is not typically expressed by glial cells; to date, it has only been observed in a small subset of spinal cord astrocytes during development (Hochstim et al.,2008). This study demonstrates that ODG cells express reelin, Dab1, and VLDLR proteins and secrete reelin into the surrounding environment. This finding is unique, in that ODG express all three components needed for reelin signaling, raising the possibility that reelin secretion may have an autocrine effect. However, any biological effects of reelin, whether through autocrine or paracrine signaling, on ODG cells has not been previously investigated.
Our data also suggests that ODG may be a source of reelin in the developing CNS. This is in agreement with two studies that also provide indirect evidence that ODGs may be a source of reelin. In this work, OPCs were ablated in mice, which generated severe cerebellar malformations. Subsequent analysis revealed decreased reelin levels and a disorganization of the cellular architecture that included impaired interneuron cell migration (Mathis et al.,2003; Collin et al.,2007). However, the possibility that cells in the ODG lineage are reelin producers has not been examined before this work.
Studies of the intracellular signaling pathways activated by reelin have been only done in neurons, and identified functional interactions between the cytoplasmic adaptor molecule Dab1, the serine-threonine kinases P35 and Cdk5, and Src family kinase (Fyn), which may yield a clue to the role reelin in the OL lineage (Howell et al.,1997; Arnaud et al.,2003; Fatermi,2005; Herz and Chen,2006). Of particular interest is the discovery that reelin binding to its receptor activates the Src family kinase Fyn (Fatermi,2005). Fyn activation is essential for the migration and differentiation of OPCs (Osterhout et al.,1999; Miyamoto et al.,2008). This suggests reelin may play a role in OPC migration and differentiation. We show that both the reelin and Dab1 proteins are at maximal expression during the progenitor phase and are quickly down regulated during OPC differentiation. This suggests that reelin may be more important for migration, rather than differentiation.
Although the role of reelin in ODG biology is undetermined, studies are currently underway utilizing mutant mouse models (reeler and scrambler) to understand how reelin influences the behavior of ODG cells. The findings from this study suggest that reelin may have a role in ODG biology, and that ODG cells may be an important source of reelin in the CNS.
The authors would like to thank Yijuan Lin for maintaining the primary OPC cultures. The authors are very grateful to Drs. Huaiyu Hu (CR-50), and Brain Howell (anti-Dab1, and VLDLR) in the Neuroscience and Physiology Department for the antibodies used this study, as well as helpful discussions of our work. The authors gratefully acknowledge Ms. Karen Gentile and the SUNY Upstate Medical University Microarray Core Facility for processing all the RNA samples, and Dr. Frank Middleton for his assistance with the microarray data analysis.