Rnd2 differentially regulates oligodendrocyte myelination at different developmental periods

INTRODUCTION

During the development of the CNS, oligodendrocyte precursor cells (OPCs) arise along the caudorostral axis of the ventricular neuroepithelium (Bradl and Lassmann, 2010; Simons and Lyons, 2013; Bergles and Richardson, 2015; Saab and Nave, 2017). OPCs then differentiate into mature oligodendrocytes, forming myelin sheaths with their multiple layers. Mature myelin sheaths protect axons from various external stresses, including physical and physiological stresses, and insulate them to increase nerve conduction velocity. Although we know of various extrinsic and intrinsic cues such as growth factors, cell adhesion molecules, and protein kinases that mediate cellular morphology and gene expression, thereby regulating the spatial and temporal patterns of oligodendrocyte myelination, the overall picture of the myelination mechanism remains to be thoroughly clarified (Zuchero and Barres, 2013).

Myelin formation begins with the first contact of oligodendrocyte processes along axons, which triggers the subsequent steps. The molecular steps controlling oligodendrocyte myelination often involve the extensive reorganization of cytoskeletal proteins, leading to morphological changes (Nawaz et al., 2015; Zuchero et al., 2015). The small guanosine triphosphatase (GTPase) family is a major regulator underlying such changes. It serves as a molecular switch by cycling between an active guanosine triphosphate (GTP)-bound form and an inactive guanosine diphosphate (GDP)-bound form (Jaffe and Hall, 2005; Rossman et al., 2005). In the active state, members of this family bind to downstream effectors to evoke diverse biological functions. For example, Rho family small GTPases Cdc42 and Rac1 act as positive molecular switches controlling oligodendrocyte differentiation and myelination (Liang et al., 2004; Thurnherr et al., 2006). The leucine-rich repeat and immunoglobulin-like domain–containing Nogo receptor-interacting protein (LINGO)-1, a component of the p75 neurotrophin receptor and tumor necrosis factor receptor family member Troy signaling complex, negatively regulates myelination through a RhoA-dependent mechanism (Mi et al., 2004). The inhibition of LINGO-1 leads to the down-regulation of RhoA activity and promotes OPC differentiation in vitro (Zhao et al., 2007).

The Rnd proteins, Rnd1, Rnd2, and Rnd3/RhoE, which constitute a unique branch of the Rho family of small GTPase, are important regulators of cell migration, morphological changes, and differentiation in various types of cells (Chardin, 2006; Riou et al., 2010). Rnd proteins have weak or no intrinsic GTPase activity and accordingly are constitutively bound to GTP. They are thought to be regulated at the expression, phosphorylation, or cellular localization levels. It is likely that Rnd proteins, especially Rnd1 and Rnd3, bind to the p190RhoGAP and increase the GTPase-activating protein (GAP) activity to decrease active GTP-bound RhoA, which antagonizes the cellular functions of RhoA and the downstream Rho kinase (Wennerberg et al., 2003). Rnd2 plays a role in morphological changes in neuronal cells (Fujita et al., 2002; Tanaka et al., 2006) and also induces neurite branching in concert with the effector protein, Rapostlin, in PC12 cells (Fujita et al., 2002). Furthermore, the transcription factor Neurogenin-2 is known to promote the radial migration of cortical neurons by inducing the expression of Rnd2 along the time axis (Heng et al., 2008).

Here, we report that Rnd2 affects oligodendrocyte myelination in different ways during different periods of development. In mice, knockout of Rnd2 in oligodendrocytes leads to decreased myelin thickness in an early developmental period soon after birth but leads to increased myelin thickness in a later developmental period, which is in keeping with the results from oligodendrocyte-neuronal cocultures. The role of Rnd2 can also be observed using oligodendrocyte-specific Rnd2 transgenic mice. These phenomena are consistent with the fact that signaling through Rho kinase occurs in a dephosphorylated state and is down-regulated during the first period of myelination but occurs in a phosphorylated state later in the myelination process. A similar observation was made in the case of Rnd2 knockout in adulthood. Together, our findings reveal that Rnd2 has different effects on the regulation of myelination in different periods of myelination, demonstrating that Rnd2 plays a unique role in the CNS.

RESULTS

Rnd2 is up-regulated during CNS development

Rnd2 is a small GTPase that has been extensively investigated in neuronal progenitor cells (Fujita et al., 2002; Tanaka et al., 2006; Heng et al., 2008). Transcriptome data related to oligodendrocyte differentiation have been reported by our own and other groups (Dugas et al., 2006). Our transcriptome analysis shows that Rnd2 is transcribed in rat primary oligodendrocytes in our transcriptome analysis (Table 1).

To examine the expression patterns of Rnd2 during CNS development, we performed immunoblotting with an anti-Rnd2 antibody using whole mouse brains. The expression levels of Rnd2 were clearly detectable from 14 d after birth (Figure 1A), the approximate time point when active myelination starts in the developing brain. Elevated expression levels of Rnd2 were detected until adulthood. As the mice developed, the expression levels of the myelin marker proteins 2′, 3′-cyclic-nucleotide 3′-phosphodiesterase (CNPase) and myelin basic protein (MBP), the major central myelin proteins, were increased. The expression levels of control actin were comparable in all stages. The expression profiles of Rnd2 and myelin marker proteins in cocultures established from the primary oligodendrocytes and dorsal root ganglion (DRG) neurons (1–21 d in vitro [DIV]) of rats were similar to those observed in developing brains (Figure 1B). Additionally, in immunoblotting using lysates from primary oligodendrocytes, the expression levels of Rnd2 were detected from DIV1 and increased as differentiation proceeded (Figure 1C). Similar patterns were observed in myelin marker proteins. These results suggest that Rnd2 is progressively expressed in differentiating oligodendrocytes and that its expression increases as myelination proceeds in the CNS.

During development, OPCs undergo a differentiation program that is traced by myelin-specific expression markers and typical changes in cell morphology. To further characterize the spatial and temporal expression of Rnd2, we performed immunohistochemical analysis using mouse brain sections. Rnd2 exhibited a dispersive expression in tissues and was costained with the platelet-derived growth factor receptor (PDGFR) α-positive oligodendrocyte precursor cells at postnatal day 7 and the premyelinating marker O4 at postnatal day 14 as well as with the myelinating oligodendrocyte marker CC1 at postnatal days 14 and 21 in the corpus callosum (Figure 1, D and E). In addition, immunostaining of the corpus callosum at postnatal day 14 using an antibody against neurofilament (NF) proteins revealed that Rnd2 was not present in NF-positive cells (Figure 1F).

Rnd2 regulates the onset of myelination

Next we investigated the effect of knockout of Rnd2 in oligodendrocytes. We generated Rnd2 conditional knockout mice that were suitable for the Cre-loxP recombination system (Figure 2A). Oligodendrocyte-specific MBP promoter–driven Cre recombinase transgenic mice were used for the deletion of loxP-flanked genes. Production of Rnd2 conditional knockout mice was confirmed by Southern blotting (Supplemental Figure S1A) and genomic PCR (Supplemental Figure S1B) with specific probes and primers, respectively, using tail genomic DNA. To check whether Cre was active in MBP promoter–expressing tissues, we designed the primer pair for the Cre-excised region and performed PCR amplification using the genome from the mouse corpus callosum (Supplemental Figure S1C). To further confirm knockout effectiveness in the mouse brain, we preformed coimmunostaining for Rnd2 with CC1 in the corpus callosum. Staining for Rnd2 was decreased in knockout mice (Supplemental Figure S1, D and E).

We first performed an immunohistochemical analysis of MBP in 14-d-old brain tissues. Knockout of Rnd2 decreased MBP staining levels compared with littermate controls (Figure 2, B and C, tentatively referred to as Rnd2^fl/fl; Mbp-Cre in the figure panels). The number of cells positive for the oligodendrocyte lineage cell marker Olig2 was unchanged in knockout mice (Figure 2, B and D). Staining for the astrocyte marker glial fibrillary acidic protein (GFAP) and the neuron marker NeuN was comparable between knockout mice and controls (Supplemental Figure S2, A–D). Recently, CNPase has been shown to be crucial for maintaining key channels in the myelin sheaths. An increase in CNPase has been found to be correlated with greater numbers of cytoplasmic channels (Snaidero et al., 2017). Therefore, we next performed immunostaining of CNPase at day 14 in Rnd2 knockout mice. The expression levels of CNPase were decreased in Rnd2 knockout mouse brains (Figure 2, E and F), suggesting a role of Rnd2 in CNPase expression levels. To further confirm the changes in myelin marker expression levels through immunoblotting, we isolated total proteins from brains. Knockout of Rnd2 decreased MBP and CNPase expression levels (Figure 2, G and H), suggesting that Rnd2 is involved in the onset of myelination in vivo.

Because Rnd2 conditional knockout mice exhibited decreased MBP and CNPase levels, we analyzed the myelin ultrastructure in the corpus callosum. Electron microscopic analysis revealed decreased myelin thickness in 15-d-old knockout mice compared with those from control mice (Figure 3A). The decreased myelin thickness in knockout mice is evident from the quantification of the average g-ratios, which are the numerical average ratios of the axon diameter to the diameter of the outermost myelinated fiber. Thinner myelin sheaths with larger average g-ratios as well as distribution of the g-ratios were detected in knockout mice (Figure 3, B–D; 0.864 ± 0.0300 in knockout mice and 0.831 ± 0.0411 in controls).

To further evaluate the role of Rnd2 in myelination, we established an oligodendrocyte-neuronal coculture system. We developed two nonoverlapping Rnd2 short hairpin RNAs (shRNAs; Rnd2#1 and Rnd2#2, which were effective for oligodendrocytes) (Figure 4A). In cocultures, knockdown of Rnd2 in oligodendrocytes decreased the amounts of MBP-positive cells and those of Olig2-positive cells expressing MBP at 21 d (Figure 4, B–D). The amounts of NF proteins were comparable in Rnd2 knocked down cells and controls. Additionally, Rnd2 knockdown decreased the formation of MBP-positive myelin segments and their length (Figure 4, E and F). Similar results were obtained through staining with an anti-CNPase antibody (Supplemental Figure S3, A–C). It seemed likely that Rnd2#2 shRNA was more effective than Rnd2#1 in their length. The number of Olig2-positive cells was not significantly different between control shRNA- and Rnd2 shRNA–infected cocultures (Figure 4G), suggesting that Rnd2 is not essential for the generation of oligodendrocyte lineage cells in vitro. Furthermore, the expression levels of MBP and CNPase were decreased by knockdown of Rnd2 in cocultures (Figure 4, H and I).

Considering these in vivo and in vitro findings together leads us to conclude that Rnd2 is involved in the onset of myelination.

Rnd2 inhibits myelination at later developmental stages

To investigate whether knockout of Rnd2 has the same effects on myelin thickness in later developmental periods, we examined the staining levels of MBP in the 28-d-old brain. It is of note that Rnd2 knockout mice exhibited increased MBP staining levels compared with littermate controls (Figure 5, A and B). In contrast, the number of Olig2-positive cells was unchanged in knockout mice (Figure 5, A and C). These results are consistent with the observable increase in MBP and CNPase levels revealed through immunoblotting (Figure 5, D and E). Additionally, the staining levels of GFAP and NeuN were comparable between knockout and control mice (Supplemental Figure S4, A–D).

Next, we assessed the myelin sheath ultrastructure in 28-d-old knockout mice. Knockout mice exhibited increased myelin thickness compared with controls (Figure 5F). These findings are evident from the quantification of the average g-ratios as well as the distribution of the g-ratios (Figure 5, G–I). At postnatal day 28, the average g-ratios were 0.787 ± 0.0890 in knockout mice compared with 0.822 ± 0.0876 in controls. Similar results were observed in 2-mo-old knockout mice (Figure 5J). The average g-ratios were 0.797 ± 0.0609 in knockout mice compared with 0.817 ± 0.0614 in controls (Figure 5, K–M), revealing that the increase in myelin thickness persists into adolescence.

On the basis of these findings, we set out to study whether Rnd2 is also involved in the regulation of later periods of oligodendrocyte myelination in vitro. We infected shRNAs for Rnd2 into oligodendrocytes and cocultured them with neurons for 28 d. Knockdown of Rnd2 in oligodendrocytes increased the amounts of MBP-positive cells and those of Olig2-positive cells expressing MBP at 28 d (Figure 6, A–C), whereas the amounts of NF proteins were comparable. The total number of MBP-positive myelin segments and their lengths were increased (Figure 6, D and E). The number of Olig2-positive cells was comparable between control and Rnd2 knocked-down cocultures (Figure 6F). Consistent with these results from immunofluorescence studies, the expression levels of MBP and CNPase were increased by Rnd2 knockdown (Figure 6, G and H). Additionally, Rnd2 shRNAs were effective in cocultures even at 28 d (Figure 6G).

Together, these findings show that Rnd2 is involved in the inhibition of myelination in later periods of the myelination process.

Rnd2 Tg mice exhibit a phenotype opposite that of Rnd2 knockout mice

To confirm the knockout phenotype of Rnd2, we produced Rnd2 transgenic (Tg) mice. We injected a linearized transgene that contained the loxP-flanked multiple polyA sequences between the CAG promoter and the Rnd2 sequence into fertilized mouse eggs according to the standard method of generating Tg mice (Supplemental Figure S5, A–C) (Miyamoto et al., 2014a). The Tg mice were crossbred with MBP promoter-controlled Cre recombinase Tg mice to generate Rnd2 conditional Tg mice (Supplemental Figure S5, A–C). In the resulting Rnd2; Mbp-Cre+/– Tg mouse line (one copy number of transgene, compared with positive control; tentatively referred to as a Rnd2^Tg mouse in the figure panels), the expression levels of Rnd2 in brains were markedly higher than those in controls (Figure 7A). The expression levels of the myelin marker proteins MBP and CNPase were markedly increased in 14-d-old Rnd2 Tg mice (Figure 7, A–C). Similarly, Rnd2 Tg mice exhibited increased CNPase staining levels compared with littermate controls (Figure 7, D and E). The number of cells positive for Olig2 was unchanged in Tg mice (Figure 7, D and F). Electron microscopic analysis of the corpus callosum revealed increased myelin thickness in 14-d-old Rnd2 Tg mice compared with control mice (Figure 7G). The average g-ratios in 14-d-old Rnd2 Tg mice were 0.830 ± 0.0446, whereas littermate controls had g-ratios of 0.857 ± 0.0376 (Figure 7, H–J).

On the other hand, Rnd2 Tg mice exhibited decreased expression levels of myelin marker proteins in brain tissues at postnatal day 28 (Figure 7, K–M). Myelin thickness in the corpus callosum was also decreased in Rnd2 Tg mice compared with controls (Figure 7, N and O; 0.864 ± 0.0500 in knockout mice and 0.811 ± 0.0786 in controls).

These results concerning the phenotypes in Tg mice support our findings in knockout mice.

Rnd2 regulates myelination signaling through Rho kinase

Rho and Rho kinase signaling regulates myelination by oligodendrocytes. Up-regulation of this signaling is associated with the inhibition of myelination, whereas down-regulation promotes it. As in the case of Rnd3 (Pacary et al., 2011), Rnd2 inhibits the RhoA and Rho kinase signaling (Foster et al., 1996). It has thus been likely that knockout of Rnd2 in mice up-regulates Rho kinase signaling and inhibits myelination at postnatal day 14. We performed immunoblotting of an active phosphorylated Rho kinase and a phosphorylated form of the Rho kinase substrate myosin-binding subunit (Mbs). At 14 d, the phosphorylation levels of Mbs and Rho kinase were increased in Rnd2 knockout mice, although the expression levels of their nonphosphorylated forms were comparable between control and knockout mice (Figure 8, A–C). At 28 d, on the other hand, the phosphorylation levels of Mbs and Rho kinase were decreased in Rnd2 knockout mice, while their expression levels remained comparable between the two genotypes (Figure 8, D–F). In keeping with these results, in Rnd2 Tg mice, the phosphorylation levels of Mbs and Rho kinase were decreased at 14 d, whereas they were increased at 28 d (Figure 8, G–L).

These findings show that, in early developmental periods, Rnd2 can inhibit Rho kinase signaling, possibly inhibiting myelination. In later periods, Rnd2 can promote Rho kinase signaling, possibly promoting myelination.

Rnd2 is involved in myelin maintenance in adulthood

Once we had discerned the role of Rnd2 in myelination during development, we next investigated whether Rnd2 is involved in myelin maintenance during adulthood. We crossbred Rnd2 floxed mice with tamoxifen (TAM)-inducible Plp1 promoter-controlled Cre recombinase transgenic mice (tentatively referred to as Rnd2^fl/fl; Plp1-Cre^ERT mice in the figure panels). Tamoxifen was administrated for 5 d to delete the rnd2 gene in 8- to 9-wk-old Rnd2 knockout and control mice. Mouse tissues were harvested at different time points, and it was confirmed that, at 2 wk after tamoxifen administration, inducible knockout of Rnd2 decreased the intensities of MBP staining in the brain compared with control mice (Figure 9, A and B). The numbers of Olig2-positive cells were comparable between Rnd2 knockout and control mice (Figure 9, A and C). In addition, knockout of Rnd2 decreased the expression levels of MBP as well as those of CNPase (Figure 9, D and E).

Next, we analyzed the myelin sheath ultrastructure by electron microscopy. We observed that Rnd2 knockout mice exhibited decreased myelin thickness in the corpus callosum (Figure 9F). The average g-ratios were 0.817 ± 0.0595 in Rnd2 knockout mice and 0.807 ± 0.0612 in controls (Figure 9, G–I). Together, these findings suggest that Rnd2 can positively regulate short-term myelin maintenance in adulthood.

To further investigate the effect of Rnd2 knockout in myelin maintenance, we performed MBP staining at 4 wk after tamoxifen administration. Knockout of Rnd2 increased MBP intensity in the brain (Figure 10, A and B), whereas the number of Olig2-positive cells was unchanged (Figure 10, A and C). Immunoblotting revealed that the expression levels of MBP and CNPase were greater in Rnd2 knockout mice that in control mice (Figure 10, D and E).

We analyzed the myelin sheath ultrastructure and observed that inducible knockout mice exhibited increased myelin thickness in the corpus callosum compared with controls (Figure 10F). The average g-ratios were 0.775 ± 0.0854 in knockout mice and 0.807 ± 0.0701 in controls (Figure 10, G–I).

Collectively, these results suggest that the effects of Rnd2 on myelination during development persist during myelin maintenance.

DISCUSSION

Oligodendrocyte myelination requires dramatic growth of the plasma membranes such that they form multiple layers around axons. Surprisingly, the collective areas of these differentiated plasma membranes often increase to more than one hundred times the areas of the plasma membranes of the oligodendrocytes before myelination. Myelination requires continuous and dynamic oligodendrocyte morphogenesis, in which a variety of intracellular as well as extracellular signaling molecules are involved in controlling cell morphological changes (Bradl and Lassmann, 2010; Simons and Lyons, 2013; Bergles and Richardson, 2015; Saab and Nave, 2017). It has been proposed that the switching of mechanisms on and off by their signaling molecules is required to achieve the precise morphological changes that are necessary (Bradl and Lassmann, 2010; Simons and Lyons, 2013; Bergles and Richardson, 2015; Saab and Nave, 2017).

Among the small GTPase family proteins controlling the intracellular molecules that switch these mechanisms on and off, increasing evidence indicates that, in addition to the Rho GTPases, including Rnd2, the Ras GTPases also play important roles in oligodendrocyte myelination, acting through the activation of many signal transduction pathways (Bennett et al., 2003). Interestingly, the downstream signaling of Rho GTPases is quite similar to that of Ras GTPases (Devreotes and Horwitz, 2015). For example, in addition to the prototypic Ras proteins such as H-, K-, and N-Ras, R-Ras1/2 have been reported to be critical for oligodendrocyte differentiation and myelination (Sanz-Rodriguez et al., 2018). R-Ras1/2 are expressed in oligodendrocytes at almost equal levels. Mice knocking out both R-Ras1/2 exhibit decreased myelin thickness in the optic nerves. The phenotypes seen in double-knockout mice are presumed to be due to the down-regulation of phosphatidylinositol-3 kinase (PI3K) forming 3-phosphoinositides (Sanz-Rodriguez et al., 2018). Signaling through PI3K stimulates a molecular pathway that is indispensable for myelination and that includes the unique Ras family members Rheb1/2 as well as their regulators such as the tuberous sclerosis complex 1/2 (TSC1/2). Rheb1/2 are involved in the regulation of cell spreading and cell volumes in oligodendrocytes as well as their energy metabolisms. Indeed, deleting Rheb1 in mice results in delayed oligodendrocyte differentiation and decreased myelin formation (Zou et al., 2014). Additionally, the final amounts of 3-phosphoinositides in cells are based on the degree of balance between their phosphorylation and dephosphorylation. Sufficient balance between the activities of PI3K and those of the phosphatases dephosphorylating 3-phosphoinositides is needed to properly trigger myelination. Actually, conditional knockout of phosphatase and the tensin homologue deleted from chromosome 10 (PTEN) in mice typically results in enhanced myelin thickness leading to excessive myelin membrane formation (Goebbels et al., 2010). The presence of regulatory networks underlying Ras and Rho GTPases as well as the related on-and-off signaling mechanisms that operate in myelination during development reminds us of the presence of other small GTPase networks, which have turned out to be more complicated than earlier studies on myelination had indicated (Thurnherr et al., 2006). It is conceivable that molecules belonging to identified or unidentified networks positively and negatively control continuous and dynamic myelination at different developmental periods as needed.

This study is the first to show that Rnd2, a unique Rnd subfamily molecule belonging to the Rho family of small GTPases, acts as one of the regulators of oligodendrocyte myelination in mice. Our findings reveal that Rnd2 has different effects on myelination at different developmental periods. In earlier developmental stages, that is, shortly after birth, oligodendrocyte-specific knockout of Rnd2 has a negative effect on myelin sheath formation without affecting the number of oligodendrocyte lineage cells. The effect of Rnd2 on myelination could in part be explained by the activities of Rho kinase, which is the negative regulator of myelination (Zhao et al., 2007), given that Rnd2 knockout increased the amounts of active phosphorylated forms of Rho kinase. Oligodendrocyte-specific transgenic mice of Rnd2, on the other hand, exhibited enhanced myelin thickness. As we now know, Rnd2 plays a role in inhibiting myelination in later periods, possibly by breaking down excessively thick myelin membranes, as suggested by the fact that Rnd2 knockout and transgenic mice exhibit increased and decreased myelin thickness, respectively. Rnd2 knockout also decreases the phosphorylation levels of Rho kinase (these results are summarized in Table 2 and Figure 11). The variable effects of Rnd2 on myelination depending on developmental period have also been noted in studies knocking out Rnd2 specifically in adult mice. The phenomenon whereby Rnd2 regulates myelination differently in different periods is likely to be reflected in Rnd2’s effects on myelin maintenance in adult mice. It may be of interest that Rnd2 has two roles in myelination during development, one as a negative regulator and one as a positive regulator.

It remains unclear how Rnd2 affects Rho kinase and other possible signaling molecules differentially to regulate myelination. It may be because Rnd2 affects Rho kinase by the on-and-off switching mechanism of RhoA through different effectors at different periods. It is well known that p190RhoGAP A/B, which inactivate RhoA, are the effectors of the Rnd2 homologues Rnd1/3, whose interaction results in antagonizes the activities of RhoA, and in turn, Rho kinase (Wennerberg et al., 2003). Because Rnd2 seems to bind directly to p190RhoGAP proteins in silico (URL: https://thebiogrid.org/), it is possible that Rnd2 inactivates RhoA and Rho kinase through p190RhoGAP proteins, leading to a positive effect on myelination in earlier developmental periods. Additionally, Rnd2 can indirectly activate RhoA through the Rnd2 effector Pragmin (also called pseudopodium-enriched atypical kinase 1-related, kinase-activating pseudokinase 1 [PRAG1], Tanaka et al., 2006). Rnd2’s interaction with Pragmin may have a negative regulatory effect on myelination in later developmental periods; although the expression levels of Rnd2 increase continuously during development, those of Pragmin in the brain remain steady and are not up-regulated from the embryonic stage to adulthood (Tanaka et al., 2006). It is reasonable to presume that one of the molecular mechanisms underlying the differential effects of Rnd2 on myelination is related to differences in the levels of identified or unidentified Rnd2 effector proteins at different developmental periods.

Further studies will promote our understanding not only of the detailed mechanisms by which Rnd2 is differentially involved in the regulation of oligodendrocyte myelination at different developmental periods but also of how Rnd2 regulates myelin maintenance. Such studies will yield information that may elucidate positive and negative regulatory mechanisms triggering oligodendrocyte myelination.

MATERIALS AND METHODS

Request a protocol through Bio-protocol.

Antibodies

The following antibodies were purchased: rabbit polyclonal anti-Rnd2 from Proteintech (Rosemont, IL); mouse monoclonal anti-­myelin basic protein (MBP), and rabbit polyclonal anti-glial fibrillary acidic protein (GFAP) from Covance (Princeton, NJ); mouse monoclonal anti–β-actin from Fujifilm (Tokyo, Japan); goat polyclonal anti–platelet-derived growth factor receptor (PDGFR) α from Santa Cruz Biotechnology (Santa Cruz, CA); mouse monoclonal anti-MBP, mouse monoclonal anti–adenomatus polyposis coli (APC) (also called CC1), mouse monoclonal anti-O4, and mouse monoclonal anti-NeuN from Merck-Millipore (Billerica, MA); rabbit polyclonal anti–phosphorylated Thr696-Mbs and rabbit polyclonal anti-Mbs from Cell Signaling Technology (Danvers, MA); rabbit monoclonal anti-Olig2, rabbit polyclonal anti–phosphorylated Ser1366-Rho kinase, which recognizes active states, and rabbit polyclonal anti-Rho kinase from Abcam (Cambridge, UK); mouse monoclonal anti-2′, 3′-cyclicnucleotide 3′-phosphodiesterase (CNPase) and rabbit polyclonal anti-neurofilament (NF) large subunit from Sigma-Aldrich (St. Louis, MO); peroxidase-conjugated secondary antibodies from MBL (Nagoya, Japan); and fluorescence-labeled secondary antibodies from Thermo Fisher Scientific (Waltham, MA).

Generation of Rnd2 conditional knockout mice

A genomic region containing exons 1–5 of rnd2 was obtained from a C57BL/6 mouse genomic library (Unitech, Chiba, Japan). The targeting vector was constructed to replace the original exons 1–5 with the diphtheria toxin, loxP, the exons 1–5, and Frt-flanked neomycin-resistant sequences. C57BL/6 embryonic stem (ES) cells were transfected with the linearized targeting vector by electroporation. G418-resistant, homologous recombination, positive transfectants were isolated, and their genomic sequences were confirmed. These ES cells were used for generating chimeric mice.

First, to identify the neo gene containing the Rnd2 knockout cassette, offspring were genotyped by genomic PCR. To evaluate the knockout cassette allele, the primers used were 5′-GCCGGCGCGTAGCCGGGACC-3′ (sense) and 5′-CGAACACCTGCAGCAACGCTGTCTTG-3′ (antisense) for the knockout cassette long-arm position and 5′-CGTGCAATCCATCTTGTTCAAT-3′ (sense) and 5′-TGAACCTGCTTTGTATCAGAAGTGC-3′ (antisense) for the knockout cassette short-arm position. In the former primer pair, the allele-harboring knockout cassette displayed ∼160 base pairs. In the later primer pair, the allele-harboring knockout cassette displayed ∼3500 base pairs, enabling us to obtain 11 chimeric mice.

Next, to remove the neo gene, offspring were mated with Unitech Flp recombinase transgenic mice (Unitech). To evaluate the neo gene deletion allele, the primers used were 5′-AAGTGGGAG­CTGTGTCCTACGTT-3′ (sense) and 5′-GAGGTCAAAAAGTCC­AGGACGAC-3′ (antisense) for the long-arm position and 5′-­GAACA­AGATGGATTGCACGCAGGTTCTCCG-3′ (sense) and 5′-GTAGCCAACGCTATGTCCTGATAG-3′ (antisense) for the neo gene. In the former primer pair, the knockout allele with the neo gene, the knockout allele without the neo gene (also called the floxed allele), and the wild-type allele displayed ∼2020, ∼470, and ∼310 base pairs, respectively. In the later primer pair, the knockout allele with the neo gene displayed ∼670 base pairs, enabling us to obtain a single kind of ES cell–derived 7 hemizygotic floxed knockout mice.

Finally, to remove the flp gene, these mice were further mated with wild-type C57BL/6 mice. The primers used were 5′-TAGTTTGCAATTACAGTTCGAATCA-3′ (sense) and 5′-AGCCTTGTTGTACGATCTGACTAAG-3′ (antisense) for the flp gene, whose removed allele displayed ∼500 base pairs.

The primers used for Rnd2 conditional knockout mouse genotyping were 5′-AAGTGGGAGCTGTGTCCTACGTT-3′ (sense) and 5′-GAGGTCAAAAAGTCCAGGACGAC-3′ (antisense). The floxed and the wild-type alleles displayed ∼470 and ∼310 base pairs, respectively. Southern blotting, using the 5′ or 3′ region of exons of the gene encoding Rnd2 as the specific probe, confirmed that the floxed and wild-type alleles were ∼7.5 and ∼9.8 kbp for the 5′ probe or ∼7.6 and ∼3.7 kbp for the 3′ probe, respectively.

To delete the gene specifically in oligodendrocytes, Rnd2 conditional knockout mice were mated with MBP promoter– Cre recombinase transgenic mice (RBRC No. 01461; Riken BioResource Center, Ibaraki, Japan) or tamoxifen-inducible PLP1 promoter–driven Cre recombinase transgenic mice (JAX No. 005975; The Jackson Laboratory, Bar Harbor, ME). These transgenic mice were mated with wild-type C57BL/6 mice for more than 12 generations to achieve the same background as the other genetically modified mice. Genomic PCRs were performed according to established standard protocols (Riken BioResource Center or The Jackson Laboratory). The primer pair for Cre recombinase transgenic mice was 5′-CCACCACCTCTCCATTGCAC-3′ (sense) and 5′-ATGTTTAGCTGGCCCAAATG-3′ (antisense), displaying ∼500 base pairs.

To show that Cre was active in MBP-expressing tissues, we performed PCR amplification using the mouse corpus callosum genome. The primer pair was 5′-CTTTCCACTCAGTACTTGGCTAGAG-3′ (sense) and 5′-AGAGAAGGGATAAGGGTGTAGAGAA-3′ (antisense). The floxed and deleted alleles displayed ∼3.9 and ∼1.0 kbp, respectively.

Mouse tail genetic PCR amplification (KAPA mouse genotyping kit; NIPPON Genetics Co., Ltd., Tokyo, Japan) was routinely performed in 35 cycles, with each cycle consisting of denaturation at 94°C for 10 or 15 s, annealing at 58 to 68°C (depending on each primer’s T_m value) for 15 s, and extension at 68 to 72°C for 15 or 30 s (depending on product length).

When it was possible to determine their sex, male mice were used for experiments. All homozygous mice, as well as their heterozygous mice, were fertile under standard breeding conditions.

Generation of Rnd2 conditional transgenic mice

An ∼4.5 kbp DNA fragment containing the CAG promoter, loxP, triple SV40 polyAs, loxP, human Rnd2, and single SV40 polyA sequences (Vector Builder, Chicago, IL) was digested from the backbone vector VB180809 with ApaLI restriction enzyme and injected into fertilized BDF1 oocytes (Miyamoto et al., 2014a). Transgenic founder mice and established transgenic mice were identified by both genomic PCR (840 base pairs of length) using tail DNA with the specific primer pair (5′-CAGGTGTTCGCCAAGGACGCCTATC-3′ [sense] and 5′-CCT­CCCACACCTCCCCCTGAAC-3′ [antisense] for Rnd2) and by Southern blotting with EcoRI-digested tail DNA hybridized to a radioisotope-labeled probe for the internal Rnd2 sequence of the transgene (∼700 base pairs of hybridized band). Genomic PCR was performed in 35 cycles, each consisting of denaturation at 94°C for 15 s, annealing at 63°C for 15 s, and extension at 72°C for 15 s. We succeeded in generating five founder mice, each harboring one or more transgenes, which were mated to wild-type C57BL/6J mice. The transgenes in two transgenic lines were stably maintained for at least three generations. Among them, one line was used in the following experiments. The transgenic mice were fertile and engaged in normal behavior under standard breeding conditions. Male mice were used for experiments when their gender was distinguishable.

To induce the gene specifically in oligodendrocytes, Rnd2 conditional transgenic mice were mated with MBP promoter–driven Cre recombinase transgenic mice. The primer pair for genomic PCR was 5′-CCACCACCTCTCCATTGCAC-3′ (sense) and 5′-ATGTTTAGCTGGCCCAAATG-3′ (antisense), displaying ∼500 base pairs (The Jackson Laboratory).

Tamoxifen treatment in vivo

Tamoxifen (TAM; Sigma-Aldrich) was dissolved in a mixture of 90% corn oil and 10% ethanol at a concentration of 10 mg/ml. Eight- to nine-week-old mice were given a daily intraperitoneal injection of 1 mg (100 μl) TAM for 5 consecutive days (Miyamoto et al., 2016). The first day of injections was designated as day 0.

Preparation of plasmids containing short-hairpin RNAs

All nucleotide sequences were confirmed by sequencing. The sense and antisense oligonucleotides, which were used for duplex production, were as follows. Rnd2#1 (starting from nucleotide 339 of rat Rnd2) shRNA is composed of 5′-GATCCGGTGGTGTTGGTTGGTTGTCTGTGAAGCCACAGATGGGACAGCCAACCAGCACCACCTTTTTTACGCGTAT-3′ (sense) and 5′-CGATACGCGTAAAAAAGGTGGTGCTGGTTGGCTGTCCCATCTGTGGCTTCACAGACAACCAACCAACACCACCG-3′ (antisense); Rnd2#2 (starting from nucleotide 405 of rat Rnd2) shRNA is composed of 5′-GATCCGTAGAGGCTTATCTCCGTCCTGTGAA­GCCACAGATGGGGACGGGGATAAGTCTCTGCTTTTTTACGCGTAT-3′ (sense) and 5′-CGATACGCGTAAAAAAGCAGAGACTTATCCCCGTCCCCATCTGTGGCTTCACAGGACGGAGATAAGCGTCTACG-3′; and the control Photinus pyralis luciferase shRNA is composed of 5′-GATCCGGCCATTCTATCCTCTAGACTGTGAAGCCACAGATGGGTCTAGAGGATAGAATGGCCTTTTTTACGCGTAT-3′ (sense) and 5′-CGATACGCGTAAAAAA GGCCATTCTATCCTCTAGACCCATCTGTGGCTTCACAGTCTAGAGGATAGAATGGCCG-3′ (antisense). The annealed duplexes were ligated into the retrovirus vector, pSINmU6 (Takara Bio, Shiga, Japan).

Primary oligodendrocyte cultures

Oligodendrocyte precursor cells were isolated from embryonic day 15 (E15) Sprague Dawley rats (Miyamoto et al., 2007, 2008, 2014b). Briefly, cerebral cortices were dissected, dissociated with 0.25% trypsin, triturated, and passed through mesh with 70 μm pores. Cells were collected, resuspended in MEM containing 10% fetal bovine serum, 50 U/ml penicillin, and 50 μg/ml streptomycin, and seeded on poly-l-lysine–coated dishes. After two passages, the cells were cultured on noncoated Petri dishes (Thermo Scientific). On the second day of culture, the medium was changed to a DMEM-based serum-free growth medium containing the growth factors PDGF and FGF (PDGF-AA and bFGF; Peprotech, Rocky Hill, NJ) and N2 (Life Technologies), and cells were cultured for an additional 2 d. These cells were then used as oligodendrocyte precursor cells (a proliferating state before differentiation). The isolated cells indeed possessed the oligodendrocyte precursor cell marker PDGF receptor α (GEO accession No. GSE114957; Miyamoto et al., 2016). They also had the oligodendrocyte lineage cell markers Olig1 and Olig2 but not the astrocyte markers Nfia, Sox9, Gfap, and Aqp4 (see Table 1). To induce their differentiation, cells were cultured with a differentiation medium containing 20 ng/ml triodothyronine and 20 ng/ml thyroxine but no growth factor. After 3 d, these cells gradually differentiated into mature oligodendrocytes with MBP-positive myelin web-like structures with myelin markers (see Table 1). To confirm the viability of the attached cells, cells were stained with 0.4% trypan blue. Trypan blue–incorporating cells numbered fewer than 5% in each experiment in this study.

Production of retroviruses and retrovirus-mediated DNA transfection

Using a CalPhos Transfection kit (Takara Bio), each retrovirus vector was cotransfected into G3Thi cells with vectors pVSVG and pGP, which were provided with the Takara Bio retroviral production kit. After 2 d, each culture supernatant was centrifuged at 10,000 rpm for 8 h to concentrate the recombinant retroviruses. The virus pellets were suspended in each culture medium. Because retroviruses are infected into proliferating cells, they were used for transfection (infection) into growing, primary oligodendrocyte precursor cells before being cocultured with neurons (Miyamoto et al., 2008, 2014b).

Oligodendrocyte-neuronal cocultures

DRG neurons, whose axons are generally connected from peripheral tissues to the CNS via the spinal cord, were isolated from E15 Sprague Dawley rats and then dissociated and plated onto collagen (type I)-coated 22-mm coverslips (Chan et al., 2004; Yamauchi et al., 2012; Miyamoto et al., 2013). Nonneuronal growing cells were eliminated by cycling three times with medium containing 5-fluorodeoxyuridine and uridine. Cocultures were established by seeding ∼200,000 purified oligodendrocyte precursor cells on neurons. Cocultures were maintained for 3–4 wk. Medium was replaced every 3–4 d.

Immunoblotting

Tissues or cells were lysed in lysis buffer (50 mM HEPES-NaOH [pH 7.5], 20 mM MgCl₂, 150 mM NaCl, 1 mM dithiothreitol, 1 mM phenylmethane sulfonylfluoride, 1 μg/ml leupeptin, 1 mM EDTA, 1 mM Na₃VO₄, 10 mM NaF, 0.5% NP-40, 1% CHAPS, and 0.1% or 0.3% SDS) to elute myelin segment proteins and were centrifuged in a microcentrifuge to obtain clear supernatants (Chan et al., 2004; Yamauchi et al., 2012; Miyamoto et al., 2013, 2018). These lysates (10 μg per sample on average) were denatured and subjected into SDS–polyacrylamide gels. The electrophoretically separated proteins were transferred to a polyvinylidene difluoride membrane, blocked with Blocking One reagent (Nacalai Tesque, Kyoto, Japan), and immunoblotted first with primary antibodies and then with peroxidase-conjugated secondary antibodies. The bound antibodies were detected using a normal- or high-sensitivity chemiluminescence detection kit (FUJIFILM Wako Pure Chemical Corporation, Tokyo, Japan). Littermate mice (genetically modified mice and control mice) were harvested on different days, and at least three sets were prepared for experiments. All photos in the figures are representative of multiple experiments.

Immunofluorescence

Cells on coverslips or dishes were normally fixed with 4% paraformaldehyde (PFA). Four percent PFA followed by 100% cold methanol was used for myelin segment staining (Chan et al., 2004; Yamauchi et al., 2012; Miyamoto et al., 2013, 2018). Cells were permeabilized with phosphate-buffered saline (PBS) containing 0.1% Triton X-100, blocked with Blocking One reagent, and incubated first with primary antibodies and then with fluorescence-­labeled secondary antibodies. The coverslips or dishes were mounted with Vectashield reagent with 4’,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA). The fluorescence images were captured with a DMI4000B fluorescence microscope system (Leica, Wetzlar, Germany) and analyzed with AF6000 software (Leica). Culture dishes (Rnd2 shRNA and control shRNA) were harvested on different days, and at least three sets were prepared for experiments. All photos in the figures are representative of multiple experiments.

Immunohistochemistry

Tissues were perfused first with PBS and then with PBS containing 4% PFA. The tissues were postfixed with 4% PFA, replaced with 20% sucrose, and embedded in Tissue-Tek reagent (Sakura Finetechnical, Tokyo, Japan) (Chan et al., 2004; Yamauchi et al., 2012; Miyamoto et al., 2013, 2018). The cryosections for immunostaining with an antibody for O4 were prepared according to the manufacturer’s instructions (Merck-Millipore). Microtome sections on glass slides were blocked with Blocking One reagent and incubated first with primary antibodies and then with fluorescence-labeled secondary antibodies. The glass slides were mounted with Vectashield reagent. The fluorescent images were collected with an IX81 microscope system (Olympus, Tokyo, Japan) equipped with a laser-scanning FV500 or FV1000D (Olympus) and analyzed with Fluoview software (Olympus). Littermate mice (genetically modified mice and control mice) were harvested on different days, and at least three sets were prepared for experiments. All photos in the figures are representative of multiple experiments.

Electron microscopy

Tissues were fixed with 2% PFA and 2% glutaraldehyde in 0.1% cacodylate-containing buffer. They were then contrasted with 2% osmium tetroxide, dehydrated with an ethanol gradient, and treated with propylene oxide. Finally, samples were infiltrated and embedded in pure epoxy resin. Ultrathin sections were stained with uranyl acetate and lead staining solution. Images were taken with a JEM-1200EX or JEM-1400 Plus electron microscope system (JOEL, Tokyo, Japan) by the Hana-ichi Ultrastructure Research Institute (Nagoya, Japan). The g-ratio is the ratio of the diameter of the axon proper (i.e., not including any myelin) to the outer diameter of the myelinated axon. These ratios were plotted to determine axon diameter. Their distributions are also shown in the graph.

Transcriptome analysis

We previously registered data for primary rat oligodendroglial precursor cell mRNAs before (day 0) and after (day 3) the induction of differentiation (GEO Acc. No. GSE114957; Miyamoto et al., 2016). Briefly, total RNA was labeled with Cy5 using an Amino Allyl MessageAMP II aRNA Amplification Kit (Thermo Fisher Scientific). A 3D-Gene chip (rat oligo chip 20k; Toray, Tokyo, Japan) was used for microarray analysis by the Toray 3D-Gene chip service (URL: http://www.3d-gene.com/). The signals, which were hybridized to Cy5-­labeled RNA pools, were obtained using a Toray 3D-Gene scanner. The detected signals for the respective genes were normalized according to the global normalization method and were processed by the Kamakura-Techno 3D-Gene analytical service (URL: http://www.kamakura-ts.co.jp/). The median value of the detected signal intensities was adjusted to 25. Values greater than 100 were considered valid.

Statistical analysis

Values shown in the figure panels represent the means ± SD from separate experiments. Comparisons between two experimental groups were made using Student’s t test (**, p< 0.01, *, p< 0.05). A one-way analysis of variance (ANOVA) was followed by a Fisher’s protected least significant difference (PLSD) test as a post hoc comparison (**, p < 0.01, *, p < 0.05).

Animal studies

Genetically modified/unmodified mice were handled in accordance with a protocol approved by the Japanese National Research Institute for Child Health and Development Animal Care Committee. Gene recombination processes were handled in accordance with a protocol approved by the Japanese National Research Institute for Child Health and Development Gene Recombination Committee.

FOOTNOTES

ACKNOWLEDGMENTS