Myelin Transcription Factor 1 (Myt1) Expression inDemyelinated Lesions of Rodent and Human CNSADAM C. VANA,1,2 CLAUDIA F. LUCCHINETTI,3 TUAN Q. LE,2 AND REGINA C. ARMSTRONG1,2*1Program in Neuroscience, Uniformed Services University of the Health Sciences, Bethesda, Maryland2Department of Anatomy, Physiology, and Genetics, Uniformed Services University of the Health Sciences, Bethesda, Maryland3Department of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota

KEY WORDSremyelination; oligodendroglia; oligodendrocyte progenitor;myelin; multiple sclerosis

ABSTRACTMyelin transcription factor 1 (Myt1) is a zinc-finger DNAbinding protein that influences developing oligodendrocyteprogenitor (OP) cell proliferation, differentiation, and mye-lin gene transcription in vitro. The potential of Myt1 toplay a role in OP responses leading to remyelination wasexamined using murine hepatitis virus strain A59 (MHV)to induce spinal cord demyelination and potential relevanceto human pathology was evaluated in multiple sclerosis(MS) lesions. In MHV-infected mice, the density of Myt1expressing cells markedly increased in lesioned areas of spi-nal cord white matter. Myt1 expressing cells proliferatedmost extensively during active demyelination and subse-quently accumulated to maximal levels during early remye-lination. Cells with nuclear Myt1 immunoreactivity weremainly OP cells, identified by co-localization with platelet-derived growth factor alpha receptor, with additional phe-notypes being either oligodendrocytes or neural stem cells,identified by CC1 antigen and Musashi1, respectively. Thedensity of OP cells expressing Myt1 was significantlyincreased in white matter of MHV-infected mice during de-myelination and early remyelination then as remyelinationadvanced the values returned to levels comparable to PBS-injected control mice. In MHV lesions, Myt1 was notexpressed in astrocytes, lymphocytes, or macrophage/micro-glial cells. MS lesions demonstrated increased Myt1 expres-sion in both the periplaque white matter adjacent to lesionsand within early remyelinating lesions. These resultssuggest a potential role for Myt1 in the regeneration of oli-godendrocyte lineage cells in response to demyelination.VVC 2007 Wiley-Liss, Inc.

INTRODUCTION

Demyelination and oligodendrocyte loss in the CNScan result from multiple sclerosis (MS), spinal cordinjury, toxic insults, and neurodegenerative disorders.Demyelination leads to slowing or blockade of actionpotential conduction. In addition, without the protectiveeffect of myelin, denuded axons are more vulnerable totransection that can lead to permanent neurologicalimpairment (Ferguson et al., 1997; Trapp et al., 1998).

Spontaneous remyelination has been observed in MSlesions following initial episodes of demyelination, yetthe capacity for remyelination becomes limited with sub-

sequent prolonged or repeated episodes of demyelination(Bruck et al., 2003). Animal models have shown thatremyelination requires oligodendrocyte progenitor (OP)cells in the proximity of demyelinated lesions to prolifer-ate and then differentiate into mature oligodendrocytesthat can then myelinate denuded axons (Reynolds et al.,2001; Watanabe et al., 2002). This sequence of cellularresponses must be orchestrated through transcriptionalcontrol of the complex gene expression patterns that reg-ulate OP proliferation, differentiation, and eventual syn-thesis of myelin components.

Several previous studies implicate myelin transcrip-tion factor 1 (Myt1) as potentially important in regulat-ing OP responses in the developing and adult CNS.Myt1 is a zinc-finger DNA binding protein that was ori-ginally identified based upon binding affinity within thepromoter region of the proteolipid (PLP) gene, the mostabundantly transcribed CNS myelin gene (Kim andHudson, 1992). During development of the oligodendro-cyte lineage, Myt1 is localized within nuclei of immaturecells and then downregulated after terminal differentia-tion and accumulation of myelin proteins in mature oli-godendrocytes (Armstrong et al., 1995). Myt1 continuesto be expressed in germinal zones of the adult CNS andis upregulated in gliomas (Armstrong et al., 1997) andfollowing spinal cord traumatic injury (Wrathall et al.,1998). In vitro, expression of a dominant negative formof Myt1 showed that Myt1 can regulate a critical transi-tion in oligodendrocyte lineage cell development by mod-ulating OP proliferation relative to terminal differentia-tion and upregulation of myelin gene transcription(Nielsen et al., 2004).

To examine the potential for Myt1 to play a role inregulating the OP responses required for remyelination,the current study determines the expression of Myt1relative to the proliferative status and phenotype of cellsin white matter during the progression of experimentaldemyelination, followed by spontaneous remyelination.

Grant sponsor: USUHS; Grant numbers: RO70PC; Grant sponsor: NationalInstitutes of Health; Grant numbers: NS39293National Multiple Sclerosis Society;Grant numbers: RG3515 and RG 3185-A2, B3.

*Correspondence to: Regina C. Armstrong, Ph.D., Department of Anatomy, Phys-iology, and Genetics, Uniformed Services University of the Health Sciences, 4301Jones Bridge Rd, Bethesda, MD 20814, USA. E-mail: rarmstrong@usuhs.mil

Received 28 August 2006; Revised 5 January 2007; Accepted 10 January 2007

DOI 10.1002/glia.20492

Published online 1 March 2007 in Wiley InterScience (www.interscience.wiley.com).

GLIA 55:687–697 (2007)

VVC 2007 Wiley-Liss, Inc.

Murine hepatitis virus strain A59 (MHV) infection wasused to induce focal demyelinated lesions of spinal cordwhite matter from lytic infection of oligodendrocyteswith subsequent lymphocytic infiltration and virus clear-ance (Matthews et al., 2002; Redwine and Armstrong,1998). Furthermore, MS lesions were examined for Myt1immunoreactivity relative to stage of disease pathologyand control tissues to predict the potential for Myt1 tohave a role in human demyelinating disease.

MATERIALS AND METHODSAnimals

C57Bl/6 mice (Jackson Laboratories, Bar Harbor, ME)were maintained in the Uniformed Services Universityof the Health Sciences (USUHS) animal housing facilityin accordance with guidelines of the National Institutesof Health and the USUHS Institutional Animal Careand Use Committee.

MHV Model of Spinal Cord Demyelination

As previously described (Armstrong et al., 2005; Red-wine and Armstrong, 1998), 4-week-old female C57Bl/6mice were anesthetized and injected intracranially withMHV diluted in sterile PBS to 1,000 plaque formingunits per 10 lL injection volume. Control animals wereinjected with 10 lL of sterile PBS.

A hang time assay was used to quantify motor impair-ment and recovery of function that corresponds with his-topathological findings of demyelination and remyelina-tion (Armstrong et al., 2005; Frost et al., 2003; Redwineand Armstrong, 1998). Duration of hanging from a cagetop was recorded on days 0, 7, 10, 14, 17, 21, 24, 28, andthen every 7 days for up to 8 weeks post infection (wpi).On the same days a clinical score was assigned as fol-lows: 0 for no evidence of paresis/paralysis, 1–5 for pare-sis/paralysis in 1–5 appendages, and 6 for morbidity(Armstrong et al., 2005; Redwine and Armstrong, 1998).Each mouse used in the study had a clinical score of 2or greater, indicating impaired limb movement orstrength, and hang time values characteristic of severedisability, followed by spontaneous recovery (Frost et al.,2003; Redwine and Armstrong, 1998).

At 2, 4, and 8 wpi, mice were perfused with 4% para-formaldehyde (Sigma, St. Louis, MO). Spinal cords werepostfixed in 4% paraformaldehyde at 4�C overnight andcryoprotected prior to cutting as 15 lm transverse sec-tions and mounting onto Superfrost Plus slides (Fisher,Pittsburgh, PA).

In Situ Hybridization of Mouse Tissue Sections

In situ hybridization and preparation of digoxigenin-labeled riboprobes were performed with modifications ofmethods previously detailed (Redwine and Armstrong,1998). The antisense riboprobe to detect Myt1 mRNA

transcripts was synthesized from a cDNA template (giftfrom Dr. Lynn Hudson; National Institutes of Health) byin vitro transcription to be complementary to a 1.2 kbmRNA fragment encoding the 30 untranslated region ofmurine Myt1 (Kim et al., 1997). Spinal cord cryosectionswere digested with 20 U/mL proteinase K (Sigma, St.Louis, MO) at 37�C for 40 min prior to hybridization at53�C overnight. Digoxigenin was detected with an alka-line phosphatase-conjugated sheep anti-digoxigenin anti-body (Roche, Indianapolis, IN), followed by reaction withNBT/BCIP overnight (DAKO, Carpinteria, CA).

BrdU Incorporation and Detection in Mice

At 4 and 2 h prior to perfusion, mice were injected in-traperitoneally with 200 mg/kg bromodeoxyuridine(BrdU) (Sigma, St. Louis, MO)(Armstrong et al., 2002).Following in situ hybridization detection of Myt1 tran-scripts, sections were immunostained with anti-BrdUantibody directly conjugated with horseradish peroxi-dase (Roche, Indianapolis, IN) and detected with DAB(Vector Labs, Burlingame, CA).

Immunohistochemistry of Mouse Tissue Sections

Mouse transverse spinal cord tissue sections were im-munostained with aMyt1-His rabbit polyclonal antibody(1:100; gift from Dr. Lynn Hudson; National Institutes ofHealth), which has been characterized in a previousstudy (Armstrong et al., 1995). OP cells were immunola-beled for platelet-derived growth factor alpha receptor(PDGFaR) antibody (APA5, 1:200; BD Biosciences Phar-mingen, San Diego, CA) and mature oligodendrocyteswere labeled with CC1 monoclonal antibody (1:20; Onco-gene Research Products, Cambridge, MA). Neural stemcells were identified with a monoclonal antibody againstMusashi1 (Msi1, 1:2,000; gift from Dr. Hideyuki Okano;Keio University School of Medicine, Tokyo, Japan;Kaneko et al., 2000). Astrocytes were immunostained forglial fibrillary acidic protein (GFAP, 1:20; Roche, India-napolis, IN). Immune cells were identified with Ox-52pan T-cell antibody (1:1,000) and CD45R antibody for B-cells (1:25) (both from BD Biosciences Pharmingen, SanDiego, CA), and Mac 1 antibody for macrophages/micro-glia (1:100; Roche, Indianapolis, IN).

All secondary antibodies were purchased from JacksonImmunoresearch (West Grove, PA). PDGFaR, Msi1,Ox52, and Mac1 signals were amplified with fluoresceintyramide signal amplification (Perkin Elmer LifeSciences, Boston, MA). Sections were incubated withDAPI (Sigma, St. Louis, MO) to stain nuclei.

Quantitative Analysis in Mouse Tissue Sections

Areas within transverse spinal cord sections weremeasured using Spot2 software and all labeled cellswithin each area were counted with a 403 objective to

688 VANA ET AL.

GLIA DOI 10.1002/glia

determine cells per mm2. For each condition and timepoint examined, at least three mice were analyzed withat least three sections included for each mouse. To distin-guish significant differences between groups, unpairedStudent’s t-tests were used for single time point compari-sons and one-way analysis of variance (ANOVA withTukey’s post hoc test) was used between control mice andmultiple MHV disease time points.

MS and Control Tissues

The pathological analysis of MS autopsy and biopsysamples was approved by the Mayo Clinic InstitutionalReview Board (IRB No. 2067-99). Eight cases wereselected from a larger series of MS cases with detailedclinical follow up (n 5 130) derived from the MS LesionProject cohort (Lucchinetti; NMSS grant RG 3185-A2;B3). Control tissue consisted of adult human brain sub-cortical white matter areas chosen from autopsy caseswith no known neurological involvement. Two autopsycontrol cases were obtained from the Brain and TissueBank for Developmental Disorders (Baltimore, MD) andone was obtained from the Mayo Clinic.

Histopathology and Staging of DemyelinatingActivity in MS Tissues

Specimens were fixed in 4% paraformaldehyde andembedded in paraffin. Sections 4 lm thick were stainedwith haematoxylin and eosin, Luxol-fast blue, and peri-odic acid-Schiff as well as Bielschowsky’s silver impreg-nation. Immunohistochemical staining was performedwith a biotin-avidin or an alkaline phosphatase/anti-alkaline phosphatase technique. The primary antibodiesused were anti-myelin basic protein (MBP, BoehringerMannheim, Germany), anti-CD3 (T cells, DAKO, Den-mark), anti-CD8 (cytotoxic T cells, DAKO, Denmark),anti-CD20 (B cells, DAKO, Denmark), anti-KiM1P (mac-rophages/microglial cells, Dr. Radzun, University ofG€ottingen, Germany), and anti-MRP 14 (early activatedmacrophages, BMA Biomedicals, Switzerland). Thedemyelinating activity was classified according to thepresence of minor myelin proteins (MOG, CNPase,MAG) in macrophages (early active lesions), the pre-sence of early remyelination, and the absence of bothactive demyelination and remyelination (inactive demye-linated) (Br€uck et al., 1995). As white matter lesionsmight show more than one of these features, a totalnumber of 8 different lesion areas were investigated (3early active; 3 inactive; and 2 remyelinated).

Myt1 Immunohistochemistry in Human Tissues

Immunostaining with aMyt1-His rabbit polyclonalantibody (1:50; see above) was detected using the EliteABC kit (Vector Labs; Burlingame, CA) with DAB as thesubstrate. The specificity of aMyt1-His immunoreactiv-

ity in human tissues was characterized by pre-absorp-tion with excess antigen (Armstrong et al., 1997). As afurther demonstration of specificity of the immunostain-ing reaction, selected tissues were similarly immuno-stained with the aMyt1L-His antibody that recognizesMyt1L, a Myt1 family member that is expressed in neu-ronal populations but not in oligodendrocyte lineagecells (Kim et al., 1998). Sections were counterstainedwith eosin.

In Situ Hybridization in Human Tissues

As previously described (Wong et al., 2000), sectionsprocessed for in situ hybridization with digoxigenin-labeled riboprobes to detect mRNA transcripts. Antisenseriboprobes for human Myt1 (Wrathall et al., 1998) andPLP (Hudson et al., 1987) were generated from plasmidsprovided by Dr. Lynn Hudson (National Institutes ofHealth, Bethesda, MD). For combining detection of PLPmRNA and Myt1 protein, sections were processed forPLP mRNA in situ hybridization, followed by immuno-staining for Myt1, as described above, without eosincounterstaining.

Human Tissue Analysis

All human tissue sections were analyzed qualitativelyand tissues immunostained for Myt1 that exhibited spe-cific and consistent immunoreactivity were also assessedquantitatively. Cells were counted using a 1003 oilobjective and 10 fields were sampled within each tissuearea (i.e. active, remyelinated, or inactive lesion, adja-cent periplaque white matter—PPWM, normal whitematter—NWM). Characteristics of cell morphology andanatomic localization were used to exclude cells thatwere expected to be distinct from oligodendrocyte line-age cells, such as lymphocytes, based on comparisonswith immunostaining for myelin proteins.

Imaging of Mouse and Human Tissue Sections

All images were acquired with a Spot 2 digital cameramounted to an IX-70 microscope (Olympus, Melville,NY) with narrow band pass and triple band pass filtersfor fluorescence detection and co-localization studies.Images were prepared as panels using Adobe Photoshop(Mountain View, CA).

RESULTSMyt1 Expressing Cells Proliferate in Lesions

During Demyelination and Early Remyelination

MHV infection in C57Bl/6 mice induces focal areas ofdemyelination throughout the spinal cord, with subse-quent spontaneous remyelination (Frost et al., 2003;Redwine and Armstrong, 1998). To determine whether

689Myt1 EXPRESSION IN RESPONSE TO CNS DEMYELINATION

GLIA DOI 10.1002/glia

Myt1 expression was upregulated in correlation with aspecific stage of disease progression, cellular expressionof Myt1 mRNA transcripts was detected by in situhybridization in spinal cord sections (Fig. 1). In MHV-infected mice, the density of Myt1 expressing cells inthe white matter was increased during the demyelin-ation phase examined at 2 wpi, as compared with PBS-injected controls. This Myt1 expression peaked duringthe active remyelination phase, which correspondedwith 4 wpi in the MHV-infected mice. By 8 wpi, asremyelination progressed in MHV-infected mice the den-sity of cells expressing Myt1 declined to near that ofPBS-injected controls.

In models of experimental demyelination, proliferationof OP cells precedes the generation of new oligodendro-cytes and remyelination (Reynolds et al., 2001; Wata-nabe et al., 2002). To evaluate a potential role of Myt1in proliferation in response to demyelination, in situhybridization for Myt1 was combined with incorporationof BrdU (Fig. 1). Mice were administered BrdU duringthe final 4 h before sacrifice, which allowed detection ofa sufficient proportion of cells undergoing DNA synthe-sis while minimizing the potential for differentiationsubsequent to BrdU incorporation. The density of cellsdouble labeled for Myt1 mRNA combined with BrdUimmunohistochemistry is significantly increased duringthe active demyelination phase at 2 wpi (Fig. 1B). By 4wpi, the density of Myt1 and BrdU double labeled cellsdeclined slightly yet remained significantly increasedover PBS-injected controls (Fig. 1B). Incorporation ofBrdU into cells without detectable Myt1 (Fig. 1B) maycorrespond with proliferation of astrocytes in MHVlesions (see below; Redwine and Armstrong, 1998). By 8wpi, BrdU incorporation in sections from MHV-infectedmice has declined to the levels observed in sections fromPBS-injected control mice (Fig. 1B).

Amplification of Myt1 expressing cells is specificallyassociated with MHV lesion areas (Fig. 1F). Lesions inspinal cord white matter were identified using dark fieldand phase contrast microscopy to image areas of myelinloss and vacuolation (not shown; Armstrong, 2000; Red-wine and Armstrong, 1998). The overall density of cellsexpressing Myt1 mRNA is greatly increased withinlesions. In contrast, normal appearing white matter(NAWM) of MHV-infected mice had a relatively low den-sity of cells expressing Myt1, which was similar tovalues from the white matter of PBS-injected controls.

Myt1 is Localized to the Nuclei of Immature CellTypes in MHV Lesions

Immunohistochemistry for Myt1 demonstrated thatMyt1 protein was increased in cells associated withwhite matter lesions of MHV-infected mice during demye-lination (2 wpi) and early remyelination (4 wpi) (Fig. 2).Double immunolabeling for cell type-specific antigenswas used to characterize cell types with Myt1 immunor-eactivity (Figs. 2, 3, and 4). The nuclear pattern of Myt1immunoreactivity was confirmed by co-localization with

DAPI staining of nuclei (data not shown). As expectedduring demyelination and early remyelination in MHV-infected mice, the density of oligodendrocytes was signif-icantly lower in lesioned areas (Fig. 2B) and corre-sponded with a reduction in the density of CC11 oligo-dendrocytes in the total white matter area, as comparedwith PBS controls (Fig. 3C). Among the mature oligo-dendrocytes immunolabeled with CC1 only a small pro-portion exhibited Myt1 immunoreactivity (Figs. 3C and4). The small proportion of Myt11 CC11 double labeledcells is consistent with our previous study showing thatduring CNS development Myt1 immunoreactivity isdownregulated after OP cells undergo terminal differen-tiation into mature oligodendrocytes (Armstrong et al.,1995).

Our results correlating Myt1 transcript expressionwith a proliferative population of cells in MHV lesions(Fig. 1) indicated that immature cell types may be corre-lated with Myt1 expression in MHV lesions. Indeed, inspinal cord white matter, the majority of OP cells identi-fied by immunolabeling for PDGFaR; (Fig. 2D) exhibitednuclear Myt1 immunoreactivity (Fig. 3B; 85.2% in PBSWM; 77.4% in MHV WM). The density of cells doubleimmunolabeled for Myt1 and PDGFaR was significantlyincreased in sections from MHV mice at 2 and 4 wpi, ascompared to PBS-injected control mice (Fig. 4).

A substantial population of cells expressing Myt1 didnot appear to be OP cells (i.e. Myt11 PDGFaR2) yetcould not be accounted for by the relatively few Myt11cells that had acquired CC1 immunolabeling as oligoden-drocytes (i.e. Myt11 CC11) (Fig. 3). Therefore, expres-sion of Myt1 at a more immature lineage stage wasexamined. Immunolabeling for Musashi1 (Msi1) wasused as a marker of adult neural stem cells prior to theOP cell stage (Kaneko et al., 2000; Pincus et al., 1998;Sakakibara and Okano, 1997). A population of cells withnuclear immunoreactivity for Myt1 was immunolabeledfor Msi1 (Fig. 2, panel C). The cell density of theseMyt11 Msi11 cells was increased in lesioned white mat-ter, particularly early in the disease progression (Fig. 4).Msi1 is known to be expressed in reactive astrocytes(Sakakibara and Okano, 1997), which was confirmed inthe MHV lesioned tissues using double immunolabelingfor glial fibrillary acidic protein (GFAP) to identify reac-tive astrocytes (data not shown). However, reactive astro-cytes did not exhibit Myt1 immunoreactivity (data notshown) and so did not complicate the interpretation ofthe Myt11 Msi11 cells as neural stem cells.

The overall population of Myt1 expressing cells can beappreciated from the cell densities quantified for eachcell type from white matter of MHV-infected micethroughout the disease course as compared to PBS con-trol white matter (Fig. 4). Myt1 expression appears tobegin during the neural stem cell stage (Myt11 Msi11)and continue throughout the OP cell stage (Myt11PDGFaR1), which comprises the majority of cells withnuclear Myt1 immunoreactivity. In contrast, Myt1 isexpressed in only a small proportion of the mature oligo-dendrocyte population (Myt11 CC11), possibly reflect-ing expression during a transitional stage of OP differ-

GLIA DOI 10.1002/glia

690 VANA ET AL.

Fig. 1. Myt1 expression and BrdU incorporation in MHV lesions duringdemyelination and remyelination. Mice injected with MHV or PBS vehiclewere perfused at 2 weeks postinjection (wpi) (active demyelination), at 4 wpi(early remyelination), and at 8 wpi (advanced remyelination). (A) Quantifi-cation of Myt1 mRNA expression in white matter of cervical transverse sec-tions throughout MHV disease progression. In MHV-injected mice, as com-pared to PBS control mice, the density of cells labeled by in situ hybridiza-tion for Myt1 mRNA is increased by 2 wpi (P5 0.0323), peaks at 4 wpi (P50.0020), and then declines by 8 wpi but is still elevated from controls (P 50.0297). (B) Quantification of Myt1 mRNA expression relative to BrdU incor-poration throughout MHV disease progression. BrdU was administered dur-ing the final 4 h prior to sacrifice to identify actively proliferating cells(BrdU1) among the population of cells expressing Myt1 mRNA (Myt11).The Myt11 BrdU1 cell density peaks during active demyelination (2 wpi; P5 0.016) then remains significantly increased at 4 wpi (P 5 0.003), relativeto the very low density (0.4 cells/mm2) in PBS-injected mice. By 8 weeks,Myt11 BrdU1 cell density has returned to control levels (P 5 0.0791).BrdU1 cells that do not have detectable Myt1 mRNA (Myt12) are increasedover control levels at 2 wpi (P 5 0.016) and at 4 wpi (P 5 0.0007), but arestill slightly elevated at 8 wks (P 5 0.017). For panels A and B: PBS, n 5 3mice; MHV 2 wpi, n 5 4 mice; MHV 4 wpi, n 5 6 mice; MHV 8 wpi, n 5 4mice. (C) Representative image of dorsal column area (border denoted byblack dashed line) in transverse section of cervical spinal cord from anMHV-infected mouse at 4 wpi. In situ hybridization showing Myt1 mRNA tran-scripts (blue) combined with detection of BrdU (brown) to assess prolifera-tion. Increased BrdU incorporation in lesioned areas of the dorsal columns(upper part of image) corresponds with an increased density of cells expres-sing Myt1, compared to the normal appearing areas in the deep dorsal col-umn white matter. GM 5 gray matter. Scale bar5 50 lm. (D) Enlargementof area boxed in panel (C). Cells with detectable Myt1 mRNA are small withan elongated cell body, characteristic of OP cells, and could be found labeled

with BrdU (Myt11 BrdU1, black arrowheads) or may not have incorpo-rated BrdU during this 4-h terminal pulse (Myt11 BrdU2, black arrows).BrdU labeled cells that do not express Myt1 (Myt12 BrdU1, brown arrow-head) are also common, such as the example with a large oval nucleus thatis characteristic of an astrocyte. Dorsal column border with gray matter(GM) is denoted by black dash lines. Scale bar5 25 lm. (E) Disease progres-sion among MHV-infected mice used in the data shown in panels (A) and(B). Motor impairment and recovery was monitored with the hang time testof grip strength and coordination. Control mice injected with PBS can read-ily hang from bars of a wire cage top for 60 s. Following MHV injection,severely affected mice had significantly reduced hang times when tested ondays 7–17. In the following weeks, the hang time scores improve graduallyuntil reaching control levels. ANOVA with repeated measures: **P < 0.01;*P < 0.05; days 7–14: n 5 23 mice for MHV, n 5 12 mice for PBS; days 17–28: n 5 17 mice for MHV, n 5 9 mice for PBS; days 35–56: n 5 9 mice forMHV, n5 6 mice for PBS. (F) Quantification of Myt1 mRNA expression andBrdU incorporation within normal versus lesioned white matter areas.Within transverse spinal cord sections, regions were demarcated as whitematter areas of demyelination (MHV lesioned WM), adjacent normalappearing white matter (MHVNAWM), or white matter of PBS-injected con-trols (PBS WM). Mice were analyzed at 4 wpi after MHV or PBS injection.An increased density of cells with detectable levels of Myt1 mRNA tran-scripts is mainly restricted to MHV lesion areas, which are significantly ele-vated compared with NAWM (P 5 0.0005) or with PBS WM (P 5 0.0081).Myt11 BrdU1 proliferating cells are also significantly increased specificallywithin lesion areas relative to NAWM (P 5 0.0009) and PBS WM (P 50.0003). Myt12 BrdU1 cells within MHV lesions are also significantlyincreased over both NAWM (P 5 0.0002) and PBS WM (P 5 0.0038).NAWM was not significantly different from PBS WM for any of the valuesshown. PBS, n5 3 mice; MHV, n5 6 mice.

691Myt1 EXPRESSION IN RESPONSE TO CNS DEMYELINATION

GLIA DOI 10.1002/glia

entiation into mature oligodendrocyte with subsequentdown regulation. The combined densities of Myt11 cellsis lower than the density of oligodendrocytes in non-

lesioned tissue (Fig. 3C) but could be sufficient forregenerating the observed oligodendrocyte populationwith proliferation in subsequent weeks.

Fig. 2. Characterization of cell types expressing Myt1. Myt1 immuno-detection alone (A) and in combination with immunostaining for CC1 toidentify mature oligodendrocytes (B), Msi1 to identify immature neuralstem cells (C), and PDGFaR to identify OP cells (D). (A) Low power of atransverse hemi-section of cervical spinal cord during the early remyeli-nation stage at 4 wpi of MHV. The approximate boundary of the dorsalcolumns is denoted by white dash marks. The density of Myt1 immunore-active cells is highest in lesioned areas of the superficial dorsal columnsand the ventrolateral white matter. Scale bar 5 100 lm. (B) In a highermagnification of dorsal column white matter from a 2 wpi mouse, the dor-sal column border with the gray matter (GM) is noted by white dashmarks. Normal CC1 immunolabeling (green) of oligodendrocytes is seenin the deep dorsal column white matter, which is not lesioned. In themore superficial dorsal column region, the lesion area is evident by thelack of CC1 immunolabeled oligodendrocytes (top right of panel) where ahigh density of Myt11 CC12 cells (red) are present (examples at arrow-heads). Along the lesion border, CC1 immunoreactivity is detected in cellswith nuclear Myt1 (CC11 Myt11 examples at arrows). Scale bar 5 50 lm.The inset shows an enlarged view of several CC11 cells and a pair of

CC11 Myt11 cells. Scale bar 5 10 lm. (C) Msi1 immunoreactivity (green)is increased in reactive astrocytes and neural stem cells in the su-perficial part of the dorsal column white matter of a 2 wpi mouse. Cellsexpressing Myt1 (red; arrowheads) are more prevalent in the lesion areathan in the nonlesioned deep dorsal column white matter. Reactiveastrocytes express Msi1 (detected with immunostaining for glial fibril-lary acidic protein; data not shown). However, Msi11 Myt11 cells(examples at arrows) are likely to be neural stem cells since in these tis-sues Myt1 was not co-expressed in astrocytes (data not shown). Scalebar5 50 lm. The inset shows an enlarged view of two cells with cytoplas-mic Msi1 and nuclear Myt1 as well as three cells with only nuclear Myt1immunoreactivity. Scale bar5 10 lm. (D) The density of OP cells expres-sing PDGFaR (green) is increased in the lesioned area of the superficialdorsal column in a 4 wpi mouse. The density of cells with nuclear Myt1immunoreactivity (red) is also increased in the lesion area. Examples ofcells double labeled for PDGFaR andMyt1 are indicated by arrows whilePDGFaR- Myt11 cells are shown by arrowheads. Scale bar 5 50 lm.The inset shows a pair of PDGFaR1 cells with nuclear Myt1 immunor-eactivity. Scale bar5 10 lm.

GLIA DOI 10.1002/glia

692 VANA ET AL.

MHV lesions also contain substantial populations ofinfiltrating and endogenous immune cells (Matthewset al., 2002; Redwine and Armstrong, 1998). Myt1immunoreactivity was not detected in Mac1 immunola-beled macrophage/microglial cells, or within lymphocytesimmunostained with either Ox-52 for T-cells or CD-45receptor for B-cells (data not shown).

Myt1 Expression in MS Lesions and Control Cases

MS lesions of differing stages of demyelinating activ-ity (active, inactive, and remyelinated) and control caseswere analyzed by immunostaining and in situ hybridiza-tion to detect Myt1 expression (Fig. 5 and Table 1). InMS cases, Myt1 nuclear immunoreactivity was fre-quently observed in cells located in the PPWM adjacentto active or inactive lesions, or within early remyelinat-ing lesions (Fig. 5A and Table 1). In control adult whitematter, nuclear Myt1 immunostaining was much lessfrequently observed (Fig. 5B and Table 1), consistentwith our previous report (Armstrong et al., 1997). In aminority of lesions, nuclear Myt1 immunoreactivity wasdetected among infiltrating lymphocytes (data notshown) or reactive astrocytes had cytoplasmic Myt1immunoreactivity (Fig. 5C). No immunostaining wasdetected above background levels when the Myt1 pri-mary antibody was omitted or was replaced with anantibody against Myt1L, a closely related Myt1 familymember that is not expressed in oligodendrocytes (Kimet al., 1998 and data not shown). The increased fre-quency of Myt1 immunostaining in MS tissues was con-firmed using in situ hybridization to detect Myt1 mRNAtranscripts (Fig. 5D).

Myt1 expression may activate transcription of myelin-specific genes as OP cells differentiate into mature oligo-dendrocytes (Kim and Hudson, 1992; Nielsen et al.,2004). To identify cells with nuclear Myt1 localizationrelative to myelin-specific gene transcription, immuno-

Fig. 4. Comparison among cell types exhibiting nuclear Myt1 immu-noreactivity in PBS control mice and MHV mice throughout diseaseprogression. OP cells immunolabeled with PDGFaR were the principalcell type that expressed Myt1. In addition, the population of Myt11PDGFaR1 OP cells increased significantly at 2 wpi (P < 0.001) and 4wpi (P < 0.001) before returning to PBS control levels at 8 wpi (p >0.05). Musashi1 (Msi1) neural stem cell marker detected a smallerpopulation of Myt1 expressing cells that was greatest at 2 wpi. Rela-tively few cells continued to exhibit nuclear Myt1 immunoreactivity af-ter differentiating into mature oligodendrocytes, as indicated by CC1immunolabeling. PBS, n 5 3 mice; MHV, n 5 3 mice.

Fig. 3. Quantitative analysis of Myt1 expression relative to cell typespecific markers. At 4 wpi with MHV or PBS vehicle, mice were sacri-ficed and the white matter (WM) of transverse cervical sections was an-alyzed to determine the density of Myt1 immunolabeled cells in combi-nation with cell type-specific markers. (A) Myt1 expressing immatureneural stem cells were detected with Musashi1 (Myt11 Msi11). Astro-cytes express Msi1 but not Myt1 (i.e. Myt12 Msi1) and so can be dis-tinguished from neural stem cells (see text). (B) The majority of OPcells expressing PDGFaR are mostly also immunolabeled for Myt1 (i.e.Myt11 PDGFaR1 relative to Myt12 PDGFaR1). (C) The density ofmature oligodendrocytes, identified by CC1 immunolabeling, is veryhigh in PBS WM but greatly reduced in MHV WM. Relatively fewmature oligodendrocytes have detectable Myt1 expression (i.e. Myt11CC11 relative to Myt12 CC11). PBS, n 5 3 mice; MHV, n 5 3 mice.

693Myt1 EXPRESSION IN RESPONSE TO CNS DEMYELINATION

GLIA DOI 10.1002/glia

staining for Myt1 was combined with in situ hybridiza-tion for PLP mRNA (Figs. 5E,F). In the combined proto-col, specific signal was maintained although the numberof cells identified was reduced, relative to the separatedetection methods. Individual cells could be clearly iden-

tified as double labeled for Myt1 immunoreactivity andPLP mRNA transcripts in MS tissues (Fig. 5E andinset), which was not observed in the white matter ofcontrol cases (Fig. 5F). A similar protocol to double labelOP cells for PDGFaR mRNA and Myt1 was not feasible

Fig. 5. Myt1 expression in multiple sclerosis (MS) and control cases.Representative results are shown from adult human brain sections thatwere immunostained for Myt1 (A–C), processed for in situ hybridiza-tion for Myt1 mRNA (D), or analyzed by immunostaining for Myt1 com-bined with in situ hybridization for proteolipid protein (PLP) mRNA asa marker of oligodendrocytes (E,F). Nuclear immunostaining for Myt1(brown DAB with pink eosin tissue counter stain) was seen in smallround cells in the periplaque white matter (PPWM) adjacent to MSlesions (A, PPWM; case 9, Table 1), at an increased density relative towhite matter control areas (B; case 2, Table 1). In some MS lesions (C,lesion area; case 6, Table 1), reactive astrocytes exhibited cytoplasmicimmunoreactivity for Myt1 (brown DAB with pink eosin tissue counter

stain), which was never seen in control cases. In situ hybridization forMyt1 mRNA (D, lesion edge; dark blue; case 4, Table 1) in MS lesionsindicated expression of Myt1 in a pattern similar to Myt1 immunoreac-tivity (A). Within MS lesions, some cells with nuclear Myt1 expressedPLP mRNA transcripts (E, lesion; Myt1 brown, PLP dark blue; case 10,Table 1). Co-labeling for Myt1 among oligodendrocytes expressing PLPmRNA was generally not observed in adult human control white matter(F; case 3, Table 1). Scale bars 5 50 lm for (A–F), as shown in (F). (A–F) Arrows indicate examples of cells expressing Myt1 protein (A–C,E)or mRNA transcripts (D). Arrowheads (B,E,F) indicate examples of cellsthat are not expressing Myt1.

694 VANA ET AL.

GLIA DOI 10.1002/glia

because of the low abundance of PDGFaR transcriptsper cell. Importantly, as in our mouse data of Myt11CC11 cells (Figs. 2–4), Myt1 expression in a subset ofoligodendrocytes identified by PLP mRNA may indicateongoing OP differentiation into mature oligodendrocytesand support a potential role for Myt1 in oligodendrocyteregeneration in MS lesions.

DISCUSSION

In experimental and human demyelinating diseases,the CNS is capable of effective oligodendrocyte regenera-tion, remyelination, and recovery of function in viableaxons. However, this capacity for repair becomes limitedwith repeated or chronic episodes of demyelination. Tobetter understand how successful remyelination can beaccomplished, the current study examines the cellularresponses during spontaneous remyelination followingtransient experimental demyelination. Recent studieshave identified an emerging array of transcription fac-tors that are associated with OP proliferation, inductionof differentiation toward an oligodendrocyte phenotype,and elevated transcription of myelin-specific genes dur-ing myelination (Gohkan et al., 2005; Ligon et al., 2006).Analysis of developmental myelination has predictedspecific transcription factors that have been shown to beinvolved in oligodendrocyte lineage responses to demye-lination (Arnett et al., 2004; Fancy et al., 2004; Wata-nabe et al., 2004).

Myt1 warranted analysis as a transcription factor reg-ulating oligodendrocyte regeneration and remyelinationbased upon findings indicating a role in proliferation,differentiation, and activation of myelin-specific genetranscription during the generation of oligodendrocytesin the developing CNS (Armstrong et al., 1995; Nielsenet al., 2004). A role for Myt1 in human development hasbeen predicted from the pattern of Myt1 expression inthe developing human brain and from correlation ofMyt1 immunolabeling with the attempted regenerativeresponse of immature oligodendrocyte lineage cells incases of periventricular leukomalacia (Hirayama et al.,2003). In the adult human CNS, Myt1 is expressed ingerminal zones and is upregulated in gliomas, indicatingpotential function in immature proliferative cell types(Armstrong et al., 1997).

We report increased expression of Myt1 as a localwhite matter response to demyelination. We used theMHV model of experimental demyelination to assess theoligodendrocyte regenerative response in the context ofa complex lesion environment that includes gliosis,inflammation, and breakdown of the blood-brain barrierto reflect the complex pathology of MS lesions (Lucchi-netti et al., 2005; Morales et al., 2006). Myt1 expressioncorresponded with incorporation of BrdU, indicative of aproliferative phenotype. Consistent with this finding,the majority of cells with nuclei immunolabeled forMyt1 were double immunolabeled for PDGFaR, amarker of OP cells, with a substantial response alsonoted in cells immunolabeled for Msi1, a marker ofneural stem cells. Furthermore, a role for Myt1 in theoligodendrocyte lineage response in human injury anddisease is suggested by our observation of increasedexpression of Myt1 associated with MS lesions.

In the current study, a role for Myt1 in remyelinationis supported by correlation with increased expression ofMyt1 in proliferating cells and localization of Myt1 innuclei of neural stem cells and OP cells. Interestingly,only nuclear immunoreactivity for Myt1 was observed inthe adult mouse tissues, even though a transition tocytoplasmic localization was observed in rat developingoligodendrocytes (Armstrong et al., 1995). The presentquantitative cellular analysis demonstrating Myt1expression in OP cells extends an earlier report ofincreased abundance of mRNA transcripts for both Myt1and PDGFaR during remyelination following ethidiumbromide induced demyelination in rats (Sim et al.,2002). In addition, similar to our current quantificationof Myt1 and BrdU incorporation, our previous MHVstudy demonstrated that PDGFaR1 cell proliferation inresponse to demyelination was localized within and nearlesions (Redwine and Armstrong, 1998). However, Myt1is expressed in additional populations of responding cellsthat were not identified by PDGFaR immunolabeling.We identified a subset of Myt1 expressing cells by co-immunolabeling for Msi1. Msi1 is an RNA-binding pro-tein that has been used as a marker of neural stem cellsin the developing and adult CNS of rodents and humans(Kaneko et al., 2000; Pincus et al., 1998; Sakakibaraand Okano, 1997). Therefore, the initial expression ofMyt1 in MHV lesions is expected to be in neural stemcells that may represent a relatively small but poten-

TABLE 1. Density of Cells with Myt1 Nuclear Immunoreactivity in White Matter of Human Controls and MS Lesions

Case Lesion stage Lesion cells (mm2) PPWM cells (mm2) NWM cells (mm2)

1 Control 152 Control 1013 Control 864 Demyelinated inactive 101 3215 Demyelinated inactive 113 2846 Demyelinated inactive 57 3097 Early active 193 3718 Early active 113 3659 Early active 38 309

10 Early remyelination 536 n/a

PPWM, periplaque white matter adjacent to lesions in MS tissues. NWM, normal white matter in non-neurological controls. n/a, samples are each from a biopsy of thelesion area so PPWM tissue is not available.

695Myt1 EXPRESSION IN RESPONSE TO CNS DEMYELINATION

GLIA DOI 10.1002/glia

tially important population of cells that is responsive todemyelination.

While the current analysis has not directly demon-strated a lineage progression for the expression of Myt1during MHV disease progression, a possible scenariofrom current data and the available literature is thatMyt1 is expressed in Msi11 neural stem cells that maydifferentiate into PDGFaR1 OP cells then amplify inresponse to demyelination to generate CC11 oligoden-drocytes to accomplish remyelination. Myt1 expressionprior to the OP stage has been shown with in vitro anal-ysis of cultured neonatal cells (Armstrong et al., 1995).In vitro, Myt1 continued to be expressed at the OPstage, followed by downregulation in mature oligoden-drocytes after accumulation of myelin proteins (Arm-strong et al., 1995). In vivo, the proliferative responseduring demyelination has been clearly associated withthe OP stage in MHV lesions, other models of experi-mental demyelination, and MS lesions (Redwine andArmstrong, 1998; Reynolds et al., 2001; Solanky et al.,2001; Watanabe et al., 2002). In this MHV model, cellsthat incorporated thymidine analogues, as indicators ofproliferation, during demyelination have been shown toexpress PDGFaR and then maintain labeling during thetransition into oligodendrocytes in the remyelinationphase (Godfraind et al., 1989; Redwine and Armstrong,1998). In the current MHV study, expression of Myt1was observed mainly in PDGFaR1 OP cells and in onlya small proportion of CC11 oligodendrocytes, whichmay be cells that recently differentiated and had not yetdownregulated Myt1.

The current finding that the increased density ofMyt1 cells is specifically localized to areas of white mat-ter lesions is consistent with the local generation of oli-godendrocytes reported in several studies of remyelina-tion. In focally demyelinated spinal cord, newly gener-ated remyelinating cells appeared to be generated froma local ring of normal tissue surrounding the lesion(Franklin et al., 1997). Similarly, retroviral labeling ofremyelinating cells indicated recruitment from less than500 lm from the lesion site in adult rat corpus callosum(Gensert and Goldman, 1997). A preliminary study ofmice administered cuprizone indicated that Myt1expression is upregulated during acute demyelination inthe corpus callosum in areas of increased OP prolifera-tion (data not shown). This comparison indicates thatincreased Myt1 expression may occur during OP amplifi-cation in response to diverse demyelinating pathologies.

Analysis of MS lesions demonstrated the highest den-sity of Myt1 expressing cells in early remyelinatinglesions, consistent with our findings in MHV and cupri-zone. Furthermore, Myt1 expression was increased inPPWM relative to active or inactive lesion areas. Thevariability of Myt1 expression in MS lesions may reflectthe reported differences of immature OP populations(Chang et al., 2002; Maeda et al., 2001). In contrast toMHV lesions, some MS lesions exhibited Myt1 immunor-eactivity in lymphocytes and astrocytes. In the unusualcases in which astrocytes expressed Myt1, the immunor-eactivity was localized in the cytoplasm. Similarly, Myt1

immunoreactivity in astrocyte cytoplasm was observedin highly reactive astrocytes of epileptic foci in humantemporal lobe tissue (Armstrong et al., 1997). It is notclear whether this astrocytic expression of Myt1 is theresult of a species difference or possibly due to differ-ences in the pathology of some MS lesions.

Further work is required to understand the regulationof Myt1 in development and pathology, and the functionof Myt1 in each context. Myt1 acted as a transcriptionalactivator to increase luciferase levels from the promoterof a myelin-specific gene, 20,30-cyclic nucleotide 30-phos-phodiesterase, in transient transfection assays (Nielsenet al., 2004). However, Myt1 may act in concert withother transcription factors or co-factors that contributeto effects on transcriptional controls (Koyano-Nakagawaet al., 1999; Schneider et al., 2001). In addition, Myt1may form complexes with Sin3B, a protein that med-iates repression by recruiting histone deacetylases(Romm et al., 2005). Therefore, the function of Myt1may depend upon the levels of other nuclear proteinsthat may be differentially regulated in cells relative tostage of differentiation or environmental context. Thecurrent study documenting Myt1 expression correspond-ing to regenerative responses of the oligodendrocyte line-age in animal models and MS lesions suggests thatMyt1 may have functional importance in remyelination.

ACKNOWLEDGMENTS

Dr. Lynn D. Hudson, National Institutes of Health,generously provided the Myt1 antibody and plasmids forMyt1 and PLP.

REFERENCES

Armstrong RC. 2000. Potential roles of trophic factors in CNS develop-ment and recovery from demyelinating disease. In: Mocchetti I, edi-tor. Neurotrophic factors. Graham Publishing, Mountain Home, TN.pp 417–437.

Armstrong RC, Kim JG, Hudson LD. 1995. Expression of myelin tran-scription factor I (MyTI), a ‘‘zinc-finger’’ DNA-binding protein, indeveloping oligodendrocytes. Glia 14:303–321.

Armstrong RC, Le TQ, Frost EE, Borke RC, Vana AC. 2002. Absence offibroblast growth factor 2 promotes oligodendroglial repopulation ofdemyelinated white matter. J Neurosci 22:8574–8585.

Armstrong RC, Migneault A, Shegog ML, Kim JG, Hudson LD, HesslerRB. 1997. High-grade human brain tumors exhibit increased expres-sion of myelin transcription factor 1 (MYT1), a zinc finger DNA-bind-ing protein. J Neuropathol Exp Neurol 56:772–781.

Armstrong RC, Redwine JM, Messersmith DJ. 2005.Coronavirus-induced demyelination and spontaneous remyelination: Growth factorexpression and function. In: Lavi E, Constantinescu CS, editors.Experimental models of multiple sclerosis. Norwell, MA: SpringerScience. p 2005.

Arnett HA, Fancy SP, Alberta JA, Zhao C, Plant SR, Kaing S, RaineCS, Rowitch DH, Franklin RJ, Stiles CD. 2004. bHLH transcriptionfactor Olig1 is required to repair demyelinated lesions in the CNS.Science 306:2111–2115.

Bruck W, Kuhlmann T, Stadelmann C. 2003. Remyelination in multiplesclerosis. J Neurol Sci 206:181–185.

Bruck W, Porada P, Poser S, Rieckmann P, Hanefeld F, KretzschmarHA, Lassmann H. 1995. Monocyte/macrophage differentiation inearly multiple sclerosis lesions. Ann Neurol 38:788–796.

Chang A, Tourtellotte WW, Rudick R, Trapp BD. 2002. Premyelinatingoligodendrocytes in chronic lesions of multiple sclerosis. N Engl JMed 346:165–173.

696 VANA ET AL.

GLIA DOI 10.1002/glia

Fancy SP, Zhao C, Franklin RJ. 2004. Increased expression of Nkx2.2 andOlig2 identifies reactive oligodendrocyte progenitor cells responding todemyelination in the adult CNS. Mol Cell Neurosci 27:247–254.

Ferguson B, Matyszak MK, Esiri MM, Perry VH. 1997. Axonal damagein acute multiple sclerosis lesions. Brain 120(Part 3):393–399.

Franklin RJ, Gilson JM, Blakemore WF. 1997. Local recruitment ofremyelinating cells in the repair of demyelination in the central nerv-ous system. J Neurosci Res 50:337–344.

Frost EE, Nielsen JA, Le TQ, Armstrong RC. 2003. PDGF and FGF2regulate oligodendrocyte progenitor responses to demyelination. JNeurobiol 54:457–472.

Gensert JM, Goldman JE. 1997. Endogenous progenitors remyelinatedemyelinated axons in the adult CNS. Neuron 19:197–203.

Godfraind C, Friedrich VL, Holmes KV, Dubois-Dalcq. 1989. In vivoanalysis of glial cell phenotypes during a viral demyelinating diseasein mice. J Cell Biol 109:2405–2416.

Gokhan S, Marin-Husstege M, Yung SY, Fontanez D, Casaccia-BonnefilP, Mehler MF. 2005. Combinatorial profiles of oligodendrocyte-selec-tive classes of transcriptional regulators differentially modulate mye-lin basic protein gene expression. J Neurosci 25:8311–8321.

Hirayama A, Oka A, Ito M, Tanaka F, Okoshi Y, Takashima S. 2003.Myelin transcription factor 1 (MyT1) immunoreactivity in infantswith periventricular leukomalacia. Brain Res Dev Brain Res 140:85–92.

Hudson LD, Berndt JA, Puckett C, Kozak CA, Lazzarini RA. 1987.Aberrant splicing of proteolipid protein mRNA in the dysmyelinatingjimpy mutant mouse. Proc Natl Acad Sci USA 84:1454–1458.

Kaneko Y, Sakakibara S, Imai T, Suzuki A, Nakamura Y, Sawamoto K,Ogawa Y, Toyama Y, Miyata T, Okano H. 2000. Musashi1: An evolu-tionally conserved marker for CNS progenitor cells including neuralstem cells. Dev Neurosci 22:139–153.

Kim JG, Armstrong RC, Berndt JA, Kim NW, Hudson LD. 1998. Asecreted DNA-binding protein that is translated through an internalribosome entry site (IRES) and distributed in a discrete pattern inthe central nervous system. Mol Cell Neurosci 12:119–140.

Kim JG, Armstrong RC, v Agoston D, Robinsky A, Wiese C, Nagle J,Hudson LD. 1997. Myelin transcription factor 1 (Myt1) of the oligo-dendrocyte lineage, along with a closely related CCHC zinc finger, isexpressed in developing neurons in the mammalian central nervoussystem. J Neurosci Res 50:272–290.

Kim JG, Hudson LD. 1992. Novel member of the zinc finger superfam-ily: A C2-HC finger that recognizes a glia-specific gene. Mol Cell Biol12:5632–5639.

Koyano-Nakagawa N, Wettstein D, Kintner C. 1999. Activation ofXenopus genes required for lateral inhibition and neuronal differ-entiation during primary neurogenesis. Mol Cell Neurosci 14:327–339.

Ligon KL, Kesari S, Kitada M, Sun T, Arnett HA, Alberta JA, AndersonDJ, Stiles CD, Rowitch DH. 2006. Development of NG2 neural pro-genitor cells requires Olig gene function. Proc Natl Acad Sci USA103:7853–7858.

Lucchinetti CF, Parisi J, Bruck W. 2005. The pathology of multiple scle-rosis. Neurol Clin 23:77–105.

Maeda Y, Solanky M, Menonna J, Chapin J, Li W, Dowling P. 2001.Platelet-derived growth factor-a receptor-positive oligodendroglia arefrequent in multiple sclerosis lesions. Ann Neurol 49:776–785.

Matthews AE, Weiss SR, Paterson Y. 2002. Murine hepatitis virus–amodel for virus-induced CNS demyelination. J Neurovirol 8:76–85.

Morales Y, Parisi JE, Lucchinetti CF. 2006. The pathology of multiplesclerosis: Evidence for heterogeneity. Adv Neurol 98:27–45.

Nielsen JA, Berndt JA, Hudson LD, Armstrong RC. 2004. Myelin tran-scription factor 1 (Myt1) modulates the proliferation and differentia-tion of oligodendrocyte lineage cells. Mol Cell Neurosci 25:111–123.

Pincus DW, Keyoung HM, Harrison-Restelli C, Goodman RR, FraserRA, Edgar M, Sakakibara S, Okano H, Nedergaard M, Goldman SA.1998. Fibroblast growth factor-2/brain-derived neurotrophic factor-associated maturation of new neurons generated from adult humansubependymal cells. Ann Neurol 43:576–585.

Redwine JM, Armstrong RC. 1998. In vivo proliferation of oligodendro-cyte progenitors expressing PDGFaR during early remyelination. JNeurobiol 37:413–428.

Reynolds R, Cenci di Bello I, Dawson M, Levine J. 2001. The responseof adult oligodendrocyte progenitors to demyelination in EAE. ProgBrain Res 132:165–174.

Romm E, Nielsen JA, Kim JG, Hudson LD. 2005. Myt1 family recruitshistone deacetylase to regulate neural transcription. J Neurochem93:1444–1453.

Sakakibara S, Okano H. 1997. Expression of neural RNA-binding pro-teins in the postnatal CNS: Implications of their roles in neuronaland glial cell development. J Neurosci 17:8300–8312.

Schneider ML, Turner DL, Vetter ML. 2001. Notch signaling can in-hibit Xath 5 function in the neural plate and developing retina. MolCell Neurosci 18:458–472.

Sim FJ, Zhao C, Penderis J, Franklin RJ. 2002. The age-relateddecrease in CNS remyelination efficiency is attributable to an impair-ment of both oligodendrocyte progenitor recruitment and differentia-tion. J Neurosci 22:2451–2459.

Solanky M, Maeda Y, Ming X, Husar W, Li W, Cook S, Dowling P. 2001.Proliferating oligodendrocytes are present in both active and chronicinactive multiple sclerosis plaques. J Neurosci Res 65:308–317.

Trapp BD, Peterson J, Ransohoff RM, Rudick R, Mork S, Bo L. 1998.Axonal transection in the lesions of multiple sclerosis. N Engl J Med338:278–285.

Watanabe M, Hadzic T, Nishiyama A. 2004. Transient upregulation ofNkx2.2 expression in oligodendrocyte lineage cells during remyelina-tion. Glia 46:311–322.

Watanabe M, Toyama Y, Nishiyama A. 2002. Differentiation of prolifer-ated NG2-positive glial progenitor cells in a remyelinating lesion.J Neurosci Res 69:826–836.

Wong K, Armstrong RC, Gyure KA, Morrison AL, Rodriguez D, Mata-lon R, Johnson AB, Wollmann R, Gilbert E, Le TQ, Bradley CA,Crutchfield K, Schiffmann R. 2000. Foamy cells with oligodendroglialphenotype in childhood ataxia with diffuse central nervous systemhypomyelination syndrome. Acta Neuropathol (Berl) 100:635–646.

Wrathall JR, Li W, Hudson LD. 1998. Myelin gene expression after ex-perimental contusive spinal cord injury. J Neurosci 18:8780–8793.

697Myt1 EXPRESSION IN RESPONSE TO CNS DEMYELINATION

GLIA DOI 10.1002/glia