DIMINISHED DEGRADATION OF MYELIN BASIC PROTEIN BYANTI-SULFATIDE ANTIBODY AND INTERFERON-γ IN MYELINFROM GLIA MATURATION FACTOR-DEFICIENT MICE

Krishnakumar Menon1,2, Yanghong Wu2, Joel Haas2, Shailendra K. Sahu2,3, Baoli Yang4,and Asgar Zaheer1,2,51Veterans Affair Medical Center, Iowa City, IA

2Division of Neurochemistry and Neurobiology, Department of Neurology, University of Iowa, Iowa City, IA52242

3Department of Neurosurgery, University of Iowa, Iowa City, IA 52242

4Department of Obstetric and Gynecology, University of Iowa, Iowa City, IA 52242

AbstractIn this study we show the effect of anti-sulfatide (RmAb) antibodies and inflammatory cytokines,TNF-α and IFN-γ in inducing myelin basic protein (MBP) degradation in myelin isolated from controlwild type (WT) and glia maturation factor (GMF)-deficient (GMF-KO) mice. GMF was not detectedin isolated myelin from WT and GMF-KO mice although it is present in brains of WT mice. Ourresults show that calcium dependent neutral protease activity caused significantly elevateddegradation of 18.5 and/or 17.5 kDa isoforms of MBP in WT-myelin treated with RmAb or IFN-γ.In contrast, MBP degradation in isolated myelin from GMF-KO mice remained unaffected followingtreatment with RmAb, IFN-γ, or GM-CSF. Neither the 14 kDa isoform of MBP or proteolipid protein(PLP) showed an elevated degradation compared to controls. A virtual absence of GM-CSF, TNF-α and IFN-γ in GMF-KO brain compared to WT was also evident when the animals were challengedwith MOG 35-55. Additionally, the myelin from GMF-KO mice showed difference in distributionof myelin oligodendrocyte glycoprotein (MOG) and β-tubulin in a sucrose density gradient myelin-axolemmal fractions compared to WT. Taken together, our data suggests a role for GMF in thebiochemical organization of myelin and thereby its effect on MBP degradation induced by RmAband IFN-γ.

KeywordsGlia maturation factor; inflammatory cytokines; MBP; myelin degradation

INTRODUCTIONMultiple sclerosis (MS) is a demyelinating disorder characterized by an autoimmune responseto myelin antigens resulting in a widespread myelin destruction accompanied by damage tothe underlying axon. This event apparently thought to be the result of a combined autoimmune

5Address correspondence to: Asgar Zaheer, Department of Neurology, The University of Iowa, 200 Hawkins Drive, Iowa City, IA 52242,Tel. 319-335-8529; Fax 319-335-8528 E-Mail: asgar-zaheer@uiowa.eduPublisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resultingproof before it is published in its final citable form. Please note that during the production process errors may be discovered which couldaffect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public AccessAuthor ManuscriptNeurosci Res. Author manuscript; available in PMC 2008 June 1.

Published in final edited form as:Neurosci Res. 2007 June ; 58(2): 156–163.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

response to some of the myelin components (Glynn and Linington 1989;Bernard et al.1997;Steinman and Zamvil 2005;Gold et al. 2006). Research efforts in recent years on gliamaturation factor (GMF), a highly conserved brain-specific protein, isolated, sequenced andcloned in our laboratory (Lim et al. 1989;Lim et al. 1990;Kaplan et al. 1991;Zaheer et al.1993), have demonstrated an immunomodulatory function for GMF. Recently, it has beenestablished that overexpression of GMF in astrocytes leads to immune activation of microgliathrough secretion of granulocyte-macrophage-colony stimulating factor (GM-CSF) (Zaheer etal. J. Neurochem. 2006 In press). Moreover, on survey of gene expression by DNA microarrayanalysis (Zaheer et al. 2002), we have found a significant increase in the expression of severalgenes, such as major histocompatibility complex (MHC) proteins, IL-1 beta, MIP-1 β, all ofwhich have been associated with the development of EAE. We also reported the stimulationof p38 MAP kinase pathway (Lim and Zaheer 1996;Zaheer and Lim 1996,1998) and NF-kB(Lim et al. 2000) by GMF in astrocytes. Based on GMF’s ability to activate microglia andinduce several well-established pro-inflammatory mediators, we hypothesize that intracellularGMF is involved in the pathogenesis of inflammatory demyelinating disease of the centralnervous system such as MS and EAE. Our recent experiments to test this hypothesis usingGMF-deficient (GMF-knockout) mice, prepared in our laboratory (Lim et al. 2004),demonstrated a significant decrease in incidence, delay in onset, and reduced severity of EAEin GMF-knockout mice (Zaheer et al. 2006 In press). Whether, lack of an active immuneresponse at the level of proinflammatory cytokine production in GMFKO animals or theresistance of myelin itself to immune attack is critical in understanding the biochemistry behinddemyelination. Though the biochemical mechanism leading to myelinolysis is poorlyunderstood, previous studies, both in vitro (de Rosbo and Bernard 1989;Johns et al.1995;Menon et al. 1997) and in EAE animals implicated an important role for the calciumactivated neutral proteases (CANP) in MBP degradation (Schaecher et al. 2001;Schaecher etal. 2002). Injection of anti-GalC antiserum into the normal rabbit vitreous body resulted indemyelination of epiretinal myelinated fibers (Ozawa et al., 1989) and that intrathecal injectionof anti-GalC (galactocerebroside) along with complement induced focal demyelination(Keirstead et al., 1998). Similarly, implantation of anti-GalC secreting hybridoma cells in brainled to extensive demyelination (Rosenbluth et al., 2003). These studies implicate the significantinvolvement of antibodies in inducing myelin damage. How the myelin damage is broughtabout remained unanswered. We previously demonstrated that anti-MOG and anti-GalCinduces myelin breakdown by either activation of CANP or increased accessibility of MBP toCANP (de Rosbo and Bernard 1989;Menon et al., 1997). Subsequent studies on culturedoligodendrocytes using anti-MOG antibodies led to activation of signaling moleculessuggesting that signaling mediated events are involved in antibody induced oligodendrocyteand myelin damage (Menon et al., 1997;Marta et al., 2003). Thus our aim is to test the abilityof antibodies to sulfated galactocerebroside (R-mAb) which have been shown to induce adown-regulation of MBP and patching of the oligodendrocyte membrane (Dyer and Benjamins1988;Bansal and Pfeiffer 1989) as well as cytokines such as TNF-α or IFN-γ, which are knownto induce oligodendrocyte damage (Pouly et al. 2000;Jurewicz et al. 2005;Balabanov et al.2006 Zaheer et al. J. Neurochem. 2006 In press) in inducing an MBP degradation in myelinisolated from WT and GMF-KO animals. In addition, we will also test the ability of GMCSF,a chemokine that plays a major role in development of EAE in inducing MBP degradation inmyelin isolated from WT and GMF-KO animals.

MATERIALS AND METHODSAntibodies and cytokines

The antibody and the various cytokines used with myelin incubations include anti-sulfatideantibody (R-mAb) (a gift from Prof. Barbara Ranchst), IFN-γ and TNF-α (Chemicon, USA)and GM-CSF (Prospec Techno gene, IL). Following antibodies were used for the detection of

Menon et al. Page 2

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

proteins following electrophoresis and western blotting. Anti-MBP and anti-PLP (Chemicon,USA), anti-CNP (Calbiochem, USA), anti-MOG (clone 8-18C5, a gift from Prof. ClaudeBernard, Monash University, Australia), anti-β tubulin (Hybridoma facility, University ofIowa). Species specific secondary antibodies, goat anti-mouse or Goat anti-rabbit (Calbiochem,USA) were used at 1:2000 dilutions.

Induction of EAE and Quantitative estimation of mRNA by real-time RT-PCRFor active induction of EAE, mice (8-10 week-old, male) received 100 μg myelinoligodendrocyte glycoprotein peptide 35-55 in 100 μl PBS and mixed with 100 μl of completeFreund’s adjuvant (Sigma-Aldrich, St. Louis, MO). Animals injected with complete Freund’sadjuvant (CFA) were used as controls. Next day each animal received i.p. injection of 200 ngpertussis toxin. The animals were cared for in accordance with the guidelines approved by theIACUC and National Institutes of Health. Total RNA was extracted from four mice eachexperimental group (GMF+/+, wild type and GMF-/-, GMF-knockout), ten days postimmunization. Reverse transcription-polymerase chain reaction (RT-PCR) was carried out asdescribed earlier (Zaheer et al. 1993;Zaheer et al. 1995). The primer pairs were selected toyield a single amplicon based on dissociation curves and analysis by acrylamide gelelectrophoresis. The following oligonucleotide primers were used: GM-CSF, (5′-GGCCTTGGAAGCATGTAGAGG-3′ and 5′-GGAGAACTCGTTAGAGACGACTT-3′);TNF-α, (5′-CATCTTCTCAAAATTCGAGTGACAA-3′ and 5′-TGGGAGTAGACAAGGTACAACCC-3′); IFN-γ,(TCAAGTGGCATAGATGTGGAAGAA-3′ and 5′-TGGCTCTGCAGGATTTTCATG-3′);18S, (5′-TAA GTC CCT GCC CTT TGT ACA CA-3′ and 5′- GAT CCG AGG GCC TCACTA AAC-3′). Real-time quantitative PCR was performed (Zaheer et al. 2006, J. Neurochem.In press) in a model ABI Prism 7000 sequence detection system. The values were normalizedusing 18S rRNA as an endogenous internal standard. A serial dilution of a standard was runon each plate for each mRNA and used to calculate the relative levels of mRNA.

Myelin IsolationMyelin was isolated by the method of Norton and Podusolo (Norton and Poduslo 1973). Fivebatches of myelin were made from both wild type and GMFKO mice. Following isolation,myelin samples were snap frozen and kept at -80°C until used. In order to analyze thedistribution pattern of myelin from GMF-KO vs WT, we fractionated myelin according to themethod described by Menon et al. (Menon et al. 2003) for the isolation of myelin-axolemmalcomplexes. Briefly, brains were homogenized in 0.32 M sucrose/EGTA solution and layeredon top of 0.85M sucrose/EGTA solution in an SW 28 ultraclear tube and centrifuged at140,000g in an SW28 rotor for 1 h. The myelin at the interface and the dispersions beneath themyelin up to the pellet at the bottom were collected and given osmotic shock in 140 mL 10mMEGTA in milliQ water pH7.2. The entire material was centrifuged and the resultant pellethomogenized in 0.85M sucrose and layered over 1.0M sucrose/EGTA solution. This isfollowed by layering an equal amount of 0.32M sucrose/EGTA solution over the 0.85Msucrose-myelin homogenate in an SW 28 ultraclear tube. The entire material was centrifugedat 140,000g in an SW28 rotor for 1.5 h. The myelin at the interface and the dispersions beneaththe myelin up to 1M sucrose solution were collected and given the osmotic shock in 140 mL2mM EGTA in milliQ water pH7.2. The entire material was centrifuged and the resultant pellethomogenized in 0.85M sucrose and transferred to an SW 28 ultraclear centrifuge tube.Subsequently, 0.75M sucrose and 0.32M sucrose/EGTA solutions were layered in sequenceover 0.85M myelin-sucrose homogenate solution. The entire material was centrifuged at140,000g in an SW28 rotor for 15 h. Following ultracentrifugation, 1 mL fractions werecollected. The fractions include, 10 to 13 the low-density band (LDB), 14, 15, 16 and 17 themain compact myelin band, 18 to 25 the light dispersion (LD), 26 to 29 the heavy dispersion(HD), 30 to 33 the A and finally 34 to 37 as B. Fractions corresponding to LDB, LD, HD, A

Menon et al. Page 3

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

and B were pooled, given the osmotic shock and centrifuged. Pooling of these fractions wasperformed as described by Menon et al. (Menon et al. 2003). Fractions 14 to 17 wereindividually given osmotic shock in 2 volumes of 2mM EGTA in milliQ water pH7.2andcentrifuged. The pellets obtained following centrifugation were suspended in 50mM Tris pH7.4 containing protease cocktail inhibitor (Roche diagnostics, Mannheim, Germany) and keptfrozen at -80°C.

Myelin incubationsMyelin samples (120μg of total myelin proteins) isolated from WT and GMF-KO mice wereincubated in 25mM HEPES buffer at pH 7.4 in the presence or absence of either RmAb antibodyor control Igs (1:40 dilution) or cytokines (GM-CSF, 40ng/mL; TNF-α, 100ng/mL; IFN-γ,100ng/mL) in a final volume of 120 μL containing 30μM ATP at 37°C. In order to confirmthat calcium plays a key role in MBP degradation, we incubated WT myelin with RmAb or3mM CaCl2in presence or absence of 5mM EGTA. The reactions were started by the additionof myelin on gentle vortex. At different time intervals such as 23s and 60min, a 40μL aliquotwas taken and the reaction was stopped by mixing with an equal volume of 10%SDS on vortex.The incubation for 23 s served as the 0 time control.

Gel electrophoresis and immunoblottingSamples were prepared for electrophoresis according to Menon et al., (Menon et al. 2003) inurea buffer. 1 μg of samples were electrophoresed to detect the extent of MBP and PLPdegradation in a 14% SDS polyacrylamide gels (Laemmli 1970) and electroblotted onto PVDFImmobilon-P membranes (Amersham, Piscataway, NJ, USA). Blots were blocked and probedwith different antibodies in 10% solution of non-fat dry milk in Tris-buffered saline (TBS; 25mM Tris, 2.6 mM KCl, 0.14 M NaCl, pH 8.0) according to described by Menon et al.,(2003). Briefly, bands for the corresponding proteins were detected by probing with specificprimary antibodies at respective dilutions (above) in 10% non-fat dry milk for 1.5h in TBS,washed three times (TBS, 0.2%TW20), incubated with respective secondary antibodies for45min, washed as before and the bands were visualized using supersignal west pico ECLsolution (Pierce Biotechnology, Inc, IL, USA) on a blue basic autorad double emulsion film(ISC Bio Express, Kaysville, UT, USA). In order to detect the levels of PLP on the same blotsthat were probed for MBP, the respective blots were re probed (as above) following strippingof antibodies with 100mM glycine at pH 2. For the detection of MBP and PLP, a 1:1000 and1:500 dilutions of anti-MBP and anti-PLP antibodies were used. For the detection of MOGand β-tubulin a 5μg of myelin proteins was electrophoresed on a 14% gel and immunoblottedas described previously. Subsequently, the blots were probed with anti-MOG (1:500) and β-tubulin (1:2) in 10% milk.

Quantification of MBPThe extent of MBP degradation in 60 min incubated myelin samples compared to 0 minincubated samples was measured by scanning film images within the linear ranges of graindensity followed by measuring the mean band densities using NIH image analyzer (NationalInstitute of Health, Bethesda, MD, USA). The values were normalized to the DM20 (a splicevariant of PLP) values in the same blot and % degradation of MBP isoforms (18.5, 17.5 and14 kDa) or PLP was measured after 60 min of incubation from 0 min incubation.

Statistical analysisThe values given for incubations of each cytokine or antibody with myelin are means ± SEMof five different batches of myelin unless otherwise mentioned. Student’s t-test was used forthe analysis of data and p-values less than 0.05 is considered statistically significant.

Menon et al. Page 4

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

RESULTSAbsence of GM-CSF, TNF-α and INF-γ expression in GMF deficient mice

In order to test the expression profile of major cytokines such as TNF-α and IFN-γ that areknown to induce oligodendrocyte damage, we challenged WT and GMF-KO mice with MOG35-55 peptide. Subsequently, we analyzed the levels of GM-CSF, TNF-α and INF-γ levels inWT and GMF-KO mice by quantitative real time RT-PCR on RNA isolated from brain. Asshown in Fig. 1, the expression levels of GM-CSF, TNF-α and INF-γ in GMF-KO brain weresignificantly lower compared to WT.

MBP degradation in samples incubated with different cytokinesIn order to detect the MBP degradation in the incubated myelin samples from WT and GMF-KO, the different MBP isoforms were visualized by immunoblotting and quantified asdescribed in materials and methods. To measure the basal level of MBP degradation following60 min incubation, myelin samples from WT and GMF-KO were incubated with vehicle orcontrol Igs only. The extent of degradation of 18.5 kDa MBP isoform from five different myelinis 18.8 ± 3.5% (WT) and 21.6 ± 8.0% (GMF-KO) (Fig. 2A); for 17.5 kDa MBP isoform is 25.0± 5.8% (WT) and 25.6 ± 6.7% (GMF-KO) (Fig. 2B); for 14 kDa MBP isoform is 28.4 ± 3.6%(WT) and 32.8 ± 5.0% (GMF-KO) (Fig. 2C). In presence of control Igs, the percent decreasein MBP for 18.5 kDa MBP is 26.6 ± 4.9% (WT) and 11 ± 4.1% (GMF-KO), for 17 kDa MBPis 17.3 ± 8.6% (WT) and 21 ± 0.5 % (GMF-KO); for 14 kDa MBP is 37.6 ± 4.8% (WT) and22.3 ± 11% (GMF-KO) respectively. On incubation of myelin with GM-CSF resulted inminimal MBP degradation in all the MBP isoforms analyzed, a level similar to that seen incontrols from WT or GMF-KO myelin (Fig. 2). In contrast, IFN-γ treatment of myelin fromWT led to a significantly elevated degradation of 18.5 kDa MBP isoform (30.0 ± 2.2%; p=0.01)when compared to the same isoform from control myelin (18.8 ± 3.5%) (Fig. 2A). However,IFN-γ treatment did not affect the levels of either17.5 kDa (WT, 15.4 ± 5.7%; GMF-KO, 19.4± 4.7%; (Fig. 2B) or 14 kDa (WT, 34.0 ± 8.3%; GMF-KO, 28.2 ± 8.1%) isoforms of MBP inWT or GMF-KO myelin when compared to controls (Fig. 2C). On the other hand, TNF-αtreatment showed no change in 18.5 kDa isoform levels in WT (22.6 ± 5.8%) or GMF-KOmyelin (22.4 ± 6.5%) compared to controls (WT, 18.8 ± 3.5%; GMF-KO, 21.6 ± 8.0%) (Fig.2A). Notably, 17.5 kDa (21.2 ± 6.1%) and 14 kDa (23.8 ± 7.5%) isoforms of MBP from WTor GMF-KO (17.5 kDa, 18.8 ± 4.8%; 14 kDa, 27.6 ± 5.7%) did not show a difference in MBPdegradation compared to the respective WT controls (WT 17.5 kDa, 25 ± 5.8%; WT 14 kDa,32.8 ± 5.0%) or GMF-KO controls (17.5 kDa, 25.6 ± 6.7%; 14 kDa, 32.8 ± 5.0%) (Fig. 2B,C).

MBP degradation in samples incubated with anti-sulfatide RmAb antibodiesIn a similar manner to the cytokine incubations with myelin, we incubated myelin isolated fromboth WT and GMF-KO mice with anti-sulfatide RmAb antibody. As expected, RmAb antibodyinduced a significantly elevated degradation of both 18.5 kDa (47.0 ± 5.2%; p=0.002 vs controlsham, p=0.01 vs control Igs) and 17.5 kDa (48.6 ± 7.1%; p=0.01 vs control sham and controlIgs) isoforms of MBP in myelin from WT background (Fig. 2 A, B). In contrast, a surprisinglylow 23 ± 8.8 and 21± 8.2 % degradation of 18.5 and 17.5 kDa MBP isoforms is observed inthe incubation of RmAb with GMF-KO myelin (Fig. 2 A, B). Notably, the levels of 18.5 and17.5 kDa isoforms of MBP in RmAb treated WT myelin is significantly different (18.5 kDa,p=0.01) (17.5 kDa, p=0.02) when compared to RmAb treated GMF-KO myelin (Fig. 2 A, B).On the contrary, 14 kDa MBP isoform from WT and GMF-KO myelin showed similar levelsof degradation to that of controls with RmAb antibody (Fig. 2 C).

Menon et al. Page 5

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

PLP degradation in samples incubated with anti-sulfatide R-mAb antibodies or cytokinesTo identify whether another major structural protein like MBP undergoes degradation, weanalyzed the extent of PLP degradation in the same samples analyzed for MBP degradation onthe same blots. Following incubation of myelin from WT and GMF-KO animals with anti-sulfatide antibodies or cytokines tested above showed no degradation of PLP (less than 8%)(Fig. 3). Nonetheless, the extent of degradation with RmAb is significantly different comparedto control (p=0.03, Fig. 3). On the contrary, GMF-KO myelin failed to show any degradation(Fig. 3).

Role of Calcium in MBP degradationTo further analyze the effect of RmAb mediated degradation of MBP is indeed orchestratedby calcium, we tested the degradation of 18.5, 17.5 and 14 kDa isoforms of MBP in presenceof EGTA. The results (Fig. 4) show that EGTA significantly blocked the RmAb induceddegradation of both 17.5 and 18.5 kDa MBP to less than 10% values (Fig. 4). Calcium aloneinduced more than 80% degradation of 18.5 and 17.5 kDa MBP. This effect was abolished byEGTA as shown by a less than 4% degradation of MBP even in presence of added calcium.Surprisingly no such degradation of 14 kDa MBP is evident in calcium treated samples (Fig.4) compared to controls and that presence of EGTA blocked the degradation of 14 kDa MBPsignificantly (Fig. 4).

Discrepancies in distribution of myelin proteins in myelin-axolemmal complexes isolatedfrom GMF-KO compared to WT

In order identify whether the distribution pattern of myelin proteins across myelin are affectedin GMF-KO mice, we fractionated myelin from both WT and GMF-KO mice according to themethod of Menon et al. (Menon et al. 2003). The major myelin proteins MBP, PLP and CNPfollowed a pattern similar to that seen before in WT, whereby MBP, CNP and PLP showedsignificant levels in compact myelin fractions (fractions 14-17) and minimal levels in heavyfraction B and P (Fig. 5, CNP, PLP, MBP and MOG panel) (Menon et al. 2003). Although,GMF-KO myelin exhibited similar pattern of distribution of MBP and CNP compared to WT,the heavy fraction B of PLP displayed lower levels of PLP (Fig. 5, boxed PLP panel). On theother hand, MOG and β-tubulin showed discrepancies in their distribution across fractions (Fig.5, MOG and β-tubulin boxed panel). Levels of MOG were higher compared to WT in LDBand fraction B (Fig. 5, boxed fractions LDB and B). In contrast, fraction 16 and 17 (one of thefractions in which the compact myelin floats) the levels of MOG in GMF-KO are reducedcompared to WT (Fig. 5) boxed fraction 16 and 17). The distribution of β-tubulin into fraction15 and 16 in GMF-KO myelin is strikingly higher compared to the same fractions in WT (Fig.5, boxed fractions 15, 16). These results suggest that myelin proteins MOG, PLP and β-tubulinare indeed affected in their distribution pattern across GMF-KO myelin.

DISCUSSIONMyelin is a multilamellar complex membranous structure found to contain a variety of proteinsthat are involved in metabolic turnover, signaling and the organization of myelin (Taylor et al.2004). Immune mediated attack on myelin is implicated as a major event in the myelinbreakdown in MS. Recently we demonstrated that GMF deficient mice fail to showdemyelination compared to WT when challenged with MOG (Zaheer et al., 2007). The absenceof demyelination observed in GMF-KO mice suggests that the immune response may becompromised in these animals. Thus, at first we analyzed the major cytokines that are knownto cause damage to oligodendrocytes from WT and GMF-KO animals challenged with MOG35-55 peptide. The levels of GM-CSF, TNF-α and INF-γ in GMF-KO mice brain weresignificantly lower compared to WT (Fig. 1). This suggests that the extremely low levels ofinflammatory cytokine response may have contributed to the absence of demyelination in

Menon et al. Page 6

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

GMF-KO mice. However, any biochemical difference in myelin of GMF-KO animalscompared to WT contributed to the lack of antibody mediated MBP breakdown observed inGMF-KO myelin remained unanswered. Thus, at first we incubated myelin isolated from WTand GMF-KO animals with anti-sulfatide (RmAb; an antibody directed to sulfatedgalactocerebroside) or different cytokines such as TNF-α and IFN-γ or GMCSF. Here wedemonstrate that anti-sulfatide antibody is capable of inducing an MBP degradation of both18.5 and 17.5 kDa isoforms in WT myelin like the anti-MOG and anti-GalC antibodies (Menonet al. 1997). This effect parallels the effects of RmAb on cultured oligodendrocytes and anti-GalC on myelin (Dyer and Benjamins 1988;Bansal and Pfeiffer 1989;Menon et al. 1997).Degradation of major MBP isoforms such as 18.5 and 17.5 kDa is important as 18.5 kDaisoform plays a major role in the compaction of myelin (Baumann and Pham-Dinh 2001;Harauzet al. 2004;Boggs 2006) and 17.5 kDa is implicated in the formation of compact myelin (Kimuraet al. 1998). Thus a significant break down of MBP could lead to myelinolysis and therebyinitiating the process of demyelination. No such MBP degradation was observed in myelinfrom GMF-KO mice (Fig. 2 A, B and C) suggesting that GMF-KO myelin is resistant to MBPdegradation. Although the cause behind the resistance of GMF-KO myelin to MBP degradationis not known, we hypothesize that myelin from GMF-KO mice may possess different structuraland biochemical properties. In fact extensive fractionation of GMF-KO myelin shows analtered distribution of some of the myelin proteins across fractions (Fig. 5) compared to WT.A major compact myelin fraction (Fig. 5, boxed MOG panel, fraction 16) from GMF-KO micerepresented low levels of MOG compared to WT. Similarly the LDB and fraction B (Fig. 5)also showed differential distribution of MOG compared to WT. Notably in GMF-KO, themyelin fractions 15 and 16 (one of the fractions where MOG is differentially distributedcompared to WT) (Fig. 5) showed higher levels of β-tubulin in contrast to WT (Fig. 5, β- tubulinpanel, fraction 15 and 16). Presence of β-tubulin in the compact myelin fractions confirms theassociation of β-tubulin to myelin demonstrated by proteomic analysis of myelin (Taylor et al.2004). The functional significance of MOG-tubulin association was indicated due to anti-MOGantibody induced β-tubulin dephosphorylation and repartition of MOG and β-tubulin into lipidrafts following MOG signaling in cultured oligodendrocytes (Marta et al. 2003).Dephosphorylation of β-tubulin has been shown to interfere with the microtubule stability(Atashi et al. 1992) and may be involved in the organization of myelin. Here we show that bothMOG and β-tubulin are differentially distributed in WT and GMF-KO myelin fractions.Floatation of β-tubulin into lighter density fractions, as we observed here, is similar to that seenwith Na channels and Caspr in ceramide galactosyl transferase (CGT) null mice compared toWT (Menon et al. 2003) suggesting that organization of myelin is affected. Similarly, changesassociated with myelin fractions have been shown in EAE (Salvati et al., 1989). Analysis ofmyelin light fractions following density gradient centrifugation from EAE induced animalsshowed abnormalities in myelin morphology compared to WT (Salvati et al., 1989). Themorphological abnormality is accompanied by a decrease in 2′:3′-cyclic nucleotide 3′-phosphodiesterase (CNPase) specific activity and the amount of myelin basic protein (Salvatiet al., 1989). Moreover, GMF-KO mice showed impaired motor performance and learning(Lim et al. 2004). Taken together these findings suggest that the biochemical organizationalcomplexity might be compromised in GMF-KO myelin. Nonetheless, the absence of MBPdegradation in the presence of added cytokines in GMF-KO myelin implies that biochemicalorganizational complexity in GMF-KO myelin may play a major role in the absence of MBPdegradation. Thus these two aspects together might have contributed to the lack ofdemyelination observed in in vivo.

Treatment of myelin with IFN-γ led to a significant increase in degradation of 18.5 kDa MBPisoform in WT compared to controls (Fig. 2 A). The result with IFN-γ in myelin is in line withthe exacerbation observed in MS patients following IFN-γ administration (Panitch et al.1987). This suggests that treatment with IFN-γ may have adverse effect on myelin.Interestingly, the results with IFN-γ on GMF-KO myelin is consistent with the results obtained

Menon et al. Page 7

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

with RmAb as the level of degradation of the same 18.5 kDa MBP isoform is significantlylower (Fig. 2 A) in GMF-KO myelin. This once again confirms the resistance of GMF-KOmyelin to MBP degradation. Incubation of myelin with TNF-α did not show any degradationof MBP either in WT or in GMF-KO myelin compared to control suggesting that mechanismresponsible for oligodendrocyte damage with TNF-α may not be operating in myelin. Thissuggests that the MBP degradation observed with RmAb and IFN-γ in WT myelin might bedue to specific mechanisms that respond to some of the immune components such as RmAband IFN-γ. It is clear from our study with EGTA and RmAb that the degradation of 18.5 and17.5 kDa MBP is mediated by calcium induced neutral proteases present in myelin (Fig. 4)and that the levels of 14 kDa isoform remained unaffected by calcium. Surprisingly, EGTAinhibited the degradation of 14 kDa MBP. It is likely that calcium may have its effect througha different mechanism other than antibody mediated signaling regarding 14 kDa MBPdegradation. Further studies are required to understand the mechanism by which how 14kDaMBP degradation is controlled by calcium. In this context the level of PLP, another importantstructural component of myelin following RmAb and cytokine treatment is noteworthy (Fig.3). Although WT myelin treated with RmAb showed a significant PLP degradation comparedto control, the extent of degradation remained below 10% suggesting that PLP remainedrelatively unaffected by the treatment with anti-sulfatide RmAb antibody or any of the cytokineanalyzed (Fig. 3). Interestingly, the level of degradation of PLP in GMF-KO myelin treatedwith RmAb is significantly lower compared to WT myelin treated with RmAb (Fig. 3); anobservation again consistent with the resistance of GMF-KO myelin to MBP degradationshown in this study.

In summary, we carried out experiments to understand the role of GMF in the induction ofEAE and cytokine response and its effect on MBP degradation in isolated myelin from WTand GMF-KO animals. We conclude that GMF-KO myelin is resistant to MBP degradationand is consistent with the absence of demyelination observed in GMF-KO animals exposed toMOG 35-55 peptide. Absence of an inflammatory response in GMF-KO myelin could havecontributed to the in vivo absence of demyelination. Our results on myelin fractionation indicatethat the biochemical organizational complexity could be one aspect that affected thedegradation of MBP in GMF-KO myelin. However, the mechanisms behind the resistance ofGMF-KO myelin to MBP degradation induced by RmAb and IFN-γ remains to be solved.Further investigations are essential to decipher the precise role played by GMF in the MBPbreakdown in GMF-KO myelin.

Acknowledgements

GMF-deficient mice were produced at the University of Iowa Gene Targeting Core Facility. We thank Ashna Zaheer,Scott Knight and Marcus Ahrens for excellent technical assistance. This work was supported by the Department ofVeterans Affairs Merit Review award (to A.Z.) and by the National Institute of Neurological Disorders and Strokegrant NS-47145 (to A.Z.).

The abbreviations used areGMF, glia maturation factor; GM-CSF, granulocytemacrophage-colony stimulating factor;TNF, tumor necrosis factor; IFN, interferon; GMF-KO, glia maturation factor-deficient mice;RmAb, anti-sulfatide antibodies; MBP, myelin basic protein; PLP, proteolipid protein; MOG,myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; EAE, experimental autoimmuneencephalomyelitis; CANP, calcium activated neutral proteases; RT-PCR, reversetranscription-polymerase chain reaction.

Menon et al. Page 8

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

REFERENCESAtashi JR, Klinz SG, Ingraham CA, Matten WT, Schachner M, Maness PF. Neural cell adhesion

molecules modulate tyrosine phosphorylation of tubulin in nerve growth cone membranes. Neuron1992;8:831–842. [PubMed: 1375036]

Balabanov R, Strand K, Kemper A, Lee JY, Popko B. Suppressor of cytokine signaling 1 expressionprotects oligodendrocytes from the deleterious effects of interferon-gamma. J Neurosci 2006;26:5143–5152. [PubMed: 16687505]

Bansal R, Pfeiffer SE. Reversible inhibition of oligodendrocyte progenitor differentiation by amonoclonal antibody against surface galactolipids. Proceedings of the National Academy of Sciencesof the United States of America 1989;86:6181–6185. [PubMed: 2668957]

Baumann N, Pham-Dinh D. Biology of oligodendrocyte and myelin in the mammalian central nervoussystem. Physiol Rev 2001;81:871–927. [PubMed: 11274346]

Bernard CC, Johns TG, Slavin A, Ichikawa M, Ewing C, Liu J, Bettadapura J. Myelin oligodendrocyteglycoprotein: a novel candidate autoantigen in multiple sclerosis. J Mol Med 1997;75:77–88.[PubMed: 9083925]

Boggs JM. Myelin basic protein: a multifunctional protein. Cell Mol Life Sci 2006;63:1945–1961.[PubMed: 16794783]

de Rosbo NK, Bernard CC. Multiple sclerosis brain immunoglobulins stimulate myelin basic proteindegradation in human myelin: a new cause of demyelination. Journal of neurochemistry 1989;53:513–518. [PubMed: 2473168]

Dyer CA, Benjamins JA. Antibody to galactocerebroside alters organization of oligodendroglialmembrane sheets in culture. J Neurosci 1988;8:4307–4318. [PubMed: 2460596]

Glynn P, Linington C. Cellular and molecular mechanisms of autoimmune demyelination in the centralnervous system. Critical reviews in neurobiology 1989;4:367–385. [PubMed: 2655941]

Gold R, Linington C, Lassmann H. Understanding pathogenesis and therapy of multiple sclerosis viaanimal models: 70 years of merits and culprits in experimental autoimmune encephalomyelitisresearch. Brain 2006;129:1953–1971. [PubMed: 16632554]

Harauz G, Ishiyama N, Hill CM, Bates IR, Libich DS, Fares C. Myelin basic protein-diverseconformational states of an intrinsically unstructured protein and its roles in myelin assembly andmultiple sclerosis. Micron 2004;35:503–542. [PubMed: 15219899]

Johns TG, Kerlero de Rosbo N, Menon KK, Abo S, Gonzales MF, Bernard CC. Myelin oligodendrocyteglycoprotein induces a demyelinating encephalomyelitis resembling multiple sclerosis. J Immunol1995;154:5536–5541. [PubMed: 7537310]

Jurewicz A, Matysiak M, Tybor K, Kilianek L, Raine CS, Selmaj K. Tumour necrosis factor-induceddeath of adult human oligodendrocytes is mediated by apoptosis inducing factor. Brain2005;128:2675–2688. [PubMed: 16219674]

Kaplan R, Zaheer A, Jaye M, Lim R. Molecular cloning and expression of biologically active human gliamaturation factor-beta. Journal of neurochemistry 1991;57:483–490. [PubMed: 1712830]

Keirstead HS, Hughes HC, Blakemore WF. A quantifiable model of axonal regeneration in thedemyelinated adult rat spinal cord. Experimental Neurology 1998;151:303–313. [PubMed: 9628765]

Kimura M, Sato M, Akatsuka A, Saito S, Ando K, Yokoyama M, Katsuki M. Overexpression of a minorcomponent of myelin basic protein isoform (17.2 kDa) can restore myelinogenesis in transgenicshiverer mice. Brain Res 1998;785:245–252. [PubMed: 9518636]

Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature1970;227:680–685. [PubMed: 5432063]

Lim R, Zaheer A. In vitro enhancement of p38 mitogen-activated protein kinase activity byphosphorylated glia maturation factor. The Journal of biological chemistry 1996;271:22953–22956.[PubMed: 8798479]

Lim R, Miller JF, Zaheer A. Purification and characterization of glia maturation factor beta: a growthregulator for neurons and glia. Proceedings of the National Academy of Sciences of the United Statesof America 1989;86:3901–3905. [PubMed: 2726756]

Menon et al. Page 9

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Lim R, Zaheer A, Lane WS. Complete amino acid sequence of bovine glia maturation factor beta.Proceedings of the National Academy of Sciences of the United States of America 1990;87:5233–5237. [PubMed: 2196564]

Lim R, Zaheer A, Yorek MA, Darby CJ, Oberley LW. Activation of nuclear factor-kappaB in C6 ratglioma cells after transfection with glia maturation factor. Journal of neurochemistry 2000;74:596–602. [PubMed: 10646510]

Lim R, Zaheer A, Khosravi H, Freeman JH Jr. Halverson HE, Wemmie JA, Yang B. Impaired motorperformance and learning in glia maturation factorknockout mice. Brain Res 2004;1024:225–232.[PubMed: 15451385]

Marta CB, Taylor CM, Coetzee T, Kim T, Winkler S, Bansal R, Pfeiffer SE. Antibody cross-linking ofmyelin oligodendrocyte glycoprotein leads to its rapid repartitioning into detergent-insolublefractions, and altered protein phosphorylation and cell morphology. J Neurosci 2003;23:5461–5471.[PubMed: 12843245]

Menon K, Rasband MN, Taylor CM, Brophy P, Bansal R, Pfeiffer SE. The myelin-axolemmal complex:biochemical dissection and the role of galactosphingolipids. Journal of neurochemistry 2003;87:995–1009. [PubMed: 14622129]

Menon KK, Piddlesden SJ, Bernard CC. Demyelinating antibodies to myelin oligodendrocyteglycoprotein and galactocerebroside induce degradation of myelin basic protein in isolated humanmyelin. Journal of neurochemistry 1997;69:214–222. [PubMed: 9202313]

Norton WT, Poduslo SE. Myelination in rat brain: method of myelin isolation. Journal of neurochemistry1973;21:749–757. [PubMed: 4271082]

Ozawa K, Saida T, Saida K, Nishitani H, Kameyama M. In vivo CNS demyelinationmediated by anti-galactocerebroside antibody. Acta Neuropathologica (Berlin) 1989;7:621–628. [PubMed: 2750480]

Panitch HS, Hirsch RL, Haley AS, Johnson KP. Exacerbations of multiple sclerosis in patients treatedwith gamma interferon. Lancet 1987;1:893–895. [PubMed: 2882294]

Pouly S, Becher B, Blain M, Antel JP. Interferon-gamma modulates human oligodendrocytesusceptibility to Fas-mediated apoptosis. J Neuropathol Exp Neurol 2000;59:280–286. [PubMed:10759183]

Rosenbluth J, Schiff R, Liang WL, Dou W. Antibody-mediated CNS demyelination II. Focal spinal cordlesions induced by implantation of an IgM antisulfatide-secreting hybridoma. J Neurocytol2003;32:265–76. [PubMed: 14724389]

Salvati S, D’Urso D, Conti Devirgiliis L, Serlupi Crescenzi G. Biochemical changes in central nervoussystem membranes in experimental allergic encephalomyelitis. J Neurochem 1986;47:239–244.[PubMed: 3011992]

Schaecher K, Rocchini A, Dinkins J, Matzelle DD, Banik NL. Calpain expression and infiltration ofactivated T cells in experimental allergic encephalomyelitis over time: increased calpain activitybegins with onset of disease. J Neuroimmunol 2002;129:1–9. [PubMed: 12161014]

Schaecher KE, Shields DC, Banik NL. Mechanism of myelin breakdown in experimental demyelination:a putative role for calpain. Neurochemical research 2001;26:731–737. [PubMed: 11519732]

Steinman L, Zamvil SS. Virtues and pitfalls of EAE for the development of therapies for multiplesclerosis. Trends Immunol 2005;26:565–571. [PubMed: 16153891]

Taylor CM, Marta CB, Claycomb RJ, Han DK, Rasband MN, Coetzee T, Pfeiffer SE. Proteomic mappingprovides powerful insights into functional myelin biology. Proceedings of the National Academy ofSciences of the United States of America 2004;101:4643–4648. [PubMed: 15070771]

Zaheer A, Fink BD, Lim R. Expression of glia maturation factor beta mRNA and protein in rat organsand cells. Journal of neurochemistry 1993;60:914–920. [PubMed: 8436977]

Zaheer A, Zhong W, Uc EY, Moser DR, Lim R. Expression of mRNAs of multiple growth factors andreceptors by astrocytes and glioma cells: detection with reverse transcription-polymerase chainreaction. Cellular and molecular neurobiology 1995;15:221–237. [PubMed: 8590453]

Zaheer A, Lim R. In vitro inhibition of MAP kinase (ERK1/ERK2) activity by phosphorylated gliamaturation factor (GMF). Biochemistry 1996;35:6283–6288. [PubMed: 8639570]

Zaheer A, Lim R. Overexpression of glia maturation factor (GMF) in PC12 pheochromocytoma cellsactivates p38 MAP kinase, MAPKAP kinase-2, and tyrosine hydroxylase. Biochemical andbiophysical research communications 1998;250:278–282. [PubMed: 9753620]

Menon et al. Page 10

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Zaheer A, Mathur SN, Lim R. Overexpression of glia maturation factor in astrocytes leads to immuneactivation of microglia through secretion of granulocyte-macrophage-colony stimulating factor.Biochemical and biophysical research communications 2002;294:238–244. [PubMed: 12051700]

Zaheer A, Zaheer S, Sahu SK, Yang B, Lim R. Reduced Severity of Experimental AutoimmuneEncephalomyelitis in GMF-Deficient Mice. Neurochem Res 2007;32:39–47. [PubMed: 17151915]

Menon et al. Page 11

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig 1.Expression profiles of GM-CSF, TNF-α and IFN-γ in brain of WT and GMF-KO micemeasured by quantitative real-time RT-PCR. Note the significantly lower levels of GM-CSF,TNF-α and IFN-γ in brain from GMF-KO mice challenged with MOG 35-55 compared to WTmice.

Menon et al. Page 12

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig 2.The extent of MBP degradation in myelin following 60 min incubation of WT and GMF-KOmyelin with anti-sulfatide RmAb antibody or with different cytokines was assessed byelectrophoresis of 1 μg of total myelin proteins. Data are average of five different sets of myelinexcept in case of control and GMF-KO Igs treated myelin where the data was the average ofthree. (A) The extent of 18.5 kDa isoform MBP degradation with RmAb, GM-CSF, TNF-αand IFN-γ. Note the significant degradation of 18.5 kDa isoform of MBP with RmAb (*p=0.002 vs control sham, 0.02 vs control Igs) and INF-γ (# p=0.01 vs control) in WT myelin.(B) The extent of 17.5 kDa isoform of MBP degradation with RmAb, GM-CSF, TNF-α andIFN-γ. An elevated degradation of 17.5 kDa MBP is seen with RmAb (** p=0.01 vs control

Menon et al. Page 13

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

sham and control Igs). Notably, GMF-KO myelin did not show an elevated degradation of18.5, 17.5 or 14 kDa isoforms of MBP compared to controls (panel A, B and C).

Menon et al. Page 14

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig 3.The extent of PLP degradation in myelin following 60 min incubation of WT and GMF-KOmyelin with different antibodies and cytokines. A 1μg of total myelin proteins wereelectrophoresed. Data are average of five different sets of myelin except in case of control andGMF-KO Igs treated myelin where the data was the average of three. (A) Histograms showinglevels of PLP degradation in myelin incubated with GM-CSF, TNF-α, IFN-γ, anti-sulfatideRmAb or control antibody following 60 min incubation. Although an elevated degradation ofPLP could be seen in RmAb treated myelin compared to controls, the levels of MBPdegradation is below 10%. Note the low levels of degradation in PLP compared to MBP. (B)Representative immunoblot showing levels of PLP degradation following 60 min incubationof WT and GMF-KO myelin with various cytokines and antibodies.

Menon et al. Page 15

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig 4.MBP degradation following incubation of myelin with RmAb in presence or absence of EGTAwas assessed by electrophoresis of 1 μg of total myelin proteins. Data are average of threedifferent sets of myelin. Note the significantly low degradation of MBP in presence of EGTA.Level of significance for different isoforms of MBP are (a) for 18.5 kDa MBP (Control IgGvs RmAb * p=0.01; RmAb vs RmAb EGTA **p=0.005; Ca vs Ca EGTA ***p=<0.00001) (b)for 17.5 kDa MBP (Control IgG vs RmAb #p=0.01; RmAb vs RmAb EGTA ##p=0.002; Cavs Ca EGTA ###p=0.0007) and (c) for 14 kDa MBP (Control IgG vs RmAb +p=0.15; RmAbvs RmAb EGTA ++p=0.1; Ca vs Ca EGTA +++p=0.002) respectively.

Menon et al. Page 16

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig 5.A.Characterization of WT (+) and GMF-KO (-) myelin to identify the distribution pattern ofmyelin proteins in myelin-axolemmal fractions by western blotting. The boxed areas representfractions showing differential distribution of proteins between WT and GMF-KO myelin. Notethat β-tubulin in GMF-deficient myelin floats into much lighter density fractions than in WT.The panel representing MOG, tubulin, CNP, PLP and MBP are from the same blot. LDB, low-density band; LD, Light dispersion; HD, heavy dispersion; A and B heavy density fractions;P, pellet. B. Schematic representation of the myelin distribution and the fractions analyzed inpanel A following 15 h density gradient centrifugation.The data shown are representative of at least three experiments.

Menon et al. Page 17

Neurosci Res. Author manuscript; available in PMC 2008 June 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript