B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r. com/ l oca te /b ra in res

Research Report

Increased phosphorylation of caveolin-1 in the sciatic nerves ofLewis rats with experimental autoimmune neuritis

Heechul Kima,1, Changjong Moonb,1, Meejung Ahna, Yoh Matsumotoc,Chang-Sung Kohd, Moon-Doo Kime, Taekyun Shina,⁎aDepartment of Veterinary Medicine, Cheju National University, Jeju 690-756, South KoreabDepartment of Veterinary Anatomy, College of Veterinary Medicine, Chonnam National University, Gwangju 500-757, South KoreacDepartment of Molecular Neuropathology, Tokyo Metropolitan Institute for Neuroscience, Fuchu, Tokyo 183, JapandDepartment of Biomedical Laboratory Sciences, Shinshu University School of Health Sciences, 3-1-1 Asahi, Matsumoto 390-8621, JapaneDepartment of Psychiatry, Cheju National University College of Medicine, Jeju 690-756, South Korea

A R T I C L E I N F O

⁎ Corresponding author. Fax: +64 7563354.E-mail address: shint@cheju.ac.kr (T. Shin

1 The first two authors contributed to this

0006-8993/$ – see front matter © 2006 Elsevidoi:10.1016/j.brainres.2006.12.034

A B S T R A C T

Article history:Accepted 13 December 2006Available online 20 December 2006

The levels of phosphorylated caveolin-1 (p-caveolin-1) were analyzed in the sciatic nerves ofLewis rats with experimental autoimmune neuritis (EAN). Western blot analysis showedthat the phosphorylation of caveolin-1 increased significantly in the sciatic nerves of EAN-affected rats at the paralytic stage of EAN on day 14 post-immunization (PI) (P<0.05) anddeclined slightly thereafter during the recovery stage. Immunohistochemistry showedintense p-caveolin-1 immunostaining in some inflammatory macrophages, as well as in T-cells in individual nerve fascicles at the peak stage of EAN, while p-caveolin-1 was weaklyexpressed in some of the vascular endothelial cells and Schwann cells of normal sciaticnerves. The inflammatory cells with intense p-caveolin-1 expression in the EAN-affectedindividual nerve fascicles were not positive for terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL), while the TUNEL-positive apoptotic cells in theperineurium, where infiltration initially occurred, were weakly positive for p-caveolin-1.Based on these findings, we postulate that caveolin-1 is phosphorylated in inflammatorycells soon after they infiltrate the sciatic nerve, as well as in the perineurium, and thatp-caveolin-1 activates intracellular signaling in inflammatory cells, leading to cell death,which ultimately eliminates the infiltrating inflammatory cells from the sciatic nerves ofanimals with EAN.

© 2006 Elsevier B.V. All rights reserved.

Keywords:Caveolin-1Experimental autoimmune neuritisMacrophagePhosphorylationSchwann cell

1. Introduction

Experimental autoimmune neuritis (EAN), which is a T-cell-mediated autoimmune disease of the peripheral nervoussystem (PNS), is used as a model of human demyelinatingdiseases, such as Guillain–Barré Syndrome (Hartung and

).work equally.

er B.V. All rights reserved

Toyka, 1990; Zhu et al., 1998). The clinical course of EAN ischaracterized by weight loss, ascending progressive paralysis,and spontaneous recovery. It has been proposed that inflam-matorymediators produced in the affected spinal nerves rootsand sciatic nerves are involved in the pathogenesis of EAN(Zhu et al., 1998). EAN lesions in susceptible animals are

.

154 B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

characterized by T cells and macrophages infiltrating into thesciatic nerves during the peak stages of the disease (Moon etal., 2005, 2006). In addition, the apoptotic elimination ofinflammatory cells is a common feature of autoimmunediseases, such as EAN (Moon et al., 2005). Although manyfactors are involved in the process of apoptosis in EAN lesions(Conti et al., 1998; Moon et al., 2005), little is known about theactivation of lipid raft proteins, such as caveolins.

Caveolin is a transmembrane adaptor molecule thatrecognizes glycosylphosphatidyl inositol-linked proteins andinteracts with downstream cytoplasmic signaling molecules,which include the Src-family tyrosine kinases and hetero-trimeric G proteins (Li et al., 1996; Okamoto et al., 1998). Undercertain activation conditions, caveolin is phosphorylated, andthe phosphorylation of caveolin at Tyr-14, Ser-88, and otherresidues in v-Src-transformed cells leads to the flattening,aggregation, and fusion of caveolae and caveolae-derivedvesicles (Li et al., 1996). In addition, the phosphorylated formof caveolin-1 (p-caveolin-1) is thought to be a downstreamelement of p38 mitogen-activated protein (MAP) kinase and c-Src in NIH 3T3 cells (Volonte et al., 2001), which are alsoimportant signals in inflammatory diseases.

In autoimmune disease models, we have reported that theexpression of caveolins is significantly elevated in the spinal

Fig. 1 – Histopathological examination of normal and EAN-affecpresent in the sciatic nerves of the normal control rats (A). On danerves (B). In the EAN-affected sciatic nerve (D14PI), the majorityT cells (C) and ED1-positive macrophages (D) in the individual neHematoxylin–eosin staining. (C and D) Immunostained with eithScale bars: in A and B, 80 μm; in C and D, 40 μm; in inset, 30 μm

cord in experimental autoimmune encephalomyelitis (EAE)(Shin et al., 2005) and in the sciatic nerve in EAN (Ahn et al.,2006). Subsequently, we have found that the level of caveolin-1 phosphorylation increases in the spinal cords of EAEanimals, particularly in the activated microglia and macro-phages present in the EAE lesions (Kim et al., 2006). In EAE,caveolin-1 is not phosphorylated in all caveolin-1-positivecells (Shin et al., 2005). These findings suggest that caveolin-1is phosphorylated in a limited cell population (mainly macro-phages and activated microglia) in rat EAE. Although a varietyof cells in the EAN lesions express caveolin-1 (Ahn et al., 2006),little is known about the status of caveolin-1 phosphorylationin the sciatic nerves of rats with EAN. Thus, we examined thephosphorylation of caveolin-1 in the PNS in autoimmune-targeted sciatic nerves.

2. Results

2.1. Clinical progression of experimental autoimmuneneuritis and histopathological findings

Rats with EAN immunized with the SP26 peptide developedfloppy tails (grade 1, G1) on days 10–12 post-immunization (PI)

ted sciatic nerves of rats. No inflammatory cells werey 14 PI, many inflammatory cells were found in the sciaticof the inflammatory cells were R73 (TCR αβ)-positiverve fascicles and perineurium (insets, arrow). (A and B)er R73 (C) or ED1 (D) and counterstained with hematoxylin..

155B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

and exhibited progressive hindlimb paralysis (G2 or G3) ondays 14–16 PI. All of the rats recovered on days 21–30 PI(recovery 0, R0).

Histopathological examination showed no infiltrating cellsin the sciatic nerves of the normal controls (Fig. 1A). With theonset of tail weakness and hindlimb paralysis, inflammatorycells infiltrated the perineurium of the sciatic nerve, as well asthe individual nerve fascicles (day 14 PI) (Fig. 1B). In the EANlesions, most of the inflammatory cells consisted of R73 (TCRαβ)-positive T cells (Fig. 1C) and ED1-positive macrophages(Fig. 1D). Most of the inflammatory cells in the EAN lesionsafter recovery from paralysis were gradually eliminatedthrough apoptosis (data not shown). The clinical observationsand histological findings related to EAN largely correspond tothose described previously (Moon et al., 2006).

2.2. Increased phosphorylation of caveolin-1 inEAN-affected sciatic nerves

The level of the phosphorylated form of caveolin-1 (p-caveolin-1) was evaluated semiquantitatively in the sciaticnerves during the course of EAN using Western blot analysis.The phosphorylation of caveolin-1 increased significantly inthe sciatic nerve in EAN at the early (day 10 PI; about 3.7-fold)and peak (day 14 PI; about 4.9-fold) stages compared to theexpression in normal controls (P<0.05, n=3/time point). Thelevel of p-caveolin-1 in the sciatic nerves in EAN at day 21 PIwas still higher than that of normal controls (P<0.05; about

Fig. 2 – Western blot with anti-p-caveolin-1 (Tyr-14)antibody of the sciatic nerves of normal control rats (Normal)and EAN-affected rats. Representative photographs ofWestern blots for p-caveolin-1 and β-actin. Lane 1, normalcontrol; lane 2, day 10 PI; lane 3, day 14 PI; lane 4, day 21 PI;lane 5, day 30 PI. Arrowheads indicate the positions ofp-caveolin-1 (approximately 22 kDa) and β-actin (45 kDa).Bar graph: the results of the densitometric data analysis(mean±SE, n=3 rats/group). The relative expression ofp-caveolin-1 is calculated after normalization to the β-actinbands from three different samples; the level of expression inthe normal controls is defined arbitrarily as 1. *P<0.05 vs.normal controls.

3.9-fold) (Fig. 2). At day 30 PI, the phosphorylation decreasedclose to that of normal controls.

2.3. Localization of p-caveolin-1 in EAN-affected rat sciaticnerves

Immunohistochemically, p-caveolin-1 was seen constitu-tively in a few vascular endothelial cells (Fig. 3A, arrowhead)and Schwann cells (Fig. 3A, arrow) in the sciatic nerves ofnormal rats. At the peak stage of EAN (day 14 PI), theinflammatory cells in the individual nerve fascicles wereimmunopositive for p-caveolin-1 (Fig. 3B, arrows), while the p-caveolin-1 immunoreactivity had not changed in the Schwanncells (Fig. 3C, arrow) or vascular endothelial cells (Fig. 3C,arrowhead) of EAN-affected rats compared to the controls.These findings suggest that the level of phosphorylation in theSchwann cells and vascular endothelial cells of EAN rats isnegligible. In the recovery stage of EAN (day 21 PI), theindividual nerve fascicles contained fewer inflammatorycells than at the peak stage, and some inflammatory cellswere positive for p-caveolin-1 (Fig. 3D).

2.4. Cell phenotypes associated with p-caveolin-1 inEAN-affected rat sciatic nerves

In the individual nerve fascicles of EAN lesions (day 14 PI), p-caveolin-1 immunoreactivity (Fig. 4A; green, arrows) co-localized with some ED1-positive cells (Fig. 4B; red, arrows),which suggests that macrophages express p-caveolin-1 in thesciatic nerves of EAN rats (Fig. 4C; merge, arrows). Further-more, in the EAN lesions (individual nerve fascicles), some IB4-positive vascular endothelial cells (arrow) and IB4-positivemacrophages (arrowheads) (Figs. 4D–F) were immunopositivefor p-caveolin-1. In the sciatic nerves, p-caveolin-1 immunor-eactivity (Fig. 4G; green, arrows) occasionally co-localized withGFAP-positive Schwann cells (Fig. 4H; red, arrows), whichindicates that very few Schwann cells express p-caveolin-1(Fig. 4I; merge, arrows).

In frozen sections that reactedwith rabbit anti-p-caveolin-1and R73, p-caveolin-1 immunoreactivity was co-localized inR73-positive cells, which suggests that the majority of inflam-matory cells are positive for p-caveolin-1. These findingsmatch the single-immunostaining results (data not shown).

2.5. Relationship between apoptosis and caveolin-1phosphorylation in EAN lesions

The relationship between apoptosis and the phosphorylationof caveolin-1 was examined in EAN-affected rat sciatic nervesbecause the majority of inflammatory cells in EAN areeliminated by apoptosis (Conti et al., 1998; Moon et al., 2005),which leads to spontaneous recovery from paralysis. On day14 PI, TUNEL-positive cells were detected in the sciatic nerves,and many more TUNEL-positive cells were seen in theperineurium than in the individual nerve fascicles (Fig. 5B).Moreover, some p-caveolin-1 immunoreactivity was observedin these areas (Fig. 5A). Importantly, we found a massiveinfiltration of inflammatory cells (ED1 and/or R73-positivecells) in the individual nerve fascicles and the involved cellswere strongly positive for p-caveolin-1 (Fig. 5D), althoughmost

Fig. 3 – Immunohistochemical detection of p-caveolin-1 in the sciatic nerves of rats with EAN. In the normal controls,p-caveolin-1 was present at low intensity in some vessels (A, arrowhead) and Schwann cells (A, arrow) of the sciatic nerves. InEAN-affected sciatic nerves on day 14 PI (B and C), p-caveolin-1 immunoreactivity was localized mainly to the inflammatorycells (B, arrows), although some Schwann cells (C, arrow) and vessels (C, arrowhead) were weakly positive for p-caveolin-1. InEAN-affected sciatic nerves on day 21 PI (D), some p-caveolin-1 immunoreactive cells remained (D, arrows). Counterstainedwith hematoxylin. Scale bars=40 μm.

156 B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

of these cells were not TUNEL-positive (Figs. 5E, F, merge).However, in the perineurium, we observed many TUNEL-positive cells (Fig. 5A, inset) that were very weakly positive forp-caveolin-1 (Figs. 5B, C, inset).

In summary, the phosphorylation of caveolin-1 in inflam-matory cells was inversely related to TUNEL positivity, whichsuggests that early activated inflammatory cells expressp-caveolin-1 and then undergo apoptosis. This may explainwhy the TUNEL-positive cells in the perineurium expressedp-caveolin-1 very weakly or not at all.

3. Discussion

In accordance with our previous report on the increasedexpression of caveolin-1 in EAN lesions (Ahn et al., 2006), wehave confirmed that increased phosphorylation of caveolin-1occurs in EAN lesions, particularly at the peak stage of EANwhen inflammatory cells infiltrate the sciatic nerve.

Experimental autoimmune neuritis lesions in susceptibleanimals are characterized by T cells and macrophagesinfiltrating the sciatic nerves during the peak stage of thedisease (Moon et al., 2005, 2006), and these cells were affectedby pro-inflammatory cytokines at the early and peak stages(Zhu et al., 1997). In particular, more cells expressed inter-leukin-1β (IL-1β) mRNA at the early stage than during therecovery stage of EAN (Zhu et al., 1997). IL-1β induces

phosphorylation of caveolin-1 in HIT-T15 cells (Veluthakal etal., 2005). Consequently, it is highly probable that the IL-1βinduces the phosphorylation of caveolin-1 in T cells ormacrophages, leading to the inflammatory events in EAN.

Although the biological relevance of caveolin-1 phosphor-ylation in inflammatory cells remains to be determined, ageneral consensus exists that the phosphorylation of caveo-lin-1 represents a downstream element of p38MAP kinase andc-Src in NIH 3T3 cells (Volonte et al., 2001), which is one of theimportant MAP kinases in the course of rat EAN.

Regarding the fate of inflammatory cells in rat EAN, someinflammatory cells are eliminated through apoptosis via theincreased expression of the phosphorylated form of p38 (Moonet al., 2005). Matching the related results of increases in both p-caveolin-1 and phosphorylation of p38 described above, itseems likely that the increased phosphorylation of p38 ininflammatory cells (presumably T cells and macrophages)stimulates the phosphorylation of caveolin-1 in the sciaticnerves of rats with EAN, leading to the apoptosis of inflam-matory cells.

With respect to the phosphorylation of caveolin-1 inSchwann cells in EAN rats, we found that a few Schwanncells phosphorylated caveolin-1 constitutively and that fewapoptotic Schwann cells were present. Since the activation ofthe cell survival MAP kinases, which include the extracellularsignal-regulated kinase (ERK), is reported to be enhancedgreatly in EAN lesions (Ahn et al., 2004) and the expression of

Fig. 4 – Identification of p-caveolin-1-positive cells in the sciatic nerves of Lewis rats with EAN (day 14 PI). In the individualnerve fascicles of the EAN-affected sciatic nerves, some p-caveolin-1 immunopositive inflammatory cells were co-localized inED1-positive macrophages (A–C, arrows) or isolectin B4 (IB4)-positive macrophages (D–F, arrowheads). In addition, somep-caveolin-1-positive cells were co-localized in IB4-positive vascular endothelial cells (D–F, arrows) or GFAP-positive Schwanncells (G–J, arrows). C, F, and I are merged images. Scale bars=20 μm.

157B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

p-p38 in Schwann cells is limited (Moon et al., 2005), wepostulate that Schwann cell death is limited, if it occurs at all,through the activation of ERK. However, we cannot excludethe possibility that some Schwann cells were eliminatedthrough apoptosis in EAN, even with the activation ofcaveolin-1 in Schwann cells when death signals were subse-quently activated.

Reviewing the previous studies on the relationshipbetween p-caveolin-1 and either p38 MAP kinase or ERK, asdescribed above, we postulate that p-caveolin-1 expression bySchwann cells is very rarely linked to the apoptosis viaactivation of the p38 MAPK pathway of Schwann cells in thesciatic nerves of rats with EAN (Conti et al., 1998; Volonte et al.,2001; Ahn et al., 2004; Moon et al., 2005).

In the EAN lesions, we found intense p-caveolin-1 immu-noreactivity in inflammatory cells within the individual nervefascicles but very weak immunoreactivity in the perineurium(Fig. 5). This finding was inversely related to the apoptosisvisualized using TUNEL. We postulate that autoreactiveinflammatory cells initially infiltrate the perineurium in EANand express p-caveolin-1 and that this is followed by

apoptosis. Inflammatory cells in the individual nerve fasci-cles meet the same fate at a later time point. This explainsthe inverse relationship between the TUNEL positivity andp-caveolin-1 immunoreactivity observed in the EAN lesionsat the peak stage of the disease.

In summary, we postulate that the increased phosphoryla-tion of caveolin-1 in EAN rats is closely associated with theactivation of inflammatory cell death, which leads to func-tional recovery from EAN paralysis. The precise role of theincreased phosphorylation of caveolin-1 in autoimmunediseases of the PNS, including Guillain–Barré Syndrome andits animal model EAN, remains to be elucidated.

4. Experimental procedures

4.1. Animals

Lewis rats were obtained from Harlan (Indianapolis, IN, USA)and bred in our animal facility. Female rats aged 7–12 weeksand weighing 160–200 g were used in the experiments. This

Fig. 5 – Double staining for apoptosis and p-caveolin-1 expression of EAN-affected sciatic nerves at day 14 PI. The TUNELreaction and p-caveolin-1 immunoreactivity were detected in the perineurium and individual nerve fascicles of the sciaticnerves of EAN rats (A–C). Weak p-caveolin-1 immunoreactivity is detected in the perineurium (A, inset), whereas manyTUNEL-positive cells were observed (B, inset), and p-caveolin-1 and TUNEL were co-localized in a few cells (C, inset). Thep-caveolin-1 immunoreactivitywas localizedmainly to the inflammatory cells (D, green) in the individual nerve fascicles, whilevery few of these cells were TUNEL-positive (E, red; F, merge). Scale bars: in A–C, 40 μm; in D–I and insets, 20 μm. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

158 B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

research was conducted in accordance with the internation-ally accepted principles for laboratory animal use and care, asdescribed in the NIH guidelines (Bethesda, MD, USA).

4.2. EAN induction

Active EAN was induced in the Lewis rats, as describedpreviously (Ahn et al., 2004; Moon et al., 2005). Each rat wasinjected in both hind footpads with an emulsion thatcontained 100 μg of SP26 (Shimadzu, Kyoto, Japan) andFreund's complete adjuvant (CFA; M. tuberculosis H37Ra,5 mg/ml) and was evaluated clinically. Each rat was treatedwith 500 ng of pertussis toxin (Sigma, St. Louis, MO, USA) ondays 0 and 2 post-immunization (PI). The progress of EAN wasdivided into seven clinical stages: Grade (G) 0, no signs; G1,floppy tail; G2, mild paraparesis; G3, severe paraparesis; G4,tetraparesis; G5, moribund condition or death; and R0, therecovery stage (Matsumoto et al., 2000). On days 10, 14, 21, and30 PI, rats were killed under ether anesthesia and 5 cm of thesciatic nerve was removed bilaterally. Control rats wereimmunized with CFA only.

4.3. Antibodies and antisera

Rabbit polyclonal anti-p-caveolin-1 (Tyr-14, catalog #sc-14037-R; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and mousemonoclonal anti-β-actin (Sigma) antibodies were used in thisstudy. To identify the cell phenotypes, the following mono-

clonal antibodies were used: mouse monoclonal anti-glialfibrillary acidic protein (GFAP; Sigma); mouse monoclonalanti-rat macrophage (ED1; Serotec, London, UK); and mousemonoclonal anti-T-cell receptor αβ (R73; Blackthorn, Bicester,UK). Biotinylated isolectin B4 (IB4) derived from Griffoniasimplicifolia (Sigma) was used to visualize vascular endothelialcells and macrophages (Kim et al., 2006).

4.4. Tissue sampling

To study the phosphorylation of caveolin-1 in rats with EAN,tissues were sampled on days 10, 14, 21, and 30 PI. The sciaticnerves were obtained from CFA-immunized control rats onday 14 PI. The sciatic nerves of three rats were removed andfrozen at −70 °C for protein analysis. To prepare frozensections, three rats were killed at each stage PI, and the sciaticnerves were snap-frozen in OCT compound (Sakura, Tokyo,Japan). Eight-micrometer-thick sections were cut in a cryostatand stored at −70 °C until use. In addition, the sciatic nerves ofthree rats were embedded in paraffin after fixation in 4%paraformaldehyde in phosphate-buffered saline (PBS; pH 7.4).

4.5. Immunoblotting

The sciatic nerves were dissected out, minced, homogenized,and lysed in a buffer that contained 40 mM Tris–HCl (pH 7.4),120 mM NaCl, 2 mM sodium orthovanadate, and 0.1% NonidetP-40 (polyoxyethylene [9] p-t-octyl phenol) supplemented with

159B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

the protease inhibitors leupeptin (0.5 μg/ml), PMSF (1mM), andaprotinin (5 μg/ml). The homogenates were transferred tomicrotubes and centrifuged at 14,000 rpm for 20 min, and thesupernatant was harvested. For the immunoblot assay, super-natants that contained 40 μg of protein were loaded into thelanes of 12% SDS–PAGE gels, electrophoresed, and immuno-blotted onto nitrocellulose membranes (Schleicher andSchuell, Keene, NH, USA). The residual binding sites on themembrane were blocked by incubation with 5% nonfat milk inTris-buffered saline (TBS; 10 mM Tris–HCl [pH 7.4], 150 mMNaCl) for 1 h and then incubated with rabbit polyclonal anti-p-caveolin-1 (Tyr-14) antibody for 2 h.

The blots were washed three times in TBS that contained0.1% Tween 20 and then incubated with horseradish perox-idase-conjugated anti-rabbit IgG (Vector Laboratories, Burlin-game, CA, USA) for 1 h. The bound antibody was detectedusing enhanced chemiluminescence reagents (AmershamBiosciences, Arlington Heights, IL, USA) according to themanufacturer's instructions. After imaging, the membraneswere stripped and re-probedwith themonoclonal anti-β-actinantibody as the primary antibody (Sigma). The density (OD/mm2) of each band was measured with a scanning laserdensitometer (GS-700; Bio-Rad, Hercules, CA, USA). The resultswere analyzed statistically using one-way analysis of variancefollowed by the post hoc Student–Newman–Keuls t-test formultiple comparisons. In all cases, P<0.05 was considered tobe statistically significant.

4.6. Immunohistochemistry

Briefly, paraffin-embedded sciatic nerves sections (5 μm) weredeparaffinized, treated with citrate buffer (0.01 M, pH 6.0) in amicrowave for 3 min, and then treated with 0.3% hydrogenperoxide in methyl alcohol for 20 min to block endogenousperoxidase activity. After three washes with PBS, the sectionswere incubated with 10% normal goat serum and thenincubated for 1 h at room temperature (RT) with primaryantibodies, which included the polyclonal anti-p-caveolin-1,ED1, and R73 antibodies. Immunoreactivity was visualizedwith the avidin–biotin peroxidase complex (Vector Elite;Vector Laboratories). The peroxidase reaction was developedusing a diaminobenzidine substrate kit (Vector Laboratories).

Paraffin sections were used to examine the doubleimmunofluorescence of p-caveolin-1 and either rat macro-phages or Schwann cells, while frozen sections were used forthe double-labeling with p-caveolin-1 and R73. The paraffinsections were deparaffinized and hydrated. The frozensections were air-dried, fixed in 4% paraformaldehyde inPBS (pH 7.4) for 20 min, and hydrated. Both the frozen andparaffin sections were washed with PBS three times andincubated with the listed reagents in the following order:10% normal goat serum for 1 h at RT, rabbit polyclonal anti-p-caveolin-1 antibody overnight at 4 °C, and fluoresceinisothiocyanate (FITC)-labeled goat anti-rabbit IgG (Sigma) for1 h at RT. The sections were then washed and blocked with10% normal goat serum for 1 h at RT, incubated with eitherED1, monoclonal anti-GFAP, or R73 overnight at 4 °C,washed, and incubated with tetramethyl rhodamine isothio-cyanate (TRITC)-labeled goat anti-mouse IgG (Sigma). Tovisualize the co-localization of p-caveolin-1 and IB4 in the

EAN lesions, the sections were reacted with biotinylated IB4(Sigma) followed by TRITC-labeled streptavidin (Zymed, SanFrancisco, CA, USA). The sections were then reacted with theanti-p-caveolin-1 antibody followed by FITC-labeled goatanti-rabbit IgG (Sigma).

To reduce or eliminate lipofuscin autofluorescence, thesections were washed three times for 1 h each in PBS at RTthen dipped briefly in distilled H2O, treated with 10 mM CuSO4

in ammonium acetate buffer (50 mM CH3COONH4 [pH 5.0]) for20 min, dipped briefly again in distilled H2O, and returned toPBS. The double-immunofluorescence-stained specimenswere examined under an FV500 laser confocal microscope(Olympus, Tokyo, Japan).

4.7. Detection of apoptosis

DNA fragmentation was detected using in situ terminaldeoxynucleotidyl transferase (TdT)-mediated dUTP nick end-labeling (TUNEL), which was performed with the ApopTag insitu Apoptosis Detection Kit (Intergen, Purchase, NY, USA),according to the manufacturer's instructions. Co-localizationof the TUNEL reaction and p-caveolin-1 immunoreactivity wasexamined using double-immunofluorescence labeling of thesame section. Briefly, after finishing the TUNEL reaction,which was allowed to react with TRITC-labeled anti-digox-igenin antibody, double immunofluorescence was carried outwith FITC-labeled goat anti-rabbit IgG (Sigma) secondaryantibodies to co-localize the TUNEL reaction and the indivi-dual protein antigens in the same cell.

R E F E R E N C E S

Ahn, M., Moon, C., Lee, Y., Koh, C.S., Kohyama, K., Tanuma, N.,Matsumoto, Y., Kim, H.M., Kim, S.R., Shin, T., 2004. Activationof extracellular signal-regulated kinases in the sciatic nerves ofrats with experimental autoimmune neuritis. Neurosci. Lett.372, 57–61.

Ahn, M., Moon, C., Kim, H., Lee, J., Koh, C.S., Matsumoto, Y., Shin,T., 2006. Immunohistochemical study of caveolin-1 in thesciatic nerves of Lewis rats with experimental autoimmuneneuritis. Brain Res. 1102, 86–91.

Conti, G., Scarpini, E., Rostami, A., Livraghi, S., Baron, P.L., Pleasure,D., Scarlato, G., 1998. Schwann cell undergoes apoptosis duringexperimental allergic neuritis (EAN). J. Neurol. Sci. 161 (1),29–35.

Hartung, H.P., Toyka, K.V., 1990. T-cell and macrophage activationin experimental autoimmune neuritis and Guillain–Barrésyndrome. Ann. Neurol. 27 Suppl., S57–S63 Review.

Kim, H., Ahn, M., Lee, J., Moon, C., Matsumoto, Y., Koh, C.S., Shin,T., 2006. Increased phosphorylation of caveolin-1 in the spinalcord of Lewis rats with experimental autoimmuneencephalomyelitis. Neurosci. Lett. 402, 76–80.

Li, S., Seitz, R., Lisanti, M.P., 1996. Phosphorylation of caveolin bysrc tyrosine kinases. The alpha-isoform of caveolin isselectively phosphorylated by v-Src in vivo. J. Biol. Chem. 271,3863–3868.

Matsumoto, Y., Kim, G., Tanuma, N., 2000. Characterization of Tcell receptor associated with the development of P2peptide-induced autoimmune neuritis. J. Neuroimmunol. 102,67–72.

Moon, C., Ahn, M., Kim, H., Lee, Y., Koh, C.S., Matsumoto, Y., Shin,T., 2005. Activation of p38 mitogen-activated protein kinase in

160 B R A I N R E S E A R C H 1 1 3 7 ( 2 0 0 7 ) 1 5 3 – 1 6 0

the early and peak phases of autoimmune neuritis in rat sciaticnerves. Brain Res. 1040, 208–213.

Moon, C., Kim, H., Ahn, M., Jin, J.K., Wang, H., Matsumoto, Y., Shin,T., 2006. Enhanced expression of netrin-1 protein in the sciaticnerves of Lewis rats with experimental autoimmune neuritis:possible role of the netrin-1/DCC binding pathway in anautoimmune PNS disorder. J. Neuroimmunol. 172, 66–72.

Okamoto, T., Schlegel, A., Scherer, P.E., Lisanti, M.P., 1998.Caveolins, a family of scaffolding proteins for organizing“preassembled signaling complexes” at the plasmamembrane.J. Biol. Chem. 273, 5419–5422.

Shin, T., Kim, H., Jin, J.K., Moon, C., Ahn, M., Tanuma, N.,Matsumoto, Y., 2005. Expression of caveolin-1,-2, and-3 in thespinal cords of Lewis rats with experimental autoimmuneencephalomyelitis. J. Neuroimmunol. 165, 11–20.

Veluthakal, R., Chvyrkova, I., Tannous, M., McDonald, P., Amin, R.,Hadden, T., Thurmond, D.C., Quon, M.J., Kowluru, A., 2005.Essential role for membrane lipid rafts in

interleukin-1beta-induced nitric oxide release frominsulin-secreting cells: potential regulation by caveolin-1+.Diabetes 54, 2576–2585.

Volonte, D., Galbiati, F., Pestell, R.G., Lisanti, M.P., 2001. Cellularstress induces the tyrosine phosphorylation of caveolin-1 (Tyr(14)) via activation of p38mitogen-activated protein kinase andc-Src kinase. Evidence for caveolae, the actin cytoskeleton, andfocal adhesions asmechanical sensors of osmotic stress. J. Biol.Chem. 276, 8094–8103.

Zhu, J., Bai, X.F., Mix, E., Link, H., 1997. Cytokine dichotomy inperipheral nervous system influences the outcome ofexperimental allergic neuritis: dynamics of mRNA expressionfor IL-1 beta, IL-6, IL-10, IL-12, TNF-alpha, TNF-beta, andcytolysin. Clin. Immunol. Immunopathol.84, 85–94.

Zhu, J., Mix, E., Link, H., 1998. Cytokine production and thepathogenesis of experimental autoimmune neuritis andGuillain–Barré syndrome. J. Neuroimmunol. 84, 40–52.