Short communication

Effects of polyclonal immunoglobulins and other immunomodulatory

agents on microglial phagocytosis of apoptotic inflammatory T-cells

Andrew Chan*,1, Christina Papadimitriou1,2, Wolfgang Graf, Klaus V. Toyka, Ralf Gold

Department of Neurology, Clinical Research Group for Multiple Sclerosis and Neuroimmunology, Julius-Maximilians-University,

D-97080 Wurzburg, Germany

Received 3 September 2002; received in revised form 25 November 2002; accepted 25 November 2002

Abstract

T-cell apoptosis in the CNS is an effective mechanism for the noninflammatory resolution of autoimmune T-cell infiltrates. Ingestion of

apoptotic leukocytes by microglia results in an efficient clearance of the inflammatory infiltrate, followed by a profound downregulation of

proinflammatory phagocyte immune functions. The effects of different immunomodulatory agents on Lewis rat microglial phagocytosis of

apoptotic autologous thymocytes or myelin-basic protein (MBP)-specific, encephalitogenic T-cells were investigated using a standardized,

light microscopical in vitro phagocytosis assay. Pretreatment of microglia with polyclonal 7S immunoglobulins (IVIg) decreased the

phagocytosis of apoptotic thymocytes by 38.2% ( p < 0.0001). Also, immunoglobulin F(abV)2 fragments decreased microglial phagocytosis,

suggesting an Fc receptor-independent mechanism. Similar results were obtained using MBP-specific T-cells. Pretreatment of microglia with

IFN-g increased the phagocytosis of apoptotic cells by 65.4%, which was to a large extent counteracted by IVIg. Glatiramer acetate (GLAT)

did not exert an effect on microglial phagocytosis, while methylprednisolone (MP) induced microglial apoptosis in vitro. These results

indicate that IVIg has a high potential to inhibit microglial phagocytosis of apoptotic inflammatory T-cells even under proinflammatory

conditions and extend our view of the complex immunomodulatory effects of IVIg.

D 2002 Elsevier Science B.V. All rights reserved.

Keywords: T-cell apoptosis; Multiple sclerosis; Experimental autoimmune encephalomyelitis; Glatiramer acetate; Glucocorticosteroids

1. Introduction

Apoptosis of 30–50% of all invading T-cells represents a

crucial mechanism in the termination of autoimmune T-cell-

mediated inflammation in the human and rodent CNS,

contributing to clinical recovery in experimental autoim-

mune encephalomyelitis (EAE) and acute disseminated

leukoencephalomyelitis in man (ADEM) (Bauer et al.,

2001; Pender and Rist, 2001).

A key event in the resolution of an inflammatory infiltrate

is the nonphlogistic and thus safe phagocytic clearance of

apoptotic leukocytes by tissue-specific phagocytes (Fadok et

al., 2001). Phagocytosis of apoptotic lymphocytes by macro-

phages/microglia, oligodendrocytes and astrocytes has been

described in situ in Lewis rat EAE (Nguyen and Pender,

1998). Lewis rat microglia efficiently phagocytoses apop-

totic, encephalitogenic MBP-specific T-cells in vitro, differ-

entially regulated by Th1-/Th2-type cytokines (Chan et al.,

2001). The phagocytosis of apoptotic T-cells by Lewis rat

microglia is more efficient than by astrocytes and leads to a

profound downregulation of microglial immune functions,

making this process an attractive target for therapeutic

interventions (Magnus et al., 2002). The immunomodulatory

agent interferon-beta augments phagocytosis specifically of

apoptotic inflammatory cells, and also adult human microglia

obtained from normal brain tissue phagocytoses apoptotic

inflammatory cells in vitro (Chan et al., 2002). Here we set

out to investigate the effects of other therapeutically used

immunomodulatory agents on Lewis rat microglial phagocy-

tosis of apoptotic inflammatory cells.

0165-5728/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.

doi:10.1016/S0165-5728(02)00433-2

* Corresponding author. Neurologische Universitatsklinik, Josef-

Schneider-Straße 11, D-97080 Wurzburg, Germany. Tel.: +49-931-201-

24621; fax: +49-931-201-23488.

E-mail address: AndrewHKChan@hotmail.com (A. Chan).1 Equally contributing authors.2 Present address: Department of Neurology, University Clinic, Ahepa

Hospital, Thessaloniki, Greece.

www.elsevier.com/locate/jneuroim

Journal of Neuroimmunology 135 (2003) 161–165

2. Materials and methods

2.1. Cell culture

All cell culture media and supplements were obtained

from Gibco/BRL (Eggenstein, Germany) unless otherwise

noted. Microglial cells from neonatal Lewis rats (P0–P2,

Charles River, Sulzfeld, Germany) with a purity of consis-

tently >97% were isolated as described before (Chan et al.,

2001). Apoptosis of autologous thymocytes and of the MBP-

specific, encephalitogenic CD4+ T-cell line MBP13 was

induced by methylprednisolone (MP, Aventis Pharma, Bad

Soden, Germany) as described in detail before (Chan et al.,

2001). MP-treated thymocytes had a proportion of

37.8F 2.6% (meanF S.E.M.) annexin V single-positive

(early apoptotic) and 5.6F 0.6% annexin V/PI-positive (late

apoptotic/necrotic) cells, untreated control thymocytes were

12F 2.6% annexin V-positive and 3.2F 0.5% annexin V/PI-

positive (Chan et al., 2001). The proportion of viable thy-

mocytes or MBP13 cells excluding trypan blue was consis-

tently >97%.

2.2. In vitro phagocytosis assay

The standardized, microscopically quantified microglial

phagocytosis assay has been described and illustrated in

detail before (Chan et al., 2001). Four hundred microliters

of a 0.75� 106/ml suspension of microglial cells per well

was seeded in 48-well plates (Costar) and cultured over-

night at 37 jC/5% CO2 (BME, 10% FCS (Sigma Aldrich

Chemie, Steinheim, Germany), 50 U/ml penicillin, 50 Ag/ml streptomycin). Triplicate wells of microglial cells were

then incubated with the respective concentrations of IVIg

(SandoglobulinR, Novartis Pharma, Nurnberg, Germany),

F(abV)2 fragments (GammaveninR, Aventis Behring, Mar-

burg, Germany), glatiramer acetate (GLAT, TEVA Pharma/

Aventis Pharma), MP or BME for 20–24 h. Immunoglo-

bulin preparations were dialyzed (H2O or BME) to remove

stabilising agents that are potentially toxic in cell culture.

Human albumin (Octapharma, Langenfeld, Germany; DRK

Blutspendedienst, Baden-Baden, Germany) or ovalbumin

(Sigma) were used as control proteins, respectively. In case

of combined IFN-g (30 U/ml, R&D, Minneapolis, MN,

USA))/IVIg pretreatment, microglia was either preincu-

bated simultaneously or in another set of experiments

sequentially treated with IFN-g (8 h) followed by IVIg

(12 h). For RGDS/RGES peptide inhibition experiments,

target cells were preincubated with the peptides (1–2 mM,

BME, 15 min, 4 jC) and subsequently added to the

microglia without further washing. Thymocytes (500 Al,20� 106/ml in BME) or MBP13 T-cells (500 Al, 10� 106/

ml) were co-cultured with the microglia (2 h, 37 jC, 5%CO2) followed by vigorous washing with cold PBS (4 jC)(Chan et al., 2001). After trypsinization, a separate cyto-

centrifuge preparation was obtained for each well and

stained with May-Giemsa (Merck, Darmstadt, Germany)

(Chan et al., 2001). An average of 500 microglial cells per

slide were counted in a blinded fashion by light micro-

scopy. In some experiments, data is additionally given as

phagocytic index (percent of phagocytosing microglia

multiplied with the average number of ingested target cells

per microglia) (Chan et al., 2001). All values are expressed

as meanF S.E.M. Statistical significance was evaluated

using Student’s t-test (GraphPad Software, San Diego,

USA).

3. Results

3.1. IVIg and F(abV)2 fragments decrease microglial

phagocytosis of apoptotic inflammatory cells

As reported before, microglia has a high capacity to

phagocytose apoptotic thymocytes or CNS autoantigen-

specific, encephalitogenic T-cells in contrast to non-apop-

totic target cells (Chan et al., 2001) (Figs. 1A and 2). As

illustrated in Figs. 1A,B and 2, IVIg pretreatment (20 mg/

ml) decreased the phagocytosis of apoptotic thymocytes

by 38.2F 5.2% (meanF S.E.M.) in comparison to

untreated microglia ( p < 0.0001). Pretreatment with human

albumin (HA) did not exert an effect (Fig. 2). The

decrease of phagocytosis was dose-dependent and reached

a plateau with 20 mg/ml IVIg (percent inhibition, 10 mg/

ml: 26.3F 6.25, p < 0.01; 30 mg/ml: 39.5F 7.2%, p <

0.001). IVIg not only decreased the phagocytosis of

corticosteroid-treated, apoptotic target cells, but also the

much lower baseline phagocytosis of nontreated thymo-

cytes, albeit to a lesser extent (Fig. 2, 27.5F 3.4%, p <

0.0001). The stronger inhibitory IVIg effect on phagocy-

tosis of apoptotic thymocytes was even more pronounced

in the phagocytic index, which reflects the phagocytic

capacity of individual microglial cells (IVIg 20 mg/ml,

percent inhibition for apoptotic thymocytes: 32.6F 7.5%,

p < 0.001; for non-corticosteroid-treated cells: 18.2F 6%,

p < 0.05). IVIg (20 mg/ml) also decreased the phagocyto-

sis of corticosteroid-treated, apoptotic, encephalitogenic

MBP-specific MBP13 T-cells (percent inhibition 20.7F14.7%) and of non-corticosteroid-treated T-cells (13.6F5.6%).

To investigate which portion of IVIg mediated the inhib-

ition of phagocytosis, microglia was pretreated using F(abV)2immunoglobulin fragments. Again, F(abV)2-mediated inhib-

ition was more pronounced for the uptake of apoptotic

thymocytes (49.1F 3%, p < 0.0001) than for non-cortico-

steroid-treated cells (17.8F 1.6%, p < 0.001). F(abV)2 frag-

ments also decreased microglial phagocytosis of apoptotic

MBP13 T-cells by 48.3F 7% ( p < 0.0001), whereas no clear

inhibition could be observed for non-glucocorticosteroid-

treated T-cells. These results indicated that the suppression of

phagocytosis of apoptotic and non-apoptotic cells was at

least to a great part independent of Fc-receptor-mediated

mechanisms.

A. Chan et al. / Journal of Neuroimmunology 135 (2003) 161–165162

3.2. IVIg partially counteracts the phagocytosis-promoting

effect of interferon-gamma (IFN-c)

As reported before, IFN-g increases microglial phagocy-

tosis of apoptotic thymocytes, whereas the uptake of non-

apoptotic target cells is not altered (Chan et al., 2001). IFN-g

(30 U/ml) augmented the microglial phagocytosis rate for

apoptotic thymocytes by 65.4F 18.5% above the untreated

controls (Fig. 3, p < 0.01), similar to values reported pre-

viously (Chan et al., 2001). Simultaneous preincubation of

microglia with IFN-g and IVIg partially reversed this effect

in an IVIg dose-dependent manner (percent inhibition of the

IFN-g-augmented phagocytosis rate, IVIg 10 mg/ml:

33.2F 8.4%; IVIg 20 mg/ml: 49.4F 2.8%, p < 0.05). No

change was observed with the combination of IFN-g and HA

(Fig. 3). The inhibitory effect was even more pronounced

using the phagocytic index. Whereas IFN-g alone increased

the phagocytic index by 88.4F 29% ( p < 0.05), this effect

was reduced by IVIg (20 mg/ml) by 69.4F 2% ( p < 0.05).

3.3. Glatiramer acetate (GLAT) does not have an effect on

microglial phagocytosis while methylprednisolone (MP)

induces microglial apoptosis in vitro

Pretreatment with GLAT (10–50 Ag/ml) did not show any

specific effects on microglial phagocytosis of apoptotic or

Fig. 1. Photomicrographs of untreated (A) or IVIg-pretreated (20 mg/ml)

Lewis rat microglia (B) after 2 h interaction with autologous apoptotic

thymocytes, which show typical apoptotic morphology with condensed

chromatin (arrows). May-Giemsa stain. Bar = 10 Am.

Fig. 2. Phagocytosis of autologous thymocytes by untreated and IVIg-

pretreated microglia. Microglia had a higher capacity for the uptake of

corticosteroid-treated, apoptotic (+) thymocytes than for non-corticosteroid-

treated (� ) thymocytes. IVIg decreased phagocytosis of apoptotic target

cells and to a lesser extent of non-corticosteroid-treated cells in comparison

to the untreated microglia, while human albumin (HA) did not have an effect

on phagocytosis. Phagocytosis rate: given as percentage of the mean of the

untreated controls (+) + S.E.M. 42.1F 4.7% (meanF S.E.M.) of the

untreated microglia were capable of phagocytosing apoptotic thymocytes.

Five independent experiments, each performed in triplicates.

Fig. 3. Microglial phagocytosis of apoptotic thymocytes after pretreatment

with IFN-g, IFN-g/IVIg or IFN-g/human albumin. IFN-g (30 IU/ml)

increased the phagocytosis rate for apoptotic target cells (+), which was

partially counteracted by the combinationwith IVIg (IFN-g/IVIg, 20mg/ml).

Combined pretreatment with IFN-g and human albumin (IFN-g/HA, 20 mg/

ml) did not alter the phagocytosis-promoting effect of IFN-g. Phagocytosis

rate: given as percentage of the mean of the untreated controls (+) + S.E.M.

34.9F 5.9% (meanF S.E.M.) of the untreated microglia were capable of

phagocytosing apoptotic thymocytes. Two independent experiments, each

performed in triplicates.

A. Chan et al. / Journal of Neuroimmunology 135 (2003) 161–165 163

non-corticosteroid-treated thymocytes in comparison to

ovalbumine (data not shown). Pretreatment of microglia

with MP (50–1000 nM) resulted in decreased phagocytosis

of thymocytes. Immunohistochemistry demonstrated a pro-

found induction of microglial apoptosis even by low con-

centrations of MP with all cells showing annexin V/PI

positivity after 3 h of MP treatment (200 nM, data not

shown).

4. Discussion

Although high dose IVIg are widely used in neurological

diseases of presumed autoimmune etiology, their pleiotropic

mechanisms of action are only incompletely understood

(Kazatchkine and Kaveri, 2001; Wiles et al., 2002). More

recently, a possible influence of IVIg also on local CNS

immune reactions and remyelination has been suggested

(Stangel et al., 2000a; Stangel and Compston, 2001, War-

rington et al., 2000).

Here, we have demonstrated that IVIg in concentrations

corresponding to serum levels during IVIg treatment inhibit

microglial phagocytosis of apoptotic thymocytes as well as

MBP-specific, encephalitogenic T-cells. This inhibition

appears to be mediated to a great part via the immunoglo-

bulin Fab-portion. However, the exact mechanisms and

potential microglial recognition molecules involved remain

elusive. A broad range of specialized receptors have been

implied in the phagocytosis of apoptotic cells and the

subsequent modulation of phagocyte immune functions

(Fadok et al., 2001). Lectin-, integrin- and phosphatidylser-

ine-dependent mechanisms have been described in the phag-

ocytosis of apoptotic targets by rodent microglia in vitro (De

Simone et al., 2002; Witting et al., 2000). IVIg were recently

demonstrated to inhibit leukocyte adhesion by antibodies

against the RGD adhesion motif (Vassilev et al., 1999). In

our system, RGDS peptides did not specifically inhibit

microglial phagocytosis of apoptotic thymocytes, arguing

against an integrin-mediated recognition/uptake mechanism

(data not shown). Also, complement-components have been

demonstrated in the phagocytosis of apoptotic cells (Fadok et

al., 2001). Since in our study sera were heat-inactivated and

all phagocytosis experiments were performed under serum-

free conditions, microglial phagocytosis was not dependent

on complement factors.

Whatever mechanism involved, the inhibition of micro-

glial phagocytosis of apoptotic cells by IVIg appears to be

very potent, since even the strong phagocytosis-promoting

effect of IFN-gwas largely reversed by IVIg. An interference

with IFN-g stimulation by anti-interferon antibodies in IVIg

(Ross et al., 1995) was excluded by sequential IFN-g and

IVIg pretreatment (data not shown). The ‘‘baseline’’ phago-

cytosis of non-steroid-treated cells and its inhibition by IVIg

can partly be explained by the minor proportion of apoptotic

cells inevitably present in the cell preparations. IVIg-medi-

ated inhibition of unspecific phagocytosis mechanisms could

additionally play a role (Stangel et al., 2000b). Recently,

IVIg have been demonstrated to increase Fc-receptor-medi-

ated PNS-myelin phagocytosis by macrophages, while

microglial CNS-myelin phagocytosis was not affected

(Kuhlmann et al., 2002). However, the in vivo significance

of these findings is unknown (Stangel et al., 2000b). Our data

indicate that in autoimmune CNS-inflammation IVIg could

interfere with the removal of the inflammatory infiltrate, at

least during stages with a high prevalence of apoptotic

inflammatory cells. Whether IVIg also inhibit the phagocy-

tosis of apoptotic neurons and oligodendrocytes is currently

unknown.

In addition to several other presumed mechanisms of

action, GLAT has also been demonstrated to alter macro-

phage effector functions in vitro (Siglienti et al., 2000). Here,

we could not demonstrate an effect of GLAT on microglial

phagocytosis of apoptotic or non-corticosteroid-treated thy-

mocytes. MP increases T-cell apoptosis in situ in EAE but

does not appear to affect glial cells (Schmidt et al., 2000).

Moreover, glucocorticosteroids promote the phagocytosis of

apoptotic granulocytes by human monocyte-derived macro-

phages in vitro (Liu et al., 1999). Here, MP led to rapid

microglial apoptosis in vitro. Previous studies have shown a

reduction in the number of rodent corpus callosum microglia

after glucocorticosteroid injections (Wu et al., 2001). Thus,

the effects of MP-pulse therapy during EAE on potential

microglial apoptosis and possible influences on the phag-

ocytosis of apoptotic cells merit further investigations.

In conclusion, our data add to the growing notion that, in

addition to peripheral mechanisms at least under conditions

of an impaired blood–brain barrier, IVIg could also exert

local immunomodulatory effects in the inflamed CNS.

However, the complex interplay between these mechanisms

remains to be elucidated in vivo.

Acknowledgements

The authors thank Annette Horn for excellent technical

support. We are indebted to Dr. Jack Antel for many

stimulating discussions. We thank Dr. Martin Stangel for his

expert advice on IVIg action on glial cells and Prof. Ioannis

Milonas for his continuous support of C.P. Supported by

grants from the Deutsche Forschungsgemeinschaft (DFG Go

459/8-3), funds from the state of Bavaria and a fellowship

grant of the European Neurological Society to C.P.

References

Bauer, J., Rauschka, H., Lassmann, H., 2001. Inflammation in the nervous

system: the human perspective. Glia 36, 235–243.

Chan, A., Magnus, T., Gold, R., 2001. Phagocytosis of apoptotic inflam-

matory cells by microglia and modulation by different cytokines: mech-

anism for removal of apoptotic cells in the inflamed nervous system.

Glia 33, 87–95.

Chan, A., Seguin, R., Magnus, T., Antel, J.P., Toyka, K.V., Gold, R., 2002.

A. Chan et al. / Journal of Neuroimmunology 135 (2003) 161–165164

Phagocytic clearance of apoptotic inflammatory cells by human and

rodent microglia and effects of different immunomodulatory agents.

Glia S1, S82.

De Simone, R., Ajmone-Cat, M.A., Nicolini, A., Minghetti, L., 2002. Ex-

pression of phosphatidylserine receptor and down-regulation of pro-

inflammatory molecule production by its natural ligand in rat microglial

cultures. J. Neuropathol. Exp. Neurol. 61, 237–244.

Fadok, V.A., Bratton, D.L., Henson, P.M., 2001. Phagocyte receptors for

apoptotic cells: recognition, uptake, and consequences. J. Clin. Invest.

108, 957–962.

Kazatchkine, M.D., Kaveri, S.V., 2001. Immunomodulation of autoimmune

and inflammatory diseases with intravenous immune globulin. N. Engl.

J. Med. 345, 747–755.

Kuhlmann, T., Wendling, U., Nolte, C., Zipp, F., Maruschak, B., Stadel-

mann, C., Siebert, H., Bruck, W., 2002. Differential regulation of mye-

lin phagocytosis by macrophages/microglia, involvement of target

myelin, Fc receptors and activation by intravenous immunoglobulins.

J. Neurosci. Res. 67, 185–190.

Liu, Y., Cousin, J.M., Hughes, J., van Damme, J., Seckl, J.R., Haslett, C.,

Dransfield, I., Savill, J., Rossi, A.G., 1999. Glucocorticoids promote

nonphlogistic phagocytosis of apoptotic leukocytes. J. Immunol. 162,

3639–3646.

Magnus, T., Chan, A., Savill, J., Toyka, K.V., Gold, R., 2002. Phagocytosis

of apoptotic, inflammatory lymphocytes in the central nervous system

(CNS) by microglia and its functional implications. J. Neuroimmunol.

130, 1–9.

Nguyen, K.B., Pender, M.P., 1998. Phagocytosis of apoptotic lymphocytes

by oligodendrocytes in experimental autoimmune encephalomyelitis.

Acta Neuropathol. 95, 40–46.

Pender, M.P., Rist, M.J., 2001. Apoptosis of inflammatory cells in immune

control of the nervous system: role of glia. Glia 36, 137–144.

Ross, C., Svenson, M., Hansen, M.B., Vejlsgaard, G.L., Bendtzen, K.,

1995. High avidity IFN-neutralizing antibodies in pharmaceutically pre-

pared human IgG. J. Clin. Invest. 95, 1974–1978.

Schmidt, J., Gold, R., Zettl, U.K., Hartung, H.P., Toyka, K.V., 2000. T-cell

apoptosis in situ in experimental autoimmune encephalomyelitis follow-

ing methylprednisolone pulse therapy. Brain 123, 1431–1441.

Siglienti, I., Toyka, K.V., Scarpini, E., Jung, S., 2000. Copolymer-1 effects

on cytokine production by Lewis rat peritoneal macrophages. J. Neurol.

247, III/203.

Stangel, M., Compston, A., 2001. Polyclonal immunoglobulins (IVIg)

modulate nitric oxide production and microglial functions in vitro via

Fc receptors. J. Neuroimmunol. 112, 63–71.

Stangel, M., Compston, A., Scolding, N.J., 2000a. Oligodendroglia are

protected from antibody-mediated complement injury by normal immu-

noglobulins (‘‘IVIg’’). J. Neuroimmunol. 103, 195–201.

Stangel, M., Joly, E., Scolding, N.J., Compston, D.A., 2000b. Normal

polyclonal immunoglobulins (‘IVIg’) inhibit microglial phagocytosis

in vitro. J. Neuroimmunol. 106, 137–144.

Vassilev, T., Kazatchkine, M.D., Van Huyen, J.P.D., Mekrache, M., Bon-

nin, E., Mani, J.C., Lecroubier, C., Korinth, D., Baruch, D., Schriever,

F., Kaveri, S., 1999. Inhibition of cell adhesion by antibodies to Arg-

Gly-Asp (RGD) in normal immunoglobulin for therapeutic use (intra-

venous immunoglobulin IVIg). Blood 93, 3624–3631.

Warrington, A.E., Asakura, K., Bieber, A., Ciric, B., van Keulen, V., Ka-

veri, S.V., Kyle, R.A., Pease, L.R., Rodriguez, M., 2000. Human mono-

clonal antibodies reactive to oligodendrocytes promote remyelination in

a model of multiple sclerosis. Proc. Natl. Acad. Sci. 97, 6820–6825.

Wiles, C.M., Brown, P., Chapel, H., Guerrini, R., Hughes, R.A.C., Martin,

T.D., McCrone, P., Newsom-Davis, J., Palace, J., Rees, J.H., Rose, M.R.,

Scolding, N., Webster, A.D.B., 2002. Intravenous immunoglobulin in

neurological disease: a specialist review. J. Neurol. Neurosurg. Psychia-

try 72, 440–448.

Witting, A., Muller, P., Herrmann, A., Kettenmann, H., Nolte, C., 2000.

Phagocytic clearance of apoptotic neurons by microglia/brain macro-

phages in vitro: involvement of lectin-, integrin-, and phosphatidylser-

ine-mediated recognition. J. Neurochem. 75, 1060–1070.

Wu, C.H., Chien, H.F., Chang, C.Y., Chen, S.H., Huang, Y.S., 2001. Re-

sponse of amoeboid and differentiating ramified microglia to glucocor-

ticoids in postnatal rats: a lectin histochemical and ultrastructural study.

Neurosci. Res. 40, 235–244.

A. Chan et al. / Journal of Neuroimmunology 135 (2003) 161–165 165