Microglial and astroglial reactions to inflammatory lesions of experimental autoimmune encephalomyelitis in the rat central nervous system

Journal of Neuroimmunology, 37 (1992) 23-33

23

© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-5728/92/$05.00 JNI 02129

Microglial and astroglial reactions to inflammatory lesions of experimental autoimmune encephalomyelitis in the rat central nervous system Yoh Matsumoto, Katsutoshi Ohmori and Michio Fujiwara Department of Immunology, Niigata Unicersity School of Medicine, Niigata, Japan

(Received 1 July 1991) (Revised, received3 October 1991) (Accepted 3 October 1991)

Key words: Experimental autoimmune encephalomyelitis;Microglia;Astrocyte; Lesion repair

Summary Gliosis is a repair process of lesions appearing in the central nervous system (CNS). Although gliosis by astrocytes (astrocytic gliosis) has been well documented, that by microglia (microglial gliosis) remains poorly understood. In the present study we induced experimental autoimmune encephalomyelitis (EAE) in Lewis rats and examined microglial and astroglial reactions to EAE lesions at various stages of the disease by immunohistochemistry. For the demonstration of microglia and astrocytes, antibodies against complement receptor type 3 (OX42) and glial fibrillary acidic protein (GFAP) were used, respectively. It was revealed that the whole course of microgliai and astroglial reactions to EAE lesions is divisible into three stages, i.e., initial, peak and recovery stages. Microglial and astroglial reactions to EAE lesions at each stage correspond well with the clinical and histological stages of EAE. At the initial stage, rats showed mild clinical signs and a few inflammatory foci were found in the CNS. Microglia were increased in number in close association with inflammatory cell aggregates, whereas astrocytes showed no significant reaction in spite of the presence of infammatory cells. At the peak stage, rats showed full-blown E A E and the number of inflammatory cells reached maximum. The most characteristic finding at this stage was 'encasement' of inflammatory lesions by astrocytic fibers. Microglia were increased in number, but association of microglia with lesions was prevented by astrocytes. Interestingly, however, such characteristic distribution of microglia and astrocytes was not observed at the recovery stage. Residual inflammatory cell aggregates were intermingled with dense microglial and astrocytic gliosis, forming 'micro-astroglial scars'. Double immunofluorescence staining with anti-GFAP and antibromodeoxyuridine (BrdU), or with OX42 and anti-BrdU revealed that BrdU-incorporated microglia, but not astrocytes, were present mainly at the initial and peak stages, suggesting that microglia would proliferate by cell division to create gliosis, whereas astrocytic gliosis would be a result of migration of astrocytes a n d / o r up-regulation of expression of G F A P molecule. Taken together with previous in vitro

Correspondence to: Dr. Yoh Matsumoto, Department of Immunology, Niigata University School of Medicine, Asahimachi-l, Niigata 951, Japan.

Supported by a grant from the Neuroimmunological Research Committee, Ministryof Health and Welfare, Japan.

24 findings that microglia, but not astrocytes, stimulate encephalitogenic T ceil proliferation, these in vivo findings suggest that microglia augment, whereas astrocytcs suppress, inflammatory processes in the CNS.

Introduction

It has recently been established that brain cells such as microglia and astrocytcs are involved in immunological reactions in the central nervous system (CNS) (sec Fontana ct al., 1987 for review). This notion comes from several lines of evidence. First, during immunological disorders in the CNS such as experimental autoimmunc encephalomyelitis (EAE) (Matsumoto and Fujiwara, 1986; Matsumoto et al., 1986; Vass et al., 1986), multiple sclerosis (Lampson and Hickey, 1986; Woodrofe et al., 1986; Hayes ct al., 1987; Boyle and McGeer, 1990) and other neurological diseases (Hickey and Kimura, 1987), brain cells in situ, mainly microglia, express la antigens in close association with inflammatory lesions. In addition, cultured brain cells (e.g. astrocytes, microglia and cerebral endothelial cells) express la antigens after treatment with y-interferon (IFNy) (Fontana ct al., 1984; Fierz et al., 1985; Male et al., 1987; Massa ct al., 1987; Sasaki et al., 1990). Second, in vitro studies have shown that la-expressing brain cells have antigen-presenting ability (Fontana ct al., 1984; Ficrz et al., 1985; McCarron ct al., 1985; Frci ct al., 1987; Wilcox et al., 1987). Taken together, it is generally assumed that resident cells in the CNS up-regulate the proliferation of infiltrating inflammatory cells, especially T cells. However, our previous studies did not always support this speculation. For example, when passive E A E was induced in irradiated recipients, no proliferating cells were observed in EAE lesions by bromodcoxyuridinc (BrdU)-anti-BrdU immunohistochemistry (Matsumoto et al., 1988). Also in ordinary active and passive EAE, infiltrating T cells do not prolifcrate vigorously in the CNS parcnchyma (Ohmori ct al., 1992). Furthermore, cultured mixed glial cells under in vivo-mimicking conditions showed very weak antigen-presenting ability, although microglia isolated from this population presented antigen to T cells (Matsumoto, Y., Ohmori, K.

and Fujiwara, M., manuscript submitted for publication). To better understand the real immunological functions of brain cells and the interaction between brain cells and infiltrating inflammatory cells, morphological studies are indispensable. In the present study we induced active and passive EAE in Lewis rats and examined microglial and astroglial reactions to EAE lesions at various stages of EAE. Consequently, wc found that microglia reacted to inflammatory foci at the very early stage, whereas astrocytes did not. At the peak stage of EAE, astrocytes encased the inflammatory foci so that the association of microglia with the lesions seen at the initial stage was prevented. These findings are well consistent with our in vitro studies suggesting that microglia, but not astrocytes, augment T cell proliferation in an antigen-specific manner, and suggest that brain cells comprising different typcs of cell generally suppress inflammatory reactions in the CNS and protect the brain from damages.

Materials and methods

Animals Lewis (LEW) rats were obtained from Charles River Japan (Kanagawa). Female rats at thc age of 8-12 weeks were used. Antibodies and reagents Monoclonal antibodies (MAb) used in the present study were OX42 which recognizes type 3 complement receptors and stains macrophages/ microglia positively (Robinson et al., 1986), OX6 against la antigens (RTI.B) and anti-BrdU MAb (Becton-Dickinson, Mountain View, CA, USA). All MAbs except anti-BrdU MAb were obtained from Scrotec (Blackthorn, Bicester. Bucks., UK). Anti-glial fibrillary acidic protein (GFAP) antibody was obtained from Dako (Copenhagen, Denmark). BrdU was purchased from Sigma (St. l,ouis, MO, USA). Guinea pig myelin basic pro-

25 tein (GPBP) was prepared by the method of Deibler et al. (1972).

Induction of actiee and passive EAE Active and passive EAE was induced as described previously (Matsumoto et al., 1990a, b). In brief, active EAE was induced in LEW rats by immunization with 100/zg GPBP in Mycobacterium tuberculosis H37RA-enriched complete Freund's adjuvant (MT-enriched CFA). Passive EAE was induced by injection of a mixture of spleen cells and lymph node cells from GPBP-immunized rats. Before transfer, lymphoid cells were cultured for 3 days in the presence of 5 /xg/ml GPBP and 2 × 107 viable cells per rat were given intravenously. BrdU labeling and tissue sampling For the demonstration of DNA-synthesizing cells in the CNS, the BrdU-anti-BrdU immunohistochemistry was employed (Matsumoto et al., 1988; Ohmori et al., 1992). 45 min before killing, each rat received 50 mg/kg BrdU intravenously. Several segments of the lumbar spinal cord were removed and snap frozen. Tissue sampling was done between days 5 and 49 post-immunization and between days 2 and 11 after cell transfer. lmmunoperoxidase staining procedures Single immunoperoxidase staining for OX42, GFAP, OX6 and BrdU was performed as described previously (Matsumoto and Fujiwara, 1987, 1988). Frozen sections were air-dried and fixed in ether for 10 min. After incubation with normal sheep serum, sections were allowed to react with the first Ab, biotinylated anti-mouse or anti-rabbit Ig (Amersham, Amersham, UK) and horseradish peroxidase (HRP)-Iabeled streptavidin (Amersham). For the demonstration of BrdU-positive cells, DNA denaturation was done by incubating sections with 1 N HC1 (Matsumoto et al., 1988) or with 4% paraformaldehyde for 45 min at 70°C (Magaud et al., 1989). Double-immunofluorescence staining procedures Double-immunofluorescence staining with OX42 (IgG2a) and anti-GFAP (rabbit serum), OX42 and anti-BrdU (IgG1), or anti-GFAP and anti-BrdU was performed as described previously

(Matsumoto and Fujiwara, 1988; Ohmori et al., 1992). Fixed sections were incubated with the first antibody in the first step, followed by biotinylated secondary antibody and fluorescein isothiocyanate (FITC)-Iabeled streptavidin (Amersham). Then, the slides were incubated with 4% paraformaidehyde for 45 min at 70°C for subsequent BrdU staining. After washing, the slides were incubated with the first antibody in the second step followed by rhodamine-labeled secondary antibody.

Results

Active EAE Microglial reactions. Microglia under normal and pathologic conditions were identified on the basis of their morphology and reactivity with MAb OX42 (CDllb). In the previous studies using mice, we showed that anti-CD1 lb antibody recognized fetal brain macrophages (Matsumoto and Ikuta, 1985) and microglia in the adult brain with cold injury (Matsumoto et al., 1985). As shown in Fig. 1A, microglia in the normal rat CNS were stained positively by OX42. Under pathologic conditions such as cold injury (unpublished observation) and autoimmune encephalomyelitis (Matsumoto et al., 1986), infiltrating blood-borne macrophages were also stained positively with OX42. In such cases, however, the morphology of positive cells is helpful for distinguishing microglia from macrophages. As demonstrated by bone marrow chimera studies (Matsumoto and Fujiwara, 1987), OX42-positive cells with dendritic morphology are microglia, whereas round or oval OX42-positive cells are bone marrow-derived macrophages. Based on these findings, we followed up microglial reactions (the distribution of OX42-positive cells with dendritic morphology) to EAE lesions. In the normal spinal cord, OX42-positive microglia are evenly distributed in both the gray and white matter (Fig. 1A). On day 10 post-immunization, EAE lesions became apparent mainly in the white matter. As shown in Fig. 1B, microglia were increased in number in close association with inflammatory cell aggregates. Microglia at this stage became rod-shaped compared with

2~

those under normal conditions. In the CNS with full-blown EAE (between days 12 and 15), microglia were increased in number and showed the so-called 'activated' form (Fig. 1C). It is noteworthy, however, that in some lesions, microglia did not attach to inflammatory cell aggregates (Fig. 1C). This charactcristic finding will bc described later in more detail. On day 17 and thereafter, E A E lesions decreased in number. Microglia again showed the 'dendritic' morphology and intermingled with still existing inflammatory cells (Fig. I D) along with astrocytes (scc below). This

•

''

-

"

.;-,

•

,.,X,,.,.

',"

. : ~ , .~" _ . . - ~ , ~ . ; , •

.

; ~ . ~

t-

._,

.~ .. ~..

mixture of different cell types mentioned above forms a glial scar. Astroglial reactions. Astroglial reactions to infiltrating inflammatory, cells occurred later than microglia[ reactions. In the white matter of the normal CNS, cell processes of GFAP-positivc astrocytes ran parallel, as shown in Fig. 2 A. At the early stage of EAE (on day 10), the distribution of GFAP-positivc astrocytes was essentially the same as that seen in the normal CNS, although some fibers ran wavily duc to the prescncc of inflammatory cells (Fig. 2B). However, on days

~ ,

"~ .

,

,

,.'5

Fig. 1. Immunopcroxidase staining of OX42-positive microglia in the white manor (A-I)). Under normal conditions (A), microglia with dendritic morphology distribute evenly in the white matter as well as the gray matter (not shown) of the spinal cord. At the initial stage of EAE. microglia in the white matter (asterisks indicate the subarachnoid space) mainly increase in number in close association with infiltrating inflammatory cells (B). The number of microglia at the peak stage reaches maximum. Note that most mieroglia at this stage become rod-shaped and exist slightly away from ~m inflammatory lesion (C). At the recovery stage, mieroglia form glial scar with residual inflammatory cells (D). A, normal, x 180; B, day II) post-immunization (p.i.), grade I, × 180: (". day 12 p.i., grade 4, ×240; D, day 41 p.i., grade I), × 180.

27

Interaction of microglia and astrocytes. Based on the data obtained by single-immunoperoxidase staining, we decided to examine the CNS with EAE on days 12 and 21 in more detail. This was because on day 12 astrocytes surrounded inflammatory cell aggregates, whereas microglia seemed to be distributed slightly distant from the lesions. From day 21 onwards, on the other hand, both astrocytes and microglia intermingled with each other to form glial scars, along with still-existing inflammatory cells. The interaction of both cell types was examined by double-immunofluorescencc staining with OX42 and anti-GFAP. As

12 and 15 post-immunization when rats showed full-blown EAE, characteristic findings were observed in the spinal cord. More than half of the EAE lesions were surrounded by dense processes of astrocytes (Fig. 2C). In addition, astrocytic reactions occurred mainly at the margin of the lesions. Astrocytes between lesions, on the other hand, showed minimal changes such as slight increase in staining intcnsity. This is in sharp contrast to microglial reactions, in which microglia between lesions increase drastically in number. From day 21 onwards, astrocytes formed glial scars with microglia (Fig. 1D).

,.

.,,.

:/'.

,

",,~

:t..

.

<...:

~,,,:"

t.

'

q

~

~:

"

).

")

~

~.

,

"~'il,.

~',~

.

.

.

.

.

.

.

~ -.'. " '~"

'

;.. ~ "

'

"

*

,v), "

'

"

f~-~~,:.~T~_~.,:S::-l?:I~ p.i-';:~ k ' . ' L . " " ~.l,'Y ~ , . ~ / . ) . I ~ -.,,.,

.,

j,.'.,~l

~"'~ < ".. ,l

,"

"iz-

" .

.'~.

~':

•

V',~:'..!j~ir

~-.-..i~l ti ...-l,~(

....

.

-

,; Y " '~ " *~"'"

~.~

--.

.~ --i ,~¢~: .

,i~', "

"

~'~. •

,

-"

."

~-:

,

^

•

".L. ,0,~,",~-"h

:,

~)~'-~

-" ~ ~ . ,

~,,~.

.

~.

.

.

B;

.

~;'!<.-"!I"

-... ,~-:,i~. - . . ,. " '-~ "" ~

. '..~ "

;:"T :.li~<=.;:~,/i . "

.~ ,. i i _ i l . , . ' . : ~ , . ' ;.0~..~-'I=£~ ~

-,

, 1 ~''" .

.

~

~.,~ ,

- ,, ..%.;+-. I . T , i

.~.', " ' ~ 4-' I~..~

k :--

. .- ,~ "" .~

f '

f;"

"

~

.

ID,.~.....-. .-:. . . . " "

i

"7,

•

'

'

tI. 1 7

•

,

- . ' i ..k-~ • -"

"

,,.

.~,, ,

..

"-

.-~,-

..;~ . ,,. . i~

:~ ,, i

,

": .l: .',i.,~

..,

.>-{.

.....-.r.

:

..

r:.-~ .--:-..,~.

,,. ~....~< . , ~

. . ~ _ ~ i . ~ . ~ ~ . l ! ~ ,

.

.li

'.;

._"Ir~...:-,~.-~. ::,::~,.,

~ ~ ,".:~ . " ""~~ ~' ,~,,

:. :,.

.b,". ,,

'

i: "..'.i ,'- J . . ~

"i', \

"~i~.'~.:~" .....

~

.... ;

,~.,

i

'

' '.

. . . .

" •

,~..,-~>,t+f~l :-,~ ,II." * ~ " I" ,i. 7 . 1 1 ~

""~/r..

- Is. < ' . l~t~.// ~ .....

• ;

,

•

-.-.~s;ld~ .:".':li.-o..~'( ./,.'il'<~.,,,,~'..~i, z';'.,l, ;L.:"I~ "It'. ,,~ ".~.~ ~ " ~" ,, "~ " "

,.

.

~'.-,.~:,'~i" "~'-"

~.

"" "

,'~#

" 7

~- .

~ " ;,

Pill. 2. lmmunoperoxidase staining o f OFAP-positivc aslrocytes ( A - D ) . As well as in the normal CNS (A). astro~ytes ai lhe initial stage o f E A E ( B ) show no significant reaction to inflammatory lesions. A t the peak siagc, however, astrocytes react to infiltrating inflammatory cells, encase lhc lesion ( C ) and subsequenlly form glial scar al the recovery stage (D). A, normal; B, day 10 p.i., grade 1; C, day 12 p.i., grade 4; D, day 49 p.i., grade (i. A - D , ×240.

2~ s h o w n in Fig. 3 A , a s t r o c y t c s e n c a s e d i n t ] a m m a tory foci o n day 12, so t h a t m i c r o g l i a did not a t t a c h to t h e lesions. O n day 21, on t h e o t h e r h a n d , b o t h cell t y p e s i n t c r m i n g l c d wit.h e a c h o t h c r (Fig. 3 B ) . T a k e n t o g e t h c r , t h c s e f i n d i n g s s u g g e s t t h a t b r a i n cclls r c a c t to E A E lesions as follows. A t t h e early s t a g e o f E A E ( a r o u n d day 10), m i e r o g l i a r e a c t m a i n l y to i n f l a m m a t o r y loci a n d p r o l i f e r a t e . A s t r o c y t c s , on t h e o t h e r h a n d , s c e m to be still at a r c s t i n g state, a n d s h o w no signifi-

/

c a n t r e a c t i o n at this stage. A s inflammatory, cells i n c r e a s e in n u m b c r , a s t r o c y t c s react a n d e n c a s e the lcsions so that m i c r o g l i a a r c s e p a r a t e d f r o m i n f l a m m a t o r y ccll a g g r e g a t e s . H o w e v e r , t h e two ecll t y p e s i n t e r m i n g l e with e a c h o t h e r a n d f o r m a glial scar at the r e c o v e r y stage. Cell khtetics oJ" microglia and astrocTtes. T o i n v e s t i g a t e t h e k i n e t i c s o f m i c r o g l i a l a n d astroglial p r o l i f c r a t i o n , B r d U - l a b c l c d m i e r o g l i a a n d a s t r o c y t c s w e r c i d c n t i f i e d by d o u b l e - i m m u n o -

,/

A

B

C

D

Fig. 3. Double-immunofluoresccnce staining with OX42 and anti-GFAP (A and B), OX42 and anti-BrdU (C) or anti-GFAP and anti-BrdU (D). At the peak stage, astrocytes (green) encase an inflammatory cell aggregatc (red). so that microglia (red cells with dendritic morphology indicated by arrows) do not attach to the lesion (A). Then, both microglia (red) and astroc3'tcs (green) form glial scars at the recover stage (B). Double-immunofluorescence staining also dcmonstratcs that some microglia are labclcd with BrdU (C). On thc other hand. no GFAP-positive astrocytes with BrdU-positivc nuclei ~,rc present (D). An ~,strocyte in the center is not labeled with BrdtJ because a process of the astroeytc (arrows) is superimposed on a nucleus, indicating that the Iwo components do not belting to a single cell. ,4, day 12. gradc 4. ×360: B. day 21, grade 0, × 180: (". day 21. grade 0, ×360, D, day 12. grade 3, ×360.

29

fluorescence staining with anti-BrdU and OX42 or with anti-BrdU and anti-GFAP. The results are shown in Table 1 and Fig. 3C and D. About half of the microglia at the early stage of EAE (day 10) were labeled with BrdU. The percentages of BrdU-labeled microglia became lower on days 12 and 17 when rats were at the peak and recovery stages, respectively. Since the number of la-positivc microglia in and around EAE lesions in the spinal cord reached maximum at the peak stage (day 12) (Matsumoto and Fujiwara, 1986), the proliferation of microglia would precede this period. We were interested in the kinetics of astrocytes in the CNS of rats with EAE, because astrocytes are known to react to various types of lesions formed in the CNS (Ikuta et ai., 1983b). In spite of careful observation, we failed to find any GFAP-positive astrocytes labeled with BrdU throughout the course of EAE (Fig. 3D). Therefore, astroglial reactions to EAE lesions such as encasement of inflammatory cell aggregates by GFAP-positive astrocytes at the peak stage and astroglial scar formation at the later stages might be attributable to up-regulation of G F A P molecules instead of cell proliferation.

Passice E A E Microglia and astroglial reactions to inflammatory cell aggregates in passive EAE lesions were

similar to those seen in active EAE lesions. However, they were relatively mild. On day 3 of cell transfer, the earliest lesion was observed. Microglial but not astroglial reactions were observed on days 3 and 4. On days 5 and 6 the number of inflammatory cells reached maximum. In contrast to active EAE, encasement of inflammatory lesions was incomplete, and the increase of microglia was not so marked. In the recovery phase there were very few inflammatory cells, and both residual astrogliosis and microgliosis were minimal.

Discussion

It is well known that some types of brain cell proliferate or migrate in response to various pathological conditions in the CNS, and subsequently repair the lesions (lkuta et al., 1983a, b). This phenomenon is called gliosis or astrogliosis. Recent attempts to elucidate the mechanism of this phenomenon have provided several important findings. The first is related to the types of responding brain cell. Astrocytes have been well documented as cells that react to various lesions (Duffy, 1983). Also, astrocytic reactions to infiltrating inflammatory cells in the CNS of rats and mice with EAE have been investigated (Smith et al., 1983; Webster et al., 1985; Smith et al., 1987;

TABLE 1 Q U A N T I T A T I V E ANALYSIS OF BrdU-LABELED MICROGLIA AND M A C R O P H A G E S IN T H E CNS WITH ACTIVE OR PASSIVE EAE Frozen sections of the lumbar spinal cords taken from Lewis rats that had been previously injected with 50 m g / k g BrdU were double-stained with OX42 and anti-BrdU antibodies. Single- and double-positive cells with the morphological features of microglia and macrophages were enumerated separately. Percentages in parentheses are calculated as BrdU-positive microglia per total microglia. EAE

Day

OX42 ÷ mieroglia

OX42" macrophage

OX42-

examined

BrdU ÷

BrdU -

BrdU ÷

BrdU "

BrdU +

Active

10 12 17

11 (48%) 6 (24%) 6 (8%)

12 19 74

22 15 8

135 145 270

12 16 16

Passive

4 11

12 (24%) 5 (12%)

37 38

17 13

132 188

12 4

30

Cammer et al., 1990). This was mainly because astrocytes contain astrocyte-specific intermediatc fibrils. Therefore, one could easily identify astrocytcs by staining these fibrils using classical staining methods or immunostaining. On the other hand, microglia, which were first reported by Rio-Hortega (1932) almost 60 years ago, still remain poorly understood in tcrms of their localization and role under normal and pathological conditions. This is because no method for identifying this cell type has been available until recently, in 1986, we (Matsumoto and Fujiwara, 1986: Matsumoto et al., 1986) and others (Vass et al., 1986) first reported that microglia express la antigens in the CNS with EAE lesions and suggested that they are immunocompetent cells which respond rapidly to various types of CNS lesions. In the present study we have shown using immunostaining for either microglia or astrocytes, or in combination, that in the rat CNS with EAE, (1) microglia responded to inflammatory lesions earlier than astrocytes: (2) microglia first proliferated in close association with lesions, whereas astrocytes encased the lesions, thus preventing the inflammatory cell-microglia interaction; and (3) at later stages both cell types existed in residual inflammatory cell aggregates to form microastroglial scars. The most interesting finding obtained in the present study was the differential localization of microglia and astrocytes at the peak stage of EAE and the subsequent coexistence of both cell types at latcr stages. A previous in vitro study done by us revealed that microglia augment, whereas astrocytes suppress, T cell proliferation. Taken together, it is reasonable to speculate that astrocytcs may prevent microgliadriven T cell proliferation by separating microglia from infiltrating T cells. With regard to astroglial reactions, the findings obtained in the present study are slightly different from those reported previously (Smith et al., 1983, 1987; Goldmuntz et al., 1986; Aquino et al., 1988). In the present study astrogliai reactions were restricted to the sites of inflammation and were noted at the peak stage of EAE and thereafter. On the other hand, it was reported that astroglial reactions characterized by increase in the G F A P staining intensity on astrocytes in the gray matter and in the G F A P content oc-

curred extensively at early stagcs. It should bc noted, however, that as shown in both present and previous (Smith et at., 1987) studies, vigorous proliferation of astrocytes was not recognized in the CNS with EAE. There are several explanations for these discrepancies. First, it is attributable to the difference in the antigen used. Myelin basic protein (MBP) was used in this study, whereas whole myelin was used in all the above cited rcports. Since more severe EA[! is inducible by crude antigen than by purified one (Feurcr et al., 1985; our unpublished observation), it is possible to speculate that EAE in our system is milder than that reported before, thus resulting in mild astroglial reactions in sites remote from inflammatory foci. Total neural tissue destruction produced by cold injury induces much faster and stronger astroglial responses than autoimmune inflammation (Matsumoto et al., 1985). Second, there is difference in staining procedures of GFAP. Compared with paraffin-embedded sections, properly fixed frozen sections generally show more intense staining especially on protoplasmic astrocytes (our unpublished observation). Therefore, differencc in staining intensity between normal controls and EAE sections becomes less prominent in our system. Taken together with previous findings, it is strongly suggested that the magnitude of astroglial reactions is variable depending on the nature and severity of damage. Furthermore, it seems that stronger reaction occurs to tissue destruction such as infarction and demyelination rather than inflammatory cell infiltration without significant brain damage. In sharp contrast, microglia react with both types of lesion quickly. There has been considerable progrcss made in our understanding of the cytokine network (Balkwill and Burke, 1989). It is reported that cultured microglia (Giulian and Baker, 1985; Giulian and Lachman, 1985) and astrocytes (Malipiero et al., 1990; Ohno ct al., 1990) release factors that are mitogenic for each other. Sincc microglia and astrocytes in the CNS with EAE lesions were activated, both types of cell might release mitogens and proliferate in situ. BrdU immunohistochemistry revealed that microglia, but not astrocytes, proliferated in response to EAE lesions. Therefore, the latter finding suggests that astro-

31 cytic responses to E A E lesions were u n d e r t a k e n by migration of astrocytes in association with u p - r e g u l a t i o n of G F A P molecules instead of cell proliferation. However, o n e must be wary in concluding that astrocytic reactions always occur without cell proliferation. In fact, it is r e p o r t e d that some astrocytes u n d e r g o mitosis w h e n relatively severe lesions such as infarction a n d cold injury ( I k u t a et al., 1983a) are p r o d u c e d in the CNS. F u r t h e r m o r e , we employed single-pulse labeling of B r d U in o r d e r to detect B r d U - i n c o r p o rated cells in situ at a given time. T h e r e f o r e , there is a possibility that a small p o p u l a t i o n of astrocytes in S phase were not labeled by this m e t h o d . T a k i n g these possibilities into consideration, it can be said that hypertrophy of astrocytes a n d u p - r e g u l a t i o n of G F A P molecules are mainly responsible for the f o r m a t i o n of astrogliosis. Micro-astroglial scars f o u n d at later stages of E A E were an u n e x p e c t e d b u t i n t e r e s t i n g finding. A l t h o u g h we do not have any direct evidence to explain this feature, some h u m o r a l factors released by astrocytes or residual i n f l a m m a t o r y cells may play a role. As m e n t i o n e d above, m a n y cytokines have b e e n shown to be mitogenic factors for b r a i n ceils. However, chemotactic factors which i n d u c e a certain type of b r a i n cell to migrate to c e r t a i n locations r e m a i n u n k n o w n . T h e characteristic response p a t t e r n s shown by microglia a n d astrocytes at each stage of the inflammatory process in the CNS suggest that more t h a n o n e chemotactic factor is p r o d u c e d for regulation of the different stages of i n f l a m m a t i o n . In summary, we have p r o d u c e d E A E lesions in the CNS of Lewis rats a n d e x a m i n e d the resulting responses of microglia a n d astrocytes. It was f o u n d that microglia initially, a n d astrocytes subsequently, react to the lesions a n d that both cell types finally form a mixed gliai scar. Since i m m u nizations with brain-specific a n t i g e n are able to r e p r o d u c e similar i n f l a m m a t o r y processes constantly, this m e t h o d is very useful for analyzing the m e c h a n i s m of lesion repair by b r a i n cells.

References Aquino, D.A., Chiu, F.-C., Brosnan, C.F. and Norton, W.T. (1988) Glial fibrillary acidic protein increases in the spinal

cord of Lewis rats with acute experimental autoimmune encephalomyelitis. J. Neurochem. 51, 1085-1(196. Balkwill, F.R. and Burke, F. (1989) The cytokine network. Immunol. Today 10, 299-304. Boyle, E.A. and McGeer, P.L. (1990) Cellular immune response in multiple sclerosis plaques. Am. J. Pathol. 137, 575-584. Cammer, W., Tansey, F.A. and Brosnan, C.F. (1990) Reactive gliosis in the brains of Lewis rats with experimental allergic encephalomyelitis. J. Neuroimmunol. 27, 111-120. Deibler, G.E., Martenson, R.E. and Kies, M.W. (1972) Large scale preparation of myelin basic protein from central nervous tissue of several mammalian species. Prep. Biochem. 2, 139-165. Duffy, P.E. (1983) Astrocytes: Normal, Reactive and Neoplastic, Raven Press, New York, pp. 87-99. Feurer, C., Prentice, D.E. and Cammisuli, S. (1985) Chronic relapsing experimental allergic encephalomyelitis in the Lewis rat. J. Neuroimmunol. 10, 159-166. Fierz, W., Endler, B., Reske, K., Wekerle, H. and Fontana, A. (1985) Astrocytes as antigen-presenting cells. I. Induction of la antigen expression on astrocytes by T cells via immune interferon and its effect on antigen presentation. J. lmmunol. 134, 3785-3793. Fontana, A., Fierz, W. and Wekerle, H. (1984) Astrocytes present myelin basic protein to encephalitogenic T-cell lines. Nature 273-276. Fontana, A., Frei, K., Bodmer, S. and Hofer, E. (1987) Immune-mediated encephalitis: on the role of antigen-presenting cells in brain tissue. Immunol. Rev. 100, 185-201. Frei, K., Siepl, C., Groscurth, P., l:k~dmer, S., Schwerdel, C. and Fontana, A. (1987) Antigen presentation and tumor cytotoxicity by interferon-y-treated microglial cells. Eur. J. lmmunol. 17, 1271-1278. Giulian, D. and Baker, T.J. (1985) Peptides released by ameboid microglia regulate astroglial proliferation. J. Cell Biol. 101,2411-2415. Giulian, D. and Lachman, L.B. (1985) Interleukin-I stimulation of astroglial proliferation after brain injury. Science 228, 497-499. Goldmuntz, E., Brosnan, C.F., Chiu, F.-C. and Norton, W.T. (1986) Astrocytic reactivity and intermediate filament metabolism in experimental autoimmune encephalomyelitis: the effect of suppression with prazosin. Brain Res. 397, 16-26. Hayes, G.M., Woodroofe, M.N. and Cuzner, M.L. (1987) Microglia are the major cell type expressing MHC class II in human white matter. J. Neurol. Sci. 80, 25-37. Hickey, W.F. and Kimura, H. (1987) Graft-vs.-host disease elicits expression of cla~ I and class II histocompatibility antigens and the presence of scattered T lymphocytes in rat central nervous system. Proc. Natl. Acad. Sci. USA 84, 2082- 2086. Ikuta, F., Ohama, E., Yoshida, Y., Takeda, S. and Oyanagi, K. (1983a) Mitosis of the astrocyte in the cerebral lesions: its significance on the production of reactive astrocytes and lesion-repair. Prog. Neurol. Res. (Tokyo) 27, 839-856 (in Japanese with English abstract).

32 lkuta, F., Yoshida. Y., Ohama, E., Oyanagi, K., Takeda, S., Yamazaki, K. and Watabe, K. (lt~83b) Revised pathophysiology on BBB damage: the edema as an ingeniously provided condition for cell motility and lesion repair. Acta Neuropathol. Suppl. 8. 103-110. l,~tmpson, L.A. and Hickey, W.F. (19861 Monoclonal antibody analysis of MHC expression in human brain biopsies: tissue ranging from 'histologically normal' to that showing different levels of glial tumor involvement. J. lmmunol. 136, 4(154-4062. Magaud, J.-P., Sargent, I., Clarke, P.J., French, M., Rimokh, R. and Mason, D.Y. (1989) Double immunocytochemical labeling of cell and tissue samples with monoclonal antibromodeoxy-uridine. J. Histochem. Cytochcm. 37, 1517.-1527. Male. D. and Pryce, G. (19881 Kinetics of MItC gene expression and mRNA synthesis in brain endothelium, immunology 63, 37-42. Male, D.K., Prycc, G. and Hughes, C.C.W. (1987) Antigen presentation in brain: MHC induction on brain endothelium and astrocytes compared. Immunology 6~1, 453-459. Malipiero, U.V., Frei, K. and Fontana, A. (199(I) Production of hemopoietic colony-stimulating factors by astrot,3,tes. J. Immunol. 144, 3816--3821. Massa, P.T., tcr Meulen. V. and Fontana, A. (19871 Hyperinducibility of la antigen on astrocytes correlates with strain-specific susceptibility to experimental autoimmune encephalomyelitis. Proc. Natl. Acad. Sci. USA 84, 42194223. Matsumoto, Y. and Fujiwara, M. (1986) In situ detection of class I and II major histocompatibility complex antigens in the rat central nervous system during experimcntal allergic encephalomyelitis. An immunohistochemical study. J. Ncuroimmunol, 12, 265-277. Matsumoto, Y. and Fujiwara, M. ( 19871 Absencc of donor-type major histocompatibility complex class 1 antigen-bearing microglia in the rat central nervous system of radiation bone marrow chimeras. J. Neuroimmunol. 17, 71-82. Matsumoto, Y. and Fujiwara, M. (19881 Adoptively transferred experimental allergic encephalomyelitis in chimeric rats: identification of transferred cells in the lesions of the central nervous system. Immunology 65, 23-29. Matsumoto, Y. and Ikuta, F. (19851 Appearance and distribution of fetal brain maerophages in mice. Immunohistochemical study with a monoclonal antitx~dy. (;ell Tissue. Res. 239, 271-278. Matsumoto, Y., Watabe. K. and lkuta, F. (1985) lmmunohistochemical study on neuroglia identified by the monoclonal antibody against a macrophage differentiation antigcn (Mac-l). J. Neuroimmunol. 9, 379-389. Matsumoto, Y., Hara, N., Tanaka, R. and Fujiwara, M. (19861 Immunohistochemical analysis of the rat central nervous system during experimental allergic encephalomyelitis, with special referencc to la-positive cells with dendritic morphology. J. Immunol. 136, 3668-3676. Matsumoto, Y., Kawai, K. and Fujiwara, M. (1988) ttemorrhagie autoimmune encephalomyelitis induced by adoptive

transfer of activated semiallogeneic spleen cells into irradiated rats. Am. J. Pathol. 133, 3(16-315. Matsumoto. Y., Kawai. K. and Fujiwara, M. (1990a)Analysis of the T cell repertoire for myelin basic protein in thymus-grafted and other types of chimera: evidence that major histocompatibility complex molecules on accessory cells rather than T cell specificity mainly regulate susceptibility to autoimmune encephalomyelitis. Eur. J. lmmunol. 2t), 2119-2126. Matsumoto, Y., Kawai, K., Tomita, Y. and Fujiwara, M, (199(lb) Limiting-dilution analysis of the frequency of myelin basic protein-reactive T cells in Lewis, P V G / c and BN rats. Implication for susceptibility to autoimmune encephalomyelitis. Immunology 69, 215-221. MeCarron, R.M., Kempski, O., Spatz, M. and MeFarlin, D.E. (1985) Presentation of myelin basic protein by murine cerebral vascular endothelial cells. J. lmmunol. 134, 31(1(I3103. Ohmori. K., Hong, Y., Fujiwara, M. and Matsumoto, Y. (1992) In situ demonstration of proliferating cells in the rat central nervous system during experimental autoimmune encephalomyelitis. Evidence suggesting that most infiltrating T cells do not proliferate in the large! organ. Lab. Invest. (in press). Ohno, K., Suzumura, A., Sawada, M. and Marunouch, T. (19901 Production of granulocyte/macrophage colonystimulating factor by cultured astrocytes. Biochem. Biophys. Res. Commun. 169, 719-724. Rio-Hortega, del P. (19321 Microglia. In: W. Penfield tEd.), (.3,tology and Cellular Pathology of the Nervous System, Vol. 2, Iloeber, New York, pp. 483-534. Risau, W., Engelhardt, B, and Wekerle, II. (19~11 Immune function of the blood-brain barrier: incomplete presentation of protein (auto-)antigens hy rat brain mierovaseular endothelium in vitro. J. (.'ell Biol. 1111, 1757- 1766. Robinson, A.P., White, T M . and Mason, W. (19861 Maerophage heterogeneity in the rat as delineated by two monoclonal antibodies MRC OX-41 and MRC OX-42, the latter recognizing complement receptor type 3. Immunology 57, 239-247. Sasaki, A., l,evison, S.W. and Ting, J.P.-Y. (19901 Diflcrential suppression of interferon-y-induced la antigen expression on cultured rat astroglia and microglia by second messengers. J. Neuroimmunol. 29, 213-222. Smith, M.E. and Eng, L.F. (19871 Glial fibrillary acidic protein in chronic relapsing experimental allergic encephalomyelitis in S J L / J micc. J. Neurosci. Res. 18, 2(132O8. Smith, M.D.. Somera, F.P. and Eng, L.F. (1983) lmmunocytochemical staining for glial fibrillary acidic protein and the metabolism of cytoskeletal proteins in experimental allergic encephalomyelitis. Brain Res. 264, 241-253. Smith, M.E., Gibbs. M.A., Forno, LS. and Eng, L.F. (1987) [3H]Thymidine labeling of astrocytes in experimental allergic encephalomyelitis. J. Neuroimmunol. 15, 309-321. Vass, K., I,assmann, tl.. Wekerle, H. and Wisniewski, H . M (19861 The distribution or la antigen in the lesions of rat

33 acute experimental allergic encephalomyelitis. Acta Neuropathol. 70, 149-160. Webster, H.F., Shii, H. and Lassmann, H. (1985) Immunocytochemical study of myelin-associated glycoprotein (MAG), basic protein (BF), and glial fibrillary acidic protein (GFAP) in chronic relapsing experimental allergic encephalomyelitis (EAE). Acta Neuropathol. 65, 177-189. Wilcox, C.E., Healey, D.G., Baker, D., Willoughby, D.A. and

Turk, J.L (1989) Presentation of myelin basic protein by normal guinea-pig brain endothelial cells and its relevance to experimental allergic encephalomyelitis. Immunology 67, 435-440. Woodrofe, M.N., Bellamy, A.S., Feldmann, M., Davison, A.N. and Cuzner, M.L. (1986) lmmunocytochemical characterisation of the immune reaction in the central nervous system in multiple sclerosis. J. Neurol. Sci. 74, 135-152.