b7/bb-1 antigen expression on adult human microglia studied in vitro and in situ

Eur. J. Immunol. 1994. 24: 3031-3037 Microglia B7/BB-1 expression 3031

Kenneth WilliamsA, Elling Ulvestad. and Jack P. AnteP

Department of Neurology and NeurosurgeryA, Neuroimmunology Unit, Montreal Neurological Institute, McGill University, Montreal and Department of Microbiology and Immunology., The Gade Institute, University of Bergen, Bergen

B7/BB-l antigen expression on adult human microglia studied in vitro and in situ*

In this study, we have examined the expression and function of B7/BB-1 on individual glial cells, by utilizing surgically resected adult human central nervous system (CNS) tissues, tissues derived from fetal human CNS, and pathology material from cases of multiple sclerosis (MS) . Immunofluorescence analysis using enriched adult human derived cultures of microglia and oligodendrocytes, and mixed microglia/astrocyte cultures, demonstrated that B7/BB-1 was expressed on microglia. Adult human-derived oligodendrocytes and astrocytes, and human fetal astrocytes were B7/BB-1 negative under all culture conditions. Flow cytometry studies demonstrated a low basal level of B7B3B-1 expression on microglia that was up-regulated following incubation with interferon-y (IFN-y). Co-culture of purified fresh allogeneic CD4+ T cells with microglia for 24 h resulted in clustering of T cells around microglia and microglial B7/BB-1 expression. Preincubation of microglia with an anti BB-1 monoclonal antibody (mAb) prior to microglia: CD4+ T cell co-cultures resulted in partial inhibition of the ability of microglia both to present recall antigen to autologous CD4+ T cells and to present antigen to allogeneic CD4+ T cells in primary mixed lymphocyte reaction (1"MLR). The CTLA-4 Ig fusion protein inhibited the ability of microglia to present antigen in both antigen presentation assays to an even greater extent than did the anti BB-1 mAb.The BB-1 antibody also inhibited the ability of microglia to stimulate previously activated T cells in a secondary 2" MLR. In sections of multiple sclerosis brain, B7/BB-1 expression was observed on activated microglia in select parenchymal lesions, and on perivascular cells and infiltrating monocytes. B7/BB-1 immunoreactivity was not found in normal appearing white matter from MS brain or from non-inflammatory brain specimens. Our results indicate that the B7/BB-1 molecule plays a functional role in the capacity of microglia to serve as CNS antigen-presenting cells that can both initiate and perpetuate CD4+ T cell activation.

1 Introduction

The B7/BB-1 Ag is constitutively expressed on dendritic cells and on activated B cells and macrophages [l-31. B7/BB-1 expression is upregulated on B cells following stimulation with anti-IgG [l] and on monocytes following stimulation with interferon-y (IFN-y) [4]. B7 binds to CD28 [5] expressed on the majority of CD4+ T lymphocytes [6] and to the CD28 homologue CTLA-4 expressed on activated Tcells [7]. Ligation of CD28 by B7 has been demonstrated to augment T cell activation by a T cell receptor-independent mechanism, resulting in the induction of IL-2 transcription and mRNA stabilization [8, 91.

[I 129881 * This work was supported by grants from the Multiple Sclerosis

Society of Canada,The Multiple Sclerosis Society of Bergen, and the Medical Research Council of Canada.

Correspondence: Kenneth Williams, Department of Pathology, Dartmouth Medical School, One Medical Center Drive, Lebanon, NH 03756-0001, USA (Fax: +1-603 6506120)

Abbreviations: CNS: Central nervous system MNC: Mononu- clear cells MS: Multiple sclerosis

Key words: B7/BB-1/ Microglia / Central nervous system / MuItiple sclerosis

We have previously reported that adult human derived microglia can present recall antigen to ex vivo CD4+ T cells with resultant T cell proliferation and present antigen in a primary mixed lymphocyte reaction (1" MLR) [lo, 111. Human fetal astrocytes were unable to present antigen in the 1" MLR, but could support the continued proliferation of previously activated CD4+ T cells [12]. Microglia and astrocytes both express an array of immune accessory/adhesion molecules including ICAM-1 and LFA-3.The role of these molecules contributing to the antigen presentation capability of microglia and astrocytes is not defined. In this report we have investigated the expression and functional significance of B7/BB-1 by microglia in vitro and in situ. Zn vitro, a low level of B7/BB-1 Ag expression was demonstrated on microglia. The basal level of B7/BB-1 expression on microglia could be augmented following IFN-y incubation or co-culture of microglia with CD4+ T cells. Pre- incubation of microglia with an anti-BB-1 mAb inhibited the ability of microglia to present recall antigen and to present antigen in a 1" MLR, and inhibited the ability of microglia to support the proliferation of previously activated CD4+ T cells. The CTLA-4 Ig fusion protein inhibited both the recall and 1" MLR response by microglia to an even greater extent than the anti-BB-1 mAb. In situ, expression of B7BB-1 Ag on microglia was found in active MS lesions but not in parenchymal material outside of and adjacent to active 1esions.These data are discussed in terms of microglial cells functioning as CNS antigen-presenting cells (APC) and the role of B7 in augmenting CNS immune activity.

0 VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1994 0014-2980/94/1212-3031$10.00 + .25/0

3032 K. Williams, E. Ulvestad and J. P. Ante1 Eur. J. Immunol. 1994.24: 3031-3037

2 Materials and methods 2.3 CD4+ lymphocyte preparations

2.1 Tissue

CNS tissue was removed from patients undergoing surgical resections for treatment of intractable epilepsy (n = 6) (age range 22-36 years) (pathology ranged from mild to severe gliosis without evidence of inflammation or malignancy). Tissue was removed by Cavitron ultrasonic aspiration resulting in 2-3 mm3 pieces, some of which were blocked in OCT cryomount (Tissue-Tek, Elkhart, IN), frozen in iso- pentane previously cooled in liquid nitrogen, and then sectioned (4-6 pm) on a cryostat. Brain specimens from two patients with multiple sclerosis (MS) were obtained at autopsy and lesion and non-lesion areas were similarly frozen in liquid nitrogen and sectioned. Fetal human CNS tissues served as the source of astrocytes and were prepared from 8-12 week-old specimens ( n = 3) following Medical Research Council of Canada approved guidelines.

2.2 Glial cell cultures

Isolated cells from adult human tissue were obtained by a previously described protocol [ 121. In brief, aspirated tissue was washed with PBS and dissociated using 0.25 YO trypsin in the presence of 20 pg/ml DNase (Gibco, Burlington, Ontario) for 1 h at 37 "C. The resulting cell suspension was then centrifuged in 30 YO Percoll at 4 "C. Glial cells were recovered, washed in PBS and seeded in Eagles MEM supplemented with 5 % FCS, gentamicin (20 pg/ml), glu- cose (1 pg/ml) and 2 mM glutamine (Gibco) into 25-cm2 culture flasks (Nunc, Burlington, Ontario) or directly onto 9 mm Aclar flurocarbon plastic coverslips (S. Kim, Van- couver, British Columbia) previously coated with 10 pg/ml poly-L-lysine (Sigma, St Louis, MO). Cells in flasks were left undisturbed overnight at 37 "C in a humidified incuba- tor with 5 YO CO2. The following day, culture flasks were lightly shaken and floating cells consisting of > 80 % oligodendrocytes were removed and seeded onto coverslips as described above. The remaining cells consisting of mixed microglia and astrocytes were either retrypsinized and seeded onto coverslips as mixed microglia-astrocyte cultures or they were allowed to differentiate morphologically for 1 week. After 1 week, enriched microglia cultures were obtained by subjecting mixed microglia-astrocytes cultures to 5 h shaking on an orbital shaker at 150 rpm. The less-adherent astrocytes floated off and remaining microglia (> 95 % Leu-M5+ cells) were either trypsinized, seeded onto coverslips for immunofluorescence studies, and stained for flow cytometry, or seeded into 94-well microtiter plates for APC studies.

Fetal astrocyte cultures were prepared by stripping CNS materials of meninges and blood vessels and dissociating the tissue mechanically and then using 0.25 % trypsin in the presence of DNase (50 pg/ml) at 37°C for 45 min. The tissues were passed through a 130 km nylon mesh and then washed 2X in PBS. Dissociated cells were then plated onto poly-L-lysine coated culture dishes in culture medium described above. Cultures were split every 2 weeks when confluent and experiments conducted after the second or third passage when neurons and microglia were no longer apparent.

Mononuclear cells (MNC) and CD4+ T cell populations were isolated and prepared from individuals whose resected CNS tissue was used to establish microglia cultures or from healthy normal individuals as previously described [lo, 111. Briefly, MNC were isolated by Ficoll-Hypaque density centrifugation from heparinized peripheral blood samples.Tota1 T cells were isolated by rosetting h4NC with S-(2-aminoethyl isothiouronium bromide) (AET)-treated sheep red blood cells [13, 141. CD4+ Tcells were then prepared by complement-mediated lysis of CDW cells using OKT8 mAb (hybridoma obtained from ATCC, Rockville, MD) followed by the addition of rabbit complement (Cedarlane Laboratories, Hornby, Ontario). T cells were cultured in medium consisting of RPMI supplemented with 5 % human AB serum (Pel-Frez, Deer Born, WI), 2.5 pglml penicillin, and 2.5 pglml streptomycin.The mean purity of CD4+ T cells, assessed by FACScan analysis of T cells immunostained with phycoerythrein-conjugated anti-CD4 mAb (Leu3a) (Becton Dickinson, San Jose, CA) and fluorescein-labeled anti-CD8 mAb (Leu2), was > 96%.

2.4 Antigen presentation assays

2.4.1 Presentation of recall antigens

Autologous peripheral blood-derived CD4+ T cells (1 x lo5) were co-cultured in 96-well microtiter plates with 1 x lo4 irradiated microglia (2500 rad) in 200 pl of culture medium.Thereafter, 10 pl of a tetanus toxoid antigen stock solution (1 : 100 v/v dilution of a 50 protein unit per ml stock) was added to individual wells at the beginning of the assay.

2.4.2 Primary mixed lymphocyte cultures (1" MLR)

One hundred thousand allogeneic CD4+ T cells (from two different donors per experiment) were co-incubated with irradiated microglia that had been previously seeded in 96-well microtiter plates at a concentration of 2.5 x lo4 cells/well. In some experiments, non-rosetting mononuclear cells (E- cells) consisting of monocytes and B lymphocytes, were irradiated (2500 rad) and used as stimulator cells (1 x lo5 cells/well) in the MLR. Medium for all glial: T cell antigen presentation assays was RPMI 1640 supplemented with 5 % human AB serum (Pel-Frez). Recall antigen and MLR assays were carried out for 7 and 5 days, respectively. At the end of each assay, T cell proliferation was measured following a 5-h pulse with [3H] thymidine (1 pCi/well) (ICN Flow Laboratories, Mississaga, Ontar- io). Cells were harvested and radioactive counts determined using a beta liquid scintillation counter (LKB, Fisher, Montreal, Quebec). Results are expressed as mean cpm of triplicate culture wells.

2.4.3 Activated T cell cultures (2" MLR)

For these studies, purified CD4+ T cells (1 x lo7) were co-incubated with confluent allogeneic astrocyte or microglial cultures in 25-cm2 Nunc tissue culture flasks in the presence of exogenously added human rIL-2 (50 U/ml ,

Eur. J. Immunol. 1994. 24: 3031-3037 Microglia B7lBB-1 expression 3033

Genzyme, Boston, MA). Following 5 days of co-culture, diluted in PBS, pH 7.2 for 30 rnin at room temperature T cells were recovered, washed, and cultured with corres- followed by overnight incubation at 4°C in a humid ponding sister microglia or astrocytes in a MLR system as chamber with mAb diluted in 10 % rabbit serum. Biotiny- described above. lated rabbit anti-mouse immunoglobulin diluted 1 : 300

APC inhibition assays were performed by preincubating stimulator cells with either anti-BB-1 mAb (IgM, 10-100 pg/ml) (a gift from Dr. J. Ledbetter, Bristol Meyers Squibb, Seattle, WA) a CTLA-4 Ig fusion protein (2-10 vglml) (a gift from Dr. J. Ledbetter), or a HLA-DR blocking mAb L243 (IgG2a, 20 yg/ml) (a gift from Dr. R. Sekaly, CRIM, Montreal, Quebec) at 4°C for 30 min. The percent of inhibition was calculated relative to the basal response of unblocked control cultures. An anti Thy-1 mAb (IgM, 10-100 pg/ml (a gift from Dr. M. Ratt- cliffe, Montreal, Quebec) and a chimeric molecule CD40- HuIgGl (Dr. M. Kehry, Boehringer Ingelheim Corp., CT) served as control antibodies for the anti-BB-1 and CTLA-4 Ig blocking experiments respectively.

2.5 Immunofluoresence analysis

Glial cells on coverslips were analyzed by an indirect immunofluoresence labeling technique as previously described [lo]. Purity of individual glial cell populations was determined using either mouse anti-CDllc (Leu-M5, microglia) (1 : 10, Becton Dickinson), rabbit anti-glial fibril- lary acid protein (GFAP) (astrocytes) (1: 100, Dako, West Chester, PA), or mouse anti-GalC (H8H9, oligodendrocytes) (1:10, a gift from Dr S Kim, Vancouver, British Columbia) followed by rhodamine-conjugated goat anti- mouse antibodies (1 : 150, Cappel, Lexington, MA) or FITC-conjugated goat anti-rabbit antibodies (1 : 150, Cap- pel). To immunostain enriched microglia and oligodendro- cyte cultures for B7/BB-1, live cells on coverslips were first incubated with the anti-B7/BB-l mAb (1: 100, Becton Dickinson) for 1 h at 4°C followed by rhodamine-conjugated goat anti-mouse Ab for 45 min at 4 "C, and then cells were fixed using acid alcohol ( 5 % glacial acetic acid: 95 % absolute ethanol, v/v) at - 20°C for 15 min. Mixed microglia-astrocyte cultures under basal culture conditions and following activation with human rIFN-y (100 U/ml, Gen- zyme) were double stained using a B71GFAP protocol where viable cells on coverslips were first incubated with anti-B7/BB-l (1:100, Becton Dickinson) for 1 h at 4°C followed by rhodamine-conjugated goat anti-mouse Ab for 45 min at 4"C, and then fixed and stained for GFAP as described above. Enriched microglial cultures were also analyzed for basal and IF"-y (100 U/ml) induced levels of B7/BB-1 expression compared with basal and induced HLA-DR (anti HLA-DR-FITC, Becton Dickinson, 10 yglmi) by flow cytometry. For flow cytometry staining, microglia were trypsinized from culture flasks, washed twice with 1 % BSA in PBS, incubated with primary mAb for 45 min on ice, washed twice, and then incubated with FITC-conjugated goat anti-mouse IgG for 45 min on ice. Cells were then fixed with a 1 % solution of paraformalde- hyde in PBS. Controls consisted of unstained cells, the omission of primary antibody, and irrelevant isotype controls.

2.6 Immunohistochemistry

Sections were fixed in cold acetone for 5 min. Following fixation, sections were incubated with 10 % rabbit serum

were applied for 30 min. Avidin-bgtin-horseradish peroxidase complex (ABC complexMRP; Dakopatts, Glostrup, Denmark)) was prepared according to the manufacturer's recommendation and allowed to react with sections for 30 min. Thereafter sections were treated with a 3-amino- 9-ethyl-carbazole containing buffer and the colored reaction product monitored. Sections were lightly counter- stained with hematoxylin, mounted in glycergel (Dako- patts) and examined by light microscopy. In addition to staining with B7/BB-1 mAb, alternate serial sections were stained with Leu-M3 mAb (CD14, Becton Dickinson) and subjected to the nonspecific esterase (NSE) reaction as previously described [lo] to differentiate between infiltrating monocytes (Leu-M3+, NSE+) and microglia (Leu- M3-) [15, 161.

3 Results

3.1 Expression and regulation of B7/BB-1 on microglia in vitro

Enriched cultures of adult human-derived microglia (> 95 % Leu-M5+) (Fig. 1), oligodendrocytes (> 80 % GalC+), and mixed microglidastrocyte cultures were esta- blished as previously reported [lo-131. Microglia in either

Figure 1 . Adult human-derived microglia in vitro (enriched cultures > 95 % purity). Leu-MS immunoreactivity was demonstrated by fixing cells with ice-cold acetone for 5 rnin followed by incubation of Leu-M5 mAb ( 1 : l O ) for 1 h. A biotinylated rabbit anti-mouse immunoglobulin (1 : 300) was thenincubated for 30 rnin followed by an avidin-biotin-horseradish peroxidase complex (ABC/HRF') for 30 min. Control stains on sister microglia cultures consisting of an IgG2a isotype control and omision of the primary antibody were performed and no reaction product was seen. The colored reaction product was developed using 3-amino-9-ethyl- carbazole containing buffer (160X).

3034 K. Williams, E. Ulvestad and J. P. Ante1 Eur. J. Immunol. 1994.24: 3031-3037

enriched cultures or in mixed astrocyte/microglial cultures were B7/BB-1 positive following incubation with IFN-y as assessed using fluorescence microscopy. Oligodendrocytes and astrocytes were B7/BB-1 negative under all culture conditions. Results from flow cytometry studies demonstrated a low basal level of B7/BB-1 by microglia with variability of expression between individual brain specimens (n = 3); (Fig. 2). HLA-DR was more readily detected on microglia under basal culture conditions than was B7/BB-1 using flow cytometry (Fig. 2); both B7IBB-1 and HL A-DR expression was increased on microglia following stimulation with IFN-y (100 U/ml, 24 h; Fig. 2). Co- incubation of 1 x lo5 allogeneic CD4+ T cells with 1 x lo3 microglia for 18 h resulted in clustering of T cells on microglia and the induction of B7/BB-1 expression on microglia (Fig. 3).

3.2 The ability of microglia to present antigen is inhibited by anti-B7/BB-1 antibody and CTLA-4 Ig

As previously demonstrated [lo, 111 microglia were able to present tetanus toxoid proteins to ex vivo autologous CD4+ T cells, and present antigen in 1" MLR. Preincubation of microglia with the anti-BB-1 mAb (10-100 pg/ml) partially inhibited the ability of microglia to present antigen in a 1" MLR (Table 1). Pre-incubation of microglia with the CTLA-4 Ig fusion protein (2-10 pg/ml) inhibited the ability of microglia to present antigen in both the recall assay and the 1" MLR to an extent greater than the anti BB-1 mAb (Table 1). The inhibition of both the recaIl and 1" MLR responses by microglia using the CTLA-4 Ig fusion protein

"i" IFN M W

Figure2. B7/BB-1 and HLA- DR antigen expression on microglia in vitro. Microglia were stained either with anti-BB-1 mAb followed by FITC-labeled goat-anti-mouse or with a FITC- labeled anti-HLA-DR mAb. Histograms represent 10000 ev- ents anaIyzed showing (A) basal B7/BB-1 expression on enriched microglial cultures compared with an isotype control. (B) B7/BB-1 on IFNy (100 U/ml, 24 h) stimulated microglia compared to isotype control. (C) Basal and IFN-y-stimulated HLA-DR expression on microglia .

was similar to the inhibition seen using the anti-HLA-DR mAb L243 (Table l).The anti BB-l/B7 mAb did not inhibit the ability of non-rosetting E- cells consisting of monocytes and B cells to stimulate allogeneic T cells in the MLR (39861 f 4407 vs. 36660 k 1934; n = 3). Pre-incubation of microglia with the anti-BB-1 mAb and the HLA-DR blocking mAb inhibited the ability of microglia to stimulate previously activated CD4+ T cells in a 2" MLR (Table 1).

3.3 MS brain B'IIBB-1 expression

Immunohistochemical analysis of tissue sections from the same material used to establish enriched microglial cultures

Figure 3. B7/BB-1 expression on microglia is induced with IFN-y by co-culturing microglia with allogeneic CD4+ T cells. (A) Phase contrast micrograph of microglia (1 X lo3) seeded onto coverslips and co-cultured with CD4+ Tcells and compared with sister microglia cultures that were cultured without T cells (inset) (120X). (B) Same field as (A) demonstrating B7/BB-1 expression on microglia co-cultured with CD4+ T cells over that of non treated microglia [inset, same field as above, inset (A)] (120X). (C) Incubation of microglial cultures with IFN-y for 24 h induces B7/BB-1 antigen expression (300X).

Eur. J. Immunol. 1994.24: 3031-3037 Microglia B7BB-1 expression 3035

demonstrated that these tissues were negative for B7/BB-1. Within active MS lesions B7/BB-1 immunoreactivity was found in selective areas on reactive microglia (Fig. 4). I n alternate serial sections the parenchymal cells were found to be Leu-M3- and NSE- discriminating them as microglia rather than recently infiltrated monocytes/macrophages [lo, 15, 171. No B7A3B-1 immunoreactivity was found within non-lesion parenchymal MS brain material, though perivascular microglia and perivascular infiltrates, most likely representing T cells and infiltrating monocytes, were also positive for the reaction product (Fig. 4).

Table 1. Inhibition of the antigen presentation capacity of adult human-derived microglia with anti BB-1 mAb, and the CTLA-4 Ig fusion proteina)

Antigen presentation Inhibition assay Blocking Ab Mean Range

Recall antigen Anti BB-1 CTLA-4 Ig

Anti-HLA-DR 1" MLR Anti BB-1

CTLA-4 Ig Anti-HLA-DR

Anti-HLA-DR 2" MLR Anti BB-1

37 % 51 % 67 %

22 % 73 % 71 %

59 % 61 %

19-55 % (n = 8) 20-67 % (n = 8) 56-72% (n = 3) 15-42% (a = 6) 62-77% (n = 3) 65-76% (n = 3) 53-65% (n = 2) 53-69 % (n = 2) 4 Discussion

a) Percent inhibition expressed as the quotient of cpm of APC + T cells + blocking mAblcpm of non-treated controls. Mean cpm of recall antigen = 31 105 f 2479 cpm (n = 8); 1" MLR = 24 210 f 1200 cpm ( n = 6); 2" MLR = 25 233 f 1382 (n = 2). Isotype and fusion protein controls resulted in a maximal inhibition of 8% for anti-Thy-1 (IgM) (n = 4) and 12% for hIgG1-mCD40 (a = 4) for 1" MLR. CD4+ T cells alone or microglia alone < 900 cpm; () denotes number of experiments.

Ligation of theTcR by an APC is a minimal requirement for antigen-restricted T cell stimulation. Secondary signals in addition to that provided by MHC class II molecules on the APC are required for resultant T cell activation [18, 191. Whether the second signals are cytokines [20,21] or cell surface adhesion molecules [22,23] or both is not resolved. Several studies have shown that accessory molecules including LFA-3,VCAM-1, B7/BB-1, and ICAM-1 expressed on the APC contribute critical signals to T cells. The microglia and astrocytes in our assay system both express the immune accessoryladhesion molecules MHC class 11, ICAM-1 and LFA-3, but only microglia expressed B7/BB-1. B7/BB-1 was up-regulated on microglia following incubation with IFN-y and co-cultivation of microglia with purified CD4+ T cells. The selective contribution of each of these molecules for primary versus secondary Tcell responses is, however, not clearly defined although several studies emphasize the importance of B7/BB-1 and CD28 interac- tion in CD4+ T cell stimulation [24-271. The ability of anti B7/BB-1 mAb to partially inhibit allogeneic T cell responses has been demonstrated using dendritic cells [3, 281, EBV-transformed B cells [29] or T cells [30,31] as the stimulator cells. Though activated human peripheral blood monocytes have been demonstrated to express B71BB-1 antigens in vitro, inhibition studies using anti B71BB-1 mAb do not demonstrate a functional role for B7/BB-1 expression on these cells [4]. We have previously demonstrated that adult human-derived microglia are able to present recall antigen to T cells [lo], to present antigen to ex vivo CD4+ T cells in 1" MLR [ l l ] , and to support the proliferation of previously sensitized T cells in a 2" MLR. Human fetal astrocytes are able to support the continued proliferation of previously activated T cells, but can not present antigen in a 1" MLR without the addition of either human rIL-2, rIL-1, or non-priming "contaminating" con- centrations of microglia [14]. In our system, the human fetal astrocytes may be similar to keratinocytes that also express high levels of ICAM-1 and €&A-DR antigens following exposure to IFN-y, but are B7-negative cells [32]. When keratinocytes are used as stimulators in primary APC assays the activation of T cells does not occur and instead the T cells become anergic or apoptotic [32-341.

Figure 4 . Immune peroxidase demonstration of B7/BB-1 on frozen tissue sections from MS brain. (A) Reaction product on the surface of perivascular cells (arrows, 240 X). (3) Reaction product on infiltrating cells in a perivascular cuff (arrows; 240X). (C) Reaction product on reactive microglia in the parenchyma of an active M S lesion (240 X> . "

Functionally, we could demonstrate a partial inhibition of both recall antigen responses and antigen presentation in the 1"MLR by pre-incubating microglia with an anti- B7/BB-1 mAb and the CTLA-4 Ig fusion protein, with more convincing numbers using the CTLA-4 Ig protein. The limited and variable capacity of the anti-B7/BB-l mAb comDared to the CTLA-4 Ie to inhibit the ability of

3036 K. Williams, E. Ulvestad and J. P. Antel Eur. J. Immunol. 1994. 24: 3031-3037

microglia to function as an APC might be explained by a 20-fold greater affinity for the B7/BB-1 Ag by the CTLA4- Ig [S]. In addition, CTLA4-Ig has the capacity to bind alternative ligands. Azuma et al. [35] recently identified a second ligand for CTLA4 and CD28, the B70 antigen, which is constitutively expressed on monocytes and dendritic cells. A mAb reactive with the B70 antigen efficiently inhibited 1" MLR whereas a mAb reactive with B7/BB-1 only partially inhibited the MLR. We can therefore not exclude the possibility that microglia also express B70 or similar yet-to-be defined CTLA-4 ligands. Previous reports have demonstrated B7 Ag expression on professional APC including skin Langerhans cells, veiled cells of lymph node sinuses, and interdigitating cells of spleen and lymph nodes on non-pathologic tissues and the expression of B7/BB-1 that was induced on macrophages and epithelial cells in granulomatous inflammation [36].

In the current study, using frozen sections of MS brain material, BB-1 antigen expression was demonstrated on activated microglia in select lesions. We did not detect B7/BB-1 expression within adjacent non-lesion sections from MS brain or from the resected tissues from which microglial cultures were derived. B7/BB-1 expression was noted on perivascular cells and infiltrates resembling monocytes and T cells indicating some level of activation of immune cells infiltrating the CNS.

CD4+ T cells have been demonstrated to play a critical role in MS and EAE pathology. Rodent studies have demonstrated that activated T cells preferentially enter the CNS irrespective of MHC class restriction or antigen specificity [37]. Whether T cells in the CNS are further stimulated to proliferate or instead become anergic or apoptotic [38-401 might depend both upon the brain microenviroment and putative CNS APC. Both astrocytes [41, 421 and microglia [43,44] have been demonstrated to function as APC in vitro, though microglia are the most likely cell to function as resident CNS APC in situ [45-471. One report has demonstrated that perivascular microglia can present antigen in vivo [47]. Early studies have suggested that there might exist a correlation between the induction of rodent MHC class I1 expression and susceptibility to EAE [48] although these data recently been questioned [49]. Our data suggest that T cell immune reactivity in the CNS might be as much a result of aberrant B7/BB-1 expression as MHC class I1 induction. Our previous data demonstrate that adult human derived microglia represent a glial cell type that can present recall antigen and present antigen in a MLR to ex vivo CD4+ T cells and suggest that in vivo, microglia might function to support either the initial or continued activation of T cells that cross blood brain barrier. Our data implicate the expression of B7/BB-1 on microglia in these responses.

Received March 4,1994; in revised form August 15,1994; accepted September 7, 1994.

5 References

1 Freedman, A. S., Freeman, G., Horowitz, J. C., Daley, J. and

2 Yokochi,T., Holly, R. D. and Clark, E. A., J. Immunol. 1981. Nadler, M., J. Immunol. 1987. 139: 3260.

128: 823.

3 Young, J. W., Koulova, L., Soergel, S. A., Clark, E. A., Steinmann, R. M. and Dupont, B., J. Clin. Immunol. 1992. 90: 229.

4 Freedman, A. S., Freeman, G. J., Rhynhart, K. and Nadler, L. M . , Cell. Immunol. 1991. 137: 427.

5 Linlsley, P. S., Clark, E. A. and Ledbetter, J. A., Proc. Natl. Acad. Sci. USA 1990. 87: 5031.

6 Hara,T., Fu, S. M. and Hanses, J. A., J. Exp. Med. 1991.161: 1513.

7 Linsley, P. S., Brady,W., Urnes, M., Grosmaire, L. S., Damle, N. K. and Ledbetter, J. A., J. Exp. Med. 1991. 174: 561.

8 Fraser, J. D., Irving, B. A., Crabtree, G. R. and Weiss, A,, Science 1991. 251: 313.

9 Lindstein, T., June, C. H., Ledbetter, J. A., Stella, G. and Thompson, C. B., Science 1989. 244: 339.

10 Williams, K., Bar-Or, A,, Ulvestad, E., Olivier, A,, Antel, J. P. and Yong,V. W., J. Neuropath. Exp. Neurol. 1992. 51: 538.

11 Williams, K., Ulvestad, E., Cragg, L., Blain, M. and Antel, J. P., J. Neurosci. Res. 1993. 36: 382.

12 Williams, K., Dooley, N. P., Ulvestad, E., Waage, A., Yong, V. W. and Antel, J., J. Neurosci. 1994, in press.

13 Yong,V. W. and Antel, J. F?, in Richardson, A. and Fedoroff, S. (Eds.) Protocols for neutral cell culture. Humana Press, St. Louis, 1992, 81.

14 Pellegrino, M. A., Ferrone, S. and Theofilopoulos, A. N., J. Immunol. Methods 1976. 11: 273.

15 Ulvestad, E.,Williams, K., Matre, R., Nyland, H., Olivier, A. and Antel, J. P., J. Neuropath. Exp. Neurol. 1993. 53: 27.

16 Ulvestad, E. ,Williams, K., Mork, S., Antel, J. and Nyland, H., J. Neuropath.Exp. Neurol. 1994. 53: 492.

17 Ulvestad, E.,Williams, K., Matre, R., Nyland, H. and Antel, J. P., J. Neurol. Sci. 1994. 121: 125.

18 Mueller, D. L., Jenkins, M. K. and Schwartz, R. H., Annu. Rev. Immunol. 1989. 7: 445.

19 Schwartz, R. H., Science 1990. 248: 1349. 20 Kurt-Jones, E. A., Beller, D. I., Mizel, S. B. and Unanue,

21 Koide, S. and Steinman, R. M . , J. Exp. Med. 1989. 168: 409. 22 Springer, T. A., Nature 1990. 346: 425. 23 Steinman, R. M. and Young, J. W., Curr. Opin. Immunol.

24 Azuma, M., Cayabyab, M., Buck, D., Philips, H. and Lanier,

25 Norton, S. D., Zuckerman, L., Urdahl, K. D., Shefner, R.,

26 Damle, N. K., Klussman, K., Linsley, P. S. and Arrufo, A., J.

27 Damle, N. K., Klussman, K., Linsley, P. S., Aruffo, A. and

E. R., Proc. Natl. Acad. Sci. USA 1985. 82: 1204.

1991. 3: 361.

L. L., J. Exp. Med. 1992. 175: 353.

Miller, J. and Jenkins, M. K., J. Immunol. 1992. 149: 1556.

Immunol. 1992. 148: 1985.

Ledbetter. J.. J. Immunol. 1992. 149: 2541. 28 Symington, F. W., Brady, W. and Linsley, P. S., J. Immunol.

1993. 150: 1286. 29 Koulova, L., Clark, E. A., Shu, G. and Dupont, B., J. Exp.

Med. 1991. 73: 759. 30 Azuma, M., Yessel, H., Phillips, J. H., Spits, H. and Lanier,

L. L., J. Exp. Med. 1993. 177: 845. 31 Wyss-Coray, T., Mauri-Hellweg, T., Baumann, K., Betters, F.,

Grunow, R. and Pichler,W., Eur. J. Immunol. 1993.23: 2175. 32 Nicholoff, B. J., Mitra, R. S., Lee, K. ,Turka, L. A,, Green, J.,

Thompson, C. and Shimizu,Y., Am. J. Pathol. 1993.142: 1029. 33 Gaspari, A. A., Jenkins, M. Y and Katz, S. I., J. Immunol.

1988. 141: 2216. 34 Bal,V., McIndoe, A., Denton, G., Hudson, D., Lombardi, G.,

Lamb, J. and Lechler, R., Eur. J. Immunol. 1990. 20: 1893. 35 Azuma, M . , Ito, D.,Yagita, H., Okumura, K., Phillips, J. H.,

Lanier, L. L. and Somoza, C., Nature 1993. 366: 76. 36 Vandenbergh, P., Delabie, J., de Boer, M., De Wolf-Peeters, C.

and Ceuppens, J. L., Int. Immunol. 1993. 5: 317. 37 Hickey, W. F., Hsu, B. L. and Kimura, H., J. Neurosci. Res.

1991. 28: 254. 38 Pender, M. I?, McCombe, P. A.,Yoong, G. and Nguyen, K. B.,

J. Autoimmun. 1992. 5: 401.

Eur. J. Immunol. 1994. 24: 3031-3037

39 Schmied, M., Breitschopf, H., Gold, R., Zischler, H., Rothe, G., Wekerle, H. and Lassmann, H. , Am. J. Pathol. 1993. 43: 446.

40 Matsumoto, Y., Hanawa, H. , Tsuchida, M. and Abo, T., Immunol. 1993. 390: 81.

41 Fontana, A. , Fierz,W. and Wekerle, H., Nature 1984.307: 273. 42 Fierz,W., Endler, B., Reske, K.,Wekerle, H. and Fontana, A. ,

43 Frei, K., Siepl, D., Grosscarth, D., Bodmer, S., Schwerdel, C.

44 Matsumoto, Y., Ohmori, K. and Fujiwara, M., immunology

J. Immunol. 1985. 134: 3785.

and Fontana, A,, Eur. J. Immunol. 1987. 17: 1271.

1992. 76: 209.

Microglia B7/BB-1 expression 3037

45 Hayes, G. M. ,Woodroofe, M. N. and Cuzner, M. L., 1. Neurol.

46 Matsumoto, Y., Hara, N., Tanaka, R. and Fujiwara, M., J.

47 Hickey, W. F. and Kimura, H. , Science 1988. 239: 290. 48 Massa, F! T., ter Meulen,V. and Fontana, A., Proc. Natl. Acad.

49 Sedgwick, J. D., Schwender, R., Gregersen, R., Dorries, R.

Sci. 1987. 80: 25.

Immunol. 1986. 136: 3668.

Sci. USA 1987. 84: 4219.

and ter Meulen,V., J. Exp. Med. 1993. 177: 1145.

b7/bb-1 antigen expression on adult human microglia studied in vitro and in situ

Documents