Immunopathology of the lesion in chronic relapsing experimental autoimmune encephalomyelitis in the mouse

CELLULAR

IMMUNOLOGY

99,

395-410 (1986)

lmmunopathology of the Lesion in Chronic Relapsing Experimental Autoimmune Encephalomyelitis in the Mouse U. TBAUGOTT’,

D. E. MCFARLIN,

AND C. S. RAINE

Departments of Pathology (Neuropathology), Neurology, and Neuroscienceand the Rose F. Kennedy Centerfor Research in Mental Retardation and Human Development, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, N. Y. 10461; and the Neuroimmunology Branch, NINCDS, Bethesda, Maryland 20205 Received August 29, 1985; acceptedJanuary 6, I986 To analyze immunopathologic events within the central nervous system (CNS) during various stagesof actively induced chronic relapsing EAE in SJL/J mice, animals were sampled at various timepoints post inoculation (PI) and T cells, T-cell subsets,Ia+ cells and Ig+ cells, albumin, and Ig deposits were localized in frozen sections by immunocytochemical techniques. Furthermore, sections were stained for the demonstration of Ia antigen, myelin basic protein (MBP), and galactocerebroside(CC) on endothelial cells and astrocytes.During the acute phaseof the disease, large numbers of all types of inflammatory cells studied (Lyt- 1.2+,L3T4+, Lyt-2+, Ia+, Ig+) were randomly distributed throughout lesions, a finding similar to that described previously for acute EAE. A more distinct distribution pattern of infiltrating T cells was found during active chronic diseasein that L3T4+ cells predominated within the CNS parenchyma, while Lyt-2+ cells were more numerous in meningeal and perivascular areas.During all chronic stages,a low-grade diffuse infiltration of the neuraxis by hematogeneous cells was present. Ia and myelin antigens were detectableon some endothelial cells and astrocytes.Damage to the blood-brain barrier, as indicated by albumin and Ig deposits, was more extensive during the acute than during chronic stagesof the disease.Taken in concert, the resultsfurther support the possibility of local antigen presentation on endothelial and astroglial cells and an essential involvement of helper (L3T4+) T cells in CNS lesion formation. These findings correlate well with events reported previously in acute and chronic multiple sclerosislesions. 0 1986 Academic Ress, 1~ INTRODUCTION

Although it is well established that experimental (autoimmune) allergic encephalomyelitis (EAE) is a T-cell-mediated condition (l-3), detailed analysis of the immunopathologic changesoccurring within the central nervous system(CNS), the target organ of this diseaseprocesshasbeen hampered by the unavailability of specific markers applicable to the localization in situ of components of the immune system. The induction of reproducible models of acute and chronic EAE in the mouse (4-7), the speciesmost widely studied immunologically, has afforded the opportunity to apply well-characterized monoclonal antibodies (mAb) against various components of the immune system to the investigation of CNS lesion formation in this autoimmune ’ To whom correspondence should be addressed: Department of Pathology (Neuropathology), Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, N.Y. 10461. 395 0008-8749/86 $3.00 Copyright Q 1986 by Academic Press,Inc. All rights of reproduction in any form reserved.

396

TRAUGOTT,

MC FARLIN, AND RAINE

demyelinating disease (8, 9). In a recent study on the distribution of T cells, T-cell subsets,B cells, and Ia+ cells in CNS tissue from mice with acute EAE, Lyt-l+, and Lyt-2+ cells were demonstrated in perivascular and meningeal infiltrates as well as throughout the white matter parenchyma (IO, 11). Lyt- l+ cells seemedto invade the CNS parenchyma earlier than Lyt-2+ cells, while B cells and Ig+ macrophagesappeared even later. The observed presence of Ia and myelin antigens (myelin basic proteinMBP; galactocerebroside-GC) on endothelial cells early on was considered suggestive of local antigen presentation (10). Since acute EAE, with its monophasic course, has only limited applicability to multiple sclerosis (MS), the human prototype of demyelinating diseases(12) the present study was undertaken on chronic relapsing EAE in the mouse. The results suggestdistinct distribution patterns of inflammatory cells in acute and chronic lesions reminiscent of the findings on acute and chronic MS lesions ( 13, 14) and the presence of small numbers of inflammatory cells throughout the CNS during all stagesof chronic relapsing EAE, might reflect continuous diffuse involvement of the neuraxis. MATERIAL

AND METHODS

Mice inoculated for chronic relapsing EAE were kept in groups of 3-5 per cage,fed chow and water ad libitum, and checked daily for neurologic signs. At various timepoints post inoculation (PI; latent period, first episode of clinical signs, remission, and relapse), mice were anesthetized with ether and perfused through the heart with 20 ml of ice-cold phosphate-buffered saline (PBS). Induction of chronic relapsing EAE. For the induction of chronic relapsing EAE, a group of 20 female SJL/J mice (8 weeks old; Jackson Laboratories, Bar Harbor, Me.) were inoculated subcutaneously twice at 7-day intervals in the flank in multiple sites with a total of 0.3 ml of an emulsion consisting of 1 mg lyophilized spinal cord in 0.15 ml PBS and 0.15 ml of complete Freund’s adjuvant (CFA-H37RA Difco Laboratories) containing 0.03 mg killed M. tuberculosis hominis (4). Tissue preparation. After perfusion with PBS, the CNS was removed and frozen blocks were prepared from brain, cerebellum, brain stem, and three levels each of the cervical, thoracic, and lumbar spinal cord using OCT (optimal cooling temperatureTiter Tek) compound and an acetone dry/ice bath. Frozen sections were cut on a cryostat at 10 pm and stored at -20°C until used. Clinical rating. Mice were evaluated neurologically according to a 5 point scoring system (1 = normal; 2 = decreasedtail tonus; 3 = hind limb paresis and incontinence; 4 = quadriparesis; 5 = moribund). Pathologic evaluation. For the evaluation of diseaseactivity according to the degree of inflammation, sections from each block were stained with H and E. Immunocytochemical staining procedures.For this study, various fixation procedures were tested to obtain optimal results. These included acetone; methanol; ethanol; buffered Formalin, and glutaraldehyde. The latter three fixatives were applied at various concentrations and the staining procedure was then performed with or without an attempt to unmask antigenic sites by exposure to enzymes. Acetone gave the best results for staining of T cells, B cells, and GC, while Ia antigen and MBP showed better preservation with 95% ethanol. To control for possible variations in the staining reaction, sections from normal lymph node or brain were carried along each time experimental tissue was processed.

IMMUNOPATHOLOGY

OF CHRONIC EAE

397

Avidin-Biotin Complex (ABC) technique. The ABC technique was applied for labeling of Lyt- 1.2+ cells (pan T cells), Lyt- 1+ cells (T cells, some B cells: results not shown), L3T4+ (helper/inducer) T cells, Lyt-2’ (suppressor/cytotoxic) T cells, la+ cells (B cells, macrophages, activated T cells, some endothelial cells), lg+ cells (B cells, lgcontaining macrophages), and for the localization of MBP and GC in myelin and possibly on some endothelial cells ( 15, 16). For this, frozen sections were air-dried for 20 min and fixed in acetone (Lyt-1.2+, L3T4+, Lyt-2+, B cells, galactocerebroside) or 95% ethanol (la, MBP) for 5-10 min. Endogenous peroxidase was inactivated and Fc receptors were blocked by consecutive exposure for 30 min each to 0.1% Hz02 in Tris buffer containing 0.15 M NaCl(Tris/saline, pH 7.6) and Tris/saline containing 5% normal serum from the species from which the biotinylated antibody was derived. This was then followed by incubation with rat monoclonal antibodies against T cells, T-cell subsets,and la+ cells diluted 1:lOO-1:500 in 5% normal rabbit serum-containing Trislsaline for 90 min. Afterwards, sections were exposed consecutively for 60 min each to the biotinylated antibody (raised in rabbits), diluted 1:200 followed by the avidin-biotin-peroxidase complex (ABC) freshly prepared according to the company’s (Vectastain) specifications. A brown color reaction was then obtained on specifically labeled cells by exposure to 3’,3’-diaminobenzidine at a concentration of 50 mg % in Tris buffer (pH 7.6) containing 0.01% H202. After a positive reaction product was obtained, sectionswere rinsed in PBS(pH 7.4) and one ofthe slides,stained in duplicate, was counterstained with hematoxylin for nuclear staining to facilitate morphologic evaluation. All sections were then dehydrated in increasing concentrations of alcohol, cleared in xylene, and mounted in Flo-Texx (Lerner Laboratories). The intensity of inflammation was rated on a four point scale (++ to -), based upon the number of labeled cells per CNS section. Monoclonal antibodies and antisera. Rat monoclonal antibodies to Lyt-1 and Lyt-2 were purchased from Becton-Dickinson (Calif.). Rat clones, producing anti-IaS’ and anti-L3T4 mAb were obtained from American Type Culture Collection (Md.) and maintained in this laboratory. For the localization of MBP and GC, rabbit antisera raised in this laboratory (17) were used. Use of the latter antisera required the introduction of a biotinylated sheep antibody as the first link antibody. All other steps of the staining procedure were identical to those described above for rat mAb. Protein-A column. IgG was isolated from the culture supernate of the GKI.5 (antiL3T4) clone and the anti-laSf producing clone by affinity chromatography using a Protein A cloumn (18). For this, the pH of the culture supemate was adjusted to 8.6 and this was then passedthrough the Protein A column. After a rinse with Tris buffer (pH 8.6) a step-wiseelution at pH 7.0, pH 5.0, and 4.3 was performed and the column was then regenerated by a rinse with glycine buffer at pH 2.3. Protein-containing fractions were pooled, dialyzed against Tris buffer (pH 8.2) and concentrated to an IgG content of 1 mg/ml. Direct fluorescence. Deposits of albumin and lg (IgG + IgA + IgM) were demonstrated by direct fluorescence. For this, sections were air dried for 20 min, fixed in 95% ethanol for 10 min, rinsed in PBS, and then incubated for 60 min with FITCconjugated F(ab’)* fragment of sheep anti-mouse lg (IgG + IgA + IgM, heavy and light chain specific; Tago Labs) or FITC-conjugated IgG fraction of goat anti-mouse albumin (Cappel Labs) diluted 1:40 and 1:100, respectively, in PBS containing 5% normal sheep or goat serum. After appropriate rinses in PBS, sections were mounted

398

TRAUGOTT,

MC FARLIN, AND RAINE

in PBS-buffered glycerine and evaluated on a Zeiss Photomicroscope III, equipped for phase and fluorescence microscopy. RESULTS Clinical Findings The majority of mice ( 17 out of 20) inoculated for chronic relapsing EAE developed neurologic signs (loss of tail tonus, incontinence, hind limb paresis) between Days 12 and 2 1 PI from which most recovered completely within 7- 10 days. At various time intervals thereafter, some mice developed up to two spontaneous relapses. After a relapse, the neurologic deficit tended to remit only partially. Two mice remained clinically healthy for up to 15 months after they had recovered from the initial episode of the disease.One mouse, sacrificed at 14 days PI had not developed neurologic signs. Pathology Latent period. In H and E stained sections,mild to moderate inflammatory infiltrates, consisting of mononuclear cells and polymorphonuclear leukocytes (PMNs) were seen in the meninges and Virchow-Robin spacesand a few inflammatory cells could also be detected within the CNS parenchyma. Infiltrates were more intensive in brain, brainstem, and cerebellum than in the spinal cord. First episode of signs. With onset of signs (Day 12 PI), the overall density of inflammatory infiltrates increased (Fig. 1). While at Day 12 PI, rostra1 CNS areas were

FIG. 1. Brain, sampled 7 days PI, stained with H + E. Inflammatory cells are seen in perivascular areas and in the parenchyma (ependyma on top). X200.

IMMUNOPATHOLOGY

OF CHRONIC EAE

399

predominantly affected,from Day 17 PI onwards, the spinal cord becameprogressively involved displaying infiltrating cells in the meninges, in perivascular areas,and within the parenchyma. Relapse and chronic progressive disease. CNS tissue from mice sampled during a relapse or during chronic progressive diseaseshowed moderate numbers of infiltrating cells sometimes with a lesion-like pattern. Remission and clinically silent disease. Only a few hematogenous cells were detectable in the meninges and Virchow-Robin spacesin CNS tissue from mice sampled after neurologic signs had resolved and no new signs had developed for up to 15 months. Immunopathology Scoring system. The scoring system used for the evaluation of immunopathologic changesis based upon the density and distribution of labeled cells which in numerical terms has been found to correspond to the following: - = none; + = 95.3 f 48.6 (mean f 1SD); + = 451.0 * 275.9; and ++ = 1396.3 + 370.6 labeled cells per brain section. When cell numbers were compared in meningeal (perivascular) and parenchymal areas, a ratio of 0.96 f 0.4, and 2.13 * 1.00 was found for ++, +, and + scores,respectively. This reflected a predilection of infiltrating cells for the meninges over the parenchyma in low-grade (&) inflammation while hematogeneouscells seemed to be more equally distributed in the meninges and parenchyma with increasing density of infiltrates.

FIG. 2. Cerebellum, sampled 7 days PI, stained with anti-Lyt-1.2 mAb. Pan T cells are detectable in the meninges (top), in perivascular cuffs (large arrow), and in the parenchyma (small arrows). X200.

8 months

9

6

2

3

2-3

2

1

1

Clin score

Stable

Chronic progr

Relapses (X2)

Onset

Preclinical

Normal

Disease source

B LC

B LC B LC

B LC

B LC

B LC

Tissue

k -+

It +

k

f

f

k -

* -

+ +

+

+

+

+

++

++

tt

++

++

+

++

++

tt

++

+

f -

++ ++ +

+ +

++ +

-

-

++

-

-

-

-

-

f

-

+

-

k

-

-

-

++

++

+

+

-

-

par

f -

+

++

+

+

+

-

pv

par

pv

Lyt-2

par

L3T4

pv

Lyt-1.2

+

+

+

x!Y

-

+

++

++

++

+

pv

+

+

+

f

f

+

++

++

++

++

par

Iaf cells

f

+ f tt +

tt

+

++

++

++t

iI

+

+

+

+

+

5 -

* -

+ -

dz -

f -

-

-

+

-

Endo

MBP

f +

+

t

+

+

+ +

+

t

-

-

GC

Endothelium

Astro

Ia antigen

Note. Immunopathologic analysis of CNS changes in chronic relapsing EAE in the SJL mouse. B, brain, brainstem, and cerebellum; LC, lumbar cord; pv, perivascular and meningeal; par, parenchymal; Astro, astrocytes; Endo, endothelial cells; MBP, myelin basic protein; CC, galactocerebroside. Rating score for labeled cells: (-) none, (k) 95.3 + 48.6, (+) 451 + 275.9, (++) 1396.3 + 370.6 labeled cells per brain section.

3- 15 months

2

3 months

Active chronic

Inactive chronic

2

12 to 21 days

Onset

1

7 days

Latent

2

0

Time PI

Normal

Diseasestage

No. of animals

T cells

Synopsis of the Immunopathology of Mouse Chronic Relapsing EAE

TABLE 1

IMMUNOPATHOLOGY

OF CHRONIC EAE

401

FIG. 3. Cervical spinal cord, sampled 7 days PI, stained with anti-mouse Ig. Meninges to left. Moderate numbers of Ig+ mononuclear cells are seenin the parenchyma and some astrocytesare also labeled (arrows). x300.

Latent period. At Day 7 PI (the earliest timepoint investigated), considerable numbers of infiltrating T cells could be detected in meningeal, perivascular, and parenchymal areas in forebrain, brainstem, and cervical cord (Table 1 and Fig. 2). In more caudal regions, T cells were less numerous and were mainly found in the meninges. All cell types studied (Lyt- 1.2+, L3T4+, Lyt-2+, Ia+, and Ig+ cells) were similarly distributed (Fig. 3). In addition to its presence on hematogenous cells, Ia antigen was also demonstrable on some endothelial cells within the brain and a few astrocytes in the spinal cord (Table 1 and Fig. 4). Albumin and Ig deposits were most pronounced in the forebrain and brainstem and were less intense in the spinal cord (Fig. 4). MBP was found on a few endothelial cells in the brain only, whereas GC was present throughout the neuraxis. First episode of signs. In mice displaying the first signs of EAE ( 12-2 1 days PI), the density of infiltrating cells was somewhat higher than at Day 7 PI and all cell types were distributed evenly throughout lesions in the brain and spinal cord (Table 1 and Fig. 4). While Lyt-1.2+ and Lyt-2+ cells were found in both the meninges and CNS parenchyma, L3T4+ cells were occasionally more numerous in the meninges (Figs. 5-7). Ia+ and Ig+ cells were also present throughout the neuraxis and displayed a slightly higher density in the lower spinal cord (Table 1). Ia was found on some endothelial cells and astrocytes throughout the CNS and Ia+ astrocytes were sometimes detectable at the edge of lesions (Fig. 8). Deposits of albumin and Ig were more pronounced. The distribution of GC and MBP on endothelium was comparable to that seenduring the latent period. With time PI, the immunopathologic changesextended from the brain caudally to the brainstem, the cervical, and lumbar spinal cord. Thoracic levels were commonly spared initially.

402

TRAUGGTT,

MC FARLIN, AND RAINE

NORMAL WM

NORMAL WM

*

: W

NORMAL WM

FIG. 4. Schematicsummary of immunopathologic changesin chronic relapsing EAE. Key: 0 = mononuclear cells, which represent from top to bottom: Lyt-1.2+ cells, L3T4+ cells, Lyt-2+ cells, Ia+ cells, Ig+ cells; O-000 equals +-++. v = vessel, MBP = myeliti basic protein; GC = galactocerebroside; shaded areas surrounding vessels= albumin and IgG deposits,width reflectsintensity of deposits.t = Ia-negative astrccytes; * = Ia-positive astrocytes.Upper panel: Acute EAE lesion. All types of inflammatory cells are evenly distributed throughout the lesion. Ia is detectable on endothelial cells but not on astrocytes; myelin antigens are demonstrable on some endothelial cells. Extensive deposits of albumin and IgG are seen.Middlepanel: Active chronic EAE lesion. L3T4+ cells predominate within the lesion, while Lyt-2+ cells are virtually absent. Ia is present on endothelial cells and astrocytes,the latter being detectable in greatest numbers at the lesion edge. Myelin antigens are found on endothelial cells. Moderate degree of perivascular albumin and IgG deposits occurs. Lower panel: Silent chronic EAE. Low numbers of all types of inflammatory cells are detectable in old lesions and normal-appearing white matter. Ia is found on some endothelial cells but not on astrocytes. Myelin antigens are present on a few endothelial cells. Low-grade albumin and IgG deposits occur. In CNS tissue sampled during both the acute and chronic stagesof EAE, low numbers of all types of inflammatory cells were found within normal-appearing white matter.

Relapsing and chronic progressive disease. Mice sampled between 3 and 15 months PI, (Table 1 and Fig. 4) after a relapsing-remitting or chronic progressivediseasecourse, were mildly (grade 2) to moderately (grade 3) impaired neurologically at the time of sacrifice. Low numbers of all types of inflammatory cells were consistently detected at all levels of the CNS in both meningeal and parenchymal areas. Superimposed upon this low-grade diffuse inflammation, a focal accumulation of infiltrating cells was occasionally detectable in relapsing disease. Lyt-1.2+, L3T4+, Ia+, and Ig’ cells were found predominantly within the CNS parenchyma (Fig. 9), while Lyt-2+ cells, present in lower numbers, were more common in meningeal and perivascular areas

IMMUNOPATHOLOGY

OF CHRONIC EAE

403

FIG. 5. Cerebellum, sampled 12 days PI, stained with anti-Lyt- 1.2 mAb. Dense T-cell infiltrates are demonstrated in a perivascular cuff (left) and in the white matter parenchyma. X300.

FIG.6. Lumbar spinal cord, sampled 12days PI, stainedwith anti-L3T4 mAb. L3T4+ cellsare predominantly located within the CNS parenchyma and only a few are found in perivascular areas (upper left comer). X480.

404

TRAUGOTT,

MC FARLIN, AND RAINE

FIG. 7. Brain, sampled 17days PI, stainedwith anti-Lyt-2 mAb. Lyt-2+ cells are numerous in the perivascular cuff (right), but are also seen in the parenchyma. X200.

FIG. 8. Lumbar spinal cord, sampled 12 days PI, stained with anti-Ia mAb. Ia+ mononuclear cells are shown in a lesion (meninges on top). At the edge of the lesion, a few Ia+ astrocytes can be seen (arrows). X480.

IMMUNOPATHOLOGY

OF CHRONIC EAE

405

FIG. 9. Lumbar spinal cord, sampled after chronic progressive diseaseat 8 months PI, stained with antiL3T4 mAb. L3T4+ cells are seen in the fibrous meninges (top) and throughout a subpial zone of white matter. X480.

FIG. 10. Lumbar spinal cord, sampled after chronic progressivediseaseat 8 months PI, stained with antiLyt-2 mAb. Lyt-2+ cells are predominantly present in the meninges. X300.

406

TRAUGOTT,

MC FARLIN, AND RAINE

(Fig. 10). In chronic progressivedisease,L3T4+ cells were present in moderate numbers and were more widespread within the parenchyma of brain and lumbar spinal cord, while Lyt-2+ cells were virtually absent. Ia+ and Ig+ cells were detectable in moderate densities in both parenchyma and perivascular cuffs. Ia antigen was found on a few endothelial cells throughout the CNS (Fig. 1l), and Ia+ astrocytes were occasionally present. GC was demonstrable more frequently on endothelial cells than MBP (Fig. 12). Albumin and Ig deposits continued to show some predilection for rostra1 areas but decreasedin intensity with time PI. Remissionand clinically silent disease.CNS tissue from an animal which had failed to develop clinical signsby Day 14 PI and from others which had recovered completely after the initial episode and had remained healthy thereafter, displayed a low degree of infiltration by hematogenouscells throughout the CNS with thoracic levels frequently showing the least involvement. Ia, MBP, and GC were found on a few endothelial cells only and albumin and Ig deposits were scarce. Normal mouseCNS tissue.In CNS tissue from normal SJL/J mice, no T cells, Tcell subsets, or Ia+ cells could be detected within the parenchyma. Endothelial cells and astrocytes did not react with anti-Ia mAb or anti-MBP and anti-GC antisera (Table 1, Fig. 13). DISCUSSION In the present study, a selection of immunologic parameters was investigated in situ within the CNS of SJL mice during various stagesof chronic relapsing EAE using immunocytochemical techniques. While hematogeneous cells were not detectable in CNS tissue from normal SJL mice, they were consistently found in tissue from mice

FIG. 11. Brain, sampled at 8 months PI after chronic progressive disease, stained with anti-Ia mAb. Discontinuous staining of Ia antigen is seen on some endothelial cells in gray matter. X300.

IMMUNOPATHOLOGY

OF CHRONIC EAE

407

FIG. 12.Brain, sampled at 8 months PI after chronic progressivedisease,stainedwith anti-galactocerebroside (GC) antiserum. GC is demonstrated on the luminal surface of endothelial cells. X300.

with chronic EAE, though in varying numbers. During the latent period and the first episode of clinical signs, all types of inflammatory cells studied (Lyt-1.2+, L3T4+, Lyt-2+, Ia+, Ig+) were randomly distributed throughout lesions and with onset of signs the density of infiltrates increased further. CNS tissue sampled during silent chronic stagesdisplayed low-grade diffuse inflammation. During active chronic disease,some focal accumulation of infiltrating cells in which L3T4+ cells were more common than Lyt-2+ cells was superimposed upon a mild, diffuse CNS involvement. Ia was demonstrable more frequently on endothelial cells than on astrocytesand myelin antigens (MBP, GC) were detectable on endothelium. The degree of damage to the bloodbrain barrier, a prominent feature during the acute phase, was lesspronounced during chronic stages. In the present study, two types of CNS lesions were distinguishable by immunopathologic criteria. During the latent period and the first episode of clinical signs, lesions contained diffuse, unstructured infiltrates in which all types of hematogenous cells were intermixed. This finding is reminiscent of that described previously for acute EAE lesions in the mouse (10) and acute MS lesions (13, 14). Lesions found during chronic stagesof the disease,presented with a distinct distribution pattern of inflammatory cells in which L3T4+ cells predominated within the parenchyma while Lyt-2+ cells were more common in perivascular and meningeal areas,The latter picture is similar to that described previously for active chronic MS lesions (13, 14, 19, 20). A similar difference in the distribution pattern of inflammatory cells (T cells, macrophages), to that described here for BAE lesions, has been documented previously for leprosy lesions (2 1). In tuberculoid leprosy lesions, well organized infiltrates consisting of T cells and macrophages were considered to reflect the elimination of the

408

TRAUGOTT,

MC FARLIN, AND RAINE

FIG. 13. Normal brain, stained with anti-Ia mAb. There is no labeling of cellular components. Arrows indicate vessel. X300.

causative agent (M. leprue) by the immune system. The diffuse mixture of all types of infiltrating cells in lepromatous lesions was thought to be associated with antigen persistence. Similarly, in chronic EAE and MS, the two distinct lesion types observed (acute and chronic) might reflect the more or lesssuccessfulelimination of the antigen(s) in question rather than basic differences in pathogenetic mechanism. In agreement with previous reports (lo), the density of cellular infiltrates (Lyt- 1.2+, L3T4+, Lyt-2+, Ia+, Ig+ cells) in the present study, correlated well with the clinical activity of the disease process. The observed latency period of 5 days between the appearanceof infiltrates (7 days PI) and onset of signs ( 12 days PI), observedpreviously also for acute EAE (lo), might suggestthat hematogenous cells per se do not cause clinical signs but initiate a cascadeof immunologic events leading to tissue damage and functional impairment. That during chronic stages,L3T4+ cells were more common than Ia+ or Ig+ cells and that L3T4+ cells were even more numerous and widespread in chronic progressive than in exacerbating-remitting disease(where they were found within the lesion only), underlines the importance of helper/inducer T cells in lesion pathogenesis. These findings are in accord with previous observations showing that lesion progression in active chronic MS appearedto be associatedwith the presence of T4+ cells (14, 19). The observed moderate numbers of Ia+ and Ig+ macrophages in chronic EAE lesions in the mouse contrast with active chronic MS lesions in which Ia+ macrophagesare abundant (14, 19). Together with the previously reported presence of polymorphonuclear leukocytes in actively induced chronic EAE lesions (4) the low density of Ia+ macrophages might suggestthat different mechanisms are operative in mouse EAE and MS. A diffuse infiltration of the CNS by low numbers of hematogenous cells, similar to that described previously in chronic MS (14, 19), was also detectable during later

IMMLJNOPATHOLGGY

OF CHRONIC EAE

409

stagesin mouse chronic EAE. This continuous low-grade inflammation might indicate that in EAE, and perhaps also in MS, the diseaseprocesspersistsfor prolonged periods, albeit at low level. Various hitherto unknown factors such as polyclonal stimulation of the immune system might reactivate the diseaseprocessleading to new lesion formation and clinical worsening. The change in permeability of the blood-brain barrier with time as indicated by the decreaseof Ig and albumin deposits paralleled the intensity of cellular infiltrates. This damage to the blood-brain barrier might have been a consequence of local presentation of specific antigen (MBP, GC) on endothelial cells within the CNS to circulating T cells and the subsequent releaseof vasoactive amines, as has been suggested previously ( 10, 11,22,23). The detectability of Ia and myelin antigen(s) on endothelium in CNS tissue from mice with chronic relapsing EAE might constitute basic requirements for local antigen presentation. The minor variability in Ia and myelin antigen expression by endothelial cells in both white and gray matter during active and inactive stages of this white matter diseasemight suggestthat a variety of factors are required for new lesion formation and reactivation of the diseaseprocess.While antigen presentation on endothelial cells might be essential for initiating the diseaseprocess,astrocytes, which have been demonstrated to present MBP to specifically sensitized T cells in tissue culture (24), might play a more important role as antigen-presenting cells within established lesions and might be involved in lesion progression from the periphery. Similar mechanisms have been suggestedin MS (25). While the above data have analyzed certain cell-mediated phenomena in acute and chronic lesions in mouse EAE, more detailed information on early events, as well as the demonstration of soluble mediators of the immune system will be required for a fuller insight into the immunopathogenesis of this demyelinating diseaseprocess.Such knowledge might expedite the search for a successful approach for therapeutic immunomodulation applicable to MS. ACKNOWLEDGMENTS The authors thank Dr. Ethan Shevach (NIAID) for supplying some rat clones; Patricia Kennedy, Robyn Balf and Howard Finch for technical assistance,and Michele Griebel for typing the manuscript. Supported in part by National Multiple SclerosisSociety Grants RG 1664-B-1 and RG 1004-E-5;and NIH Grants NS 11920 and NS 08952.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

Gonatas, N., and Howard, J. C., Science 186, 839, 1974. Ortiz-Ortiz, L., Nakamura, R. M., and Weigle, W. O., J. Immunol. 117,576, 1976. Pettinelli, C. B., and McFarlin, D. E., J. Immunol. 127, 1420, 1981. Brown, A., McFarlin, D. E., and Raine, C. S., Lab. Invest. 46, 171, 1982. Lublin, F. D., J. Neural. Sci. 57, 105, 1982. Raine, C. S., Bamett, L. B., Brown, A., Behar, T., and McFarlin, D. E., Lab. Invest. 43, 150, 1980. Linthicum, D. S., Munoz, J. J., and Blaskett, A.,Cell Immunol. 73, 299, 1982. Ledbetter, J. A., Evans, R. L., Lipinski, M., Cunningham-Rundles, C., Good. R. A., and Herzenberg, L. A., J. Exp. Med. 153, 310, 1981. 9. Sriram, S., Soloman, D., Rouse, R. V., and Steinman, L., J. Immunol. 129, 1649, 1982. 10. Traugott, U., Raine, C. S., and McFarlin, D. E., Cell Immunol. 91, 240, 1985. 11. Traugott, U., Springer Semin. Immunopathol. 8, 7 1, 1984. 12. Raine, C. S., and Stone, S. H., NY State J. Med. 77( 1 I), 1693. 1977. 13. Traugott, U., Reinherz, E. L., and Raine, C. S., J. Neuroimmunol. 4,201, 1983.

410

TRAUGGTT,

MC FARLIN, AND RAINE

14. Traugott, U., and Raine, C. S., In “Multiple Sclerosis:Experimental and Clinical Aspects,” Ann. NY Acad. Sci. 436, 163, 1984. 15. Hsu, S., and Raine, L., J. Histochem. Cytochem. 29, 1349, 1981. 16. Guesdon, J., Temynck, T., and Avrameas, S., J. Histochem. Cytochem. 27, 1131, 1979. 17. Raine, C. S., Johnson, A. B., Marcus, D., Suzuki, A., and Bomstein, M. B., J. Neural. Sci. 52, 117, 1981. 18. Oi, V. T., and Herxenber& L. A., In “Selected Methods in Cellular Immunology,” 368, 1980. 19. Traugott, U., Reinherz, E. L., and Raine, C. S., Science 129, 308, 1983. 20. Traugott, U., Reinhetz, E. L., and Raine, C. S., J. Neuroimmunol. 3,365, 1982. 21. Modlin, R. L., Hofman, F., Meyer, P. R., Sharma, 0. P., Taylor, C. R., and Rea, T. H., C/in. Exp. Zmmunol. 51,430, 1983. 22. Burger, D. R., Ford, D., Vetto, R. M., Hamblin, A., Goldstein, A., Hubbard, M., and Dumonde, D. C., Human Zmmunol. 3,209, 1981. 23. Burger, D. R., and Vetto, R. M., Cell. Zmmunol. 70, 357, 1981. 24, Fontana, A,, Fien, W., and Werkerle, H., Nature (London) 307, 273, 1984. 25. Traugott, U., Scheinberg, L. C., and Raine, C. S., J. Neuroimmunol. 8, 1, 1985.