Download - Exogenous antigen containing perivascular phagocytes induce a non-encephalitogenic extravasation of primed lymphocytes

Ž .Journal of Neuroimmunology 117 2001 30–42www.elsevier.comrlocaterjneuroin

Exogenous antigen containing perivascular phagocytes induce anon-encephalitogenic extravasation of primed lymphocytes

Michael Walther a, Anastas Popratiloff b, Nina Lachnit a, Nils Hofmann a, Michael Streppel b,Orlando Guntinas-Lichius b, Wolfram F. Neiss a, Doychin N. Angelov a,)

a Institut I fur Anatomie der UniÕersitat zu Koln, Joseph-Stelzmann-Strasse 9, D-50931 Cologne, Germany¨ ¨ ¨b Klinik fur Hals-, Nasen- und Ohrenheilkunde der UniÕersitat zu Koln, D-50931 Cologne, Germany¨ ¨ ¨

Received 20 October 2000; received in revised form 28 December 2000; received in revised form 9 March 2001; accepted 14 March 2001

Abstract

Recent evidence suggests that T-lymphocyte extravasation and CNS-parenchymal infiltration during autoimmune disease might beŽ q. Ž .regulated by antigen-presenting ED2 cerebralrspinal perivascular phagocytes CPPrSPP . Since the massive erythrocytic and

leukocytic infiltrates in the CNS of rats with experimental allergic encephalomyelitis do not allow a precise differentiation betweenCPPrSPP and the invading cells in the Virchow–Robin space, we developed a new immune-response model whereby the extravasation

Ž .of T-lymphocytes was not followed by other blood cells. Adult Lewis rats were sensitized to horseradish peroxidase HRP . SubsequentŽ . Ž . Ž . Ž .intracerebroventricular i.c.v. injections of HRP andror Fluoro-Emerald FE served to: 1 challenge the primed T-lymphocytes and 2

label the CPPrSPP for additional immunocytochemical analysis. We found that 24 h and 3 days after single, double, or triple antigenŽ q q q.boosting T-lymphocytes R73 , W3r25 , OX50 entered the Virchow–Robin space but did not break through the astrocytic glia

Ž q q q q q.limitans. Instead they adhered to HRP-containing activated CPPrSPP mabs OX-6 , SILK6 , CD40 , CD80 , CD86 . This selectiveŽ . Žcontact was mediated neither by cell adhesion molecules P-selectin, ICAM-1, VCAM-1 , nor promoted by chemokine receptors CCR1,

. Ž Ž . .CCR5 or chemokines monocyte chemoattractant protein MCP -1, MIP-1a , MIP-1b, RANTES . This non-inflammatory, but antigen-dependent lymphocyte extravasation provides optimal conditions to further study the CNS immune response. q 2001 Elsevier ScienceB.V. All rights reserved.

Keywords: Rat; Antigen-presenting cells; Lymphocytes; Chemokines; Cell adhesion molecules

1. Introduction

The extravasation of lymphocytes into the CNS intercel-Ž .lular perivascular Virchow–Robin space and the subse-

quent leukocytic infiltration of the brain-parenchyma arekey steps in the pathogenesis of any autoimmune neurolog-

Žical disease Lassmann et al., 1986; Kawamata et al., 1992;Engelhardt et al., 1993; Schluesener et al., 1996; Raivich

.et al., 1998 . Despite voluminous work on the molecularbiology of the secondary immune response in the CNS, thegeneral mechanisms regulating these processes are still amatter of controversy.

According to a wide-spread theory, local neural degen-eration promotes the secretion of proinflammatory

Žmolecules histamine, intrerleukin 1, tumor necrosis factor

) Corresponding author. Tel.: q49-2214785654; fax: q49-2214786711.Ž .E-mail address: [email protected] D.N. Angelov .

.a , complement, prostaglandins, leukotrienes, chemokineswhich in turn contract the smooth vascular muscles and

Ž .render the blood–brain barrier BBB permeable to circu-Žlating B- and T-lymphocytes Knopf et al., 1998; Serpe et

.al., 1999 .Another possibility is that certain antigens may have

remained AhiddenB for the immune system behind anintact BBB for prolonged periods of time and are able toevoke an inflammatory immune response much later after

Žperipheral sensitization Matyszak and Perry, 1995, 1996,.1998; Perry et al., 1995 . Accumulating evidence shows

that, in this case, the extravasation of lymphocytes may notbe promoted by proinflammatory cytokines, but can also

Žbe initiated by resident cells Cross et al., 1990, 1993;Gehrmann et al., 1993; Morimoto et al., 1996; Hu et al.,

. Ž1998; Deiniger et al., 1999 or B-lymphocytes Litzen-.burger et al., 1998 . Current knowledge shows that these

Žare the cerebral antigen-presenting cells APC; Matsumoto

0165-5728r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0165-5728 01 00302-2

( )M. Walther et al.rJournal of Neuroimmunology 117 2001 30–42 31

et al., 1986; Gautam and Glynn, 1989; Perry, 1998; Arche-.los and Hartung, 2000 .

Since it is well-known that, under the conditions of astructurally intact BBB, the cerebral perivascular phago-

Ž . Žcytes CPP are the APC of the brain Hickey and Kimura,1988; Graeber et al., 1989, 1992; Streit et al., 1989; Mato

.et al., 1996 , we decided to study their role in lymphocyteextravasation during experimental allergic encephalomyeli-

Ž .tis EAE . However, numerous and densely packed cells inthe Virchow–Robin space rendered the differentiationbetween CPPrSPP and extravasating cells very dif-ficult. This is why we developed a Acleaned upB modelof an immune response, whereby the extravasation ofT-lymphocytes was not followed by erythrocytes androrother leukocytes.

Employing this model, we found that primed T-lymphocytes crossed the cerebral endothelium exclusivelyat sites where the Virchow–Robin space contained CPP.This T-lymphocyte extravasation in the cerebral perivascu-lar space was independent of molecules promoting rolling,activation, and adherence. We conclude that, indicating thesites for lymphocytic extravasation, CPP play a major rolein the initiation of local inflammatory processes.

2. Materials and methods

2.1. Experimental animal groups

Ž .Forty-four male Lewis strain LEWrHan Rij Hsd ,distributed in seven experimental groups, were used. Be-fore and after experiments, all rats were kept on standard

Ž .laboratory food Ssniff, D-59494 Soest, Germany and tapwater ad libitum with an artificial light–dark cycle of 12 hlight on, 12 h light off. All experiments were conducted inaccordance with the German law on Animal Protection, allprocedures used were approved by The Local Animal-Pro-

Žtection Committee Bezirksregierung Koln, Az. 23.203.2-K¨.35, 33r98 vom 21.12.1998 .

Ž .The animals of group 1 eight rats were used tooptimize the labeling of cerebral perivascular phagocytesŽ . Ž .CPP and spinal perivascular phagocytes SPP using

Ž .intracerebroventricular i.c.v. injection of horseradish per-Ž .oxidase HRP; four rats and of fluorochrome-conjugatedŽ .dextrans four rats . Longitudinal sections through the

lumbar spinal cord were used to quantitatively estimate thedensity of SPP per mm2.

Each of the groups 2–4 consisted of eight rats whichwere sensitized to HRP and fixed by perfusion after oneŽ . Ž . Ž .group 2 , two group 3 , or three group 4 i.c.v. chal-

Ž . Ž .lenges AboostingsB with the antigen HRP . Thereby fourrats in each group survived for 24 h, and the other fourrats, for 3 days after i.c.v. boosting with HRP. Theseanimals served to study the immune response followingsingle and multiple i.c.v. challenge with the antigen. Lon-

gitudinal sections of the spinal cord were used to estimatethe SPP-density.

Ž .The rats of group 5 eight animals were sensitized forŽ .HRP but received a mixture of HRP antigen boosting

Žplus Fluoro-Emerald to visualize CPPrSPP in green fluo-.rescence together in one i.c.v. injection. These rats were

used to immunocytochemically demonstrate the putativeŽ .molecules visualized in red fluorescence promoting

CPPrSPP activation and lymphocyte extravasation.Ž .The rats of group 6 four animals served as controls to

the primed animals in groups 2–5. They were not sensi-tized for HRP and received subcutaneous injections ofFreund’s complete adjuvant only.

2.2. Labeling of CPP and SPP employing intracerebroÕen-( )tricular i.c.Õ. injections

( )2.2.1. Horseradish peroxidase HRPAfter an intraperitoneal injection of Ketamin plus Xy-

Ž w wlazin 100 mg Ketanest plus 5 mg Rompun per kilo-.gram body weight rats were fixed in a stereotactic appara-

Ž . Žtus and 35 ml of a 6.7% wrv solution of HRP Type VI.A, Sigma, No. P 6782 in 0.9% saline were injected into

the right lateral cerebral ventricle over a period of 5 minŽ . Ž .Wagner et al., 1974 using a very thin 10 mm thick glass

Ž .pipette to minimize tissue damage Mingheti et al., 1999 .

( )2.2.2. Fluorescein-conjugated dextran Fluoro-EmeraldUnder narcosis, the rats were fixed in a stereotactic

apparatus and received 35 ml of a 5% solution of Fluoro-Ž .Emerald FE; Molecular Probes, Cat. No. D-1820 in

distilled water into the right lateral cerebral ventricle.

2.2.3. Mixture of HRP- and FE-containing solutionsUnder narcosis, HRP-primed rats were fixed in a stereo-

tactic apparatus and received 35 ml of a solution contain-Žing 6.7% HRP employed as boosting antigen to recruit

. Žcirculating lymphocytes and 5% FE to visualize CPP in.green fluorescence into the right lateral cerebral ventricle.

( )2.3. Sensitizing priming of rats to HRP

Under KetaminrXylazin narcosis rats received a subcu-Ž .taneous injection in the hind footpad of 2 mg HRP

Ž .Sigma, No. P 6782 in 200 ml 0.1 M phosphate buffer pHŽ .7.4 emulsified 50:50 in Freund’s complete adjuvant

Ž .FCA; Sigma, F5881 . Booster injection of the antigenŽ .HRP in the lateral cerebral ventricle followed 30 days

Ž .later Baloyannis and Gonatas, 1979 .

2.4. Single or multiple boosting with HRP

This procedure was identical with that of labeling ofŽ .CPP and SPP after i.c.v. injection of HRP see above .

Each subsequent boosting followed in intervals of 30 days.

( )M. Walther et al.rJournal of Neuroimmunology 117 2001 30–4232

Animals were fixed by perfusion 24 h or 3 days after thelast i.c.v. injection.

2.5. Tissue processing

2.5.1. FixationAll rats were anesthetized with ether, their vascular

system rinsed for 60 s with 0.9% NaCl saline and fixed forat least 40 min by transcardial perfusion with 1.0 l of

Ž . Žperiodate–lysine–paraformaldehyde PLP fixative Mc-.Lean and Nakane, 1974 . The PLP fixative was mixed

from premade stem stock solutions immediately prior toŽperfusion. The CNS cerebrum, cerebellum, brainstem,

. Ž .spinal cord were removed and stored in a 4% wrvparaformaldehyde in 0.1 M phosphate buffer, pH 7.4.

2.5.2. Visualizing of HRP-labeled CPP and SPPŽ .Vibratome sections 50 mm thick of the brain were

Ž . Xincubated for 20 min in 0.75% wrv 3,3 -diaminobenzi-Ždine tetrahydrochloride 150 mg in 200 ml buffer; DAB,

. Ž .Sigma, D 5637 plus 0.01% vrv H O plus 0.075%2 2Ž .wrv nickel chloride in 0.05 M Tris–HCl buffer, pH 7.6.This procedure yielded a very intensive dark-purple toblack reaction product exclusively in the cytoplasm of the

ŽCPP. Control tissue sections from brains which received.no HRP injections contained no reaction product.

For electron microscopy some of these sections wereIII Ž .postfixed in 1% OsO q1.5% K Fe CN , dehydrated in4 3 6

graded acetones, and embedded flat in Araldite CY212Ž .Fluka, No. 44610 .

2.5.3. Density of lumbar SPPThe lumbar ventral horn consists of 12–15 longitudinal

Ž .vibratome sections 50 mm thick . Counts of SPP andmeasurements of the areas in mm2 were performed atmagnification =10 in three non-overlapping fields in ev-

Ž .ery second section i.e. at least 18 fields per animal withthe Optimas 6.5. System. A one-way ANOVA test was

Žused to compare the SPP-density of control animals group.1 to those obtained from HRP-sensitized and boosted ratsŽ .groups 2–4 .

2.5.4. Visualizing of Fluoro-Emerald-labeled CPP andSPP

Ž .Vibratome sections 50 mm thick of the brainŽwere observed through filter set 10 excitation BP 450–

.490, beamsplitter FT 510, emission BP 515–565 or byŽthe combined filter set 25 excitation TBP400r495r

570, beamsplitter FT410r505r585, emission TBP460r.530r610 of a Carl Zeiss fluorescence microscope.

2.6. Immunocytochemistry

2.6.1. Reagents to differentiate CPPrSPP from astrocytesand from microglia

Ž . Ž1 Polyclonal anti-GFAP from guinea pig Progen,. Ž .GP52, 1:3000 was used to stain astrocytes; 2 Isolectin

Ž .IB from Griffonia simplicifolia GSA I-B was em-4 4Ž .ployed to stain microglia Streit, 1990 using 20 mgrml

Ž .biotinylated GSA I-B Sigma, L-2140 . The reaction-4

product detection was achieved employing FluoroLinkeŽCy3e—Streptavidin Amersham Life Science, Cat. No.

. Ž . ŽPA 43001 . 3 Mouse monoclonal ED2 Serotec, MCA.342, 1:250 , directed against rat macrophage membrane

Žantigen was used to label CPPrSPP Dijkstra et al., 1985;.Graeber et al., 1989 .

2.6.2. Primary antibodies to estimate the actiÕation degreeof CPPrSPP

Ž . Ž .1 Mouse monoclonal OX-6 Serotec, MCA 46, 1:1000Ž . Žrecognizing rat MHC class II Ia antigen McMaster and

. Ž . ŽWilliams, 1979 . 2 Mouse monoclonal SILK 6 Serotec,.MCA-1397, 1:500 recognizing recombinant rat IL-1b

Ž . Ž .Angelov et al., 1998b . 3 Armenian hamster monoclonalŽ .anti-mouse CD40 BD PharMingen, 09401A, 1:100 to test

for presence of the costimulatory molecule CD40 onŽ . ŽCPPrSPP. 4 Mouse monoclonal CD80 BD PharMingen,

.22661D, 1:100 to test for presence of the co-stimulatoryŽ .molecule B7-1 on CPPrSPP. 5 Mouse monoclonal CD86

Ž .BD PharMingen, 22671D, 1:100 to test for presence ofthe co-stimulatory molecule B7-2 on CPPrSPP.

2.6.3. Primary antibodies to establish the nature of theadhering blood cells

As already briefly mentioned in the Introduction, someblood cells approached and contacted the antigen-contain-ing CPPrSPP after HRP-priming and boosting. To estab-lish the immunological pertinence of these blood cells, weperformed a detailed immunocytochemical analysis using:Ž . Ž1 Monoclonal mouse anti-rat TCR alpharbeta clone

.R73, Serotec MCA 453G, 1:20 recognizing a constantŽ .determinant of the arb T cell receptor. 2 Monoclonal

Žmouse anti-rat CD4 clone number W3r25, Serotec.MCA55G, 1:50 recognizing the CD4 cell surface glyco-

Ž .protein expressed by helper T cells and thymocytes. 3ŽMonoclonal mouse anti-rat CD44 clone OX-50, Serotec

.MCA643XZ, 1:50 recognizing the CD44 cell surfaceantigen expressed by T cells, B cells, macrophages, andthymocytes.

2.6.4. Primary antibodies to elucidate the lymphocyte at-tracting potential of CPP

Ž . Ž1 Monoclonal mouse anti-human CD162 clone PL2,.Serotec, MCA 1727, 1:50 recognizing a human CD162

cell surface glycoprotein, known as P-Selectin Glycopro-Ž .tein Ligand PSGL-1 . CD162 is the major counter recep-

tor for P-selectin and is involved in neutrophil attachmentŽ .and rolling to endothelium. 2 Polyclonal goat anti-rat

Ž .ICAM-1 M-19; Santa Cruz Biotechnology, sc-1511, 1:50Ž .recognizing rat intercellular adhesion molecule-1. 3 Poly-

Žclonal goat anti-rat VCAM-1 C-19; Santa Cruz Biotech-.nology, sc-1504, 1:50 recognizing rat vascular cell adhe-

Ž .sion molecule-1. 4 Monoclonal mouse anti-rat MCP-1Ž .clone C4, Pharmingen, No. 24011D, 1:50 recognizing the


Ž .CC-chemokine Amonocyte chemoattractant protein MCP -1B which promotes the directed migration of inflammatory

Ž . Žcells. 5 Polyclonal rabbit anti-rat MCP-1 Cedarlane, No.. Ž .CL9576AP, 1:500 . 6 Monoclonal mouse anti-human

ŽCCR-1 clone 53504.111, R and D Systems, MAB145,. Ž .1:50 . 7 Monoclonal mouse anti-human CCR-5 purified

Ž .clone 2D7rCCR5 Pharmingen, No. 36460D, dilution.1:50 recognizing the chemokine receptor CCR5 which

regulates lymphocyte chemotaxis activation and tran-Ž .sendothelial migration during inflammation. 8 Polyclonal

Ž .goat anti-human MIP-1a RBI, No. M-256, 1:20 recog-nizing the chemokine Amacrophage inflammatory protein-

Ž . Ž1a B. 9 Polyclonal goat anti-human MIP-1b RBI, No..M-257, 1:20 recognizing the chemokine Amacrophage

Ž .inflammatory protein-1bB. 10 Polyclonal goat anti-hu-Žman RANTES neutralizing, affinity purified RBI, No.

. Ž .R-125, 1:20 recognizing the chemokine RANTES. 11ŽPolyclonal rabbit anti-rat RANTES Cedarlane, No. CL

.9578AP, 1:30,000 .The antibodies recognizing human CD162, CCR-1,

CCR-5, and RANTES showed a constant cross reactivitywith the corresponding antigens in the rat brain. However,the antibodies recognizing MIP-1a and MIP-1b showedno cross reactivity with rat tissue: we failed to immunos-tain any structures in the rat brain employing the antibod-ies recognizing these human chemokines. This is why werepeated the experiments with the only recently availableŽ . Ž12 rabbit anti-rat MIP-1a Cedarlane, No. CL9577AP,

.1:2000 .

2.6.5. Secondary antibodiesŽ . Ž1 Biotinylated goat anti-rabbit IgG DAKO, No.

. Ž . ŽE0432 . 2 Biotinylated goat anti-mouse IgG Fab-specific,. Ž . ŽSigma, B-0529 . 3 Biotinylated goat anti-mouse IgG Fc

specific, pre-absorbed with human IgG and rat serum. Ž .proteins, Sigma, No. B-9904 . 4 Biotinylated rabbit anti-

Ž .goat IgG DAKO, No. E0466 . In order to minimize thebackground staining of neurons the biotinylated IgGs werealways pre-absorbed with the individual rat’s spleen pro-

Ž . Ž .tein Angelov et al., 1996 . 5 Biotin-conjugated mouseŽ .anti-hamster IgG cocktail; PharMingen, 12102D .

Red fluorescent staining of the biotinylated IgGs wasachieved by 1:100 diluted FluoroLinke Cy3e—Strep-

Ž .tavidin Amersham Life Science, Cat. No. PA 43001 .

2.6.6. Standard incubation protocolFree floating vibratome sections were immunostained

on a shaker at room temperature through the followingŽ . Ž . Žsteps: 1 0.6% vrv H O in buffer to block the en-2 2

. Ž . Ž .dogenous peroxidase activity for 30 min; 2 5.0% wrvŽ .bovine serum albumin BSA, Sigma, No. A-9647 in 0.1Ž . Ž .M Tris-buffered saline TBS pH 7.6 for 60 min; 3

Ž .primary antibody see above diluted in TBS plus 0.8%Ž . Ž . Ž .wrv BSA for 2 h; 4 5.0% vrv normal goat serumŽ . ŽNGS, Vector No. S-1000 or normal rabbit serum NRS,

.DAKO, No. X0902 plus 0.8% BSA in TBS for 15 min.

Ž .5 1:400 dilution of the corresponding biotinylated sec-Žondary antibody raised in goat, rabbit, or hamster see

. Ž .above in TBS for 1 h; 6 1:100 diluted Cy3-streptavidinŽ . Ž .for 1 h. Steps 1 and 5 were followed by two 10 min

Ž . Ž .washes in TBS. Steps 3 and 6 were followed by four10-min washes in TBS. Finally, sections were dehydrated

Ž .with ethanol and Histoclear a nontoxic xylene-substituteand coverslipped.

2.6.7. ControlsOmission of the primary or secondary antibody yielded

blank sections in which only the green CPPrSPP fluo-resced. Incubation of sections with non-relevant biotinyl-

Žated secondary antibodies e.g. goat anti-rabbit IgG for.recognition of mouse primary antibodies yielded also

blank sections.

2.6.8. Fluorescence microscopyAs already mentioned, all FE-labeled CPPrSPP were

visualized through filter set 10 of a Carl Zeiss fluorescenceŽ .microscope. The immunopositive CY3-positive cells were

Žobserved through Zeiss filter set 15 excitation BP546r12,.beamsplitter FT580, emission LP590 . Both FE-labeled

CPPrSPP and CY3-immunopositive cells were simultane-Žously observed employing Zeiss filter set 25 excitation

TBP400r495r570, beamsplitter FT410r505r585, emis-.sion TBP460r530r610 .

3. Results

( )3.1. Labeling of periÕascular phagocytes CPPrSPP inintact rats

There was no hazy staining indicative of free markersubstance in the CNS 24 h after injection of horseradish

Ž . Ž .peroxidase HRP or Fluoro-Emerald FE in the rightlateral ventricle. Labeled CPP, neurons and numerous small

Ž .round cells with a visible nucleus leukocytes were read-ily observed only in the restricted area around the injection

Ž .site parietal cortex and hippocampus . The rest of thebrain- and spinal-cord parenchyma were devoid of HRP

Žand only the perivascular phagocytes were labeled arrows.in Fig. 1A .

ŽThis specific and very reliable way of labeling cf..Zhang et al., 1992; Kida et al., 1993 facilitated the

differentiation between CPP and parenchymal microgliaŽ .Fig. 1B . No DAB-stained blood cells were ever observedin the brain parenchyma of perfusion-fixed animals. NoDAB-staining was observed in the brains of rats that didnot receive an i.c.v. injection of HRP.

Employing the Optimas 6.5. software, we found that 24h after an i.c.v. injection of HRP, the lumbar ventral hornof intact rats contained 31"5 HRP- or FE-labeled SPP

2 Ž .per mm mean"SD; ns8 rats .


Ž . Ž .Fig. 1. Identification of cerebral perivascular phagocytes CPP employing the method of intracerebroventricular i.c.v. injection of markers in theŽ . Ž .CSF-filled brain interstitium. A Low-power view of a coronal section through rat cerebral hemispheres showing the distribution of CPP arrows after

Ž .i.c.v. injection of 35 ml, 6.7% solution of HRP. A 50-mm vibratome section; =50. B Overall distribution of microglia in the same region of the rat brainŽ . Ž . Ž .as in A after GSA-IB staining Streit, 1990 . A 50-mm vibratome section; =50. C Electron micrograph showing HRP-labeled CPP in the vicinity of a4

Ž . Ž . Ž .pericyte P . The CPP is located between the vascular basal lamina small arrows and the membrana limitans gliae perivascularis arrowheads . TheŽ .reaction product is observed in the lysosomes arrows and as granular material in the cytoplasm =7000.

Observations of serial ultrathin sections with the trans-mission electron microscope revealed the typical localiza-tion of CPP between two basal laminae: the one lining thevascular endothelium and the other one covering the super-

Ž .ficial glia Fig. 1C .The HRP-labeling of CPPrSPP was not a long-lasting

one. Only single perivascular phagocytes were found la-beled 4 days after i.c.v. injection of HRP. Five days afterthe i.c.v. injection, there were only fading traces of themarker. For this reason, we restricted our observations upto 3 days after HRP-application.

3.2. The CNS periÕascular space after chronic challengeof the immune system

Rats were sensitized by a single subcutaneous injectionŽ .of HRP in complete Freund’s adjuvant CFA . Thirty days

later, we challenged the animals with an i.c.v. injection ofHRP, which triggered a secondary immune response. Inthis way, the exogenous protein HRP served two purposes:Ž . Ž .i as an established label for phagocytosis and as ii anantigen to be presented.

ŽSimilar to the observations in non-sensitized rats group.1 , the CPPrSPP phagocytized HRP and removed the

marker substance from the CNS within 24 h after its i.c.v.application.

No infiltrated blood cells were observed in the CNS-parenchyma at 24 h or 3 days after i.c.v. antigen challengeŽ . Žgroup 2 . Numerous leukocytes left the circulation ex-

.travasated at sites adjacent to the HRP-labeled CPP in theŽ .Virchow–Robin space Fig. 2A . Our counts showed that

24 h after this first boosting, the lumbar spinal cord2 Ž .contained 38"3 SPP per mm mean"SD; ns4 rats .

In the other subgroup, where the rats survived 3 days afterthe i.c.v. application of HRP, we counted 29"6 SPP per

2 Ž .mm mean"SD; ns4 rats . These values of SPP-den-sity were not significantly different from the value estab-

Ž .lished in non-sensitized animals 31"5; see above .Principally the same findings were observed after the

Ž .second i.c.v. application of HRP group 3 . The SPPŽ .density was 32"3 24 h survival after boosting and

Ž .28"7 3 days after boosting . A lot of leukocytes ex-travasated at sites adjacent to the HRP-labeled CPPrSPP.However, none of them were observed in the adjacent

Ž .brain parenchyma Fig. 2B .There were also numerous blood-borne cells which

extravasated and contacted the CPPrSPP at 24 h or 3 daysŽ .after the third i.c.v. application of HRP group 4 . No

cellular infiltrations in the surrounding parenchyma wereever observed. The electron microscopic analysis revealedthat the blood cells did not simply adhere to the vessel’swall, but penetrated through widened inter-endothelial


Ž . Ž . Ž .Fig. 2. Contacts between CPP and blood-borne cells in sensitized primed rats. A At least four contacts between HRP-containing CPP arrowheads andŽ . Ž .accumulated small round cells with scanty cytoplasm and large round nuclei arrows 24 h after i.c.v. application of the antigen HRP . A 50-mm

Ž . Ž .vibratome section; =400. B Close proximity between a HRP-labeled CPP and an extravasated cell lymphocyte; Ly in the Virchow–Robin space 3 daysŽ .after second i.c.v. application of HRP. Note that the surrounding brain parenchyma contains no other blood cells. A 1-mm-thick plastic section; =400. C

Ž . Ž . Ž .A blood cell BC penetrating through the cerebral endothelium Et and establishing contact arrow with HRP-labeled cerebral perivascular phagocyte.Ž . Ž . Ž .No counterstaining. =7000. D Electron micrograph showing a HRP-labeled CPP empty arrow next to an extravasated lymphocyte Ly , =12,000.

Ž .clefts Fig. 2C,D . The SPP density was 35"6 and 31"8Ž .after 24 h and 3 days survival, respectively . The fact thateven after three boostings HRP-primed cells still ap-proached and contacted the antigen-containing CPP showsthat the cell-contacting capacity of these cells did notdecrease. In contrast to the situation during EAE, theimmune systems obviously did not develop a tolerance to

Žsubsequent challenge by the same antigens Weller et al.,.1996 .

Despite the rather constant SPP-density, which remainsŽ 2lower than in animals with EAE 57"11 SPP per mm ;

.ns10 rats; own unpublished results in all three chal-lenged groups, the occurrence of a rapid recruitment ofED1q blood macrophages in the perivascular space cannot

Ž .be excluded Bauer et al., 1995 . This means, however,that cells must have also left: otherwise the Virchow–Robinspace would have been filled up within a very short time.One explanation is that SPP have undergone apoptosisupon challenge. However, apoptotic perivascular cells have

Ž .been observed neither in humans Kosel et al., 1997 nor¨Ž .in rats Bauer et al., 1995; Bechmann et al., 2000 . An

alternate explanation for a constant number of perivascularcells, despite a continuous influx, is a migration from theperivascular spaces through the glia limitans into the brainparenchyma or to the peripheral lymphoid organs. Indeed,a slow turnover of microglial cells has been observed in

Ž .rats Hickey et al., 1992; Lassmann et al., 1993 and mice

Ž .Kennedy and Abkowitz, 1997 , but it remains unclearwhether such cells undergo a transformation from bloodmonocytes via perivascular cells to resident microglia ordirectly transform from monocytesrmacrophages to mi-croglial cells. Alternatively, preliminary data indicate thatperivascular spaces harbour a population of OX-62-posi-

Ž .tive dendritic cells. Their migration into cervical lymphnodes upon sensitization accompanied by an equal influxof macrophages may account for stable total numbers of

Žcells in this compartment I. Bechmann, personal commu-.nication .

Ž .The constitutively ED2-positive CPPrSPP Fig. 3A,Cresponded with activation after each contact to HRP.Twenty-four hours after the i.c.v. injection, manyCPPrSPP upregulated the expression of MHC class II

Ž .glycoproteins Fig. 3D,F , started the synthesis of IL1-bŽ . ŽFig. 3G , and were immunoreactive for the CD40- Fig.. Ž . Ž .3J , B7-1- Fig. 3M , and B7-2 molecules Fig. 3P .

CPPrSPP of non-sensitized but HRP-injected rats did notŽ .synthetise IL1-b Fig. 3I and costimulatory molecules

Ž .Fig. 3L,O,R .The extravasated leukocytes could be stained by the

Žmab R73 recognizing a constant determinant of the rat T.cell receptor, arrows in Fig. 4B , by the mab W3r25

Žrecognizing CD4 cell surface glycoprotein expressed by T.helper cells and thymocytes, arrows in Fig. 4D , and by the

Žmab OX-50 recognizing rat CD44 cell surface antigen


Ž . Ž .Fig. 3. Comparison between activated CPPrSPP of HRP-sensitized rats left column with CPPrSPP in non-sensitized rats right column 24 h after ani.c.v. injection of HRP. The middle column depicts FE-labeled CPP and represents a control with omission of the primary antibody. Most CPP, which have

Ž . Ž . qbeen labeled after an i.c.v. application of HRPrFE fluorescent in B , are also recognized by the mab ED2 fluorescence by CY3 in A. The ED2 CPP inŽ . Ž .non-sensitized rats are shown in C. All CPP labeled by FE fluorescent in E are also OX6-positive fluorescent in D , i.e. they express immunocytochemi-

cally detectable amounts of MHC class II glycoproteins. Some MHC class IIq CPP in non-sensitized rats are shown in F. All CPP labeled by FEŽ . Ž . Ž .fluorescent in H are also immunoreactive for Interleukin 1b fluorescent in G . FE-labeled CPP in non-sensitized rats are immunonegative for SILK6 I .

Ž . Ž .Most of the FE-labeled CPP fluorescent in K, N, Q express also the CD40-, B7-1-, and B7-2 molecules fluorescence in J, M, P . FE-labeled CPP inŽ . Ž . Ž .non-sensitized rats are immunonegative for CD40 L , CD80 O , and for CD86 R . 50 mm vibratome sections; =70. Scale bars150 mm.

expressed by T cells, B cells, macrophages, and thymo-.cytes, arrows in Fig. 4F . These data show that the bulk of

the cells which contacted the CPP were T-lymphocytes.The controls with omission of primary or secondary anti-bodies and the control incubations with non-relevant iso-type-matched biotinylated secondary antibodies containedonly FE-labeled perivascular phagocytes, identical withthose shown in Fig. 4A,C,E.

Two important issues must be underlined here. First, noleukocytes were ever observed to contact CPPrSPP in ratswhich had received only CFA in the sensitizing injection.Second, all lymphocytes which had penetrated the vascularwall did not leave the circulation in a Anon-specificB orAapparently randomB manner. Instead, 24 h and 3 daysafter the boosting injection, they always established con-tacts with the CPPrSPP.


Fig. 4. Identification of the adhering blood cells and expression of adhesion molecules at the blood–brain interface zone 24 h after i.c.v. application ofŽ . Ž . Ž .6.7% HRP plus 5% FE in sensitized rats. A–F FE-labeled CPP fluorescent in A, C, E contact T-lymphocytes arrows identified by the red

Ž .CY3-fluorescence after incubation with the T-cell receptor specific antibody R73 B , by the CD4 cell surface glycoprotein recognizing antibody W3r25Ž . Ž . Ž . ŽD , or by the rat CD44 cell surface antigen recognizing antibody OX-50 F . A 50-mm vibratome section; =100. G–L FE-labeled CPP fluorescent in

. Ž . Ž . Ž . Ž . ŽG, I, K in the vicinity of endothel cells arrows expressing CD162 H , the ICAM-1 J , and VCAM-1 L . Some CPP are also immunopositive open.arrow in H and L . A 50-mm vibratome section; =150.

3.3. CPP act as adhesion site for the extraÕasated lympho-cytes

To elucidate which molecular mechanisms might gov-ern the extravasation of lymphocytes and their adhesion toCPPrSPP in the model employed, we investigated whetherCPPrSPP expressed molecules promoting leukocyterolling, activation, and adherence.

3.3.1. Adhesion moleculesDespite the common view that even during ongoing

inflammation the expression of adhesion molecules by thecerebral endothelium is not sufficient to support leukocyte

Žrecruitment Engelhardt et al., 1994; Bell and Perry, 1995;.Perry et al., 1995 , we examined whether Fluoro-Emerald

Ž .FE labeled CPP expressed the adhesion molecule P-Ž .selectin CD162 on their surface. P-selectin is found on

endothelial cells and mediates adhesion to specific mucinsŽ . Žcarbohydrate ligands on many leukocytes Frenette and

.Wagner, 1996 . Employing anti-human serum, recognizingthe cell surface glycoprotein P-Selectin Glycoprotein Lig-and in dilution 1:50, we observed cross-reacting im-

Ž .munopositive cerebral endothelial cells arrow in Fig. 4HŽ .and CPP empty arrow in Fig. 4G 24 h after i.c.v.

boosting with HRP. An identical finding was detected innon-sensitized animals.

Ž .Employing a polyclonal anti-ICAM-1 M-19 in a dilu-tion of 1:50, we observed immunopositive endothelial cellsin the brain of both non-primed as well as primed and

Ž .boosted rats arrow in Fig. 4J 24 h after HRP-challenge.Ž .Some CPPrSPP, identified by their FE-labeling Fig. 4I

were also immunopositive.Ž .Likewise using polyclonal anti-VCAM-1 C-19 , we also

observed immunoreactive endothelial cells in non-sensi-

Ž .tized and in primed and boosted rats arrow in Fig. 4L 24h after HRP-boosting. The CPPrSPP, however, were im-

Ž .munonegative Fig. 4K . This result is in agreement with arecent work demonstrating expression of VCAM-1 mRNAin cerebral perivascular cells only in the acute phase of

Žimmune-mediated injury of the rat nervous system Jander.et al., 1996 .

3.3.2. Chemokines and chemokine receptorsChemokines are proposed to play a role in CNS inflam-

matory disease at the stage of leukocyte recruitment in theperivascular space in response to activated antigen-specific

Ž .T-lymphocytes Ransohoff et al., 1996, 1997 . Togetherwith their receptors, the chemokines build a complex andsophisticated biochemical signaling system involved in the

Žregulation of leukocyte and lymphocyte trafficking Harri-.son et al., 1998; Lavi et al., 1998 .

The chemokine receptor CCR-1 specifically binds nu-merous CC chemokines which control the attraction of

Ž .leukocytes to tissues Luster, 1998 . Incubation of sectionsfrom the brains of HRP-primed and boosted rats with 1:50anti-human CCR-1 revealed cross-reacting immunoposi-tive cells. Most of them were localized on the brain surfaceand thus were associated with the pia mater encephali. TheCCPrSPP, identifiable by their FE-labeling, were im-

Ž .munonegative Fig. 5A . The immunostaining of brainsections from animals which had undergone no HRP-prim-ing and subsequent i.c.v. challenge yielded identical resultsŽ .Fig. 5C .

The CC-chemokine monocyte chemoattractant proteinŽ .MCP -1 plays a crucial role in recruiting macrophages

Ž .and monocytes Calvo et al., 1996 . Employing the mono-clonal mouse anti-rat MCP-1 in dilution 1:50, we observedMCP-1 faintly positive endothelium in both non-sensitized


ŽFig. 5. Identification of chemokines and chemokine receptors in the pia mater 24 h after i.c.v. application of 6.7% HRP plus 5% FE in sensitized rats A,. Ž .C, E, G . All pictures in the middle column B, E, H, K are taken from sections in which the primary antibody has been omitted or a non-relevant

biotinylated secondary antibody has been used. All pictures in the right column are taken from the brain of non-sensitized but HRPrFE-injected animals.Ž .All images have been captured through Filter set 25 of Zeiss. A, C, D, F FE-labeled CPP in the vicinity of pial cells expressing the chemokine receptor

Ž . Ž . Ž .CCR-1 A, C and the CC-chemokine MCP-1 D, F . The CPP themselves are immunonegative. G I, J, L FE-labeled CPP expressing the chemokineŽ . Ž .receptor CCR-5 G, I and the CC chemokine RANTES J, L . 50 mm vibratome sections; =60.

Ž . Ž .Fig. 5F as well as in primed and boosted Fig. 5D rats.The FE-labeled CPPrSPP were immunonegative.

Incubation of sections with an antibody recognizing thechemokine receptor CCR-5 revealed cross-reacting en-

Ž .dothelial cells and CPPrSPP in non-sensitized Fig. 5IŽ .and in sensitized and HRP-challenged Fig. 5G rats. Most

of these immunopositive cells were also localized in piamater encephali.


Since CCR-5 belongs to the b chemokine receptorfamily and is known to provide directional cues in leuko-

Žcytic invasion and primary infection Bleul et al., 1997;.Woodroofe et al., 1999; Simpson et al., 2000 , we addi-

tionally incubated sections in antisera against chemokinesŽ .which specifically bind to CCR5: i the macrophage

Ž . Ž .inflammatory protein 1a MIP-1a , ii the macrophageŽ . Ž .inflammatory protein 1b MIP-1b , and iii RANTES

ŽCombadiere et al., 1996; Samson et al., 1996; Hvas et al.,.1997; Glabinski et al., 1998 .

Observations of sections from the brains of HRP-primedand boosted rats, which had been incubated with anti-hu-man and anti-rat MIP-1a or anti-human MIP-1b, showed

Ž .no immunopositive cells see also Materials and methods .However, observations of sections which had been incu-bated with anti-RANTES showed that many CPPrSPP

Ž .expressed this chemokine Fig. 5J . An identical pattern ofimmunostaining, though with a definitely lower intensity,was observed in control rats which had not been subjected

Ž .to HRP-priming and i.c.v. boosting Fig. 5L .

4. Discussion

The first major finding of this work is the developmentof a new model to study in vivo et situ lymphocytes, which

Žhad extravasated into the CNS perivascular Virchow–.Robin space after antigen challenge, but did not enter the

cerebral parenchyma. This compound model combines onŽ .one side the well-known i sensitization of rats to the

Ž .exogenous protein HRP Baloyannis and Gonatas, 1979Ž .and ii the intracerebroventricular injection of HRP to

Ž .label the perivascular phagocytes Wagner et al., 1974Ž .with iii an immunofluorescent staining of readily identifi-Ž . Žable fluorescent-labeled cells Angelov et al., 1995, 1996,.1998a .

Employing this model, we were able to study therelationship between CPPrSPP and extravasated lympho-cytes. This in turn provided the second major finding ofthe present report: CPPrSPP can indicate the sites oflymphocytic extravasation.

Earlier work has shown that the nervous system iscontinuously patrolled by T-lymphocytes which can cross

Žthe BBB only in an activated state Lassmann et al., 1986,.1991a,b; Raine et al., 1990 . When these T-lymphocytes

recognize an antigen presented in association with theMHC class II molecule, they initiate an inflammatoryreaction through the secretion of pro-inflammatory cy-

Ž .tokines Berger et al., 1997; Delves and Roitt, 2000 . Themost probable candidate for antigen presentation in the ratCNS are the ED2q cerebral perivascular phagocytesŽ .Graeber et al., 1992; Archelos and Hartung, 2000 . Acti-vating naive CD4q T-lymphocytes through secretion of

Žcytokines Martin and Resch, 1988; Croft and Dubey,.1997 and co-stimulatory molecules, the CPPrSPP could

play a key role in the repetitive lymphocytic invasions andŽleukocyte recruitment Sun and Wekerle, 1986; Matsumoto

.et al., 1992; Brosnan and Raine, 1996 .Confirming these suggestions, our present results show

Žthat antigen-containing perivascular phagocytes CPPr.SPP can indicate the locus of lymphocyte extravasation

Ž .though under non-inflammatory conditions . At first sight,this precisely located extravasation looks contradictory toearlier reports stating that T-lymphocytes aggressive to

Ž .myelin basic protein MBP or radically activated againstŽ .non-CNS antigens e.g. ovalbumin Apenetrate the blood–

Ž .brain barrier non-specificallyB Wekerle et al., 1986 or inŽ .an Aapparently random mannerB Hickey et al., 1991 .

These statements, however, concern the process of infiltra-tion of leukocytes and are thus valid only under inflamma-tory conditions. Furthermore, they hold true only for thisportion of the Virchow–Robin space which does not con-tain perivascular phagocytes, i.e. only for the space aroundcapillaries and venules. Whether the inflammatory extrava-sation of lymphocytes through the wall of arteries andarterioles, which are lined by perivascular phagocytesŽ .Weller et al., 1992, 1996 , is also random or mediated byantigen-containing perivascular phagocytes remains so farunknown.

Our results provide no direct evidence for antigen-specificity of this extravasation. Nevertheless, the onlyAtargetsB for lymphocytes, which get activated after chal-lenge, are the antigen-containing CPPrSPP. They upregu-late the expression of MHC class II glycoproteins, start thesynthesis of interleukin 1b, CD40- and B7-molecules. Onthe contrary, the expression of molecules promoting gen-

Žeral leukocyte rolling, activation, and adherence P-selec-.tin, ICAM-1, and VCAM-1 as well of those providing

Ždirectional cues in leukocytic invasion MCP-1, MIP-1a ,.MIP-1b, and RANTES remains at basal level. All this

allows us to conclude that there should be a certain degreeof antigen-specificity during this type of immune response.

It is well-known that, during inflammatory immuneresponse, the extravasation of lymphocytes is followed bya massive invasion of blood cells into the tissue, a processtermed AinfiltrationB. Whereas the pertinence of CPPrSPPto the group of antigen-presenting cells could explain theircapability to indicate the site for T-lymphocyte extravasa-tion, the reasons which we can point out for the lackinginfiltration of blood cells into the CNS-parenchyma arepurely speculative. A very plausible reason for this Amis-sedB infiltration could be that the injection of HRP in CFAinduces a predominately B-cell response, which in general,does not develop into a local inflammation. This explana-tion is supported by hard evidence for presence of antibod-ies to HRP and antibody-producing cells in the circulationŽBaloyannis and Gonatas, 1979; Geldof et al., 1983; Geldof

.and van de Ende, 1984 . Our results, however, show thatmost of the cells, which establish contact to CPPrSPPat24 h after single or chronic antigen boosting are T-

Ž .lymphocytes Fig. 4B,D . Whereas there is no doubt that


these T-lymphocytes have been accompanied by B-Ž .lymphocytes stained also by mab OX50 in Fig. 4F , we

cannot claim that the portion of the latter was dominatingover that of the T-lymphocytes.

A second reason may be that the antigen HRP merelyhas not the necessary antigenicity to develop an inflamma-tory immune response. Two major lines of evidence, how-ever, do not support this argument. First, the HRP-protein

Žhas been repeatedly employed as sensitizing antigen Weber.et al., 1974; Eikelenboom, 1978; Sminia et al., 1983 .

Second, although a structural breakdown of the blood–brainŽ .barrier BBB has not been identified, the widened inter-

Žcellular clefts between the endothelial cells containing the.activated leukocytes, Fig. 2C,D provide hard evidence for

a AselectiveB permeability. Such a disruption, however, hasbeen reported to occur only after specific antigen recogni-

Žtion Linnington et al., 1988; Seeldrayers et al., 1993;.Raivich et al., 1998, 1999 . Thus, the possibility that the

antigenicity of HRP might be insufficient to cause animmune response should be ruled out.

Finally, a third reason for the AmissedB infiltrationmight be that the HRP-labeled CPP not only fail to acti-vate, but even impede the blood cells from crossing thelamina superficialis gliae limitans and enter theparenchyma. The structures, which actually AstopB theinvasion of blood cells into the brain, are identical withthose which initiate the immune response by the presenta-tion of antigen. What is the biological meaning of this, atfirst site paradoxical event?

In our opinion, these two activities of CPPrSPP, anti-gen presentation and prevention of infiltrates, might becomponents of a finely synchronized and perfectly bal-

Ž .anced protective mechanism s . On one side, CPP areŽ .under conditions of an intact BBB the only brain phago-cytotic cells which can AsenseB any changes in the CSF-filled cerebral interstitium, cleave up neuronal debris andinfectious pathogens, and present the processed antigens tothe immune system in the context of MHC. On the otherside, preventing the entry of blood cells, including acti-vated lymphocytes into the brain parenchyma, the CPP

Ž .obviously fulfil other very important role s associatedwith neuronal protection.

Acknowledgements

This work has been financially supported by theŽ .Imhoff-Foundation D.N.A., M.W. , the Jean Uhrmacher-

Ž .Foundation A.P., M.S., O.G.-L. , and the Megapharm,Ž .Germany M.S. . The critical reading and useful comments

Ž . Ž .of C.D. Dijkstra Amsterdam , M.B. Graeber LondonŽ .and V.H. Perry Southampton are highly appreciated. The

authors are thankful for the skillful technical assistance ofI. Rohrmann and D. Felder and for the fine photographicalwork of I. Koch.

References

Angelov, D.N., Gunkel, A., Stennert, E., Neiss, W.F., 1995. Phagocyticmicroglia during delayed neuronal loss in the facial nucleus of the rat.Time course of the neuronofugal migration of brain macrophages.Glia 13, 113–129.

Angelov, D.N., Neiss, W.F., Streppel, M., Walther, M., Guntinas-Lichius,O., Stennert, E., 1996. ED2-positive perivascular cells act as neu-ronophages during delayed neuronal loss in the facial nucleus of therat. Glia 16, 129–139.

Angelov, D.N., Walther, M., Streppel, M., Guntinas-Lichius, O., Neiss,W.F., 1998a. The cerebral perivascular cells. Advances in Anatomy,Embryology and Cell Biology, vol. 147, Springer Verlag, Berlin.

Angelov, D.N., Streppel, M., Walther, M., Guntinas-Lichius, O., vanDam, A.M., Stennert, E., Neiss, W.F., 1998b. ED2-positive perivascu-lar neuronophages produce interleukin-1b during delayed neuronalloss in the facial nucleus of the rat. J. Neurosci. Res. 54, 820–827.

Archelos, J.J., Hartung, H.-P., 2000. Pathogenetic role of autoantibodiesin neurological disease. TINS 23, 317–327.

Baloyannis, S., Gonatas, N.K., 1979. Distribution of anti-HRP antibodiesin the central nervous system of immunized rats after disruption of theblood brain barrier. J. Neuropathol. Exp. Neurol. 38, 519–531.

Bauer, J., Huitinga, I., Zhao, W., Lassmann, H., Hickey, W., Dijkstra,C.D., 1995. The role of macrophages, perivascular cells, and mi-croglial cells in the pathogenesis of experimental autoimmune en-cephalomyelitis. Glia 15, 437–446.

Bechmann, I., Lossau, S., Steiner, B., Mor, G., Gimsa, U., Nitsch, R.,Ž . Ž .2000. Reactive astrocytes upregulate Fas CD95 and FasL CD95L

expression, but do not undergo programmed cell death in the courseof anterograde degeneration. Glia 32, 25–41.

Bell, M.D., Perry, V.H., 1995. Adhesion molecule expression on murinecerebral endothelium following the injection of a proinflammogen orduring acute neuronal degeneration. J. Neurocytol. 24, 695–710.

Berger, T., Weerth, S., Kojima, K., Linington, C., Wekerle, H., Lass-mann, H., 1997. Experimental autoimmune encephalomyelitis: theantigen specificity of T-lymphocytes determines the topography oflesions in the central and peripheral nervous system. Lab. Invest. 76,355–364.

Bleul, C.C., Wu, L., Hoxie, J.A., Springer, T.A., Mackay, C.R., 1997.The HIV coreceptors CXCR4 and CCR5 are differentially expressedand regulated on human T lymphocytes. Proc. Natl. Acad. Sci. U. S.A. 94, 1925–1930.

Brosnan, C.F., Raine, C.S., 1996. Mechanisms of immune injury inmultiple sclerosis. Brain Pathol. 6, 243–257.

Calvo, C.F., Yoshimura, T., Gelman, M., Mallat, M., 1996. Production ofmonocyte chemotactic protein-1 by rat brain macrophages. Eur. J.Neurosci. 8, 1725–1734.

Combadiere, C., Ahuja, S.K., Tiffany, H.L., 1996. Cloning and functionalexpression of CC CKR5, a human monocyte CC chemokine receptor

Ž . Ž .selective for MIP-1 alpha , MIP-1 beta and RANTES. J. LeukocyteBiol. 60, 147–152.

Croft, M., Dubey, C., 1997. Accessory molecule and costimulationrequirements for CD4 T cell response. Crit. Rev. Immunol. 17,89–118.

Cross, A.H., Cannella, B., Brosnan, C.F., Raine, C.S., 1990. Homing tocentral nervous system vasculature by antigen specific lymphocytes: I.Localization of 14C-labeled cells during acute, chronic and relapsingexperimental allergic encephalomyelitis. Lab. Invest. 63, 162–170.

Cross, A.H., O’Mara, T., Raine, C.S., 1993. Chronologic localization ofmyelin-reactive cells in the lesions of relapsing EAE: implications forthe study of multiple sclerosis. Neurology 43, 1028–1033.

Deiniger, M.H., Zhao, Y., Schluesener, H.J., 1999. CP-10, a chemotacticpeptide, is expressed in lesions of experimental autoimmune en-cephalomyelitis, neuritis, uveitis and in C6 gliomas. J. Neuroim-munol. 93, 156–163.

Delves, P.J., Roitt, I.M., 2000. The immune system. N. Engl. J. Med.343, 37–49.


Dijkstra, C.D., Dopp, E.A., Joling, P., Kraal, G., 1985. The heterogeneity¨of mononuclear phagocytes in lymphoid organs: distinct macrophagesubpopulations in the rat recognized by monoclonal antibodies ED1,ED2 and ED3. Immunology 54, 589–599.

Eikelenboom, P., 1978. Dendritic cells in the rat spleen follicles. Acombined immuno- and enzyme histochemical study. Cell Tissue Res.190, 79–87.

Engelhardt, J.I., Tajti, J., Appel, S.H., 1993. Lymphocytic infiltrates inthe spinal cord in amyotrophic lateral sclerosis. Arch. Neurol. 50,30–36.

Engelhardt, B., Conley, F.K., Butcher, E.C., 1994. Cell-adhesionmolecules on vessels during inflammation in the mouse central ner-vous system. J. Neuroimmunol. 51, 199–208.

Frenette, P.S., Wagner, D.D., 1996. Adhesion molecules. N. Engl. J.Med. 334, 1526–1529.

Gautam, A.M., Glynn, P., 1989. Lewis rat lymphoid dendritic cells canefficiently present homologous myelin basic protein to encephalito-genic lymphocytes. J. Neuroimmunol. 22, 113–121.

Gehrmann, J., Gold, R., Linington, C., Lannes-Vieira, J., Wekerle, H.,Kreutzberg, G.W., 1993. Microglial involvement in experimentalautoimmune inflammation of the central and peripheral nervous sys-tem. Glia 7, 50–59.

Geldof, A.A., van de Ende, M.B., 1984. Migration of specific antibody-forming cells during the secondary immune response against

w xhorseradish peroxidase. Virchows Arch. Cell Pathol. 45, 107–112.Geldof, A.A., van de Ende, M.B., Janse, E.M., Sminia, T., 1983. Specific

antibody formation in mouse spleen. Histology and kinetics of thesecondary immune response against HRP. Cell Tissue Res. 231,135–142.

Glabinski, A.R., Tuohy, V.K., Ransohoff, R.M., 1998. Expression ofchemokines RANTES, MIP-1alpha and GRO-alpha correlates withinflammation in acute experimental autoimmune encephalomyelitis.Neuroimmunomodulation 5, 166–171.

Graeber, M.B., Streit, W.J., Kreutzberg, G.W., 1989. Identity of ED2-positive perivascular cells in rat brain. J. Neurosci. Res. 22, 103–106.

Graeber, M.B., Streit, W.J., Buringer, D., Larry Sparks, D., Kreutzberg,¨G.W., 1992. Ultrastructural location of major histocompatibility com-

Ž .plex MHC class II positive perivascular cells in histologicallynormal human brain. J. Neuropathol. Exp. Neurol. 51, 303–311.

Harrison, J.K., Jiang, Y., Chen, S., Xia, Y., Macijewski, D., McNamara,R.K., Streit, W.J., Salafranca, M.N., Adhikari, S., Thompson, D.A.,Botti, P., Bacon, K.B., Feng, L., 1998. Role for neuronally derivedfractalkine in mediating interactions between neurons and CX3CR1-expressing microglia. Proc. Natl. Acad. Sci. U. S. A. 95, 10896–10901.

Hickey, W.F., Kimura, H., 1988. Perivascular microglial cells of the CNSare bone marrow derived and present antigen in vivo. Science 239,290–292.

Hickey, W.F., Hsu, B.L., Kimura, H., 1991. T-lymphocyte entry into thecentral nervous system. J. Neurosci. Res. 28, 254–260.

Hickey, W.F., Vass, K., Lassmann, K., 1992. Bone marrow-derivedelements in the central nervous system: an immunohistochemical andultrastructural survey of rat chimeras. J. Neuropathol. Exp. Neurol. 5Ž .1 , 246–256.

Hu, P., Pollard, J., Hunt, N., Chan-Ling, T., 1998. Microvascular andcellular responses in the retina of rats with acute experimental allergic

Ž .encephalomyelitis EAE . Brain Pathol. 8, 487–498.Hvas, J., McLean, C., Justesen, J., Kannourakis, G., Steinman, L.,

Oksenberg, J.R., Bernard, C.C., 1997. Perivascular T cells express thepro-inflammatory chemokine RANTES mRNA in multiple sclerosislesions. Scand. J. Immunol. 46, 195–203.

Jander, S., Pohl, J., Gillen, C., Schroeter, M., Stoll, G., 1996. Vascularcell adhesion molecule-1 mRNA is expressed in immune-mediatedand ischemic injury of the rat nervous system. J. Neuroimmunol. 70,75–80.

Kawamata, T., Akiyama, A., Yamada, T., McGeer, P.L., 1992. Immuno-

logic reactions in amyotrophic lateral sclerosis brain and spinal cord.Am. J. Pathol. 140, 691–707.

Kennedy, D.W., Abkowitz, J.L., 1997. Kinetics of the central nervoussystem microglial and macrophage engraftment: analysis using atransgenic bone marrow transplantation model. Blood 90, 986–993.

Kida, S., Steart, P.V., Zhang, E.T., Weller, R.O., 1993. Perivascular cellsact as scavengers in the cerebral perivascular spaces and remaindistinct from pericytes, microglia and macrophages. Acta Neu-ropathol. 85, 646–652.

Knopf, P.M., Harling-Berg, C.J., Cserr, H.F., Basu, D., Sirulnick, E.J.,Nolan, S.C., Park, J.T., Keir, G., Thompson, E.J., Hickey, W.F.,1998. Antigen-dependent intrathecal synthesis in the normal rat brain:tissue entry and local retention of antigen specific B cells. J. Im-munol. 161, 692–701.

Kosel, S., Egensperger, R., Bise, K., Arbogast, S., Mehraein, P., Graeber,¨M.B., 1997. Long-lasting perivascular accumulation of major histo-compatibility complex class II-positive lipophages in the spinal cordof stroke patients: possible relevance for the immune privilege of the

Ž .brain. Acta Neuropathol. Berlin 94, 532–538.Lassmann, H., Vass, K., Brunner, C., Seutelberger, F., 1986. Characteri-

zation of inflammatory infiltrates in experimental allergic en-cephalomyelitis. Prog. Neuropathol. 6, 33–62.

Lassmann, H., Zimprich, F., Vass, K., Hickey, W.F., 1991a. Microglialcells are a component of the perivascular glia limitans. J. Neurosci.Res. 28, 236–243.

Lassmann, H., Zimprich, F., Rossler, K., Vass, K., 1991b. Inflamation in¨Ž .the nervous system. Rev. Neurol. Paris 147, 763–781.

Lassmann, H., Schmied, M., Vass, K., Hickey, W.F., 1993. Bone marrowderived elements and resident microglia in brain inflammation. Glia 7,19–24.

Lavi, E., Kolson, D.L., Ulrich, A.M., Fu, L., Gonzalez-Scarano, F., 1998.Chemokine receptors in the human brain and their relationship to HIVinfection. J. NeuroVirol. 4, 301–311.

Linnington, C., Bradl, M., Lassmann, H., Brunner, C., Vass, K., 1988.Augmentation of demyelination of rat acute allergic encephalomyeli-tis by circulating mouse monoclonal antibodies directed against amyelinroligodendrocyte glycoprotein. Am. J. Pathol. 130, 443–454.

Litzenburger, T., Fassler, R., Bauer, J., Lassmann, H., Linington, C.,¨Wekerle, H., Iglesias, A., 1998. B lymphocytes producing demyelinat-ing autoantibodies: development and function in gene-targeted trans-genic mice. J. Exp. Med. 188, 169–180.

Luster, A.D., 1998. Chemokines—chemotactic cytokines that mediateinflammation. N. Engl. J. Med. 338, 436–445.

Martin, M., Resch, K., 1988. Interleukin 1: more than a mediator betweenleukocytes. TINS 9, 171–177.

Mato, N., Ookara, S., Sakamoto, A., Aikawa, E., Ogawa, T., Mitsuhashi,U., Masuzawa, T., Suzuki, H., Honda, M., Yazaki, Y., Watanabe, E.,Luoma, J., Yla-Herttuala, S., Fraser, I., Gordon, S., Kodama, T.,1996. Involvement of specific macrophage-lineage cell surroundingarterioles in barrier and scavenger function in brain cortex. Proc. Natl.Acad. Sci. U. S. A. 93, 3269–3274.

Matsumoto, Y., Hara, N., Tanaka, R., Fujihara, M., 1986. Immunohisto-chemical analysis of the rat central nervous system during experimen-tal allergic encephalomyelitis, with special reference to Ia-positivecells with dendritic morphology. J. Neuroimmunol. 136, 3668–3676.

Matsumoto, Y., Ohmori, J., Fujihara, M., 1992. Immune regulation bybrain cells in the central nervous system: microglia but not astrocytespresent myelin basic protein to encephalitogenic T-cells under in vivomimicking conditions. Immunology 76, 209–216.

Matyszak, M.K., Perry, V.H., 1995. Demyelination in the central nervoussystem following a delayed-type hypersensitivity response to bacillusCalmette–Guerin. Neuroscience 64, 967–977.

Matyszak, M.K., Perry, V.H., 1996. Delayed type hypersensitivity lesionsin the central nervous system are prevented by inhibitors of metallo-proteinases. J. Neuroimmunol. 69, 141–149.

Matyszak, M.K., Perry, V.H., 1998. Bacillus Calmette–Guerin se-


questered in the brain parenchyma escapes immune recognition. J.Neuroimmunol. 82, 73–80.

McLean, I.W., Nakane, P.K., 1974. Periodate–lysine–paraformaldehydefixative a new fixative for immunoelectron microscopy. J. Histochem.Cytochem. 22, 1077–1087.

McMaster, W.R., Williams, A.F., 1979. Identification of Ia glycoproteinsin rat thymus and purification from rat spleen. Eur. J. Immunol. 9,426–433.

Mingheti, L., Walsh, D.T., Levi, G., Perry, V.H., 1999. In vivo expres-sion of cyclooxygenase-2 in rat brain following intraparenchymalinjection of bacterial endotoxin and inflammatory cytokines. J. Neu-ropathol. Exp. Neurol. 58, 1184–1191.

Morimoto, K., Craig Hooper, D., Bornhorst, A., Corisdeo, S., Bette, M.,Fu, Z.F., Schafer, M.K.-H., Koprowsky, H., Weihe, E., Dietzschold,¨B., 1996. Intrinsic responses to Borna disease virus infection of thecentral nervous system. Proc. Natl. Acad. Sci. U. S. A. 93, 13345–13350.

Perry, V.H., 1998. A revised view of the central nervous system microen-vironment and major histocompatibility complex class II antigenpresentation. J. Neuroimmunol. 90, 113–121.

Perry, V.H., Bell, M.D., Brown, H.C., Matyszak, M.K., 1995. Inflama-tion in the nervous system. Curr. Opin. Neurobiol. 5, 636–641.

Raine, C.S., Cannella, B., Duijvestijn, A.M., Cross, A.H., 1990. Homingto central nervous system vasculature by antigen-specific lympho-cytes: II. Lymphocyterendothelial cell adhesion during the initialstages of autoimmune demyelination. Lab. Invest. 63, 476–489.

Raivich, G., Jones, L.L., Kloss, C.U.A., Werner, A., Neumann, H.,Kreutzberg, G.W., 1998. Immune surveillance in the injured nervoussystem: T-lymphocytes invade the axotomized mouse facial motornucleus and aggregate around sites of neuronal degeneration. J.Neurosci. 18, 5804–5816.

Raivich, G., Bohatscheck, M., Kloss, C.U.A., Werner, A., Neumann, H.,Jones, L.L., Kreutzberg, G.W., 1999. Neuroglial activation repertoirein the injured brain: graded response, molecular mechanisms and cuesto physiological function. Brain Res. Rev. 30, 77–105.

Ransohoff, R.M., Glabinski, A.R., Tani, M., 1996. Chemokines in im-mune-mediated inflammation of the central nervous system. CytokineGrowth Factor Rev. 7, 35–46.

Ransohoff, R.M., Tani, M., Glabinski, A.R., Chernosky, A., Krivacic, K.,Peterson, J.W., Chien, H.-F., Trapp, B.D., 1997. Chemokines andchemokine receptors in model neurological pathologies: molecularand immunohistochemical approaches. Methods Enzymol. 287, 319–348.

Samson, M., Labbe, O., Mollereau, C., Vassart, G., Parmentier, M., 1996.Molecular cloning and functional expression of a new human CC-chemokine receptor gene. Biochemistry 35, 3362–3367.

Schluesener, H.J., Seid, K., Kretzschmar, J., Meyermann, R., 1996.Leukocyte chemotactic factor, a natural ligand to CD4, is expressedby lymphocytes and microglial cells of the MS plaque. J. Neurosci.Res. 44, 606–611.

Seeldrayers, P.A., Syha, J., Morrissey, S.P., Stodal, H., Vass, K., Jung,S., Gneiting, T., Lassmann, H., Haase, A., Hartung, H.P., 1993.Magnetic resonance imaging investigation of blood–brain-barrierdamage in adoptive transfer of experimental autoimmune en-cephalomyeitis. J. Neuroimmunol. 46, 199–206.

Serpe, C.J., Kohm, A.P., Huppenbauer, C.B., Sanders, V.M., Jones, K.J.,1999. Exacerbation of facial motoneuron loss after facial nerve tran-

Ž .section in severe combined immunodeficient scid mice. J. Neurosci.19RCN, 1–5.

Simpson, J., Rezaie, P., Newcombe, J., Luise Cuzner, M., Male, D.,Woodroofe, M.N., 2000. Expression of the b-chemokine receptorsCCR2, CCR3 and CCR5 in multiple sclerosis central nervous systemtissue. J. Neuroimmunol. 108, 192–200.

Sminia, T., Delemarre, F., Janse, E.M., 1983. Histological observationson the intestinal immune response towards horseradish peroxidase inrats. Immunology 50, 53–56.

Streit, W.J., 1990. An improved staining method for rat microglial cellsŽ .using the lectin from Griffonia simplicifolia GSA I-B4 . J. His-

tochem. Cytochem. 38, 1683–1686.Streit, W.J., Graeber, M.B., Kreutzberg, G.W., 1989. Expression of Ia

Antigen on perivascular and microglial cells after sublethal and lethalmotor neuron injury. Exp. Neurol. 105, 115–126.

Sun, D., Wekerle, H., 1986. Ia restricted encephalitogenic T lymphocytesmediating EAE lyse autoantigen-presenting astrocytes. Nature 320,70–72.

Wagner, H.J., Pilgrim, C., Brandl, J., 1974. Penetration and removal ofhorseradish peroxidase injected into the cerebrospinal fluid: role ofcerebral perivascular spaces, endothelium and microglia. Acta Neu-ropathol. 27, 299–315.

Weber, K., Ueki, H., Wolff, H.H., Braun-Falco, O., 1974. Reversedpassive Arthus reaction using horseradish peroxidase as antigen: II.Light and electron microscopical observations. Arch. Dermatol.Forsch. 250, 15–32.

Wekerle, H., Linington, C., Lassmann, H., Meyermann, R., 1986. Cellu-lar immune reactivity within the CNS. TINS 9, 271–277.

Weller, R.O., Kida, S., Zhang, E.T., 1992. Pathways of fluid drainagefrom the brain: morphological aspects and immunological signifi-cance in rat and in man. Brain Pathol. 2, 277–284.

Weller, R.O., Engelhardt, B., Phillips, M.J., 1996. Lymphocyte targetingof the central nervous system: a review of afferent and efferentCNS-immune pathways. Brain Pathol. 6, 275–288.

Woodroofe, M.N., Cross, A.K., Harkness, K.A., Simpson, J.E., 1999.The role of chemokines in the pathogenesis of multiple sclerosis.Adv. Exp. Med. Biol. 468, 135–150.

Zhang, E.T., Richards, H.K., Kida, S., Weller, R.O., 1992. Directionaland compartmentalized drainage of interstitial fluid from the rat brain.Acta Neuropathol. 83, 233–239.