In situ processing and distribution of intracerebrally injected OVA in the CNS

Journal of Neuroimmunology 141 (2003) 90 – 98 www.elsevier.com/locate/jneuroim

In situ processing and distribution of intracerebrally injected OVA $ in the CNS Changying Ling, Matyas Sandor, Zsuzsa Fabry* Department of Pathology and Laboratory Medicine, University of Wisconsin, 1300 University Avenue, 6130 MSC, Madison, WI 53706, USA Received 28 February 2003; received in revised form 17 June 2003; accepted 18 June 2003

Abstract Drainage and retention of brain-derived antigens are important factors in initiating and regulating immune responses in the central nervous system (CNS). We investigated distribution, immunological processing and retention of intracerebrally infused protein antigen, ovalbumin (OVA), and the subsequent recruitment of CD8+ T cells into the CNS. We found that protein antigens infused into the CNS can drain rapidly into the cervical lymph node and initiate antigen-specific immune response in the periphery. A portion of the antigens are also retained by CD11b/MAC-1+ cells in the brain parenchyma where they are recognized by antigen-specific CD8+ T cells. D 2003 Elsevier B.V. All rights reserved. Keywords: Protein antigen; Ovalbumin; Antigen retention; Preferential distribution; Brain parenchyma; CD8+ T cell

1. Introduction The central nervous system (CNS) has been regarded as an immunologically privileged organ due to the absence of classically defined lymphatics and the presence of the blood –brain barrier (BBB). However, communication between the CNS and the immune system occurs, even in the presence of an intact BBB and the absence of CNS diseases and infections. The cerebral extracellular fluid provides the opportunity for the communication by delivering CNS-derived antigens to peripheral lymphoid organs (Cserr and Ostrach, 1974; Ghersi-Egea et al., 1996; Hochwald et al., 1988; Ichimura et al., 1991; Szentistvanyi et al., 1984; Widner et al., 1988). Through the efflux of extracellular fluids, brain-derived protein antigens drain rapidly into peripheral lymph nodes after administration, notably cervical lymphatics (Boulton et al., 1998; Bradbury et al., 1981; Yamada et al., 1991). Through such drainage, brain-derived protein antigens elicit an antigen-specific humoral immune response, as evidenced by the generation of antibody-secreting plasma cells in cervical lymph nodes (CLN) and the $ This work was supported by National Institute of Health, Grant RO1-NS 37570-01A2 to Z. Fabry. * Corresponding author. Tel.: +1-608-265-8716; fax: +1-608-2653301. E-mail address: [email protected] (Z. Fabry).

0165-5728/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0165-5728(03)00249-2

appearance of antibodies in blood (Gordon et al., 1992; Hochwald et al., 1988; Knopf et al., 1995; Widner et al., 1988). Although eliciting a robust serum antibody response, the intracerebral infusion of protein antigens fails to induce a specific DTH response (Harling-Berg et al., 1999). It has also been reported that following an intracerebral infusion, the majority of infused soluble protein antigens drain into peripheral lymphoid organs and only a small, but immunologically significant, portion of the antigen may be retained in the CNS (Cserr et al., 1992). The longest retention of soluble protein antigens that has been reported was only 96 h (Oehmichen and Huber, 1978). By contrast, following intracerebral administration, Mycobacterium bovis (BCG) is phagocytosed by macrophages and retained in the brain parenchyma for up to 1 year (Perry, 2000). To further clarify the contribution of drainage and local persistence of intracerebrally infused protein antigens, we analyzed the in situ processing of a protein antigen, ovalbumin (OVA), in the CNS and its capability to induce antigen-specific CD8+ T cell responses in the periphery. Here we demonstrate that the intracerebrally infused OVA is processed locally and distributed preferentially in the CNS. The infused OVA partially drains to the CLN and activates antigen-specific CD8+ T cells, and these activated CD8+ T cells traffic into the brain. These data further emphasize the importance of local processing and retention of protein antigens in recruitment of antigen-specific CD8+ T cells into the CNS.

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

2. Materials and methods 2.1. Mice C57BL/6 (B6), 5- to 7-week-old female mice were purchased from Jackson Laboratory (Bar Harbor, ME). Mice transgenic for a TCR recognizing OVA257 – 264 (SIINFEKL) peptide bound to H2-Kb (OT-1) on the B6 background (Hogquist et al., 1994; Kelly et al., 1993) were provided by Dr. Kris Hoguist (University of Minnesota). Mice of this strain were used at 6 –10 weeks of age. All mice were housed in a pathogen-free facility at the University of Wisconsin, Medical School Animal Care Unit. All protocols involved in these experiments were approved by the Committee on Animal Care, University of Wisconsin-Madison. 2.2. Tetramer, antibodies, and antigens

91

mg/kg of ketamine and 10 mg/kg of xylazine. Antigen (DQOVA or OVA, 60 Ag diluted into 20 Al sterile PBS) or equal volume PBS was injected into the right frontal lobe through an insulin syringe attached to a penetrating depth controller. The injection was restricted to the ventral-posterior region of the frontal lobe, and the penetrating depth of the syringe was 1.5 mm from the surface of the brain. For each IC injection, the solution was delivered slowly, and then the syringe was held in place for an additional minute to reduce backfilling of injected solution. In some cases, IC immunizations were repeated 14 days following the primary injection. 2.4. Lymphocyte isolation One to four weeks after the last immunization, immunized mice and age-matched control mice were anesthetized and perfused transcardially with PBS. Cervical lymph nodes (CLN) and brains were dissected and transferred into cold Hanks’ buffered saline solution (HBSS). The CLN was processed to single cell suspensions by standard methods (Qing et al., 2000). Brains were transferred into MediCons disposable units and processed to single cell suspensions with a Medimachine (both from Becton-Dickinson). Cell suspensions were centrifuged, resuspended in 2-ml 50% Percoll (Pharmacia, Piscataway, NY), and overlaid with 30% Percoll. After centrifuging at 2500  g for 30 min, the lymphocytes were harvested from the interface between the two Percoll layers. The amount of contamination of the brain lymphocyte preparation with peripheral cells was assessed using proflavin staining (Brabb et al., 2000; Rodriguez-Peralta, 1968). In a separate experiment, mice received an intravenous administration of proflavin hydrochloride (20 mg/ml, 200 Al/mouse). More than 85% of lymphocytes isolated from brain were negative for proflavin (data not shown). Isolated lymphocytes were washed with HBSS twice and resuspended in FACS staining buffer (balanced salt solution containing 5% fetal calf serum and 0.1% NaN3).

MHC I-Kb/SIINFEKL tetrameric complexes conjugated with streptavidin-APC (SIINFEKL-APC) were provided by MHC Tetramer Core Facility, National Institute of Allergy and Infectious Disease (Atlanta, GA). MHC I-Kb SIINFEKL tetrameric complexes conjugated with streptavidinPE (SIINFEKL-PE) were purchased from Beckman Coulter (Fullerton, CA). Each batch of SIINFEKL tetramer complexes was titrated and normalized using OT-1 splenic T cells before being used in experiments. Antibodies purchased from PharMingen (San Diego, CA) included anti-CD8a (conjugated to PE, or APC). Hybridoma lines producing anti-CD11b/MAC-1 (clone M1/70, Rat IgG2b) and anti-CD45R/B220 (clone RA36B2, Rat IgG2a) were obtained from ATCC (American Type Culture Collection, Rockville, MD), and used to produce monoclonal antibodies. The purified antibodies were conjugated with FITC or Cy-5 as described previously (Fabry et al., 1992). Ovalbumin (OVA, chicken egg albumin) was purchased from Sigma (St Louis, MO). DQk ovalbumin (DQ-OVA) was purchased from Molecular Probes (Eugene, OR). DQOVA is a self-quenched conjugate of OVA and BODIPY, a pH-insensitive fluorogenic substrate for proteases. It is designed especially for the study of antigen processing and presentation. In intact DQ-OVA, fluorescence is quenched, with very low background signal. Upon hydrolysis of the DQ-OVA to single, dye-labeled peptides by proteases, the quenching is relieved, and brightly fluorescent products with emission maximum of 515 nm (green) are released. When the digested fragments of DQ-OVA accumulate in organelles at a high concentration, the BODIPY FL fluorophores can potentially form excimers, resulting in a shift of the fluorophore’s emission maximum from 515 nm (green) to f 620 nm (red).

For quantitative identification of activated CD8+ T cells, 1  106 cells isolated from the CLN, or f 5  104 mononuclear cells isolated from the brain were stained with monoclonal antibodies, anti-CD8a (PE), -CD11b/MAC-1 (FITC), CD45R/B220 (FITC) and SIINFEKL-APC for flow cytometric analysis. Antibody incubations included unlabeled antibody specific for Fcg III/II receptor, clone 2.4G2, to block binding to Fc receptors. Cell surface staining was measured using a FACS CaliburR instrument (Becton-Dickinson) and four-color FACSR data analyzed using CellQuest (Becton-Dickinson).

2.3. Intracerebral antigen delivery

2.6. Immunohistochemistry and confocal microscopy

For intracerebral (IC) antigen delivery, mice were anesthetized by intraperitoneal (IP) injection of a mixture of 90

Experimental or control mice were perfused with sterile PBS under deep anesthesia, the cervical lymph nodes and

2.5. Flow cytometry

92

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

brains were removed, immediately frozen on dry ice, and sectioned with a cryostat microtome at an interval of 8 Am. In some cases, the brains were removed, post-fixed in 4% paraformaldehyde in PBS, and sectioned with a vibrotome at an interval of 50 Am. CD11b/MAC-1+ cells, CD8+ T cells and SIINFEKL+ CD8+ T cells in the frozen sections, or in the vibrotome sections were detected immunocytochemically using anti-CD11b/MAC-1 antibody (conjugated with Cy-5), anti-CD8a antibody (conjugated with APC) and SIINFEKL-PE tetramer, according to the protocol provided by NIAID MHC tetramer core facility. The presence of CD8+ and SIINFEKL+ cells in the brain was confirmed, and their localization was examined and imaged by a confocal laser scanning microscope (Bio-Rad MRC 1000). The presence and distribution of proteolytic DQ-OVA in both the brain and the cervical lymph node was also visualized and imaged by the confocal microscope. Digital data were exported into Adobe Photoshop version 5.0 for the Macintosh, for further analysis and presentation. Fig. 1. The distribution of proteolytically degraded DQ-OVA in the CLN of experimental animals following intracerebral immunization. Mice received a single intracerebral injection of sterile DQ-OVA and were sacrificed 2, 4 or 8 h, or 7 days later. The CLN was dissected, embedded in O.C.T., and sectioned by a cryostat (at 8-Am interval). On each section, the processed DQ-OVA is visualized by emission of green fluorescence (indicated by arrows). Scale bar: 200 Am.

2.7. Statistics Data were analyzed statistically with two-tailed Student’s t-test and ANOVA. All values are expressed as mean F standard error (S.E.).

Fig. 2. Proteolyzed DQ-OVA in the brain 2 h (left panel), 8 h (center panel) and 7 days (right panel) following a single intracerebral immunization. Mice received a single intracerebral injection of sterile DQ-OVA. Two hours, eight hours and seven days following this immunization, the brain was dissected by a vibrotome (at 50-Am interval). All sections were examined and imaged using confocal microscopy. DQ-OVA is a fluorogenic substrate for proteases. The intact DQ-OVA is nonfluorescent due to auto-quenching. Upon proteolytic digestion, the quenching is lost, and dye-conjugated peptides emit bright fluorescence with emission maximum of 515 nm (green). When the digested fragments accumulate in organelles at a high concentration, the BODIPY FL fluorophores can potentially form excimers, resulting in a shift of the fluorophore’s emission maximum from 515 nm (green) to f 620 nm (red). For each image, green fluorescence indicates the hydrolysis of DQ-OVA to single, dye-labeled peptides by proteases, while red fluorescence indicates increased eximer formation. Top row: red channel; lower row: green channel. Arrows indicate injection tracks. D: dorsal; L: lateral. Scale bar: 250 Am.

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

93

3. Results 3.1. Drainage of DQ-OVA into cervical lymph nodes after IC antigen delivery To investigate the drainage of protein antigen from the CNS to the periphery, mice received an intracerebral DQOVA injection. Intact DQ-OVA is nonfluorescent due to auto-quenching. After proteolytic digestion, such as after internalization by mononuclear cells, DQ-OVA emits green fluorescence at relatively low concentration and red fluorescence at high concentration. Utilizing these specific features of DQ-OVA, we examined the drainage of intracerebrally (IC) administered DQ-OVA into the CLN at different times after administration. As shown in Fig. 1, the processed DQ-OVA is detected in the CLN as early as 2 h after injection, where it is largely restricted to the afferent lymphatic vessels around the surface of the CLN (indicated by arrows in the upper-left panel in Fig. 1). As draining and processing continued, the fluorescent intensity of DQ-OVA in the CLN reaches a peak during the next 2 h, and the densely accumulated DQ-OVA is seen in both the afferent lymphatic vessels and in the cortical area (indicated by arrows in the upper-right panel of Fig. 1). Eight hours after IC injection, the processed DQ-OVA was visualized as scattered granules in the CLN (indicated by arrows in lower-left panel of Fig. 1). A similar distribution of the processed DQ-OVA is seen 7 days later (lower-right panel in Fig. 1), but the density of fluorescent granules is reduced. In

Fig. 4. DQ-OVA (green) internalized by CD11b/MAC-1+ cells (blue) at the injection site following an intracerebral DQ-OVA injection. Mice received a single intracerebral injection of sterile DQ-OVA. Seven days or four weeks following this injection, animals were sacrificed, and their brains were dissected and sectioned. Sections were stained with anti-CD11b/MAC-1 antibody (conjugated with Cy5), examined, and imaged using a confocal microscopy. Arrowheads: CD11b/MAC-1+ cells containing DQ-OVA in their cytoplasm; small arrows in A: the processed DQ-OVA at the injection site; large arrows in B: CD11b/MAC-1+ cell without internalized DQ-OVA. Scale bar: 200 Am.

contrast, no fluorescence was detected in any CLNs of naı¨ve mice or mice with an intracerebral PBS injection (data not shown). These observations indicate that protein antigens drain into the CLN following intracerebral antigen infusion. 3.2. DQ-OVA is processed in situ and localized preferentially in the brain

Fig. 3. Preferential distribution of proteolyzed DQ-OVA in the brain seven days following intracerebral immunization. Left: distribution of the proteolyzed DQ-OVA in the brain at low magnification (100  ). Right: Image of proteolyzed DQ-OVA at the bottom of injection track at higher magnification (  400) shows an antigen-free region in the white matter (arrows). Top right: confocal microscopy image; bottom right: a Nomarski interference-contrast image of the same area. D: dorsal; L: lateral. Scale bar: 250 Am in the left panel; 100 Am in the right panels of B.

Our next experiment was aimed at characterizing the processing of intracerebrally infused protein antigens. We examined the processed DQ-OVA in the brain at different times after IC DQ-OVA delivery. Fig. 2 shows that during the first 2 h after injection, only a small portion of the infused DQ-OVA emits detectable fluorescence, and the fluorescence is largely restricted to the injection track (left panels in Fig. 2), indicating that the majority of infused DQ-OVA has not been proteolyzed yet. Eight hours following IC antigen delivery, there is a significant increase in the fluorescence intensity in both green and red channels in the injection site, and the fluorescent label disperses in the ventrolateral direction, indicating increased proteolysis and directed dispersion of the processed OVA in the brain parenchyma (center panels in Fig. 2). Seven days after IC injection, the fluorescent intensity in the injection site is reduced significantly (right panels in Fig. 2), indicating that only a small portion of infused OVA is retained in the CNS. Interestingly, a small amount of proteolyzed OVA is distrib-

94

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

uted ventrolaterally along the ipsilateral external capsule, up to 4 mm from the original injection site (Fig. 3). This localized dispersion of DQ-OVA antigens leaves an antigen-free region in the white matter immediately beneath the injection tract (arrows in Fig. 3). These observations indicate that protein antigen can be processed locally in the brain and suggest a preferential drainage, rather than free diffusion, of the brain-derived antigen in the brain parenchyma.

analysis demonstrates that IC-OVA delivery induces a significant increase in both the percentage of CD8+ T cells within the lymphocyte population and the percentage of tetramer+ cells within CD8+ T cell population in the CLN and the brain (Fig. 5). The peak SIINFEKL+ percentage is seen in the CLN 3 days after IC injection, but occurs in the brain 4 days later, when the SIINFEKL+ percentage has returned to baseline in the CLN (Fig. 5). Consistently, the absolute number of Kb/SIINFEKL+ CD8+ T cells increased

3.3. DQ-OVA is internalized by CD11b/MAC-1+ cells in the CNS following IC antigen delivery In previous experiments, we showed that DQ-OVA can be processed locally in the CNS. To further confirm localized antigen processing, we analyzed potential antigen-presenting cells, CD11b/MAC-1+ cells, in the brain. Brain sections (8 Am) were prepared, and CD11b/MAC-1+ cells in the brain were detected immunocytochemically using Cy5-conjugated anti-CD11b/MAC-1 antibodies. In the brain of naı¨ve mice, only a few scattered CD11b/ MAC-1+ cells are seen (data not shown). An intracerebral administration of DQ-OVA results in a significant increase in CD11b/MAC-1+ cell population in the cortical area ipsilateral to the injection and with the highest concentration at the injection site. Seven days after injection, CD11b/ MAC-1+ cells (blue color, also indicated by arrowheads in Fig. 4) accumulate densely in the injection site and are surrounded by DQ-OVA (green, also indicated by small arrows in Fig. 4). The CD11b/MAC-1+ cells (arrowheads in Fig. 4A) contain punctate DQ-OVA (green granules) in their cytoplasm, indicating internalization of DQ-OVA. The density of DQ-OVA is significantly reduced 4 weeks later at the injection site, but many CD11b/MAC-1+ cells remain localized to the injection site (Fig. 4B). Many of these CD11b/ MAC-1+ cells are co-stained with DQ-OVA, although a few single positive CD11b/MAC-1+ cells (arrows in Fig. 4B) are also observed. These data indicate that protein antigen, DQOVA, can be internalized by CD11b/MAC-1+ cells and also provide the first evidence of the long-term retention of brain-derived protein antigens in the CNS. 3.4. Antigen-specific CD8+ T cells in the periphery respond to intracerebrally infused OVA antigen We have shown that a portion of DQ-OVA drains into the CLN after intracerebral administration, and a portion of administered DQ-OVA is recognized by CD11b/MAC-1+ cells and retained for up to 4 weeks in the brain parenchyma. Whether the drainage of the processed DQ-OVA in the CLN or the retained DQ-OVA in the brain parenchyma can initiate an antigen-specific CD8+ T cell response was our next question. We investigated the presence of antigenspecific CD8+ T cells in the CLN and CNS following ICOVA delivery. Lymphocytes isolated from the brain and the CLN were stained with SIINFEKL-APC to detect CD8+ T cells specific for SIINFEKL peptide. Flow cytometric

Fig. 5. Accumulation of antigen-specific (Kb/SIINFEKL+) CD8+ T cells in the CLN and the brain after an intracerebral antigen delivery. Mice received intracerebral OVA immunization(s) at day 0 and day 14 as indicated by the arrows in B and were perfused on days 3, 7, 21, 28 or 42 after the primary injection. Lymphocytes were isolated from the CLN and the brain and stained with Kb/SIINFEKL-APC tetramer and antibodies to MAC-1, B220 and CD8, followed by flow cytometric analysis. For each brain, the total number of isolated lymphocytes was counted, and the percentage and the absolute number of antigen-specific cells that were CD8a and SIINFEKL tetramer+ but MAC-1 and B220 negative were calculated and analyzed statistically. (A) Dot plots shown are gated on CD8+, CD45R/B220 and CD11b/MAC-1 lymphocytes. The left value in each plot is the percentage of CD8+ cells within the CD45R/B220 and CD11b/MAC-1 lymphocyte population, and the right value is the percentage of SIINFEKL-APC+ cells within the CD8+ T cell population. (B) Changes in the percentage of SIINFEKL-APC+/CD8+ T cells following IC OVA immunization. Each point represents the mean percentage F S.E. of SIINFEKL-APC+ cells within CD8+ T cell populations derived from at least four mice.

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

95

Fig. 6. Co-localization of CD8+ T cells and SIINFEKL+ (OVA-specific) cells in the brain of experimental animals 7 days following a single intracerebral OVA injection. Mice were sacrificed, and their brains were dissected and prepared for sectioning. (A, B) A brain section (50 Am) containing the IC-injection site was double-stained with anti-CD8 (Cy5, blue in A) and Kb/SIINFEKL tetramer (PE, red in B). The majority of CD8+ cells (blue) that accumulate densely at the injection site are also Kb/SIINFEKL+ (red). (C) A cryostat section of a different brain (8 Am) containing the IC-injection site was stained with H&E. Large arrow: injection site; small arrows: lymphocytes. Scale bar: 150 Am in A and B; 200 Am in C.

10-fold in the CLN 3 days after IC-OVA injection and over 180-fold in the brain 7 days after injection. Comparatively, an IC-PBS injection did not induce any increase in the number of CD8+ T cells or SIINFEKL+ T cells in both the CLN and the brain (data not shown), suggesting that this effect was not merely due to damage of the BBB by injection. Kb/SIINFEKL+ CD8+ T cells in the brain were further demonstrated by immunocytochemical staining experiments. Fig. 6 shows that the majority of cells at the injection site are CD8 (blue) and Kb/SIINFEKL (red) double positive 7 days after IC immunization (Fig. 6A,B). Similarly, H&E staining shows the accumulation of infiltrated lymphocytes at the injection site (Fig. 6C), but not in other areas of the brain, including those adjacent to the injection (data not shown). These data indicate that antigen-specific CD8+ T cells traffic into the brain towards their specific antigens.

4. Discussion The present study was designed to evaluate processing and distribution of a soluble protein antigen in the CNS following a direct IC injection and the systemic antigenspecific immune response in the periphery following this injection. Our data show that intracerebrally infused soluble protein antigens can be processed locally and distributed preferentially in the brain parenchyma. The antigen can also drain into the CLN and activate an antigen-specific CD8+ T cell response in the periphery, resulting in recruitment of antigen-specific CD8+ T cells into the brain. Localization, efflux and immunologic processing of brain-derived protein antigens are important factors in the initiation and regulation of immune response in the CNS. There is a great amount of information about the efflux of

these antigens into peripheral lymphoid organs and their capability to initiate a significant humoral response (Gordon et al., 1992; Harling-Berg et al., 1999; Hochwald et al., 1988; Knopf et al., 1995). However, there is a lack of information concerning the distribution, immunologic processing and long-term retention of these antigens in the brain parenchyma following an intracerebral infusion. In the present study, we monitored the distribution of immunologically processed DQ-OVA, a protein antigen, following intracerebal administration in C57BL/6 mice. Emitting detectable fluorescence only upon proteolysis, DQ-OVA gives a unique opportunity to investigate the drainage, distribution and retention of the processed protein antigen both in the CNS and in the peripheral lymphoid organ. In this study, the processed DQ-OVA is seen at the injection site in the brain and in the afferent lymphatic vessel of the CLN as early as 2 h after a single IC administration (Figs. 1 and 3). This is the first observation that brain-derived protein antigen can be proteolyzed locally in the CNS, in addition to draining into the CLN. It has been demonstrated previously that intracerebrally injected protein antigens rapidly drain into the CLN (Harling-Berg et al., 1999; Qing et al., 2000; Yamada et al., 1991). Surprisingly, our data showed that a significant amount of the antigen is retained in the CNS and is taken up by CD11b/MAC-1+ cells (Fig. 4). Whether these CD11b/ MAC-1+ cells play a role in antigen presentation in the CNS is controversial. It has been suggested that CD11b/MAC-1+ cells in the perivascular space of the brain play an important role as first-line CNS scavengers and as antigen-presenting cells (Aloisi et al., 2000; Graeber et al., 1992; Weller, 1998). It has also been reported that brain resident microglia can progressively acquire a macrophage phenotype, upregulate adhesion molecules, and express co-stimulatory molecules (such as CD40, CD80 and CD86) that promote APC-

96

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

dependent T cell activation in response to CNS inflammation and neuronal damage (De Simone et al., 1995; Gerritse et al., 1996; Issazadeh et al., 1998; Kreutzberg, 1996; Li et al., 1996). In the present study, only limited inflammation and damage was observed at the injection site (Fig. 6), where CD11b/MAC-1+ cells accumulated. As presented in Fig. 4, these CD11b/MAC-1+ cells (dark blue) contain DQOVA in their cytoplasm (light blue granules), or in contact with DQ-OVA at their surfaces, suggesting that they are involved in antigen presentation. However, we cannot exclude either the possibility that some CD11b/MAC-1 cells may be also involved in antigen presentation, or the possibility that a portion of DQ-OVA can be degraded, nonspecifically, by enzymes located in extracellular space of the CNS. We also observed that, in addition to being processed in situ, antigen is also distributed preferentially along the ipsilateral external capsule. It is generally accepted that within the cranium interstitial fluid can flow through the space between nerve fiber tracts in white matter (HarlingBerg et al., 1999). When India ink is injected deeply into the rat brain parenchyma, carbon particles are distributed diffusely through the white mater, but spread selectively along the perivascular space in the grey matter (Ohata et al., 1990; Zhang et al., 1992). It was also noted that 1 week after IC injection many carbon particles are taken up by macrophages in the white matter and in the injection site, but the data were not shown (Zhang et al., 1992). In this study, we used a soluble protein, DQ-OVA, and the injection was restricted to the cortex, the grey matter. No processed DQOVA accumulates along any perivascular space at any time point that we have examined up to 7 days after IC injection. Although the processed antigen is also seen in the white matter, it is restricted to only the ipsilateral external capsule that is ventrolateral to the injection. No free diffusion is detected in any other brain area, including the corpus callosum (white matter, dorsal – medial to the injection), anterior commissure (white matter, distal to the injection), and the cortical area (grey matter) adjacent to the injection, except the remnant of DQ-OVA within the injection track. Interestingly, CD11b/MAC-1+ cells containing internalized DQ-OVA accumulate along the injection site (Fig. 4), but are barely seen in the external capsule. These data suggest that there may be a specific draining pathway for brainderived, soluble protein antigen along the external capsule. Whether CD11b/MAC-1+ cells are required for this specific drainage is still an open question. It is well known that the drainage of brain-derived protein antigens can generate antibody-secreting plasma cells in the CLN, resulting in the appearance of antibodies in blood (Gordon et al., 1992; Hochwald et al., 1988; Knopf et al., 1995; Widner et al., 1988). Although the drainage of brain-derived protein antigens in the CLN fails to induce a specific DTH response (Harling-Berg et al., 1999), it elicits an increase in the number of antigenspecific T cells in the CSF (Qing et al., 2000). In this

study, OVA also drains into the CLN within 2 h after IC delivery, and the number of antigen-specific CD8+ T cells increases significantly in the CLN 3 days later. By contrast, although the IC injected OVA can be processed locally within 2 h, a bulk recruitment of antigen-specific CD8+ T cells occurs 7 days later when the number of antigen-specific CD8+ T cells has dropped to baseline in the CLN. These data suggest that drainage of brain-derived protein antigens can induce an antigen-specific T cell response both in the periphery and in the CNS, although the peripheral response to the secondary IC boosting seems different from ones induced by a regular peripheral immunization with an adjuvant. The number of antigen-specific CD8+ T cells in the cervical lymph node is actually lower after the boost compared to the primary inoculation. In this study, OVA antigen was injected into the CNS without any adjuvant. It has been suggested that when APCs are activated without adjuvants or microbial products to indicate infectious-nonself (Janeway, 1992) or danger (Matzinger, 1994), primed T cells are not able to engage co-stimulation-dependent cell functions such as the CD40 CD40L/IL12/IFN-g cascade, and the local antigen recognition does not result in the development of an inflammatory response. Instead, self-limiting production of cytokines occurs, perhaps followed by T cell apoptosis (Bauer et al., 1998; Lehmann et al., 1998; Pender et al., 1991). Although this type of limited cell response has been detected in the periphery following a noninfectious peripheral immunization, we cannot exclude the possibility of such a process in the brain following our noninfectious intracerebral immunization. Also, we previously demonstrated (Hofstetter et al., 2003) that a different cytokine expression pattern might be engaged in the primed T cells during sterile CNS inflammation. This is similar to the outcome of stimulating T cells with altered peptide ligands (Brocke et al., 1996). This different cytokine pattern might lead to T cell or APC apoptosis in the CLN or other peripheral tissues. Origin of significantly increased OVA-specific CD8+ T cells in the brain following intracerebral delivery of protein antigens is another interesting question. In this study a significant increase of OVA-specific CD8+ T cells appears first in the periphery, suggesting that the initiation of the response is most likely to occur in the periphery, and activated antigen-specific CD8+ T cells preferentially traffic into the brain. Since the number of CD8+ T cells in the CNS (f 8  103/brain) is significantly lower than that in the peripheral lymphoid tissue (>1  106/CLN or >5  106/ spleen) following IC-OVA immunization (data not shown), the increased percentage of antigen-specific T cells/CD8+ T in the CNS may indicate antigen-specific recruitment of T cells, but does not necessarily represent a significantly higher number of cells compared to the periphery. Whether primed antigen-specific CD8+ T cells proliferate in the brain in response to local antigen presentation is still an open question, and further investigation is underway in our laboratory.

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

In our experimental model, an IC injection, similar to a cerebral stab or neural cell transplantation, can result in trauma and limited damage in the BBB as reported previously (Dunnett et al., 1997; Wenkel et al., 2000). It could be argued that disruption of the BBB allows T cell recruitment into the brain. However, in this study, when PBS instead of OVA was injected into the mouse brain, no increase in the CD8+ T cell recruitment is detected. Consistently, there is also evidence from other studies that systemic immune deviation in the brain is not dependent on BBB integrity (Wenkel et al., 2000). Similarly, cannula implantation into the mouse brain followed by 2 weeks rest to allow the BBB to reestablish itself before intraventricular injection of protein antigens also allows rapid drainage of the antigen from the ventricle to the CLN and induces antigen-specific T cell response in the CSF (Qing et al., 2000). Altogether, our findings indicate that protein antigens delivered into the CNS can be processed locally in the brain parenchyma and distributed preferentially along the external capsule ipsilateral to the injection site. Furthermore, we demonstrate the long-term retention of the protein antigens by local CD11b/MAC-1+ cells in the brain parenchyma following IC injection. Finally, we identify that the drainage of protein antigens from the brain parenchyma results in the traffic of antigen-specific CD8+ T cells from the periphery to the CNS. The peripheral activation of antigen-specific T cells in response to brain-derived protein antigens may have a great significance in the initiation of CNS inflammatory diseases. These data are compatible with a model in which brain-derived antigens translocate to the cervical lymph nodes where T cell priming may occur. The processed antigen in the brain may be instrumental to reactivate and retain the specific CD8+ T cells in the brain. To substantiate this mechanism, further investigations are required.

Acknowledgements The authors thank Khen Macvilay for his flow cytometric expertise, Toshi Kinoshita for tissue processing for confocal microscopy and immunohistochemistry, and Dr. Laura Hogan and Dominic O. Co for critical reading of the manuscript and valuable discussions.

References Aloisi, F., Ria, F., Adorini, L., 2000. Regulation of T-cell responses by CNS antigen-presenting cells: different roles for microglia and astrocytes. Immunol. Today 21, 141 – 147. Bauer, J., Bradl, M., Hickley, W.F., Forss-Petter, S., Breitschopf, H., Linington, C., Wekerle, H., Lassmann, H., 1998. T-cell apoptosis in inflammatory brain lesions: destruction of T cells does not depend on antigen recognition. Am. J. Pathol. 153, 715 – 724. Boulton, M., Flessner, M., Armstrong, D., Hay, J., Johnston, M., 1998. Determination of volumetric cerebrospinal fluid absorption into extracranial lymphatics in sheep. Am. J. Physiol. 274, R88 – R96.

97

Brabb, T., von Dassow, P., Ordonez, N., Schnabel, B., Duke, B., Goverman, J., 2000. In situ tolerance within the central nervous system as a mechanism for preventing autoimmunity. J. Exp. Med. 192, 871 – 880. Bradbury, M.W., Cserr, H.F., Westrop, R.J., 1981. Drainage of cerebral interstitial fluid into deep cervical lymph of the rabbit. Am. J. Physiol. 240, F329 – F336. Brocke, S., Gijbels, K., Allegretta, M., Ferber, I., Piercy, C., Blankenstein, T., Martin, R., Utz, U., Karin, N., Mitchell, D., Veromaa, T., Waisman, A., Gaur, A., Conlon, P., Ling, N., Fairchild, P.J., Wraith, D.C., O’Garra, A., Fathman, C.G., Steinman, L., 1996. Treatment of experimental encephalomyelitis with a peptide analogue of myelin basic protein. Nature 379, 343 – 346. Cserr, H.F., Ostrach, L.H., 1974. Bulk flow of interstitial fluid after intracranial injection of blue dextran. Exp. Neurol. 45, 50 – 60. Cserr, H.F., DePasquale, M., Harling-Berg, C.J., Park, J.T., Knopf, P.M., 1992. Afferent and efferent arms of the humoral immune response to CSF-administered albumins in a rat model with normal blood – brain barrier permeability. J. Neuroimmunol. 41, 195 – 202. De Simone, R., Giampaolo, A., Giometto, B., Gallo, P., Levi, G., Peschle, C., Aloisi, F., 1995. The costimulatory molecule B7 is expressed on human microglia in culture and in multiple sclerosis acute lesions. J. Neuropathol. Exp. Neurol. 54, 175 – 187. Dunnett, S.B., Kendall, A.L., Watts, C., Torres, E.M., 1997. Neuronal cell transplantation for Parkinson’s and Huntington’s diseases. Br. Med. Bull. 53, 757 – 776. Fabry, Z., Waldschmidt, M.M., Hendrickson, D., Keiner, J., Love-Homan, L., Takei, F., Hart, M.N., 1992. Adhesion molecules on murine brain microvascular endothelial cells: expression and regulation of ICAM-1 and Lgp 55. J. Neuroimmunol. 36, 1 – 11. Gerritse, K., Laman, J.D., Noelle, R.J., Aruffo, A., Ledbetter, J.A., Boersma, W.J., Claassen, E., 1996. CD40 CD40 ligand interactions in experimental allergic encephalomyelitis and multiple sclerosis. Proc. Natl. Acad. Sci. U. S. A. 93, 2499 – 2504. Ghersi-Egea, J.F., Gorevic, P.D., Ghiso, J., Frangione, B., Patlak, C.S., Fenstermacher, J.D., 1996. Fate of cerebrospinal fluid-borne amyloid beta-peptide: rapid clearance into blood and appreciable accumulation by cerebral arteries. J. Neurochem. 67, 880 – 883. Gordon, L.B., Knopf, P.M., Cserr, H.F., 1992. Ovalbumin is more immunogenic when introduced into brain or cerebrospinal fluid than into extracerebral sites. J. Neuroimmunol. 40, 81 – 87. Graeber, M.B., Streit, W.J., Buringer, D., Sparks, D.L., Kreutzberg, G.W., 1992. Ultrastructural location of major histocompatibility complex (MHC) class II positive perivascular cells in histologically normal human brain. J. Neuropathol. Exp. Neurol. 51, 303 – 311. Harling-Berg, C.J., Park, T.J., Knopf, P.M., 1999. Role of the cervical lymphatics in the Th2-type hierarchy of CNS immune regulation. J. Neuroimmunol. 101, 111 – 127. Hochwald, G.M., Van Driel, A., Robinson, M.E., Thorbecke, G.J., 1988. Immune response in draining lymph nodes and spleen after intraventricular injection of antigen. Int. J. Neurosci. 39, 299 – 306. Hofstetter, H.H., Sewell, D.L., Liu, F., Sandor, M., Forsthuber, T., Lehmann, P.V., Fabry, Z., 2003. Autoreactive T cells promote post-traumatic healing in the central nervous system. J. Neuroimmunol. 134, 25 – 34. Hogquist, K.A., Jameson, S.C., Heath, W.R., Howard, J.L., Bevan, M.J., Carbone, F.R., 1994. T cell receptor antagonist peptides induce positive selection. Cell 76, 17 – 27. Ichimura, T., Fraser, P.A., Cserr, H.F., 1991. Distribution of extracellular tracers in perivascular spaces of the rat brain. Brain Res. 545, 103 – 113. Issazadeh, S., Navikas, V., Schaub, M., Sayegh, M., Khoury, S., 1998. Kinetics of expression of costimulatory molecules and their ligands in murine relapsing experimental autoimmune encephalomyelitis in vivo. J. Immunol. 161, 1104 – 1112. Janeway Jr., C.A., 1992. The immune system evolved to discriminate infectious nonself from noninfectious self. Immunol. Today 13, 11 – 16. Kelly, J.M., Sterry, S.J., Cose, S., Turner, S.J., Fecondo, J., Rodda, S., Fink, P.J., Carbone, F.R., 1993. Identification of conserved T cell receptor

98

C. Ling et al. / Journal of Neuroimmunology 141 (2003) 90–98

CDR3 residues contacting known exposed peptide side chains from a major histocompatibility complex class I-bound determinant. Eur. J. Immunol. 23, 3318 – 3326. Knopf, P.M., Cserr, H.F., Nolan, S.C., Wu, T.Y., Harling-Berg, C.J., 1995. Physiology and immunology of lymphatic drainage of interstitial and cerebrospinal fluid from the brain. Neuropathol. Appl. Neurobiol. 21, 175 – 180. Kreutzberg, G.W., 1996. Microglia: a sensor for pathological events in the CNS. Trends Neurosci. 19, 312 – 318. Lehmann, P.V., Targoni, O.S., Forsthuber, T.G., 1998. Shifting T-cell activation thresholds in autoimmunity and determinant spreading. Immunol. Rev. 164, 53 – 61. Li, H., Cuzner, M.L., Newcombe, J., 1996. Microglia-derived macrophages in early multiple sclerosis plaques. Neuropathol. Appl. Neurobiol. 22, 207 – 215. Matzinger, P., 1994. Tolerance, danger, and the extended family. Annu. Rev. Immunol. 12, 991 – 1045. Oehmichen, M., Huber, H., 1978. Supplementary cytodiagnostic analyses of mononuclear cells of the cerebrospinal fluid using cytological markers. J. Neurol. 218, 187 – 196. Ohata, K., Marmarou, A., Povlishock, J.T., 1990. An immunocytochemical study of protein clearance in brain infusion edema. Acta Neuropathol. (Berl.) 81, 162 – 177. Pender, M.P., Nguyen, K.B., McCombe, P.A., Kerr, J.F., 1991. Apoptosis in the nervous system in experimental allergic encephalomyelitis. J. Neurol. Sci. 104, 81 – 87.

Perry, V.H., 2000. Persistent pathogens in the parenchyma of the brain. J. Neurovirology 6 (Suppl. 1), S86 – S89. Qing, Z., Sewell, D., Sandor, M., Fabry, Z., 2000. Antigen-specific T cell trafficking into the central nervous system. J. Neuroimmunol. 105, 169 – 178. Rodriguez-Peralta, L.A., 1968. Hematic and fluid barriers of the retina and vitreous body. J. Comp. Neurol. 132, 109 – 124. Szentistvanyi, I., Patlak, C.S., Ellis, R.A., Cserr, H.F., 1984. Drainage of interstitial fluid from different regions of rat brain. Am. J. Physiol. 246, F835 – F844. Weller, R.O., 1998. Pathology of cerebrospinal fluid and interstitial fluid of the CNS: significance for Alzheimer disease, prion disorders and multiple sclerosis. J. Neuropathol. Exp. Neurol. 57, 885 – 894. Wenkel, H., Streilein, J.W., Young, M.J., 2000. Systemic immune deviation in the brain that does not depend on the integrity of the blood – brain barrier. J. Immunol. 164, 5125 – 5131. Widner, H., Moller, G., Johansson, B.B., 1988. Immune response in deep cervical lymph nodes and spleen in the mouse after antigen deposition in different intracerebral sites. Scand. J. Immunol. 28, 563 – 571. Yamada, S., DePasquale, M., Patlak, C.S., Cserr, H.F., 1991. Albumin outflow into deep cervical lymph from different regions of rabbit brain. Am. J. Physiol. 261, H1197 – H1204. Zhang, E.T., Richards, H.K., Kida, S., Weller, R.O., 1992. Directional and compartmentalised drainage of interstitial fluid and cerebrospinal fluid from the rat brain. Acta Neuropathol. (Berl.) 83, 233 – 239.