Neutrophils mediate blood-spinal cord barrier disruption in demyelinating neuroinflammatory diseases.

Neutrophils Mediate Blood−Spinal Cord Barrier Disruption in Demyelinating Neuroinflammatory Diseases This information is current as of August 13, 2014.

J Immunol published online 21 July 2014 http://www.jimmunol.org/content/early/2014/07/25/jimmun ol.1400401

Supplementary Material

http://www.jimmunol.org/content/suppl/2014/07/18/content.1400401. DCSupplemental.html

Subscriptions

Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions

Permissions Email Alerts

Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

Downloaded from http://www.jimmunol.org/ at York Univ Scott Lib/Serials Sec on August 13, 2014

Benoit Aubé, Sébastien A. Lévesque, Alexandre Paré, Émilie Chamma, Hania Kébir, Roser Gorina, Marc-André Lécuyer, Jorge I. Alvarez, Yves De Koninck, Britta Engelhardt, Alexandre Prat, Daniel Côté and Steve Lacroix

Published July 25, 2014, doi:10.4049/jimmunol.1400401 The Journal of Immunology

Neutrophils Mediate Blood–Spinal Cord Barrier Disruption in Demyelinating Neuroinflammatory Diseases Benoit Aube´,*,†,‡,x Se´bastien A. Le´vesque,*,† Alexandre Pare´,*,† E´milie Chamma,‡,x Hania Ke´bir,{ Roser Gorina,‖ Marc-Andre´ Lećuyer,{ Jorge I. Alvarez,{ Yves De Koninck,‡ Britta Engelhardt,‖ Alexandre Prat,{ Daniel Coˆte´,‡,x and Steve Lacroix*,†

L

eukocyte entry into the healthy mammalian CNS is strictly controlled by the blood–brain and the blood–spinal cord barriers (BBB and BSCB, respectively), complex vascular gatekeepers that maintain CNS homeostasis by regulating the passage of soluble compounds and leukocytes from the periphery into the parenchyma (1, 2). Dysfunction of the BBB and BSCB is *Centre de Recherche du Centre Hospitalier Universitaire de Que´bec–Centre Hospitalier de l’Universite´ Laval, Quebec, Quebec G1V 4G2, Canada; †De´partement de Me´decine Molećulaire, Faculte´ de Me´decine, Universite´ Laval, Quebec, Quebec G1V 0A6, Canada; ‡Centre de Recherche de l’Institut Universitaire en Sante´ Mentale de Que´bec, Universite´ Laval, Quebec, Quebec G1J 2G3, Canada; xCentre d’Optique, Photonique et Laser, Universite´ Laval, Quebec, Quebec G1V 0A6, Canada; {Unite´ de Neuroimmunologie, Centre d’Excellence en Neuromique, Centre de Recherche du Centre Hospitalier de l’Universite´ de Montreál, Faculte´ de Me´decine, Universite´ de Montreál, Montreál, Quebec H3C 3J7, Canada; and ‖Theodor Kocher Institute, University of Bern, 3012 Bern, Switzerland ORCID: 0000-0003-0781-6371 (S.L.). Received for publication February 11, 2014. Accepted for publication June 27, 2014. This work was supported by an Emerging Team grant from the Canadian Institutes of Health Research (to Y.D.K., A. Prat, D.C., and S.L.), an operating grant from the Multiple Sclerosis Society of Canada (to S.L.), and Swiss National Science Foundation Grant 133092 (to B.E.). Salary support was provided by the Multiple Sclerosis Society of Canada (to S.A.L., A. Pare´, and M.A.L.), the Canada Research Chairs program (to D.C.), and the Fonds de Recherche du Que´bec en Sante´ (to A. Prat and S.L.). Address correspondence and reprint requests to Dr. Steve Lacroix, Centre de Recherche du Centre Hospitalier Universitaire de Que´bec—Centre Hospitalier Universite´ Laval, 2705 Boulevard Laurier, Quebec, Quebec G1V 4G2, Canada. E-mail address: [email protected] The online version of this article contains supplemental material. Abbreviations used in this article: Alexa-594, Alexa Fluor 594 hydrazide sodium salt; BBB, blood–brain barrier; BSCB, blood–spinal cord barrier; d.p.i., day(s) postimmunization; EAE, experimental autoimmune encephalomyelitis; EC, endothelial cell; ko, knockout; NaFl, fluorescein sodium salt; PMN, polymorphonuclear neutrophil; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; NMO, neuromyelitis optica; PTX, pertussis toxin. Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1400401

a common feature of several neurologic conditions, including traumatic brain and spinal cord injury, stroke, neuromyelitis optica (NMO), and multiple sclerosis (MS) (3–5). In the animal model of MS, experimental autoimmune encephalomyelitis (EAE), the BSCB is compromised, thereby exposing the fragile CNS environment to the immune system and its cellular arsenal. This leads to immune cell invasion, formation of demyelinated lesions, and axonal damage (4, 6–8). Early breakdown of the BSCB in EAE has been widely documented, but the precise timing and triggers of this disruption are still a matter of debate (9–12). This has prompted us to look further into the kinetics of BSCB disruption and the putative cellular candidates involved in this process. The abundance of polymorphonuclear neutrophils (PMNs) and their capacity to be rapidly deployed to sites of inflammation make these effector cells particularly well suited to participate in inflammatory cascades that may result in early disruption of the BSCB, in the context of neuroinflammation. Indeed, the prompt activation of meningeal mast cells in EAE mice was shown to elicit sustained neutrophil recruitment, alter BSCB integrity, and promote subsequent leukocyte infiltration into the CNS (13). This is in agreement with data showing that mast cells control early neutrophil influx via the secretion of chemokines CXCL1 and CXCL2 (14). Interestingly, Kroenke et al. (15) showed that these same chemokines were upregulated in the brain and spinal cord of EAE mice when the disease was induced by adoptive transfer of IL-23–modulated T cells. This correlated with extensive inflammatory infiltrates predominantly composed of neutrophils. Segal and colleagues (16) also showed that CXCL1 and CXCL2 were abundantly transcribed in the spinal cord of naive mice injected with Th17 cells, shedding light on the involvement of these cells with the ELR+ CXC chemokine pathway in EAE. Analyses of mRNA levels in the spinal cord of EAE mice suggest


Disruption of the blood–brain and blood–spinal cord barriers (BBB and BSCB, respectively) and immune cell infiltration are early pathophysiological hallmarks of multiple sclerosis (MS), its animal model experimental autoimmune encephalomyelitis (EAE), and neuromyelitis optica (NMO). However, their contribution to disease initiation and development remains unclear. In this study, we induced EAE in lys-eGFP-ki mice and performed single, nonterminal intravital imaging to investigate BSCB permeability simultaneously with the kinetics of GFP+ myeloid cell infiltration. We observed a loss in BSCB integrity within a day of disease onset, which paralleled the infiltration of GFP+ cells into the CNS and lasted for ∼4 d. Neutrophils accounted for a significant proportion of the circulating and CNS-infiltrating myeloid cells during the preclinical phase of EAE, and their depletion delayed the onset and reduced the severity of EAE while maintaining BSCB integrity. We also show that neutrophils collected from the blood or bone marrow of EAE mice transmigrate more efficiently than do neutrophils of naive animals in a BBB cell culture model. Moreover, using intravital videomicroscopy, we demonstrate that the IL-1 receptor type 1 governs the firm adhesion of neutrophils to the inflamed spinal cord vasculature. Finally, immunostaining of postmortem CNS material obtained from an acutely ill multiple sclerosis patient and two neuromyelitis optica patients revealed instances of infiltrated neutrophils associated with regions of BBB or BSCB leakage. Taken together, our data provide evidence that neutrophils are involved in the initial events that take place during EAE and that they are intimately linked with the status of the BBB/BSCB. The Journal of Immunology, 2014, 193: 000–000.

2

NEUTROPHILS DISRUPT BSCB INTEGRITY IN NEUROINFLAMMATION

Materials and Methods Animals All animal procedures were approved by the Animal Welfare Committees of Laval University, the Centre Hospitalier de l’Universite´ de Montreál, and the Veterinary Office of the Kanton Bern (Bern, Switzerland), in strict accordance with guidelines of the Canadian and Swiss Council on Animal Care. Female C57BL/6 mice were purchased from Charles River Laboratories (St. Constant, QC, Canada) at 8–10 wk of age. Breeders of the lys-eGFP-ki strain were obtained from Dr. Gregory Dekaban (Robarts Institute, London, ON, Canada), with authorization from Dr. Thomas Graf (Barcelona, Spain). Both heterozygous male and female mice of this strain were used for EAE induction and imaging experiments when they were .8 wk of age. No differences were observed in terms of induction rate and disease outcome between either gender or strain used. IL-1R1–knockout (ko) mice on a C57BL/6 background and their wild-type littermates were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice had ad libidum access to food and water.

EAE induction and clinical evaluation EAE was induced by s.c. injections of 100 mg myelin oligodendrocyte glycoprotein (MOG)35–55 (MEVGWYRSPFSRVVHLYRNGK; AnaSpec, Fremont, CA) in CFA (IFA containing 4 mg/ml heat-inactivated Mycobacterium tuberculosis H37Ra; BD Biosciences, Mississauga, ON, Canada). An i.v. injection of 200 ng pertussis toxin (PTX; List Biological Laboratories, Campbell, CA) was also administered on days 0 and 2 of the immunization. The severity of EAE was scored daily using a grading scale of 0–5, following recommendations of Stromnes and Goverman (36): 0, unaffected; 0.5, partially limp tail; 1, paralyzed tail; 2, hindlimb paresis and loss in coordinated movement; 2.5, one hindlimb paralyzed; 3, both hindlimbs paralyzed; 3.5, hindlimbs paralyzed and weakness in forelimbs; 4, forelimbs paralyzed; and 5, moribund/death. Mice displaying a score of $2 received daily manual bladder evacuation, and those with a score .3 received daily s.c. injections of sterile saline.

Surgical procedure Optical access to the spinal cord was achieved by means of a single laminectomy at the vertebral L1 level, except for the experiment in which we studied in real time the adhesion of systemically infused neutrophils to the cervical (C5 level) spinal cord microvasculature (see below). Isoflurane was used to induce anesthesia (4% v/v) and subsequently throughout the surgery and imaging session (2% v/v in oxygen). Briefly, a midline incision ∼1 cm in length was performed on the shaved back of mice. The skin was then retracted to expose the muscles and tissues covering the vertebra of interest, after which the former were carefully separated from the vertebral column. Upon removal of the dorsal aspect of the vertebra, the dura mater was carefully removed. Pilot experiments revealed this step was necessary in animals displaying heavy clinical burden, because many GFP+ cells accumulated in the meninges and prevented acceptable image clarity under the microscope. We confirmed that dura removal did not affect either BSCB leakage or cellular infiltration during a 90-min observation period (data no shown). Animals were administered 1% vascular tracer via the tail vein to visualize the status of the BSCB. C57BL/6 mice were injected with 376 Da fluorescein sodium salt (NaFl; Sigma-Aldrich Canada, Oakville, ON, Canada, catalog No. F6377), and lys-eGFP-ki mice were injected with 760 Da Alexa Fluor 594 hydrazide, sodium salt (Alexa-594; Life Technologies, Carlsbad, CA, catalog No. A-10438), and both tracers were diluted in sterile saline. Alexa-594 was used because of its size comparable to NaFl and its emission spectrum not overlapping with GFP as opposed to the former. Following surgery, mice were placed on a custom-made stabilization device on a heating pad maintained at 37˚C. Gel-Seal (GE Healthcare, Baie d’Urfe´, QC, Canada) was applied on the tissue surrounding the spinal cord to create a watertight cavity filled with sterile HBSS. Once the imaging session was completed (∼50 min), the surgical area was carefully cleaned of Gel-Seal. The muscular layers were then sutured and the cutaneous layers stapled as before (37), after which mice received 150 ml sterile saline s.c. and were placed in their cage for recovery.

Two-photon intravital microscopy and data acquisition We used an Olympus FV1000 MPE microscope equipped with a Mai Tai DeepSee pulsed laser (Spectra-Physics; Newport, Santa Clara, CA) for all intravital imaging experiments. Two-photon excitation was generated at 930 nm for NaFl in C57BL/6 mice and 840 nm for Alexa-594 in lys-eGFP-ki mice. Previous independent experiments confirmed that 840 nm was optimally providing simultaneous excitation of both Alexa-594 and GFP without spectral overlap. Of note, laser power was kept to a minimum to avoid photodamage, that is, ∼5–10 mW at the sample. Imaging was performed with an Olympus Ultra 325 MPE water immersion objective (1.05 numerical aperture, working distance of 2 mm) at a 512 3 512 resolution, using a pixel dwell time of 2 ms. The frame rate thus corresponds to ∼2 frames/s. For every animal, all the visible blood vessels were imaged during the imaging session. In general, we observed that leakage occurred between 10 and 30 min following tracer administration. Therefore, we made sure to sample all vessels within this time frame. Interestingly, no noticeable changes in BSCB integrity were apparent after 30 min. Because animals were not sacrificed following imaging sessions, the latter did not last .60 min. For time restriction considerations, the dorsal vein was thus partially imaged; its sole purpose was to provide spatial context for subsequent reconstruction of the vascular arborization. Images were acquired as z-stacks with 2- to 4-mm increments, after which they were exported in Tiff format for analysis. Time-lapse movies were processed with the Intravital imaging toolbox using ImageJ (37).

Intravital images analysis Analyses were carried out on maximal intensity projections using ImageJ 1.46n (Wayne Rasband, National Institutes of Health, Bethesda, MD). For


that neutrophils are among the first inflammatory cells recruited into the CNS, which is corroborated by studies showing the presence of PMNs in the meninges before the onset of clinical symptoms (13, 17, 18). PMNs are also increasingly recognized as having pivotal functions in driving inflammatory processes within the target organ, extending their role beyond that of bystander cells of the adaptive immune response (19–21). Recent findings present neutrophils as potent modulators of dendritic cell recruitment and function, which themselves are responsible for Ag presentation to encephalitogenic Th1 and Th17 cells following the latter’s transmigration from the perivascular space into the parenchyma (22– 27). The activity of IFN-g–producing Th1 lymphocytes is suppressed by neutrophils in the CNS of EAE mice, and neutrophilderived myeloperoxidase was shown to inhibit dendritic cell activation, therefore dampening T cell–driven inflammation (28, 29). Aside from their capacity to secrete a vast array of proinflammatory mediators, neutrophils are now recognized for their ability to prime Ag-specific Th1 and Th17 responses (30, 31). It was also reported that the depletion of circulating granulocytes led to a marked reduction in the number of relapses, attenuated disease severity, and abrogated the increase in BBB permeability typically observed during EAE in SJL mice (16, 32). Collectively, these data suggest an important role for neutrophils in the events that occur early in the course of EAE, among which BSCB breakdown is a crucial component warranting closer examination. Simultaneous analyses of the time course of immune cell recruitment into the CNS and BSCB disruption are hampered in EAE by the paucity of tools available for tracking cellular populations with minimal disturbance of their physiological environment. Intravital optical imaging enables sampling/monitoring of a substantial volume of tissue within the anatomical and functional cell microenvironment (33–35). In this study, we performed intravital imaging in lys-eGFP-ki mice to investigate the kinetics of BSCB disruption and CNS myeloid cell infiltration during active EAE. Our data show that leakage of small fluorescent tracers (,800 Da) across the BSCB precedes the onset of neurologic deficits and is transient. Interestingly, the temporal pattern of infiltration of LysM+ cells coincides with an increase in BSCB permeability. In the present study, we demonstrate that these infiltrated GFP+ cells are majorly composed of neutrophils and that their depletion leads to a marked decrease of vascular leakage in EAE mice as compared with control animals. Moreover, and in accordance with previous reports, we found the severity of the disease to be reduced in neutrophildepleted animals (16, 27, 32). Finally, immunohistochemical analysis of tissue sections from MS brain and NMO spinal cord reveals that neutrophils are closely apposed to the CNS vasculature, in areas associated with increased BBB/BSCB leakage.

The Journal of Immunology

Immunohistochemistry on murine EAE spinal cords Immunohistochemical analyses were performed on lys-eGFP-ki and C57BL/6 EAE mice that did not undergo surgery or imaging (with the exception of the experiment in Supplemental Fig. 2C, 2D). In brief, mice were overdosed with ketamine/xylazine and transcardially perfused with 4% paraformaldehyde (pH 7.4) in PBS. Spinal cords were dissected out, postfixed overnight at 4˚C, and then transferred for 1 d into PBS containing 20% sucrose. Cervical, thoracic, and lumbar spinal segments, corresponding respectively to spinal levels C4–6, T5–12, and L1–4, were isolated and cut in seven to nine series of 35-mm-thick coronal sections using a cryostat or microtome, as described before (38). Immunoperoxidase labeling of Ly6G and CD3 was used to visualize neutrophils and lymphocytes, respectively, in spinal cord tissue sections from EAE mice. Because none of the tested Abs against F4/80, Iba1, CD68, and Galectin-3 could successfully distinguish macrophages from activated microglia in immunohistochemistry, the infiltration of monocyte-derived macrophages was assessed using flow cytometry, taking advantage of the fact that the latter cells express higher levels of CD45 than do their microglial counterparts (39, 40). Primary Abs used in this study are from the following sources: rat anti-mouse Ly6G (1:2500, BD Biosciences), rat anti-mouse CD3 (1:500, BD Biosciences), rat anti-mouse F4/80 (1:200, AbD Serotec, Raleigh, NC), rabbit anti-mouse Iba1 (1:750, Wako Chemicals, Richmond, VA), rat anti-mouse CD68 (1:2500, AbD Serotec), and rat anti-mouse Galectin-3 (1:500, American Type Culture Collection, Manassas, VA). Primary Abs were detected using a biotinylated anti-rat or anti-rabbit secondary Ab (Vector Laboratories, Burlington, ON, Canada) in conjunction with an avidin-biotin-peroxidase amplification system (VectaStain ABC Kit, Vector Laboratories) and 3,3-diaminobenzidine. The spatial location of GFP+ cells with respect to the BSCB in lyseGFP-ki mice with EAE and lys-eGFP-ki mice treated with PTX was determined by confocal immunofluorescence labeling of the endothelial and parenchymal basement membranes using a rabbit anti–pan-laminin polyclonal Ab (1:500, Dako Canada, Burlington, ON, Canada). The identity of GFP+ cells as to whether they are neutrophils (or M1 monocytes) or not was verified using the rat anti-mouse 7/4 Ab (1:1000, AbD Serotec). A goat anti–IL-1R1 polyclonal Ab (1:100, R&D Systems, Minneapolis, MN) was used to identify cells expressing the IL-1R1 in the normal spinal cord. Alexa Fluor secondary Ab conjugates (1:200, Life Technologies) were used as secondary Abs, whereas DAPI (Life Technologies) was used for nuclear counterstaining. Immunofluorescence labeling was performed according to our previously published methods (41). Sections were observed and imaged on a IX81 inverted confocal microscope system equipped with Ar 488, HeNe1 543, and HeNe2 633 laser lines (Olympus Canada).

Blood sample preparation Flow cytometric analysis of whole blood was performed in both naive and immunized (at 8 d postimmunization [d.p.i.]) lys-eGFP-ki mice to identify leukocyte populations expressing GFP. Additionally, peripheral blood leukocytes were analyzed by flow cytometry 1 d prior to neutrophil depletion and 1, 2, and 3 d later to confirm the effectiveness and specificity of the depletion strategy in C57BL/6 mice. Blood was harvested with heparinized syringes via cardiac puncture in mice deeply anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg). Blood samples were immediately placed in EDTA-coated tubes (Sarstedt, Montreal, QC, Canada) in an agitator at room temperature until further processing. Cervical dislocation was performed to ensure euthanasia. Cell

suspensions were washed in PBS/2% FBS, after which they were incubated on ice with mouse Fc block (i.e., purified anti-mouse CD16/CD32; BD Biosciences) for 15 min to prevent nonspecific binding. They were then incubated on ice for 30 min with the following fluorochromeconjugated Abs (all from BD Biosciences, except where noted) for multicolor analysis: anti-CD45 PerCP (dilution: 1:100), anti-CD11b Alexa Fluor 700 (1:100), anti-7/4 PE (1:40, AbD Serotec), anti-Ly6C V450 (1:150), anti-Ly6G PE-Cy7 (1:100), and anti-B220 Alexa Fluor 488 (1:100). After a 5-min wash, erythrocytes were lysed (Beckman Coulter Canada, Mississauga, ON, Canada, catalog No. 6603152), and the Live/ Dead fixable yellow dead cell stain kit (Life Technologies) was used to distinguish live from dead cells. Cells were then washed twice in PBS without serum before data collection.

Spinal cords preparation Mice of the lys-eGFP strain were immunized for EAE and their lumbar spinal cord was harvested shortly before or at disease onset for flow cytometric analysis of infiltrated leukocytes to confirm the presence of GFP+ neutrophils, following our previously published method (42). Briefly, animals were transcardially perfused with cold HBSS to remove immune cells from the vasculature, their spinal cords were dissected out, and lumbar spinal cords were homogenized with a Potter–Elvehjem tissue grinder. The tissue was then digested with an enzymatic mixture (0.25% [w/v] collagenase type IV (Worthington Biochemical, Lakewood Township, NJ), 1 U/ml elastase (Worthington Biochemical), 0.025 U/ml DNAse I (Worthington Biochemical), 0.1 mg/ml Na-tosyl-L-lysine chloromethyl ketone hydrochloride (Sigma-Aldrich Canada), and 20 mM HEPES in HBSS) at 37˚C for 30 min. After a wash in HBSS, cells were filtered through a 70-mm nylon mesh cell strainer (BD Biosciences), centrifuged at 300 3 g for 10 min, and washed again with HBSS. Specific removal of myelin debris was performed by incubating single-cell suspensions with Myelin Removal Beads II (Miltenyi Biotec, Auburn, CA), according to the manufacturer’s instructions. For multicolor immunofluorescent labeling, cells were incubated on ice with mouse Fc block (i.e., purified anti-mouse CD16/CD32; BD Biosciences) for 15 min to prevent nonspecific binding, followed by labeling for 30 min on ice with the following fluorescently conjugated primary Abs (all from BD Biosciences except where noted): anti-CD45 PerCP (dilution, 1:100), anti-CD11b Alexa Fluor 700 (1:100), anti-Ly6C BD Horizon v450 (1:167), anti-Ly6G PE-Cy7 (1:100), anti-F4/80 allophycocyanin (1:25), anti-CD3e PE-CF594 (1:100), and anti-7/4 PE (1:40, AbD Serotec) (for a full description of some of these primary Abs, please refer to our published work; see Ref. 43). As for the blood samples, the Live/Dead fixable yellow dead cell stain kit (Life Technologies) was used to distinguish live from dead cells.

Flow cytometry and gating strategy Cells were first excluded from debris and erythrocytes according to their forward and side scatter characteristics, after which doublets and dead cells were discarded. Myeloid cells were identified as follows: neutrophils (CD45hi, B2202, CD11b+, 7/4dim, Ly6Cdim, Ly6G+), proinflammatory M1 monocytes (CD45hi, B2202, CD11b+, 7/4hi, Ly6Chi, Ly6G2), and antiinflammatory M2 monocytes (CD45hi, B2202, CD11b+, 7/42, Ly6Clo, Ly6G2). B cells were unaffected by the depletion of neutrophils (see Fig. 3B and Ref. 44), and therefore cell counts were normalized relative to this population, identified as (CD45hi, B220+, CD11b2, 7/42). Of note, the Alexa Fluor 488–conjugated anti-B220 was only used in mice of the C57BL/6 strain in neutrophil depletion experiments. Data were acquired for 1 min with an LSR II special order flow cytometer (BD Biosciences) and analyzed with FlowJo software (version 9.2; Tree Star, Ashland, OR). Fluorescence minus one controls were used to establish gating boundaries for every fluorochrome of the staining mixture (45).

Neutrophil depletion and anakinra treatment The 1A8 mAb was used to deplete neutrophils in EAE mice, whereas the 2A3 mAb served as the isotype control (both from BioXCell, West Lebanon, NH). The 1A8 clone was chosen over the anti–Gr-1 mAb (clone RB6-8C5) because in addition to the Ly6G Ag the latter binds Ly6C, expressed on monocytes, dendritic cells, and lymphocytes as well as on neutrophils (46, 47). In contrast, 1A8 recognizes only Ly6G and as such is specific to neutrophils (48, 49). As described in Results, compounds were administered i.p. in 100 ml sterile saline because the i.v route resulted in significant mortality of EAE animals. For flow cytometry experiments an additional control group was injected with 100 ml saline. The depletion treatment was initiated either at 5 or 7 d.p.i. with an injection of 100 mg Ab or control IgG in sterile saline. Afterward mice received 10 mg every other day up to 10 d later.


evaluation of BSCB leakage, every vessel was manually outlined with the freehand line tracer tool, and leaky vessels were identified based on the presence of dye outside blood vessels. The extent of permeability was calculated as the ratio of “leaky vessels length” over the “total vessel length” for a given animal. We routinely imaged .10,000 mm in total vessel length. For cellular infiltration studies, the analysis strategy was based on the parenchymal surface occupied by GFP+ cells. Channels from RGB images were split, after which the red channel (corresponding to the vasculature) was thresholded so that blood vessels were assigned a pixel value of 0 (black) and the parenchyma was assigned a value of 255 (white). In doing so, a vasculature mask was created and the number of white pixels was used as the measure of parenchymal surface that could potentially be occupied by GFP+ cells. Then the thresholded red channel was subtracted from the green channel of the original RGB image so that the latter was devoid of blood vessels, thus containing only GFP+ cells that had extravasated. To eliminate signal differences between experiments, the resulting image was thresholded and the area occupied by cells measured, regardless of fluorescence intensity. Hence, cellular infiltration was calculated as the ratio of the surface covered by GFP+ cells over the parenchymal surface.

3

4


For the experiments in which we investigated the role of IL-1R1 in neutrophil adhesion to the inflamed spinal cord microvasculature, EAEimmunized mice were injected i.v. daily with 100 ml of the IL-1R1 antagonist anakinra (Kineret; Swedish Orphan Biovitrum). Anakinra was injected at a concentration of 20 mg/ml in saline for a dose of 0.1 mg/g body weight, starting the day before MOG35–55 immunization until 14 d.p.i.

Purification of blood neutrophils Blood from naive and EAE mice was collected as described above and erythrocytes were lysed with 0.2% NaCl. Osmolarity was restored upon addition of 1.6% NaCl. Neutrophils were isolated by positive selection using immunomagnetic anti-Ly6G microbeads (Miltenyi Biotec) according to the manufacturer’s instructions. Eluted cells were washed, resuspended in culture medium, and their purity, assessed by flow cytometry, was on average 98%.

Purification of bone marrow neutrophils

Primary cultures of mouse BBB endothelial cells and neutrophil transmigration assay Primary cultures of mouse brain capillary endothelial cells (ECs) were prepared from 6- to 8-wk-old female C57BL/6 mice. Mouse brain tissue, free of meninges, was minced, homogenized, and digested in a mixture of 0.7 mg/ml collagenase type II and 39 U/ml DNase I in DMEM for 75 min at 37˚C. Myelin was removed by centrifugation at 1000 3 g for 20 min in 20% BSA-DMEM. The cell pellet was then incubated for another hour at 37˚C with a mixture of 1 mg/ml collagenase-dispase and 39 U/ml DNAse I in DMEM. Microvascular ECs were separated on a 33% continuous Percoll gradient, collected, and plated on culture dishes coated with 10 mg/ml collagen type IV and 5 mg/ml gelatin. Mouse brain capillary ECs were grown in DMEM supplemented with 20% (v/v) FBS, 1 ng/ml basic fibroblast growth factor, 100 mg/ml heparin, 1.4 mM hydrocortisone, and 13 antibiotic-antimycotic solution. Puromycin (10 mg/ml) was added to the media for the first 2 d of culture. On the third day, media was changed for fresh culture media containing 4 mg/ml puromycin. Cultures expressed vascular endothelial–cadherin protein. No immune reactivity for a-smooth muscle actin, glial fibrillary acidic protein, or neuronal nuclei protein could be detected, confirming the absence of contaminating smooth muscle cells, astrocytes, and neurons, respectively. Primary cultures of mouse BBB-ECs were used to generate an in vitro model of the BBB. After reaching confluence (typically 4–6 d), mouse BBB-ECs were seeded onto collagen type IV/gelatin-coated Transwell permeable inserts (3-mm pore size), at a density of 5 3 104 cells per well in EC culture media. Brain ECs were allowed to grow for 4 d to form a confluent monolayer. The upper compartment of each well of the 24-well Transwell was loaded with 200 ml cell suspension containing 1 3 106 purified mouse neutrophils obtained from the blood or bone marrow of naive and EAE mice at 7 d.p.i. Neutrophil migration was assessed by counting the absolute number of cells that transmigrated to the lower chamber after 18 h.

Real-time intravital videomicroscopy Neutrophils were purified from the bone marrow of EAE mice at 10 d.p.i., as described above. Neutrophils were stained with the CellTracker green fluorescent probe (Life Technologies, 1:2000 dilution), washed with HBSS, resuspended in sterile saline at a final concentration of 1 3 106 cells per 100 ml volume and then systemically injected via the right carotid artery. A total of three injections were made (each injected during 1 min) at 2-min intervals in IL-1R1–ko and wild-type (i.e., C57BL/6) mice with EAE at 15 d.p.i. Detailed methods describing the surgical procedure, intravital microscopy using the epi-illumination technique, as well as procedures for off-line quantitative analysis of neutrophil interactions in spinal cord white matter postcapillary venules at the lower cervical (C5) level have been published elsewhere (50, 51). The intravital microscopy imaging was performed on a custom-made Mikron IVM500 epifluorescent microscope (Mikron Instruments, San Marcos, CA) equipped with a low light–imaging camera VE-1000 silicone-intensified target system (DAGE-MTI, Michigan

Postmortem brain and spinal cord tissue from MS and NMO patients MS patient. An autopsy was performed on a 32-y-old MS female. Postmortem diagnosis was severe inflammatory rebound following cessation of natalizumab, as confirmed by Dr. Wolfgang Bruck and Dr. Imke Metz (Go¨ttingen University Medical School, Go¨ttingen, Germany). Because the patient was off natalizumab for .4 mo, she was not treated with plasma exchange. JC virus Ab serology was weakly positive, but JC virus PCRs on CSF and brain tissue were negative (four times). Pathological analyses of CNS frozen material revealed innumerable inflammatory infiltrates in the brain, brainstem, cerebellum, and spinal cord. Final pathological diagnosis was acute and severe exacerbation of MS with multiple actively demyelinating lesions following cessation of natalizumab. Whether this MS patient suffered from an immune reconstitution inflammatory syndrome remains a possibility, but this was not the diagnosis made by the neuropathologists (W. Br€uck and I. Metz, personal communication). NMO patients. Formalin-fixed, paraffin-embedded thoracic spinal cord material from two clinically suspected and autopsy-confirmed NMO patients were retrieved from the Pathology Department archival material at the Centre Hospitalier de l’Universite´ de Montreál. Patients were 68 and 50 y old. Disease duration was 5 and 14 y, respectively, and the cause of death was pneumonia in both cases. Clinical and pathological diagnoses were confirmed to be opticospinal disease, also known as NMO or Devic’s disease. Because the year of death/autopsy were 1988 and 1995, anti– aquaporin 4 Ab status is not available.

Immunostaining of MS and NMO tissue sections CNS material from one MS patient affected by an acute and severe episode of relapses (see above) was collected at autopsy and snap-frozen in n-methyl butane (270˚C). Brain sections (n = 3) were fixed in acetone for 10 min and transferred to ethanol for 5 min, hydrated in PBS, and blocked with 10% donkey serum at room temperature for 30 min. Sections were then incubated for 60 min with primary Abs diluted in 3% donkey serum. The following two primary Abs were used: mouse anti-human elastase (Dako Canada, 1:300 dilution) and rabbit anti-human fibrinogen (Innovative Research, 1:500). Next, sections were washed with PBS and 0.05% Tween 20 and incubated with secondary Abs at room temperature for 30 min. Secondary Abs were donkey anti-mouse Alexa-488 (1:500) and donkey anti-rabbit rhodamine red-X (1:500), both from Jackson ImmunoResearch Laboratories. Finally, sections were mounted using Gelvatol containing TO-PRO-3 (Life Technologies, 1:300). Each experiment included a negative control in which the primary Ab was omitted. Formalin-fixed paraffin-embedded CNS material from two clinically and pathologically confirmed NMO patients was also used. For immunohistochemistry and fluorescence staining, 3-mm-thick tissue sections (n = 4) were deparaffinized in three successive changes of toluene and rehydrated in 100 and 95% ethanol, water, and PBS. Slides were mounted with Permount. All reagents were from Sigma-Aldrich Canada. For neutrophil defensins (subtypes 1, 2, and 3) and fibrinogen stains, Ag retrieval was performed with sodium citrate at 95˚C for 30 min, cooled at room temperature, immersed for 3 min in PBS/Tween 20, and blocked in 10% rabbit serum. Sections were incubated with mouse anti-human neutrophil defensins (Leica Biosystems, 1:150) for 1 h at 37˚C and then washed in PBS/Tween 20. Secondary Ab (rabbit anti-mouse HRP; from Dako Canada, 1:150) was incubated at room temperature for 30 min and the immunoperoxidase reaction was developed using 3,3-diaminobenzidine as a chromogen. For the immunofluorescence stainings, the mouse anti-human neutrophil defensins (1:150) and rabbit anti-human fibrinogen (1:300) primary Abs were combined in the same incubation solution, following the method described above. In all cases, control stainings were performed omitting the primary Ab, and no immunopositive cells could be detected. Staining was visualized using either a Leica DM6000 microscope with OpenLab software or a Leica SP5 confocal microscope and analyzed using the Leica LAS AF software.

Statistical analysis Statistical analyses pertaining to EAE courses were performed using the nonparametric Mann–Whitney U test. Otherwise, data were analyzed with


Neutrophils were also purified from the bone marrow of naive and EAE mice whose peripheral blood was collected. After ensuring proper euthanasia, femurs and tibias were removed and the bone marrow was flushed with HBSS using a 25-gauge needle. Erythrocytes were lysed with ammonium chloride and cells were then passed through a 70- to 100-mm nylon mesh. Cells were washed and neutrophils isolated using two consecutive Percoll gradients of 64.8 and 61.5%, respectively. They were then washed in culture medium and their purity was assessed using flow cytometry as described above. The average neutrophil purity from bone marrow preparations was 97%.

City, IN). Firmly adherent neutrophils were identified as fluorescent cells that are stuck to the vessel wall without moving or detaching from the endothelium. Firm neutrophil adhesion was quantified at 10, 30, and 60 min after the first injection in four different fields of view (at 310 original magnification) containing a sufficient number of postcapillary venules. Importantly, the fields of view were selected before performing the first injection and contained approximately the same number of postcapillary venules.


5

a two-tailed unpaired t test, one-way ANOVA, or two-way repeatedmeasures ANOVA, followed by the Bonferroni posttest. A p value ,0.05 was considered statistically significant, and data are expressed as means 6 SEM unless otherwise noted. All analyses were performed with Prism software (GraphPad Software, San Diego, CA).

1G). These data therefore demonstrate that the surgery/imaging strategy can be implemented in EAE studies in lys-eGFP-ki mice and permits the collection of data prior to and following disease onset without compromising the pathophysiological events taking place at any time during disease development.

Results

BSCB disruption is transient and initiated shortly before EAE onset

Nonterminal intravital imaging sessions do not influence the initiation, progression, or severity of EAE in mice

BSCB permeability to low–molecular mass fluorescent tracers (NaFl, 376 Da; Alexa-594, 760 Da) was evaluated in C57BL/6 (NaFl; n = 28) and lys-eGFP-ki (Alexa-594; n = 30) mice at different time points throughout disease course (Fig. 1A). Animals were allowed to survive following imaging to associate the level of permeability with the day of disease onset and thus establish the temporal dynamics of BSCB disruption. On average for both mouse lines (n = 58), the integrity of the BSCB was undistinguishable among animals yet unaffected by motor deficits up to the day before the manifestation of symptoms, as demonstrated by a marginal permeability 4 d up to 2 d prior to disease onset (Fig. 1B). In lys-eGFP-ki mice, a small increase in permeability was detected on the day before onset when compared with other pre-onset time points (Fig. 1C). The compromised status of the BSCB reached a maximum during a 3-d period starting at disease onset, during which we observed tracer leakage in ∼40% of the blood vessels (Fig. 1A–C). After this peak period, the permeability steadily declined toward baseline values. Six days after onset, the extent of vascular leakage was about a fifth of the value observed during the peak period and by day 8 postonset it returned to pre-onset values (Fig. 1B). Because of the limited number of transgenic lys-eGFP-ki mice available, we chose to direct our imaging efforts in those animals around disease onset (n = 30 between 24 and 4 d versus onset; Fig. 1C). Hence, no data were collected later than 4 d postonset in lys-eGFP-ki mice. No permeability has been observed for either NaFl or Alexa-594 in the spinal cord of naive C57BL/6 and lys-eGFP-ki mice (data not shown). The surgical/imaging strategy we have implemented enables us to express BSCB permeability data relative to the day of disease onset, which is not possible with histological tissue preparations and terminal imaging sessions. Interestingly, linear regression analysis revealed that the clinical score attributed to individual animals was not correlated with the extent of tracer leakage measured (r2 = 0.135, Fig. 1D). However, it was evident that mice displaying partial or total limb paralysis (EAE score $2) presented high BSCB disruption. Hence, in our hands the EAE score is only partially related to the extent of BSCB disruption in the lumbar spinal cord and as such is not the optimal point of reference for reporting BSCB permeability. This slightly differs from other studies reporting a correlation between clinical severity and the extent of BSCB disruption, although they used different assays, immunization protocols, or mouse strains (4, 55). Many active EAE induction protocols, including ours, involve administration of PTX on the day of immunization and 24 or 48 h later (0 and 1 or 2 d.p.i.). It has been proposed that the toxin momentarily alters tight junction architecture via vasoactive amine sensitization, resulting in transient opening of the BBB in susceptible mouse strains (56–58). Therefore, we assessed whether breaches in the BSCB were detectable early in the course of the pathology, that is, 24 h following each i.v. PTX injection in immunized animals. None of the eight mice imaged at 1 or 3 d.p.i. (n = 4 mice/day, disease incidence = 100%) displayed fluorescent tracer accumulation outside blood vessels (data not shown). Additional control experiments were conducted in mice that received a combination of adjuvants (without MOG) to ascertain that the


In MOG-induced chronic EAE in C57BL/6 mice, disease progression follows a predictable monophasic pattern once animals start exhibiting clinical deficits. The most unpredictable parameter remains the onset of disease, which can occur anytime between 8 and 12 d.p.i. This poses considerable uncertainty in the investigation of pathological events taking place prior to disease onset, because it is difficult to establish correlation between observations made before onset and subsequent changes in behavior or disease course for instance. To alleviate the uncertainty pertaining to the day of EAE onset, we implemented an intravital imaging strategy in which mice are imaged once and allowed to survive until termination of the experimental protocol. This enables collecting data about cellular infiltration and status of the BSCB in asymptomatic animals and relating them to the exact timing of disease onset. Chronic implants have been described and would be the optimal strategy to put forward to remedy the situation, because they allow repetitive imaging of the same animal in a longitudinal fashion (52, 53). However, in our hands they are not suitable in the study of autoimmune diseases because they require animals to be administered with immunosuppressant or anti-inflammatory drugs to limit fibrosis over the surface of the spinal cord, a critical step to achieve acceptable success rates and in-depth high-resolution imaging. Before performing permeability or cellular infiltration studies, we validated that the surgery and imaging session did not affect the initiation, development, progression, or severity of the disease. The premise was to image animals at selected time points in the course of EAE and compare different parameters to establish whether indeed the surgery/imaging exerted an influence on any of them. Prior to immunization, female C57BL/6 (n = 26) mice between 8 and 10 wk of age were randomly assigned to one of six groups according to the time at which they would be imaged. Imaging sessions took place at days 1, 3, 7, 10, 14, and 17 following immunization. To avoid bias, animals were imaged regardless of their presenting EAE symptoms or not. All mice developed EAE (100% incidence), so based upon their individual EAE course they were attributed either the group “Pre onset” (n = 16) or “Post onset” (n = 10), depending on whether they were imaged before or after first manifesting EAE symptoms, respectively. The mean and median EAE scores were similar between both groups at all times throughout the entire duration of the experiment (Supplemental Fig. 1A, 1B), as were the mean days of onset (Supplemental Fig. 1C) and peak of disease (Supplemental Fig. 1D). Hence, the surgical protocol and imaging session do not influence the day when animals first display symptoms or reach their highest clinical score. To evaluate the total severity of EAE, the area under the curve and maximal clinical score attained were used as summary statistics to compare individual animals (54). Both parameters were identical between the two groups (Supplemental Fig. 1E, 1F), indicating that conducting a single imaging session does not affect the extent at which animals develop EAE. Moreover, comparison of the EAE course between C57BL/6 mice and heterozygous lys-eGFP-ki mice in the C57BL/6 background in separate experiments revealed no differences in terms of days of onset and peak of disease or the severity of EAE (Supplemental Fig.

6


vascular leakage measured in EAE animals was not caused by the adjuvants themselves. For this purpose, 16 animals were randomly assigned to one of four groups (n = 4/group) receiving different components of the emulsion. Mice were either injected with PTX alone, PBS plus CFA, PTX plus CFA (categorized as control groups), or PTX plus CFA plus MOG (i.e., the EAE group). By virtue of the results presented in Fig. 1B, imaging sessions took place as mice from the EAE group first displayed clinical symptoms (at ∼10 d.p.i.; disease incidence was 100%) to compare data acquired in a similar time frame. None of the animals from the

three control groups exhibited symptoms at any time during the protocol (data not shown) nor did they present evidence of BSCB permeability (Supplemental Fig. 2A). In sharp contrast, in all four mice from the EAE group we could measure a significant loss of BSCB integrity. To definitively rule out the possibility that PTX induces BSCB disruption, and to study the behavior of myeloid cells in response to PTX treatment, we examined BSBC leakage in lys-eGFP-ki mice. For this purpose, both saline-treated (n = 2) and PTX-treated (n = 4) mice were imaged at 6 h postinjection by means of two-photon intravital microscopy and then allowed to


FIGURE 1. In vivo permeability of the BSCB to low molecular mass fluorescent tracers shortly precedes EAE onset and coincides with the infiltration of LysM+ cells. (A) Representative projections of intravital images acquired in the lumbar spinal cord of C57BL/6 mice using NaFl (green, top) and in lyseGFP-ki mice using Alexa-594 (red, bottom). Scale bars, 50 mm. (B) Quantification of BSCB leakage in both mouse strains pooled together (n = 58). An increase in permeability is apparent 1 d prior to disease onset. Maximal leakage occurs during a 3-d period starting at disease onset, after which it declines over a few days. (C) Quantification of BSCB leakage and GFP+ infiltration in lys-eGFP-ki mice (n = 30). Infiltration takes place 1 d prior to disease onset, concomitantly with BSCB disruption. (D) The extent of BSCB permeability is not correlated with the clinical EAE score. However, animals displaying moderate or heavy signs of paralysis (score $2) have a much more leaky BSCB than animals that do not (score ,2). Data are pooled from both mouse strains (n = 58). (E) Confocal image of a spinal cord from an EAE lys-eGFP-ki mouse. GFP+ cells were found in the lumen of blood vessels, between the endothelial and parenchymal basement membranes (pan-laminin staining in red), migrating across the parenchymal basement membrane (arrows), as well as in the spinal cord parenchyma. Leukocyte infiltration causes a distention of the perivascular space (*). Scale bar, 10 mm. (F) Individual images from an intravital time-lapse video showing GFP+ cells (white arrows) leaving the vasculature (Alexa -594, red) and entering the perivascular space in a mouse one day prior to EAE onset. Scale bar, 25 mm. All data are represented as means 6 SEM. **p , 0.01 by two-tailed unpaired t test (D).


Myeloid cell infiltration in the spinal cord precedes EAE onset and is concomitant with BSCB disruption Because lys-eGFP-ki mice are reporters of mature granulomyelomonocytic cells (60), we investigated the temporal dynamics of their infiltration in the spinal cord during the course of acute EAE. Accumulation of GFP+ myeloid cells outside blood vessels was not observed in asymptomatic animals until the eve of disease onset (Fig. 1A, 1C, 1E). Before that, we could only visualize cells circulating in the bloodstream with occasional instances of perivascular macrophages or extravasated cells, as previously reported in the normal spinal cord of lys-eGFP-ki mice (61). However, a considerable shift in the infiltration pattern was evident concomitantly with an increase in BSCB permeability 1 d prior to onset (n = 4, Fig. 1A, 1C). Interestingly, a large number of GFP+ cells were bordering the outer surface or in the close vicinity of blood vessels inside the CNS, suggesting that they were in the process of transmigration across the BSCB or had recently entered the perivascular space or spinal cord parenchyma (Fig. 1A, onset 21). We are inclined to think that all of these possibilities are likely to be true for the following reasons: 1) GFP+ myeloid cells were seen crossing the spinal cord endothelium in vivo, as demonstrated by a series of images from time-lapse movies (Fig. 1F); and 2) GFP+ myeloid cells were found in the lumen of blood vessels, between the endothelial and parenchymal basement membranes, and in the spinal cord parenchyma upon inspection of spinal cord sections by laser-scanning confocal microscopy (Fig. 1E). Notably, some GFP+ cells were observed in the process of migrating across the parenchymal basement membrane, most likely toward the parenchyma (see arrow-pointed cells in Fig. 1E). Animals imaged on the day of clinical onset presented extensive infiltration as demonstrated by the high CNS coverage we measured, reaching a maximal value 1 d following onset (n = 6, Fig. 1C). To our surprise, the presence of GFP+ myeloid cells in the spinal cord parenchyma was not as long-lived as we expected. Indeed, their surface coverage started declining 2 d after the appearance of symptoms (n = 4) and by 4 d postonset was a third of

the maximum value measured on the day after onset (Fig. 1C). The temporal course of GFP+ cell infiltration shared many similarities with that of BSCB disruption. Notably, both were detectable shortly before the appearance of motor deficits and reached a maximum at disease onset for a 3-d period, after which they declined with a slightly different dynamics. The most noticeable difference was that BSCB breakdown is considerable on day 2 postonset whereas cellular infiltration is already diminished by that time (Fig. 1C). This infiltration pattern is reminiscent of a short-lived myeloid population, which together with their high number and early course of action relative to BSCB disruption led us to hypothesize that GFP+ infiltrates are mainly composed of neutrophils. This hypothesis is supported by earlier studies that showed that neutrophil infiltration increases during EAE onset, remains high at the peak of disease, and dramatically declines thereafter (18, 62). Flow cytometric analysis of blood from lys-eGFP mice shortly before EAE onset (8 d.p.i.) confirmed that neutrophils are significantly more numerous than M1 and M2 monocytes, representing ∼40% of all CD45+ leukocytes (Fig. 2A). Neutrophils constituted ∼75% of all GFP+ cells in the blood at this time point (Fig. 2B), and the mean GFP fluorescence intensity in neutrophils was at least 3-fold higher than in either monocyte subsets (Fig. 2C, 2D). Examination of leukocytes in the spinal cord of lys-eGFP mice revealed infiltration from Ly6G+ neutrophils at similar levels than F4/80lo and F4/80hi macrophages and CD3+ T lymphocytes both before and on the day of EAE onset (Fig. 2E, 2G). Interestingly, infiltrated neutrophils from mice not yet displaying symptoms were forming a high proportion of GFP+ cells (Fig. 2F), consistent with data obtained from the blood (Fig. 2B). Once mice started exhibiting locomotor deficits, however, macrophage subsets and neutrophils were indistinguishable in terms of their contribution to the GFP+ population (Fig. 2H). Of note, no macrophages expressing M2 markers were detected at any time point in the spinal cord, and T cells did not express GFP in the spinal cord and blood (Fig. 2F, 2H and data not shown). In addition to flow cytometry experiments, immunohistochemical stainings were performed on lumbar spinal cord sections to confirm that neutrophils are indeed present in the CNS shortly before disease onset in EAE mice. As mentioned in Materials and Methods, none of the Abs we tested allowed satisfactory discrimination of infiltrating macrophages from resident microglia, and hence only immunostainings for neutrophils and T cells were quantified (Fig. 2I, 2J). At 7 d.p.i., Ly6G+ neutrophils were significantly more numerous than CD3+ T cells, whereas both cell types were present at similar levels at 14 and 21 d.p.i. Taken together, these results support other studies pointing toward an early involvement of neutrophils in CNS pathophysiology (15, 16, 18, 27), as they infiltrate the spinal cord parenchyma during the preclinical stage of EAE, consistent with our intravital imaging data (Fig. 1A–C). This suggests an intricate relationship between neutrophils and events taking place early in the course of the disease. By virtue of the results from our intravital studies (Fig. 1A–C), we hypothesized that BSCB disruption is such an event deserving further investigation. Neutrophils are eliminated from the circulation of EAE mice administered the anti-Ly6G 1A8 mAb To test this hypothesis, we performed depletion experiments in EAE mice and characterized BSCB integrity in animals devoid of circulating neutrophils. To selectively deplete the neutrophil population, we administered anti-Ly6G mAb (clone 1A8) prior to EAE onset. The efficiency of the treatment in eliminating neutrophils and sparing monocytes was verified in whole blood using


survive until day 2. Animals were then killed by transcardiac perfusion and their spinal cords processed, immunostained, and imaged by confocal microscopy. No permeability was observed for Alexa-594 in the spinal cord of saline-treated and PTX-treated lys-eGFP-ki mice at 6 h postinjection (Supplemental Fig. 2B). However, intravital imaging revealed the presence of GFP+ cells firmly adhered to the endothelium and sometimes crawling along spinal cord microvessels after PTX treatment, corroborating an earlier immunohistochemical study by Richard et al. (59). The only occasional GFP+ cells that were observed in control animals had a spindle-shaped morphology and were located in the perivascular space, reminiscent of perivascular macrophages. Importantly, the myeloid cells that adhered to the spinal cord endothelium of PTX-treated mice did not penetrate into the parenchyma and appeared to be anatomically restricted to the meningeal vessels (Supplemental Fig. 2C, 2D). Collectively, our data rule out the contribution of adjuvants in disrupting the BSCB shortly following immunization or disease onset in EAE animals. As far as PTX is concerned, this corroborates findings from other groups who also failed to detect an effect of the toxin in disrupting the BSCB (9, 11). Note, however, that PTX clearly induces changes at the level of the spinal cord endothelium, changes that seem to be related to the recruitment and adhesion of immune cells. Therefore, the permeability observed is elicited by the combination of adjuvants and the inflammatory process induced by myelin fragments as opposed to an action of the adjuvants alone.

7

8


flow cytometric analysis 24 h following the first Ab injection. Cells were first gated according to their scattering properties and then doublets were discarded using the forward and side light scatter

parameters (Fig. 3A, left). After selecting live CD45+ cells (i.e., live leukocytes; Fig. 3A, middle), neutrophils and monocyte subsets were identified according to their lack of B220 expression


FIGURE 2. Neutrophils infiltrate the spinal cord during preclinical EAE, constituting most GFP+ cells in lys-eGFP-ki mice blood and spinal cord during the preonset period. (A–D) Flow cytometric analysis of lys-eGFP mouse blood at 8 d.p.i. (n = 4), which is before disease onset. Cells were gated so that doublets were excluded and only live CD45-expressing leukocytes were considered. Individual populations were identified based on their expression (or lack thereof) of GFP as well as CD11b, 7/4, Ly6G, and Ly6C markers. Neutrophils are more numerous than M1 and M2 monocytes relative to live leukocytes (A) and form the dominant population expressing GFP (B) during the preclinical stage of EAE. Histogram and quantification of the mean GFP fluorescence intensity show that GFP is expressed at a higher level in neutrophils compared with either monocyte subsets (C and D). (E–H) Flow cytometric analysis of lys-eGFP mouse spinal cord before and on the day of EAE onset (n = 3 mice/time point). Live CD45+ cells were gated as described above, and microglia (CD45dim) were distinguished from infiltrating blood-derived leukocytes (CD45hi) based on their levels of CD45 expression. The phenotype of neutrophils, macrophages (Mf; which includes both hematogenous and perivascular macrophages), and T cells was confirmed with Ly6G, F4/80, and CD3, respectively, in addition to CD11b, 7/4, and Ly6C. Note that monocytes that have recently emigrated from the bloodstream into the spinal cord perivascular spaces express F4/80 on their surface, but the level is lower than on perivascular and spinal cord–infiltrated macrophages (F4/80lo versus F4/80hi). All macrophages detected were Ly6Chi, which suggests that no M2 macrophages were present. (E) Neutrophils, macrophages, and T cells infiltrate the lower spinal cord (T9–S4) at similar levels before EAE onset. (F) Neutrophils appear to constitute the dominant population expressing GFP at this time point, similar to what has been observed in the blood (B). (G) Neutrophils, macrophages, and T cells infiltrate the spinal cord at similar levels, which are higher than in mice in the preclinical stage of the disease. (H) Neutrophils and macrophages contribute equally to the GFP+ cellular population. (I) Immunohistochemical stainings of C57BL/6 mouse spinal cords at 7 (left), 14 (right), and 21 (not shown) d.p.i. (n = 3/time point). Scale bar, 200 mm. (J) Quantification reveals that neutrophils have infiltrated the spinal cord at 7 d.p.i., as opposed to T cells. At 14 and 21 d.p.i., both cell types are present at similar levels. All data are expressed as means 6 SEM. *p , 0.05, **p , 0.01, ***p , 0.001 with one-way ANOVA followed by Bonferroni posttest (A– H) or two-tailed unpaired t test (J).


Neutrophil depletion delays the manifestation of EAE, attenuates clinical symptoms, and prevents early BSCB disruption Neutrophil depletion has been performed in autoimmune contexts, where it was shown that the effector phase of relapsing–remitting EAE was suppressed in SJL mice that received the anti–Gr-1 mAb (clone RB6-8C5) directed against both Ly6C and Ly6G (32). Interestingly, Carlson et al. (16) demonstrated that Evans blue dye extravasation was prevented in neutrophil-depleted SJL mice during acute symptomatic and relapse episodes, in addition to a restoration of clinical symptoms following cessation of the anti– Gr-1 injections. We thus verified that neutrophil depletion using the 1A8 mAb resulted in similar effects on the EAE course of C57BL/6 animals. For that purpose, the treatment was initiated prior to the onset of clinical symptoms and continued every other day until the peak of the disease (Fig. 4A). Similar to the abovementioned studies, the manifestation of motor deficits was delayed in mice receiving the 1A8 mAb in comparison with the isotype control, as demonstrated by differences in the mean daily scores and number of days with symptoms (Fig. 4A, 4B). Moreover, mice depleted in neutrophils were either protected against EAE or exhibited attenuated motor deficits, as the maximal clinical score they reached during the protocols was significantly less than in their control littermates, which all developed EAE (Fig. 4B, middle). Finally, the total severity of the disease during the 30 d of observation was lower in the anti-Ly6G group (Fig. 4B, right).

Therefore, our data corroborate findings by other groups who observed a delayed and attenuated manifestation of neurologic and motor deficits in EAE mice depleted of Gr-1+ cells, among which the neutrophil population is the most abundant (16, 32). The delayed onset of clinical symptoms elicited by the absence of neutrophils suggests an intricate relationship with the immunological events taking place early in the course of EAE, among which BSCB disruption and immune cell infiltration are key elements (63). Therefore, we performed intravital imaging experiments around disease onset in neutrophil-depleted mice to establish whether the altered EAE courses observed were associated with changes in BSCB integrity. Considering the time frame during which we measured significant leakage of low–molecular mass fluorescent tracers in previous experiments (Fig. 1), imaging sessions were conducted when animals from the isotype control group were displaying acute clinical symptoms (i.e., on days 0 and 2 postonset). In mice that were depleted in neutrophils and subsequently imaged, we observed no (in six of eight) or negligible (in two of eight) amounts of vascular tracer outside a few blood vessels (Fig. 4C). In sharp contrast, every animal from the control group displayed evident vascular leakage, with the fraction of leaky vessels from this group being 7-fold larger than in the antiLy6G group for the same time frame during which imaging took place (Fig. 4D). Collectively, our data point toward a plausible involvement of neutrophils in creating breaches in the BSCB early in the course of EAE, after which they are suited to enter the CNS environment and promote the entry and perhaps polarization of T cells, monocytes, and APCs, thus helping the initiation of deleterious inflammatory events resulting in lesion formation and neurologic deficits. Neutrophils acquire a higher capacity to transmigrate across murine endothelial cells during the process of EAE We next chose to investigate whether the induction of EAE could affect the capacity of neutrophils to migrate across the BBB, using our previously established in vitro model of leukocyte transmigration. We found that highly purified neutrophils isolated either from the blood or the bone marrow of presymptomatic EAE animals migrate more efficiently in vitro across mouse CNS microvascular EC monolayers than do those collected from nonimmunized animals (Fig. 4E). These data demonstrate that the transmigration of neutrophils through brain microvascular ECs is significantly enhanced early during the course of EAE, and that this process can occur without the contribution of other leukocyte subsets. Firm adhesion of neutrophils to the inflamed spinal cord microvasculature is regulated by IL-1R1 during EAE Previous studies have suggested an important role of IL-1R1 in neutrophil recruitment at sites of inflammation (64, 65), as well as in the pathogenesis of EAE, MS, and NMO (66–69). Thus, we determined the expression pattern of IL-1R1 in the mouse spinal cord. We performed immunofluorescence staining on spinal cord tissue taken from naive C57BL/6 and IL-1R1–ko (negative control) mice using a polyclonal anti–IL-1R1 Ab. IL-1R1 staining was always found in close proximity to blood vessel basement membranes and observed on their luminal side where endothelial cells are found (Fig. 5A). Importantly, no immunofluorescence signal was detected in the spinal cord of IL-1R1–ko mice (Fig. 5B), confirming the specificity of the Ab used and suggesting that endothelial cells are the major source of IL-1R1 in the normal spinal cord. Because of the reported importance of IL-1R1 in neutrophil recruitment in various inflammatory conditions, and because en-


combined with their differential expression of 7/4 (Ly6B.2; Fig. 3A, right). The addition of the B220 marker enabled normalization of cell counts to the B cell population (Fig. 3B), which is not affected by the depletion treatment (44). We confirmed the phenotype of every gated cell population with the CD11b, Ly6C, and Ly6G markers. Because the depleting Ab was directed against Ly6G, this marker was not used in the gating strategy but rather to confirm that Ly6G+ cells were indeed eliminated from the blood. The B cell population was quantified for a 1-min-long acquisition, and B cell counts from the three groups (saline, control IgG, and anti-Ly6G) were identical (Fig. 3B). In contrast, quantification of the total live CD45+ leukocytes population revealed that animals injected with the anti-Ly6G mAb had significantly fewer leukocytes in their blood than did the controls (n = 5/group, Fig. 3C), thus suggesting that the Ab effectively eliminated cells instead of blocking receptors at their surface, thereby preventing their detection. Gating cells based on the expression of the 7/4 Ag confirmed that neutrophils were completely absent from the circulation in anti-Ly6G–treated mice (Fig. 3D). In contrast, neutrophil counts were similar among the isotype control and saline groups. Quantification of the Ly6Chi and Ly6Clo monocyte subsets supported the premise that the anti-Ly6G treatment did not affect any of them, indicating that 1A8 specifically depletes circulating neutrophils and spares the monocyte populations (Fig. 3E, 3F), in agreement with previous reports (43, 48). As a point of interest, the administration route was a critical factor for keeping mice healthy in depletion experiments. Indeed, previous independent experiments revealed that i.v. injection of the 1A8 Ab on day 7 postimmunization was not tolerated by EAE animals, as 4 of 14 died in the first 3 h following the initial injection (data not shown). In contrast, the i.p. route caused no problems whatsoever regarding the survival or well being of the animals, and hence for every depletion experiments whose results are presented in this study, Abs were administered i.p. Importantly, the depletion of neutrophils is long-lasting and very effective, with undetectable numbers of blood neutrophils for at least 3 d after bolus injection of the anti-Ly6G mAb (Fig. 3G).

9

10


dothelial cells are the principal source of IL-1R1 in the mouse spinal cord, we next reasoned that systemically infused neutrophils would be incapable of migrating across the BSCB in mice lacking IL-1R1 during EAE. Using intravital epifluorescence videomicroscopy, we therefore quantified the number of fluorescently labeled neutrophils permanently adhering to spinal cord postcapillary venules at different time points after neutrophil infusion in both IL-1R1–ko and control (C57BL/6) mice (Fig. 5C–E). As shown in Fig. 5C, neutrophils were found to undergo firm and sustained adhesion to the spinal cord endothelium of C57BL/6 mice with EAE. In contrast, almost no neutrophils could firmly adhere to the microvasculature of IL-1R1–ko mice, as evidenced

by a .90% drop compared with numbers seen in the C57BL/6 recipient group. Taken together, these results suggest that endothelial IL-1R1 is critical for firm adhesion of neutrophils to spinal cord microvascular walls during EAE. Finally, we sought to determine whether treatment with the IL1R1 antagonist anakinra could replicate the findings from the neutrophil depletion studies (Fig. 4), that is, that clinical signs of EAE are delayed in mice that received the depleting Ab compared with those treated with the isotype control. Similar to neutrophil depletion, IL-1R1 blockade through i.v. administration of anakinra significantly delayed the day of manifestation of clinical signs of EAE compared with saline treatment (Fig. 5F, 5G). Thus, our


FIGURE 3. Neutrophils are eliminated from the blood of EAE mice after i.p. injection of the anti-Ly6G 1A8 mAb. (A) Flow cytometric analysis of EAE mouse blood at 8 d.p.i. Cells were first gated according to their scatter characteristics (left), after which doublets were discarded and live CD45+ cells (i.e., leukocytes) were selected (middle). Then, B cells, neutrophils, and monocytes were identified based on their expression of distinct markers, including B220 and 7/4 (right). Cellular phenotypes were confirmed with other markers such as CD11b, Ly6C, and Ly6G. (B and C) Quantification of B cells (B) and live leukocytes (C) at 1 d after neutrophil depletion (n = 5 mice/group). Note that the depletion treatment was initiated at 7 d.p.i. and that cells were counted during a 1-min acquisition time. The B cell population is unaffected by the anti-Ly6G treatment, as opposed to the overall leukocyte population. This indicates that the 1A8 mAb effectively depletes cells instead of blocking surface receptors. (D) Neutrophils are eliminated from the circulation after administration of 1A8 mAb, as opposed to the isotype control and saline treatments. (E and F) Both the Ly6Chi and Ly6Clo monocyte subsets [(E) and (F), respectively] are unaltered by the 1A8 mAb injection when compared with either control group. Cell counts were normalized to the number of B cells, which were not affected by the depletion treatment (B). (G) The depletion of neutrophils in the blood is long-lasting, persisting for at least 3 d after the bolus injection of the depleting anti-Ly6G Ab (100 mg given i.p.). Data are represented as means 6 SEM. *p , 0.05, **p , 0.01, ***p , 0.001 with one-way ANOVA followed by a Bonferroni posttest.


11


FIGURE 4. Neutrophil depletion delays EAE onset and severity and prevents early BSCB permeability. (A) Time course from two independent experiments, showing that initiating the treatment at 7 (left, n = 6/group) or 5 (right, n = 8/group) d.p.i. results in delayed day of onset and reduced severity. Red arrows refer to days at which mAb injections took place. Disease incidence was 100% in both experiments for the IgG control groups. (B) Summary statistics for different parameters. Because some neutrophil-depleted animals are protected from EAE, the number of days with symptoms (left) is used as an alternative to the day of onset, which would be undefined in those animals. Neutrophil-depleted mice display a delayed onset, because they exhibit symptoms for shorter period of times than do IgG controls. The maximal clinical score is lower in mice devoid of circulating neutrophils (middle), as is the area under the curve (AUC, arbitrary units; right). Therefore, disease severity is reduced in those animals compared with the IgG controls. Data are pooled from the two experiments shown in (A) (n = 14/group in total). (C and D) Representative intravital images (C) and quantification (D) of BSCB integrity from anti-Ly6G–treated (n = 8) and control IgG-treated (n = 4) EAE mice. All animals were imaged during the same time frame, which is when IgG controls were displaying acute EAE symptoms. The depletion treatment yields a markedly reduced BSCB permeability for the time (Figure legend continues)

12


results suggest that a key mechanism by which neutrophils are recruited to the inflamed spinal cord during EAE involves signaling by IL-1R1 in ECs. The presence of neutrophils correlates with BBB leakage in acute lesions in MS brain and in the spinal cord of NMO

Discussion It is well established that activated leukocytes cross the disrupted BSCB in NMO, MS, and EAE, although the underlying molecular bases of those pathological events are ill-defined (63). In this study, we induced EAE in lys-eGFP-ki mice and used two-photon intravital imaging to monitor BSCB leakage and correlate this process with the infiltration of mature granulomyelomonocytic cells in the lumbar spinal cord during the course of the disease. We present evidence to support the notion that neutrophils are involved in the initial steps of the neuroinflammatory response in EAE. This is demonstrated by the early influx of neutrophils in the CNS of EAE mice, which precedes the onset of symptoms and coincides with an increased permeability of the BSCB to low– molecular mass fluorescent tracers. Consistent with this hypothesis, we found that 1) neutrophil depletion prevented vascular leakage in the spinal cord of EAE mice, and 2) neutrophils isolated from the bone marrow or peripheral blood of EAE mice transmigrate more efficiently across the BBB than do naive neutrophils, a phenomenon that occurs without any influence from other immune cell types. Although sparse, perivascular neutrophils were found in areas of the CNS displaying increased BBB leakage, in both MS and NMO patients. To enable disease monitoring and the association of pre-onset observations to subsequent manifestation of motor deficits, we devised an experimental paradigm in which mice are allowed to survive following the single imaging session they undergo. Using this strategy, we measured an increase in vascular leakage 1 d prior to disease onset. No leakage of fluorescent tracer outside blood vessels could be detected prior to that day. Experiments performed in the rat using low–molecular mass molecules such as mannitol or radiolabeled ionic tracers revealed a comparable pattern of dif-

period investigated. Scale bar, 50 mm. (E) Neutrophils from EAE mice have a greater propensity to transmigrate than do naive neutrophils. Neutrophils (1 3 106 cells/Transwell) were allowed to migrate for 18 h across a confluent monolayer of primary mouse BBB ECs in a Boyden chamber migration assay. Each dot represents one Boyden chamber. Neutrophils were isolated from the blood or bone marrow (BM) of naive mice (n = 30) and EAE mice (n = 10) at 7 d.p i. Data are represented as means 6 SEM. *p , 0.05, **p , 0.01 with a two-tailed Mann–Whitney U test (A and B), a two-tailed unpaired t test (D), or a oneway ANOVA followed by a Bonferroni posttest (E).


In active and chronic MS lesions, the CNS cellular infiltrates are composed mainly of monocytes/macrophages and T lymphocytes. However, the presence of neutrophils in the early process of immune cell infiltration in MS is still a matter of debate. To establish clinical relevance to our findings, we elected to evaluate the presence of neutrophils in very acute MS lesions and to correlate the extent of BBB leakage with the presence of neutrophils. Our data show that in hyperacute MS lesions, the occasional presence of perivascular elastase+ neutrophils is associated with focal areas enriched in fibrinogen, confirming that neutrophil infiltration is associated with focal breaches in the BBB (Fig. 6). This was confirmed in the spinal cord of NMO patients, an Ab-mediated disease in which neutrophils are known to play a pivotal role. In the spinal cord of these patients, defensins+ neutrophils were seen scattered through the sections, always in close contact with microvessels (Fig. 7). These neutrophils were often seen in areas immunopositive for fibrinogen (Fig. 7, lower panels). As Enzmann et al. (70) reported in acute human stroke lesions, neutrophils were not seen in CNS parenchyma.

fusion, with extravasation of ions from the lumbar spinal cord vasculature before other CNS regions (71, 72). Similar observations were made in C3H/He mice, in which ascending paralysis was closely associated with HRP extravasation in the spinal cord, whereas predominant cerebellar involvement was evident only in instances of axial rotatory EAE (11). In the EAE SJL mouse model, it was proposed that the first occurrence of BBB disruption and rabbit IgG leakage occurred in the cerebellum, and then spread to the spinal cord (9). Magnetic resonance imaging in the mouse revealed that BSCB and BBB disruption took place as animals were already displaying acute motor symptoms (10, 12). In the brain of EAE mice, Floris et al. (10) further observed that monocyte infiltration preceded vascular leakage, although assessments of BBB integrity and cellular infiltration were performed in separate animals that were sacrificed at the end of the imaging session. Nevertheless, it demonstrated that loss of vascular integrity is closely related to the initiation of the disease. In recent studies the extent of BSCB disruption was shown to correlate with the severity of EAE, as measured by fluorometric analyses of whole homogenized spinal cords (55, 73). In the present study, we did not find a direct correlation between the level of BSCB permeability and clinical score, but rather with the day of disease onset. However, we did observe that the extent of leakage was significantly higher in animals presenting signs of paralysis compared with those that did not. Of note, we observed considerable vascular leakage in vivo between 10 and 30 min following i.v. injection of the fluorescent tracer, whereas the authors of the above-mentioned studies transcardially perfused animals 10 min following the i.p administration of the tracer (55, 73). Hence, this short circulation time combined with the i.p. injection route might have resulted in underestimation of BSCB disruption in some animals. Another important conclusion stemming from our observations is that the increase in BSCB permeability is transient, peaking 2 d after disease onset and returning to baseline values ∼6 d postonset. This is consistent with findings from other groups who also reported a short-term, rather than a permanent, BSCB and BBB disruption during EAE (10, 72, 74). Interestingly, magnetic resonance imaging studies in MS patients provided evidence that temporary breakdown of the BBB occurs early in the formation of new lesions and can be associated with younger age of onset or more severe disease in relapsing–remitting patients (75, 76). Also note that we have only used small tracers for permeability studies, comparable in size to ions (360–760 Da). Therefore, the temporal changes in BBB integrity that we detected in our study do not necessarily match alterations of BBB permeability to larger molecules, such as albumin or dextrans for instance. As Kang et al. (77) demonstrated, it is erroneous to assume that the BBB behaves similarly toward small ions and large molecules, supporting the hypothesis of a hierarchical pattern of disruption (4). Therefore, it is plausible that high–molecular mass vascular tracers would extravasate according to a different dynamics than what we measured with low-molecular mass NaFl and Alexa-594. The triggers responsible for initiating BSCB breakdown in EAE are yet to be elucidated, although many cellular and molecular candidates have been shown to influence or modulate its permeability (6). Constituents of the neurovascular unit such as ECs,


13

mast cells, and pericytes have received considerable attention because of their influence on vascular stability (8, 78–80). In the periphery, cytokines and chemokines released by circulating effector cells such as T cells and neutrophils are also known to affect CNS barrier homeostasis. Notably, IL-1b has been shown to modulate BBB permeability in mice and rats and to activate human ECs (7, 81, 82). Interestingly, we found in the present study that blocking or deleting the endothelial IL-1R1 prevents the adhesion of circulating neutrophils to the spinal cord endothelium and delays EAE onset. The choroid plexuses were identified as an alternate route for Th17 cell entry into the CNS via CCR6–CCL20 signaling (83). Interestingly, neutrophils release CCL20 under inflammatory conditions, and this chemokine is implicated in a reciprocal activation and recruitment pattern between human neutrophils and Th17 cells (30, 84). Because of their capacity to produce and secrete a broad array of molecules and given their rapid deployment to sites of inflammation, neutrophils have emerged as likely contributors to the disruption of blood–CNS barriers induced in EAE (28, 63, 85–88). Recent studies suggest that myelin-specific encephalitogenic T cells, upon reactivation in the meninges by dendritic cells, act in concert with mast cells to promote neutrophil recruitment and subsequent disruption of the BBB (13, 74). The authors postulated that neutrophils promote this phenomenon, thereby facilitating massive leukocyte influx through the compromised BSCB. In support of this hypothesis, in vitro experiments by Allen et al. (89)

demonstrated that neutrophils acquire a neurotoxic phenotype upon cerebrovascular transendothelial migration, releasing soluble factors and proinflammatory cytokines, which could facilitate such an influx into the CNS. Interestingly, this neurotoxic phenotype of neutrophils was acquired when neutrophils transmigrated across an IL-1–stimulated brain endothelium. Along the same line, our in vivo data show that blockade of IL-1R1 signaling with anakinra injected i.v. delays the clinical manifestations of EAE. Furthermore, we found that activated neutrophils isolated from EAE mice are unable to firmly adhere to spinal cord microvessels, and therefore to transmigrate across the BSCB, when infused into the bloodstream of IL-R1–ko mice. IL-1 is among the most important differentiation factors for Th17 cells (67), which are necessary for induction of autoimmune encephalomyelitis (90), and it is also critical for encephalitogenicity of Th17 cells by regulating GM-CSF production and subsequent recruitment of effector myeloid cells (91, 92). Work by Segal and colleagues (16) provides further evidence of the interplay between T cells and neutrophils in the context of EAE by showing a significant upregulation of neutrophil-attractant chemokines following transfer of myelin-specific CD4+ Th17 cells in naive mice. Interestingly, these chemokines were present in the CNS of EAE animals before the clinical manifestation of symptoms. These findings are consistent with other studies reporting accumulation of neutrophils in the meninges prior to EAE onset as well as their prompt entry in the CNS upon immunological challenge, evi-


FIGURE 5. Firm adhesion of neutrophils to the inflamed spinal cord microvasculature is regulated by IL-1R1 during EAE. (A and B) Representative confocal photomicrographs showing IL-1R1 immunostaining (red) in the spinal cord of a naive C57BL/6 and IL-1R1-ko (negative control) mice. Note the close proximity between the IL-1R1 signal found at the endothelial cell surface and the pan-laminin immunostaining (green) of blood vessel basement membranes. Nuclear staining with DAPI is shown in blue. Scale bars, 15 mm. (C and D) Fields-of-view (FOV) images taken from an intravital real-time video showing neutrophils, stained with the CellTracker green fluorescent probe, adhering to the inflamed BSCB endothelium of an immunized C57BL/6 mouse (C; C57 → C57), but not of an immunized IL-1R1-ko mouse (D; C57 → IL-1R1-ko), at 30 min after cell infusion. Blood vessels are delineated by dashed red lines. (E) Quantification of firm neutrophil adhesion at 10, 30, and 60 min following systemic infusions of fluorescently stained neutrophils into C57BL/6 (n = 6) and IL-1R1-ko (n = 6) recipient EAE mice. All data are expressed as means 6 SEM, with the mean representing the average number of firmly adherent neutrophils per FOV (n = 4 FOV/mouse) using the 310 objective. **p , 0.01 with two-way repeated-measures ANOVA, followed by a Bonferroni posttest. (F and G) Mice treated with the IL-1R1 antagonist anakinra developed clinical signs of EAE significantly later than those treated with saline, similar to the response seen in neutrophil-depleted mice (see Fig. 4). Data are represented as means 6 SEM. **p , 0.01, ***p , 0.001 with twoway repeated-measures ANOVA followed by a Bonferroni posttest (F) or a two-tailed Mann–Whitney U test (G).

14

NEUTROPHILS DISRUPT BSCB INTEGRITY IN NEUROINFLAMMATION As in the case of BSCB permeability, this does not necessarily rule out disease severity as an indicator of neutrophil infiltration in the whole spinal cord. In our experimental paradigm, we systematically imaged the same lumbar region of the spinal cord in mice, thus potentially overlooking other areas of cellular infiltration or leakage. This limitation of our intravital imaging strategy is supplanted by numerous attributes, including the considerable temporal resolution and context it enables, as well as the large sampling volume achieved in the unperturbed CNS of live animals. Importantly, with standard histological techniques the temporal evolution of BSCB integrity cannot be precisely determined in


FIGURE 6. The presence of neutrophils correlates with BBB leakage in acute MS lesions. Confocal images show that fibrinogen (red) extravasation into the brain tissue correlates with the presence of elastase-expressing neutrophils (green) in the vicinity of CNS vessels. The postmortem tissue was obtained from an MS patient affected by acute and severe relapse episodes following cessation of natalizumab therapy. Although not the final diagnosis made by the two neuropathologists (see Materials and Methods), we acknowledge the possibility of an immune reconstitution inflammatory syndrome. Left panels show high magnification views of insets (A and C); right panels show high magnification views of (B) and (D). Nuclear counterstaining with TO-PRO-3 is shown in blue. Scale bars, 20 mm.

denced by their conspicuous presence in early inflammatory infiltrates (17, 18, 27, 74). In keeping with these data, we show that GFP+ neutrophils are present in the spinal cord of lys-eGFP-ki mice 1 d prior to the manifestation of clinical symptoms, before transiently infiltrating the parenchyma. Shortly before onset, a large number of GFP+ cells are detected in the close vicinity of blood vessels, as compared with later time points, where widespread infiltration is more evident. The presence of neutrophils in perivascular infiltrates prior to disease onset emphasizes their putative involvement in BSCB disruption, possibly by favoring interactions with cells forming the BSCB and myelin-specific T cells (16, 74). Our data show that the presence of neutrophils in the CNS parallels the leakage of vascular tracers over time during the course of the disease, as both phenomena follow similar temporal patterns. A sharp increase in tracer extravasation and GFP+ infiltration occurs as animals start displaying neurologic deficits, irrespective of the clinical severity of disease at that time. Indeed, the determinant factor influencing the extent of neutrophil infiltration in the CNS was the time of onset of motor deficits, suggesting that neutrophil infiltration was a critical triggering event in the development of the pathology. This reflects the unpredictable nature of EAE pathogenesis, as disease progression over time is not linear and animals most affected in terms of motor deficits (i.e., higher scores) have not necessarily been displaying symptoms for longer than their low-scoring littermates. Hence, the timing between disease onset and evaluation of BSCB integrity or cellular infiltration is paramount in the optics of establishing correlations between the manifestation of symptoms and the occurrence of pathological events.

FIGURE 7. The presence of neutrophils correlates with BSCB leakage in NMO patients. (A) Immunohistochemistry for elastase shows neutrophil infiltration (brown cells) in the spinal cord of NMO patients. The tissue was counterstained using hematoxylin. (B) Confocal images show that the recruitment of defensins+ neutrophils (green) to the BSCB correlates with fibrinogen (red) leakage in postmortem tissue obtained from NMO patients. Insets show an area where neutrophil proximity to endothelium associates with BSCB disruption. Corresponding high-magnification images are shown in left panels. Nuclear counterstaining with TO-PRO-3 is shown in blue. Scale bar, 50 mm.


Acknowledgments We thank Nadia Fortin, Martine Lessard, Nicolas Vallie`res, Heidi Tardent, and Gaby Enzmann for invaluable assistance in this work. We are also grateful to Dr. Denis Soulet (Centre de Recherche du Centre Hospitalier Universitaire de Que´bec—Centre Hospitalier Universite´ Laval) for providing access to a laser scanning confocal microscope.

Disclosures The authors have no financial conflicts of interest.

References 1. Hickey, W. F. 2001. Basic principles of immunological surveillance of the normal central nervous system. Glia 36: 118–124. 2. Ransohoff, R. M., and B. Engelhardt. 2012. The anatomical and cellular basis of immune surveillance in the central nervous system. Nat. Rev. Immunol. 12: 623– 635. 3. Zlokovic, B. V. 2008. The blood-brain barrier in health and chronic neurodegenerative disorders. Neuron 57: 178–201. 4. Bennett, J., J. Basivireddy, A. Kollar, K. E. Biron, P. Reickmann, W. A. Jefferies, and S. McQuaid. 2010. Blood-brain barrier disruption and enhanced vascular permeability in the multiple sclerosis model EAE. J. Neuroimmunol. 229: 180– 191. 5. Popescu, B. F., and C. F. Lucchinetti. 2012. Pathology of demyelinating diseases. Annu. Rev. Pathol. 7: 185–217. 6. Stamatovic, S. M., P. Shakui, R. F. Keep, B. B. Moore, S. L. Kunkel, N. Van Rooijen, and A. V. Andjelkovic. 2005. Monocyte chemoattractant protein-1 regulation of blood-brain barrier permeability. J. Cereb. Blood Flow Metab. 25: 593–606. 7. Zhu, W., N. R. London, C. C. Gibson, C. T. Davis, Z. Tong, L. K. Sorensen, D. S. Shi, J. Guo, M. C. Smith, A. H. Grossmann, et al. 2012. Interleukin receptor activates a MYD88-ARNO-ARF6 cascade to disrupt vascular stability. Nature 492: 252–255. 8. Cristante, E., S. McArthur, C. Mauro, E. Maggioli, I. A. Romero, M. Wylezinska-Arridge, P. O. Couraud, J. Lopez-Tremoleda, H. C. Christian, B. B. Weksler, et al. 2013. Identification of an essential endogenous regulator of blood-brain barrier integrity, and its pathological and therapeutic implications. Proc. Natl. Acad. Sci. USA 110: 832–841.

9. Tonra, J. R., B. S. Reiseter, R. Kolbeck, K. Nagashima, R. Robertson, B. Keyt, and R. M. Lindsay. 2001. Comparison of the timing of acute blood-brain barrier breakdown to rabbit immunoglobulin G in the cerebellum and spinal cord of mice with experimental autoimmune encephalomyelitis. J. Comp. Neurol. 430: 131–144. 10. Floris, S., E. L. Blezer, G. Schreibelt, E. Dopp, S. M. van der Pol, I. L. SchadeeEestermans, K. Nicolay, C. D. Dijkstra, and H. E. de Vries. 2004. Blood-brain barrier permeability and monocyte infiltration in experimental allergic encephalomyelitis: a quantitative MRI study. Brain 127: 616–627. 11. Muller, D. M., M. P. Pender, and J. M. Greer. 2005. Blood-brain barrier disruption and lesion localisation in experimental autoimmune encephalomyelitis with predominant cerebellar and brainstem involvement. J. Neuroimmunol. 160: 162–169. 12. Schellenberg, A. E., R. Buist, V. W. Yong, M. R. Del Bigio, and J. Peeling. 2007. Magnetic resonance imaging of blood-spinal cord barrier disruption in mice with experimental autoimmune encephalomyelitis. Magn. Reson. Med. 58: 298–305. 13. Christy, A. L., M. E. Walker, M. J. Hessner, and M. A. Brown. 2013. Mast cell activation and neutrophil recruitment promotes early and robust inflammation in the meninges in EAE. J. Autoimmun. 42: 50–61. 14. De Filippo, K., A. Dudeck, M. Hasenberg, E. Nye, N. van Rooijen, K. Hartmann, M. Gunzer, A. Roers, and N. Hogg. 2013. Mast cell and macrophage chemokines CXCL1/CXCL2 control the early stage of neutrophil recruitment during tissue inflammation. Blood 121: 4930–4937. 15. Kroenke, M. A., T. J. Carlson, A. V. Andjelkovic, and B. M. Segal. 2008. IL-12and IL-23-modulated T cells induce distinct types of EAE based on histology, CNS chemokine profile, and response to cytokine inhibition. J. Exp. Med. 205: 1535–1541. 16. Carlson, T., M. Kroenke, P. Rao, T. E. Lane, and B. Segal. 2008. The Th17-ELR+ CXC chemokine pathway is essential for the development of central nervous system autoimmune disease. J. Exp. Med. 205: 811–823. 17. Tran, E. H., E. N. Prince, and T. Owens. 2000. IFN-g shapes immune invasion of the central nervous system via regulation of chemokines. J. Immunol. 164: 2759–2768. 18. Wu, F., W. Cao, Y. Yang, and A. Liu. 2010. Extensive infiltration of neutrophils in the acute phase of experimental autoimmune encephalomyelitis in C57BL/6 mice. Histochem. Cell Biol. 133: 313–322. 19. Mantovani, A., M. A. Cassatella, C. Costantini, and S. Jaillon. 2011. Neutrophils in the activation and regulation of innate and adaptive immunity. Nat. Rev. Immunol. 11: 519–531. 20. Beyrau, M., J. V. Bodkin, and S. Nourshargh. 2012. Neutrophil heterogeneity in health and disease: a revitalized avenue in inflammation and immunity. Open Biol. 2: 120134. 21. Jaillon, S., M. R. Galdiero, D. Del Prete, M. A. Cassatella, C. Garlanda, and A. Mantovani. 2013. Neutrophils in innate and adaptive immunity. Semin. Immunopathol. 35: 377–394. 22. Bennouna, S., S. K. Bliss, T. J. Curiel, and E. Y. Denkers. 2003. Cross-talk in the innate immune system: neutrophils instruct recruitment and activation of dendritic cells during microbial infection. J. Immunol. 171: 6052–6058. 23. van Gisbergen, K. P., M. Sanchez-Hernandez, T. B. Geijtenbeek, and Y. van Kooyk. 2005. Neutrophils mediate immune modulation of dendritic cells through glycosylation-dependent interactions between Mac-1 and DC-SIGN. J. Exp. Med. 201: 1281–1292. 24. Goverman, J. 2009. Autoimmune T cell responses in the central nervous system. Nat. Rev. Immunol. 9: 393–407. 25. Pierson, E., S. B. Simmons, L. Castelli, and J. M. Goverman. 2012. Mechanisms regulating regional localization of inflammation during CNS autoimmunity. Immunol. Rev. 248: 205–215. 26. Ransohoff, R. M., and M. A. Brown. 2012. Innate immunity in the central nervous system. J. Clin. Invest. 122: 1164–1171. 27. Steinbach, K., M. Piedavent, S. Bauer, J. T. Neumann, and M. A. Friese. 2013. Neutrophils amplify autoimmune central nervous system infiltrates by maturing local APCs. J. Immunol. 191: 4531–4539. 28. Zhou, J., S. A. Stohlman, D. R. Hinton, and N. W. Marten. 2003. Neutrophils promote mononuclear cell infiltration during viral-induced encephalitis. J. Immunol. 170: 3331–3336. 29. Odobasic, D., A. R. Kitching, Y. Yang, K. M. O’Sullivan, R. C. Muljadi, K. L. Edgtton, D. S. Tan, S. A. Summers, E. F. Morand, and S. R. Holdsworth. 2013. Neutrophil myeloperoxidase regulates T-cell-driven tissue inflammation in mice by inhibiting dendritic cell function. Blood 121: 4195–4204. 30. Scapini, P., J. A. Lapinet-Vera, S. Gasperini, F. Calzetti, F. Bazzoni, and M. A. Cassatella. 2000. The neutrophil as a cellular source of chemokines. Immunol. Rev. 177: 195–203. 31. Abi Abdallah, D. S., C. E. Egan, B. A. Butcher, and E. Y. Denkers. 2011. Mouse neutrophils are professional antigen-presenting cells programmed to instruct Th1 and Th17 T-cell differentiation. Int. Immunol. 23: 317–326. 32. McColl, S. R., M. A. Staykova, A. Wozniak, S. Fordham, J. Bruce, and D. O. Willenborg. 1998. Treatment with anti-granulocyte antibodies inhibits the effector phase of experimental autoimmune encephalomyelitis. J. Immunol. 161: 6421–6426. 33. Misgeld, T., and M. Kerschensteiner. 2006. In vivo imaging of the diseased nervous system. Nat. Rev. Neurosci. 7: 449–463. 34. Davalos, D., J. K. Ryu, M. Merlini, K. M. Baeten, N. Le Moan, M. A. Petersen, T. J. Deerinck, D. S. Smirnoff, C. Bedard, H. Hakozaki, et al. 2012. Fibrinogeninduced perivascular microglial clustering is required for the development of axonal damage in neuroinflammation. Nat. Commun. 3: 1227. 35. Niesner, R., V. Siffrin, and F. Zipp. 2013. Two-photon imaging of immune cells in neural tissue. Cold Spring Harbor Protoc. 2013: pdb.prot073528.


asymptomatic animals because of the uncertainty pertaining to the timing of disease onset. Moreover, the use of low–molecular mass tracers combined with transcardial perfusion is prone to generate false negatives, as washout of the dye is a possibility, potentially resulting in underestimation of actual extravasation (93). The intravital imaging strategy we have employed circumvents some of these caveats, as it enables live assessment of BSCB integrity in relationship to the clinical outcome of the pathology. Experiments were carried out in EAE mice following neutrophil depletion to assess their contribution to the initial breakdown of the BSCB. Anti-Ly6G treatment delayed the manifestation of clinical symptoms of EAE and markedly reduced the severity of the disease. These findings are consistent with previous reports, in which a less specific neutrophil-depleting Ab was employed (i.e., clone RB6-8C5 instead of 1A8) (16, 32). Moreover, neutrophil depletion significantly affected BSCB permeability, as evidenced by the decrease of tracer extravasation in these mice. This lends further support to the hypothesis that early disruption of the BSCB is closely related with the infiltration of neutrophils into the CNS, although we did not investigate the underlying mechanisms. In summary, this study provides evidence that neutrophils are important players in the pathogenesis of EAE, because they are recruited in the lumbar spinal cord early in the inflammatory process, before disease onset. Neutrophil recruitment to the CNS directly parallels a loss of BSCB integrity toward small molecular tracers. Therefore, targeting early neutrophil trafficking into the inflamed tissue by, for example, blocking the IL-1R1 signaling pathway may provide beneficial clinical outcome and thus represent a valuable therapeutic strategy for neuroinflammatory disorders, as suggested by our observation that neutrophil infiltration in the cerebral vasculature of MS and NMO patients correlates with BBB leakage in acute lesions.

15

16

NEUTROPHILS DISRUPT BSCB INTEGRITY IN NEUROINFLAMMATION 62. Soulika, A. M., E. Lee, E. McCauley, L. Miers, P. Bannerman, and D. Pleasure. 2009. Initiation and progression of axonopathy in experimental autoimmune encephalomyelitis. J. Neurosci. 29: 14965–14979. 63. Larochelle, C., J. I. Alvarez, and A. Prat. 2011. How do immune cells overcome the blood-brain barrier in multiple sclerosis? FEBS Lett. 585: 3770–3780. 64. Chen, C. J., H. Kono, D. Golenbock, G. Reed, S. Akira, and K. L. Rock. 2007. Identification of a key pathway required for the sterile inflammatory response triggered by dying cells. Nat. Med. 13: 851–856. 65. Chen, C. J., Y. Shi, A. Hearn, K. Fitzgerald, D. Golenbock, G. Reed, S. Akira, and K. L. Rock. 2006. MyD88-dependent IL-1 receptor signaling is essential for gouty inflammation stimulated by monosodium urate crystals. J. Clin. Invest. 116: 2262–2271. 66. Schiffenbauer, J., W. J. Streit, E. Butfiloski, M. LaBow, C. Edwards, III, and L. L. Moldawer. 2000. The induction of EAE is only partially dependent on TNF receptor signaling but requires the IL-1 type I receptor. Clin. Immunol. 95: 117– 123. 67. Sutton, C., C. Brereton, B. Keogh, K. H. Mills, and E. C. Lavelle. 2006. A crucial role for interleukin (IL)-1 in the induction of IL-17-producing T cells that mediate autoimmune encephalomyelitis. J. Exp. Med. 203: 1685–1691. 68. Li, Q., N. Powell, H. Zhang, N. Belevych, S. Ching, Q. Chen, J. Sheridan, C. Whitacre, and N. Quan. 2011. Endothelial IL-1R1 is a critical mediator of EAE pathogenesis. Brain Behav. Immun. 25: 160–167. 69. Kitic, M., S. Hochmeister, I. Wimmer, J. Bauer, T. Misu, S. Mader, M. Reindl, K. Fujihara, H. Lassmann, and M. Bradl. 2013. Intrastriatal injection of interleukin-1b triggers the formation of neuromyelitis optica-like lesions in NMOIgG seropositive rats. Acta Neuropathol. Commun. 1: 5. 70. Enzmann, G., C. Mysiorek, R. Gorina, Y. J. Cheng, S. Ghavampour, M. J. Hannocks, V. Prinz, U. Dirnagl, M. Endres, M. Prinz, et al. 2013. The neurovascular unit as a selective barrier to polymorphonuclear granulocyte (PMN) infiltration into the brain after ischemic injury. Acta Neuropathol. 125: 395–412. 71. Daniel, P. M., D. K. Lam, and O. E. Pratt. 1981. Changes in the effectiveness of the blood-brain and blood-spinal cord barriers in experimental allergic encephalomyelitis. J. Neurol. Sci. 52: 211–219. 72. Juhler, M., D. I. Barry, H. Offner, G. Konat, L. Klinken, and O. B. Paulson. 1984. Blood-brain and blood-spinal cord barrier permeability during the course of experimental allergic encephalomyelitis in the rat. Brain Res. 302: 347–355. 73. Fabis, M. J., T. W. Phares, R. B. Kean, H. Koprowski, and D. C. Hooper. 2008. Blood-brain barrier changes and cell invasion differ between therapeutic immune clearance of neurotrophic virus and CNS autoimmunity. Proc. Natl. Acad. Sci. USA 105: 15511–15516. 74. Sayed, B. A., A. L. Christy, M. E. Walker, and M. A. Brown. 2010. Meningeal mast cells affect early T cell central nervous system infiltration and blood-brain barrier integrity through TNF: a role for neutrophil recruitment? J. Immunol. 184: 6891–6900. 75. Kermode, A. G., A. J. Thompson, P. Tofts, D. G. MacManus, B. E. Kendall, D. P. Kingsley, I. F. Moseley, P. Rudge, and W. I. McDonald. 1990. Breakdown of the blood-brain barrier precedes symptoms and other MRI signs of new lesions in multiple sclerosis. Pathogenetic and clinical implications. Brain 113: 1477–1489. 76. Stone, L. A., M. E. Smith, P. S. Albert, C. N. Bash, H. Maloni, J. A. Frank, and H. F. McFarland. 1995. Blood-brain barrier disruption on contrast-enhanced MRI in patients with mild relapsing-remitting multiple sclerosis: relationship to course, gender, and age. Neurology 45: 1122–1126. 77. Kang, E. J., S. Major, D. Jorks, C. Reiffurth, N. Offenhauser, A. Friedman, and J. P. Dreier. 2013. Blood-brain barrier opening to large molecules does not imply blood-brain barrier opening to small ions. Neurobiol. Dis. 52: 204–218. 78. Owens, T., I. Bechmann, and B. Engelhardt. 2008. Perivascular spaces and the two steps to neuroinflammation. J. Neuropathol. Exp. Neurol. 67: 1113–1121. 79. Armulik, A., G. Genove´, M. Maë, M. H. Nisancioglu, E. Wallgard, C. Niaudet, L. He, J. Norlin, P. Lindblom, K. Strittmatter, et al. 2010. Pericytes regulate the blood-brain barrier. Nature 468: 557–561. 80. Stanimirovic, D. B., and A. Friedman. 2012. Pathophysiology of the neurovascular unit: disease cause or consequence? J. Cereb. Blood Flow Metab. 32: 1207–1221. 81. Anthony, D. C., S. J. Bolton, S. Fearn, and V. H. Perry. 1997. Age-related effects of interleukin-1 beta on polymorphonuclear neutrophil-dependent increases in blood-brain barrier permeability in rats. Brain 120: 435–444. 82. Argaw, A. T., Y. Zhang, B. J. Snyder, M. L. Zhao, N. Kopp, S. C. Lee, C. S. Raine, C. F. Brosnan, and G. R. John. 2006. IL-1b regulates blood-brain barrier permeability via reactivation of the hypoxia-angiogenesis program. J. Immunol. 177: 5574–5584. 83. Reboldi, A., C. Coisne, D. Baumjohann, F. Benvenuto, D. Bottinelli, S. Lira, A. Uccelli, A. Lanzavecchia, B. Engelhardt, and F. Sallusto. 2009. C-C chemokine receptor 6-regulated entry of TH-17 cells into the CNS through the choroid plexus is required for the initiation of EAE. Nat. Immunol. 10: 514–523. 84. Pelletier, M., L. Maggi, A. Micheletti, E. Lazzeri, N. Tamassia, C. Costantini, L. Cosmi, C. Lunardi, F. Annunziato, S. Romagnani, and M. A. Cassatella. 2010. Evidence for a cross-talk between human neutrophils and Th17 cells. Blood 115: 335–343. 85. Nyga˚rdas, P. T., J. A. Maä¨tta¨, and A. E. Hinkkanen. 2000. Chemokine expression by central nervous system resident cells and infiltrating neutrophils during experimental autoimmune encephalomyelitis in the BALB/c mouse. Eur. J. Immunol. 30: 1911–1918. 86. Zehntner, S. P., C. Brickman, L. Bourbonniere, L. Remington, M. Caruso, and T. Owens. 2005. Neutrophils that infiltrate the central nervous system regulate T cell responses. J. Immunol. 174: 5124–5131. 87. DiStasi, M. R., and K. Ley. 2009. Opening the flood-gates: how neutrophilendothelial interactions regulate permeability. Trends Immunol. 30: 547–556.


36. Stromnes, I. M., and J. M. Goverman. 2006. Active induction of experimental allergic encephalomyelitis. Nat. Protoc. 1: 1810–1819. 37. Soulet, D., A. Pare´, J. Coste, and S. Lacroix. 2013. Automated filtering of intrinsic movement artifacts during two-photon intravital microscopy. PLoS ONE 8: e53942. 38. de Rivero Vaccari, J. P., D. Bastien, G. Yurcisin, I. Pineau, W. D. Dietrich, Y. De Koninck, R. W. Keane, and S. Lacroix. 2012. P2X4 receptors influence inflammasome activation after spinal cord injury. J. Neurosci. 32: 3058–3066. 39. Sedgwick, J. D., S. Schwender, H. Imrich, R. Do¨rries, G. W. Butcher, and V. ter Meulen. 1991. Isolation and direct characterization of resident microglial cells from the normal and inflamed central nervous system. Proc. Natl. Acad. Sci. USA 88: 7438–7442. 40. Babcock, A. A., W. A. Kuziel, S. Rivest, and T. Owens. 2003. Chemokine expression by glial cells directs leukocytes to sites of axonal injury in the CNS. J. Neurosci. 23: 7922–7930. 41. Barrette, B., M. A. He´bert, M. Filali, K. Lafortune, N. Vallie`res, G. Gowing, J. P. Julien, and S. Lacroix. 2008. Requirement of myeloid cells for axon regeneration. J. Neurosci. 28: 9363–9376. 42. Pineau, I., L. Sun, D. Bastien, and S. Lacroix. 2010. Astrocytes initiate inflammation in the injured mouse spinal cord by promoting the entry of neutrophils and inflammatory monocytes in an IL-1 receptor/MyD88-dependent fashion. Brain Behav. Immun. 24: 540–553. 43. Nadeau, S., M. Filali, J. Zhang, B. J. Kerr, S. Rivest, D. Soulet, Y. Iwakura, J. P. de Rivero Vaccari, R. W. Keane, and S. Lacroix. 2011. Functional recovery after peripheral nerve injury is dependent on the pro-inflammatory cytokines IL1b and TNF: implications for neuropathic pain. J. Neurosci. 31: 12533–12542. 44. Ribes, S., T. Regen, T. Meister, S. C. Tauber, S. Sch€utze, A. Mildner, M. Mack, U. K. Hanisch, and R. Nau. 2013. Resistance of the brain to Escherichia coli K1 infection depends on MyD88 signaling and the contribution of neutrophils and monocytes. Infect. Immun. 81: 1810–1819. 45. Tung, J. W., K. Heydari, R. Tirouvanziam, B. Sahaf, D. R. Parks, L. A. Herzenberg, and L. A. Herzenberg. 2007. Modern flow cytometry: a practical approach. Clin. Lab. Med. 27: 453–468, v (v.). 46. Jutila, M. A., F. G. Kroese, K. L. Jutila, A. M. Stall, S. Fiering, L. A. Herzenberg, E. L. Berg, and E. C. Butcher. 1988. Ly-6C is a monocyte/macrophage and endothelial cell differentiation antigen regulated by interferon-gamma. Eur. J. Immunol. 18: 1819–1826. 47. Hestdal, K., F. W. Ruscetti, J. N. Ihle, S. E. Jacobsen, C. M. Dubois, W. C. Kopp, D. L. Longo, and J. R. Keller. 1991. Characterization and regulation of RB6-8C5 antigen expression on murine bone marrow cells. J. Immunol. 147: 22–28. 48. Daley, J. M., A. A. Thomay, M. D. Connolly, J. S. Reichner, and J. E. Albina. 2008. Use of Ly6G-specific monoclonal antibody to deplete neutrophils in mice. J. Leukoc. Biol. 83: 64–70. 49. Tate, M. D., Y. M. Deng, J. E. Jones, G. P. Anderson, A. G. Brooks, and P. C. Reading. 2009. Neutrophils ameliorate lung injury and the development of severe disease during influenza infection. J. Immunol. 183: 7441–7450. 50. Coisne, C., and B. Engelhardt. 2010. Preclinical testing of strategies for therapeutic targeting of human T-cell trafficking in vivo. Methods Mol. Biol. 616: 268–281. 51. Vajkoczy, P., M. Laschinger, and B. Engelhardt. 2001. a4-integrin-VCAM-1 binding mediates G protein-independent capture of encephalitogenic T cell blasts to CNS white matter microvessels. J. Clin. Invest. 108: 557–565. 52. Farrar, M. J., I. M. Bernstein, D. H. Schlafer, T. A. Cleland, J. R. Fetcho, and C. B. Schaffer. 2012. Chronic in vivo imaging in the mouse spinal cord using an implanted chamber. Nat. Methods 9: 297–302. 53. Fenrich, K. K., P. Weber, M. Hocine, M. Zalc, G. Rougon, and F. Debarbieux. 2012. Long-term in vivo imaging of normal and pathological mouse spinal cord with subcellular resolution using implanted glass windows. J. Physiol. 590: 3665–3675. 54. Fleming, K. K., J. A. Bovaird, M. C. Mosier, M. R. Emerson, S. M. LeVine, and J. G. Marquis. 2005. Statistical analysis of data from studies on experimental autoimmune encephalomyelitis. J. Neuroimmunol. 170: 71–84. 55. Fabis, M. J., G. S. Scott, R. B. Kean, H. Koprowski, and D. C. Hooper. 2007. Loss of blood-brain barrier integrity in the spinal cord is common to experimental allergic encephalomyelitis in knockout mouse models. Proc. Natl. Acad. Sci. USA 104: 5656–5661. 56. Linthicum, D. S., J. J. Munoz, and A. Blaskett. 1982. Acute experimental autoimmune encephalomyelitis in mice. I. Adjuvant action of Bordetella pertussis is due to vasoactive amine sensitization and increased vascular permeability of the central nervous system. Cell. Immunol. 73: 299–310. 57. Sudweeks, J. D., J. A. Todd, E. P. Blankenhorn, B. B. Wardell, S. R. Woodward, N. D. Meeker, S. S. Estes, and C. Teuscher. 1993. Locus controlling Bordetella pertussis-induced histamine sensitization (Bphs), an autoimmune diseasesusceptibility gene, maps distal to T-cell receptor b-chain gene on mouse chromosome 6. Proc. Natl. Acad. Sci. USA 90: 3700–3704. 58. Yong, T., G. A. Meininger, and D. S. Linthicum. 1993. Enhancement of histamine-induced vascular leakage by pertussis toxin in SJL/J mice but not BALB/c mice. J. Neuroimmunol. 45: 47–52. 59. Richard, J. F., M. Roy, J. Audoy-Re´mus, P. Tremblay, and L. Vallie`res. 2011. Crawling phagocytes recruited in the brain vasculature after pertussis toxin exposure through IL6, ICAM1 and ITGaM. Brain Pathol. 21: 661–671. 60. Faust, N., F. Varas, L. M. Kelly, S. Heck, and T. Graf. 2000. Insertion of enhanced green fluorescent protein into the lysozyme gene creates mice with green fluorescent granulocytes and macrophages. Blood 96: 719–726. 61. Mawhinney, L. A., S. G. Thawer, W. Y. Lu, Nv. Rooijen, L. C. Weaver, A. Brown, and G. A. Dekaban. 2012. Differential detection and distribution of microglial and hematogenous macrophage populations in the injured spinal cord of lys-EGFP-ki transgenic mice. J. Neuropathol. Exp. Neurol. 71: 180–197.

The Journal of Immunology 88. Engelhardt, B., and R. M. Ransohoff. 2012. Capture, crawl, cross: the T cell code to breach the blood-brain barriers. Trends Immunol. 33: 579–589. 89. Allen, C., P. Thornton, A. Denes, B. W. McColl, A. Pierozynski, M. Monestier, E. Pinteaux, N. J. Rothwell, and S. M. Allan. 2012. Neutrophil cerebrovascular transmigration triggers rapid neurotoxicity through release of proteases associated with decondensed DNA. J. Immunol. 189: 381–392. 90. Stockinger, B., and M. Veldhoen. 2007. Differentiation and function of Th17 T cells. Curr. Opin. Immunol. 19: 281–286. 91. Codarri, L., G. Gy€ ulve´szi, V. Tosevski, L. Hesske, A. Fontana, L. Magnenat, T. Suter, and B. Becher. 2011. RORgt drives production of the cytokine GM-CSF

17 in helper T cells, which is essential for the effector phase of autoimmune neuroinflammation. Nat. Immunol. 12: 560–567. 92. El-Behi, M., B. Ciric, H. Dai, Y. Yan, M. Cullimore, F. Safavi, G. X. Zhang, B. N. Dittel, and A. Rostami. 2011. The encephalitogenicity of TH17 cells is dependent on IL-1- and IL-23-induced production of the cytokine GM-CSF. Nat. Immunol. 12: 568–575. 93. Hoffmann, A., J. Bredno, M. Wendland, N. Derugin, P. Ohara, and M. Wintermark. 2011. High and low molecular weight fluorescein isothiocyanate (FITC)-dextrans to assess blood-brain barrier disruption: technical considerations. Transl. Stroke Res. 2: 106–111.


Dexmedetomidine attenuates blood-spinal cord barrier disruption induced by spinal cord ischemia reperfusion injury in rats.

Mitochondrial dysfunction in demyelinating diseases.

Hantavirus-induced disruption of the endothelial barrier: neutrophils are on the payroll.

Cutaneous IgA deposits in bullous diseases function as ligands to mediate adherence of activated neutrophils.

Neuroinflammatory paradigms in lysosomal storage diseases.

Disruption of a Regulatory Network Consisting of Neutrophils and Platelets Fosters Persisting Inflammation in Rheumatic Diseases.

Blood-spinal cord barrier disruption contributes to early motor-neuron degeneration in ALS-model mice.

Immunological aspects of demyelinating diseases.

Melatonin treatment protects against acute spinal cord injury-induced disruption of blood spinal cord barrier in mice.

Blood-tumor barrier disruption controversies.

Nutritional factors and aging in demyelinating diseases.

Retinoic Acid Prevents Disruption of Blood-Spinal Cord Barrier by Inducing Autophagic Flux After Spinal Cord Injury.

Rac1 pathway after acute spinal cord injury.

Phenylbutyrate prevents disruption of blood-spinal cord barrier by inhibiting endoplasmic reticulum stress after spinal cord injury.

Structure and function of the ependymal barrier and diseases associated with ependyma disruption.

Recent Work on the Demyelinating Diseases.

Selective blood-tumor barrier disruption by leukotrienes.

Lamellar body secretory response to barrier disruption.

Psychological effects of calisthenic exercises on neuroinflammatory and rheumatic diseases.

Olfactory Pathology in Central Nervous System Demyelinating Diseases.

Neutrophils: neglected players in viral diseases.

Myelin Oligodendrocyte Glycoprotein: Deciphering a Target in Inflammatory Demyelinating Diseases.

Severe demyelinating diseases in childhood: a clinicopathological study.

[Neuroinflammatory diseases with a focus on multiple sclerosis - an overview].