Loss of blood–brain barrier integrity in the spinal cord is common to experimental allergic encephalomyelitis in knockout mouse models

  1. Marzena J. Fabis*,†,
  2. Gwen S. Scott*,‡,
  3. Rhonda B. Kean*,†,
  4. Hilary Koprowski*,†,§, and
  5. D. Craig Hooper*,†,§,
  1. *Center for Neurovirology, Department of Cancer Biology,
  2. †Biotechnology Foundation Laboratories, and
  3. Department of Neurological Surgery, Thomas Jefferson University, 1020 Locust Street, JAH 454, Philadelphia, PA 19107-6799

  1. Contributed by Hilary Koprowski, February 12, 2007 (received for review January 26, 2007)

 
Next Section

Abstract

Experimental allergic encephalomyelitis (EAE) is an inflammatory demyelinating disease of the CNS that is used to model certain parameters of multiple sclerosis. To establish the relative contributions of T cell reactivity, the loss of blood–brain barrier (BBB) integrity, CNS inflammation, and lesion formation toward the pathogenesis of EAE, we assessed the incidence of EAE and these parameters in mice lacking NF-κB, TNF-α, IFN-αβ receptors, IFN-γ receptors, and inducible nitric oxide synthase. Although increased myelin oligodendrocyte glycoprotein-specific T cell reactivity was generally associated with a more rapid onset or increased disease severity, the loss of BBB integrity and cell accumulation in spinal cord tissues was invariably associated with the development of neurological disease signs. Histological and real-time RT-PCR analyses revealed differences in the nature of immune/inflammatory cell accumulation in the spinal cord tissues of the different mouse strains. On the other hand, disease severity during the acute phase of EAE directly correlated with the extent of BBB permeability. Thus, the loss of BBB integrity seems to be a requisite event in the development of EAE and can occur in the absence of important inflammatory mediators.

In our previous studies of the pathogenesis of experimental allergic encephalomyelitis (EAE), a neuroinflammatory disease that recapitulates in mice certain features of multiple sclerosis (MS), we have shown that the disease can be prevented by inhibiting the peroxynitrite-dependent loss of blood–brain barrier (BBB) integrity without interfering with the development of antigen-specific immunity (1). The BBB is a major contributor to the maintenance of the CNS homeostasis by regulating soluble factor and cellular exchange between the CNS and the blood (2), and the loss of BBB integrity has been associated with several neurological diseases such as MS, EAE, ischemic stroke, traumatic brain injury, and others (3–5). However, enhanced BBB permeability is also associated in mice with the therapeutic clearance of neurotropic virus (6), indicating that the loss of BBB integrity is unlikely to be the sole determinant of pathogenesis in EAE and possibly MS.

One approach to assessing the contribution of an immunological pathway to the pathogenesis of EAE is to examine the onset, incidence, and severity of the disease in mice lacking relevant genes. Mice that are deficient in NF-κB, a transcription factor that contributes to the activity of a number of genes involved in immunity and inflammation, have been found to be less susceptible to the induction of EAE than their congenic controls (7). TNF-α knockout (KO) mice have generally been reported to develop EAE with an incidence comparable to congenic controls but with a more protracted onset and reduced severity (8–10). In mice lacking IFN-β, the incidence of EAE and the disease severity are greater (11). Similar results have been reported for mice lacking either IFN-γ or its receptor (12, 13).

The consequences of the deletion of inducible nitric oxide synthase (iNOS) on the development of EAE are somewhat paradoxical. Similar to MS, EAE involves the accumulation of inflammatory cells expressing iNOS (NOS-2) in CNS tissue lesions (14, 15). Evidence suggests that peroxynitrite, a downstream product of iNOS, can play an important role in the pathogenesis of MS and EAE (14, 16–18). However, mice lacking iNOS remain highly susceptible to the induction of EAE (19). These findings illustrate that the absence of any particular gene may not have the expected impact on EAE, possibly because of the complexity of the pathological mechanisms involved. In this case, more detailed studies of EAE in gene KO mouse models may help identify pathological mechanisms that are particularly relevant. To probe the relative contributions of myelin-specific T cell autoimmunity, CNS inflammation, and the loss of BBB integrity to the development of EAE, we have compared these parameters in congenic controls and mice lacking NF-κB, TNF-α, and iNOS as well as IFN-αβ and IFN-γ receptors (IFN-αβ-R and IFN-γ-R) after immunization with a myelin oligodendrocyte glycoprotein (MOG) peptide. To examine the impact of the absence of these products on gender differences in EAE, both male and female mice were studied.

Previous SectionNext Section

Results

Clinical Signs of EAE Vary in Incidence, Onset, and Severity Between Mice That Differ in Gender and the Expression of Genes Relevant to Immune Function.

The clinical course of EAE differs between mice of different genders and genetic backgrounds as well as between normal mice and those with targeted deletion of certain genes that encode a variety of products related to immunity and inflammation (Table 1). Normal B6.129 mice develop clinical signs of EAE several days before 129/SvEv mice, with males from both strains developing disease signs earlier than females. Pooled data from all of the mice studied revealed that, overall, females were significantly less susceptible to EAE (P < 0.001 for all mice, P < 0.01 for all mice exclusive of NF-κB KO, according to Fisher's test). Nevertheless, within each strain, the incidence and severity of the clinical signs of EAE in male and female mice are ultimately comparable when only the animals that developed the disease are compared. This relationship between males and females in onset, incidence, and severity of EAE is maintained in mice on a 129/SvEv background that lack IFN-αβ-R or IFN-γ-R despite the fact that the KO mice more rapidly develop more severe EAE than their wild-type counterparts. Clinical disease signs again develop earlier in the male mice, particularly those lacking the IFN-αβ-R. With the numbers of mice studied (minimum of 31 per strain), only the IFN-γ-R KO mice seemed to have a slightly higher incidence of EAE than the 129/SvEv controls (P = 0.047 according to Fisher's test for both males and females combined). Although male B6.129 mice developed EAE a couple of days before females and with slightly increased severity, no significant differences in onset or severity were seen between males and females in any of the KO strains on this background that were studied. Mice lacking iNOS developed EAE more rapidly than B6.129 controls but with no significant difference in incidence or severity and no difference between male and female mice. Although male NF-κB KO mice developed EAE several days after the controls and no delay was seen between female mice and controls, the incidence of EAE was, surprisingly, only reduced in NF-κB KO females. However, when either male or female NF-κB KO mice developed EAE, the severity of their disease was comparable to that seen in wild-type B6.129 mice. The most substantial delay in the onset of EAE, up to nearly 1 week for males, was seen in TNF-α KO mice. On average, EAE was less severe in the mice lacking TNF-α that developed the disease than in control B6.129 mice with EAE (P < 0.0001, males and females combined). However, the fact that the reduction in severity was greater in male than female TNF-α KO mice is likely because of the more severe disease seen in B6.129 males than females.

View this table:
Table 1.

Clinical disease signs in EAE-sensitized mice


MOG35–55-Specific Reactivity of Cells from Congenic and KO Mice in Vitro.

The development of EAE depends on the capacity of T cells to respond to particular myelin-associated antigens. We therefore assessed the ability of T cells from MOG-immunized animals to mediate an antigen-specific proliferative response in vitro. As shown in Fig. 1, there is considerable variability in the capacity of lymph node T cells from MOG-immunized mice of different strains to proliferate in response to MOG35-55 peptide in vitro. In general, cells from animals with a 129/SvEv background proliferate more vigorously to MOG35-55 peptide than those from mice with a B6.129 background. The MOG-specific proliferation of lymph node cells from mice deficient in either IFN-αβ-R or IFN-γ-R was significantly increased over that of similar cells from congenic controls (P < 0.001; Fig. 1 A). On the other hand, although the MOG-specific response was equivalent for lymph node cells from animals lacking iNOS and congenic wild-type controls, lymph node cells isolated from MOG-immunized NF-κB KO mice as well as from TNF-α KO mice showed a reduced in vitro response to the MOG35-55 peptide (Fig. 1 B).

Fig. 1.
View larger version:
Fig. 1.

MOG-specific reactivity of T cells from MOG-immunized KO mice in vitro. The MOG-specific proliferation of T cells from MOG-immunized 129/SvEv mice as well IFN-αβ-R KO and IFN-γ-R KO on a 129/SvEv background (A) and B6.129 mice as well as NF-κB KO, TNF-α KO, and iNOS KO on a B6.129 background (B) were assessed as detailed in Materials and Methods. The [3H]thymidine uptake of lymph node and spleen cells cultured in the absence (open bars) and presence (filled bars) of 10 μg of MOG is shown. Data are expressed as mean cpm ± SD incorporation of triplicate cultures. The pound sign denotes significant differences between the response of KO and wild-type congenic cells to antigen according to one-way ANOVA with post hoc Dunnet's test (P < 0.001).


BBB Permeability Changes in MOG-Immunized KO and Congenic Wild-Type Mice.

If BBB breakdown is an underlying event in the pathogenesis of EAE, it may be expected that gene KO mice that develop the disease differ in the extent of BBB permeability. We therefore compared the extent of BBB permeability changes in the different groups of KO and wild-type mice between days 15 and 28 postimmunization with MOG35-55 peptide. By this time, susceptible mice begin to exhibit clinical signs of EAE (see Table 1). The uptake of fluid-phase marker sodium fluorescein (NaF) into spinal cord tissues was assessed as an index of BBB permeability. BBB permeability to NaF was markedly increased in all of the strains of mice when significant clinical signs of EAE developed after MOG immunization (Fig. 2). The extent of NaF uptake after immunization did not significantly differ between wild-type congenic mice and animals lacking IFN-αβ-R, IFN-γ-R, or iNOS (Fig. 2). However, MOG-induced BBB permeability changes were significantly reduced in NF-κB KO as well as in TNF-α KO mice with EAE compared with wild-type congenic animals with the disease (P < 0.05; Fig. 2 B).

Fig. 2.
View larger version:
Fig. 2.

MOG immunization induces BBB permeability in gene KO mice. BBB permeability was assessed in nonimmune (open bars) and MOG-immunized (filled bars) IFN-αβ-R KO and IFN-γ-R KO mice and their background wild-type control 129/SvEv (A) as well as NF-κB KO, TNF-α KO, and iNOS KO mice and their background wild-type control B6.129 (B) by measuring the uptake of NaF from circulation into spinal cord tissue as detailed in Materials and Methods. Data are expressed as the mean uptake of NaF into spinal cord tissues ± SEM, where n = 4–13 mice per group. Statistically significant differences in NaF uptake determined by the Mann–Whitney test between nonimmune and MOG-immunized mice (∗∗∗, P < 0.001) and between MOG-immunized congenic wild-type and KO mice (#, P < 0.05) are indicated.


The results detailed above suggest that the considerable variability in the clinical course of EAE in MOG-immunized mice of the different gene KO strains may be accompanied by differences in the extent of BBB permeability changes. To determine whether this is the case, we examined whether there is a general correlation between BBB permeability and disease severity by using pooled data from all of the different KO strains (Fig. 3). In Fig. 3, the levels of NaF in the spinal cord tissues of individual animals from all strains are plotted against their disease score on the day of sampling. A close linear correlation exists between the extent of BBB permeability and the clinical score (r 2 = 0.989; Fig. 3).

Fig. 3.
View larger version:
Fig. 3.

Correlation between disease severity and the extent of BBB permeability in MOG-immunized mice. The extent of NaF uptake into spinal cord tissues for the mice described in the Fig. 2 legend is individually plotted against their disease severity score at the time of sample collection (filled circles). The line represents the linear relationship between the two variables determined by regression analysis, with r 2 = 0.989 (P < 0.0001).


Cell Infiltration into the Spinal Cord Tissues in KO and Congenic Wild-Type Mice After MOG Immunization.

In conventional EAE models, the loss of BBB integrity and cell infiltration into CNS tissues are closely associated events (20). To gain insight into the contribution of the nature of the CNS inflammatory infiltrate to the pathogenesis of EAE, we assessed CNS inflammation in the different mouse strains when the clinical signs of EAE were at their peak. As a general survey of immune/inflammatory cell accumulation, we first used real-time quantitative RT-PCR to determine the levels in spinal cord tissues of mRNAs specific for the T cell markers CD4 and CD8 as well as the monocyte marker CD11b (Mac-1, integrin-αM chain; Fig. 4). Surprisingly, mice on the 129/SvEv background commonly had considerably higher levels of these mRNAs in spinal cord tissues than those on a B6.129 background (Fig. 4). Nevertheless, elevations in CD4, CD8, and CD11b mRNAs in the spinal cord tissues were seen in association with clinical signs of EAE in most of the strains studied. Notable exceptions for CD4 mRNA were mice lacking IFN-γ-R and iNOS, in which significant increases were not seen. CD8 mRNA was not detected at higher levels in a variety of mice including those of the 129/SvEv strain and mice deficient in IFN-γ-R, NF-κB, and iNOS. CD11b mRNA was significantly elevated in diseased spinal cord tissues of all of the strains except for 129/SvEv, IFN-γ-R KO, and TNF-α KO.

Fig. 4.
View larger version:
Fig. 4.

General accumulation of immune/inflammatory cell markers in the spinal cord tissues of mice with EAE. Levels of mRNAs specific for the T cell markers CD4 and CD8, as well as the monocyte marker CD11b, were assessed in nonimmune (open bars) and MOG-immunized (filled bars) mice with targeted disruption in the genes for IFN-αβ-R, IFN-γ-R, TNF-α, NF-κB, and iNOS, and congenic wild-type animals were assessed by real-time quantitative RT-PCR as described in Materials and Methods. The data are expressed as the mean ± SEM copies of specific mRNA per copy of the housekeeping gene L13 mRNA. Statistical significance of the differences in mean copy number between samples from mice with EAE and nonimmune controls were tested by Student's t test: ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.


Lesion Formation in the Spinal Cords of KO and Congenic Wild-Type Mice After MOG Immunization.

As is the case for MS, lesion formation is a hallmark of the pathological changes seen in the neural tissues during EAE (14). To determine whether the absence of any of the genes studied impacts lesion formation in EAE, we performed histological analysis on spinal cord sections from KO and congenic wild-type mice that had developed clinical signs of EAE after immunization with MOG. Lesions were detectable in the spinal cord tissues of all mice that developed disease signs regardless of their immunological defect. The spinal cords of wild-type mice with either a 129/SvEv or B6.129 background had extensive lesions that penetrated deep into the white matter, and a similar pattern was seen for mice lacking IFN-αβ-R (Fig. 5 A Left and Center and B Left and Center). The lesions in the spinal cords of IFN-γ-R KO and iNOS KO mice, although also fairly extensive, appear to invade the white matter to a lesser extent, particularly in the iNOS KO mice, in which most lesions were limited to areas adjacent to the meninges (Fig. 5 A Left and B Left). The spinal cords of TNF-α and NF-κB KO mice showed limited lesion activity (Fig. 5 B Left).

Fig. 5.
View larger version:
Fig. 5.

Lesion activity in the spinal cords of gene KO mice with EAE. Sections from the lumbar region of spinal cord of MOG-immunized IFN-αβ-R KO and IFN-γ-R KO mice and their background wild-type control 129/SvEv (A) as well as NF-κB KO, TNF-α KO, and iNOS KO mice and their background wild-type control B6.129 (B) with a clinical score of three or more were stained with Harris' hematoxylin and eosin Y as described in Materials and Methods (Left). Arrows indicate areas of lesions. (Center) Enlargements (10×) of areas of the sections shown in Left exhibiting lesions. To assess the extent of cell infiltration into CNS tissues, levels of mRNAs specific for the T cell markers CD4 and CD8, as well as the monocyte marker CD11b, were determined by using real-time quantitative RT-PCR in 10 sections consecutive to those shown. Results are expressed as the mean ± SEM fold increase in specific mRNA level with the levels detected in similar sections from control, nonimmune mice taken as one (Right). The statistical significance of differences in mRNA copy numbers between spinal cord sections from mice with EAE and nonimmune controls was assessed by using Student's t test: ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.


To examine the composition of the immune/inflammatory cells accumulating in lesions, we assessed CD4, CD8, and CD11b mRNA levels in sections adjacent to those studied by immunohistochemistry. In all sections with lesions, CD4 and CD11b mRNA levels were significantly increased over those of nonimmune tissues. However, striking differences in CD8 mRNA accumulation were seen with all sections containing lesions, showing strong elevations in this marker except for those from mice lacking IFN-αβ-R, IFN-γ-R, and iNOS (Fig. 5 A Right).

Previous SectionNext Section

Discussion

Gene KO mice lacking the inflammatory mediators TNF-α and iNOS, lacking IFN receptors, or with more extensive deficits in immunity through deletion of the transcription factor NF-κB all differ from their wild-type congenic controls in susceptibility to EAE. By comparison with the disease in controls, we found EAE to be more severe in mice lacking IFN-γ-R or iNOS but less severe in mice lacking TNF-α or NF-κB, as shown in refs. 7, 12, 21, and 22. The substantial delay of ≈7 days in the mean onset of EAE in TNF-α-deficient mice is also reproducible (8). We were unable to find any reference to previous studies of EAE in IFN-αβ-R KO mice, which is surprising with respect to the fact that individuals with MS are treated with IFN-β (23, 24). Notably, in this regard we found that these mice develop EAE more rapidly and with increased incidence and severity than their congenic counterparts, supporting the notion that there is an IFN-αβ-driven mechanism that protects against the pathogenesis of CNS inflammation. Therefore, as discussed in previous publications from a number of groups, the differences between the various gene KO mice and their congenic controls are all likely to be the result of differences in proinflammatory and antigen-specific immunity (TNF-α and NF-κB) as well as the loss of regulatory pathways (IFN receptors, iNOS; refs. 7, 9, and 13).

Although the primary objective of this study was to identify mechanisms common to the development of EAE regardless of the phenotype of the mice developing the disease, comparison of EAE parameters in male versus female mice revealed some noteworthy differences. When appropriately immunized, the majority of both male and female mice of all of the strains that we tested developed clear signs of EAE; the only exception was female NF-κB KO mice, in which the incidence of disease was ≈30%. This low incidence of EAE in female NF-κB KO mice sharply contrasts with a 94% incidence in their male counterparts and reflects a general trend in our data toward a lower incidence of EAE in females. However, the clinical scores of mice that developed EAE did not differ between males and females of the individual strains except for B6.129 mice, in which the males developed slightly more extensive disease. Nevertheless, EAE generally developed more slowly in females than males in all of the mice with a 129/SvEv background as well as in wild-type B6.129 mice. Thus, although previous studies with a variety of mouse strains and immunogens have come to disparate conclusions as to whether EAE differs between male and female mice (25, 26), our findings indicate that such differences exist and are likely to depend on both the genetic makeup and immune status of the animals.

The differences in immune function and in the parameters of EAE between the different strains of mice provide us with the opportunity to identify mechanisms that are common to all mice developing this disease and, therefore, are of central importance to the development of CNS inflammatory pathology. An appropriate T cell response is clearly a prerequisite of EAE pathogenesis (27), and in general, lymphocytes from the strains of mice that developed more severe EAE mediated higher proliferative responses to MOG35-55 in vitro. However, it is evident that the relationship between disease severity and the capacity of T cells to respond to MOG35-55 in vitro is complex. For example, although female TNF-α KO and NF-κB KO mice are less likely to develop EAE than the other strains of mice tested and their T cells have a limited ability to respond to MOG35-55 in vitro, when EAE develops in NF-κB KO mice, the disease is as severe as that seen in congenic controls.

A general correlation between the appearance of clinical signs of EAE and BBB permeability changes has been noted in conventional EAE models (28–38). The loss of BBB integrity is likely to contribute to the pathogenesis of a neuroinflammatory disease by providing circulating cells and factors access to normally privileged CNS tissues and soluble CNS products access to the circulation. In every case in which mice developed clinical signs of EAE, enhanced BBB permeability was detected by the accumulation of the fluid-phase marker NaF in spinal cord tissues. If the loss of BBB integrity is a fundamental event in the initiation of EAE, it may be expected that differences in the extent of BBB permeability are more closely related to the severity of the disease in a given animal as opposed to its genetic makeup. To test this hypothesis, we examined the relationship between disease severity and the extent of BBB permeability for individual animals from all of the strains studied and found that there is a close linear relationship between these two parameters. Similar results were obtained when mice from the different strains were assessed separately with the exception of TNF-α KO mice, in which there was no correlation between the extent of permeability and disease severity (data not shown).

As noted above, the loss of BBB integrity may contribute to the pathogenesis of EAE by providing toxic soluble factors access to the CNS tissues or facilitating immune/inflammatory cell invasion. Although there has been some speculation that the nature of the CNS inflammatory response may dictate the severity of EAE (30), our findings suggest that the overall extent of inflammatory cell invasion likely does not correlate with disease severity. For example, 129/SvEv mice generally have higher levels of CD4, CD8, and CD11b mRNAs in their spinal cord tissues than B6.129 mice despite higher disease severity in the latter. Moreover, the levels of all three markers differed considerably, whereas disease severity was similar in IFN-αβ-R KO and IFN-γ-R KO mice.

The apparent lack of an association between immune/inflammatory cell accumulation in the spinal cord and clinical disease led us to examine lesion formation in these animals. Histological examination showed accumulations of immune/inflammatory cells in the spinal cord tissues of all mice that developed disease signs regardless of their immunological defect. Although there are differences in the phenotype of the lesions between the different strains, in general, mice from strains that develop more severe EAE have greater numbers of these lesions per spinal cord section. Analysis of mRNA levels revealed that elevations in CD4 and CD11b are common to regions with cell accumulations in all of the mouse strains, suggesting that these cells are likely to be essential contributors to lesion formation. Elevations in CD8 cells are absent from the accumulations of cells seen in spinal cord tissues from IFN-αβ-R and IFN-γ-R KO as well as iNOS KO mice. Because of the severity of EAE in these animals, we speculate that CD8 cells may play a regulatory role in lesion formation.

In conclusion, our findings indicate that the loss of BBB integrity in the spinal cord is a fundamental event in the pathogenesis of EAE regardless of the genetic makeup of the mouse strain studied and that the magnitude of the neurovascular defect is generally associated with the severity of the disease. Although the consequences of edema formation in the spinal cord may contribute to the disease (39–43), damage caused by invading CD4 and CD11b cells is also important, because lesion formation by these cells is a common feature of EAE.

Previous SectionNext Section

Materials and Methods

Animals.

Mice lacking IFN-αβ-R (n = 32) or IFN-γ-R (n = 63) on a 129/SvEv background were kindly provided by Michel Aguet (Swiss Institute for Experimental Cancer Research, Lausanne, Switzerland). Mice on a B6.129 background with targeted deletions in genes for the transcription factor NF-κB (n = 33), the proinflammatory cytokine TNF-α (n = 82), or iNOS (n = 153), an enzyme that catalyzes nitric oxide production by inflammatory cells, were obtained from The Jackson Laboratory. Control 129/SvEv mice (n = 50) were purchased from Taconic (Germantown, NY), and B6.129 mice (n = 195) were obtained from The Jackson Laboratory. The development of the mice used in this study has been detailed elsewhere (7, 44–47).

Induction of EAE.

EAE was induced in 8- to 10-week-old male and female mice by s.c. immunization in the right (day 0) and left (day 7) flanks with 150 μg of the MOG35-55 peptide (MEVGWYRSPFSRVVHLYRNGK) in complete Freund's adjuvant supplemented with 4 mg/ml Mycobacterium tuberculosis H37RA (Difco). Mice also received 400 ng of pertussis toxin in saline (List Biological Laboratories, Campbell, CA) i.p. on days 0 and 2 postimmunization. Mice were monitored daily for clinical signs of EAE scored on an expanded seven-point severity scale: 0 = normal mouse; 1 = piloerection, tail weakness; 2 = tail paralysis; 3 = tail paralysis plus hind-limb weakness; 4 = tail paralysis plus partial hind-limb paralysis; 5 = complete hind-limb paralysis; 6 = hind- and fore-limb paralysis; 7 = moribund/dead from EAE. All procedures were conducted in accordance with federal guidelines under animal protocols approved by the Thomas Jefferson University Institutional Animal Care and Use Committee.

In Vitro Assay of MOG35–55-Specific T Cell Reactivity.

Antigen-specific T cell proliferative responses were assessed in cells recovered from the inguinal and axillary lymph node and spleens of three to five female mice of each strain 18 days after immunization with MOG35-55. Single cell suspensions were prepared, and 2.5 × 106 cells per ml were cultured in 200-μl volumes in round-bottomed 96-well plates as described in ref. 1. The medium used was MEM, alpha modification (GIBCO/Life Technologies, Grand Island, NY), supplemented with 4 mM l-glutamine, 25 mM Hepes, 50 μM 2-mercaptoethanol, 10 μg of gentamicin, and 0.6% fresh mouse serum. Cultures were incubated for 72 h at 37°C in the presence or absence of 10 μg of MOG35-55 and then pulsed with 1 μCi (1 Ci = 37 GBq) of methyl-[3H]thymidine, specific activity 65 Ci/mmol (DuPont-NEN) for 4 h. Cells were then harvested on glass fiber filters by using a Mach III harvester 96 (Tomtec, Orange, CT), and the [3H]thymidine was incorporated into new DNA estimated by liquid scintillation counting by using a 1450 Microbeta Trilux counter (Wallac Oy, Turku, Finland).

Determination of BBB Permeability.

Fluid-phase BBB permeability was assessed as described previously by using NaF as a tracer molecule (14). Briefly, mice received 100 μl of 10% NaF (Sigma) in PBS i.p. and, 10 min later, were anesthetized, bled, and transcardially perfused with a minimum of 30 ml of PBS/heparin (1,000 units/liter) and PBS. The spinal cords were then removed and homogenized in 1.5 ml of cold 7.5% trichloroacetic acid and centrifuged for 10 min at 10,000 × g. After the addition of 0.25 ml of 5 M NaOH, the fluorescence of 100 μl of the spinal cord tissue supernatant was determined by using a Cytofluor II fluorimeter (PerSeptive Biosystems, Framingham, MA) at 485-nm excitation and 530-nm emission. NaF standards (125–4,000 μg/μl) were used to calculate NaF content in micrograms. NaF uptake from the circulation into spinal cord tissue is expressed as (μg fluorescence spinal cord/mg protein)/(μg fluorescence sera/μl blood) to normalize values for blood levels of the marker.

Histology.

Spinal cords were removed from terminally anesthetized, transcardially perfused mice and snap-frozen in Tissue-Tek O.C.T. Compound (Sakura Finetek, Torrance, CA). The lumbar regions were sectioned at 15 μm by using a Thermo Shandon (Pittsburgh, PA) cryostat, and sections were stained with Harris' hematoxylin (0.1%) and eosin Y (1%). Photographs were taken by using a Nikon (Melville, NY) digital camera on an Olympus (Center Valley, PA) BX-60 microscope.

Real-Time Quantitative RT-PCR Assessment of Cell Infiltration into CNS Tissues.

To estimate the extent of cell infiltration into CNS tissues, the levels of mRNAs specific for the T cell markers CD4 and CD8, as well as the monocyte marker CD11b (Mac-1, integrin-αM chain), were determined by real-time quantitative RT-PCR with primers and probes as described in ref. 6. In brief, spinal tissues were collected from transcardially perfused mice at 19–20 days postimmunization and used for RNA isolation with TRIzol B (Life Technologies) and DNA-free agent (Ambion, Austin, TX) or snap-frozen for sections. RNA was prepared from 10 consecutive sections by using the Absolutely RNA Nanoprep Kit (Stratagene). cDNA was synthesized from total RNA by using Moloney murine leukemia virus reverse transcriptase (Promega) and dT15 primer and then subjected to PCR with specific primers and probes, the TaqMan PCR core reagent kit (Applied Biosystems), and a Bio-Rad iCycler iQ real-time detection system. Data were calculated on the basis of the threshold cycle determined as the cycle with a signal higher than that of the background (signal detected in cycles: 2–10) plus 10× SD. To determine the number of copies of specific mRNAs in each sample, synthetic cDNA standards were made for each gene and used to produce standard curves. Data are expressed as the number of copies of specific mRNA per copy of the housekeeping mRNA L13 in a particular sample or as a fold increase in copy numbers over levels in control nonimmune samples, both normalized to L13 mRNA content.

Statistical Analysis.

Evaluation of the significance of differences between groups in mean EAE severity and NaF uptake into spinal cord was determined by using the Mann–Whitney test. Differences in the time of disease onset and in the levels of mRNAs for specific cell markers in the spinal cord tissues were assessed by using Student's t test. The significance of the MOG35-55-specific cell proliferative response was determined by using one way-ANOVA with posthoc Dunnet's test. Disease incidence between different strains was compared by using Fisher's test. In all cases, P < 0.05 was considered significant. Graphs were plotted and statistics were assessed by using GraphPad (San Diego, CA) Prism 3.0 software.

Previous SectionNext Section

Acknowledgments

This investigation was supported (in part) by a grant from the National Multiple Sclerosis Society (to D.C.H.) and a grant to the Biotechnology Foundation Laboratories Inc. from the Commonwealth of Pennsylvania.

Previous SectionNext Section

Footnotes

  • §To whom correspondence may be addressed. E-mail: douglas.hooper{at}jefferson.edu or hilary.koprowski{at}jefferson.edu

  • Author contributions: M.J.F., G.S.S., and D.C.H. designed research; M.J.F., G.S.S., R.B.K., and D.C.H. performed research; M.J.F., G.S.S., and D.C.H. analyzed data; and M.J.F., G.S.S., H.K., and D.C.H. wrote the paper.

  • ↵ ‡Present address: Department of Cellular and Molecular Medicine, University of Bristol, University Walk, Bristol BS8 1TB, United Kingdom.

  • The authors declare no conflict of interest.

  • Abbreviations:
    EAE,
    experimental allergic encephalomyelitis;
    MS,
    multiple sclerosis;
    BBB,
    blood–brain barrier;
    KO,
    knockout;
    iNOS,
    inducible nitric oxide synthase;
    IFN-αβ-R,
    IFN-αβ receptors;
    IFN-γ-R,
    IFN-γ receptors;
    MOG,
    myelin oligodendrocyte glycoprotein.
Previous Section
 

References

  1. ↵
    1. Kean RB,
    2. Spitsin SV,
    3. Mikheeva T,
    4. Scott GS,
    5. Hooper DC
    (2000) J Immunol 165:6511–6518.
    Abstract/FREE Full Text
  2. ↵
    1. Banks WA
    (1999) J Neurovirol 5:538–555.
    MedlineISI
  3. ↵
    1. Moore ACE,
    2. De Vries HE,
    3. De Boer AG,
    4. Breimer DD
    (1994) Biochem Pharmacol 47:1717–1724.
    CrossRefMedlineISI
  4. ↵
    1. De Vries HE,
    2. Kuiper J,
    3. De Boer AG,
    4. Van Berkel TJC,
    5. Breimer DD
    (1997) Pharmacol Rev 49:143–155.
    Abstract/FREE Full Text
  5. ↵
    1. Hawkins BT,
    2. Davis TP
    (2005) Pharmacol Rev 57:173–185.
    Abstract/FREE Full Text
  6. ↵
    1. Phares TW,
    2. Kean RB,
    3. Mikheeva T,
    4. Hooper DC
    (2006) J Immunol 176:7666–7675.
    Abstract/FREE Full Text
  7. ↵
    1. Hilliard B,
    2. Samoilova EB,
    3. Liu TS,
    4. Rostami A,
    5. Chen Y
    (1999) J Immunol 163:2937–2943.
    Abstract/FREE Full Text
  8. ↵
    1. Murphy CA,
    2. Hoek RM,
    3. Wiekowski MT,
    4. Lira SA,
    5. Sedgwick JD
    (2002) J Immunol 169:7054–7062.
    Abstract/FREE Full Text
  9. ↵
    1. Kassiotis G,
    2. Kollias G
    (2001) J Exp Med 193:427–434.
    Abstract/FREE Full Text
  10. ↵
    1. Kassiotis G,
    2. Kranidioti K,
    3. Kollias G
    (2001) J Neuroimmunol 119:239–247.
    CrossRefMedlineISI
  11. ↵
    1. Teige I,
    2. Treschow A,
    3. Teige A,
    4. Mattsson R,
    5. Navikas V,
    6. Leanderson T,
    7. Holmdahl R,
    8. Issazadeh-Navkas S
    (2003) J Immunol 170:4776–4784.
    Abstract/FREE Full Text
  12. ↵
    1. Willenborg DO,
    2. Fordham S,
    3. Bernard CCA,
    4. Cowden WB,
    5. Ramshaw IA
    (1996) J Immunol 157:3223–3227.
    Abstract
  13. ↵
    1. Tran EH,
    2. Prince EN,
    3. Owens T
    (2000) J Immunol 164:2759–2768.
    Abstract/FREE Full Text
  14. ↵
    1. Hooper DC,
    2. Scott GS,
    3. Zborek A,
    4. Mikheeva T,
    5. Kean RB,
    6. Koprowski H,
    7. Spitsin SV
    (2000) FASEB J 14:691–698.
    Abstract/FREE Full Text
  15. ↵
    1. Aboul-Enein F,
    2. Weiser P,
    3. Hoftberger R,
    4. Lassmann H,
    5. Bradl M
    (2006) Acta Neuropathol (Berl) 111:539–547.
    CrossRefMedline
  16. ↵
    1. Scott GS,
    2. Spitsin SV,
    3. Kean RB,
    4. Mikheeva T,
    5. Koprowski H,
    6. Hooper DC
    (2002) Proc Natl Acad Sci USA 99:16303–16308.
    Abstract/FREE Full Text
  17. ↵
    1. Hooper DC,
    2. Spitsin S,
    3. Kean RB,
    4. Champion JM,
    5. Dickson GM,
    6. Chaudhry I,
    7. Koprowski H
    (1998) Proc Natl Acad Sci USA 95:675–680.
    Abstract/FREE Full Text
  18. ↵
    1. Spitsin SV,
    2. Scott GS,
    3. Mikheeva T,
    4. Zborek A,
    5. Kean RB,
    6. Brimer CM,
    7. Koprowski H,
    8. Hooper DC
    (2002) Free Radic Biol Med 33:1363–1371.
    CrossRefMedlineISI
  19. ↵
    1. Dalton DK,
    2. Wittmer S
    (2005) J Neuroimmunol 160:110–121.
    CrossRefMedlineISI
  20. ↵
    1. Kirk J,
    2. Plumb J,
    3. Mirakhur M,
    4. McQuaid S
    (2003) J Pathol 201:319–327.
    CrossRefMedlineISI
  21. ↵
    1. Fenyk-Melody JE,
    2. Garrison AE,
    3. Brunnert SR,
    4. Weidner JR,
    5. Shen F,
    6. Shelton BA,
    7. Mudgett JS
    (1998) J Immunol 160:2940–2946.
    Abstract/FREE Full Text
  22. ↵
    1. Sahrbacher UC,
    2. Lechner F,
    3. Eugster H-P,
    4. Frei K,
    5. Lassman H,
    6. Fontana A
    (1998) Eur J Immunol 28:1332–1338.
    CrossRefMedlineISI
  23. ↵
    1. Chadha K,
    2. Weinstock-Guttman B,
    3. Zivadinov R,
    4. Bhasi K,
    5. Muhitch J,
    6. Feichter J,
    7. Tamano-Blanco RN,
    8. Abdelrahman N,
    9. Ambrus J,
    10. Munschauer F,
    11. et al.
    (2006) Arch Neurol 63:1579–1584.
    Abstract/FREE Full Text
  24. ↵
    1. Kraus J,
    2. Oschmann P
    (2006) Drug Discov Today 11:755–762.
    CrossRefMedlineISI
  25. ↵
    1. Okuda Y,
    2. Okuda M,
    3. Bernard CC
    (2002) Immunol Lett 81:25–29.
    CrossRefMedlineISI
  26. ↵
    1. Papenfuss TL,
    2. Rogers CJ,
    3. Gienapp I,
    4. Yurrita M,
    5. McClain M,
    6. Damico N,
    7. Valo J,
    8. Song F,
    9. Whitacre CC
    (2004) J Neuroimmunol 150:59–69.
    CrossRefMedlineISI
  27. ↵
    1. Chofflon M,
    2. Ben-Amor AF
    (2002) Clin Neurol Neurosurg 104:244–248.
    CrossRefMedlineISI
  28. ↵
    1. Espejo C,
    2. Penkowa M,
    3. Sáez-Torres I,
    4. Hidalgo J,
    5. García A,
    6. Montalban X,
    7. Martínez-Cáceres EM
    (2002) Exp Neurol 177:21–31.
    CrossRefMedlineISI
  29. ↵
    1. Willenborg DO,
    2. Staykova MA,
    3. Cowden WB
    (1999) J Neuroimmunol 100:21–35.
    CrossRefMedlineISI
  30. ↵
    1. Mohamed A,
    2. Tarhuni H,
    3. Dufan T,
    4. Benghuzzi H,
    5. Tucci M
    (2004) Biomed Sci Instrum 40:419–423.
    Medline
  31. ↵
    1. Paterson PY,
    2. Koh CJ,
    3. Kwaan HC
    (1987) Fed Proc 46:91–96.
    MedlineISI
  32. ↵