The JI Acurri Cytometers
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Giuliani, F.
Right arrow Articles by Yong, V. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Giuliani, F.
Right arrow Articles by Yong, V. W.
The Journal of Immunology, 2003, 171: 368-379.
Copyright © 2003 by The American Association of Immunologists

Vulnerability of Human Neurons to T Cell-Mediated Cytotoxicity1

Fabrizio Giuliani*, Cynthia G. Goodyer{dagger}, Jack P. Antel{ddagger} and V. Wee Yong2,*

* Departments of Oncology and Clinical Neurosciences, University of Calgary, Calgary, Alberta, Canada; {dagger} McGill University, Montreal Children’s Hospital Research Institute, Montreal, Quebec, Canada; and {ddagger} Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Axonal and neuronal loss occurs in inflammatory diseases of the CNS such as multiple sclerosis. The cause of the loss remains unclear. We report that polyclonally activated T cells align along axons and soma of cultured human neurons leading to substantial neuronal death. This occurs in an allogeneic and syngeneic manner in the absence of added Ag, requires T cells to be activated, and is mediated through cell contact-dependent mechanisms involving FasL, LFA-1, and CD40 but not MHC class I. Activated CD4+ and CD8+ T cell subsets are equally neuronal cytotoxic. In contrast to neurons, other CNS cell types (oligodendrocytes and astrocytes) are not killed by T cells. These results demonstrate for the first time the high and selective vulnerability of human neurons to T cells, and suggest that when enough activated T cells accumulate in the CNS, neuronal cytotoxicity can result through Ag-independent non-MHC class I mechanisms.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Multiple sclerosis (MS)3 is considered an autoimmune disease of the CNS and is characterized by demyelination and axonal loss. Neurological disability has been correlated with atrophy of the spinal cord, cerebellum, and cerebral cortex (1, 2). Magnetic resonance spectroscopy for N-acetyl-aspartate, a marker of axonal integrity, found reduced levels in brain tissue of MS patients. The decrease of N-acetyl-aspartate per lesion volume was significantly higher in secondary progressive (SP)MS patients with irreversible neurological disability than was in relapsing remitting (RR)MS patients with little or fixed disability (3).

Axonal injury occurs early in the course of MS and correlates with the extent of inflammation within the CNS (4, 5). This injury leads to the transection of axons and to formation of axon spheroids at their proximal ends (2). Recently, neuronal death in cortical and thalamic MS lesions was reported (6, 7); the extent of neuronal loss in the thalamus exceeded 30% compared with controls (7). In an animal model of MS, experimental autoimmune encephalomyelitis, demyelination of the optic nerve, and concomitant death of retinal ganglion neurons was observed (8).

The morphology of MS lesions suggests an important role of T cells in the disease. Active MS lesions are characterized by perivascular lymphocyte infiltration that spreads also into the parenchyma. A prevailing hypothesis of MS pathogenesis (9, 10) is that T cells are activated in the periphery by unknown Ags; both myelin and nonmyelin Ags could play a role (11, 12, 13, 14). When activated, T cells cross the blood-brain barrier where they accumulate, proliferate in response to Ag restimulation (15, 16, 17), and secrete proinflammatory cytokines that induce a chain reaction with the activation of resident microglia, infiltrating macrophages, and B cells. This inflammatory response is thought to damage axons and myelin either through Ag-specific or bystander mechanisms.

There has been limited information on whether and how T cells may kill neurons. The first evidence of the susceptibility of neurons to T cell cytotoxicity was shown using mouse peripheral nervous system neurons in vitro (18); only allogeneic T cells were cytotoxic to neurons. More recently, Ag-specific T cells were demonstrated to be toxic to syngeneic mouse neurons; however, this required the expression of MHC class I (MHC-I) on neurons by prior treatment with tetrodotoxin and IFN-{gamma} (19, 20).

Because neurons in the CNS express minimal or negligible levels of MHC-I (21, 22), and given that many infiltrating T cells in MS are Ag nonspecific (23), we have investigated other means of producing neuronal cytotoxicity. We demonstrate that polyclonally activated T cells have a potent cytotoxic effect on human neurons. This occurs in an allogeneic and syngeneic system in the absence of added Ag, requires the activation of T cells, and is mediated through contact-dependent non-MHC-I mechanisms. These results demonstrate for the first time the high and selective vulnerability of human neurons to T cells and suggest that when enough activated T cells accumulate in the CNS, neuronal cytotoxicity can result.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Isolation of T cells

Mononuclear cells were isolated from the blood of healthy volunteers by Ficoll-Hypaque centrifugation as previously described (24, 25, 26) and suspended in serum-free AIM-V medium (Life Technologies, Burlington, Ontario, Canada). To activate T cells, 1 µg/ml of an anti-CD3 Ab (OKT3) was added once for a period of 3 days. The floating cells were then removed from any adherent monocytes, and a fixed density was then used for testing cytotoxicity. Unless otherwise stated, the density of anti-CD3-activated T cells was 100,000 in 100 µl of AIM-V, and this was added to 50,000 neurons previously prepared (see below). Some mononuclear cell preparations did not receive OKT3, and the floating cells that were collected 3 days after were referred to as unactivated T cells. Flow cytometry analyses of the floating cells collected after 3 days of initiation of OKT3 treatment indicated that CD3+ T cells constituted over 90% of the total cell population; these were ~60% CD4+ and 40% CD8+ in cell ratio. B lymphocytes (CD19+) and NK cells (CD56+) consisted of the rest of the floating cell population and no monocytes (CD14+) were detected; when analyzed further, NK cells constituted <3% of the population. There was no significant difference in the proportion of the various cell subsets between the unactivated and activated lymphocyte populations. Henceforth, given that the majority of cells in the floating population are T cells, these will be referred to as T lymphocytes. T lymphocytes were also obtained from the spleen of fetuses and tested on syngeneic neurons. Each spleen was diced into small cubes and then placed on a filter containing 130-µm pore size. Tissue fragments were mechanically dissociated into single cells that were collected into PBS. Cell suspensions were then subjected to Ficoll-Hypaque centrifugation, and the T cells were then cultured and activated as previously described for cells from adult donors. Using flow cytometry analyses, the purity of CD3+ T cells after 3 days of culture was in excess of 90%.

In experiments that involved T cell subsets, mononuclear cells were first isolated from the blood of healthy adult volunteers by Ficoll-Hypaque centrifugation and then suspended in AIM-V medium as previously described. The T cells were then subdivided into CD3+, CD4+, or CD8+ populations using magnetic separation columns (Minimacs; Miltenyi Biotec, Auburn, CA). The subsets were activated by incubation for 72 h in control plates or anti-CD3-coated plates (BD PharMingen, Bedford, MA) at a concentration of 200,000 cells per well.

Human fetal neuron and astrocyte cultures

Brain and spleen tissues from human fetuses of 10–18 wk fetal age were obtained following therapeutic abortion according to guidelines approved by local institutional ethics committees. For the preparation of brain cell cultures, 5–15 g of brain tissue diced into fragments of ≤1 mm with a pair of scalpels was incubated in 40-ml volume for 15 min at 37°C with 0.25% trypsin and 200 µg/ml DNase I in PBS. The suspension was then washed through a filter of 130-µm pore size, and the filtrate was centrifuged at 1200 rpm for 10 min. The cell pellet was resuspended in PBS and centrifuged. Following a final washing step in feeding medium (see below), the pellet was suspended in feeding medium and cells were plated into T-75 flasks coated with 10 µg/ml polyornithine. Plating density was 50 million cells in 25 ml of medium. Feeding medium was MEM-supplemented with 10% FBS, 20 µg/ml gentamicin, 0.1% dextrose, 1x nonessential amino acids, 10 µM glutamine, and 1 mM sodium pyruvate. All medium constituents were from Life Technologies.

To obtain neuron-enriched cultures, cells in the T-75 flasks were subjected, immediately upon seeding, to 3 cycles of 25 µM cytosine arabinoside (Sigma-Aldrich, St. Louis, MO) to kill dividing astrocytes. Each cycle consisted of 2 days of treatment with cytosine arabinoside, followed by 3 days of noncytosine arabinoside containing feeding medium. Neuron-enriched cultures, in excess of 90% purity and containing <5% of astrocytes (27), were then plated onto 16-well Lab-tek (Nunc, Naperville, IL) slides or into wells of 96-well plates at a density of 50,000 cells per 100 µl of medium per well. Another treatment with cytosine arabinoside ensued, and cells were used for cytotoxicity assays at least 2 days after removal of cytosine arabinoside-containing medium. Astrocyte cultures of over 95% purity were obtained as previously described (27).

Isolation of human oligodendrocytes

Specimens were obtained during surgery for epilepsy. Viable cells were isolated using Percoll gradient centrifugation, and oligodendrocytes of over 95% purity were obtained as detailed elsewhere (28).

Immunocytochemistry and evaluation of cytotoxicity

Following fixation with 4% paraformaldehyde in PBS, astrocytes were identified with a rabbit polyclonal Ab to glial fibrillary acidic protein (GFAP) (1:100 dilution; DAKO, Carpinteria, CA), which is an intermediate filament protein specific to astrocytes in the CNS. Oligodendrocytes were identified with a mouse mAb to galactocerebroside (29). Neurons were labeled by a mouse mAb to microtubule-associated protein (MAP)-2 (1:5000; Sigma-Aldrich) or by a rabbit polyclonal Ab to {tau} protein (1:400; Sigma-Aldrich). MHC-I expression was identified by a mouse anti-human HLA-ABC Ab (2.5 µg/ml, W6/32 clone; Serotec, Oxford, U.K.). MHC class II (MHC-II) expression was assessed using a mouse anti-human HLA-DR, -DP, and -DQ Ab (2.5 µg/ml, Tu39 clone; Research Diagnostics, Flanders, NJ).

Appropriate secondary Abs conjugated to Cy3 or Alexa 488 (Molecular Probes, Leiden, The Netherlands) were then employed and immunofluorescence was analyzed by using a Leica (Deerfield, IL) immunofluorescence microscope. Cultures were counterstained with Hoescht dye (33258) to label all nuclei.

To evaluate cytotoxic effects on neurons and to achieve consistency in quantitation between samples, each well was assessed along its equatorial axis counting six random fields using the x40 objective. All the MAP-2-positive cells in each field were enumerated. For every condition in each set of experiments, 4 wells were counted to obtain the mean number of neurons remaining in culture. The mean value for control neuronal cultures (i.e., wells not exposed to T cells) was thus obtained, and results from experimental groups were then expressed as a percentage of that of control cultures. Finally, each bar in the figures represents the mean ± SD of the results from 4 wells, and at least 800 cells were counted in the control cultures. All results were reproduced in at least three sets of experiments.

Annexin V immunohistochemistry was used to address whether some neurons were undergoing death through apoptosis. Following their encounter with activated T cells, unfixed cultures were exposed to Annexin V-Fluos (1:50 dilution; Roche Diagnostics, Mannheim, Germany) for 15 min and photographed.

MTT viability assay

In some experiments, neuronal toxicity was evaluated using the MTT spectrophotometric assay. Neurons in 96-wells plates were exposed to test agents in 100 µl culture medium for 48 h. Twenty microliters of 0.5 mg/ml solution MTT was then added per well. Following incubation for 1 h at 37°C, and after two washes with PBS, 200 µl DMSO was added for 10 min. The amount of solubilized MTT product was read on a spectrophotometer at 550 nm.

Cytokines

Cytokines and chemokines were tested alone or in different pairs. These were recombinant human (rh)IL-1{beta} (BioSource International, Camarillo, CA), rhTNF-{alpha} (BD Biosciences, Bedford, MA), rhIFN-{gamma} (Boehringer Mannheim, Mannheim, Germany), rhIL-12 (R&D Systems, Minneapolis, MN), rhGM-CSF (Upstate Biotechnology, Lake Placid, NY), recombinant human monocyte chemoattractant protein 1 (CCL-2; R&D Systems), recombinant human monocyte chemoattractant protein 3 (CCL-7; R&D Systems), recombinant human inflammatory protein 10 (CXCL-10,; R&D Systems), and recombinant human macrophage inflammatory protein 1{alpha} (CCL-3; R&D Systems).

Abs tested for their capacity to attenuate T cell cytotoxicity

The following mAbs were used: mouse anti-human CD40 (Genzyme, Cambridge, MA), mouse anti-human CD95 ligand (FasL; BD PharMingen), mouse anti-human CD11a (LFA-1; BD PharMingen) and mouse anti-human ICAM-1 (Chemicon International, Temecula, CA). These Abs were used at a concentration of 5 µg/ml. The function-blocking mouse anti-human MHC-I (W6/32 clone) or MHC-II (Tu39) Abs were used at a final concentration of 10 µg/ml each. T cells or neurons were incubated for 30 min with the respective Ab, following which excess Abs were washed off using PBS before the coculture.

To ensure that the anti-MHC-I Ab (W6/32) was indeed a function-blocking Ab, the MHC-I-restricted killing by CD8+ T cells of the LY8 cell line, a non-Hodgkin B cell lymphoma, was used. A 51Cr release assay was used to measure cytotoxicity as previously described (30).

Statistical analysis

Because all experiments involved multiple groups, statistical analyses were conducted using one-way ANOVA with Tukey’s post hoc test. Statistical analysis was performed using SPSS version 11.0 (SPSS, Chicago, IL).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Neurons are highly susceptible to T cell cytotoxicity

We first determined whether T cells could kill neurons. Human fetal neurons were cocultured with unactivated or anti-CD3-activated T cells from adult human donors without any apparent neurological diseases. After 24 h, cultures were stained for MAP-2 because the disappearance of MAP-2 immunoreactivity is associated in vivo and in vitro with neuronal injury and death (31, 32, 33, 34, 35, 36, 37). We found that while unactivated T cells did not produce neuronal death, activated T cells consistently resulted in a significant loss of neurons (Fig. 1 and Fig. 2). The mode of cell death for some neurons was apoptotic, as these cells were annexin V-positive before their detachment from the culture substrate (Fig. 2D). Fig. 3A documents the number of neurons remaining in culture expressed as a percentage of that in control conditions (i.e., neurons not exposed to T cells). Although unactivated T cells did not produce loss of neurons, activated T cells from three adult donors caused the death of 60–75% of neurons by 24 h.



View larger version (40K):
[in this window]
[in a new window]
  		FIGURE 1. MAP-2 immunoreactivity documents the extensive loss of neurons exposed to activated T cells for 24 h. A, A high magnification micrograph of neurons alone in culture demonstrates their well-developed morphology (x1000 original magnification). The insert is an example of GFAP-positive astrocytes that are found in <5% purity in the enriched neuronal cultures. B, A low magnification micrograph of neurons alone (x200 original magnification) serves as a comparison of neuronal density for C and D. C, Neurons cocultured with unactivated T cells with similar cellular density to B (x200). D, Neurons cocultured with activated T cells (x200) where few cells remain.

 


View larger version (133K):
[in this window]
[in a new window]
  		FIGURE 2. Coculture of neurons and activated T cells visualized by phase-contrast microscopy. A, Neurons alone. B, Neurons following treatment with unactivated T cells. C, Neurons cocultured with activated T cells for 24 h where few neurons remain. Few T cells are observed in B and C because the frames were captured after the cultures had been fixed and washed, which removed the T cells. Original magnification is x200. D, Annexin V-positive neurons indicative of apoptotic cells can be found in cultures exposed to activated T cells. Original magnification is x1200.

 


View larger version (21K):
[in this window]
[in a new window]
  		FIGURE 3. Quantitative loss of neurons exposed to activated T cells. A, Comparison of the number of MAP-2 neurons that remain after 24 h of allogeneic neurons-T cells coculture. The results of activated (act.) or unactivated (unact.) T cells from three different donors on neuronal toxicity are displayed; similar results were seen in at least 10 other experiments involving different neuronal and leukocyte samples. Furthermore, while most neuronal cultures were used 2 wk after isolation, neurons that were "aged" in vitro for 1–2 mo were equally susceptible to toxicity by T cells. All results have been normalized to the number of neurons present in control cultures that were not exposed to T cells. B, Time course of neuronal loss when 100,000 allogeneic T cells were added to 50,000 neurons. C, Different numbers of activated T cells were added to 50,000 neurons and the number of surviving neurons was assessed 24 h after. A total of 100,000 unactivated T cells were also used. All results were normalized to neuronal cultures that were not exposed to T cells. One-way ANOVA, Tukey’s post hoc test: *, p < 0.005.

 
Besides MAP-2 immunoreactivity, two other means to detect cytotoxicity were used. These methods involved the detection of free nucleosomes liberated into the cytoplasm of dying cells or of cellular MTT levels. It was found that the nucleosome content was consistently increased in neurons exposed to activated, but not unactivated, T cells compared with neurons alone (data not shown). Furthermore, MTT analyses showed a marked reduction in cultures exposed to activated T cells (data not shown).

We evaluated the rapidity of the onset of T cell cytotoxicity. We found that the number of MAP-2-positive cells started to decrease significantly by 3 h of coculture and progressively declined over the 72 h of study (Fig. 3B). Thus, the manifestation of T cell toxicity occurs quickly. In contrast to activated T cells, unactivated T cells did not reduce neuronal numbers at all the time points examined. To assess the relative ratio of T cells to neurons required to produce neuronal death, 50,000 neurons were coincubated with different numbers of T cells. Although 10,000 T cells were observed to reduce neuronal numbers, a statistically significant result was obtained when the number of T cells was increased to 50,000 or 100,000 T cells with a ratio of T cells-neurons of 1:1 and 2:1, respectively (Fig. 3C).

In contrast to the T cell-mediated killing, monocytes seeded at a concentration of 100,000 cells (ratio of monocytes-neurons of 2:1) were not able to kill neurons. Activation of monocytes with LPS (100 ng/ml) did not further elicit any toxicity to neurons (data not shown).

To evaluate which T cell subset was responsible for the neuronal cytotoxicity, human fetal neurons were cocultured for 24 h with CD4+ or CD8+ subsets obtained through magnetic separation. Both subsets were able to kill neurons after anti-CD3 activation, producing a reduction of >80% of neurons compared with untreated cultures or neurons cocultured with unactivated CD4+ or CD8+ T cells (Fig. 4).



View larger version (25K):
[in this window]
[in a new window]
  		FIGURE 4. Both activated CD4+ and CD8+ T cells can promote extensive neuronal loss. Comparison between the percentage of MAP-2-positive cells that remains after 24 h of neurons-T cells coculture show that while unactivated (unact.) T cells did not kill neurons, and both the CD4+ and CD8+ activated (act.) subsets were equipotent as the CD3+ whole T cell population. Prior BrdU proliferation assays performed to assess the activation status of cells showed a significant increase in cell proliferation of microbeads-separated cells only after anti-CD3 treatment; these data suggest that microbead purification did not activate T cells nonspecifically or contribute to neuronal toxicity unless the purified populations received anti-CD3 treatment. One-way ANOVA, Tukey’s post hoc test: *, p < 0.005

 
The previous results were from fetal neurons cocultured with T cells from adult donors. The MHC mismatch might have resulted in an MHC-I-mediated killing such as that occurring in a graft-vs-host reaction. Thus, we obtained T cells from the spleen of the same fetal brain donor. Neurons cultured with activated syngeneic splenocytes demonstrated an over 60% reduction in cell numbers compared with control neurons or neurons exposed to unactivated syngeneic T cells (Fig. 5A).



View larger version (31K):
[in this window]
[in a new window]
  		FIGURE 5. A, Syngeneic-activated T cells also promote the loss of neurons. Results of splenocytes and neurons from two fetuses are displayed. As in allogeneic cocultures, unactivated T cells had no obvious toxicity. One-way ANOVA, Tukey’s post hoc test: *, p < 0.005. B, A function-blocking Ab (10 µg/ml) to MHC-I did not prevent neuronal loss attributed to activated T cell cytotoxicity. As previously noted, this Ab attenuated an MHC-I-mediated CD8+ cytotoxicity to a B cell line as evident using a 51Cr release assay. C, Consistent with the lack of protection of neurons by the anti-MHC-I Ab, there was no apparent expression of MHC-I (green) on neurons (red, neuronal soma were stained by an anti-{tau} Ab), as no yellow stain (mixture of red and green) was discernible on the surface of neurons. In contrast, the underlying astrocytes were MHC-I immunoreactive. Note that for this demonstration, a mixed culture containing abundant astrocytes was used as the astrocytes served as positive controls for the MHC-I staining.

 
To provide more compelling evidence that T cell-mediated toxicity is not through an MHC-I mechanism, neurons were treated with an MHC-I function-blocking Ab at the concentration of 10 µg/ml for 30 min before their coculture with activated T cells. Neuronal loss was not prevented by this Ab (Fig. 5B) nor was it attenuated by the MHC-II function-blocking Ab (data not shown). That the MHC-I blocking Ab is efficacious in the first place is demonstrated by its ability to attenuate an MHC-I-mediated CD8+ cytotoxicity to a B cell line as evident using a 51Cr release assay. Here, a 55% specific killing of the LY8 B cell line by CD8+ T cells is attenuated to 15% by 10 µg/ml of the anti-MHC-I Ab. To confirm the efficacy of the MHC-II Ab, we assessed its ability to block the proliferation of glatiramer acetate-specific T cell lines in response to class II-dependent Ag presentation. In these experiments the Ab was able to reduce the proliferation by 80% compared with control cocultures.

Consistent with the lack of involvement of MHC-I in the killing of neurons by T cells is the finding that human neurons in culture do not express MHC-I unlike their astrocyte counterparts (Fig. 5C). In summary, neurons are highly susceptible to toxicity mediated by T cells, as long as the T cells are activated. There is no obvious difference in neuronal death mediated by allogeneic or syngeneic T cells. Activated CD4+ and CD8+ T cell subsets are equally cytotoxic for neurons. Once activated, T cells are not dependent on MHC-dependent mechanisms of cytotoxicity.

Cell-cell interaction is required for the neuronal killing by cytotoxic T cells

Upon activation, T cells produce various soluble factors including cytokines that potentially can damage target cells. There is conflicting data on the neuroprotective vs neurotoxic effect of proinflammatory cytokines both in vivo and in vitro (38). To assess the effect of cytokines in our model, we added different concentrations and combinations of proinflammatory cytokines and chemokines to neurons. No cell death was evident by MTT assay (Fig. 6, A and B).



View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 6. Soluble factors do not mediate toxicity to neurons. A and B, Representing MTT results from two separate experiments, cytokines or chemokines in pairs (100 U/ml each) or alone did not result in neuronal toxicity. The positive control staurosporine (SP, 4 µM), a nonspecific kinase inhibitor, resulted in extensive death. One-way ANOVA, Tukey’s post hoc test: *, p < 0.05. C, The presence of activated T cells is required to kill neurons and this cannot be reproduced by conditioned medium (CM) collected from activated T cells. One-way ANOVA, Tukey’s post hoc test: *, p < 0.005.

 
Human neurons could be susceptible to a combination of inflammatory cytokines (39) and thus toxicity might not have manifested when we applied cytokines individually or in pairs. To address this, the conditioned medium from activated T cells, which contains a variety of cytokines, was tested to determine whether this was toxic to neurons. There was absence of killing when neurons exposed to the conditioned medium were assessed after 24 h of culture (Fig. 6C). Finally, in Transwell experiments where T cells and neurons were exposed to each of the other soluble factors, but where both cell types were physically separated by a filter, no killing of neurons was evident (data not shown).

The data from the cytokine and conditioned medium studies suggest that T cell cytotoxicity was likely mediated by a cell-cell contact mechanism. In support, we noted that when T cells were cocultured with neurons, they aggregated around neuronal soma and neurites; importantly, where T cells were in contact with neuronal elements, the latter lost their MAP-2 immunoreactivity (Fig. 7).



View larger version (69K):
[in this window]
[in a new window]
  		FIGURE 7. Contact between activated T cells and neuronal structures results in the disappearance of MAP-2 immunoreactivity. A, A neurite (arrow) surrounded by activated T cells (arrow) is negative for MAP-2, whereas a neurite that is not (arrowhead) continues to be MAP-2-positive. A, MAP-2 immunofluorescence. B, Corresponding phase-contrast micrograph. The morphology of T cells suggests that they are likely also killed following the encounter with neurons.

 
T cell-mediated toxicity can be neutralized

To investigate whether neuronal death induced by T cells is cell-cell contact dependent, we investigated whether blocking cell surface molecules could attenuate the T cell toxicity. We used neutralizing Abs to selected molecules and focused on those implicated in T cell-mediated killing in other systems. Thus, T cells were treated for 30 min with 5 µg/ml of function-blocking Abs to FasL, LFA-1, or CD40, then washed and cocultured with neurons for 24 h. These treatments protected neurons as 70–80% remained viable, compared with <20% in cultures not exposed to Abs (Fig. 8). The selectivity of the anti-FasL, anti-LFA-1, or anti-CD40 Abs in protecting neurons was indicated by the lack of effect of anti-ICAM-1 pretreatment of T cells to confer neuroprotection.



View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 8. The toxicity of T cells can be neutralized. Whereas activated T cells result in 80% loss of neurons in 24 h, this toxicity was attenuated by the treatment of T cells with Abs to LFA-1, FasL, and CD40, but not ICAM-1. Treatment of neurons with anti-ICAM-1 or CD40 also prevented the subsequent loss of neurons upon coculture with activated T cells. One-way ANOVA, Tukey’s post hoc test: *, p < 0.005.

 
In parallel experiments, neurons were treated for 30 min with 5 µg/ml anti-ICAM-1 or anti-CD40. These Abs were then removed through washing, and activated T cells were added to the neuronal cultures. Both Abs protected neurons from T cell-mediated cytotoxicity (Fig. 8).

Because the results of the Ab blocking experiments suggested the presence of ICAM-1 and CD40 on neurons, we used immunohistochemistry to determine whether these molecules could be detected on human neurons in culture; a mixed culture of neurons and astrocytes was used to maximize the possibility of detection of these molecules on either cell type. Fig. 9 shows that neurons were positive for CD40 although not all neurons expressed this molecule. Similarly, neurons also expressed ICAM-1, although the level tended to be lower than that of astrocytes (Fig. 9).



View larger version (56K):
[in this window]
[in a new window]
  		FIGURE 9. Human neurons in culture express CD40 and ICAM-1. A mixed culture of neurons and astrocytes was used to maximize the possibility of detection of CD40 and ICAM-1 on either cell type. A–C, Shows that some neurons are positive for CD40: phase contrast (A), MAP-2 (B), CD40 (C). Similarly, neurons also expressed ICAM-1: phase contrast (D), MAP-2 (E), ICAM-1 (F), although the levels tended to be lower than that of astrocytes (flat cell). Some neurons positive for either CD40 or ICAM-1 are indicated by arrows in B and E, respectively.

 
In summary, the toxicity of activated T cells on neurons requires cell-cell contact, and is mediated in part through FasL, LFA-1, and CD40.

T cell cytotoxicity is selective for neurons

We investigated whether activated T cells indiscriminately destroy all CNS constituents or whether toxicity was limited to neurons. Thus, T cells were incubated with human oligodendrocytes or astrocytes. No death was evident for oligodendrocytes as determined by morphologic analysis and cell counts (Fig. 10). When T cells were cocultured with astrocytes, there was no obvious reduction in astrocytic cell numbers (Fig. 10); however, the morphology of astrocytes was altered in that cells converted from a flat, fibroblast-like morphology to those with multiple stellate processes (Fig. 10D), indicative of a more reactive state.



View larger version (63K):
[in this window]
[in a new window]
  		FIGURE 10. There is no obvious reduction in cell numbers of oligodendrocytes and astrocytes when they are exposed to activated allogeneic T cells. A, Oligodendrocytes alone. B, Astrocytes alone. C, Oligodendrocytes cocultured with activated T cells for 24 h. D, Astrocytes cocultured with activated T cells for 24 h. Oligodendrocytes were identified by an Ab to galactocerebroside while astrocytes were marked by an Ab to GFAP. E, Quantitative cell counts demonstrate that oligodendrocytes and astrocytes are not lost when exposed for 24 h to unactivated or activated allogeneic T cells. All results have been normalized to oligodendrocyte or astrocyte cultures not exposed to T cells ("cells alone"). There is no statistical significance when analyzed by one-way ANOVA. A longer incubation period of 48 h with T cells also did not result in toxicity to oligodendrocytes and astrocytes.

 
Thus, the toxicity of activated T cells is selective for neurons, and oligodendrocyte and astrocyte numbers are not obviously altered upon confrontation with T cells during the period (24–48 h) of study.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
It has become evident in recent years that many inflammatory diseases of the CNS, including MS, Rasmussen’s encephalitis, and paraneoplastic encephalomyelitis, have a substantial degenerative component where neurons and axons are lost in large numbers (2, 6, 7, 40, 41, 42). This degenerative process is of crucial importance in determining clinical outcomes. In MS, the loss of neurons and axons beyond a critical threshold likely accounts for conversion of the relapsing remitting form of the disease to SPMS with permanent disabilities (43). Approximately 70% of patients with RRMS convert to SPMS at some stage in their disease (44). Furthermore, in a cohort of patients followed for 5 years the SPMS patients had a higher inflammatory activity during the time before the transformation from RRMS compared with patients still in the relapsing-remitting phase (45).

The direct cause of the neuronal and axonal injury has not been clear although this occurs in regions of accumulation of various inflammatory cell types (3, 5, 6, 46, 47, 48). Recently, it was reported that an average of 7% of CD8+ T cells within the CNS parenchyma of Rasmussen’s encephalitis were present in close apposition to neurons, and even appeared to invade them (42). Despite the correspondence of neuronal and axonal loss with inflammation, a direct cause and effect relationship of inflammatory cells destroying neurons has not been demonstrated.

In this study, by using defined sets of neural cultures, we describe that T cells are potent effectors of neuronal death. This occurs rapidly and necessitates that T cells are activated. Both CD4+ and CD8+ T cell subsets were neurotoxic. Unlike other examples where Ag-specific T cells were lethal to mouse neurons only upon the induced expression of MHC-I on neurons by prior treatment with tetrodotoxin and IFN-{gamma} (19, 20, 21), there was no such requirement for MHC-I or Ags in this study so long as the T cells have been polyclonally activated through anti-CD3 treatment. A 1:1 ratio of T cells-neurons is sufficient for toxicity to manifest. This ratio is achievable in vivo because many T cells are known to cluster around neurons in neuroinflammatory diseases of humans and experimental models (41, 42).

It is noted that the T cells used in this study were not pure populations and it is conceivable that the few contaminating non-T cell types may be responsible for the cytotoxicity observed. This is unlikely because these contaminating cell types would also be present in the non-anti-CD3-treated preparations, which were not cytotoxic. Furthermore, CD3+-selected T cells were similarly cytotoxic to neurons (Fig. 4). Moreover, human T cell lines were also able to destroy human neurons in our studies (F. Giuliani, A. Bar-Or, and V. W. Yong, unpublished observations). Nonetheless, future studies will aim to derive pure primary T cells from human blood to obviate any possibility of other cell types having a cytotoxic response on human neurons.

We note that the neuronal cultures used in this study were over 90% pure. Most of the contaminating cell types are astrocytes and it is plausible that the killing of neurons might have been mediated through astrocytes. This is unlikely for at least two reasons. First, in a few cultures in which there were virtually no astrocytes, neuronal toxicity was still elicited by activated T cells. Second, the T cell-neuron contact in Fig. 7A, which resulted in the loss of MAP-2 staining in neurites, occurred in the absence of any astrocytes.

The T cell cytotoxicity in this study requires cell-cell contact and cannot be reproduced by soluble products such as cytokines. In this regard, our experiments do not confirm the data of others that proinflammatory cytokines such as TNF-{alpha} can kill human fetal neurons in vitro (39). In support of a contact-mediated mechanism, our results show that T cells localized to the soma and processes of neurons causing the disappearance of MAP-2 staining (Fig. 7). The rapid loss of MAP-2 immunoreactivity in injured axons has been attributed to the activation of calpains, which promptly proteolyse MAP-2 leading to its nondetectability (49). Cell surface molecules that contributed to the T cell toxicity to neurons included FasL, LFA-1/ICAM-1, and CD40, but not MHC-I or MHC-II; of note, CD40 and ICAM-1 were detected on some but not all neurons (Fig. 9). We do not yet know the mechanisms by which the engagement of these molecules contribute to neuronal death. However, it has been reported that Fas-FasL interaction (50), or CD40 ligation (51), can alter intracellular pathways that regulate apoptosis of cells.

It should be noted that some neurons in vivo may express MHC molecules, particularly during a disease process. For example, neurons in Rasmussen’s encephalitis tissue (42) or following infection with Theiler’s virus (52) have been shown to express components of MHC-I, indicating that neurons in vivo can be susceptible to MHC-I-dependent killing. However, our results would emphasize that non-MHC-I-dependent mode of neuronal killing can also occur (53).

The mode of death of neurons following their encounter with T cells in this study is not clear and it is possible that both apoptotic and necrotic cell death occur to some degree. Some neurons were found to express annexin V (Fig. 2D), a marker of apoptosis, before their detachment from the culture substrate.

It is instructive to note that when autoreactive T cell lines are adoptively transferred into animals, a fate of these cells that migrate into the CNS is death; indeed, the expression of FasL by neurons is thought to account for the demise of these T cells (54, 55). Nonetheless, while the CNS has high potential for the elimination of T cell-dependent inflammation (56), this state is overwhelmed in conditions such as MS where T cells infiltrate into the CNS in significant numbers. In our experiments, we observed that many of the T cells that aggregate around neuronal elements appear pyknotic and dead (Fig. 7), in correspondence with the concept that neurons can evoke T cell death as a mechanism of neuroprotection. However, before the elimination of T cells, significant axonal and neuronal injury had already occurred. It is also possible that T cells are first killed by the neuronal contact, and that it is the dying T cells that then release cytotoxic factors to neurons. However, we have tested this by adding dying T cells to neurons but no toxicity was found (F. Giuliani, C. G. Goodyer, J. P. Antel, and V. W. Yong, unpublished observations).

The effector molecules that account for the T cell cytotoxicity may be through two different pathways, one mediated by granzymes and perforins and the other mediated by the interaction of Fas with FasL. We did not evaluate whether one or both are operational in our model. Discordant data are present in the literature where the neuronal killing of neurons by T cells could be perforin-mediated (57) or Fas-mediated (58). In a model of MS, Theiler’s murine encephalomyelitis, it was shown that perforin-deficient mice showed no clinical disease (59). Others have reported that CNS neurons are refractory to cytotoxic granule-mediated lysis (60) and that, after the induction of MHC-I expression, neuronal killing is via an apoptotic mechanism mediated by a Fas-FasL interaction (58). In our model the treatment of T cells with blocking Abs anti-FasL, anti-LFA-1, and anti-CD40 can prevent the neuronal damage. This result is in accordance with previous reports (61, 62, 63) that all the pathways mediated by those molecules are involved in T cell-mediated cytotoxicity. The block of ICAM-1 on T cells did not have any effect on killing, consistent with the lack of expression of ICAM-1 on T cells. On the other hand, the targeting of ICAM-1 present on neurons prevented the killing by activated T cells. The ICAM-1 expressed on neuronal membranes could interact with the LFA-1 over-expressed on activated T cells. Consistent with these data is the report that neurons do express ICAM-1, Fas, FasL, and CD40 (54, 55, 64, 65, 66). Human fetal neurons specifically have also been found to express Fas (67), although we were not able to demonstrate this convincingly by immunohistochemistry in our cultures (data not shown). However, we were able to document CD40 and ICAM-1 on human neurons in culture (Fig. 9).

The data that showed an anti-CD40 blocking Ab applied to neurons diminished T cell cytotoxicity is consistent with CD40 on neurons mediating the interaction with T cells to produce toxicity. When we applied the CD40 blocking Ab to T cells, this also alleviated subsequent toxicity of T cells on neurons. Although this implies that the CD40 on T cells also has a role in causing neuronal death, an alternate explanation is that the CD40 Ab treatment of T cells decreased other aspects of T cell activation; in this regard, blocking CD40 on T cells has been reported to produce their anergy (68).

Finally, our results demonstrate that the T cell cytotoxicity is selective in that oligodendrocytes and astrocytes do not appear to be susceptible; however, we must emphasize that the adult nature of the oligodendrocytes might have contributed to this differential response compared with fetal neurons. Previously, we have found that human oligodendrocytes were not susceptible to killing by activated CD4+ or CD8+ T cells (69). Thus, oligodendrocyte vulnerability is likely through other mechanisms, such as Ab-mediated cytotoxicity or soluble factors such as TNF-{alpha} (70). We noted that although astrocytes were not killed by activated T cells, a modification of their shape from flat to stellar was evident probably indicating an activation of these cells. The resistance of astrocytes to T cell killing could be facilitated by the induction of apoptosis of T cells in accordance with previous reports (71). We do not know why neurons are more vulnerable than oligodendrocytes or astrocytes to the T cell-mediated killing described in this report. This circumstance may involve a differential spectrum of adhesion molecules on neurons that facilitate T cell attachment, but the adhesion molecules on neurons that we studied are also expressed on oligodendrocytes or astrocytes (64, 65, 66). It remains possible that a neuron-specific adhesion molecule not examined in this study is indeed the predominant ligand for an interacting partner on activated T cells to evoke neuronal death.

T cells are known to patrol the CNS under normal conditions; why these do not kill neurons in the normal state could be related to their relatively low numbers in the normal state compared with diseased conditions. It is also possible that the level of activation of the T cells that patrol the CNS in the normal state is lower than that of T cells that infiltrate into the CNS in neuro-inflammatory diseases. This study highlights the need to examine more carefully the possibility that activated T cells under some conditions will be toxic to neurons in the CNS. This emerging area of study is supported by the recent demonstration in Rasmussen’s encephalitis that CD8+ T cells are in close apposition to dying neurons (42).

We would like to emphasize that fetal human neurons were used in this study, and their level of maturity might have contributed to their relative susceptibility to T cell-mediated killing. To test this, one would need to use differentiated neurons from adult brains, but these do not survive the isolation and culture process. Another means is to "age" the neurons in culture. In this regard, while most neuronal cultures were used 2 wk after isolation, neurons that were aged in vitro for 1–2 mo were equally susceptible to toxicity by T cells (Fig. 3). It has been shown that enriched fetal human cortical neurons, in vitro, undergo morphological and biochemical differentiation with the formation of synapses within 1 wk after isolation. Furthermore, they develop excitable properties between day 7 and day 21 after isolation (72). Finally, we believe that our results are relevant to neuronal death in human diseases given the recent realization that neurons and axons are lost in large numbers in MS, Rasmussen’s encephalitis, and paraneoplastic encephalomyelitis (2, 3, 7, 40, 41, 42).

In summary, this study reveals for the first time that neurons are selectively vulnerable to T cell cytotoxicity and that this could be responsible for the neurodegeneration observed in inflammatory diseases. Our results would suggest that while the initiation of MS may be attributed to T cells that are Ag-specific, the subsequent arrival of activated T cells of many specificities then damage the CNS in a cell contact-dependent non-Ag-restricted manner.


   Acknowledgments
 
We gratefully acknowledge the superb technical assistance of Tammy Wilson and Ellie McCrea.


   Footnotes
 
1 This work was supported by the Canadian Institutes for Health Research and its Interdisciplinary Health Research Team program. F.G. is supported by a clinical fellowship award from the Alberta Heritage Foundation for Medical Research. V.W.Y. is a Senior Scholar of Alberta Heritage Foundation for Medical Research and a Canadian Institutes of Health Research scientist. Back

2 Address correspondence and reprint requests to Dr. V. Wee Yong, Departments of Oncology and Clinical Neurosciences, University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta T2N 4N1, Canada. E-mail address: vyong{at}ucalgary.ca Back

3 Abbreviations used in this paper: MS, multiple sclerosis; SPMS, secondary progressive MS; RRMS, relapsing remitting MS; MAP, microtubule-associated protein; MHC-I, MHC class I; MHC-II, MHC class II; GFAP, glial fibrillary acidic protein. Back

Received for publication December 6, 2002. Accepted for publication April 22, 2003.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Davie, C. A., G. J. Barker, S. Webb, P. S. Tofts, A. J. Thompson, A. E. Harding, W. I. McDonald, D. H. Miller. 1995. Persistent functional deficit in multiple sclerosis and autosomal dominant cerebellar ataxia is associated with axon loss. Brain 118:1583.[Abstract/Free Full Text]
  2. Trapp, B. D., J. Peterson, R. M. Ransohoff, R. Rudick, S. Mork, L. Bo. 1998. Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338:278.[Abstract/Free Full Text]
  3. Tourbah, A., J. L. Stievenart, O. Gout, B. Fontaine, R. Liblau, C. Lubetzki, E. A. Cabanis, O. Lyon-Caen. 1999. Localized proton magnetic resonance spectroscopy in relapsing remitting versus secondary progressive multiple sclerosis. Neurology 53:1091.[Abstract/Free Full Text]
  4. Ferguson, B., M. K. Matyszak, M. M. Esiri, V. H. Perry. 1997. Axonal damage in acute multiple sclerosis lesions. Brain 120:393.[Abstract/Free Full Text]
  5. Kuhlmann, T., G. Lingfeld, A. Bitsch, J. Schuchardt, W. Bruck. 2002. Acute axonal damage in multiple sclerosis is most extensive in early disease stages and decreases over time. Brain 125:2202.[Abstract/Free Full Text]
  6. Peterson, J. W., L. Bo, S. Mork, A. Chang, B. D. Trapp. 2001. Transected neurites, apoptotic neurons, and reduced inflammation in cortical multiple sclerosis lesions. Ann. Neurol. 50:389.[Medline]
  7. Cifelli, A., M. Arridge, P. Jezzard, M. M. Esiri, J. Palace, P. M. Matthews. 2002. Thalamic neurodegeneration in multiple sclerosis. Ann. Neurol. 52:650.[Medline]
  8. Meyer, R., R. Weissert, R. Diem, M. K. Storch, K. L. de Graaf, B. Kramer, M. Bahr. 2001. Acute neuronal apoptosis in a rat model of multiple sclerosis. J. Neurosci. 21:6214.[Abstract/Free Full Text]
  9. Lucchinetti, C., W. Bruck, M. Rodriguez, G. Lassmann. 1996. Distinct patterns of multiple sclerosis pathology indicates heterogeneity on pathogenesis. Brain Pathol. 6:259.[Medline]
  10. Martin, R., C. S. Sturzebecher, H. F. McFarland. 2001. Immunotherapy of multiple sclerosis: where are we? where should we go?. Nat. Immun. 2:785.
  11. Wekerle, H., K. Kojima, J. Lannes-Vieira, H. Lassmann, C. Linington. 1994. Animal models. Ann. Neurol. 36:S47.
  12. Hafler, D. A., H. L. Weiner. 1995. Antigen-specific immunosuppression: oral tolerance for treatment of autoimmune disease. Chem. Immunol. 60:126.[Medline]
  13. Steinman, L.. 1996. Multiple sclerosis: a coordinated immunological attack against myelin in the central nervous system. Cell 85:299.[Medline]
  14. Genain, C. P., B. Cannella, S. L. Hauser, C. S. Raine. 1999. Identification of autoantibodies associated with myelin damage in multiple sclerosis. Nat. Med. 5:170.[Medline]
  15. Hickey, W. F., B. L. Hsu, H. Kimura. 1991. T-lymphocyte entry into the central nervous system. J. Neurosci. Res. 28:254.[Medline]
  16. Flugel, A., M. Willem, T. Berkowicz, H. Wekerle. 1999. Gene transfer into CD4+ T lymphocytes: green fluorescent protein-engineered, encephalitogenic T cells illuminate brain autoimmune responses. Nat. Med. 5:843.[Medline]
  17. Qing, Z., D. Sewell, M. Sandor, Z. Fabry. 2000. Antigen-specific T-cell trafficking into the central nervous system. J. Neuroimmunol. 105:169.[Medline]
  18. Manning, P. T., E. M. J. Johnson. 1987. MHC-specific cytotoxic T lymphocyte killing of dissociated sympathetic neuronal cultures. Am. J. Pathol. 128:395.[Abstract]
  19. Rall, G. F., L. Mucke, M. B. Oldstone. 1995. Consequences of cytotoxic T lymphocyte interaction with major histocompatibility complex class-1 expressing neurons in vivo. J. Exp. Med. 182:1201.[Abstract/Free Full Text]
  20. Medana, I., M. A. Martinic, H. Wekerle, H. Neumann. 2001. Transection of major histocompatibility complex class I-induced neurites by cytotoxic T lymphocytes. Am. J. Pathol. 159:809.[Abstract/Free Full Text]
  21. Neumann, H., A. Cavalie, D. E. Jenne, H. Wekerle. 1995. Induction of MHC class I genes in neurons. Science 269:2582.
  22. Redwine, J. M., M. J. Buchmeier, C. F. Evans. 2001. In vivo expression of major histocompatibility complex molecules on oligodendrocytes and neurons during viral infection. Am. J. Pathol. 159:1219.[Abstract/Free Full Text]
  23. Wucherpfennig, K. W., J. Newcombe, H. Li, C. Keddy, M. L. Cuzner, D. A. Hafler. 1992. T cell receptor V{alpha}-V{beta} repertoire and cytokine gene expression in active multiple sclerosis lesions. J. Exp. Med. 175:993.[Abstract/Free Full Text]
  24. Stuve, O., N. P. Dooley, J. H. Uhm, J. P. Antel, G. S. Francis, G. Williams, V. W. Yong. 1996. Interferon {beta}-1b decreases the migration of T lymphocytes in vitro: effects on matrix metalloproteinase-9. Ann. Neurol. 40:853.[Medline]
  25. Chabot, S., G. Williams, V. W. Yong. 1997. Microglial production of TNF-{alpha} is induced by activated T lymphocytes: involvement of VLA-4 and inhibition by interferon {beta}-1b. J. Clin. Invest. 100:604.[Medline]
  26. Chabot, S., G. Williams, M. Hamilton, G. Sutherland, V. W. Yong. 1999. Mechanisms of IL-10 production in human microglia-T cell interaction. J. Immunol. 162:6819.[Abstract/Free Full Text]
  27. Vecil, G. G., P. H. Larsen, S. M. Corley, L. M. Herx, A. Besson, C. G. Goodyer, V. W. Yong. 2000. Interleukin-1 is a key regulator of matrix metalloproteinase-9 expression in human neurons in culture and following mouse brain trauma in vivo. J. Neurosci. Res. 61:212.[Medline]
  28. Yong, V. W., J. Antel. 2001. Culture of glial cells from human brain biopsies. A. Richardson, and S. Fedoroff, eds. Protocols for Neural Cell Culture 129. Humana Press, St. Louis.
  29. Sommer, I., M. Schachner. 1981. Monoclonal antibodies (O1 to O4) to oligodendrocyte cell surfaces: an immunocytochemical study in the central nervous system. Dev. Biol. 83:311.[Medline]
  30. Baltuch, G. H., J. G. Villemure, E. McCrea, J. P. Antel. 1994. T cell-mediated cytotoxicity of human gliomas: a tumor necrosis factor-independent mechanism. Neurosurgery 35:450.[Medline]
  31. Matesic, D. F., R. C. Lin. 1994. Microtubule-associated protein 2 as an early indicator of ischemia-induced neurodegeneration in the gerbil forebrain. J. Neurochem. 63:1012.[Medline]
  32. Rosenstein, J. M.. 1995. Diminished expression of microtubule-associated protein (MAP-2) and {beta}-tubulin as aputative marker for ischemic injury in neocortical transplants. Cell Transplant. 4:83.[Medline]
  33. Postmantur, R. M., A. Kampfl, S. J. Liu, K. Heck, W. C. Taft, G. L. Clifton, R. L. Hayes. 1996. Cytoskeletal derangements of cortical neuronal processes three hours after traumatic brain injury in rats: an immunofluorescence study. J. Neuropathol. Exp. Neurol. 55:68.[Medline]
  34. Saatman, K. E., D. I. Graham, T. K. McIntosh. 1998. The neuronal cytoskeleton is at risk after mild and moderate brain injury. J. Neurotrauma 15:1047.[Medline]
  35. Harada, J., M. Sugimoto. 1999. Activation of caspase-3 in {beta}-amyloid-induced apoptosis of cultured rat cortical neurons. Brain Res. 842:311.[Medline]
  36. Yagita, Y., M. Matsumoto, K. Kitagawa, T. Mabuchi, T. Ohtsuki, M. Hori, T. Yanagihara. 1999. DNA cleavage and proteolysis of microtubule-associated protein 2 after cerebral ischemia of different severity. Neuroscience 92:1417.[Medline]
  37. Zhang, S. X., M. Underwood, A. Landfield, F. F. Huang, S. Gison, J. W. Geddes. 2000. Cytoskeletal disruption following contusion injury to the rat spinal cord. J. Neuropathol. Exp. Neurol. 59:287.[Medline]
  38. Yong, V. W.. 1996. Cytokines, astrogliosis and neurotrophism following CNS trauma. E. Benveniste, and R. Ransohoff, eds. Cytokines and the CNS: Development, Defense and Disease 309. CRC Press, Boca Raton.
  39. Downen, M., T. D. Amaral, L. L. Hua, M. L. Zhao, S. C. Lee. 1999. Neuronal death in cytokine-activated primary human brain cell culture: role of tumor necrosis factor-{alpha}. Glia 28:114.[Medline]
  40. Steinman, L.. 2001. Multiple sclerosis: a two-stage disease. Nat. Immun. 2:762.
  41. Bernal, F., F. Graus, A. Pifarre, A. Saiz, B. Benyahia, T. Ribalta. 2002. Immunohistochemical analysis of anti-Hu-associated paraneoplastic encephalomyelitis. Acta Neuropathol. 103:509.[Medline]
  42. Bien, C. G., J. Bauer, T. L. Deckwerth, H. Wiendl, M. Deckert, O. D. Wiestler, J. Schramm, C. E. Elger, H. Lassmann. 2002. Destruction of neurons by cytotoxic T cells: a new pathogenic mechanism in Rsmussen’s encephalitis. Ann. Neurol. 51:311.[Medline]
  43. Trapp, B. D., R. Ransohoff, R. Rudick. 1999. Axonal pathology in multiple sclerosis: relationship to neurologic disability. Curr. Opin. Neurol. 12:295.[Medline]
  44. Weinshenker, B. G., B. Bass, G. P. Rise, J. Noseworthy, W. Carriere, J. Baskerville, G. C. Ebers. 1989. The natural history of multiple sclerosis: a geographically based study. I. Clinical course and disability. Brain 112:133.[Abstract/Free Full Text]
  45. Casanova, B., F. Coret, C. Valero, L. Landete, A. Pascual, J. J. Vilchez. 2002. High clinical inflammatory activity prior the development of secondary progression: a prospective 5-year follow-up study. Mult. Scler. 8:59.[Abstract/Free Full Text]
  46. Bitsch, A., J. Schuchardt, S. Bunkowski, T. Kuhlmann, W. Bruck. 2000. Acute axonal injury in multiple sclerosis: correlation with demyelination and inflammation. Brain 123:1174.[Abstract/Free Full Text]
  47. Kornek, B., M. K. Storch, R. Weissert, E. Wallstroem, A. Stefferl, T. Olsson, C. Linington, M. Schmidbauer, H. Lassmann. 2000. Multiple sclerosis and chronic autoimmune encephalomyelitis: a comparative quantitative study of axonal injury in active, inactive, and remyelinated lesions. Am. J. Pathol. 157:267.[Abstract/Free Full Text]
  48. Smith, T., A. Groom, B. Zhu, L. Turski. 2000. Autoimmune encephalomyelitis ameliorated by AMPA antagonists. Nat. Med. 6:62.[Medline]
  49. Blomgren, K., A. McRae, E. Bona, T. C. Saido, J. O. Karlsson, H. Hagberg. 1995. Degradation of fodrin and MAP 2 after neonatal cerebral hypoxic-ischemia. Brain Res. 684:136.[Medline]
  50. Maher, S., D. Toomey, C. Condron, D. Bouchier-Hayes. 2002. Activation-induced cell death: the controversial role of Fas and Fas ligand in immune privilege and tumour counterattack. Immunol. Cell Biol. 80:131.[Medline]
  51. Szocinski, J. L., A. R. Khaled, J. Hixon, D. Halverson, S. Funakoshi, W. C. Fanslow, A. Boyd, D. D. Taub, S. K. Durum, C. B. Siegall, D. L. Longo, W. J. Murphy. 2002. Activation-induced cell death of aggressive histology lymphomas by CD40 stimulation: induction of bax. Blood 100:217.[Abstract/Free Full Text]
  52. Rivera-Quinones, C., D. McGavern, J. D. Schmelzer, S. F. Hunter, P. A. Low, M. Rodriguez. 1998. Absence of neurological deficits following extensive demyelination in class I-deficient murine model of multiple sclerosis. Nat. Med. 4:187.[Medline]
  53. Newman, T. A., S. T. Woolley, P. M. Hughes, N. R. Sibson, D. C. Anthony, V. H. Perry. 2001. T-cell- and macrophage-mediated axon damage in the absence of a CNS-specific immune response: involvement of metalloproteinases. Brain 124:2203.[Abstract/Free Full Text]
  54. Bauer, J., M. Bradl, W. F. Hickey, S. Forss-Petter, H. Breitschopf, C. Linington, H. Wekerle, H. Lassmann. 1998. T-cell apoptosis in inflammatory brain lesions: destruction of T cells does not depend on antigen recognition. Am. J. Pathol. 153:715.[Abstract/Free Full Text]
  55. Flugel, A., F. W. Schwaiger, H. Neumann, I. Medana, M. Willem, H. Wekerle, G. W. Kreutzberg, M. B. Graeber. 2000. Neuronal FasL induces cell death of encephalitogenic T lymphocytes. Brain Pathol. 10:353.[Medline]
  56. Gold, R., H. P. Hartung, H. Lassmann. 1997. T-cell apoptosis in autoimmune diseases: termination of inflammation in the nervous system and other sites with specialized immune-defense mechanisms. Trends Neurosci. 20:399.[Medline]
  57. Rensing-Ehl, A., U. Malipiero, M. Irmler, J. Tschopp, D. Constam, A. Fontana. 1996. Neurons induced to express major histocompatibility complex class I antigen are killed via the perforin and not the Fas (APO-1/CD95) pathway. Eur. J. Immunol. 26:2271.[Medline]
  58. Medana, I. M., A. Gallimore, A. Oxenius, M. M. Martinic, H. Wekerle, H. Neumann. 2000. MHC class I-restricted killing of neurons by virus-specific CD8+ T lymphocytes is effected through the Fas/FasL, but not the perforin pathway. Eur. J. Immunol. 30:3623.[Medline]
  59. Murray, P. D., D. B. McGavern, X. Lin, M. K. Njenga, J. Leibowitz, L. R. Pease, M. Rodriguez. 1998. Perforin-dependent neurologic injury in a viral model of multiple sclerosis. J. Neurosci. 18:7306.[Abstract/Free Full Text]
  60. Keane, R. W., M. W. Tallent, E. R. Podack. 1992. Resistance and susceptibility of neural cells to lysis by cytotoxic lymphocytes and by cytolytic granules. Transplantation 54:520.[Medline]
  61. Van Gool, S. W., J. Vermeiren, K. Rafiq, K. Lorr, M. de Boer, J. L. Ceuppens. 1999. Blocking CD40-CD154 and CD80/CD86-CD28 interactions during primary allogeneic stimulation results in T cell anergy and high IL-10 production. Eur. J. Immunol. 29:2367.[Medline]
  62. Andreasen, S. O., J. E. Christensen, O. Marker, A. R. Thomsen. 2000. Role of CD40 ligand and CD28 in induction and maintenance of antiviral CD8+ effector T cell responses. J. Immunol. 164:3689.[Abstract/Free Full Text]
  63. Kojima, H., M. V. Sitkovsky. 2000. Lysis of innocent bystanders by antigen-specific cytotoxic lymphocytes. M. V. Sitkovsky, and P. A. Henkart, eds. Cytotoxic Cells: Basic Mechanisms and Medical Applications 79. Lippincott Williams & Wilkins, Philadelphia.
  64. Hery, C., G. Sebire, S. Peudenier, M. Tardieu. 1995. Adhesion to human neurons and astrocytes of monocytes: the role of interaction of CR3 and ICAM-1 and modulation by cytokines. J. Neuroimmunol. 57:101.[Medline]
  65. Ferrer, I., B. Puig, J. Krupinsk, M. Carmona, R. Blanco. 2001. Fas and Fas ligand expression in Alzheimer’s disease. Acta Neuropathol. 102:121.[Medline]
  66. Tan, J., T. Town, T. Mori, D. Obregon, Y. Wu, A. DelleDonne, A. Rojiani, F. Crawford, R. A. Flavell, M. Mullan. 2002. CD40 is expressed and functional on neuronal cells. EMBO J. 21:643.[Medline]
  67. Nat, R., E. Radu, T. Regalia, L. M. Popsen. 2001. Apoptosis in human embryo development: 3: Fas-induced apoptosis in brain primary cultures. J. Cell. Mol. Med. 5:417.[Medline]
  68. Koen, H. G., I. Joosten. 2000. Blockade of CD86 and CD40 induces alloantigen-specific immunoregulatory T cells that remain anergic even after reversal of hyporesponsiveness. Blood 95:3153.[Abstract/Free Full Text]
  69. Jurewicz, A., W. E. Biddison, J. P. Antel. 1998. MHC class I-restricted lysis of human oligodendrocytes by myelin basic protein peptide-specific CD8 T lymphocytes. J. Immunol. 160:3056.[Abstract/Free Full Text]
  70. Benn, T., C. Halfpenny, N. Scolding. 2001. Glial cells as targets for cytotoxic immune mediators. Glia 36:200.[Medline]
  71. Pender, M. P., M. J. Rist. 2001. Apoptosis of inflammatory cells in immune control of the nervous system: role of glia. Glia 36:137.[Medline]
  72. Kerkovich, D. M., D. Sapp, K. Weidenheim, C. Brosnan, S. E. Pfeiffer, H. H. Yee, J. Busciglio. 1999. Fetal human cortical neurons grown in culture: morphological differentiation, biochemical correlates and development of electrical activity. Int. J. Dev. Neurosci. 17:347.[Medline]



This article has been cited by other articles:


Home page
 Am. J. Pathol.Home page
A. Brunn, O. Utermohlen, M. Carstov, M. S. Ruiz, H. Miletic, D. Schluter, and M. Deckert
CD4 T Cells Mediate Axonal Damage and Spinal Cord Motor Neuron Apoptosis in Murine P0106-125-Induced Experimental Autoimmune Neuritis
Am. J. Pathol., July 1, 2008; 173(1): 93 - 105.
[Abstract] [Full Text] [PDF]

Home page
 BloodHome page
L. Tian, J. Lappalainen, M. Autero, S. Hanninen, H. Rauvala, and C. G. Gahmberg
Shedded neuronal ICAM-5 suppresses T-cell activation
Blood, April 1, 2008; 111(7): 3615 - 3625.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
I. P. Greene, E.-Y. Lee, N. Prow, B. Ngwang, and D. E. Griffin
Protection from fatal viral encephalomyelitis: AMPA receptor antagonists have a direct effect on the inflammatory response to infection
PNAS, March 4, 2008; 105(9): 3575 - 3580.
[Abstract] [Full Text] [PDF]

Home page