An inappropriate cross talk between activated T lymphocytes infiltrating the CNS and neural cells can sustain the onset and progression of demyelination and axonal degeneration in neuroinflammatory diseases. To mimic this deleterious cross talk, we designed an experimental paradigm consisting of transient cocultures of T lymphocytes chronically activated by retrovirus infection (not virus productive) with human multipotent neural precursors or primary oligodendrocytes from rat brain. We showed that activated T lymphocytes induced apoptotic death of multipotent neural progenitors and immature oligodendrocytes after a progressive collapse of their process extensions. These effects were reminiscent of those induced by brain semaphorin on neural cells. Blockade by specific Abs of soluble CD100 (sCD100)/semaphorin 4D released by activated T cells, or treatment with rsCD100, demonstrated that this immune semaphorin has the ability to collapse oligodendrocyte process extensions and to trigger neural cell apoptosis, most likely through receptors of the plexin family. The specific presence of sCD100 in the cerebrospinal fluid and of CD100-expressing T lymphocytes in the spinal cord of patients suffering with neuroinflammatory demyelination pointed to the potential pathological effect of sCD100 in the CNS. Thus, our results show that CD100 is a new important element in the deleterious T cell-neural cell cross talk during neuroinflammation and suggest its role in demyelination or absence of remyelination in neuroinflammatory diseases including multiple sclerosis and human T lymphotropic virus type 1-associated myelopathy.

Interaction between the immune and the nervous systems may trigger an inappropriate cross talk, as suggested by demyelination and axonal loss in neuroinflammatory diseases. In attempting to understand this cross talk, two key issues need to be resolved: the entry and survival of activated T lymphocytes in the brain, and the identification of immune molecules leading to demyelination. A large and diverse molecular repertoire shared by immune and CNS, such as cytokines, chemokines, the metalloproteinase/inhibitor, and Fas/Fas ligand systems, revealing remarkable similarities between signaling molecules and cellular receptors in both systems, could participate in this cross talk (1). It also provides essential information reflecting important steps toward a more complete understanding of the functional properties of diverse molecules either in normal or pathological conditions. It is becoming clear that semaphorins, which are well known for their role in axonal steering, synapse formation, and chemorepulsion in the developing nervous system (2, 3, 4), also have roles in the immune system. In fact, CD100 (also known as semaphorin 4D (SEMA-4D))4 and SEMA-4A, two immune semaphorins, are expressed constitutively in T cells (5, 6), function as soluble ligand to regulate the humoral and cellular immune responses (7, 8, 9), and are suspected to be involved in autoimmune diseases such as experimental autoimmune encephalitis (EAE) (6, 10). SEMA-3A acts in axonal guidance, apoptosis of neural precursors or neurons, and oligodendrocyte alteration (11, 12, 13, 14), and inhibits the chemokine-induced migration of human immune cells similarly to CD100 (15). Neuropilin-1, a component of SEMA-3A receptor in brain, is also expressed in immune cells and plays a role in the establishment of contacts between naive T lymphocytes and dendritic cells (16). In this context, we propose that soluble CD100 (sCD100), which is produced upon T lymphocyte activation through proteolysis by a metalloproteinase (17), could be responsible for an inappropriate response in neural-immune interactions during inflammation.

Neuroinflammatory diseases, including multiple sclerosis (MS) and myelopathy associated with human T lymphotropic virus type 1 (HTLV-1) infection (TSP/HAM), are characterized by inflammation associated with damage of the white matter and axonal degeneration in the brain and spinal cord (18, 19, 20). Both infiltrated T lymphocytes and inflammatory mediators are suspected to participate in pathogenic mechanisms (21, 22, 23, 24, 25). The persistent demyelination in these diseases and the reported sensibility of neural progenitors to inflammatory mediators suggest the dysfunction or death of the precursors or myelinating oligodendrocytes required for remyelination (26, 27, 28). Axonal loss, detected early in the disease by magnetic resonance spectroscopy (29, 30), could result from defects in remyelination (31) and abnormal levels of molecular signals that regulate axon extension such as semaphorins (32). Assuming that immune semaphorins produced by infiltrating T lymphocytes may have a deleterious effect on neural cells, we investigated the in vitro effect of activated T cells releasing sCD100-SEMA-4D and rsCD100 protein on human pluripotent neural precursors and on rat oligodendrocytes. These cellular contacts mimic interactions occurring during neuroinflammation between infiltrating/activated T lymphocytes and oligodendrocytes or pluripotent neural precursors still capable of generating glial cells in adult brain (33). In respect to the effects of brain semaphorins on neural precursors and oligodendrocytes (12, 13), we particularly focused on process extensions and cell survival. Next, the in vivo functional relevance was examined by looking for: 1) the presence of sCD100 in the cerebrospinal fluid (CSF), and 2) the presence of cell surface CD100 on activated T lymphocytes in postmortem spinal cords from patients suffering with neuroinflammatory demyelination (TSP/HAM).

The following CD4+ T cells were used: 1) the human CD25+CD100+ T cell line C8166, which is activated by chronic retroviral infection (HTLV-I, no virus productive) (34) and released high level of sCD100 (3 ng/μl from 20 × 106 cells in 1 ml compared with 1 ng/μl from CD100-transfected Jurkat T cells in same conditions), as detected by ELISA (5); 2) the nonactivated CD4+ T cell line CEM used as control; and 3) the primary culture of T lymphocytes isolated from a patient infected with HTLV-1 (CIB, CD25+, CD69+, CD100+). These T lymphocytes were transiently cocultured, as previously described (35), with the following human and rat neural cells (ratio 1/10, respectively) growing on culture slide or in flask: 1) the human pluripotent neural precursor cell line Dev, which has the ability to differentiate into neurons, astrocytes, and oligodendrocytes (36, 37); 2) primary culture of rat glial cells (38), which contained 35–54% oligodendrocytes, the remaining cells corresponding to glial acidic fibrillary protein (GFAP)-positive astrocytes. The oligodendrocytes have distinct phenotypic stages identified with a panel of specific Abs by immunochemistry or flow cytometry (see below): early oligodendrocyte precursor (pre-oligodendrocyte) expressing the NG2 chondroitin sulfate (10–15%), immature oligodendrocyte expressing galactocerebroside (GalCer, 17–25%) and cyclic nucleotide phosphodiesterase (CNPase), mature oligodendrocyte expressing myelin-associated glycoprotein (3–15%) and myelin basic protein, and 3) human fetal neural cells cultivated from the subventricular zone identified by immunodetection of β3-tubulin. Following transient coculture (20 h) of C8166 or CEM T lymphocytes with human or rat neural cells, the nonadherent T cells were eliminated by washing of neural cultures, and their elimination was verified by the lack of CD4 Ag in flow cytometry. For some experiments, these cocultures were treated with anti-CD100-purified Abs, namely the blocking sCD100 BD16 mAb or the nonblocking BB18 mAb.

Alternatively, in the same set of experiments, the neural cells were treated with rsCD100 from the supernatant of permanently transfected Jurkat cells, used crude or after purification on anti-CD100 BB18-mAb column (1 ng/μl concentration). Supernatant from Jurkat cells transfected to produce CD27 was used in similar condition as control. Each experiment was repeated at least three times.

The CNS tissue samples from three TSP/HAM patients and two noninfected patients (Parkinson disease, car crash) were examined for CD100-expressing T cells. Paraffin-embedded spinal cords were examined using H&E-safran and Luxol Fast Blue for myelin detection. Clinical and pathological characteristics of these TSP/HAM patients are demonstrated in Table I. The CSF from patients suffering with TSP/HAM (9 patients), meningitis (4 patients), or encephalomyelitis (3 patients) were examined for the presence of sCD100 by a sandwich ELISA, as previously described (5). All TSP/HAM patients tested were positive for HTLV-1 provirus.

Table I.

Clinical and pathological characteristics of TSP/HAM patients

CaseAge/SexDuration of IllnessCause of DeathCNS Neuropathology
InflammationMyelin/axonal lossAtrophy
62 /F 15 years Pulmonary embolism +++ +++ +++ 
35 /M 4 years Mesenteric thrombosis ++ +++ ++ 
65 /F 8 years Pneumonia 
CaseAge/SexDuration of IllnessCause of DeathCNS Neuropathology
InflammationMyelin/axonal lossAtrophy
62 /F 15 years Pulmonary embolism +++ +++ +++ 
35 /M 4 years Mesenteric thrombosis ++ +++ ++ 
65 /F 8 years Pneumonia 

Oligodendrocytes were identified in the rat primary glial culture by immunofluorescence on culture slide or by flow cytometry (5, 35), with anti-CNPase (Sigma-Aldrich, St. Louis, MO), anti-GalCer, anti-NG2 (Chemicon International, Temecula, CA), anti-myelin-associated glycoprotein (Boehringer Mannheim, Indianapolis, IN), and anti-myelin basic protein (Serotec, Oxford, U.K.) mAbs. Astrocytes were identified in the same conditions with a rabbit polyclonal Ab (anti-GFAP; DAKO, Carpenteria, CA). Human fetal neural precursors were detected with anti-β3-tubulin (T8660; Sigma-Aldrich). Semaphorin receptors were detected with polyclonal anti-human plexin-B1 (N18; SantaCruz-Tebu, Le Perray, France) and anti-MAM neuropilin-1-blocking polyclonal Ab (12). Immune cells were detected in spinal cord sections from TSP/HAM patients by immunofluorescence with anti-CD4 (M0716; DAKO), anti-CD8 (M7103; DAKO), anti-CD45RO (UCHL1; DAKO), and anti-fascin (M3567; DAKO) mAbs. Membrane-bound CD100 was detected with an anti-CD100 polyclonal antiserum directed against the intracellular portion of the protein (aa 799–813). After dewaxing with toluene and ethanol, sections were incubated in blocking solution (1% BSA, 0.3% Triton X-100, 1 h), then with specific Abs (4°C/overnight). Alexa546-labeled anti-mouse or Alexa488-labeled anti-rabbit Ig Abs (Molecular Probes, Eugene, OR) were then applied.

mRNA were detected by extraction, followed by RT-PCR and Southern blotting using appropriate [33P]dATP 5′ end-labeled internal oligonucleotide probes, as previously described (38). Oligonucleotide primers were chosen from their mRNA sequences (GenBank access numbers HSU60800 for CD100, NM002663 for plexin-B1, NM012401 for plexin-B2, AF149019 for plexin-B3, L26081 for SEMA-3A, NM01101 for β-actin, used as control). One set of nucleotide primers (forward, TGGTGAAGCCAAGTGATGAG; reverse, CTGCAGAAGTTCGTGGATGA) was selected for amplification of a common 300-bp amplicon in the three plexin-B mRNA.

In each experiment, apoptotic death was detected by three methods: 1) the TUNEL method performed on culture slide (Promega, Madison, WI); 2) immunodetection of the executer caspase, active caspase-3 (rabbit serum; BD PharMingen, San Diego, CA); and 3) detection of apoptotic bodies with DNA intercal 4′,6′-diamidino-2-phenylindole (1 μg/ml; Sigma-Aldrich) staining of nucleus. Codetection of TUNEL, active caspase-3, and oligodendrocyte or astrocyte markers was also performed to identify the damaged cell population within the glial culture.

The number of total cells (nucleus staining with 4′,6′-diamidino-2-phenylindole), TUNEL, and active caspase-3-positive cells was counted in neural precusors and primary glial cells grown in culture slide and transiently cocultured with C8166 or CEM T cells (15–20 microscope fields/800-1000 cells counted per experiment, 3 independent experiments) using AnalySiS 3.2 software. The values were expressed as mean ± SEM of positive cells per field. Cell loss was calculated by counting the total cell number per field in each experimental situation or by flow cytometry after immunodetection of oligodendrocyte populations. Differences between groups were calculated with the Student’s t test. Measure of number and length of oligodendrocyte process was performed in the cocultured glial culture by using the same software (3 independent experiments).

The effect of immune semaphorin SEMA-4D/CD100 released from chronically activated T lymphocytes was investigated on human pluripotent neural precursors (Dev cells) and on rat primary oligodendrocytes within glial primoculture by analyzing neural cell morphology and survival after contact. The T cell line C8166, which is activated by chronic retroviral infection and produces high level of sCD100, was transiently cocultured with these neural cells. The non-sCD100-producing T cells, CEM, were used as control. Spontaneous apoptosis was detected in 0.41–5.96% of untreated Dev cells. Following contact with CD100-producing T cells (48 h), their level of spontaneous apoptotic death was greatly enhanced. The number of TUNEL-positive cells increased from 17.2 ± 2.4 per field in C8166-treated neural precursors vs 7 ± 2.2 in CEM-treated and 4.5 ± 2.7 in untreated cells (p < 0.001). Apoptotic bodies in TUNEL-positive cells are shown in Fig. 1,A. In addition, the executer caspase, active caspase-3, was detected by immunofluorescence. As shown in Fig. 1,B (three independent experiments gave similar results), active caspase-3 was detected in 25.5 ± 8.8 cells per field following contact with C8166 T cells vs 11.3 ± 3.7 in CEM-treated and 10.8 ± 1.7 in untreated neural precursors. Increase in cell apoptosis corroborated the decrease in cell number (Fig. 1 C). At 48 h postcontact with C8166 T cells, there was 24 ± 4.4% neural cell loss and 58.5 ± 9.2% at 72 h, p < 0.001. Interestingly, neural precursor death was observed at a same extent following treatment with culture supernatant of CD100-producing T cells, C8166, or coculture without contact (Transwell chamber) with these T cells. Thus, in contrast to T cell control, CD100-producing T cells dramatically enhanced the rate of natural death in neural precursors, probably via a soluble factor.

FIGURE 1.

CD100-induced death in human neural precursors. A–C, Apoptotic death detected at 48 h by the TUNEL method (A), and estimated by counting the number of active caspase-3-positive cells (B) and the total cell numbers (C) in human neural precursor cells transiently cocultured (20 h) with control or CD100-producing T lymphocytes (mean ± SEM cell number per field, microscopy AnalySiS, Student’s t test, representative experiment). D–F, Similar analysis on neural precursors following 48-h treatment with rsCD100 protein (0.5 ng/μl rsCD100 purified on CD100 BB18-mAb column).

FIGURE 1.

CD100-induced death in human neural precursors. A–C, Apoptotic death detected at 48 h by the TUNEL method (A), and estimated by counting the number of active caspase-3-positive cells (B) and the total cell numbers (C) in human neural precursor cells transiently cocultured (20 h) with control or CD100-producing T lymphocytes (mean ± SEM cell number per field, microscopy AnalySiS, Student’s t test, representative experiment). D–F, Similar analysis on neural precursors following 48-h treatment with rsCD100 protein (0.5 ng/μl rsCD100 purified on CD100 BB18-mAb column).

Close modal

The relevance to this observation was confirmed with additional cellular models: 1) primary culture of human fetal neural cells originating from the subventricular zone in contact with CD100-producing C8166 cells; 2) neural precursor cell line Dev in contact with primary T cells (CIB) activated by retroviral infection (CD25+, CD69+, CD100+). Human fetal precursors did not survive following contact with C8166 T cells, while contact with CEM T cell did not change their survival levels (data not shown). C8166 and primary T cells induced apoptosis in Dev cells at a similar level (6.9, 10.3, and 7.1% of apoptotic cells detected at 27, 48, and 72 h, respectively, following contact with C8166 vs 2.9, 3.5, and 5.2% after contact with CIB).

The effect of CD100-producing T cells was also examined on rat glial primary culture containing oligodendrocytes at various stages of maturation and astrocytes. Transient coculture with C8166 T cells induced cell damages of oligodendrocytes, but not of astrocytes, which remained unchanged in number and morphology. In fact, a progressive collapse and loss of process extensions were observed in glial cells identified as immature oligodendrocytes by immunodetection of the specific marker, GalCer (Fig. 2). The length and number of process extensions of these immature oligodendrocytes were progressively reduced after 24- and 48-h contact with CD100-producing T cells CEM (Fig. 2,B) compared with culture treated with control T cell (Fig. 2,A) or untreated culture (data not shown). The relative measure of the length (1 ± 0.2 μm vs 2.1 ± 0.3 μm in CEM-treated glial culture at 48 h) and the count of processes (1.3 ± 0.2 process per cell vs 5.3 ± 0.6 in CEM-treated glial culture at 48 h) were estimated in three independent experiments (Fig. 2,D). Contact with CD100-producing T cells induced collapse of cell process in immature oligodendrocytes, while the control T cells had no effect. Interestingly, the T cell-induced collapse of oligodendrocyte processes was followed by a late apoptotic death detected in the glial culture. As observed in three independent experiments, contact with CD100-producing T cells increased the number of TUNEL-positive cells (13.3 ± 0.3 per field vs 6.3 ± 0.3 in CEM-treated culture and 6 ± 0.86 in untreated culture; Fig. 3,A) and active caspase-3-positive cells (19.7 ± 8.1 per field vs 7.3 ± 5.5 in CEM-treated culture; data not shown). As shown in Fig. 3,B, apoptosis resulted in a partial loss in oligodendrocyte population, evidenced by the reduced number of oligodendrocytes per field (3.7 ± 0.7 oligodendrocytes per field vs 6.9 ± 0.5 in CEM-treated glial culture and 6.7 ± 0.8 in untreated culture) and the total oligodendrocyte number (26.8 ± 3.6% loss at 48 h following contact with C8166, p < 0.01). The number of astrocytes in the same treated cultures was not changed (10.7 ± 0.4, 12.0 ± 0.9, and 12.0 ± 0.9 astrocytes per field, respectively) (Fig. 3 B). In parallel experiments using astrocytes purified from rat glial culture or established in cell line (C8S), CD100-producing T lymphocytes never induced astrocyte death (data not shown).

FIGURE 2.

CD100-induced alteration of oligodendrocyte process. Oligodendrocytes detected in the rat primary glial culture by GalCer immunodetection (A). Morphological changes in GalCer-positive cells at 24 and 48 h following transient coculture with CD100+ T cells (B) or treatment with rsCD100 protein (C). No change after coculture with control T cells (A). D, Number and length of GalCer-positive cell process extensions after 24- and 48-h coculture with T cells or treatment with rsCD100 measured using microscopy AnalySiS (mean ± SEM of process number and length, 10–20 measures, 3 experiments).

FIGURE 2.

CD100-induced alteration of oligodendrocyte process. Oligodendrocytes detected in the rat primary glial culture by GalCer immunodetection (A). Morphological changes in GalCer-positive cells at 24 and 48 h following transient coculture with CD100+ T cells (B) or treatment with rsCD100 protein (C). No change after coculture with control T cells (A). D, Number and length of GalCer-positive cell process extensions after 24- and 48-h coculture with T cells or treatment with rsCD100 measured using microscopy AnalySiS (mean ± SEM of process number and length, 10–20 measures, 3 experiments).

Close modal
FIGURE 3.

CD100 induced the death of immature oligodendrocytes. A, Count of apoptotic (TUNEL+) cells in cocultured or rsCD100-treated glial culture (mean ± SEM positive cell number per field). B, Number of total oligodendrocytes and astrocytes in glial culture 48 h after coculture with control and CD100-producing T cells or treatment with rsCD100 (mean ± SEM cell number per field). C, Flow cytometry analysis of oligodendrocyte populations (percentage of total cell number) detected in rat glial culture by immunodetection of GalCer (immature oligodendrocytes) and NG2 (early oligodendrocyte progenitors) after coculture with CD100-producing T cells. D, Double labeling of GalCer and TUNEL identifying immature oligodendrocytes as dying cells.

FIGURE 3.

CD100 induced the death of immature oligodendrocytes. A, Count of apoptotic (TUNEL+) cells in cocultured or rsCD100-treated glial culture (mean ± SEM positive cell number per field). B, Number of total oligodendrocytes and astrocytes in glial culture 48 h after coculture with control and CD100-producing T cells or treatment with rsCD100 (mean ± SEM cell number per field). C, Flow cytometry analysis of oligodendrocyte populations (percentage of total cell number) detected in rat glial culture by immunodetection of GalCer (immature oligodendrocytes) and NG2 (early oligodendrocyte progenitors) after coculture with CD100-producing T cells. D, Double labeling of GalCer and TUNEL identifying immature oligodendrocytes as dying cells.

Close modal

Interestingly, cytometry analysis performed on glial culture following contact with CD100-producing T cells showed a dramatic decrease in the number of immature oligodendrocytes (Fig. 3,C) identified by immunodetection of GalCer (38 ± 2.4% of total oligodendrocytes vs 58.6 ± 6.6% in coculture with CEM), while population of early oligodendrocyte progenitors, identified by NG2-specific marker, was not significantly modified (47.8 ± 14.1% of total oligodendrocytes vs 30.6 ± 1.3 in coculture with CEM). In addition, double labeling of GalCer and TUNEL identified immature oligodendrocytes as the dying cells (Fig. 3 D). Collectively, these effects are reminiscent of the previously reported collapse of oligodendrocyte processes and death of neural precursors induced by secreted SEMA-3A, a brain-derived semaphorin (12, 13, 14).

We next looked at whether CD100 was involved in the T cell-induced neural cell damages. The fact that T cell supernatant reproduced the effect of T cell contact on neural cells led us to suspect sCD100 as the deleterious factor. The direct effect of CD100 was demonstrated by treating the neural precursor cells with rsCD100, released as a dimer from transfected Jurkat cells (15). Similar to the contact with CD100-producing T cells (Fig. 1, D–F), treatment with rsCD100 induced apoptotic death of neural precursors. Increase in the number of active caspase-3-positive cells was detected in rsCD100-treated neural precursors (25.3 ± 1.6 positive cells per field in culture treated with 0.5 ng/μl rsCD100 vs 10.8 ± 1.8 in untreated culture; Fig. 1,E) and was associated with a decreased cell number per field (Fig. 1,F). A prominent cell loss (43.5 ± 13.4%) was next observed at 72 h. Treatment with increasing doses of rsCD100 (0.01, 0.1, and 0.5 ng/μl) induced apoptosis of neural precursors in a dose-dependent manner (7.8 ± 1.2, 19.8 ± 3, and 21.8 ± 0.2 active caspase-3-positive cells per field, respectively, vs 6.4 ± 1.2 in culture treated with control supernatant; Fig. 4 A).

FIGURE 4.

Dose-dependent effect of CD100 on neural cells. Evaluation of apoptotic cell number expressing active caspase-3 in neural precursors (A) or glial primary culture (B) following treatment by increasing doses of rsCD100 (0.01, 0.1, and 0.5 ng/μl) (mean ± SEM active caspase-3 cells per field). rsCD100 was collected from CD100-transfected Jurkat cells and control supernatant (sControl) from CD27-transfected Jurkat cells.

FIGURE 4.

Dose-dependent effect of CD100 on neural cells. Evaluation of apoptotic cell number expressing active caspase-3 in neural precursors (A) or glial primary culture (B) following treatment by increasing doses of rsCD100 (0.01, 0.1, and 0.5 ng/μl) (mean ± SEM active caspase-3 cells per field). rsCD100 was collected from CD100-transfected Jurkat cells and control supernatant (sControl) from CD27-transfected Jurkat cells.

Close modal

Treatment of glial primary culture with rsCD100 also resulted in morphological evidence of cell damage in oligodendrocytes (Fig. 2). This treatment reduced the number of processes (1.4 ± 0.2 processes per oligodendrocyte at 48 h vs 5.3 ± 0.4 in untreated culture; Fig. 2,D) and their relative length (1 ± 0.2 μm at 48 h vs 2.1 ± 0.3 μm in untreated culture) in immature oligodendrocytes identified by GalCer immunodetection (Fig. 2,C). This morphological deterioration was followed, at 48 h, by an increased number of apoptotic cells, as shown by active caspase-3 expression (15.0 ± 2.3 positive cells per field in culture treated with 0.5 ng/μl rsCD100 vs 6.0 ± 1.9 positive cells in untreated culture; data not shown) and the number of TUNEL-positive cells (13.4 ± 0.8 per field vs 6.0 ± 0.86 in untreated culture; Fig. 3,A). Such an increase in apoptosis marker expression corroborated the decreased number of total oligodendrocytes under sCD100 treatment (3.6 ± 0.6 oligodendrocytes per field vs 6.7 ± 0.8 in untreated culture; Fig. 3,B) and resulted in 29 ± 4% loss of oligodendrocytes at 48 h posttreatment. By contrast, there was no alteration of astrocyte survival in the same cultures as shown in Fig. 3,B. In addition, increased doses of rsCD100 (0.1, 1, and 5 ng/μl) induced apoptosis in a dose-dependent manner (7.5 ± 0.7, 13.3 ± 1.8, and 16.6 ± 1.5 active caspase-3-positive cells per field, respectively, vs 6.0 ± 0.8 in untreated glial culture; Fig. 4,B). Double labeling of the glial culture with oligodendrocyte markers and TUNEL identified immature oligodendrocytes (GalCer positive) as a cell population sensitive to sCD100 treatment (data not shown). Thus, rsCD100 induced a deleterious effect on neural precursors and oligodendrocytes similarly to CD100-producing T cells, while no detectable change was observed on astrocytes. Higher rsCD100 dose (1 ng/μl) induced the death of 64% neural precursors and 37% total oligodendrocytes at 72 h. In control experiments, treatment of these human and rat neural cells with a supernatant from Jurkat stably transfected with an unrelated molecule, CD27, did not induce any change (Fig. 4 A), further demonstrating that this deleterious effect was attributable to sCD100.

Finally, involvement of CD100 in the T cell-mediated damage on neural cells was confirmed by treatment of cocultures with CD100-specific Abs. The anti-CD100 BD16 mAb, previously shown to inhibit activity of sCD100 (15), was able to antagonize the C8166 T cell-induced effects on neural precursors and immature oligodendrocytes (Fig. 5). In fact, BD16 mAb used at increasing concentrations progressively decreased, in a dose-dependent manner, the number of apoptotic neural precursors detected by the TUNEL method at 48 h postcontact with CD100-producing T cells (13.4 ± 1.4, 9.9 ± 0.2, 2.3 ± 0.6, and 3.2 ± 0.8 positive cells per field, respectively, vs 16.5 ± 0.9 in mAb-untreated coculture; Fig. 5,A). In addition, BD16 treatment reduced the T cell-mediated cell loss estimated at 72 h (41.5 ± 0.7% cell loss vs 62.0 ± 7.1% in mAb-untreated coculture). In contrast, anti-CD100 BB18 mAb, which has no ability to block sCD100 activity, did not reduce T cell-mediated damage (data not shown). Similarly, treatment of the glial primary cells with BD16 mAb along with the coculture of CD100-producing T cells reduced the T cell-mediated loss of oligodendrocytes. As shown in Fig. 5,B, BD16 treatment normalized the number of oligodendrocytes per field. BD16 treatment also decreased the number of TUNEL-positive cells compared with untreated coculture (Fig. 5, C3 vs C2). Thus, BD16 mAb exerted a protective effect toward immature oligodendrocytes as exemplified by the number of CNPase-positive cells higher in coculture under Ab treatment (Fig. 5, C3 vs C2).

FIGURE 5.

Reduction of the T cell-mediated damage by CD100-blocking BD16 mAb. Neural precursors (A) and glial culture (B and C) were cocultured with control T cells, or with CD100+ T cells and treated or not treated with anti-CD100 BD16 mAb (0.1, 0.4, 1, 2 ng/μl for neural precursors, 1 ng/μl for glial culture). Apoptotic cells were detected by the TUNEL method, and positive cell number was evaluated (mean ± SEM positive cells per field, microscopy AnalySiS). A, Reduction in a dose-dependent manner of T cell-mediated apoptosis in neural precursors by BD16 mAb. B, Count of astrocytes and oligodendrocytes in glial culture (mean ± SEM cell number per field, microscopy AnalySiS): decrease in oligodendrocyte number after coculture with CD100-producing T cells (B2) and resettled number under BD16 mAb treatment (B3). No modification in astrocyte population. C, A view of apoptosis detection by the TUNEL method (green) and of immature oligodendrocyte population by CNPase immunodetection (red) in glial primary culture. Increase in the number of TUNEL-positive cells after coculture with CD100-producing T cells (C2) compared with coculture with control T cells (C1) and reduction under BD16 mAb treatment (C3). In parallel, decrease in the number of CNPase-positive cells (C2) and limited reduction under BD16 mAb treatment (C3).

FIGURE 5.

Reduction of the T cell-mediated damage by CD100-blocking BD16 mAb. Neural precursors (A) and glial culture (B and C) were cocultured with control T cells, or with CD100+ T cells and treated or not treated with anti-CD100 BD16 mAb (0.1, 0.4, 1, 2 ng/μl for neural precursors, 1 ng/μl for glial culture). Apoptotic cells were detected by the TUNEL method, and positive cell number was evaluated (mean ± SEM positive cells per field, microscopy AnalySiS). A, Reduction in a dose-dependent manner of T cell-mediated apoptosis in neural precursors by BD16 mAb. B, Count of astrocytes and oligodendrocytes in glial culture (mean ± SEM cell number per field, microscopy AnalySiS): decrease in oligodendrocyte number after coculture with CD100-producing T cells (B2) and resettled number under BD16 mAb treatment (B3). No modification in astrocyte population. C, A view of apoptosis detection by the TUNEL method (green) and of immature oligodendrocyte population by CNPase immunodetection (red) in glial primary culture. Increase in the number of TUNEL-positive cells after coculture with CD100-producing T cells (C2) compared with coculture with control T cells (C1) and reduction under BD16 mAb treatment (C3). In parallel, decrease in the number of CNPase-positive cells (C2) and limited reduction under BD16 mAb treatment (C3).

Close modal

Semaphorins may signal in neural cells through receptors of the neuropilin and plexin families (39). Neuropilins act as coreceptors with A plexins, while B plexins alone behave as fully functional signal transducers for both transmembrane and secreted forms of semaphorins. Our previous work had shown that neuropilin-1 is present on human neural precursors and mediates SEMA-3A-induced apoptosis of these cells (12). To eliminate the involvement of neuropilin-1 and SEMA-3A in the T cell-induced apoptosis of neural precursors, we added anti-MAM neuropilin-1 Ab in the T neural cell coculture. This Ab, previously shown to block the SEMA-3A-induced neural precursor death (12), had no significant effect on the rate of death induced by CD100-producing T cells (Fig. 6,A). In addition, the remote possibility of an effect of anti-CD100 BD16 mAb on SEMA-3A activity (15) was ruled out by the absence of SEMA-3A expression in T cells, as shown by RT-PCR analysis of SEMA-3A mRNA in neural and T cells (Fig. 6,B). These observations excluded the implication of neuropilin-1/SEMA-3A in the T cell-mediated damages. Involvement of plexins, in particular plexin-B1, identified as a receptor for sCD100 in human cells (39), was next investigated. The presence of mRNA coding for plexin-B1, -B2, or -B3 in human neural precursors and human fetal cortex (used as positive control) was evidenced by RT-PCR. Taking into account the high homology of RNA sequence between these three plexins, oligomers were selected for their ability to amplify a common 300-bp amplicon. Amplicons were further identified by restriction enzymes, PstI and Sau3AI, which can generate 171- and 129-bp products in the plexin-B1 and plexin-B3 amplicon, and 205- and 95-bp products in the plexin-B1 and plexin-B2 amplicon, respectively. PstI and Sau3AI generated the expected products from a 300-bp amplicon (Fig. 6,C), showing the presence of plexin-B-coding mRNA in human neural precursors as in human fetal cortex. The presence of plexin-B1 at the cell membrane was confirmed by immunodetection on live neural precursors, and flow cytometry detected 50–54% positive cells (Fig. 6,D). The possible involvement of plexin-B1 in the CD100-producing T cell-me-diated apoptosis in neural precursors was suggested by results of treatment with anti-plexin-B1 Ab. When added throughout the coculture, anti-plexin-B1 Ab reduced the number of apoptotic neural cells (Fig. 6 A) and the rate of cell loss (5 vs 25.4% in untreated coculture at 48 h) induced by CD100-producing T cells. Treatment of rat oligodendrocytes cocultured with CD100-producing T cells did not modify the rate of cell apoptosis (data not shown), probably due to the human specificity of this anti-plexin Ab.

FIGURE 6.

Involvement of semaphorin receptor plexin-B in CD100-mediated effect on neural cells. A, Number of apoptotic/active caspase-3-positive neural precursor cells in the presence of CD100-producing T cells with or without anti-plexin-B1 and anti-neuropilin-1 Abs (mean ± SEM active caspase-3-positive cells per field, microscopy AnalySiS). Reduction of T cell-mediated apoptosis by anti-plexin-B1 Ab. B, Detection by RT-PCR and Southern blotting using [33P]dATP 5′ end-labeled internal probes of mRNA coding for SEMA-3A and β-actin (as housekeeping gene) in: 1) glial primary culture; 2) neural precursors; 3) neural precursors + CD100+ T cells; and 4) CD100+ T cells. T cells did not express SEMA-3A. C, Detection by RT-PCR of mRNA coding for a 300-bp amplicon common to plexin-B1, plexin-B2, and plexin-B3 in human neural precursors and human fetal brain. Identification of the subsequent generation of 171- and 129-bp products by PstI in plexin-B1 and -B3 amplicon, and of 205- and 95-bp products by Sau3AI in plexin-B1 and -B2 amplicon. mRNA coding for plexin-B was present in neural precursors as in fetal brain. D, Plexin-B1 detected in neural precursor cells by immunofluorescence and flow cytometry analysis (54% positive).

FIGURE 6.

Involvement of semaphorin receptor plexin-B in CD100-mediated effect on neural cells. A, Number of apoptotic/active caspase-3-positive neural precursor cells in the presence of CD100-producing T cells with or without anti-plexin-B1 and anti-neuropilin-1 Abs (mean ± SEM active caspase-3-positive cells per field, microscopy AnalySiS). Reduction of T cell-mediated apoptosis by anti-plexin-B1 Ab. B, Detection by RT-PCR and Southern blotting using [33P]dATP 5′ end-labeled internal probes of mRNA coding for SEMA-3A and β-actin (as housekeeping gene) in: 1) glial primary culture; 2) neural precursors; 3) neural precursors + CD100+ T cells; and 4) CD100+ T cells. T cells did not express SEMA-3A. C, Detection by RT-PCR of mRNA coding for a 300-bp amplicon common to plexin-B1, plexin-B2, and plexin-B3 in human neural precursors and human fetal brain. Identification of the subsequent generation of 171- and 129-bp products by PstI in plexin-B1 and -B3 amplicon, and of 205- and 95-bp products by Sau3AI in plexin-B1 and -B2 amplicon. mRNA coding for plexin-B was present in neural precursors as in fetal brain. D, Plexin-B1 detected in neural precursor cells by immunofluorescence and flow cytometry analysis (54% positive).

Close modal

The functional relevance of these experimental data was tested by examining the presence of sCD100 in CSF and CD100-positive infiltrating T lymphocytes in postmortem CNS from patients suffering with demyelination associated with neuroinflammation, TSP/HAM. CD100 was detected in spinal cord and CSF from these patients, but not from patients with noninflammatory neurological injuries (Table II and Fig. 7). Three patients suffering with TSP/HAM and exhibiting various levels of neuroinflammation evidenced by immune infiltrates were chosen for the histological study. Anatomopathology analysis of paraffin-embedded spinal cords, using standard staining (H&E-safran), Luxol Fast Blue for myelin detection, and immunofluorescence for cell identification, indicated that spinal cords from these patients displayed demyelination (Fig. 7,A) associated with moderate to marked meningial (Fig. 7, A–C) and perivascular (Fig. 7, D and E) infiltrates of immune cells. Level of neuroinflammation correlated with the extent of atrophy in the thoracic level and degeneration of the lateral corticospinal and spinocerebellar tracts, the diffuse loss of myelin and axons, and the thickness of the leptomeninges and astrocytosis, as previously observed (19) (also see Table I). Immunofluorescence analysis of spinal cords with anti-CD4, anti-CD8, anti-CD45RO, or anti-fascin Abs identified the infiltrating immune cells as primarily CD45RO-positive cells and T lymphocytes with a predominance of CD8 T cells (Fig. 7,B, case 1). As BD16 and BB18 anti-CD100 mAbs did not react with fixed cells, we developed a new polyclonal rabbit Ab recognizing an intracellular peptide of the CD100 molecule to determine whether the membrane-bound CD100 was detected on infiltrated immune cells. Double labeling using anti-CD100 and anti-CD45RO revealed the coexpression of these molecules in cell from the hemopoietic lineage. The highest density of double-labeled cells was found in meningia (Fig. 7,C) and around blood vessels (Fig. 7, E and F). Double-labeled T lymphocytes were also observed in parenchyma (Fig. 7,F) both in gray and white matters. Astrocytosis in the white matter and around blood vessels was revealed by GFAP immunostaining (Fig. 7,G) of spinal cord from patient with high inflammation level (case 1). Interestingly, we had detected the matrix metalloproteinase-9 within astrocytosis in the spinal cord of this patient (35). In contrast, the few CD45RO-positive cells present in spinal cords of nonneuroinflammatory patients were all CD100 negative (two cases studied; data not shown). Assuming that the expression and release of sCD100 may be enhanced within the inflammatory neural tissue and CSF by high levels of metalloproteinases in TSP/HAM patients (40, 41), the expression of sCD100 was analyzed in the CSF from patients with TSP/HAM (n = 9) or patients with nonneuroinflammatory diseases (n = 7) by ELISA. This analysis revealed the presence of sCD100 at a level of 97.7 ± 23.1 ng/ml in CSF from TSP/HAM patients (Table II), while levels were undetectable in the CSF from other groups of patients.

Table II.

sCD100 detected by ELISA in the CSF from patients infected with HTLV-1 and suffering with demyelination associated with virus-induced neuroinflammation (TSP/HAM)a

TSP/HAM PatientsCD100 (ng/ml)
132.6 ± 29.9 
66.6 ± 14.2 
158.3 ± 17.2 
79.6 ± 36.3 
123.9 ± 42.7 
60 ± 9.9 
62.9 ± 14.2 
57.9 ± 16.3 
75 ± 27.6 
TSP/HAM PatientsCD100 (ng/ml)
132.6 ± 29.9 
66.6 ± 14.2 
158.3 ± 17.2 
79.6 ± 36.3 
123.9 ± 42.7 
60 ± 9.9 
62.9 ± 14.2 
57.9 ± 16.3 
75 ± 27.6 
a

Patients with other neurological diseases (n = 7) were negative for CD100 detection in CSF.

FIGURE 7.

Detection of CD100-expressing T cells in the spinal cord of patients. Examination of spinal cords from three patients infected with HTLV-1 and suffering with demyelination associated with virus-induced neuroinflammation (TSP/HAM). A and D, H&E-safran staining and Luxol Fast Blue for myelin detection (histology) showed demyelination (A, open areas in blue staining, open arrow) and immune cell infiltration (filled arrow) in meningia (A) and around blood vessels (D). B, Immunodetection of the T cell markers CD4 and CD8 in meningia. E, C, and F, Immunodetection of CD100 (green) and CD45RO (red) in meningia (C), around blood vessel (E), and in parenchyma (F). Arrow triangle = coexpression (orange); arrow square = monoexpression. Moderate to marked infiltration is seen in F: F1, case 1; F2, case 2; F3, case 3. Astrocytosis identified by GFAP immunodetection (G, case 1).

FIGURE 7.

Detection of CD100-expressing T cells in the spinal cord of patients. Examination of spinal cords from three patients infected with HTLV-1 and suffering with demyelination associated with virus-induced neuroinflammation (TSP/HAM). A and D, H&E-safran staining and Luxol Fast Blue for myelin detection (histology) showed demyelination (A, open areas in blue staining, open arrow) and immune cell infiltration (filled arrow) in meningia (A) and around blood vessels (D). B, Immunodetection of the T cell markers CD4 and CD8 in meningia. E, C, and F, Immunodetection of CD100 (green) and CD45RO (red) in meningia (C), around blood vessel (E), and in parenchyma (F). Arrow triangle = coexpression (orange); arrow square = monoexpression. Moderate to marked infiltration is seen in F: F1, case 1; F2, case 2; F3, case 3. Astrocytosis identified by GFAP immunodetection (G, case 1).

Close modal

The current pathogenesis sequence of inflammatory demyelinating diseases, such as MS, starts with an inflammatory phase, which fulfills the criteria of an autoimmune disease, followed by a phase of selective demyelination, and finally a neurodegenerative phase (42, 43, 44). The initial event of inflammation is the migration of activated T lymphocytes of Th1 phenotype to the brain and spinal cord, where they release deleterious proinflammatory cytokines. There is indeed in vivo evidence that Th1-related cytokines such as TNF-α and IFN-γ are involved in lesion formation (22, 25, 45). However, heterogeneity with respect to clinical course and response to therapy has led to the concept of heterogeneous pathogenic mechanisms of demyelination (46). Additional demyelinating amplification factors are probably required to produce the large demyelinating lesions detected in patients. The present study points out the deleterious effect of the immune semaphorin CD100 on oligodendrocyte and neural precursor integrity and proposes that immune semaphorins are candidates in the pathogenic mechanism of demyelination.

We show that sCD100, reported to be highly expressed in activated T lymphocytes and released from the T cell surface through a metalloproteinase-mediated proteolytic cleavage (17), induced a progressive decrease in process extensions of immature oligodendrocytes, followed by their death, and the death of human multipotent neural precursors. Thus, sCD100 of immune origin can initiate a signaling cascade impairing neural precursor and oligodendrocyte homeostasis, similarly to the neural SEMA-3A (12, 13, 47). Interestingly, sCD100 affected mainly the immature oligodendrocytes (GalCer positive), while oligodendrocyte precursors (NG2 positive) were preserved from T cell-mediated effect and proliferate. This observation suggests the presence of a window of vulnerability to CD100 in immature oligodendrocytes and not in precursors. In fact, NG2-positive oligodendrocyte precursors are the principal dividing cell population within the perinatal and adult brain (48), but after a period of perinatal proliferation, GalCer-positive differentiating oligodendrocytes send a signal back to their precursors preventing their differentiation. The alteration of this feedback mechanism is thought to cause the rapid proliferation of NG2-positive cells during demyelinating disease. In our model, the selective loss of GalCer-positive oligodendrocytes, the negative proliferation signal, following contact with sCD100-producing T lymphocytes, could explain the sustained presence of NG2-positive oligodendrocytes in the culture. The differential susceptibility of astrocytes and oligodendrocytes to CD100-induced damage is intriguing, as both cells express plexin-B1. However, apoptosis in neural cells is strictly controlled by sequestration of active caspase, regulation of death receptor activity, and expression of apoptosis-inhibitory proteins (49). Cell type-specific regulations at these various levels could explain the differential susceptibility to CD100. The capacity of sCD100-blocking mAb, BD16, to reduce the damage to neural cells mediated by activated T lymphocytes, in contrast to the nonblocking BB18 mAb, confirms the physiological role of sCD100. In addition, detection of sCD100 in the CSF of patients suffering with neuroinflammatory demyelination (TSP/HAM) and the presence of numerous infiltrating CD100/CD45RO-positive cells in their spinal cord, in contrast to patients suffering with noninflammatory neurological injury, support the idea of a correlation between CD100 release and demyelination. The collapse of process extensions in the immature/premyelinating oligodendrocyte population and their death induced by sCD100 would dramatically compromise the capacity of remyelination in inflamed brain. The present data together with 1) the abundant expression of CD100 by activated T cells (7, 10), 2) the elevated expression of metalloproteinases within the CNS of patients suffering neuroinflammatory demyelination (23, 40), and 3) the gradual reduction in the size of the oligodendrocyte population, but the sustained presence of NG2-positive cells observed in brains of patients suffering neuroinflammatory demyelination (27, 50), strongly suggest that infiltrated/activated T cells affect myelination through release of factors toxic for neural cells, including semaphorin CD100.

In addition to its deleterious effects on neural cells, sCD100 could participate in the autoimmune response within the brain of patients suffering with neuroinflammatory demyelination (21, 51). In fact, the two murine immune semaphorins, CD100 and SEMA-4A, are expressed constitutively in T cells and function as soluble ligands to CD72 and Tim2, respectively, to regulate the humoral and cellular immune response (7, 8, 9). They are suspected to be involved in autoimmune diseases (6, 10, 52). sCD100 produced after lymphocyte activation has a potent costimulatory function to induce proliferation of B cells and to enhance the maturation of professional APCs (8, 53). This is reflected by the presence of significant levels of sCD100 in the serum of MLR/lpr mice developing autoantibodies (10). In addition, mice carrying a truncated CD100 transgene encoding an easily released sCD100 develop EAE more rapidly than normal mice (52). The relation between this autoimmune response to the CD100 transgene overexpression was proved by enhanced concentration of sCD100 in body fluids and, conversely, by the fact that the CD100−/− mice are resistant to EAE. A direct role for sCD100 in oligodendrocyte pathogenesis remains to be investigated. More recently, the implication of SEMA-4A in the differentiation and activation of T cells upon interaction with its receptor Tim2 was reported (6). Interestingly, treating mice with anti-SEMA-4A mAb blocked the development of EAE induced by the antigenic peptide derived from myelin oligodendrocyte glycoprotein. Although the mAb effect was evident when injected at an early phase of T cell responses against myelin oligodendrocyte glycoprotein, it would be interesting to look for a direct effect of SEMA-4A on oligodendrocytes. The fact that SEMA-3A and sCD100 can inhibit the migration of immune cells (15) also suggests that these semaphorins may affect migratory T lymphocytes within the CNS. Taken together, these works and our findings in humans indicate that in neuroinflammatory situation, sCD100 and SEMA-3A could participate, through binding to immune and neural cells, in a close partnership, both in the sequestration and stimulation of immune cells within the CNS and in the collapse of oligodendrocyte processes and neural precursor death.

The presence of neuropilin-1 in neural precursors and oligodendrocytes and its involvement in the class 3 semaphorin-mediated apoptosis of neural precursors and changes in oligodendrocytes were previously reported (12, 13, 16). Our present data show that neuropilin-1 is not the signaling receptor of sCD100 in neural cells, as blocking anti-neuropilin-1 Ab failed to prevent CD100-mediated apoptosis. As B plexins fully function as signal transducers for both transmembrane and secreted forms of class 4 semaphorins (39), members of this plexin family could participate in sCD100 signaling. The presence of plexin-B1 in neural precursors and the capacity of anti-plexin-B1 Ab to antagonize the T cell-induced apoptosis of neural precursors may support this hypothesis. Thus, sCD100 and SEMA-3A can exert a similar paralytic effect on immune cell migration via an identical cell surface receptor, while their deleterious effects leading to neural cell alteration and death are mediated by different receptors.

In conclusion, this original paradigm of activated T lymphocytes/neural cell interaction and the presence of sCD100- and CD100-positive infiltrating T cells in CNS of patients suffering of neuroinflammatory demyelination point to the role of the immune semaphorin CD100/SEMA-4D in the cross talk between the immune system and the CNS. Immune semaphorins could be crucial in demyelinating diseases, including MS and TSP/HAM, in which a causal relationship among inflammation, oligodendrocyte loss, axonal damage, and irreversible neurological disability is suspected (19, 25, 42, 54).

We thank Isabelle Chabert de Ponnat for FACS analysis.

1

This work was supported by grants from French Agencies of Research on Multiple Sclerosis and on HIV (Agence Recherche Selérose Plaques, Ligue Recherche Sclérose Plaques, Association Francaise Contre les Myopathies, and Agence Nationale de Recherches sur le SIDA) and Institut National de la Santé et de la Recherche Médicale funds.

4

Abbreviations used in this paper: SEMA, semaphorin; CNPase, cyclic nucleotide phosphodiesterase; CSF, cerebrospinal fluid; EAE, experimental autoimmune encephalitis; GalCer, galactocerebroside; GFAP, glial acidic fibrillary protein; HTLV-1, human T lymphotropic virus type 1; MS, multiple sclerosis; sCD100, soluble CD100; TSP/HAM, tropical spastic paraparesis/HTLV-1-associated myelopathy.

1
Boulanger, L. M., G. S. Huh, C. J. Shatz.
2001
. Neuronal plasticity and cellular immunity: shared molecular mechanisms.
Curr. Opin. Neurobiol.
11
:
568
.
2
Liu, B. P., S. M. Strittmatter.
2001
. Semaphorin-mediated axonal guidance via Rho-related G proteins.
Curr. Opin. Cell Biol.
13
:
619
.
3
Kolodkin, A. L., D. J. Matthes, T. P. O’Connor, N. H. Patel, A. Admon, D. Bentley, C. S. Goodman.
1992
. Fasciclin IV: sequence, expression, and function during growth cone guidance in the grasshopper embryo.
Neuron
9
:
831
.
4
Chisholm, A., M. Tessier-Lavigne.
1999
. Conservation and divergence of axon guidance mechanisms.
Curr. Opin. Neurobiol.
9
:
603
.
5
Hall, K. T., L. Boumsell, J. L. Schultze, V. A. Boussiotis, D. M. Dorfman, A. A. Cardoso, A. Bensussan, L. M. Nadler, G. J. Freeman.
1996
. Human CD100, a novel leukocyte semaphorin that promotes B-cell aggregation and differentiation.
Proc. Natl. Acad. Sci. USA
93
:
11780
.
6
Kumanogoh, A., S. Marukawa, K. Suzuki, N. Takegahara, C. Watanabe, E. Ch’ng, I. Ishida, H. Fujimura, S. Sakoda, K. Yoshida, H. Kikutani.
2002
. Class IV semaphorin Sema4A enhances T-cell activation and interacts with Tim-2.
Nature
419
:
629
.
7
Bismuth, G., L. Boumsell.
2002
. Controlling the immune system through semaphorins.
Sci. STKE
2002
:
RE4
.
8
Suzuki, K., A. Kumanogoh, H. Kikutani.
2003
. CD100/Sema4D, a lymphocyte semaphorin involved in the regulation of humoral and cellular immune responses.
Cytokine Growth Factor Rev.
14
:
17
.
9
Kumanogoh, A., H. Kikutani.
2001
. The CD100-CD72 interaction: a novel mechanism of immune regulation.
Trends Immunol.
22
:
670
.
10
Wang, X., A. Kumanogoh, C. Watanabe, W. Shi, K. Yoshida, H. Kikutani.
2001
. Functional soluble CD100/Sema4D released from activated lymphocytes: possible role in normal and pathologic immune responses.
Blood
97
:
3498
.
11
Goshima, Y., T. Ito, Y. Sasaki, F. Nakamura.
2002
. Semaphorins as signals for cell repulsion and invasion.
J. Clin. Invest.
109
:
993
.
12
Bagnard, D., C. Vaillant, S. T. Khuth, N. Dufay, M. Lohrum, A. W. Puschel, M. F. Belin, J. Bolz, N. Thomasset.
2001
. Semaphorin 3A-vascular endothelial growth factor-165 balance mediates migration and apoptosis of neural progenitor cells by the recruitment of shared receptor.
J. Neurosci.
21
:
3332
.
13
Ricard, D., V. Rogemond, E. Charrier, M. Aguera, D. Bagnard, M. F. Belin, N. Thomasset, J. Honnorat.
2001
. Isolation and expression pattern of human Unc-33-like phosphoprotein 6/collapsin response mediator protein 5 (Ulip6/CRMP5): coexistence with Ulip2/CRMP2 in Sema3a-sensitive oligodendrocytes.
J. Neurosci.
21
:
7203
.
14
Shirvan, A., R. Shina, I. Ziv, E. Melamed, A. Barzilai.
2000
. Induction of neuronal apoptosis by Semaphorin3A-derived peptide.
Brain Res. Mol. Brain Res.
83
:
81
.
15
Delaire, S., C. Billard, R. Tordjman, A. Chedotal, A. Elhabazi, A. Bensussan, L. Boumsell.
2001
. Biological activity of soluble CD100. II. Soluble CD100, similarly to H-SemaIII, inhibits immune cell migration.
J. Immunol.
166
:
4348
.
16
Tordjman, R., Y. Lepelletier, V. Lemarchandel, M. Cambot, P. Gaulard, O. Hermine, P. H. Romeo.
2002
. A neuronal receptor, neuropilin-1, is essential for the initiation of the primary immune response.
Nat. Immun.
3
:
477
.
17
Elhabazi, A., S. Delaire, A. Bensussan, L. Boumsell, G. Bismuth.
2001
. Biological activity of soluble CD100. I. The extracellular region of CD100 is released from the surface of T lymphocytes by regulated proteolysis.
J. Immunol.
166
:
4341
.
18
Prineas, J. W., R. O. Barnard, T. Revesz, E. E. Kwon, L. Sharer, E. S. Cho.
1993
. Multiple sclerosis: pathology of recurrent lesions.
Brain
116
:
681
.
19
Cartier, L. M., J. G. Cea, C. Vergara, F. Araya, P. Born.
1997
. Clinical and neuropathological study of six patients with spastic paraparesis associated with HTLV-I: an axomyelinic degeneration of the central nervous system.
J. Neuropathol. Exp. Neurol.
56
:
403
.
20
Osame, M., K. Arimura, M. Nakagawa, F. Umehara, K. Usuku, S. Ijichi.
1997
. HTLV-I associated myelopathy (HAM): review and recent studies.
Leukemia
11
:(Suppl. 3):
63
.
21
t Hart, B. A., H. P. Brok, S. Amor, R. E. Bontrop.
2001
. The major histocompatibility complex influences the ethiopathogenesis of MS-like disease in primates at multiple levels.
Hum. Immunol.
62
:
1371
.
22
Jacobson, S..
2002
. Immunopathogenesis of human T cell lymphotropic virus type I-associated neurologic disease.
J. Infect. Dis.
186
:(Suppl. 2):
S187
.
23
Cuzner, M. L., D. Gveric, C. Strand, A. J. Loughlin, L. Paemen, G. Opdenakker, J. Newcombe.
1996
. The expression of tissue-type plasminogen activator, matrix metalloproteases and endogenous inhibitors in the central nervous system in multiple sclerosis: comparison of stages in lesion evolution.
J. Neuropathol. Exp. Neurol.
55
:
1194
.
24
Yamano, Y., M. Nagai, M. Brennan, C. A. Mora, S. S. Soldan, U. Tomaru, N. Takenouchi, S. Izumo, M. Osame, S. Jacobson.
2002
. Correlation of human T-cell lymphotropic virus type 1 (HTLV-1) mRNA with proviral DNA load, virus-specific CD8+ T cells, and disease severity in HTLV-1-associated myelopathy (HAM/TSP).
Blood
99
:
88
.
25
Grant, C., K. Barmak, T. Alefantis, J. Yao, S. Jacobson, B. Wigdahl.
2002
. Human T cell leukemia virus type I and neurological disease: events in bone marrow, peripheral blood, and central nervous system during normal immune surveillance and neuroinflammation.
J. Cell. Physiol.
190
:
133
.
26
Blakemore, W. F., R. J. Franklin.
2000
. Transplantation options for therapeutic central nervous system remyelination.
Cell Transplant.
9
:
289
.
27
Chang, A., W. W. Tourtellotte, R. Rudick, B. D. Trapp.
2002
. Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis.
N. Engl. J. Med.
346
:
165
.
28
Horner, P. J., A. E. Power, G. Kempermann, H. G. Kuhn, T. D. Palmer, J. Winkler, L. J. Thal, F. H. Gage.
2000
. Proliferation and differentiation of progenitor cells throughout the intact adult rat spinal cord.
J. Neurosci.
20
:
2218
.
29
Narayanan, S., L. Fu, E. Pioro, N. De Stefano, D. L. Collins, G. S. Francis, J. P. Antel, P. M. Matthews, D. L. Arnold.
1997
. Imaging of axonal damage in multiple sclerosis: spatial distribution of magnetic resonance imaging lesions.
Ann. Neurol.
41
:
385
.
30
Howard, A. K., D. K. Li, J. Oger.
2003
. MRI contributes to the differentiation between MS and HTLV-I associated myelopathy in British Columbian coastal natives.
Can. J. Neurol. Sci.
30
:
41
.
31
Coleman, M. P., V. H. Perry.
2002
. Axon pathology in neurological disease: a neglected therapeutic target.
Trends Neurosci.
25
:
532
.
32
De Winter, F., M. Oudega, A. J. Lankhorst, F. P. Hamers, B. Blits, M. J. Ruitenberg, R. J. Pasterkamp, W. H. Gispen, J. Verhaagen.
2002
. Injury-induced class 3 semaphorin expression in the rat spinal cord.
Exp. Neurol.
175
:
61
.
33
Zhang, S. C., B. Ge, I. D. Duncan.
1999
. Adult brain retains the potential to generate oligodendroglial progenitors with extensive myelination capacity.
Proc. Natl. Acad. Sci. USA
96
:
4089
.
34
Popovic, M., G. Lange-Wantzin, P. S. Sarin, D. Mann, R. C. Gallo.
1983
. Transformation of human umbilical cord blood T cells by human T-cell leukemia/lymphoma virus.
Proc. Natl. Acad. Sci. USA
80
:
5402
.
35
Giraudon, P., R. Szymocha, S. Buart, A. Bernard, L. Cartier, M. F. Belin, H. Akaoka.
2000
. T lymphocytes activated by persistent viral infection differentially modify the expression of metalloproteinases and their endogenous inhibitors, TIMPs, in human astrocytes: relevance to HTLV-I-induced neurological disease.
J. Immunol.
164
:
2718
.
36
Derrington, E. A., N. Dufay, B. B. Rudkin, M. F. Belin.
1998
. Human primitive neuroectodermal tumor cells behave as multipotent neural precursors in response to FGF2.
Oncogene
17
:
1663
.
37
Buzanska, L., N. Spassky, M. F. Belin, A. Giangrande, F. Guillemot, C. Klambt, M. Labouesse, J. L. Thomas, K. Domanska-Janik, B. Zalc.
2001
. Human medulloblastoma cell line DEV is a potent tool to screen for factors influencing differentiation of neural stem cells.
J. Neurosci. Res.
65
:
17
.
38
Szymocha, R., H. Akaoka, M. Dutuit, C. Malcus, M. Didier-Bazes, M. F. Belin, P. Giraudon.
2000
. Human T-cell lymphotropic virus type 1-infected T lymphocytes impair catabolism and uptake of glutamate by astrocytes via Tax-1 and tumor necrosis factor α.
J. Virol.
74
:
6433
.
39
Tamagnone, L., S. Artigiani, H. Chen, Z. He, G. I. Ming, H. Song, A. Chedotal, M. L. Winberg, C. S. Goodman, M. Poo, et al
1999
. Plexins are a large family of receptors for transmembrane, secreted, and GPI-anchored semaphorins in vertebrates.
Cell
99
:
71
.
40
Giraudon, P., S. Buart, A. Bernard, N. Thomasset, M. F. Belin.
1996
. Extracellular matrix-remodeling metalloproteinases and infection of the central nervous system with retrovirus human T-lymphotropic virus type I (HTLV-I).
Prog. Neurobiol.
49
:
169
.
41
Umehara, F., Y. Okada, N. Fujimoto, M. Abe, S. Izumo, M. Osame.
1998
. Expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in HTLV-I-associated myelopathy.
J. Neuropathol. Exp. Neurol.
57
:
839
.
42
Neuhaus, O., J. J. Archelos, H. P. Hartung.
2003
. Immunomodulation in multiple sclerosis: from immunosuppression to neuroprotection.
Trends Pharmacol. Sci.
24
:
131
.
43
Steinman, L..
2001
. Multiple sclerosis: a two-stage disease.
Nat. Immun.
2
:
762
.
44
Bjartmar, C., B. D. Trapp.
2001
. Axonal and neuronal degeneration in multiple sclerosis: mechanisms and functional consequences.
Curr. Opin. Neurol.
14
:
271
.
45
Zhang, J., S. Markovic-Plese, B. Lacet, J. Raus, H. L. Weiner, D. A. Hafler.
1994
. Increased frequency of interleukin 2-responsive T cells specific for myelin basic protein and proteolipid protein in peripheral blood and cerebrospinal fluid of patients with multiple sclerosis.
J. Exp. Med.
179
:
973
.
46
Lassmann, H..
2002
. Mechanisms of demyelination and tissue destruction in multiple sclerosis.
Clin. Neurol. Neurosurg.
104
:
168
.
47
Pasterkamp, R. J., J. Verhaagen.
2001
. Emerging roles for semaphorins in neural regeneration.
Brain Res. Brain Res. Rev.
35
:
36
.
48
Levine, J. M., R. Reynolds, J. W. Fawcett.
2001
. The oligodendrocyte precursor cell in health and disease.
Trends Neurosci.
24
:
39
.
49
Leist, M., M. Jaattela.
2001
. Four deaths and a funeral: from caspases to alternative mechanisms.
Nat. Rev. Mol. Cell Biol.
2
:
589
.
50
Wolswijk, G..
2002
. Oligodendrocyte precursor cells in the demyelinated multiple sclerosis spinal cord.
Brain
125
:
338
.
51
Levin, M. C., M. Krichavsky, J. Berk, S. Foley, M. Rosenfeld, J. Dalmau, G. Chang, J. B. Posner, S. Jacobson.
1998
. Neuronal molecular mimicry in immune-mediated neurologic disease.
Ann. Neurol.
44
:
87
.
52
Shi, W., A. Kumanogoh, C. Watanabe, J. Uchida, X. Wang, T. Yasui, K. Yukawa, M. Ikawa, M. Okabe, J. R. Parnes, et al
2000
. The class IV semaphorin CD100 plays nonredundant roles in the immune system: defective B and T cell activation in CD100-deficient mice.
Immunity
13
:
633
.
53
Kikutani, H., A. Kumanogoh.
2003
. Semaphorins in interactions between T cells and antigen-presenting cells.
Nat. Rev. Immunol.
3
:
159
.
54
Versijpt, J., K. Van Laere, R. A. Dierckx, F. Dumont, P. P. De Deyn, G. Slegers, J. Korf.
2003
. Scintigraphic visualization of inflammation in neurodegenerative disorders.
Nucl. Med. Commun.
24
:
209
.