| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Neuroscience, and
Microbiology and Tumor Biology Center, Karolinska Institutet, Stockholm, Sweden
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NK cells normally play a crucial role in the innate and early-induced immune response against many infectious agents (reviewed in Ref. 1). Like cytotoxic T cells, they can readily kill target cells but, in contrast to cytotoxic T cells, do so without a requirement for initial priming. A role for NK cells in nerve cell killing has been implicated in the selective death of sympathetic neurons resident in the superior cervical ganglia of rats after exposure to the drug guanethidine (2, 3). Guanethidine, an adrenergic blocking agent, causes an extensive destruction of neurons in the superior cervical ganglion, a phenomenon that was previously observed to be accompanied by a mononuclear inflammatory cell infiltration (4, 5). Subsequent studies demonstrated that immunosuppression partly prevented this neuronal degeneration (6), indicating an immune-mediated mechanism. The hypothesis that NK cells might be involved in this immune-mediated neuronal destruction derives from the following observations: 1) The degeneration occurs even in athymic nude rats, which lack T lymphocytes (7). 2) Lymphocytes with a characteristic NK cell phenotype are present among the infiltrating mononuclear cells in the ganglia (8). 3) Depletion of NK cells with NK cell-specific Abs prevents the guanethidine-induced destruction of neurons (2, 3). These observations suggest that NK cells may function as a principle effector cell in neurodegenerative disorders of the nervous system. However, the mechanism of action of NK cells in this model has not been addressed in detail. In particular, a role for NK cells in direct nerve cell killing has to our knowledge not been explored.
The aim of the present study was to examine whether dorsal root ganglia (DRG)3 neurons can be targets for an NK cell-mediated cytotoxic attack, and to characterize the conditions for, and mechanisms of, a potential killing. We demonstrate that NK cells can kill mouse dorsal root ganglia neurons in culture, and that nerve cell lysis is mediated by the release of perforin.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
For preparation of DRG cultures, embryos from pregnant C57BL/6 (B6; obtained from B&K Universal AB, Sollentuna, Sweden) and ß2-microglobulin-/- (ß2m-/-) (9) mice were used. Effector cells were prepared from male B6, perforin-/- (POKB-/-) (10) and recombinase-activating gene-1-/- (11) mice. All mice were bred and maintained at the Microbiology and Tumor Biology Center when not else noted. All mice that served as a source of effector cells were 6–8 wk of age. The experiments were performed with ethical permission from Stockholm’s Norra Djurförsöksetiska Nämnd, and animal care was in accordance with institutional guidelines.
Establishment of DRG primary cultures
Cultures were prepared essentially as described previously (12). Briefly, DRG were dissected from 15- to 16-day embryonic mice. The ganglia were collected in media and mechanically dissociated by repeated suspensions through a narrowed Pasteur pipette. The dissociated cells were seeded into 35-mm petri dishes (Corning Costar, Cambridge, MA; 1.5 x 105 cells/dish), which were precoated with collagen (Vitrogen 100; Collagen Biomaterials, Palo Alto, CA). One type of culture, the high glial cell density culture, was grown in MEM supplemented with 10% horse serum, 10% FBS, 1 mM L-glutamine, 2% chicken embryo extract, 15 µg/ml gentamicin (all obtained from Life Technologies, Paisley, U.K.), 6 mg/ml glucose, and 10 ng/ml nerve growth factor (Sigma, St. Louis, MO). On the fourth day in vitro, the medium was changed to one without FBS. The other type of culture, the low glial cell density culture, was grown in dishes, which in addition to collagen had been precoated with Matrigel (Becton Dickinson Labware, Bedford, MA). The culture medium used consisted of Neurobasal medium with addition of B27 supplement (both obtained from Life Technologies), 2 mM L-glutamine, 15 µg/ml gentamicin, and nerve growth factor. The cultures were used for experiments after 2–3 and 9–12 days in vitro.
Preparation of effector cells
Effector cells were obtained from spleen cells suspended in RPMI
1640 medium (Life Technologies). Erythrocytes were depleted and the
spleen cells were resuspended in the RPMI 1640 medium supplemented with
5% FBS, 1 mM sodium pyruvate (Life Technologies), 1x nonessential
amino acids (Life Technologies), 1 mM L-glutamine, 2
x 10-7 M 2-ME (Merck, Darmstadt, Germany), and
penicillin/streptomycin. The cell suspensions were loaded onto
equilibrated nylon wool (Polyscience, Eppleheim, Germany) columns and
incubated at 37°C for 1 h. Nonadherent cells were eluted from
the column using 
-MEM, 10 mM HEPES (both from Life Technologies),
10% FBS, 2 x 10-7 M 2-ME, 1 mM
L-glutamine, and penicillin/streptomycin. These cells were
seeded in 25-cm2 culture flasks (Corning Costar),
25–30x106 cells/flask. Finally, rIL-2 (800
U/ml; PeproTech EC, London, U.K.) was added and the cells were
incubated at 37°C/10% CO2 for 4 days. The
cells were then harvested and used as effector cells. Resting NK cells
were prepared at the day of exposure, as described above.
Sorting of effector cells
Effector cells were resuspended in PBS and incubated for 30 min
on ice with a mAb directed to NK1.1 (PK136) conjugated with PE
(PharMingen, San Diego, CA). After washing in PBS,
NK1.1+ and NK1.1- cells
were sorted using a FACS Vantage cell sorter (Becton Dickinson,
Mountain View, CA). The sorted cells were collected in tubes containing
FBS, then pooled and resuspended in medium (-MEM, FBS, HEPES, 2-ME,
L-glutamine, and penicillin/streptomycin). Finally, rIL-2
(800 U/ml) was added, and the cells were incubated overnight at
37°C/10% CO2.
Exposure of DRG target cells to effector cells
Effector cells were added to target cell culture medium at different effector cell concentrations. One-half the volume of the medium in the target cell cultures was removed and replaced by effector cell suspension or medium (control). These cultures were incubated at 37°C for 2–24 h. Experiments were also performed in which the effector cells were separated from the target cells through a Transwell membrane (Corning Costar).
Immunocytochemistry
For analysis of nerve cell morphology after effector cell exposure, DRG cultures were fixed in 4% Formalin in PBS for 10 min at room temperature, followed by incubation with 5% BSA and 0.3% Triton X-100 in PBS for 30 min at room temperature. The cultures were then incubated overnight at 4°C with a primary mAb to neuron-specific class III ß-tubulin (TUJ1; Berkeley Antibody, Richmond, CA) diluted 1/250. To visualize the primary Ab, cells were incubated with donkey anti-mouse IgG conjugated to cyanine (1/200; Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 h at room temperature and then mounted in glycerol. Both the primary and secondary Abs were diluted in 2% BSA/0.3% Triton X-100 in PBS. Quantification of the NK cell-mediated cytotoxicity was performed under coded numbers. The number of nerve cells with fragmented axons was counted and compared with the total number of counted nerve cells. Ten randomly selected regions in each TUJ1-immunolabeled culture were analyzed. In each experiment, four dishes for each effector cell concentration were used and the experiments were repeated twice. Double labeling was performed using a polyclonal Ab raised against S100 (Santa Cruz Biotechnology, Santa Cruz, CA), a marker for Schwann cells, and the TUJ1 Ab. The cultures were fixed in 1% Formalin at 4°C and incubated in 5% donkey serum in PBS, followed by the S100 Ab (diluted 1/50 in 1.5% donkey serum in PBS, incubated 1 h at room temperature) and the TUJ1 Ab. The S100 Ab was detected with an Alexa488 donkey anti-goat Ab (1/100; Molecular Probes, Leiden, The Netherlands) and the TUJ1 with a Cy3 donkey anti-mouse Ab (1/500; Jackson ImmunoResearch Laboratories). To examine whether or not apoptosis in the neurons occurred, the cultures were stained with propidium iodide (4 µg/ml; Sigma), in PBS supplemented with RNase (1 µg/ml; Boehringer Mannheim, Mannheim, Germany) for 30 min at room temperature, after immunolabeling with TUJ1.
Statistical analysis
Statistical analysis was performed with Student t test using MINITAB for Windows, release 11 (Minitab, State College, PA). Values are given as percentage of fragmented neurons of the total number counted neurons or percentage of apoptotic neurons.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the first set of experiments, DRG cultures at day 9 in vitro
were exposed to syngenic B6-derived IL-2-activated NK cells. In these
cultures, the neurons were surrounded by a layer of small,
spindle-shaped cells with the morphological characteristics of Schwann
cells (13) (Fig. 1
A), which constitute the
glial cells in the peripheral nervous system. The culture also
contained large, flattened, fibroblast-like cells. Exposure to
IL-2-activated NK cells revealed no obvious effects on either the glial
cells or the neurons, as observed in the phase-contrast microscope.
Furthermore, no structural changes of the neurons or the Schwann cells
were observed, as assessed after immunolabeling with the TUJ1 mAb and
the S100 pAb, respectively (Fig. 2).
These cultures were also analyzed for signs of apoptosis, i.e., the
presence of apoptotic bodies. No sign of apoptosis was observed in
these cultures (Fig. 3). Exposure time
was extended in some experiments to 8 or 24 h, but no effects were
discernible even after this increased length of exposure to NK cells.
Cultures grown under similar conditions for 2–3 days were also exposed
to IL-2-activated NK cells. These DRG neurons appeared in clusters with
large amounts of surrounding glial cells. Similar to the older
cultures, there were no obvious effects on the neurons after this
treatment. Thus, it appears that IL-2-activated NK cells have no overt
effects on DRG neurons cultured in the presence of large amounts of
Schwann cells. To elucidate whether the expression of MHC class I
molecules on Schwann cells was associated with the protection of
neuronal cell death, high glial cell density DRG cultures generated
from ß2m-/- mice,
lacking expression of MHC class I molecules, were exposed to
IL-2-activated NK cells. No overt effects on the DRG neurons were
detected, showing that absence of MHC class I molecules does not
enhance the susceptibility of the neurons to IL-2-activated NK cells.
This experiment also excludes the possibility that presence of glial
cells would enhance levels of MHC class I molecules on the neurons to a
level that would per se be associated with resistance to NK cell
lysis.
|
|
|
Because the neurons in the initial experiments were surrounded by
glial cells, we next set out to evaluate whether IL-2-activated NK
cells could have any effects on DRG neurons with minimal admixing
numbers of Schwann cells. For this purpose, we established low glial
cell density cultures, in which most DRG neurons occurred as single
cells with no surrounding Schwann cells. These neurons extended long
axons already after 2 days in vitro (Fig. 1
B). Only a minor
portion of the neurons appeared in clusters with admixing Schwann
cells. When exposing the DRG cultures to syngenic B6-derived
IL-2-activated NK cells (2 x 106/ml), a
dramatic effect was observed with a total loss of neurons occurring
within 4 h. Immunolabeling of the neurons with TUJ1 showed severe
fragmentation of the axonal processes already after 2 h, and after
4 h only scattered remnants of axons and cell bodies were observed
(Fig. 4, B and C).
The proportion of degenerated neurons increased with the number of
effector cells (Fig. 5). Signs of
apoptosis, i.e., the presence of apoptotic bodies, were examined after
4 h of exposure to IL-2-activated NK cells (Fig. 4D).
The proportion of apoptotic neurons increased with increasing numbers
of IL-2-activated NK cells (Fig. 3). Studies with FACS-sorted
NK1.1+ IL-2-activated spleen cells as well as
IL-2-activated NK cells isolated from recombinase-activating
gene-1-/- mice, which lack B and T cells,
confirmed that the effects were mediated by NK cells. These cells also
caused a dramatic neuronal degeneration similar to the effect after
exposure to IL-2-activated bulk NK cells from B6 mice. In contrast,
resting NK cells had no cytotoxic effects on the DRG neurons (Fig. 4A), even when exposure times were extended to 16 h.
The low glial cell density cultures were grown under serum-free
conditions, while the high glial cell density cultures were grown in
serum-containing medium. To exclude the possibility that the neurons in
the low glial cell density cultures were susceptible to an NK
cell-mediated attack only because they were cultured in the absence of
serum, 10% FBS was added to these cultures 12 h before exposure
to IL-2-activated NK cells. Presence of serum did not affect the
susceptibility to NK cell-mediated nerve cell destruction. In the
serum-treated cultures, 24.9% ± 1.8% of the neurons showed signs of
apoptosis compared with 27.4% ± 1.9% in the cultures grown in
serum-free medium.
|
|
The neurodegenerative effects are dependent upon direct cell-cell contact and require perforin
To examine whether the neurodegenerative effects were mediated by
soluble factors released from the IL-2-activated NK cells or if the
killing was cell contact dependent, NK cell supernatants were added to
the low glial cell density DRG cultures. Exposure of the neurons to
supernatants for 4 h had no degenerative effects. We also exposed
the low glial cell density DRG cultures to IL-2-activated NK cells
separated by a Transwell membrane. This system allows soluble factors
from the NK cells to affect target cells, but prevents cell contact
between these cells and the effector cells. Exposure of the neurons to
IL-2-activated NK cells in this system had no effects on the DRG
neurons either as assessed by analyzing neuronal structure or the
proportion of apoptotic neurons. None of the neurons showed signs of
apoptosis after exposure to 5 x 105
effector cells/ml separated by a Transwell compared with 27.4% ±
1.9% of the neurons after exposure to the same concentration of
effector cells without Transwell separation. Taken together, these
results indicate that the neuronal destruction primarily is the result
of a direct cell-mediated process. Perforin has been shown to be a key
effector molecule for NK cell-mediated cytolysis (10). To
further analyze the mechanism by which NK cells mediate the nerve cell
destruction, low glial cell density DRG cultures were exposed to
IL-2-activated NK cells obtained from
perforin-/- mice. In contrast to NK cells from
wild-type mice, NK cells from perforin-/- mice
had no effect on the DRG neurons, demonstrating that the killing of the
neurons is perforin dependent (Fig. 6).
Control experiments revealed that the IL-2-activated NK cells from
perforin-/- mice had a blastlike morphology and
expressed the activation markers CD69 (93.8% of the B6 and 92.9% of
the perforin-/--derived
NK1.1+ cells were positive for CD69) and CD25
(30.4% of the B6 and 29.6% of the
perforin-/--derived
NK1.1+ cells were positive for CD25). This
ensured that there was no general difference in activation in the NK
cells from perforin-/- or wild-type mice. Thus,
nerve cell killing was dependent on the expression of perforin by the
NK cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DRG neurons cultured with a surrounding confluent layer of glial cells showed a decreased vulnerability to NK cell-mediated cytotoxicity. This is interesting in relation to an in vivo situation in which NK cell-sensitive MHC class I-negative lymphoma cells inoculated intracerebrally were shown to escape from NK cell-mediated rejection responses (14). It was suggested that a mechanism might have evolved to prevent rejection of MHC class I-deficient cells in the brain, either by suppression of migration of effector cells into the brain or suppression of their activity within the nervous tissue (14). The present study, demonstrating that neurons with surrounding glial cells can escape NK cell-mediated killing, favors the latter suggestion. Whether this is due to a steric hindrance mediated by the glial cells, factors secreted by the glial cells affecting the maturation and resistance of the neurons (15), or direct NK cell-inhibiting factors produced by the surrounding cells remains to be elucidated. The observed role of the surrounding cells in the prevention of NK cell-mediated DRG killing is compatible with the view that immunosuppressive molecules may be resident in Schwann cells and other endoneurial cells (reviewed in Ref. 16). This would be particularly important for DRG, since these lack an efficient barrier between the blood and the nervous tissue (17).
The NK cell-dependent death of sympathetic neurons resident in the superior cervical ganglia of rats, observed after the exposure to the drug guanethidine (2, 3), is the first in vivo disorder of the nervous system in which NK cells appears to be the principle effector cell. The pathogenic mechanism observed appeared to represent a novel type of autoimmune reaction: an exogenously/chemically induced alteration in a specific subset of cells that was suggested to target them for direct NK cell-mediated killing. This notion is strengthened by the present observation of direct NK cell-mediated killing of nerve cells in vitro, an observation that to our knowledge has not been demonstrated previously. The latter observations are of significant interest in relation to the model above. In that model, it could not be determined whether the NK cell-mediated destruction was due to direct neural cytotoxicity, or if the selective cell lysis was dependent exclusively on indirect effects. The present data suggest that, at least in an in vitro system, the former mode is possible. Our results taken together with the previous results are consistent with a model, which suggests a direct cytotoxic role for NK cells in the killing of a neuroanatomically restricted set of neurons. It cannot be excluded that this notion has a bearing on human diseases of the nervous system. However, although a number of observations suggest that immunological mediated mechanisms are involved in a number of neurodegenerative disorders, it should be noted that there is at present no evidence for an NK cell-mediated neural destruction in any of these conditions. In neuroimmunological disorders such as multiple sclerosis and experimental allergic encephelomyelitis, NK cells have been suggested to play an important regulatory role (see e.g., Refs. 18, 19, 20). The paradigm for such diseases in the peripheral nervous system is Guillian-Barré syndrome, which is characterized by demyelination, accompanied by mononuclear infiltrates. Degeneration of DRG axons projecting into the spinal cord posterior columns has been described in up to half of the deceased Guillian-Barré syndrome patients (21). Axonal degeneration also occurs during more severe forms of experimental allergic neuritis, which is the experimental counterpart to the disease (22). It would be interesting to find out whether NK cells are a component of the inflammatory cell infiltration during these conditions (23) and if a disturbed axon-glia cell relation, induced by the disease process, may expose the axons to an NK cell-mediated attack, as presently observed.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Eva Backström, Department of Neuroscience, Nobels väg 12 A, Karolinska Institutet, S-171 77 Stockholm, Sweden.
3 Abbreviations used in this paper: DRG, dorsal root ganglion; ß2m, ß2-microglobulin.
Received for publication November 16, 1999. Accepted for publication August 2, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cells in acute autoimmune encephalomyelitis. Eur. J. Immunol. 28:1681.[Medline]
This article has been cited by other articles:
|
|
 |
|
  Y. Kannan-Hayashi, K. Okamura, S. Hattori, M. Kuwamura, E. Higuchi, H. Terayama, M. Moriyama, M. Mukamoto, M. Okada, Y. Ohsugi, et al. Neuritogenic Effects of T Cell-Derived IL-3 on Mouse Splenic Sympathetic Neurons In Vivo J. Immunol., March 15, 2008; 180(6): 4227 - 4234. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Gee, A. Kalil, C. Shea, and K. J. Becker Lymphocytes: Potential Mediators of Postischemic Injury and Neuroprotection Stroke, February 1, 2007; 38(2): 783 - 788. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Kitaichi, S. Kotake, T. Morohashi, K. Onoe, S. Ohno, and A. W. Taylor Diminution of experimental autoimmune uveoretinitis (EAU) in mice depleted of NK cells J. Leukoc. Biol., December 1, 2002; 72(6): 1117 - 1121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Porter, S. Schatzberg, S. McDonough, D. Mertens, and A. de Lahunta Ganglioradiculitis (Sensory Neuronopathy) in a Dog: Clinical, Morphologic, and Immunohistochemical Findings Vet. Pathol., September 1, 2002; 39(5): 598 - 602. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
