| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online June 18, 2004 as doi:10.1096/fj.03-1199fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,,1
* Istituto di Neuroscienze CNR, Sez. Psicobiologia e Psicofarmacologia;
Laboratorio di Neurochimica, Centro di Neurobiologia Sperimentale, "Mondino-Tor Vergata-S. Lucià";
Fondazione Santa Lucia, IRCCS; and
Dipartimento di Biologia, Università di Roma "Tor Vergata", Rome, Italy
1Correspondence: Dipartimento di Biologia, Università di Roma "Tor Vergata", Via della Ricerca Scientifica, 00133 Rome, Italy. E-mail: carri{at}Bio.uniroma2.it
SPECIFIC AIM
The contribution of different cellular types to the pathogenesis of amyotrophic lateral sclerosis (ALS) is unknown; this prompted us to investigate the functional relationship between glial and neural cells in generating inflammatory response and apoptotic events in ALS. We devised a cellular model comprising human glioblastoma cells stably expressing ALS-associated mutant G93A-SOD1 (G93A-glia) cocultured with human neuroblastoma cells stably expressing mutant G93A-SOD1 (G93A-neuro).
PRINCIPAL FINDINGS
1. G93A-glia cells induce activation of caspase-1, release of cytokines, and activation of apoptotic pathways in cocultured G93A-neuro cells
SOD1 activity in the transfected lines G93A-glia and G93A-neuro is about twice that of the corresponding parental lines (U373 glioblastoma and SH-SY5Y neuroblastoma). Expression of G93A-SOD1 induces a marked increase of total intracellular reactive oxygen species (ROS) in glioblastoma and neuroblastoma cell lines and causes activation of glial cells, as demonstrated by increased expression of glial fibrillary acidic protein (GFAP) in Western blot experiments. In coculture experimental conditions, the two cell types are kept physically separated but can exchange molecular signals, and it is possible to perform biochemical determination of diffusible factor(s) and assign functional roles to them. When G93A-neuro (2 independent monoclonal cell lines, b and c) is cocultured with G93A-glia (2 independent monoclonal cell lines, a and b) for 10 days, a significant induction of caspase-3 (Fig. 1
a) and nuclear fragmentation (Fig. 1c
) takes place in G93A-neuroblastoma. Caspase-3 activation is observed when neuroblastoma cell lines expressing other fALS-SOD1s (G37R-, G85R-, and I113T-neuro) are cocultured with either the corresponding glioblastoma line (G37R-, G85R-, and I113T-glia) or G93Aa-glia (Fig. 1b
).
|
In this kind of experiment, caspase-1 is activated selectively in G93A-neuro cocultured with G93A-glia (as observed in other fALS-SOD1 cocultures) and mature IL-1ß is produced and secreted by G93A-neuro in the medium. Intriguingly, specific inhibition of caspase-1 inhibits activation of caspase-3 in G93A-neuro in coculture with G93A-glia. These data suggest that molecular signals are released in the medium by fALS-SOD1-glia, which trigger caspase-1 activation and the apoptotic pathway in fALS-SOD1-neuro.
2. Coculture with G93A-neuro causes activation of inflammatory processes in G93A-glia
When G93A-glia cells are cocultured with G93A-neuro, the intracellular level of IFN-
is dramatically increased in G93A-glia cells; release of IFN- in the medium is increased as well. Activation of G93A-glia in cocultures with G93A-neuro is indicated by the increase in expression of GFAP and cyclooxygenase2 (COX2) in Western blot experiments and by higher NF-
B activity (assayed in EMSA), specifically in G93A-glia.
3. Coculture with G93A-neuro causes activation of nNOS in G93A-glia
Induction of death pathways can be achieved in G93A-neuro cells by treatment with nitric oxide, as demonstrated by activation of caspase-3 and Hoechst staining of apoptotic nuclei. In our coculture system, expression of G93A-SOD1 induces an increase in NO production, with the most relevant effect in G93A-glioblastoma/G93A-neuroblastoma cocultures (Fig. 2
a). Neuronal NOS is undetectable in immunostaining experiments in all neuroblastoma-derived cell lines regardless of coculture conditions; however, coculture with G93A-neuroblastoma increases nNOS in G93A-glioblastoma (Fig. 2b
). Inhibition of NO synthase by L-NAME reduces released NO to nearly basal levels in all conditions (Fig. 2c
) and entirely reverts the proapoptotic effect exerted by G93A-glioblastoma on G93A-neuroblastoma cells (Fig. 2d
). Finally, inhibition of caspase-1 prevents release of NO in the medium (Fig. 2e
).
|
4. Nitric oxide, IFN-, or coculture with G93A-glia induce death of embryonic spinal motor neurones from G93A-SOD1 transgenic mice (G93A+)
As an additional control that NO and IFN- play a role as mediators of motor neuron death in G93A-SOD1-linked fALS, primary cultures from embryonic spinal cord of G93A transgenic mice were treated with proinflammatory drugs. Stimulation with IFN- and lipopolysaccharide (LPS) causes degeneration of motor neurons and an increase in NO release in G93A+ cultures. Treatment with L-NAME partially reverts death induction death, suggesting NO is among the inductors of neuronal loss in this system (Fig. 2f
). The selective death of G93A+ motor neurons is observed when G93A+ primary cultures are cocultured with G93A-glia without IFN-/LPS.
CONCLUSIONS
Several lines of evidence suggest that glial cells contribute to the pathogenesis of ALS. Reactive astrogliosis is commonly found in postmortem tissues from ALS patients, and markers of inflammation such as COX2 have been found in patients and experimental models. Microglia, resident CNS macrophages, proliferate in spinal cords of a mouse model of familial ALS during the neurodegenerative phase and microglia/macrophages are activated in ALS spinal cord tissue.
The functional relationship between glial cells and neural cells in generating the inflammatory response and apoptotic events leading to ALS is unclear. We devised a model of human origin where interactions occurring in cocultures of neuroblastoma cells and glioblastoma cells, both expressing G93A-SOD1 (or other fALS-SOD1s) at a low level, could be studied thoroughly. Our data demonstrate that this coculture condition, mimicking the situation in SOD1-linked fALS patients, triggers a cascade of events in both cell types (schematic representation, Fig. 3
). Expression of fALS-SOD1 induces activation of fALS-glioblastoma cells, which respond by increasing the level of intracellular inflammatory markers (e.g., COX2), inducing nNOS activity, and releasing NO and IFN- into the medium. Both released factors act on neuroblastoma cells expressing G93A-SOD1, inducing activation of caspase-1 and release of IL-1ß. This cytokine may diffuse to glioblastoma cells and induce NF-B-mediated up-regulation of COX2 and nNOS, thus contributing to the inflammatory response elicited by the expression of fALS-SOD1. Sequentially, NO-dependent caspase-3 activation and apoptotic death of neuroblastoma cells are triggered. A vicious cycle is generated, where the inflammatory reaction of glial cells induces neurones to enter the apoptotic pathway and release proinflammatory signals back to glia.
|
The existence of such a vicious cycle is confirmed by the observation that culture in a medium preconditioned by the presence of G93A-glioblastoma is not sufficient to induce apoptosis of G93A-neuroblastoma. Inhibition of caspase-1 abolishes caspase-3 activation, thus rescuing apoptotic death of G93A-neuroblastoma, but also prevents IL-1ß secretion and NO production from G93A-glioblastoma. These data clearly indicate that caspase-1 has a role as a pivotal factor linking inflammation to neurodegeneration in this system.
Although it is known that ROS stimulate the production of proinflammatory cytokines, those cytokines also stimulate ROS production. In our model, this kind of self-sustaining process could be triggered by the pro-oxidant activity of fALS-SOD1 acting as the primum movens, the trigger of noxious cross-talk between neurons and glia.
This molecular signaling between glia and neuronal cells is likely to occur in fALS-transgenic mice, since we have observed that NO and IFN- are mediators of cell death in primary cultures of embryonic spinal cords from G93A+ mice; nevertheless, we cannot exclude the possibility that other mediators of neuroinflammation are involved and that the relative contribution of those factors may be different in patients.
Markers of oxidative stress and neuroinflammation have been observed in sporadic ALS and in non-SOD1-linked familial ALS. Therefore, it is tempting to speculate that a molecular exchange between glia and neurones occurs in patients: upon imbalance of the interactions between the two cell types, due to increased oxidative stress or other mechanisms, a critical threshold is overcome and the pathological phenotype is triggered.
In conclusion, our data demonstrate the existence of a non-cell autonomous death of neurones induced by G93A-SOD1 and support the concept that any therapeutic trial for the treatment of ALS should be aimed at the simultaneous interception of pro-oxidant, proinflammatory, and proapoptotic signals.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.03-1199fje; doi: 10.1096/fj.03-1199fje
This article has been cited by other articles:
|
|
 |
|
  P. A. Jonsson, K. S. Graffmo, P. M. Andersen, T. Brannstrom, M. Lindberg, M. Oliveberg, and S. L. Marklund Disulphide-reduced superoxide dismutase-1 in CNS of transgenic amyotrophic lateral sclerosis models Brain, February 1, 2006; 129(2): 451 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Dobrowolny, C. Giacinti, L. Pelosi, C. Nicoletti, N. Winn, L. Barberi, M. Molinaro, N. Rosenthal, and A. Musaro Muscle expression of a local Igf-1 isoform protects motor neurons in an ALS mouse model J. Cell Biol., January 17, 2005; 168(2): 193 - 199. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
