Microglia ­ astrocyte interaction in Alzheimer's disease:
friends or foes for the nervous system?

ROMMY VON BERNHARDI AND GIGLIOLA RAMIREZ

Faculty of Medicine, Universidad de los Andes, San Carlos de Apoquindo 2200, Santiago, Chile

Corresponding author: Rommy von Bernhardi M., MD, Ph.D. Faculty of Medicine Universidad de los Andes, San Carlos de Apoquindo 2200, Las Condes, Santiago, Chile . Phone: (56 2) 214-1258. Fax: (56 2) 214-1752. E-mail: rvonb@uandes.cl

Received: March 15, 2001. Accepted: July 10, 2001.

ABSTRACT

Brain glial cells secrete several molecules that can modulate the survival of neurons after various types of damage to the CNS. Activated microglia and astrocytes closely associate to amyloid plaques in Alzheimer Disease (AD). They could have a role in the neurotoxicity observed in AD because of the inflammatory reaction they generate. There is controversy regarding the individual part played by the different glial cells, and the interrelationships between them. Both astrocytes and microglia produce several cytokines involved in the inflammatory reaction. Moreover, the same cytokines may have different effects, depending on their concentration and the type of cells in the vicinity. In turn, the events occurring in response to injury may lead to changes in the nature and relative concentration of the various factors involved. To learn about these putative glial interrelationships, we examined some effects of astrocytes on microglial activation.

Key terms: Apoptosis, ß-amyloid, cell activation, inflammatory process, neurotoxicity, senile plaques.

One of the major points in the work of Dr. Jaime Alvarez (Alvarez, 2001) is that the maintenance of neuronal structure and function is a dynamic process, reflecting the influence of other cells. In concordance to Dr. Alvarez's proposition I can say that this is not only true for neurons, but for glial cells as well. Mounting evidence suggeststhat astrocytes and microglia react in a concerted manner exerting mutual regulatory activity over each other. The reciprocal regulation of microglia and astrocytes may be central for the understanding of the ethiopathogenesis of Alzheimer's disease (AD).

When Alois Alzheimer described anatomopathological lesions in 1906, he used Bielschowsky's silver stain, which revealed only neuronal and not glial changes. Development of immunohistochemistry gave a new perspective of the architecture of the lesions. It is now known that in AD brains, microglia are abundant in senile plaques (Giulian et al., 1995; Sasaki et al., 1997) which are surrounded also by astrocytes (Itagaki et al., 1989). Microglia are highly reactive to environmental changes and respond to several types of CNS injury (Barron, 1995). They sense threats to the integrity of the CNS and may respond by releasing cytotoxic and inflammatory modulators (Giulian 1987; Dickson et al., 1993), phagocytosis and cell killing (Banati et al., 1993). Astrocytes and microglia function cooperatively generating the response of the brain to an injury (Martin et al., 1994). Astrocyte reaction may lead to the production of growth factors that could further promote microglial growth and activation (Lee et al., 1994) or could modulate its cytotoxic activity.

Astrocytes and microglial cells are interesting for understanding AD because they are highly reactive to environmental changes. Glial cells generate many molecules associated with inflammatory and immune functions. They secrete complement proteins and their regulators, inflammatory cytokines, acute phase reactants and many proteases and protease inhibitors (Table I). Thus glial cells may facilitate and amplify immune mechanisms locally within the CNS. Many of those proteins appear in association to neurodegenerative disease lesions. A strong inflammatory response may be autotoxic to neurons, exacerbating the fundamental failure underlying the neurological disorder. Reactive oxygen intermediates produced by microglia are particularly relevant. Production of oxygen free radicals can be initiated by several stimuli, including Aß. Activated microglia can generate enormous quantities of superoxide anions. While oxygen radicals are very effective against abnormal cells, normal host cells are unable to avoid being damaged. Because of that, microglia acting through these mechanisms can contribute substantially to tissue damage as a pathogenic factor of Alzheimer's disease.

However, the inflammatory response at the site of injury also represents a source of numerous growth factors and cytokines with trophic, mitogenic, chemotactic and angiogenic activities (Table I). These factors can affect cell migration, expression of cell adhesion molecules by nerve resident cells, and other phenomena necessary for tissue remodeling and axonal growth (Benveniste, 1992; Logan and Berry, 1993). The ability of astrocytes to produce neurotrophic factors is one aspect that makes glial cells attractive as therapeutic tools. Astrocytes produce their own repertoire of neurotrophic factors (Table I), including nerve growth factor (NGF), brain-derived growth factor (BDGF), and glial-derived growth factor (GDNF). Moreover, the same cytokines or growth factors may have different effects on different cells as well on the same cells, depending on their concentration, the type of cells in the vicinity, and the presence of other factors that can have suppressive, facilitator, or synergistic effects. The events occurring in response to injury may lead to changes in the nature and relative concentration of the various factors involved.

Microglia are the major source of cytokines such as IL-1ß, TGF-ß and TNF-a, which possibly are mediators of reactive gliosis. There is also evidence for induction of the same cytokines in astrocytes. Glial cells produce IL-1 and TNF-a in vitro in response to a variety of stimuli, including lipopolysaccharide (LPS), interferon g, and calcium ionophores (Lotan and Schwartz, 1994). In the mature animal, TGF-ß1 mRNA is not normally expressed in significant amounts in the nervous system. Within hours of injury, TGF-b1 is expressed locally by compromised neurons and glia and distally at multiple sites (Logan and Berry, 1993). Moreover, astrocytes produce cytokines that can act on microglia and macrophages, such as colony-stimulating factors (CSFs), thus creating a paracrine and autocrine feedback loop whereby microglia-derived factors and astroglial-derived factors regulate each other (Lotan and Schwartz, 1994). Multicolony-stimulating factor (M-CSF/IL-3) and GM-CSF injected in vivo stimulated the appearance of large numbers of microglia (Giulian and Ingemann, 1988). Astrocytes can therefore serve as a stimulus for microglial cell migration and activation after injury.

Microglia and astrocytes may coordinate their own and each other's migration and function through the release of IL-1 and TNF-a, which in turn induces IL-6, CSFs, and TGF-ß. By acting as an antagonist to some actions of IL-1 and TNF-a, TGF-ß serves as a negative feedback to limit the inflammatory reaction. Astroglial secretion of granulocyte and granulocyte-macrophage colony-stimulating factors (G- and GM-CSF) may induce mature granulocytes and macrophages to migrate into inflammatory foci within the CNS. Transforming growth factor (TGF) ß1, also produced by activated glial cells, is a multifunctional cytokine that has profound regulatory effects on numerous immune and inflammatory responses. Because of these cell interactions, mixed cultures provide a better model for studying glial activation in vitro. However, not only glial cells have a regulatory function upon each other, neurons apparently provide inhibitory factors that keep astrocytes in the normal resting state in vivo. There is increasing evidence that neurons in vitro can inhibit production of many proteins associated with reactive gliosis.

Giulian's group proposes that inflammatory cytokines and other microglia-derived factors account for neurotoxicity in gliosis while reactive-astrocytes products tend to be neuroprotective (Giulian and Corpuz, 1993; Giulian et al., 1993; 1994). Activated microglia produce factors that are detrimental to neurons in culture, whereas astrocytes promote neuronal growth and are able to attenuate microglia derived neurotoxicity. The production of neurotrophic factors and the elimination of neurotoxins by astrocytes fit well with this protective effect. Consistent with their benign activity, astrocytes, which also show some phagocytic activity, lack the potential of microglia to generate large quantities of free radicals. However, activation of microglia and astrocytes are closely linked. Microglia-derived factors secreted in response to injury can trigger reactive astrocytosis in such a way that microglia could be the real target of some stimuli that end up activating astrocytes. The complexity of the interaction between astrocytes and microglia makes it very difficult to sustain this distinction of activities.

Microglial activation is a gradual response, and it is important to keep this in mind when analyzing the possible effects of microglia in Alzheimer's disease. "Resting" or surveillance microglia sense threats to the integrity of the CNS and may respond by proliferation and the novo expression of molecules including complement receptors and MHC antigen. Activated microglia may protect against the pathological effects of noxious stimuli by performing limited action, supposedly without causing bystander damage. Phagocytosis and cell killing seem to be functions reserved to microglia-derived brain macrophages that are cytotoxic.

Microglia comprise the first barrier surrounding plaques. However, regardless of their enormous phagocytic activity, they appear to be unable to remove the amyloid fibers. Considering the scavenger function of microglial cells, we evaluated their capability to phagocytose and degrade exogenous APP and Aß constructs (von Bernhardi et al., 2001). We observed that although the degradation of Aß was slower than that of APP, microglia were able to phagocytose and degrade both APP and Aß conjugated beads. Phagocytosis was limited to microglial cells. However, there was no difference in the phagocytosis based on whether or not microglia were accompanied by astrocytes (data not shown). Thus microglial cells appear to have an enormous capacity to phagocyte APP and Aß constructs, so a failure in such a process leading to the accumulation of Aß in microglial cells might trigger an inflammatory response. We suspect that this mechanism may be involved in the pathogenesis of AD.

McGeer and McGeer (1995) proposed that neuronal damage in AD was not caused by the accumulation of Aß, but rather by the inflammatory response to it. We observed that conditioned media obtained from microglia treated with 2µM Aß increased (150%) the apoptosis of hippocampal cells in culture. In contrast, conditioned media from mixed glial cultures exposed to Aß did not induce apoptosis (Ramírez and von Bernhardi, 2000; von Bernhardi et al., in preparation). That observation suggested that astrocytes might modulate microglial activation and led us to evaluate the way in which astrocytes might affect microglial cell activation by Aß.

We analyzed changes in cell morphology, inducible Nitric Oxide Synthase (iNOS) expression and in the reductive metabolism as markers of cell activation for microglia, astrocytes and mixed glial (microglia and astrocytes, 1:1) cultures. Primary glial cultures were prepared from brains of newborn rats according to the modified procedure of Giulian and Baker (1986). The cell type of the culture was evaluated by double labeling with identity markers, lectin Griffonnia simplicifolia for microglia and GFAP for astrocytes. Microglial cultures were more than 99% pure and astrocyte cultures had 95% astrocytes. Control microglia (not exposed to Aß) were polymorph, with elongated shapes and iNOS negative (Fig. 1A). Microglia exposed to Aß showed the characteristic rounded amoeboid (activated) morphology and the expression of iNOS was up regulated (Fig. 1B). Exposure of mixed glial cultures to Aß also induced the expression of iNOS. However, rather than becoming round shaped and amoeboid, as in pure microglial cultures, the microglia frequently remained ramified or elongated in mixed glial cultures (Fig. 1C).

Figure 1: iNOS expression is up regulated in microglia exposed to Aß. Anti-iNOS labeling (FITC conjugated secondary antibody) of microglia (A, B) and mixed glial (C) cultures after 24 h in culture. (A) corresponds to a control culture and (B, C) are cells exposed to 2µM Aß. iNOS expression was similarly induced when microglia and mixed glial cell cultures were exposed to Aß. However changes of microglial morphology were more discrete when cultures also contained astrocytes. Scale bar = 50 µm.

To assess the reduction activity of the cells, we used the modified 3-[4,5-dimethylthiazol-2-yl]-2,5-dipheniltetrazolium bromide (MTT) assay to evaluate the effect of Aß (Fig. 2). Microglial, astrocyte and mixed glial cultures incubated with 10 µM Aß fibrils had a decreased ability to reduce MTT (p <0.01). However, microglial cells showed a decrease in MTT reduction at 2 µM Aß (41% decrease respect control value), while astrocyte cultures did not. Proinflammatory agents did not decrease MTT reduction activity on any of the culture composition tested (Fig. 2A). Moreover, Aß inhibited MTT reduction only when it was either conjugated to non-phagocytosable (60 µm) beads or aggregated as fibers. The fact that Aß did not inhibit MTT reduction when it was conjugated to phagocytosable (2.8 µm) beads (Fig. 2B), suggests that its presence as a non-degradable material determined the functional impairment of glial cells.

Figure 2: Effect of Aß and pro-inflammatory agents on the reduction activity of glial cells. The MTT assay was performed after the glial cell cultures were treated with the conditions indicated in the graphic, for 20-24 h. (Control) stands for cells cultured under standard conditions. Microglia present more pronounced changes than astrocyte or mixed glial cell cultures when exposed to Aß. The MTT reduction values are expressed as percentage of the control cells. Values correspond to the mean + SEM of 4-12 independent experiments performed in quadruplicate. The significance of the differences in MTT reduction was obtained by an ANOVA with Dunn's (Bonferroni) correction. Differences were assumed to be significant for p values smaller than 0.05.

The potential pathogenic role of astroglial impairments is well recognized in metabolic encephalopathies and immune-mediated diseases. A number of pathogenic processes could impair astrocytes. Among the most important we should mention are i.) persistent infection with perturbation of specialized functions, ii.) direct attack by immune cells, and iii.) exposure to soluble factors toxic for astrocytes derived from other host cells. We found that astrocytes had a modulator effect on the activation of microglial cells by Aß. Our results support the idea that astroglial dysfunction or the imbalance between astrocytic and microglial activities could participate in the neurodegeneration characteristic of AD.

ACKNOWLEDGEMENTS

R.v.B. wants to thank her friend and mentor, Dr. Jaime Alvarez, for his permanent support and advice. This work was supported by grants 1971132 and 1010146 from FONDECYT and MED 006-2000 from the Universidad de los Andes to R.v.B. We thank Dr. Heinz Döbeli from Hoffmann-La Roche, Basel, Switzerland for the Aß and APP constructs and useful discussion through the development of our work.

REFERENCES

ALVAREZ J (2001) The autonomous axon: a model based on local synthesis of proteins. Biol Res 34: 103-109

BANATI RB, GEHRMANN J, SCHUBERT P, KREUTZBERG GW (1993) Cytotoxicity of microglia. Glia 7: 111-118

BARRON KD (1995) The microglial cell. A historical review. J Neurol Sci 134: 57-68

BENVENISTE E (1992) Inflammatory cytokines within the central nervous system: sources, function, and mechanism of action. Am J Physiol 263: 1-16

DICKSON DW, LEE SC, MATTIACE LA, YEN SH, BROSNAN C (1993) Microglia and cytokines in neurological disease, with special reference to AIDS and Alzheimer's disease. Glia 7: 75-83

GIULIAN D, BAKER TJ (1986) Characterization of amoeboid microglia isolated from developing mammalian brain. J Neurosci 6: 2163-2178

GIULIAN D (1987) Amoeboid microglia as effectors of inflammation in the central nervous system. J Neurosci Res 18: 155-171

GIULIAN D, CORPUZ M (1993) Microglial secretion products and their impact on the nervous system. Adv Neurol 59: 315-320

GIULIAN D, INGEMANN JE (1988) Colony-stimulating factors as promoters of ameboid microglia. J Neurosci 8: 4707-4717

GIULIAN D, VACA K, CORPUZ M (1993) Brain glia release factors with opposing actions upon neuronal survival. J Neurosci 13: 29-37

GIULIAN D, LI J, LI X, GEORGE J, RUTECKI PA (1994) The impact of microglia-derived cytokines upon gliosis in the CNS. Dev Neurosci 16: 128-136

GIULIAN D, HAVERKAMP LJ, LI J, KARSHIN WL, YU J, TOM D, LI X KIRKPATRICK JB (1995) Senile plaques stimulate microglia to release a neurotoxin found in Alzheimer brain. Neurochem Int 27: 119-137

ITAGAKI S, MCGEER PL, AKIYAMA H, ZHU S, SELKOE D (1989) Relationship of microglia and astrocytes to amyloid deposits of Alzheimer disease. J Neuroimmunol 24: 173-182

LEE SC, LIU W, BROSNAN CF, DICKSON DW (1994) GM-CSF promotes proliferation of human fetal and adult microglia in primary cultures. Glia 12: 309-18

LOGAN A, BERRY M (1993) Transforming growth factor b1 and basic fibroblast growth factor in the injured CNS. TIPS 14: 337-343

LOTAN M, SCHWARTZ M (1994) Cross talk between the immune system and the nervous system in response to injury: implications for regeneration. FASEB J 8: 1026-1033

MARTIN LJ, PARDO CA, CORK LC, PRICE DL (1994) Synaptic pathology and glial responses to neuronal injury precede the formation of senile plaques and amyloid deposits in the aging cerebral cortex. Am J Pathol 145: 1358-1381

MCGEER PL, MCGEER EG (1995) The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Res Brain Res Rev 21: 195-218

RAMIREZ G, VON BERNHARDI R (2000) Alzheimer's disease: Neuroprotection mediated by proinflammatory molecules. Biol Res 33: R-73

SASAKI A, YAMAGUCHI H, OGAWA A, SUGIHARA S, NAKASATO Y (1997) Microglial activation in early stages of amyloid beta protein deposition. Acta Neuropathol (Berl) 94: 316-322

VON BERNHARDI R, RAMIREZ G, MATILE H, DÖBELI H (2001) Immobilised APP constructs: a tool for the in vitro screening of glial cells reactivity. E J Neurosci (in press)