The transition of microglia to a ramified phenotype is associated with the formation of stable acetylated and detyrosinated microtubules.

In situ and in vitro, microglia can have different morphologies, which are thought to reflect distinct physiological activities. Two extreme forms are ameboid and ramified microglia. To study cytoskeletal changes during differentiation, we used defined cell culture systems to yield cultures of ameboid or ramified microglia from mouse brain. With respect to proliferation, secretion, receptor-expression, and phagocytosis, ramified microglia was generally less active. We found that the transition to a ramified phenotype was accompanied by an increase in the relative amount of acetylated and detyrosinated tubulin. Whereas the modified microtubules were restricted to regions close to the microtubule-organizing centers (MTOCs) in ameboid cells, acetylated microtubules were abundant in ramified cells, where they appeared to traverse from one process to another with no apparent anchoring at MTOCs. The increase in acetylated and detyrosinated microtubules was paralleled by an increased stability against nocodazole-induced microtubule disassembly and by a lower rate of change in the length of the processes. Staining of retinal wholemounts confirmed the presence of acetylated microtubules in ramified microglia in situ. We conclude that during the transition of microglia to a ramified phenotype regulated processes exist, which cause a reorganization of microtubules and a change in composition of the microtubule skeleton resulting in a less dynamic and more stable microtubule network. Intracellular factors that are specifically involved in microtubule stabilization in ramified microglia need to be identified in future research and may provide a useful criterion for defining ramified microglia.

Complement factor C5a and epidermal growth factor trigger the activation of outward potassium currents in cultured murine microglia.

Microglia, the resident macrophages of the brain, are transformed from a quiescent into an activated phenotype in a number of pathological conditions. The signalling mechanisms which control such transformations are not yet understood. In the present study, we have characterized fast electrophysiological responses in cultured microglia, induced by two putative signalling substances, complement 5a (C5a), a chemotactic agent for macrophages and microglia, and epidermal growth factor, the receptor of which is up-regulated during pathological conditions in the brain. Both factors transiently activate an outwardly rectifying K+ conductance, while the membrane of the unstimulated microglial cell is dominated by an inwardly rectifying K+ conductance. The C5a-stimulated current developed within about 20s and decayed within a slightly slower time course. It was activated by depolarlizing voltage steps positive to the resting membrane potential of about -70 mV, and neither inactivated nor showed a delayed activation following voltage steps. The epidermal growth factor-stimulated current showed similar characteristics. When G-proteins were specifically blocked, the K+ conductance could no longer be activated by C5a or epidermal growth factor, suggesting that for both agonists an inhibitory G-protein is involved in the intracellular signalling cascade. We tested if the induction of the K+ conductance is causally linked to other C5a-induced cellular responses, like transient cytosolic Ca2+ elevation and mobility. The K+ conductance was not activated when a Ca2+ transient was induced by thapsigargin, nor did a blockade of the C5a-induced K+ conductance by K+ channel blockers affect the motility response. This implies that after activation of the C5a receptor and the G-protein, the K+ conductance activation, the Ca2+ mobilization and the motility response are governed by independent intracellular pathways, and that the K+ conductance increase must serve other functions than the control of motility.

Modulation of potassium currents in cultured murine microglial cells by receptor activation and intracellular pathways.

The electrophysiological properties of ameboid microglia from rodent brain are dominated by inwardly rectifying potassium channels and by the lack of outward currents. This channel pattern results in a distinct physiological behavior: depolarizing events, e.g. following adenosine triphosphate receptor activation, can lead to a long lasting membrane depolarization. Here we address the question whether this resting K+ channel activity can be modulated. Intracellular application of guanosine 5'-O-(3-thiotriphosphate) induced an outward current and led to a complete disappearance of the inward current inward rectifier potassium current as measured with the patch clamp technique. Moreover, an elevation in cytosolic calcium concentration (to 1.6 microM) via intracellular perfusion reversibly blocked the inward current. The inhibition of inward currents by guanosine 5'-O-(3-thiotriphosphate) could be enhanced by additional adenosine triphosphate receptor activation. Adenosine triphosphate or tumor necrosis factor receptor activation alone could lead to a transient partial block of the inward rectifier and to the transient appearance of a delayed outward current. We conclude that the activity of the microglia K+ channels and thus the physiological behavior of microglia can be modulated on a time scale of seconds by receptor activation and distinct intracellular pathways.

Extracellular ATP activates a cation conductance and a K+ conductance in cultured microglial cells from mouse brain.

Microglial cells have important functions during regenerative processes after brain injury. It is well established that they rapidly respond to damage to the brain tissue. Stages of activation are associated with changes of cellular properties such as proliferation rate or expression of surface antigens. Yet, nothing is known about signal substances leading to the rapid changes of membrane properties, which may be required to initiate the transition from one cell stage into another. From our present study, using the patch-clamp technique, we report that cultured microglial cells obtained from mouse or rat brain respond to extracellularly applied ATP with the activation of a cation conductance. Additionally, in the majority of cells an outwardly directed K+ conductance was activated with some delay. Since ADP, AMP, and adenosine (in descending order) were less potent or ineffective in inducing the cation conductance, the involvement of a P2 purinergic receptor is proposed. The receptor activation is accompanied by an increase of cytosolic Ca2+ as determined by a fura-2-based Ca(2+)-imaging system. This ATP receptor could enable microglial cells to respond to transmitter release from nerve endings with ATP as a transmitter or cotransmitter or to the death of cells with resulting leakage of ATP.

Membrane properties of ameboid microglial cells in the corpus callosum slice from early postnatal mice.

Microglial cells in culture are distinct from neurons, macroglial cells, and macrophages of tissues other than brain with respect to their membrane current pattern. To assess these cells in the intact tissue, we have applied the patch-clamp technique to study membrane currents in microglial cells from acute, whole brain slices of 6-9-d-old mice in an area of microglial cell invasion, the cingulum. As strategies to identify microglial cells prior to or after recording, we used binding and incorporation of Dil-acetylated low-density lipoproteins, binding of fluorescein isothiocyanate-coupled IgG via microglial Fc-receptors, and ultrastructural characterization. As observed previously for cultured microglial cells, depolarizing voltage steps activate only minute if any membrane currents, while hyperpolarizing voltage steps induced large inward currents. These currents exhibited properties of the inwardly rectifying K+ channel in that the reversal potential depended on the transmembrane K+ gradient, inactivation time constants decreased with hyperpolarization, and the current was blocked by tetraethylammonium (50 mM). This study represents the first attempt to assess microglial cells in situ using electrophysiological methods. It opens the possibility to address questions related to the function of microglial cells in the intact CNS.

Fragmentation of DNA in the retina of chicken embryos coincides with retinal ganglion cell death.

Neuronal cell death was studied in the developing retina of the chicken embryo. One of the most characteristic indices of the form of cell death termed apoptosis is regular, apparently internucleosomal fragmentation of DNA. When retinae of eight to seventeen day old chicken embryos were dissected out and the DNA from this tissue size fractionated on agarose gels, fragmentation typical of apoptosis was observed on day ten. The maximal amount of fragmentation was reached around day eleven and twelve and declined from day 15 to 17. These findings correlate in time with previous histological data on retinal cell death and demonstrate for the first time the occurrence of DNA fragmentation typical of apoptosis in the developing nervous system.

TNF and Plasmodium berghei ANKA-induced cerebral malaria.

The cerebral pathology observed in Plasmodium berghei ANKA-infected CBA mice has been attributed to overproduction of TNF, the mice in which this syndrome is seen being those with the highest serum TNF levels. To investigate this further, we injected recombinant human TNF into malaria-primed mice to see if we could reproduce the cerebral changes observed in P. berghei ANKA infections. A range of doses, administered as a single or repeated injections, or via osmotic pumps, failed to reproduce these changes, but did induce hypoglycaemia, midzonal liver necrosis and neutrophil adhesion in pulmonary vessels. This pathology is seen in terminal Plasmodium vinckei infections, but absent in terminal P. berghei ANKA. In addition, the permeability of the blood-brain barrier to Evan's blue, which is present in P. berghei ANKA but not in normal or P. vinckei-infected mice, was not induced by exogenous TNF. Serum levels of TNF were measured in an ELISA assay, and found to be consistently higher in P. vinckei rather than P. berghei ANKA terminal infections. This is consistent with the pathological changes we could reproduce by injecting TNF. For these reasons we suggest that the cerebral pathology seen in mice infected with P. berghei ANKA may be governed by TNF produced locally by monocytes sequestered within the cerebral blood vessels, not simply by systemic levels of this cytokine.

Retinal receptive fields under different adaptation levels studied with pattern-evoked ERG.

The pattern-evoked electroretinogram (PERG) was studied in three normal subjects under different levels of adaptation using pattern onset-offset stimulation. The occurrence of a response maximum (spatial selectivity) in the onset response at a certain spatial frequency can be interpreted as reflecting the activity of neurons having antagonistic center-surround receptive fields and allows an estimation of receptive-field center sizes. The present data suggest that in the light-adapted state, the majority of receptive fields, dominating the response can be estimated to have center diameters of 6-13 min of arc. With increasing dark adaptation, the spatial selectivity shifts from high to low spatial frequencies over 1.3-2.4 octaves, indicating that large receptive fields are active in the dark and small fields in the light. The physiological mechanisms of these changes are discussed on the basis of findings obtained from single-unit studies.

The spatial organization of retinal receptive fields in light and darkness as revealed by the pattern electroretinogram.

The spatial selectivity of the electroretinogram in response to pattern onset-offset stimuli was studied in man at several levels of adaptation ranging from scotopic to photopic levels. Under conditions of rod function the peak of the spatial selectivity based on amplitude measurements of the pattern-onset response occurs at a low spatial frequency. With increasing light adaptation a gradual shift of the selectivity to higher spatial frequencies occurs. This change in the character of the response can be explained by the assumption that antagonistic center-surround retinal receptive fields contribute to the response, which are larger under scotopic than under photopic levels of stimulation.