Carbachol increases intracellular free calcium in cultured rat microglia

Carbachol increases intracellular free calcium in cultured rat microglia

59 Brain Research, 621 (1993) 59-64 Elsevl'er Science Publishers B.V. BRES 19148 Carbachol increases intracellular free calcium in cultured rat mic...

549KB Sizes 2 Downloads 62 Views


Brain Research, 621 (1993) 59-64 Elsevl'er Science Publishers B.V.

BRES 19148

Carbachol increases intracellular free calcium in cultured rat microglia E d w a r d R. W h i t t e m o r e , A n d r e w R. K o r o t z e r , A m i r E t e b a r i a n d C a r l W. C o t m a n Iruine Research Unit in Brain Aging, University of California, Irvine, CA 92717 (USA) (Accepted 23 March 1993)

Key words: Cholinergic; Muscarinic; Norepinephrine; Fura-2; Cell culture; Alzheimer's disease

Microglia are resident macrophages in the CNS and have been shown to exhibit immune system responses common to other macrophages, including phagocytosis, secretion of superoxide anions, and secretion of regulatory and trophic factors such as interleukin-1. Phagocytosis and oxidative burst by macrophages are often reported to be preceded by an increase in cytosolic free calcium. In addition, a variety of compounds, including neuroactive peptides, have been shown to elicit such calcium responses in various macrophage preparations. The results presented demonstrate that cultured rat microglia respond to exposure to carbachol with an increase in intracellular free calcium which is atropine-sensitive and the result of the release of calcium from intracellular stores. Norepinephrine also induced increases in free calcium, whereas the metabotropic glutamate agonist 1S,3R-ACPD, serotonin, adenosine and ATP did not. These results suggest that microglia can respond to select neurotransmitters, and that there may exist a signaling loop between neurons and microglia. Furthermore, since cholinergic fibers have been shown to infiltrate neuritic plaques in Alzheimer's disease (AD) and microglia have been reported to be activated in plaques, these results suggest that interactions between select neurotransmitters and microglia may play a key role in neurodegenerative diseases.


Microglia are resident macrophages in the CNS, and may serve as the principle scavenger cells of the brain z4. As such, microglia have been shown to exhibit immune system responses common to other macrophages, including phagocytosiszl, secretion of superoxide anions (oxidative burst) z, and secretion of regulatory and trophic factors such as interleukin-11'6. Microglia have also been reported to be activated under various conditions that involve neuronal loss, such as denervated hippocampus 5, and in senile plaques of Alzheimer's disease brain 3°. Microglia may therefore play a key role in the repair response to neurodegenerative diseases. In contrast, activated microglia may confound repair systems under certain circumstances, leading to improper degradation of functional systems, or deposition of unmanageable debris. Accordingly, it is important to define the events regulating microglial activity. Phagocytosis and oxidative burst by macrophages are often reported to be preceded by an increase in cytosolic free calcium4,8,17,z3,26. In addition, a variety of compounds, including neuroactive peptides, have been shown to elicit such calcium responses in various

macrophage preparations 4'17. Thus, in this paper we have investigated the possibility that microglia may respond to exposure to neurotransmitters with an increase in intracellular free calcium. MATERIALS AND METHODS

Cell culture Mixed rat glial cultures were prepared from the cerebral cortex of 1-3-day-old rat pups as previously described 13. Briefly, the pups were rapidly decapitated and the whole brain removed and submerged into a calcium/magnesium free Hanks buffer [CMF: NaHCO 3 (4.2 mM), pyruvate (1 mM), HEPES (20 mM), BSA (3 mg/ml), in Hanks Balanced Salt Solution (Gibco)]. The meninges were then carefully removed, and extra care was taken to remove all visible blood vessels. The cortex was bluntly dissociated from the rest of the brain, and transferred to another dish of CMF (room temperature). The cortices were then sequentially transferred through 3 dishes of CMF to remove excess debris, and the tissue was triturated in a 15 ml tube with a 10 ml plastic pipette. This suspension was centrifuged (200 × g for 3 min), brought up in DMEM (Gibco) plus 15% fetal calf serum (FCS), and triturated with small-bore siliconized glass pipettes until the suspension contained no visible particles. This suspension was divided evenly into 75 ml flasks at a density of 2 cortices/flask. After 24 h, the medium was changed to DMEM plus 10% FCS, and this primary culture was fed with this medium 2 times per week. Purified cultures of microglia were then prepared as described by Giulian and Baker 6, but with several modifications. After 7-10 days, microglia were harvested following an overnight (16-24 h) shaking at 240 rpm to dislodge microglia and

Correspondence: C.W. Cotman, Irvine Research Unit in Brain Aging, University of California, Irvine, CA 92717, USA. Fax: (1) (714) 725-2071.

60 oligodendrocytes from the bed of astrocytes. The supernatant from this shaking step was transferred to another 75 ml flask, and microglia were allowed to set,tie for 30-60 rain. Non-adherent cells were removed with the supernatant, and this flask was washed 2 x in CMF. A solution of trypsin (0.25% trypsin in CMF) was added for 2 min, and the cells were mechanically detached using a cell scraper. The trypsin reaction was terminated by the addition of an equal volume of D M E M plus 10% FCS. The suspension was transferred to a 15 ml tube, centrifuged (200 X g for 3 min), and the pellet re-suspended in D M E M plus 10% FCS. Following a trituration step to dissociate the pellet, the cells were plated onto poly-L-lysine-coated glass coverslips (18 m m dia, No.l, A s s i s t e n t / F i s h e r special order; cleaned overnight with 3 M N a O H , rinsed and then sterilized), at a density of 1 × 104 cells/18 m m coverslip in 0.3 ml in 12-well plastic plates. After 1 h, the volume of each well was brought up to 1 ml with D M E M plus 10% FCS. Experiments were performed on cells 1-3 days after plating onto glass coverslips. The purity of microglial cultures was determined by staining sister plates with antibody to leukocyte common antigen (LCA). Glial Fibliary Acidic Protein (GFAP) staining was used as a negative control. Using these methods, our cultures were > 90% pure microglia. Occasionally, astrocytes and oligodendrocytes were observed; during imaging experiments, these cells could easily be avoided by morphological criteria.

Process-bearing or 'ramified' microglia could be easily distinguished from ameboid microglia by visual inspection. For the experiments reported here, process-bearing cells were defined as cells with distinct processes at least two cell body lengths long. [Note: Although we plated microglia onto glass coverslips that had been coated with poly-L-lysine, ameboid microglia, which appear to be loosly adherent, often washed away during the course of an experiment in our flow through chamber. Thus, the majority of our results are from process-bearing microglia, or strongly adherent ameboid cells.]

Fura-2 Imaging Individual plates of microglia were loaded with Fura-2 by removal of 0.5 ml of m e d i u m from an individual well, and adding this to a small volume of a F u r a - 2 / A M plus pluronic acid mixture, and this was added back to the original well such that the final concentration of F u r a - 2 / A M was 4 - 1 0 /xM and pluronic acid was 0.02%. After 30 min, this well was rinsed with 6 volumes of D M E M to remove excess F u r a - 2 / A M , and returned to the incubator for 30 min to allow for the de-esterification of the dye. This plate was then transferred to a small Leiden-type chamber (volume = 0.3 ml), and placed onto the stage of a Nikon Diaphot microscope equipped with a 4 0 × Fluor (N.A. 1.3) fluorescence objective. This chamber was continuously perfused at 1 - 2 m l / m i n with saline buffer at room


200 [- [ 5rain / "175 I 150




I 5 rain I Cell



[--o-- B,pol=,loid.~]

I t25



150 125


~* 100

75 1





25 0


(100 ~M) Carbach01 100 p.M






I $ mln I

200 , I 5 m i n I

I'''ax---- Cell#l J 300.

175 150 125




75 100

50 25 Carbachol



[100 btMl

o mini Carbachol



(100 IJ,M)

(100 p.M)

(100 ~tM)

Fig. 1. Carbachol responses in cultured microglia. Figures represent approximate intracellular calcium concentration vs time during exposures to carbachol and other compounds at the times noted. Individual cells were outlined on the digital image, and the area within the outline was sampled digitally yielding a m e a s u r e of individual cell calcium using the calibrated values noted under Methods. Ratiometric images were taken at 10 s e c / i m a g e during drug exposures and at 60-600 s / i m a g e during wash periods to avoid excessive U V exposure. A. A 4 min exposure to carbachol (100 mM) induced rapid increases in calcium in both bipolar or process-bearing microglia, and in ameboid-type microglia. B. Longer exposures (15 min) to carbachol induced long-lasting calcium increases, which recovered quickly following washout. C. Co-application of atropine blocked responses to carbachol. D. Exposure to carbachol in calcium-free saline resulted in calcium responses which were slightly reduced compared to responses in normal saline.

61 Carbachol

temperature. The saline buffer contained: NaCI 137 mM, KCI 5.3 mM, CaCI 2 3 mM, MgC12 1 mM, H E P E S 10 mM, and D-Glucose 25 mM, at p H 7.4. Calcium-free experiments utilized the same buffer but CaCl 2 was omitted and 20 /zM E G T A was added. Drugs were diluted into saline buffer, or into calcium-free saline buffer, and added to the perfusion system at the times indicated. Care was taken to place the perfusion input tube very near the cells u n d e r investigation, to avoid slow increases in drug concentrations, which may cause desensitization. Fluorescent images were obtained by a H a m a m a t s u C-2400 SIT video camera, and these images were digitized by a 386-clone computer and Image-1 (Universal Imaging Corp.) hardware and software. Approximate intracellular calcium concentrations were calculated from the ratio of emission evoked by 340 and 380 n m light 7, using the equation: [Ca 2+ ]i =

Kd[(R- Rmin)/(Rmax



(0.1 raM)

200 175 150 [Ca++l (nM) 125 100 75 50 25 0









Time (minutes)


In most of the experiments presented here, the values used were: Rmin = 0.347; Rmax = 8.29; Fmin = 45.48; Fmax = 7.75; K d = 224 nM. The compounds carbamyl choline (carbachol), norepinephrine, serotonin, adenosine 5'-triphosphate (ATP) were purchased from Sigma (St. Louis, MO). 1S,3R-l-aminocyclopentane-l,3-dicarboxylic acid (1S,3R-ACPD) was purchased from Tocris Neuramin (Bristol, UK). Fetal bovine serum (FCS) and D M E M were purchased from Gibco (Gaithersburg, MD). All other compounds were also purchased from commercial sources.

RESULTS As shown in Fig. 1A, 100 ~M carbachol induced increases in intracellular free calcium in cultured microglia. Responses to carbachol were observed at test concentrations from 1 /xM to i mM. Responses were typically rapid in onset, and recovered quickly following the removal of carbachol from the bath. Successive exposures to carbachol often failed to produce a second response, even if the bath was washed with carbachol-free saline for up to 30 min (not shown). Free intracellular calcium remained elevated throughout long duration exposures to carbachol, and recovered quickly following removal of carbachol (Fig. 1B). A larger percentage of process-bearing microglia responded to carbachol than did ameboid-type microglia, and responses in process-bearing cells were generally larger than responses in ameboid-type cells (see Fig. 1A). In 10 experiments from 10 glass coverslips of microglia, we observed 41 process-bearing microglia, and 20 ameboid-type cells; of these 75% of process-bearing cells and 40% of ameboid-type cells responded to 100 /~M carbachol (P < 0.01, Student's 2-tailed t-test). Carbachol responses in process-bearing microglia appeared to occur first in the cell processes, and later in the cell body region (see Fig. 2). In addition, the magnitude of the calcium increase was generally greater in the processes than in the cell body. Calcium responses to carbachol were blocked during co-application of carbachol with the muscarinic antagonist atropine (Fig. 1C). In 5 experiments from 5 glass coverslips, only one out of a total of 30 microglia



(0.1 raM)


25O 225 200 [Ca++l 175 (riM) 150 125 100 75 50



25 i




Time (minute~)

200 f 175


(0.1 raM)

150 [Ca++] [ (nM) 125 [






Y 0l

, 230









Time (minutes)

Fig. 2. Increases in intracellular calcium evoked by carbachol appeared in cell processes before they appeared in cell body regions. Discrete areas of individual process-bearing microglia representing processes or cell body regions were sampled and analyzed separately during exposures to carbachol. As shown, responses occured in the processes before the cell body, and increases were generally of greater magnitude in processes compared to cell body regions. Calcium concentrations were determined using the calibration values outlined in the Methods section.

displayed substantial increases in intraceltular calcium during co-application of atropine (10 IzM) and carbachol (100 /xM). Atropine alone did not induce increases in free calcium.

62 Carbachol also induced increases in intracellular calcium in calcium free medium. In 4 experiments from 4 glass coverslips, 23 of a total of 31 microglia (74%) responded to 100 /zM carbachol in calcium-free medium. These responses were often reduced in amplitude compared to responses in normal calcium-containing saline (Fig. 1D). We also investigated whether microglia responded to a variety of other neuroactive compounds, to test the specificity of this microglial calcium response. At a test concentration of 1 raM, microglia also responded to the adrenergic agonist norepinephrine (Fig. 3A). In 4 experiments from 4 glass coverslips, 15 of a total of 23 microglia (65%) responded to 1 mM norepinephrine. Responses to norepinephrine were frequently longlasting, and in many cases failed to recover despite extensive washing. We did not further characterize the pharmacological specificity of this norepinephrine response. Microglia did not show a detectable response to 1 mM concentrations of the metabotropic glutamate agonist lS,3R-l-aminocyclopentane-l,3-dicarboxylic acid (1S,3R-ACPD) (Fig. 3B; 4 coverslips, 0 / 3 0 cells responded), serotonin (Fig. 3C; 2 coverslips, 0 / 1 0 cells responded), and adenosine (not shown; 2 coverslips, 0 / 1 5 cells responded). Responses were also not observed to 5 mM adenosine-5'-triphosphate (ATP) (not shown; 3 coverslips, 0 / 1 5 cells responded).




5 min

I f e l l #1


I00 5O Cell #'2.

Norepinephrine (1 mM )



Bsmln I •



~25 at +


50 25


(1 mM)





I100 p.Ml

(1.00 gMl

I 5 mi. ]




-==o-=- Cell#I


The results presented demonstrate that cultured microglia respond to exposure to carbachol with an increase in intracellular free calcium. This calcium response is atropine-sensitive and appears to be the result, at least in part, of release of calcium from intracellular stores. These results suggest that microglia in culture express functional muscarinic receptors, which may be coupled to the inositiol phosphate second messenger system as reported for muscarinic responses in a variety of other systems 15. Microglia also responded to norepinephrine, but not to serotonin, the metabotropic glutamate agonist 1S,3R-ACPD, adenosine or ATP. Thus, ligand-induced calcium responses in microglia appear to be selective for cholinergic and adrenergic systems, among the group of compounds assayed here. Interestingly, ATP has recently been reported to induce depolarizing currents in cultured microglia using electrophysiological techniques 11. However, removal of extracellular calcium did not change these depolarizing currents 29 in support of our finding that ATP does not induce a direct change in intracellular free calcium.







I 0

[ I Serotonin [1 mMl

Carbachol [100 gMl

Fig. 3. Microglia also responded to norepinephrine, but not to 1S,3R-ACPD nor serotonin. Microglia were exposed to test compounds at the concentration of 1 mM for 4 min. Responses in individual cells were assayed as noted in the Fig. 1 legend, and approximate calcium concentrations were determined using the calibration values outlined in the Methods section.

The data presented also demonstrate that muscarinic responses in cultured microglia are related to cell morphology, in that process-bearing cells were more likely to respond to carbachol than ameboid-type cells. Since it has been suggested that microglial morphology in vivo and in vitro reflects the activation state of the cell 6, expression of transmitter receptors in microglia may be related to activation state. This ex-

63 tends previous findings from our laboratory demonstrating a relationship between expression of voltagedependent currents and microglial morphology 12. These results suggest that microglia can respond to select neurotransmitters, and that there may exist a signaling loop between neurons and microglia. Although our data do not directly address the nature of such a possible loop, theoretically microglia could respond to signals from neurons either actively or in response to injury or tissue damage. To our knowledge, active signaling between microglia and neurons has never been demonstrated directly. However, microglial processes have been reported to be associated with synapses anatomically 19, and it has also been suggested that microglia may respond to 'volume neurotransmission '28, in which transmitters may exist and act extrasynaptically. Thus, direct signals between neurons and microglia may exist, but the nature of such a hypothetical role remains uncertain. In contrast, the signal between neurons and microglia may involve the response to tissue damage and subsequent repair and cleanup by microglia. As noted in the introduction, microglia are closely related to macrophages, and as resident macrophages in the CNS are thought to perform many functions common to other macrophages. These include phagocytosis 21, generation of superoxide anions 2 and secretion of regulatory and trophic factors 1,6. In macrophages, these activities have been shown to be preceded by an increase in intracellular free calcium 4'8'17'23'26. Thus, one possibility is that carbachol and norepinephrine, which we have shown cause an increase in intracellular free calcium in microglia, may serve as a signal for microglia to initiate one or more of these functions, as a response to neuron damage. For example, damaged neurons may release these transmitters, initiating phagocytosis of damaged tissue by microglia. Such a r e p a i r / c l e a n - u p signal is consistent with previous data showing that microglia are activated in denervated hippocampus 5 and are present in the senile plaques in Alzheimer disease brain 3°, conditions in which damaged neurons may release transmitter. Furthermore, microglia have been reported to be involved in synaptic remodeling in the neurohypophysis 19. The failure of serotonin and the glutamate agonist 1S,3R-ACPD to induce calcium responses argues that this is not a generalized repair response to damage of any neuron. The connection between the calcium response and a phagocytic initiation signal is further strengthened by the demonstration above that carbachol-induced increases in intracellular calcium appeared initially in microglial processes. This supports the idea that process-bearing or 'resting' microglia may utilize a calcium

signal prior to morphological transformation from process-bearing, resting microglia to ameboid, activated microglia. Such morphological conversion has been demonstrated for cultured microglia in response to phorbol esters and lipopolysaccharide, following 72 h exposure to these compounds 27. The data presented do not address this issue, but experiments are currently underway to investigate morphological effects of long exposures to carbachol, to test this transformation hypothesis. Additionally, it is known that neuritic plaques in AD are also associated with regenerating neuritic processes 2°. Specifically, cholinergic fibers have been shown to exhibit reactive synaptogenesis in the hippocampus in AD, and cholinergic fibers are known to infiltrate neuritic plaques 2°'25. In addition, a variety of reports have shown that reactive microglia are associated with plaques m,14. Thus, another possiblility is that these abnormal cholinergic inputs in plaques result in the activation of microglia via muscarinic receptor activation and the increases in intracellular calcium reported here. Activated microglia may then begin aft inappropriate or abnormal repair response, which would serve to damage healthy tissue. Possible examples include the superoxide-mediated damage of healthy neurons, improper processing of /3-amyloid peptides 21, or the superoxide-induced increase in assembly of/3-amyloid into aggregates 3 within plaques. In contrast, signaling between neurons and microglia in disease states may take a more active form. For example, microglia have been shown to release regulatory and trophic factors ~'6, and signals from neurons in the form of neurotransmitters may initiate such activities 9'22. In addition, it has recently been shown that a kidney cell line releases /3-amyloid precursor derivatives following muscarinic receptor activation J6. Since microglia and cholinergic systems are associated with plaques in Alzheimer brains 1°A4'2°'25, it may be that release of acetylcholine in AD brain induces excessive release of/3-amyloid precursor derivatives and concomitant activation of microglia as suggested by the data presented here. Subsequent improper processing of /3-amyloid may then lead to deposition of aggregated /3-amyloid peptides. Since /3-amyloid peptides have been shown to be toxic to neurons in culture 18'2°, this connection has special relevance to understanding the etiology of Alzheimer's Disease. REFERENCES 1 Araujo, D.M. and Cotman, C.W.,/3-Amyloidstimulates glial cells in vitro to produce growth factors that accumulate in senile plaques in Alzheimer's Disease, Brain Res., 569 (1992) 141-145. 2 Colton, C.A. and Gilbert, D.L., Production of superoxide anions










11 12 13


15 16

by a CNS macrophage, the microglia, FEBS Lett., 223 (1987) 284-288. Dryks, T., Dryks, E., Hartman, T., Masters, C. and Beyreuther, K., Amyloidogenicity of /3A4-bearing amyloid protein precursor fragments by metal-catalyzed oxidation, J. Biol. Chem., 267 (1992) 18210-18217. Drysdale, B., Yapundich, R.A., Shin, M.L. and Shin, H.S., Lipopolysaccharide-mediated macrophage activation: the role of calcium in the generation of tumoricidal activity, 3. Immunol., 138 (1987) 951-956. Fagan, A.M. and Gage, F.H., Cholinergic sprouting in the hippocampus: a proposed role for IL-1, Exp. Neurol., 110 (1990) 105-120. Giulian, D. and Baker, T.J., Characterization of ameboid microglia isolated from developing mammalian brain, J. Neurosci., 6 (1986) 2163-2178. Grynkiewicz, G., Poenie, M. and Tsien, R.Y., A new generation of Ca 2+ indicators with greatly improved fluorescence properties, J. Biol. Chem., 260 (1985) 3440-3450. Holian, A. and Daniele, R.P., The role of calcium in the initiation of superoxide release from alveolar macrophages, J. Cell. Physiol., 113 (1982) 87-93. Hu, X., Goldmuntz, E.A. and Brosnan, C.F., The effect of norepinephrine in endotoxin-mediated macrophage activation, J. Neuroimmunol., 31 (1991) 35-42. Itagaki, S., McGeer, P.L., Akiyama, H., Yhu, S. and Selkoe, D., Relationship of microglia and astrocytes to amyloid deposits of Alzheimer's disease, J. Neuroimmunol., 24 (1989) 173-182. Kettenmann, H., Banati, R. and Walz, W., Electrophysiological behavior of microglia, Glia, 7 (1993) 93-101. Korotzer, A.R. and Cotman, C.W., Voltage-gated currents expressed by rat microglia in culture, Glia, 6 (1992) 81-88. McCarthy, K.D. and De Vellis, J., Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol., 85 (1980) 890-902. McGeer, P.L., Itagaki, S., Tago, H. and McGeer, E.G., Reactive microglia in patients with senile dementia of the Alzheimer type are positive for the histocompatability glycoprotein HLA-DR, Neurosci. Lett., 79 (1987) 195-200. Nathanson, N.M., Molecular properties of the muscarinic acetylcholine receptor, Annu. Reu. Neurosci., 10 (1987) 195-263. Nitsch, R.M., Slack, B.E., Wurtman, R.J. and Growdon, J.H., Release of Alzheimer Amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors, Science, 258 (1992) 304-307.

17 Ohmori, Y. and Hamilton, T.A., Ca 2+ and calmodulin selectively regulate lipopolysaccharide-inducible cytokine mRNA expression in murine peritoneal macrophages, Z Immunol., 148 (1992) 538545. 18 Pike, C.J., Cummings, C.J. and Cotman, C.W., /3-Amyloid induces neuritic dystrophy in vitro: similarities with Alzheimer pathology, NeuroReport, 3 (1992) 769-772. 19 Pow, D.V., Perry, V.H., Morris, J.F. and Gordon, S., Microglia in the neurohypophysis associate with and endocytose terminal portions of neurosecretory neurons, Neuroscience, 33 (1989) 567-578. 20 Selkoe, D.J, The molecular pathology of Alzheimer's disease, Neuron, 6 (1991) 487-498. 21 Shigematsu, K., McGeer, P.L., Walker, D.G., Ishii, T. and McGeer, E.G., Reactive microglia/macrophages phagocytose amyloid precursor protein produced by neurons following neural damage, J. Neurosci. Res., 31 (1992) 443-453. 22 Spengler, R.N., Allen, R.M., Remick, D.G., Streiter, R.M. and Kunkel, S.L., Stimulation of a-adrenergic receptor augments the production of macrophage-derived tumor necrosis factor, J. Immunol., 145 (1990) 1430-1434. 23 Stickle, D.F., Daniele, R.P. and Holian, A., Cytosolic calcium, calcium fluxes, and regulation of alveolar macrophage superoxide anion production, J. Cell. PhysioL, 121 (1984) 458-466. 24 Streit, W.J., Graeber, M.B. and Kreutzberg, G.W., Functional plasticity of microglia: a review, Glia, 1 (1988) 301-307. 25 Strubble, R.G., Cofrk, L.C., Whitehouse, P.J. and Price, D.L., Cholinergic innervation of neuritic plaques, Science, 216 (1982) 413-415. 26 Sullivan, R., Fredette, J.P., Griffin, J.D., Leavitt, J.L., Simons, E.R. and Melnick, D.A., An elevation in the concentration of free cytosolic calcium is sufficient to activate the oxidative burst of granulocytes primed with recombinant human granulocytemacrophage colony-stimulating factor, J. Biol. Chem., 264 (1989) 6302-6309. 27 Suzumura, A., Marunouchi, T. and Yamamoto, H., Morphological transformation of microglia in vitro, Brain Res., 545 (1991) 301-306. 28 Thomas, W.E., Brain macrophages: evaluation of microglia and their functions, Brain Res. Rev., 17 (1992) 61-74. 29 Walz, W., Banati, R. and Kettenmann, H., Extracellular ATP opens cation channels in cultured microglial cells, Soc. Neurosci. Abstr., 18 (1992) 633. 30 Wisniewski, H.M. and Terry, R.D., Reexamination of the pathogenesis of the senile plaque, Prog. Neuropathol., II (1973) 1-26.