blue; Griffonia simplicifolia lectin; nucleoside diphosphatase ... isolated retinae were incubated with the Griffonia simplicifolia (GS) lectin or reacted for.
GLIA 19:91–103 (1997)
Early Activation of Microglia in the Pathogenesis of Fatal Murine Cerebral Malaria ISABELLE M. MEDANA,1 NICHOLAS H. HUNT,1 AND TAILOI CHAN-LING2* of Pathology University of Sydney, Sydney, New South Wales 2006, Australia 2Department of Anatomy and Histology, University of Sydney, Sydney, New South Wales 2006, Australia
1Department
KEY WORDS
blood-brain barrier; retina; immunopathology; monocyte; Monastral blue; Griffonia simplicifolia lectin; nucleoside diphosphatase
ABSTRACT Microglia are pluripotent members of the macrophage/monocyte lineage that can respond in several ways to pathological changes in the central nervous system. To determine their role in the pathogenesis of fatal murine cerebral malaria (FMCM) we have conducted a detailed study of the changes in morphology and distribution of retinal microglia during the progression of the disease. Adult CBA/T6 mice were inoculated with Plasmodium berghei ANKA. These mice died 7 days post inoculation (p.i.) with the parasite while exhibiting cerebral symptoms, increased permeability of the blood-brain barrier, and monocyte adherence to the vascular endothelium. Mice were injected i.v. with Monastral blue 2 h prior to sacrifice to identify ‘‘activated’’ monocytes, and their isolated retinae were incubated with the Griffonia simplicifolia (GS) lectin or reacted for the nucleoside diphosphatase enzyme to visualize microglia and the vasculature. Changes in microglial morphology were seen within 2–3 days p.i., that is, at least 3 days prior to the onset of cerebral symptoms and 4 days before death. Morphological changes included retraction of ramified processes, soma enlargement, an increasingly amoeboid appearance, and vacuolation. There was also increased staining intensity and redistribution of ‘‘activated’’ microglia toward retinal vessels, but no increase in density of NDPase-positive cells. The GS lectin only labeled a small population of microglia in the uninfected adult mouse retina. However, there was a striking increase in the focal density of GS-positive microglia during the progression of the disease. Extravasation of monocytes also was observed prior to the onset of cerebral symptoms. These results provide the first evidence that microglial activation is a critical component of the pathological process during FMCM. GLIA 19:91–103, 1997. r 1997 Wiley-Liss, Inc.
INTRODUCTION Cerebral malaria (CM) is a major life-threatening complication of Plasmodium falciparum infection in humans, responsible for at least 2 million deaths annually. The occurrence of CM is increasing, due to the difficulty of controlling malaria through insecticides and anti-malarial drugs. A greater understanding of the disease process is necessary if effective modes of treatment or prevention are to be found. The pathogenesis of human CM is still a matter of debate. Postmortem human studies have the disadvantage that they reveal only the outcome of the disease rather than r 1997 Wiley-Liss, Inc.
the important events occurring during its course. Experimental murine models of CM have provided an alternative approach to this problem since they are reproducible and allow experimental manipulations (Chan-Ling et al., 1992; Grau et al., 1990; Neill and Hunt, 1992; Thumwood et al., 1988). The most widely used murine model of CM is a ‘‘fatal’’ model developed
Received 10 May, 1996; accepted 1 September, 1996. *Correspondence to: Dr. Tailoi Chan-Ling, Department of Anatomy and Histology (F13), University of Sydney, Sydney, N.S.W. 2006, Australia. Contract grant sponsor: National Health and Medical Research Council of Australia and Sydney University Research Grant Scheme.
92
MEDANA ET AL.
by Rest (1982): CBA/T6 mice infected with Plasmodium berghei ANKA (CBA-PbA) die 7 days after parasite inoculation, exhibiting neurological involvement and histological changes similar in many respects to the human condition (Chan-Ling et al., 1992; Maegraith and Fletcher, 1971; Neill and Hunt, 1992; Rest, 1982; Thumwood et al., 1988; Toro and Roman, 1978), although with some differences. Fatal murine cerebral malaria (FMCM) is dependent on CD41 T cells (Finley et al., 1983; Grau et al., 1986) and is prevented by administration of anti-oxidants (Thumwood et al., 1989) or antibodies against certain cytokines (Grau et al., 1987, 1989a, 1990), suggesting that it is an immunopathological process. To date, it has not been established whether the pathogenesis of CM is due to immune reactions occurring in the periphery, which result in non-specific damage to the cerebral vasculature, or whether central nervous system (CNS)specific cells such as microglia also actively participate. It is well established that microglia are the most likely CNS-specific cells to be involved in a local immune response. Activation of microglia has been reported in numerous CNS pathologies including Alzheimer’s disease (Giulian et al., 1995; McGeer et al., 1994; Nieto and Mora, 1994), AIDS encephalopathy (Gelman, 1993; Vazeux, 1991), multiple sclerosis (Newcombe et al., 1994; Selmaj et al., 1991), and its animal model experimental autoimmune encephalomyelitis (Bauer et al., 1995; Renno et al., 1995). In this activated condition, microglia may assume a macrophage-like morphology characterized by a larger, irregularly shaped soma with pseudopodia, lacking the ramified processes of the resting form found in normal CNS tissue (Thomas, 1992; Vrabec, 1975). In the activated state microglia are thought to produce reactive oxygen species (Colton and Gilbert, 1987; Sonderer et al., 1987) and secrete, as well as respond to, several cytokines (Benveniste, 1992), which, as mentioned earlier, are essential to the development of FMCM. However, the production of cytokines by microglia in FMCM has not been reported as yet. For these reasons we wished to examine changes in the three-dimensional spatial relationship of microglia to other cellular elements, particularly the microvasculature, during FMCM. This cannot be properly assessed using brain sections, but we have circumvented this problem by using the retinal wholemount technique (Chan-Ling et al., 1992; Medana et al., 1996; Neill et al., 1993). The retina is functionally and structurally like the brain and when used as a wholemount preparation allows visualization of the entire vascular plexus, preserving the spatial relationship of the microcirculation with other tissue components. We have previously shown that the retina undergoes cellular changes during FMCM that reflect those in the brain (Chan-Ling et al., 1992; Medana et al., 1996; Neill et al., 1993). Taking advantage of the sensitivity of nucleoside diphosphatase (NDPase) and Griffonia simplicifolia (GS) lectin histochemistry and the retinal wholemount technique, we have detected striking morphological changes, upregulation of GS lectin binding sites, loss of
regular distribution, and migration of microglia towards retinal vessels in FMCM, providing, for the first time to our knowledge, direct evidence implicating these cells in the processes that lead to the neurological manifestations in this condition.
MATERIALS AND METHODS Fatal Murine Cerebral Malaria Model Inoculation procedure CBA/T6 mice (6–8 weeks old) weighing 20–25 g were obtained from the Blackburn Animal House, University of Sydney. Mice were given i.p. injections of 106 parasitized erythrocytes suspended in 200 µl of phosphatebuffered saline (PBS). Uninfected mice were used as controls. Twelve mice were sacrificed on days 2, 3, 4, 5, and 7 post-inoculation (p.i.) with the parasite and the retinae prepared for GS lectin histochemistry and NDPase histochemistry as detailed below. Prior to sacrifice, blood smears were taken and the parasitemias were determined. To ensure that the mouse/ parasite combinations followed the usual course of the disease, an inoculated group of mice from each series was allowed to progress through the course of infection until death, which invariably occurred between days 6 and 8 p.i.
Visualization of Microglia Griffonia simplicifolia lectin histochemistry Microglia were visualized using the isolectin B4 (GS lectin), which binds to a-D-galactose. This lectin has been shown to label microglia, cells of the monocyte/ macrophage lineage, vascular precursor cells, and vascular endothelium (Chan-Ling et al., 1990; Streit and Kreutzberg, 1987). Earlier studies have shown that macrophages and microglia when activated show increased staining intensity with the GS lectin (Htain et al., 1994; Maddox et al., 1982). After an immersion fixation time of 8 min in 1% (v/v) glutaraldehyde, the retinal wholemounts were placed in PBS 1 1% Triton X-100 (PBS 1 1% T) for 72 h at 4°C. They were then incubated with the GS lectin conjugated to biotin (1:200, Sigma, St. Louis, MO) for another 72 h at 4°C. The tissue was then washed for 30 min in PBS 1 1% T and then incubated in ExtrAvidin-HRP (1:100, diluted in PBS; Sigma E-2886) for 24 h at 4°C. The HRP was then visualized using diaminobenzidine tetrahydrochloride (DAB, 0.05% [w/v]) and H2O2 (0.2% [v/v]) for approximately 2–3 min in Nickel Tris-buffered saline (NiTBS) with constant observation under a dissecting microscope. To terminate development the tissue was transferred to NiTBS and washed for another 20 min. The preparations were then mounted in Aquamount (BDH, Poole, England) and examined under transmitted light illumination.
MICROGLIAL RESPONSE TO CEREBRAL MALARIA
Nucleoside diphosphatase histochemistry Microglia also were visualized using NDPase histochemistry. Previous investigators (Penfold et al., 1993; Sanyal and De Ruiter, 1985) have observed that retinal NDPase activity is localized as intense and selective staining in blood vessels and in cells resembling macrophages and microglia. In contrast, there is an absence of staining on astrocytes, Mu¨ller cells, neurons, and pigmented endothelium. This specific and stainable enzyme activity has been exploited as a histochemical marker. Penfold and colleagues (1993) have shown that NDPase-positive microglia in the human retina also express CD45 (leukocyte common antigen) and MHC class II. Briefly, retinae were immersion fixed in 4% (w/v) paraformaldehyde (in 0.1 M sodium cacodylate buffer, pH 7.4) with 8% (w/v) sucrose and 5% (v/v) dimethyl sulphoxide (DMSO) for 3 h at 4°C. Retinae were then washed several times overnight at 4°C in 0.1 M sodium cacodylate buffer (pH 7.2) with 8% (w/v) sucrose. The retinae were then placed in the incubation medium (0.2 M Tris-maleate buffer (pH 7.2), 1% (w/v) lead nitrate, 25 mM MgCl2, DMSO, and 10 mM inosine diphosphate), for 30 min at 37°C. Retinae were then rinsed in distilled water, immersed in 2% (v/v) aqueous ammonium sulphide for 2 min, rinsed again in distilled water, placed on slides, and mounted in glycerol.
Visualization of ‘‘activated’’ monocytes Mice were injected with Monastral blue 2 h prior to sacrifice to visualize monocytes (Neill and Hunt, 1992) and then killed by cervical dislocation. Their eyes were dissected free from the optic cup and then immersion fixed in either 4% (w/v) paraformaldehyde or 1% (v/v) glutaraldehyde; the retinae were then subjected to GS lectin or NDPase histochemistry to allow covisualization of microglia and monocytes.
Evaluation of Microglial Density Outlines of retinae were drawn at 103 magnification using a graticule (Olympus, Japan). The density of GS lectin- or NDPase-positive microglia within each 1 mm2 region was counted, at a magnification of 403. These values were then multiplied by a magnification correction factor (316) and converted into a grading scale. A representative retina was mapped for each selected time point. This technique does not demonstrate accurate changes in cell numbers within the retina during the progression of the disease, but rather illustrates whether the response of microglia occurs over the whole retina or in focal regions, as well as indicating the degree of change in their density.
93
Analysis of Microglial Distribution Photomicrographs of retinae were taken at 8003 total magnification using an Olympus Vanox microscope equipped with Normarski optics. Photographic montages were assembled, and drawings of the retinal microglia and vasculature were made onto transparencies.
Distance to Nearest Neighbor Analysis A rectangular area spanning the mid-peripheral to peripheral region between two radial incisions was defined. The area was positioned to exclude the cutting edge and to include a minimum of 60 microglia. This process was repeated on three retinae on each day p.i. Using the Magellan software of Halasz and Martin (1984), the positions of the microglial nuclei within the rectangle and in the surrounding border were digitized. The distance from each microglial nucleus within the rectangle to its nearest neighbor, whether within the rectangle or within the border region, was then determined by the computer. The mean of these nearest neighbor distances was then calculated for each time point. Student’s t-test was used to analyze differences in the mean nearest neighbor distances at the different time points.
RESULTS Retinal Microglia Have a Ramified Morphology in Uninfected CBA Mice NDPase histochemistry In the uninfected CBA mouse retina, NDPase-labeled microglia displayed a ramified morphology with numerous varicosities on their processes (Fig. 1A–C). Microglia were predominantly found in three layers, the ganglion cell layer (Fig. 1A), the inner plexiform layer (Fig. 1B), and the outer plexiform layer (Fig. 1C). This layering of microglia was also evident on retinal sections (Fig. 1D) but became less distinct at the terminal stage of FMCM (Fig. 1E). While retinal microglia are found predominantly in these three layers, the processes of individual microglia can ramify through a number of retinal layers (Fig. 1D,E). Using retinal wholemounts, in which the preparations are photographed en face, it is impossible to have all processes in focus simultaneously.
Griffonia simplicifolia lectin-B4 histochemistry In uninfected CBA mouse retinae, faintly labeled microglia were detected in small numbers in the ganglion cell layer only (Fig. 2A). These cells displayed a ramified morphology characteristic of resting microglia in mouse, rat, rabbit, human and cat retinae described
94
MEDANA ET AL.
Fig. 1.
MICROGLIAL RESPONSE TO CEREBRAL MALARIA
by earlier workers (Boya et al., 1991; Chan-Ling, 1994; Perry et al., 1985; Schnitzer, 1989; Penfold et al; 1993). In contrast to the NDPase-labeled resting microglia, varicosities were not prominent on their processes. While the microglia within the parenchyma never contacted neighboring microglia, perivascular microglia were frequently observed with their processes entwined with neighboring cells (seen investing the vessel running horizontally along the lower half of Fig. 2B). Early Morphological Changes in NDPaseand GS Lectin-Labeled Microglia During the Progression of FMCM From day 3 p.i. it was evident from NDPase staining that small sub-populations of microglia had retracted their ramified processes and increased their staining intensity relative to other microglia on the same preparation (compare Fig. 1A and F). Between days 4 and 5 p.i., small numbers of microglia had adopted a more amoeboid morphology with vacuoles evident along their processes (Fig. 1G). These changes were first detectable among microglia in close association with vessels, and they increased in severity and frequency until the terminal stage (day 7 p.i.) of the disease (Fig. 1H,I). These morphological changes in microglia appeared in focal regions, and microglia with ‘‘resting’’ morphology or microglia at an advanced stage of morphological change could be found on the same retina. At days 2–3 p.i. there was an increase in GS lectin staining intensity of microglia, so that the small distensions on their processes were more prominent (Fig. 2B). Between days 3 and 4 p.i., some microglia showed a decrease in process length and an increase in soma size; this phenomenon was found predominantly among the microglia in close association with the retinal blood vessels. From day 5 p.i., some microglia showed small vacuoles within their processes; this feature was usually found in those microglia with shorter processes (Fig. 2C). All these morphological changes, and the increased staining intensity with the GS lectin, pro-
Fig. 1. Retinal microglia visualized using nucleoside diphosphatase (NDPase) enzyme histochemistry. Microglia visualized with NDPase had a greater number of terminal arborizations and more distensions than microglia visualized with the Griffonia simplicifolia lectin. A–C: The same region of a wholemounted retina from an uninfected CBA/T6 mouse in the (A) ganglion cell, (B) inner plexiform, and (C) outer plexiform layers. D: That microglia were predominantly found in three main layers of the retina in control animals was also evident on retinal sections. E: In contrast, during the progression of fatal murine cerebral malaria (FMCM), microglia became dispersed throughout these layers F–I: Microglial changes in the ganglion cell layer of wholemounted retina during the progression of FMCM. F: From day 3 p.i. microglia displayed regional variations in staining intensity and a slight reduction in the length of processes. G: From day 4–5 p.i., focal regions of microglia changed from the resting, ramified morphology to a more amoeboid shape. H: At day 7 p.i. (the terminal stage of the disease), numerous amoeboid microglia with vacuoles were evident. I: Amoeboid microglia were found at the terminal stage of FMCM to concentrate in regions juxtaposed with major retinal vessels, predominantly veins.
95
gressed in severity and frequency until the death of the infected mice (Fig. 2D). Around days 6–7 p.i. there appeared to be four different morphological types of cells that were GS lectin positive. These were: 1) intensely staining, ramified microglia with prominent distensions (Fig. 2E) found in focal regions of the retina; 2) a population with stout processes, large soma, and extensive vacuolization (Fig. 2D,F); 3) a small number of round cells with diameters greater than 15 µm (Fig. 2D); and 4) a second, less frequent, population of round cells with a diameter of approx 10 µm (arrow in Fig. 2D). Although microglia in advanced stages of morphological change were found extensively at days 6–7 p.i., many microglial elements with normal morphology, or with less advanced changes, could still be found on the same retina.
Redistribution of Microglia Toward Retinal Vessels During FMCM In control animals, microglia were distributed regularly over the entire retina (Figs. 1A–C, 4A). Figure 4A shows the apparently regular distribution of NDPaselabeled microglia in the region of the major arteries and veins in control animals. Confirming previous reports, microglia in mice displayed non-contact behaviour in the retina (Hume et al., 1983) and cerebral cortex (Perry and Gordon, 1988). Figure 3 shows schematically the density distributions of microglia labeled with GS (Fig. 3B) and NDPase (Fig. 3A) during the progression of FMCM. Using NDPase to visualize the entire population of microglia during the progression of the disease, the average density of microglia over the entire retina was calculated for each observation period during the course of FMCM. In control animals, the density of microglia was 114 cells/mm2. From days 3–7 p.i., the average microglial density ranged between 99 and 118 cells/ mm2. In contrast to those results using NDPase visualization, microglial density determined using GS lectin showed a massively increased expression of a-Dgalactose residues in focal regions (Fig. 3B). Two striking features are evident from these density maps. First, GS lectin only labels a very small sub-population of microglia in control CBA mouse retinae. Second, there was an increase in density of GS1 microglia in focal regions of the retina during the progression of the disease process in FMCM, resulting in a tenfold increase in some regions by the terminal stage of the disease.
Distance to nearest neighbor Because NDPase visualizes the entire microglial population, we elected to conduct an analysis of the changes in distribution and the distance to the nearest neighbor in FMCM at various days p.i. At the terminal stage of the disease, more microglia seemed to be in close association with the major retinal vessels (with a
96
MEDANA ET AL.
Fig. 2. Microglial response during the progression of FMCM visualized with the Griffonia simplicifolia lectin. A: In uninfected CBA/T6 mice, retinal microglia were weakly labeled with the GS lectin and displayed long, fine processes typical of resting microglia. In contrast to NDPase histochemistry, the GS lectin only labeled a small number of ramified microglia in control animals. B: From day 3 p.i., more microglia were labeled with the GS lectin, their processes becoming thicker and distensions on their processes became more evident (arrow). C: From day 4 p.i., microglia in close proximity to retinal
vessels displayed retracted processes and enlarged somas in focal regions of the retina. D: At the terminal stage of the disease (day 7 p.i.), GS lectin-labeled amoeboid microglia, as well as a small number of monocyte-like cells (arrow), were evident. E: GS lectin-labeled microglia at the terminal stage of the disease showed a range of morphologies including a ramified shape with prominent distensions along the processes (arrows) F: A high magnification view of microglia with stout processes, large soma, and extensive vacuolization (arrows), at day 7 p.i.
more marked response surrounding veins than arteries), and regions in the parenchyma appeared sparsely populated with microglia (Fig. 4B). This contrasted with the distribution of microglia found in uninfected mouse retinae, where microglia were seemingly in a fairly regular array over the entire retina (Fig. 4A). To investigate this interpretation further, the mean distance of each microglia to its nearest neighboring microglia was determined at days 3, 4, and 7 p.i. At day 3 p.i. there was a significant decrease in the distance to
Fig. 3. Maps showing representative densities of Griffonia simplicifolia lectin-labeled/NDPase-positive microglia during the progression of FMCM. The number of labeled microglia was determined within a 1 mm2 region, and the cells/mm2 were converted to a gray scale. A: During the progression of FMCM, there was no change in the average density of microglia visualized using NDPase histochemistry. B: In contrast, in control CBA/T6 mice, GS lectin only labeled very low numbers of resting microglia. During the course of the disease, microglia displayed a significant upregulation of expression of a-D-galactose residues, visualized using G. simplicifolia lectin histochemistry. By day 3 p.i., there was a detectable increase in the numbers of GS lectin-labeled microglia, progressing to a tenfold increase in some focal regions by the terminal stage of the disease.
MICROGLIAL RESPONSE TO CEREBRAL MALARIA
Fig. 3.
97
Fig. 4. A,B: Tracings of photographic montages showing the distribution of NDPase-labeled microglia in relation to the retinal vasculature, in the ganglion cell layer. A: In control CBA/T6 uninfected mice, ramified microglia within the retina were evenly distributed throughout the retina. B: At the terminal stage of the disease, day 7 p.i., microglia with an activated morphology were concentrated in the vicinity of major retinal veins. The arteries are the narrower caliber vessels with red coloration, and the veins are the wider caliber vessels with the blue coloration. C,D: Retinal wholemounts from FMCM mice (days 5–7 p.i.), labeled with the GS lectin conjugated to horseradish peroxidase. C: High-power view of adherent monocytes containing
phagocytosed Monastral blue particles within a retinal vein at day 7 p.i. Phagocytosed parasites are evident as small dark particles within the monocytes, erythrocytes, and vessel lumen. An extravasated monocyte (arrow) is evident in the lower part of the field of view. D: High-power view of a larger monocyte, containing Monastral blue, in the retinal parenchyma, accompanied by two GS-labeled cells, one containing small amounts of Monastral blue and the other not. These data show that a number of these round GS-labeled 1 cells are derived from circulating monocytes, since only intravascular cells had access to the Monastral blue.
MICROGLIAL RESPONSE TO CEREBRAL MALARIA
99
Fig. 5. Retinal wholemounts from FMCM mice (days 5–7 p.i.), labeled with the GS lectin conjugated to horseradish peroxidase. A: Cells could be identified as monocytes by the presence of phagocytosed parasite debris and their GS lectin reactivity. A variation in the intensity of GS lectin labeling was evident among monocytes, some expressing high levels (arrow) while others expressed low levels (arrowheads). B: GS lectin-labeled monocytes appear to be leaving the
retinal vessel and entering the parenchyma (arrows) at day 7 p.i. C,D: A number of GS lectin-labeled cells in the parenchyma, with varying morphology, which cannot be specifically identified. C: A subpopulation of GS lectin-labeled cells characterized by the absence of processes and a cell diameter . 15 µm. D: Monocyte-like cells in the retinal parenchyma (small arrows) adjacent to a large cell with a ‘‘fried egg’’ morphology, of unknown origin (large arrow).
the nearest neighbor (64 vs. 71 µm; P , 0.01). This decrease in distance to nearest neighbor became more marked with progression of the disease, reaching a maximum at the fatal stage of the disease (56 vs. 71 µm; P , 0.001).
first was characterized by the absence of processes, a cell diameter . 15 µm and frequent vacuolation (Fig. 5C); such cells were found only at the terminal stage of FMCM. The second population, approximately 10 µm in diameter, contained no vacuoles and was often seen in close association with retinal vessels, suggesting passage through the retinal endothelium (Fig. 5B). This population of GS lectin-labeled cells could be found in small numbers in the retinal parenchyma throughout the progression of FMCM, the numbers increasing between days 6 and 7 p.i. (Fig. 5D). These cells without processes could be interpreted either as activated microglia or extravasated monocytes. To examine whether some of these cells were extravasated monocytes, animals were given an i.v. injection of the colloidal dye Monastral blue 2 h prior to sacrifice. Since Monastral blue was administered intravenously, it would only be accessible to cells originally found inside the vascular lumen. Extravasation of monocytes during FMCM was confirmed when round, GS lectin-labeled cells containing phagocytosed Monastral blue particles were found, albeit infrequently, within the retinal parenchyma (Fig. 4D). While there exist numerous reports of monocyte margination and
Extravasation of Monocytes During FMCM Monocytosis and monocyte margination As previously described (Chan-Ling et al., 1992; Neill and Hunt, 1992), there were increased numbers of monocytic cells containing Monastral blue in the retinal circulation at day 4 p.i. in FMCM. At day 5 p.i., monocyte margination, predominantly in venules, was also evident. Monocytosis and monocyte adherence to retinal vessels peaked at the terminal stage of FMCM (day 7 p.i., Fig. 4C; Chan-Ling et al., 1992; Neill and Hunt, 1992). GS lectin-labeled monocytes with malaria parasites inside the phagolysozomes were evident from days 5–7 (Fig. 5A). However, the intensity of GS lectin reactivity on monocytes was variable (arrows in Fig. 5A). Besides the presence of process-bearing microglia labeled with GS lectin, two other GS lectin-labeled populations of cells were evident during FMCM. The
100
MEDANA ET AL.
adhesion in FMCM, this is the first demonstration of monocyte extravasation in this disease.
DISCUSSION Essential components of the immunopathogenesis associated with FMCM are T lymphocytes (Finley et al., 1983; Grau et al., 1986), monocytes (Chan-Ling et al., 1992; Neill and Hunt, 1992; Neill et al., 1993; Thumwood et al., 1988), and cytokines (in particular tumor necrosis factor-alpha [TNF-a]; Grau et al., 1987, 1989a,b, 1990). Concepts such as the importance of TNF-a in the pathogenesis of FMCM have later been shown to be relevant to the human condition (Clark et al., 1987; Grau et al., 1990; Kwiatkowski et al., 1990). Thus it has been hypothesized that FMCM is the result of immune reactions in the periphery that result in non-specific damage to the cerebral vasculature, which may contribute to cerebral complications and death (Grau et al., 1989b; Hunt et al., 1992; Neill and Hunt, 1992). The current study suggests that this hypothesis be expanded to include active participation of microglia in a local CNS immune response, resulting in the cerebral complications associated with murine malaria. Microglial changes are evident very early in FMCM, well before the onset of cerebral symptoms. As early as day 2–3 p.i., although the majority of microglia were of normal size and shape, some showed an increase in GS lectin staining intensity and distensions on their processes, progressing to a decrease in process length and an increase in soma size. These less ramified microglia increased in numbers until the terminal stage of the disease. Such morphological changes are generally considered to reflect progression to a functionally ‘‘activated’’ phenotype (Thomas, 1992; Vrabec, 1975). An apparent redistribution of microglia toward the retinal vessels began around day 3 p.i. and was most clearly seen at the later stages of the disease process. Since there was no increase in density of retinal microglia demonstrated by the NDPase technique at any stage of the disease process, such redistribution could not be due to cell proliferation or to unmasking of already existing cells. Since these microglial responses are among the earliest observable events in FMCM, it is clear that they are unlikely to be a consequence of tissue damage but are, rather, a potential cause of the cerebral complications. Activation of microglia is normally a consequence of increased permeability of the CNS barrier to macromolecules or of the action of cytokines. Previous studies (Chan-Ling et al., 1992; Neill et al., 1993) have shown a general increase in permeability of the blood-retinal barrier as early as day 2–3 p.i. This change in CNS barrier function could be induced by direct actions of malaria parasite products on the endothelium. There is a precedent for this suggestion, since vasoactive factors released by P. falciparum have been shown to induce nitric oxide production by endothelial cells (Ghigo et al., 1995; Schofield et al., 1996). The early vascular changes in FMCM coincide with the observed microglial re-
sponses so it is possible that movement of plasma molecules across the CNS barrier initiates the morphological changes and accumulation of microglia toward the retinal vessels. It has been suggested that the resting state of ramified microglia is due to their lack of exposure to plasma proteins (Perry and Gordon, 1988, 1991). An increase in CNS barrier permeability also would allow the passage of inflammatory mediators, such as parasite exoantigens, into the CNS. Soluble parasite exoantigens are discharged from infected erythrocytes at the time of merozoite release and some, at least from P. falciparum, stimulate the release of cytokines such as interleukin-1 and TNF-a from macrophages (Riley et al., 1991; Taverne et al., 1990). If exoantigens from P. berghei had similar effects they could stimulate microglia to release potentially harmful mediators. Activated microglia have the potential to secrete reactive oxygen species (Colton and Gilbert, 1987; Sonderer et al., 1987), cytokines (Appel et al., 1995; Giulian et al., 1994; Renno et al., 1995), and nitric oxide (Boje and Arora, 1992; Merrill et al., 1993). All these factors previously have been implicated in the pathogenesis of FMCM (Grau et al., 1987, 1989b, 1990; Rockett et al., 1992; Thumwood et al., 1989). At later stages of the infection, cytokines released by circulating monocytes or lymphocytes could also enter the CNS through the compromised barrier and stimulate microglia to differentiate and release inflammatory mediators that contribute to the derangement of CNS function in FMCM. Cytokines play a major role in the development of the cerebral complications associated with this condition (Grau et al., 1987, 1989b, 1990). Serum cytokine levels, particularly those of TNF-a, are elevated in humans and mice at the time of the neurological syndrome, and treatment of CM-susceptible mice with anti-cytokine antibodies, such as antiTNF-a, prevents all neuropathological abnormalities (Clark et al., 1987; Grau et al., 1987, 1990; Kwiatkowski et al., 1990). However, higher plasma levels of TNF-a are found in other murine models of malaria that do not develop cerebral symptoms, which has led to the suggestion that TNF-a produced locally within the CNS may contribute more to the sequelae than does circulating TNF-a (Clark et al., 1987). The results of our current experiments suggest that microglia may be involved in local cytokine production. Consistent with this, we have found, by in situ hybridization and immunohistochemistry, that cells with the characteristics of microglia produce TNF-a during FMCM (Medana, Hunt, and Chaudhri, ms. submitted). Further studies of the properties of microglia isolated from mice at various stages of FMCM are in progress in our laboratory. We found an increase in focal density of G. simplicifolia-positive cells during the progression of FMCM, starting as early as day 3 p.i. From the NDPase maps it is clear that microglia are always present in these layers but under normal conditions they do not possess enough a-D-galactose residues to be visualized with the GS lectin. Other researchers have reported finding weak (Htain et al., 1994) or undetectable (Striet et al.,
MICROGLIAL RESPONSE TO CEREBRAL MALARIA
1988) GS labeling of ramified microglia in the normal adult CNS. This contrasts with activated microglia in the adult mouse, or amoeboid microglia in neonatal mice, where strong GS labeling can be achieved. The upregulation of these molecules during the progression of FMCM, allowing more cells to be visualized with the GS lectin, may reflect a change in the activity of the microglia, possibly aiding in cell-cell interactions. Furthermore, the activation of macrophages is associated with an increase in a-D-galactose residues (Maddox et al., 1982), and ligation of these residues leads to TNF-a production (Tabor et al., 1992; Warfel and ZuckerFranklin, 1992). Thus the increase in a-D-galactose residues, as a result of microglia activation, could increase the potential for TNF-a production by these cells. Human patients with cutaneous leishmaniasis and chronic Chagas’ disease have antibodies that recognize a-D-galactose residues on the surface of glycoproteins or glycolipids of the parasites. Warfel and ZuckerFranklin (1992) have hypothesized that these antibodies will bind to a-D-galactose epitopes on stimulated monocytes or macrophages and thus alter their biological activity. This scenario can be adapted to human cerebral malaria, since antibodies against a-D-galactose residues have been found in the cerebrospinal fluid of patients suffering from human cerebral malaria (Ravindran and Das, 1992). Interaction of these anti-galactosyl antibodies with microglia could influence the functions of these cells, for example, by promoting TNF-a production, and could contribute to the cerebral complications associated with human CM. Local production of cytokines by microglia is also implied by the finding that astrogliosis occurs in FMCM (Medana et al., 1996). Co-visualization of monocytes and astrocytes with Monastral blue and anti-GFAP, respectively, showed that vessels in close association with gliotic patches were usually free of marginating monocytes. Thus, if cytokines are involved in the astrogliosis the most likely source is the microglia. Giulian and colleagues (1985, 1986) have suggested that astrogliosis is, in part, the result of factors released by ‘‘activated’’ microglia. Cytokines issuing from the parenchymal side of the blood-brain barrier are thought to play a role in the development of bacterial meningitis (Saez-Llorens et al., 1990). Fragments of the bacterial cell wall stimulate cytokine production from the parenchymal side of the blood-brain barrier, initiating leukocyte adhesion to the endothelium and subsequent diapedesis. The leukocytes, once within the brain parenchyma, are thought to enhance cytokine production and accelerate the disruption of the blood-brain barrier (Saez-Llorens et al., 1990). A similar mechanism might be occurring during FMCM. In many disease states a major portion of the ‘‘activated’’ macrophage population within the CNS is derived from monocytes. The ultimate contribution of these cells to the activated macrophage population within the retina during FMCM cannot be proved as there is no specific stain that will differentiate between
101
murine monocytes and microglia. Monocytes could enter the CNS at sites of hemorrhage, or via binding to adhesion molecules and emigration. Petechial hemorrhages (Chan-Ling et al., 1992; Thumwood et al., 1988) and upregulation of endothelial adhesion molecules (Grau et al., 1990; Ma et al., 1996) are both features of FMCM, particularly during the later stages. However, massive, widespread emigration of monocytes was not strikingly obvious at any stage. G. simplicifolia-labeled cells that had the same size and morphology as the monocytes seen in the lumen of the vessels were found in the retinal parenchyma, predominantly at the onset of cerebral symptoms. Some of these cells also contained Monastral blue that had been administered intravenously. The origin of the larger, round GS1 cells is hard to determine without a specific marker, although they have the characteristic morphology of activated microglia described in many pathological conditions (Thomas, 1992). Thus there are at least two possible origins of the GS lectin-labeled cells with no processes: 1) reactive microglia derived from the transformation of ramified microglia; and 2) extravasated monocytes. It seems likely that most of these cells are microglia since numerous transitional forms can be observed throughout the disease process, even before margination and adhesion to the vascular endothelium of monocytes was visible. It is well established that the interaction between monocytes and the CNS microvascular endothelium is an important determinant in the pathogenesis of FMCM (Chan-Ling et al., 1992; Neill and Hunt, 1992; Neill et al., 1993), and since significant numbers of monocytes have been found lining the walls of the blood vessels, it is conceivable that they could produce factors that contribute to the vascular and astrocyte damage associated with the disease. Electron micrographs showing monocytes adjacent to endothelial cell damage (Thumwood et al., 1988) and immunohistochemical studies showing adherent monocytes adjacent to vascular regions lacking astrocyte ensheathment (Medana et al., 1996) in the later stages of the disease support this interpretation. In conclusion, using the retinal wholemount technique, we have provided the first evidence that in FMCM there is early activation of microglia, occurring substantially before the onset of cerebral symptoms, which is consistent with a role for these cells in an immunopathological process. Changes in the morphology of the microglia may be initiated by an early increase in the permeability of the CNS barrier to macromolecules, induced by products of the circulating malaria parasites. Functional changes in microglia, for example, the release of inflammatory mediators, occurring later in the disease process could be induced, or reinforced, by cytokines derived from circulating leukocytes or from monocytes adhering to the CNS microvascular endothelium. The combination of increased permeability of the CNS barrier and secretion of various factors by microglia and monocytes would then interfere with CNS functions, leading to cerebral symptoms and death.
102
MEDANA ET AL.
ACKNOWLEDGMENTS This work was supported by grants from the National Health and Medical Research Council and Sydney University Research Grants Scheme. Isabelle Medana was supported by an Australian Postgraduate Research Award. The authors would also like to thank Clive Jeffery for his photographic assistance.
REFERENCES Appel, K., Honegger, P., and Gebicke, H.P. (1995) Expression of interleukin-3 and tumor necrosis factor-beta mRNAs in cultured microglia. J. Neuroimmunol., 60:83–91. Bauer, J., Huitinga, I., Zhao, W., Lassmann, H., Hickey, W.F., and Dijkstra, C.D. (1995) The role of macrophages, perivascular cells, and microglial cells in the pathogenesis of experimental autoimmune encephalomyelitis. Glia, 15:437–446. Benveniste, E.N. (1992) Inflammatory cytokines within the central nervous system: Sources, function, and mechanism of action. Am. J. Physiol., 263:C1–C16. Boje, K.M. and Arora, P.K. (1992) Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death. Brain Res., 587:250–256. Boya, J., Carbonell, A.L., Calvo, J.L., and Borregon, A. (1991) Microglial cells in the central nervous system of the rabbit and rat: Cytochemical identification using two different lectins. Acta Anat., 140:250–253. Chan-Ling, T. (1994) Glial, neuronal and vascular interactions in the mammalian retina. In: Progress in Retinal Research. N. Osborne and G. Chader, eds. Pergamon Press, Oxford, pp. 357–389. Chan-Ling, T., Halasz, P., and Stone, J. (1990) Development of retinal vasculature in the cat: Processes and mechanisms. J. Comp. Neurol., 9:459–478. Chan-Ling, T., Neill, A.L., and Hunt, N.H. (1992) Early microvascular changes in murine cerebral malaria detected in retinal wholemounts. Am. J. Pathol., 140:1121–1130. Clark, I.A., Cowden, W.B., Butcher, G.A., and Hunt, N.H. (1987) Possible roles of tumor necrosis factor in the pathology of malaria. Am. J. Pathol., 129:192–199. Colton, C.A. and Gilbert, D.L. (1987) Production of superoxide anions by a CNS macrophage, the microglia. FEBS Lett., 223:284–288. Finley, R., Weintraub, J., Lewis, J.A., Engers, H.D., Zubler, R., and Lambert, P.-H. (1983) Prevention of cerebral malaria by adoptive transfer of malaria specific cultured T cells into mice infected with Plasmodium berghei. J. Immunol., 131:1522–1526. Gelman, B.B. (1993) Diffuse microgliosis associated with cerebral atrophy in the acquired immunodeficiency syndrome. Ann. Neurol., 34:65–70. Ghigo, D., Todde, R., Ginsburg, H., Costamagna, C., Gautret, P., Bussolino, F., Ulliers, D., Giribaldi, G., Deharo, E., Gabrielli, G. (1995) Erythrocyte stages of Plasmodium falciparum exhibit a high nitric oxide synthase (NOS) activity and release an NOS-inducing soluble factor. J. Exp. Med., 182:677–688. Giulian, D. and Baker, T.J. (1985) Peptides released by ameboid microglia regulate astroglial proliferation. J. Cell Biol., 101:2411– 2415. Giulian, D., Baker, T.J., Shih, L.-C., and Lachman, L.B. (1986) Interleukin-1 of the central nervous system is produced by amoeboid microglia. J. Exp. Med., 164:594–604. Giulian, D., Li, J., Li, X., George, J., and Rutecki, P.A. (1994) The impact of microglia-derived cytokines upon gliosis in the CNS. Dev. Neurosci., 16:128–136. Giulian, D., Haverkamp, L.J., Li, J., Karshin, W.L., Yu, J., Tom, D., Li, X., and Kirkpatrick, J.B. (1995) Senile plaques stimulate microglia to release a neurotoxin found in Alzheimer brain. Neurochem. Int., 27:119–137. Grau, G.E., Piguet, P.-F., Engers, H.D., Louis, J.A., Vassali, P., and Lambert, P.-H. (1986) L3T41 T lymphocytes play a major role in the pathogenesis of murine cerebral malaria. J. Immunol., 137:2348– 2354. Grau, G.E., Fagardo, L.F., Piguet, P.-F., Lambert, P.-H., and Vassali, P. (1987) Tumor necrosis factor (cachectin) as an essential mediator in murine malaria. Science, 237:1210–1212. Grau, G.E., Heremans, H., Piguet, P.-F., Pointaire, P., Lambert, P.-H., Billiau, A., and Vassalli, P. (1989a) Monoclonal antibody against interferon gamma can prevent experimental cerebral malaria and
its associated overproduction of tumor necrosis factor. Proc. Natl. Acad. Sci. USA., 86:5572–5574. Grau, G.E., Piguet, P.-F., Vassali, P., and Lambert, P.-H. (1989b) Tumor-necrosis factor and other cytokines in cerebral malaria: Experimental and clinical data. Immunol. Rev., 112:49–70. Grau, G.E., Bieler, G., Pointaire, P., De Kossodo, S., Tacchini-Cotier, F., Vassali, P., Piguet, P.F., and Lambert, P.-H. (1990) Significance of cytokine production and adhesion molecules in malarial immunopathology. Immunol. Lett., 25:189–194. Halasz, P. and Martin, P. (1984) A microcomputer based system for semiautomatic analysis of histochemical sections. R. Microsc. Soc. Proc., 9:312. Htain, W.-W., Leong, S.-K., and Ling, E.-A. (1994) A comparative Mac-1 immunocytochemical and lectin histochemical study of microglial cells in the normal and athymic mice. Glia, 12:44–51. Hume, D.A., Perry, V.H., and Gordon, S. (1983) Immunohistochemical localisation of a macrophage-specific antigen in developing mouse retina: Phagocytosis of dying neurons and differentiation of microglial cells to form a regular array in the plexiform layers. J. Cell Biol., 97:253–257. Hunt, N.H., Kopp, M., and Stocker, R. (1992) Free radicals and antioxidants in malaria. In: Lipid-soluble Antioxidants: Biochemical and Clinical Applications. S. Ong and L. Packer, eds. Birkhauser, Basel, pp. 337–354. Kwiatkowski, D., Hill, A.V.S., Sanbou, I., Twumasi, P., Castrane, J., Manogue, K.R., Cerami, A., Brewster, D.R., and Greenwood, B.M. (1990) TNF concentration in fatal cerebral malaria. Lancet, 336: 1201–1204. Ma, N., Hunt, N.H., Madigan, M.C., and Chan-Ling, T. (1996) Correlation between enhanced vascular permeability, upregulation of cell adhesion molecules and monocyte adhesion in retina during fatal murine cerebral malaria. Am. J. Pathol. 149:1745–1762. Maddox, D.E., Shibata, S., and Goldstein, I.J. (1982) Stimulated macrophages express a new glycoprotein receptor reactive with Griffonia simplicifolia I-B4 isolectin. Proc. Natl. Acad. Sci. USA, 79:166–170. Maegraith, B. and Fletcher, A. (1971) The pathogenesis of mammalian malaria. Adv. Parasitol., 10:42–57. McGeer, P.L., Walker, D.G., Akiyama, H., Yasuhara, O., and McGeer, E.G. (1994) Involvement of microglia in Alzheimer’s disease. Neuropathol. Appl. Neurobiol., 20:191–192. Medana, I.M., Chan-Ling, T., and Hunt, N.H. (1996) Redistribution and degeneration of retinal astrocytes in experimental murine cerebral malaria: Relationship to disruption of the blood-retinal barrier. Glia, 16:51–64. Merrill, J.E., Ignarro, L.J., Sherman, M.P., Melinek, J., and Lane, T.E. (1993) Microglial cell cytotoxicity of oligodendrocytes is mediated through nitric oxide. J. Immunol., 151:2132–2141. Neill, A.L. and Hunt, N.H. (1992) Pathology of fatal and resolving cerebral malaria. Parasitology, 105:165–175. Neill, A.L., Chan-Ling, T., and Hunt, N.H. (1993) Comparisons between microvascular changes in cerebral and non-cerebral malaria mice, using the retinal wholemount technique. Parasitology, 107: 477–487. Newcombe, J., Li, H., and Cuzner, M.L. (1994) Low density lipoprotein uptake by macrophages in multiple sclerosis plaques: Implications for pathogenesis. Neuropathol. Appl. Neurobiol., 20:152–162. Nieto, S.M. and Mora, F. (1994) Active microglia, sick astroglia and Alzheimer type dementias. Neuroreport, 5:375–380. Penfold, P.L., Provis, J.M., and Liew, S.C.K. (1993) Human retinal microglia express phenotypic characteristics in common with dendritic antigen-presenting cells. J. Neuroimmunol., 45:183–192. Perry, V.H. and Gordon, S. (1988) Macrophages and microglia in the nervous system. Trends Neurosci., 11:273–277. Perry, V.H. and Gordon, S. (1991) Macrophages and the nervous system. Int. Rev. Cytol., 125:203–244. Perry, V.H., Hume, D.A., and Gordon, S. (1985) Immunohistochemical localization of macrophages and microglia in the adult and developing mouse brain. Neuroscience, 15:313–326. Ravindran, B. and Das, B.K. (1992) Human cerebral malaria: a-Galactosyl antibodies in cerebrospinal fluid. Trans. R. Soc. Trop. Med. Hyg., 86:132–133. Renno, T., Krakowski, M., Piccirillo, C., Lin, J.Y., and Owens, T. (1995) TNF-a expression by resident microglia and infiltrating leukocytes in the central nervous system of mice with experimental allergic encephalomyelitis. Regulation by Th1 cytokines. J. Immunol., 154: 944–953. Rest, J.R. (1982) Cerebral malaria in inbred mice. I. A new model and its pathology. Trans. R. Soc. Trop. Med. Hyg., 76:410–415. Riley, E.M., Jakobsen, P.H., Allen, S.J., Wheeler, J.G., Bennett, Soren, J., and Greenwood, B.M. (1991) Immune response to soluble exoantigens of Plasmodium falciparum may contribute to both pathogen-
MICROGLIAL RESPONSE TO CEREBRAL MALARIA esis and protection in clinical malaria: Evidence from longitudinal, prospective study of semi-immune African children. Eur. J. Immunol., 21:1019–1025. Rockett, K.A., Awburn, M.M., Aggarwal, B.B., Cowden, W.B., and Clark, I.A. (1992) In vivo induction of nitrite and nitrate by tumor necrosis factor, lymphotoxin, and interleukin-1: Possible roles in malaria infection and immunity. Infect. Immun. 60:3725–3720. Saez-Llorens, X., Ramilo, O., Mustafa, M., Mertsola, J., and McCracken, G.H. (1990) Molecular pathophysiology of bacterial meningitis: Current concepts and therapeutic implications. J. Pediatr., 116:671–684. Sanyal, S. and De Ruiter, A. (1985) Inosine diphosphatase as a histochemical marker of retinal microvasculature, with special reference to transformation of microglia. Cell Tissue Res., 241:291– 297. Schnitzer, J. (1989) Enzyme-histochemical demonstration of microglial cells in the adult and postnatal rabbit retina. J. Comp. Neurol., 282:249–263. Schofield, L., Novakovic, S., Gerold, P., Schwarz, R.T., McConville, M.J., Tachado, S.D. (1996) Glycosylphosphatidylinosito toxin of Plasmodium upregulates intercellular adhesion molecule-1, vascular adhesion molecule-1, and E-selectin expression in vascular endothelial cells and increases leukocyte and parasite cytoadherence via tyrosine kinase-dependent signal transduction. J. Immunol., 156:1886–1896. Selmaj, K., Raine, C.S., Cannella, B., and Brosnan, C.F. (1991) Identification of lymphotoxin and tumor necrosis factor in multiple sclerosis lesions. J. Clin. Invest., 87:949–954. Sonderer, B., Wild, P., Wyler, R., Fontana, A., Peterhans, E., and Schwyzer, M. (1987) Murine glia cells in culture can be stimulated to generate reactive oxygen. J. Leukocyte Biol., 42:463–473.
103
Streit, W.J. and Kreutzberg, G.W. (1987) Lectin binding by resting and reactive microglia. J. Neurocytol., 16:249–260. Streit, W.J., Graeber, M.B., and Kreutzberg, G.W. (1988) Functional plasticity of microglia: A review. Glia, 1:301–307. Tabor, D.R., Theus, S.A., Barnett, J.B., and Jacobs, R.F. (1992) Macrophage membrane glycoprotein binding of Griffonia simplicifolia I-B4 induces TNF-a production and a tumoricidal response. J. Cell. Physiol., 152:500–506. Taverne, J., Bate, C.A.W., and Playfair, J.H.L. (1990) Malaria exoantigens induce TNF, are toxic and are blocked by T-independent antibody. Immunol. Lett., 25:207–212. Thomas, W.E. (1992) Brain macrophages: Evaluation of microglia and their functions. Brain Res. Rev., 17:61–74. Thumwood, C.M., Hunt, N.H., Clark, I.A., and Cowden, W.B. (1988) Breakdown of the blood-brain barrier in murine cerebral malaria. Parasitology, 96:579–589. Thumwood, C.M., Hunt, N.H., Cowden, W.B., and Clark, I.A. (1989) Antioxidants can prevent cerebral malaria in Plasmodium bergheiinfected mice. Br. J. Exp. Pathol., 70:293–303. Toro, G. and Roman, G. (1978) Cerebral malaria: A disseminated vasculomyelopathy. Arch. Neurol., 35:271–275. Vazeux, R. (1991) AIDS encephalopathy and tropism of HIV for brain monocytes/macrophages and microglial cells. Pathobiology, 59:214– 218. Vrabec, F. (1975) Activated human microglia under pathological conditions. Arch. Klin. Exp. Ophthal., 196:49–60. Warfel, A.H. and Zucker-Franklin, D. (1992) Specific ligation of surface a-D-galactosyl epitopes markedly affects the quantity of four major proteins secreted by macrophages. J. Leukocyte Biol., 52:80–83.