The bisphosphonate clodronate depletes microglial cells in excitotoxically injured organotypic hippocampal slice cultures

Available online at www.sciencedirect.com R

Experimental Neurology 181 (2003) 1–11

www.elsevier.com/locate/yexnr

The bisphosphonate clodronate depletes microglial cells in excitotoxically injured organotypic hippocampal slice cultures A. Kohl,a,b,1 F. Dehghani,b,1 H.-W. Korf,b and N.P. Hailera,* a b

University Hospital for Orthopaedic Surgery Friedrichsheim, Marienburgstrasse 2, D-60528 Frankfurt am Main, Federal Republic of Germany Institute of Anatomy II, Johann Wolfgang Goethe-University, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Federal Republic of Germany Received 5 March 2002; revised 22 October 2002; accepted 12 November 2002

Abstract The bisphosphonate clodronate, clinically used in the treatment of osteoporosis, is known to deplete cells of the monocytic lineage. Using an in vitro model of excitotoxic damage in organotypic hippocampal slice cultures (OHSC), we investigated whether clodronate can also prevent microglial activation that occurs in CNS pathologies. Lesioning of OHSC was performed by application of 50 ␮M N-methyl-Daspartate (NMDA) for 4 h after 6 days in vitro (div). Treatment of lesioned OHSC with clodronate (1000, 100, or 10 ␮g/ml) resulted in an almost complete abrogation of the microglial reaction after 3 further div: Confocal laser scanning microscopy showed that the number of Griffonia simplicifolia isolectin B4-labeled (IB⫹ 4 ) microglial cells in the dentate gyrus (DG) was reduced to 4.25% compared with OHSC treated with NMDA alone. Continuous treatment with clodronate (100 or 10 ␮g/ml) of lesioned OHSC for 9 days resulted in a further reduction in the number of microglial cells (reduction to 2.72%). The number of degenerating, propidium iodide-labeled (PI⫹) neurons in lesioned OHSC that received clodronate treatment between 6 and 9 div was not significantly different from OHSC treated with NMDA alone. However, the number of PI⫹ neurons in lesioned OHSC that received continuous clodronate treatment for 9 div was significantly higher when compared to NMDA-lesioned OHSC. In summary, clodronate is able to reduce microglial activation induced by excitotoxic neuronal injury. Our results demonstrate that clodronate is a useful tool in the investigation of neuron– glia interactions because it induces an efficient depletion of microglial cells that are activated after excitotoxic CNS injury. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Apoptosis; Clodronic acid; Excitotoxicity; N-Methyl-D-aspartate; Neuron; Neurotoxicity; Rat

Introduction In the adult CNS microglial cells display a ramified morphology with numerous branched cellular processes and a relatively small soma (Rio-Hortega, 1932; Streit et al., 1988). These cells are considered “resting” cells with many characteristics of immunologically dormant tissue macrophages (Czapiga and Colton, 1999). Microglial cells quickly respond to all kinds of CNS pathologies and play an important role in diseases or injuries (Glass and Wesselingh, 2001; Marx et al., 1998). As a consequence of CNS lesions, microglial cells are rapidly activated. This activation is * Corresponding author. Fax: ⫹49-69-6705375. E-mail address: [email protected] (N.P. Hailer). 1 Both authors contributed equally to this work.

characterized by the morphological transformation of ramified cells into cells that display large, amoeboid somata with few pseudopodia (Csuka et al., 2000; Kreutzberg, 1996; Raivich et al., 1994). Through the secretion of neurotoxic or neuroprotective agents, stimulated macrophages and microglial cells react to CNS pathologies and determine the pattern and degree of functional recovery (Moore and Thanos, 1996). The bisphosphonate clodronate is clinically used in the treatment of osteoporosis, Paget’s disease, and hypercalcaemia in hyperparathyroidism or metastatic bone disease (Russell and Rogers, 1999). Bisphosphonates inhibit the activation of osteoclasts by inducing osteoclast apoptosis (Benford et al., 2001) and thus by inhibiting bone resorption (Frith et al., 2001). Apart from its effects on osteoclasts, clodronate has been shown to inhibit peripheral macrophage

0014-4886/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0014-4886(02)00049-3

2

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

function and to induce apoptotic death in peritoneal macrophages (Van Rooijen et al., 1996). Consequently, clodronate has been used in different experimental approaches to investigate effects of macrophage depletion. The intravenous injection of liposome-encapsulated clodronate has been shown to reduce the severity of LPS-induced neurodegeneration (Zito et al., 2001), and clodronate treatment reduced parenchymal macrophage invasion (Koennecke et al., 1999). The intention of our study was to investigate whether clodronate also exerts effects on the macrophages residing in the CNS, i.e., microglial cells. The organotypic hippocampal slice culture (OHSC) was chosen as the experimental model because it contains resident microglial cells only and an influx of monocytes from the blood stream does not occur in this in vitro preparation. Several in vitro lesions have been performed in the OHSC, e.g., excitotoxic injury by application of N-methylD-aspartate (NMDA) or perforant path transection. Treatment with NMDA has been found to cause a massive loss of granule cells in the dentate gyrus (DG), followed by the rapid accumulation of activated, amoeboid microglial cells at sites of neuronal injury (Heppner et al., 1998). Lesioning of OHSC by NMDA application was performed after 6 days in vitro (div), because all microglial cells in the inner layers of the slice culture are inactive at this time (Hailer et al., 1996). In this study, OHSC lesioned with NMDA alone were compared with OHSC that received additional clodronate treatment after or prior to lesioning. Microglial cells were stained with FITC-conjugated Griffonia simplicifolia isolectin B4 (IB4) (Streit, 1990; Streit and Kreutzberg, 1987), and their number and distribution were determined by confocal laser scanning microscopy. To quantify the amount of neuronal injury, the number of propidium iodidelabeled (PI⫹), degenerating neurons was determined in the granule cell layer of the dentate gyrus (Fig. 1B). Additional staining with hematoxylin/IB4 (H/IB4) was used to assess the hippocampal cytoarchitecture.

Materials and methods To prepare OHSC, 8-day-old Wistar rats were decapitated and their brains were dissected under sterile conditions (Ga¨ hwiler et al., 1997). The frontal pole and the cerebellum were removed and the brains were placed in minimal essential medium (MEM, Gibco BRL Life Technologies, Eggenstein, Germany), containing 1% glutamine (Gibco) at 4°C. Approximately 1 mm of the ventral surface was removed and 350-␮m-thick slices were prepared using a sliding vibratome (Vibratome 1000 Classic, St. Louis, MO, USA). From each brain six to eight OHSC were obtained and used for further experiments. OHSC were transferred into cell culture inserts (Becton Dickinson, Franklin Lakes, NJ, USA; pore size 0.4 ␮m) that were placed in 6-well culture dishes (Becton Dickinson), containing 1 ml culture

medium per well. Culture medium consisted of 50% MEM, 25% Hanks’ balanced salt solution (HBSS; Gibco), 25% normal horse serum (NHS; Gibco), 2% glutamine, 1␮g/ml insulin (Boehringer, Mannheim, Germany), 1.2 mg/ml glucose (Braun, Melsungen, Germany), 0.1 mg/ml streptomycin (Sigma Chemicals, Deisenhofen, Germany), 100 U/ml penicilline (Sigma), and 0.8 ␮g/ml vitamin C (Sigma), pH ⫽ 7.4. The culture dishes were incubated at 35°C in a fully humidified atmosphere with 5% CO2. OHSC were divided into six groups and treated according to the following protocols (Fig. 1A): (A) Unlesioned OHSC were incubated in control medium (H/IB4, n ⫽ 38; fluorescence staining, n ⫽ 22) from 0 div until 9 div. (B) OHSC were incubated in control medium for 6 days and then lesioned with 50 ␮M NMDA for 4 h and cultured in control medium for another 3 days (H/IB4, n ⫽ 32; fluorescence staining, n ⫽ 24). (C) Lesioned OHSC were incubated in clodronate-containing medium (1000 ␮g/ml: H/IB4 n ⫽ 3; fluorescence staining, n ⫽ 4; or 100 ␮g/ml: H/IB4, n ⫽ 17; fluorescence staining, n ⫽ 16; or 10 ␮g/ml: H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 4) from 6 div until 9 div. (D) Unlesioned OHSC were incubated in clodronate-containing medium (1000 ␮g/ml: H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 4; 100 ␮g/ml: H/IB4, n ⫽ 6; fluorescence staining, n ⫽ 9; 10 ␮g/ml:H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 4) from 6 div until 9 div. (E) Lesioned OHSC were cultured in clodronate-containing medium (100 ␮g/ml: H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 7; 10 ␮g/ml: H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 4) from 0 div until 9 div. (F) Unlesioned OHSC were cultured in clodronate-containing medium (100 ␮g/ml: H/IB4, n ⫽ 5; fluorescence staining, n ⫽ 6; 10 ␮g/ml; H/IB4, n ⫽ 3; fluorescence staining, n ⫽ 4) from 0 div until 9 div. OHSC of each group were incubated with medium containing PI (5 ␮g/ml) 2 h prior to fixation. PI is a fluorescent dye that intercalates in the DNA of cells with permeable or lysed cell membranes. Thus, PI labeling is a valid indicator of cell death in general (Noraberg et al., 1999). The uptake of PI to identify degenerating neurons in OHSC has been established by Pozzo-Miller et al. (1994), who examined hippocampal slice cultures under different culture conditions. All OHSC were fixed with a mixture of 4% paraformaldehyde, 15% picric acid, and 0.1% glutaraldehyde in 0.1 M phosphate buffer (PB) for 15 min, washed with 0.1 M PB, and postfixed in the same mixture without glutaraldehyde for 1 h. OHSC were finally washed with 0.1 M PB for 1 h and carefully removed from the cell culture membrane. For lectin histochemistry OHSC were placed into 0.8 M sucrose solution containing 10% NaN3 for 2 days and cut into 14-␮m-thick sections in a Microm HM 560 cryostat (Microm, Walldorf, Germany) at ⫺21°C. The sections were mounted on gelatin-coated glass slides. After endogenous peroxidase quenching with methanol and H2O2 (0.03%), they were washed with phosphate-buffered saline (PBS) and incubated with NHS (diluted 1:20 in PBS containing 0.03%

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

3

Fig. 1. (A) Different experimental groups and treatments in overview: CTL, unlesioned OHSC were treated with control medium for 9 days. NMDA, OHSC were incubated in control medium for 6 days, then lesioned with 50 ␮M NMDA for 4 h and cultured in control medium for another 3 days. NMDA⫹CLO 6 div, lesioned OHSC were treated with clodronate (1000, 100, or 10 ␮g/ml) from 6 until 9 div. CLO, unlesioned OHSC were treated with clodronate (1000, 100, or 10 ␮g/ml) from 6 until 9 div. NMDA⫹CLO 0 div, lesioned OHSC were treated with clodronate (100 or 10 ␮g/ml) from 0 until 9 div. CLO 0 div, unlesioned OHSC were treated with clodronate (100 or 10 ␮g/ml) from 0 until 9 div. (B) Anatomical regions in the hippocampal formation: CA1, CA2, and CA3 belong to the pyramidal layer of the cornu ammonis (CA); the granule cell layer (GCL) and the molecular layers (ML) compose the dentate gyrus; HI points to the hilar area, SUB marks the subiculum, and EC indicates the entorhinal cortex; the framed area indicates the region shown in Figs. 2 and 3; only the highlighted region of the GCL was considered for statistical analysis.

Triton) for 30 min. Microglial cells were labeled with biotinylated IB4 (Sigma, Taufkirchen, Germany; diluted 1:50 in PBS–Triton with 5% bovine albumin serum) and incubated with avidin– biotin complex (diluted 1:100 in PBS– Triton) and 3,3⬘-diaminobenzidine (DAB; Sigma) as a chromogene. After washing, the sections were counterstained with hematoxylin and coverslipped with Entellan (Merck, Darmstadt, Germany). For confocal laser scanning microscopy PI-treated OHSC were washed with PBS–Triton, incubated with normal goat serum (NGS; diluted 1:20 in PBS–Triton) for 30

min, stained with FITC-conjugated IB4 (diluted 1:20 in PBS–Triton; Sigma, Deisenhofen, Germany) for 24 h, washed with PBS–Triton and distilled water for 10 min, respectively, and coverslipped with DAKO fluorescent mounting medium (DAKO Diagnostika GmbH, Hamburg, Germany). OHSC were analyzed using a Zeiss confocal laser scanning microscope (LSM 510, Zeiss, Go¨ ttingen, Germany). In order to visualize IB⫹ 4 microglial cells, monochromatic light at 488 nm with a dichroic beam splitter (FT 488/543) and an emission bandpass filter of 505–530 nm were used. For detection of PI⫹ cells, monochromatic light

4

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

at 543 nm and an emission bandpass filter of 585– 615 nm were used. Using Z-mode of the LSM 510 whole OHSC were cut optically into 2-␮m-thick sections. The number of PI⫹ or IB⫹ 4 cells was determined in every third optical section in the granule cell layer (GCL) of the dentate gyrus (cells/ GCL) (Hailer et al., 2001). The hilar area and the CA3 region of the hippocampus were not considered for the quantitative analysis (Fig. 1B). Confocal images were obtained at 160-fold magnification and were converted into a binary image that was segmented and measured by the Zeiss KS 400 System (Zeiss, Hallbergmoos, Germany). The mean cell number in fluorescent labeled OHSC was calculated for each group and statistical analysis was performed. The oneway ANOVA test was used to determine whether the effect of the treatment on the number of either IB⫹ 4 microglial cells or PI⫹ nuclei was statistically significant (P ⬍ 0.05). After significant differences were found with the one-way ANOVA test, the Dunnett test for multiple comparisons was performed. Unlesioned OHSC treated with clodronate (mean values) were compared to control OHSC (mean values); lesioned OHSC treated with clodronate (mean values) were compared to OHSC treated with NMDA alone (mean values). To illustrate the effect of NMDA lesioning, we compared control OHSC to OHSC that were treated with NMDA alone.

Results Group A: unlesioned OHSC cultured for 9 days contain ramified microglial cells and show an excellent neuronal preservation DAB-labeled IB4 preparations of unlesioned control OHSC showed that few microglial cells were distributed in the different layers of the DG, mainly in the molecular and plexiform layers. The GCL contained a very small number of IB⫹ 4 microglial cells. The majority of these microglial cells possessed branched, usually tender cytoplasmic processes that extended in all directions. The morphological appearance of microglial cells was thus consistent with the ramified phenotype, indicating a resting state (Fig. 2A). Confocal laser scanning microscopy confirmed that IB⫹ 4 microglial cells were localized mainly in the molecular and

plexiform layers displaying a ramified morphology. Furthermore, only a very small number of IB⫹ 4 microglial cells (7.5/GCL; Fig. 2G; Table 1 ) was detected in the GCL of the DG (Fig. 2B). Conventional hematoxylin preparations of control OHSC showed an intact cytoarchitecture of the hippocampal formation. The stratification of the DG remained intact: The molecular layers, the hilus, and the GCL could be readily distinguished and showed a regular distribution of principal cells and interneurons (Fig. 2A). These findings obtained by bright field microscopy were confirmed by confocal laser scanning microscopy and quantitative morphometry. Very few PI⫹ nuclei were found in the GCL of control OHSC (15.06/GCL; Fig. 2H; Table 1), indicating good neuronal preservation and a healthy microenvironment (Fig. 2B). Group B: lesioning of OHSC with NMDA results in an accumulation of activated microglial cells and a loss of granule cells A robust microglial activation was observed in lesioned OHSC, especially in the GCL where microglial cells accumulated at sites of neuronal injury. The adjacent layers of the DG also contained large numbers of IB⫹ 4 microglial cells. The morphology of these IB⫹ 4 cells was round or amoeboid, reflecting an activated state of the cells. Some microglial cells lost all cellular processes, and others retained some short and plump processes (Fig. 2C). Confocal laser scanning microscopy showed microglial cells throughout all layers of the hippocampal formation. Compared with unlesioned OHSC, microglial cells accumulated in lesioned OHSC at sites of neuronal injury, i.e., in the GCL of the DG (Fig. 2D). Quantitative analysis of the microglial response toward neuronal injury revealed that the number of IB⫹ 4 microglial cells was significantly higher in lesioned OHSC (36.81/GCL; P ⬍ 0.01, Fig. 2G; Table 1) than in unlesioned OHSC (7.5/GCL). The FITC-IB⫹ 4 microglial cells also displayed a round or amoeboid morphology typical of activated cells. Lesioned OHSC showed a massive destruction of the regular hippocampal cytoarchitecture. Numerous intensely stained, often condensed nuclei were observed in the GCL of the DG, indicating damage of neurons after lesioning. Most of these nuclei were fragmented, suggesting karyorhexis of injured neurons (Fig. 2C). PI labeling, confocal

Fig. 2. Treatment of lesioned OHSC with clodronate (100 ␮g/ml) from 6 div until 9 div reduced the number of microglial cells (A–F); statistical analysis ⫹ of the mean number of IB⫹ 4 microglial cells and PI neurons (G, H). (A, B) CTL: Few ramified microglial cells (green) were located mainly in the noncellular ⫹ layers of the DG and very low numbers of PI neurons (red) in the GCL indicated an excellent neuronal preservation. (C, D) NMDA: The number of ⫹ amoeboid IB⫹ 4 microglial cells and the number of PI neurons in the GCL dramatically increased after incubation with NMDA. (E, F) NMDA⫹CLO 6 div: A strong reduction in the number of microglial cells was observed after additional treatment with clodronate, as shown by the significantly lower number ⫹ of amoeboid IB⫹ 4 microglial cells. (G) The mean number of IB4 microglial cells was significantly reduced in clodronate-treated OHSC when compared to OHSC that were treated with NMDA alone. (H) Treatment with clodronate did not affect the mean number of PI⫹ nuclei. HI indicates the hilar area; GCL indicates the granule cell layer; arrowheads show the borders between the GCL and the molecular layer of the DG; arrows point to IB⫹ 4 microglial cells, open arrowheads indicate degenerating neurons; scale bars in overview, 50 ␮m; scale bars in higher magnification, 10 ␮m; asterisks in G and H indicate statistical significance (P ⬍ 0.01) versus treatment with NMDA alone; error bars in G and H represent the SEM.

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

5

6

.

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

7

Table 1 ⫹ Number of IB⫹ nuclei in the different experimental groups 4 microglial cells and PI CTL

⫹ 4 ⫹

IB microglial cells PI nuclei

7.5 ⫾ 0.89 15.06 ⫾ 3.815 NMDA

⫹ 4 ⫹

IB microglial cells PI nuclei

36.81 ⫾ 2.302 125.8 ⫾ 9.901

CLO 6 div [␮g/ml]

CLO 0 div [␮g/ml]

[1000]

[100]

[10]

[100]

[10]

1.5 ⫾ 0.5 24.88 ⫾ 5.397

3.26 ⫾ 0.756 9.11 ⫾ 2.111

3 ⫾ 0.598 0

0.84 ⫾ 0.579 8.83 ⫾ 3.358

1.67 ⫾ 0.802 19.5 ⫾ 5.359

NMDA⫹CLO 6 div [␮g/ml]

NMDA⫹CLO 0 div [␮g/ml]

[1000]

[100]

[10]

[100]

[10]

1.53 ⫾ 0.496* 158.1 ⫾ 29.03

4.98 ⫾ 0.628* 165.8 ⫾ 14.29

15.67 ⫾ 2.728* 203.5 ⫾ 14.52

1 ⫾ 0.262* 250.8 ⫾ 30.56*

2.9 ⫾ 1.136* 329.3 ⫾ 57.37*

Note. CTL, control OHSC; CLO 6 div, OHSC treated with clodronate from 6 div until 9 div; CLO 0 div, OHSC treated with clodronate from 0 div until 9 div; NMDA, lesioned OHSC; NMDA⫹CLO 6 div, lesioned OHSC treated with clodronate from 6 div until 9 div; NMDA⫹CLO 0 div, lesioned OHSC treated with clodronate from 0 div until 9 div; asterisks indicate statistically significant results (P ⬍ 0.01) between lesioned OHSC and lesioned OHSC that were treated with clodronate.

laser scanning microscopy, and quantitative morphometry proved that the number of damaged neurons increased dramatically after lesioning. PI-labeled nuclei were mainly located in the cellular layer of the DG, reflecting NMDAinduced neuronal injury of granule cells in the DG (Fig. 2D). In lesioned OHSC the number of PI⫹ nuclei in the GCL of the DG (125.8/GCL) was significantly higher than in control OHSC (15.06/GCL; P ⬍ 0.01; Fig. 2H; Table 1). Group C: treatment of lesioned OHSC with clodronate from 6 div until 9 div results in an inhibition of microglial activation Treatment of lesioned OHSC with clodronate resulted in a strong reduction of microglial cell numbers. The few remaining microglial cells were mainly located in the GCL of the DG, but some cells were present in the molecular layer of the DG and in the hilar area. The IB⫹ 4 microglial cells had a round or amoeboid morphology similar to the microglial cells observed in OHSC treated with NMDA alone (Fig. 2E). Confocal laser scanning microscopy and quantitative morphometry confirmed that the number of IB⫹ 4 microglial cells was significantly reduced when clodronate was added (1000 ␮g/ml: 1.53/GCL; 100 ␮g/ml: 4.98/GCL; 10 ␮g/ml: 15.67/GCL; P ⬍ 0.01 for all concentrations used; Fig. 2G; Table 1). The few remaining microglial cells displayed a round or amoeboid morphology (Fig. 2F). Treatment with clodronate did not prevent the NMDAinduced deterioration of the hippocampal cytoarchitecture. Hematoxylin staining showed many condensed nuclei ob-

viously belonging to damaged cells in the cellular layer of the DG (Fig. 2E). As shown by confocal laser scanning microscopy the number of PI⫹ nuclei in OHSC treated with three different concentrations of clodronate (1000 ␮g/ml: 158.1/GCL; 100 ␮g/ml: 165.8/GCL; 10 ␮g/ml: 203.5/GCL) did not significantly differ from the number in OHSC treated with NMDA alone (125.8/GCL; Fig. 2H; Table 1). The PI⫹ neurons were mainly located in the GCL of the DG, although single PI⫹ neurons were found in the hilar area or in the molecular layer. The latter may represent damaged interneurons (Fig. 2F). Group D: treatment of unlesioned OHSC with clodronate from 6 div until 9 div does not affect the number and distribution of microglial cells Unlesioned OHSC treated with clodronate from 6 until 9 div contained a very low number of IB⫹ 4 microglial cells (1000 ␮g/ml: 1.5/GCL; 100 ␮g/ml: 3.26/GCL; 10 ␮g/ml: 3.0/GCL; Table 1). At all concentrations tested the numbers of IB⫹ 4 microglial cells did not differ significantly from control OHSC (Table 1). The morphology and distribution of IB⫹ 4 microglial cells were also similar in clodronatetreated and control OHSC: these cells displayed a ramified morphology with multiple, branched cytoplasmic processes. Some cells possessed a few, extremely long cytoplasmic processes. In both clodronate-treated and control OHSC, microglial cells were mainly found in the noncellular layers of the DG; only a few cells were observed in the GCL of the DG.

Fig. 3. Continuous treatment of lesioned OHSC with clodronate (100 ␮g/ml) from 0 div until 9 div was even more potent in depleting microglial cells (A–D): ⫹ statistical analysis of the mean number of IB⫹ neurons (E, F). (A, B) NMDA. (C, D) NMDA ⫹ CLO 0 div: OHSC additionally 4 microglial cells and PI treated with clodronate showed massive neuronal cell death; however, very few amoeboid IB⫹ 4 microglial cells (green) were found. (E) Clodronate treatment from 0 div induced a significantly lower mean number of IB⫹ 4 microglial cells compared to OHSC treated with NMDA alone. (F) Treatment with clodronate from 0 div resulted in a significantly higher mean number of PI⫹ nuclei (red) when compared to OHSC treated with NMDA alone. HI indicates the hilar area; GCL indicates the granule cell layer; arrowheads show the borders between the GCL and the molecular layer of the DG; arrows point to IB⫹ 4 microglial cells, open arrowheads indicate degenerating neurons; scale bars in overview, 50 ␮m: scale bars in higher magnification, 10 ␮m; asterisks in E and F indicate statistical significance (P ⬍ 0.01) versus treatment with NMDA alone; error bars in E and F represent the SEM.

8

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

OHSC treated with 1000, 100, or 10 ␮g/ml clodronate did not show any significant difference in the number of PI⫹ neurons when compared to control OHSC (1000 ␮g/ml: 24.88/GCL; 100 ␮g/ml: 9.11/GCL; 10 ␮g/ml: 0/GCL; Table 1). Group E: treatment of lesioned OHSC with clodronate from 0 div to 9 div results in a nearly complete elimination of microglial cells and increased neuronal damage Microglial cells were virtually absent in lesioned OHSC that were treated with clodronate from 0 div until 9 div. The very few IB⫹ 4 microglial cells that were present showed an amoeboid morphology. The cells were randomly distributed in the hilar area and in the cellular and noncellular layers of the DG (Fig. 3C). Confocal laser scanning microscopy and quantitative morphometry showed a very low number of IB⫹ 4 microglial cells in clodronate-treated lesioned OHSC (100 ␮g/ml: 1.0/GCL: 10 ␮g/ml: 2.9/GCL: Fig. 3D and F; Table 1) when compared to lesioned OHSC that were not treated with clodronate (36.8/GCL). As seen in OHSC treated with NMDA alone many damaged neurons were observed in the GCL of lesioned OHSC treated with clodronate. The total number of viable neurons was reduced, resulting in cytoarchitectural destruction (Fig. 3C). Confocal laser scanning microscopy and quantitative morphometry showed that the number of PI⫹ nuclei increased significantly in lesioned clodronate-treated OHSC (100 ␮g/ml: 250.8/GCL; 10 ␮g/ml: 329.3/GCL; P ⬍ 0.01; Fig. 3D and F; Table 1). Group F: treatment of unlesioned OHSC with clodronate from 0 div until 9 div does not affect microglial cell numbers or neuronal preservation Application of clodronate to unlesioned OHSC did not influence the number of IB⫹ 4 microglial cells (100 ␮g/ml: 0.84/GCL; 10 ␮g/ml: 1.67/GCL) and of PI⫹ nuclei (100 ␮g/ml: 8.83/GCL; 10 ␮g/ml: 19.5/GCL), indicating an intact environment and an excellent neuronal preservation. The numbers were similar to those found in control OHSC (Table 1). Independent of the applied concentration of clodronate, microglial cells displayed the ramified morphology that is typical of the resting state.

Discussion The role of microglia is of special interest in neuronal injury and axonal regeneration. As a consequence of CNS lesions such as stroke or trauma, the phenotype of microglial cells changes from a ramified into an amoeboid form (Watanabe et al., 2000; Weisserbock et al., 2000). More importantly, activated microglial cells secrete a cocktail of factors that have been characterized as neurotoxic, such as

proinflammatory cytokines, NO (Bal-Price and Brown, 2001) and proteases (Lokensgard et al., 2001). Such activated microglial cells are accompanied by macrophages invading the CNS from the blood stream, and both cell types become virtually indistinguishable from one another. Several experimental approaches have therefore been undertaken to modulate microglial activation or macrophage invasion in CNS pathologies. The aim of this study was to elucidate the role of microglial cells in the early phase of excitotoxic neuronal injury and to investigate the effects of the bisphosphonate clodronate on microglial cells in a complex organotypic hippocampal slice culture. Former investigations indicated that clodronate has an inhibitory effect on cells of the monocytic lineage, such as peripheral macrophages (Selander et al., 1996; Van Rooijen et al., 1996). As the microglial cell population is also derived from the monocytic lineage and little is known about the effect of bisphosphonates on microglial cells, we investigated the effects of clodronate on this cell population and compared NMDA-lesioned OHSC with lesioned OHSC that received additional treatment with clodronate. The preservation of the hippocampal cytoarchitecture, the number of IB⫹ 4 microglial cells, and the number of degenerating neurons were analyzed. Although there are no data available on clodronate concentrations in the cerebrospinal fluid, the clodronate concentrations used in our in vitro experiments appear comparable to those recorded in patients’ serum. After an intravenous injection of clodronate patients showed peak serum concentrations of 10.1 ⫾ 2.8 ␮g/ml (Hanhijarvi et al., 1989). Virtually no IB⫹ 4 -activated microglial cells were detected in lesioned OHSC that were treated with clodronate after the lesion (from 6 div until 9 div), whereas OHSC treated with NMDA alone showed a robust activation of microglial cells and a significant increase in the number of PI⫹ nuclei. This indicates that clodronate potently reduces the number of activated microglial cells in lesioned OHSC. The ability of clodronate to deplete activated macrophages has also been observed by Schmidt-Weber et al. (1996), who reported that activated monocytes were more efficient in incorporating clodronate liposomes than resting cells and showed increased signs of apoptosis. The ablation of microglial cells in our model of OHSC containing only the intrinsic population of CNS macrophages indicates that clodronate exerts direct effects on activated microglial cells. The effects of clodronate treatment on microglial cells were previously unclear, as former experiments using in vivo lesion models did not allow one to differentiate the effects of clodronate on invading hematogenous macrophages from direct effects on microglial cells (Koennecke et al., 1999; Liu et al., 2000a, 2000b). Furthermore, the application of clodronate in vivo seemed to ablate only perivascular macrophages, but not microglial cells (Polfliet et al., 2001). The exact mechanism of action of clodronate still remains unclear, although RAW 264 macrophages seem to metabolize clodronate into an active

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

metabolite. These results indicate that clodronate acts as a prodrug (Monkkonen et al., 2001). Similar results on the effects of clodronate on microglial cells have not yet been shown. In contrast to our earlier findings and relevant data on the effects of clodronate in CNS lesion models (LazarovSpiegler et al., 1998; Van Rooijen et al., 1996), the present results suggest that microglial depletion does not always have a positive effect on neuronal preservation. After microglial depletion by the application of clodronate to lesioned OHSC from 6 div until 9 div, the number of PI⫹ nuclei was not significantly different from OHSC treated with NMDA alone. On the contrary, lesioned OHSC that received clodronate over the entire culture period contained even more PI⫹ nuclei than OHSC treated with NMDA alone. A neurotoxic effect of clodronate on its own can probably be excluded since the number of PI⫹ nuclei in unlesioned OHSC treated with clodronate was not significantly different from the number in control OHSC. Our earlier findings of improved neuronal survival in excitotoxic injury after the addition of astrocyte-derived, microglia-deactivating factors supported the hypothesis of increased neuronal survival after an inhibition of microglial activation: Treatment of OHSC with astrocyte-derived factors following an excitotoxic neuronal injury resulted in an inhibition of the microglial reaction and a concomitant decrease in neuronal damage (Hailer et al., 2001). Studies on isolated microglial cells showed that astrocytic factors induced a ramified morphology, reduced adhesion molecule expression and proinflammatory cytokine secretion, and changed the pattern of potassium channel expression (Eder et al., 1999; Hailer et al., 2001). The hypothesis that inhibition of microglial activation reduces neuronal injury was further supported by experiments with the immunosuppressant mycophenolate mofetil: This study showed that NMDA-induced neuronal injury was attenuated following treatment with mycophenolate mofetil, while the number of activated microglial cells in the GCL was significantly reduced. This effect seemed to be due to a powerful suppression of the glial cell proliferation, as the number of bromodesoxyuridine-labeled microglial and astroglial nuclei was reduced by 91.5% (Hischebeth et al., 2001). The observation of an increased number of PI⫹ nuclei and deteriorated hippocampal cytoarchitecture after the almost complete elimination of activated microglial cells points to the ambivalent role of macrophages or microglial cells after CNS injury. On the one hand, activated macrophages were shown to elicit neuroprotective and neuriteoutgrowth-promoting effects and to play a central role in tissue healing by phagocytosis of cell debris or myelin (Chan et al., 2001). On the other hand, activated macrophages and microglial cells are known to release a wide variety of cytotoxic substances such as the proinflammatory cytokine tumor necrosis factor (TNF)-␣, interleukin-1 (Eriksson et al., 2000; Giulian and Baker, 1986), and interleukin-6 as well as nitric oxide (NO) (Chao et al., 1992;

9

Possel et al., 2000). Thus, microglial cells are supposed to contribute to the process of secondary damage, which occurs after ischemic and traumatic lesions (Kim and Ko, 1998; Schwab and Bartholdi, 1996). Pharmacological suppression of macrophage and microglial activation has therefore been discussed as a therapeutic option to reduce the amount of secondary damage following CNS lesions. Both steroids and nonsteroidal anti-inflammatory agents have been shown to improve the functional outcome after spinal cord injury (Bethea et al., 1999; Bracken et al., 1997; Mabon et al., 2000; Oudega et al., 1999): In several in vivo experiments, reduction in neuronal or axonal damage correlated with a reduced number of infiltrating inflammatory cells (Blight, 1994; Popovich et al., 1999). Although the understanding of the cellular and molecular mechanisms underlying regeneration in the CNS has increased, it is still largely unclear whether and to what extent various nonneuronal populations support neuronal recovery, or whether some cell types even counteract the limited potential for CNS regeneration. In summary, our experiments show that clodronate is able to reduce the number of activated, amoeboid microglial cells after excitotoxic neuronal injury induced by NMDA. The abrogation of activated microglial cells by application of clodronate will be a useful tool in the investigation of neuron– glia interactions following CNS lesions, particularly with regard to the question of whether activated microglial cells can cause further neuronal damage or support neuronal regeneration.

Acknowledgments This study was supported by the Stiftung Friedrichsheim, the Manja und Ernst Mordhorst-Stiftung, the Paul und Ursula Klein-Stiftung, and the Medical Faculty of the Johann Wolfgang Goethe-University. The authors gratefully acknowledge the expert technical assistance by Mrs. Nadine Roser-Bloh and Mr. Shawn Leslie.

References Bal-Price, A., Brown, G.C., 2001. Inflammatory neurodegeneration mediated by nitric oxide from activated glia-inhibiting neuronal respiration, causing glutamate release and excitotoxicity. J. Neurosci. 21, 6480 – 6491. Benford, H.L., McGowan, N.W., Helfrich, M.H., Nuttall, M.E., Rogers, M.J., 2001. Visualization of bisphosphonate-induced caspase-3 activity in apoptotic osteoclasts in vitro. Bone 28, 465– 473. Bethea, J.R., Nagashima, H., Acosta, M.C., Briceno, C., Gomez, F., Marcillo, A.E., Loor, K., Green, J., Dietrich, W.D., 1999. Systemically administered interleukin-10 reduces tumor necrosis factor-alpha production and significantly improves functional recovery following traumatic spinal cord injury in rats. J. Neurotrauma 16, 851– 863. Blight, A.R., 1994. Effects of silica on the outcome from experimental spinal cord injury: implication of macrophages in secondary tissue damage. Neuroscience 60, 263–273. Bracken, M.B., Shepard, M.J., Holford, T.R., Leo-Summers, L., Aldrich, E.F., Fazl, M., Fehlings, M., Herr, D.L., Hitchon, P.W., Marshall, L.F.,

10

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11

Nockels, R.P., Pascale, V., Perot Jr., P.L., Piepmeier, J., Sonntag, V.K., Wagner, F., Wilberger, J.E., Winn, H.R., Young, W., 1997. Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the Third National Acute Spinal Cord Injury Randomized Controlled Trial. National Acute Spinal Cord Injury Study. JAMA 277, 1597–1604. Chan, A., Magnus, T., Gold, R., 2001. Phagocytosis of apoptotic inflammatory cells by microglia and modulation by different cytokines: mechanism for removal of apoptotic cells in the inflamed nervous system. Glia 33, 87–95. Chao, C.C., Hu, S., Molitor, T.W., Shaskan, E.G., Peterson, P.K., 1992. Activated microglia mediate neuronal cell injury via a nitric oxide mechanism. J. Immunol. 149, 2736 –2741. Csuka, E., Hans, V.H., Ammann, E., Trentz, O., Kossmann, T., MorgantiKossmann, M.C., 2000. Cell activation and inflammatory response following traumatic axonal injury in the rat. Neuroreport 11, 2587–2590. Czapiga, M., Colton, C.A., 1999. Function of microglia in organotypic slice cultures. J. Neurosci. Res. 56, 644 – 651. Eder, C., Schilling, T., Heinemann, U., Haas, D., Hailer, N.P., Nitsch, R., 1999. Morphological, immunophenotypical and electrophysiological properties of resting microglia in vitro. Eur. J. Neurosci. 11, 4251– 4261. Eriksson, C., Nobel, S., Winblad, B., Schultzberg, M., 2000. Expression of interleukin 1 alpha and beta, and interleukin 1 receptor antagonist mRNA in the rat central nervous system after peripheral administration of lipopolysaccharides. Cytokine 12, 423– 431. Frith, J.C., Monkkonen, J., Auriola, S., Monkkonen, H., Rogers, M.J., 2001. The molecular mechanism of action of the antiresorptive and antiinflammatory drug clodronate: evidence for the formation in vivo of a metabolite that inhibits bone resorption and causes osteoclast and macrophage apoptosis. Arthritis Rheum. 44, 2201–2210. Ga¨ hwiler, B.H., Capogna, M., Debanne, D., McKinney, R.A., Thompson, S.M., 1997. Organotypic slice cultures: a technique has come of age. Trends Neurosci. 20, 471– 477. Giulian, D., Baker, T.J., 1986. Characterization of amoeboid microglia isolated from developing mammalian brain. J. Neurosci. 6, 2163–2178. Glass, J.D., Wesselingh, S.L., 2001. Microglia in HIV-associated neurological diseases. Microsc. Res. Tech. 54, 95–105. Hailer, N.P., Ja¨ rhult, J.D., Nitsch, R., 1996. Resting microglial cells in vitro: analysis of morphology and adhesion molecule expression in organotypic hippocampal slice cultures. Glia 18, 319 –331. Hailer, N.P., Wirjatijasa, F., Roser, N., Hischebeth, G.T.R., Korf, H.W., Dehghani, F., 2001. Astrocytic factors protect neuronal integrity and reduce microglial activation in an in vitro model of N-methyl-D-aspartate-induced excitotoxic injury in organotypic hippocampal slice cultures. Eur. J. Neurosci. 14, 315–326. Hanhijarvi, H., Elomaa, I., Karlsson, M., Lauren, L., 1989. Pharmacokinetics of disodium clodronate after daily intravenous infusions during five consecutive days. Int. J. Clin. Pharmacol. Ther. Toxicol. 27, 602– 606. Heppner, F.L., Skutella, T., Hailer, N.P., Haas, D., Nitsch, R., 1998. Activated microglial cells migrate towards sites of excitotoxic neuronal injury inside organotypic hippocampal slice cultures. Eur. J. Neurosci. 10, 3284 –3290. Hischebeth, G.T.R., Dehghani, F., Hailer, N.P., 2001. In vitro neuroprotection and inhibition of glial activation by the immunosuppressant mycophenolate-mofetil. Akt Neurol. 28, 195. Kim, W.K., KO, K.H., 1998. Potentiation of N-methyl-D-aspartate-mediated neurotoxicity by immunostimulated murine microglia. J. Neurosci. Res. 54, 17–26. Koennecke, L.A., Zito, M.A., Proescholdt, M.G., Van Rooijen, N., Heyes, M.P., 1999. Depletion of systemic macrophages by liposome-encapsulated clodronate attenuates increases in brain quinolinic acid during CNS-localized and systemic immune activation. J. Neurochem. 73, 770 –779. Kreutzberg, G.W., 1996. Microglia: a sensor for pathological events in the CNS. Trends Neurosci. 19, 312–318.

Lazarov-Spiegler, O., Solomon, A.S., Schwartz, M., 1998. Peripheral nerve-stimulated macrophages simulate a peripheral nerve-like regenerative response in rat transected optic nerve. Glia 24, 329 –337. Liu, B., Du, L., Kong, L.Y., Hudson, P.M., Wilson, B.C., Chang, R.C., Abel, H.H., Hong, J.S., 2000a. Reduction by naloxone of lipopolysaccharide-induced neurotoxicity in mouse cortical neuron-glia co-cultures. Neuroscience 97, 749 –756. Liu, T., Van Rooijen, N., Tracey, D.J., 2000b. Depletion of macrophages reduces axonal degeneration and hyperalgesia following nerve injury. Pain 86, 25–32. Lokensgard, J.R., Hu, S., Sheng, W., Vanoijen, M., Cox, D., Cheeran, M.C., Peterson, P.K., 2001. Robust expression of TNF-alpha, IL-1beta, RANTES, and IP-10 by human microglial cells during nonproductive infection with herpes simplex virus. J. Neurovirol. 7, 208 –219. Mabon, P.J., Weaver, L.C., Dekaban, G.A., 2000. Inhibition of monocyte/ macrophage migration to a spinal cord injury site by an antibody to the integrin-␣D: a potential new anti-inflammatory treatment. Exp. Neurol. 166, 52– 64. Marx, F., Blasko, I., Pavelka, M., Grubeck-Loebenstein, B., 1998. The possible role of the immune system in Alzheimer’s disease. Exp. Gerontol. 33, 871– 881. Monkkonen, H., Rogers, M.J., Makkonen, N., Niva, S., Auriola, S., Monkkonen, J., 2001. The cellular uptake and metabolism of clodronate in RAW 264 macrophages. Pharm. Res. 18, 1550 –1555. Moore, S., Thanos, S., 1996. The concept of microglia in relation to central nervous system disease and regeneration. Prog. Neurobiol. 48, 441– 460. Noraberg, J., Kristensen, B.W., Zimmer, J., 1999. Markers for neuronal degeneration in organotypic slice cultures. Brain Res. Brain Res. Protoc. 3, 278 –290. Oudega, M., Vargas, C.G., Weber, A.B., Kleitman, N., Bunge, M.B., 1999. Long-term effects of methylprednisolone following transection of adult rat spinal cord. Eur. J. Neurosci. 11, 2453–2464. Polfliet, M.M., Goede, P.H., Kesteren-Hendrikx, E.M., Van Rooijen, N., Dijkstra, C.D., Van Den Berg, T.K., 2001. A method for the selective depletion of perivascular and meningeal macrophages in the central nervous system. J. Neuroimmunol. 116, 188 –195. Popovich, P.G., Guan, Z., Wei, P., Huitinga, I., Van Rooijen, N., Stokes, B.T., 1999. Depletion of hematogenous macrophages promotes partial hindlimb recovery and neuroanatomical repair after experimental spinal cord injury. Exp. Neurol. 158, 351–365. Possel, H., Noack, H., Putzke, J., Wolf, G., Sies, H., 2000. Selective upregulation of inducible nitric oxide synthase (iNOS) by lipopolysaccharide (LPS) and cytokines in microglia: in vitro and in vivo studies. Glia 32, 51–59. Pozzo-Miller, L.D., Mahanty, N.K., Connor, J.A., Landis, D.M., 1994. Spontaneous pyramidal cell death in organotypic slice cultures from rat hippocampus is prevented by glutamate receptor antagonists. Neuroscience 63, 471– 487. Raivich, G., Moreno Flores, M.T., Moller, J.C., Kreutzberg, G.W., 1994. Inhibition of posttraumatic microglial proliferation in a genetic model of macrophage colony-stimulating factor deficiency in the mouse. Eur. J. Neurosci. 6, 1615–1618. Rio-Hortega, P.D., 1932. Microglia, in: Penfield, W. (Ed.), Cytology and Cellular Pathology of the Nervous System, Paul B. Hoeber, New York, pp. 482–534. Russell, R.G., Rogers, M.J., 1999. Bisphosphonates: from the laboratory to the clinic and back again. Bone 25, 97–106. Schmidt-Weber, C.B., Rittig, M., Buchner, E., Hauser, I., Schmidt, I., Palombo-Kinne, E., Emmrich, F., Kinne, R.W., 1996. Apoptotic cell death in activated monocytes following incorporation of clodronateliposomes. J. Leukocyte Biol. 60, 230 –244. Schwab, M.E., Bartholdi, D., 1996. Degeneration and regeneration of axons in the lesioned spinal cord. Physiol. Rev. 76, 319 –370. Selander, K.S., Monkkonen, J., Karhukorpi, E.K., Harkonen, P., Hannuniemi, R., Vaananen, H.K., 1996. Characteristics of clodronate-induced apoptosis in osteoclasts and macrophages. Mol. Pharmacol. 50, 1127– 1138.

A. Kohl et al. / Experimental Neurology 181 (2003) 1–11 Streit, W.J., 1990. An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4). J. Histochem. Cytochem. 38, 1683–1686. Streit, W.J., Graeber, M.B., Kreutzberg, G.W., 1988. Functional plasticity of microglia: a review. Glia 1, 301–307. Streit, W.J., Kreutzberg, G.W., 1987. Lectin binding by resting and reactive microglia. J. Neurocytol. 16, 249 –260. Van Rooijen, N., Sanders, A., Van Den Berg, T.K., 1996. Apoptosis of macrophages induced by liposome-mediated intracellular delivery of clodronate and propamidine. J. Immunol. Methods 193, 93–99.

11

Watanabe, H., Abe, H., Takeuchi, S., Tanaka, R., 2000. Protective effect of microglial conditioning medium on neuronal damage induced by glutamate. Neurosci. Lett. 289, 53–56. Weissenbock, H., Hornig, M., Hickey, W.F., Lipkin, W.I., 2000. Microglial activation and neuronal apoptosis in Bornavirus infected neonatal Lewis rats. Brain Pathol. 10, 260 –272. Zito, M.A., Koennecke, L.A., McAuliffe, M.J., McNally, B., Van Rooijen, N., Heyes, M.P., 2001. Depletion of systemic macrophages by liposome-encapsulated clodronate attenuates striatal macrophage invasion and neurodegeneration following local endotoxin infusion in gerbils. Brain Res. 892, 13–26.