Neuroscience Letters 416 (2007) 87–91
Minocycline does not affect amyloid  phagocytosis by human microglial cells Atoosa Familian a,b,∗ , Piet Eikelenboom a,e , Robert Veerhuis a,b,c,d a
Department of Psychiatry, VU University Medical Center, Amsterdam, The Netherlands Department of Pathology, VU University Medical Center, Amsterdam, The Netherlands c Departments of Clinical Chemistry, VU University Medical Center, Amsterdam, The Netherlands d Alzheimer center, VU University Medical Center, Amsterdam, The Netherlands e Department of Neurology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands b
Received 26 October 2006; received in revised form 23 January 2007; accepted 23 January 2007
Abstract Activated microglia accumulate in amyloid  (A) plaques containing amyloid associated factors SAP and C1q in Alzheimer’s disease (AD) brain. Microglia are involved in AD pathogenesis by promoting A plaque formation and production of pro-inflammatory cytokines. On the other hand, phagocytosis of A by activated microglia may prevent A-mediated neurotoxicity and A plaque formation. Minocycline, a tetracycline derivative, is neuroprotective in various neurodegenerative models as well as human chronic neurological disorders. Minocycline attenuates the release of TNF-␣ by human microglia upon exposure to a mixture of A, SAP and C1q. Here, we demonstrate that minocycline down-regulates the production of pro-inflammatory cytokines by human microglia without affecting their beneficial activity, phagocytosis of amyloid  fibrils. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Minocycline; Microglia; Phagocytosis; Amyloid ; SAP; C1q
Clusters of activated microglia can be found in fibrillar amyloid  plaques in early stages of Alzheimer’s disease (AD) [3,23]. These activated microglia have been implicated in AD pathogenesis by promoting A plaque formation [25], as well as release of a broad range of chemokines, pro-inflammatory cytokines, and neurotoxic agents such as reactive oxygen/nitrogen species [21,11], which eventually may lead to local inflammatory processes including chemotactic response, complement activation, and cellular toxicity [1]. On the other hand, microglial cells as phagocytes of the central nervous system are involved in the phagocytosis of A. In culture, microglial cells can remove soluble A from the medium as well as A attached to the surface of culture dishes [18,2]. Also in animal studies, microglia were found to exert phagocytic activities leading to uptake of A [26]. Microglia express receptors proposed to be involved in the binding and uptake of A, including scavenger receptors [12,7]. Therefore, ∗ Corresponding author at: Department of Pathology, VU University Medical Center, P.O. Box 7057, 1007 MB Amsterdam, The Netherlands. Tel.: +31 20 444 4096; fax.: +31 20 444 2964. E-mail address:
[email protected] (A. Familian).
0304-3940/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2007.01.052
phagocytic activities of microglia may be essential for degradation of A plaques [15,4]. In a previous study, we reported that activated (MHC-class II positive) microglia in AD plaques co-localize with serum amyloid P component (SAP) and complement factor C1q [23]. These amyloid associated proteins enhance A fibril formation and the release of pro-inflammatory cytokines by human microglia in vitro, substantiating that the A-mediated microglial activation is dependent on the degree of A fibril formation [8,22]. Minocycline, a semisynthetic derivative of tetracycline with anti-inflammatory properties, acts as a neuroprotective drug in a variety of neurodegenerative models [17]. Recent human trials also pointed to beneficial effects of minocycline on chronic neurological disorders, such as amyotrophic lateral sclerosis, Huntington’s disease, and multiple sclerosis [10,19,14]. Possible mechanisms for neuroprotective activities of minocycline include inhibition of microglial activation and proliferation [20] and inhibition of caspase dependent and independent neuronal cell death through blocking the release of pro-apoptotic factors by mitochondria, or through up-regulation of the anti-apoptotic protein, Bcl-2 [17,6]. Furthermore, minocycline, like other tetracyclines, prevents cross- sheet
88
A. Familian et al. / Neuroscience Letters 416 (2007) 87–91
formation of A in vitro [9]. Recently, we reported that minocycline dose-dependently inhibits human microglial activation at concentrations between 250 nM and 25 M, whereas inhibitory effects on A fibril formation were observed only at concentrations of 25 M or higher [8]. In a recent study with amyloid precursor protein transgenic mice (APP-tg), minocycline was shown to improve cognitive performance and induced a small increase in A deposition in the hippocampus of young mice [16]. This study suggests that suppression of microglial activation in early stages of the AD process may lead to an increase in A deposition due to the reduced capacity of microglia to phagocytize or degrade A. In the present study we investigated if minocycline in the dosage used to prevent secretion of pro-inflammatory cytokines affects the phagocytic capacity of the microglia. A1–42 (Bachem, Bubendorf, Switzerland) dissolved (250 M) in cold hexafluoroisopropanol (HFIP, Fluka) was shaken overnight at room temperature, aliquoted, dried using speed vacuum for 30 min, and stored at −20 ◦ C until use. Fluo-A was prepared by mixing unlabeled A1–42 and fluorescein conjugated A1–42 (NEN Life Science products, Boston, MA) at a ratio of 65:1. Serum amyloid P component (SAP) was obtained from Calbiochem (La Jolla, CA). Human C1q was isolated from the Cohn I fraction of pooled human plasma, as described before [23]. Human microglia were isolated from subcortical white matter specimens of AD or control cases, as described [23,5]. Human brain specimens with a short post mortem delay time were obtained at autopsy through the Netherlands Brain Bank (coordinator Dr. R. Ravid). Human monocyte-derived macrophage (MDM) were isolated from buffycoats using a Percoll gradient followed by adherence to plastic in culture flasks (Costar, Corning, NY). Isolated microglial cells were cultured in a mixture (1/1, v/v) of Dulbecco’s modified Eagle’s medium and the nutrient mixture HAM-F10 (DMEM/HamF10), MDM in RPMI-1640 (all from GIBCO Life Technologies, Breda, The Netherlands), at 37 ◦ C, 5% CO2 for 7–10 days before experiments. Culture media contained l-glutamine (GIBCO Life Technologies), sodium-penicillin G (100 IU/ml), streptomycin sulfate (50 g/ml; both from Sigma, St. Louis, MO), 10% (v/v) fetal calf serum (FCS) (ICN Biomedicals, Amsterdam, The Netherlands), and in addition recombinant GM-CSF (50 ng/ml; Leucomax, Sandoz, The Netherlands). For stimulation and phagocytosis experiments, isolated MDM or microglial cells were detached using trypsin (Sigma, St. Louis, MO), transferred to 24-well plates (Costar) at cell densities of 7 × 104 cell/well for MDM and 5–7 × 104 for microglia, and allowed to settle for 24 h. Fluo-A1–42 peptides at a final concentration of 25 M, alone or in combination with SAP and C1q, were incubated overnight at room temperature in FCSfree culture medium to allow fibril formation in the absence or presence of minocycline-HCl (Sigma, St. Louis. MO). The formation of A fibrils was confirmed with electron microscopy. Pre-incubated mixtures were diluted 2.5 times with culture medium containing FCS and added to cultured cells to be incubated at 37 ◦ C, 5% CO2 for 24 h. Final concentrations of A, SAP, C1q, and FCS were 10 M, 85 nM, 5 nM, and 0.1%, v/v,
respectively, and 25 M, 2.5 M, and 250 nM for minocycline. Cytochalasin B (Sigma, St. Louis., MO) was used as a specific inhibitor of phagocytosis. Cell free culture supernatants were collected, and stored at −20 ◦ C, until assayed for the presence of cytokines. Next, the MDM or microglial cells were detached using trypsin, washed extensively with cold PBS containing 0.25% BSA, and immediately placed on ice before analysis by flow cytometry (FACSCalibur, Becton Dickinson). A minimum of 5000 cells were accepted for FACS analysis. Cells were gated based on morphological characteristics in forward and side scatter using CellQuestTM software (Beckton Dickinson). The degree of phagocytosis of A was expressed as the percentage of Fluo-A positive cells (cells that had taken up Fluo-A) by setting a gate in such way that 1% of medium treated cells were fluorescence-positive in every experiment. Levels of IL-6 and TNF-␣ in cell culture supernatants were measured in sandwich enzyme immunoassays (ELISA), according to the manufacturer’s (Sanquin, Amsterdam, The Netherlands) instructions. Detection limits of the ELISA’s were 0.2 and 1 pg/ml for IL-6 and TNF-␣, respectively. SPSS (release 9.0.1) for Windows was used for statistical analysis. Differences in degree of phagocytosis by microglia or MDM in control and treated groups were determined by repeated measures analysis of variance (ANOVA). Bonferroni post hoc analysis was performed to evaluate the effect of minocycline at different concentrations. p values
