This protocol presents an in vitro live-imaging phagocytosis assay to measure the phagocytic capacity of astrocytes. Purified rat astrocytes and microglia are used along with pH indicator-conjugated synaptosomes. This method can detect real-time engulfment and degradation kinetics and provides a suitable screening platform to identify factors modulating astrocyte phagocytosis.
Astrocytes are the major cell type in the brain and directly contact synapses and blood vessels. Although microglial cells have been considered the major immune cells and only phagocytes in the brain, recent studies have shown that astrocytes also participate in various phagocytic processes, such as developmental synapse elimination and clearance of amyloid beta plaques in Alzheimer's disease (AD). Despite these findings, the efficiency of astrocyte engulfment and degradation of their targets is unclear compared with that of microglia. This lack of information is mostly due to the lack of an assay system in which the kinetics of astrocyte- and microglia-mediated phagocytosis are easily comparable. To achieve this goal, we have developed a long-term live-imaging in vitro phagocytosis assay to evaluate the phagocytic capacity of purified astrocytes and microglia. In this assay, real-time detection of engulfment and degradation is possible using pH indicator-conjugated synaptosomes, which emit bright red fluorescence in acidic organelles, such as lysosomes. Our novel assay provides simple and effective detection of phagocytosis through live-imaging. In addition, this in vitro phagocytosis assay can be used as a screening platform to identify chemicals and compounds that can enhance or inhibit the phagocytic capacity of astrocytes. As synaptic pruning malfunction and pathogenic protein accumulation have been shown to cause mental disorders or neurodegenerative diseases, chemicals and compounds that modulate the phagocytic capacity of glial cells should be helpful in treating various neurological disorders.
Glial cells, which refer to non-excitable cells in the brain, are the major cell type in the central nervous system (CNS). Previously, glial cells were regarded as mere supporting cells that mainly play passive roles in maintaining neuronal survival and basal synaptic properties. However, emerging evidence has revealed that glial cells play more active roles in various aspects of neurobiology, such as maintaining brain homeostasis, mediating synapse formation1,2,3 and synapse elimination4,5, and modulating synaptic plasticity6,7. Glial cells in the CNS include astrocytes, microglia, and oligodendrocytes. Among these cells, astrocytes and microglia have been shown to play phagocytic roles by engulfing synapses4,5, apoptotic cells8, neural debris9, and pathogenic proteins, such as amyloid beta plaques10,11. In the developing brain, astrocytes eliminate synapses in the dorsal lateral geniculate nucleus (dLGN) through MERTK- and MEGF10-dependent phagocytosis4. Similarly, microglia also eliminate C1q-coated synapses during developmental stages through the classical complement cascade5. Interestingly, it has been suggested that defects in synapse pruning can be one of the initiators of several neurological disorders. For example, it has been shown that mutations in complement component 4 (C4), which increases complement-mediated synapse pruning by microglia, are strongly associated with the prevalence of schizophrenia in humans12. A recent paper has also shown that the classical complement pathway is hyperactivated in the initiation stages of AD and induces early synapse loss in this disease13.
Compared with microglia-mediated phagocytosis, whether astrocyte-mediated phagocytosis contributes to the initiation and progression of various neurological disorders is less clear. However, a recent paper suggests that factors that alter the rate of normal synapse pruning by astrocytes may disrupt brain homeostasis and contribute to AD susceptibility and pathology14. The rate of synapse pruning by astrocytes is powerfully controlled by ApoE isomers, with a protective allele for AD (ApoE2) strongly enhancing the rate and a risk allele for AD (ApoE4) significantly lowering the rate. Moreover, transgenic mice expressing ApoE4 accumulated much more synaptic C1q than control or ApoE2 mice14. These data suggest that impaired astrocyte-mediated phagocytosis in the early AD brain may induce the accumulation of senescent C1q-coated synapses/synaptic debris that activates complement-mediated microglial phagocytosis, driving synaptic degeneration. The impaired phagocytic capacity of astrocytes in ApoE4 carriers may also contribute to the uncontrolled accumulation of amyloid beta plaques in AD-affected brains.
In addition, it has been shown that glial cells in the aged Drosophila brain lose their phagocytic capacity due to decreased translation of Draper, a homolog of Megf10 that astrocytes use for phagocytosing synapses. Restoring Draper levels rescued the phagocytic capacity of glial cells, which efficiently cleared damaged axonal debris in the aged brain to a similar extent as that in the young brain, indicating that aging-induced alterations in the phagocytic capacity of astrocytes may contribute to disruption of brain homeostasis15.
Based on these new findings, modulating the phagocytic capacity of astrocytes may be an attractive therapeutic strategy to prevent and treat various neurological disorders. In this regard, there have been several attempts to enhance the phagocytic capacity of astrocytes, for example, by inducing acidification of lysosomes with acidic nanoparticles16 and overexpressing transcription factor EB (TFEB), which can enhance lysosome biogenesis17. Despite these attempts, it is still unclear how astrocytes and microglial cells differ in their phagocytic kinetics and whether we should increase or decrease their phagocytic capacities in various diseases.
In this paper, we present a novel in vitro assay for detecting the phagocytic capacity of astrocytes in real-time. The data show different kinetics of engulfment and degradation in astrocytes and microglia. Astrocyte-conditioned medium (ACM), which contains secreted factors from astrocytes, is essential for effective phagocytosis of both astrocytes and microglia. Furthermore, Megf10, a phagocytic receptor in astrocytes and a homolog of Ced-1 and Draper, plays critical roles in astrocyte-mediated phagocytosis8,18.
All methods described here have been approved by The Korea Advanced Institute of Science and Technology Institutional Animal Care and Use Committee (IACUC), KA2016-08.
1. Synaptosome Purification
NOTE: These procedures are adapted from a previously published paper19 with several modifications to improve the yield of purified synaptosomes (Figure 1).
2. pH Indicator Conjugation
3. Astrocytes and Microglia Purification
NOTE: This protocol for astrocyte purification is adapted from a previously published paper21. In this protocol, the purification of rat astrocytes and microglia is described. Specific procedures for purifying mouse astrocytes and microglia are also described in the notes.
4. Collect IP-ACM
5. Phagocytosis Live-imaging Assay (Figure 2)
In this in vitro phagocytosis assay with long-term live-imaging, we used synaptosomes from adult mouse brain homogenates, which were separated in the gradient solution between 23% gradient solution and 10% gradient solution by ultracentrifugation (Figure 3). After the preparation, synaptosomes exposed PS in their outer membrane (Figure 4), suggesting that they lost their function and could be recognized by PS receptors in astrocytes and microglia. As shown in Figure 5, pH indicator-conjugated synaptosomes emitted bright red fluorescence when they were engulfed by astrocytes. Real-time comparison of the engulfment and degradation capacities of glial cells is possible by taking multiple images of a ROI every 1 or 2 h (Supplementary Movie 1, Supplementary Movie 2). With this method, we demonstrated different kinetics of astrocyte- and microglia-mediated phagocytosis (Figure 6). To quantify phagocytosis of glial cells, the area (µm2) of red fluorescence signal was measured, which is referred to phagocytic index in Figure 6B and Figure 7. Although astrocytes appeared to be efficient in phagocytosing large amounts of pH indicator-conjugated synaptosomes, microglia were faster at engulfing and degrading synaptosomes (Figure 6B). Microglia showed maximum pH indicator intensity at 26 h after pH indicator-conjugated synaptosome treatment, whereas astrocytes showed their maximum at 45 h (p-value < 0.05, two-way ANOVA between astrocyte with 1X ACM and microglia with 1X ACM). Likewise, microglial cells showed a 20.7% reduction in total pH indicator intensity 40 h after the peak point, whereas astrocytes showed a 17% reduction in total pH indicator intensity during the same time period. Interestingly, our data showed that astrocyte-secreted factors, which were contained in ACM, are essential for increasing both astrocyte- and microglia-mediated phagocytosis (Figure 6B). Astrocytes have been shown to release bridging molecules, such as MFGE8, GAS6, and ProteinS, which can bridge and induce interactions between phagocytic receptors and "eat-me" signals such as PS23. As mentioned above, astrocytes eliminate synapses via the MERTK and MEGF10 pathways4. MEGF10 is only expressed by astrocytes in the brain and participates in synapse engulfment through recognizing "eat-me" signals with unknown identities. In agreement with previous findings, the assay showed that compared with wild-type (WT) mouse astrocytes, Megf10 knock-out (KO) mouse astrocytes possessed significantly impaired phagocytic capacity (Figure 7). Compared with WT astrocytes, Megf10 KO astrocytes showed an approximately 40% reduction in total pH indicator intensity at the peak point (31 h) (Figure 7).
Figure 1. Schematic of astrocyte purification using immunopanning methods. Please click here to view a larger version of this figure.
Figure 2. Schematic of phagocytosis live-imaging assay. Please click here to view a larger version of this figure.
Figure 3. Representative brain homogenate fractionation in the gradient solutions. Please click here to view a larger version of this figure.
Figure 4. Representative images of PS-exposed synaptosomes. (A) A bright field image of astrocytes with synaptosomes. Synaptosomes are attached to astrocytes. (B) A fluorescent image of tdTomato-positive synaptosomes, which are purified from tdTomato-expressing mouse brains. (C) pSIVA binds to PS, which is exposed to outer membrane of synaptosome, and emits green fluorescence. (D) PS detected by pSIVA (green) is co-localized with tdTomato-positive synaptosomes (red). Scale bar: 20 µm Please click here to view a larger version of this figure.
Figure 5. Representative bright field and fluorescent images of astrocytes with pH indicator-conjugated synaptosomes at two time points. At 11 h after treatment (bottom panel), pH indicator-conjugated synaptosomes are engulfed by astrocytes and emit red fluorescence whereas they do not at 1 h after treatment (upper panel). Scale bar: 50 µm. Please click here to view a larger version of this figure.
Figure 6. Phagocytic kinetics of rat astrocytes and microglia through long-term live-imaging. (A) A schematic diagram of an in vitro phagocytosis assay using purified astrocytes and microglia along with pH indicator-conjugated synaptosomes. Note that before live images are taken, unbound synaptosomes are washed away after 40 min of incubation. (B) Representative graphs showing the engulfment and degradation kinetics of astrocytes and microglia. ACM, which contains astrocyte-secreted factors, significantly enhances both astrocyte- and microglia-mediated synaptosome uptake. Astrocyte: Control vs. Astrocyte with ACM 1X, ****, Tukey's multiple comparisons test. Microglia: Control vs. Microglia with ACM 1X, ****, Tukey's multiple comparison test. Error bars indicate S.E.M. *p ≤ 0.05, **** p ≤ 0.0001, two-way ANOVA. Please click here to view a larger version of this figure.
Figure 7. Decreased phagocytic capacity of Megf10 KO mouse astrocytes with 1X ACM compared with that of WT mouse astrocytes with 1X ACM. WT vs. Megf10 KO astrocytes with 1X ACM, ****, two-way ANOVA. Error bars indicate S.E.M. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 two-way ANOVA. Please click here to view a larger version of this figure.
Table 1: Solution recipes. Please click here to download this file.
Supplemental Movie 1. A representative live-imaging video showing the phagocytosis of pH indicator-conjugated synaptosomes by astrocytes. Please click here to download this file.
Supplemental Movie 2. A representative live-imaging video showing phagocytosis of pH indicator-conjugated synaptosomes by microglial cells. Please click here to download this file.
In this article, we present methods for a long-term live-imaging in vitro phagocytosis assay using purified glial cells and pH indicator-conjugated synaptosomes. We show that compared with microglia, astrocytes possess different engulfment and degradation capacity during the phagocytosis of synaptosomes. In addition, our data suggest that astrocyte-secreted factors, which contain bridging molecules such as GAS6, ProteinS, and MEGE8, are essential for efficient PS-dependent phagocytosis of glial cells in the brain. Furthermore, Megf10 KO astrocytes show defective phagocytosis of synaptosomes.
To successfully perform this experiment, the media and DPBS need to be carefully exchanged during the washing steps (Section 5). Primary cultured astrocytes and microglia, especially microglia, are vulnerable to air exposure. Therefore, reducing time intervals between washing steps is critical for maintaining live cells. For setting up long-term live-imaging with live-imaging instruments, manual focus is recommended since auto focus may increase laser/LED exposure time, which could damage the cells.
In this method, the synaptosome purification protocol19 is slightly modified. The original paper separates gradient solutions into 3%, 10%, 15%, and 23%. However, we set up 3%, 10% and 23% gradient solutions to increase the yield of purified synaptosomes. The instructions for immunopanning astrocytes and microglial cells are also modified in this method. The goal of the original panning protocol21 is to purify only astrocytes and use BSL-1, secondary-only, CD45, O4, and Itgb5 (HepaCAM for mouse) as the immunopanning order. Since BSL-1 and secondary-only plates will remove endothelial cells as well as microglia, we changed the order to CD45, secondary-only, BSL-1, O4, and Itgb5 (HepaCAM for mouse) to increase the yield of microglia from the CD45 panning dish. In this protocol, we also perfused SD rat pups (~ P7-P10) with DPBS through the circulatory system to remove various blood cells to minimize contamination from CD45-positive non-microglial populations.
There are several advantages of the in vitro phagocytosis assay. The pH indicator used in synaptosome conjugation only emits red fluorescence in low pH conditions. Therefore, when processing data, we can directly quantify red fluorescence as a signal of phagocytic events without a quenching procedure to remove background signals or complicated imaging analysis to obtain co-localized signals of engulfed material inside of phagocytes. In addition, this method allows real-time tracking of glial phagocytosis by taking images of pH indicator intensity within a given ROI at multiple time points. By generating a graph with pH indicator intensity up to 100 h, the engulfment as well as degradation capacities can be easily monitored between different cells and conditions. Another advantage is the use of synaptosomes. Since astrocytes ensheathe synapses all the time with their fine processes and have been shown to engulf synapses in vivo4, using synaptosomes is very suitable for measuring the in vitro phagocytic capacity of astrocytes. Finally, this in vitro phagocytosis assay has the potential to be developed to study glial phagocytosis of myelin debris or amyloid beta, which is related to various neurological disorders.
With increased life expectancy, a dramatic increase in the number of patients with neurodegenerative diseases is inevitable. Synapse loss through glial cells is one of the leading factors in several neurodegenerative diseases13,14. In addition, abnormal synapse pruning and an imbalance of brain homeostasis can be associated with glial phagocytosis defects. Therefore, identifying factors and compounds that can control the engulfment and degradation capacities of astrocytes could be critical for finding successful treatments for various neurological disorders. Since this in vitro phagocytosis assay can be easily scaled up with multiwell plates, this method can be used as a suitable platform for various screenings to identify such factors.
The authors have nothing to disclose.
The authors thank Yeon-Joo Jung for her experimental support during synaptosome purification and Jungjoo Park for images of synaptosomes with PS exposure. In addition, we thank all members in Chung's laboratory for helpful discussion. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (NRF-2016M3C7A1905391 and NRF-2016R1C1B3006969) (W.-S. C).
Synaptosome purification | |||
Percoll | GE healthcare life sciences | 17-0891-01 | |
Quick Start Bradford Protein Assay Kit 2 | BIO-RAD | 5000202 | |
pH indicator conjugation | |||
Dimethyl sulfoxide(DMSO) | LPS solution | DMSO100 | |
pHrodo red, succinimidyl ester | Molecula probes | P36600 | |
Immunopanning | |||
10X Earle’s balanced salt solution (EBSS) | Sigma | E7510 | |
Bovine serum albumin | Bovogen | BSA025 | |
Deoxyrebonuclease 1 (DNase) | Worthington | Is002007 | |
(DMEM) | Gibco | 11960-044 | |
(dPBS) | Welgene | LB001-02 | |
Fetal bovine serum (FBS) | Gibco | 16000-044 | |
Griffonia Simplicifolia Lectin(BSL-1) | Vector Labs | L-1100 | |
Goat anti-mouse IgG+IgM(H+L) | Jackson ImmunoResearch | 115-005-044 | |
Goat anti-mouse IgM (μ-chain) | Jackson ImmunoResearch | 115-005-020 | |
Heparin-binding epidermal growth factor | Sigma | E4643 | |
Human HepaCAM antibody | R&D systems | MAB4108 | |
Integrin beta 5 monoclonal antibody (KN52) | eBioscience | 14-0497-82 | |
L-cysteine | Sigma | C7880 | |
L-glutamate | Gibco | 25030-081 | |
N-acetly-L-cyteine (NAC) | Sigma | A8199 | |
Neurobasal media | Gibco | 21103-049 | |
O4 hybridoma supernatant(mouse IgM) | Bansal et al.23 | ||
Papain | Worthington | Is003126 | |
Penicillin/streptomycin | Gibco | 15140-122 | |
Pluristrainer 20 μm | PluriSelect | 43-50020-03 | |
Poly-D-lysine | Sigma | P6407 | |
Progesterone | Sigma | P8783 | |
Putrescine dihydrochloride | Sigma | P5780 | |
Purified rat anti-mouse CD45 | BD Pharmingen | 550539 | |
Purified mouse anti-rat CD45 | BD Pharmingen | 554875 | |
Sodium pyruvate | Gibco | 11360-070 | |
Sodium selenite | Sigma | S5261 | |
Transferrin | Sigma | T1147 | |
Trypsin | Sigma | T9935 | |
Trypsin inhibitor | Worthington | LS003086 | |
Ultra-clear tube (Tube, Thinwall, Ultra-Clear) | Beckman Coulter | 344059 | |
Collect IP-ACM | |||
Macrosep Advance Centrifugal Devices with Omega Membrane (10k) | PALL | MAP010C37 | |
Macrosep Advance Centrifugal Devices with Omega Membrane (30k) | PALL | MAP030C37 | |
Phagocytosis live imaging assay | |||
Juli stage | NanoEntek | ||
Time Series Analyzer V3 plugins | https://imagej.nih.gov/ij/plugins/time-series.html |