Isolation of Mouse Primary Microglia by Magnetic-Activated Cell Sorting in Animal Models of Demyelination

Summary

Here, we present a protocol to isolate and purify primary microglia in animal models of demyelinating diseases, utilizing columnar magnetic-activated cell sorting.

Abstract

Microglia, the resident innate immune cells in the brain, are the primary responders to inflammation or injury in the central nervous system (CNS). Microglia can be divided into resting state and activated state and can rapidly change state in response to the microenvironment of the brain. Microglia will be activated under different pathological conditions and exhibit different phenotypes. In addition, there are many different subgroups of activated microglia and great heterogeneity between different subgroups. The heterogeneity mainly depends on the molecular specificity of microglia. Studies have revealed that microglia will be activated and play an important role in the pathological process of inflammatory demyelination. To better understand the characteristics of microglia in inflammatory demyelinating diseases such as multiple sclerosis and neuromyelitis optica spectrum disorder, we propose a perilesional primary microglial sorting protocol. This protocol utilizes columnar magnetic-activated cell sorting (MACS) to obtain highly purified primary microglia and preserve the molecular characteristics of microglia to investigate the potential effects of microglia in inflammatory demyelinating diseases.

Introduction

Microglia originate from yolk-sac progenitors, which reach the embryonic brain very early and participate in the development of the CNS^1,2. For instance, they are involved in synaptic pruning³ and regulating axonal growth⁴. They secrete factors that promote neuronal survival and help neuronal localization⁵. At the same time, they are involved in removing abnormal cells and apoptotic cells to ensure normal brain development⁶. In addition, as the immune-competent cells of the brain, microglia continually surveil the brain parenchyma to clear dead cells, dysfunctional synapses, and cellular debris⁷. It has been demonstrated that microglial activation plays an important role in a variety of diseases, including inflammatory demyelinating diseases, neurodegenerative diseases, and brain tumors. Activated microglia in multiple sclerosis (MS) contribute to the differentiation of oligodendrocyte precursor cells (OPCs) and regeneration of myelin by engulfing myelin debris⁸.

In Alzheimer's disease (AD), accumulation of amyloid beta (Aβ) activates microglia, which affects the phagocytic and inflammatory functions of microglia⁹. Activated microglia in the glioma tissue, called glioma-associated microglia (GAM), can regulate the progression of glioma and ultimately affect the prognosis of patients¹⁰. The activation profoundly alters the microglial transcriptome, resulting in morphological changes, expression of immune receptors, increased phagocytic activity, and enhanced cytokine secretion¹¹. There are different subsets of activated microglia in neurodegenerative diseases such as disease-associated microglia (DAM), activated response microglia (ARM), and microglial neurodegenerative phenotype (MGnD)⁸.

Similarly, multiple dynamic functional subsets of microglia also coexist in the brain in inflammatory demyelinating diseases¹². Understanding the heterogeneity between different subsets of microglia is essential to investigate the pathogenesis of inflammatory demyelinating diseases and to find their potential therapeutic strategies. The heterogeneity of microglia mainly depends on the molecular specificity⁸. It is essential to describe the molecular alterations of microglia accurately for the study of the heterogeneity. Advances in single-cell RNA-sequencing (RNA-seq) technology have enabled the identification of the molecular characteristics of activated microglia in pathological conditions¹³. Therefore, the ability to isolate cell populations is critical for further investigation of these target cells under specific conditions.

Studies performed to understand the characteristics and functions of microglia are usually in vitro studies, since it has been found that large numbers of primary microglia can be prepared and cultured from mouse pup brains (1-3 days old), which attach to the culture flasks and grow on the plastic surface with other mixed glial cells. Subsequently, pure microglia can be isolated based on the different adhesiveness of mixed glial cells^14,15. However, this method can only isolate microglia from the perinatal brain and takes several weeks. Potential variables in cell culture may influence microglial characteristics such as molecular expression¹⁶. Moreover, microglia isolated by these methods can only participate in in vitro experiments by simulating the conditions of CNS diseases and cannot represent the characteristics and functions of microglia in in vivo disease states. Therefore, it is necessary to develop methods for isolating microglia from the adult mouse brain.

Fluorescence-activated cell sorting (FACS) and magnetic separation are two widely used methods, although they have their own different limitations^16,17,18,19. Their respective advantages and disadvantages will be contrasted in the discussion section. The maturation of MACS technology offers the possibility to rapidly purify cells. Huang et al. have developed a convenient method to label demyelinating lesions in brains²⁰. Combining these two technical approaches, we propose a rapid and efficient columnar CD11b magnetic separation protocol, providing a step-by-step description to isolate microglia around demyelinating lesions in adult mouse brains and preserve the molecular characteristics of microglia. The focal demyelinating lesions were caused by the stereotactic injection of 2 µL of lysolecithin solution (1% LPC in 0.9% NaCl) in the corpus callosum 3 days before starting the protocol²¹. This protocol lays the foundation to perform the next step in in vitro experiments. Moreover, this protocol saves time and remains feasible for widespread use in various experiments.

Protocol

All the animal procedures have been approved by the Institute of Animal Care Committee of Tongji Medical College, Huazhong University of Science and Technology, China.

1. Materials

1. Prepare the following solutions before beginning the protocol.
   1. Prepare the loading buffer by adding fetal bovine serum (FBS, 2%) to phosphate buffered saline (PBS).
   2. Add neutral red (NR) dye (final 1%) to PBS.
2. Prepare the following solutions using the commercially available Adult Brain Dissociation Kit (see the Table of Materials).
   1. To prepare enzyme mix 1, pipette 1.9 mL of Buffer Z and 50 µL of Enzyme P into a 15 mL centrifuge tube.
   2. To prepare enzyme mix 2, pipette 20 µL of Buffer Y and 10 µL of Enzyme A into a 1.5 mL microcentrifuge tube.
   3. Prepare the debris removal solution.

2. Mice perfusion and dissection

1. Intraperitoneally (i.p.) inject the mouse with 500 µL of 1% NR dye in PBS 2-3 h before anesthesia.
   NOTE: Intraperitoneal injection of NR in demyelinating mouse models helps discern the lesions (Figure 1).
2. Anesthetize the mouse with pentobarbital (50 mg/kg, i.p.), and ensure that the mouse is successfully anesthetized by examining the pain responses.
3. Open the thoracic cavity of the mouse and expose the heart.
4. Cut off a small corner of the right atrium carefully and intracardially perfuse the mouse with 20-30 mL of cold PBS from the left ventricle.
5. Decapitate the mouse under anesthesia.
6. Open the skull of the mouse and carefully free the brain from the skull.
7. Transfer the brain to the mouse brain slice mold and cut the brain into 0.1 cm slices.
8. Remove the brain slices and place them in cold PBS in a Petri dish (60/15 mm). Microdissect the lesions labeled by NR dye around the corpus callosum under a stereomicroscope using microsurgical forceps.
   NOTE: Ensure that the dissected tissue is as complete as possible to facilitate the subsequent transfer; the total dissection progress should be completed in 2 h.
9. Transfer the dissected tissue to 15 mL centrifuge tubes containing the appropriate amount of cold loading buffer.
   ​NOTE: Pool the corpus callosum tissue from 2-3 mice together in one tube as one sample.

3. Tissue dissociation

1. Centrifuge the dissected tissue at 300 ×g for 30 s (after reaching this speed) to collect the sample at the bottom of the tube.
2. Preheat enzyme mix 1 and enzyme mix 2 to 37 °C in an incubator.
3. Add 1,950 µL of preheated enzyme mix 1 to one sample and digest in an incubator at 37 °C for 5 min.
4. Add 30 µL of preheated enzyme mix 2 and mix gently.
5. Digest in an incubator at 37 °C for 15 min and mix gently every 5 min.
6. Add 4 mL of cold PBS to the tube after digestion and shake gently.
7. Place 70 µm filters on 50 mL centrifuge tubes and prewet the filters with 500 µL of cold PBS. Filter the dissociated tissue by passing the digested tissue samples through the filters, and transfer the filtered suspension in the 50 mL centrifuge tube to a 15 mL centrifuge tube.
   NOTE: Skip this step if the amount of tissue is too small.
8. Centrifuge the filtered tissue samples at 300 ×g for 10 min at 4 °C and aspirate the supernatant slowly and completely.
   NOTE: All the steps should be performed on ice except for the enzymatic digestion.

4. Debris removal

1. Resuspend the cell pellet gently with 1,550 µL of cold PBS.
2. Add 450 µL of cold solution for debris removal and mix well.
3. Overlay the above mixture very slowly and gently with 2 x 1 mL of cold PBS using a 1,000 µL pipette.
   NOTE: Tilt the centrifuge tube by 45 ° and slowly add PBS along the tube wall with the pipette.
4. Transfer the tube slowly and gently to a centrifuge and spin at 4 °C and 3,000 × g for 10 min.
5. Look for three layers after centrifugation. Aspirate the two top layers completely by using a 1,000 µL pipette (Figure 2).
6. Fill the tube up to 5 mL with cold loading buffer and gently invert the tube three times.
7. Centrifuge at 4 °C and 1,000 × g for 10 min. Aspirate the supernatant completely and avoid disrupting the cell pellet.

5. Magnetic separation of microglial cells

1. Resuspend the cell pellet with 90 µL of loading buffer and add 10 µL of CD11b (Microglia) beads.
2. Mix well and incubate for 15 min at 4 °C.
3. Add 1 mL of loading buffer and pipette the liquid gently up and down with a 1,000 µL pipette to wash the cells after the incubation. Centrifuge the cells at 4 °C and 300 × g for 10 min and aspirate the supernatant completely to remove the unbound beads.
4. Resuspend the cells in 500 µL of loading buffer.
5. Place the MS column with its separator for positive selection in the magnetic field.
6. Rinse the column with 500 µL of loading buffer to protect the cells and ensure the efficiency of magnetic sorting based on the protocol of the manufacturer.
7. Apply the cell suspension onto the MS column and discard the flowthrough containing unlabeled cells.
8. Add 500 µL of loading buffer to the column three times to wash away cells that adhere to the column wall and discard the flowthrough.
   NOTE: Only add new loading buffer for washing when the column reservoir is completely empty.
9. Remove the column from the separator after washing the column and place it on a 15 mL centrifuge tube.
10. Add 1 mL of loading buffer into the column and push the plunger to the bottom of the column to flush out the magnetically labeled cells. Repeat this step three times to completely collect the magnetically labeled cells.

Representative Results

Microglia isolated using CD11b beads have high purity
Microglial cells around the lesions in demyelination mouse models were isolated using the above-mentioned protocol and tested by flow cytometry. Cells are fluorescently labeled with CD11b-fluorescein isothiocyanate (FITC) and CD45-allophycocyanin (APC) to determine microglia in flow cytometry according to the manufacturer's instructions. There are multiple literatures demonstrating that CD11b and CD45 antibodies are enough to check for the purity of isolated microglia^19,22. The fluorescence compensation was completed on the flow cytometer using unstained and single-stained control samples. Cells were gated by forward scatter-height (FSC-H) and side scatter-height (SSC-H); single cells were gated by forward scatter-area (FSC-A) and FSC-H. The proportions of microglia (identified as CD11b⁺CD45^Intermediate) before and after the magnetic separation were tested by flow cytometry. The back-loop technique was used to adjust the gating strategy of flow cytometry analysis by determining the gate for the microglia. Compared with approximately 3 × 10⁴ cells and 6 × 10³ microglia before magnetic separation of each mouse brain (Figure 3A), MACS generally yields approximately 2.5 × 10³ cells and 2 × 10³ microglia per mouse brain (Figure 3B), which was approximately 85% of all single cells. However, nearly 3% of myeloid cells (identified as CD11b⁺CD45^high) were present in the MACS-separated cells and were difficult to remove from the CD11b population in this protocol. The purity of the isolated microglia can be increased to 95% using two magnetic columns (Supplemental Figure S1). MACS sorting could capture approximately 33% of microglia in the brain mixture sample.

Microglia isolated using CD11b beads have high cell viability
Fluorochrome 7-aminoactinomycin D (7-AAD) is often used to show the cell viability in the flow cytometry. 7-AAD⁺ cells represent dead cells¹⁹. The viability of MACS-separated microglia was approximately 95% by cell staining with 7-AAD (Figure 4).

Morphological analysis of microglia around demyelinating lesions sorted by MACS
Morphological analysis showed that microglia were not activated by MACS. Microglia exhibit a typical longitudinal bipolar cell body after MACS, which is similar to microglial morphology after FACS¹⁹. Microglia around demyelinating lesions will be activated in the model of LPC-induced demyelination²¹. Microglia stained by ionized calcium binding adaptor molecule 1 (Iba-1) showed typical amebic morphology of activated microglia in this protocol. P2ry12 is a typical molecule of microglial resting state. There is a small fraction of branched and P2ry12⁺ microglia before MACS due to the expansion of the extent of the dissociated lesions. However, the existence of P2ry12⁺ microglia after MACS also indicates that MACS will not activate microglia (Supplemental Figure S2).

Figure 1: Images of the focal demyelinating lesions labeled by the neutral red dye. Please click here to view a larger version of this figure.

Figure 2: Images of the three layers formed by centrifugation. The top two layers (layer 1 + layer 2) need to be removed in the debris removal process (see protocol step 4.5). Please click here to view a larger version of this figure.

Figure 3: Flow cytometry analysis of microglia isolated from demyelination lesions of adult mice before and after MACS. (A) Schematic representation of gating strategy used in flow cytometry analysis before MACS: FSC-H and SSC-H for selecting cells, FSC-A and FSC-H for selecting single cells, and CD11b-FITC and CD45-APC for selecting myeloid cells (up) and microglia (down). (B) Schematic representation of the gating strategy used in flow cytometry analysis sample after MACS. The proportion of microglia has increased significantly after MACS: FSC-H and SSC-H for selecting cells, FSC-A and FSC-H for selecting single cells, and CD11b-FITC and CD45-APC for selecting myeloid cells (up) and microglia (down). Abbreviations: SSC-H = side scatter-height; FSC-H = forward scatter-height; FSC-A = forward scatter-area; CD45-APC = allophycocyanin-labeled CD45; CD11b-FITC = fluorescein isothiocyanate-labeled CD11b; MACS = magnetic-activated cell sorting. Please click here to view a larger version of this figure.

Figure 4: FACS gating strategy for living/dead single cells after MACS. 7-AAD and SSC-H for selecting living cells after MACS. Abbreviations: SSC-H = side scatter-height; 7-AAD = 7-aminoactinomycin D; MACS = magnetic-activated cell sorting; FACS = fluorescence-activated cell sorting. Please click here to view a larger version of this figure.

Supplemental Figure S1: Flow cytometry analysis of microglia isolated from adult demyelination mice. (A) Schematic representation of the gating strategy used in flow cytometry analysis sample before MACS. (B) Schematic representation of the gating strategy used in flow cytometry analysis sample after MACS using two magnetic columns. (C) Single-parameter histogram of CD11b-FITC and CD45-APC in flow cytometry after MACS using two magnetic columns. Abbreviations: SSC-H = side scatter-height; FSC-H = forward scatter-height; FSC-A = forward scatter-area; CD45-APC = allophycocyanin-labeled CD45; CD11b-FITC = fluorescein isothiocyanate-labeled CD11b; MACS = magnetic-activated cell sorting; CD45^Int = intermediate CD45 expression. Please click here to download this File.

Supplemental Figure S2: Immunofluorescence staining of microglia with Iba-1 and P2ry12 before and after MACS. (A) Images of microglia before MACS. (B) Images of isolated microglia after MACS. Scale bars = 20 µm. Abbreviations: MACS = magnetic-activated cell sorting; Iba-1 = ionized calcium binding adaptor molecule 1; DAPI = 4',6-diamidino-2-phenylindole. Please click here to download this File.

Discussion

The protocol proposes a method to isolate microglia around the demyelinating lesions, which can help study the functional characteristics of microglia in inflammatory demyelinating diseases. Microglia captured using CD11b beads exhibit high purity and viability. Critical steps in the protocol include the precise localization of foci and optimal microglial purification. In protocol step 2.1, it is necessary to inject the NR solution 2 h before sacrificing the mouse to ensure that the lesions can be accurately displayed²⁰. In protocol step 2.8, care was taken to place brain tissue sections on an ice box for dissection and complete this step as soon as possible, thereby improving cell viability. In step 2.9, it is necessary to select appropriate tissue as one sample for both enzyme digestion and debris removal. Step 3.7 cannot be skipped when the effect of debris removal is not satisfactory. In step 4.3, the addition of PBS must be gentle to form different layers, ensuring that as much debris is removed as possible. All the steps in the protocol, except enzymatic digestion, should be performed on ice to ensure cell viability and reduce the transcriptional response of the microglia. Each step in the protocol is mild and retains high viability, so it is unnecessary to use the dead cells removal solution^22,23.

The enzymatic dissociation method for mouse brain in the protocol has proven suitable for single-cell RNA-seq²³. Although the enzymes in the protocol are mild, Marsh et al. have found that various enzymatic hydrolysis protocols at 37 °C may significantly alter the transcriptome of microglia²⁴. Setting a healthy control in the experiment for comparison can help to identify the differentially expressed genes of demyelination, thus eliminating the abnormal transcriptional response in the protocol. Meanwhile, toprevent this abnormal transcriptional state after enzymatic digestion, the protocol can be optimized by adding various transcription and translation inhibitors such as actinomycin D, triptolide and anisomycin at different steps. The use of these inhibitors can reveal the real gene expression of microglia in vivo without the setting of a control. Therefore, the addition of these inhibitors to the protocol can reduce the number of animals or specimens required in the experiment and is more economical for downstream analysis. However, it should be noted that some genes of microglia will be modulated by these inhibitors, which may limit the utilization of these inhibitors for future work²⁴. Using the specialized dissociator instrument after the addition of enzyme mix 1 and enzyme mix 2 for tissue dissociation can also limit transcriptional responses in protocol step 3²³.

This protocol has some limitations. The purity of microglia is only approximately 85% in the protocol, which fails to comprise more than 90% of all single cells similar to other literatures^19,22. The replacement of LS column with MS column, the absence of the filters, and the quality of microbeads may be the potential causes. If the purity of the isolated microglia fails to meet the requirements, prolonging the magnetic incubation time or passing the sorted cells through a second magnetic column will improve purity (Supplemental Figure S1). If the number of cells is low, tissue should be harvested from more mice and cell loss due to careless aspiration avoided. If the cell viability is low, replacing PBS in the loading buffer with RPMI 1640 medium could significantly improve the cell viability. It is also beneficial for cell viability to ensure that all the solutions used in the protocol are fresh, and the protocol is completed as soon as possible. It is important for users to balance the purity and the number of microglia obtained using this protocol.

The protocol utilizes the binding of microglia with CD11b beads and the enrichment in a magnetic field to purify the microglia. Therefore, purified microglia contain the contamination of myeloid cells, which is inevitable in this protocol. In contrast to FACS-the gold standard for sorting cells-the purity of microglia isolated by MACS is not sufficiently high for some elaborate applications such as deep sequencing, which limits its widespread usage. In addition, FACS for sorting microglia can eliminate the contamination of myeloid cells. However, MACS is a gentler method, retains more original morphological characteristics of glial cells especially astrocytes, takes less time for isolating microglia, and has a higher cell yield than FACS, suggesting that it can reduce the number of mice to be sacrificed and may be more suitable for time-sensitive experiments^19,25. Meanwhile, MACS is more cost-effective and widely available and has lower technical requirements for experimenters. MACS has been proven the most reliable and consistent method to isolate microglia. Morphological analysis showed that microglia would not be activated by MACS, and RNA-seq has demonstrated that microglia isolated by this method maintain a resting state^19,26.

Overall, the protocol is important for the survey of microglia in demyelinating diseases, and microglia isolated by this protocol can be utilized for various downstream applications such as quantitative RT-PCR and RNA-seq. The protocol utilizes neutral red dye to locate demyelinating lesions, which ensures the accuracy of lesion locations and reduces any confusion with normal tissue during sampling. Accurate localization of lesions can help focus on the effects of diseases and study the unique molecular phenotype and functional characteristics of microglia in demyelinating diseases. This will provide a basis for studying the mechanism of microglia in demyelinating diseases. This method of isolating microglia using magnetic CD11b beads can purify microglia with high cell viability and yield in a short time, with minimal requirements for experimental conditions, which allows widespread usage for studying microglia in different conditions.

Materials

Name	Company	Catalog Number	Comments
1.5 mL Micro Centrifuge Tubes	BIOFIL	CFT001015
15 mL Centrifuge Tubes	BIOFIL	CFT011150
50 mL Centrifuge Tubes	BIOFIL	CFT011500
70 µm Filter	Miltenyi Biotec	130-095-823
Adult Brain Dissociation Kit, mouse and rat	Miltenyi Biotec	130-107-677
C57BL/6J Mice	SJA Labs
CD11b (Microglia) Beads, human and mouse	Miltenyi Biotec	130-093-634
Fetal Bovine Serum	BOSTER	PYG0001
FlowJo	BD Biosciences	V10
MACS MultiStand	Miltenyi Biotec	130-042-303
MiniMACS Separator	Miltenyi Biotec	130-042-102
MS columns	Miltenyi Biotec	130-042-201
Neutral Red	Sigma-Aldrich	1013690025
NovoCyte Flow Cytometer	Agilent		A system consisting of various parts
NovoExpress	Agilent	1.4.1
PBS	BOSTER	PYG0021
Pentobarbital	Sigma-Aldrich	P-010
Stereomicroscope	MshOt	MZ62