This study uses flow cytometry and two different gating strategies on isolated perfused mice brain choroid plexuses; this protocol identifies the main immune cell subsets that populate this brain structure.
The brain is no longer considered as an organ functioning in isolation; accumulating evidence suggests that changes in the peripheral immune system can indirectly shape brain function. At the interface between the brain and the systemic circulation, the choroid plexuses (CP), which constitute the blood-cerebrospinal fluid barrier, have been highlighted as a key site of periphery-to-brain communication. CP produce the cerebrospinal fluid, neurotrophic factors, and signaling molecules that can shape brain homeostasis. CP are also an active immunological niche. In contrast to the brain parenchyma, which is populated mainly by microglia under physiological conditions, the heterogeneity of CP immune cells recapitulates the diversity found in other peripheral organs. The CP immune cell diversity and activity change with aging, stress, and disease and modulate the activity of the CP epithelium, thereby indirectly shaping brain function. The goal of this protocol is to isolate murine CP and identify about 90% of the main immune subsets that populate them. This method is a tool to characterize CP immune cells and understand their function in orchestrating periphery-to-brain communication. The proposed protocol may help decipher how CP immune cells indirectly modulate brain function in health and across various disease conditions.
Since the discovery of the blood-brain barrier by Paul Erhlich in the late 19th century, the brain has been considered virtually separated from the other organs and the bloodstream. Yet, this last decade has seen the emergence of the concept that brain function is shaped by various biological factors, such as gut microbiota and systemic immune cells and signals1,2,3,4. In parallel, other brain borders such as meninges and choroid plexuses (CP) have been identified as interfaces of active immune-brain cross talk rather than inert barrier tissues5,6,7,8.
The CP constitute the blood-cerebrospinal fluid barrier, one of the borders separating the brain and the periphery. They are located in each of the four ventricles of the brain, i.e., the third, the fourth, and both lateral ventricles, and are adjacent to areas involved in neurogenesis such as the subventricular zone and subgranular zone of the hippocampus3. Structurally, the CP are composed of a network of fenestrated blood capillaries enclosed by a monolayer of epithelial cells, which are interconnected by tight and adherens junctions9,10. Major physiological roles of the CP epithelium involve the production of cerebrospinal fluid, which flushes the brain from waste metabolites and protein aggregates, and the production and controlled blood-to-brain passage of various signaling molecules including hormones and neurotrophic factors11,12,13. Secreted molecules from CP shape brain's activity, i.e., by modulating neurogenesis and microglial function14,15,16,17,18,19, which makes CP crucial for brain homeostasis. CP also engage in various immune activities; whereas the main immune cell type in the brain parenchyma under non-pathological conditions is microglia, the diversity of CP immune cell populations is as broad as in peripheral organs3,7, suggesting that various channels of immune regulation and signaling are at work at the CP.
The space between the endothelial and epithelial cells, the CP stroma, is mainly populated by border-associated macrophages (BAM), which express pro-inflammatory cytokines and molecules related to antigen presentation in response to inflammatory signals3. Another subtype of macrophages, Kolmer's epiplexus cells, are present on the apical surface of the CP epithelium20. CP stroma is also a niche for dendritic cells, B cells, mast cells, basophils, neutrophils, innate lymphoid cells, and T cells which are mostly effector memory T cells able to recognize central nervous system antigens7,21,22,23,24. In addition, the composition and activity of immune cell populations at the CP changes upon systemic or brain perturbation, for example, during aging10,14,15,21,25, microbiota perturbation7, stress26, and disease27,28. Notably, these changes were suggested to indirectly shape brain function, i.e., a shift of CP CD4+ T cells towards Th2 inflammation occurs in brain aging and triggers immune signaling from the CP that may shape aging-associated cognitive decline14,15,21,25,29. Illuminating the properties of the CP immune cells would thus be crucial to better understand their regulatory function on CP epithelium physiology and secretion and thereby decipher their indirect impact on brain function under healthy and disease conditions.
CP are small structures that contain only a few immune cells. Their isolation requires microdissection after a preliminary step of perfusion; immune cells in the bloodstream would otherwise constitute major contaminants. This protocol aims to characterize the myeloid and T cell subsets of the CP using flow cytometry. This method identifies about 90% of the immune cell populations that compose mouse CP under non-inflammatory conditions, in accordance with recently published works using other methods to dissect immune CP heterogeneity7,10,28. This protocol could be applied to characterize changes in the CP immune cell compartment with disease and other experimental paradigms in vivo.
All the procedures agreed with the guidelines of the European Commission for the handling of laboratory animals, directive 86/609/EEC. They were approved by the ethical committees No. 59, by the CETEA/CEEA No. 089, under the number dap210067 and APAFIS #32382-2021070917055505 v1.
1. Preparation of the materials
2. Housing of C57BL/6 mice
3. PBS perfusion and brain dissection
4. Dissection of the Choroid Plexus from the brain
5. Preparation of samples for flow cytometry analysis
6. Flow cytometry
NOTE: The flow cytometer used in this protocol is equipped with the following 5 lasers: a 355 nm UV laser, a 405 nm Violet laser, a 488 nm Blue laser, a 561 nm Yellow-Green laser and a 637 nm Red laser.
7. CP myeloid cells gating
8. CP T cells gating
The flow cytometry analyses presented here successfully revealed the major subsets of myeloid and T cells (Figure 1 and Figure 2, respectively), and their relative total number per mouse in a highly reproducible manner (Figure 3).
The flow cytometry analysis of myeloid cells showed that CP are populated by CD11b+ CX3CR1+ F4/80high BAM, representing almost 80% of the CD45+ immune cells at the CP. These BAM were divided into two different populations according to their expression of MHC-II: the IA-IE+ BAM that constituted the major group (72% of the CD45+ immune cells), in line with previously published data7, and the IA-IE– BAM. The CD11bhigh F4/80intermediate CX3CR1+ IA-IE– population of Kolmer's epiplexus macrophages7,2 was also identified and only represented about 1.2% of the CD45+ CP immune cells, consistently with the literature7. Among the CD11b+ F4/80– immune cells, two different populations of Ly6C– IA-IE+ cells were characterized that likely correspond to dendritic cells, and that can be divided into CD11bhigh CX3CR1low and CD11b+ CX3CR1– populations which constituted around 3.1% and 3.7% of the CD45+ CP immune cells, respectively. CD45+ CD11b+ F4/80– IA-IE– Ly6C+ cell population including both monocytes and neutrophils constituted only 0.5% of CD45+ immune cells in accordance with previous reports and attesting for the high efficacy of the PBS perfusion3. These results are largely consistent with previously published data assessing CP immune populations by single-cell RNA sequencing7.
The analysis and gating strategy of T cells highlighted the presence of the minor population of CD45+ CD11b– TCRβ+ cells. Among them, about 30-50 CD8– CD4+ and CD8+ CD4– T cells per mouse populated CP under physiological conditions, representing 1.3% and 0.9% of the CD45+ CP immune cells, respectively. CD4+ cells were slightly more abundant than CD8+ T cells, which is consistent with previous results in terms of both number and proportion21,30. This gating strategy also revealed a population of CD45+ CD11b– TCRβ+ CD4– CD8– cells, which may include NKT cells, and a recently described regulatory T cell subset10,28,31. The proposed flow cytometry analysis and gating strategies allowed the cell type annotation of almost 90% of the immune cells (CD45+) that compose the mouse CP.
Figure 1: Flow cytometry analysis and gating strategy of myeloid cells from CP of perfused mice. Cells are gated based on size using FSC-A vs. SSC-A. Singlet cells are selected using FSC-A vs. FSC-H. Live cells are gated by excluding DAPI+ cells on DAPI vs. FSC-A plot. CD45+ immune cells are then identified on a CD45 vs. FSC-A plot. Gating CD11b vs. F4/80, the CD11b+ F4/80+ and the CD11b+ F4/80– populations are selected. Gating CD11b+ F4/80+ population using CX3CR1 vs. IA-IE, both the CD11b+ F4/80hi CX3CR1+ IA-IE+ BAM and the CX3CR1+ IA-IE– populations are determined. This latter population is then gated using F4/80 vs. CD11b to better separate CD11b+ F4/80high CX3CR1+ IA-IE– BAM and CD11bhigh F4/80intermediate CX3CR1+ IA-IE– Kolmer's epiplexus macrophages. Gating Ly6C vs. IA-IE on CD11b+ F4/80–, both CD11b+ F4/80– IA-IE– Ly6C+ cells that mainly correspond to monocytes and neutrophils and CD11b+ F4/80– IA-IE+ Ly6C– cells that likely represent dendritic cells population are selected. CD11b+ F4/80– IA-IE+ Ly6C– cells are gated using CD11b vs. CX3CR1 allowing identification of CD11bhigh CX3CR1low and CD11b+ CX3CR1– cells. If not specified, values below each gated population represent the frequency of the parent population. hi: high; int: intermediate; lo: low. Please click here to view a larger version of this figure.
Figure 2: Flow cytometry analysis and gating strategy of T cells from CP of perfused mice. Cells are gated based on size using FSC-A vs. SSC-A. Single cells are selected using FSC-A vs. FSC-H. Live cells are identified excluding DAPI+ cells on DAPI vs. FSC-A plot. CD45+ immune cells are then gated on a CD45 vs. FSC-A plot. Gating CD11b vs. TCRβ, CD11b– TCRβ+ are selected. Gating CD4 vs. CD8, the proportions of CD8– CD4+ and CD4– CD8+ T cells are determined. If not specified, values below each gated population represent the frequency of the parent population. Please click here to view a larger version of this figure.
Figure 3: Numbers of the major immune cell subsets at the CP per mouse. Mean of the total number of cells per mouse of the different immune subsets identified in the four CP pooled together. Representation: Mean + standard deviation. Please click here to view a larger version of this figure.
Studies aiming to understand the immunological contributions to brain homeostasis and disease have mainly focused on cells residing within the brain parenchyma, neglecting brain borders such as CP, which are nevertheless crucial contributors to brain function2,3. The analysis of immune cell populations at CP is challenging due to the small size of CP, low numbers of resident immune cells, and complicated access to this tissue. Flow cytometry performed on total brain immune cells (CD45+) does not allow the characterization of the rare immune populations that reside in the CP. To characterize the immune composition of the CP with high precision, flow cytometry analyses were conducted on CP dissected from PBS-perfused mice. This approach allows the exclusion of circulating CD45+ cells and CP CD45– cells such as pericytes, endothelial and epithelial cells32. The gating strategies focusing on myeloid and T cell populations identify the main CP immune subsets.
A critical step of the current protocol is the PBS perfusion. As the CP immune cells are rare, blood contamination could completely mask their local heterogeneity. Liver and brain discoloration is always checked as control, and therefore blood contamination is minimal. Consistently, CD11b+ F4/80– Ly6C+, which includes both monocytes and neutrophils present in the blood rather than the brain in non-inflammatory conditions3, accounted for only 0.5% of the CD45+ cells, demonstrating a high efficacy of PBS perfusion. Another helpful control for PBS perfusion efficacy might be the incorporation of antibodies against Ter119, a marker specific for erythrocytes. Finally, an additional analysis of the CP from PBS-perfused mouse brain sections by immunostaining and microscopy may also be useful to verify whether any blood immune cells that may be strongly adherent to the endothelium have resisted the PBS perfusion and remain attached at the lumen of the vasculature.
The myeloid gating strategy revealed two populations of macrophages: CD11b+ F4/80high CX3CR1+ IA-IE+/-BAM and CD11b+ F4/80intermediate CX3CR1+ IA-IE– Kolmer's epiplexus cells, which express high levels of CD11b and low levels of F4/807. The gates of Kolmer's epiplexus cells7,20 and IA-IE– BAM were not clearly separated using CD11b and F4/80 markers. Incorporation of the CD206 marker into this gating strategy may help to better differentiate CD206+ IA-IE– BAM from CD206low Kolmer's epiplexus cells7.
Almost 7% of the CP immune cells appeared to be CD11b+ F4/80– Ly6C– IA-IE+. This subset was composed of two populations according to their levels of CD11b and CX3CR1 expression. These cells likely correspond to dendritic cells, but this remains to be confirmed by analyzing their CD11c expression either by FACS3 or immunofluorescence33. The CD11b+ F4/80– Ly6C– IA-IE+ population also likely includes mast cells. Incorporating antibodies for the detection of CD117 and c-kit into the gating strategy would help to distinguish them from CD11c+ dendritic cells3,34. The proportion of CD45+ CD11b+ F4/80+ Ly6C+ cells among CP immune cells was determined. However, they include both monocytes and neutrophils, and the addition of the Ly6G marker to the gating strategy of CD45+ CD11b+ F4/80+ Ly6C+ cells would determine the ratio of Ly6G– monocytes to Ly6G+ neutrophils. Beyond the described CD45+ CD11b– TCRβ+ population, the gating strategy on CD45+ CD11b– populations may be enriched using NK1.1 and CD11c markers to identify NK cells, TCRγδ to analyze γδ T cells, CD19 and/or B220 to characterize B cells, and CD138 to identify plasma cells3. Expression of reporter genes can also be included in this analysis. For example, using Foxp3-GFP transgenic mice could be helpful in identifying CP-resident T regulatory (Treg) cells3,30. Lastly, this protocol can also incorporate intracellular staining of transcription factors or cytokines21.
Whereas the immune cells of the brain parenchyma have been extensively studied, much less is known about immune cells populating brains barriers, such as the CP3,35,36. CP in mice are small tissues located in four different areas of the brain, and their isolation requires a quite complex microdissection. Because of this, the CP were mostly studied using imaging. Staining of CP immune cells from whole CP dissected tissue or from brain sections followed by microscopic analysis provided major advances in the understanding of CP immune mechanisms14,15,21,22,26,27,29,30,33. However, this method does not allow for extensive analysis of the CP immune cell diversity as only four to six markers can be analyzed at a time. In addition, some subpopulations such as CD4+ or CD8+ T cells are quite rare at the CP, and quantitative analysis of their phenotypes with microscopy would be challenging. The flow cytometry approach allows for the analysis of tens of markers in parallel for every immune cell, providing a better-suited tool for quantitative assessment of the CP immune composition.
More recently, single nuclei transcriptomic technology (snRNA-seq) came out as a powerful method to study CP heterogeneity10,28. However, it is not well-suited for analyzing CP immune cells as they constitute only a small fraction of the CP cells. Since snRNA-seq samples the cells at the proportion they are present in the tissue, the immune cells are underrepresented in the snRNA-seq, and their low numbers pose an obstacle for in-depth analysis10,28.
Single-cell transcriptomics (scRNA-seq) has been recently applied to precisely map the immune cell heterogeneity of isolated CP immune cells7. While scRNA-seq is quite complex and costly to perform, flow cytometry may serve as a more accessible way to survey immune subpopulations at the CP. Lastly, this protocol can serve as part of a cell sorting protocol, where the selected subtypes of cells are sorted using FACS for analysis with molecular methods, such as scRNA-seq or bulk RNA analysis.
The authors have nothing to disclose.
We thank the Institut Pasteur Animalerie Centrale and the CB-UTechS facility members for their help. This work was supported financially by Institut Pasteur.
anti-mouse CD16/CD32 | BD Biosciences | 553142 | Flow cytometry antibody |
Albumin, bovine | MP Biomedicals | 160069 | Blocking reagent |
APC anti-mouse CX3CR1 | BioLegend | 149008 | Flow cytometry antibody |
APC anti-mouse TCRb | BioLegend | 109212 | Flow cytometry antibody |
APC-Cy7 anti-mouse CD4 | BioLegend | 100414 | Flow cytometry antibody |
APC-Cy7 anti-mouse IA-IE | BioLegend | 107628 | Flow cytometry antibody |
BD FACSymphony A5 Cell Analyzer | BD Biosciences | Flow cytometry analyzer | |
BV711 anti-mouse Ly6C | BioLegend | 128037 | Flow cytometry antibody |
Collagenase IV | Gibco | 17104-019 | Enzyme to dissociate CP tissue |
DAPI | Thermo Scientific | 62248 | Live/dead marker |
EDTA | Ion chelator | ||
fine scissors | FST | 14058-11 | Dissection tool |
FITC anti-mouse CD45 | BioLegend | 103108 | Flow cytometry antibody |
Flow controller infusion inset | CareFusion | RG-3-C | Blood perfusion inset |
FlowJo software | BD Biosciences | Analysis software | |
forceps | FST | 11018-12 | Dissection tool |
Heparin | Sigma-Aldrich | H3149-10KU | Anticoagulant |
Imalgene | Boehringer Ingelheim | Ketamine, anesthesic | |
OneComp eBeads | Invitrogen | 01-1111-42 | Control beads to realize compensation |
PBS-/- | Gibco | 14190-094 | Buffer |
PBS+/+ | Gibco | 14040-091 | Buffer |
PE anti-mouse CD8a | BioLegend | 100708 | Flow cytometry antibody |
PE anti-mouse F4/80 | BioLegend | 123110 | Flow cytometry antibody |
PE-Dazzle 594 anti-mouse CD11b | BioLegend | 101256 | Flow cytometry antibody |
Rompun | Bayer | Xylazine, anesthesic | |
thin forceps | Dumoxel Biology | 11242-40 | Dissection tool |
Vetergesic | Ceva | Buprenorphin, analgesic |