The choroid plexus acts as an immune cell reservoir and brain entry site in experimental autoimmune encephalomyelitis

Abstract

The choroid plexus (ChP) has been suggested as an alternative central nervous system (CNS) entry site for CCR6⁺ Th17 cells during the initiation of experimental autoimmune encephalomyelitis (EAE), an animal model for multiple sclerosis (MS). To advance our understanding of the importance of the ChP in orchestrating CNS immune cell entry during neuroinflammation, we here directly compared the accumulation of CD45⁺ immune cell subsets in the ChP, the brain and spinal cord at different stages of EAE by flow cytometry. We found that the ChP harbors high numbers of CD45^int resident innate but also of CD45^hi adaptive immune cell subsets including CCR6⁺ Th17 cells. With the exception to tissue-resident myeloid cells and B cells, numbers of CD45⁺ immune cells and specifically of CD4⁺ T cells increased in the ChP prior to EAE onset and remained elevated while declining in brain and spinal cord during chronic disease. Increased numbers of ChP immune cells preceded their increase in the cerebrospinal fluid (CSF). Th17 but also other CD4⁺ effector T-cell subsets could migrate from the basolateral to the apical side of the blood-cerebrospinal fluid barrier (BCSFB) in vitro, however, diapedesis of effector Th cells including that of Th17 cells did not require interaction of CCR6 with BCSFB derived CCL20. Our data underscore the important role of the ChP as CNS immune cell entry site in the context of autoimmune neuroinflammation.

Keywords: CCL20; Choroid plexus; Experimental autoimmune encephalomyelitis; T cell trafficking; Th17 cells.

Fig. 1

CD45⁺ immune cell subsets change after induction of aEAE in brain, spinal cord, and choroid plexus. Immune cells were isolated from the choroid plexus (ChP) of the fourth and both lateral ventricles, brains, and spinal cords (SC) of healthy C57BL/6J mice and C57BL/6J mice suffering from aEAE at the different disease time points as outlined in Supplementary Fig. 1 and analyzed by multi-color flow cytometry for CD45⁺ immune cell subsets. (A) Representative dot plots (SSC versus CD45) show the ungated events acquired from ChP, brain and SC and the gating strategy for CD45⁺ immune cells. Gate bordered by the red line depicts 100% of CD45⁺ immune cells, which is divided by the black line into two gates for CD45^int and CD45^hi immune cells subsets. Numbers within the gates depict the percentages of CD45^int and CD45^hi immune cells, respectively. (B) Quantification of the percentages of CD45^int and CD45^hi immune cell subsets in the ChP, brain and SC during EAE. (C) Absolute numbers of CD45⁺ as well as CD45^int and CD45^hi immune cell subsets per mouse in the ChP, brain and SC in healthy C57BL/6J mice and at the investigated time points during aEAE are shown. The graphs in B and C show means ± SEM of six independent EAE experiments as outlined in Supplementary Fig. 1B. Significant differences between ChP and SC or ChP and brain are shown with brown and black stars, respectively. Statistical analysis: one-way ANOVA (p < 0.05 = *, p < 0.01 = **, p < 0.001 = ***, p < 0.0001 = ****)

Fig. 2

The choroid plexus harbors high numbers of CD4⁺ and CD8⁺ T cells as well as CD4⁺CCR6⁺ T cells when compared to brain and spinal cord. (A) Absolute numbers of CD45^hi CD4⁺ and CD45^hi CD8⁺ T cells per mouse in the choroid plexus (ChP), brain and spinal cord (SC) of healthy mice and mice during progression of aEAE as determined by multi-color flow cytometry. The graphs show means ± SEM of six independent experiments as outlined in Supplementary Fig. 1. (B) Contour plots for CD4 and CCR6 of CD45^hi immune cells isolated from ChP, brain and SC are shown. Numbers depict percentages of cells detected in the respective quadrant. Data are representative for two experiments (healthy, pre-clinical, before onset) and three experiments (peak, chronic and late chronic stages). Significant differences between ChP and SC or ChP and brain are shown in brown and black stars, respectively. Statistical analysis: one-way ANOVA (p < 0.05 = *, p < 0.01 = **, p < 0.001 = ***, p < 0.0001 = ****)

Fig. 3

The composition of the immune cells in choroid plexus, brain, and spinal cord during aEAE progression. Absolute numbers of B220⁺ B cells (A) and CD11b⁺ myeloid subsets including Ly6G⁺ neutrophils, Ly6C^hi monocytes (B), CD11c⁺ dendritic cells and F4/80⁺ macrophages (C) per mouse in the choroid plexus (ChP), brain and spinal cord (SC) of healthy mice and mice suffering from aEAE, were acquired by flow cytometry as shown. The graphs show means ± SEM of six independent experiments. Significant differences between ChP and SC or ChP and brain are shown in brown and black stars, respectively. Statistical analysis: (A-C) one-way ANOVA (p < 0.05 = *, p < 0.01 = **, p < 0.001 = ***, p < 0.0001 = ****)

Fig. 4

The numbers of CD45⁺ cells in the CSF increase after aEAE induction. (A) CSF tapping was performed by puncture of the cisterna magna of PBS perfused mice. On average a total of 15 µl of CSF was harvested and pooled from three mice, sedimented via cytospin and double-stained for CD45 (green) and CD4 (red). The CSF was collected from healthy mice and mice suffering from aEAE before onset of the disease (d 13 p.i.), at peak (d 18 p.i.) and during the chronic phase (d 30 p.i.). Scale bars = 50 μm. (B) Quantification of the absolute numbers of CSF derived CD45⁺CD4⁺ (yellow) and CD45⁺ (green) immune cells on the slides from healthy mice and mice suffering from aEAE.

Fig. 5

Mouse effector/memory CD4⁺ T cells can migrate across the non-stimulated and cytokine stimulated BCSFB in vitro. (A) The migration of MOG_{35 − 55}-specific T helper cell subsets (Th0, Th1, Th2, Th17) across unstimulated pmCPEC from the basolateral to apical side towards 100 ng/mL CXCL12 after 8 h is shown. The number of T cells added to the assay (imput = 4 × 10⁵ cells) was set as 100% and the numbers of migrated T cells was assessed by flow cytometry. Bars show the mean % ± SD of two experiments with three filters per condition. (B) The migration of MOG_{35 − 55}-specific T helper cell subsets (Th0, Th1, Th2, Th17) across unstimulated or either TNFα or IFNγ (10 ng/mL, 24 h) stimulated pmCPEC from the basolateral to apical side after 8 h is shown. The number of T cells added to the assay (imput = 4 × 10⁵ cells) was set as 100% and the numbers of migrated T cells was assessed by flow cytometry. Bars show the mean % ± SD of two experiments with three filters per condition. Statistical analysis: (A) one-way ANOVA and (B) two-way ANOVA (p < 0.05 = *, p < 0.01 = **, p < 0.001 = ***, p < 0.0001 = ****)

Fig. 6

The migration of mouse Th17 cells across the BCSFB in vitro does not require CCR6. Primary mouse choroid plexus epithelial cells (pmCPECs) were used as in vitro model for the BCSFB. (A) Confluent monolayers of non-stimulated or TNFα or IFNγ (10 ng/mL, 16 h) stimulated pmCPECs were double-stained for CCL20 (green) and for E-Cadherin (red) and for nuclei (DAPI, blue). The bar graph shows the quantification of the fluorescence intensity for CCL20 (green) relative to unstimulated pmCPECs. Scale bars = 50 μm. (B) Total amount (pg) of CCL20 secretion of unstimulated and 16 h pro-inflammatory cytokine (10 ng/mL of TNFα) stimulated pmCPECs monolayers towards the basolateral (top compartment) and apical (bottom compartment) side as determined by ELISA. Bar graphs show the mean ± SD of two independent experiments measured in triplicates. (C) The migration of MOG_{35 − 55}-specific WT or CCR6 KO Th17 cells across unstimulated or cytokine stimulated (10 ng/mL of TNFα or INFγ for 24 h prior to assay) pmCPECs from the basolateral to the apical side after 8 h is shown. The number of T cells added to the assay (imput = 4 × 10⁵ T cells) was set as 100% and the numbers of migrated T cells was assessed by flow cytometry. Bars show the mean % ± SD of 4 experiments with three filters per condition. Statistical analysis: (B, C) two-way ANOVA (p < 0.05 = *, p < 0.01 = **, p < 0.001 = ***, p < 0.0001 = ****). In B the statistical analysis is based on the combined values of the mean and the upper and lower SD.

Fig. 7

Human effector/memory Th17 cells preferentially migrate across the BSCFB but in a CCR6/CCL20 independent manner. (A) Total amount (pg) of CCL20 secretion towards the basolateral (top compartment) and apical (bottom compartment) side of unstimulated and 16 h pro-inflammatory cytokine (76IU/mL TNFα + 20IU/mL IFNγ) stimulated HIBCPP monolayers as determined by ELISA. Bar graph shows the mean ± SD of three independent experiments measured in triplicates. (B) Percentage of human Th17 cells migrated across laminin coated Millicell® filters towards the lower compartment in the presence or absence of 500 ng/mL human recombinant CCL20. Bar graph shows the mean ± SD of three independent experiments measured in triplicates. (C) Percentage of human Th1, Th1*, Th2 and Th17 cells that migrated from the basolateral to the apical side across 16 h pro-inflammatory cytokine (76IU/mL TNFα + 20IU/mL IFNγ)-stimulated HIBCPP monolayers after CCR6 inhibition or vehicle control treatment. Dot plot shows the mean ± SD of 2 independent experiments done in triplicates with two different human Th17-cell donors. (D) Percentage of human Th17 cells that migrated from the basolateral to the apical side across 16 h pro-inflammatory cytokine (76IU/mL TNFα + 20IU/mL IFNγ)-stimulated HIBCPP monolayers in the presence or absence of additional 500 ng/mL human recombinant CCL20 in the basolateral (top compartment) or apical (bottom compartment) compartment, and in the presence or absence of CCR6 inhibition and vehicle control treatment. Bar graph shows the mean ± SD of three independent experiments. Statistical analysis: (B) one-way ANOVA and (A, C and D) two-way ANOVA (*p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001)