CNS-associated macrophages contribute to intracerebral aneurysm pathophysiology

Abstract

Intracerebral aneurysms (IAs) are pathological dilatations of cerebral arteries whose rupture leads to subarachnoid hemorrhage, a significant cause of disability and death. Inflammation is recognized as a critical contributor to the formation, growth, and rupture of IAs; however, its precise actors have not yet been fully elucidated. Here, we report CNS-associated macrophages (CAMs), also known as border-associated macrophages, as one of the key players in IA pathogenesis, acting as critical mediators of inflammatory processes related to IA ruptures. Using a new mouse model of middle cerebral artery (MCA) aneurysms we show that CAMs accumulate in the IA walls. This finding was confirmed in a human MCA aneurysm obtained after surgical clipping, together with other pathological characteristics found in the experimental model including morphological changes and inflammatory cell infiltration. In addition, in vivo longitudinal molecular MRI studies revealed vascular inflammation strongly associated with the aneurysm area, i.e., high expression of VCAM-1 and P-selectin adhesion molecules, which precedes and predicts the bleeding extent in the case of IA rupture. Specific CAM depletion by intracerebroventricular injection of clodronate liposomes prior to IA induction reduced IA formation and rupture rate. Moreover, the absence of CAMs ameliorated the outcome severity of IA ruptures resulting in smaller hemorrhages, accompanied by reduced neutrophil infiltration. Our data shed light on the unexplored role of CAMs as main actors orchestrating the progression of IAs towards a rupture-prone state.

Keywords: CNS-associated macrophages; Intracerebral aneurysms; MRI; Middle cerebral artery; Vascular inflammation.

Fig. 1

A new mouse model of IAs at the middle cerebral artery. a Schematic representation of experimental design and the timeline scale. b Representative images of formation, growth and rupture assessed by longitudinal MRI. Contrast-enhanced T1 weighted (CE-T1-w) MRI was used to confirm the presence of IAs (yellow arrow). T2*-w MRI was performed to confirm the rupture of IAs with the presence of hemorrhage (red arrow). Once IA ruptures, it is no longer observable with T1-w MRI, bleeding appears as hypointense signal on T2*-w MRI. c Representative T1-w images, with the location of the IA at the lateral part of the brain. Three-dimensional IA reconstruction created from the MR scans revealed the saccular type of IAs. d Occurrences of ruptured and unruptured aneurysms at different time points revealed that many aneurysms appear already at the Day 5, and most of them are ruptured by the Day 15 (presented in e). f Kaplan–Meier survival graph. g Average number of IA per animal quantified from T1-w MRI during the 15-days timeline (left), average number of IAs per animal assessed per histological analysis at the end of the protocol (right). Scale bar: 500 µm. h Hemorrhage volumes measured from T2*-w scans during the experimental timeline

Fig. 2

Histological and immunofluorescence analysis of vessel morphological changes in the mouse IA samples. a Schematic representation of experimental design. b Modified Verhoeff-Van Gieson staining on cryosections from the mouse brain, continuously infused with Ang II for 14 days. Samples were obtained at Day 15. Elastic laminae are visualized by the dark purple lines in the media of the artery. Collagen appears red/purple, nuclei in black and other structures in yellow. IA vessel shows visible degradation of internal elastic lamina, collagen turnover in red, and cell accumulation (nuclei in black). Closer look at the aneurysm (20× magnification, on the right). Representative section of IA in greater detail revealed multi-lobed nucleus characteristic for neutrophils—found at the lumen of the IA vessel. Scale bar: 100 µm. c Representation of how the measurements were carried out (Scale bar: 50 μm). d Intima media thickness and wall-to-lumen ratio. Based on the histological images, the widest vessel and vessel lumen diameters of both IAs and the contra arteries were measured. Vessel wall diameter was defined as the difference in intima-media thickness. Both intima-media thickness and the wall-to-lumen ratio were significantly increased in the aneurysms compared to the contra arteries. (****P < 0.0001, N = 11 mice, n = 60 aneurysms/n = 41 contra arteries, Mann–Whitney U test). e Representative images of Verhoeff–Van Gieson staining with IAs assigned different stages of the progression according to the manifested morphological changes. Inflammatory cell infiltration in the IA walls and at the luminal surface are indicated with the arrow. Scale bar: 50 µm f Vessel wall thickness—aneurysm vessel wall shows substantially increased diameter in compared to the healthy, contralateral artery. g Weak and dispersed laminin signal in the IAs as a sign of internal laminae degradation. Visible difference in laminin expression between the healthy contralateral artery and the IA. h Confirmation of Col IV turnover (characteristic for IAs) with Col IV positive signal in the aneurysm vessel. i SMC proliferation and migration. j Some IAs presented a complete degradation of SMCs. k Damage of the endothelium characteristic for IAs. CD31 (platelet endothelial cell adhesion molecule; PECAM-1) marker reveals discontinuous signal in the IA vessel wall with no cell–cell junctions were clearly visible. l Fibrinogen accumulation on the luminal side of the IA. m, n Immunofluorescence images with visible P-selectin signal (in turquoise blue, m) and VCAM-1 (magenta pink, n) signal, characteristic of increased endothelial activation found within the vessel lumen. Scale bar: 100 µm. Statistical analysis was performed using Mann–Whitney U test; **P < 0.01, ***P < 0.001; NS, not significant. Further detailed information on statistics is provided in the Methods and Materials

Fig. 3

Histological analysis of vessel morphological changes in the human IA samples. a Schematic representation of experimental design. A human sample was resected after microsurgical clipping from a partially thrombosed aneurysm arising from the bifurcation of the right middle cerebral artery. b Hematoxylin eosin staining of human IA sample obtained after clipping reveals thick, hypocellular wall structure with thrombus on the luminal surface. Increased thickness of the vessel wall and the wall decellularization (loss of mural cells) confirms the findings from the mice MCA aneurysms. Scale bar: 10 mm and 200µ. c Coronal view of the preoperative digital subtraction angiography (on the left) shows the filling lumen of the IA (marked with an arrow), while preoperative T1-w black blood MRI sequence without (in d) and with gadolinium contrast (in e). Wall enhancement observed after gadolinium enhancement is marked with arrow in e. f Hematoxylin eosin staining (0.7× magnification) demonstrates that the thickness of IA wall is largely composed of thrombus lining the luminal surface (marked with T). The border between the actual wall and the thrombus is depicted with an interrupted line. Higher magnification microphotographs from a CD31 immunofluorescence in g and from a CD34 immunofluorescence in (h). g, h with 20× magnification and 40× magnification, respectively. The presence of inflammatory cell infiltration is marked with i and capillaries are marked with c (both presented in g). i Adventitial inflammatory cell infiltration within the wall. The dark rectangle in b corresponds to the wall region from with the higher magnification microphotographs in panel i are taken. Similarly, the rectangle in i correspond to the wall regions from which the higher magnification microphotographs were taken. Scale bars: 500 µm, and 200 µm respectively

Fig. 4

Molecular MRI reveals vascular inflammation in the IA area and predicts the outcome severity. a Schematic representation of the experimental approach. b Representative high-resolution T2*-w images (coronal view) before and after MPIO-αP-selectin intravenous injection in mice at Day 3 post-IA induction. MPIO-αP-selectin–induced hypointense signal in the T2*-w sequence in the vicinity of the IA area. Arrowheads indicate the MPIO-specific hypointensities. P-selectin endothelial activation is increased in the ipsilateral hemisphere and restricted to the IA area. (**P < 0.01; N = 5 mice/group; Mann–Whitney U test). c Representative high-resolution T2*-w images (coronal view) after MPIO-αVCAM-1 intravenous injection in the mice subjected to IA induction, at Day3. Arrowheads indicate the MPIO-specific hypointensities concentrated mainly around the IA area. d Corresponding signal void quantification for the MPIO-αP-selectin (**P < 0.01; N = 5 mice/group; Mann–Whitney U test). e Corresponding signal void quantification for the MPIO-αVCAM-1. VCAM-1 endothelial activation significantly increased in the ipsilateral hemisphere and restricted to the IA area (**P < 0.01; N = 5 mice/per group; Mann–Whitney U test). f Representative three-dimensional reconstruction of MR scans (T2*-w sequence) with MPIO signal voids and their distribution across the brain. g Correlation between the P-selectin signal from the MR scans at Day3 and hemorrhage volumes at Day5 (N = 5, R² = 0.91). h Correlation between the VCAM-1 signal from the MR scans at Day3 and hemorrhage volumes at Day5 (N = 5, R² = 0.85). i Representative images of P-selectin MPIO signal void at earlier time points (Day 3) and the hemorrhage volumes after the IA ruptures (Day 5). A smaller volume of signal void results in a small hemorrhage, whereas a larger volume results in a substantially bigger hemorrhage. j Colocalization of IA location at Day5 and the P-selectin signal from the MR scans at the Day3. Statistical analysis was performed using Mann–Whitney U test; **P < 0.01, ***P < 0.001; NS, not significant. Further detailed information on statistics is provided in the Methods & Materials

Fig. 5

CD163⁺ Cd11b⁺ macrophage accumulation in the human IA wall. a, b Histological analysis of human MCA aneurysm wall reveals cells expressing the scavenger receptor CD163 as well as the CD11b⁺ cell presence in the IA wall. These represent the most abundant inflammatory cell types in our human IA sample. c A population of HLA-DR⁺ increased within the wall whereas some of these cells were also CD68 positive (as shown in d). e CD163 and Cd11b double-staining reveals colocalization of these cells in the human IA wall

Fig. 6

CD206⁺ CAM accumulation in the mouse IA wall. a An increased number of CD206⁺ macrophages found in the ipsilateral hemisphere at Day5 post-aneurysm induction, especially around the IAs. This was significantly higher than the number of CD206⁺ cells found around the control vessels. b Quantification of accumulated CAMs confirmed the significant accumulation of these cells in the IA area (****P < 0.0001, N = 11 mice, n = 47 IAs/n = 23 contra arteries, Mann–Whitney U test). c CAMs (CD206⁺ cells) are positive for the lysosomal marker CD68. Scale bar: 100 μm. Scale bar: 100 μm. d Quantification of CD68⁺ cells in the IA area. (****P < 0.0001, N = 11 mice, n = 47 IAs/n = 23 contra arteries, Mann–Whitney U test). Statistical analysis was performed using Mann–Whitney U test; **P < 0.01, ***P < 0.001; NS, not significant. Further detailed information on statistics is provided in the Methods and Materials

Fig. 7

Inflammatory cell accumulation in the mouse IA wall at Day 15 post-aneurysm induction. a Representative image of CD3⁺ cell accumulation around the IA vessel. Scale bar: 100 μm. b Quantification confirmed increased infiltration of CD3⁺ T-cells in the IA area. N = 11 mice, n = 53 IAs/n = 11 contra arteries. ****P < 0.0001, Mann–Whitney U test. c Representative image of Ly6G⁺ cells infiltrating the IA vessel wall. Scale bar: 100 μm. d Quantification confirmed increased infiltration of Ly6G⁺ neutrophils in the IA area. N = 11 mice; n = 51 IAs/n = 11 contra arteries. ****P < 0.0001, Mann–Whitney U test. e, f Infiltrating neutrophils co-localize with MPO and H3Cit positive signal. Scale bar: 50 μm. g Proportion of Ly6G⁺ MPO⁺ cells colocalization in the human MCA IA sample. h Proportion of Ly6G⁺ H3Cit⁺ cells in the human MCA IA sample. i Iba⁺ cell quantification shows increase in the number of microglia around the IA area. ****P < 0.0001. N = 6 mice, n = 29 IAs/n = 11 contra arteries, Mann–Whitney U test. j Microglial activation in the IA area confirmed with the expression of specific microglial activation markers. Scale bar: 100 μm and 10 μm, respectively. CD68 lysosomal marker was used to confirm the active microglia. Unlike the resting form of microglia found around the arteries on the contralateral hemisphere with long branched processes, microglia near the IA site show round shape, with decreased process extension. Closer to the IA, the expression of P2Y12R is found to be decreased, while expression of Iba1 marker is still highly present. k, l Unbiased automatic analysis of microglia morphological confirms the activation in the area of IAs. A significant decrease in the branch volume, branching and ending nodes and the increase in the sphericity was found ipsilaterally, in compared to the contralateral side. (****P < 0.0001, N = 5 mice/group. Kruskal–Wallis test followed by Dunn’s multiple comparisons test). Further detailed information on statistics is provided in the Methods and Materials

Fig. 8

CAM depletion reduces IA formation and ameliorates severity of IA ruptures. a Schematic representation of the experimental design. b Representative flow cytometry dot-plots and gating strategy used for quantification of CD206⁺ macrophages and microglia, with vehicle or clodronate-liposomes. c CLO-liposomes significantly reduce the number of CD206⁺ macrophages even 10 days post-depletion, whereas microglia number stayed unaltered. Flow cytometry quantification of CD206⁺ macrophages and Iba1⁺ microglia (d), treated either with vehicle or CLO-liposomes (N = 4, *P < 0.05, Mann–Whitney U test). e Average number of IAs developed per animal. Immunofluorescence analysis data (on the left) show that CLO-treated mice develop significantly fewer IAs compared to the control group (*P < 0.05, N = 6 CLO-treated mice/N = 4 PBS-treated mice, Mann–Whitney U test). Longitudinal MR studies over the timeline of 15 days confirmed the existing difference (on the right). f CLO-treated mice have decreased global rupture rate (*P < 0.05, n = 14 PBS-treated mice/n = 13 CLO-treated mice; PBS versus CLO: 72% versus 43%, Fisher exact test). g Aneurysm repartition between the PBS and CLO-treated mice during the period of 15 days (assessed on MR scans). h CAM depletion significantly reduces the hemorrhage volumes after the IA rupture, assessed by the MRI scans from T2*-w sequences (*P < 0.01; Mann–Whitney U test). i, j Wall-to-lumen ratio and intima-media thickness comparison between the group revealed no significant difference in dimensions of IAs (**P < 0.01, ***P < 0.001, N = 4 CLO-treated mice, n = 7 IAs/n = 6 contra arteries; N = 4 PBS-treated mice, n = 16 IAs/n = 8 contra arteries. Two-way ANOVA). Statistical analysis was performed using two-way analysis of variance (ANOVA) with post hoc Dunnett adjustment; *P < 0.05, **P < 0.01, ***P < 0.001. Further detailed information on statistics is provided in the Methods & Materials

Fig. 9

CD206⁺ and Ly6G⁺ cell quantification in IA walls of PBS and CLO-treated mice. a Representative image of CAM accumulation in PBS-treated mice. b Representative image of CAM absence in CLO-treated mice. c Quantification of CAM number around the IA vessel wall and the contra arteries at Day5 post-aneurysm induction (N = 4 CLO-treated mice, n = 7 IAs/n = 6 contra arteries. N = 4 PBS-treated mice, n = 16 IAs/n = 7 contra arteries). d Representative image of neutrophil accumulation in PBS-treated mice. e Representative image of the reduced neutrophil number in CLO-treated mice. f Quantification of neutrophil number in the IA vessel wall at Day5 post-aneurysm induction (N = 4 CLO-treated mice, n = 5 IAs/n = 5 contra arteries. N = 4 PBS-treated mice, n = 14 IAs/n = 8 contra arteries. *P < 0.05, **P < 0.01, two-way ANOVA. Scale bar: 100 μm). Statistical analysis was performed using two-way analysis of variance (ANOVA) with post hoc Dunnett adjustment. Further detailed information on statistics is provided in the Methods and Materials