Adventitial adaptive immune cells are associated with ascending aortic dilatation in patients with a bicuspid aortic valve

Abstract

Background: Bicuspid aortic valve (BAV) is associated with ascending aorta aneurysms and dissections. Presently, genetic factors and pathological flow patterns are considered responsible for aneurysm formation in BAV while the exact role of inflammatory processes remains unknown.

Methods: In order to objectify inflammation, we employ a highly sensitive, quantitative immunohistochemistry approach. Whole slides of dissected, dilated and non-dilated ascending aortas from BAV patients were quantitatively analyzed.

Results: Dilated aortas show a 4-fold increase of lymphocytes and a 25-fold increase in B lymphocytes in the adventitia compared to non-dilated aortas. Tertiary lymphoid structures with B cell follicles and helper T cell expansion were identified in dilated and dissected aortas. Dilated aortas were associated with an increase in M1-like macrophages in the aorta media, in contrast the number of M2-like macrophages did not change significantly.

Conclusion: This study finds unexpected large numbers of immune cells in dilating aortas of BAV patients. These findings raise the question whether immune cells in BAV aortopathy are innocent bystanders or contribute to the deterioration of the aortic wall.

Keywords: aortic dissection; auto-inflammation; bicuspid aortc valve; inflammation; multiplex immunohistochemistry; tertiary lymphoid structures; thoracic aorta aneurysm.

Figure 1

Image processing and data analysis methods. (A) All markers were stained consecutively and were simultaneously scanned with a fluorescence microscope and visualized with inForm. All separate channels of the adaptive immune cell panel (except autofluorescence) are shown to visualize the quality of the unmixing step. (B) inForm machine learning based tissue segmentation was used to distinguish the adventitia (A), media (M) and intima (I) and exclude the background (Ba) and blood (Bl) from the analysis. (C) Cell segmentation was based on identification of the nucleus on the DAPI signal and membrane identification on CD3, CD8 and CD20. (D) Single cell data visualized with FlowJo. Marker expression of the segmented single cells was assessed with a flowcytometry-like method. This resulted in cell populations that could be reproducibly gated as shown here for the adaptive immune cell panel. (E) Inter-observer agreement was excellent as measured with the intraclass correlation coefficient (ICC) thus cell population gating shows good reproducibility between observers (n = 27 for each cell type).

Figure 2

Adventitia of dilated aortas showed increased numbers of lymphocytes. (A) Representative cross-section of the adventitia of a non-dilated BAV ascending aorta (n = 13). Note the helper T cells, cytotoxic T cells and cDC2s which are mainly situated around the vasa vasorum (arrows). (B) Representative cross-section of the adventitia of a dilated aorta (n = 15). Note the strong increase of T cells and the appearance of a significant number of B cells (arrows). (C) Quantification of adaptive immune system cells in the adventitia of non-dilated (n = 8) and aneurysmatic (n = 15) BAV aortas, total analyzed 20×-views = 6,318. Data shown as median ± inter-quartile range, *p < 0.05, **p < 0.01 as calculated with the independent Kolmogrov-Smirnov Test. (D) Representative cross-section of the adventitia of a dissected BAV ascending aorta (n = 4). The adventitia has largely been destroyed by the force of the entering blood. However, a large number of lymphocytes can be seen scattered close to the media and at the edge of the adventitia (arrow).

Figure 3

Tertiary lymphoid structures (TLS) were only identified in the adventitia of dilated aortas and dissections. (A) Left, a 20x magnified image of a tertiary lymphoid structure showing the separate B and T cell zones and a germinal center. Right, result of tissue segmentation of this 20x magnified image with a machine learning algorithm. (B) 20x magnification brightfield hematoxylin and eosin stain of a TLS in a dilated aorta. Plasma cells, not stained by markers in our two antibody panels, are visible. (C) Representative image of a TLS found in the adventitia of an aneurysmatic aorta, germinal center indicated with an arrow. (D) Although most dissected BAV ascending aortas had a non-representative adventitia due to the disruption of the entering blood, TLS with germinal centers (white arrow) were observed.

Figure 4

The media of dissected aortas showed an increased number of helper T cells and cDC2s. (A) Representative cross-section of the media of a non-dilated BAV ascending aorta (n = 8). Sparse helper T cells (red arrow) were found in outer 2/3th of the media often in blood vessels. (B) Representative cross-section of the media of a dilated BAV ascending aorta (n = 15). A slight increase in number of T cells, now extravasated into the tissue, was observed in the media of dilated aorta samples. Cytotoxic T cells are indicated with a cyan arrow, cDC2 cells are indicated with a magenta arrow. Note the decrease of elastin density and elastin fiber breaks. (C) Infiltration of helper T cells (red arrow), and the occasional cytotoxic T cell (cyan arrow) in the aortic media of dissected aortas (n = 4). We also observed an increased number of cDC2s (magenta arrow) in these samples. Note the decrease of elastin density and elastin fiber breaks. (D) Quantification of adaptive immune system cells in the media of non-dilated (n = 8, dots), dilated (n =15, squares) and dissected (n = 4, triangles) BAV aortas. Data shown as median (interquartile range), *p < 0.05, **p < 0.01 as calculated with the independent Kruskal-Wallis Test adjusted with Bonferroni correction for multiple tests. Total analyzed 20x-views = 6318.

Figure 5

Dilated aortas and dissections showed increased numbers of macrophages in the aorta media. (A) Full thickness sample of a non-dilated BAV ascending aorta (n = 8). Note sparse M1-like macrophages (red arrow) in the media, closely situated to the sparse vasa vasorum (endothelium in white). The adventitia showed mainly M2-like macrophages (green arrow). Neutrophils were found intra-vascular (cyan arrow). (B) Full thickness sample of a dilated BAV ascending aorta (n = 15). Note the increase in M1-like macrophages and some disruption of the media elastin fibers (grey arrow). (C) Full thickness sample of a dissected BAV ascending aorta (n = 4). Note the increase in M1-like macrophages (red arrow) in the media and strong infiltration of neutrophils (cyan arrow) in the adventitia alone. (D) Quantification of innate immune system cells in the media and adventitia of non-dilated (n = 8), dilated (n = 15) and dissections (n = 4), total analyzed 20×-views = 5,931. Data shown as median ± inter-quartile range, *p < 0.05, **p < 0.01 as calculated with the independent Kruskal-Wallis Test adjusted with Bonferroni correction for multiple tests.

Figure 6

Nearest neighbor analysis of the distance between M1-like macrophages and endothelium did not show macrophages in dissections closer to the vasa vasorum. (A) Visualization of the nearest neighbor analysis on a single 20× view. M1-like macrophages are marked in red, endothelium in cyan, the distance between a M1-like macrophage and the closest endothelial cell was marked with a dashed white line. Cells are only phenotyped in the presence of a DAPI stained nucleus, which is an important quality control mechanism to prevent analysis of artefact staining. CD68 staining (in red) of the irregular membrane of macrophages can be seen without the cells’ nucleus in the section, consequently these signals are not analyzed as a cell. Note that the full analysis was done on a whole slide level, not on separate 20× views. (B) Quantification of the median distance between M1-like macrophages and endothelium. We do not observe a significant difference in this distance between non-aneurysmatic (n = 5), aneurysmatic (n = 9) and dissected (n = 4) samples as calculated with the Kruskal Wallis test. Data shown as median ± inter-quartile range.