Endothelial to mesenchymal transition in the interleukin-1 pathway during aortic aneurysm formation

Abstract

Objective: Endothelial to mesenchymal transition may represent a key link between inflammatory stress and endothelial dysfunction seen in aortic aneurysm disease. Endothelial to mesenchymal transition is regulated by interleukin-1β, and previous work has demonstrated an essential role of interleukin-1 signaling in experimental aortic aneurysm models. We hypothesize that endothelial to mesenchymal transition is present in murine aortic aneurysms, and loss of interleukin-1 signaling attenuates this process.

Methods: Murine aortic aneurysms were created in novel CDH5-Cre lineage tracking mice by treating the intact aorta with peri-adventitial elastase. Endothelial to mesenchymal transition transcription factors as well as endothelial and mesenchymal cell markers were analyzed via immunohistochemistry and immunofluorescence (n = 10/group). To determine the role of interleukin-1 signaling, endothelial-specific interleukin-1 receptor 1 knockout and wild-type mice (n = 10/group) were treated with elastase. Additionally, C57/BL6 mice were treated with the interleukin-1 receptor 1 antagonist Anakinra (n = 7) or vehicle (n = 8).

Results: Elastase treatment yielded greater aortic dilation compared with controls (elastase 97.0% ± 34.0%; control 5.3% ± 4.8%; P < .001). Genetic deletion of interleukin-1 receptor 1 attenuated aortic dilation (control 126.7% ± 38.7%; interleukin-1 receptor 1 knockout 35.2% ± 14.7%; P < .001), as did pharmacologic inhibition of interleukin-1 receptor 1 with Anakinra (vehicle 146.3% ± 30.1%; Anakinra 63.5% ± 23.3%; P < .001). Elastase treatment resulted in upregulation of endothelial to mesenchymal transition transcription factors (Snail, Slug, Twist, ZNF) and mesenchymal cell markers (S100, alpha smooth muscle actin) and loss of endothelial cell markers (vascular endothelial cadherin, endothelial nitric oxide synthase, von Willebrand factor). These changes were attenuated by interleukin-1 receptor 1 knockout and Anakinra treatment.

Conclusions: Endothelial to mesenchymal transition occurs in aortic aneurysm disease and is attenuated by loss of interleukin-1 signaling. Endothelial dysfunction through endothelial to mesenchymal transition represents a new and novel pathway in understanding aortic aneurysm disease and may be a potential target for future treatment.

Keywords: aortic aneurysm; inflammation.

FIGURE 1.

Murine models of AA demonstrate increased elastin degradation, increased EndMT transcription factor and mesenchymal cell marker staining, and decreased EC marker staining after topical elastase. A, Abdominal aortic dilation in CDH5-Cre (VEC-Cre) lineage tracking mice exposed to elastase was significantly greater than CDH5-Cre (VEC-Cre) mice exposed to saline after 14 days (P <.001). B, By immunohistochemistry, representative aortic cross-sections demonstrated increased staining of the EndMT transcription factor Twist and the mesenchymal marker S100 in CDH5-Cre elastase treated mice. Additionally, the samples demonstrated decreased staining of the endothelial marker VE-cadherin and elastin in CDH5-Cre elastase-treated mice. Scale bar, 100 μM. C, Histologic analysis using positive cell detection demonstrated significant increase in EndMT transcription factors (Snail [P <.001], Slug [P <.001], Twist [P <.001], ZNF281 [P <.001]) and mesenchymal cell markers (S100 [P <.001], ⍺SMA [P <.001]) in CDH5-Cre elastase-treated mice. Conversely, EC markers (VE-cadherin [P <.001], eNOS [P <.001], VWF [P <.001]) were preserved in CDH5-Cre saline-treated mice. Percent positive pixel detection demonstrated decreased elastin in CDH5-Cre elastase-treated mice (P = .015). *For all graphs, a horizontal line represents the median of the respective data. AAA, Abdominal aortic aneurysm; αSMa, alpha smooth muscle actin; eNOS, endothelial nitric oxide synthase; VWF, von Willebrand factor.

FIGURE 2.

Endothelial lineage tracing demonstrates increased endothelial-originating cells with loss of EC markers and expression of mesenchymal cell markers (Transitioned) or co-expression of endothelial and mesenchymal cell markers (Transitioning) after topical elastase. A, Immunofluorescence and confocal microscopy performed on CDH5-Cre (VEC-Cre) eYFP +/+ endothelial specific lineage tracking mice. Cell nuclei (DAPI) are shown in blue, CDH5-Cre linage marker in purple, VWF in green, and αSMA in red. Merged image demonstrates αSMA colocalization with VWF and endothelial lineage marker in CDH5-Cre elastase-treated mice but not in CDH5-Cre saline-treated mice. B, Histologic analysis using multiplexed analysis demonstrated significant increase in endothelial to mesenchymal Transitioned and Transitioning cells using the EC markers VWF (Transitioned, P = .001; Transitioning, P = .002) and VE-cadherin (Transitioned, P<.001; Transitioning, P = .010) in CDH5-Cre elastase-treated mice. For all graphs, a horizontal line represents the median of the respective data. AAA, Abdominal aortic aneurysm; VE-Cadherin, vascular endothelial cadherin; H&E, hematoxylin–eosin; VWF, von Willebrand factor.

FIGURE 3.

IL-1R1 KO mice demonstrate decreased elastin degradation, decreased EndMT transcription factor and mesenchymal cell marker staining, and increased EC marker staining after topical elastase. A, Abdominal aortic dilation in CDH5 IL-1R KO (flx/flx) mice was significantly less than CDH5 IL-1R wild-type (wt/wt) mice 14 days after exposure to elastase (P <.0001). B, By immunohistochemistry, representative aortic cross-sections demonstrated increased staining of the EndMT transcription factor Snail and the mesenchymal marker ⍺SMA in CDH5 IL-1R wild-type (wt/wt) mice. Additionally, the samples demonstrated decreased staining of the endothelial marker VE-cadherin and elastin in CDH5 IL-1R KO (wt/wt) mice. Scale bar, 100 μM. C, Histologic analysis using positive cell detection demonstrated significant increase in EndMT transcription factors (Snail [P = .006], Slug [P = .004], Twist [P <.001], ZNF281 [P <.001]) and mesenchymal cell markers (S100 [P = .036], ⍺SMA [P = .019]) in CDH5 IL-1R wild-type (wt/wt) mice. Conversely, EC markers (VE-cadherin [P = .012], eNOS [P = .005], and VWF [P<.001]) were preserved in CDH5 IL-1R KO (flx/flx) mice. Percent positive pixel detection demonstrated decreased elastin in CDH5 IL-1R wild-type (wt/wt) mice (P = .025). *For all graphs, a horizontal line represents the median of the respective data. VVG, Verhoeff-Van Gieson; αSMa, alpha smooth muscle actin; eNOS, endothelial nitric oxide synthase; VWF, von Willebrand factor.

FIGURE 4.

IL-1R1 KO mice demonstrate decreased endothelial-originating cells with loss of EC markers and expression of mesenchymal cell markers (Transitioned) or co-expression of endothelial and mesenchymal cell markers (Transitioning) after topical elastase. A, Immunofluorescence and confocal microscopy performed on CDH5 IL-1R1 KO (flx/flx) and wild-type (wt/wt) lineage tracing mice. Cell nuclei (DAPI) are shown in blue, CDH5-Cre linage marker in purple, VWF in green, and αSMA in red. Merged image demonstrates αSMA colocalization with VWF and endothelial lineage marker in CDH5 IL-1R wild-type (wt/wt) mice but not in CDH5 IL-1R1 KO (flx/flx) mice. B, Histologic analysis using multiplexed analysis demonstrated significant decrease in endothelial to mesenchymal Transitioned and Transitioning cells using the EC markers VWF (Transitioned, P < .001; Transitioning, P = .030) and VE-cadherin (Transitioned, P = .006; Transitioning, P = .020) in CDH5 IL-1R1 KO (flx/flx) mice. For all graphs, a horizontal line represents the median of the respective data. VE-Cadherin, Vascular endothelial cadherin; αSMa, alpha smooth muscle actin; H&E, hematoxylin–eosin; VWF, von Willebrand factor.

FIGURE 5.

Murine models of AA demonstrate decreased elastin degradation, decreased EndMT transcription factor and mesenchymal cell marker staining, and increased EC marker staining after treatment with Anakinra. A, Abdominal aortic dilation in C57BL/6 WT mice who received Anakinra 14 days after elastase exposure was less than WT mice who received Vehicle (saline) (P<.001). B, By immunohistochemistry, representative aortic cross-sections demonstrated decreased staining of the EndMT transcription factor ZNF and the mesenchymal marker S100A4 in WTAnakinra-treated mice. Additionally, the samples demonstrated increased staining of the endothelial marker eNOS and elastin in WT Anakinra-treated mice. Scale bar, 100 μM. C, Histologic analysis using positive cell detection demonstrated significant decrease in EndMT transcription factors (Snail [P = .006], Slug [P = .036], Twist [P = .027], ZNF281 [P = .047]) and mesenchymal cell markers (S100A4 [P = .002], ⍺SMA [P = .006]) in WTAnakinra-treated mice. Additionally, EC markers (VE-cadherin [P = .003], eNOS [P = .040], VWF [P <.001]) were preserved in WT Anakinra-treated mice. Percent positive pixel detection demonstrated decreased elastin in WT vehicle-treated mice (P = .038). *For all graphs, a horizontal line represents the median of the respective data. VVG, Verhoeff-Van Gieson; αSMa, alpha smooth muscle actin; VE-Cadherin, vascular endothelial cadherin; eNOS, endothelial nitric oxide synthase; VWF, von Willebrand factor.

FIGURE 6.

Human aneurysms demonstrate increased EndMT transcription factor and differences in mesenchymal cell marker and EC markers staining. A, By immunohistochemistry, human AA cross-sections, regardless of region, demonstrated increased staining of the EndMT transcription factors Snail and Slug. AAA demonstrated increased staining of the mesenchymal cell marker S100 and decreased staining of the EC marker VWF, whereas aTAA and dTAA tissue demonstrated little difference in staining for S100 and VWF compared with controls. Scale bar, 100 μM. B, Histologic analysis using positive cell detection demonstrated significant increase of the EndMT transcription factors Snail (aTAA: P <.001; dTAA: P <.001; AAA: P = .04) and Slug (aTAA: P = .02; dTAA: P = .02; AAA: 26.8 P = .0001) compared with controls. AAA tissue also demonstrated decreased staining of the EC marker VWF (P = .01) and increased staining of the mesenchymal cell marker S100 (P <.001) compared with controls. However, aTAA and dTAA tissue demonstrated no difference in staining for VWF (aTAA: P = .16; dTAA: P>.9) and S100 (aTAA: P>.9; dTAA: P>.9) compared with controls. VWF, Von Willebrand factor.