Selective inhibition of soluble tumor necrosis factor signaling reduces abdominal aortic aneurysm progression

Abstract

Background: Tumor necrosis factor (TNF) is pathologically elevated in human abdominal aortic aneurysms (AAA). Non-selective TNF inhibition-based therapeutics are approved for human use but have been linked to several side effects. Compounds that target the proinflammatory soluble form of TNF (solTNF) but preserve the immunomodulatory capabilities of the transmembrane form of TNF (tmTNF) may prevent these side effects. We hypothesize that inhibition of solTNF signaling prevents AAA expansion.

Methods: The effect of the selective solTNF inhibitor, XPro1595, and the non-selective TNF inhibitor, Etanercept (ETN) was examined in porcine pancreatic elastase (PPE) induced AAA mice, and findings with XPro1595 was confirmed in angiotensin II (ANGII) induced AAA in hyperlipidemic apolipoprotein E (Apoe) ^-/- mice.

Results: XPro1595 treatment significantly reduced AAA expansion in both models, and a similar trend (p = 0.06) was observed in PPE-induced AAA in ETN-treated mice. In the PPE aneurysm wall, XPro1595 improved elastin integrity scores. In aneurysms, mean TNFR1 levels reduced non-significantly (p = 0.07) by 50% after TNF inhibition, but the histological location in murine AAAs was unaffected and similar to that in human AAAs. Semi-quantification of infiltrating leucocytes, macrophages, T-cells, and neutrophils in the aneurysm wall were unaffected by TNF inhibition. XPro1595 increased systemic TNF levels, while ETN increased systemic IL-10 levels. In ANGII-induced AAA mice, XPro1595 increased systemic TNF and IL-5 levels. In early AAA development, proteomic analyses revealed that XPro1595 significantly upregulated ontology terms including "platelet aggregation" and "coagulation" related to the fibrinogen complex, from which several proteins were among the top regulated proteins. Downregulated ontology terms were associated with metabolic processes.

Conclusion: In conclusion, selective inhibition of solTNF signaling reduced aneurysm expansion in mice, supporting its potential as an attractive treatment option for AAA patients.

Keywords: abdominal aortic aneurysm; cardiovascular disease; inflammation; translational research; tumor necrosis factor inhibitor; vascular inflammation.

FIGURE 1

Selective inhibition of solTNF prevents expansion in PPE-induced AAA. (A) Representative macroscopic images displaying PPE-induced AAAs after 14 days of treatment with vehicle (saline), ETN (20 mg/kg), or XPro1595 (20 mg/kg). (B) In vivo AAA expansion measured as percentage change in the maximal outer diameter of the abdominal aorta (OAD) from day 0 to day 14 (n = 15–16). (C) AAA cross-sections stained with Miller’s elastic stain (black) after TNF inhibition (n = 3–5). Breaks in elastic fibers are marked with arrows. (D) Percentage distribution of elastic fiber integrity (grades 1–4) according to treatment groups (n = 3–5). (E,F) The number of MMP-9-positive cells/mm² cross-sectional aneurysmal area from mice treated with vehicle-, ETN-, or XPro1595 (n = 3–5). Data are shown as mean SEM. * Indicates p < 0.05 analyzed by one-way ANOVA using Bonferroni test for multiple comparisons. Contingency table data in E was analyzed by Fisher exact probability test with Freeman-Halton extension. M, media; Ad, adventitia.

FIGURE 2

TNF receptor protein levels and receptor localization in murine and human abdominal aortic tissue. (A) Aneurysmal protein levels of TNFR1 and TNFR2 from PPE- induced aneurysm in mice treated with vehicle (saline), ETN (20 mg/kg), or XPro1595 (20 mg/kg) (n = 3–5). Data are shown as mean SEM. A non-significant trend was detected in TNFR1 levels by one-way ANOVA using Bonferroni test for multiple comparisons. (B) Representative distribution of TNFR1 and TNFR2 in the PPE-induced AAA wall after treatment with saline, ETN, or XPro1595 for 14 days (n = 3–5). (C) TNFR1 distribution in unchallenged thoracic murine aorta. (D) Distribution of TNFR1 in human AAA specimens (n = 2) and human ascending aortas (n = 2). Black arrowheads indicate positive stained cells. M, media; Ad, adventitia.

FIGURE 3

Effect of TNF inhibition on infiltrating immune cells 14 days after PPE-induced AAA formation. Representative micrographs after TNF inhibition showing the distribution of infiltrating immune cells (left) and semi-quantification (right, number of positive cells/mm²) of CD45-positive leukocytes (A), CD206-positive M2-like macrophages (B), CD3-positive T-cells (C), and Ly6G-positive neutrophils (D) in the aneurysm wall after TNF inhibition (n = 3–5). Black arrowheads indicate positive stained cells. Data are shown as mean ± SEM. None of the semi-quantifications reached significance on one-way ANOVA using Bonferroni test for multiple comparisons. M, media; Ad, adventitia.

FIGURE 4

Pro- and anti-inflammatory cytokine profile systemically or locally produced in PPE-induced aneurysms after TNF inhibition. Changes in aneurysmal cytokine levels of TNF (n = 5–8) (A), IFN-γ (n = 6–8) (B), and IL-10 (n = 6–8) (C) 14 days after PPE-induced AAA mice treated with vehicle (saline), ETN (20 mg/kg), or XPro1595 (20 mg/kg). Changes in plasma levels of TNF (n = 7–8) (D), IL-1β (n = 8) (E), and IL-10 (n = 7–8) (F). Data are shown as mean SEM. No significance was observed by one-way ANOVA using Bonferroni test for multiple comparisons. P-values are denoted *p < 0.05, ***p < 0.001, and ^# denotes p < 0.05 for unpaired Student’s t-test for two-group comparisons.

FIGURE 5

The fibronectin/fibrinogen complex is upregulated in aneurysms after inhibition of solTNF by XPro1595. (A) Dot plots show the top 10 enriched biological function terms from proteins up- (left) or down-regulated (right) in PPE-induced aneurysms from XPro1595-treated mice (n = 7) vs. vehicle (n = 6) at day 7. Dot size indicates number of proteins found to be regulated by XPro1595 for a given term. The ratio shows the proportion of a functional term covered by proteins regulated by XPro1595, and the dot color indicates the level of significance. (B) Volcano plot of aneurysmal proteins and their differential expression in aneurysms from XPro1595-treated mice vs. vehicle. Horizontal dashed line denotes a 0.05 cut-off p-value, and vertical dashed lines denote a log2(fold-change) of 0.5 in either direction. Fibronectin (FN1), fibrinogen chains FGA, FGB, and FGG and their protein-protein interactions are illustrated by use of the STRING resource (74), showing interactions determined experimentally (pink lines), from curated databases (blue lines), and/or from text-mining (yellow lines) (left upper corner of volcano plot). (C) Fibronectin contents in representative abdominal aortic aneurysmal tissue sections from mice infused with elastase and treated with XPro1595 (20 mg/kg) or vehicle (n = 3–5). Black arrowheads indicate positive stained cells. Proteomics data were analyzed by unpaired t-test for each protein followed by fdr correction for multiple testing. M, media; Ad, adventitia.

FIGURE 6

Inhibition of solTNF prevents ANGII-induced AAA expansion. (A) Representative macroscopic images showing AAA expansion in vehicle (saline) or XPro1595 (2 mg/kg) ANGII-infused Apoe^–/– mice. (B) Representative ultrasound images of the abdominal aorta at its maximal inner circumference at day 0 and day 28 of the ANGII-induced AAA from vehicle-treated and XPro1595-treated mice. (C) ANGII-induced AAA expansion assessed by inner abdominal aortic circumference. (D) The percentage change in the maximal inner abdominal aortic circumference from day 0 to day 28 in vehicle- and XPro1595-treated mice assessed by ultrasound€ (E) Ex vivo-measured maximal outer aortic diameter in the two groups. (F) Ex vivo-measured surface area of the AAA treated with vehicle and XPro1595. (G) Representative micrographs of AAA cross-sections stained with Miller’s elastin stain (black) in vehicle- and XPro1595-treated AAAs. Arrows point to rupture of elastic fibers. (H) Medial elastin degradation semi-quantified by a 4-grade system in aneurysms from XPro1595- and vehicle-treated mice. (I) MMP-9 immunoreactivity in cross-sections of vehicle- and XPro1595-treated aneurysms and the representative semi-quantification (number of positive cells/mm²). Black arrowheads indicate positive stained cells. IgG isoform was used as negative control. Results are shown as mean SEM if data passed the D’Agostino and Pearson omnibus normality tests. Mann–Whitney test was applied in D and presented as median with interquartile range as data were non-normally distributed. *Denotes p < 0.05 analyzed by unpaired student t-test for two-group comparison or two-way ANOVA with Bonferroni test for multiple comparisons. Contingency table data in H was analyzed by Fisher exact probability test with Freeman-Halton extension. Outliers were identified by Grubb’s outlier test for normally distributed data and indicated by red squares () and excluded from the statistical analyses. M, media; Ad, adventitia.

FIGURE 7

Characterization of aneurysm wall composition after inhibition of solTNF in ANGII-induced abdominal aortic aneurysms. Representative micrographs of CD45-positive leukocytes (A), Ly6G-positive neutrophils (B), CD3-positive T cells (C) (n = 13), TNFR1 localization (D), TNFR2 localization (E) (n = 6) and their respective semi-quantification (number of positive cells/mm² or % positive staining) in cross-sections of ANGII-induced AAA treated with vehicle or XPro1595 (2 mg/kg). Black arrows point to examples of positive stained cells. Circulating TNF plasma levels (n = 4–5) (F) and IL-5 (n = 6–7) (G) in ANGII-induced AAA treated with vehicle or XPro1595. Data are shown as mean SEM if data passed the D’Agostino and Pearson omnibus normality tests (B,C,E–G). *Denotes p < 0.05 analyzed by unpaired Student’s t-test for normally distributed data, otherwise by Mann-Whitney test represented as median with interquartile range (A,D). M, media; Ad, adventitia.