β-Aminopropionitrile monofumarate induces thoracic aortic dissection in C57BL/6 mice

Thoracic aortic dissection (TAD) is a catastrophic disease with high mortality and morbidity, characterized by fragmentation of elastin and loss of smooth muscle cells. However, the underlying pathological mechanisms of this disease remain elusive because there are no appropriate animal models, limiting discovery of effective therapeutic strategies. We treated mice on C57BL/6 and FVB genetic backgrounds with β-aminopropionitrile monofumarate (BAPN), an irreversible inhibitor of lysyl oxidase, for 4 wk, followed by angiotensin II (Ang II) infusion for 24 h. We found that the BAPN plus Ang II treatment induced formation of aortic dissections in 100% of mice on both genetic backgrounds. BAPN without Ang II caused dissections in few FVB mice, but caused 87% of C57BL/6 mice to develop TAD, with 37% dying from rupture of the aortic dissection. Moreover, a lower dose of BAPN induced TAD formation and rupture earlier with fewer effects on body weight. Therefore, we have generated a reliable and convenient TAD model in C57BL/6 mice for studying the pathological process and exploring therapeutic targets of TAD.

also attenuated body weight gains (Fig. 1c,d) and significantly decreased plasma triglyceride and cholesterol levels (Fig. 1e,f) in both FVB and C57BL/6 mice.
Induction of TAD with BAPN treatment. We next examined the effects of BAPN treatment on the incidence of TAD. To determine whether Ang II was also required for TAD development, the mice were sacrificed for autopsy after 4 wk of BAPN treatment, with or without 24 h of Ang II infusion. Consistent with the previous report 8 , administration of BAPN plus Ang II induced TAD in all mice, while approximately 75% of FVB mice treated with BAPN alone did not develop TAD ( Fig. 2a and Table 1). However, the incidence of TAD in C57BL/6 mice treated with only BAPN reached 87% ( Fig. 2a and Table 1) and 45% of mice in this group died of aortic rupture. TAD was also observed in all C57BL/6 mice treated with BAPN plus Ang II, with 50% of these having aortic rupture within 24 h of Ang II infusion. Aortas of C57BL/6 mice given BAPN, with or without Ang II infusion, were enlarged from the root to the thoracic segment and, in some cases, the abdominal segment was also involved. Hematomas were observed in the lesions, indicating thrombosis (Fig. 2a). Haematoxylin and eosin staining showed tearing of the aortic wall and thrombi in the false lumens (Fig. 2b). Massive fragmentation and depletion of elastic fibres and smooth muscle cell loss were confirmed by Elastica van Gieson (EVG) staining (Fig. 2c). All these pathological changes resembled those observed histopathologically in humans (Fig. 2d). (a,b) Diastolic and systolic blood pressures (n = 6 per group); *P < 0.05. (c,d) Body weights with bar graph showing changes during the study (n = 6 per group); **P < 0.01. (e,f) Levels of plasma cholesterol (CHO) (n = 11 for FVB Ctrl, n = 16 for FVB BAPN, n = 6 for C57BL/6 Ctrl, n = 15 for C57BL/6 BAPN) and triglyceride (TG) (n = 12 for FVB Ctrl, n = 6 for FVB BAPN, n = 7 for C57BL/6 Ctrl, n = 15 for C57BL/6 BAPN); *P < 0.05, **P < 0.01.

BAPN dose optimization for TAD induction.
To further investigate the causal effects of medial degeneration on TAD formation, we applied different doses of BAPN by feeding 3-wk-old C57BL/6 male mice with diets containing 0, 0.4, 1.0 or 1.5 g BAPN per 100 g mouse chow for 4 wk. Body weights were lower with increased BAPN doses (Fig. 3a). All six mice fed with the 0.4 g BAPN per 100 g diet developed TAD and five died of dissection ruptures by 2 to 4 wk after BAPN administration. Of six mice fed with the 1.0 g BAPN diet, two had TAD at the end of the treatment, but no ruptures occurred. Most surprisingly, no TAD formation was observed in mice fed with the 1.5 g BAPN diet (Fig. 3b).

Molecular phenotypic features of BAPN-induced TAD.
Because BAPN-induced TAD exhibited typical histological features of the human disease, we next examined whether expression of TAD-related genes were also changed in the media of aortas. A panel of genes known to be dysregulated during medial degradation in TAD formation were selected for analysis. These were matrix metalloproteinases (MMPs, MMP2/3/9) 5,8,11,12 and cathepsins (cathepsin S/K/L) 13 (that degrade extracellular matrix), collagen I α1 (COL1α1) and connective tissue growth factor (CTGF) (target genes indicating activation of the TGF-β signalling pathway in LDS) 14 , α-smooth muscle actin (α-SMA) and β-myosin heavy chain (β-MHC) (associated with familial thoracic aortic aneurysm and dissection syndrome) 15 . Expression of these genes was compared in control and BAPN-treated C57BL/6 mice. In the BAPN-treated group, compared with the control, MMP2 was significantly upregulated (Fig. 4a), while MMP3 and MMP9 were downregulated (Fig. 4b,c). Cathepsin S and cathepsin K levels were no different in the two groups (Fig. 4d,e), while cathepsin L was significantly decreased in the BAPN group (Fig. 4f). Both COL1α1 and α-SMA expression were dramatically decreased with BAPN treatment (Fig. 4g,h), while CTGF and β-MHC levels were not changed (Fig. 4i,j). These results revealed that BAPN-induced TAD was associated with typical ECM degradation, possibly via MMP2, and loss of SMC leading to decreased α-SMA, effects consistent with previous observations in humans and mouse models.

Discussion
Thoracic aortic dissection, with or without rupture, represents a structural and functional failure of the aortic wall and can occur when integrity of the thoracic aortic wall is impaired 15 . In our study, we found that at the proper dose (0.4 g BAPN per 100 g diet), BAPN disrupted the aortic medial layer without substantially affecting metabolism and body growth, effectively inducing TAD in C57BL/6 mice, a widely used strain. We showed that TAD in C57BL/6 mice resembled clinically observed TAD, with a high incidence of rupture and mortality.
Clinically, TAD, without exception, exhibits medial degeneration, which results from aortic wall stressors such as hypertension, genetics and inflammatory conditions 15 . Medial degeneration is, therefore, essential for rupture of the intima or haemorrhage within the media. Thus, acute blood pressure elevation, such as hypertension induced by Ang II perfusion, is likely to trigger, but not serve as a pathological mediator of, TAD and rupture. Moreover, in a previous report, norepinephrine infusion in BAPN-treated mice failed to induce TAD,   (d-f) qPCR analysis of Cathepsin S/K/L mRNA levels (n = 8 per group); *P < 0.05, **P < 0.01. (g-j) qPCR analysis of COL1α1, α-SMA, CTGF and β-MHC mRNA levels (n = 8 per group); *P < 0.05, **P < 0.01, ***P < 0.0001.
despite elevating blood pressure similarly to Ang II infusion. This suggested that the effects of Ang II on triggering TAD onset were not caused by a blood pressure change alone 8 . To further support this, we found that 87% mice treated only with BAPN developed thoracic aortic dissection and 37% had spontaneous rupture without elevation of systolic blood pressure. As expected, thoracic aortic dissection and rupture were observed in 100% of mice receiving both BAPN and Ang II on both genetic backgrounds, supporting a triggering effect of Ang II after aortic structural disruption by BAPN. Loss of elasticity resulted in decreased diastolic blood pressure 16 , explaining our observation of reduced diastolic blood pressure in mice of both genetic backgrounds given BAPN alone, as compared with controls. Interestingly, spontaneous rupture occurred with mice given the lower doses, but not the higher doses, of BAPN. These mice also showed limited weight increases over time, suggesting that a sufficient hemodynamic load was required for the progression of TAD. Further investigation will be needed to understand the mechanisms involved in this mouse model for BAPN-induced TAD. Three-week-old male mice were fed a normal diet and administered freshly prepared BAPN (Sigma-Aldrich, St. Louis, MO, USA) solution dissolved in the drinking water (1 g/kg/d) for 4 wk, as described previously 8 . Blood pressure was measured before and after BAPN administration for 4 wk, using the tail-cuff method. Interventions lasted 4 wk and body weights were measured weekly. As previously reported, at 7 wk old, osmotic mini pumps (Alzet, Cupertino, CA, USA) administering 1 μg/kg per min Ang II (Sigma-Aldrich, St. Louis, MO, USA) were implanted subcutaneously and mice were euthanized 24 h after implantation 17 . All mice died before expected end time of the experiment were autopsied immediately, and Blood clots were found in the thoracic cavities of these mice. Mice surviving at the end of the experiment were sacrificed by an overdose of sodium pentobarbital and their blood and tissue samples were collected for further analyses.

Patient specimens and ethics statement.
Histopathological analysis. Complete gross and histopathological evaluations were performed with samples from control and BAPN-treated mice. After euthanasia, normal and dissected aortas were harvested from the ascending aorta to the iliac artery and were fixed in 10% buffered formalin, as were human tissues. Fixed, paraffin-embedded tissues were cut at 5 μm thickness, stained with haematoxylin and eosin following standard procedures and examined under light microscopy, as previously described 4 .

Statistical analysis.
In all experiments, data were collected form more than 6 mice per group to calculate means ± standard deviation (SD). Student's t-test was used for statistical analysis and P < 0.05 was considered statistically significant.