**4. Characteristics of BAV-associated ascending aortopathy**

As described above, an association between impaired embryonic signaling between different cardiac progenitor cells and the formation of a BAV has been suggested, likely contributing to aortopathogenesis. The cardiac progenitor cells involved in valvulogenesis migrate and populate the ascending aortic media [18]. The literature has as such focused on the aortic media as causative of aortopathy. The adult ascending aortic media is laminarly structured with VSMCs sandwiched between load-bearing elastin and collagen [34]. Albeit BAV has been known as a risk factor for disease since 1844 [35], it was first during 1984, in necropsy studies by Larson et al., that structural differences between BAV-associated aortopathy and degenerative ascending aortic aneurysm were proposed in light of vastly different rates of acute aortic syndromes [36]. Histologically, one can observe very small differences in the ascending aorta of BAV patients with or without aneurysm, i.e., the elastin is intact, there is VSMC apoptosis, although notably without mucoid extracellular matrix accumulation (MEMA), and the aortic intima-media exhibits few signs of inflammation [10]. Still, the nondilated aorta of individuals with BAV displays a seemingly thinner intima [37].

In past years, researchers have focused on the aortic media in BAV-associated aortopathy with findings of differential VSMC phenotypes in BAV and TAV aortopathies [11, 38]. Not only have these cells been found to undergo apoptosis [39] without apparent MEMA [10], but BAV VSMCs also exhibit distinct morphology. In BAV patients, VSMCs are less differentiated, indicating a defect in the phenotypic switch process, leading to significantly lower expression of differentiated, contractile VSMC markers, such as smoothelin, calponin, and SM22alpha [40, 41]. Additionally, VSMC dissociated from aneurysmal tissue exhibit differences in proliferation and migration comparing BAV and TAV VSMC. Specifically, in an ORIS migration assay, TAV VSMCs showed a faster migration and a higher proliferation rate than BAV VSMCs [10]. Although these cells exhibit such characteristics and behavior in aneurysmal tissue, the less differentiated and immature VSMCs are observed in both nondilated and dilated BAV populations [40], leading the mind to wonder if this VSMC phenotype might itself not be driving aneurysm development.

Interestingly, we and others have observed a mesenchymal-like state of endothelial cells in the ascending aorta of BAV patients, even prior to aortic dilatation [7, 42, 43]. Moreover, the expression of the endothelial-specific marker CD31 is decreased in nondilated BAV aorta, indicating a less differentiated endothelial phenotype [10]. Also, there are signs of a compromised basal membrane, with decreased expression of laminin gamma 1 [10], the main monomer in laminin trimers of large artery basal membranes [44]. This, together with reports of alterations in endothelial junction protein expression in nondilated BAV, such as increased protein turnover of CDH5, decreased expression of CLDN5, and increased mRNA expression of *CDH2* with dilatation compared to TAV patients [45], indeed implicates dysfunctional endothelium in BAV. Electron microscopy further strengthens this observation with signs of junctional degradation and a less intact endothelium in nondilated BAV individuals compared with TAV [45]. The genetic variants and missense mutations of ROBO4 found to associate with BAV aortopathy further strengthen the role of the endothelium in the AscAA development of BAV patients [46], as ROBO4 is an arbiter of vascular integrity and endothelial barrier function [47, 48]. The study by Gould et al. demonstrates the endothelial barrier impairment by infiltration of albumin into the ascending aortic wall [46], but we, too, have observed this to be a general characteristic of BAV ascending aortas, no matter if they are dilated or not [10].

Thus, the endothelium, too, has a seemingly important role in the distinct aortic wall phenotype observed in BAV patients. This is further supported by numerous animal models with endothelial-specific mutations producing offspring with a higher prevalence of BAVs. Most notably, in regards to endothelial function, mice lacking eNOS result in 40% BAV progeny [25]. Mice with GATA5−/− (with 25% BAV progeny) indicate that dysfunctional endothelial phenotype, as they also display lower expression of endothelial-specific markers – CDH5, TIE2, and eNOS, is related to BAV [49]. However, a drawback of these studies is that the prevalence of, or propensity to develop, aortopathy was not investigated. Nonetheless, the endothelial-specificity of eNOS, and the fact that GATA5 is mostly restricted to the endocardium, disappearing at mid-gestation and is required for early differentiation of cardiac progenitors into endothelial/endocardial cells, suggests a connection between the BAV phenotype and disturbed endothelial function.

An observed consequence of impaired endothelial function related to BAV is increased permeability. Ascending aortas of nondilated and dilated BAV patients exhibit greater infiltration of albumin in the aortic intima-media compared to TAV patients, which in nondilated state have a normal functional endothelium [10, 46]. The infiltration of plasma proteins into the aortic wall of BAV patients and potential consequences to VSMC phenotype thereof has to our knowledge, not been investigated. It is, however, a promising line of research in search of circulating biomarkers influencing the cellular phenotype of ascending aortas in BAV patients. One might speculate that such a biomarker might guide practices of surveillance and indications of ascending aortic surgery.
