Copper-Binding Domain Variation in a Novel Murine Lysyl Oxidase Model Produces Structurally Inferior Aortic Elastic Fibers Whose Failure Is Modified by Age, Sex, and Blood Pressure

Lysyl oxidase (LOX) is a copper-binding enzyme that cross-links elastin and collagen. The dominant LOX variation contributes to familial thoracic aortic aneurysm. Previously reported murine Lox mutants had a mild phenotype and did not dilate without drug-induced provocation. Here, we present a new, more severe mutant, Loxb2b370.2Clo (c.G854T; p.Cys285Phe), whose mutation falls just N-terminal to the copper-binding domain. Unlike the other mutants, the C285F Lox protein was stably produced/secreted, and male C57Bl/6J Lox+/C285F mice exhibit increased systolic blood pressure (BP; p < 0.05) and reduced caliber aortas (p < 0.01 at 100mmHg) at 3 months that independently dilate by 6 months (p < 0.0001). Multimodal imaging reveals markedly irregular elastic sheets in the mutant (p = 2.8 × 10−8 for breaks by histology) that become increasingly disrupted with age (p < 0.05) and breeding into a high BP background (p = 6.8 × 10−4). Aortic dilation was amplified in males vs. females (p < 0.0001 at 100mmHg) and ameliorated by castration. The transcriptome of young Lox mutants showed alteration in dexamethasone (p = 9.83 × 10−30) and TGFβ-responsive genes (p = 7.42 × 10−29), and aortas from older C57Bl/6J Lox+/C285F mice showed both enhanced susceptibility to elastase (p < 0.01 by ANOVA) and increased deposition of aggrecan (p < 0.05). These findings suggest that the secreted Lox+/C285F mutants produce dysfunctional elastic fibers that show increased susceptibility to proteolytic damage. Over time, the progressive weakening of the connective tissue, modified by sex and blood pressure, leads to worsening aortic disease.


Introduction
Lysyl oxidase (LOX) is an extracellular copper-dependent enzyme that catalyzes the oxidative deamination of lysine residues in collagen and elastin, resulting in spontaneous condensation with adjacent aldehydes to form inter-and intramolecule covalent crosslinks [1][2][3]. It belongs to a family of five closely related copper-dependent enzymes (LOX, LOXL1, LOXL2, LOXL3, and LOXL4). Proteins in this family have a conserved carboxy-terminal catalytic domain paired with variant amino-terminal domains. The LOX and LOXL1 propeptides must be proteolytically cleaved by procollagen C-peptidases to form an active enzyme of 30 kDa, while LOXL2-4 appear to be active in both processed and nonprocessed forms [4].
For optimal enzyme activity, three histidines in the catalytic domain must coordinate a single copper ion, causing a conformational change that enables the formation of the lysine-tyrosyl-quinone (LTQ) cofactor from lysine and tyrosine downstream from the copper-binding region [5]. In LOX, five pairs of cysteines are involved in intramolecular disulfide bonds. All the Lox cystines are evolutionarily conserved and are thought to be responsible for positioning the copper-binding domain and the LTQ near one another and stabilizing the structure [5,6].
Homozygous disruption of the Lox gene in mice results in death during the perinatal period due to aortic aneurysm [7,8]. These fetuses possess fragmented elastic fibers and discontinuous smooth muscle cell layers [8]. In humans, both nonsense and missense variation in LOX, mostly concentrated around the copper-binding domain, has been associated with FTAA in a small but growing number of pedigrees [8][9][10][11][12][13]. In general, people with LOX-related FTAA (using the dyadic system of naming) [14] do not exhibit vascular disease at birth but begin to develop features of aortic dilation and aneurysm over time. Dilation has been described in the reported patients as early as 6 years of age [11], but most are ascertained due to aneurysm in middle age. While most affected individuals exhibit isolated aortic aneurysm, an individual with a p.Cys291Ser variant [13] was described with multiple dissections of the aorta as well as aneurysm and dilation in extra-aortic vessels, such as the renal and iliac arteries. The orthologous amino acid (p.Cys285) was mutated to phenylalanine (Phe) as part of a mouse ENU screen identifying homozygous variants causing cardiovascular disease manifestations [15]. Our study uses this new mouse model, Lox b2b370.2Clo (c.G854T; p.Cys285Phe), to investigate the progression of Lox-mediated disruption of elastic fibers over the lifetime and to evaluate potential interactive effects of sex and mechanical stress on the condition.

Lox +/C285F Dilation Rate Is Modified by Sex and Blood Pressure
Because hypertension has been shown to be a modifier of dilation rate in other models of aortic dilation [17] (and even in other Lox mutant lines [18]), we crossed the C57 Lox +/C285F onto a congenic high blood pressure (HBP) background. As expected, HBP background increases SBP (Figure 2A, HBP effect p < 0.0001 by two-way ANOVA, see figure for multiple comparison testing (Sidak)), with an average increase of 12 mmHg at 3 months of age. Likewise, the aortic diameter in the male HBP Lox +/C285F is larger at all pressures tested than the C57 Lox +/C285F ( Figure 2B, p < 0.05 or better at each pressure) and appears dilated as compared to C57 Lox +/+ and HBP Lox +/+ mice at a subset of pressures. Of note, no diameter change is seen in the HBP Lox +/+ aortas.
When similar testing is done in female mice, a different pattern emerges. The female C57 Lox +/C285F mice do not show dilation until 12 months of age ( Figure 3A). Adding the HBP background yields both HBP and Lox effects to SBP ( Figure 3B, two-way ANOVA, p < 0.01 and p < 0.001, respectively, multiple comparison testing results shown in graph, Sidak). However, the HBP-mediated SBP increase is less robust in females and does not increase aortic caliber in the 3-month HBP Lox +/C285F ( Figure 3C).
When similar testing is done in female mice, a different pattern emerges. The female C57 Lox +/C285F mice do not show dilation until 12 months of age ( Figure 3A). Adding the HBP background yields both HBP and Lox effects to SBP ( Figure 3B, two-way ANOVA, p < 0.01 and p < 0.001, respectively, multiple comparison testing results shown in graph, Sidak). However, the HBP-mediated SBP increase is less robust in females and does not increase aortic caliber in the 3-month HBP Lox +/C285F ( Figure 3C).
To look for a sex hormone effect, we castrated male HBP Lox +/C285F mice. We found lower SBP in the castrated male HBP Lox +/C285F mutants, relative to the sham HBP Lox +/C285F mice ( Figure 3D, p < 0.05), and similar to HBP Lox +/C285F females. Aortic caliber followed a similar trend ( Figure 3E).

Increased Number of Breaks and Irregular Elastic Lamellar Sheets in Lox +/C285F Mice
Looking at the mutant aortas histologically using EVG, which stains elastin black, we noted a somewhat disorganized elastic lamellar structure in the (male) C57 Lox +/C285F mice ( Figure 4A). No difference in lamellar number was noted, regardless of the Lox genotype, genetic background, or age ( Figure 4B). However, breaks were present in all elastic lamellae, including the internal elastic lamina (IEL). By two-way ANOVA and subsequent Sidak's multiple comparison testing, the Lox +/C285F effect on breaks is not statistically significant in the C57 background at 3 months ( Figure 4C) but is more obvious in the dilated vessels: C57 Lox +/C285F at 6 months (p < 0.05) and HBP background at both 3 months (p < 0.01) and 6 months (p < 0.0001). While an increase in breaks is seen as early as 3 months, increased medial thickness is not noted until 6 months in the HBP Lox +/C285F ( Figure 4D, p < 0.0001). Multivariable linear regression shows that breaks are most strongly influenced by Lox genotype (p = 2.8 × 10 −8 ) with more minor effects from age (p < 0.05) and genetic background (p = 6.8 × 10 −4 , See Supplemental Table S1). Age had the strongest effect on wall thickness (p = 2.5 × 10 −5 ), followed by Lox genotype (p = 1.5 × 10 −3 ). For both phenotypes, a synergistic effect was seen among Lox +/C285F status, HBP background, and older age (Supplemental Table S1).

Increased Number of Breaks and Irregular Elastic Lamellar Sheets in Lox +/C285F Mice
Looking at the mutant aortas histologically using EVG, which stains elastin black, we noted a somewhat disorganized elastic lamellar structure in the (male) C57 Lox +/C285F mice ( Figure 4A). No difference in lamellar number was noted, regardless of the Lox genotype, genetic background, or age ( Figure 4B). However, breaks were present in all elastic lamellae, including the internal elastic lamina (IEL). By two-way ANOVA and subsequent Sidak's multiple comparison testing, the Lox +/C285F effect on breaks is not statistically significant in the C57 background at 3 months ( Figure 4C) but is more obvious in the dilated vessels: C57 Lox +/C285F at 6 months (p < 0.05) and HBP background at both 3 months (p < 0.01) and 6 months (p < 0.0001). While an increase in breaks is seen as early as 3 months, increased medial thickness is not noted until 6 months in the HBP Lox +/C285F ( Figure 4D, p < 0.0001). Multivariable linear regression shows that breaks are most strongly influenced by Lox genotype (p = 2.8 × 10 −8 ) with more minor effects from age (p < 0.05) and genetic background (p = 6.8 × 10 −4 , See Supplemental Table S1). Age had the strongest effect on wall thickness (p = 2.5 × 10 −5 ), followed by Lox genotype (p = 1.5 × 10 −3 ). For both phenotypes, a synergistic effect was seen among Lox +/C285F status, HBP background, and older age (Supplemental Table S1).  and HBP backgrounds showed an increasingly disorganized and thick vascular wall, most obvious in the 6-month HBP Lox +/C285F (Lox effect p < 0.0001, age effect p < 0.05 and HBP effect <0.001, multivariable regression, see Supplemental Table S1. * p < 0.05; ** p < 0.01; **** p < 0.0001. To further characterize the eroding aortic wall, fresh ascending aorta of 3-and 6-month HBP Lox +/+ and Lox +/C285F were imaged using two-photon en-face imaging. The image stacks were reconstructed to produce a volumetric representation of the internal elastic lamina and the subsequent 2-3 lamellae. HBP Lox +/+ elastic lamina ( Figure 5A) are smooth, nearly continuous elastic sheets with only occasional small fenestrae. However, like the light microscopy imaging, HBP Lox +/C285F mice reveal increased fenestrations (or breaks as seen in 2D) ( Figure 5A). Cross-sectional images of the XY and XZ planes confirm that these fenestrations do not merely represent an invagination of the elastic layer ( Figure 5B and Supplemental Videos S1-S6). The fenestrae become more numerous with increasing age. In addition, even at 3 months, the lamellar structure in the Lox +/C285F is noticeably more disorganized, with a frayed appearance that becomes more obvious with age. Similar findings are seen deeper in the vessel wall as well (Supplemental Figure S1).
ing age. In addition, even at 3 months, the lamellar structure in the Lox +/C285F is noticeably more disorganized, with a frayed appearance that becomes more obvious with age. Similar findings are seen deeper in the vessel wall as well (Supplemental Figure S1). When viewed at the ultrastructural level using FIB-SEM, 6-month-old HBP Lox +/+ mice have a smooth and continuous internal elastic lamina (IEL, Figure 5C, and Supplemental Video S7), while the HBP Lox +/C285F IEL shows more discontinuity ( Figure 5D,E, and Supplemental Video S8) with frequent partial thickness invaginations, fenestrations, and even disassociated "floating" elastin segments. The video reconstructions show 0 vs. 7 fenestrations in the IEL alone in a similar cross-sectional space for HBP Lox +/+ and Lox +/C285F , respectively. In some regions, the HBP Lox +/C285F IEL is relatively intact, while in others, it lacks almost any organizational structure ( Figure 5F and Supplemental Video S9). Overall, the Lox +/C285F lamellae appear to be less tightly woven, with some fibers appearing fractured or split internally ( Figure 5E and seen more clearly in Supplemental Video S8). The cells in the HBP Lox +/+ aortas are cuboidal and closely packed, while those from the HBP Lox +/C285F are irregular with additional surrounding accumulated nonfibrillar extracellular material. In some HBP Lox +/C285F cases, multiple flattened SMCs are seen layered on one another without an intervening elastic lamella.  Figure S1 for images of the second lamella). Representative images, n = 3. (B), Still images demonstrate fenestrae in 3D by crossing XY (10 µm) and XZ (5 µm) planes of the images in Lox +/C285F mice (see also Supplemental Videos S1-S6).
When viewed at the ultrastructural level using FIB-SEM, 6-month-old HBP Lox +/+ mice have a smooth and continuous internal elastic lamina (IEL, Figure 5C, and Supplemental Video S7), while the HBP Lox +/C285F IEL shows more discontinuity ( Figure 5D,E, and Supplemental Video S8) with frequent partial thickness invaginations, fenestrations, and even disassociated "floating" elastin segments. The video reconstructions show 0 vs. 7 fenestrations in the IEL alone in a similar cross-sectional space for HBP Lox +/+ and Lox +/C285F , respectively. In some regions, the HBP Lox +/C285F IEL is relatively intact, while in others, it lacks almost any organizational structure ( Figure 5F and Supplemental Video S9). Overall, the Lox +/C285F lamellae appear to be less tightly woven, with some fibers appearing fractured or split internally ( Figure 5E and seen more clearly in Supplemental Video S8). The cells in the HBP Lox +/+ aortas are cuboidal and closely packed, while those from the HBP Lox +/C285F are irregular with additional surrounding accumulated nonfibrillar extracellular material. In some HBP Lox +/C285F cases, multiple flattened SMCs are seen layered on one another without an intervening elastic lamella.
Collagen detection using fluorescence microscopy shows no obvious increase in collagen fibers (data not shown). Likewise, no difference in total insoluble elastin or collagen content was detectable by amino acid analysis (AAA) ( Figure 6A,B) of hydrolyzed tissue. (C-F), Still images from FIB-SEM show an intact IEL in Lox +/+ (C). In contrast, Lox +/C285F mice exhibit pathology ranging from significant disruptions/holes (arrows), disconnected "floating" segments (circled), and even sheared (red dotted line) stretches (D,E) to a severe, completely disorganized IEL (F) (See 3D reconstructed movies in Supplemental Videos S7-S9).
Collagen detection using fluorescence microscopy shows no obvious increase in collagen fibers (data not shown). Likewise, no difference in total insoluble elastin or collagen content was detectable by amino acid analysis (AAA) ( Figure 6A,B) of hydrolyzed tissue.

Mechanism of Lox-Mediated Disease
To better understand the mechanism by which the copper-binding domain Lox +/C285F mutant produces the phenotypes above, we evaluated gene expression and protein production. Quantitative PCR reveals similar amounts of Lox transcript in p14 aortas from HBP Lox +/+ and Lox +/C285F mice ( Figure 7A). Similarly, there is no appreciable compensatory upregulation of any of the other Lox-L genes ( Figure 7A).

Mechanism of Lox-Mediated Disease
To better understand the mechanism by which the copper-binding domain Lox +/C285F mutant produces the phenotypes above, we evaluated gene expression and protein production. Quantitative PCR reveals similar amounts of Lox transcript in p14 aortas from HBP Lox +/+ and Lox +/C285F mice ( Figure 7A). Similarly, there is no appreciable compensatory upregulation of any of the other Lox-L genes ( Figure 7A).
Protein lysates from aortas were then assessed for Lox activity using Amplex Red [19]. Lox enzyme activity was readily detected in the 3-month-old HBP aortas and showed linear kinetics (Supplemental Figure S2). In aortas from Lox +/C285F mice, the rate of substrate oxidation was lower by 46% as compared to Lox +/+ ( Figure 7B, p < 0.0001), indicating lower enzymatic activity.
To determine if the decreased Lox activity was due to lower total Lox protein in the C285F mutants, aortas were collected from Lox +/+ , Lox +/C285F , and/or Lox C285F/C285F animals at the age of E19, P15 (Lox +/+ and Lox +/C285F only), and P90 (Lox +/+ and Lox +/C285F only). Protein lysates prepared from these samples were used for the characterization of Lox synthesis and secretion. Lox protein is synthesized in the form of inactive zymogens. Its activation requires a functional secretory pathway and the proteolytic removal of the N-terminal propeptide to generate the 30 kDa active enzyme. Comparing genotypes by Western blot (Figure 7C), at age E19, the Lox C285F/C285F aortas exhibit 14% of the average Lox +/+ quantity of mature Lox ( Figure 7D, p < 0.01 in Lox +/+ vs. Lox C285F/C285F ). However, there is a 2.5-fold increase in proLox protein detectable in Lox C285F/C285F aortic tissue ( Figure 7D, p < 0.01 in Lox C285F/C285F vs. Lox +/+ ). For the Lox +/C285F aortas at all three ages, mature Lox remains lower than in the Lox +/+ vessels ( Figure 7D, p < 0.05 at E19; Figure 7E, p < 0.01 at P15; and Figure 7F, p < 0.05 at P60). There is no statistically significant difference in proLox, although an upward trend is noted in the three comparisons. , Lox +/C285F aorta showed 46% reduction in Lox enzyme activity in Lox +/285F mice as measured by Amplex Red (t-test; n = 8). **** p < 0.0001. (C), Total lysate from frozen aortas at age E19, P15, and P90 demonstrated higher quantities of 50 kDa proLox in Lox C285F/C285F as compared to Lox +/+ at age E19 only. However, a decreased amount of 30 kDa mature Lox in Lox C285F/C285F and/or Lox +/C285F was found in all three age groups. Graphs show the quantification of proLox and mature Lox at age (D), E19, (E), P15 and (F), P90 across all samples using total protein as normalization ((D) one-way ANOVA (p < 0.01 in both comparisons), Dunnet comparisons shown in figure; (E-F) Paired t-test). * p < 0.05; ** p < 0.01. (G) In conditioned medium collected from MEF isolated from Lox wild-type and mutant mice (n = 3), comparable amount of secreted mature Lox protein was detected. (Right panel, one-way ANOVA, Dunn).
Protein lysates from aortas were then assessed for Lox activity using Amplex Red [19]. Lox enzyme activity was readily detected in the 3-month-old HBP aortas and showed linear kinetics (Supplemental Figure S2). In aortas from Lox +/C285F mice, the rate of substrate oxidation was lower by 46% as compared to Lox +/+ ( Figure 7B, p < 0.0001), indicating lower enzymatic activity.
To determine if the decreased Lox activity was due to lower total Lox protein in the C285F mutants, aortas were collected from Lox +/+ , Lox +/C285F , and/or Lox C285F/C285F animals at the age of E19, P15 (Lox +/+ and Lox +/C285F only), and P90 (Lox +/+ and Lox +/C285F only). Protein lysates prepared from these samples were used for the characterization of Lox synthesis and secretion. Lox protein is synthesized in the form of inactive zymogens. Its activation requires a functional secretory pathway and the proteolytic removal of the N-terminal propeptide to generate the 30 kDa active enzyme. Comparing genotypes by Western blot (Figure 7C), at age E19, the Lox C285F/C285F aortas exhibit 14% of the average Lox +/+ quantity of mature Lox ( Figure 7D, p < 0.01 in Lox +/+ vs. Lox C285F/C285F ). However, there is a 2.5-fold increase in proLox protein detectable in Lox C285F/C285F aortic tissue ( Figure 7D, p < 0.01 in Figure 7. Mutation of Lox at C285F did not cause reduction in gene expression and Lox protein secretion as well as enzyme activity. (A), Lox and LoxL family members had similar mRNA expression in both Lox +/C285F and Lox +/+ mice as measured by quantitative PCR (t-test). (B), Lox +/C285F aorta showed 46% reduction in Lox enzyme activity in Lox +/285F mice as measured by Amplex Red (t-test; n = 8). **** p < 0.0001. (C), Total lysate from frozen aortas at age E19, P15, and P90 demonstrated higher quantities of 50 kDa proLox in Lox C285F/C285F as compared to Lox +/+ at age E19 only. However, a decreased amount of 30 kDa mature Lox in Lox C285F/C285F and/or Lox +/C285F was found in all three age groups. In aortic tissue, it is not possible to distinguish intra-vs. extracellular protein. To examine whether the mutant Lox protein is adequately secreted, conditioned media were collected from Lox +/+ , Lox +/C285F , and Lox C285F/C285F MEF lines, and Lox protein was quantified ( Figure 7G). Minimal proLox was seen in the conditioned media, and the amount of 30 kDa mature Lox was similar across genotype groups (quantification in Figure 7G), indicating equivalent secretion of the C285F-mutated Lox protein.

Lox +/C285F Vessels Are Susceptible to Elastolytic Damage and Show Increased Elastase Production
Together with the imaging depicting looser, less organized elastic lamellae and abnormal cell morphology, the combination of normal to reduced extracellular levels of mature Lox (depending on the system) and normal elastin and collagen concentrations suggest a change in the quality, rather than quantity, of elastic fibers and predict that the elastin deposited by Lox +/C285F mutants is structurally inferior, leading to abnormal cell-matrix interactions and a more rapid failure rate.
To test quality of the elastic matrix, we placed vessels from C57 Lox +/+ and Lox +/C285F mice on a pressure myograph fitted with a flutter valve to allow a pulsatile flow in the vessel. We then bathed the vessel in a low concentration of elastase and monitored the vessel diameter. When assessed at pressures in the elastic portion of the curve [20,21], we found that C57 Lox +/C285F mutant vessels had a more rapid increase in diameter with elastase treatment (Figure 8A (aorta; two-way ANOVA genotype effect p < 0.01) and B (carotid; two-way ANOVA genotype effect p < 0.01)). gest a change in the quality, rather than quantity, of elastic fibers and predict t elastin deposited by Lox +/C285F mutants is structurally inferior, leading to abnorm matrix interactions and a more rapid failure rate.
To test quality of the elastic matrix, we placed vessels from C57 Lox +/+ and L mice on a pressure myograph fitted with a flutter valve to allow a pulsatile flow vessel. We then bathed the vessel in a low concentration of elastase and monito vessel diameter. When assessed at pressures in the elastic portion of the curve [20, found that C57 Lox +/C285F mutant vessels had a more rapid increase in diameter wi tase treatment ( Figure 8A (aorta; two-way ANOVA genotype effect p < 0.01) and rotid; two-way ANOVA genotype effect p < 0.01)).

Lox +/C285F Gene Expression Pattern Reveals Increased Matrix Remodeling
To identify additional differentially regulated pathways between Lox +/+ and Lox +/C285F aortas, we performed RNAseq on HBP p14 aortic tissue ( Figure 9A, Supplemental Table S2). Gene set enrichment showed increased expression of genes associated with the matrisome (Naba matrisome q = 4.9 × 10 −14 and Naba matrisome associated q = 1.14 × 10 −11 ) [22] along with a series of cytokine and cellular signaling transduction pathways (Supplemental Table S3) in the Lox +/C285F mutants. As predicted by our protein studies, Cela1, the gene encoding elastase-1, was upregulated in the Lox +/C285F aortas. In addition, several fibroblast growth factors and serpin species, as well as the proteoglycan aggrecan, were noted in the matrix and matrix-associated sublist of upregulated genes.

Discussion
Lysyl oxidase is a copper-binding enzyme known to cross-link elastin and collagen. The clinical literature has associated a variation in this gene with dominantly inherited FTAA (MIM#617168). In most cases of FTAA, no obvious aortic phenotype is described at birth but develops over the following years to decades. Multiple pedigrees with missense variants in close proximity to the copper-binding domain have been described, with a recently published case demonstrating an individual with a p.Cys291Ser variant, whose case was particularly severe. The mouse model used here (Lox +/c.G854T (b2b370.2Clo); Lox +/C285F ) impacts the analogous Cys in mice and can be used to probe disease mechanisms and to decipher the synergistic impact of influences, such as age, sex, and elevated blood pressure.
Compared to previously published Lox mutants [7,8], the vascular features in the Lox +/C285F mouse are more pronounced. Like the others, the aortas are initially of smaller Aggrecan, a molecule previously noted to be increased in aneurysmal tissues [23], increases asymmetrically in the HBP 6-month Lox +/C285F animals around the time we detected aortic thickening ( Figure 9B,C). Looking at the genes that are decreased in Lox +/C285F mutants (Supplemental Table S4), we again see enrichment for genes in the matrisome, but the significance is less robust (q = 1.70 × 10 −3 ). The downregulated list includes molecules involved in insulin signaling and a set of metallopeptidases and metallopeptidase inhibitors. Interestingly, when the differentially expressed genes were analyzed using Ingenuity Pathway Analysis (IPA) to look for patterns of expression based on upstream regulators, we saw relative inhibition of genes known to be controlled by dexamethasone in Lox +/C285F (z = −3.97, p = 9.83 × 10 −30 ) as well as relative activation of genes influenced by TGFβ (z = 2.44, p = 7.42 × 10 −29 ). See Supplemental Table S5 for a full list of regulators.

Discussion
Lysyl oxidase is a copper-binding enzyme known to cross-link elastin and collagen. The clinical literature has associated a variation in this gene with dominantly inherited FTAA (MIM#617168). In most cases of FTAA, no obvious aortic phenotype is described at birth but develops over the following years to decades. Multiple pedigrees with missense variants in close proximity to the copper-binding domain have been described, with a recently published case demonstrating an individual with a p.Cys291Ser variant, whose case was particularly severe. The mouse model used here (Lox +/c.G854T (b2b370.2Clo); Lox +/C285F ) impacts the analogous Cys in mice and can be used to probe disease mechanisms and to decipher the synergistic impact of influences, such as age, sex, and elevated blood pressure.
Compared to previously published Lox mutants [7,8], the vascular features in the Lox +/C285F mouse are more pronounced. Like the others, the aortas are initially of smaller caliber and display evidence of vascular stiffness (increased pulse pressure and reduced slope of the pressure-diameter curve, although the difference is noted down to 75 mmHg in the Lox +/C285F mutant and only detectable at 150 mmHg and above in the previous models). Systolic blood pressure is also increased in the Lox +/C285F relative to Lox +/+ males. Over time and without pharmacologic provocation, the Lox +/C285F aorta dilates. The largercaliber vessel is first apparent by 6 months in males and continues to progress, although frank aneurysm does not occur. This dilation may be accelerated by the introduction of a small amount of 129X1/Sv × genetic material on chromosome 1 in an otherwise C57Bl/6 genetic background. This area had been linked to a QTL for blood pressure in previous studies [24] and does raise the blood pressure of the male mice. Hypertension is a known risk factor for aneurysm progression [25], and lowering the blood pressure by treatment with either beta blockers or angiotensin receptor blockers slowed the aneurysm progression rate in patients with Marfan syndrome [26]. Likewise, as in humans [27], female Lox +/C285F mice do not develop dilation as quickly as males and are less influenced by the HBP background. Castration of male mice leads to milder phenotypes, similar to females, suggesting either a provocative role for testosterone or a protective one for estrogen in the dilation phenotype [28]. Previous studies in the Col3A1-related vascular Ehlers-Danlos mouse model also show sex-based differences in aneurysm [29]. In this case, both androgen and perinatal oxytocin were linked to the outcomes. We did not specifically study pregnancy as a modifier of aortic outcomes in this study, but we do note that while unmated females were used for the 3-and 6-month experiments, retired breeders were used for the 12-month studies, suggesting a possibility for similar mechanisms to play a role in the late dilation seen in the female Lox +/C285F animals.
On the molecular level, our studies showed that the mutant message is stable, and full-length protein is produced by cells. Moreover, the Lox +/C285F protein was adequately produced, secreted, and cleaved such that the conditioned media from MEFs contained normal to increased amounts of mature 30 kDa protein. Lox +/C285F aortic tissue, on the other hand, contained relatively less mature enzyme, suggesting the possibility of increased turnover of the mutant mature protein in the in vivo setting.
Previous studies have shown the importance of Lox's Cys pairs for positioning the molecule's CBD relative to the LTQ [6,30]. As such, even if increased turnover does not occur, the loss of a Cys proximal to the CBD can reasonably be expected to directly impact activity. Indeed, the activity assay shows a marked reduction (−46%) in Lox activity in heterozygous aortic tissue. Nevertheless, altered kinetics of enzyme activation or release from the developing elastic fiber could also be considered. To become active, Lox must be targeted to the elastic fiber by way of its prodomain [31]. Once there, its prodomain is removed by a procollagen C peptidase, activating it. It was previously reported that BMP1 s ability to process type 1 procollagen is enhanced by its binding to fibronectin [32], and fibronectin is known to interact with the C-terminal portion of Lox [33]. As such, it is possible that, in addition to influencing the catalytic function of the enzyme directly, the variation that impacts the C-terminus may also impact the complex interaction among the molecules required for either Lox's activation or the duration of its occupancy on the elastic fiber, invoking a dominant negative mechanism.
However, how does that decreased activity produce the aneurysm phenotype? Considering our light, two-photon, and scanning electron microscopy together, we first see increased elastic lamellar fenestrations in the Lox +/C285F mutant, followed by later vessel wall thickening and loss of normal smooth muscle cell appearance. These holes are subtle in young animals and increase with age and HBP influences. Studies in the Fbn1 C1039/+ Marfan mice also show increased fenestrations [34], but unlike the Marfan mice, in which the elastic lamellae remain smooth and regular by two-photon imaging, the Lox +/C285F lamellae appear increasingly ragged. The elastic sheet appears less tightly woven, with a disorganized and frayed façade, even at early ages, suggesting that although the apparent vascular defect is mild at young ages, the elastic fibers are not deposited normally. That initial defect is then amplified with increasing time and pressure that unravels the imperfectly woven elastic sheet.
In addition, our protease experiments reveal the Lox +/C285F mutants to be overly susceptible to proteolytic damage, with more rapid dilation in response to elastase treatment than in Lox +/+ aortas. Intriguingly, the Lox +/C285F vessels themselves exhibit increased elastase. In many aneurysm models where the ECM is obviously abnormal, the tissue attempts to compensate by remodeling. Correspondingly, our gene expression studies show enrichment for matrisome genes, as well as genes known to be regulated by TGFβ and dexamethasone. Perturbation of TGFβ [35,36] is well-known in aneurysm models, and dexamethosone [37,38] is known to influence elastic fiber assembly.
The FIB-SEM imaging takes this a step further, revealing not only fenestrations, but also disconnected and fractured sections of elastic fibers in some areas of the Lox +/C285F aorta and a complete loss of the lamellar structure in others. What is also prominent in these samples from older mice is the loss of normal smooth muscle cellular structure and the accumulation of a nonelastin ground substance in the interlamellar space. Studies have suggested that proteoglycan accumulation is associated with vessel rupture and dissection risk [23,[39][40][41]. Aggrecan, in particular, was found to be accumulated in a mouse model of severe Marfan syndrome [23,42]. These mice died from thoracic aortic aneurysm and dissection (TADD) early with an average survival of 2.5 months and demonstrated increased accumulation of aggrecan when compared to their wild-type littermates. Most importantly, the mice that died of aortic rupture exhibited the highest amount of aggrecan accumulation, which spanned the full thickness of the ascending aorta. This demonstrated that the degree of aggrecan accumulation, either due to increased production or decreased degradation, correlates with the severity of aortic dilation and rupture [23]. In our immunofluorescence experiments, increased aggrecan deposition was detected in tissues of 6-month-old HBP Lox +/C285F mice, corresponding with the period of prominent aortic wall thickening and dilation in this genotype. The aggrecan in the Lox +/C285F mutant is deposited asymmetrically, similar to what is seen in mice with severe Marfan syndrome [23] as well as in Eln +/mice [43] where its location was thought to be impacted by differential mechanical forces, suggesting an attempt at a compensatory response to the elastic fiber destruction [23].
Taken together, these studies extend the body of data linking missense Lox variants proximal to the copper-binding domain with aortic disease. The data support a mechanism ( Figure 10) whereby the Lox +/C285F mutants deposit structurally incompetent elastic fibers that are more susceptible to degradation with time and environmental stressors. Increases in elastase and aggrecan accumulation in the extracellular matrix further disrupt the incompetent elastic fibers and cause aortic dilation, resulting in aortic diseases. As such, patients with pathogenic variation within this domain are at increased risk of aortic disease that increases with age and may benefit from disease-modifying therapies, such as blood pressure control [27], with consideration for potential hormonal strategies requiring further investigation. Figure 10. Lox C285F model results in structurally incompetent elastic fibers that are more susceptible to proteolytic damage causing progressive elastic lamellar damage and altered cell-matrix interactions. In the top row, Lox +/+ mice exhibit structurally competent elastic lamella (green Lox enzyme) with the expected mild degradation that accompanies aging and a lifetime of the mechanical stress associated with repetitive stretch-recoil cycles. The Lox +/C285F mutant mice, however, deposit structurally incompetent elastic fibers due to the mutated form of the enzyme (smaller purple Lox C285F with reduced enzyme activity). The abnormal lamellae are unusually susceptible to proteolytic damage and undergo increased destruction that is amplified by elevated blood pressure, male sex, and aging. An uptick of elastase (yellow lightning bolt), a morphologic change in the smooth muscle cells, and infiltration of the elastic lamellae with proteoglycans (blue dots) occur as part of the process. Of note, the aggrecan deposition occurs somewhat later, after elastic fiber breaks are more numerous. Although the deposition of proteoglycans may initially be a compensatory response, the molecules may ultimately further disrupt cell-matrix interactions leading to further vascular wall dysfunction.
Our study demonstrated a potential mechanism by which Lox variants near the copper-binding domain induce changes to the aortic architecture and function. The process is impacted by complex interactions between elastic fibers and smooth muscle cells over time and in response to ongoing physiological stress. To truly understand this process, it would be informative to perform multimodal physiological and imaging analyses in the same animals over time; such work may illustrate important associations between realtime physiological parameters on aortic outcomes. Of particular interest in understanding the sex effects would be an investigation of pregnancy-associated outcomes in females. On the molecular side, previous investigators have noted changes in multiple Lox subtypes in the walls of aneurysmal vessels [44]; our study does not show changes in mRNA expression of the various Lox types in this mutant, but protein level quantification would be needed to rule this out. Likewise, proteomic investigation of the vessel wall may provide further insight into the impact of proteolytic damage to the vessel wall. Information about the quantity and quality of elastin crosslinks remains unknown.
In conclusion, missense variants near the copper-binding domain produce highly penetrant, early-onset autosomal dominant thoracic aortic aneurysm in humans. Work presented here shows that mutations of this type in mice produce a secreted protein that generates irregular elastic fibers that are increasingly susceptible to proteolytic damage over the lifetime of the animal, leading to aortic dilation. Complex physiological factors, such as sex, blood pressure, and age, influence these effects. Additional work is needed at Figure 10. Lox C285F model results in structurally incompetent elastic fibers that are more susceptible to proteolytic damage causing progressive elastic lamellar damage and altered cell-matrix interactions. In the top row, Lox +/+ mice exhibit structurally competent elastic lamella (green Lox enzyme) with the expected mild degradation that accompanies aging and a lifetime of the mechanical stress associated with repetitive stretch-recoil cycles. The Lox +/C285F mutant mice, however, deposit structurally incompetent elastic fibers due to the mutated form of the enzyme (smaller purple Lox C285F with reduced enzyme activity). The abnormal lamellae are unusually susceptible to proteolytic damage and undergo increased destruction that is amplified by elevated blood pressure, male sex, and aging. An uptick of elastase (yellow lightning bolt), a morphologic change in the smooth muscle cells, and infiltration of the elastic lamellae with proteoglycans (blue dots) occur as part of the process. Of note, the aggrecan deposition occurs somewhat later, after elastic fiber breaks are more numerous. Although the deposition of proteoglycans may initially be a compensatory response, the molecules may ultimately further disrupt cell-matrix interactions leading to further vascular wall dysfunction.
Our study demonstrated a potential mechanism by which Lox variants near the copper-binding domain induce changes to the aortic architecture and function. The process is impacted by complex interactions between elastic fibers and smooth muscle cells over time and in response to ongoing physiological stress. To truly understand this process, it would be informative to perform multimodal physiological and imaging analyses in the same animals over time; such work may illustrate important associations between real-time physiological parameters on aortic outcomes. Of particular interest in understanding the sex effects would be an investigation of pregnancy-associated outcomes in females. On the molecular side, previous investigators have noted changes in multiple Lox subtypes in the walls of aneurysmal vessels [44]; our study does not show changes in mRNA expression of the various Lox types in this mutant, but protein level quantification would be needed to rule this out. Likewise, proteomic investigation of the vessel wall may provide further insight into the impact of proteolytic damage to the vessel wall. Information about the quantity and quality of elastin crosslinks remains unknown.
In conclusion, missense variants near the copper-binding domain produce highly penetrant, early-onset autosomal dominant thoracic aortic aneurysm in humans. Work presented here shows that mutations of this type in mice produce a secreted protein that generates irregular elastic fibers that are increasingly susceptible to proteolytic damage over the lifetime of the animal, leading to aortic dilation. Complex physiological factors, such as sex, blood pressure, and age, influence these effects. Additional work is needed at the population level to ascertain the impact of variants further away from the copper-binding domain.

Mouse Strains and Breeding
Experiments were approved by the NHLBI animal studies committee. The Lox +/c.G854T mouse (MGI 5313524(b2b370.2Clo); JAX #013616-C57Bl/6J-b2b370Clo; Lox +/C285F ) was created through the Bench to Bassinet program using ENU mutagenesis of C57Bl/6 mice [15]. Two mutations were present in the parental line (b2b370.1Clo and b2b370.2Clo, http://www.informatics.jax.org/allele/MGI:5313524, accessed on 22 April 2022). The Lox variant (b2b370.2), when present in homozygous state, causes stenosis of the great arteries, cardiac hypertrophy, and diaphragmatic hernia. The identity of b2b370.1Clo was never identified genetically; mice homozygous for the b2b370.1Clo allele exhibited holoprosencephaly, right-sided aortic arch, and hypoplastic proximal arteries. Sperm from males that had produced Lox G854T/G854T offspring were stored at Jackson labs for further study. It is unknown whether these fathers also carried the unknown b2b370.1Clo allele. The line studied here was rederived at Jax by fertilization of C57Bl/6J eggs with these sperm. The resultant pups were genotyped using TaqMan SNP Genotyping Assay kit (ThermoFisher Scientific, Waltham, MA, USA, Assay number 4332077 custom: mLoxG854T) and were backcrossed further to C57Bl6/J to decrease the likelihood of cotransmission of the second unknown allele. After more than 10 generations of backcrossing or sibling breeding, no mice with holoprosencephaly were identified. Consequently, the line reported here is thought to represent only the effect of Lox variation. We refer to this mouse throughout the paper as C57 Lox +/C285F .
To increase blood pressure, the C57 Lox +/C285F mouse was bred to a majority C57Bl/6 mouse strain that also carries 129X1/Sv material in a specific region of chromosome 1 (minimal interval for 129/Sv material Chr1: 086182722-137503552 and maximum interval Chr1:082250512-140519860 (NCBI37/mm9 numbering)) based on SNP genotyping in the region. Mice with 129X1/Sv genetic material in this region were shown in quantitative trait locus studies to have higher blood pressure, with homozygosity for 129X1/Sv raising blood pressure by 15-20 mm Hg over C57Bl/6J [24]. Mice carrying homozygous 129X1/Sv material on chromosome 1 in this area are referred to as HBP Lox +/+ or HBP Lox +/C285F .
For the described experiments, mice were phenotyped at three, six, or twelve months of age. Unmated littermates were used whenever possible, although not all animals could be assessed for all phenotypes. Animals were housed in group cages under standard conditions.

Systemic Blood Pressure and Heart Rate Measurement
Blood pressure was measured as previously described [45]. Briefly, once a level plane of anesthesia was achieved with isoflurane (Isoflurane florane, Baxter, Deerfield, IL, USA), a pressure catheter (1.0-F, model SPR-1000, Millar Instruments, Houston, TX, USA) was inserted into the right carotid and advanced to the ascending aorta. Systolic and diastolic pressures were recorded using Chart 5 software (ADInstruments, Sydney, Australia). Animals were monitored closely for discomfort or over-sedation.

Castration
At postnatal day 20, mice were anesthetized with 1-3% isoflurane. Hair was removed from the abdomen. The skin was prepared with surgical scrub/alcohol, and aseptically draped. A midline skin incision was made on the lower ventral abdomen to expose the abdominal cavity. Each vas deferens was then identified to locate the testicles. Testicles were moved into the abdominal cavity and dissected free from the interior scrotal wall. The artery and vein were cauterized proximal to the testicles, and the testicles were removed. The abdomen was closed using a simple uninterrupted suture and the skin closed with clips. Mice were given 1mg/kg Buprenorphine subcutaneously pre-operatively (ZooPharm SR-LAB, Fort Collins, CO, USA) and 4-5 drops of Bupivacaine (Fresenius Kabi, 460417, Lake Zurich, IL, USA) on the wound prior to closing the skin and allowed to recover. All animals were monitored post-op for pain or discomfort.

Histology
The ascending aorta was perfused with phosphate buffered saline (PBS), excised, and fixed in 1 mL of 10% buffered formalin (Fisher Scientific, SF100-4, Waltham, MA, USA) for 24 h, before dehydration in ethanol. Vessels were then embedded in paraffin and crosssectional rings cut from just proximal to the innominate down to the root. Representative sections along this stretch were stained with Elastic Tissue Fibers-Verhoeff Van Gieson (EVG, Poly Scientific R&D, k059, New York, NY, USA) stain according to the manufacturer's instructions to visualize elastin. Slides were scanned on a NanoZoomer 2.0-RS digital slide scanner (Hamamatsu Photonics, Hamamatsu City, Japan) and analyzed using NDP.view2 (Hamamatsu Photonics, Hamamatsu City, Japan) viewing software. In each quadrant, lamellar number was manually counted and wall thickness measured using a preinstalled ruler in the NDP.view2 software. Total number of breaks in the elastic lamella were also manually quantified in each section.

Two-Photon Microscopy Imaging and Analysis
A 120 µm thick adhesive spacer (Electron Microscopy Sciences, 70327-8S, Hatfield, PA, USA) was placed on a glass slide, and 9 µL of PBS was pipetted in the center. The ascending aorta was then dissected from each mouse and rinsed with PBS to remove blood. The vessel was immediately cut lengthwise, opened flat, and mounted with intima (endothelial cell side) facing up. A #0 cover slip (Electron Microscopy Sciences, 72198) was placed gently on top on the tissue, which allowed en-face imaging of the vessel surface. PBS was used as mounting media. With the optimized setup [46], acquisition of images consisted of an en-face z-stack of images running from intima to about 100 µm depth (or about 4 elastic lamellae) where the elastin autofluorescence signal became too low for detection.
Two-photon microscopy imaging using an inverted Leica SP5 five channel confocal and multiphoton MP-OPO system (Leica Microsystems, Mannheim, Germany) was performed as previously described [46]. Two-photon mode was used with a pulsed femtosecond Titanium:Sapphire (Ti:Sa) laser, (Chameleon Vision II, Coherent, Santa Clara, CA, USA) tunable for excitation from 680 to 1080 nm. Imaging of freshly prepared, whole mounts "en-face" aorta preparations was performed using Leica HC-PL-IRAPO 40X/1.1 NA water immersion objective (WD = 0.6 mm). Two-photon excitation at 860nm was used to reveal structural information by intrinsic contrast imaging of second harmonic generated signal (SHG) collected via 525/40 nm emission filter on nondescanned detectors 1 (NDD1). Aorta autofluorescence from elastin was collected with 525/40 nm emission filter on NDD2.
For 3D volume rendering, series of xyz images (typically 1 × 1 × 1.5 µm 3 voxel size) were collected along the z-axis at 1.5 µm intervals over a range of depths (80-120 µm) throughout the depth of whole mount tissue and over large regions using the tile function of the Leica LAS-AF software to automatically generate stitched volumes comprising an area of approximately 2.0 × 1.2 mm 2 (x-y) and 100 µm (z). For 3D renderings and quantitative image analyses, we used Imaris v 9.5.1 software (Bitplane Inc., Zurich, Switzerland).

Focused Ion Beam-Scanning Electron Microscopy (FIB-SEM)
For the aorta samples, Lox +/+ and Lox +/C285F were processed as previously described [47], with the exception that a 30 min room temperature pyrogallol incubation (320 mM aqueous solution, pH 4.1, Alfa Aesar, Fisher Scientific, 44152-09) was used instead of thiocarbohydrazide in the staining process. The samples were imaged using a Zeiss Crossbeam 540 FIB-SEM microscope (Carl Zeiss Microscopy GmbH, Jena, Germany). Platinum and carbon pads were deposited over the region of interest (ROI), and the run was set up and controlled by Atlas software (Fibics Incorporated, Ottawa, ON, Canada). ROI were selected from Lox +/+ (n = 1) and Lox +/C285F aortas (n = 2). SEM settings: 1.5 kV; 1.5 nA; milling probe: 700 pA. The slice thickness and the pixel size were set to 9 nm.

Advanced Imaging and Analysis
The FIB-SEM images were aligned using Atlas software. The data were then imported into Fiji software (Image J) [48] and binned 3X, to 27 × 27 × 27 nm isotropic voxels. The contrast was then normalized using the Enhance Local Contrast (CLAHE3D) [49] plugin in ImageJ [50]. Images and videos were rendered using Imaris v 9.3.1 (Bitplane Inc., Zurich, Switzerland).

Biochemical Analysis on Elastin and Collagen Contents
Aortas from the aortic root to the innominate were dissected from WT and mutant mice and stored at −80 • C prior to processing. Briefly, specimens (about 20 mg wet weight) were thawed and digested with high purity bacterial collagenase (Sigma, C0773; 100 U/mL, 37 • C, 18 h). After centrifugation, the soluble fractions containing collagen were hydrolyzed in 6 N HCl at 110 • C for 24 h and subjected to amino acid analysis (AAA) on a Biochrom 30 amino acid analyzer according to standard machine specific protocols (Biochrom, UK) using specified amino acid standards (Onken, Germany). The respective protein content of the different fractions was calculated as the sum of molecular weights of each amino acid (aa) in the AAA corrected for the molecular weight of water released during peptide bond formation. Collagen content was calculated based on a content of 14 mg hydroxyproline in 100 mg collagen. The insoluble fraction after collagenase digestion was extracted by hot alkali (0.1 N NaOH, 95 • C, 45 min). After centrifugation, an aliquot of the supernatant containing noncollagenous, nonelastin proteins and an aliquot of the insoluble residue containing insoluble elastin were subjected to hydrolysis and AAA as outlined above. For analysis, both collagen and elastin were normed to total protein, which was calculated as the sum of elastin, collagen, and noncollagenous, nonelastin proteins in the aorta.

RNA Preparation for QPCR and RNA Seq
Lox +/+ or Lox +/C285F mouse aortas were isolated by dissection from 2-week-old males and flash frozen in liquid nitrogen and stored at −80 • C. RNA was isolated using a Qiagen RNA Extraction kit (RNeasy, Qiagen 74104, Hilden, Germany). Briefly, frozen tissues were thawed in a supplied lysis buffer also containing β-mercaptoethanol. Tissues were disrupted in a bead homogenizer (Bead Ruptor 4, Omni International, 25-010, with 2 mL bead tubes Omni International, 19-628) and then processed according to the manufacturer's protocol. RNAs were validated for quantity by nanodrop and for integrity by bioanalyzer. For quantitative PCR (qPCR) analysis, cDNAs were generated from RNA by reverse transcription according to the manufacturer's protocol (High-Capacity cDNA Reverse Transcription kit ThermoFisher Scientific, 4368814). Three Lox +/+ and three Lox +/C285F cDNA samples were assayed by qPCR for Lox, Loxl1, Loxl2, Loxl3, and Loxl4 (see Supplemental  Table S6 for primer sequences). qPCR assays were performed on a QuantStudio3 real-time PCR system (ThermoFisher Scientific) and analyzed by ∆∆Ct with Hprt and 18S RNA as endogenous controls. For RNA seq, sequencing libraries were constructed from 300 ng of total RNA using the TruSeq Stranded Total RNA kit with Ribo-Zero Globin (Illumina 20020612, San Diego, CA, USA) following the manufacturer's instructions. The fragment size of RNAseq libraries was verified using the Agilent 2100 Bioanalyzer (Agilent G2939BA, Santa Clara, CA, USA), and the concentrations were determined using Qubit fluorometer (Life Technologies Q33216, Carlsbad, CA, USA). The libraries were loaded onto Illumina HiSeq 3000 for 2 × 75 bp paired-end read sequencing. Fastq files were generated using the bcl2fastq software (Illumina) for further analysis. Sequence reads were aligned to mouse reference genome M16 by STAR [51], indexed using samtools, and counts were performed using the featureCounts utility of the Subread package [52], and raw counts were normalized and analyzed for differential expression using DESeq2 [53]. Genes were selected as differentially expressed with |log 2 (fold change)| > 0.58 (1.5 fold change) and false discovery rate < 0.10. Gene Set Enrichment Analysis (GSEA) of that set for genes in canonical pathways was performed using the GSEA tool and molecular signatures database v7.1 [54,55]. A total of 859 differentially expressed genes are mapped as gene symbols to 600 genes in the GSEA database. Using GSEA, p values were computed using the hypergeometric distribution (Fisher's exact test) and adjusted with the Benjamini and Hochberg [56] method. Upstream regulator prediction from differentially expressed genes was performed using IPA (Qiagen Inc., https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis, accessed on 3 January 2020). Expression data were mapped into the IPA database using Ensembl IDs, matching 857 of 859 differentially expressed genes.

Lysyl Oxidase Enzyme Activity Assay
The aorta, from the root to diaphragm, was dissected out and snap frozen in liquid nitrogen. Lox enzyme activity was measured in tissues as described by Trackman and Bais [19] by measuring the production of hydrogen peroxidase through oxidation of Amplex Red (ThermoFisher Scientific, A12222, Waltham, MA, USA), which results in the generation of highly fluorescent resorufin for detection. Briefly, the frozen aortas were homogenized using the Bead Ruptor 4 Homogenizer (Omni International) in 250 µL of buffer containing 6M urea and 50 mM borate (pH 8.2). One hundred microliters of samples were then added to the reaction buffer with final concentration of 1.2M urea, 50 mM borate (pH 8.2), 1 unit/mL of horseradish peroxidase, 12.5 µM Amplex Red, and 12.5 mM 1,5-diaminopentane. Each sample was tested in duplicate. Parallel assays were prepared with 625 mM β-aminopropionitrile (BAPN) to inhibit the Lox activity. All reactions were incubated at 37 • C and measured for fluorescence at excitation wavelength of 563 nm and emission wavelength at 587 nm continuously for 60 min in CLARIOstar microplate reader (BMG Labtech). Reported activity is the average slope of the line generated from plotting the resorufin fluorescence by time done in n = 8.

Isolation and Protein Extraction from Mouse Aorta for Western Blot
Mice were sacrificed by CO 2 inhalation, and the vasculature was perfused with phosphate-buffered saline, via the left ventricle, to remove all blood. The aortas from E19, P15, and P90 were then carefully excised (free of fat) from the root down to the diaphragm. Vessels were snap frozen in liquid nitrogen and stored at −80 • C. For Western blot analysis, the frozen aortas were then homogenized using the Bead Ruptor 4 Homogenizer (Omni International) in 200 µL of RIPA buffer (Milliporesigma, R0278, Burlington, MA, USA) containing complete EDTA-free Protease Inhibitor Cocktail (MilliporeSigma, 11873580001, Burlington, MA, USA), and protein was quantified with Pierce Rapid Gold BCA Protein Assay kit (Thermo Scientific, A53226). A total of 30 µg of total protein was loaded on TGX stain-free protein gels (Bio-Rad, 17000927, Hercules, CA, USA) for blotting.

Isolation of Mouse Embryonic Fibroblasts (MEFs)
A pregnant Lox +/C285F mouse was sacrificed at 13 days postcoitum. The uterus was then dissected out and rinsed in 70% ethanol and placed in sterile PBS. Each embryo was then separated away from the uterus under sterile conditions and decapitated. The head and internal organs were removed, and the remaining carcass was washed in PBS and finely minced using a sterile razor until segments were small enough to pipette. The minced fragments were then digested with 0.05% trypsin/EDTA (ThermoFisher Scientific, 25300054, Waltham, MA, USA) and 100 Kunitz units of DNase I (ThermoFisher Scientific 18047019, Waltham, MA, USA) at 37 • C for 15 min. After that, trypsin was inactivated with MEF medium, containing 10% fetal bovine serum (FBS) (GE Healthcare, SH30071.03HI, Chicago, IL, USA), 1% of Penicillin-streptomycin (ThermoFisher Scientific, 15070063, Waltham, MA, USA) and 1% of nonessential amino acid (ThermoFisher Scientific, 11140050) in DMEM with 4.5 g/L glucose. The slurry was centrifuged at low speed (300× g) for 5 min. Supernatant was discarded, and the cell pellet was resuspended in MEF medium and seeded in one T75 flask coated with 0.1% gelatin (Stemcell technologies, #07903, Vancouver, BC, Canada). Cells reached 80-90% confluent after 48 h and were split in 1:4 ratio to obtain P1 MEFs. P1 MEFs were collected for DNA extraction to confirm the genotype and were frozen for future experiments.