Female C57BL/6N mice are a viable model of aortic aging in women

Abstract

The aorta stiffens with aging in both men and women, which predicts cardiovascular mortality. Aortic wall structural and extracellular matrix (ECM) remodeling, induced in part by chronic low-grade inflammation, contribute to aortic stiffening. Male mice are an established model of aortic aging. However, there is little information regarding whether female mice are an appropriate model of aortic aging in women, which we aimed to elucidate in the present study. We assessed two strains of mice and found that in C57BL/6N mice, in vivo aortic stiffness (pulse wave velocity, PWV) was higher with aging in both sexes, whereas in B6D2F1 mice, PWV was higher in old versus young male mice, but not in old versus young female mice. Because the age-related stiffening that occurs in men and women was reflected in male and female C57BL/6N mice, we examined the mechanisms of stiffening in this strain. In both sexes, aortic modulus of elasticity (pin myography) was lower in old mice, occurred in conjunction with and was related to higher plasma levels of the elastin-degrading enzyme matrix metalloproteinase-9 (MMP-9), and was accompanied by higher numbers of aortic elastin breaks and higher abundance of adventitial collagen-1. Plasma levels of the inflammatory cytokines interferon-γ, interleukin 6, and monocyte chemoattractant protein-1 were higher in both sexes of old mice. In conclusion, female C57BL/6N mice exhibit aortic stiffening, reduced modulus of elasticity and structural/ECM remodeling, and associated increases in MMP-9 and systemic inflammation with aging, and thus are an appropriate model of aortic aging in women.

NEW & NOTEWORTHY Our study demonstrates that with aging, female C57BL/6N mice exhibit higher in vivo aortic stiffness, reduced modulus of elasticity, aortic wall structural and extracellular matrix remodeling, and elevations in systemic inflammation. These changes are largely reflective of those that occur with aging in women. Thus, female C57BL/6N mice are a viable model of human aortic aging and the utility of these animals should be considered in future biomedical investigations.

INTRODUCTION

The aorta is the primary blood vessel responsible for delivering blood from the heart to the systemic circulation (1). The aorta is a large artery that is, under normal physiological conditions, highly elastic and dampens the pulsatile blood flow ejected from the heart (1, 2). This dampening effect is protective of the microvasculature and target organs, particularly in organs that receive large volumes of blood flow and have delicate, low-resistance vascular beds, such as the brain and kidneys (1). With advancing age, increases in aortic stiffness, as assessed in vivo by the reference standard measure carotid-femoral pulse wave velocity (c-fPWV), occur in both men and women (3–6). This increase in stiffness directly contributes to the development of cardiovascular diseases (CVD) (3), and increases in c-fPWV can predict (4–7) future incidence of CVD. Furthermore, c-fPWV is an independent predictor of numerous age-related conditions in men and women, including cognitive decline (8, 9), chronic kidney disease (10, 11), and impaired glucose tolerance and type 2 diabetes mellitus (11–13).

Structural changes to the arterial wall largely contribute to the increase in aortic stiffness in vivo. Fragmentation and degradation of the arterial structural protein elastin and consequently decreased elastic energy lead to aortic dilation/enlargement and a compensatory increase in the deposition of the structural protein collagen that provides rigidity to the arterial wall (1–3, 14). Other age-related changes in the arterial structure include an increase in the thickness of the intimal and medial layers (1–3). These adverse, yet adaptive, arterial structural changes are in part due to an increase in inflammation that occurs with aging (1, 15).

As the demographics in the United States and other developed nations shift toward a more aged population (16, 17), the prevalence of CVD is projected to increase by 30–35% (18). Despite extensive research in this field, there is still a need for effective interventions that prevent or treat age-related aortic stiffening and aortic wall structural and extracellular matrix remodeling. Use of appropriate animal models in biomedical research is essential for various reasons including 1) screening interventions for safety and efficacy (19, 20) and 2) providing mechanistic insight that is difficult, or impossible, to obtain in humans (19–22).

A key feature necessary for the use of animal models in biomedical research is that the model must completely, or mostly, reflect the human condition (19). Despite data demonstrating that age-related aortic stiffening and CVD occur in both men and women (e.g., 2, 21, 22), the majority of research to date examining mechanisms of and treatments for age-related aortic stiffening have been conducted in male rodent models (23–29). Some recent studies have begun to examine mechanisms of arterial function and aging in female mice (30–33), but there is still a paucity of information regarding the utility of female mice as an appropriate model of age-related aortic stiffening and structural remodeling and their relevance to aortic aging in women, especially late in life.

The goals of this study were to first determine whether female mice are an appropriate model (i.e., reflective of the human condition) of age-related aortic stiffening in vivo (aortic PWV, the rodent corollary to c-fPWV in humans) using two mouse models that are commonly studied in the context of conditions and diseases of aging, the B6D2F1 and C57BL/6N strains (25, 34, 35). In the strain(s) that were reflective of the human condition, we next sought to determine whether in vivo aortic stiffening with aging is accompanied by and related to aortic wall structural and extracellular matrix remodeling and systemic inflammation. Finally, we aimed to determine whether there were any sex differences in these outcomes between young and old mice.

MATERIALS AND METHODS

Ethical Approval

All protocols and procedures for this study were submitted to and approved by the University of Colorado Boulder Institutional Animal Care and Use Committee (Protocol No. 08-03-LES-01 and 2618) and adhered to the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals (36).

Study Design and Experimental Animals

We conducted a cross-sectional study in young and old, male and female, B6D2F1 and C57BL/6N mice to determine whether female mice were a viable model of age-related changes in aortic stiffness, aortic wall structural and extracellular matrix remodeling, and systemic (plasma) inflammation that occur in humans. Mice were studied at the following ages based on the age of physical maturity in young mice and the age of median lifespan for each sex in old mice (34). Young (studied at 6 mo of age) male and female B6D2F1 mice (n = 6 or 7/sex) were obtained from Charles River (Wilmington, MA). Old male (studied at 31 mo of age; n = 9) and old female (studied at 28 mo of age; n = 9) B6D2F1 mice were obtained from the National Institute of Aging colony (maintained by Charles River). Young male and female (studied at 6–7 mo of age; n = 10/sex) C57BL/6N mice were obtained from Charles River, and old male (studied at 27 mo of age, n = 20) and female (studied at 29 mo of age, n = 20) C57BL/6N mice were obtained from the National Institute of Aging colony. C57BL/6N mice (7 old males and 7 old females) died before euthanasia because of natural aging processes, resulting in final group sizes of n = 13/sex. All female mice were virgin, and we did not control for estrous cycle to increase generalizability of our findings. Mice were given ≥4 wk to adjust to our facilities at the University of Colorado Boulder before any testing. Mice were group-housed with a 12-h:12-h light/dark cycle and given ad libitum access to pelleted and irradiated NIH-31 Open formula mouse/rat sterilizable diet (stored at room temperature, Teklad 7917; Envigo, Indianapolis, IN) and drinking water.

Experimental Procedures

Investigators were blinded to the age and sex of the mice for biochemical assessments and data analyses.

In vivo aortic stiffness.

Aortic stiffness was assessed in vivo as aortic PWV via Doppler ultrasonography (Doppler Signal Processing Workstation, Indus Instruments, Webster, TX) as previously described by our laboratory (26, 27, 37–39). Mice were placed supine on a heating pad under light inhaled isoflurane anesthesia (1.5–2.0%) in O₂. Front and hind limbs were secured to electrodes to monitor heart rate, which was maintained between ∼400 and 500 beats/min. Doppler probes were placed on the transverse aortic arch and abdominal aorta. Three consecutive 2-s ultrasound tracings were recorded, and the time between the ECG R wave to the beginning of the Doppler signal (i.e., average pre-ejection time) was determined for each location. PWV (in cm/s) was calculated by dividing the distance between the two probes by the difference in the pre-ejection times of the aortic arch and abdominal aorta (time_abdominal – time_arch).

Euthanasia and tissue collection.

Mice were euthanized using a method approved under the American Veterinary Medical Association guidelines (40). Mice were anesthetized under inhaled anesthesia (open-drop method) and euthanized via cardiac exsanguination. Whole blood was heparinized and centrifuged at 10,000 rpm (9,391 rcf) for 10 min at room temperature to obtain plasma, which was then stored at −80°C for analyses described in Plasma MMP-9 and Plasma proinflammatory cytokines/chemokines. The heart was removed, cleaned, and weighed. The aorta was excised and rinsed in physiological saline solution (PSS), cleared of perivascular adipose tissue, and sectioned and stored as described in Ex vivo aortic modulus of elasticity—elastin-dominant region, Aortic diameter and IMT, and Elastin breaks, media cell count, and adventitila collagen-1 abundance.

Ex vivo aortic modulus of elasticity—elastin-dominant region.

Two 1 mm segments of thoracic aorta were collected, stored in PSS, and frozen at −80°C. Aortic modulus of elasticity (resistance to deformation within the elastic region of the elastic modulus) in the elastin-dominant region was assessed via pin (force) myography (DMT, Arhaus, Denmark) as previously described (37, 41, 42). Briefly, thawed aorta rings were mounted onto two pins in a phosphate-buffered saline (PBS) bath heated to 37°C. After three rounds of prestretching (pins displaced to 1 mm), the aortic diameter was increased until a force of 1 mN was reached and then was incrementally increased by 50 μm every 3 min until the vessel reached mechanical failure. The force corresponding to each stretching interval was recorded and used to calculate stress and strain and to construct a stress-strain curve:

$$\text{Strain }\left(\lambda \right)=\Delta d/d_{\text{i}}$$

where d is diameter and d_i is the initial diameter, and

$$\text{Stress}\left(t\right)=\left(\lambda \text{L}\right)/\text{2}\left(HD\right)$$

where L is one-dimensional load, H is intima-media thickness, and D is vessel length.

The borders of the low-force, elastin-dominant region of the stress-strain curve were established by fitting a seventh polynomial equation to the data (r² > 0.99; RStudio, Boston, MA) and then computing the roots of the equation. The first root was considered the boundary between the very low-force region and the point at which elastin fibers are initially engaged, and the second root was considered the upper boundary of the elastin-dominant region (43, 44). The modulus of elasticity was defined as the slope of a linear equation fitted to the stress-strain data between the first and second roots where the curvature is ∼0.

Plasma MMP-9.

Abundance of plasma matrix metalloproteinase-9 (MMP-9) in plasma was assessed using a quantikine ELISA kit per the manufacturer’s instructions (R&D Systems, Minneapolis, MN; Cat. No. MMPT90).

Aortic diameter and IMT.

The aortic diameter and intima-media thickness (IMT) were assessed in 1-mm aortic rings that were frozen in optimal cutting temperature (OCT) compound (Tissue-Tek) in liquid nitrogen-cooled methylbutane and stored at −80°C until the time of cryostat sectioning (Leica CM1520, Leica biosystems, Weltzar, Germany). Sections (7 μm) were plated on microscope slides and imaged under a bright-field microscope (Nikon Eclipse TS100; 4× magnification), and ImageJ software (National Institutes of Health, Bethesda, MA; RRID:SCR_003070) was used to quantify the aortic diameter and IMT (45). The differentiation between the medial and adventitial layers was determined as the point between the regular banding patterns of the external elastic lamellae and the diffuse pattern of the adventitia.

Elastin breaks, media cell count, and adventitial collagen-1 abundance.

Sections (∼1 mm) of thoracic aorta were excised and frozen in OCT compound, as described in the Aortic diameter and IMT. Sections (7 μm; Leica CM1520) were plated on poly-l-lysine-coated microscope slides, fixed in 2% paraformaldehyde for 10 min, washed with PBS, and permeabilized for 15 min (0.1% Triton X-100). Slides were blocked with 5% donkey serum for 1 h in a humidified chamber, incubated with an anti-collagen-1 primary antibody, (1:200; Cat. No. 1310-01; RRID: AB_2753206; Southern Biotech, Birmingham, AL) for 1 h, washed with PBS, and incubated with a species-specific fluorescent secondary antibody (1:200; AlexaFluor 647; Invitrogen, Waltham, MA) for 30 min. Slides were washed, stained with DAPI (1:1,000; Cat. No. 62248; ThermoFisher, Waltham, MA) for 5 min, and cured overnight with Invitrogen ProLong Gold mounting media (Cat. No. P36934; Invitrogen). The slides were then imaged using an EVOS m7000 (Cat. No. AMF7000; ThermoFisher) fluorescence microscope at 20× magnification under identical conditions and analyzed using Invitrogen Celleste 5.0 Image Analysis Software (RRID:SCR_022692). Images for assessing elastin breaks were obtained between 470 and 525 nm, where the elastic lamellae autofluorescence, as previously described (46, 47). Elastin breaks were determined as points in which there were large discontinuous regions in the elastic lamellae. These breaks were confirmed across two to three images per animal to reduce the likelihood of determining an elastin break when the lamellae were merely out of the focal plane of the microscope. The number of elastin breaks was averaged across n = 2 or 3 samples/animal. Media cell count was obtained by automatically counting the number of cells (identified with DAPI) within the borders of the internal and external elastic lamellae. The adventitial abundance of collagen-1 was determined as the average intensity (in arbitrary units, AU) of the collagen-positive area across n = 4–6 samples/animal. Specificity of the collagen-1 antibody was determined using negative controls (secondary antibody-only conditions) in a subset of the samples.

Plasma proinflammatory cytokines/chemokines.

Levels of interleukin (IL)-6, interferon (IFN)-γ, monocyte chemoattractant protein (MCP)-1, tumor necrosis factor (TNF)-α, and IL-1β were determined in plasma samples using a custom mouse G-Series semiquantitative cytokine antibody array (Cat. No. AAM-CUST-GS; RayBiotech, Peachtree Corners, GA). Plasma was diluted 2× and loaded into each microplate well. The assay was performed per the manufacturer’s instructions. The array slides were shipped to the manufacturer for scanning and data extraction, and data were sorted using the Excel-based GSM5 S4 Software, which was provided by the manufacturer.

Arterial blood pressure.

Systolic blood pressure (SBP) and diastolic BP (DBP) were measured using the CODA noninvasive tail-cuff system (Kent Scientific, Torrington, CT) as we have previously described (27, 38, 39). The average pressure readings were calculated from 20 collection cycles (after 5 acclimatization cycles) obtained each day for 3 consecutive days.

Statistics

Power calculations were performed using G*power 3.1 (RRID: SCR_013726) for our primary outcome variable, aortic PWV. Previously, our laboratory obtained effect sizes of 1.35 when comparing aortic PWV between young and old male mice. With this effect size, n = 6 mice per condition were required to achieve 99% statistical power. Additional mice were studied in each group to ensure sufficient PWV traces were obtained and to account for age-related attrition.

Statistical analyses were conducted using GraphPad Prism version 9.4.0 (GraphPad Software, San Diego, CA; RRID:SCR_002798). Data were assessed for statistical outliers (ROUT test; Q = 1%), and outliers were excluded from final analyses. All variables were assessed using two-way (age × sex) analysis of variance (ANOVA) with Šidák’s post hoc test when a main effect was observed. Relations between specific variables were assessed using simple linear regression. Data incorporated into these regressions were first assessed for normal distribution (Shapiro–Wilk normality test) and were log-transformed to account for skewness when P < 0.05. Statistical significance was set to α = 0.05. Data are presented as means ± SD.

RESULTS

In Vivo Aortic Stiffness

In B6D2F1 mice, we observed that aortic PWV was higher in males (P = 0.016 vs. young) but not in females (P > 0.999 vs. young) with aging (Fig. 1A), whereas both male and female C57BL/6N mice exhibited elevations in aortic PWV with aging (both P < 0.001 vs. young within sex; Fig. 1B). There were sex differences in aortic PWV between old B6D2F1 mice (P = 0.027), but no sex differences between young B6D2F1 (P = 0.866) or young (P = 0.992) and old (P = 0.629) C57BL/6N mice. Given that the age-related changes that occur in men and women were accurately reflected in male and female C57BL/6N mice, we further examined mechanisms of aortic stiffening in this strain.

Figure 1.

In vivo aortic stiffness is higher with aging in both male and female C57BL/6N mice, but only in male B6D2F1 mice. Aortic pulse wave velocity (PWV) in male and female, young (6–7 mo) and old (28–31 mo) B6D2F1 (A) and young (6–7 mo) and old (27–29 mo) C57BL/6N (B) mice. Data are represented as means ± SD; n = 6–15/group. Statistics are two-way mixed (age × sex) ANOVA with Šidák’s post hoc test. *P < 0.05; ***P < 0.001; ****P < 0.0001.

Aortic Wall Extracellular Matrix and Structural Remodeling

In both male and female C57BL/6N mice, we observed lower aortic modulus of elasticity with aging (Fig. 2A; both P < 0.001 vs. young within sex), which was inversely related to in vivo aortic stiffness (Fig. 2B; r² = 0.27; P = 0.006) and accompanied by higher numbers of aortic elastin breaks (Fig. 2, C and D; both P < 0.03 vs. young within sex). In addition, plasma levels of a key elastin degrading enzyme, matrix metalloproteinase-9 (MMP-9) (48, 49), were elevated with aging (Fig. 2E; both P < 0.01 vs. young within sex), and MMP-9 levels were positively related to aortic stiffness (Fig. 2F; r² = 0.20; P = 0.008) and inversely related to aortic modulus of elasticity (Fig. 2G; r² = 0.29; P = 0.005). There were no sex differences in any of these outcomes (all P > 0.288).

Figure 2.

Aortic modulus of elasticity is lower and elastin breaks and matrix metalloproteinase-9 (MMP-9) are higher with aging in both male and female mice. In male and female, young (6–7 mo) and old (27–29 mo) C57BL/6N mice, aortic modulus of elasticity (elastin dominant, low-force region of the stress-strain curve) (A), aortic modulus of elasticity is inversely related to aortic stiffness (aortic pulse wave velocity, PWV) (B), number of aortic elastin breaks (C) with representative images below (D) [within each aortic ring, there are zoomed in visuals (white circles) and arrows denoting the elastin breaks], and plasma levels of MMP-9 (E), and plasma MMP-9 levels are positively related to aortic stiffness (PWV; F) and inversely related to aortic modulus of elasticity (G). Data are represented as means ± SD; n = 4–9/group unless otherwise noted. Statistics are two-way mixed (age × sex) ANOVA with Šidák’s post hoc test and simple linear regression. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Aortic diameter tended to be higher in young males than young females (P = 0.060) and was higher in old versus young female mice (P = 0.016) but was not different between old versus young males (P = 0.444; Fig. 3, A and C). Aortic IMT was higher in both old versus young male and female mice (Fig. 3, B and C; both P < 0.001), likely due to medial cell hypertrophy (vs. hyperplasia), as there were no differences in medial cell count between male or female old versus young mice (Fig. 3D; both P > 0.352 vs. young), similar to previous reports (23). These structural changes were accompanied by a higher adventitial abundance of collagen-1 [primary isoform of collagen in arteries (50)] in both old male (Fig. 3, E and F; P = 0.049) and female mice (P = 0.081), with a similar average magnitude of age-related difference in both sexes. There were no sex differences in aortic IMT, medial cell count, nor adventitial abundance of collagen-1 (all P > 0.190).

Figure 3.

Aortic wall structural and extracellular matrix remodeling in male and female mice with aging. In male and female, young (6–7 mo) and old (27–29 mo) C57BL/6N mice, aortic diameter (A) and aortic intima-media thickness (IMT; B), measured in 7 μm segments of thoracic aorta; representative images of aortic diameter and IMT (C); media cell count using DAPI staining (D); and adventitial abundance of collagen-1 (E), measured via immunofluorescence in 7 μm segments of thoracic aorta; representative images of DAPI (blue) and collagen-1 (red) (F). Data are represented as means ± SD; n = 5–13/group. Statistics are two-way mixed (age × sex) ANOVA with Šidák’s post hoc test. *P < 0.05; ***P < 0.001; ****P < 0.0001.

Circulating Markers of Inflammation

Circulating (plasma) levels of the proinflammatory cytokines/chemokines interferon-γ (IFN-γ), interleukin 6 (IL-6), and monocyte chemoattractant protein-1 (MCP-1) were elevated with aging in both male and female mice, whereas there were no effects of age on levels of interleukin-1β (IL-1β) or tumor necrosis factor-α (TNF-α) (Table 1). There was a significant interaction between age and sex for IL-1β, but there was no significant main effect of either age or sex (Table 1). There were no differences across sex for any of the other circulating inflammatory markers (Table 1).

Table 1.

Plasma cytokine/chemokine levels

	Male		Female		P Value
Cytokine	Young	Old	Young	Old	Age	Sex	Interaction
IFN-γ	1,006 ± 607	2,128 ± 1,301	1,626 ± 1,033	2,432 ± 2,276	0.0373	0.3081	0.7264
IL-1β	354 ± 43	463 ± 114	516 ± 194	415 ± 135	0.9166	0.1953	0.0203
IL-6	2,970 ± 724	6,320 ± 3,132	2,794 ± 572	8,261 ± 4,197	<0.0001	0.3162	0.2309
MCP-1	727 ± 134	1,088 ± 526	891 ± 131	1,427 ± 855	0.0121	0.1489	0.6118
TNF-α	583 ± 171	603 ± 165	648 ± 114	610 ± 176	0.4700	0.8624	0.5581

Values are means ± SD (n = 9–13/group) and are expressed as arbitrary units of intensity relative to an internal positive control. Plasma was analyzed by cytokine array. IFN-γ, interferon-γ; IL-1β, interleukin-1β; IL-6, interleukin 6; MCP-1, monocyte chemoattractant protein-1; TNF-α, tumor necrosis factor-α. Statistics are two-way mixed (age × sex) ANOVA with Sidák’s post hoc test. Boldface indicates significant value.

Relations were assessed between the aortic modulus of elasticity or in vivo stiffness and the circulating (plasma) markers of inflammation that were higher with aging (Table 1). Inflammatory markers were log-transformed to account for skewness (Shapiro–Wilk normality test: all P < 0.05). Aortic modulus of elasticity was inversely related to IL-6 (Fig. 4B; r² = 0.491; P < 0.001) and MCP-1 (Fig. 4C; r² = 0.122; P = 0.058). Aortic stiffness was positively associated with IL-6 (Fig. 4E; r² = 0.120; P = 0.018). There were no relations between aortic modulus of elasticity and IFN-γ (Fig. 4A; r² = 0.076; P = 0.140) or between aortic stiffness and IFN-γ (Fig. 4D; r² = 0.016; P = 0.402) or MCP-1 (Fig. 4F; r² = 0.000; P = 0.881).

Figure 4.

Relations between aortic modulus of elasticity and in vivo stiffness [pulse wave velocity (PWV)] and select inflammatory markers that were higher with aging. In male and female, young (6–7 mo) and old (27–29 mo) C57BL/6N mice, relations between aortic modulus of elasticity and plasma levels of interferon-γ (IFN-γ; A), interleukin-6 (IL-6; B), and monocyte chemoattractant protein-1 (MCP-1; C) and between aortic stiffness and plasma levels of IFN-γ (D), IL-6 (E), and MCP-1 (F); n = 10–14/group. Plasma was analyzed by cytokine array and reported as arbitrary units (AU) of intensity. All cytokines/chemokines were log-transformed to account for skewness (Shapiro–Wilk normality test; all P < 0.05). Statistics are simple linear regression.

Blood Pressure

DBP was lower in old versus young male mice (Fig. 5B; P = 0.002) and tended to be lower in old versus young female mice (Fig. 5B; P = 0.096). SBP was lower in old versus young male and female mice (Fig. 5A; P < 0.001), and there was no effect of age on PP (Fig. 4C; P = 0.554) in either male or female mice. There were no sex differences in these outcomes (all P > 0.347).

Figure 5.

Systolic blood pressure (SBP) and diastolic BP (DBP) are lower in old vs. young male and female mice, and pulse pressure (PP) is not different. Tail-cuff BP in young (6–7 mo) and old (27–29 mo), male and female C57BL/6N mice: SBP (A); DBP (B); and PP (C). Data are represented as means ± SD; n = 6–10/group. Statistics are two-way mixed (age × sex) ANOVA with Šidák’s post hoc test. *P < 0.05; **P < 0.01.

Heart and Left Ventricular Masses

These age-related aortic functional and structural changes were accompanied by elevations in the mass of the heart (Fig. 6A) and left ventricle (LV) (Fig. 6B) (normalized to body mass) in old versus young male and female mice (all P < 0.015). There were no sex differences in heart or LV mass (both P > 0.106).

Figure 6.

Heart and left ventricular (LV) masses are higher in old vs. young male and female mice. Ratio of heart (A) and left ventricular (B) masses to body mass in young (6–7 mo) and old (27–29 mo), male and female C57BL/6N mice is shown. Data are represented as means ± SD; n = 9–14/group. Statistics are two-way mixed (age × sex) ANOVA with Šidák’s post hoc test. *P < 0.05; ***P < 0.001; ****P < 0.0001.

DISCUSSION

Our results demonstrate that in vivo aortic stiffness, as assessed by aortic PWV, is higher in old versus young both male and female C57BL/6N mice, and in male but not female B6D2F1 mice. As such, female C57BL/6N mice appear to be an appropriate model of in vivo aortic stiffening with aging in women. Subsequently, in C57BL/6N mice, we explored whether mechanisms that contribute to aortic stiffening with aging were apparent in these mice, and observed reduced aortic modulus of elasticity, aortic wall structural and extracellular matrix remodeling, and increases in systemic inflammation in both sexes.

Female Mice in Vascular Aging Research

Aspects of CVD risk, trajectories, presentation, and responsiveness to/effectiveness of various cardiovascular-targeted interventions differ between men and women. Accordingly, organizations such as the National Institutes of Health, the American Heart Association, and the American Physiological Society have emphasized the need to consider sex as a biological variable to improve the quality and generalizability of biomedical research (51–56). Male mice have been established as a model of aortic aging, and the majority of vascular aging research (e.g., 23–27, 57, 58), and cardiovascular research in general (54, 55, 59), to date has been conducted solely using male mouse models. Recently, investigators have begun to examine sex differences in 1) mechanisms of vascular function using transgenic mouse models and 2) impairments in vascular function in middle to early late life (30–33). These previous studies have demonstrated that there are sex differences in normal arterial function and the time course of arterial aging in midlife, which can be attributed to sexually dimorphic arterial abundance of elastin and mineralocorticoid and estrogen receptors. We sought to contribute to this emerging body of literature by assessing changes in aortic function and structure in female mice in late life, their relevance to human aortic aging, and thus viability as a model for screening interventions for improving vascular aging in late life, which to our knowledge have not been thoroughly characterized. This study provides the necessary validation of female C57BL/6N mice as a model of aortic aging in women, which will aid the vascular aging research community in addressing important biomedical research questions that pertain to sex as a biological variable. Furthermore, our study demonstrates that female mice of different strains exhibit variations in aortic aging, suggesting strain-specific utility of mice as models for aortic aging in humans.

Trajectories of Aortic Aging in Humans and Mouse Models

In humans, c-fPWV is higher in older compared with young men and women, with a similar magnitude of increase with aging in both sexes (3, 60), although increases in c-fPWV may be delayed temporarily during midlife in women (57). We observed that the in vivo aortic stiffening that occurs with aging in men and women was reflected in male and female C57BL/6N mice, and in male but not in female B6D2F1 mice, suggesting that C57BL/6N mice are a more appropriate model of age-related aortic stiffening in humans. As such, we sought to determine if the mechanistic phenotypes, including aortic wall structural changes and systemic inflammation, that accompany and likely contribute to in vivo aortic stiffening were similar between old male and female C57BL/6N mice, and similar to those that occur in humans.

In humans, aortic compliance, which reflects the elastic properties of the artery, can be estimated in vivo (61). These data show reductions in aortic compliance with aging in both men and women, which occurs to a greater extent in men (2, 62). Mechanical and tensile properties of human arteries can be assessed ex vivo using numerous techniques (63) and collectively demonstrate altered mechanical properties (64–66) and reduced aortic elastic energy with aging (67). These studies did not report data separately by sex. We assessed the elastin dominant component of the aortic elastic modulus by determining the slope of the linear portion of the low-force region of the stress-strain curve (i.e., modulus of elasticity). Consistent with data in humans and previous findings in mice (37, 41, 44), modulus of elasticity in this region was lower with aging in the present study, and we did not observe any sex differences.

In line with the findings discussed earlier, we observed an inverse relation between aortic stiffness (PWV) and the low-force, elastin-dominant modulus of elasticity. The Moens–Korteweg equation predicts a positive relation between PWV and elastic modulus. However, elastic modulus in this equation is typically obtained from the high-force region of the stress-strain curve and may encompass artery stiffness that would not typically be observed within the normal physiological pressure range. Thus, we believe assessing the relation between PWV and the low-force region of the elastic modulus curve is relevant and does not contradict this equation and previous findings, but rather is an alternate way to compare these other relevant aspects of arterial mechanics and physiology.

Increased aortic stiffness and reduced modulus of elasticity are largely due to aortic wall structural remodeling (1, 3, 68, 69). Data in humans collected during biopsies and in postmortem tissues demonstrate that aortic elastin fragmentation/disorganization and collagen content increase with age (64, 70–73), consistent with what we observed in the present study. These data in humans are not reported separately by sex, and we did not observe any sex differences in these outcomes.

MMP-9 is a key enzyme that degrades elastin and is indicated in aortic stiffening (1, 74–78). Circulating levels of MMP-9 are higher with aging in male mice (79), whereas in humans, conflicting data have been reported concerning changes with aging and differences across sex (80–82). We observed an increase in circulating MMP-9 with aging in both male and female mice, which was positively related to aortic stiffness and inversely related to aortic modulus of elasticity, suggesting a potential role of MMP-9-mediated elastin degradation in increased aortic stiffness and reduced modulus of elasticity in the C57BL/6N mouse model.

In humans, aortic root diameter and IMT are higher with advanced age in both men and women, but aortic diameter is larger in men at any given age (2, 3, 83). Aortic wall thickness has been reported to be higher in men with aging, although this has not always been observed (2, 83, 84). We similarly observed an age-related increase in aortic diameter and IMT in male and female mice but did not observe any sex differences.

Serum concentrations of inflammatory cytokines are elevated with aging and associated with and/or implicated in the development of aortic stiffening in humans and male C57BL/6N mice (1, 27, 85–87). As such, we measured an array of circulating (plasma) inflammatory cytokines and observed elevated levels of IFN-γ, IL-6, and MCP-1 with aging in mice, reflecting the increase in systemic inflammation observed in humans with aging. Furthermore, aortic stiffness was positively related to IL-6, and aortic modulus of elasticity was inversely related to IL-6 and MCP-1, suggesting a potential role for systemic inflammation in mediating increases in stiffness and decreases in modulus of elasticity in the aging aorta of these mice, but future studies should investigate whether this relation is causal in both sexes.

Blood Pressure and Cardiac Hypertrophy

In humans, noninvasive assessments of brachial artery BP show that SBP increases and DBP is unchanged or decreases with aging, resulting in a widening of PP (3), and this increase in SBP is linked to arterial stiffening (88). In contrast, when using a noninvasive corollary of this measure (tail-cuff plethysmography) in C57BL/6N mice, we have not observed similar increases in SBP and PP with aging (27, 44), consistent with the present study. Direct, invasive intra-arterial assessments of BP are more precise and indicate age-related increases in SBP and PP in male C57BL/6N and transgenic mice (89, 90) and in male and female C57BL/6J mice (33). Therefore, we note that tail-cuff plethysmography may not accurately capture age-related changes in central blood pressure, which is a limitation of the study.

Increases in aortic stiffness and consequent increases in aortic impedance contribute to adverse cardiac remodeling, in particular left ventricular hypertrophy (91–93). We measured heart and left ventricular masses normalized to body mass and observed increases in both measures in male and female mice, consistent with what is observed with aging in men and women (94).

Experimental Considerations Regarding Sex Hormones

Sex hormones influence aspects of vascular function and aging. For example, PWV varies across the estrous cycle in young female C57BL/6N mice (95). We did not control for estrous cycle in the present study to increase the generalizability of our findings. However, it is possible that in our study, fluctuations in hormone levels in young female mice and consequent changes in PWV could have contributed to the lack of an age-related difference in PWV in female B6D2F1 mice.

In addition, our study examined mice during early adulthood and late life but did not include assessments in midlife. The estrous cycle of female mice becomes irregular and typically ceases in midlife (96), which may influence the timing of age-related changes in vascular function in a sexually dimorphic way. For example, female relative to male C57BL/6J mice have delayed aortic stiffening in midlife (12–18 mo of age) (31). However, intracarotid PWV is not different between male and female C57BL/6 mice at 13 mo of age (32), and aortic PWV is equivalent in male and female wild-type mice (unknown strain) at 24 mo of age (30). The age of female mice relative to follicular failure and potential sex differences in the timing of age-related changes in vascular function should be considered in future studies and when translating findings from mice to humans. These considerations are particularly relevant in the context of identifying mechanisms of or treatments for aortic/arterial aging in pre- versus postmenopausal women.

Conclusions

Our findings demonstrate that female C57BL/6N mice generally reflect the major aortic structural and functional changes that occur with aging in women, and therefore are an appropriate model for use in biomedical investigations. Future studies should consider the use of both male and female C57BL/6N mice in the context of identifying sex differences in mechanisms of aortic aging and identifying or developing treatment strategies for preventing or improving aortic aging.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by National Institutes of Health Grants T32 AG000279 (to A.G.L.), F31 HL160173 (to A.G.L.), F32 HL151022 (to Z.S.C), and K99 HL159241 (to Z.S.C.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.G.L., M.C.Z., and Z.S.C. conceived and designed research; A.G.L., R.V., M.C.Z., A.J.L., S.A.M., N.T.G., N.S.V., and Z.S.C. performed experiments; A.G.L., A.J.L., K.R.L., and Z.S.C. analyzed data; A.G.L. and Z.S.C. interpreted results of experiments; A.G.L. prepared figures; A.G.L. drafted manuscript; A.G.L., R.V., M.C.Z., A.J.L., S.A.M., N.T.G., N.S.V., K.R.L., K.L.M., D.R.S., and Z.S.C.edited and revised manuscript; A.G.L., R.V., M.C.Z., A.J.L., S.A.M., N.T.G., N.S.V., K.R.L., K.L.M., D.R.S., and Z.S.C. approved final version of manuscript.