Relative Contributions of Age and Atherosclerosis to Vascular Stiffness

Abstract

To determine the relative contributions of aging and atherosclerosis to vascular stiffness, we studied aortic stiffness, plaque, and elastin in 8‐, 16‐, 25‐, and 34‐week‐old male ApoE‐KO and C57BL/6J control mice (N = 48). Stiffness increased gradually in both strains up to 25 weeks (p < 0.05), and dramatically between 25 and 34 weeks in ApoE‐KO (p < 0.001). Aging ApoE‐KO demonstrated increased plaque (p = 0.02), medial thickening (p < 0.001), and severe elastin fragmentation (p < 0.001). We conclude that the contribution of aging to vascular stiffness is relatively minor compared with the influence of atherosclerosis. However, the effect of atherosclerosis on stiffness is significant only with advanced stages of plaque formation.

Introduction

It is well established that remodeling of the arterial wall occurs with age and leads to increased vascular stiffness. Specifically, studies indicate that the elastic content of the tunica media decreases with age, resulting in reduced vessel compliance. ¹ Progressive vascular stiffening occurs in aging individuals free from traditional cardiovascular disease (CVD) risk factors, ² while also being associated with atherosclerosis. ³ , ⁴ Furthermore, there is evidence that the incidence and severity of atherosclerosis increases with age. ⁵ Some suggest that age and atherosclerosis contribute synergistically to increase vascular stiffness, ⁶ whereas others demonstrate that age is the predominant correlate of increased stiffness. ² , ⁷ It has also been suggested that stiffness is an independent predictor of cardiovascular events and morbidity. ⁸ Although the etiology of vascular stiffening is not well understood, numerous studies propose that the mechanism by which the vessel stiffens is different in aging versus atherosclerotic vessels. ⁹ , ¹⁰ , ¹¹ However, the relative contributions of aging and atherosclerosis to increased vascular stiffness have not been determined.

The C57BL/6J mouse, used extensively in animal studies of CVD, exhibits an increase in vascular stiffness with age that is comparable to that of a healthy human ¹² and develops plaques when fed a Western‐type, or atherogenic diet. ¹³ The apolipoprotein E‐deficient (ApoE‐KO) mouse, possessing a C57BL/6J background, is widely used as a model of atherosclerosis. ¹⁴ This mouse lacks the polypeptide responsible for clearance of circulating cholesterol and this deficiency leads to severe hyperlipidemia and increased aortic stiffness with age. ⁴ , ¹⁵ Plaques develop at an accelerated rate when these mice are fed an atherogenic diet. ¹⁶ Furthermore, this strain develops plaques similar in appearance and distribution to those observed in humans, demonstrates all known phases of atherogenesis (from monocyte adhesion to the development of fatty streaks and fibrous plaques), and has a predilection for development of plaque at branch points, outflow tracts, and bifurcations, with early lesions forming within the aortic sinus and arch. ¹⁷ , ¹⁸

Our objective was to study the relative contributions of age and atherosclerosis to aortic stiffening in both C57BL/6J and ApoE‐KO mice by assessing stiffness and evaluating vessel histology as a function of age. Specifically, we aimed to (1) measure aortic stiffness using a highly accurate ex vivo approach, (2) assess plaque size and composition, and (3) evaluate the elastic content of the aortic wall.

Methods

Animals

We conducted blinded experiments in 8‐, 16‐, 25‐, and 34‐week‐old male ApoE‐KO and age‐matched control C57BL/6J mice (n= 6 per age per strain, total N= 48) obtained from Jackson Laboratories (002052 and 000664; Bar Harbor, ME, USA). Beginning at 5 weeks old, we fed both strains an atherogenic diet ad libitum (45 kcal% fat, 1.25% cholesterol, and 0.5% cholic acid; D04081305A1, Research Diets, New Brunswick, NJ, USA). Mice were housed at the University of Pittsburgh under standard conditions and all experimental protocols were approved by the Institutional Animal Care and Use Committee.

Ex vivo measurement of aortic stiffness

After euthanizing mice via carbon dioxide inhalation, we harvested thoracic aortas using the descending aortic arch and celiac trunk as landmarks to define the beginning and end of the segment. Before excising, we measured the in vivo length of the segment. Immediately after, we removed the connective tissue surrounding the adventitial surface and stored the aorta in 4°C phosphate buffered saline (PBS; pH 7.3; BP665‐1, Fisher Scientific, Pittsburgh, PA, USA) until testing.

We used an ex vivo vascular testing system to measure stiffness of the aorta; we previously customized this system to accommodate the size of the mouse aorta (approximately 1‐mm outer diameter, and 18‐mm long). ¹⁹ As described in detail elsewhere, the system consists of a temperature‐controlled tissue bath, a pressurization loop, and a laser micrometer. ¹⁹ , ²⁰ , ²¹ The system functions by applying a dynamic intraluminal pressure waveform to an isolated vascular segment while precisely measuring the change in vessel outer diameter (accurate to 0.4 μm).

We mounted each aorta at in vivo length onto a set of proximal and distal microcannulae within the tissue bath. We then pressurized the segment intraluminally by driving PBS through the proximal microcannulae and estimated the average pressure within the vessel (P _avg) by averaging the pressures measured upstream and downstream (P ₁ and P ₂, respectively) of the vessel (i.e., P _avg= (P ₁+P ₂)/2). We positioned the laser micrometer beam at the middle of the segment length and measured vessel outer diameter (OD) as P _avg was cycled from 80 to 130 mmHg, the diastolic and systolic pressures of a healthy mouse. ²² Before collecting data, we preconditioned the vessel by repeating the pressure‐loading cycles 10 times. We then subjected the vessel to continuous pressurization cycles while recording the pressure–diameter response.

To derive a value for aortic stiffness, we smoothed the preconditioned pressure and diameter data using a moving average (n= 5), and analyzed 10 continuous cycles demonstrating a set of uniform pressurizations. Using the data from the loading portion of each cycle, we calculated Peterson's pressure–strain elastic modulus (E _p), an index of vascular stiffness, which is independent of vessel size, for each specimen (Equation (1)) ²³ :

(1)

where P _max and P _min represent the maximum and minimum P _avg values, respectively, and D _max and D _min represent the maximum and minimum OD values, respectively, of each of the 10 chosen cycles. We report E _p in mmHg and larger values indicate stiffer vessels.

Evaluation of plaque and elastin in aortic cross‐sections

Cryopreservation and sectioning.

After testing, we dismounted each aorta from the microcannulae, filled the vessel lumen with OCT freezing medium (Sakura Finetek, 4583, Torrance, CA, USA), embedded the vessel vertically in 5 mL of OCT, and snap froze it over liquid nitrogen. We cut serial, 10‐μm sections from the proximal aorta using a cryostat (Shandon Lipshaw, 620E, Pittsburgh, PA, USA) and mounted sections onto glass slides (VWR, 48311‐703, West Chester, PA, USA).

Staining and microscopy.

Before staining, we allowed sections to dry at room temperature. We stained for lipids using Oil Red O and elastin with Verhoeff's stain (Newcomer Supply, 1277B and 9116B, Middleton, WI, USA). To evaluate smooth muscle cells, we counterstained the Oil Red O slides with Harris hematoxylin (Newcomer Supply, 1201A). After staining, we coverslipped sections using an aqueous mounting solution (Newcomer Supply, 6271A) and stored at 4°C. Using a light microscope equipped with a digital camera (Olympus, Provis I, Center Valley, PA, USA) we viewed slides.

Morphometric analysis.

To evaluate plaque size, we imported the digital images of Oil Red O stained sections into MetaMorph (Molecular Devices Corporation, Version 7, Sunnyvale, CA, USA), an image analysis software program. We used the “trace region” function to measure lumen and plaque areas, and to normalize for differences in vessel or lumen size, we calculated plaque size as a percentage of the total cross‐sectional area (Plaque % Area; i.e., percent of the aorta cross‐section that is covered by plaque) (Equation (2)):

(2)

where A _lumen and A _plaque represent the areas of the lumen and plaque (bounded by the internal elastic lamina and lumen), respectively. In order to evaluate plaque composition, we examined the images for evidence of cholesterol crystals, acellular regions, and fibrous caps.

To evaluate elastin, we inspected the digital images of Verhoeff stained sections for torn fibers and loss of fiber organization. Using a scale described by Sukhova et al., we graded elastin preservation as follows ²⁴ : grade 1, intact and well‐organized fibers; grade 2, some interruptions and breaks in fibers; grade 3, severe fiber fragmentation or loss. We also imported elastin images into MetaMorph to determine media thickness (distance between the internal and external elastic laminae), calculated as the average of four measurements made at equidistant points along the cross‐section.

Statistical analysis

We evaluated all parameters for normal distribution and unequal variance using Shapiro–Wilk and F‐tests. To test statistical significance, we used single‐factor ANOVAs and unpaired Student's t‐tests. Data are expressed as means ± one standard deviation and p‐values of <0.05 are reported as significant. We used STATA (Version 9.1; StataCorp, College Station, PA, USA) and Excel (Version 10.1.7, Microsoft Corporation, Redmond, WA, USA) to generate all plots and statistical comparisons.

Results

Mechanical testing

In aging C57BL/6J mice (n= 6 per group), E _p increased progressively with age (Figure 1, Table 1) from 719.0 ± 135.6 mmHg at 8 weeks to 1009.7 ± 122.0 mmHg at 34 weeks; this increase was significant between the 8‐ and 25‐week time points (p= 0.02) but not between 25 and 34 weeks. Aging ApoE‐KO mice (n= 6 per group) also demonstrated increased stiffness with age (Figure 1, Table 1) with E _pvalues increasing from 768.6 ± 224.9 mmHg at 8 weeks to 2571.7 ± 599.5 mmHg at 34 weeks; this increase was significant between 8 and 25 weeks (p= 0.005), and between 25 and 34 weeks (p= 0.001).

Figure 1.

Mean stiffness (E _p, mmHg) and standard deviations for 8‐, 16‐, 25‐, and 34‐week‐old C57BL/6J and ApoE‐KO mice (n= 6 per time point/strain, total N= 48). For both strains, the data demonstrate an increase in stiffness with age. As denoted by the asterisks, stiffness increased significantly in C57BL/6J mice between 8 and 25 weeks (p= 0.02) and in ApoE‐KO mice between 8 and 25 weeks (p= 0.005) and 25 and 34 weeks (p= 0.001). Actual E_p values are given in Table 1.

Table 1.

Stiffness and histological data for ApoE‐KO and C57BL/6J strains at 8, 16, 25, and 34 weeks (n = 6 per age group per strain, total N = 48).

Strain	Age(weeks)	n	Ep Stiffness (mmHg)	Plaque present % (n/group)	Plaque area* (%)	Media thickness (μm)	Elastin grade(1–3)
ApoE‐KO	8	6	768.6 ± 224.9	0% (0/6)**	–	29.11 ± 1.71	1.17 ± 0.41
ApoE‐KO	16	6	1058.7 ± 387.5	33% (2/6)	5.79 ± 5.71	33.95 ± 1.60	1.67 ± 0.58
ApoE‐KO	25	6	1186.2 ± 166.4‡	83% (5/6)	38.42 ± 17.96	39.96 ± 4.54‡,#	2.33 ± 0.82‡
ApoE‐KO	34	6	2571.7 ± 599.5†,§,#	100% (6/6)	52.60 ± 28.58†	40.43 ± 2.99†,#	2.83 ± 0.41†,#
C57BL/6J	8	6	719.0 ± 135.6	0% (0/6)	–	28.41 ± 1.76	1.00 ± 0.00
C57BL/6J	16	6	853.3 ± 168.8	0% (0/6)	–	34.36 ± 1.42	1.17 ± 0.41
C57BL/6J	25	6	1051.1 ± 248.3‡	0% (0/6)	–	31.50 ± 1.70#	1.60 ± 0.55
C57BL/6J	34	6	1009.7 ± 122.0†,#	0% (0/6)**	–	30.36 ± 1.35#	2.00 ± 0.00†,#

*Average taken only of samples with plaque; ^†Significant difference between 8 and 34 weeks within strain; ^‡Significant difference between 8 and 25 weeks within strain; ^§Significant difference between 25 and 34 weeks within strain; ^#Significant difference between strains; **Fatty streak present.

When comparing E _p values between strains (Figure 2), our data demonstrated that stiffness was slightly greater in the ApoE‐KO strain at 8, 16, and 25 weeks (p= NS), but significantly greater at 34 weeks (p < 0.001).

Figure 2.

Mean stiffness (E_p, mmHg) and plaque percent area results for 8‐, 16‐, 25‐, and 34‐week‐old ApoE‐KO and C57BL/6J mice. As denoted by the asterisk, stiffness (lines) increased significantly between 25‐ and 34‐week‐old ApoE‐KO (p < 0.001). No plaques were present in C57BL/6J and plaque percent area (bars) increased progressively between 8 and 34 weeks in ApoE‐KO (p= 0.02). Actual E_p and plaque percent area values are given in Table 1.

It should, however, be noted that with age, vessel diameter did not significantly increase within each strain or differ between strains; for C57BL/6J and ApoE‐KO vessels, actual mean outer diameter values, respectively, were 0.81 mm and 0.80 mm at 8 weeks, 0.86 mm and 0.82 mm at 16 weeks, 0.89 mm and 0.86 mm at 25 weeks, and 0.90 mm and 0.93 mm at 34 weeks.

Histology

No plaques were found at any time point in C57BL/6J mice, though fatty streaks were visible in 33% of animals at 34 weeks. As illustrated in 2, 3, and detailed in 1, 2, plaques developed progressively with age in the ApoE‐KO mice. At 8 weeks, fatty streaks were visible in 50% of animals. By 16 weeks, small plaques had developed in 33% of animals with plaque size averaging 5.8% of the luminal cross‐sectional area. By 25 weeks, 83% of animals had plaques with an average size of 38.4% whose composition included cholesterol crystals and acellular regions. By 34 weeks, 100% of animals had developed large plaques with an average size of 52.6%, half had a fibrous cap, and all contained cholesterol crystals and acellular regions. In ApoE‐KO mice, plaque size did not increase significantly between 16 and 25 weeks, or between 25 and 34 weeks; however, the increase was significant when we compared the amount of plaque present at 8 weeks to that at 34 weeks (p= 0.02).

Figure 3.

Oil Red O and hematoxylin staining of aortic cross‐sections from 8‐, 16‐, 25‐, and 34‐week‐old ApoE‐KO mice shows progression from a fatty streak at 8 weeks (A) a small plaque at 16 weeks (B) a larger plaque at 25 weeks (C)with cholesterol crystals, acellular regions, and lipid droplets throughout, and a fully developed plaque at 34 weeks (D), complete with fibrous cap, acellular regions, cholesterol crystals, and lipid droplets. C57BL/6J mice did not develop plaques; however, some fatty streaks were visible at 34 weeks (not shown). Actual plaque size and composition data are given in 1, 2, respectively. Sections are shown at the same magnification (black arrow indicates lipid; L: vessel lumen; bar: 100 μm).

Table 2.

Composition of plaques in 8‐, 16‐, 25‐, and 34‐week ApoE‐KO mice (n = 6 for each age group).

Age (weeks)	Plaque present	Acellular regions	Fibrous cap	Cholesterol crystals
8	0% (0/6)	–	–	–
16	33% (2/6)	0% (0/6)	0% (0/6)	0% (0/6)
25	83% (5/6)	50% (3/6)	0% (0/6)	66% (4/6)
34	100% (6/6)	100% (6/6)	50% (3/6)	100% (6/6)

Between 8 and 34 weeks, elastin fibers were significantly more preserved in the C57BL/6J mouse as compared to the ApoE‐KO. In ApoE‐KO (Figure 4, Table 1), elastin grades ranged from 1.17 at 8 weeks to 2.83 at 34 weeks and the thickness of the tunica media increased from 29.11 μm at 8 weeks to 40.43 μm at 34 weeks; elastin fragmentation and media thickening were significant between 8 and 25 weeks (p= 0.02 and p= 0.002, respectively). Conversely, in C57BL/6J (Figure 4) elastin grades ranged from 1.00 at 8 weeks to 2.00 at 34 weeks (p < 0.001) and there was no significant change in media thickness (8 weeks: 28.41 μm, 34 weeks: 30.36 μm, p= NS).

Figure 4.

Verhoeff's staining of aortic cross‐sections from 8‐ and 34‐week‐old ApoE‐KO mice illustrates significant thickening of the tunica media and elastin fiber fragmentation (A: 8 weeks, B: 34 weeks). C57BL/6J mice exhibited some elastin fiber fragmentation, though to a lesser degree, and media thickness did not change significantly with age (not shown). Actual media thicknesses and elastin grades are given in Table 1. Sections are shown at the same magnification (black arrows indicate breaks in elastin fibers; L: vessel lumen; bar: 100 μm).

Discussion

In this study, we fed an atherogenic diet to a group of aging ApoE‐KO and C57BL/6J mice and evaluated the relative contributions of age and atherosclerosis to vascular stiffness. In particular, we assessed histological changes within the lumen (plaque) and aortic wall (elastin) with respect to significant increases in aortic stiffness.

To measure stiffness of the mouse aorta, we utilized an ex vivo vascular testing system. This system is particularly advantageous given its ability to produce pulsatile pressurization of the vessel within a physiologic range, its use of a laser micrometer to precisely measure vessel outer diameter, its ability to control the biochemical environment of the vessel, and the maintenance of structural and cellular integrity of the vessel wall during and after testing. ¹⁹ , ²⁰ , ²¹ Furthermore, our previous studies indicate that compared to pulse wave velocity (PWV), this system provides a more accurate and sensitive approach to the measurement of small caliber vessel stiffness. ¹⁹

We found that aortic stiffness increased significantly with age in both strains, though more dramatically in ApoE‐KO mice. This finding is consistent with previous reports that ApoE‐KO mice experience a significant increase in aortic stiffness, as Wang et al. have documented between 16‐ and 56‐week‐old ApoE‐KO mice. ⁴ Similarly, Reddy et al. demonstrated that healthy C57BL/6J mice experience age‐related stiffening between 34 and 125 weeks. ¹²

Stiffness at early time points was approximately equal between strains through 25 weeks, and ApoE‐KO did not demonstrate significantly greater stiffness than C57BL/6J until 34 weeks. Fatty streaks and plaques were present in 8‐, 16‐, and 25‐week ApoE‐KO, while no plaques were present in C57BL/6J. Additionally, we observed media thickening and elastin fiber fragmentation in the ApoE‐KO mice at these time points, while in the C57BL/6J mice, elastin fibers fragmented to a lesser degree and media thickness did not change significantly with age. In spite of these distinct differences in histology through 25 weeks, stiffness was similar between strains. At 34 weeks, however, the significant increase in stiffness occurring in ApoE‐KO mice occurred alongside further advancement of the plaque (as evidenced by its significantly increased size, appearance of a fibrous cap, and increased acellular regions and cholesterol crystals), damage to elastin fibers, and media thickening. These observations suggest that the significant increase in stiffness between 8 and 25 weeks in both strains is attributed to age‐related causes while the increase seen in ApoE‐KO mice between 25 and 34 weeks correlates directly to advanced atherosclerotic disease and is characterized by significant narrowing of the lumen and injury to the wall near the plaque. While the developing plaque and initial changes in the media may account for the slightly higher stiffness values in ApoE‐KO up to 25 weeks, our data indicate that early increases in stiffness in this strain are not driven by plaque development.

Increasing media thickness and elastin fragmentation in the ApoE‐KO are consistent (though perhaps accelerated due to the atherogenic diet) with what has been previously reported. ⁴ , ¹⁷ , ¹⁸ Since we did not find a significant increase in media thickness or elastin grade coinciding with the dramatic increase in stiffness between 25‐ and 34‐week ApoE‐KO mice, we conclude that thickening and fragmentation alone are not responsible for the increase in stiffness. Though elastin fiber disruption occurring alongside plaque development is likely a mechanism by which plaque‐burden influences and directly decreases the elasticity of the vessel, the amount that can be attributed specifically to plaque or elastin disruption remains to be explored.

While increased plaque size, wall thickness, and elastin fiber fragmentation have been previously reported in ApoE‐KO, until now, studies of aortic stiffness in this mouse have utilized techniques that may underestimate stiffness (such as PWV ²⁵ ) and furthermore, have looked at time points that are too far apart to be able to differentiate which increases are age‐related and which are atherosclerotic disease‐related. Though several investigators have shown that arterial stiffness increases progressively without the presence of atherosclerotic plaque in aging healthy humans, studies have yet to elucidate the mechanism by which this may occur.

Though our ex vivo approach did not allow for true longitudinal analysis, this study provides clear evidence that increases in vascular stiffness are not only due to plaque development. This finding points to the existence of additional and more complex mechanisms, to which other studies have also alluded. The understanding of these processes has several clinical implications, including the interpretation of aortic stiffness data. Since increased stiffness may only be detected with advanced disease, as observed in this study, more aggressive treatments and monitoring procedures should be instituted for patients at higher risk for developing accelerated vascular stiffness.

Conclusion

Our results represent the first assessment of aortic stiffness and histology as a function of age in ApoE‐KO and C57BL/6J mice fed an atherogenic diet. Through this study, we evaluated the relative contributions of age and atherosclerosis to vascular stiffness. Though we found that a distinct difference in histology exists between aging and atherosclerotic vessels, our data suggest that early increases in stiffness in the atherosclerotic‐prone mouse model are age‐related and that plaque is not a major contributor to stiffness until advanced stages of atherosclerotic disease.